nld-latn-100mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-36000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.265602322206096,
  "eval_steps": 3000,
  "global_step": 36000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.774228096008301,
      "epoch": 0.00045355587808418,
      "grad_norm": 17.375,
      "learning_rate": 2e-06,
      "loss": 14.4464,
      "mean_token_accuracy": 7.704160525463521e-05,
      "num_tokens": 10827.0,
      "step": 5
    },
    {
      "entropy": 4.7717701435089115,
      "epoch": 0.00090711175616836,
      "grad_norm": 18.375,
      "learning_rate": 4.5e-06,
      "loss": 14.3304,
      "mean_token_accuracy": 0.0,
      "num_tokens": 21025.0,
      "step": 10
    },
    {
      "entropy": 4.839741182327271,
      "epoch": 0.00136066763425254,
      "grad_norm": 24.125,
      "learning_rate": 7e-06,
      "loss": 14.0962,
      "mean_token_accuracy": 7.401924231089651e-05,
      "num_tokens": 31678.0,
      "step": 15
    },
    {
      "entropy": 5.192698621749878,
      "epoch": 0.00181422351233672,
      "grad_norm": 32.0,
      "learning_rate": 9.5e-06,
      "loss": 13.3294,
      "mean_token_accuracy": 0.0,
      "num_tokens": 41025.0,
      "step": 20
    },
    {
      "entropy": 7.903842544555664,
      "epoch": 0.0022677793904208998,
      "grad_norm": 10.8125,
      "learning_rate": 1.2e-05,
      "loss": 11.653,
      "mean_token_accuracy": 0.0,
      "num_tokens": 51270.0,
      "step": 25
    },
    {
      "entropy": 10.564220905303955,
      "epoch": 0.00272133526850508,
      "grad_norm": 3.453125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.7764,
      "mean_token_accuracy": 0.0030453664774540814,
      "num_tokens": 62142.0,
      "step": 30
    },
    {
      "entropy": 10.698297691345214,
      "epoch": 0.00317489114658926,
      "grad_norm": 3.203125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.5251,
      "mean_token_accuracy": 0.017070478899404405,
      "num_tokens": 72709.0,
      "step": 35
    },
    {
      "entropy": 10.688808727264405,
      "epoch": 0.00362844702467344,
      "grad_norm": 2.609375,
      "learning_rate": 1.95e-05,
      "loss": 10.1613,
      "mean_token_accuracy": 0.029810004867613316,
      "num_tokens": 82313.0,
      "step": 40
    },
    {
      "entropy": 10.627019691467286,
      "epoch": 0.00408200290275762,
      "grad_norm": 2.640625,
      "learning_rate": 2.2e-05,
      "loss": 9.8529,
      "mean_token_accuracy": 0.03834186047315598,
      "num_tokens": 92450.0,
      "step": 45
    },
    {
      "entropy": 10.572821617126465,
      "epoch": 0.0045355587808417995,
      "grad_norm": 2.09375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.7295,
      "mean_token_accuracy": 0.0418919712305069,
      "num_tokens": 103978.0,
      "step": 50
    },
    {
      "entropy": 10.587093353271484,
      "epoch": 0.00498911465892598,
      "grad_norm": 2.09375,
      "learning_rate": 2.7e-05,
      "loss": 9.5671,
      "mean_token_accuracy": 0.04565211050212383,
      "num_tokens": 114134.0,
      "step": 55
    },
    {
      "entropy": 10.595294857025147,
      "epoch": 0.00544267053701016,
      "grad_norm": 2.15625,
      "learning_rate": 2.95e-05,
      "loss": 9.5346,
      "mean_token_accuracy": 0.04470008127391338,
      "num_tokens": 125037.0,
      "step": 60
    },
    {
      "entropy": 10.570393085479736,
      "epoch": 0.00589622641509434,
      "grad_norm": 2.015625,
      "learning_rate": 3.2e-05,
      "loss": 9.3921,
      "mean_token_accuracy": 0.04514850843697786,
      "num_tokens": 134721.0,
      "step": 65
    },
    {
      "entropy": 10.546493816375733,
      "epoch": 0.00634978229317852,
      "grad_norm": 2.140625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.3112,
      "mean_token_accuracy": 0.04249380584806204,
      "num_tokens": 144448.0,
      "step": 70
    },
    {
      "entropy": 10.550293254852296,
      "epoch": 0.006803338171262699,
      "grad_norm": 2.125,
      "learning_rate": 3.7e-05,
      "loss": 9.2335,
      "mean_token_accuracy": 0.04518970362842083,
      "num_tokens": 154343.0,
      "step": 75
    },
    {
      "entropy": 10.532429504394532,
      "epoch": 0.00725689404934688,
      "grad_norm": 2.078125,
      "learning_rate": 3.95e-05,
      "loss": 9.2049,
      "mean_token_accuracy": 0.041810613870620725,
      "num_tokens": 165650.0,
      "step": 80
    },
    {
      "entropy": 10.50885066986084,
      "epoch": 0.007710449927431059,
      "grad_norm": 1.875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.2095,
      "mean_token_accuracy": 0.03873651716858149,
      "num_tokens": 177098.0,
      "step": 85
    },
    {
      "entropy": 10.484589099884033,
      "epoch": 0.00816400580551524,
      "grad_norm": 2.0625,
      "learning_rate": 4.45e-05,
      "loss": 8.9579,
      "mean_token_accuracy": 0.04810894541442394,
      "num_tokens": 187903.0,
      "step": 90
    },
    {
      "entropy": 10.423638439178466,
      "epoch": 0.00861756168359942,
      "grad_norm": 1.890625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.8843,
      "mean_token_accuracy": 0.04500575102865696,
      "num_tokens": 198077.0,
      "step": 95
    },
    {
      "entropy": 10.382922554016114,
      "epoch": 0.009071117561683599,
      "grad_norm": 1.828125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.8487,
      "mean_token_accuracy": 0.04283392727375031,
      "num_tokens": 208641.0,
      "step": 100
    },
    {
      "entropy": 10.327754020690918,
      "epoch": 0.009524673439767779,
      "grad_norm": 1.84375,
      "learning_rate": 5.2e-05,
      "loss": 8.6499,
      "mean_token_accuracy": 0.0530441515147686,
      "num_tokens": 218398.0,
      "step": 105
    },
    {
      "entropy": 10.238056182861328,
      "epoch": 0.00997822931785196,
      "grad_norm": 1.6015625,
      "learning_rate": 5.45e-05,
      "loss": 8.501,
      "mean_token_accuracy": 0.05147750936448574,
      "num_tokens": 229059.0,
      "step": 110
    },
    {
      "entropy": 10.118829822540283,
      "epoch": 0.01043178519593614,
      "grad_norm": 1.6953125,
      "learning_rate": 5.7e-05,
      "loss": 8.4107,
      "mean_token_accuracy": 0.05338236689567566,
      "num_tokens": 239775.0,
      "step": 115
    },
    {
      "entropy": 9.956042957305907,
      "epoch": 0.01088534107402032,
      "grad_norm": 1.734375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.2389,
      "mean_token_accuracy": 0.053830433264374734,
      "num_tokens": 249954.0,
      "step": 120
    },
    {
      "entropy": 9.754269218444824,
      "epoch": 0.011338896952104499,
      "grad_norm": 1.3515625,
      "learning_rate": 6.2e-05,
      "loss": 8.1364,
      "mean_token_accuracy": 0.05633715465664864,
      "num_tokens": 260056.0,
      "step": 125
    },
    {
      "entropy": 9.600945854187012,
      "epoch": 0.01179245283018868,
      "grad_norm": 1.2265625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.0344,
      "mean_token_accuracy": 0.05928211845457554,
      "num_tokens": 269904.0,
      "step": 130
    },
    {
      "entropy": 9.370787525177002,
      "epoch": 0.01224600870827286,
      "grad_norm": 1.171875,
      "learning_rate": 6.7e-05,
      "loss": 7.9224,
      "mean_token_accuracy": 0.05375306829810143,
      "num_tokens": 280900.0,
      "step": 135
    },
    {
      "entropy": 9.10864839553833,
      "epoch": 0.01269956458635704,
      "grad_norm": 1.0,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.8234,
      "mean_token_accuracy": 0.055295905098319054,
      "num_tokens": 290844.0,
      "step": 140
    },
    {
      "entropy": 8.816210174560547,
      "epoch": 0.013153120464441219,
      "grad_norm": 0.87890625,
      "learning_rate": 7.2e-05,
      "loss": 7.7326,
      "mean_token_accuracy": 0.062234926223754886,
      "num_tokens": 301345.0,
      "step": 145
    },
    {
      "entropy": 8.610615539550782,
      "epoch": 0.013606676342525399,
      "grad_norm": 1.0546875,
      "learning_rate": 7.45e-05,
      "loss": 7.6332,
      "mean_token_accuracy": 0.06342925988137722,
      "num_tokens": 312681.0,
      "step": 150
    },
    {
      "entropy": 8.336861324310302,
      "epoch": 0.01406023222060958,
      "grad_norm": 0.98046875,
      "learning_rate": 7.7e-05,
      "loss": 7.5646,
      "mean_token_accuracy": 0.06853245571255684,
      "num_tokens": 323492.0,
      "step": 155
    },
    {
      "entropy": 8.240836334228515,
      "epoch": 0.01451378809869376,
      "grad_norm": 1.0625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.564,
      "mean_token_accuracy": 0.06723525486886502,
      "num_tokens": 334521.0,
      "step": 160
    },
    {
      "entropy": 8.05573558807373,
      "epoch": 0.014967343976777939,
      "grad_norm": 1.2109375,
      "learning_rate": 8.2e-05,
      "loss": 7.5656,
      "mean_token_accuracy": 0.07214355692267418,
      "num_tokens": 345468.0,
      "step": 165
    },
    {
      "entropy": 8.033101844787598,
      "epoch": 0.015420899854862119,
      "grad_norm": 1.8046875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.4802,
      "mean_token_accuracy": 0.07146222665905952,
      "num_tokens": 355911.0,
      "step": 170
    },
    {
      "entropy": 7.941764783859253,
      "epoch": 0.0158744557329463,
      "grad_norm": 0.953125,
      "learning_rate": 8.7e-05,
      "loss": 7.505,
      "mean_token_accuracy": 0.07519448474049568,
      "num_tokens": 367537.0,
      "step": 175
    },
    {
      "entropy": 7.970178508758545,
      "epoch": 0.01632801161103048,
      "grad_norm": 1.03125,
      "learning_rate": 8.95e-05,
      "loss": 7.6193,
      "mean_token_accuracy": 0.07340684309601783,
      "num_tokens": 377918.0,
      "step": 180
    },
    {
      "entropy": 7.922939395904541,
      "epoch": 0.01678156748911466,
      "grad_norm": 1.078125,
      "learning_rate": 9.2e-05,
      "loss": 7.4722,
      "mean_token_accuracy": 0.07753909192979336,
      "num_tokens": 388461.0,
      "step": 185
    },
    {
      "entropy": 7.8546586513519285,
      "epoch": 0.01723512336719884,
      "grad_norm": 1.15625,
      "learning_rate": 9.45e-05,
      "loss": 7.4354,
      "mean_token_accuracy": 0.08010556474328041,
      "num_tokens": 399771.0,
      "step": 190
    },
    {
      "entropy": 7.820327281951904,
      "epoch": 0.01768867924528302,
      "grad_norm": 1.1640625,
      "learning_rate": 9.7e-05,
      "loss": 7.3185,
      "mean_token_accuracy": 0.07953049913048744,
      "num_tokens": 410254.0,
      "step": 195
    },
    {
      "entropy": 7.83513126373291,
      "epoch": 0.018142235123367198,
      "grad_norm": 1.3984375,
      "learning_rate": 9.95e-05,
      "loss": 7.3297,
      "mean_token_accuracy": 0.07624107114970684,
      "num_tokens": 419792.0,
      "step": 200
    },
    {
      "entropy": 7.7827085018157955,
      "epoch": 0.018595791001451378,
      "grad_norm": 1.609375,
      "learning_rate": 0.000102,
      "loss": 7.3324,
      "mean_token_accuracy": 0.08292475119233131,
      "num_tokens": 429264.0,
      "step": 205
    },
    {
      "entropy": 7.716277408599853,
      "epoch": 0.019049346879535557,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.2853,
      "mean_token_accuracy": 0.08389051035046577,
      "num_tokens": 439624.0,
      "step": 210
    },
    {
      "entropy": 7.717530107498169,
      "epoch": 0.01950290275761974,
      "grad_norm": 1.15625,
      "learning_rate": 0.000107,
      "loss": 7.267,
      "mean_token_accuracy": 0.0840241178870201,
      "num_tokens": 449556.0,
      "step": 215
    },
    {
      "entropy": 7.703034257888794,
      "epoch": 0.01995645863570392,
      "grad_norm": 1.25,
      "learning_rate": 0.0001095,
      "loss": 7.2846,
      "mean_token_accuracy": 0.08519925326108932,
      "num_tokens": 460243.0,
      "step": 220
    },
    {
      "entropy": 7.728843021392822,
      "epoch": 0.0204100145137881,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000112,
      "loss": 7.3326,
      "mean_token_accuracy": 0.0805255938321352,
      "num_tokens": 469767.0,
      "step": 225
    },
    {
      "entropy": 7.720327663421631,
      "epoch": 0.02086357039187228,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0001145,
      "loss": 7.2493,
      "mean_token_accuracy": 0.08286545500159263,
      "num_tokens": 480329.0,
      "step": 230
    },
    {
      "entropy": 7.666221618652344,
      "epoch": 0.02131712626995646,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.214,
      "mean_token_accuracy": 0.07896244674921035,
      "num_tokens": 490096.0,
      "step": 235
    },
    {
      "entropy": 7.731700372695923,
      "epoch": 0.02177068214804064,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.2592,
      "mean_token_accuracy": 0.07835591658949852,
      "num_tokens": 501231.0,
      "step": 240
    },
    {
      "entropy": 7.684773302078247,
      "epoch": 0.022224238026124818,
      "grad_norm": 1.609375,
      "learning_rate": 0.000122,
      "loss": 7.1609,
      "mean_token_accuracy": 0.08823157548904419,
      "num_tokens": 510722.0,
      "step": 245
    },
    {
      "entropy": 7.513065958023072,
      "epoch": 0.022677793904208998,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0001245,
      "loss": 7.2055,
      "mean_token_accuracy": 0.08414689600467681,
      "num_tokens": 521452.0,
      "step": 250
    },
    {
      "entropy": 7.660957479476929,
      "epoch": 0.023131349782293177,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000127,
      "loss": 7.211,
      "mean_token_accuracy": 0.08599318936467171,
      "num_tokens": 532397.0,
      "step": 255
    },
    {
      "entropy": 7.555563879013062,
      "epoch": 0.02358490566037736,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0001295,
      "loss": 7.1818,
      "mean_token_accuracy": 0.08441983088850975,
      "num_tokens": 542271.0,
      "step": 260
    },
    {
      "entropy": 7.615720558166504,
      "epoch": 0.02403846153846154,
      "grad_norm": 1.15625,
      "learning_rate": 0.000132,
      "loss": 7.1579,
      "mean_token_accuracy": 0.09066328033804893,
      "num_tokens": 552411.0,
      "step": 265
    },
    {
      "entropy": 7.520418405532837,
      "epoch": 0.02449201741654572,
      "grad_norm": 0.9375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.1502,
      "mean_token_accuracy": 0.08552343845367431,
      "num_tokens": 561518.0,
      "step": 270
    },
    {
      "entropy": 7.588455486297607,
      "epoch": 0.0249455732946299,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1831,
      "mean_token_accuracy": 0.08754325062036514,
      "num_tokens": 571361.0,
      "step": 275
    },
    {
      "entropy": 7.492841577529907,
      "epoch": 0.02539912917271408,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0001395,
      "loss": 7.0901,
      "mean_token_accuracy": 0.09500023052096367,
      "num_tokens": 580926.0,
      "step": 280
    },
    {
      "entropy": 7.54066162109375,
      "epoch": 0.025852685050798258,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1773,
      "mean_token_accuracy": 0.08410000652074814,
      "num_tokens": 592282.0,
      "step": 285
    },
    {
      "entropy": 7.580021572113037,
      "epoch": 0.026306240928882438,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0001445,
      "loss": 7.1784,
      "mean_token_accuracy": 0.08815034925937652,
      "num_tokens": 602767.0,
      "step": 290
    },
    {
      "entropy": 7.532192230224609,
      "epoch": 0.026759796806966617,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000147,
      "loss": 7.0936,
      "mean_token_accuracy": 0.08808856308460236,
      "num_tokens": 612763.0,
      "step": 295
    },
    {
      "entropy": 7.443887853622437,
      "epoch": 0.027213352685050797,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001495,
      "loss": 7.0434,
      "mean_token_accuracy": 0.09205911681056023,
      "num_tokens": 622796.0,
      "step": 300
    },
    {
      "entropy": 7.466957712173462,
      "epoch": 0.027666908563134977,
      "grad_norm": 1.140625,
      "learning_rate": 0.000152,
      "loss": 7.0302,
      "mean_token_accuracy": 0.09625576436519623,
      "num_tokens": 633136.0,
      "step": 305
    },
    {
      "entropy": 7.39753270149231,
      "epoch": 0.02812046444121916,
      "grad_norm": 1.078125,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0236,
      "mean_token_accuracy": 0.09388496801257133,
      "num_tokens": 643347.0,
      "step": 310
    },
    {
      "entropy": 7.445803928375244,
      "epoch": 0.02857402031930334,
      "grad_norm": 1.125,
      "learning_rate": 0.000157,
      "loss": 7.0109,
      "mean_token_accuracy": 0.09590190947055817,
      "num_tokens": 653096.0,
      "step": 315
    },
    {
      "entropy": 7.36529951095581,
      "epoch": 0.02902757619738752,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001595,
      "loss": 7.04,
      "mean_token_accuracy": 0.09056715369224548,
      "num_tokens": 663557.0,
      "step": 320
    },
    {
      "entropy": 7.485755491256714,
      "epoch": 0.0294811320754717,
      "grad_norm": 1.234375,
      "learning_rate": 0.000162,
      "loss": 7.1043,
      "mean_token_accuracy": 0.09179518148303031,
      "num_tokens": 674395.0,
      "step": 325
    },
    {
      "entropy": 7.3991217613220215,
      "epoch": 0.029934687953555878,
      "grad_norm": 1.109375,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0039,
      "mean_token_accuracy": 0.0983194276690483,
      "num_tokens": 684776.0,
      "step": 330
    },
    {
      "entropy": 7.310597705841064,
      "epoch": 0.030388243831640058,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9818,
      "mean_token_accuracy": 0.09324271008372306,
      "num_tokens": 694886.0,
      "step": 335
    },
    {
      "entropy": 7.388318014144898,
      "epoch": 0.030841799709724237,
      "grad_norm": 1.1875,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0598,
      "mean_token_accuracy": 0.09387328922748565,
      "num_tokens": 705426.0,
      "step": 340
    },
    {
      "entropy": 7.296148014068604,
      "epoch": 0.03129535558780842,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.0644,
      "mean_token_accuracy": 0.09106143489480019,
      "num_tokens": 716271.0,
      "step": 345
    },
    {
      "entropy": 7.398095321655274,
      "epoch": 0.0317489114658926,
      "grad_norm": 1.234375,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0437,
      "mean_token_accuracy": 0.09280406758189201,
      "num_tokens": 726875.0,
      "step": 350
    },
    {
      "entropy": 7.330491304397583,
      "epoch": 0.032202467343976776,
      "grad_norm": 1.171875,
      "learning_rate": 0.000177,
      "loss": 6.9817,
      "mean_token_accuracy": 0.09162108749151229,
      "num_tokens": 737825.0,
      "step": 355
    },
    {
      "entropy": 7.354272747039795,
      "epoch": 0.03265602322206096,
      "grad_norm": 1.0,
      "learning_rate": 0.0001795,
      "loss": 6.9502,
      "mean_token_accuracy": 0.09616798982024193,
      "num_tokens": 747424.0,
      "step": 360
    },
    {
      "entropy": 7.289549160003662,
      "epoch": 0.033109579100145135,
      "grad_norm": 1.0625,
      "learning_rate": 0.000182,
      "loss": 6.9635,
      "mean_token_accuracy": 0.09623874425888061,
      "num_tokens": 757649.0,
      "step": 365
    },
    {
      "entropy": 7.327360343933106,
      "epoch": 0.03356313497822932,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0001845,
      "loss": 6.947,
      "mean_token_accuracy": 0.09573540687561036,
      "num_tokens": 768556.0,
      "step": 370
    },
    {
      "entropy": 7.2021252632141115,
      "epoch": 0.034016690856313495,
      "grad_norm": 1.46875,
      "learning_rate": 0.000187,
      "loss": 6.9449,
      "mean_token_accuracy": 0.0929239995777607,
      "num_tokens": 778331.0,
      "step": 375
    },
    {
      "entropy": 7.242694997787476,
      "epoch": 0.03447024673439768,
      "grad_norm": 1.34375,
      "learning_rate": 0.0001895,
      "loss": 6.9322,
      "mean_token_accuracy": 0.09555335119366645,
      "num_tokens": 788065.0,
      "step": 380
    },
    {
      "entropy": 7.3150105476379395,
      "epoch": 0.03492380261248186,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000192,
      "loss": 6.9711,
      "mean_token_accuracy": 0.09826483204960823,
      "num_tokens": 798912.0,
      "step": 385
    },
    {
      "entropy": 7.2783135890960695,
      "epoch": 0.03537735849056604,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001945,
      "loss": 6.9073,
      "mean_token_accuracy": 0.10042868107557297,
      "num_tokens": 809580.0,
      "step": 390
    },
    {
      "entropy": 7.212211751937867,
      "epoch": 0.03583091436865022,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.996,
      "mean_token_accuracy": 0.09339204281568528,
      "num_tokens": 820207.0,
      "step": 395
    },
    {
      "entropy": 7.088284826278686,
      "epoch": 0.036284470246734396,
      "grad_norm": 1.375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.826,
      "mean_token_accuracy": 0.09160310104489326,
      "num_tokens": 831038.0,
      "step": 400
    },
    {
      "entropy": 7.272746944427491,
      "epoch": 0.03673802612481858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000202,
      "loss": 6.9068,
      "mean_token_accuracy": 0.09824135601520538,
      "num_tokens": 842144.0,
      "step": 405
    },
    {
      "entropy": 7.274405145645142,
      "epoch": 0.037191582002902755,
      "grad_norm": 1.25,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.9044,
      "mean_token_accuracy": 0.10123971402645111,
      "num_tokens": 851315.0,
      "step": 410
    },
    {
      "entropy": 7.09048376083374,
      "epoch": 0.03764513788098694,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000207,
      "loss": 6.921,
      "mean_token_accuracy": 0.09870171621441841,
      "num_tokens": 861233.0,
      "step": 415
    },
    {
      "entropy": 7.206084251403809,
      "epoch": 0.038098693759071114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002095,
      "loss": 6.9004,
      "mean_token_accuracy": 0.09542403146624565,
      "num_tokens": 871131.0,
      "step": 420
    },
    {
      "entropy": 7.1921416282653805,
      "epoch": 0.0385522496371553,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000212,
      "loss": 6.8717,
      "mean_token_accuracy": 0.09407154396176338,
      "num_tokens": 882195.0,
      "step": 425
    },
    {
      "entropy": 7.185799217224121,
      "epoch": 0.03900580551523948,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002145,
      "loss": 6.8489,
      "mean_token_accuracy": 0.10136798098683357,
      "num_tokens": 891806.0,
      "step": 430
    },
    {
      "entropy": 7.238110542297363,
      "epoch": 0.03945936139332366,
      "grad_norm": 1.203125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.9054,
      "mean_token_accuracy": 0.09595799744129181,
      "num_tokens": 901350.0,
      "step": 435
    },
    {
      "entropy": 7.158117580413818,
      "epoch": 0.03991291727140784,
      "grad_norm": 1.125,
      "learning_rate": 0.0002195,
      "loss": 6.8187,
      "mean_token_accuracy": 0.09831003770232201,
      "num_tokens": 912061.0,
      "step": 440
    },
    {
      "entropy": 7.09474549293518,
      "epoch": 0.040366473149492016,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000222,
      "loss": 6.7766,
      "mean_token_accuracy": 0.1011108048260212,
      "num_tokens": 922220.0,
      "step": 445
    },
    {
      "entropy": 7.1174585819244385,
      "epoch": 0.0408200290275762,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002245,
      "loss": 6.8018,
      "mean_token_accuracy": 0.10038985908031464,
      "num_tokens": 932319.0,
      "step": 450
    },
    {
      "entropy": 7.0610755443572994,
      "epoch": 0.041273584905660375,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8413,
      "mean_token_accuracy": 0.09963766038417816,
      "num_tokens": 943254.0,
      "step": 455
    },
    {
      "entropy": 7.112555265426636,
      "epoch": 0.04172714078374456,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.8619,
      "mean_token_accuracy": 0.0950029470026493,
      "num_tokens": 953921.0,
      "step": 460
    },
    {
      "entropy": 7.016997766494751,
      "epoch": 0.042180696661828734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7679,
      "mean_token_accuracy": 0.10482859760522842,
      "num_tokens": 963832.0,
      "step": 465
    },
    {
      "entropy": 7.03766975402832,
      "epoch": 0.04263425253991292,
      "grad_norm": 1.296875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.866,
      "mean_token_accuracy": 0.09811323508620262,
      "num_tokens": 974066.0,
      "step": 470
    },
    {
      "entropy": 7.129855823516846,
      "epoch": 0.0430878084179971,
      "grad_norm": 1.234375,
      "learning_rate": 0.000237,
      "loss": 6.8551,
      "mean_token_accuracy": 0.10151214599609375,
      "num_tokens": 984402.0,
      "step": 475
    },
    {
      "entropy": 7.089972972869873,
      "epoch": 0.04354136429608128,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002395,
      "loss": 6.7652,
      "mean_token_accuracy": 0.09788535833358765,
      "num_tokens": 993555.0,
      "step": 480
    },
    {
      "entropy": 7.085727262496948,
      "epoch": 0.04399492017416546,
      "grad_norm": 1.265625,
      "learning_rate": 0.000242,
      "loss": 6.7472,
      "mean_token_accuracy": 0.10479735285043716,
      "num_tokens": 1002920.0,
      "step": 485
    },
    {
      "entropy": 7.040792322158813,
      "epoch": 0.044448476052249636,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002445,
      "loss": 6.8392,
      "mean_token_accuracy": 0.09684377312660217,
      "num_tokens": 1013611.0,
      "step": 490
    },
    {
      "entropy": 6.981588220596313,
      "epoch": 0.04490203193033382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000247,
      "loss": 6.7247,
      "mean_token_accuracy": 0.09907414987683297,
      "num_tokens": 1024121.0,
      "step": 495
    },
    {
      "entropy": 7.122877836227417,
      "epoch": 0.045355587808417995,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002495,
      "loss": 6.9627,
      "mean_token_accuracy": 0.10345524102449417,
      "num_tokens": 1033845.0,
      "step": 500
    },
    {
      "entropy": 7.049224853515625,
      "epoch": 0.04580914368650218,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000252,
      "loss": 6.7454,
      "mean_token_accuracy": 0.10035812482237816,
      "num_tokens": 1044642.0,
      "step": 505
    },
    {
      "entropy": 7.04058837890625,
      "epoch": 0.046262699564586354,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002545,
      "loss": 6.8115,
      "mean_token_accuracy": 0.1038434311747551,
      "num_tokens": 1054728.0,
      "step": 510
    },
    {
      "entropy": 6.993410062789917,
      "epoch": 0.04671625544267054,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000257,
      "loss": 6.7643,
      "mean_token_accuracy": 0.09849002212285995,
      "num_tokens": 1065743.0,
      "step": 515
    },
    {
      "entropy": 6.9517449855804445,
      "epoch": 0.04716981132075472,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002595,
      "loss": 6.686,
      "mean_token_accuracy": 0.10116103440523147,
      "num_tokens": 1076798.0,
      "step": 520
    },
    {
      "entropy": 7.008026170730591,
      "epoch": 0.047623367198838897,
      "grad_norm": 1.1875,
      "learning_rate": 0.000262,
      "loss": 6.8212,
      "mean_token_accuracy": 0.10135331153869628,
      "num_tokens": 1087445.0,
      "step": 525
    },
    {
      "entropy": 6.9268933773040775,
      "epoch": 0.04807692307692308,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7326,
      "mean_token_accuracy": 0.10336857363581657,
      "num_tokens": 1098379.0,
      "step": 530
    },
    {
      "entropy": 7.028698968887329,
      "epoch": 0.048530478955007256,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.809,
      "mean_token_accuracy": 0.10176651552319527,
      "num_tokens": 1108016.0,
      "step": 535
    },
    {
      "entropy": 6.998309707641601,
      "epoch": 0.04898403483309144,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7018,
      "mean_token_accuracy": 0.10114236250519752,
      "num_tokens": 1117887.0,
      "step": 540
    },
    {
      "entropy": 6.907250595092774,
      "epoch": 0.049437590711175615,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6942,
      "mean_token_accuracy": 0.09997687637805938,
      "num_tokens": 1128497.0,
      "step": 545
    },
    {
      "entropy": 6.936363220214844,
      "epoch": 0.0498911465892598,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002745,
      "loss": 6.7284,
      "mean_token_accuracy": 0.10731528326869011,
      "num_tokens": 1138168.0,
      "step": 550
    },
    {
      "entropy": 6.892492866516113,
      "epoch": 0.050344702467343974,
      "grad_norm": 1.328125,
      "learning_rate": 0.000277,
      "loss": 6.6977,
      "mean_token_accuracy": 0.10209395885467529,
      "num_tokens": 1148357.0,
      "step": 555
    },
    {
      "entropy": 6.959150648117065,
      "epoch": 0.05079825834542816,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002795,
      "loss": 6.719,
      "mean_token_accuracy": 0.10157941356301307,
      "num_tokens": 1159964.0,
      "step": 560
    },
    {
      "entropy": 6.994518995285034,
      "epoch": 0.05125181422351233,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.7767,
      "mean_token_accuracy": 0.10542606040835381,
      "num_tokens": 1170683.0,
      "step": 565
    },
    {
      "entropy": 6.960160112380981,
      "epoch": 0.051705370101596516,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002845,
      "loss": 6.7378,
      "mean_token_accuracy": 0.09894839897751809,
      "num_tokens": 1181680.0,
      "step": 570
    },
    {
      "entropy": 6.797056579589844,
      "epoch": 0.0521589259796807,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000287,
      "loss": 6.6327,
      "mean_token_accuracy": 0.10385172069072723,
      "num_tokens": 1191619.0,
      "step": 575
    },
    {
      "entropy": 6.813391923904419,
      "epoch": 0.052612481857764876,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002895,
      "loss": 6.6332,
      "mean_token_accuracy": 0.103890310972929,
      "num_tokens": 1202264.0,
      "step": 580
    },
    {
      "entropy": 6.810958385467529,
      "epoch": 0.05306603773584906,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000292,
      "loss": 6.6287,
      "mean_token_accuracy": 0.104695063829422,
      "num_tokens": 1212119.0,
      "step": 585
    },
    {
      "entropy": 6.923195457458496,
      "epoch": 0.053519593613933235,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002945,
      "loss": 6.764,
      "mean_token_accuracy": 0.10014403387904167,
      "num_tokens": 1222317.0,
      "step": 590
    },
    {
      "entropy": 6.910294437408448,
      "epoch": 0.05397314949201742,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000297,
      "loss": 6.703,
      "mean_token_accuracy": 0.10804183036088943,
      "num_tokens": 1232914.0,
      "step": 595
    },
    {
      "entropy": 6.811793613433838,
      "epoch": 0.054426705370101594,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002995,
      "loss": 6.5542,
      "mean_token_accuracy": 0.11040094196796417,
      "num_tokens": 1244214.0,
      "step": 600
    },
    {
      "entropy": 6.8492508888244625,
      "epoch": 0.05488026124818578,
      "grad_norm": 1.171875,
      "learning_rate": 0.000302,
      "loss": 6.6376,
      "mean_token_accuracy": 0.10812766700983048,
      "num_tokens": 1254626.0,
      "step": 605
    },
    {
      "entropy": 6.818840599060058,
      "epoch": 0.05533381712626995,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003045,
      "loss": 6.7416,
      "mean_token_accuracy": 0.10723426789045334,
      "num_tokens": 1265461.0,
      "step": 610
    },
    {
      "entropy": 6.867012405395508,
      "epoch": 0.055787373004354136,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000307,
      "loss": 6.6166,
      "mean_token_accuracy": 0.10740614756941795,
      "num_tokens": 1275810.0,
      "step": 615
    },
    {
      "entropy": 6.771302795410156,
      "epoch": 0.05624092888243832,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003095,
      "loss": 6.63,
      "mean_token_accuracy": 0.11132507100701332,
      "num_tokens": 1285170.0,
      "step": 620
    },
    {
      "entropy": 6.833296775817871,
      "epoch": 0.056694484760522496,
      "grad_norm": 1.203125,
      "learning_rate": 0.000312,
      "loss": 6.5804,
      "mean_token_accuracy": 0.10996494144201278,
      "num_tokens": 1295460.0,
      "step": 625
    },
    {
      "entropy": 6.802701473236084,
      "epoch": 0.05714804063860668,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003145,
      "loss": 6.7121,
      "mean_token_accuracy": 0.10138657614588738,
      "num_tokens": 1306656.0,
      "step": 630
    },
    {
      "entropy": 6.8321558952331545,
      "epoch": 0.057601596516690855,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000317,
      "loss": 6.6778,
      "mean_token_accuracy": 0.10745546817779542,
      "num_tokens": 1318019.0,
      "step": 635
    },
    {
      "entropy": 6.660759735107422,
      "epoch": 0.05805515239477504,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003195,
      "loss": 6.6174,
      "mean_token_accuracy": 0.1097694255411625,
      "num_tokens": 1328120.0,
      "step": 640
    },
    {
      "entropy": 6.918548679351806,
      "epoch": 0.058508708272859214,
      "grad_norm": 1.15625,
      "learning_rate": 0.000322,
      "loss": 6.6987,
      "mean_token_accuracy": 0.10450451076030731,
      "num_tokens": 1338354.0,
      "step": 645
    },
    {
      "entropy": 6.80439510345459,
      "epoch": 0.0589622641509434,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6586,
      "mean_token_accuracy": 0.10693488866090775,
      "num_tokens": 1348666.0,
      "step": 650
    },
    {
      "entropy": 6.861764764785766,
      "epoch": 0.05941582002902757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.705,
      "mean_token_accuracy": 0.10773044675588608,
      "num_tokens": 1359713.0,
      "step": 655
    },
    {
      "entropy": 6.759540653228759,
      "epoch": 0.059869375907111756,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5522,
      "mean_token_accuracy": 0.110796270519495,
      "num_tokens": 1369507.0,
      "step": 660
    },
    {
      "entropy": 6.832991409301758,
      "epoch": 0.06032293178519594,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.715,
      "mean_token_accuracy": 0.10180763974785804,
      "num_tokens": 1380726.0,
      "step": 665
    },
    {
      "entropy": 6.828793239593506,
      "epoch": 0.060776487663280115,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.642,
      "mean_token_accuracy": 0.11072484701871872,
      "num_tokens": 1391423.0,
      "step": 670
    },
    {
      "entropy": 6.827483797073365,
      "epoch": 0.0612300435413643,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000337,
      "loss": 6.6225,
      "mean_token_accuracy": 0.1150636076927185,
      "num_tokens": 1402492.0,
      "step": 675
    },
    {
      "entropy": 6.715798854827881,
      "epoch": 0.061683599419448475,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003395,
      "loss": 6.6714,
      "mean_token_accuracy": 0.11112096533179283,
      "num_tokens": 1412497.0,
      "step": 680
    },
    {
      "entropy": 6.875859308242798,
      "epoch": 0.06213715529753266,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000342,
      "loss": 6.6295,
      "mean_token_accuracy": 0.11218903288245201,
      "num_tokens": 1423214.0,
      "step": 685
    },
    {
      "entropy": 6.707458257675171,
      "epoch": 0.06259071117561683,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.6269,
      "mean_token_accuracy": 0.10577443093061448,
      "num_tokens": 1434044.0,
      "step": 690
    },
    {
      "entropy": 6.722873497009277,
      "epoch": 0.06304426705370102,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000347,
      "loss": 6.5825,
      "mean_token_accuracy": 0.10959424301981927,
      "num_tokens": 1444387.0,
      "step": 695
    },
    {
      "entropy": 6.695552062988281,
      "epoch": 0.0634978229317852,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003495,
      "loss": 6.5396,
      "mean_token_accuracy": 0.10780341550707817,
      "num_tokens": 1455157.0,
      "step": 700
    },
    {
      "entropy": 6.774385118484497,
      "epoch": 0.06395137880986937,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000352,
      "loss": 6.6271,
      "mean_token_accuracy": 0.11427033320069313,
      "num_tokens": 1466145.0,
      "step": 705
    },
    {
      "entropy": 6.722665452957154,
      "epoch": 0.06440493468795355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003545,
      "loss": 6.6276,
      "mean_token_accuracy": 0.111500333994627,
      "num_tokens": 1476713.0,
      "step": 710
    },
    {
      "entropy": 6.750820589065552,
      "epoch": 0.06485849056603774,
      "grad_norm": 1.1875,
      "learning_rate": 0.000357,
      "loss": 6.6316,
      "mean_token_accuracy": 0.11196371614933014,
      "num_tokens": 1486908.0,
      "step": 715
    },
    {
      "entropy": 6.727015495300293,
      "epoch": 0.06531204644412192,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003595,
      "loss": 6.5495,
      "mean_token_accuracy": 0.11699655205011368,
      "num_tokens": 1498185.0,
      "step": 720
    },
    {
      "entropy": 6.688741731643677,
      "epoch": 0.0657656023222061,
      "grad_norm": 1.1875,
      "learning_rate": 0.000362,
      "loss": 6.5868,
      "mean_token_accuracy": 0.11200948804616928,
      "num_tokens": 1508376.0,
      "step": 725
    },
    {
      "entropy": 6.779923057556152,
      "epoch": 0.06621915820029027,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003645,
      "loss": 6.6764,
      "mean_token_accuracy": 0.10785475298762322,
      "num_tokens": 1519953.0,
      "step": 730
    },
    {
      "entropy": 6.637065696716308,
      "epoch": 0.06667271407837445,
      "grad_norm": 1.015625,
      "learning_rate": 0.000367,
      "loss": 6.5419,
      "mean_token_accuracy": 0.11318475008010864,
      "num_tokens": 1529793.0,
      "step": 735
    },
    {
      "entropy": 6.70322322845459,
      "epoch": 0.06712626995645864,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003695,
      "loss": 6.5429,
      "mean_token_accuracy": 0.11152931004762649,
      "num_tokens": 1540859.0,
      "step": 740
    },
    {
      "entropy": 6.675182676315307,
      "epoch": 0.06757982583454282,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000372,
      "loss": 6.4683,
      "mean_token_accuracy": 0.11367111355066299,
      "num_tokens": 1549783.0,
      "step": 745
    },
    {
      "entropy": 6.616857862472534,
      "epoch": 0.06803338171262699,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003745,
      "loss": 6.5441,
      "mean_token_accuracy": 0.10823952555656433,
      "num_tokens": 1560522.0,
      "step": 750
    },
    {
      "entropy": 6.6370926856994625,
      "epoch": 0.06848693759071117,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000377,
      "loss": 6.5206,
      "mean_token_accuracy": 0.11319050565361977,
      "num_tokens": 1570668.0,
      "step": 755
    },
    {
      "entropy": 6.6942650318145756,
      "epoch": 0.06894049346879536,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003795,
      "loss": 6.6439,
      "mean_token_accuracy": 0.11470866054296494,
      "num_tokens": 1581523.0,
      "step": 760
    },
    {
      "entropy": 6.626244068145752,
      "epoch": 0.06939404934687954,
      "grad_norm": 1.171875,
      "learning_rate": 0.000382,
      "loss": 6.5478,
      "mean_token_accuracy": 0.10985084623098373,
      "num_tokens": 1591356.0,
      "step": 765
    },
    {
      "entropy": 6.59173264503479,
      "epoch": 0.06984760522496372,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003845,
      "loss": 6.5349,
      "mean_token_accuracy": 0.11471892967820167,
      "num_tokens": 1601779.0,
      "step": 770
    },
    {
      "entropy": 6.733083581924438,
      "epoch": 0.07030116110304789,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.5258,
      "mean_token_accuracy": 0.11292637586593628,
      "num_tokens": 1611995.0,
      "step": 775
    },
    {
      "entropy": 6.631600999832154,
      "epoch": 0.07075471698113207,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.5418,
      "mean_token_accuracy": 0.11008346080780029,
      "num_tokens": 1621834.0,
      "step": 780
    },
    {
      "entropy": 6.629804039001465,
      "epoch": 0.07120827285921626,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.596,
      "mean_token_accuracy": 0.10650180131196976,
      "num_tokens": 1632881.0,
      "step": 785
    },
    {
      "entropy": 6.621839952468872,
      "epoch": 0.07166182873730044,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.5746,
      "mean_token_accuracy": 0.11100505664944649,
      "num_tokens": 1642870.0,
      "step": 790
    },
    {
      "entropy": 6.605715703964234,
      "epoch": 0.07211538461538461,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.441,
      "mean_token_accuracy": 0.12122759222984314,
      "num_tokens": 1653390.0,
      "step": 795
    },
    {
      "entropy": 6.608106994628907,
      "epoch": 0.07256894049346879,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003995,
      "loss": 6.5597,
      "mean_token_accuracy": 0.11547763496637345,
      "num_tokens": 1663382.0,
      "step": 800
    },
    {
      "entropy": 6.710092163085937,
      "epoch": 0.07302249637155298,
      "grad_norm": 1.078125,
      "learning_rate": 0.000402,
      "loss": 6.5794,
      "mean_token_accuracy": 0.11190545409917832,
      "num_tokens": 1673601.0,
      "step": 805
    },
    {
      "entropy": 6.551822805404663,
      "epoch": 0.07347605224963716,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004045,
      "loss": 6.4868,
      "mean_token_accuracy": 0.11824472174048424,
      "num_tokens": 1683565.0,
      "step": 810
    },
    {
      "entropy": 6.574606466293335,
      "epoch": 0.07392960812772134,
      "grad_norm": 2.078125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4743,
      "mean_token_accuracy": 0.11708612516522407,
      "num_tokens": 1693636.0,
      "step": 815
    },
    {
      "entropy": 6.7191204071044925,
      "epoch": 0.07438316400580551,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004095,
      "loss": 6.5787,
      "mean_token_accuracy": 0.1106836386024952,
      "num_tokens": 1703415.0,
      "step": 820
    },
    {
      "entropy": 6.614391088485718,
      "epoch": 0.0748367198838897,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000412,
      "loss": 6.547,
      "mean_token_accuracy": 0.11781998202204705,
      "num_tokens": 1713579.0,
      "step": 825
    },
    {
      "entropy": 6.650404357910157,
      "epoch": 0.07529027576197388,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004145,
      "loss": 6.5694,
      "mean_token_accuracy": 0.11126149147748947,
      "num_tokens": 1724583.0,
      "step": 830
    },
    {
      "entropy": 6.56261305809021,
      "epoch": 0.07574383164005806,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000417,
      "loss": 6.4685,
      "mean_token_accuracy": 0.11994619444012641,
      "num_tokens": 1735293.0,
      "step": 835
    },
    {
      "entropy": 6.594982242584228,
      "epoch": 0.07619738751814223,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004195,
      "loss": 6.5008,
      "mean_token_accuracy": 0.11226540133357048,
      "num_tokens": 1745623.0,
      "step": 840
    },
    {
      "entropy": 6.603390169143677,
      "epoch": 0.07665094339622641,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000422,
      "loss": 6.52,
      "mean_token_accuracy": 0.11589490175247193,
      "num_tokens": 1755713.0,
      "step": 845
    },
    {
      "entropy": 6.6110374450683596,
      "epoch": 0.0771044992743106,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004245,
      "loss": 6.5272,
      "mean_token_accuracy": 0.11683018952608108,
      "num_tokens": 1765264.0,
      "step": 850
    },
    {
      "entropy": 6.545265436172485,
      "epoch": 0.07755805515239478,
      "grad_norm": 1.1875,
      "learning_rate": 0.000427,
      "loss": 6.5236,
      "mean_token_accuracy": 0.11411887183785438,
      "num_tokens": 1775769.0,
      "step": 855
    },
    {
      "entropy": 6.497287750244141,
      "epoch": 0.07801161103047896,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004295,
      "loss": 6.4583,
      "mean_token_accuracy": 0.11649047806859017,
      "num_tokens": 1786245.0,
      "step": 860
    },
    {
      "entropy": 6.541773176193237,
      "epoch": 0.07846516690856313,
      "grad_norm": 1.125,
      "learning_rate": 0.000432,
      "loss": 6.4934,
      "mean_token_accuracy": 0.1203076533973217,
      "num_tokens": 1796040.0,
      "step": 865
    },
    {
      "entropy": 6.557205772399902,
      "epoch": 0.07891872278664731,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004345,
      "loss": 6.5032,
      "mean_token_accuracy": 0.11635191962122918,
      "num_tokens": 1806007.0,
      "step": 870
    },
    {
      "entropy": 6.6393091678619385,
      "epoch": 0.0793722786647315,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000437,
      "loss": 6.5811,
      "mean_token_accuracy": 0.10934827327728272,
      "num_tokens": 1816935.0,
      "step": 875
    },
    {
      "entropy": 6.550049209594727,
      "epoch": 0.07982583454281568,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004395,
      "loss": 6.4333,
      "mean_token_accuracy": 0.1151725821197033,
      "num_tokens": 1827419.0,
      "step": 880
    },
    {
      "entropy": 6.5682214260101315,
      "epoch": 0.08027939042089985,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000442,
      "loss": 6.5467,
      "mean_token_accuracy": 0.11440723314881325,
      "num_tokens": 1837933.0,
      "step": 885
    },
    {
      "entropy": 6.561804389953613,
      "epoch": 0.08073294629898403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004445,
      "loss": 6.5946,
      "mean_token_accuracy": 0.11757715567946433,
      "num_tokens": 1848551.0,
      "step": 890
    },
    {
      "entropy": 6.58006534576416,
      "epoch": 0.08118650217706821,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000447,
      "loss": 6.4703,
      "mean_token_accuracy": 0.11650048196315765,
      "num_tokens": 1858380.0,
      "step": 895
    },
    {
      "entropy": 6.507673358917236,
      "epoch": 0.0816400580551524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3925,
      "mean_token_accuracy": 0.12206198945641518,
      "num_tokens": 1868291.0,
      "step": 900
    },
    {
      "entropy": 6.512743759155273,
      "epoch": 0.08209361393323658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.4936,
      "mean_token_accuracy": 0.11916299089789391,
      "num_tokens": 1879219.0,
      "step": 905
    },
    {
      "entropy": 6.576422262191772,
      "epoch": 0.08254716981132075,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4837,
      "mean_token_accuracy": 0.11856920123100281,
      "num_tokens": 1889809.0,
      "step": 910
    },
    {
      "entropy": 6.512048292160034,
      "epoch": 0.08300072568940493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4555,
      "mean_token_accuracy": 0.11503883451223373,
      "num_tokens": 1899986.0,
      "step": 915
    },
    {
      "entropy": 6.468104887008667,
      "epoch": 0.08345428156748912,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4304,
      "mean_token_accuracy": 0.1187589332461357,
      "num_tokens": 1909314.0,
      "step": 920
    },
    {
      "entropy": 6.530900621414185,
      "epoch": 0.0839078374455733,
      "grad_norm": 1.109375,
      "learning_rate": 0.000462,
      "loss": 6.4927,
      "mean_token_accuracy": 0.11667392775416374,
      "num_tokens": 1918443.0,
      "step": 925
    },
    {
      "entropy": 6.4985747814178465,
      "epoch": 0.08436139332365747,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004645,
      "loss": 6.4426,
      "mean_token_accuracy": 0.12001407295465469,
      "num_tokens": 1928683.0,
      "step": 930
    },
    {
      "entropy": 6.341750717163086,
      "epoch": 0.08481494920174165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000467,
      "loss": 6.3506,
      "mean_token_accuracy": 0.11788592040538788,
      "num_tokens": 1938293.0,
      "step": 935
    },
    {
      "entropy": 6.479642105102539,
      "epoch": 0.08526850507982583,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004695,
      "loss": 6.4465,
      "mean_token_accuracy": 0.11793545484542847,
      "num_tokens": 1948233.0,
      "step": 940
    },
    {
      "entropy": 6.530252504348755,
      "epoch": 0.08572206095791002,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000472,
      "loss": 6.3522,
      "mean_token_accuracy": 0.12431273758411407,
      "num_tokens": 1958611.0,
      "step": 945
    },
    {
      "entropy": 6.493687391281128,
      "epoch": 0.0861756168359942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004745,
      "loss": 6.5117,
      "mean_token_accuracy": 0.12384502068161965,
      "num_tokens": 1968957.0,
      "step": 950
    },
    {
      "entropy": 6.461392259597778,
      "epoch": 0.08662917271407837,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000477,
      "loss": 6.3574,
      "mean_token_accuracy": 0.12506917640566825,
      "num_tokens": 1978503.0,
      "step": 955
    },
    {
      "entropy": 6.35320930480957,
      "epoch": 0.08708272859216255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004795,
      "loss": 6.3587,
      "mean_token_accuracy": 0.12145144119858742,
      "num_tokens": 1988458.0,
      "step": 960
    },
    {
      "entropy": 6.526880693435669,
      "epoch": 0.08753628447024674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000482,
      "loss": 6.4944,
      "mean_token_accuracy": 0.11722690165042877,
      "num_tokens": 1999984.0,
      "step": 965
    },
    {
      "entropy": 6.340011119842529,
      "epoch": 0.08798984034833092,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004845,
      "loss": 6.3565,
      "mean_token_accuracy": 0.12677358314394951,
      "num_tokens": 2010645.0,
      "step": 970
    },
    {
      "entropy": 6.513621473312378,
      "epoch": 0.08844339622641509,
      "grad_norm": 1.125,
      "learning_rate": 0.000487,
      "loss": 6.4595,
      "mean_token_accuracy": 0.1156634084880352,
      "num_tokens": 2020196.0,
      "step": 975
    },
    {
      "entropy": 6.416438961029053,
      "epoch": 0.08889695210449927,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004895,
      "loss": 6.3612,
      "mean_token_accuracy": 0.11996159106492996,
      "num_tokens": 2030101.0,
      "step": 980
    },
    {
      "entropy": 6.428423595428467,
      "epoch": 0.08935050798258345,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000492,
      "loss": 6.4427,
      "mean_token_accuracy": 0.12014320492744446,
      "num_tokens": 2040454.0,
      "step": 985
    },
    {
      "entropy": 6.405238437652588,
      "epoch": 0.08980406386066764,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945,
      "loss": 6.3829,
      "mean_token_accuracy": 0.1194852776825428,
      "num_tokens": 2050378.0,
      "step": 990
    },
    {
      "entropy": 6.48555850982666,
      "epoch": 0.09025761973875182,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000497,
      "loss": 6.4375,
      "mean_token_accuracy": 0.1193937435746193,
      "num_tokens": 2060282.0,
      "step": 995
    },
    {
      "entropy": 6.425385618209839,
      "epoch": 0.09071117561683599,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995,
      "loss": 6.4225,
      "mean_token_accuracy": 0.1208973728120327,
      "num_tokens": 2070563.0,
      "step": 1000
    },
    {
      "entropy": 6.519330835342407,
      "epoch": 0.09116473149492017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999985110217,
      "loss": 6.4932,
      "mean_token_accuracy": 0.1223867803812027,
      "num_tokens": 2080153.0,
      "step": 1005
    },
    {
      "entropy": 6.329967784881592,
      "epoch": 0.09161828737300436,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999999924620471,
      "loss": 6.3445,
      "mean_token_accuracy": 0.12452480345964431,
      "num_tokens": 2090590.0,
      "step": 1010
    },
    {
      "entropy": 6.515519332885742,
      "epoch": 0.09207184325108854,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999999817600155,
      "loss": 6.417,
      "mean_token_accuracy": 0.11619279980659485,
      "num_tokens": 2101945.0,
      "step": 1015
    },
    {
      "entropy": 6.467217254638672,
      "epoch": 0.09252539912917271,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999999664049268,
      "loss": 6.4102,
      "mean_token_accuracy": 0.12234249636530876,
      "num_tokens": 2112514.0,
      "step": 1020
    },
    {
      "entropy": 6.314993619918823,
      "epoch": 0.09297895500725689,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999999463967816,
      "loss": 6.3675,
      "mean_token_accuracy": 0.12220757827162743,
      "num_tokens": 2122296.0,
      "step": 1025
    },
    {
      "entropy": 6.554604530334473,
      "epoch": 0.09343251088534107,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049999992173558,
      "loss": 6.5145,
      "mean_token_accuracy": 0.10880634039640427,
      "num_tokens": 2134512.0,
      "step": 1030
    },
    {
      "entropy": 6.392258501052856,
      "epoch": 0.09388606676342526,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999998924213228,
      "loss": 6.4275,
      "mean_token_accuracy": 0.12143760100007057,
      "num_tokens": 2143934.0,
      "step": 1035
    },
    {
      "entropy": 6.342760133743286,
      "epoch": 0.09433962264150944,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999998584540105,
      "loss": 6.3158,
      "mean_token_accuracy": 0.11730768010020257,
      "num_tokens": 2155730.0,
      "step": 1040
    },
    {
      "entropy": 6.444509124755859,
      "epoch": 0.09479317851959361,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999998198336437,
      "loss": 6.35,
      "mean_token_accuracy": 0.11236922815442085,
      "num_tokens": 2166422.0,
      "step": 1045
    },
    {
      "entropy": 6.4285173416137695,
      "epoch": 0.09524673439767779,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999997765602233,
      "loss": 6.4565,
      "mean_token_accuracy": 0.11630769446492195,
      "num_tokens": 2177842.0,
      "step": 1050
    },
    {
      "entropy": 6.47231822013855,
      "epoch": 0.09570029027576198,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999997286337502,
      "loss": 6.5277,
      "mean_token_accuracy": 0.11602402850985527,
      "num_tokens": 2188614.0,
      "step": 1055
    },
    {
      "entropy": 6.416055536270141,
      "epoch": 0.09615384615384616,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999996760542254,
      "loss": 6.3069,
      "mean_token_accuracy": 0.12384679168462753,
      "num_tokens": 2197886.0,
      "step": 1060
    },
    {
      "entropy": 6.2871458530426025,
      "epoch": 0.09660740203193033,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049999961882165,
      "loss": 6.3647,
      "mean_token_accuracy": 0.12341554090380669,
      "num_tokens": 2209930.0,
      "step": 1065
    },
    {
      "entropy": 6.457307767868042,
      "epoch": 0.09706095791001451,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999995569360251,
      "loss": 6.39,
      "mean_token_accuracy": 0.11931687965989113,
      "num_tokens": 2219767.0,
      "step": 1070
    },
    {
      "entropy": 6.368694543838501,
      "epoch": 0.0975145137880987,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499999490397352,
      "loss": 6.3183,
      "mean_token_accuracy": 0.12491154968738556,
      "num_tokens": 2230082.0,
      "step": 1075
    },
    {
      "entropy": 6.255974912643433,
      "epoch": 0.09796806966618288,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999994192056321,
      "loss": 6.3344,
      "mean_token_accuracy": 0.1294071279466152,
      "num_tokens": 2240066.0,
      "step": 1080
    },
    {
      "entropy": 6.41351056098938,
      "epoch": 0.09842162554426705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999993433608669,
      "loss": 6.3732,
      "mean_token_accuracy": 0.121247049421072,
      "num_tokens": 2250732.0,
      "step": 1085
    },
    {
      "entropy": 6.363176012039185,
      "epoch": 0.09887518142235123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999992628630579,
      "loss": 6.2529,
      "mean_token_accuracy": 0.12264698669314385,
      "num_tokens": 2261452.0,
      "step": 1090
    },
    {
      "entropy": 6.3054649353027346,
      "epoch": 0.09932873730043541,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999991777122069,
      "loss": 6.3573,
      "mean_token_accuracy": 0.12316267415881157,
      "num_tokens": 2271929.0,
      "step": 1095
    },
    {
      "entropy": 6.356408214569091,
      "epoch": 0.0997822931785196,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999990879083156,
      "loss": 6.3355,
      "mean_token_accuracy": 0.12437993809580802,
      "num_tokens": 2282987.0,
      "step": 1100
    },
    {
      "entropy": 6.351103973388672,
      "epoch": 0.10023584905660378,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999989934513857,
      "loss": 6.3068,
      "mean_token_accuracy": 0.13111748546361923,
      "num_tokens": 2292796.0,
      "step": 1105
    },
    {
      "entropy": 6.383915185928345,
      "epoch": 0.10068940493468795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999988943414193,
      "loss": 6.3617,
      "mean_token_accuracy": 0.12342413067817688,
      "num_tokens": 2303057.0,
      "step": 1110
    },
    {
      "entropy": 6.292774343490601,
      "epoch": 0.10114296081277213,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999987905784184,
      "loss": 6.3395,
      "mean_token_accuracy": 0.12345447167754173,
      "num_tokens": 2314168.0,
      "step": 1115
    },
    {
      "entropy": 6.347012901306153,
      "epoch": 0.10159651669085631,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999986821623853,
      "loss": 6.3135,
      "mean_token_accuracy": 0.12161569744348526,
      "num_tokens": 2325241.0,
      "step": 1120
    },
    {
      "entropy": 6.389744472503662,
      "epoch": 0.1020500725689405,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999985690933219,
      "loss": 6.3344,
      "mean_token_accuracy": 0.12111192867159844,
      "num_tokens": 2334866.0,
      "step": 1125
    },
    {
      "entropy": 6.2507038593292235,
      "epoch": 0.10250362844702467,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999984513712311,
      "loss": 6.3401,
      "mean_token_accuracy": 0.11989274248480797,
      "num_tokens": 2345422.0,
      "step": 1130
    },
    {
      "entropy": 6.4327864170074465,
      "epoch": 0.10295718432510885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999983289961146,
      "loss": 6.3632,
      "mean_token_accuracy": 0.12089978456497193,
      "num_tokens": 2356285.0,
      "step": 1135
    },
    {
      "entropy": 6.264034223556519,
      "epoch": 0.10341074020319303,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999982019679755,
      "loss": 6.2759,
      "mean_token_accuracy": 0.12783560231328012,
      "num_tokens": 2366861.0,
      "step": 1140
    },
    {
      "entropy": 6.348967456817627,
      "epoch": 0.10386429608127722,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999980702868164,
      "loss": 6.3095,
      "mean_token_accuracy": 0.13093309253454208,
      "num_tokens": 2376851.0,
      "step": 1145
    },
    {
      "entropy": 6.375752353668213,
      "epoch": 0.1043178519593614,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999979339526396,
      "loss": 6.3741,
      "mean_token_accuracy": 0.12913595587015153,
      "num_tokens": 2386857.0,
      "step": 1150
    },
    {
      "entropy": 6.366842555999756,
      "epoch": 0.10477140783744557,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999977929654484,
      "loss": 6.2937,
      "mean_token_accuracy": 0.12512771040201187,
      "num_tokens": 2397035.0,
      "step": 1155
    },
    {
      "entropy": 6.386923742294312,
      "epoch": 0.10522496371552975,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999976473252453,
      "loss": 6.3748,
      "mean_token_accuracy": 0.12434866130352021,
      "num_tokens": 2408177.0,
      "step": 1160
    },
    {
      "entropy": 6.419609022140503,
      "epoch": 0.10567851959361393,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999974970320338,
      "loss": 6.3057,
      "mean_token_accuracy": 0.12743960320949554,
      "num_tokens": 2418445.0,
      "step": 1165
    },
    {
      "entropy": 6.237172079086304,
      "epoch": 0.10613207547169812,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999973420858165,
      "loss": 6.3161,
      "mean_token_accuracy": 0.1236274503171444,
      "num_tokens": 2428990.0,
      "step": 1170
    },
    {
      "entropy": 6.369281435012818,
      "epoch": 0.10658563134978229,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999971824865968,
      "loss": 6.3461,
      "mean_token_accuracy": 0.12022999972105027,
      "num_tokens": 2439643.0,
      "step": 1175
    },
    {
      "entropy": 6.437968683242798,
      "epoch": 0.10703918722786647,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999970182343782,
      "loss": 6.3961,
      "mean_token_accuracy": 0.12295096069574356,
      "num_tokens": 2450691.0,
      "step": 1180
    },
    {
      "entropy": 6.316388273239136,
      "epoch": 0.10749274310595065,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999968493291638,
      "loss": 6.3304,
      "mean_token_accuracy": 0.12266101613640785,
      "num_tokens": 2460706.0,
      "step": 1185
    },
    {
      "entropy": 6.174196100234985,
      "epoch": 0.10794629898403484,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999966757709573,
      "loss": 6.2314,
      "mean_token_accuracy": 0.12887799739837646,
      "num_tokens": 2470947.0,
      "step": 1190
    },
    {
      "entropy": 6.283685493469238,
      "epoch": 0.10839985486211902,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999964975597622,
      "loss": 6.1927,
      "mean_token_accuracy": 0.13100447580218316,
      "num_tokens": 2481132.0,
      "step": 1195
    },
    {
      "entropy": 6.299878740310669,
      "epoch": 0.10885341074020319,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999963146955821,
      "loss": 6.2239,
      "mean_token_accuracy": 0.12988563030958175,
      "num_tokens": 2491701.0,
      "step": 1200
    },
    {
      "entropy": 6.244841003417969,
      "epoch": 0.10930696661828737,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499996127178421,
      "loss": 6.313,
      "mean_token_accuracy": 0.12518261447548867,
      "num_tokens": 2502360.0,
      "step": 1205
    },
    {
      "entropy": 6.283074140548706,
      "epoch": 0.10976052249637155,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999959350082825,
      "loss": 6.1874,
      "mean_token_accuracy": 0.12541429623961448,
      "num_tokens": 2512140.0,
      "step": 1210
    },
    {
      "entropy": 6.323525142669678,
      "epoch": 0.11021407837445574,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999957381851709,
      "loss": 6.3119,
      "mean_token_accuracy": 0.12578000873327255,
      "num_tokens": 2522721.0,
      "step": 1215
    },
    {
      "entropy": 6.276627159118652,
      "epoch": 0.1106676342525399,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999955367090901,
      "loss": 6.177,
      "mean_token_accuracy": 0.13011879548430444,
      "num_tokens": 2531786.0,
      "step": 1220
    },
    {
      "entropy": 6.302189302444458,
      "epoch": 0.11112119013062409,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999953305800441,
      "loss": 6.2937,
      "mean_token_accuracy": 0.12109724506735801,
      "num_tokens": 2542895.0,
      "step": 1225
    },
    {
      "entropy": 6.235093212127685,
      "epoch": 0.11157474600870827,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999951197980374,
      "loss": 6.2842,
      "mean_token_accuracy": 0.12432139739394188,
      "num_tokens": 2554120.0,
      "step": 1230
    },
    {
      "entropy": 6.333096551895141,
      "epoch": 0.11202830188679246,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999949043630744,
      "loss": 6.3243,
      "mean_token_accuracy": 0.12348744869232178,
      "num_tokens": 2563935.0,
      "step": 1235
    },
    {
      "entropy": 6.225638771057129,
      "epoch": 0.11248185776487664,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999946842751593,
      "loss": 6.2389,
      "mean_token_accuracy": 0.12366029843688012,
      "num_tokens": 2574465.0,
      "step": 1240
    },
    {
      "entropy": 6.264844942092895,
      "epoch": 0.11293541364296081,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999944595342968,
      "loss": 6.1886,
      "mean_token_accuracy": 0.13071749806404115,
      "num_tokens": 2584694.0,
      "step": 1245
    },
    {
      "entropy": 6.227580261230469,
      "epoch": 0.11338896952104499,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999942301404916,
      "loss": 6.1615,
      "mean_token_accuracy": 0.1314219854772091,
      "num_tokens": 2594540.0,
      "step": 1250
    },
    {
      "entropy": 6.110536909103393,
      "epoch": 0.11384252539912917,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999939960937484,
      "loss": 6.103,
      "mean_token_accuracy": 0.13548171445727347,
      "num_tokens": 2604896.0,
      "step": 1255
    },
    {
      "entropy": 6.294047069549561,
      "epoch": 0.11429608127721336,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999937573940721,
      "loss": 6.2735,
      "mean_token_accuracy": 0.1363769695162773,
      "num_tokens": 2614143.0,
      "step": 1260
    },
    {
      "entropy": 6.137121486663818,
      "epoch": 0.11474963715529753,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999935140414674,
      "loss": 6.18,
      "mean_token_accuracy": 0.12688928544521333,
      "num_tokens": 2623999.0,
      "step": 1265
    },
    {
      "entropy": 6.380985927581787,
      "epoch": 0.11520319303338171,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999932660359395,
      "loss": 6.3721,
      "mean_token_accuracy": 0.12538434714078903,
      "num_tokens": 2636208.0,
      "step": 1270
    },
    {
      "entropy": 6.249299001693726,
      "epoch": 0.11565674891146589,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999930133774936,
      "loss": 6.2537,
      "mean_token_accuracy": 0.13482384383678436,
      "num_tokens": 2645854.0,
      "step": 1275
    },
    {
      "entropy": 6.322882556915284,
      "epoch": 0.11611030478955008,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999927560661348,
      "loss": 6.2406,
      "mean_token_accuracy": 0.1355570986866951,
      "num_tokens": 2655877.0,
      "step": 1280
    },
    {
      "entropy": 6.303021097183228,
      "epoch": 0.11656386066763426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999924941018685,
      "loss": 6.257,
      "mean_token_accuracy": 0.12821097895503045,
      "num_tokens": 2666157.0,
      "step": 1285
    },
    {
      "entropy": 6.149188089370727,
      "epoch": 0.11701741654571843,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999922274847,
      "loss": 6.1708,
      "mean_token_accuracy": 0.13158801943063736,
      "num_tokens": 2676647.0,
      "step": 1290
    },
    {
      "entropy": 6.261129522323609,
      "epoch": 0.11747097242380261,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499991956214635,
      "loss": 6.2095,
      "mean_token_accuracy": 0.14001647382974625,
      "num_tokens": 2686417.0,
      "step": 1295
    },
    {
      "entropy": 6.232670593261719,
      "epoch": 0.1179245283018868,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499991680291679,
      "loss": 6.2224,
      "mean_token_accuracy": 0.1243120551109314,
      "num_tokens": 2697540.0,
      "step": 1300
    },
    {
      "entropy": 6.192728900909424,
      "epoch": 0.11837808417997098,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999913997158378,
      "loss": 6.2607,
      "mean_token_accuracy": 0.12081601396203041,
      "num_tokens": 2709397.0,
      "step": 1305
    },
    {
      "entropy": 6.286089134216309,
      "epoch": 0.11883164005805515,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999911144871169,
      "loss": 6.2128,
      "mean_token_accuracy": 0.12598777934908867,
      "num_tokens": 2720430.0,
      "step": 1310
    },
    {
      "entropy": 6.178103065490722,
      "epoch": 0.11928519593613933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999908246055226,
      "loss": 6.1701,
      "mean_token_accuracy": 0.12547316178679466,
      "num_tokens": 2730648.0,
      "step": 1315
    },
    {
      "entropy": 6.184869623184204,
      "epoch": 0.11973875181422351,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999905300710607,
      "loss": 6.1406,
      "mean_token_accuracy": 0.1283143639564514,
      "num_tokens": 2739979.0,
      "step": 1320
    },
    {
      "entropy": 6.296625852584839,
      "epoch": 0.1201923076923077,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999902308837373,
      "loss": 6.2786,
      "mean_token_accuracy": 0.12489124983549119,
      "num_tokens": 2750329.0,
      "step": 1325
    },
    {
      "entropy": 6.146040248870849,
      "epoch": 0.12064586357039188,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999899270435584,
      "loss": 6.1522,
      "mean_token_accuracy": 0.13480892404913902,
      "num_tokens": 2760535.0,
      "step": 1330
    },
    {
      "entropy": 6.269543600082398,
      "epoch": 0.12109941944847605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999896185505307,
      "loss": 6.1902,
      "mean_token_accuracy": 0.12844027504324912,
      "num_tokens": 2770470.0,
      "step": 1335
    },
    {
      "entropy": 6.204631614685058,
      "epoch": 0.12155297532656023,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999893054046603,
      "loss": 6.2205,
      "mean_token_accuracy": 0.12515390738844873,
      "num_tokens": 2781506.0,
      "step": 1340
    },
    {
      "entropy": 6.161186933517456,
      "epoch": 0.12200653120464441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999889876059537,
      "loss": 6.0858,
      "mean_token_accuracy": 0.13881843388080597,
      "num_tokens": 2791245.0,
      "step": 1345
    },
    {
      "entropy": 6.227107095718384,
      "epoch": 0.1224600870827286,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999886651544176,
      "loss": 6.1082,
      "mean_token_accuracy": 0.13308778330683707,
      "num_tokens": 2801680.0,
      "step": 1350
    },
    {
      "entropy": 6.124445152282715,
      "epoch": 0.12291364296081277,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999883380500584,
      "loss": 6.1789,
      "mean_token_accuracy": 0.1382739506661892,
      "num_tokens": 2812805.0,
      "step": 1355
    },
    {
      "entropy": 6.302854251861572,
      "epoch": 0.12336719883889695,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999880062928831,
      "loss": 6.2143,
      "mean_token_accuracy": 0.1299120791256428,
      "num_tokens": 2822834.0,
      "step": 1360
    },
    {
      "entropy": 6.152606868743897,
      "epoch": 0.12382075471698113,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999876698828985,
      "loss": 6.1596,
      "mean_token_accuracy": 0.13407627418637275,
      "num_tokens": 2832312.0,
      "step": 1365
    },
    {
      "entropy": 6.221649599075318,
      "epoch": 0.12427431059506532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999873288201116,
      "loss": 6.1793,
      "mean_token_accuracy": 0.1410585030913353,
      "num_tokens": 2842998.0,
      "step": 1370
    },
    {
      "entropy": 6.259373044967651,
      "epoch": 0.1247278664731495,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004999869831045294,
      "loss": 6.1999,
      "mean_token_accuracy": 0.13460145518183708,
      "num_tokens": 2854218.0,
      "step": 1375
    },
    {
      "entropy": 6.121993970870972,
      "epoch": 0.12518142235123367,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999866327361589,
      "loss": 6.096,
      "mean_token_accuracy": 0.13676728904247284,
      "num_tokens": 2864484.0,
      "step": 1380
    },
    {
      "entropy": 6.224773120880127,
      "epoch": 0.12563497822931785,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999862777150077,
      "loss": 6.2064,
      "mean_token_accuracy": 0.1293405920267105,
      "num_tokens": 2873800.0,
      "step": 1385
    },
    {
      "entropy": 6.2337018013000485,
      "epoch": 0.12608853410740203,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999859180410829,
      "loss": 6.2067,
      "mean_token_accuracy": 0.12882139533758163,
      "num_tokens": 2884247.0,
      "step": 1390
    },
    {
      "entropy": 6.138038253784179,
      "epoch": 0.12654208998548622,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999855537143919,
      "loss": 6.1086,
      "mean_token_accuracy": 0.13381363451480865,
      "num_tokens": 2894739.0,
      "step": 1395
    },
    {
      "entropy": 6.129825210571289,
      "epoch": 0.1269956458635704,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999851847349425,
      "loss": 6.1729,
      "mean_token_accuracy": 0.1333797574043274,
      "num_tokens": 2904376.0,
      "step": 1400
    },
    {
      "entropy": 6.258524942398071,
      "epoch": 0.12744920174165458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999848111027419,
      "loss": 6.2175,
      "mean_token_accuracy": 0.1292863242328167,
      "num_tokens": 2914304.0,
      "step": 1405
    },
    {
      "entropy": 6.214796018600464,
      "epoch": 0.12790275761973874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999844328177983,
      "loss": 6.1028,
      "mean_token_accuracy": 0.13051475137472152,
      "num_tokens": 2924400.0,
      "step": 1410
    },
    {
      "entropy": 6.328621816635132,
      "epoch": 0.12835631349782292,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999840498801191,
      "loss": 6.3489,
      "mean_token_accuracy": 0.1303907461464405,
      "num_tokens": 2935040.0,
      "step": 1415
    },
    {
      "entropy": 6.081346130371093,
      "epoch": 0.1288098693759071,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999836622897126,
      "loss": 6.1337,
      "mean_token_accuracy": 0.13372014090418816,
      "num_tokens": 2945536.0,
      "step": 1420
    },
    {
      "entropy": 6.166437578201294,
      "epoch": 0.1292634252539913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999832700465865,
      "loss": 6.0475,
      "mean_token_accuracy": 0.13463784158229827,
      "num_tokens": 2954969.0,
      "step": 1425
    },
    {
      "entropy": 6.1763997077941895,
      "epoch": 0.12971698113207547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999828731507491,
      "loss": 6.1646,
      "mean_token_accuracy": 0.12978584319353104,
      "num_tokens": 2966218.0,
      "step": 1430
    },
    {
      "entropy": 6.1608034610748295,
      "epoch": 0.13017053701015965,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999824716022085,
      "loss": 6.2042,
      "mean_token_accuracy": 0.12838013991713523,
      "num_tokens": 2976640.0,
      "step": 1435
    },
    {
      "entropy": 6.1678853034973145,
      "epoch": 0.13062409288824384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999820654009731,
      "loss": 6.1584,
      "mean_token_accuracy": 0.13406179323792458,
      "num_tokens": 2985883.0,
      "step": 1440
    },
    {
      "entropy": 6.187309789657593,
      "epoch": 0.13107764876632802,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999816545470513,
      "loss": 6.1702,
      "mean_token_accuracy": 0.1267956852912903,
      "num_tokens": 2996695.0,
      "step": 1445
    },
    {
      "entropy": 6.194827270507813,
      "epoch": 0.1315312046444122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999812390404514,
      "loss": 6.1609,
      "mean_token_accuracy": 0.13191465437412261,
      "num_tokens": 3007500.0,
      "step": 1450
    },
    {
      "entropy": 6.0264750003814695,
      "epoch": 0.13198476052249636,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999808188811823,
      "loss": 6.0252,
      "mean_token_accuracy": 0.13417760878801346,
      "num_tokens": 3017299.0,
      "step": 1455
    },
    {
      "entropy": 6.300446462631226,
      "epoch": 0.13243831640058054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999803940692524,
      "loss": 6.3135,
      "mean_token_accuracy": 0.12158919349312783,
      "num_tokens": 3028389.0,
      "step": 1460
    },
    {
      "entropy": 6.197498369216919,
      "epoch": 0.13289187227866472,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999799646046707,
      "loss": 6.1173,
      "mean_token_accuracy": 0.13784717693924903,
      "num_tokens": 3038883.0,
      "step": 1465
    },
    {
      "entropy": 6.177143383026123,
      "epoch": 0.1333454281567489,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999795304874461,
      "loss": 6.2414,
      "mean_token_accuracy": 0.13582047894597055,
      "num_tokens": 3049904.0,
      "step": 1470
    },
    {
      "entropy": 6.133323240280151,
      "epoch": 0.1337989840348331,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999790917175873,
      "loss": 6.078,
      "mean_token_accuracy": 0.13267199248075484,
      "num_tokens": 3059788.0,
      "step": 1475
    },
    {
      "entropy": 6.19330358505249,
      "epoch": 0.13425253991291727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999786482951036,
      "loss": 6.0752,
      "mean_token_accuracy": 0.1339847445487976,
      "num_tokens": 3070413.0,
      "step": 1480
    },
    {
      "entropy": 6.126073980331421,
      "epoch": 0.13470609579100146,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999782002200041,
      "loss": 6.1987,
      "mean_token_accuracy": 0.1323849603533745,
      "num_tokens": 3081093.0,
      "step": 1485
    },
    {
      "entropy": 6.15864782333374,
      "epoch": 0.13515965166908564,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999777474922982,
      "loss": 6.124,
      "mean_token_accuracy": 0.12964514642953873,
      "num_tokens": 3090594.0,
      "step": 1490
    },
    {
      "entropy": 6.221482515335083,
      "epoch": 0.13561320754716982,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999772901119951,
      "loss": 6.2361,
      "mean_token_accuracy": 0.1309862792491913,
      "num_tokens": 3101061.0,
      "step": 1495
    },
    {
      "entropy": 6.122462558746338,
      "epoch": 0.13606676342525398,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999768280791043,
      "loss": 6.1558,
      "mean_token_accuracy": 0.13689188212156295,
      "num_tokens": 3111715.0,
      "step": 1500
    },
    {
      "entropy": 6.163290500640869,
      "epoch": 0.13652031930333816,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999763613936352,
      "loss": 6.0644,
      "mean_token_accuracy": 0.13736301958560942,
      "num_tokens": 3121378.0,
      "step": 1505
    },
    {
      "entropy": 6.096087884902954,
      "epoch": 0.13697387518142234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999758900555979,
      "loss": 5.9987,
      "mean_token_accuracy": 0.1399257391691208,
      "num_tokens": 3131229.0,
      "step": 1510
    },
    {
      "entropy": 6.0675342082977295,
      "epoch": 0.13742743105950653,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999754140650017,
      "loss": 6.1017,
      "mean_token_accuracy": 0.13268654868006707,
      "num_tokens": 3140859.0,
      "step": 1515
    },
    {
      "entropy": 6.186848449707031,
      "epoch": 0.1378809869375907,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999749334218567,
      "loss": 6.1,
      "mean_token_accuracy": 0.13572516441345214,
      "num_tokens": 3151692.0,
      "step": 1520
    },
    {
      "entropy": 6.060773086547852,
      "epoch": 0.1383345428156749,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999744481261726,
      "loss": 6.1658,
      "mean_token_accuracy": 0.1306878425180912,
      "num_tokens": 3161879.0,
      "step": 1525
    },
    {
      "entropy": 6.109863138198852,
      "epoch": 0.13878809869375908,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999739581779597,
      "loss": 6.0243,
      "mean_token_accuracy": 0.1430966079235077,
      "num_tokens": 3171282.0,
      "step": 1530
    },
    {
      "entropy": 6.160919523239135,
      "epoch": 0.13924165457184326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499973463577228,
      "loss": 6.0849,
      "mean_token_accuracy": 0.13672013431787491,
      "num_tokens": 3181398.0,
      "step": 1535
    },
    {
      "entropy": 6.1300629615783695,
      "epoch": 0.13969521044992744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999729643239877,
      "loss": 6.2069,
      "mean_token_accuracy": 0.1347118280827999,
      "num_tokens": 3191431.0,
      "step": 1540
    },
    {
      "entropy": 6.1940147399902346,
      "epoch": 0.1401487663280116,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999724604182492,
      "loss": 6.2112,
      "mean_token_accuracy": 0.12552011385560036,
      "num_tokens": 3202147.0,
      "step": 1545
    },
    {
      "entropy": 6.215757179260254,
      "epoch": 0.14060232220609578,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999719518600229,
      "loss": 6.069,
      "mean_token_accuracy": 0.13700326383113862,
      "num_tokens": 3212428.0,
      "step": 1550
    },
    {
      "entropy": 6.07675952911377,
      "epoch": 0.14105587808417996,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999714386493192,
      "loss": 6.1366,
      "mean_token_accuracy": 0.1377066530287266,
      "num_tokens": 3222099.0,
      "step": 1555
    },
    {
      "entropy": 6.184482622146606,
      "epoch": 0.14150943396226415,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999709207861489,
      "loss": 6.0864,
      "mean_token_accuracy": 0.13630361706018448,
      "num_tokens": 3232079.0,
      "step": 1560
    },
    {
      "entropy": 6.0271689891815186,
      "epoch": 0.14196298984034833,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999703982705227,
      "loss": 6.0002,
      "mean_token_accuracy": 0.14721214026212692,
      "num_tokens": 3241762.0,
      "step": 1565
    },
    {
      "entropy": 6.100229215621948,
      "epoch": 0.1424165457184325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999698711024513,
      "loss": 6.0699,
      "mean_token_accuracy": 0.13954419642686844,
      "num_tokens": 3252571.0,
      "step": 1570
    },
    {
      "entropy": 6.0743285655975345,
      "epoch": 0.1428701015965167,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999693392819456,
      "loss": 6.1225,
      "mean_token_accuracy": 0.13330407217144966,
      "num_tokens": 3263542.0,
      "step": 1575
    },
    {
      "entropy": 6.2108501434326175,
      "epoch": 0.14332365747460088,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999688028090166,
      "loss": 6.1965,
      "mean_token_accuracy": 0.13395313844084739,
      "num_tokens": 3273295.0,
      "step": 1580
    },
    {
      "entropy": 6.2388311386108395,
      "epoch": 0.14377721335268506,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999682616836755,
      "loss": 6.2142,
      "mean_token_accuracy": 0.13328842967748641,
      "num_tokens": 3283605.0,
      "step": 1585
    },
    {
      "entropy": 6.087570333480835,
      "epoch": 0.14423076923076922,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999677159059334,
      "loss": 6.0438,
      "mean_token_accuracy": 0.14040226265788078,
      "num_tokens": 3293837.0,
      "step": 1590
    },
    {
      "entropy": 5.992144441604614,
      "epoch": 0.1446843251088534,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999671654758016,
      "loss": 6.0574,
      "mean_token_accuracy": 0.14208618625998498,
      "num_tokens": 3304131.0,
      "step": 1595
    },
    {
      "entropy": 6.161336088180542,
      "epoch": 0.14513788098693758,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004999666103932915,
      "loss": 6.0925,
      "mean_token_accuracy": 0.13309103175997733,
      "num_tokens": 3313543.0,
      "step": 1600
    },
    {
      "entropy": 6.0388938903808596,
      "epoch": 0.14559143686502177,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999660506584145,
      "loss": 6.0621,
      "mean_token_accuracy": 0.13285203129053116,
      "num_tokens": 3324487.0,
      "step": 1605
    },
    {
      "entropy": 6.177606534957886,
      "epoch": 0.14604499274310595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999654862711823,
      "loss": 6.085,
      "mean_token_accuracy": 0.13585741370916365,
      "num_tokens": 3335094.0,
      "step": 1610
    },
    {
      "entropy": 6.041245794296264,
      "epoch": 0.14649854862119013,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999649172316065,
      "loss": 6.0213,
      "mean_token_accuracy": 0.13612035140395165,
      "num_tokens": 3344252.0,
      "step": 1615
    },
    {
      "entropy": 6.07570447921753,
      "epoch": 0.14695210449927432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499964343539699,
      "loss": 6.1063,
      "mean_token_accuracy": 0.1323385626077652,
      "num_tokens": 3354265.0,
      "step": 1620
    },
    {
      "entropy": 6.204560565948486,
      "epoch": 0.1474056603773585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999637651954714,
      "loss": 6.2092,
      "mean_token_accuracy": 0.13653515949845313,
      "num_tokens": 3364616.0,
      "step": 1625
    },
    {
      "entropy": 6.1248493671417235,
      "epoch": 0.14785921625544268,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999631821989358,
      "loss": 6.1551,
      "mean_token_accuracy": 0.13333765044808388,
      "num_tokens": 3373650.0,
      "step": 1630
    },
    {
      "entropy": 6.091972589492798,
      "epoch": 0.14831277213352684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999625945501043,
      "loss": 6.0473,
      "mean_token_accuracy": 0.13804969042539597,
      "num_tokens": 3384590.0,
      "step": 1635
    },
    {
      "entropy": 6.157991361618042,
      "epoch": 0.14876632801161102,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499962002248989,
      "loss": 6.0816,
      "mean_token_accuracy": 0.13664070069789885,
      "num_tokens": 3395616.0,
      "step": 1640
    },
    {
      "entropy": 6.1123932838439945,
      "epoch": 0.1492198838896952,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999614052956024,
      "loss": 6.0892,
      "mean_token_accuracy": 0.13248431012034417,
      "num_tokens": 3405810.0,
      "step": 1645
    },
    {
      "entropy": 6.143494653701782,
      "epoch": 0.1496734397677794,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999608036899563,
      "loss": 6.1237,
      "mean_token_accuracy": 0.13100728020071983,
      "num_tokens": 3415932.0,
      "step": 1650
    },
    {
      "entropy": 6.0603516578674315,
      "epoch": 0.15012699564586357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999601974320636,
      "loss": 6.0703,
      "mean_token_accuracy": 0.13310141935944558,
      "num_tokens": 3425626.0,
      "step": 1655
    },
    {
      "entropy": 6.097725772857666,
      "epoch": 0.15058055152394775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999595865219367,
      "loss": 6.0662,
      "mean_token_accuracy": 0.1289604127407074,
      "num_tokens": 3434837.0,
      "step": 1660
    },
    {
      "entropy": 6.11887493133545,
      "epoch": 0.15103410740203194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999589709595881,
      "loss": 6.0818,
      "mean_token_accuracy": 0.14125847369432448,
      "num_tokens": 3445238.0,
      "step": 1665
    },
    {
      "entropy": 6.03086633682251,
      "epoch": 0.15148766328011612,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999583507450308,
      "loss": 6.0619,
      "mean_token_accuracy": 0.1350873090326786,
      "num_tokens": 3456738.0,
      "step": 1670
    },
    {
      "entropy": 6.022101402282715,
      "epoch": 0.1519412191582003,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999577258782774,
      "loss": 5.9519,
      "mean_token_accuracy": 0.13770247548818587,
      "num_tokens": 3466296.0,
      "step": 1675
    },
    {
      "entropy": 6.160738801956176,
      "epoch": 0.15239477503628446,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499957096359341,
      "loss": 6.1438,
      "mean_token_accuracy": 0.13199611157178878,
      "num_tokens": 3476989.0,
      "step": 1680
    },
    {
      "entropy": 6.0629199028015135,
      "epoch": 0.15284833091436864,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999564621882343,
      "loss": 5.9823,
      "mean_token_accuracy": 0.14023435190320016,
      "num_tokens": 3487050.0,
      "step": 1685
    },
    {
      "entropy": 6.138583707809448,
      "epoch": 0.15330188679245282,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999558233649708,
      "loss": 6.1963,
      "mean_token_accuracy": 0.1306483641266823,
      "num_tokens": 3497777.0,
      "step": 1690
    },
    {
      "entropy": 6.148853921890259,
      "epoch": 0.153755442670537,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999551798895636,
      "loss": 6.0907,
      "mean_token_accuracy": 0.133428006619215,
      "num_tokens": 3509066.0,
      "step": 1695
    },
    {
      "entropy": 6.038267374038696,
      "epoch": 0.1542089985486212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999545317620258,
      "loss": 6.0262,
      "mean_token_accuracy": 0.13769044652581214,
      "num_tokens": 3519567.0,
      "step": 1700
    },
    {
      "entropy": 6.028669166564941,
      "epoch": 0.15466255442670537,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999538789823711,
      "loss": 6.0521,
      "mean_token_accuracy": 0.14178896844387054,
      "num_tokens": 3529402.0,
      "step": 1705
    },
    {
      "entropy": 6.16622405052185,
      "epoch": 0.15511611030478956,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999532215506127,
      "loss": 6.0718,
      "mean_token_accuracy": 0.14612792134284974,
      "num_tokens": 3539805.0,
      "step": 1710
    },
    {
      "entropy": 5.986016368865966,
      "epoch": 0.15556966618287374,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999525594667645,
      "loss": 5.9352,
      "mean_token_accuracy": 0.14136730134487152,
      "num_tokens": 3550477.0,
      "step": 1715
    },
    {
      "entropy": 5.952186918258667,
      "epoch": 0.15602322206095792,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049995189273084,
      "loss": 6.0427,
      "mean_token_accuracy": 0.1397734761238098,
      "num_tokens": 3560729.0,
      "step": 1720
    },
    {
      "entropy": 6.097905158996582,
      "epoch": 0.15647677793904208,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499951221342853,
      "loss": 6.1137,
      "mean_token_accuracy": 0.12626646235585212,
      "num_tokens": 3571538.0,
      "step": 1725
    },
    {
      "entropy": 6.125197029113769,
      "epoch": 0.15693033381712626,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999505453028174,
      "loss": 6.1197,
      "mean_token_accuracy": 0.1374218739569187,
      "num_tokens": 3583128.0,
      "step": 1730
    },
    {
      "entropy": 6.121435117721558,
      "epoch": 0.15738388969521044,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999498646107472,
      "loss": 6.0475,
      "mean_token_accuracy": 0.1424655109643936,
      "num_tokens": 3594358.0,
      "step": 1735
    },
    {
      "entropy": 6.065750217437744,
      "epoch": 0.15783744557329463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999491792666566,
      "loss": 5.9753,
      "mean_token_accuracy": 0.14380643963813783,
      "num_tokens": 3604345.0,
      "step": 1740
    },
    {
      "entropy": 6.030856895446777,
      "epoch": 0.1582910014513788,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999484892705597,
      "loss": 6.0287,
      "mean_token_accuracy": 0.13884891346096992,
      "num_tokens": 3616129.0,
      "step": 1745
    },
    {
      "entropy": 6.111476278305053,
      "epoch": 0.158744557329463,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999477946224705,
      "loss": 6.0824,
      "mean_token_accuracy": 0.13742278516292572,
      "num_tokens": 3627458.0,
      "step": 1750
    },
    {
      "entropy": 6.081455039978027,
      "epoch": 0.15919811320754718,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999470953224037,
      "loss": 5.9647,
      "mean_token_accuracy": 0.13912637010216714,
      "num_tokens": 3636047.0,
      "step": 1755
    },
    {
      "entropy": 5.941955947875977,
      "epoch": 0.15965166908563136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999463913703735,
      "loss": 5.956,
      "mean_token_accuracy": 0.14151325970888137,
      "num_tokens": 3646930.0,
      "step": 1760
    },
    {
      "entropy": 6.053501844406128,
      "epoch": 0.16010522496371554,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999456827663948,
      "loss": 6.0337,
      "mean_token_accuracy": 0.14219933450222016,
      "num_tokens": 3657279.0,
      "step": 1765
    },
    {
      "entropy": 6.029218292236328,
      "epoch": 0.1605587808417997,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499944969510482,
      "loss": 5.948,
      "mean_token_accuracy": 0.14011043757200242,
      "num_tokens": 3667070.0,
      "step": 1770
    },
    {
      "entropy": 6.026233434677124,
      "epoch": 0.16101233671988388,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999442516026498,
      "loss": 5.9741,
      "mean_token_accuracy": 0.13762375786900521,
      "num_tokens": 3677560.0,
      "step": 1775
    },
    {
      "entropy": 5.963876390457154,
      "epoch": 0.16146589259796806,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999435290429133,
      "loss": 5.934,
      "mean_token_accuracy": 0.14428269788622855,
      "num_tokens": 3686701.0,
      "step": 1780
    },
    {
      "entropy": 6.075827646255493,
      "epoch": 0.16191944847605225,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999428018312872,
      "loss": 6.035,
      "mean_token_accuracy": 0.14326198846101762,
      "num_tokens": 3696338.0,
      "step": 1785
    },
    {
      "entropy": 6.04256386756897,
      "epoch": 0.16237300435413643,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999420699677867,
      "loss": 6.0059,
      "mean_token_accuracy": 0.141026820987463,
      "num_tokens": 3705752.0,
      "step": 1790
    },
    {
      "entropy": 6.1249024868011475,
      "epoch": 0.1628265602322206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999413334524269,
      "loss": 6.0907,
      "mean_token_accuracy": 0.13355766385793685,
      "num_tokens": 3716225.0,
      "step": 1795
    },
    {
      "entropy": 5.983501863479614,
      "epoch": 0.1632801161103048,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999405922852229,
      "loss": 5.9908,
      "mean_token_accuracy": 0.13743126392364502,
      "num_tokens": 3726593.0,
      "step": 1800
    },
    {
      "entropy": 6.018549251556396,
      "epoch": 0.16373367198838898,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999398464661901,
      "loss": 6.0353,
      "mean_token_accuracy": 0.13529310375452042,
      "num_tokens": 3737563.0,
      "step": 1805
    },
    {
      "entropy": 6.175122833251953,
      "epoch": 0.16418722786647316,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999390959953441,
      "loss": 6.0825,
      "mean_token_accuracy": 0.1384120099246502,
      "num_tokens": 3747752.0,
      "step": 1810
    },
    {
      "entropy": 5.976292610168457,
      "epoch": 0.16464078374455732,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999383408727002,
      "loss": 6.0091,
      "mean_token_accuracy": 0.13851915821433067,
      "num_tokens": 3758795.0,
      "step": 1815
    },
    {
      "entropy": 6.05274600982666,
      "epoch": 0.1650943396226415,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999375810982741,
      "loss": 6.0132,
      "mean_token_accuracy": 0.1393853485584259,
      "num_tokens": 3768690.0,
      "step": 1820
    },
    {
      "entropy": 5.971596431732178,
      "epoch": 0.16554789550072568,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999368166720815,
      "loss": 6.0593,
      "mean_token_accuracy": 0.13606920316815377,
      "num_tokens": 3779018.0,
      "step": 1825
    },
    {
      "entropy": 6.16007022857666,
      "epoch": 0.16600145137880987,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999360475941382,
      "loss": 5.9574,
      "mean_token_accuracy": 0.14323117956519127,
      "num_tokens": 3788985.0,
      "step": 1830
    },
    {
      "entropy": 6.033173227310181,
      "epoch": 0.16645500725689405,
      "grad_norm": 1.0,
      "learning_rate": 0.00049993527386446,
      "loss": 6.1029,
      "mean_token_accuracy": 0.1309938281774521,
      "num_tokens": 3800787.0,
      "step": 1835
    },
    {
      "entropy": 6.060845899581909,
      "epoch": 0.16690856313497823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999344954830633,
      "loss": 5.9746,
      "mean_token_accuracy": 0.13664656654000282,
      "num_tokens": 3811381.0,
      "step": 1840
    },
    {
      "entropy": 6.05128288269043,
      "epoch": 0.16736211901306242,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999337124499638,
      "loss": 6.0943,
      "mean_token_accuracy": 0.14225277155637742,
      "num_tokens": 3822055.0,
      "step": 1845
    },
    {
      "entropy": 6.048533248901367,
      "epoch": 0.1678156748911466,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999329247651777,
      "loss": 6.0429,
      "mean_token_accuracy": 0.13500285595655442,
      "num_tokens": 3831900.0,
      "step": 1850
    },
    {
      "entropy": 6.054124593734741,
      "epoch": 0.16826923076923078,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999321324287213,
      "loss": 6.0071,
      "mean_token_accuracy": 0.1393166773021221,
      "num_tokens": 3842702.0,
      "step": 1855
    },
    {
      "entropy": 6.126334619522095,
      "epoch": 0.16872278664731494,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999313354406114,
      "loss": 5.9721,
      "mean_token_accuracy": 0.13867918774485588,
      "num_tokens": 3852690.0,
      "step": 1860
    },
    {
      "entropy": 5.936737442016602,
      "epoch": 0.16917634252539912,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499930533800864,
      "loss": 5.9698,
      "mean_token_accuracy": 0.14831455647945405,
      "num_tokens": 3863040.0,
      "step": 1865
    },
    {
      "entropy": 6.028145837783813,
      "epoch": 0.1696298984034833,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999297275094958,
      "loss": 6.0554,
      "mean_token_accuracy": 0.13361043334007264,
      "num_tokens": 3873613.0,
      "step": 1870
    },
    {
      "entropy": 5.9600602149963375,
      "epoch": 0.1700834542815675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999289165665235,
      "loss": 5.8696,
      "mean_token_accuracy": 0.14620799273252488,
      "num_tokens": 3883936.0,
      "step": 1875
    },
    {
      "entropy": 5.9179552555084225,
      "epoch": 0.17053701015965167,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999281009719639,
      "loss": 5.8189,
      "mean_token_accuracy": 0.14930711612105368,
      "num_tokens": 3893788.0,
      "step": 1880
    },
    {
      "entropy": 5.955077457427978,
      "epoch": 0.17099056603773585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999272807258339,
      "loss": 5.9045,
      "mean_token_accuracy": 0.14696305319666864,
      "num_tokens": 3903434.0,
      "step": 1885
    },
    {
      "entropy": 5.98124589920044,
      "epoch": 0.17144412191582004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999264558281504,
      "loss": 5.9164,
      "mean_token_accuracy": 0.14665744602680206,
      "num_tokens": 3913476.0,
      "step": 1890
    },
    {
      "entropy": 5.882257986068725,
      "epoch": 0.17189767779390422,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999256262789305,
      "loss": 5.8757,
      "mean_token_accuracy": 0.14397282898426056,
      "num_tokens": 3923301.0,
      "step": 1895
    },
    {
      "entropy": 6.017929792404175,
      "epoch": 0.1723512336719884,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999247920781913,
      "loss": 6.0171,
      "mean_token_accuracy": 0.1397230200469494,
      "num_tokens": 3933335.0,
      "step": 1900
    },
    {
      "entropy": 6.048897361755371,
      "epoch": 0.17280478955007256,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999239532259502,
      "loss": 6.0409,
      "mean_token_accuracy": 0.13537009730935096,
      "num_tokens": 3944171.0,
      "step": 1905
    },
    {
      "entropy": 6.089568996429444,
      "epoch": 0.17325834542815674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999231097222244,
      "loss": 6.1116,
      "mean_token_accuracy": 0.13979827240109444,
      "num_tokens": 3954160.0,
      "step": 1910
    },
    {
      "entropy": 6.0698545455932615,
      "epoch": 0.17371190130624092,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999222615670312,
      "loss": 5.9036,
      "mean_token_accuracy": 0.14813649281859398,
      "num_tokens": 3964110.0,
      "step": 1915
    },
    {
      "entropy": 5.989406299591065,
      "epoch": 0.1741654571843251,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999214087603885,
      "loss": 6.0022,
      "mean_token_accuracy": 0.13940656036138535,
      "num_tokens": 3974293.0,
      "step": 1920
    },
    {
      "entropy": 5.934039068222046,
      "epoch": 0.1746190130624093,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999205513023135,
      "loss": 5.9553,
      "mean_token_accuracy": 0.14510801061987877,
      "num_tokens": 3984346.0,
      "step": 1925
    },
    {
      "entropy": 5.9414143562316895,
      "epoch": 0.17507256894049347,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999196891928245,
      "loss": 5.8849,
      "mean_token_accuracy": 0.1501953735947609,
      "num_tokens": 3992897.0,
      "step": 1930
    },
    {
      "entropy": 5.9804035186767575,
      "epoch": 0.17552612481857766,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999188224319388,
      "loss": 5.9295,
      "mean_token_accuracy": 0.1484602600336075,
      "num_tokens": 4002575.0,
      "step": 1935
    },
    {
      "entropy": 5.99068865776062,
      "epoch": 0.17597968069666184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999179510196746,
      "loss": 5.9908,
      "mean_token_accuracy": 0.1398739770054817,
      "num_tokens": 4013752.0,
      "step": 1940
    },
    {
      "entropy": 6.032836437225342,
      "epoch": 0.17643323657474602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999170749560498,
      "loss": 5.8863,
      "mean_token_accuracy": 0.14885374158620834,
      "num_tokens": 4023718.0,
      "step": 1945
    },
    {
      "entropy": 5.925063753128052,
      "epoch": 0.17688679245283018,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999161942410827,
      "loss": 5.9889,
      "mean_token_accuracy": 0.1374013051390648,
      "num_tokens": 4034210.0,
      "step": 1950
    },
    {
      "entropy": 6.060517311096191,
      "epoch": 0.17734034833091436,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999153088747913,
      "loss": 5.9787,
      "mean_token_accuracy": 0.139659071713686,
      "num_tokens": 4044868.0,
      "step": 1955
    },
    {
      "entropy": 5.909337139129638,
      "epoch": 0.17779390420899854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499914418857194,
      "loss": 5.9873,
      "mean_token_accuracy": 0.1421900287270546,
      "num_tokens": 4055049.0,
      "step": 1960
    },
    {
      "entropy": 6.143319940567016,
      "epoch": 0.17824746008708273,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999135241883092,
      "loss": 6.0254,
      "mean_token_accuracy": 0.14073289632797242,
      "num_tokens": 4065889.0,
      "step": 1965
    },
    {
      "entropy": 5.994856882095337,
      "epoch": 0.1787010159651669,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999126248681555,
      "loss": 6.0474,
      "mean_token_accuracy": 0.13780821710824967,
      "num_tokens": 4077288.0,
      "step": 1970
    },
    {
      "entropy": 6.0060615062713625,
      "epoch": 0.1791545718432511,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999117208967513,
      "loss": 5.9939,
      "mean_token_accuracy": 0.14116482213139533,
      "num_tokens": 4087402.0,
      "step": 1975
    },
    {
      "entropy": 6.124054670333862,
      "epoch": 0.17960812772133528,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999108122741154,
      "loss": 6.0457,
      "mean_token_accuracy": 0.13824116289615632,
      "num_tokens": 4098150.0,
      "step": 1980
    },
    {
      "entropy": 5.931825828552246,
      "epoch": 0.18006168359941946,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999098990002666,
      "loss": 5.9012,
      "mean_token_accuracy": 0.14494443386793138,
      "num_tokens": 4108134.0,
      "step": 1985
    },
    {
      "entropy": 6.072109079360962,
      "epoch": 0.18051523947750364,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999089810752237,
      "loss": 6.0357,
      "mean_token_accuracy": 0.13650458827614784,
      "num_tokens": 4118251.0,
      "step": 1990
    },
    {
      "entropy": 5.921803188323975,
      "epoch": 0.1809687953555878,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499908058499006,
      "loss": 5.8887,
      "mean_token_accuracy": 0.1494055435061455,
      "num_tokens": 4129228.0,
      "step": 1995
    },
    {
      "entropy": 5.943948698043823,
      "epoch": 0.18142235123367198,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999071312716321,
      "loss": 5.9474,
      "mean_token_accuracy": 0.14696409404277802,
      "num_tokens": 4139391.0,
      "step": 2000
    },
    {
      "entropy": 6.006927824020385,
      "epoch": 0.18187590711175616,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999061993931215,
      "loss": 5.8893,
      "mean_token_accuracy": 0.14896611645817756,
      "num_tokens": 4148891.0,
      "step": 2005
    },
    {
      "entropy": 6.163501930236817,
      "epoch": 0.18232946298984035,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999052628634934,
      "loss": 6.2293,
      "mean_token_accuracy": 0.13529689162969588,
      "num_tokens": 4159551.0,
      "step": 2010
    },
    {
      "entropy": 6.137354469299316,
      "epoch": 0.18278301886792453,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999043216827671,
      "loss": 6.0245,
      "mean_token_accuracy": 0.13915295749902726,
      "num_tokens": 4170055.0,
      "step": 2015
    },
    {
      "entropy": 5.894563341140747,
      "epoch": 0.1832365747460087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999033758509621,
      "loss": 5.8787,
      "mean_token_accuracy": 0.1426935002207756,
      "num_tokens": 4180767.0,
      "step": 2020
    },
    {
      "entropy": 5.976743268966675,
      "epoch": 0.1836901306240929,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499902425368098,
      "loss": 6.0046,
      "mean_token_accuracy": 0.14720579907298087,
      "num_tokens": 4190654.0,
      "step": 2025
    },
    {
      "entropy": 6.0072839736938475,
      "epoch": 0.18414368650217708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999014702341944,
      "loss": 5.9162,
      "mean_token_accuracy": 0.14252835661172866,
      "num_tokens": 4200753.0,
      "step": 2030
    },
    {
      "entropy": 5.982507371902466,
      "epoch": 0.18459724238026126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999005104492712,
      "loss": 6.0081,
      "mean_token_accuracy": 0.14058929979801177,
      "num_tokens": 4210994.0,
      "step": 2035
    },
    {
      "entropy": 6.036164665222168,
      "epoch": 0.18505079825834542,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998995460133481,
      "loss": 5.9517,
      "mean_token_accuracy": 0.14202935472130776,
      "num_tokens": 4220883.0,
      "step": 2040
    },
    {
      "entropy": 5.995604372024536,
      "epoch": 0.1855043541364296,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998985769264451,
      "loss": 5.9858,
      "mean_token_accuracy": 0.1352172464132309,
      "num_tokens": 4231269.0,
      "step": 2045
    },
    {
      "entropy": 5.994415187835694,
      "epoch": 0.18595791001451378,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998976031885821,
      "loss": 5.9821,
      "mean_token_accuracy": 0.13850782588124275,
      "num_tokens": 4241559.0,
      "step": 2050
    },
    {
      "entropy": 5.946982669830322,
      "epoch": 0.18641146589259797,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998966247997794,
      "loss": 5.8821,
      "mean_token_accuracy": 0.14087381660938264,
      "num_tokens": 4250870.0,
      "step": 2055
    },
    {
      "entropy": 5.9092240810394285,
      "epoch": 0.18686502177068215,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998956417600572,
      "loss": 5.8321,
      "mean_token_accuracy": 0.14701732099056244,
      "num_tokens": 4260848.0,
      "step": 2060
    },
    {
      "entropy": 5.90106840133667,
      "epoch": 0.18731857764876633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998946540694358,
      "loss": 5.9497,
      "mean_token_accuracy": 0.14662184566259384,
      "num_tokens": 4270973.0,
      "step": 2065
    },
    {
      "entropy": 6.0752662181854244,
      "epoch": 0.18777213352685052,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998936617279357,
      "loss": 6.0441,
      "mean_token_accuracy": 0.13860629945993425,
      "num_tokens": 4281211.0,
      "step": 2070
    },
    {
      "entropy": 5.913724660873413,
      "epoch": 0.1882256894049347,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998926647355773,
      "loss": 5.903,
      "mean_token_accuracy": 0.14046634137630462,
      "num_tokens": 4291425.0,
      "step": 2075
    },
    {
      "entropy": 6.071384477615356,
      "epoch": 0.18867924528301888,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998916630923813,
      "loss": 5.994,
      "mean_token_accuracy": 0.14140329733490944,
      "num_tokens": 4301079.0,
      "step": 2080
    },
    {
      "entropy": 5.993822479248047,
      "epoch": 0.18913280116110304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998906567983685,
      "loss": 5.9682,
      "mean_token_accuracy": 0.14236864894628526,
      "num_tokens": 4312374.0,
      "step": 2085
    },
    {
      "entropy": 5.950992774963379,
      "epoch": 0.18958635703918722,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998896458535595,
      "loss": 5.8985,
      "mean_token_accuracy": 0.14630423933267594,
      "num_tokens": 4322810.0,
      "step": 2090
    },
    {
      "entropy": 5.95527663230896,
      "epoch": 0.1900399129172714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998886302579752,
      "loss": 5.8899,
      "mean_token_accuracy": 0.14624738991260527,
      "num_tokens": 4332082.0,
      "step": 2095
    },
    {
      "entropy": 6.014179372787476,
      "epoch": 0.19049346879535559,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499887610011637,
      "loss": 5.976,
      "mean_token_accuracy": 0.1419885538518429,
      "num_tokens": 4342632.0,
      "step": 2100
    },
    {
      "entropy": 5.916087007522583,
      "epoch": 0.19094702467343977,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998865851145656,
      "loss": 5.9117,
      "mean_token_accuracy": 0.14088311344385146,
      "num_tokens": 4353544.0,
      "step": 2105
    },
    {
      "entropy": 5.980039072036743,
      "epoch": 0.19140058055152395,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998855555667822,
      "loss": 6.0379,
      "mean_token_accuracy": 0.1379381775856018,
      "num_tokens": 4363636.0,
      "step": 2110
    },
    {
      "entropy": 6.0455575466156,
      "epoch": 0.19185413642960814,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998845213683084,
      "loss": 5.9077,
      "mean_token_accuracy": 0.14140428155660628,
      "num_tokens": 4373672.0,
      "step": 2115
    },
    {
      "entropy": 5.835667896270752,
      "epoch": 0.19230769230769232,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998834825191652,
      "loss": 5.88,
      "mean_token_accuracy": 0.14386406689882278,
      "num_tokens": 4383199.0,
      "step": 2120
    },
    {
      "entropy": 6.009014415740967,
      "epoch": 0.1927612481857765,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998824390193743,
      "loss": 5.9413,
      "mean_token_accuracy": 0.14438435882329942,
      "num_tokens": 4394446.0,
      "step": 2125
    },
    {
      "entropy": 6.04556131362915,
      "epoch": 0.19321480406386066,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998813908689573,
      "loss": 5.9947,
      "mean_token_accuracy": 0.14130873903632163,
      "num_tokens": 4405882.0,
      "step": 2130
    },
    {
      "entropy": 5.920941781997681,
      "epoch": 0.19366835994194484,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998803380679359,
      "loss": 5.8482,
      "mean_token_accuracy": 0.14850495457649232,
      "num_tokens": 4415280.0,
      "step": 2135
    },
    {
      "entropy": 5.956070137023926,
      "epoch": 0.19412191582002902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998792806163317,
      "loss": 6.0573,
      "mean_token_accuracy": 0.13813181966543198,
      "num_tokens": 4425755.0,
      "step": 2140
    },
    {
      "entropy": 5.9665940284729,
      "epoch": 0.1945754716981132,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998782185141667,
      "loss": 5.9163,
      "mean_token_accuracy": 0.13927923887968063,
      "num_tokens": 4436259.0,
      "step": 2145
    },
    {
      "entropy": 5.970826911926269,
      "epoch": 0.1950290275761974,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998771517614629,
      "loss": 5.9388,
      "mean_token_accuracy": 0.13907137587666513,
      "num_tokens": 4446423.0,
      "step": 2150
    },
    {
      "entropy": 5.963461065292359,
      "epoch": 0.19548258345428157,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998760803582422,
      "loss": 5.8887,
      "mean_token_accuracy": 0.1487796500325203,
      "num_tokens": 4456678.0,
      "step": 2155
    },
    {
      "entropy": 5.92654185295105,
      "epoch": 0.19593613933236576,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998750043045269,
      "loss": 5.8878,
      "mean_token_accuracy": 0.14657580107450485,
      "num_tokens": 4467031.0,
      "step": 2160
    },
    {
      "entropy": 5.977866840362549,
      "epoch": 0.19638969521044994,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998739236003392,
      "loss": 5.9626,
      "mean_token_accuracy": 0.1362947404384613,
      "num_tokens": 4477771.0,
      "step": 2165
    },
    {
      "entropy": 5.975367212295533,
      "epoch": 0.1968432510885341,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998728382457014,
      "loss": 5.9837,
      "mean_token_accuracy": 0.1417712986469269,
      "num_tokens": 4488426.0,
      "step": 2170
    },
    {
      "entropy": 6.073901844024658,
      "epoch": 0.19729680696661828,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998717482406362,
      "loss": 5.9613,
      "mean_token_accuracy": 0.1399609178304672,
      "num_tokens": 4499735.0,
      "step": 2175
    },
    {
      "entropy": 5.915829515457153,
      "epoch": 0.19775036284470246,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998706535851657,
      "loss": 5.9059,
      "mean_token_accuracy": 0.14285302758216858,
      "num_tokens": 4509500.0,
      "step": 2180
    },
    {
      "entropy": 5.927563905715942,
      "epoch": 0.19820391872278664,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998695542793129,
      "loss": 5.9101,
      "mean_token_accuracy": 0.14308948516845704,
      "num_tokens": 4520431.0,
      "step": 2185
    },
    {
      "entropy": 5.991705322265625,
      "epoch": 0.19865747460087083,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998684503231003,
      "loss": 6.0086,
      "mean_token_accuracy": 0.14144350960850716,
      "num_tokens": 4532294.0,
      "step": 2190
    },
    {
      "entropy": 5.943409013748169,
      "epoch": 0.199111030478955,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499867341716551,
      "loss": 5.9129,
      "mean_token_accuracy": 0.14232037886977195,
      "num_tokens": 4542488.0,
      "step": 2195
    },
    {
      "entropy": 5.885735511779785,
      "epoch": 0.1995645863570392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998662284596878,
      "loss": 5.8553,
      "mean_token_accuracy": 0.14037457183003427,
      "num_tokens": 4552933.0,
      "step": 2200
    },
    {
      "entropy": 5.944642543792725,
      "epoch": 0.20001814223512338,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998651105525335,
      "loss": 5.846,
      "mean_token_accuracy": 0.14879003316164016,
      "num_tokens": 4562373.0,
      "step": 2205
    },
    {
      "entropy": 5.890647745132446,
      "epoch": 0.20047169811320756,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998639879951115,
      "loss": 5.9585,
      "mean_token_accuracy": 0.14374665394425393,
      "num_tokens": 4572715.0,
      "step": 2210
    },
    {
      "entropy": 6.039751386642456,
      "epoch": 0.2009252539912917,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499862860787445,
      "loss": 5.8561,
      "mean_token_accuracy": 0.14138285890221597,
      "num_tokens": 4582782.0,
      "step": 2215
    },
    {
      "entropy": 5.931440114974976,
      "epoch": 0.2013788098693759,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998617289295571,
      "loss": 5.9624,
      "mean_token_accuracy": 0.1358715571463108,
      "num_tokens": 4594108.0,
      "step": 2220
    },
    {
      "entropy": 5.978836011886597,
      "epoch": 0.20183236574746008,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998605924214714,
      "loss": 5.9241,
      "mean_token_accuracy": 0.14495790153741836,
      "num_tokens": 4604732.0,
      "step": 2225
    },
    {
      "entropy": 5.946112108230591,
      "epoch": 0.20228592162554426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998594512632114,
      "loss": 6.0523,
      "mean_token_accuracy": 0.13024092167615892,
      "num_tokens": 4614998.0,
      "step": 2230
    },
    {
      "entropy": 5.972310304641724,
      "epoch": 0.20273947750362845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998583054548006,
      "loss": 5.8754,
      "mean_token_accuracy": 0.1418835252523422,
      "num_tokens": 4624665.0,
      "step": 2235
    },
    {
      "entropy": 5.947936773300171,
      "epoch": 0.20319303338171263,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998571549962628,
      "loss": 5.8824,
      "mean_token_accuracy": 0.14991916120052337,
      "num_tokens": 4633731.0,
      "step": 2240
    },
    {
      "entropy": 5.932089471817017,
      "epoch": 0.2036465892597968,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998559998876216,
      "loss": 5.8978,
      "mean_token_accuracy": 0.1406530499458313,
      "num_tokens": 4643421.0,
      "step": 2245
    },
    {
      "entropy": 5.991897392272949,
      "epoch": 0.204100145137881,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998548401289012,
      "loss": 5.9815,
      "mean_token_accuracy": 0.14005577862262725,
      "num_tokens": 4653795.0,
      "step": 2250
    },
    {
      "entropy": 6.006911277770996,
      "epoch": 0.20455370101596518,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998536757201253,
      "loss": 5.9461,
      "mean_token_accuracy": 0.14425258636474608,
      "num_tokens": 4664840.0,
      "step": 2255
    },
    {
      "entropy": 5.921462535858154,
      "epoch": 0.20500725689404933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998525066613181,
      "loss": 5.9605,
      "mean_token_accuracy": 0.14682526290416717,
      "num_tokens": 4675508.0,
      "step": 2260
    },
    {
      "entropy": 5.9264942646026615,
      "epoch": 0.20546081277213352,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998513329525038,
      "loss": 5.8727,
      "mean_token_accuracy": 0.1501593768596649,
      "num_tokens": 4685995.0,
      "step": 2265
    },
    {
      "entropy": 5.995488023757934,
      "epoch": 0.2059143686502177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998501545937066,
      "loss": 5.9227,
      "mean_token_accuracy": 0.14225359484553338,
      "num_tokens": 4696659.0,
      "step": 2270
    },
    {
      "entropy": 5.867328739166259,
      "epoch": 0.20636792452830188,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499848971584951,
      "loss": 5.8615,
      "mean_token_accuracy": 0.15242719650268555,
      "num_tokens": 4706438.0,
      "step": 2275
    },
    {
      "entropy": 5.889807510375976,
      "epoch": 0.20682148040638607,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998477839262611,
      "loss": 5.8707,
      "mean_token_accuracy": 0.1458761513233185,
      "num_tokens": 4716402.0,
      "step": 2280
    },
    {
      "entropy": 5.955387878417969,
      "epoch": 0.20727503628447025,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499846591617662,
      "loss": 5.9585,
      "mean_token_accuracy": 0.14434442222118377,
      "num_tokens": 4728488.0,
      "step": 2285
    },
    {
      "entropy": 5.912747049331665,
      "epoch": 0.20772859216255443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998453946591779,
      "loss": 5.9712,
      "mean_token_accuracy": 0.14680054485797883,
      "num_tokens": 4738693.0,
      "step": 2290
    },
    {
      "entropy": 5.982892036437988,
      "epoch": 0.20818214804063861,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998441930508339,
      "loss": 5.867,
      "mean_token_accuracy": 0.14620437175035478,
      "num_tokens": 4748895.0,
      "step": 2295
    },
    {
      "entropy": 5.956034708023071,
      "epoch": 0.2086357039187228,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998429867926547,
      "loss": 5.9875,
      "mean_token_accuracy": 0.14012246131896972,
      "num_tokens": 4760135.0,
      "step": 2300
    },
    {
      "entropy": 6.089736557006836,
      "epoch": 0.20908925979680695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998417758846651,
      "loss": 6.0374,
      "mean_token_accuracy": 0.14575955420732498,
      "num_tokens": 4769935.0,
      "step": 2305
    },
    {
      "entropy": 6.0489821434021,
      "epoch": 0.20954281567489114,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998405603268903,
      "loss": 5.9447,
      "mean_token_accuracy": 0.14609623178839684,
      "num_tokens": 4779520.0,
      "step": 2310
    },
    {
      "entropy": 5.840287733078003,
      "epoch": 0.20999637155297532,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998393401193554,
      "loss": 5.8684,
      "mean_token_accuracy": 0.14978107139468194,
      "num_tokens": 4791256.0,
      "step": 2315
    },
    {
      "entropy": 5.842484188079834,
      "epoch": 0.2104499274310595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998381152620857,
      "loss": 5.8286,
      "mean_token_accuracy": 0.1479355052113533,
      "num_tokens": 4801334.0,
      "step": 2320
    },
    {
      "entropy": 5.891949033737182,
      "epoch": 0.21090348330914369,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998368857551067,
      "loss": 5.7918,
      "mean_token_accuracy": 0.15084001272916794,
      "num_tokens": 4811361.0,
      "step": 2325
    },
    {
      "entropy": 5.865246343612671,
      "epoch": 0.21135703918722787,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998356515984433,
      "loss": 5.8649,
      "mean_token_accuracy": 0.14442769065499306,
      "num_tokens": 4821190.0,
      "step": 2330
    },
    {
      "entropy": 5.967099905014038,
      "epoch": 0.21181059506531205,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998344127921213,
      "loss": 5.9012,
      "mean_token_accuracy": 0.14606177806854248,
      "num_tokens": 4830694.0,
      "step": 2335
    },
    {
      "entropy": 5.83759298324585,
      "epoch": 0.21226415094339623,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998331693361666,
      "loss": 5.8882,
      "mean_token_accuracy": 0.14482705742120744,
      "num_tokens": 4841344.0,
      "step": 2340
    },
    {
      "entropy": 5.916370058059693,
      "epoch": 0.21271770682148042,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998319212306046,
      "loss": 5.8601,
      "mean_token_accuracy": 0.14071092009544373,
      "num_tokens": 4851161.0,
      "step": 2345
    },
    {
      "entropy": 5.892922449111938,
      "epoch": 0.21317126269956457,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499830668475461,
      "loss": 5.8446,
      "mean_token_accuracy": 0.14808950796723366,
      "num_tokens": 4860740.0,
      "step": 2350
    },
    {
      "entropy": 5.941983938217163,
      "epoch": 0.21362481857764876,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499829411070762,
      "loss": 5.8718,
      "mean_token_accuracy": 0.1465384006500244,
      "num_tokens": 4871733.0,
      "step": 2355
    },
    {
      "entropy": 5.912432241439819,
      "epoch": 0.21407837445573294,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998281490165335,
      "loss": 5.9292,
      "mean_token_accuracy": 0.14041552618145942,
      "num_tokens": 4883617.0,
      "step": 2360
    },
    {
      "entropy": 5.977138614654541,
      "epoch": 0.21453193033381712,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998268823128016,
      "loss": 5.8841,
      "mean_token_accuracy": 0.14232782125473023,
      "num_tokens": 4894519.0,
      "step": 2365
    },
    {
      "entropy": 5.894336938858032,
      "epoch": 0.2149854862119013,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998256109595923,
      "loss": 5.8185,
      "mean_token_accuracy": 0.14750942587852478,
      "num_tokens": 4904965.0,
      "step": 2370
    },
    {
      "entropy": 5.793227291107177,
      "epoch": 0.2154390420899855,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998243349569323,
      "loss": 5.749,
      "mean_token_accuracy": 0.15446996241807937,
      "num_tokens": 4914618.0,
      "step": 2375
    },
    {
      "entropy": 5.894077157974243,
      "epoch": 0.21589259796806967,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998230543048477,
      "loss": 5.8212,
      "mean_token_accuracy": 0.15240628868341446,
      "num_tokens": 4924341.0,
      "step": 2380
    },
    {
      "entropy": 5.813013172149658,
      "epoch": 0.21634615384615385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499821769003365,
      "loss": 5.7744,
      "mean_token_accuracy": 0.14865444898605346,
      "num_tokens": 4934231.0,
      "step": 2385
    },
    {
      "entropy": 5.808397579193115,
      "epoch": 0.21679970972423804,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998204790525109,
      "loss": 5.7992,
      "mean_token_accuracy": 0.14798300564289094,
      "num_tokens": 4944303.0,
      "step": 2390
    },
    {
      "entropy": 5.8492330551147464,
      "epoch": 0.2172532656023222,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998191844523119,
      "loss": 5.8847,
      "mean_token_accuracy": 0.14911072105169296,
      "num_tokens": 4954571.0,
      "step": 2395
    },
    {
      "entropy": 5.985284900665283,
      "epoch": 0.21770682148040638,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499817885202795,
      "loss": 5.8453,
      "mean_token_accuracy": 0.14537417143583298,
      "num_tokens": 4964418.0,
      "step": 2400
    },
    {
      "entropy": 5.873136949539185,
      "epoch": 0.21816037735849056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998165813039869,
      "loss": 5.8541,
      "mean_token_accuracy": 0.14498028308153152,
      "num_tokens": 4975036.0,
      "step": 2405
    },
    {
      "entropy": 5.87366943359375,
      "epoch": 0.21861393323657474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998152727559145,
      "loss": 5.9677,
      "mean_token_accuracy": 0.14030904546380044,
      "num_tokens": 4985272.0,
      "step": 2410
    },
    {
      "entropy": 6.023105239868164,
      "epoch": 0.21906748911465893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499813959558605,
      "loss": 5.8715,
      "mean_token_accuracy": 0.14627939462661743,
      "num_tokens": 4995187.0,
      "step": 2415
    },
    {
      "entropy": 5.9488426685333256,
      "epoch": 0.2195210449927431,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998126417120856,
      "loss": 5.9311,
      "mean_token_accuracy": 0.1463703356683254,
      "num_tokens": 5005013.0,
      "step": 2420
    },
    {
      "entropy": 5.997941732406616,
      "epoch": 0.2199746008708273,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998113192163835,
      "loss": 5.9439,
      "mean_token_accuracy": 0.14230974465608598,
      "num_tokens": 5016789.0,
      "step": 2425
    },
    {
      "entropy": 5.901341581344605,
      "epoch": 0.22042815674891147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499809992071526,
      "loss": 5.9122,
      "mean_token_accuracy": 0.14620402008295058,
      "num_tokens": 5026913.0,
      "step": 2430
    },
    {
      "entropy": 5.943270683288574,
      "epoch": 0.22088171262699566,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998086602775406,
      "loss": 5.9933,
      "mean_token_accuracy": 0.14236516878008842,
      "num_tokens": 5037501.0,
      "step": 2435
    },
    {
      "entropy": 5.9212024211883545,
      "epoch": 0.2213352685050798,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998073238344547,
      "loss": 5.8343,
      "mean_token_accuracy": 0.14206503182649613,
      "num_tokens": 5048478.0,
      "step": 2440
    },
    {
      "entropy": 5.894640779495239,
      "epoch": 0.221788824383164,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998059827422962,
      "loss": 5.8923,
      "mean_token_accuracy": 0.1398580104112625,
      "num_tokens": 5059118.0,
      "step": 2445
    },
    {
      "entropy": 5.949969148635864,
      "epoch": 0.22224238026124818,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998046370010926,
      "loss": 5.8746,
      "mean_token_accuracy": 0.15106075257062912,
      "num_tokens": 5070180.0,
      "step": 2450
    },
    {
      "entropy": 5.862327241897583,
      "epoch": 0.22269593613933236,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998032866108718,
      "loss": 5.8561,
      "mean_token_accuracy": 0.14829719066619873,
      "num_tokens": 5079912.0,
      "step": 2455
    },
    {
      "entropy": 6.009226226806641,
      "epoch": 0.22314949201741655,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998019315716618,
      "loss": 5.9424,
      "mean_token_accuracy": 0.14838250279426574,
      "num_tokens": 5090549.0,
      "step": 2460
    },
    {
      "entropy": 5.816465187072754,
      "epoch": 0.22360304789550073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998005718834905,
      "loss": 5.7914,
      "mean_token_accuracy": 0.15030866116285324,
      "num_tokens": 5100880.0,
      "step": 2465
    },
    {
      "entropy": 5.939529895782471,
      "epoch": 0.2240566037735849,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997992075463862,
      "loss": 5.908,
      "mean_token_accuracy": 0.1464825004339218,
      "num_tokens": 5111796.0,
      "step": 2470
    },
    {
      "entropy": 5.948081302642822,
      "epoch": 0.2245101596516691,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997978385603769,
      "loss": 5.8941,
      "mean_token_accuracy": 0.14777286797761918,
      "num_tokens": 5123468.0,
      "step": 2475
    },
    {
      "entropy": 5.865221118927002,
      "epoch": 0.22496371552975328,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997964649254911,
      "loss": 5.8243,
      "mean_token_accuracy": 0.15206352174282073,
      "num_tokens": 5132925.0,
      "step": 2480
    },
    {
      "entropy": 5.941302299499512,
      "epoch": 0.22541727140783743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997950866417572,
      "loss": 5.9058,
      "mean_token_accuracy": 0.1450276717543602,
      "num_tokens": 5143445.0,
      "step": 2485
    },
    {
      "entropy": 5.9101451396942135,
      "epoch": 0.22587082728592162,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004997937037092034,
      "loss": 5.8269,
      "mean_token_accuracy": 0.14438713937997819,
      "num_tokens": 5153176.0,
      "step": 2490
    },
    {
      "entropy": 5.911853361129761,
      "epoch": 0.2263243831640058,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997923161278587,
      "loss": 5.8592,
      "mean_token_accuracy": 0.1432860165834427,
      "num_tokens": 5162765.0,
      "step": 2495
    },
    {
      "entropy": 5.9264219284057615,
      "epoch": 0.22677793904208998,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997909238977514,
      "loss": 5.9191,
      "mean_token_accuracy": 0.14676568061113357,
      "num_tokens": 5173390.0,
      "step": 2500
    },
    {
      "entropy": 5.89142746925354,
      "epoch": 0.22723149492017417,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997895270189108,
      "loss": 5.8215,
      "mean_token_accuracy": 0.14164483100175856,
      "num_tokens": 5184212.0,
      "step": 2505
    },
    {
      "entropy": 5.806804370880127,
      "epoch": 0.22768505079825835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997881254913653,
      "loss": 5.8189,
      "mean_token_accuracy": 0.14264959916472436,
      "num_tokens": 5194311.0,
      "step": 2510
    },
    {
      "entropy": 5.885697603225708,
      "epoch": 0.22813860667634253,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997867193151441,
      "loss": 5.8002,
      "mean_token_accuracy": 0.1477848395705223,
      "num_tokens": 5204708.0,
      "step": 2515
    },
    {
      "entropy": 5.83394422531128,
      "epoch": 0.22859216255442671,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997853084902763,
      "loss": 5.7834,
      "mean_token_accuracy": 0.14696393460035323,
      "num_tokens": 5214566.0,
      "step": 2520
    },
    {
      "entropy": 5.841207504272461,
      "epoch": 0.2290457184325109,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997838930167911,
      "loss": 5.8174,
      "mean_token_accuracy": 0.15246393531560898,
      "num_tokens": 5224762.0,
      "step": 2525
    },
    {
      "entropy": 5.891236639022827,
      "epoch": 0.22949927431059505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997824728947176,
      "loss": 5.8883,
      "mean_token_accuracy": 0.1477118618786335,
      "num_tokens": 5235146.0,
      "step": 2530
    },
    {
      "entropy": 5.933800506591797,
      "epoch": 0.22995283018867924,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997810481240853,
      "loss": 5.8496,
      "mean_token_accuracy": 0.14163196831941605,
      "num_tokens": 5243867.0,
      "step": 2535
    },
    {
      "entropy": 5.862011909484863,
      "epoch": 0.23040638606676342,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997796187049238,
      "loss": 5.7643,
      "mean_token_accuracy": 0.1487149715423584,
      "num_tokens": 5254279.0,
      "step": 2540
    },
    {
      "entropy": 5.967965698242187,
      "epoch": 0.2308599419448476,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997781846372623,
      "loss": 5.8745,
      "mean_token_accuracy": 0.1466490976512432,
      "num_tokens": 5264769.0,
      "step": 2545
    },
    {
      "entropy": 5.881758451461792,
      "epoch": 0.23131349782293179,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997767459211307,
      "loss": 5.8353,
      "mean_token_accuracy": 0.14295921698212624,
      "num_tokens": 5274291.0,
      "step": 2550
    },
    {
      "entropy": 5.835591793060303,
      "epoch": 0.23176705370101597,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997753025565587,
      "loss": 5.7787,
      "mean_token_accuracy": 0.150295127928257,
      "num_tokens": 5284366.0,
      "step": 2555
    },
    {
      "entropy": 5.887094306945801,
      "epoch": 0.23222060957910015,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997738545435763,
      "loss": 5.8779,
      "mean_token_accuracy": 0.14753376841545104,
      "num_tokens": 5294494.0,
      "step": 2560
    },
    {
      "entropy": 5.842414712905883,
      "epoch": 0.23267416545718433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997724018822133,
      "loss": 5.8907,
      "mean_token_accuracy": 0.1439896434545517,
      "num_tokens": 5305606.0,
      "step": 2565
    },
    {
      "entropy": 5.894366121292114,
      "epoch": 0.23312772133526852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997709445724996,
      "loss": 5.8728,
      "mean_token_accuracy": 0.14524707421660424,
      "num_tokens": 5315677.0,
      "step": 2570
    },
    {
      "entropy": 5.966664171218872,
      "epoch": 0.23358127721335267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997694826144656,
      "loss": 5.9577,
      "mean_token_accuracy": 0.14343089759349822,
      "num_tokens": 5326453.0,
      "step": 2575
    },
    {
      "entropy": 5.89670033454895,
      "epoch": 0.23403483309143686,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997680160081414,
      "loss": 5.8366,
      "mean_token_accuracy": 0.1491502456367016,
      "num_tokens": 5337591.0,
      "step": 2580
    },
    {
      "entropy": 5.8483922481536865,
      "epoch": 0.23448838896952104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997665447535573,
      "loss": 5.8308,
      "mean_token_accuracy": 0.15065386444330214,
      "num_tokens": 5349400.0,
      "step": 2585
    },
    {
      "entropy": 5.93424882888794,
      "epoch": 0.23494194484760522,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004997650688507438,
      "loss": 5.8278,
      "mean_token_accuracy": 0.14633653834462165,
      "num_tokens": 5359984.0,
      "step": 2590
    },
    {
      "entropy": 5.778865098953247,
      "epoch": 0.2353955007256894,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004997635882997314,
      "loss": 5.8775,
      "mean_token_accuracy": 0.15031310468912124,
      "num_tokens": 5369254.0,
      "step": 2595
    },
    {
      "entropy": 5.970759725570678,
      "epoch": 0.2358490566037736,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997621031005507,
      "loss": 5.8118,
      "mean_token_accuracy": 0.14642709940671922,
      "num_tokens": 5379648.0,
      "step": 2600
    },
    {
      "entropy": 5.777373361587524,
      "epoch": 0.23630261248185777,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997606132532325,
      "loss": 5.78,
      "mean_token_accuracy": 0.15143134370446204,
      "num_tokens": 5390321.0,
      "step": 2605
    },
    {
      "entropy": 5.845559740066529,
      "epoch": 0.23675616835994195,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004997591187578075,
      "loss": 5.8326,
      "mean_token_accuracy": 0.15119464695453644,
      "num_tokens": 5401078.0,
      "step": 2610
    },
    {
      "entropy": 5.831754875183106,
      "epoch": 0.23720972423802614,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997576196143067,
      "loss": 5.791,
      "mean_token_accuracy": 0.15142232924699783,
      "num_tokens": 5411977.0,
      "step": 2615
    },
    {
      "entropy": 5.869622707366943,
      "epoch": 0.2376632801161103,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499756115822761,
      "loss": 5.7833,
      "mean_token_accuracy": 0.14541143029928208,
      "num_tokens": 5422967.0,
      "step": 2620
    },
    {
      "entropy": 5.9165294647216795,
      "epoch": 0.23811683599419448,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997546073832015,
      "loss": 5.8272,
      "mean_token_accuracy": 0.15650950670242308,
      "num_tokens": 5432260.0,
      "step": 2625
    },
    {
      "entropy": 5.808580589294434,
      "epoch": 0.23857039187227866,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997530942956595,
      "loss": 5.802,
      "mean_token_accuracy": 0.1496383234858513,
      "num_tokens": 5441706.0,
      "step": 2630
    },
    {
      "entropy": 5.923821020126343,
      "epoch": 0.23902394775036284,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997515765601662,
      "loss": 5.7981,
      "mean_token_accuracy": 0.1539872854948044,
      "num_tokens": 5452397.0,
      "step": 2635
    },
    {
      "entropy": 5.806406927108765,
      "epoch": 0.23947750362844702,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997500541767531,
      "loss": 5.7578,
      "mean_token_accuracy": 0.15279131680727004,
      "num_tokens": 5463762.0,
      "step": 2640
    },
    {
      "entropy": 5.8301534175872805,
      "epoch": 0.2399310595065312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997485271454516,
      "loss": 5.8275,
      "mean_token_accuracy": 0.14339211732149124,
      "num_tokens": 5474622.0,
      "step": 2645
    },
    {
      "entropy": 5.869717407226562,
      "epoch": 0.2403846153846154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997469954662933,
      "loss": 5.812,
      "mean_token_accuracy": 0.14975824058055878,
      "num_tokens": 5484697.0,
      "step": 2650
    },
    {
      "entropy": 5.789428615570069,
      "epoch": 0.24083817126269957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997454591393097,
      "loss": 5.8108,
      "mean_token_accuracy": 0.14811063259840013,
      "num_tokens": 5496433.0,
      "step": 2655
    },
    {
      "entropy": 5.954967594146728,
      "epoch": 0.24129172714078376,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997439181645329,
      "loss": 5.9288,
      "mean_token_accuracy": 0.14460990205407143,
      "num_tokens": 5507917.0,
      "step": 2660
    },
    {
      "entropy": 5.975463628768921,
      "epoch": 0.2417452830188679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997423725419945,
      "loss": 5.9088,
      "mean_token_accuracy": 0.14406671226024628,
      "num_tokens": 5519159.0,
      "step": 2665
    },
    {
      "entropy": 5.845655298233032,
      "epoch": 0.2421988388969521,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997408222717267,
      "loss": 5.8678,
      "mean_token_accuracy": 0.15028798580169678,
      "num_tokens": 5529180.0,
      "step": 2670
    },
    {
      "entropy": 5.850816488265991,
      "epoch": 0.24265239477503628,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997392673537612,
      "loss": 5.7844,
      "mean_token_accuracy": 0.1593550056219101,
      "num_tokens": 5539363.0,
      "step": 2675
    },
    {
      "entropy": 5.830254602432251,
      "epoch": 0.24310595065312046,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997377077881306,
      "loss": 5.894,
      "mean_token_accuracy": 0.14436824321746827,
      "num_tokens": 5549841.0,
      "step": 2680
    },
    {
      "entropy": 5.904850482940674,
      "epoch": 0.24355950653120464,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997361435748668,
      "loss": 5.7754,
      "mean_token_accuracy": 0.15043861716985701,
      "num_tokens": 5560299.0,
      "step": 2685
    },
    {
      "entropy": 5.894953012466431,
      "epoch": 0.24401306240928883,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997345747140024,
      "loss": 5.8383,
      "mean_token_accuracy": 0.14846533983945848,
      "num_tokens": 5571081.0,
      "step": 2690
    },
    {
      "entropy": 5.8214860439300535,
      "epoch": 0.244466618287373,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997330012055695,
      "loss": 5.7963,
      "mean_token_accuracy": 0.1535398468375206,
      "num_tokens": 5581389.0,
      "step": 2695
    },
    {
      "entropy": 5.8494048595428465,
      "epoch": 0.2449201741654572,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997314230496011,
      "loss": 5.8218,
      "mean_token_accuracy": 0.14826998114585876,
      "num_tokens": 5592347.0,
      "step": 2700
    },
    {
      "entropy": 5.818295621871949,
      "epoch": 0.24537373004354138,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997298402461296,
      "loss": 5.7857,
      "mean_token_accuracy": 0.1517721101641655,
      "num_tokens": 5603411.0,
      "step": 2705
    },
    {
      "entropy": 5.848494911193848,
      "epoch": 0.24582728592162553,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997282527951876,
      "loss": 5.787,
      "mean_token_accuracy": 0.15339138209819794,
      "num_tokens": 5614615.0,
      "step": 2710
    },
    {
      "entropy": 5.791625308990478,
      "epoch": 0.24628084179970972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997266606968081,
      "loss": 5.7528,
      "mean_token_accuracy": 0.14949269592761993,
      "num_tokens": 5624582.0,
      "step": 2715
    },
    {
      "entropy": 5.821265411376953,
      "epoch": 0.2467343976777939,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997250639510241,
      "loss": 5.7641,
      "mean_token_accuracy": 0.14658788964152336,
      "num_tokens": 5634310.0,
      "step": 2720
    },
    {
      "entropy": 5.86549563407898,
      "epoch": 0.24718795355587808,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997234625578684,
      "loss": 5.8658,
      "mean_token_accuracy": 0.14597384184598922,
      "num_tokens": 5644018.0,
      "step": 2725
    },
    {
      "entropy": 5.881694889068603,
      "epoch": 0.24764150943396226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997218565173742,
      "loss": 5.7995,
      "mean_token_accuracy": 0.14841866344213486,
      "num_tokens": 5653592.0,
      "step": 2730
    },
    {
      "entropy": 5.89327039718628,
      "epoch": 0.24809506531204645,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004997202458295748,
      "loss": 5.8404,
      "mean_token_accuracy": 0.1519847184419632,
      "num_tokens": 5662812.0,
      "step": 2735
    },
    {
      "entropy": 5.868380165100097,
      "epoch": 0.24854862119013063,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997186304945034,
      "loss": 5.8389,
      "mean_token_accuracy": 0.1472091943025589,
      "num_tokens": 5672939.0,
      "step": 2740
    },
    {
      "entropy": 5.809984922409058,
      "epoch": 0.24900217706821481,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997170105121935,
      "loss": 5.8024,
      "mean_token_accuracy": 0.15122157633304595,
      "num_tokens": 5683350.0,
      "step": 2745
    },
    {
      "entropy": 5.945703458786011,
      "epoch": 0.249455732946299,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997153858826786,
      "loss": 5.8884,
      "mean_token_accuracy": 0.14479928761720656,
      "num_tokens": 5694516.0,
      "step": 2750
    },
    {
      "entropy": 5.877769088745117,
      "epoch": 0.24990928882438315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997137566059922,
      "loss": 5.7253,
      "mean_token_accuracy": 0.1561413288116455,
      "num_tokens": 5705443.0,
      "step": 2755
    },
    {
      "entropy": 5.7393638610839846,
      "epoch": 0.25036284470246734,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997121226821681,
      "loss": 5.8265,
      "mean_token_accuracy": 0.1438466988503933,
      "num_tokens": 5715933.0,
      "step": 2760
    },
    {
      "entropy": 5.926178312301635,
      "epoch": 0.25081640058055155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049971048411124,
      "loss": 5.859,
      "mean_token_accuracy": 0.14209090322256088,
      "num_tokens": 5726630.0,
      "step": 2765
    },
    {
      "entropy": 5.964452457427979,
      "epoch": 0.2512699564586357,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997088408932418,
      "loss": 5.8065,
      "mean_token_accuracy": 0.1468511700630188,
      "num_tokens": 5736890.0,
      "step": 2770
    },
    {
      "entropy": 5.693834829330444,
      "epoch": 0.25172351233671986,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997071930282076,
      "loss": 5.7317,
      "mean_token_accuracy": 0.1594724640250206,
      "num_tokens": 5747042.0,
      "step": 2775
    },
    {
      "entropy": 5.852517175674438,
      "epoch": 0.25217706821480407,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997055405161712,
      "loss": 5.8433,
      "mean_token_accuracy": 0.14911026507616043,
      "num_tokens": 5757367.0,
      "step": 2780
    },
    {
      "entropy": 5.884478521347046,
      "epoch": 0.2526306240928882,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997038833571671,
      "loss": 5.9121,
      "mean_token_accuracy": 0.14246488213539124,
      "num_tokens": 5769557.0,
      "step": 2785
    },
    {
      "entropy": 5.899376440048218,
      "epoch": 0.25308417997097243,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997022215512295,
      "loss": 5.7988,
      "mean_token_accuracy": 0.1549243524670601,
      "num_tokens": 5780615.0,
      "step": 2790
    },
    {
      "entropy": 5.844459962844849,
      "epoch": 0.2535377358490566,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997005550983925,
      "loss": 5.7778,
      "mean_token_accuracy": 0.14659300297498704,
      "num_tokens": 5791271.0,
      "step": 2795
    },
    {
      "entropy": 5.723701143264771,
      "epoch": 0.2539912917271408,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000499698883998691,
      "loss": 5.7122,
      "mean_token_accuracy": 0.15313654989004136,
      "num_tokens": 5801117.0,
      "step": 2800
    },
    {
      "entropy": 5.828806066513062,
      "epoch": 0.25444484760522496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996972082521591,
      "loss": 5.7524,
      "mean_token_accuracy": 0.14977902621030809,
      "num_tokens": 5811696.0,
      "step": 2805
    },
    {
      "entropy": 5.825814199447632,
      "epoch": 0.25489840348330917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996955278588319,
      "loss": 5.768,
      "mean_token_accuracy": 0.15769439488649367,
      "num_tokens": 5822207.0,
      "step": 2810
    },
    {
      "entropy": 5.754099035263062,
      "epoch": 0.2553519593613933,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996938428187437,
      "loss": 5.7277,
      "mean_token_accuracy": 0.1581838384270668,
      "num_tokens": 5831927.0,
      "step": 2815
    },
    {
      "entropy": 5.834832811355591,
      "epoch": 0.2558055152394775,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996921531319297,
      "loss": 5.7944,
      "mean_token_accuracy": 0.1462201587855816,
      "num_tokens": 5842093.0,
      "step": 2820
    },
    {
      "entropy": 5.81879677772522,
      "epoch": 0.2562590711175617,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004996904587984246,
      "loss": 5.7148,
      "mean_token_accuracy": 0.15955326408147813,
      "num_tokens": 5850551.0,
      "step": 2825
    },
    {
      "entropy": 5.7429585456848145,
      "epoch": 0.25671262699564584,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996887598182636,
      "loss": 5.7179,
      "mean_token_accuracy": 0.15299306213855743,
      "num_tokens": 5860651.0,
      "step": 2830
    },
    {
      "entropy": 5.741921520233154,
      "epoch": 0.25716618287373005,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996870561914819,
      "loss": 5.7073,
      "mean_token_accuracy": 0.15380395650863649,
      "num_tokens": 5871500.0,
      "step": 2835
    },
    {
      "entropy": 5.753065347671509,
      "epoch": 0.2576197387518142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996853479181145,
      "loss": 5.7024,
      "mean_token_accuracy": 0.14877581000328063,
      "num_tokens": 5881523.0,
      "step": 2840
    },
    {
      "entropy": 5.855972957611084,
      "epoch": 0.2580732946298984,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996836349981969,
      "loss": 5.8495,
      "mean_token_accuracy": 0.14823734387755394,
      "num_tokens": 5891162.0,
      "step": 2845
    },
    {
      "entropy": 5.766469144821167,
      "epoch": 0.2585268505079826,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996819174317645,
      "loss": 5.7406,
      "mean_token_accuracy": 0.15299083068966865,
      "num_tokens": 5901818.0,
      "step": 2850
    },
    {
      "entropy": 5.832020998001099,
      "epoch": 0.2589804063860668,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996801952188528,
      "loss": 5.841,
      "mean_token_accuracy": 0.1450772315263748,
      "num_tokens": 5913889.0,
      "step": 2855
    },
    {
      "entropy": 5.846477127075195,
      "epoch": 0.25943396226415094,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996784683594973,
      "loss": 5.7912,
      "mean_token_accuracy": 0.15060947686433793,
      "num_tokens": 5923429.0,
      "step": 2860
    },
    {
      "entropy": 5.9096756935119625,
      "epoch": 0.2598875181422351,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996767368537338,
      "loss": 5.8223,
      "mean_token_accuracy": 0.1533066600561142,
      "num_tokens": 5933847.0,
      "step": 2865
    },
    {
      "entropy": 5.770804500579834,
      "epoch": 0.2603410740203193,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996750007015983,
      "loss": 5.7053,
      "mean_token_accuracy": 0.15112933814525603,
      "num_tokens": 5943874.0,
      "step": 2870
    },
    {
      "entropy": 5.81883487701416,
      "epoch": 0.26079462989840346,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996732599031264,
      "loss": 5.7723,
      "mean_token_accuracy": 0.1524694785475731,
      "num_tokens": 5953653.0,
      "step": 2875
    },
    {
      "entropy": 5.84723162651062,
      "epoch": 0.2612481857764877,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996715144583543,
      "loss": 5.724,
      "mean_token_accuracy": 0.1565740928053856,
      "num_tokens": 5964039.0,
      "step": 2880
    },
    {
      "entropy": 5.858557844161988,
      "epoch": 0.26170174165457183,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499669764367318,
      "loss": 5.8498,
      "mean_token_accuracy": 0.14639762043952942,
      "num_tokens": 5974083.0,
      "step": 2885
    },
    {
      "entropy": 5.799863529205322,
      "epoch": 0.26215529753265604,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996680096300537,
      "loss": 5.7846,
      "mean_token_accuracy": 0.14885008633136748,
      "num_tokens": 5984629.0,
      "step": 2890
    },
    {
      "entropy": 5.76010479927063,
      "epoch": 0.2626088534107402,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996662502465978,
      "loss": 5.6993,
      "mean_token_accuracy": 0.16074687242507935,
      "num_tokens": 5994295.0,
      "step": 2895
    },
    {
      "entropy": 5.767422151565552,
      "epoch": 0.2630624092888244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996644862169864,
      "loss": 5.7305,
      "mean_token_accuracy": 0.1524806246161461,
      "num_tokens": 6004156.0,
      "step": 2900
    },
    {
      "entropy": 5.783775281906128,
      "epoch": 0.26351596516690856,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996627175412563,
      "loss": 5.827,
      "mean_token_accuracy": 0.15107300579547883,
      "num_tokens": 6014188.0,
      "step": 2905
    },
    {
      "entropy": 5.905015087127685,
      "epoch": 0.2639695210449927,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499660944219444,
      "loss": 5.8832,
      "mean_token_accuracy": 0.1467879518866539,
      "num_tokens": 6024879.0,
      "step": 2910
    },
    {
      "entropy": 5.935989856719971,
      "epoch": 0.2644230769230769,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996591662515861,
      "loss": 5.936,
      "mean_token_accuracy": 0.14425073117017745,
      "num_tokens": 6036374.0,
      "step": 2915
    },
    {
      "entropy": 5.863274049758911,
      "epoch": 0.2648766328011611,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996573836377193,
      "loss": 5.7858,
      "mean_token_accuracy": 0.14999294206500052,
      "num_tokens": 6046432.0,
      "step": 2920
    },
    {
      "entropy": 5.885637998580933,
      "epoch": 0.2653301886792453,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996555963778806,
      "loss": 5.8023,
      "mean_token_accuracy": 0.1489254727959633,
      "num_tokens": 6056864.0,
      "step": 2925
    },
    {
      "entropy": 5.9313414096832275,
      "epoch": 0.26578374455732945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499653804472107,
      "loss": 5.8136,
      "mean_token_accuracy": 0.15608129501342774,
      "num_tokens": 6067294.0,
      "step": 2930
    },
    {
      "entropy": 5.811369609832764,
      "epoch": 0.26623730043541366,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996520079204353,
      "loss": 5.8308,
      "mean_token_accuracy": 0.14605820178985596,
      "num_tokens": 6077587.0,
      "step": 2935
    },
    {
      "entropy": 5.876870012283325,
      "epoch": 0.2666908563134978,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499650206722903,
      "loss": 5.7673,
      "mean_token_accuracy": 0.15060035437345504,
      "num_tokens": 6089306.0,
      "step": 2940
    },
    {
      "entropy": 5.725161027908325,
      "epoch": 0.267144412191582,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996484008795471,
      "loss": 5.7388,
      "mean_token_accuracy": 0.14864355698227882,
      "num_tokens": 6099651.0,
      "step": 2945
    },
    {
      "entropy": 5.796488952636719,
      "epoch": 0.2675979680696662,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499646590390405,
      "loss": 5.718,
      "mean_token_accuracy": 0.14930034428834915,
      "num_tokens": 6110063.0,
      "step": 2950
    },
    {
      "entropy": 5.782108545303345,
      "epoch": 0.26805152394775034,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996447752555142,
      "loss": 5.7581,
      "mean_token_accuracy": 0.15502049624919892,
      "num_tokens": 6120031.0,
      "step": 2955
    },
    {
      "entropy": 5.821945476531982,
      "epoch": 0.26850507982583455,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996429554749122,
      "loss": 5.8167,
      "mean_token_accuracy": 0.15453848391771316,
      "num_tokens": 6131274.0,
      "step": 2960
    },
    {
      "entropy": 5.907453536987305,
      "epoch": 0.2689586357039187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996411310486367,
      "loss": 5.8431,
      "mean_token_accuracy": 0.1503222480416298,
      "num_tokens": 6140878.0,
      "step": 2965
    },
    {
      "entropy": 5.831985235214233,
      "epoch": 0.2694121915820029,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996393019767253,
      "loss": 5.8372,
      "mean_token_accuracy": 0.1460697442293167,
      "num_tokens": 6152229.0,
      "step": 2970
    },
    {
      "entropy": 5.816186428070068,
      "epoch": 0.26986574746008707,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996374682592158,
      "loss": 5.731,
      "mean_token_accuracy": 0.15502249598503112,
      "num_tokens": 6162311.0,
      "step": 2975
    },
    {
      "entropy": 5.855098247528076,
      "epoch": 0.2703193033381713,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996356298961463,
      "loss": 5.8514,
      "mean_token_accuracy": 0.14927736073732376,
      "num_tokens": 6172806.0,
      "step": 2980
    },
    {
      "entropy": 5.843709278106689,
      "epoch": 0.27077285921625543,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996337868875549,
      "loss": 5.803,
      "mean_token_accuracy": 0.15235266238451003,
      "num_tokens": 6183479.0,
      "step": 2985
    },
    {
      "entropy": 5.852834177017212,
      "epoch": 0.27122641509433965,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996319392334792,
      "loss": 5.7794,
      "mean_token_accuracy": 0.14965162798762321,
      "num_tokens": 6194201.0,
      "step": 2990
    },
    {
      "entropy": 5.759162425994873,
      "epoch": 0.2716799709724238,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499630086933958,
      "loss": 5.7323,
      "mean_token_accuracy": 0.14620673805475234,
      "num_tokens": 6204774.0,
      "step": 2995
    },
    {
      "entropy": 5.8088103294372555,
      "epoch": 0.27213352685050796,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996282299890292,
      "loss": 5.7233,
      "mean_token_accuracy": 0.14723710864782333,
      "num_tokens": 6215056.0,
      "step": 3000
    },
    {
      "epoch": 0.27213352685050796,
      "eval_entropy": 5.67995113377118,
      "eval_loss": 5.81314754486084,
      "eval_mean_token_accuracy": 0.1565627772883592,
      "eval_num_tokens": 6215056.0,
      "eval_runtime": 26.5641,
      "eval_samples_per_second": 1331.12,
      "eval_steps_per_second": 166.39,
      "step": 3000
    },
    {
      "entropy": 5.872125720977783,
      "epoch": 0.27258708272859217,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996263683987316,
      "loss": 5.8407,
      "mean_token_accuracy": 0.1507551245391369,
      "num_tokens": 6226264.0,
      "step": 3005
    },
    {
      "entropy": 5.812562322616577,
      "epoch": 0.2730406386066763,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996245021631033,
      "loss": 5.7505,
      "mean_token_accuracy": 0.15249016731977463,
      "num_tokens": 6235137.0,
      "step": 3010
    },
    {
      "entropy": 5.790219068527222,
      "epoch": 0.27349419448476053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499622631282183,
      "loss": 5.7793,
      "mean_token_accuracy": 0.15375427156686783,
      "num_tokens": 6245411.0,
      "step": 3015
    },
    {
      "entropy": 5.873235607147217,
      "epoch": 0.2739477503628447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996207557560095,
      "loss": 5.8397,
      "mean_token_accuracy": 0.14681091234087945,
      "num_tokens": 6255633.0,
      "step": 3020
    },
    {
      "entropy": 5.818582916259766,
      "epoch": 0.2744013062409289,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996188755846216,
      "loss": 5.7906,
      "mean_token_accuracy": 0.1482468992471695,
      "num_tokens": 6265109.0,
      "step": 3025
    },
    {
      "entropy": 5.841054964065552,
      "epoch": 0.27485486211901305,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499616990768058,
      "loss": 5.8363,
      "mean_token_accuracy": 0.14764970988035203,
      "num_tokens": 6275241.0,
      "step": 3030
    },
    {
      "entropy": 5.906813526153565,
      "epoch": 0.27530841799709727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996151013063579,
      "loss": 5.8813,
      "mean_token_accuracy": 0.14210136905312537,
      "num_tokens": 6286603.0,
      "step": 3035
    },
    {
      "entropy": 5.801692199707031,
      "epoch": 0.2757619738751814,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996132071995601,
      "loss": 5.7166,
      "mean_token_accuracy": 0.1628328561782837,
      "num_tokens": 6297219.0,
      "step": 3040
    },
    {
      "entropy": 5.741138601303101,
      "epoch": 0.2762155297532656,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004996113084477042,
      "loss": 5.7775,
      "mean_token_accuracy": 0.1547319009900093,
      "num_tokens": 6307694.0,
      "step": 3045
    },
    {
      "entropy": 5.750327444076538,
      "epoch": 0.2766690856313498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996094050508291,
      "loss": 5.6979,
      "mean_token_accuracy": 0.15616930574178695,
      "num_tokens": 6317384.0,
      "step": 3050
    },
    {
      "entropy": 5.863014984130859,
      "epoch": 0.27712264150943394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996074970089741,
      "loss": 5.8308,
      "mean_token_accuracy": 0.14901937842369078,
      "num_tokens": 6327558.0,
      "step": 3055
    },
    {
      "entropy": 5.802416276931763,
      "epoch": 0.27757619738751815,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004996055843221789,
      "loss": 5.6831,
      "mean_token_accuracy": 0.16094797402620314,
      "num_tokens": 6337828.0,
      "step": 3060
    },
    {
      "entropy": 5.804899978637695,
      "epoch": 0.2780297532656023,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996036669904831,
      "loss": 5.7488,
      "mean_token_accuracy": 0.1513446494936943,
      "num_tokens": 6348678.0,
      "step": 3065
    },
    {
      "entropy": 5.784188747406006,
      "epoch": 0.2784833091436865,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996017450139261,
      "loss": 5.834,
      "mean_token_accuracy": 0.1463714897632599,
      "num_tokens": 6359670.0,
      "step": 3070
    },
    {
      "entropy": 5.87315411567688,
      "epoch": 0.2789368650217707,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995998183925478,
      "loss": 5.7926,
      "mean_token_accuracy": 0.15490294694900514,
      "num_tokens": 6371034.0,
      "step": 3075
    },
    {
      "entropy": 5.775865983963013,
      "epoch": 0.2793904208998549,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995978871263881,
      "loss": 5.7395,
      "mean_token_accuracy": 0.1566627189517021,
      "num_tokens": 6380672.0,
      "step": 3080
    },
    {
      "entropy": 5.811061096191406,
      "epoch": 0.27984397677793904,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004995959512154868,
      "loss": 5.8517,
      "mean_token_accuracy": 0.1400656595826149,
      "num_tokens": 6390799.0,
      "step": 3085
    },
    {
      "entropy": 5.772877502441406,
      "epoch": 0.2802975326560232,
      "grad_norm": 1.125,
      "learning_rate": 0.000499594010659884,
      "loss": 5.724,
      "mean_token_accuracy": 0.15223535746335984,
      "num_tokens": 6400965.0,
      "step": 3090
    },
    {
      "entropy": 5.814068937301636,
      "epoch": 0.2807510885341074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995920654596199,
      "loss": 5.7517,
      "mean_token_accuracy": 0.14599630907177924,
      "num_tokens": 6411567.0,
      "step": 3095
    },
    {
      "entropy": 5.83027400970459,
      "epoch": 0.28120464441219156,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995901156147345,
      "loss": 5.804,
      "mean_token_accuracy": 0.1567014843225479,
      "num_tokens": 6421193.0,
      "step": 3100
    },
    {
      "entropy": 5.850267601013184,
      "epoch": 0.2816582002902758,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995881611252684,
      "loss": 5.8144,
      "mean_token_accuracy": 0.15475278794765474,
      "num_tokens": 6430281.0,
      "step": 3105
    },
    {
      "entropy": 5.7939496517181395,
      "epoch": 0.28211175616835993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995862019912619,
      "loss": 5.8048,
      "mean_token_accuracy": 0.146951512247324,
      "num_tokens": 6441055.0,
      "step": 3110
    },
    {
      "entropy": 5.824093437194824,
      "epoch": 0.28256531204644414,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995842382127555,
      "loss": 5.7252,
      "mean_token_accuracy": 0.15018295422196387,
      "num_tokens": 6451082.0,
      "step": 3115
    },
    {
      "entropy": 5.802210283279419,
      "epoch": 0.2830188679245283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995822697897897,
      "loss": 5.8054,
      "mean_token_accuracy": 0.14790797382593154,
      "num_tokens": 6461617.0,
      "step": 3120
    },
    {
      "entropy": 5.898024320602417,
      "epoch": 0.2834724238026125,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995802967224056,
      "loss": 5.8146,
      "mean_token_accuracy": 0.14873994886875153,
      "num_tokens": 6472780.0,
      "step": 3125
    },
    {
      "entropy": 5.7456702709198,
      "epoch": 0.28392597968069666,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995783190106434,
      "loss": 5.7096,
      "mean_token_accuracy": 0.15184671580791473,
      "num_tokens": 6484078.0,
      "step": 3130
    },
    {
      "entropy": 5.8592980861663815,
      "epoch": 0.2843795355587808,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995763366545445,
      "loss": 5.8203,
      "mean_token_accuracy": 0.14368489608168603,
      "num_tokens": 6494366.0,
      "step": 3135
    },
    {
      "entropy": 5.727130985260009,
      "epoch": 0.284833091436865,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995743496541497,
      "loss": 5.7561,
      "mean_token_accuracy": 0.15323618352413176,
      "num_tokens": 6505461.0,
      "step": 3140
    },
    {
      "entropy": 5.864217281341553,
      "epoch": 0.2852866473149492,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995723580095001,
      "loss": 5.7622,
      "mean_token_accuracy": 0.15169936046004295,
      "num_tokens": 6516141.0,
      "step": 3145
    },
    {
      "entropy": 5.699955415725708,
      "epoch": 0.2857402031930334,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499570361720637,
      "loss": 5.7494,
      "mean_token_accuracy": 0.15026842653751374,
      "num_tokens": 6525536.0,
      "step": 3150
    },
    {
      "entropy": 5.858482599258423,
      "epoch": 0.28619375907111755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995683607876015,
      "loss": 5.807,
      "mean_token_accuracy": 0.1474149063229561,
      "num_tokens": 6535852.0,
      "step": 3155
    },
    {
      "entropy": 5.877813768386841,
      "epoch": 0.28664731494920176,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995663552104352,
      "loss": 5.7854,
      "mean_token_accuracy": 0.15658673495054246,
      "num_tokens": 6546353.0,
      "step": 3160
    },
    {
      "entropy": 5.71932168006897,
      "epoch": 0.2871008708272859,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995643449891793,
      "loss": 5.7043,
      "mean_token_accuracy": 0.15023708939552308,
      "num_tokens": 6557782.0,
      "step": 3165
    },
    {
      "entropy": 5.821673774719239,
      "epoch": 0.2875544267053701,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995623301238757,
      "loss": 5.7329,
      "mean_token_accuracy": 0.15660330504179001,
      "num_tokens": 6568100.0,
      "step": 3170
    },
    {
      "entropy": 5.765221214294433,
      "epoch": 0.2880079825834543,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995603106145657,
      "loss": 5.8021,
      "mean_token_accuracy": 0.14751623272895814,
      "num_tokens": 6578698.0,
      "step": 3175
    },
    {
      "entropy": 5.8275402069091795,
      "epoch": 0.28846153846153844,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995582864612913,
      "loss": 5.7745,
      "mean_token_accuracy": 0.14643221199512482,
      "num_tokens": 6588908.0,
      "step": 3180
    },
    {
      "entropy": 5.737450218200683,
      "epoch": 0.28891509433962265,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995562576640943,
      "loss": 5.6387,
      "mean_token_accuracy": 0.15332392156124114,
      "num_tokens": 6599445.0,
      "step": 3185
    },
    {
      "entropy": 5.767725610733033,
      "epoch": 0.2893686502177068,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995542242230167,
      "loss": 5.7407,
      "mean_token_accuracy": 0.15147417038679123,
      "num_tokens": 6610246.0,
      "step": 3190
    },
    {
      "entropy": 5.761809396743774,
      "epoch": 0.289822206095791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995521861381006,
      "loss": 5.7459,
      "mean_token_accuracy": 0.1523432895541191,
      "num_tokens": 6620501.0,
      "step": 3195
    },
    {
      "entropy": 5.912834644317627,
      "epoch": 0.29027576197387517,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499550143409388,
      "loss": 5.8203,
      "mean_token_accuracy": 0.14403355419635772,
      "num_tokens": 6631206.0,
      "step": 3200
    },
    {
      "entropy": 5.721426582336425,
      "epoch": 0.2907293178519594,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995480960369211,
      "loss": 5.7518,
      "mean_token_accuracy": 0.14767736345529556,
      "num_tokens": 6641600.0,
      "step": 3205
    },
    {
      "entropy": 5.808769845962525,
      "epoch": 0.29118287373004353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995460440207425,
      "loss": 5.8272,
      "mean_token_accuracy": 0.14944822043180467,
      "num_tokens": 6652319.0,
      "step": 3210
    },
    {
      "entropy": 5.869615316390991,
      "epoch": 0.29163642960812775,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995439873608945,
      "loss": 5.7555,
      "mean_token_accuracy": 0.1475253716111183,
      "num_tokens": 6663702.0,
      "step": 3215
    },
    {
      "entropy": 5.792519235610962,
      "epoch": 0.2920899854862119,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995419260574196,
      "loss": 5.751,
      "mean_token_accuracy": 0.1453598067164421,
      "num_tokens": 6673908.0,
      "step": 3220
    },
    {
      "entropy": 5.787820816040039,
      "epoch": 0.29254354136429606,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995398601103603,
      "loss": 5.7812,
      "mean_token_accuracy": 0.15197647958993912,
      "num_tokens": 6683237.0,
      "step": 3225
    },
    {
      "entropy": 5.817029237747192,
      "epoch": 0.29299709724238027,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995377895197597,
      "loss": 5.7149,
      "mean_token_accuracy": 0.15499222725629808,
      "num_tokens": 6693451.0,
      "step": 3230
    },
    {
      "entropy": 5.665473747253418,
      "epoch": 0.2934506531204644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995357142856602,
      "loss": 5.654,
      "mean_token_accuracy": 0.15413334965705872,
      "num_tokens": 6703479.0,
      "step": 3235
    },
    {
      "entropy": 5.710675239562988,
      "epoch": 0.29390420899854863,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499533634408105,
      "loss": 5.6716,
      "mean_token_accuracy": 0.15576022118330002,
      "num_tokens": 6715321.0,
      "step": 3240
    },
    {
      "entropy": 5.783253860473633,
      "epoch": 0.2943577648766328,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499531549887137,
      "loss": 5.7149,
      "mean_token_accuracy": 0.15394483953714372,
      "num_tokens": 6725541.0,
      "step": 3245
    },
    {
      "entropy": 5.759927129745483,
      "epoch": 0.294811320754717,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004995294607227993,
      "loss": 5.7693,
      "mean_token_accuracy": 0.1588973805308342,
      "num_tokens": 6735165.0,
      "step": 3250
    },
    {
      "entropy": 5.8403795719146725,
      "epoch": 0.29526487663280115,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995273669151351,
      "loss": 5.7616,
      "mean_token_accuracy": 0.15932275354862213,
      "num_tokens": 6744245.0,
      "step": 3255
    },
    {
      "entropy": 5.75037317276001,
      "epoch": 0.29571843251088537,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499525268464188,
      "loss": 5.7239,
      "mean_token_accuracy": 0.1572644904255867,
      "num_tokens": 6754894.0,
      "step": 3260
    },
    {
      "entropy": 5.779234743118286,
      "epoch": 0.2961719883889695,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995231653700009,
      "loss": 5.7984,
      "mean_token_accuracy": 0.14348074272274972,
      "num_tokens": 6764442.0,
      "step": 3265
    },
    {
      "entropy": 5.824223375320434,
      "epoch": 0.2966255442670537,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995210576326176,
      "loss": 5.7582,
      "mean_token_accuracy": 0.14877086579799653,
      "num_tokens": 6775054.0,
      "step": 3270
    },
    {
      "entropy": 5.795979833602905,
      "epoch": 0.2970791001451379,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995189452520816,
      "loss": 5.7278,
      "mean_token_accuracy": 0.1630966395139694,
      "num_tokens": 6784825.0,
      "step": 3275
    },
    {
      "entropy": 5.796776056289673,
      "epoch": 0.29753265602322204,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004995168282284366,
      "loss": 5.7803,
      "mean_token_accuracy": 0.15851086676120757,
      "num_tokens": 6795928.0,
      "step": 3280
    },
    {
      "entropy": 5.693297910690307,
      "epoch": 0.29798621190130625,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995147065617263,
      "loss": 5.6995,
      "mean_token_accuracy": 0.1500129833817482,
      "num_tokens": 6805456.0,
      "step": 3285
    },
    {
      "entropy": 5.815392637252808,
      "epoch": 0.2984397677793904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995125802519946,
      "loss": 5.7757,
      "mean_token_accuracy": 0.15002186894416808,
      "num_tokens": 6816215.0,
      "step": 3290
    },
    {
      "entropy": 5.834223985671997,
      "epoch": 0.2988933236574746,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995104492992857,
      "loss": 5.8169,
      "mean_token_accuracy": 0.14767192974686622,
      "num_tokens": 6826545.0,
      "step": 3295
    },
    {
      "entropy": 5.70528507232666,
      "epoch": 0.2993468795355588,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004995083137036434,
      "loss": 5.7226,
      "mean_token_accuracy": 0.15581652522087097,
      "num_tokens": 6836184.0,
      "step": 3300
    },
    {
      "entropy": 5.8633002758026125,
      "epoch": 0.299800435413643,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995061734651121,
      "loss": 5.7961,
      "mean_token_accuracy": 0.15215000808238982,
      "num_tokens": 6846646.0,
      "step": 3305
    },
    {
      "entropy": 5.830773496627808,
      "epoch": 0.30025399129172714,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995040285837358,
      "loss": 5.7157,
      "mean_token_accuracy": 0.16062179654836656,
      "num_tokens": 6857597.0,
      "step": 3310
    },
    {
      "entropy": 5.789421701431275,
      "epoch": 0.3007075471698113,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995018790595589,
      "loss": 5.796,
      "mean_token_accuracy": 0.15099546015262605,
      "num_tokens": 6867523.0,
      "step": 3315
    },
    {
      "entropy": 5.900367307662964,
      "epoch": 0.3011611030478955,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994997248926261,
      "loss": 5.8924,
      "mean_token_accuracy": 0.14628595039248465,
      "num_tokens": 6878190.0,
      "step": 3320
    },
    {
      "entropy": 5.762005186080932,
      "epoch": 0.30161465892597966,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994975660829817,
      "loss": 5.6861,
      "mean_token_accuracy": 0.16212552934885024,
      "num_tokens": 6888483.0,
      "step": 3325
    },
    {
      "entropy": 5.688748073577881,
      "epoch": 0.3020682148040639,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994954026306705,
      "loss": 5.6749,
      "mean_token_accuracy": 0.15575181618332862,
      "num_tokens": 6898337.0,
      "step": 3330
    },
    {
      "entropy": 5.831185674667358,
      "epoch": 0.30252177068214803,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994932345357371,
      "loss": 5.7539,
      "mean_token_accuracy": 0.1522991269826889,
      "num_tokens": 6909131.0,
      "step": 3335
    },
    {
      "entropy": 5.75893816947937,
      "epoch": 0.30297532656023224,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994910617982264,
      "loss": 5.7326,
      "mean_token_accuracy": 0.14802005887031555,
      "num_tokens": 6919790.0,
      "step": 3340
    },
    {
      "entropy": 5.829823637008667,
      "epoch": 0.3034288824383164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994888844181833,
      "loss": 5.7614,
      "mean_token_accuracy": 0.1517559289932251,
      "num_tokens": 6930072.0,
      "step": 3345
    },
    {
      "entropy": 5.741778373718262,
      "epoch": 0.3038824383164006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994867023956529,
      "loss": 5.6478,
      "mean_token_accuracy": 0.15023863464593887,
      "num_tokens": 6940646.0,
      "step": 3350
    },
    {
      "entropy": 5.6631731510162355,
      "epoch": 0.30433599419448476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994845157306803,
      "loss": 5.6513,
      "mean_token_accuracy": 0.15208848789334298,
      "num_tokens": 6951645.0,
      "step": 3355
    },
    {
      "entropy": 5.856181573867798,
      "epoch": 0.3047895500725689,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994823244233107,
      "loss": 5.7988,
      "mean_token_accuracy": 0.15186134427785875,
      "num_tokens": 6963020.0,
      "step": 3360
    },
    {
      "entropy": 5.821042013168335,
      "epoch": 0.3052431059506531,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004994801284735895,
      "loss": 5.7291,
      "mean_token_accuracy": 0.15716724395751952,
      "num_tokens": 6972895.0,
      "step": 3365
    },
    {
      "entropy": 5.8223480701446535,
      "epoch": 0.3056966618287373,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994779278815619,
      "loss": 5.8715,
      "mean_token_accuracy": 0.14308758080005646,
      "num_tokens": 6985363.0,
      "step": 3370
    },
    {
      "entropy": 5.8458006381988525,
      "epoch": 0.3061502177068215,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004994757226472736,
      "loss": 5.8187,
      "mean_token_accuracy": 0.15202625691890717,
      "num_tokens": 6997040.0,
      "step": 3375
    },
    {
      "entropy": 5.806244993209839,
      "epoch": 0.30660377358490565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994735127707702,
      "loss": 5.7438,
      "mean_token_accuracy": 0.15390410125255585,
      "num_tokens": 7007591.0,
      "step": 3380
    },
    {
      "entropy": 5.826634693145752,
      "epoch": 0.30705732946298986,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004994712982520972,
      "loss": 5.7401,
      "mean_token_accuracy": 0.16007999330759048,
      "num_tokens": 7017858.0,
      "step": 3385
    },
    {
      "entropy": 5.678028774261475,
      "epoch": 0.307510885341074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994690790913008,
      "loss": 5.704,
      "mean_token_accuracy": 0.15252023860812186,
      "num_tokens": 7027365.0,
      "step": 3390
    },
    {
      "entropy": 5.743245077133179,
      "epoch": 0.3079644412191582,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994668552884265,
      "loss": 5.7079,
      "mean_token_accuracy": 0.15393923968076706,
      "num_tokens": 7038728.0,
      "step": 3395
    },
    {
      "entropy": 5.860477924346924,
      "epoch": 0.3084179970972424,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004994646268435204,
      "loss": 5.649,
      "mean_token_accuracy": 0.15698569566011428,
      "num_tokens": 7048736.0,
      "step": 3400
    },
    {
      "entropy": 5.672683954238892,
      "epoch": 0.30887155297532654,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994623937566286,
      "loss": 5.7635,
      "mean_token_accuracy": 0.15398403406143188,
      "num_tokens": 7059205.0,
      "step": 3405
    },
    {
      "entropy": 5.810827445983887,
      "epoch": 0.30932510885341075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994601560277974,
      "loss": 5.7705,
      "mean_token_accuracy": 0.151111601293087,
      "num_tokens": 7069513.0,
      "step": 3410
    },
    {
      "entropy": 5.794088840484619,
      "epoch": 0.3097786647314949,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499457913657073,
      "loss": 5.6867,
      "mean_token_accuracy": 0.15149736255407334,
      "num_tokens": 7079940.0,
      "step": 3415
    },
    {
      "entropy": 5.805775165557861,
      "epoch": 0.3102322206095791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994556666445016,
      "loss": 5.6806,
      "mean_token_accuracy": 0.1570603534579277,
      "num_tokens": 7090792.0,
      "step": 3420
    },
    {
      "entropy": 5.662250566482544,
      "epoch": 0.31068577648766327,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994534149901299,
      "loss": 5.6457,
      "mean_token_accuracy": 0.15090260058641433,
      "num_tokens": 7101964.0,
      "step": 3425
    },
    {
      "entropy": 5.749277830123901,
      "epoch": 0.3111393323657475,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004994511586940044,
      "loss": 5.7098,
      "mean_token_accuracy": 0.15125103592872619,
      "num_tokens": 7113332.0,
      "step": 3430
    },
    {
      "entropy": 5.871686697006226,
      "epoch": 0.31159288824383163,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004994488977561717,
      "loss": 5.8774,
      "mean_token_accuracy": 0.14723959863185881,
      "num_tokens": 7123481.0,
      "step": 3435
    },
    {
      "entropy": 5.696882629394532,
      "epoch": 0.31204644412191584,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994466321766786,
      "loss": 5.6277,
      "mean_token_accuracy": 0.15849945396184922,
      "num_tokens": 7133694.0,
      "step": 3440
    },
    {
      "entropy": 5.741990947723389,
      "epoch": 0.3125,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499444361955572,
      "loss": 5.7319,
      "mean_token_accuracy": 0.15803271979093553,
      "num_tokens": 7144491.0,
      "step": 3445
    },
    {
      "entropy": 5.9043221950531,
      "epoch": 0.31295355587808416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994420870928988,
      "loss": 5.8197,
      "mean_token_accuracy": 0.14943068027496337,
      "num_tokens": 7154716.0,
      "step": 3450
    },
    {
      "entropy": 5.81868691444397,
      "epoch": 0.31340711175616837,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994398075887059,
      "loss": 5.8234,
      "mean_token_accuracy": 0.14441180527210234,
      "num_tokens": 7164258.0,
      "step": 3455
    },
    {
      "entropy": 5.822063636779785,
      "epoch": 0.3138606676342525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994375234430407,
      "loss": 5.6777,
      "mean_token_accuracy": 0.15146763920783995,
      "num_tokens": 7175098.0,
      "step": 3460
    },
    {
      "entropy": 5.750956344604492,
      "epoch": 0.31431422351233673,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994352346559504,
      "loss": 5.7521,
      "mean_token_accuracy": 0.15109349340200423,
      "num_tokens": 7186480.0,
      "step": 3465
    },
    {
      "entropy": 5.7146683692932125,
      "epoch": 0.3147677793904209,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004994329412274821,
      "loss": 5.7403,
      "mean_token_accuracy": 0.15527350157499314,
      "num_tokens": 7197056.0,
      "step": 3470
    },
    {
      "entropy": 5.814749813079834,
      "epoch": 0.3152213352685051,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004994306431576835,
      "loss": 5.8274,
      "mean_token_accuracy": 0.14993275180459023,
      "num_tokens": 7206040.0,
      "step": 3475
    },
    {
      "entropy": 5.851417589187622,
      "epoch": 0.31567489114658925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499428340446602,
      "loss": 5.694,
      "mean_token_accuracy": 0.15531422793865204,
      "num_tokens": 7216344.0,
      "step": 3480
    },
    {
      "entropy": 5.784599637985229,
      "epoch": 0.31612844702467346,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994260330942851,
      "loss": 5.7747,
      "mean_token_accuracy": 0.1495263546705246,
      "num_tokens": 7226729.0,
      "step": 3485
    },
    {
      "entropy": 5.67893795967102,
      "epoch": 0.3165820029027576,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994237211007808,
      "loss": 5.7149,
      "mean_token_accuracy": 0.15380621552467347,
      "num_tokens": 7237616.0,
      "step": 3490
    },
    {
      "entropy": 5.821342611312867,
      "epoch": 0.3170355587808418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994214044661367,
      "loss": 5.6561,
      "mean_token_accuracy": 0.1568485826253891,
      "num_tokens": 7247857.0,
      "step": 3495
    },
    {
      "entropy": 5.650297212600708,
      "epoch": 0.317489114658926,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994190831904007,
      "loss": 5.6209,
      "mean_token_accuracy": 0.16020147651433944,
      "num_tokens": 7257858.0,
      "step": 3500
    },
    {
      "entropy": 5.652521848678589,
      "epoch": 0.31794267053701014,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499416757273621,
      "loss": 5.6358,
      "mean_token_accuracy": 0.15801683515310289,
      "num_tokens": 7268440.0,
      "step": 3505
    },
    {
      "entropy": 5.8561008930206295,
      "epoch": 0.31839622641509435,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004994144267158454,
      "loss": 5.7149,
      "mean_token_accuracy": 0.15808832198381423,
      "num_tokens": 7278560.0,
      "step": 3510
    },
    {
      "entropy": 5.721616554260254,
      "epoch": 0.3188497822931785,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994120915171224,
      "loss": 5.7152,
      "mean_token_accuracy": 0.15598661005496978,
      "num_tokens": 7288422.0,
      "step": 3515
    },
    {
      "entropy": 5.822176361083985,
      "epoch": 0.3193033381712627,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994097516775002,
      "loss": 5.7688,
      "mean_token_accuracy": 0.1543071150779724,
      "num_tokens": 7298613.0,
      "step": 3520
    },
    {
      "entropy": 5.81471004486084,
      "epoch": 0.3197568940493469,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994074071970271,
      "loss": 5.8157,
      "mean_token_accuracy": 0.1467634655535221,
      "num_tokens": 7309509.0,
      "step": 3525
    },
    {
      "entropy": 5.774685192108154,
      "epoch": 0.3202104499274311,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004994050580757515,
      "loss": 5.7244,
      "mean_token_accuracy": 0.15079469233751297,
      "num_tokens": 7319015.0,
      "step": 3530
    },
    {
      "entropy": 5.718948554992676,
      "epoch": 0.32066400580551524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994027043137223,
      "loss": 5.6906,
      "mean_token_accuracy": 0.16023388206958772,
      "num_tokens": 7329026.0,
      "step": 3535
    },
    {
      "entropy": 5.813447046279907,
      "epoch": 0.3211175616835994,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994003459109879,
      "loss": 5.7859,
      "mean_token_accuracy": 0.1511634275317192,
      "num_tokens": 7339638.0,
      "step": 3540
    },
    {
      "entropy": 5.781355905532837,
      "epoch": 0.3215711175616836,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993979828675972,
      "loss": 5.6939,
      "mean_token_accuracy": 0.15422995537519454,
      "num_tokens": 7349979.0,
      "step": 3545
    },
    {
      "entropy": 5.67366590499878,
      "epoch": 0.32202467343976776,
      "grad_norm": 1.1875,
      "learning_rate": 0.000499395615183599,
      "loss": 5.6205,
      "mean_token_accuracy": 0.1652266263961792,
      "num_tokens": 7359398.0,
      "step": 3550
    },
    {
      "entropy": 5.656069612503051,
      "epoch": 0.32247822931785197,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993932428590422,
      "loss": 5.671,
      "mean_token_accuracy": 0.15403880178928375,
      "num_tokens": 7370418.0,
      "step": 3555
    },
    {
      "entropy": 5.780758047103882,
      "epoch": 0.3229317851959361,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004993908658939762,
      "loss": 5.7088,
      "mean_token_accuracy": 0.1502725899219513,
      "num_tokens": 7380461.0,
      "step": 3560
    },
    {
      "entropy": 5.806817102432251,
      "epoch": 0.32338534107402034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993884842884496,
      "loss": 5.7122,
      "mean_token_accuracy": 0.15863013118505478,
      "num_tokens": 7390646.0,
      "step": 3565
    },
    {
      "entropy": 5.747936010360718,
      "epoch": 0.3238388969521045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993860980425121,
      "loss": 5.8076,
      "mean_token_accuracy": 0.155539433658123,
      "num_tokens": 7402106.0,
      "step": 3570
    },
    {
      "entropy": 5.781191968917847,
      "epoch": 0.3242924528301887,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993837071562131,
      "loss": 5.7951,
      "mean_token_accuracy": 0.15347181856632233,
      "num_tokens": 7412936.0,
      "step": 3575
    },
    {
      "entropy": 5.744540643692017,
      "epoch": 0.32474600870827286,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004993813116296017,
      "loss": 5.654,
      "mean_token_accuracy": 0.15949824452400208,
      "num_tokens": 7422303.0,
      "step": 3580
    },
    {
      "entropy": 5.784517669677735,
      "epoch": 0.325199564586357,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004993789114627275,
      "loss": 5.7187,
      "mean_token_accuracy": 0.15309756696224214,
      "num_tokens": 7432829.0,
      "step": 3585
    },
    {
      "entropy": 5.81943531036377,
      "epoch": 0.3256531204644412,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004993765066556403,
      "loss": 5.753,
      "mean_token_accuracy": 0.15031561404466628,
      "num_tokens": 7443218.0,
      "step": 3590
    },
    {
      "entropy": 5.7325328350067135,
      "epoch": 0.3261066763425254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993740972083899,
      "loss": 5.6859,
      "mean_token_accuracy": 0.15189058780670167,
      "num_tokens": 7454531.0,
      "step": 3595
    },
    {
      "entropy": 5.796178579330444,
      "epoch": 0.3265602322206096,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004993716831210259,
      "loss": 5.7621,
      "mean_token_accuracy": 0.1558135211467743,
      "num_tokens": 7464993.0,
      "step": 3600
    },
    {
      "entropy": 5.710817050933838,
      "epoch": 0.32701378809869375,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993692643935983,
      "loss": 5.7732,
      "mean_token_accuracy": 0.15145019814372063,
      "num_tokens": 7475755.0,
      "step": 3605
    },
    {
      "entropy": 5.791586971282959,
      "epoch": 0.32746734397677796,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004993668410261571,
      "loss": 5.7697,
      "mean_token_accuracy": 0.15512488633394242,
      "num_tokens": 7486778.0,
      "step": 3610
    },
    {
      "entropy": 5.883192777633667,
      "epoch": 0.3279208998548621,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004993644130187525,
      "loss": 5.7449,
      "mean_token_accuracy": 0.15393500328063964,
      "num_tokens": 7497439.0,
      "step": 3615
    },
    {
      "entropy": 5.713404703140259,
      "epoch": 0.3283744557329463,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004993619803714347,
      "loss": 5.7857,
      "mean_token_accuracy": 0.149221783131361,
      "num_tokens": 7508274.0,
      "step": 3620
    },
    {
      "entropy": 5.824280738830566,
      "epoch": 0.3288280116110305,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993595430842539,
      "loss": 5.7501,
      "mean_token_accuracy": 0.15261871069669725,
      "num_tokens": 7519315.0,
      "step": 3625
    },
    {
      "entropy": 5.868468761444092,
      "epoch": 0.32928156748911463,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993571011572606,
      "loss": 5.7546,
      "mean_token_accuracy": 0.15521898567676545,
      "num_tokens": 7529910.0,
      "step": 3630
    },
    {
      "entropy": 5.664683103561401,
      "epoch": 0.32973512336719885,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004993546545905053,
      "loss": 5.681,
      "mean_token_accuracy": 0.15814965665340425,
      "num_tokens": 7538966.0,
      "step": 3635
    },
    {
      "entropy": 5.843475103378296,
      "epoch": 0.330188679245283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993522033840386,
      "loss": 5.7227,
      "mean_token_accuracy": 0.14933252111077308,
      "num_tokens": 7549683.0,
      "step": 3640
    },
    {
      "entropy": 5.716665935516358,
      "epoch": 0.3306422351233672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499349747537911,
      "loss": 5.6994,
      "mean_token_accuracy": 0.15621910542249678,
      "num_tokens": 7560629.0,
      "step": 3645
    },
    {
      "entropy": 5.719961214065552,
      "epoch": 0.33109579100145137,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993472870521736,
      "loss": 5.6413,
      "mean_token_accuracy": 0.1616758957505226,
      "num_tokens": 7570613.0,
      "step": 3650
    },
    {
      "entropy": 5.70309534072876,
      "epoch": 0.3315493468795356,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004993448219268771,
      "loss": 5.6207,
      "mean_token_accuracy": 0.15617789477109909,
      "num_tokens": 7580640.0,
      "step": 3655
    },
    {
      "entropy": 5.7273846626281735,
      "epoch": 0.33200290275761973,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004993423521620726,
      "loss": 5.7108,
      "mean_token_accuracy": 0.1586635485291481,
      "num_tokens": 7590164.0,
      "step": 3660
    },
    {
      "entropy": 5.818833351135254,
      "epoch": 0.33245645863570394,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004993398777578109,
      "loss": 5.7807,
      "mean_token_accuracy": 0.15034616962075234,
      "num_tokens": 7600173.0,
      "step": 3665
    },
    {
      "entropy": 5.7085833072662355,
      "epoch": 0.3329100145137881,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993373987141435,
      "loss": 5.5698,
      "mean_token_accuracy": 0.16563706398010253,
      "num_tokens": 7609958.0,
      "step": 3670
    },
    {
      "entropy": 5.732572078704834,
      "epoch": 0.33336357039187225,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993349150311214,
      "loss": 5.7178,
      "mean_token_accuracy": 0.15159249901771546,
      "num_tokens": 7620237.0,
      "step": 3675
    },
    {
      "entropy": 5.698968887329102,
      "epoch": 0.33381712626995647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993324267087962,
      "loss": 5.6101,
      "mean_token_accuracy": 0.1608090043067932,
      "num_tokens": 7629886.0,
      "step": 3680
    },
    {
      "entropy": 5.755589199066162,
      "epoch": 0.3342706821480406,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993299337472192,
      "loss": 5.6803,
      "mean_token_accuracy": 0.15914253294467925,
      "num_tokens": 7640057.0,
      "step": 3685
    },
    {
      "entropy": 5.7210606098175045,
      "epoch": 0.33472423802612483,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499327436146442,
      "loss": 5.6389,
      "mean_token_accuracy": 0.15623899549245834,
      "num_tokens": 7649766.0,
      "step": 3690
    },
    {
      "entropy": 5.704188871383667,
      "epoch": 0.335177793904209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993249339065162,
      "loss": 5.6988,
      "mean_token_accuracy": 0.15833286568522453,
      "num_tokens": 7660890.0,
      "step": 3695
    },
    {
      "entropy": 5.789506959915161,
      "epoch": 0.3356313497822932,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004993224270274937,
      "loss": 5.7184,
      "mean_token_accuracy": 0.15361475795507432,
      "num_tokens": 7670737.0,
      "step": 3700
    },
    {
      "entropy": 5.6635422706604,
      "epoch": 0.33608490566037735,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993199155094263,
      "loss": 5.6866,
      "mean_token_accuracy": 0.15698174983263016,
      "num_tokens": 7681407.0,
      "step": 3705
    },
    {
      "entropy": 5.730187511444091,
      "epoch": 0.33653846153846156,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993173993523658,
      "loss": 5.6158,
      "mean_token_accuracy": 0.16286221295595169,
      "num_tokens": 7691758.0,
      "step": 3710
    },
    {
      "entropy": 5.675654602050781,
      "epoch": 0.3369920174165457,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004993148785563643,
      "loss": 5.6525,
      "mean_token_accuracy": 0.15679143369197845,
      "num_tokens": 7702934.0,
      "step": 3715
    },
    {
      "entropy": 5.694839334487915,
      "epoch": 0.3374455732946299,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499312353121474,
      "loss": 5.6491,
      "mean_token_accuracy": 0.15497201532125474,
      "num_tokens": 7713001.0,
      "step": 3720
    },
    {
      "entropy": 5.801605463027954,
      "epoch": 0.3378991291727141,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004993098230477471,
      "loss": 5.7128,
      "mean_token_accuracy": 0.14621546119451523,
      "num_tokens": 7724005.0,
      "step": 3725
    },
    {
      "entropy": 5.722197771072388,
      "epoch": 0.33835268505079824,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004993072883352358,
      "loss": 5.6576,
      "mean_token_accuracy": 0.15557765886187552,
      "num_tokens": 7733743.0,
      "step": 3730
    },
    {
      "entropy": 5.782563161849976,
      "epoch": 0.33880624092888245,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004993047489839927,
      "loss": 5.8425,
      "mean_token_accuracy": 0.1471106566488743,
      "num_tokens": 7743762.0,
      "step": 3735
    },
    {
      "entropy": 5.88274621963501,
      "epoch": 0.3392597968069666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993022049940702,
      "loss": 5.7398,
      "mean_token_accuracy": 0.15454336851835251,
      "num_tokens": 7753758.0,
      "step": 3740
    },
    {
      "entropy": 5.700439834594727,
      "epoch": 0.3397133526850508,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992996563655209,
      "loss": 5.6715,
      "mean_token_accuracy": 0.15768471136689186,
      "num_tokens": 7764692.0,
      "step": 3745
    },
    {
      "entropy": 5.702702379226684,
      "epoch": 0.340166908563135,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992971030983975,
      "loss": 5.718,
      "mean_token_accuracy": 0.155170801281929,
      "num_tokens": 7776892.0,
      "step": 3750
    },
    {
      "entropy": 5.838658714294434,
      "epoch": 0.3406204644412192,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992945451927531,
      "loss": 5.7769,
      "mean_token_accuracy": 0.15279003530740737,
      "num_tokens": 7787143.0,
      "step": 3755
    },
    {
      "entropy": 5.725220155715943,
      "epoch": 0.34107402031930334,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992919826486402,
      "loss": 5.674,
      "mean_token_accuracy": 0.1564330369234085,
      "num_tokens": 7798537.0,
      "step": 3760
    },
    {
      "entropy": 5.685366582870484,
      "epoch": 0.3415275761973875,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992894154661118,
      "loss": 5.6518,
      "mean_token_accuracy": 0.1586100533604622,
      "num_tokens": 7808434.0,
      "step": 3765
    },
    {
      "entropy": 5.7612800121307375,
      "epoch": 0.3419811320754717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004992868436452213,
      "loss": 5.7717,
      "mean_token_accuracy": 0.15543903708457946,
      "num_tokens": 7819592.0,
      "step": 3770
    },
    {
      "entropy": 5.801908349990844,
      "epoch": 0.34243468795355586,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004992842671860217,
      "loss": 5.6409,
      "mean_token_accuracy": 0.16048139780759813,
      "num_tokens": 7830602.0,
      "step": 3775
    },
    {
      "entropy": 5.761849641799927,
      "epoch": 0.34288824383164007,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004992816860885662,
      "loss": 5.8073,
      "mean_token_accuracy": 0.1533203274011612,
      "num_tokens": 7840850.0,
      "step": 3780
    },
    {
      "entropy": 5.884047603607177,
      "epoch": 0.3433417997097242,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004992791003529084,
      "loss": 5.7605,
      "mean_token_accuracy": 0.14962777942419053,
      "num_tokens": 7852371.0,
      "step": 3785
    },
    {
      "entropy": 5.735295009613037,
      "epoch": 0.34379535558780844,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992765099791017,
      "loss": 5.7693,
      "mean_token_accuracy": 0.1477210432291031,
      "num_tokens": 7863290.0,
      "step": 3790
    },
    {
      "entropy": 5.721421813964843,
      "epoch": 0.3442489114658926,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004992739149671995,
      "loss": 5.6176,
      "mean_token_accuracy": 0.15730728954076767,
      "num_tokens": 7873927.0,
      "step": 3795
    },
    {
      "entropy": 5.7242011547088625,
      "epoch": 0.3447024673439768,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992713153172556,
      "loss": 5.638,
      "mean_token_accuracy": 0.15784906297922136,
      "num_tokens": 7884977.0,
      "step": 3800
    },
    {
      "entropy": 5.627726364135742,
      "epoch": 0.34515602322206096,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004992687110293238,
      "loss": 5.7255,
      "mean_token_accuracy": 0.16205305680632592,
      "num_tokens": 7894620.0,
      "step": 3805
    },
    {
      "entropy": 5.740850877761841,
      "epoch": 0.3456095791001451,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992661021034579,
      "loss": 5.6443,
      "mean_token_accuracy": 0.15828863084316253,
      "num_tokens": 7904541.0,
      "step": 3810
    },
    {
      "entropy": 5.705569267272949,
      "epoch": 0.3460631349782293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992634885397118,
      "loss": 5.6798,
      "mean_token_accuracy": 0.1567293658852577,
      "num_tokens": 7915435.0,
      "step": 3815
    },
    {
      "entropy": 5.717358255386353,
      "epoch": 0.3465166908563135,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992608703381396,
      "loss": 5.6525,
      "mean_token_accuracy": 0.15847330242395402,
      "num_tokens": 7925870.0,
      "step": 3820
    },
    {
      "entropy": 5.707509326934814,
      "epoch": 0.3469702467343977,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004992582474987955,
      "loss": 5.6621,
      "mean_token_accuracy": 0.151695054769516,
      "num_tokens": 7937046.0,
      "step": 3825
    },
    {
      "entropy": 5.682017850875854,
      "epoch": 0.34742380261248185,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992556200217337,
      "loss": 5.6083,
      "mean_token_accuracy": 0.15659248381853103,
      "num_tokens": 7947150.0,
      "step": 3830
    },
    {
      "entropy": 5.776818370819091,
      "epoch": 0.34787735849056606,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004992529879070085,
      "loss": 5.6931,
      "mean_token_accuracy": 0.15769558846950532,
      "num_tokens": 7956815.0,
      "step": 3835
    },
    {
      "entropy": 5.83626880645752,
      "epoch": 0.3483309143686502,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992503511546745,
      "loss": 5.743,
      "mean_token_accuracy": 0.15489998310804368,
      "num_tokens": 7968114.0,
      "step": 3840
    },
    {
      "entropy": 5.773981189727783,
      "epoch": 0.3487844702467344,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004992477097647859,
      "loss": 5.6626,
      "mean_token_accuracy": 0.15163280665874482,
      "num_tokens": 7977840.0,
      "step": 3845
    },
    {
      "entropy": 5.754621744155884,
      "epoch": 0.3492380261248186,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992450637373976,
      "loss": 5.7765,
      "mean_token_accuracy": 0.1560756802558899,
      "num_tokens": 7988663.0,
      "step": 3850
    },
    {
      "entropy": 5.654491233825683,
      "epoch": 0.34969158200290273,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004992424130725642,
      "loss": 5.5985,
      "mean_token_accuracy": 0.15517864674329757,
      "num_tokens": 7999693.0,
      "step": 3855
    },
    {
      "entropy": 5.813797044754028,
      "epoch": 0.35014513788098695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992397577703406,
      "loss": 5.7652,
      "mean_token_accuracy": 0.1493635967373848,
      "num_tokens": 8010154.0,
      "step": 3860
    },
    {
      "entropy": 5.859886837005615,
      "epoch": 0.3505986937590711,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004992370978307816,
      "loss": 5.7515,
      "mean_token_accuracy": 0.15075698494911194,
      "num_tokens": 8020921.0,
      "step": 3865
    },
    {
      "entropy": 5.7340294361114506,
      "epoch": 0.3510522496371553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992344332539424,
      "loss": 5.6605,
      "mean_token_accuracy": 0.1617124542593956,
      "num_tokens": 8031404.0,
      "step": 3870
    },
    {
      "entropy": 5.685090446472168,
      "epoch": 0.35150580551523947,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992317640398779,
      "loss": 5.6695,
      "mean_token_accuracy": 0.16097560971975328,
      "num_tokens": 8041310.0,
      "step": 3875
    },
    {
      "entropy": 5.719081544876099,
      "epoch": 0.3519593613933237,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004992290901886434,
      "loss": 5.6399,
      "mean_token_accuracy": 0.15827004760503768,
      "num_tokens": 8051054.0,
      "step": 3880
    },
    {
      "entropy": 5.7040183544158936,
      "epoch": 0.35241291727140783,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499226411700294,
      "loss": 5.6566,
      "mean_token_accuracy": 0.16001638025045395,
      "num_tokens": 8061063.0,
      "step": 3885
    },
    {
      "entropy": 5.758534955978393,
      "epoch": 0.35286647314949204,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004992237285748855,
      "loss": 5.6567,
      "mean_token_accuracy": 0.15772127062082292,
      "num_tokens": 8071715.0,
      "step": 3890
    },
    {
      "entropy": 5.806764888763428,
      "epoch": 0.3533200290275762,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992210408124731,
      "loss": 5.7353,
      "mean_token_accuracy": 0.14987001866102218,
      "num_tokens": 8082679.0,
      "step": 3895
    },
    {
      "entropy": 5.712429571151733,
      "epoch": 0.35377358490566035,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992183484131123,
      "loss": 5.7682,
      "mean_token_accuracy": 0.1527717225253582,
      "num_tokens": 8093785.0,
      "step": 3900
    },
    {
      "entropy": 5.746983051300049,
      "epoch": 0.35422714078374457,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992156513768591,
      "loss": 5.6654,
      "mean_token_accuracy": 0.1591253861784935,
      "num_tokens": 8105822.0,
      "step": 3905
    },
    {
      "entropy": 5.720091438293457,
      "epoch": 0.3546806966618287,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000499212949703769,
      "loss": 5.5759,
      "mean_token_accuracy": 0.1609518736600876,
      "num_tokens": 8115521.0,
      "step": 3910
    },
    {
      "entropy": 5.691662406921386,
      "epoch": 0.35513425253991293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992102433938981,
      "loss": 5.672,
      "mean_token_accuracy": 0.1599155269563198,
      "num_tokens": 8125524.0,
      "step": 3915
    },
    {
      "entropy": 5.70084958076477,
      "epoch": 0.3555878084179971,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004992075324473021,
      "loss": 5.5852,
      "mean_token_accuracy": 0.16763251274824142,
      "num_tokens": 8135552.0,
      "step": 3920
    },
    {
      "entropy": 5.827421236038208,
      "epoch": 0.3560413642960813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004992048168640371,
      "loss": 5.7055,
      "mean_token_accuracy": 0.15760034322738647,
      "num_tokens": 8145617.0,
      "step": 3925
    },
    {
      "entropy": 5.698878335952759,
      "epoch": 0.35649492017416545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992020966441595,
      "loss": 5.7267,
      "mean_token_accuracy": 0.15130315870046615,
      "num_tokens": 8156729.0,
      "step": 3930
    },
    {
      "entropy": 5.716935110092163,
      "epoch": 0.35694847605224966,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991993717877254,
      "loss": 5.6959,
      "mean_token_accuracy": 0.15315268635749818,
      "num_tokens": 8167637.0,
      "step": 3935
    },
    {
      "entropy": 5.789372968673706,
      "epoch": 0.3574020319303338,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004991966422947911,
      "loss": 5.6835,
      "mean_token_accuracy": 0.16331233233213424,
      "num_tokens": 8177354.0,
      "step": 3940
    },
    {
      "entropy": 5.685820865631103,
      "epoch": 0.357855587808418,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004991939081654132,
      "loss": 5.6194,
      "mean_token_accuracy": 0.1660743162035942,
      "num_tokens": 8187580.0,
      "step": 3945
    },
    {
      "entropy": 5.716048860549927,
      "epoch": 0.3583091436865022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499191169399648,
      "loss": 5.7133,
      "mean_token_accuracy": 0.15225303024053574,
      "num_tokens": 8199382.0,
      "step": 3950
    },
    {
      "entropy": 5.73907208442688,
      "epoch": 0.35876269956458634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991884259975525,
      "loss": 5.6285,
      "mean_token_accuracy": 0.15740728080272676,
      "num_tokens": 8208484.0,
      "step": 3955
    },
    {
      "entropy": 5.742117643356323,
      "epoch": 0.35921625544267055,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499185677959183,
      "loss": 5.6756,
      "mean_token_accuracy": 0.15358153730630875,
      "num_tokens": 8219056.0,
      "step": 3960
    },
    {
      "entropy": 5.73494143486023,
      "epoch": 0.3596698113207547,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004991829252845966,
      "loss": 5.686,
      "mean_token_accuracy": 0.152549646794796,
      "num_tokens": 8230149.0,
      "step": 3965
    },
    {
      "entropy": 5.730507802963257,
      "epoch": 0.3601233671988389,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991801679738502,
      "loss": 5.7519,
      "mean_token_accuracy": 0.14736442565917968,
      "num_tokens": 8240803.0,
      "step": 3970
    },
    {
      "entropy": 5.757702302932739,
      "epoch": 0.3605769230769231,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004991774060270008,
      "loss": 5.7139,
      "mean_token_accuracy": 0.14796190112829208,
      "num_tokens": 8251563.0,
      "step": 3975
    },
    {
      "entropy": 5.680777549743652,
      "epoch": 0.3610304789550073,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991746394441055,
      "loss": 5.6508,
      "mean_token_accuracy": 0.15206940174102784,
      "num_tokens": 8261342.0,
      "step": 3980
    },
    {
      "entropy": 5.735341548919678,
      "epoch": 0.36148403483309144,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004991718682252216,
      "loss": 5.7576,
      "mean_token_accuracy": 0.15018298774957656,
      "num_tokens": 8271540.0,
      "step": 3985
    },
    {
      "entropy": 5.823741292953491,
      "epoch": 0.3619375907111756,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004991690923704063,
      "loss": 5.662,
      "mean_token_accuracy": 0.15624243170022964,
      "num_tokens": 8280593.0,
      "step": 3990
    },
    {
      "entropy": 5.720953178405762,
      "epoch": 0.3623911465892598,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004991663118797171,
      "loss": 5.6835,
      "mean_token_accuracy": 0.15814144015312195,
      "num_tokens": 8290126.0,
      "step": 3995
    },
    {
      "entropy": 5.742925214767456,
      "epoch": 0.36284470246734396,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991635267532113,
      "loss": 5.637,
      "mean_token_accuracy": 0.1533143252134323,
      "num_tokens": 8301416.0,
      "step": 4000
    },
    {
      "entropy": 5.761494159698486,
      "epoch": 0.36329825834542817,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991607369909468,
      "loss": 5.6985,
      "mean_token_accuracy": 0.1542692556977272,
      "num_tokens": 8311034.0,
      "step": 4005
    },
    {
      "entropy": 5.755333518981933,
      "epoch": 0.3637518142235123,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991579425929811,
      "loss": 5.7185,
      "mean_token_accuracy": 0.1533808723092079,
      "num_tokens": 8322203.0,
      "step": 4010
    },
    {
      "entropy": 5.734006357192993,
      "epoch": 0.36420537010159654,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000499155143559372,
      "loss": 5.6028,
      "mean_token_accuracy": 0.15972371846437455,
      "num_tokens": 8330981.0,
      "step": 4015
    },
    {
      "entropy": 5.595567846298218,
      "epoch": 0.3646589259796807,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991523398901773,
      "loss": 5.6455,
      "mean_token_accuracy": 0.15632233768701553,
      "num_tokens": 8341294.0,
      "step": 4020
    },
    {
      "entropy": 5.888984680175781,
      "epoch": 0.3651124818577649,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991495315854553,
      "loss": 5.7736,
      "mean_token_accuracy": 0.14876558110117913,
      "num_tokens": 8351824.0,
      "step": 4025
    },
    {
      "entropy": 5.706782484054566,
      "epoch": 0.36556603773584906,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004991467186452637,
      "loss": 5.6474,
      "mean_token_accuracy": 0.1557074546813965,
      "num_tokens": 8361975.0,
      "step": 4030
    },
    {
      "entropy": 5.656030082702637,
      "epoch": 0.3660195936139332,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991439010696609,
      "loss": 5.5978,
      "mean_token_accuracy": 0.16090150475502013,
      "num_tokens": 8372145.0,
      "step": 4035
    },
    {
      "entropy": 5.77086763381958,
      "epoch": 0.3664731494920174,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991410788587051,
      "loss": 5.6861,
      "mean_token_accuracy": 0.1525834932923317,
      "num_tokens": 8382542.0,
      "step": 4040
    },
    {
      "entropy": 5.766021776199341,
      "epoch": 0.3669267053701016,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991382520124547,
      "loss": 5.695,
      "mean_token_accuracy": 0.15353819727897644,
      "num_tokens": 8393450.0,
      "step": 4045
    },
    {
      "entropy": 5.739769554138183,
      "epoch": 0.3673802612481858,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991354205309681,
      "loss": 5.6296,
      "mean_token_accuracy": 0.15330485254526138,
      "num_tokens": 8404929.0,
      "step": 4050
    },
    {
      "entropy": 5.684649848937989,
      "epoch": 0.36783381712626995,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499132584414304,
      "loss": 5.6407,
      "mean_token_accuracy": 0.1600140005350113,
      "num_tokens": 8415703.0,
      "step": 4055
    },
    {
      "entropy": 5.716530799865723,
      "epoch": 0.36828737300435416,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004991297436625207,
      "loss": 5.7259,
      "mean_token_accuracy": 0.15426832735538482,
      "num_tokens": 8426220.0,
      "step": 4060
    },
    {
      "entropy": 5.657416963577271,
      "epoch": 0.3687409288824383,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004991268982756774,
      "loss": 5.619,
      "mean_token_accuracy": 0.1579454243183136,
      "num_tokens": 8436220.0,
      "step": 4065
    },
    {
      "entropy": 5.7623128414154055,
      "epoch": 0.3691944847605225,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004991240482538327,
      "loss": 5.7639,
      "mean_token_accuracy": 0.1493281066417694,
      "num_tokens": 8446724.0,
      "step": 4070
    },
    {
      "entropy": 5.762617349624634,
      "epoch": 0.3696480406386067,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004991211935970455,
      "loss": 5.6143,
      "mean_token_accuracy": 0.15606819838285446,
      "num_tokens": 8457270.0,
      "step": 4075
    },
    {
      "entropy": 5.686959457397461,
      "epoch": 0.37010159651669083,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004991183343053748,
      "loss": 5.6888,
      "mean_token_accuracy": 0.15353127270936967,
      "num_tokens": 8468137.0,
      "step": 4080
    },
    {
      "entropy": 5.741979169845581,
      "epoch": 0.37055515239477504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00049911547037888,
      "loss": 5.6281,
      "mean_token_accuracy": 0.15946190059185028,
      "num_tokens": 8477574.0,
      "step": 4085
    },
    {
      "entropy": 5.722776651382446,
      "epoch": 0.3710087082728592,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004991126018176202,
      "loss": 5.7124,
      "mean_token_accuracy": 0.15095608532428742,
      "num_tokens": 8487268.0,
      "step": 4090
    },
    {
      "entropy": 5.846265125274658,
      "epoch": 0.3714622641509434,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991097286216545,
      "loss": 5.785,
      "mean_token_accuracy": 0.15681469887495042,
      "num_tokens": 8497785.0,
      "step": 4095
    },
    {
      "entropy": 5.764314413070679,
      "epoch": 0.37191582002902757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004991068507910427,
      "loss": 5.637,
      "mean_token_accuracy": 0.1572082445025444,
      "num_tokens": 8508169.0,
      "step": 4100
    },
    {
      "entropy": 5.668509101867675,
      "epoch": 0.3723693759071118,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499103968325844,
      "loss": 5.6267,
      "mean_token_accuracy": 0.15181283801794052,
      "num_tokens": 8519201.0,
      "step": 4105
    },
    {
      "entropy": 5.773807764053345,
      "epoch": 0.37282293178519593,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004991010812261181,
      "loss": 5.7172,
      "mean_token_accuracy": 0.15204160511493683,
      "num_tokens": 8531093.0,
      "step": 4110
    },
    {
      "entropy": 5.656918621063232,
      "epoch": 0.37327648766328014,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990981894919248,
      "loss": 5.5981,
      "mean_token_accuracy": 0.15744011104106903,
      "num_tokens": 8541730.0,
      "step": 4115
    },
    {
      "entropy": 5.677169275283814,
      "epoch": 0.3737300435413643,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990952931233238,
      "loss": 5.6771,
      "mean_token_accuracy": 0.15572711825370789,
      "num_tokens": 8552570.0,
      "step": 4120
    },
    {
      "entropy": 5.765596532821656,
      "epoch": 0.37418359941944845,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000499092392120375,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1613634631037712,
      "num_tokens": 8562340.0,
      "step": 4125
    },
    {
      "entropy": 5.75891342163086,
      "epoch": 0.37463715529753266,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990894864831385,
      "loss": 5.7786,
      "mean_token_accuracy": 0.14843543320894242,
      "num_tokens": 8573240.0,
      "step": 4130
    },
    {
      "entropy": 5.706006050109863,
      "epoch": 0.3750907111756168,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004990865762116743,
      "loss": 5.6169,
      "mean_token_accuracy": 0.16231092363595961,
      "num_tokens": 8583194.0,
      "step": 4135
    },
    {
      "entropy": 5.76061520576477,
      "epoch": 0.37554426705370103,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990836613060426,
      "loss": 5.6887,
      "mean_token_accuracy": 0.15307701751589775,
      "num_tokens": 8593603.0,
      "step": 4140
    },
    {
      "entropy": 5.7104448795318605,
      "epoch": 0.3759978229317852,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004990807417663036,
      "loss": 5.6975,
      "mean_token_accuracy": 0.15583582669496537,
      "num_tokens": 8603176.0,
      "step": 4145
    },
    {
      "entropy": 5.770629453659057,
      "epoch": 0.3764513788098694,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990778175925179,
      "loss": 5.7169,
      "mean_token_accuracy": 0.15750074684619902,
      "num_tokens": 8614801.0,
      "step": 4150
    },
    {
      "entropy": 5.818771934509277,
      "epoch": 0.37690493468795355,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990748887847457,
      "loss": 5.7293,
      "mean_token_accuracy": 0.15358437448740006,
      "num_tokens": 8624999.0,
      "step": 4155
    },
    {
      "entropy": 5.697445487976074,
      "epoch": 0.37735849056603776,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990719553430477,
      "loss": 5.5989,
      "mean_token_accuracy": 0.16195421367883683,
      "num_tokens": 8635358.0,
      "step": 4160
    },
    {
      "entropy": 5.751416826248169,
      "epoch": 0.3778120464441219,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004990690172674847,
      "loss": 5.6686,
      "mean_token_accuracy": 0.1510378047823906,
      "num_tokens": 8645430.0,
      "step": 4165
    },
    {
      "entropy": 5.774804925918579,
      "epoch": 0.3782656023222061,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990660745581172,
      "loss": 5.6322,
      "mean_token_accuracy": 0.16380622684955598,
      "num_tokens": 8655115.0,
      "step": 4170
    },
    {
      "entropy": 5.644221448898316,
      "epoch": 0.3787191582002903,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990631272150062,
      "loss": 5.6125,
      "mean_token_accuracy": 0.1576675742864609,
      "num_tokens": 8665391.0,
      "step": 4175
    },
    {
      "entropy": 5.724270439147949,
      "epoch": 0.37917271407837444,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990601752382127,
      "loss": 5.722,
      "mean_token_accuracy": 0.1531612053513527,
      "num_tokens": 8676819.0,
      "step": 4180
    },
    {
      "entropy": 5.810973405838013,
      "epoch": 0.37962626995645865,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004990572186277975,
      "loss": 5.6377,
      "mean_token_accuracy": 0.16196503043174743,
      "num_tokens": 8687052.0,
      "step": 4185
    },
    {
      "entropy": 5.741652250289917,
      "epoch": 0.3800798258345428,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990542573838221,
      "loss": 5.7152,
      "mean_token_accuracy": 0.14744316637516022,
      "num_tokens": 8697756.0,
      "step": 4190
    },
    {
      "entropy": 5.6437897205352785,
      "epoch": 0.380533381712627,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004990512915063475,
      "loss": 5.6266,
      "mean_token_accuracy": 0.16152960807085037,
      "num_tokens": 8708388.0,
      "step": 4195
    },
    {
      "entropy": 5.822674703598023,
      "epoch": 0.38098693759071117,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004990483209954352,
      "loss": 5.736,
      "mean_token_accuracy": 0.14884860962629318,
      "num_tokens": 8719475.0,
      "step": 4200
    },
    {
      "entropy": 5.752110195159912,
      "epoch": 0.3814404934687954,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990453458511464,
      "loss": 5.5806,
      "mean_token_accuracy": 0.16833294183015823,
      "num_tokens": 8729395.0,
      "step": 4205
    },
    {
      "entropy": 5.708733367919922,
      "epoch": 0.38189404934687954,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004990423660735428,
      "loss": 5.6509,
      "mean_token_accuracy": 0.15839240252971648,
      "num_tokens": 8740351.0,
      "step": 4210
    },
    {
      "entropy": 5.7155640602111815,
      "epoch": 0.3823476052249637,
      "grad_norm": 1.328125,
      "learning_rate": 0.000499039381662686,
      "loss": 5.6512,
      "mean_token_accuracy": 0.16292448043823243,
      "num_tokens": 8749260.0,
      "step": 4215
    },
    {
      "entropy": 5.792791223526001,
      "epoch": 0.3828011611030479,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004990363926186377,
      "loss": 5.7406,
      "mean_token_accuracy": 0.1454516589641571,
      "num_tokens": 8759604.0,
      "step": 4220
    },
    {
      "entropy": 5.777758598327637,
      "epoch": 0.38325471698113206,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004990333989414597,
      "loss": 5.7626,
      "mean_token_accuracy": 0.15234508663415908,
      "num_tokens": 8769815.0,
      "step": 4225
    },
    {
      "entropy": 5.728119850158691,
      "epoch": 0.38370827285921627,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990304006312138,
      "loss": 5.6821,
      "mean_token_accuracy": 0.1551062658429146,
      "num_tokens": 8780557.0,
      "step": 4230
    },
    {
      "entropy": 5.733680391311646,
      "epoch": 0.3841618287373004,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004990273976879622,
      "loss": 5.6469,
      "mean_token_accuracy": 0.1641158416867256,
      "num_tokens": 8790993.0,
      "step": 4235
    },
    {
      "entropy": 5.759084987640381,
      "epoch": 0.38461538461538464,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990243901117669,
      "loss": 5.6933,
      "mean_token_accuracy": 0.15187538415193558,
      "num_tokens": 8800657.0,
      "step": 4240
    },
    {
      "entropy": 5.6477912902832035,
      "epoch": 0.3850689404934688,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990213779026903,
      "loss": 5.6959,
      "mean_token_accuracy": 0.1512046605348587,
      "num_tokens": 8812983.0,
      "step": 4245
    },
    {
      "entropy": 5.851211357116699,
      "epoch": 0.385522496371553,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990183610607944,
      "loss": 5.6557,
      "mean_token_accuracy": 0.1610846310853958,
      "num_tokens": 8822980.0,
      "step": 4250
    },
    {
      "entropy": 5.713478326797485,
      "epoch": 0.38597605224963716,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004990153395861415,
      "loss": 5.6438,
      "mean_token_accuracy": 0.161063851416111,
      "num_tokens": 8833821.0,
      "step": 4255
    },
    {
      "entropy": 5.584740400314331,
      "epoch": 0.3864296081277213,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004990123134787946,
      "loss": 5.5909,
      "mean_token_accuracy": 0.15470565110445023,
      "num_tokens": 8843929.0,
      "step": 4260
    },
    {
      "entropy": 5.697726917266846,
      "epoch": 0.3868831640058055,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004990092827388159,
      "loss": 5.5817,
      "mean_token_accuracy": 0.1604609251022339,
      "num_tokens": 8854180.0,
      "step": 4265
    },
    {
      "entropy": 5.631396293640137,
      "epoch": 0.3873367198838897,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004990062473662681,
      "loss": 5.5968,
      "mean_token_accuracy": 0.16301485896110535,
      "num_tokens": 8865018.0,
      "step": 4270
    },
    {
      "entropy": 5.728056001663208,
      "epoch": 0.3877902757619739,
      "grad_norm": 1.234375,
      "learning_rate": 0.000499003207361214,
      "loss": 5.7851,
      "mean_token_accuracy": 0.1480627566576004,
      "num_tokens": 8875423.0,
      "step": 4275
    },
    {
      "entropy": 5.879735326766967,
      "epoch": 0.38824383164005805,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990001627237165,
      "loss": 5.7072,
      "mean_token_accuracy": 0.15677697360515594,
      "num_tokens": 8887850.0,
      "step": 4280
    },
    {
      "entropy": 5.671970367431641,
      "epoch": 0.38869738751814226,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004989971134538386,
      "loss": 5.614,
      "mean_token_accuracy": 0.15710368007421494,
      "num_tokens": 8897429.0,
      "step": 4285
    },
    {
      "entropy": 5.6937713623046875,
      "epoch": 0.3891509433962264,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004989940595516432,
      "loss": 5.6128,
      "mean_token_accuracy": 0.15821060836315154,
      "num_tokens": 8907048.0,
      "step": 4290
    },
    {
      "entropy": 5.72869644165039,
      "epoch": 0.3896044992743106,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004989910010171938,
      "loss": 5.688,
      "mean_token_accuracy": 0.15163719207048415,
      "num_tokens": 8917987.0,
      "step": 4295
    },
    {
      "entropy": 5.723460674285889,
      "epoch": 0.3900580551523948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989879378505532,
      "loss": 5.587,
      "mean_token_accuracy": 0.15756893455982207,
      "num_tokens": 8929339.0,
      "step": 4300
    },
    {
      "entropy": 5.622553491592408,
      "epoch": 0.39051161103047893,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004989848700517849,
      "loss": 5.598,
      "mean_token_accuracy": 0.155763378739357,
      "num_tokens": 8939473.0,
      "step": 4305
    },
    {
      "entropy": 5.663455677032471,
      "epoch": 0.39096516690856314,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004989817976209526,
      "loss": 5.62,
      "mean_token_accuracy": 0.1591288760304451,
      "num_tokens": 8949173.0,
      "step": 4310
    },
    {
      "entropy": 5.7296106815338135,
      "epoch": 0.3914187227866473,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004989787205581196,
      "loss": 5.6071,
      "mean_token_accuracy": 0.1606694459915161,
      "num_tokens": 8959879.0,
      "step": 4315
    },
    {
      "entropy": 5.717441272735596,
      "epoch": 0.3918722786647315,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004989756388633496,
      "loss": 5.6899,
      "mean_token_accuracy": 0.15467706620693206,
      "num_tokens": 8970643.0,
      "step": 4320
    },
    {
      "entropy": 5.686373376846314,
      "epoch": 0.39232583454281567,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004989725525367063,
      "loss": 5.5633,
      "mean_token_accuracy": 0.1609751895070076,
      "num_tokens": 8979882.0,
      "step": 4325
    },
    {
      "entropy": 5.7686194896698,
      "epoch": 0.3927793904208999,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004989694615782534,
      "loss": 5.6911,
      "mean_token_accuracy": 0.15228137522935867,
      "num_tokens": 8990925.0,
      "step": 4330
    },
    {
      "entropy": 5.746078300476074,
      "epoch": 0.39323294629898403,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004989663659880552,
      "loss": 5.7046,
      "mean_token_accuracy": 0.1520710125565529,
      "num_tokens": 9001365.0,
      "step": 4335
    },
    {
      "entropy": 5.773279905319214,
      "epoch": 0.3936865021770682,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004989632657661754,
      "loss": 5.7065,
      "mean_token_accuracy": 0.15934449434280396,
      "num_tokens": 9011942.0,
      "step": 4340
    },
    {
      "entropy": 5.776633024215698,
      "epoch": 0.3941400580551524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000498960160912678,
      "loss": 5.7333,
      "mean_token_accuracy": 0.15434861183166504,
      "num_tokens": 9022610.0,
      "step": 4345
    },
    {
      "entropy": 5.718952655792236,
      "epoch": 0.39459361393323655,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004989570514276276,
      "loss": 5.621,
      "mean_token_accuracy": 0.15780188590288163,
      "num_tokens": 9033034.0,
      "step": 4350
    },
    {
      "entropy": 5.699564695358276,
      "epoch": 0.39504716981132076,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004989539373110883,
      "loss": 5.6721,
      "mean_token_accuracy": 0.15199518352746963,
      "num_tokens": 9043418.0,
      "step": 4355
    },
    {
      "entropy": 5.812909030914307,
      "epoch": 0.3955007256894049,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004989508185631245,
      "loss": 5.7588,
      "mean_token_accuracy": 0.1541099339723587,
      "num_tokens": 9052708.0,
      "step": 4360
    },
    {
      "entropy": 5.673247146606445,
      "epoch": 0.39595428156748913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004989476951838006,
      "loss": 5.6199,
      "mean_token_accuracy": 0.1546905130147934,
      "num_tokens": 9062883.0,
      "step": 4365
    },
    {
      "entropy": 5.740717458724975,
      "epoch": 0.3964078374455733,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004989445671731814,
      "loss": 5.5977,
      "mean_token_accuracy": 0.1623426467180252,
      "num_tokens": 9072830.0,
      "step": 4370
    },
    {
      "entropy": 5.605792236328125,
      "epoch": 0.3968613933236575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989414345313315,
      "loss": 5.5413,
      "mean_token_accuracy": 0.15586361438035964,
      "num_tokens": 9082749.0,
      "step": 4375
    },
    {
      "entropy": 5.7155884265899655,
      "epoch": 0.39731494920174165,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989382972583156,
      "loss": 5.6329,
      "mean_token_accuracy": 0.15358513742685317,
      "num_tokens": 9094188.0,
      "step": 4380
    },
    {
      "entropy": 5.6907202243804935,
      "epoch": 0.3977685050798258,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004989351553541986,
      "loss": 5.6623,
      "mean_token_accuracy": 0.15161869078874587,
      "num_tokens": 9105023.0,
      "step": 4385
    },
    {
      "entropy": 5.616463232040405,
      "epoch": 0.39822206095791,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004989320088190456,
      "loss": 5.611,
      "mean_token_accuracy": 0.15799364000558852,
      "num_tokens": 9115563.0,
      "step": 4390
    },
    {
      "entropy": 5.691041946411133,
      "epoch": 0.3986756168359942,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989288576529216,
      "loss": 5.6362,
      "mean_token_accuracy": 0.15247485041618347,
      "num_tokens": 9125394.0,
      "step": 4395
    },
    {
      "entropy": 5.73917670249939,
      "epoch": 0.3991291727140784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989257018558917,
      "loss": 5.6292,
      "mean_token_accuracy": 0.15650943666696548,
      "num_tokens": 9137061.0,
      "step": 4400
    },
    {
      "entropy": 5.713806009292602,
      "epoch": 0.39958272859216254,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004989225414280213,
      "loss": 5.5985,
      "mean_token_accuracy": 0.1589561104774475,
      "num_tokens": 9148196.0,
      "step": 4405
    },
    {
      "entropy": 5.646256589889527,
      "epoch": 0.40003628447024675,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004989193763693756,
      "loss": 5.604,
      "mean_token_accuracy": 0.16273730546236037,
      "num_tokens": 9158013.0,
      "step": 4410
    },
    {
      "entropy": 5.649022006988526,
      "epoch": 0.4004898403483309,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004989162066800202,
      "loss": 5.5375,
      "mean_token_accuracy": 0.15886689722537994,
      "num_tokens": 9167813.0,
      "step": 4415
    },
    {
      "entropy": 5.624588108062744,
      "epoch": 0.4009433962264151,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004989130323600205,
      "loss": 5.6008,
      "mean_token_accuracy": 0.16580986827611924,
      "num_tokens": 9177783.0,
      "step": 4420
    },
    {
      "entropy": 5.660892677307129,
      "epoch": 0.40139695210449927,
      "grad_norm": 1.25,
      "learning_rate": 0.0004989098534094424,
      "loss": 5.6514,
      "mean_token_accuracy": 0.1579426884651184,
      "num_tokens": 9187556.0,
      "step": 4425
    },
    {
      "entropy": 5.729719161987305,
      "epoch": 0.4018505079825834,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004989066698283512,
      "loss": 5.7008,
      "mean_token_accuracy": 0.15310224890708923,
      "num_tokens": 9198659.0,
      "step": 4430
    },
    {
      "entropy": 5.774732494354248,
      "epoch": 0.40230406386066764,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004989034816168132,
      "loss": 5.6671,
      "mean_token_accuracy": 0.16033934354782103,
      "num_tokens": 9209546.0,
      "step": 4435
    },
    {
      "entropy": 5.666764545440674,
      "epoch": 0.4027576197387518,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004989002887748941,
      "loss": 5.5867,
      "mean_token_accuracy": 0.16027882397174836,
      "num_tokens": 9219339.0,
      "step": 4440
    },
    {
      "entropy": 5.718165111541748,
      "epoch": 0.403211175616836,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004988970913026601,
      "loss": 5.7698,
      "mean_token_accuracy": 0.15176644548773766,
      "num_tokens": 9230364.0,
      "step": 4445
    },
    {
      "entropy": 5.744258975982666,
      "epoch": 0.40366473149492016,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004988938892001771,
      "loss": 5.6168,
      "mean_token_accuracy": 0.1577327385544777,
      "num_tokens": 9240732.0,
      "step": 4450
    },
    {
      "entropy": 5.651133108139038,
      "epoch": 0.40411828737300437,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004988906824675113,
      "loss": 5.5603,
      "mean_token_accuracy": 0.1604924499988556,
      "num_tokens": 9251049.0,
      "step": 4455
    },
    {
      "entropy": 5.6878711700439455,
      "epoch": 0.4045718432510885,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004988874711047293,
      "loss": 5.6033,
      "mean_token_accuracy": 0.16183407455682755,
      "num_tokens": 9261660.0,
      "step": 4460
    },
    {
      "entropy": 5.727019739151001,
      "epoch": 0.40502539912917274,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004988842551118972,
      "loss": 5.6668,
      "mean_token_accuracy": 0.16216951757669448,
      "num_tokens": 9272114.0,
      "step": 4465
    },
    {
      "entropy": 5.788266038894653,
      "epoch": 0.4054789550072569,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004988810344890818,
      "loss": 5.6439,
      "mean_token_accuracy": 0.15519319474697113,
      "num_tokens": 9282125.0,
      "step": 4470
    },
    {
      "entropy": 5.740830755233764,
      "epoch": 0.40593251088534105,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004988778092363494,
      "loss": 5.6799,
      "mean_token_accuracy": 0.15653618425130844,
      "num_tokens": 9294078.0,
      "step": 4475
    },
    {
      "entropy": 5.690979671478272,
      "epoch": 0.40638606676342526,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004988745793537671,
      "loss": 5.6334,
      "mean_token_accuracy": 0.15408705770969391,
      "num_tokens": 9304026.0,
      "step": 4480
    },
    {
      "entropy": 5.729645156860352,
      "epoch": 0.4068396226415094,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004988713448414011,
      "loss": 5.6865,
      "mean_token_accuracy": 0.16053041964769363,
      "num_tokens": 9313933.0,
      "step": 4485
    },
    {
      "entropy": 5.719425439834595,
      "epoch": 0.4072931785195936,
      "grad_norm": 1.296875,
      "learning_rate": 0.000498868105699319,
      "loss": 5.6998,
      "mean_token_accuracy": 0.15748196542263032,
      "num_tokens": 9323932.0,
      "step": 4490
    },
    {
      "entropy": 5.604436445236206,
      "epoch": 0.4077467343976778,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004988648619275872,
      "loss": 5.6043,
      "mean_token_accuracy": 0.171552175283432,
      "num_tokens": 9333467.0,
      "step": 4495
    },
    {
      "entropy": 5.745215559005738,
      "epoch": 0.408200290275762,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004988616135262729,
      "loss": 5.6731,
      "mean_token_accuracy": 0.15075961947441102,
      "num_tokens": 9344909.0,
      "step": 4500
    },
    {
      "entropy": 5.866862964630127,
      "epoch": 0.40865384615384615,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004988583604954436,
      "loss": 5.7127,
      "mean_token_accuracy": 0.152850903570652,
      "num_tokens": 9355630.0,
      "step": 4505
    },
    {
      "entropy": 5.665213871002197,
      "epoch": 0.40910740203193036,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004988551028351663,
      "loss": 5.6624,
      "mean_token_accuracy": 0.15456879287958145,
      "num_tokens": 9365700.0,
      "step": 4510
    },
    {
      "entropy": 5.628423118591309,
      "epoch": 0.4095609579100145,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004988518405455084,
      "loss": 5.745,
      "mean_token_accuracy": 0.1484757512807846,
      "num_tokens": 9377088.0,
      "step": 4515
    },
    {
      "entropy": 5.7334541320800785,
      "epoch": 0.41001451378809867,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004988485736265374,
      "loss": 5.5316,
      "mean_token_accuracy": 0.16351988762617112,
      "num_tokens": 9386855.0,
      "step": 4520
    },
    {
      "entropy": 5.731301927566529,
      "epoch": 0.4104680696661829,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000498845302078321,
      "loss": 5.6519,
      "mean_token_accuracy": 0.15784264355897903,
      "num_tokens": 9398661.0,
      "step": 4525
    },
    {
      "entropy": 5.735857725143433,
      "epoch": 0.41092162554426703,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004988420259009266,
      "loss": 5.6382,
      "mean_token_accuracy": 0.1608929917216301,
      "num_tokens": 9408670.0,
      "step": 4530
    },
    {
      "entropy": 5.7451990127563475,
      "epoch": 0.41137518142235124,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000498838745094422,
      "loss": 5.6583,
      "mean_token_accuracy": 0.15440730154514312,
      "num_tokens": 9418740.0,
      "step": 4535
    },
    {
      "entropy": 5.706992340087891,
      "epoch": 0.4118287373004354,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004988354596588751,
      "loss": 5.6468,
      "mean_token_accuracy": 0.15828332751989366,
      "num_tokens": 9429339.0,
      "step": 4540
    },
    {
      "entropy": 5.724893856048584,
      "epoch": 0.4122822931785196,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004988321695943539,
      "loss": 5.6011,
      "mean_token_accuracy": 0.16610953062772751,
      "num_tokens": 9438996.0,
      "step": 4545
    },
    {
      "entropy": 5.666243076324463,
      "epoch": 0.41273584905660377,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004988288749009264,
      "loss": 5.6138,
      "mean_token_accuracy": 0.16006203293800353,
      "num_tokens": 9448669.0,
      "step": 4550
    },
    {
      "entropy": 5.654767560958862,
      "epoch": 0.413189404934688,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004988255755786608,
      "loss": 5.5987,
      "mean_token_accuracy": 0.16067843437194823,
      "num_tokens": 9458857.0,
      "step": 4555
    },
    {
      "entropy": 5.69340295791626,
      "epoch": 0.41364296081277213,
      "grad_norm": 1.25,
      "learning_rate": 0.0004988222716276251,
      "loss": 5.5957,
      "mean_token_accuracy": 0.16045788377523423,
      "num_tokens": 9470241.0,
      "step": 4560
    },
    {
      "entropy": 5.792037296295166,
      "epoch": 0.4140965166908563,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498818963047888,
      "loss": 5.7455,
      "mean_token_accuracy": 0.1541075199842453,
      "num_tokens": 9480528.0,
      "step": 4565
    },
    {
      "entropy": 5.792665910720825,
      "epoch": 0.4145500725689405,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004988156498395176,
      "loss": 5.6614,
      "mean_token_accuracy": 0.15965667963027955,
      "num_tokens": 9491000.0,
      "step": 4570
    },
    {
      "entropy": 5.717759847640991,
      "epoch": 0.41500362844702465,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004988123320025825,
      "loss": 5.6374,
      "mean_token_accuracy": 0.16161510795354844,
      "num_tokens": 9501612.0,
      "step": 4575
    },
    {
      "entropy": 5.71904559135437,
      "epoch": 0.41545718432510886,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004988090095371514,
      "loss": 5.7505,
      "mean_token_accuracy": 0.15754791498184204,
      "num_tokens": 9513999.0,
      "step": 4580
    },
    {
      "entropy": 5.719426965713501,
      "epoch": 0.415910740203193,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004988056824432928,
      "loss": 5.6116,
      "mean_token_accuracy": 0.15234914422035217,
      "num_tokens": 9524392.0,
      "step": 4585
    },
    {
      "entropy": 5.748761129379273,
      "epoch": 0.41636429608127723,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004988023507210758,
      "loss": 5.6438,
      "mean_token_accuracy": 0.15767379999160766,
      "num_tokens": 9534227.0,
      "step": 4590
    },
    {
      "entropy": 5.687965345382691,
      "epoch": 0.4168178519593614,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004987990143705692,
      "loss": 5.6305,
      "mean_token_accuracy": 0.16447580456733704,
      "num_tokens": 9544242.0,
      "step": 4595
    },
    {
      "entropy": 5.6869677066802975,
      "epoch": 0.4172714078374456,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004987956733918418,
      "loss": 5.5491,
      "mean_token_accuracy": 0.15838005393743515,
      "num_tokens": 9553921.0,
      "step": 4600
    },
    {
      "entropy": 5.648387908935547,
      "epoch": 0.41772496371552975,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004987923277849629,
      "loss": 5.6058,
      "mean_token_accuracy": 0.1557765543460846,
      "num_tokens": 9564768.0,
      "step": 4605
    },
    {
      "entropy": 5.73125262260437,
      "epoch": 0.4181785195936139,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004987889775500017,
      "loss": 5.6643,
      "mean_token_accuracy": 0.15236683785915375,
      "num_tokens": 9576360.0,
      "step": 4610
    },
    {
      "entropy": 5.753176498413086,
      "epoch": 0.4186320754716981,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004987856226870272,
      "loss": 5.6521,
      "mean_token_accuracy": 0.15639690160751343,
      "num_tokens": 9587623.0,
      "step": 4615
    },
    {
      "entropy": 5.70543532371521,
      "epoch": 0.4190856313497823,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004987822631961092,
      "loss": 5.6943,
      "mean_token_accuracy": 0.15374598652124405,
      "num_tokens": 9597837.0,
      "step": 4620
    },
    {
      "entropy": 5.706055736541748,
      "epoch": 0.4195391872278665,
      "grad_norm": 1.25,
      "learning_rate": 0.000498778899077317,
      "loss": 5.6236,
      "mean_token_accuracy": 0.15881578475236893,
      "num_tokens": 9609509.0,
      "step": 4625
    },
    {
      "entropy": 5.7174797534942625,
      "epoch": 0.41999274310595064,
      "grad_norm": 1.59375,
      "learning_rate": 0.00049877553033072,
      "loss": 5.6172,
      "mean_token_accuracy": 0.15826885551214218,
      "num_tokens": 9619954.0,
      "step": 4630
    },
    {
      "entropy": 5.679054498672485,
      "epoch": 0.42044629898403485,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004987721569563882,
      "loss": 5.6179,
      "mean_token_accuracy": 0.15558764338493347,
      "num_tokens": 9629742.0,
      "step": 4635
    },
    {
      "entropy": 5.683841419219971,
      "epoch": 0.420899854862119,
      "grad_norm": 1.5,
      "learning_rate": 0.000498768778954391,
      "loss": 5.6418,
      "mean_token_accuracy": 0.16237951517105104,
      "num_tokens": 9641378.0,
      "step": 4640
    },
    {
      "entropy": 5.763059043884278,
      "epoch": 0.4213534107402032,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004987653963247985,
      "loss": 5.6998,
      "mean_token_accuracy": 0.16142413020133972,
      "num_tokens": 9652279.0,
      "step": 4645
    },
    {
      "entropy": 5.770412397384644,
      "epoch": 0.42180696661828737,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004987620090676805,
      "loss": 5.7089,
      "mean_token_accuracy": 0.1528452843427658,
      "num_tokens": 9663416.0,
      "step": 4650
    },
    {
      "entropy": 5.7631096839904785,
      "epoch": 0.4222605224963715,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004987586171831072,
      "loss": 5.662,
      "mean_token_accuracy": 0.15435225814580916,
      "num_tokens": 9674647.0,
      "step": 4655
    },
    {
      "entropy": 5.690637016296387,
      "epoch": 0.42271407837445574,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004987552206711487,
      "loss": 5.515,
      "mean_token_accuracy": 0.15983799695968628,
      "num_tokens": 9685084.0,
      "step": 4660
    },
    {
      "entropy": 5.581456327438355,
      "epoch": 0.4231676342525399,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004987518195318752,
      "loss": 5.5734,
      "mean_token_accuracy": 0.15758123695850373,
      "num_tokens": 9695661.0,
      "step": 4665
    },
    {
      "entropy": 5.736875104904175,
      "epoch": 0.4236211901306241,
      "grad_norm": 1.296875,
      "learning_rate": 0.000498748413765357,
      "loss": 5.6232,
      "mean_token_accuracy": 0.15394429713487626,
      "num_tokens": 9705773.0,
      "step": 4670
    },
    {
      "entropy": 5.76428861618042,
      "epoch": 0.42407474600870826,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004987450033716646,
      "loss": 5.6723,
      "mean_token_accuracy": 0.15749599486589433,
      "num_tokens": 9715799.0,
      "step": 4675
    },
    {
      "entropy": 5.718335390090942,
      "epoch": 0.42452830188679247,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004987415883508686,
      "loss": 5.6476,
      "mean_token_accuracy": 0.16066358387470245,
      "num_tokens": 9727012.0,
      "step": 4680
    },
    {
      "entropy": 5.74974217414856,
      "epoch": 0.4249818577648766,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004987381687030394,
      "loss": 5.6351,
      "mean_token_accuracy": 0.15791599750518798,
      "num_tokens": 9737044.0,
      "step": 4685
    },
    {
      "entropy": 5.817522478103638,
      "epoch": 0.42543541364296084,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004987347444282479,
      "loss": 5.7391,
      "mean_token_accuracy": 0.14861048609018326,
      "num_tokens": 9747648.0,
      "step": 4690
    },
    {
      "entropy": 5.656645727157593,
      "epoch": 0.425888969521045,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004987313155265649,
      "loss": 5.637,
      "mean_token_accuracy": 0.15445875972509385,
      "num_tokens": 9757812.0,
      "step": 4695
    },
    {
      "entropy": 5.77165412902832,
      "epoch": 0.42634252539912915,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004987278819980613,
      "loss": 5.6382,
      "mean_token_accuracy": 0.15808448195457458,
      "num_tokens": 9768337.0,
      "step": 4700
    },
    {
      "entropy": 5.789901208877564,
      "epoch": 0.42679608127721336,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004987244438428081,
      "loss": 5.6806,
      "mean_token_accuracy": 0.14943017363548278,
      "num_tokens": 9779433.0,
      "step": 4705
    },
    {
      "entropy": 5.658951282501221,
      "epoch": 0.4272496371552975,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004987210010608764,
      "loss": 5.5923,
      "mean_token_accuracy": 0.1585081249475479,
      "num_tokens": 9790069.0,
      "step": 4710
    },
    {
      "entropy": 5.642023754119873,
      "epoch": 0.4277031930333817,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004987175536523373,
      "loss": 5.5616,
      "mean_token_accuracy": 0.16363361924886705,
      "num_tokens": 9799669.0,
      "step": 4715
    },
    {
      "entropy": 5.639230585098266,
      "epoch": 0.4281567489114659,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004987141016172622,
      "loss": 5.5899,
      "mean_token_accuracy": 0.16178745478391648,
      "num_tokens": 9809615.0,
      "step": 4720
    },
    {
      "entropy": 5.680355024337769,
      "epoch": 0.4286103047895501,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004987106449557225,
      "loss": 5.5722,
      "mean_token_accuracy": 0.15792733579874038,
      "num_tokens": 9819666.0,
      "step": 4725
    },
    {
      "entropy": 5.779202842712403,
      "epoch": 0.42906386066763424,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004987071836677896,
      "loss": 5.7204,
      "mean_token_accuracy": 0.15531140714883804,
      "num_tokens": 9831015.0,
      "step": 4730
    },
    {
      "entropy": 5.685993623733521,
      "epoch": 0.42951741654571846,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004987037177535353,
      "loss": 5.597,
      "mean_token_accuracy": 0.15545632392168046,
      "num_tokens": 9841354.0,
      "step": 4735
    },
    {
      "entropy": 5.646900415420532,
      "epoch": 0.4299709724238026,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004987002472130309,
      "loss": 5.5875,
      "mean_token_accuracy": 0.153756807744503,
      "num_tokens": 9850467.0,
      "step": 4740
    },
    {
      "entropy": 5.784751653671265,
      "epoch": 0.43042452830188677,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004986967720463485,
      "loss": 5.6621,
      "mean_token_accuracy": 0.1529627487063408,
      "num_tokens": 9861115.0,
      "step": 4745
    },
    {
      "entropy": 5.768982076644898,
      "epoch": 0.430878084179971,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004986932922535598,
      "loss": 5.6608,
      "mean_token_accuracy": 0.15878958255052567,
      "num_tokens": 9871172.0,
      "step": 4750
    },
    {
      "entropy": 5.647015523910523,
      "epoch": 0.43133164005805513,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004986898078347368,
      "loss": 5.5618,
      "mean_token_accuracy": 0.15951561331748962,
      "num_tokens": 9881997.0,
      "step": 4755
    },
    {
      "entropy": 5.6569263458251955,
      "epoch": 0.43178519593613934,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004986863187899516,
      "loss": 5.5653,
      "mean_token_accuracy": 0.16232846826314926,
      "num_tokens": 9892310.0,
      "step": 4760
    },
    {
      "entropy": 5.64468412399292,
      "epoch": 0.4322387518142235,
      "grad_norm": 1.625,
      "learning_rate": 0.0004986828251192764,
      "loss": 5.6226,
      "mean_token_accuracy": 0.15712934285402297,
      "num_tokens": 9902664.0,
      "step": 4765
    },
    {
      "entropy": 5.7097142219543455,
      "epoch": 0.4326923076923077,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004986793268227833,
      "loss": 5.5963,
      "mean_token_accuracy": 0.1599283143877983,
      "num_tokens": 9912491.0,
      "step": 4770
    },
    {
      "entropy": 5.671255397796631,
      "epoch": 0.43314586357039186,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004986758239005448,
      "loss": 5.5598,
      "mean_token_accuracy": 0.16305184960365296,
      "num_tokens": 9922057.0,
      "step": 4775
    },
    {
      "entropy": 5.632063341140747,
      "epoch": 0.4335994194484761,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004986723163526331,
      "loss": 5.5983,
      "mean_token_accuracy": 0.15621958673000336,
      "num_tokens": 9933661.0,
      "step": 4780
    },
    {
      "entropy": 5.657237720489502,
      "epoch": 0.43405297532656023,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004986688041791209,
      "loss": 5.5713,
      "mean_token_accuracy": 0.15300980508327483,
      "num_tokens": 9945177.0,
      "step": 4785
    },
    {
      "entropy": 5.710802459716797,
      "epoch": 0.4345065312046444,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000498665287380081,
      "loss": 5.6062,
      "mean_token_accuracy": 0.15459759831428527,
      "num_tokens": 9956226.0,
      "step": 4790
    },
    {
      "entropy": 5.701862812042236,
      "epoch": 0.4349600870827286,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004986617659555859,
      "loss": 5.6691,
      "mean_token_accuracy": 0.15917366743087769,
      "num_tokens": 9965877.0,
      "step": 4795
    },
    {
      "entropy": 5.690611982345581,
      "epoch": 0.43541364296081275,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004986582399057085,
      "loss": 5.6244,
      "mean_token_accuracy": 0.1667877346277237,
      "num_tokens": 9975592.0,
      "step": 4800
    },
    {
      "entropy": 5.775518417358398,
      "epoch": 0.43586719883889696,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004986547092305216,
      "loss": 5.65,
      "mean_token_accuracy": 0.15770362168550492,
      "num_tokens": 9986094.0,
      "step": 4805
    },
    {
      "entropy": 5.735955429077149,
      "epoch": 0.4363207547169811,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004986511739300984,
      "loss": 5.6806,
      "mean_token_accuracy": 0.1539325311779976,
      "num_tokens": 9997241.0,
      "step": 4810
    },
    {
      "entropy": 5.640730094909668,
      "epoch": 0.43677431059506533,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004986476340045119,
      "loss": 5.597,
      "mean_token_accuracy": 0.15729670077562333,
      "num_tokens": 10008091.0,
      "step": 4815
    },
    {
      "entropy": 5.661537456512451,
      "epoch": 0.4372278664731495,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004986440894538354,
      "loss": 5.6726,
      "mean_token_accuracy": 0.15475862249732017,
      "num_tokens": 10018794.0,
      "step": 4820
    },
    {
      "entropy": 5.731957864761353,
      "epoch": 0.4376814223512337,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004986405402781421,
      "loss": 5.579,
      "mean_token_accuracy": 0.1584927648305893,
      "num_tokens": 10028414.0,
      "step": 4825
    },
    {
      "entropy": 5.750116586685181,
      "epoch": 0.43813497822931785,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004986369864775054,
      "loss": 5.6496,
      "mean_token_accuracy": 0.15727464854717255,
      "num_tokens": 10038749.0,
      "step": 4830
    },
    {
      "entropy": 5.720384407043457,
      "epoch": 0.438588534107402,
      "grad_norm": 1.25,
      "learning_rate": 0.0004986334280519989,
      "loss": 5.604,
      "mean_token_accuracy": 0.1561151698231697,
      "num_tokens": 10049866.0,
      "step": 4835
    },
    {
      "entropy": 5.676847696304321,
      "epoch": 0.4390420899854862,
      "grad_norm": 1.59375,
      "learning_rate": 0.000498629865001696,
      "loss": 5.6276,
      "mean_token_accuracy": 0.16333256363868714,
      "num_tokens": 10060680.0,
      "step": 4840
    },
    {
      "entropy": 5.776250123977661,
      "epoch": 0.43949564586357037,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004986262973266706,
      "loss": 5.7157,
      "mean_token_accuracy": 0.1555226355791092,
      "num_tokens": 10070931.0,
      "step": 4845
    },
    {
      "entropy": 5.693648958206177,
      "epoch": 0.4399492017416546,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004986227250269964,
      "loss": 5.6633,
      "mean_token_accuracy": 0.1558353751897812,
      "num_tokens": 10080189.0,
      "step": 4850
    },
    {
      "entropy": 5.716169309616089,
      "epoch": 0.44040275761973874,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004986191481027472,
      "loss": 5.5571,
      "mean_token_accuracy": 0.1630585163831711,
      "num_tokens": 10088813.0,
      "step": 4855
    },
    {
      "entropy": 5.719408226013184,
      "epoch": 0.44085631349782295,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004986155665539971,
      "loss": 5.5975,
      "mean_token_accuracy": 0.15667377263307572,
      "num_tokens": 10098125.0,
      "step": 4860
    },
    {
      "entropy": 5.71316180229187,
      "epoch": 0.4413098693759071,
      "grad_norm": 1.328125,
      "learning_rate": 0.00049861198038082,
      "loss": 5.6213,
      "mean_token_accuracy": 0.15492214560508727,
      "num_tokens": 10109125.0,
      "step": 4865
    },
    {
      "entropy": 5.656546688079834,
      "epoch": 0.4417634252539913,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004986083895832903,
      "loss": 5.5859,
      "mean_token_accuracy": 0.15925385057926178,
      "num_tokens": 10120182.0,
      "step": 4870
    },
    {
      "entropy": 5.681136751174927,
      "epoch": 0.44221698113207547,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004986047941614821,
      "loss": 5.6628,
      "mean_token_accuracy": 0.15356798693537713,
      "num_tokens": 10130941.0,
      "step": 4875
    },
    {
      "entropy": 5.768857765197754,
      "epoch": 0.4426705370101596,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004986011941154696,
      "loss": 5.6395,
      "mean_token_accuracy": 0.15759876668453215,
      "num_tokens": 10141708.0,
      "step": 4880
    },
    {
      "entropy": 5.660665369033813,
      "epoch": 0.44312409288824384,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004985975894453275,
      "loss": 5.5811,
      "mean_token_accuracy": 0.1579940661787987,
      "num_tokens": 10151816.0,
      "step": 4885
    },
    {
      "entropy": 5.685306549072266,
      "epoch": 0.443577648766328,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004985939801511304,
      "loss": 5.6903,
      "mean_token_accuracy": 0.15800520777702332,
      "num_tokens": 10162470.0,
      "step": 4890
    },
    {
      "entropy": 5.720489025115967,
      "epoch": 0.4440312046444122,
      "grad_norm": 1.25,
      "learning_rate": 0.0004985903662329527,
      "loss": 5.5777,
      "mean_token_accuracy": 0.1637508600950241,
      "num_tokens": 10173171.0,
      "step": 4895
    },
    {
      "entropy": 5.6964959621429445,
      "epoch": 0.44448476052249636,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004985867476908693,
      "loss": 5.6135,
      "mean_token_accuracy": 0.16020630449056625,
      "num_tokens": 10184185.0,
      "step": 4900
    },
    {
      "entropy": 5.72540864944458,
      "epoch": 0.44493831640058057,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004985831245249549,
      "loss": 5.6767,
      "mean_token_accuracy": 0.15526049435138703,
      "num_tokens": 10195516.0,
      "step": 4905
    },
    {
      "entropy": 5.693022298812866,
      "epoch": 0.4453918722786647,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004985794967352846,
      "loss": 5.6153,
      "mean_token_accuracy": 0.1589369460940361,
      "num_tokens": 10204936.0,
      "step": 4910
    },
    {
      "entropy": 5.758884763717651,
      "epoch": 0.44584542815674894,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004985758643219334,
      "loss": 5.6387,
      "mean_token_accuracy": 0.15786018073558808,
      "num_tokens": 10216218.0,
      "step": 4915
    },
    {
      "entropy": 5.800872135162353,
      "epoch": 0.4462989840348331,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004985722272849762,
      "loss": 5.6395,
      "mean_token_accuracy": 0.15207151025533677,
      "num_tokens": 10226370.0,
      "step": 4920
    },
    {
      "entropy": 5.575708436965942,
      "epoch": 0.44675253991291725,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004985685856244884,
      "loss": 5.5611,
      "mean_token_accuracy": 0.16749200075864792,
      "num_tokens": 10236170.0,
      "step": 4925
    },
    {
      "entropy": 5.689274263381958,
      "epoch": 0.44720609579100146,
      "grad_norm": 1.25,
      "learning_rate": 0.0004985649393405453,
      "loss": 5.6399,
      "mean_token_accuracy": 0.15733460038900376,
      "num_tokens": 10246119.0,
      "step": 4930
    },
    {
      "entropy": 5.70348482131958,
      "epoch": 0.4476596516690856,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004985612884332223,
      "loss": 5.6505,
      "mean_token_accuracy": 0.15208822041749953,
      "num_tokens": 10256762.0,
      "step": 4935
    },
    {
      "entropy": 5.65702223777771,
      "epoch": 0.4481132075471698,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004985576329025949,
      "loss": 5.5674,
      "mean_token_accuracy": 0.16472743451595306,
      "num_tokens": 10267317.0,
      "step": 4940
    },
    {
      "entropy": 5.779310035705566,
      "epoch": 0.448566763425254,
      "grad_norm": 1.25,
      "learning_rate": 0.0004985539727487385,
      "loss": 5.744,
      "mean_token_accuracy": 0.14801670163869857,
      "num_tokens": 10279111.0,
      "step": 4945
    },
    {
      "entropy": 5.705371475219726,
      "epoch": 0.4490203193033382,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004985503079717292,
      "loss": 5.5892,
      "mean_token_accuracy": 0.16035750806331633,
      "num_tokens": 10289383.0,
      "step": 4950
    },
    {
      "entropy": 5.721146535873413,
      "epoch": 0.44947387518142234,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004985466385716425,
      "loss": 5.5991,
      "mean_token_accuracy": 0.16319389194250106,
      "num_tokens": 10299806.0,
      "step": 4955
    },
    {
      "entropy": 5.655878496170044,
      "epoch": 0.44992743105950656,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004985429645485543,
      "loss": 5.6265,
      "mean_token_accuracy": 0.15566550940275192,
      "num_tokens": 10310348.0,
      "step": 4960
    },
    {
      "entropy": 5.761146450042725,
      "epoch": 0.4503809869375907,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004985392859025407,
      "loss": 5.6387,
      "mean_token_accuracy": 0.16084989458322524,
      "num_tokens": 10320354.0,
      "step": 4965
    },
    {
      "entropy": 5.67796630859375,
      "epoch": 0.45083454281567487,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004985356026336776,
      "loss": 5.5461,
      "mean_token_accuracy": 0.1607133001089096,
      "num_tokens": 10330292.0,
      "step": 4970
    },
    {
      "entropy": 5.622106838226318,
      "epoch": 0.4512880986937591,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004985319147420414,
      "loss": 5.5092,
      "mean_token_accuracy": 0.16529798954725267,
      "num_tokens": 10340449.0,
      "step": 4975
    },
    {
      "entropy": 5.677884149551391,
      "epoch": 0.45174165457184323,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004985282222277082,
      "loss": 5.586,
      "mean_token_accuracy": 0.15833908766508104,
      "num_tokens": 10350637.0,
      "step": 4980
    },
    {
      "entropy": 5.6070215702056885,
      "epoch": 0.45219521044992744,
      "grad_norm": 1.5,
      "learning_rate": 0.0004985245250907544,
      "loss": 5.5264,
      "mean_token_accuracy": 0.16214152425527573,
      "num_tokens": 10361548.0,
      "step": 4985
    },
    {
      "entropy": 5.6438604354858395,
      "epoch": 0.4526487663280116,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004985208233312564,
      "loss": 5.5343,
      "mean_token_accuracy": 0.15428854525089264,
      "num_tokens": 10371765.0,
      "step": 4990
    },
    {
      "entropy": 5.7256732940673825,
      "epoch": 0.4531023222060958,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004985171169492908,
      "loss": 5.6525,
      "mean_token_accuracy": 0.16112247854471207,
      "num_tokens": 10382108.0,
      "step": 4995
    },
    {
      "entropy": 5.667687511444091,
      "epoch": 0.45355587808417996,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004985134059449344,
      "loss": 5.5963,
      "mean_token_accuracy": 0.16138881593942642,
      "num_tokens": 10392390.0,
      "step": 5000
    },
    {
      "entropy": 5.645096778869629,
      "epoch": 0.4540094339622642,
      "grad_norm": 1.5,
      "learning_rate": 0.0004985096903182637,
      "loss": 5.6477,
      "mean_token_accuracy": 0.16399622708559036,
      "num_tokens": 10402094.0,
      "step": 5005
    },
    {
      "entropy": 5.71685528755188,
      "epoch": 0.45446298984034833,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004985059700693557,
      "loss": 5.601,
      "mean_token_accuracy": 0.15897753685712815,
      "num_tokens": 10413350.0,
      "step": 5010
    },
    {
      "entropy": 5.757471227645874,
      "epoch": 0.4549165457184325,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004985022451982872,
      "loss": 5.7296,
      "mean_token_accuracy": 0.14736578613519669,
      "num_tokens": 10424292.0,
      "step": 5015
    },
    {
      "entropy": 5.805013990402221,
      "epoch": 0.4553701015965167,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004984985157051354,
      "loss": 5.5862,
      "mean_token_accuracy": 0.15852102637290955,
      "num_tokens": 10433724.0,
      "step": 5020
    },
    {
      "entropy": 5.681409645080566,
      "epoch": 0.45582365747460085,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004984947815899774,
      "loss": 5.6568,
      "mean_token_accuracy": 0.15463263988494874,
      "num_tokens": 10444724.0,
      "step": 5025
    },
    {
      "entropy": 5.689796447753906,
      "epoch": 0.45627721335268506,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984910428528902,
      "loss": 5.5847,
      "mean_token_accuracy": 0.16503604203462602,
      "num_tokens": 10455086.0,
      "step": 5030
    },
    {
      "entropy": 5.637588691711426,
      "epoch": 0.4567307692307692,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004984872994939514,
      "loss": 5.534,
      "mean_token_accuracy": 0.1577824205160141,
      "num_tokens": 10465076.0,
      "step": 5035
    },
    {
      "entropy": 5.698536729812622,
      "epoch": 0.45718432510885343,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984835515132382,
      "loss": 5.5831,
      "mean_token_accuracy": 0.16402017027139665,
      "num_tokens": 10474817.0,
      "step": 5040
    },
    {
      "entropy": 5.690934658050537,
      "epoch": 0.4576378809869376,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004984797989108282,
      "loss": 5.6017,
      "mean_token_accuracy": 0.15358709692955017,
      "num_tokens": 10485735.0,
      "step": 5045
    },
    {
      "entropy": 5.629631233215332,
      "epoch": 0.4580914368650218,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984760416867991,
      "loss": 5.6895,
      "mean_token_accuracy": 0.1563015177845955,
      "num_tokens": 10496731.0,
      "step": 5050
    },
    {
      "entropy": 5.68785285949707,
      "epoch": 0.45854499274310595,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004984722798412286,
      "loss": 5.5529,
      "mean_token_accuracy": 0.1587032288312912,
      "num_tokens": 10506787.0,
      "step": 5055
    },
    {
      "entropy": 5.730269145965576,
      "epoch": 0.4589985486211901,
      "grad_norm": 1.25,
      "learning_rate": 0.0004984685133741941,
      "loss": 5.6188,
      "mean_token_accuracy": 0.16047224551439285,
      "num_tokens": 10517170.0,
      "step": 5060
    },
    {
      "entropy": 5.651216459274292,
      "epoch": 0.4594521044992743,
      "grad_norm": 1.390625,
      "learning_rate": 0.000498464742285774,
      "loss": 5.6124,
      "mean_token_accuracy": 0.15768254101276397,
      "num_tokens": 10527342.0,
      "step": 5065
    },
    {
      "entropy": 5.672451400756836,
      "epoch": 0.45990566037735847,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000498460966576046,
      "loss": 5.5809,
      "mean_token_accuracy": 0.1630593493580818,
      "num_tokens": 10537253.0,
      "step": 5070
    },
    {
      "entropy": 5.677005481719971,
      "epoch": 0.4603592162554427,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004984571862450884,
      "loss": 5.6169,
      "mean_token_accuracy": 0.160668021440506,
      "num_tokens": 10547249.0,
      "step": 5075
    },
    {
      "entropy": 5.756367826461792,
      "epoch": 0.46081277213352684,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004984534012929791,
      "loss": 5.643,
      "mean_token_accuracy": 0.1614480659365654,
      "num_tokens": 10558174.0,
      "step": 5080
    },
    {
      "entropy": 5.753311634063721,
      "epoch": 0.46126632801161105,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004984496117197967,
      "loss": 5.6044,
      "mean_token_accuracy": 0.16208672225475312,
      "num_tokens": 10567610.0,
      "step": 5085
    },
    {
      "entropy": 5.646350717544555,
      "epoch": 0.4617198838896952,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004984458175256193,
      "loss": 5.5971,
      "mean_token_accuracy": 0.15640690922737122,
      "num_tokens": 10578494.0,
      "step": 5090
    },
    {
      "entropy": 5.7145490646362305,
      "epoch": 0.4621734397677794,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004984420187105253,
      "loss": 5.709,
      "mean_token_accuracy": 0.1579895183444023,
      "num_tokens": 10589482.0,
      "step": 5095
    },
    {
      "entropy": 5.78198881149292,
      "epoch": 0.46262699564586357,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004984382152745936,
      "loss": 5.6597,
      "mean_token_accuracy": 0.15277581810951232,
      "num_tokens": 10599965.0,
      "step": 5100
    },
    {
      "entropy": 5.71567816734314,
      "epoch": 0.4630805515239477,
      "grad_norm": 1.25,
      "learning_rate": 0.0004984344072179026,
      "loss": 5.6952,
      "mean_token_accuracy": 0.15894951820373535,
      "num_tokens": 10610500.0,
      "step": 5105
    },
    {
      "entropy": 5.731478691101074,
      "epoch": 0.46353410740203194,
      "grad_norm": 1.5,
      "learning_rate": 0.0004984305945405312,
      "loss": 5.6397,
      "mean_token_accuracy": 0.15854529589414595,
      "num_tokens": 10619843.0,
      "step": 5110
    },
    {
      "entropy": 5.73313536643982,
      "epoch": 0.4639876632801161,
      "grad_norm": 1.328125,
      "learning_rate": 0.000498426777242558,
      "loss": 5.5447,
      "mean_token_accuracy": 0.16565094739198685,
      "num_tokens": 10631026.0,
      "step": 5115
    },
    {
      "entropy": 5.546000003814697,
      "epoch": 0.4644412191582003,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004984229553240623,
      "loss": 5.5721,
      "mean_token_accuracy": 0.16317055374383926,
      "num_tokens": 10641130.0,
      "step": 5120
    },
    {
      "entropy": 5.704706859588623,
      "epoch": 0.46489477503628446,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984191287851228,
      "loss": 5.6195,
      "mean_token_accuracy": 0.15852320194244385,
      "num_tokens": 10652305.0,
      "step": 5125
    },
    {
      "entropy": 5.781041049957276,
      "epoch": 0.46534833091436867,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984152976258188,
      "loss": 5.6487,
      "mean_token_accuracy": 0.16050119698047638,
      "num_tokens": 10663083.0,
      "step": 5130
    },
    {
      "entropy": 5.711477136611938,
      "epoch": 0.4658018867924528,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004984114618462296,
      "loss": 5.6963,
      "mean_token_accuracy": 0.15250758677721024,
      "num_tokens": 10673602.0,
      "step": 5135
    },
    {
      "entropy": 5.701936101913452,
      "epoch": 0.46625544267053703,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004984076214464343,
      "loss": 5.5087,
      "mean_token_accuracy": 0.1663011431694031,
      "num_tokens": 10684622.0,
      "step": 5140
    },
    {
      "entropy": 5.676464653015136,
      "epoch": 0.4667089985486212,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004984037764265126,
      "loss": 5.6355,
      "mean_token_accuracy": 0.15928329527378082,
      "num_tokens": 10695129.0,
      "step": 5145
    },
    {
      "entropy": 5.592328977584839,
      "epoch": 0.46716255442670535,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004983999267865438,
      "loss": 5.4749,
      "mean_token_accuracy": 0.1723786011338234,
      "num_tokens": 10705236.0,
      "step": 5150
    },
    {
      "entropy": 5.653973770141602,
      "epoch": 0.46761611030478956,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004983960725266078,
      "loss": 5.6002,
      "mean_token_accuracy": 0.16147854626178743,
      "num_tokens": 10715748.0,
      "step": 5155
    },
    {
      "entropy": 5.631850051879883,
      "epoch": 0.4680696661828737,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004983922136467838,
      "loss": 5.5327,
      "mean_token_accuracy": 0.16219267398118972,
      "num_tokens": 10725799.0,
      "step": 5160
    },
    {
      "entropy": 5.688706350326538,
      "epoch": 0.4685232220609579,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000498388350147152,
      "loss": 5.6346,
      "mean_token_accuracy": 0.15641454756259918,
      "num_tokens": 10735668.0,
      "step": 5165
    },
    {
      "entropy": 5.798784685134888,
      "epoch": 0.4689767779390421,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004983844820277922,
      "loss": 5.7451,
      "mean_token_accuracy": 0.15428698062896729,
      "num_tokens": 10746466.0,
      "step": 5170
    },
    {
      "entropy": 5.826989364624024,
      "epoch": 0.4694303338171263,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004983806092887843,
      "loss": 5.6568,
      "mean_token_accuracy": 0.1553424969315529,
      "num_tokens": 10756724.0,
      "step": 5175
    },
    {
      "entropy": 5.670391082763672,
      "epoch": 0.46988388969521044,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004983767319302086,
      "loss": 5.6984,
      "mean_token_accuracy": 0.15869035348296165,
      "num_tokens": 10766919.0,
      "step": 5180
    },
    {
      "entropy": 5.573981809616089,
      "epoch": 0.47033744557329465,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004983728499521451,
      "loss": 5.4832,
      "mean_token_accuracy": 0.16845543682575226,
      "num_tokens": 10777550.0,
      "step": 5185
    },
    {
      "entropy": 5.641195344924927,
      "epoch": 0.4707910014513788,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004983689633546741,
      "loss": 5.5391,
      "mean_token_accuracy": 0.15530664175748826,
      "num_tokens": 10788456.0,
      "step": 5190
    },
    {
      "entropy": 5.667837238311767,
      "epoch": 0.47124455732946297,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004983650721378761,
      "loss": 5.5794,
      "mean_token_accuracy": 0.15722571164369584,
      "num_tokens": 10800049.0,
      "step": 5195
    },
    {
      "entropy": 5.690346145629883,
      "epoch": 0.4716981132075472,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004983611763018314,
      "loss": 5.6121,
      "mean_token_accuracy": 0.15779779106378555,
      "num_tokens": 10810743.0,
      "step": 5200
    },
    {
      "entropy": 5.8374152183532715,
      "epoch": 0.47215166908563133,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004983572758466208,
      "loss": 5.7976,
      "mean_token_accuracy": 0.15054364502429962,
      "num_tokens": 10821897.0,
      "step": 5205
    },
    {
      "entropy": 5.762923002243042,
      "epoch": 0.47260522496371554,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004983533707723247,
      "loss": 5.6686,
      "mean_token_accuracy": 0.15647778511047364,
      "num_tokens": 10831589.0,
      "step": 5210
    },
    {
      "entropy": 5.68774905204773,
      "epoch": 0.4730587808417997,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004983494610790241,
      "loss": 5.5639,
      "mean_token_accuracy": 0.15626207888126373,
      "num_tokens": 10841404.0,
      "step": 5215
    },
    {
      "entropy": 5.646328639984131,
      "epoch": 0.4735123367198839,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004983455467667996,
      "loss": 5.6074,
      "mean_token_accuracy": 0.15539631098508835,
      "num_tokens": 10851577.0,
      "step": 5220
    },
    {
      "entropy": 5.65619740486145,
      "epoch": 0.47396589259796806,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004983416278357322,
      "loss": 5.5374,
      "mean_token_accuracy": 0.16633753776550292,
      "num_tokens": 10861403.0,
      "step": 5225
    },
    {
      "entropy": 5.6497701644897464,
      "epoch": 0.4744194484760523,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004983377042859032,
      "loss": 5.5732,
      "mean_token_accuracy": 0.1568841502070427,
      "num_tokens": 10871933.0,
      "step": 5230
    },
    {
      "entropy": 5.693173456192016,
      "epoch": 0.47487300435413643,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004983337761173936,
      "loss": 5.6452,
      "mean_token_accuracy": 0.1571093514561653,
      "num_tokens": 10881800.0,
      "step": 5235
    },
    {
      "entropy": 5.7815196990966795,
      "epoch": 0.4753265602322206,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004983298433302843,
      "loss": 5.6001,
      "mean_token_accuracy": 0.16028908640146255,
      "num_tokens": 10891830.0,
      "step": 5240
    },
    {
      "entropy": 5.634613370895385,
      "epoch": 0.4757801161103048,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004983259059246572,
      "loss": 5.5268,
      "mean_token_accuracy": 0.16579733341932296,
      "num_tokens": 10901557.0,
      "step": 5245
    },
    {
      "entropy": 5.5637064456939695,
      "epoch": 0.47623367198838895,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004983219639005934,
      "loss": 5.6438,
      "mean_token_accuracy": 0.1479102648794651,
      "num_tokens": 10912155.0,
      "step": 5250
    },
    {
      "entropy": 5.665174531936645,
      "epoch": 0.47668722786647316,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004983180172581745,
      "loss": 5.5462,
      "mean_token_accuracy": 0.16128475666046144,
      "num_tokens": 10922291.0,
      "step": 5255
    },
    {
      "entropy": 5.754207944869995,
      "epoch": 0.4771407837445573,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004983140659974819,
      "loss": 5.6172,
      "mean_token_accuracy": 0.15364252179861068,
      "num_tokens": 10933176.0,
      "step": 5260
    },
    {
      "entropy": 5.72802848815918,
      "epoch": 0.47759433962264153,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004983101101185977,
      "loss": 5.7016,
      "mean_token_accuracy": 0.15314582884311675,
      "num_tokens": 10944049.0,
      "step": 5265
    },
    {
      "entropy": 5.698485708236694,
      "epoch": 0.4780478955007257,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004983061496216035,
      "loss": 5.5636,
      "mean_token_accuracy": 0.16470168977975846,
      "num_tokens": 10954347.0,
      "step": 5270
    },
    {
      "entropy": 5.628724670410156,
      "epoch": 0.4785014513788099,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004983021845065812,
      "loss": 5.609,
      "mean_token_accuracy": 0.16007599085569382,
      "num_tokens": 10965123.0,
      "step": 5275
    },
    {
      "entropy": 5.607919788360595,
      "epoch": 0.47895500725689405,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004982982147736128,
      "loss": 5.5918,
      "mean_token_accuracy": 0.16313560605049132,
      "num_tokens": 10976395.0,
      "step": 5280
    },
    {
      "entropy": 5.689217376708984,
      "epoch": 0.4794085631349782,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004982942404227805,
      "loss": 5.5793,
      "mean_token_accuracy": 0.1574721962213516,
      "num_tokens": 10986797.0,
      "step": 5285
    },
    {
      "entropy": 5.7448052883148195,
      "epoch": 0.4798621190130624,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004982902614541662,
      "loss": 5.6325,
      "mean_token_accuracy": 0.15480213314294816,
      "num_tokens": 10997229.0,
      "step": 5290
    },
    {
      "entropy": 5.632512140274048,
      "epoch": 0.48031567489114657,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004982862778678526,
      "loss": 5.5496,
      "mean_token_accuracy": 0.16841159462928773,
      "num_tokens": 11008180.0,
      "step": 5295
    },
    {
      "entropy": 5.69216685295105,
      "epoch": 0.4807692307692308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004982822896639218,
      "loss": 5.5897,
      "mean_token_accuracy": 0.16379621475934983,
      "num_tokens": 11019248.0,
      "step": 5300
    },
    {
      "entropy": 5.631191778182983,
      "epoch": 0.48122278664731494,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004982782968424563,
      "loss": 5.5139,
      "mean_token_accuracy": 0.16136889904737473,
      "num_tokens": 11029760.0,
      "step": 5305
    },
    {
      "entropy": 5.595670461654663,
      "epoch": 0.48167634252539915,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004982742994035388,
      "loss": 5.5861,
      "mean_token_accuracy": 0.1590326741337776,
      "num_tokens": 11039211.0,
      "step": 5310
    },
    {
      "entropy": 5.633118486404419,
      "epoch": 0.4821298984034833,
      "grad_norm": 1.375,
      "learning_rate": 0.0004982702973472518,
      "loss": 5.5684,
      "mean_token_accuracy": 0.1575867161154747,
      "num_tokens": 11049892.0,
      "step": 5315
    },
    {
      "entropy": 5.752400588989258,
      "epoch": 0.4825834542815675,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004982662906736783,
      "loss": 5.6219,
      "mean_token_accuracy": 0.15781162232160567,
      "num_tokens": 11060823.0,
      "step": 5320
    },
    {
      "entropy": 5.606657981872559,
      "epoch": 0.48303701015965167,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004982622793829009,
      "loss": 5.5955,
      "mean_token_accuracy": 0.15996954441070557,
      "num_tokens": 11072353.0,
      "step": 5325
    },
    {
      "entropy": 5.680554294586182,
      "epoch": 0.4834905660377358,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004982582634750027,
      "loss": 5.6099,
      "mean_token_accuracy": 0.15869979858398436,
      "num_tokens": 11082655.0,
      "step": 5330
    },
    {
      "entropy": 5.679764032363892,
      "epoch": 0.48394412191582004,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004982542429500667,
      "loss": 5.614,
      "mean_token_accuracy": 0.16280234158039092,
      "num_tokens": 11093315.0,
      "step": 5335
    },
    {
      "entropy": 5.555749750137329,
      "epoch": 0.4843976777939042,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004982502178081761,
      "loss": 5.5235,
      "mean_token_accuracy": 0.16128282994031906,
      "num_tokens": 11103074.0,
      "step": 5340
    },
    {
      "entropy": 5.700632524490357,
      "epoch": 0.4848512336719884,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004982461880494141,
      "loss": 5.5623,
      "mean_token_accuracy": 0.16219093054533004,
      "num_tokens": 11113636.0,
      "step": 5345
    },
    {
      "entropy": 5.656824111938477,
      "epoch": 0.48530478955007256,
      "grad_norm": 1.25,
      "learning_rate": 0.000498242153673864,
      "loss": 5.5104,
      "mean_token_accuracy": 0.16639752238988875,
      "num_tokens": 11123509.0,
      "step": 5350
    },
    {
      "entropy": 5.632119607925415,
      "epoch": 0.48575834542815677,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004982381146816094,
      "loss": 5.5862,
      "mean_token_accuracy": 0.1564563274383545,
      "num_tokens": 11133405.0,
      "step": 5355
    },
    {
      "entropy": 5.721227788925171,
      "epoch": 0.4862119013062409,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004982340710727336,
      "loss": 5.6453,
      "mean_token_accuracy": 0.15045734122395515,
      "num_tokens": 11144725.0,
      "step": 5360
    },
    {
      "entropy": 5.666970729827881,
      "epoch": 0.48666545718432513,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004982300228473203,
      "loss": 5.5482,
      "mean_token_accuracy": 0.1582345709204674,
      "num_tokens": 11155446.0,
      "step": 5365
    },
    {
      "entropy": 5.663780879974365,
      "epoch": 0.4871190130624093,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004982259700054533,
      "loss": 5.6098,
      "mean_token_accuracy": 0.16389473974704744,
      "num_tokens": 11165380.0,
      "step": 5370
    },
    {
      "entropy": 5.712706756591797,
      "epoch": 0.48757256894049344,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004982219125472163,
      "loss": 5.6538,
      "mean_token_accuracy": 0.1558239132165909,
      "num_tokens": 11175641.0,
      "step": 5375
    },
    {
      "entropy": 5.7976789474487305,
      "epoch": 0.48802612481857766,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004982178504726933,
      "loss": 5.8096,
      "mean_token_accuracy": 0.1573277808725834,
      "num_tokens": 11186257.0,
      "step": 5380
    },
    {
      "entropy": 5.710126447677612,
      "epoch": 0.4884796806966618,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004982137837819682,
      "loss": 5.6183,
      "mean_token_accuracy": 0.1532593734562397,
      "num_tokens": 11197284.0,
      "step": 5385
    },
    {
      "entropy": 5.758053302764893,
      "epoch": 0.488933236574746,
      "grad_norm": 1.375,
      "learning_rate": 0.0004982097124751252,
      "loss": 5.596,
      "mean_token_accuracy": 0.16171942353248597,
      "num_tokens": 11208413.0,
      "step": 5390
    },
    {
      "entropy": 5.615036535263061,
      "epoch": 0.4893867924528302,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004982056365522486,
      "loss": 5.6043,
      "mean_token_accuracy": 0.16109529733657837,
      "num_tokens": 11219016.0,
      "step": 5395
    },
    {
      "entropy": 5.637114524841309,
      "epoch": 0.4898403483309144,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004982015560134225,
      "loss": 5.5702,
      "mean_token_accuracy": 0.1623409003019333,
      "num_tokens": 11228934.0,
      "step": 5400
    },
    {
      "entropy": 5.671698760986328,
      "epoch": 0.49029390420899854,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004981974708587312,
      "loss": 5.5081,
      "mean_token_accuracy": 0.16479768455028534,
      "num_tokens": 11239596.0,
      "step": 5405
    },
    {
      "entropy": 5.5890604019165036,
      "epoch": 0.49074746008708275,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004981933810882594,
      "loss": 5.4798,
      "mean_token_accuracy": 0.1588309571146965,
      "num_tokens": 11249560.0,
      "step": 5410
    },
    {
      "entropy": 5.677786207199096,
      "epoch": 0.4912010159651669,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004981892867020917,
      "loss": 5.6842,
      "mean_token_accuracy": 0.15366458296775817,
      "num_tokens": 11259787.0,
      "step": 5415
    },
    {
      "entropy": 5.762712717056274,
      "epoch": 0.49165457184325106,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004981851877003125,
      "loss": 5.6203,
      "mean_token_accuracy": 0.16309960931539536,
      "num_tokens": 11269717.0,
      "step": 5420
    },
    {
      "entropy": 5.639627838134766,
      "epoch": 0.4921081277213353,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004981810840830068,
      "loss": 5.6055,
      "mean_token_accuracy": 0.16716673821210862,
      "num_tokens": 11279863.0,
      "step": 5425
    },
    {
      "entropy": 5.685359144210816,
      "epoch": 0.49256168359941943,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004981769758502596,
      "loss": 5.7219,
      "mean_token_accuracy": 0.14935771152377128,
      "num_tokens": 11291161.0,
      "step": 5430
    },
    {
      "entropy": 5.661555337905884,
      "epoch": 0.49301523947750364,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981728630021555,
      "loss": 5.5338,
      "mean_token_accuracy": 0.161973038315773,
      "num_tokens": 11301151.0,
      "step": 5435
    },
    {
      "entropy": 5.631797122955322,
      "epoch": 0.4934687953555878,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004981687455387797,
      "loss": 5.5399,
      "mean_token_accuracy": 0.1609097644686699,
      "num_tokens": 11311466.0,
      "step": 5440
    },
    {
      "entropy": 5.667022657394409,
      "epoch": 0.493922351233672,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981646234602173,
      "loss": 5.5048,
      "mean_token_accuracy": 0.16592205464839935,
      "num_tokens": 11321522.0,
      "step": 5445
    },
    {
      "entropy": 5.712038326263428,
      "epoch": 0.49437590711175616,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004981604967665539,
      "loss": 5.6251,
      "mean_token_accuracy": 0.1631213590502739,
      "num_tokens": 11332317.0,
      "step": 5450
    },
    {
      "entropy": 5.616385984420776,
      "epoch": 0.4948294629898404,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004981563654578743,
      "loss": 5.5578,
      "mean_token_accuracy": 0.15613383799791336,
      "num_tokens": 11343024.0,
      "step": 5455
    },
    {
      "entropy": 5.598169612884521,
      "epoch": 0.49528301886792453,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004981522295342643,
      "loss": 5.5237,
      "mean_token_accuracy": 0.1637219548225403,
      "num_tokens": 11353114.0,
      "step": 5460
    },
    {
      "entropy": 5.679735422134399,
      "epoch": 0.4957365747460087,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004981480889958092,
      "loss": 5.5801,
      "mean_token_accuracy": 0.16203076988458634,
      "num_tokens": 11363446.0,
      "step": 5465
    },
    {
      "entropy": 5.638913345336914,
      "epoch": 0.4961901306240929,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004981439438425949,
      "loss": 5.5839,
      "mean_token_accuracy": 0.16449467986822128,
      "num_tokens": 11373413.0,
      "step": 5470
    },
    {
      "entropy": 5.596277523040771,
      "epoch": 0.49664368650217705,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000498139794074707,
      "loss": 5.4583,
      "mean_token_accuracy": 0.16666698157787324,
      "num_tokens": 11383142.0,
      "step": 5475
    },
    {
      "entropy": 5.6968464851379395,
      "epoch": 0.49709724238026126,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004981356396922313,
      "loss": 5.6029,
      "mean_token_accuracy": 0.16057899594306946,
      "num_tokens": 11394635.0,
      "step": 5480
    },
    {
      "entropy": 5.642231130599976,
      "epoch": 0.4975507982583454,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004981314806952535,
      "loss": 5.5018,
      "mean_token_accuracy": 0.16755957901477814,
      "num_tokens": 11404442.0,
      "step": 5485
    },
    {
      "entropy": 5.574909734725952,
      "epoch": 0.49800435413642963,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00049812731708386,
      "loss": 5.4602,
      "mean_token_accuracy": 0.16344375908374786,
      "num_tokens": 11414217.0,
      "step": 5490
    },
    {
      "entropy": 5.713902139663697,
      "epoch": 0.4984579100145138,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004981231488581367,
      "loss": 5.6039,
      "mean_token_accuracy": 0.1595131129026413,
      "num_tokens": 11423834.0,
      "step": 5495
    },
    {
      "entropy": 5.563067150115967,
      "epoch": 0.498911465892598,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004981189760181699,
      "loss": 5.479,
      "mean_token_accuracy": 0.16150647401809692,
      "num_tokens": 11433613.0,
      "step": 5500
    },
    {
      "entropy": 5.617917442321778,
      "epoch": 0.49936502177068215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004981147985640456,
      "loss": 5.5508,
      "mean_token_accuracy": 0.1645784094929695,
      "num_tokens": 11443915.0,
      "step": 5505
    },
    {
      "entropy": 5.798919534683227,
      "epoch": 0.4998185776487663,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004981106164958506,
      "loss": 5.5962,
      "mean_token_accuracy": 0.1555358052253723,
      "num_tokens": 11454588.0,
      "step": 5510
    },
    {
      "entropy": 5.619657421112061,
      "epoch": 0.5002721335268505,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004981064298136712,
      "loss": 5.6519,
      "mean_token_accuracy": 0.15784326046705247,
      "num_tokens": 11465533.0,
      "step": 5515
    },
    {
      "entropy": 5.677310466766357,
      "epoch": 0.5007256894049347,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004981022385175939,
      "loss": 5.5938,
      "mean_token_accuracy": 0.1527741253376007,
      "num_tokens": 11475776.0,
      "step": 5520
    },
    {
      "entropy": 5.728420829772949,
      "epoch": 0.5011792452830188,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004980980426077054,
      "loss": 5.567,
      "mean_token_accuracy": 0.16829029023647307,
      "num_tokens": 11486200.0,
      "step": 5525
    },
    {
      "entropy": 5.639609622955322,
      "epoch": 0.5016328011611031,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004980938420840926,
      "loss": 5.5837,
      "mean_token_accuracy": 0.1582867532968521,
      "num_tokens": 11496768.0,
      "step": 5530
    },
    {
      "entropy": 5.645894479751587,
      "epoch": 0.5020863570391872,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004980896369468423,
      "loss": 5.5549,
      "mean_token_accuracy": 0.16483985483646393,
      "num_tokens": 11507096.0,
      "step": 5535
    },
    {
      "entropy": 5.664897155761719,
      "epoch": 0.5025399129172714,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004980854271960415,
      "loss": 5.6059,
      "mean_token_accuracy": 0.15945658534765245,
      "num_tokens": 11517301.0,
      "step": 5540
    },
    {
      "entropy": 5.730061626434326,
      "epoch": 0.5029934687953556,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004980812128317772,
      "loss": 5.6605,
      "mean_token_accuracy": 0.16360202729701995,
      "num_tokens": 11528616.0,
      "step": 5545
    },
    {
      "entropy": 5.749828433990478,
      "epoch": 0.5034470246734397,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004980769938541364,
      "loss": 5.5917,
      "mean_token_accuracy": 0.16737237572669983,
      "num_tokens": 11539344.0,
      "step": 5550
    },
    {
      "entropy": 5.668318414688111,
      "epoch": 0.503900580551524,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004980727702632066,
      "loss": 5.6596,
      "mean_token_accuracy": 0.15578173995018005,
      "num_tokens": 11549511.0,
      "step": 5555
    },
    {
      "entropy": 5.656678009033203,
      "epoch": 0.5043541364296081,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004980685420590752,
      "loss": 5.5735,
      "mean_token_accuracy": 0.15664219707250596,
      "num_tokens": 11559425.0,
      "step": 5560
    },
    {
      "entropy": 5.761865758895874,
      "epoch": 0.5048076923076923,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004980643092418294,
      "loss": 5.6885,
      "mean_token_accuracy": 0.15531755238771439,
      "num_tokens": 11570406.0,
      "step": 5565
    },
    {
      "entropy": 5.6483618259429935,
      "epoch": 0.5052612481857764,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004980600718115568,
      "loss": 5.5607,
      "mean_token_accuracy": 0.1617144003510475,
      "num_tokens": 11580509.0,
      "step": 5570
    },
    {
      "entropy": 5.649977540969848,
      "epoch": 0.5057148040638607,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004980558297683452,
      "loss": 5.5774,
      "mean_token_accuracy": 0.162077334523201,
      "num_tokens": 11590860.0,
      "step": 5575
    },
    {
      "entropy": 5.675752067565918,
      "epoch": 0.5061683599419449,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004980515831122822,
      "loss": 5.5906,
      "mean_token_accuracy": 0.16118255853652955,
      "num_tokens": 11601954.0,
      "step": 5580
    },
    {
      "entropy": 5.675711917877197,
      "epoch": 0.506621915820029,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004980473318434555,
      "loss": 5.5955,
      "mean_token_accuracy": 0.15900284945964813,
      "num_tokens": 11612581.0,
      "step": 5585
    },
    {
      "entropy": 5.614867401123047,
      "epoch": 0.5070754716981132,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004980430759619532,
      "loss": 5.5274,
      "mean_token_accuracy": 0.16224440485239028,
      "num_tokens": 11622599.0,
      "step": 5590
    },
    {
      "entropy": 5.6130211353302,
      "epoch": 0.5075290275761973,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004980388154678633,
      "loss": 5.5505,
      "mean_token_accuracy": 0.159950090944767,
      "num_tokens": 11633611.0,
      "step": 5595
    },
    {
      "entropy": 5.590432405471802,
      "epoch": 0.5079825834542816,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004980345503612738,
      "loss": 5.5174,
      "mean_token_accuracy": 0.16368415355682372,
      "num_tokens": 11643159.0,
      "step": 5600
    },
    {
      "entropy": 5.688422393798828,
      "epoch": 0.5084361393323658,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004980302806422729,
      "loss": 5.4963,
      "mean_token_accuracy": 0.16657138168811797,
      "num_tokens": 11653174.0,
      "step": 5605
    },
    {
      "entropy": 5.689723253250122,
      "epoch": 0.5088896952104499,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004980260063109491,
      "loss": 5.5478,
      "mean_token_accuracy": 0.16194719970226287,
      "num_tokens": 11662621.0,
      "step": 5610
    },
    {
      "entropy": 5.531690263748169,
      "epoch": 0.5093432510885341,
      "grad_norm": 1.375,
      "learning_rate": 0.0004980217273673906,
      "loss": 5.5082,
      "mean_token_accuracy": 0.1628020793199539,
      "num_tokens": 11673538.0,
      "step": 5615
    },
    {
      "entropy": 5.670006513595581,
      "epoch": 0.5097968069666183,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004980174438116859,
      "loss": 5.5158,
      "mean_token_accuracy": 0.16596334427595139,
      "num_tokens": 11683203.0,
      "step": 5620
    },
    {
      "entropy": 5.5967333793640135,
      "epoch": 0.5102503628447025,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004980131556439236,
      "loss": 5.529,
      "mean_token_accuracy": 0.16600242257118225,
      "num_tokens": 11693313.0,
      "step": 5625
    },
    {
      "entropy": 5.628219795227051,
      "epoch": 0.5107039187227866,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004980088628641925,
      "loss": 5.5999,
      "mean_token_accuracy": 0.16239438205957413,
      "num_tokens": 11704194.0,
      "step": 5630
    },
    {
      "entropy": 5.6606667041778564,
      "epoch": 0.5111574746008708,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004980045654725812,
      "loss": 5.4617,
      "mean_token_accuracy": 0.16957058012485504,
      "num_tokens": 11715109.0,
      "step": 5635
    },
    {
      "entropy": 5.587406253814697,
      "epoch": 0.511611030478955,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004980002634691787,
      "loss": 5.5063,
      "mean_token_accuracy": 0.16856465935707093,
      "num_tokens": 11724961.0,
      "step": 5640
    },
    {
      "entropy": 5.55975661277771,
      "epoch": 0.5120645863570392,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000497995956854074,
      "loss": 5.5203,
      "mean_token_accuracy": 0.16654812693595886,
      "num_tokens": 11735334.0,
      "step": 5645
    },
    {
      "entropy": 5.60379810333252,
      "epoch": 0.5125181422351234,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004979916456273561,
      "loss": 5.5212,
      "mean_token_accuracy": 0.15964554995298386,
      "num_tokens": 11746858.0,
      "step": 5650
    },
    {
      "entropy": 5.744812774658203,
      "epoch": 0.5129716981132075,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004979873297891141,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1556268513202667,
      "num_tokens": 11757819.0,
      "step": 5655
    },
    {
      "entropy": 5.700674200057984,
      "epoch": 0.5134252539912917,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004979830093394372,
      "loss": 5.5531,
      "mean_token_accuracy": 0.16144313663244247,
      "num_tokens": 11767470.0,
      "step": 5660
    },
    {
      "entropy": 5.652323913574219,
      "epoch": 0.513878809869376,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004979786842784149,
      "loss": 5.485,
      "mean_token_accuracy": 0.17092576026916503,
      "num_tokens": 11777060.0,
      "step": 5665
    },
    {
      "entropy": 5.565343761444092,
      "epoch": 0.5143323657474601,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004979743546061366,
      "loss": 5.5276,
      "mean_token_accuracy": 0.1668785184621811,
      "num_tokens": 11787176.0,
      "step": 5670
    },
    {
      "entropy": 5.527346277236939,
      "epoch": 0.5147859216255443,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004979700203226918,
      "loss": 5.4265,
      "mean_token_accuracy": 0.16941867619752884,
      "num_tokens": 11796419.0,
      "step": 5675
    },
    {
      "entropy": 5.734120464324951,
      "epoch": 0.5152394775036284,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004979656814281702,
      "loss": 5.6449,
      "mean_token_accuracy": 0.15707888454198837,
      "num_tokens": 11807279.0,
      "step": 5680
    },
    {
      "entropy": 5.68532247543335,
      "epoch": 0.5156930333817126,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004979613379226615,
      "loss": 5.5324,
      "mean_token_accuracy": 0.16275421977043153,
      "num_tokens": 11817669.0,
      "step": 5685
    },
    {
      "entropy": 5.599370241165161,
      "epoch": 0.5161465892597968,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004979569898062554,
      "loss": 5.5532,
      "mean_token_accuracy": 0.16634066104888917,
      "num_tokens": 11827247.0,
      "step": 5690
    },
    {
      "entropy": 5.653990411758423,
      "epoch": 0.516600145137881,
      "grad_norm": 2.15625,
      "learning_rate": 0.000497952637079042,
      "loss": 5.533,
      "mean_token_accuracy": 0.15918290764093398,
      "num_tokens": 11838201.0,
      "step": 5695
    },
    {
      "entropy": 5.628846836090088,
      "epoch": 0.5170537010159652,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004979482797411114,
      "loss": 5.4696,
      "mean_token_accuracy": 0.17014151364564895,
      "num_tokens": 11847561.0,
      "step": 5700
    },
    {
      "entropy": 5.6203492164611815,
      "epoch": 0.5175072568940493,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004979439177925534,
      "loss": 5.5928,
      "mean_token_accuracy": 0.16502239555120468,
      "num_tokens": 11858272.0,
      "step": 5705
    },
    {
      "entropy": 5.693549585342407,
      "epoch": 0.5179608127721336,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004979395512334584,
      "loss": 5.6287,
      "mean_token_accuracy": 0.15832090377807617,
      "num_tokens": 11869660.0,
      "step": 5710
    },
    {
      "entropy": 5.728826475143433,
      "epoch": 0.5184143686502177,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004979351800639166,
      "loss": 5.6433,
      "mean_token_accuracy": 0.15534287691116333,
      "num_tokens": 11880239.0,
      "step": 5715
    },
    {
      "entropy": 5.7823890209197994,
      "epoch": 0.5188679245283019,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004979308042840186,
      "loss": 5.6272,
      "mean_token_accuracy": 0.15747978836297988,
      "num_tokens": 11890630.0,
      "step": 5720
    },
    {
      "entropy": 5.736940622329712,
      "epoch": 0.519321480406386,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004979264238938547,
      "loss": 5.6771,
      "mean_token_accuracy": 0.15597223192453386,
      "num_tokens": 11900948.0,
      "step": 5725
    },
    {
      "entropy": 5.611868143081665,
      "epoch": 0.5197750362844702,
      "grad_norm": 1.375,
      "learning_rate": 0.0004979220388935155,
      "loss": 5.5449,
      "mean_token_accuracy": 0.1591492623090744,
      "num_tokens": 11912528.0,
      "step": 5730
    },
    {
      "entropy": 5.612641048431397,
      "epoch": 0.5202285921625545,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004979176492830918,
      "loss": 5.4925,
      "mean_token_accuracy": 0.16271282583475113,
      "num_tokens": 11922211.0,
      "step": 5735
    },
    {
      "entropy": 5.666022205352784,
      "epoch": 0.5206821480406386,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004979132550626742,
      "loss": 5.5457,
      "mean_token_accuracy": 0.16144200712442397,
      "num_tokens": 11932520.0,
      "step": 5740
    },
    {
      "entropy": 5.680260562896729,
      "epoch": 0.5211357039187228,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004979088562323538,
      "loss": 5.4854,
      "mean_token_accuracy": 0.1675450712442398,
      "num_tokens": 11943312.0,
      "step": 5745
    },
    {
      "entropy": 5.60998158454895,
      "epoch": 0.5215892597968069,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004979044527922214,
      "loss": 5.6338,
      "mean_token_accuracy": 0.1593760222196579,
      "num_tokens": 11954455.0,
      "step": 5750
    },
    {
      "entropy": 5.657192420959473,
      "epoch": 0.5220428156748912,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004979000447423681,
      "loss": 5.5176,
      "mean_token_accuracy": 0.16424417793750762,
      "num_tokens": 11964650.0,
      "step": 5755
    },
    {
      "entropy": 5.650221443176269,
      "epoch": 0.5224963715529753,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004978956320828851,
      "loss": 5.5597,
      "mean_token_accuracy": 0.16050074696540834,
      "num_tokens": 11975134.0,
      "step": 5760
    },
    {
      "entropy": 5.661907148361206,
      "epoch": 0.5229499274310595,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004978912148138637,
      "loss": 5.5968,
      "mean_token_accuracy": 0.15903744250535964,
      "num_tokens": 11986486.0,
      "step": 5765
    },
    {
      "entropy": 5.7855431079864506,
      "epoch": 0.5234034833091437,
      "grad_norm": 1.34375,
      "learning_rate": 0.000497886792935395,
      "loss": 5.6066,
      "mean_token_accuracy": 0.1529120087623596,
      "num_tokens": 11998045.0,
      "step": 5770
    },
    {
      "entropy": 5.768364191055298,
      "epoch": 0.5238570391872278,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004978823664475707,
      "loss": 5.5903,
      "mean_token_accuracy": 0.1632813185453415,
      "num_tokens": 12008130.0,
      "step": 5775
    },
    {
      "entropy": 5.649330568313599,
      "epoch": 0.5243105950653121,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004978779353504824,
      "loss": 5.662,
      "mean_token_accuracy": 0.1525845743715763,
      "num_tokens": 12018753.0,
      "step": 5780
    },
    {
      "entropy": 5.679066848754883,
      "epoch": 0.5247641509433962,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004978734996442214,
      "loss": 5.5586,
      "mean_token_accuracy": 0.16855039298534394,
      "num_tokens": 12028202.0,
      "step": 5785
    },
    {
      "entropy": 5.69135422706604,
      "epoch": 0.5252177068214804,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004978690593288798,
      "loss": 5.6229,
      "mean_token_accuracy": 0.16048222556710243,
      "num_tokens": 12039652.0,
      "step": 5790
    },
    {
      "entropy": 5.696301746368408,
      "epoch": 0.5256712626995645,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004978646144045491,
      "loss": 5.5322,
      "mean_token_accuracy": 0.1572120249271393,
      "num_tokens": 12051269.0,
      "step": 5795
    },
    {
      "entropy": 5.570094633102417,
      "epoch": 0.5261248185776488,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004978601648713215,
      "loss": 5.4186,
      "mean_token_accuracy": 0.16403459310531615,
      "num_tokens": 12060647.0,
      "step": 5800
    },
    {
      "entropy": 5.623483562469483,
      "epoch": 0.526578374455733,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004978557107292889,
      "loss": 5.5752,
      "mean_token_accuracy": 0.16484973430633545,
      "num_tokens": 12071669.0,
      "step": 5805
    },
    {
      "entropy": 5.572929716110229,
      "epoch": 0.5270319303338171,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004978512519785433,
      "loss": 5.4485,
      "mean_token_accuracy": 0.16810268610715867,
      "num_tokens": 12081971.0,
      "step": 5810
    },
    {
      "entropy": 5.587448024749756,
      "epoch": 0.5274854862119013,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000497846788619177,
      "loss": 5.5893,
      "mean_token_accuracy": 0.1635736644268036,
      "num_tokens": 12092465.0,
      "step": 5815
    },
    {
      "entropy": 5.7142359733581545,
      "epoch": 0.5279390420899854,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004978423206512825,
      "loss": 5.5248,
      "mean_token_accuracy": 0.1615642175078392,
      "num_tokens": 12101752.0,
      "step": 5820
    },
    {
      "entropy": 5.731019115447998,
      "epoch": 0.5283925979680697,
      "grad_norm": 1.46875,
      "learning_rate": 0.000497837848074952,
      "loss": 5.6075,
      "mean_token_accuracy": 0.1558190815150738,
      "num_tokens": 12111805.0,
      "step": 5825
    },
    {
      "entropy": 5.712796831130982,
      "epoch": 0.5288461538461539,
      "grad_norm": 1.328125,
      "learning_rate": 0.000497833370890278,
      "loss": 5.6231,
      "mean_token_accuracy": 0.16041593104600907,
      "num_tokens": 12122511.0,
      "step": 5830
    },
    {
      "entropy": 5.646581935882568,
      "epoch": 0.529299709724238,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004978288890973529,
      "loss": 5.5107,
      "mean_token_accuracy": 0.16504603922367095,
      "num_tokens": 12132510.0,
      "step": 5835
    },
    {
      "entropy": 5.560458707809448,
      "epoch": 0.5297532656023222,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004978244026962698,
      "loss": 5.5272,
      "mean_token_accuracy": 0.160628604888916,
      "num_tokens": 12142419.0,
      "step": 5840
    },
    {
      "entropy": 5.665068292617798,
      "epoch": 0.5302068214804064,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004978199116871212,
      "loss": 5.5653,
      "mean_token_accuracy": 0.15784812569618226,
      "num_tokens": 12152759.0,
      "step": 5845
    },
    {
      "entropy": 5.674602031707764,
      "epoch": 0.5306603773584906,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00049781541607,
      "loss": 5.5238,
      "mean_token_accuracy": 0.16014756113290787,
      "num_tokens": 12162758.0,
      "step": 5850
    },
    {
      "entropy": 5.650653648376465,
      "epoch": 0.5311139332365747,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004978109158449993,
      "loss": 5.5829,
      "mean_token_accuracy": 0.15944256633520126,
      "num_tokens": 12172664.0,
      "step": 5855
    },
    {
      "entropy": 5.6479345798492435,
      "epoch": 0.5315674891146589,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004978064110122121,
      "loss": 5.5361,
      "mean_token_accuracy": 0.1643272027373314,
      "num_tokens": 12182028.0,
      "step": 5860
    },
    {
      "entropy": 5.698169708251953,
      "epoch": 0.532021044992743,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004978019015717314,
      "loss": 5.5836,
      "mean_token_accuracy": 0.16326664686203002,
      "num_tokens": 12192168.0,
      "step": 5865
    },
    {
      "entropy": 5.559962463378906,
      "epoch": 0.5324746008708273,
      "grad_norm": 1.375,
      "learning_rate": 0.0004977973875236508,
      "loss": 5.4842,
      "mean_token_accuracy": 0.16042792201042175,
      "num_tokens": 12202889.0,
      "step": 5870
    },
    {
      "entropy": 5.602790117263794,
      "epoch": 0.5329281567489115,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004977928688680634,
      "loss": 5.5518,
      "mean_token_accuracy": 0.16603052914142608,
      "num_tokens": 12212514.0,
      "step": 5875
    },
    {
      "entropy": 5.641613245010376,
      "epoch": 0.5333817126269956,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004977883456050626,
      "loss": 5.4052,
      "mean_token_accuracy": 0.1707530677318573,
      "num_tokens": 12222353.0,
      "step": 5880
    },
    {
      "entropy": 5.693383884429932,
      "epoch": 0.5338352685050798,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000497783817734742,
      "loss": 5.6637,
      "mean_token_accuracy": 0.1538747400045395,
      "num_tokens": 12232845.0,
      "step": 5885
    },
    {
      "entropy": 5.656322717666626,
      "epoch": 0.534288824383164,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004977792852571954,
      "loss": 5.611,
      "mean_token_accuracy": 0.15664656311273575,
      "num_tokens": 12242950.0,
      "step": 5890
    },
    {
      "entropy": 5.683221960067749,
      "epoch": 0.5347423802612482,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004977747481725164,
      "loss": 5.5828,
      "mean_token_accuracy": 0.15758552253246308,
      "num_tokens": 12253132.0,
      "step": 5895
    },
    {
      "entropy": 5.648868751525879,
      "epoch": 0.5351959361393324,
      "grad_norm": 1.375,
      "learning_rate": 0.0004977702064807987,
      "loss": 5.5352,
      "mean_token_accuracy": 0.16616235822439193,
      "num_tokens": 12263214.0,
      "step": 5900
    },
    {
      "entropy": 5.588494825363159,
      "epoch": 0.5356494920174165,
      "grad_norm": 1.75,
      "learning_rate": 0.0004977656601821366,
      "loss": 5.4564,
      "mean_token_accuracy": 0.17042534202337264,
      "num_tokens": 12273250.0,
      "step": 5905
    },
    {
      "entropy": 5.783517217636108,
      "epoch": 0.5361030478955007,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004977611092766238,
      "loss": 5.6551,
      "mean_token_accuracy": 0.16542720496654512,
      "num_tokens": 12283767.0,
      "step": 5910
    },
    {
      "entropy": 5.619170093536377,
      "epoch": 0.5365566037735849,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004977565537643544,
      "loss": 5.496,
      "mean_token_accuracy": 0.15744243562221527,
      "num_tokens": 12294612.0,
      "step": 5915
    },
    {
      "entropy": 5.574744081497192,
      "epoch": 0.5370101596516691,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004977519936454227,
      "loss": 5.5902,
      "mean_token_accuracy": 0.16173869967460633,
      "num_tokens": 12305614.0,
      "step": 5920
    },
    {
      "entropy": 5.6041184425354,
      "epoch": 0.5374637155297532,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004977474289199231,
      "loss": 5.4586,
      "mean_token_accuracy": 0.1602635309100151,
      "num_tokens": 12316152.0,
      "step": 5925
    },
    {
      "entropy": 5.607742023468018,
      "epoch": 0.5379172714078374,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004977428595879499,
      "loss": 5.5545,
      "mean_token_accuracy": 0.16442461162805558,
      "num_tokens": 12325940.0,
      "step": 5930
    },
    {
      "entropy": 5.586103534698486,
      "epoch": 0.5383708272859217,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004977382856495975,
      "loss": 5.5444,
      "mean_token_accuracy": 0.1628282383084297,
      "num_tokens": 12335482.0,
      "step": 5935
    },
    {
      "entropy": 5.7526442527771,
      "epoch": 0.5388243831640058,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004977337071049606,
      "loss": 5.6234,
      "mean_token_accuracy": 0.1566042810678482,
      "num_tokens": 12346386.0,
      "step": 5940
    },
    {
      "entropy": 5.743467712402344,
      "epoch": 0.53927793904209,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000497729123954134,
      "loss": 5.5752,
      "mean_token_accuracy": 0.16442530900239943,
      "num_tokens": 12356282.0,
      "step": 5945
    },
    {
      "entropy": 5.566780138015747,
      "epoch": 0.5397314949201741,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004977245361972123,
      "loss": 5.5747,
      "mean_token_accuracy": 0.16320088356733323,
      "num_tokens": 12366994.0,
      "step": 5950
    },
    {
      "entropy": 5.675820970535279,
      "epoch": 0.5401850507982583,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004977199438342904,
      "loss": 5.4543,
      "mean_token_accuracy": 0.16420960426330566,
      "num_tokens": 12377524.0,
      "step": 5955
    },
    {
      "entropy": 5.647736883163452,
      "epoch": 0.5406386066763426,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004977153468654634,
      "loss": 5.6035,
      "mean_token_accuracy": 0.16150177270174026,
      "num_tokens": 12387593.0,
      "step": 5960
    },
    {
      "entropy": 5.665026760101318,
      "epoch": 0.5410921625544267,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004977107452908262,
      "loss": 5.557,
      "mean_token_accuracy": 0.16269752979278565,
      "num_tokens": 12398732.0,
      "step": 5965
    },
    {
      "entropy": 5.676980781555176,
      "epoch": 0.5415457184325109,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000497706139110474,
      "loss": 5.5696,
      "mean_token_accuracy": 0.16847805306315422,
      "num_tokens": 12409821.0,
      "step": 5970
    },
    {
      "entropy": 5.6549646854400635,
      "epoch": 0.541999274310595,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004977015283245021,
      "loss": 5.5842,
      "mean_token_accuracy": 0.15639931112527847,
      "num_tokens": 12420044.0,
      "step": 5975
    },
    {
      "entropy": 5.526405477523804,
      "epoch": 0.5424528301886793,
      "grad_norm": 1.375,
      "learning_rate": 0.0004976969129330059,
      "loss": 5.5117,
      "mean_token_accuracy": 0.15921124070882797,
      "num_tokens": 12430980.0,
      "step": 5980
    },
    {
      "entropy": 5.622525978088379,
      "epoch": 0.5429063860667634,
      "grad_norm": 1.5,
      "learning_rate": 0.0004976922929360808,
      "loss": 5.5557,
      "mean_token_accuracy": 0.16727575212717055,
      "num_tokens": 12441999.0,
      "step": 5985
    },
    {
      "entropy": 5.723354196548462,
      "epoch": 0.5433599419448476,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004976876683338222,
      "loss": 5.6019,
      "mean_token_accuracy": 0.15907001495361328,
      "num_tokens": 12452837.0,
      "step": 5990
    },
    {
      "entropy": 5.6485847473144535,
      "epoch": 0.5438134978229318,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000497683039126326,
      "loss": 5.5139,
      "mean_token_accuracy": 0.16824455857276915,
      "num_tokens": 12462142.0,
      "step": 5995
    },
    {
      "entropy": 5.643194198608398,
      "epoch": 0.5442670537010159,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004976784053136878,
      "loss": 5.5616,
      "mean_token_accuracy": 0.16665686517953873,
      "num_tokens": 12472756.0,
      "step": 6000
    },
    {
      "epoch": 0.5442670537010159,
      "eval_entropy": 5.472823539039129,
      "eval_loss": 5.589262962341309,
      "eval_mean_token_accuracy": 0.16745942614938877,
      "eval_num_tokens": 12472756.0,
      "eval_runtime": 26.5027,
      "eval_samples_per_second": 1334.206,
      "eval_steps_per_second": 166.776,
      "step": 6000
    },
    {
      "entropy": 5.670133972167969,
      "epoch": 0.5447206095791002,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004976737668960034,
      "loss": 5.6728,
      "mean_token_accuracy": 0.1545624852180481,
      "num_tokens": 12483115.0,
      "step": 6005
    },
    {
      "entropy": 5.66155514717102,
      "epoch": 0.5451741654571843,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004976691238733688,
      "loss": 5.5282,
      "mean_token_accuracy": 0.1589282423257828,
      "num_tokens": 12493280.0,
      "step": 6010
    },
    {
      "entropy": 5.693755388259888,
      "epoch": 0.5456277213352685,
      "grad_norm": 1.40625,
      "learning_rate": 0.00049766447624588,
      "loss": 5.6572,
      "mean_token_accuracy": 0.16183248460292815,
      "num_tokens": 12503665.0,
      "step": 6015
    },
    {
      "entropy": 5.695195531845092,
      "epoch": 0.5460812772133526,
      "grad_norm": 1.40625,
      "learning_rate": 0.000497659824013633,
      "loss": 5.5691,
      "mean_token_accuracy": 0.1570957064628601,
      "num_tokens": 12513381.0,
      "step": 6020
    },
    {
      "entropy": 5.753922653198242,
      "epoch": 0.5465348330914369,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004976551671767241,
      "loss": 5.5945,
      "mean_token_accuracy": 0.16156743764877318,
      "num_tokens": 12524121.0,
      "step": 6025
    },
    {
      "entropy": 5.648118257522583,
      "epoch": 0.5469883889695211,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004976505057352496,
      "loss": 5.5566,
      "mean_token_accuracy": 0.16416458934545516,
      "num_tokens": 12533972.0,
      "step": 6030
    },
    {
      "entropy": 5.698191070556641,
      "epoch": 0.5474419448476052,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004976458396893059,
      "loss": 5.6674,
      "mean_token_accuracy": 0.1542829692363739,
      "num_tokens": 12544110.0,
      "step": 6035
    },
    {
      "entropy": 5.691080808639526,
      "epoch": 0.5478955007256894,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004976411690389896,
      "loss": 5.4975,
      "mean_token_accuracy": 0.15873684734106064,
      "num_tokens": 12554273.0,
      "step": 6040
    },
    {
      "entropy": 5.705283308029175,
      "epoch": 0.5483490566037735,
      "grad_norm": 1.53125,
      "learning_rate": 0.000497636493784397,
      "loss": 5.5805,
      "mean_token_accuracy": 0.16140856519341468,
      "num_tokens": 12564633.0,
      "step": 6045
    },
    {
      "entropy": 5.645505380630493,
      "epoch": 0.5488026124818578,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004976318139256251,
      "loss": 5.5151,
      "mean_token_accuracy": 0.1661514475941658,
      "num_tokens": 12573790.0,
      "step": 6050
    },
    {
      "entropy": 5.6294355392456055,
      "epoch": 0.549256168359942,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004976271294627706,
      "loss": 5.5808,
      "mean_token_accuracy": 0.15535681694746017,
      "num_tokens": 12584265.0,
      "step": 6055
    },
    {
      "entropy": 5.621347236633301,
      "epoch": 0.5497097242380261,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004976224403959303,
      "loss": 5.4564,
      "mean_token_accuracy": 0.16849451810121535,
      "num_tokens": 12594413.0,
      "step": 6060
    },
    {
      "entropy": 5.657793998718262,
      "epoch": 0.5501632801161103,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004976177467252012,
      "loss": 5.5255,
      "mean_token_accuracy": 0.1673310860991478,
      "num_tokens": 12604637.0,
      "step": 6065
    },
    {
      "entropy": 5.661483812332153,
      "epoch": 0.5506168359941945,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004976130484506804,
      "loss": 5.5974,
      "mean_token_accuracy": 0.1647314742207527,
      "num_tokens": 12614873.0,
      "step": 6070
    },
    {
      "entropy": 5.6395495414733885,
      "epoch": 0.5510703918722787,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004976083455724649,
      "loss": 5.5349,
      "mean_token_accuracy": 0.1589552417397499,
      "num_tokens": 12626182.0,
      "step": 6075
    },
    {
      "entropy": 5.614484357833862,
      "epoch": 0.5515239477503628,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004976036380906523,
      "loss": 5.4679,
      "mean_token_accuracy": 0.16557910591363906,
      "num_tokens": 12635418.0,
      "step": 6080
    },
    {
      "entropy": 5.648619842529297,
      "epoch": 0.551977503628447,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004975989260053395,
      "loss": 5.6148,
      "mean_token_accuracy": 0.1543217584490776,
      "num_tokens": 12645942.0,
      "step": 6085
    },
    {
      "entropy": 5.660171985626221,
      "epoch": 0.5524310595065312,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004975942093166242,
      "loss": 5.5313,
      "mean_token_accuracy": 0.16052016615867615,
      "num_tokens": 12656460.0,
      "step": 6090
    },
    {
      "entropy": 5.695255470275879,
      "epoch": 0.5528846153846154,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004975894880246041,
      "loss": 5.4757,
      "mean_token_accuracy": 0.16347223967313768,
      "num_tokens": 12666916.0,
      "step": 6095
    },
    {
      "entropy": 5.601450824737549,
      "epoch": 0.5533381712626996,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004975847621293766,
      "loss": 5.5431,
      "mean_token_accuracy": 0.1696389064192772,
      "num_tokens": 12677562.0,
      "step": 6100
    },
    {
      "entropy": 5.63624677658081,
      "epoch": 0.5537917271407837,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004975800316310394,
      "loss": 5.5016,
      "mean_token_accuracy": 0.16171834021806716,
      "num_tokens": 12687349.0,
      "step": 6105
    },
    {
      "entropy": 5.649888229370117,
      "epoch": 0.5542452830188679,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004975752965296906,
      "loss": 5.5659,
      "mean_token_accuracy": 0.1648343488574028,
      "num_tokens": 12697850.0,
      "step": 6110
    },
    {
      "entropy": 5.612441205978394,
      "epoch": 0.5546988388969522,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004975705568254278,
      "loss": 5.4159,
      "mean_token_accuracy": 0.16573707461357118,
      "num_tokens": 12708283.0,
      "step": 6115
    },
    {
      "entropy": 5.648651742935181,
      "epoch": 0.5551523947750363,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004975658125183493,
      "loss": 5.5105,
      "mean_token_accuracy": 0.1632908210158348,
      "num_tokens": 12718201.0,
      "step": 6120
    },
    {
      "entropy": 5.60323486328125,
      "epoch": 0.5556059506531205,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000497561063608553,
      "loss": 5.4824,
      "mean_token_accuracy": 0.1683592066168785,
      "num_tokens": 12728860.0,
      "step": 6125
    },
    {
      "entropy": 5.626165914535522,
      "epoch": 0.5560595065312046,
      "grad_norm": 1.375,
      "learning_rate": 0.0004975563100961372,
      "loss": 5.5331,
      "mean_token_accuracy": 0.15690662562847138,
      "num_tokens": 12739007.0,
      "step": 6130
    },
    {
      "entropy": 5.606911945343017,
      "epoch": 0.5565130624092888,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004975515519812002,
      "loss": 5.3911,
      "mean_token_accuracy": 0.17006937563419341,
      "num_tokens": 12749004.0,
      "step": 6135
    },
    {
      "entropy": 5.596025609970093,
      "epoch": 0.556966618287373,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004975467892638402,
      "loss": 5.5562,
      "mean_token_accuracy": 0.16381375193595887,
      "num_tokens": 12760085.0,
      "step": 6140
    },
    {
      "entropy": 5.645767450332642,
      "epoch": 0.5574201741654572,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004975420219441561,
      "loss": 5.5069,
      "mean_token_accuracy": 0.17016789615154265,
      "num_tokens": 12770303.0,
      "step": 6145
    },
    {
      "entropy": 5.661552429199219,
      "epoch": 0.5578737300435413,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004975372500222462,
      "loss": 5.5206,
      "mean_token_accuracy": 0.16546058803796768,
      "num_tokens": 12780334.0,
      "step": 6150
    },
    {
      "entropy": 5.6559959888458256,
      "epoch": 0.5583272859216255,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004975324734982094,
      "loss": 5.5742,
      "mean_token_accuracy": 0.16141054183244705,
      "num_tokens": 12790980.0,
      "step": 6155
    },
    {
      "entropy": 5.643654870986938,
      "epoch": 0.5587808417997098,
      "grad_norm": 3.84375,
      "learning_rate": 0.0004975276923721442,
      "loss": 5.604,
      "mean_token_accuracy": 0.15867158025503159,
      "num_tokens": 12801713.0,
      "step": 6160
    },
    {
      "entropy": 5.699330234527588,
      "epoch": 0.5592343976777939,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004975229066441496,
      "loss": 5.6061,
      "mean_token_accuracy": 0.15926154106855392,
      "num_tokens": 12813040.0,
      "step": 6165
    },
    {
      "entropy": 5.693249845504761,
      "epoch": 0.5596879535558781,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004975181163143247,
      "loss": 5.5658,
      "mean_token_accuracy": 0.15429083853960038,
      "num_tokens": 12824048.0,
      "step": 6170
    },
    {
      "entropy": 5.656903409957886,
      "epoch": 0.5601415094339622,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004975133213827684,
      "loss": 5.5134,
      "mean_token_accuracy": 0.15892634391784669,
      "num_tokens": 12834297.0,
      "step": 6175
    },
    {
      "entropy": 5.745335054397583,
      "epoch": 0.5605950653120464,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004975085218495798,
      "loss": 5.5787,
      "mean_token_accuracy": 0.16047458797693254,
      "num_tokens": 12844574.0,
      "step": 6180
    },
    {
      "entropy": 5.560383129119873,
      "epoch": 0.5610486211901307,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004975037177148585,
      "loss": 5.5142,
      "mean_token_accuracy": 0.1610662505030632,
      "num_tokens": 12854670.0,
      "step": 6185
    },
    {
      "entropy": 5.561988258361817,
      "epoch": 0.5615021770682148,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004974989089787035,
      "loss": 5.5378,
      "mean_token_accuracy": 0.1686451852321625,
      "num_tokens": 12865281.0,
      "step": 6190
    },
    {
      "entropy": 5.702982187271118,
      "epoch": 0.561955732946299,
      "grad_norm": 2.5,
      "learning_rate": 0.0004974940956412143,
      "loss": 5.5535,
      "mean_token_accuracy": 0.1641027018427849,
      "num_tokens": 12875222.0,
      "step": 6195
    },
    {
      "entropy": 5.816201686859131,
      "epoch": 0.5624092888243831,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004974892777024905,
      "loss": 5.6098,
      "mean_token_accuracy": 0.15752922147512435,
      "num_tokens": 12885079.0,
      "step": 6200
    },
    {
      "entropy": 5.67298469543457,
      "epoch": 0.5628628447024674,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004974844551626318,
      "loss": 5.5498,
      "mean_token_accuracy": 0.1641029268503189,
      "num_tokens": 12896402.0,
      "step": 6205
    },
    {
      "entropy": 5.614017629623413,
      "epoch": 0.5633164005805515,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004974796280217379,
      "loss": 5.526,
      "mean_token_accuracy": 0.16557157337665557,
      "num_tokens": 12906403.0,
      "step": 6210
    },
    {
      "entropy": 5.71490626335144,
      "epoch": 0.5637699564586357,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004974747962799086,
      "loss": 5.5168,
      "mean_token_accuracy": 0.1579867884516716,
      "num_tokens": 12917061.0,
      "step": 6215
    },
    {
      "entropy": 5.656363534927368,
      "epoch": 0.5642235123367199,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004974699599372437,
      "loss": 5.5807,
      "mean_token_accuracy": 0.16344264447689055,
      "num_tokens": 12927182.0,
      "step": 6220
    },
    {
      "entropy": 5.624368143081665,
      "epoch": 0.564677068214804,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004974651189938434,
      "loss": 5.5315,
      "mean_token_accuracy": 0.16056101471185685,
      "num_tokens": 12937371.0,
      "step": 6225
    },
    {
      "entropy": 5.70891375541687,
      "epoch": 0.5651306240928883,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004974602734498079,
      "loss": 5.5941,
      "mean_token_accuracy": 0.15544345378875732,
      "num_tokens": 12948569.0,
      "step": 6230
    },
    {
      "entropy": 5.741915559768676,
      "epoch": 0.5655841799709724,
      "grad_norm": 1.875,
      "learning_rate": 0.000497455423305237,
      "loss": 5.5432,
      "mean_token_accuracy": 0.15569789856672286,
      "num_tokens": 12958503.0,
      "step": 6235
    },
    {
      "entropy": 5.718269109725952,
      "epoch": 0.5660377358490566,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004974505685602314,
      "loss": 5.5621,
      "mean_token_accuracy": 0.15982589423656463,
      "num_tokens": 12968714.0,
      "step": 6240
    },
    {
      "entropy": 5.644350957870484,
      "epoch": 0.5664912917271407,
      "grad_norm": 1.625,
      "learning_rate": 0.0004974457092148912,
      "loss": 5.5907,
      "mean_token_accuracy": 0.16113034933805465,
      "num_tokens": 12978857.0,
      "step": 6245
    },
    {
      "entropy": 5.7991985321044925,
      "epoch": 0.566944847605225,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004974408452693172,
      "loss": 5.7043,
      "mean_token_accuracy": 0.1546899065375328,
      "num_tokens": 12989863.0,
      "step": 6250
    },
    {
      "entropy": 5.7228484630584715,
      "epoch": 0.5673984034833092,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004974359767236099,
      "loss": 5.5004,
      "mean_token_accuracy": 0.16300520449876785,
      "num_tokens": 13000307.0,
      "step": 6255
    },
    {
      "entropy": 5.664002227783203,
      "epoch": 0.5678519593613933,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004974311035778698,
      "loss": 5.5672,
      "mean_token_accuracy": 0.16984948217868806,
      "num_tokens": 13009053.0,
      "step": 6260
    },
    {
      "entropy": 5.61906361579895,
      "epoch": 0.5683055152394775,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004974262258321979,
      "loss": 5.5952,
      "mean_token_accuracy": 0.15511150807142257,
      "num_tokens": 13019136.0,
      "step": 6265
    },
    {
      "entropy": 5.743262004852295,
      "epoch": 0.5687590711175616,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004974213434866949,
      "loss": 5.6184,
      "mean_token_accuracy": 0.15917437374591828,
      "num_tokens": 13030356.0,
      "step": 6270
    },
    {
      "entropy": 5.682684850692749,
      "epoch": 0.5692126269956459,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004974164565414617,
      "loss": 5.6411,
      "mean_token_accuracy": 0.16206429302692413,
      "num_tokens": 13040477.0,
      "step": 6275
    },
    {
      "entropy": 5.6901579856872555,
      "epoch": 0.56966618287373,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004974115649965998,
      "loss": 5.5217,
      "mean_token_accuracy": 0.1673882707953453,
      "num_tokens": 13051755.0,
      "step": 6280
    },
    {
      "entropy": 5.739024353027344,
      "epoch": 0.5701197387518142,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004974066688522098,
      "loss": 5.5282,
      "mean_token_accuracy": 0.17134420275688172,
      "num_tokens": 13061670.0,
      "step": 6285
    },
    {
      "entropy": 5.634558582305909,
      "epoch": 0.5705732946298984,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004974017681083934,
      "loss": 5.511,
      "mean_token_accuracy": 0.1640141114592552,
      "num_tokens": 13070917.0,
      "step": 6290
    },
    {
      "entropy": 5.620419692993164,
      "epoch": 0.5710268505079826,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004973968627652516,
      "loss": 5.5586,
      "mean_token_accuracy": 0.16330747306346893,
      "num_tokens": 13080577.0,
      "step": 6295
    },
    {
      "entropy": 5.691771507263184,
      "epoch": 0.5714804063860668,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000497391952822886,
      "loss": 5.5913,
      "mean_token_accuracy": 0.16383007168769836,
      "num_tokens": 13091446.0,
      "step": 6300
    },
    {
      "entropy": 5.720631074905396,
      "epoch": 0.5719339622641509,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004973870382813983,
      "loss": 5.6143,
      "mean_token_accuracy": 0.1649190291762352,
      "num_tokens": 13102752.0,
      "step": 6305
    },
    {
      "entropy": 5.738799524307251,
      "epoch": 0.5723875181422351,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004973821191408899,
      "loss": 5.6321,
      "mean_token_accuracy": 0.1576658844947815,
      "num_tokens": 13113944.0,
      "step": 6310
    },
    {
      "entropy": 5.651340627670288,
      "epoch": 0.5728410740203193,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004973771954014626,
      "loss": 5.5738,
      "mean_token_accuracy": 0.15872492641210556,
      "num_tokens": 13124316.0,
      "step": 6315
    },
    {
      "entropy": 5.741772079467774,
      "epoch": 0.5732946298984035,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004973722670632182,
      "loss": 5.6688,
      "mean_token_accuracy": 0.15541704148054122,
      "num_tokens": 13133567.0,
      "step": 6320
    },
    {
      "entropy": 5.67438702583313,
      "epoch": 0.5737481857764877,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004973673341262587,
      "loss": 5.5689,
      "mean_token_accuracy": 0.16231633573770524,
      "num_tokens": 13144279.0,
      "step": 6325
    },
    {
      "entropy": 5.728409671783448,
      "epoch": 0.5742017416545718,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000497362396590686,
      "loss": 5.6256,
      "mean_token_accuracy": 0.16435034647583963,
      "num_tokens": 13154463.0,
      "step": 6330
    },
    {
      "entropy": 5.754732084274292,
      "epoch": 0.574655297532656,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004973574544566024,
      "loss": 5.603,
      "mean_token_accuracy": 0.15627192705869675,
      "num_tokens": 13165565.0,
      "step": 6335
    },
    {
      "entropy": 5.630088138580322,
      "epoch": 0.5751088534107403,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004973525077241098,
      "loss": 5.5327,
      "mean_token_accuracy": 0.16437099426984786,
      "num_tokens": 13175839.0,
      "step": 6340
    },
    {
      "entropy": 5.612335538864135,
      "epoch": 0.5755624092888244,
      "grad_norm": 3.078125,
      "learning_rate": 0.0004973475563933109,
      "loss": 5.504,
      "mean_token_accuracy": 0.16129760444164276,
      "num_tokens": 13186372.0,
      "step": 6345
    },
    {
      "entropy": 5.63255467414856,
      "epoch": 0.5760159651669086,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004973426004643077,
      "loss": 5.5579,
      "mean_token_accuracy": 0.16001515984535217,
      "num_tokens": 13196259.0,
      "step": 6350
    },
    {
      "entropy": 5.689817476272583,
      "epoch": 0.5764695210449927,
      "grad_norm": 2.734375,
      "learning_rate": 0.0004973376399372029,
      "loss": 5.5211,
      "mean_token_accuracy": 0.16678900122642518,
      "num_tokens": 13206497.0,
      "step": 6355
    },
    {
      "entropy": 5.654291391372681,
      "epoch": 0.5769230769230769,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004973326748120991,
      "loss": 5.5664,
      "mean_token_accuracy": 0.16469015628099443,
      "num_tokens": 13217188.0,
      "step": 6360
    },
    {
      "entropy": 5.670374917984009,
      "epoch": 0.5773766328011611,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004973277050890988,
      "loss": 5.5756,
      "mean_token_accuracy": 0.16094538420438767,
      "num_tokens": 13226544.0,
      "step": 6365
    },
    {
      "entropy": 5.710937643051148,
      "epoch": 0.5778301886792453,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004973227307683049,
      "loss": 5.6146,
      "mean_token_accuracy": 0.15699274688959122,
      "num_tokens": 13236572.0,
      "step": 6370
    },
    {
      "entropy": 5.80208945274353,
      "epoch": 0.5782837445573294,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004973177518498204,
      "loss": 5.6121,
      "mean_token_accuracy": 0.16148673593997956,
      "num_tokens": 13247958.0,
      "step": 6375
    },
    {
      "entropy": 5.642516231536865,
      "epoch": 0.5787373004354136,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004973127683337482,
      "loss": 5.5127,
      "mean_token_accuracy": 0.1633951857686043,
      "num_tokens": 13258255.0,
      "step": 6380
    },
    {
      "entropy": 5.5757979393005375,
      "epoch": 0.5791908563134979,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004973077802201911,
      "loss": 5.5415,
      "mean_token_accuracy": 0.1615252360701561,
      "num_tokens": 13268416.0,
      "step": 6385
    },
    {
      "entropy": 5.58531641960144,
      "epoch": 0.579644412191582,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004973027875092526,
      "loss": 5.5174,
      "mean_token_accuracy": 0.16312111020088196,
      "num_tokens": 13278087.0,
      "step": 6390
    },
    {
      "entropy": 5.604562997817993,
      "epoch": 0.5800979680696662,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004972977902010357,
      "loss": 5.424,
      "mean_token_accuracy": 0.1713629350066185,
      "num_tokens": 13288121.0,
      "step": 6395
    },
    {
      "entropy": 5.772391843795776,
      "epoch": 0.5805515239477503,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004972927882956439,
      "loss": 5.644,
      "mean_token_accuracy": 0.15260053873062135,
      "num_tokens": 13298620.0,
      "step": 6400
    },
    {
      "entropy": 5.630117321014405,
      "epoch": 0.5810050798258345,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004972877817931807,
      "loss": 5.4559,
      "mean_token_accuracy": 0.16051941812038423,
      "num_tokens": 13308413.0,
      "step": 6405
    },
    {
      "entropy": 5.5901875495910645,
      "epoch": 0.5814586357039188,
      "grad_norm": 2.875,
      "learning_rate": 0.0004972827706937493,
      "loss": 5.5598,
      "mean_token_accuracy": 0.16214399337768554,
      "num_tokens": 13319387.0,
      "step": 6410
    },
    {
      "entropy": 5.653078556060791,
      "epoch": 0.5819121915820029,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004972777549974537,
      "loss": 5.5052,
      "mean_token_accuracy": 0.16626724004745483,
      "num_tokens": 13330494.0,
      "step": 6415
    },
    {
      "entropy": 5.667278242111206,
      "epoch": 0.5823657474600871,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004972727347043975,
      "loss": 5.5175,
      "mean_token_accuracy": 0.15966289341449738,
      "num_tokens": 13341009.0,
      "step": 6420
    },
    {
      "entropy": 5.6079059600830075,
      "epoch": 0.5828193033381712,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004972677098146845,
      "loss": 5.4678,
      "mean_token_accuracy": 0.16285288631916045,
      "num_tokens": 13351269.0,
      "step": 6425
    },
    {
      "entropy": 5.573035478591919,
      "epoch": 0.5832728592162555,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004972626803284185,
      "loss": 5.5728,
      "mean_token_accuracy": 0.16400626450777053,
      "num_tokens": 13362160.0,
      "step": 6430
    },
    {
      "entropy": 5.701131343841553,
      "epoch": 0.5837264150943396,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004972576462457038,
      "loss": 5.6297,
      "mean_token_accuracy": 0.16086410582065583,
      "num_tokens": 13372497.0,
      "step": 6435
    },
    {
      "entropy": 5.788417720794678,
      "epoch": 0.5841799709724238,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004972526075666442,
      "loss": 5.5724,
      "mean_token_accuracy": 0.16689816117286682,
      "num_tokens": 13381665.0,
      "step": 6440
    },
    {
      "entropy": 5.620719575881958,
      "epoch": 0.584633526850508,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004972475642913442,
      "loss": 5.4744,
      "mean_token_accuracy": 0.16386179476976395,
      "num_tokens": 13392507.0,
      "step": 6445
    },
    {
      "entropy": 5.537435913085938,
      "epoch": 0.5850870827285921,
      "grad_norm": 1.625,
      "learning_rate": 0.0004972425164199078,
      "loss": 5.4755,
      "mean_token_accuracy": 0.17011575847864152,
      "num_tokens": 13402171.0,
      "step": 6450
    },
    {
      "entropy": 5.713273334503174,
      "epoch": 0.5855406386066764,
      "grad_norm": 3.921875,
      "learning_rate": 0.0004972374639524396,
      "loss": 5.6784,
      "mean_token_accuracy": 0.15556367337703705,
      "num_tokens": 13412933.0,
      "step": 6455
    },
    {
      "entropy": 5.7077325820922855,
      "epoch": 0.5859941944847605,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000497232406889044,
      "loss": 5.5258,
      "mean_token_accuracy": 0.15624898225069045,
      "num_tokens": 13423563.0,
      "step": 6460
    },
    {
      "entropy": 5.683661603927613,
      "epoch": 0.5864477503628447,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004972273452298255,
      "loss": 5.5487,
      "mean_token_accuracy": 0.16698916330933572,
      "num_tokens": 13433998.0,
      "step": 6465
    },
    {
      "entropy": 5.607936191558838,
      "epoch": 0.5869013062409288,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004972222789748888,
      "loss": 5.489,
      "mean_token_accuracy": 0.16828617155551912,
      "num_tokens": 13443878.0,
      "step": 6470
    },
    {
      "entropy": 5.618822765350342,
      "epoch": 0.5873548621190131,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004972172081243389,
      "loss": 5.5781,
      "mean_token_accuracy": 0.1563350200653076,
      "num_tokens": 13456051.0,
      "step": 6475
    },
    {
      "entropy": 5.6659345626831055,
      "epoch": 0.5878084179970973,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004972121326782806,
      "loss": 5.5814,
      "mean_token_accuracy": 0.16245667785406112,
      "num_tokens": 13467276.0,
      "step": 6480
    },
    {
      "entropy": 5.6519945621490475,
      "epoch": 0.5882619738751814,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004972070526368186,
      "loss": 5.5502,
      "mean_token_accuracy": 0.1641760379076004,
      "num_tokens": 13478331.0,
      "step": 6485
    },
    {
      "entropy": 5.698674488067627,
      "epoch": 0.5887155297532656,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004972019680000582,
      "loss": 5.4949,
      "mean_token_accuracy": 0.16153709441423417,
      "num_tokens": 13489879.0,
      "step": 6490
    },
    {
      "entropy": 5.666459035873413,
      "epoch": 0.5891690856313497,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004971968787681044,
      "loss": 5.6105,
      "mean_token_accuracy": 0.16374465972185134,
      "num_tokens": 13500879.0,
      "step": 6495
    },
    {
      "entropy": 5.701348543167114,
      "epoch": 0.589622641509434,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004971917849410627,
      "loss": 5.5873,
      "mean_token_accuracy": 0.1603960059583187,
      "num_tokens": 13511719.0,
      "step": 6500
    },
    {
      "entropy": 5.652278709411621,
      "epoch": 0.5900761973875182,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004971866865190381,
      "loss": 5.5155,
      "mean_token_accuracy": 0.16301714330911637,
      "num_tokens": 13521850.0,
      "step": 6505
    },
    {
      "entropy": 5.681592130661011,
      "epoch": 0.5905297532656023,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004971815835021362,
      "loss": 5.5479,
      "mean_token_accuracy": 0.16761482805013656,
      "num_tokens": 13532095.0,
      "step": 6510
    },
    {
      "entropy": 5.670617389678955,
      "epoch": 0.5909833091436865,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004971764758904626,
      "loss": 5.5305,
      "mean_token_accuracy": 0.1592058703303337,
      "num_tokens": 13543133.0,
      "step": 6515
    },
    {
      "entropy": 5.653409767150879,
      "epoch": 0.5914368650217707,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004971713636841229,
      "loss": 5.6048,
      "mean_token_accuracy": 0.1548130914568901,
      "num_tokens": 13553771.0,
      "step": 6520
    },
    {
      "entropy": 5.633445310592651,
      "epoch": 0.5918904208998549,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004971662468832228,
      "loss": 5.5414,
      "mean_token_accuracy": 0.16742569208145142,
      "num_tokens": 13564548.0,
      "step": 6525
    },
    {
      "entropy": 5.644088172912598,
      "epoch": 0.592343976777939,
      "grad_norm": 2.0,
      "learning_rate": 0.000497161125487868,
      "loss": 5.5248,
      "mean_token_accuracy": 0.1607619509100914,
      "num_tokens": 13574567.0,
      "step": 6530
    },
    {
      "entropy": 5.584167718887329,
      "epoch": 0.5927975326560232,
      "grad_norm": 2.0,
      "learning_rate": 0.0004971559994981645,
      "loss": 5.5618,
      "mean_token_accuracy": 0.15815457254648208,
      "num_tokens": 13585302.0,
      "step": 6535
    },
    {
      "entropy": 5.720269870758057,
      "epoch": 0.5932510885341074,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004971508689142184,
      "loss": 5.5935,
      "mean_token_accuracy": 0.15577241480350495,
      "num_tokens": 13594754.0,
      "step": 6540
    },
    {
      "entropy": 5.678057336807251,
      "epoch": 0.5937046444121916,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004971457337361356,
      "loss": 5.5339,
      "mean_token_accuracy": 0.16067788600921631,
      "num_tokens": 13604691.0,
      "step": 6545
    },
    {
      "entropy": 5.523071575164795,
      "epoch": 0.5941582002902758,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004971405939640225,
      "loss": 5.4263,
      "mean_token_accuracy": 0.16877733170986176,
      "num_tokens": 13613911.0,
      "step": 6550
    },
    {
      "entropy": 5.77616057395935,
      "epoch": 0.5946117561683599,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004971354495979854,
      "loss": 5.6417,
      "mean_token_accuracy": 0.1572772741317749,
      "num_tokens": 13624285.0,
      "step": 6555
    },
    {
      "entropy": 5.626784420013427,
      "epoch": 0.5950653120464441,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004971303006381305,
      "loss": 5.4779,
      "mean_token_accuracy": 0.16492559462785722,
      "num_tokens": 13635434.0,
      "step": 6560
    },
    {
      "entropy": 5.600918865203857,
      "epoch": 0.5955188679245284,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004971251470845643,
      "loss": 5.5534,
      "mean_token_accuracy": 0.16005088835954667,
      "num_tokens": 13645582.0,
      "step": 6565
    },
    {
      "entropy": 5.643901062011719,
      "epoch": 0.5959724238026125,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004971199889373935,
      "loss": 5.4986,
      "mean_token_accuracy": 0.16063801646232606,
      "num_tokens": 13655029.0,
      "step": 6570
    },
    {
      "entropy": 5.631298017501831,
      "epoch": 0.5964259796806967,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004971148261967249,
      "loss": 5.5052,
      "mean_token_accuracy": 0.1626351296901703,
      "num_tokens": 13665204.0,
      "step": 6575
    },
    {
      "entropy": 5.613391780853272,
      "epoch": 0.5968795355587808,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004971096588626649,
      "loss": 5.6002,
      "mean_token_accuracy": 0.15829678177833556,
      "num_tokens": 13676281.0,
      "step": 6580
    },
    {
      "entropy": 5.623476934432984,
      "epoch": 0.597333091436865,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004971044869353206,
      "loss": 5.5719,
      "mean_token_accuracy": 0.16425087451934814,
      "num_tokens": 13687144.0,
      "step": 6585
    },
    {
      "entropy": 5.665033578872681,
      "epoch": 0.5977866473149492,
      "grad_norm": 1.4375,
      "learning_rate": 0.000497099310414799,
      "loss": 5.5261,
      "mean_token_accuracy": 0.1613985627889633,
      "num_tokens": 13696371.0,
      "step": 6590
    },
    {
      "entropy": 5.68272442817688,
      "epoch": 0.5982402031930334,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004970941293012069,
      "loss": 5.4862,
      "mean_token_accuracy": 0.16283626109361649,
      "num_tokens": 13706900.0,
      "step": 6595
    },
    {
      "entropy": 5.613545656204224,
      "epoch": 0.5986937590711175,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004970889435946518,
      "loss": 5.5497,
      "mean_token_accuracy": 0.1644856095314026,
      "num_tokens": 13717874.0,
      "step": 6600
    },
    {
      "entropy": 5.616415214538574,
      "epoch": 0.5991473149492017,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004970837532952406,
      "loss": 5.5268,
      "mean_token_accuracy": 0.15892142057418823,
      "num_tokens": 13728020.0,
      "step": 6605
    },
    {
      "entropy": 5.709966087341309,
      "epoch": 0.599600870827286,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004970785584030808,
      "loss": 5.5682,
      "mean_token_accuracy": 0.16402234882116318,
      "num_tokens": 13738007.0,
      "step": 6610
    },
    {
      "entropy": 5.693265438079834,
      "epoch": 0.6000544267053701,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004970733589182799,
      "loss": 5.5832,
      "mean_token_accuracy": 0.15396313965320588,
      "num_tokens": 13749429.0,
      "step": 6615
    },
    {
      "entropy": 5.636207914352417,
      "epoch": 0.6005079825834543,
      "grad_norm": 1.625,
      "learning_rate": 0.0004970681548409452,
      "loss": 5.4346,
      "mean_token_accuracy": 0.17165576219558715,
      "num_tokens": 13760279.0,
      "step": 6620
    },
    {
      "entropy": 5.659836006164551,
      "epoch": 0.6009615384615384,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004970629461711846,
      "loss": 5.6745,
      "mean_token_accuracy": 0.15390522927045822,
      "num_tokens": 13770575.0,
      "step": 6625
    },
    {
      "entropy": 5.6224652290344235,
      "epoch": 0.6014150943396226,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004970577329091057,
      "loss": 5.508,
      "mean_token_accuracy": 0.16576763540506362,
      "num_tokens": 13781068.0,
      "step": 6630
    },
    {
      "entropy": 5.711998653411865,
      "epoch": 0.6018686502177069,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004970525150548162,
      "loss": 5.586,
      "mean_token_accuracy": 0.16630152910947799,
      "num_tokens": 13790745.0,
      "step": 6635
    },
    {
      "entropy": 5.654412794113159,
      "epoch": 0.602322206095791,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004970472926084242,
      "loss": 5.5201,
      "mean_token_accuracy": 0.16336238980293274,
      "num_tokens": 13800196.0,
      "step": 6640
    },
    {
      "entropy": 5.580491256713867,
      "epoch": 0.6027757619738752,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004970420655700375,
      "loss": 5.4165,
      "mean_token_accuracy": 0.17028857320547103,
      "num_tokens": 13809954.0,
      "step": 6645
    },
    {
      "entropy": 5.61370964050293,
      "epoch": 0.6032293178519593,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004970368339397643,
      "loss": 5.5829,
      "mean_token_accuracy": 0.1555069014430046,
      "num_tokens": 13820281.0,
      "step": 6650
    },
    {
      "entropy": 5.714426946640015,
      "epoch": 0.6036828737300436,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004970315977177129,
      "loss": 5.5259,
      "mean_token_accuracy": 0.16427477449178696,
      "num_tokens": 13830614.0,
      "step": 6655
    },
    {
      "entropy": 5.630530929565429,
      "epoch": 0.6041364296081277,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004970263569039914,
      "loss": 5.4939,
      "mean_token_accuracy": 0.1645730271935463,
      "num_tokens": 13840551.0,
      "step": 6660
    },
    {
      "entropy": 5.715430402755738,
      "epoch": 0.6045899854862119,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004970211114987083,
      "loss": 5.6767,
      "mean_token_accuracy": 0.15893951207399368,
      "num_tokens": 13851830.0,
      "step": 6665
    },
    {
      "entropy": 5.650357341766357,
      "epoch": 0.6050435413642961,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004970158615019719,
      "loss": 5.4799,
      "mean_token_accuracy": 0.16365663409233094,
      "num_tokens": 13862299.0,
      "step": 6670
    },
    {
      "entropy": 5.756088066101074,
      "epoch": 0.6054970972423802,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004970106069138912,
      "loss": 5.6015,
      "mean_token_accuracy": 0.15317101627588273,
      "num_tokens": 13872852.0,
      "step": 6675
    },
    {
      "entropy": 5.660994815826416,
      "epoch": 0.6059506531204645,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004970053477345743,
      "loss": 5.5423,
      "mean_token_accuracy": 0.16634648889303208,
      "num_tokens": 13883601.0,
      "step": 6680
    },
    {
      "entropy": 5.639829063415528,
      "epoch": 0.6064042089985486,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004970000839641304,
      "loss": 5.5516,
      "mean_token_accuracy": 0.16657450199127197,
      "num_tokens": 13894246.0,
      "step": 6685
    },
    {
      "entropy": 5.607128047943116,
      "epoch": 0.6068577648766328,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004969948156026681,
      "loss": 5.5731,
      "mean_token_accuracy": 0.16131135672330857,
      "num_tokens": 13904823.0,
      "step": 6690
    },
    {
      "entropy": 5.694171953201294,
      "epoch": 0.6073113207547169,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004969895426502965,
      "loss": 5.5796,
      "mean_token_accuracy": 0.16069512516260148,
      "num_tokens": 13915338.0,
      "step": 6695
    },
    {
      "entropy": 5.619790935516358,
      "epoch": 0.6077648766328012,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004969842651071246,
      "loss": 5.533,
      "mean_token_accuracy": 0.16402424424886702,
      "num_tokens": 13925838.0,
      "step": 6700
    },
    {
      "entropy": 5.7219775199890135,
      "epoch": 0.6082184325108854,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004969789829732615,
      "loss": 5.5567,
      "mean_token_accuracy": 0.15887922644615174,
      "num_tokens": 13936341.0,
      "step": 6705
    },
    {
      "entropy": 5.68441743850708,
      "epoch": 0.6086719883889695,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004969736962488164,
      "loss": 5.548,
      "mean_token_accuracy": 0.1660606235265732,
      "num_tokens": 13945438.0,
      "step": 6710
    },
    {
      "entropy": 5.621346664428711,
      "epoch": 0.6091255442670537,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004969684049338987,
      "loss": 5.4361,
      "mean_token_accuracy": 0.16126258224248885,
      "num_tokens": 13956029.0,
      "step": 6715
    },
    {
      "entropy": 5.666759252548218,
      "epoch": 0.6095791001451378,
      "grad_norm": 1.578125,
      "learning_rate": 0.000496963109028618,
      "loss": 5.5515,
      "mean_token_accuracy": 0.1595468908548355,
      "num_tokens": 13966498.0,
      "step": 6720
    },
    {
      "entropy": 5.627069568634033,
      "epoch": 0.6100326560232221,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004969578085330834,
      "loss": 5.5007,
      "mean_token_accuracy": 0.16338904947042465,
      "num_tokens": 13976997.0,
      "step": 6725
    },
    {
      "entropy": 5.680052900314331,
      "epoch": 0.6104862119013063,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004969525034474049,
      "loss": 5.5073,
      "mean_token_accuracy": 0.1636262834072113,
      "num_tokens": 13986664.0,
      "step": 6730
    },
    {
      "entropy": 5.596657180786133,
      "epoch": 0.6109397677793904,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000496947193771692,
      "loss": 5.5189,
      "mean_token_accuracy": 0.1638951927423477,
      "num_tokens": 13997828.0,
      "step": 6735
    },
    {
      "entropy": 5.608260631561279,
      "epoch": 0.6113933236574746,
      "grad_norm": 1.75,
      "learning_rate": 0.0004969418795060548,
      "loss": 5.4759,
      "mean_token_accuracy": 0.15902622193098068,
      "num_tokens": 14008293.0,
      "step": 6740
    },
    {
      "entropy": 5.629490756988526,
      "epoch": 0.6118468795355588,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004969365606506027,
      "loss": 5.444,
      "mean_token_accuracy": 0.16772356033325195,
      "num_tokens": 14018031.0,
      "step": 6745
    },
    {
      "entropy": 5.596383762359619,
      "epoch": 0.612300435413643,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004969312372054461,
      "loss": 5.4327,
      "mean_token_accuracy": 0.17152134627103804,
      "num_tokens": 14028013.0,
      "step": 6750
    },
    {
      "entropy": 5.552363443374634,
      "epoch": 0.6127539912917271,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004969259091706948,
      "loss": 5.5647,
      "mean_token_accuracy": 0.16048205494880677,
      "num_tokens": 14039787.0,
      "step": 6755
    },
    {
      "entropy": 5.588634395599366,
      "epoch": 0.6132075471698113,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004969205765464593,
      "loss": 5.4468,
      "mean_token_accuracy": 0.1649627611041069,
      "num_tokens": 14049223.0,
      "step": 6760
    },
    {
      "entropy": 5.660885524749756,
      "epoch": 0.6136611030478955,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004969152393328497,
      "loss": 5.493,
      "mean_token_accuracy": 0.16500439047813414,
      "num_tokens": 14059004.0,
      "step": 6765
    },
    {
      "entropy": 5.674933958053589,
      "epoch": 0.6141146589259797,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004969098975299764,
      "loss": 5.592,
      "mean_token_accuracy": 0.15621163845062255,
      "num_tokens": 14069615.0,
      "step": 6770
    },
    {
      "entropy": 5.6336719512939455,
      "epoch": 0.6145682148040639,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004969045511379499,
      "loss": 5.5297,
      "mean_token_accuracy": 0.16255359649658202,
      "num_tokens": 14078886.0,
      "step": 6775
    },
    {
      "entropy": 5.673841714859009,
      "epoch": 0.615021770682148,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004968992001568806,
      "loss": 5.5974,
      "mean_token_accuracy": 0.1564471200108528,
      "num_tokens": 14089021.0,
      "step": 6780
    },
    {
      "entropy": 5.651300048828125,
      "epoch": 0.6154753265602322,
      "grad_norm": 2.6875,
      "learning_rate": 0.0004968938445868795,
      "loss": 5.4176,
      "mean_token_accuracy": 0.17256750017404557,
      "num_tokens": 14098316.0,
      "step": 6785
    },
    {
      "entropy": 5.591040754318238,
      "epoch": 0.6159288824383164,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004968884844280569,
      "loss": 5.5163,
      "mean_token_accuracy": 0.1584145352244377,
      "num_tokens": 14108874.0,
      "step": 6790
    },
    {
      "entropy": 5.641624498367309,
      "epoch": 0.6163824383164006,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004968831196805242,
      "loss": 5.5646,
      "mean_token_accuracy": 0.16046148389577866,
      "num_tokens": 14119613.0,
      "step": 6795
    },
    {
      "entropy": 5.676394128799439,
      "epoch": 0.6168359941944848,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004968777503443918,
      "loss": 5.5095,
      "mean_token_accuracy": 0.16080328971147537,
      "num_tokens": 14129085.0,
      "step": 6800
    },
    {
      "entropy": 5.512157869338989,
      "epoch": 0.6172895500725689,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004968723764197711,
      "loss": 5.4132,
      "mean_token_accuracy": 0.1770444855093956,
      "num_tokens": 14138461.0,
      "step": 6805
    },
    {
      "entropy": 5.594473648071289,
      "epoch": 0.6177431059506531,
      "grad_norm": 1.640625,
      "learning_rate": 0.000496866997906773,
      "loss": 5.4897,
      "mean_token_accuracy": 0.16651539504528046,
      "num_tokens": 14149126.0,
      "step": 6810
    },
    {
      "entropy": 5.635282516479492,
      "epoch": 0.6181966618287373,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004968616148055089,
      "loss": 5.4823,
      "mean_token_accuracy": 0.1690201073884964,
      "num_tokens": 14160068.0,
      "step": 6815
    },
    {
      "entropy": 5.627933597564697,
      "epoch": 0.6186502177068215,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004968562271160901,
      "loss": 5.5869,
      "mean_token_accuracy": 0.166331148147583,
      "num_tokens": 14170288.0,
      "step": 6820
    },
    {
      "entropy": 5.728124380111694,
      "epoch": 0.6191037735849056,
      "grad_norm": 1.59375,
      "learning_rate": 0.000496850834838628,
      "loss": 5.5757,
      "mean_token_accuracy": 0.16276340931653976,
      "num_tokens": 14181231.0,
      "step": 6825
    },
    {
      "entropy": 5.669750642776489,
      "epoch": 0.6195573294629898,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004968454379732342,
      "loss": 5.49,
      "mean_token_accuracy": 0.16779735684394836,
      "num_tokens": 14191715.0,
      "step": 6830
    },
    {
      "entropy": 5.602129220962524,
      "epoch": 0.6200108853410741,
      "grad_norm": 2.03125,
      "learning_rate": 0.00049684003652002,
      "loss": 5.4878,
      "mean_token_accuracy": 0.1612647980451584,
      "num_tokens": 14203416.0,
      "step": 6835
    },
    {
      "entropy": 5.548123359680176,
      "epoch": 0.6204644412191582,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004968346304790973,
      "loss": 5.4266,
      "mean_token_accuracy": 0.17077218294143676,
      "num_tokens": 14213658.0,
      "step": 6840
    },
    {
      "entropy": 5.668492031097412,
      "epoch": 0.6209179970972424,
      "grad_norm": 4.78125,
      "learning_rate": 0.0004968292198505781,
      "loss": 5.6062,
      "mean_token_accuracy": 0.15588149279356003,
      "num_tokens": 14223956.0,
      "step": 6845
    },
    {
      "entropy": 5.668875312805175,
      "epoch": 0.6213715529753265,
      "grad_norm": 2.015625,
      "learning_rate": 0.000496823804634574,
      "loss": 5.612,
      "mean_token_accuracy": 0.1597575768828392,
      "num_tokens": 14236801.0,
      "step": 6850
    },
    {
      "entropy": 5.602407693862915,
      "epoch": 0.6218251088534107,
      "grad_norm": 2.125,
      "learning_rate": 0.0004968183848311972,
      "loss": 5.4082,
      "mean_token_accuracy": 0.17220478355884553,
      "num_tokens": 14247382.0,
      "step": 6855
    },
    {
      "entropy": 5.683701372146606,
      "epoch": 0.622278664731495,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004968129604405596,
      "loss": 5.4997,
      "mean_token_accuracy": 0.16317506432533263,
      "num_tokens": 14258627.0,
      "step": 6860
    },
    {
      "entropy": 5.559516525268554,
      "epoch": 0.6227322206095791,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004968075314627734,
      "loss": 5.5014,
      "mean_token_accuracy": 0.16878147274255753,
      "num_tokens": 14269013.0,
      "step": 6865
    },
    {
      "entropy": 5.602084016799926,
      "epoch": 0.6231857764876633,
      "grad_norm": 2.171875,
      "learning_rate": 0.000496802097897951,
      "loss": 5.4929,
      "mean_token_accuracy": 0.1642188236117363,
      "num_tokens": 14279620.0,
      "step": 6870
    },
    {
      "entropy": 5.647443389892578,
      "epoch": 0.6236393323657474,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004967966597462046,
      "loss": 5.506,
      "mean_token_accuracy": 0.17011392563581468,
      "num_tokens": 14289448.0,
      "step": 6875
    },
    {
      "entropy": 5.62725305557251,
      "epoch": 0.6240928882438317,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004967912170076469,
      "loss": 5.5123,
      "mean_token_accuracy": 0.16029064506292343,
      "num_tokens": 14299999.0,
      "step": 6880
    },
    {
      "entropy": 5.5908074378967285,
      "epoch": 0.6245464441219158,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004967857696823901,
      "loss": 5.488,
      "mean_token_accuracy": 0.1604388862848282,
      "num_tokens": 14310384.0,
      "step": 6885
    },
    {
      "entropy": 5.666422319412232,
      "epoch": 0.625,
      "grad_norm": 1.625,
      "learning_rate": 0.0004967803177705472,
      "loss": 5.5893,
      "mean_token_accuracy": 0.15651167333126068,
      "num_tokens": 14320644.0,
      "step": 6890
    },
    {
      "entropy": 5.684583044052124,
      "epoch": 0.6254535558780842,
      "grad_norm": 1.75,
      "learning_rate": 0.0004967748612722308,
      "loss": 5.5494,
      "mean_token_accuracy": 0.16236335188150405,
      "num_tokens": 14330077.0,
      "step": 6895
    },
    {
      "entropy": 5.7078063011169435,
      "epoch": 0.6259071117561683,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004967694001875538,
      "loss": 5.6323,
      "mean_token_accuracy": 0.1552937150001526,
      "num_tokens": 14341549.0,
      "step": 6900
    },
    {
      "entropy": 5.68909273147583,
      "epoch": 0.6263606676342526,
      "grad_norm": 1.890625,
      "learning_rate": 0.000496763934516629,
      "loss": 5.5059,
      "mean_token_accuracy": 0.1664724200963974,
      "num_tokens": 14352146.0,
      "step": 6905
    },
    {
      "entropy": 5.614871978759766,
      "epoch": 0.6268142235123367,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004967584642595696,
      "loss": 5.5244,
      "mean_token_accuracy": 0.15754454135894774,
      "num_tokens": 14362521.0,
      "step": 6910
    },
    {
      "entropy": 5.698942518234253,
      "epoch": 0.6272677793904209,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004967529894164887,
      "loss": 5.5712,
      "mean_token_accuracy": 0.16412983238697051,
      "num_tokens": 14373216.0,
      "step": 6915
    },
    {
      "entropy": 5.63185224533081,
      "epoch": 0.627721335268505,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004967475099874992,
      "loss": 5.4773,
      "mean_token_accuracy": 0.17082804292440415,
      "num_tokens": 14383104.0,
      "step": 6920
    },
    {
      "entropy": 5.556405639648437,
      "epoch": 0.6281748911465893,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000496742025972715,
      "loss": 5.5112,
      "mean_token_accuracy": 0.16977826058864592,
      "num_tokens": 14393854.0,
      "step": 6925
    },
    {
      "entropy": 5.622954559326172,
      "epoch": 0.6286284470246735,
      "grad_norm": 2.03125,
      "learning_rate": 0.000496736537372249,
      "loss": 5.4498,
      "mean_token_accuracy": 0.17209360599517823,
      "num_tokens": 14403213.0,
      "step": 6930
    },
    {
      "entropy": 5.703037691116333,
      "epoch": 0.6290820029027576,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004967310441862151,
      "loss": 5.6666,
      "mean_token_accuracy": 0.1583329074084759,
      "num_tokens": 14412792.0,
      "step": 6935
    },
    {
      "entropy": 5.676712512969971,
      "epoch": 0.6295355587808418,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004967255464147265,
      "loss": 5.5133,
      "mean_token_accuracy": 0.16898928731679916,
      "num_tokens": 14423061.0,
      "step": 6940
    },
    {
      "entropy": 5.653528928756714,
      "epoch": 0.6299891146589259,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004967200440578972,
      "loss": 5.4613,
      "mean_token_accuracy": 0.16783155202865602,
      "num_tokens": 14433338.0,
      "step": 6945
    },
    {
      "entropy": 5.568450927734375,
      "epoch": 0.6304426705370102,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004967145371158408,
      "loss": 5.4547,
      "mean_token_accuracy": 0.17319378405809402,
      "num_tokens": 14444124.0,
      "step": 6950
    },
    {
      "entropy": 5.613980054855347,
      "epoch": 0.6308962264150944,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004967090255886714,
      "loss": 5.5855,
      "mean_token_accuracy": 0.16500539034605027,
      "num_tokens": 14453868.0,
      "step": 6955
    },
    {
      "entropy": 5.628482532501221,
      "epoch": 0.6313497822931785,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004967035094765028,
      "loss": 5.4726,
      "mean_token_accuracy": 0.16420088410377504,
      "num_tokens": 14465015.0,
      "step": 6960
    },
    {
      "entropy": 5.596653985977173,
      "epoch": 0.6318033381712627,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004966979887794491,
      "loss": 5.437,
      "mean_token_accuracy": 0.17249543368816375,
      "num_tokens": 14475523.0,
      "step": 6965
    },
    {
      "entropy": 5.6447957992553714,
      "epoch": 0.6322568940493469,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004966924634976245,
      "loss": 5.5063,
      "mean_token_accuracy": 0.1661619022488594,
      "num_tokens": 14485523.0,
      "step": 6970
    },
    {
      "entropy": 5.697253131866455,
      "epoch": 0.6327104499274311,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004966869336311433,
      "loss": 5.5762,
      "mean_token_accuracy": 0.1666168376803398,
      "num_tokens": 14495237.0,
      "step": 6975
    },
    {
      "entropy": 5.640621852874756,
      "epoch": 0.6331640058055152,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004966813991801197,
      "loss": 5.6977,
      "mean_token_accuracy": 0.15432378351688386,
      "num_tokens": 14506431.0,
      "step": 6980
    },
    {
      "entropy": 5.727398633956909,
      "epoch": 0.6336175616835994,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004966758601446684,
      "loss": 5.6151,
      "mean_token_accuracy": 0.15618521571159363,
      "num_tokens": 14517445.0,
      "step": 6985
    },
    {
      "entropy": 5.673369693756103,
      "epoch": 0.6340711175616836,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004966703165249038,
      "loss": 5.5485,
      "mean_token_accuracy": 0.1577815368771553,
      "num_tokens": 14527818.0,
      "step": 6990
    },
    {
      "entropy": 5.660148620605469,
      "epoch": 0.6345246734397678,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004966647683209406,
      "loss": 5.6308,
      "mean_token_accuracy": 0.1515124946832657,
      "num_tokens": 14538235.0,
      "step": 6995
    },
    {
      "entropy": 5.7893064498901365,
      "epoch": 0.634978229317852,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004966592155328935,
      "loss": 5.6421,
      "mean_token_accuracy": 0.15409057587385178,
      "num_tokens": 14549784.0,
      "step": 7000
    },
    {
      "entropy": 5.675354909896851,
      "epoch": 0.6354317851959361,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004966536581608771,
      "loss": 5.6258,
      "mean_token_accuracy": 0.15651679188013076,
      "num_tokens": 14561057.0,
      "step": 7005
    },
    {
      "entropy": 5.6757800579071045,
      "epoch": 0.6358853410740203,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004966480962050069,
      "loss": 5.5419,
      "mean_token_accuracy": 0.1694130152463913,
      "num_tokens": 14570859.0,
      "step": 7010
    },
    {
      "entropy": 5.654591512680054,
      "epoch": 0.6363388969521045,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004966425296653974,
      "loss": 5.548,
      "mean_token_accuracy": 0.16952096819877624,
      "num_tokens": 14581066.0,
      "step": 7015
    },
    {
      "entropy": 5.601092910766601,
      "epoch": 0.6367924528301887,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000496636958542164,
      "loss": 5.5125,
      "mean_token_accuracy": 0.16194411963224412,
      "num_tokens": 14590709.0,
      "step": 7020
    },
    {
      "entropy": 5.6308215141296385,
      "epoch": 0.6372460087082729,
      "grad_norm": 1.375,
      "learning_rate": 0.0004966313828354218,
      "loss": 5.5009,
      "mean_token_accuracy": 0.1659430295228958,
      "num_tokens": 14600849.0,
      "step": 7025
    },
    {
      "entropy": 5.604932880401611,
      "epoch": 0.637699564586357,
      "grad_norm": 1.71875,
      "learning_rate": 0.000496625802545286,
      "loss": 5.3908,
      "mean_token_accuracy": 0.16500746309757233,
      "num_tokens": 14611427.0,
      "step": 7030
    },
    {
      "entropy": 5.651730728149414,
      "epoch": 0.6381531204644412,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004966202176718723,
      "loss": 5.5443,
      "mean_token_accuracy": 0.16733615100383759,
      "num_tokens": 14623860.0,
      "step": 7035
    },
    {
      "entropy": 5.621776580810547,
      "epoch": 0.6386066763425254,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004966146282152959,
      "loss": 5.4668,
      "mean_token_accuracy": 0.1645897924900055,
      "num_tokens": 14633481.0,
      "step": 7040
    },
    {
      "entropy": 5.639799451828003,
      "epoch": 0.6390602322206096,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004966090341756726,
      "loss": 5.4407,
      "mean_token_accuracy": 0.173050656914711,
      "num_tokens": 14643519.0,
      "step": 7045
    },
    {
      "entropy": 5.589098262786865,
      "epoch": 0.6395137880986937,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000496603435553118,
      "loss": 5.4535,
      "mean_token_accuracy": 0.16662810295820235,
      "num_tokens": 14654570.0,
      "step": 7050
    },
    {
      "entropy": 5.610861587524414,
      "epoch": 0.6399673439767779,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004965978323477478,
      "loss": 5.4955,
      "mean_token_accuracy": 0.15878622084856034,
      "num_tokens": 14665596.0,
      "step": 7055
    },
    {
      "entropy": 5.670983934402466,
      "epoch": 0.6404208998548622,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000496592224559678,
      "loss": 5.4431,
      "mean_token_accuracy": 0.17061298489570617,
      "num_tokens": 14675660.0,
      "step": 7060
    },
    {
      "entropy": 5.668635034561158,
      "epoch": 0.6408744557329463,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004965866121890245,
      "loss": 5.6024,
      "mean_token_accuracy": 0.161976358294487,
      "num_tokens": 14685229.0,
      "step": 7065
    },
    {
      "entropy": 5.66521725654602,
      "epoch": 0.6413280116110305,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004965809952359034,
      "loss": 5.5157,
      "mean_token_accuracy": 0.15676263421773912,
      "num_tokens": 14695672.0,
      "step": 7070
    },
    {
      "entropy": 5.630953550338745,
      "epoch": 0.6417815674891146,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004965753737004309,
      "loss": 5.4415,
      "mean_token_accuracy": 0.17113818228244781,
      "num_tokens": 14705160.0,
      "step": 7075
    },
    {
      "entropy": 5.616978645324707,
      "epoch": 0.6422351233671988,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004965697475827231,
      "loss": 5.4768,
      "mean_token_accuracy": 0.1638179302215576,
      "num_tokens": 14715487.0,
      "step": 7080
    },
    {
      "entropy": 5.648258352279663,
      "epoch": 0.6426886792452831,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004965641168828965,
      "loss": 5.5295,
      "mean_token_accuracy": 0.16361365467309952,
      "num_tokens": 14726987.0,
      "step": 7085
    },
    {
      "entropy": 5.580990219116211,
      "epoch": 0.6431422351233672,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004965584816010676,
      "loss": 5.4892,
      "mean_token_accuracy": 0.16957174092531205,
      "num_tokens": 14737810.0,
      "step": 7090
    },
    {
      "entropy": 5.700578737258911,
      "epoch": 0.6435957910014514,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004965528417373528,
      "loss": 5.5038,
      "mean_token_accuracy": 0.15920886397361755,
      "num_tokens": 14748997.0,
      "step": 7095
    },
    {
      "entropy": 5.654855442047119,
      "epoch": 0.6440493468795355,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004965471972918688,
      "loss": 5.52,
      "mean_token_accuracy": 0.1612129181623459,
      "num_tokens": 14760439.0,
      "step": 7100
    },
    {
      "entropy": 5.593002033233643,
      "epoch": 0.6445029027576198,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004965415482647322,
      "loss": 5.4461,
      "mean_token_accuracy": 0.16583144515752793,
      "num_tokens": 14770828.0,
      "step": 7105
    },
    {
      "entropy": 5.570415592193603,
      "epoch": 0.6449564586357039,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00049653589465606,
      "loss": 5.5725,
      "mean_token_accuracy": 0.16377214789390565,
      "num_tokens": 14781753.0,
      "step": 7110
    },
    {
      "entropy": 5.717770576477051,
      "epoch": 0.6454100145137881,
      "grad_norm": 1.9609375,
      "learning_rate": 0.000496530236465969,
      "loss": 5.5756,
      "mean_token_accuracy": 0.159645214676857,
      "num_tokens": 14792241.0,
      "step": 7115
    },
    {
      "entropy": 5.6190179824829105,
      "epoch": 0.6458635703918723,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004965245736945763,
      "loss": 5.414,
      "mean_token_accuracy": 0.17340738624334334,
      "num_tokens": 14802580.0,
      "step": 7120
    },
    {
      "entropy": 5.597347974777222,
      "epoch": 0.6463171262699564,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004965189063419989,
      "loss": 5.6069,
      "mean_token_accuracy": 0.16257513761520387,
      "num_tokens": 14813103.0,
      "step": 7125
    },
    {
      "entropy": 5.628874015808106,
      "epoch": 0.6467706821480407,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004965132344083541,
      "loss": 5.5043,
      "mean_token_accuracy": 0.16554489135742187,
      "num_tokens": 14823334.0,
      "step": 7130
    },
    {
      "entropy": 5.747768974304199,
      "epoch": 0.6472242380261248,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004965075578937593,
      "loss": 5.5237,
      "mean_token_accuracy": 0.15984877645969392,
      "num_tokens": 14834850.0,
      "step": 7135
    },
    {
      "entropy": 5.638379621505737,
      "epoch": 0.647677793904209,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004965018767983316,
      "loss": 5.5399,
      "mean_token_accuracy": 0.1638460412621498,
      "num_tokens": 14844276.0,
      "step": 7140
    },
    {
      "entropy": 5.518471670150757,
      "epoch": 0.6481313497822931,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004964961911221886,
      "loss": 5.394,
      "mean_token_accuracy": 0.17712354212999343,
      "num_tokens": 14854352.0,
      "step": 7145
    },
    {
      "entropy": 5.6490014553070065,
      "epoch": 0.6485849056603774,
      "grad_norm": 1.8125,
      "learning_rate": 0.000496490500865448,
      "loss": 5.5667,
      "mean_token_accuracy": 0.1627000540494919,
      "num_tokens": 14864395.0,
      "step": 7150
    },
    {
      "entropy": 5.6571167469024655,
      "epoch": 0.6490384615384616,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004964848060282275,
      "loss": 5.4477,
      "mean_token_accuracy": 0.17021121829748154,
      "num_tokens": 14874373.0,
      "step": 7155
    },
    {
      "entropy": 5.620741987228394,
      "epoch": 0.6494920174165457,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004964791066106447,
      "loss": 5.4425,
      "mean_token_accuracy": 0.1613517627120018,
      "num_tokens": 14884932.0,
      "step": 7160
    },
    {
      "entropy": 5.63075442314148,
      "epoch": 0.6499455732946299,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004964734026128175,
      "loss": 5.6013,
      "mean_token_accuracy": 0.1577137976884842,
      "num_tokens": 14896254.0,
      "step": 7165
    },
    {
      "entropy": 5.713843965530396,
      "epoch": 0.650399129172714,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004964676940348638,
      "loss": 5.5376,
      "mean_token_accuracy": 0.16059393137693406,
      "num_tokens": 14907392.0,
      "step": 7170
    },
    {
      "entropy": 5.73220853805542,
      "epoch": 0.6508526850507983,
      "grad_norm": 1.8828125,
      "learning_rate": 0.000496461980876902,
      "loss": 5.6427,
      "mean_token_accuracy": 0.16424211263656616,
      "num_tokens": 14917676.0,
      "step": 7175
    },
    {
      "entropy": 5.627636861801148,
      "epoch": 0.6513062409288825,
      "grad_norm": 2.0,
      "learning_rate": 0.0004964562631390498,
      "loss": 5.4909,
      "mean_token_accuracy": 0.16617691963911058,
      "num_tokens": 14928500.0,
      "step": 7180
    },
    {
      "entropy": 5.609093856811524,
      "epoch": 0.6517597968069666,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004964505408214257,
      "loss": 5.449,
      "mean_token_accuracy": 0.17101617455482482,
      "num_tokens": 14938970.0,
      "step": 7185
    },
    {
      "entropy": 5.616778039932251,
      "epoch": 0.6522133526850508,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000496444813924148,
      "loss": 5.5116,
      "mean_token_accuracy": 0.16331736892461776,
      "num_tokens": 14949017.0,
      "step": 7190
    },
    {
      "entropy": 5.654310798645019,
      "epoch": 0.652666908563135,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004964390824473351,
      "loss": 5.4473,
      "mean_token_accuracy": 0.1724479764699936,
      "num_tokens": 14958529.0,
      "step": 7195
    },
    {
      "entropy": 5.597932052612305,
      "epoch": 0.6531204644412192,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004964333463911056,
      "loss": 5.5422,
      "mean_token_accuracy": 0.15932074636220933,
      "num_tokens": 14968044.0,
      "step": 7200
    },
    {
      "entropy": 5.65596113204956,
      "epoch": 0.6535740203193033,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004964276057555779,
      "loss": 5.566,
      "mean_token_accuracy": 0.16014891862869263,
      "num_tokens": 14977867.0,
      "step": 7205
    },
    {
      "entropy": 5.783083295822143,
      "epoch": 0.6540275761973875,
      "grad_norm": 1.34375,
      "learning_rate": 0.000496421860540871,
      "loss": 5.5475,
      "mean_token_accuracy": 0.16266472786664962,
      "num_tokens": 14986936.0,
      "step": 7210
    },
    {
      "entropy": 5.61876163482666,
      "epoch": 0.6544811320754716,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004964161107471036,
      "loss": 5.5228,
      "mean_token_accuracy": 0.1681314930319786,
      "num_tokens": 14996681.0,
      "step": 7215
    },
    {
      "entropy": 5.598669624328613,
      "epoch": 0.6549346879535559,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004964103563743946,
      "loss": 5.5116,
      "mean_token_accuracy": 0.1631477177143097,
      "num_tokens": 15007100.0,
      "step": 7220
    },
    {
      "entropy": 5.694092845916748,
      "epoch": 0.6553882438316401,
      "grad_norm": 1.46875,
      "learning_rate": 0.000496404597422863,
      "loss": 5.5868,
      "mean_token_accuracy": 0.1644970640540123,
      "num_tokens": 15017452.0,
      "step": 7225
    },
    {
      "entropy": 5.620909023284912,
      "epoch": 0.6558417997097242,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000496398833892628,
      "loss": 5.4344,
      "mean_token_accuracy": 0.16442154794931413,
      "num_tokens": 15028362.0,
      "step": 7230
    },
    {
      "entropy": 5.638983058929443,
      "epoch": 0.6562953555878084,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004963930657838086,
      "loss": 5.5293,
      "mean_token_accuracy": 0.1629357695579529,
      "num_tokens": 15038691.0,
      "step": 7235
    },
    {
      "entropy": 5.651097059249878,
      "epoch": 0.6567489114658926,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004963872930965242,
      "loss": 5.5059,
      "mean_token_accuracy": 0.16842897832393647,
      "num_tokens": 15050037.0,
      "step": 7240
    },
    {
      "entropy": 5.642683744430542,
      "epoch": 0.6572024673439768,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004963815158308941,
      "loss": 5.4735,
      "mean_token_accuracy": 0.16414920687675477,
      "num_tokens": 15059793.0,
      "step": 7245
    },
    {
      "entropy": 5.633060359954834,
      "epoch": 0.657656023222061,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004963757339870379,
      "loss": 5.5821,
      "mean_token_accuracy": 0.1542697638273239,
      "num_tokens": 15070500.0,
      "step": 7250
    },
    {
      "entropy": 5.718131399154663,
      "epoch": 0.6581095791001451,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004963699475650751,
      "loss": 5.625,
      "mean_token_accuracy": 0.15848404318094253,
      "num_tokens": 15081109.0,
      "step": 7255
    },
    {
      "entropy": 5.629377222061157,
      "epoch": 0.6585631349782293,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004963641565651255,
      "loss": 5.5291,
      "mean_token_accuracy": 0.16712549030780793,
      "num_tokens": 15092315.0,
      "step": 7260
    },
    {
      "entropy": 5.615194702148438,
      "epoch": 0.6590166908563135,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004963583609873084,
      "loss": 5.5868,
      "mean_token_accuracy": 0.16172214299440385,
      "num_tokens": 15103957.0,
      "step": 7265
    },
    {
      "entropy": 5.7270472049713135,
      "epoch": 0.6594702467343977,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004963525608317443,
      "loss": 5.5518,
      "mean_token_accuracy": 0.1632919952273369,
      "num_tokens": 15114996.0,
      "step": 7270
    },
    {
      "entropy": 5.7020470142364506,
      "epoch": 0.6599238026124818,
      "grad_norm": 1.5,
      "learning_rate": 0.0004963467560985527,
      "loss": 5.5199,
      "mean_token_accuracy": 0.16006214767694474,
      "num_tokens": 15125098.0,
      "step": 7275
    },
    {
      "entropy": 5.632985973358155,
      "epoch": 0.660377358490566,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004963409467878538,
      "loss": 5.496,
      "mean_token_accuracy": 0.15927427560091018,
      "num_tokens": 15134746.0,
      "step": 7280
    },
    {
      "entropy": 5.706460189819336,
      "epoch": 0.6608309143686503,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004963351328997678,
      "loss": 5.5321,
      "mean_token_accuracy": 0.15826075226068498,
      "num_tokens": 15145698.0,
      "step": 7285
    },
    {
      "entropy": 5.623006582260132,
      "epoch": 0.6612844702467344,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004963293144344148,
      "loss": 5.5806,
      "mean_token_accuracy": 0.15674502849578859,
      "num_tokens": 15156026.0,
      "step": 7290
    },
    {
      "entropy": 5.632738924026489,
      "epoch": 0.6617380261248186,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004963234913919151,
      "loss": 5.4957,
      "mean_token_accuracy": 0.15496615469455718,
      "num_tokens": 15165900.0,
      "step": 7295
    },
    {
      "entropy": 5.635572147369385,
      "epoch": 0.6621915820029027,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004963176637723893,
      "loss": 5.572,
      "mean_token_accuracy": 0.1640890881419182,
      "num_tokens": 15176693.0,
      "step": 7300
    },
    {
      "entropy": 5.696575164794922,
      "epoch": 0.6626451378809869,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004963118315759578,
      "loss": 5.6022,
      "mean_token_accuracy": 0.14927273318171502,
      "num_tokens": 15186535.0,
      "step": 7305
    },
    {
      "entropy": 5.7064004898071286,
      "epoch": 0.6630986937590712,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004963059948027411,
      "loss": 5.5269,
      "mean_token_accuracy": 0.1611385740339756,
      "num_tokens": 15196019.0,
      "step": 7310
    },
    {
      "entropy": 5.552914619445801,
      "epoch": 0.6635522496371553,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004963001534528601,
      "loss": 5.3816,
      "mean_token_accuracy": 0.1719992369413376,
      "num_tokens": 15205273.0,
      "step": 7315
    },
    {
      "entropy": 5.627516508102417,
      "epoch": 0.6640058055152395,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004962943075264355,
      "loss": 5.5184,
      "mean_token_accuracy": 0.16372389495372772,
      "num_tokens": 15216716.0,
      "step": 7320
    },
    {
      "entropy": 5.636356401443481,
      "epoch": 0.6644593613933236,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004962884570235884,
      "loss": 5.5794,
      "mean_token_accuracy": 0.15782008916139603,
      "num_tokens": 15227306.0,
      "step": 7325
    },
    {
      "entropy": 5.564645528793335,
      "epoch": 0.6649129172714079,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004962826019444394,
      "loss": 5.455,
      "mean_token_accuracy": 0.16822835057973862,
      "num_tokens": 15237302.0,
      "step": 7330
    },
    {
      "entropy": 5.635433626174927,
      "epoch": 0.665366473149492,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004962767422891099,
      "loss": 5.4674,
      "mean_token_accuracy": 0.16430780440568923,
      "num_tokens": 15247064.0,
      "step": 7335
    },
    {
      "entropy": 5.672469520568848,
      "epoch": 0.6658200290275762,
      "grad_norm": 1.453125,
      "learning_rate": 0.000496270878057721,
      "loss": 5.5503,
      "mean_token_accuracy": 0.16559139490127564,
      "num_tokens": 15257006.0,
      "step": 7340
    },
    {
      "entropy": 5.680461406707764,
      "epoch": 0.6662735849056604,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004962650092503939,
      "loss": 5.5328,
      "mean_token_accuracy": 0.1673779770731926,
      "num_tokens": 15266720.0,
      "step": 7345
    },
    {
      "entropy": 5.6296360969543455,
      "epoch": 0.6667271407837445,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004962591358672502,
      "loss": 5.5296,
      "mean_token_accuracy": 0.16174665689468384,
      "num_tokens": 15277330.0,
      "step": 7350
    },
    {
      "entropy": 5.600244045257568,
      "epoch": 0.6671806966618288,
      "grad_norm": 1.40625,
      "learning_rate": 0.000496253257908411,
      "loss": 5.4762,
      "mean_token_accuracy": 0.16490522772073746,
      "num_tokens": 15287401.0,
      "step": 7355
    },
    {
      "entropy": 5.532177495956421,
      "epoch": 0.6676342525399129,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004962473753739981,
      "loss": 5.4373,
      "mean_token_accuracy": 0.16928917169570923,
      "num_tokens": 15297426.0,
      "step": 7360
    },
    {
      "entropy": 5.651616144180298,
      "epoch": 0.6680878084179971,
      "grad_norm": 1.6875,
      "learning_rate": 0.000496241488264133,
      "loss": 5.4184,
      "mean_token_accuracy": 0.16742407381534577,
      "num_tokens": 15308412.0,
      "step": 7365
    },
    {
      "entropy": 5.6203423023223875,
      "epoch": 0.6685413642960812,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004962355965789377,
      "loss": 5.4491,
      "mean_token_accuracy": 0.16204553544521333,
      "num_tokens": 15318821.0,
      "step": 7370
    },
    {
      "entropy": 5.587005376815796,
      "epoch": 0.6689949201741655,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004962297003185338,
      "loss": 5.4998,
      "mean_token_accuracy": 0.1618948683142662,
      "num_tokens": 15328737.0,
      "step": 7375
    },
    {
      "entropy": 5.670815324783325,
      "epoch": 0.6694484760522497,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004962237994830433,
      "loss": 5.6049,
      "mean_token_accuracy": 0.1564209431409836,
      "num_tokens": 15339695.0,
      "step": 7380
    },
    {
      "entropy": 5.707769441604614,
      "epoch": 0.6699020319303338,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004962178940725883,
      "loss": 5.55,
      "mean_token_accuracy": 0.16244056969881057,
      "num_tokens": 15350270.0,
      "step": 7385
    },
    {
      "entropy": 5.664539051055908,
      "epoch": 0.670355587808418,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004962119840872908,
      "loss": 5.5434,
      "mean_token_accuracy": 0.1645205706357956,
      "num_tokens": 15360986.0,
      "step": 7390
    },
    {
      "entropy": 5.645076084136963,
      "epoch": 0.6708091436865021,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004962060695272733,
      "loss": 5.5294,
      "mean_token_accuracy": 0.16415667831897734,
      "num_tokens": 15371786.0,
      "step": 7395
    },
    {
      "entropy": 5.63773102760315,
      "epoch": 0.6712626995645864,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004962001503926577,
      "loss": 5.5013,
      "mean_token_accuracy": 0.16971592009067535,
      "num_tokens": 15382960.0,
      "step": 7400
    },
    {
      "entropy": 5.576283693313599,
      "epoch": 0.6717162554426706,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004961942266835667,
      "loss": 5.4904,
      "mean_token_accuracy": 0.1673077717423439,
      "num_tokens": 15392807.0,
      "step": 7405
    },
    {
      "entropy": 5.650079441070557,
      "epoch": 0.6721698113207547,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004961882984001227,
      "loss": 5.5379,
      "mean_token_accuracy": 0.1599237620830536,
      "num_tokens": 15403712.0,
      "step": 7410
    },
    {
      "entropy": 5.676372385025024,
      "epoch": 0.6726233671988389,
      "grad_norm": 1.625,
      "learning_rate": 0.0004961823655424484,
      "loss": 5.5965,
      "mean_token_accuracy": 0.16035202890634537,
      "num_tokens": 15414839.0,
      "step": 7415
    },
    {
      "entropy": 5.642154216766357,
      "epoch": 0.6730769230769231,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004961764281106664,
      "loss": 5.4736,
      "mean_token_accuracy": 0.1691312834620476,
      "num_tokens": 15425320.0,
      "step": 7420
    },
    {
      "entropy": 5.652766180038452,
      "epoch": 0.6735304789550073,
      "grad_norm": 1.75,
      "learning_rate": 0.0004961704861048995,
      "loss": 5.4818,
      "mean_token_accuracy": 0.16118890792131424,
      "num_tokens": 15435476.0,
      "step": 7425
    },
    {
      "entropy": 5.64156289100647,
      "epoch": 0.6739840348330914,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004961645395252705,
      "loss": 5.5884,
      "mean_token_accuracy": 0.16087498962879182,
      "num_tokens": 15445786.0,
      "step": 7430
    },
    {
      "entropy": 5.700147104263306,
      "epoch": 0.6744375907111756,
      "grad_norm": 1.625,
      "learning_rate": 0.0004961585883719025,
      "loss": 5.5913,
      "mean_token_accuracy": 0.1652859181165695,
      "num_tokens": 15455915.0,
      "step": 7435
    },
    {
      "entropy": 5.735110139846801,
      "epoch": 0.6748911465892597,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004961526326449186,
      "loss": 5.5602,
      "mean_token_accuracy": 0.16309167593717575,
      "num_tokens": 15466523.0,
      "step": 7440
    },
    {
      "entropy": 5.655019807815552,
      "epoch": 0.675344702467344,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004961466723444419,
      "loss": 5.5459,
      "mean_token_accuracy": 0.16104988008737564,
      "num_tokens": 15476821.0,
      "step": 7445
    },
    {
      "entropy": 5.603958559036255,
      "epoch": 0.6757982583454282,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004961407074705955,
      "loss": 5.4841,
      "mean_token_accuracy": 0.16059614568948746,
      "num_tokens": 15487182.0,
      "step": 7450
    },
    {
      "entropy": 5.692904949188232,
      "epoch": 0.6762518142235123,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004961347380235031,
      "loss": 5.4569,
      "mean_token_accuracy": 0.16992294639348984,
      "num_tokens": 15497600.0,
      "step": 7455
    },
    {
      "entropy": 5.706664991378784,
      "epoch": 0.6767053701015965,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004961287640032879,
      "loss": 5.6197,
      "mean_token_accuracy": 0.15421608984470367,
      "num_tokens": 15508278.0,
      "step": 7460
    },
    {
      "entropy": 5.7377471923828125,
      "epoch": 0.6771589259796807,
      "grad_norm": 1.375,
      "learning_rate": 0.0004961227854100734,
      "loss": 5.6387,
      "mean_token_accuracy": 0.15761489421129227,
      "num_tokens": 15518458.0,
      "step": 7465
    },
    {
      "entropy": 5.603558444976807,
      "epoch": 0.6776124818577649,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004961168022439835,
      "loss": 5.4402,
      "mean_token_accuracy": 0.16726915091276168,
      "num_tokens": 15528738.0,
      "step": 7470
    },
    {
      "entropy": 5.601382780075073,
      "epoch": 0.6780660377358491,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004961108145051417,
      "loss": 5.4898,
      "mean_token_accuracy": 0.16797880381345748,
      "num_tokens": 15538656.0,
      "step": 7475
    },
    {
      "entropy": 5.526081609725952,
      "epoch": 0.6785195936139332,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000496104822193672,
      "loss": 5.5025,
      "mean_token_accuracy": 0.16627522855997084,
      "num_tokens": 15549285.0,
      "step": 7480
    },
    {
      "entropy": 5.69929609298706,
      "epoch": 0.6789731494920174,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004960988253096982,
      "loss": 5.5587,
      "mean_token_accuracy": 0.17137698233127593,
      "num_tokens": 15559766.0,
      "step": 7485
    },
    {
      "entropy": 5.658153438568116,
      "epoch": 0.6794267053701016,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004960928238533443,
      "loss": 5.4357,
      "mean_token_accuracy": 0.1714092329144478,
      "num_tokens": 15570873.0,
      "step": 7490
    },
    {
      "entropy": 5.671041440963745,
      "epoch": 0.6798802612481858,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004960868178247344,
      "loss": 5.6779,
      "mean_token_accuracy": 0.1592846006155014,
      "num_tokens": 15581229.0,
      "step": 7495
    },
    {
      "entropy": 5.626502847671508,
      "epoch": 0.68033381712627,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004960808072239929,
      "loss": 5.5501,
      "mean_token_accuracy": 0.15719763189554214,
      "num_tokens": 15591686.0,
      "step": 7500
    },
    {
      "entropy": 5.696853733062744,
      "epoch": 0.6807873730043541,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004960747920512439,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1583081491291523,
      "num_tokens": 15601869.0,
      "step": 7505
    },
    {
      "entropy": 5.6992940425872805,
      "epoch": 0.6812409288824384,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000496068772306612,
      "loss": 5.5539,
      "mean_token_accuracy": 0.1609680771827698,
      "num_tokens": 15612946.0,
      "step": 7510
    },
    {
      "entropy": 5.662940216064453,
      "epoch": 0.6816944847605225,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004960627479902214,
      "loss": 5.5927,
      "mean_token_accuracy": 0.15721766501665116,
      "num_tokens": 15623761.0,
      "step": 7515
    },
    {
      "entropy": 5.627185249328614,
      "epoch": 0.6821480406386067,
      "grad_norm": 1.578125,
      "learning_rate": 0.000496056719102197,
      "loss": 5.5756,
      "mean_token_accuracy": 0.15465673059225082,
      "num_tokens": 15634500.0,
      "step": 7520
    },
    {
      "entropy": 5.67201452255249,
      "epoch": 0.6826015965166908,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004960506856426631,
      "loss": 5.446,
      "mean_token_accuracy": 0.16561726480722427,
      "num_tokens": 15644896.0,
      "step": 7525
    },
    {
      "entropy": 5.644251871109009,
      "epoch": 0.683055152394775,
      "grad_norm": 2.125,
      "learning_rate": 0.0004960446476117448,
      "loss": 5.5551,
      "mean_token_accuracy": 0.16353006064891815,
      "num_tokens": 15655008.0,
      "step": 7530
    },
    {
      "entropy": 5.605746555328369,
      "epoch": 0.6835087082728593,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004960386050095668,
      "loss": 5.4758,
      "mean_token_accuracy": 0.16005973517894745,
      "num_tokens": 15665944.0,
      "step": 7535
    },
    {
      "entropy": 5.689500379562378,
      "epoch": 0.6839622641509434,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004960325578362543,
      "loss": 5.5606,
      "mean_token_accuracy": 0.1629498139023781,
      "num_tokens": 15677099.0,
      "step": 7540
    },
    {
      "entropy": 5.708564949035645,
      "epoch": 0.6844158200290276,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000496026506091932,
      "loss": 5.5212,
      "mean_token_accuracy": 0.16468921154737473,
      "num_tokens": 15687531.0,
      "step": 7545
    },
    {
      "entropy": 5.658125495910644,
      "epoch": 0.6848693759071117,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004960204497767253,
      "loss": 5.5427,
      "mean_token_accuracy": 0.16106893569231034,
      "num_tokens": 15697002.0,
      "step": 7550
    },
    {
      "entropy": 5.572294759750366,
      "epoch": 0.685322931785196,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004960143888907594,
      "loss": 5.4228,
      "mean_token_accuracy": 0.16344174593687058,
      "num_tokens": 15707249.0,
      "step": 7555
    },
    {
      "entropy": 5.641739702224731,
      "epoch": 0.6857764876632801,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004960083234341595,
      "loss": 5.4955,
      "mean_token_accuracy": 0.16488028764724733,
      "num_tokens": 15718764.0,
      "step": 7560
    },
    {
      "entropy": 5.649561786651612,
      "epoch": 0.6862300435413643,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004960022534070513,
      "loss": 5.4482,
      "mean_token_accuracy": 0.17171214669942855,
      "num_tokens": 15728854.0,
      "step": 7565
    },
    {
      "entropy": 5.563173294067383,
      "epoch": 0.6866835994194485,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00049599617880956,
      "loss": 5.4517,
      "mean_token_accuracy": 0.16561209410429,
      "num_tokens": 15739682.0,
      "step": 7570
    },
    {
      "entropy": 5.589160346984864,
      "epoch": 0.6871371552975326,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004959900996418115,
      "loss": 5.459,
      "mean_token_accuracy": 0.16965826749801635,
      "num_tokens": 15749628.0,
      "step": 7575
    },
    {
      "entropy": 5.614845037460327,
      "epoch": 0.6875907111756169,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004959840159039315,
      "loss": 5.5172,
      "mean_token_accuracy": 0.1624355211853981,
      "num_tokens": 15760137.0,
      "step": 7580
    },
    {
      "entropy": 5.6442975997924805,
      "epoch": 0.688044267053701,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004959779275960456,
      "loss": 5.4938,
      "mean_token_accuracy": 0.1618069216609001,
      "num_tokens": 15769889.0,
      "step": 7585
    },
    {
      "entropy": 5.643933296203613,
      "epoch": 0.6884978229317852,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00049597183471828,
      "loss": 5.5783,
      "mean_token_accuracy": 0.15737499594688414,
      "num_tokens": 15780912.0,
      "step": 7590
    },
    {
      "entropy": 5.657950305938721,
      "epoch": 0.6889513788098693,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004959657372707604,
      "loss": 5.4573,
      "mean_token_accuracy": 0.17339664250612258,
      "num_tokens": 15791719.0,
      "step": 7595
    },
    {
      "entropy": 5.65258297920227,
      "epoch": 0.6894049346879536,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004959596352536132,
      "loss": 5.5068,
      "mean_token_accuracy": 0.16826928555965423,
      "num_tokens": 15802551.0,
      "step": 7600
    },
    {
      "entropy": 5.662902116775513,
      "epoch": 0.6898584905660378,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004959535286669643,
      "loss": 5.572,
      "mean_token_accuracy": 0.15824030786752702,
      "num_tokens": 15813551.0,
      "step": 7605
    },
    {
      "entropy": 5.67209939956665,
      "epoch": 0.6903120464441219,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004959474175109402,
      "loss": 5.5381,
      "mean_token_accuracy": 0.16531185656785966,
      "num_tokens": 15824262.0,
      "step": 7610
    },
    {
      "entropy": 5.7075118064880375,
      "epoch": 0.6907656023222061,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004959413017856672,
      "loss": 5.5786,
      "mean_token_accuracy": 0.16659896671772004,
      "num_tokens": 15835493.0,
      "step": 7615
    },
    {
      "entropy": 5.561695861816406,
      "epoch": 0.6912191582002902,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004959351814912716,
      "loss": 5.4769,
      "mean_token_accuracy": 0.16452614217996597,
      "num_tokens": 15846224.0,
      "step": 7620
    },
    {
      "entropy": 5.615854978561401,
      "epoch": 0.6916727140783745,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004959290566278803,
      "loss": 5.5017,
      "mean_token_accuracy": 0.1664932370185852,
      "num_tokens": 15855622.0,
      "step": 7625
    },
    {
      "entropy": 5.7003296375274655,
      "epoch": 0.6921262699564587,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004959229271956199,
      "loss": 5.5494,
      "mean_token_accuracy": 0.16689909547567366,
      "num_tokens": 15865835.0,
      "step": 7630
    },
    {
      "entropy": 5.669817972183227,
      "epoch": 0.6925798258345428,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004959167931946169,
      "loss": 5.4935,
      "mean_token_accuracy": 0.16737326830625535,
      "num_tokens": 15875915.0,
      "step": 7635
    },
    {
      "entropy": 5.633016347885132,
      "epoch": 0.693033381712627,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004959106546249984,
      "loss": 5.4911,
      "mean_token_accuracy": 0.16724033057689666,
      "num_tokens": 15887111.0,
      "step": 7640
    },
    {
      "entropy": 5.636037111282349,
      "epoch": 0.6934869375907112,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004959045114868913,
      "loss": 5.6119,
      "mean_token_accuracy": 0.15588464289903642,
      "num_tokens": 15899528.0,
      "step": 7645
    },
    {
      "entropy": 5.619613742828369,
      "epoch": 0.6939404934687954,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004958983637804225,
      "loss": 5.5178,
      "mean_token_accuracy": 0.17082595229148864,
      "num_tokens": 15910881.0,
      "step": 7650
    },
    {
      "entropy": 5.7317766666412355,
      "epoch": 0.6943940493468795,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004958922115057194,
      "loss": 5.564,
      "mean_token_accuracy": 0.16362126916646957,
      "num_tokens": 15921189.0,
      "step": 7655
    },
    {
      "entropy": 5.708053493499756,
      "epoch": 0.6948476052249637,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004958860546629089,
      "loss": 5.5302,
      "mean_token_accuracy": 0.17005573958158493,
      "num_tokens": 15931276.0,
      "step": 7660
    },
    {
      "entropy": 5.700485134124756,
      "epoch": 0.6953011611030478,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004958798932521186,
      "loss": 5.6323,
      "mean_token_accuracy": 0.15413284301757812,
      "num_tokens": 15941506.0,
      "step": 7665
    },
    {
      "entropy": 5.591758680343628,
      "epoch": 0.6957547169811321,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004958737272734759,
      "loss": 5.4652,
      "mean_token_accuracy": 0.16640574038028716,
      "num_tokens": 15951158.0,
      "step": 7670
    },
    {
      "entropy": 5.682381105422974,
      "epoch": 0.6962082728592163,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004958675567271081,
      "loss": 5.4934,
      "mean_token_accuracy": 0.16660627275705336,
      "num_tokens": 15960852.0,
      "step": 7675
    },
    {
      "entropy": 5.649328660964966,
      "epoch": 0.6966618287373004,
      "grad_norm": 1.640625,
      "learning_rate": 0.000495861381613143,
      "loss": 5.5042,
      "mean_token_accuracy": 0.1625627711415291,
      "num_tokens": 15971291.0,
      "step": 7680
    },
    {
      "entropy": 5.6603899002075195,
      "epoch": 0.6971153846153846,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004958552019317082,
      "loss": 5.5533,
      "mean_token_accuracy": 0.16272665113210677,
      "num_tokens": 15982180.0,
      "step": 7685
    },
    {
      "entropy": 5.628291273117066,
      "epoch": 0.6975689404934688,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004958490176829316,
      "loss": 5.4536,
      "mean_token_accuracy": 0.16861288696527482,
      "num_tokens": 15992363.0,
      "step": 7690
    },
    {
      "entropy": 5.6698826313018795,
      "epoch": 0.698022496371553,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000495842828866941,
      "loss": 5.5444,
      "mean_token_accuracy": 0.16244934797286986,
      "num_tokens": 16003477.0,
      "step": 7695
    },
    {
      "entropy": 5.652411031723022,
      "epoch": 0.6984760522496372,
      "grad_norm": 3.421875,
      "learning_rate": 0.0004958366354838645,
      "loss": 5.5218,
      "mean_token_accuracy": 0.16245862245559692,
      "num_tokens": 16013747.0,
      "step": 7700
    },
    {
      "entropy": 5.642243480682373,
      "epoch": 0.6989296081277213,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00049583043753383,
      "loss": 5.5493,
      "mean_token_accuracy": 0.16651039719581603,
      "num_tokens": 16025662.0,
      "step": 7705
    },
    {
      "entropy": 5.605423450469971,
      "epoch": 0.6993831640058055,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004958242350169658,
      "loss": 5.5869,
      "mean_token_accuracy": 0.16867056488990784,
      "num_tokens": 16037315.0,
      "step": 7710
    },
    {
      "entropy": 5.584050416946411,
      "epoch": 0.6998367198838897,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004958180279334001,
      "loss": 5.4296,
      "mean_token_accuracy": 0.16454358100891114,
      "num_tokens": 16047297.0,
      "step": 7715
    },
    {
      "entropy": 5.66097617149353,
      "epoch": 0.7002902757619739,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004958118162832614,
      "loss": 5.4167,
      "mean_token_accuracy": 0.16869923025369643,
      "num_tokens": 16057573.0,
      "step": 7720
    },
    {
      "entropy": 5.605273962020874,
      "epoch": 0.700743831640058,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004958056000666781,
      "loss": 5.5764,
      "mean_token_accuracy": 0.15875106006860734,
      "num_tokens": 16067926.0,
      "step": 7725
    },
    {
      "entropy": 5.6579488754272464,
      "epoch": 0.7011973875181422,
      "grad_norm": 4.46875,
      "learning_rate": 0.0004957993792837786,
      "loss": 5.5062,
      "mean_token_accuracy": 0.16394148617982865,
      "num_tokens": 16077938.0,
      "step": 7730
    },
    {
      "entropy": 5.610279369354248,
      "epoch": 0.7016509433962265,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004957931539346917,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1661431223154068,
      "num_tokens": 16088163.0,
      "step": 7735
    },
    {
      "entropy": 5.670271682739258,
      "epoch": 0.7021044992743106,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004957869240195462,
      "loss": 5.6106,
      "mean_token_accuracy": 0.15643369257450104,
      "num_tokens": 16100640.0,
      "step": 7740
    },
    {
      "entropy": 5.63419885635376,
      "epoch": 0.7025580551523948,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004957806895384707,
      "loss": 5.4382,
      "mean_token_accuracy": 0.16354862451553345,
      "num_tokens": 16111018.0,
      "step": 7745
    },
    {
      "entropy": 5.5481712341308596,
      "epoch": 0.7030116110304789,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004957744504915944,
      "loss": 5.471,
      "mean_token_accuracy": 0.1648949444293976,
      "num_tokens": 16120873.0,
      "step": 7750
    },
    {
      "entropy": 5.671518754959107,
      "epoch": 0.7034651669085631,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004957682068790463,
      "loss": 5.5001,
      "mean_token_accuracy": 0.1649119570851326,
      "num_tokens": 16131178.0,
      "step": 7755
    },
    {
      "entropy": 5.646680402755737,
      "epoch": 0.7039187227866474,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004957619587009553,
      "loss": 5.4848,
      "mean_token_accuracy": 0.16286496371030806,
      "num_tokens": 16141156.0,
      "step": 7760
    },
    {
      "entropy": 5.633469486236573,
      "epoch": 0.7043722786647315,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004957557059574507,
      "loss": 5.5988,
      "mean_token_accuracy": 0.15549362897872926,
      "num_tokens": 16150077.0,
      "step": 7765
    },
    {
      "entropy": 5.645452451705933,
      "epoch": 0.7048258345428157,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004957494486486618,
      "loss": 5.5126,
      "mean_token_accuracy": 0.16329682469367982,
      "num_tokens": 16160514.0,
      "step": 7770
    },
    {
      "entropy": 5.671679973602295,
      "epoch": 0.7052793904208998,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004957431867747182,
      "loss": 5.558,
      "mean_token_accuracy": 0.16956402212381363,
      "num_tokens": 16170811.0,
      "step": 7775
    },
    {
      "entropy": 5.577692174911499,
      "epoch": 0.7057329462989841,
      "grad_norm": 1.5,
      "learning_rate": 0.0004957369203357492,
      "loss": 5.4453,
      "mean_token_accuracy": 0.16849376410245895,
      "num_tokens": 16181553.0,
      "step": 7780
    },
    {
      "entropy": 5.606373643875122,
      "epoch": 0.7061865021770682,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004957306493318844,
      "loss": 5.4884,
      "mean_token_accuracy": 0.17001904994249345,
      "num_tokens": 16191308.0,
      "step": 7785
    },
    {
      "entropy": 5.617648077011109,
      "epoch": 0.7066400580551524,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004957243737632535,
      "loss": 5.4393,
      "mean_token_accuracy": 0.17012592554092407,
      "num_tokens": 16201311.0,
      "step": 7790
    },
    {
      "entropy": 5.680548620223999,
      "epoch": 0.7070936139332366,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004957180936299863,
      "loss": 5.4894,
      "mean_token_accuracy": 0.1714714452624321,
      "num_tokens": 16211105.0,
      "step": 7795
    },
    {
      "entropy": 5.652728271484375,
      "epoch": 0.7075471698113207,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004957118089322127,
      "loss": 5.5114,
      "mean_token_accuracy": 0.16141459792852403,
      "num_tokens": 16220836.0,
      "step": 7800
    },
    {
      "entropy": 5.632328701019287,
      "epoch": 0.708000725689405,
      "grad_norm": 1.625,
      "learning_rate": 0.0004957055196700628,
      "loss": 5.4226,
      "mean_token_accuracy": 0.170921191573143,
      "num_tokens": 16231630.0,
      "step": 7805
    },
    {
      "entropy": 5.648638772964477,
      "epoch": 0.7084542815674891,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004956992258436663,
      "loss": 5.521,
      "mean_token_accuracy": 0.16436462551355363,
      "num_tokens": 16242272.0,
      "step": 7810
    },
    {
      "entropy": 5.563827037811279,
      "epoch": 0.7089078374455733,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004956929274531536,
      "loss": 5.4206,
      "mean_token_accuracy": 0.16884785741567612,
      "num_tokens": 16252555.0,
      "step": 7815
    },
    {
      "entropy": 5.5962522506713865,
      "epoch": 0.7093613933236574,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000495686624498655,
      "loss": 5.5488,
      "mean_token_accuracy": 0.16552574336528778,
      "num_tokens": 16261964.0,
      "step": 7820
    },
    {
      "entropy": 5.690983247756958,
      "epoch": 0.7098149492017417,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004956803169803007,
      "loss": 5.5803,
      "mean_token_accuracy": 0.1607639327645302,
      "num_tokens": 16274543.0,
      "step": 7825
    },
    {
      "entropy": 5.619023227691651,
      "epoch": 0.7102685050798259,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004956740048982212,
      "loss": 5.4861,
      "mean_token_accuracy": 0.16002560406923294,
      "num_tokens": 16285124.0,
      "step": 7830
    },
    {
      "entropy": 5.624927282333374,
      "epoch": 0.71072206095791,
      "grad_norm": 2.125,
      "learning_rate": 0.0004956676882525472,
      "loss": 5.5772,
      "mean_token_accuracy": 0.1634851261973381,
      "num_tokens": 16295727.0,
      "step": 7835
    },
    {
      "entropy": 5.696490097045898,
      "epoch": 0.7111756168359942,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000495661367043409,
      "loss": 5.4756,
      "mean_token_accuracy": 0.17260087430477142,
      "num_tokens": 16305496.0,
      "step": 7840
    },
    {
      "entropy": 5.607415437698364,
      "epoch": 0.7116291727140783,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004956550412709376,
      "loss": 5.4603,
      "mean_token_accuracy": 0.16924340277910233,
      "num_tokens": 16315112.0,
      "step": 7845
    },
    {
      "entropy": 5.623329830169678,
      "epoch": 0.7120827285921626,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004956487109352637,
      "loss": 5.4983,
      "mean_token_accuracy": 0.17477993816137313,
      "num_tokens": 16325613.0,
      "step": 7850
    },
    {
      "entropy": 5.665192794799805,
      "epoch": 0.7125362844702468,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004956423760365181,
      "loss": 5.4333,
      "mean_token_accuracy": 0.16873666346073152,
      "num_tokens": 16335317.0,
      "step": 7855
    },
    {
      "entropy": 5.609439182281494,
      "epoch": 0.7129898403483309,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004956360365748321,
      "loss": 5.5845,
      "mean_token_accuracy": 0.15723467022180557,
      "num_tokens": 16345003.0,
      "step": 7860
    },
    {
      "entropy": 5.658202457427978,
      "epoch": 0.7134433962264151,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004956296925503367,
      "loss": 5.5558,
      "mean_token_accuracy": 0.1622123345732689,
      "num_tokens": 16356108.0,
      "step": 7865
    },
    {
      "entropy": 5.6005700588226315,
      "epoch": 0.7138969521044993,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004956233439631628,
      "loss": 5.3675,
      "mean_token_accuracy": 0.16724979281425476,
      "num_tokens": 16364733.0,
      "step": 7870
    },
    {
      "entropy": 5.619435358047485,
      "epoch": 0.7143505079825835,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004956169908134421,
      "loss": 5.4829,
      "mean_token_accuracy": 0.16559068113565445,
      "num_tokens": 16375229.0,
      "step": 7875
    },
    {
      "entropy": 5.59596996307373,
      "epoch": 0.7148040638606676,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004956106331013057,
      "loss": 5.5706,
      "mean_token_accuracy": 0.1656107187271118,
      "num_tokens": 16386366.0,
      "step": 7880
    },
    {
      "entropy": 5.6192371368408205,
      "epoch": 0.7152576197387518,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004956042708268853,
      "loss": 5.4437,
      "mean_token_accuracy": 0.16840773224830627,
      "num_tokens": 16397067.0,
      "step": 7885
    },
    {
      "entropy": 5.605705690383911,
      "epoch": 0.715711175616836,
      "grad_norm": 1.75,
      "learning_rate": 0.0004955979039903123,
      "loss": 5.4123,
      "mean_token_accuracy": 0.1653614744544029,
      "num_tokens": 16407256.0,
      "step": 7890
    },
    {
      "entropy": 5.532815551757812,
      "epoch": 0.7161647314949202,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004955915325917183,
      "loss": 5.4582,
      "mean_token_accuracy": 0.16202937364578246,
      "num_tokens": 16418701.0,
      "step": 7895
    },
    {
      "entropy": 5.560326719284058,
      "epoch": 0.7166182873730044,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004955851566312354,
      "loss": 5.4014,
      "mean_token_accuracy": 0.1720588907599449,
      "num_tokens": 16428019.0,
      "step": 7900
    },
    {
      "entropy": 5.603496789932251,
      "epoch": 0.7170718432510885,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004955787761089952,
      "loss": 5.4644,
      "mean_token_accuracy": 0.1611000120639801,
      "num_tokens": 16440122.0,
      "step": 7905
    },
    {
      "entropy": 5.698283433914185,
      "epoch": 0.7175253991291727,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004955723910251296,
      "loss": 5.5366,
      "mean_token_accuracy": 0.1604372724890709,
      "num_tokens": 16449801.0,
      "step": 7910
    },
    {
      "entropy": 5.602445983886719,
      "epoch": 0.717978955007257,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004955660013797709,
      "loss": 5.4201,
      "mean_token_accuracy": 0.16884026378393174,
      "num_tokens": 16460155.0,
      "step": 7915
    },
    {
      "entropy": 5.65273084640503,
      "epoch": 0.7184325108853411,
      "grad_norm": 1.578125,
      "learning_rate": 0.000495559607173051,
      "loss": 5.6034,
      "mean_token_accuracy": 0.15278467386960984,
      "num_tokens": 16470656.0,
      "step": 7920
    },
    {
      "entropy": 5.678693532943726,
      "epoch": 0.7188860667634253,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004955532084051022,
      "loss": 5.5052,
      "mean_token_accuracy": 0.16545139700174333,
      "num_tokens": 16481449.0,
      "step": 7925
    },
    {
      "entropy": 5.689681911468506,
      "epoch": 0.7193396226415094,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004955468050760568,
      "loss": 5.5232,
      "mean_token_accuracy": 0.1603458657860756,
      "num_tokens": 16492032.0,
      "step": 7930
    },
    {
      "entropy": 5.605563116073609,
      "epoch": 0.7197931785195936,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004955403971860474,
      "loss": 5.4559,
      "mean_token_accuracy": 0.16635042279958726,
      "num_tokens": 16502221.0,
      "step": 7935
    },
    {
      "entropy": 5.575072383880615,
      "epoch": 0.7202467343976778,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004955339847352063,
      "loss": 5.4508,
      "mean_token_accuracy": 0.1656094402074814,
      "num_tokens": 16512021.0,
      "step": 7940
    },
    {
      "entropy": 5.602277755737305,
      "epoch": 0.720700290275762,
      "grad_norm": 7.3125,
      "learning_rate": 0.0004955275677236662,
      "loss": 5.4922,
      "mean_token_accuracy": 0.16556978672742845,
      "num_tokens": 16522060.0,
      "step": 7945
    },
    {
      "entropy": 5.636832046508789,
      "epoch": 0.7211538461538461,
      "grad_norm": 4.25,
      "learning_rate": 0.0004955211461515599,
      "loss": 5.4857,
      "mean_token_accuracy": 0.16849270910024644,
      "num_tokens": 16532827.0,
      "step": 7950
    },
    {
      "entropy": 5.637872695922852,
      "epoch": 0.7216074020319303,
      "grad_norm": 1.734375,
      "learning_rate": 0.00049551472001902,
      "loss": 5.5172,
      "mean_token_accuracy": 0.1674077406525612,
      "num_tokens": 16542838.0,
      "step": 7955
    },
    {
      "entropy": 5.592290878295898,
      "epoch": 0.7220609579100146,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004955082893261796,
      "loss": 5.4789,
      "mean_token_accuracy": 0.17216830253601073,
      "num_tokens": 16552615.0,
      "step": 7960
    },
    {
      "entropy": 5.716319561004639,
      "epoch": 0.7225145137880987,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004955018540731714,
      "loss": 5.5494,
      "mean_token_accuracy": 0.15874373018741608,
      "num_tokens": 16561777.0,
      "step": 7965
    },
    {
      "entropy": 5.671530818939209,
      "epoch": 0.7229680696661829,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004954954142601288,
      "loss": 5.5132,
      "mean_token_accuracy": 0.16330315917730331,
      "num_tokens": 16572111.0,
      "step": 7970
    },
    {
      "entropy": 5.677050161361694,
      "epoch": 0.723421625544267,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004954889698871849,
      "loss": 5.5445,
      "mean_token_accuracy": 0.16566748470067977,
      "num_tokens": 16581674.0,
      "step": 7975
    },
    {
      "entropy": 5.697398138046265,
      "epoch": 0.7238751814223512,
      "grad_norm": 1.5,
      "learning_rate": 0.0004954825209544728,
      "loss": 5.5406,
      "mean_token_accuracy": 0.17010942101478577,
      "num_tokens": 16591802.0,
      "step": 7980
    },
    {
      "entropy": 5.579377937316894,
      "epoch": 0.7243287373004355,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000495476067462126,
      "loss": 5.4295,
      "mean_token_accuracy": 0.17563054859638214,
      "num_tokens": 16602772.0,
      "step": 7985
    },
    {
      "entropy": 5.614494657516479,
      "epoch": 0.7247822931785196,
      "grad_norm": 1.359375,
      "learning_rate": 0.000495469609410278,
      "loss": 5.4919,
      "mean_token_accuracy": 0.16485417038202285,
      "num_tokens": 16613526.0,
      "step": 7990
    },
    {
      "entropy": 5.538692665100098,
      "epoch": 0.7252358490566038,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004954631467990622,
      "loss": 5.5094,
      "mean_token_accuracy": 0.16715267151594163,
      "num_tokens": 16623583.0,
      "step": 7995
    },
    {
      "entropy": 5.623130130767822,
      "epoch": 0.7256894049346879,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004954566796286124,
      "loss": 5.5403,
      "mean_token_accuracy": 0.1612747296690941,
      "num_tokens": 16635068.0,
      "step": 8000
    },
    {
      "entropy": 5.659089231491089,
      "epoch": 0.7261429608127722,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004954502078990623,
      "loss": 5.4492,
      "mean_token_accuracy": 0.16616036742925644,
      "num_tokens": 16645021.0,
      "step": 8005
    },
    {
      "entropy": 5.618288230895996,
      "epoch": 0.7265965166908563,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004954437316105457,
      "loss": 5.5206,
      "mean_token_accuracy": 0.16156897693872452,
      "num_tokens": 16656021.0,
      "step": 8010
    },
    {
      "entropy": 5.596080017089844,
      "epoch": 0.7270500725689405,
      "grad_norm": 1.625,
      "learning_rate": 0.0004954372507631965,
      "loss": 5.4269,
      "mean_token_accuracy": 0.17329126596450806,
      "num_tokens": 16665236.0,
      "step": 8015
    },
    {
      "entropy": 5.621707201004028,
      "epoch": 0.7275036284470247,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004954307653571489,
      "loss": 5.6036,
      "mean_token_accuracy": 0.15767206251621246,
      "num_tokens": 16676690.0,
      "step": 8020
    },
    {
      "entropy": 5.598207330703735,
      "epoch": 0.7279571843251088,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004954242753925367,
      "loss": 5.4999,
      "mean_token_accuracy": 0.16131179332733153,
      "num_tokens": 16687288.0,
      "step": 8025
    },
    {
      "entropy": 5.669662570953369,
      "epoch": 0.7284107402031931,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004954177808694946,
      "loss": 5.5124,
      "mean_token_accuracy": 0.15983750075101852,
      "num_tokens": 16697873.0,
      "step": 8030
    },
    {
      "entropy": 5.608494615554809,
      "epoch": 0.7288642960812772,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004954112817881565,
      "loss": 5.5089,
      "mean_token_accuracy": 0.16026637256145476,
      "num_tokens": 16709002.0,
      "step": 8035
    },
    {
      "entropy": 5.5762261867523195,
      "epoch": 0.7293178519593614,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004954047781486568,
      "loss": 5.4862,
      "mean_token_accuracy": 0.16284284740686417,
      "num_tokens": 16718989.0,
      "step": 8040
    },
    {
      "entropy": 5.748119974136353,
      "epoch": 0.7297714078374455,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004953982699511303,
      "loss": 5.5813,
      "mean_token_accuracy": 0.16285916417837143,
      "num_tokens": 16730425.0,
      "step": 8045
    },
    {
      "entropy": 5.667424249649048,
      "epoch": 0.7302249637155298,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004953917571957114,
      "loss": 5.4673,
      "mean_token_accuracy": 0.16531988382339477,
      "num_tokens": 16740216.0,
      "step": 8050
    },
    {
      "entropy": 5.566254091262818,
      "epoch": 0.730678519593614,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004953852398825346,
      "loss": 5.4612,
      "mean_token_accuracy": 0.16659636050462723,
      "num_tokens": 16752436.0,
      "step": 8055
    },
    {
      "entropy": 5.577506065368652,
      "epoch": 0.7311320754716981,
      "grad_norm": 2.15625,
      "learning_rate": 0.000495378718011735,
      "loss": 5.478,
      "mean_token_accuracy": 0.1658300280570984,
      "num_tokens": 16762631.0,
      "step": 8060
    },
    {
      "entropy": 5.710255908966064,
      "epoch": 0.7315856313497823,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004953721915834473,
      "loss": 5.5083,
      "mean_token_accuracy": 0.16436021476984025,
      "num_tokens": 16772613.0,
      "step": 8065
    },
    {
      "entropy": 5.665786218643189,
      "epoch": 0.7320391872278664,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004953656605978067,
      "loss": 5.5044,
      "mean_token_accuracy": 0.1608337089419365,
      "num_tokens": 16782963.0,
      "step": 8070
    },
    {
      "entropy": 5.600820016860962,
      "epoch": 0.7324927431059507,
      "grad_norm": 1.359375,
      "learning_rate": 0.000495359125054948,
      "loss": 5.449,
      "mean_token_accuracy": 0.16788298934698104,
      "num_tokens": 16793892.0,
      "step": 8075
    },
    {
      "entropy": 5.526678943634034,
      "epoch": 0.7329462989840348,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004953525849550063,
      "loss": 5.4313,
      "mean_token_accuracy": 0.16043167561292648,
      "num_tokens": 16803388.0,
      "step": 8080
    },
    {
      "entropy": 5.590819931030273,
      "epoch": 0.733399854862119,
      "grad_norm": 1.75,
      "learning_rate": 0.0004953460402981172,
      "loss": 5.3868,
      "mean_token_accuracy": 0.17100578546524048,
      "num_tokens": 16813201.0,
      "step": 8085
    },
    {
      "entropy": 5.688029956817627,
      "epoch": 0.7338534107402032,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004953394910844157,
      "loss": 5.6513,
      "mean_token_accuracy": 0.1624013900756836,
      "num_tokens": 16823265.0,
      "step": 8090
    },
    {
      "entropy": 5.64172945022583,
      "epoch": 0.7343069666182874,
      "grad_norm": 1.625,
      "learning_rate": 0.0004953329373140374,
      "loss": 5.4315,
      "mean_token_accuracy": 0.16412428468465806,
      "num_tokens": 16833057.0,
      "step": 8095
    },
    {
      "entropy": 5.550912380218506,
      "epoch": 0.7347605224963716,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004953263789871179,
      "loss": 5.3976,
      "mean_token_accuracy": 0.16362497657537461,
      "num_tokens": 16843488.0,
      "step": 8100
    },
    {
      "entropy": 5.552389669418335,
      "epoch": 0.7352140783744557,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004953198161037925,
      "loss": 5.4474,
      "mean_token_accuracy": 0.17534587383270264,
      "num_tokens": 16852448.0,
      "step": 8105
    },
    {
      "entropy": 5.507678890228272,
      "epoch": 0.7356676342525399,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004953132486641975,
      "loss": 5.4647,
      "mean_token_accuracy": 0.16817467659711838,
      "num_tokens": 16863944.0,
      "step": 8110
    },
    {
      "entropy": 5.596392822265625,
      "epoch": 0.736121190130624,
      "grad_norm": 1.421875,
      "learning_rate": 0.000495306676668468,
      "loss": 5.3928,
      "mean_token_accuracy": 0.1675754576921463,
      "num_tokens": 16873923.0,
      "step": 8115
    },
    {
      "entropy": 5.61533260345459,
      "epoch": 0.7365747460087083,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004953001001167406,
      "loss": 5.4943,
      "mean_token_accuracy": 0.16641771644353867,
      "num_tokens": 16884373.0,
      "step": 8120
    },
    {
      "entropy": 5.675483465194702,
      "epoch": 0.7370283018867925,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004952935190091508,
      "loss": 5.6098,
      "mean_token_accuracy": 0.16353694051504136,
      "num_tokens": 16895869.0,
      "step": 8125
    },
    {
      "entropy": 5.536702632904053,
      "epoch": 0.7374818577648766,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004952869333458349,
      "loss": 5.4735,
      "mean_token_accuracy": 0.16639091521501542,
      "num_tokens": 16906737.0,
      "step": 8130
    },
    {
      "entropy": 5.584896087646484,
      "epoch": 0.7379354136429608,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004952803431269292,
      "loss": 5.4054,
      "mean_token_accuracy": 0.17622085362672807,
      "num_tokens": 16916083.0,
      "step": 8135
    },
    {
      "entropy": 5.665721368789673,
      "epoch": 0.738388969521045,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004952737483525699,
      "loss": 5.5641,
      "mean_token_accuracy": 0.16562667936086656,
      "num_tokens": 16926419.0,
      "step": 8140
    },
    {
      "entropy": 5.680433750152588,
      "epoch": 0.7388425253991292,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004952671490228932,
      "loss": 5.5005,
      "mean_token_accuracy": 0.16464401185512542,
      "num_tokens": 16936067.0,
      "step": 8145
    },
    {
      "entropy": 5.651733016967773,
      "epoch": 0.7392960812772134,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004952605451380357,
      "loss": 5.4472,
      "mean_token_accuracy": 0.1683550000190735,
      "num_tokens": 16946571.0,
      "step": 8150
    },
    {
      "entropy": 5.558815574645996,
      "epoch": 0.7397496371552975,
      "grad_norm": 1.5,
      "learning_rate": 0.0004952539366981342,
      "loss": 5.5204,
      "mean_token_accuracy": 0.16246189475059508,
      "num_tokens": 16956764.0,
      "step": 8155
    },
    {
      "entropy": 5.61862416267395,
      "epoch": 0.7402031930333817,
      "grad_norm": 2.0,
      "learning_rate": 0.000495247323703325,
      "loss": 5.4277,
      "mean_token_accuracy": 0.17167850583791733,
      "num_tokens": 16966002.0,
      "step": 8160
    },
    {
      "entropy": 5.65856704711914,
      "epoch": 0.7406567489114659,
      "grad_norm": 1.4375,
      "learning_rate": 0.000495240706153745,
      "loss": 5.5475,
      "mean_token_accuracy": 0.165491583943367,
      "num_tokens": 16975705.0,
      "step": 8165
    },
    {
      "entropy": 5.617011165618896,
      "epoch": 0.7411103047895501,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004952340840495312,
      "loss": 5.4249,
      "mean_token_accuracy": 0.16945848613977432,
      "num_tokens": 16985945.0,
      "step": 8170
    },
    {
      "entropy": 5.5943726062774655,
      "epoch": 0.7415638606676342,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004952274573908203,
      "loss": 5.4007,
      "mean_token_accuracy": 0.16729725301265716,
      "num_tokens": 16995475.0,
      "step": 8175
    },
    {
      "entropy": 5.578772687911988,
      "epoch": 0.7420174165457184,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004952208261777496,
      "loss": 5.4676,
      "mean_token_accuracy": 0.1648031771183014,
      "num_tokens": 17005028.0,
      "step": 8180
    },
    {
      "entropy": 5.645346689224243,
      "epoch": 0.7424709724238027,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000495214190410456,
      "loss": 5.5031,
      "mean_token_accuracy": 0.16819154769182204,
      "num_tokens": 17014786.0,
      "step": 8185
    },
    {
      "entropy": 5.651758861541748,
      "epoch": 0.7429245283018868,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004952075500890768,
      "loss": 5.5106,
      "mean_token_accuracy": 0.16572055369615554,
      "num_tokens": 17024431.0,
      "step": 8190
    },
    {
      "entropy": 5.568734359741211,
      "epoch": 0.743378084179971,
      "grad_norm": 1.75,
      "learning_rate": 0.0004952009052137493,
      "loss": 5.4671,
      "mean_token_accuracy": 0.1714123785495758,
      "num_tokens": 17034368.0,
      "step": 8195
    },
    {
      "entropy": 5.56487398147583,
      "epoch": 0.7438316400580551,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004951942557846112,
      "loss": 5.523,
      "mean_token_accuracy": 0.16194852590560913,
      "num_tokens": 17043818.0,
      "step": 8200
    },
    {
      "entropy": 5.666324090957642,
      "epoch": 0.7442851959361393,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004951876018017997,
      "loss": 5.5143,
      "mean_token_accuracy": 0.16839117258787156,
      "num_tokens": 17054843.0,
      "step": 8205
    },
    {
      "entropy": 5.672155237197876,
      "epoch": 0.7447387518142236,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004951809432654524,
      "loss": 5.5133,
      "mean_token_accuracy": 0.1668528288602829,
      "num_tokens": 17065539.0,
      "step": 8210
    },
    {
      "entropy": 5.580246162414551,
      "epoch": 0.7451923076923077,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000495174280175707,
      "loss": 5.4133,
      "mean_token_accuracy": 0.16898000985383987,
      "num_tokens": 17075443.0,
      "step": 8215
    },
    {
      "entropy": 5.681068229675293,
      "epoch": 0.7456458635703919,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004951676125327015,
      "loss": 5.5488,
      "mean_token_accuracy": 0.16850530058145524,
      "num_tokens": 17085490.0,
      "step": 8220
    },
    {
      "entropy": 5.652074337005615,
      "epoch": 0.746099419448476,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004951609403365737,
      "loss": 5.5403,
      "mean_token_accuracy": 0.17004851400852203,
      "num_tokens": 17094803.0,
      "step": 8225
    },
    {
      "entropy": 5.583604717254639,
      "epoch": 0.7465529753265603,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004951542635874615,
      "loss": 5.3929,
      "mean_token_accuracy": 0.16981043070554733,
      "num_tokens": 17104373.0,
      "step": 8230
    },
    {
      "entropy": 5.644840860366822,
      "epoch": 0.7470065312046444,
      "grad_norm": 1.625,
      "learning_rate": 0.0004951475822855029,
      "loss": 5.496,
      "mean_token_accuracy": 0.16572268605232238,
      "num_tokens": 17115469.0,
      "step": 8235
    },
    {
      "entropy": 5.609163236618042,
      "epoch": 0.7474600870827286,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004951408964308364,
      "loss": 5.4631,
      "mean_token_accuracy": 0.16574133783578873,
      "num_tokens": 17125443.0,
      "step": 8240
    },
    {
      "entropy": 5.555290651321411,
      "epoch": 0.7479136429608128,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004951342060236,
      "loss": 5.4202,
      "mean_token_accuracy": 0.1661999449133873,
      "num_tokens": 17136253.0,
      "step": 8245
    },
    {
      "entropy": 5.700448513031006,
      "epoch": 0.7483671988388969,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004951275110639321,
      "loss": 5.486,
      "mean_token_accuracy": 0.16710744500160218,
      "num_tokens": 17145544.0,
      "step": 8250
    },
    {
      "entropy": 5.659205961227417,
      "epoch": 0.7488207547169812,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004951208115519712,
      "loss": 5.4875,
      "mean_token_accuracy": 0.1646759882569313,
      "num_tokens": 17155043.0,
      "step": 8255
    },
    {
      "entropy": 5.596213150024414,
      "epoch": 0.7492743105950653,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004951141074878558,
      "loss": 5.5352,
      "mean_token_accuracy": 0.1620617538690567,
      "num_tokens": 17164885.0,
      "step": 8260
    },
    {
      "entropy": 5.673506355285644,
      "epoch": 0.7497278664731495,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004951073988717246,
      "loss": 5.524,
      "mean_token_accuracy": 0.16366185396909713,
      "num_tokens": 17175681.0,
      "step": 8265
    },
    {
      "entropy": 5.618485164642334,
      "epoch": 0.7501814223512336,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004951006857037163,
      "loss": 5.4009,
      "mean_token_accuracy": 0.16720621883869172,
      "num_tokens": 17186002.0,
      "step": 8270
    },
    {
      "entropy": 5.59273247718811,
      "epoch": 0.7506349782293179,
      "grad_norm": 1.625,
      "learning_rate": 0.0004950939679839696,
      "loss": 5.4413,
      "mean_token_accuracy": 0.17087941765785217,
      "num_tokens": 17195499.0,
      "step": 8275
    },
    {
      "entropy": 5.597292947769165,
      "epoch": 0.7510885341074021,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004950872457126237,
      "loss": 5.4594,
      "mean_token_accuracy": 0.16754258424043655,
      "num_tokens": 17204684.0,
      "step": 8280
    },
    {
      "entropy": 5.606363010406494,
      "epoch": 0.7515420899854862,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004950805188898175,
      "loss": 5.4918,
      "mean_token_accuracy": 0.16287693679332732,
      "num_tokens": 17213987.0,
      "step": 8285
    },
    {
      "entropy": 5.635597372055054,
      "epoch": 0.7519956458635704,
      "grad_norm": 1.40625,
      "learning_rate": 0.00049507378751569,
      "loss": 5.5086,
      "mean_token_accuracy": 0.1621052622795105,
      "num_tokens": 17224512.0,
      "step": 8290
    },
    {
      "entropy": 5.718463325500489,
      "epoch": 0.7524492017416545,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004950670515903805,
      "loss": 5.5242,
      "mean_token_accuracy": 0.16537761837244033,
      "num_tokens": 17234362.0,
      "step": 8295
    },
    {
      "entropy": 5.61590986251831,
      "epoch": 0.7529027576197388,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004950603111140283,
      "loss": 5.517,
      "mean_token_accuracy": 0.16930110305547713,
      "num_tokens": 17244241.0,
      "step": 8300
    },
    {
      "entropy": 5.533387899398804,
      "epoch": 0.753356313497823,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004950535660867728,
      "loss": 5.4593,
      "mean_token_accuracy": 0.17571018189191817,
      "num_tokens": 17254689.0,
      "step": 8305
    },
    {
      "entropy": 5.653801774978637,
      "epoch": 0.7538098693759071,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004950468165087535,
      "loss": 5.5217,
      "mean_token_accuracy": 0.1659944862127304,
      "num_tokens": 17265506.0,
      "step": 8310
    },
    {
      "entropy": 5.675744581222534,
      "epoch": 0.7542634252539913,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004950400623801098,
      "loss": 5.5988,
      "mean_token_accuracy": 0.1633657842874527,
      "num_tokens": 17276376.0,
      "step": 8315
    },
    {
      "entropy": 5.673819494247437,
      "epoch": 0.7547169811320755,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004950333037009818,
      "loss": 5.5015,
      "mean_token_accuracy": 0.1547602504491806,
      "num_tokens": 17287005.0,
      "step": 8320
    },
    {
      "entropy": 5.646138334274292,
      "epoch": 0.7551705370101597,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004950265404715089,
      "loss": 5.4998,
      "mean_token_accuracy": 0.16027978658676148,
      "num_tokens": 17297595.0,
      "step": 8325
    },
    {
      "entropy": 5.5359149932861325,
      "epoch": 0.7556240928882438,
      "grad_norm": 1.78125,
      "learning_rate": 0.000495019772691831,
      "loss": 5.3759,
      "mean_token_accuracy": 0.16829735040664673,
      "num_tokens": 17307319.0,
      "step": 8330
    },
    {
      "entropy": 5.618014287948609,
      "epoch": 0.756077648766328,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004950130003620882,
      "loss": 5.5992,
      "mean_token_accuracy": 0.16626063138246536,
      "num_tokens": 17317978.0,
      "step": 8335
    },
    {
      "entropy": 5.620723533630371,
      "epoch": 0.7565312046444121,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004950062234824204,
      "loss": 5.4916,
      "mean_token_accuracy": 0.16590883433818818,
      "num_tokens": 17328339.0,
      "step": 8340
    },
    {
      "entropy": 5.63666729927063,
      "epoch": 0.7569847605224964,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004949994420529679,
      "loss": 5.5351,
      "mean_token_accuracy": 0.16365355849266053,
      "num_tokens": 17338668.0,
      "step": 8345
    },
    {
      "entropy": 5.61228494644165,
      "epoch": 0.7574383164005806,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004949926560738709,
      "loss": 5.4883,
      "mean_token_accuracy": 0.16132984161376954,
      "num_tokens": 17348609.0,
      "step": 8350
    },
    {
      "entropy": 5.669183969497681,
      "epoch": 0.7578918722786647,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004949858655452697,
      "loss": 5.4883,
      "mean_token_accuracy": 0.16427465975284578,
      "num_tokens": 17358014.0,
      "step": 8355
    },
    {
      "entropy": 5.620366477966309,
      "epoch": 0.7583454281567489,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004949790704673048,
      "loss": 5.4967,
      "mean_token_accuracy": 0.16788626164197923,
      "num_tokens": 17367975.0,
      "step": 8360
    },
    {
      "entropy": 5.630387926101685,
      "epoch": 0.7587989840348331,
      "grad_norm": 1.75,
      "learning_rate": 0.0004949722708401166,
      "loss": 5.4505,
      "mean_token_accuracy": 0.16372619867324828,
      "num_tokens": 17377960.0,
      "step": 8365
    },
    {
      "entropy": 5.629988765716552,
      "epoch": 0.7592525399129173,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004949654666638457,
      "loss": 5.5134,
      "mean_token_accuracy": 0.16987081617116928,
      "num_tokens": 17387334.0,
      "step": 8370
    },
    {
      "entropy": 5.618249797821045,
      "epoch": 0.7597060957910015,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000494958657938633,
      "loss": 5.4907,
      "mean_token_accuracy": 0.16165588349103927,
      "num_tokens": 17396521.0,
      "step": 8375
    },
    {
      "entropy": 5.584721469879151,
      "epoch": 0.7601596516690856,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004949518446646192,
      "loss": 5.4699,
      "mean_token_accuracy": 0.16930657476186753,
      "num_tokens": 17407217.0,
      "step": 8380
    },
    {
      "entropy": 5.659988975524902,
      "epoch": 0.7606132075471698,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004949450268419451,
      "loss": 5.5628,
      "mean_token_accuracy": 0.1590219631791115,
      "num_tokens": 17416982.0,
      "step": 8385
    },
    {
      "entropy": 5.568830490112305,
      "epoch": 0.761066763425254,
      "grad_norm": 1.421875,
      "learning_rate": 0.000494938204470752,
      "loss": 5.418,
      "mean_token_accuracy": 0.16622348129749298,
      "num_tokens": 17427636.0,
      "step": 8390
    },
    {
      "entropy": 5.612373018264771,
      "epoch": 0.7615203193033382,
      "grad_norm": 1.625,
      "learning_rate": 0.0004949313775511806,
      "loss": 5.5272,
      "mean_token_accuracy": 0.1630595237016678,
      "num_tokens": 17437488.0,
      "step": 8395
    },
    {
      "entropy": 5.589271116256714,
      "epoch": 0.7619738751814223,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004949245460833723,
      "loss": 5.3345,
      "mean_token_accuracy": 0.17178670167922974,
      "num_tokens": 17447838.0,
      "step": 8400
    },
    {
      "entropy": 5.6152825355529785,
      "epoch": 0.7624274310595065,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004949177100674684,
      "loss": 5.5397,
      "mean_token_accuracy": 0.16557554602622987,
      "num_tokens": 17457257.0,
      "step": 8405
    },
    {
      "entropy": 5.63554835319519,
      "epoch": 0.7628809869375908,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004949108695036102,
      "loss": 5.4666,
      "mean_token_accuracy": 0.1702419102191925,
      "num_tokens": 17467729.0,
      "step": 8410
    },
    {
      "entropy": 5.643184089660645,
      "epoch": 0.7633345428156749,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004949040243919392,
      "loss": 5.4834,
      "mean_token_accuracy": 0.15880836993455888,
      "num_tokens": 17477692.0,
      "step": 8415
    },
    {
      "entropy": 5.608482503890992,
      "epoch": 0.7637880986937591,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004948971747325968,
      "loss": 5.4943,
      "mean_token_accuracy": 0.16530852615833283,
      "num_tokens": 17488730.0,
      "step": 8420
    },
    {
      "entropy": 5.662958192825317,
      "epoch": 0.7642416545718432,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000494890320525725,
      "loss": 5.565,
      "mean_token_accuracy": 0.16616504043340682,
      "num_tokens": 17498374.0,
      "step": 8425
    },
    {
      "entropy": 5.644989442825318,
      "epoch": 0.7646952104499274,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004948834617714652,
      "loss": 5.4229,
      "mean_token_accuracy": 0.17348451167345047,
      "num_tokens": 17508601.0,
      "step": 8430
    },
    {
      "entropy": 5.626499938964844,
      "epoch": 0.7651487663280117,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004948765984699595,
      "loss": 5.4435,
      "mean_token_accuracy": 0.1681686297059059,
      "num_tokens": 17518164.0,
      "step": 8435
    },
    {
      "entropy": 5.593100833892822,
      "epoch": 0.7656023222060958,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004948697306213494,
      "loss": 5.4401,
      "mean_token_accuracy": 0.17205666303634642,
      "num_tokens": 17527244.0,
      "step": 8440
    },
    {
      "entropy": 5.661113691329956,
      "epoch": 0.76605587808418,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004948628582257775,
      "loss": 5.5026,
      "mean_token_accuracy": 0.16697869300842286,
      "num_tokens": 17538045.0,
      "step": 8445
    },
    {
      "entropy": 5.594349479675293,
      "epoch": 0.7665094339622641,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004948559812833857,
      "loss": 5.4576,
      "mean_token_accuracy": 0.16509982347488403,
      "num_tokens": 17547584.0,
      "step": 8450
    },
    {
      "entropy": 5.607390213012695,
      "epoch": 0.7669629898403484,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004948490997943162,
      "loss": 5.4958,
      "mean_token_accuracy": 0.16601382941007614,
      "num_tokens": 17558925.0,
      "step": 8455
    },
    {
      "entropy": 5.617490577697754,
      "epoch": 0.7674165457184325,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004948422137587111,
      "loss": 5.5893,
      "mean_token_accuracy": 0.15748300403356552,
      "num_tokens": 17569355.0,
      "step": 8460
    },
    {
      "entropy": 5.650879049301148,
      "epoch": 0.7678701015965167,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004948353231767131,
      "loss": 5.4606,
      "mean_token_accuracy": 0.16597110778093338,
      "num_tokens": 17579782.0,
      "step": 8465
    },
    {
      "entropy": 5.610150527954102,
      "epoch": 0.7683236574746009,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004948284280484645,
      "loss": 5.499,
      "mean_token_accuracy": 0.1709799662232399,
      "num_tokens": 17589772.0,
      "step": 8470
    },
    {
      "entropy": 5.549994325637817,
      "epoch": 0.768777213352685,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004948215283741081,
      "loss": 5.4234,
      "mean_token_accuracy": 0.16429323256015776,
      "num_tokens": 17599613.0,
      "step": 8475
    },
    {
      "entropy": 5.6233827590942385,
      "epoch": 0.7692307692307693,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004948146241537866,
      "loss": 5.4971,
      "mean_token_accuracy": 0.16073604077100753,
      "num_tokens": 17609056.0,
      "step": 8480
    },
    {
      "entropy": 5.569718742370606,
      "epoch": 0.7696843251088534,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004948077153876425,
      "loss": 5.4345,
      "mean_token_accuracy": 0.16776058673858643,
      "num_tokens": 17618620.0,
      "step": 8485
    },
    {
      "entropy": 5.547171401977539,
      "epoch": 0.7701378809869376,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004948008020758188,
      "loss": 5.3973,
      "mean_token_accuracy": 0.1737499088048935,
      "num_tokens": 17627630.0,
      "step": 8490
    },
    {
      "entropy": 5.569094276428222,
      "epoch": 0.7705914368650217,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004947938842184585,
      "loss": 5.4028,
      "mean_token_accuracy": 0.16861821115016937,
      "num_tokens": 17638903.0,
      "step": 8495
    },
    {
      "entropy": 5.597677564620971,
      "epoch": 0.771044992743106,
      "grad_norm": 1.5,
      "learning_rate": 0.0004947869618157048,
      "loss": 5.4617,
      "mean_token_accuracy": 0.1655205011367798,
      "num_tokens": 17649257.0,
      "step": 8500
    },
    {
      "entropy": 5.566603136062622,
      "epoch": 0.7714985486211902,
      "grad_norm": 1.625,
      "learning_rate": 0.0004947800348677006,
      "loss": 5.4307,
      "mean_token_accuracy": 0.1743499293923378,
      "num_tokens": 17659495.0,
      "step": 8505
    },
    {
      "entropy": 5.5201982975006105,
      "epoch": 0.7719521044992743,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004947731033745893,
      "loss": 5.4148,
      "mean_token_accuracy": 0.1677752047777176,
      "num_tokens": 17669671.0,
      "step": 8510
    },
    {
      "entropy": 5.661861324310303,
      "epoch": 0.7724056603773585,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004947661673365142,
      "loss": 5.5803,
      "mean_token_accuracy": 0.15949635207653046,
      "num_tokens": 17680927.0,
      "step": 8515
    },
    {
      "entropy": 5.592693090438843,
      "epoch": 0.7728592162554426,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004947592267536188,
      "loss": 5.452,
      "mean_token_accuracy": 0.169338421523571,
      "num_tokens": 17691378.0,
      "step": 8520
    },
    {
      "entropy": 5.6139641284942625,
      "epoch": 0.7733127721335269,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004947522816260466,
      "loss": 5.5283,
      "mean_token_accuracy": 0.16320052444934846,
      "num_tokens": 17700942.0,
      "step": 8525
    },
    {
      "entropy": 5.638260126113892,
      "epoch": 0.773766328011611,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004947453319539411,
      "loss": 5.4434,
      "mean_token_accuracy": 0.1692434623837471,
      "num_tokens": 17710076.0,
      "step": 8530
    },
    {
      "entropy": 5.60617470741272,
      "epoch": 0.7742198838896952,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004947383777374462,
      "loss": 5.4754,
      "mean_token_accuracy": 0.16550959944725036,
      "num_tokens": 17721009.0,
      "step": 8535
    },
    {
      "entropy": 5.536574983596802,
      "epoch": 0.7746734397677794,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004947314189767058,
      "loss": 5.4831,
      "mean_token_accuracy": 0.16459177136421205,
      "num_tokens": 17731821.0,
      "step": 8540
    },
    {
      "entropy": 5.5489661693573,
      "epoch": 0.7751269956458636,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004947244556718635,
      "loss": 5.3689,
      "mean_token_accuracy": 0.17378001660108566,
      "num_tokens": 17741425.0,
      "step": 8545
    },
    {
      "entropy": 5.602510833740235,
      "epoch": 0.7755805515239478,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004947174878230635,
      "loss": 5.422,
      "mean_token_accuracy": 0.17464102655649186,
      "num_tokens": 17752381.0,
      "step": 8550
    },
    {
      "entropy": 5.51142692565918,
      "epoch": 0.7760341074020319,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004947105154304498,
      "loss": 5.4165,
      "mean_token_accuracy": 0.16561309546232222,
      "num_tokens": 17762768.0,
      "step": 8555
    },
    {
      "entropy": 5.576867341995239,
      "epoch": 0.7764876632801161,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004947035384941667,
      "loss": 5.444,
      "mean_token_accuracy": 0.16490090787410736,
      "num_tokens": 17774032.0,
      "step": 8560
    },
    {
      "entropy": 5.681223964691162,
      "epoch": 0.7769412191582002,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004946965570143584,
      "loss": 5.5794,
      "mean_token_accuracy": 0.1618152603507042,
      "num_tokens": 17785098.0,
      "step": 8565
    },
    {
      "entropy": 5.647640323638916,
      "epoch": 0.7773947750362845,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004946895709911695,
      "loss": 5.4623,
      "mean_token_accuracy": 0.16696828752756118,
      "num_tokens": 17795487.0,
      "step": 8570
    },
    {
      "entropy": 5.615338945388794,
      "epoch": 0.7778483309143687,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004946825804247442,
      "loss": 5.5768,
      "mean_token_accuracy": 0.15942772179841996,
      "num_tokens": 17806045.0,
      "step": 8575
    },
    {
      "entropy": 5.659988880157471,
      "epoch": 0.7783018867924528,
      "grad_norm": 2.078125,
      "learning_rate": 0.000494675585315227,
      "loss": 5.5455,
      "mean_token_accuracy": 0.1624840170145035,
      "num_tokens": 17816948.0,
      "step": 8580
    },
    {
      "entropy": 5.621607828140259,
      "epoch": 0.778755442670537,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004946685856627628,
      "loss": 5.4557,
      "mean_token_accuracy": 0.16741457730531692,
      "num_tokens": 17826929.0,
      "step": 8585
    },
    {
      "entropy": 5.656989812850952,
      "epoch": 0.7792089985486212,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004946615814674963,
      "loss": 5.5614,
      "mean_token_accuracy": 0.16943255215883254,
      "num_tokens": 17837780.0,
      "step": 8590
    },
    {
      "entropy": 5.49801287651062,
      "epoch": 0.7796625544267054,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004946545727295724,
      "loss": 5.3401,
      "mean_token_accuracy": 0.1726154550909996,
      "num_tokens": 17847835.0,
      "step": 8595
    },
    {
      "entropy": 5.560629606246948,
      "epoch": 0.7801161103047896,
      "grad_norm": 1.375,
      "learning_rate": 0.0004946475594491359,
      "loss": 5.4298,
      "mean_token_accuracy": 0.1720295011997223,
      "num_tokens": 17858324.0,
      "step": 8600
    },
    {
      "entropy": 5.627698230743408,
      "epoch": 0.7805696661828737,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004946405416263319,
      "loss": 5.4823,
      "mean_token_accuracy": 0.16047568321228028,
      "num_tokens": 17868455.0,
      "step": 8605
    },
    {
      "entropy": 5.667699670791626,
      "epoch": 0.7810232220609579,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004946335192613054,
      "loss": 5.4701,
      "mean_token_accuracy": 0.16973807662725449,
      "num_tokens": 17879038.0,
      "step": 8610
    },
    {
      "entropy": 5.498501920700074,
      "epoch": 0.7814767779390421,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004946264923542019,
      "loss": 5.3344,
      "mean_token_accuracy": 0.17062352299690248,
      "num_tokens": 17888153.0,
      "step": 8615
    },
    {
      "entropy": 5.567123794555664,
      "epoch": 0.7819303338171263,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004946194609051664,
      "loss": 5.4711,
      "mean_token_accuracy": 0.17214930728077887,
      "num_tokens": 17898289.0,
      "step": 8620
    },
    {
      "entropy": 5.640062475204468,
      "epoch": 0.7823838896952104,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004946124249143447,
      "loss": 5.4642,
      "mean_token_accuracy": 0.16954964846372605,
      "num_tokens": 17909192.0,
      "step": 8625
    },
    {
      "entropy": 5.533799076080323,
      "epoch": 0.7828374455732946,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000494605384381882,
      "loss": 5.4139,
      "mean_token_accuracy": 0.1630323737859726,
      "num_tokens": 17918937.0,
      "step": 8630
    },
    {
      "entropy": 5.554599857330322,
      "epoch": 0.7832910014513788,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004945983393079239,
      "loss": 5.4596,
      "mean_token_accuracy": 0.17361268699169158,
      "num_tokens": 17929346.0,
      "step": 8635
    },
    {
      "entropy": 5.599385929107666,
      "epoch": 0.783744557329463,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004945912896926163,
      "loss": 5.4545,
      "mean_token_accuracy": 0.16786740422248841,
      "num_tokens": 17939479.0,
      "step": 8640
    },
    {
      "entropy": 5.6467255592346195,
      "epoch": 0.7841981132075472,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004945842355361049,
      "loss": 5.5292,
      "mean_token_accuracy": 0.16834045499563216,
      "num_tokens": 17949841.0,
      "step": 8645
    },
    {
      "entropy": 5.562479829788208,
      "epoch": 0.7846516690856313,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004945771768385355,
      "loss": 5.4616,
      "mean_token_accuracy": 0.168412184715271,
      "num_tokens": 17960334.0,
      "step": 8650
    },
    {
      "entropy": 5.593608140945435,
      "epoch": 0.7851052249637155,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000494570113600054,
      "loss": 5.4063,
      "mean_token_accuracy": 0.1693262502551079,
      "num_tokens": 17970595.0,
      "step": 8655
    },
    {
      "entropy": 5.642229700088501,
      "epoch": 0.7855587808417998,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004945630458208068,
      "loss": 5.5532,
      "mean_token_accuracy": 0.15822365134954453,
      "num_tokens": 17981301.0,
      "step": 8660
    },
    {
      "entropy": 5.552031660079956,
      "epoch": 0.7860123367198839,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004945559735009397,
      "loss": 5.378,
      "mean_token_accuracy": 0.17165354639291763,
      "num_tokens": 17991262.0,
      "step": 8665
    },
    {
      "entropy": 5.594365787506104,
      "epoch": 0.7864658925979681,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004945488966405993,
      "loss": 5.4197,
      "mean_token_accuracy": 0.17020548284053802,
      "num_tokens": 18001212.0,
      "step": 8670
    },
    {
      "entropy": 5.615402460098267,
      "epoch": 0.7869194484760522,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004945418152399317,
      "loss": 5.442,
      "mean_token_accuracy": 0.17426688373088836,
      "num_tokens": 18010432.0,
      "step": 8675
    },
    {
      "entropy": 5.61946291923523,
      "epoch": 0.7873730043541364,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004945347292990833,
      "loss": 5.4735,
      "mean_token_accuracy": 0.16698182374238968,
      "num_tokens": 18020902.0,
      "step": 8680
    },
    {
      "entropy": 5.618869018554688,
      "epoch": 0.7878265602322206,
      "grad_norm": 3.140625,
      "learning_rate": 0.0004945276388182008,
      "loss": 5.3486,
      "mean_token_accuracy": 0.16942863166332245,
      "num_tokens": 18031133.0,
      "step": 8685
    },
    {
      "entropy": 5.64780101776123,
      "epoch": 0.7882801161103048,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004945205437974309,
      "loss": 5.4595,
      "mean_token_accuracy": 0.17045676559209824,
      "num_tokens": 18042050.0,
      "step": 8690
    },
    {
      "entropy": 5.468453788757325,
      "epoch": 0.788733671988389,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004945134442369201,
      "loss": 5.4341,
      "mean_token_accuracy": 0.17557145655155182,
      "num_tokens": 18051642.0,
      "step": 8695
    },
    {
      "entropy": 5.660046672821045,
      "epoch": 0.7891872278664731,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004945063401368154,
      "loss": 5.4728,
      "mean_token_accuracy": 0.15648096203804016,
      "num_tokens": 18062180.0,
      "step": 8700
    },
    {
      "entropy": 5.644154119491577,
      "epoch": 0.7896407837445574,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004944992314972637,
      "loss": 5.4416,
      "mean_token_accuracy": 0.16445865631103515,
      "num_tokens": 18072302.0,
      "step": 8705
    },
    {
      "entropy": 5.597486782073974,
      "epoch": 0.7900943396226415,
      "grad_norm": 3.796875,
      "learning_rate": 0.0004944921183184118,
      "loss": 5.4514,
      "mean_token_accuracy": 0.1722353458404541,
      "num_tokens": 18081824.0,
      "step": 8710
    },
    {
      "entropy": 5.574193096160888,
      "epoch": 0.7905478955007257,
      "grad_norm": 1.859375,
      "learning_rate": 0.000494485000600407,
      "loss": 5.5003,
      "mean_token_accuracy": 0.17043693363666534,
      "num_tokens": 18092696.0,
      "step": 8715
    },
    {
      "entropy": 5.570696210861206,
      "epoch": 0.7910014513788098,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004944778783433965,
      "loss": 5.4079,
      "mean_token_accuracy": 0.17257224768400192,
      "num_tokens": 18101790.0,
      "step": 8720
    },
    {
      "entropy": 5.677782106399536,
      "epoch": 0.791455007256894,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004944707515475276,
      "loss": 5.5331,
      "mean_token_accuracy": 0.16561739444732665,
      "num_tokens": 18111572.0,
      "step": 8725
    },
    {
      "entropy": 5.673485660552979,
      "epoch": 0.7919085631349783,
      "grad_norm": 3.9375,
      "learning_rate": 0.0004944636202129475,
      "loss": 5.501,
      "mean_token_accuracy": 0.17471941858530043,
      "num_tokens": 18120650.0,
      "step": 8730
    },
    {
      "entropy": 5.583009815216064,
      "epoch": 0.7923621190130624,
      "grad_norm": 3.015625,
      "learning_rate": 0.0004944564843398039,
      "loss": 5.3872,
      "mean_token_accuracy": 0.1823519989848137,
      "num_tokens": 18131055.0,
      "step": 8735
    },
    {
      "entropy": 5.659669351577759,
      "epoch": 0.7928156748911466,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004944493439282443,
      "loss": 5.4851,
      "mean_token_accuracy": 0.1695306792855263,
      "num_tokens": 18141423.0,
      "step": 8740
    },
    {
      "entropy": 5.525578498840332,
      "epoch": 0.7932692307692307,
      "grad_norm": 2.828125,
      "learning_rate": 0.0004944421989784162,
      "loss": 5.4181,
      "mean_token_accuracy": 0.17386405616998674,
      "num_tokens": 18152015.0,
      "step": 8745
    },
    {
      "entropy": 5.573276948928833,
      "epoch": 0.793722786647315,
      "grad_norm": 1.875,
      "learning_rate": 0.0004944350494904676,
      "loss": 5.5188,
      "mean_token_accuracy": 0.16456999033689498,
      "num_tokens": 18161934.0,
      "step": 8750
    },
    {
      "entropy": 5.623917818069458,
      "epoch": 0.7941763425253991,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004944278954645463,
      "loss": 5.4113,
      "mean_token_accuracy": 0.17385263442993165,
      "num_tokens": 18172287.0,
      "step": 8755
    },
    {
      "entropy": 5.685469770431519,
      "epoch": 0.7946298984034833,
      "grad_norm": 2.828125,
      "learning_rate": 0.0004944207369008002,
      "loss": 5.4886,
      "mean_token_accuracy": 0.17678707838058472,
      "num_tokens": 18181015.0,
      "step": 8760
    },
    {
      "entropy": 5.643927192687988,
      "epoch": 0.7950834542815675,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004944135737993773,
      "loss": 5.5563,
      "mean_token_accuracy": 0.1640169993042946,
      "num_tokens": 18192382.0,
      "step": 8765
    },
    {
      "entropy": 5.567079830169678,
      "epoch": 0.7955370101596516,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004944064061604258,
      "loss": 5.4184,
      "mean_token_accuracy": 0.17833883166313172,
      "num_tokens": 18202597.0,
      "step": 8770
    },
    {
      "entropy": 5.666832399368286,
      "epoch": 0.7959905660377359,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004943992339840939,
      "loss": 5.4887,
      "mean_token_accuracy": 0.16130754947662354,
      "num_tokens": 18213756.0,
      "step": 8775
    },
    {
      "entropy": 5.580977487564087,
      "epoch": 0.79644412191582,
      "grad_norm": 2.03125,
      "learning_rate": 0.00049439205727053,
      "loss": 5.4197,
      "mean_token_accuracy": 0.1659858763217926,
      "num_tokens": 18223669.0,
      "step": 8780
    },
    {
      "entropy": 5.619691276550293,
      "epoch": 0.7968976777939042,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004943848760198823,
      "loss": 5.445,
      "mean_token_accuracy": 0.1736903518438339,
      "num_tokens": 18234884.0,
      "step": 8785
    },
    {
      "entropy": 5.579283094406128,
      "epoch": 0.7973512336719883,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004943776902322995,
      "loss": 5.3905,
      "mean_token_accuracy": 0.16901502311229705,
      "num_tokens": 18245155.0,
      "step": 8790
    },
    {
      "entropy": 5.499757480621338,
      "epoch": 0.7978047895500726,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004943704999079302,
      "loss": 5.3245,
      "mean_token_accuracy": 0.17517054378986358,
      "num_tokens": 18254686.0,
      "step": 8795
    },
    {
      "entropy": 5.565064907073975,
      "epoch": 0.7982583454281568,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004943633050469231,
      "loss": 5.3535,
      "mean_token_accuracy": 0.17353624999523162,
      "num_tokens": 18265473.0,
      "step": 8800
    },
    {
      "entropy": 5.671705102920532,
      "epoch": 0.7987119013062409,
      "grad_norm": 7.375,
      "learning_rate": 0.0004943561056494269,
      "loss": 5.5584,
      "mean_token_accuracy": 0.16333960890769958,
      "num_tokens": 18276024.0,
      "step": 8805
    },
    {
      "entropy": 5.6350846767425535,
      "epoch": 0.7991654571843251,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004943489017155906,
      "loss": 5.5076,
      "mean_token_accuracy": 0.15510011464357376,
      "num_tokens": 18287261.0,
      "step": 8810
    },
    {
      "entropy": 5.618888854980469,
      "epoch": 0.7996190130624092,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004943416932455631,
      "loss": 5.523,
      "mean_token_accuracy": 0.16325552016496658,
      "num_tokens": 18299374.0,
      "step": 8815
    },
    {
      "entropy": 5.56708722114563,
      "epoch": 0.8000725689404935,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004943344802394935,
      "loss": 5.4844,
      "mean_token_accuracy": 0.16325221210718155,
      "num_tokens": 18309344.0,
      "step": 8820
    },
    {
      "entropy": 5.679027700424195,
      "epoch": 0.8005261248185777,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004943272626975309,
      "loss": 5.4854,
      "mean_token_accuracy": 0.16476835906505585,
      "num_tokens": 18319778.0,
      "step": 8825
    },
    {
      "entropy": 5.721097755432129,
      "epoch": 0.8009796806966618,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004943200406198246,
      "loss": 5.4801,
      "mean_token_accuracy": 0.1666264146566391,
      "num_tokens": 18329787.0,
      "step": 8830
    },
    {
      "entropy": 5.5708975315094,
      "epoch": 0.801433236574746,
      "grad_norm": 2.515625,
      "learning_rate": 0.000494312814006524,
      "loss": 5.4358,
      "mean_token_accuracy": 0.17473117858171464,
      "num_tokens": 18340670.0,
      "step": 8835
    },
    {
      "entropy": 5.561055850982666,
      "epoch": 0.8018867924528302,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004943055828577784,
      "loss": 5.4336,
      "mean_token_accuracy": 0.16693542152643204,
      "num_tokens": 18352484.0,
      "step": 8840
    },
    {
      "entropy": 5.611057615280151,
      "epoch": 0.8023403483309144,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004942983471737376,
      "loss": 5.387,
      "mean_token_accuracy": 0.1695320188999176,
      "num_tokens": 18363103.0,
      "step": 8845
    },
    {
      "entropy": 5.557615613937378,
      "epoch": 0.8027939042089985,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004942911069545511,
      "loss": 5.4384,
      "mean_token_accuracy": 0.17061060220003127,
      "num_tokens": 18373925.0,
      "step": 8850
    },
    {
      "entropy": 5.607903099060058,
      "epoch": 0.8032474600870827,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004942838622003686,
      "loss": 5.5177,
      "mean_token_accuracy": 0.16096073687076567,
      "num_tokens": 18384081.0,
      "step": 8855
    },
    {
      "entropy": 5.642839336395264,
      "epoch": 0.8037010159651669,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004942766129113399,
      "loss": 5.4022,
      "mean_token_accuracy": 0.16844492107629777,
      "num_tokens": 18392510.0,
      "step": 8860
    },
    {
      "entropy": 5.689938163757324,
      "epoch": 0.8041545718432511,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004942693590876151,
      "loss": 5.5265,
      "mean_token_accuracy": 0.16685045510530472,
      "num_tokens": 18402809.0,
      "step": 8865
    },
    {
      "entropy": 5.6905910015106205,
      "epoch": 0.8046081277213353,
      "grad_norm": 1.71875,
      "learning_rate": 0.000494262100729344,
      "loss": 5.5423,
      "mean_token_accuracy": 0.16762113571166992,
      "num_tokens": 18413342.0,
      "step": 8870
    },
    {
      "entropy": 5.62622389793396,
      "epoch": 0.8050616835994194,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004942548378366768,
      "loss": 5.5462,
      "mean_token_accuracy": 0.16534917205572128,
      "num_tokens": 18424735.0,
      "step": 8875
    },
    {
      "entropy": 5.619510841369629,
      "epoch": 0.8055152394775036,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004942475704097635,
      "loss": 5.4241,
      "mean_token_accuracy": 0.1706240862607956,
      "num_tokens": 18434797.0,
      "step": 8880
    },
    {
      "entropy": 5.576655149459839,
      "epoch": 0.8059687953555879,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004942402984487548,
      "loss": 5.4938,
      "mean_token_accuracy": 0.16197436153888703,
      "num_tokens": 18445456.0,
      "step": 8885
    },
    {
      "entropy": 5.545839595794678,
      "epoch": 0.806422351233672,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004942330219538008,
      "loss": 5.4537,
      "mean_token_accuracy": 0.163727068901062,
      "num_tokens": 18456728.0,
      "step": 8890
    },
    {
      "entropy": 5.749781274795533,
      "epoch": 0.8068759071117562,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000494225740925052,
      "loss": 5.5981,
      "mean_token_accuracy": 0.16090889275074005,
      "num_tokens": 18468813.0,
      "step": 8895
    },
    {
      "entropy": 5.674880886077881,
      "epoch": 0.8073294629898403,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004942184553626591,
      "loss": 5.4916,
      "mean_token_accuracy": 0.1678992658853531,
      "num_tokens": 18477912.0,
      "step": 8900
    },
    {
      "entropy": 5.607471370697022,
      "epoch": 0.8077830188679245,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004942111652667726,
      "loss": 5.4312,
      "mean_token_accuracy": 0.17500946670770645,
      "num_tokens": 18488115.0,
      "step": 8905
    },
    {
      "entropy": 5.58686318397522,
      "epoch": 0.8082365747460087,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004942038706375433,
      "loss": 5.4746,
      "mean_token_accuracy": 0.16537439674139023,
      "num_tokens": 18498853.0,
      "step": 8910
    },
    {
      "entropy": 5.577048015594483,
      "epoch": 0.8086901306240929,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004941965714751222,
      "loss": 5.5233,
      "mean_token_accuracy": 0.16133568957448005,
      "num_tokens": 18510047.0,
      "step": 8915
    },
    {
      "entropy": 5.638028335571289,
      "epoch": 0.809143686502177,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004941892677796601,
      "loss": 5.4787,
      "mean_token_accuracy": 0.17180612534284592,
      "num_tokens": 18519673.0,
      "step": 8920
    },
    {
      "entropy": 5.619547414779663,
      "epoch": 0.8095972423802612,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004941819595513081,
      "loss": 5.5156,
      "mean_token_accuracy": 0.16819463521242142,
      "num_tokens": 18530090.0,
      "step": 8925
    },
    {
      "entropy": 5.719571781158447,
      "epoch": 0.8100507982583455,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004941746467902173,
      "loss": 5.5421,
      "mean_token_accuracy": 0.15966865420341492,
      "num_tokens": 18541862.0,
      "step": 8930
    },
    {
      "entropy": 5.710291719436645,
      "epoch": 0.8105043541364296,
      "grad_norm": 2.0625,
      "learning_rate": 0.000494167329496539,
      "loss": 5.5275,
      "mean_token_accuracy": 0.15767375528812408,
      "num_tokens": 18551624.0,
      "step": 8935
    },
    {
      "entropy": 5.617796134948731,
      "epoch": 0.8109579100145138,
      "grad_norm": 2.0,
      "learning_rate": 0.0004941600076704244,
      "loss": 5.4768,
      "mean_token_accuracy": 0.1689252883195877,
      "num_tokens": 18561784.0,
      "step": 8940
    },
    {
      "entropy": 5.647332811355591,
      "epoch": 0.8114114658925979,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004941526813120251,
      "loss": 5.5331,
      "mean_token_accuracy": 0.1632409393787384,
      "num_tokens": 18572110.0,
      "step": 8945
    },
    {
      "entropy": 5.661430311203003,
      "epoch": 0.8118650217706821,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004941453504214924,
      "loss": 5.4947,
      "mean_token_accuracy": 0.16330064535140992,
      "num_tokens": 18582012.0,
      "step": 8950
    },
    {
      "entropy": 5.565213871002197,
      "epoch": 0.8123185776487664,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004941380149989781,
      "loss": 5.3471,
      "mean_token_accuracy": 0.17561429738998413,
      "num_tokens": 18592850.0,
      "step": 8955
    },
    {
      "entropy": 5.594003438949585,
      "epoch": 0.8127721335268505,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004941306750446336,
      "loss": 5.509,
      "mean_token_accuracy": 0.16701007038354873,
      "num_tokens": 18602590.0,
      "step": 8960
    },
    {
      "entropy": 5.596885013580322,
      "epoch": 0.8132256894049347,
      "grad_norm": 1.625,
      "learning_rate": 0.000494123330558611,
      "loss": 5.4151,
      "mean_token_accuracy": 0.16790961176157,
      "num_tokens": 18612607.0,
      "step": 8965
    },
    {
      "entropy": 5.572688865661621,
      "epoch": 0.8136792452830188,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004941159815410622,
      "loss": 5.3953,
      "mean_token_accuracy": 0.17194055318832396,
      "num_tokens": 18622936.0,
      "step": 8970
    },
    {
      "entropy": 5.582172870635986,
      "epoch": 0.8141328011611031,
      "grad_norm": 2.640625,
      "learning_rate": 0.000494108627992139,
      "loss": 5.5593,
      "mean_token_accuracy": 0.15717708468437194,
      "num_tokens": 18633997.0,
      "step": 8975
    },
    {
      "entropy": 5.608917236328125,
      "epoch": 0.8145863570391872,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004941012699119936,
      "loss": 5.3884,
      "mean_token_accuracy": 0.17790337949991225,
      "num_tokens": 18644347.0,
      "step": 8980
    },
    {
      "entropy": 5.617583227157593,
      "epoch": 0.8150399129172714,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004940939073007779,
      "loss": 5.4721,
      "mean_token_accuracy": 0.16982878595590592,
      "num_tokens": 18654989.0,
      "step": 8985
    },
    {
      "entropy": 5.628685855865479,
      "epoch": 0.8154934687953556,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004940865401586446,
      "loss": 5.4521,
      "mean_token_accuracy": 0.16858104318380357,
      "num_tokens": 18665526.0,
      "step": 8990
    },
    {
      "entropy": 5.599122714996338,
      "epoch": 0.8159470246734397,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004940791684857458,
      "loss": 5.4407,
      "mean_token_accuracy": 0.1744929790496826,
      "num_tokens": 18676226.0,
      "step": 8995
    },
    {
      "entropy": 5.596544647216797,
      "epoch": 0.816400580551524,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004940717922822339,
      "loss": 5.3821,
      "mean_token_accuracy": 0.1740715578198433,
      "num_tokens": 18686499.0,
      "step": 9000
    },
    {
      "epoch": 0.816400580551524,
      "eval_entropy": 5.499293960273535,
      "eval_loss": 5.492147922515869,
      "eval_mean_token_accuracy": 0.17267374004494415,
      "eval_num_tokens": 18686499.0,
      "eval_runtime": 26.5091,
      "eval_samples_per_second": 1333.882,
      "eval_steps_per_second": 166.735,
      "step": 9000
    },
    {
      "entropy": 5.648325109481812,
      "epoch": 0.8168541364296081,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004940644115482615,
      "loss": 5.4541,
      "mean_token_accuracy": 0.16429938971996308,
      "num_tokens": 18696543.0,
      "step": 9005
    },
    {
      "entropy": 5.581717205047608,
      "epoch": 0.8173076923076923,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004940570262839812,
      "loss": 5.5077,
      "mean_token_accuracy": 0.1612074851989746,
      "num_tokens": 18707029.0,
      "step": 9010
    },
    {
      "entropy": 5.606132745742798,
      "epoch": 0.8177612481857764,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004940496364895457,
      "loss": 5.4899,
      "mean_token_accuracy": 0.16660574823617935,
      "num_tokens": 18716937.0,
      "step": 9015
    },
    {
      "entropy": 5.674174022674561,
      "epoch": 0.8182148040638607,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004940422421651082,
      "loss": 5.4624,
      "mean_token_accuracy": 0.16896364241838455,
      "num_tokens": 18727478.0,
      "step": 9020
    },
    {
      "entropy": 5.573435878753662,
      "epoch": 0.8186683599419449,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004940348433108211,
      "loss": 5.4041,
      "mean_token_accuracy": 0.16320528239011764,
      "num_tokens": 18737339.0,
      "step": 9025
    },
    {
      "entropy": 5.609128522872925,
      "epoch": 0.819121915820029,
      "grad_norm": 2.375,
      "learning_rate": 0.0004940274399268377,
      "loss": 5.5113,
      "mean_token_accuracy": 0.16753003001213074,
      "num_tokens": 18747810.0,
      "step": 9030
    },
    {
      "entropy": 5.60022087097168,
      "epoch": 0.8195754716981132,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004940200320133109,
      "loss": 5.4681,
      "mean_token_accuracy": 0.16738731861114503,
      "num_tokens": 18757564.0,
      "step": 9035
    },
    {
      "entropy": 5.6804650783538815,
      "epoch": 0.8200290275761973,
      "grad_norm": 1.625,
      "learning_rate": 0.0004940126195703941,
      "loss": 5.5088,
      "mean_token_accuracy": 0.16182428747415542,
      "num_tokens": 18768196.0,
      "step": 9040
    },
    {
      "entropy": 5.6730348587036135,
      "epoch": 0.8204825834542816,
      "grad_norm": 3.28125,
      "learning_rate": 0.0004940052025982405,
      "loss": 5.4786,
      "mean_token_accuracy": 0.17686939835548401,
      "num_tokens": 18778350.0,
      "step": 9045
    },
    {
      "entropy": 5.607815074920654,
      "epoch": 0.8209361393323658,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004939977810970035,
      "loss": 5.435,
      "mean_token_accuracy": 0.16083383411169053,
      "num_tokens": 18789825.0,
      "step": 9050
    },
    {
      "entropy": 5.555605840682984,
      "epoch": 0.8213896952104499,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004939903550668366,
      "loss": 5.4238,
      "mean_token_accuracy": 0.16935580968856812,
      "num_tokens": 18799793.0,
      "step": 9055
    },
    {
      "entropy": 5.564256954193115,
      "epoch": 0.8218432510885341,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004939829245078934,
      "loss": 5.4685,
      "mean_token_accuracy": 0.16753165423870087,
      "num_tokens": 18809771.0,
      "step": 9060
    },
    {
      "entropy": 5.619735240936279,
      "epoch": 0.8222968069666183,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004939754894203275,
      "loss": 5.4987,
      "mean_token_accuracy": 0.16463650315999984,
      "num_tokens": 18820529.0,
      "step": 9065
    },
    {
      "entropy": 5.5516197204589846,
      "epoch": 0.8227503628447025,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004939680498042926,
      "loss": 5.416,
      "mean_token_accuracy": 0.17473565191030502,
      "num_tokens": 18830797.0,
      "step": 9070
    },
    {
      "entropy": 5.579305028915405,
      "epoch": 0.8232039187227866,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004939606056599428,
      "loss": 5.4075,
      "mean_token_accuracy": 0.16576789766550065,
      "num_tokens": 18840526.0,
      "step": 9075
    },
    {
      "entropy": 5.575499153137207,
      "epoch": 0.8236574746008708,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004939531569874316,
      "loss": 5.4431,
      "mean_token_accuracy": 0.16471141427755356,
      "num_tokens": 18851201.0,
      "step": 9080
    },
    {
      "entropy": 5.503150367736817,
      "epoch": 0.824111030478955,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004939457037869135,
      "loss": 5.259,
      "mean_token_accuracy": 0.18129919618368148,
      "num_tokens": 18861676.0,
      "step": 9085
    },
    {
      "entropy": 5.5292870044708256,
      "epoch": 0.8245645863570392,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004939382460585424,
      "loss": 5.4173,
      "mean_token_accuracy": 0.1645190566778183,
      "num_tokens": 18872669.0,
      "step": 9090
    },
    {
      "entropy": 5.5579780578613285,
      "epoch": 0.8250181422351234,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004939307838024726,
      "loss": 5.4602,
      "mean_token_accuracy": 0.16422475427389144,
      "num_tokens": 18882130.0,
      "step": 9095
    },
    {
      "entropy": 5.5297197818756105,
      "epoch": 0.8254716981132075,
      "grad_norm": 1.625,
      "learning_rate": 0.0004939233170188585,
      "loss": 5.4326,
      "mean_token_accuracy": 0.16917428374290466,
      "num_tokens": 18892457.0,
      "step": 9100
    },
    {
      "entropy": 5.6344566822052,
      "epoch": 0.8259252539912917,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004939158457078543,
      "loss": 5.4585,
      "mean_token_accuracy": 0.16588743925094604,
      "num_tokens": 18903152.0,
      "step": 9105
    },
    {
      "entropy": 5.5791308879852295,
      "epoch": 0.826378809869376,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004939083698696146,
      "loss": 5.429,
      "mean_token_accuracy": 0.16439961045980453,
      "num_tokens": 18913740.0,
      "step": 9110
    },
    {
      "entropy": 5.640698719024658,
      "epoch": 0.8268323657474601,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000493900889504294,
      "loss": 5.4674,
      "mean_token_accuracy": 0.17006396055221557,
      "num_tokens": 18923501.0,
      "step": 9115
    },
    {
      "entropy": 5.62080750465393,
      "epoch": 0.8272859216255443,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004938934046120473,
      "loss": 5.4622,
      "mean_token_accuracy": 0.16710171848535538,
      "num_tokens": 18933255.0,
      "step": 9120
    },
    {
      "entropy": 5.551970100402832,
      "epoch": 0.8277394775036284,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004938859151930293,
      "loss": 5.4103,
      "mean_token_accuracy": 0.1690342456102371,
      "num_tokens": 18943649.0,
      "step": 9125
    },
    {
      "entropy": 5.584992265701294,
      "epoch": 0.8281930333817126,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004938784212473947,
      "loss": 5.3262,
      "mean_token_accuracy": 0.17703648358583451,
      "num_tokens": 18953187.0,
      "step": 9130
    },
    {
      "entropy": 5.631279706954956,
      "epoch": 0.8286465892597968,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004938709227752985,
      "loss": 5.4605,
      "mean_token_accuracy": 0.16520616114139558,
      "num_tokens": 18963015.0,
      "step": 9135
    },
    {
      "entropy": 5.565444612503052,
      "epoch": 0.829100145137881,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000493863419776896,
      "loss": 5.4996,
      "mean_token_accuracy": 0.16376311779022218,
      "num_tokens": 18974184.0,
      "step": 9140
    },
    {
      "entropy": 5.549704885482788,
      "epoch": 0.8295537010159652,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004938559122523421,
      "loss": 5.4252,
      "mean_token_accuracy": 0.16940660178661346,
      "num_tokens": 18985036.0,
      "step": 9145
    },
    {
      "entropy": 5.625360918045044,
      "epoch": 0.8300072568940493,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004938484002017921,
      "loss": 5.5363,
      "mean_token_accuracy": 0.16343602538108826,
      "num_tokens": 18995237.0,
      "step": 9150
    },
    {
      "entropy": 5.653959465026856,
      "epoch": 0.8304608127721336,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004938408836254015,
      "loss": 5.5356,
      "mean_token_accuracy": 0.16166142523288726,
      "num_tokens": 19006272.0,
      "step": 9155
    },
    {
      "entropy": 5.618923091888428,
      "epoch": 0.8309143686502177,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004938333625233258,
      "loss": 5.442,
      "mean_token_accuracy": 0.16922508031129838,
      "num_tokens": 19016728.0,
      "step": 9160
    },
    {
      "entropy": 5.653804159164428,
      "epoch": 0.8313679245283019,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004938258368957202,
      "loss": 5.4964,
      "mean_token_accuracy": 0.16299268156290053,
      "num_tokens": 19028024.0,
      "step": 9165
    },
    {
      "entropy": 5.588284826278686,
      "epoch": 0.831821480406386,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004938183067427406,
      "loss": 5.4522,
      "mean_token_accuracy": 0.17027496546506882,
      "num_tokens": 19038218.0,
      "step": 9170
    },
    {
      "entropy": 5.565521621704102,
      "epoch": 0.8322750362844702,
      "grad_norm": 7.5625,
      "learning_rate": 0.0004938107720645426,
      "loss": 5.4822,
      "mean_token_accuracy": 0.16394347697496414,
      "num_tokens": 19048574.0,
      "step": 9175
    },
    {
      "entropy": 5.626050996780395,
      "epoch": 0.8327285921625545,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004938032328612821,
      "loss": 5.4424,
      "mean_token_accuracy": 0.16529956907033921,
      "num_tokens": 19057721.0,
      "step": 9180
    },
    {
      "entropy": 5.588996124267578,
      "epoch": 0.8331821480406386,
      "grad_norm": 2.484375,
      "learning_rate": 0.000493795689133115,
      "loss": 5.4157,
      "mean_token_accuracy": 0.1627140760421753,
      "num_tokens": 19067582.0,
      "step": 9185
    },
    {
      "entropy": 5.563428068161011,
      "epoch": 0.8336357039187228,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004937881408801974,
      "loss": 5.3863,
      "mean_token_accuracy": 0.1715189129114151,
      "num_tokens": 19077896.0,
      "step": 9190
    },
    {
      "entropy": 5.598232936859131,
      "epoch": 0.8340892597968069,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000493780588102685,
      "loss": 5.4768,
      "mean_token_accuracy": 0.1720636621117592,
      "num_tokens": 19088082.0,
      "step": 9195
    },
    {
      "entropy": 5.667886781692505,
      "epoch": 0.8345428156748912,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004937730308007346,
      "loss": 5.5003,
      "mean_token_accuracy": 0.16846424341201782,
      "num_tokens": 19097176.0,
      "step": 9200
    },
    {
      "entropy": 5.573349237442017,
      "epoch": 0.8349963715529753,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004937654689745021,
      "loss": 5.4085,
      "mean_token_accuracy": 0.1709039628505707,
      "num_tokens": 19107835.0,
      "step": 9205
    },
    {
      "entropy": 5.536523103713989,
      "epoch": 0.8354499274310595,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004937579026241439,
      "loss": 5.4,
      "mean_token_accuracy": 0.17664874494075775,
      "num_tokens": 19118028.0,
      "step": 9210
    },
    {
      "entropy": 5.637387037277222,
      "epoch": 0.8359034833091437,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004937503317498165,
      "loss": 5.437,
      "mean_token_accuracy": 0.1704498365521431,
      "num_tokens": 19129223.0,
      "step": 9215
    },
    {
      "entropy": 5.603167486190796,
      "epoch": 0.8363570391872278,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004937427563516765,
      "loss": 5.5498,
      "mean_token_accuracy": 0.16333572715520858,
      "num_tokens": 19140048.0,
      "step": 9220
    },
    {
      "entropy": 5.6831597805023195,
      "epoch": 0.8368105950653121,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004937351764298805,
      "loss": 5.5027,
      "mean_token_accuracy": 0.16168099492788315,
      "num_tokens": 19150853.0,
      "step": 9225
    },
    {
      "entropy": 5.675702953338623,
      "epoch": 0.8372641509433962,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004937275919845855,
      "loss": 5.4946,
      "mean_token_accuracy": 0.16548409759998323,
      "num_tokens": 19161645.0,
      "step": 9230
    },
    {
      "entropy": 5.678725910186768,
      "epoch": 0.8377177068214804,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004937200030159481,
      "loss": 5.5681,
      "mean_token_accuracy": 0.16480537950992585,
      "num_tokens": 19173276.0,
      "step": 9235
    },
    {
      "entropy": 5.632607173919678,
      "epoch": 0.8381712626995645,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004937124095241252,
      "loss": 5.4703,
      "mean_token_accuracy": 0.17256869822740556,
      "num_tokens": 19182949.0,
      "step": 9240
    },
    {
      "entropy": 5.630819845199585,
      "epoch": 0.8386248185776488,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004937048115092739,
      "loss": 5.4557,
      "mean_token_accuracy": 0.16706619113683702,
      "num_tokens": 19193504.0,
      "step": 9245
    },
    {
      "entropy": 5.610013198852539,
      "epoch": 0.839078374455733,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004936972089715515,
      "loss": 5.5351,
      "mean_token_accuracy": 0.15942700803279877,
      "num_tokens": 19203406.0,
      "step": 9250
    },
    {
      "entropy": 5.647968816757202,
      "epoch": 0.8395319303338171,
      "grad_norm": 2.03125,
      "learning_rate": 0.000493689601911115,
      "loss": 5.484,
      "mean_token_accuracy": 0.16736481189727784,
      "num_tokens": 19214325.0,
      "step": 9255
    },
    {
      "entropy": 5.619275045394898,
      "epoch": 0.8399854862119013,
      "grad_norm": 2.625,
      "learning_rate": 0.0004936819903281219,
      "loss": 5.5216,
      "mean_token_accuracy": 0.15449996888637543,
      "num_tokens": 19226466.0,
      "step": 9260
    },
    {
      "entropy": 5.677626037597657,
      "epoch": 0.8404390420899854,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004936743742227295,
      "loss": 5.5348,
      "mean_token_accuracy": 0.16590409278869628,
      "num_tokens": 19237032.0,
      "step": 9265
    },
    {
      "entropy": 5.7064573764801025,
      "epoch": 0.8408925979680697,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004936667535950952,
      "loss": 5.5256,
      "mean_token_accuracy": 0.16213936507701873,
      "num_tokens": 19248009.0,
      "step": 9270
    },
    {
      "entropy": 5.655757761001587,
      "epoch": 0.8413461538461539,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004936591284453768,
      "loss": 5.574,
      "mean_token_accuracy": 0.16394732594490052,
      "num_tokens": 19259363.0,
      "step": 9275
    },
    {
      "entropy": 5.610392808914185,
      "epoch": 0.841799709724238,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004936514987737319,
      "loss": 5.4925,
      "mean_token_accuracy": 0.16612009406089784,
      "num_tokens": 19269191.0,
      "step": 9280
    },
    {
      "entropy": 5.631801128387451,
      "epoch": 0.8422532656023222,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004936438645803182,
      "loss": 5.4471,
      "mean_token_accuracy": 0.16355190873146058,
      "num_tokens": 19280278.0,
      "step": 9285
    },
    {
      "entropy": 5.6273674964904785,
      "epoch": 0.8427068214804064,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004936362258652938,
      "loss": 5.4404,
      "mean_token_accuracy": 0.16636256724596024,
      "num_tokens": 19290130.0,
      "step": 9290
    },
    {
      "entropy": 5.745988178253174,
      "epoch": 0.8431603773584906,
      "grad_norm": 1.375,
      "learning_rate": 0.0004936285826288165,
      "loss": 5.558,
      "mean_token_accuracy": 0.16738905161619186,
      "num_tokens": 19301669.0,
      "step": 9295
    },
    {
      "entropy": 5.549793291091919,
      "epoch": 0.8436139332365747,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004936209348710444,
      "loss": 5.4185,
      "mean_token_accuracy": 0.1677660569548607,
      "num_tokens": 19312490.0,
      "step": 9300
    },
    {
      "entropy": 5.5131879329681395,
      "epoch": 0.8440674891146589,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004936132825921355,
      "loss": 5.4514,
      "mean_token_accuracy": 0.1659799888730049,
      "num_tokens": 19321743.0,
      "step": 9305
    },
    {
      "entropy": 5.606175374984741,
      "epoch": 0.844521044992743,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004936056257922484,
      "loss": 5.3578,
      "mean_token_accuracy": 0.1735738232731819,
      "num_tokens": 19332149.0,
      "step": 9310
    },
    {
      "entropy": 5.589995050430298,
      "epoch": 0.8449746008708273,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004935979644715412,
      "loss": 5.46,
      "mean_token_accuracy": 0.16639793962240218,
      "num_tokens": 19341655.0,
      "step": 9315
    },
    {
      "entropy": 5.573223495483399,
      "epoch": 0.8454281567489115,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004935902986301724,
      "loss": 5.503,
      "mean_token_accuracy": 0.16116373836994172,
      "num_tokens": 19351404.0,
      "step": 9320
    },
    {
      "entropy": 5.657394313812256,
      "epoch": 0.8458817126269956,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004935826282683005,
      "loss": 5.5445,
      "mean_token_accuracy": 0.16180313229560853,
      "num_tokens": 19361788.0,
      "step": 9325
    },
    {
      "entropy": 5.650906705856324,
      "epoch": 0.8463352685050798,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004935749533860842,
      "loss": 5.5018,
      "mean_token_accuracy": 0.16616404056549072,
      "num_tokens": 19372902.0,
      "step": 9330
    },
    {
      "entropy": 5.5491042137146,
      "epoch": 0.846788824383164,
      "grad_norm": 1.6875,
      "learning_rate": 0.000493567273983682,
      "loss": 5.3628,
      "mean_token_accuracy": 0.16852847784757613,
      "num_tokens": 19382181.0,
      "step": 9335
    },
    {
      "entropy": 5.611289405822754,
      "epoch": 0.8472423802612482,
      "grad_norm": 1.890625,
      "learning_rate": 0.000493559590061253,
      "loss": 5.437,
      "mean_token_accuracy": 0.16242333501577377,
      "num_tokens": 19393692.0,
      "step": 9340
    },
    {
      "entropy": 5.629254102706909,
      "epoch": 0.8476959361393324,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004935519016189559,
      "loss": 5.4866,
      "mean_token_accuracy": 0.17058742940425872,
      "num_tokens": 19404587.0,
      "step": 9345
    },
    {
      "entropy": 5.630183506011963,
      "epoch": 0.8481494920174165,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004935442086569498,
      "loss": 5.4523,
      "mean_token_accuracy": 0.16475782990455629,
      "num_tokens": 19415296.0,
      "step": 9350
    },
    {
      "entropy": 5.683138704299926,
      "epoch": 0.8486030478955007,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004935365111753939,
      "loss": 5.519,
      "mean_token_accuracy": 0.1648883879184723,
      "num_tokens": 19426119.0,
      "step": 9355
    },
    {
      "entropy": 5.616968631744385,
      "epoch": 0.8490566037735849,
      "grad_norm": 1.75,
      "learning_rate": 0.0004935288091744472,
      "loss": 5.515,
      "mean_token_accuracy": 0.16678015291690826,
      "num_tokens": 19436828.0,
      "step": 9360
    },
    {
      "entropy": 5.645136785507202,
      "epoch": 0.8495101596516691,
      "grad_norm": 1.609375,
      "learning_rate": 0.000493521102654269,
      "loss": 5.4853,
      "mean_token_accuracy": 0.16086220145225524,
      "num_tokens": 19446583.0,
      "step": 9365
    },
    {
      "entropy": 5.590010356903076,
      "epoch": 0.8499637155297532,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004935133916150187,
      "loss": 5.3769,
      "mean_token_accuracy": 0.1739301025867462,
      "num_tokens": 19456757.0,
      "step": 9370
    },
    {
      "entropy": 5.503833961486817,
      "epoch": 0.8504172714078374,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004935056760568558,
      "loss": 5.4446,
      "mean_token_accuracy": 0.16398343741893767,
      "num_tokens": 19467624.0,
      "step": 9375
    },
    {
      "entropy": 5.5467448234558105,
      "epoch": 0.8508708272859217,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004934979559799399,
      "loss": 5.4102,
      "mean_token_accuracy": 0.16686284244060517,
      "num_tokens": 19477759.0,
      "step": 9380
    },
    {
      "entropy": 5.6368509292602536,
      "epoch": 0.8513243831640058,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004934902313844306,
      "loss": 5.4104,
      "mean_token_accuracy": 0.16647852212190628,
      "num_tokens": 19488877.0,
      "step": 9385
    },
    {
      "entropy": 5.565980243682861,
      "epoch": 0.85177793904209,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004934825022704875,
      "loss": 5.418,
      "mean_token_accuracy": 0.16903312206268312,
      "num_tokens": 19499170.0,
      "step": 9390
    },
    {
      "entropy": 5.551848745346069,
      "epoch": 0.8522314949201741,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004934747686382707,
      "loss": 5.439,
      "mean_token_accuracy": 0.15809385925531388,
      "num_tokens": 19509944.0,
      "step": 9395
    },
    {
      "entropy": 5.6449051856994625,
      "epoch": 0.8526850507982583,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004934670304879398,
      "loss": 5.4074,
      "mean_token_accuracy": 0.16593187153339387,
      "num_tokens": 19519014.0,
      "step": 9400
    },
    {
      "entropy": 5.57209005355835,
      "epoch": 0.8531386066763426,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004934592878196553,
      "loss": 5.4133,
      "mean_token_accuracy": 0.16572651267051697,
      "num_tokens": 19528586.0,
      "step": 9405
    },
    {
      "entropy": 5.6237119197845455,
      "epoch": 0.8535921625544267,
      "grad_norm": 2.171875,
      "learning_rate": 0.000493451540633577,
      "loss": 5.5419,
      "mean_token_accuracy": 0.16077084690332413,
      "num_tokens": 19539912.0,
      "step": 9410
    },
    {
      "entropy": 5.611822319030762,
      "epoch": 0.8540457184325109,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004934437889298651,
      "loss": 5.5116,
      "mean_token_accuracy": 0.15974553525447846,
      "num_tokens": 19549860.0,
      "step": 9415
    },
    {
      "entropy": 5.672833824157715,
      "epoch": 0.854499274310595,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004934360327086799,
      "loss": 5.4546,
      "mean_token_accuracy": 0.16649190634489058,
      "num_tokens": 19559446.0,
      "step": 9420
    },
    {
      "entropy": 5.714667272567749,
      "epoch": 0.8549528301886793,
      "grad_norm": 2.265625,
      "learning_rate": 0.000493428271970182,
      "loss": 5.499,
      "mean_token_accuracy": 0.17092570215463637,
      "num_tokens": 19570282.0,
      "step": 9425
    },
    {
      "entropy": 5.611190223693848,
      "epoch": 0.8554063860667634,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004934205067145318,
      "loss": 5.4067,
      "mean_token_accuracy": 0.16731165051460267,
      "num_tokens": 19581087.0,
      "step": 9430
    },
    {
      "entropy": 5.564599466323853,
      "epoch": 0.8558599419448476,
      "grad_norm": 2.0,
      "learning_rate": 0.0004934127369418898,
      "loss": 5.5082,
      "mean_token_accuracy": 0.16411774158477782,
      "num_tokens": 19592100.0,
      "step": 9435
    },
    {
      "entropy": 5.570101594924926,
      "epoch": 0.8563134978229318,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004934049626524168,
      "loss": 5.4782,
      "mean_token_accuracy": 0.16585781723260878,
      "num_tokens": 19602557.0,
      "step": 9440
    },
    {
      "entropy": 5.676563405990601,
      "epoch": 0.8567670537010159,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004933971838462734,
      "loss": 5.5149,
      "mean_token_accuracy": 0.1598167359828949,
      "num_tokens": 19613643.0,
      "step": 9445
    },
    {
      "entropy": 5.647255325317383,
      "epoch": 0.8572206095791002,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004933894005236207,
      "loss": 5.4243,
      "mean_token_accuracy": 0.16713989078998565,
      "num_tokens": 19624180.0,
      "step": 9450
    },
    {
      "entropy": 5.606472158432007,
      "epoch": 0.8576741654571843,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004933816126846195,
      "loss": 5.4553,
      "mean_token_accuracy": 0.16742375642061233,
      "num_tokens": 19634799.0,
      "step": 9455
    },
    {
      "entropy": 5.561668920516968,
      "epoch": 0.8581277213352685,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004933738203294309,
      "loss": 5.4312,
      "mean_token_accuracy": 0.1714523419737816,
      "num_tokens": 19644618.0,
      "step": 9460
    },
    {
      "entropy": 5.56659164428711,
      "epoch": 0.8585812772133526,
      "grad_norm": 2.78125,
      "learning_rate": 0.000493366023458216,
      "loss": 5.4427,
      "mean_token_accuracy": 0.1622690811753273,
      "num_tokens": 19655550.0,
      "step": 9465
    },
    {
      "entropy": 5.621352195739746,
      "epoch": 0.8590348330914369,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004933582220711362,
      "loss": 5.494,
      "mean_token_accuracy": 0.17273536175489426,
      "num_tokens": 19665734.0,
      "step": 9470
    },
    {
      "entropy": 5.668038558959961,
      "epoch": 0.8594883889695211,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004933504161683526,
      "loss": 5.4391,
      "mean_token_accuracy": 0.16332752257585526,
      "num_tokens": 19677130.0,
      "step": 9475
    },
    {
      "entropy": 5.57708215713501,
      "epoch": 0.8599419448476052,
      "grad_norm": 2.953125,
      "learning_rate": 0.0004933426057500269,
      "loss": 5.3749,
      "mean_token_accuracy": 0.17379804104566574,
      "num_tokens": 19687374.0,
      "step": 9480
    },
    {
      "entropy": 5.641341924667358,
      "epoch": 0.8603955007256894,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004933347908163203,
      "loss": 5.4666,
      "mean_token_accuracy": 0.16353899091482163,
      "num_tokens": 19698182.0,
      "step": 9485
    },
    {
      "entropy": 5.643030023574829,
      "epoch": 0.8608490566037735,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004933269713673947,
      "loss": 5.4334,
      "mean_token_accuracy": 0.16726512908935548,
      "num_tokens": 19707664.0,
      "step": 9490
    },
    {
      "entropy": 5.599136781692505,
      "epoch": 0.8613026124818578,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004933191474034116,
      "loss": 5.4788,
      "mean_token_accuracy": 0.1638665184378624,
      "num_tokens": 19717647.0,
      "step": 9495
    },
    {
      "entropy": 5.622937393188477,
      "epoch": 0.861756168359942,
      "grad_norm": 1.546875,
      "learning_rate": 0.000493311318924533,
      "loss": 5.5182,
      "mean_token_accuracy": 0.1666355237364769,
      "num_tokens": 19727444.0,
      "step": 9500
    },
    {
      "entropy": 5.624603986740112,
      "epoch": 0.8622097242380261,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004933034859309208,
      "loss": 5.535,
      "mean_token_accuracy": 0.15709710568189622,
      "num_tokens": 19738719.0,
      "step": 9505
    },
    {
      "entropy": 5.7038360118865965,
      "epoch": 0.8626632801161103,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004932956484227367,
      "loss": 5.4874,
      "mean_token_accuracy": 0.16988746225833892,
      "num_tokens": 19749786.0,
      "step": 9510
    },
    {
      "entropy": 5.607921314239502,
      "epoch": 0.8631168359941945,
      "grad_norm": 3.3125,
      "learning_rate": 0.000493287806400143,
      "loss": 5.4634,
      "mean_token_accuracy": 0.17672932147979736,
      "num_tokens": 19759623.0,
      "step": 9515
    },
    {
      "entropy": 5.638581037521362,
      "epoch": 0.8635703918722787,
      "grad_norm": 2.0,
      "learning_rate": 0.0004932799598633018,
      "loss": 5.446,
      "mean_token_accuracy": 0.16770731210708617,
      "num_tokens": 19770117.0,
      "step": 9520
    },
    {
      "entropy": 5.627360916137695,
      "epoch": 0.8640239477503628,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004932721088123753,
      "loss": 5.5331,
      "mean_token_accuracy": 0.1628728300333023,
      "num_tokens": 19780748.0,
      "step": 9525
    },
    {
      "entropy": 5.646074390411377,
      "epoch": 0.864477503628447,
      "grad_norm": 2.25,
      "learning_rate": 0.0004932642532475262,
      "loss": 5.5089,
      "mean_token_accuracy": 0.16957747638225557,
      "num_tokens": 19791816.0,
      "step": 9530
    },
    {
      "entropy": 5.695994901657104,
      "epoch": 0.8649310595065312,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004932563931689168,
      "loss": 5.5442,
      "mean_token_accuracy": 0.16334045678377151,
      "num_tokens": 19801096.0,
      "step": 9535
    },
    {
      "entropy": 5.635082721710205,
      "epoch": 0.8653846153846154,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004932485285767093,
      "loss": 5.4149,
      "mean_token_accuracy": 0.16868644654750825,
      "num_tokens": 19811258.0,
      "step": 9540
    },
    {
      "entropy": 5.658902931213379,
      "epoch": 0.8658381712626996,
      "grad_norm": 3.453125,
      "learning_rate": 0.0004932406594710667,
      "loss": 5.4879,
      "mean_token_accuracy": 0.1752654105424881,
      "num_tokens": 19821765.0,
      "step": 9545
    },
    {
      "entropy": 5.6280457973480225,
      "epoch": 0.8662917271407837,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004932327858521517,
      "loss": 5.3739,
      "mean_token_accuracy": 0.17672177404165268,
      "num_tokens": 19831537.0,
      "step": 9550
    },
    {
      "entropy": 5.535717916488648,
      "epoch": 0.8667452830188679,
      "grad_norm": 2.828125,
      "learning_rate": 0.000493224907720127,
      "loss": 5.4003,
      "mean_token_accuracy": 0.16636214405298233,
      "num_tokens": 19841830.0,
      "step": 9555
    },
    {
      "entropy": 5.635063886642456,
      "epoch": 0.8671988388969522,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004932170250751556,
      "loss": 5.4935,
      "mean_token_accuracy": 0.15869371294975282,
      "num_tokens": 19851896.0,
      "step": 9560
    },
    {
      "entropy": 5.665247392654419,
      "epoch": 0.8676523947750363,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004932091379174006,
      "loss": 5.5211,
      "mean_token_accuracy": 0.15944714993238449,
      "num_tokens": 19861735.0,
      "step": 9565
    },
    {
      "entropy": 5.637991523742675,
      "epoch": 0.8681059506531205,
      "grad_norm": 2.078125,
      "learning_rate": 0.000493201246247025,
      "loss": 5.4278,
      "mean_token_accuracy": 0.17442315816879272,
      "num_tokens": 19872333.0,
      "step": 9570
    },
    {
      "entropy": 5.607363271713257,
      "epoch": 0.8685595065312046,
      "grad_norm": 2.765625,
      "learning_rate": 0.000493193350064192,
      "loss": 5.4307,
      "mean_token_accuracy": 0.16355780214071275,
      "num_tokens": 19882896.0,
      "step": 9575
    },
    {
      "entropy": 5.6081709384918215,
      "epoch": 0.8690130624092888,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004931854493690649,
      "loss": 5.3899,
      "mean_token_accuracy": 0.16593107730150222,
      "num_tokens": 19893767.0,
      "step": 9580
    },
    {
      "entropy": 5.592438220977783,
      "epoch": 0.869466618287373,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004931775441618071,
      "loss": 5.3853,
      "mean_token_accuracy": 0.17099117934703828,
      "num_tokens": 19904242.0,
      "step": 9585
    },
    {
      "entropy": 5.536753511428833,
      "epoch": 0.8699201741654572,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004931696344425821,
      "loss": 5.5062,
      "mean_token_accuracy": 0.16561063528060913,
      "num_tokens": 19913898.0,
      "step": 9590
    },
    {
      "entropy": 5.674816989898682,
      "epoch": 0.8703737300435413,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004931617202115535,
      "loss": 5.4887,
      "mean_token_accuracy": 0.1691025123000145,
      "num_tokens": 19923756.0,
      "step": 9595
    },
    {
      "entropy": 5.6317566394805905,
      "epoch": 0.8708272859216255,
      "grad_norm": 2.75,
      "learning_rate": 0.0004931538014688849,
      "loss": 5.4302,
      "mean_token_accuracy": 0.17173125892877578,
      "num_tokens": 19932720.0,
      "step": 9600
    },
    {
      "entropy": 5.670051956176758,
      "epoch": 0.8712808417997098,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004931458782147401,
      "loss": 5.6214,
      "mean_token_accuracy": 0.15633623227477073,
      "num_tokens": 19942962.0,
      "step": 9605
    },
    {
      "entropy": 5.6734381198883055,
      "epoch": 0.8717343976777939,
      "grad_norm": 5.28125,
      "learning_rate": 0.000493137950449283,
      "loss": 5.5007,
      "mean_token_accuracy": 0.16917594224214555,
      "num_tokens": 19953735.0,
      "step": 9610
    },
    {
      "entropy": 5.670583915710449,
      "epoch": 0.8721879535558781,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004931300181726775,
      "loss": 5.4712,
      "mean_token_accuracy": 0.1696719706058502,
      "num_tokens": 19964416.0,
      "step": 9615
    },
    {
      "entropy": 5.657572889328003,
      "epoch": 0.8726415094339622,
      "grad_norm": 2.734375,
      "learning_rate": 0.0004931220813850875,
      "loss": 5.4892,
      "mean_token_accuracy": 0.16641833484172822,
      "num_tokens": 19974531.0,
      "step": 9620
    },
    {
      "entropy": 5.650475597381591,
      "epoch": 0.8730950653120464,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004931141400866773,
      "loss": 5.4961,
      "mean_token_accuracy": 0.17093998789787293,
      "num_tokens": 19984964.0,
      "step": 9625
    },
    {
      "entropy": 5.606709909439087,
      "epoch": 0.8735486211901307,
      "grad_norm": 3.1875,
      "learning_rate": 0.0004931061942776112,
      "loss": 5.4341,
      "mean_token_accuracy": 0.1740843877196312,
      "num_tokens": 19995568.0,
      "step": 9630
    },
    {
      "entropy": 5.735747480392456,
      "epoch": 0.8740021770682148,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004930982439580534,
      "loss": 5.604,
      "mean_token_accuracy": 0.15779380351305008,
      "num_tokens": 20005856.0,
      "step": 9635
    },
    {
      "entropy": 5.714074563980103,
      "epoch": 0.874455732946299,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004930902891281683,
      "loss": 5.5229,
      "mean_token_accuracy": 0.16212179958820344,
      "num_tokens": 20015683.0,
      "step": 9640
    },
    {
      "entropy": 5.566607093811035,
      "epoch": 0.8749092888243831,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004930823297881205,
      "loss": 5.3496,
      "mean_token_accuracy": 0.17044655978679657,
      "num_tokens": 20025998.0,
      "step": 9645
    },
    {
      "entropy": 5.638138914108277,
      "epoch": 0.8753628447024674,
      "grad_norm": 2.5,
      "learning_rate": 0.0004930743659380745,
      "loss": 5.5097,
      "mean_token_accuracy": 0.1626819744706154,
      "num_tokens": 20036486.0,
      "step": 9650
    },
    {
      "entropy": 5.618236541748047,
      "epoch": 0.8758164005805515,
      "grad_norm": 1.8125,
      "learning_rate": 0.000493066397578195,
      "loss": 5.4014,
      "mean_token_accuracy": 0.17181899398565292,
      "num_tokens": 20046784.0,
      "step": 9655
    },
    {
      "entropy": 5.622888803482056,
      "epoch": 0.8762699564586357,
      "grad_norm": 2.25,
      "learning_rate": 0.0004930584247086468,
      "loss": 5.4641,
      "mean_token_accuracy": 0.1695317029953003,
      "num_tokens": 20056999.0,
      "step": 9660
    },
    {
      "entropy": 5.560177183151245,
      "epoch": 0.8767235123367199,
      "grad_norm": 2.125,
      "learning_rate": 0.000493050447329595,
      "loss": 5.3476,
      "mean_token_accuracy": 0.1711304858326912,
      "num_tokens": 20067612.0,
      "step": 9665
    },
    {
      "entropy": 5.6600298404693605,
      "epoch": 0.877177068214804,
      "grad_norm": 2.953125,
      "learning_rate": 0.0004930424654412043,
      "loss": 5.5128,
      "mean_token_accuracy": 0.1657562106847763,
      "num_tokens": 20077330.0,
      "step": 9670
    },
    {
      "entropy": 5.663735485076904,
      "epoch": 0.8776306240928883,
      "grad_norm": 4.09375,
      "learning_rate": 0.0004930344790436398,
      "loss": 5.4952,
      "mean_token_accuracy": 0.1622760236263275,
      "num_tokens": 20086801.0,
      "step": 9675
    },
    {
      "entropy": 5.621307373046875,
      "epoch": 0.8780841799709724,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004930264881370666,
      "loss": 5.4479,
      "mean_token_accuracy": 0.16382185965776444,
      "num_tokens": 20097047.0,
      "step": 9680
    },
    {
      "entropy": 5.62794976234436,
      "epoch": 0.8785377358490566,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004930184927216504,
      "loss": 5.4856,
      "mean_token_accuracy": 0.1664504364132881,
      "num_tokens": 20107185.0,
      "step": 9685
    },
    {
      "entropy": 5.686764240264893,
      "epoch": 0.8789912917271407,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004930104927975559,
      "loss": 5.43,
      "mean_token_accuracy": 0.16731444001197815,
      "num_tokens": 20118493.0,
      "step": 9690
    },
    {
      "entropy": 5.630422067642212,
      "epoch": 0.879444847605225,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004930024883649489,
      "loss": 5.4647,
      "mean_token_accuracy": 0.1667337730526924,
      "num_tokens": 20129294.0,
      "step": 9695
    },
    {
      "entropy": 5.5502057552337645,
      "epoch": 0.8798984034833092,
      "grad_norm": 2.3125,
      "learning_rate": 0.000492994479423995,
      "loss": 5.4366,
      "mean_token_accuracy": 0.17128972113132476,
      "num_tokens": 20139194.0,
      "step": 9700
    },
    {
      "entropy": 5.601175928115845,
      "epoch": 0.8803519593613933,
      "grad_norm": 3.96875,
      "learning_rate": 0.0004929864659748597,
      "loss": 5.47,
      "mean_token_accuracy": 0.1654967784881592,
      "num_tokens": 20149681.0,
      "step": 9705
    },
    {
      "entropy": 5.680785179138184,
      "epoch": 0.8808055152394775,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004929784480177087,
      "loss": 5.5689,
      "mean_token_accuracy": 0.1593937858939171,
      "num_tokens": 20160557.0,
      "step": 9710
    },
    {
      "entropy": 5.672782897949219,
      "epoch": 0.8812590711175616,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004929704255527078,
      "loss": 5.446,
      "mean_token_accuracy": 0.16927874982357025,
      "num_tokens": 20171047.0,
      "step": 9715
    },
    {
      "entropy": 5.688721179962158,
      "epoch": 0.8817126269956459,
      "grad_norm": 2.21875,
      "learning_rate": 0.000492962398580023,
      "loss": 5.5869,
      "mean_token_accuracy": 0.1585509419441223,
      "num_tokens": 20182148.0,
      "step": 9720
    },
    {
      "entropy": 5.5953943729400635,
      "epoch": 0.88216618287373,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004929543670998203,
      "loss": 5.3564,
      "mean_token_accuracy": 0.1713259622454643,
      "num_tokens": 20192131.0,
      "step": 9725
    },
    {
      "entropy": 5.6180421829223635,
      "epoch": 0.8826197387518142,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004929463311122657,
      "loss": 5.4797,
      "mean_token_accuracy": 0.16329212933778764,
      "num_tokens": 20202805.0,
      "step": 9730
    },
    {
      "entropy": 5.596766090393066,
      "epoch": 0.8830732946298984,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004929382906175253,
      "loss": 5.4089,
      "mean_token_accuracy": 0.17018195986747742,
      "num_tokens": 20212702.0,
      "step": 9735
    },
    {
      "entropy": 5.612695646286011,
      "epoch": 0.8835268505079826,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004929302456157657,
      "loss": 5.4377,
      "mean_token_accuracy": 0.16651211231946944,
      "num_tokens": 20222661.0,
      "step": 9740
    },
    {
      "entropy": 5.642892456054687,
      "epoch": 0.8839804063860668,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004929221961071531,
      "loss": 5.4873,
      "mean_token_accuracy": 0.16402020901441575,
      "num_tokens": 20233012.0,
      "step": 9745
    },
    {
      "entropy": 5.647194147109985,
      "epoch": 0.8844339622641509,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004929141420918538,
      "loss": 5.4908,
      "mean_token_accuracy": 0.16944866329431535,
      "num_tokens": 20242918.0,
      "step": 9750
    },
    {
      "entropy": 5.610573482513428,
      "epoch": 0.8848875181422351,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004929060835700347,
      "loss": 5.4291,
      "mean_token_accuracy": 0.17089887857437133,
      "num_tokens": 20253254.0,
      "step": 9755
    },
    {
      "entropy": 5.621883535385132,
      "epoch": 0.8853410740203193,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000492898020541862,
      "loss": 5.4811,
      "mean_token_accuracy": 0.16487018764019012,
      "num_tokens": 20263355.0,
      "step": 9760
    },
    {
      "entropy": 5.606476926803589,
      "epoch": 0.8857946298984035,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004928899530075029,
      "loss": 5.4199,
      "mean_token_accuracy": 0.16572795957326888,
      "num_tokens": 20274716.0,
      "step": 9765
    },
    {
      "entropy": 5.545453405380249,
      "epoch": 0.8862481857764877,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004928818809671241,
      "loss": 5.399,
      "mean_token_accuracy": 0.16600928008556365,
      "num_tokens": 20285631.0,
      "step": 9770
    },
    {
      "entropy": 5.629395246505737,
      "epoch": 0.8867017416545718,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004928738044208923,
      "loss": 5.4646,
      "mean_token_accuracy": 0.1744563788175583,
      "num_tokens": 20295717.0,
      "step": 9775
    },
    {
      "entropy": 5.5827960014343265,
      "epoch": 0.887155297532656,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004928657233689748,
      "loss": 5.437,
      "mean_token_accuracy": 0.1701323866844177,
      "num_tokens": 20306637.0,
      "step": 9780
    },
    {
      "entropy": 5.611489343643188,
      "epoch": 0.8876088534107403,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004928576378115387,
      "loss": 5.4,
      "mean_token_accuracy": 0.17060102224349977,
      "num_tokens": 20316629.0,
      "step": 9785
    },
    {
      "entropy": 5.7080353736877445,
      "epoch": 0.8880624092888244,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004928495477487509,
      "loss": 5.597,
      "mean_token_accuracy": 0.1562129482626915,
      "num_tokens": 20327935.0,
      "step": 9790
    },
    {
      "entropy": 5.635652685165406,
      "epoch": 0.8885159651669086,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004928414531807792,
      "loss": 5.4061,
      "mean_token_accuracy": 0.16605042219161986,
      "num_tokens": 20339142.0,
      "step": 9795
    },
    {
      "entropy": 5.605915451049805,
      "epoch": 0.8889695210449927,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004928333541077906,
      "loss": 5.4254,
      "mean_token_accuracy": 0.17367625385522842,
      "num_tokens": 20348474.0,
      "step": 9800
    },
    {
      "entropy": 5.644237422943116,
      "epoch": 0.8894230769230769,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004928252505299527,
      "loss": 5.4722,
      "mean_token_accuracy": 0.1646744877099991,
      "num_tokens": 20359440.0,
      "step": 9805
    },
    {
      "entropy": 5.6727293014526365,
      "epoch": 0.8898766328011611,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004928171424474332,
      "loss": 5.5075,
      "mean_token_accuracy": 0.16782969534397124,
      "num_tokens": 20370355.0,
      "step": 9810
    },
    {
      "entropy": 5.615191698074341,
      "epoch": 0.8903301886792453,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004928090298603996,
      "loss": 5.4158,
      "mean_token_accuracy": 0.17068232148885726,
      "num_tokens": 20380214.0,
      "step": 9815
    },
    {
      "entropy": 5.6379311084747314,
      "epoch": 0.8907837445573294,
      "grad_norm": 2.90625,
      "learning_rate": 0.0004928009127690199,
      "loss": 5.431,
      "mean_token_accuracy": 0.168188439309597,
      "num_tokens": 20389190.0,
      "step": 9820
    },
    {
      "entropy": 5.605611085891724,
      "epoch": 0.8912373004354136,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004927927911734617,
      "loss": 5.4853,
      "mean_token_accuracy": 0.16537884026765823,
      "num_tokens": 20399531.0,
      "step": 9825
    },
    {
      "entropy": 5.658443212509155,
      "epoch": 0.8916908563134979,
      "grad_norm": 3.984375,
      "learning_rate": 0.0004927846650738931,
      "loss": 5.4623,
      "mean_token_accuracy": 0.1641345277428627,
      "num_tokens": 20410201.0,
      "step": 9830
    },
    {
      "entropy": 5.617740869522095,
      "epoch": 0.892144412191582,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004927765344704821,
      "loss": 5.4169,
      "mean_token_accuracy": 0.17194532752037048,
      "num_tokens": 20421579.0,
      "step": 9835
    },
    {
      "entropy": 5.618694734573364,
      "epoch": 0.8925979680696662,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004927683993633969,
      "loss": 5.4892,
      "mean_token_accuracy": 0.1712222546339035,
      "num_tokens": 20432048.0,
      "step": 9840
    },
    {
      "entropy": 5.697521066665649,
      "epoch": 0.8930515239477503,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004927602597528057,
      "loss": 5.5362,
      "mean_token_accuracy": 0.16094869524240493,
      "num_tokens": 20441752.0,
      "step": 9845
    },
    {
      "entropy": 5.665847206115723,
      "epoch": 0.8935050798258345,
      "grad_norm": 4.125,
      "learning_rate": 0.0004927521156388769,
      "loss": 5.3811,
      "mean_token_accuracy": 0.1692692369222641,
      "num_tokens": 20451693.0,
      "step": 9850
    },
    {
      "entropy": 5.636432886123657,
      "epoch": 0.8939586357039188,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004927439670217788,
      "loss": 5.4927,
      "mean_token_accuracy": 0.15978458672761917,
      "num_tokens": 20462312.0,
      "step": 9855
    },
    {
      "entropy": 5.613775014877319,
      "epoch": 0.8944121915820029,
      "grad_norm": 2.546875,
      "learning_rate": 0.00049273581390168,
      "loss": 5.4099,
      "mean_token_accuracy": 0.1645998850464821,
      "num_tokens": 20472546.0,
      "step": 9860
    },
    {
      "entropy": 5.650700902938842,
      "epoch": 0.8948657474600871,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000492727656278749,
      "loss": 5.4614,
      "mean_token_accuracy": 0.16577322036027908,
      "num_tokens": 20483266.0,
      "step": 9865
    },
    {
      "entropy": 5.587610387802124,
      "epoch": 0.8953193033381712,
      "grad_norm": 4.6875,
      "learning_rate": 0.0004927194941531547,
      "loss": 5.4024,
      "mean_token_accuracy": 0.168634133040905,
      "num_tokens": 20492490.0,
      "step": 9870
    },
    {
      "entropy": 5.582613468170166,
      "epoch": 0.8957728592162555,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004927113275250657,
      "loss": 5.4107,
      "mean_token_accuracy": 0.17028550952672958,
      "num_tokens": 20503077.0,
      "step": 9875
    },
    {
      "entropy": 5.590944814682007,
      "epoch": 0.8962264150943396,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004927031563946511,
      "loss": 5.4557,
      "mean_token_accuracy": 0.16236511170864104,
      "num_tokens": 20513771.0,
      "step": 9880
    },
    {
      "entropy": 5.610186958312989,
      "epoch": 0.8966799709724238,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004926949807620796,
      "loss": 5.5565,
      "mean_token_accuracy": 0.16225252002477647,
      "num_tokens": 20523616.0,
      "step": 9885
    },
    {
      "entropy": 5.625984811782837,
      "epoch": 0.897133526850508,
      "grad_norm": 3.734375,
      "learning_rate": 0.0004926868006275204,
      "loss": 5.3434,
      "mean_token_accuracy": 0.1741590067744255,
      "num_tokens": 20533613.0,
      "step": 9890
    },
    {
      "entropy": 5.634837245941162,
      "epoch": 0.8975870827285921,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004926786159911427,
      "loss": 5.436,
      "mean_token_accuracy": 0.16801166236400605,
      "num_tokens": 20543812.0,
      "step": 9895
    },
    {
      "entropy": 5.569585657119751,
      "epoch": 0.8980406386066764,
      "grad_norm": 2.21875,
      "learning_rate": 0.000492670426853116,
      "loss": 5.3903,
      "mean_token_accuracy": 0.16847120970487595,
      "num_tokens": 20553785.0,
      "step": 9900
    },
    {
      "entropy": 5.655517578125,
      "epoch": 0.8984941944847605,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004926622332136092,
      "loss": 5.4781,
      "mean_token_accuracy": 0.1724136456847191,
      "num_tokens": 20563326.0,
      "step": 9905
    },
    {
      "entropy": 5.577728033065796,
      "epoch": 0.8989477503628447,
      "grad_norm": 4.375,
      "learning_rate": 0.0004926540350727918,
      "loss": 5.3547,
      "mean_token_accuracy": 0.173677958548069,
      "num_tokens": 20573643.0,
      "step": 9910
    },
    {
      "entropy": 5.55289888381958,
      "epoch": 0.8994013062409288,
      "grad_norm": 2.375,
      "learning_rate": 0.0004926458324308337,
      "loss": 5.3721,
      "mean_token_accuracy": 0.17063824236392974,
      "num_tokens": 20583916.0,
      "step": 9915
    },
    {
      "entropy": 5.659201669692993,
      "epoch": 0.8998548621190131,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004926376252879043,
      "loss": 5.5467,
      "mean_token_accuracy": 0.16558234691619872,
      "num_tokens": 20593984.0,
      "step": 9920
    },
    {
      "entropy": 5.634729337692261,
      "epoch": 0.9003084179970973,
      "grad_norm": 2.75,
      "learning_rate": 0.0004926294136441733,
      "loss": 5.3592,
      "mean_token_accuracy": 0.17320367842912673,
      "num_tokens": 20603927.0,
      "step": 9925
    },
    {
      "entropy": 5.62198920249939,
      "epoch": 0.9007619738751814,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004926211974998106,
      "loss": 5.4945,
      "mean_token_accuracy": 0.170858134329319,
      "num_tokens": 20614147.0,
      "step": 9930
    },
    {
      "entropy": 5.669112634658814,
      "epoch": 0.9012155297532656,
      "grad_norm": 2.265625,
      "learning_rate": 0.000492612976854986,
      "loss": 5.4631,
      "mean_token_accuracy": 0.16913086771965027,
      "num_tokens": 20623631.0,
      "step": 9935
    },
    {
      "entropy": 5.569746732711792,
      "epoch": 0.9016690856313497,
      "grad_norm": 2.0,
      "learning_rate": 0.0004926047517098697,
      "loss": 5.3488,
      "mean_token_accuracy": 0.17397348433732987,
      "num_tokens": 20632566.0,
      "step": 9940
    },
    {
      "entropy": 5.617513370513916,
      "epoch": 0.902122641509434,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004925965220646315,
      "loss": 5.4507,
      "mean_token_accuracy": 0.16440732777118683,
      "num_tokens": 20643246.0,
      "step": 9945
    },
    {
      "entropy": 5.643515872955322,
      "epoch": 0.9025761973875182,
      "grad_norm": 2.0,
      "learning_rate": 0.0004925882879194419,
      "loss": 5.5369,
      "mean_token_accuracy": 0.16406310051679612,
      "num_tokens": 20653372.0,
      "step": 9950
    },
    {
      "entropy": 5.657034015655517,
      "epoch": 0.9030297532656023,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004925800492744712,
      "loss": 5.4803,
      "mean_token_accuracy": 0.16646226197481157,
      "num_tokens": 20662687.0,
      "step": 9955
    },
    {
      "entropy": 5.6161580085754395,
      "epoch": 0.9034833091436865,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004925718061298895,
      "loss": 5.3914,
      "mean_token_accuracy": 0.17280880659818648,
      "num_tokens": 20673168.0,
      "step": 9960
    },
    {
      "entropy": 5.7037781238555905,
      "epoch": 0.9039368650217707,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004925635584858675,
      "loss": 5.5148,
      "mean_token_accuracy": 0.16221273839473724,
      "num_tokens": 20683100.0,
      "step": 9965
    },
    {
      "entropy": 5.718763256072998,
      "epoch": 0.9043904208998549,
      "grad_norm": 2.734375,
      "learning_rate": 0.0004925553063425755,
      "loss": 5.5259,
      "mean_token_accuracy": 0.16453297883272172,
      "num_tokens": 20694097.0,
      "step": 9970
    },
    {
      "entropy": 5.639046764373779,
      "epoch": 0.904843976777939,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004925470497001844,
      "loss": 5.4937,
      "mean_token_accuracy": 0.16766450703144073,
      "num_tokens": 20704640.0,
      "step": 9975
    },
    {
      "entropy": 5.629187488555909,
      "epoch": 0.9052975326560232,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004925387885588648,
      "loss": 5.4272,
      "mean_token_accuracy": 0.16981081664562225,
      "num_tokens": 20714419.0,
      "step": 9980
    },
    {
      "entropy": 5.687669944763184,
      "epoch": 0.9057510885341074,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004925305229187877,
      "loss": 5.5088,
      "mean_token_accuracy": 0.1662205457687378,
      "num_tokens": 20724705.0,
      "step": 9985
    },
    {
      "entropy": 5.703696823120117,
      "epoch": 0.9062046444121916,
      "grad_norm": 1.953125,
      "learning_rate": 0.000492522252780124,
      "loss": 5.5996,
      "mean_token_accuracy": 0.16465610563755034,
      "num_tokens": 20734753.0,
      "step": 9990
    },
    {
      "entropy": 5.616272354125977,
      "epoch": 0.9066582002902758,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004925139781430446,
      "loss": 5.3871,
      "mean_token_accuracy": 0.1729523003101349,
      "num_tokens": 20745389.0,
      "step": 9995
    },
    {
      "entropy": 5.631888151168823,
      "epoch": 0.9071117561683599,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004925056990077207,
      "loss": 5.4276,
      "mean_token_accuracy": 0.1671442434191704,
      "num_tokens": 20755702.0,
      "step": 10000
    },
    {
      "entropy": 5.613785123825073,
      "epoch": 0.9075653120464441,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004924974153743235,
      "loss": 5.3549,
      "mean_token_accuracy": 0.174472938477993,
      "num_tokens": 20766046.0,
      "step": 10005
    },
    {
      "entropy": 5.628176593780518,
      "epoch": 0.9080188679245284,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004924891272430244,
      "loss": 5.4566,
      "mean_token_accuracy": 0.16712419539690018,
      "num_tokens": 20776654.0,
      "step": 10010
    },
    {
      "entropy": 5.670589637756348,
      "epoch": 0.9084724238026125,
      "grad_norm": 2.921875,
      "learning_rate": 0.0004924808346139948,
      "loss": 5.5143,
      "mean_token_accuracy": 0.17070617973804475,
      "num_tokens": 20787066.0,
      "step": 10015
    },
    {
      "entropy": 5.666522789001465,
      "epoch": 0.9089259796806967,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004924725374874059,
      "loss": 5.411,
      "mean_token_accuracy": 0.1702822580933571,
      "num_tokens": 20796822.0,
      "step": 10020
    },
    {
      "entropy": 5.683545827865601,
      "epoch": 0.9093795355587808,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004924642358634297,
      "loss": 5.4697,
      "mean_token_accuracy": 0.16714202910661696,
      "num_tokens": 20807602.0,
      "step": 10025
    },
    {
      "entropy": 5.644869184494018,
      "epoch": 0.909833091436865,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004924559297422376,
      "loss": 5.4424,
      "mean_token_accuracy": 0.16788566261529922,
      "num_tokens": 20819046.0,
      "step": 10030
    },
    {
      "entropy": 5.640174627304077,
      "epoch": 0.9102866473149492,
      "grad_norm": 3.140625,
      "learning_rate": 0.0004924476191240014,
      "loss": 5.4676,
      "mean_token_accuracy": 0.164874167740345,
      "num_tokens": 20828989.0,
      "step": 10035
    },
    {
      "entropy": 5.627811527252197,
      "epoch": 0.9107402031930334,
      "grad_norm": 2.90625,
      "learning_rate": 0.000492439304008893,
      "loss": 5.4784,
      "mean_token_accuracy": 0.16124556064605713,
      "num_tokens": 20839366.0,
      "step": 10040
    },
    {
      "entropy": 5.675979280471802,
      "epoch": 0.9111937590711175,
      "grad_norm": 4.28125,
      "learning_rate": 0.0004924309843970847,
      "loss": 5.4463,
      "mean_token_accuracy": 0.1698861002922058,
      "num_tokens": 20849513.0,
      "step": 10045
    },
    {
      "entropy": 5.709691905975342,
      "epoch": 0.9116473149492017,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004924226602887479,
      "loss": 5.5092,
      "mean_token_accuracy": 0.16104661524295807,
      "num_tokens": 20861406.0,
      "step": 10050
    },
    {
      "entropy": 5.598871183395386,
      "epoch": 0.912100870827286,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004924143316840552,
      "loss": 5.4258,
      "mean_token_accuracy": 0.1793083667755127,
      "num_tokens": 20871190.0,
      "step": 10055
    },
    {
      "entropy": 5.660051965713501,
      "epoch": 0.9125544267053701,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004924059985831787,
      "loss": 5.551,
      "mean_token_accuracy": 0.16229840219020844,
      "num_tokens": 20881589.0,
      "step": 10060
    },
    {
      "entropy": 5.66608419418335,
      "epoch": 0.9130079825834543,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004923976609862908,
      "loss": 5.4908,
      "mean_token_accuracy": 0.1605662927031517,
      "num_tokens": 20892072.0,
      "step": 10065
    },
    {
      "entropy": 5.685091018676758,
      "epoch": 0.9134615384615384,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004923893188935638,
      "loss": 5.4763,
      "mean_token_accuracy": 0.16243244409561158,
      "num_tokens": 20902616.0,
      "step": 10070
    },
    {
      "entropy": 5.653387975692749,
      "epoch": 0.9139150943396226,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004923809723051702,
      "loss": 5.4772,
      "mean_token_accuracy": 0.16339652985334396,
      "num_tokens": 20913832.0,
      "step": 10075
    },
    {
      "entropy": 5.518843936920166,
      "epoch": 0.9143686502177069,
      "grad_norm": 2.25,
      "learning_rate": 0.0004923726212212829,
      "loss": 5.3731,
      "mean_token_accuracy": 0.16908359676599502,
      "num_tokens": 20923322.0,
      "step": 10080
    },
    {
      "entropy": 5.597713804244995,
      "epoch": 0.914822206095791,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004923642656420742,
      "loss": 5.474,
      "mean_token_accuracy": 0.16452524811029434,
      "num_tokens": 20933334.0,
      "step": 10085
    },
    {
      "entropy": 5.64165244102478,
      "epoch": 0.9152757619738752,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004923559055677173,
      "loss": 5.4313,
      "mean_token_accuracy": 0.1654943734407425,
      "num_tokens": 20944171.0,
      "step": 10090
    },
    {
      "entropy": 5.658243036270141,
      "epoch": 0.9157293178519593,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004923475409983848,
      "loss": 5.4761,
      "mean_token_accuracy": 0.16481802463531495,
      "num_tokens": 20954200.0,
      "step": 10095
    },
    {
      "entropy": 5.592943096160889,
      "epoch": 0.9161828737300436,
      "grad_norm": 3.046875,
      "learning_rate": 0.0004923391719342497,
      "loss": 5.4813,
      "mean_token_accuracy": 0.16222574710845947,
      "num_tokens": 20964935.0,
      "step": 10100
    },
    {
      "entropy": 5.679191446304321,
      "epoch": 0.9166364296081277,
      "grad_norm": 4.8125,
      "learning_rate": 0.0004923307983754852,
      "loss": 5.4798,
      "mean_token_accuracy": 0.16879493445158006,
      "num_tokens": 20975499.0,
      "step": 10105
    },
    {
      "entropy": 5.59067120552063,
      "epoch": 0.9170899854862119,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004923224203222645,
      "loss": 5.4001,
      "mean_token_accuracy": 0.16911139339208603,
      "num_tokens": 20986199.0,
      "step": 10110
    },
    {
      "entropy": 5.612363624572754,
      "epoch": 0.9175435413642961,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004923140377747606,
      "loss": 5.4742,
      "mean_token_accuracy": 0.16479670703411103,
      "num_tokens": 20996313.0,
      "step": 10115
    },
    {
      "entropy": 5.621811199188232,
      "epoch": 0.9179970972423802,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004923056507331471,
      "loss": 5.4475,
      "mean_token_accuracy": 0.16965973675251006,
      "num_tokens": 21006133.0,
      "step": 10120
    },
    {
      "entropy": 5.699572801589966,
      "epoch": 0.9184506531204645,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004922972591975974,
      "loss": 5.4407,
      "mean_token_accuracy": 0.16748667806386947,
      "num_tokens": 21016448.0,
      "step": 10125
    },
    {
      "entropy": 5.622897386550903,
      "epoch": 0.9189042089985486,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004922888631682849,
      "loss": 5.3551,
      "mean_token_accuracy": 0.16948991268873215,
      "num_tokens": 21026458.0,
      "step": 10130
    },
    {
      "entropy": 5.5447803974151615,
      "epoch": 0.9193577648766328,
      "grad_norm": 3.0625,
      "learning_rate": 0.0004922804626453834,
      "loss": 5.4068,
      "mean_token_accuracy": 0.1717316135764122,
      "num_tokens": 21037111.0,
      "step": 10135
    },
    {
      "entropy": 5.529819822311401,
      "epoch": 0.9198113207547169,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004922720576290665,
      "loss": 5.4657,
      "mean_token_accuracy": 0.1738602638244629,
      "num_tokens": 21048041.0,
      "step": 10140
    },
    {
      "entropy": 5.581639623641967,
      "epoch": 0.9202648766328012,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004922636481195082,
      "loss": 5.3495,
      "mean_token_accuracy": 0.18189457952976226,
      "num_tokens": 21058099.0,
      "step": 10145
    },
    {
      "entropy": 5.649187707901001,
      "epoch": 0.9207184325108854,
      "grad_norm": 2.828125,
      "learning_rate": 0.0004922552341168823,
      "loss": 5.4764,
      "mean_token_accuracy": 0.1608864963054657,
      "num_tokens": 21068137.0,
      "step": 10150
    },
    {
      "entropy": 5.555316972732544,
      "epoch": 0.9211719883889695,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004922468156213627,
      "loss": 5.3595,
      "mean_token_accuracy": 0.17540239095687865,
      "num_tokens": 21077556.0,
      "step": 10155
    },
    {
      "entropy": 5.5683581829071045,
      "epoch": 0.9216255442670537,
      "grad_norm": 3.4375,
      "learning_rate": 0.0004922383926331236,
      "loss": 5.3939,
      "mean_token_accuracy": 0.17052381485700607,
      "num_tokens": 21087524.0,
      "step": 10160
    },
    {
      "entropy": 5.541804695129395,
      "epoch": 0.9220791001451378,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004922299651523392,
      "loss": 5.4537,
      "mean_token_accuracy": 0.16598909497261047,
      "num_tokens": 21098446.0,
      "step": 10165
    },
    {
      "entropy": 5.590465116500854,
      "epoch": 0.9225326560232221,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004922215331791837,
      "loss": 5.4725,
      "mean_token_accuracy": 0.16691119074821473,
      "num_tokens": 21109291.0,
      "step": 10170
    },
    {
      "entropy": 5.647093009948731,
      "epoch": 0.9229862119013063,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004922130967138316,
      "loss": 5.3542,
      "mean_token_accuracy": 0.17282696664333344,
      "num_tokens": 21120043.0,
      "step": 10175
    },
    {
      "entropy": 5.6608076095581055,
      "epoch": 0.9234397677793904,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004922046557564574,
      "loss": 5.478,
      "mean_token_accuracy": 0.16004344522953035,
      "num_tokens": 21130339.0,
      "step": 10180
    },
    {
      "entropy": 5.618777227401734,
      "epoch": 0.9238933236574746,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004921962103072355,
      "loss": 5.5418,
      "mean_token_accuracy": 0.16347383558750153,
      "num_tokens": 21142307.0,
      "step": 10185
    },
    {
      "entropy": 5.5993763446807865,
      "epoch": 0.9243468795355588,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004921877603663406,
      "loss": 5.4336,
      "mean_token_accuracy": 0.1706625923514366,
      "num_tokens": 21151843.0,
      "step": 10190
    },
    {
      "entropy": 5.5873476505279545,
      "epoch": 0.924800435413643,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004921793059339474,
      "loss": 5.4312,
      "mean_token_accuracy": 0.16952639818191528,
      "num_tokens": 21162574.0,
      "step": 10195
    },
    {
      "entropy": 5.616939735412598,
      "epoch": 0.9252539912917271,
      "grad_norm": 3.421875,
      "learning_rate": 0.0004921708470102309,
      "loss": 5.3868,
      "mean_token_accuracy": 0.171491040289402,
      "num_tokens": 21172637.0,
      "step": 10200
    },
    {
      "entropy": 5.596718406677246,
      "epoch": 0.9257075471698113,
      "grad_norm": 2.0,
      "learning_rate": 0.000492162383595366,
      "loss": 5.3917,
      "mean_token_accuracy": 0.17129719704389573,
      "num_tokens": 21182272.0,
      "step": 10205
    },
    {
      "entropy": 5.637783479690552,
      "epoch": 0.9261611030478955,
      "grad_norm": 1.875,
      "learning_rate": 0.0004921539156895275,
      "loss": 5.5522,
      "mean_token_accuracy": 0.16203432977199556,
      "num_tokens": 21193712.0,
      "step": 10210
    },
    {
      "entropy": 5.6768382549285885,
      "epoch": 0.9266146589259797,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004921454432928908,
      "loss": 5.536,
      "mean_token_accuracy": 0.16012393087148666,
      "num_tokens": 21205104.0,
      "step": 10215
    },
    {
      "entropy": 5.695552110671997,
      "epoch": 0.9270682148040639,
      "grad_norm": 2.359375,
      "learning_rate": 0.000492136966405631,
      "loss": 5.4251,
      "mean_token_accuracy": 0.16568379253149032,
      "num_tokens": 21215495.0,
      "step": 10220
    },
    {
      "entropy": 5.629831695556641,
      "epoch": 0.927521770682148,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004921284850279233,
      "loss": 5.4298,
      "mean_token_accuracy": 0.17006988748908042,
      "num_tokens": 21226047.0,
      "step": 10225
    },
    {
      "entropy": 5.6132285594940186,
      "epoch": 0.9279753265602322,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004921199991599433,
      "loss": 5.4882,
      "mean_token_accuracy": 0.164862522482872,
      "num_tokens": 21235784.0,
      "step": 10230
    },
    {
      "entropy": 5.594307994842529,
      "epoch": 0.9284288824383164,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004921115088018665,
      "loss": 5.4503,
      "mean_token_accuracy": 0.16549347192049027,
      "num_tokens": 21245493.0,
      "step": 10235
    },
    {
      "entropy": 5.748114490509034,
      "epoch": 0.9288824383164006,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000492103013953868,
      "loss": 5.5554,
      "mean_token_accuracy": 0.16245089769363402,
      "num_tokens": 21255541.0,
      "step": 10240
    },
    {
      "entropy": 5.609032535552979,
      "epoch": 0.9293359941944848,
      "grad_norm": 1.875,
      "learning_rate": 0.0004920945146161242,
      "loss": 5.4162,
      "mean_token_accuracy": 0.17053214609622955,
      "num_tokens": 21265427.0,
      "step": 10245
    },
    {
      "entropy": 5.5645238876342775,
      "epoch": 0.9297895500725689,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004920860107888105,
      "loss": 5.4028,
      "mean_token_accuracy": 0.17055170685052873,
      "num_tokens": 21276597.0,
      "step": 10250
    },
    {
      "entropy": 5.585609579086304,
      "epoch": 0.9302431059506531,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004920775024721026,
      "loss": 5.4108,
      "mean_token_accuracy": 0.16735020279884338,
      "num_tokens": 21286416.0,
      "step": 10255
    },
    {
      "entropy": 5.60588436126709,
      "epoch": 0.9306966618287373,
      "grad_norm": 3.59375,
      "learning_rate": 0.0004920689896661767,
      "loss": 5.4395,
      "mean_token_accuracy": 0.1755400061607361,
      "num_tokens": 21296064.0,
      "step": 10260
    },
    {
      "entropy": 5.556747627258301,
      "epoch": 0.9311502177068215,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004920604723712088,
      "loss": 5.463,
      "mean_token_accuracy": 0.16989123821258545,
      "num_tokens": 21306836.0,
      "step": 10265
    },
    {
      "entropy": 5.569533777236939,
      "epoch": 0.9316037735849056,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004920519505873751,
      "loss": 5.414,
      "mean_token_accuracy": 0.17076023817062377,
      "num_tokens": 21316936.0,
      "step": 10270
    },
    {
      "entropy": 5.570468902587891,
      "epoch": 0.9320573294629898,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004920434243148515,
      "loss": 5.3685,
      "mean_token_accuracy": 0.17534030824899674,
      "num_tokens": 21326892.0,
      "step": 10275
    },
    {
      "entropy": 5.61998815536499,
      "epoch": 0.9325108853410741,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004920348935538147,
      "loss": 5.4101,
      "mean_token_accuracy": 0.16621700972318648,
      "num_tokens": 21337270.0,
      "step": 10280
    },
    {
      "entropy": 5.621613168716431,
      "epoch": 0.9329644412191582,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000492026358304441,
      "loss": 5.47,
      "mean_token_accuracy": 0.16533970683813096,
      "num_tokens": 21347968.0,
      "step": 10285
    },
    {
      "entropy": 5.577658987045288,
      "epoch": 0.9334179970972424,
      "grad_norm": 1.953125,
      "learning_rate": 0.000492017818566907,
      "loss": 5.4204,
      "mean_token_accuracy": 0.1734221026301384,
      "num_tokens": 21358312.0,
      "step": 10290
    },
    {
      "entropy": 5.544279146194458,
      "epoch": 0.9338715529753265,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004920092743413892,
      "loss": 5.3949,
      "mean_token_accuracy": 0.1729722023010254,
      "num_tokens": 21368794.0,
      "step": 10295
    },
    {
      "entropy": 5.659960556030273,
      "epoch": 0.9343251088534107,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004920007256280642,
      "loss": 5.4765,
      "mean_token_accuracy": 0.16844243556261063,
      "num_tokens": 21380080.0,
      "step": 10300
    },
    {
      "entropy": 5.648020219802857,
      "epoch": 0.934778664731495,
      "grad_norm": 2.171875,
      "learning_rate": 0.000491992172427109,
      "loss": 5.4247,
      "mean_token_accuracy": 0.1703952133655548,
      "num_tokens": 21389961.0,
      "step": 10305
    },
    {
      "entropy": 5.596118021011352,
      "epoch": 0.9352322206095791,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004919836147387002,
      "loss": 5.4857,
      "mean_token_accuracy": 0.1696673095226288,
      "num_tokens": 21399869.0,
      "step": 10310
    },
    {
      "entropy": 5.640964031219482,
      "epoch": 0.9356857764876633,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004919750525630151,
      "loss": 5.4528,
      "mean_token_accuracy": 0.1662815734744072,
      "num_tokens": 21410343.0,
      "step": 10315
    },
    {
      "entropy": 5.635088062286377,
      "epoch": 0.9361393323657474,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004919664859002306,
      "loss": 5.4672,
      "mean_token_accuracy": 0.16929827481508256,
      "num_tokens": 21420915.0,
      "step": 10320
    },
    {
      "entropy": 5.637465095520019,
      "epoch": 0.9365928882438317,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004919579147505239,
      "loss": 5.4098,
      "mean_token_accuracy": 0.1708474949002266,
      "num_tokens": 21430879.0,
      "step": 10325
    },
    {
      "entropy": 5.602361822128296,
      "epoch": 0.9370464441219158,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004919493391140721,
      "loss": 5.488,
      "mean_token_accuracy": 0.16649596393108368,
      "num_tokens": 21440843.0,
      "step": 10330
    },
    {
      "entropy": 5.674230480194092,
      "epoch": 0.9375,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004919407589910529,
      "loss": 5.6487,
      "mean_token_accuracy": 0.15944903194904328,
      "num_tokens": 21452132.0,
      "step": 10335
    },
    {
      "entropy": 5.667742013931274,
      "epoch": 0.9379535558780842,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004919321743816434,
      "loss": 5.4311,
      "mean_token_accuracy": 0.16516388356685638,
      "num_tokens": 21461861.0,
      "step": 10340
    },
    {
      "entropy": 5.591098022460938,
      "epoch": 0.9384071117561683,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004919235852860213,
      "loss": 5.4364,
      "mean_token_accuracy": 0.1637754425406456,
      "num_tokens": 21472244.0,
      "step": 10345
    },
    {
      "entropy": 5.635477733612061,
      "epoch": 0.9388606676342526,
      "grad_norm": 1.5,
      "learning_rate": 0.0004919149917043641,
      "loss": 5.49,
      "mean_token_accuracy": 0.1652377024292946,
      "num_tokens": 21482394.0,
      "step": 10350
    },
    {
      "entropy": 5.654922580718994,
      "epoch": 0.9393142235123367,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004919063936368496,
      "loss": 5.5379,
      "mean_token_accuracy": 0.16441430747509003,
      "num_tokens": 21493334.0,
      "step": 10355
    },
    {
      "entropy": 5.699097156524658,
      "epoch": 0.9397677793904209,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004918977910836557,
      "loss": 5.4858,
      "mean_token_accuracy": 0.1640081062912941,
      "num_tokens": 21504443.0,
      "step": 10360
    },
    {
      "entropy": 5.647023248672485,
      "epoch": 0.940221335268505,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004918891840449601,
      "loss": 5.5114,
      "mean_token_accuracy": 0.165273255109787,
      "num_tokens": 21513908.0,
      "step": 10365
    },
    {
      "entropy": 5.655364942550659,
      "epoch": 0.9406748911465893,
      "grad_norm": 1.890625,
      "learning_rate": 0.000491880572520941,
      "loss": 5.5268,
      "mean_token_accuracy": 0.15919932425022126,
      "num_tokens": 21524349.0,
      "step": 10370
    },
    {
      "entropy": 5.7412327289581295,
      "epoch": 0.9411284470246735,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004918719565117763,
      "loss": 5.5616,
      "mean_token_accuracy": 0.16886789351701736,
      "num_tokens": 21535003.0,
      "step": 10375
    },
    {
      "entropy": 5.734970474243164,
      "epoch": 0.9415820029027576,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004918633360176444,
      "loss": 5.5939,
      "mean_token_accuracy": 0.15955100655555726,
      "num_tokens": 21545045.0,
      "step": 10380
    },
    {
      "entropy": 5.5925661563873295,
      "epoch": 0.9420355587808418,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004918547110387234,
      "loss": 5.4073,
      "mean_token_accuracy": 0.16955724358558655,
      "num_tokens": 21554830.0,
      "step": 10385
    },
    {
      "entropy": 5.615862941741943,
      "epoch": 0.9424891146589259,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004918460815751916,
      "loss": 5.3668,
      "mean_token_accuracy": 0.16941992640495301,
      "num_tokens": 21564693.0,
      "step": 10390
    },
    {
      "entropy": 5.583719253540039,
      "epoch": 0.9429426705370102,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004918374476272277,
      "loss": 5.3961,
      "mean_token_accuracy": 0.16966528594493865,
      "num_tokens": 21575475.0,
      "step": 10395
    },
    {
      "entropy": 5.5933574676513675,
      "epoch": 0.9433962264150944,
      "grad_norm": 1.75,
      "learning_rate": 0.0004918288091950101,
      "loss": 5.4287,
      "mean_token_accuracy": 0.16622391939163209,
      "num_tokens": 21586944.0,
      "step": 10400
    },
    {
      "entropy": 5.691214942932129,
      "epoch": 0.9438497822931785,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004918201662787174,
      "loss": 5.453,
      "mean_token_accuracy": 0.16582643538713454,
      "num_tokens": 21598170.0,
      "step": 10405
    },
    {
      "entropy": 5.575848245620728,
      "epoch": 0.9443033381712627,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004918115188785285,
      "loss": 5.4645,
      "mean_token_accuracy": 0.1633247345685959,
      "num_tokens": 21609170.0,
      "step": 10410
    },
    {
      "entropy": 5.630935955047607,
      "epoch": 0.9447568940493469,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004918028669946221,
      "loss": 5.5295,
      "mean_token_accuracy": 0.16391694992780687,
      "num_tokens": 21620171.0,
      "step": 10415
    },
    {
      "entropy": 5.694646644592285,
      "epoch": 0.9452104499274311,
      "grad_norm": 2.0,
      "learning_rate": 0.0004917942106271771,
      "loss": 5.4801,
      "mean_token_accuracy": 0.17085294872522355,
      "num_tokens": 21631072.0,
      "step": 10420
    },
    {
      "entropy": 5.559463977813721,
      "epoch": 0.9456640058055152,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004917855497763727,
      "loss": 5.3375,
      "mean_token_accuracy": 0.17428645938634874,
      "num_tokens": 21641543.0,
      "step": 10425
    },
    {
      "entropy": 5.57183518409729,
      "epoch": 0.9461175616835994,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004917768844423878,
      "loss": 5.455,
      "mean_token_accuracy": 0.16970002353191377,
      "num_tokens": 21651342.0,
      "step": 10430
    },
    {
      "entropy": 5.677303838729858,
      "epoch": 0.9465711175616836,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004917682146254017,
      "loss": 5.4307,
      "mean_token_accuracy": 0.1728857398033142,
      "num_tokens": 21661043.0,
      "step": 10435
    },
    {
      "entropy": 5.538600015640259,
      "epoch": 0.9470246734397678,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004917595403255938,
      "loss": 5.3096,
      "mean_token_accuracy": 0.17871788889169693,
      "num_tokens": 21671197.0,
      "step": 10440
    },
    {
      "entropy": 5.501651239395142,
      "epoch": 0.947478229317852,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004917508615431431,
      "loss": 5.3921,
      "mean_token_accuracy": 0.17176435738801957,
      "num_tokens": 21682066.0,
      "step": 10445
    },
    {
      "entropy": 5.654246997833252,
      "epoch": 0.9479317851959361,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004917421782782296,
      "loss": 5.4209,
      "mean_token_accuracy": 0.1663619577884674,
      "num_tokens": 21692117.0,
      "step": 10450
    },
    {
      "entropy": 5.694399499893189,
      "epoch": 0.9483853410740203,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004917334905310325,
      "loss": 5.5368,
      "mean_token_accuracy": 0.15789986699819564,
      "num_tokens": 21703544.0,
      "step": 10455
    },
    {
      "entropy": 5.5399058818817135,
      "epoch": 0.9488388969521045,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004917247983017317,
      "loss": 5.352,
      "mean_token_accuracy": 0.17476127594709395,
      "num_tokens": 21712850.0,
      "step": 10460
    },
    {
      "entropy": 5.614771795272827,
      "epoch": 0.9492924528301887,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004917161015905067,
      "loss": 5.5185,
      "mean_token_accuracy": 0.16575202494859695,
      "num_tokens": 21723015.0,
      "step": 10465
    },
    {
      "entropy": 5.598294687271118,
      "epoch": 0.9497460087082729,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004917074003975376,
      "loss": 5.4745,
      "mean_token_accuracy": 0.16261564195156097,
      "num_tokens": 21732453.0,
      "step": 10470
    },
    {
      "entropy": 5.646253156661987,
      "epoch": 0.950199564586357,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004916986947230042,
      "loss": 5.5315,
      "mean_token_accuracy": 0.1652168944478035,
      "num_tokens": 21743251.0,
      "step": 10475
    },
    {
      "entropy": 5.618532323837281,
      "epoch": 0.9506531204644412,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004916899845670866,
      "loss": 5.4457,
      "mean_token_accuracy": 0.1703191101551056,
      "num_tokens": 21752891.0,
      "step": 10480
    },
    {
      "entropy": 5.6609643459320065,
      "epoch": 0.9511066763425254,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004916812699299649,
      "loss": 5.4653,
      "mean_token_accuracy": 0.16858232170343398,
      "num_tokens": 21763230.0,
      "step": 10485
    },
    {
      "entropy": 5.6318341255187985,
      "epoch": 0.9515602322206096,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004916725508118193,
      "loss": 5.4042,
      "mean_token_accuracy": 0.16566513776779174,
      "num_tokens": 21773663.0,
      "step": 10490
    },
    {
      "entropy": 5.627205228805542,
      "epoch": 0.9520137880986937,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004916638272128301,
      "loss": 5.4661,
      "mean_token_accuracy": 0.16785551458597184,
      "num_tokens": 21784265.0,
      "step": 10495
    },
    {
      "entropy": 5.554153251647949,
      "epoch": 0.9524673439767779,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004916550991331777,
      "loss": 5.4123,
      "mean_token_accuracy": 0.17339823246002198,
      "num_tokens": 21795056.0,
      "step": 10500
    },
    {
      "entropy": 5.578380060195923,
      "epoch": 0.9529208998548622,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004916463665730427,
      "loss": 5.4303,
      "mean_token_accuracy": 0.16940593868494033,
      "num_tokens": 21806204.0,
      "step": 10505
    },
    {
      "entropy": 5.603130865097046,
      "epoch": 0.9533744557329463,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004916376295326057,
      "loss": 5.4633,
      "mean_token_accuracy": 0.16669130325317383,
      "num_tokens": 21816993.0,
      "step": 10510
    },
    {
      "entropy": 5.5685646533966064,
      "epoch": 0.9538280116110305,
      "grad_norm": 3.078125,
      "learning_rate": 0.0004916288880120472,
      "loss": 5.3222,
      "mean_token_accuracy": 0.17768334001302719,
      "num_tokens": 21827457.0,
      "step": 10515
    },
    {
      "entropy": 5.594205617904663,
      "epoch": 0.9542815674891146,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004916201420115482,
      "loss": 5.4239,
      "mean_token_accuracy": 0.16775002628564833,
      "num_tokens": 21837555.0,
      "step": 10520
    },
    {
      "entropy": 5.697423982620239,
      "epoch": 0.9547351233671988,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004916113915312895,
      "loss": 5.4504,
      "mean_token_accuracy": 0.17421349436044692,
      "num_tokens": 21847074.0,
      "step": 10525
    },
    {
      "entropy": 5.668843555450439,
      "epoch": 0.9551886792452831,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004916026365714519,
      "loss": 5.4371,
      "mean_token_accuracy": 0.16593135595321656,
      "num_tokens": 21858333.0,
      "step": 10530
    },
    {
      "entropy": 5.541938638687133,
      "epoch": 0.9556422351233672,
      "grad_norm": 5.28125,
      "learning_rate": 0.0004915938771322165,
      "loss": 5.458,
      "mean_token_accuracy": 0.16816020756959915,
      "num_tokens": 21869132.0,
      "step": 10535
    },
    {
      "entropy": 5.560562515258789,
      "epoch": 0.9560957910014514,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004915851132137646,
      "loss": 5.4986,
      "mean_token_accuracy": 0.16549153625965118,
      "num_tokens": 21879153.0,
      "step": 10540
    },
    {
      "entropy": 5.666241884231567,
      "epoch": 0.9565493468795355,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004915763448162774,
      "loss": 5.3652,
      "mean_token_accuracy": 0.17085392624139786,
      "num_tokens": 21889063.0,
      "step": 10545
    },
    {
      "entropy": 5.685885953903198,
      "epoch": 0.9570029027576198,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004915675719399362,
      "loss": 5.498,
      "mean_token_accuracy": 0.16560063809156417,
      "num_tokens": 21899005.0,
      "step": 10550
    },
    {
      "entropy": 5.639123916625977,
      "epoch": 0.9574564586357039,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004915587945849225,
      "loss": 5.3559,
      "mean_token_accuracy": 0.17045865058898926,
      "num_tokens": 21908391.0,
      "step": 10555
    },
    {
      "entropy": 5.613085746765137,
      "epoch": 0.9579100145137881,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004915500127514176,
      "loss": 5.4354,
      "mean_token_accuracy": 0.1698034629225731,
      "num_tokens": 21917655.0,
      "step": 10560
    },
    {
      "entropy": 5.7150092124938965,
      "epoch": 0.9583635703918723,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004915412264396034,
      "loss": 5.5009,
      "mean_token_accuracy": 0.1664179176092148,
      "num_tokens": 21928038.0,
      "step": 10565
    },
    {
      "entropy": 5.629287338256836,
      "epoch": 0.9588171262699564,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004915324356496613,
      "loss": 5.3813,
      "mean_token_accuracy": 0.1750314638018608,
      "num_tokens": 21938106.0,
      "step": 10570
    },
    {
      "entropy": 5.575638914108277,
      "epoch": 0.9592706821480407,
      "grad_norm": 4.78125,
      "learning_rate": 0.0004915236403817733,
      "loss": 5.3991,
      "mean_token_accuracy": 0.17556844651699066,
      "num_tokens": 21947484.0,
      "step": 10575
    },
    {
      "entropy": 5.603404188156128,
      "epoch": 0.9597242380261248,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004915148406361213,
      "loss": 5.4683,
      "mean_token_accuracy": 0.1709429606795311,
      "num_tokens": 21957139.0,
      "step": 10580
    },
    {
      "entropy": 5.618351602554322,
      "epoch": 0.960177793904209,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004915060364128872,
      "loss": 5.4586,
      "mean_token_accuracy": 0.16406761407852172,
      "num_tokens": 21967030.0,
      "step": 10585
    },
    {
      "entropy": 5.593357563018799,
      "epoch": 0.9606313497822931,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004914972277122532,
      "loss": 5.421,
      "mean_token_accuracy": 0.1659909650683403,
      "num_tokens": 21976418.0,
      "step": 10590
    },
    {
      "entropy": 5.595254135131836,
      "epoch": 0.9610849056603774,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004914884145344012,
      "loss": 5.4269,
      "mean_token_accuracy": 0.17377042919397354,
      "num_tokens": 21987284.0,
      "step": 10595
    },
    {
      "entropy": 5.500611209869385,
      "epoch": 0.9615384615384616,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004914795968795137,
      "loss": 5.3406,
      "mean_token_accuracy": 0.17092138677835464,
      "num_tokens": 21997106.0,
      "step": 10600
    },
    {
      "entropy": 5.594227981567383,
      "epoch": 0.9619920174165457,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004914707747477728,
      "loss": 5.3284,
      "mean_token_accuracy": 0.17086504995822907,
      "num_tokens": 22007589.0,
      "step": 10605
    },
    {
      "entropy": 5.621239805221558,
      "epoch": 0.9624455732946299,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004914619481393613,
      "loss": 5.4618,
      "mean_token_accuracy": 0.16725393682718276,
      "num_tokens": 22017669.0,
      "step": 10610
    },
    {
      "entropy": 5.732152843475342,
      "epoch": 0.962899129172714,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004914531170544615,
      "loss": 5.6239,
      "mean_token_accuracy": 0.16082653999328614,
      "num_tokens": 22029330.0,
      "step": 10615
    },
    {
      "entropy": 5.58074631690979,
      "epoch": 0.9633526850507983,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004914442814932561,
      "loss": 5.3937,
      "mean_token_accuracy": 0.16925872266292571,
      "num_tokens": 22039877.0,
      "step": 10620
    },
    {
      "entropy": 5.604477071762085,
      "epoch": 0.9638062409288825,
      "grad_norm": 2.125,
      "learning_rate": 0.0004914354414559278,
      "loss": 5.439,
      "mean_token_accuracy": 0.17077770978212356,
      "num_tokens": 22049579.0,
      "step": 10625
    },
    {
      "entropy": 5.504826402664184,
      "epoch": 0.9642597968069666,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004914265969426594,
      "loss": 5.3331,
      "mean_token_accuracy": 0.1832723319530487,
      "num_tokens": 22059608.0,
      "step": 10630
    },
    {
      "entropy": 5.55698938369751,
      "epoch": 0.9647133526850508,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004914177479536338,
      "loss": 5.4679,
      "mean_token_accuracy": 0.16628993898630143,
      "num_tokens": 22070036.0,
      "step": 10635
    },
    {
      "entropy": 5.5715710639953615,
      "epoch": 0.965166908563135,
      "grad_norm": 2.125,
      "learning_rate": 0.000491408894489034,
      "loss": 5.4605,
      "mean_token_accuracy": 0.16512693017721175,
      "num_tokens": 22080679.0,
      "step": 10640
    },
    {
      "entropy": 5.610262632369995,
      "epoch": 0.9656204644412192,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004914000365490432,
      "loss": 5.3036,
      "mean_token_accuracy": 0.1728259280323982,
      "num_tokens": 22092192.0,
      "step": 10645
    },
    {
      "entropy": 5.620035028457641,
      "epoch": 0.9660740203193033,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004913911741338445,
      "loss": 5.5059,
      "mean_token_accuracy": 0.1626466490328312,
      "num_tokens": 22102722.0,
      "step": 10650
    },
    {
      "entropy": 5.625868129730224,
      "epoch": 0.9665275761973875,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004913823072436211,
      "loss": 5.3747,
      "mean_token_accuracy": 0.17529655545949935,
      "num_tokens": 22113521.0,
      "step": 10655
    },
    {
      "entropy": 5.64664397239685,
      "epoch": 0.9669811320754716,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004913734358785565,
      "loss": 5.4623,
      "mean_token_accuracy": 0.16819260120391846,
      "num_tokens": 22124891.0,
      "step": 10660
    },
    {
      "entropy": 5.53647198677063,
      "epoch": 0.9674346879535559,
      "grad_norm": 3.078125,
      "learning_rate": 0.0004913645600388341,
      "loss": 5.4714,
      "mean_token_accuracy": 0.1639208272099495,
      "num_tokens": 22135972.0,
      "step": 10665
    },
    {
      "entropy": 5.657358455657959,
      "epoch": 0.9678882438316401,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004913556797246374,
      "loss": 5.5193,
      "mean_token_accuracy": 0.15633328408002853,
      "num_tokens": 22146623.0,
      "step": 10670
    },
    {
      "entropy": 5.737673711776734,
      "epoch": 0.9683417997097242,
      "grad_norm": 2.5,
      "learning_rate": 0.0004913467949361503,
      "loss": 5.4368,
      "mean_token_accuracy": 0.1660117581486702,
      "num_tokens": 22157619.0,
      "step": 10675
    },
    {
      "entropy": 5.6835862159729,
      "epoch": 0.9687953555878084,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004913379056735561,
      "loss": 5.4414,
      "mean_token_accuracy": 0.16401823461055756,
      "num_tokens": 22168655.0,
      "step": 10680
    },
    {
      "entropy": 5.6037169933319095,
      "epoch": 0.9692489114658926,
      "grad_norm": 2.0625,
      "learning_rate": 0.000491329011937039,
      "loss": 5.4783,
      "mean_token_accuracy": 0.16528795212507247,
      "num_tokens": 22178500.0,
      "step": 10685
    },
    {
      "entropy": 5.677508354187012,
      "epoch": 0.9697024673439768,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004913201137267828,
      "loss": 5.469,
      "mean_token_accuracy": 0.16165651381015778,
      "num_tokens": 22188873.0,
      "step": 10690
    },
    {
      "entropy": 5.586783695220947,
      "epoch": 0.970156023222061,
      "grad_norm": 3.25,
      "learning_rate": 0.0004913112110429715,
      "loss": 5.3817,
      "mean_token_accuracy": 0.1676276743412018,
      "num_tokens": 22199743.0,
      "step": 10695
    },
    {
      "entropy": 5.557729578018188,
      "epoch": 0.9706095791001451,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004913023038857892,
      "loss": 5.3564,
      "mean_token_accuracy": 0.17372817248106004,
      "num_tokens": 22210301.0,
      "step": 10700
    },
    {
      "entropy": 5.558896493911743,
      "epoch": 0.9710631349782293,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004912933922554203,
      "loss": 5.3625,
      "mean_token_accuracy": 0.17335860878229142,
      "num_tokens": 22220771.0,
      "step": 10705
    },
    {
      "entropy": 5.609603786468506,
      "epoch": 0.9715166908563135,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004912844761520487,
      "loss": 5.4487,
      "mean_token_accuracy": 0.16578324362635613,
      "num_tokens": 22231100.0,
      "step": 10710
    },
    {
      "entropy": 5.63015341758728,
      "epoch": 0.9719702467343977,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004912755555758592,
      "loss": 5.4431,
      "mean_token_accuracy": 0.16642088890075685,
      "num_tokens": 22240997.0,
      "step": 10715
    },
    {
      "entropy": 5.607190656661987,
      "epoch": 0.9724238026124818,
      "grad_norm": 2.015625,
      "learning_rate": 0.000491266630527036,
      "loss": 5.4656,
      "mean_token_accuracy": 0.16432343423366547,
      "num_tokens": 22251795.0,
      "step": 10720
    },
    {
      "entropy": 5.6826615810394285,
      "epoch": 0.972877358490566,
      "grad_norm": 3.296875,
      "learning_rate": 0.0004912577010057638,
      "loss": 5.4957,
      "mean_token_accuracy": 0.16548267304897307,
      "num_tokens": 22262465.0,
      "step": 10725
    },
    {
      "entropy": 5.744708251953125,
      "epoch": 0.9733309143686503,
      "grad_norm": 2.375,
      "learning_rate": 0.0004912487670122273,
      "loss": 5.5933,
      "mean_token_accuracy": 0.16162458509206773,
      "num_tokens": 22272399.0,
      "step": 10730
    },
    {
      "entropy": 5.687432813644409,
      "epoch": 0.9737844702467344,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004912398285466111,
      "loss": 5.516,
      "mean_token_accuracy": 0.16217903792858124,
      "num_tokens": 22282545.0,
      "step": 10735
    },
    {
      "entropy": 5.626268434524536,
      "epoch": 0.9742380261248186,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004912308856091002,
      "loss": 5.4557,
      "mean_token_accuracy": 0.16961072236299515,
      "num_tokens": 22293154.0,
      "step": 10740
    },
    {
      "entropy": 5.633498001098633,
      "epoch": 0.9746915820029027,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004912219381998795,
      "loss": 5.3943,
      "mean_token_accuracy": 0.17631157487630844,
      "num_tokens": 22302309.0,
      "step": 10745
    },
    {
      "entropy": 5.608208513259887,
      "epoch": 0.9751451378809869,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004912129863191339,
      "loss": 5.5102,
      "mean_token_accuracy": 0.15686720460653306,
      "num_tokens": 22312316.0,
      "step": 10750
    },
    {
      "entropy": 5.652306604385376,
      "epoch": 0.9755986937590712,
      "grad_norm": 2.125,
      "learning_rate": 0.0004912040299670487,
      "loss": 5.4865,
      "mean_token_accuracy": 0.16555508375167846,
      "num_tokens": 22322387.0,
      "step": 10755
    },
    {
      "entropy": 5.691938161849976,
      "epoch": 0.9760522496371553,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004911950691438091,
      "loss": 5.4573,
      "mean_token_accuracy": 0.1755552962422371,
      "num_tokens": 22332759.0,
      "step": 10760
    },
    {
      "entropy": 5.665207719802856,
      "epoch": 0.9765058055152395,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004911861038496004,
      "loss": 5.5144,
      "mean_token_accuracy": 0.16096826121211052,
      "num_tokens": 22344056.0,
      "step": 10765
    },
    {
      "entropy": 5.541015195846557,
      "epoch": 0.9769593613933236,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004911771340846079,
      "loss": 5.4595,
      "mean_token_accuracy": 0.16349495202302933,
      "num_tokens": 22353877.0,
      "step": 10770
    },
    {
      "entropy": 5.6556745052337645,
      "epoch": 0.9774129172714079,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004911681598490172,
      "loss": 5.4795,
      "mean_token_accuracy": 0.1672011524438858,
      "num_tokens": 22363549.0,
      "step": 10775
    },
    {
      "entropy": 5.737138319015503,
      "epoch": 0.977866473149492,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004911591811430139,
      "loss": 5.5342,
      "mean_token_accuracy": 0.1661311835050583,
      "num_tokens": 22374040.0,
      "step": 10780
    },
    {
      "entropy": 5.660326099395752,
      "epoch": 0.9783200290275762,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004911501979667836,
      "loss": 5.5915,
      "mean_token_accuracy": 0.16359687969088554,
      "num_tokens": 22384819.0,
      "step": 10785
    },
    {
      "entropy": 5.646378850936889,
      "epoch": 0.9787735849056604,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004911412103205122,
      "loss": 5.4162,
      "mean_token_accuracy": 0.1688998430967331,
      "num_tokens": 22395130.0,
      "step": 10790
    },
    {
      "entropy": 5.616553831100464,
      "epoch": 0.9792271407837445,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004911322182043855,
      "loss": 5.4576,
      "mean_token_accuracy": 0.1648877426981926,
      "num_tokens": 22405093.0,
      "step": 10795
    },
    {
      "entropy": 5.657214879989624,
      "epoch": 0.9796806966618288,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004911232216185893,
      "loss": 5.4734,
      "mean_token_accuracy": 0.16335293501615525,
      "num_tokens": 22415008.0,
      "step": 10800
    },
    {
      "entropy": 5.672890996932983,
      "epoch": 0.9801342525399129,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004911142205633099,
      "loss": 5.5315,
      "mean_token_accuracy": 0.16531921476125716,
      "num_tokens": 22426486.0,
      "step": 10805
    },
    {
      "entropy": 5.619123888015747,
      "epoch": 0.9805878084179971,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004911052150387334,
      "loss": 5.4504,
      "mean_token_accuracy": 0.16570301055908204,
      "num_tokens": 22436844.0,
      "step": 10810
    },
    {
      "entropy": 5.696360635757446,
      "epoch": 0.9810413642960812,
      "grad_norm": 2.25,
      "learning_rate": 0.0004910962050450458,
      "loss": 5.5439,
      "mean_token_accuracy": 0.1659548982977867,
      "num_tokens": 22446630.0,
      "step": 10815
    },
    {
      "entropy": 5.66130223274231,
      "epoch": 0.9814949201741655,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004910871905824338,
      "loss": 5.3869,
      "mean_token_accuracy": 0.17135941237211227,
      "num_tokens": 22456406.0,
      "step": 10820
    },
    {
      "entropy": 5.541624641418457,
      "epoch": 0.9819484760522497,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004910781716510837,
      "loss": 5.3782,
      "mean_token_accuracy": 0.1720363527536392,
      "num_tokens": 22466210.0,
      "step": 10825
    },
    {
      "entropy": 5.620612621307373,
      "epoch": 0.9824020319303338,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004910691482511817,
      "loss": 5.4841,
      "mean_token_accuracy": 0.16932221949100495,
      "num_tokens": 22476978.0,
      "step": 10830
    },
    {
      "entropy": 5.7275959014892575,
      "epoch": 0.982855587808418,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004910601203829148,
      "loss": 5.4462,
      "mean_token_accuracy": 0.1642207011580467,
      "num_tokens": 22488613.0,
      "step": 10835
    },
    {
      "entropy": 5.5737100601196286,
      "epoch": 0.9833091436865021,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004910510880464696,
      "loss": 5.3546,
      "mean_token_accuracy": 0.17655791491270065,
      "num_tokens": 22498312.0,
      "step": 10840
    },
    {
      "entropy": 5.546988391876221,
      "epoch": 0.9837626995645864,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004910420512420327,
      "loss": 5.4741,
      "mean_token_accuracy": 0.16846197545528413,
      "num_tokens": 22508131.0,
      "step": 10845
    },
    {
      "entropy": 5.67532377243042,
      "epoch": 0.9842162554426706,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004910330099697913,
      "loss": 5.449,
      "mean_token_accuracy": 0.17251691222190857,
      "num_tokens": 22517842.0,
      "step": 10850
    },
    {
      "entropy": 5.655220031738281,
      "epoch": 0.9846698113207547,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004910239642299321,
      "loss": 5.4366,
      "mean_token_accuracy": 0.1687016546726227,
      "num_tokens": 22528239.0,
      "step": 10855
    },
    {
      "entropy": 5.625150299072265,
      "epoch": 0.9851233671988389,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004910149140226422,
      "loss": 5.4232,
      "mean_token_accuracy": 0.16798526495695115,
      "num_tokens": 22539734.0,
      "step": 10860
    },
    {
      "entropy": 5.620212841033935,
      "epoch": 0.9855769230769231,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004910058593481089,
      "loss": 5.4372,
      "mean_token_accuracy": 0.17239609360694885,
      "num_tokens": 22549796.0,
      "step": 10865
    },
    {
      "entropy": 5.554531002044678,
      "epoch": 0.9860304789550073,
      "grad_norm": 2.734375,
      "learning_rate": 0.0004909968002065195,
      "loss": 5.4103,
      "mean_token_accuracy": 0.17158298045396805,
      "num_tokens": 22559934.0,
      "step": 10870
    },
    {
      "entropy": 5.662750625610352,
      "epoch": 0.9864840348330914,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000490987736598061,
      "loss": 5.5042,
      "mean_token_accuracy": 0.15988094210624695,
      "num_tokens": 22569864.0,
      "step": 10875
    },
    {
      "entropy": 5.611508750915528,
      "epoch": 0.9869375907111756,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004909786685229211,
      "loss": 5.4641,
      "mean_token_accuracy": 0.17335709184408188,
      "num_tokens": 22580687.0,
      "step": 10880
    },
    {
      "entropy": 5.648767518997192,
      "epoch": 0.9873911465892597,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004909695959812875,
      "loss": 5.4814,
      "mean_token_accuracy": 0.16742359101772308,
      "num_tokens": 22590507.0,
      "step": 10885
    },
    {
      "entropy": 5.681576299667358,
      "epoch": 0.987844702467344,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004909605189733474,
      "loss": 5.5178,
      "mean_token_accuracy": 0.16965399086475372,
      "num_tokens": 22601901.0,
      "step": 10890
    },
    {
      "entropy": 5.578258848190307,
      "epoch": 0.9882982583454282,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004909514374992889,
      "loss": 5.4577,
      "mean_token_accuracy": 0.1665840521454811,
      "num_tokens": 22613003.0,
      "step": 10895
    },
    {
      "entropy": 5.582123470306397,
      "epoch": 0.9887518142235123,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004909423515592996,
      "loss": 5.3163,
      "mean_token_accuracy": 0.17244535386562349,
      "num_tokens": 22623401.0,
      "step": 10900
    },
    {
      "entropy": 5.639914417266846,
      "epoch": 0.9892053701015965,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004909332611535675,
      "loss": 5.4738,
      "mean_token_accuracy": 0.1673109322786331,
      "num_tokens": 22633147.0,
      "step": 10905
    },
    {
      "entropy": 5.634562587738037,
      "epoch": 0.9896589259796807,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004909241662822804,
      "loss": 5.3143,
      "mean_token_accuracy": 0.18473917990922928,
      "num_tokens": 22642990.0,
      "step": 10910
    },
    {
      "entropy": 5.662867212295533,
      "epoch": 0.9901124818577649,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004909150669456265,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16324340552091599,
      "num_tokens": 22652734.0,
      "step": 10915
    },
    {
      "entropy": 5.61931586265564,
      "epoch": 0.9905660377358491,
      "grad_norm": 2.203125,
      "learning_rate": 0.000490905963143794,
      "loss": 5.5545,
      "mean_token_accuracy": 0.1586074098944664,
      "num_tokens": 22662968.0,
      "step": 10920
    },
    {
      "entropy": 5.570623779296875,
      "epoch": 0.9910195936139332,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004908968548769713,
      "loss": 5.369,
      "mean_token_accuracy": 0.17906321883201598,
      "num_tokens": 22673157.0,
      "step": 10925
    },
    {
      "entropy": 5.661365270614624,
      "epoch": 0.9914731494920174,
      "grad_norm": 3.09375,
      "learning_rate": 0.0004908877421453466,
      "loss": 5.5042,
      "mean_token_accuracy": 0.16541301161050798,
      "num_tokens": 22683105.0,
      "step": 10930
    },
    {
      "entropy": 5.677806234359741,
      "epoch": 0.9919267053701016,
      "grad_norm": 2.65625,
      "learning_rate": 0.0004908786249491082,
      "loss": 5.4559,
      "mean_token_accuracy": 0.1736716389656067,
      "num_tokens": 22693793.0,
      "step": 10935
    },
    {
      "entropy": 5.736819887161255,
      "epoch": 0.9923802612481858,
      "grad_norm": 2.015625,
      "learning_rate": 0.000490869503288445,
      "loss": 5.4957,
      "mean_token_accuracy": 0.16935908496379853,
      "num_tokens": 22703884.0,
      "step": 10940
    },
    {
      "entropy": 5.64982419013977,
      "epoch": 0.99283381712627,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004908603771635454,
      "loss": 5.5239,
      "mean_token_accuracy": 0.1698847383260727,
      "num_tokens": 22713699.0,
      "step": 10945
    },
    {
      "entropy": 5.6703019618988035,
      "epoch": 0.9932873730043541,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004908512465745982,
      "loss": 5.3645,
      "mean_token_accuracy": 0.1739519238471985,
      "num_tokens": 22722871.0,
      "step": 10950
    },
    {
      "entropy": 5.618347883224487,
      "epoch": 0.9937409288824384,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004908421115217922,
      "loss": 5.3935,
      "mean_token_accuracy": 0.1663510337471962,
      "num_tokens": 22733314.0,
      "step": 10955
    },
    {
      "entropy": 5.541477346420288,
      "epoch": 0.9941944847605225,
      "grad_norm": 4.0,
      "learning_rate": 0.0004908329720053164,
      "loss": 5.3755,
      "mean_token_accuracy": 0.1778403714299202,
      "num_tokens": 22743255.0,
      "step": 10960
    },
    {
      "entropy": 5.504152488708496,
      "epoch": 0.9946480406386067,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004908238280253597,
      "loss": 5.3771,
      "mean_token_accuracy": 0.17440174520015717,
      "num_tokens": 22753857.0,
      "step": 10965
    },
    {
      "entropy": 5.617768621444702,
      "epoch": 0.9951015965166908,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004908146795821112,
      "loss": 5.3717,
      "mean_token_accuracy": 0.166821451485157,
      "num_tokens": 22764326.0,
      "step": 10970
    },
    {
      "entropy": 5.653025674819946,
      "epoch": 0.995555152394775,
      "grad_norm": 4.0625,
      "learning_rate": 0.0004908055266757602,
      "loss": 5.4636,
      "mean_token_accuracy": 0.1688439816236496,
      "num_tokens": 22774958.0,
      "step": 10975
    },
    {
      "entropy": 5.584131574630737,
      "epoch": 0.9960087082728593,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004907963693064958,
      "loss": 5.3461,
      "mean_token_accuracy": 0.16801189184188842,
      "num_tokens": 22785616.0,
      "step": 10980
    },
    {
      "entropy": 5.6416544914245605,
      "epoch": 0.9964622641509434,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004907872074745075,
      "loss": 5.5217,
      "mean_token_accuracy": 0.1628538802266121,
      "num_tokens": 22795967.0,
      "step": 10985
    },
    {
      "entropy": 5.594269466400147,
      "epoch": 0.9969158200290276,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004907780411799849,
      "loss": 5.5457,
      "mean_token_accuracy": 0.16120059192180633,
      "num_tokens": 22806335.0,
      "step": 10990
    },
    {
      "entropy": 5.702027273178101,
      "epoch": 0.9973693759071117,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004907688704231173,
      "loss": 5.446,
      "mean_token_accuracy": 0.1731191873550415,
      "num_tokens": 22816390.0,
      "step": 10995
    },
    {
      "entropy": 5.7088724136352536,
      "epoch": 0.997822931785196,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004907596952040945,
      "loss": 5.523,
      "mean_token_accuracy": 0.1675207257270813,
      "num_tokens": 22826395.0,
      "step": 11000
    },
    {
      "entropy": 5.679268312454224,
      "epoch": 0.9982764876632801,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004907505155231063,
      "loss": 5.5056,
      "mean_token_accuracy": 0.16937604695558547,
      "num_tokens": 22836790.0,
      "step": 11005
    },
    {
      "entropy": 5.599216890335083,
      "epoch": 0.9987300435413643,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004907413313803424,
      "loss": 5.3913,
      "mean_token_accuracy": 0.16755914837121963,
      "num_tokens": 22846687.0,
      "step": 11010
    },
    {
      "entropy": 5.633000278472901,
      "epoch": 0.9991835994194485,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004907321427759928,
      "loss": 5.4314,
      "mean_token_accuracy": 0.17038674354553224,
      "num_tokens": 22857208.0,
      "step": 11015
    },
    {
      "entropy": 5.656092023849487,
      "epoch": 0.9996371552975326,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004907229497102475,
      "loss": 5.5455,
      "mean_token_accuracy": 0.16008800864219666,
      "num_tokens": 22868556.0,
      "step": 11020
    },
    {
      "entropy": 5.639604282379151,
      "epoch": 1.0000907111756168,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004907137521832966,
      "loss": 5.505,
      "mean_token_accuracy": 0.16944388896226883,
      "num_tokens": 22880081.0,
      "step": 11025
    },
    {
      "entropy": 5.6398672580719,
      "epoch": 1.000544267053701,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004907045501953304,
      "loss": 5.3217,
      "mean_token_accuracy": 0.1732504278421402,
      "num_tokens": 22892141.0,
      "step": 11030
    },
    {
      "entropy": 5.637064456939697,
      "epoch": 1.0009978229317853,
      "grad_norm": 2.234375,
      "learning_rate": 0.000490695343746539,
      "loss": 5.3847,
      "mean_token_accuracy": 0.1710570126771927,
      "num_tokens": 22902688.0,
      "step": 11035
    },
    {
      "entropy": 5.497150611877442,
      "epoch": 1.0014513788098693,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004906861328371131,
      "loss": 5.3066,
      "mean_token_accuracy": 0.17424043715000154,
      "num_tokens": 22912453.0,
      "step": 11040
    },
    {
      "entropy": 5.631495571136474,
      "epoch": 1.0019049346879536,
      "grad_norm": 1.875,
      "learning_rate": 0.0004906769174672429,
      "loss": 5.438,
      "mean_token_accuracy": 0.16529137939214705,
      "num_tokens": 22923167.0,
      "step": 11045
    },
    {
      "entropy": 5.729016590118408,
      "epoch": 1.0023584905660377,
      "grad_norm": 1.9140625,
      "learning_rate": 0.000490667697637119,
      "loss": 5.43,
      "mean_token_accuracy": 0.16472474336624146,
      "num_tokens": 22934753.0,
      "step": 11050
    },
    {
      "entropy": 5.587201690673828,
      "epoch": 1.002812046444122,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004906584733469324,
      "loss": 5.3371,
      "mean_token_accuracy": 0.17595323026180268,
      "num_tokens": 22945277.0,
      "step": 11055
    },
    {
      "entropy": 5.613748550415039,
      "epoch": 1.0032656023222062,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004906492445968735,
      "loss": 5.4152,
      "mean_token_accuracy": 0.16204020231962205,
      "num_tokens": 22955797.0,
      "step": 11060
    },
    {
      "entropy": 5.683445882797241,
      "epoch": 1.0037191582002902,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004906400113871332,
      "loss": 5.404,
      "mean_token_accuracy": 0.16932731568813325,
      "num_tokens": 22965969.0,
      "step": 11065
    },
    {
      "entropy": 5.689712619781494,
      "epoch": 1.0041727140783745,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004906307737179026,
      "loss": 5.3642,
      "mean_token_accuracy": 0.17712394148111343,
      "num_tokens": 22975323.0,
      "step": 11070
    },
    {
      "entropy": 5.61050066947937,
      "epoch": 1.0046262699564585,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004906215315893726,
      "loss": 5.364,
      "mean_token_accuracy": 0.17275502830743789,
      "num_tokens": 22985063.0,
      "step": 11075
    },
    {
      "entropy": 5.614771890640259,
      "epoch": 1.0050798258345428,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004906122850017344,
      "loss": 5.5039,
      "mean_token_accuracy": 0.1602733001112938,
      "num_tokens": 22996001.0,
      "step": 11080
    },
    {
      "entropy": 5.636023378372192,
      "epoch": 1.005533381712627,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004906030339551792,
      "loss": 5.397,
      "mean_token_accuracy": 0.1702730357646942,
      "num_tokens": 23006896.0,
      "step": 11085
    },
    {
      "entropy": 5.630139017105103,
      "epoch": 1.0059869375907111,
      "grad_norm": 2.875,
      "learning_rate": 0.0004905937784498983,
      "loss": 5.3757,
      "mean_token_accuracy": 0.17295269072055816,
      "num_tokens": 23017080.0,
      "step": 11090
    },
    {
      "entropy": 5.649996185302735,
      "epoch": 1.0064404934687954,
      "grad_norm": 6.0,
      "learning_rate": 0.0004905845184860831,
      "loss": 5.3337,
      "mean_token_accuracy": 0.1737719804048538,
      "num_tokens": 23027168.0,
      "step": 11095
    },
    {
      "entropy": 5.61572232246399,
      "epoch": 1.0068940493468794,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004905752540639252,
      "loss": 5.375,
      "mean_token_accuracy": 0.16635738760232927,
      "num_tokens": 23037364.0,
      "step": 11100
    },
    {
      "entropy": 5.531828784942627,
      "epoch": 1.0073476052249637,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004905659851836161,
      "loss": 5.2905,
      "mean_token_accuracy": 0.17861230075359344,
      "num_tokens": 23047609.0,
      "step": 11105
    },
    {
      "entropy": 5.608002805709839,
      "epoch": 1.007801161103048,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004905567118453475,
      "loss": 5.4224,
      "mean_token_accuracy": 0.17069784104824065,
      "num_tokens": 23057942.0,
      "step": 11110
    },
    {
      "entropy": 5.608658170700073,
      "epoch": 1.008254716981132,
      "grad_norm": 2.0,
      "learning_rate": 0.0004905474340493111,
      "loss": 5.3283,
      "mean_token_accuracy": 0.17128928005695343,
      "num_tokens": 23069119.0,
      "step": 11115
    },
    {
      "entropy": 5.557733488082886,
      "epoch": 1.0087082728592163,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004905381517956988,
      "loss": 5.3342,
      "mean_token_accuracy": 0.1733451947569847,
      "num_tokens": 23080060.0,
      "step": 11120
    },
    {
      "entropy": 5.568910789489746,
      "epoch": 1.0091618287373005,
      "grad_norm": 6.4375,
      "learning_rate": 0.0004905288650847027,
      "loss": 5.3607,
      "mean_token_accuracy": 0.16494861245155334,
      "num_tokens": 23090645.0,
      "step": 11125
    },
    {
      "entropy": 5.657036018371582,
      "epoch": 1.0096153846153846,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004905195739165146,
      "loss": 5.3811,
      "mean_token_accuracy": 0.16976771950721742,
      "num_tokens": 23100449.0,
      "step": 11130
    },
    {
      "entropy": 5.637253189086914,
      "epoch": 1.0100689404934688,
      "grad_norm": 1.8125,
      "learning_rate": 0.000490510278291327,
      "loss": 5.3585,
      "mean_token_accuracy": 0.1682575449347496,
      "num_tokens": 23110415.0,
      "step": 11135
    },
    {
      "entropy": 5.54347448348999,
      "epoch": 1.010522496371553,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004905009782093318,
      "loss": 5.3396,
      "mean_token_accuracy": 0.1727476641535759,
      "num_tokens": 23120336.0,
      "step": 11140
    },
    {
      "entropy": 5.569380283355713,
      "epoch": 1.0109760522496372,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004904916736707214,
      "loss": 5.4353,
      "mean_token_accuracy": 0.1605848491191864,
      "num_tokens": 23130568.0,
      "step": 11145
    },
    {
      "entropy": 5.645925760269165,
      "epoch": 1.0114296081277214,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004904823646756885,
      "loss": 5.384,
      "mean_token_accuracy": 0.16973509043455123,
      "num_tokens": 23140080.0,
      "step": 11150
    },
    {
      "entropy": 5.605403327941895,
      "epoch": 1.0118831640058055,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004904730512244252,
      "loss": 5.3285,
      "mean_token_accuracy": 0.17408903390169145,
      "num_tokens": 23150360.0,
      "step": 11155
    },
    {
      "entropy": 5.547512102127075,
      "epoch": 1.0123367198838897,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004904637333171243,
      "loss": 5.3093,
      "mean_token_accuracy": 0.16765036731958388,
      "num_tokens": 23160620.0,
      "step": 11160
    },
    {
      "entropy": 5.625001668930054,
      "epoch": 1.0127902757619738,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004904544109539785,
      "loss": 5.4058,
      "mean_token_accuracy": 0.17393511682748794,
      "num_tokens": 23171920.0,
      "step": 11165
    },
    {
      "entropy": 5.6038806438446045,
      "epoch": 1.013243831640058,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004904450841351805,
      "loss": 5.3717,
      "mean_token_accuracy": 0.17274330407381058,
      "num_tokens": 23182929.0,
      "step": 11170
    },
    {
      "entropy": 5.6337761878967285,
      "epoch": 1.0136973875181423,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004904357528609236,
      "loss": 5.3278,
      "mean_token_accuracy": 0.1682947263121605,
      "num_tokens": 23193020.0,
      "step": 11175
    },
    {
      "entropy": 5.560843801498413,
      "epoch": 1.0141509433962264,
      "grad_norm": 2.375,
      "learning_rate": 0.0004904264171314,
      "loss": 5.363,
      "mean_token_accuracy": 0.17510967701673508,
      "num_tokens": 23204073.0,
      "step": 11180
    },
    {
      "entropy": 5.544686508178711,
      "epoch": 1.0146044992743106,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004904170769468035,
      "loss": 5.2702,
      "mean_token_accuracy": 0.16796393543481827,
      "num_tokens": 23214376.0,
      "step": 11185
    },
    {
      "entropy": 5.650951957702636,
      "epoch": 1.0150580551523947,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004904077323073268,
      "loss": 5.4104,
      "mean_token_accuracy": 0.16763273030519485,
      "num_tokens": 23224304.0,
      "step": 11190
    },
    {
      "entropy": 5.610902404785156,
      "epoch": 1.015511611030479,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004903983832131633,
      "loss": 5.4216,
      "mean_token_accuracy": 0.16713034361600876,
      "num_tokens": 23235324.0,
      "step": 11195
    },
    {
      "entropy": 5.605709981918335,
      "epoch": 1.0159651669085632,
      "grad_norm": 2.125,
      "learning_rate": 0.0004903890296645063,
      "loss": 5.4045,
      "mean_token_accuracy": 0.1695600464940071,
      "num_tokens": 23245774.0,
      "step": 11200
    },
    {
      "entropy": 5.634368467330932,
      "epoch": 1.0164187227866472,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004903796716615495,
      "loss": 5.3366,
      "mean_token_accuracy": 0.16800456941127778,
      "num_tokens": 23256274.0,
      "step": 11205
    },
    {
      "entropy": 5.590472078323364,
      "epoch": 1.0168722786647315,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004903703092044861,
      "loss": 5.3707,
      "mean_token_accuracy": 0.16843401342630387,
      "num_tokens": 23266304.0,
      "step": 11210
    },
    {
      "entropy": 5.500264310836792,
      "epoch": 1.0173258345428158,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004903609422935097,
      "loss": 5.2473,
      "mean_token_accuracy": 0.17742429822683334,
      "num_tokens": 23276552.0,
      "step": 11215
    },
    {
      "entropy": 5.577713441848755,
      "epoch": 1.0177793904208998,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004903515709288143,
      "loss": 5.387,
      "mean_token_accuracy": 0.16179130971431732,
      "num_tokens": 23287799.0,
      "step": 11220
    },
    {
      "entropy": 5.572709417343139,
      "epoch": 1.018232946298984,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004903421951105933,
      "loss": 5.2383,
      "mean_token_accuracy": 0.17750613689422606,
      "num_tokens": 23297958.0,
      "step": 11225
    },
    {
      "entropy": 5.593379259109497,
      "epoch": 1.0186865021770681,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004903328148390411,
      "loss": 5.418,
      "mean_token_accuracy": 0.16655783653259276,
      "num_tokens": 23307523.0,
      "step": 11230
    },
    {
      "entropy": 5.576770782470703,
      "epoch": 1.0191400580551524,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004903234301143513,
      "loss": 5.4078,
      "mean_token_accuracy": 0.17188440263271332,
      "num_tokens": 23317970.0,
      "step": 11235
    },
    {
      "entropy": 5.5847412109375,
      "epoch": 1.0195936139332367,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004903140409367181,
      "loss": 5.379,
      "mean_token_accuracy": 0.17394014745950698,
      "num_tokens": 23328612.0,
      "step": 11240
    },
    {
      "entropy": 5.572464323043823,
      "epoch": 1.0200471698113207,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004903046473063357,
      "loss": 5.3366,
      "mean_token_accuracy": 0.17184139490127565,
      "num_tokens": 23338104.0,
      "step": 11245
    },
    {
      "entropy": 5.6089866161346436,
      "epoch": 1.020500725689405,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004902952492233983,
      "loss": 5.2717,
      "mean_token_accuracy": 0.18489170372486113,
      "num_tokens": 23348105.0,
      "step": 11250
    },
    {
      "entropy": 5.556183958053589,
      "epoch": 1.020954281567489,
      "grad_norm": 2.90625,
      "learning_rate": 0.0004902858466881002,
      "loss": 5.2456,
      "mean_token_accuracy": 0.17444435209035875,
      "num_tokens": 23357150.0,
      "step": 11255
    },
    {
      "entropy": 5.600876045227051,
      "epoch": 1.0214078374455733,
      "grad_norm": 2.390625,
      "learning_rate": 0.000490276439700636,
      "loss": 5.422,
      "mean_token_accuracy": 0.17182350754737855,
      "num_tokens": 23367223.0,
      "step": 11260
    },
    {
      "entropy": 5.600956726074219,
      "epoch": 1.0218613933236576,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004902670282612002,
      "loss": 5.372,
      "mean_token_accuracy": 0.1688934974372387,
      "num_tokens": 23377664.0,
      "step": 11265
    },
    {
      "entropy": 5.52209243774414,
      "epoch": 1.0223149492017416,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004902576123699873,
      "loss": 5.2688,
      "mean_token_accuracy": 0.1766052410006523,
      "num_tokens": 23388057.0,
      "step": 11270
    },
    {
      "entropy": 5.466941595077515,
      "epoch": 1.0227685050798259,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004902481920271922,
      "loss": 5.218,
      "mean_token_accuracy": 0.17293768748641014,
      "num_tokens": 23397808.0,
      "step": 11275
    },
    {
      "entropy": 5.534582138061523,
      "epoch": 1.02322206095791,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004902387672330097,
      "loss": 5.3019,
      "mean_token_accuracy": 0.17146763652563096,
      "num_tokens": 23406979.0,
      "step": 11280
    },
    {
      "entropy": 5.599915313720703,
      "epoch": 1.0236756168359942,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004902293379876345,
      "loss": 5.3288,
      "mean_token_accuracy": 0.1721256121993065,
      "num_tokens": 23416545.0,
      "step": 11285
    },
    {
      "entropy": 5.600649118423462,
      "epoch": 1.0241291727140784,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004902199042912618,
      "loss": 5.369,
      "mean_token_accuracy": 0.16481490582227706,
      "num_tokens": 23427464.0,
      "step": 11290
    },
    {
      "entropy": 5.6426581859588625,
      "epoch": 1.0245827285921625,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004902104661440868,
      "loss": 5.3902,
      "mean_token_accuracy": 0.1715350404381752,
      "num_tokens": 23436690.0,
      "step": 11295
    },
    {
      "entropy": 5.587870264053345,
      "epoch": 1.0250362844702468,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004902010235463042,
      "loss": 5.3171,
      "mean_token_accuracy": 0.17325413078069687,
      "num_tokens": 23446701.0,
      "step": 11300
    },
    {
      "entropy": 5.644993925094605,
      "epoch": 1.025489840348331,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004901915764981098,
      "loss": 5.3975,
      "mean_token_accuracy": 0.16523970365524293,
      "num_tokens": 23456966.0,
      "step": 11305
    },
    {
      "entropy": 5.623632907867432,
      "epoch": 1.025943396226415,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004901821249996988,
      "loss": 5.3707,
      "mean_token_accuracy": 0.17261502742767335,
      "num_tokens": 23467300.0,
      "step": 11310
    },
    {
      "entropy": 5.639683961868286,
      "epoch": 1.0263969521044993,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004901726690512665,
      "loss": 5.4102,
      "mean_token_accuracy": 0.1675952449440956,
      "num_tokens": 23478457.0,
      "step": 11315
    },
    {
      "entropy": 5.6091742515563965,
      "epoch": 1.0268505079825834,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004901632086530087,
      "loss": 5.3229,
      "mean_token_accuracy": 0.1715937301516533,
      "num_tokens": 23488066.0,
      "step": 11320
    },
    {
      "entropy": 5.596414279937744,
      "epoch": 1.0273040638606676,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004901537438051207,
      "loss": 5.3757,
      "mean_token_accuracy": 0.1635369509458542,
      "num_tokens": 23498856.0,
      "step": 11325
    },
    {
      "entropy": 5.591539669036865,
      "epoch": 1.027757619738752,
      "grad_norm": 3.03125,
      "learning_rate": 0.0004901442745077986,
      "loss": 5.3552,
      "mean_token_accuracy": 0.1709854334592819,
      "num_tokens": 23509845.0,
      "step": 11330
    },
    {
      "entropy": 5.6809183120727536,
      "epoch": 1.028211175616836,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004901348007612381,
      "loss": 5.4858,
      "mean_token_accuracy": 0.16896604150533676,
      "num_tokens": 23520322.0,
      "step": 11335
    },
    {
      "entropy": 5.678867530822754,
      "epoch": 1.0286647314949202,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004901253225656351,
      "loss": 5.3826,
      "mean_token_accuracy": 0.17234744727611542,
      "num_tokens": 23530030.0,
      "step": 11340
    },
    {
      "entropy": 5.676992511749267,
      "epoch": 1.0291182873730043,
      "grad_norm": 16.125,
      "learning_rate": 0.0004901158399211855,
      "loss": 5.4445,
      "mean_token_accuracy": 0.17008337378501892,
      "num_tokens": 23541063.0,
      "step": 11345
    },
    {
      "entropy": 5.572519445419312,
      "epoch": 1.0295718432510885,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004901063528280856,
      "loss": 5.3659,
      "mean_token_accuracy": 0.16454305946826936,
      "num_tokens": 23551693.0,
      "step": 11350
    },
    {
      "entropy": 5.64924726486206,
      "epoch": 1.0300253991291728,
      "grad_norm": 3.078125,
      "learning_rate": 0.0004900968612865314,
      "loss": 5.3979,
      "mean_token_accuracy": 0.17524353563785552,
      "num_tokens": 23561758.0,
      "step": 11355
    },
    {
      "entropy": 5.573123121261597,
      "epoch": 1.0304789550072568,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004900873652967195,
      "loss": 5.3216,
      "mean_token_accuracy": 0.16829973012208937,
      "num_tokens": 23571912.0,
      "step": 11360
    },
    {
      "entropy": 5.592647314071655,
      "epoch": 1.030932510885341,
      "grad_norm": 3.59375,
      "learning_rate": 0.0004900778648588459,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1677442193031311,
      "num_tokens": 23582600.0,
      "step": 11365
    },
    {
      "entropy": 5.618772172927857,
      "epoch": 1.0313860667634251,
      "grad_norm": 4.0,
      "learning_rate": 0.0004900683599731073,
      "loss": 5.25,
      "mean_token_accuracy": 0.17848899364471435,
      "num_tokens": 23592761.0,
      "step": 11370
    },
    {
      "entropy": 5.710945415496826,
      "epoch": 1.0318396226415094,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004900588506397002,
      "loss": 5.4284,
      "mean_token_accuracy": 0.17176624089479448,
      "num_tokens": 23603119.0,
      "step": 11375
    },
    {
      "entropy": 5.533707857131958,
      "epoch": 1.0322931785195937,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004900493368588213,
      "loss": 5.3342,
      "mean_token_accuracy": 0.1732027918100357,
      "num_tokens": 23613916.0,
      "step": 11380
    },
    {
      "entropy": 5.600461626052857,
      "epoch": 1.0327467343976777,
      "grad_norm": 2.0,
      "learning_rate": 0.0004900398186306672,
      "loss": 5.4758,
      "mean_token_accuracy": 0.16654439717531205,
      "num_tokens": 23623956.0,
      "step": 11385
    },
    {
      "entropy": 5.6156196117401125,
      "epoch": 1.033200290275762,
      "grad_norm": 1.734375,
      "learning_rate": 0.000490030295955435,
      "loss": 5.3703,
      "mean_token_accuracy": 0.16991166919469833,
      "num_tokens": 23634666.0,
      "step": 11390
    },
    {
      "entropy": 5.598365688323975,
      "epoch": 1.0336538461538463,
      "grad_norm": 4.53125,
      "learning_rate": 0.0004900207688333215,
      "loss": 5.3274,
      "mean_token_accuracy": 0.16624072343111038,
      "num_tokens": 23645601.0,
      "step": 11395
    },
    {
      "entropy": 5.633153963088989,
      "epoch": 1.0341074020319303,
      "grad_norm": 5.84375,
      "learning_rate": 0.0004900112372645236,
      "loss": 5.3869,
      "mean_token_accuracy": 0.17123257666826247,
      "num_tokens": 23656339.0,
      "step": 11400
    },
    {
      "entropy": 5.604528951644897,
      "epoch": 1.0345609579100146,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004900017012492386,
      "loss": 5.383,
      "mean_token_accuracy": 0.17367076873779297,
      "num_tokens": 23667433.0,
      "step": 11405
    },
    {
      "entropy": 5.644150257110596,
      "epoch": 1.0350145137880986,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004899921607876634,
      "loss": 5.3984,
      "mean_token_accuracy": 0.16625724285840987,
      "num_tokens": 23678186.0,
      "step": 11410
    },
    {
      "entropy": 5.614169263839722,
      "epoch": 1.0354680696661829,
      "grad_norm": 2.890625,
      "learning_rate": 0.0004899826158799958,
      "loss": 5.3541,
      "mean_token_accuracy": 0.1678687408566475,
      "num_tokens": 23688317.0,
      "step": 11415
    },
    {
      "entropy": 5.572248554229736,
      "epoch": 1.0359216255442671,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004899730665264328,
      "loss": 5.2754,
      "mean_token_accuracy": 0.1699757918715477,
      "num_tokens": 23697752.0,
      "step": 11420
    },
    {
      "entropy": 5.4876162052154545,
      "epoch": 1.0363751814223512,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004899635127271721,
      "loss": 5.2807,
      "mean_token_accuracy": 0.1753618225455284,
      "num_tokens": 23707803.0,
      "step": 11425
    },
    {
      "entropy": 5.552979755401611,
      "epoch": 1.0368287373004355,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004899539544824111,
      "loss": 5.2867,
      "mean_token_accuracy": 0.17673207223415374,
      "num_tokens": 23717630.0,
      "step": 11430
    },
    {
      "entropy": 5.65933198928833,
      "epoch": 1.0372822931785195,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004899443917923476,
      "loss": 5.4196,
      "mean_token_accuracy": 0.16775544807314874,
      "num_tokens": 23728836.0,
      "step": 11435
    },
    {
      "entropy": 5.553336429595947,
      "epoch": 1.0377358490566038,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004899348246571792,
      "loss": 5.3869,
      "mean_token_accuracy": 0.16835874766111375,
      "num_tokens": 23740145.0,
      "step": 11440
    },
    {
      "entropy": 5.576502132415771,
      "epoch": 1.038189404934688,
      "grad_norm": 2.109375,
      "learning_rate": 0.000489925253077104,
      "loss": 5.3435,
      "mean_token_accuracy": 0.1710005670785904,
      "num_tokens": 23751162.0,
      "step": 11445
    },
    {
      "entropy": 5.638379001617432,
      "epoch": 1.038642960812772,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004899156770523196,
      "loss": 5.378,
      "mean_token_accuracy": 0.16817111074924468,
      "num_tokens": 23761634.0,
      "step": 11450
    },
    {
      "entropy": 5.633851337432861,
      "epoch": 1.0390965166908563,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004899060965830244,
      "loss": 5.3457,
      "mean_token_accuracy": 0.17330029457807541,
      "num_tokens": 23772483.0,
      "step": 11455
    },
    {
      "entropy": 5.59468264579773,
      "epoch": 1.0395500725689404,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004898965116694162,
      "loss": 5.3102,
      "mean_token_accuracy": 0.1680698201060295,
      "num_tokens": 23782298.0,
      "step": 11460
    },
    {
      "entropy": 5.569092273712158,
      "epoch": 1.0400036284470247,
      "grad_norm": 3.875,
      "learning_rate": 0.0004898869223116935,
      "loss": 5.374,
      "mean_token_accuracy": 0.16889462172985076,
      "num_tokens": 23792492.0,
      "step": 11465
    },
    {
      "entropy": 5.699532079696655,
      "epoch": 1.040457184325109,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004898773285100546,
      "loss": 5.5137,
      "mean_token_accuracy": 0.1637313649058342,
      "num_tokens": 23802999.0,
      "step": 11470
    },
    {
      "entropy": 5.603556537628174,
      "epoch": 1.040910740203193,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004898677302646975,
      "loss": 5.3201,
      "mean_token_accuracy": 0.1761540174484253,
      "num_tokens": 23812541.0,
      "step": 11475
    },
    {
      "entropy": 5.5463886737823485,
      "epoch": 1.0413642960812772,
      "grad_norm": 2.25,
      "learning_rate": 0.0004898581275758211,
      "loss": 5.434,
      "mean_token_accuracy": 0.16390567123889924,
      "num_tokens": 23823244.0,
      "step": 11480
    },
    {
      "entropy": 5.54480562210083,
      "epoch": 1.0418178519593615,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004898485204436239,
      "loss": 5.2795,
      "mean_token_accuracy": 0.17889072000980377,
      "num_tokens": 23832346.0,
      "step": 11485
    },
    {
      "entropy": 5.650099802017212,
      "epoch": 1.0422714078374455,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004898389088683047,
      "loss": 5.4214,
      "mean_token_accuracy": 0.17537481784820558,
      "num_tokens": 23842554.0,
      "step": 11490
    },
    {
      "entropy": 5.580894851684571,
      "epoch": 1.0427249637155298,
      "grad_norm": 3.890625,
      "learning_rate": 0.0004898292928500619,
      "loss": 5.2566,
      "mean_token_accuracy": 0.17736405432224273,
      "num_tokens": 23852938.0,
      "step": 11495
    },
    {
      "entropy": 5.598966026306153,
      "epoch": 1.0431785195936139,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004898196723890947,
      "loss": 5.2988,
      "mean_token_accuracy": 0.17443305552005767,
      "num_tokens": 23862367.0,
      "step": 11500
    },
    {
      "entropy": 5.558216524124146,
      "epoch": 1.0436320754716981,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004898100474856019,
      "loss": 5.389,
      "mean_token_accuracy": 0.17014889568090438,
      "num_tokens": 23873249.0,
      "step": 11505
    },
    {
      "entropy": 5.5902259349823,
      "epoch": 1.0440856313497824,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004898004181397824,
      "loss": 5.3899,
      "mean_token_accuracy": 0.16259795129299165,
      "num_tokens": 23883830.0,
      "step": 11510
    },
    {
      "entropy": 5.622036981582641,
      "epoch": 1.0445391872278664,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004897907843518358,
      "loss": 5.321,
      "mean_token_accuracy": 0.17028684467077254,
      "num_tokens": 23894359.0,
      "step": 11515
    },
    {
      "entropy": 5.569010543823242,
      "epoch": 1.0449927431059507,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004897811461219609,
      "loss": 5.3212,
      "mean_token_accuracy": 0.17068181782960892,
      "num_tokens": 23903779.0,
      "step": 11520
    },
    {
      "entropy": 5.6016926765441895,
      "epoch": 1.0454462989840347,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004897715034503571,
      "loss": 5.4009,
      "mean_token_accuracy": 0.16589456498622895,
      "num_tokens": 23914523.0,
      "step": 11525
    },
    {
      "entropy": 5.55782413482666,
      "epoch": 1.045899854862119,
      "grad_norm": 2.40625,
      "learning_rate": 0.000489761856337224,
      "loss": 5.371,
      "mean_token_accuracy": 0.17274439632892608,
      "num_tokens": 23924996.0,
      "step": 11530
    },
    {
      "entropy": 5.594377183914185,
      "epoch": 1.0463534107402033,
      "grad_norm": 2.625,
      "learning_rate": 0.0004897522047827609,
      "loss": 5.3201,
      "mean_token_accuracy": 0.1750901833176613,
      "num_tokens": 23935843.0,
      "step": 11535
    },
    {
      "entropy": 5.613669109344483,
      "epoch": 1.0468069666182873,
      "grad_norm": 3.8125,
      "learning_rate": 0.0004897425487871676,
      "loss": 5.3105,
      "mean_token_accuracy": 0.1711737558245659,
      "num_tokens": 23945044.0,
      "step": 11540
    },
    {
      "entropy": 5.5151208400726315,
      "epoch": 1.0472605224963716,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004897328883506436,
      "loss": 5.3485,
      "mean_token_accuracy": 0.16891540288925172,
      "num_tokens": 23955933.0,
      "step": 11545
    },
    {
      "entropy": 5.636126756668091,
      "epoch": 1.0477140783744556,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004897232234733887,
      "loss": 5.4154,
      "mean_token_accuracy": 0.16608502417802812,
      "num_tokens": 23966320.0,
      "step": 11550
    },
    {
      "entropy": 5.627421760559082,
      "epoch": 1.04816763425254,
      "grad_norm": 2.25,
      "learning_rate": 0.0004897135541556028,
      "loss": 5.3217,
      "mean_token_accuracy": 0.17145986109972,
      "num_tokens": 23976233.0,
      "step": 11555
    },
    {
      "entropy": 5.628482675552368,
      "epoch": 1.0486211901306242,
      "grad_norm": 2.734375,
      "learning_rate": 0.000489703880397486,
      "loss": 5.3832,
      "mean_token_accuracy": 0.17113422602415085,
      "num_tokens": 23987854.0,
      "step": 11560
    },
    {
      "entropy": 5.56703929901123,
      "epoch": 1.0490747460087082,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004896942021992382,
      "loss": 5.3753,
      "mean_token_accuracy": 0.1662612795829773,
      "num_tokens": 23997973.0,
      "step": 11565
    },
    {
      "entropy": 5.584162759780884,
      "epoch": 1.0495283018867925,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004896845195610595,
      "loss": 5.379,
      "mean_token_accuracy": 0.1711257964372635,
      "num_tokens": 24007313.0,
      "step": 11570
    },
    {
      "entropy": 5.626845645904541,
      "epoch": 1.0499818577648767,
      "grad_norm": 3.59375,
      "learning_rate": 0.0004896748324831504,
      "loss": 5.4163,
      "mean_token_accuracy": 0.1633000910282135,
      "num_tokens": 24018254.0,
      "step": 11575
    },
    {
      "entropy": 5.651694965362549,
      "epoch": 1.0504354136429608,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004896651409657109,
      "loss": 5.406,
      "mean_token_accuracy": 0.16983163505792617,
      "num_tokens": 24029799.0,
      "step": 11580
    },
    {
      "entropy": 5.550058126449585,
      "epoch": 1.050888969521045,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004896554450089417,
      "loss": 5.2925,
      "mean_token_accuracy": 0.17730078995227813,
      "num_tokens": 24040972.0,
      "step": 11585
    },
    {
      "entropy": 5.540880584716797,
      "epoch": 1.051342525399129,
      "grad_norm": 2.5,
      "learning_rate": 0.0004896457446130431,
      "loss": 5.2262,
      "mean_token_accuracy": 0.17580059617757798,
      "num_tokens": 24050781.0,
      "step": 11590
    },
    {
      "entropy": 5.529133701324463,
      "epoch": 1.0517960812772134,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004896360397782159,
      "loss": 5.3523,
      "mean_token_accuracy": 0.16468161195516587,
      "num_tokens": 24060184.0,
      "step": 11595
    },
    {
      "entropy": 5.647056913375854,
      "epoch": 1.0522496371552976,
      "grad_norm": 3.34375,
      "learning_rate": 0.0004896263305046607,
      "loss": 5.4138,
      "mean_token_accuracy": 0.1736113965511322,
      "num_tokens": 24070777.0,
      "step": 11600
    },
    {
      "entropy": 5.654244089126587,
      "epoch": 1.0527031930333817,
      "grad_norm": 3.921875,
      "learning_rate": 0.0004896166167925782,
      "loss": 5.4383,
      "mean_token_accuracy": 0.16724486500024796,
      "num_tokens": 24080878.0,
      "step": 11605
    },
    {
      "entropy": 5.6524169921875,
      "epoch": 1.053156748911466,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004896068986421693,
      "loss": 5.3879,
      "mean_token_accuracy": 0.17224199324846268,
      "num_tokens": 24090038.0,
      "step": 11610
    },
    {
      "entropy": 5.6196681499481205,
      "epoch": 1.05361030478955,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004895971760536351,
      "loss": 5.4199,
      "mean_token_accuracy": 0.17333686649799346,
      "num_tokens": 24100425.0,
      "step": 11615
    },
    {
      "entropy": 5.627913570404052,
      "epoch": 1.0540638606676342,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004895874490271768,
      "loss": 5.4929,
      "mean_token_accuracy": 0.16355858445167543,
      "num_tokens": 24112182.0,
      "step": 11620
    },
    {
      "entropy": 5.603630542755127,
      "epoch": 1.0545174165457185,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004895777175629952,
      "loss": 5.3344,
      "mean_token_accuracy": 0.17166843712329866,
      "num_tokens": 24122182.0,
      "step": 11625
    },
    {
      "entropy": 5.614436769485474,
      "epoch": 1.0549709724238026,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004895679816612917,
      "loss": 5.3341,
      "mean_token_accuracy": 0.17125269323587416,
      "num_tokens": 24131930.0,
      "step": 11630
    },
    {
      "entropy": 5.65815486907959,
      "epoch": 1.0554245283018868,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004895582413222678,
      "loss": 5.338,
      "mean_token_accuracy": 0.18034985512495041,
      "num_tokens": 24141575.0,
      "step": 11635
    },
    {
      "entropy": 5.679404544830322,
      "epoch": 1.0558780841799709,
      "grad_norm": 2.0,
      "learning_rate": 0.0004895484965461247,
      "loss": 5.4652,
      "mean_token_accuracy": 0.16713959127664565,
      "num_tokens": 24151720.0,
      "step": 11640
    },
    {
      "entropy": 5.5869152545928955,
      "epoch": 1.0563316400580551,
      "grad_norm": 2.46875,
      "learning_rate": 0.000489538747333064,
      "loss": 5.3863,
      "mean_token_accuracy": 0.17306092977523804,
      "num_tokens": 24161881.0,
      "step": 11645
    },
    {
      "entropy": 5.538717746734619,
      "epoch": 1.0567851959361394,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004895289936832874,
      "loss": 5.3187,
      "mean_token_accuracy": 0.17443083673715593,
      "num_tokens": 24171474.0,
      "step": 11650
    },
    {
      "entropy": 5.6506819248199465,
      "epoch": 1.0572387518142234,
      "grad_norm": 2.953125,
      "learning_rate": 0.0004895192355969966,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1692167952656746,
      "num_tokens": 24182186.0,
      "step": 11655
    },
    {
      "entropy": 5.636696863174438,
      "epoch": 1.0576923076923077,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004895094730743932,
      "loss": 5.3517,
      "mean_token_accuracy": 0.17387636750936508,
      "num_tokens": 24192318.0,
      "step": 11660
    },
    {
      "entropy": 5.633651304244995,
      "epoch": 1.058145863570392,
      "grad_norm": 2.875,
      "learning_rate": 0.0004894997061156793,
      "loss": 5.4218,
      "mean_token_accuracy": 0.1694432780146599,
      "num_tokens": 24202918.0,
      "step": 11665
    },
    {
      "entropy": 5.5955101490020756,
      "epoch": 1.058599419448476,
      "grad_norm": 2.75,
      "learning_rate": 0.0004894899347210567,
      "loss": 5.3118,
      "mean_token_accuracy": 0.17389714866876602,
      "num_tokens": 24213218.0,
      "step": 11670
    },
    {
      "entropy": 5.598051071166992,
      "epoch": 1.0590529753265603,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004894801588907278,
      "loss": 5.3139,
      "mean_token_accuracy": 0.17599815875291824,
      "num_tokens": 24223229.0,
      "step": 11675
    },
    {
      "entropy": 5.652417850494385,
      "epoch": 1.0595065312046443,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004894703786248944,
      "loss": 5.3791,
      "mean_token_accuracy": 0.1630527600646019,
      "num_tokens": 24233637.0,
      "step": 11680
    },
    {
      "entropy": 5.634524631500244,
      "epoch": 1.0599600870827286,
      "grad_norm": 2.4375,
      "learning_rate": 0.000489460593923759,
      "loss": 5.4139,
      "mean_token_accuracy": 0.17403268963098525,
      "num_tokens": 24244251.0,
      "step": 11685
    },
    {
      "entropy": 5.592057085037231,
      "epoch": 1.0604136429608129,
      "grad_norm": 3.25,
      "learning_rate": 0.0004894508047875237,
      "loss": 5.3287,
      "mean_token_accuracy": 0.1699887201189995,
      "num_tokens": 24254380.0,
      "step": 11690
    },
    {
      "entropy": 5.460594224929809,
      "epoch": 1.060867198838897,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004894410112163913,
      "loss": 5.2479,
      "mean_token_accuracy": 0.1714046984910965,
      "num_tokens": 24264841.0,
      "step": 11695
    },
    {
      "entropy": 5.500969696044922,
      "epoch": 1.0613207547169812,
      "grad_norm": 2.875,
      "learning_rate": 0.0004894312132105641,
      "loss": 5.2987,
      "mean_token_accuracy": 0.17383984178304673,
      "num_tokens": 24274383.0,
      "step": 11700
    },
    {
      "entropy": 5.636331558227539,
      "epoch": 1.0617743105950652,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004894214107702447,
      "loss": 5.3576,
      "mean_token_accuracy": 0.16789450645446777,
      "num_tokens": 24284140.0,
      "step": 11705
    },
    {
      "entropy": 5.7018595218658445,
      "epoch": 1.0622278664731495,
      "grad_norm": 3.046875,
      "learning_rate": 0.000489411603895636,
      "loss": 5.4674,
      "mean_token_accuracy": 0.1698828309774399,
      "num_tokens": 24295044.0,
      "step": 11710
    },
    {
      "entropy": 5.641237354278564,
      "epoch": 1.0626814223512338,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004894017925869406,
      "loss": 5.3514,
      "mean_token_accuracy": 0.17169802039861679,
      "num_tokens": 24305080.0,
      "step": 11715
    },
    {
      "entropy": 5.598285913467407,
      "epoch": 1.0631349782293178,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004893919768443615,
      "loss": 5.3968,
      "mean_token_accuracy": 0.17039976567029952,
      "num_tokens": 24316306.0,
      "step": 11720
    },
    {
      "entropy": 5.482102537155152,
      "epoch": 1.063588534107402,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004893821566681017,
      "loss": 5.3808,
      "mean_token_accuracy": 0.17302751988172532,
      "num_tokens": 24326740.0,
      "step": 11725
    },
    {
      "entropy": 5.577399110794067,
      "epoch": 1.0640420899854863,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004893723320583644,
      "loss": 5.3968,
      "mean_token_accuracy": 0.1635014906525612,
      "num_tokens": 24337621.0,
      "step": 11730
    },
    {
      "entropy": 5.549607038497925,
      "epoch": 1.0644956458635704,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004893625030153525,
      "loss": 5.2499,
      "mean_token_accuracy": 0.17473862320184708,
      "num_tokens": 24347732.0,
      "step": 11735
    },
    {
      "entropy": 5.5875303745269775,
      "epoch": 1.0649492017416546,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004893526695392694,
      "loss": 5.3078,
      "mean_token_accuracy": 0.18272399604320527,
      "num_tokens": 24359173.0,
      "step": 11740
    },
    {
      "entropy": 5.5956933975219725,
      "epoch": 1.0654027576197387,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004893428316303185,
      "loss": 5.2961,
      "mean_token_accuracy": 0.17074852734804152,
      "num_tokens": 24369465.0,
      "step": 11745
    },
    {
      "entropy": 5.559246158599853,
      "epoch": 1.065856313497823,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004893329892887031,
      "loss": 5.3271,
      "mean_token_accuracy": 0.17232367992401124,
      "num_tokens": 24378899.0,
      "step": 11750
    },
    {
      "entropy": 5.540904426574707,
      "epoch": 1.0663098693759072,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004893231425146271,
      "loss": 5.4532,
      "mean_token_accuracy": 0.165650475025177,
      "num_tokens": 24388919.0,
      "step": 11755
    },
    {
      "entropy": 5.540589141845703,
      "epoch": 1.0667634252539913,
      "grad_norm": 2.65625,
      "learning_rate": 0.0004893132913082937,
      "loss": 5.3144,
      "mean_token_accuracy": 0.1708405449986458,
      "num_tokens": 24399632.0,
      "step": 11760
    },
    {
      "entropy": 5.621877193450928,
      "epoch": 1.0672169811320755,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004893034356699069,
      "loss": 5.3509,
      "mean_token_accuracy": 0.17433714866638184,
      "num_tokens": 24410817.0,
      "step": 11765
    },
    {
      "entropy": 5.619347333908081,
      "epoch": 1.0676705370101596,
      "grad_norm": 5.28125,
      "learning_rate": 0.0004892935755996705,
      "loss": 5.3632,
      "mean_token_accuracy": 0.17336190938949586,
      "num_tokens": 24421058.0,
      "step": 11770
    },
    {
      "entropy": 5.566294860839844,
      "epoch": 1.0681240928882438,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004892837110977883,
      "loss": 5.2779,
      "mean_token_accuracy": 0.17225192338228226,
      "num_tokens": 24431522.0,
      "step": 11775
    },
    {
      "entropy": 5.544850015640259,
      "epoch": 1.068577648766328,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004892738421644644,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1740766227245331,
      "num_tokens": 24441746.0,
      "step": 11780
    },
    {
      "entropy": 5.505834913253784,
      "epoch": 1.0690312046444121,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004892639687999027,
      "loss": 5.2143,
      "mean_token_accuracy": 0.17782359719276428,
      "num_tokens": 24451916.0,
      "step": 11785
    },
    {
      "entropy": 5.546728324890137,
      "epoch": 1.0694847605224964,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004892540910043076,
      "loss": 5.3487,
      "mean_token_accuracy": 0.1773233100771904,
      "num_tokens": 24461935.0,
      "step": 11790
    },
    {
      "entropy": 5.585660028457641,
      "epoch": 1.0699383164005805,
      "grad_norm": 2.625,
      "learning_rate": 0.0004892442087778833,
      "loss": 5.2718,
      "mean_token_accuracy": 0.17202629297971725,
      "num_tokens": 24472437.0,
      "step": 11795
    },
    {
      "entropy": 5.589559173583984,
      "epoch": 1.0703918722786647,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004892343221208342,
      "loss": 5.3387,
      "mean_token_accuracy": 0.1752584308385849,
      "num_tokens": 24481726.0,
      "step": 11800
    },
    {
      "entropy": 5.60119514465332,
      "epoch": 1.070845428156749,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004892244310333646,
      "loss": 5.4531,
      "mean_token_accuracy": 0.1676913768053055,
      "num_tokens": 24492340.0,
      "step": 11805
    },
    {
      "entropy": 5.546935272216797,
      "epoch": 1.071298984034833,
      "grad_norm": 3.296875,
      "learning_rate": 0.0004892145355156793,
      "loss": 5.3146,
      "mean_token_accuracy": 0.1705259934067726,
      "num_tokens": 24502138.0,
      "step": 11810
    },
    {
      "entropy": 5.560450696945191,
      "epoch": 1.0717525399129173,
      "grad_norm": 2.625,
      "learning_rate": 0.0004892046355679829,
      "loss": 5.3683,
      "mean_token_accuracy": 0.1682853877544403,
      "num_tokens": 24512723.0,
      "step": 11815
    },
    {
      "entropy": 5.604918575286865,
      "epoch": 1.0722060957910013,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004891947311904799,
      "loss": 5.4154,
      "mean_token_accuracy": 0.16189062148332595,
      "num_tokens": 24523766.0,
      "step": 11820
    },
    {
      "entropy": 5.623622941970825,
      "epoch": 1.0726596516690856,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004891848223833753,
      "loss": 5.3718,
      "mean_token_accuracy": 0.17246003299951554,
      "num_tokens": 24534571.0,
      "step": 11825
    },
    {
      "entropy": 5.653265762329101,
      "epoch": 1.0731132075471699,
      "grad_norm": 1.9375,
      "learning_rate": 0.000489174909146874,
      "loss": 5.3589,
      "mean_token_accuracy": 0.17910315543413163,
      "num_tokens": 24545265.0,
      "step": 11830
    },
    {
      "entropy": 5.6582252979278564,
      "epoch": 1.073566763425254,
      "grad_norm": 5.90625,
      "learning_rate": 0.0004891649914811811,
      "loss": 5.4295,
      "mean_token_accuracy": 0.16662697345018387,
      "num_tokens": 24555524.0,
      "step": 11835
    },
    {
      "entropy": 5.51617956161499,
      "epoch": 1.0740203193033382,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004891550693865014,
      "loss": 5.3135,
      "mean_token_accuracy": 0.17839529514312744,
      "num_tokens": 24565420.0,
      "step": 11840
    },
    {
      "entropy": 5.609707641601562,
      "epoch": 1.0744738751814222,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004891451428630405,
      "loss": 5.366,
      "mean_token_accuracy": 0.1649378016591072,
      "num_tokens": 24576794.0,
      "step": 11845
    },
    {
      "entropy": 5.624512720108032,
      "epoch": 1.0749274310595065,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004891352119110033,
      "loss": 5.3623,
      "mean_token_accuracy": 0.17520020753145218,
      "num_tokens": 24586320.0,
      "step": 11850
    },
    {
      "entropy": 5.593661451339722,
      "epoch": 1.0753809869375908,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004891252765305954,
      "loss": 5.3464,
      "mean_token_accuracy": 0.17566470056772232,
      "num_tokens": 24596524.0,
      "step": 11855
    },
    {
      "entropy": 5.621684932708741,
      "epoch": 1.0758345428156748,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004891153367220222,
      "loss": 5.2914,
      "mean_token_accuracy": 0.17703112959861755,
      "num_tokens": 24605877.0,
      "step": 11860
    },
    {
      "entropy": 5.645792245864868,
      "epoch": 1.076288098693759,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004891053924854894,
      "loss": 5.4918,
      "mean_token_accuracy": 0.1607166647911072,
      "num_tokens": 24616719.0,
      "step": 11865
    },
    {
      "entropy": 5.580338191986084,
      "epoch": 1.0767416545718433,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004890954438212024,
      "loss": 5.2589,
      "mean_token_accuracy": 0.17941452264785768,
      "num_tokens": 24626882.0,
      "step": 11870
    },
    {
      "entropy": 5.61026782989502,
      "epoch": 1.0771952104499274,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004890854907293671,
      "loss": 5.3584,
      "mean_token_accuracy": 0.16531697511672974,
      "num_tokens": 24637884.0,
      "step": 11875
    },
    {
      "entropy": 5.553039932250977,
      "epoch": 1.0776487663280117,
      "grad_norm": 2.6875,
      "learning_rate": 0.0004890755332101893,
      "loss": 5.3509,
      "mean_token_accuracy": 0.16965146064758302,
      "num_tokens": 24647389.0,
      "step": 11880
    },
    {
      "entropy": 5.574040079116822,
      "epoch": 1.0781023222060957,
      "grad_norm": 3.140625,
      "learning_rate": 0.0004890655712638749,
      "loss": 5.3804,
      "mean_token_accuracy": 0.16506000161170958,
      "num_tokens": 24657595.0,
      "step": 11885
    },
    {
      "entropy": 5.613958406448364,
      "epoch": 1.07855587808418,
      "grad_norm": 2.28125,
      "learning_rate": 0.00048905560489063,
      "loss": 5.2976,
      "mean_token_accuracy": 0.17296027392148972,
      "num_tokens": 24667683.0,
      "step": 11890
    },
    {
      "entropy": 5.658120727539062,
      "epoch": 1.0790094339622642,
      "grad_norm": 2.375,
      "learning_rate": 0.0004890456340906605,
      "loss": 5.4328,
      "mean_token_accuracy": 0.16448845267295836,
      "num_tokens": 24678041.0,
      "step": 11895
    },
    {
      "entropy": 5.5773170471191404,
      "epoch": 1.0794629898403483,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004890356588641729,
      "loss": 5.4171,
      "mean_token_accuracy": 0.17782746106386185,
      "num_tokens": 24688569.0,
      "step": 11900
    },
    {
      "entropy": 5.594895219802856,
      "epoch": 1.0799165457184325,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004890256792113733,
      "loss": 5.3854,
      "mean_token_accuracy": 0.1775667577981949,
      "num_tokens": 24699913.0,
      "step": 11905
    },
    {
      "entropy": 5.661224460601806,
      "epoch": 1.0803701015965168,
      "grad_norm": 2.5,
      "learning_rate": 0.0004890156951324681,
      "loss": 5.3877,
      "mean_token_accuracy": 0.17179592847824096,
      "num_tokens": 24710682.0,
      "step": 11910
    },
    {
      "entropy": 5.72910704612732,
      "epoch": 1.0808236574746009,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004890057066276636,
      "loss": 5.4722,
      "mean_token_accuracy": 0.1660769134759903,
      "num_tokens": 24721087.0,
      "step": 11915
    },
    {
      "entropy": 5.5810271263122555,
      "epoch": 1.0812772133526851,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004889957136971668,
      "loss": 5.3203,
      "mean_token_accuracy": 0.17739502489566802,
      "num_tokens": 24731323.0,
      "step": 11920
    },
    {
      "entropy": 5.583218431472778,
      "epoch": 1.0817307692307692,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000488985716341184,
      "loss": 5.3736,
      "mean_token_accuracy": 0.16614993065595626,
      "num_tokens": 24742559.0,
      "step": 11925
    },
    {
      "entropy": 5.639229249954224,
      "epoch": 1.0821843251088534,
      "grad_norm": 2.921875,
      "learning_rate": 0.0004889757145599222,
      "loss": 5.3668,
      "mean_token_accuracy": 0.17185461968183519,
      "num_tokens": 24752307.0,
      "step": 11930
    },
    {
      "entropy": 5.526152944564819,
      "epoch": 1.0826378809869377,
      "grad_norm": 2.078125,
      "learning_rate": 0.000488965708353588,
      "loss": 5.2896,
      "mean_token_accuracy": 0.16761810183525086,
      "num_tokens": 24762792.0,
      "step": 11935
    },
    {
      "entropy": 5.657108974456787,
      "epoch": 1.0830914368650217,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004889556977223883,
      "loss": 5.4208,
      "mean_token_accuracy": 0.16339073926210404,
      "num_tokens": 24774377.0,
      "step": 11940
    },
    {
      "entropy": 5.6454047679901125,
      "epoch": 1.083544992743106,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004889456826665304,
      "loss": 5.3641,
      "mean_token_accuracy": 0.17016541510820388,
      "num_tokens": 24785919.0,
      "step": 11945
    },
    {
      "entropy": 5.626445770263672,
      "epoch": 1.08399854862119,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004889356631862211,
      "loss": 5.3822,
      "mean_token_accuracy": 0.1718483343720436,
      "num_tokens": 24796594.0,
      "step": 11950
    },
    {
      "entropy": 5.576580619812011,
      "epoch": 1.0844521044992743,
      "grad_norm": 2.34375,
      "learning_rate": 0.000488925639281668,
      "loss": 5.3857,
      "mean_token_accuracy": 0.17252150028944016,
      "num_tokens": 24807322.0,
      "step": 11955
    },
    {
      "entropy": 5.555296468734741,
      "epoch": 1.0849056603773586,
      "grad_norm": 2.6875,
      "learning_rate": 0.0004889156109530781,
      "loss": 5.276,
      "mean_token_accuracy": 0.1731458216905594,
      "num_tokens": 24817593.0,
      "step": 11960
    },
    {
      "entropy": 5.590374898910523,
      "epoch": 1.0853592162554426,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004889055782006588,
      "loss": 5.2842,
      "mean_token_accuracy": 0.17565004229545594,
      "num_tokens": 24827577.0,
      "step": 11965
    },
    {
      "entropy": 5.574050760269165,
      "epoch": 1.085812772133527,
      "grad_norm": 3.015625,
      "learning_rate": 0.0004888955410246176,
      "loss": 5.3387,
      "mean_token_accuracy": 0.17546096444129944,
      "num_tokens": 24838148.0,
      "step": 11970
    },
    {
      "entropy": 5.567724370956421,
      "epoch": 1.086266328011611,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004888854994251621,
      "loss": 5.2811,
      "mean_token_accuracy": 0.17895624935626983,
      "num_tokens": 24848451.0,
      "step": 11975
    },
    {
      "entropy": 5.546176195144653,
      "epoch": 1.0867198838896952,
      "grad_norm": 2.734375,
      "learning_rate": 0.0004888754534025,
      "loss": 5.4116,
      "mean_token_accuracy": 0.1617180287837982,
      "num_tokens": 24859619.0,
      "step": 11980
    },
    {
      "entropy": 5.605387496948242,
      "epoch": 1.0871734397677795,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004888654029568391,
      "loss": 5.4454,
      "mean_token_accuracy": 0.1611100822687149,
      "num_tokens": 24869863.0,
      "step": 11985
    },
    {
      "entropy": 5.651314067840576,
      "epoch": 1.0876269956458635,
      "grad_norm": 3.1875,
      "learning_rate": 0.0004888553480883871,
      "loss": 5.3644,
      "mean_token_accuracy": 0.16844964623451233,
      "num_tokens": 24880524.0,
      "step": 11990
    },
    {
      "entropy": 5.573370265960693,
      "epoch": 1.0880805515239478,
      "grad_norm": 4.3125,
      "learning_rate": 0.000488845288797352,
      "loss": 5.3276,
      "mean_token_accuracy": 0.17795932441949844,
      "num_tokens": 24891514.0,
      "step": 11995
    },
    {
      "entropy": 5.537796258926392,
      "epoch": 1.0885341074020318,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004888352250839418,
      "loss": 5.3006,
      "mean_token_accuracy": 0.18009179681539536,
      "num_tokens": 24900842.0,
      "step": 12000
    },
    {
      "epoch": 1.0885341074020318,
      "eval_entropy": 5.3571358324715455,
      "eval_loss": 5.452709674835205,
      "eval_mean_token_accuracy": 0.17524898895715696,
      "eval_num_tokens": 24900842.0,
      "eval_runtime": 26.4984,
      "eval_samples_per_second": 1334.418,
      "eval_steps_per_second": 166.802,
      "step": 12000
    },
    {
      "entropy": 5.507623529434204,
      "epoch": 1.088987663280116,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004888251569483647,
      "loss": 5.2846,
      "mean_token_accuracy": 0.1762351632118225,
      "num_tokens": 24911845.0,
      "step": 12005
    },
    {
      "entropy": 5.526329374313354,
      "epoch": 1.0894412191582004,
      "grad_norm": 1.875,
      "learning_rate": 0.0004888150843908288,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1737929806113243,
      "num_tokens": 24922211.0,
      "step": 12010
    },
    {
      "entropy": 5.6169493198394775,
      "epoch": 1.0898947750362844,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004888050074115426,
      "loss": 5.4176,
      "mean_token_accuracy": 0.16657098829746247,
      "num_tokens": 24932129.0,
      "step": 12015
    },
    {
      "entropy": 5.646394968032837,
      "epoch": 1.0903483309143687,
      "grad_norm": 3.015625,
      "learning_rate": 0.0004887949260107143,
      "loss": 5.4192,
      "mean_token_accuracy": 0.16553286761045455,
      "num_tokens": 24942869.0,
      "step": 12020
    },
    {
      "entropy": 5.553748750686646,
      "epoch": 1.0908018867924527,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004887848401885524,
      "loss": 5.3774,
      "mean_token_accuracy": 0.1705244481563568,
      "num_tokens": 24952621.0,
      "step": 12025
    },
    {
      "entropy": 5.690526676177979,
      "epoch": 1.091255442670537,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004887747499452656,
      "loss": 5.4618,
      "mean_token_accuracy": 0.15994815826416015,
      "num_tokens": 24963587.0,
      "step": 12030
    },
    {
      "entropy": 5.621628618240356,
      "epoch": 1.0917089985486212,
      "grad_norm": 6.0625,
      "learning_rate": 0.0004887646552810624,
      "loss": 5.3065,
      "mean_token_accuracy": 0.17487196773290634,
      "num_tokens": 24974122.0,
      "step": 12035
    },
    {
      "entropy": 5.610294485092163,
      "epoch": 1.0921625544267053,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004887545561961517,
      "loss": 5.3965,
      "mean_token_accuracy": 0.175234654545784,
      "num_tokens": 24985145.0,
      "step": 12040
    },
    {
      "entropy": 5.546441316604614,
      "epoch": 1.0926161103047896,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004887444526907424,
      "loss": 5.3249,
      "mean_token_accuracy": 0.17132711559534072,
      "num_tokens": 24995511.0,
      "step": 12045
    },
    {
      "entropy": 5.580210113525391,
      "epoch": 1.0930696661828738,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004887343447650432,
      "loss": 5.3329,
      "mean_token_accuracy": 0.16668287366628648,
      "num_tokens": 25006193.0,
      "step": 12050
    },
    {
      "entropy": 5.643644237518311,
      "epoch": 1.0935232220609579,
      "grad_norm": 2.734375,
      "learning_rate": 0.0004887242324192634,
      "loss": 5.3783,
      "mean_token_accuracy": 0.17420458793640137,
      "num_tokens": 25016444.0,
      "step": 12055
    },
    {
      "entropy": 5.66413426399231,
      "epoch": 1.0939767779390421,
      "grad_norm": 2.828125,
      "learning_rate": 0.0004887141156536119,
      "loss": 5.3817,
      "mean_token_accuracy": 0.16439399123191833,
      "num_tokens": 25027898.0,
      "step": 12060
    },
    {
      "entropy": 5.569183969497681,
      "epoch": 1.0944303338171262,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004887039944682981,
      "loss": 5.2603,
      "mean_token_accuracy": 0.17389585971832275,
      "num_tokens": 25038201.0,
      "step": 12065
    },
    {
      "entropy": 5.657640790939331,
      "epoch": 1.0948838896952104,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004886938688635313,
      "loss": 5.4831,
      "mean_token_accuracy": 0.16164977103471756,
      "num_tokens": 25048466.0,
      "step": 12070
    },
    {
      "entropy": 5.706678771972657,
      "epoch": 1.0953374455732947,
      "grad_norm": 1.625,
      "learning_rate": 0.0004886837388395207,
      "loss": 5.4017,
      "mean_token_accuracy": 0.1703506514430046,
      "num_tokens": 25059064.0,
      "step": 12075
    },
    {
      "entropy": 5.64215497970581,
      "epoch": 1.0957910014513788,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000488673604396476,
      "loss": 5.4788,
      "mean_token_accuracy": 0.1606872409582138,
      "num_tokens": 25068834.0,
      "step": 12080
    },
    {
      "entropy": 5.586708450317383,
      "epoch": 1.096244557329463,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004886634655346067,
      "loss": 5.4397,
      "mean_token_accuracy": 0.162100912630558,
      "num_tokens": 25080058.0,
      "step": 12085
    },
    {
      "entropy": 5.591971731185913,
      "epoch": 1.0966981132075473,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004886533222541225,
      "loss": 5.3074,
      "mean_token_accuracy": 0.17501784563064576,
      "num_tokens": 25089751.0,
      "step": 12090
    },
    {
      "entropy": 5.6128082275390625,
      "epoch": 1.0971516690856313,
      "grad_norm": 3.0625,
      "learning_rate": 0.0004886431745552331,
      "loss": 5.371,
      "mean_token_accuracy": 0.17031979709863662,
      "num_tokens": 25099055.0,
      "step": 12095
    },
    {
      "entropy": 5.634336948394775,
      "epoch": 1.0976052249637156,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004886330224381485,
      "loss": 5.4672,
      "mean_token_accuracy": 0.16404568701982497,
      "num_tokens": 25109793.0,
      "step": 12100
    },
    {
      "entropy": 5.648942136764527,
      "epoch": 1.0980587808417996,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004886228659030785,
      "loss": 5.3978,
      "mean_token_accuracy": 0.170868156850338,
      "num_tokens": 25119814.0,
      "step": 12105
    },
    {
      "entropy": 5.655860137939453,
      "epoch": 1.098512336719884,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004886127049502331,
      "loss": 5.4157,
      "mean_token_accuracy": 0.16878984421491622,
      "num_tokens": 25130747.0,
      "step": 12110
    },
    {
      "entropy": 5.593351268768311,
      "epoch": 1.0989658925979682,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004886025395798228,
      "loss": 5.3377,
      "mean_token_accuracy": 0.16716485768556594,
      "num_tokens": 25142235.0,
      "step": 12115
    },
    {
      "entropy": 5.505036449432373,
      "epoch": 1.0994194484760522,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004885923697920573,
      "loss": 5.2672,
      "mean_token_accuracy": 0.17605393528938293,
      "num_tokens": 25152047.0,
      "step": 12120
    },
    {
      "entropy": 5.526382112503052,
      "epoch": 1.0998730043541365,
      "grad_norm": 3.0,
      "learning_rate": 0.0004885821955871472,
      "loss": 5.327,
      "mean_token_accuracy": 0.17646963745355607,
      "num_tokens": 25162678.0,
      "step": 12125
    },
    {
      "entropy": 5.637505960464478,
      "epoch": 1.1003265602322205,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004885720169653028,
      "loss": 5.3282,
      "mean_token_accuracy": 0.17307656705379487,
      "num_tokens": 25172973.0,
      "step": 12130
    },
    {
      "entropy": 5.677087020874024,
      "epoch": 1.1007801161103048,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004885618339267348,
      "loss": 5.4393,
      "mean_token_accuracy": 0.163906592130661,
      "num_tokens": 25183458.0,
      "step": 12135
    },
    {
      "entropy": 5.552436780929566,
      "epoch": 1.101233671988389,
      "grad_norm": 2.375,
      "learning_rate": 0.0004885516464716537,
      "loss": 5.3436,
      "mean_token_accuracy": 0.17302643954753877,
      "num_tokens": 25193827.0,
      "step": 12140
    },
    {
      "entropy": 5.6194024085998535,
      "epoch": 1.101687227866473,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048854145460027,
      "loss": 5.3603,
      "mean_token_accuracy": 0.1634839802980423,
      "num_tokens": 25204448.0,
      "step": 12145
    },
    {
      "entropy": 5.667161560058593,
      "epoch": 1.1021407837445574,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004885312583127946,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1770660623908043,
      "num_tokens": 25215041.0,
      "step": 12150
    },
    {
      "entropy": 5.618182945251465,
      "epoch": 1.1025943396226414,
      "grad_norm": 3.34375,
      "learning_rate": 0.0004885210576094385,
      "loss": 5.375,
      "mean_token_accuracy": 0.16552863866090775,
      "num_tokens": 25225937.0,
      "step": 12155
    },
    {
      "entropy": 5.5022780418396,
      "epoch": 1.1030478955007257,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004885108524904124,
      "loss": 5.268,
      "mean_token_accuracy": 0.17433019280433654,
      "num_tokens": 25235270.0,
      "step": 12160
    },
    {
      "entropy": 5.5538805484771725,
      "epoch": 1.10350145137881,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004885006429559275,
      "loss": 5.3828,
      "mean_token_accuracy": 0.16601527482271194,
      "num_tokens": 25247080.0,
      "step": 12165
    },
    {
      "entropy": 5.626804065704346,
      "epoch": 1.103955007256894,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004884904290061949,
      "loss": 5.3279,
      "mean_token_accuracy": 0.16205299496650696,
      "num_tokens": 25256718.0,
      "step": 12170
    },
    {
      "entropy": 5.646880435943603,
      "epoch": 1.1044085631349783,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004884802106414258,
      "loss": 5.4578,
      "mean_token_accuracy": 0.16500043272972106,
      "num_tokens": 25266632.0,
      "step": 12175
    },
    {
      "entropy": 5.603062582015991,
      "epoch": 1.1048621190130623,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004884699878618316,
      "loss": 5.336,
      "mean_token_accuracy": 0.17417868971824646,
      "num_tokens": 25275598.0,
      "step": 12180
    },
    {
      "entropy": 5.654306173324585,
      "epoch": 1.1053156748911466,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004884597606676236,
      "loss": 5.4418,
      "mean_token_accuracy": 0.16754518151283265,
      "num_tokens": 25286598.0,
      "step": 12185
    },
    {
      "entropy": 5.577902507781983,
      "epoch": 1.1057692307692308,
      "grad_norm": 3.234375,
      "learning_rate": 0.0004884495290590134,
      "loss": 5.3423,
      "mean_token_accuracy": 0.1773931473493576,
      "num_tokens": 25296730.0,
      "step": 12190
    },
    {
      "entropy": 5.557775259017944,
      "epoch": 1.1062227866473149,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004884392930362124,
      "loss": 5.2596,
      "mean_token_accuracy": 0.18149112164974213,
      "num_tokens": 25306326.0,
      "step": 12195
    },
    {
      "entropy": 5.612017250061035,
      "epoch": 1.1066763425253991,
      "grad_norm": 3.28125,
      "learning_rate": 0.0004884290525994326,
      "loss": 5.3665,
      "mean_token_accuracy": 0.17338535636663438,
      "num_tokens": 25316342.0,
      "step": 12200
    },
    {
      "entropy": 5.66097674369812,
      "epoch": 1.1071298984034832,
      "grad_norm": 2.625,
      "learning_rate": 0.0004884188077488855,
      "loss": 5.4022,
      "mean_token_accuracy": 0.16372329443693162,
      "num_tokens": 25326129.0,
      "step": 12205
    },
    {
      "entropy": 5.616948175430298,
      "epoch": 1.1075834542815675,
      "grad_norm": 4.375,
      "learning_rate": 0.0004884085584847831,
      "loss": 5.389,
      "mean_token_accuracy": 0.16632895767688752,
      "num_tokens": 25337156.0,
      "step": 12210
    },
    {
      "entropy": 5.704233694076538,
      "epoch": 1.1080370101596517,
      "grad_norm": 2.890625,
      "learning_rate": 0.0004883983048073375,
      "loss": 5.4791,
      "mean_token_accuracy": 0.1600780501961708,
      "num_tokens": 25347632.0,
      "step": 12215
    },
    {
      "entropy": 5.585485649108887,
      "epoch": 1.1084905660377358,
      "grad_norm": 8.5625,
      "learning_rate": 0.0004883880467167603,
      "loss": 5.2639,
      "mean_token_accuracy": 0.17368678599596024,
      "num_tokens": 25359186.0,
      "step": 12220
    },
    {
      "entropy": 5.620640993118286,
      "epoch": 1.10894412191582,
      "grad_norm": 4.78125,
      "learning_rate": 0.000488377784213264,
      "loss": 5.3948,
      "mean_token_accuracy": 0.16820335239171982,
      "num_tokens": 25368993.0,
      "step": 12225
    },
    {
      "entropy": 5.585221481323242,
      "epoch": 1.1093976777939043,
      "grad_norm": 3.75,
      "learning_rate": 0.0004883675172970608,
      "loss": 5.3393,
      "mean_token_accuracy": 0.16990748345851897,
      "num_tokens": 25379880.0,
      "step": 12230
    },
    {
      "entropy": 5.603193998336792,
      "epoch": 1.1098512336719883,
      "grad_norm": 2.59375,
      "learning_rate": 0.000488357245968363,
      "loss": 5.5198,
      "mean_token_accuracy": 0.15841159969568253,
      "num_tokens": 25389942.0,
      "step": 12235
    },
    {
      "entropy": 5.587256574630738,
      "epoch": 1.1103047895500726,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004883469702273829,
      "loss": 5.317,
      "mean_token_accuracy": 0.17499275505542755,
      "num_tokens": 25400836.0,
      "step": 12240
    },
    {
      "entropy": 5.609990501403809,
      "epoch": 1.1107583454281567,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004883366900743331,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17449965476989746,
      "num_tokens": 25410502.0,
      "step": 12245
    },
    {
      "entropy": 5.545148801803589,
      "epoch": 1.111211901306241,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004883264055094262,
      "loss": 5.284,
      "mean_token_accuracy": 0.18040893524885177,
      "num_tokens": 25420301.0,
      "step": 12250
    },
    {
      "entropy": 5.57139105796814,
      "epoch": 1.1116654571843252,
      "grad_norm": 2.0,
      "learning_rate": 0.0004883161165328749,
      "loss": 5.3728,
      "mean_token_accuracy": 0.16551127135753632,
      "num_tokens": 25431107.0,
      "step": 12255
    },
    {
      "entropy": 5.595264530181884,
      "epoch": 1.1121190130624092,
      "grad_norm": 3.015625,
      "learning_rate": 0.000488305823144892,
      "loss": 5.3804,
      "mean_token_accuracy": 0.16993945240974426,
      "num_tokens": 25441256.0,
      "step": 12260
    },
    {
      "entropy": 5.624409151077271,
      "epoch": 1.1125725689404935,
      "grad_norm": 5.8125,
      "learning_rate": 0.0004882955253456902,
      "loss": 5.3217,
      "mean_token_accuracy": 0.17479237616062165,
      "num_tokens": 25451627.0,
      "step": 12265
    },
    {
      "entropy": 5.656954574584961,
      "epoch": 1.1130261248185778,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004882852231354826,
      "loss": 5.418,
      "mean_token_accuracy": 0.16552453935146333,
      "num_tokens": 25461273.0,
      "step": 12270
    },
    {
      "entropy": 5.587964868545532,
      "epoch": 1.1134796806966618,
      "grad_norm": 2.25,
      "learning_rate": 0.0004882749165144822,
      "loss": 5.3118,
      "mean_token_accuracy": 0.17116914242506026,
      "num_tokens": 25471949.0,
      "step": 12275
    },
    {
      "entropy": 5.645139741897583,
      "epoch": 1.113933236574746,
      "grad_norm": 1.625,
      "learning_rate": 0.0004882646054829023,
      "loss": 5.3336,
      "mean_token_accuracy": 0.17258032262325287,
      "num_tokens": 25482175.0,
      "step": 12280
    },
    {
      "entropy": 5.604651880264282,
      "epoch": 1.1143867924528301,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048825429004095597,
      "loss": 5.3999,
      "mean_token_accuracy": 0.1717308834195137,
      "num_tokens": 25491786.0,
      "step": 12285
    },
    {
      "entropy": 5.566480207443237,
      "epoch": 1.1148403483309144,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004882439701888566,
      "loss": 5.3316,
      "mean_token_accuracy": 0.17830322086811065,
      "num_tokens": 25502103.0,
      "step": 12290
    },
    {
      "entropy": 5.580591869354248,
      "epoch": 1.1152939042089987,
      "grad_norm": 2.703125,
      "learning_rate": 0.00048823364592681754,
      "loss": 5.3025,
      "mean_token_accuracy": 0.1707880452275276,
      "num_tokens": 25512530.0,
      "step": 12295
    },
    {
      "entropy": 5.5863017559051515,
      "epoch": 1.1157474600870827,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048822331725505234,
      "loss": 5.3082,
      "mean_token_accuracy": 0.17238383442163469,
      "num_tokens": 25522905.0,
      "step": 12300
    },
    {
      "entropy": 5.592488145828247,
      "epoch": 1.116201015965167,
      "grad_norm": 2.546875,
      "learning_rate": 0.00048821298417377456,
      "loss": 5.2789,
      "mean_token_accuracy": 0.17962971776723863,
      "num_tokens": 25532963.0,
      "step": 12305
    },
    {
      "entropy": 5.485294008255005,
      "epoch": 1.116654571843251,
      "grad_norm": 3.6875,
      "learning_rate": 0.000488202646683198,
      "loss": 5.364,
      "mean_token_accuracy": 0.16723102033138276,
      "num_tokens": 25542583.0,
      "step": 12310
    },
    {
      "entropy": 5.586943578720093,
      "epoch": 1.1171081277213353,
      "grad_norm": 2.25,
      "learning_rate": 0.0004881923047835364,
      "loss": 5.3379,
      "mean_token_accuracy": 0.17322114557027818,
      "num_tokens": 25553367.0,
      "step": 12315
    },
    {
      "entropy": 5.731902122497559,
      "epoch": 1.1175616835994195,
      "grad_norm": 3.390625,
      "learning_rate": 0.0004881819584750036,
      "loss": 5.429,
      "mean_token_accuracy": 0.1654035449028015,
      "num_tokens": 25564823.0,
      "step": 12320
    },
    {
      "entropy": 5.634766054153443,
      "epoch": 1.1180152394775036,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048817160775781356,
      "loss": 5.4069,
      "mean_token_accuracy": 0.16874194741249085,
      "num_tokens": 25574946.0,
      "step": 12325
    },
    {
      "entropy": 5.64937686920166,
      "epoch": 1.1184687953555879,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048816125263218023,
      "loss": 5.392,
      "mean_token_accuracy": 0.16683305501937867,
      "num_tokens": 25584779.0,
      "step": 12330
    },
    {
      "entropy": 5.7065942764282225,
      "epoch": 1.118922351233672,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048815089309831803,
      "loss": 5.4394,
      "mean_token_accuracy": 0.16809054017066954,
      "num_tokens": 25594725.0,
      "step": 12335
    },
    {
      "entropy": 5.621603631973267,
      "epoch": 1.1193759071117562,
      "grad_norm": 2.46875,
      "learning_rate": 0.000488140529156441,
      "loss": 5.4307,
      "mean_token_accuracy": 0.16446011364459992,
      "num_tokens": 25605499.0,
      "step": 12340
    },
    {
      "entropy": 5.562032842636109,
      "epoch": 1.1198294629898404,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004881301608067635,
      "loss": 5.3678,
      "mean_token_accuracy": 0.17318881750106813,
      "num_tokens": 25616443.0,
      "step": 12345
    },
    {
      "entropy": 5.619059038162232,
      "epoch": 1.1202830188679245,
      "grad_norm": 2.265625,
      "learning_rate": 0.00048811978804949997,
      "loss": 5.3491,
      "mean_token_accuracy": 0.17076736688613892,
      "num_tokens": 25627762.0,
      "step": 12350
    },
    {
      "entropy": 5.66572413444519,
      "epoch": 1.1207365747460087,
      "grad_norm": 2.4375,
      "learning_rate": 0.000488109410884865,
      "loss": 5.4323,
      "mean_token_accuracy": 0.16833171248435974,
      "num_tokens": 25637904.0,
      "step": 12355
    },
    {
      "entropy": 5.605768728256225,
      "epoch": 1.1211901306240928,
      "grad_norm": 2.53125,
      "learning_rate": 0.00048809902931307294,
      "loss": 5.2871,
      "mean_token_accuracy": 0.1745767042040825,
      "num_tokens": 25649058.0,
      "step": 12360
    },
    {
      "entropy": 5.570656871795654,
      "epoch": 1.121643686502177,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004880886433343388,
      "loss": 5.3577,
      "mean_token_accuracy": 0.17595502585172654,
      "num_tokens": 25659161.0,
      "step": 12365
    },
    {
      "entropy": 5.612319993972778,
      "epoch": 1.1220972423802613,
      "grad_norm": 3.15625,
      "learning_rate": 0.0004880782529488771,
      "loss": 5.2741,
      "mean_token_accuracy": 0.17882423251867294,
      "num_tokens": 25669336.0,
      "step": 12370
    },
    {
      "entropy": 5.6442542552948,
      "epoch": 1.1225507982583454,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004880678581569029,
      "loss": 5.4047,
      "mean_token_accuracy": 0.1716239035129547,
      "num_tokens": 25680443.0,
      "step": 12375
    },
    {
      "entropy": 5.590997934341431,
      "epoch": 1.1230043541364296,
      "grad_norm": 2.53125,
      "learning_rate": 0.00048805745895863105,
      "loss": 5.3011,
      "mean_token_accuracy": 0.1769050568342209,
      "num_tokens": 25690202.0,
      "step": 12380
    },
    {
      "entropy": 5.585967111587524,
      "epoch": 1.1234579100145137,
      "grad_norm": 2.78125,
      "learning_rate": 0.0004880470553542767,
      "loss": 5.366,
      "mean_token_accuracy": 0.17231284379959105,
      "num_tokens": 25699255.0,
      "step": 12385
    },
    {
      "entropy": 5.604588222503662,
      "epoch": 1.123911465892598,
      "grad_norm": 3.046875,
      "learning_rate": 0.00048803664734405495,
      "loss": 5.3599,
      "mean_token_accuracy": 0.1651730254292488,
      "num_tokens": 25710714.0,
      "step": 12390
    },
    {
      "entropy": 5.6293251514434814,
      "epoch": 1.1243650217706822,
      "grad_norm": 2.984375,
      "learning_rate": 0.000488026234928181,
      "loss": 5.3303,
      "mean_token_accuracy": 0.17343648225069047,
      "num_tokens": 25721722.0,
      "step": 12395
    },
    {
      "entropy": 5.607216262817383,
      "epoch": 1.1248185776487662,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004880158181068703,
      "loss": 5.3202,
      "mean_token_accuracy": 0.17717739641666413,
      "num_tokens": 25732096.0,
      "step": 12400
    },
    {
      "entropy": 5.627986001968384,
      "epoch": 1.1252721335268505,
      "grad_norm": 2.546875,
      "learning_rate": 0.00048800539688033807,
      "loss": 5.3671,
      "mean_token_accuracy": 0.17440272569656373,
      "num_tokens": 25741499.0,
      "step": 12405
    },
    {
      "entropy": 5.606373643875122,
      "epoch": 1.1257256894049348,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004879949712488001,
      "loss": 5.4597,
      "mean_token_accuracy": 0.17155432403087617,
      "num_tokens": 25751913.0,
      "step": 12410
    },
    {
      "entropy": 5.6275887966156,
      "epoch": 1.1261792452830188,
      "grad_norm": 2.71875,
      "learning_rate": 0.00048798454121247177,
      "loss": 5.3916,
      "mean_token_accuracy": 0.16975918412208557,
      "num_tokens": 25762602.0,
      "step": 12415
    },
    {
      "entropy": 5.5750587463378904,
      "epoch": 1.126632801161103,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004879741067715688,
      "loss": 5.3769,
      "mean_token_accuracy": 0.1725024700164795,
      "num_tokens": 25773872.0,
      "step": 12420
    },
    {
      "entropy": 5.6519242286682125,
      "epoch": 1.1270863570391871,
      "grad_norm": 2.53125,
      "learning_rate": 0.00048796366792630715,
      "loss": 5.3783,
      "mean_token_accuracy": 0.1695870876312256,
      "num_tokens": 25784723.0,
      "step": 12425
    },
    {
      "entropy": 5.5859214782714846,
      "epoch": 1.1275399129172714,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004879532246769025,
      "loss": 5.3973,
      "mean_token_accuracy": 0.17278240323066713,
      "num_tokens": 25794897.0,
      "step": 12430
    },
    {
      "entropy": 5.559264755249023,
      "epoch": 1.1279934687953557,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004879427770235709,
      "loss": 5.2915,
      "mean_token_accuracy": 0.1784004509449005,
      "num_tokens": 25804445.0,
      "step": 12435
    },
    {
      "entropy": 5.629906272888183,
      "epoch": 1.1284470246734397,
      "grad_norm": 2.234375,
      "learning_rate": 0.00048793232496652846,
      "loss": 5.4376,
      "mean_token_accuracy": 0.16687726825475693,
      "num_tokens": 25815495.0,
      "step": 12440
    },
    {
      "entropy": 5.6425799369812015,
      "epoch": 1.128900580551524,
      "grad_norm": 3.171875,
      "learning_rate": 0.00048792186850599124,
      "loss": 5.3636,
      "mean_token_accuracy": 0.17704028636217117,
      "num_tokens": 25826120.0,
      "step": 12445
    },
    {
      "entropy": 5.57022442817688,
      "epoch": 1.1293541364296082,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004879114076421755,
      "loss": 5.3516,
      "mean_token_accuracy": 0.1738788291811943,
      "num_tokens": 25835926.0,
      "step": 12450
    },
    {
      "entropy": 5.601430511474609,
      "epoch": 1.1298076923076923,
      "grad_norm": 2.28125,
      "learning_rate": 0.00048790094237529763,
      "loss": 5.4008,
      "mean_token_accuracy": 0.16673785895109178,
      "num_tokens": 25846095.0,
      "step": 12455
    },
    {
      "entropy": 5.64870138168335,
      "epoch": 1.1302612481857766,
      "grad_norm": 7.21875,
      "learning_rate": 0.000487890472705574,
      "loss": 5.397,
      "mean_token_accuracy": 0.16596953570842743,
      "num_tokens": 25858541.0,
      "step": 12460
    },
    {
      "entropy": 5.645746803283691,
      "epoch": 1.1307148040638606,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048787999863322113,
      "loss": 5.3072,
      "mean_token_accuracy": 0.17562833726406096,
      "num_tokens": 25867664.0,
      "step": 12465
    },
    {
      "entropy": 5.601834630966186,
      "epoch": 1.1311683599419449,
      "grad_norm": 3.875,
      "learning_rate": 0.0004878695201584557,
      "loss": 5.3741,
      "mean_token_accuracy": 0.1729818180203438,
      "num_tokens": 25877300.0,
      "step": 12470
    },
    {
      "entropy": 5.557867574691772,
      "epoch": 1.1316219158200291,
      "grad_norm": 5.78125,
      "learning_rate": 0.0004878590372814943,
      "loss": 5.3525,
      "mean_token_accuracy": 0.1747228294610977,
      "num_tokens": 25888009.0,
      "step": 12475
    },
    {
      "entropy": 5.63745608329773,
      "epoch": 1.1320754716981132,
      "grad_norm": 2.9375,
      "learning_rate": 0.00048784855000255374,
      "loss": 5.4187,
      "mean_token_accuracy": 0.1693594604730606,
      "num_tokens": 25897285.0,
      "step": 12480
    },
    {
      "entropy": 5.55023045539856,
      "epoch": 1.1325290275761974,
      "grad_norm": 2.296875,
      "learning_rate": 0.00048783805832185097,
      "loss": 5.2962,
      "mean_token_accuracy": 0.17955439239740373,
      "num_tokens": 25907355.0,
      "step": 12485
    },
    {
      "entropy": 5.582215213775635,
      "epoch": 1.1329825834542815,
      "grad_norm": 3.03125,
      "learning_rate": 0.0004878275622396029,
      "loss": 5.378,
      "mean_token_accuracy": 0.17706575989723206,
      "num_tokens": 25918566.0,
      "step": 12490
    },
    {
      "entropy": 5.625691747665405,
      "epoch": 1.1334361393323658,
      "grad_norm": 4.84375,
      "learning_rate": 0.00048781706175602665,
      "loss": 5.4249,
      "mean_token_accuracy": 0.16875527054071426,
      "num_tokens": 25929145.0,
      "step": 12495
    },
    {
      "entropy": 5.602830982208252,
      "epoch": 1.13388969521045,
      "grad_norm": 2.0,
      "learning_rate": 0.00048780655687133927,
      "loss": 5.2591,
      "mean_token_accuracy": 0.1770707294344902,
      "num_tokens": 25938163.0,
      "step": 12500
    },
    {
      "entropy": 5.572844696044922,
      "epoch": 1.134343251088534,
      "grad_norm": 2.6875,
      "learning_rate": 0.00048779604758575806,
      "loss": 5.3331,
      "mean_token_accuracy": 0.168691024184227,
      "num_tokens": 25948962.0,
      "step": 12505
    },
    {
      "entropy": 5.530844593048096,
      "epoch": 1.1347968069666183,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004877855338995004,
      "loss": 5.3057,
      "mean_token_accuracy": 0.16639216989278793,
      "num_tokens": 25959653.0,
      "step": 12510
    },
    {
      "entropy": 5.650057125091553,
      "epoch": 1.1352503628447024,
      "grad_norm": 4.75,
      "learning_rate": 0.00048777501581278367,
      "loss": 5.3823,
      "mean_token_accuracy": 0.16966882944107056,
      "num_tokens": 25970268.0,
      "step": 12515
    },
    {
      "entropy": 5.64359221458435,
      "epoch": 1.1357039187227866,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004877644933258254,
      "loss": 5.352,
      "mean_token_accuracy": 0.16738999187946318,
      "num_tokens": 25980321.0,
      "step": 12520
    },
    {
      "entropy": 5.593331241607666,
      "epoch": 1.136157474600871,
      "grad_norm": 2.375,
      "learning_rate": 0.0004877539664388432,
      "loss": 5.3433,
      "mean_token_accuracy": 0.1679622158408165,
      "num_tokens": 25992466.0,
      "step": 12525
    },
    {
      "entropy": 5.567163467407227,
      "epoch": 1.136611030478955,
      "grad_norm": 2.734375,
      "learning_rate": 0.00048774343515205464,
      "loss": 5.3395,
      "mean_token_accuracy": 0.1757104605436325,
      "num_tokens": 26003553.0,
      "step": 12530
    },
    {
      "entropy": 5.583215379714966,
      "epoch": 1.1370645863570392,
      "grad_norm": 2.125,
      "learning_rate": 0.0004877328994656778,
      "loss": 5.3569,
      "mean_token_accuracy": 0.17268076688051223,
      "num_tokens": 26013299.0,
      "step": 12535
    },
    {
      "entropy": 5.577181673049926,
      "epoch": 1.1375181422351233,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004877223593799303,
      "loss": 5.3547,
      "mean_token_accuracy": 0.17435978204011918,
      "num_tokens": 26022913.0,
      "step": 12540
    },
    {
      "entropy": 5.560851335525513,
      "epoch": 1.1379716981132075,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004877118148950302,
      "loss": 5.3523,
      "mean_token_accuracy": 0.17192198634147643,
      "num_tokens": 26033191.0,
      "step": 12545
    },
    {
      "entropy": 5.666209173202515,
      "epoch": 1.1384252539912918,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048770126601119555,
      "loss": 5.4801,
      "mean_token_accuracy": 0.1612833708524704,
      "num_tokens": 26043517.0,
      "step": 12550
    },
    {
      "entropy": 5.682900476455688,
      "epoch": 1.1388788098693758,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048769071272864454,
      "loss": 5.4265,
      "mean_token_accuracy": 0.16507856845855712,
      "num_tokens": 26054173.0,
      "step": 12555
    },
    {
      "entropy": 5.73878493309021,
      "epoch": 1.13933236574746,
      "grad_norm": 2.859375,
      "learning_rate": 0.00048768015504759536,
      "loss": 5.4091,
      "mean_token_accuracy": 0.16610510051250457,
      "num_tokens": 26064929.0,
      "step": 12560
    },
    {
      "entropy": 5.569660615921021,
      "epoch": 1.1397859216255442,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004876695929682665,
      "loss": 5.3208,
      "mean_token_accuracy": 0.17363446354866027,
      "num_tokens": 26074872.0,
      "step": 12565
    },
    {
      "entropy": 5.557733154296875,
      "epoch": 1.1402394775036284,
      "grad_norm": 2.6875,
      "learning_rate": 0.00048765902649087605,
      "loss": 5.2466,
      "mean_token_accuracy": 0.1786336600780487,
      "num_tokens": 26086328.0,
      "step": 12570
    },
    {
      "entropy": 5.626145648956299,
      "epoch": 1.1406930333817127,
      "grad_norm": 2.390625,
      "learning_rate": 0.00048764845561564285,
      "loss": 5.3443,
      "mean_token_accuracy": 0.1702533707022667,
      "num_tokens": 26096477.0,
      "step": 12575
    },
    {
      "entropy": 5.601234483718872,
      "epoch": 1.1411465892597967,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048763788034278545,
      "loss": 5.3234,
      "mean_token_accuracy": 0.17736123502254486,
      "num_tokens": 26106183.0,
      "step": 12580
    },
    {
      "entropy": 5.620738458633423,
      "epoch": 1.141600145137881,
      "grad_norm": 3.3125,
      "learning_rate": 0.0004876273006725225,
      "loss": 5.3801,
      "mean_token_accuracy": 0.17188373506069182,
      "num_tokens": 26117932.0,
      "step": 12585
    },
    {
      "entropy": 5.573065900802613,
      "epoch": 1.1420537010159653,
      "grad_norm": 3.875,
      "learning_rate": 0.00048761671660507266,
      "loss": 5.3286,
      "mean_token_accuracy": 0.17553336024284363,
      "num_tokens": 26128616.0,
      "step": 12590
    },
    {
      "entropy": 5.598615646362305,
      "epoch": 1.1425072568940493,
      "grad_norm": 2.359375,
      "learning_rate": 0.000487606128140655,
      "loss": 5.325,
      "mean_token_accuracy": 0.17197948098182678,
      "num_tokens": 26138758.0,
      "step": 12595
    },
    {
      "entropy": 5.534066200256348,
      "epoch": 1.1429608127721336,
      "grad_norm": 3.234375,
      "learning_rate": 0.0004875955352794885,
      "loss": 5.3307,
      "mean_token_accuracy": 0.17095252275466918,
      "num_tokens": 26149199.0,
      "step": 12600
    },
    {
      "entropy": 5.591140508651733,
      "epoch": 1.1434143686502176,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004875849380217921,
      "loss": 5.4152,
      "mean_token_accuracy": 0.17149858623743058,
      "num_tokens": 26160276.0,
      "step": 12605
    },
    {
      "entropy": 5.6406333446502686,
      "epoch": 1.1438679245283019,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000487574336367785,
      "loss": 5.2991,
      "mean_token_accuracy": 0.1732035234570503,
      "num_tokens": 26169564.0,
      "step": 12610
    },
    {
      "entropy": 5.576666784286499,
      "epoch": 1.1443214804063861,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004875637303176865,
      "loss": 5.2783,
      "mean_token_accuracy": 0.17607302516698836,
      "num_tokens": 26179135.0,
      "step": 12615
    },
    {
      "entropy": 5.617819881439209,
      "epoch": 1.1447750362844702,
      "grad_norm": 2.84375,
      "learning_rate": 0.00048755311987171585,
      "loss": 5.4096,
      "mean_token_accuracy": 0.17113402783870696,
      "num_tokens": 26190116.0,
      "step": 12620
    },
    {
      "entropy": 5.6403364658355715,
      "epoch": 1.1452285921625545,
      "grad_norm": 1.96875,
      "learning_rate": 0.00048754250503009244,
      "loss": 5.4832,
      "mean_token_accuracy": 0.16499626636505127,
      "num_tokens": 26200847.0,
      "step": 12625
    },
    {
      "entropy": 5.711669445037842,
      "epoch": 1.1456821480406387,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00048753188579303596,
      "loss": 5.4402,
      "mean_token_accuracy": 0.16778092607855796,
      "num_tokens": 26211151.0,
      "step": 12630
    },
    {
      "entropy": 5.638349628448486,
      "epoch": 1.1461357039187228,
      "grad_norm": 4.0625,
      "learning_rate": 0.0004875212621607659,
      "loss": 5.3565,
      "mean_token_accuracy": 0.18072609156370162,
      "num_tokens": 26221273.0,
      "step": 12635
    },
    {
      "entropy": 5.598391008377075,
      "epoch": 1.146589259796807,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048751063413350203,
      "loss": 5.4555,
      "mean_token_accuracy": 0.1663660630583763,
      "num_tokens": 26231770.0,
      "step": 12640
    },
    {
      "entropy": 5.6362285137176515,
      "epoch": 1.147042815674891,
      "grad_norm": 2.796875,
      "learning_rate": 0.000487500001711464,
      "loss": 5.409,
      "mean_token_accuracy": 0.16883457452058792,
      "num_tokens": 26241386.0,
      "step": 12645
    },
    {
      "entropy": 5.752071523666382,
      "epoch": 1.1474963715529753,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004874893648948719,
      "loss": 5.485,
      "mean_token_accuracy": 0.16835420429706574,
      "num_tokens": 26251991.0,
      "step": 12650
    },
    {
      "entropy": 5.724516439437866,
      "epoch": 1.1479499274310596,
      "grad_norm": 2.609375,
      "learning_rate": 0.00048747872368394546,
      "loss": 5.5087,
      "mean_token_accuracy": 0.16761526167392732,
      "num_tokens": 26262331.0,
      "step": 12655
    },
    {
      "entropy": 5.678382444381714,
      "epoch": 1.1484034833091437,
      "grad_norm": 2.9375,
      "learning_rate": 0.00048746807807890496,
      "loss": 5.3966,
      "mean_token_accuracy": 0.16898454427719117,
      "num_tokens": 26272939.0,
      "step": 12660
    },
    {
      "entropy": 5.591447019577027,
      "epoch": 1.148857039187228,
      "grad_norm": 2.546875,
      "learning_rate": 0.00048745742807997046,
      "loss": 5.359,
      "mean_token_accuracy": 0.17616231739521027,
      "num_tokens": 26282849.0,
      "step": 12665
    },
    {
      "entropy": 5.560210514068603,
      "epoch": 1.149310595065312,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048744677368736216,
      "loss": 5.3798,
      "mean_token_accuracy": 0.17931461036205293,
      "num_tokens": 26292449.0,
      "step": 12670
    },
    {
      "entropy": 5.534169292449951,
      "epoch": 1.1497641509433962,
      "grad_norm": 2.328125,
      "learning_rate": 0.00048743611490130044,
      "loss": 5.3128,
      "mean_token_accuracy": 0.17279691994190216,
      "num_tokens": 26302506.0,
      "step": 12675
    },
    {
      "entropy": 5.6572449684143065,
      "epoch": 1.1502177068214805,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048742545172200576,
      "loss": 5.3816,
      "mean_token_accuracy": 0.17448493242263793,
      "num_tokens": 26312346.0,
      "step": 12680
    },
    {
      "entropy": 5.699370336532593,
      "epoch": 1.1506712626995645,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004874147841496986,
      "loss": 5.3932,
      "mean_token_accuracy": 0.17356102615594865,
      "num_tokens": 26323173.0,
      "step": 12685
    },
    {
      "entropy": 5.6548998832702635,
      "epoch": 1.1511248185776488,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004874041121845996,
      "loss": 5.4472,
      "mean_token_accuracy": 0.16051401942968369,
      "num_tokens": 26334371.0,
      "step": 12690
    },
    {
      "entropy": 5.611789703369141,
      "epoch": 1.1515783744557329,
      "grad_norm": 3.046875,
      "learning_rate": 0.0004873934358269294,
      "loss": 5.3853,
      "mean_token_accuracy": 0.17125811725854873,
      "num_tokens": 26344506.0,
      "step": 12695
    },
    {
      "entropy": 5.668434953689575,
      "epoch": 1.1520319303338171,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004873827550769088,
      "loss": 5.3772,
      "mean_token_accuracy": 0.16571766287088394,
      "num_tokens": 26354534.0,
      "step": 12700
    },
    {
      "entropy": 5.651076793670654,
      "epoch": 1.1524854862119014,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004873720699347587,
      "loss": 5.3628,
      "mean_token_accuracy": 0.169284388422966,
      "num_tokens": 26365682.0,
      "step": 12705
    },
    {
      "entropy": 5.684403753280639,
      "epoch": 1.1529390420899854,
      "grad_norm": 2.125,
      "learning_rate": 0.0004873613804007001,
      "loss": 5.4637,
      "mean_token_accuracy": 0.17238470315933227,
      "num_tokens": 26376702.0,
      "step": 12710
    },
    {
      "entropy": 5.590936088562012,
      "epoch": 1.1533925979680697,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004873506864749541,
      "loss": 5.2834,
      "mean_token_accuracy": 0.17693755626678467,
      "num_tokens": 26387154.0,
      "step": 12715
    },
    {
      "entropy": 5.658411073684692,
      "epoch": 1.1538461538461537,
      "grad_norm": 2.0,
      "learning_rate": 0.0004873399881577417,
      "loss": 5.4446,
      "mean_token_accuracy": 0.16750529110431672,
      "num_tokens": 26397883.0,
      "step": 12720
    },
    {
      "entropy": 5.500313663482666,
      "epoch": 1.154299709724238,
      "grad_norm": 2.375,
      "learning_rate": 0.0004873292854492842,
      "loss": 5.2677,
      "mean_token_accuracy": 0.1807516932487488,
      "num_tokens": 26407613.0,
      "step": 12725
    },
    {
      "entropy": 5.5440812587738035,
      "epoch": 1.1547532656023223,
      "grad_norm": 2.625,
      "learning_rate": 0.0004873185783498031,
      "loss": 5.279,
      "mean_token_accuracy": 0.17283984422683715,
      "num_tokens": 26417221.0,
      "step": 12730
    },
    {
      "entropy": 5.603883457183838,
      "epoch": 1.1552068214804063,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004873078668595197,
      "loss": 5.3655,
      "mean_token_accuracy": 0.17053250819444657,
      "num_tokens": 26427976.0,
      "step": 12735
    },
    {
      "entropy": 5.618934535980225,
      "epoch": 1.1556603773584906,
      "grad_norm": 2.984375,
      "learning_rate": 0.00048729715097865534,
      "loss": 5.4054,
      "mean_token_accuracy": 0.16815861016511918,
      "num_tokens": 26439213.0,
      "step": 12740
    },
    {
      "entropy": 5.586228656768799,
      "epoch": 1.1561139332365746,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004872864307074319,
      "loss": 5.3517,
      "mean_token_accuracy": 0.17429283857345582,
      "num_tokens": 26449763.0,
      "step": 12745
    },
    {
      "entropy": 5.653108072280884,
      "epoch": 1.156567489114659,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048727570604607096,
      "loss": 5.4084,
      "mean_token_accuracy": 0.1733284428715706,
      "num_tokens": 26460133.0,
      "step": 12750
    },
    {
      "entropy": 5.630918216705322,
      "epoch": 1.1570210449927432,
      "grad_norm": 2.375,
      "learning_rate": 0.00048726497699479437,
      "loss": 5.3283,
      "mean_token_accuracy": 0.17489899545907975,
      "num_tokens": 26470164.0,
      "step": 12755
    },
    {
      "entropy": 5.633225536346435,
      "epoch": 1.1574746008708272,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000487254243553824,
      "loss": 5.3728,
      "mean_token_accuracy": 0.1685292363166809,
      "num_tokens": 26480314.0,
      "step": 12760
    },
    {
      "entropy": 5.556787967681885,
      "epoch": 1.1579281567489115,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004872435057233817,
      "loss": 5.3397,
      "mean_token_accuracy": 0.1793245404958725,
      "num_tokens": 26491184.0,
      "step": 12765
    },
    {
      "entropy": 5.519004535675049,
      "epoch": 1.1583817126269957,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048723276350368957,
      "loss": 5.2352,
      "mean_token_accuracy": 0.17802902013063432,
      "num_tokens": 26501432.0,
      "step": 12770
    },
    {
      "entropy": 5.612009906768799,
      "epoch": 1.1588352685050798,
      "grad_norm": 2.0,
      "learning_rate": 0.00048722201689496984,
      "loss": 5.3534,
      "mean_token_accuracy": 0.1688969388604164,
      "num_tokens": 26510379.0,
      "step": 12775
    },
    {
      "entropy": 5.59565372467041,
      "epoch": 1.159288824383164,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004872112658974447,
      "loss": 5.3205,
      "mean_token_accuracy": 0.17860374599695206,
      "num_tokens": 26519454.0,
      "step": 12780
    },
    {
      "entropy": 5.543451499938965,
      "epoch": 1.159742380261248,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048720051051133657,
      "loss": 5.2781,
      "mean_token_accuracy": 0.17807969748973845,
      "num_tokens": 26531325.0,
      "step": 12785
    },
    {
      "entropy": 5.564303970336914,
      "epoch": 1.1601959361393324,
      "grad_norm": 7.5625,
      "learning_rate": 0.0004871897507368678,
      "loss": 5.3582,
      "mean_token_accuracy": 0.17585408091545104,
      "num_tokens": 26541510.0,
      "step": 12790
    },
    {
      "entropy": 5.657733917236328,
      "epoch": 1.1606494920174166,
      "grad_norm": 2.578125,
      "learning_rate": 0.00048717898657426093,
      "loss": 5.5384,
      "mean_token_accuracy": 0.16009921431541443,
      "num_tokens": 26553400.0,
      "step": 12795
    },
    {
      "entropy": 5.641717624664307,
      "epoch": 1.1611030478955007,
      "grad_norm": 2.0,
      "learning_rate": 0.00048716821802373856,
      "loss": 5.3703,
      "mean_token_accuracy": 0.16997628957033156,
      "num_tokens": 26564340.0,
      "step": 12800
    },
    {
      "entropy": 5.6257671356201175,
      "epoch": 1.161556603773585,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00048715744508552334,
      "loss": 5.3427,
      "mean_token_accuracy": 0.1764123722910881,
      "num_tokens": 26573615.0,
      "step": 12805
    },
    {
      "entropy": 5.6515766143798825,
      "epoch": 1.1620101596516692,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004871466677598381,
      "loss": 5.4905,
      "mean_token_accuracy": 0.1637832596898079,
      "num_tokens": 26584149.0,
      "step": 12810
    },
    {
      "entropy": 5.609748268127442,
      "epoch": 1.1624637155297532,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004871358860469057,
      "loss": 5.3884,
      "mean_token_accuracy": 0.16375558823347092,
      "num_tokens": 26595064.0,
      "step": 12815
    },
    {
      "entropy": 5.586152744293213,
      "epoch": 1.1629172714078375,
      "grad_norm": 2.65625,
      "learning_rate": 0.00048712509994694903,
      "loss": 5.3905,
      "mean_token_accuracy": 0.16831739246845245,
      "num_tokens": 26605636.0,
      "step": 12820
    },
    {
      "entropy": 5.642949008941651,
      "epoch": 1.1633708272859216,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004871143094601914,
      "loss": 5.4768,
      "mean_token_accuracy": 0.16458763927221298,
      "num_tokens": 26616065.0,
      "step": 12825
    },
    {
      "entropy": 5.6962902545928955,
      "epoch": 1.1638243831640058,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004871035145868557,
      "loss": 5.4149,
      "mean_token_accuracy": 0.1654929742217064,
      "num_tokens": 26627246.0,
      "step": 12830
    },
    {
      "entropy": 5.566495084762574,
      "epoch": 1.16427793904209,
      "grad_norm": 1.78125,
      "learning_rate": 0.00048709271532716524,
      "loss": 5.3828,
      "mean_token_accuracy": 0.16515607237815857,
      "num_tokens": 26637630.0,
      "step": 12835
    },
    {
      "entropy": 5.634595155715942,
      "epoch": 1.1647314949201741,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004870819116813434,
      "loss": 5.4468,
      "mean_token_accuracy": 0.17313675433397294,
      "num_tokens": 26648683.0,
      "step": 12840
    },
    {
      "entropy": 5.702267360687256,
      "epoch": 1.1651850507982584,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00048707110364961363,
      "loss": 5.3989,
      "mean_token_accuracy": 0.167598594725132,
      "num_tokens": 26659573.0,
      "step": 12845
    },
    {
      "entropy": 5.5648540496826175,
      "epoch": 1.1656386066763424,
      "grad_norm": 2.328125,
      "learning_rate": 0.00048706029123219937,
      "loss": 5.3016,
      "mean_token_accuracy": 0.17497960329055787,
      "num_tokens": 26669384.0,
      "step": 12850
    },
    {
      "entropy": 5.638215684890747,
      "epoch": 1.1660921625544267,
      "grad_norm": 2.328125,
      "learning_rate": 0.00048704947442932426,
      "loss": 5.462,
      "mean_token_accuracy": 0.16881868094205857,
      "num_tokens": 26679721.0,
      "step": 12855
    },
    {
      "entropy": 5.654860210418701,
      "epoch": 1.166545718432511,
      "grad_norm": 2.421875,
      "learning_rate": 0.00048703865324121196,
      "loss": 5.3309,
      "mean_token_accuracy": 0.179543574154377,
      "num_tokens": 26689923.0,
      "step": 12860
    },
    {
      "entropy": 5.519535636901855,
      "epoch": 1.166999274310595,
      "grad_norm": 2.203125,
      "learning_rate": 0.00048702782766808637,
      "loss": 5.3272,
      "mean_token_accuracy": 0.1707490310072899,
      "num_tokens": 26699953.0,
      "step": 12865
    },
    {
      "entropy": 5.60990538597107,
      "epoch": 1.1674528301886793,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004870169977101711,
      "loss": 5.5146,
      "mean_token_accuracy": 0.16255164593458177,
      "num_tokens": 26712099.0,
      "step": 12870
    },
    {
      "entropy": 5.675438261032104,
      "epoch": 1.1679063860667633,
      "grad_norm": 2.640625,
      "learning_rate": 0.00048700616336769037,
      "loss": 5.3905,
      "mean_token_accuracy": 0.17527406960725783,
      "num_tokens": 26722334.0,
      "step": 12875
    },
    {
      "entropy": 5.627234649658203,
      "epoch": 1.1683599419448476,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048699532464086817,
      "loss": 5.3362,
      "mean_token_accuracy": 0.1743389457464218,
      "num_tokens": 26732470.0,
      "step": 12880
    },
    {
      "entropy": 5.580811023712158,
      "epoch": 1.1688134978229319,
      "grad_norm": 2.65625,
      "learning_rate": 0.00048698448152992865,
      "loss": 5.3575,
      "mean_token_accuracy": 0.17156845331192017,
      "num_tokens": 26742015.0,
      "step": 12885
    },
    {
      "entropy": 5.554739141464234,
      "epoch": 1.169267053701016,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048697363403509605,
      "loss": 5.3612,
      "mean_token_accuracy": 0.1754850775003433,
      "num_tokens": 26753231.0,
      "step": 12890
    },
    {
      "entropy": 5.592912435531616,
      "epoch": 1.1697206095791002,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004869627821565946,
      "loss": 5.3295,
      "mean_token_accuracy": 0.17159301191568374,
      "num_tokens": 26763895.0,
      "step": 12895
    },
    {
      "entropy": 5.644183826446533,
      "epoch": 1.1701741654571842,
      "grad_norm": 3.9375,
      "learning_rate": 0.0004869519258946489,
      "loss": 5.4291,
      "mean_token_accuracy": 0.16817592978477477,
      "num_tokens": 26774208.0,
      "step": 12900
    },
    {
      "entropy": 5.620671081542969,
      "epoch": 1.1706277213352685,
      "grad_norm": 2.375,
      "learning_rate": 0.00048694106524948324,
      "loss": 5.3838,
      "mean_token_accuracy": 0.1723668321967125,
      "num_tokens": 26785554.0,
      "step": 12905
    },
    {
      "entropy": 5.626670503616333,
      "epoch": 1.1710812772133528,
      "grad_norm": 2.65625,
      "learning_rate": 0.00048693020022132243,
      "loss": 5.3172,
      "mean_token_accuracy": 0.17295462340116502,
      "num_tokens": 26795488.0,
      "step": 12910
    },
    {
      "entropy": 5.5940498352050785,
      "epoch": 1.1715348330914368,
      "grad_norm": 1.765625,
      "learning_rate": 0.000486919330810391,
      "loss": 5.4447,
      "mean_token_accuracy": 0.16420766115188598,
      "num_tokens": 26806801.0,
      "step": 12915
    },
    {
      "entropy": 5.669963407516479,
      "epoch": 1.171988388969521,
      "grad_norm": 3.859375,
      "learning_rate": 0.0004869084570169139,
      "loss": 5.37,
      "mean_token_accuracy": 0.17365534603595734,
      "num_tokens": 26816869.0,
      "step": 12920
    },
    {
      "entropy": 5.6147373676300045,
      "epoch": 1.172441944847605,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048689757884111585,
      "loss": 5.3312,
      "mean_token_accuracy": 0.1720971494913101,
      "num_tokens": 26827003.0,
      "step": 12925
    },
    {
      "entropy": 5.533352851867676,
      "epoch": 1.1728955007256894,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00048688669628322185,
      "loss": 5.319,
      "mean_token_accuracy": 0.17673178762197495,
      "num_tokens": 26837065.0,
      "step": 12930
    },
    {
      "entropy": 5.516833209991455,
      "epoch": 1.1733490566037736,
      "grad_norm": 2.53125,
      "learning_rate": 0.000486875809343457,
      "loss": 5.2684,
      "mean_token_accuracy": 0.17590547055006028,
      "num_tokens": 26847487.0,
      "step": 12935
    },
    {
      "entropy": 5.595408296585083,
      "epoch": 1.1738026124818577,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048686491802204645,
      "loss": 5.3469,
      "mean_token_accuracy": 0.18359509110450745,
      "num_tokens": 26858407.0,
      "step": 12940
    },
    {
      "entropy": 5.6458361625671385,
      "epoch": 1.174256168359942,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004868540223192154,
      "loss": 5.3765,
      "mean_token_accuracy": 0.17196135371923446,
      "num_tokens": 26868469.0,
      "step": 12945
    },
    {
      "entropy": 5.61460919380188,
      "epoch": 1.1747097242380262,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004868431222351892,
      "loss": 5.4117,
      "mean_token_accuracy": 0.17240348756313323,
      "num_tokens": 26879580.0,
      "step": 12950
    },
    {
      "entropy": 5.578996324539185,
      "epoch": 1.1751632801161103,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048683221777019313,
      "loss": 5.3781,
      "mean_token_accuracy": 0.17557137757539748,
      "num_tokens": 26889513.0,
      "step": 12955
    },
    {
      "entropy": 5.623287105560303,
      "epoch": 1.1756168359941945,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048682130892445293,
      "loss": 5.4709,
      "mean_token_accuracy": 0.1664819225668907,
      "num_tokens": 26900792.0,
      "step": 12960
    },
    {
      "entropy": 5.594268465042115,
      "epoch": 1.1760703918722786,
      "grad_norm": 1.890625,
      "learning_rate": 0.000486810395698194,
      "loss": 5.3259,
      "mean_token_accuracy": 0.17498817890882493,
      "num_tokens": 26911298.0,
      "step": 12965
    },
    {
      "entropy": 5.617185020446778,
      "epoch": 1.1765239477503628,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00048679947809164223,
      "loss": 5.404,
      "mean_token_accuracy": 0.1651250973343849,
      "num_tokens": 26921545.0,
      "step": 12970
    },
    {
      "entropy": 5.6643743991851805,
      "epoch": 1.176977503628447,
      "grad_norm": 3.578125,
      "learning_rate": 0.0004867885561050232,
      "loss": 5.3593,
      "mean_token_accuracy": 0.18152386844158172,
      "num_tokens": 26932140.0,
      "step": 12975
    },
    {
      "entropy": 5.673213815689087,
      "epoch": 1.1774310595065312,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004867776297385629,
      "loss": 5.3986,
      "mean_token_accuracy": 0.16779907494783403,
      "num_tokens": 26942634.0,
      "step": 12980
    },
    {
      "entropy": 5.594217157363891,
      "epoch": 1.1778846153846154,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048676669899248723,
      "loss": 5.3643,
      "mean_token_accuracy": 0.17593518346548082,
      "num_tokens": 26952758.0,
      "step": 12985
    },
    {
      "entropy": 5.538848495483398,
      "epoch": 1.1783381712626997,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004867557638670222,
      "loss": 5.3164,
      "mean_token_accuracy": 0.17142992168664933,
      "num_tokens": 26963439.0,
      "step": 12990
    },
    {
      "entropy": 5.598716974258423,
      "epoch": 1.1787917271407837,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004867448243623941,
      "loss": 5.3558,
      "mean_token_accuracy": 0.16914114207029343,
      "num_tokens": 26974331.0,
      "step": 12995
    },
    {
      "entropy": 5.58670334815979,
      "epoch": 1.179245283018868,
      "grad_norm": 4.1875,
      "learning_rate": 0.000486733880478829,
      "loss": 5.2362,
      "mean_token_accuracy": 0.17722886502742768,
      "num_tokens": 26983416.0,
      "step": 13000
    },
    {
      "entropy": 5.567386436462402,
      "epoch": 1.179698838896952,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048672293221655333,
      "loss": 5.3649,
      "mean_token_accuracy": 0.17718790024518966,
      "num_tokens": 26994176.0,
      "step": 13005
    },
    {
      "entropy": 5.563203144073486,
      "epoch": 1.1801523947750363,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004867119795757935,
      "loss": 5.3592,
      "mean_token_accuracy": 0.16978540718555452,
      "num_tokens": 27003295.0,
      "step": 13010
    },
    {
      "entropy": 5.594437217712402,
      "epoch": 1.1806059506531206,
      "grad_norm": 2.203125,
      "learning_rate": 0.000486701022556776,
      "loss": 5.3139,
      "mean_token_accuracy": 0.17459460943937302,
      "num_tokens": 27014018.0,
      "step": 13015
    },
    {
      "entropy": 5.588791704177856,
      "epoch": 1.1810595065312046,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004866900611597273,
      "loss": 5.3457,
      "mean_token_accuracy": 0.17550429105758666,
      "num_tokens": 27024706.0,
      "step": 13020
    },
    {
      "entropy": 5.626877975463867,
      "epoch": 1.1815130624092889,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048667909538487427,
      "loss": 5.3552,
      "mean_token_accuracy": 0.17770835310220717,
      "num_tokens": 27035801.0,
      "step": 13025
    },
    {
      "entropy": 5.519720411300659,
      "epoch": 1.181966618287373,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004866681252324436,
      "loss": 5.2525,
      "mean_token_accuracy": 0.1783450052142143,
      "num_tokens": 27046596.0,
      "step": 13030
    },
    {
      "entropy": 5.595010089874267,
      "epoch": 1.1824201741654572,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00048665715070266214,
      "loss": 5.333,
      "mean_token_accuracy": 0.16773384809494019,
      "num_tokens": 27057057.0,
      "step": 13035
    },
    {
      "entropy": 5.4910078048706055,
      "epoch": 1.1828737300435415,
      "grad_norm": 2.3125,
      "learning_rate": 0.00048664617179575685,
      "loss": 5.3074,
      "mean_token_accuracy": 0.17854873538017274,
      "num_tokens": 27067947.0,
      "step": 13040
    },
    {
      "entropy": 5.592524147033691,
      "epoch": 1.1833272859216255,
      "grad_norm": 2.90625,
      "learning_rate": 0.0004866351885119548,
      "loss": 5.3447,
      "mean_token_accuracy": 0.17544471025466918,
      "num_tokens": 27077503.0,
      "step": 13045
    },
    {
      "entropy": 5.6903480052947994,
      "epoch": 1.1837808417997098,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004866242008514832,
      "loss": 5.4714,
      "mean_token_accuracy": 0.16892702728509904,
      "num_tokens": 27088298.0,
      "step": 13050
    },
    {
      "entropy": 5.549736833572387,
      "epoch": 1.1842343976777938,
      "grad_norm": 2.421875,
      "learning_rate": 0.00048661320881456904,
      "loss": 5.3289,
      "mean_token_accuracy": 0.16736720353364945,
      "num_tokens": 27097841.0,
      "step": 13055
    },
    {
      "entropy": 5.486719179153442,
      "epoch": 1.184687953555878,
      "grad_norm": 3.21875,
      "learning_rate": 0.0004866022124014399,
      "loss": 5.2632,
      "mean_token_accuracy": 0.17707863748073577,
      "num_tokens": 27108263.0,
      "step": 13060
    },
    {
      "entropy": 5.576300191879272,
      "epoch": 1.1851415094339623,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048659121161232313,
      "loss": 5.3333,
      "mean_token_accuracy": 0.17358799129724503,
      "num_tokens": 27119339.0,
      "step": 13065
    },
    {
      "entropy": 5.637163829803467,
      "epoch": 1.1855950653120464,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004865802064474461,
      "loss": 5.3895,
      "mean_token_accuracy": 0.16541709303855895,
      "num_tokens": 27130393.0,
      "step": 13070
    },
    {
      "entropy": 5.6292026996612545,
      "epoch": 1.1860486211901307,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048656919690703653,
      "loss": 5.3833,
      "mean_token_accuracy": 0.16610905081033706,
      "num_tokens": 27141156.0,
      "step": 13075
    },
    {
      "entropy": 5.6156259059906,
      "epoch": 1.1865021770682147,
      "grad_norm": 2.125,
      "learning_rate": 0.000486558182991322,
      "loss": 5.4215,
      "mean_token_accuracy": 0.17152254432439804,
      "num_tokens": 27151630.0,
      "step": 13080
    },
    {
      "entropy": 5.613323926925659,
      "epoch": 1.186955732946299,
      "grad_norm": 2.375,
      "learning_rate": 0.0004865471647005304,
      "loss": 5.2724,
      "mean_token_accuracy": 0.17786764204502106,
      "num_tokens": 27160734.0,
      "step": 13085
    },
    {
      "entropy": 5.577333402633667,
      "epoch": 1.1874092888243832,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048653614203488947,
      "loss": 5.3561,
      "mean_token_accuracy": 0.17039074152708053,
      "num_tokens": 27171584.0,
      "step": 13090
    },
    {
      "entropy": 5.536734628677368,
      "epoch": 1.1878628447024673,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004865251149946273,
      "loss": 5.2787,
      "mean_token_accuracy": 0.1836892694234848,
      "num_tokens": 27180817.0,
      "step": 13095
    },
    {
      "entropy": 5.608707857131958,
      "epoch": 1.1883164005805515,
      "grad_norm": 2.1875,
      "learning_rate": 0.00048651408357997173,
      "loss": 5.3573,
      "mean_token_accuracy": 0.16944698989391327,
      "num_tokens": 27190363.0,
      "step": 13100
    },
    {
      "entropy": 5.6198090553283695,
      "epoch": 1.1887699564586356,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048650304779115114,
      "loss": 5.3158,
      "mean_token_accuracy": 0.17619554400444032,
      "num_tokens": 27202181.0,
      "step": 13105
    },
    {
      "entropy": 5.55724778175354,
      "epoch": 1.1892235123367199,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004864920076283935,
      "loss": 5.3772,
      "mean_token_accuracy": 0.16659168899059296,
      "num_tokens": 27212514.0,
      "step": 13110
    },
    {
      "entropy": 5.585368537902832,
      "epoch": 1.1896770682148041,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004864809630919273,
      "loss": 5.3495,
      "mean_token_accuracy": 0.17874616682529448,
      "num_tokens": 27223352.0,
      "step": 13115
    },
    {
      "entropy": 5.586872148513794,
      "epoch": 1.1901306240928882,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004864699141819809,
      "loss": 5.2721,
      "mean_token_accuracy": 0.17861711233854294,
      "num_tokens": 27232643.0,
      "step": 13120
    },
    {
      "entropy": 5.561257791519165,
      "epoch": 1.1905841799709724,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048645886089878276,
      "loss": 5.3325,
      "mean_token_accuracy": 0.17624130398035048,
      "num_tokens": 27243793.0,
      "step": 13125
    },
    {
      "entropy": 5.592378807067871,
      "epoch": 1.1910377358490567,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048644780324256153,
      "loss": 5.3494,
      "mean_token_accuracy": 0.17617304623126984,
      "num_tokens": 27254554.0,
      "step": 13130
    },
    {
      "entropy": 5.608823251724243,
      "epoch": 1.1914912917271407,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004864367412135458,
      "loss": 5.3982,
      "mean_token_accuracy": 0.17065580040216446,
      "num_tokens": 27265842.0,
      "step": 13135
    },
    {
      "entropy": 5.6333879947662355,
      "epoch": 1.191944847605225,
      "grad_norm": 1.953125,
      "learning_rate": 0.00048642567481196446,
      "loss": 5.4066,
      "mean_token_accuracy": 0.16704277396202089,
      "num_tokens": 27277617.0,
      "step": 13140
    },
    {
      "entropy": 5.576776742935181,
      "epoch": 1.192398403483309,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004864146040380462,
      "loss": 5.3291,
      "mean_token_accuracy": 0.17662875801324845,
      "num_tokens": 27287450.0,
      "step": 13145
    },
    {
      "entropy": 5.597208642959595,
      "epoch": 1.1928519593613933,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004864035288920202,
      "loss": 5.3378,
      "mean_token_accuracy": 0.17345335930585862,
      "num_tokens": 27297010.0,
      "step": 13150
    },
    {
      "entropy": 5.553128576278686,
      "epoch": 1.1933055152394776,
      "grad_norm": 2.9375,
      "learning_rate": 0.00048639244937411516,
      "loss": 5.3894,
      "mean_token_accuracy": 0.17080822885036467,
      "num_tokens": 27306874.0,
      "step": 13155
    },
    {
      "entropy": 5.692587232589721,
      "epoch": 1.1937590711175616,
      "grad_norm": 2.3125,
      "learning_rate": 0.00048638136548456045,
      "loss": 5.4039,
      "mean_token_accuracy": 0.1690503478050232,
      "num_tokens": 27318762.0,
      "step": 13160
    },
    {
      "entropy": 5.568741178512573,
      "epoch": 1.194212626995646,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048637027722358536,
      "loss": 5.2515,
      "mean_token_accuracy": 0.18417855054140092,
      "num_tokens": 27327231.0,
      "step": 13165
    },
    {
      "entropy": 5.503830003738403,
      "epoch": 1.1946661828737302,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048635918459141893,
      "loss": 5.3024,
      "mean_token_accuracy": 0.17428579032421113,
      "num_tokens": 27338517.0,
      "step": 13170
    },
    {
      "entropy": 5.6223094940185545,
      "epoch": 1.1951197387518142,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048634808758829073,
      "loss": 5.3705,
      "mean_token_accuracy": 0.16856776028871537,
      "num_tokens": 27348861.0,
      "step": 13175
    },
    {
      "entropy": 5.590210628509522,
      "epoch": 1.1955732946298985,
      "grad_norm": 2.953125,
      "learning_rate": 0.00048633698621443024,
      "loss": 5.306,
      "mean_token_accuracy": 0.18056849539279937,
      "num_tokens": 27358610.0,
      "step": 13180
    },
    {
      "entropy": 5.535129499435425,
      "epoch": 1.1960268505079825,
      "grad_norm": 1.921875,
      "learning_rate": 0.000486325880470067,
      "loss": 5.4086,
      "mean_token_accuracy": 0.16984545737504958,
      "num_tokens": 27369707.0,
      "step": 13185
    },
    {
      "entropy": 5.574238729476929,
      "epoch": 1.1964804063860668,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004863147703554307,
      "loss": 5.3555,
      "mean_token_accuracy": 0.1724462628364563,
      "num_tokens": 27381030.0,
      "step": 13190
    },
    {
      "entropy": 5.636377620697021,
      "epoch": 1.196933962264151,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048630365587075115,
      "loss": 5.3831,
      "mean_token_accuracy": 0.17001646161079406,
      "num_tokens": 27390866.0,
      "step": 13195
    },
    {
      "entropy": 5.607774686813355,
      "epoch": 1.197387518142235,
      "grad_norm": 1.984375,
      "learning_rate": 0.00048629253701625816,
      "loss": 5.2915,
      "mean_token_accuracy": 0.17824333310127258,
      "num_tokens": 27400724.0,
      "step": 13200
    },
    {
      "entropy": 5.584063959121704,
      "epoch": 1.1978410740203194,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004862814137921816,
      "loss": 5.3277,
      "mean_token_accuracy": 0.17415580600500108,
      "num_tokens": 27411212.0,
      "step": 13205
    },
    {
      "entropy": 5.567069482803345,
      "epoch": 1.1982946298984034,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00048627028619875157,
      "loss": 5.3742,
      "mean_token_accuracy": 0.16939998269081116,
      "num_tokens": 27421279.0,
      "step": 13210
    },
    {
      "entropy": 5.595053100585938,
      "epoch": 1.1987481857764877,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004862591542361982,
      "loss": 5.3536,
      "mean_token_accuracy": 0.17195852249860763,
      "num_tokens": 27432520.0,
      "step": 13215
    },
    {
      "entropy": 5.583989191055298,
      "epoch": 1.199201741654572,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004862480179047516,
      "loss": 5.2826,
      "mean_token_accuracy": 0.17873045057058334,
      "num_tokens": 27442246.0,
      "step": 13220
    },
    {
      "entropy": 5.600679969787597,
      "epoch": 1.199655297532656,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004862368772046423,
      "loss": 5.3125,
      "mean_token_accuracy": 0.17643520534038543,
      "num_tokens": 27452134.0,
      "step": 13225
    },
    {
      "entropy": 5.548134994506836,
      "epoch": 1.2001088534107403,
      "grad_norm": 2.328125,
      "learning_rate": 0.00048622573213610057,
      "loss": 5.3689,
      "mean_token_accuracy": 0.16806789636611938,
      "num_tokens": 27462253.0,
      "step": 13230
    },
    {
      "entropy": 5.59502592086792,
      "epoch": 1.2005624092888243,
      "grad_norm": 3.15625,
      "learning_rate": 0.00048621458269935676,
      "loss": 5.3172,
      "mean_token_accuracy": 0.17468537986278534,
      "num_tokens": 27473131.0,
      "step": 13235
    },
    {
      "entropy": 5.609209251403809,
      "epoch": 1.2010159651669086,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048620342889464153,
      "loss": 5.3309,
      "mean_token_accuracy": 0.17891620099544525,
      "num_tokens": 27483683.0,
      "step": 13240
    },
    {
      "entropy": 5.522462654113769,
      "epoch": 1.2014695210449928,
      "grad_norm": 1.75,
      "learning_rate": 0.00048619227072218565,
      "loss": 5.3043,
      "mean_token_accuracy": 0.17087059319019318,
      "num_tokens": 27493679.0,
      "step": 13245
    },
    {
      "entropy": 5.583718633651733,
      "epoch": 1.2019230769230769,
      "grad_norm": 2.125,
      "learning_rate": 0.0004861811081822198,
      "loss": 5.306,
      "mean_token_accuracy": 0.18382486552000046,
      "num_tokens": 27503897.0,
      "step": 13250
    },
    {
      "entropy": 5.600480222702027,
      "epoch": 1.2023766328011611,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004861699412749748,
      "loss": 5.402,
      "mean_token_accuracy": 0.1658663794398308,
      "num_tokens": 27514560.0,
      "step": 13255
    },
    {
      "entropy": 5.528561019897461,
      "epoch": 1.2028301886792452,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004861587700006816,
      "loss": 5.3039,
      "mean_token_accuracy": 0.17578955888748168,
      "num_tokens": 27525594.0,
      "step": 13260
    },
    {
      "entropy": 5.517774438858032,
      "epoch": 1.2032837445573294,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048614759435957114,
      "loss": 5.2611,
      "mean_token_accuracy": 0.18220822513103485,
      "num_tokens": 27535676.0,
      "step": 13265
    },
    {
      "entropy": 5.53112564086914,
      "epoch": 1.2037373004354137,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004861364143518747,
      "loss": 5.2766,
      "mean_token_accuracy": 0.16874281764030458,
      "num_tokens": 27545520.0,
      "step": 13270
    },
    {
      "entropy": 5.659768342971802,
      "epoch": 1.2041908563134978,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004861252299778234,
      "loss": 5.3829,
      "mean_token_accuracy": 0.16583643853664398,
      "num_tokens": 27556866.0,
      "step": 13275
    },
    {
      "entropy": 5.575717735290527,
      "epoch": 1.204644412191582,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048611404123764855,
      "loss": 5.2993,
      "mean_token_accuracy": 0.17845721989870073,
      "num_tokens": 27566951.0,
      "step": 13280
    },
    {
      "entropy": 5.564585971832275,
      "epoch": 1.205097968069666,
      "grad_norm": 1.875,
      "learning_rate": 0.00048610284813158154,
      "loss": 5.4067,
      "mean_token_accuracy": 0.17460761070251465,
      "num_tokens": 27576796.0,
      "step": 13285
    },
    {
      "entropy": 5.496961212158203,
      "epoch": 1.2055515239477503,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004860916506598538,
      "loss": 5.2609,
      "mean_token_accuracy": 0.17227614670991898,
      "num_tokens": 27586902.0,
      "step": 13290
    },
    {
      "entropy": 5.658858919143677,
      "epoch": 1.2060050798258346,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048608044882269695,
      "loss": 5.4709,
      "mean_token_accuracy": 0.16562795490026475,
      "num_tokens": 27599232.0,
      "step": 13295
    },
    {
      "entropy": 5.589589643478393,
      "epoch": 1.2064586357039186,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004860692426203427,
      "loss": 5.2515,
      "mean_token_accuracy": 0.17265625596046447,
      "num_tokens": 27609022.0,
      "step": 13300
    },
    {
      "entropy": 5.64921293258667,
      "epoch": 1.206912191582003,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004860580320530226,
      "loss": 5.4129,
      "mean_token_accuracy": 0.17320955991744996,
      "num_tokens": 27619829.0,
      "step": 13305
    },
    {
      "entropy": 5.578033590316773,
      "epoch": 1.2073657474600872,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004860468171209687,
      "loss": 5.3955,
      "mean_token_accuracy": 0.1777842402458191,
      "num_tokens": 27629337.0,
      "step": 13310
    },
    {
      "entropy": 5.591690254211426,
      "epoch": 1.2078193033381712,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004860355978244128,
      "loss": 5.4108,
      "mean_token_accuracy": 0.17662533670663833,
      "num_tokens": 27639377.0,
      "step": 13315
    },
    {
      "entropy": 5.6097251892089846,
      "epoch": 1.2082728592162555,
      "grad_norm": 2.90625,
      "learning_rate": 0.000486024374163587,
      "loss": 5.2975,
      "mean_token_accuracy": 0.17844983637332917,
      "num_tokens": 27649989.0,
      "step": 13320
    },
    {
      "entropy": 5.612568140029907,
      "epoch": 1.2087264150943395,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048601314613872334,
      "loss": 5.3911,
      "mean_token_accuracy": 0.16680487841367722,
      "num_tokens": 27659967.0,
      "step": 13325
    },
    {
      "entropy": 5.595348501205445,
      "epoch": 1.2091799709724238,
      "grad_norm": 1.9609375,
      "learning_rate": 0.000486001913750054,
      "loss": 5.3346,
      "mean_token_accuracy": 0.1766873061656952,
      "num_tokens": 27670099.0,
      "step": 13330
    },
    {
      "entropy": 5.535204839706421,
      "epoch": 1.209633526850508,
      "grad_norm": 2.96875,
      "learning_rate": 0.0004859906769978113,
      "loss": 5.3939,
      "mean_token_accuracy": 0.16544295102357864,
      "num_tokens": 27681497.0,
      "step": 13335
    },
    {
      "entropy": 5.67044677734375,
      "epoch": 1.2100870827285921,
      "grad_norm": 2.78125,
      "learning_rate": 0.00048597943588222776,
      "loss": 5.4846,
      "mean_token_accuracy": 0.16721213459968567,
      "num_tokens": 27691828.0,
      "step": 13340
    },
    {
      "entropy": 5.623655319213867,
      "epoch": 1.2105406386066764,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048596819040353564,
      "loss": 5.2936,
      "mean_token_accuracy": 0.17424602806568146,
      "num_tokens": 27701426.0,
      "step": 13345
    },
    {
      "entropy": 5.616913557052612,
      "epoch": 1.2109941944847606,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004859569405619676,
      "loss": 5.2861,
      "mean_token_accuracy": 0.176700422167778,
      "num_tokens": 27711112.0,
      "step": 13350
    },
    {
      "entropy": 5.588740968704224,
      "epoch": 1.2114477503628447,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004859456863577562,
      "loss": 5.4105,
      "mean_token_accuracy": 0.16844981163740158,
      "num_tokens": 27723160.0,
      "step": 13355
    },
    {
      "entropy": 5.626598691940307,
      "epoch": 1.211901306240929,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048593442779113437,
      "loss": 5.376,
      "mean_token_accuracy": 0.16700208634138108,
      "num_tokens": 27733143.0,
      "step": 13360
    },
    {
      "entropy": 5.588902187347412,
      "epoch": 1.212354862119013,
      "grad_norm": 3.203125,
      "learning_rate": 0.0004859231648623348,
      "loss": 5.4027,
      "mean_token_accuracy": 0.17078619003295897,
      "num_tokens": 27743956.0,
      "step": 13365
    },
    {
      "entropy": 5.615359687805176,
      "epoch": 1.2128084179970973,
      "grad_norm": 2.625,
      "learning_rate": 0.00048591189757159035,
      "loss": 5.3447,
      "mean_token_accuracy": 0.18196104913949968,
      "num_tokens": 27754192.0,
      "step": 13370
    },
    {
      "entropy": 5.623263359069824,
      "epoch": 1.2132619738751815,
      "grad_norm": 6.09375,
      "learning_rate": 0.0004859006259191342,
      "loss": 5.4072,
      "mean_token_accuracy": 0.16230515986680985,
      "num_tokens": 27763939.0,
      "step": 13375
    },
    {
      "entropy": 5.609740257263184,
      "epoch": 1.2137155297532656,
      "grad_norm": 2.65625,
      "learning_rate": 0.0004858893499051993,
      "loss": 5.3483,
      "mean_token_accuracy": 0.17944246083498,
      "num_tokens": 27774743.0,
      "step": 13380
    },
    {
      "entropy": 5.629550981521606,
      "epoch": 1.2141690856313498,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004858780695300189,
      "loss": 5.3426,
      "mean_token_accuracy": 0.17732920199632646,
      "num_tokens": 27785344.0,
      "step": 13385
    },
    {
      "entropy": 5.607261800765992,
      "epoch": 1.2146226415094339,
      "grad_norm": 5.1875,
      "learning_rate": 0.0004858667847938263,
      "loss": 5.4238,
      "mean_token_accuracy": 0.16874755024909974,
      "num_tokens": 27797101.0,
      "step": 13390
    },
    {
      "entropy": 5.522930717468261,
      "epoch": 1.2150761973875182,
      "grad_norm": 2.453125,
      "learning_rate": 0.00048585549569685486,
      "loss": 5.2861,
      "mean_token_accuracy": 0.18136875480413436,
      "num_tokens": 27807581.0,
      "step": 13395
    },
    {
      "entropy": 5.539710998535156,
      "epoch": 1.2155297532656024,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00048584420223933805,
      "loss": 5.2781,
      "mean_token_accuracy": 0.17657115757465364,
      "num_tokens": 27818170.0,
      "step": 13400
    },
    {
      "entropy": 5.550384378433227,
      "epoch": 1.2159833091436865,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004858329044215094,
      "loss": 5.2987,
      "mean_token_accuracy": 0.1752548485994339,
      "num_tokens": 27828153.0,
      "step": 13405
    },
    {
      "entropy": 5.57208662033081,
      "epoch": 1.2164368650217707,
      "grad_norm": 1.828125,
      "learning_rate": 0.00048582160224360253,
      "loss": 5.2436,
      "mean_token_accuracy": 0.18096850961446762,
      "num_tokens": 27837681.0,
      "step": 13410
    },
    {
      "entropy": 5.543278360366822,
      "epoch": 1.2168904208998548,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004858102957058513,
      "loss": 5.3849,
      "mean_token_accuracy": 0.17566452026367188,
      "num_tokens": 27847447.0,
      "step": 13415
    },
    {
      "entropy": 5.492272233963012,
      "epoch": 1.217343976777939,
      "grad_norm": 2.34375,
      "learning_rate": 0.00048579898480848933,
      "loss": 5.2476,
      "mean_token_accuracy": 0.17472253888845443,
      "num_tokens": 27858586.0,
      "step": 13420
    },
    {
      "entropy": 5.552965974807739,
      "epoch": 1.2177975326560233,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004857876695517507,
      "loss": 5.3531,
      "mean_token_accuracy": 0.17455514818429946,
      "num_tokens": 27868832.0,
      "step": 13425
    },
    {
      "entropy": 5.57877631187439,
      "epoch": 1.2182510885341074,
      "grad_norm": 2.453125,
      "learning_rate": 0.00048577634993586923,
      "loss": 5.3512,
      "mean_token_accuracy": 0.17423500567674638,
      "num_tokens": 27879339.0,
      "step": 13430
    },
    {
      "entropy": 5.622534132003784,
      "epoch": 1.2187046444121916,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048576502596107924,
      "loss": 5.3645,
      "mean_token_accuracy": 0.17551299035549164,
      "num_tokens": 27889964.0,
      "step": 13435
    },
    {
      "entropy": 5.6464944839477536,
      "epoch": 1.2191582002902757,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004857536976276148,
      "loss": 5.4034,
      "mean_token_accuracy": 0.16876127123832702,
      "num_tokens": 27901004.0,
      "step": 13440
    },
    {
      "entropy": 5.556020641326905,
      "epoch": 1.21961175616836,
      "grad_norm": 2.921875,
      "learning_rate": 0.0004857423649357102,
      "loss": 5.3146,
      "mean_token_accuracy": 0.17484554201364516,
      "num_tokens": 27910788.0,
      "step": 13445
    },
    {
      "entropy": 5.619220113754272,
      "epoch": 1.2200653120464442,
      "grad_norm": 3.296875,
      "learning_rate": 0.00048573102788559974,
      "loss": 5.3619,
      "mean_token_accuracy": 0.17394628673791884,
      "num_tokens": 27921360.0,
      "step": 13450
    },
    {
      "entropy": 5.560455989837647,
      "epoch": 1.2205188679245282,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004857196864775179,
      "loss": 5.3402,
      "mean_token_accuracy": 0.1805930107831955,
      "num_tokens": 27931889.0,
      "step": 13455
    },
    {
      "entropy": 5.541463375091553,
      "epoch": 1.2209724238026125,
      "grad_norm": 1.90625,
      "learning_rate": 0.00048570834071169926,
      "loss": 5.3039,
      "mean_token_accuracy": 0.17708689570426941,
      "num_tokens": 27942049.0,
      "step": 13460
    },
    {
      "entropy": 5.594935178756714,
      "epoch": 1.2214259796806965,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00048569699058837856,
      "loss": 5.4451,
      "mean_token_accuracy": 0.17203628569841384,
      "num_tokens": 27952416.0,
      "step": 13465
    },
    {
      "entropy": 5.566387414932251,
      "epoch": 1.2218795355587808,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004856856361077903,
      "loss": 5.3097,
      "mean_token_accuracy": 0.17856546491384506,
      "num_tokens": 27962009.0,
      "step": 13470
    },
    {
      "entropy": 5.639000082015992,
      "epoch": 1.222333091436865,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004856742772701694,
      "loss": 5.3614,
      "mean_token_accuracy": 0.17219636142253875,
      "num_tokens": 27972175.0,
      "step": 13475
    },
    {
      "entropy": 5.633493041992187,
      "epoch": 1.2227866473149491,
      "grad_norm": 5.25,
      "learning_rate": 0.00048566291407575073,
      "loss": 5.3809,
      "mean_token_accuracy": 0.17799410670995713,
      "num_tokens": 27982899.0,
      "step": 13480
    },
    {
      "entropy": 5.639202070236206,
      "epoch": 1.2232402031930334,
      "grad_norm": 3.046875,
      "learning_rate": 0.00048565154652476937,
      "loss": 5.3189,
      "mean_token_accuracy": 0.17795234322547912,
      "num_tokens": 27993231.0,
      "step": 13485
    },
    {
      "entropy": 5.58167986869812,
      "epoch": 1.2236937590711174,
      "grad_norm": 5.125,
      "learning_rate": 0.0004856401746174603,
      "loss": 5.2742,
      "mean_token_accuracy": 0.17736472338438034,
      "num_tokens": 28003623.0,
      "step": 13490
    },
    {
      "entropy": 5.592493391036987,
      "epoch": 1.2241473149492017,
      "grad_norm": 2.203125,
      "learning_rate": 0.00048562879835405874,
      "loss": 5.2873,
      "mean_token_accuracy": 0.1699166476726532,
      "num_tokens": 28013878.0,
      "step": 13495
    },
    {
      "entropy": 5.539293432235718,
      "epoch": 1.224600870827286,
      "grad_norm": 2.5,
      "learning_rate": 0.0004856174177348,
      "loss": 5.3688,
      "mean_token_accuracy": 0.17718744724988938,
      "num_tokens": 28024224.0,
      "step": 13500
    },
    {
      "entropy": 5.568283557891846,
      "epoch": 1.22505442670537,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004856060327599194,
      "loss": 5.2821,
      "mean_token_accuracy": 0.17411323636770248,
      "num_tokens": 28034819.0,
      "step": 13505
    },
    {
      "entropy": 5.569042110443116,
      "epoch": 1.2255079825834543,
      "grad_norm": 2.25,
      "learning_rate": 0.00048559464342965233,
      "loss": 5.3446,
      "mean_token_accuracy": 0.1745307981967926,
      "num_tokens": 28044240.0,
      "step": 13510
    },
    {
      "entropy": 5.626883220672608,
      "epoch": 1.2259615384615385,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004855832497442344,
      "loss": 5.495,
      "mean_token_accuracy": 0.1612863466143608,
      "num_tokens": 28055046.0,
      "step": 13515
    },
    {
      "entropy": 5.562499094009399,
      "epoch": 1.2264150943396226,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004855718517039012,
      "loss": 5.2842,
      "mean_token_accuracy": 0.17647689431905747,
      "num_tokens": 28065272.0,
      "step": 13520
    },
    {
      "entropy": 5.628909778594971,
      "epoch": 1.2268686502177069,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048556044930888844,
      "loss": 5.2803,
      "mean_token_accuracy": 0.1847059026360512,
      "num_tokens": 28076255.0,
      "step": 13525
    },
    {
      "entropy": 5.5902173042297365,
      "epoch": 1.2273222060957911,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004855490425594319,
      "loss": 5.3973,
      "mean_token_accuracy": 0.1718058779835701,
      "num_tokens": 28086258.0,
      "step": 13530
    },
    {
      "entropy": 5.577108526229859,
      "epoch": 1.2277757619738752,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004855376314557675,
      "loss": 5.3148,
      "mean_token_accuracy": 0.17808407545089722,
      "num_tokens": 28096497.0,
      "step": 13535
    },
    {
      "entropy": 5.576297807693481,
      "epoch": 1.2282293178519594,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004855262159981313,
      "loss": 5.3236,
      "mean_token_accuracy": 0.17770994156599046,
      "num_tokens": 28106282.0,
      "step": 13540
    },
    {
      "entropy": 5.512001419067383,
      "epoch": 1.2286828737300435,
      "grad_norm": 2.34375,
      "learning_rate": 0.00048551479618675925,
      "loss": 5.3752,
      "mean_token_accuracy": 0.17105621099472046,
      "num_tokens": 28115771.0,
      "step": 13545
    },
    {
      "entropy": 5.646786832809449,
      "epoch": 1.2291364296081277,
      "grad_norm": 2.484375,
      "learning_rate": 0.00048550337202188756,
      "loss": 5.3236,
      "mean_token_accuracy": 0.1757451429963112,
      "num_tokens": 28125477.0,
      "step": 13550
    },
    {
      "entropy": 5.666293478012085,
      "epoch": 1.229589985486212,
      "grad_norm": 2.875,
      "learning_rate": 0.0004854919435037525,
      "loss": 5.3482,
      "mean_token_accuracy": 0.1717785194516182,
      "num_tokens": 28135800.0,
      "step": 13555
    },
    {
      "entropy": 5.525774145126343,
      "epoch": 1.230043541364296,
      "grad_norm": 2.625,
      "learning_rate": 0.0004854805106325904,
      "loss": 5.2391,
      "mean_token_accuracy": 0.17402883917093276,
      "num_tokens": 28146779.0,
      "step": 13560
    },
    {
      "entropy": 5.555139207839966,
      "epoch": 1.2304970972423803,
      "grad_norm": 4.4375,
      "learning_rate": 0.0004854690734086378,
      "loss": 5.3687,
      "mean_token_accuracy": 0.1721974492073059,
      "num_tokens": 28156819.0,
      "step": 13565
    },
    {
      "entropy": 5.613695383071899,
      "epoch": 1.2309506531204644,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004854576318321311,
      "loss": 5.3707,
      "mean_token_accuracy": 0.17164791524410247,
      "num_tokens": 28168553.0,
      "step": 13570
    },
    {
      "entropy": 5.61791467666626,
      "epoch": 1.2314042089985486,
      "grad_norm": 1.890625,
      "learning_rate": 0.00048544618590330685,
      "loss": 5.326,
      "mean_token_accuracy": 0.17907538563013076,
      "num_tokens": 28178064.0,
      "step": 13575
    },
    {
      "entropy": 5.644110298156738,
      "epoch": 1.231857764876633,
      "grad_norm": 2.71875,
      "learning_rate": 0.00048543473562240194,
      "loss": 5.4637,
      "mean_token_accuracy": 0.1726391851902008,
      "num_tokens": 28188580.0,
      "step": 13580
    },
    {
      "entropy": 5.630799150466919,
      "epoch": 1.232311320754717,
      "grad_norm": 5.75,
      "learning_rate": 0.000485423280989653,
      "loss": 5.4237,
      "mean_token_accuracy": 0.16658966690301896,
      "num_tokens": 28199955.0,
      "step": 13585
    },
    {
      "entropy": 5.65954737663269,
      "epoch": 1.2327648766328012,
      "grad_norm": 2.984375,
      "learning_rate": 0.000485411822005297,
      "loss": 5.4003,
      "mean_token_accuracy": 0.1691377192735672,
      "num_tokens": 28210506.0,
      "step": 13590
    },
    {
      "entropy": 5.604473209381103,
      "epoch": 1.2332184325108853,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048540035866957097,
      "loss": 5.3207,
      "mean_token_accuracy": 0.1821328490972519,
      "num_tokens": 28220486.0,
      "step": 13595
    },
    {
      "entropy": 5.648378658294678,
      "epoch": 1.2336719883889695,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004853888909827118,
      "loss": 5.4542,
      "mean_token_accuracy": 0.16308983862400056,
      "num_tokens": 28232668.0,
      "step": 13600
    },
    {
      "entropy": 5.587193727493286,
      "epoch": 1.2341255442670538,
      "grad_norm": 3.21875,
      "learning_rate": 0.0004853774189449568,
      "loss": 5.3769,
      "mean_token_accuracy": 0.17219879627227783,
      "num_tokens": 28243733.0,
      "step": 13605
    },
    {
      "entropy": 5.651928329467774,
      "epoch": 1.2345791001451378,
      "grad_norm": 2.390625,
      "learning_rate": 0.00048536594255654315,
      "loss": 5.4177,
      "mean_token_accuracy": 0.17333347350358963,
      "num_tokens": 28254953.0,
      "step": 13610
    },
    {
      "entropy": 5.61195182800293,
      "epoch": 1.235032656023222,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004853544618177083,
      "loss": 5.3674,
      "mean_token_accuracy": 0.17475463896989823,
      "num_tokens": 28264959.0,
      "step": 13615
    },
    {
      "entropy": 5.641897296905517,
      "epoch": 1.2354862119013061,
      "grad_norm": 2.3125,
      "learning_rate": 0.00048534297672868945,
      "loss": 5.3504,
      "mean_token_accuracy": 0.17468301206827164,
      "num_tokens": 28275509.0,
      "step": 13620
    },
    {
      "entropy": 5.570536947250366,
      "epoch": 1.2359397677793904,
      "grad_norm": 2.265625,
      "learning_rate": 0.00048533148728972434,
      "loss": 5.3008,
      "mean_token_accuracy": 0.17754732072353363,
      "num_tokens": 28285983.0,
      "step": 13625
    },
    {
      "entropy": 5.600645112991333,
      "epoch": 1.2363933236574747,
      "grad_norm": 3.234375,
      "learning_rate": 0.00048531999350105037,
      "loss": 5.4008,
      "mean_token_accuracy": 0.16587357074022294,
      "num_tokens": 28296178.0,
      "step": 13630
    },
    {
      "entropy": 5.589850425720215,
      "epoch": 1.2368468795355587,
      "grad_norm": 2.546875,
      "learning_rate": 0.00048530849536290537,
      "loss": 5.3475,
      "mean_token_accuracy": 0.1683177575469017,
      "num_tokens": 28305741.0,
      "step": 13635
    },
    {
      "entropy": 5.583351135253906,
      "epoch": 1.237300435413643,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004852969928755271,
      "loss": 5.3042,
      "mean_token_accuracy": 0.1744295671582222,
      "num_tokens": 28315607.0,
      "step": 13640
    },
    {
      "entropy": 5.553523206710816,
      "epoch": 1.237753991291727,
      "grad_norm": 2.609375,
      "learning_rate": 0.00048528548603915344,
      "loss": 5.3328,
      "mean_token_accuracy": 0.17730236947536468,
      "num_tokens": 28327349.0,
      "step": 13645
    },
    {
      "entropy": 5.583177471160889,
      "epoch": 1.2382075471698113,
      "grad_norm": 2.515625,
      "learning_rate": 0.00048527397485402236,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1699524223804474,
      "num_tokens": 28339949.0,
      "step": 13650
    },
    {
      "entropy": 5.632159280776977,
      "epoch": 1.2386611030478956,
      "grad_norm": 2.375,
      "learning_rate": 0.00048526245932037186,
      "loss": 5.365,
      "mean_token_accuracy": 0.17508769482374192,
      "num_tokens": 28351340.0,
      "step": 13655
    },
    {
      "entropy": 5.51812219619751,
      "epoch": 1.2391146589259796,
      "grad_norm": 2.828125,
      "learning_rate": 0.00048525093943844013,
      "loss": 5.2131,
      "mean_token_accuracy": 0.17740751802921295,
      "num_tokens": 28360929.0,
      "step": 13660
    },
    {
      "entropy": 5.525761604309082,
      "epoch": 1.2395682148040639,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004852394152084654,
      "loss": 5.3223,
      "mean_token_accuracy": 0.17984029799699783,
      "num_tokens": 28370477.0,
      "step": 13665
    },
    {
      "entropy": 5.5181183338165285,
      "epoch": 1.240021770682148,
      "grad_norm": 2.46875,
      "learning_rate": 0.000485227886630686,
      "loss": 5.2394,
      "mean_token_accuracy": 0.18653616011142732,
      "num_tokens": 28380876.0,
      "step": 13670
    },
    {
      "entropy": 5.502352857589722,
      "epoch": 1.2404753265602322,
      "grad_norm": 2.375,
      "learning_rate": 0.0004852163537053403,
      "loss": 5.233,
      "mean_token_accuracy": 0.1757717177271843,
      "num_tokens": 28390685.0,
      "step": 13675
    },
    {
      "entropy": 5.56557879447937,
      "epoch": 1.2409288824383164,
      "grad_norm": 3.296875,
      "learning_rate": 0.0004852048164326669,
      "loss": 5.325,
      "mean_token_accuracy": 0.17249781042337417,
      "num_tokens": 28400236.0,
      "step": 13680
    },
    {
      "entropy": 5.60340313911438,
      "epoch": 1.2413824383164005,
      "grad_norm": 1.75,
      "learning_rate": 0.0004851932748129043,
      "loss": 5.3708,
      "mean_token_accuracy": 0.17916882783174515,
      "num_tokens": 28410353.0,
      "step": 13685
    },
    {
      "entropy": 5.585198068618775,
      "epoch": 1.2418359941944848,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004851817288462912,
      "loss": 5.3003,
      "mean_token_accuracy": 0.1741696134209633,
      "num_tokens": 28420241.0,
      "step": 13690
    },
    {
      "entropy": 5.54184923171997,
      "epoch": 1.242289550072569,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004851701785330663,
      "loss": 5.3786,
      "mean_token_accuracy": 0.16846685856580734,
      "num_tokens": 28430610.0,
      "step": 13695
    },
    {
      "entropy": 5.5550336837768555,
      "epoch": 1.242743105950653,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004851586238734687,
      "loss": 5.2928,
      "mean_token_accuracy": 0.17815205305814744,
      "num_tokens": 28440026.0,
      "step": 13700
    },
    {
      "entropy": 5.591141557693481,
      "epoch": 1.2431966618287373,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004851470648677371,
      "loss": 5.3697,
      "mean_token_accuracy": 0.16559183597564697,
      "num_tokens": 28450707.0,
      "step": 13705
    },
    {
      "entropy": 5.624572372436523,
      "epoch": 1.2436502177068216,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004851355015161107,
      "loss": 5.3057,
      "mean_token_accuracy": 0.17137494087219238,
      "num_tokens": 28462378.0,
      "step": 13710
    },
    {
      "entropy": 5.650377941131592,
      "epoch": 1.2441037735849056,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004851239338188286,
      "loss": 5.3955,
      "mean_token_accuracy": 0.16669329702854158,
      "num_tokens": 28473084.0,
      "step": 13715
    },
    {
      "entropy": 5.557840061187744,
      "epoch": 1.24455732946299,
      "grad_norm": 2.1875,
      "learning_rate": 0.00048511236177612994,
      "loss": 5.3068,
      "mean_token_accuracy": 0.17402911484241484,
      "num_tokens": 28482767.0,
      "step": 13720
    },
    {
      "entropy": 5.5816985130310055,
      "epoch": 1.245010885341074,
      "grad_norm": 2.796875,
      "learning_rate": 0.00048510078538825413,
      "loss": 5.3159,
      "mean_token_accuracy": 0.1766592636704445,
      "num_tokens": 28493274.0,
      "step": 13725
    },
    {
      "entropy": 5.6507079124450685,
      "epoch": 1.2454644412191582,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048508920465544056,
      "loss": 5.312,
      "mean_token_accuracy": 0.1739680737257004,
      "num_tokens": 28503823.0,
      "step": 13730
    },
    {
      "entropy": 5.45209813117981,
      "epoch": 1.2459179970972425,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004850776195779286,
      "loss": 5.3981,
      "mean_token_accuracy": 0.17718724757432938,
      "num_tokens": 28514209.0,
      "step": 13735
    },
    {
      "entropy": 5.613685607910156,
      "epoch": 1.2463715529753265,
      "grad_norm": 2.796875,
      "learning_rate": 0.000485066030155958,
      "loss": 5.3563,
      "mean_token_accuracy": 0.16931882351636887,
      "num_tokens": 28524665.0,
      "step": 13740
    },
    {
      "entropy": 5.625153398513794,
      "epoch": 1.2468251088534108,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004850544363897684,
      "loss": 5.2785,
      "mean_token_accuracy": 0.17885896563529968,
      "num_tokens": 28533639.0,
      "step": 13745
    },
    {
      "entropy": 5.497832870483398,
      "epoch": 1.2472786647314948,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048504283827959947,
      "loss": 5.3084,
      "mean_token_accuracy": 0.1729617953300476,
      "num_tokens": 28544469.0,
      "step": 13750
    },
    {
      "entropy": 5.615324258804321,
      "epoch": 1.2477322206095791,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004850312358256912,
      "loss": 5.283,
      "mean_token_accuracy": 0.17643998712301254,
      "num_tokens": 28554637.0,
      "step": 13755
    },
    {
      "entropy": 5.542525815963745,
      "epoch": 1.2481857764876634,
      "grad_norm": 2.65625,
      "learning_rate": 0.0004850196290282834,
      "loss": 5.3119,
      "mean_token_accuracy": 0.17558214962482452,
      "num_tokens": 28564752.0,
      "step": 13760
    },
    {
      "entropy": 5.5779043674469,
      "epoch": 1.2486393323657474,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00048500801788761613,
      "loss": 5.4338,
      "mean_token_accuracy": 0.17437762171030044,
      "num_tokens": 28575007.0,
      "step": 13765
    },
    {
      "entropy": 5.650508308410645,
      "epoch": 1.2490928882438317,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004849964024039296,
      "loss": 5.3277,
      "mean_token_accuracy": 0.17274573892354966,
      "num_tokens": 28584907.0,
      "step": 13770
    },
    {
      "entropy": 5.670906925201416,
      "epoch": 1.2495464441219157,
      "grad_norm": 2.765625,
      "learning_rate": 0.00048498478257746387,
      "loss": 5.398,
      "mean_token_accuracy": 0.18125055581331254,
      "num_tokens": 28594244.0,
      "step": 13775
    },
    {
      "entropy": 5.625990772247315,
      "epoch": 1.25,
      "grad_norm": 2.375,
      "learning_rate": 0.00048497315840845934,
      "loss": 5.3649,
      "mean_token_accuracy": 0.17237695455551147,
      "num_tokens": 28603816.0,
      "step": 13780
    },
    {
      "entropy": 5.62138671875,
      "epoch": 1.2504535558780843,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004849615298971565,
      "loss": 5.4171,
      "mean_token_accuracy": 0.17015549838542937,
      "num_tokens": 28613067.0,
      "step": 13785
    },
    {
      "entropy": 5.592409896850586,
      "epoch": 1.2509071117561683,
      "grad_norm": 2.1875,
      "learning_rate": 0.00048494989704379556,
      "loss": 5.3479,
      "mean_token_accuracy": 0.17158761173486708,
      "num_tokens": 28623752.0,
      "step": 13790
    },
    {
      "entropy": 5.540713834762573,
      "epoch": 1.2513606676342526,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004849382598486173,
      "loss": 5.3007,
      "mean_token_accuracy": 0.17873719930648804,
      "num_tokens": 28633190.0,
      "step": 13795
    },
    {
      "entropy": 5.698855304718018,
      "epoch": 1.2518142235123366,
      "grad_norm": 2.203125,
      "learning_rate": 0.00048492661831186243,
      "loss": 5.4055,
      "mean_token_accuracy": 0.16575995683670045,
      "num_tokens": 28644567.0,
      "step": 13800
    },
    {
      "entropy": 5.646363115310669,
      "epoch": 1.2522677793904209,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004849149724337715,
      "loss": 5.3863,
      "mean_token_accuracy": 0.16907308250665665,
      "num_tokens": 28654785.0,
      "step": 13805
    },
    {
      "entropy": 5.526299047470093,
      "epoch": 1.2527213352685052,
      "grad_norm": 2.75,
      "learning_rate": 0.00048490332221458543,
      "loss": 5.2843,
      "mean_token_accuracy": 0.1709463194012642,
      "num_tokens": 28665286.0,
      "step": 13810
    },
    {
      "entropy": 5.588380002975464,
      "epoch": 1.2531748911465892,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004848916676545452,
      "loss": 5.3955,
      "mean_token_accuracy": 0.17111329585313798,
      "num_tokens": 28675830.0,
      "step": 13815
    },
    {
      "entropy": 5.665946960449219,
      "epoch": 1.2536284470246735,
      "grad_norm": 2.234375,
      "learning_rate": 0.00048488000875389174,
      "loss": 5.4392,
      "mean_token_accuracy": 0.16648067981004716,
      "num_tokens": 28686640.0,
      "step": 13820
    },
    {
      "entropy": 5.603734540939331,
      "epoch": 1.2540820029027575,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004848683455128663,
      "loss": 5.3293,
      "mean_token_accuracy": 0.17375490069389343,
      "num_tokens": 28697677.0,
      "step": 13825
    },
    {
      "entropy": 5.594356060028076,
      "epoch": 1.2545355587808418,
      "grad_norm": 2.4375,
      "learning_rate": 0.00048485667793170994,
      "loss": 5.3213,
      "mean_token_accuracy": 0.17539387792348862,
      "num_tokens": 28708831.0,
      "step": 13830
    },
    {
      "entropy": 5.5950311660766605,
      "epoch": 1.254989114658926,
      "grad_norm": 2.515625,
      "learning_rate": 0.000484845006010664,
      "loss": 5.3646,
      "mean_token_accuracy": 0.1703229323029518,
      "num_tokens": 28718959.0,
      "step": 13835
    },
    {
      "entropy": 5.591436815261841,
      "epoch": 1.25544267053701,
      "grad_norm": 2.25,
      "learning_rate": 0.0004848333297499699,
      "loss": 5.3165,
      "mean_token_accuracy": 0.1799551621079445,
      "num_tokens": 28728719.0,
      "step": 13840
    },
    {
      "entropy": 5.615367221832275,
      "epoch": 1.2558962264150944,
      "grad_norm": 2.6875,
      "learning_rate": 0.00048482164914986903,
      "loss": 5.4626,
      "mean_token_accuracy": 0.16355045288801193,
      "num_tokens": 28740032.0,
      "step": 13845
    },
    {
      "entropy": 5.569112777709961,
      "epoch": 1.2563497822931784,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048480996421060307,
      "loss": 5.3018,
      "mean_token_accuracy": 0.17964765876531602,
      "num_tokens": 28749374.0,
      "step": 13850
    },
    {
      "entropy": 5.566294431686401,
      "epoch": 1.2568033381712627,
      "grad_norm": 1.9375,
      "learning_rate": 0.00048479827493241347,
      "loss": 5.3306,
      "mean_token_accuracy": 0.17251615673303605,
      "num_tokens": 28760038.0,
      "step": 13855
    },
    {
      "entropy": 5.597704887390137,
      "epoch": 1.257256894049347,
      "grad_norm": 2.25,
      "learning_rate": 0.00048478658131554216,
      "loss": 5.3488,
      "mean_token_accuracy": 0.17368148714303971,
      "num_tokens": 28770247.0,
      "step": 13860
    },
    {
      "entropy": 5.581388139724732,
      "epoch": 1.2577104499274312,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004847748833602308,
      "loss": 5.3331,
      "mean_token_accuracy": 0.17661478668451308,
      "num_tokens": 28780251.0,
      "step": 13865
    },
    {
      "entropy": 5.544677972793579,
      "epoch": 1.2581640058055152,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004847631810667215,
      "loss": 5.3024,
      "mean_token_accuracy": 0.18099457025527954,
      "num_tokens": 28790829.0,
      "step": 13870
    },
    {
      "entropy": 5.493164587020874,
      "epoch": 1.2586175616835993,
      "grad_norm": 2.875,
      "learning_rate": 0.0004847514744352561,
      "loss": 5.2461,
      "mean_token_accuracy": 0.17944961339235305,
      "num_tokens": 28801036.0,
      "step": 13875
    },
    {
      "entropy": 5.480087518692017,
      "epoch": 1.2590711175616836,
      "grad_norm": 2.96875,
      "learning_rate": 0.0004847397634660768,
      "loss": 5.2869,
      "mean_token_accuracy": 0.1799040913581848,
      "num_tokens": 28811013.0,
      "step": 13880
    },
    {
      "entropy": 5.576053714752197,
      "epoch": 1.2595246734397678,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048472804815942575,
      "loss": 5.2989,
      "mean_token_accuracy": 0.17565877139568328,
      "num_tokens": 28821367.0,
      "step": 13885
    },
    {
      "entropy": 5.593064785003662,
      "epoch": 1.259978229317852,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048471632851554517,
      "loss": 5.333,
      "mean_token_accuracy": 0.17698910236358642,
      "num_tokens": 28831077.0,
      "step": 13890
    },
    {
      "entropy": 5.57422251701355,
      "epoch": 1.2604317851959361,
      "grad_norm": 2.125,
      "learning_rate": 0.0004847046045346775,
      "loss": 5.3456,
      "mean_token_accuracy": 0.1693313956260681,
      "num_tokens": 28841490.0,
      "step": 13895
    },
    {
      "entropy": 5.591898250579834,
      "epoch": 1.2608853410740204,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048469287621706516,
      "loss": 5.3676,
      "mean_token_accuracy": 0.16982273012399673,
      "num_tokens": 28852510.0,
      "step": 13900
    },
    {
      "entropy": 5.57936429977417,
      "epoch": 1.2613388969521044,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004846811435629507,
      "loss": 5.3424,
      "mean_token_accuracy": 0.17442412674427032,
      "num_tokens": 28863388.0,
      "step": 13905
    },
    {
      "entropy": 5.494202566146851,
      "epoch": 1.2617924528301887,
      "grad_norm": 2.125,
      "learning_rate": 0.0004846694065725768,
      "loss": 5.3144,
      "mean_token_accuracy": 0.17659474164247513,
      "num_tokens": 28873798.0,
      "step": 13910
    },
    {
      "entropy": 5.588933515548706,
      "epoch": 1.262246008708273,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004846576652461861,
      "loss": 5.3071,
      "mean_token_accuracy": 0.17719524800777436,
      "num_tokens": 28883482.0,
      "step": 13915
    },
    {
      "entropy": 5.571336030960083,
      "epoch": 1.262699564586357,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048464591958402147,
      "loss": 5.288,
      "mean_token_accuracy": 0.17565980404615403,
      "num_tokens": 28893573.0,
      "step": 13920
    },
    {
      "entropy": 5.652734994888306,
      "epoch": 1.2631531204644413,
      "grad_norm": 5.59375,
      "learning_rate": 0.00048463416958632573,
      "loss": 5.4434,
      "mean_token_accuracy": 0.17066749930381775,
      "num_tokens": 28904064.0,
      "step": 13925
    },
    {
      "entropy": 5.637394189834595,
      "epoch": 1.2636066763425253,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048462241525334206,
      "loss": 5.3734,
      "mean_token_accuracy": 0.17052133977413178,
      "num_tokens": 28915329.0,
      "step": 13930
    },
    {
      "entropy": 5.586847114562988,
      "epoch": 1.2640602322206096,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004846106565853133,
      "loss": 5.4485,
      "mean_token_accuracy": 0.1688288614153862,
      "num_tokens": 28925228.0,
      "step": 13935
    },
    {
      "entropy": 5.589742755889892,
      "epoch": 1.2645137880986939,
      "grad_norm": 2.5,
      "learning_rate": 0.00048459889358248284,
      "loss": 5.2948,
      "mean_token_accuracy": 0.17173005044460296,
      "num_tokens": 28935891.0,
      "step": 13940
    },
    {
      "entropy": 5.602288722991943,
      "epoch": 1.264967343976778,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048458712624509383,
      "loss": 5.2801,
      "mean_token_accuracy": 0.17830158472061158,
      "num_tokens": 28946372.0,
      "step": 13945
    },
    {
      "entropy": 5.611473178863525,
      "epoch": 1.2654208998548622,
      "grad_norm": 2.078125,
      "learning_rate": 0.00048457535457338964,
      "loss": 5.3434,
      "mean_token_accuracy": 0.17337343245744705,
      "num_tokens": 28957874.0,
      "step": 13950
    },
    {
      "entropy": 5.6025785446167,
      "epoch": 1.2658744557329462,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004845635785676137,
      "loss": 5.3402,
      "mean_token_accuracy": 0.1646830677986145,
      "num_tokens": 28968023.0,
      "step": 13955
    },
    {
      "entropy": 5.545039701461792,
      "epoch": 1.2663280116110305,
      "grad_norm": 2.46875,
      "learning_rate": 0.00048455179822800957,
      "loss": 5.2344,
      "mean_token_accuracy": 0.1761086106300354,
      "num_tokens": 28977744.0,
      "step": 13960
    },
    {
      "entropy": 5.6339164733886715,
      "epoch": 1.2667815674891147,
      "grad_norm": 2.5,
      "learning_rate": 0.0004845400135548208,
      "loss": 5.4829,
      "mean_token_accuracy": 0.16729221642017364,
      "num_tokens": 28988875.0,
      "step": 13965
    },
    {
      "entropy": 5.566168642044067,
      "epoch": 1.2672351233671988,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004845282245482912,
      "loss": 5.2828,
      "mean_token_accuracy": 0.17748188078403473,
      "num_tokens": 28998562.0,
      "step": 13970
    },
    {
      "entropy": 5.605121660232544,
      "epoch": 1.267688679245283,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004845164312086646,
      "loss": 5.3243,
      "mean_token_accuracy": 0.18036383986473084,
      "num_tokens": 29008783.0,
      "step": 13975
    },
    {
      "entropy": 5.507620191574096,
      "epoch": 1.268142235123367,
      "grad_norm": 2.65625,
      "learning_rate": 0.0004845046335361847,
      "loss": 5.2348,
      "mean_token_accuracy": 0.17694439440965654,
      "num_tokens": 29019303.0,
      "step": 13980
    },
    {
      "entropy": 5.590313148498535,
      "epoch": 1.2685957910014514,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048449283153109566,
      "loss": 5.4508,
      "mean_token_accuracy": 0.16927626430988313,
      "num_tokens": 29029843.0,
      "step": 13985
    },
    {
      "entropy": 5.647569561004639,
      "epoch": 1.2690493468795356,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004844810251936413,
      "loss": 5.3811,
      "mean_token_accuracy": 0.1699263259768486,
      "num_tokens": 29040603.0,
      "step": 13990
    },
    {
      "entropy": 5.578702926635742,
      "epoch": 1.2695029027576197,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004844692145240661,
      "loss": 5.3094,
      "mean_token_accuracy": 0.17437443286180496,
      "num_tokens": 29050599.0,
      "step": 13995
    },
    {
      "entropy": 5.595240592956543,
      "epoch": 1.269956458635704,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004844573995226142,
      "loss": 5.3587,
      "mean_token_accuracy": 0.17279112488031387,
      "num_tokens": 29061292.0,
      "step": 14000
    },
    {
      "entropy": 5.576063728332519,
      "epoch": 1.270410014513788,
      "grad_norm": 2.4375,
      "learning_rate": 0.00048444558018952984,
      "loss": 5.2917,
      "mean_token_accuracy": 0.18096554577350615,
      "num_tokens": 29071779.0,
      "step": 14005
    },
    {
      "entropy": 5.5674886226654055,
      "epoch": 1.2708635703918723,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048443375652505754,
      "loss": 5.2612,
      "mean_token_accuracy": 0.17743561863899232,
      "num_tokens": 29082073.0,
      "step": 14010
    },
    {
      "entropy": 5.5836893081665036,
      "epoch": 1.2713171262699565,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004844219285294418,
      "loss": 5.2429,
      "mean_token_accuracy": 0.17464204579591752,
      "num_tokens": 29092032.0,
      "step": 14015
    },
    {
      "entropy": 5.536452293395996,
      "epoch": 1.2717706821480406,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048441009620292714,
      "loss": 5.2909,
      "mean_token_accuracy": 0.18139055967330933,
      "num_tokens": 29102936.0,
      "step": 14020
    },
    {
      "entropy": 5.584950637817383,
      "epoch": 1.2722242380261248,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004843982595457584,
      "loss": 5.3206,
      "mean_token_accuracy": 0.1767016664147377,
      "num_tokens": 29113385.0,
      "step": 14025
    },
    {
      "entropy": 5.538218832015991,
      "epoch": 1.2726777939042089,
      "grad_norm": 2.421875,
      "learning_rate": 0.00048438641855818026,
      "loss": 5.3484,
      "mean_token_accuracy": 0.17236390858888626,
      "num_tokens": 29125075.0,
      "step": 14030
    },
    {
      "entropy": 5.606749868392944,
      "epoch": 1.2731313497822931,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004843745732404376,
      "loss": 5.2838,
      "mean_token_accuracy": 0.17292884588241578,
      "num_tokens": 29135098.0,
      "step": 14035
    },
    {
      "entropy": 5.621056032180786,
      "epoch": 1.2735849056603774,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004843627235927755,
      "loss": 5.3877,
      "mean_token_accuracy": 0.176491641998291,
      "num_tokens": 29145577.0,
      "step": 14040
    },
    {
      "entropy": 5.535854721069336,
      "epoch": 1.2740384615384617,
      "grad_norm": 3.34375,
      "learning_rate": 0.0004843508696154388,
      "loss": 5.3335,
      "mean_token_accuracy": 0.17485394179821015,
      "num_tokens": 29155905.0,
      "step": 14045
    },
    {
      "entropy": 5.540239381790161,
      "epoch": 1.2744920174165457,
      "grad_norm": 5.1875,
      "learning_rate": 0.00048433901130867283,
      "loss": 5.2407,
      "mean_token_accuracy": 0.17786966264247894,
      "num_tokens": 29164926.0,
      "step": 14050
    },
    {
      "entropy": 5.565501737594604,
      "epoch": 1.2749455732946298,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004843271486727228,
      "loss": 5.2761,
      "mean_token_accuracy": 0.17572084963321685,
      "num_tokens": 29175050.0,
      "step": 14055
    },
    {
      "entropy": 5.542440462112427,
      "epoch": 1.275399129172714,
      "grad_norm": 2.09375,
      "learning_rate": 0.00048431528170783395,
      "loss": 5.3585,
      "mean_token_accuracy": 0.16913863271474838,
      "num_tokens": 29186255.0,
      "step": 14060
    },
    {
      "entropy": 5.603945970535278,
      "epoch": 1.2758526850507983,
      "grad_norm": 4.78125,
      "learning_rate": 0.0004843034104142517,
      "loss": 5.3226,
      "mean_token_accuracy": 0.18160293996334076,
      "num_tokens": 29195639.0,
      "step": 14065
    },
    {
      "entropy": 5.615475225448608,
      "epoch": 1.2763062409288826,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004842915347922216,
      "loss": 5.3545,
      "mean_token_accuracy": 0.16717389822006226,
      "num_tokens": 29206482.0,
      "step": 14070
    },
    {
      "entropy": 5.617571115493774,
      "epoch": 1.2767597968069666,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004842796548419893,
      "loss": 5.319,
      "mean_token_accuracy": 0.17026246339082718,
      "num_tokens": 29217433.0,
      "step": 14075
    },
    {
      "entropy": 5.620957708358764,
      "epoch": 1.2772133526850509,
      "grad_norm": 2.25,
      "learning_rate": 0.00048426777056380035,
      "loss": 5.3775,
      "mean_token_accuracy": 0.165461266040802,
      "num_tokens": 29226193.0,
      "step": 14080
    },
    {
      "entropy": 5.516555881500244,
      "epoch": 1.277666908563135,
      "grad_norm": 2.25,
      "learning_rate": 0.0004842558819579006,
      "loss": 5.3186,
      "mean_token_accuracy": 0.17778911292552949,
      "num_tokens": 29237061.0,
      "step": 14085
    },
    {
      "entropy": 5.512198257446289,
      "epoch": 1.2781204644412192,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004842439890245358,
      "loss": 5.3114,
      "mean_token_accuracy": 0.1845606416463852,
      "num_tokens": 29247689.0,
      "step": 14090
    },
    {
      "entropy": 5.520582437515259,
      "epoch": 1.2785740203193035,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048423209176395204,
      "loss": 5.273,
      "mean_token_accuracy": 0.17814113199710846,
      "num_tokens": 29258235.0,
      "step": 14095
    },
    {
      "entropy": 5.5709623336792,
      "epoch": 1.2790275761973875,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004842201901763954,
      "loss": 5.2842,
      "mean_token_accuracy": 0.1762159213423729,
      "num_tokens": 29268884.0,
      "step": 14100
    },
    {
      "entropy": 5.5821925640106205,
      "epoch": 1.2794811320754718,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004842082842621118,
      "loss": 5.4247,
      "mean_token_accuracy": 0.16802155822515488,
      "num_tokens": 29280484.0,
      "step": 14105
    },
    {
      "entropy": 5.636690616607666,
      "epoch": 1.2799346879535558,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004841963740213476,
      "loss": 5.3665,
      "mean_token_accuracy": 0.17644825279712678,
      "num_tokens": 29291199.0,
      "step": 14110
    },
    {
      "entropy": 5.670208406448364,
      "epoch": 1.28038824383164,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004841844594543491,
      "loss": 5.3411,
      "mean_token_accuracy": 0.1748999461531639,
      "num_tokens": 29302007.0,
      "step": 14115
    },
    {
      "entropy": 5.590766382217407,
      "epoch": 1.2808417997097243,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004841725405613626,
      "loss": 5.3102,
      "mean_token_accuracy": 0.18238283842802047,
      "num_tokens": 29312127.0,
      "step": 14120
    },
    {
      "entropy": 5.5790143489837645,
      "epoch": 1.2812953555878084,
      "grad_norm": 2.75,
      "learning_rate": 0.0004841606173426347,
      "loss": 5.4435,
      "mean_token_accuracy": 0.17037187814712523,
      "num_tokens": 29323691.0,
      "step": 14125
    },
    {
      "entropy": 5.574641561508178,
      "epoch": 1.2817489114658926,
      "grad_norm": 3.078125,
      "learning_rate": 0.00048414868979841197,
      "loss": 5.356,
      "mean_token_accuracy": 0.1665747731924057,
      "num_tokens": 29335203.0,
      "step": 14130
    },
    {
      "entropy": 5.678859043121338,
      "epoch": 1.2822024673439767,
      "grad_norm": 3.703125,
      "learning_rate": 0.00048413675792894105,
      "loss": 5.394,
      "mean_token_accuracy": 0.17578777819871902,
      "num_tokens": 29346422.0,
      "step": 14135
    },
    {
      "entropy": 5.594743442535401,
      "epoch": 1.282656023222061,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004841248217344686,
      "loss": 5.3079,
      "mean_token_accuracy": 0.17514521181583403,
      "num_tokens": 29356084.0,
      "step": 14140
    },
    {
      "entropy": 5.506899070739746,
      "epoch": 1.2831095791001452,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004841128812152415,
      "loss": 5.2926,
      "mean_token_accuracy": 0.17588884979486466,
      "num_tokens": 29366093.0,
      "step": 14145
    },
    {
      "entropy": 5.573301553726196,
      "epoch": 1.2835631349782293,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004841009363715068,
      "loss": 5.3549,
      "mean_token_accuracy": 0.1706303671002388,
      "num_tokens": 29376621.0,
      "step": 14150
    },
    {
      "entropy": 5.647098398208618,
      "epoch": 1.2840166908563135,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004840889872035115,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1716342732310295,
      "num_tokens": 29387119.0,
      "step": 14155
    },
    {
      "entropy": 5.5658587455749515,
      "epoch": 1.2844702467343976,
      "grad_norm": 4.5625,
      "learning_rate": 0.0004840770337115026,
      "loss": 5.3185,
      "mean_token_accuracy": 0.17690143287181853,
      "num_tokens": 29396981.0,
      "step": 14160
    },
    {
      "entropy": 5.581387710571289,
      "epoch": 1.2849238026124818,
      "grad_norm": 2.4375,
      "learning_rate": 0.00048406507589572747,
      "loss": 5.3006,
      "mean_token_accuracy": 0.17793996781110763,
      "num_tokens": 29406737.0,
      "step": 14165
    },
    {
      "entropy": 5.620014095306397,
      "epoch": 1.2853773584905661,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004840531137564331,
      "loss": 5.3334,
      "mean_token_accuracy": 0.17301466912031174,
      "num_tokens": 29417300.0,
      "step": 14170
    },
    {
      "entropy": 5.620226097106934,
      "epoch": 1.2858309143686502,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004840411472938672,
      "loss": 5.2304,
      "mean_token_accuracy": 0.1822351783514023,
      "num_tokens": 29426749.0,
      "step": 14175
    },
    {
      "entropy": 5.50346884727478,
      "epoch": 1.2862844702467344,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004840291765082771,
      "loss": 5.2608,
      "mean_token_accuracy": 0.1771259680390358,
      "num_tokens": 29437264.0,
      "step": 14180
    },
    {
      "entropy": 5.537852430343628,
      "epoch": 1.2867380261248185,
      "grad_norm": 2.375,
      "learning_rate": 0.00048401720139991036,
      "loss": 5.2905,
      "mean_token_accuracy": 0.17025234550237656,
      "num_tokens": 29448192.0,
      "step": 14185
    },
    {
      "entropy": 5.609713506698609,
      "epoch": 1.2871915820029027,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004840052219690146,
      "loss": 5.2026,
      "mean_token_accuracy": 0.18591743111610412,
      "num_tokens": 29457923.0,
      "step": 14190
    },
    {
      "entropy": 5.487228536605835,
      "epoch": 1.287645137880987,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004839932382158376,
      "loss": 5.273,
      "mean_token_accuracy": 0.18596986830234527,
      "num_tokens": 29466914.0,
      "step": 14195
    },
    {
      "entropy": 5.531790113449096,
      "epoch": 1.288098693759071,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004839812501406272,
      "loss": 5.3683,
      "mean_token_accuracy": 0.1714567095041275,
      "num_tokens": 29477648.0,
      "step": 14200
    },
    {
      "entropy": 5.63369517326355,
      "epoch": 1.2885522496371553,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004839692577436313,
      "loss": 5.3379,
      "mean_token_accuracy": 0.17652663439512253,
      "num_tokens": 29487947.0,
      "step": 14205
    },
    {
      "entropy": 5.712900400161743,
      "epoch": 1.2890058055152394,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004839572610250978,
      "loss": 5.4421,
      "mean_token_accuracy": 0.1679430440068245,
      "num_tokens": 29497667.0,
      "step": 14210
    },
    {
      "entropy": 5.595472431182861,
      "epoch": 1.2894593613933236,
      "grad_norm": 3.296875,
      "learning_rate": 0.00048394525998527496,
      "loss": 5.3046,
      "mean_token_accuracy": 0.17251573652029037,
      "num_tokens": 29507065.0,
      "step": 14215
    },
    {
      "entropy": 5.613528299331665,
      "epoch": 1.2899129172714079,
      "grad_norm": 3.6875,
      "learning_rate": 0.0004839332546244109,
      "loss": 5.4146,
      "mean_token_accuracy": 0.1717883124947548,
      "num_tokens": 29517752.0,
      "step": 14220
    },
    {
      "entropy": 5.629713726043701,
      "epoch": 1.2903664731494922,
      "grad_norm": 2.75,
      "learning_rate": 0.00048392124494275385,
      "loss": 5.3221,
      "mean_token_accuracy": 0.17212332785129547,
      "num_tokens": 29527894.0,
      "step": 14225
    },
    {
      "entropy": 5.587793636322021,
      "epoch": 1.2908200290275762,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004839092309405523,
      "loss": 5.3546,
      "mean_token_accuracy": 0.16875882893800737,
      "num_tokens": 29538136.0,
      "step": 14230
    },
    {
      "entropy": 5.598608922958374,
      "epoch": 1.2912735849056602,
      "grad_norm": 3.046875,
      "learning_rate": 0.00048389721261805457,
      "loss": 5.318,
      "mean_token_accuracy": 0.1747458204627037,
      "num_tokens": 29549193.0,
      "step": 14235
    },
    {
      "entropy": 5.530500507354736,
      "epoch": 1.2917271407837445,
      "grad_norm": 1.9375,
      "learning_rate": 0.00048388518997550926,
      "loss": 5.2241,
      "mean_token_accuracy": 0.1818700060248375,
      "num_tokens": 29558889.0,
      "step": 14240
    },
    {
      "entropy": 5.57679009437561,
      "epoch": 1.2921806966618288,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048387316301316493,
      "loss": 5.3605,
      "mean_token_accuracy": 0.1677996590733528,
      "num_tokens": 29569691.0,
      "step": 14245
    },
    {
      "entropy": 5.595895290374756,
      "epoch": 1.292634252539913,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004838611317312705,
      "loss": 5.425,
      "mean_token_accuracy": 0.1709785968065262,
      "num_tokens": 29580459.0,
      "step": 14250
    },
    {
      "entropy": 5.58565616607666,
      "epoch": 1.293087808417997,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004838490961300745,
      "loss": 5.3424,
      "mean_token_accuracy": 0.170879827439785,
      "num_tokens": 29590652.0,
      "step": 14255
    },
    {
      "entropy": 5.574022960662842,
      "epoch": 1.2935413642960814,
      "grad_norm": 2.125,
      "learning_rate": 0.00048383705620982606,
      "loss": 5.3303,
      "mean_token_accuracy": 0.18456486910581588,
      "num_tokens": 29600353.0,
      "step": 14260
    },
    {
      "entropy": 5.573683357238769,
      "epoch": 1.2939949201741654,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048382501197077405,
      "loss": 5.32,
      "mean_token_accuracy": 0.18218221813440322,
      "num_tokens": 29610561.0,
      "step": 14265
    },
    {
      "entropy": 5.56094446182251,
      "epoch": 1.2944484760522497,
      "grad_norm": 2.125,
      "learning_rate": 0.0004838129634131675,
      "loss": 5.3254,
      "mean_token_accuracy": 0.1744862377643585,
      "num_tokens": 29621741.0,
      "step": 14270
    },
    {
      "entropy": 5.595221090316772,
      "epoch": 1.294902031930334,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004838009105372558,
      "loss": 5.2455,
      "mean_token_accuracy": 0.1733754977583885,
      "num_tokens": 29633097.0,
      "step": 14275
    },
    {
      "entropy": 5.595546388626099,
      "epoch": 1.295355587808418,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048378885334328793,
      "loss": 5.3642,
      "mean_token_accuracy": 0.16239650547504425,
      "num_tokens": 29644290.0,
      "step": 14280
    },
    {
      "entropy": 5.664061069488525,
      "epoch": 1.2958091436865022,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004837767918315135,
      "loss": 5.3916,
      "mean_token_accuracy": 0.16826958060264588,
      "num_tokens": 29654977.0,
      "step": 14285
    },
    {
      "entropy": 5.617649936676026,
      "epoch": 1.2962626995645863,
      "grad_norm": 2.46875,
      "learning_rate": 0.00048376472600218174,
      "loss": 5.3937,
      "mean_token_accuracy": 0.1715206876397133,
      "num_tokens": 29665020.0,
      "step": 14290
    },
    {
      "entropy": 5.614870643615722,
      "epoch": 1.2967162554426706,
      "grad_norm": 2.75,
      "learning_rate": 0.0004837526558555422,
      "loss": 5.3494,
      "mean_token_accuracy": 0.17420814484357833,
      "num_tokens": 29675098.0,
      "step": 14295
    },
    {
      "entropy": 5.557895946502685,
      "epoch": 1.2971698113207548,
      "grad_norm": 3.46875,
      "learning_rate": 0.0004837405813918446,
      "loss": 5.2849,
      "mean_token_accuracy": 0.17971768379211425,
      "num_tokens": 29685450.0,
      "step": 14300
    },
    {
      "entropy": 5.542715930938721,
      "epoch": 1.2976233671988389,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048372850261133856,
      "loss": 5.2964,
      "mean_token_accuracy": 0.17652963995933532,
      "num_tokens": 29694818.0,
      "step": 14305
    },
    {
      "entropy": 5.61685094833374,
      "epoch": 1.2980769230769231,
      "grad_norm": 2.515625,
      "learning_rate": 0.00048371641951427393,
      "loss": 5.3715,
      "mean_token_accuracy": 0.17161611020565032,
      "num_tokens": 29705573.0,
      "step": 14310
    },
    {
      "entropy": 5.616137933731079,
      "epoch": 1.2985304789550072,
      "grad_norm": 2.28125,
      "learning_rate": 0.00048370433210090055,
      "loss": 5.3119,
      "mean_token_accuracy": 0.16940421611070633,
      "num_tokens": 29716366.0,
      "step": 14315
    },
    {
      "entropy": 5.55403962135315,
      "epoch": 1.2989840348330914,
      "grad_norm": 2.0,
      "learning_rate": 0.0004836922403714683,
      "loss": 5.29,
      "mean_token_accuracy": 0.17545078992843627,
      "num_tokens": 29727181.0,
      "step": 14320
    },
    {
      "entropy": 5.5547308921813965,
      "epoch": 1.2994375907111757,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048368014432622746,
      "loss": 5.3114,
      "mean_token_accuracy": 0.17786243706941604,
      "num_tokens": 29736960.0,
      "step": 14325
    },
    {
      "entropy": 5.564285802841186,
      "epoch": 1.2998911465892597,
      "grad_norm": 2.796875,
      "learning_rate": 0.000483668043965428,
      "loss": 5.2126,
      "mean_token_accuracy": 0.1873780980706215,
      "num_tokens": 29746354.0,
      "step": 14330
    },
    {
      "entropy": 5.573829126358032,
      "epoch": 1.300344702467344,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004836559392893203,
      "loss": 5.3315,
      "mean_token_accuracy": 0.1724160149693489,
      "num_tokens": 29756627.0,
      "step": 14335
    },
    {
      "entropy": 5.5727660179138185,
      "epoch": 1.300798258345428,
      "grad_norm": 3.75,
      "learning_rate": 0.0004836438302981545,
      "loss": 5.2877,
      "mean_token_accuracy": 0.18192365765571594,
      "num_tokens": 29766684.0,
      "step": 14340
    },
    {
      "entropy": 5.5299266338348385,
      "epoch": 1.3012518142235123,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004836317169921811,
      "loss": 5.3585,
      "mean_token_accuracy": 0.17365316897630692,
      "num_tokens": 29776556.0,
      "step": 14345
    },
    {
      "entropy": 5.653978776931763,
      "epoch": 1.3017053701015966,
      "grad_norm": 3.109375,
      "learning_rate": 0.00048361959937165075,
      "loss": 5.4434,
      "mean_token_accuracy": 0.17234432697296143,
      "num_tokens": 29786896.0,
      "step": 14350
    },
    {
      "entropy": 5.641994857788086,
      "epoch": 1.3021589259796806,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004836074774368138,
      "loss": 5.2559,
      "mean_token_accuracy": 0.1803387701511383,
      "num_tokens": 29797332.0,
      "step": 14355
    },
    {
      "entropy": 5.605131053924561,
      "epoch": 1.302612481857765,
      "grad_norm": 3.40625,
      "learning_rate": 0.00048359535118792116,
      "loss": 5.3357,
      "mean_token_accuracy": 0.17664601057767867,
      "num_tokens": 29806428.0,
      "step": 14360
    },
    {
      "entropy": 5.558047676086426,
      "epoch": 1.303066037735849,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004835832206252235,
      "loss": 5.3291,
      "mean_token_accuracy": 0.17150520533323288,
      "num_tokens": 29816794.0,
      "step": 14365
    },
    {
      "entropy": 5.636795949935913,
      "epoch": 1.3035195936139332,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004835710857489716,
      "loss": 5.3679,
      "mean_token_accuracy": 0.17814830988645552,
      "num_tokens": 29827077.0,
      "step": 14370
    },
    {
      "entropy": 5.582400703430176,
      "epoch": 1.3039731494920175,
      "grad_norm": 2.25,
      "learning_rate": 0.0004835589465594165,
      "loss": 5.317,
      "mean_token_accuracy": 0.17814186364412307,
      "num_tokens": 29837868.0,
      "step": 14375
    },
    {
      "entropy": 5.614265441894531,
      "epoch": 1.3044267053701015,
      "grad_norm": 2.328125,
      "learning_rate": 0.00048354680305680926,
      "loss": 5.3611,
      "mean_token_accuracy": 0.17542893588542938,
      "num_tokens": 29846929.0,
      "step": 14380
    },
    {
      "entropy": 5.527841424942016,
      "epoch": 1.3048802612481858,
      "grad_norm": 1.9375,
      "learning_rate": 0.00048353465524140105,
      "loss": 5.2826,
      "mean_token_accuracy": 0.1783314049243927,
      "num_tokens": 29857905.0,
      "step": 14385
    },
    {
      "entropy": 5.55200047492981,
      "epoch": 1.3053338171262698,
      "grad_norm": 3.71875,
      "learning_rate": 0.000483522503113443,
      "loss": 5.3379,
      "mean_token_accuracy": 0.1743301570415497,
      "num_tokens": 29868230.0,
      "step": 14390
    },
    {
      "entropy": 5.52324366569519,
      "epoch": 1.305787373004354,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048351034667318634,
      "loss": 5.3065,
      "mean_token_accuracy": 0.17327929586172103,
      "num_tokens": 29879224.0,
      "step": 14395
    },
    {
      "entropy": 5.626120519638062,
      "epoch": 1.3062409288824384,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004834981859208827,
      "loss": 5.4481,
      "mean_token_accuracy": 0.16675183176994324,
      "num_tokens": 29890416.0,
      "step": 14400
    },
    {
      "entropy": 5.66015248298645,
      "epoch": 1.3066944847605226,
      "grad_norm": 3.234375,
      "learning_rate": 0.00048348602085678346,
      "loss": 5.3793,
      "mean_token_accuracy": 0.17032576054334642,
      "num_tokens": 29901064.0,
      "step": 14405
    },
    {
      "entropy": 5.730725383758545,
      "epoch": 1.3071480406386067,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048347385148114004,
      "loss": 5.4284,
      "mean_token_accuracy": 0.17141589671373367,
      "num_tokens": 29911084.0,
      "step": 14410
    },
    {
      "entropy": 5.5188093185424805,
      "epoch": 1.3076015965166907,
      "grad_norm": 1.96875,
      "learning_rate": 0.00048346167779420436,
      "loss": 5.2753,
      "mean_token_accuracy": 0.17884353399276734,
      "num_tokens": 29921104.0,
      "step": 14415
    },
    {
      "entropy": 5.587260150909424,
      "epoch": 1.308055152394775,
      "grad_norm": 2.96875,
      "learning_rate": 0.00048344949979622806,
      "loss": 5.3196,
      "mean_token_accuracy": 0.17668967247009276,
      "num_tokens": 29932014.0,
      "step": 14420
    },
    {
      "entropy": 5.558705139160156,
      "epoch": 1.3085087082728593,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048343731748746294,
      "loss": 5.3262,
      "mean_token_accuracy": 0.17766952514648438,
      "num_tokens": 29941707.0,
      "step": 14425
    },
    {
      "entropy": 5.539162826538086,
      "epoch": 1.3089622641509435,
      "grad_norm": 2.828125,
      "learning_rate": 0.000483425130868161,
      "loss": 5.3094,
      "mean_token_accuracy": 0.17264337688684464,
      "num_tokens": 29952082.0,
      "step": 14430
    },
    {
      "entropy": 5.52536096572876,
      "epoch": 1.3094158200290276,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004834129399385742,
      "loss": 5.1823,
      "mean_token_accuracy": 0.1933560460805893,
      "num_tokens": 29961744.0,
      "step": 14435
    },
    {
      "entropy": 5.60400242805481,
      "epoch": 1.3098693759071118,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004834007446989548,
      "loss": 5.301,
      "mean_token_accuracy": 0.18017879426479338,
      "num_tokens": 29972611.0,
      "step": 14440
    },
    {
      "entropy": 5.549803972244263,
      "epoch": 1.3103229317851959,
      "grad_norm": 1.984375,
      "learning_rate": 0.00048338854514955474,
      "loss": 5.219,
      "mean_token_accuracy": 0.1845446929335594,
      "num_tokens": 29982660.0,
      "step": 14445
    },
    {
      "entropy": 5.550955057144165,
      "epoch": 1.3107764876632801,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048337634129062655,
      "loss": 5.348,
      "mean_token_accuracy": 0.16491775959730148,
      "num_tokens": 29993349.0,
      "step": 14450
    },
    {
      "entropy": 5.618444108963013,
      "epoch": 1.3112300435413644,
      "grad_norm": 2.640625,
      "learning_rate": 0.00048336413312242255,
      "loss": 5.4154,
      "mean_token_accuracy": 0.16765930950641633,
      "num_tokens": 30004474.0,
      "step": 14455
    },
    {
      "entropy": 5.631626129150391,
      "epoch": 1.3116835994194485,
      "grad_norm": 1.875,
      "learning_rate": 0.00048335192064519514,
      "loss": 5.3838,
      "mean_token_accuracy": 0.1729219973087311,
      "num_tokens": 30015189.0,
      "step": 14460
    },
    {
      "entropy": 5.615939140319824,
      "epoch": 1.3121371552975327,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048333970385919695,
      "loss": 5.2741,
      "mean_token_accuracy": 0.17484485656023024,
      "num_tokens": 30025491.0,
      "step": 14465
    },
    {
      "entropy": 5.5642561435699465,
      "epoch": 1.3125907111756168,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048332748276468053,
      "loss": 5.3209,
      "mean_token_accuracy": 0.17495635747909546,
      "num_tokens": 30034683.0,
      "step": 14470
    },
    {
      "entropy": 5.520941114425659,
      "epoch": 1.313044267053701,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004833152573618987,
      "loss": 5.2814,
      "mean_token_accuracy": 0.1753089800477028,
      "num_tokens": 30044135.0,
      "step": 14475
    },
    {
      "entropy": 5.5774375915527346,
      "epoch": 1.3134978229317853,
      "grad_norm": 2.796875,
      "learning_rate": 0.00048330302765110436,
      "loss": 5.4012,
      "mean_token_accuracy": 0.17630247473716737,
      "num_tokens": 30054520.0,
      "step": 14480
    },
    {
      "entropy": 5.620419597625732,
      "epoch": 1.3139513788098693,
      "grad_norm": 2.609375,
      "learning_rate": 0.00048329079363255024,
      "loss": 5.3758,
      "mean_token_accuracy": 0.1691085249185562,
      "num_tokens": 30065439.0,
      "step": 14485
    },
    {
      "entropy": 5.61645827293396,
      "epoch": 1.3144049346879536,
      "grad_norm": 2.234375,
      "learning_rate": 0.00048327855530648934,
      "loss": 5.3568,
      "mean_token_accuracy": 0.17591703236103057,
      "num_tokens": 30075362.0,
      "step": 14490
    },
    {
      "entropy": 5.524699974060058,
      "epoch": 1.3148584905660377,
      "grad_norm": 2.140625,
      "learning_rate": 0.000483266312673175,
      "loss": 5.3228,
      "mean_token_accuracy": 0.17203290313482283,
      "num_tokens": 30084913.0,
      "step": 14495
    },
    {
      "entropy": 5.588080596923828,
      "epoch": 1.315312046444122,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004832540657328602,
      "loss": 5.3294,
      "mean_token_accuracy": 0.17741949260234832,
      "num_tokens": 30095704.0,
      "step": 14500
    },
    {
      "entropy": 5.555439853668213,
      "epoch": 1.3157656023222062,
      "grad_norm": 3.21875,
      "learning_rate": 0.0004832418144857982,
      "loss": 5.3048,
      "mean_token_accuracy": 0.17525641918182372,
      "num_tokens": 30105981.0,
      "step": 14505
    },
    {
      "entropy": 5.537001419067383,
      "epoch": 1.3162191582002902,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048322955893224245,
      "loss": 5.3488,
      "mean_token_accuracy": 0.17036312073469162,
      "num_tokens": 30115772.0,
      "step": 14510
    },
    {
      "entropy": 5.579875469207764,
      "epoch": 1.3166727140783745,
      "grad_norm": 2.984375,
      "learning_rate": 0.00048321729907244626,
      "loss": 5.3018,
      "mean_token_accuracy": 0.1734536573290825,
      "num_tokens": 30127050.0,
      "step": 14515
    },
    {
      "entropy": 5.598989963531494,
      "epoch": 1.3171262699564585,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004832050349066634,
      "loss": 5.3443,
      "mean_token_accuracy": 0.17404793351888656,
      "num_tokens": 30138221.0,
      "step": 14520
    },
    {
      "entropy": 5.625929832458496,
      "epoch": 1.3175798258345428,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004831927664351473,
      "loss": 5.4014,
      "mean_token_accuracy": 0.1705172449350357,
      "num_tokens": 30150181.0,
      "step": 14525
    },
    {
      "entropy": 5.58912501335144,
      "epoch": 1.318033381712627,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004831804936581518,
      "loss": 5.3828,
      "mean_token_accuracy": 0.17455049604177475,
      "num_tokens": 30160521.0,
      "step": 14530
    },
    {
      "entropy": 5.627479076385498,
      "epoch": 1.3184869375907111,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004831682165759305,
      "loss": 5.3468,
      "mean_token_accuracy": 0.17037987262010573,
      "num_tokens": 30170611.0,
      "step": 14535
    },
    {
      "entropy": 5.647074174880982,
      "epoch": 1.3189404934687954,
      "grad_norm": 2.625,
      "learning_rate": 0.00048315593518873757,
      "loss": 5.3137,
      "mean_token_accuracy": 0.17692942619323732,
      "num_tokens": 30179914.0,
      "step": 14540
    },
    {
      "entropy": 5.575990104675293,
      "epoch": 1.3193940493468794,
      "grad_norm": 4.15625,
      "learning_rate": 0.0004831436494968268,
      "loss": 5.3291,
      "mean_token_accuracy": 0.17755057066679,
      "num_tokens": 30191426.0,
      "step": 14545
    },
    {
      "entropy": 5.53160662651062,
      "epoch": 1.3198476052249637,
      "grad_norm": 1.875,
      "learning_rate": 0.00048313135950045234,
      "loss": 5.2747,
      "mean_token_accuracy": 0.17838724404573442,
      "num_tokens": 30202813.0,
      "step": 14550
    },
    {
      "entropy": 5.514211893081665,
      "epoch": 1.320301161103048,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048311906519986824,
      "loss": 5.2181,
      "mean_token_accuracy": 0.18271743953227998,
      "num_tokens": 30213044.0,
      "step": 14555
    },
    {
      "entropy": 5.59300012588501,
      "epoch": 1.320754716981132,
      "grad_norm": 4.03125,
      "learning_rate": 0.000483106766595329,
      "loss": 5.3015,
      "mean_token_accuracy": 0.17218518853187562,
      "num_tokens": 30223419.0,
      "step": 14560
    },
    {
      "entropy": 5.600766897201538,
      "epoch": 1.3212082728592163,
      "grad_norm": 2.5,
      "learning_rate": 0.0004830944636870887,
      "loss": 5.3658,
      "mean_token_accuracy": 0.16912604719400406,
      "num_tokens": 30233203.0,
      "step": 14565
    },
    {
      "entropy": 5.626376008987426,
      "epoch": 1.3216618287373003,
      "grad_norm": 3.09375,
      "learning_rate": 0.00048308215647540183,
      "loss": 5.3529,
      "mean_token_accuracy": 0.16886424869298935,
      "num_tokens": 30244604.0,
      "step": 14570
    },
    {
      "entropy": 5.618494510650635,
      "epoch": 1.3221153846153846,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048306984496052297,
      "loss": 5.4112,
      "mean_token_accuracy": 0.17318980395793915,
      "num_tokens": 30255355.0,
      "step": 14575
    },
    {
      "entropy": 5.62944016456604,
      "epoch": 1.3225689404934688,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048305752914270665,
      "loss": 5.4275,
      "mean_token_accuracy": 0.1672905758023262,
      "num_tokens": 30265665.0,
      "step": 14580
    },
    {
      "entropy": 5.630397844314575,
      "epoch": 1.3230224963715531,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048304520902220767,
      "loss": 5.3813,
      "mean_token_accuracy": 0.1761367589235306,
      "num_tokens": 30274347.0,
      "step": 14585
    },
    {
      "entropy": 5.607367658615113,
      "epoch": 1.3234760522496372,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004830328845992808,
      "loss": 5.2409,
      "mean_token_accuracy": 0.18162842839956284,
      "num_tokens": 30284243.0,
      "step": 14590
    },
    {
      "entropy": 5.469470977783203,
      "epoch": 1.3239296081277212,
      "grad_norm": 2.546875,
      "learning_rate": 0.00048302055587418074,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1771307483315468,
      "num_tokens": 30294327.0,
      "step": 14595
    },
    {
      "entropy": 5.516725826263428,
      "epoch": 1.3243831640058055,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004830082228471626,
      "loss": 5.3568,
      "mean_token_accuracy": 0.17763551324605942,
      "num_tokens": 30304188.0,
      "step": 14600
    },
    {
      "entropy": 5.618366765975952,
      "epoch": 1.3248367198838897,
      "grad_norm": 4.71875,
      "learning_rate": 0.00048299588551848146,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1809214636683464,
      "num_tokens": 30314917.0,
      "step": 14605
    },
    {
      "entropy": 5.578002119064331,
      "epoch": 1.325290275761974,
      "grad_norm": 2.734375,
      "learning_rate": 0.00048298354388839237,
      "loss": 5.293,
      "mean_token_accuracy": 0.17485674619674682,
      "num_tokens": 30325221.0,
      "step": 14610
    },
    {
      "entropy": 5.6580933094024655,
      "epoch": 1.325743831640058,
      "grad_norm": 2.875,
      "learning_rate": 0.0004829711979571506,
      "loss": 5.4004,
      "mean_token_accuracy": 0.16982516497373581,
      "num_tokens": 30336196.0,
      "step": 14615
    },
    {
      "entropy": 5.53471188545227,
      "epoch": 1.3261973875181423,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004829588477250115,
      "loss": 5.1964,
      "mean_token_accuracy": 0.18727820217609406,
      "num_tokens": 30345546.0,
      "step": 14620
    },
    {
      "entropy": 5.517544174194336,
      "epoch": 1.3266509433962264,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004829464931922304,
      "loss": 5.3429,
      "mean_token_accuracy": 0.1726578325033188,
      "num_tokens": 30355672.0,
      "step": 14625
    },
    {
      "entropy": 5.6081907749176025,
      "epoch": 1.3271044992743106,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048293413435906285,
      "loss": 5.3194,
      "mean_token_accuracy": 0.17273131310939788,
      "num_tokens": 30366233.0,
      "step": 14630
    },
    {
      "entropy": 5.629884862899781,
      "epoch": 1.327558055152395,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004829217712257644,
      "loss": 5.3218,
      "mean_token_accuracy": 0.17876290082931517,
      "num_tokens": 30377366.0,
      "step": 14635
    },
    {
      "entropy": 5.5762341022491455,
      "epoch": 1.328011611030479,
      "grad_norm": 2.25,
      "learning_rate": 0.00048290940379259073,
      "loss": 5.3719,
      "mean_token_accuracy": 0.17302599251270295,
      "num_tokens": 30388022.0,
      "step": 14640
    },
    {
      "entropy": 5.579840612411499,
      "epoch": 1.3284651669085632,
      "grad_norm": 7.53125,
      "learning_rate": 0.00048289703205979756,
      "loss": 5.259,
      "mean_token_accuracy": 0.17411655634641648,
      "num_tokens": 30399235.0,
      "step": 14645
    },
    {
      "entropy": 5.672154331207276,
      "epoch": 1.3289187227866472,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048288465602764084,
      "loss": 5.3999,
      "mean_token_accuracy": 0.16967085152864456,
      "num_tokens": 30409112.0,
      "step": 14650
    },
    {
      "entropy": 5.557979011535645,
      "epoch": 1.3293722786647315,
      "grad_norm": 2.53125,
      "learning_rate": 0.00048287227569637645,
      "loss": 5.3128,
      "mean_token_accuracy": 0.1765489861369133,
      "num_tokens": 30420202.0,
      "step": 14655
    },
    {
      "entropy": 5.613364887237549,
      "epoch": 1.3298258345428158,
      "grad_norm": 2.59375,
      "learning_rate": 0.00048285989106626036,
      "loss": 5.3527,
      "mean_token_accuracy": 0.17810803204774855,
      "num_tokens": 30431210.0,
      "step": 14660
    },
    {
      "entropy": 5.600608921051025,
      "epoch": 1.3302793904208998,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004828475021375489,
      "loss": 5.2802,
      "mean_token_accuracy": 0.17770425826311112,
      "num_tokens": 30441291.0,
      "step": 14665
    },
    {
      "entropy": 5.547167205810547,
      "epoch": 1.330732946298984,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004828351089104981,
      "loss": 5.2168,
      "mean_token_accuracy": 0.18351980298757553,
      "num_tokens": 30450285.0,
      "step": 14670
    },
    {
      "entropy": 5.5321026802062985,
      "epoch": 1.3311865021770681,
      "grad_norm": 3.15625,
      "learning_rate": 0.00048282271138536424,
      "loss": 5.3239,
      "mean_token_accuracy": 0.1722715437412262,
      "num_tokens": 30460311.0,
      "step": 14675
    },
    {
      "entropy": 5.680997323989868,
      "epoch": 1.3316400580551524,
      "grad_norm": 4.8125,
      "learning_rate": 0.0004828103095624037,
      "loss": 5.3652,
      "mean_token_accuracy": 0.16947381645441056,
      "num_tokens": 30470188.0,
      "step": 14680
    },
    {
      "entropy": 5.572077417373658,
      "epoch": 1.3320936139332367,
      "grad_norm": 2.53125,
      "learning_rate": 0.000482797903441873,
      "loss": 5.2888,
      "mean_token_accuracy": 0.17391078025102616,
      "num_tokens": 30481406.0,
      "step": 14685
    },
    {
      "entropy": 5.479278564453125,
      "epoch": 1.3325471698113207,
      "grad_norm": 2.75,
      "learning_rate": 0.0004827854930240288,
      "loss": 5.297,
      "mean_token_accuracy": 0.17948488891124725,
      "num_tokens": 30492314.0,
      "step": 14690
    },
    {
      "entropy": 5.58660249710083,
      "epoch": 1.333000725689405,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004827730783091276,
      "loss": 5.43,
      "mean_token_accuracy": 0.16671850234270097,
      "num_tokens": 30503273.0,
      "step": 14695
    },
    {
      "entropy": 5.646798706054687,
      "epoch": 1.333454281567489,
      "grad_norm": 2.6875,
      "learning_rate": 0.00048276065929742616,
      "loss": 5.3089,
      "mean_token_accuracy": 0.17752225548028946,
      "num_tokens": 30513582.0,
      "step": 14700
    },
    {
      "entropy": 5.576589775085449,
      "epoch": 1.3339078374455733,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004827482359891814,
      "loss": 5.2514,
      "mean_token_accuracy": 0.17876120358705522,
      "num_tokens": 30523261.0,
      "step": 14705
    },
    {
      "entropy": 5.557085752487183,
      "epoch": 1.3343613933236576,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00048273580838465016,
      "loss": 5.3653,
      "mean_token_accuracy": 0.17147661447525026,
      "num_tokens": 30533774.0,
      "step": 14710
    },
    {
      "entropy": 5.613748359680176,
      "epoch": 1.3348149492017416,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004827233764840894,
      "loss": 5.3878,
      "mean_token_accuracy": 0.17482657581567765,
      "num_tokens": 30544580.0,
      "step": 14715
    },
    {
      "entropy": 5.625464963912964,
      "epoch": 1.3352685050798259,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004827109402877564,
      "loss": 5.271,
      "mean_token_accuracy": 0.1773221969604492,
      "num_tokens": 30555105.0,
      "step": 14720
    },
    {
      "entropy": 5.568336868286133,
      "epoch": 1.33572206095791,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004826984997959082,
      "loss": 5.3779,
      "mean_token_accuracy": 0.16772764623165132,
      "num_tokens": 30564912.0,
      "step": 14725
    },
    {
      "entropy": 5.546491813659668,
      "epoch": 1.3361756168359942,
      "grad_norm": 4.3125,
      "learning_rate": 0.000482686055008802,
      "loss": 5.2995,
      "mean_token_accuracy": 0.1742671862244606,
      "num_tokens": 30575862.0,
      "step": 14730
    },
    {
      "entropy": 5.564079713821411,
      "epoch": 1.3366291727140784,
      "grad_norm": 2.34375,
      "learning_rate": 0.00048267360592669535,
      "loss": 5.4293,
      "mean_token_accuracy": 0.1716134801506996,
      "num_tokens": 30586951.0,
      "step": 14735
    },
    {
      "entropy": 5.5788383960723875,
      "epoch": 1.3370827285921625,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004826611525498456,
      "loss": 5.3205,
      "mean_token_accuracy": 0.17495224326848985,
      "num_tokens": 30597604.0,
      "step": 14740
    },
    {
      "entropy": 5.642521047592163,
      "epoch": 1.3375362844702468,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004826486948785103,
      "loss": 5.3907,
      "mean_token_accuracy": 0.17303643524646758,
      "num_tokens": 30608157.0,
      "step": 14745
    },
    {
      "entropy": 5.613686227798462,
      "epoch": 1.3379898403483308,
      "grad_norm": 2.421875,
      "learning_rate": 0.00048263623291294706,
      "loss": 5.3343,
      "mean_token_accuracy": 0.17162981629371643,
      "num_tokens": 30618527.0,
      "step": 14750
    },
    {
      "entropy": 5.613192796707153,
      "epoch": 1.338443396226415,
      "grad_norm": 3.203125,
      "learning_rate": 0.00048262376665341355,
      "loss": 5.4055,
      "mean_token_accuracy": 0.16945113241672516,
      "num_tokens": 30629253.0,
      "step": 14755
    },
    {
      "entropy": 5.595995950698852,
      "epoch": 1.3388969521044993,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004826112961001677,
      "loss": 5.3819,
      "mean_token_accuracy": 0.1626031368970871,
      "num_tokens": 30639015.0,
      "step": 14760
    },
    {
      "entropy": 5.691329002380371,
      "epoch": 1.3393505079825834,
      "grad_norm": 2.28125,
      "learning_rate": 0.00048259882125346735,
      "loss": 5.461,
      "mean_token_accuracy": 0.1610155239701271,
      "num_tokens": 30650035.0,
      "step": 14765
    },
    {
      "entropy": 5.590901136398315,
      "epoch": 1.3398040638606676,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048258634211357045,
      "loss": 5.3593,
      "mean_token_accuracy": 0.17194414436817168,
      "num_tokens": 30659813.0,
      "step": 14770
    },
    {
      "entropy": 5.570027923583984,
      "epoch": 1.3402576197387517,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004825738586807351,
      "loss": 5.3172,
      "mean_token_accuracy": 0.17335225641727448,
      "num_tokens": 30670243.0,
      "step": 14775
    },
    {
      "entropy": 5.546255826950073,
      "epoch": 1.340711175616836,
      "grad_norm": 2.703125,
      "learning_rate": 0.00048256137095521945,
      "loss": 5.2837,
      "mean_token_accuracy": 0.1731187880039215,
      "num_tokens": 30680056.0,
      "step": 14780
    },
    {
      "entropy": 5.580109119415283,
      "epoch": 1.3411647314949202,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048254887893728175,
      "loss": 5.2441,
      "mean_token_accuracy": 0.1779971808195114,
      "num_tokens": 30690536.0,
      "step": 14785
    },
    {
      "entropy": 5.542200756072998,
      "epoch": 1.3416182873730045,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004825363826271804,
      "loss": 5.3193,
      "mean_token_accuracy": 0.1733024775981903,
      "num_tokens": 30700200.0,
      "step": 14790
    },
    {
      "entropy": 5.511820363998413,
      "epoch": 1.3420718432510885,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004825238820251737,
      "loss": 5.2197,
      "mean_token_accuracy": 0.17725178897380828,
      "num_tokens": 30710576.0,
      "step": 14795
    },
    {
      "entropy": 5.565164995193482,
      "epoch": 1.3425253991291728,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048251137713152027,
      "loss": 5.3504,
      "mean_token_accuracy": 0.16702319532632828,
      "num_tokens": 30721495.0,
      "step": 14800
    },
    {
      "entropy": 5.569557046890258,
      "epoch": 1.3429789550072568,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004824988679464787,
      "loss": 5.3651,
      "mean_token_accuracy": 0.1734679028391838,
      "num_tokens": 30731700.0,
      "step": 14805
    },
    {
      "entropy": 5.629995059967041,
      "epoch": 1.343432510885341,
      "grad_norm": 2.59375,
      "learning_rate": 0.00048248635447030767,
      "loss": 5.3526,
      "mean_token_accuracy": 0.16986653208732605,
      "num_tokens": 30743421.0,
      "step": 14810
    },
    {
      "entropy": 5.567506456375122,
      "epoch": 1.3438860667634254,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004824738367032659,
      "loss": 5.2583,
      "mean_token_accuracy": 0.1790758728981018,
      "num_tokens": 30754238.0,
      "step": 14815
    },
    {
      "entropy": 5.579785871505737,
      "epoch": 1.3443396226415094,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004824613146456123,
      "loss": 5.3647,
      "mean_token_accuracy": 0.17224185913801193,
      "num_tokens": 30763938.0,
      "step": 14820
    },
    {
      "entropy": 5.6040191650390625,
      "epoch": 1.3447931785195937,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004824487882976059,
      "loss": 5.4162,
      "mean_token_accuracy": 0.1700545772910118,
      "num_tokens": 30773654.0,
      "step": 14825
    },
    {
      "entropy": 5.66545557975769,
      "epoch": 1.3452467343976777,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048243625765950564,
      "loss": 5.3793,
      "mean_token_accuracy": 0.16727219223976136,
      "num_tokens": 30784659.0,
      "step": 14830
    },
    {
      "entropy": 5.543287754058838,
      "epoch": 1.345700290275762,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004824237227315708,
      "loss": 5.2645,
      "mean_token_accuracy": 0.16864370703697204,
      "num_tokens": 30795194.0,
      "step": 14835
    },
    {
      "entropy": 5.476270246505737,
      "epoch": 1.3461538461538463,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004824111835140604,
      "loss": 5.1946,
      "mean_token_accuracy": 0.18628957420587539,
      "num_tokens": 30805623.0,
      "step": 14840
    },
    {
      "entropy": 5.5201915264129635,
      "epoch": 1.3466074020319303,
      "grad_norm": 2.09375,
      "learning_rate": 0.000482398640007234,
      "loss": 5.236,
      "mean_token_accuracy": 0.18110164999961853,
      "num_tokens": 30814899.0,
      "step": 14845
    },
    {
      "entropy": 5.555194616317749,
      "epoch": 1.3470609579100146,
      "grad_norm": 2.125,
      "learning_rate": 0.00048238609221135076,
      "loss": 5.3811,
      "mean_token_accuracy": 0.16786562353372575,
      "num_tokens": 30825546.0,
      "step": 14850
    },
    {
      "entropy": 5.499044179916382,
      "epoch": 1.3475145137880986,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004823735401266704,
      "loss": 5.2674,
      "mean_token_accuracy": 0.17715806365013123,
      "num_tokens": 30836406.0,
      "step": 14855
    },
    {
      "entropy": 5.559240770339966,
      "epoch": 1.3479680696661829,
      "grad_norm": 3.25,
      "learning_rate": 0.00048236098375345227,
      "loss": 5.3598,
      "mean_token_accuracy": 0.17311821430921553,
      "num_tokens": 30847966.0,
      "step": 14860
    },
    {
      "entropy": 5.66569414138794,
      "epoch": 1.3484216255442671,
      "grad_norm": 2.625,
      "learning_rate": 0.0004823484230919563,
      "loss": 5.3667,
      "mean_token_accuracy": 0.17487807273864747,
      "num_tokens": 30856831.0,
      "step": 14865
    },
    {
      "entropy": 5.487578392028809,
      "epoch": 1.3488751814223512,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048233585814244195,
      "loss": 5.177,
      "mean_token_accuracy": 0.1836571589112282,
      "num_tokens": 30866076.0,
      "step": 14870
    },
    {
      "entropy": 5.502755832672119,
      "epoch": 1.3493287373004355,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004823232889051693,
      "loss": 5.266,
      "mean_token_accuracy": 0.18168337643146515,
      "num_tokens": 30876845.0,
      "step": 14875
    },
    {
      "entropy": 5.527269411087036,
      "epoch": 1.3497822931785195,
      "grad_norm": 2.125,
      "learning_rate": 0.00048231071538039816,
      "loss": 5.3345,
      "mean_token_accuracy": 0.17084167152643204,
      "num_tokens": 30887448.0,
      "step": 14880
    },
    {
      "entropy": 5.607729387283325,
      "epoch": 1.3502358490566038,
      "grad_norm": 2.484375,
      "learning_rate": 0.00048229813756838863,
      "loss": 5.3009,
      "mean_token_accuracy": 0.1751062110066414,
      "num_tokens": 30898414.0,
      "step": 14885
    },
    {
      "entropy": 5.596131372451782,
      "epoch": 1.350689404934688,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004822855554694007,
      "loss": 5.3269,
      "mean_token_accuracy": 0.17376189231872557,
      "num_tokens": 30908559.0,
      "step": 14890
    },
    {
      "entropy": 5.57922043800354,
      "epoch": 1.351142960812772,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048227296908369475,
      "loss": 5.3113,
      "mean_token_accuracy": 0.17814643383026124,
      "num_tokens": 30918655.0,
      "step": 14895
    },
    {
      "entropy": 5.540636777877808,
      "epoch": 1.3515965166908563,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00048226037841153104,
      "loss": 5.3275,
      "mean_token_accuracy": 0.17941174656152725,
      "num_tokens": 30929188.0,
      "step": 14900
    },
    {
      "entropy": 5.617151594161987,
      "epoch": 1.3520500725689404,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004822477834531698,
      "loss": 5.3898,
      "mean_token_accuracy": 0.17542280852794648,
      "num_tokens": 30939902.0,
      "step": 14905
    },
    {
      "entropy": 5.570489454269409,
      "epoch": 1.3525036284470247,
      "grad_norm": 3.390625,
      "learning_rate": 0.0004822351842088716,
      "loss": 5.3336,
      "mean_token_accuracy": 0.17203953862190247,
      "num_tokens": 30951289.0,
      "step": 14910
    },
    {
      "entropy": 5.6825377464294435,
      "epoch": 1.352957184325109,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004822225806788969,
      "loss": 5.4447,
      "mean_token_accuracy": 0.16396759450435638,
      "num_tokens": 30961233.0,
      "step": 14915
    },
    {
      "entropy": 5.647076797485352,
      "epoch": 1.353410740203193,
      "grad_norm": 3.625,
      "learning_rate": 0.00048220997286350654,
      "loss": 5.268,
      "mean_token_accuracy": 0.18247477412223817,
      "num_tokens": 30971922.0,
      "step": 14920
    },
    {
      "entropy": 5.567844486236572,
      "epoch": 1.3538642960812772,
      "grad_norm": 2.390625,
      "learning_rate": 0.00048219736076296114,
      "loss": 5.2989,
      "mean_token_accuracy": 0.17440518289804458,
      "num_tokens": 30981358.0,
      "step": 14925
    },
    {
      "entropy": 5.580735588073731,
      "epoch": 1.3543178519593613,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004821847443775214,
      "loss": 5.4159,
      "mean_token_accuracy": 0.1733836606144905,
      "num_tokens": 30991984.0,
      "step": 14930
    },
    {
      "entropy": 5.5941376209259035,
      "epoch": 1.3547714078374455,
      "grad_norm": 2.34375,
      "learning_rate": 0.00048217212370744853,
      "loss": 5.3063,
      "mean_token_accuracy": 0.17704341262578965,
      "num_tokens": 31002790.0,
      "step": 14935
    },
    {
      "entropy": 5.499529409408569,
      "epoch": 1.3552249637155298,
      "grad_norm": 2.578125,
      "learning_rate": 0.00048215949875300327,
      "loss": 5.1621,
      "mean_token_accuracy": 0.18547530621290206,
      "num_tokens": 31012827.0,
      "step": 14940
    },
    {
      "entropy": 5.4738743782043455,
      "epoch": 1.3556785195936139,
      "grad_norm": 3.65625,
      "learning_rate": 0.0004821468695144467,
      "loss": 5.2781,
      "mean_token_accuracy": 0.17176304012537003,
      "num_tokens": 31023075.0,
      "step": 14945
    },
    {
      "entropy": 5.5304771900177006,
      "epoch": 1.3561320754716981,
      "grad_norm": 2.78125,
      "learning_rate": 0.00048213423599204015,
      "loss": 5.3205,
      "mean_token_accuracy": 0.18099177926778792,
      "num_tokens": 31032454.0,
      "step": 14950
    },
    {
      "entropy": 5.6834893226623535,
      "epoch": 1.3565856313497822,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048212159818604477,
      "loss": 5.4345,
      "mean_token_accuracy": 0.16614234894514085,
      "num_tokens": 31042498.0,
      "step": 14955
    },
    {
      "entropy": 5.565841150283814,
      "epoch": 1.3570391872278664,
      "grad_norm": 2.671875,
      "learning_rate": 0.000482108956096722,
      "loss": 5.3214,
      "mean_token_accuracy": 0.1715633600950241,
      "num_tokens": 31053188.0,
      "step": 14960
    },
    {
      "entropy": 5.568301200866699,
      "epoch": 1.3574927431059507,
      "grad_norm": 3.265625,
      "learning_rate": 0.00048209630972433326,
      "loss": 5.2872,
      "mean_token_accuracy": 0.18262624591588975,
      "num_tokens": 31062840.0,
      "step": 14965
    },
    {
      "entropy": 5.6303094863891605,
      "epoch": 1.357946298984035,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048208365906914,
      "loss": 5.4185,
      "mean_token_accuracy": 0.17723245173692703,
      "num_tokens": 31072646.0,
      "step": 14970
    },
    {
      "entropy": 5.539926385879516,
      "epoch": 1.358399854862119,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048207100413140397,
      "loss": 5.2543,
      "mean_token_accuracy": 0.17492770701646804,
      "num_tokens": 31082700.0,
      "step": 14975
    },
    {
      "entropy": 5.608774471282959,
      "epoch": 1.3588534107402033,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004820583449113868,
      "loss": 5.3587,
      "mean_token_accuracy": 0.16961142644286156,
      "num_tokens": 31093743.0,
      "step": 14980
    },
    {
      "entropy": 5.567367315292358,
      "epoch": 1.3593069666182873,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048204568140935023,
      "loss": 5.2959,
      "mean_token_accuracy": 0.1769140988588333,
      "num_tokens": 31104293.0,
      "step": 14985
    },
    {
      "entropy": 5.527382898330688,
      "epoch": 1.3597605224963716,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004820330136255563,
      "loss": 5.3082,
      "mean_token_accuracy": 0.17653436809778214,
      "num_tokens": 31114244.0,
      "step": 14990
    },
    {
      "entropy": 5.552898788452149,
      "epoch": 1.3602140783744558,
      "grad_norm": 2.65625,
      "learning_rate": 0.00048202034156026683,
      "loss": 5.2771,
      "mean_token_accuracy": 0.17202102094888688,
      "num_tokens": 31123596.0,
      "step": 14995
    },
    {
      "entropy": 5.605963468551636,
      "epoch": 1.36066763425254,
      "grad_norm": 3.296875,
      "learning_rate": 0.00048200766521374396,
      "loss": 5.3559,
      "mean_token_accuracy": 0.16639875918626784,
      "num_tokens": 31134258.0,
      "step": 15000
    },
    {
      "epoch": 1.36066763425254,
      "eval_entropy": 5.434485418872057,
      "eval_loss": 5.401484966278076,
      "eval_mean_token_accuracy": 0.1792781375904833,
      "eval_num_tokens": 31134258.0,
      "eval_runtime": 26.4259,
      "eval_samples_per_second": 1338.082,
      "eval_steps_per_second": 167.26,
      "step": 15000
    },
    {
      "entropy": 5.5745984554290775,
      "epoch": 1.3611211901306242,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004819949845862498,
      "loss": 5.3457,
      "mean_token_accuracy": 0.17296434789896012,
      "num_tokens": 31144806.0,
      "step": 15005
    },
    {
      "entropy": 5.673216056823731,
      "epoch": 1.3615747460087082,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004819822996780466,
      "loss": 5.4258,
      "mean_token_accuracy": 0.1662140116095543,
      "num_tokens": 31155642.0,
      "step": 15010
    },
    {
      "entropy": 5.636314439773559,
      "epoch": 1.3620283018867925,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004819696104893967,
      "loss": 5.2508,
      "mean_token_accuracy": 0.17902350872755052,
      "num_tokens": 31165034.0,
      "step": 15015
    },
    {
      "entropy": 5.550515604019165,
      "epoch": 1.3624818577648767,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004819569170205625,
      "loss": 5.3465,
      "mean_token_accuracy": 0.17229864448308946,
      "num_tokens": 31176491.0,
      "step": 15020
    },
    {
      "entropy": 5.592280721664428,
      "epoch": 1.3629354136429608,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00048194421927180654,
      "loss": 5.3655,
      "mean_token_accuracy": 0.16831783801317216,
      "num_tokens": 31186198.0,
      "step": 15025
    },
    {
      "entropy": 5.650245571136475,
      "epoch": 1.363388969521045,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004819315172433914,
      "loss": 5.3276,
      "mean_token_accuracy": 0.17557564228773118,
      "num_tokens": 31197094.0,
      "step": 15030
    },
    {
      "entropy": 5.568355751037598,
      "epoch": 1.363842525399129,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048191881093557977,
      "loss": 5.3338,
      "mean_token_accuracy": 0.17247996628284454,
      "num_tokens": 31207505.0,
      "step": 15035
    },
    {
      "entropy": 5.654825258255005,
      "epoch": 1.3642960812772134,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048190610034863434,
      "loss": 5.4671,
      "mean_token_accuracy": 0.16520622223615647,
      "num_tokens": 31218809.0,
      "step": 15040
    },
    {
      "entropy": 5.662196636199951,
      "epoch": 1.3647496371552976,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004818933854828181,
      "loss": 5.3782,
      "mean_token_accuracy": 0.16452924609184266,
      "num_tokens": 31230281.0,
      "step": 15045
    },
    {
      "entropy": 5.559078693389893,
      "epoch": 1.3652031930333817,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004818806663383939,
      "loss": 5.2601,
      "mean_token_accuracy": 0.18405363112688064,
      "num_tokens": 31239573.0,
      "step": 15050
    },
    {
      "entropy": 5.536240577697754,
      "epoch": 1.365656748911466,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004818679429156248,
      "loss": 5.2976,
      "mean_token_accuracy": 0.17511320114135742,
      "num_tokens": 31249671.0,
      "step": 15055
    },
    {
      "entropy": 5.551848411560059,
      "epoch": 1.36611030478955,
      "grad_norm": 3.15625,
      "learning_rate": 0.00048185521521477385,
      "loss": 5.3252,
      "mean_token_accuracy": 0.17376567125320436,
      "num_tokens": 31259601.0,
      "step": 15060
    },
    {
      "entropy": 5.617363595962525,
      "epoch": 1.3665638606676342,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004818424832361044,
      "loss": 5.3063,
      "mean_token_accuracy": 0.17559281289577483,
      "num_tokens": 31269085.0,
      "step": 15065
    },
    {
      "entropy": 5.51986255645752,
      "epoch": 1.3670174165457185,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004818297469798797,
      "loss": 5.3441,
      "mean_token_accuracy": 0.16626193523406982,
      "num_tokens": 31280177.0,
      "step": 15070
    },
    {
      "entropy": 5.529023170471191,
      "epoch": 1.3674709724238026,
      "grad_norm": 2.421875,
      "learning_rate": 0.00048181700644636314,
      "loss": 5.2457,
      "mean_token_accuracy": 0.1768273264169693,
      "num_tokens": 31290545.0,
      "step": 15075
    },
    {
      "entropy": 5.524762964248657,
      "epoch": 1.3679245283018868,
      "grad_norm": 2.71875,
      "learning_rate": 0.00048180426163581816,
      "loss": 5.2029,
      "mean_token_accuracy": 0.17964352667331696,
      "num_tokens": 31301143.0,
      "step": 15080
    },
    {
      "entropy": 5.523261833190918,
      "epoch": 1.3683780841799709,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004817915125485083,
      "loss": 5.2809,
      "mean_token_accuracy": 0.1761244624853134,
      "num_tokens": 31311071.0,
      "step": 15085
    },
    {
      "entropy": 5.573470497131348,
      "epoch": 1.3688316400580551,
      "grad_norm": 2.53125,
      "learning_rate": 0.00048177875918469735,
      "loss": 5.3924,
      "mean_token_accuracy": 0.17012052834033967,
      "num_tokens": 31322263.0,
      "step": 15090
    },
    {
      "entropy": 5.599078416824341,
      "epoch": 1.3692851959361394,
      "grad_norm": 2.28125,
      "learning_rate": 0.00048176600154464897,
      "loss": 5.2938,
      "mean_token_accuracy": 0.17614298611879348,
      "num_tokens": 31332975.0,
      "step": 15095
    },
    {
      "entropy": 5.710826683044433,
      "epoch": 1.3697387518142234,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004817532396286269,
      "loss": 5.4319,
      "mean_token_accuracy": 0.16938230842351915,
      "num_tokens": 31345252.0,
      "step": 15100
    },
    {
      "entropy": 5.595699644088745,
      "epoch": 1.3701923076923077,
      "grad_norm": 1.75,
      "learning_rate": 0.0004817404734368952,
      "loss": 5.2707,
      "mean_token_accuracy": 0.177989599108696,
      "num_tokens": 31355479.0,
      "step": 15105
    },
    {
      "entropy": 5.629980230331421,
      "epoch": 1.3706458635703918,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048172770296971785,
      "loss": 5.3924,
      "mean_token_accuracy": 0.1703348308801651,
      "num_tokens": 31366020.0,
      "step": 15110
    },
    {
      "entropy": 5.597391510009766,
      "epoch": 1.371099419448476,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004817149282273589,
      "loss": 5.3346,
      "mean_token_accuracy": 0.17335227727890015,
      "num_tokens": 31376055.0,
      "step": 15115
    },
    {
      "entropy": 5.60340986251831,
      "epoch": 1.3715529753265603,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004817021492100825,
      "loss": 5.4352,
      "mean_token_accuracy": 0.16931031346321107,
      "num_tokens": 31386179.0,
      "step": 15120
    },
    {
      "entropy": 5.526395082473755,
      "epoch": 1.3720065312046443,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000481689365918153,
      "loss": 5.2594,
      "mean_token_accuracy": 0.17771146893501283,
      "num_tokens": 31395850.0,
      "step": 15125
    },
    {
      "entropy": 5.564254808425903,
      "epoch": 1.3724600870827286,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004816765783518348,
      "loss": 5.2124,
      "mean_token_accuracy": 0.18355452567338942,
      "num_tokens": 31405053.0,
      "step": 15130
    },
    {
      "entropy": 5.543607997894287,
      "epoch": 1.3729136429608126,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004816637865113923,
      "loss": 5.2378,
      "mean_token_accuracy": 0.1825024738907814,
      "num_tokens": 31415153.0,
      "step": 15135
    },
    {
      "entropy": 5.5463470935821535,
      "epoch": 1.373367198838897,
      "grad_norm": 2.25,
      "learning_rate": 0.00048165099039708994,
      "loss": 5.3421,
      "mean_token_accuracy": 0.17285471111536027,
      "num_tokens": 31425451.0,
      "step": 15140
    },
    {
      "entropy": 5.545378684997559,
      "epoch": 1.3738207547169812,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004816381900091925,
      "loss": 5.3119,
      "mean_token_accuracy": 0.17135562151670455,
      "num_tokens": 31436298.0,
      "step": 15145
    },
    {
      "entropy": 5.59074182510376,
      "epoch": 1.3742743105950654,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048162538534796465,
      "loss": 5.2986,
      "mean_token_accuracy": 0.1795909121632576,
      "num_tokens": 31446932.0,
      "step": 15150
    },
    {
      "entropy": 5.624588823318481,
      "epoch": 1.3747278664731495,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004816125764136711,
      "loss": 5.368,
      "mean_token_accuracy": 0.16332074552774428,
      "num_tokens": 31457582.0,
      "step": 15155
    },
    {
      "entropy": 5.70192289352417,
      "epoch": 1.3751814223512338,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004815997632065769,
      "loss": 5.3375,
      "mean_token_accuracy": 0.1737053543329239,
      "num_tokens": 31468168.0,
      "step": 15160
    },
    {
      "entropy": 5.614826250076294,
      "epoch": 1.3756349782293178,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048158694572694694,
      "loss": 5.33,
      "mean_token_accuracy": 0.1763288915157318,
      "num_tokens": 31478524.0,
      "step": 15165
    },
    {
      "entropy": 5.563558292388916,
      "epoch": 1.376088534107402,
      "grad_norm": 2.0,
      "learning_rate": 0.00048157412397504623,
      "loss": 5.3015,
      "mean_token_accuracy": 0.17530810981988906,
      "num_tokens": 31488608.0,
      "step": 15170
    },
    {
      "entropy": 5.573767614364624,
      "epoch": 1.3765420899854863,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004815612979511401,
      "loss": 5.3195,
      "mean_token_accuracy": 0.17861174196004867,
      "num_tokens": 31498984.0,
      "step": 15175
    },
    {
      "entropy": 5.645355367660523,
      "epoch": 1.3769956458635704,
      "grad_norm": 2.34375,
      "learning_rate": 0.00048154846765549353,
      "loss": 5.3348,
      "mean_token_accuracy": 0.17623123973608018,
      "num_tokens": 31509110.0,
      "step": 15180
    },
    {
      "entropy": 5.556439161300659,
      "epoch": 1.3774492017416546,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048153563308837215,
      "loss": 5.331,
      "mean_token_accuracy": 0.1748549982905388,
      "num_tokens": 31519266.0,
      "step": 15185
    },
    {
      "entropy": 5.5317702293396,
      "epoch": 1.3779027576197387,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048152279425004124,
      "loss": 5.3347,
      "mean_token_accuracy": 0.17385509461164475,
      "num_tokens": 31529043.0,
      "step": 15190
    },
    {
      "entropy": 5.626209259033203,
      "epoch": 1.378356313497823,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004815099511407663,
      "loss": 5.3307,
      "mean_token_accuracy": 0.1783709317445755,
      "num_tokens": 31538650.0,
      "step": 15195
    },
    {
      "entropy": 5.559007930755615,
      "epoch": 1.3788098693759072,
      "grad_norm": 2.609375,
      "learning_rate": 0.000481497103760813,
      "loss": 5.2555,
      "mean_token_accuracy": 0.17701199203729628,
      "num_tokens": 31549392.0,
      "step": 15200
    },
    {
      "entropy": 5.649675130844116,
      "epoch": 1.3792634252539913,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048148425211044684,
      "loss": 5.4402,
      "mean_token_accuracy": 0.1625445678830147,
      "num_tokens": 31559999.0,
      "step": 15205
    },
    {
      "entropy": 5.675137090682983,
      "epoch": 1.3797169811320755,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004814713961899338,
      "loss": 5.3382,
      "mean_token_accuracy": 0.17124535590410234,
      "num_tokens": 31570108.0,
      "step": 15210
    },
    {
      "entropy": 5.577499771118164,
      "epoch": 1.3801705370101596,
      "grad_norm": 2.609375,
      "learning_rate": 0.00048145853599953966,
      "loss": 5.2925,
      "mean_token_accuracy": 0.17274475395679473,
      "num_tokens": 31579698.0,
      "step": 15215
    },
    {
      "entropy": 5.536963653564453,
      "epoch": 1.3806240928882438,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048144567153953036,
      "loss": 5.3544,
      "mean_token_accuracy": 0.168358650803566,
      "num_tokens": 31589292.0,
      "step": 15220
    },
    {
      "entropy": 5.642316579818726,
      "epoch": 1.381077648766328,
      "grad_norm": 2.53125,
      "learning_rate": 0.00048143280281017197,
      "loss": 5.4077,
      "mean_token_accuracy": 0.17338637858629227,
      "num_tokens": 31599848.0,
      "step": 15225
    },
    {
      "entropy": 5.610063982009888,
      "epoch": 1.3815312046444121,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004814199298117307,
      "loss": 5.3442,
      "mean_token_accuracy": 0.1768866792321205,
      "num_tokens": 31611109.0,
      "step": 15230
    },
    {
      "entropy": 5.5713691234588625,
      "epoch": 1.3819847605224964,
      "grad_norm": 2.0,
      "learning_rate": 0.00048140705254447256,
      "loss": 5.2647,
      "mean_token_accuracy": 0.1775397315621376,
      "num_tokens": 31621140.0,
      "step": 15235
    },
    {
      "entropy": 5.6416410446167,
      "epoch": 1.3824383164005805,
      "grad_norm": 2.0,
      "learning_rate": 0.00048139417100866405,
      "loss": 5.3621,
      "mean_token_accuracy": 0.17323873788118363,
      "num_tokens": 31631268.0,
      "step": 15240
    },
    {
      "entropy": 5.590442085266114,
      "epoch": 1.3828918722786647,
      "grad_norm": 2.5625,
      "learning_rate": 0.00048138128520457146,
      "loss": 5.3099,
      "mean_token_accuracy": 0.17548275887966155,
      "num_tokens": 31640758.0,
      "step": 15245
    },
    {
      "entropy": 5.523724031448364,
      "epoch": 1.383345428156749,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004813683951324613,
      "loss": 5.312,
      "mean_token_accuracy": 0.17180445194244384,
      "num_tokens": 31651060.0,
      "step": 15250
    },
    {
      "entropy": 5.610667324066162,
      "epoch": 1.383798984034833,
      "grad_norm": 2.015625,
      "learning_rate": 0.00048135550079260014,
      "loss": 5.4182,
      "mean_token_accuracy": 0.17218077480792998,
      "num_tokens": 31662282.0,
      "step": 15255
    },
    {
      "entropy": 5.599547386169434,
      "epoch": 1.3842525399129173,
      "grad_norm": 2.96875,
      "learning_rate": 0.0004813426021852547,
      "loss": 5.2902,
      "mean_token_accuracy": 0.1699610948562622,
      "num_tokens": 31672364.0,
      "step": 15260
    },
    {
      "entropy": 5.606910181045532,
      "epoch": 1.3847060957910013,
      "grad_norm": 2.0,
      "learning_rate": 0.00048132969931069156,
      "loss": 5.3522,
      "mean_token_accuracy": 0.17120744436979293,
      "num_tokens": 31683094.0,
      "step": 15265
    },
    {
      "entropy": 5.5920312881469725,
      "epoch": 1.3851596516690856,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004813167921691777,
      "loss": 5.3078,
      "mean_token_accuracy": 0.1756906419992447,
      "num_tokens": 31693087.0,
      "step": 15270
    },
    {
      "entropy": 5.621430730819702,
      "epoch": 1.3856132075471699,
      "grad_norm": 2.484375,
      "learning_rate": 0.00048130388076098003,
      "loss": 5.2988,
      "mean_token_accuracy": 0.1698193743824959,
      "num_tokens": 31703708.0,
      "step": 15275
    },
    {
      "entropy": 5.527341222763061,
      "epoch": 1.386066763425254,
      "grad_norm": 1.90625,
      "learning_rate": 0.00048129096508636547,
      "loss": 5.2413,
      "mean_token_accuracy": 0.18231841176748276,
      "num_tokens": 31713640.0,
      "step": 15280
    },
    {
      "entropy": 5.554745435714722,
      "epoch": 1.3865203193033382,
      "grad_norm": 1.78125,
      "learning_rate": 0.00048127804514560126,
      "loss": 5.3275,
      "mean_token_accuracy": 0.1750309646129608,
      "num_tokens": 31723526.0,
      "step": 15285
    },
    {
      "entropy": 5.556029653549194,
      "epoch": 1.3869738751814222,
      "grad_norm": 3.796875,
      "learning_rate": 0.00048126512093895447,
      "loss": 5.2687,
      "mean_token_accuracy": 0.17604386061429977,
      "num_tokens": 31733265.0,
      "step": 15290
    },
    {
      "entropy": 5.6346104621887205,
      "epoch": 1.3874274310595065,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004812521924666925,
      "loss": 5.2664,
      "mean_token_accuracy": 0.17434800118207933,
      "num_tokens": 31743832.0,
      "step": 15295
    },
    {
      "entropy": 5.6342267990112305,
      "epoch": 1.3878809869375908,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004812392597290825,
      "loss": 5.3295,
      "mean_token_accuracy": 0.18450314104557036,
      "num_tokens": 31754443.0,
      "step": 15300
    },
    {
      "entropy": 5.520844793319702,
      "epoch": 1.3883345428156748,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004812263227263921,
      "loss": 5.2943,
      "mean_token_accuracy": 0.17682955265045167,
      "num_tokens": 31765304.0,
      "step": 15305
    },
    {
      "entropy": 5.557877731323242,
      "epoch": 1.388788098693759,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004812133814588888,
      "loss": 5.3401,
      "mean_token_accuracy": 0.16958561837673186,
      "num_tokens": 31775809.0,
      "step": 15310
    },
    {
      "entropy": 5.647122001647949,
      "epoch": 1.3892416545718431,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004812004359268402,
      "loss": 5.3284,
      "mean_token_accuracy": 0.17818181365728378,
      "num_tokens": 31786271.0,
      "step": 15315
    },
    {
      "entropy": 5.608468866348266,
      "epoch": 1.3896952104499274,
      "grad_norm": 3.0625,
      "learning_rate": 0.0004811874861305141,
      "loss": 5.371,
      "mean_token_accuracy": 0.1740392342209816,
      "num_tokens": 31797358.0,
      "step": 15320
    },
    {
      "entropy": 5.553903198242187,
      "epoch": 1.3901487663280117,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004811745320701782,
      "loss": 5.2925,
      "mean_token_accuracy": 0.18476718217134475,
      "num_tokens": 31808303.0,
      "step": 15325
    },
    {
      "entropy": 5.538501644134522,
      "epoch": 1.390602322206096,
      "grad_norm": 2.125,
      "learning_rate": 0.0004811615737461004,
      "loss": 5.2993,
      "mean_token_accuracy": 0.17466944307088852,
      "num_tokens": 31818837.0,
      "step": 15330
    },
    {
      "entropy": 5.636811304092407,
      "epoch": 1.39105587808418,
      "grad_norm": 3.3125,
      "learning_rate": 0.00048114861115854877,
      "loss": 5.329,
      "mean_token_accuracy": 0.17781821489334107,
      "num_tokens": 31829729.0,
      "step": 15335
    },
    {
      "entropy": 5.565691328048706,
      "epoch": 1.3915094339622642,
      "grad_norm": 4.09375,
      "learning_rate": 0.00048113564430779136,
      "loss": 5.2527,
      "mean_token_accuracy": 0.17771869152784348,
      "num_tokens": 31839996.0,
      "step": 15340
    },
    {
      "entropy": 5.498915433883667,
      "epoch": 1.3919629898403483,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004811226731940962,
      "loss": 5.2297,
      "mean_token_accuracy": 0.1784809947013855,
      "num_tokens": 31850298.0,
      "step": 15345
    },
    {
      "entropy": 5.579777479171753,
      "epoch": 1.3924165457184325,
      "grad_norm": 2.015625,
      "learning_rate": 0.00048110969781773173,
      "loss": 5.3388,
      "mean_token_accuracy": 0.18225645273923874,
      "num_tokens": 31862170.0,
      "step": 15350
    },
    {
      "entropy": 5.622347116470337,
      "epoch": 1.3928701015965168,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00048109671817896614,
      "loss": 5.3822,
      "mean_token_accuracy": 0.17012085616588593,
      "num_tokens": 31872081.0,
      "step": 15355
    },
    {
      "entropy": 5.575254678726196,
      "epoch": 1.3933236574746009,
      "grad_norm": 6.78125,
      "learning_rate": 0.00048108373427806797,
      "loss": 5.3265,
      "mean_token_accuracy": 0.1732793316245079,
      "num_tokens": 31882715.0,
      "step": 15360
    },
    {
      "entropy": 5.5584324359893795,
      "epoch": 1.3937772133526851,
      "grad_norm": 3.6875,
      "learning_rate": 0.00048107074611530555,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1840381905436516,
      "num_tokens": 31891925.0,
      "step": 15365
    },
    {
      "entropy": 5.595956611633301,
      "epoch": 1.3942307692307692,
      "grad_norm": 2.875,
      "learning_rate": 0.00048105775369094756,
      "loss": 5.3252,
      "mean_token_accuracy": 0.1715623542666435,
      "num_tokens": 31902320.0,
      "step": 15370
    },
    {
      "entropy": 5.548269557952881,
      "epoch": 1.3946843251088534,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004810447570052628,
      "loss": 5.2137,
      "mean_token_accuracy": 0.1813837006688118,
      "num_tokens": 31912726.0,
      "step": 15375
    },
    {
      "entropy": 5.569794702529907,
      "epoch": 1.3951378809869377,
      "grad_norm": 2.296875,
      "learning_rate": 0.00048103175605851997,
      "loss": 5.3303,
      "mean_token_accuracy": 0.17515553086996077,
      "num_tokens": 31924039.0,
      "step": 15380
    },
    {
      "entropy": 5.472156763076782,
      "epoch": 1.3955914368650217,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004810187508509879,
      "loss": 5.2058,
      "mean_token_accuracy": 0.18504221886396408,
      "num_tokens": 31933280.0,
      "step": 15385
    },
    {
      "entropy": 5.535130214691162,
      "epoch": 1.396044992743106,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004810057413829355,
      "loss": 5.4285,
      "mean_token_accuracy": 0.17255861461162567,
      "num_tokens": 31943182.0,
      "step": 15390
    },
    {
      "entropy": 5.693903160095215,
      "epoch": 1.39649854862119,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048099272765463193,
      "loss": 5.3381,
      "mean_token_accuracy": 0.17597001641988755,
      "num_tokens": 31953256.0,
      "step": 15395
    },
    {
      "entropy": 5.708860874176025,
      "epoch": 1.3969521044992743,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004809797096663463,
      "loss": 5.4561,
      "mean_token_accuracy": 0.16812326312065123,
      "num_tokens": 31964231.0,
      "step": 15400
    },
    {
      "entropy": 5.588603925704956,
      "epoch": 1.3974056603773586,
      "grad_norm": 2.921875,
      "learning_rate": 0.0004809666874183477,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1737409710884094,
      "num_tokens": 31974322.0,
      "step": 15405
    },
    {
      "entropy": 5.5504100799560545,
      "epoch": 1.3978592162554426,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004809536609109056,
      "loss": 5.3642,
      "mean_token_accuracy": 0.1764423131942749,
      "num_tokens": 31984239.0,
      "step": 15410
    },
    {
      "entropy": 5.534015321731568,
      "epoch": 1.398312772133527,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048094063014428925,
      "loss": 5.3364,
      "mean_token_accuracy": 0.17192057371139527,
      "num_tokens": 31994366.0,
      "step": 15415
    },
    {
      "entropy": 5.638738203048706,
      "epoch": 1.398766328011611,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004809275951187682,
      "loss": 5.403,
      "mean_token_accuracy": 0.1663533017039299,
      "num_tokens": 32004921.0,
      "step": 15420
    },
    {
      "entropy": 5.594494581222534,
      "epoch": 1.3992198838896952,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048091455583461204,
      "loss": 5.3624,
      "mean_token_accuracy": 0.1772454112768173,
      "num_tokens": 32016385.0,
      "step": 15425
    },
    {
      "entropy": 5.6230103969573975,
      "epoch": 1.3996734397677795,
      "grad_norm": 2.0,
      "learning_rate": 0.00048090151229209036,
      "loss": 5.3396,
      "mean_token_accuracy": 0.17050480097532272,
      "num_tokens": 32027398.0,
      "step": 15430
    },
    {
      "entropy": 5.602575016021729,
      "epoch": 1.4001269956458635,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004808884644914729,
      "loss": 5.3231,
      "mean_token_accuracy": 0.17583974748849868,
      "num_tokens": 32038326.0,
      "step": 15435
    },
    {
      "entropy": 5.559352207183838,
      "epoch": 1.4005805515239478,
      "grad_norm": 1.875,
      "learning_rate": 0.0004808754124330296,
      "loss": 5.3992,
      "mean_token_accuracy": 0.1730026677250862,
      "num_tokens": 32048744.0,
      "step": 15440
    },
    {
      "entropy": 5.585540294647217,
      "epoch": 1.4010341074020318,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004808623561170302,
      "loss": 5.2881,
      "mean_token_accuracy": 0.18045868575572968,
      "num_tokens": 32059133.0,
      "step": 15445
    },
    {
      "entropy": 5.5799158096313475,
      "epoch": 1.401487663280116,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004808492955437449,
      "loss": 5.3193,
      "mean_token_accuracy": 0.17536248713731767,
      "num_tokens": 32069277.0,
      "step": 15450
    },
    {
      "entropy": 5.563054609298706,
      "epoch": 1.4019412191582004,
      "grad_norm": 4.0,
      "learning_rate": 0.00048083623071344374,
      "loss": 5.2615,
      "mean_token_accuracy": 0.17921701073646545,
      "num_tokens": 32079802.0,
      "step": 15455
    },
    {
      "entropy": 5.611124515533447,
      "epoch": 1.4023947750362844,
      "grad_norm": 2.75,
      "learning_rate": 0.00048082316162639685,
      "loss": 5.3896,
      "mean_token_accuracy": 0.17074760794639587,
      "num_tokens": 32090962.0,
      "step": 15460
    },
    {
      "entropy": 5.5569041728973385,
      "epoch": 1.4028483309143687,
      "grad_norm": 2.53125,
      "learning_rate": 0.00048081008828287447,
      "loss": 5.34,
      "mean_token_accuracy": 0.1742991641163826,
      "num_tokens": 32101058.0,
      "step": 15465
    },
    {
      "entropy": 5.618633031845093,
      "epoch": 1.4033018867924527,
      "grad_norm": 2.984375,
      "learning_rate": 0.00048079701068314706,
      "loss": 5.3498,
      "mean_token_accuracy": 0.1745060220360756,
      "num_tokens": 32111763.0,
      "step": 15470
    },
    {
      "entropy": 5.642866802215576,
      "epoch": 1.403755442670537,
      "grad_norm": 2.484375,
      "learning_rate": 0.00048078392882748507,
      "loss": 5.4221,
      "mean_token_accuracy": 0.1726563900709152,
      "num_tokens": 32123556.0,
      "step": 15475
    },
    {
      "entropy": 5.61187105178833,
      "epoch": 1.4042089985486212,
      "grad_norm": 2.671875,
      "learning_rate": 0.00048077084271615895,
      "loss": 5.2953,
      "mean_token_accuracy": 0.17316671907901765,
      "num_tokens": 32134012.0,
      "step": 15480
    },
    {
      "entropy": 5.5063676834106445,
      "epoch": 1.4046625544267053,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004807577523494394,
      "loss": 5.2639,
      "mean_token_accuracy": 0.18363068103790284,
      "num_tokens": 32144540.0,
      "step": 15485
    },
    {
      "entropy": 5.578015422821045,
      "epoch": 1.4051161103047896,
      "grad_norm": 1.90625,
      "learning_rate": 0.00048074465772759707,
      "loss": 5.3814,
      "mean_token_accuracy": 0.17300562858581542,
      "num_tokens": 32154655.0,
      "step": 15490
    },
    {
      "entropy": 5.6178028106689455,
      "epoch": 1.4055696661828736,
      "grad_norm": 3.25,
      "learning_rate": 0.00048073155885090276,
      "loss": 5.2791,
      "mean_token_accuracy": 0.17553114145994186,
      "num_tokens": 32165191.0,
      "step": 15495
    },
    {
      "entropy": 5.618647575378418,
      "epoch": 1.4060232220609579,
      "grad_norm": 4.21875,
      "learning_rate": 0.00048071845571962733,
      "loss": 5.3689,
      "mean_token_accuracy": 0.16695779114961623,
      "num_tokens": 32175961.0,
      "step": 15500
    },
    {
      "entropy": 5.590060710906982,
      "epoch": 1.4064767779390421,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004807053483340419,
      "loss": 5.3289,
      "mean_token_accuracy": 0.17218441516160965,
      "num_tokens": 32187073.0,
      "step": 15505
    },
    {
      "entropy": 5.581760215759277,
      "epoch": 1.4069303338171264,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004806922366944175,
      "loss": 5.3245,
      "mean_token_accuracy": 0.17853282392024994,
      "num_tokens": 32197426.0,
      "step": 15510
    },
    {
      "entropy": 5.655210065841675,
      "epoch": 1.4073838896952104,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048067912080102513,
      "loss": 5.2993,
      "mean_token_accuracy": 0.18646540343761445,
      "num_tokens": 32206847.0,
      "step": 15515
    },
    {
      "entropy": 5.50341420173645,
      "epoch": 1.4078374455732947,
      "grad_norm": 2.125,
      "learning_rate": 0.0004806660006541362,
      "loss": 5.3373,
      "mean_token_accuracy": 0.1751742720603943,
      "num_tokens": 32217647.0,
      "step": 15520
    },
    {
      "entropy": 5.497707319259644,
      "epoch": 1.4082910014513788,
      "grad_norm": 3.6875,
      "learning_rate": 0.0004806528762540219,
      "loss": 5.1927,
      "mean_token_accuracy": 0.1859515830874443,
      "num_tokens": 32227133.0,
      "step": 15525
    },
    {
      "entropy": 5.622465038299561,
      "epoch": 1.408744557329463,
      "grad_norm": 3.03125,
      "learning_rate": 0.00048063974760095374,
      "loss": 5.4168,
      "mean_token_accuracy": 0.16662675887346268,
      "num_tokens": 32237378.0,
      "step": 15530
    },
    {
      "entropy": 5.5286284446716305,
      "epoch": 1.4091981132075473,
      "grad_norm": 2.203125,
      "learning_rate": 0.00048062661469520327,
      "loss": 5.2801,
      "mean_token_accuracy": 0.17990752756595613,
      "num_tokens": 32247213.0,
      "step": 15535
    },
    {
      "entropy": 5.636937093734741,
      "epoch": 1.4096516690856313,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048061347753704193,
      "loss": 5.3351,
      "mean_token_accuracy": 0.17815128117799758,
      "num_tokens": 32257100.0,
      "step": 15540
    },
    {
      "entropy": 5.616419076919556,
      "epoch": 1.4101052249637156,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004806003361267415,
      "loss": 5.2919,
      "mean_token_accuracy": 0.17090966403484345,
      "num_tokens": 32267456.0,
      "step": 15545
    },
    {
      "entropy": 5.505732631683349,
      "epoch": 1.4105587808417996,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004805871904645737,
      "loss": 5.2654,
      "mean_token_accuracy": 0.182029390335083,
      "num_tokens": 32278220.0,
      "step": 15550
    },
    {
      "entropy": 5.497928047180176,
      "epoch": 1.411012336719884,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004805740405508103,
      "loss": 5.3202,
      "mean_token_accuracy": 0.18124406337738036,
      "num_tokens": 32289036.0,
      "step": 15555
    },
    {
      "entropy": 5.575134563446045,
      "epoch": 1.4114658925979682,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004805608863857236,
      "loss": 5.2423,
      "mean_token_accuracy": 0.18197584450244902,
      "num_tokens": 32298909.0,
      "step": 15560
    },
    {
      "entropy": 5.594216060638428,
      "epoch": 1.4119194484760522,
      "grad_norm": 2.859375,
      "learning_rate": 0.00048054772796958517,
      "loss": 5.3029,
      "mean_token_accuracy": 0.17125291228294373,
      "num_tokens": 32309851.0,
      "step": 15565
    },
    {
      "entropy": 5.519925212860107,
      "epoch": 1.4123730043541365,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004805345653026675,
      "loss": 5.2483,
      "mean_token_accuracy": 0.18098585307598114,
      "num_tokens": 32319336.0,
      "step": 15570
    },
    {
      "entropy": 5.622329330444336,
      "epoch": 1.4128265602322205,
      "grad_norm": 1.9375,
      "learning_rate": 0.00048052139838524263,
      "loss": 5.3212,
      "mean_token_accuracy": 0.17362552583217622,
      "num_tokens": 32329898.0,
      "step": 15575
    },
    {
      "entropy": 5.560531425476074,
      "epoch": 1.4132801161103048,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048050822721758274,
      "loss": 5.1962,
      "mean_token_accuracy": 0.1757473111152649,
      "num_tokens": 32340100.0,
      "step": 15580
    },
    {
      "entropy": 5.568778800964355,
      "epoch": 1.413733671988389,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004804950517999604,
      "loss": 5.2988,
      "mean_token_accuracy": 0.17427062392234802,
      "num_tokens": 32349663.0,
      "step": 15585
    },
    {
      "entropy": 5.604432106018066,
      "epoch": 1.414187227866473,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004804818721326481,
      "loss": 5.3486,
      "mean_token_accuracy": 0.17533576786518096,
      "num_tokens": 32360461.0,
      "step": 15590
    },
    {
      "entropy": 5.591039228439331,
      "epoch": 1.4146407837445574,
      "grad_norm": 3.125,
      "learning_rate": 0.0004804686882159183,
      "loss": 5.3362,
      "mean_token_accuracy": 0.17856885343790055,
      "num_tokens": 32370910.0,
      "step": 15595
    },
    {
      "entropy": 5.656470441818238,
      "epoch": 1.4150943396226414,
      "grad_norm": 1.921875,
      "learning_rate": 0.00048045550005004355,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18431166857481002,
      "num_tokens": 32381669.0,
      "step": 15600
    },
    {
      "entropy": 5.5526059627532955,
      "epoch": 1.4155478955007257,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004804423076352968,
      "loss": 5.2689,
      "mean_token_accuracy": 0.17448710501194,
      "num_tokens": 32392024.0,
      "step": 15605
    },
    {
      "entropy": 5.522314548492432,
      "epoch": 1.41600145137881,
      "grad_norm": 2.296875,
      "learning_rate": 0.00048042911097195075,
      "loss": 5.2797,
      "mean_token_accuracy": 0.18201130628585815,
      "num_tokens": 32401872.0,
      "step": 15610
    },
    {
      "entropy": 5.5263749122619625,
      "epoch": 1.416455007256894,
      "grad_norm": 2.25,
      "learning_rate": 0.0004804159100602784,
      "loss": 5.2461,
      "mean_token_accuracy": 0.17806998491287232,
      "num_tokens": 32411749.0,
      "step": 15615
    },
    {
      "entropy": 5.669288444519043,
      "epoch": 1.4169085631349783,
      "grad_norm": 2.90625,
      "learning_rate": 0.00048040270490055264,
      "loss": 5.4103,
      "mean_token_accuracy": 0.1736241862177849,
      "num_tokens": 32421358.0,
      "step": 15620
    },
    {
      "entropy": 5.633365726470947,
      "epoch": 1.4173621190130623,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004803894954930465,
      "loss": 5.3222,
      "mean_token_accuracy": 0.17544811815023423,
      "num_tokens": 32432322.0,
      "step": 15625
    },
    {
      "entropy": 5.585404014587402,
      "epoch": 1.4178156748911466,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004803762818380334,
      "loss": 5.3279,
      "mean_token_accuracy": 0.17793203443288802,
      "num_tokens": 32440930.0,
      "step": 15630
    },
    {
      "entropy": 5.603235101699829,
      "epoch": 1.4182692307692308,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048036306393578644,
      "loss": 5.3553,
      "mean_token_accuracy": 0.17149794697761536,
      "num_tokens": 32451124.0,
      "step": 15635
    },
    {
      "entropy": 5.602581548690796,
      "epoch": 1.4187227866473149,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004803498417865789,
      "loss": 5.2993,
      "mean_token_accuracy": 0.1734090492129326,
      "num_tokens": 32460636.0,
      "step": 15640
    },
    {
      "entropy": 5.621681833267212,
      "epoch": 1.4191763425253991,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004803366153906844,
      "loss": 5.3625,
      "mean_token_accuracy": 0.17160202860832213,
      "num_tokens": 32470501.0,
      "step": 15645
    },
    {
      "entropy": 5.639283990859985,
      "epoch": 1.4196298984034832,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004803233847483763,
      "loss": 5.4442,
      "mean_token_accuracy": 0.1704925261437893,
      "num_tokens": 32481559.0,
      "step": 15650
    },
    {
      "entropy": 5.751777648925781,
      "epoch": 1.4200834542815675,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048031014985992825,
      "loss": 5.4855,
      "mean_token_accuracy": 0.16743257343769075,
      "num_tokens": 32493000.0,
      "step": 15655
    },
    {
      "entropy": 5.607096147537232,
      "epoch": 1.4205370101596517,
      "grad_norm": 2.9375,
      "learning_rate": 0.000480296910725614,
      "loss": 5.2847,
      "mean_token_accuracy": 0.1876971110701561,
      "num_tokens": 32502773.0,
      "step": 15660
    },
    {
      "entropy": 5.587019681930542,
      "epoch": 1.4209905660377358,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004802836673457074,
      "loss": 5.3758,
      "mean_token_accuracy": 0.1735771968960762,
      "num_tokens": 32513430.0,
      "step": 15665
    },
    {
      "entropy": 5.603146076202393,
      "epoch": 1.42144412191582,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048027041972048225,
      "loss": 5.288,
      "mean_token_accuracy": 0.1782058894634247,
      "num_tokens": 32524409.0,
      "step": 15670
    },
    {
      "entropy": 5.640337085723877,
      "epoch": 1.421897677793904,
      "grad_norm": 2.96875,
      "learning_rate": 0.0004802571678502124,
      "loss": 5.343,
      "mean_token_accuracy": 0.17156349271535873,
      "num_tokens": 32534696.0,
      "step": 15675
    },
    {
      "entropy": 5.616892147064209,
      "epoch": 1.4223512336719883,
      "grad_norm": 2.25,
      "learning_rate": 0.000480243911735172,
      "loss": 5.3285,
      "mean_token_accuracy": 0.1795778289437294,
      "num_tokens": 32544782.0,
      "step": 15680
    },
    {
      "entropy": 5.5560774326324465,
      "epoch": 1.4228047895500726,
      "grad_norm": 2.125,
      "learning_rate": 0.00048023065137563534,
      "loss": 5.3402,
      "mean_token_accuracy": 0.17047480195760728,
      "num_tokens": 32554818.0,
      "step": 15685
    },
    {
      "entropy": 5.578672075271607,
      "epoch": 1.4232583454281569,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004802173867718764,
      "loss": 5.3625,
      "mean_token_accuracy": 0.17898129522800446,
      "num_tokens": 32565007.0,
      "step": 15690
    },
    {
      "entropy": 5.603560495376587,
      "epoch": 1.423711901306241,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004802041179241697,
      "loss": 5.3101,
      "mean_token_accuracy": 0.17625815272331238,
      "num_tokens": 32575657.0,
      "step": 15695
    },
    {
      "entropy": 5.542565488815308,
      "epoch": 1.4241654571843252,
      "grad_norm": 3.203125,
      "learning_rate": 0.0004801908448327895,
      "loss": 5.3549,
      "mean_token_accuracy": 0.17317947298288344,
      "num_tokens": 32586676.0,
      "step": 15700
    },
    {
      "entropy": 5.619441938400269,
      "epoch": 1.4246190130624092,
      "grad_norm": 7.625,
      "learning_rate": 0.0004801775674980103,
      "loss": 5.2914,
      "mean_token_accuracy": 0.1770435929298401,
      "num_tokens": 32596191.0,
      "step": 15705
    },
    {
      "entropy": 5.668187046051026,
      "epoch": 1.4250725689404935,
      "grad_norm": 1.921875,
      "learning_rate": 0.00048016428592010676,
      "loss": 5.3634,
      "mean_token_accuracy": 0.16804969906806946,
      "num_tokens": 32607178.0,
      "step": 15710
    },
    {
      "entropy": 5.646436548233032,
      "epoch": 1.4255261248185778,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00048015100009935347,
      "loss": 5.2839,
      "mean_token_accuracy": 0.17549687027931213,
      "num_tokens": 32617301.0,
      "step": 15715
    },
    {
      "entropy": 5.513381099700927,
      "epoch": 1.4259796806966618,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004801377100360253,
      "loss": 5.3019,
      "mean_token_accuracy": 0.17838069498538972,
      "num_tokens": 32627395.0,
      "step": 15720
    },
    {
      "entropy": 5.55570387840271,
      "epoch": 1.426433236574746,
      "grad_norm": 1.7890625,
      "learning_rate": 0.000480124415730397,
      "loss": 5.3239,
      "mean_token_accuracy": 0.17893136590719222,
      "num_tokens": 32638460.0,
      "step": 15725
    },
    {
      "entropy": 5.607681703567505,
      "epoch": 1.4268867924528301,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004801111171827434,
      "loss": 5.3668,
      "mean_token_accuracy": 0.1706285983324051,
      "num_tokens": 32649319.0,
      "step": 15730
    },
    {
      "entropy": 5.5936291217803955,
      "epoch": 1.4273403483309144,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004800978143933398,
      "loss": 5.3139,
      "mean_token_accuracy": 0.1776212051510811,
      "num_tokens": 32659976.0,
      "step": 15735
    },
    {
      "entropy": 5.531074953079224,
      "epoch": 1.4277939042089987,
      "grad_norm": 2.734375,
      "learning_rate": 0.00048008450736246094,
      "loss": 5.2415,
      "mean_token_accuracy": 0.17658588886260987,
      "num_tokens": 32670772.0,
      "step": 15740
    },
    {
      "entropy": 5.591675853729248,
      "epoch": 1.4282474600870827,
      "grad_norm": 2.25,
      "learning_rate": 0.00048007119609038237,
      "loss": 5.3828,
      "mean_token_accuracy": 0.16657683551311492,
      "num_tokens": 32682368.0,
      "step": 15745
    },
    {
      "entropy": 5.528488111495972,
      "epoch": 1.428701015965167,
      "grad_norm": 2.09375,
      "learning_rate": 0.00048005788057737907,
      "loss": 5.2248,
      "mean_token_accuracy": 0.1756732389330864,
      "num_tokens": 32692649.0,
      "step": 15750
    },
    {
      "entropy": 5.635058259963989,
      "epoch": 1.429154571843251,
      "grad_norm": 2.96875,
      "learning_rate": 0.0004800445608237266,
      "loss": 5.4533,
      "mean_token_accuracy": 0.17150902152061462,
      "num_tokens": 32704075.0,
      "step": 15755
    },
    {
      "entropy": 5.668446588516235,
      "epoch": 1.4296081277213353,
      "grad_norm": 4.03125,
      "learning_rate": 0.00048003123682970045,
      "loss": 5.4404,
      "mean_token_accuracy": 0.16416675150394439,
      "num_tokens": 32714670.0,
      "step": 15760
    },
    {
      "entropy": 5.5901451587677,
      "epoch": 1.4300616835994195,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048001790859557596,
      "loss": 5.2649,
      "mean_token_accuracy": 0.17869900017976761,
      "num_tokens": 32724601.0,
      "step": 15765
    },
    {
      "entropy": 5.548739814758301,
      "epoch": 1.4305152394775036,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004800045761216289,
      "loss": 5.2442,
      "mean_token_accuracy": 0.17866247594356538,
      "num_tokens": 32735015.0,
      "step": 15770
    },
    {
      "entropy": 5.554587030410767,
      "epoch": 1.4309687953555879,
      "grad_norm": 1.8984375,
      "learning_rate": 0.000479991239408135,
      "loss": 5.3425,
      "mean_token_accuracy": 0.1783635586500168,
      "num_tokens": 32745899.0,
      "step": 15775
    },
    {
      "entropy": 5.592035913467408,
      "epoch": 1.431422351233672,
      "grad_norm": 3.15625,
      "learning_rate": 0.00047997789845537003,
      "loss": 5.2862,
      "mean_token_accuracy": 0.18256122171878814,
      "num_tokens": 32756052.0,
      "step": 15780
    },
    {
      "entropy": 5.5443182468414305,
      "epoch": 1.4318759071117562,
      "grad_norm": 2.953125,
      "learning_rate": 0.0004799645532636098,
      "loss": 5.3106,
      "mean_token_accuracy": 0.17663087844848632,
      "num_tokens": 32767118.0,
      "step": 15785
    },
    {
      "entropy": 5.511271047592163,
      "epoch": 1.4323294629898404,
      "grad_norm": 4.625,
      "learning_rate": 0.00047995120383313044,
      "loss": 5.1913,
      "mean_token_accuracy": 0.18685914129018782,
      "num_tokens": 32777069.0,
      "step": 15790
    },
    {
      "entropy": 5.591581344604492,
      "epoch": 1.4327830188679245,
      "grad_norm": 2.171875,
      "learning_rate": 0.00047993785016420796,
      "loss": 5.4052,
      "mean_token_accuracy": 0.1662734940648079,
      "num_tokens": 32786967.0,
      "step": 15795
    },
    {
      "entropy": 5.616228342056274,
      "epoch": 1.4332365747460087,
      "grad_norm": 2.6875,
      "learning_rate": 0.00047992449225711854,
      "loss": 5.2708,
      "mean_token_accuracy": 0.17899943441152572,
      "num_tokens": 32797563.0,
      "step": 15800
    },
    {
      "entropy": 5.600063610076904,
      "epoch": 1.4336901306240928,
      "grad_norm": 2.15625,
      "learning_rate": 0.00047991113011213837,
      "loss": 5.2473,
      "mean_token_accuracy": 0.18090436160564421,
      "num_tokens": 32808730.0,
      "step": 15805
    },
    {
      "entropy": 5.550161266326905,
      "epoch": 1.434143686502177,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004798977637295438,
      "loss": 5.2362,
      "mean_token_accuracy": 0.1853669688105583,
      "num_tokens": 32818050.0,
      "step": 15810
    },
    {
      "entropy": 5.594446659088135,
      "epoch": 1.4345972423802613,
      "grad_norm": 3.140625,
      "learning_rate": 0.0004798843931096113,
      "loss": 5.2931,
      "mean_token_accuracy": 0.176404969394207,
      "num_tokens": 32828993.0,
      "step": 15815
    },
    {
      "entropy": 5.584015035629273,
      "epoch": 1.4350507982583454,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004798710182526173,
      "loss": 5.3105,
      "mean_token_accuracy": 0.1836951866745949,
      "num_tokens": 32839331.0,
      "step": 15820
    },
    {
      "entropy": 5.5648500442504885,
      "epoch": 1.4355043541364296,
      "grad_norm": 2.546875,
      "learning_rate": 0.00047985763915883847,
      "loss": 5.3051,
      "mean_token_accuracy": 0.17554595619440078,
      "num_tokens": 32849946.0,
      "step": 15825
    },
    {
      "entropy": 5.49817967414856,
      "epoch": 1.4359579100145137,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004798442558285514,
      "loss": 5.2389,
      "mean_token_accuracy": 0.17933938950300216,
      "num_tokens": 32859960.0,
      "step": 15830
    },
    {
      "entropy": 5.563261795043945,
      "epoch": 1.436411465892598,
      "grad_norm": 3.4375,
      "learning_rate": 0.0004798308682620329,
      "loss": 5.2698,
      "mean_token_accuracy": 0.175748310983181,
      "num_tokens": 32870317.0,
      "step": 15835
    },
    {
      "entropy": 5.663985538482666,
      "epoch": 1.4368650217706822,
      "grad_norm": 1.984375,
      "learning_rate": 0.00047981747645955996,
      "loss": 5.4236,
      "mean_token_accuracy": 0.1691619411110878,
      "num_tokens": 32880906.0,
      "step": 15840
    },
    {
      "entropy": 5.64682354927063,
      "epoch": 1.4373185776487662,
      "grad_norm": 3.328125,
      "learning_rate": 0.0004798040804214094,
      "loss": 5.4475,
      "mean_token_accuracy": 0.17020842283964158,
      "num_tokens": 32891038.0,
      "step": 15845
    },
    {
      "entropy": 5.621930074691773,
      "epoch": 1.4377721335268505,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004797906801478582,
      "loss": 5.4192,
      "mean_token_accuracy": 0.1774629458785057,
      "num_tokens": 32901013.0,
      "step": 15850
    },
    {
      "entropy": 5.620538663864136,
      "epoch": 1.4382256894049346,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004797772756391836,
      "loss": 5.3323,
      "mean_token_accuracy": 0.17698168009519577,
      "num_tokens": 32911592.0,
      "step": 15855
    },
    {
      "entropy": 5.605400419235229,
      "epoch": 1.4386792452830188,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004797638668956627,
      "loss": 5.219,
      "mean_token_accuracy": 0.17972943037748337,
      "num_tokens": 32922087.0,
      "step": 15860
    },
    {
      "entropy": 5.619326210021972,
      "epoch": 1.439132801161103,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004797504539175729,
      "loss": 5.3123,
      "mean_token_accuracy": 0.18107516467571258,
      "num_tokens": 32932039.0,
      "step": 15865
    },
    {
      "entropy": 5.575647926330566,
      "epoch": 1.4395863570391874,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004797370367051915,
      "loss": 5.2951,
      "mean_token_accuracy": 0.1767032504081726,
      "num_tokens": 32941896.0,
      "step": 15870
    },
    {
      "entropy": 5.573793888092041,
      "epoch": 1.4400399129172714,
      "grad_norm": 2.140625,
      "learning_rate": 0.000479723615258796,
      "loss": 5.2917,
      "mean_token_accuracy": 0.17015568614006044,
      "num_tokens": 32951675.0,
      "step": 15875
    },
    {
      "entropy": 5.635651445388794,
      "epoch": 1.4404934687953557,
      "grad_norm": 1.9375,
      "learning_rate": 0.00047971018957866396,
      "loss": 5.3938,
      "mean_token_accuracy": 0.18106958419084548,
      "num_tokens": 32961378.0,
      "step": 15880
    },
    {
      "entropy": 5.613216304779053,
      "epoch": 1.4409470246734397,
      "grad_norm": 2.515625,
      "learning_rate": 0.00047969675966507314,
      "loss": 5.2749,
      "mean_token_accuracy": 0.1759553387761116,
      "num_tokens": 32971770.0,
      "step": 15885
    },
    {
      "entropy": 5.5462287902832035,
      "epoch": 1.441400580551524,
      "grad_norm": 2.03125,
      "learning_rate": 0.000479683325518301,
      "loss": 5.3205,
      "mean_token_accuracy": 0.17446187734603882,
      "num_tokens": 32982934.0,
      "step": 15890
    },
    {
      "entropy": 5.548264455795288,
      "epoch": 1.4418541364296082,
      "grad_norm": 2.78125,
      "learning_rate": 0.0004796698871386257,
      "loss": 5.3128,
      "mean_token_accuracy": 0.17143182158470155,
      "num_tokens": 32993252.0,
      "step": 15895
    },
    {
      "entropy": 5.592018747329712,
      "epoch": 1.4423076923076923,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004796564445263249,
      "loss": 5.3397,
      "mean_token_accuracy": 0.17018009275197982,
      "num_tokens": 33003610.0,
      "step": 15900
    },
    {
      "entropy": 5.622594690322876,
      "epoch": 1.4427612481857766,
      "grad_norm": 2.625,
      "learning_rate": 0.00047964299768167665,
      "loss": 5.2915,
      "mean_token_accuracy": 0.17889841198921203,
      "num_tokens": 33014037.0,
      "step": 15905
    },
    {
      "entropy": 5.565037584304809,
      "epoch": 1.4432148040638606,
      "grad_norm": 2.875,
      "learning_rate": 0.0004796295466049591,
      "loss": 5.3084,
      "mean_token_accuracy": 0.17223209887742996,
      "num_tokens": 33023299.0,
      "step": 15910
    },
    {
      "entropy": 5.607126760482788,
      "epoch": 1.4436683599419449,
      "grad_norm": 2.828125,
      "learning_rate": 0.00047961609129645036,
      "loss": 5.3548,
      "mean_token_accuracy": 0.17241258025169373,
      "num_tokens": 33034194.0,
      "step": 15915
    },
    {
      "entropy": 5.567903280258179,
      "epoch": 1.4441219158200291,
      "grad_norm": 3.375,
      "learning_rate": 0.00047960263175642867,
      "loss": 5.2368,
      "mean_token_accuracy": 0.1804424375295639,
      "num_tokens": 33044715.0,
      "step": 15920
    },
    {
      "entropy": 5.507661199569702,
      "epoch": 1.4445754716981132,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004795891679851724,
      "loss": 5.2687,
      "mean_token_accuracy": 0.18387278616428376,
      "num_tokens": 33055214.0,
      "step": 15925
    },
    {
      "entropy": 5.556830310821534,
      "epoch": 1.4450290275761974,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004795756999829601,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1800214797258377,
      "num_tokens": 33065026.0,
      "step": 15930
    },
    {
      "entropy": 5.664533948898315,
      "epoch": 1.4454825834542815,
      "grad_norm": 2.4375,
      "learning_rate": 0.00047956222775007016,
      "loss": 5.3306,
      "mean_token_accuracy": 0.17116337567567824,
      "num_tokens": 33075138.0,
      "step": 15935
    },
    {
      "entropy": 5.6261053562164305,
      "epoch": 1.4459361393323658,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004795487512867812,
      "loss": 5.3253,
      "mean_token_accuracy": 0.1768587648868561,
      "num_tokens": 33086625.0,
      "step": 15940
    },
    {
      "entropy": 5.632616567611694,
      "epoch": 1.44638969521045,
      "grad_norm": 2.5,
      "learning_rate": 0.00047953527059337194,
      "loss": 5.3719,
      "mean_token_accuracy": 0.16886406689882277,
      "num_tokens": 33097076.0,
      "step": 15945
    },
    {
      "entropy": 5.582935428619384,
      "epoch": 1.446843251088534,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004795217856701212,
      "loss": 5.3089,
      "mean_token_accuracy": 0.17497887164354325,
      "num_tokens": 33107168.0,
      "step": 15950
    },
    {
      "entropy": 5.6572021484375,
      "epoch": 1.4472968069666183,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00047950829651730774,
      "loss": 5.3593,
      "mean_token_accuracy": 0.17300489842891692,
      "num_tokens": 33117862.0,
      "step": 15955
    },
    {
      "entropy": 5.6208662509918215,
      "epoch": 1.4477503628447024,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004794948031352107,
      "loss": 5.2648,
      "mean_token_accuracy": 0.18001434355974197,
      "num_tokens": 33127144.0,
      "step": 15960
    },
    {
      "entropy": 5.6349778175354,
      "epoch": 1.4482039187227866,
      "grad_norm": 4.5625,
      "learning_rate": 0.00047948130552410897,
      "loss": 5.4069,
      "mean_token_accuracy": 0.16833055764436722,
      "num_tokens": 33139180.0,
      "step": 15965
    },
    {
      "entropy": 5.642861032485962,
      "epoch": 1.448657474600871,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00047946780368428175,
      "loss": 5.311,
      "mean_token_accuracy": 0.17056434005498886,
      "num_tokens": 33149467.0,
      "step": 15970
    },
    {
      "entropy": 5.566349029541016,
      "epoch": 1.449111030478955,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004794542976160081,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19035275876522065,
      "num_tokens": 33158752.0,
      "step": 15975
    },
    {
      "entropy": 5.505987596511841,
      "epoch": 1.4495645863570392,
      "grad_norm": 1.78125,
      "learning_rate": 0.00047944078731956765,
      "loss": 5.2334,
      "mean_token_accuracy": 0.1816803738474846,
      "num_tokens": 33168242.0,
      "step": 15980
    },
    {
      "entropy": 5.496522617340088,
      "epoch": 1.4500181422351233,
      "grad_norm": 2.296875,
      "learning_rate": 0.00047942727279523957,
      "loss": 5.2612,
      "mean_token_accuracy": 0.17875745594501496,
      "num_tokens": 33179134.0,
      "step": 15985
    },
    {
      "entropy": 5.591103553771973,
      "epoch": 1.4504716981132075,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004794137540433034,
      "loss": 5.3602,
      "mean_token_accuracy": 0.17043962478637695,
      "num_tokens": 33188388.0,
      "step": 15990
    },
    {
      "entropy": 5.630726957321167,
      "epoch": 1.4509252539912918,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004794002310640386,
      "loss": 5.2841,
      "mean_token_accuracy": 0.17821843773126603,
      "num_tokens": 33198750.0,
      "step": 15995
    },
    {
      "entropy": 5.706152105331421,
      "epoch": 1.4513788098693758,
      "grad_norm": 3.328125,
      "learning_rate": 0.00047938670385772503,
      "loss": 5.4086,
      "mean_token_accuracy": 0.16741834580898285,
      "num_tokens": 33209532.0,
      "step": 16000
    },
    {
      "entropy": 5.591556978225708,
      "epoch": 1.45183236574746,
      "grad_norm": 2.5625,
      "learning_rate": 0.00047937317242464244,
      "loss": 5.2805,
      "mean_token_accuracy": 0.1794445186853409,
      "num_tokens": 33219675.0,
      "step": 16005
    },
    {
      "entropy": 5.555029487609863,
      "epoch": 1.4522859216255442,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004793596367650704,
      "loss": 5.3558,
      "mean_token_accuracy": 0.1733049526810646,
      "num_tokens": 33230919.0,
      "step": 16010
    },
    {
      "entropy": 5.573984956741333,
      "epoch": 1.4527394775036284,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004793460968792891,
      "loss": 5.2754,
      "mean_token_accuracy": 0.18246083408594133,
      "num_tokens": 33241146.0,
      "step": 16015
    },
    {
      "entropy": 5.604196310043335,
      "epoch": 1.4531930333817127,
      "grad_norm": 3.0625,
      "learning_rate": 0.00047933255276757836,
      "loss": 5.3213,
      "mean_token_accuracy": 0.17690571993589402,
      "num_tokens": 33251434.0,
      "step": 16020
    },
    {
      "entropy": 5.641094398498535,
      "epoch": 1.4536465892597967,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004793190044302184,
      "loss": 5.4151,
      "mean_token_accuracy": 0.1653292953968048,
      "num_tokens": 33261868.0,
      "step": 16025
    },
    {
      "entropy": 5.588883399963379,
      "epoch": 1.454100145137881,
      "grad_norm": 2.25,
      "learning_rate": 0.00047930545186748936,
      "loss": 5.2332,
      "mean_token_accuracy": 0.1792287826538086,
      "num_tokens": 33271798.0,
      "step": 16030
    },
    {
      "entropy": 5.5627909183502195,
      "epoch": 1.454553701015965,
      "grad_norm": 2.40625,
      "learning_rate": 0.00047929189507967153,
      "loss": 5.35,
      "mean_token_accuracy": 0.17282679229974746,
      "num_tokens": 33282614.0,
      "step": 16035
    },
    {
      "entropy": 5.519088888168335,
      "epoch": 1.4550072568940493,
      "grad_norm": 2.65625,
      "learning_rate": 0.00047927833406704516,
      "loss": 5.2636,
      "mean_token_accuracy": 0.17347875535488128,
      "num_tokens": 33292906.0,
      "step": 16040
    },
    {
      "entropy": 5.560147380828857,
      "epoch": 1.4554608127721336,
      "grad_norm": 2.875,
      "learning_rate": 0.0004792647688298909,
      "loss": 5.283,
      "mean_token_accuracy": 0.17912703156471252,
      "num_tokens": 33302720.0,
      "step": 16045
    },
    {
      "entropy": 5.5463025093078615,
      "epoch": 1.4559143686502178,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004792511993684891,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18150248974561692,
      "num_tokens": 33312959.0,
      "step": 16050
    },
    {
      "entropy": 5.561960506439209,
      "epoch": 1.4563679245283019,
      "grad_norm": 2.328125,
      "learning_rate": 0.00047923762568312053,
      "loss": 5.2332,
      "mean_token_accuracy": 0.17864568084478377,
      "num_tokens": 33322737.0,
      "step": 16055
    },
    {
      "entropy": 5.515981245040893,
      "epoch": 1.4568214804063861,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004792240477740657,
      "loss": 5.2328,
      "mean_token_accuracy": 0.1829989343881607,
      "num_tokens": 33333185.0,
      "step": 16060
    },
    {
      "entropy": 5.579063129425049,
      "epoch": 1.4572750362844702,
      "grad_norm": 2.796875,
      "learning_rate": 0.0004792104656416056,
      "loss": 5.3443,
      "mean_token_accuracy": 0.1680214062333107,
      "num_tokens": 33343414.0,
      "step": 16065
    },
    {
      "entropy": 5.570414733886719,
      "epoch": 1.4577285921625545,
      "grad_norm": 2.0625,
      "learning_rate": 0.00047919687928602104,
      "loss": 5.2843,
      "mean_token_accuracy": 0.17771936655044557,
      "num_tokens": 33352527.0,
      "step": 16070
    },
    {
      "entropy": 5.618252801895141,
      "epoch": 1.4581821480406387,
      "grad_norm": 2.0,
      "learning_rate": 0.00047918328870759293,
      "loss": 5.3403,
      "mean_token_accuracy": 0.17243436127901077,
      "num_tokens": 33363526.0,
      "step": 16075
    },
    {
      "entropy": 5.573766613006592,
      "epoch": 1.4586357039187228,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004791696939066024,
      "loss": 5.3551,
      "mean_token_accuracy": 0.17930236905813218,
      "num_tokens": 33374302.0,
      "step": 16080
    },
    {
      "entropy": 5.607821607589722,
      "epoch": 1.459089259796807,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004791560948833306,
      "loss": 5.3181,
      "mean_token_accuracy": 0.1794634371995926,
      "num_tokens": 33384786.0,
      "step": 16085
    },
    {
      "entropy": 5.6230028629302975,
      "epoch": 1.459542815674891,
      "grad_norm": 3.4375,
      "learning_rate": 0.0004791424916380586,
      "loss": 5.2941,
      "mean_token_accuracy": 0.18244815915822982,
      "num_tokens": 33394464.0,
      "step": 16090
    },
    {
      "entropy": 5.609851598739624,
      "epoch": 1.4599963715529753,
      "grad_norm": 5.0,
      "learning_rate": 0.000479128884171068,
      "loss": 5.3248,
      "mean_token_accuracy": 0.17386998534202575,
      "num_tokens": 33405476.0,
      "step": 16095
    },
    {
      "entropy": 5.5651795864105225,
      "epoch": 1.4604499274310596,
      "grad_norm": 3.875,
      "learning_rate": 0.00047911527248264003,
      "loss": 5.3592,
      "mean_token_accuracy": 0.1755985587835312,
      "num_tokens": 33415758.0,
      "step": 16100
    },
    {
      "entropy": 5.567941427230835,
      "epoch": 1.4609034833091437,
      "grad_norm": 3.125,
      "learning_rate": 0.00047910165657305613,
      "loss": 5.3129,
      "mean_token_accuracy": 0.17710541039705277,
      "num_tokens": 33424845.0,
      "step": 16105
    },
    {
      "entropy": 5.64595594406128,
      "epoch": 1.461357039187228,
      "grad_norm": 2.515625,
      "learning_rate": 0.00047908803644259804,
      "loss": 5.3537,
      "mean_token_accuracy": 0.17239485532045365,
      "num_tokens": 33435928.0,
      "step": 16110
    },
    {
      "entropy": 5.583413457870483,
      "epoch": 1.461810595065312,
      "grad_norm": 2.265625,
      "learning_rate": 0.00047907441209154726,
      "loss": 5.2591,
      "mean_token_accuracy": 0.1753436580300331,
      "num_tokens": 33447307.0,
      "step": 16115
    },
    {
      "entropy": 5.573362064361572,
      "epoch": 1.4622641509433962,
      "grad_norm": 2.328125,
      "learning_rate": 0.00047906078352018575,
      "loss": 5.3087,
      "mean_token_accuracy": 0.17144439816474916,
      "num_tokens": 33458116.0,
      "step": 16120
    },
    {
      "entropy": 5.438568162918091,
      "epoch": 1.4627177068214805,
      "grad_norm": 3.125,
      "learning_rate": 0.0004790471507287952,
      "loss": 5.1791,
      "mean_token_accuracy": 0.18206090480089188,
      "num_tokens": 33469065.0,
      "step": 16125
    },
    {
      "entropy": 5.564370965957641,
      "epoch": 1.4631712626995645,
      "grad_norm": 3.671875,
      "learning_rate": 0.00047903351371765757,
      "loss": 5.2955,
      "mean_token_accuracy": 0.17744127959012984,
      "num_tokens": 33479118.0,
      "step": 16130
    },
    {
      "entropy": 5.6320398330688475,
      "epoch": 1.4636248185776488,
      "grad_norm": 2.328125,
      "learning_rate": 0.00047901987248705484,
      "loss": 5.3446,
      "mean_token_accuracy": 0.17615798711776734,
      "num_tokens": 33489642.0,
      "step": 16135
    },
    {
      "entropy": 5.518034648895264,
      "epoch": 1.4640783744557329,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004790062270372693,
      "loss": 5.3485,
      "mean_token_accuracy": 0.16897599697113036,
      "num_tokens": 33499771.0,
      "step": 16140
    },
    {
      "entropy": 5.5140910148620605,
      "epoch": 1.4645319303338171,
      "grad_norm": 2.921875,
      "learning_rate": 0.0004789925773685828,
      "loss": 5.3086,
      "mean_token_accuracy": 0.17108865678310395,
      "num_tokens": 33509975.0,
      "step": 16145
    },
    {
      "entropy": 5.639580917358399,
      "epoch": 1.4649854862119014,
      "grad_norm": 2.625,
      "learning_rate": 0.0004789789234812779,
      "loss": 5.3503,
      "mean_token_accuracy": 0.17270530462265016,
      "num_tokens": 33520017.0,
      "step": 16150
    },
    {
      "entropy": 5.647706842422485,
      "epoch": 1.4654390420899854,
      "grad_norm": 2.3125,
      "learning_rate": 0.00047896526537563686,
      "loss": 5.4004,
      "mean_token_accuracy": 0.16492098644375802,
      "num_tokens": 33531977.0,
      "step": 16155
    },
    {
      "entropy": 5.600175666809082,
      "epoch": 1.4658925979680697,
      "grad_norm": 2.4375,
      "learning_rate": 0.00047895160305194223,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1743601605296135,
      "num_tokens": 33543660.0,
      "step": 16160
    },
    {
      "entropy": 5.58420958518982,
      "epoch": 1.4663461538461537,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004789379365104764,
      "loss": 5.2946,
      "mean_token_accuracy": 0.17684556394815446,
      "num_tokens": 33553814.0,
      "step": 16165
    },
    {
      "entropy": 5.58945369720459,
      "epoch": 1.466799709724238,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004789242657515221,
      "loss": 5.3104,
      "mean_token_accuracy": 0.1810485988855362,
      "num_tokens": 33564250.0,
      "step": 16170
    },
    {
      "entropy": 5.591405248641967,
      "epoch": 1.4672532656023223,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004789105907753621,
      "loss": 5.3146,
      "mean_token_accuracy": 0.17684276252985,
      "num_tokens": 33574755.0,
      "step": 16175
    },
    {
      "entropy": 5.568609046936035,
      "epoch": 1.4677068214804063,
      "grad_norm": 4.8125,
      "learning_rate": 0.000478896911582279,
      "loss": 5.2334,
      "mean_token_accuracy": 0.1777913749217987,
      "num_tokens": 33585268.0,
      "step": 16180
    },
    {
      "entropy": 5.60820894241333,
      "epoch": 1.4681603773584906,
      "grad_norm": 8.125,
      "learning_rate": 0.0004788832281725558,
      "loss": 5.3672,
      "mean_token_accuracy": 0.17676627784967422,
      "num_tokens": 33595753.0,
      "step": 16185
    },
    {
      "entropy": 5.5418128490448,
      "epoch": 1.4686139332365746,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004788695405464755,
      "loss": 5.2353,
      "mean_token_accuracy": 0.17875579595565796,
      "num_tokens": 33605270.0,
      "step": 16190
    },
    {
      "entropy": 5.58299298286438,
      "epoch": 1.469067489114659,
      "grad_norm": 2.578125,
      "learning_rate": 0.00047885584870432127,
      "loss": 5.3464,
      "mean_token_accuracy": 0.1789108082652092,
      "num_tokens": 33615096.0,
      "step": 16195
    },
    {
      "entropy": 5.584152793884277,
      "epoch": 1.4695210449927432,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047884215264637606,
      "loss": 5.218,
      "mean_token_accuracy": 0.18002844452857972,
      "num_tokens": 33624770.0,
      "step": 16200
    },
    {
      "entropy": 5.621974849700928,
      "epoch": 1.4699746008708272,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004788284523729232,
      "loss": 5.3718,
      "mean_token_accuracy": 0.17021364867687225,
      "num_tokens": 33634545.0,
      "step": 16205
    },
    {
      "entropy": 5.653466320037841,
      "epoch": 1.4704281567489115,
      "grad_norm": 1.890625,
      "learning_rate": 0.00047881474788424595,
      "loss": 5.343,
      "mean_token_accuracy": 0.17663231790065764,
      "num_tokens": 33644950.0,
      "step": 16210
    },
    {
      "entropy": 5.624150896072388,
      "epoch": 1.4708817126269955,
      "grad_norm": 2.375,
      "learning_rate": 0.00047880103918062776,
      "loss": 5.2631,
      "mean_token_accuracy": 0.18180761039257048,
      "num_tokens": 33654369.0,
      "step": 16215
    },
    {
      "entropy": 5.492585563659668,
      "epoch": 1.4713352685050798,
      "grad_norm": 2.21875,
      "learning_rate": 0.00047878732626235225,
      "loss": 5.314,
      "mean_token_accuracy": 0.17613308727741242,
      "num_tokens": 33664687.0,
      "step": 16220
    },
    {
      "entropy": 5.543760108947754,
      "epoch": 1.471788824383164,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004787736091297028,
      "loss": 5.2851,
      "mean_token_accuracy": 0.18668262660503387,
      "num_tokens": 33675832.0,
      "step": 16225
    },
    {
      "entropy": 5.690488052368164,
      "epoch": 1.4722423802612483,
      "grad_norm": 3.15625,
      "learning_rate": 0.0004787598877829633,
      "loss": 5.3534,
      "mean_token_accuracy": 0.1757723182439804,
      "num_tokens": 33686788.0,
      "step": 16230
    },
    {
      "entropy": 5.585574007034301,
      "epoch": 1.4726959361393324,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004787461622224173,
      "loss": 5.2653,
      "mean_token_accuracy": 0.17948741614818572,
      "num_tokens": 33696384.0,
      "step": 16235
    },
    {
      "entropy": 5.533924245834351,
      "epoch": 1.4731494920174166,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004787324324483488,
      "loss": 5.284,
      "mean_token_accuracy": 0.1765810877084732,
      "num_tokens": 33707758.0,
      "step": 16240
    },
    {
      "entropy": 5.556590366363525,
      "epoch": 1.4736030478955007,
      "grad_norm": 3.390625,
      "learning_rate": 0.0004787186984610416,
      "loss": 5.2523,
      "mean_token_accuracy": 0.18424173295497895,
      "num_tokens": 33717747.0,
      "step": 16245
    },
    {
      "entropy": 5.554596042633056,
      "epoch": 1.474056603773585,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00047870496026077984,
      "loss": 5.2671,
      "mean_token_accuracy": 0.1738946631550789,
      "num_tokens": 33727960.0,
      "step": 16250
    },
    {
      "entropy": 5.513382291793823,
      "epoch": 1.4745101596516692,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004786912178478476,
      "loss": 5.2469,
      "mean_token_accuracy": 0.17737340927124023,
      "num_tokens": 33738874.0,
      "step": 16255
    },
    {
      "entropy": 5.519195222854615,
      "epoch": 1.4749637155297532,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004786774712225291,
      "loss": 5.2399,
      "mean_token_accuracy": 0.18178923577070236,
      "num_tokens": 33749071.0,
      "step": 16260
    },
    {
      "entropy": 5.522031307220459,
      "epoch": 1.4754172714078375,
      "grad_norm": 2.96875,
      "learning_rate": 0.0004786637203851086,
      "loss": 5.2716,
      "mean_token_accuracy": 0.17326659858226776,
      "num_tokens": 33759509.0,
      "step": 16265
    },
    {
      "entropy": 5.620713329315185,
      "epoch": 1.4758708272859216,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004786499653358705,
      "loss": 5.3245,
      "mean_token_accuracy": 0.17957060188055038,
      "num_tokens": 33769974.0,
      "step": 16270
    },
    {
      "entropy": 5.679375267028808,
      "epoch": 1.4763243831640058,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004786362060750991,
      "loss": 5.3274,
      "mean_token_accuracy": 0.179963868111372,
      "num_tokens": 33780403.0,
      "step": 16275
    },
    {
      "entropy": 5.59434061050415,
      "epoch": 1.47677793904209,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047862244260307913,
      "loss": 5.3694,
      "mean_token_accuracy": 0.17553676217794417,
      "num_tokens": 33791364.0,
      "step": 16280
    },
    {
      "entropy": 5.551088047027588,
      "epoch": 1.4772314949201741,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004786086749200951,
      "loss": 5.2785,
      "mean_token_accuracy": 0.1771296203136444,
      "num_tokens": 33801964.0,
      "step": 16285
    },
    {
      "entropy": 5.580364656448364,
      "epoch": 1.4776850507982584,
      "grad_norm": 4.65625,
      "learning_rate": 0.0004785949030264319,
      "loss": 5.3024,
      "mean_token_accuracy": 0.18014794141054152,
      "num_tokens": 33813132.0,
      "step": 16290
    },
    {
      "entropy": 5.551224994659424,
      "epoch": 1.4781386066763424,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004785811269223741,
      "loss": 5.2772,
      "mean_token_accuracy": 0.18419003039598464,
      "num_tokens": 33823069.0,
      "step": 16295
    },
    {
      "entropy": 5.5988410949707035,
      "epoch": 1.4785921625544267,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004785673466082068,
      "loss": 5.2768,
      "mean_token_accuracy": 0.17857310473918914,
      "num_tokens": 33832626.0,
      "step": 16300
    },
    {
      "entropy": 5.527375602722168,
      "epoch": 1.479045718432511,
      "grad_norm": 2.09375,
      "learning_rate": 0.00047855356208421484,
      "loss": 5.2751,
      "mean_token_accuracy": 0.17451235949993132,
      "num_tokens": 33843384.0,
      "step": 16305
    },
    {
      "entropy": 5.521923542022705,
      "epoch": 1.479499274310595,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004785397733506834,
      "loss": 5.2638,
      "mean_token_accuracy": 0.18057364225387573,
      "num_tokens": 33854286.0,
      "step": 16310
    },
    {
      "entropy": 5.563215065002441,
      "epoch": 1.4799528301886793,
      "grad_norm": 3.296875,
      "learning_rate": 0.0004785259804078976,
      "loss": 5.2343,
      "mean_token_accuracy": 0.17960118055343627,
      "num_tokens": 33863379.0,
      "step": 16315
    },
    {
      "entropy": 5.647448635101318,
      "epoch": 1.4804063860667633,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004785121832561426,
      "loss": 5.4498,
      "mean_token_accuracy": 0.17131491154432296,
      "num_tokens": 33875641.0,
      "step": 16320
    },
    {
      "entropy": 5.594933700561524,
      "epoch": 1.4808599419448476,
      "grad_norm": 2.09375,
      "learning_rate": 0.00047849838189570375,
      "loss": 5.2479,
      "mean_token_accuracy": 0.18560532629489898,
      "num_tokens": 33885023.0,
      "step": 16325
    },
    {
      "entropy": 5.64141526222229,
      "epoch": 1.4813134978229319,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004784845763268666,
      "loss": 5.3737,
      "mean_token_accuracy": 0.16750590056180953,
      "num_tokens": 33895562.0,
      "step": 16330
    },
    {
      "entropy": 5.5561380863189695,
      "epoch": 1.481767053701016,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004784707665499165,
      "loss": 5.3235,
      "mean_token_accuracy": 0.1791413202881813,
      "num_tokens": 33907095.0,
      "step": 16335
    },
    {
      "entropy": 5.506449413299561,
      "epoch": 1.4822206095791002,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004784569525651391,
      "loss": 5.1278,
      "mean_token_accuracy": 0.18808289617300034,
      "num_tokens": 33917475.0,
      "step": 16340
    },
    {
      "entropy": 5.528838109970093,
      "epoch": 1.4826741654571842,
      "grad_norm": 3.203125,
      "learning_rate": 0.00047844313437282005,
      "loss": 5.284,
      "mean_token_accuracy": 0.17913316786289216,
      "num_tokens": 33927740.0,
      "step": 16345
    },
    {
      "entropy": 5.531650066375732,
      "epoch": 1.4831277213352685,
      "grad_norm": 1.90625,
      "learning_rate": 0.00047842931197324525,
      "loss": 5.2368,
      "mean_token_accuracy": 0.17942603081464767,
      "num_tokens": 33937416.0,
      "step": 16350
    },
    {
      "entropy": 5.520142269134522,
      "epoch": 1.4835812772133528,
      "grad_norm": 2.375,
      "learning_rate": 0.00047841548536670033,
      "loss": 5.2052,
      "mean_token_accuracy": 0.18642400950193405,
      "num_tokens": 33946968.0,
      "step": 16355
    },
    {
      "entropy": 5.655274677276611,
      "epoch": 1.4840348330914368,
      "grad_norm": 2.34375,
      "learning_rate": 0.00047840165455347136,
      "loss": 5.3511,
      "mean_token_accuracy": 0.17068121284246446,
      "num_tokens": 33957877.0,
      "step": 16360
    },
    {
      "entropy": 5.633646440505982,
      "epoch": 1.484488388969521,
      "grad_norm": 1.765625,
      "learning_rate": 0.00047838781953384436,
      "loss": 5.3413,
      "mean_token_accuracy": 0.17185924500226973,
      "num_tokens": 33967890.0,
      "step": 16365
    },
    {
      "entropy": 5.523364925384522,
      "epoch": 1.484941944847605,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004783739803081054,
      "loss": 5.2751,
      "mean_token_accuracy": 0.17618156224489212,
      "num_tokens": 33977117.0,
      "step": 16370
    },
    {
      "entropy": 5.5796263217926025,
      "epoch": 1.4853955007256894,
      "grad_norm": 2.578125,
      "learning_rate": 0.00047836013687654075,
      "loss": 5.3761,
      "mean_token_accuracy": 0.17150522023439407,
      "num_tokens": 33986943.0,
      "step": 16375
    },
    {
      "entropy": 5.536896848678589,
      "epoch": 1.4858490566037736,
      "grad_norm": 2.71875,
      "learning_rate": 0.00047834628923943654,
      "loss": 5.2697,
      "mean_token_accuracy": 0.1773731917142868,
      "num_tokens": 33996499.0,
      "step": 16380
    },
    {
      "entropy": 5.632357501983643,
      "epoch": 1.4863026124818577,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004783324373970793,
      "loss": 5.3896,
      "mean_token_accuracy": 0.17493083626031875,
      "num_tokens": 34006749.0,
      "step": 16385
    },
    {
      "entropy": 5.640044116973877,
      "epoch": 1.486756168359942,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004783185813497555,
      "loss": 5.3833,
      "mean_token_accuracy": 0.16676943600177765,
      "num_tokens": 34017194.0,
      "step": 16390
    },
    {
      "entropy": 5.614242506027222,
      "epoch": 1.487209724238026,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004783047210977516,
      "loss": 5.3221,
      "mean_token_accuracy": 0.17638596296310424,
      "num_tokens": 34027386.0,
      "step": 16395
    },
    {
      "entropy": 5.6528648853302,
      "epoch": 1.4876632801161103,
      "grad_norm": 2.171875,
      "learning_rate": 0.00047829085664135424,
      "loss": 5.3531,
      "mean_token_accuracy": 0.17121404707431792,
      "num_tokens": 34037540.0,
      "step": 16400
    },
    {
      "entropy": 5.598683261871338,
      "epoch": 1.4881168359941945,
      "grad_norm": 2.234375,
      "learning_rate": 0.00047827698798085015,
      "loss": 5.286,
      "mean_token_accuracy": 0.17973822355270386,
      "num_tokens": 34048008.0,
      "step": 16405
    },
    {
      "entropy": 5.65091724395752,
      "epoch": 1.4885703918722788,
      "grad_norm": 3.109375,
      "learning_rate": 0.0004782631151165262,
      "loss": 5.3232,
      "mean_token_accuracy": 0.17462074160575866,
      "num_tokens": 34058009.0,
      "step": 16410
    },
    {
      "entropy": 5.624415445327759,
      "epoch": 1.4890239477503628,
      "grad_norm": 2.765625,
      "learning_rate": 0.00047824923804866914,
      "loss": 5.2612,
      "mean_token_accuracy": 0.1797480031847954,
      "num_tokens": 34067625.0,
      "step": 16415
    },
    {
      "entropy": 5.484868764877319,
      "epoch": 1.4894775036284469,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00047823535677756607,
      "loss": 5.2853,
      "mean_token_accuracy": 0.17014513462781905,
      "num_tokens": 34077349.0,
      "step": 16420
    },
    {
      "entropy": 5.636375665664673,
      "epoch": 1.4899310595065312,
      "grad_norm": 3.3125,
      "learning_rate": 0.0004782214713035041,
      "loss": 5.4101,
      "mean_token_accuracy": 0.17359355092048645,
      "num_tokens": 34088636.0,
      "step": 16425
    },
    {
      "entropy": 5.621540069580078,
      "epoch": 1.4903846153846154,
      "grad_norm": 2.71875,
      "learning_rate": 0.00047820758162677026,
      "loss": 5.3214,
      "mean_token_accuracy": 0.17569074034690857,
      "num_tokens": 34098806.0,
      "step": 16430
    },
    {
      "entropy": 5.541275310516357,
      "epoch": 1.4908381712626997,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00047819368774765187,
      "loss": 5.3129,
      "mean_token_accuracy": 0.1821640908718109,
      "num_tokens": 34109713.0,
      "step": 16435
    },
    {
      "entropy": 5.498097467422485,
      "epoch": 1.4912917271407837,
      "grad_norm": 2.328125,
      "learning_rate": 0.00047817978966643626,
      "loss": 5.2795,
      "mean_token_accuracy": 0.1815725937485695,
      "num_tokens": 34119585.0,
      "step": 16440
    },
    {
      "entropy": 5.655697011947632,
      "epoch": 1.491745283018868,
      "grad_norm": 2.75,
      "learning_rate": 0.0004781658873834108,
      "loss": 5.3175,
      "mean_token_accuracy": 0.17075912952423095,
      "num_tokens": 34131245.0,
      "step": 16445
    },
    {
      "entropy": 5.602940273284912,
      "epoch": 1.492198838896952,
      "grad_norm": 2.171875,
      "learning_rate": 0.00047815198089886296,
      "loss": 5.2904,
      "mean_token_accuracy": 0.17867336869239808,
      "num_tokens": 34141229.0,
      "step": 16450
    },
    {
      "entropy": 5.5427995204925535,
      "epoch": 1.4926523947750363,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004781380702130805,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1748720094561577,
      "num_tokens": 34151781.0,
      "step": 16455
    },
    {
      "entropy": 5.576898097991943,
      "epoch": 1.4931059506531206,
      "grad_norm": 2.046875,
      "learning_rate": 0.00047812415532635087,
      "loss": 5.2823,
      "mean_token_accuracy": 0.17601683735847473,
      "num_tokens": 34162305.0,
      "step": 16460
    },
    {
      "entropy": 5.6581902503967285,
      "epoch": 1.4935595065312046,
      "grad_norm": 3.53125,
      "learning_rate": 0.000478110236238962,
      "loss": 5.3855,
      "mean_token_accuracy": 0.1733548030257225,
      "num_tokens": 34173613.0,
      "step": 16465
    },
    {
      "entropy": 5.573139190673828,
      "epoch": 1.4940130624092889,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00047809631295120164,
      "loss": 5.2777,
      "mean_token_accuracy": 0.17567175030708312,
      "num_tokens": 34183540.0,
      "step": 16470
    },
    {
      "entropy": 5.517584943771363,
      "epoch": 1.494466618287373,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047808238546335786,
      "loss": 5.3064,
      "mean_token_accuracy": 0.1748233824968338,
      "num_tokens": 34193292.0,
      "step": 16475
    },
    {
      "entropy": 5.626152753829956,
      "epoch": 1.4949201741654572,
      "grad_norm": 2.046875,
      "learning_rate": 0.00047806845377571856,
      "loss": 5.3561,
      "mean_token_accuracy": 0.17944617867469786,
      "num_tokens": 34202697.0,
      "step": 16480
    },
    {
      "entropy": 5.58451018333435,
      "epoch": 1.4953737300435415,
      "grad_norm": 2.4375,
      "learning_rate": 0.00047805451788857187,
      "loss": 5.2439,
      "mean_token_accuracy": 0.1796257123351097,
      "num_tokens": 34213396.0,
      "step": 16485
    },
    {
      "entropy": 5.6154790878295895,
      "epoch": 1.4958272859216255,
      "grad_norm": 1.90625,
      "learning_rate": 0.000478040577802206,
      "loss": 5.2518,
      "mean_token_accuracy": 0.17889929562807083,
      "num_tokens": 34223020.0,
      "step": 16490
    },
    {
      "entropy": 5.58239336013794,
      "epoch": 1.4962808417997098,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004780266335169093,
      "loss": 5.2909,
      "mean_token_accuracy": 0.18215290009975432,
      "num_tokens": 34233249.0,
      "step": 16495
    },
    {
      "entropy": 5.640660333633423,
      "epoch": 1.4967343976777938,
      "grad_norm": 2.53125,
      "learning_rate": 0.00047801268503297,
      "loss": 5.3395,
      "mean_token_accuracy": 0.17162637710571288,
      "num_tokens": 34244091.0,
      "step": 16500
    },
    {
      "entropy": 5.58487639427185,
      "epoch": 1.497187953555878,
      "grad_norm": 2.28125,
      "learning_rate": 0.00047799873235067667,
      "loss": 5.3264,
      "mean_token_accuracy": 0.1807827115058899,
      "num_tokens": 34255060.0,
      "step": 16505
    },
    {
      "entropy": 5.5405360698699955,
      "epoch": 1.4976415094339623,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004779847754703178,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1885210543870926,
      "num_tokens": 34264361.0,
      "step": 16510
    },
    {
      "entropy": 5.515546989440918,
      "epoch": 1.4980950653120464,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004779708143921821,
      "loss": 5.2877,
      "mean_token_accuracy": 0.17634256482124328,
      "num_tokens": 34274480.0,
      "step": 16515
    },
    {
      "entropy": 5.554407453536987,
      "epoch": 1.4985486211901307,
      "grad_norm": 2.6875,
      "learning_rate": 0.00047795684911655824,
      "loss": 5.3123,
      "mean_token_accuracy": 0.17488722801208495,
      "num_tokens": 34285951.0,
      "step": 16520
    },
    {
      "entropy": 5.583678245544434,
      "epoch": 1.4990021770682147,
      "grad_norm": 2.046875,
      "learning_rate": 0.000477942879643735,
      "loss": 5.3555,
      "mean_token_accuracy": 0.170409631729126,
      "num_tokens": 34296384.0,
      "step": 16525
    },
    {
      "entropy": 5.550891017913818,
      "epoch": 1.499455732946299,
      "grad_norm": 2.015625,
      "learning_rate": 0.00047792890597400125,
      "loss": 5.2763,
      "mean_token_accuracy": 0.18009277135133744,
      "num_tokens": 34306776.0,
      "step": 16530
    },
    {
      "entropy": 5.602018213272094,
      "epoch": 1.4999092888243832,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047791492810764604,
      "loss": 5.2945,
      "mean_token_accuracy": 0.1780122771859169,
      "num_tokens": 34317047.0,
      "step": 16535
    },
    {
      "entropy": 5.590870904922485,
      "epoch": 1.5003628447024675,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004779009460449584,
      "loss": 5.2613,
      "mean_token_accuracy": 0.17889285385608672,
      "num_tokens": 34326392.0,
      "step": 16540
    },
    {
      "entropy": 5.602251243591309,
      "epoch": 1.5008164005805515,
      "grad_norm": 2.359375,
      "learning_rate": 0.00047788695978622756,
      "loss": 5.3045,
      "mean_token_accuracy": 0.18073899447917938,
      "num_tokens": 34336580.0,
      "step": 16545
    },
    {
      "entropy": 5.501368427276612,
      "epoch": 1.5012699564586356,
      "grad_norm": 2.34375,
      "learning_rate": 0.00047787296933174265,
      "loss": 5.3342,
      "mean_token_accuracy": 0.17427235543727876,
      "num_tokens": 34346792.0,
      "step": 16550
    },
    {
      "entropy": 5.574243545532227,
      "epoch": 1.5017235123367199,
      "grad_norm": 1.765625,
      "learning_rate": 0.000477858974681793,
      "loss": 5.3302,
      "mean_token_accuracy": 0.17644186317920685,
      "num_tokens": 34356675.0,
      "step": 16555
    },
    {
      "entropy": 5.607922840118408,
      "epoch": 1.5021770682148041,
      "grad_norm": 2.0,
      "learning_rate": 0.0004778449758366681,
      "loss": 5.3096,
      "mean_token_accuracy": 0.1823266088962555,
      "num_tokens": 34366357.0,
      "step": 16560
    },
    {
      "entropy": 5.556922483444214,
      "epoch": 1.5026306240928884,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004778309727966574,
      "loss": 5.2998,
      "mean_token_accuracy": 0.18184888809919358,
      "num_tokens": 34377229.0,
      "step": 16565
    },
    {
      "entropy": 5.538240528106689,
      "epoch": 1.5030841799709724,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00047781696556205045,
      "loss": 5.2325,
      "mean_token_accuracy": 0.18401357531547546,
      "num_tokens": 34387873.0,
      "step": 16570
    },
    {
      "entropy": 5.6033775329589846,
      "epoch": 1.5035377358490565,
      "grad_norm": 2.75,
      "learning_rate": 0.00047780295413313696,
      "loss": 5.2549,
      "mean_token_accuracy": 0.1794172301888466,
      "num_tokens": 34398889.0,
      "step": 16575
    },
    {
      "entropy": 5.5521012306213375,
      "epoch": 1.5039912917271407,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047778893851020676,
      "loss": 5.2141,
      "mean_token_accuracy": 0.1822761982679367,
      "num_tokens": 34408508.0,
      "step": 16580
    },
    {
      "entropy": 5.530418968200683,
      "epoch": 1.504444847605225,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004777749186935496,
      "loss": 5.3692,
      "mean_token_accuracy": 0.17065185755491258,
      "num_tokens": 34418928.0,
      "step": 16585
    },
    {
      "entropy": 5.582806015014649,
      "epoch": 1.5048984034833093,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004777608946834555,
      "loss": 5.3654,
      "mean_token_accuracy": 0.17221508473157882,
      "num_tokens": 34429789.0,
      "step": 16590
    },
    {
      "entropy": 5.666883563995361,
      "epoch": 1.5053519593613933,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004777468664802144,
      "loss": 5.4349,
      "mean_token_accuracy": 0.16979475766420365,
      "num_tokens": 34440795.0,
      "step": 16595
    },
    {
      "entropy": 5.654188966751098,
      "epoch": 1.5058055152394774,
      "grad_norm": 2.5,
      "learning_rate": 0.00047773283408411636,
      "loss": 5.328,
      "mean_token_accuracy": 0.1737726464867592,
      "num_tokens": 34450772.0,
      "step": 16600
    },
    {
      "entropy": 5.575477361679077,
      "epoch": 1.5062590711175616,
      "grad_norm": 2.046875,
      "learning_rate": 0.00047771879749545175,
      "loss": 5.2768,
      "mean_token_accuracy": 0.1772533357143402,
      "num_tokens": 34461273.0,
      "step": 16605
    },
    {
      "entropy": 5.56493992805481,
      "epoch": 1.506712626995646,
      "grad_norm": 2.21875,
      "learning_rate": 0.00047770475671451073,
      "loss": 5.3393,
      "mean_token_accuracy": 0.17333926856517792,
      "num_tokens": 34470942.0,
      "step": 16610
    },
    {
      "entropy": 5.601726055145264,
      "epoch": 1.5071661828737302,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004776907117415837,
      "loss": 5.2672,
      "mean_token_accuracy": 0.17901651412248612,
      "num_tokens": 34481117.0,
      "step": 16615
    },
    {
      "entropy": 5.553108215332031,
      "epoch": 1.5076197387518142,
      "grad_norm": 2.046875,
      "learning_rate": 0.00047767666257696103,
      "loss": 5.243,
      "mean_token_accuracy": 0.1862640082836151,
      "num_tokens": 34490953.0,
      "step": 16620
    },
    {
      "entropy": 5.5408392429351805,
      "epoch": 1.5080732946298983,
      "grad_norm": 4.9375,
      "learning_rate": 0.00047766260922093337,
      "loss": 5.1904,
      "mean_token_accuracy": 0.18672532886266707,
      "num_tokens": 34502765.0,
      "step": 16625
    },
    {
      "entropy": 5.625203323364258,
      "epoch": 1.5085268505079825,
      "grad_norm": 1.75,
      "learning_rate": 0.0004776485516737913,
      "loss": 5.3938,
      "mean_token_accuracy": 0.17629127204418182,
      "num_tokens": 34514473.0,
      "step": 16630
    },
    {
      "entropy": 5.620902109146118,
      "epoch": 1.5089804063860668,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004776344899358256,
      "loss": 5.3892,
      "mean_token_accuracy": 0.17226543873548508,
      "num_tokens": 34524218.0,
      "step": 16635
    },
    {
      "entropy": 5.596171808242798,
      "epoch": 1.509433962264151,
      "grad_norm": 2.25,
      "learning_rate": 0.0004776204240073269,
      "loss": 5.2888,
      "mean_token_accuracy": 0.17800034135580062,
      "num_tokens": 34535697.0,
      "step": 16640
    },
    {
      "entropy": 5.5919922351837155,
      "epoch": 1.509887518142235,
      "grad_norm": 1.78125,
      "learning_rate": 0.00047760635388858624,
      "loss": 5.3183,
      "mean_token_accuracy": 0.1809437707066536,
      "num_tokens": 34545130.0,
      "step": 16645
    },
    {
      "entropy": 5.546190023422241,
      "epoch": 1.5103410740203191,
      "grad_norm": 2.640625,
      "learning_rate": 0.00047759227957989465,
      "loss": 5.2707,
      "mean_token_accuracy": 0.17763865739107132,
      "num_tokens": 34554974.0,
      "step": 16650
    },
    {
      "entropy": 5.6263288974761965,
      "epoch": 1.5107946298984034,
      "grad_norm": 2.140625,
      "learning_rate": 0.000477578201081543,
      "loss": 5.3108,
      "mean_token_accuracy": 0.17559502869844437,
      "num_tokens": 34565660.0,
      "step": 16655
    },
    {
      "entropy": 5.610887241363526,
      "epoch": 1.5112481857764877,
      "grad_norm": 1.75,
      "learning_rate": 0.0004775641183938225,
      "loss": 5.3297,
      "mean_token_accuracy": 0.17595831602811812,
      "num_tokens": 34576447.0,
      "step": 16660
    },
    {
      "entropy": 5.63410062789917,
      "epoch": 1.511701741654572,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047755003151702443,
      "loss": 5.4658,
      "mean_token_accuracy": 0.1745808869600296,
      "num_tokens": 34588333.0,
      "step": 16665
    },
    {
      "entropy": 5.544179439544678,
      "epoch": 1.512155297532656,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047753594045144015,
      "loss": 5.1948,
      "mean_token_accuracy": 0.1790759325027466,
      "num_tokens": 34598166.0,
      "step": 16670
    },
    {
      "entropy": 5.543941068649292,
      "epoch": 1.5126088534107403,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004775218451973609,
      "loss": 5.2438,
      "mean_token_accuracy": 0.176773039996624,
      "num_tokens": 34608615.0,
      "step": 16675
    },
    {
      "entropy": 5.5469011783599855,
      "epoch": 1.5130624092888243,
      "grad_norm": 3.078125,
      "learning_rate": 0.0004775077457550784,
      "loss": 5.294,
      "mean_token_accuracy": 0.17714096754789352,
      "num_tokens": 34618618.0,
      "step": 16680
    },
    {
      "entropy": 5.554601240158081,
      "epoch": 1.5135159651669086,
      "grad_norm": 1.84375,
      "learning_rate": 0.00047749364212488403,
      "loss": 5.3085,
      "mean_token_accuracy": 0.17146841138601304,
      "num_tokens": 34629412.0,
      "step": 16685
    },
    {
      "entropy": 5.625255870819092,
      "epoch": 1.5139695210449928,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004774795343070696,
      "loss": 5.3893,
      "mean_token_accuracy": 0.17256869822740556,
      "num_tokens": 34640487.0,
      "step": 16690
    },
    {
      "entropy": 5.608859443664551,
      "epoch": 1.5144230769230769,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004774654223019268,
      "loss": 5.3179,
      "mean_token_accuracy": 0.17825735062360765,
      "num_tokens": 34650126.0,
      "step": 16695
    },
    {
      "entropy": 5.506993341445923,
      "epoch": 1.5148766328011611,
      "grad_norm": 3.046875,
      "learning_rate": 0.0004774513061097474,
      "loss": 5.1655,
      "mean_token_accuracy": 0.18520084172487258,
      "num_tokens": 34659721.0,
      "step": 16700
    },
    {
      "entropy": 5.506360483169556,
      "epoch": 1.5153301886792452,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004774371857308235,
      "loss": 5.2796,
      "mean_token_accuracy": 0.18183324486017227,
      "num_tokens": 34670611.0,
      "step": 16705
    },
    {
      "entropy": 5.592843151092529,
      "epoch": 1.5157837445573294,
      "grad_norm": 2.328125,
      "learning_rate": 0.00047742306116544693,
      "loss": 5.2808,
      "mean_token_accuracy": 0.18022513538599014,
      "num_tokens": 34680173.0,
      "step": 16710
    },
    {
      "entropy": 5.578129291534424,
      "epoch": 1.5162373004354137,
      "grad_norm": 2.8125,
      "learning_rate": 0.00047740893241390994,
      "loss": 5.3306,
      "mean_token_accuracy": 0.17745720744132995,
      "num_tokens": 34690176.0,
      "step": 16715
    },
    {
      "entropy": 5.615120267868042,
      "epoch": 1.516690856313498,
      "grad_norm": 2.875,
      "learning_rate": 0.0004773947994765046,
      "loss": 5.2518,
      "mean_token_accuracy": 0.18413628488779069,
      "num_tokens": 34700481.0,
      "step": 16720
    },
    {
      "entropy": 5.533414506912232,
      "epoch": 1.517144412191582,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004773806623535232,
      "loss": 5.4327,
      "mean_token_accuracy": 0.16931308805942535,
      "num_tokens": 34710643.0,
      "step": 16725
    },
    {
      "entropy": 5.613033962249756,
      "epoch": 1.517597968069666,
      "grad_norm": 2.296875,
      "learning_rate": 0.00047736652104525816,
      "loss": 5.2806,
      "mean_token_accuracy": 0.17312693744897842,
      "num_tokens": 34720998.0,
      "step": 16730
    },
    {
      "entropy": 5.621651649475098,
      "epoch": 1.5180515239477503,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004773523755520019,
      "loss": 5.3491,
      "mean_token_accuracy": 0.17128076255321503,
      "num_tokens": 34731071.0,
      "step": 16735
    },
    {
      "entropy": 5.523481225967407,
      "epoch": 1.5185050798258346,
      "grad_norm": 1.9375,
      "learning_rate": 0.000477338225874047,
      "loss": 5.2485,
      "mean_token_accuracy": 0.18011671900749207,
      "num_tokens": 34741346.0,
      "step": 16740
    },
    {
      "entropy": 5.5440820217132565,
      "epoch": 1.5189586357039189,
      "grad_norm": 3.28125,
      "learning_rate": 0.00047732407201168595,
      "loss": 5.2611,
      "mean_token_accuracy": 0.18415879160165788,
      "num_tokens": 34751092.0,
      "step": 16745
    },
    {
      "entropy": 5.6409985542297365,
      "epoch": 1.519412191582003,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004773099139652116,
      "loss": 5.3761,
      "mean_token_accuracy": 0.1740630939602852,
      "num_tokens": 34761477.0,
      "step": 16750
    },
    {
      "entropy": 5.555561208724976,
      "epoch": 1.519865747460087,
      "grad_norm": 3.15625,
      "learning_rate": 0.0004772957517349167,
      "loss": 5.2831,
      "mean_token_accuracy": 0.1745086506009102,
      "num_tokens": 34772491.0,
      "step": 16755
    },
    {
      "entropy": 5.526907444000244,
      "epoch": 1.5203193033381712,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004772815853210939,
      "loss": 5.3315,
      "mean_token_accuracy": 0.17033641189336776,
      "num_tokens": 34783192.0,
      "step": 16760
    },
    {
      "entropy": 5.562852621078491,
      "epoch": 1.5207728592162555,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047726741472403655,
      "loss": 5.3082,
      "mean_token_accuracy": 0.17706797420978546,
      "num_tokens": 34794164.0,
      "step": 16765
    },
    {
      "entropy": 5.617201662063598,
      "epoch": 1.5212264150943398,
      "grad_norm": 2.734375,
      "learning_rate": 0.00047725323994403753,
      "loss": 5.3357,
      "mean_token_accuracy": 0.17154910415410995,
      "num_tokens": 34804390.0,
      "step": 16770
    },
    {
      "entropy": 5.684087181091309,
      "epoch": 1.5216799709724238,
      "grad_norm": 1.875,
      "learning_rate": 0.00047723906098138993,
      "loss": 5.4289,
      "mean_token_accuracy": 0.17009544670581817,
      "num_tokens": 34814751.0,
      "step": 16775
    },
    {
      "entropy": 5.569220066070557,
      "epoch": 1.5221335268505078,
      "grad_norm": 3.5,
      "learning_rate": 0.00047722487783638697,
      "loss": 5.2781,
      "mean_token_accuracy": 0.17727454602718354,
      "num_tokens": 34824977.0,
      "step": 16780
    },
    {
      "entropy": 5.546183729171753,
      "epoch": 1.5225870827285921,
      "grad_norm": 2.0625,
      "learning_rate": 0.000477210690509322,
      "loss": 5.3394,
      "mean_token_accuracy": 0.1772815078496933,
      "num_tokens": 34835540.0,
      "step": 16785
    },
    {
      "entropy": 5.587404680252075,
      "epoch": 1.5230406386066764,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00047719649900048843,
      "loss": 5.2986,
      "mean_token_accuracy": 0.175293929874897,
      "num_tokens": 34845233.0,
      "step": 16790
    },
    {
      "entropy": 5.701802206039429,
      "epoch": 1.5234941944847606,
      "grad_norm": 6.28125,
      "learning_rate": 0.00047718230331017975,
      "loss": 5.3962,
      "mean_token_accuracy": 0.171320341527462,
      "num_tokens": 34855756.0,
      "step": 16795
    },
    {
      "entropy": 5.628451108932495,
      "epoch": 1.5239477503628447,
      "grad_norm": 2.4375,
      "learning_rate": 0.00047716810343868956,
      "loss": 5.3301,
      "mean_token_accuracy": 0.16580800712108612,
      "num_tokens": 34867869.0,
      "step": 16800
    },
    {
      "entropy": 5.653856992721558,
      "epoch": 1.5244013062409287,
      "grad_norm": 1.828125,
      "learning_rate": 0.00047715389938631143,
      "loss": 5.3882,
      "mean_token_accuracy": 0.17076897621154785,
      "num_tokens": 34878625.0,
      "step": 16805
    },
    {
      "entropy": 5.616347408294677,
      "epoch": 1.524854862119013,
      "grad_norm": 4.78125,
      "learning_rate": 0.00047713969115333916,
      "loss": 5.3545,
      "mean_token_accuracy": 0.1697998657822609,
      "num_tokens": 34889408.0,
      "step": 16810
    },
    {
      "entropy": 5.580618810653687,
      "epoch": 1.5253084179970973,
      "grad_norm": 3.21875,
      "learning_rate": 0.0004771254787400666,
      "loss": 5.3732,
      "mean_token_accuracy": 0.1722514197230339,
      "num_tokens": 34899602.0,
      "step": 16815
    },
    {
      "entropy": 5.568227958679199,
      "epoch": 1.5257619738751815,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004771112621467876,
      "loss": 5.2232,
      "mean_token_accuracy": 0.18233878016471863,
      "num_tokens": 34910215.0,
      "step": 16820
    },
    {
      "entropy": 5.531526851654053,
      "epoch": 1.5262155297532656,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004770970413737961,
      "loss": 5.3007,
      "mean_token_accuracy": 0.17985867857933044,
      "num_tokens": 34919641.0,
      "step": 16825
    },
    {
      "entropy": 5.580641222000122,
      "epoch": 1.5266690856313496,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004770828164213865,
      "loss": 5.2023,
      "mean_token_accuracy": 0.18557829558849334,
      "num_tokens": 34930243.0,
      "step": 16830
    },
    {
      "entropy": 5.59973292350769,
      "epoch": 1.5271226415094339,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004770685872898525,
      "loss": 5.3673,
      "mean_token_accuracy": 0.17803116887807846,
      "num_tokens": 34941057.0,
      "step": 16835
    },
    {
      "entropy": 5.566740322113037,
      "epoch": 1.5275761973875182,
      "grad_norm": 3.28125,
      "learning_rate": 0.00047705435397948876,
      "loss": 5.2151,
      "mean_token_accuracy": 0.1883818805217743,
      "num_tokens": 34951150.0,
      "step": 16840
    },
    {
      "entropy": 5.564435195922852,
      "epoch": 1.5280297532656024,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004770401164905895,
      "loss": 5.3213,
      "mean_token_accuracy": 0.17725121974945068,
      "num_tokens": 34961648.0,
      "step": 16845
    },
    {
      "entropy": 5.581844282150269,
      "epoch": 1.5284833091436865,
      "grad_norm": 1.796875,
      "learning_rate": 0.00047702587482344915,
      "loss": 5.327,
      "mean_token_accuracy": 0.17188748121261596,
      "num_tokens": 34971573.0,
      "step": 16850
    },
    {
      "entropy": 5.64786171913147,
      "epoch": 1.5289368650217707,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004770116289783622,
      "loss": 5.3231,
      "mean_token_accuracy": 0.17190929353237153,
      "num_tokens": 34981984.0,
      "step": 16855
    },
    {
      "entropy": 5.580846977233887,
      "epoch": 1.5293904208998548,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004769973789556233,
      "loss": 5.3202,
      "mean_token_accuracy": 0.1771014928817749,
      "num_tokens": 34990861.0,
      "step": 16860
    },
    {
      "entropy": 5.561270618438721,
      "epoch": 1.529843976777939,
      "grad_norm": 2.96875,
      "learning_rate": 0.00047698312475552714,
      "loss": 5.3304,
      "mean_token_accuracy": 0.17301484793424607,
      "num_tokens": 35001504.0,
      "step": 16865
    },
    {
      "entropy": 5.602186632156372,
      "epoch": 1.5302975326560233,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004769688663783684,
      "loss": 5.2379,
      "mean_token_accuracy": 0.18345279097557068,
      "num_tokens": 35012425.0,
      "step": 16870
    },
    {
      "entropy": 5.590351009368897,
      "epoch": 1.5307510885341074,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004769546038244421,
      "loss": 5.2534,
      "mean_token_accuracy": 0.18211067467927933,
      "num_tokens": 35022854.0,
      "step": 16875
    },
    {
      "entropy": 5.582770156860351,
      "epoch": 1.5312046444121916,
      "grad_norm": 2.09375,
      "learning_rate": 0.00047694033709404313,
      "loss": 5.3516,
      "mean_token_accuracy": 0.17412721812725068,
      "num_tokens": 35032694.0,
      "step": 16880
    },
    {
      "entropy": 5.568556594848633,
      "epoch": 1.5316582002902757,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00047692606618746655,
      "loss": 5.2811,
      "mean_token_accuracy": 0.18107204437255858,
      "num_tokens": 35042217.0,
      "step": 16885
    },
    {
      "entropy": 5.57844033241272,
      "epoch": 1.53211175616836,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00047691179110500735,
      "loss": 5.362,
      "mean_token_accuracy": 0.1685964971780777,
      "num_tokens": 35052831.0,
      "step": 16890
    },
    {
      "entropy": 5.621140861511231,
      "epoch": 1.5325653120464442,
      "grad_norm": 2.296875,
      "learning_rate": 0.00047689751184696097,
      "loss": 5.3793,
      "mean_token_accuracy": 0.17402787059545516,
      "num_tokens": 35061955.0,
      "step": 16895
    },
    {
      "entropy": 5.6145707130432125,
      "epoch": 1.5330188679245285,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004768832284136226,
      "loss": 5.2834,
      "mean_token_accuracy": 0.17781832665205002,
      "num_tokens": 35072526.0,
      "step": 16900
    },
    {
      "entropy": 5.554147148132325,
      "epoch": 1.5334724238026125,
      "grad_norm": 1.75,
      "learning_rate": 0.0004768689408052876,
      "loss": 5.2487,
      "mean_token_accuracy": 0.18554634004831314,
      "num_tokens": 35082415.0,
      "step": 16905
    },
    {
      "entropy": 5.499531841278076,
      "epoch": 1.5339259796806965,
      "grad_norm": 2.765625,
      "learning_rate": 0.00047685464902225144,
      "loss": 5.2629,
      "mean_token_accuracy": 0.18011668771505357,
      "num_tokens": 35092805.0,
      "step": 16910
    },
    {
      "entropy": 5.5594557762146,
      "epoch": 1.5343795355587808,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004768403530648097,
      "loss": 5.3083,
      "mean_token_accuracy": 0.18167798221111298,
      "num_tokens": 35102963.0,
      "step": 16915
    },
    {
      "entropy": 5.6697916984558105,
      "epoch": 1.534833091436865,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047682605293325806,
      "loss": 5.3201,
      "mean_token_accuracy": 0.17642893642187119,
      "num_tokens": 35114140.0,
      "step": 16920
    },
    {
      "entropy": 5.64707236289978,
      "epoch": 1.5352866473149493,
      "grad_norm": 2.203125,
      "learning_rate": 0.00047681174862789215,
      "loss": 5.3939,
      "mean_token_accuracy": 0.17817157804965972,
      "num_tokens": 35125314.0,
      "step": 16925
    },
    {
      "entropy": 5.559857320785523,
      "epoch": 1.5357402031930334,
      "grad_norm": 2.15625,
      "learning_rate": 0.00047679744014900795,
      "loss": 5.2884,
      "mean_token_accuracy": 0.17577572166919708,
      "num_tokens": 35134934.0,
      "step": 16930
    },
    {
      "entropy": 5.547251462936401,
      "epoch": 1.5361937590711174,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004767831274969011,
      "loss": 5.2498,
      "mean_token_accuracy": 0.18336026966571808,
      "num_tokens": 35145341.0,
      "step": 16935
    },
    {
      "entropy": 5.522019290924073,
      "epoch": 1.5366473149492017,
      "grad_norm": 2.015625,
      "learning_rate": 0.00047676881067186793,
      "loss": 5.226,
      "mean_token_accuracy": 0.17458679229021073,
      "num_tokens": 35156046.0,
      "step": 16940
    },
    {
      "entropy": 5.601333141326904,
      "epoch": 1.537100870827286,
      "grad_norm": 1.96875,
      "learning_rate": 0.00047675448967420426,
      "loss": 5.3841,
      "mean_token_accuracy": 0.17411985397338867,
      "num_tokens": 35168232.0,
      "step": 16945
    },
    {
      "entropy": 5.671712207794189,
      "epoch": 1.5375544267053702,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004767401645042064,
      "loss": 5.3395,
      "mean_token_accuracy": 0.17694685906171798,
      "num_tokens": 35178585.0,
      "step": 16950
    },
    {
      "entropy": 5.574842834472657,
      "epoch": 1.5380079825834543,
      "grad_norm": 2.046875,
      "learning_rate": 0.00047672583516217043,
      "loss": 5.2223,
      "mean_token_accuracy": 0.17769193202257155,
      "num_tokens": 35189417.0,
      "step": 16955
    },
    {
      "entropy": 5.557365894317627,
      "epoch": 1.5384615384615383,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004767115016483928,
      "loss": 5.2567,
      "mean_token_accuracy": 0.1797560289502144,
      "num_tokens": 35200167.0,
      "step": 16960
    },
    {
      "entropy": 5.53791241645813,
      "epoch": 1.5389150943396226,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004766971639631699,
      "loss": 5.272,
      "mean_token_accuracy": 0.18110861927270888,
      "num_tokens": 35210711.0,
      "step": 16965
    },
    {
      "entropy": 5.538530111312866,
      "epoch": 1.5393686502177069,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047668282210679825,
      "loss": 5.2978,
      "mean_token_accuracy": 0.17265373170375825,
      "num_tokens": 35220866.0,
      "step": 16970
    },
    {
      "entropy": 5.528390216827392,
      "epoch": 1.5398222060957911,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004766684760795745,
      "loss": 5.2583,
      "mean_token_accuracy": 0.18271117210388182,
      "num_tokens": 35231478.0,
      "step": 16975
    },
    {
      "entropy": 5.5061047077178955,
      "epoch": 1.5402757619738752,
      "grad_norm": 1.703125,
      "learning_rate": 0.00047665412588179524,
      "loss": 5.2052,
      "mean_token_accuracy": 0.18749460577964783,
      "num_tokens": 35242233.0,
      "step": 16980
    },
    {
      "entropy": 5.520451068878174,
      "epoch": 1.5407293178519592,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004766397715137573,
      "loss": 5.2708,
      "mean_token_accuracy": 0.17564469277858735,
      "num_tokens": 35253097.0,
      "step": 16985
    },
    {
      "entropy": 5.666391658782959,
      "epoch": 1.5411828737300435,
      "grad_norm": 2.09375,
      "learning_rate": 0.00047662541297575743,
      "loss": 5.4189,
      "mean_token_accuracy": 0.17499935775995254,
      "num_tokens": 35264258.0,
      "step": 16990
    },
    {
      "entropy": 5.691465854644775,
      "epoch": 1.5416364296081277,
      "grad_norm": 2.078125,
      "learning_rate": 0.00047661105026809274,
      "loss": 5.3943,
      "mean_token_accuracy": 0.16887064427137374,
      "num_tokens": 35273501.0,
      "step": 16995
    },
    {
      "entropy": 5.5808107376098635,
      "epoch": 1.542089985486212,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004765966833910601,
      "loss": 5.3423,
      "mean_token_accuracy": 0.17378796488046647,
      "num_tokens": 35284125.0,
      "step": 17000
    },
    {
      "entropy": 5.630951166152954,
      "epoch": 1.542543541364296,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004765823123449567,
      "loss": 5.3427,
      "mean_token_accuracy": 0.1758963629603386,
      "num_tokens": 35296039.0,
      "step": 17005
    },
    {
      "entropy": 5.628085708618164,
      "epoch": 1.54299709724238,
      "grad_norm": 2.0625,
      "learning_rate": 0.00047656793713007965,
      "loss": 5.3045,
      "mean_token_accuracy": 0.17300825119018554,
      "num_tokens": 35306178.0,
      "step": 17010
    },
    {
      "entropy": 5.5611155986785885,
      "epoch": 1.5434506531204644,
      "grad_norm": 2.546875,
      "learning_rate": 0.00047655355774672636,
      "loss": 5.3179,
      "mean_token_accuracy": 0.17878797948360442,
      "num_tokens": 35316445.0,
      "step": 17015
    },
    {
      "entropy": 5.603832817077636,
      "epoch": 1.5439042089985486,
      "grad_norm": 2.25,
      "learning_rate": 0.00047653917419519413,
      "loss": 5.3145,
      "mean_token_accuracy": 0.17284477651119232,
      "num_tokens": 35327274.0,
      "step": 17020
    },
    {
      "entropy": 5.590578985214234,
      "epoch": 1.544357764876633,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004765247864757804,
      "loss": 5.3056,
      "mean_token_accuracy": 0.17986350059509276,
      "num_tokens": 35339007.0,
      "step": 17025
    },
    {
      "entropy": 5.634578847885132,
      "epoch": 1.544811320754717,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004765103945887827,
      "loss": 5.4072,
      "mean_token_accuracy": 0.16666809171438218,
      "num_tokens": 35349303.0,
      "step": 17030
    },
    {
      "entropy": 5.6126580238342285,
      "epoch": 1.5452648766328012,
      "grad_norm": 2.125,
      "learning_rate": 0.0004764959985344987,
      "loss": 5.2363,
      "mean_token_accuracy": 0.17778971791267395,
      "num_tokens": 35359556.0,
      "step": 17035
    },
    {
      "entropy": 5.5355448722839355,
      "epoch": 1.5457184325108853,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004764815983132261,
      "loss": 5.2575,
      "mean_token_accuracy": 0.18408314734697342,
      "num_tokens": 35370043.0,
      "step": 17040
    },
    {
      "entropy": 5.5542285442352295,
      "epoch": 1.5461719883889695,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004764671939252628,
      "loss": 5.2558,
      "mean_token_accuracy": 0.1832486718893051,
      "num_tokens": 35380031.0,
      "step": 17045
    },
    {
      "entropy": 5.588816547393799,
      "epoch": 1.5466255442670538,
      "grad_norm": 1.875,
      "learning_rate": 0.0004764527853709064,
      "loss": 5.3541,
      "mean_token_accuracy": 0.17472671270370482,
      "num_tokens": 35390481.0,
      "step": 17050
    },
    {
      "entropy": 5.570986318588257,
      "epoch": 1.5470791001451378,
      "grad_norm": 4.40625,
      "learning_rate": 0.0004764383726504552,
      "loss": 5.3398,
      "mean_token_accuracy": 0.17840237468481063,
      "num_tokens": 35401756.0,
      "step": 17055
    },
    {
      "entropy": 5.613854169845581,
      "epoch": 1.547532656023222,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004764239557642071,
      "loss": 5.2371,
      "mean_token_accuracy": 0.1823093980550766,
      "num_tokens": 35411045.0,
      "step": 17060
    },
    {
      "entropy": 5.605848693847657,
      "epoch": 1.5479862119013061,
      "grad_norm": 2.25,
      "learning_rate": 0.0004764095347124603,
      "loss": 5.2905,
      "mean_token_accuracy": 0.17576333582401277,
      "num_tokens": 35421256.0,
      "step": 17065
    },
    {
      "entropy": 5.597743940353394,
      "epoch": 1.5484397677793904,
      "grad_norm": 4.3125,
      "learning_rate": 0.0004763951094955129,
      "loss": 5.3175,
      "mean_token_accuracy": 0.179447965323925,
      "num_tokens": 35430798.0,
      "step": 17070
    },
    {
      "entropy": 5.554385614395142,
      "epoch": 1.5488933236574747,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004763806801136634,
      "loss": 5.2534,
      "mean_token_accuracy": 0.18434806764125825,
      "num_tokens": 35440459.0,
      "step": 17075
    },
    {
      "entropy": 5.549894142150879,
      "epoch": 1.549346879535559,
      "grad_norm": 3.03125,
      "learning_rate": 0.0004763662465672101,
      "loss": 5.3106,
      "mean_token_accuracy": 0.17880339175462723,
      "num_tokens": 35451586.0,
      "step": 17080
    },
    {
      "entropy": 5.563257074356079,
      "epoch": 1.549800435413643,
      "grad_norm": 2.3125,
      "learning_rate": 0.00047635180885645146,
      "loss": 5.3352,
      "mean_token_accuracy": 0.17747626900672914,
      "num_tokens": 35462197.0,
      "step": 17085
    },
    {
      "entropy": 5.6337559700012205,
      "epoch": 1.550253991291727,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004763373669816862,
      "loss": 5.2451,
      "mean_token_accuracy": 0.18264649212360382,
      "num_tokens": 35471737.0,
      "step": 17090
    },
    {
      "entropy": 5.647986507415771,
      "epoch": 1.5507075471698113,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004763229209432128,
      "loss": 5.3353,
      "mean_token_accuracy": 0.17153648287057877,
      "num_tokens": 35482016.0,
      "step": 17095
    },
    {
      "entropy": 5.62834792137146,
      "epoch": 1.5511611030478956,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004763084707413301,
      "loss": 5.4019,
      "mean_token_accuracy": 0.16543370336294175,
      "num_tokens": 35492622.0,
      "step": 17100
    },
    {
      "entropy": 5.677309513092041,
      "epoch": 1.5516146589259798,
      "grad_norm": 2.46875,
      "learning_rate": 0.00047629401637633696,
      "loss": 5.4614,
      "mean_token_accuracy": 0.17138228416442872,
      "num_tokens": 35504468.0,
      "step": 17105
    },
    {
      "entropy": 5.58835186958313,
      "epoch": 1.5520682148040639,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00047627955784853226,
      "loss": 5.2715,
      "mean_token_accuracy": 0.1747351497411728,
      "num_tokens": 35514272.0,
      "step": 17110
    },
    {
      "entropy": 5.584002685546875,
      "epoch": 1.552521770682148,
      "grad_norm": 2.25,
      "learning_rate": 0.000476265095158215,
      "loss": 5.2727,
      "mean_token_accuracy": 0.17862669080495835,
      "num_tokens": 35524110.0,
      "step": 17115
    },
    {
      "entropy": 5.568962144851684,
      "epoch": 1.5529753265602322,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004762506283056843,
      "loss": 5.2178,
      "mean_token_accuracy": 0.18359458297491074,
      "num_tokens": 35535504.0,
      "step": 17120
    },
    {
      "entropy": 5.681602811813354,
      "epoch": 1.5534288824383164,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004762361572912393,
      "loss": 5.3915,
      "mean_token_accuracy": 0.1724989205598831,
      "num_tokens": 35544885.0,
      "step": 17125
    },
    {
      "entropy": 5.617863512039184,
      "epoch": 1.5538824383164007,
      "grad_norm": 3.09375,
      "learning_rate": 0.00047622168211517926,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1779982030391693,
      "num_tokens": 35555485.0,
      "step": 17130
    },
    {
      "entropy": 5.504045915603638,
      "epoch": 1.5543359941944848,
      "grad_norm": 2.609375,
      "learning_rate": 0.00047620720277780353,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18138349801301956,
      "num_tokens": 35567197.0,
      "step": 17135
    },
    {
      "entropy": 5.61895523071289,
      "epoch": 1.5547895500725688,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047619271927941165,
      "loss": 5.2464,
      "mean_token_accuracy": 0.18175357580184937,
      "num_tokens": 35577576.0,
      "step": 17140
    },
    {
      "entropy": 5.592403268814087,
      "epoch": 1.555243105950653,
      "grad_norm": 2.390625,
      "learning_rate": 0.000476178231620303,
      "loss": 5.3149,
      "mean_token_accuracy": 0.1787196397781372,
      "num_tokens": 35588012.0,
      "step": 17145
    },
    {
      "entropy": 5.527973175048828,
      "epoch": 1.5556966618287373,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004761637398007774,
      "loss": 5.3134,
      "mean_token_accuracy": 0.17254016548395157,
      "num_tokens": 35598076.0,
      "step": 17150
    },
    {
      "entropy": 5.468479108810425,
      "epoch": 1.5561502177068216,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004761492438211342,
      "loss": 5.1114,
      "mean_token_accuracy": 0.18612807393074035,
      "num_tokens": 35607170.0,
      "step": 17155
    },
    {
      "entropy": 5.554360485076904,
      "epoch": 1.5566037735849056,
      "grad_norm": 2.0625,
      "learning_rate": 0.00047613474368167355,
      "loss": 5.2596,
      "mean_token_accuracy": 0.17387451678514482,
      "num_tokens": 35617865.0,
      "step": 17160
    },
    {
      "entropy": 5.577972364425659,
      "epoch": 1.5570573294629897,
      "grad_norm": 1.90625,
      "learning_rate": 0.000476120239382695,
      "loss": 5.3282,
      "mean_token_accuracy": 0.17546191215515136,
      "num_tokens": 35629329.0,
      "step": 17165
    },
    {
      "entropy": 5.560886716842651,
      "epoch": 1.557510885341074,
      "grad_norm": 4.15625,
      "learning_rate": 0.00047610573092449877,
      "loss": 5.3166,
      "mean_token_accuracy": 0.17200313806533812,
      "num_tokens": 35640044.0,
      "step": 17170
    },
    {
      "entropy": 5.594679880142212,
      "epoch": 1.5579644412191582,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004760912183073847,
      "loss": 5.3574,
      "mean_token_accuracy": 0.17837274372577666,
      "num_tokens": 35650435.0,
      "step": 17175
    },
    {
      "entropy": 5.593170928955078,
      "epoch": 1.5584179970972425,
      "grad_norm": 3.375,
      "learning_rate": 0.000476076701531653,
      "loss": 5.2432,
      "mean_token_accuracy": 0.17320944517850875,
      "num_tokens": 35659613.0,
      "step": 17180
    },
    {
      "entropy": 5.55014271736145,
      "epoch": 1.5588715529753265,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004760621805976039,
      "loss": 5.2971,
      "mean_token_accuracy": 0.17954155951738357,
      "num_tokens": 35670987.0,
      "step": 17185
    },
    {
      "entropy": 5.536005926132202,
      "epoch": 1.5593251088534106,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00047604765550553766,
      "loss": 5.2769,
      "mean_token_accuracy": 0.1802051156759262,
      "num_tokens": 35682505.0,
      "step": 17190
    },
    {
      "entropy": 5.5957588195800785,
      "epoch": 1.5597786647314948,
      "grad_norm": 2.0,
      "learning_rate": 0.00047603312625575466,
      "loss": 5.2413,
      "mean_token_accuracy": 0.18120294213294982,
      "num_tokens": 35693056.0,
      "step": 17195
    },
    {
      "entropy": 5.565271949768066,
      "epoch": 1.5602322206095791,
      "grad_norm": 2.1875,
      "learning_rate": 0.00047601859284855544,
      "loss": 5.2175,
      "mean_token_accuracy": 0.1850488379597664,
      "num_tokens": 35703659.0,
      "step": 17200
    },
    {
      "entropy": 5.509706449508667,
      "epoch": 1.5606857764876634,
      "grad_norm": 2.5,
      "learning_rate": 0.0004760040552842403,
      "loss": 5.2884,
      "mean_token_accuracy": 0.17964380979537964,
      "num_tokens": 35714367.0,
      "step": 17205
    },
    {
      "entropy": 5.532325649261475,
      "epoch": 1.5611393323657474,
      "grad_norm": 2.15625,
      "learning_rate": 0.00047598951356311035,
      "loss": 5.2359,
      "mean_token_accuracy": 0.17535126358270645,
      "num_tokens": 35724773.0,
      "step": 17210
    },
    {
      "entropy": 5.620828866958618,
      "epoch": 1.5615928882438317,
      "grad_norm": 2.5,
      "learning_rate": 0.0004759749676854658,
      "loss": 5.3944,
      "mean_token_accuracy": 0.17658211141824723,
      "num_tokens": 35735622.0,
      "step": 17215
    },
    {
      "entropy": 5.581115961074829,
      "epoch": 1.5620464441219157,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004759604176516079,
      "loss": 5.2613,
      "mean_token_accuracy": 0.1816563919186592,
      "num_tokens": 35744967.0,
      "step": 17220
    },
    {
      "entropy": 5.601293611526489,
      "epoch": 1.5625,
      "grad_norm": 2.6875,
      "learning_rate": 0.0004759458634618372,
      "loss": 5.3021,
      "mean_token_accuracy": 0.17240564227104188,
      "num_tokens": 35755647.0,
      "step": 17225
    },
    {
      "entropy": 5.60336537361145,
      "epoch": 1.5629535558780843,
      "grad_norm": 2.25,
      "learning_rate": 0.0004759313051164549,
      "loss": 5.3285,
      "mean_token_accuracy": 0.17064625173807144,
      "num_tokens": 35765595.0,
      "step": 17230
    },
    {
      "entropy": 5.625713872909546,
      "epoch": 1.5634071117561683,
      "grad_norm": 1.921875,
      "learning_rate": 0.000475916742615762,
      "loss": 5.2853,
      "mean_token_accuracy": 0.18276906609535218,
      "num_tokens": 35775804.0,
      "step": 17235
    },
    {
      "entropy": 5.52908730506897,
      "epoch": 1.5638606676342526,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004759021759600597,
      "loss": 5.2186,
      "mean_token_accuracy": 0.18050986230373384,
      "num_tokens": 35786227.0,
      "step": 17240
    },
    {
      "entropy": 5.604998302459717,
      "epoch": 1.5643142235123366,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004758876051496492,
      "loss": 5.361,
      "mean_token_accuracy": 0.17663424760103225,
      "num_tokens": 35797511.0,
      "step": 17245
    },
    {
      "entropy": 5.566086053848267,
      "epoch": 1.5647677793904209,
      "grad_norm": 2.125,
      "learning_rate": 0.00047587303018483177,
      "loss": 5.302,
      "mean_token_accuracy": 0.17646126598119735,
      "num_tokens": 35807248.0,
      "step": 17250
    },
    {
      "entropy": 5.596436214447022,
      "epoch": 1.5652213352685052,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047585845106590897,
      "loss": 5.3078,
      "mean_token_accuracy": 0.1742606595158577,
      "num_tokens": 35818164.0,
      "step": 17255
    },
    {
      "entropy": 5.593039321899414,
      "epoch": 1.5656748911465894,
      "grad_norm": 1.765625,
      "learning_rate": 0.00047584386779318214,
      "loss": 5.2675,
      "mean_token_accuracy": 0.1831199660897255,
      "num_tokens": 35828746.0,
      "step": 17260
    },
    {
      "entropy": 5.593298482894897,
      "epoch": 1.5661284470246735,
      "grad_norm": 1.640625,
      "learning_rate": 0.000475829280366953,
      "loss": 5.3289,
      "mean_token_accuracy": 0.16948552280664445,
      "num_tokens": 35839325.0,
      "step": 17265
    },
    {
      "entropy": 5.5756721019744875,
      "epoch": 1.5665820029027575,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047581468878752313,
      "loss": 5.2886,
      "mean_token_accuracy": 0.18293060213327408,
      "num_tokens": 35849406.0,
      "step": 17270
    },
    {
      "entropy": 5.516935157775879,
      "epoch": 1.5670355587808418,
      "grad_norm": 1.890625,
      "learning_rate": 0.00047580009305519427,
      "loss": 5.2053,
      "mean_token_accuracy": 0.17819298654794694,
      "num_tokens": 35858715.0,
      "step": 17275
    },
    {
      "entropy": 5.541009855270386,
      "epoch": 1.567489114658926,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004757854931702684,
      "loss": 5.3025,
      "mean_token_accuracy": 0.17792823910713196,
      "num_tokens": 35869063.0,
      "step": 17280
    },
    {
      "entropy": 5.5267918586730955,
      "epoch": 1.5679426705370103,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047577088913304733,
      "loss": 5.1724,
      "mean_token_accuracy": 0.18229275494813918,
      "num_tokens": 35878428.0,
      "step": 17285
    },
    {
      "entropy": 5.623327541351318,
      "epoch": 1.5683962264150944,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00047575628094383317,
      "loss": 5.3094,
      "mean_token_accuracy": 0.17703061550855637,
      "num_tokens": 35888801.0,
      "step": 17290
    },
    {
      "entropy": 5.522741317749023,
      "epoch": 1.5688497822931784,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004757416686029279,
      "loss": 5.3066,
      "mean_token_accuracy": 0.17580777704715728,
      "num_tokens": 35899381.0,
      "step": 17295
    },
    {
      "entropy": 5.602077341079712,
      "epoch": 1.5693033381712627,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004757270521106338,
      "loss": 5.3393,
      "mean_token_accuracy": 0.17475106418132783,
      "num_tokens": 35909546.0,
      "step": 17300
    },
    {
      "entropy": 5.620920038223266,
      "epoch": 1.569756894049347,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004757124314672531,
      "loss": 5.3115,
      "mean_token_accuracy": 0.18136712163686752,
      "num_tokens": 35919301.0,
      "step": 17305
    },
    {
      "entropy": 5.69102635383606,
      "epoch": 1.5702104499274312,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047569780667308816,
      "loss": 5.3695,
      "mean_token_accuracy": 0.17574509978294373,
      "num_tokens": 35928603.0,
      "step": 17310
    },
    {
      "entropy": 5.594991970062256,
      "epoch": 1.5706640058055152,
      "grad_norm": 1.953125,
      "learning_rate": 0.00047568317772844147,
      "loss": 5.3171,
      "mean_token_accuracy": 0.1822368785738945,
      "num_tokens": 35939005.0,
      "step": 17315
    },
    {
      "entropy": 5.609203100204468,
      "epoch": 1.5711175616835993,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047566854463361544,
      "loss": 5.2874,
      "mean_token_accuracy": 0.18577074110507966,
      "num_tokens": 35948471.0,
      "step": 17320
    },
    {
      "entropy": 5.551702642440796,
      "epoch": 1.5715711175616836,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004756539073889128,
      "loss": 5.3309,
      "mean_token_accuracy": 0.176429383456707,
      "num_tokens": 35958488.0,
      "step": 17325
    },
    {
      "entropy": 5.589171600341797,
      "epoch": 1.5720246734397678,
      "grad_norm": 2.046875,
      "learning_rate": 0.00047563926599463623,
      "loss": 5.3759,
      "mean_token_accuracy": 0.17729507237672806,
      "num_tokens": 35968291.0,
      "step": 17330
    },
    {
      "entropy": 5.6353535652160645,
      "epoch": 1.572478229317852,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004756246204510886,
      "loss": 5.2955,
      "mean_token_accuracy": 0.1771715834736824,
      "num_tokens": 35978565.0,
      "step": 17335
    },
    {
      "entropy": 5.591199827194214,
      "epoch": 1.5729317851959361,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047560997075857253,
      "loss": 5.2565,
      "mean_token_accuracy": 0.1825004905462265,
      "num_tokens": 35987483.0,
      "step": 17340
    },
    {
      "entropy": 5.575090217590332,
      "epoch": 1.5733853410740202,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004755953169173913,
      "loss": 5.3181,
      "mean_token_accuracy": 0.18051265180110931,
      "num_tokens": 35998355.0,
      "step": 17345
    },
    {
      "entropy": 5.531962537765503,
      "epoch": 1.5738388969521044,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004755806589278477,
      "loss": 5.2813,
      "mean_token_accuracy": 0.17916443645954133,
      "num_tokens": 36008365.0,
      "step": 17350
    },
    {
      "entropy": 5.5999908447265625,
      "epoch": 1.5742924528301887,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00047556599679024497,
      "loss": 5.3596,
      "mean_token_accuracy": 0.17237817049026488,
      "num_tokens": 36018880.0,
      "step": 17355
    },
    {
      "entropy": 5.611241340637207,
      "epoch": 1.574746008708273,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00047555133050488637,
      "loss": 5.3667,
      "mean_token_accuracy": 0.17026831060647965,
      "num_tokens": 36029335.0,
      "step": 17360
    },
    {
      "entropy": 5.546616649627685,
      "epoch": 1.575199564586357,
      "grad_norm": 2.546875,
      "learning_rate": 0.00047553666007207516,
      "loss": 5.2698,
      "mean_token_accuracy": 0.18537819534540176,
      "num_tokens": 36039774.0,
      "step": 17365
    },
    {
      "entropy": 5.628086233139038,
      "epoch": 1.575653120464441,
      "grad_norm": 1.984375,
      "learning_rate": 0.00047552198549211463,
      "loss": 5.3945,
      "mean_token_accuracy": 0.17898861020803453,
      "num_tokens": 36049864.0,
      "step": 17370
    },
    {
      "entropy": 5.5565690994262695,
      "epoch": 1.5761066763425253,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004755073067653084,
      "loss": 5.3135,
      "mean_token_accuracy": 0.1732003539800644,
      "num_tokens": 36059224.0,
      "step": 17375
    },
    {
      "entropy": 5.469857931137085,
      "epoch": 1.5765602322206096,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004754926238919599,
      "loss": 5.2459,
      "mean_token_accuracy": 0.18510266691446303,
      "num_tokens": 36069243.0,
      "step": 17380
    },
    {
      "entropy": 5.633704423904419,
      "epoch": 1.5770137880986939,
      "grad_norm": 2.71875,
      "learning_rate": 0.0004754779368723729,
      "loss": 5.2884,
      "mean_token_accuracy": 0.17391278743743896,
      "num_tokens": 36079833.0,
      "step": 17385
    },
    {
      "entropy": 5.6392497539520265,
      "epoch": 1.577467343976778,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004754632457068511,
      "loss": 5.2979,
      "mean_token_accuracy": 0.17892367243766785,
      "num_tokens": 36090673.0,
      "step": 17390
    },
    {
      "entropy": 5.573913955688477,
      "epoch": 1.5779208998548622,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004754485503956982,
      "loss": 5.3217,
      "mean_token_accuracy": 0.1820672109723091,
      "num_tokens": 36101719.0,
      "step": 17395
    },
    {
      "entropy": 5.5665816307067875,
      "epoch": 1.5783744557329462,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004754338509392183,
      "loss": 5.2912,
      "mean_token_accuracy": 0.17392912358045579,
      "num_tokens": 36112352.0,
      "step": 17400
    },
    {
      "entropy": 5.511636781692505,
      "epoch": 1.5788280116110305,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004754191473377153,
      "loss": 5.1663,
      "mean_token_accuracy": 0.18536356687545777,
      "num_tokens": 36121797.0,
      "step": 17405
    },
    {
      "entropy": 5.571134471893311,
      "epoch": 1.5792815674891147,
      "grad_norm": 1.796875,
      "learning_rate": 0.00047540443959149316,
      "loss": 5.3009,
      "mean_token_accuracy": 0.18212876766920089,
      "num_tokens": 36132094.0,
      "step": 17410
    },
    {
      "entropy": 5.53169093132019,
      "epoch": 1.5797351233671988,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004753897277008562,
      "loss": 5.2829,
      "mean_token_accuracy": 0.17925245463848113,
      "num_tokens": 36142860.0,
      "step": 17415
    },
    {
      "entropy": 5.603940391540528,
      "epoch": 1.580188679245283,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004753750116661086,
      "loss": 5.2793,
      "mean_token_accuracy": 0.172691947221756,
      "num_tokens": 36154555.0,
      "step": 17420
    },
    {
      "entropy": 5.636639738082886,
      "epoch": 1.580642235123367,
      "grad_norm": 2.359375,
      "learning_rate": 0.00047536029148755477,
      "loss": 5.3935,
      "mean_token_accuracy": 0.16756432205438615,
      "num_tokens": 36165479.0,
      "step": 17425
    },
    {
      "entropy": 5.569704675674439,
      "epoch": 1.5810957910014514,
      "grad_norm": 1.78125,
      "learning_rate": 0.00047534556716549893,
      "loss": 5.3134,
      "mean_token_accuracy": 0.17519205659627915,
      "num_tokens": 36175873.0,
      "step": 17430
    },
    {
      "entropy": 5.564962673187256,
      "epoch": 1.5815493468795356,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00047533083870024575,
      "loss": 5.2909,
      "mean_token_accuracy": 0.1830425202846527,
      "num_tokens": 36186792.0,
      "step": 17435
    },
    {
      "entropy": 5.71272349357605,
      "epoch": 1.58200290275762,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004753161060920998,
      "loss": 5.4148,
      "mean_token_accuracy": 0.1697296068072319,
      "num_tokens": 36196691.0,
      "step": 17440
    },
    {
      "entropy": 5.646224737167358,
      "epoch": 1.582456458635704,
      "grad_norm": 2.921875,
      "learning_rate": 0.0004753013693413657,
      "loss": 5.2924,
      "mean_token_accuracy": 0.18126584440469742,
      "num_tokens": 36207173.0,
      "step": 17445
    },
    {
      "entropy": 5.503301572799683,
      "epoch": 1.582910014513788,
      "grad_norm": 2.140625,
      "learning_rate": 0.00047528662844834827,
      "loss": 5.3065,
      "mean_token_accuracy": 0.1785816118121147,
      "num_tokens": 36217912.0,
      "step": 17450
    },
    {
      "entropy": 5.511010074615479,
      "epoch": 1.5833635703918723,
      "grad_norm": 2.40625,
      "learning_rate": 0.00047527188341335233,
      "loss": 5.2025,
      "mean_token_accuracy": 0.18541764318943024,
      "num_tokens": 36227714.0,
      "step": 17455
    },
    {
      "entropy": 5.610887336730957,
      "epoch": 1.5838171262699565,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004752571342366828,
      "loss": 5.2977,
      "mean_token_accuracy": 0.18052321076393127,
      "num_tokens": 36239114.0,
      "step": 17460
    },
    {
      "entropy": 5.588019895553589,
      "epoch": 1.5842706821480408,
      "grad_norm": 3.421875,
      "learning_rate": 0.0004752423809186446,
      "loss": 5.3045,
      "mean_token_accuracy": 0.17453892529010773,
      "num_tokens": 36249436.0,
      "step": 17465
    },
    {
      "entropy": 5.5782060623168945,
      "epoch": 1.5847242380261248,
      "grad_norm": 2.296875,
      "learning_rate": 0.00047522762345954305,
      "loss": 5.2498,
      "mean_token_accuracy": 0.17963732779026031,
      "num_tokens": 36259541.0,
      "step": 17470
    },
    {
      "entropy": 5.591192674636841,
      "epoch": 1.5851777939042089,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004752128618596832,
      "loss": 5.2999,
      "mean_token_accuracy": 0.17359114587306976,
      "num_tokens": 36270689.0,
      "step": 17475
    },
    {
      "entropy": 5.667599773406982,
      "epoch": 1.5856313497822931,
      "grad_norm": 2.421875,
      "learning_rate": 0.00047519809611937036,
      "loss": 5.3599,
      "mean_token_accuracy": 0.1725229352712631,
      "num_tokens": 36281191.0,
      "step": 17480
    },
    {
      "entropy": 5.512209844589234,
      "epoch": 1.5860849056603774,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004751833262389098,
      "loss": 5.2572,
      "mean_token_accuracy": 0.18060265630483627,
      "num_tokens": 36291022.0,
      "step": 17485
    },
    {
      "entropy": 5.583451604843139,
      "epoch": 1.5865384615384617,
      "grad_norm": 2.65625,
      "learning_rate": 0.00047516855221860715,
      "loss": 5.3302,
      "mean_token_accuracy": 0.17770405411720275,
      "num_tokens": 36301237.0,
      "step": 17490
    },
    {
      "entropy": 5.619327545166016,
      "epoch": 1.5869920174165457,
      "grad_norm": 2.0625,
      "learning_rate": 0.00047515377405876774,
      "loss": 5.2977,
      "mean_token_accuracy": 0.1720043033361435,
      "num_tokens": 36311792.0,
      "step": 17495
    },
    {
      "entropy": 5.516510963439941,
      "epoch": 1.5874455732946298,
      "grad_norm": 3.046875,
      "learning_rate": 0.0004751389917596973,
      "loss": 5.2878,
      "mean_token_accuracy": 0.1841294512152672,
      "num_tokens": 36321847.0,
      "step": 17500
    },
    {
      "entropy": 5.568793821334839,
      "epoch": 1.587899129172714,
      "grad_norm": 2.0625,
      "learning_rate": 0.00047512420532170154,
      "loss": 5.3532,
      "mean_token_accuracy": 0.1710747703909874,
      "num_tokens": 36332735.0,
      "step": 17505
    },
    {
      "entropy": 5.658964586257935,
      "epoch": 1.5883526850507983,
      "grad_norm": 1.75,
      "learning_rate": 0.0004751094147450862,
      "loss": 5.3638,
      "mean_token_accuracy": 0.17166862785816192,
      "num_tokens": 36343149.0,
      "step": 17510
    },
    {
      "entropy": 5.611267757415772,
      "epoch": 1.5888062409288826,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004750946200301571,
      "loss": 5.2531,
      "mean_token_accuracy": 0.18371012955904006,
      "num_tokens": 36353081.0,
      "step": 17515
    },
    {
      "entropy": 5.568607902526855,
      "epoch": 1.5892597968069666,
      "grad_norm": 2.3125,
      "learning_rate": 0.00047507982117722035,
      "loss": 5.3269,
      "mean_token_accuracy": 0.17549228072166442,
      "num_tokens": 36363624.0,
      "step": 17520
    },
    {
      "entropy": 5.636461448669434,
      "epoch": 1.5897133526850507,
      "grad_norm": 2.140625,
      "learning_rate": 0.000475065018186582,
      "loss": 5.3438,
      "mean_token_accuracy": 0.17758631110191345,
      "num_tokens": 36374081.0,
      "step": 17525
    },
    {
      "entropy": 5.626317834854126,
      "epoch": 1.590166908563135,
      "grad_norm": 2.78125,
      "learning_rate": 0.00047505021105854794,
      "loss": 5.4075,
      "mean_token_accuracy": 0.17428145855665206,
      "num_tokens": 36384360.0,
      "step": 17530
    },
    {
      "entropy": 5.628881740570068,
      "epoch": 1.5906204644412192,
      "grad_norm": 2.71875,
      "learning_rate": 0.00047503539979342456,
      "loss": 5.3439,
      "mean_token_accuracy": 0.17727007120847701,
      "num_tokens": 36395924.0,
      "step": 17535
    },
    {
      "entropy": 5.535763072967529,
      "epoch": 1.5910740203193035,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004750205843915182,
      "loss": 5.2084,
      "mean_token_accuracy": 0.1836434781551361,
      "num_tokens": 36405246.0,
      "step": 17540
    },
    {
      "entropy": 5.574892234802246,
      "epoch": 1.5915275761973875,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004750057648531351,
      "loss": 5.2757,
      "mean_token_accuracy": 0.1820523962378502,
      "num_tokens": 36414964.0,
      "step": 17545
    },
    {
      "entropy": 5.5513739585876465,
      "epoch": 1.5919811320754715,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004749909411785819,
      "loss": 5.3342,
      "mean_token_accuracy": 0.1754442036151886,
      "num_tokens": 36425011.0,
      "step": 17550
    },
    {
      "entropy": 5.535242652893066,
      "epoch": 1.5924346879535558,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000474976113368165,
      "loss": 5.2292,
      "mean_token_accuracy": 0.18569039106369017,
      "num_tokens": 36434734.0,
      "step": 17555
    },
    {
      "entropy": 5.601694536209107,
      "epoch": 1.59288824383164,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004749612814221912,
      "loss": 5.3077,
      "mean_token_accuracy": 0.17970757335424423,
      "num_tokens": 36444748.0,
      "step": 17560
    },
    {
      "entropy": 5.6988465785980225,
      "epoch": 1.5933417997097243,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047494644534096703,
      "loss": 5.3249,
      "mean_token_accuracy": 0.17141301333904266,
      "num_tokens": 36455076.0,
      "step": 17565
    },
    {
      "entropy": 5.560309743881225,
      "epoch": 1.5937953555878084,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004749316051247995,
      "loss": 5.2536,
      "mean_token_accuracy": 0.18055053651332856,
      "num_tokens": 36465302.0,
      "step": 17570
    },
    {
      "entropy": 5.4911620140075685,
      "epoch": 1.5942489114658926,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00047491676077399543,
      "loss": 5.3652,
      "mean_token_accuracy": 0.17181792855262756,
      "num_tokens": 36476666.0,
      "step": 17575
    },
    {
      "entropy": 5.5574909210205075,
      "epoch": 1.5947024673439767,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004749019122888618,
      "loss": 5.221,
      "mean_token_accuracy": 0.1810115858912468,
      "num_tokens": 36487379.0,
      "step": 17580
    },
    {
      "entropy": 5.612413740158081,
      "epoch": 1.595156023222061,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047488705966970565,
      "loss": 5.2257,
      "mean_token_accuracy": 0.1844322144985199,
      "num_tokens": 36497897.0,
      "step": 17585
    },
    {
      "entropy": 5.567325162887573,
      "epoch": 1.5956095791001452,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004748722029168342,
      "loss": 5.3587,
      "mean_token_accuracy": 0.1787263885140419,
      "num_tokens": 36508528.0,
      "step": 17590
    },
    {
      "entropy": 5.609759950637818,
      "epoch": 1.5960631349782293,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004748573420305546,
      "loss": 5.3361,
      "mean_token_accuracy": 0.1720815747976303,
      "num_tokens": 36519196.0,
      "step": 17595
    },
    {
      "entropy": 5.578181552886963,
      "epoch": 1.5965166908563135,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047484247701117427,
      "loss": 5.3221,
      "mean_token_accuracy": 0.17435511499643325,
      "num_tokens": 36529871.0,
      "step": 17600
    },
    {
      "entropy": 5.666941547393799,
      "epoch": 1.5969702467343976,
      "grad_norm": 2.125,
      "learning_rate": 0.0004748276078590006,
      "loss": 5.3767,
      "mean_token_accuracy": 0.16987912803888322,
      "num_tokens": 36541188.0,
      "step": 17605
    },
    {
      "entropy": 5.643979072570801,
      "epoch": 1.5974238026124818,
      "grad_norm": 2.828125,
      "learning_rate": 0.0004748127345743411,
      "loss": 5.3701,
      "mean_token_accuracy": 0.1743515834212303,
      "num_tokens": 36551968.0,
      "step": 17610
    },
    {
      "entropy": 5.576541948318481,
      "epoch": 1.5978773584905661,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004747978571575033,
      "loss": 5.2964,
      "mean_token_accuracy": 0.17886736392974853,
      "num_tokens": 36562870.0,
      "step": 17615
    },
    {
      "entropy": 5.5086695671081545,
      "epoch": 1.5983309143686504,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004747829756087949,
      "loss": 5.2388,
      "mean_token_accuracy": 0.181069815158844,
      "num_tokens": 36573295.0,
      "step": 17620
    },
    {
      "entropy": 5.5391926765441895,
      "epoch": 1.5987844702467344,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047476808992852364,
      "loss": 5.234,
      "mean_token_accuracy": 0.1844158500432968,
      "num_tokens": 36583830.0,
      "step": 17625
    },
    {
      "entropy": 5.574624872207641,
      "epoch": 1.5992380261248185,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004747532001169974,
      "loss": 5.2999,
      "mean_token_accuracy": 0.18300977796316148,
      "num_tokens": 36594052.0,
      "step": 17630
    },
    {
      "entropy": 5.568613290786743,
      "epoch": 1.5996915820029027,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00047473830617452415,
      "loss": 5.3418,
      "mean_token_accuracy": 0.17662158906459807,
      "num_tokens": 36604378.0,
      "step": 17635
    },
    {
      "entropy": 5.6372918605804445,
      "epoch": 1.600145137880987,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047472340810141163,
      "loss": 5.4337,
      "mean_token_accuracy": 0.16617694050073623,
      "num_tokens": 36615588.0,
      "step": 17640
    },
    {
      "entropy": 5.697837495803833,
      "epoch": 1.6005986937590713,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047470850589796833,
      "loss": 5.332,
      "mean_token_accuracy": 0.17311546355485916,
      "num_tokens": 36626369.0,
      "step": 17645
    },
    {
      "entropy": 5.628124618530274,
      "epoch": 1.6010522496371553,
      "grad_norm": 2.90625,
      "learning_rate": 0.00047469359956450215,
      "loss": 5.3521,
      "mean_token_accuracy": 0.17312052249908447,
      "num_tokens": 36636565.0,
      "step": 17650
    },
    {
      "entropy": 5.626499462127685,
      "epoch": 1.6015058055152394,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00047467868910132144,
      "loss": 5.4231,
      "mean_token_accuracy": 0.1726812720298767,
      "num_tokens": 36647721.0,
      "step": 17655
    },
    {
      "entropy": 5.586396646499634,
      "epoch": 1.6019593613933236,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004746637745087346,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18292518854141235,
      "num_tokens": 36657887.0,
      "step": 17660
    },
    {
      "entropy": 5.646685934066772,
      "epoch": 1.6024129172714079,
      "grad_norm": 2.421875,
      "learning_rate": 0.00047464885578705,
      "loss": 5.3146,
      "mean_token_accuracy": 0.1757381498813629,
      "num_tokens": 36667918.0,
      "step": 17665
    },
    {
      "entropy": 5.493096399307251,
      "epoch": 1.6028664731494922,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00047463393293657616,
      "loss": 5.2628,
      "mean_token_accuracy": 0.186036416888237,
      "num_tokens": 36678491.0,
      "step": 17670
    },
    {
      "entropy": 5.557955932617188,
      "epoch": 1.6033200290275762,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004746190059576217,
      "loss": 5.3431,
      "mean_token_accuracy": 0.17912136465311052,
      "num_tokens": 36688638.0,
      "step": 17675
    },
    {
      "entropy": 5.63788104057312,
      "epoch": 1.6037735849056602,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004746040748504954,
      "loss": 5.3287,
      "mean_token_accuracy": 0.1824111372232437,
      "num_tokens": 36699344.0,
      "step": 17680
    },
    {
      "entropy": 5.643543863296509,
      "epoch": 1.6042271407837445,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047458913961550593,
      "loss": 5.2645,
      "mean_token_accuracy": 0.17989160120487213,
      "num_tokens": 36709606.0,
      "step": 17685
    },
    {
      "entropy": 5.517097997665405,
      "epoch": 1.6046806966618288,
      "grad_norm": 5.65625,
      "learning_rate": 0.0004745742002529622,
      "loss": 5.24,
      "mean_token_accuracy": 0.18074310421943665,
      "num_tokens": 36718740.0,
      "step": 17690
    },
    {
      "entropy": 5.574347686767578,
      "epoch": 1.605134252539913,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004745592567631731,
      "loss": 5.3205,
      "mean_token_accuracy": 0.17144889682531356,
      "num_tokens": 36729960.0,
      "step": 17695
    },
    {
      "entropy": 5.488494920730591,
      "epoch": 1.605587808417997,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00047454430914644773,
      "loss": 5.2222,
      "mean_token_accuracy": 0.18233761340379714,
      "num_tokens": 36740221.0,
      "step": 17700
    },
    {
      "entropy": 5.498316669464112,
      "epoch": 1.6060413642960811,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004745293574030952,
      "loss": 5.2444,
      "mean_token_accuracy": 0.18478453457355498,
      "num_tokens": 36750841.0,
      "step": 17705
    },
    {
      "entropy": 5.532949447631836,
      "epoch": 1.6064949201741654,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047451440153342477,
      "loss": 5.2846,
      "mean_token_accuracy": 0.17422025948762893,
      "num_tokens": 36760944.0,
      "step": 17710
    },
    {
      "entropy": 5.56636266708374,
      "epoch": 1.6069484760522497,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004744994415377457,
      "loss": 5.2795,
      "mean_token_accuracy": 0.17990392446517944,
      "num_tokens": 36770922.0,
      "step": 17715
    },
    {
      "entropy": 5.611558437347412,
      "epoch": 1.607402031930334,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047448447741636724,
      "loss": 5.3065,
      "mean_token_accuracy": 0.1748896449804306,
      "num_tokens": 36780702.0,
      "step": 17720
    },
    {
      "entropy": 5.623646545410156,
      "epoch": 1.607855587808418,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047446950916959905,
      "loss": 5.3507,
      "mean_token_accuracy": 0.16943768560886383,
      "num_tokens": 36791868.0,
      "step": 17725
    },
    {
      "entropy": 5.638291215896606,
      "epoch": 1.608309143686502,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004744545367977505,
      "loss": 5.326,
      "mean_token_accuracy": 0.17830570936203002,
      "num_tokens": 36802330.0,
      "step": 17730
    },
    {
      "entropy": 5.633695125579834,
      "epoch": 1.6087626995645863,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00047443956030113134,
      "loss": 5.3549,
      "mean_token_accuracy": 0.17336882650852203,
      "num_tokens": 36812936.0,
      "step": 17735
    },
    {
      "entropy": 5.566090774536133,
      "epoch": 1.6092162554426706,
      "grad_norm": 2.1875,
      "learning_rate": 0.00047442457968005123,
      "loss": 5.3129,
      "mean_token_accuracy": 0.17918934375047685,
      "num_tokens": 36823123.0,
      "step": 17740
    },
    {
      "entropy": 5.525178480148315,
      "epoch": 1.6096698113207548,
      "grad_norm": 2.0,
      "learning_rate": 0.00047440959493481996,
      "loss": 5.2073,
      "mean_token_accuracy": 0.18745797276496887,
      "num_tokens": 36833050.0,
      "step": 17745
    },
    {
      "entropy": 5.62385926246643,
      "epoch": 1.6101233671988389,
      "grad_norm": 2.25,
      "learning_rate": 0.00047439460606574745,
      "loss": 5.341,
      "mean_token_accuracy": 0.177030511200428,
      "num_tokens": 36842520.0,
      "step": 17750
    },
    {
      "entropy": 5.570717525482178,
      "epoch": 1.6105769230769231,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004743796130731437,
      "loss": 5.2881,
      "mean_token_accuracy": 0.17283373326063156,
      "num_tokens": 36853024.0,
      "step": 17755
    },
    {
      "entropy": 5.590139389038086,
      "epoch": 1.6110304789550072,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004743646159573188,
      "loss": 5.2394,
      "mean_token_accuracy": 0.18364014625549316,
      "num_tokens": 36862534.0,
      "step": 17760
    },
    {
      "entropy": 5.526464986801147,
      "epoch": 1.6114840348330914,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004743496147185828,
      "loss": 5.2277,
      "mean_token_accuracy": 0.18336836099624634,
      "num_tokens": 36872356.0,
      "step": 17765
    },
    {
      "entropy": 5.522708702087402,
      "epoch": 1.6119375907111757,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047433460935724595,
      "loss": 5.2853,
      "mean_token_accuracy": 0.1796351119875908,
      "num_tokens": 36882643.0,
      "step": 17770
    },
    {
      "entropy": 5.602589464187622,
      "epoch": 1.6123911465892597,
      "grad_norm": 1.875,
      "learning_rate": 0.0004743195998736186,
      "loss": 5.2876,
      "mean_token_accuracy": 0.1787036269903183,
      "num_tokens": 36893349.0,
      "step": 17775
    },
    {
      "entropy": 5.7006911277771,
      "epoch": 1.612844702467344,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047430458626801107,
      "loss": 5.2958,
      "mean_token_accuracy": 0.1805070474743843,
      "num_tokens": 36903713.0,
      "step": 17780
    },
    {
      "entropy": 5.531917667388916,
      "epoch": 1.613298258345428,
      "grad_norm": 1.9375,
      "learning_rate": 0.000474289568540734,
      "loss": 5.2138,
      "mean_token_accuracy": 0.1790839374065399,
      "num_tokens": 36912999.0,
      "step": 17785
    },
    {
      "entropy": 5.590320634841919,
      "epoch": 1.6137518142235123,
      "grad_norm": 1.9375,
      "learning_rate": 0.00047427454669209785,
      "loss": 5.3387,
      "mean_token_accuracy": 0.17709368020296096,
      "num_tokens": 36923598.0,
      "step": 17790
    },
    {
      "entropy": 5.533603811264038,
      "epoch": 1.6142053701015966,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047425952072241326,
      "loss": 5.2477,
      "mean_token_accuracy": 0.17710369527339936,
      "num_tokens": 36933036.0,
      "step": 17795
    },
    {
      "entropy": 5.593995475769043,
      "epoch": 1.6146589259796809,
      "grad_norm": 2.828125,
      "learning_rate": 0.000474244490631991,
      "loss": 5.3479,
      "mean_token_accuracy": 0.17640879899263381,
      "num_tokens": 36943137.0,
      "step": 17800
    },
    {
      "entropy": 5.715324115753174,
      "epoch": 1.615112481857765,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004742294564211419,
      "loss": 5.4907,
      "mean_token_accuracy": 0.1650419756770134,
      "num_tokens": 36955619.0,
      "step": 17805
    },
    {
      "entropy": 5.565889978408814,
      "epoch": 1.615566037735849,
      "grad_norm": 1.5,
      "learning_rate": 0.0004742144180901769,
      "loss": 5.2551,
      "mean_token_accuracy": 0.18613714426755906,
      "num_tokens": 36966377.0,
      "step": 17810
    },
    {
      "entropy": 5.570078420639038,
      "epoch": 1.6160195936139332,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047419937563940697,
      "loss": 5.2984,
      "mean_token_accuracy": 0.18170032799243926,
      "num_tokens": 36977100.0,
      "step": 17815
    },
    {
      "entropy": 5.603732490539551,
      "epoch": 1.6164731494920175,
      "grad_norm": 2.0,
      "learning_rate": 0.00047418432906914314,
      "loss": 5.3783,
      "mean_token_accuracy": 0.17650981843471528,
      "num_tokens": 36987120.0,
      "step": 17820
    },
    {
      "entropy": 5.64243483543396,
      "epoch": 1.6169267053701017,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004741692783796967,
      "loss": 5.3445,
      "mean_token_accuracy": 0.17881186157464982,
      "num_tokens": 36996885.0,
      "step": 17825
    },
    {
      "entropy": 5.535692262649536,
      "epoch": 1.6173802612481858,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00047415422357137876,
      "loss": 5.2628,
      "mean_token_accuracy": 0.1851559340953827,
      "num_tokens": 37006570.0,
      "step": 17830
    },
    {
      "entropy": 5.625456809997559,
      "epoch": 1.6178338171262698,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004741391646445008,
      "loss": 5.3481,
      "mean_token_accuracy": 0.1750331312417984,
      "num_tokens": 37017185.0,
      "step": 17835
    },
    {
      "entropy": 5.59498028755188,
      "epoch": 1.618287373004354,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004741241015993741,
      "loss": 5.3155,
      "mean_token_accuracy": 0.18061118572950363,
      "num_tokens": 37028386.0,
      "step": 17840
    },
    {
      "entropy": 5.515576648712158,
      "epoch": 1.6187409288824384,
      "grad_norm": 1.625,
      "learning_rate": 0.00047410903443631026,
      "loss": 5.209,
      "mean_token_accuracy": 0.18232527375221252,
      "num_tokens": 37038326.0,
      "step": 17845
    },
    {
      "entropy": 5.644190692901612,
      "epoch": 1.6191944847605226,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004740939631556209,
      "loss": 5.3191,
      "mean_token_accuracy": 0.18096337467432022,
      "num_tokens": 37048570.0,
      "step": 17850
    },
    {
      "entropy": 5.522991895675659,
      "epoch": 1.6196480406386067,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004740788877576176,
      "loss": 5.2092,
      "mean_token_accuracy": 0.18529599010944367,
      "num_tokens": 37058909.0,
      "step": 17855
    },
    {
      "entropy": 5.496800136566162,
      "epoch": 1.6201015965166907,
      "grad_norm": 3.703125,
      "learning_rate": 0.0004740638082426121,
      "loss": 5.2806,
      "mean_token_accuracy": 0.18698763698339463,
      "num_tokens": 37068957.0,
      "step": 17860
    },
    {
      "entropy": 5.56817479133606,
      "epoch": 1.620555152394775,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00047404872461091643,
      "loss": 5.2307,
      "mean_token_accuracy": 0.1811065137386322,
      "num_tokens": 37080187.0,
      "step": 17865
    },
    {
      "entropy": 5.564313364028931,
      "epoch": 1.6210087082728593,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00047403363686284236,
      "loss": 5.3005,
      "mean_token_accuracy": 0.17366160899400712,
      "num_tokens": 37090314.0,
      "step": 17870
    },
    {
      "entropy": 5.6116862297058105,
      "epoch": 1.6214622641509435,
      "grad_norm": 2.28125,
      "learning_rate": 0.00047401854499870197,
      "loss": 5.34,
      "mean_token_accuracy": 0.17598073333501815,
      "num_tokens": 37099907.0,
      "step": 17875
    },
    {
      "entropy": 5.585748243331909,
      "epoch": 1.6219158200290276,
      "grad_norm": 2.4375,
      "learning_rate": 0.00047400344901880735,
      "loss": 5.3567,
      "mean_token_accuracy": 0.1745814263820648,
      "num_tokens": 37110495.0,
      "step": 17880
    },
    {
      "entropy": 5.629552793502808,
      "epoch": 1.6223693759071116,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004739883489234707,
      "loss": 5.3478,
      "mean_token_accuracy": 0.1769860118627548,
      "num_tokens": 37121957.0,
      "step": 17885
    },
    {
      "entropy": 5.581470394134522,
      "epoch": 1.6228229317851959,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047397324471300424,
      "loss": 5.2287,
      "mean_token_accuracy": 0.18291717767715454,
      "num_tokens": 37132743.0,
      "step": 17890
    },
    {
      "entropy": 5.596113395690918,
      "epoch": 1.6232764876632801,
      "grad_norm": 2.25,
      "learning_rate": 0.00047395813638772047,
      "loss": 5.3497,
      "mean_token_accuracy": 0.17141922414302826,
      "num_tokens": 37143947.0,
      "step": 17895
    },
    {
      "entropy": 5.7439638614654545,
      "epoch": 1.6237300435413644,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047394302394793164,
      "loss": 5.4288,
      "mean_token_accuracy": 0.16941222846508025,
      "num_tokens": 37154316.0,
      "step": 17900
    },
    {
      "entropy": 5.599140977859497,
      "epoch": 1.6241835994194485,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004739279073939504,
      "loss": 5.3237,
      "mean_token_accuracy": 0.17644609957933427,
      "num_tokens": 37164999.0,
      "step": 17905
    },
    {
      "entropy": 5.592998313903808,
      "epoch": 1.6246371552975325,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004739127867260894,
      "loss": 5.3324,
      "mean_token_accuracy": 0.17743028402328492,
      "num_tokens": 37174392.0,
      "step": 17910
    },
    {
      "entropy": 5.546643924713135,
      "epoch": 1.6250907111756168,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004738976619446613,
      "loss": 5.3433,
      "mean_token_accuracy": 0.1760048747062683,
      "num_tokens": 37185341.0,
      "step": 17915
    },
    {
      "entropy": 5.698810815811157,
      "epoch": 1.625544267053701,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00047388253304997874,
      "loss": 5.3308,
      "mean_token_accuracy": 0.17272518426179886,
      "num_tokens": 37196725.0,
      "step": 17920
    },
    {
      "entropy": 5.569040679931641,
      "epoch": 1.6259978229317853,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004738674000423548,
      "loss": 5.24,
      "mean_token_accuracy": 0.1813740223646164,
      "num_tokens": 37206877.0,
      "step": 17925
    },
    {
      "entropy": 5.455147933959961,
      "epoch": 1.6264513788098693,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004738522629221023,
      "loss": 5.2057,
      "mean_token_accuracy": 0.17897325456142427,
      "num_tokens": 37217277.0,
      "step": 17930
    },
    {
      "entropy": 5.488034582138061,
      "epoch": 1.6269049346879536,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004738371216895344,
      "loss": 5.2069,
      "mean_token_accuracy": 0.18273129016160966,
      "num_tokens": 37227716.0,
      "step": 17935
    },
    {
      "entropy": 5.536426973342896,
      "epoch": 1.6273584905660377,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004738219763449641,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1750743418931961,
      "num_tokens": 37238178.0,
      "step": 17940
    },
    {
      "entropy": 5.571993780136109,
      "epoch": 1.627812046444122,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004738068268887047,
      "loss": 5.3395,
      "mean_token_accuracy": 0.18040392845869063,
      "num_tokens": 37248408.0,
      "step": 17945
    },
    {
      "entropy": 5.604945516586303,
      "epoch": 1.6282656023222062,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004737916733210694,
      "loss": 5.4303,
      "mean_token_accuracy": 0.17124892622232438,
      "num_tokens": 37259176.0,
      "step": 17950
    },
    {
      "entropy": 5.5989540100097654,
      "epoch": 1.6287191582002902,
      "grad_norm": 2.015625,
      "learning_rate": 0.00047377651564237165,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1876084938645363,
      "num_tokens": 37269205.0,
      "step": 17955
    },
    {
      "entropy": 5.592734527587891,
      "epoch": 1.6291727140783745,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047376135385292494,
      "loss": 5.3179,
      "mean_token_accuracy": 0.186687870323658,
      "num_tokens": 37278841.0,
      "step": 17960
    },
    {
      "entropy": 5.568622350692749,
      "epoch": 1.6296262699564585,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004737461879530428,
      "loss": 5.3332,
      "mean_token_accuracy": 0.17898489087820052,
      "num_tokens": 37288487.0,
      "step": 17965
    },
    {
      "entropy": 5.611185836791992,
      "epoch": 1.6300798258345428,
      "grad_norm": 1.5,
      "learning_rate": 0.00047373101794303884,
      "loss": 5.3326,
      "mean_token_accuracy": 0.17773114144802094,
      "num_tokens": 37300086.0,
      "step": 17970
    },
    {
      "entropy": 5.626567459106445,
      "epoch": 1.630533381712627,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004737158438232267,
      "loss": 5.3554,
      "mean_token_accuracy": 0.16666506230831146,
      "num_tokens": 37310072.0,
      "step": 17975
    },
    {
      "entropy": 5.6384035587310795,
      "epoch": 1.6309869375907113,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047370066559392033,
      "loss": 5.3129,
      "mean_token_accuracy": 0.18591950684785843,
      "num_tokens": 37320278.0,
      "step": 17980
    },
    {
      "entropy": 5.508388423919678,
      "epoch": 1.6314404934687954,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00047368548325543366,
      "loss": 5.2351,
      "mean_token_accuracy": 0.17428006529808043,
      "num_tokens": 37330741.0,
      "step": 17985
    },
    {
      "entropy": 5.567923736572266,
      "epoch": 1.6318940493468794,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004736702968080804,
      "loss": 5.333,
      "mean_token_accuracy": 0.17588341534137725,
      "num_tokens": 37341213.0,
      "step": 17990
    },
    {
      "entropy": 5.548464250564575,
      "epoch": 1.6323476052249637,
      "grad_norm": 3.6875,
      "learning_rate": 0.00047365510625217483,
      "loss": 5.2098,
      "mean_token_accuracy": 0.1842402622103691,
      "num_tokens": 37350528.0,
      "step": 17995
    },
    {
      "entropy": 5.552949333190918,
      "epoch": 1.632801161103048,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00047363991158803117,
      "loss": 5.2984,
      "mean_token_accuracy": 0.17459098100662232,
      "num_tokens": 37359938.0,
      "step": 18000
    },
    {
      "epoch": 1.632801161103048,
      "eval_entropy": 5.3826052208292,
      "eval_loss": 5.35751485824585,
      "eval_mean_token_accuracy": 0.1835274959834318,
      "eval_num_tokens": 37359938.0,
      "eval_runtime": 26.2138,
      "eval_samples_per_second": 1348.91,
      "eval_steps_per_second": 168.614,
      "step": 18000
    },
    {
      "entropy": 5.566889381408691,
      "epoch": 1.6332547169811322,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047362471281596334,
      "loss": 5.2894,
      "mean_token_accuracy": 0.179564867913723,
      "num_tokens": 37369744.0,
      "step": 18005
    },
    {
      "entropy": 5.616558074951172,
      "epoch": 1.6337082728592163,
      "grad_norm": 1.984375,
      "learning_rate": 0.00047360950993628585,
      "loss": 5.3804,
      "mean_token_accuracy": 0.17757397443056105,
      "num_tokens": 37380578.0,
      "step": 18010
    },
    {
      "entropy": 5.581391096115112,
      "epoch": 1.6341618287373003,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004735943029493132,
      "loss": 5.2796,
      "mean_token_accuracy": 0.1792287603020668,
      "num_tokens": 37390237.0,
      "step": 18015
    },
    {
      "entropy": 5.625637674331665,
      "epoch": 1.6346153846153846,
      "grad_norm": 2.109375,
      "learning_rate": 0.00047357909185535965,
      "loss": 5.3371,
      "mean_token_accuracy": 0.17637136280536653,
      "num_tokens": 37400261.0,
      "step": 18020
    },
    {
      "entropy": 5.587732934951783,
      "epoch": 1.6350689404934688,
      "grad_norm": 1.625,
      "learning_rate": 0.0004735638766547399,
      "loss": 5.2931,
      "mean_token_accuracy": 0.17909928411245346,
      "num_tokens": 37410012.0,
      "step": 18025
    },
    {
      "entropy": 5.593369436264038,
      "epoch": 1.6355224963715531,
      "grad_norm": 1.75,
      "learning_rate": 0.00047354865734776857,
      "loss": 5.4023,
      "mean_token_accuracy": 0.16935993283987044,
      "num_tokens": 37420530.0,
      "step": 18030
    },
    {
      "entropy": 5.6103901863098145,
      "epoch": 1.6359760522496372,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004735334339347605,
      "loss": 5.2793,
      "mean_token_accuracy": 0.1897657722234726,
      "num_tokens": 37431146.0,
      "step": 18035
    },
    {
      "entropy": 5.597890567779541,
      "epoch": 1.6364296081277212,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00047351820641603035,
      "loss": 5.2254,
      "mean_token_accuracy": 0.18372856229543685,
      "num_tokens": 37441253.0,
      "step": 18040
    },
    {
      "entropy": 5.501802921295166,
      "epoch": 1.6368831640058055,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004735029747918931,
      "loss": 5.2963,
      "mean_token_accuracy": 0.17532598376274108,
      "num_tokens": 37451326.0,
      "step": 18045
    },
    {
      "entropy": 5.575360250473023,
      "epoch": 1.6373367198838897,
      "grad_norm": 2.015625,
      "learning_rate": 0.00047348773906266374,
      "loss": 5.2617,
      "mean_token_accuracy": 0.18154544830322267,
      "num_tokens": 37462462.0,
      "step": 18050
    },
    {
      "entropy": 5.539795541763306,
      "epoch": 1.637790275761974,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004734724992286574,
      "loss": 5.2359,
      "mean_token_accuracy": 0.1820863112807274,
      "num_tokens": 37473016.0,
      "step": 18055
    },
    {
      "entropy": 5.582274293899536,
      "epoch": 1.638243831640058,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004734572552901892,
      "loss": 5.3675,
      "mean_token_accuracy": 0.1775408133864403,
      "num_tokens": 37483737.0,
      "step": 18060
    },
    {
      "entropy": 5.573561429977417,
      "epoch": 1.638697387518142,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004734420072475743,
      "loss": 5.2524,
      "mean_token_accuracy": 0.18024279475212096,
      "num_tokens": 37493070.0,
      "step": 18065
    },
    {
      "entropy": 5.566709852218628,
      "epoch": 1.6391509433962264,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004734267551011282,
      "loss": 5.2708,
      "mean_token_accuracy": 0.1766637772321701,
      "num_tokens": 37503696.0,
      "step": 18070
    },
    {
      "entropy": 5.509756183624267,
      "epoch": 1.6396044992743106,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004734114988511662,
      "loss": 5.2977,
      "mean_token_accuracy": 0.1736341282725334,
      "num_tokens": 37513468.0,
      "step": 18075
    },
    {
      "entropy": 5.514056777954101,
      "epoch": 1.640058055152395,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004733962384980039,
      "loss": 5.1594,
      "mean_token_accuracy": 0.18871453702449797,
      "num_tokens": 37523631.0,
      "step": 18080
    },
    {
      "entropy": 5.628369569778442,
      "epoch": 1.640511611030479,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004733809740419568,
      "loss": 5.3339,
      "mean_token_accuracy": 0.17943619936704636,
      "num_tokens": 37534602.0,
      "step": 18085
    },
    {
      "entropy": 5.517502737045288,
      "epoch": 1.640965166908563,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047336570548334053,
      "loss": 5.2718,
      "mean_token_accuracy": 0.17896320074796676,
      "num_tokens": 37544547.0,
      "step": 18090
    },
    {
      "entropy": 5.587415075302124,
      "epoch": 1.6414187227866472,
      "grad_norm": 3.015625,
      "learning_rate": 0.00047335043282247105,
      "loss": 5.2761,
      "mean_token_accuracy": 0.17600781172513963,
      "num_tokens": 37554903.0,
      "step": 18095
    },
    {
      "entropy": 5.57360258102417,
      "epoch": 1.6418722786647315,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047333515605966396,
      "loss": 5.2041,
      "mean_token_accuracy": 0.17939600497484207,
      "num_tokens": 37565586.0,
      "step": 18100
    },
    {
      "entropy": 5.5247461795806885,
      "epoch": 1.6423258345428158,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047331987519523533,
      "loss": 5.3059,
      "mean_token_accuracy": 0.1821260705590248,
      "num_tokens": 37575996.0,
      "step": 18105
    },
    {
      "entropy": 5.527936220169067,
      "epoch": 1.6427793904208998,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004733045902295012,
      "loss": 5.1956,
      "mean_token_accuracy": 0.17629784792661668,
      "num_tokens": 37586441.0,
      "step": 18110
    },
    {
      "entropy": 5.601177120208741,
      "epoch": 1.643232946298984,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047328930116277755,
      "loss": 5.3553,
      "mean_token_accuracy": 0.17627686262130737,
      "num_tokens": 37598226.0,
      "step": 18115
    },
    {
      "entropy": 5.570518350601196,
      "epoch": 1.6436865021770681,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004732740079953806,
      "loss": 5.315,
      "mean_token_accuracy": 0.17447395324707032,
      "num_tokens": 37609102.0,
      "step": 18120
    },
    {
      "entropy": 5.555383777618408,
      "epoch": 1.6441400580551524,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004732587107276267,
      "loss": 5.3162,
      "mean_token_accuracy": 0.1735365092754364,
      "num_tokens": 37619349.0,
      "step": 18125
    },
    {
      "entropy": 5.654542684555054,
      "epoch": 1.6445936139332367,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00047324340935983225,
      "loss": 5.3008,
      "mean_token_accuracy": 0.1872055783867836,
      "num_tokens": 37629844.0,
      "step": 18130
    },
    {
      "entropy": 5.546751642227173,
      "epoch": 1.6450471698113207,
      "grad_norm": 2.25,
      "learning_rate": 0.00047322810389231333,
      "loss": 5.2119,
      "mean_token_accuracy": 0.17782589495182038,
      "num_tokens": 37639339.0,
      "step": 18135
    },
    {
      "entropy": 5.502415990829467,
      "epoch": 1.645500725689405,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004732127943253868,
      "loss": 5.2318,
      "mean_token_accuracy": 0.18662937581539155,
      "num_tokens": 37648836.0,
      "step": 18140
    },
    {
      "entropy": 5.569932126998902,
      "epoch": 1.645954281567489,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004731974806593693,
      "loss": 5.2464,
      "mean_token_accuracy": 0.17925968319177626,
      "num_tokens": 37658424.0,
      "step": 18145
    },
    {
      "entropy": 5.57012152671814,
      "epoch": 1.6464078374455733,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00047318216289457727,
      "loss": 5.2616,
      "mean_token_accuracy": 0.1818898141384125,
      "num_tokens": 37669849.0,
      "step": 18150
    },
    {
      "entropy": 5.56253604888916,
      "epoch": 1.6468613933236576,
      "grad_norm": 1.75,
      "learning_rate": 0.00047316684103132766,
      "loss": 5.2359,
      "mean_token_accuracy": 0.1830328658223152,
      "num_tokens": 37678987.0,
      "step": 18155
    },
    {
      "entropy": 5.634511566162109,
      "epoch": 1.6473149492017418,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004731515150699372,
      "loss": 5.2769,
      "mean_token_accuracy": 0.1763278976082802,
      "num_tokens": 37689692.0,
      "step": 18160
    },
    {
      "entropy": 5.579637479782105,
      "epoch": 1.6477685050798259,
      "grad_norm": 1.875,
      "learning_rate": 0.00047313618501072303,
      "loss": 5.4282,
      "mean_token_accuracy": 0.17266168296337128,
      "num_tokens": 37700092.0,
      "step": 18165
    },
    {
      "entropy": 5.543984889984131,
      "epoch": 1.64822206095791,
      "grad_norm": 1.953125,
      "learning_rate": 0.00047312085085400194,
      "loss": 5.3262,
      "mean_token_accuracy": 0.17469842433929444,
      "num_tokens": 37710305.0,
      "step": 18170
    },
    {
      "entropy": 5.69915623664856,
      "epoch": 1.6486756168359942,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047310551260009126,
      "loss": 5.3981,
      "mean_token_accuracy": 0.16923402100801468,
      "num_tokens": 37720451.0,
      "step": 18175
    },
    {
      "entropy": 5.628125381469727,
      "epoch": 1.6491291727140784,
      "grad_norm": 2.34375,
      "learning_rate": 0.000473090170249308,
      "loss": 5.3064,
      "mean_token_accuracy": 0.18191324174404144,
      "num_tokens": 37731952.0,
      "step": 18180
    },
    {
      "entropy": 5.510310649871826,
      "epoch": 1.6495827285921627,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004730748238019696,
      "loss": 5.2289,
      "mean_token_accuracy": 0.1747705176472664,
      "num_tokens": 37742201.0,
      "step": 18185
    },
    {
      "entropy": 5.515878391265869,
      "epoch": 1.6500362844702468,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047305947325839337,
      "loss": 5.2697,
      "mean_token_accuracy": 0.18233695179224013,
      "num_tokens": 37753317.0,
      "step": 18190
    },
    {
      "entropy": 5.6299241065979,
      "epoch": 1.6504898403483308,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004730441186188968,
      "loss": 5.2952,
      "mean_token_accuracy": 0.1811891257762909,
      "num_tokens": 37763452.0,
      "step": 18195
    },
    {
      "entropy": 5.571076822280884,
      "epoch": 1.650943396226415,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047302875988379734,
      "loss": 5.2246,
      "mean_token_accuracy": 0.18177709132432937,
      "num_tokens": 37772804.0,
      "step": 18200
    },
    {
      "entropy": 5.594820785522461,
      "epoch": 1.6513969521044993,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047301339705341266,
      "loss": 5.3076,
      "mean_token_accuracy": 0.17330503314733506,
      "num_tokens": 37784441.0,
      "step": 18205
    },
    {
      "entropy": 5.586144876480103,
      "epoch": 1.6518505079825836,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047299803012806056,
      "loss": 5.3179,
      "mean_token_accuracy": 0.17605923414230346,
      "num_tokens": 37795103.0,
      "step": 18210
    },
    {
      "entropy": 5.524984836578369,
      "epoch": 1.6523040638606676,
      "grad_norm": 3.90625,
      "learning_rate": 0.0004729826591080586,
      "loss": 5.3086,
      "mean_token_accuracy": 0.1751462921500206,
      "num_tokens": 37806654.0,
      "step": 18215
    },
    {
      "entropy": 5.582616853713989,
      "epoch": 1.6527576197387517,
      "grad_norm": 2.328125,
      "learning_rate": 0.00047296728399372483,
      "loss": 5.3085,
      "mean_token_accuracy": 0.18192391246557235,
      "num_tokens": 37816673.0,
      "step": 18220
    },
    {
      "entropy": 5.57633228302002,
      "epoch": 1.653211175616836,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047295190478537725,
      "loss": 5.3142,
      "mean_token_accuracy": 0.17700045853853225,
      "num_tokens": 37827389.0,
      "step": 18225
    },
    {
      "entropy": 5.638291645050049,
      "epoch": 1.6536647314949202,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047293652148333376,
      "loss": 5.3156,
      "mean_token_accuracy": 0.18137261420488357,
      "num_tokens": 37838087.0,
      "step": 18230
    },
    {
      "entropy": 5.471645641326904,
      "epoch": 1.6541182873730045,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047292113408791266,
      "loss": 5.1993,
      "mean_token_accuracy": 0.18105009347200393,
      "num_tokens": 37848630.0,
      "step": 18235
    },
    {
      "entropy": 5.4655442237854,
      "epoch": 1.6545718432510885,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00047290574259943195,
      "loss": 5.1808,
      "mean_token_accuracy": 0.18966717422008514,
      "num_tokens": 37858689.0,
      "step": 18240
    },
    {
      "entropy": 5.6084226131439205,
      "epoch": 1.6550253991291726,
      "grad_norm": 2.109375,
      "learning_rate": 0.00047289034701821013,
      "loss": 5.3403,
      "mean_token_accuracy": 0.1753557175397873,
      "num_tokens": 37869049.0,
      "step": 18245
    },
    {
      "entropy": 5.552996587753296,
      "epoch": 1.6554789550072568,
      "grad_norm": 9.875,
      "learning_rate": 0.00047287494734456545,
      "loss": 5.2567,
      "mean_token_accuracy": 0.1756283462047577,
      "num_tokens": 37879495.0,
      "step": 18250
    },
    {
      "entropy": 5.57558650970459,
      "epoch": 1.655932510885341,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047285954357881646,
      "loss": 5.3051,
      "mean_token_accuracy": 0.17791603207588197,
      "num_tokens": 37889442.0,
      "step": 18255
    },
    {
      "entropy": 5.643783712387085,
      "epoch": 1.6563860667634254,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004728441357212817,
      "loss": 5.4194,
      "mean_token_accuracy": 0.16828967183828353,
      "num_tokens": 37900147.0,
      "step": 18260
    },
    {
      "entropy": 5.64457893371582,
      "epoch": 1.6568396226415094,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004728287237722798,
      "loss": 5.2567,
      "mean_token_accuracy": 0.17521301060914993,
      "num_tokens": 37910307.0,
      "step": 18265
    },
    {
      "entropy": 5.5179225444793705,
      "epoch": 1.6572931785195935,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004728133077321295,
      "loss": 5.2236,
      "mean_token_accuracy": 0.1793442949652672,
      "num_tokens": 37920838.0,
      "step": 18270
    },
    {
      "entropy": 5.567683267593384,
      "epoch": 1.6577467343976777,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047279788760114954,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18116903752088548,
      "num_tokens": 37930717.0,
      "step": 18275
    },
    {
      "entropy": 5.62808256149292,
      "epoch": 1.658200290275762,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004727824633796589,
      "loss": 5.3177,
      "mean_token_accuracy": 0.17190706580877305,
      "num_tokens": 37941254.0,
      "step": 18280
    },
    {
      "entropy": 5.54310245513916,
      "epoch": 1.6586538461538463,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004727670350679765,
      "loss": 5.2962,
      "mean_token_accuracy": 0.184939706325531,
      "num_tokens": 37952132.0,
      "step": 18285
    },
    {
      "entropy": 5.533212327957154,
      "epoch": 1.6591074020319303,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004727516026664214,
      "loss": 5.2822,
      "mean_token_accuracy": 0.18514151871204376,
      "num_tokens": 37962026.0,
      "step": 18290
    },
    {
      "entropy": 5.599540662765503,
      "epoch": 1.6595609579100146,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004727361661753128,
      "loss": 5.2158,
      "mean_token_accuracy": 0.1775220274925232,
      "num_tokens": 37972911.0,
      "step": 18295
    },
    {
      "entropy": 5.613226413726807,
      "epoch": 1.6600145137880986,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004727207255949699,
      "loss": 5.2423,
      "mean_token_accuracy": 0.18424174338579177,
      "num_tokens": 37983577.0,
      "step": 18300
    },
    {
      "entropy": 5.576841115951538,
      "epoch": 1.6604680696661829,
      "grad_norm": 3.78125,
      "learning_rate": 0.00047270528092571197,
      "loss": 5.3496,
      "mean_token_accuracy": 0.17141730040311814,
      "num_tokens": 37994294.0,
      "step": 18305
    },
    {
      "entropy": 5.528322219848633,
      "epoch": 1.6609216255442671,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004726898321678584,
      "loss": 5.3082,
      "mean_token_accuracy": 0.1792794644832611,
      "num_tokens": 38005144.0,
      "step": 18310
    },
    {
      "entropy": 5.5663378715515135,
      "epoch": 1.6613751814223512,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004726743793217288,
      "loss": 5.3263,
      "mean_token_accuracy": 0.17675687968730927,
      "num_tokens": 38015277.0,
      "step": 18315
    },
    {
      "entropy": 5.615156793594361,
      "epoch": 1.6618287373004355,
      "grad_norm": 2.15625,
      "learning_rate": 0.00047265892238764265,
      "loss": 5.2468,
      "mean_token_accuracy": 0.1804247483611107,
      "num_tokens": 38027530.0,
      "step": 18320
    },
    {
      "entropy": 5.617325973510742,
      "epoch": 1.6622822931785195,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004726434613659196,
      "loss": 5.3483,
      "mean_token_accuracy": 0.17979378402233123,
      "num_tokens": 38036895.0,
      "step": 18325
    },
    {
      "entropy": 5.633412885665893,
      "epoch": 1.6627358490566038,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004726279962568794,
      "loss": 5.3769,
      "mean_token_accuracy": 0.1729107305407524,
      "num_tokens": 38047001.0,
      "step": 18330
    },
    {
      "entropy": 5.5717181205749515,
      "epoch": 1.663189404934688,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004726125270608419,
      "loss": 5.2579,
      "mean_token_accuracy": 0.18555023074150084,
      "num_tokens": 38056815.0,
      "step": 18335
    },
    {
      "entropy": 5.570738792419434,
      "epoch": 1.6636429608127723,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00047259705377812703,
      "loss": 5.3442,
      "mean_token_accuracy": 0.18098867386579515,
      "num_tokens": 38067139.0,
      "step": 18340
    },
    {
      "entropy": 5.574991464614868,
      "epoch": 1.6640965166908563,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004725815764090546,
      "loss": 5.256,
      "mean_token_accuracy": 0.18270086795091628,
      "num_tokens": 38076824.0,
      "step": 18345
    },
    {
      "entropy": 5.561831712722778,
      "epoch": 1.6645500725689404,
      "grad_norm": 1.703125,
      "learning_rate": 0.00047256609495394494,
      "loss": 5.2857,
      "mean_token_accuracy": 0.17946920841932296,
      "num_tokens": 38087517.0,
      "step": 18350
    },
    {
      "entropy": 5.608994722366333,
      "epoch": 1.6650036284470247,
      "grad_norm": 1.5,
      "learning_rate": 0.000472550609413118,
      "loss": 5.3047,
      "mean_token_accuracy": 0.17631544619798661,
      "num_tokens": 38098499.0,
      "step": 18355
    },
    {
      "entropy": 5.544527435302735,
      "epoch": 1.665457184325109,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004725351197868942,
      "loss": 5.2794,
      "mean_token_accuracy": 0.18367933481931686,
      "num_tokens": 38108327.0,
      "step": 18360
    },
    {
      "entropy": 5.661221170425415,
      "epoch": 1.6659107402031932,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00047251962607559373,
      "loss": 5.3021,
      "mean_token_accuracy": 0.17399619221687318,
      "num_tokens": 38118667.0,
      "step": 18365
    },
    {
      "entropy": 5.5493536472320555,
      "epoch": 1.6663642960812772,
      "grad_norm": 1.9375,
      "learning_rate": 0.00047250412827953717,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1796363115310669,
      "num_tokens": 38128571.0,
      "step": 18370
    },
    {
      "entropy": 5.559793758392334,
      "epoch": 1.6668178519593613,
      "grad_norm": 1.625,
      "learning_rate": 0.0004724886263990448,
      "loss": 5.3726,
      "mean_token_accuracy": 0.17228264659643172,
      "num_tokens": 38138921.0,
      "step": 18375
    },
    {
      "entropy": 5.620505905151367,
      "epoch": 1.6672714078374455,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004724731204344372,
      "loss": 5.3113,
      "mean_token_accuracy": 0.17792647927999497,
      "num_tokens": 38148858.0,
      "step": 18380
    },
    {
      "entropy": 5.505112886428833,
      "epoch": 1.6677249637155298,
      "grad_norm": 1.875,
      "learning_rate": 0.00047245761038603536,
      "loss": 5.2294,
      "mean_token_accuracy": 0.1830008417367935,
      "num_tokens": 38159107.0,
      "step": 18385
    },
    {
      "entropy": 5.65545072555542,
      "epoch": 1.668178519593614,
      "grad_norm": 2.03125,
      "learning_rate": 0.00047244209625415975,
      "loss": 5.355,
      "mean_token_accuracy": 0.17429672330617904,
      "num_tokens": 38169228.0,
      "step": 18390
    },
    {
      "entropy": 5.6219032287597654,
      "epoch": 1.6686320754716981,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004724265780391312,
      "loss": 5.3011,
      "mean_token_accuracy": 0.17925980687141418,
      "num_tokens": 38179175.0,
      "step": 18395
    },
    {
      "entropy": 5.560810899734497,
      "epoch": 1.6690856313497822,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004724110557412708,
      "loss": 5.3073,
      "mean_token_accuracy": 0.18161923736333846,
      "num_tokens": 38188537.0,
      "step": 18400
    },
    {
      "entropy": 5.577822256088257,
      "epoch": 1.6695391872278664,
      "grad_norm": 2.015625,
      "learning_rate": 0.00047239552936089943,
      "loss": 5.3058,
      "mean_token_accuracy": 0.17439125180244447,
      "num_tokens": 38198972.0,
      "step": 18405
    },
    {
      "entropy": 5.554353475570679,
      "epoch": 1.6699927431059507,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00047237999889833816,
      "loss": 5.1963,
      "mean_token_accuracy": 0.18053891509771347,
      "num_tokens": 38209472.0,
      "step": 18410
    },
    {
      "entropy": 5.41317310333252,
      "epoch": 1.670446298984035,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047236446435390826,
      "loss": 5.2108,
      "mean_token_accuracy": 0.18047342747449874,
      "num_tokens": 38219977.0,
      "step": 18415
    },
    {
      "entropy": 5.584135246276856,
      "epoch": 1.670899854862119,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047234892572793093,
      "loss": 5.2875,
      "mean_token_accuracy": 0.1785808429121971,
      "num_tokens": 38230492.0,
      "step": 18420
    },
    {
      "entropy": 5.575874185562133,
      "epoch": 1.671353410740203,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004723333830207275,
      "loss": 5.2545,
      "mean_token_accuracy": 0.17139156311750411,
      "num_tokens": 38240799.0,
      "step": 18425
    },
    {
      "entropy": 5.5515962600708,
      "epoch": 1.6718069666182873,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047231783623261954,
      "loss": 5.2473,
      "mean_token_accuracy": 0.17644179463386536,
      "num_tokens": 38251439.0,
      "step": 18430
    },
    {
      "entropy": 5.641415929794311,
      "epoch": 1.6722605224963716,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004723022853639283,
      "loss": 5.3796,
      "mean_token_accuracy": 0.17785269916057586,
      "num_tokens": 38261795.0,
      "step": 18435
    },
    {
      "entropy": 5.512445640563965,
      "epoch": 1.6727140783744558,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004722867304149756,
      "loss": 5.2422,
      "mean_token_accuracy": 0.18040314465761184,
      "num_tokens": 38272172.0,
      "step": 18440
    },
    {
      "entropy": 5.568269395828247,
      "epoch": 1.67316763425254,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004722711713860831,
      "loss": 5.352,
      "mean_token_accuracy": 0.16600459218025207,
      "num_tokens": 38282829.0,
      "step": 18445
    },
    {
      "entropy": 5.637834787368774,
      "epoch": 1.673621190130624,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004722556082775724,
      "loss": 5.3674,
      "mean_token_accuracy": 0.1798107236623764,
      "num_tokens": 38293793.0,
      "step": 18450
    },
    {
      "entropy": 5.637474393844604,
      "epoch": 1.6740747460087082,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004722400410897655,
      "loss": 5.3152,
      "mean_token_accuracy": 0.18018442541360855,
      "num_tokens": 38302577.0,
      "step": 18455
    },
    {
      "entropy": 5.622150182723999,
      "epoch": 1.6745283018867925,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004722244698229842,
      "loss": 5.2849,
      "mean_token_accuracy": 0.17926738709211348,
      "num_tokens": 38312571.0,
      "step": 18460
    },
    {
      "entropy": 5.5422343730926515,
      "epoch": 1.6749818577648767,
      "grad_norm": 1.890625,
      "learning_rate": 0.00047220889447755073,
      "loss": 5.2249,
      "mean_token_accuracy": 0.18368929773569107,
      "num_tokens": 38323544.0,
      "step": 18465
    },
    {
      "entropy": 5.596741151809693,
      "epoch": 1.6754354136429608,
      "grad_norm": 2.0,
      "learning_rate": 0.0004721933150537869,
      "loss": 5.3047,
      "mean_token_accuracy": 0.17704518139362335,
      "num_tokens": 38334509.0,
      "step": 18470
    },
    {
      "entropy": 5.51951265335083,
      "epoch": 1.675888969521045,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00047217773155201526,
      "loss": 5.2129,
      "mean_token_accuracy": 0.1830393850803375,
      "num_tokens": 38344747.0,
      "step": 18475
    },
    {
      "entropy": 5.596103239059448,
      "epoch": 1.676342525399129,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047216214397255776,
      "loss": 5.316,
      "mean_token_accuracy": 0.17654080241918563,
      "num_tokens": 38355671.0,
      "step": 18480
    },
    {
      "entropy": 5.622044658660888,
      "epoch": 1.6767960812772134,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047214655231573693,
      "loss": 5.3154,
      "mean_token_accuracy": 0.17879847884178163,
      "num_tokens": 38366368.0,
      "step": 18485
    },
    {
      "entropy": 5.606190395355225,
      "epoch": 1.6772496371552976,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004721309565818752,
      "loss": 5.3763,
      "mean_token_accuracy": 0.1776167184114456,
      "num_tokens": 38377322.0,
      "step": 18490
    },
    {
      "entropy": 5.597412872314453,
      "epoch": 1.6777031930333817,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00047211535677129496,
      "loss": 5.357,
      "mean_token_accuracy": 0.1756614163517952,
      "num_tokens": 38389204.0,
      "step": 18495
    },
    {
      "entropy": 5.583812570571899,
      "epoch": 1.678156748911466,
      "grad_norm": 2.03125,
      "learning_rate": 0.00047209975288431893,
      "loss": 5.3116,
      "mean_token_accuracy": 0.17512716203927994,
      "num_tokens": 38399117.0,
      "step": 18500
    },
    {
      "entropy": 5.5568798065185545,
      "epoch": 1.67861030478955,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047208414492126986,
      "loss": 5.2726,
      "mean_token_accuracy": 0.17218129336833954,
      "num_tokens": 38410341.0,
      "step": 18505
    },
    {
      "entropy": 5.610870599746704,
      "epoch": 1.6790638606676342,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047206853288247043,
      "loss": 5.3283,
      "mean_token_accuracy": 0.17318085581064224,
      "num_tokens": 38422742.0,
      "step": 18510
    },
    {
      "entropy": 5.588087177276611,
      "epoch": 1.6795174165457185,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00047205291676824345,
      "loss": 5.3458,
      "mean_token_accuracy": 0.17394431978464125,
      "num_tokens": 38433022.0,
      "step": 18515
    },
    {
      "entropy": 5.608362245559692,
      "epoch": 1.6799709724238028,
      "grad_norm": 1.609375,
      "learning_rate": 0.000472037296578912,
      "loss": 5.3292,
      "mean_token_accuracy": 0.1713524654507637,
      "num_tokens": 38444325.0,
      "step": 18520
    },
    {
      "entropy": 5.580276966094971,
      "epoch": 1.6804245283018868,
      "grad_norm": 1.671875,
      "learning_rate": 0.000472021672314799,
      "loss": 5.2275,
      "mean_token_accuracy": 0.18263774067163469,
      "num_tokens": 38454576.0,
      "step": 18525
    },
    {
      "entropy": 5.604031944274903,
      "epoch": 1.6808780841799709,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00047200604397622763,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18348122090101243,
      "num_tokens": 38465090.0,
      "step": 18530
    },
    {
      "entropy": 5.648733806610108,
      "epoch": 1.6813316400580551,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004719904115635211,
      "loss": 5.4666,
      "mean_token_accuracy": 0.1666980892419815,
      "num_tokens": 38477253.0,
      "step": 18535
    },
    {
      "entropy": 5.635500478744507,
      "epoch": 1.6817851959361394,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004719747750770026,
      "loss": 5.415,
      "mean_token_accuracy": 0.16887594908475875,
      "num_tokens": 38488040.0,
      "step": 18540
    },
    {
      "entropy": 5.577012395858764,
      "epoch": 1.6822387518142237,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004719591345169956,
      "loss": 5.2773,
      "mean_token_accuracy": 0.18415339142084122,
      "num_tokens": 38498016.0,
      "step": 18545
    },
    {
      "entropy": 5.6353143692016605,
      "epoch": 1.6826923076923077,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047194348988382345,
      "loss": 5.3736,
      "mean_token_accuracy": 0.1762053921818733,
      "num_tokens": 38508996.0,
      "step": 18550
    },
    {
      "entropy": 5.545180749893189,
      "epoch": 1.6831458635703918,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004719278411778097,
      "loss": 5.3379,
      "mean_token_accuracy": 0.1800803929567337,
      "num_tokens": 38519608.0,
      "step": 18555
    },
    {
      "entropy": 5.585157775878907,
      "epoch": 1.683599419448476,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047191218839927817,
      "loss": 5.2971,
      "mean_token_accuracy": 0.175361005961895,
      "num_tokens": 38530737.0,
      "step": 18560
    },
    {
      "entropy": 5.556194639205932,
      "epoch": 1.6840529753265603,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004718965315485523,
      "loss": 5.2265,
      "mean_token_accuracy": 0.1820900097489357,
      "num_tokens": 38540925.0,
      "step": 18565
    },
    {
      "entropy": 5.544345188140869,
      "epoch": 1.6845065312046446,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004718808706259561,
      "loss": 5.3206,
      "mean_token_accuracy": 0.17487624436616897,
      "num_tokens": 38553176.0,
      "step": 18570
    },
    {
      "entropy": 5.635834360122681,
      "epoch": 1.6849600870827286,
      "grad_norm": 2.109375,
      "learning_rate": 0.00047186520563181325,
      "loss": 5.3131,
      "mean_token_accuracy": 0.1755726858973503,
      "num_tokens": 38563394.0,
      "step": 18575
    },
    {
      "entropy": 5.652438545227051,
      "epoch": 1.6854136429608126,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004718495365664477,
      "loss": 5.3099,
      "mean_token_accuracy": 0.17899992763996125,
      "num_tokens": 38573243.0,
      "step": 18580
    },
    {
      "entropy": 5.568324232101441,
      "epoch": 1.685867198838897,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00047183386343018364,
      "loss": 5.3067,
      "mean_token_accuracy": 0.17865951210260392,
      "num_tokens": 38584014.0,
      "step": 18585
    },
    {
      "entropy": 5.561781787872315,
      "epoch": 1.6863207547169812,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004718181862233451,
      "loss": 5.2467,
      "mean_token_accuracy": 0.17909508794546128,
      "num_tokens": 38594467.0,
      "step": 18590
    },
    {
      "entropy": 5.625349330902099,
      "epoch": 1.6867743105950654,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004718025049462563,
      "loss": 5.3185,
      "mean_token_accuracy": 0.1729271039366722,
      "num_tokens": 38604407.0,
      "step": 18595
    },
    {
      "entropy": 5.630964612960815,
      "epoch": 1.6872278664731495,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004717868195992416,
      "loss": 5.3159,
      "mean_token_accuracy": 0.18313867300748826,
      "num_tokens": 38614508.0,
      "step": 18600
    },
    {
      "entropy": 5.630541133880615,
      "epoch": 1.6876814223512335,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004717711301826253,
      "loss": 5.3104,
      "mean_token_accuracy": 0.17731861919164657,
      "num_tokens": 38626244.0,
      "step": 18605
    },
    {
      "entropy": 5.524689149856568,
      "epoch": 1.6881349782293178,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047175543669673186,
      "loss": 5.2412,
      "mean_token_accuracy": 0.18064275234937668,
      "num_tokens": 38636694.0,
      "step": 18610
    },
    {
      "entropy": 5.649219942092896,
      "epoch": 1.688588534107402,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00047173973914188587,
      "loss": 5.3773,
      "mean_token_accuracy": 0.1762804090976715,
      "num_tokens": 38647414.0,
      "step": 18615
    },
    {
      "entropy": 5.586320209503174,
      "epoch": 1.6890420899854863,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047172403751841196,
      "loss": 5.2854,
      "mean_token_accuracy": 0.17749737203121185,
      "num_tokens": 38657425.0,
      "step": 18620
    },
    {
      "entropy": 5.611736297607422,
      "epoch": 1.6894956458635704,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047170833182663485,
      "loss": 5.3022,
      "mean_token_accuracy": 0.1797708287835121,
      "num_tokens": 38667751.0,
      "step": 18625
    },
    {
      "entropy": 5.496054935455322,
      "epoch": 1.6899492017416544,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004716926220668793,
      "loss": 5.2111,
      "mean_token_accuracy": 0.18331532031297684,
      "num_tokens": 38676918.0,
      "step": 18630
    },
    {
      "entropy": 5.543702459335327,
      "epoch": 1.6904027576197387,
      "grad_norm": 2.015625,
      "learning_rate": 0.00047167690823947015,
      "loss": 5.278,
      "mean_token_accuracy": 0.18322616815567017,
      "num_tokens": 38685887.0,
      "step": 18635
    },
    {
      "entropy": 5.551270389556885,
      "epoch": 1.690856313497823,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004716611903447325,
      "loss": 5.2901,
      "mean_token_accuracy": 0.18363554477691652,
      "num_tokens": 38696179.0,
      "step": 18640
    },
    {
      "entropy": 5.575514554977417,
      "epoch": 1.6913098693759072,
      "grad_norm": 1.625,
      "learning_rate": 0.0004716454683829912,
      "loss": 5.2549,
      "mean_token_accuracy": 0.176232248544693,
      "num_tokens": 38705777.0,
      "step": 18645
    },
    {
      "entropy": 5.5657590389251705,
      "epoch": 1.6917634252539913,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047162974235457163,
      "loss": 5.3018,
      "mean_token_accuracy": 0.18033342361450194,
      "num_tokens": 38717709.0,
      "step": 18650
    },
    {
      "entropy": 5.577129936218261,
      "epoch": 1.6922169811320755,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004716140122597988,
      "loss": 5.2465,
      "mean_token_accuracy": 0.1850181519985199,
      "num_tokens": 38729517.0,
      "step": 18655
    },
    {
      "entropy": 5.544319105148316,
      "epoch": 1.6926705370101596,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004715982780989981,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1807475909590721,
      "num_tokens": 38740878.0,
      "step": 18660
    },
    {
      "entropy": 5.527649021148681,
      "epoch": 1.6931240928882438,
      "grad_norm": 3.734375,
      "learning_rate": 0.00047158253987249487,
      "loss": 5.227,
      "mean_token_accuracy": 0.18430681377649308,
      "num_tokens": 38751291.0,
      "step": 18665
    },
    {
      "entropy": 5.478517150878906,
      "epoch": 1.693577648766328,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047156679758061465,
      "loss": 5.1944,
      "mean_token_accuracy": 0.19257359206676483,
      "num_tokens": 38760940.0,
      "step": 18670
    },
    {
      "entropy": 5.595829772949219,
      "epoch": 1.6940312046444121,
      "grad_norm": 2.0625,
      "learning_rate": 0.00047155105122368297,
      "loss": 5.2111,
      "mean_token_accuracy": 0.1829608976840973,
      "num_tokens": 38770240.0,
      "step": 18675
    },
    {
      "entropy": 5.494606685638428,
      "epoch": 1.6944847605224964,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004715353008020255,
      "loss": 5.2452,
      "mean_token_accuracy": 0.18168769031763077,
      "num_tokens": 38780452.0,
      "step": 18680
    },
    {
      "entropy": 5.566636085510254,
      "epoch": 1.6949383164005805,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047151954631596794,
      "loss": 5.3306,
      "mean_token_accuracy": 0.1782773658633232,
      "num_tokens": 38790763.0,
      "step": 18685
    },
    {
      "entropy": 5.546055841445923,
      "epoch": 1.6953918722786647,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047150378776583603,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18536781370639802,
      "num_tokens": 38802164.0,
      "step": 18690
    },
    {
      "entropy": 5.571982479095459,
      "epoch": 1.695845428156749,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004714880251519557,
      "loss": 5.2732,
      "mean_token_accuracy": 0.17710047215223312,
      "num_tokens": 38812901.0,
      "step": 18695
    },
    {
      "entropy": 5.574962520599366,
      "epoch": 1.6962989840348333,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000471472258474653,
      "loss": 5.2505,
      "mean_token_accuracy": 0.18173828423023225,
      "num_tokens": 38823305.0,
      "step": 18700
    },
    {
      "entropy": 5.549403953552246,
      "epoch": 1.6967525399129173,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047145648773425384,
      "loss": 5.2227,
      "mean_token_accuracy": 0.1767922043800354,
      "num_tokens": 38833602.0,
      "step": 18705
    },
    {
      "entropy": 5.5870680809021,
      "epoch": 1.6972060957910013,
      "grad_norm": 2.109375,
      "learning_rate": 0.00047144071293108456,
      "loss": 5.3336,
      "mean_token_accuracy": 0.17575362473726272,
      "num_tokens": 38843950.0,
      "step": 18710
    },
    {
      "entropy": 5.576045417785645,
      "epoch": 1.6976596516690856,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004714249340654712,
      "loss": 5.2218,
      "mean_token_accuracy": 0.1785681962966919,
      "num_tokens": 38853753.0,
      "step": 18715
    },
    {
      "entropy": 5.462394952774048,
      "epoch": 1.6981132075471699,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004714091511377402,
      "loss": 5.1915,
      "mean_token_accuracy": 0.17795278131961823,
      "num_tokens": 38863394.0,
      "step": 18720
    },
    {
      "entropy": 5.552939319610596,
      "epoch": 1.6985667634252541,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00047139336414821784,
      "loss": 5.2475,
      "mean_token_accuracy": 0.18154885470867158,
      "num_tokens": 38872972.0,
      "step": 18725
    },
    {
      "entropy": 5.578620958328247,
      "epoch": 1.6990203193033382,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047137757309723073,
      "loss": 5.2506,
      "mean_token_accuracy": 0.187425135076046,
      "num_tokens": 38883251.0,
      "step": 18730
    },
    {
      "entropy": 5.499541568756103,
      "epoch": 1.6994738751814222,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004713617779851053,
      "loss": 5.2706,
      "mean_token_accuracy": 0.178276988863945,
      "num_tokens": 38893274.0,
      "step": 18735
    },
    {
      "entropy": 5.598380041122437,
      "epoch": 1.6999274310595065,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004713459788121683,
      "loss": 5.371,
      "mean_token_accuracy": 0.1740770995616913,
      "num_tokens": 38903870.0,
      "step": 18740
    },
    {
      "entropy": 5.608860635757447,
      "epoch": 1.7003809869375908,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004713301755787464,
      "loss": 5.3201,
      "mean_token_accuracy": 0.1764300212264061,
      "num_tokens": 38914921.0,
      "step": 18745
    },
    {
      "entropy": 5.602755308151245,
      "epoch": 1.700834542815675,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004713143682851665,
      "loss": 5.2627,
      "mean_token_accuracy": 0.18174388259649277,
      "num_tokens": 38924630.0,
      "step": 18750
    },
    {
      "entropy": 5.52225661277771,
      "epoch": 1.701288098693759,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004712985569317553,
      "loss": 5.2547,
      "mean_token_accuracy": 0.1825944498181343,
      "num_tokens": 38934926.0,
      "step": 18755
    },
    {
      "entropy": 5.637522268295288,
      "epoch": 1.7017416545718431,
      "grad_norm": 1.625,
      "learning_rate": 0.0004712827415188401,
      "loss": 5.411,
      "mean_token_accuracy": 0.1700553700327873,
      "num_tokens": 38944513.0,
      "step": 18760
    },
    {
      "entropy": 5.5701672554016115,
      "epoch": 1.7021952104499274,
      "grad_norm": 1.625,
      "learning_rate": 0.0004712669220467476,
      "loss": 5.2662,
      "mean_token_accuracy": 0.1797216847538948,
      "num_tokens": 38955918.0,
      "step": 18765
    },
    {
      "entropy": 5.593360567092896,
      "epoch": 1.7026487663280117,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047125109851580533,
      "loss": 5.2938,
      "mean_token_accuracy": 0.1808386817574501,
      "num_tokens": 38966221.0,
      "step": 18770
    },
    {
      "entropy": 5.617777585983276,
      "epoch": 1.703102322206096,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047123527092634025,
      "loss": 5.2545,
      "mean_token_accuracy": 0.17875444442033767,
      "num_tokens": 38976463.0,
      "step": 18775
    },
    {
      "entropy": 5.505704116821289,
      "epoch": 1.70355587808418,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004712194392786797,
      "loss": 5.1754,
      "mean_token_accuracy": 0.18302649706602098,
      "num_tokens": 38986375.0,
      "step": 18780
    },
    {
      "entropy": 5.587542247772217,
      "epoch": 1.704009433962264,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047120360357315127,
      "loss": 5.3334,
      "mean_token_accuracy": 0.17832759916782379,
      "num_tokens": 38996116.0,
      "step": 18785
    },
    {
      "entropy": 5.574239587783813,
      "epoch": 1.7044629898403483,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047118776381008216,
      "loss": 5.279,
      "mean_token_accuracy": 0.18264166116714478,
      "num_tokens": 39005846.0,
      "step": 18790
    },
    {
      "entropy": 5.59389853477478,
      "epoch": 1.7049165457184325,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047117191998980026,
      "loss": 5.2791,
      "mean_token_accuracy": 0.17569634169340134,
      "num_tokens": 39016744.0,
      "step": 18795
    },
    {
      "entropy": 5.540518474578858,
      "epoch": 1.7053701015965168,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047115607211263295,
      "loss": 5.3385,
      "mean_token_accuracy": 0.1739113599061966,
      "num_tokens": 39026725.0,
      "step": 18800
    },
    {
      "entropy": 5.598899602890015,
      "epoch": 1.7058236574746009,
      "grad_norm": 2.0,
      "learning_rate": 0.00047114022017890814,
      "loss": 5.2759,
      "mean_token_accuracy": 0.17662841379642485,
      "num_tokens": 39036851.0,
      "step": 18805
    },
    {
      "entropy": 5.535115385055542,
      "epoch": 1.706277213352685,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004711243641889536,
      "loss": 5.2133,
      "mean_token_accuracy": 0.17613203078508377,
      "num_tokens": 39047179.0,
      "step": 18810
    },
    {
      "entropy": 5.497289657592773,
      "epoch": 1.7067307692307692,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00047110850414309714,
      "loss": 5.1307,
      "mean_token_accuracy": 0.19041862338781357,
      "num_tokens": 39057758.0,
      "step": 18815
    },
    {
      "entropy": 5.523637008666992,
      "epoch": 1.7071843251088534,
      "grad_norm": 2.59375,
      "learning_rate": 0.00047109264004166696,
      "loss": 5.3238,
      "mean_token_accuracy": 0.17678047418594361,
      "num_tokens": 39068142.0,
      "step": 18820
    },
    {
      "entropy": 5.561571788787842,
      "epoch": 1.7076378809869377,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004710767718849909,
      "loss": 5.2473,
      "mean_token_accuracy": 0.17571136802434922,
      "num_tokens": 39078157.0,
      "step": 18825
    },
    {
      "entropy": 5.5462456226348875,
      "epoch": 1.7080914368650217,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004710608996733973,
      "loss": 5.2893,
      "mean_token_accuracy": 0.1832023337483406,
      "num_tokens": 39089090.0,
      "step": 18830
    },
    {
      "entropy": 5.539453172683716,
      "epoch": 1.708544992743106,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00047104502340721433,
      "loss": 5.3249,
      "mean_token_accuracy": 0.17220360338687896,
      "num_tokens": 39099079.0,
      "step": 18835
    },
    {
      "entropy": 5.564136266708374,
      "epoch": 1.70899854862119,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047102914308677033,
      "loss": 5.2541,
      "mean_token_accuracy": 0.17446735203266145,
      "num_tokens": 39109927.0,
      "step": 18840
    },
    {
      "entropy": 5.586283254623413,
      "epoch": 1.7094521044992743,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00047101325871239363,
      "loss": 5.2559,
      "mean_token_accuracy": 0.1797412544488907,
      "num_tokens": 39119806.0,
      "step": 18845
    },
    {
      "entropy": 5.616036748886108,
      "epoch": 1.7099056603773586,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004709973702844128,
      "loss": 5.3459,
      "mean_token_accuracy": 0.17455461472272873,
      "num_tokens": 39130147.0,
      "step": 18850
    },
    {
      "entropy": 5.503326511383056,
      "epoch": 1.7103592162554426,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047098147780315645,
      "loss": 5.1874,
      "mean_token_accuracy": 0.1860576882958412,
      "num_tokens": 39139937.0,
      "step": 18855
    },
    {
      "entropy": 5.566011714935303,
      "epoch": 1.710812772133527,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004709655812689532,
      "loss": 5.2909,
      "mean_token_accuracy": 0.18068781346082688,
      "num_tokens": 39149977.0,
      "step": 18860
    },
    {
      "entropy": 5.528696298599243,
      "epoch": 1.711266328011611,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004709496806821318,
      "loss": 5.2908,
      "mean_token_accuracy": 0.1796919137239456,
      "num_tokens": 39161076.0,
      "step": 18865
    },
    {
      "entropy": 5.594974613189697,
      "epoch": 1.7117198838896952,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00047093377604302106,
      "loss": 5.3362,
      "mean_token_accuracy": 0.1726291596889496,
      "num_tokens": 39170776.0,
      "step": 18870
    },
    {
      "entropy": 5.575620126724243,
      "epoch": 1.7121734397677795,
      "grad_norm": 2.5,
      "learning_rate": 0.0004709178673519499,
      "loss": 5.289,
      "mean_token_accuracy": 0.1782035067677498,
      "num_tokens": 39181134.0,
      "step": 18875
    },
    {
      "entropy": 5.568337106704712,
      "epoch": 1.7126269956458637,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004709019546092473,
      "loss": 5.3753,
      "mean_token_accuracy": 0.1725681707262993,
      "num_tokens": 39193331.0,
      "step": 18880
    },
    {
      "entropy": 5.663220977783203,
      "epoch": 1.7130805515239478,
      "grad_norm": 1.8125,
      "learning_rate": 0.00047088603781524237,
      "loss": 5.3428,
      "mean_token_accuracy": 0.17419660091400146,
      "num_tokens": 39204372.0,
      "step": 18885
    },
    {
      "entropy": 5.560180521011352,
      "epoch": 1.7135341074020318,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004708701169702642,
      "loss": 5.2583,
      "mean_token_accuracy": 0.18472733050584794,
      "num_tokens": 39214725.0,
      "step": 18890
    },
    {
      "entropy": 5.462447023391723,
      "epoch": 1.713987663280116,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004708541920746422,
      "loss": 5.2319,
      "mean_token_accuracy": 0.18085746467113495,
      "num_tokens": 39225119.0,
      "step": 18895
    },
    {
      "entropy": 5.557621812820434,
      "epoch": 1.7144412191582004,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004708382631287055,
      "loss": 5.2181,
      "mean_token_accuracy": 0.17973356395959855,
      "num_tokens": 39235021.0,
      "step": 18900
    },
    {
      "entropy": 5.605642986297608,
      "epoch": 1.7148947750362846,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004708223301327837,
      "loss": 5.3885,
      "mean_token_accuracy": 0.17347122579813004,
      "num_tokens": 39245830.0,
      "step": 18905
    },
    {
      "entropy": 5.624499320983887,
      "epoch": 1.7153483309143687,
      "grad_norm": 1.796875,
      "learning_rate": 0.00047080639308720617,
      "loss": 5.3615,
      "mean_token_accuracy": 0.18051055818796158,
      "num_tokens": 39255264.0,
      "step": 18910
    },
    {
      "entropy": 5.602855014801025,
      "epoch": 1.7158018867924527,
      "grad_norm": 1.734375,
      "learning_rate": 0.00047079045199230254,
      "loss": 5.2388,
      "mean_token_accuracy": 0.18300697803497315,
      "num_tokens": 39264966.0,
      "step": 18915
    },
    {
      "entropy": 5.54524245262146,
      "epoch": 1.716255442670537,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047077450684840247,
      "loss": 5.3462,
      "mean_token_accuracy": 0.17479234635829927,
      "num_tokens": 39275082.0,
      "step": 18920
    },
    {
      "entropy": 5.5645805358886715,
      "epoch": 1.7167089985486212,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004707585576558357,
      "loss": 5.4145,
      "mean_token_accuracy": 0.16926276087760925,
      "num_tokens": 39285426.0,
      "step": 18925
    },
    {
      "entropy": 5.703546857833862,
      "epoch": 1.7171625544267055,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004707426044149321,
      "loss": 5.3645,
      "mean_token_accuracy": 0.17415008395910264,
      "num_tokens": 39296633.0,
      "step": 18930
    },
    {
      "entropy": 5.556660509109497,
      "epoch": 1.7176161103047896,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00047072664712602153,
      "loss": 5.1685,
      "mean_token_accuracy": 0.19140244424343109,
      "num_tokens": 39306775.0,
      "step": 18935
    },
    {
      "entropy": 5.492854309082031,
      "epoch": 1.7180696661828736,
      "grad_norm": 1.953125,
      "learning_rate": 0.000470710685789434,
      "loss": 5.2975,
      "mean_token_accuracy": 0.17646047919988633,
      "num_tokens": 39317917.0,
      "step": 18940
    },
    {
      "entropy": 5.578994560241699,
      "epoch": 1.7185232220609579,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047069472040549965,
      "loss": 5.2931,
      "mean_token_accuracy": 0.17650235146284105,
      "num_tokens": 39328387.0,
      "step": 18945
    },
    {
      "entropy": 5.48938136100769,
      "epoch": 1.7189767779390421,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004706787509745486,
      "loss": 5.1453,
      "mean_token_accuracy": 0.19174888879060745,
      "num_tokens": 39337639.0,
      "step": 18950
    },
    {
      "entropy": 5.520390748977661,
      "epoch": 1.7194303338171264,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047066277749691115,
      "loss": 5.2971,
      "mean_token_accuracy": 0.1763899177312851,
      "num_tokens": 39347971.0,
      "step": 18955
    },
    {
      "entropy": 5.607920837402344,
      "epoch": 1.7198838896952104,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004706467999729175,
      "loss": 5.2964,
      "mean_token_accuracy": 0.17695277333259582,
      "num_tokens": 39358038.0,
      "step": 18960
    },
    {
      "entropy": 5.605505084991455,
      "epoch": 1.7203374455732945,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004706308184028983,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1813315063714981,
      "num_tokens": 39367834.0,
      "step": 18965
    },
    {
      "entropy": 5.586561489105224,
      "epoch": 1.7207910014513788,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004706148327871839,
      "loss": 5.3062,
      "mean_token_accuracy": 0.18175091296434404,
      "num_tokens": 39378924.0,
      "step": 18970
    },
    {
      "entropy": 5.650143194198608,
      "epoch": 1.721244557329463,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047059884312610487,
      "loss": 5.4039,
      "mean_token_accuracy": 0.1711315095424652,
      "num_tokens": 39390082.0,
      "step": 18975
    },
    {
      "entropy": 5.575082778930664,
      "epoch": 1.7216981132075473,
      "grad_norm": 1.75,
      "learning_rate": 0.00047058284941999185,
      "loss": 5.3163,
      "mean_token_accuracy": 0.18062724769115449,
      "num_tokens": 39400485.0,
      "step": 18980
    },
    {
      "entropy": 5.607129669189453,
      "epoch": 1.7221516690856313,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004705668516691758,
      "loss": 5.3223,
      "mean_token_accuracy": 0.18061004877090453,
      "num_tokens": 39411601.0,
      "step": 18985
    },
    {
      "entropy": 5.583995151519775,
      "epoch": 1.7226052249637154,
      "grad_norm": 1.8125,
      "learning_rate": 0.00047055084987398735,
      "loss": 5.2127,
      "mean_token_accuracy": 0.18552757203578948,
      "num_tokens": 39421625.0,
      "step": 18990
    },
    {
      "entropy": 5.549121522903443,
      "epoch": 1.7230587808417996,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004705348440347575,
      "loss": 5.2895,
      "mean_token_accuracy": 0.1728318840265274,
      "num_tokens": 39432033.0,
      "step": 18995
    },
    {
      "entropy": 5.5157335758209225,
      "epoch": 1.723512336719884,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047051883415181724,
      "loss": 5.2475,
      "mean_token_accuracy": 0.17824530601501465,
      "num_tokens": 39442514.0,
      "step": 19000
    },
    {
      "entropy": 5.543750953674317,
      "epoch": 1.7239658925979682,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004705028202254977,
      "loss": 5.1482,
      "mean_token_accuracy": 0.18811102509498595,
      "num_tokens": 39452497.0,
      "step": 19005
    },
    {
      "entropy": 5.537263345718384,
      "epoch": 1.7244194484760522,
      "grad_norm": 1.734375,
      "learning_rate": 0.00047048680225612996,
      "loss": 5.2789,
      "mean_token_accuracy": 0.1814630761742592,
      "num_tokens": 39462173.0,
      "step": 19010
    },
    {
      "entropy": 5.581815147399903,
      "epoch": 1.7248730043541363,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004704707802440454,
      "loss": 5.3394,
      "mean_token_accuracy": 0.1827238157391548,
      "num_tokens": 39473123.0,
      "step": 19015
    },
    {
      "entropy": 5.563253545761109,
      "epoch": 1.7253265602322205,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00047045475418957524,
      "loss": 5.3055,
      "mean_token_accuracy": 0.16820675134658813,
      "num_tokens": 39484142.0,
      "step": 19020
    },
    {
      "entropy": 5.577175426483154,
      "epoch": 1.7257801161103048,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004704387240930509,
      "loss": 5.3061,
      "mean_token_accuracy": 0.1735610067844391,
      "num_tokens": 39494926.0,
      "step": 19025
    },
    {
      "entropy": 5.574322891235352,
      "epoch": 1.726233671988389,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004704226899548041,
      "loss": 5.255,
      "mean_token_accuracy": 0.17994279861450196,
      "num_tokens": 39505153.0,
      "step": 19030
    },
    {
      "entropy": 5.474737215042114,
      "epoch": 1.726687227866473,
      "grad_norm": 1.734375,
      "learning_rate": 0.00047040665177516616,
      "loss": 5.2334,
      "mean_token_accuracy": 0.1800874724984169,
      "num_tokens": 39515905.0,
      "step": 19035
    },
    {
      "entropy": 5.5191967487335205,
      "epoch": 1.7271407837445574,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00047039060955446886,
      "loss": 5.2468,
      "mean_token_accuracy": 0.1814584270119667,
      "num_tokens": 39526762.0,
      "step": 19040
    },
    {
      "entropy": 5.530157899856567,
      "epoch": 1.7275943396226414,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000470374563293044,
      "loss": 5.2366,
      "mean_token_accuracy": 0.18080762326717376,
      "num_tokens": 39537340.0,
      "step": 19045
    },
    {
      "entropy": 5.556510877609253,
      "epoch": 1.7280478955007257,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004703585129912234,
      "loss": 5.2555,
      "mean_token_accuracy": 0.17898902893066407,
      "num_tokens": 39547144.0,
      "step": 19050
    },
    {
      "entropy": 5.668293237686157,
      "epoch": 1.72850145137881,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047034245864933894,
      "loss": 5.3842,
      "mean_token_accuracy": 0.17208921015262604,
      "num_tokens": 39557707.0,
      "step": 19055
    },
    {
      "entropy": 5.625163841247558,
      "epoch": 1.728955007256894,
      "grad_norm": 2.1875,
      "learning_rate": 0.00047032640026772274,
      "loss": 5.3333,
      "mean_token_accuracy": 0.17836526334285735,
      "num_tokens": 39567629.0,
      "step": 19060
    },
    {
      "entropy": 5.538285589218139,
      "epoch": 1.7294085631349783,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004703103378467067,
      "loss": 5.2599,
      "mean_token_accuracy": 0.17980674058198928,
      "num_tokens": 39577918.0,
      "step": 19065
    },
    {
      "entropy": 5.554657220840454,
      "epoch": 1.7298621190130623,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047029427138662314,
      "loss": 5.3085,
      "mean_token_accuracy": 0.1831940859556198,
      "num_tokens": 39587838.0,
      "step": 19070
    },
    {
      "entropy": 5.541716909408569,
      "epoch": 1.7303156748911466,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004702782008878042,
      "loss": 5.1667,
      "mean_token_accuracy": 0.18785454332828522,
      "num_tokens": 39597353.0,
      "step": 19075
    },
    {
      "entropy": 5.515530300140381,
      "epoch": 1.7307692307692308,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004702621263505824,
      "loss": 5.2166,
      "mean_token_accuracy": 0.18336535394191741,
      "num_tokens": 39607349.0,
      "step": 19080
    },
    {
      "entropy": 5.595157766342163,
      "epoch": 1.731222786647315,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004702460477752901,
      "loss": 5.3077,
      "mean_token_accuracy": 0.17635563910007476,
      "num_tokens": 39617810.0,
      "step": 19085
    },
    {
      "entropy": 5.562395668029785,
      "epoch": 1.7316763425253991,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004702299651622597,
      "loss": 5.1939,
      "mean_token_accuracy": 0.18990109115839005,
      "num_tokens": 39626805.0,
      "step": 19090
    },
    {
      "entropy": 5.485794019699097,
      "epoch": 1.7321298984034832,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004702138785118239,
      "loss": 5.2392,
      "mean_token_accuracy": 0.1839746430516243,
      "num_tokens": 39637839.0,
      "step": 19095
    },
    {
      "entropy": 5.5044605255126955,
      "epoch": 1.7325834542815675,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004701977878243153,
      "loss": 5.2107,
      "mean_token_accuracy": 0.18512007743120193,
      "num_tokens": 39648697.0,
      "step": 19100
    },
    {
      "entropy": 5.511010646820068,
      "epoch": 1.7330370101596517,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004701816931000668,
      "loss": 5.2329,
      "mean_token_accuracy": 0.18457728028297424,
      "num_tokens": 39658154.0,
      "step": 19105
    },
    {
      "entropy": 5.56726713180542,
      "epoch": 1.733490566037736,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047016559433941106,
      "loss": 5.2851,
      "mean_token_accuracy": 0.17813993692398072,
      "num_tokens": 39667939.0,
      "step": 19110
    },
    {
      "entropy": 5.4743834972381595,
      "epoch": 1.73394412191582,
      "grad_norm": 1.375,
      "learning_rate": 0.0004701494915426812,
      "loss": 5.1854,
      "mean_token_accuracy": 0.18131975531578065,
      "num_tokens": 39678289.0,
      "step": 19115
    },
    {
      "entropy": 5.5168215274810795,
      "epoch": 1.734397677793904,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047013338471021,
      "loss": 5.245,
      "mean_token_accuracy": 0.18065883070230485,
      "num_tokens": 39689901.0,
      "step": 19120
    },
    {
      "entropy": 5.5856832504272464,
      "epoch": 1.7348512336719883,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047011727384233074,
      "loss": 5.2835,
      "mean_token_accuracy": 0.17925552874803544,
      "num_tokens": 39700831.0,
      "step": 19125
    },
    {
      "entropy": 5.596054220199585,
      "epoch": 1.7353047895500726,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004701011589393765,
      "loss": 5.307,
      "mean_token_accuracy": 0.1762596920132637,
      "num_tokens": 39711794.0,
      "step": 19130
    },
    {
      "entropy": 5.545517158508301,
      "epoch": 1.7357583454281569,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047008504000168054,
      "loss": 5.245,
      "mean_token_accuracy": 0.17974234074354173,
      "num_tokens": 39721889.0,
      "step": 19135
    },
    {
      "entropy": 5.509635591506958,
      "epoch": 1.736211901306241,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004700689170295763,
      "loss": 5.338,
      "mean_token_accuracy": 0.17838057428598403,
      "num_tokens": 39732562.0,
      "step": 19140
    },
    {
      "entropy": 5.649844360351563,
      "epoch": 1.736665457184325,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047005279002339717,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18086100220680237,
      "num_tokens": 39742286.0,
      "step": 19145
    },
    {
      "entropy": 5.609019660949707,
      "epoch": 1.7371190130624092,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047003665898347663,
      "loss": 5.2915,
      "mean_token_accuracy": 0.18152936697006225,
      "num_tokens": 39754336.0,
      "step": 19150
    },
    {
      "entropy": 5.514881753921509,
      "epoch": 1.7375725689404935,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004700205239101483,
      "loss": 5.1798,
      "mean_token_accuracy": 0.18925442844629287,
      "num_tokens": 39763545.0,
      "step": 19155
    },
    {
      "entropy": 5.6054558753967285,
      "epoch": 1.7380261248185778,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004700043848037458,
      "loss": 5.3292,
      "mean_token_accuracy": 0.17266523838043213,
      "num_tokens": 39774536.0,
      "step": 19160
    },
    {
      "entropy": 5.589639568328858,
      "epoch": 1.7384796806966618,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004699882416646029,
      "loss": 5.2683,
      "mean_token_accuracy": 0.18399334698915482,
      "num_tokens": 39785377.0,
      "step": 19165
    },
    {
      "entropy": 5.545966911315918,
      "epoch": 1.7389332365747459,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004699720944930535,
      "loss": 5.3546,
      "mean_token_accuracy": 0.16535880267620087,
      "num_tokens": 39796692.0,
      "step": 19170
    },
    {
      "entropy": 5.505479621887207,
      "epoch": 1.7393867924528301,
      "grad_norm": 2.15625,
      "learning_rate": 0.00046995594328943147,
      "loss": 5.2637,
      "mean_token_accuracy": 0.18172186464071274,
      "num_tokens": 39807927.0,
      "step": 19175
    },
    {
      "entropy": 5.682715892791748,
      "epoch": 1.7398403483309144,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004699397880540709,
      "loss": 5.4467,
      "mean_token_accuracy": 0.1752352386713028,
      "num_tokens": 39817617.0,
      "step": 19180
    },
    {
      "entropy": 5.569416284561157,
      "epoch": 1.7402939042089987,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004699236287873058,
      "loss": 5.1626,
      "mean_token_accuracy": 0.1836479425430298,
      "num_tokens": 39827092.0,
      "step": 19185
    },
    {
      "entropy": 5.480993032455444,
      "epoch": 1.7407474600870827,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004699074654894705,
      "loss": 5.2234,
      "mean_token_accuracy": 0.1864486739039421,
      "num_tokens": 39837378.0,
      "step": 19190
    },
    {
      "entropy": 5.538658571243286,
      "epoch": 1.7412010159651667,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000469891298160899,
      "loss": 5.2385,
      "mean_token_accuracy": 0.17462693005800248,
      "num_tokens": 39848191.0,
      "step": 19195
    },
    {
      "entropy": 5.51838960647583,
      "epoch": 1.741654571843251,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046987512680192586,
      "loss": 5.2255,
      "mean_token_accuracy": 0.18287772089242935,
      "num_tokens": 39857886.0,
      "step": 19200
    },
    {
      "entropy": 5.615221357345581,
      "epoch": 1.7421081277213353,
      "grad_norm": 1.578125,
      "learning_rate": 0.00046985895141288546,
      "loss": 5.3315,
      "mean_token_accuracy": 0.17384516149759294,
      "num_tokens": 39868348.0,
      "step": 19205
    },
    {
      "entropy": 5.611031532287598,
      "epoch": 1.7425616835994195,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046984277199411224,
      "loss": 5.3426,
      "mean_token_accuracy": 0.18068430721759796,
      "num_tokens": 39879129.0,
      "step": 19210
    },
    {
      "entropy": 5.531503677368164,
      "epoch": 1.7430152394775036,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00046982658854594083,
      "loss": 5.2345,
      "mean_token_accuracy": 0.17873648405075074,
      "num_tokens": 39890239.0,
      "step": 19215
    },
    {
      "entropy": 5.509018278121948,
      "epoch": 1.7434687953555879,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046981040106870595,
      "loss": 5.1389,
      "mean_token_accuracy": 0.19047822803258896,
      "num_tokens": 39900907.0,
      "step": 19220
    },
    {
      "entropy": 5.570319557189942,
      "epoch": 1.743922351233672,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004697942095627423,
      "loss": 5.2691,
      "mean_token_accuracy": 0.18196940571069717,
      "num_tokens": 39911564.0,
      "step": 19225
    },
    {
      "entropy": 5.546303558349609,
      "epoch": 1.7443759071117562,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004697780140283848,
      "loss": 5.2055,
      "mean_token_accuracy": 0.18187068700790404,
      "num_tokens": 39922168.0,
      "step": 19230
    },
    {
      "entropy": 5.547090148925781,
      "epoch": 1.7448294629898404,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004697618144659683,
      "loss": 5.2505,
      "mean_token_accuracy": 0.18501388430595397,
      "num_tokens": 39932690.0,
      "step": 19235
    },
    {
      "entropy": 5.631829357147216,
      "epoch": 1.7452830188679245,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004697456108758278,
      "loss": 5.3913,
      "mean_token_accuracy": 0.17561757266521455,
      "num_tokens": 39942970.0,
      "step": 19240
    },
    {
      "entropy": 5.5298703670501705,
      "epoch": 1.7457365747460087,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004697294032582986,
      "loss": 5.229,
      "mean_token_accuracy": 0.18262505978345872,
      "num_tokens": 39953255.0,
      "step": 19245
    },
    {
      "entropy": 5.536805582046509,
      "epoch": 1.7461901306240928,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00046971319161371564,
      "loss": 5.1759,
      "mean_token_accuracy": 0.1837702512741089,
      "num_tokens": 39963521.0,
      "step": 19250
    },
    {
      "entropy": 5.554080057144165,
      "epoch": 1.746643686502177,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046969697594241426,
      "loss": 5.2427,
      "mean_token_accuracy": 0.180544213950634,
      "num_tokens": 39973494.0,
      "step": 19255
    },
    {
      "entropy": 5.560983657836914,
      "epoch": 1.7470972423802613,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004696807562447298,
      "loss": 5.3885,
      "mean_token_accuracy": 0.17059191316366196,
      "num_tokens": 39984652.0,
      "step": 19260
    },
    {
      "entropy": 5.61394476890564,
      "epoch": 1.7475507982583456,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00046966453252099763,
      "loss": 5.3149,
      "mean_token_accuracy": 0.18026408851146697,
      "num_tokens": 39995490.0,
      "step": 19265
    },
    {
      "entropy": 5.508791112899781,
      "epoch": 1.7480043541364296,
      "grad_norm": 1.9375,
      "learning_rate": 0.00046964830477155335,
      "loss": 5.1006,
      "mean_token_accuracy": 0.2010169193148613,
      "num_tokens": 40006447.0,
      "step": 19270
    },
    {
      "entropy": 5.5989467144012455,
      "epoch": 1.7484579100145137,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046963207299673257,
      "loss": 5.3439,
      "mean_token_accuracy": 0.1783621534705162,
      "num_tokens": 40015733.0,
      "step": 19275
    },
    {
      "entropy": 5.601906251907349,
      "epoch": 1.748911465892598,
      "grad_norm": 2.296875,
      "learning_rate": 0.00046961583719687085,
      "loss": 5.211,
      "mean_token_accuracy": 0.18360452204942704,
      "num_tokens": 40025844.0,
      "step": 19280
    },
    {
      "entropy": 5.554730749130249,
      "epoch": 1.7493650217706822,
      "grad_norm": 1.71875,
      "learning_rate": 0.00046959959737230414,
      "loss": 5.2601,
      "mean_token_accuracy": 0.18667895942926407,
      "num_tokens": 40035942.0,
      "step": 19285
    },
    {
      "entropy": 5.49570746421814,
      "epoch": 1.7498185776487665,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046958335352336806,
      "loss": 5.1496,
      "mean_token_accuracy": 0.19050020724534988,
      "num_tokens": 40045551.0,
      "step": 19290
    },
    {
      "entropy": 5.550003147125244,
      "epoch": 1.7502721335268505,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004695671056503987,
      "loss": 5.2669,
      "mean_token_accuracy": 0.18744464814662934,
      "num_tokens": 40055803.0,
      "step": 19295
    },
    {
      "entropy": 5.479763746261597,
      "epoch": 1.7507256894049346,
      "grad_norm": 1.9140625,
      "learning_rate": 0.000469550853753732,
      "loss": 5.2512,
      "mean_token_accuracy": 0.17667979449033738,
      "num_tokens": 40066004.0,
      "step": 19300
    },
    {
      "entropy": 5.552136325836182,
      "epoch": 1.7511792452830188,
      "grad_norm": 1.375,
      "learning_rate": 0.000469534597833704,
      "loss": 5.252,
      "mean_token_accuracy": 0.18481021374464035,
      "num_tokens": 40075522.0,
      "step": 19305
    },
    {
      "entropy": 5.572956705093384,
      "epoch": 1.751632801161103,
      "grad_norm": 1.59375,
      "learning_rate": 0.000469518337890651,
      "loss": 5.2481,
      "mean_token_accuracy": 0.19009540379047393,
      "num_tokens": 40084839.0,
      "step": 19310
    },
    {
      "entropy": 5.589095878601074,
      "epoch": 1.7520863570391874,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046950207392490917,
      "loss": 5.2827,
      "mean_token_accuracy": 0.17998334169387817,
      "num_tokens": 40094392.0,
      "step": 19315
    },
    {
      "entropy": 5.4875177383422855,
      "epoch": 1.7525399129172714,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00046948580593681496,
      "loss": 5.2341,
      "mean_token_accuracy": 0.18356212824583054,
      "num_tokens": 40105990.0,
      "step": 19320
    },
    {
      "entropy": 5.467586660385132,
      "epoch": 1.7529934687953554,
      "grad_norm": 1.625,
      "learning_rate": 0.0004694695339267047,
      "loss": 5.1446,
      "mean_token_accuracy": 0.187503619492054,
      "num_tokens": 40116143.0,
      "step": 19325
    },
    {
      "entropy": 5.625067710876465,
      "epoch": 1.7534470246734397,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00046945325789491483,
      "loss": 5.3358,
      "mean_token_accuracy": 0.17721422910690307,
      "num_tokens": 40126849.0,
      "step": 19330
    },
    {
      "entropy": 5.541573524475098,
      "epoch": 1.753900580551524,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004694369778417822,
      "loss": 5.3077,
      "mean_token_accuracy": 0.17925147116184234,
      "num_tokens": 40137259.0,
      "step": 19335
    },
    {
      "entropy": 5.47883563041687,
      "epoch": 1.7543541364296082,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004694206937676431,
      "loss": 5.1569,
      "mean_token_accuracy": 0.18007537871599197,
      "num_tokens": 40148578.0,
      "step": 19340
    },
    {
      "entropy": 5.564558601379394,
      "epoch": 1.7548076923076923,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004694044056728347,
      "loss": 5.3145,
      "mean_token_accuracy": 0.1732069283723831,
      "num_tokens": 40159754.0,
      "step": 19345
    },
    {
      "entropy": 5.594398355484008,
      "epoch": 1.7552612481857763,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046938811355769363,
      "loss": 5.2795,
      "mean_token_accuracy": 0.1757770761847496,
      "num_tokens": 40169388.0,
      "step": 19350
    },
    {
      "entropy": 5.519630813598633,
      "epoch": 1.7557148040638606,
      "grad_norm": 2.09375,
      "learning_rate": 0.00046937181742255673,
      "loss": 5.2793,
      "mean_token_accuracy": 0.17731226980686188,
      "num_tokens": 40179271.0,
      "step": 19355
    },
    {
      "entropy": 5.506827116012573,
      "epoch": 1.7561683599419449,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004693555172677612,
      "loss": 5.2943,
      "mean_token_accuracy": 0.17774943709373475,
      "num_tokens": 40189292.0,
      "step": 19360
    },
    {
      "entropy": 5.557961082458496,
      "epoch": 1.7566219158200291,
      "grad_norm": 1.59375,
      "learning_rate": 0.000469339213093644,
      "loss": 5.2624,
      "mean_token_accuracy": 0.1828974589705467,
      "num_tokens": 40199271.0,
      "step": 19365
    },
    {
      "entropy": 5.560376262664795,
      "epoch": 1.7570754716981132,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046932290490054235,
      "loss": 5.3034,
      "mean_token_accuracy": 0.17741362750530243,
      "num_tokens": 40210386.0,
      "step": 19370
    },
    {
      "entropy": 5.583480978012085,
      "epoch": 1.7575290275761972,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004693065926887936,
      "loss": 5.2502,
      "mean_token_accuracy": 0.18022023737430573,
      "num_tokens": 40220892.0,
      "step": 19375
    },
    {
      "entropy": 5.593159770965576,
      "epoch": 1.7579825834542815,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00046929027645873484,
      "loss": 5.2671,
      "mean_token_accuracy": 0.17794726490974427,
      "num_tokens": 40231737.0,
      "step": 19380
    },
    {
      "entropy": 5.537965583801269,
      "epoch": 1.7584361393323658,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004692739562107038,
      "loss": 5.3114,
      "mean_token_accuracy": 0.17342878878116608,
      "num_tokens": 40242218.0,
      "step": 19385
    },
    {
      "entropy": 5.532397747039795,
      "epoch": 1.75888969521045,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046925763194503775,
      "loss": 5.252,
      "mean_token_accuracy": 0.18378586024045945,
      "num_tokens": 40253659.0,
      "step": 19390
    },
    {
      "entropy": 5.511300277709961,
      "epoch": 1.759343251088534,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004692413036620744,
      "loss": 5.1221,
      "mean_token_accuracy": 0.18775354623794555,
      "num_tokens": 40264328.0,
      "step": 19395
    },
    {
      "entropy": 5.456592988967896,
      "epoch": 1.7597968069666183,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004692249713621514,
      "loss": 5.2254,
      "mean_token_accuracy": 0.18251916617155076,
      "num_tokens": 40275188.0,
      "step": 19400
    },
    {
      "entropy": 5.5984574317932125,
      "epoch": 1.7602503628447024,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046920863504560647,
      "loss": 5.4024,
      "mean_token_accuracy": 0.17412373274564744,
      "num_tokens": 40287094.0,
      "step": 19405
    },
    {
      "entropy": 5.633415842056275,
      "epoch": 1.7607039187227866,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004691922947127775,
      "loss": 5.2116,
      "mean_token_accuracy": 0.17909178584814073,
      "num_tokens": 40296551.0,
      "step": 19410
    },
    {
      "entropy": 5.497962713241577,
      "epoch": 1.761157474600871,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004691759503640024,
      "loss": 5.2316,
      "mean_token_accuracy": 0.17792341262102127,
      "num_tokens": 40306920.0,
      "step": 19415
    },
    {
      "entropy": 5.5028623104095455,
      "epoch": 1.761611030478955,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004691596019996191,
      "loss": 5.1484,
      "mean_token_accuracy": 0.18782503753900529,
      "num_tokens": 40316797.0,
      "step": 19420
    },
    {
      "entropy": 5.458573246002198,
      "epoch": 1.7620645863570392,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004691432496199658,
      "loss": 5.1844,
      "mean_token_accuracy": 0.17946340441703795,
      "num_tokens": 40327002.0,
      "step": 19425
    },
    {
      "entropy": 5.624983215332032,
      "epoch": 1.7625181422351233,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00046912689322538064,
      "loss": 5.37,
      "mean_token_accuracy": 0.17723644971847535,
      "num_tokens": 40337247.0,
      "step": 19430
    },
    {
      "entropy": 5.67900915145874,
      "epoch": 1.7629716981132075,
      "grad_norm": 1.5,
      "learning_rate": 0.00046911053281620187,
      "loss": 5.315,
      "mean_token_accuracy": 0.181755630671978,
      "num_tokens": 40346960.0,
      "step": 19435
    },
    {
      "entropy": 5.586133432388306,
      "epoch": 1.7634252539912918,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046909416839276775,
      "loss": 5.3053,
      "mean_token_accuracy": 0.17899185568094253,
      "num_tokens": 40357169.0,
      "step": 19440
    },
    {
      "entropy": 5.398183727264405,
      "epoch": 1.763878809869376,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004690777999554168,
      "loss": 5.1656,
      "mean_token_accuracy": 0.18723757266998292,
      "num_tokens": 40367660.0,
      "step": 19445
    },
    {
      "entropy": 5.473207807540893,
      "epoch": 1.76433236574746,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004690614275044875,
      "loss": 5.154,
      "mean_token_accuracy": 0.1830536738038063,
      "num_tokens": 40378217.0,
      "step": 19450
    },
    {
      "entropy": 5.522790956497192,
      "epoch": 1.7647859216255442,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004690450510403184,
      "loss": 5.2581,
      "mean_token_accuracy": 0.18336454927921295,
      "num_tokens": 40388249.0,
      "step": 19455
    },
    {
      "entropy": 5.5764532566070555,
      "epoch": 1.7652394775036284,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004690286705632483,
      "loss": 5.3456,
      "mean_token_accuracy": 0.1749739870429039,
      "num_tokens": 40398239.0,
      "step": 19460
    },
    {
      "entropy": 5.580385494232178,
      "epoch": 1.7656930333817127,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004690122860736157,
      "loss": 5.3148,
      "mean_token_accuracy": 0.1809021547436714,
      "num_tokens": 40408473.0,
      "step": 19465
    },
    {
      "entropy": 5.616230916976929,
      "epoch": 1.766146589259797,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00046899589757175964,
      "loss": 5.3802,
      "mean_token_accuracy": 0.1767401620745659,
      "num_tokens": 40418674.0,
      "step": 19470
    },
    {
      "entropy": 5.525962543487549,
      "epoch": 1.766600145137881,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000468979505058019,
      "loss": 5.2303,
      "mean_token_accuracy": 0.17888405323028564,
      "num_tokens": 40429815.0,
      "step": 19475
    },
    {
      "entropy": 5.616372394561767,
      "epoch": 1.767053701015965,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046896310853273274,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1797700360417366,
      "num_tokens": 40439809.0,
      "step": 19480
    },
    {
      "entropy": 5.643142890930176,
      "epoch": 1.7675072568940493,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046894670799624,
      "loss": 5.3083,
      "mean_token_accuracy": 0.17907944917678834,
      "num_tokens": 40449503.0,
      "step": 19485
    },
    {
      "entropy": 5.500616121292114,
      "epoch": 1.7679608127721336,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046893030344887995,
      "loss": 5.3338,
      "mean_token_accuracy": 0.1706432968378067,
      "num_tokens": 40459794.0,
      "step": 19490
    },
    {
      "entropy": 5.545575094223023,
      "epoch": 1.7684143686502178,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046891389489099166,
      "loss": 5.2096,
      "mean_token_accuracy": 0.1881048485636711,
      "num_tokens": 40470044.0,
      "step": 19495
    },
    {
      "entropy": 5.571574783325195,
      "epoch": 1.7688679245283019,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046889748232291476,
      "loss": 5.2515,
      "mean_token_accuracy": 0.18086734414100647,
      "num_tokens": 40480709.0,
      "step": 19500
    },
    {
      "entropy": 5.488039159774781,
      "epoch": 1.769321480406386,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046888106574498843,
      "loss": 5.0894,
      "mean_token_accuracy": 0.18742610067129134,
      "num_tokens": 40490906.0,
      "step": 19505
    },
    {
      "entropy": 5.43992657661438,
      "epoch": 1.7697750362844702,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004688646451575522,
      "loss": 5.2014,
      "mean_token_accuracy": 0.1891149565577507,
      "num_tokens": 40501971.0,
      "step": 19510
    },
    {
      "entropy": 5.538615131378174,
      "epoch": 1.7702285921625545,
      "grad_norm": 1.828125,
      "learning_rate": 0.00046884822056094587,
      "loss": 5.2998,
      "mean_token_accuracy": 0.1767584875226021,
      "num_tokens": 40512685.0,
      "step": 19515
    },
    {
      "entropy": 5.532645177841187,
      "epoch": 1.7706821480406387,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046883179195550883,
      "loss": 5.2331,
      "mean_token_accuracy": 0.18290726840496063,
      "num_tokens": 40523460.0,
      "step": 19520
    },
    {
      "entropy": 5.528083086013794,
      "epoch": 1.7711357039187228,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046881535934158094,
      "loss": 5.2123,
      "mean_token_accuracy": 0.18141518235206605,
      "num_tokens": 40533594.0,
      "step": 19525
    },
    {
      "entropy": 5.538231372833252,
      "epoch": 1.7715892597968068,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004687989227195021,
      "loss": 5.2666,
      "mean_token_accuracy": 0.17919520735740663,
      "num_tokens": 40543504.0,
      "step": 19530
    },
    {
      "entropy": 5.525389719009399,
      "epoch": 1.772042815674891,
      "grad_norm": 1.578125,
      "learning_rate": 0.00046878248208961204,
      "loss": 5.2334,
      "mean_token_accuracy": 0.17830566316843033,
      "num_tokens": 40553711.0,
      "step": 19535
    },
    {
      "entropy": 5.517330455780029,
      "epoch": 1.7724963715529753,
      "grad_norm": 1.8984375,
      "learning_rate": 0.000468766037452251,
      "loss": 5.2627,
      "mean_token_accuracy": 0.18021900802850724,
      "num_tokens": 40563731.0,
      "step": 19540
    },
    {
      "entropy": 5.57808895111084,
      "epoch": 1.7729499274310596,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046874958880775886,
      "loss": 5.2937,
      "mean_token_accuracy": 0.17990937381982802,
      "num_tokens": 40574235.0,
      "step": 19545
    },
    {
      "entropy": 5.563354253768921,
      "epoch": 1.7734034833091437,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004687331361564758,
      "loss": 5.1992,
      "mean_token_accuracy": 0.18843076527118682,
      "num_tokens": 40584459.0,
      "step": 19550
    },
    {
      "entropy": 5.5117463111877445,
      "epoch": 1.7738570391872277,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004687166794987422,
      "loss": 5.2402,
      "mean_token_accuracy": 0.18069267868995667,
      "num_tokens": 40594268.0,
      "step": 19555
    },
    {
      "entropy": 5.522432661056518,
      "epoch": 1.774310595065312,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004687002188348983,
      "loss": 5.2304,
      "mean_token_accuracy": 0.18841561377048494,
      "num_tokens": 40604673.0,
      "step": 19560
    },
    {
      "entropy": 5.4379181385040285,
      "epoch": 1.7747641509433962,
      "grad_norm": 2.0625,
      "learning_rate": 0.00046868375416528443,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18093203455209733,
      "num_tokens": 40614998.0,
      "step": 19565
    },
    {
      "entropy": 5.566183042526245,
      "epoch": 1.7752177068214805,
      "grad_norm": 1.75,
      "learning_rate": 0.00046866728549024123,
      "loss": 5.2626,
      "mean_token_accuracy": 0.18395310044288635,
      "num_tokens": 40624729.0,
      "step": 19570
    },
    {
      "entropy": 5.608444166183472,
      "epoch": 1.7756712626995645,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004686508128101092,
      "loss": 5.3227,
      "mean_token_accuracy": 0.1760799318552017,
      "num_tokens": 40635673.0,
      "step": 19575
    },
    {
      "entropy": 5.607580327987671,
      "epoch": 1.7761248185776488,
      "grad_norm": 1.625,
      "learning_rate": 0.000468634336125229,
      "loss": 5.3041,
      "mean_token_accuracy": 0.17305496633052825,
      "num_tokens": 40645990.0,
      "step": 19580
    },
    {
      "entropy": 5.5750750541687015,
      "epoch": 1.7765783744557329,
      "grad_norm": 1.890625,
      "learning_rate": 0.00046861785543594143,
      "loss": 5.2591,
      "mean_token_accuracy": 0.18518468290567397,
      "num_tokens": 40656339.0,
      "step": 19585
    },
    {
      "entropy": 5.555278539657593,
      "epoch": 1.7770319303338171,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004686013707425873,
      "loss": 5.1535,
      "mean_token_accuracy": 0.18803519904613494,
      "num_tokens": 40666371.0,
      "step": 19590
    },
    {
      "entropy": 5.568233585357666,
      "epoch": 1.7774854862119014,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046858488204550735,
      "loss": 5.2158,
      "mean_token_accuracy": 0.18010344356298447,
      "num_tokens": 40676423.0,
      "step": 19595
    },
    {
      "entropy": 5.518277502059936,
      "epoch": 1.7779390420899854,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004685683893450428,
      "loss": 5.1887,
      "mean_token_accuracy": 0.18854024559259414,
      "num_tokens": 40686331.0,
      "step": 19600
    },
    {
      "entropy": 5.446909666061401,
      "epoch": 1.7783925979680697,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004685518926415346,
      "loss": 5.2267,
      "mean_token_accuracy": 0.19035279750823975,
      "num_tokens": 40697408.0,
      "step": 19605
    },
    {
      "entropy": 5.464632987976074,
      "epoch": 1.7788461538461537,
      "grad_norm": 1.46875,
      "learning_rate": 0.000468535391935324,
      "loss": 5.1337,
      "mean_token_accuracy": 0.1886301651597023,
      "num_tokens": 40707078.0,
      "step": 19610
    },
    {
      "entropy": 5.565882682800293,
      "epoch": 1.779299709724238,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004685188872267521,
      "loss": 5.2403,
      "mean_token_accuracy": 0.18142314553260802,
      "num_tokens": 40718116.0,
      "step": 19615
    },
    {
      "entropy": 5.538724613189697,
      "epoch": 1.7797532656023223,
      "grad_norm": 1.625,
      "learning_rate": 0.0004685023785161603,
      "loss": 5.2622,
      "mean_token_accuracy": 0.1779429718852043,
      "num_tokens": 40727737.0,
      "step": 19620
    },
    {
      "entropy": 5.554495000839234,
      "epoch": 1.7802068214804065,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004684858658038901,
      "loss": 5.2281,
      "mean_token_accuracy": 0.1805749401450157,
      "num_tokens": 40737455.0,
      "step": 19625
    },
    {
      "entropy": 5.628190994262695,
      "epoch": 1.7806603773584906,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004684693490902828,
      "loss": 5.2626,
      "mean_token_accuracy": 0.18136932849884033,
      "num_tokens": 40746715.0,
      "step": 19630
    },
    {
      "entropy": 5.420752477645874,
      "epoch": 1.7811139332365746,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046845282837568,
      "loss": 5.1233,
      "mean_token_accuracy": 0.18474067747592926,
      "num_tokens": 40756432.0,
      "step": 19635
    },
    {
      "entropy": 5.39605484008789,
      "epoch": 1.781567489114659,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046843630366042355,
      "loss": 5.2212,
      "mean_token_accuracy": 0.1809815764427185,
      "num_tokens": 40766629.0,
      "step": 19640
    },
    {
      "entropy": 5.6358452320098875,
      "epoch": 1.7820210449927432,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046841977494485506,
      "loss": 5.3336,
      "mean_token_accuracy": 0.17904380708932877,
      "num_tokens": 40776650.0,
      "step": 19645
    },
    {
      "entropy": 5.639852666854859,
      "epoch": 1.7824746008708274,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004684032422293162,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18358127027750015,
      "num_tokens": 40786722.0,
      "step": 19650
    },
    {
      "entropy": 5.404448652267456,
      "epoch": 1.7829281567489115,
      "grad_norm": 1.625,
      "learning_rate": 0.000468386705514149,
      "loss": 5.0985,
      "mean_token_accuracy": 0.1923267960548401,
      "num_tokens": 40797058.0,
      "step": 19655
    },
    {
      "entropy": 5.476280593872071,
      "epoch": 1.7833817126269955,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004683701647996956,
      "loss": 5.276,
      "mean_token_accuracy": 0.18550340533256532,
      "num_tokens": 40808207.0,
      "step": 19660
    },
    {
      "entropy": 5.630779457092285,
      "epoch": 1.7838352685050798,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004683536200862978,
      "loss": 5.3293,
      "mean_token_accuracy": 0.17899418920278548,
      "num_tokens": 40819655.0,
      "step": 19665
    },
    {
      "entropy": 5.5990461826324465,
      "epoch": 1.784288824383164,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004683370713742979,
      "loss": 5.3352,
      "mean_token_accuracy": 0.17808464616537095,
      "num_tokens": 40831015.0,
      "step": 19670
    },
    {
      "entropy": 5.533109188079834,
      "epoch": 1.7847423802612483,
      "grad_norm": 1.625,
      "learning_rate": 0.00046832051866403814,
      "loss": 5.2024,
      "mean_token_accuracy": 0.18750581294298171,
      "num_tokens": 40841763.0,
      "step": 19675
    },
    {
      "entropy": 5.5117333889007565,
      "epoch": 1.7851959361393324,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004683039619558607,
      "loss": 5.2199,
      "mean_token_accuracy": 0.179233381152153,
      "num_tokens": 40852479.0,
      "step": 19680
    },
    {
      "entropy": 5.535496473312378,
      "epoch": 1.7856494920174164,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046828740125010813,
      "loss": 5.254,
      "mean_token_accuracy": 0.18790972828865052,
      "num_tokens": 40862486.0,
      "step": 19685
    },
    {
      "entropy": 5.601308870315552,
      "epoch": 1.7861030478955007,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004682708365471228,
      "loss": 5.3283,
      "mean_token_accuracy": 0.17832344472408296,
      "num_tokens": 40872182.0,
      "step": 19690
    },
    {
      "entropy": 5.592092037200928,
      "epoch": 1.786556603773585,
      "grad_norm": 2.21875,
      "learning_rate": 0.00046825426784724734,
      "loss": 5.3008,
      "mean_token_accuracy": 0.18158154040575028,
      "num_tokens": 40883036.0,
      "step": 19695
    },
    {
      "entropy": 5.564819240570069,
      "epoch": 1.7870101596516692,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004682376951508244,
      "loss": 5.3334,
      "mean_token_accuracy": 0.16950711458921433,
      "num_tokens": 40893684.0,
      "step": 19700
    },
    {
      "entropy": 5.601074838638306,
      "epoch": 1.7874637155297532,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00046822111845819665,
      "loss": 5.3213,
      "mean_token_accuracy": 0.17773591428995134,
      "num_tokens": 40904540.0,
      "step": 19705
    },
    {
      "entropy": 5.630629777908325,
      "epoch": 1.7879172714078373,
      "grad_norm": 2.28125,
      "learning_rate": 0.00046820453776970697,
      "loss": 5.3048,
      "mean_token_accuracy": 0.18278117179870607,
      "num_tokens": 40915269.0,
      "step": 19710
    },
    {
      "entropy": 5.587602615356445,
      "epoch": 1.7883708272859216,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00046818795308569815,
      "loss": 5.296,
      "mean_token_accuracy": 0.1754431188106537,
      "num_tokens": 40925987.0,
      "step": 19715
    },
    {
      "entropy": 5.53252215385437,
      "epoch": 1.7888243831640058,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004681713644065133,
      "loss": 5.1944,
      "mean_token_accuracy": 0.18986007273197175,
      "num_tokens": 40936130.0,
      "step": 19720
    },
    {
      "entropy": 5.506086111068726,
      "epoch": 1.78927793904209,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004681547717324954,
      "loss": 5.1294,
      "mean_token_accuracy": 0.18584294766187667,
      "num_tokens": 40945476.0,
      "step": 19725
    },
    {
      "entropy": 5.381852388381958,
      "epoch": 1.7897314949201741,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046813817506398755,
      "loss": 5.1382,
      "mean_token_accuracy": 0.1898008942604065,
      "num_tokens": 40956492.0,
      "step": 19730
    },
    {
      "entropy": 5.565962076187134,
      "epoch": 1.7901850507982582,
      "grad_norm": 2.25,
      "learning_rate": 0.00046812157440133303,
      "loss": 5.3292,
      "mean_token_accuracy": 0.17620390951633452,
      "num_tokens": 40966990.0,
      "step": 19735
    },
    {
      "entropy": 5.60609655380249,
      "epoch": 1.7906386066763424,
      "grad_norm": 1.5,
      "learning_rate": 0.0004681049697448752,
      "loss": 5.1921,
      "mean_token_accuracy": 0.18829306960105896,
      "num_tokens": 40977065.0,
      "step": 19740
    },
    {
      "entropy": 5.522910070419312,
      "epoch": 1.7910921625544267,
      "grad_norm": 1.5,
      "learning_rate": 0.00046808836109495734,
      "loss": 5.2437,
      "mean_token_accuracy": 0.1861900046467781,
      "num_tokens": 40987207.0,
      "step": 19745
    },
    {
      "entropy": 5.483324384689331,
      "epoch": 1.791545718432511,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004680717484519229,
      "loss": 5.2059,
      "mean_token_accuracy": 0.18658143579959868,
      "num_tokens": 40997541.0,
      "step": 19750
    },
    {
      "entropy": 5.508931016921997,
      "epoch": 1.791999274310595,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004680551318161156,
      "loss": 5.2203,
      "mean_token_accuracy": 0.187997467815876,
      "num_tokens": 41009044.0,
      "step": 19755
    },
    {
      "entropy": 5.608010005950928,
      "epoch": 1.7924528301886793,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046803851118787897,
      "loss": 5.2403,
      "mean_token_accuracy": 0.1851588800549507,
      "num_tokens": 41018165.0,
      "step": 19760
    },
    {
      "entropy": 5.509888553619385,
      "epoch": 1.7929063860667633,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004680218865675567,
      "loss": 5.1926,
      "mean_token_accuracy": 0.18521979600191116,
      "num_tokens": 41028972.0,
      "step": 19765
    },
    {
      "entropy": 5.480887842178345,
      "epoch": 1.7933599419448476,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004680052579554926,
      "loss": 5.2315,
      "mean_token_accuracy": 0.18455054461956025,
      "num_tokens": 41039552.0,
      "step": 19770
    },
    {
      "entropy": 5.566290664672851,
      "epoch": 1.7938134978229319,
      "grad_norm": 2.78125,
      "learning_rate": 0.0004679886253520307,
      "loss": 5.3153,
      "mean_token_accuracy": 0.17608019858598709,
      "num_tokens": 41049808.0,
      "step": 19775
    },
    {
      "entropy": 5.538585090637207,
      "epoch": 1.794267053701016,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046797198875751474,
      "loss": 5.198,
      "mean_token_accuracy": 0.17646675258874894,
      "num_tokens": 41060574.0,
      "step": 19780
    },
    {
      "entropy": 5.604112720489502,
      "epoch": 1.7947206095791002,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004679553481722889,
      "loss": 5.365,
      "mean_token_accuracy": 0.1758938416838646,
      "num_tokens": 41071239.0,
      "step": 19785
    },
    {
      "entropy": 5.626893615722656,
      "epoch": 1.7951741654571842,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046793870359669733,
      "loss": 5.3664,
      "mean_token_accuracy": 0.17729438543319703,
      "num_tokens": 41080819.0,
      "step": 19790
    },
    {
      "entropy": 5.611803197860718,
      "epoch": 1.7956277213352685,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004679220550310842,
      "loss": 5.1728,
      "mean_token_accuracy": 0.1937761977314949,
      "num_tokens": 41090457.0,
      "step": 19795
    },
    {
      "entropy": 5.501735782623291,
      "epoch": 1.7960812772133528,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004679054024757937,
      "loss": 5.2356,
      "mean_token_accuracy": 0.18729056268930436,
      "num_tokens": 41101210.0,
      "step": 19800
    },
    {
      "entropy": 5.481209087371826,
      "epoch": 1.796534833091437,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004678887459311704,
      "loss": 5.2573,
      "mean_token_accuracy": 0.18591037392616272,
      "num_tokens": 41110566.0,
      "step": 19805
    },
    {
      "entropy": 5.498694133758545,
      "epoch": 1.796988388969521,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004678720853975587,
      "loss": 5.1717,
      "mean_token_accuracy": 0.18803776949644088,
      "num_tokens": 41120000.0,
      "step": 19810
    },
    {
      "entropy": 5.472893524169922,
      "epoch": 1.797441944847605,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00046785542087530307,
      "loss": 5.157,
      "mean_token_accuracy": 0.18770240545272826,
      "num_tokens": 41129944.0,
      "step": 19815
    },
    {
      "entropy": 5.505154132843018,
      "epoch": 1.7978955007256894,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004678387523647483,
      "loss": 5.2271,
      "mean_token_accuracy": 0.18277432024478912,
      "num_tokens": 41138803.0,
      "step": 19820
    },
    {
      "entropy": 5.582071685791016,
      "epoch": 1.7983490566037736,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00046782207986623885,
      "loss": 5.3234,
      "mean_token_accuracy": 0.17467063218355178,
      "num_tokens": 41150365.0,
      "step": 19825
    },
    {
      "entropy": 5.6089507102966305,
      "epoch": 1.798802612481858,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00046780540338011974,
      "loss": 5.2746,
      "mean_token_accuracy": 0.1776033028960228,
      "num_tokens": 41160310.0,
      "step": 19830
    },
    {
      "entropy": 5.503898429870605,
      "epoch": 1.799256168359942,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004677887229067358,
      "loss": 5.1493,
      "mean_token_accuracy": 0.18875844776630402,
      "num_tokens": 41170506.0,
      "step": 19835
    },
    {
      "entropy": 5.536115837097168,
      "epoch": 1.799709724238026,
      "grad_norm": 1.734375,
      "learning_rate": 0.00046777203844643184,
      "loss": 5.2779,
      "mean_token_accuracy": 0.17575109452009202,
      "num_tokens": 41181293.0,
      "step": 19840
    },
    {
      "entropy": 5.500993394851685,
      "epoch": 1.8001632801161103,
      "grad_norm": 1.984375,
      "learning_rate": 0.00046775534999955303,
      "loss": 5.2225,
      "mean_token_accuracy": 0.18243959695100784,
      "num_tokens": 41191637.0,
      "step": 19845
    },
    {
      "entropy": 5.4735126972198485,
      "epoch": 1.8006168359941945,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004677386575664445,
      "loss": 5.209,
      "mean_token_accuracy": 0.18708751797676088,
      "num_tokens": 41201833.0,
      "step": 19850
    },
    {
      "entropy": 5.619037199020386,
      "epoch": 1.8010703918722788,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046772196114745133,
      "loss": 5.2648,
      "mean_token_accuracy": 0.18672931641340257,
      "num_tokens": 41211243.0,
      "step": 19855
    },
    {
      "entropy": 5.5941568374633786,
      "epoch": 1.8015239477503628,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004677052607429189,
      "loss": 5.2416,
      "mean_token_accuracy": 0.18227624595165254,
      "num_tokens": 41221111.0,
      "step": 19860
    },
    {
      "entropy": 5.490173625946045,
      "epoch": 1.8019775036284469,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004676885563531927,
      "loss": 5.2319,
      "mean_token_accuracy": 0.18557505905628205,
      "num_tokens": 41231578.0,
      "step": 19865
    },
    {
      "entropy": 5.506578350067139,
      "epoch": 1.8024310595065312,
      "grad_norm": 1.6875,
      "learning_rate": 0.00046767184797861795,
      "loss": 5.3078,
      "mean_token_accuracy": 0.18385085165500642,
      "num_tokens": 41242130.0,
      "step": 19870
    },
    {
      "entropy": 5.5875694274902346,
      "epoch": 1.8028846153846154,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046765513561954026,
      "loss": 5.2651,
      "mean_token_accuracy": 0.17687568366527556,
      "num_tokens": 41252885.0,
      "step": 19875
    },
    {
      "entropy": 5.607456636428833,
      "epoch": 1.8033381712626997,
      "grad_norm": 1.78125,
      "learning_rate": 0.00046763841927630534,
      "loss": 5.2889,
      "mean_token_accuracy": 0.17704417407512665,
      "num_tokens": 41263532.0,
      "step": 19880
    },
    {
      "entropy": 5.606713914871216,
      "epoch": 1.8037917271407837,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046762169894925876,
      "loss": 5.2548,
      "mean_token_accuracy": 0.18106650859117507,
      "num_tokens": 41273469.0,
      "step": 19885
    },
    {
      "entropy": 5.545895338058472,
      "epoch": 1.8042452830188678,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004676049746387464,
      "loss": 5.2551,
      "mean_token_accuracy": 0.180370756983757,
      "num_tokens": 41283663.0,
      "step": 19890
    },
    {
      "entropy": 5.474084424972534,
      "epoch": 1.804698838896952,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004675882463451141,
      "loss": 5.1632,
      "mean_token_accuracy": 0.1879630744457245,
      "num_tokens": 41294657.0,
      "step": 19895
    },
    {
      "entropy": 5.537800359725952,
      "epoch": 1.8051523947750363,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004675715140687078,
      "loss": 5.2789,
      "mean_token_accuracy": 0.1823866233229637,
      "num_tokens": 41305124.0,
      "step": 19900
    },
    {
      "entropy": 5.49251537322998,
      "epoch": 1.8056059506531206,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046755477780987344,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19205471575260163,
      "num_tokens": 41315839.0,
      "step": 19905
    },
    {
      "entropy": 5.5995934963226315,
      "epoch": 1.8060595065312046,
      "grad_norm": 2.453125,
      "learning_rate": 0.00046753803756895726,
      "loss": 5.2714,
      "mean_token_accuracy": 0.18146696984767913,
      "num_tokens": 41326388.0,
      "step": 19910
    },
    {
      "entropy": 5.555939340591431,
      "epoch": 1.8065130624092887,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004675212933463054,
      "loss": 5.2882,
      "mean_token_accuracy": 0.17351327538490297,
      "num_tokens": 41337333.0,
      "step": 19915
    },
    {
      "entropy": 5.5659082412719725,
      "epoch": 1.806966618287373,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004675045451422641,
      "loss": 5.3095,
      "mean_token_accuracy": 0.17297808527946473,
      "num_tokens": 41348353.0,
      "step": 19920
    },
    {
      "entropy": 5.569571924209595,
      "epoch": 1.8074201741654572,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046748779295717973,
      "loss": 5.2879,
      "mean_token_accuracy": 0.18718119263648986,
      "num_tokens": 41358797.0,
      "step": 19925
    },
    {
      "entropy": 5.544040393829346,
      "epoch": 1.8078737300435415,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004674710367913989,
      "loss": 5.2707,
      "mean_token_accuracy": 0.1829158902168274,
      "num_tokens": 41368818.0,
      "step": 19930
    },
    {
      "entropy": 5.552248859405518,
      "epoch": 1.8083272859216255,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004674542766452678,
      "loss": 5.2539,
      "mean_token_accuracy": 0.18378699868917464,
      "num_tokens": 41378949.0,
      "step": 19935
    },
    {
      "entropy": 5.563677549362183,
      "epoch": 1.8087808417997098,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00046743751251913327,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18650126904249192,
      "num_tokens": 41388892.0,
      "step": 19940
    },
    {
      "entropy": 5.53504581451416,
      "epoch": 1.8092343976777938,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00046742074441334197,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18386920690536498,
      "num_tokens": 41399197.0,
      "step": 19945
    },
    {
      "entropy": 5.542147779464722,
      "epoch": 1.809687953555878,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004674039723282406,
      "loss": 5.2647,
      "mean_token_accuracy": 0.1727042779326439,
      "num_tokens": 41409551.0,
      "step": 19950
    },
    {
      "entropy": 5.502180433273315,
      "epoch": 1.8101415094339623,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046738719626417607,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1805973768234253,
      "num_tokens": 41418900.0,
      "step": 19955
    },
    {
      "entropy": 5.514769172668457,
      "epoch": 1.8105950653120464,
      "grad_norm": 2.765625,
      "learning_rate": 0.00046737041622149536,
      "loss": 5.1922,
      "mean_token_accuracy": 0.18147982954978942,
      "num_tokens": 41428204.0,
      "step": 19960
    },
    {
      "entropy": 5.503611755371094,
      "epoch": 1.8110486211901307,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004673536322005454,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18150213807821275,
      "num_tokens": 41437922.0,
      "step": 19965
    },
    {
      "entropy": 5.589338207244873,
      "epoch": 1.8115021770682147,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004673368442016732,
      "loss": 5.2325,
      "mean_token_accuracy": 0.17826240956783296,
      "num_tokens": 41447411.0,
      "step": 19970
    },
    {
      "entropy": 5.62417778968811,
      "epoch": 1.811955732946299,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004673200522252261,
      "loss": 5.3389,
      "mean_token_accuracy": 0.17650886327028276,
      "num_tokens": 41458756.0,
      "step": 19975
    },
    {
      "entropy": 5.580564355850219,
      "epoch": 1.8124092888243832,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004673032562715514,
      "loss": 5.2343,
      "mean_token_accuracy": 0.18187149316072465,
      "num_tokens": 41468610.0,
      "step": 19980
    },
    {
      "entropy": 5.544586133956909,
      "epoch": 1.8128628447024675,
      "grad_norm": 1.734375,
      "learning_rate": 0.00046728645634099627,
      "loss": 5.3798,
      "mean_token_accuracy": 0.16945139914751053,
      "num_tokens": 41479849.0,
      "step": 19985
    },
    {
      "entropy": 5.531964445114136,
      "epoch": 1.8133164005805515,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046726965243390825,
      "loss": 5.2685,
      "mean_token_accuracy": 0.17659684270620346,
      "num_tokens": 41490945.0,
      "step": 19990
    },
    {
      "entropy": 5.545766687393188,
      "epoch": 1.8137699564586356,
      "grad_norm": 2.734375,
      "learning_rate": 0.0004672528445506348,
      "loss": 5.1796,
      "mean_token_accuracy": 0.18117519319057465,
      "num_tokens": 41501479.0,
      "step": 19995
    },
    {
      "entropy": 5.583986473083496,
      "epoch": 1.8142235123367199,
      "grad_norm": 1.375,
      "learning_rate": 0.0004672360326915236,
      "loss": 5.1845,
      "mean_token_accuracy": 0.19188567399978637,
      "num_tokens": 41512447.0,
      "step": 20000
    },
    {
      "entropy": 5.444471073150635,
      "epoch": 1.8146770682148041,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004672192168569222,
      "loss": 5.1563,
      "mean_token_accuracy": 0.1923826664686203,
      "num_tokens": 41521905.0,
      "step": 20005
    },
    {
      "entropy": 5.462075090408325,
      "epoch": 1.8151306240928884,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00046720239704717845,
      "loss": 5.2665,
      "mean_token_accuracy": 0.1857066348195076,
      "num_tokens": 41533143.0,
      "step": 20010
    },
    {
      "entropy": 5.5456404209136965,
      "epoch": 1.8155841799709724,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046718557326264,
      "loss": 5.1876,
      "mean_token_accuracy": 0.18167518824338913,
      "num_tokens": 41542234.0,
      "step": 20015
    },
    {
      "entropy": 5.617145967483521,
      "epoch": 1.8160377358490565,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004671687455036551,
      "loss": 5.3304,
      "mean_token_accuracy": 0.17571862190961837,
      "num_tokens": 41552632.0,
      "step": 20020
    },
    {
      "entropy": 5.562243890762329,
      "epoch": 1.8164912917271407,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004671519137705715,
      "loss": 5.2694,
      "mean_token_accuracy": 0.18459609746932984,
      "num_tokens": 41563723.0,
      "step": 20025
    },
    {
      "entropy": 5.5164870262146,
      "epoch": 1.816944847605225,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046713507806373736,
      "loss": 5.2216,
      "mean_token_accuracy": 0.18303583264350892,
      "num_tokens": 41573680.0,
      "step": 20030
    },
    {
      "entropy": 5.527782154083252,
      "epoch": 1.8173984034833093,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046711823838350086,
      "loss": 5.3013,
      "mean_token_accuracy": 0.17864342480897905,
      "num_tokens": 41584722.0,
      "step": 20035
    },
    {
      "entropy": 5.5861505508422855,
      "epoch": 1.8178519593613933,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046710139473021025,
      "loss": 5.3178,
      "mean_token_accuracy": 0.17666983753442764,
      "num_tokens": 41595785.0,
      "step": 20040
    },
    {
      "entropy": 5.525841999053955,
      "epoch": 1.8183055152394774,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046708454710421376,
      "loss": 5.3248,
      "mean_token_accuracy": 0.1767853543162346,
      "num_tokens": 41606023.0,
      "step": 20045
    },
    {
      "entropy": 5.576890563964843,
      "epoch": 1.8187590711175616,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046706769550586,
      "loss": 5.2515,
      "mean_token_accuracy": 0.1810461699962616,
      "num_tokens": 41616010.0,
      "step": 20050
    },
    {
      "entropy": 5.548904466629028,
      "epoch": 1.819212626995646,
      "grad_norm": 1.640625,
      "learning_rate": 0.00046705083993549723,
      "loss": 5.205,
      "mean_token_accuracy": 0.1847629129886627,
      "num_tokens": 41625105.0,
      "step": 20055
    },
    {
      "entropy": 5.493886709213257,
      "epoch": 1.8196661828737302,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004670339803934743,
      "loss": 5.2656,
      "mean_token_accuracy": 0.1856745570898056,
      "num_tokens": 41634989.0,
      "step": 20060
    },
    {
      "entropy": 5.527575302124023,
      "epoch": 1.8201197387518142,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004670171168801396,
      "loss": 5.224,
      "mean_token_accuracy": 0.18196288049221038,
      "num_tokens": 41645736.0,
      "step": 20065
    },
    {
      "entropy": 5.635229110717773,
      "epoch": 1.8205732946298983,
      "grad_norm": 1.625,
      "learning_rate": 0.000467000249395842,
      "loss": 5.2359,
      "mean_token_accuracy": 0.18672096133232116,
      "num_tokens": 41655505.0,
      "step": 20070
    },
    {
      "entropy": 5.5040473461151125,
      "epoch": 1.8210268505079825,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004669833779409303,
      "loss": 5.218,
      "mean_token_accuracy": 0.18544519692659378,
      "num_tokens": 41665732.0,
      "step": 20075
    },
    {
      "entropy": 5.495439052581787,
      "epoch": 1.8214804063860668,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004669665025157535,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1785016492009163,
      "num_tokens": 41676050.0,
      "step": 20080
    },
    {
      "entropy": 5.603182649612426,
      "epoch": 1.821933962264151,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00046694962312066046,
      "loss": 5.3714,
      "mean_token_accuracy": 0.17878033667802812,
      "num_tokens": 41686787.0,
      "step": 20085
    },
    {
      "entropy": 5.506361055374145,
      "epoch": 1.822387518142235,
      "grad_norm": 2.5,
      "learning_rate": 0.00046693273975600036,
      "loss": 5.1851,
      "mean_token_accuracy": 0.18441618233919144,
      "num_tokens": 41697168.0,
      "step": 20090
    },
    {
      "entropy": 5.479551696777344,
      "epoch": 1.8228410740203191,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004669158524221222,
      "loss": 5.2716,
      "mean_token_accuracy": 0.1856548711657524,
      "num_tokens": 41707524.0,
      "step": 20095
    },
    {
      "entropy": 5.571998167037964,
      "epoch": 1.8232946298984034,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046689896111937537,
      "loss": 5.295,
      "mean_token_accuracy": 0.18541192412376403,
      "num_tokens": 41718109.0,
      "step": 20100
    },
    {
      "entropy": 5.636944913864136,
      "epoch": 1.8237481857764877,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004668820658481091,
      "loss": 5.2772,
      "mean_token_accuracy": 0.17956832498311998,
      "num_tokens": 41728862.0,
      "step": 20105
    },
    {
      "entropy": 5.600374841690064,
      "epoch": 1.824201741654572,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046686516660867283,
      "loss": 5.3324,
      "mean_token_accuracy": 0.17926641702651977,
      "num_tokens": 41739024.0,
      "step": 20110
    },
    {
      "entropy": 5.544338655471802,
      "epoch": 1.824655297532656,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004668482634014159,
      "loss": 5.2734,
      "mean_token_accuracy": 0.1798844516277313,
      "num_tokens": 41748979.0,
      "step": 20115
    },
    {
      "entropy": 5.544781923294067,
      "epoch": 1.8251088534107403,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004668313562266882,
      "loss": 5.3036,
      "mean_token_accuracy": 0.18296718895435332,
      "num_tokens": 41759972.0,
      "step": 20120
    },
    {
      "entropy": 5.536408233642578,
      "epoch": 1.8255624092888243,
      "grad_norm": 1.53125,
      "learning_rate": 0.000466814445084839,
      "loss": 5.1243,
      "mean_token_accuracy": 0.19248746633529662,
      "num_tokens": 41770150.0,
      "step": 20125
    },
    {
      "entropy": 5.50967583656311,
      "epoch": 1.8260159651669086,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004667975299762183,
      "loss": 5.2093,
      "mean_token_accuracy": 0.18377644419670106,
      "num_tokens": 41780405.0,
      "step": 20130
    },
    {
      "entropy": 5.5058728694915775,
      "epoch": 1.8264695210449928,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046678061090117576,
      "loss": 5.146,
      "mean_token_accuracy": 0.18893924206495286,
      "num_tokens": 41791269.0,
      "step": 20135
    },
    {
      "entropy": 5.451051187515259,
      "epoch": 1.8269230769230769,
      "grad_norm": 1.75,
      "learning_rate": 0.0004667636878600613,
      "loss": 5.1517,
      "mean_token_accuracy": 0.18863540887832642,
      "num_tokens": 41802513.0,
      "step": 20140
    },
    {
      "entropy": 5.549851036071777,
      "epoch": 1.8273766328011611,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004667467608532249,
      "loss": 5.3123,
      "mean_token_accuracy": 0.17471880465745926,
      "num_tokens": 41814149.0,
      "step": 20145
    },
    {
      "entropy": 5.596239519119263,
      "epoch": 1.8278301886792452,
      "grad_norm": 1.375,
      "learning_rate": 0.0004667298298810167,
      "loss": 5.2704,
      "mean_token_accuracy": 0.17963476330041886,
      "num_tokens": 41825763.0,
      "step": 20150
    },
    {
      "entropy": 5.52785358428955,
      "epoch": 1.8282837445573294,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004667128949437867,
      "loss": 5.2243,
      "mean_token_accuracy": 0.18647098988294603,
      "num_tokens": 41834543.0,
      "step": 20155
    },
    {
      "entropy": 5.484307718276978,
      "epoch": 1.8287373004354137,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004666959560418852,
      "loss": 5.1763,
      "mean_token_accuracy": 0.18895630836486815,
      "num_tokens": 41844940.0,
      "step": 20160
    },
    {
      "entropy": 5.456264209747315,
      "epoch": 1.829190856313498,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046667901317566245,
      "loss": 5.1617,
      "mean_token_accuracy": 0.1915537565946579,
      "num_tokens": 41855532.0,
      "step": 20165
    },
    {
      "entropy": 5.559297657012939,
      "epoch": 1.829644412191582,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004666620663454689,
      "loss": 5.256,
      "mean_token_accuracy": 0.17944580763578416,
      "num_tokens": 41865661.0,
      "step": 20170
    },
    {
      "entropy": 5.548720169067383,
      "epoch": 1.830097968069666,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046664511555165496,
      "loss": 5.2583,
      "mean_token_accuracy": 0.18172841519117355,
      "num_tokens": 41875662.0,
      "step": 20175
    },
    {
      "entropy": 5.516387701034546,
      "epoch": 1.8305515239477503,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00046662816079457114,
      "loss": 5.2219,
      "mean_token_accuracy": 0.19133128374814987,
      "num_tokens": 41886211.0,
      "step": 20180
    },
    {
      "entropy": 5.572917032241821,
      "epoch": 1.8310050798258346,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004666112020745682,
      "loss": 5.3237,
      "mean_token_accuracy": 0.1788841664791107,
      "num_tokens": 41896737.0,
      "step": 20185
    },
    {
      "entropy": 5.574762582778931,
      "epoch": 1.8314586357039189,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004665942393919967,
      "loss": 5.2245,
      "mean_token_accuracy": 0.18119091987609864,
      "num_tokens": 41907768.0,
      "step": 20190
    },
    {
      "entropy": 5.528245067596435,
      "epoch": 1.831912191582003,
      "grad_norm": 2.1875,
      "learning_rate": 0.00046657727274720755,
      "loss": 5.1893,
      "mean_token_accuracy": 0.17866142690181733,
      "num_tokens": 41917587.0,
      "step": 20195
    },
    {
      "entropy": 5.450700855255127,
      "epoch": 1.832365747460087,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004665603021405516,
      "loss": 5.1639,
      "mean_token_accuracy": 0.1917879670858383,
      "num_tokens": 41927842.0,
      "step": 20200
    },
    {
      "entropy": 5.503348541259766,
      "epoch": 1.8328193033381712,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004665433275723798,
      "loss": 5.2228,
      "mean_token_accuracy": 0.18814008086919784,
      "num_tokens": 41939899.0,
      "step": 20205
    },
    {
      "entropy": 5.48411808013916,
      "epoch": 1.8332728592162555,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004665263490430431,
      "loss": 5.2134,
      "mean_token_accuracy": 0.17579588145017624,
      "num_tokens": 41950932.0,
      "step": 20210
    },
    {
      "entropy": 5.578576469421387,
      "epoch": 1.8337264150943398,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004665093665528928,
      "loss": 5.283,
      "mean_token_accuracy": 0.1814352825284004,
      "num_tokens": 41963512.0,
      "step": 20215
    },
    {
      "entropy": 5.597681379318237,
      "epoch": 1.8341799709724238,
      "grad_norm": 3.125,
      "learning_rate": 0.00046649238010227983,
      "loss": 5.286,
      "mean_token_accuracy": 0.18207603245973586,
      "num_tokens": 41974677.0,
      "step": 20220
    },
    {
      "entropy": 5.4647767543792725,
      "epoch": 1.8346335268505078,
      "grad_norm": 1.875,
      "learning_rate": 0.0004664753896915558,
      "loss": 5.1813,
      "mean_token_accuracy": 0.18435967117547988,
      "num_tokens": 41984582.0,
      "step": 20225
    },
    {
      "entropy": 5.513673543930054,
      "epoch": 1.8350870827285921,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004664583953210719,
      "loss": 5.1717,
      "mean_token_accuracy": 0.18482727259397508,
      "num_tokens": 41994624.0,
      "step": 20230
    },
    {
      "entropy": 5.564163446426392,
      "epoch": 1.8355406386066764,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004664413969911795,
      "loss": 5.177,
      "mean_token_accuracy": 0.18555971086025239,
      "num_tokens": 42004835.0,
      "step": 20235
    },
    {
      "entropy": 5.490981340408325,
      "epoch": 1.8359941944847606,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004664243947022303,
      "loss": 5.2035,
      "mean_token_accuracy": 0.18475724756717682,
      "num_tokens": 42014727.0,
      "step": 20240
    },
    {
      "entropy": 5.536380577087402,
      "epoch": 1.8364477503628447,
      "grad_norm": 1.78125,
      "learning_rate": 0.00046640738845457586,
      "loss": 5.2537,
      "mean_token_accuracy": 0.18352523297071457,
      "num_tokens": 42023628.0,
      "step": 20245
    },
    {
      "entropy": 5.617591857910156,
      "epoch": 1.8369013062409287,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004663903782485678,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1802628979086876,
      "num_tokens": 42033587.0,
      "step": 20250
    },
    {
      "entropy": 5.624684143066406,
      "epoch": 1.837354862119013,
      "grad_norm": 1.71875,
      "learning_rate": 0.00046637336408455793,
      "loss": 5.2779,
      "mean_token_accuracy": 0.17954636961221696,
      "num_tokens": 42043765.0,
      "step": 20255
    },
    {
      "entropy": 5.592569351196289,
      "epoch": 1.8378084179970973,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046635634596289816,
      "loss": 5.3473,
      "mean_token_accuracy": 0.17787133157253265,
      "num_tokens": 42054205.0,
      "step": 20260
    },
    {
      "entropy": 5.468901348114014,
      "epoch": 1.8382619738751815,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004663393238839404,
      "loss": 5.1352,
      "mean_token_accuracy": 0.18333043456077575,
      "num_tokens": 42064512.0,
      "step": 20265
    },
    {
      "entropy": 5.58189582824707,
      "epoch": 1.8387155297532656,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004663222978480366,
      "loss": 5.2362,
      "mean_token_accuracy": 0.1801536649465561,
      "num_tokens": 42074954.0,
      "step": 20270
    },
    {
      "entropy": 5.595988941192627,
      "epoch": 1.8391690856313496,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000466305267855539,
      "loss": 5.2548,
      "mean_token_accuracy": 0.18269609212875365,
      "num_tokens": 42084914.0,
      "step": 20275
    },
    {
      "entropy": 5.557289981842041,
      "epoch": 1.8396226415094339,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004662882339067996,
      "loss": 5.2753,
      "mean_token_accuracy": 0.18714795857667924,
      "num_tokens": 42095454.0,
      "step": 20280
    },
    {
      "entropy": 5.539448547363281,
      "epoch": 1.8400761973875182,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004662711960021709,
      "loss": 5.3133,
      "mean_token_accuracy": 0.17826288491487502,
      "num_tokens": 42106960.0,
      "step": 20285
    },
    {
      "entropy": 5.560920524597168,
      "epoch": 1.8405297532656024,
      "grad_norm": 2.1875,
      "learning_rate": 0.00046625415414200513,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1847240149974823,
      "num_tokens": 42117009.0,
      "step": 20290
    },
    {
      "entropy": 5.590720176696777,
      "epoch": 1.8409833091436865,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004662371083266546,
      "loss": 5.2306,
      "mean_token_accuracy": 0.18851496577262877,
      "num_tokens": 42126371.0,
      "step": 20295
    },
    {
      "entropy": 5.573096704483032,
      "epoch": 1.8414368650217707,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000466220058556472,
      "loss": 5.2338,
      "mean_token_accuracy": 0.1865280643105507,
      "num_tokens": 42135890.0,
      "step": 20300
    },
    {
      "entropy": 5.648856782913208,
      "epoch": 1.8418904208998548,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00046620300483180986,
      "loss": 5.3684,
      "mean_token_accuracy": 0.17492888420820235,
      "num_tokens": 42145696.0,
      "step": 20305
    },
    {
      "entropy": 5.580995368957519,
      "epoch": 1.842343976777939,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004661859471530209,
      "loss": 5.279,
      "mean_token_accuracy": 0.18526298701763153,
      "num_tokens": 42156236.0,
      "step": 20310
    },
    {
      "entropy": 5.591434097290039,
      "epoch": 1.8427975326560233,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004661688855204578,
      "loss": 5.2995,
      "mean_token_accuracy": 0.181219682097435,
      "num_tokens": 42166179.0,
      "step": 20315
    },
    {
      "entropy": 5.504329252243042,
      "epoch": 1.8432510885341074,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046615181993447344,
      "loss": 5.1646,
      "mean_token_accuracy": 0.18525262624025346,
      "num_tokens": 42176643.0,
      "step": 20320
    },
    {
      "entropy": 5.599558305740357,
      "epoch": 1.8437046444121916,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004661347503954207,
      "loss": 5.4314,
      "mean_token_accuracy": 0.17119999676942826,
      "num_tokens": 42188328.0,
      "step": 20325
    },
    {
      "entropy": 5.633039140701294,
      "epoch": 1.8441582002902757,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00046611767690365265,
      "loss": 5.2716,
      "mean_token_accuracy": 0.18542708158493043,
      "num_tokens": 42198110.0,
      "step": 20330
    },
    {
      "entropy": 5.506770753860474,
      "epoch": 1.84461175616836,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004661005994595223,
      "loss": 5.2344,
      "mean_token_accuracy": 0.1758541375398636,
      "num_tokens": 42208438.0,
      "step": 20335
    },
    {
      "entropy": 5.524604749679566,
      "epoch": 1.8450653120464442,
      "grad_norm": 2.75,
      "learning_rate": 0.0004660835180633829,
      "loss": 5.1958,
      "mean_token_accuracy": 0.18411289602518083,
      "num_tokens": 42220052.0,
      "step": 20340
    },
    {
      "entropy": 5.513128042221069,
      "epoch": 1.8455188679245285,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004660664327155876,
      "loss": 5.2119,
      "mean_token_accuracy": 0.1856969490647316,
      "num_tokens": 42229497.0,
      "step": 20345
    },
    {
      "entropy": 5.529416799545288,
      "epoch": 1.8459724238026125,
      "grad_norm": 1.625,
      "learning_rate": 0.0004660493434164898,
      "loss": 5.2683,
      "mean_token_accuracy": 0.1755200669169426,
      "num_tokens": 42240725.0,
      "step": 20350
    },
    {
      "entropy": 5.581337976455688,
      "epoch": 1.8464259796806965,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004660322501664429,
      "loss": 5.2734,
      "mean_token_accuracy": 0.18036877512931823,
      "num_tokens": 42251114.0,
      "step": 20355
    },
    {
      "entropy": 5.559364318847656,
      "epoch": 1.8468795355587808,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004660151529658004,
      "loss": 5.4498,
      "mean_token_accuracy": 0.17488445192575455,
      "num_tokens": 42262825.0,
      "step": 20360
    },
    {
      "entropy": 5.614293575286865,
      "epoch": 1.847333091436865,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004659980518149158,
      "loss": 5.2721,
      "mean_token_accuracy": 0.19035911560058594,
      "num_tokens": 42272176.0,
      "step": 20365
    },
    {
      "entropy": 5.621585416793823,
      "epoch": 1.8477866473149493,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046598094671414285,
      "loss": 5.3112,
      "mean_token_accuracy": 0.18066442608833314,
      "num_tokens": 42281934.0,
      "step": 20370
    },
    {
      "entropy": 5.487783479690552,
      "epoch": 1.8482402031930334,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046596383766383533,
      "loss": 5.1848,
      "mean_token_accuracy": 0.18560474514961242,
      "num_tokens": 42291927.0,
      "step": 20375
    },
    {
      "entropy": 5.478231143951416,
      "epoch": 1.8486937590711174,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004659467246643469,
      "loss": 5.1888,
      "mean_token_accuracy": 0.1868455946445465,
      "num_tokens": 42301407.0,
      "step": 20380
    },
    {
      "entropy": 5.461211490631103,
      "epoch": 1.8491473149492017,
      "grad_norm": 1.6875,
      "learning_rate": 0.00046592960771603154,
      "loss": 5.1349,
      "mean_token_accuracy": 0.1933228477835655,
      "num_tokens": 42311199.0,
      "step": 20385
    },
    {
      "entropy": 5.506057548522949,
      "epoch": 1.849600870827286,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004659124868192433,
      "loss": 5.2118,
      "mean_token_accuracy": 0.18900125920772554,
      "num_tokens": 42320725.0,
      "step": 20390
    },
    {
      "entropy": 5.608890724182129,
      "epoch": 1.8500544267053702,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004658953619743361,
      "loss": 5.3137,
      "mean_token_accuracy": 0.18401983380317688,
      "num_tokens": 42330857.0,
      "step": 20395
    },
    {
      "entropy": 5.472943067550659,
      "epoch": 1.8505079825834543,
      "grad_norm": 1.6875,
      "learning_rate": 0.00046587823318166427,
      "loss": 5.1588,
      "mean_token_accuracy": 0.18216922581195832,
      "num_tokens": 42340241.0,
      "step": 20400
    },
    {
      "entropy": 5.5945055961608885,
      "epoch": 1.8509615384615383,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046586110044158184,
      "loss": 5.3654,
      "mean_token_accuracy": 0.17415248453617097,
      "num_tokens": 42351756.0,
      "step": 20405
    },
    {
      "entropy": 5.559737920761108,
      "epoch": 1.8514150943396226,
      "grad_norm": 1.921875,
      "learning_rate": 0.00046584396375444335,
      "loss": 5.1887,
      "mean_token_accuracy": 0.18865942806005478,
      "num_tokens": 42361649.0,
      "step": 20410
    },
    {
      "entropy": 5.520679521560669,
      "epoch": 1.8518686502177069,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046582682312060295,
      "loss": 5.218,
      "mean_token_accuracy": 0.18609192669391633,
      "num_tokens": 42371273.0,
      "step": 20415
    },
    {
      "entropy": 5.502820301055908,
      "epoch": 1.8523222060957911,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004658096785404152,
      "loss": 5.2181,
      "mean_token_accuracy": 0.1833156406879425,
      "num_tokens": 42382275.0,
      "step": 20420
    },
    {
      "entropy": 5.615487718582154,
      "epoch": 1.8527757619738752,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046579253001423473,
      "loss": 5.2527,
      "mean_token_accuracy": 0.18078850507736205,
      "num_tokens": 42392360.0,
      "step": 20425
    },
    {
      "entropy": 5.543534564971924,
      "epoch": 1.8532293178519592,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046577537754241604,
      "loss": 5.259,
      "mean_token_accuracy": 0.1723129242658615,
      "num_tokens": 42403695.0,
      "step": 20430
    },
    {
      "entropy": 5.5188923358917235,
      "epoch": 1.8536828737300435,
      "grad_norm": 1.5,
      "learning_rate": 0.000465758221125314,
      "loss": 5.1709,
      "mean_token_accuracy": 0.18989308476448058,
      "num_tokens": 42414199.0,
      "step": 20435
    },
    {
      "entropy": 5.524149227142334,
      "epoch": 1.8541364296081277,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00046574106076328333,
      "loss": 5.2994,
      "mean_token_accuracy": 0.1790919542312622,
      "num_tokens": 42423655.0,
      "step": 20440
    },
    {
      "entropy": 5.496356773376465,
      "epoch": 1.854589985486212,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004657238964566789,
      "loss": 5.1996,
      "mean_token_accuracy": 0.1867751270532608,
      "num_tokens": 42434235.0,
      "step": 20445
    },
    {
      "entropy": 5.46200795173645,
      "epoch": 1.855043541364296,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004657067282058557,
      "loss": 5.2289,
      "mean_token_accuracy": 0.18581778705120086,
      "num_tokens": 42444701.0,
      "step": 20450
    },
    {
      "entropy": 5.482279872894287,
      "epoch": 1.85549709724238,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004656895560111687,
      "loss": 5.2113,
      "mean_token_accuracy": 0.18465594202280045,
      "num_tokens": 42454499.0,
      "step": 20455
    },
    {
      "entropy": 5.625277996063232,
      "epoch": 1.8559506531204644,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00046567237987297323,
      "loss": 5.1988,
      "mean_token_accuracy": 0.17584859281778337,
      "num_tokens": 42465391.0,
      "step": 20460
    },
    {
      "entropy": 5.545505380630493,
      "epoch": 1.8564042089985486,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004656551997916243,
      "loss": 5.202,
      "mean_token_accuracy": 0.17676693350076675,
      "num_tokens": 42476126.0,
      "step": 20465
    },
    {
      "entropy": 5.514323377609253,
      "epoch": 1.856857764876633,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046563801576747716,
      "loss": 5.2755,
      "mean_token_accuracy": 0.1834051012992859,
      "num_tokens": 42486833.0,
      "step": 20470
    },
    {
      "entropy": 5.65627269744873,
      "epoch": 1.857311320754717,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046562082780088733,
      "loss": 5.3052,
      "mean_token_accuracy": 0.17678626626729965,
      "num_tokens": 42496986.0,
      "step": 20475
    },
    {
      "entropy": 5.556107854843139,
      "epoch": 1.8577648766328012,
      "grad_norm": 1.875,
      "learning_rate": 0.0004656036358922103,
      "loss": 5.213,
      "mean_token_accuracy": 0.18136774003505707,
      "num_tokens": 42507352.0,
      "step": 20480
    },
    {
      "entropy": 5.524391555786133,
      "epoch": 1.8582184325108853,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004655864400418014,
      "loss": 5.2536,
      "mean_token_accuracy": 0.18613822609186173,
      "num_tokens": 42517655.0,
      "step": 20485
    },
    {
      "entropy": 5.463539075851441,
      "epoch": 1.8586719883889695,
      "grad_norm": 2.625,
      "learning_rate": 0.0004655692402500163,
      "loss": 5.16,
      "mean_token_accuracy": 0.19564370959997177,
      "num_tokens": 42527502.0,
      "step": 20490
    },
    {
      "entropy": 5.480543327331543,
      "epoch": 1.8591255442670538,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046555203651721085,
      "loss": 5.1984,
      "mean_token_accuracy": 0.1866522029042244,
      "num_tokens": 42538893.0,
      "step": 20495
    },
    {
      "entropy": 5.385703039169312,
      "epoch": 1.8595791001451378,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004655348288437407,
      "loss": 5.096,
      "mean_token_accuracy": 0.19266476184129716,
      "num_tokens": 42550372.0,
      "step": 20500
    },
    {
      "entropy": 5.598840999603271,
      "epoch": 1.860032656023222,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004655176172299618,
      "loss": 5.3181,
      "mean_token_accuracy": 0.18092742413282395,
      "num_tokens": 42560814.0,
      "step": 20505
    },
    {
      "entropy": 5.595803260803223,
      "epoch": 1.8604862119013061,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046550040167622986,
      "loss": 5.3212,
      "mean_token_accuracy": 0.17554933130741118,
      "num_tokens": 42570993.0,
      "step": 20510
    },
    {
      "entropy": 5.532810640335083,
      "epoch": 1.8609397677793904,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046548318218290124,
      "loss": 5.1699,
      "mean_token_accuracy": 0.18102711141109468,
      "num_tokens": 42581357.0,
      "step": 20515
    },
    {
      "entropy": 5.488310623168945,
      "epoch": 1.8613933236574747,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004654659587503318,
      "loss": 5.2866,
      "mean_token_accuracy": 0.1743515819311142,
      "num_tokens": 42591653.0,
      "step": 20520
    },
    {
      "entropy": 5.604928827285766,
      "epoch": 1.861846879535559,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004654487313788778,
      "loss": 5.3321,
      "mean_token_accuracy": 0.1765903800725937,
      "num_tokens": 42602232.0,
      "step": 20525
    },
    {
      "entropy": 5.572270917892456,
      "epoch": 1.862300435413643,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00046543150006889546,
      "loss": 5.2579,
      "mean_token_accuracy": 0.18489742875099183,
      "num_tokens": 42613283.0,
      "step": 20530
    },
    {
      "entropy": 5.494172048568726,
      "epoch": 1.862753991291727,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004654142648207412,
      "loss": 5.1656,
      "mean_token_accuracy": 0.18415360897779465,
      "num_tokens": 42623594.0,
      "step": 20535
    },
    {
      "entropy": 5.539736413955689,
      "epoch": 1.8632075471698113,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004653970256347715,
      "loss": 5.2462,
      "mean_token_accuracy": 0.18466486632823945,
      "num_tokens": 42633587.0,
      "step": 20540
    },
    {
      "entropy": 5.573883771896362,
      "epoch": 1.8636611030478956,
      "grad_norm": 1.5,
      "learning_rate": 0.00046537978251134265,
      "loss": 5.2294,
      "mean_token_accuracy": 0.18109232783317566,
      "num_tokens": 42643426.0,
      "step": 20545
    },
    {
      "entropy": 5.471459817886353,
      "epoch": 1.8641146589259798,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004653625354508115,
      "loss": 5.2239,
      "mean_token_accuracy": 0.1836502358317375,
      "num_tokens": 42653521.0,
      "step": 20550
    },
    {
      "entropy": 5.536389636993408,
      "epoch": 1.8645682148040639,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004653452844535346,
      "loss": 5.2601,
      "mean_token_accuracy": 0.18279604911804198,
      "num_tokens": 42664818.0,
      "step": 20555
    },
    {
      "entropy": 5.620320606231689,
      "epoch": 1.865021770682148,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046532802951986863,
      "loss": 5.3007,
      "mean_token_accuracy": 0.18213828057050704,
      "num_tokens": 42674738.0,
      "step": 20560
    },
    {
      "entropy": 5.546884775161743,
      "epoch": 1.8654753265602322,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004653107706501705,
      "loss": 5.1951,
      "mean_token_accuracy": 0.19272361546754838,
      "num_tokens": 42685003.0,
      "step": 20565
    },
    {
      "entropy": 5.480933475494385,
      "epoch": 1.8659288824383164,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004652935078447973,
      "loss": 5.2451,
      "mean_token_accuracy": 0.18348952978849412,
      "num_tokens": 42695428.0,
      "step": 20570
    },
    {
      "entropy": 5.563340377807617,
      "epoch": 1.8663824383164007,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046527624110410574,
      "loss": 5.2125,
      "mean_token_accuracy": 0.18958747684955596,
      "num_tokens": 42704901.0,
      "step": 20575
    },
    {
      "entropy": 5.562381219863892,
      "epoch": 1.8668359941944848,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000465258970428453,
      "loss": 5.2944,
      "mean_token_accuracy": 0.1768243595957756,
      "num_tokens": 42714379.0,
      "step": 20580
    },
    {
      "entropy": 5.614674282073975,
      "epoch": 1.8672895500725688,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046524169581819633,
      "loss": 5.4114,
      "mean_token_accuracy": 0.17499799132347107,
      "num_tokens": 42724813.0,
      "step": 20585
    },
    {
      "entropy": 5.530964994430542,
      "epoch": 1.867743105950653,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004652244172736929,
      "loss": 5.2408,
      "mean_token_accuracy": 0.18225323557853698,
      "num_tokens": 42736590.0,
      "step": 20590
    },
    {
      "entropy": 5.554121589660644,
      "epoch": 1.8681966618287373,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004652071347953,
      "loss": 5.3195,
      "mean_token_accuracy": 0.1787085637450218,
      "num_tokens": 42746718.0,
      "step": 20595
    },
    {
      "entropy": 5.547988033294677,
      "epoch": 1.8686502177068216,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046518984838337513,
      "loss": 5.2076,
      "mean_token_accuracy": 0.17813289612531663,
      "num_tokens": 42756484.0,
      "step": 20600
    },
    {
      "entropy": 5.5786045551300045,
      "epoch": 1.8691037735849056,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004651725580382757,
      "loss": 5.325,
      "mean_token_accuracy": 0.17880180478096008,
      "num_tokens": 42768512.0,
      "step": 20605
    },
    {
      "entropy": 5.50294861793518,
      "epoch": 1.8695573294629897,
      "grad_norm": 1.84375,
      "learning_rate": 0.00046515526376035934,
      "loss": 5.162,
      "mean_token_accuracy": 0.18474295884370803,
      "num_tokens": 42779812.0,
      "step": 20610
    },
    {
      "entropy": 5.442275619506836,
      "epoch": 1.870010885341074,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004651379655499837,
      "loss": 5.1835,
      "mean_token_accuracy": 0.18176996409893037,
      "num_tokens": 42791163.0,
      "step": 20615
    },
    {
      "entropy": 5.475088500976563,
      "epoch": 1.8704644412191582,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00046512066340750636,
      "loss": 5.218,
      "mean_token_accuracy": 0.18224275559186937,
      "num_tokens": 42801810.0,
      "step": 20620
    },
    {
      "entropy": 5.615349435806275,
      "epoch": 1.8709179970972425,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004651033573332853,
      "loss": 5.2478,
      "mean_token_accuracy": 0.18207058161497117,
      "num_tokens": 42813358.0,
      "step": 20625
    },
    {
      "entropy": 5.449173450469971,
      "epoch": 1.8713715529753265,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004650860473276784,
      "loss": 5.0819,
      "mean_token_accuracy": 0.19518740773200988,
      "num_tokens": 42823519.0,
      "step": 20630
    },
    {
      "entropy": 5.527100992202759,
      "epoch": 1.8718251088534106,
      "grad_norm": 1.375,
      "learning_rate": 0.0004650687333910436,
      "loss": 5.2377,
      "mean_token_accuracy": 0.18469234555959702,
      "num_tokens": 42833257.0,
      "step": 20635
    },
    {
      "entropy": 5.599443006515503,
      "epoch": 1.8722786647314948,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00046505141552373897,
      "loss": 5.3049,
      "mean_token_accuracy": 0.17883420139551162,
      "num_tokens": 42843536.0,
      "step": 20640
    },
    {
      "entropy": 5.514290142059326,
      "epoch": 1.8727322206095791,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004650340937261226,
      "loss": 5.2207,
      "mean_token_accuracy": 0.18087659031152725,
      "num_tokens": 42853911.0,
      "step": 20645
    },
    {
      "entropy": 5.60651330947876,
      "epoch": 1.8731857764876634,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004650167679985528,
      "loss": 5.3366,
      "mean_token_accuracy": 0.17289844006299973,
      "num_tokens": 42864173.0,
      "step": 20650
    },
    {
      "entropy": 5.530979204177856,
      "epoch": 1.8736393323657474,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004649994383413877,
      "loss": 5.1846,
      "mean_token_accuracy": 0.18983518630266188,
      "num_tokens": 42873470.0,
      "step": 20655
    },
    {
      "entropy": 5.496775817871094,
      "epoch": 1.8740928882438317,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00046498210475498595,
      "loss": 5.1681,
      "mean_token_accuracy": 0.19233807772397996,
      "num_tokens": 42882496.0,
      "step": 20660
    },
    {
      "entropy": 5.484168577194214,
      "epoch": 1.8745464441219157,
      "grad_norm": 1.75,
      "learning_rate": 0.00046496476723970574,
      "loss": 5.173,
      "mean_token_accuracy": 0.18492518812417985,
      "num_tokens": 42893163.0,
      "step": 20665
    },
    {
      "entropy": 5.52030577659607,
      "epoch": 1.875,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004649474257959058,
      "loss": 5.2091,
      "mean_token_accuracy": 0.18306216299533845,
      "num_tokens": 42903406.0,
      "step": 20670
    },
    {
      "entropy": 5.575620365142822,
      "epoch": 1.8754535558780843,
      "grad_norm": 2.125,
      "learning_rate": 0.0004649300804239447,
      "loss": 5.195,
      "mean_token_accuracy": 0.1850738450884819,
      "num_tokens": 42913702.0,
      "step": 20675
    },
    {
      "entropy": 5.460291337966919,
      "epoch": 1.8759071117561683,
      "grad_norm": 1.71875,
      "learning_rate": 0.00046491273112418105,
      "loss": 5.1594,
      "mean_token_accuracy": 0.1850622072815895,
      "num_tokens": 42923877.0,
      "step": 20680
    },
    {
      "entropy": 5.460482025146485,
      "epoch": 1.8763606676342526,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046489537789697386,
      "loss": 5.1882,
      "mean_token_accuracy": 0.18675815165042878,
      "num_tokens": 42934393.0,
      "step": 20685
    },
    {
      "entropy": 5.53875765800476,
      "epoch": 1.8768142235123366,
      "grad_norm": 2.625,
      "learning_rate": 0.00046487802074268177,
      "loss": 5.2797,
      "mean_token_accuracy": 0.18384454399347305,
      "num_tokens": 42945803.0,
      "step": 20690
    },
    {
      "entropy": 5.494478225708008,
      "epoch": 1.8772677793904209,
      "grad_norm": 1.375,
      "learning_rate": 0.00046486065966166377,
      "loss": 5.1943,
      "mean_token_accuracy": 0.18626582771539688,
      "num_tokens": 42955829.0,
      "step": 20695
    },
    {
      "entropy": 5.55991563796997,
      "epoch": 1.8777213352685052,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004648432946542791,
      "loss": 5.2553,
      "mean_token_accuracy": 0.1786755234003067,
      "num_tokens": 42967080.0,
      "step": 20700
    },
    {
      "entropy": 5.47931981086731,
      "epoch": 1.8781748911465894,
      "grad_norm": 1.890625,
      "learning_rate": 0.00046482592572088654,
      "loss": 5.2202,
      "mean_token_accuracy": 0.17966048866510392,
      "num_tokens": 42978082.0,
      "step": 20705
    },
    {
      "entropy": 5.505255126953125,
      "epoch": 1.8786284470246735,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00046480855286184565,
      "loss": 5.2762,
      "mean_token_accuracy": 0.18295015543699264,
      "num_tokens": 42988982.0,
      "step": 20710
    },
    {
      "entropy": 5.519764137268067,
      "epoch": 1.8790820029027575,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046479117607751535,
      "loss": 5.2282,
      "mean_token_accuracy": 0.1744844749569893,
      "num_tokens": 42998982.0,
      "step": 20715
    },
    {
      "entropy": 5.550038433074951,
      "epoch": 1.8795355587808418,
      "grad_norm": 3.453125,
      "learning_rate": 0.00046477379536825534,
      "loss": 5.1694,
      "mean_token_accuracy": 0.18558911085128785,
      "num_tokens": 43009574.0,
      "step": 20720
    },
    {
      "entropy": 5.518360567092896,
      "epoch": 1.879989114658926,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004647564107344247,
      "loss": 5.2008,
      "mean_token_accuracy": 0.18252825886011123,
      "num_tokens": 43019335.0,
      "step": 20725
    },
    {
      "entropy": 5.431141567230225,
      "epoch": 1.8804426705370103,
      "grad_norm": 1.859375,
      "learning_rate": 0.00046473902217638327,
      "loss": 5.2575,
      "mean_token_accuracy": 0.1854771465063095,
      "num_tokens": 43029264.0,
      "step": 20730
    },
    {
      "entropy": 5.478175449371338,
      "epoch": 1.8808962264150944,
      "grad_norm": 2.015625,
      "learning_rate": 0.00046472162969449037,
      "loss": 5.2327,
      "mean_token_accuracy": 0.18347976952791215,
      "num_tokens": 43040519.0,
      "step": 20735
    },
    {
      "entropy": 5.61260027885437,
      "epoch": 1.8813497822931784,
      "grad_norm": 1.9375,
      "learning_rate": 0.00046470423328910594,
      "loss": 5.2332,
      "mean_token_accuracy": 0.18816636353731156,
      "num_tokens": 43051146.0,
      "step": 20740
    },
    {
      "entropy": 5.592605495452881,
      "epoch": 1.8818033381712627,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00046468683296058957,
      "loss": 5.3344,
      "mean_token_accuracy": 0.18182787895202637,
      "num_tokens": 43061339.0,
      "step": 20745
    },
    {
      "entropy": 5.472495126724243,
      "epoch": 1.882256894049347,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046466942870930126,
      "loss": 5.2368,
      "mean_token_accuracy": 0.18735036700963975,
      "num_tokens": 43072624.0,
      "step": 20750
    },
    {
      "entropy": 5.489013576507569,
      "epoch": 1.8827104499274312,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004646520205356007,
      "loss": 5.1712,
      "mean_token_accuracy": 0.18891848623752594,
      "num_tokens": 43082120.0,
      "step": 20755
    },
    {
      "entropy": 5.598401260375977,
      "epoch": 1.8831640058055152,
      "grad_norm": 2.15625,
      "learning_rate": 0.00046463460843984805,
      "loss": 5.3233,
      "mean_token_accuracy": 0.17959915697574616,
      "num_tokens": 43093202.0,
      "step": 20760
    },
    {
      "entropy": 5.714721918106079,
      "epoch": 1.8836175616835993,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004646171924224034,
      "loss": 5.4204,
      "mean_token_accuracy": 0.17964942008256912,
      "num_tokens": 43104801.0,
      "step": 20765
    },
    {
      "entropy": 5.578269529342651,
      "epoch": 1.8840711175616836,
      "grad_norm": 2.15625,
      "learning_rate": 0.00046459977248362687,
      "loss": 5.2167,
      "mean_token_accuracy": 0.18485645949840546,
      "num_tokens": 43114855.0,
      "step": 20770
    },
    {
      "entropy": 5.508686590194702,
      "epoch": 1.8845246734397678,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004645823486238788,
      "loss": 5.1921,
      "mean_token_accuracy": 0.17956898212432862,
      "num_tokens": 43124759.0,
      "step": 20775
    },
    {
      "entropy": 5.472143268585205,
      "epoch": 1.884978229317852,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004645649208435193,
      "loss": 5.1911,
      "mean_token_accuracy": 0.1906578779220581,
      "num_tokens": 43135526.0,
      "step": 20780
    },
    {
      "entropy": 5.512765026092529,
      "epoch": 1.8854317851959361,
      "grad_norm": 1.90625,
      "learning_rate": 0.00046454748914290906,
      "loss": 5.3132,
      "mean_token_accuracy": 0.18096586167812348,
      "num_tokens": 43145605.0,
      "step": 20785
    },
    {
      "entropy": 5.59318208694458,
      "epoch": 1.8858853410740202,
      "grad_norm": 1.703125,
      "learning_rate": 0.00046453005352240835,
      "loss": 5.3712,
      "mean_token_accuracy": 0.17699321806430818,
      "num_tokens": 43156578.0,
      "step": 20790
    },
    {
      "entropy": 5.647248935699463,
      "epoch": 1.8863388969521044,
      "grad_norm": 1.4375,
      "learning_rate": 0.000464512613982378,
      "loss": 5.2455,
      "mean_token_accuracy": 0.18705827742815018,
      "num_tokens": 43166870.0,
      "step": 20795
    },
    {
      "entropy": 5.462701511383057,
      "epoch": 1.8867924528301887,
      "grad_norm": 3.484375,
      "learning_rate": 0.0004644951705231783,
      "loss": 5.1848,
      "mean_token_accuracy": 0.19335444420576095,
      "num_tokens": 43176406.0,
      "step": 20800
    },
    {
      "entropy": 5.561164665222168,
      "epoch": 1.887246008708273,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00046447772314517024,
      "loss": 5.3852,
      "mean_token_accuracy": 0.1807613953948021,
      "num_tokens": 43187502.0,
      "step": 20805
    },
    {
      "entropy": 5.503824710845947,
      "epoch": 1.887699564586357,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00046446027184871464,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19027903974056243,
      "num_tokens": 43197928.0,
      "step": 20810
    },
    {
      "entropy": 5.514868211746216,
      "epoch": 1.888153120464441,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046444281663417236,
      "loss": 5.2243,
      "mean_token_accuracy": 0.18992340713739395,
      "num_tokens": 43209083.0,
      "step": 20815
    },
    {
      "entropy": 5.540303659439087,
      "epoch": 1.8886066763425253,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004644253575019043,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1820124864578247,
      "num_tokens": 43219089.0,
      "step": 20820
    },
    {
      "entropy": 5.585845184326172,
      "epoch": 1.8890602322206096,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004644078944522716,
      "loss": 5.2612,
      "mean_token_accuracy": 0.18278798609972,
      "num_tokens": 43229440.0,
      "step": 20825
    },
    {
      "entropy": 5.533747768402099,
      "epoch": 1.8895137880986939,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004643904274856353,
      "loss": 5.2256,
      "mean_token_accuracy": 0.1858087047934532,
      "num_tokens": 43239995.0,
      "step": 20830
    },
    {
      "entropy": 5.4650720119476315,
      "epoch": 1.889967343976778,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00046437295660235674,
      "loss": 5.2033,
      "mean_token_accuracy": 0.1913838118314743,
      "num_tokens": 43249722.0,
      "step": 20835
    },
    {
      "entropy": 5.494582223892212,
      "epoch": 1.8904208998548622,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004643554818027972,
      "loss": 5.2596,
      "mean_token_accuracy": 0.18329727947711943,
      "num_tokens": 43260029.0,
      "step": 20840
    },
    {
      "entropy": 5.568980360031128,
      "epoch": 1.8908744557329462,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004643380030873181,
      "loss": 5.1954,
      "mean_token_accuracy": 0.1860044538974762,
      "num_tokens": 43270571.0,
      "step": 20845
    },
    {
      "entropy": 5.540264654159546,
      "epoch": 1.8913280116110305,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004643205204562808,
      "loss": 5.188,
      "mean_token_accuracy": 0.18714713007211686,
      "num_tokens": 43281310.0,
      "step": 20850
    },
    {
      "entropy": 5.515167760848999,
      "epoch": 1.8917815674891147,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00046430303391004687,
      "loss": 5.2265,
      "mean_token_accuracy": 0.1839125856757164,
      "num_tokens": 43292904.0,
      "step": 20855
    },
    {
      "entropy": 5.550775766372681,
      "epoch": 1.8922351233671988,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004642855434489779,
      "loss": 5.2191,
      "mean_token_accuracy": 0.1820638194680214,
      "num_tokens": 43303745.0,
      "step": 20860
    },
    {
      "entropy": 5.537549018859863,
      "epoch": 1.892688679245283,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046426804907343574,
      "loss": 5.249,
      "mean_token_accuracy": 0.18073318898677826,
      "num_tokens": 43314652.0,
      "step": 20865
    },
    {
      "entropy": 5.495028924942017,
      "epoch": 1.893142235123367,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046425055078378206,
      "loss": 5.1072,
      "mean_token_accuracy": 0.195041286945343,
      "num_tokens": 43324242.0,
      "step": 20870
    },
    {
      "entropy": 5.466238164901734,
      "epoch": 1.8935957910014514,
      "grad_norm": 2.078125,
      "learning_rate": 0.00046423304858037885,
      "loss": 5.2177,
      "mean_token_accuracy": 0.190965473651886,
      "num_tokens": 43334054.0,
      "step": 20875
    },
    {
      "entropy": 5.444868373870849,
      "epoch": 1.8940493468795356,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004642155424635878,
      "loss": 5.1182,
      "mean_token_accuracy": 0.1934479683637619,
      "num_tokens": 43343389.0,
      "step": 20880
    },
    {
      "entropy": 5.531835508346558,
      "epoch": 1.89450290275762,
      "grad_norm": 1.71875,
      "learning_rate": 0.00046419803243377125,
      "loss": 5.1847,
      "mean_token_accuracy": 0.19177436530590058,
      "num_tokens": 43353892.0,
      "step": 20885
    },
    {
      "entropy": 5.437585544586182,
      "epoch": 1.894956458635704,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046418051849129115,
      "loss": 5.2127,
      "mean_token_accuracy": 0.18835393488407134,
      "num_tokens": 43363912.0,
      "step": 20890
    },
    {
      "entropy": 5.478995847702026,
      "epoch": 1.895410014513788,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004641630006365097,
      "loss": 5.2763,
      "mean_token_accuracy": 0.17987176924943923,
      "num_tokens": 43374719.0,
      "step": 20895
    },
    {
      "entropy": 5.563891839981079,
      "epoch": 1.8958635703918723,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00046414547886978926,
      "loss": 5.2553,
      "mean_token_accuracy": 0.18430898934602738,
      "num_tokens": 43385549.0,
      "step": 20900
    },
    {
      "entropy": 5.580017709732056,
      "epoch": 1.8963171262699565,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046412795319149204,
      "loss": 5.2648,
      "mean_token_accuracy": 0.1904647931456566,
      "num_tokens": 43395503.0,
      "step": 20905
    },
    {
      "entropy": 5.514829444885254,
      "epoch": 1.8967706821480408,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00046411042360198053,
      "loss": 5.2925,
      "mean_token_accuracy": 0.17414944022893905,
      "num_tokens": 43406913.0,
      "step": 20910
    },
    {
      "entropy": 5.540830993652344,
      "epoch": 1.8972242380261248,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004640928901016173,
      "loss": 5.1667,
      "mean_token_accuracy": 0.18329977691173555,
      "num_tokens": 43416459.0,
      "step": 20915
    },
    {
      "entropy": 5.500973892211914,
      "epoch": 1.8976777939042089,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046407535269076493,
      "loss": 5.2479,
      "mean_token_accuracy": 0.18456374257802963,
      "num_tokens": 43426280.0,
      "step": 20920
    },
    {
      "entropy": 5.586113452911377,
      "epoch": 1.8981313497822931,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046405781136978606,
      "loss": 5.2756,
      "mean_token_accuracy": 0.18635608851909638,
      "num_tokens": 43436676.0,
      "step": 20925
    },
    {
      "entropy": 5.648950481414795,
      "epoch": 1.8985849056603774,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046404026613904357,
      "loss": 5.2731,
      "mean_token_accuracy": 0.18088891357183456,
      "num_tokens": 43447019.0,
      "step": 20930
    },
    {
      "entropy": 5.556041574478149,
      "epoch": 1.8990384615384617,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004640227169989001,
      "loss": 5.225,
      "mean_token_accuracy": 0.1835897296667099,
      "num_tokens": 43456609.0,
      "step": 20935
    },
    {
      "entropy": 5.499090671539307,
      "epoch": 1.8994920174165457,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004640051639497187,
      "loss": 5.2742,
      "mean_token_accuracy": 0.17909349650144576,
      "num_tokens": 43467616.0,
      "step": 20940
    },
    {
      "entropy": 5.468207359313965,
      "epoch": 1.8999455732946298,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004639876069918624,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18101399391889572,
      "num_tokens": 43478209.0,
      "step": 20945
    },
    {
      "entropy": 5.5190938949584964,
      "epoch": 1.900399129172714,
      "grad_norm": 2.4375,
      "learning_rate": 0.00046397004612569416,
      "loss": 5.1273,
      "mean_token_accuracy": 0.18639697432518004,
      "num_tokens": 43487278.0,
      "step": 20950
    },
    {
      "entropy": 5.4787108421325685,
      "epoch": 1.9008526850507983,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00046395248135157715,
      "loss": 5.1719,
      "mean_token_accuracy": 0.19059406071901322,
      "num_tokens": 43497457.0,
      "step": 20955
    },
    {
      "entropy": 5.534517574310303,
      "epoch": 1.9013062409288826,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046393491266987484,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1835586130619049,
      "num_tokens": 43507761.0,
      "step": 20960
    },
    {
      "entropy": 5.468868160247803,
      "epoch": 1.9017597968069666,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004639173400809502,
      "loss": 5.1854,
      "mean_token_accuracy": 0.17885498851537704,
      "num_tokens": 43517198.0,
      "step": 20965
    },
    {
      "entropy": 5.559218311309815,
      "epoch": 1.9022133526850507,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004638997635851669,
      "loss": 5.2798,
      "mean_token_accuracy": 0.18072159886360167,
      "num_tokens": 43527429.0,
      "step": 20970
    },
    {
      "entropy": 5.600560903549194,
      "epoch": 1.902666908563135,
      "grad_norm": 1.84375,
      "learning_rate": 0.00046388218318288844,
      "loss": 5.2762,
      "mean_token_accuracy": 0.1842792123556137,
      "num_tokens": 43537867.0,
      "step": 20975
    },
    {
      "entropy": 5.530772686004639,
      "epoch": 1.9031204644412192,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004638645988744782,
      "loss": 5.2275,
      "mean_token_accuracy": 0.1826688051223755,
      "num_tokens": 43547284.0,
      "step": 20980
    },
    {
      "entropy": 5.536803483963013,
      "epoch": 1.9035740203193035,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004638470106603,
      "loss": 5.266,
      "mean_token_accuracy": 0.18454342037439347,
      "num_tokens": 43557693.0,
      "step": 20985
    },
    {
      "entropy": 5.547591972351074,
      "epoch": 1.9040275761973875,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004638294185407174,
      "loss": 5.2315,
      "mean_token_accuracy": 0.19201000183820724,
      "num_tokens": 43566909.0,
      "step": 20990
    },
    {
      "entropy": 5.470988130569458,
      "epoch": 1.9044811320754715,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004638118225160944,
      "loss": 5.2923,
      "mean_token_accuracy": 0.18890233635902404,
      "num_tokens": 43576963.0,
      "step": 20995
    },
    {
      "entropy": 5.439335203170776,
      "epoch": 1.9049346879535558,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00046379422258679477,
      "loss": 5.2011,
      "mean_token_accuracy": 0.1840580567717552,
      "num_tokens": 43586649.0,
      "step": 21000
    },
    {
      "epoch": 1.9049346879535558,
      "eval_entropy": 5.3283459365637595,
      "eval_loss": 5.299813270568848,
      "eval_mean_token_accuracy": 0.18859136360802803,
      "eval_num_tokens": 43586649.0,
      "eval_runtime": 26.5301,
      "eval_samples_per_second": 1332.827,
      "eval_steps_per_second": 166.603,
      "step": 21000
    },
    {
      "entropy": 5.513235473632813,
      "epoch": 1.90538824383164,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004637766187531825,
      "loss": 5.1966,
      "mean_token_accuracy": 0.18809781819581986,
      "num_tokens": 43596506.0,
      "step": 21005
    },
    {
      "entropy": 5.539356136322022,
      "epoch": 1.9058417997097243,
      "grad_norm": 1.75,
      "learning_rate": 0.00046375901101562163,
      "loss": 5.2597,
      "mean_token_accuracy": 0.1829143211245537,
      "num_tokens": 43607188.0,
      "step": 21010
    },
    {
      "entropy": 5.541349077224732,
      "epoch": 1.9062953555878084,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004637413993744763,
      "loss": 5.1684,
      "mean_token_accuracy": 0.1943049266934395,
      "num_tokens": 43617465.0,
      "step": 21015
    },
    {
      "entropy": 5.457902097702027,
      "epoch": 1.9067489114658926,
      "grad_norm": 2.390625,
      "learning_rate": 0.00046372378383011077,
      "loss": 5.187,
      "mean_token_accuracy": 0.18874061703681946,
      "num_tokens": 43627692.0,
      "step": 21020
    },
    {
      "entropy": 5.53016300201416,
      "epoch": 1.9072024673439767,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046370616438288933,
      "loss": 5.2752,
      "mean_token_accuracy": 0.17710960805416107,
      "num_tokens": 43637350.0,
      "step": 21025
    },
    {
      "entropy": 5.573669195175171,
      "epoch": 1.907656023222061,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00046368854103317624,
      "loss": 5.2366,
      "mean_token_accuracy": 0.18209994733333587,
      "num_tokens": 43647092.0,
      "step": 21030
    },
    {
      "entropy": 5.548462581634522,
      "epoch": 1.9081095791001452,
      "grad_norm": 2.109375,
      "learning_rate": 0.00046367091378133615,
      "loss": 5.2059,
      "mean_token_accuracy": 0.18755248188972473,
      "num_tokens": 43657237.0,
      "step": 21035
    },
    {
      "entropy": 5.504029369354248,
      "epoch": 1.9085631349782293,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004636532826277333,
      "loss": 5.2187,
      "mean_token_accuracy": 0.18418193608522415,
      "num_tokens": 43667966.0,
      "step": 21040
    },
    {
      "entropy": 5.48861346244812,
      "epoch": 1.9090166908563135,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00046363564757273264,
      "loss": 5.2091,
      "mean_token_accuracy": 0.186693774163723,
      "num_tokens": 43678059.0,
      "step": 21045
    },
    {
      "entropy": 5.526192617416382,
      "epoch": 1.9094702467343976,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046361800861669867,
      "loss": 5.2212,
      "mean_token_accuracy": 0.18879299759864807,
      "num_tokens": 43687556.0,
      "step": 21050
    },
    {
      "entropy": 5.573375558853149,
      "epoch": 1.9099238026124818,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00046360036575999623,
      "loss": 5.2544,
      "mean_token_accuracy": 0.18251536339521407,
      "num_tokens": 43697791.0,
      "step": 21055
    },
    {
      "entropy": 5.499818277359009,
      "epoch": 1.9103773584905661,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004635827190029902,
      "loss": 5.1423,
      "mean_token_accuracy": 0.18465647250413894,
      "num_tokens": 43707390.0,
      "step": 21060
    },
    {
      "entropy": 5.590476322174072,
      "epoch": 1.9108309143686504,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004635650683460454,
      "loss": 5.2958,
      "mean_token_accuracy": 0.17967903316020967,
      "num_tokens": 43716842.0,
      "step": 21065
    },
    {
      "entropy": 5.510005283355713,
      "epoch": 1.9112844702467344,
      "grad_norm": 1.71875,
      "learning_rate": 0.000463547413789527,
      "loss": 5.2195,
      "mean_token_accuracy": 0.18445561081171036,
      "num_tokens": 43726967.0,
      "step": 21070
    },
    {
      "entropy": 5.567798900604248,
      "epoch": 1.9117380261248185,
      "grad_norm": 1.84375,
      "learning_rate": 0.00046352975533379997,
      "loss": 5.2265,
      "mean_token_accuracy": 0.18739120811223983,
      "num_tokens": 43737425.0,
      "step": 21075
    },
    {
      "entropy": 5.398691177368164,
      "epoch": 1.9121915820029027,
      "grad_norm": 1.953125,
      "learning_rate": 0.00046351209297922964,
      "loss": 5.1553,
      "mean_token_accuracy": 0.19358859062194825,
      "num_tokens": 43747395.0,
      "step": 21080
    },
    {
      "entropy": 5.539512300491333,
      "epoch": 1.912645137880987,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046349442672618114,
      "loss": 5.2528,
      "mean_token_accuracy": 0.17830377966165542,
      "num_tokens": 43756348.0,
      "step": 21085
    },
    {
      "entropy": 5.5730705738067625,
      "epoch": 1.9130986937590713,
      "grad_norm": 2.0,
      "learning_rate": 0.00046347675657501986,
      "loss": 5.21,
      "mean_token_accuracy": 0.18450499325990677,
      "num_tokens": 43767629.0,
      "step": 21090
    },
    {
      "entropy": 5.473129224777222,
      "epoch": 1.9135522496371553,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004634590825261113,
      "loss": 5.1338,
      "mean_token_accuracy": 0.19277871996164322,
      "num_tokens": 43777398.0,
      "step": 21095
    },
    {
      "entropy": 5.514169120788575,
      "epoch": 1.9140058055152394,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046344140457982075,
      "loss": 5.2521,
      "mean_token_accuracy": 0.18332654982805252,
      "num_tokens": 43787605.0,
      "step": 21100
    },
    {
      "entropy": 5.527050924301148,
      "epoch": 1.9144593613933236,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000463423722736514,
      "loss": 5.2181,
      "mean_token_accuracy": 0.1874515727162361,
      "num_tokens": 43796989.0,
      "step": 21105
    },
    {
      "entropy": 5.564722681045533,
      "epoch": 1.9149129172714079,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004634060369965567,
      "loss": 5.2408,
      "mean_token_accuracy": 0.18568496704101561,
      "num_tokens": 43806653.0,
      "step": 21110
    },
    {
      "entropy": 5.486712312698364,
      "epoch": 1.9153664731494922,
      "grad_norm": 2.328125,
      "learning_rate": 0.00046338834736031446,
      "loss": 5.1588,
      "mean_token_accuracy": 0.18663265705108642,
      "num_tokens": 43817152.0,
      "step": 21115
    },
    {
      "entropy": 5.4952113151550295,
      "epoch": 1.9158200290275762,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00046337065382815324,
      "loss": 5.1616,
      "mean_token_accuracy": 0.18696332424879075,
      "num_tokens": 43827341.0,
      "step": 21120
    },
    {
      "entropy": 5.469807243347168,
      "epoch": 1.9162735849056602,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046335295640043887,
      "loss": 5.1794,
      "mean_token_accuracy": 0.19245145916938783,
      "num_tokens": 43837195.0,
      "step": 21125
    },
    {
      "entropy": 5.539794540405273,
      "epoch": 1.9167271407837445,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046333525507753743,
      "loss": 5.1645,
      "mean_token_accuracy": 0.18886090070009232,
      "num_tokens": 43847473.0,
      "step": 21130
    },
    {
      "entropy": 5.459527587890625,
      "epoch": 1.9171806966618288,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004633175498598148,
      "loss": 5.1891,
      "mean_token_accuracy": 0.18221555054187774,
      "num_tokens": 43857590.0,
      "step": 21135
    },
    {
      "entropy": 5.536103773117065,
      "epoch": 1.917634252539913,
      "grad_norm": 1.703125,
      "learning_rate": 0.00046329984074763733,
      "loss": 5.334,
      "mean_token_accuracy": 0.17719638645648955,
      "num_tokens": 43867085.0,
      "step": 21140
    },
    {
      "entropy": 5.550732803344727,
      "epoch": 1.918087808417997,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004632821277413711,
      "loss": 5.1635,
      "mean_token_accuracy": 0.18615642338991165,
      "num_tokens": 43877153.0,
      "step": 21145
    },
    {
      "entropy": 5.631152820587158,
      "epoch": 1.9185413642960811,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004632644108413825,
      "loss": 5.3048,
      "mean_token_accuracy": 0.16588189005851744,
      "num_tokens": 43887977.0,
      "step": 21150
    },
    {
      "entropy": 5.468219327926636,
      "epoch": 1.9189949201741654,
      "grad_norm": 2.078125,
      "learning_rate": 0.00046324669004803803,
      "loss": 5.2006,
      "mean_token_accuracy": 0.1856427848339081,
      "num_tokens": 43898649.0,
      "step": 21155
    },
    {
      "entropy": 5.552229976654052,
      "epoch": 1.9194484760522497,
      "grad_norm": 1.640625,
      "learning_rate": 0.000463228965361704,
      "loss": 5.3003,
      "mean_token_accuracy": 0.18038928359746934,
      "num_tokens": 43909369.0,
      "step": 21160
    },
    {
      "entropy": 5.552200508117676,
      "epoch": 1.919902031930334,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004632112367827469,
      "loss": 5.2116,
      "mean_token_accuracy": 0.18755275756120682,
      "num_tokens": 43919941.0,
      "step": 21165
    },
    {
      "entropy": 5.567319679260254,
      "epoch": 1.920355587808418,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004631935043115335,
      "loss": 5.2666,
      "mean_token_accuracy": 0.1831866979598999,
      "num_tokens": 43929710.0,
      "step": 21170
    },
    {
      "entropy": 5.504434299468994,
      "epoch": 1.920809143686502,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00046317576794843056,
      "loss": 5.183,
      "mean_token_accuracy": 0.1818172186613083,
      "num_tokens": 43939850.0,
      "step": 21175
    },
    {
      "entropy": 5.501026773452759,
      "epoch": 1.9212626995645863,
      "grad_norm": 2.0,
      "learning_rate": 0.00046315802769380474,
      "loss": 5.1288,
      "mean_token_accuracy": 0.18796159476041793,
      "num_tokens": 43950067.0,
      "step": 21180
    },
    {
      "entropy": 5.483568906784058,
      "epoch": 1.9217162554426706,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004631402835480229,
      "loss": 5.2061,
      "mean_token_accuracy": 0.19073150157928467,
      "num_tokens": 43960029.0,
      "step": 21185
    },
    {
      "entropy": 5.49980411529541,
      "epoch": 1.9221698113207548,
      "grad_norm": 2.03125,
      "learning_rate": 0.00046312253551145213,
      "loss": 5.222,
      "mean_token_accuracy": 0.186762972176075,
      "num_tokens": 43971221.0,
      "step": 21190
    },
    {
      "entropy": 5.531082344055176,
      "epoch": 1.9226233671988389,
      "grad_norm": 2.546875,
      "learning_rate": 0.00046310478358445935,
      "loss": 5.2576,
      "mean_token_accuracy": 0.17998868525028228,
      "num_tokens": 43981579.0,
      "step": 21195
    },
    {
      "entropy": 5.605056953430176,
      "epoch": 1.9230769230769231,
      "grad_norm": 1.75,
      "learning_rate": 0.0004630870277674117,
      "loss": 5.3399,
      "mean_token_accuracy": 0.17578468471765518,
      "num_tokens": 43991132.0,
      "step": 21200
    },
    {
      "entropy": 5.515637159347534,
      "epoch": 1.9235304789550072,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004630692680606764,
      "loss": 5.138,
      "mean_token_accuracy": 0.19137988090515137,
      "num_tokens": 44001255.0,
      "step": 21205
    },
    {
      "entropy": 5.422712230682373,
      "epoch": 1.9239840348330914,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004630515044646207,
      "loss": 5.0633,
      "mean_token_accuracy": 0.1926187038421631,
      "num_tokens": 44011275.0,
      "step": 21210
    },
    {
      "entropy": 5.4819597721099855,
      "epoch": 1.9244375907111757,
      "grad_norm": 1.9453125,
      "learning_rate": 0.000463033736979612,
      "loss": 5.234,
      "mean_token_accuracy": 0.1890595003962517,
      "num_tokens": 44022524.0,
      "step": 21215
    },
    {
      "entropy": 5.490801906585693,
      "epoch": 1.9248911465892597,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046301596560601767,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1831893026828766,
      "num_tokens": 44033955.0,
      "step": 21220
    },
    {
      "entropy": 5.520248651504517,
      "epoch": 1.925344702467344,
      "grad_norm": 2.0,
      "learning_rate": 0.0004629981903442052,
      "loss": 5.2663,
      "mean_token_accuracy": 0.1813955321907997,
      "num_tokens": 44045149.0,
      "step": 21225
    },
    {
      "entropy": 5.555677604675293,
      "epoch": 1.925798258345428,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004629804111945424,
      "loss": 5.2353,
      "mean_token_accuracy": 0.18255945593118666,
      "num_tokens": 44055280.0,
      "step": 21230
    },
    {
      "entropy": 5.521226930618286,
      "epoch": 1.9262518142235123,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004629626281573966,
      "loss": 5.2015,
      "mean_token_accuracy": 0.18622561395168305,
      "num_tokens": 44065578.0,
      "step": 21235
    },
    {
      "entropy": 5.447855758666992,
      "epoch": 1.9267053701015966,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004629448412331359,
      "loss": 5.1626,
      "mean_token_accuracy": 0.1926008477807045,
      "num_tokens": 44075454.0,
      "step": 21240
    },
    {
      "entropy": 5.480479288101196,
      "epoch": 1.9271589259796809,
      "grad_norm": 4.125,
      "learning_rate": 0.0004629270504221279,
      "loss": 5.2317,
      "mean_token_accuracy": 0.18337714970111846,
      "num_tokens": 44085680.0,
      "step": 21245
    },
    {
      "entropy": 5.447093677520752,
      "epoch": 1.927612481857765,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004629092557247407,
      "loss": 5.1562,
      "mean_token_accuracy": 0.1901509553194046,
      "num_tokens": 44097177.0,
      "step": 21250
    },
    {
      "entropy": 5.529837894439697,
      "epoch": 1.928066037735849,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004628914571413422,
      "loss": 5.2814,
      "mean_token_accuracy": 0.18412837982177735,
      "num_tokens": 44107251.0,
      "step": 21255
    },
    {
      "entropy": 5.536059093475342,
      "epoch": 1.9285195936139332,
      "grad_norm": 2.265625,
      "learning_rate": 0.00046287365467230035,
      "loss": 5.2364,
      "mean_token_accuracy": 0.18229172080755235,
      "num_tokens": 44117943.0,
      "step": 21260
    },
    {
      "entropy": 5.592858934402466,
      "epoch": 1.9289731494920175,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046285584831798354,
      "loss": 5.277,
      "mean_token_accuracy": 0.18355348110198974,
      "num_tokens": 44128608.0,
      "step": 21265
    },
    {
      "entropy": 5.557268238067627,
      "epoch": 1.9294267053701017,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046283803807875996,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1861183002591133,
      "num_tokens": 44138809.0,
      "step": 21270
    },
    {
      "entropy": 5.505529975891113,
      "epoch": 1.9298802612481858,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046282022395499785,
      "loss": 5.2205,
      "mean_token_accuracy": 0.18793853521347045,
      "num_tokens": 44148532.0,
      "step": 21275
    },
    {
      "entropy": 5.534604597091675,
      "epoch": 1.9303338171262698,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004628024059470656,
      "loss": 5.2616,
      "mean_token_accuracy": 0.1845951110124588,
      "num_tokens": 44158628.0,
      "step": 21280
    },
    {
      "entropy": 5.463634729385376,
      "epoch": 1.930787373004354,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004627845840553318,
      "loss": 5.1112,
      "mean_token_accuracy": 0.19445031583309175,
      "num_tokens": 44168574.0,
      "step": 21285
    },
    {
      "entropy": 5.549312162399292,
      "epoch": 1.9312409288824384,
      "grad_norm": 1.625,
      "learning_rate": 0.00046276675828016493,
      "loss": 5.2938,
      "mean_token_accuracy": 0.18451836705207825,
      "num_tokens": 44180207.0,
      "step": 21290
    },
    {
      "entropy": 5.5893511295318605,
      "epoch": 1.9316944847605226,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004627489286219336,
      "loss": 5.3368,
      "mean_token_accuracy": 0.1727435678243637,
      "num_tokens": 44190473.0,
      "step": 21295
    },
    {
      "entropy": 5.6309246063232425,
      "epoch": 1.9321480406386067,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004627310950810066,
      "loss": 5.2779,
      "mean_token_accuracy": 0.18745171129703522,
      "num_tokens": 44201150.0,
      "step": 21300
    },
    {
      "entropy": 5.489469718933106,
      "epoch": 1.9326015965166907,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004627132576577528,
      "loss": 5.2174,
      "mean_token_accuracy": 0.18337173163890838,
      "num_tokens": 44211272.0,
      "step": 21305
    },
    {
      "entropy": 5.539886045455932,
      "epoch": 1.933055152394775,
      "grad_norm": 1.9609375,
      "learning_rate": 0.000462695416352541,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1773107022047043,
      "num_tokens": 44223556.0,
      "step": 21310
    },
    {
      "entropy": 5.544064664840699,
      "epoch": 1.9335087082728593,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00046267757116574,
      "loss": 5.1911,
      "mean_token_accuracy": 0.18819284737110137,
      "num_tokens": 44233911.0,
      "step": 21315
    },
    {
      "entropy": 5.5880101203918455,
      "epoch": 1.9339622641509435,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004626597220977191,
      "loss": 5.2665,
      "mean_token_accuracy": 0.18593984097242355,
      "num_tokens": 44244108.0,
      "step": 21320
    },
    {
      "entropy": 5.422260904312134,
      "epoch": 1.9344158200290276,
      "grad_norm": 2.984375,
      "learning_rate": 0.00046264186914884737,
      "loss": 5.1659,
      "mean_token_accuracy": 0.1846998780965805,
      "num_tokens": 44253604.0,
      "step": 21325
    },
    {
      "entropy": 5.44390082359314,
      "epoch": 1.9348693759071116,
      "grad_norm": 1.640625,
      "learning_rate": 0.00046262401231949386,
      "loss": 5.2707,
      "mean_token_accuracy": 0.17849959135055543,
      "num_tokens": 44263573.0,
      "step": 21330
    },
    {
      "entropy": 5.544204139709473,
      "epoch": 1.9353229317851959,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046260615161002803,
      "loss": 5.2109,
      "mean_token_accuracy": 0.1856840506196022,
      "num_tokens": 44274071.0,
      "step": 21335
    },
    {
      "entropy": 5.616547775268555,
      "epoch": 1.9357764876632801,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004625882870208192,
      "loss": 5.2376,
      "mean_token_accuracy": 0.1823873668909073,
      "num_tokens": 44283997.0,
      "step": 21340
    },
    {
      "entropy": 5.499124670028687,
      "epoch": 1.9362300435413644,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046257041855223674,
      "loss": 5.2125,
      "mean_token_accuracy": 0.1839566260576248,
      "num_tokens": 44295024.0,
      "step": 21345
    },
    {
      "entropy": 5.4667338848114015,
      "epoch": 1.9366835994194485,
      "grad_norm": 1.765625,
      "learning_rate": 0.00046255254620465033,
      "loss": 5.1581,
      "mean_token_accuracy": 0.1919988974928856,
      "num_tokens": 44305476.0,
      "step": 21350
    },
    {
      "entropy": 5.497404718399048,
      "epoch": 1.9371371552975325,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046253466997842936,
      "loss": 5.1455,
      "mean_token_accuracy": 0.18601612448692323,
      "num_tokens": 44316101.0,
      "step": 21355
    },
    {
      "entropy": 5.451179313659668,
      "epoch": 1.9375907111756168,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004625167898739436,
      "loss": 5.0936,
      "mean_token_accuracy": 0.19413564801216127,
      "num_tokens": 44325058.0,
      "step": 21360
    },
    {
      "entropy": 5.667487001419067,
      "epoch": 1.938044267053701,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000462498905891563,
      "loss": 5.4136,
      "mean_token_accuracy": 0.17001090198755264,
      "num_tokens": 44336585.0,
      "step": 21365
    },
    {
      "entropy": 5.569233512878418,
      "epoch": 1.9384978229317853,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004624810180316571,
      "loss": 5.2186,
      "mean_token_accuracy": 0.1870625749230385,
      "num_tokens": 44346911.0,
      "step": 21370
    },
    {
      "entropy": 5.461955165863037,
      "epoch": 1.9389513788098693,
      "grad_norm": 1.859375,
      "learning_rate": 0.00046246312629459596,
      "loss": 5.1864,
      "mean_token_accuracy": 0.18599557429552077,
      "num_tokens": 44358040.0,
      "step": 21375
    },
    {
      "entropy": 5.525521516799927,
      "epoch": 1.9394049346879536,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004624452306807497,
      "loss": 5.202,
      "mean_token_accuracy": 0.19044436067342757,
      "num_tokens": 44368781.0,
      "step": 21380
    },
    {
      "entropy": 5.447435283660889,
      "epoch": 1.9398584905660377,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004624273311904882,
      "loss": 5.1439,
      "mean_token_accuracy": 0.1890501469373703,
      "num_tokens": 44378809.0,
      "step": 21385
    },
    {
      "entropy": 5.569868564605713,
      "epoch": 1.940312046444122,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046240942782418184,
      "loss": 5.2841,
      "mean_token_accuracy": 0.18139278888702393,
      "num_tokens": 44388517.0,
      "step": 21390
    },
    {
      "entropy": 5.59320821762085,
      "epoch": 1.9407656023222062,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004623915205822007,
      "loss": 5.2568,
      "mean_token_accuracy": 0.18316439986228944,
      "num_tokens": 44399364.0,
      "step": 21395
    },
    {
      "entropy": 5.503791809082031,
      "epoch": 1.9412191582002902,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00046237360946491514,
      "loss": 5.1625,
      "mean_token_accuracy": 0.18004566729068755,
      "num_tokens": 44409506.0,
      "step": 21400
    },
    {
      "entropy": 5.551319122314453,
      "epoch": 1.9416727140783745,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004623556944726956,
      "loss": 5.2241,
      "mean_token_accuracy": 0.1860983654856682,
      "num_tokens": 44419446.0,
      "step": 21405
    },
    {
      "entropy": 5.455884265899658,
      "epoch": 1.9421262699564585,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046233777560591256,
      "loss": 5.2091,
      "mean_token_accuracy": 0.18064884841442108,
      "num_tokens": 44429560.0,
      "step": 21410
    },
    {
      "entropy": 5.532247018814087,
      "epoch": 1.9425798258345428,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00046231985286493654,
      "loss": 5.1908,
      "mean_token_accuracy": 0.18096156865358354,
      "num_tokens": 44439640.0,
      "step": 21415
    },
    {
      "entropy": 5.557065629959107,
      "epoch": 1.943033381712627,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046230192625013836,
      "loss": 5.2247,
      "mean_token_accuracy": 0.1828388050198555,
      "num_tokens": 44450480.0,
      "step": 21420
    },
    {
      "entropy": 5.45677080154419,
      "epoch": 1.9434869375907113,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046228399576188844,
      "loss": 5.0998,
      "mean_token_accuracy": 0.185494127869606,
      "num_tokens": 44460633.0,
      "step": 21425
    },
    {
      "entropy": 5.537350416183472,
      "epoch": 1.9439404934687954,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046226606140055784,
      "loss": 5.1769,
      "mean_token_accuracy": 0.18674484491348267,
      "num_tokens": 44470917.0,
      "step": 21430
    },
    {
      "entropy": 5.520677661895752,
      "epoch": 1.9443940493468794,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046224812316651736,
      "loss": 5.2523,
      "mean_token_accuracy": 0.1788308650255203,
      "num_tokens": 44482538.0,
      "step": 21435
    },
    {
      "entropy": 5.53683123588562,
      "epoch": 1.9448476052249637,
      "grad_norm": 1.84375,
      "learning_rate": 0.000462230181060138,
      "loss": 5.2684,
      "mean_token_accuracy": 0.18148650974035263,
      "num_tokens": 44493518.0,
      "step": 21440
    },
    {
      "entropy": 5.541353797912597,
      "epoch": 1.945301161103048,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046221223508179077,
      "loss": 5.2467,
      "mean_token_accuracy": 0.18390317410230636,
      "num_tokens": 44503685.0,
      "step": 21445
    },
    {
      "entropy": 5.492373752593994,
      "epoch": 1.9457547169811322,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00046219428523184685,
      "loss": 5.1856,
      "mean_token_accuracy": 0.1904973417520523,
      "num_tokens": 44514015.0,
      "step": 21450
    },
    {
      "entropy": 5.502481460571289,
      "epoch": 1.9462082728592163,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004621763315106774,
      "loss": 5.2258,
      "mean_token_accuracy": 0.18745078295469284,
      "num_tokens": 44525187.0,
      "step": 21455
    },
    {
      "entropy": 5.457600545883179,
      "epoch": 1.9466618287373003,
      "grad_norm": 1.90625,
      "learning_rate": 0.00046215837391865364,
      "loss": 5.1872,
      "mean_token_accuracy": 0.18981099277734756,
      "num_tokens": 44535964.0,
      "step": 21460
    },
    {
      "entropy": 5.449596881866455,
      "epoch": 1.9471153846153846,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00046214041245614713,
      "loss": 5.1807,
      "mean_token_accuracy": 0.18191886991262435,
      "num_tokens": 44546985.0,
      "step": 21465
    },
    {
      "entropy": 5.543500328063965,
      "epoch": 1.9475689404934688,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004621224471235291,
      "loss": 5.2566,
      "mean_token_accuracy": 0.17886385172605515,
      "num_tokens": 44556888.0,
      "step": 21470
    },
    {
      "entropy": 5.512652111053467,
      "epoch": 1.9480224963715531,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046210447792117114,
      "loss": 5.198,
      "mean_token_accuracy": 0.18476185202598572,
      "num_tokens": 44566857.0,
      "step": 21475
    },
    {
      "entropy": 5.559841680526733,
      "epoch": 1.9484760522496372,
      "grad_norm": 2.1875,
      "learning_rate": 0.000462086504849445,
      "loss": 5.2444,
      "mean_token_accuracy": 0.18033508956432343,
      "num_tokens": 44578152.0,
      "step": 21480
    },
    {
      "entropy": 5.49734935760498,
      "epoch": 1.9489296081277212,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004620685279087222,
      "loss": 5.1862,
      "mean_token_accuracy": 0.19175361096858978,
      "num_tokens": 44588814.0,
      "step": 21485
    },
    {
      "entropy": 5.459387683868409,
      "epoch": 1.9493831640058055,
      "grad_norm": 2.59375,
      "learning_rate": 0.00046205054709937474,
      "loss": 5.1827,
      "mean_token_accuracy": 0.18409885466098785,
      "num_tokens": 44599366.0,
      "step": 21490
    },
    {
      "entropy": 5.625522041320801,
      "epoch": 1.9498367198838897,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004620325624217741,
      "loss": 5.295,
      "mean_token_accuracy": 0.1776812732219696,
      "num_tokens": 44609962.0,
      "step": 21495
    },
    {
      "entropy": 5.601441192626953,
      "epoch": 1.950290275761974,
      "grad_norm": 1.5,
      "learning_rate": 0.0004620145738762925,
      "loss": 5.2453,
      "mean_token_accuracy": 0.18335151225328444,
      "num_tokens": 44619917.0,
      "step": 21500
    },
    {
      "entropy": 5.421952390670777,
      "epoch": 1.950743831640058,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004619965814633019,
      "loss": 5.2078,
      "mean_token_accuracy": 0.18688548654317855,
      "num_tokens": 44629723.0,
      "step": 21505
    },
    {
      "entropy": 5.497846031188965,
      "epoch": 1.951197387518142,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004619785851831742,
      "loss": 5.2599,
      "mean_token_accuracy": 0.1862933650612831,
      "num_tokens": 44641011.0,
      "step": 21510
    },
    {
      "entropy": 5.530605983734131,
      "epoch": 1.9516509433962264,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046196058503628176,
      "loss": 5.176,
      "mean_token_accuracy": 0.18789939880371093,
      "num_tokens": 44651154.0,
      "step": 21515
    },
    {
      "entropy": 5.556767559051513,
      "epoch": 1.9521044992743106,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004619425810229968,
      "loss": 5.2816,
      "mean_token_accuracy": 0.18097535520792007,
      "num_tokens": 44661672.0,
      "step": 21520
    },
    {
      "entropy": 5.472743940353394,
      "epoch": 1.952558055152395,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004619245731436917,
      "loss": 5.1654,
      "mean_token_accuracy": 0.19322258979082108,
      "num_tokens": 44671449.0,
      "step": 21525
    },
    {
      "entropy": 5.477061319351196,
      "epoch": 1.953011611030479,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046190656139873863,
      "loss": 5.1555,
      "mean_token_accuracy": 0.19122230261564255,
      "num_tokens": 44681977.0,
      "step": 21530
    },
    {
      "entropy": 5.459119653701782,
      "epoch": 1.953465166908563,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004618885457885103,
      "loss": 5.191,
      "mean_token_accuracy": 0.18866928815841674,
      "num_tokens": 44691946.0,
      "step": 21535
    },
    {
      "entropy": 5.520205068588257,
      "epoch": 1.9539187227866472,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004618705263133793,
      "loss": 5.2294,
      "mean_token_accuracy": 0.18194324523210526,
      "num_tokens": 44702012.0,
      "step": 21540
    },
    {
      "entropy": 5.501115560531616,
      "epoch": 1.9543722786647315,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004618525029737182,
      "loss": 5.1658,
      "mean_token_accuracy": 0.18788416534662247,
      "num_tokens": 44711605.0,
      "step": 21545
    },
    {
      "entropy": 5.524523448944092,
      "epoch": 1.9548258345428158,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046183447576989966,
      "loss": 5.2717,
      "mean_token_accuracy": 0.1871534764766693,
      "num_tokens": 44722765.0,
      "step": 21550
    },
    {
      "entropy": 5.537755489349365,
      "epoch": 1.9552793904208998,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00046181644470229656,
      "loss": 5.199,
      "mean_token_accuracy": 0.18745887130498887,
      "num_tokens": 44732082.0,
      "step": 21555
    },
    {
      "entropy": 5.434156990051269,
      "epoch": 1.955732946298984,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004617984097712817,
      "loss": 5.1179,
      "mean_token_accuracy": 0.19392033517360688,
      "num_tokens": 44742054.0,
      "step": 21560
    },
    {
      "entropy": 5.482811117172242,
      "epoch": 1.9561865021770681,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004617803709772282,
      "loss": 5.2437,
      "mean_token_accuracy": 0.18608320355415345,
      "num_tokens": 44752954.0,
      "step": 21565
    },
    {
      "entropy": 5.531443166732788,
      "epoch": 1.9566400580551524,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004617623283205091,
      "loss": 5.2853,
      "mean_token_accuracy": 0.1799483448266983,
      "num_tokens": 44764277.0,
      "step": 21570
    },
    {
      "entropy": 5.549055480957032,
      "epoch": 1.9570936139332367,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046174428180149733,
      "loss": 5.2815,
      "mean_token_accuracy": 0.18125036507844924,
      "num_tokens": 44775429.0,
      "step": 21575
    },
    {
      "entropy": 5.587586402893066,
      "epoch": 1.9575471698113207,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004617262314205663,
      "loss": 5.2603,
      "mean_token_accuracy": 0.1768353208899498,
      "num_tokens": 44786390.0,
      "step": 21580
    },
    {
      "entropy": 5.512666463851929,
      "epoch": 1.958000725689405,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004617081771780893,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19096047580242156,
      "num_tokens": 44796473.0,
      "step": 21585
    },
    {
      "entropy": 5.498332118988037,
      "epoch": 1.958454281567489,
      "grad_norm": 2.1875,
      "learning_rate": 0.00046169011907443944,
      "loss": 5.1718,
      "mean_token_accuracy": 0.18823871165513992,
      "num_tokens": 44806379.0,
      "step": 21590
    },
    {
      "entropy": 5.4884007453918455,
      "epoch": 1.9589078374455733,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004616720571099904,
      "loss": 5.2534,
      "mean_token_accuracy": 0.18040031790733338,
      "num_tokens": 44817659.0,
      "step": 21595
    },
    {
      "entropy": 5.541465091705322,
      "epoch": 1.9593613933236576,
      "grad_norm": 2.0,
      "learning_rate": 0.0004616539912851157,
      "loss": 5.1602,
      "mean_token_accuracy": 0.179518423974514,
      "num_tokens": 44827073.0,
      "step": 21600
    },
    {
      "entropy": 5.507293081283569,
      "epoch": 1.9598149492017418,
      "grad_norm": 1.71875,
      "learning_rate": 0.00046163592160018886,
      "loss": 5.1765,
      "mean_token_accuracy": 0.18562255203723907,
      "num_tokens": 44837019.0,
      "step": 21605
    },
    {
      "entropy": 5.4569305896759035,
      "epoch": 1.9602685050798259,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004616178480555836,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1859275445342064,
      "num_tokens": 44846670.0,
      "step": 21610
    },
    {
      "entropy": 5.3797026634216305,
      "epoch": 1.96072206095791,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004615997706516737,
      "loss": 5.1242,
      "mean_token_accuracy": 0.18921856135129927,
      "num_tokens": 44857016.0,
      "step": 21615
    },
    {
      "entropy": 5.588710498809815,
      "epoch": 1.9611756168359942,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000461581689388833,
      "loss": 5.2875,
      "mean_token_accuracy": 0.17903646677732468,
      "num_tokens": 44868180.0,
      "step": 21620
    },
    {
      "entropy": 5.512429237365723,
      "epoch": 1.9616291727140784,
      "grad_norm": 1.375,
      "learning_rate": 0.00046156360426743536,
      "loss": 5.1475,
      "mean_token_accuracy": 0.19138599038124085,
      "num_tokens": 44877864.0,
      "step": 21625
    },
    {
      "entropy": 5.478376817703247,
      "epoch": 1.9620827285921627,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004615455152878549,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19185684025287628,
      "num_tokens": 44887737.0,
      "step": 21630
    },
    {
      "entropy": 5.4673535346984865,
      "epoch": 1.9625362844702468,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004615274224504657,
      "loss": 5.2137,
      "mean_token_accuracy": 0.18118021339178086,
      "num_tokens": 44898669.0,
      "step": 21635
    },
    {
      "entropy": 5.54384298324585,
      "epoch": 1.9629898403483308,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046150932575564173,
      "loss": 5.2791,
      "mean_token_accuracy": 0.18486236184835433,
      "num_tokens": 44909074.0,
      "step": 21640
    },
    {
      "entropy": 5.526303625106811,
      "epoch": 1.963443396226415,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046149122520375746,
      "loss": 5.1595,
      "mean_token_accuracy": 0.190170057117939,
      "num_tokens": 44919468.0,
      "step": 21645
    },
    {
      "entropy": 5.548718976974487,
      "epoch": 1.9638969521044993,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004614731207951871,
      "loss": 5.2759,
      "mean_token_accuracy": 0.1777361050248146,
      "num_tokens": 44930470.0,
      "step": 21650
    },
    {
      "entropy": 5.495882225036621,
      "epoch": 1.9643505079825836,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004614550125303051,
      "loss": 5.2741,
      "mean_token_accuracy": 0.1827964887022972,
      "num_tokens": 44940611.0,
      "step": 21655
    },
    {
      "entropy": 5.563466548919678,
      "epoch": 1.9648040638606676,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004614369004094859,
      "loss": 5.255,
      "mean_token_accuracy": 0.19044745415449144,
      "num_tokens": 44950985.0,
      "step": 21660
    },
    {
      "entropy": 5.5223125457763675,
      "epoch": 1.9652576197387517,
      "grad_norm": 2.359375,
      "learning_rate": 0.00046141878443310415,
      "loss": 5.2619,
      "mean_token_accuracy": 0.18291483968496322,
      "num_tokens": 44961858.0,
      "step": 21665
    },
    {
      "entropy": 5.569842052459717,
      "epoch": 1.965711175616836,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046140066460153444,
      "loss": 5.2202,
      "mean_token_accuracy": 0.18667929917573928,
      "num_tokens": 44972048.0,
      "step": 21670
    },
    {
      "entropy": 5.525922155380249,
      "epoch": 1.9661647314949202,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046138254091515136,
      "loss": 5.1401,
      "mean_token_accuracy": 0.18954193741083145,
      "num_tokens": 44981835.0,
      "step": 21675
    },
    {
      "entropy": 5.446562004089356,
      "epoch": 1.9666182873730045,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046136441337432996,
      "loss": 5.1055,
      "mean_token_accuracy": 0.19480181336402894,
      "num_tokens": 44992108.0,
      "step": 21680
    },
    {
      "entropy": 5.574547243118286,
      "epoch": 1.9670718432510885,
      "grad_norm": 2.125,
      "learning_rate": 0.00046134628197944496,
      "loss": 5.2964,
      "mean_token_accuracy": 0.17923678308725358,
      "num_tokens": 45002105.0,
      "step": 21685
    },
    {
      "entropy": 5.499050569534302,
      "epoch": 1.9675253991291726,
      "grad_norm": 1.9375,
      "learning_rate": 0.00046132814673087143,
      "loss": 5.2392,
      "mean_token_accuracy": 0.17889176458120346,
      "num_tokens": 45013934.0,
      "step": 21690
    },
    {
      "entropy": 5.5273411750793455,
      "epoch": 1.9679789550072568,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004613100076289842,
      "loss": 5.2109,
      "mean_token_accuracy": 0.18237763941287993,
      "num_tokens": 45023851.0,
      "step": 21695
    },
    {
      "entropy": 5.477634859085083,
      "epoch": 1.968432510885341,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004612918646741586,
      "loss": 5.1263,
      "mean_token_accuracy": 0.18345651775598526,
      "num_tokens": 45034140.0,
      "step": 21700
    },
    {
      "entropy": 5.4941943168640135,
      "epoch": 1.9688860667634254,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004612737178667698,
      "loss": 5.1891,
      "mean_token_accuracy": 0.17893602550029755,
      "num_tokens": 45044452.0,
      "step": 21705
    },
    {
      "entropy": 5.461782312393188,
      "epoch": 1.9693396226415094,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00046125556720719304,
      "loss": 5.2,
      "mean_token_accuracy": 0.17875810861587524,
      "num_tokens": 45053845.0,
      "step": 21710
    },
    {
      "entropy": 5.476531362533569,
      "epoch": 1.9697931785195935,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004612374126958037,
      "loss": 5.1391,
      "mean_token_accuracy": 0.18899961709976196,
      "num_tokens": 45063973.0,
      "step": 21715
    },
    {
      "entropy": 5.565526342391967,
      "epoch": 1.9702467343976777,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004612192543329771,
      "loss": 5.3449,
      "mean_token_accuracy": 0.17847554236650467,
      "num_tokens": 45074879.0,
      "step": 21720
    },
    {
      "entropy": 5.518007135391235,
      "epoch": 1.970700290275762,
      "grad_norm": 2.015625,
      "learning_rate": 0.000461201092119089,
      "loss": 5.1156,
      "mean_token_accuracy": 0.19124460369348525,
      "num_tokens": 45085161.0,
      "step": 21725
    },
    {
      "entropy": 5.532666873931885,
      "epoch": 1.9711538461538463,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004611829260545148,
      "loss": 5.2733,
      "mean_token_accuracy": 0.17564905285835267,
      "num_tokens": 45095862.0,
      "step": 21730
    },
    {
      "entropy": 5.530684757232666,
      "epoch": 1.9716074020319303,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00046116475613963026,
      "loss": 5.1927,
      "mean_token_accuracy": 0.18311181366443635,
      "num_tokens": 45106396.0,
      "step": 21735
    },
    {
      "entropy": 5.535721778869629,
      "epoch": 1.9720609579100146,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00046114658237481114,
      "loss": 5.1256,
      "mean_token_accuracy": 0.19127557873725892,
      "num_tokens": 45117178.0,
      "step": 21740
    },
    {
      "entropy": 5.381462430953979,
      "epoch": 1.9725145137880986,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004611284047604332,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19378728568553924,
      "num_tokens": 45126571.0,
      "step": 21745
    },
    {
      "entropy": 5.520886182785034,
      "epoch": 1.9729680696661829,
      "grad_norm": 1.953125,
      "learning_rate": 0.00046111022329687246,
      "loss": 5.2569,
      "mean_token_accuracy": 0.18602931946516038,
      "num_tokens": 45137247.0,
      "step": 21750
    },
    {
      "entropy": 5.6217869281768795,
      "epoch": 1.9734216255442671,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00046109203798450483,
      "loss": 5.2198,
      "mean_token_accuracy": 0.18413451910018921,
      "num_tokens": 45147531.0,
      "step": 21755
    },
    {
      "entropy": 5.4621110439300535,
      "epoch": 1.9738751814223512,
      "grad_norm": 1.5,
      "learning_rate": 0.0004610738488237065,
      "loss": 5.1803,
      "mean_token_accuracy": 0.18786286562681198,
      "num_tokens": 45158312.0,
      "step": 21760
    },
    {
      "entropy": 5.458483457565308,
      "epoch": 1.9743287373004355,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004610556558148535,
      "loss": 5.2495,
      "mean_token_accuracy": 0.18937471956014634,
      "num_tokens": 45169754.0,
      "step": 21765
    },
    {
      "entropy": 5.5523888111114506,
      "epoch": 1.9747822931785195,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004610374589583222,
      "loss": 5.2641,
      "mean_token_accuracy": 0.17835654318332672,
      "num_tokens": 45180428.0,
      "step": 21770
    },
    {
      "entropy": 5.537300729751587,
      "epoch": 1.9752358490566038,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004610192582544888,
      "loss": 5.1719,
      "mean_token_accuracy": 0.18875014334917067,
      "num_tokens": 45191558.0,
      "step": 21775
    },
    {
      "entropy": 5.462187337875366,
      "epoch": 1.975689404934688,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004610010537037297,
      "loss": 5.1472,
      "mean_token_accuracy": 0.19207779169082642,
      "num_tokens": 45201567.0,
      "step": 21780
    },
    {
      "entropy": 5.551715087890625,
      "epoch": 1.9761429608127723,
      "grad_norm": 2.78125,
      "learning_rate": 0.0004609828453064214,
      "loss": 5.2905,
      "mean_token_accuracy": 0.1800078883767128,
      "num_tokens": 45211359.0,
      "step": 21785
    },
    {
      "entropy": 5.61629695892334,
      "epoch": 1.9765965166908563,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046096463306294057,
      "loss": 5.2085,
      "mean_token_accuracy": 0.1805325448513031,
      "num_tokens": 45222106.0,
      "step": 21790
    },
    {
      "entropy": 5.468383646011352,
      "epoch": 1.9770500725689404,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046094641697366366,
      "loss": 5.2197,
      "mean_token_accuracy": 0.18467198312282562,
      "num_tokens": 45233169.0,
      "step": 21795
    },
    {
      "entropy": 5.55444712638855,
      "epoch": 1.9775036284470247,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004609281970389675,
      "loss": 5.2706,
      "mean_token_accuracy": 0.17949533611536025,
      "num_tokens": 45243627.0,
      "step": 21800
    },
    {
      "entropy": 5.543450880050659,
      "epoch": 1.977957184325109,
      "grad_norm": 1.6875,
      "learning_rate": 0.00046090997325922887,
      "loss": 5.1513,
      "mean_token_accuracy": 0.188946533203125,
      "num_tokens": 45253730.0,
      "step": 21805
    },
    {
      "entropy": 5.478025674819946,
      "epoch": 1.9784107402031932,
      "grad_norm": 1.625,
      "learning_rate": 0.00046089174563482456,
      "loss": 5.2599,
      "mean_token_accuracy": 0.18899038434028625,
      "num_tokens": 45263443.0,
      "step": 21810
    },
    {
      "entropy": 5.4623500347137455,
      "epoch": 1.9788642960812772,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004608735141661316,
      "loss": 5.1867,
      "mean_token_accuracy": 0.18740262389183043,
      "num_tokens": 45275298.0,
      "step": 21815
    },
    {
      "entropy": 5.5028759956359865,
      "epoch": 1.9793178519593613,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046085527885352694,
      "loss": 5.2379,
      "mean_token_accuracy": 0.1850697726011276,
      "num_tokens": 45285628.0,
      "step": 21820
    },
    {
      "entropy": 5.460802602767944,
      "epoch": 1.9797714078374455,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004608370396973878,
      "loss": 5.1326,
      "mean_token_accuracy": 0.19123319983482362,
      "num_tokens": 45295559.0,
      "step": 21825
    },
    {
      "entropy": 5.478798723220825,
      "epoch": 1.9802249637155298,
      "grad_norm": 1.859375,
      "learning_rate": 0.00046081879669809133,
      "loss": 5.1365,
      "mean_token_accuracy": 0.19272014796733855,
      "num_tokens": 45305395.0,
      "step": 21830
    },
    {
      "entropy": 5.485418081283569,
      "epoch": 1.980678519593614,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004608005498560148,
      "loss": 5.1991,
      "mean_token_accuracy": 0.1881477028131485,
      "num_tokens": 45316763.0,
      "step": 21835
    },
    {
      "entropy": 5.498616504669189,
      "epoch": 1.9811320754716981,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004607822991715356,
      "loss": 5.3155,
      "mean_token_accuracy": 0.1813168480992317,
      "num_tokens": 45326484.0,
      "step": 21840
    },
    {
      "entropy": 5.529428768157959,
      "epoch": 1.9815856313497822,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000460764044645031,
      "loss": 5.2391,
      "mean_token_accuracy": 0.18152078986167908,
      "num_tokens": 45335792.0,
      "step": 21845
    },
    {
      "entropy": 5.523342847824097,
      "epoch": 1.9820391872278664,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004607457862768787,
      "loss": 5.1704,
      "mean_token_accuracy": 0.1787888690829277,
      "num_tokens": 45347247.0,
      "step": 21850
    },
    {
      "entropy": 5.579998779296875,
      "epoch": 1.9824927431059507,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004607275240674562,
      "loss": 5.2923,
      "mean_token_accuracy": 0.17809874415397645,
      "num_tokens": 45357994.0,
      "step": 21855
    },
    {
      "entropy": 5.4561107635498045,
      "epoch": 1.982946298984035,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004607092580171411,
      "loss": 5.2164,
      "mean_token_accuracy": 0.18890070021152497,
      "num_tokens": 45368170.0,
      "step": 21860
    },
    {
      "entropy": 5.513271903991699,
      "epoch": 1.983399854862119,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004606909881263113,
      "loss": 5.2614,
      "mean_token_accuracy": 0.18403886258602142,
      "num_tokens": 45378696.0,
      "step": 21865
    },
    {
      "entropy": 5.527182722091675,
      "epoch": 1.983853410740203,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046067271439534445,
      "loss": 5.2809,
      "mean_token_accuracy": 0.18343165069818496,
      "num_tokens": 45388578.0,
      "step": 21870
    },
    {
      "entropy": 5.518750429153442,
      "epoch": 1.9843069666182873,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046065443682461864,
      "loss": 5.2466,
      "mean_token_accuracy": 0.18810704201459885,
      "num_tokens": 45398835.0,
      "step": 21875
    },
    {
      "entropy": 5.53296971321106,
      "epoch": 1.9847605224963716,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046063615541451177,
      "loss": 5.2281,
      "mean_token_accuracy": 0.1875179648399353,
      "num_tokens": 45409032.0,
      "step": 21880
    },
    {
      "entropy": 5.630390214920044,
      "epoch": 1.9852140783744558,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046061787016540184,
      "loss": 5.3755,
      "mean_token_accuracy": 0.1721666693687439,
      "num_tokens": 45420661.0,
      "step": 21885
    },
    {
      "entropy": 5.483451223373413,
      "epoch": 1.98566763425254,
      "grad_norm": 1.765625,
      "learning_rate": 0.00046059958107766706,
      "loss": 5.0746,
      "mean_token_accuracy": 0.19587773978710174,
      "num_tokens": 45429751.0,
      "step": 21890
    },
    {
      "entropy": 5.50592794418335,
      "epoch": 1.986121190130624,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004605812881516857,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19116167575120926,
      "num_tokens": 45439551.0,
      "step": 21895
    },
    {
      "entropy": 5.426746559143067,
      "epoch": 1.9865747460087082,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00046056299138783595,
      "loss": 5.1034,
      "mean_token_accuracy": 0.20268187075853347,
      "num_tokens": 45449543.0,
      "step": 21900
    },
    {
      "entropy": 5.531546688079834,
      "epoch": 1.9870283018867925,
      "grad_norm": 1.375,
      "learning_rate": 0.0004605446907864963,
      "loss": 5.2118,
      "mean_token_accuracy": 0.1883380874991417,
      "num_tokens": 45459069.0,
      "step": 21905
    },
    {
      "entropy": 5.521339845657349,
      "epoch": 1.9874818577648767,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004605263863480452,
      "loss": 5.292,
      "mean_token_accuracy": 0.17791853100061417,
      "num_tokens": 45468987.0,
      "step": 21910
    },
    {
      "entropy": 5.569965887069702,
      "epoch": 1.9879354136429608,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046050807807286105,
      "loss": 5.2377,
      "mean_token_accuracy": 0.18455829173326493,
      "num_tokens": 45479781.0,
      "step": 21915
    },
    {
      "entropy": 5.574013137817383,
      "epoch": 1.988388969521045,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00046048976596132265,
      "loss": 5.338,
      "mean_token_accuracy": 0.1717803433537483,
      "num_tokens": 45490318.0,
      "step": 21920
    },
    {
      "entropy": 5.504294061660767,
      "epoch": 1.988842525399129,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046047145001380856,
      "loss": 5.1867,
      "mean_token_accuracy": 0.18546608984470367,
      "num_tokens": 45501514.0,
      "step": 21925
    },
    {
      "entropy": 5.517447471618652,
      "epoch": 1.9892960812772134,
      "grad_norm": 2.421875,
      "learning_rate": 0.00046045313023069764,
      "loss": 5.1986,
      "mean_token_accuracy": 0.18642459958791732,
      "num_tokens": 45511727.0,
      "step": 21930
    },
    {
      "entropy": 5.549218416213989,
      "epoch": 1.9897496371552976,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004604348066123688,
      "loss": 5.2773,
      "mean_token_accuracy": 0.1820951759815216,
      "num_tokens": 45521807.0,
      "step": 21935
    },
    {
      "entropy": 5.566455554962158,
      "epoch": 1.9902031930333817,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004604164791592008,
      "loss": 5.3117,
      "mean_token_accuracy": 0.17808157801628113,
      "num_tokens": 45532033.0,
      "step": 21940
    },
    {
      "entropy": 5.501931619644165,
      "epoch": 1.990656748911466,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046039814787157283,
      "loss": 5.2339,
      "mean_token_accuracy": 0.18212053477764129,
      "num_tokens": 45542806.0,
      "step": 21945
    },
    {
      "entropy": 5.523515701293945,
      "epoch": 1.99111030478955,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004603798127498639,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19989240914583206,
      "num_tokens": 45551982.0,
      "step": 21950
    },
    {
      "entropy": 5.5064740657806395,
      "epoch": 1.9915638606676342,
      "grad_norm": 1.375,
      "learning_rate": 0.0004603614737944532,
      "loss": 5.216,
      "mean_token_accuracy": 0.18705193549394608,
      "num_tokens": 45561998.0,
      "step": 21955
    },
    {
      "entropy": 5.462306261062622,
      "epoch": 1.9920174165457185,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00046034313100571994,
      "loss": 5.2272,
      "mean_token_accuracy": 0.18855227530002594,
      "num_tokens": 45572246.0,
      "step": 21960
    },
    {
      "entropy": 5.564800596237182,
      "epoch": 1.9924709724238028,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004603247843840436,
      "loss": 5.2444,
      "mean_token_accuracy": 0.18371566087007524,
      "num_tokens": 45582251.0,
      "step": 21965
    },
    {
      "entropy": 5.523930072784424,
      "epoch": 1.9929245283018868,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004603064339298034,
      "loss": 5.1366,
      "mean_token_accuracy": 0.18662743419408798,
      "num_tokens": 45592749.0,
      "step": 21970
    },
    {
      "entropy": 5.44760046005249,
      "epoch": 1.9933780841799709,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004602880796433789,
      "loss": 5.1681,
      "mean_token_accuracy": 0.19081476032733918,
      "num_tokens": 45602412.0,
      "step": 21975
    },
    {
      "entropy": 5.497353839874267,
      "epoch": 1.9938316400580551,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046026972152514976,
      "loss": 5.2342,
      "mean_token_accuracy": 0.18509870022535324,
      "num_tokens": 45612736.0,
      "step": 21980
    },
    {
      "entropy": 5.540775537490845,
      "epoch": 1.9942851959361394,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004602513595754956,
      "loss": 5.229,
      "mean_token_accuracy": 0.1859084039926529,
      "num_tokens": 45622952.0,
      "step": 21985
    },
    {
      "entropy": 5.545159101486206,
      "epoch": 1.9947387518142237,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046023299379479606,
      "loss": 5.2297,
      "mean_token_accuracy": 0.18694642037153245,
      "num_tokens": 45633569.0,
      "step": 21990
    },
    {
      "entropy": 5.4640419483184814,
      "epoch": 1.9951923076923077,
      "grad_norm": 1.375,
      "learning_rate": 0.00046021462418343104,
      "loss": 5.1903,
      "mean_token_accuracy": 0.1895455837249756,
      "num_tokens": 45644347.0,
      "step": 21995
    },
    {
      "entropy": 5.528571128845215,
      "epoch": 1.9956458635703918,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004601962507417804,
      "loss": 5.2569,
      "mean_token_accuracy": 0.19117714762687682,
      "num_tokens": 45655268.0,
      "step": 22000
    },
    {
      "entropy": 5.508088254928589,
      "epoch": 1.996099419448476,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00046017787347022403,
      "loss": 5.1792,
      "mean_token_accuracy": 0.18839693814516068,
      "num_tokens": 45664900.0,
      "step": 22005
    },
    {
      "entropy": 5.567244195938111,
      "epoch": 1.9965529753265603,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046015949236914217,
      "loss": 5.2745,
      "mean_token_accuracy": 0.18514120131731032,
      "num_tokens": 45675252.0,
      "step": 22010
    },
    {
      "entropy": 5.563499927520752,
      "epoch": 1.9970065312046446,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004601411074389147,
      "loss": 5.1846,
      "mean_token_accuracy": 0.19242769926786424,
      "num_tokens": 45686027.0,
      "step": 22015
    },
    {
      "entropy": 5.474637889862061,
      "epoch": 1.9974600870827286,
      "grad_norm": 1.609375,
      "learning_rate": 0.000460122718679922,
      "loss": 5.1855,
      "mean_token_accuracy": 0.1859956130385399,
      "num_tokens": 45696436.0,
      "step": 22020
    },
    {
      "entropy": 5.578459882736206,
      "epoch": 1.9979136429608126,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004601043260925443,
      "loss": 5.296,
      "mean_token_accuracy": 0.18878642469644547,
      "num_tokens": 45706509.0,
      "step": 22025
    },
    {
      "entropy": 5.589174222946167,
      "epoch": 1.998367198838897,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046008592967716197,
      "loss": 5.3174,
      "mean_token_accuracy": 0.1794729545712471,
      "num_tokens": 45717933.0,
      "step": 22030
    },
    {
      "entropy": 5.515478515625,
      "epoch": 1.9988207547169812,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004600675294341554,
      "loss": 5.2486,
      "mean_token_accuracy": 0.1764887973666191,
      "num_tokens": 45727966.0,
      "step": 22035
    },
    {
      "entropy": 5.5202751636505125,
      "epoch": 1.9992743105950654,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046004912536390526,
      "loss": 5.204,
      "mean_token_accuracy": 0.18472840934991835,
      "num_tokens": 45737649.0,
      "step": 22040
    },
    {
      "entropy": 5.508662223815918,
      "epoch": 1.9997278664731495,
      "grad_norm": 1.484375,
      "learning_rate": 0.000460030717466792,
      "loss": 5.1753,
      "mean_token_accuracy": 0.18483225554227828,
      "num_tokens": 45747561.0,
      "step": 22045
    },
    {
      "entropy": 5.44984917640686,
      "epoch": 2.0001814223512335,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004600123057431963,
      "loss": 5.1466,
      "mean_token_accuracy": 0.18811834305524827,
      "num_tokens": 45758183.0,
      "step": 22050
    },
    {
      "entropy": 5.475098657608032,
      "epoch": 2.000634978229318,
      "grad_norm": 1.359375,
      "learning_rate": 0.000459993890193499,
      "loss": 5.0933,
      "mean_token_accuracy": 0.2021588057279587,
      "num_tokens": 45768006.0,
      "step": 22055
    },
    {
      "entropy": 5.566107940673828,
      "epoch": 2.001088534107402,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004599754708180809,
      "loss": 5.2448,
      "mean_token_accuracy": 0.18215628415346147,
      "num_tokens": 45777981.0,
      "step": 22060
    },
    {
      "entropy": 5.512448024749756,
      "epoch": 2.0015420899854863,
      "grad_norm": 1.515625,
      "learning_rate": 0.000459957047617323,
      "loss": 5.1836,
      "mean_token_accuracy": 0.18159467577934266,
      "num_tokens": 45789083.0,
      "step": 22065
    },
    {
      "entropy": 5.514123487472534,
      "epoch": 2.0019956458635706,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00045993862059160617,
      "loss": 5.1135,
      "mean_token_accuracy": 0.18725121021270752,
      "num_tokens": 45799750.0,
      "step": 22070
    },
    {
      "entropy": 5.479077577590942,
      "epoch": 2.0024492017416544,
      "grad_norm": 1.625,
      "learning_rate": 0.0004599201897413116,
      "loss": 5.1159,
      "mean_token_accuracy": 0.18990031629800797,
      "num_tokens": 45810321.0,
      "step": 22075
    },
    {
      "entropy": 5.444447374343872,
      "epoch": 2.0029027576197387,
      "grad_norm": 1.5,
      "learning_rate": 0.00045990175506682024,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19680560380220413,
      "num_tokens": 45820581.0,
      "step": 22080
    },
    {
      "entropy": 5.509267568588257,
      "epoch": 2.003356313497823,
      "grad_norm": 1.859375,
      "learning_rate": 0.00045988331656851355,
      "loss": 5.1732,
      "mean_token_accuracy": 0.18819838166236877,
      "num_tokens": 45831504.0,
      "step": 22085
    },
    {
      "entropy": 5.557058906555175,
      "epoch": 2.003809869375907,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00045986487424677284,
      "loss": 5.2148,
      "mean_token_accuracy": 0.18467621952295304,
      "num_tokens": 45841100.0,
      "step": 22090
    },
    {
      "entropy": 5.51617693901062,
      "epoch": 2.0042634252539915,
      "grad_norm": 2.796875,
      "learning_rate": 0.0004598464281019793,
      "loss": 5.1852,
      "mean_token_accuracy": 0.19077038168907165,
      "num_tokens": 45851647.0,
      "step": 22095
    },
    {
      "entropy": 5.4919788360595705,
      "epoch": 2.0047169811320753,
      "grad_norm": 2.15625,
      "learning_rate": 0.00045982797813451463,
      "loss": 5.1146,
      "mean_token_accuracy": 0.1922835722565651,
      "num_tokens": 45862995.0,
      "step": 22100
    },
    {
      "entropy": 5.492366313934326,
      "epoch": 2.0051705370101596,
      "grad_norm": 1.53125,
      "learning_rate": 0.00045980952434476014,
      "loss": 5.0987,
      "mean_token_accuracy": 0.1907983273267746,
      "num_tokens": 45872360.0,
      "step": 22105
    },
    {
      "entropy": 5.493115425109863,
      "epoch": 2.005624092888244,
      "grad_norm": 1.625,
      "learning_rate": 0.00045979106673309765,
      "loss": 5.1917,
      "mean_token_accuracy": 0.18106362968683243,
      "num_tokens": 45883440.0,
      "step": 22110
    },
    {
      "entropy": 5.457197523117065,
      "epoch": 2.006077648766328,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045977260529990883,
      "loss": 5.1423,
      "mean_token_accuracy": 0.18632368594408036,
      "num_tokens": 45894323.0,
      "step": 22115
    },
    {
      "entropy": 5.475085735321045,
      "epoch": 2.0065312046444124,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004597541400455755,
      "loss": 5.1896,
      "mean_token_accuracy": 0.18448591679334642,
      "num_tokens": 45904544.0,
      "step": 22120
    },
    {
      "entropy": 5.551220893859863,
      "epoch": 2.006984760522496,
      "grad_norm": 1.375,
      "learning_rate": 0.00045973567097047933,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20483342558145523,
      "num_tokens": 45913225.0,
      "step": 22125
    },
    {
      "entropy": 5.483597993850708,
      "epoch": 2.0074383164005805,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045971719807500253,
      "loss": 5.1936,
      "mean_token_accuracy": 0.19039818346500398,
      "num_tokens": 45923306.0,
      "step": 22130
    },
    {
      "entropy": 5.470352458953857,
      "epoch": 2.0078918722786647,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000459698721359527,
      "loss": 5.1565,
      "mean_token_accuracy": 0.18772485703229905,
      "num_tokens": 45933243.0,
      "step": 22135
    },
    {
      "entropy": 5.444543123245239,
      "epoch": 2.008345428156749,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004596802408244348,
      "loss": 5.2078,
      "mean_token_accuracy": 0.18851405531167983,
      "num_tokens": 45944498.0,
      "step": 22140
    },
    {
      "entropy": 5.436703205108643,
      "epoch": 2.0087989840348333,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004596617564701081,
      "loss": 5.0498,
      "mean_token_accuracy": 0.19302451908588408,
      "num_tokens": 45954550.0,
      "step": 22145
    },
    {
      "entropy": 5.554660320281982,
      "epoch": 2.009252539912917,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045964326829692925,
      "loss": 5.1883,
      "mean_token_accuracy": 0.18482790440320968,
      "num_tokens": 45965425.0,
      "step": 22150
    },
    {
      "entropy": 5.4861479759216305,
      "epoch": 2.0097060957910013,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045962477630528064,
      "loss": 5.0826,
      "mean_token_accuracy": 0.19791797995567323,
      "num_tokens": 45975637.0,
      "step": 22155
    },
    {
      "entropy": 5.425532722473145,
      "epoch": 2.0101596516690856,
      "grad_norm": 1.5,
      "learning_rate": 0.00045960628049554453,
      "loss": 5.1961,
      "mean_token_accuracy": 0.17972565591335296,
      "num_tokens": 45985268.0,
      "step": 22160
    },
    {
      "entropy": 5.452089309692383,
      "epoch": 2.01061320754717,
      "grad_norm": 3.765625,
      "learning_rate": 0.00045958778086810354,
      "loss": 5.086,
      "mean_token_accuracy": 0.18859538286924363,
      "num_tokens": 45995335.0,
      "step": 22165
    },
    {
      "entropy": 5.51650562286377,
      "epoch": 2.011066763425254,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004595692774233402,
      "loss": 5.1924,
      "mean_token_accuracy": 0.18055151998996735,
      "num_tokens": 46006037.0,
      "step": 22170
    },
    {
      "entropy": 5.520053911209106,
      "epoch": 2.011520319303338,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045955077016163713,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19210627377033235,
      "num_tokens": 46017016.0,
      "step": 22175
    },
    {
      "entropy": 5.563284778594971,
      "epoch": 2.0119738751814222,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045953225908337717,
      "loss": 5.2267,
      "mean_token_accuracy": 0.18513831943273545,
      "num_tokens": 46026797.0,
      "step": 22180
    },
    {
      "entropy": 5.41064076423645,
      "epoch": 2.0124274310595065,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045951374418894304,
      "loss": 5.0801,
      "mean_token_accuracy": 0.1916659578680992,
      "num_tokens": 46037799.0,
      "step": 22185
    },
    {
      "entropy": 5.470755195617675,
      "epoch": 2.0128809869375908,
      "grad_norm": 1.703125,
      "learning_rate": 0.00045949522547871765,
      "loss": 5.0807,
      "mean_token_accuracy": 0.18824495673179625,
      "num_tokens": 46048617.0,
      "step": 22190
    },
    {
      "entropy": 5.5553285598754885,
      "epoch": 2.013334542815675,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000459476702953084,
      "loss": 5.2056,
      "mean_token_accuracy": 0.18449336290359497,
      "num_tokens": 46060122.0,
      "step": 22195
    },
    {
      "entropy": 5.539176177978516,
      "epoch": 2.013788098693759,
      "grad_norm": 1.59375,
      "learning_rate": 0.000459458176612425,
      "loss": 5.1619,
      "mean_token_accuracy": 0.1897239089012146,
      "num_tokens": 46071488.0,
      "step": 22200
    },
    {
      "entropy": 5.531408500671387,
      "epoch": 2.014241654571843,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045943964645712407,
      "loss": 5.1969,
      "mean_token_accuracy": 0.18442261219024658,
      "num_tokens": 46081954.0,
      "step": 22205
    },
    {
      "entropy": 5.4700360774993895,
      "epoch": 2.0146952104499274,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004594211124875642,
      "loss": 5.1248,
      "mean_token_accuracy": 0.1934116840362549,
      "num_tokens": 46092488.0,
      "step": 22210
    },
    {
      "entropy": 5.446074438095093,
      "epoch": 2.0151487663280117,
      "grad_norm": 1.5,
      "learning_rate": 0.0004594025747041287,
      "loss": 5.1176,
      "mean_token_accuracy": 0.19655249267816544,
      "num_tokens": 46102451.0,
      "step": 22215
    },
    {
      "entropy": 5.440164852142334,
      "epoch": 2.015602322206096,
      "grad_norm": 1.515625,
      "learning_rate": 0.000459384033107201,
      "loss": 5.1656,
      "mean_token_accuracy": 0.19374954998493193,
      "num_tokens": 46113862.0,
      "step": 22220
    },
    {
      "entropy": 5.592874956130982,
      "epoch": 2.01605587808418,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004593654876971645,
      "loss": 5.2277,
      "mean_token_accuracy": 0.17921980023384093,
      "num_tokens": 46124517.0,
      "step": 22225
    },
    {
      "entropy": 5.449488925933838,
      "epoch": 2.016509433962264,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004593469384744027,
      "loss": 5.1497,
      "mean_token_accuracy": 0.18633628338575364,
      "num_tokens": 46134221.0,
      "step": 22230
    },
    {
      "entropy": 5.509955930709839,
      "epoch": 2.0169629898403483,
      "grad_norm": 1.6875,
      "learning_rate": 0.00045932838543929924,
      "loss": 5.2158,
      "mean_token_accuracy": 0.18649559170007707,
      "num_tokens": 46144512.0,
      "step": 22235
    },
    {
      "entropy": 5.555418300628662,
      "epoch": 2.0174165457184325,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004593098285922378,
      "loss": 5.1264,
      "mean_token_accuracy": 0.1886797159910202,
      "num_tokens": 46154373.0,
      "step": 22240
    },
    {
      "entropy": 5.566738271713257,
      "epoch": 2.017870101596517,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004592912679336021,
      "loss": 5.2701,
      "mean_token_accuracy": 0.18273737132549286,
      "num_tokens": 46164834.0,
      "step": 22245
    },
    {
      "entropy": 5.4686572551727295,
      "epoch": 2.018323657474601,
      "grad_norm": 1.875,
      "learning_rate": 0.0004592727034637761,
      "loss": 5.1054,
      "mean_token_accuracy": 0.19309171587228774,
      "num_tokens": 46174815.0,
      "step": 22250
    },
    {
      "entropy": 5.436180257797242,
      "epoch": 2.018777213352685,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045925413518314363,
      "loss": 5.1367,
      "mean_token_accuracy": 0.1855654701590538,
      "num_tokens": 46185155.0,
      "step": 22255
    },
    {
      "entropy": 5.509893608093262,
      "epoch": 2.019230769230769,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004592355630920886,
      "loss": 5.1877,
      "mean_token_accuracy": 0.19145286828279495,
      "num_tokens": 46194537.0,
      "step": 22260
    },
    {
      "entropy": 5.566504240036011,
      "epoch": 2.0196843251088534,
      "grad_norm": 1.78125,
      "learning_rate": 0.00045921698719099527,
      "loss": 5.2195,
      "mean_token_accuracy": 0.1876019075512886,
      "num_tokens": 46205494.0,
      "step": 22265
    },
    {
      "entropy": 5.4441827774047855,
      "epoch": 2.0201378809869377,
      "grad_norm": 1.5625,
      "learning_rate": 0.00045919840748024764,
      "loss": 5.1222,
      "mean_token_accuracy": 0.1913737252354622,
      "num_tokens": 46216399.0,
      "step": 22270
    },
    {
      "entropy": 5.410929727554321,
      "epoch": 2.020591436865022,
      "grad_norm": 1.828125,
      "learning_rate": 0.00045917982396023,
      "loss": 5.1064,
      "mean_token_accuracy": 0.19204334318637847,
      "num_tokens": 46226515.0,
      "step": 22275
    },
    {
      "entropy": 5.548749542236328,
      "epoch": 2.021044992743106,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045916123663132676,
      "loss": 5.2051,
      "mean_token_accuracy": 0.17982069849967958,
      "num_tokens": 46236688.0,
      "step": 22280
    },
    {
      "entropy": 5.475341176986694,
      "epoch": 2.02149854862119,
      "grad_norm": 1.765625,
      "learning_rate": 0.00045914264549392215,
      "loss": 5.1332,
      "mean_token_accuracy": 0.19538284987211227,
      "num_tokens": 46247166.0,
      "step": 22285
    },
    {
      "entropy": 5.497462415695191,
      "epoch": 2.0219521044992743,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004591240505484007,
      "loss": 5.1422,
      "mean_token_accuracy": 0.18391215801239014,
      "num_tokens": 46258902.0,
      "step": 22290
    },
    {
      "entropy": 5.5082234859466555,
      "epoch": 2.0224056603773586,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045910545179514694,
      "loss": 5.2693,
      "mean_token_accuracy": 0.1813361331820488,
      "num_tokens": 46269819.0,
      "step": 22295
    },
    {
      "entropy": 5.59643177986145,
      "epoch": 2.022859216255443,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004590868492345455,
      "loss": 5.2096,
      "mean_token_accuracy": 0.18898700624704362,
      "num_tokens": 46280715.0,
      "step": 22300
    },
    {
      "entropy": 5.551849746704102,
      "epoch": 2.0233127721335267,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004590682428669813,
      "loss": 5.2205,
      "mean_token_accuracy": 0.1784398928284645,
      "num_tokens": 46291079.0,
      "step": 22305
    },
    {
      "entropy": 5.5302534103393555,
      "epoch": 2.023766328011611,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004590496326928387,
      "loss": 5.1653,
      "mean_token_accuracy": 0.1897205203771591,
      "num_tokens": 46301362.0,
      "step": 22310
    },
    {
      "entropy": 5.487811660766601,
      "epoch": 2.024219883889695,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00045903101871250295,
      "loss": 5.1346,
      "mean_token_accuracy": 0.19536945819854737,
      "num_tokens": 46311820.0,
      "step": 22315
    },
    {
      "entropy": 5.517128276824951,
      "epoch": 2.0246734397677795,
      "grad_norm": 1.375,
      "learning_rate": 0.0004590124009263587,
      "loss": 5.2409,
      "mean_token_accuracy": 0.18514203429222106,
      "num_tokens": 46321992.0,
      "step": 22320
    },
    {
      "entropy": 5.510509347915649,
      "epoch": 2.0251269956458637,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004589937793347912,
      "loss": 5.173,
      "mean_token_accuracy": 0.18724144250154495,
      "num_tokens": 46331537.0,
      "step": 22325
    },
    {
      "entropy": 5.484735822677612,
      "epoch": 2.0255805515239476,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045897515393818547,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1869571253657341,
      "num_tokens": 46342214.0,
      "step": 22330
    },
    {
      "entropy": 5.45034384727478,
      "epoch": 2.026034107402032,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004589565247369266,
      "loss": 5.1519,
      "mean_token_accuracy": 0.18482770174741744,
      "num_tokens": 46352471.0,
      "step": 22335
    },
    {
      "entropy": 5.471403980255127,
      "epoch": 2.026487663280116,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004589378917314,
      "loss": 5.1939,
      "mean_token_accuracy": 0.1885033592581749,
      "num_tokens": 46363221.0,
      "step": 22340
    },
    {
      "entropy": 5.42208685874939,
      "epoch": 2.0269412191582004,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045891925492199087,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20296556651592254,
      "num_tokens": 46373454.0,
      "step": 22345
    },
    {
      "entropy": 5.581568098068237,
      "epoch": 2.0273947750362846,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004589006143090847,
      "loss": 5.3179,
      "mean_token_accuracy": 0.18480725437402726,
      "num_tokens": 46384403.0,
      "step": 22350
    },
    {
      "entropy": 5.473208665847778,
      "epoch": 2.0278483309143684,
      "grad_norm": 1.5,
      "learning_rate": 0.00045888196989306697,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20522022545337676,
      "num_tokens": 46393362.0,
      "step": 22355
    },
    {
      "entropy": 5.540653371810913,
      "epoch": 2.0283018867924527,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004588633216743232,
      "loss": 5.2731,
      "mean_token_accuracy": 0.18473271727561952,
      "num_tokens": 46405100.0,
      "step": 22360
    },
    {
      "entropy": 5.433826208114624,
      "epoch": 2.028755442670537,
      "grad_norm": 1.5625,
      "learning_rate": 0.00045884466965323914,
      "loss": 5.0822,
      "mean_token_accuracy": 0.1881657972931862,
      "num_tokens": 46415058.0,
      "step": 22365
    },
    {
      "entropy": 5.493132209777832,
      "epoch": 2.0292089985486212,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045882601383020045,
      "loss": 5.1421,
      "mean_token_accuracy": 0.19210312962532045,
      "num_tokens": 46425863.0,
      "step": 22370
    },
    {
      "entropy": 5.52985520362854,
      "epoch": 2.0296625544267055,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004588073542055929,
      "loss": 5.1539,
      "mean_token_accuracy": 0.18793633282184602,
      "num_tokens": 46436172.0,
      "step": 22375
    },
    {
      "entropy": 5.5440568923950195,
      "epoch": 2.0301161103047893,
      "grad_norm": 1.375,
      "learning_rate": 0.00045878869077980246,
      "loss": 5.18,
      "mean_token_accuracy": 0.18524127155542375,
      "num_tokens": 46446717.0,
      "step": 22380
    },
    {
      "entropy": 5.495404672622681,
      "epoch": 2.0305696661828736,
      "grad_norm": 1.5625,
      "learning_rate": 0.00045877002355321504,
      "loss": 5.1959,
      "mean_token_accuracy": 0.18270741403102875,
      "num_tokens": 46456944.0,
      "step": 22385
    },
    {
      "entropy": 5.533405685424805,
      "epoch": 2.031023222060958,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004587513525262168,
      "loss": 5.2341,
      "mean_token_accuracy": 0.18903636187314987,
      "num_tokens": 46467231.0,
      "step": 22390
    },
    {
      "entropy": 5.5295196056365965,
      "epoch": 2.031476777939042,
      "grad_norm": 1.5,
      "learning_rate": 0.00045873267769919364,
      "loss": 5.1004,
      "mean_token_accuracy": 0.18582315295934676,
      "num_tokens": 46477107.0,
      "step": 22395
    },
    {
      "entropy": 5.474626636505127,
      "epoch": 2.0319303338171264,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004587139990725319,
      "loss": 5.1566,
      "mean_token_accuracy": 0.19170880168676377,
      "num_tokens": 46487351.0,
      "step": 22400
    },
    {
      "entropy": 5.496630144119263,
      "epoch": 2.0323838896952107,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004586953166466179,
      "loss": 5.1764,
      "mean_token_accuracy": 0.1846473127603531,
      "num_tokens": 46497685.0,
      "step": 22405
    },
    {
      "entropy": 5.490532302856446,
      "epoch": 2.0328374455732945,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004586766304218379,
      "loss": 5.1896,
      "mean_token_accuracy": 0.17795229703187943,
      "num_tokens": 46507866.0,
      "step": 22410
    },
    {
      "entropy": 5.44972038269043,
      "epoch": 2.0332910014513788,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045865794039857846,
      "loss": 5.0615,
      "mean_token_accuracy": 0.19314694702625274,
      "num_tokens": 46518797.0,
      "step": 22415
    },
    {
      "entropy": 5.490714502334595,
      "epoch": 2.033744557329463,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004586392465772259,
      "loss": 5.1405,
      "mean_token_accuracy": 0.1916805237531662,
      "num_tokens": 46528759.0,
      "step": 22420
    },
    {
      "entropy": 5.504214143753051,
      "epoch": 2.0341981132075473,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045862054895816703,
      "loss": 5.2792,
      "mean_token_accuracy": 0.18013475090265274,
      "num_tokens": 46539014.0,
      "step": 22425
    },
    {
      "entropy": 5.494124460220337,
      "epoch": 2.0346516690856316,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004586018475417884,
      "loss": 5.1416,
      "mean_token_accuracy": 0.1913141831755638,
      "num_tokens": 46548856.0,
      "step": 22430
    },
    {
      "entropy": 5.485150098800659,
      "epoch": 2.0351052249637154,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004585831423284766,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1942108079791069,
      "num_tokens": 46559164.0,
      "step": 22435
    },
    {
      "entropy": 5.513820028305053,
      "epoch": 2.0355587808417996,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004585644333186188,
      "loss": 5.215,
      "mean_token_accuracy": 0.18316422700881957,
      "num_tokens": 46569793.0,
      "step": 22440
    },
    {
      "entropy": 5.483968210220337,
      "epoch": 2.036012336719884,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045854572051260166,
      "loss": 5.1699,
      "mean_token_accuracy": 0.18796504586935042,
      "num_tokens": 46580257.0,
      "step": 22445
    },
    {
      "entropy": 5.4946702003479,
      "epoch": 2.036465892597968,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045852700391081234,
      "loss": 5.167,
      "mean_token_accuracy": 0.18257856518030166,
      "num_tokens": 46590297.0,
      "step": 22450
    },
    {
      "entropy": 5.51847095489502,
      "epoch": 2.0369194484760524,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004585082835136378,
      "loss": 5.1215,
      "mean_token_accuracy": 0.1949250340461731,
      "num_tokens": 46600647.0,
      "step": 22455
    },
    {
      "entropy": 5.478297996520996,
      "epoch": 2.0373730043541363,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004584895593214652,
      "loss": 5.1204,
      "mean_token_accuracy": 0.18793613463640213,
      "num_tokens": 46610478.0,
      "step": 22460
    },
    {
      "entropy": 5.472389364242554,
      "epoch": 2.0378265602322205,
      "grad_norm": 1.71875,
      "learning_rate": 0.00045847083133468164,
      "loss": 5.1973,
      "mean_token_accuracy": 0.18593110591173173,
      "num_tokens": 46620159.0,
      "step": 22465
    },
    {
      "entropy": 5.504296398162841,
      "epoch": 2.038280116110305,
      "grad_norm": 2.265625,
      "learning_rate": 0.00045845209955367464,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19230891913175582,
      "num_tokens": 46630555.0,
      "step": 22470
    },
    {
      "entropy": 5.502232360839844,
      "epoch": 2.038733671988389,
      "grad_norm": 1.890625,
      "learning_rate": 0.00045843336397883146,
      "loss": 5.2096,
      "mean_token_accuracy": 0.18655613362789153,
      "num_tokens": 46640188.0,
      "step": 22475
    },
    {
      "entropy": 5.468849468231201,
      "epoch": 2.0391872278664733,
      "grad_norm": 1.5,
      "learning_rate": 0.0004584146246105396,
      "loss": 5.1325,
      "mean_token_accuracy": 0.1890099823474884,
      "num_tokens": 46650284.0,
      "step": 22480
    },
    {
      "entropy": 5.529465532302856,
      "epoch": 2.039640783744557,
      "grad_norm": 1.765625,
      "learning_rate": 0.00045839588144918643,
      "loss": 5.1543,
      "mean_token_accuracy": 0.19062524884939194,
      "num_tokens": 46661466.0,
      "step": 22485
    },
    {
      "entropy": 5.522082948684693,
      "epoch": 2.0400943396226414,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004583771344951598,
      "loss": 5.15,
      "mean_token_accuracy": 0.18721580505371094,
      "num_tokens": 46671683.0,
      "step": 22490
    },
    {
      "entropy": 5.496673488616944,
      "epoch": 2.0405478955007257,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045835838374884733,
      "loss": 5.1875,
      "mean_token_accuracy": 0.18664518892765045,
      "num_tokens": 46681787.0,
      "step": 22495
    },
    {
      "entropy": 5.491007280349732,
      "epoch": 2.04100145137881,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00045833962921063665,
      "loss": 5.1368,
      "mean_token_accuracy": 0.1908806711435318,
      "num_tokens": 46692752.0,
      "step": 22500
    },
    {
      "entropy": 5.488840818405151,
      "epoch": 2.041455007256894,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045832087088091574,
      "loss": 5.1807,
      "mean_token_accuracy": 0.18543701916933059,
      "num_tokens": 46702980.0,
      "step": 22505
    },
    {
      "entropy": 5.476349925994873,
      "epoch": 2.041908563134978,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045830210876007257,
      "loss": 5.0901,
      "mean_token_accuracy": 0.18785773664712907,
      "num_tokens": 46712281.0,
      "step": 22510
    },
    {
      "entropy": 5.487309741973877,
      "epoch": 2.0423621190130623,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000458283342848495,
      "loss": 5.1363,
      "mean_token_accuracy": 0.19154182970523834,
      "num_tokens": 46722097.0,
      "step": 22515
    },
    {
      "entropy": 5.422311639785766,
      "epoch": 2.0428156748911466,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004582645731465713,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19707242399454117,
      "num_tokens": 46733056.0,
      "step": 22520
    },
    {
      "entropy": 5.444029903411865,
      "epoch": 2.043269230769231,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045824579965468944,
      "loss": 5.1144,
      "mean_token_accuracy": 0.19041707515716552,
      "num_tokens": 46742608.0,
      "step": 22525
    },
    {
      "entropy": 5.441998767852783,
      "epoch": 2.043722786647315,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00045822702237323777,
      "loss": 5.0926,
      "mean_token_accuracy": 0.1987132266163826,
      "num_tokens": 46751762.0,
      "step": 22530
    },
    {
      "entropy": 5.4432408809661865,
      "epoch": 2.044176342525399,
      "grad_norm": 1.5,
      "learning_rate": 0.00045820824130260457,
      "loss": 5.0947,
      "mean_token_accuracy": 0.19529207944869995,
      "num_tokens": 46762252.0,
      "step": 22535
    },
    {
      "entropy": 5.479282426834106,
      "epoch": 2.044629898403483,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004581894564431783,
      "loss": 5.1543,
      "mean_token_accuracy": 0.18822377324104309,
      "num_tokens": 46772172.0,
      "step": 22540
    },
    {
      "entropy": 5.47203950881958,
      "epoch": 2.0450834542815675,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004581706677953473,
      "loss": 5.179,
      "mean_token_accuracy": 0.18773523718118668,
      "num_tokens": 46782894.0,
      "step": 22545
    },
    {
      "entropy": 5.504971218109131,
      "epoch": 2.0455370101596517,
      "grad_norm": 1.59375,
      "learning_rate": 0.00045815187535950034,
      "loss": 5.1517,
      "mean_token_accuracy": 0.18961043655872345,
      "num_tokens": 46793097.0,
      "step": 22550
    },
    {
      "entropy": 5.537362146377563,
      "epoch": 2.045990566037736,
      "grad_norm": 1.375,
      "learning_rate": 0.00045813307913602586,
      "loss": 5.2498,
      "mean_token_accuracy": 0.17678584903478622,
      "num_tokens": 46804196.0,
      "step": 22555
    },
    {
      "entropy": 5.516360282897949,
      "epoch": 2.04644412191582,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004581142791253126,
      "loss": 5.1398,
      "mean_token_accuracy": 0.18865461349487306,
      "num_tokens": 46813985.0,
      "step": 22560
    },
    {
      "entropy": 5.439433765411377,
      "epoch": 2.046897677793904,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045809547532774943,
      "loss": 5.0609,
      "mean_token_accuracy": 0.1871512696146965,
      "num_tokens": 46824125.0,
      "step": 22565
    },
    {
      "entropy": 5.461756896972656,
      "epoch": 2.0473512336719883,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00045807666774372514,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19090001434087753,
      "num_tokens": 46834982.0,
      "step": 22570
    },
    {
      "entropy": 5.469232273101807,
      "epoch": 2.0478047895500726,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045805785637362874,
      "loss": 5.0736,
      "mean_token_accuracy": 0.19850182831287383,
      "num_tokens": 46844669.0,
      "step": 22575
    },
    {
      "entropy": 5.47543306350708,
      "epoch": 2.048258345428157,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004580390412178492,
      "loss": 5.1753,
      "mean_token_accuracy": 0.18986421227455139,
      "num_tokens": 46855108.0,
      "step": 22580
    },
    {
      "entropy": 5.528331279754639,
      "epoch": 2.048711901306241,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045802022227677566,
      "loss": 5.0957,
      "mean_token_accuracy": 0.19028275609016418,
      "num_tokens": 46864878.0,
      "step": 22585
    },
    {
      "entropy": 5.471252584457398,
      "epoch": 2.049165457184325,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045800139955079727,
      "loss": 5.1808,
      "mean_token_accuracy": 0.18617268204689025,
      "num_tokens": 46874916.0,
      "step": 22590
    },
    {
      "entropy": 5.503813552856445,
      "epoch": 2.0496190130624092,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004579825730403033,
      "loss": 5.2178,
      "mean_token_accuracy": 0.18394626528024674,
      "num_tokens": 46885167.0,
      "step": 22595
    },
    {
      "entropy": 5.531022787094116,
      "epoch": 2.0500725689404935,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004579637427456831,
      "loss": 5.1741,
      "mean_token_accuracy": 0.18382207453250884,
      "num_tokens": 46895526.0,
      "step": 22600
    },
    {
      "entropy": 5.450811719894409,
      "epoch": 2.0505261248185778,
      "grad_norm": 2.046875,
      "learning_rate": 0.00045794490866732607,
      "loss": 5.1483,
      "mean_token_accuracy": 0.18905539214611053,
      "num_tokens": 46905426.0,
      "step": 22605
    },
    {
      "entropy": 5.4773355484008786,
      "epoch": 2.050979680696662,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004579260708056217,
      "loss": 5.1756,
      "mean_token_accuracy": 0.18745941966772078,
      "num_tokens": 46915728.0,
      "step": 22610
    },
    {
      "entropy": 5.457514810562134,
      "epoch": 2.051433236574746,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004579072291609596,
      "loss": 5.1081,
      "mean_token_accuracy": 0.19365862458944322,
      "num_tokens": 46925972.0,
      "step": 22615
    },
    {
      "entropy": 5.468387317657471,
      "epoch": 2.05188679245283,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045788838373372935,
      "loss": 5.1437,
      "mean_token_accuracy": 0.1918352246284485,
      "num_tokens": 46937108.0,
      "step": 22620
    },
    {
      "entropy": 5.433141374588013,
      "epoch": 2.0523403483309144,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00045786953452432076,
      "loss": 5.1175,
      "mean_token_accuracy": 0.18781979829072953,
      "num_tokens": 46947684.0,
      "step": 22625
    },
    {
      "entropy": 5.536245727539063,
      "epoch": 2.0527939042089987,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004578506815331236,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1833452105522156,
      "num_tokens": 46958089.0,
      "step": 22630
    },
    {
      "entropy": 5.442163133621216,
      "epoch": 2.053247460087083,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004578318247605277,
      "loss": 5.1141,
      "mean_token_accuracy": 0.1955268681049347,
      "num_tokens": 46968618.0,
      "step": 22635
    },
    {
      "entropy": 5.47055287361145,
      "epoch": 2.0537010159651667,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004578129642069231,
      "loss": 5.132,
      "mean_token_accuracy": 0.19417261630296706,
      "num_tokens": 46978791.0,
      "step": 22640
    },
    {
      "entropy": 5.432848834991455,
      "epoch": 2.054154571843251,
      "grad_norm": 1.625,
      "learning_rate": 0.0004577940998726998,
      "loss": 5.0783,
      "mean_token_accuracy": 0.20038242489099503,
      "num_tokens": 46988706.0,
      "step": 22645
    },
    {
      "entropy": 5.533437299728393,
      "epoch": 2.0546081277213353,
      "grad_norm": 1.59375,
      "learning_rate": 0.00045777523175824796,
      "loss": 5.2747,
      "mean_token_accuracy": 0.17530401647090912,
      "num_tokens": 47000140.0,
      "step": 22650
    },
    {
      "entropy": 5.523886871337891,
      "epoch": 2.0550616835994195,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004577563598639577,
      "loss": 5.1215,
      "mean_token_accuracy": 0.19129243195056916,
      "num_tokens": 47009773.0,
      "step": 22655
    },
    {
      "entropy": 5.5639630317687985,
      "epoch": 2.055515239477504,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045773748419021944,
      "loss": 5.1912,
      "mean_token_accuracy": 0.1872617170214653,
      "num_tokens": 47020326.0,
      "step": 22660
    },
    {
      "entropy": 5.414564561843872,
      "epoch": 2.0559687953555876,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045771860473742347,
      "loss": 5.1089,
      "mean_token_accuracy": 0.18562285602092743,
      "num_tokens": 47030410.0,
      "step": 22665
    },
    {
      "entropy": 5.488576030731201,
      "epoch": 2.056422351233672,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045769972150596006,
      "loss": 5.1659,
      "mean_token_accuracy": 0.1881070226430893,
      "num_tokens": 47040546.0,
      "step": 22670
    },
    {
      "entropy": 5.4914137840271,
      "epoch": 2.056875907111756,
      "grad_norm": 1.703125,
      "learning_rate": 0.00045768083449621987,
      "loss": 5.1289,
      "mean_token_accuracy": 0.19496792703866958,
      "num_tokens": 47050210.0,
      "step": 22675
    },
    {
      "entropy": 5.5842938899993895,
      "epoch": 2.0573294629898404,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045766194370859347,
      "loss": 5.2672,
      "mean_token_accuracy": 0.17446165457367896,
      "num_tokens": 47060641.0,
      "step": 22680
    },
    {
      "entropy": 5.519540691375733,
      "epoch": 2.0577830188679247,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004576430491434716,
      "loss": 5.2027,
      "mean_token_accuracy": 0.18346914947032927,
      "num_tokens": 47070737.0,
      "step": 22685
    },
    {
      "entropy": 5.573321390151977,
      "epoch": 2.0582365747460085,
      "grad_norm": 1.703125,
      "learning_rate": 0.00045762415080124487,
      "loss": 5.168,
      "mean_token_accuracy": 0.18899074345827102,
      "num_tokens": 47080884.0,
      "step": 22690
    },
    {
      "entropy": 5.460345792770386,
      "epoch": 2.058690130624093,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045760524868230417,
      "loss": 5.1965,
      "mean_token_accuracy": 0.18607835620641708,
      "num_tokens": 47090865.0,
      "step": 22695
    },
    {
      "entropy": 5.479660511016846,
      "epoch": 2.059143686502177,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004575863427870404,
      "loss": 5.159,
      "mean_token_accuracy": 0.19261547476053237,
      "num_tokens": 47100570.0,
      "step": 22700
    },
    {
      "entropy": 5.47812328338623,
      "epoch": 2.0595972423802613,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004575674331158446,
      "loss": 5.1545,
      "mean_token_accuracy": 0.1940849766135216,
      "num_tokens": 47111199.0,
      "step": 22705
    },
    {
      "entropy": 5.465964841842651,
      "epoch": 2.0600507982583456,
      "grad_norm": 1.625,
      "learning_rate": 0.0004575485196691076,
      "loss": 5.1532,
      "mean_token_accuracy": 0.1897718533873558,
      "num_tokens": 47121191.0,
      "step": 22710
    },
    {
      "entropy": 5.5575226783752445,
      "epoch": 2.0605043541364294,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004575296024472209,
      "loss": 5.2285,
      "mean_token_accuracy": 0.18596497178077698,
      "num_tokens": 47131608.0,
      "step": 22715
    },
    {
      "entropy": 5.550099277496338,
      "epoch": 2.0609579100145137,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004575106814505753,
      "loss": 5.1749,
      "mean_token_accuracy": 0.18474768549203874,
      "num_tokens": 47141492.0,
      "step": 22720
    },
    {
      "entropy": 5.4462934017181395,
      "epoch": 2.061411465892598,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004574917566795624,
      "loss": 5.0573,
      "mean_token_accuracy": 0.19715633988380432,
      "num_tokens": 47152576.0,
      "step": 22725
    },
    {
      "entropy": 5.486327981948852,
      "epoch": 2.061865021770682,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045747282813457345,
      "loss": 5.177,
      "mean_token_accuracy": 0.18754634857177735,
      "num_tokens": 47163863.0,
      "step": 22730
    },
    {
      "entropy": 5.485260534286499,
      "epoch": 2.0623185776487665,
      "grad_norm": 1.53125,
      "learning_rate": 0.000457453895816,
      "loss": 5.126,
      "mean_token_accuracy": 0.18809847980737687,
      "num_tokens": 47174140.0,
      "step": 22735
    },
    {
      "entropy": 5.538044404983521,
      "epoch": 2.0627721335268503,
      "grad_norm": 2.15625,
      "learning_rate": 0.00045743495972423346,
      "loss": 5.1874,
      "mean_token_accuracy": 0.18672025054693223,
      "num_tokens": 47184708.0,
      "step": 22740
    },
    {
      "entropy": 5.443181037902832,
      "epoch": 2.0632256894049346,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004574160198596653,
      "loss": 5.1957,
      "mean_token_accuracy": 0.1833674877882004,
      "num_tokens": 47196286.0,
      "step": 22745
    },
    {
      "entropy": 5.438728141784668,
      "epoch": 2.063679245283019,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00045739707622268757,
      "loss": 5.1171,
      "mean_token_accuracy": 0.19421084523200988,
      "num_tokens": 47206774.0,
      "step": 22750
    },
    {
      "entropy": 5.4780984878540036,
      "epoch": 2.064132801161103,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004573781288136917,
      "loss": 5.0294,
      "mean_token_accuracy": 0.19541225284337999,
      "num_tokens": 47217009.0,
      "step": 22755
    },
    {
      "entropy": 5.461385488510132,
      "epoch": 2.0645863570391874,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045735917763306967,
      "loss": 5.1623,
      "mean_token_accuracy": 0.1906869351863861,
      "num_tokens": 47226718.0,
      "step": 22760
    },
    {
      "entropy": 5.429342460632324,
      "epoch": 2.065039912917271,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004573402226812134,
      "loss": 5.0421,
      "mean_token_accuracy": 0.19337306320667266,
      "num_tokens": 47237666.0,
      "step": 22765
    },
    {
      "entropy": 5.455378770828247,
      "epoch": 2.0654934687953554,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045732126395851483,
      "loss": 5.2095,
      "mean_token_accuracy": 0.18235008269548417,
      "num_tokens": 47249230.0,
      "step": 22770
    },
    {
      "entropy": 5.483230113983154,
      "epoch": 2.0659470246734397,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004573023014653661,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19401631504297256,
      "num_tokens": 47259232.0,
      "step": 22775
    },
    {
      "entropy": 5.5046226501464846,
      "epoch": 2.066400580551524,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045728333520215933,
      "loss": 5.1533,
      "mean_token_accuracy": 0.18902986049652098,
      "num_tokens": 47269006.0,
      "step": 22780
    },
    {
      "entropy": 5.510041189193726,
      "epoch": 2.0668541364296082,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004572643651692867,
      "loss": 5.2568,
      "mean_token_accuracy": 0.17894799411296844,
      "num_tokens": 47280048.0,
      "step": 22785
    },
    {
      "entropy": 5.460446119308472,
      "epoch": 2.0673076923076925,
      "grad_norm": 2.4375,
      "learning_rate": 0.00045724539136714063,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19163165241479874,
      "num_tokens": 47290019.0,
      "step": 22790
    },
    {
      "entropy": 5.453388833999634,
      "epoch": 2.0677612481857763,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004572264137961133,
      "loss": 5.1063,
      "mean_token_accuracy": 0.1931236907839775,
      "num_tokens": 47300573.0,
      "step": 22795
    },
    {
      "entropy": 5.414437961578369,
      "epoch": 2.0682148040638606,
      "grad_norm": 1.75,
      "learning_rate": 0.00045720743245659734,
      "loss": 5.1416,
      "mean_token_accuracy": 0.18426740616559983,
      "num_tokens": 47311795.0,
      "step": 22800
    },
    {
      "entropy": 5.420792198181152,
      "epoch": 2.068668359941945,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004571884473489853,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20625277161598204,
      "num_tokens": 47321768.0,
      "step": 22805
    },
    {
      "entropy": 5.506304025650024,
      "epoch": 2.069121915820029,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004571694584736696,
      "loss": 5.2365,
      "mean_token_accuracy": 0.18673464506864548,
      "num_tokens": 47332662.0,
      "step": 22810
    },
    {
      "entropy": 5.458328437805176,
      "epoch": 2.0695754716981134,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004571504658310432,
      "loss": 5.1229,
      "mean_token_accuracy": 0.19072562754154204,
      "num_tokens": 47344355.0,
      "step": 22815
    },
    {
      "entropy": 5.501658725738525,
      "epoch": 2.0700290275761972,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004571314694214987,
      "loss": 5.0791,
      "mean_token_accuracy": 0.19009438455104827,
      "num_tokens": 47354228.0,
      "step": 22820
    },
    {
      "entropy": 5.493122959136963,
      "epoch": 2.0704825834542815,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004571124692454291,
      "loss": 5.1443,
      "mean_token_accuracy": 0.19632193744182586,
      "num_tokens": 47364477.0,
      "step": 22825
    },
    {
      "entropy": 5.516074323654175,
      "epoch": 2.0709361393323658,
      "grad_norm": 1.65625,
      "learning_rate": 0.00045709346530322715,
      "loss": 5.2118,
      "mean_token_accuracy": 0.1879907563328743,
      "num_tokens": 47375745.0,
      "step": 22830
    },
    {
      "entropy": 5.464600133895874,
      "epoch": 2.07138969521045,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004570744575952859,
      "loss": 5.2357,
      "mean_token_accuracy": 0.19553990215063094,
      "num_tokens": 47385341.0,
      "step": 22835
    },
    {
      "entropy": 5.45963487625122,
      "epoch": 2.0718432510885343,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004570554461219985,
      "loss": 5.1614,
      "mean_token_accuracy": 0.19064618051052093,
      "num_tokens": 47396266.0,
      "step": 22840
    },
    {
      "entropy": 5.498890399932861,
      "epoch": 2.072296806966618,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00045703643088375806,
      "loss": 5.1559,
      "mean_token_accuracy": 0.18707890063524246,
      "num_tokens": 47406772.0,
      "step": 22845
    },
    {
      "entropy": 5.516940641403198,
      "epoch": 2.0727503628447024,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045701741188095793,
      "loss": 5.2364,
      "mean_token_accuracy": 0.18621406555175782,
      "num_tokens": 47416833.0,
      "step": 22850
    },
    {
      "entropy": 5.465582466125488,
      "epoch": 2.0732039187227866,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004569983891139912,
      "loss": 5.136,
      "mean_token_accuracy": 0.19301635026931763,
      "num_tokens": 47426954.0,
      "step": 22855
    },
    {
      "entropy": 5.4051250457763675,
      "epoch": 2.073657474600871,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045697936258325155,
      "loss": 5.0991,
      "mean_token_accuracy": 0.19658381044864653,
      "num_tokens": 47437520.0,
      "step": 22860
    },
    {
      "entropy": 5.4565520763397215,
      "epoch": 2.074111030478955,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045696033228913226,
      "loss": 5.1158,
      "mean_token_accuracy": 0.18996300548315048,
      "num_tokens": 47447977.0,
      "step": 22865
    },
    {
      "entropy": 5.500593471527099,
      "epoch": 2.074564586357039,
      "grad_norm": 2.21875,
      "learning_rate": 0.00045694129823202684,
      "loss": 5.1688,
      "mean_token_accuracy": 0.18784139156341553,
      "num_tokens": 47458458.0,
      "step": 22870
    },
    {
      "entropy": 5.505392742156983,
      "epoch": 2.0750181422351233,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004569222604123291,
      "loss": 5.1748,
      "mean_token_accuracy": 0.18677269518375397,
      "num_tokens": 47469048.0,
      "step": 22875
    },
    {
      "entropy": 5.528280591964721,
      "epoch": 2.0754716981132075,
      "grad_norm": 1.625,
      "learning_rate": 0.0004569032188304326,
      "loss": 5.214,
      "mean_token_accuracy": 0.18188802897930145,
      "num_tokens": 47480353.0,
      "step": 22880
    },
    {
      "entropy": 5.406366395950317,
      "epoch": 2.075925253991292,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004568841734867313,
      "loss": 5.0383,
      "mean_token_accuracy": 0.19404292553663255,
      "num_tokens": 47489951.0,
      "step": 22885
    },
    {
      "entropy": 5.404531621932984,
      "epoch": 2.076378809869376,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004568651243816188,
      "loss": 5.0947,
      "mean_token_accuracy": 0.19256189167499543,
      "num_tokens": 47499015.0,
      "step": 22890
    },
    {
      "entropy": 5.53515133857727,
      "epoch": 2.07683236574746,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00045684607151548923,
      "loss": 5.145,
      "mean_token_accuracy": 0.1932419493794441,
      "num_tokens": 47508694.0,
      "step": 22895
    },
    {
      "entropy": 5.6298990726470945,
      "epoch": 2.077285921625544,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045682701488873655,
      "loss": 5.3124,
      "mean_token_accuracy": 0.1737601324915886,
      "num_tokens": 47521336.0,
      "step": 22900
    },
    {
      "entropy": 5.523710107803344,
      "epoch": 2.0777394775036284,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004568079545017549,
      "loss": 5.2965,
      "mean_token_accuracy": 0.18147259205579758,
      "num_tokens": 47531857.0,
      "step": 22905
    },
    {
      "entropy": 5.540319108963013,
      "epoch": 2.0781930333817127,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004567888903549385,
      "loss": 5.1807,
      "mean_token_accuracy": 0.17992958277463914,
      "num_tokens": 47542224.0,
      "step": 22910
    },
    {
      "entropy": 5.4791257858276365,
      "epoch": 2.078646589259797,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004567698224486813,
      "loss": 5.1927,
      "mean_token_accuracy": 0.18805686980485917,
      "num_tokens": 47552370.0,
      "step": 22915
    },
    {
      "entropy": 5.615638542175293,
      "epoch": 2.0791001451378808,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045675075078337803,
      "loss": 5.137,
      "mean_token_accuracy": 0.18939531445503235,
      "num_tokens": 47561997.0,
      "step": 22920
    },
    {
      "entropy": 5.51215181350708,
      "epoch": 2.079553701015965,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004567316753594229,
      "loss": 5.1658,
      "mean_token_accuracy": 0.19293537139892578,
      "num_tokens": 47571968.0,
      "step": 22925
    },
    {
      "entropy": 5.472894477844238,
      "epoch": 2.0800072568940493,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00045671259617721037,
      "loss": 5.1852,
      "mean_token_accuracy": 0.18640337139368057,
      "num_tokens": 47583459.0,
      "step": 22930
    },
    {
      "entropy": 5.4187843799591064,
      "epoch": 2.0804608127721336,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000456693513237135,
      "loss": 5.1005,
      "mean_token_accuracy": 0.1941688686609268,
      "num_tokens": 47594026.0,
      "step": 22935
    },
    {
      "entropy": 5.523876619338989,
      "epoch": 2.080914368650218,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004566744265395916,
      "loss": 5.1829,
      "mean_token_accuracy": 0.18704994022846222,
      "num_tokens": 47605246.0,
      "step": 22940
    },
    {
      "entropy": 5.553946161270142,
      "epoch": 2.081367924528302,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045665533608497467,
      "loss": 5.227,
      "mean_token_accuracy": 0.18494464159011842,
      "num_tokens": 47616702.0,
      "step": 22945
    },
    {
      "entropy": 5.596485996246338,
      "epoch": 2.081821480406386,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045663624187367914,
      "loss": 5.2216,
      "mean_token_accuracy": 0.181157748401165,
      "num_tokens": 47627616.0,
      "step": 22950
    },
    {
      "entropy": 5.465940809249878,
      "epoch": 2.08227503628447,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004566171439060998,
      "loss": 5.1185,
      "mean_token_accuracy": 0.1917110338807106,
      "num_tokens": 47637287.0,
      "step": 22955
    },
    {
      "entropy": 5.380648565292359,
      "epoch": 2.0827285921625545,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004565980421826317,
      "loss": 5.14,
      "mean_token_accuracy": 0.19133764803409575,
      "num_tokens": 47647474.0,
      "step": 22960
    },
    {
      "entropy": 5.552446985244751,
      "epoch": 2.0831821480406387,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004565789367036697,
      "loss": 5.2011,
      "mean_token_accuracy": 0.18841667771339415,
      "num_tokens": 47657036.0,
      "step": 22965
    },
    {
      "entropy": 5.511623191833496,
      "epoch": 2.083635703918723,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004565598274696091,
      "loss": 5.0745,
      "mean_token_accuracy": 0.18941321969032288,
      "num_tokens": 47668026.0,
      "step": 22970
    },
    {
      "entropy": 5.556039524078369,
      "epoch": 2.084089259796807,
      "grad_norm": 1.921875,
      "learning_rate": 0.00045654071448084506,
      "loss": 5.2416,
      "mean_token_accuracy": 0.1839629352092743,
      "num_tokens": 47679508.0,
      "step": 22975
    },
    {
      "entropy": 5.476069831848145,
      "epoch": 2.084542815674891,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045652159773777273,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19029734134674073,
      "num_tokens": 47688474.0,
      "step": 22980
    },
    {
      "entropy": 5.604936504364014,
      "epoch": 2.0849963715529753,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004565024772407875,
      "loss": 5.3292,
      "mean_token_accuracy": 0.17761608213186264,
      "num_tokens": 47698675.0,
      "step": 22985
    },
    {
      "entropy": 5.471998643875122,
      "epoch": 2.0854499274310596,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004564833529902848,
      "loss": 5.1134,
      "mean_token_accuracy": 0.1993321508169174,
      "num_tokens": 47709214.0,
      "step": 22990
    },
    {
      "entropy": 5.532295274734497,
      "epoch": 2.085903483309144,
      "grad_norm": 1.71875,
      "learning_rate": 0.00045646422498666,
      "loss": 5.2462,
      "mean_token_accuracy": 0.18700508922338485,
      "num_tokens": 47721000.0,
      "step": 22995
    },
    {
      "entropy": 5.4448590755462645,
      "epoch": 2.0863570391872277,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00045644509323030893,
      "loss": 5.1353,
      "mean_token_accuracy": 0.1867292881011963,
      "num_tokens": 47731218.0,
      "step": 23000
    },
    {
      "entropy": 5.466960334777832,
      "epoch": 2.086810595065312,
      "grad_norm": 1.375,
      "learning_rate": 0.00045642595772162704,
      "loss": 5.1274,
      "mean_token_accuracy": 0.18860442489385604,
      "num_tokens": 47742339.0,
      "step": 23005
    },
    {
      "entropy": 5.522220087051392,
      "epoch": 2.0872641509433962,
      "grad_norm": 2.421875,
      "learning_rate": 0.00045640681846101014,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18443699628114701,
      "num_tokens": 47752858.0,
      "step": 23010
    },
    {
      "entropy": 5.3967344760894775,
      "epoch": 2.0877177068214805,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004563876754488539,
      "loss": 5.0845,
      "mean_token_accuracy": 0.19560057818889617,
      "num_tokens": 47763303.0,
      "step": 23015
    },
    {
      "entropy": 5.439888381958008,
      "epoch": 2.0881712626995648,
      "grad_norm": 1.625,
      "learning_rate": 0.0004563685286855544,
      "loss": 5.1261,
      "mean_token_accuracy": 0.19129272699356079,
      "num_tokens": 47773212.0,
      "step": 23020
    },
    {
      "entropy": 5.540351438522339,
      "epoch": 2.0886248185776486,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004563493781715075,
      "loss": 5.2699,
      "mean_token_accuracy": 0.18439988046884537,
      "num_tokens": 47783376.0,
      "step": 23025
    },
    {
      "entropy": 5.453643035888672,
      "epoch": 2.089078374455733,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004563302239071093,
      "loss": 5.1453,
      "mean_token_accuracy": 0.19186050593852996,
      "num_tokens": 47794859.0,
      "step": 23030
    },
    {
      "entropy": 5.519970655441284,
      "epoch": 2.089531930333817,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045631106589275576,
      "loss": 5.158,
      "mean_token_accuracy": 0.19167364984750748,
      "num_tokens": 47804721.0,
      "step": 23035
    },
    {
      "entropy": 5.452056550979615,
      "epoch": 2.0899854862119014,
      "grad_norm": 2.015625,
      "learning_rate": 0.00045629190412884324,
      "loss": 5.0388,
      "mean_token_accuracy": 0.19643571823835373,
      "num_tokens": 47814701.0,
      "step": 23040
    },
    {
      "entropy": 5.490446519851685,
      "epoch": 2.0904390420899857,
      "grad_norm": 2.03125,
      "learning_rate": 0.00045627273861576786,
      "loss": 5.0862,
      "mean_token_accuracy": 0.19314375668764114,
      "num_tokens": 47823941.0,
      "step": 23045
    },
    {
      "entropy": 5.481373882293701,
      "epoch": 2.0908925979680695,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045625356935392614,
      "loss": 5.1664,
      "mean_token_accuracy": 0.18839575052261354,
      "num_tokens": 47835793.0,
      "step": 23050
    },
    {
      "entropy": 5.5196010112762455,
      "epoch": 2.0913461538461537,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004562343963437144,
      "loss": 5.1957,
      "mean_token_accuracy": 0.18791369050741197,
      "num_tokens": 47845677.0,
      "step": 23055
    },
    {
      "entropy": 5.513240337371826,
      "epoch": 2.091799709724238,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004562152195855291,
      "loss": 5.1358,
      "mean_token_accuracy": 0.193760547041893,
      "num_tokens": 47855611.0,
      "step": 23060
    },
    {
      "entropy": 5.441588449478149,
      "epoch": 2.0922532656023223,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004561960390797669,
      "loss": 5.1382,
      "mean_token_accuracy": 0.19349662363529205,
      "num_tokens": 47865115.0,
      "step": 23065
    },
    {
      "entropy": 5.5522225856781,
      "epoch": 2.0927068214804065,
      "grad_norm": 1.75,
      "learning_rate": 0.00045617685482682454,
      "loss": 5.2298,
      "mean_token_accuracy": 0.18527799993753433,
      "num_tokens": 47876690.0,
      "step": 23070
    },
    {
      "entropy": 5.479308986663819,
      "epoch": 2.0931603773584904,
      "grad_norm": 1.5,
      "learning_rate": 0.0004561576668270986,
      "loss": 5.1637,
      "mean_token_accuracy": 0.18955969363451003,
      "num_tokens": 47887709.0,
      "step": 23075
    },
    {
      "entropy": 5.475148439407349,
      "epoch": 2.0936139332365746,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045613847508098587,
      "loss": 5.1561,
      "mean_token_accuracy": 0.18796104788780213,
      "num_tokens": 47898209.0,
      "step": 23080
    },
    {
      "entropy": 5.497140264511108,
      "epoch": 2.094067489114659,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045611927958888335,
      "loss": 5.0234,
      "mean_token_accuracy": 0.19765589833259584,
      "num_tokens": 47907494.0,
      "step": 23085
    },
    {
      "entropy": 5.452366685867309,
      "epoch": 2.094521044992743,
      "grad_norm": 1.609375,
      "learning_rate": 0.00045610008035118797,
      "loss": 5.1565,
      "mean_token_accuracy": 0.18953470885753632,
      "num_tokens": 47916923.0,
      "step": 23090
    },
    {
      "entropy": 5.4416608810424805,
      "epoch": 2.0949746008708274,
      "grad_norm": 2.328125,
      "learning_rate": 0.00045608087736829687,
      "loss": 5.0986,
      "mean_token_accuracy": 0.19400160908699035,
      "num_tokens": 47927043.0,
      "step": 23095
    },
    {
      "entropy": 5.536021757125854,
      "epoch": 2.0954281567489113,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000456061670640607,
      "loss": 5.1404,
      "mean_token_accuracy": 0.19012030810117722,
      "num_tokens": 47937296.0,
      "step": 23100
    },
    {
      "entropy": 5.510265779495239,
      "epoch": 2.0958817126269955,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004560424601685157,
      "loss": 5.1854,
      "mean_token_accuracy": 0.19302549362182617,
      "num_tokens": 47948268.0,
      "step": 23105
    },
    {
      "entropy": 5.447001123428345,
      "epoch": 2.09633526850508,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004560232459524201,
      "loss": 5.0949,
      "mean_token_accuracy": 0.18754720836877822,
      "num_tokens": 47958375.0,
      "step": 23110
    },
    {
      "entropy": 5.494138622283936,
      "epoch": 2.096788824383164,
      "grad_norm": 1.671875,
      "learning_rate": 0.00045600402799271766,
      "loss": 5.1894,
      "mean_token_accuracy": 0.19686047583818436,
      "num_tokens": 47968469.0,
      "step": 23115
    },
    {
      "entropy": 5.462585067749023,
      "epoch": 2.0972423802612483,
      "grad_norm": 1.25,
      "learning_rate": 0.0004559848062898058,
      "loss": 5.0963,
      "mean_token_accuracy": 0.19786120802164078,
      "num_tokens": 47979630.0,
      "step": 23120
    },
    {
      "entropy": 5.45167555809021,
      "epoch": 2.097695936139332,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004559655808440821,
      "loss": 5.1492,
      "mean_token_accuracy": 0.18776504844427108,
      "num_tokens": 47989828.0,
      "step": 23125
    },
    {
      "entropy": 5.505210733413696,
      "epoch": 2.0981494920174164,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045594635165594404,
      "loss": 5.2684,
      "mean_token_accuracy": 0.17908042669296265,
      "num_tokens": 47999219.0,
      "step": 23130
    },
    {
      "entropy": 5.528690147399902,
      "epoch": 2.0986030478955007,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045592711872578925,
      "loss": 5.143,
      "mean_token_accuracy": 0.195282880961895,
      "num_tokens": 48009244.0,
      "step": 23135
    },
    {
      "entropy": 5.4871889591217045,
      "epoch": 2.099056603773585,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045590788205401557,
      "loss": 5.0987,
      "mean_token_accuracy": 0.19328412115573884,
      "num_tokens": 48019066.0,
      "step": 23140
    },
    {
      "entropy": 5.398633718490601,
      "epoch": 2.099510159651669,
      "grad_norm": 1.75,
      "learning_rate": 0.00045588864164102086,
      "loss": 5.1031,
      "mean_token_accuracy": 0.1925794705748558,
      "num_tokens": 48028704.0,
      "step": 23145
    },
    {
      "entropy": 5.498190927505493,
      "epoch": 2.0999637155297535,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045586939748720284,
      "loss": 5.143,
      "mean_token_accuracy": 0.18692257404327392,
      "num_tokens": 48039210.0,
      "step": 23150
    },
    {
      "entropy": 5.556540250778198,
      "epoch": 2.1004172714078373,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00045585014959295965,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1835030198097229,
      "num_tokens": 48048446.0,
      "step": 23155
    },
    {
      "entropy": 5.467826700210571,
      "epoch": 2.1008708272859216,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045583089795868934,
      "loss": 5.1595,
      "mean_token_accuracy": 0.19161137640476228,
      "num_tokens": 48058880.0,
      "step": 23160
    },
    {
      "entropy": 5.5096235275268555,
      "epoch": 2.101324383164006,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004558116425847899,
      "loss": 5.1807,
      "mean_token_accuracy": 0.19925292432308198,
      "num_tokens": 48070224.0,
      "step": 23165
    },
    {
      "entropy": 5.497380781173706,
      "epoch": 2.10177793904209,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004557923834716596,
      "loss": 5.1969,
      "mean_token_accuracy": 0.18574568778276443,
      "num_tokens": 48080914.0,
      "step": 23170
    },
    {
      "entropy": 5.451671028137207,
      "epoch": 2.1022314949201744,
      "grad_norm": 1.25,
      "learning_rate": 0.00045577312061969677,
      "loss": 5.1566,
      "mean_token_accuracy": 0.18882377147674562,
      "num_tokens": 48090873.0,
      "step": 23175
    },
    {
      "entropy": 5.60972957611084,
      "epoch": 2.102685050798258,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004557538540292997,
      "loss": 5.2093,
      "mean_token_accuracy": 0.1849701389670372,
      "num_tokens": 48101885.0,
      "step": 23180
    },
    {
      "entropy": 5.529236507415772,
      "epoch": 2.1031386066763424,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004557345837008669,
      "loss": 5.2052,
      "mean_token_accuracy": 0.1868951916694641,
      "num_tokens": 48111472.0,
      "step": 23185
    },
    {
      "entropy": 5.378008270263672,
      "epoch": 2.1035921625544267,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045571530963479684,
      "loss": 5.0697,
      "mean_token_accuracy": 0.19511385262012482,
      "num_tokens": 48121351.0,
      "step": 23190
    },
    {
      "entropy": 5.5516884326934814,
      "epoch": 2.104045718432511,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045569603183148814,
      "loss": 5.1498,
      "mean_token_accuracy": 0.19740214943885803,
      "num_tokens": 48131104.0,
      "step": 23195
    },
    {
      "entropy": 5.543045711517334,
      "epoch": 2.1044992743105952,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004556767502913395,
      "loss": 5.2464,
      "mean_token_accuracy": 0.1838505744934082,
      "num_tokens": 48142895.0,
      "step": 23200
    },
    {
      "entropy": 5.394887256622314,
      "epoch": 2.104952830188679,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045565746501474946,
      "loss": 5.0225,
      "mean_token_accuracy": 0.19256235659122467,
      "num_tokens": 48153294.0,
      "step": 23205
    },
    {
      "entropy": 5.516482830047607,
      "epoch": 2.1054063860667633,
      "grad_norm": 1.53125,
      "learning_rate": 0.00045563817600211714,
      "loss": 5.2079,
      "mean_token_accuracy": 0.18306470811367034,
      "num_tokens": 48164319.0,
      "step": 23210
    },
    {
      "entropy": 5.488808345794678,
      "epoch": 2.1058599419448476,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045561888325384124,
      "loss": 5.1711,
      "mean_token_accuracy": 0.1947334960103035,
      "num_tokens": 48173918.0,
      "step": 23215
    },
    {
      "entropy": 5.524158954620361,
      "epoch": 2.106313497822932,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004555995867703209,
      "loss": 5.1596,
      "mean_token_accuracy": 0.18677118718624114,
      "num_tokens": 48184127.0,
      "step": 23220
    },
    {
      "entropy": 5.42077522277832,
      "epoch": 2.106767053701016,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045558028655195504,
      "loss": 5.1301,
      "mean_token_accuracy": 0.185513998568058,
      "num_tokens": 48194198.0,
      "step": 23225
    },
    {
      "entropy": 5.420413827896118,
      "epoch": 2.1072206095791,
      "grad_norm": 1.53125,
      "learning_rate": 0.00045556098259914293,
      "loss": 5.1066,
      "mean_token_accuracy": 0.18546139001846312,
      "num_tokens": 48204654.0,
      "step": 23230
    },
    {
      "entropy": 5.544856977462769,
      "epoch": 2.1076741654571842,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045554167491228357,
      "loss": 5.2649,
      "mean_token_accuracy": 0.17661546170711517,
      "num_tokens": 48215322.0,
      "step": 23235
    },
    {
      "entropy": 5.508966159820557,
      "epoch": 2.1081277213352685,
      "grad_norm": 2.09375,
      "learning_rate": 0.00045552236349177637,
      "loss": 5.0458,
      "mean_token_accuracy": 0.2006254807114601,
      "num_tokens": 48226142.0,
      "step": 23240
    },
    {
      "entropy": 5.442447662353516,
      "epoch": 2.1085812772133528,
      "grad_norm": 1.5,
      "learning_rate": 0.00045550304833802066,
      "loss": 5.0858,
      "mean_token_accuracy": 0.1918679177761078,
      "num_tokens": 48236896.0,
      "step": 23245
    },
    {
      "entropy": 5.410110139846802,
      "epoch": 2.109034833091437,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000455483729451416,
      "loss": 5.1059,
      "mean_token_accuracy": 0.19121052622795104,
      "num_tokens": 48247425.0,
      "step": 23250
    },
    {
      "entropy": 5.535222673416138,
      "epoch": 2.109488388969521,
      "grad_norm": 1.671875,
      "learning_rate": 0.00045546440683236173,
      "loss": 5.2068,
      "mean_token_accuracy": 0.18610516041517258,
      "num_tokens": 48258560.0,
      "step": 23255
    },
    {
      "entropy": 5.564100313186645,
      "epoch": 2.109941944847605,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045544508048125763,
      "loss": 5.2594,
      "mean_token_accuracy": 0.18516540080308913,
      "num_tokens": 48269469.0,
      "step": 23260
    },
    {
      "entropy": 5.458248472213745,
      "epoch": 2.1103955007256894,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004554257503985032,
      "loss": 5.1399,
      "mean_token_accuracy": 0.1948898807168007,
      "num_tokens": 48278847.0,
      "step": 23265
    },
    {
      "entropy": 5.422713041305542,
      "epoch": 2.1108490566037736,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045540641658449834,
      "loss": 5.0757,
      "mean_token_accuracy": 0.1891564682126045,
      "num_tokens": 48288091.0,
      "step": 23270
    },
    {
      "entropy": 5.438827180862427,
      "epoch": 2.111302612481858,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045538707903964275,
      "loss": 5.0795,
      "mean_token_accuracy": 0.18970006108283996,
      "num_tokens": 48297946.0,
      "step": 23275
    },
    {
      "entropy": 5.490026998519897,
      "epoch": 2.1117561683599417,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045536773776433647,
      "loss": 5.1393,
      "mean_token_accuracy": 0.19388551712036134,
      "num_tokens": 48307650.0,
      "step": 23280
    },
    {
      "entropy": 5.5355151176452635,
      "epoch": 2.112209724238026,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045534839275897937,
      "loss": 5.2571,
      "mean_token_accuracy": 0.17679874897003173,
      "num_tokens": 48318824.0,
      "step": 23285
    },
    {
      "entropy": 5.475306224822998,
      "epoch": 2.1126632801161103,
      "grad_norm": 1.5,
      "learning_rate": 0.00045532904402397157,
      "loss": 5.1123,
      "mean_token_accuracy": 0.18936043679714204,
      "num_tokens": 48329909.0,
      "step": 23290
    },
    {
      "entropy": 5.515163612365723,
      "epoch": 2.1131168359941945,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004553096915597132,
      "loss": 5.2656,
      "mean_token_accuracy": 0.18386084735393524,
      "num_tokens": 48340142.0,
      "step": 23295
    },
    {
      "entropy": 5.520307111740112,
      "epoch": 2.113570391872279,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004552903353666044,
      "loss": 5.1132,
      "mean_token_accuracy": 0.19528384804725646,
      "num_tokens": 48350540.0,
      "step": 23300
    },
    {
      "entropy": 5.438007640838623,
      "epoch": 2.114023947750363,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045527097544504556,
      "loss": 5.0668,
      "mean_token_accuracy": 0.19666175693273544,
      "num_tokens": 48360153.0,
      "step": 23305
    },
    {
      "entropy": 5.4257824420928955,
      "epoch": 2.114477503628447,
      "grad_norm": 2.46875,
      "learning_rate": 0.000455251611795437,
      "loss": 5.1229,
      "mean_token_accuracy": 0.19187461733818054,
      "num_tokens": 48370088.0,
      "step": 23310
    },
    {
      "entropy": 5.564465618133545,
      "epoch": 2.114931059506531,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004552322444181792,
      "loss": 5.1697,
      "mean_token_accuracy": 0.1852997362613678,
      "num_tokens": 48381325.0,
      "step": 23315
    },
    {
      "entropy": 5.510464906692505,
      "epoch": 2.1153846153846154,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004552128733136726,
      "loss": 5.1494,
      "mean_token_accuracy": 0.1880299836397171,
      "num_tokens": 48393077.0,
      "step": 23320
    },
    {
      "entropy": 5.44994478225708,
      "epoch": 2.1158381712626997,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004551934984823179,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19198026657104492,
      "num_tokens": 48403260.0,
      "step": 23325
    },
    {
      "entropy": 5.452255725860596,
      "epoch": 2.116291727140784,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004551741199245157,
      "loss": 5.1365,
      "mean_token_accuracy": 0.1914306327700615,
      "num_tokens": 48413616.0,
      "step": 23330
    },
    {
      "entropy": 5.51603741645813,
      "epoch": 2.1167452830188678,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00045515473764066685,
      "loss": 5.1495,
      "mean_token_accuracy": 0.19222956001758576,
      "num_tokens": 48424189.0,
      "step": 23335
    },
    {
      "entropy": 5.499660587310791,
      "epoch": 2.117198838896952,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045513535163117205,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1861589252948761,
      "num_tokens": 48433603.0,
      "step": 23340
    },
    {
      "entropy": 5.507645082473755,
      "epoch": 2.1176523947750363,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004551159618964323,
      "loss": 5.1695,
      "mean_token_accuracy": 0.1872040942311287,
      "num_tokens": 48443043.0,
      "step": 23345
    },
    {
      "entropy": 5.434371995925903,
      "epoch": 2.1181059506531206,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004550965684368486,
      "loss": 5.0385,
      "mean_token_accuracy": 0.2034829556941986,
      "num_tokens": 48452765.0,
      "step": 23350
    },
    {
      "entropy": 5.449034833908081,
      "epoch": 2.118559506531205,
      "grad_norm": 1.125,
      "learning_rate": 0.0004550771712528219,
      "loss": 5.1089,
      "mean_token_accuracy": 0.1937694489955902,
      "num_tokens": 48463191.0,
      "step": 23355
    },
    {
      "entropy": 5.47435245513916,
      "epoch": 2.1190130624092887,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004550577703447534,
      "loss": 5.107,
      "mean_token_accuracy": 0.1956036388874054,
      "num_tokens": 48473280.0,
      "step": 23360
    },
    {
      "entropy": 5.450800514221191,
      "epoch": 2.119466618287373,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045503836571304443,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19549233764410018,
      "num_tokens": 48483465.0,
      "step": 23365
    },
    {
      "entropy": 5.413967418670654,
      "epoch": 2.119920174165457,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004550189573580961,
      "loss": 5.0837,
      "mean_token_accuracy": 0.19415228068828583,
      "num_tokens": 48494497.0,
      "step": 23370
    },
    {
      "entropy": 5.428210020065308,
      "epoch": 2.1203737300435415,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004549995452803098,
      "loss": 5.1543,
      "mean_token_accuracy": 0.19250529855489731,
      "num_tokens": 48504106.0,
      "step": 23375
    },
    {
      "entropy": 5.541994094848633,
      "epoch": 2.1208272859216257,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004549801294800872,
      "loss": 5.1852,
      "mean_token_accuracy": 0.18428828567266464,
      "num_tokens": 48515729.0,
      "step": 23380
    },
    {
      "entropy": 5.528612232208252,
      "epoch": 2.1212808417997095,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004549607099578295,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1846257284283638,
      "num_tokens": 48526611.0,
      "step": 23385
    },
    {
      "entropy": 5.424442768096924,
      "epoch": 2.121734397677794,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004549412867139386,
      "loss": 5.0911,
      "mean_token_accuracy": 0.1911414623260498,
      "num_tokens": 48537016.0,
      "step": 23390
    },
    {
      "entropy": 5.478576803207398,
      "epoch": 2.122187953555878,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045492185974881587,
      "loss": 5.172,
      "mean_token_accuracy": 0.18557947278022766,
      "num_tokens": 48546786.0,
      "step": 23395
    },
    {
      "entropy": 5.451839113235474,
      "epoch": 2.1226415094339623,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004549024290628634,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19080354273319244,
      "num_tokens": 48556306.0,
      "step": 23400
    },
    {
      "entropy": 5.541786575317383,
      "epoch": 2.1230950653120466,
      "grad_norm": 1.375,
      "learning_rate": 0.0004548829946564827,
      "loss": 5.2016,
      "mean_token_accuracy": 0.1899572104215622,
      "num_tokens": 48566303.0,
      "step": 23405
    },
    {
      "entropy": 5.5550390720367435,
      "epoch": 2.1235486211901304,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004548635565300759,
      "loss": 5.1602,
      "mean_token_accuracy": 0.19345995485782624,
      "num_tokens": 48576313.0,
      "step": 23410
    },
    {
      "entropy": 5.4529375553131105,
      "epoch": 2.1240021770682147,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004548441146840449,
      "loss": 5.136,
      "mean_token_accuracy": 0.19145744293928146,
      "num_tokens": 48587133.0,
      "step": 23415
    },
    {
      "entropy": 5.527559280395508,
      "epoch": 2.124455732946299,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00045482466911879184,
      "loss": 5.2792,
      "mean_token_accuracy": 0.1834419399499893,
      "num_tokens": 48598524.0,
      "step": 23420
    },
    {
      "entropy": 5.451411199569702,
      "epoch": 2.1249092888243832,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045480521983471875,
      "loss": 5.1106,
      "mean_token_accuracy": 0.19435191303491592,
      "num_tokens": 48609870.0,
      "step": 23425
    },
    {
      "entropy": 5.4907615184783936,
      "epoch": 2.1253628447024675,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004547857668322279,
      "loss": 5.1635,
      "mean_token_accuracy": 0.1947512537240982,
      "num_tokens": 48619828.0,
      "step": 23430
    },
    {
      "entropy": 5.454772567749023,
      "epoch": 2.1258164005805513,
      "grad_norm": 1.375,
      "learning_rate": 0.0004547663101117215,
      "loss": 5.1005,
      "mean_token_accuracy": 0.19613279700279235,
      "num_tokens": 48630361.0,
      "step": 23435
    },
    {
      "entropy": 5.431140470504761,
      "epoch": 2.1262699564586356,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045474684967360216,
      "loss": 5.1299,
      "mean_token_accuracy": 0.18157708048820495,
      "num_tokens": 48641113.0,
      "step": 23440
    },
    {
      "entropy": 5.567628240585327,
      "epoch": 2.12672351233672,
      "grad_norm": 1.34375,
      "learning_rate": 0.000454727385518272,
      "loss": 5.2173,
      "mean_token_accuracy": 0.17845577895641326,
      "num_tokens": 48651777.0,
      "step": 23445
    },
    {
      "entropy": 5.5861986637115475,
      "epoch": 2.127177068214804,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004547079176461338,
      "loss": 5.256,
      "mean_token_accuracy": 0.18326229602098465,
      "num_tokens": 48663237.0,
      "step": 23450
    },
    {
      "entropy": 5.42162857055664,
      "epoch": 2.1276306240928884,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004546884460575901,
      "loss": 5.1441,
      "mean_token_accuracy": 0.19412964284420015,
      "num_tokens": 48673932.0,
      "step": 23455
    },
    {
      "entropy": 5.432135009765625,
      "epoch": 2.1280841799709727,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004546689707530434,
      "loss": 5.0725,
      "mean_token_accuracy": 0.19587133824825287,
      "num_tokens": 48684304.0,
      "step": 23460
    },
    {
      "entropy": 5.404352140426636,
      "epoch": 2.1285377358490565,
      "grad_norm": 1.609375,
      "learning_rate": 0.00045464949173289676,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19175645262002944,
      "num_tokens": 48695569.0,
      "step": 23465
    },
    {
      "entropy": 5.404065752029419,
      "epoch": 2.1289912917271407,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004546300089975528,
      "loss": 5.1231,
      "mean_token_accuracy": 0.19128832668066026,
      "num_tokens": 48706550.0,
      "step": 23470
    },
    {
      "entropy": 5.585313844680786,
      "epoch": 2.129444847605225,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045461052254741445,
      "loss": 5.1916,
      "mean_token_accuracy": 0.18274935632944106,
      "num_tokens": 48716852.0,
      "step": 23475
    },
    {
      "entropy": 5.464076185226441,
      "epoch": 2.1298984034833093,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045459103238288474,
      "loss": 5.0818,
      "mean_token_accuracy": 0.19232735335826873,
      "num_tokens": 48726607.0,
      "step": 23480
    },
    {
      "entropy": 5.4788977146148685,
      "epoch": 2.130351959361393,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004545715385043667,
      "loss": 5.3022,
      "mean_token_accuracy": 0.18026136010885238,
      "num_tokens": 48738011.0,
      "step": 23485
    },
    {
      "entropy": 5.5107849597930905,
      "epoch": 2.1308055152394774,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004545520409122635,
      "loss": 5.1851,
      "mean_token_accuracy": 0.18437408208847045,
      "num_tokens": 48747899.0,
      "step": 23490
    },
    {
      "entropy": 5.552368640899658,
      "epoch": 2.1312590711175616,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004545325396069783,
      "loss": 5.1921,
      "mean_token_accuracy": 0.18261688351631164,
      "num_tokens": 48758719.0,
      "step": 23495
    },
    {
      "entropy": 5.560001564025879,
      "epoch": 2.131712626995646,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004545130345889145,
      "loss": 5.2115,
      "mean_token_accuracy": 0.18920855075120926,
      "num_tokens": 48769677.0,
      "step": 23500
    },
    {
      "entropy": 5.458061075210571,
      "epoch": 2.13216618287373,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004544935258584754,
      "loss": 5.1566,
      "mean_token_accuracy": 0.1932092934846878,
      "num_tokens": 48778440.0,
      "step": 23505
    },
    {
      "entropy": 5.561990404129029,
      "epoch": 2.1326197387518144,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004544740134160644,
      "loss": 5.2548,
      "mean_token_accuracy": 0.18236325681209564,
      "num_tokens": 48788328.0,
      "step": 23510
    },
    {
      "entropy": 5.553691101074219,
      "epoch": 2.1330732946298983,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004544544972620851,
      "loss": 5.2118,
      "mean_token_accuracy": 0.18092956095933915,
      "num_tokens": 48799218.0,
      "step": 23515
    },
    {
      "entropy": 5.517909193038941,
      "epoch": 2.1335268505079825,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004544349773969411,
      "loss": 5.1333,
      "mean_token_accuracy": 0.18315283805131913,
      "num_tokens": 48809648.0,
      "step": 23520
    },
    {
      "entropy": 5.498174142837525,
      "epoch": 2.133980406386067,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004544154538210359,
      "loss": 5.1668,
      "mean_token_accuracy": 0.18409568518400193,
      "num_tokens": 48819965.0,
      "step": 23525
    },
    {
      "entropy": 5.587610864639283,
      "epoch": 2.134433962264151,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004543959265347735,
      "loss": 5.2379,
      "mean_token_accuracy": 0.1871068298816681,
      "num_tokens": 48829862.0,
      "step": 23530
    },
    {
      "entropy": 5.506388092041016,
      "epoch": 2.1348875181422353,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045437639553855757,
      "loss": 5.1669,
      "mean_token_accuracy": 0.18522842228412628,
      "num_tokens": 48840551.0,
      "step": 23535
    },
    {
      "entropy": 5.393575525283813,
      "epoch": 2.135341074020319,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004543568608327921,
      "loss": 5.1518,
      "mean_token_accuracy": 0.19597553461790085,
      "num_tokens": 48851032.0,
      "step": 23540
    },
    {
      "entropy": 5.48370475769043,
      "epoch": 2.1357946298984034,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045433732241788105,
      "loss": 5.2281,
      "mean_token_accuracy": 0.18249870240688323,
      "num_tokens": 48861773.0,
      "step": 23545
    },
    {
      "entropy": 5.527497577667236,
      "epoch": 2.1362481857764877,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045431778029422846,
      "loss": 5.2586,
      "mean_token_accuracy": 0.18606623113155366,
      "num_tokens": 48872430.0,
      "step": 23550
    },
    {
      "entropy": 5.4979816436767575,
      "epoch": 2.136701741654572,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004542982344622385,
      "loss": 5.0547,
      "mean_token_accuracy": 0.194625224173069,
      "num_tokens": 48882575.0,
      "step": 23555
    },
    {
      "entropy": 5.4439356327056885,
      "epoch": 2.137155297532656,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045427868492231533,
      "loss": 5.1529,
      "mean_token_accuracy": 0.18669850677251815,
      "num_tokens": 48893337.0,
      "step": 23560
    },
    {
      "entropy": 5.589477634429931,
      "epoch": 2.13760885341074,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00045425913167486333,
      "loss": 5.2981,
      "mean_token_accuracy": 0.182708303630352,
      "num_tokens": 48904675.0,
      "step": 23565
    },
    {
      "entropy": 5.4386005878448485,
      "epoch": 2.1380624092888243,
      "grad_norm": 1.25,
      "learning_rate": 0.00045423957472028677,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20119546353816986,
      "num_tokens": 48914831.0,
      "step": 23570
    },
    {
      "entropy": 5.383177804946899,
      "epoch": 2.1385159651669086,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004542200140589901,
      "loss": 5.065,
      "mean_token_accuracy": 0.19772401154041291,
      "num_tokens": 48924574.0,
      "step": 23575
    },
    {
      "entropy": 5.49116063117981,
      "epoch": 2.138969521044993,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045420044969137787,
      "loss": 5.212,
      "mean_token_accuracy": 0.18901256173849107,
      "num_tokens": 48933921.0,
      "step": 23580
    },
    {
      "entropy": 5.50197401046753,
      "epoch": 2.139423076923077,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004541808816178547,
      "loss": 5.1662,
      "mean_token_accuracy": 0.18785565942525864,
      "num_tokens": 48944571.0,
      "step": 23585
    },
    {
      "entropy": 5.504676580429077,
      "epoch": 2.139876632801161,
      "grad_norm": 2.109375,
      "learning_rate": 0.00045416130983882524,
      "loss": 5.1956,
      "mean_token_accuracy": 0.18762335181236267,
      "num_tokens": 48954569.0,
      "step": 23590
    },
    {
      "entropy": 5.455940294265747,
      "epoch": 2.140330188679245,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045414173435469427,
      "loss": 5.1342,
      "mean_token_accuracy": 0.18446247577667235,
      "num_tokens": 48965886.0,
      "step": 23595
    },
    {
      "entropy": 5.544996690750122,
      "epoch": 2.1407837445573294,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045412215516586654,
      "loss": 5.1638,
      "mean_token_accuracy": 0.18827107846736907,
      "num_tokens": 48976219.0,
      "step": 23600
    },
    {
      "entropy": 5.45710334777832,
      "epoch": 2.1412373004354137,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000454102572272747,
      "loss": 5.1431,
      "mean_token_accuracy": 0.187054742872715,
      "num_tokens": 48986322.0,
      "step": 23605
    },
    {
      "entropy": 5.432085990905762,
      "epoch": 2.141690856313498,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004540829856757406,
      "loss": 5.0783,
      "mean_token_accuracy": 0.19683008193969725,
      "num_tokens": 48996297.0,
      "step": 23610
    },
    {
      "entropy": 5.4643988609313965,
      "epoch": 2.142144412191582,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004540633953752525,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19332904368638992,
      "num_tokens": 49006165.0,
      "step": 23615
    },
    {
      "entropy": 5.494270658493042,
      "epoch": 2.142597968069666,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004540438013716876,
      "loss": 5.097,
      "mean_token_accuracy": 0.18242512941360473,
      "num_tokens": 49017278.0,
      "step": 23620
    },
    {
      "entropy": 5.380304288864136,
      "epoch": 2.1430515239477503,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004540242036654514,
      "loss": 5.1134,
      "mean_token_accuracy": 0.18941081166267396,
      "num_tokens": 49027088.0,
      "step": 23625
    },
    {
      "entropy": 5.5036155700683596,
      "epoch": 2.1435050798258346,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000454004602256949,
      "loss": 5.0652,
      "mean_token_accuracy": 0.1909342497587204,
      "num_tokens": 49038156.0,
      "step": 23630
    },
    {
      "entropy": 5.531635665893555,
      "epoch": 2.143958635703919,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004539849971465858,
      "loss": 5.1665,
      "mean_token_accuracy": 0.18742911219596864,
      "num_tokens": 49047480.0,
      "step": 23635
    },
    {
      "entropy": 5.498327589035034,
      "epoch": 2.1444121915820027,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045396538833476725,
      "loss": 5.1949,
      "mean_token_accuracy": 0.1937381163239479,
      "num_tokens": 49057373.0,
      "step": 23640
    },
    {
      "entropy": 5.481448984146118,
      "epoch": 2.144865747460087,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004539457758218989,
      "loss": 5.1786,
      "mean_token_accuracy": 0.18988621532917022,
      "num_tokens": 49068380.0,
      "step": 23645
    },
    {
      "entropy": 5.443109607696533,
      "epoch": 2.1453193033381712,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004539261596083863,
      "loss": 5.1522,
      "mean_token_accuracy": 0.19043559283018113,
      "num_tokens": 49078646.0,
      "step": 23650
    },
    {
      "entropy": 5.460894346237183,
      "epoch": 2.1457728592162555,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004539065396946351,
      "loss": 5.0806,
      "mean_token_accuracy": 0.19664311408996582,
      "num_tokens": 49089281.0,
      "step": 23655
    },
    {
      "entropy": 5.535195970535279,
      "epoch": 2.1462264150943398,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004538869160810511,
      "loss": 5.1764,
      "mean_token_accuracy": 0.18650639802217484,
      "num_tokens": 49099514.0,
      "step": 23660
    },
    {
      "entropy": 5.522756910324096,
      "epoch": 2.146679970972424,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045386728876804,
      "loss": 5.185,
      "mean_token_accuracy": 0.18645163327455522,
      "num_tokens": 49110247.0,
      "step": 23665
    },
    {
      "entropy": 5.520337009429932,
      "epoch": 2.147133526850508,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045384765775600787,
      "loss": 5.25,
      "mean_token_accuracy": 0.18433319330215453,
      "num_tokens": 49122492.0,
      "step": 23670
    },
    {
      "entropy": 5.49483003616333,
      "epoch": 2.147587082728592,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004538280230453606,
      "loss": 5.142,
      "mean_token_accuracy": 0.19005344957113265,
      "num_tokens": 49132395.0,
      "step": 23675
    },
    {
      "entropy": 5.524321174621582,
      "epoch": 2.1480406386066764,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004538083846365042,
      "loss": 5.1836,
      "mean_token_accuracy": 0.1911781206727028,
      "num_tokens": 49142909.0,
      "step": 23680
    },
    {
      "entropy": 5.465768194198608,
      "epoch": 2.1484941944847606,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004537887425298448,
      "loss": 5.1289,
      "mean_token_accuracy": 0.18385403901338576,
      "num_tokens": 49152933.0,
      "step": 23685
    },
    {
      "entropy": 5.430705308914185,
      "epoch": 2.1489477503628445,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045376909672578875,
      "loss": 5.1256,
      "mean_token_accuracy": 0.1892481490969658,
      "num_tokens": 49163595.0,
      "step": 23690
    },
    {
      "entropy": 5.4096959114074705,
      "epoch": 2.1494013062409287,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004537494472247422,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19391730129718782,
      "num_tokens": 49174664.0,
      "step": 23695
    },
    {
      "entropy": 5.501069259643555,
      "epoch": 2.149854862119013,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004537297940271115,
      "loss": 5.1804,
      "mean_token_accuracy": 0.190680693089962,
      "num_tokens": 49185105.0,
      "step": 23700
    },
    {
      "entropy": 5.382751893997193,
      "epoch": 2.1503084179970973,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004537101371333031,
      "loss": 5.0744,
      "mean_token_accuracy": 0.19528792798519135,
      "num_tokens": 49194679.0,
      "step": 23705
    },
    {
      "entropy": 5.396391201019287,
      "epoch": 2.1507619738751815,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045369047654372356,
      "loss": 5.0621,
      "mean_token_accuracy": 0.19378271996974944,
      "num_tokens": 49204563.0,
      "step": 23710
    },
    {
      "entropy": 5.4904028415679935,
      "epoch": 2.151215529753266,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004536708122587795,
      "loss": 5.1224,
      "mean_token_accuracy": 0.1913180962204933,
      "num_tokens": 49213759.0,
      "step": 23715
    },
    {
      "entropy": 5.413399934768677,
      "epoch": 2.1516690856313496,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004536511442788774,
      "loss": 5.0364,
      "mean_token_accuracy": 0.19912656098604203,
      "num_tokens": 49223604.0,
      "step": 23720
    },
    {
      "entropy": 5.433433866500854,
      "epoch": 2.152122641509434,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004536314726044241,
      "loss": 5.1699,
      "mean_token_accuracy": 0.1935412049293518,
      "num_tokens": 49234000.0,
      "step": 23725
    },
    {
      "entropy": 5.497545957565308,
      "epoch": 2.152576197387518,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045361179723582654,
      "loss": 5.1293,
      "mean_token_accuracy": 0.19192034155130386,
      "num_tokens": 49244874.0,
      "step": 23730
    },
    {
      "entropy": 5.559602403640747,
      "epoch": 2.1530297532656024,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004535921181734914,
      "loss": 5.3143,
      "mean_token_accuracy": 0.18501890003681182,
      "num_tokens": 49255495.0,
      "step": 23735
    },
    {
      "entropy": 5.4241667747497555,
      "epoch": 2.1534833091436867,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004535724354178257,
      "loss": 5.0892,
      "mean_token_accuracy": 0.19287949204444885,
      "num_tokens": 49265357.0,
      "step": 23740
    },
    {
      "entropy": 5.5038289546966555,
      "epoch": 2.1539368650217705,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045355274896923664,
      "loss": 5.2141,
      "mean_token_accuracy": 0.17828187793493272,
      "num_tokens": 49276396.0,
      "step": 23745
    },
    {
      "entropy": 5.462511396408081,
      "epoch": 2.1543904208998548,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004535330588281312,
      "loss": 5.0563,
      "mean_token_accuracy": 0.192010760307312,
      "num_tokens": 49285925.0,
      "step": 23750
    },
    {
      "entropy": 5.393792343139649,
      "epoch": 2.154843976777939,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004535133649949166,
      "loss": 5.1239,
      "mean_token_accuracy": 0.1915045738220215,
      "num_tokens": 49296493.0,
      "step": 23755
    },
    {
      "entropy": 5.412475109100342,
      "epoch": 2.1552975326560233,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004534936674700001,
      "loss": 5.0832,
      "mean_token_accuracy": 0.19377999007701874,
      "num_tokens": 49307273.0,
      "step": 23760
    },
    {
      "entropy": 5.4462419033050535,
      "epoch": 2.1557510885341076,
      "grad_norm": 1.984375,
      "learning_rate": 0.00045347396625378914,
      "loss": 5.0791,
      "mean_token_accuracy": 0.2007341206073761,
      "num_tokens": 49317055.0,
      "step": 23765
    },
    {
      "entropy": 5.456128978729248,
      "epoch": 2.1562046444121914,
      "grad_norm": 1.25,
      "learning_rate": 0.00045345426134669103,
      "loss": 5.1413,
      "mean_token_accuracy": 0.18606617599725722,
      "num_tokens": 49327420.0,
      "step": 23770
    },
    {
      "entropy": 5.459082508087159,
      "epoch": 2.1566582002902757,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00045343455274911325,
      "loss": 5.2281,
      "mean_token_accuracy": 0.18362988233566285,
      "num_tokens": 49338323.0,
      "step": 23775
    },
    {
      "entropy": 5.510842418670654,
      "epoch": 2.15711175616836,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004534148404614635,
      "loss": 5.1843,
      "mean_token_accuracy": 0.18918455094099046,
      "num_tokens": 49348364.0,
      "step": 23780
    },
    {
      "entropy": 5.549263763427734,
      "epoch": 2.157565312046444,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00045339512448414943,
      "loss": 5.201,
      "mean_token_accuracy": 0.18591219186782837,
      "num_tokens": 49358274.0,
      "step": 23785
    },
    {
      "entropy": 5.45104947090149,
      "epoch": 2.1580188679245285,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004533754048175787,
      "loss": 5.0487,
      "mean_token_accuracy": 0.1889199882745743,
      "num_tokens": 49368651.0,
      "step": 23790
    },
    {
      "entropy": 5.4266400814056395,
      "epoch": 2.1584724238026123,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004533556814621591,
      "loss": 5.0721,
      "mean_token_accuracy": 0.18779059499502182,
      "num_tokens": 49378363.0,
      "step": 23795
    },
    {
      "entropy": 5.390038251876831,
      "epoch": 2.1589259796806965,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045333595441829865,
      "loss": 5.1002,
      "mean_token_accuracy": 0.19344583600759507,
      "num_tokens": 49388241.0,
      "step": 23800
    },
    {
      "entropy": 5.463345146179199,
      "epoch": 2.159379535558781,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004533162236864051,
      "loss": 5.1075,
      "mean_token_accuracy": 0.19428814351558685,
      "num_tokens": 49397678.0,
      "step": 23805
    },
    {
      "entropy": 5.505478191375732,
      "epoch": 2.159833091436865,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004532964892668867,
      "loss": 5.2178,
      "mean_token_accuracy": 0.18762192577123643,
      "num_tokens": 49407368.0,
      "step": 23810
    },
    {
      "entropy": 5.446516609191894,
      "epoch": 2.1602866473149493,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045327675116015146,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1908867195248604,
      "num_tokens": 49417926.0,
      "step": 23815
    },
    {
      "entropy": 5.543562221527099,
      "epoch": 2.1607402031930336,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004532570093666075,
      "loss": 5.253,
      "mean_token_accuracy": 0.17862159311771392,
      "num_tokens": 49427899.0,
      "step": 23820
    },
    {
      "entropy": 5.480096387863159,
      "epoch": 2.1611937590711174,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004532372638866633,
      "loss": 5.0945,
      "mean_token_accuracy": 0.1988714024424553,
      "num_tokens": 49439102.0,
      "step": 23825
    },
    {
      "entropy": 5.423920059204102,
      "epoch": 2.1616473149492017,
      "grad_norm": 1.5,
      "learning_rate": 0.00045321751472072693,
      "loss": 5.047,
      "mean_token_accuracy": 0.19765320271253586,
      "num_tokens": 49448013.0,
      "step": 23830
    },
    {
      "entropy": 5.363828134536743,
      "epoch": 2.162100870827286,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045319776186920695,
      "loss": 5.1058,
      "mean_token_accuracy": 0.19787309020757676,
      "num_tokens": 49457620.0,
      "step": 23835
    },
    {
      "entropy": 5.370371389389038,
      "epoch": 2.1625544267053702,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00045317800533251193,
      "loss": 5.1421,
      "mean_token_accuracy": 0.18510669618844985,
      "num_tokens": 49468038.0,
      "step": 23840
    },
    {
      "entropy": 5.413502836227417,
      "epoch": 2.163007982583454,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045315824511105027,
      "loss": 5.0587,
      "mean_token_accuracy": 0.19356884062290192,
      "num_tokens": 49477714.0,
      "step": 23845
    },
    {
      "entropy": 5.494656372070312,
      "epoch": 2.1634615384615383,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00045313848120523073,
      "loss": 5.0968,
      "mean_token_accuracy": 0.19027163237333297,
      "num_tokens": 49488214.0,
      "step": 23850
    },
    {
      "entropy": 5.470267200469971,
      "epoch": 2.1639150943396226,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045311871361546205,
      "loss": 5.2027,
      "mean_token_accuracy": 0.19198982268571854,
      "num_tokens": 49498444.0,
      "step": 23855
    },
    {
      "entropy": 5.380479192733764,
      "epoch": 2.164368650217707,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004530989423421529,
      "loss": 5.0453,
      "mean_token_accuracy": 0.19954425394535064,
      "num_tokens": 49508759.0,
      "step": 23860
    },
    {
      "entropy": 5.328580284118653,
      "epoch": 2.164822206095791,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045307916738571223,
      "loss": 5.0013,
      "mean_token_accuracy": 0.2014836221933365,
      "num_tokens": 49518787.0,
      "step": 23865
    },
    {
      "entropy": 5.39508023262024,
      "epoch": 2.1652757619738754,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004530593887465491,
      "loss": 5.0943,
      "mean_token_accuracy": 0.1997660756111145,
      "num_tokens": 49528130.0,
      "step": 23870
    },
    {
      "entropy": 5.569505739212036,
      "epoch": 2.165729317851959,
      "grad_norm": 1.25,
      "learning_rate": 0.00045303960642507235,
      "loss": 5.2652,
      "mean_token_accuracy": 0.18460616916418077,
      "num_tokens": 49539128.0,
      "step": 23875
    },
    {
      "entropy": 5.390063524246216,
      "epoch": 2.1661828737300435,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045301982042169123,
      "loss": 5.1604,
      "mean_token_accuracy": 0.1925584152340889,
      "num_tokens": 49550571.0,
      "step": 23880
    },
    {
      "entropy": 5.437206983566284,
      "epoch": 2.1666364296081277,
      "grad_norm": 1.59375,
      "learning_rate": 0.00045300003073681485,
      "loss": 5.1443,
      "mean_token_accuracy": 0.19219372123479844,
      "num_tokens": 49561106.0,
      "step": 23885
    },
    {
      "entropy": 5.420179319381714,
      "epoch": 2.167089985486212,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00045298023737085243,
      "loss": 5.0468,
      "mean_token_accuracy": 0.19049613624811174,
      "num_tokens": 49571389.0,
      "step": 23890
    },
    {
      "entropy": 5.418888759613037,
      "epoch": 2.1675435413642963,
      "grad_norm": 1.375,
      "learning_rate": 0.0004529604403242134,
      "loss": 5.1378,
      "mean_token_accuracy": 0.1905090406537056,
      "num_tokens": 49582224.0,
      "step": 23895
    },
    {
      "entropy": 5.412771320343017,
      "epoch": 2.16799709724238,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045294063959730707,
      "loss": 5.1524,
      "mean_token_accuracy": 0.18502840101718904,
      "num_tokens": 49592413.0,
      "step": 23900
    },
    {
      "entropy": 5.519330215454102,
      "epoch": 2.1684506531204644,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045292083519054297,
      "loss": 5.2404,
      "mean_token_accuracy": 0.17348009645938872,
      "num_tokens": 49602796.0,
      "step": 23905
    },
    {
      "entropy": 5.479265117645264,
      "epoch": 2.1689042089985486,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004529010271043307,
      "loss": 5.1106,
      "mean_token_accuracy": 0.1971450239419937,
      "num_tokens": 49613625.0,
      "step": 23910
    },
    {
      "entropy": 5.528969383239746,
      "epoch": 2.169357764876633,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045288121533907977,
      "loss": 5.2352,
      "mean_token_accuracy": 0.18479463458061218,
      "num_tokens": 49624719.0,
      "step": 23915
    },
    {
      "entropy": 5.460185956954956,
      "epoch": 2.169811320754717,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00045286139989520006,
      "loss": 5.1884,
      "mean_token_accuracy": 0.1846339538693428,
      "num_tokens": 49634916.0,
      "step": 23920
    },
    {
      "entropy": 5.440398931503296,
      "epoch": 2.170264876632801,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004528415807731012,
      "loss": 5.0836,
      "mean_token_accuracy": 0.19349978864192963,
      "num_tokens": 49645215.0,
      "step": 23925
    },
    {
      "entropy": 5.442452669143677,
      "epoch": 2.1707184325108853,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00045282175797319317,
      "loss": 5.1476,
      "mean_token_accuracy": 0.19261400401592255,
      "num_tokens": 49655805.0,
      "step": 23930
    },
    {
      "entropy": 5.483649635314942,
      "epoch": 2.1711719883889695,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004528019314958859,
      "loss": 5.1404,
      "mean_token_accuracy": 0.1908984288573265,
      "num_tokens": 49665264.0,
      "step": 23935
    },
    {
      "entropy": 5.523253965377807,
      "epoch": 2.171625544267054,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00045278210134158935,
      "loss": 5.2008,
      "mean_token_accuracy": 0.18546196222305297,
      "num_tokens": 49676152.0,
      "step": 23940
    },
    {
      "entropy": 5.480348777770996,
      "epoch": 2.172079100145138,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004527622675107137,
      "loss": 5.1415,
      "mean_token_accuracy": 0.18608943670988082,
      "num_tokens": 49686762.0,
      "step": 23945
    },
    {
      "entropy": 5.475543642044068,
      "epoch": 2.172532656023222,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045274243000366897,
      "loss": 5.1619,
      "mean_token_accuracy": 0.1941516578197479,
      "num_tokens": 49697504.0,
      "step": 23950
    },
    {
      "entropy": 5.468310165405273,
      "epoch": 2.172986211901306,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004527225888208655,
      "loss": 5.2114,
      "mean_token_accuracy": 0.18153633326292037,
      "num_tokens": 49709197.0,
      "step": 23955
    },
    {
      "entropy": 5.494181966781616,
      "epoch": 2.1734397677793904,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004527027439627137,
      "loss": 5.1624,
      "mean_token_accuracy": 0.1861064165830612,
      "num_tokens": 49719764.0,
      "step": 23960
    },
    {
      "entropy": 5.4839029788970945,
      "epoch": 2.1738933236574747,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045268289542962377,
      "loss": 5.1278,
      "mean_token_accuracy": 0.188662613928318,
      "num_tokens": 49730363.0,
      "step": 23965
    },
    {
      "entropy": 5.47813048362732,
      "epoch": 2.174346879535559,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004526630432220063,
      "loss": 5.1836,
      "mean_token_accuracy": 0.18097906708717346,
      "num_tokens": 49741037.0,
      "step": 23970
    },
    {
      "entropy": 5.4488122940063475,
      "epoch": 2.1748004354136428,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004526431873402719,
      "loss": 5.0545,
      "mean_token_accuracy": 0.1943922683596611,
      "num_tokens": 49750127.0,
      "step": 23975
    },
    {
      "entropy": 5.547024965286255,
      "epoch": 2.175253991291727,
      "grad_norm": 1.4375,
      "learning_rate": 0.000452623327784831,
      "loss": 5.2482,
      "mean_token_accuracy": 0.1798812448978424,
      "num_tokens": 49760298.0,
      "step": 23980
    },
    {
      "entropy": 5.483637857437134,
      "epoch": 2.1757075471698113,
      "grad_norm": 1.5,
      "learning_rate": 0.0004526034645560945,
      "loss": 5.158,
      "mean_token_accuracy": 0.18692175149917603,
      "num_tokens": 49770841.0,
      "step": 23985
    },
    {
      "entropy": 5.423768854141235,
      "epoch": 2.1761611030478956,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045258359765447306,
      "loss": 5.1357,
      "mean_token_accuracy": 0.193781316280365,
      "num_tokens": 49781000.0,
      "step": 23990
    },
    {
      "entropy": 5.4700172424316404,
      "epoch": 2.17661465892598,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004525637270803776,
      "loss": 5.1344,
      "mean_token_accuracy": 0.19144668728113173,
      "num_tokens": 49790886.0,
      "step": 23995
    },
    {
      "entropy": 5.425890684127808,
      "epoch": 2.1770682148040637,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045254385283421896,
      "loss": 5.1278,
      "mean_token_accuracy": 0.19399382025003434,
      "num_tokens": 49801477.0,
      "step": 24000
    },
    {
      "epoch": 2.1770682148040637,
      "eval_entropy": 5.229644330702216,
      "eval_loss": 5.241685390472412,
      "eval_mean_token_accuracy": 0.19425802878293785,
      "eval_num_tokens": 49801477.0,
      "eval_runtime": 26.5345,
      "eval_samples_per_second": 1332.607,
      "eval_steps_per_second": 166.576,
      "step": 24000
    },
    {
      "entropy": 5.444155263900757,
      "epoch": 2.177521770682148,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004525239749164082,
      "loss": 5.1792,
      "mean_token_accuracy": 0.1902041807770729,
      "num_tokens": 49812044.0,
      "step": 24005
    },
    {
      "entropy": 5.401356887817383,
      "epoch": 2.177975326560232,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004525040933273564,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19280415773391724,
      "num_tokens": 49821559.0,
      "step": 24010
    },
    {
      "entropy": 5.519820213317871,
      "epoch": 2.1784288824383164,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045248420806747483,
      "loss": 5.2228,
      "mean_token_accuracy": 0.18425508290529252,
      "num_tokens": 49832937.0,
      "step": 24015
    },
    {
      "entropy": 5.522712230682373,
      "epoch": 2.1788824383164007,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004524643191371745,
      "loss": 5.148,
      "mean_token_accuracy": 0.1910183236002922,
      "num_tokens": 49843176.0,
      "step": 24020
    },
    {
      "entropy": 5.413462829589844,
      "epoch": 2.179335994194485,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045244442653686684,
      "loss": 5.0532,
      "mean_token_accuracy": 0.2020078867673874,
      "num_tokens": 49852522.0,
      "step": 24025
    },
    {
      "entropy": 5.388050365447998,
      "epoch": 2.179789550072569,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045242453026696325,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20043592900037766,
      "num_tokens": 49863143.0,
      "step": 24030
    },
    {
      "entropy": 5.534803342819214,
      "epoch": 2.180243105950653,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004524046303278752,
      "loss": 5.1951,
      "mean_token_accuracy": 0.18813924938440324,
      "num_tokens": 49873465.0,
      "step": 24035
    },
    {
      "entropy": 5.542555093765259,
      "epoch": 2.1806966618287373,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004523847267200141,
      "loss": 5.2452,
      "mean_token_accuracy": 0.18234845548868178,
      "num_tokens": 49883690.0,
      "step": 24040
    },
    {
      "entropy": 5.479370498657227,
      "epoch": 2.1811502177068216,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045236481944379165,
      "loss": 5.1995,
      "mean_token_accuracy": 0.189257849752903,
      "num_tokens": 49894418.0,
      "step": 24045
    },
    {
      "entropy": 5.458526277542115,
      "epoch": 2.1816037735849054,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004523449084996196,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19717839658260344,
      "num_tokens": 49903927.0,
      "step": 24050
    },
    {
      "entropy": 5.472884702682495,
      "epoch": 2.1820573294629897,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004523249938879096,
      "loss": 5.1632,
      "mean_token_accuracy": 0.19136575162410735,
      "num_tokens": 49914203.0,
      "step": 24055
    },
    {
      "entropy": 5.44833984375,
      "epoch": 2.182510885341074,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00045230507560907354,
      "loss": 5.1336,
      "mean_token_accuracy": 0.19491467922925948,
      "num_tokens": 49924320.0,
      "step": 24060
    },
    {
      "entropy": 5.388657808303833,
      "epoch": 2.1829644412191582,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045228515366352334,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19549995809793472,
      "num_tokens": 49934634.0,
      "step": 24065
    },
    {
      "entropy": 5.528589677810669,
      "epoch": 2.1834179970972425,
      "grad_norm": 1.890625,
      "learning_rate": 0.00045226522805167095,
      "loss": 5.1561,
      "mean_token_accuracy": 0.19179681241512297,
      "num_tokens": 49944586.0,
      "step": 24070
    },
    {
      "entropy": 5.480982065200806,
      "epoch": 2.1838715529753268,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004522452987739286,
      "loss": 5.143,
      "mean_token_accuracy": 0.19525347650051117,
      "num_tokens": 49955441.0,
      "step": 24075
    },
    {
      "entropy": 5.516525506973267,
      "epoch": 2.1843251088534106,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00045222536583070823,
      "loss": 5.2628,
      "mean_token_accuracy": 0.1895440936088562,
      "num_tokens": 49966718.0,
      "step": 24080
    },
    {
      "entropy": 5.5220033645629885,
      "epoch": 2.184778664731495,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045220542922242205,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18243651539087297,
      "num_tokens": 49979483.0,
      "step": 24085
    },
    {
      "entropy": 5.466339778900147,
      "epoch": 2.185232220609579,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045218548894948254,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19454994201660156,
      "num_tokens": 49989976.0,
      "step": 24090
    },
    {
      "entropy": 5.408672571182251,
      "epoch": 2.1856857764876634,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045216554501230187,
      "loss": 5.0275,
      "mean_token_accuracy": 0.193217995762825,
      "num_tokens": 50000069.0,
      "step": 24095
    },
    {
      "entropy": 5.45632381439209,
      "epoch": 2.1861393323657476,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004521455974112927,
      "loss": 5.1396,
      "mean_token_accuracy": 0.1914377674460411,
      "num_tokens": 50009970.0,
      "step": 24100
    },
    {
      "entropy": 5.486239385604859,
      "epoch": 2.1865928882438315,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00045212564614686735,
      "loss": 5.133,
      "mean_token_accuracy": 0.19490534216165542,
      "num_tokens": 50019590.0,
      "step": 24105
    },
    {
      "entropy": 5.476805353164673,
      "epoch": 2.1870464441219157,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004521056912194386,
      "loss": 5.142,
      "mean_token_accuracy": 0.18195036351680755,
      "num_tokens": 50030092.0,
      "step": 24110
    },
    {
      "entropy": 5.45863184928894,
      "epoch": 2.1875,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004520857326294189,
      "loss": 5.1611,
      "mean_token_accuracy": 0.18786808252334594,
      "num_tokens": 50040318.0,
      "step": 24115
    },
    {
      "entropy": 5.5085655689239506,
      "epoch": 2.1879535558780843,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00045206577037722114,
      "loss": 5.2681,
      "mean_token_accuracy": 0.17772439867258072,
      "num_tokens": 50050507.0,
      "step": 24120
    },
    {
      "entropy": 5.6405433177947994,
      "epoch": 2.1884071117561685,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004520458044632582,
      "loss": 5.2177,
      "mean_token_accuracy": 0.18510869145393372,
      "num_tokens": 50059727.0,
      "step": 24125
    },
    {
      "entropy": 5.495986080169677,
      "epoch": 2.1888606676342524,
      "grad_norm": 2.71875,
      "learning_rate": 0.00045202583488794285,
      "loss": 5.1572,
      "mean_token_accuracy": 0.18939761966466903,
      "num_tokens": 50070812.0,
      "step": 24130
    },
    {
      "entropy": 5.483536052703857,
      "epoch": 2.1893142235123366,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004520058616516881,
      "loss": 5.2322,
      "mean_token_accuracy": 0.18377099186182022,
      "num_tokens": 50081175.0,
      "step": 24135
    },
    {
      "entropy": 5.476307964324951,
      "epoch": 2.189767779390421,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045198588475490716,
      "loss": 5.1156,
      "mean_token_accuracy": 0.1915664479136467,
      "num_tokens": 50092246.0,
      "step": 24140
    },
    {
      "entropy": 5.496579694747925,
      "epoch": 2.190221335268505,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004519659041980129,
      "loss": 5.1876,
      "mean_token_accuracy": 0.19071128368377685,
      "num_tokens": 50101934.0,
      "step": 24145
    },
    {
      "entropy": 5.456558084487915,
      "epoch": 2.1906748911465894,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004519459199814187,
      "loss": 5.1906,
      "mean_token_accuracy": 0.18769328892230988,
      "num_tokens": 50112159.0,
      "step": 24150
    },
    {
      "entropy": 5.4201079368591305,
      "epoch": 2.1911284470246732,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004519259321055377,
      "loss": 5.1735,
      "mean_token_accuracy": 0.18650781661272048,
      "num_tokens": 50123079.0,
      "step": 24155
    },
    {
      "entropy": 5.4661633491516115,
      "epoch": 2.1915820029027575,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004519059405707834,
      "loss": 5.1482,
      "mean_token_accuracy": 0.18935231417417525,
      "num_tokens": 50133097.0,
      "step": 24160
    },
    {
      "entropy": 5.58074984550476,
      "epoch": 2.1920355587808418,
      "grad_norm": 4.5625,
      "learning_rate": 0.00045188594537756916,
      "loss": 5.2191,
      "mean_token_accuracy": 0.18929015845060349,
      "num_tokens": 50143062.0,
      "step": 24165
    },
    {
      "entropy": 5.567539691925049,
      "epoch": 2.192489114658926,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045186594652630853,
      "loss": 5.1554,
      "mean_token_accuracy": 0.19044677913188934,
      "num_tokens": 50154482.0,
      "step": 24170
    },
    {
      "entropy": 5.5074669361114506,
      "epoch": 2.1929426705370103,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000451845944017415,
      "loss": 5.1249,
      "mean_token_accuracy": 0.19497748017311095,
      "num_tokens": 50164880.0,
      "step": 24175
    },
    {
      "entropy": 5.408302545547485,
      "epoch": 2.1933962264150946,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045182593785130225,
      "loss": 5.0825,
      "mean_token_accuracy": 0.19561868906021118,
      "num_tokens": 50173896.0,
      "step": 24180
    },
    {
      "entropy": 5.467991018295288,
      "epoch": 2.1938497822931784,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004518059280283842,
      "loss": 5.2018,
      "mean_token_accuracy": 0.18498082906007768,
      "num_tokens": 50185078.0,
      "step": 24185
    },
    {
      "entropy": 5.519409656524658,
      "epoch": 2.1943033381712627,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045178591454907434,
      "loss": 5.1802,
      "mean_token_accuracy": 0.1905406266450882,
      "num_tokens": 50195500.0,
      "step": 24190
    },
    {
      "entropy": 5.426968383789062,
      "epoch": 2.194756894049347,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00045176589741378675,
      "loss": 5.0955,
      "mean_token_accuracy": 0.1922988310456276,
      "num_tokens": 50206103.0,
      "step": 24195
    },
    {
      "entropy": 5.431756687164307,
      "epoch": 2.195210449927431,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004517458766229354,
      "loss": 5.0836,
      "mean_token_accuracy": 0.19348585307598115,
      "num_tokens": 50216522.0,
      "step": 24200
    },
    {
      "entropy": 5.4912103652954105,
      "epoch": 2.195664005805515,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045172585217693426,
      "loss": 5.1554,
      "mean_token_accuracy": 0.1980164870619774,
      "num_tokens": 50226721.0,
      "step": 24205
    },
    {
      "entropy": 5.509608554840088,
      "epoch": 2.1961175616835993,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045170582407619747,
      "loss": 5.2265,
      "mean_token_accuracy": 0.18313349336385726,
      "num_tokens": 50236327.0,
      "step": 24210
    },
    {
      "entropy": 5.419965553283691,
      "epoch": 2.1965711175616836,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004516857923211392,
      "loss": 5.086,
      "mean_token_accuracy": 0.19851548820734025,
      "num_tokens": 50245729.0,
      "step": 24215
    },
    {
      "entropy": 5.472430086135864,
      "epoch": 2.197024673439768,
      "grad_norm": 1.9375,
      "learning_rate": 0.00045166575691217375,
      "loss": 5.2428,
      "mean_token_accuracy": 0.18163604587316512,
      "num_tokens": 50257024.0,
      "step": 24220
    },
    {
      "entropy": 5.502130222320557,
      "epoch": 2.197478229317852,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045164571784971536,
      "loss": 5.0869,
      "mean_token_accuracy": 0.19257978051900865,
      "num_tokens": 50266910.0,
      "step": 24225
    },
    {
      "entropy": 5.466563129425049,
      "epoch": 2.1979317851959363,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045162567513417864,
      "loss": 5.1434,
      "mean_token_accuracy": 0.18989861905574798,
      "num_tokens": 50278233.0,
      "step": 24230
    },
    {
      "entropy": 5.426043748855591,
      "epoch": 2.19838534107402,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045160562876597776,
      "loss": 5.1221,
      "mean_token_accuracy": 0.18596071898937225,
      "num_tokens": 50289276.0,
      "step": 24235
    },
    {
      "entropy": 5.49346227645874,
      "epoch": 2.1988388969521044,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004515855787455276,
      "loss": 5.1185,
      "mean_token_accuracy": 0.1889270305633545,
      "num_tokens": 50299336.0,
      "step": 24240
    },
    {
      "entropy": 5.465900182723999,
      "epoch": 2.1992924528301887,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004515655250732426,
      "loss": 5.0929,
      "mean_token_accuracy": 0.1954464852809906,
      "num_tokens": 50308308.0,
      "step": 24245
    },
    {
      "entropy": 5.4557208061218265,
      "epoch": 2.199746008708273,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045154546774953764,
      "loss": 5.1466,
      "mean_token_accuracy": 0.1918361708521843,
      "num_tokens": 50318295.0,
      "step": 24250
    },
    {
      "entropy": 5.449669075012207,
      "epoch": 2.2001995645863572,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004515254067748274,
      "loss": 5.1371,
      "mean_token_accuracy": 0.19506621360778809,
      "num_tokens": 50329296.0,
      "step": 24255
    },
    {
      "entropy": 5.460938596725464,
      "epoch": 2.200653120464441,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004515053421495267,
      "loss": 5.2006,
      "mean_token_accuracy": 0.18895391523838043,
      "num_tokens": 50339389.0,
      "step": 24260
    },
    {
      "entropy": 5.497395610809326,
      "epoch": 2.2011066763425253,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00045148527387405056,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18312782794237137,
      "num_tokens": 50349647.0,
      "step": 24265
    },
    {
      "entropy": 5.541597938537597,
      "epoch": 2.2015602322206096,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004514652019488141,
      "loss": 5.2116,
      "mean_token_accuracy": 0.18577232360839843,
      "num_tokens": 50361292.0,
      "step": 24270
    },
    {
      "entropy": 5.4578571796417235,
      "epoch": 2.202013788098694,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004514451263742321,
      "loss": 5.128,
      "mean_token_accuracy": 0.18848206102848053,
      "num_tokens": 50372293.0,
      "step": 24275
    },
    {
      "entropy": 5.4311384677886965,
      "epoch": 2.202467343976778,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004514250471507201,
      "loss": 5.0863,
      "mean_token_accuracy": 0.19815784841775894,
      "num_tokens": 50382462.0,
      "step": 24280
    },
    {
      "entropy": 5.495049047470093,
      "epoch": 2.202920899854862,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00045140496427869306,
      "loss": 5.1826,
      "mean_token_accuracy": 0.1939759522676468,
      "num_tokens": 50393157.0,
      "step": 24285
    },
    {
      "entropy": 5.343840169906616,
      "epoch": 2.203374455732946,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004513848777585664,
      "loss": 5.0839,
      "mean_token_accuracy": 0.19871456325054168,
      "num_tokens": 50403474.0,
      "step": 24290
    },
    {
      "entropy": 5.478339910507202,
      "epoch": 2.2038280116110305,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004513647875907556,
      "loss": 5.1434,
      "mean_token_accuracy": 0.18997347354888916,
      "num_tokens": 50414337.0,
      "step": 24295
    },
    {
      "entropy": 5.566988611221314,
      "epoch": 2.2042815674891147,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045134469377567597,
      "loss": 5.2666,
      "mean_token_accuracy": 0.179206046462059,
      "num_tokens": 50424863.0,
      "step": 24300
    },
    {
      "entropy": 5.5240490436553955,
      "epoch": 2.204735123367199,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004513245963137432,
      "loss": 5.1883,
      "mean_token_accuracy": 0.1892564445734024,
      "num_tokens": 50435982.0,
      "step": 24305
    },
    {
      "entropy": 5.425048732757569,
      "epoch": 2.205188679245283,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004513044952053728,
      "loss": 5.087,
      "mean_token_accuracy": 0.19000388383865358,
      "num_tokens": 50445112.0,
      "step": 24310
    },
    {
      "entropy": 5.396102237701416,
      "epoch": 2.205642235123367,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045128439045098045,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20149641782045363,
      "num_tokens": 50455266.0,
      "step": 24315
    },
    {
      "entropy": 5.497623538970947,
      "epoch": 2.2060957910014514,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045126428205098207,
      "loss": 5.2228,
      "mean_token_accuracy": 0.1776631385087967,
      "num_tokens": 50466930.0,
      "step": 24320
    },
    {
      "entropy": 5.464960956573487,
      "epoch": 2.2065493468795356,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004512441700057934,
      "loss": 5.1452,
      "mean_token_accuracy": 0.19454966932535173,
      "num_tokens": 50477534.0,
      "step": 24325
    },
    {
      "entropy": 5.4735182762146,
      "epoch": 2.20700290275762,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004512240543158305,
      "loss": 5.1094,
      "mean_token_accuracy": 0.18959185779094695,
      "num_tokens": 50487352.0,
      "step": 24330
    },
    {
      "entropy": 5.427239799499512,
      "epoch": 2.2074564586357037,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045120393498150905,
      "loss": 5.0404,
      "mean_token_accuracy": 0.19953638762235643,
      "num_tokens": 50496227.0,
      "step": 24335
    },
    {
      "entropy": 5.402370595932007,
      "epoch": 2.207910014513788,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045118381200324537,
      "loss": 5.1563,
      "mean_token_accuracy": 0.18833740353584288,
      "num_tokens": 50507536.0,
      "step": 24340
    },
    {
      "entropy": 5.451164579391479,
      "epoch": 2.2083635703918723,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004511636853814556,
      "loss": 5.1565,
      "mean_token_accuracy": 0.19057951420545577,
      "num_tokens": 50517659.0,
      "step": 24345
    },
    {
      "entropy": 5.4926141738891605,
      "epoch": 2.2088171262699565,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045114355511655595,
      "loss": 5.1369,
      "mean_token_accuracy": 0.18686085641384126,
      "num_tokens": 50528717.0,
      "step": 24350
    },
    {
      "entropy": 5.4405303478240965,
      "epoch": 2.209270682148041,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045112342120896267,
      "loss": 5.0512,
      "mean_token_accuracy": 0.192780701816082,
      "num_tokens": 50538449.0,
      "step": 24355
    },
    {
      "entropy": 5.483138704299927,
      "epoch": 2.2097242380261246,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045110328365909215,
      "loss": 5.22,
      "mean_token_accuracy": 0.18563944101333618,
      "num_tokens": 50549098.0,
      "step": 24360
    },
    {
      "entropy": 5.512986898422241,
      "epoch": 2.210177793904209,
      "grad_norm": 1.375,
      "learning_rate": 0.0004510831424673609,
      "loss": 5.1842,
      "mean_token_accuracy": 0.1889504998922348,
      "num_tokens": 50560047.0,
      "step": 24365
    },
    {
      "entropy": 5.5371678352355955,
      "epoch": 2.210631349782293,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004510629976341853,
      "loss": 5.1228,
      "mean_token_accuracy": 0.1842415526509285,
      "num_tokens": 50569704.0,
      "step": 24370
    },
    {
      "entropy": 5.430223894119263,
      "epoch": 2.2110849056603774,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045104284915998215,
      "loss": 5.1626,
      "mean_token_accuracy": 0.19332678765058517,
      "num_tokens": 50580006.0,
      "step": 24375
    },
    {
      "entropy": 5.385150480270386,
      "epoch": 2.2115384615384617,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045102269704516793,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19803228378295898,
      "num_tokens": 50589593.0,
      "step": 24380
    },
    {
      "entropy": 5.4745392322540285,
      "epoch": 2.211992017416546,
      "grad_norm": 1.25,
      "learning_rate": 0.0004510025412901595,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19442939013242722,
      "num_tokens": 50598774.0,
      "step": 24385
    },
    {
      "entropy": 5.322408819198609,
      "epoch": 2.2124455732946298,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045098238189537367,
      "loss": 5.0664,
      "mean_token_accuracy": 0.19591669887304305,
      "num_tokens": 50608815.0,
      "step": 24390
    },
    {
      "entropy": 5.375529384613037,
      "epoch": 2.212899129172714,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045096221886122733,
      "loss": 5.2374,
      "mean_token_accuracy": 0.19473819732666015,
      "num_tokens": 50619255.0,
      "step": 24395
    },
    {
      "entropy": 5.527984380722046,
      "epoch": 2.2133526850507983,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004509420521881375,
      "loss": 5.149,
      "mean_token_accuracy": 0.1931985393166542,
      "num_tokens": 50629483.0,
      "step": 24400
    },
    {
      "entropy": 5.4565435409545895,
      "epoch": 2.2138062409288826,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004509218818765211,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19298033267259598,
      "num_tokens": 50638552.0,
      "step": 24405
    },
    {
      "entropy": 5.514334678649902,
      "epoch": 2.2142597968069664,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00045090170792679544,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19257598072290422,
      "num_tokens": 50648752.0,
      "step": 24410
    },
    {
      "entropy": 5.446681308746338,
      "epoch": 2.2147133526850507,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045088153033937765,
      "loss": 5.1396,
      "mean_token_accuracy": 0.19519342482089996,
      "num_tokens": 50658373.0,
      "step": 24415
    },
    {
      "entropy": 5.370953369140625,
      "epoch": 2.215166908563135,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00045086134911468496,
      "loss": 5.1688,
      "mean_token_accuracy": 0.1906610459089279,
      "num_tokens": 50668529.0,
      "step": 24420
    },
    {
      "entropy": 5.465918779373169,
      "epoch": 2.215620464441219,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004508411642531347,
      "loss": 5.0888,
      "mean_token_accuracy": 0.19472404569387436,
      "num_tokens": 50678038.0,
      "step": 24425
    },
    {
      "entropy": 5.461847162246704,
      "epoch": 2.2160740203193035,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004508209757551445,
      "loss": 5.1556,
      "mean_token_accuracy": 0.19134274423122405,
      "num_tokens": 50687788.0,
      "step": 24430
    },
    {
      "entropy": 5.460209321975708,
      "epoch": 2.2165275761973877,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045080078362113153,
      "loss": 5.1369,
      "mean_token_accuracy": 0.18793174177408217,
      "num_tokens": 50697748.0,
      "step": 24435
    },
    {
      "entropy": 5.374427843093872,
      "epoch": 2.2169811320754715,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045078058785151363,
      "loss": 5.0604,
      "mean_token_accuracy": 0.1983875513076782,
      "num_tokens": 50707776.0,
      "step": 24440
    },
    {
      "entropy": 5.475234794616699,
      "epoch": 2.217434687953556,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004507603884467085,
      "loss": 5.1918,
      "mean_token_accuracy": 0.18520963191986084,
      "num_tokens": 50717547.0,
      "step": 24445
    },
    {
      "entropy": 5.473999500274658,
      "epoch": 2.21788824383164,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004507401854071336,
      "loss": 5.1006,
      "mean_token_accuracy": 0.1948298215866089,
      "num_tokens": 50728446.0,
      "step": 24450
    },
    {
      "entropy": 5.4087354183197025,
      "epoch": 2.2183417997097243,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045071997873320687,
      "loss": 5.0593,
      "mean_token_accuracy": 0.1917655274271965,
      "num_tokens": 50737909.0,
      "step": 24455
    },
    {
      "entropy": 5.374684619903564,
      "epoch": 2.2187953555878086,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045069976842534635,
      "loss": 5.1564,
      "mean_token_accuracy": 0.19300876557826996,
      "num_tokens": 50749745.0,
      "step": 24460
    },
    {
      "entropy": 5.499877786636352,
      "epoch": 2.2192489114658924,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004506795544839697,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19205449819564818,
      "num_tokens": 50760073.0,
      "step": 24465
    },
    {
      "entropy": 5.513488245010376,
      "epoch": 2.2197024673439767,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004506593369094951,
      "loss": 5.0936,
      "mean_token_accuracy": 0.19292904138565065,
      "num_tokens": 50769261.0,
      "step": 24470
    },
    {
      "entropy": 5.5142271518707275,
      "epoch": 2.220156023222061,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045063911570234074,
      "loss": 5.2214,
      "mean_token_accuracy": 0.17965349555015564,
      "num_tokens": 50779572.0,
      "step": 24475
    },
    {
      "entropy": 5.441055107116699,
      "epoch": 2.2206095791001452,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004506188908629247,
      "loss": 5.0716,
      "mean_token_accuracy": 0.19284136444330216,
      "num_tokens": 50789893.0,
      "step": 24480
    },
    {
      "entropy": 5.366277647018433,
      "epoch": 2.2210631349782295,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004505986623916652,
      "loss": 5.0258,
      "mean_token_accuracy": 0.19234238117933272,
      "num_tokens": 50799177.0,
      "step": 24485
    },
    {
      "entropy": 5.432749080657959,
      "epoch": 2.2215166908563133,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004505784302889806,
      "loss": 5.1217,
      "mean_token_accuracy": 0.1969163179397583,
      "num_tokens": 50810093.0,
      "step": 24490
    },
    {
      "entropy": 5.5434122562408445,
      "epoch": 2.2219702467343976,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045055819455528935,
      "loss": 5.2719,
      "mean_token_accuracy": 0.1868878871202469,
      "num_tokens": 50820272.0,
      "step": 24495
    },
    {
      "entropy": 5.501110982894898,
      "epoch": 2.222423802612482,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004505379551910099,
      "loss": 5.2115,
      "mean_token_accuracy": 0.19033110290765762,
      "num_tokens": 50830301.0,
      "step": 24500
    },
    {
      "entropy": 5.511803674697876,
      "epoch": 2.222877358490566,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004505177121965608,
      "loss": 5.1799,
      "mean_token_accuracy": 0.19044588208198548,
      "num_tokens": 50839937.0,
      "step": 24505
    },
    {
      "entropy": 5.434956884384155,
      "epoch": 2.2233309143686504,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045049746557236067,
      "loss": 5.117,
      "mean_token_accuracy": 0.1928895816206932,
      "num_tokens": 50850295.0,
      "step": 24510
    },
    {
      "entropy": 5.495173025131225,
      "epoch": 2.223784470246734,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045047721531882827,
      "loss": 5.1631,
      "mean_token_accuracy": 0.18797722309827805,
      "num_tokens": 50861889.0,
      "step": 24515
    },
    {
      "entropy": 5.499255084991455,
      "epoch": 2.2242380261248185,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004504569614363823,
      "loss": 5.1252,
      "mean_token_accuracy": 0.18922070562839508,
      "num_tokens": 50872011.0,
      "step": 24520
    },
    {
      "entropy": 5.546734189987182,
      "epoch": 2.2246915820029027,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004504367039254416,
      "loss": 5.178,
      "mean_token_accuracy": 0.18667405992746353,
      "num_tokens": 50882559.0,
      "step": 24525
    },
    {
      "entropy": 5.392483139038086,
      "epoch": 2.225145137880987,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045041644278642523,
      "loss": 5.2225,
      "mean_token_accuracy": 0.18459372967481613,
      "num_tokens": 50894132.0,
      "step": 24530
    },
    {
      "entropy": 5.5087847232818605,
      "epoch": 2.2255986937590713,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045039617801975214,
      "loss": 5.1566,
      "mean_token_accuracy": 0.19386180937290193,
      "num_tokens": 50904337.0,
      "step": 24535
    },
    {
      "entropy": 5.555236530303955,
      "epoch": 2.2260522496371555,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045037590962584143,
      "loss": 5.1921,
      "mean_token_accuracy": 0.1832629770040512,
      "num_tokens": 50914961.0,
      "step": 24540
    },
    {
      "entropy": 5.498615503311157,
      "epoch": 2.2265058055152394,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004503556376051121,
      "loss": 5.1358,
      "mean_token_accuracy": 0.18782947212457657,
      "num_tokens": 50925682.0,
      "step": 24545
    },
    {
      "entropy": 5.43533148765564,
      "epoch": 2.2269593613933236,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045033536195798355,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20296464413404464,
      "num_tokens": 50935809.0,
      "step": 24550
    },
    {
      "entropy": 5.551101398468018,
      "epoch": 2.227412917271408,
      "grad_norm": 1.515625,
      "learning_rate": 0.000450315082684875,
      "loss": 5.2636,
      "mean_token_accuracy": 0.1834515377879143,
      "num_tokens": 50946098.0,
      "step": 24555
    },
    {
      "entropy": 5.542461204528808,
      "epoch": 2.227866473149492,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004502947997862059,
      "loss": 5.2021,
      "mean_token_accuracy": 0.18616910725831987,
      "num_tokens": 50956240.0,
      "step": 24560
    },
    {
      "entropy": 5.479972505569458,
      "epoch": 2.228320029027576,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004502745132623957,
      "loss": 5.1121,
      "mean_token_accuracy": 0.19794540256261825,
      "num_tokens": 50966870.0,
      "step": 24565
    },
    {
      "entropy": 5.456949853897095,
      "epoch": 2.2287735849056602,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004502542231138639,
      "loss": 5.1208,
      "mean_token_accuracy": 0.18681297153234483,
      "num_tokens": 50977528.0,
      "step": 24570
    },
    {
      "entropy": 5.490008878707886,
      "epoch": 2.2292271407837445,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045023392934103005,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1922762468457222,
      "num_tokens": 50988565.0,
      "step": 24575
    },
    {
      "entropy": 5.451269865036011,
      "epoch": 2.2296806966618288,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045021363194431397,
      "loss": 5.2016,
      "mean_token_accuracy": 0.19042182117700576,
      "num_tokens": 50998410.0,
      "step": 24580
    },
    {
      "entropy": 5.507575941085816,
      "epoch": 2.230134252539913,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004501933309241352,
      "loss": 5.1724,
      "mean_token_accuracy": 0.1864386260509491,
      "num_tokens": 51008890.0,
      "step": 24585
    },
    {
      "entropy": 5.481625699996949,
      "epoch": 2.2305878084179973,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004501730262809138,
      "loss": 5.1178,
      "mean_token_accuracy": 0.18682246059179305,
      "num_tokens": 51020264.0,
      "step": 24590
    },
    {
      "entropy": 5.477860927581787,
      "epoch": 2.231041364296081,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004501527180150696,
      "loss": 5.042,
      "mean_token_accuracy": 0.20203485041856767,
      "num_tokens": 51029892.0,
      "step": 24595
    },
    {
      "entropy": 5.55087308883667,
      "epoch": 2.2314949201741654,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004501324061270224,
      "loss": 5.2614,
      "mean_token_accuracy": 0.18485147804021834,
      "num_tokens": 51040499.0,
      "step": 24600
    },
    {
      "entropy": 5.402206802368164,
      "epoch": 2.2319484760522497,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004501120906171926,
      "loss": 5.0529,
      "mean_token_accuracy": 0.19466286301612853,
      "num_tokens": 51050395.0,
      "step": 24605
    },
    {
      "entropy": 5.436091184616089,
      "epoch": 2.232402031930334,
      "grad_norm": 1.25,
      "learning_rate": 0.00045009177148600004,
      "loss": 5.0665,
      "mean_token_accuracy": 0.19699629545211791,
      "num_tokens": 51060316.0,
      "step": 24610
    },
    {
      "entropy": 5.389241886138916,
      "epoch": 2.232855587808418,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045007144873386514,
      "loss": 5.1052,
      "mean_token_accuracy": 0.19449352324008942,
      "num_tokens": 51070032.0,
      "step": 24615
    },
    {
      "entropy": 5.582000637054444,
      "epoch": 2.233309143686502,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045005112236120795,
      "loss": 5.2554,
      "mean_token_accuracy": 0.18259712308645248,
      "num_tokens": 51080738.0,
      "step": 24620
    },
    {
      "entropy": 5.51475944519043,
      "epoch": 2.2337626995645863,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000450030792368449,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1982099309563637,
      "num_tokens": 51090287.0,
      "step": 24625
    },
    {
      "entropy": 5.415401649475098,
      "epoch": 2.2342162554426706,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004500104587560086,
      "loss": 5.1321,
      "mean_token_accuracy": 0.1916848123073578,
      "num_tokens": 51101188.0,
      "step": 24630
    },
    {
      "entropy": 5.385489654541016,
      "epoch": 2.234669811320755,
      "grad_norm": 1.375,
      "learning_rate": 0.0004499901215243074,
      "loss": 5.0376,
      "mean_token_accuracy": 0.1953576147556305,
      "num_tokens": 51112310.0,
      "step": 24635
    },
    {
      "entropy": 5.510780572891235,
      "epoch": 2.235123367198839,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044996978067376575,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18807010352611542,
      "num_tokens": 51122481.0,
      "step": 24640
    },
    {
      "entropy": 5.4312060356140135,
      "epoch": 2.235576923076923,
      "grad_norm": 1.546875,
      "learning_rate": 0.00044994943620480453,
      "loss": 5.1055,
      "mean_token_accuracy": 0.19047617465257644,
      "num_tokens": 51131918.0,
      "step": 24645
    },
    {
      "entropy": 5.4547926902771,
      "epoch": 2.236030478955007,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044992908811784446,
      "loss": 5.1911,
      "mean_token_accuracy": 0.18242136985063553,
      "num_tokens": 51142861.0,
      "step": 24650
    },
    {
      "entropy": 5.5100555419921875,
      "epoch": 2.2364840348330914,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044990873641330615,
      "loss": 5.1706,
      "mean_token_accuracy": 0.19059670269489287,
      "num_tokens": 51153956.0,
      "step": 24655
    },
    {
      "entropy": 5.498290872573852,
      "epoch": 2.2369375907111757,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004498883810916107,
      "loss": 5.0568,
      "mean_token_accuracy": 0.1932290628552437,
      "num_tokens": 51164499.0,
      "step": 24660
    },
    {
      "entropy": 5.49201078414917,
      "epoch": 2.23739114658926,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004498680221531789,
      "loss": 5.1986,
      "mean_token_accuracy": 0.18876997232437134,
      "num_tokens": 51175722.0,
      "step": 24665
    },
    {
      "entropy": 5.466195774078369,
      "epoch": 2.237844702467344,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004498476595984319,
      "loss": 5.1229,
      "mean_token_accuracy": 0.19188192188739778,
      "num_tokens": 51184597.0,
      "step": 24670
    },
    {
      "entropy": 5.418113565444946,
      "epoch": 2.238298258345428,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004498272934277906,
      "loss": 5.1734,
      "mean_token_accuracy": 0.1913633167743683,
      "num_tokens": 51194751.0,
      "step": 24675
    },
    {
      "entropy": 5.499386787414551,
      "epoch": 2.2387518142235123,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044980692364167646,
      "loss": 5.0891,
      "mean_token_accuracy": 0.1986499920487404,
      "num_tokens": 51204184.0,
      "step": 24680
    },
    {
      "entropy": 5.456738471984863,
      "epoch": 2.2392053701015966,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004497865502405105,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19446427524089813,
      "num_tokens": 51213761.0,
      "step": 24685
    },
    {
      "entropy": 5.467603158950806,
      "epoch": 2.239658925979681,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044976617322471424,
      "loss": 5.0963,
      "mean_token_accuracy": 0.1944295272231102,
      "num_tokens": 51224150.0,
      "step": 24690
    },
    {
      "entropy": 5.552486753463745,
      "epoch": 2.2401124818577647,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000449745792594709,
      "loss": 5.236,
      "mean_token_accuracy": 0.18574569821357728,
      "num_tokens": 51235133.0,
      "step": 24695
    },
    {
      "entropy": 5.541725206375122,
      "epoch": 2.240566037735849,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044972540835091617,
      "loss": 5.1445,
      "mean_token_accuracy": 0.18979519009590148,
      "num_tokens": 51245632.0,
      "step": 24700
    },
    {
      "entropy": 5.456617021560669,
      "epoch": 2.241019593613933,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00044970502049375736,
      "loss": 5.1804,
      "mean_token_accuracy": 0.19018363058567048,
      "num_tokens": 51255747.0,
      "step": 24705
    },
    {
      "entropy": 5.435675001144409,
      "epoch": 2.2414731494920175,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004496846290236542,
      "loss": 5.12,
      "mean_token_accuracy": 0.1964313268661499,
      "num_tokens": 51265531.0,
      "step": 24710
    },
    {
      "entropy": 5.453099107742309,
      "epoch": 2.2419267053701017,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044966423394102845,
      "loss": 5.159,
      "mean_token_accuracy": 0.19531475007534027,
      "num_tokens": 51275319.0,
      "step": 24715
    },
    {
      "entropy": 5.461896753311157,
      "epoch": 2.2423802612481856,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044964383524630176,
      "loss": 5.1473,
      "mean_token_accuracy": 0.19841596633195877,
      "num_tokens": 51285655.0,
      "step": 24720
    },
    {
      "entropy": 5.370176219940186,
      "epoch": 2.24283381712627,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004496234329398961,
      "loss": 5.0635,
      "mean_token_accuracy": 0.18823614120483398,
      "num_tokens": 51295938.0,
      "step": 24725
    },
    {
      "entropy": 5.492547607421875,
      "epoch": 2.243287373004354,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004496030270222333,
      "loss": 5.0916,
      "mean_token_accuracy": 0.1946649506688118,
      "num_tokens": 51306483.0,
      "step": 24730
    },
    {
      "entropy": 5.471787166595459,
      "epoch": 2.2437409288824384,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044958261749373547,
      "loss": 5.1279,
      "mean_token_accuracy": 0.19319992065429686,
      "num_tokens": 51316446.0,
      "step": 24735
    },
    {
      "entropy": 5.3465500831604,
      "epoch": 2.2441944847605226,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00044956220435482465,
      "loss": 5.157,
      "mean_token_accuracy": 0.1929304078221321,
      "num_tokens": 51326840.0,
      "step": 24740
    },
    {
      "entropy": 5.478730916976929,
      "epoch": 2.244648040638607,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044954178760592283,
      "loss": 5.2019,
      "mean_token_accuracy": 0.18205519318580626,
      "num_tokens": 51337573.0,
      "step": 24745
    },
    {
      "entropy": 5.563036346435547,
      "epoch": 2.2451015965166907,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044952136724745254,
      "loss": 5.1628,
      "mean_token_accuracy": 0.18987581729888917,
      "num_tokens": 51347827.0,
      "step": 24750
    },
    {
      "entropy": 5.475693321228027,
      "epoch": 2.245555152394775,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004495009432798358,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1926501587033272,
      "num_tokens": 51358396.0,
      "step": 24755
    },
    {
      "entropy": 5.500465250015258,
      "epoch": 2.2460087082728593,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00044948051570349503,
      "loss": 5.1454,
      "mean_token_accuracy": 0.18267306238412856,
      "num_tokens": 51368882.0,
      "step": 24760
    },
    {
      "entropy": 5.487477540969849,
      "epoch": 2.2464622641509435,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004494600845188528,
      "loss": 5.1858,
      "mean_token_accuracy": 0.18938415199518205,
      "num_tokens": 51378717.0,
      "step": 24765
    },
    {
      "entropy": 5.51063756942749,
      "epoch": 2.2469158200290273,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044943964972633154,
      "loss": 5.1459,
      "mean_token_accuracy": 0.1887650966644287,
      "num_tokens": 51389270.0,
      "step": 24770
    },
    {
      "entropy": 5.451701593399048,
      "epoch": 2.2473693759071116,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004494192113263539,
      "loss": 5.1494,
      "mean_token_accuracy": 0.19050061851739883,
      "num_tokens": 51400025.0,
      "step": 24775
    },
    {
      "entropy": 5.417097663879394,
      "epoch": 2.247822931785196,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044939876931934255,
      "loss": 5.1384,
      "mean_token_accuracy": 0.19511894583702089,
      "num_tokens": 51410903.0,
      "step": 24780
    },
    {
      "entropy": 5.406564998626709,
      "epoch": 2.24827648766328,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004493783237057201,
      "loss": 5.0833,
      "mean_token_accuracy": 0.19614691138267518,
      "num_tokens": 51420246.0,
      "step": 24785
    },
    {
      "entropy": 5.531072378158569,
      "epoch": 2.2487300435413644,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004493578744859095,
      "loss": 5.2455,
      "mean_token_accuracy": 0.18949256092309952,
      "num_tokens": 51431309.0,
      "step": 24790
    },
    {
      "entropy": 5.533622980117798,
      "epoch": 2.2491835994194487,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004493374216603338,
      "loss": 5.2033,
      "mean_token_accuracy": 0.1898050755262375,
      "num_tokens": 51443142.0,
      "step": 24795
    },
    {
      "entropy": 5.5814002513885494,
      "epoch": 2.2496371552975325,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004493169652294156,
      "loss": 5.3726,
      "mean_token_accuracy": 0.1769214615225792,
      "num_tokens": 51454499.0,
      "step": 24800
    },
    {
      "entropy": 5.515808534622193,
      "epoch": 2.2500907111756168,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044929650519357823,
      "loss": 5.1544,
      "mean_token_accuracy": 0.19280312359333038,
      "num_tokens": 51464667.0,
      "step": 24805
    },
    {
      "entropy": 5.474452066421509,
      "epoch": 2.250544267053701,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004492760415532447,
      "loss": 5.1624,
      "mean_token_accuracy": 0.19315065443515778,
      "num_tokens": 51473627.0,
      "step": 24810
    },
    {
      "entropy": 5.355597639083863,
      "epoch": 2.2509978229317853,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004492555743088382,
      "loss": 5.0074,
      "mean_token_accuracy": 0.19308802336454392,
      "num_tokens": 51483052.0,
      "step": 24815
    },
    {
      "entropy": 5.417157697677612,
      "epoch": 2.2514513788098696,
      "grad_norm": 1.375,
      "learning_rate": 0.000449235103460782,
      "loss": 5.132,
      "mean_token_accuracy": 0.18932186663150788,
      "num_tokens": 51492422.0,
      "step": 24820
    },
    {
      "entropy": 5.453801536560059,
      "epoch": 2.2519049346879534,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004492146290094995,
      "loss": 5.092,
      "mean_token_accuracy": 0.19248677790164948,
      "num_tokens": 51503566.0,
      "step": 24825
    },
    {
      "entropy": 5.471642971038818,
      "epoch": 2.2523584905660377,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00044919415095541406,
      "loss": 5.0895,
      "mean_token_accuracy": 0.19247982352972032,
      "num_tokens": 51513699.0,
      "step": 24830
    },
    {
      "entropy": 5.463069772720337,
      "epoch": 2.252812046444122,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044917366929894914,
      "loss": 5.1354,
      "mean_token_accuracy": 0.18790675103664398,
      "num_tokens": 51524745.0,
      "step": 24835
    },
    {
      "entropy": 5.430230140686035,
      "epoch": 2.253265602322206,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044915318404052846,
      "loss": 5.0751,
      "mean_token_accuracy": 0.19084617793560027,
      "num_tokens": 51536516.0,
      "step": 24840
    },
    {
      "entropy": 5.439774513244629,
      "epoch": 2.2537191582002905,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044913269518057547,
      "loss": 5.1801,
      "mean_token_accuracy": 0.18542472422122955,
      "num_tokens": 51547063.0,
      "step": 24845
    },
    {
      "entropy": 5.455276298522949,
      "epoch": 2.2541727140783743,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000449112202719514,
      "loss": 5.1375,
      "mean_token_accuracy": 0.18838786482810974,
      "num_tokens": 51557244.0,
      "step": 24850
    },
    {
      "entropy": 5.511529350280762,
      "epoch": 2.2546262699564585,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004490917066577678,
      "loss": 5.1708,
      "mean_token_accuracy": 0.19215657711029052,
      "num_tokens": 51567630.0,
      "step": 24855
    },
    {
      "entropy": 5.421895170211792,
      "epoch": 2.255079825834543,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044907120699576075,
      "loss": 5.1746,
      "mean_token_accuracy": 0.19485049396753312,
      "num_tokens": 51577951.0,
      "step": 24860
    },
    {
      "entropy": 5.489801216125488,
      "epoch": 2.255533381712627,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044905070373391684,
      "loss": 5.1047,
      "mean_token_accuracy": 0.18912877440452575,
      "num_tokens": 51587533.0,
      "step": 24865
    },
    {
      "entropy": 5.526775884628296,
      "epoch": 2.2559869375907113,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004490301968726599,
      "loss": 5.1048,
      "mean_token_accuracy": 0.1997961089015007,
      "num_tokens": 51597212.0,
      "step": 24870
    },
    {
      "entropy": 5.437281942367553,
      "epoch": 2.256440493468795,
      "grad_norm": 1.46875,
      "learning_rate": 0.00044900968641241426,
      "loss": 5.1678,
      "mean_token_accuracy": 0.19154802709817886,
      "num_tokens": 51608520.0,
      "step": 24875
    },
    {
      "entropy": 5.462450838088989,
      "epoch": 2.2568940493468794,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004489891723536039,
      "loss": 5.2121,
      "mean_token_accuracy": 0.18629295974969864,
      "num_tokens": 51619651.0,
      "step": 24880
    },
    {
      "entropy": 5.4257289409637455,
      "epoch": 2.2573476052249637,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004489686546966531,
      "loss": 5.041,
      "mean_token_accuracy": 0.19435139298439025,
      "num_tokens": 51630184.0,
      "step": 24885
    },
    {
      "entropy": 5.39904899597168,
      "epoch": 2.257801161103048,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044894813344198626,
      "loss": 5.0416,
      "mean_token_accuracy": 0.20058251321315765,
      "num_tokens": 51641293.0,
      "step": 24890
    },
    {
      "entropy": 5.444141864776611,
      "epoch": 2.2582547169811322,
      "grad_norm": 1.375,
      "learning_rate": 0.0004489276085900276,
      "loss": 5.128,
      "mean_token_accuracy": 0.19071456044912338,
      "num_tokens": 51652002.0,
      "step": 24895
    },
    {
      "entropy": 5.357690668106079,
      "epoch": 2.2587082728592165,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044890708014120174,
      "loss": 4.9928,
      "mean_token_accuracy": 0.20423184037208558,
      "num_tokens": 51661668.0,
      "step": 24900
    },
    {
      "entropy": 5.5418647766113285,
      "epoch": 2.2591618287373003,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004488865480959331,
      "loss": 5.194,
      "mean_token_accuracy": 0.18604316860437392,
      "num_tokens": 51672879.0,
      "step": 24905
    },
    {
      "entropy": 5.47110595703125,
      "epoch": 2.2596153846153846,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004488660124546464,
      "loss": 5.1354,
      "mean_token_accuracy": 0.1894693836569786,
      "num_tokens": 51683878.0,
      "step": 24910
    },
    {
      "entropy": 5.41967453956604,
      "epoch": 2.260068940493469,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004488454732177662,
      "loss": 5.1407,
      "mean_token_accuracy": 0.19257080554962158,
      "num_tokens": 51693971.0,
      "step": 24915
    },
    {
      "entropy": 5.459742069244385,
      "epoch": 2.260522496371553,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044882493038571737,
      "loss": 5.048,
      "mean_token_accuracy": 0.1983783170580864,
      "num_tokens": 51703685.0,
      "step": 24920
    },
    {
      "entropy": 5.4330925941467285,
      "epoch": 2.260976052249637,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044880438395892457,
      "loss": 5.1647,
      "mean_token_accuracy": 0.18871901631355287,
      "num_tokens": 51714469.0,
      "step": 24925
    },
    {
      "entropy": 5.474190521240234,
      "epoch": 2.261429608127721,
      "grad_norm": 2.25,
      "learning_rate": 0.0004487838339378129,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18198044449090958,
      "num_tokens": 51725283.0,
      "step": 24930
    },
    {
      "entropy": 5.475726652145386,
      "epoch": 2.2618831640058055,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004487632803228072,
      "loss": 5.1108,
      "mean_token_accuracy": 0.19528569132089615,
      "num_tokens": 51735628.0,
      "step": 24935
    },
    {
      "entropy": 5.365825223922729,
      "epoch": 2.2623367198838897,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044874272311433265,
      "loss": 4.9987,
      "mean_token_accuracy": 0.20071550607681274,
      "num_tokens": 51745991.0,
      "step": 24940
    },
    {
      "entropy": 5.385110998153687,
      "epoch": 2.262790275761974,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004487221623128143,
      "loss": 5.0514,
      "mean_token_accuracy": 0.19837261885404586,
      "num_tokens": 51756306.0,
      "step": 24945
    },
    {
      "entropy": 5.4251621723175045,
      "epoch": 2.2632438316400583,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044870159791867743,
      "loss": 5.1581,
      "mean_token_accuracy": 0.19399850964546203,
      "num_tokens": 51765466.0,
      "step": 24950
    },
    {
      "entropy": 5.45842924118042,
      "epoch": 2.263697387518142,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044868102993234716,
      "loss": 5.1419,
      "mean_token_accuracy": 0.1932751014828682,
      "num_tokens": 51775526.0,
      "step": 24955
    },
    {
      "entropy": 5.464655494689941,
      "epoch": 2.2641509433962264,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044866045835424896,
      "loss": 5.1178,
      "mean_token_accuracy": 0.18620866537094116,
      "num_tokens": 51785434.0,
      "step": 24960
    },
    {
      "entropy": 5.453844928741455,
      "epoch": 2.2646044992743106,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004486398831848082,
      "loss": 5.1762,
      "mean_token_accuracy": 0.1820254296064377,
      "num_tokens": 51796047.0,
      "step": 24965
    },
    {
      "entropy": 5.452254295349121,
      "epoch": 2.265058055152395,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00044861930442445037,
      "loss": 5.1407,
      "mean_token_accuracy": 0.18732326477766037,
      "num_tokens": 51806173.0,
      "step": 24970
    },
    {
      "entropy": 5.552246809005737,
      "epoch": 2.2655116110304787,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004485987220736012,
      "loss": 5.3165,
      "mean_token_accuracy": 0.18157914727926255,
      "num_tokens": 51817644.0,
      "step": 24975
    },
    {
      "entropy": 5.508435440063477,
      "epoch": 2.265965166908563,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004485781361326861,
      "loss": 5.2378,
      "mean_token_accuracy": 0.18558070957660674,
      "num_tokens": 51829746.0,
      "step": 24980
    },
    {
      "entropy": 5.487930202484131,
      "epoch": 2.2664187227866472,
      "grad_norm": 1.203125,
      "learning_rate": 0.000448557546602131,
      "loss": 5.1838,
      "mean_token_accuracy": 0.18341616690158843,
      "num_tokens": 51840868.0,
      "step": 24985
    },
    {
      "entropy": 5.522884035110474,
      "epoch": 2.2668722786647315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044853695348236156,
      "loss": 5.0563,
      "mean_token_accuracy": 0.1956033319234848,
      "num_tokens": 51851385.0,
      "step": 24990
    },
    {
      "entropy": 5.427377986907959,
      "epoch": 2.2673258345428158,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044851635677380374,
      "loss": 5.1003,
      "mean_token_accuracy": 0.18714677840471267,
      "num_tokens": 51861021.0,
      "step": 24995
    },
    {
      "entropy": 5.433924579620362,
      "epoch": 2.2677793904209,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004484957564768834,
      "loss": 5.2019,
      "mean_token_accuracy": 0.18545889407396315,
      "num_tokens": 51870688.0,
      "step": 25000
    },
    {
      "entropy": 5.413428401947021,
      "epoch": 2.268232946298984,
      "grad_norm": 1.25,
      "learning_rate": 0.00044847515259202657,
      "loss": 5.1397,
      "mean_token_accuracy": 0.19354881197214127,
      "num_tokens": 51880439.0,
      "step": 25005
    },
    {
      "entropy": 5.441373348236084,
      "epoch": 2.268686502177068,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044845454511965947,
      "loss": 5.1001,
      "mean_token_accuracy": 0.19453007280826567,
      "num_tokens": 51890861.0,
      "step": 25010
    },
    {
      "entropy": 5.529043292999267,
      "epoch": 2.2691400580551524,
      "grad_norm": 1.46875,
      "learning_rate": 0.00044843393406020823,
      "loss": 5.1063,
      "mean_token_accuracy": 0.1950567990541458,
      "num_tokens": 51900687.0,
      "step": 25015
    },
    {
      "entropy": 5.471779251098633,
      "epoch": 2.2695936139332367,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044841331941409884,
      "loss": 5.2851,
      "mean_token_accuracy": 0.1790032908320427,
      "num_tokens": 51911805.0,
      "step": 25020
    },
    {
      "entropy": 5.440558910369873,
      "epoch": 2.270047169811321,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044839270118175793,
      "loss": 5.0738,
      "mean_token_accuracy": 0.20187791138887407,
      "num_tokens": 51922520.0,
      "step": 25025
    },
    {
      "entropy": 5.497613620758057,
      "epoch": 2.2705007256894048,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004483720793636118,
      "loss": 5.2147,
      "mean_token_accuracy": 0.18295258432626724,
      "num_tokens": 51934056.0,
      "step": 25030
    },
    {
      "entropy": 5.492384147644043,
      "epoch": 2.270954281567489,
      "grad_norm": 1.53125,
      "learning_rate": 0.00044835145396008685,
      "loss": 5.203,
      "mean_token_accuracy": 0.18846085667610168,
      "num_tokens": 51944696.0,
      "step": 25035
    },
    {
      "entropy": 5.512798595428467,
      "epoch": 2.2714078374455733,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004483308249716097,
      "loss": 5.1803,
      "mean_token_accuracy": 0.18951540142297746,
      "num_tokens": 51955111.0,
      "step": 25040
    },
    {
      "entropy": 5.4325175285339355,
      "epoch": 2.2718613933236576,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004483101923986068,
      "loss": 5.1265,
      "mean_token_accuracy": 0.19310408532619477,
      "num_tokens": 51966036.0,
      "step": 25045
    },
    {
      "entropy": 5.38664140701294,
      "epoch": 2.272314949201742,
      "grad_norm": 1.71875,
      "learning_rate": 0.00044828955624150505,
      "loss": 5.1418,
      "mean_token_accuracy": 0.1888946309685707,
      "num_tokens": 51976060.0,
      "step": 25050
    },
    {
      "entropy": 5.426778554916382,
      "epoch": 2.2727685050798256,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004482689165007311,
      "loss": 5.1125,
      "mean_token_accuracy": 0.19152620434761047,
      "num_tokens": 51986016.0,
      "step": 25055
    },
    {
      "entropy": 5.4252441883087155,
      "epoch": 2.27322206095791,
      "grad_norm": 1.484375,
      "learning_rate": 0.00044824827317671177,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20576469004154205,
      "num_tokens": 51994794.0,
      "step": 25060
    },
    {
      "entropy": 5.386856412887573,
      "epoch": 2.273675616835994,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000448227626269874,
      "loss": 5.0492,
      "mean_token_accuracy": 0.19599704295396805,
      "num_tokens": 52004976.0,
      "step": 25065
    },
    {
      "entropy": 5.466947841644287,
      "epoch": 2.2741291727140784,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044820697578064485,
      "loss": 5.1603,
      "mean_token_accuracy": 0.18359825760126114,
      "num_tokens": 52016380.0,
      "step": 25070
    },
    {
      "entropy": 5.461364841461181,
      "epoch": 2.2745827285921627,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004481863217094513,
      "loss": 5.1024,
      "mean_token_accuracy": 0.19454454630613327,
      "num_tokens": 52026439.0,
      "step": 25075
    },
    {
      "entropy": 5.411532974243164,
      "epoch": 2.2750362844702465,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004481656640567204,
      "loss": 5.0883,
      "mean_token_accuracy": 0.1927181139588356,
      "num_tokens": 52037111.0,
      "step": 25080
    },
    {
      "entropy": 5.446225643157959,
      "epoch": 2.275489840348331,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004481450028228795,
      "loss": 5.1311,
      "mean_token_accuracy": 0.1914452075958252,
      "num_tokens": 52047463.0,
      "step": 25085
    },
    {
      "entropy": 5.491949653625488,
      "epoch": 2.275943396226415,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004481243380083558,
      "loss": 5.1602,
      "mean_token_accuracy": 0.19231636375188826,
      "num_tokens": 52058081.0,
      "step": 25090
    },
    {
      "entropy": 5.4350251197814945,
      "epoch": 2.2763969521044993,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004481036696135767,
      "loss": 5.1609,
      "mean_token_accuracy": 0.19299952387809755,
      "num_tokens": 52069367.0,
      "step": 25095
    },
    {
      "entropy": 5.476330232620239,
      "epoch": 2.2768505079825836,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004480829976389695,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19288607239723204,
      "num_tokens": 52080023.0,
      "step": 25100
    },
    {
      "entropy": 5.380625772476196,
      "epoch": 2.277304063860668,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044806232208496194,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1926248699426651,
      "num_tokens": 52089184.0,
      "step": 25105
    },
    {
      "entropy": 5.518385934829712,
      "epoch": 2.2777576197387517,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004480416429519814,
      "loss": 5.2068,
      "mean_token_accuracy": 0.18830369412899017,
      "num_tokens": 52098711.0,
      "step": 25110
    },
    {
      "entropy": 5.502094745635986,
      "epoch": 2.278211175616836,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004480209602404556,
      "loss": 5.1035,
      "mean_token_accuracy": 0.19798117727041245,
      "num_tokens": 52108826.0,
      "step": 25115
    },
    {
      "entropy": 5.38461709022522,
      "epoch": 2.27866473149492,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004480002739508122,
      "loss": 5.0849,
      "mean_token_accuracy": 0.19654273092746735,
      "num_tokens": 52118934.0,
      "step": 25120
    },
    {
      "entropy": 5.37301287651062,
      "epoch": 2.2791182873730045,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004479795840834791,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19918226152658464,
      "num_tokens": 52129803.0,
      "step": 25125
    },
    {
      "entropy": 5.369647645950318,
      "epoch": 2.2795718432510883,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044795889063888413,
      "loss": 5.0038,
      "mean_token_accuracy": 0.19680616408586502,
      "num_tokens": 52140070.0,
      "step": 25130
    },
    {
      "entropy": 5.582769918441772,
      "epoch": 2.2800253991291726,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004479381936174552,
      "loss": 5.3233,
      "mean_token_accuracy": 0.18247663378715515,
      "num_tokens": 52150462.0,
      "step": 25135
    },
    {
      "entropy": 5.524487543106079,
      "epoch": 2.280478955007257,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004479174930196204,
      "loss": 5.1585,
      "mean_token_accuracy": 0.19400387704372407,
      "num_tokens": 52160552.0,
      "step": 25140
    },
    {
      "entropy": 5.353589105606079,
      "epoch": 2.280932510885341,
      "grad_norm": 1.25,
      "learning_rate": 0.00044789678884580767,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20069977939128875,
      "num_tokens": 52169780.0,
      "step": 25145
    },
    {
      "entropy": 5.46290225982666,
      "epoch": 2.2813860667634254,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00044787608109644536,
      "loss": 5.148,
      "mean_token_accuracy": 0.1983001261949539,
      "num_tokens": 52180055.0,
      "step": 25150
    },
    {
      "entropy": 5.462240886688233,
      "epoch": 2.2818396226415096,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004478553697719616,
      "loss": 5.0951,
      "mean_token_accuracy": 0.1976122200489044,
      "num_tokens": 52189772.0,
      "step": 25155
    },
    {
      "entropy": 5.455110931396485,
      "epoch": 2.2822931785195935,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004478346548727848,
      "loss": 5.0812,
      "mean_token_accuracy": 0.2011331796646118,
      "num_tokens": 52199608.0,
      "step": 25160
    },
    {
      "entropy": 5.468139219284057,
      "epoch": 2.2827467343976777,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004478139363993433,
      "loss": 5.1295,
      "mean_token_accuracy": 0.18921409398317338,
      "num_tokens": 52208236.0,
      "step": 25165
    },
    {
      "entropy": 5.475308847427368,
      "epoch": 2.283200290275762,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004477932143520655,
      "loss": 5.1528,
      "mean_token_accuracy": 0.19372028410434722,
      "num_tokens": 52218403.0,
      "step": 25170
    },
    {
      "entropy": 5.496800804138184,
      "epoch": 2.2836538461538463,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044777248873138005,
      "loss": 5.1173,
      "mean_token_accuracy": 0.19576524645090104,
      "num_tokens": 52228255.0,
      "step": 25175
    },
    {
      "entropy": 5.5027515411376955,
      "epoch": 2.2841074020319305,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00044775175953771547,
      "loss": 5.1538,
      "mean_token_accuracy": 0.18925679326057435,
      "num_tokens": 52237732.0,
      "step": 25180
    },
    {
      "entropy": 5.494544315338135,
      "epoch": 2.2845609579100143,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004477310267715005,
      "loss": 5.2934,
      "mean_token_accuracy": 0.1798497512936592,
      "num_tokens": 52249226.0,
      "step": 25185
    },
    {
      "entropy": 5.564633464813232,
      "epoch": 2.2850145137880986,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004477102904331639,
      "loss": 5.1183,
      "mean_token_accuracy": 0.1956036865711212,
      "num_tokens": 52259550.0,
      "step": 25190
    },
    {
      "entropy": 5.510807180404663,
      "epoch": 2.285468069666183,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044768955052313443,
      "loss": 5.1548,
      "mean_token_accuracy": 0.18989001214504242,
      "num_tokens": 52269670.0,
      "step": 25195
    },
    {
      "entropy": 5.443511247634888,
      "epoch": 2.285921625544267,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044766880704184104,
      "loss": 5.1368,
      "mean_token_accuracy": 0.19318526089191437,
      "num_tokens": 52280135.0,
      "step": 25200
    },
    {
      "entropy": 5.400499439239502,
      "epoch": 2.2863751814223514,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004476480599897128,
      "loss": 5.0688,
      "mean_token_accuracy": 0.19580385237932205,
      "num_tokens": 52291806.0,
      "step": 25205
    },
    {
      "entropy": 5.37821774482727,
      "epoch": 2.2868287373004352,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004476273093671786,
      "loss": 5.0878,
      "mean_token_accuracy": 0.19349646121263503,
      "num_tokens": 52302118.0,
      "step": 25210
    },
    {
      "entropy": 5.525963544845581,
      "epoch": 2.2872822931785195,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044760655517466766,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18764487504959107,
      "num_tokens": 52312762.0,
      "step": 25215
    },
    {
      "entropy": 5.531395483016968,
      "epoch": 2.2877358490566038,
      "grad_norm": 1.4375,
      "learning_rate": 0.00044758579741260916,
      "loss": 5.1638,
      "mean_token_accuracy": 0.18612560480833054,
      "num_tokens": 52323367.0,
      "step": 25220
    },
    {
      "entropy": 5.500016832351685,
      "epoch": 2.288189404934688,
      "grad_norm": 1.84375,
      "learning_rate": 0.00044756503608143247,
      "loss": 5.2088,
      "mean_token_accuracy": 0.18317812979221343,
      "num_tokens": 52335231.0,
      "step": 25225
    },
    {
      "entropy": 5.483082056045532,
      "epoch": 2.2886429608127723,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044754427118156675,
      "loss": 5.0967,
      "mean_token_accuracy": 0.19350934624671937,
      "num_tokens": 52345066.0,
      "step": 25230
    },
    {
      "entropy": 5.439260625839234,
      "epoch": 2.289096516690856,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004475235027134416,
      "loss": 5.1253,
      "mean_token_accuracy": 0.1990129455924034,
      "num_tokens": 52356383.0,
      "step": 25235
    },
    {
      "entropy": 5.464977025985718,
      "epoch": 2.2895500725689404,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004475027306774865,
      "loss": 5.2314,
      "mean_token_accuracy": 0.19565231651067733,
      "num_tokens": 52366013.0,
      "step": 25240
    },
    {
      "entropy": 5.525763320922851,
      "epoch": 2.2900036284470247,
      "grad_norm": 1.59375,
      "learning_rate": 0.00044748195507413095,
      "loss": 5.1659,
      "mean_token_accuracy": 0.19332284480333328,
      "num_tokens": 52376298.0,
      "step": 25245
    },
    {
      "entropy": 5.41515531539917,
      "epoch": 2.290457184325109,
      "grad_norm": 1.375,
      "learning_rate": 0.00044746117590380456,
      "loss": 5.0545,
      "mean_token_accuracy": 0.1933938294649124,
      "num_tokens": 52387134.0,
      "step": 25250
    },
    {
      "entropy": 5.426428127288818,
      "epoch": 2.290910740203193,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004474403931669371,
      "loss": 5.0392,
      "mean_token_accuracy": 0.19825795143842698,
      "num_tokens": 52397389.0,
      "step": 25255
    },
    {
      "entropy": 5.430628728866577,
      "epoch": 2.2913642960812775,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004474196068639584,
      "loss": 5.0844,
      "mean_token_accuracy": 0.1967586949467659,
      "num_tokens": 52407456.0,
      "step": 25260
    },
    {
      "entropy": 5.422469854354858,
      "epoch": 2.2918178519593613,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004473988169952983,
      "loss": 5.137,
      "mean_token_accuracy": 0.19848017245531083,
      "num_tokens": 52417967.0,
      "step": 25265
    },
    {
      "entropy": 5.5211633205413815,
      "epoch": 2.2922714078374455,
      "grad_norm": 2.484375,
      "learning_rate": 0.00044737802356138674,
      "loss": 5.2332,
      "mean_token_accuracy": 0.19205388873815538,
      "num_tokens": 52428757.0,
      "step": 25270
    },
    {
      "entropy": 5.500093269348144,
      "epoch": 2.29272496371553,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004473572265626537,
      "loss": 5.1336,
      "mean_token_accuracy": 0.19593432992696763,
      "num_tokens": 52438973.0,
      "step": 25275
    },
    {
      "entropy": 5.408484697341919,
      "epoch": 2.293178519593614,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004473364259995293,
      "loss": 5.0649,
      "mean_token_accuracy": 0.19317463785409927,
      "num_tokens": 52449323.0,
      "step": 25280
    },
    {
      "entropy": 5.44531512260437,
      "epoch": 2.293632075471698,
      "grad_norm": 1.25,
      "learning_rate": 0.0004473156218724436,
      "loss": 5.1159,
      "mean_token_accuracy": 0.19877298027276993,
      "num_tokens": 52459299.0,
      "step": 25285
    },
    {
      "entropy": 5.409148073196411,
      "epoch": 2.294085631349782,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044729481418182704,
      "loss": 5.0866,
      "mean_token_accuracy": 0.18947478532791137,
      "num_tokens": 52470026.0,
      "step": 25290
    },
    {
      "entropy": 5.488951349258423,
      "epoch": 2.2945391872278664,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044727400292810985,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19429329186677932,
      "num_tokens": 52479975.0,
      "step": 25295
    },
    {
      "entropy": 5.478442907333374,
      "epoch": 2.2949927431059507,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044725318811172227,
      "loss": 5.2257,
      "mean_token_accuracy": 0.1849939689040184,
      "num_tokens": 52491067.0,
      "step": 25300
    },
    {
      "entropy": 5.353318357467652,
      "epoch": 2.295446298984035,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000447232369733095,
      "loss": 5.0934,
      "mean_token_accuracy": 0.19986770451068878,
      "num_tokens": 52500671.0,
      "step": 25305
    },
    {
      "entropy": 5.424852275848389,
      "epoch": 2.2958998548621192,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004472115477926583,
      "loss": 5.0671,
      "mean_token_accuracy": 0.19497156888246536,
      "num_tokens": 52509910.0,
      "step": 25310
    },
    {
      "entropy": 5.476299381256103,
      "epoch": 2.296353410740203,
      "grad_norm": 1.4375,
      "learning_rate": 0.00044719072229084295,
      "loss": 5.1003,
      "mean_token_accuracy": 0.19486893266439437,
      "num_tokens": 52520517.0,
      "step": 25315
    },
    {
      "entropy": 5.433670425415039,
      "epoch": 2.2968069666182873,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004471698932280796,
      "loss": 5.195,
      "mean_token_accuracy": 0.19461120516061783,
      "num_tokens": 52530760.0,
      "step": 25320
    },
    {
      "entropy": 5.507651948928833,
      "epoch": 2.2972605224963716,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000447149060604799,
      "loss": 5.204,
      "mean_token_accuracy": 0.18476898223161697,
      "num_tokens": 52541791.0,
      "step": 25325
    },
    {
      "entropy": 5.57869701385498,
      "epoch": 2.297714078374456,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044712822442143194,
      "loss": 5.2053,
      "mean_token_accuracy": 0.18511044830083848,
      "num_tokens": 52551885.0,
      "step": 25330
    },
    {
      "entropy": 5.418063068389893,
      "epoch": 2.2981676342525397,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044710738467840927,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19595830589532853,
      "num_tokens": 52563080.0,
      "step": 25335
    },
    {
      "entropy": 5.407884407043457,
      "epoch": 2.298621190130624,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004470865413761621,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2023771211504936,
      "num_tokens": 52573564.0,
      "step": 25340
    },
    {
      "entropy": 5.4415772438049315,
      "epoch": 2.299074746008708,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044706569451512143,
      "loss": 5.1022,
      "mean_token_accuracy": 0.19369253665208816,
      "num_tokens": 52583801.0,
      "step": 25345
    },
    {
      "entropy": 5.512453556060791,
      "epoch": 2.2995283018867925,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044704484409571825,
      "loss": 5.1485,
      "mean_token_accuracy": 0.1873415008187294,
      "num_tokens": 52594908.0,
      "step": 25350
    },
    {
      "entropy": 5.409165143966675,
      "epoch": 2.2999818577648767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044702399011838393,
      "loss": 5.0624,
      "mean_token_accuracy": 0.1870497852563858,
      "num_tokens": 52604861.0,
      "step": 25355
    },
    {
      "entropy": 5.383542108535766,
      "epoch": 2.300435413642961,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004470031325835497,
      "loss": 5.1146,
      "mean_token_accuracy": 0.1936957821249962,
      "num_tokens": 52616203.0,
      "step": 25360
    },
    {
      "entropy": 5.409993457794189,
      "epoch": 2.300888969521045,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004469822714916468,
      "loss": 5.055,
      "mean_token_accuracy": 0.1960047572851181,
      "num_tokens": 52625838.0,
      "step": 25365
    },
    {
      "entropy": 5.376798295974732,
      "epoch": 2.301342525399129,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004469614068431067,
      "loss": 5.0879,
      "mean_token_accuracy": 0.19759478121995927,
      "num_tokens": 52635741.0,
      "step": 25370
    },
    {
      "entropy": 5.381258964538574,
      "epoch": 2.3017960812772134,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004469405386383608,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19404916167259217,
      "num_tokens": 52645948.0,
      "step": 25375
    },
    {
      "entropy": 5.442795658111573,
      "epoch": 2.3022496371552976,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044691966687784087,
      "loss": 5.1117,
      "mean_token_accuracy": 0.19442020803689958,
      "num_tokens": 52655073.0,
      "step": 25380
    },
    {
      "entropy": 5.41845064163208,
      "epoch": 2.302703193033382,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044689879156197833,
      "loss": 5.0955,
      "mean_token_accuracy": 0.20042895078659057,
      "num_tokens": 52664429.0,
      "step": 25385
    },
    {
      "entropy": 5.554607391357422,
      "epoch": 2.3031567489114657,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044687791269120504,
      "loss": 5.1979,
      "mean_token_accuracy": 0.19066931903362275,
      "num_tokens": 52674890.0,
      "step": 25390
    },
    {
      "entropy": 5.498943662643432,
      "epoch": 2.30361030478955,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044685703026595266,
      "loss": 5.2063,
      "mean_token_accuracy": 0.1921961635351181,
      "num_tokens": 52685673.0,
      "step": 25395
    },
    {
      "entropy": 5.556607294082641,
      "epoch": 2.3040638606676342,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00044683614428665314,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18344938457012178,
      "num_tokens": 52696696.0,
      "step": 25400
    },
    {
      "entropy": 5.502080392837525,
      "epoch": 2.3045174165457185,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00044681525475373843,
      "loss": 5.1643,
      "mean_token_accuracy": 0.19055817276239395,
      "num_tokens": 52707662.0,
      "step": 25405
    },
    {
      "entropy": 5.4467936038970945,
      "epoch": 2.3049709724238028,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004467943616676403,
      "loss": 5.1117,
      "mean_token_accuracy": 0.19793922305107117,
      "num_tokens": 52718438.0,
      "step": 25410
    },
    {
      "entropy": 5.4210127830505375,
      "epoch": 2.3054245283018866,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004467734650287911,
      "loss": 5.0895,
      "mean_token_accuracy": 0.1952739730477333,
      "num_tokens": 52729267.0,
      "step": 25415
    },
    {
      "entropy": 5.36895260810852,
      "epoch": 2.305878084179971,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004467525648376229,
      "loss": 5.0476,
      "mean_token_accuracy": 0.20604359209537507,
      "num_tokens": 52738754.0,
      "step": 25420
    },
    {
      "entropy": 5.409491491317749,
      "epoch": 2.306331640058055,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004467316610945679,
      "loss": 5.1142,
      "mean_token_accuracy": 0.19882371425628662,
      "num_tokens": 52749086.0,
      "step": 25425
    },
    {
      "entropy": 5.523556280136108,
      "epoch": 2.3067851959361394,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044671075380005825,
      "loss": 5.204,
      "mean_token_accuracy": 0.18855140656232833,
      "num_tokens": 52758961.0,
      "step": 25430
    },
    {
      "entropy": 5.510347318649292,
      "epoch": 2.3072387518142237,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004466898429545266,
      "loss": 5.1541,
      "mean_token_accuracy": 0.19264985322952272,
      "num_tokens": 52769447.0,
      "step": 25435
    },
    {
      "entropy": 5.470531702041626,
      "epoch": 2.3076923076923075,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044666892855840513,
      "loss": 5.216,
      "mean_token_accuracy": 0.18563998341560364,
      "num_tokens": 52780306.0,
      "step": 25440
    },
    {
      "entropy": 5.434840250015259,
      "epoch": 2.3081458635703918,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004466480106121266,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19298110902309418,
      "num_tokens": 52791263.0,
      "step": 25445
    },
    {
      "entropy": 5.497778749465942,
      "epoch": 2.308599419448476,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004466270891161234,
      "loss": 5.1305,
      "mean_token_accuracy": 0.18794185519218445,
      "num_tokens": 52802194.0,
      "step": 25450
    },
    {
      "entropy": 5.455404996871948,
      "epoch": 2.3090529753265603,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044660616407082827,
      "loss": 5.1141,
      "mean_token_accuracy": 0.20189228802919387,
      "num_tokens": 52812391.0,
      "step": 25455
    },
    {
      "entropy": 5.4229851245880125,
      "epoch": 2.3095065312046446,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044658523547667385,
      "loss": 5.1041,
      "mean_token_accuracy": 0.18781272321939468,
      "num_tokens": 52823540.0,
      "step": 25460
    },
    {
      "entropy": 5.365628623962403,
      "epoch": 2.309960087082729,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004465643033340931,
      "loss": 5.0455,
      "mean_token_accuracy": 0.20173140615224838,
      "num_tokens": 52833238.0,
      "step": 25465
    },
    {
      "entropy": 5.47570161819458,
      "epoch": 2.3104136429608126,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044654336764351886,
      "loss": 5.1983,
      "mean_token_accuracy": 0.1868945598602295,
      "num_tokens": 52844510.0,
      "step": 25470
    },
    {
      "entropy": 5.467204523086548,
      "epoch": 2.310867198838897,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044652242840538406,
      "loss": 5.1666,
      "mean_token_accuracy": 0.198001466691494,
      "num_tokens": 52856311.0,
      "step": 25475
    },
    {
      "entropy": 5.348452091217041,
      "epoch": 2.311320754716981,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004465014856201218,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19415535032749176,
      "num_tokens": 52866295.0,
      "step": 25480
    },
    {
      "entropy": 5.4596820831298825,
      "epoch": 2.3117743105950654,
      "grad_norm": 1.40625,
      "learning_rate": 0.000446480539288165,
      "loss": 5.1439,
      "mean_token_accuracy": 0.19614551663398744,
      "num_tokens": 52876818.0,
      "step": 25485
    },
    {
      "entropy": 5.45586051940918,
      "epoch": 2.3122278664731493,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004464595894099469,
      "loss": 5.089,
      "mean_token_accuracy": 0.19112872928380967,
      "num_tokens": 52886173.0,
      "step": 25490
    },
    {
      "entropy": 5.388339710235596,
      "epoch": 2.3126814223512335,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044643863598590094,
      "loss": 5.0688,
      "mean_token_accuracy": 0.19009080976247789,
      "num_tokens": 52895875.0,
      "step": 25495
    },
    {
      "entropy": 5.452623128890991,
      "epoch": 2.313134978229318,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044641767901646025,
      "loss": 5.1957,
      "mean_token_accuracy": 0.18912117630243303,
      "num_tokens": 52906465.0,
      "step": 25500
    },
    {
      "entropy": 5.51167631149292,
      "epoch": 2.313588534107402,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004463967185020583,
      "loss": 5.1019,
      "mean_token_accuracy": 0.19650138020515442,
      "num_tokens": 52916378.0,
      "step": 25505
    },
    {
      "entropy": 5.457577323913574,
      "epoch": 2.3140420899854863,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004463757544431285,
      "loss": 5.188,
      "mean_token_accuracy": 0.19375678896903992,
      "num_tokens": 52926698.0,
      "step": 25510
    },
    {
      "entropy": 5.464661121368408,
      "epoch": 2.3144956458635706,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004463547868401044,
      "loss": 5.1171,
      "mean_token_accuracy": 0.19122205078601837,
      "num_tokens": 52936469.0,
      "step": 25515
    },
    {
      "entropy": 5.453711986541748,
      "epoch": 2.3149492017416544,
      "grad_norm": 1.375,
      "learning_rate": 0.00044633381569341964,
      "loss": 5.1309,
      "mean_token_accuracy": 0.1920606091618538,
      "num_tokens": 52947006.0,
      "step": 25520
    },
    {
      "entropy": 5.361803913116455,
      "epoch": 2.3154027576197387,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004463128410035079,
      "loss": 5.0138,
      "mean_token_accuracy": 0.19698380380868913,
      "num_tokens": 52956663.0,
      "step": 25525
    },
    {
      "entropy": 5.415988206863403,
      "epoch": 2.315856313497823,
      "grad_norm": 1.609375,
      "learning_rate": 0.00044629186277080307,
      "loss": 5.045,
      "mean_token_accuracy": 0.19653597176074983,
      "num_tokens": 52967383.0,
      "step": 25530
    },
    {
      "entropy": 5.3944323539733885,
      "epoch": 2.316309869375907,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044627088099573884,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20411770790815353,
      "num_tokens": 52976567.0,
      "step": 25535
    },
    {
      "entropy": 5.387534856796265,
      "epoch": 2.3167634252539915,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044624989567874907,
      "loss": 5.1312,
      "mean_token_accuracy": 0.19269508123397827,
      "num_tokens": 52986159.0,
      "step": 25540
    },
    {
      "entropy": 5.459084415435791,
      "epoch": 2.3172169811320753,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044622890682026787,
      "loss": 5.1488,
      "mean_token_accuracy": 0.18669390976428984,
      "num_tokens": 52996028.0,
      "step": 25545
    },
    {
      "entropy": 5.5031208992004395,
      "epoch": 2.3176705370101596,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004462079144207292,
      "loss": 5.1263,
      "mean_token_accuracy": 0.1904450386762619,
      "num_tokens": 53006446.0,
      "step": 25550
    },
    {
      "entropy": 5.483714056015015,
      "epoch": 2.318124092888244,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044618691848056726,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19108579009771348,
      "num_tokens": 53016497.0,
      "step": 25555
    },
    {
      "entropy": 5.443817186355591,
      "epoch": 2.318577648766328,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004461659190002162,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19942113012075424,
      "num_tokens": 53027811.0,
      "step": 25560
    },
    {
      "entropy": 5.437132740020752,
      "epoch": 2.3190312046444124,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044614491598011027,
      "loss": 5.1153,
      "mean_token_accuracy": 0.18713532835245134,
      "num_tokens": 53038206.0,
      "step": 25565
    },
    {
      "entropy": 5.498736333847046,
      "epoch": 2.319484760522496,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004461239094206839,
      "loss": 5.1602,
      "mean_token_accuracy": 0.1945645436644554,
      "num_tokens": 53047974.0,
      "step": 25570
    },
    {
      "entropy": 5.575474500656128,
      "epoch": 2.3199383164005805,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044610289932237154,
      "loss": 5.1651,
      "mean_token_accuracy": 0.19981667995452881,
      "num_tokens": 53057904.0,
      "step": 25575
    },
    {
      "entropy": 5.430214691162109,
      "epoch": 2.3203918722786647,
      "grad_norm": 1.4375,
      "learning_rate": 0.00044608188568560756,
      "loss": 5.1074,
      "mean_token_accuracy": 0.19397639930248262,
      "num_tokens": 53068447.0,
      "step": 25580
    },
    {
      "entropy": 5.4285989761352536,
      "epoch": 2.320845428156749,
      "grad_norm": 2.125,
      "learning_rate": 0.0004460608685108266,
      "loss": 5.1348,
      "mean_token_accuracy": 0.18913317918777467,
      "num_tokens": 53079369.0,
      "step": 25585
    },
    {
      "entropy": 5.5516095638275145,
      "epoch": 2.3212989840348333,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044603984779846334,
      "loss": 5.2159,
      "mean_token_accuracy": 0.18623747527599335,
      "num_tokens": 53090794.0,
      "step": 25590
    },
    {
      "entropy": 5.480582284927368,
      "epoch": 2.321752539912917,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004460188235489524,
      "loss": 5.0398,
      "mean_token_accuracy": 0.19461479038000107,
      "num_tokens": 53100802.0,
      "step": 25595
    },
    {
      "entropy": 5.362231779098511,
      "epoch": 2.3222060957910013,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004459977957627285,
      "loss": 5.1625,
      "mean_token_accuracy": 0.18649092316627502,
      "num_tokens": 53111950.0,
      "step": 25600
    },
    {
      "entropy": 5.3397032737731935,
      "epoch": 2.3226596516690856,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044597676444022683,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2010385736823082,
      "num_tokens": 53121908.0,
      "step": 25605
    },
    {
      "entropy": 5.488620328903198,
      "epoch": 2.32311320754717,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044595572958188194,
      "loss": 5.1134,
      "mean_token_accuracy": 0.1952907234430313,
      "num_tokens": 53133065.0,
      "step": 25610
    },
    {
      "entropy": 5.393708992004394,
      "epoch": 2.323566763425254,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004459346911881291,
      "loss": 5.07,
      "mean_token_accuracy": 0.19479803293943404,
      "num_tokens": 53144135.0,
      "step": 25615
    },
    {
      "entropy": 5.421140909194946,
      "epoch": 2.3240203193033384,
      "grad_norm": 1.125,
      "learning_rate": 0.0004459136492594032,
      "loss": 5.0527,
      "mean_token_accuracy": 0.20202367901802062,
      "num_tokens": 53153916.0,
      "step": 25620
    },
    {
      "entropy": 5.356860494613647,
      "epoch": 2.3244738751814222,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004458926037961394,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20291237086057662,
      "num_tokens": 53163937.0,
      "step": 25625
    },
    {
      "entropy": 5.49337100982666,
      "epoch": 2.3249274310595065,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004458715547987732,
      "loss": 5.1894,
      "mean_token_accuracy": 0.1851067066192627,
      "num_tokens": 53174327.0,
      "step": 25630
    },
    {
      "entropy": 5.497378587722778,
      "epoch": 2.3253809869375908,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004458505022677396,
      "loss": 5.1311,
      "mean_token_accuracy": 0.1879854530096054,
      "num_tokens": 53185946.0,
      "step": 25635
    },
    {
      "entropy": 5.469008731842041,
      "epoch": 2.325834542815675,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004458294462034741,
      "loss": 5.0798,
      "mean_token_accuracy": 0.19626264572143554,
      "num_tokens": 53195944.0,
      "step": 25640
    },
    {
      "entropy": 5.410752248764038,
      "epoch": 2.326288098693759,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004458083866064121,
      "loss": 5.1343,
      "mean_token_accuracy": 0.19780405312776567,
      "num_tokens": 53206054.0,
      "step": 25645
    },
    {
      "entropy": 5.452144575119019,
      "epoch": 2.326741654571843,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004457873234769891,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19352302253246306,
      "num_tokens": 53216923.0,
      "step": 25650
    },
    {
      "entropy": 5.418481826782227,
      "epoch": 2.3271952104499274,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044576625681564086,
      "loss": 5.1075,
      "mean_token_accuracy": 0.19262817800045012,
      "num_tokens": 53226788.0,
      "step": 25655
    },
    {
      "entropy": 5.468872547149658,
      "epoch": 2.3276487663280117,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044574518662280276,
      "loss": 5.1496,
      "mean_token_accuracy": 0.18912499845027925,
      "num_tokens": 53237573.0,
      "step": 25660
    },
    {
      "entropy": 5.511098909378052,
      "epoch": 2.328102322206096,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004457241128989108,
      "loss": 5.1563,
      "mean_token_accuracy": 0.18842088133096696,
      "num_tokens": 53247724.0,
      "step": 25665
    },
    {
      "entropy": 5.527591943740845,
      "epoch": 2.32855587808418,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004457030356444006,
      "loss": 5.0989,
      "mean_token_accuracy": 0.19520705938339233,
      "num_tokens": 53258774.0,
      "step": 25670
    },
    {
      "entropy": 5.438469791412354,
      "epoch": 2.329009433962264,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004456819548597082,
      "loss": 5.0689,
      "mean_token_accuracy": 0.2020641639828682,
      "num_tokens": 53268550.0,
      "step": 25675
    },
    {
      "entropy": 5.281864976882934,
      "epoch": 2.3294629898403483,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044566087054526946,
      "loss": 5.054,
      "mean_token_accuracy": 0.198075370490551,
      "num_tokens": 53277975.0,
      "step": 25680
    },
    {
      "entropy": 5.514032983779908,
      "epoch": 2.3299165457184325,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004456397827015204,
      "loss": 5.1998,
      "mean_token_accuracy": 0.19070678651332856,
      "num_tokens": 53287939.0,
      "step": 25685
    },
    {
      "entropy": 5.51838231086731,
      "epoch": 2.330370101596517,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004456186913288971,
      "loss": 5.1447,
      "mean_token_accuracy": 0.1948044329881668,
      "num_tokens": 53298355.0,
      "step": 25690
    },
    {
      "entropy": 5.497270345687866,
      "epoch": 2.3308236574746006,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004455975964278359,
      "loss": 5.2105,
      "mean_token_accuracy": 0.1851953685283661,
      "num_tokens": 53308513.0,
      "step": 25695
    },
    {
      "entropy": 5.525723171234131,
      "epoch": 2.331277213352685,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044557649799877285,
      "loss": 5.1467,
      "mean_token_accuracy": 0.18975103944540023,
      "num_tokens": 53318249.0,
      "step": 25700
    },
    {
      "entropy": 5.366728973388672,
      "epoch": 2.331730769230769,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004455553960421443,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19502483010292054,
      "num_tokens": 53328460.0,
      "step": 25705
    },
    {
      "entropy": 5.432518482208252,
      "epoch": 2.3321843251088534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044553429055838685,
      "loss": 5.1053,
      "mean_token_accuracy": 0.1982534185051918,
      "num_tokens": 53338984.0,
      "step": 25710
    },
    {
      "entropy": 5.498274612426758,
      "epoch": 2.3326378809869377,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044551318154793676,
      "loss": 5.1597,
      "mean_token_accuracy": 0.19266138672828675,
      "num_tokens": 53348092.0,
      "step": 25715
    },
    {
      "entropy": 5.3915853023529055,
      "epoch": 2.333091436865022,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044549206901123056,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1961325839161873,
      "num_tokens": 53357343.0,
      "step": 25720
    },
    {
      "entropy": 5.393789196014405,
      "epoch": 2.333544992743106,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000445470952948705,
      "loss": 5.018,
      "mean_token_accuracy": 0.19801161885261537,
      "num_tokens": 53367343.0,
      "step": 25725
    },
    {
      "entropy": 5.506641626358032,
      "epoch": 2.33399854862119,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044544983336079665,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19134739190340042,
      "num_tokens": 53379093.0,
      "step": 25730
    },
    {
      "entropy": 5.418594455718994,
      "epoch": 2.3344521044992743,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004454287102479423,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19462050944566728,
      "num_tokens": 53388883.0,
      "step": 25735
    },
    {
      "entropy": 5.4462274551391605,
      "epoch": 2.3349056603773586,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004454075836105788,
      "loss": 5.1066,
      "mean_token_accuracy": 0.19678357392549514,
      "num_tokens": 53399393.0,
      "step": 25740
    },
    {
      "entropy": 5.39352650642395,
      "epoch": 2.335359216255443,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044538645344914307,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20513215512037278,
      "num_tokens": 53408545.0,
      "step": 25745
    },
    {
      "entropy": 5.4003349304199215,
      "epoch": 2.3358127721335267,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004453653197640721,
      "loss": 5.132,
      "mean_token_accuracy": 0.19357440918684005,
      "num_tokens": 53418325.0,
      "step": 25750
    },
    {
      "entropy": 5.450744533538819,
      "epoch": 2.336266328011611,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044534418255580286,
      "loss": 5.1267,
      "mean_token_accuracy": 0.18739283382892608,
      "num_tokens": 53427984.0,
      "step": 25755
    },
    {
      "entropy": 5.545875692367554,
      "epoch": 2.336719883889695,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044532304182477245,
      "loss": 5.2169,
      "mean_token_accuracy": 0.1835709497332573,
      "num_tokens": 53438765.0,
      "step": 25760
    },
    {
      "entropy": 5.503445482254028,
      "epoch": 2.3371734397677795,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004453018975714182,
      "loss": 5.1232,
      "mean_token_accuracy": 0.1994517117738724,
      "num_tokens": 53448792.0,
      "step": 25765
    },
    {
      "entropy": 5.430860900878907,
      "epoch": 2.3376269956458637,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004452807497961773,
      "loss": 5.1755,
      "mean_token_accuracy": 0.19010599702596664,
      "num_tokens": 53459334.0,
      "step": 25770
    },
    {
      "entropy": 5.424499416351319,
      "epoch": 2.3380805515239476,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044525959849948704,
      "loss": 5.2021,
      "mean_token_accuracy": 0.18933138102293015,
      "num_tokens": 53469841.0,
      "step": 25775
    },
    {
      "entropy": 5.500175046920776,
      "epoch": 2.338534107402032,
      "grad_norm": 1.75,
      "learning_rate": 0.00044523844368178503,
      "loss": 5.1591,
      "mean_token_accuracy": 0.18785994350910187,
      "num_tokens": 53480660.0,
      "step": 25780
    },
    {
      "entropy": 5.486942338943481,
      "epoch": 2.338987663280116,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004452172853435085,
      "loss": 5.0635,
      "mean_token_accuracy": 0.19624347537755965,
      "num_tokens": 53490276.0,
      "step": 25785
    },
    {
      "entropy": 5.301886940002442,
      "epoch": 2.3394412191582004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004451961234850953,
      "loss": 5.0351,
      "mean_token_accuracy": 0.2061780422925949,
      "num_tokens": 53501100.0,
      "step": 25790
    },
    {
      "entropy": 5.3816040515899655,
      "epoch": 2.3398947750362846,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004451749581069827,
      "loss": 5.0892,
      "mean_token_accuracy": 0.19910448044538498,
      "num_tokens": 53511680.0,
      "step": 25795
    },
    {
      "entropy": 5.485548686981201,
      "epoch": 2.3403483309143684,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004451537892096087,
      "loss": 5.1805,
      "mean_token_accuracy": 0.19454847574234008,
      "num_tokens": 53521298.0,
      "step": 25800
    },
    {
      "entropy": 5.415682983398438,
      "epoch": 2.3408018867924527,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044513261679341094,
      "loss": 5.012,
      "mean_token_accuracy": 0.19942183047533035,
      "num_tokens": 53530413.0,
      "step": 25805
    },
    {
      "entropy": 5.370277452468872,
      "epoch": 2.341255442670537,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004451114408588273,
      "loss": 5.1143,
      "mean_token_accuracy": 0.19681460559368133,
      "num_tokens": 53540959.0,
      "step": 25810
    },
    {
      "entropy": 5.438940238952637,
      "epoch": 2.3417089985486212,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00044509026140629584,
      "loss": 5.1196,
      "mean_token_accuracy": 0.1964701995253563,
      "num_tokens": 53551265.0,
      "step": 25815
    },
    {
      "entropy": 5.494240379333496,
      "epoch": 2.3421625544267055,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044506907843625427,
      "loss": 5.1399,
      "mean_token_accuracy": 0.18944817632436753,
      "num_tokens": 53562666.0,
      "step": 25820
    },
    {
      "entropy": 5.4044805526733395,
      "epoch": 2.34261611030479,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044504789194914086,
      "loss": 5.2162,
      "mean_token_accuracy": 0.18839894235134125,
      "num_tokens": 53573688.0,
      "step": 25825
    },
    {
      "entropy": 5.488436126708985,
      "epoch": 2.3430696661828736,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044502670194539374,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19930662959814072,
      "num_tokens": 53584055.0,
      "step": 25830
    },
    {
      "entropy": 5.330423355102539,
      "epoch": 2.343523222060958,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00044500550842545104,
      "loss": 4.987,
      "mean_token_accuracy": 0.20091248154640198,
      "num_tokens": 53593765.0,
      "step": 25835
    },
    {
      "entropy": 5.33390245437622,
      "epoch": 2.343976777939042,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004449843113897512,
      "loss": 5.0294,
      "mean_token_accuracy": 0.20045515149831772,
      "num_tokens": 53603480.0,
      "step": 25840
    },
    {
      "entropy": 5.440939855575562,
      "epoch": 2.3444303338171264,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004449631108387325,
      "loss": 5.0954,
      "mean_token_accuracy": 0.1971719294786453,
      "num_tokens": 53613797.0,
      "step": 25845
    },
    {
      "entropy": 5.3853270530700685,
      "epoch": 2.34488388969521,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004449419067728333,
      "loss": 5.0643,
      "mean_token_accuracy": 0.19647799134254457,
      "num_tokens": 53623646.0,
      "step": 25850
    },
    {
      "entropy": 5.382241535186767,
      "epoch": 2.3453374455732945,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044492069919249206,
      "loss": 5.0661,
      "mean_token_accuracy": 0.1901763215661049,
      "num_tokens": 53634118.0,
      "step": 25855
    },
    {
      "entropy": 5.466392946243286,
      "epoch": 2.3457910014513788,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004448994880981476,
      "loss": 5.0961,
      "mean_token_accuracy": 0.19478834718465804,
      "num_tokens": 53644370.0,
      "step": 25860
    },
    {
      "entropy": 5.461419725418091,
      "epoch": 2.346244557329463,
      "grad_norm": 1.25,
      "learning_rate": 0.00044487827349023833,
      "loss": 5.1508,
      "mean_token_accuracy": 0.1879341185092926,
      "num_tokens": 53655560.0,
      "step": 25865
    },
    {
      "entropy": 5.419775915145874,
      "epoch": 2.3466981132075473,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044485705536920315,
      "loss": 5.2019,
      "mean_token_accuracy": 0.18899722546339034,
      "num_tokens": 53666401.0,
      "step": 25870
    },
    {
      "entropy": 5.5321417331695555,
      "epoch": 2.3471516690856316,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044483583373548075,
      "loss": 5.1706,
      "mean_token_accuracy": 0.20052152127027512,
      "num_tokens": 53676133.0,
      "step": 25875
    },
    {
      "entropy": 5.42283525466919,
      "epoch": 2.3476052249637154,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004448146085895101,
      "loss": 5.1212,
      "mean_token_accuracy": 0.18502788841724396,
      "num_tokens": 53687683.0,
      "step": 25880
    },
    {
      "entropy": 5.411877107620239,
      "epoch": 2.3480587808417996,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004447933799317299,
      "loss": 5.1825,
      "mean_token_accuracy": 0.18695258349180222,
      "num_tokens": 53698343.0,
      "step": 25885
    },
    {
      "entropy": 5.486205387115478,
      "epoch": 2.348512336719884,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004447721477625795,
      "loss": 5.1048,
      "mean_token_accuracy": 0.18943817168474197,
      "num_tokens": 53709014.0,
      "step": 25890
    },
    {
      "entropy": 5.416701984405518,
      "epoch": 2.348965892597968,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044475091208249774,
      "loss": 5.1023,
      "mean_token_accuracy": 0.19184874445199968,
      "num_tokens": 53718838.0,
      "step": 25895
    },
    {
      "entropy": 5.399969577789307,
      "epoch": 2.3494194484760524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004447296728919239,
      "loss": 5.0582,
      "mean_token_accuracy": 0.19905736446380615,
      "num_tokens": 53728526.0,
      "step": 25900
    },
    {
      "entropy": 5.4348925113677975,
      "epoch": 2.3498730043541363,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004447084301912972,
      "loss": 5.174,
      "mean_token_accuracy": 0.1973975643515587,
      "num_tokens": 53739261.0,
      "step": 25905
    },
    {
      "entropy": 5.450429439544678,
      "epoch": 2.3503265602322205,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004446871839810568,
      "loss": 5.1665,
      "mean_token_accuracy": 0.19391213059425355,
      "num_tokens": 53749599.0,
      "step": 25910
    },
    {
      "entropy": 5.455526971817017,
      "epoch": 2.350780116110305,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044466593426164224,
      "loss": 5.149,
      "mean_token_accuracy": 0.18947076052427292,
      "num_tokens": 53759898.0,
      "step": 25915
    },
    {
      "entropy": 5.434930276870728,
      "epoch": 2.351233671988389,
      "grad_norm": 1.203125,
      "learning_rate": 0.000444644681033493,
      "loss": 5.0933,
      "mean_token_accuracy": 0.1871253341436386,
      "num_tokens": 53770561.0,
      "step": 25920
    },
    {
      "entropy": 5.4039702892303465,
      "epoch": 2.3516872278664733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044462342429704845,
      "loss": 5.1038,
      "mean_token_accuracy": 0.19413934797048568,
      "num_tokens": 53780408.0,
      "step": 25925
    },
    {
      "entropy": 5.342476272583008,
      "epoch": 2.352140783744557,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044460216405274827,
      "loss": 5.028,
      "mean_token_accuracy": 0.20675642490386964,
      "num_tokens": 53789125.0,
      "step": 25930
    },
    {
      "entropy": 5.41141185760498,
      "epoch": 2.3525943396226414,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044458090030103216,
      "loss": 5.0657,
      "mean_token_accuracy": 0.19853561669588088,
      "num_tokens": 53799164.0,
      "step": 25935
    },
    {
      "entropy": 5.388838100433349,
      "epoch": 2.3530478955007257,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044455963304233976,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20183418393135072,
      "num_tokens": 53810710.0,
      "step": 25940
    },
    {
      "entropy": 5.341407918930054,
      "epoch": 2.35350145137881,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000444538362277111,
      "loss": 5.0807,
      "mean_token_accuracy": 0.19059078842401506,
      "num_tokens": 53820692.0,
      "step": 25945
    },
    {
      "entropy": 5.4672163963317875,
      "epoch": 2.353955007256894,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004445170880057857,
      "loss": 5.1452,
      "mean_token_accuracy": 0.19001904726028443,
      "num_tokens": 53831082.0,
      "step": 25950
    },
    {
      "entropy": 5.487230730056763,
      "epoch": 2.354408563134978,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044449581022880375,
      "loss": 5.1961,
      "mean_token_accuracy": 0.18312811255455017,
      "num_tokens": 53842602.0,
      "step": 25955
    },
    {
      "entropy": 5.411869525909424,
      "epoch": 2.3548621190130623,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004444745289466053,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1928018346428871,
      "num_tokens": 53853784.0,
      "step": 25960
    },
    {
      "entropy": 5.347913551330566,
      "epoch": 2.3553156748911466,
      "grad_norm": 1.25,
      "learning_rate": 0.00044445324415963043,
      "loss": 4.988,
      "mean_token_accuracy": 0.2043285220861435,
      "num_tokens": 53863526.0,
      "step": 25965
    },
    {
      "entropy": 5.342478704452515,
      "epoch": 2.355769230769231,
      "grad_norm": 1.375,
      "learning_rate": 0.0004444319558683192,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19219793230295182,
      "num_tokens": 53874594.0,
      "step": 25970
    },
    {
      "entropy": 5.47093710899353,
      "epoch": 2.356222786647315,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044441066407311205,
      "loss": 5.1119,
      "mean_token_accuracy": 0.18826810717582704,
      "num_tokens": 53884771.0,
      "step": 25975
    },
    {
      "entropy": 5.454582262039184,
      "epoch": 2.3566763425253994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004443893687744492,
      "loss": 5.093,
      "mean_token_accuracy": 0.19586560875177383,
      "num_tokens": 53895181.0,
      "step": 25980
    },
    {
      "entropy": 5.337407350540161,
      "epoch": 2.357129898403483,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044436806997277096,
      "loss": 4.9931,
      "mean_token_accuracy": 0.20317072719335555,
      "num_tokens": 53904153.0,
      "step": 25985
    },
    {
      "entropy": 5.375059652328491,
      "epoch": 2.3575834542815675,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044434676766851787,
      "loss": 5.0715,
      "mean_token_accuracy": 0.19876565784215927,
      "num_tokens": 53914064.0,
      "step": 25990
    },
    {
      "entropy": 5.437718868255615,
      "epoch": 2.3580370101596517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044432546186213054,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19215816557407378,
      "num_tokens": 53924240.0,
      "step": 25995
    },
    {
      "entropy": 5.4520305633544925,
      "epoch": 2.358490566037736,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044430415255404944,
      "loss": 5.1653,
      "mean_token_accuracy": 0.19797309637069702,
      "num_tokens": 53935395.0,
      "step": 26000
    },
    {
      "entropy": 5.394330739974976,
      "epoch": 2.35894412191582,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044428283974471533,
      "loss": 5.0887,
      "mean_token_accuracy": 0.19030908048152922,
      "num_tokens": 53945768.0,
      "step": 26005
    },
    {
      "entropy": 5.3491339683532715,
      "epoch": 2.359397677793904,
      "grad_norm": 1.234375,
      "learning_rate": 0.000444261523434569,
      "loss": 5.086,
      "mean_token_accuracy": 0.19660570472478867,
      "num_tokens": 53956467.0,
      "step": 26010
    },
    {
      "entropy": 5.362010431289673,
      "epoch": 2.3598512336719883,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044424020362405126,
      "loss": 5.0224,
      "mean_token_accuracy": 0.200508451461792,
      "num_tokens": 53967587.0,
      "step": 26015
    },
    {
      "entropy": 5.37784218788147,
      "epoch": 2.3603047895500726,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004442188803136029,
      "loss": 5.0693,
      "mean_token_accuracy": 0.1974543049931526,
      "num_tokens": 53977691.0,
      "step": 26020
    },
    {
      "entropy": 5.362749147415161,
      "epoch": 2.360758345428157,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000444197553503665,
      "loss": 5.1678,
      "mean_token_accuracy": 0.1857205241918564,
      "num_tokens": 53987608.0,
      "step": 26025
    },
    {
      "entropy": 5.471940517425537,
      "epoch": 2.361211901306241,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044417622319467857,
      "loss": 5.2325,
      "mean_token_accuracy": 0.18502036184072496,
      "num_tokens": 53998010.0,
      "step": 26030
    },
    {
      "entropy": 5.450388813018799,
      "epoch": 2.361665457184325,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044415488938708475,
      "loss": 5.1164,
      "mean_token_accuracy": 0.19691625237464905,
      "num_tokens": 54009852.0,
      "step": 26035
    },
    {
      "entropy": 5.412683916091919,
      "epoch": 2.3621190130624092,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004441335520813247,
      "loss": 5.0447,
      "mean_token_accuracy": 0.19751207977533342,
      "num_tokens": 54019635.0,
      "step": 26040
    },
    {
      "entropy": 5.428099012374878,
      "epoch": 2.3625725689404935,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044411221127783966,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1961450085043907,
      "num_tokens": 54030046.0,
      "step": 26045
    },
    {
      "entropy": 5.4522477149963375,
      "epoch": 2.3630261248185778,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004440908669770711,
      "loss": 5.1031,
      "mean_token_accuracy": 0.1905611574649811,
      "num_tokens": 54041197.0,
      "step": 26050
    },
    {
      "entropy": 5.47782187461853,
      "epoch": 2.3634796806966616,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004440695191794602,
      "loss": 5.0967,
      "mean_token_accuracy": 0.19779825508594512,
      "num_tokens": 54051592.0,
      "step": 26055
    },
    {
      "entropy": 5.407860136032104,
      "epoch": 2.363933236574746,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004440481678854487,
      "loss": 5.1598,
      "mean_token_accuracy": 0.18468309938907623,
      "num_tokens": 54062893.0,
      "step": 26060
    },
    {
      "entropy": 5.46908745765686,
      "epoch": 2.36438679245283,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044402681309547786,
      "loss": 5.1738,
      "mean_token_accuracy": 0.1882178857922554,
      "num_tokens": 54073651.0,
      "step": 26065
    },
    {
      "entropy": 5.409016036987305,
      "epoch": 2.3648403483309144,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044400545480998956,
      "loss": 5.0854,
      "mean_token_accuracy": 0.2030894085764885,
      "num_tokens": 54084889.0,
      "step": 26070
    },
    {
      "entropy": 5.34979395866394,
      "epoch": 2.3652939042089987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044398409302942533,
      "loss": 5.1027,
      "mean_token_accuracy": 0.19335113018751143,
      "num_tokens": 54095724.0,
      "step": 26075
    },
    {
      "entropy": 5.413653945922851,
      "epoch": 2.365747460087083,
      "grad_norm": 1.28125,
      "learning_rate": 0.000443962727754227,
      "loss": 5.07,
      "mean_token_accuracy": 0.19431506395339965,
      "num_tokens": 54106581.0,
      "step": 26080
    },
    {
      "entropy": 5.473599481582641,
      "epoch": 2.3662010159651667,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004439413589848364,
      "loss": 5.1402,
      "mean_token_accuracy": 0.19437065124511718,
      "num_tokens": 54116074.0,
      "step": 26085
    },
    {
      "entropy": 5.433373022079468,
      "epoch": 2.366654571843251,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004439199867216955,
      "loss": 5.1507,
      "mean_token_accuracy": 0.186528542637825,
      "num_tokens": 54126902.0,
      "step": 26090
    },
    {
      "entropy": 5.396596813201905,
      "epoch": 2.3671081277213353,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004438986109652462,
      "loss": 5.0712,
      "mean_token_accuracy": 0.1985920786857605,
      "num_tokens": 54137554.0,
      "step": 26095
    },
    {
      "entropy": 5.422725963592529,
      "epoch": 2.3675616835994195,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004438772317159306,
      "loss": 5.0762,
      "mean_token_accuracy": 0.1973506659269333,
      "num_tokens": 54148035.0,
      "step": 26100
    },
    {
      "entropy": 5.455394601821899,
      "epoch": 2.368015239477504,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044385584897419075,
      "loss": 5.0787,
      "mean_token_accuracy": 0.2049710839986801,
      "num_tokens": 54157417.0,
      "step": 26105
    },
    {
      "entropy": 5.410461378097534,
      "epoch": 2.3684687953555876,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000443834462740469,
      "loss": 5.126,
      "mean_token_accuracy": 0.19010598212480545,
      "num_tokens": 54168353.0,
      "step": 26110
    },
    {
      "entropy": 5.3784486770629885,
      "epoch": 2.368922351233672,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004438130730152074,
      "loss": 5.158,
      "mean_token_accuracy": 0.18750004172325135,
      "num_tokens": 54180182.0,
      "step": 26115
    },
    {
      "entropy": 5.468604469299317,
      "epoch": 2.369375907111756,
      "grad_norm": 1.609375,
      "learning_rate": 0.00044379167979884847,
      "loss": 5.1105,
      "mean_token_accuracy": 0.19235606789588927,
      "num_tokens": 54190973.0,
      "step": 26120
    },
    {
      "entropy": 5.465696668624878,
      "epoch": 2.3698294629898404,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004437702830918347,
      "loss": 5.1885,
      "mean_token_accuracy": 0.1887810170650482,
      "num_tokens": 54202077.0,
      "step": 26125
    },
    {
      "entropy": 5.4419056415557865,
      "epoch": 2.3702830188679247,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004437488828946084,
      "loss": 5.0974,
      "mean_token_accuracy": 0.19470890015363693,
      "num_tokens": 54211778.0,
      "step": 26130
    },
    {
      "entropy": 5.422237491607666,
      "epoch": 2.3707365747460085,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004437274792076122,
      "loss": 4.9796,
      "mean_token_accuracy": 0.20606542527675628,
      "num_tokens": 54220941.0,
      "step": 26135
    },
    {
      "entropy": 5.391664218902588,
      "epoch": 2.371190130624093,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044370607203128874,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18865661919116974,
      "num_tokens": 54231602.0,
      "step": 26140
    },
    {
      "entropy": 5.43750376701355,
      "epoch": 2.371643686502177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004436846613660807,
      "loss": 5.0972,
      "mean_token_accuracy": 0.19441655576229094,
      "num_tokens": 54241797.0,
      "step": 26145
    },
    {
      "entropy": 5.489793872833252,
      "epoch": 2.3720972423802613,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004436632472124308,
      "loss": 5.1764,
      "mean_token_accuracy": 0.18921824246644975,
      "num_tokens": 54253085.0,
      "step": 26150
    },
    {
      "entropy": 5.5350738048553465,
      "epoch": 2.3725507982583456,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004436418295707821,
      "loss": 5.1205,
      "mean_token_accuracy": 0.2002403736114502,
      "num_tokens": 54262164.0,
      "step": 26155
    },
    {
      "entropy": 5.381980514526367,
      "epoch": 2.3730043541364294,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004436204084415774,
      "loss": 5.136,
      "mean_token_accuracy": 0.192525514960289,
      "num_tokens": 54271922.0,
      "step": 26160
    },
    {
      "entropy": 5.412452554702758,
      "epoch": 2.3734579100145137,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004435989838252596,
      "loss": 5.0756,
      "mean_token_accuracy": 0.196290822327137,
      "num_tokens": 54281738.0,
      "step": 26165
    },
    {
      "entropy": 5.443496322631836,
      "epoch": 2.373911465892598,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004435775557222719,
      "loss": 5.0978,
      "mean_token_accuracy": 0.1957110270857811,
      "num_tokens": 54292252.0,
      "step": 26170
    },
    {
      "entropy": 5.41145658493042,
      "epoch": 2.374365021770682,
      "grad_norm": 1.65625,
      "learning_rate": 0.00044355612413305736,
      "loss": 5.1477,
      "mean_token_accuracy": 0.19168460816144944,
      "num_tokens": 54302518.0,
      "step": 26175
    },
    {
      "entropy": 5.398676824569702,
      "epoch": 2.3748185776487665,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004435346890580593,
      "loss": 5.1236,
      "mean_token_accuracy": 0.19442239701747893,
      "num_tokens": 54312917.0,
      "step": 26180
    },
    {
      "entropy": 5.430923700332642,
      "epoch": 2.3752721335268507,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044351325049772085,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20230170637369155,
      "num_tokens": 54322552.0,
      "step": 26185
    },
    {
      "entropy": 5.421678781509399,
      "epoch": 2.3757256894049346,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044349180845248536,
      "loss": 5.1019,
      "mean_token_accuracy": 0.19752747267484666,
      "num_tokens": 54332917.0,
      "step": 26190
    },
    {
      "entropy": 5.370342779159546,
      "epoch": 2.376179245283019,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004434703629227964,
      "loss": 5.0467,
      "mean_token_accuracy": 0.19748874455690385,
      "num_tokens": 54342791.0,
      "step": 26195
    },
    {
      "entropy": 5.356838035583496,
      "epoch": 2.376632801161103,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004434489139090975,
      "loss": 4.9835,
      "mean_token_accuracy": 0.20938123017549515,
      "num_tokens": 54352821.0,
      "step": 26200
    },
    {
      "entropy": 5.485013771057129,
      "epoch": 2.3770863570391874,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044342746141183206,
      "loss": 5.177,
      "mean_token_accuracy": 0.18855869621038437,
      "num_tokens": 54363095.0,
      "step": 26205
    },
    {
      "entropy": 5.356549787521362,
      "epoch": 2.377539912917271,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004434060054314438,
      "loss": 5.074,
      "mean_token_accuracy": 0.19063431173563003,
      "num_tokens": 54373841.0,
      "step": 26210
    },
    {
      "entropy": 5.35684380531311,
      "epoch": 2.3779934687953554,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044338454596837637,
      "loss": 5.0381,
      "mean_token_accuracy": 0.1961337670683861,
      "num_tokens": 54385012.0,
      "step": 26215
    },
    {
      "entropy": 5.3921064853668215,
      "epoch": 2.3784470246734397,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004433630830230737,
      "loss": 5.1183,
      "mean_token_accuracy": 0.1948670744895935,
      "num_tokens": 54395298.0,
      "step": 26220
    },
    {
      "entropy": 5.420003986358642,
      "epoch": 2.378900580551524,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004433416165959795,
      "loss": 5.1054,
      "mean_token_accuracy": 0.1937075212597847,
      "num_tokens": 54405752.0,
      "step": 26225
    },
    {
      "entropy": 5.481740856170655,
      "epoch": 2.3793541364296082,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004433201466875378,
      "loss": 5.1065,
      "mean_token_accuracy": 0.19339539706707,
      "num_tokens": 54416607.0,
      "step": 26230
    },
    {
      "entropy": 5.431913709640503,
      "epoch": 2.3798076923076925,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004432986732981926,
      "loss": 5.0842,
      "mean_token_accuracy": 0.19884953647851944,
      "num_tokens": 54426873.0,
      "step": 26235
    },
    {
      "entropy": 5.336875915527344,
      "epoch": 2.3802612481857763,
      "grad_norm": 1.109375,
      "learning_rate": 0.000443277196428388,
      "loss": 5.0168,
      "mean_token_accuracy": 0.2055708423256874,
      "num_tokens": 54437611.0,
      "step": 26240
    },
    {
      "entropy": 5.456114149093628,
      "epoch": 2.3807148040638606,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000443255716078568,
      "loss": 5.1068,
      "mean_token_accuracy": 0.19867335110902787,
      "num_tokens": 54448992.0,
      "step": 26245
    },
    {
      "entropy": 5.386185503005981,
      "epoch": 2.381168359941945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004432342322491769,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20108321756124498,
      "num_tokens": 54459172.0,
      "step": 26250
    },
    {
      "entropy": 5.341157245635986,
      "epoch": 2.381621915820029,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00044321274494065903,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19448653161525725,
      "num_tokens": 54469660.0,
      "step": 26255
    },
    {
      "entropy": 5.451760005950928,
      "epoch": 2.3820754716981134,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044319125415345886,
      "loss": 5.1852,
      "mean_token_accuracy": 0.18912017196416855,
      "num_tokens": 54479567.0,
      "step": 26260
    },
    {
      "entropy": 5.408758163452148,
      "epoch": 2.3825290275761972,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004431697598880205,
      "loss": 5.1475,
      "mean_token_accuracy": 0.19001547247171402,
      "num_tokens": 54489690.0,
      "step": 26265
    },
    {
      "entropy": 5.398011302947998,
      "epoch": 2.3829825834542815,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044314826214478877,
      "loss": 5.1325,
      "mean_token_accuracy": 0.19321519881486893,
      "num_tokens": 54500876.0,
      "step": 26270
    },
    {
      "entropy": 5.44443187713623,
      "epoch": 2.3834361393323658,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044312676092420804,
      "loss": 5.1535,
      "mean_token_accuracy": 0.18925971537828445,
      "num_tokens": 54510988.0,
      "step": 26275
    },
    {
      "entropy": 5.447180938720703,
      "epoch": 2.38388969521045,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004431052562267231,
      "loss": 5.0691,
      "mean_token_accuracy": 0.1987309768795967,
      "num_tokens": 54522149.0,
      "step": 26280
    },
    {
      "entropy": 5.488087320327759,
      "epoch": 2.3843432510885343,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044308374805277867,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18216657340526582,
      "num_tokens": 54533482.0,
      "step": 26285
    },
    {
      "entropy": 5.4569367408752445,
      "epoch": 2.384796806966618,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004430622364028194,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19010511338710784,
      "num_tokens": 54543058.0,
      "step": 26290
    },
    {
      "entropy": 5.464120101928711,
      "epoch": 2.3852503628447024,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044304072127729035,
      "loss": 5.091,
      "mean_token_accuracy": 0.19569306671619416,
      "num_tokens": 54552655.0,
      "step": 26295
    },
    {
      "entropy": 5.412784385681152,
      "epoch": 2.3857039187227866,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044301920267663633,
      "loss": 5.024,
      "mean_token_accuracy": 0.1938367962837219,
      "num_tokens": 54562696.0,
      "step": 26300
    },
    {
      "entropy": 5.390468263626099,
      "epoch": 2.386157474600871,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004429976806013024,
      "loss": 5.0765,
      "mean_token_accuracy": 0.20204827189445496,
      "num_tokens": 54572803.0,
      "step": 26305
    },
    {
      "entropy": 5.393419694900513,
      "epoch": 2.386611030478955,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004429761550517336,
      "loss": 5.1273,
      "mean_token_accuracy": 0.1923670545220375,
      "num_tokens": 54582450.0,
      "step": 26310
    },
    {
      "entropy": 5.344720697402954,
      "epoch": 2.387064586357039,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044295462602837504,
      "loss": 5.0531,
      "mean_token_accuracy": 0.19711477160453797,
      "num_tokens": 54591868.0,
      "step": 26315
    },
    {
      "entropy": 5.44606237411499,
      "epoch": 2.3875181422351233,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004429330935316721,
      "loss": 5.085,
      "mean_token_accuracy": 0.1916671797633171,
      "num_tokens": 54601858.0,
      "step": 26320
    },
    {
      "entropy": 5.450952529907227,
      "epoch": 2.3879716981132075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044291155756206996,
      "loss": 5.1341,
      "mean_token_accuracy": 0.1891488790512085,
      "num_tokens": 54612243.0,
      "step": 26325
    },
    {
      "entropy": 5.3951558589935305,
      "epoch": 2.388425253991292,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044289001812001404,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20383111387491226,
      "num_tokens": 54622365.0,
      "step": 26330
    },
    {
      "entropy": 5.3801429748535154,
      "epoch": 2.388878809869376,
      "grad_norm": 1.125,
      "learning_rate": 0.0004428684752059497,
      "loss": 5.0045,
      "mean_token_accuracy": 0.20077110677957535,
      "num_tokens": 54632356.0,
      "step": 26335
    },
    {
      "entropy": 5.468726253509521,
      "epoch": 2.3893323657474603,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004428469288203226,
      "loss": 5.1178,
      "mean_token_accuracy": 0.18975288420915604,
      "num_tokens": 54642976.0,
      "step": 26340
    },
    {
      "entropy": 5.4899507522583,
      "epoch": 2.389785921625544,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004428253789635783,
      "loss": 5.1213,
      "mean_token_accuracy": 0.19858651608228683,
      "num_tokens": 54652597.0,
      "step": 26345
    },
    {
      "entropy": 5.425835227966308,
      "epoch": 2.3902394775036284,
      "grad_norm": 1.125,
      "learning_rate": 0.0004428038256361622,
      "loss": 5.219,
      "mean_token_accuracy": 0.18807751387357713,
      "num_tokens": 54662930.0,
      "step": 26350
    },
    {
      "entropy": 5.598465251922607,
      "epoch": 2.3906930333817127,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004427822688385203,
      "loss": 5.3352,
      "mean_token_accuracy": 0.18742639869451522,
      "num_tokens": 54674562.0,
      "step": 26355
    },
    {
      "entropy": 5.474493455886841,
      "epoch": 2.391146589259797,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004427607085710984,
      "loss": 5.1277,
      "mean_token_accuracy": 0.1935278907418251,
      "num_tokens": 54684432.0,
      "step": 26360
    },
    {
      "entropy": 5.432281541824341,
      "epoch": 2.3916001451378808,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044273914483434226,
      "loss": 5.1246,
      "mean_token_accuracy": 0.193709097802639,
      "num_tokens": 54694060.0,
      "step": 26365
    },
    {
      "entropy": 5.3489076614379885,
      "epoch": 2.392053701015965,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044271757762869777,
      "loss": 5.1361,
      "mean_token_accuracy": 0.19431476891040803,
      "num_tokens": 54704411.0,
      "step": 26370
    },
    {
      "entropy": 5.4083600521087645,
      "epoch": 2.3925072568940493,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004426960069546112,
      "loss": 5.1122,
      "mean_token_accuracy": 0.19368250370025636,
      "num_tokens": 54715185.0,
      "step": 26375
    },
    {
      "entropy": 5.4495643138885494,
      "epoch": 2.3929608127721336,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004426744328125284,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20014284998178483,
      "num_tokens": 54725529.0,
      "step": 26380
    },
    {
      "entropy": 5.503159666061402,
      "epoch": 2.393414368650218,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004426528552028955,
      "loss": 5.186,
      "mean_token_accuracy": 0.19484029710292816,
      "num_tokens": 54736992.0,
      "step": 26385
    },
    {
      "entropy": 5.375679588317871,
      "epoch": 2.393867924528302,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044263127412615897,
      "loss": 5.0607,
      "mean_token_accuracy": 0.19604058861732482,
      "num_tokens": 54746951.0,
      "step": 26390
    },
    {
      "entropy": 5.400032329559326,
      "epoch": 2.394321480406386,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004426096895827649,
      "loss": 5.0327,
      "mean_token_accuracy": 0.2025053009390831,
      "num_tokens": 54757272.0,
      "step": 26395
    },
    {
      "entropy": 5.450964164733887,
      "epoch": 2.39477503628447,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044258810157315977,
      "loss": 5.1723,
      "mean_token_accuracy": 0.1917475551366806,
      "num_tokens": 54767835.0,
      "step": 26400
    },
    {
      "entropy": 5.437030506134033,
      "epoch": 2.3952285921625545,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044256651009779,
      "loss": 5.122,
      "mean_token_accuracy": 0.20004142373800277,
      "num_tokens": 54778432.0,
      "step": 26405
    },
    {
      "entropy": 5.473431062698364,
      "epoch": 2.3956821480406387,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044254491515710213,
      "loss": 5.1726,
      "mean_token_accuracy": 0.1882229208946228,
      "num_tokens": 54788582.0,
      "step": 26410
    },
    {
      "entropy": 5.508546924591064,
      "epoch": 2.3961357039187225,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004425233167515426,
      "loss": 5.1485,
      "mean_token_accuracy": 0.19360743463039398,
      "num_tokens": 54798848.0,
      "step": 26415
    },
    {
      "entropy": 5.379193925857544,
      "epoch": 2.396589259796807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004425017148815583,
      "loss": 5.0336,
      "mean_token_accuracy": 0.19564675986766816,
      "num_tokens": 54809867.0,
      "step": 26420
    },
    {
      "entropy": 5.421932744979858,
      "epoch": 2.397042815674891,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044248010954759575,
      "loss": 5.1019,
      "mean_token_accuracy": 0.1924191489815712,
      "num_tokens": 54819905.0,
      "step": 26425
    },
    {
      "entropy": 5.445680284500122,
      "epoch": 2.3974963715529753,
      "grad_norm": 1.578125,
      "learning_rate": 0.00044245850075010193,
      "loss": 5.1022,
      "mean_token_accuracy": 0.19531032294034958,
      "num_tokens": 54829723.0,
      "step": 26430
    },
    {
      "entropy": 5.384658527374268,
      "epoch": 2.3979499274310596,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044243688848952364,
      "loss": 5.0399,
      "mean_token_accuracy": 0.19864729344844817,
      "num_tokens": 54840008.0,
      "step": 26435
    },
    {
      "entropy": 5.4060498714447025,
      "epoch": 2.398403483309144,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004424152727663078,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19077180624008178,
      "num_tokens": 54851693.0,
      "step": 26440
    },
    {
      "entropy": 5.458563327789307,
      "epoch": 2.3988570391872277,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004423936535809015,
      "loss": 5.0652,
      "mean_token_accuracy": 0.19485504776239396,
      "num_tokens": 54862086.0,
      "step": 26445
    },
    {
      "entropy": 5.5464818477630615,
      "epoch": 2.399310595065312,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004423720309337517,
      "loss": 5.2007,
      "mean_token_accuracy": 0.19614293575286865,
      "num_tokens": 54872883.0,
      "step": 26450
    },
    {
      "entropy": 5.445002698898316,
      "epoch": 2.3997641509433962,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044235040482530576,
      "loss": 5.1215,
      "mean_token_accuracy": 0.18725504726171494,
      "num_tokens": 54882815.0,
      "step": 26455
    },
    {
      "entropy": 5.249258327484131,
      "epoch": 2.4002177068214805,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004423287752560108,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2050217017531395,
      "num_tokens": 54892480.0,
      "step": 26460
    },
    {
      "entropy": 5.325466728210449,
      "epoch": 2.4006712626995648,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044230714222631397,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2067038357257843,
      "num_tokens": 54901607.0,
      "step": 26465
    },
    {
      "entropy": 5.532425403594971,
      "epoch": 2.4011248185776486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044228550573666294,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1901920586824417,
      "num_tokens": 54913545.0,
      "step": 26470
    },
    {
      "entropy": 5.5140917778015135,
      "epoch": 2.401578374455733,
      "grad_norm": 1.453125,
      "learning_rate": 0.000442263865787505,
      "loss": 5.1458,
      "mean_token_accuracy": 0.1924404263496399,
      "num_tokens": 54924392.0,
      "step": 26475
    },
    {
      "entropy": 5.413170099258423,
      "epoch": 2.402031930333817,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004422422223792877,
      "loss": 5.0862,
      "mean_token_accuracy": 0.19667893052101135,
      "num_tokens": 54934114.0,
      "step": 26480
    },
    {
      "entropy": 5.45251145362854,
      "epoch": 2.4024854862119014,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004422205755124586,
      "loss": 5.1236,
      "mean_token_accuracy": 0.1966942712664604,
      "num_tokens": 54944268.0,
      "step": 26485
    },
    {
      "entropy": 5.466082191467285,
      "epoch": 2.4029390420899857,
      "grad_norm": 1.25,
      "learning_rate": 0.00044219892518746536,
      "loss": 5.1372,
      "mean_token_accuracy": 0.18885668218135834,
      "num_tokens": 54954877.0,
      "step": 26490
    },
    {
      "entropy": 5.416239261627197,
      "epoch": 2.4033925979680695,
      "grad_norm": 1.125,
      "learning_rate": 0.00044217727140475585,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20059396475553512,
      "num_tokens": 54964745.0,
      "step": 26495
    },
    {
      "entropy": 5.375434446334839,
      "epoch": 2.4038461538461537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044215561416477764,
      "loss": 4.9923,
      "mean_token_accuracy": 0.1994026631116867,
      "num_tokens": 54975256.0,
      "step": 26500
    },
    {
      "entropy": 5.453321266174316,
      "epoch": 2.404299709724238,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004421339534679788,
      "loss": 5.1384,
      "mean_token_accuracy": 0.18749998956918718,
      "num_tokens": 54986226.0,
      "step": 26505
    },
    {
      "entropy": 5.463255882263184,
      "epoch": 2.4047532656023223,
      "grad_norm": 1.25,
      "learning_rate": 0.0004421122893148072,
      "loss": 5.1184,
      "mean_token_accuracy": 0.18850539326667787,
      "num_tokens": 54997350.0,
      "step": 26510
    },
    {
      "entropy": 5.448295402526855,
      "epoch": 2.4052068214804065,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044209062170571086,
      "loss": 5.0944,
      "mean_token_accuracy": 0.19584467709064485,
      "num_tokens": 55007947.0,
      "step": 26515
    },
    {
      "entropy": 5.416891288757324,
      "epoch": 2.4056603773584904,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004420689506411379,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19409723877906798,
      "num_tokens": 55018789.0,
      "step": 26520
    },
    {
      "entropy": 5.414243698120117,
      "epoch": 2.4061139332365746,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044204727612153643,
      "loss": 5.0465,
      "mean_token_accuracy": 0.1924987643957138,
      "num_tokens": 55028934.0,
      "step": 26525
    },
    {
      "entropy": 5.439186143875122,
      "epoch": 2.406567489114659,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044202559814735475,
      "loss": 5.0634,
      "mean_token_accuracy": 0.20147013813257217,
      "num_tokens": 55039395.0,
      "step": 26530
    },
    {
      "entropy": 5.383830595016479,
      "epoch": 2.407021044992743,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004420039167190411,
      "loss": 5.0103,
      "mean_token_accuracy": 0.1953554317355156,
      "num_tokens": 55049932.0,
      "step": 26535
    },
    {
      "entropy": 5.423081731796264,
      "epoch": 2.4074746008708274,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000441982231837044,
      "loss": 5.1225,
      "mean_token_accuracy": 0.18865152150392533,
      "num_tokens": 55061532.0,
      "step": 26540
    },
    {
      "entropy": 5.4721489429473875,
      "epoch": 2.4079281567489117,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004419605435018117,
      "loss": 5.1615,
      "mean_token_accuracy": 0.19393083155155183,
      "num_tokens": 55072368.0,
      "step": 26545
    },
    {
      "entropy": 5.44450535774231,
      "epoch": 2.4083817126269955,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004419388517137928,
      "loss": 5.107,
      "mean_token_accuracy": 0.18910344690084457,
      "num_tokens": 55083128.0,
      "step": 26550
    },
    {
      "entropy": 5.45556960105896,
      "epoch": 2.40883526850508,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044191715647343594,
      "loss": 5.1155,
      "mean_token_accuracy": 0.1948249712586403,
      "num_tokens": 55094071.0,
      "step": 26555
    },
    {
      "entropy": 5.399063396453857,
      "epoch": 2.409288824383164,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044189545778118973,
      "loss": 5.0464,
      "mean_token_accuracy": 0.19844317585229873,
      "num_tokens": 55104504.0,
      "step": 26560
    },
    {
      "entropy": 5.400847434997559,
      "epoch": 2.4097423802612483,
      "grad_norm": 1.265625,
      "learning_rate": 0.000441873755637503,
      "loss": 5.111,
      "mean_token_accuracy": 0.19591050297021867,
      "num_tokens": 55115722.0,
      "step": 26565
    },
    {
      "entropy": 5.401132488250733,
      "epoch": 2.410195936139332,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004418520500428244,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20051231384277343,
      "num_tokens": 55126628.0,
      "step": 26570
    },
    {
      "entropy": 5.38998498916626,
      "epoch": 2.4106494920174164,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044183034099760287,
      "loss": 5.096,
      "mean_token_accuracy": 0.19565919786691666,
      "num_tokens": 55137352.0,
      "step": 26575
    },
    {
      "entropy": 5.4466643810272215,
      "epoch": 2.4111030478955007,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004418086285022874,
      "loss": 5.0896,
      "mean_token_accuracy": 0.2021987944841385,
      "num_tokens": 55147145.0,
      "step": 26580
    },
    {
      "entropy": 5.431712913513183,
      "epoch": 2.411556603773585,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044178691255732707,
      "loss": 5.1076,
      "mean_token_accuracy": 0.19431793987751006,
      "num_tokens": 55157386.0,
      "step": 26585
    },
    {
      "entropy": 5.433710527420044,
      "epoch": 2.412010159651669,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004417651931631708,
      "loss": 5.199,
      "mean_token_accuracy": 0.1916285499930382,
      "num_tokens": 55168969.0,
      "step": 26590
    },
    {
      "entropy": 5.393306827545166,
      "epoch": 2.4124637155297535,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004417434703202678,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1882381334900856,
      "num_tokens": 55180123.0,
      "step": 26595
    },
    {
      "entropy": 5.394544649124145,
      "epoch": 2.4129172714078373,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004417217440290673,
      "loss": 5.062,
      "mean_token_accuracy": 0.1969819337129593,
      "num_tokens": 55190668.0,
      "step": 26600
    },
    {
      "entropy": 5.379046154022217,
      "epoch": 2.4133708272859216,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004417000142900188,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20026251077651977,
      "num_tokens": 55200852.0,
      "step": 26605
    },
    {
      "entropy": 5.482087087631226,
      "epoch": 2.413824383164006,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004416782811035715,
      "loss": 5.2305,
      "mean_token_accuracy": 0.18375695496797562,
      "num_tokens": 55211174.0,
      "step": 26610
    },
    {
      "entropy": 5.3385344505310055,
      "epoch": 2.41427793904209,
      "grad_norm": 1.125,
      "learning_rate": 0.00044165654447017485,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20232074111700057,
      "num_tokens": 55222042.0,
      "step": 26615
    },
    {
      "entropy": 5.421162843704224,
      "epoch": 2.4147314949201744,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044163480439027845,
      "loss": 5.1294,
      "mean_token_accuracy": 0.19431929737329484,
      "num_tokens": 55231688.0,
      "step": 26620
    },
    {
      "entropy": 5.460363531112671,
      "epoch": 2.415185050798258,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004416130608643317,
      "loss": 5.1298,
      "mean_token_accuracy": 0.19735439121723175,
      "num_tokens": 55241089.0,
      "step": 26625
    },
    {
      "entropy": 5.445112657546997,
      "epoch": 2.4156386066763424,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004415913138927844,
      "loss": 5.0794,
      "mean_token_accuracy": 0.19574034512042998,
      "num_tokens": 55251511.0,
      "step": 26630
    },
    {
      "entropy": 5.344516658782959,
      "epoch": 2.4160921625544267,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004415695634760864,
      "loss": 5.082,
      "mean_token_accuracy": 0.19186503440141678,
      "num_tokens": 55262855.0,
      "step": 26635
    },
    {
      "entropy": 5.370210886001587,
      "epoch": 2.416545718432511,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004415478096146873,
      "loss": 4.9817,
      "mean_token_accuracy": 0.20217658430337906,
      "num_tokens": 55273035.0,
      "step": 26640
    },
    {
      "entropy": 5.384596061706543,
      "epoch": 2.4169992743105952,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000441526052309037,
      "loss": 5.074,
      "mean_token_accuracy": 0.2002120643854141,
      "num_tokens": 55284033.0,
      "step": 26645
    },
    {
      "entropy": 5.392719173431397,
      "epoch": 2.417452830188679,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004415042915595856,
      "loss": 5.0824,
      "mean_token_accuracy": 0.2002329096198082,
      "num_tokens": 55294290.0,
      "step": 26650
    },
    {
      "entropy": 5.441633462905884,
      "epoch": 2.4179063860667633,
      "grad_norm": 1.390625,
      "learning_rate": 0.000441482527366783,
      "loss": 5.1243,
      "mean_token_accuracy": 0.19733967185020446,
      "num_tokens": 55305681.0,
      "step": 26655
    },
    {
      "entropy": 5.343516159057617,
      "epoch": 2.4183599419448476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004414607597310793,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19777437001466752,
      "num_tokens": 55316647.0,
      "step": 26660
    },
    {
      "entropy": 5.466746473312378,
      "epoch": 2.418813497822932,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004414389886529247,
      "loss": 5.2082,
      "mean_token_accuracy": 0.18845624774694442,
      "num_tokens": 55328318.0,
      "step": 26665
    },
    {
      "entropy": 5.516735696792603,
      "epoch": 2.419267053701016,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004414172141327693,
      "loss": 5.119,
      "mean_token_accuracy": 0.18654351085424423,
      "num_tokens": 55338483.0,
      "step": 26670
    },
    {
      "entropy": 5.411861801147461,
      "epoch": 2.4197206095791,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004413954361710636,
      "loss": 5.171,
      "mean_token_accuracy": 0.19133355766534804,
      "num_tokens": 55348648.0,
      "step": 26675
    },
    {
      "entropy": 5.361277675628662,
      "epoch": 2.4201741654571842,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044137365476825785,
      "loss": 5.0428,
      "mean_token_accuracy": 0.1985986977815628,
      "num_tokens": 55358599.0,
      "step": 26680
    },
    {
      "entropy": 5.472253656387329,
      "epoch": 2.4206277213352685,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044135186992480253,
      "loss": 5.1266,
      "mean_token_accuracy": 0.1944972664117813,
      "num_tokens": 55369847.0,
      "step": 26685
    },
    {
      "entropy": 5.427027225494385,
      "epoch": 2.4210812772133528,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044133008164114814,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1974634125828743,
      "num_tokens": 55381235.0,
      "step": 26690
    },
    {
      "entropy": 5.395779418945312,
      "epoch": 2.421534833091437,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044130828991774526,
      "loss": 5.1203,
      "mean_token_accuracy": 0.19005934596061708,
      "num_tokens": 55391527.0,
      "step": 26695
    },
    {
      "entropy": 5.4578094482421875,
      "epoch": 2.4219883889695213,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044128649475504454,
      "loss": 5.1604,
      "mean_token_accuracy": 0.19557902216911316,
      "num_tokens": 55401528.0,
      "step": 26700
    },
    {
      "entropy": 5.516239833831787,
      "epoch": 2.422441944847605,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004412646961534968,
      "loss": 5.1503,
      "mean_token_accuracy": 0.188229301571846,
      "num_tokens": 55411411.0,
      "step": 26705
    },
    {
      "entropy": 5.374529552459717,
      "epoch": 2.4228955007256894,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004412428941135528,
      "loss": 4.9796,
      "mean_token_accuracy": 0.20231729447841645,
      "num_tokens": 55423164.0,
      "step": 26710
    },
    {
      "entropy": 5.408137941360474,
      "epoch": 2.4233490566037736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004412210886356633,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20495155304670334,
      "num_tokens": 55433397.0,
      "step": 26715
    },
    {
      "entropy": 5.382529449462891,
      "epoch": 2.423802612481858,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044119927972027936,
      "loss": 5.078,
      "mean_token_accuracy": 0.18939927071332932,
      "num_tokens": 55442898.0,
      "step": 26720
    },
    {
      "entropy": 5.369056177139282,
      "epoch": 2.4242561683599417,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044117746736785195,
      "loss": 5.0956,
      "mean_token_accuracy": 0.1970010668039322,
      "num_tokens": 55452226.0,
      "step": 26725
    },
    {
      "entropy": 5.420422554016113,
      "epoch": 2.424709724238026,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004411556515788322,
      "loss": 5.073,
      "mean_token_accuracy": 0.19588782340288163,
      "num_tokens": 55462503.0,
      "step": 26730
    },
    {
      "entropy": 5.4552630424499515,
      "epoch": 2.4251632801161103,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004411338323536712,
      "loss": 5.2254,
      "mean_token_accuracy": 0.1896193116903305,
      "num_tokens": 55472789.0,
      "step": 26735
    },
    {
      "entropy": 5.378664255142212,
      "epoch": 2.4256168359941945,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004411120096928203,
      "loss": 5.041,
      "mean_token_accuracy": 0.19714638143777846,
      "num_tokens": 55482572.0,
      "step": 26740
    },
    {
      "entropy": 5.454841947555542,
      "epoch": 2.426070391872279,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004410901835967306,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20143741369247437,
      "num_tokens": 55493033.0,
      "step": 26745
    },
    {
      "entropy": 5.405610466003418,
      "epoch": 2.426523947750363,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044106835406585354,
      "loss": 5.0498,
      "mean_token_accuracy": 0.19565775394439697,
      "num_tokens": 55503574.0,
      "step": 26750
    },
    {
      "entropy": 5.454274368286133,
      "epoch": 2.426977503628447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044104652110064067,
      "loss": 5.1312,
      "mean_token_accuracy": 0.1904119148850441,
      "num_tokens": 55514119.0,
      "step": 26755
    },
    {
      "entropy": 5.380870628356933,
      "epoch": 2.427431059506531,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004410246847015435,
      "loss": 5.0476,
      "mean_token_accuracy": 0.19407933801412583,
      "num_tokens": 55524973.0,
      "step": 26760
    },
    {
      "entropy": 5.419319677352905,
      "epoch": 2.4278846153846154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004410028448690134,
      "loss": 5.1477,
      "mean_token_accuracy": 0.1949432834982872,
      "num_tokens": 55534847.0,
      "step": 26765
    },
    {
      "entropy": 5.473853302001953,
      "epoch": 2.4283381712626997,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004409810016035023,
      "loss": 5.1203,
      "mean_token_accuracy": 0.18820348232984543,
      "num_tokens": 55544886.0,
      "step": 26770
    },
    {
      "entropy": 5.410084295272827,
      "epoch": 2.4287917271407835,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004409591549054617,
      "loss": 5.1785,
      "mean_token_accuracy": 0.18957405388355256,
      "num_tokens": 55554534.0,
      "step": 26775
    },
    {
      "entropy": 5.426144933700561,
      "epoch": 2.4292452830188678,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044093730477534355,
      "loss": 5.1346,
      "mean_token_accuracy": 0.18693427592515946,
      "num_tokens": 55564302.0,
      "step": 26780
    },
    {
      "entropy": 5.337580728530884,
      "epoch": 2.429698838896952,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004409154512135996,
      "loss": 5.0537,
      "mean_token_accuracy": 0.19700943529605866,
      "num_tokens": 55574543.0,
      "step": 26785
    },
    {
      "entropy": 5.325474119186401,
      "epoch": 2.4301523947750363,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004408935942206818,
      "loss": 5.0502,
      "mean_token_accuracy": 0.19998756498098375,
      "num_tokens": 55583849.0,
      "step": 26790
    },
    {
      "entropy": 5.409649991989136,
      "epoch": 2.4306059506531206,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004408717337970423,
      "loss": 5.1345,
      "mean_token_accuracy": 0.19450366199016572,
      "num_tokens": 55594294.0,
      "step": 26795
    },
    {
      "entropy": 5.449091291427612,
      "epoch": 2.431059506531205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044084986994313304,
      "loss": 5.1082,
      "mean_token_accuracy": 0.19243838191032409,
      "num_tokens": 55604185.0,
      "step": 26800
    },
    {
      "entropy": 5.388963222503662,
      "epoch": 2.4315130624092887,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004408280026594062,
      "loss": 5.0569,
      "mean_token_accuracy": 0.1974897190928459,
      "num_tokens": 55615203.0,
      "step": 26805
    },
    {
      "entropy": 5.378930854797363,
      "epoch": 2.431966618287373,
      "grad_norm": 1.390625,
      "learning_rate": 0.000440806131946314,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19528093487024306,
      "num_tokens": 55626157.0,
      "step": 26810
    },
    {
      "entropy": 5.484310960769653,
      "epoch": 2.432420174165457,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044078425780430877,
      "loss": 5.1042,
      "mean_token_accuracy": 0.18910741060972214,
      "num_tokens": 55636862.0,
      "step": 26815
    },
    {
      "entropy": 5.407122802734375,
      "epoch": 2.4328737300435415,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004407623802338428,
      "loss": 5.0282,
      "mean_token_accuracy": 0.20061464458703995,
      "num_tokens": 55647646.0,
      "step": 26820
    },
    {
      "entropy": 5.40669994354248,
      "epoch": 2.4333272859216257,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044074049923536855,
      "loss": 5.0976,
      "mean_token_accuracy": 0.1979583442211151,
      "num_tokens": 55659847.0,
      "step": 26825
    },
    {
      "entropy": 5.3606555461883545,
      "epoch": 2.4337808417997095,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004407186148093386,
      "loss": 5.1162,
      "mean_token_accuracy": 0.19633566439151764,
      "num_tokens": 55670626.0,
      "step": 26830
    },
    {
      "entropy": 5.4348513126373295,
      "epoch": 2.434234397677794,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044069672695620543,
      "loss": 5.165,
      "mean_token_accuracy": 0.1936669185757637,
      "num_tokens": 55681368.0,
      "step": 26835
    },
    {
      "entropy": 5.368651819229126,
      "epoch": 2.434687953555878,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004406748356764217,
      "loss": 5.0204,
      "mean_token_accuracy": 0.19066446125507355,
      "num_tokens": 55691606.0,
      "step": 26840
    },
    {
      "entropy": 5.422337055206299,
      "epoch": 2.4351415094339623,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044065294097044017,
      "loss": 5.1897,
      "mean_token_accuracy": 0.19016959965229036,
      "num_tokens": 55702223.0,
      "step": 26845
    },
    {
      "entropy": 5.409996795654297,
      "epoch": 2.4355950653120466,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004406310428387136,
      "loss": 5.0099,
      "mean_token_accuracy": 0.19840503931045533,
      "num_tokens": 55712729.0,
      "step": 26850
    },
    {
      "entropy": 5.389606904983521,
      "epoch": 2.4360486211901304,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004406091412816949,
      "loss": 5.1085,
      "mean_token_accuracy": 0.19879755675792693,
      "num_tokens": 55722863.0,
      "step": 26855
    },
    {
      "entropy": 5.446271324157715,
      "epoch": 2.4365021770682147,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044058723629983694,
      "loss": 5.1756,
      "mean_token_accuracy": 0.20013254284858703,
      "num_tokens": 55733124.0,
      "step": 26860
    },
    {
      "entropy": 5.532567548751831,
      "epoch": 2.436955732946299,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004405653278935927,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20330767780542375,
      "num_tokens": 55743438.0,
      "step": 26865
    },
    {
      "entropy": 5.382174348831176,
      "epoch": 2.4374092888243832,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044054341606341533,
      "loss": 4.9697,
      "mean_token_accuracy": 0.19963061064481735,
      "num_tokens": 55753602.0,
      "step": 26870
    },
    {
      "entropy": 5.358104085922241,
      "epoch": 2.4378628447024675,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004405215008097579,
      "loss": 5.032,
      "mean_token_accuracy": 0.20289485901594162,
      "num_tokens": 55763473.0,
      "step": 26875
    },
    {
      "entropy": 5.3607066631317135,
      "epoch": 2.4383164005805513,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00044049958213307364,
      "loss": 5.0605,
      "mean_token_accuracy": 0.1991008922457695,
      "num_tokens": 55772848.0,
      "step": 26880
    },
    {
      "entropy": 5.325480222702026,
      "epoch": 2.4387699564586356,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044047766003381584,
      "loss": 5.0284,
      "mean_token_accuracy": 0.201642444729805,
      "num_tokens": 55783468.0,
      "step": 26885
    },
    {
      "entropy": 5.399738454818726,
      "epoch": 2.43922351233672,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044045573451243793,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20219780653715133,
      "num_tokens": 55793619.0,
      "step": 26890
    },
    {
      "entropy": 5.464264440536499,
      "epoch": 2.439677068214804,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044043380556939326,
      "loss": 5.0807,
      "mean_token_accuracy": 0.1955818846821785,
      "num_tokens": 55803869.0,
      "step": 26895
    },
    {
      "entropy": 5.492811155319214,
      "epoch": 2.4401306240928884,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004404118732051353,
      "loss": 5.1166,
      "mean_token_accuracy": 0.19291312843561173,
      "num_tokens": 55814754.0,
      "step": 26900
    },
    {
      "entropy": 5.373237419128418,
      "epoch": 2.4405841799709727,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004403899374201176,
      "loss": 5.052,
      "mean_token_accuracy": 0.19373751282691956,
      "num_tokens": 55824862.0,
      "step": 26905
    },
    {
      "entropy": 5.29146556854248,
      "epoch": 2.4410377358490565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044036799821479397,
      "loss": 5.0463,
      "mean_token_accuracy": 0.19518034607172013,
      "num_tokens": 55835506.0,
      "step": 26910
    },
    {
      "entropy": 5.488911581039429,
      "epoch": 2.4414912917271407,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044034605558961786,
      "loss": 5.1685,
      "mean_token_accuracy": 0.18968981206417085,
      "num_tokens": 55846244.0,
      "step": 26915
    },
    {
      "entropy": 5.4152215957641605,
      "epoch": 2.441944847605225,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044032410954504325,
      "loss": 5.0472,
      "mean_token_accuracy": 0.2026146486401558,
      "num_tokens": 55856192.0,
      "step": 26920
    },
    {
      "entropy": 5.3748856544494625,
      "epoch": 2.4423984034833093,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044030216008152394,
      "loss": 5.0806,
      "mean_token_accuracy": 0.19297774732112885,
      "num_tokens": 55866509.0,
      "step": 26925
    },
    {
      "entropy": 5.431516838073731,
      "epoch": 2.442851959361393,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044028020719951376,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1964048609137535,
      "num_tokens": 55877321.0,
      "step": 26930
    },
    {
      "entropy": 5.430860233306885,
      "epoch": 2.4433055152394774,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004402582508994669,
      "loss": 5.1105,
      "mean_token_accuracy": 0.19690361618995667,
      "num_tokens": 55887933.0,
      "step": 26935
    },
    {
      "entropy": 5.35083532333374,
      "epoch": 2.4437590711175616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044023629118183726,
      "loss": 5.0269,
      "mean_token_accuracy": 0.19567239731550218,
      "num_tokens": 55897274.0,
      "step": 26940
    },
    {
      "entropy": 5.460517549514771,
      "epoch": 2.444212626995646,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044021432804707895,
      "loss": 5.1731,
      "mean_token_accuracy": 0.19314413666725158,
      "num_tokens": 55907393.0,
      "step": 26945
    },
    {
      "entropy": 5.398646974563599,
      "epoch": 2.44466618287373,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004401923614956463,
      "loss": 5.0802,
      "mean_token_accuracy": 0.1940239831805229,
      "num_tokens": 55918429.0,
      "step": 26950
    },
    {
      "entropy": 5.422098112106323,
      "epoch": 2.4451197387518144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004401703915279935,
      "loss": 5.0725,
      "mean_token_accuracy": 0.19506197720766066,
      "num_tokens": 55928733.0,
      "step": 26955
    },
    {
      "entropy": 5.416505432128906,
      "epoch": 2.4455732946298983,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044014841814457486,
      "loss": 5.0315,
      "mean_token_accuracy": 0.1993398368358612,
      "num_tokens": 55939128.0,
      "step": 26960
    },
    {
      "entropy": 5.438691091537476,
      "epoch": 2.4460268505079825,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004401264413458449,
      "loss": 5.1932,
      "mean_token_accuracy": 0.1920777827501297,
      "num_tokens": 55950237.0,
      "step": 26965
    },
    {
      "entropy": 5.413142251968384,
      "epoch": 2.446480406386067,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004401044611322581,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2023296669125557,
      "num_tokens": 55960436.0,
      "step": 26970
    },
    {
      "entropy": 5.385692024230957,
      "epoch": 2.446933962264151,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004400824775042689,
      "loss": 4.9903,
      "mean_token_accuracy": 0.19951148331165314,
      "num_tokens": 55970446.0,
      "step": 26975
    },
    {
      "entropy": 5.393396711349487,
      "epoch": 2.447387518142235,
      "grad_norm": 1.203125,
      "learning_rate": 0.000440060490462332,
      "loss": 5.1222,
      "mean_token_accuracy": 0.1920037552714348,
      "num_tokens": 55980706.0,
      "step": 26980
    },
    {
      "entropy": 5.4532723903656,
      "epoch": 2.447841074020319,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004400385000069021,
      "loss": 5.1706,
      "mean_token_accuracy": 0.18979728519916533,
      "num_tokens": 55990669.0,
      "step": 26985
    },
    {
      "entropy": 5.521352863311767,
      "epoch": 2.4482946298984034,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044001650613843395,
      "loss": 5.1868,
      "mean_token_accuracy": 0.18724606335163116,
      "num_tokens": 56000709.0,
      "step": 26990
    },
    {
      "entropy": 5.437122535705567,
      "epoch": 2.4487481857764877,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004399945088573824,
      "loss": 5.0869,
      "mean_token_accuracy": 0.19765886068344116,
      "num_tokens": 56011435.0,
      "step": 26995
    },
    {
      "entropy": 5.394946098327637,
      "epoch": 2.449201741654572,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043997250816420246,
      "loss": 5.0849,
      "mean_token_accuracy": 0.20457582771778107,
      "num_tokens": 56021264.0,
      "step": 27000
    },
    {
      "epoch": 2.449201741654572,
      "eval_entropy": 5.179353702230151,
      "eval_loss": 5.184836387634277,
      "eval_mean_token_accuracy": 0.19986100867896328,
      "eval_num_tokens": 56021264.0,
      "eval_runtime": 26.5764,
      "eval_samples_per_second": 1330.506,
      "eval_steps_per_second": 166.313,
      "step": 27000
    },
    {
      "entropy": 5.367761898040771,
      "epoch": 2.449655297532656,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004399505040593489,
      "loss": 5.1128,
      "mean_token_accuracy": 0.19100664258003236,
      "num_tokens": 56031463.0,
      "step": 27005
    },
    {
      "entropy": 5.364358711242676,
      "epoch": 2.45010885341074,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00043992849654327695,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2020670533180237,
      "num_tokens": 56041195.0,
      "step": 27010
    },
    {
      "entropy": 5.41200065612793,
      "epoch": 2.4505624092888243,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004399064856164416,
      "loss": 5.1221,
      "mean_token_accuracy": 0.1924453094601631,
      "num_tokens": 56051840.0,
      "step": 27015
    },
    {
      "entropy": 5.452530288696289,
      "epoch": 2.4510159651669086,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043988447127929817,
      "loss": 5.1469,
      "mean_token_accuracy": 0.19324837625026703,
      "num_tokens": 56062535.0,
      "step": 27020
    },
    {
      "entropy": 5.415385961532593,
      "epoch": 2.451469521044993,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043986245353230173,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20218175202608107,
      "num_tokens": 56072177.0,
      "step": 27025
    },
    {
      "entropy": 5.3668488502502445,
      "epoch": 2.451923076923077,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043984043237590785,
      "loss": 5.1158,
      "mean_token_accuracy": 0.1964294731616974,
      "num_tokens": 56081607.0,
      "step": 27030
    },
    {
      "entropy": 5.435453176498413,
      "epoch": 2.452376632801161,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043981840781057185,
      "loss": 5.0863,
      "mean_token_accuracy": 0.19493989646434784,
      "num_tokens": 56093371.0,
      "step": 27035
    },
    {
      "entropy": 5.396610450744629,
      "epoch": 2.452830188679245,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00043979637983674913,
      "loss": 5.0791,
      "mean_token_accuracy": 0.19490621387958526,
      "num_tokens": 56103973.0,
      "step": 27040
    },
    {
      "entropy": 5.404712152481079,
      "epoch": 2.4532837445573294,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004397743484548953,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19436376094818114,
      "num_tokens": 56113906.0,
      "step": 27045
    },
    {
      "entropy": 5.38924388885498,
      "epoch": 2.4537373004354137,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004397523136654659,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19301296323537825,
      "num_tokens": 56124172.0,
      "step": 27050
    },
    {
      "entropy": 5.389764976501465,
      "epoch": 2.454190856313498,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043973027546891677,
      "loss": 4.9509,
      "mean_token_accuracy": 0.20552754253149033,
      "num_tokens": 56133088.0,
      "step": 27055
    },
    {
      "entropy": 5.335010623931884,
      "epoch": 2.4546444121915822,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004397082338657034,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20288709849119185,
      "num_tokens": 56143098.0,
      "step": 27060
    },
    {
      "entropy": 5.414164018630982,
      "epoch": 2.455097968069666,
      "grad_norm": 1.125,
      "learning_rate": 0.00043968618885628185,
      "loss": 5.06,
      "mean_token_accuracy": 0.20028484463691712,
      "num_tokens": 56153889.0,
      "step": 27065
    },
    {
      "entropy": 5.39607253074646,
      "epoch": 2.4555515239477503,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043966414044110805,
      "loss": 5.0325,
      "mean_token_accuracy": 0.19601753652095794,
      "num_tokens": 56164164.0,
      "step": 27070
    },
    {
      "entropy": 5.434161281585693,
      "epoch": 2.4560050798258346,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043964208862063775,
      "loss": 5.1193,
      "mean_token_accuracy": 0.19328295588493347,
      "num_tokens": 56173172.0,
      "step": 27075
    },
    {
      "entropy": 5.444272565841675,
      "epoch": 2.456458635703919,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043962003339532714,
      "loss": 5.0837,
      "mean_token_accuracy": 0.2015843003988266,
      "num_tokens": 56183127.0,
      "step": 27080
    },
    {
      "entropy": 5.4163237571716305,
      "epoch": 2.4569121915820027,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043959797476563226,
      "loss": 5.1046,
      "mean_token_accuracy": 0.1959637850522995,
      "num_tokens": 56192528.0,
      "step": 27085
    },
    {
      "entropy": 5.367453670501709,
      "epoch": 2.457365747460087,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043957591273200943,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1981686383485794,
      "num_tokens": 56203239.0,
      "step": 27090
    },
    {
      "entropy": 5.4585553169250485,
      "epoch": 2.4578193033381712,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004395538472949146,
      "loss": 5.1362,
      "mean_token_accuracy": 0.18804559409618377,
      "num_tokens": 56214068.0,
      "step": 27095
    },
    {
      "entropy": 5.499663925170898,
      "epoch": 2.4582728592162555,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004395317784548045,
      "loss": 5.109,
      "mean_token_accuracy": 0.18797148615121842,
      "num_tokens": 56224829.0,
      "step": 27100
    },
    {
      "entropy": 5.4448432445526125,
      "epoch": 2.4587264150943398,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004395097062121351,
      "loss": 5.1216,
      "mean_token_accuracy": 0.19244975745677947,
      "num_tokens": 56235412.0,
      "step": 27105
    },
    {
      "entropy": 5.449646234512329,
      "epoch": 2.459179970972424,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043948763056736315,
      "loss": 5.1646,
      "mean_token_accuracy": 0.19293936043977739,
      "num_tokens": 56246785.0,
      "step": 27110
    },
    {
      "entropy": 5.448668003082275,
      "epoch": 2.459633526850508,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000439465551520945,
      "loss": 5.1389,
      "mean_token_accuracy": 0.19386271834373475,
      "num_tokens": 56256701.0,
      "step": 27115
    },
    {
      "entropy": 5.428870582580567,
      "epoch": 2.460087082728592,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043944346907333744,
      "loss": 5.0926,
      "mean_token_accuracy": 0.19903660267591478,
      "num_tokens": 56267418.0,
      "step": 27120
    },
    {
      "entropy": 5.3856123924255375,
      "epoch": 2.4605406386066764,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000439421383224997,
      "loss": 5.0096,
      "mean_token_accuracy": 0.19608869552612304,
      "num_tokens": 56278232.0,
      "step": 27125
    },
    {
      "entropy": 5.448611450195313,
      "epoch": 2.4609941944847606,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004393992939763804,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19906105399131774,
      "num_tokens": 56289866.0,
      "step": 27130
    },
    {
      "entropy": 5.381925344467163,
      "epoch": 2.4614477503628445,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043937720132794457,
      "loss": 5.0533,
      "mean_token_accuracy": 0.19801431894302368,
      "num_tokens": 56299595.0,
      "step": 27135
    },
    {
      "entropy": 5.450376224517822,
      "epoch": 2.4619013062409287,
      "grad_norm": 1.25,
      "learning_rate": 0.00043935510528014625,
      "loss": 5.1375,
      "mean_token_accuracy": 0.19975775182247163,
      "num_tokens": 56310152.0,
      "step": 27140
    },
    {
      "entropy": 5.463697338104248,
      "epoch": 2.462354862119013,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004393330058334426,
      "loss": 5.1858,
      "mean_token_accuracy": 0.18612506240606308,
      "num_tokens": 56320288.0,
      "step": 27145
    },
    {
      "entropy": 5.474953603744507,
      "epoch": 2.4628084179970973,
      "grad_norm": 1.125,
      "learning_rate": 0.00043931090298829047,
      "loss": 5.1145,
      "mean_token_accuracy": 0.1944684848189354,
      "num_tokens": 56330345.0,
      "step": 27150
    },
    {
      "entropy": 5.483683919906616,
      "epoch": 2.4632619738751815,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043928879674514694,
      "loss": 5.2313,
      "mean_token_accuracy": 0.19648835211992263,
      "num_tokens": 56340996.0,
      "step": 27155
    },
    {
      "entropy": 5.287700271606445,
      "epoch": 2.463715529753266,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004392666871044692,
      "loss": 4.9511,
      "mean_token_accuracy": 0.20781915485858918,
      "num_tokens": 56350326.0,
      "step": 27160
    },
    {
      "entropy": 5.3975060939788815,
      "epoch": 2.4641690856313496,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043924457406671456,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20122376531362535,
      "num_tokens": 56360556.0,
      "step": 27165
    },
    {
      "entropy": 5.434667110443115,
      "epoch": 2.464622641509434,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004392224576323403,
      "loss": 5.0541,
      "mean_token_accuracy": 0.19569238573312758,
      "num_tokens": 56369980.0,
      "step": 27170
    },
    {
      "entropy": 5.420522832870484,
      "epoch": 2.465076197387518,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043920033780180367,
      "loss": 5.1125,
      "mean_token_accuracy": 0.19162882566452027,
      "num_tokens": 56379974.0,
      "step": 27175
    },
    {
      "entropy": 5.405654954910278,
      "epoch": 2.4655297532656024,
      "grad_norm": 1.125,
      "learning_rate": 0.0004391782145755623,
      "loss": 5.0155,
      "mean_token_accuracy": 0.2022262454032898,
      "num_tokens": 56390542.0,
      "step": 27180
    },
    {
      "entropy": 5.472509288787842,
      "epoch": 2.4659833091436867,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004391560879540736,
      "loss": 5.221,
      "mean_token_accuracy": 0.18512699007987976,
      "num_tokens": 56401755.0,
      "step": 27185
    },
    {
      "entropy": 5.4126060962677,
      "epoch": 2.4664368650217705,
      "grad_norm": 1.125,
      "learning_rate": 0.0004391339579377952,
      "loss": 5.0996,
      "mean_token_accuracy": 0.19068958908319472,
      "num_tokens": 56412223.0,
      "step": 27190
    },
    {
      "entropy": 5.441014814376831,
      "epoch": 2.4668904208998548,
      "grad_norm": 1.125,
      "learning_rate": 0.00043911182452718476,
      "loss": 5.1479,
      "mean_token_accuracy": 0.19299090951681136,
      "num_tokens": 56422947.0,
      "step": 27195
    },
    {
      "entropy": 5.377451038360595,
      "epoch": 2.467343976777939,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043908968772269984,
      "loss": 5.092,
      "mean_token_accuracy": 0.20025220811367034,
      "num_tokens": 56433189.0,
      "step": 27200
    },
    {
      "entropy": 5.437313079833984,
      "epoch": 2.4677975326560233,
      "grad_norm": 1.796875,
      "learning_rate": 0.00043906754752479836,
      "loss": 5.1148,
      "mean_token_accuracy": 0.19322952777147293,
      "num_tokens": 56444460.0,
      "step": 27205
    },
    {
      "entropy": 5.471961402893067,
      "epoch": 2.4682510885341076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004390454039339383,
      "loss": 5.1278,
      "mean_token_accuracy": 0.19390352815389633,
      "num_tokens": 56454368.0,
      "step": 27210
    },
    {
      "entropy": 5.346301507949829,
      "epoch": 2.4687046444121914,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043902325695057743,
      "loss": 5.0343,
      "mean_token_accuracy": 0.19764251708984376,
      "num_tokens": 56464153.0,
      "step": 27215
    },
    {
      "entropy": 5.351440668106079,
      "epoch": 2.4691582002902757,
      "grad_norm": 1.125,
      "learning_rate": 0.00043900110657517377,
      "loss": 5.0426,
      "mean_token_accuracy": 0.20216597616672516,
      "num_tokens": 56473627.0,
      "step": 27220
    },
    {
      "entropy": 5.3471275806427006,
      "epoch": 2.46961175616836,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004389789528081855,
      "loss": 5.0445,
      "mean_token_accuracy": 0.19809859544038771,
      "num_tokens": 56483660.0,
      "step": 27225
    },
    {
      "entropy": 5.4496983051300045,
      "epoch": 2.470065312046444,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004389567956500707,
      "loss": 5.0566,
      "mean_token_accuracy": 0.2026229277253151,
      "num_tokens": 56492653.0,
      "step": 27230
    },
    {
      "entropy": 5.499282264709473,
      "epoch": 2.4705188679245285,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004389346351012876,
      "loss": 5.2091,
      "mean_token_accuracy": 0.1865890309214592,
      "num_tokens": 56503289.0,
      "step": 27235
    },
    {
      "entropy": 5.496986961364746,
      "epoch": 2.4709724238026123,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043891247116229436,
      "loss": 5.16,
      "mean_token_accuracy": 0.18831212371587752,
      "num_tokens": 56513843.0,
      "step": 27240
    },
    {
      "entropy": 5.511669063568116,
      "epoch": 2.4714259796806965,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043889030383354957,
      "loss": 5.1174,
      "mean_token_accuracy": 0.19084074944257737,
      "num_tokens": 56524175.0,
      "step": 27245
    },
    {
      "entropy": 5.379713439941407,
      "epoch": 2.471879535558781,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004388681331155114,
      "loss": 5.0359,
      "mean_token_accuracy": 0.20516521036624907,
      "num_tokens": 56534976.0,
      "step": 27250
    },
    {
      "entropy": 5.420804500579834,
      "epoch": 2.472333091436865,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004388459590086386,
      "loss": 5.1267,
      "mean_token_accuracy": 0.19212482124567032,
      "num_tokens": 56545967.0,
      "step": 27255
    },
    {
      "entropy": 5.444358444213867,
      "epoch": 2.4727866473149493,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043882378151338956,
      "loss": 5.1268,
      "mean_token_accuracy": 0.19512087404727935,
      "num_tokens": 56555527.0,
      "step": 27260
    },
    {
      "entropy": 5.479075860977173,
      "epoch": 2.4732402031930336,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000438801600630223,
      "loss": 5.1093,
      "mean_token_accuracy": 0.18782702535390855,
      "num_tokens": 56565874.0,
      "step": 27265
    },
    {
      "entropy": 5.377644300460815,
      "epoch": 2.4736937590711174,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043877941635959755,
      "loss": 5.0247,
      "mean_token_accuracy": 0.19658600389957429,
      "num_tokens": 56575452.0,
      "step": 27270
    },
    {
      "entropy": 5.3261970520019535,
      "epoch": 2.4741473149492017,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043875722870197214,
      "loss": 5.0244,
      "mean_token_accuracy": 0.2032456010580063,
      "num_tokens": 56586176.0,
      "step": 27275
    },
    {
      "entropy": 5.402632808685302,
      "epoch": 2.474600870827286,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004387350376578054,
      "loss": 5.0763,
      "mean_token_accuracy": 0.1968648165464401,
      "num_tokens": 56595473.0,
      "step": 27280
    },
    {
      "entropy": 5.417623567581177,
      "epoch": 2.4750544267053702,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004387128432275563,
      "loss": 5.127,
      "mean_token_accuracy": 0.19077519476413726,
      "num_tokens": 56606016.0,
      "step": 27285
    },
    {
      "entropy": 5.4374754428863525,
      "epoch": 2.475507982583454,
      "grad_norm": 1.421875,
      "learning_rate": 0.000438690645411684,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19719622135162354,
      "num_tokens": 56617674.0,
      "step": 27290
    },
    {
      "entropy": 5.4310996532440186,
      "epoch": 2.4759615384615383,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043866844421064745,
      "loss": 5.0823,
      "mean_token_accuracy": 0.19438118487596512,
      "num_tokens": 56628077.0,
      "step": 27295
    },
    {
      "entropy": 5.4286762237548825,
      "epoch": 2.4764150943396226,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043864623962490575,
      "loss": 5.0374,
      "mean_token_accuracy": 0.19493543058633805,
      "num_tokens": 56638691.0,
      "step": 27300
    },
    {
      "entropy": 5.381913805007935,
      "epoch": 2.476868650217707,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004386240316549181,
      "loss": 5.079,
      "mean_token_accuracy": 0.19408734738826752,
      "num_tokens": 56650338.0,
      "step": 27305
    },
    {
      "entropy": 5.404742050170898,
      "epoch": 2.477322206095791,
      "grad_norm": 1.125,
      "learning_rate": 0.00043860182030114374,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20047744363546371,
      "num_tokens": 56660328.0,
      "step": 27310
    },
    {
      "entropy": 5.450244760513305,
      "epoch": 2.4777757619738754,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043857960556404214,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19589452743530272,
      "num_tokens": 56670674.0,
      "step": 27315
    },
    {
      "entropy": 5.402787160873413,
      "epoch": 2.478229317851959,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004385573874440725,
      "loss": 5.0721,
      "mean_token_accuracy": 0.19866886138916015,
      "num_tokens": 56680656.0,
      "step": 27320
    },
    {
      "entropy": 5.4160302639007565,
      "epoch": 2.4786828737300435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043853516594169446,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20092224776744844,
      "num_tokens": 56691039.0,
      "step": 27325
    },
    {
      "entropy": 5.423696231842041,
      "epoch": 2.4791364296081277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004385129410573675,
      "loss": 5.084,
      "mean_token_accuracy": 0.1938493564724922,
      "num_tokens": 56703101.0,
      "step": 27330
    },
    {
      "entropy": 5.389424753189087,
      "epoch": 2.479589985486212,
      "grad_norm": 1.5,
      "learning_rate": 0.00043849071279155124,
      "loss": 5.1265,
      "mean_token_accuracy": 0.1999271109700203,
      "num_tokens": 56712351.0,
      "step": 27335
    },
    {
      "entropy": 5.385401344299316,
      "epoch": 2.480043541364296,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043846848114470537,
      "loss": 5.1088,
      "mean_token_accuracy": 0.20244877338409423,
      "num_tokens": 56722836.0,
      "step": 27340
    },
    {
      "entropy": 5.473705053329468,
      "epoch": 2.48049709724238,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043844624611728974,
      "loss": 5.0909,
      "mean_token_accuracy": 0.19198721498250962,
      "num_tokens": 56734032.0,
      "step": 27345
    },
    {
      "entropy": 5.41905722618103,
      "epoch": 2.4809506531204644,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043842400770976406,
      "loss": 5.0582,
      "mean_token_accuracy": 0.199035581946373,
      "num_tokens": 56743536.0,
      "step": 27350
    },
    {
      "entropy": 5.34170355796814,
      "epoch": 2.4814042089985486,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004384017659225882,
      "loss": 5.0874,
      "mean_token_accuracy": 0.19496845155954362,
      "num_tokens": 56753475.0,
      "step": 27355
    },
    {
      "entropy": 5.377800607681275,
      "epoch": 2.481857764876633,
      "grad_norm": 1.734375,
      "learning_rate": 0.00043837952075622224,
      "loss": 5.0216,
      "mean_token_accuracy": 0.20081770271062852,
      "num_tokens": 56763602.0,
      "step": 27360
    },
    {
      "entropy": 5.460765218734741,
      "epoch": 2.482311320754717,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004383572722111261,
      "loss": 5.0112,
      "mean_token_accuracy": 0.20665454715490342,
      "num_tokens": 56773859.0,
      "step": 27365
    },
    {
      "entropy": 5.38170108795166,
      "epoch": 2.482764876632801,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043833502028775995,
      "loss": 5.0763,
      "mean_token_accuracy": 0.19952169358730315,
      "num_tokens": 56783810.0,
      "step": 27370
    },
    {
      "entropy": 5.43801064491272,
      "epoch": 2.4832184325108853,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043831276498658395,
      "loss": 5.1629,
      "mean_token_accuracy": 0.1875397816300392,
      "num_tokens": 56794088.0,
      "step": 27375
    },
    {
      "entropy": 5.344621467590332,
      "epoch": 2.4836719883889695,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004382905063080584,
      "loss": 5.0327,
      "mean_token_accuracy": 0.1973615363240242,
      "num_tokens": 56804623.0,
      "step": 27380
    },
    {
      "entropy": 5.353916072845459,
      "epoch": 2.484125544267054,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043826824425264357,
      "loss": 5.096,
      "mean_token_accuracy": 0.19025506526231767,
      "num_tokens": 56813842.0,
      "step": 27385
    },
    {
      "entropy": 5.48569688796997,
      "epoch": 2.484579100145138,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004382459788207997,
      "loss": 5.0842,
      "mean_token_accuracy": 0.1904136136174202,
      "num_tokens": 56823501.0,
      "step": 27390
    },
    {
      "entropy": 5.404024934768676,
      "epoch": 2.485032656023222,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043822371001298754,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19624642729759217,
      "num_tokens": 56834601.0,
      "step": 27395
    },
    {
      "entropy": 5.324504423141479,
      "epoch": 2.485486211901306,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004382014378296674,
      "loss": 5.0898,
      "mean_token_accuracy": 0.19539990425109863,
      "num_tokens": 56844701.0,
      "step": 27400
    },
    {
      "entropy": 5.367699766159058,
      "epoch": 2.4859397677793904,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004381791622713,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19920360296964645,
      "num_tokens": 56854867.0,
      "step": 27405
    },
    {
      "entropy": 5.462157297134399,
      "epoch": 2.4863933236574747,
      "grad_norm": 1.265625,
      "learning_rate": 0.000438156883338346,
      "loss": 5.1131,
      "mean_token_accuracy": 0.20046492367982865,
      "num_tokens": 56865688.0,
      "step": 27410
    },
    {
      "entropy": 5.451690816879273,
      "epoch": 2.486846879535559,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043813460103126595,
      "loss": 5.1404,
      "mean_token_accuracy": 0.19504693895578384,
      "num_tokens": 56876504.0,
      "step": 27415
    },
    {
      "entropy": 5.348641347885132,
      "epoch": 2.487300435413643,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043811231535052076,
      "loss": 5.0535,
      "mean_token_accuracy": 0.2017896592617035,
      "num_tokens": 56886083.0,
      "step": 27420
    },
    {
      "entropy": 5.436499547958374,
      "epoch": 2.487753991291727,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004380900262965714,
      "loss": 5.1513,
      "mean_token_accuracy": 0.19297454953193666,
      "num_tokens": 56897079.0,
      "step": 27425
    },
    {
      "entropy": 5.461305141448975,
      "epoch": 2.4882075471698113,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004380677338698787,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20076536536216735,
      "num_tokens": 56906962.0,
      "step": 27430
    },
    {
      "entropy": 5.432268762588501,
      "epoch": 2.4886611030478956,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043804543807090374,
      "loss": 5.1732,
      "mean_token_accuracy": 0.1936105713248253,
      "num_tokens": 56918156.0,
      "step": 27435
    },
    {
      "entropy": 5.361149692535401,
      "epoch": 2.48911465892598,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043802313890010757,
      "loss": 5.0709,
      "mean_token_accuracy": 0.19617870151996614,
      "num_tokens": 56928888.0,
      "step": 27440
    },
    {
      "entropy": 5.436874151229858,
      "epoch": 2.4895682148040637,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004380008363579513,
      "loss": 5.0682,
      "mean_token_accuracy": 0.1950429141521454,
      "num_tokens": 56939929.0,
      "step": 27445
    },
    {
      "entropy": 5.427339887619018,
      "epoch": 2.490021770682148,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043797853044489623,
      "loss": 5.0112,
      "mean_token_accuracy": 0.19927417039871215,
      "num_tokens": 56950992.0,
      "step": 27450
    },
    {
      "entropy": 5.424587917327881,
      "epoch": 2.490475326560232,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004379562211614036,
      "loss": 5.1749,
      "mean_token_accuracy": 0.19401944428682327,
      "num_tokens": 56962118.0,
      "step": 27455
    },
    {
      "entropy": 5.384808206558228,
      "epoch": 2.4909288824383164,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004379339085079348,
      "loss": 5.0873,
      "mean_token_accuracy": 0.19636207222938537,
      "num_tokens": 56972837.0,
      "step": 27460
    },
    {
      "entropy": 5.413208055496216,
      "epoch": 2.4913824383164007,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004379115924849512,
      "loss": 5.0711,
      "mean_token_accuracy": 0.1978232055902481,
      "num_tokens": 56982794.0,
      "step": 27465
    },
    {
      "entropy": 5.516129112243652,
      "epoch": 2.491835994194485,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004378892730929145,
      "loss": 5.2424,
      "mean_token_accuracy": 0.19208418130874633,
      "num_tokens": 56993979.0,
      "step": 27470
    },
    {
      "entropy": 5.433774852752686,
      "epoch": 2.492289550072569,
      "grad_norm": 1.375,
      "learning_rate": 0.00043786695033228604,
      "loss": 5.0853,
      "mean_token_accuracy": 0.1930247187614441,
      "num_tokens": 57004505.0,
      "step": 27475
    },
    {
      "entropy": 5.425076532363891,
      "epoch": 2.492743105950653,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043784462420352747,
      "loss": 5.1183,
      "mean_token_accuracy": 0.19356049448251725,
      "num_tokens": 57014121.0,
      "step": 27480
    },
    {
      "entropy": 5.444897270202636,
      "epoch": 2.4931966618287373,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043782229470710067,
      "loss": 5.0943,
      "mean_token_accuracy": 0.19372855126857758,
      "num_tokens": 57023748.0,
      "step": 27485
    },
    {
      "entropy": 5.366559553146362,
      "epoch": 2.4936502177068216,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004377999618434673,
      "loss": 5.0706,
      "mean_token_accuracy": 0.19405495822429658,
      "num_tokens": 57034782.0,
      "step": 27490
    },
    {
      "entropy": 5.359586906433106,
      "epoch": 2.4941037735849054,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004377776256130892,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20250319093465804,
      "num_tokens": 57045136.0,
      "step": 27495
    },
    {
      "entropy": 5.4207357406616214,
      "epoch": 2.4945573294629897,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043775528601642833,
      "loss": 5.2315,
      "mean_token_accuracy": 0.18663275241851807,
      "num_tokens": 57055893.0,
      "step": 27500
    },
    {
      "entropy": 5.417135381698609,
      "epoch": 2.495010885341074,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043773294305394665,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19476390182971953,
      "num_tokens": 57067189.0,
      "step": 27505
    },
    {
      "entropy": 5.461691665649414,
      "epoch": 2.4954644412191582,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043771059672610636,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20191597044467927,
      "num_tokens": 57077374.0,
      "step": 27510
    },
    {
      "entropy": 5.401415014266968,
      "epoch": 2.4959179970972425,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043768824703336933,
      "loss": 5.0744,
      "mean_token_accuracy": 0.1908416897058487,
      "num_tokens": 57088168.0,
      "step": 27515
    },
    {
      "entropy": 5.3731687545776365,
      "epoch": 2.4963715529753268,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004376658939761979,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20157163590192795,
      "num_tokens": 57098743.0,
      "step": 27520
    },
    {
      "entropy": 5.350860548019409,
      "epoch": 2.4968251088534106,
      "grad_norm": 1.25,
      "learning_rate": 0.0004376435375550544,
      "loss": 5.1867,
      "mean_token_accuracy": 0.19047414213418962,
      "num_tokens": 57109854.0,
      "step": 27525
    },
    {
      "entropy": 5.356866025924683,
      "epoch": 2.497278664731495,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043762117777040106,
      "loss": 5.0121,
      "mean_token_accuracy": 0.20137630105018617,
      "num_tokens": 57120296.0,
      "step": 27530
    },
    {
      "entropy": 5.3842058181762695,
      "epoch": 2.497732220609579,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004375988146227003,
      "loss": 5.0438,
      "mean_token_accuracy": 0.1976532071828842,
      "num_tokens": 57131295.0,
      "step": 27535
    },
    {
      "entropy": 5.429742431640625,
      "epoch": 2.4981857764876634,
      "grad_norm": 1.375,
      "learning_rate": 0.0004375764481124147,
      "loss": 5.1083,
      "mean_token_accuracy": 0.1965476393699646,
      "num_tokens": 57141936.0,
      "step": 27540
    },
    {
      "entropy": 5.344677400588989,
      "epoch": 2.4986393323657476,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043755407824000664,
      "loss": 5.0434,
      "mean_token_accuracy": 0.2039247304201126,
      "num_tokens": 57152510.0,
      "step": 27545
    },
    {
      "entropy": 5.399235439300537,
      "epoch": 2.4990928882438315,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043753170500593884,
      "loss": 5.0845,
      "mean_token_accuracy": 0.19281034618616105,
      "num_tokens": 57162546.0,
      "step": 27550
    },
    {
      "entropy": 5.322568035125732,
      "epoch": 2.4995464441219157,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043750932841067397,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2042108252644539,
      "num_tokens": 57174658.0,
      "step": 27555
    },
    {
      "entropy": 5.450945138931274,
      "epoch": 2.5,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043748694845467483,
      "loss": 5.2189,
      "mean_token_accuracy": 0.18601366728544236,
      "num_tokens": 57185157.0,
      "step": 27560
    },
    {
      "entropy": 5.408987617492675,
      "epoch": 2.5004535558780843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043746456513840405,
      "loss": 4.9839,
      "mean_token_accuracy": 0.19429257214069368,
      "num_tokens": 57195277.0,
      "step": 27565
    },
    {
      "entropy": 5.451808166503906,
      "epoch": 2.5009071117561685,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004374421784623248,
      "loss": 5.1098,
      "mean_token_accuracy": 0.19439501613378524,
      "num_tokens": 57205374.0,
      "step": 27570
    },
    {
      "entropy": 5.38882098197937,
      "epoch": 2.501360667634253,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00043741978842689987,
      "loss": 5.0679,
      "mean_token_accuracy": 0.19867158383131028,
      "num_tokens": 57215700.0,
      "step": 27575
    },
    {
      "entropy": 5.401166009902954,
      "epoch": 2.5018142235123366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004373973950325923,
      "loss": 5.1214,
      "mean_token_accuracy": 0.19389134496450425,
      "num_tokens": 57225765.0,
      "step": 27580
    },
    {
      "entropy": 5.41153769493103,
      "epoch": 2.502267779390421,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043737499827986527,
      "loss": 5.0941,
      "mean_token_accuracy": 0.19553497731685637,
      "num_tokens": 57235913.0,
      "step": 27585
    },
    {
      "entropy": 5.366300487518311,
      "epoch": 2.502721335268505,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043735259816918184,
      "loss": 5.0079,
      "mean_token_accuracy": 0.20154172480106353,
      "num_tokens": 57246204.0,
      "step": 27590
    },
    {
      "entropy": 5.379629135131836,
      "epoch": 2.5031748911465894,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004373301947010053,
      "loss": 5.1172,
      "mean_token_accuracy": 0.1906338229775429,
      "num_tokens": 57256167.0,
      "step": 27595
    },
    {
      "entropy": 5.343207120895386,
      "epoch": 2.5036284470246732,
      "grad_norm": 1.53125,
      "learning_rate": 0.00043730778787579895,
      "loss": 5.0544,
      "mean_token_accuracy": 0.2013910785317421,
      "num_tokens": 57266621.0,
      "step": 27600
    },
    {
      "entropy": 5.415818405151367,
      "epoch": 2.5040820029027575,
      "grad_norm": 1.78125,
      "learning_rate": 0.00043728537769402627,
      "loss": 5.0982,
      "mean_token_accuracy": 0.19761655628681182,
      "num_tokens": 57277282.0,
      "step": 27605
    },
    {
      "entropy": 5.433597421646118,
      "epoch": 2.5045355587808418,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004372629641561506,
      "loss": 5.167,
      "mean_token_accuracy": 0.18893757462501526,
      "num_tokens": 57287541.0,
      "step": 27610
    },
    {
      "entropy": 5.41346697807312,
      "epoch": 2.504989114658926,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043724054726263547,
      "loss": 5.11,
      "mean_token_accuracy": 0.19085682481527327,
      "num_tokens": 57298158.0,
      "step": 27615
    },
    {
      "entropy": 5.4388025283813475,
      "epoch": 2.5054426705370103,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004372181270139444,
      "loss": 5.1024,
      "mean_token_accuracy": 0.1938956469297409,
      "num_tokens": 57308356.0,
      "step": 27620
    },
    {
      "entropy": 5.400198316574096,
      "epoch": 2.5058962264150946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043719570341054127,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20556032359600068,
      "num_tokens": 57317905.0,
      "step": 27625
    },
    {
      "entropy": 5.425683069229126,
      "epoch": 2.5063497822931784,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043717327645288956,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19852415174245835,
      "num_tokens": 57328277.0,
      "step": 27630
    },
    {
      "entropy": 5.49717435836792,
      "epoch": 2.5068033381712627,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043715084614145316,
      "loss": 5.1494,
      "mean_token_accuracy": 0.19903456717729567,
      "num_tokens": 57338975.0,
      "step": 27635
    },
    {
      "entropy": 5.397202730178833,
      "epoch": 2.507256894049347,
      "grad_norm": 1.0625,
      "learning_rate": 0.000437128412476696,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19960702806711197,
      "num_tokens": 57349439.0,
      "step": 27640
    },
    {
      "entropy": 5.378646421432495,
      "epoch": 2.507710449927431,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004371059754590819,
      "loss": 5.1248,
      "mean_token_accuracy": 0.1993075802922249,
      "num_tokens": 57359414.0,
      "step": 27645
    },
    {
      "entropy": 5.513858985900879,
      "epoch": 2.508164005805515,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043708353508907494,
      "loss": 5.1522,
      "mean_token_accuracy": 0.192555470764637,
      "num_tokens": 57369711.0,
      "step": 27650
    },
    {
      "entropy": 5.377740812301636,
      "epoch": 2.5086175616835993,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004370610913671392,
      "loss": 5.0134,
      "mean_token_accuracy": 0.20826903581619263,
      "num_tokens": 57380001.0,
      "step": 27655
    },
    {
      "entropy": 5.421742391586304,
      "epoch": 2.5090711175616836,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00043703864429373875,
      "loss": 5.1978,
      "mean_token_accuracy": 0.1860716924071312,
      "num_tokens": 57390794.0,
      "step": 27660
    },
    {
      "entropy": 5.4729656219482425,
      "epoch": 2.509524673439768,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043701619386933793,
      "loss": 5.1328,
      "mean_token_accuracy": 0.18730636686086655,
      "num_tokens": 57401975.0,
      "step": 27665
    },
    {
      "entropy": 5.362549734115601,
      "epoch": 2.509978229317852,
      "grad_norm": 1.125,
      "learning_rate": 0.00043699374009440085,
      "loss": 4.959,
      "mean_token_accuracy": 0.21097642928361893,
      "num_tokens": 57411249.0,
      "step": 27670
    },
    {
      "entropy": 5.347667121887207,
      "epoch": 2.5104317851959363,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004369712829693919,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20445141047239304,
      "num_tokens": 57422679.0,
      "step": 27675
    },
    {
      "entropy": 5.389731931686401,
      "epoch": 2.51088534107402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004369488224947757,
      "loss": 5.0796,
      "mean_token_accuracy": 0.20091069787740706,
      "num_tokens": 57433265.0,
      "step": 27680
    },
    {
      "entropy": 5.39980993270874,
      "epoch": 2.5113388969521044,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004369263586710164,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19611217379570006,
      "num_tokens": 57444095.0,
      "step": 27685
    },
    {
      "entropy": 5.3841423988342285,
      "epoch": 2.5117924528301887,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004369038914985789,
      "loss": 5.0429,
      "mean_token_accuracy": 0.2003264382481575,
      "num_tokens": 57454059.0,
      "step": 27690
    },
    {
      "entropy": 5.38724274635315,
      "epoch": 2.512246008708273,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004368814209779276,
      "loss": 5.0381,
      "mean_token_accuracy": 0.19818007946014404,
      "num_tokens": 57463940.0,
      "step": 27695
    },
    {
      "entropy": 5.334531927108765,
      "epoch": 2.512699564586357,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043685894710952724,
      "loss": 5.0743,
      "mean_token_accuracy": 0.19673750251531602,
      "num_tokens": 57474731.0,
      "step": 27700
    },
    {
      "entropy": 5.385692882537842,
      "epoch": 2.513153120464441,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004368364698938426,
      "loss": 4.9985,
      "mean_token_accuracy": 0.2046023800969124,
      "num_tokens": 57485096.0,
      "step": 27705
    },
    {
      "entropy": 5.414707374572754,
      "epoch": 2.5136066763425253,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004368139893313385,
      "loss": 5.0746,
      "mean_token_accuracy": 0.19757582247257233,
      "num_tokens": 57496371.0,
      "step": 27710
    },
    {
      "entropy": 5.445034742355347,
      "epoch": 2.5140602322206096,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043679150542248,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19327596127986907,
      "num_tokens": 57506012.0,
      "step": 27715
    },
    {
      "entropy": 5.402710199356079,
      "epoch": 2.514513788098694,
      "grad_norm": 1.375,
      "learning_rate": 0.00043676901816773174,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1950491860508919,
      "num_tokens": 57516536.0,
      "step": 27720
    },
    {
      "entropy": 5.360824966430664,
      "epoch": 2.514967343976778,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004367465275675591,
      "loss": 5.0105,
      "mean_token_accuracy": 0.20159698277711868,
      "num_tokens": 57527450.0,
      "step": 27725
    },
    {
      "entropy": 5.459282350540161,
      "epoch": 2.5154208998548624,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043672403362242704,
      "loss": 5.1298,
      "mean_token_accuracy": 0.1972229391336441,
      "num_tokens": 57538234.0,
      "step": 27730
    },
    {
      "entropy": 5.418883752822876,
      "epoch": 2.515874455732946,
      "grad_norm": 1.125,
      "learning_rate": 0.00043670153633280066,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20331853181123732,
      "num_tokens": 57547938.0,
      "step": 27735
    },
    {
      "entropy": 5.40467939376831,
      "epoch": 2.5163280116110305,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004366790356991454,
      "loss": 5.0956,
      "mean_token_accuracy": 0.19832125157117844,
      "num_tokens": 57560553.0,
      "step": 27740
    },
    {
      "entropy": 5.337105798721313,
      "epoch": 2.5167815674891147,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004366565317219265,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20166149586439133,
      "num_tokens": 57570670.0,
      "step": 27745
    },
    {
      "entropy": 5.522860860824585,
      "epoch": 2.5172351233671986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043663402440160924,
      "loss": 5.2755,
      "mean_token_accuracy": 0.18531838655471802,
      "num_tokens": 57582377.0,
      "step": 27750
    },
    {
      "entropy": 5.402630281448364,
      "epoch": 2.517688679245283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004366115137386592,
      "loss": 4.976,
      "mean_token_accuracy": 0.20786523669958115,
      "num_tokens": 57592764.0,
      "step": 27755
    },
    {
      "entropy": 5.349255609512329,
      "epoch": 2.518142235123367,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004365889997335419,
      "loss": 5.0636,
      "mean_token_accuracy": 0.19647965729236602,
      "num_tokens": 57603768.0,
      "step": 27760
    },
    {
      "entropy": 5.32365779876709,
      "epoch": 2.5185957910014514,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043656648238672285,
      "loss": 5.0479,
      "mean_token_accuracy": 0.2032769575715065,
      "num_tokens": 57614346.0,
      "step": 27765
    },
    {
      "entropy": 5.496939373016358,
      "epoch": 2.5190493468795356,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004365439616986678,
      "loss": 5.2029,
      "mean_token_accuracy": 0.1945197820663452,
      "num_tokens": 57625015.0,
      "step": 27770
    },
    {
      "entropy": 5.482428455352784,
      "epoch": 2.51950290275762,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043652143766984247,
      "loss": 5.1574,
      "mean_token_accuracy": 0.18995508253574372,
      "num_tokens": 57635476.0,
      "step": 27775
    },
    {
      "entropy": 5.357411956787109,
      "epoch": 2.519956458635704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043649891030071264,
      "loss": 5.0416,
      "mean_token_accuracy": 0.1986810088157654,
      "num_tokens": 57645512.0,
      "step": 27780
    },
    {
      "entropy": 5.3772680282592775,
      "epoch": 2.520410014513788,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043647637959174425,
      "loss": 5.0516,
      "mean_token_accuracy": 0.196814925968647,
      "num_tokens": 57655766.0,
      "step": 27785
    },
    {
      "entropy": 5.440534925460815,
      "epoch": 2.5208635703918723,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043645384554340315,
      "loss": 5.0791,
      "mean_token_accuracy": 0.19864285588264466,
      "num_tokens": 57667325.0,
      "step": 27790
    },
    {
      "entropy": 5.363713169097901,
      "epoch": 2.5213171262699565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043643130815615536,
      "loss": 5.0579,
      "mean_token_accuracy": 0.1933349534869194,
      "num_tokens": 57677884.0,
      "step": 27795
    },
    {
      "entropy": 5.440874147415161,
      "epoch": 2.521770682148041,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043640876743046694,
      "loss": 5.1116,
      "mean_token_accuracy": 0.2007884368300438,
      "num_tokens": 57687765.0,
      "step": 27800
    },
    {
      "entropy": 5.388621282577515,
      "epoch": 2.5222242380261246,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004363862233668041,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1926063448190689,
      "num_tokens": 57697777.0,
      "step": 27805
    },
    {
      "entropy": 5.371286392211914,
      "epoch": 2.522677793904209,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000436363675965633,
      "loss": 5.0705,
      "mean_token_accuracy": 0.19636721014976502,
      "num_tokens": 57708765.0,
      "step": 27810
    },
    {
      "entropy": 5.423065042495727,
      "epoch": 2.523131349782293,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004363411252274201,
      "loss": 5.0967,
      "mean_token_accuracy": 0.20002916157245637,
      "num_tokens": 57719658.0,
      "step": 27815
    },
    {
      "entropy": 5.396345567703247,
      "epoch": 2.5235849056603774,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043631857115263153,
      "loss": 5.0615,
      "mean_token_accuracy": 0.20458050966262817,
      "num_tokens": 57730018.0,
      "step": 27820
    },
    {
      "entropy": 5.3088868141174315,
      "epoch": 2.5240384615384617,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004362960137417338,
      "loss": 5.0156,
      "mean_token_accuracy": 0.20366084575653076,
      "num_tokens": 57741251.0,
      "step": 27825
    },
    {
      "entropy": 5.389115858078003,
      "epoch": 2.524492017416546,
      "grad_norm": 1.125,
      "learning_rate": 0.0004362734529951933,
      "loss": 5.1061,
      "mean_token_accuracy": 0.19563048481941223,
      "num_tokens": 57751020.0,
      "step": 27830
    },
    {
      "entropy": 5.390361785888672,
      "epoch": 2.5249455732946298,
      "grad_norm": 1.375,
      "learning_rate": 0.00043625088891347685,
      "loss": 5.0533,
      "mean_token_accuracy": 0.20215084105730058,
      "num_tokens": 57760989.0,
      "step": 27835
    },
    {
      "entropy": 5.4776496410369875,
      "epoch": 2.525399129172714,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043622832149705086,
      "loss": 5.1307,
      "mean_token_accuracy": 0.19919588714838027,
      "num_tokens": 57770882.0,
      "step": 27840
    },
    {
      "entropy": 5.434988355636596,
      "epoch": 2.5258526850507983,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043620575074638206,
      "loss": 5.1477,
      "mean_token_accuracy": 0.19579990655183793,
      "num_tokens": 57781812.0,
      "step": 27845
    },
    {
      "entropy": 5.413989019393921,
      "epoch": 2.5263062409288826,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043618317666193734,
      "loss": 5.0274,
      "mean_token_accuracy": 0.2040962517261505,
      "num_tokens": 57792009.0,
      "step": 27850
    },
    {
      "entropy": 5.336366367340088,
      "epoch": 2.5267597968069664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043616059924418346,
      "loss": 5.0286,
      "mean_token_accuracy": 0.20862455517053605,
      "num_tokens": 57802109.0,
      "step": 27855
    },
    {
      "entropy": 5.411025667190552,
      "epoch": 2.5272133526850507,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004361380184935872,
      "loss": 5.08,
      "mean_token_accuracy": 0.19713224917650224,
      "num_tokens": 57812374.0,
      "step": 27860
    },
    {
      "entropy": 5.374635934829712,
      "epoch": 2.527666908563135,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004361154344106158,
      "loss": 4.9813,
      "mean_token_accuracy": 0.20326960682868958,
      "num_tokens": 57822936.0,
      "step": 27865
    },
    {
      "entropy": 5.510503530502319,
      "epoch": 2.528120464441219,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043609284699573604,
      "loss": 5.2491,
      "mean_token_accuracy": 0.19052855670452118,
      "num_tokens": 57832693.0,
      "step": 27870
    },
    {
      "entropy": 5.452493143081665,
      "epoch": 2.5285740203193035,
      "grad_norm": 1.25,
      "learning_rate": 0.00043607025624941523,
      "loss": 5.1111,
      "mean_token_accuracy": 0.19441504329442977,
      "num_tokens": 57843161.0,
      "step": 27875
    },
    {
      "entropy": 5.410515165328979,
      "epoch": 2.5290275761973877,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004360476621721204,
      "loss": 5.0563,
      "mean_token_accuracy": 0.20482183992862701,
      "num_tokens": 57853027.0,
      "step": 27880
    },
    {
      "entropy": 5.456954622268677,
      "epoch": 2.5294811320754715,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004360250647643189,
      "loss": 5.073,
      "mean_token_accuracy": 0.2024420380592346,
      "num_tokens": 57862821.0,
      "step": 27885
    },
    {
      "entropy": 5.366050434112549,
      "epoch": 2.529934687953556,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043600246402647805,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20375704616308213,
      "num_tokens": 57874065.0,
      "step": 27890
    },
    {
      "entropy": 5.383389663696289,
      "epoch": 2.53038824383164,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004359798599590652,
      "loss": 5.0944,
      "mean_token_accuracy": 0.19408720284700393,
      "num_tokens": 57885190.0,
      "step": 27895
    },
    {
      "entropy": 5.417861413955689,
      "epoch": 2.5308417997097243,
      "grad_norm": 1.125,
      "learning_rate": 0.00043595725256254786,
      "loss": 5.0303,
      "mean_token_accuracy": 0.19702100455760957,
      "num_tokens": 57894234.0,
      "step": 27900
    },
    {
      "entropy": 5.443327808380127,
      "epoch": 2.531295355587808,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004359346418373935,
      "loss": 5.2019,
      "mean_token_accuracy": 0.19020499289035797,
      "num_tokens": 57905401.0,
      "step": 27905
    },
    {
      "entropy": 5.363066816329956,
      "epoch": 2.5317489114658924,
      "grad_norm": 1.65625,
      "learning_rate": 0.00043591202778406975,
      "loss": 5.0359,
      "mean_token_accuracy": 0.19688884019851685,
      "num_tokens": 57916097.0,
      "step": 27910
    },
    {
      "entropy": 5.305980968475342,
      "epoch": 2.5322024673439767,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004358894104030442,
      "loss": 4.9594,
      "mean_token_accuracy": 0.20557164251804352,
      "num_tokens": 57926871.0,
      "step": 27915
    },
    {
      "entropy": 5.352419185638428,
      "epoch": 2.532656023222061,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043586678969478464,
      "loss": 5.0229,
      "mean_token_accuracy": 0.20886398553848268,
      "num_tokens": 57937344.0,
      "step": 27920
    },
    {
      "entropy": 5.480348873138428,
      "epoch": 2.5331095791001452,
      "grad_norm": 1.109375,
      "learning_rate": 0.000435844165659759,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19853079468011856,
      "num_tokens": 57947633.0,
      "step": 27925
    },
    {
      "entropy": 5.381428289413452,
      "epoch": 2.5335631349782295,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043582153829843493,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20024688690900802,
      "num_tokens": 57957948.0,
      "step": 27930
    },
    {
      "entropy": 5.4201703548431395,
      "epoch": 2.5340166908563138,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004357989076112805,
      "loss": 5.0647,
      "mean_token_accuracy": 0.19535111337900163,
      "num_tokens": 57967719.0,
      "step": 27935
    },
    {
      "entropy": 5.448672723770142,
      "epoch": 2.5344702467343976,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004357762735987637,
      "loss": 5.136,
      "mean_token_accuracy": 0.190582075715065,
      "num_tokens": 57978218.0,
      "step": 27940
    },
    {
      "entropy": 5.379021215438843,
      "epoch": 2.534923802612482,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004357536362613525,
      "loss": 4.9817,
      "mean_token_accuracy": 0.20436299294233323,
      "num_tokens": 57987860.0,
      "step": 27945
    },
    {
      "entropy": 5.4070827007293705,
      "epoch": 2.535377358490566,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043573099559951516,
      "loss": 5.0841,
      "mean_token_accuracy": 0.20014401227235795,
      "num_tokens": 57997872.0,
      "step": 27950
    },
    {
      "entropy": 5.437116527557373,
      "epoch": 2.5358309143686504,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004357083516137199,
      "loss": 5.1434,
      "mean_token_accuracy": 0.19429580271244049,
      "num_tokens": 58007771.0,
      "step": 27955
    },
    {
      "entropy": 5.427969360351563,
      "epoch": 2.536284470246734,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000435685704304435,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19337427318096162,
      "num_tokens": 58018972.0,
      "step": 27960
    },
    {
      "entropy": 5.372975301742554,
      "epoch": 2.5367380261248185,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043566305367212884,
      "loss": 4.9571,
      "mean_token_accuracy": 0.19993621110916138,
      "num_tokens": 58028581.0,
      "step": 27965
    },
    {
      "entropy": 5.350557279586792,
      "epoch": 2.5371915820029027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043564039971726964,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20058995485305786,
      "num_tokens": 58038877.0,
      "step": 27970
    },
    {
      "entropy": 5.391227579116821,
      "epoch": 2.537645137880987,
      "grad_norm": 1.765625,
      "learning_rate": 0.00043561774244032614,
      "loss": 5.0639,
      "mean_token_accuracy": 0.1994498386979103,
      "num_tokens": 58048777.0,
      "step": 27975
    },
    {
      "entropy": 5.449144077301026,
      "epoch": 2.5380986937590713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043559508184176675,
      "loss": 5.1416,
      "mean_token_accuracy": 0.1962524563074112,
      "num_tokens": 58060270.0,
      "step": 27980
    },
    {
      "entropy": 5.385779619216919,
      "epoch": 2.5385522496371555,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043557241792206014,
      "loss": 5.0868,
      "mean_token_accuracy": 0.1990053579211235,
      "num_tokens": 58070838.0,
      "step": 27985
    },
    {
      "entropy": 5.4378368854522705,
      "epoch": 2.5390058055152394,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043554975068167505,
      "loss": 5.1385,
      "mean_token_accuracy": 0.1986191302537918,
      "num_tokens": 58079933.0,
      "step": 27990
    },
    {
      "entropy": 5.38010458946228,
      "epoch": 2.5394593613933236,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004355270801210802,
      "loss": 5.088,
      "mean_token_accuracy": 0.19801419526338576,
      "num_tokens": 58090673.0,
      "step": 27995
    },
    {
      "entropy": 5.363089275360108,
      "epoch": 2.539912917271408,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004355044062407444,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21173454821109772,
      "num_tokens": 58100172.0,
      "step": 28000
    },
    {
      "entropy": 5.3438694953918455,
      "epoch": 2.540366473149492,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004354817290411366,
      "loss": 5.0627,
      "mean_token_accuracy": 0.20235658437013626,
      "num_tokens": 58108965.0,
      "step": 28005
    },
    {
      "entropy": 5.4009614944458,
      "epoch": 2.540820029027576,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043545904852272573,
      "loss": 5.0473,
      "mean_token_accuracy": 0.19140686839818954,
      "num_tokens": 58120094.0,
      "step": 28010
    },
    {
      "entropy": 5.459799671173096,
      "epoch": 2.5412735849056602,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043543636468598083,
      "loss": 5.1137,
      "mean_token_accuracy": 0.18963236212730408,
      "num_tokens": 58131525.0,
      "step": 28015
    },
    {
      "entropy": 5.379569864273071,
      "epoch": 2.5417271407837445,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043541367753137107,
      "loss": 5.0279,
      "mean_token_accuracy": 0.201113161444664,
      "num_tokens": 58141986.0,
      "step": 28020
    },
    {
      "entropy": 5.393915176391602,
      "epoch": 2.5421806966618288,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004353909870593656,
      "loss": 5.1019,
      "mean_token_accuracy": 0.1985570564866066,
      "num_tokens": 58154122.0,
      "step": 28025
    },
    {
      "entropy": 5.442276859283448,
      "epoch": 2.542634252539913,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004353682932704336,
      "loss": 5.0066,
      "mean_token_accuracy": 0.21012976467609407,
      "num_tokens": 58164148.0,
      "step": 28030
    },
    {
      "entropy": 5.32414698600769,
      "epoch": 2.5430878084179973,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004353455961650444,
      "loss": 5.0039,
      "mean_token_accuracy": 0.19818162471055983,
      "num_tokens": 58173836.0,
      "step": 28035
    },
    {
      "entropy": 5.3203369140625,
      "epoch": 2.543541364296081,
      "grad_norm": 1.125,
      "learning_rate": 0.0004353228957436675,
      "loss": 5.0668,
      "mean_token_accuracy": 0.1969632625579834,
      "num_tokens": 58184598.0,
      "step": 28040
    },
    {
      "entropy": 5.335827970504761,
      "epoch": 2.5439949201741654,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043530019200677215,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20004291236400604,
      "num_tokens": 58194820.0,
      "step": 28045
    },
    {
      "entropy": 5.396665811538696,
      "epoch": 2.5444484760522497,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043527748495482806,
      "loss": 5.1608,
      "mean_token_accuracy": 0.1930520251393318,
      "num_tokens": 58206589.0,
      "step": 28050
    },
    {
      "entropy": 5.442803525924683,
      "epoch": 2.544902031930334,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043525477458830473,
      "loss": 5.0888,
      "mean_token_accuracy": 0.19549889415502547,
      "num_tokens": 58216750.0,
      "step": 28055
    },
    {
      "entropy": 5.371806383132935,
      "epoch": 2.5453555878084178,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004352320609076718,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19790212512016297,
      "num_tokens": 58228496.0,
      "step": 28060
    },
    {
      "entropy": 5.376483821868897,
      "epoch": 2.545809143686502,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043520934391339894,
      "loss": 5.0235,
      "mean_token_accuracy": 0.19632636457681657,
      "num_tokens": 58237844.0,
      "step": 28065
    },
    {
      "entropy": 5.33961033821106,
      "epoch": 2.5462626995645863,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004351866236059561,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20029880851507187,
      "num_tokens": 58248522.0,
      "step": 28070
    },
    {
      "entropy": 5.360802173614502,
      "epoch": 2.5467162554426706,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043516389998581305,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2022661730647087,
      "num_tokens": 58258492.0,
      "step": 28075
    },
    {
      "entropy": 5.368976402282715,
      "epoch": 2.547169811320755,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004351411730534398,
      "loss": 5.0394,
      "mean_token_accuracy": 0.1958549961447716,
      "num_tokens": 58268683.0,
      "step": 28080
    },
    {
      "entropy": 5.40429162979126,
      "epoch": 2.547623367198839,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004351184428093062,
      "loss": 5.1173,
      "mean_token_accuracy": 0.1895530954003334,
      "num_tokens": 58279479.0,
      "step": 28085
    },
    {
      "entropy": 5.457309007644653,
      "epoch": 2.5480769230769234,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004350957092538825,
      "loss": 5.1558,
      "mean_token_accuracy": 0.1901289626955986,
      "num_tokens": 58290751.0,
      "step": 28090
    },
    {
      "entropy": 5.461227226257324,
      "epoch": 2.548530478955007,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004350729723876386,
      "loss": 5.1366,
      "mean_token_accuracy": 0.18946841210126877,
      "num_tokens": 58302163.0,
      "step": 28095
    },
    {
      "entropy": 5.388751792907715,
      "epoch": 2.5489840348330914,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043505023221104497,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19568925350904465,
      "num_tokens": 58311488.0,
      "step": 28100
    },
    {
      "entropy": 5.380607461929321,
      "epoch": 2.5494375907111757,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043502748872457166,
      "loss": 5.0525,
      "mean_token_accuracy": 0.1964387282729149,
      "num_tokens": 58321599.0,
      "step": 28105
    },
    {
      "entropy": 5.362976598739624,
      "epoch": 2.5498911465892595,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004350047419286892,
      "loss": 5.0234,
      "mean_token_accuracy": 0.19896968454122543,
      "num_tokens": 58330696.0,
      "step": 28110
    },
    {
      "entropy": 5.328888273239135,
      "epoch": 2.550344702467344,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00043498199182386775,
      "loss": 5.0922,
      "mean_token_accuracy": 0.19908124059438706,
      "num_tokens": 58341793.0,
      "step": 28115
    },
    {
      "entropy": 5.421567344665528,
      "epoch": 2.550798258345428,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043495923841057807,
      "loss": 5.1785,
      "mean_token_accuracy": 0.18872070014476777,
      "num_tokens": 58352260.0,
      "step": 28120
    },
    {
      "entropy": 5.538152456283569,
      "epoch": 2.5512518142235123,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043493648168929053,
      "loss": 5.2021,
      "mean_token_accuracy": 0.18601759374141694,
      "num_tokens": 58363052.0,
      "step": 28125
    },
    {
      "entropy": 5.47307243347168,
      "epoch": 2.5517053701015966,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004349137216604758,
      "loss": 5.1054,
      "mean_token_accuracy": 0.19336642622947692,
      "num_tokens": 58374039.0,
      "step": 28130
    },
    {
      "entropy": 5.324053192138672,
      "epoch": 2.552158925979681,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004348909583246045,
      "loss": 4.9168,
      "mean_token_accuracy": 0.2083549052476883,
      "num_tokens": 58383441.0,
      "step": 28135
    },
    {
      "entropy": 5.347598266601563,
      "epoch": 2.552612481857765,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004348681916821475,
      "loss": 5.0372,
      "mean_token_accuracy": 0.19972455501556396,
      "num_tokens": 58394164.0,
      "step": 28140
    },
    {
      "entropy": 5.404113101959228,
      "epoch": 2.553066037735849,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004348454217335755,
      "loss": 5.0761,
      "mean_token_accuracy": 0.19857614934444429,
      "num_tokens": 58404610.0,
      "step": 28145
    },
    {
      "entropy": 5.37535982131958,
      "epoch": 2.553519593613933,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043482264847935953,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20031402707099916,
      "num_tokens": 58414512.0,
      "step": 28150
    },
    {
      "entropy": 5.444642639160156,
      "epoch": 2.5539731494920175,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043479987191997045,
      "loss": 5.1548,
      "mean_token_accuracy": 0.19256415069103242,
      "num_tokens": 58424669.0,
      "step": 28155
    },
    {
      "entropy": 5.3608136653900145,
      "epoch": 2.5544267053701017,
      "grad_norm": 1.25,
      "learning_rate": 0.0004347770920558792,
      "loss": 4.9746,
      "mean_token_accuracy": 0.2056750625371933,
      "num_tokens": 58435348.0,
      "step": 28160
    },
    {
      "entropy": 5.4270326614379885,
      "epoch": 2.5548802612481856,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043475430888755706,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19812069982290267,
      "num_tokens": 58445534.0,
      "step": 28165
    },
    {
      "entropy": 5.361429882049561,
      "epoch": 2.55533381712627,
      "grad_norm": 1.375,
      "learning_rate": 0.0004347315224154751,
      "loss": 5.0968,
      "mean_token_accuracy": 0.1960151121020317,
      "num_tokens": 58456718.0,
      "step": 28170
    },
    {
      "entropy": 5.504360818862915,
      "epoch": 2.555787373004354,
      "grad_norm": 1.25,
      "learning_rate": 0.0004347087326401045,
      "loss": 5.1385,
      "mean_token_accuracy": 0.19003151953220368,
      "num_tokens": 58467176.0,
      "step": 28175
    },
    {
      "entropy": 5.411391353607177,
      "epoch": 2.5562409288824384,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043468593956191663,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20418331772089005,
      "num_tokens": 58477642.0,
      "step": 28180
    },
    {
      "entropy": 5.374653387069702,
      "epoch": 2.5566944847605226,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004346631431813829,
      "loss": 5.0721,
      "mean_token_accuracy": 0.1949272483587265,
      "num_tokens": 58489055.0,
      "step": 28185
    },
    {
      "entropy": 5.430748319625854,
      "epoch": 2.557148040638607,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004346403434989746,
      "loss": 5.0824,
      "mean_token_accuracy": 0.1927397817373276,
      "num_tokens": 58499062.0,
      "step": 28190
    },
    {
      "entropy": 5.358388757705688,
      "epoch": 2.5576015965166907,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004346175405151633,
      "loss": 5.0784,
      "mean_token_accuracy": 0.19377515316009522,
      "num_tokens": 58509154.0,
      "step": 28195
    },
    {
      "entropy": 5.372982549667358,
      "epoch": 2.558055152394775,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004345947342304207,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20020240545272827,
      "num_tokens": 58520056.0,
      "step": 28200
    },
    {
      "entropy": 5.412386465072632,
      "epoch": 2.5585087082728593,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043457192464521827,
      "loss": 5.0595,
      "mean_token_accuracy": 0.19752898514270784,
      "num_tokens": 58529934.0,
      "step": 28205
    },
    {
      "entropy": 5.333535194396973,
      "epoch": 2.5589622641509435,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004345491117600278,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20144687443971634,
      "num_tokens": 58540013.0,
      "step": 28210
    },
    {
      "entropy": 5.455470180511474,
      "epoch": 2.5594158200290273,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043452629557532097,
      "loss": 5.0705,
      "mean_token_accuracy": 0.19543950259685516,
      "num_tokens": 58550282.0,
      "step": 28215
    },
    {
      "entropy": 5.453017950057983,
      "epoch": 2.5598693759071116,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043450347609156974,
      "loss": 5.1056,
      "mean_token_accuracy": 0.18887785077095032,
      "num_tokens": 58560490.0,
      "step": 28220
    },
    {
      "entropy": 5.361964559555053,
      "epoch": 2.560322931785196,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043448065330924595,
      "loss": 5.084,
      "mean_token_accuracy": 0.19403723776340484,
      "num_tokens": 58570371.0,
      "step": 28225
    },
    {
      "entropy": 5.459331512451172,
      "epoch": 2.56077648766328,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004344578272288216,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20478795915842057,
      "num_tokens": 58580701.0,
      "step": 28230
    },
    {
      "entropy": 5.444489622116089,
      "epoch": 2.5612300435413644,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043443499785076885,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19664188772439956,
      "num_tokens": 58590903.0,
      "step": 28235
    },
    {
      "entropy": 5.389712285995484,
      "epoch": 2.5616835994194487,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004344121651755596,
      "loss": 5.0521,
      "mean_token_accuracy": 0.197299325466156,
      "num_tokens": 58600924.0,
      "step": 28240
    },
    {
      "entropy": 5.4122151851654055,
      "epoch": 2.5621371552975325,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043438932920366615,
      "loss": 5.1432,
      "mean_token_accuracy": 0.19789200127124787,
      "num_tokens": 58611369.0,
      "step": 28245
    },
    {
      "entropy": 5.4746373176574705,
      "epoch": 2.5625907111756168,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004343664899355608,
      "loss": 5.1239,
      "mean_token_accuracy": 0.20029057413339615,
      "num_tokens": 58622725.0,
      "step": 28250
    },
    {
      "entropy": 5.381461048126221,
      "epoch": 2.563044267053701,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043434364737171587,
      "loss": 5.0681,
      "mean_token_accuracy": 0.19515618085861205,
      "num_tokens": 58634157.0,
      "step": 28255
    },
    {
      "entropy": 5.363257122039795,
      "epoch": 2.5634978229317853,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043432080151260366,
      "loss": 4.9823,
      "mean_token_accuracy": 0.19521068185567855,
      "num_tokens": 58643384.0,
      "step": 28260
    },
    {
      "entropy": 5.383065176010132,
      "epoch": 2.563951378809869,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043429795235869667,
      "loss": 5.0299,
      "mean_token_accuracy": 0.2024572178721428,
      "num_tokens": 58653748.0,
      "step": 28265
    },
    {
      "entropy": 5.390145206451416,
      "epoch": 2.5644049346879534,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043427509991046746,
      "loss": 5.0703,
      "mean_token_accuracy": 0.19453651756048201,
      "num_tokens": 58664619.0,
      "step": 28270
    },
    {
      "entropy": 5.345679807662964,
      "epoch": 2.5648584905660377,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043425224416838857,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21045442372560502,
      "num_tokens": 58674506.0,
      "step": 28275
    },
    {
      "entropy": 5.447764158248901,
      "epoch": 2.565312046444122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004342293851329327,
      "loss": 5.1471,
      "mean_token_accuracy": 0.19011306166648864,
      "num_tokens": 58686252.0,
      "step": 28280
    },
    {
      "entropy": 5.4000989437103275,
      "epoch": 2.565765602322206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043420652280457255,
      "loss": 5.0529,
      "mean_token_accuracy": 0.19241922795772554,
      "num_tokens": 58696477.0,
      "step": 28285
    },
    {
      "entropy": 5.308055591583252,
      "epoch": 2.5662191582002905,
      "grad_norm": 1.125,
      "learning_rate": 0.00043418365718378096,
      "loss": 4.9522,
      "mean_token_accuracy": 0.2064671829342842,
      "num_tokens": 58706664.0,
      "step": 28290
    },
    {
      "entropy": 5.351271343231201,
      "epoch": 2.5666727140783747,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004341607882710307,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20037147253751755,
      "num_tokens": 58716950.0,
      "step": 28295
    },
    {
      "entropy": 5.366053342819214,
      "epoch": 2.5671262699564585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004341379160667949,
      "loss": 5.0476,
      "mean_token_accuracy": 0.19751273542642594,
      "num_tokens": 58727723.0,
      "step": 28300
    },
    {
      "entropy": 5.37007155418396,
      "epoch": 2.567579825834543,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004341150405715464,
      "loss": 5.1417,
      "mean_token_accuracy": 0.1936953142285347,
      "num_tokens": 58737956.0,
      "step": 28305
    },
    {
      "entropy": 5.391983556747436,
      "epoch": 2.568033381712627,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043409216178575826,
      "loss": 4.999,
      "mean_token_accuracy": 0.2095926061272621,
      "num_tokens": 58748476.0,
      "step": 28310
    },
    {
      "entropy": 5.42882719039917,
      "epoch": 2.5684869375907113,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004340692797099037,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20324602723121643,
      "num_tokens": 58759604.0,
      "step": 28315
    },
    {
      "entropy": 5.40426287651062,
      "epoch": 2.568940493468795,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004340463943444559,
      "loss": 5.037,
      "mean_token_accuracy": 0.20370614975690843,
      "num_tokens": 58769781.0,
      "step": 28320
    },
    {
      "entropy": 5.396906042098999,
      "epoch": 2.5693940493468794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004340235056898882,
      "loss": 5.0801,
      "mean_token_accuracy": 0.20026499032974243,
      "num_tokens": 58779384.0,
      "step": 28325
    },
    {
      "entropy": 5.484280729293824,
      "epoch": 2.5698476052249637,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043400061374667374,
      "loss": 5.1625,
      "mean_token_accuracy": 0.1915605276823044,
      "num_tokens": 58791953.0,
      "step": 28330
    },
    {
      "entropy": 5.348475122451783,
      "epoch": 2.570301161103048,
      "grad_norm": 1.375,
      "learning_rate": 0.0004339777185152862,
      "loss": 5.0578,
      "mean_token_accuracy": 0.20144594758749007,
      "num_tokens": 58802415.0,
      "step": 28335
    },
    {
      "entropy": 5.463997650146484,
      "epoch": 2.5707547169811322,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004339548199961989,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20100287199020386,
      "num_tokens": 58812582.0,
      "step": 28340
    },
    {
      "entropy": 5.400321960449219,
      "epoch": 2.5712082728592165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043393191818988543,
      "loss": 5.1088,
      "mean_token_accuracy": 0.1959491491317749,
      "num_tokens": 58822987.0,
      "step": 28345
    },
    {
      "entropy": 5.384636497497558,
      "epoch": 2.5716618287373003,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004339090130968193,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20268218964338303,
      "num_tokens": 58832651.0,
      "step": 28350
    },
    {
      "entropy": 5.376557397842407,
      "epoch": 2.5721153846153846,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004338861047174743,
      "loss": 5.0888,
      "mean_token_accuracy": 0.19533269852399826,
      "num_tokens": 58842493.0,
      "step": 28355
    },
    {
      "entropy": 5.371889972686768,
      "epoch": 2.572568940493469,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004338631930523242,
      "loss": 4.9757,
      "mean_token_accuracy": 0.20563452690839767,
      "num_tokens": 58852428.0,
      "step": 28360
    },
    {
      "entropy": 5.34083890914917,
      "epoch": 2.573022496371553,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004338402781018429,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20419654101133347,
      "num_tokens": 58863234.0,
      "step": 28365
    },
    {
      "entropy": 5.360314178466797,
      "epoch": 2.573476052249637,
      "grad_norm": 1.109375,
      "learning_rate": 0.000433817359866504,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20386557132005692,
      "num_tokens": 58872632.0,
      "step": 28370
    },
    {
      "entropy": 5.4650053024292,
      "epoch": 2.573929608127721,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004337944383467817,
      "loss": 5.241,
      "mean_token_accuracy": 0.1824258491396904,
      "num_tokens": 58882774.0,
      "step": 28375
    },
    {
      "entropy": 5.346919012069702,
      "epoch": 2.5743831640058055,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043377151354314996,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20454225987195968,
      "num_tokens": 58892380.0,
      "step": 28380
    },
    {
      "entropy": 5.423246383666992,
      "epoch": 2.5748367198838897,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004337485854560829,
      "loss": 5.0587,
      "mean_token_accuracy": 0.19338658899068834,
      "num_tokens": 58902659.0,
      "step": 28385
    },
    {
      "entropy": 5.3649310111999515,
      "epoch": 2.575290275761974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004337256540860545,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20085662454366685,
      "num_tokens": 58913979.0,
      "step": 28390
    },
    {
      "entropy": 5.396544647216797,
      "epoch": 2.5757438316400583,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004337027194335393,
      "loss": 5.1054,
      "mean_token_accuracy": 0.1991891086101532,
      "num_tokens": 58924481.0,
      "step": 28395
    },
    {
      "entropy": 5.42782998085022,
      "epoch": 2.576197387518142,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043367978149901125,
      "loss": 5.0877,
      "mean_token_accuracy": 0.19446210712194442,
      "num_tokens": 58936287.0,
      "step": 28400
    },
    {
      "entropy": 5.443893003463745,
      "epoch": 2.5766509433962264,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043365684028294493,
      "loss": 5.0774,
      "mean_token_accuracy": 0.19998634308576585,
      "num_tokens": 58946070.0,
      "step": 28405
    },
    {
      "entropy": 5.415624761581421,
      "epoch": 2.5771044992743106,
      "grad_norm": 1.25,
      "learning_rate": 0.00043363389578581483,
      "loss": 5.1261,
      "mean_token_accuracy": 0.18869203180074692,
      "num_tokens": 58956083.0,
      "step": 28410
    },
    {
      "entropy": 5.4103868961334225,
      "epoch": 2.577558055152395,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004336109480080953,
      "loss": 5.0748,
      "mean_token_accuracy": 0.1873430222272873,
      "num_tokens": 58967557.0,
      "step": 28415
    },
    {
      "entropy": 5.391271543502808,
      "epoch": 2.5780116110304787,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004335879969502608,
      "loss": 5.0237,
      "mean_token_accuracy": 0.200988008081913,
      "num_tokens": 58977509.0,
      "step": 28420
    },
    {
      "entropy": 5.3128430366516115,
      "epoch": 2.578465166908563,
      "grad_norm": 1.125,
      "learning_rate": 0.00043356504261278625,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21559532582759858,
      "num_tokens": 58987498.0,
      "step": 28425
    },
    {
      "entropy": 5.448055744171143,
      "epoch": 2.5789187227866472,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004335420849961462,
      "loss": 5.1966,
      "mean_token_accuracy": 0.1862797111272812,
      "num_tokens": 58998503.0,
      "step": 28430
    },
    {
      "entropy": 5.379032135009766,
      "epoch": 2.5793722786647315,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004335191241008154,
      "loss": 5.0409,
      "mean_token_accuracy": 0.1974438950419426,
      "num_tokens": 59007384.0,
      "step": 28435
    },
    {
      "entropy": 5.4840991497039795,
      "epoch": 2.5798258345428158,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004334961599272687,
      "loss": 5.1756,
      "mean_token_accuracy": 0.19217898249626159,
      "num_tokens": 59017739.0,
      "step": 28440
    },
    {
      "entropy": 5.433402824401855,
      "epoch": 2.5802793904209,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004334731924759811,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19666444063186644,
      "num_tokens": 59028439.0,
      "step": 28445
    },
    {
      "entropy": 5.404414367675781,
      "epoch": 2.5807329462989843,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004334502217474274,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2011512592434883,
      "num_tokens": 59039250.0,
      "step": 28450
    },
    {
      "entropy": 5.469974851608276,
      "epoch": 2.581186502177068,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043342724774208277,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19834097176790239,
      "num_tokens": 59049375.0,
      "step": 28455
    },
    {
      "entropy": 5.477208137512207,
      "epoch": 2.5816400580551524,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043340427046042237,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19694374799728392,
      "num_tokens": 59059873.0,
      "step": 28460
    },
    {
      "entropy": 5.407156038284302,
      "epoch": 2.5820936139332367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004333812899029212,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2045658275485039,
      "num_tokens": 59069288.0,
      "step": 28465
    },
    {
      "entropy": 5.3969491481781,
      "epoch": 2.5825471698113205,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043335830607005456,
      "loss": 5.0723,
      "mean_token_accuracy": 0.19560055583715438,
      "num_tokens": 59079339.0,
      "step": 28470
    },
    {
      "entropy": 5.382390642166138,
      "epoch": 2.5830007256894048,
      "grad_norm": 1.25,
      "learning_rate": 0.0004333353189622979,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19603794068098068,
      "num_tokens": 59089883.0,
      "step": 28475
    },
    {
      "entropy": 5.387158393859863,
      "epoch": 2.583454281567489,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004333123285801264,
      "loss": 5.1562,
      "mean_token_accuracy": 0.1993858054280281,
      "num_tokens": 59100532.0,
      "step": 28480
    },
    {
      "entropy": 5.398775482177735,
      "epoch": 2.5839078374455733,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004332893349240156,
      "loss": 5.0628,
      "mean_token_accuracy": 0.1983143851161003,
      "num_tokens": 59111295.0,
      "step": 28485
    },
    {
      "entropy": 5.460626888275146,
      "epoch": 2.5843613933236576,
      "grad_norm": 1.125,
      "learning_rate": 0.0004332663379944411,
      "loss": 5.0928,
      "mean_token_accuracy": 0.19365076422691346,
      "num_tokens": 59120808.0,
      "step": 28490
    },
    {
      "entropy": 5.4197570323944095,
      "epoch": 2.584814949201742,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043324333779187844,
      "loss": 5.1243,
      "mean_token_accuracy": 0.2015594318509102,
      "num_tokens": 59131438.0,
      "step": 28495
    },
    {
      "entropy": 5.340666389465332,
      "epoch": 2.585268505079826,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043322033431680314,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20191137939691545,
      "num_tokens": 59141701.0,
      "step": 28500
    },
    {
      "entropy": 5.351067543029785,
      "epoch": 2.58572206095791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000433197327569691,
      "loss": 5.0333,
      "mean_token_accuracy": 0.20057248324155807,
      "num_tokens": 59152184.0,
      "step": 28505
    },
    {
      "entropy": 5.386619663238525,
      "epoch": 2.586175616835994,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043317431755101785,
      "loss": 5.0102,
      "mean_token_accuracy": 0.2051554873585701,
      "num_tokens": 59162177.0,
      "step": 28510
    },
    {
      "entropy": 5.479893064498901,
      "epoch": 2.5866291727140784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004331513042612594,
      "loss": 5.1575,
      "mean_token_accuracy": 0.19613301753997803,
      "num_tokens": 59171964.0,
      "step": 28515
    },
    {
      "entropy": 5.375412368774414,
      "epoch": 2.5870827285921627,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004331282877008918,
      "loss": 5.04,
      "mean_token_accuracy": 0.20047928243875504,
      "num_tokens": 59182224.0,
      "step": 28520
    },
    {
      "entropy": 5.379922389984131,
      "epoch": 2.5875362844702465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004331052678703909,
      "loss": 5.047,
      "mean_token_accuracy": 0.19802451133728027,
      "num_tokens": 59191912.0,
      "step": 28525
    },
    {
      "entropy": 5.39660611152649,
      "epoch": 2.587989840348331,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004330822447702327,
      "loss": 5.0309,
      "mean_token_accuracy": 0.20051221400499344,
      "num_tokens": 59200984.0,
      "step": 28530
    },
    {
      "entropy": 5.471497249603272,
      "epoch": 2.588443396226415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043305921840089347,
      "loss": 5.0998,
      "mean_token_accuracy": 0.19892797172069548,
      "num_tokens": 59210672.0,
      "step": 28535
    },
    {
      "entropy": 5.367439699172974,
      "epoch": 2.5888969521044993,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043303618876284935,
      "loss": 5.1002,
      "mean_token_accuracy": 0.20148296803236007,
      "num_tokens": 59220230.0,
      "step": 28540
    },
    {
      "entropy": 5.306804275512695,
      "epoch": 2.5893505079825836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043301315585657645,
      "loss": 4.9932,
      "mean_token_accuracy": 0.20207904577255248,
      "num_tokens": 59230615.0,
      "step": 28545
    },
    {
      "entropy": 5.483529043197632,
      "epoch": 2.589804063860668,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004329901196825513,
      "loss": 5.1172,
      "mean_token_accuracy": 0.19827897250652313,
      "num_tokens": 59241717.0,
      "step": 28550
    },
    {
      "entropy": 5.51355996131897,
      "epoch": 2.5902576197387517,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043296708024125025,
      "loss": 5.128,
      "mean_token_accuracy": 0.19317781180143356,
      "num_tokens": 59251791.0,
      "step": 28555
    },
    {
      "entropy": 5.3451377868652346,
      "epoch": 2.590711175616836,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043294403753314965,
      "loss": 4.9584,
      "mean_token_accuracy": 0.20797019004821776,
      "num_tokens": 59262434.0,
      "step": 28560
    },
    {
      "entropy": 5.360514163970947,
      "epoch": 2.59116473149492,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004329209915587261,
      "loss": 5.1322,
      "mean_token_accuracy": 0.19719347208738328,
      "num_tokens": 59272138.0,
      "step": 28565
    },
    {
      "entropy": 5.4229034900665285,
      "epoch": 2.5916182873730045,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043289794231845624,
      "loss": 5.0971,
      "mean_token_accuracy": 0.1973898380994797,
      "num_tokens": 59283470.0,
      "step": 28570
    },
    {
      "entropy": 5.352436637878418,
      "epoch": 2.5920718432510883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004328748898128168,
      "loss": 5.0593,
      "mean_token_accuracy": 0.2008277326822281,
      "num_tokens": 59293670.0,
      "step": 28575
    },
    {
      "entropy": 5.390877103805542,
      "epoch": 2.5925253991291726,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004328518340422842,
      "loss": 5.0618,
      "mean_token_accuracy": 0.19778138697147368,
      "num_tokens": 59303728.0,
      "step": 28580
    },
    {
      "entropy": 5.349004554748535,
      "epoch": 2.592978955007257,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043282877500733555,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20152160674333572,
      "num_tokens": 59314271.0,
      "step": 28585
    },
    {
      "entropy": 5.4001892566680905,
      "epoch": 2.593432510885341,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004328057127084476,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20261212140321733,
      "num_tokens": 59324581.0,
      "step": 28590
    },
    {
      "entropy": 5.43292727470398,
      "epoch": 2.5938860667634254,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043278264714609733,
      "loss": 5.0415,
      "mean_token_accuracy": 0.19951809644699098,
      "num_tokens": 59333912.0,
      "step": 28595
    },
    {
      "entropy": 5.372965621948242,
      "epoch": 2.5943396226415096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043275957832076173,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20148476660251619,
      "num_tokens": 59344245.0,
      "step": 28600
    },
    {
      "entropy": 5.38805890083313,
      "epoch": 2.5947931785195935,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043273650623291775,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19702895134687423,
      "num_tokens": 59353825.0,
      "step": 28605
    },
    {
      "entropy": 5.421781539916992,
      "epoch": 2.5952467343976777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004327134308830427,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19377930015325545,
      "num_tokens": 59364901.0,
      "step": 28610
    },
    {
      "entropy": 5.363695907592773,
      "epoch": 2.595700290275762,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004326903522716137,
      "loss": 5.0351,
      "mean_token_accuracy": 0.2008543699979782,
      "num_tokens": 59374772.0,
      "step": 28615
    },
    {
      "entropy": 5.310523843765258,
      "epoch": 2.5961538461538463,
      "grad_norm": 1.15625,
      "learning_rate": 0.000432667270399108,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21552883684635163,
      "num_tokens": 59383640.0,
      "step": 28620
    },
    {
      "entropy": 5.427074909210205,
      "epoch": 2.59660740203193,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043264418526600305,
      "loss": 5.0022,
      "mean_token_accuracy": 0.2025330886244774,
      "num_tokens": 59392928.0,
      "step": 28625
    },
    {
      "entropy": 5.462540864944458,
      "epoch": 2.5970609579100143,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004326210968727761,
      "loss": 5.1185,
      "mean_token_accuracy": 0.18946553170681,
      "num_tokens": 59402969.0,
      "step": 28630
    },
    {
      "entropy": 5.435910272598266,
      "epoch": 2.5975145137880986,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004325980052199047,
      "loss": 5.1255,
      "mean_token_accuracy": 0.1902220591902733,
      "num_tokens": 59413439.0,
      "step": 28635
    },
    {
      "entropy": 5.4020778179168705,
      "epoch": 2.597968069666183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043257491030786644,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19200607389211655,
      "num_tokens": 59423403.0,
      "step": 28640
    },
    {
      "entropy": 5.465194034576416,
      "epoch": 2.598421625544267,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043255181213713883,
      "loss": 5.15,
      "mean_token_accuracy": 0.19386085569858552,
      "num_tokens": 59434519.0,
      "step": 28645
    },
    {
      "entropy": 5.4347724437713625,
      "epoch": 2.5988751814223514,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004325287107081996,
      "loss": 5.0735,
      "mean_token_accuracy": 0.1908993348479271,
      "num_tokens": 59445056.0,
      "step": 28650
    },
    {
      "entropy": 5.3549394607543945,
      "epoch": 2.5993287373004357,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004325056060215265,
      "loss": 5.0463,
      "mean_token_accuracy": 0.19591470658779145,
      "num_tokens": 59455037.0,
      "step": 28655
    },
    {
      "entropy": 5.3882036209106445,
      "epoch": 2.5997822931785195,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004324824980775973,
      "loss": 5.0844,
      "mean_token_accuracy": 0.1962878465652466,
      "num_tokens": 59465093.0,
      "step": 28660
    },
    {
      "entropy": 5.414080667495727,
      "epoch": 2.6002358490566038,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043245938687688995,
      "loss": 5.1402,
      "mean_token_accuracy": 0.1982944577932358,
      "num_tokens": 59475449.0,
      "step": 28665
    },
    {
      "entropy": 5.422813367843628,
      "epoch": 2.600689404934688,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043243627241988235,
      "loss": 5.0838,
      "mean_token_accuracy": 0.199821275472641,
      "num_tokens": 59484990.0,
      "step": 28670
    },
    {
      "entropy": 5.453827667236328,
      "epoch": 2.6011429608127723,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043241315470705254,
      "loss": 5.0821,
      "mean_token_accuracy": 0.1926708087325096,
      "num_tokens": 59495739.0,
      "step": 28675
    },
    {
      "entropy": 5.42250428199768,
      "epoch": 2.601596516690856,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043239003373887855,
      "loss": 5.0843,
      "mean_token_accuracy": 0.1983434244990349,
      "num_tokens": 59506432.0,
      "step": 28680
    },
    {
      "entropy": 5.394314384460449,
      "epoch": 2.6020500725689404,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004323669095158385,
      "loss": 5.0602,
      "mean_token_accuracy": 0.19577138721942902,
      "num_tokens": 59516585.0,
      "step": 28685
    },
    {
      "entropy": 5.392453002929687,
      "epoch": 2.6025036284470247,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004323437820384107,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19662971645593644,
      "num_tokens": 59526824.0,
      "step": 28690
    },
    {
      "entropy": 5.461570453643799,
      "epoch": 2.602957184325109,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004323206513070734,
      "loss": 5.111,
      "mean_token_accuracy": 0.19246241599321365,
      "num_tokens": 59537749.0,
      "step": 28695
    },
    {
      "entropy": 5.420508003234863,
      "epoch": 2.603410740203193,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000432297517322305,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1931758388876915,
      "num_tokens": 59548310.0,
      "step": 28700
    },
    {
      "entropy": 5.390656280517578,
      "epoch": 2.6038642960812775,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004322743800845838,
      "loss": 5.1376,
      "mean_token_accuracy": 0.19388025104999543,
      "num_tokens": 59558191.0,
      "step": 28705
    },
    {
      "entropy": 5.460296392440796,
      "epoch": 2.6043178519593613,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043225123959438836,
      "loss": 5.1192,
      "mean_token_accuracy": 0.1880297690629959,
      "num_tokens": 59568961.0,
      "step": 28710
    },
    {
      "entropy": 5.43510160446167,
      "epoch": 2.6047714078374455,
      "grad_norm": 1.125,
      "learning_rate": 0.00043222809585219716,
      "loss": 5.0659,
      "mean_token_accuracy": 0.1954086184501648,
      "num_tokens": 59579137.0,
      "step": 28715
    },
    {
      "entropy": 5.243053293228149,
      "epoch": 2.60522496371553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043220494885848896,
      "loss": 4.8746,
      "mean_token_accuracy": 0.20860507041215898,
      "num_tokens": 59589612.0,
      "step": 28720
    },
    {
      "entropy": 5.440770483016967,
      "epoch": 2.605678519593614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004321817986137423,
      "loss": 5.1014,
      "mean_token_accuracy": 0.1928172841668129,
      "num_tokens": 59600332.0,
      "step": 28725
    },
    {
      "entropy": 5.35710391998291,
      "epoch": 2.606132075471698,
      "grad_norm": 1.25,
      "learning_rate": 0.000432158645118436,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20773691534996033,
      "num_tokens": 59610491.0,
      "step": 28730
    },
    {
      "entropy": 5.495172357559204,
      "epoch": 2.606585631349782,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004321354883730488,
      "loss": 5.2337,
      "mean_token_accuracy": 0.18104492127895355,
      "num_tokens": 59620729.0,
      "step": 28735
    },
    {
      "entropy": 5.391151237487793,
      "epoch": 2.6070391872278664,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043211232837805973,
      "loss": 5.0149,
      "mean_token_accuracy": 0.19994231164455414,
      "num_tokens": 59631243.0,
      "step": 28740
    },
    {
      "entropy": 5.36821026802063,
      "epoch": 2.6074927431059507,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043208916513394767,
      "loss": 5.0528,
      "mean_token_accuracy": 0.20390599519014357,
      "num_tokens": 59642055.0,
      "step": 28745
    },
    {
      "entropy": 5.45222225189209,
      "epoch": 2.607946298984035,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043206599864119164,
      "loss": 5.1519,
      "mean_token_accuracy": 0.1929887890815735,
      "num_tokens": 59652144.0,
      "step": 28750
    },
    {
      "entropy": 5.424995183944702,
      "epoch": 2.6083998548621192,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043204282890027074,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20789813995361328,
      "num_tokens": 59662367.0,
      "step": 28755
    },
    {
      "entropy": 5.413872861862183,
      "epoch": 2.608853410740203,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043201965591166413,
      "loss": 5.0802,
      "mean_token_accuracy": 0.2003683030605316,
      "num_tokens": 59672729.0,
      "step": 28760
    },
    {
      "entropy": 5.44765272140503,
      "epoch": 2.6093069666182873,
      "grad_norm": 1.046875,
      "learning_rate": 0.000431996479675851,
      "loss": 5.1098,
      "mean_token_accuracy": 0.20238879024982454,
      "num_tokens": 59683282.0,
      "step": 28765
    },
    {
      "entropy": 5.455147838592529,
      "epoch": 2.6097605224963716,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043197330019331074,
      "loss": 5.1662,
      "mean_token_accuracy": 0.19005318284034728,
      "num_tokens": 59694297.0,
      "step": 28770
    },
    {
      "entropy": 5.341357660293579,
      "epoch": 2.610214078374456,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043195011746452244,
      "loss": 5.0007,
      "mean_token_accuracy": 0.20065861344337463,
      "num_tokens": 59704149.0,
      "step": 28775
    },
    {
      "entropy": 5.433978271484375,
      "epoch": 2.6106676342525397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043192693148996585,
      "loss": 5.1163,
      "mean_token_accuracy": 0.19925116002559662,
      "num_tokens": 59714811.0,
      "step": 28780
    },
    {
      "entropy": 5.407086992263794,
      "epoch": 2.611121190130624,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004319037422701203,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20311776399612427,
      "num_tokens": 59724950.0,
      "step": 28785
    },
    {
      "entropy": 5.431817626953125,
      "epoch": 2.611574746008708,
      "grad_norm": 1.125,
      "learning_rate": 0.0004318805498054654,
      "loss": 5.1764,
      "mean_token_accuracy": 0.18610777854919433,
      "num_tokens": 59735905.0,
      "step": 28790
    },
    {
      "entropy": 5.485265445709229,
      "epoch": 2.6120283018867925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043185735409648077,
      "loss": 5.1229,
      "mean_token_accuracy": 0.1969365209341049,
      "num_tokens": 59746743.0,
      "step": 28795
    },
    {
      "entropy": 5.4247386932373045,
      "epoch": 2.6124818577648767,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043183415514364606,
      "loss": 5.0571,
      "mean_token_accuracy": 0.19677528142929077,
      "num_tokens": 59757205.0,
      "step": 28800
    },
    {
      "entropy": 5.32027235031128,
      "epoch": 2.612935413642961,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004318109529474411,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2033400610089302,
      "num_tokens": 59767691.0,
      "step": 28805
    },
    {
      "entropy": 5.392648935317993,
      "epoch": 2.6133889695210453,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004317877475083456,
      "loss": 5.0832,
      "mean_token_accuracy": 0.19693247228860855,
      "num_tokens": 59778618.0,
      "step": 28810
    },
    {
      "entropy": 5.451924419403076,
      "epoch": 2.613842525399129,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043176453882683955,
      "loss": 5.1108,
      "mean_token_accuracy": 0.19042819142341613,
      "num_tokens": 59789423.0,
      "step": 28815
    },
    {
      "entropy": 5.497757434844971,
      "epoch": 2.6142960812772134,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043174132690340287,
      "loss": 5.1664,
      "mean_token_accuracy": 0.1804507553577423,
      "num_tokens": 59800021.0,
      "step": 28820
    },
    {
      "entropy": 5.398585605621338,
      "epoch": 2.6147496371552976,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004317181117385156,
      "loss": 5.0262,
      "mean_token_accuracy": 0.19479135870933534,
      "num_tokens": 59810006.0,
      "step": 28825
    },
    {
      "entropy": 5.388393068313599,
      "epoch": 2.6152031930333814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043169489333265795,
      "loss": 5.0678,
      "mean_token_accuracy": 0.1934066191315651,
      "num_tokens": 59819505.0,
      "step": 28830
    },
    {
      "entropy": 5.419868278503418,
      "epoch": 2.6156567489114657,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043167167168630977,
      "loss": 5.151,
      "mean_token_accuracy": 0.19172338396310806,
      "num_tokens": 59830439.0,
      "step": 28835
    },
    {
      "entropy": 5.404328680038452,
      "epoch": 2.61611030478955,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004316484467999516,
      "loss": 5.0651,
      "mean_token_accuracy": 0.1979897975921631,
      "num_tokens": 59841281.0,
      "step": 28840
    },
    {
      "entropy": 5.453585958480835,
      "epoch": 2.6165638606676342,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043162521867406376,
      "loss": 5.1273,
      "mean_token_accuracy": 0.187883959710598,
      "num_tokens": 59853581.0,
      "step": 28845
    },
    {
      "entropy": 5.500374460220337,
      "epoch": 2.6170174165457185,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004316019873091263,
      "loss": 5.1267,
      "mean_token_accuracy": 0.19512415379285813,
      "num_tokens": 59864787.0,
      "step": 28850
    },
    {
      "entropy": 5.406511545181274,
      "epoch": 2.6174709724238028,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004315787527056199,
      "loss": 5.0822,
      "mean_token_accuracy": 0.1996711015701294,
      "num_tokens": 59875244.0,
      "step": 28855
    },
    {
      "entropy": 5.365839195251465,
      "epoch": 2.617924528301887,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004315555148640251,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1951599106192589,
      "num_tokens": 59886535.0,
      "step": 28860
    },
    {
      "entropy": 5.403475713729859,
      "epoch": 2.618378084179971,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004315322737848222,
      "loss": 5.0557,
      "mean_token_accuracy": 0.19702994376420974,
      "num_tokens": 59896298.0,
      "step": 28865
    },
    {
      "entropy": 5.317899942398071,
      "epoch": 2.618831640058055,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000431509029468492,
      "loss": 5.0265,
      "mean_token_accuracy": 0.201338329911232,
      "num_tokens": 59907010.0,
      "step": 28870
    },
    {
      "entropy": 5.457853746414185,
      "epoch": 2.6192851959361394,
      "grad_norm": 1.25,
      "learning_rate": 0.0004314857819155153,
      "loss": 5.0353,
      "mean_token_accuracy": 0.2052634373307228,
      "num_tokens": 59916247.0,
      "step": 28875
    },
    {
      "entropy": 5.3326927661895756,
      "epoch": 2.6197387518142237,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043146253112637267,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20580073148012162,
      "num_tokens": 59926680.0,
      "step": 28880
    },
    {
      "entropy": 5.355069923400879,
      "epoch": 2.6201923076923075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043143927710154504,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2108958899974823,
      "num_tokens": 59936346.0,
      "step": 28885
    },
    {
      "entropy": 5.4029919624328615,
      "epoch": 2.6206458635703918,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004314160198415133,
      "loss": 5.086,
      "mean_token_accuracy": 0.20173701345920564,
      "num_tokens": 59947007.0,
      "step": 28890
    },
    {
      "entropy": 5.438334178924561,
      "epoch": 2.621099419448476,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043139275934675844,
      "loss": 5.1156,
      "mean_token_accuracy": 0.19181872457265853,
      "num_tokens": 59957123.0,
      "step": 28895
    },
    {
      "entropy": 5.392713499069214,
      "epoch": 2.6215529753265603,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004313694956177613,
      "loss": 5.093,
      "mean_token_accuracy": 0.2022766277194023,
      "num_tokens": 59966566.0,
      "step": 28900
    },
    {
      "entropy": 5.3785864353179935,
      "epoch": 2.6220065312046446,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043134622865500334,
      "loss": 5.0558,
      "mean_token_accuracy": 0.19994070380926132,
      "num_tokens": 59976065.0,
      "step": 28905
    },
    {
      "entropy": 5.4725048542022705,
      "epoch": 2.622460087082729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004313229584589655,
      "loss": 5.0664,
      "mean_token_accuracy": 0.19647751599550248,
      "num_tokens": 59986201.0,
      "step": 28910
    },
    {
      "entropy": 5.3776157855987545,
      "epoch": 2.6229136429608126,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431299685030129,
      "loss": 5.0746,
      "mean_token_accuracy": 0.19658943712711335,
      "num_tokens": 59996574.0,
      "step": 28915
    },
    {
      "entropy": 5.403197431564331,
      "epoch": 2.623367198838897,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043127640836897516,
      "loss": 5.091,
      "mean_token_accuracy": 0.1968645617365837,
      "num_tokens": 60006082.0,
      "step": 28920
    },
    {
      "entropy": 5.410541200637818,
      "epoch": 2.623820754716981,
      "grad_norm": 1.125,
      "learning_rate": 0.0004312531284759854,
      "loss": 5.0669,
      "mean_token_accuracy": 0.19631859958171843,
      "num_tokens": 60015483.0,
      "step": 28925
    },
    {
      "entropy": 5.385684061050415,
      "epoch": 2.6242743105950654,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00043122984535164116,
      "loss": 5.0037,
      "mean_token_accuracy": 0.20849861800670624,
      "num_tokens": 60025984.0,
      "step": 28930
    },
    {
      "entropy": 5.447122526168823,
      "epoch": 2.6247278664731493,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043120655899642386,
      "loss": 5.0676,
      "mean_token_accuracy": 0.19414779543876648,
      "num_tokens": 60036109.0,
      "step": 28935
    },
    {
      "entropy": 5.501953887939453,
      "epoch": 2.6251814223512335,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043118326941081513,
      "loss": 5.1349,
      "mean_token_accuracy": 0.18942345380783082,
      "num_tokens": 60046546.0,
      "step": 28940
    },
    {
      "entropy": 5.434204196929931,
      "epoch": 2.625634978229318,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004311599765952966,
      "loss": 5.1216,
      "mean_token_accuracy": 0.199664805829525,
      "num_tokens": 60056768.0,
      "step": 28945
    },
    {
      "entropy": 5.465752649307251,
      "epoch": 2.626088534107402,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043113668055034995,
      "loss": 5.1756,
      "mean_token_accuracy": 0.19442594647407532,
      "num_tokens": 60066962.0,
      "step": 28950
    },
    {
      "entropy": 5.444908618927002,
      "epoch": 2.6265420899854863,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004311133812764569,
      "loss": 5.0659,
      "mean_token_accuracy": 0.1959977701306343,
      "num_tokens": 60076477.0,
      "step": 28955
    },
    {
      "entropy": 5.357529354095459,
      "epoch": 2.6269956458635706,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043109007877409946,
      "loss": 4.9524,
      "mean_token_accuracy": 0.2063591867685318,
      "num_tokens": 60088060.0,
      "step": 28960
    },
    {
      "entropy": 5.411446666717529,
      "epoch": 2.6274492017416544,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043106677304375935,
      "loss": 5.0789,
      "mean_token_accuracy": 0.19770629405975343,
      "num_tokens": 60098677.0,
      "step": 28965
    },
    {
      "entropy": 5.428258466720581,
      "epoch": 2.6279027576197387,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004310434640859186,
      "loss": 5.1239,
      "mean_token_accuracy": 0.1919982612133026,
      "num_tokens": 60109054.0,
      "step": 28970
    },
    {
      "entropy": 5.3919861793518065,
      "epoch": 2.628356313497823,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004310201519010592,
      "loss": 5.007,
      "mean_token_accuracy": 0.20709054917097092,
      "num_tokens": 60119951.0,
      "step": 28975
    },
    {
      "entropy": 5.2599201679229735,
      "epoch": 2.628809869375907,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043099683648966335,
      "loss": 4.9367,
      "mean_token_accuracy": 0.20691928416490554,
      "num_tokens": 60130026.0,
      "step": 28980
    },
    {
      "entropy": 5.2954676151275635,
      "epoch": 2.629263425253991,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004309735178522132,
      "loss": 4.9705,
      "mean_token_accuracy": 0.20936646759510041,
      "num_tokens": 60140748.0,
      "step": 28985
    },
    {
      "entropy": 5.391308546066284,
      "epoch": 2.6297169811320753,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004309501959891909,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20332566201686858,
      "num_tokens": 60151042.0,
      "step": 28990
    },
    {
      "entropy": 5.425695943832397,
      "epoch": 2.6301705370101596,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043092687090107883,
      "loss": 5.049,
      "mean_token_accuracy": 0.1945054277777672,
      "num_tokens": 60160758.0,
      "step": 28995
    },
    {
      "entropy": 5.380381155014038,
      "epoch": 2.630624092888244,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043090354258835934,
      "loss": 5.0738,
      "mean_token_accuracy": 0.19239141643047333,
      "num_tokens": 60170249.0,
      "step": 29000
    },
    {
      "entropy": 5.386873817443847,
      "epoch": 2.631077648766328,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043088021105151483,
      "loss": 5.0832,
      "mean_token_accuracy": 0.19918040931224823,
      "num_tokens": 60181216.0,
      "step": 29005
    },
    {
      "entropy": 5.388864374160766,
      "epoch": 2.6315312046444124,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004308568762910279,
      "loss": 5.033,
      "mean_token_accuracy": 0.20024227499961852,
      "num_tokens": 60190559.0,
      "step": 29010
    },
    {
      "entropy": 5.483091735839844,
      "epoch": 2.6319847605224966,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043083353830738097,
      "loss": 5.1709,
      "mean_token_accuracy": 0.18734700381755828,
      "num_tokens": 60201268.0,
      "step": 29015
    },
    {
      "entropy": 5.336070013046265,
      "epoch": 2.6324383164005805,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043081019710105683,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2044358178973198,
      "num_tokens": 60210907.0,
      "step": 29020
    },
    {
      "entropy": 5.361450433731079,
      "epoch": 2.6328918722786647,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004307868526725381,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20446448922157287,
      "num_tokens": 60220298.0,
      "step": 29025
    },
    {
      "entropy": 5.320916748046875,
      "epoch": 2.633345428156749,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004307635050223075,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2078732043504715,
      "num_tokens": 60230811.0,
      "step": 29030
    },
    {
      "entropy": 5.400971508026123,
      "epoch": 2.6337989840348333,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000430740154150848,
      "loss": 4.964,
      "mean_token_accuracy": 0.20207771211862563,
      "num_tokens": 60241211.0,
      "step": 29035
    },
    {
      "entropy": 5.322976160049438,
      "epoch": 2.634252539912917,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004307168000586423,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20417421609163283,
      "num_tokens": 60251506.0,
      "step": 29040
    },
    {
      "entropy": 5.311602973937989,
      "epoch": 2.6347060957910013,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004306934427461736,
      "loss": 5.0066,
      "mean_token_accuracy": 0.21196245700120925,
      "num_tokens": 60261416.0,
      "step": 29045
    },
    {
      "entropy": 5.419220590591431,
      "epoch": 2.6351596516690856,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004306700822139248,
      "loss": 5.0694,
      "mean_token_accuracy": 0.1990858420729637,
      "num_tokens": 60272315.0,
      "step": 29050
    },
    {
      "entropy": 5.40738697052002,
      "epoch": 2.63561320754717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000430646718462379,
      "loss": 5.1358,
      "mean_token_accuracy": 0.19730154424905777,
      "num_tokens": 60282722.0,
      "step": 29055
    },
    {
      "entropy": 5.362531137466431,
      "epoch": 2.636066763425254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004306233514920194,
      "loss": 5.0364,
      "mean_token_accuracy": 0.195218725502491,
      "num_tokens": 60292636.0,
      "step": 29060
    },
    {
      "entropy": 5.386047744750977,
      "epoch": 2.6365203193033384,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043059998130332933,
      "loss": 5.048,
      "mean_token_accuracy": 0.20121754109859466,
      "num_tokens": 60302979.0,
      "step": 29065
    },
    {
      "entropy": 5.4258135795593265,
      "epoch": 2.6369738751814222,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043057660789679194,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19930365532636643,
      "num_tokens": 60313181.0,
      "step": 29070
    },
    {
      "entropy": 5.449718713760376,
      "epoch": 2.6374274310595065,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004305532312728906,
      "loss": 5.0204,
      "mean_token_accuracy": 0.1929544135928154,
      "num_tokens": 60323401.0,
      "step": 29075
    },
    {
      "entropy": 5.42025203704834,
      "epoch": 2.6378809869375908,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004305298514321088,
      "loss": 5.1323,
      "mean_token_accuracy": 0.1945386439561844,
      "num_tokens": 60333478.0,
      "step": 29080
    },
    {
      "entropy": 5.37011022567749,
      "epoch": 2.638334542815675,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043050646837493014,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20285616517066957,
      "num_tokens": 60343980.0,
      "step": 29085
    },
    {
      "entropy": 5.38485689163208,
      "epoch": 2.638788098693759,
      "grad_norm": 1.125,
      "learning_rate": 0.0004304830821018381,
      "loss": 5.0234,
      "mean_token_accuracy": 0.19826746731996536,
      "num_tokens": 60354262.0,
      "step": 29090
    },
    {
      "entropy": 5.4336809635162355,
      "epoch": 2.639241654571843,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004304596926133162,
      "loss": 5.1034,
      "mean_token_accuracy": 0.19824122339487077,
      "num_tokens": 60364137.0,
      "step": 29095
    },
    {
      "entropy": 5.363964319229126,
      "epoch": 2.6396952104499274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043043629990984823,
      "loss": 5.1027,
      "mean_token_accuracy": 0.19456753581762315,
      "num_tokens": 60374528.0,
      "step": 29100
    },
    {
      "entropy": 5.386783504486084,
      "epoch": 2.6401487663280117,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043041290399191797,
      "loss": 5.0851,
      "mean_token_accuracy": 0.2028406575322151,
      "num_tokens": 60384880.0,
      "step": 29105
    },
    {
      "entropy": 5.433948564529419,
      "epoch": 2.640602322206096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004303895048600093,
      "loss": 5.0209,
      "mean_token_accuracy": 0.1957051932811737,
      "num_tokens": 60395543.0,
      "step": 29110
    },
    {
      "entropy": 5.394303035736084,
      "epoch": 2.64105587808418,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004303661025146061,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20056162178516387,
      "num_tokens": 60405432.0,
      "step": 29115
    },
    {
      "entropy": 5.430849361419678,
      "epoch": 2.641509433962264,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004303426969561922,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1890409007668495,
      "num_tokens": 60416360.0,
      "step": 29120
    },
    {
      "entropy": 5.425290632247925,
      "epoch": 2.6419629898403483,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004303192881852518,
      "loss": 5.0743,
      "mean_token_accuracy": 0.19614432603120804,
      "num_tokens": 60426371.0,
      "step": 29125
    },
    {
      "entropy": 5.4100257396698,
      "epoch": 2.6424165457184325,
      "grad_norm": 1.53125,
      "learning_rate": 0.00043029587620226897,
      "loss": 5.071,
      "mean_token_accuracy": 0.19666725099086763,
      "num_tokens": 60437505.0,
      "step": 29130
    },
    {
      "entropy": 5.38546986579895,
      "epoch": 2.642870101596517,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004302724610077278,
      "loss": 5.0745,
      "mean_token_accuracy": 0.19494210183620453,
      "num_tokens": 60448170.0,
      "step": 29135
    },
    {
      "entropy": 5.453001546859741,
      "epoch": 2.6433236574746006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004302490426021126,
      "loss": 5.1531,
      "mean_token_accuracy": 0.18887126445770264,
      "num_tokens": 60458776.0,
      "step": 29140
    },
    {
      "entropy": 5.388426780700684,
      "epoch": 2.643777213352685,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004302256209859076,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20911710262298583,
      "num_tokens": 60468279.0,
      "step": 29145
    },
    {
      "entropy": 5.36911039352417,
      "epoch": 2.644230769230769,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043020219615959727,
      "loss": 5.124,
      "mean_token_accuracy": 0.19389037787914276,
      "num_tokens": 60478507.0,
      "step": 29150
    },
    {
      "entropy": 5.529068231582642,
      "epoch": 2.6446843251088534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004301787681236659,
      "loss": 5.1736,
      "mean_token_accuracy": 0.18983519971370696,
      "num_tokens": 60489016.0,
      "step": 29155
    },
    {
      "entropy": 5.515232276916504,
      "epoch": 2.6451378809869377,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004301553368785981,
      "loss": 5.1156,
      "mean_token_accuracy": 0.19229382425546646,
      "num_tokens": 60500169.0,
      "step": 29160
    },
    {
      "entropy": 5.336970615386963,
      "epoch": 2.645591436865022,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004301319024248784,
      "loss": 5.0099,
      "mean_token_accuracy": 0.2014343723654747,
      "num_tokens": 60511109.0,
      "step": 29165
    },
    {
      "entropy": 5.3483765602111815,
      "epoch": 2.6460449927431062,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043010846476299145,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20195322483778,
      "num_tokens": 60521735.0,
      "step": 29170
    },
    {
      "entropy": 5.45266432762146,
      "epoch": 2.64649854862119,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004300850238934219,
      "loss": 5.1053,
      "mean_token_accuracy": 0.1949572056531906,
      "num_tokens": 60532071.0,
      "step": 29175
    },
    {
      "entropy": 5.4453325271606445,
      "epoch": 2.6469521044992743,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043006157981665446,
      "loss": 5.0562,
      "mean_token_accuracy": 0.2034536436200142,
      "num_tokens": 60541909.0,
      "step": 29180
    },
    {
      "entropy": 5.3410557270050045,
      "epoch": 2.6474056603773586,
      "grad_norm": 1.25,
      "learning_rate": 0.00043003813253317415,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20621223747730255,
      "num_tokens": 60551361.0,
      "step": 29185
    },
    {
      "entropy": 5.347345542907715,
      "epoch": 2.6478592162554424,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043001468204346564,
      "loss": 5.0818,
      "mean_token_accuracy": 0.19990906864404678,
      "num_tokens": 60562112.0,
      "step": 29190
    },
    {
      "entropy": 5.442684745788574,
      "epoch": 2.6483127721335267,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042999122834801406,
      "loss": 5.0944,
      "mean_token_accuracy": 0.19456730484962464,
      "num_tokens": 60572827.0,
      "step": 29195
    },
    {
      "entropy": 5.369096946716309,
      "epoch": 2.648766328011611,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004299677714473044,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2036827862262726,
      "num_tokens": 60582222.0,
      "step": 29200
    },
    {
      "entropy": 5.398920011520386,
      "epoch": 2.649219883889695,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004299443113418217,
      "loss": 5.1012,
      "mean_token_accuracy": 0.20191438645124435,
      "num_tokens": 60592943.0,
      "step": 29205
    },
    {
      "entropy": 5.4006226539611815,
      "epoch": 2.6496734397677795,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042992084803205124,
      "loss": 5.0759,
      "mean_token_accuracy": 0.19711816310882568,
      "num_tokens": 60603206.0,
      "step": 29210
    },
    {
      "entropy": 5.399846410751342,
      "epoch": 2.6501269956458637,
      "grad_norm": 1.1875,
      "learning_rate": 0.000429897381518478,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20322587639093398,
      "num_tokens": 60614776.0,
      "step": 29215
    },
    {
      "entropy": 5.401335668563843,
      "epoch": 2.650580551523948,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042987391180158757,
      "loss": 5.0307,
      "mean_token_accuracy": 0.19667024165391922,
      "num_tokens": 60625914.0,
      "step": 29220
    },
    {
      "entropy": 5.371406507492066,
      "epoch": 2.651034107402032,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042985043888186516,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20150816291570664,
      "num_tokens": 60636934.0,
      "step": 29225
    },
    {
      "entropy": 5.371162080764771,
      "epoch": 2.651487663280116,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004298269627597962,
      "loss": 5.092,
      "mean_token_accuracy": 0.19261346161365508,
      "num_tokens": 60647707.0,
      "step": 29230
    },
    {
      "entropy": 5.319137096405029,
      "epoch": 2.6519412191582004,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042980348343586624,
      "loss": 5.0086,
      "mean_token_accuracy": 0.1996026322245598,
      "num_tokens": 60658691.0,
      "step": 29235
    },
    {
      "entropy": 5.376399087905884,
      "epoch": 2.6523947750362846,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042978000091056074,
      "loss": 5.0983,
      "mean_token_accuracy": 0.19866568893194197,
      "num_tokens": 60668479.0,
      "step": 29240
    },
    {
      "entropy": 5.380235099792481,
      "epoch": 2.6528483309143684,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004297565151843653,
      "loss": 4.9623,
      "mean_token_accuracy": 0.20203416049480438,
      "num_tokens": 60679158.0,
      "step": 29245
    },
    {
      "entropy": 5.4198463439941404,
      "epoch": 2.6533018867924527,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004297330262577659,
      "loss": 5.1749,
      "mean_token_accuracy": 0.1919507384300232,
      "num_tokens": 60689363.0,
      "step": 29250
    },
    {
      "entropy": 5.353726053237915,
      "epoch": 2.653755442670537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004297095341312479,
      "loss": 4.9445,
      "mean_token_accuracy": 0.20843117833137512,
      "num_tokens": 60699650.0,
      "step": 29255
    },
    {
      "entropy": 5.350832319259643,
      "epoch": 2.6542089985486212,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004296860388052974,
      "loss": 5.0267,
      "mean_token_accuracy": 0.19964946061372757,
      "num_tokens": 60709921.0,
      "step": 29260
    },
    {
      "entropy": 5.393980932235718,
      "epoch": 2.6546625544267055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042966254028040015,
      "loss": 5.0268,
      "mean_token_accuracy": 0.20660987943410875,
      "num_tokens": 60720112.0,
      "step": 29265
    },
    {
      "entropy": 5.327514553070069,
      "epoch": 2.65511611030479,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042963903855704216,
      "loss": 4.9751,
      "mean_token_accuracy": 0.1972302868962288,
      "num_tokens": 60730604.0,
      "step": 29270
    },
    {
      "entropy": 5.316270971298218,
      "epoch": 2.6555696661828736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042961553363570946,
      "loss": 5.0079,
      "mean_token_accuracy": 0.20136940032243728,
      "num_tokens": 60740896.0,
      "step": 29275
    },
    {
      "entropy": 5.415946292877197,
      "epoch": 2.656023222060958,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004295920255168881,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2006162941455841,
      "num_tokens": 60751771.0,
      "step": 29280
    },
    {
      "entropy": 5.439013051986694,
      "epoch": 2.656476777939042,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042956851420106423,
      "loss": 5.0825,
      "mean_token_accuracy": 0.1956859201192856,
      "num_tokens": 60762802.0,
      "step": 29285
    },
    {
      "entropy": 5.324782991409302,
      "epoch": 2.6569303338171264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004295449996887242,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20423976182937623,
      "num_tokens": 60774102.0,
      "step": 29290
    },
    {
      "entropy": 5.3456377506256105,
      "epoch": 2.65738388969521,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004295214819803541,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20100480169057847,
      "num_tokens": 60785383.0,
      "step": 29295
    },
    {
      "entropy": 5.374890470504761,
      "epoch": 2.6578374455732945,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042949796107644053,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20624749958515168,
      "num_tokens": 60795173.0,
      "step": 29300
    },
    {
      "entropy": 5.409754991531372,
      "epoch": 2.6582910014513788,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042947443697746956,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20446381121873855,
      "num_tokens": 60805621.0,
      "step": 29305
    },
    {
      "entropy": 5.463833236694336,
      "epoch": 2.658744557329463,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004294509096839279,
      "loss": 5.1817,
      "mean_token_accuracy": 0.1841423124074936,
      "num_tokens": 60816799.0,
      "step": 29310
    },
    {
      "entropy": 5.323073291778565,
      "epoch": 2.6591981132075473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042942737919630215,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20168149322271348,
      "num_tokens": 60826827.0,
      "step": 29315
    },
    {
      "entropy": 5.397406959533692,
      "epoch": 2.6596516690856316,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042940384551507876,
      "loss": 5.0608,
      "mean_token_accuracy": 0.1949218064546585,
      "num_tokens": 60836567.0,
      "step": 29320
    },
    {
      "entropy": 5.379178905487061,
      "epoch": 2.6601052249637154,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004293803086407447,
      "loss": 5.0565,
      "mean_token_accuracy": 0.1927633836865425,
      "num_tokens": 60846723.0,
      "step": 29325
    },
    {
      "entropy": 5.4065728187561035,
      "epoch": 2.6605587808417996,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004293567685737863,
      "loss": 5.1443,
      "mean_token_accuracy": 0.19115439355373381,
      "num_tokens": 60857434.0,
      "step": 29330
    },
    {
      "entropy": 5.436360931396484,
      "epoch": 2.661012336719884,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042933322531469074,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19862047135829924,
      "num_tokens": 60868780.0,
      "step": 29335
    },
    {
      "entropy": 5.473215198516845,
      "epoch": 2.661465892597968,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042930967886394456,
      "loss": 5.1123,
      "mean_token_accuracy": 0.19893196523189544,
      "num_tokens": 60879037.0,
      "step": 29340
    },
    {
      "entropy": 5.3245964527130125,
      "epoch": 2.661919448476052,
      "grad_norm": 1.03125,
      "learning_rate": 0.000429286129222035,
      "loss": 5.0244,
      "mean_token_accuracy": 0.20444056987762452,
      "num_tokens": 60889543.0,
      "step": 29345
    },
    {
      "entropy": 5.319226312637329,
      "epoch": 2.6623730043541363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000429262576389449,
      "loss": 5.0348,
      "mean_token_accuracy": 0.19940050989389418,
      "num_tokens": 60900141.0,
      "step": 29350
    },
    {
      "entropy": 5.430813217163086,
      "epoch": 2.6628265602322205,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004292390203666736,
      "loss": 5.0828,
      "mean_token_accuracy": 0.2022681638598442,
      "num_tokens": 60910244.0,
      "step": 29355
    },
    {
      "entropy": 5.363577079772949,
      "epoch": 2.663280116110305,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000429215461154196,
      "loss": 5.0043,
      "mean_token_accuracy": 0.2093103602528572,
      "num_tokens": 60919908.0,
      "step": 29360
    },
    {
      "entropy": 5.370628738403321,
      "epoch": 2.663733671988389,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042919189875250324,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20272407978773116,
      "num_tokens": 60929096.0,
      "step": 29365
    },
    {
      "entropy": 5.372285413742065,
      "epoch": 2.6641872278664733,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042916833316208284,
      "loss": 5.0427,
      "mean_token_accuracy": 0.19441961795091628,
      "num_tokens": 60939346.0,
      "step": 29370
    },
    {
      "entropy": 5.421276378631592,
      "epoch": 2.6646407837445576,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004291447643834219,
      "loss": 5.0714,
      "mean_token_accuracy": 0.19666676074266434,
      "num_tokens": 60950364.0,
      "step": 29375
    },
    {
      "entropy": 5.385446214675904,
      "epoch": 2.6650943396226414,
      "grad_norm": 1.171875,
      "learning_rate": 0.000429121192417008,
      "loss": 5.0394,
      "mean_token_accuracy": 0.2035601705312729,
      "num_tokens": 60960905.0,
      "step": 29380
    },
    {
      "entropy": 5.43964114189148,
      "epoch": 2.6655478955007257,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042909761726332864,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20789116770029067,
      "num_tokens": 60971281.0,
      "step": 29385
    },
    {
      "entropy": 5.360865879058838,
      "epoch": 2.66600145137881,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004290740389228712,
      "loss": 5.0766,
      "mean_token_accuracy": 0.19949969202280043,
      "num_tokens": 60982522.0,
      "step": 29390
    },
    {
      "entropy": 5.374077224731446,
      "epoch": 2.6664550072568938,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042905045739612347,
      "loss": 5.0739,
      "mean_token_accuracy": 0.1971197634935379,
      "num_tokens": 60992738.0,
      "step": 29395
    },
    {
      "entropy": 5.474529838562011,
      "epoch": 2.666908563134978,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042902687268357297,
      "loss": 5.1729,
      "mean_token_accuracy": 0.19621514678001403,
      "num_tokens": 61003665.0,
      "step": 29400
    },
    {
      "entropy": 5.537272214889526,
      "epoch": 2.6673621190130623,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042900328478570745,
      "loss": 5.1869,
      "mean_token_accuracy": 0.1940956622362137,
      "num_tokens": 61014663.0,
      "step": 29405
    },
    {
      "entropy": 5.389990663528442,
      "epoch": 2.6678156748911466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004289796937030148,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2043056532740593,
      "num_tokens": 61024916.0,
      "step": 29410
    },
    {
      "entropy": 5.299099159240723,
      "epoch": 2.668269230769231,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042895609943598286,
      "loss": 4.9885,
      "mean_token_accuracy": 0.19833351969718932,
      "num_tokens": 61035043.0,
      "step": 29415
    },
    {
      "entropy": 5.387543058395385,
      "epoch": 2.668722786647315,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004289325019850995,
      "loss": 5.1096,
      "mean_token_accuracy": 0.19310559928417206,
      "num_tokens": 61046153.0,
      "step": 29420
    },
    {
      "entropy": 5.432633876800537,
      "epoch": 2.6691763425253994,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004289089013508528,
      "loss": 5.0229,
      "mean_token_accuracy": 0.20363078266382217,
      "num_tokens": 61057009.0,
      "step": 29425
    },
    {
      "entropy": 5.469908809661865,
      "epoch": 2.669629898403483,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004288852975337308,
      "loss": 5.1834,
      "mean_token_accuracy": 0.18653122186660767,
      "num_tokens": 61068514.0,
      "step": 29430
    },
    {
      "entropy": 5.412503576278686,
      "epoch": 2.6700834542815675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042886169053422165,
      "loss": 5.0649,
      "mean_token_accuracy": 0.2000389501452446,
      "num_tokens": 61078902.0,
      "step": 29435
    },
    {
      "entropy": 5.392664432525635,
      "epoch": 2.6705370101596517,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004288380803528135,
      "loss": 5.0854,
      "mean_token_accuracy": 0.2012828603386879,
      "num_tokens": 61089216.0,
      "step": 29440
    },
    {
      "entropy": 5.3389026641845705,
      "epoch": 2.670990566037736,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042881446698999465,
      "loss": 4.9848,
      "mean_token_accuracy": 0.20387418568134308,
      "num_tokens": 61098470.0,
      "step": 29445
    },
    {
      "entropy": 5.501800155639648,
      "epoch": 2.67144412191582,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004287908504462534,
      "loss": 5.2037,
      "mean_token_accuracy": 0.20261537581682204,
      "num_tokens": 61108694.0,
      "step": 29450
    },
    {
      "entropy": 5.369113302230835,
      "epoch": 2.671897677793904,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042876723072207825,
      "loss": 5.0388,
      "mean_token_accuracy": 0.2019443228840828,
      "num_tokens": 61119389.0,
      "step": 29455
    },
    {
      "entropy": 5.468770694732666,
      "epoch": 2.6723512336719883,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042874360781795757,
      "loss": 5.1461,
      "mean_token_accuracy": 0.19186817854642868,
      "num_tokens": 61130491.0,
      "step": 29460
    },
    {
      "entropy": 5.428310108184815,
      "epoch": 2.6728047895500726,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042871998173437987,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20356832295656205,
      "num_tokens": 61140404.0,
      "step": 29465
    },
    {
      "entropy": 5.43123140335083,
      "epoch": 2.673258345428157,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042869635247183373,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19860687255859374,
      "num_tokens": 61150777.0,
      "step": 29470
    },
    {
      "entropy": 5.395602560043335,
      "epoch": 2.673711901306241,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042867272003080793,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20620987862348555,
      "num_tokens": 61160685.0,
      "step": 29475
    },
    {
      "entropy": 5.468594264984131,
      "epoch": 2.674165457184325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004286490844117911,
      "loss": 5.0559,
      "mean_token_accuracy": 0.1977478563785553,
      "num_tokens": 61170553.0,
      "step": 29480
    },
    {
      "entropy": 5.451732444763183,
      "epoch": 2.6746190130624092,
      "grad_norm": 1.25,
      "learning_rate": 0.00042862544561527203,
      "loss": 5.0791,
      "mean_token_accuracy": 0.19771808981895447,
      "num_tokens": 61181396.0,
      "step": 29485
    },
    {
      "entropy": 5.369629859924316,
      "epoch": 2.6750725689404935,
      "grad_norm": 1.125,
      "learning_rate": 0.0004286018036417396,
      "loss": 5.0898,
      "mean_token_accuracy": 0.1982921838760376,
      "num_tokens": 61192315.0,
      "step": 29490
    },
    {
      "entropy": 5.38681116104126,
      "epoch": 2.6755261248185778,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042857815849168277,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20519535094499589,
      "num_tokens": 61202719.0,
      "step": 29495
    },
    {
      "entropy": 5.312905216217041,
      "epoch": 2.6759796806966616,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004285545101655904,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21166519224643707,
      "num_tokens": 61212703.0,
      "step": 29500
    },
    {
      "entropy": 5.302706384658814,
      "epoch": 2.676433236574746,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004285308586639517,
      "loss": 4.9044,
      "mean_token_accuracy": 0.21491365134716034,
      "num_tokens": 61222406.0,
      "step": 29505
    },
    {
      "entropy": 5.371212434768677,
      "epoch": 2.67688679245283,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042850720398725575,
      "loss": 5.1384,
      "mean_token_accuracy": 0.19220755100250245,
      "num_tokens": 61232758.0,
      "step": 29510
    },
    {
      "entropy": 5.3618083000183105,
      "epoch": 2.6773403483309144,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042848354613599164,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2065221682190895,
      "num_tokens": 61243715.0,
      "step": 29515
    },
    {
      "entropy": 5.435568237304688,
      "epoch": 2.6777939042089987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042845988511064863,
      "loss": 5.0634,
      "mean_token_accuracy": 0.20038834363222122,
      "num_tokens": 61254669.0,
      "step": 29520
    },
    {
      "entropy": 5.365223121643067,
      "epoch": 2.678247460087083,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042843622091171625,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20495359599590302,
      "num_tokens": 61264377.0,
      "step": 29525
    },
    {
      "entropy": 5.373605394363404,
      "epoch": 2.6787010159651667,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004284125535396836,
      "loss": 5.0471,
      "mean_token_accuracy": 0.1964878559112549,
      "num_tokens": 61275348.0,
      "step": 29530
    },
    {
      "entropy": 5.334469509124756,
      "epoch": 2.679154571843251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042838888299504024,
      "loss": 4.9108,
      "mean_token_accuracy": 0.2116192027926445,
      "num_tokens": 61285579.0,
      "step": 29535
    },
    {
      "entropy": 5.360258007049561,
      "epoch": 2.6796081277213353,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042836520927827575,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20019187927246093,
      "num_tokens": 61296037.0,
      "step": 29540
    },
    {
      "entropy": 5.375100612640381,
      "epoch": 2.6800616835994195,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004283415323898797,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21222453117370604,
      "num_tokens": 61305882.0,
      "step": 29545
    },
    {
      "entropy": 5.362612199783325,
      "epoch": 2.6805152394775034,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042831785233034156,
      "loss": 4.9843,
      "mean_token_accuracy": 0.20564391762018203,
      "num_tokens": 61317117.0,
      "step": 29550
    },
    {
      "entropy": 5.396396827697754,
      "epoch": 2.6809687953555876,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042829416910015116,
      "loss": 5.0849,
      "mean_token_accuracy": 0.19745588451623916,
      "num_tokens": 61328190.0,
      "step": 29555
    },
    {
      "entropy": 5.3447518825531,
      "epoch": 2.681422351233672,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042827048269979834,
      "loss": 5.052,
      "mean_token_accuracy": 0.20459520518779756,
      "num_tokens": 61337978.0,
      "step": 29560
    },
    {
      "entropy": 5.42297511100769,
      "epoch": 2.681875907111756,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004282467931297729,
      "loss": 5.1129,
      "mean_token_accuracy": 0.19447076469659805,
      "num_tokens": 61348523.0,
      "step": 29565
    },
    {
      "entropy": 5.377737712860108,
      "epoch": 2.6823294629898404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004282231003905647,
      "loss": 4.9682,
      "mean_token_accuracy": 0.2079920008778572,
      "num_tokens": 61359250.0,
      "step": 29570
    },
    {
      "entropy": 5.3170124053955075,
      "epoch": 2.6827830188679247,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042819940448266376,
      "loss": 4.9557,
      "mean_token_accuracy": 0.20387499779462814,
      "num_tokens": 61369999.0,
      "step": 29575
    },
    {
      "entropy": 5.410287380218506,
      "epoch": 2.683236574746009,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042817570540656003,
      "loss": 5.078,
      "mean_token_accuracy": 0.19439928680658342,
      "num_tokens": 61380279.0,
      "step": 29580
    },
    {
      "entropy": 5.361128282546997,
      "epoch": 2.683690130624093,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004281520031627438,
      "loss": 5.055,
      "mean_token_accuracy": 0.2052997976541519,
      "num_tokens": 61389794.0,
      "step": 29585
    },
    {
      "entropy": 5.387273597717285,
      "epoch": 2.684143686502177,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042812829775170497,
      "loss": 5.177,
      "mean_token_accuracy": 0.19572146385908126,
      "num_tokens": 61401136.0,
      "step": 29590
    },
    {
      "entropy": 5.404409456253052,
      "epoch": 2.6845972423802613,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042810458917393404,
      "loss": 5.1216,
      "mean_token_accuracy": 0.19688222110271453,
      "num_tokens": 61412830.0,
      "step": 29595
    },
    {
      "entropy": 5.407378196716309,
      "epoch": 2.6850507982583456,
      "grad_norm": 3.1875,
      "learning_rate": 0.0004280808774299211,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21036689579486847,
      "num_tokens": 61422168.0,
      "step": 29600
    },
    {
      "entropy": 5.328280591964722,
      "epoch": 2.6855043541364294,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042805716252015665,
      "loss": 4.9949,
      "mean_token_accuracy": 0.2063021868467331,
      "num_tokens": 61432471.0,
      "step": 29605
    },
    {
      "entropy": 5.377523756027221,
      "epoch": 2.6859579100145137,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042803344444513106,
      "loss": 5.0354,
      "mean_token_accuracy": 0.20649911612272262,
      "num_tokens": 61443402.0,
      "step": 29610
    },
    {
      "entropy": 5.422032165527344,
      "epoch": 2.686411465892598,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004280097232053349,
      "loss": 5.0186,
      "mean_token_accuracy": 0.1980150669813156,
      "num_tokens": 61454465.0,
      "step": 29615
    },
    {
      "entropy": 5.389312887191773,
      "epoch": 2.686865021770682,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004279859988012586,
      "loss": 5.0656,
      "mean_token_accuracy": 0.20141032189130784,
      "num_tokens": 61464671.0,
      "step": 29620
    },
    {
      "entropy": 5.353705406188965,
      "epoch": 2.6873185776487665,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004279622712333929,
      "loss": 5.027,
      "mean_token_accuracy": 0.2074280709028244,
      "num_tokens": 61474940.0,
      "step": 29625
    },
    {
      "entropy": 5.384049797058106,
      "epoch": 2.6877721335268507,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004279385405022284,
      "loss": 5.0636,
      "mean_token_accuracy": 0.2008373558521271,
      "num_tokens": 61485240.0,
      "step": 29630
    },
    {
      "entropy": 5.352193593978882,
      "epoch": 2.6882256894049346,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000427914806608256,
      "loss": 5.0029,
      "mean_token_accuracy": 0.20355838984251023,
      "num_tokens": 61495382.0,
      "step": 29635
    },
    {
      "entropy": 5.325720930099488,
      "epoch": 2.688679245283019,
      "grad_norm": 1.125,
      "learning_rate": 0.0004278910695519663,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2082790032029152,
      "num_tokens": 61505372.0,
      "step": 29640
    },
    {
      "entropy": 5.393371629714966,
      "epoch": 2.689132801161103,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042786732933385046,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20508154779672622,
      "num_tokens": 61516039.0,
      "step": 29645
    },
    {
      "entropy": 5.3500806331634525,
      "epoch": 2.6895863570391874,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004278435859543992,
      "loss": 5.1005,
      "mean_token_accuracy": 0.20176443755626677,
      "num_tokens": 61525435.0,
      "step": 29650
    },
    {
      "entropy": 5.410941934585571,
      "epoch": 2.690039912917271,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004278198394141037,
      "loss": 5.046,
      "mean_token_accuracy": 0.19835190922021867,
      "num_tokens": 61534714.0,
      "step": 29655
    },
    {
      "entropy": 5.48473801612854,
      "epoch": 2.6904934687953554,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042779608971345496,
      "loss": 5.1797,
      "mean_token_accuracy": 0.19354444295167922,
      "num_tokens": 61545841.0,
      "step": 29660
    },
    {
      "entropy": 5.338754272460937,
      "epoch": 2.6909470246734397,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042777233685294407,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2124696284532547,
      "num_tokens": 61555355.0,
      "step": 29665
    },
    {
      "entropy": 5.367818927764892,
      "epoch": 2.691400580551524,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042774858083306246,
      "loss": 5.074,
      "mean_token_accuracy": 0.19443438947200775,
      "num_tokens": 61565474.0,
      "step": 29670
    },
    {
      "entropy": 5.386504697799682,
      "epoch": 2.6918541364296082,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042772482165430117,
      "loss": 4.9736,
      "mean_token_accuracy": 0.20769481509923934,
      "num_tokens": 61575631.0,
      "step": 29675
    },
    {
      "entropy": 5.342767906188965,
      "epoch": 2.6923076923076925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042770105931715167,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20122816115617753,
      "num_tokens": 61586398.0,
      "step": 29680
    },
    {
      "entropy": 5.35707802772522,
      "epoch": 2.6927612481857763,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004276772938221054,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21171467304229735,
      "num_tokens": 61596630.0,
      "step": 29685
    },
    {
      "entropy": 5.41720495223999,
      "epoch": 2.6932148040638606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004276535251696537,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20652303993701934,
      "num_tokens": 61605965.0,
      "step": 29690
    },
    {
      "entropy": 5.4366669178009035,
      "epoch": 2.693668359941945,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004276297533602883,
      "loss": 5.1166,
      "mean_token_accuracy": 0.1934044450521469,
      "num_tokens": 61616973.0,
      "step": 29695
    },
    {
      "entropy": 5.449663162231445,
      "epoch": 2.694121915820029,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042760597839450057,
      "loss": 5.094,
      "mean_token_accuracy": 0.19788265824317933,
      "num_tokens": 61627420.0,
      "step": 29700
    },
    {
      "entropy": 5.329347419738769,
      "epoch": 2.694575471698113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004275822002727824,
      "loss": 5.0343,
      "mean_token_accuracy": 0.2023436188697815,
      "num_tokens": 61638315.0,
      "step": 29705
    },
    {
      "entropy": 5.320310068130493,
      "epoch": 2.6950290275761972,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042755841899562544,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20413169711828233,
      "num_tokens": 61649294.0,
      "step": 29710
    },
    {
      "entropy": 5.4761292934417725,
      "epoch": 2.6954825834542815,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042753463456352155,
      "loss": 5.0832,
      "mean_token_accuracy": 0.19653009474277497,
      "num_tokens": 61660312.0,
      "step": 29715
    },
    {
      "entropy": 5.411499071121216,
      "epoch": 2.6959361393323658,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004275108469769625,
      "loss": 5.0481,
      "mean_token_accuracy": 0.2053380072116852,
      "num_tokens": 61670539.0,
      "step": 29720
    },
    {
      "entropy": 5.388102293014526,
      "epoch": 2.69638969521045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042748705623644025,
      "loss": 4.9708,
      "mean_token_accuracy": 0.2100500002503395,
      "num_tokens": 61679613.0,
      "step": 29725
    },
    {
      "entropy": 5.3933491706848145,
      "epoch": 2.6968432510885343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004274632623424468,
      "loss": 5.0557,
      "mean_token_accuracy": 0.19715509563684464,
      "num_tokens": 61689714.0,
      "step": 29730
    },
    {
      "entropy": 5.4408026218414305,
      "epoch": 2.6972968069666186,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042743946529547435,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19654424488544464,
      "num_tokens": 61700701.0,
      "step": 29735
    },
    {
      "entropy": 5.4277674674987795,
      "epoch": 2.6977503628447024,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004274156650960148,
      "loss": 5.0897,
      "mean_token_accuracy": 0.1987374946475029,
      "num_tokens": 61711403.0,
      "step": 29740
    },
    {
      "entropy": 5.2684002876281735,
      "epoch": 2.6982039187227866,
      "grad_norm": 1.125,
      "learning_rate": 0.0004273918617445605,
      "loss": 4.8834,
      "mean_token_accuracy": 0.21251602619886398,
      "num_tokens": 61721787.0,
      "step": 29745
    },
    {
      "entropy": 5.40399432182312,
      "epoch": 2.698657474600871,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042736805524160363,
      "loss": 5.1569,
      "mean_token_accuracy": 0.19641713351011275,
      "num_tokens": 61732702.0,
      "step": 29750
    },
    {
      "entropy": 5.386903524398804,
      "epoch": 2.6991110304789547,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042734424558763656,
      "loss": 5.0545,
      "mean_token_accuracy": 0.20251042246818543,
      "num_tokens": 61743477.0,
      "step": 29755
    },
    {
      "entropy": 5.440708494186401,
      "epoch": 2.699564586357039,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004273204327831516,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19471192061901094,
      "num_tokens": 61753620.0,
      "step": 29760
    },
    {
      "entropy": 5.388537311553955,
      "epoch": 2.7000181422351233,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004272966168286414,
      "loss": 5.0076,
      "mean_token_accuracy": 0.2067195326089859,
      "num_tokens": 61763678.0,
      "step": 29765
    },
    {
      "entropy": 5.41415958404541,
      "epoch": 2.7004716981132075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042727279772459837,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1993306264281273,
      "num_tokens": 61774419.0,
      "step": 29770
    },
    {
      "entropy": 5.463683938980102,
      "epoch": 2.700925253991292,
      "grad_norm": 1.15625,
      "learning_rate": 0.000427248975471515,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19924687147140502,
      "num_tokens": 61785546.0,
      "step": 29775
    },
    {
      "entropy": 5.4191771984100345,
      "epoch": 2.701378809869376,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042722515006988397,
      "loss": 5.0049,
      "mean_token_accuracy": 0.20281445235013962,
      "num_tokens": 61795846.0,
      "step": 29780
    },
    {
      "entropy": 5.332764339447022,
      "epoch": 2.7018323657474603,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004272013215201982,
      "loss": 5.0707,
      "mean_token_accuracy": 0.19666958898305892,
      "num_tokens": 61806440.0,
      "step": 29785
    },
    {
      "entropy": 5.365705823898315,
      "epoch": 2.702285921625544,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042717748982295015,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20192903727293016,
      "num_tokens": 61815371.0,
      "step": 29790
    },
    {
      "entropy": 5.420378494262695,
      "epoch": 2.7027394775036284,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004271536549786329,
      "loss": 4.9931,
      "mean_token_accuracy": 0.2025228515267372,
      "num_tokens": 61825225.0,
      "step": 29795
    },
    {
      "entropy": 5.370684242248535,
      "epoch": 2.7031930333817127,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004271298169877393,
      "loss": 5.0735,
      "mean_token_accuracy": 0.20411968976259232,
      "num_tokens": 61836705.0,
      "step": 29800
    },
    {
      "entropy": 5.408667755126953,
      "epoch": 2.703646589259797,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271059758507624,
      "loss": 5.0261,
      "mean_token_accuracy": 0.20023907274007796,
      "num_tokens": 61847082.0,
      "step": 29805
    },
    {
      "entropy": 5.405517196655273,
      "epoch": 2.7041001451378808,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042708213156819505,
      "loss": 5.0519,
      "mean_token_accuracy": 0.19896311014890672,
      "num_tokens": 61857130.0,
      "step": 29810
    },
    {
      "entropy": 5.454390811920166,
      "epoch": 2.704553701015965,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004270582841405305,
      "loss": 5.143,
      "mean_token_accuracy": 0.1941450297832489,
      "num_tokens": 61866796.0,
      "step": 29815
    },
    {
      "entropy": 5.456648063659668,
      "epoch": 2.7050072568940493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004270344335682618,
      "loss": 5.0805,
      "mean_token_accuracy": 0.19387909919023513,
      "num_tokens": 61876356.0,
      "step": 29820
    },
    {
      "entropy": 5.353257083892823,
      "epoch": 2.7054608127721336,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042701057985188235,
      "loss": 5.0861,
      "mean_token_accuracy": 0.1959335207939148,
      "num_tokens": 61887374.0,
      "step": 29825
    },
    {
      "entropy": 5.4384105682373045,
      "epoch": 2.705914368650218,
      "grad_norm": 1.125,
      "learning_rate": 0.0004269867229918854,
      "loss": 5.1195,
      "mean_token_accuracy": 0.19748747646808623,
      "num_tokens": 61897376.0,
      "step": 29830
    },
    {
      "entropy": 5.479804372787475,
      "epoch": 2.706367924528302,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042696286298876424,
      "loss": 5.1416,
      "mean_token_accuracy": 0.19525735676288605,
      "num_tokens": 61908212.0,
      "step": 29835
    },
    {
      "entropy": 5.4614969253540036,
      "epoch": 2.706821480406386,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004269389998430124,
      "loss": 5.1521,
      "mean_token_accuracy": 0.187445168197155,
      "num_tokens": 61920699.0,
      "step": 29840
    },
    {
      "entropy": 5.415454912185669,
      "epoch": 2.70727503628447,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004269151335551233,
      "loss": 5.0726,
      "mean_token_accuracy": 0.197402124106884,
      "num_tokens": 61931556.0,
      "step": 29845
    },
    {
      "entropy": 5.466433954238892,
      "epoch": 2.7077285921625545,
      "grad_norm": 1.125,
      "learning_rate": 0.0004268912641255905,
      "loss": 5.1813,
      "mean_token_accuracy": 0.20088098645210267,
      "num_tokens": 61941483.0,
      "step": 29850
    },
    {
      "entropy": 5.389068365097046,
      "epoch": 2.7081821480406387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042686739155490763,
      "loss": 5.0079,
      "mean_token_accuracy": 0.20291488617658615,
      "num_tokens": 61951262.0,
      "step": 29855
    },
    {
      "entropy": 5.469057321548462,
      "epoch": 2.7086357039187225,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004268435158435684,
      "loss": 5.097,
      "mean_token_accuracy": 0.19506601095199586,
      "num_tokens": 61962423.0,
      "step": 29860
    },
    {
      "entropy": 5.494082403182984,
      "epoch": 2.709089259796807,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042681963699206667,
      "loss": 5.1647,
      "mean_token_accuracy": 0.18815541714429856,
      "num_tokens": 61973276.0,
      "step": 29865
    },
    {
      "entropy": 5.366739273071289,
      "epoch": 2.709542815674891,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042679575500089605,
      "loss": 5.0083,
      "mean_token_accuracy": 0.20364255756139754,
      "num_tokens": 61984369.0,
      "step": 29870
    },
    {
      "entropy": 5.352225971221924,
      "epoch": 2.7099963715529753,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004267718698705506,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20219548493623735,
      "num_tokens": 61994224.0,
      "step": 29875
    },
    {
      "entropy": 5.3619543552398685,
      "epoch": 2.7104499274310596,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004267479816015242,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20482718646526338,
      "num_tokens": 62004546.0,
      "step": 29880
    },
    {
      "entropy": 5.418005037307739,
      "epoch": 2.710903483309144,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004267240901943109,
      "loss": 5.0614,
      "mean_token_accuracy": 0.20377166122198104,
      "num_tokens": 62013924.0,
      "step": 29885
    },
    {
      "entropy": 5.403272008895874,
      "epoch": 2.7113570391872277,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042670019564940474,
      "loss": 5.0434,
      "mean_token_accuracy": 0.19787343889474868,
      "num_tokens": 62026150.0,
      "step": 29890
    },
    {
      "entropy": 5.4024840831756595,
      "epoch": 2.711810595065312,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004266762979672998,
      "loss": 5.0558,
      "mean_token_accuracy": 0.19646582305431365,
      "num_tokens": 62036203.0,
      "step": 29895
    },
    {
      "entropy": 5.412879180908203,
      "epoch": 2.7122641509433962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042665239714849044,
      "loss": 5.0632,
      "mean_token_accuracy": 0.19663473963737488,
      "num_tokens": 62046961.0,
      "step": 29900
    },
    {
      "entropy": 5.346556615829468,
      "epoch": 2.7127177068214805,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042662849319347073,
      "loss": 5.0155,
      "mean_token_accuracy": 0.19838072657585143,
      "num_tokens": 62057365.0,
      "step": 29905
    },
    {
      "entropy": 5.386085796356201,
      "epoch": 2.7131712626995643,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042660458610273527,
      "loss": 5.0097,
      "mean_token_accuracy": 0.19780435264110566,
      "num_tokens": 62067934.0,
      "step": 29910
    },
    {
      "entropy": 5.455262804031372,
      "epoch": 2.7136248185776486,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004265806758767783,
      "loss": 5.1169,
      "mean_token_accuracy": 0.1929931715130806,
      "num_tokens": 62077203.0,
      "step": 29915
    },
    {
      "entropy": 5.421120643615723,
      "epoch": 2.714078374455733,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042655676251609435,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19920509159564972,
      "num_tokens": 62087262.0,
      "step": 29920
    },
    {
      "entropy": 5.356379985809326,
      "epoch": 2.714531930333817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042653284602117795,
      "loss": 5.0493,
      "mean_token_accuracy": 0.1990089550614357,
      "num_tokens": 62097931.0,
      "step": 29925
    },
    {
      "entropy": 5.407110738754272,
      "epoch": 2.7149854862119014,
      "grad_norm": 1.0,
      "learning_rate": 0.0004265089263925236,
      "loss": 5.1289,
      "mean_token_accuracy": 0.19158044904470445,
      "num_tokens": 62109478.0,
      "step": 29930
    },
    {
      "entropy": 5.492278432846069,
      "epoch": 2.7154390420899857,
      "grad_norm": 1.015625,
      "learning_rate": 0.000426485003630626,
      "loss": 5.1202,
      "mean_token_accuracy": 0.19543081521987915,
      "num_tokens": 62120735.0,
      "step": 29935
    },
    {
      "entropy": 5.4279931545257565,
      "epoch": 2.71589259796807,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004264610777359801,
      "loss": 5.0361,
      "mean_token_accuracy": 0.1993958458304405,
      "num_tokens": 62130950.0,
      "step": 29940
    },
    {
      "entropy": 5.383908224105835,
      "epoch": 2.7163461538461537,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004264371487090804,
      "loss": 5.0646,
      "mean_token_accuracy": 0.19991304725408554,
      "num_tokens": 62141351.0,
      "step": 29945
    },
    {
      "entropy": 5.297012567520142,
      "epoch": 2.716799709724238,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042641321655042185,
      "loss": 4.9921,
      "mean_token_accuracy": 0.204030142724514,
      "num_tokens": 62151388.0,
      "step": 29950
    },
    {
      "entropy": 5.446522378921509,
      "epoch": 2.7172532656023223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004263892812604995,
      "loss": 5.1761,
      "mean_token_accuracy": 0.19742177575826644,
      "num_tokens": 62162809.0,
      "step": 29955
    },
    {
      "entropy": 5.418376684188843,
      "epoch": 2.7177068214804065,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042636534283980814,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20520640909671783,
      "num_tokens": 62172738.0,
      "step": 29960
    },
    {
      "entropy": 5.46491265296936,
      "epoch": 2.7181603773584904,
      "grad_norm": 1.03125,
      "learning_rate": 0.000426341401288843,
      "loss": 5.1905,
      "mean_token_accuracy": 0.19229784309864045,
      "num_tokens": 62182508.0,
      "step": 29965
    },
    {
      "entropy": 5.427996349334717,
      "epoch": 2.7186139332365746,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042631745660809905,
      "loss": 5.1541,
      "mean_token_accuracy": 0.19107009917497636,
      "num_tokens": 62192873.0,
      "step": 29970
    },
    {
      "entropy": 5.446498441696167,
      "epoch": 2.719067489114659,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042629350879807155,
      "loss": 5.1675,
      "mean_token_accuracy": 0.19017426371574403,
      "num_tokens": 62204727.0,
      "step": 29975
    },
    {
      "entropy": 5.448894071578979,
      "epoch": 2.719521044992743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042626955785925576,
      "loss": 5.0578,
      "mean_token_accuracy": 0.1933468297123909,
      "num_tokens": 62214268.0,
      "step": 29980
    },
    {
      "entropy": 5.4572559833526615,
      "epoch": 2.7199746008708274,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042624560379214693,
      "loss": 5.1084,
      "mean_token_accuracy": 0.20052724927663804,
      "num_tokens": 62224756.0,
      "step": 29985
    },
    {
      "entropy": 5.413329076766968,
      "epoch": 2.7204281567489117,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004262216465972405,
      "loss": 5.0356,
      "mean_token_accuracy": 0.19778827875852584,
      "num_tokens": 62234065.0,
      "step": 29990
    },
    {
      "entropy": 5.42058572769165,
      "epoch": 2.7208817126269955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004261976862750318,
      "loss": 5.0666,
      "mean_token_accuracy": 0.19908198863267898,
      "num_tokens": 62243732.0,
      "step": 29995
    },
    {
      "entropy": 5.488332748413086,
      "epoch": 2.72133526850508,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004261737228260165,
      "loss": 5.0826,
      "mean_token_accuracy": 0.20270526260137559,
      "num_tokens": 62253461.0,
      "step": 30000
    },
    {
      "epoch": 2.72133526850508,
      "eval_entropy": 5.200946805056404,
      "eval_loss": 5.126651763916016,
      "eval_mean_token_accuracy": 0.2046460328412946,
      "eval_num_tokens": 62253461.0,
      "eval_runtime": 26.3929,
      "eval_samples_per_second": 1339.753,
      "eval_steps_per_second": 167.469,
      "step": 30000
    },
    {
      "entropy": 5.424751615524292,
      "epoch": 2.721788824383164,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004261497562506901,
      "loss": 5.1027,
      "mean_token_accuracy": 0.19926908612251282,
      "num_tokens": 62263830.0,
      "step": 30005
    },
    {
      "entropy": 5.32047119140625,
      "epoch": 2.7222423802612483,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004261257865495483,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20667739808559418,
      "num_tokens": 62274383.0,
      "step": 30010
    },
    {
      "entropy": 5.4616859436035154,
      "epoch": 2.722695936139332,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004261018137230865,
      "loss": 5.2008,
      "mean_token_accuracy": 0.1833724021911621,
      "num_tokens": 62286476.0,
      "step": 30015
    },
    {
      "entropy": 5.527337265014649,
      "epoch": 2.7231494920174164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042607783777180085,
      "loss": 5.0999,
      "mean_token_accuracy": 0.20044250041246414,
      "num_tokens": 62296192.0,
      "step": 30020
    },
    {
      "entropy": 5.360300302505493,
      "epoch": 2.7236030478955007,
      "grad_norm": 1.125,
      "learning_rate": 0.00042605385869618693,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20860530883073808,
      "num_tokens": 62305300.0,
      "step": 30025
    },
    {
      "entropy": 5.408338308334351,
      "epoch": 2.724056603773585,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004260298764967408,
      "loss": 5.0508,
      "mean_token_accuracy": 0.19820448160171508,
      "num_tokens": 62316802.0,
      "step": 30030
    },
    {
      "entropy": 5.329595041275025,
      "epoch": 2.724510159651669,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004260058911739583,
      "loss": 5.0415,
      "mean_token_accuracy": 0.19691911339759827,
      "num_tokens": 62326444.0,
      "step": 30035
    },
    {
      "entropy": 5.387649583816528,
      "epoch": 2.7249637155297535,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042598190272833547,
      "loss": 4.9939,
      "mean_token_accuracy": 0.204490627348423,
      "num_tokens": 62336872.0,
      "step": 30040
    },
    {
      "entropy": 5.259207153320313,
      "epoch": 2.7254172714078373,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004259579111603685,
      "loss": 4.9244,
      "mean_token_accuracy": 0.20555949360132217,
      "num_tokens": 62346771.0,
      "step": 30045
    },
    {
      "entropy": 5.361576890945434,
      "epoch": 2.7258708272859216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042593391647055334,
      "loss": 5.0749,
      "mean_token_accuracy": 0.20057656168937682,
      "num_tokens": 62357503.0,
      "step": 30050
    },
    {
      "entropy": 5.35308804512024,
      "epoch": 2.726324383164006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004259099186593864,
      "loss": 4.9413,
      "mean_token_accuracy": 0.20961254686117173,
      "num_tokens": 62367664.0,
      "step": 30055
    },
    {
      "entropy": 5.41559066772461,
      "epoch": 2.72677793904209,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004258859177273639,
      "loss": 5.0195,
      "mean_token_accuracy": 0.19873745739459991,
      "num_tokens": 62376924.0,
      "step": 30060
    },
    {
      "entropy": 5.299186944961548,
      "epoch": 2.727231494920174,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004258619136749821,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2051345959305763,
      "num_tokens": 62387529.0,
      "step": 30065
    },
    {
      "entropy": 5.3408973693847654,
      "epoch": 2.727685050798258,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004258379065027375,
      "loss": 5.017,
      "mean_token_accuracy": 0.203294575214386,
      "num_tokens": 62397833.0,
      "step": 30070
    },
    {
      "entropy": 5.486776351928711,
      "epoch": 2.7281386066763424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004258138962111266,
      "loss": 5.0907,
      "mean_token_accuracy": 0.19701809138059617,
      "num_tokens": 62409022.0,
      "step": 30075
    },
    {
      "entropy": 5.371695518493652,
      "epoch": 2.7285921625544267,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004257898828006459,
      "loss": 5.0653,
      "mean_token_accuracy": 0.19014487266540528,
      "num_tokens": 62419105.0,
      "step": 30080
    },
    {
      "entropy": 5.382728052139282,
      "epoch": 2.729045718432511,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004257658662717919,
      "loss": 5.0413,
      "mean_token_accuracy": 0.20359178483486176,
      "num_tokens": 62428371.0,
      "step": 30085
    },
    {
      "entropy": 5.412524700164795,
      "epoch": 2.7294992743105952,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004257418466250615,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20227133631706237,
      "num_tokens": 62439301.0,
      "step": 30090
    },
    {
      "entropy": 5.3786461353302,
      "epoch": 2.7299528301886795,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042571782386095124,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2025826469063759,
      "num_tokens": 62449450.0,
      "step": 30095
    },
    {
      "entropy": 5.380969762802124,
      "epoch": 2.7304063860667633,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042569379797995793,
      "loss": 5.0383,
      "mean_token_accuracy": 0.19370868653059006,
      "num_tokens": 62461051.0,
      "step": 30100
    },
    {
      "entropy": 5.370656204223633,
      "epoch": 2.7308599419448476,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004256697689825785,
      "loss": 4.9826,
      "mean_token_accuracy": 0.20760337859392167,
      "num_tokens": 62471755.0,
      "step": 30105
    },
    {
      "entropy": 5.339439487457275,
      "epoch": 2.731313497822932,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042564573686930987,
      "loss": 5.0137,
      "mean_token_accuracy": 0.1987733796238899,
      "num_tokens": 62482060.0,
      "step": 30110
    },
    {
      "entropy": 5.335133790969849,
      "epoch": 2.7317670537010157,
      "grad_norm": 1.078125,
      "learning_rate": 0.000425621701640649,
      "loss": 5.0651,
      "mean_token_accuracy": 0.20036448091268538,
      "num_tokens": 62493735.0,
      "step": 30115
    },
    {
      "entropy": 5.407992601394653,
      "epoch": 2.7322206095791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000425597663297093,
      "loss": 5.0728,
      "mean_token_accuracy": 0.19690533876419067,
      "num_tokens": 62504582.0,
      "step": 30120
    },
    {
      "entropy": 5.359456872940063,
      "epoch": 2.7326741654571842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004255736218391389,
      "loss": 5.01,
      "mean_token_accuracy": 0.19697056859731674,
      "num_tokens": 62516665.0,
      "step": 30125
    },
    {
      "entropy": 5.396124458312988,
      "epoch": 2.7331277213352685,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042554957726728394,
      "loss": 5.0529,
      "mean_token_accuracy": 0.2011127531528473,
      "num_tokens": 62527745.0,
      "step": 30130
    },
    {
      "entropy": 5.394494724273682,
      "epoch": 2.7335812772133528,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004255255295820253,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20533617585897446,
      "num_tokens": 62538622.0,
      "step": 30135
    },
    {
      "entropy": 5.346411418914795,
      "epoch": 2.734034833091437,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042550147878386044,
      "loss": 5.035,
      "mean_token_accuracy": 0.20152650773525238,
      "num_tokens": 62548302.0,
      "step": 30140
    },
    {
      "entropy": 5.32673978805542,
      "epoch": 2.7344883889695213,
      "grad_norm": 1.25,
      "learning_rate": 0.0004254774248732867,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21034851521253586,
      "num_tokens": 62557150.0,
      "step": 30145
    },
    {
      "entropy": 5.500696992874145,
      "epoch": 2.734941944847605,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042545336785080126,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1868739441037178,
      "num_tokens": 62568291.0,
      "step": 30150
    },
    {
      "entropy": 5.404809284210205,
      "epoch": 2.7353955007256894,
      "grad_norm": 1.171875,
      "learning_rate": 0.000425429307716902,
      "loss": 5.0421,
      "mean_token_accuracy": 0.20029497891664505,
      "num_tokens": 62579351.0,
      "step": 30155
    },
    {
      "entropy": 5.3221172332763675,
      "epoch": 2.7358490566037736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042540524447208625,
      "loss": 4.9086,
      "mean_token_accuracy": 0.2086418852210045,
      "num_tokens": 62588614.0,
      "step": 30160
    },
    {
      "entropy": 5.428528070449829,
      "epoch": 2.736302612481858,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042538117811685174,
      "loss": 5.0479,
      "mean_token_accuracy": 0.2026985540986061,
      "num_tokens": 62598844.0,
      "step": 30165
    },
    {
      "entropy": 5.394393825531006,
      "epoch": 2.7367561683599417,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042535710865169616,
      "loss": 5.0681,
      "mean_token_accuracy": 0.19816955029964448,
      "num_tokens": 62608722.0,
      "step": 30170
    },
    {
      "entropy": 5.35942325592041,
      "epoch": 2.737209724238026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004253330360771173,
      "loss": 4.9796,
      "mean_token_accuracy": 0.1999138817191124,
      "num_tokens": 62618712.0,
      "step": 30175
    },
    {
      "entropy": 5.4106433391571045,
      "epoch": 2.7376632801161103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042530896039361284,
      "loss": 5.1415,
      "mean_token_accuracy": 0.18953516632318496,
      "num_tokens": 62629329.0,
      "step": 30180
    },
    {
      "entropy": 5.422923469543457,
      "epoch": 2.7381168359941945,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042528488160168077,
      "loss": 5.0845,
      "mean_token_accuracy": 0.1972768098115921,
      "num_tokens": 62639644.0,
      "step": 30185
    },
    {
      "entropy": 5.428293657302857,
      "epoch": 2.738570391872279,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042526079970181915,
      "loss": 5.1148,
      "mean_token_accuracy": 0.19065061956644058,
      "num_tokens": 62650979.0,
      "step": 30190
    },
    {
      "entropy": 5.471181058883667,
      "epoch": 2.739023947750363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004252367146945259,
      "loss": 5.139,
      "mean_token_accuracy": 0.1963496595621109,
      "num_tokens": 62661163.0,
      "step": 30195
    },
    {
      "entropy": 5.425654602050781,
      "epoch": 2.739477503628447,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042521262658029905,
      "loss": 5.1477,
      "mean_token_accuracy": 0.1930776208639145,
      "num_tokens": 62672371.0,
      "step": 30200
    },
    {
      "entropy": 5.366004800796508,
      "epoch": 2.739931059506531,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042518853535963673,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20186078548431396,
      "num_tokens": 62682521.0,
      "step": 30205
    },
    {
      "entropy": 5.33939208984375,
      "epoch": 2.7403846153846154,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042516444103303736,
      "loss": 4.9337,
      "mean_token_accuracy": 0.20942243933677673,
      "num_tokens": 62692420.0,
      "step": 30210
    },
    {
      "entropy": 5.3551328659057615,
      "epoch": 2.7408381712626997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042514034360099907,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20382513701915742,
      "num_tokens": 62702686.0,
      "step": 30215
    },
    {
      "entropy": 5.381569242477417,
      "epoch": 2.7412917271407835,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042511624306402015,
      "loss": 5.0473,
      "mean_token_accuracy": 0.20121122002601624,
      "num_tokens": 62713262.0,
      "step": 30220
    },
    {
      "entropy": 5.3795140266418455,
      "epoch": 2.7417452830188678,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004250921394225992,
      "loss": 5.005,
      "mean_token_accuracy": 0.2035805642604828,
      "num_tokens": 62724460.0,
      "step": 30225
    },
    {
      "entropy": 5.360591220855713,
      "epoch": 2.742198838896952,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004250680326772343,
      "loss": 5.0932,
      "mean_token_accuracy": 0.20079419314861296,
      "num_tokens": 62734190.0,
      "step": 30230
    },
    {
      "entropy": 5.405256605148315,
      "epoch": 2.7426523947750363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004250439228284244,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2002414047718048,
      "num_tokens": 62744323.0,
      "step": 30235
    },
    {
      "entropy": 5.390026664733886,
      "epoch": 2.7431059506531206,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042501980987666807,
      "loss": 5.0409,
      "mean_token_accuracy": 0.1987066835165024,
      "num_tokens": 62753868.0,
      "step": 30240
    },
    {
      "entropy": 5.414296197891235,
      "epoch": 2.743559506531205,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004249956938224636,
      "loss": 5.1303,
      "mean_token_accuracy": 0.19862023293972014,
      "num_tokens": 62763954.0,
      "step": 30245
    },
    {
      "entropy": 5.382428121566773,
      "epoch": 2.7440130624092887,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042497157466631015,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20916296541690826,
      "num_tokens": 62774116.0,
      "step": 30250
    },
    {
      "entropy": 5.496476840972901,
      "epoch": 2.744466618287373,
      "grad_norm": 1.0,
      "learning_rate": 0.00042494745240870626,
      "loss": 5.0933,
      "mean_token_accuracy": 0.19698828756809234,
      "num_tokens": 62784939.0,
      "step": 30255
    },
    {
      "entropy": 5.375816392898559,
      "epoch": 2.744920174165457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042492332705015075,
      "loss": 5.0531,
      "mean_token_accuracy": 0.2006991282105446,
      "num_tokens": 62794988.0,
      "step": 30260
    },
    {
      "entropy": 5.369754648208618,
      "epoch": 2.7453737300435415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042489919859114273,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20816653370857238,
      "num_tokens": 62804984.0,
      "step": 30265
    },
    {
      "entropy": 5.391978120803833,
      "epoch": 2.7458272859216253,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042487506703218104,
      "loss": 5.0196,
      "mean_token_accuracy": 0.19114813357591628,
      "num_tokens": 62816401.0,
      "step": 30270
    },
    {
      "entropy": 5.3766913414001465,
      "epoch": 2.7462808417997095,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004248509323737648,
      "loss": 4.996,
      "mean_token_accuracy": 0.20215995758771896,
      "num_tokens": 62826411.0,
      "step": 30275
    },
    {
      "entropy": 5.422786378860474,
      "epoch": 2.746734397677794,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004248267946163931,
      "loss": 5.0879,
      "mean_token_accuracy": 0.1961979329586029,
      "num_tokens": 62837228.0,
      "step": 30280
    },
    {
      "entropy": 5.326909923553467,
      "epoch": 2.747187953555878,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042480265376056506,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20826188176870347,
      "num_tokens": 62847852.0,
      "step": 30285
    },
    {
      "entropy": 5.341286087036133,
      "epoch": 2.7476415094339623,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042477850980678,
      "loss": 5.0866,
      "mean_token_accuracy": 0.19376596361398696,
      "num_tokens": 62859516.0,
      "step": 30290
    },
    {
      "entropy": 5.368163061141968,
      "epoch": 2.7480950653120466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004247543627555372,
      "loss": 5.0452,
      "mean_token_accuracy": 0.19852111041545867,
      "num_tokens": 62869757.0,
      "step": 30295
    },
    {
      "entropy": 5.330967664718628,
      "epoch": 2.748548621190131,
      "grad_norm": 1.109375,
      "learning_rate": 0.000424730212607336,
      "loss": 5.1133,
      "mean_token_accuracy": 0.19507840275764465,
      "num_tokens": 62879618.0,
      "step": 30300
    },
    {
      "entropy": 5.434990262985229,
      "epoch": 2.7490021770682147,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004247060593626758,
      "loss": 5.07,
      "mean_token_accuracy": 0.1960070699453354,
      "num_tokens": 62890491.0,
      "step": 30305
    },
    {
      "entropy": 5.462742519378662,
      "epoch": 2.749455732946299,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004246819030220562,
      "loss": 5.0839,
      "mean_token_accuracy": 0.2095190778374672,
      "num_tokens": 62899954.0,
      "step": 30310
    },
    {
      "entropy": 5.366754770278931,
      "epoch": 2.7499092888243832,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004246577435859767,
      "loss": 5.0913,
      "mean_token_accuracy": 0.20218194872140885,
      "num_tokens": 62911857.0,
      "step": 30315
    },
    {
      "entropy": 5.379948425292969,
      "epoch": 2.7503628447024675,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004246335810549369,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20580071359872817,
      "num_tokens": 62922023.0,
      "step": 30320
    },
    {
      "entropy": 5.44872145652771,
      "epoch": 2.7508164005805513,
      "grad_norm": 1.125,
      "learning_rate": 0.00042460941542943656,
      "loss": 5.0985,
      "mean_token_accuracy": 0.20783040076494216,
      "num_tokens": 62932845.0,
      "step": 30325
    },
    {
      "entropy": 5.3937506675720215,
      "epoch": 2.7512699564586356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004245852467099753,
      "loss": 5.0261,
      "mean_token_accuracy": 0.19893017411231995,
      "num_tokens": 62942881.0,
      "step": 30330
    },
    {
      "entropy": 5.313350057601928,
      "epoch": 2.75172351233672,
      "grad_norm": 1.0,
      "learning_rate": 0.0004245610748970531,
      "loss": 5.0113,
      "mean_token_accuracy": 0.19873238503932952,
      "num_tokens": 62953957.0,
      "step": 30335
    },
    {
      "entropy": 5.338680982589722,
      "epoch": 2.752177068214804,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042453689999116974,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20560264438390732,
      "num_tokens": 62964803.0,
      "step": 30340
    },
    {
      "entropy": 5.3736348152160645,
      "epoch": 2.7526306240928884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004245127219928252,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20111618489027022,
      "num_tokens": 62975477.0,
      "step": 30345
    },
    {
      "entropy": 5.356233930587768,
      "epoch": 2.7530841799709727,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00042448854090251944,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20252890139818192,
      "num_tokens": 62985430.0,
      "step": 30350
    },
    {
      "entropy": 5.3529942512512205,
      "epoch": 2.7535377358490565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042446435672075254,
      "loss": 5.045,
      "mean_token_accuracy": 0.19784435331821443,
      "num_tokens": 62997493.0,
      "step": 30355
    },
    {
      "entropy": 5.444988107681274,
      "epoch": 2.7539912917271407,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042444016944802467,
      "loss": 5.1455,
      "mean_token_accuracy": 0.20066531747579575,
      "num_tokens": 63007573.0,
      "step": 30360
    },
    {
      "entropy": 5.44176926612854,
      "epoch": 2.754444847605225,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042441597908483603,
      "loss": 5.0921,
      "mean_token_accuracy": 0.20032460391521453,
      "num_tokens": 63018372.0,
      "step": 30365
    },
    {
      "entropy": 5.391333198547363,
      "epoch": 2.7548984034833093,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004243917856316869,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20917322784662246,
      "num_tokens": 63028632.0,
      "step": 30370
    },
    {
      "entropy": 5.388815975189209,
      "epoch": 2.755351959361393,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042436758908907747,
      "loss": 5.0294,
      "mean_token_accuracy": 0.1985765278339386,
      "num_tokens": 63038510.0,
      "step": 30375
    },
    {
      "entropy": 5.476664686203003,
      "epoch": 2.7558055152394774,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042434338945750833,
      "loss": 5.2009,
      "mean_token_accuracy": 0.18543557822704315,
      "num_tokens": 63049375.0,
      "step": 30380
    },
    {
      "entropy": 5.416744804382324,
      "epoch": 2.7562590711175616,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042431918673747974,
      "loss": 5.0452,
      "mean_token_accuracy": 0.19913794845342636,
      "num_tokens": 63059422.0,
      "step": 30385
    },
    {
      "entropy": 5.472788381576538,
      "epoch": 2.756712626995646,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042429498092949245,
      "loss": 5.1412,
      "mean_token_accuracy": 0.18727853894233704,
      "num_tokens": 63071101.0,
      "step": 30390
    },
    {
      "entropy": 5.34202389717102,
      "epoch": 2.75716618287373,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004242707720340468,
      "loss": 4.9824,
      "mean_token_accuracy": 0.20348337441682815,
      "num_tokens": 63081104.0,
      "step": 30395
    },
    {
      "entropy": 5.377484226226807,
      "epoch": 2.7576197387518144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004242465600516436,
      "loss": 5.0702,
      "mean_token_accuracy": 0.19953351467847824,
      "num_tokens": 63091123.0,
      "step": 30400
    },
    {
      "entropy": 5.380934429168701,
      "epoch": 2.7580732946298983,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042422234498278353,
      "loss": 5.039,
      "mean_token_accuracy": 0.19966279119253158,
      "num_tokens": 63101541.0,
      "step": 30405
    },
    {
      "entropy": 5.438534164428711,
      "epoch": 2.7585268505079825,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042419812682796727,
      "loss": 5.1162,
      "mean_token_accuracy": 0.1966940641403198,
      "num_tokens": 63112012.0,
      "step": 30410
    },
    {
      "entropy": 5.387514400482178,
      "epoch": 2.758980406386067,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004241739055876957,
      "loss": 5.0297,
      "mean_token_accuracy": 0.1969226136803627,
      "num_tokens": 63122021.0,
      "step": 30415
    },
    {
      "entropy": 5.446099138259887,
      "epoch": 2.759433962264151,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042414968126246985,
      "loss": 5.1446,
      "mean_token_accuracy": 0.1947627618908882,
      "num_tokens": 63132286.0,
      "step": 30420
    },
    {
      "entropy": 5.372425174713134,
      "epoch": 2.759887518142235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042412545385279055,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20841199457645415,
      "num_tokens": 63142159.0,
      "step": 30425
    },
    {
      "entropy": 5.361631774902344,
      "epoch": 2.760341074020319,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042410122335915883,
      "loss": 4.9849,
      "mean_token_accuracy": 0.1992568001151085,
      "num_tokens": 63152225.0,
      "step": 30430
    },
    {
      "entropy": 5.359405374526977,
      "epoch": 2.7607946298984034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004240769897820758,
      "loss": 5.0211,
      "mean_token_accuracy": 0.20392679125070573,
      "num_tokens": 63162522.0,
      "step": 30435
    },
    {
      "entropy": 5.386137533187866,
      "epoch": 2.7612481857764877,
      "grad_norm": 1.0,
      "learning_rate": 0.00042405275312204266,
      "loss": 4.981,
      "mean_token_accuracy": 0.20562264323234558,
      "num_tokens": 63172094.0,
      "step": 30440
    },
    {
      "entropy": 5.35512752532959,
      "epoch": 2.761701741654572,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042402851337956066,
      "loss": 5.062,
      "mean_token_accuracy": 0.19950058311223984,
      "num_tokens": 63182122.0,
      "step": 30445
    },
    {
      "entropy": 5.349746561050415,
      "epoch": 2.762155297532656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042400427055513094,
      "loss": 5.0359,
      "mean_token_accuracy": 0.19539693742990494,
      "num_tokens": 63193174.0,
      "step": 30450
    },
    {
      "entropy": 5.438201570510865,
      "epoch": 2.7626088534107405,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000423980024649255,
      "loss": 4.9782,
      "mean_token_accuracy": 0.20892860740423203,
      "num_tokens": 63203698.0,
      "step": 30455
    },
    {
      "entropy": 5.2768162250518795,
      "epoch": 2.7630624092888243,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004239557756624342,
      "loss": 4.9123,
      "mean_token_accuracy": 0.20494503825902938,
      "num_tokens": 63213343.0,
      "step": 30460
    },
    {
      "entropy": 5.408652114868164,
      "epoch": 2.7635159651669086,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042393152359516994,
      "loss": 5.1434,
      "mean_token_accuracy": 0.19391841441392899,
      "num_tokens": 63223962.0,
      "step": 30465
    },
    {
      "entropy": 5.383771753311157,
      "epoch": 2.763969521044993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042390726844796384,
      "loss": 4.9894,
      "mean_token_accuracy": 0.2049087628722191,
      "num_tokens": 63234279.0,
      "step": 30470
    },
    {
      "entropy": 5.40609884262085,
      "epoch": 2.7644230769230766,
      "grad_norm": 1.125,
      "learning_rate": 0.0004238830102213174,
      "loss": 5.0912,
      "mean_token_accuracy": 0.1993947669863701,
      "num_tokens": 63245567.0,
      "step": 30475
    },
    {
      "entropy": 5.377705144882202,
      "epoch": 2.764876632801161,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004238587489157325,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2075589895248413,
      "num_tokens": 63256227.0,
      "step": 30480
    },
    {
      "entropy": 5.311401510238648,
      "epoch": 2.765330188679245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004238344845317107,
      "loss": 4.9424,
      "mean_token_accuracy": 0.20513349324464797,
      "num_tokens": 63266618.0,
      "step": 30485
    },
    {
      "entropy": 5.288667345046997,
      "epoch": 2.7657837445573294,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004238102170697539,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20921492725610732,
      "num_tokens": 63277130.0,
      "step": 30490
    },
    {
      "entropy": 5.2971635341644285,
      "epoch": 2.7662373004354137,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004237859465303637,
      "loss": 4.9046,
      "mean_token_accuracy": 0.21150925159454345,
      "num_tokens": 63287351.0,
      "step": 30495
    },
    {
      "entropy": 5.2888147830963135,
      "epoch": 2.766690856313498,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042376167291404234,
      "loss": 4.9978,
      "mean_token_accuracy": 0.2002965897321701,
      "num_tokens": 63298573.0,
      "step": 30500
    },
    {
      "entropy": 5.395222949981689,
      "epoch": 2.7671444121915822,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042373739622129166,
      "loss": 5.0485,
      "mean_token_accuracy": 0.19586751610040665,
      "num_tokens": 63310740.0,
      "step": 30505
    },
    {
      "entropy": 5.399692153930664,
      "epoch": 2.767597968069666,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042371311645261376,
      "loss": 5.0232,
      "mean_token_accuracy": 0.19878321439027785,
      "num_tokens": 63321323.0,
      "step": 30510
    },
    {
      "entropy": 5.382269239425659,
      "epoch": 2.7680515239477503,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004236888336085107,
      "loss": 5.0327,
      "mean_token_accuracy": 0.2059201866388321,
      "num_tokens": 63332322.0,
      "step": 30515
    },
    {
      "entropy": 5.335413026809692,
      "epoch": 2.7685050798258346,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004236645476894846,
      "loss": 4.9344,
      "mean_token_accuracy": 0.20956376194953918,
      "num_tokens": 63342204.0,
      "step": 30520
    },
    {
      "entropy": 5.259674406051635,
      "epoch": 2.768958635703919,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004236402586960378,
      "loss": 5.0024,
      "mean_token_accuracy": 0.20758628994226455,
      "num_tokens": 63352460.0,
      "step": 30525
    },
    {
      "entropy": 5.293515634536743,
      "epoch": 2.7694121915820027,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004236159666286726,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20447613894939423,
      "num_tokens": 63362474.0,
      "step": 30530
    },
    {
      "entropy": 5.407142496109008,
      "epoch": 2.769865747460087,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042359167148789135,
      "loss": 5.0264,
      "mean_token_accuracy": 0.21135727167129517,
      "num_tokens": 63372537.0,
      "step": 30535
    },
    {
      "entropy": 5.319940137863159,
      "epoch": 2.7703193033381712,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004235673732741964,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21414869725704194,
      "num_tokens": 63382985.0,
      "step": 30540
    },
    {
      "entropy": 5.3473803997039795,
      "epoch": 2.7707728592162555,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042354307198809037,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20319103002548217,
      "num_tokens": 63393424.0,
      "step": 30545
    },
    {
      "entropy": 5.261241245269775,
      "epoch": 2.7712264150943398,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004235187676300757,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21018621623516082,
      "num_tokens": 63403160.0,
      "step": 30550
    },
    {
      "entropy": 5.406363821029663,
      "epoch": 2.771679970972424,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004234944602006552,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20101663768291472,
      "num_tokens": 63413001.0,
      "step": 30555
    },
    {
      "entropy": 5.361946725845337,
      "epoch": 2.772133526850508,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042347014970033124,
      "loss": 4.999,
      "mean_token_accuracy": 0.20931318253278733,
      "num_tokens": 63422735.0,
      "step": 30560
    },
    {
      "entropy": 5.269633436203003,
      "epoch": 2.772587082728592,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004234458361296069,
      "loss": 4.9989,
      "mean_token_accuracy": 0.20032371580600739,
      "num_tokens": 63432566.0,
      "step": 30565
    },
    {
      "entropy": 5.360549402236939,
      "epoch": 2.7730406386066764,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042342151948898476,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20033189803361892,
      "num_tokens": 63442412.0,
      "step": 30570
    },
    {
      "entropy": 5.379809045791626,
      "epoch": 2.7734941944847606,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042339719977896784,
      "loss": 5.0198,
      "mean_token_accuracy": 0.20590965896844865,
      "num_tokens": 63453018.0,
      "step": 30575
    },
    {
      "entropy": 5.424623298645019,
      "epoch": 2.7739477503628445,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000423372877000059,
      "loss": 5.0544,
      "mean_token_accuracy": 0.19870160520076752,
      "num_tokens": 63464286.0,
      "step": 30580
    },
    {
      "entropy": 5.330277824401856,
      "epoch": 2.7744013062409287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042334855115276117,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20232771337032318,
      "num_tokens": 63475138.0,
      "step": 30585
    },
    {
      "entropy": 5.418788576126099,
      "epoch": 2.774854862119013,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042332422223757756,
      "loss": 5.0931,
      "mean_token_accuracy": 0.1992061674594879,
      "num_tokens": 63486104.0,
      "step": 30590
    },
    {
      "entropy": 5.317217779159546,
      "epoch": 2.7753084179970973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042329989025501123,
      "loss": 5.0061,
      "mean_token_accuracy": 0.2019389048218727,
      "num_tokens": 63496490.0,
      "step": 30595
    },
    {
      "entropy": 5.352649784088134,
      "epoch": 2.7757619738751815,
      "grad_norm": 1.125,
      "learning_rate": 0.00042327555520556536,
      "loss": 4.986,
      "mean_token_accuracy": 0.201349675655365,
      "num_tokens": 63506562.0,
      "step": 30600
    },
    {
      "entropy": 5.389174509048462,
      "epoch": 2.776215529753266,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042325121708974324,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20629602372646333,
      "num_tokens": 63516641.0,
      "step": 30605
    },
    {
      "entropy": 5.369472455978394,
      "epoch": 2.7766690856313496,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004232268759080482,
      "loss": 5.02,
      "mean_token_accuracy": 0.20426793694496154,
      "num_tokens": 63526864.0,
      "step": 30610
    },
    {
      "entropy": 5.3795839786529545,
      "epoch": 2.777122641509434,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042320253166098354,
      "loss": 5.1055,
      "mean_token_accuracy": 0.19832750856876374,
      "num_tokens": 63537160.0,
      "step": 30615
    },
    {
      "entropy": 5.40336012840271,
      "epoch": 2.777576197387518,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042317818434905275,
      "loss": 4.9863,
      "mean_token_accuracy": 0.20359795093536376,
      "num_tokens": 63547485.0,
      "step": 30620
    },
    {
      "entropy": 5.392403030395508,
      "epoch": 2.7780297532656024,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004231538339727594,
      "loss": 4.9872,
      "mean_token_accuracy": 0.19988429248332978,
      "num_tokens": 63559369.0,
      "step": 30625
    },
    {
      "entropy": 5.471197080612183,
      "epoch": 2.7784833091436862,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042312948053260704,
      "loss": 5.1901,
      "mean_token_accuracy": 0.18629950731992723,
      "num_tokens": 63570033.0,
      "step": 30630
    },
    {
      "entropy": 5.408815145492554,
      "epoch": 2.7789368650217705,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004231051240290992,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2050995022058487,
      "num_tokens": 63581274.0,
      "step": 30635
    },
    {
      "entropy": 5.392179822921753,
      "epoch": 2.7793904208998548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042308076446273963,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2015829563140869,
      "num_tokens": 63590903.0,
      "step": 30640
    },
    {
      "entropy": 5.277794265747071,
      "epoch": 2.779843976777939,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042305640183403214,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20488465428352357,
      "num_tokens": 63600982.0,
      "step": 30645
    },
    {
      "entropy": 5.290684080123901,
      "epoch": 2.7802975326560233,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042303203614348057,
      "loss": 4.9254,
      "mean_token_accuracy": 0.20209805220365523,
      "num_tokens": 63612411.0,
      "step": 30650
    },
    {
      "entropy": 5.343329191207886,
      "epoch": 2.7807510885341076,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004230076673915888,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20815595984458923,
      "num_tokens": 63622128.0,
      "step": 30655
    },
    {
      "entropy": 5.380663967132568,
      "epoch": 2.781204644412192,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042298329557886074,
      "loss": 5.0565,
      "mean_token_accuracy": 0.2009834423661232,
      "num_tokens": 63632047.0,
      "step": 30660
    },
    {
      "entropy": 5.461785888671875,
      "epoch": 2.7816582002902757,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004229589207058003,
      "loss": 5.0709,
      "mean_token_accuracy": 0.2011624738574028,
      "num_tokens": 63642490.0,
      "step": 30665
    },
    {
      "entropy": 5.340253400802612,
      "epoch": 2.78211175616836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042293454277291177,
      "loss": 4.9945,
      "mean_token_accuracy": 0.20552384108304977,
      "num_tokens": 63653206.0,
      "step": 30670
    },
    {
      "entropy": 5.270850658416748,
      "epoch": 2.782565312046444,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042291016178069916,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20215390473604203,
      "num_tokens": 63663427.0,
      "step": 30675
    },
    {
      "entropy": 5.279891490936279,
      "epoch": 2.7830188679245285,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004228857777296668,
      "loss": 4.9834,
      "mean_token_accuracy": 0.2071673035621643,
      "num_tokens": 63674269.0,
      "step": 30680
    },
    {
      "entropy": 5.338767337799072,
      "epoch": 2.7834724238026123,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004228613906203188,
      "loss": 4.9749,
      "mean_token_accuracy": 0.20204235166311263,
      "num_tokens": 63683989.0,
      "step": 30685
    },
    {
      "entropy": 5.367225122451782,
      "epoch": 2.7839259796806965,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042283700045315955,
      "loss": 4.9996,
      "mean_token_accuracy": 0.2013174518942833,
      "num_tokens": 63694540.0,
      "step": 30690
    },
    {
      "entropy": 5.357919311523437,
      "epoch": 2.784379535558781,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042281260722869355,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20067740231752396,
      "num_tokens": 63705149.0,
      "step": 30695
    },
    {
      "entropy": 5.298944807052612,
      "epoch": 2.784833091436865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042278821094742495,
      "loss": 4.9719,
      "mean_token_accuracy": 0.20063435137271882,
      "num_tokens": 63714061.0,
      "step": 30700
    },
    {
      "entropy": 5.418653774261474,
      "epoch": 2.7852866473149493,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042276381160985874,
      "loss": 5.0511,
      "mean_token_accuracy": 0.20065810680389404,
      "num_tokens": 63724732.0,
      "step": 30705
    },
    {
      "entropy": 5.383792495727539,
      "epoch": 2.7857402031930336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004227394092164991,
      "loss": 5.0315,
      "mean_token_accuracy": 0.20057410150766372,
      "num_tokens": 63734942.0,
      "step": 30710
    },
    {
      "entropy": 5.398646640777588,
      "epoch": 2.7861937590711174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004227150037678509,
      "loss": 5.1524,
      "mean_token_accuracy": 0.2071177914738655,
      "num_tokens": 63745326.0,
      "step": 30715
    },
    {
      "entropy": 5.393946123123169,
      "epoch": 2.7866473149492017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004226905952644188,
      "loss": 5.0172,
      "mean_token_accuracy": 0.2009531795978546,
      "num_tokens": 63755355.0,
      "step": 30720
    },
    {
      "entropy": 5.3841265678405765,
      "epoch": 2.787100870827286,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042266618370670743,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20144355148077012,
      "num_tokens": 63765112.0,
      "step": 30725
    },
    {
      "entropy": 5.309381580352783,
      "epoch": 2.7875544267053702,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042264176909522183,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21113115549087524,
      "num_tokens": 63774680.0,
      "step": 30730
    },
    {
      "entropy": 5.381954765319824,
      "epoch": 2.788007982583454,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004226173514304669,
      "loss": 5.082,
      "mean_token_accuracy": 0.20701963305473328,
      "num_tokens": 63784939.0,
      "step": 30735
    },
    {
      "entropy": 5.4267865180969235,
      "epoch": 2.7884615384615383,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004225929307129474,
      "loss": 5.1033,
      "mean_token_accuracy": 0.1897610157728195,
      "num_tokens": 63795836.0,
      "step": 30740
    },
    {
      "entropy": 5.384806680679321,
      "epoch": 2.7889150943396226,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042256850694316844,
      "loss": 4.9649,
      "mean_token_accuracy": 0.20194617807865142,
      "num_tokens": 63805607.0,
      "step": 30745
    },
    {
      "entropy": 5.310363626480102,
      "epoch": 2.789368650217707,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004225440801216353,
      "loss": 4.9678,
      "mean_token_accuracy": 0.20706472396850586,
      "num_tokens": 63815081.0,
      "step": 30750
    },
    {
      "entropy": 5.27111291885376,
      "epoch": 2.789822206095791,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004225196502488529,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2050422251224518,
      "num_tokens": 63825932.0,
      "step": 30755
    },
    {
      "entropy": 5.349942302703857,
      "epoch": 2.7902757619738754,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042249521732532657,
      "loss": 4.9484,
      "mean_token_accuracy": 0.20355030298233032,
      "num_tokens": 63835653.0,
      "step": 30760
    },
    {
      "entropy": 5.459928274154663,
      "epoch": 2.790729317851959,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004224707813515615,
      "loss": 5.1652,
      "mean_token_accuracy": 0.19274606108665465,
      "num_tokens": 63847657.0,
      "step": 30765
    },
    {
      "entropy": 5.450480222702026,
      "epoch": 2.7911828737300435,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042244634232806324,
      "loss": 5.0712,
      "mean_token_accuracy": 0.19759095907211305,
      "num_tokens": 63858476.0,
      "step": 30770
    },
    {
      "entropy": 5.309613037109375,
      "epoch": 2.7916364296081277,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042242190025533694,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21136646419763566,
      "num_tokens": 63867970.0,
      "step": 30775
    },
    {
      "entropy": 5.360978364944458,
      "epoch": 2.792089985486212,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004223974551338881,
      "loss": 4.9948,
      "mean_token_accuracy": 0.2140187904238701,
      "num_tokens": 63878294.0,
      "step": 30780
    },
    {
      "entropy": 5.304713487625122,
      "epoch": 2.792543541364296,
      "grad_norm": 1.0,
      "learning_rate": 0.0004223730069642224,
      "loss": 4.9934,
      "mean_token_accuracy": 0.199824757874012,
      "num_tokens": 63889315.0,
      "step": 30785
    },
    {
      "entropy": 5.310572147369385,
      "epoch": 2.79299709724238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004223485557468454,
      "loss": 4.98,
      "mean_token_accuracy": 0.19991356432437896,
      "num_tokens": 63899923.0,
      "step": 30790
    },
    {
      "entropy": 5.2911420345306395,
      "epoch": 2.7934506531204644,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042232410148226277,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20941994935274125,
      "num_tokens": 63909813.0,
      "step": 30795
    },
    {
      "entropy": 5.319617986679077,
      "epoch": 2.7939042089985486,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042229964417098007,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20020392835140227,
      "num_tokens": 63920931.0,
      "step": 30800
    },
    {
      "entropy": 5.300818014144897,
      "epoch": 2.794357764876633,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042227518381350316,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2057108610868454,
      "num_tokens": 63930799.0,
      "step": 30805
    },
    {
      "entropy": 5.492636728286743,
      "epoch": 2.794811320754717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042225072041033804,
      "loss": 5.1331,
      "mean_token_accuracy": 0.1944277748465538,
      "num_tokens": 63941763.0,
      "step": 30810
    },
    {
      "entropy": 5.458992290496826,
      "epoch": 2.7952648766328014,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042222625396199043,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19169605225324632,
      "num_tokens": 63952775.0,
      "step": 30815
    },
    {
      "entropy": 5.404214239120483,
      "epoch": 2.7957184325108853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004222017844689663,
      "loss": 5.0637,
      "mean_token_accuracy": 0.20549387633800506,
      "num_tokens": 63964223.0,
      "step": 30820
    },
    {
      "entropy": 5.374739313125611,
      "epoch": 2.7961719883889695,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004221773119317719,
      "loss": 5.1194,
      "mean_token_accuracy": 0.19921797811985015,
      "num_tokens": 63976413.0,
      "step": 30825
    },
    {
      "entropy": 5.463623428344727,
      "epoch": 2.796625544267054,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042215283635091313,
      "loss": 5.0822,
      "mean_token_accuracy": 0.20404281318187714,
      "num_tokens": 63986578.0,
      "step": 30830
    },
    {
      "entropy": 5.412996053695679,
      "epoch": 2.7970791001451376,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004221283577268962,
      "loss": 5.0482,
      "mean_token_accuracy": 0.2092985913157463,
      "num_tokens": 63996856.0,
      "step": 30835
    },
    {
      "entropy": 5.274520397186279,
      "epoch": 2.797532656023222,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042210387606022724,
      "loss": 5.0011,
      "mean_token_accuracy": 0.20856949537992478,
      "num_tokens": 64007700.0,
      "step": 30840
    },
    {
      "entropy": 5.430122661590576,
      "epoch": 2.797986211901306,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004220793913514128,
      "loss": 5.1657,
      "mean_token_accuracy": 0.19249868541955947,
      "num_tokens": 64018771.0,
      "step": 30845
    },
    {
      "entropy": 5.390881633758545,
      "epoch": 2.7984397677793904,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004220549036009589,
      "loss": 5.0321,
      "mean_token_accuracy": 0.19879861623048783,
      "num_tokens": 64029381.0,
      "step": 30850
    },
    {
      "entropy": 5.279818868637085,
      "epoch": 2.7988933236574747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042203041280937224,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2056998059153557,
      "num_tokens": 64039596.0,
      "step": 30855
    },
    {
      "entropy": 5.420748472213745,
      "epoch": 2.799346879535559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004220059189771592,
      "loss": 5.0562,
      "mean_token_accuracy": 0.21141218841075898,
      "num_tokens": 64049123.0,
      "step": 30860
    },
    {
      "entropy": 5.414463520050049,
      "epoch": 2.799800435413643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042198142210482623,
      "loss": 5.0821,
      "mean_token_accuracy": 0.19914474189281464,
      "num_tokens": 64059426.0,
      "step": 30865
    },
    {
      "entropy": 5.299511098861695,
      "epoch": 2.800253991291727,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042195692219287997,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21435992419719696,
      "num_tokens": 64069227.0,
      "step": 30870
    },
    {
      "entropy": 5.355284309387207,
      "epoch": 2.8007075471698113,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004219324192418272,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20360470414161683,
      "num_tokens": 64078482.0,
      "step": 30875
    },
    {
      "entropy": 5.378703641891479,
      "epoch": 2.8011611030478956,
      "grad_norm": 1.125,
      "learning_rate": 0.00042190791325217447,
      "loss": 5.0012,
      "mean_token_accuracy": 0.204270339012146,
      "num_tokens": 64088664.0,
      "step": 30880
    },
    {
      "entropy": 5.3556914806365965,
      "epoch": 2.80161465892598,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004218834042244287,
      "loss": 4.951,
      "mean_token_accuracy": 0.20531627535820007,
      "num_tokens": 64099316.0,
      "step": 30885
    },
    {
      "entropy": 5.312828922271729,
      "epoch": 2.8020682148040637,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042185889215909665,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20428589582443238,
      "num_tokens": 64109786.0,
      "step": 30890
    },
    {
      "entropy": 5.390570735931396,
      "epoch": 2.802521770682148,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042183437705668535,
      "loss": 4.9935,
      "mean_token_accuracy": 0.2059012994170189,
      "num_tokens": 64120771.0,
      "step": 30895
    },
    {
      "entropy": 5.431223297119141,
      "epoch": 2.802975326560232,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042180985891770176,
      "loss": 5.0478,
      "mean_token_accuracy": 0.19412563741207123,
      "num_tokens": 64130833.0,
      "step": 30900
    },
    {
      "entropy": 5.287931823730469,
      "epoch": 2.8034288824383164,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004217853377426528,
      "loss": 4.9676,
      "mean_token_accuracy": 0.20415856838226318,
      "num_tokens": 64139672.0,
      "step": 30905
    },
    {
      "entropy": 5.474739837646484,
      "epoch": 2.8038824383164007,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004217608135320456,
      "loss": 5.1626,
      "mean_token_accuracy": 0.19852155447006226,
      "num_tokens": 64150383.0,
      "step": 30910
    },
    {
      "entropy": 5.3344518661499025,
      "epoch": 2.804335994194485,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004217362862863875,
      "loss": 4.9656,
      "mean_token_accuracy": 0.2110673874616623,
      "num_tokens": 64160786.0,
      "step": 30915
    },
    {
      "entropy": 5.4119409084320065,
      "epoch": 2.804789550072569,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042171175600618546,
      "loss": 5.0648,
      "mean_token_accuracy": 0.1983549028635025,
      "num_tokens": 64172057.0,
      "step": 30920
    },
    {
      "entropy": 5.362578392028809,
      "epoch": 2.805243105950653,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042168722269194704,
      "loss": 5.045,
      "mean_token_accuracy": 0.20970076620578765,
      "num_tokens": 64182599.0,
      "step": 30925
    },
    {
      "entropy": 5.349286365509033,
      "epoch": 2.8056966618287373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004216626863441794,
      "loss": 5.0604,
      "mean_token_accuracy": 0.1993937835097313,
      "num_tokens": 64193356.0,
      "step": 30930
    },
    {
      "entropy": 5.434128522872925,
      "epoch": 2.8061502177068216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042163814696339014,
      "loss": 5.1315,
      "mean_token_accuracy": 0.19517428576946258,
      "num_tokens": 64203625.0,
      "step": 30935
    },
    {
      "entropy": 5.391817903518676,
      "epoch": 2.8066037735849054,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004216136045500866,
      "loss": 5.135,
      "mean_token_accuracy": 0.20016538500785827,
      "num_tokens": 64214742.0,
      "step": 30940
    },
    {
      "entropy": 5.395909070968628,
      "epoch": 2.8070573294629897,
      "grad_norm": 1.125,
      "learning_rate": 0.0004215890591047763,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20213039070367814,
      "num_tokens": 64223811.0,
      "step": 30945
    },
    {
      "entropy": 5.377744436264038,
      "epoch": 2.807510885341074,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042156451062796686,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21088848114013672,
      "num_tokens": 64234854.0,
      "step": 30950
    },
    {
      "entropy": 5.401691436767578,
      "epoch": 2.8079644412191582,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042153995912016606,
      "loss": 5.0728,
      "mean_token_accuracy": 0.20080698728561402,
      "num_tokens": 64245914.0,
      "step": 30955
    },
    {
      "entropy": 5.403627681732178,
      "epoch": 2.8084179970972425,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004215154045818816,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19835127294063568,
      "num_tokens": 64256217.0,
      "step": 30960
    },
    {
      "entropy": 5.413418292999268,
      "epoch": 2.8088715529753268,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004214908470136212,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19280827790498734,
      "num_tokens": 64267211.0,
      "step": 30965
    },
    {
      "entropy": 5.4342793941497805,
      "epoch": 2.8093251088534106,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004214662864158927,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20344641357660292,
      "num_tokens": 64277770.0,
      "step": 30970
    },
    {
      "entropy": 5.326209688186646,
      "epoch": 2.809778664731495,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042144172278920413,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21010780930519105,
      "num_tokens": 64287973.0,
      "step": 30975
    },
    {
      "entropy": 5.322775936126709,
      "epoch": 2.810232220609579,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004214171561340634,
      "loss": 5.038,
      "mean_token_accuracy": 0.19826909005641938,
      "num_tokens": 64297898.0,
      "step": 30980
    },
    {
      "entropy": 5.37241153717041,
      "epoch": 2.8106857764876634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042139258645097863,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21481916606426238,
      "num_tokens": 64308114.0,
      "step": 30985
    },
    {
      "entropy": 5.3655096054077145,
      "epoch": 2.811139332365747,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042136801374045775,
      "loss": 5.0001,
      "mean_token_accuracy": 0.2011181727051735,
      "num_tokens": 64319586.0,
      "step": 30990
    },
    {
      "entropy": 5.371907138824463,
      "epoch": 2.8115928882438315,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004213434380030091,
      "loss": 5.0849,
      "mean_token_accuracy": 0.19811442196369172,
      "num_tokens": 64330612.0,
      "step": 30995
    },
    {
      "entropy": 5.435436725616455,
      "epoch": 2.8120464441219157,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004213188592391409,
      "loss": 5.1203,
      "mean_token_accuracy": 0.19466065615415573,
      "num_tokens": 64341934.0,
      "step": 31000
    },
    {
      "entropy": 5.412488222122192,
      "epoch": 2.8125,
      "grad_norm": 1.0,
      "learning_rate": 0.0004212942774493614,
      "loss": 4.9878,
      "mean_token_accuracy": 0.20534046739339828,
      "num_tokens": 64352166.0,
      "step": 31005
    },
    {
      "entropy": 5.352151536941529,
      "epoch": 2.8129535558780843,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042126969263417885,
      "loss": 5.0427,
      "mean_token_accuracy": 0.19819290935993195,
      "num_tokens": 64361995.0,
      "step": 31010
    },
    {
      "entropy": 5.303421354293823,
      "epoch": 2.8134071117561685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042124510479410183,
      "loss": 4.9938,
      "mean_token_accuracy": 0.20216063857078553,
      "num_tokens": 64372336.0,
      "step": 31015
    },
    {
      "entropy": 5.371252965927124,
      "epoch": 2.813860667634253,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042122051392963885,
      "loss": 5.034,
      "mean_token_accuracy": 0.2122174918651581,
      "num_tokens": 64382309.0,
      "step": 31020
    },
    {
      "entropy": 5.421134948730469,
      "epoch": 2.8143142235123366,
      "grad_norm": 1.25,
      "learning_rate": 0.0004211959200412984,
      "loss": 5.0794,
      "mean_token_accuracy": 0.19973201155662537,
      "num_tokens": 64393205.0,
      "step": 31025
    },
    {
      "entropy": 5.363537883758545,
      "epoch": 2.814767779390421,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004211713231295889,
      "loss": 5.041,
      "mean_token_accuracy": 0.20355946570634842,
      "num_tokens": 64403068.0,
      "step": 31030
    },
    {
      "entropy": 5.3884913444519045,
      "epoch": 2.815221335268505,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004211467231950194,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20245780497789384,
      "num_tokens": 64414296.0,
      "step": 31035
    },
    {
      "entropy": 5.353057527542115,
      "epoch": 2.8156748911465894,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042112212023809824,
      "loss": 5.0313,
      "mean_token_accuracy": 0.19993753880262374,
      "num_tokens": 64424748.0,
      "step": 31040
    },
    {
      "entropy": 5.374426412582397,
      "epoch": 2.8161284470246732,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004210975142593345,
      "loss": 4.9772,
      "mean_token_accuracy": 0.20455421656370162,
      "num_tokens": 64434723.0,
      "step": 31045
    },
    {
      "entropy": 5.432885980606079,
      "epoch": 2.8165820029027575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004210729052592369,
      "loss": 5.1054,
      "mean_token_accuracy": 0.19649850577116013,
      "num_tokens": 64445864.0,
      "step": 31050
    },
    {
      "entropy": 5.307337379455566,
      "epoch": 2.8170355587808418,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004210482932383144,
      "loss": 4.9609,
      "mean_token_accuracy": 0.20217827409505845,
      "num_tokens": 64456704.0,
      "step": 31055
    },
    {
      "entropy": 5.31566367149353,
      "epoch": 2.817489114658926,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000421023678197076,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20389137268066407,
      "num_tokens": 64466870.0,
      "step": 31060
    },
    {
      "entropy": 5.379441165924073,
      "epoch": 2.8179426705370103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004209990601360307,
      "loss": 5.0381,
      "mean_token_accuracy": 0.19891558289527894,
      "num_tokens": 64477069.0,
      "step": 31065
    },
    {
      "entropy": 5.410980129241944,
      "epoch": 2.8183962264150946,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042097443905568765,
      "loss": 5.0359,
      "mean_token_accuracy": 0.19656723588705063,
      "num_tokens": 64487165.0,
      "step": 31070
    },
    {
      "entropy": 5.306865262985229,
      "epoch": 2.8188497822931784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042094981495655596,
      "loss": 4.913,
      "mean_token_accuracy": 0.20712795555591584,
      "num_tokens": 64497977.0,
      "step": 31075
    },
    {
      "entropy": 5.368498086929321,
      "epoch": 2.8193033381712627,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000420925187839145,
      "loss": 5.0381,
      "mean_token_accuracy": 0.20034489184617996,
      "num_tokens": 64508767.0,
      "step": 31080
    },
    {
      "entropy": 5.3168642044067385,
      "epoch": 2.819756894049347,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042090055770396383,
      "loss": 4.9311,
      "mean_token_accuracy": 0.204688860476017,
      "num_tokens": 64519040.0,
      "step": 31085
    },
    {
      "entropy": 5.315719890594482,
      "epoch": 2.820210449927431,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042087592455152206,
      "loss": 4.9657,
      "mean_token_accuracy": 0.21236419081687927,
      "num_tokens": 64529721.0,
      "step": 31090
    },
    {
      "entropy": 5.35269250869751,
      "epoch": 2.820664005805515,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000420851288382329,
      "loss": 5.1266,
      "mean_token_accuracy": 0.1918093279004097,
      "num_tokens": 64541299.0,
      "step": 31095
    },
    {
      "entropy": 5.446705532073975,
      "epoch": 2.8211175616835993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042082664919689406,
      "loss": 4.9816,
      "mean_token_accuracy": 0.20644061714410783,
      "num_tokens": 64550831.0,
      "step": 31100
    },
    {
      "entropy": 5.375873374938965,
      "epoch": 2.8215711175616836,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004208020069957269,
      "loss": 5.0143,
      "mean_token_accuracy": 0.19909945279359817,
      "num_tokens": 64562076.0,
      "step": 31105
    },
    {
      "entropy": 5.276387929916382,
      "epoch": 2.822024673439768,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000420777361779337,
      "loss": 5.0106,
      "mean_token_accuracy": 0.20155130326747894,
      "num_tokens": 64572955.0,
      "step": 31110
    },
    {
      "entropy": 5.441110610961914,
      "epoch": 2.822478229317852,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042075271354823427,
      "loss": 5.0439,
      "mean_token_accuracy": 0.205308173596859,
      "num_tokens": 64584376.0,
      "step": 31115
    },
    {
      "entropy": 5.438935279846191,
      "epoch": 2.8229317851959363,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004207280623029281,
      "loss": 5.0583,
      "mean_token_accuracy": 0.19917130917310716,
      "num_tokens": 64594522.0,
      "step": 31120
    },
    {
      "entropy": 5.359533977508545,
      "epoch": 2.82338534107402,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042070340804392853,
      "loss": 5.0754,
      "mean_token_accuracy": 0.2002228245139122,
      "num_tokens": 64604625.0,
      "step": 31125
    },
    {
      "entropy": 5.411371183395386,
      "epoch": 2.8238388969521044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004206787507717454,
      "loss": 5.0286,
      "mean_token_accuracy": 0.20436375588178635,
      "num_tokens": 64614184.0,
      "step": 31130
    },
    {
      "entropy": 5.420125818252563,
      "epoch": 2.8242924528301887,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042065409048688856,
      "loss": 5.0912,
      "mean_token_accuracy": 0.19892250895500183,
      "num_tokens": 64623019.0,
      "step": 31135
    },
    {
      "entropy": 5.331463956832886,
      "epoch": 2.824746008708273,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000420629427189868,
      "loss": 5.0245,
      "mean_token_accuracy": 0.1984967142343521,
      "num_tokens": 64634017.0,
      "step": 31140
    },
    {
      "entropy": 5.343720579147339,
      "epoch": 2.825199564586357,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042060476088119376,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20749538838863374,
      "num_tokens": 64645007.0,
      "step": 31145
    },
    {
      "entropy": 5.460946416854858,
      "epoch": 2.825653120464441,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000420580091561376,
      "loss": 5.0661,
      "mean_token_accuracy": 0.1985458806157112,
      "num_tokens": 64655705.0,
      "step": 31150
    },
    {
      "entropy": 5.468452453613281,
      "epoch": 2.8261066763425253,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042055541923092485,
      "loss": 5.1262,
      "mean_token_accuracy": 0.19944604784250258,
      "num_tokens": 64666094.0,
      "step": 31155
    },
    {
      "entropy": 5.402296495437622,
      "epoch": 2.8265602322206096,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004205307438903505,
      "loss": 5.096,
      "mean_token_accuracy": 0.19497907757759095,
      "num_tokens": 64676107.0,
      "step": 31160
    },
    {
      "entropy": 5.340319061279297,
      "epoch": 2.827013788098694,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004205060655401633,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20272487550973892,
      "num_tokens": 64685846.0,
      "step": 31165
    },
    {
      "entropy": 5.302958631515503,
      "epoch": 2.827467343976778,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004204813841808736,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20358201414346694,
      "num_tokens": 64695959.0,
      "step": 31170
    },
    {
      "entropy": 5.365285634994507,
      "epoch": 2.8279208998548624,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004204566998129917,
      "loss": 5.0016,
      "mean_token_accuracy": 0.20407870709896087,
      "num_tokens": 64705594.0,
      "step": 31175
    },
    {
      "entropy": 5.421918344497681,
      "epoch": 2.828374455732946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042043201243702834,
      "loss": 5.0518,
      "mean_token_accuracy": 0.20004392117261888,
      "num_tokens": 64715974.0,
      "step": 31180
    },
    {
      "entropy": 5.3130663394927975,
      "epoch": 2.8288280116110305,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004204073220534938,
      "loss": 4.9705,
      "mean_token_accuracy": 0.2060365706682205,
      "num_tokens": 64725846.0,
      "step": 31185
    },
    {
      "entropy": 5.3385429859161375,
      "epoch": 2.8292815674891147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004203826286628988,
      "loss": 4.958,
      "mean_token_accuracy": 0.20894140899181365,
      "num_tokens": 64735789.0,
      "step": 31190
    },
    {
      "entropy": 5.336319828033448,
      "epoch": 2.8297351233671986,
      "grad_norm": 1.15625,
      "learning_rate": 0.000420357932265754,
      "loss": 4.9268,
      "mean_token_accuracy": 0.21818263679742814,
      "num_tokens": 64745738.0,
      "step": 31195
    },
    {
      "entropy": 5.328480815887451,
      "epoch": 2.830188679245283,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004203332328625701,
      "loss": 5.0141,
      "mean_token_accuracy": 0.19954464435577393,
      "num_tokens": 64756523.0,
      "step": 31200
    },
    {
      "entropy": 5.396979665756225,
      "epoch": 2.830642235123367,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042030853045385796,
      "loss": 5.1157,
      "mean_token_accuracy": 0.19429542869329453,
      "num_tokens": 64767367.0,
      "step": 31205
    },
    {
      "entropy": 5.392879056930542,
      "epoch": 2.8310957910014514,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042028382504012837,
      "loss": 5.0394,
      "mean_token_accuracy": 0.202104452252388,
      "num_tokens": 64777729.0,
      "step": 31210
    },
    {
      "entropy": 5.454670143127442,
      "epoch": 2.8315493468795356,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042025911662189224,
      "loss": 5.0553,
      "mean_token_accuracy": 0.20532247722148894,
      "num_tokens": 64788223.0,
      "step": 31215
    },
    {
      "entropy": 5.319871330261231,
      "epoch": 2.83200290275762,
      "grad_norm": 1.125,
      "learning_rate": 0.00042023440519966057,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2010360762476921,
      "num_tokens": 64797367.0,
      "step": 31220
    },
    {
      "entropy": 5.364941167831421,
      "epoch": 2.832456458635704,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042020969077394434,
      "loss": 5.0343,
      "mean_token_accuracy": 0.20020794570446016,
      "num_tokens": 64807179.0,
      "step": 31225
    },
    {
      "entropy": 5.406053113937378,
      "epoch": 2.832910014513788,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042018497334525475,
      "loss": 4.9303,
      "mean_token_accuracy": 0.20334082692861558,
      "num_tokens": 64817276.0,
      "step": 31230
    },
    {
      "entropy": 5.3379417896270756,
      "epoch": 2.8333635703918723,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004201602529141029,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20207542628049852,
      "num_tokens": 64827509.0,
      "step": 31235
    },
    {
      "entropy": 5.363300848007202,
      "epoch": 2.8338171262699565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042013552948100003,
      "loss": 5.0252,
      "mean_token_accuracy": 0.20012049973011017,
      "num_tokens": 64836578.0,
      "step": 31240
    },
    {
      "entropy": 5.3187971115112305,
      "epoch": 2.834270682148041,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004201108030464574,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20998306721448898,
      "num_tokens": 64846518.0,
      "step": 31245
    },
    {
      "entropy": 5.41255464553833,
      "epoch": 2.8347242380261246,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042008607361098646,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20297278314828873,
      "num_tokens": 64856304.0,
      "step": 31250
    },
    {
      "entropy": 5.330134105682373,
      "epoch": 2.835177793904209,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004200613411750985,
      "loss": 4.9408,
      "mean_token_accuracy": 0.212796089053154,
      "num_tokens": 64867381.0,
      "step": 31255
    },
    {
      "entropy": 5.335889387130737,
      "epoch": 2.835631349782293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000420036605739305,
      "loss": 5.0659,
      "mean_token_accuracy": 0.19474900811910628,
      "num_tokens": 64877306.0,
      "step": 31260
    },
    {
      "entropy": 5.3152453899383545,
      "epoch": 2.8360849056603774,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004200118673041176,
      "loss": 4.9752,
      "mean_token_accuracy": 0.20220406502485275,
      "num_tokens": 64886932.0,
      "step": 31265
    },
    {
      "entropy": 5.452041482925415,
      "epoch": 2.8365384615384617,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041998712587004776,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19744213670492172,
      "num_tokens": 64897387.0,
      "step": 31270
    },
    {
      "entropy": 5.393964338302612,
      "epoch": 2.836992017416546,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041996238143760726,
      "loss": 5.0592,
      "mean_token_accuracy": 0.2080727517604828,
      "num_tokens": 64907251.0,
      "step": 31275
    },
    {
      "entropy": 5.280246686935425,
      "epoch": 2.8374455732946298,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041993763400730776,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2036980301141739,
      "num_tokens": 64917300.0,
      "step": 31280
    },
    {
      "entropy": 5.3257866382598875,
      "epoch": 2.837899129172714,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004199128835796611,
      "loss": 4.9527,
      "mean_token_accuracy": 0.20799694955348969,
      "num_tokens": 64927133.0,
      "step": 31285
    },
    {
      "entropy": 5.40493221282959,
      "epoch": 2.8383526850507983,
      "grad_norm": 1.109375,
      "learning_rate": 0.000419888130155179,
      "loss": 5.017,
      "mean_token_accuracy": 0.20187835097312928,
      "num_tokens": 64937789.0,
      "step": 31290
    },
    {
      "entropy": 5.441811227798462,
      "epoch": 2.8388062409288826,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004198633737343735,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20225802510976792,
      "num_tokens": 64948127.0,
      "step": 31295
    },
    {
      "entropy": 5.391612148284912,
      "epoch": 2.8392597968069664,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004198386143177565,
      "loss": 5.0826,
      "mean_token_accuracy": 0.2066267430782318,
      "num_tokens": 64959632.0,
      "step": 31300
    },
    {
      "entropy": 5.233128786087036,
      "epoch": 2.8397133526850507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041981385190584003,
      "loss": 4.8549,
      "mean_token_accuracy": 0.21165927350521088,
      "num_tokens": 64970411.0,
      "step": 31305
    },
    {
      "entropy": 5.272853660583496,
      "epoch": 2.840166908563135,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004197890864991362,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20278060287237168,
      "num_tokens": 64980748.0,
      "step": 31310
    },
    {
      "entropy": 5.3584846496582035,
      "epoch": 2.840620464441219,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004197643180981572,
      "loss": 4.9432,
      "mean_token_accuracy": 0.2006258025765419,
      "num_tokens": 64990818.0,
      "step": 31315
    },
    {
      "entropy": 5.408952522277832,
      "epoch": 2.8410740203193035,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004197395467034152,
      "loss": 4.9956,
      "mean_token_accuracy": 0.20695394426584243,
      "num_tokens": 65001993.0,
      "step": 31320
    },
    {
      "entropy": 5.306002044677735,
      "epoch": 2.8415275761973877,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004197147723154225,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21345075517892836,
      "num_tokens": 65013007.0,
      "step": 31325
    },
    {
      "entropy": 5.414487409591675,
      "epoch": 2.8419811320754715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004196899949346914,
      "loss": 4.9927,
      "mean_token_accuracy": 0.19674643129110336,
      "num_tokens": 65023595.0,
      "step": 31330
    },
    {
      "entropy": 5.315158128738403,
      "epoch": 2.842434687953556,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004196652145617343,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20264245718717575,
      "num_tokens": 65033631.0,
      "step": 31335
    },
    {
      "entropy": 5.379789209365844,
      "epoch": 2.84288824383164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004196404311970638,
      "loss": 5.0259,
      "mean_token_accuracy": 0.2058860957622528,
      "num_tokens": 65044059.0,
      "step": 31340
    },
    {
      "entropy": 5.337030124664307,
      "epoch": 2.8433417997097243,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004196156448411922,
      "loss": 5.007,
      "mean_token_accuracy": 0.20163046419620514,
      "num_tokens": 65053931.0,
      "step": 31345
    },
    {
      "entropy": 5.369784116744995,
      "epoch": 2.843795355587808,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041959085549463234,
      "loss": 5.0172,
      "mean_token_accuracy": 0.2084933713078499,
      "num_tokens": 65064011.0,
      "step": 31350
    },
    {
      "entropy": 5.447152948379516,
      "epoch": 2.8442489114658924,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004195660631578966,
      "loss": 5.0905,
      "mean_token_accuracy": 0.19944891929626465,
      "num_tokens": 65074417.0,
      "step": 31355
    },
    {
      "entropy": 5.324445056915283,
      "epoch": 2.8447024673439767,
      "grad_norm": 1.09375,
      "learning_rate": 0.000419541267831498,
      "loss": 4.9893,
      "mean_token_accuracy": 0.20809907615184783,
      "num_tokens": 65084593.0,
      "step": 31360
    },
    {
      "entropy": 5.318591356277466,
      "epoch": 2.845156023222061,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000419516469515949,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20541626513004302,
      "num_tokens": 65095047.0,
      "step": 31365
    },
    {
      "entropy": 5.404474496841431,
      "epoch": 2.8456095791001452,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041949166821176264,
      "loss": 5.0929,
      "mean_token_accuracy": 0.20315452367067338,
      "num_tokens": 65104434.0,
      "step": 31370
    },
    {
      "entropy": 5.454703998565674,
      "epoch": 2.8460631349782295,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041946686391945183,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20594650506973267,
      "num_tokens": 65114039.0,
      "step": 31375
    },
    {
      "entropy": 5.480854272842407,
      "epoch": 2.8465166908563138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004194420566395294,
      "loss": 5.1454,
      "mean_token_accuracy": 0.189396670460701,
      "num_tokens": 65125387.0,
      "step": 31380
    },
    {
      "entropy": 5.3963460445404055,
      "epoch": 2.8469702467343976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004194172463725084,
      "loss": 5.047,
      "mean_token_accuracy": 0.20932589322328568,
      "num_tokens": 65136351.0,
      "step": 31385
    },
    {
      "entropy": 5.356893682479859,
      "epoch": 2.847423802612482,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041939243311890195,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20001642853021623,
      "num_tokens": 65147343.0,
      "step": 31390
    },
    {
      "entropy": 5.347734022140503,
      "epoch": 2.847877358490566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041936761687922324,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21186278760433197,
      "num_tokens": 65157097.0,
      "step": 31395
    },
    {
      "entropy": 5.352808570861816,
      "epoch": 2.8483309143686504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041934279765398535,
      "loss": 5.0283,
      "mean_token_accuracy": 0.20221714675426483,
      "num_tokens": 65167546.0,
      "step": 31400
    },
    {
      "entropy": 5.379040622711182,
      "epoch": 2.848784470246734,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004193179754437017,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2000724494457245,
      "num_tokens": 65178497.0,
      "step": 31405
    },
    {
      "entropy": 5.334487771987915,
      "epoch": 2.8492380261248185,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004192931502488855,
      "loss": 4.9614,
      "mean_token_accuracy": 0.20753635019063948,
      "num_tokens": 65189530.0,
      "step": 31410
    },
    {
      "entropy": 5.350640535354614,
      "epoch": 2.8496915820029027,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041926832207005025,
      "loss": 4.9895,
      "mean_token_accuracy": 0.2018150046467781,
      "num_tokens": 65199836.0,
      "step": 31415
    },
    {
      "entropy": 5.257211112976075,
      "epoch": 2.850145137880987,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004192434909077092,
      "loss": 4.861,
      "mean_token_accuracy": 0.2097110405564308,
      "num_tokens": 65210208.0,
      "step": 31420
    },
    {
      "entropy": 5.336671590805054,
      "epoch": 2.8505986937590713,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041921865676237605,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20641868114471434,
      "num_tokens": 65220954.0,
      "step": 31425
    },
    {
      "entropy": 5.379059219360352,
      "epoch": 2.8510522496371555,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041919381963456437,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20617053806781768,
      "num_tokens": 65230870.0,
      "step": 31430
    },
    {
      "entropy": 5.4085164070129395,
      "epoch": 2.8515058055152394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004191689795247877,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20910748988389968,
      "num_tokens": 65241913.0,
      "step": 31435
    },
    {
      "entropy": 5.43391637802124,
      "epoch": 2.8519593613933236,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004191441364335598,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20666832774877547,
      "num_tokens": 65252513.0,
      "step": 31440
    },
    {
      "entropy": 5.29032039642334,
      "epoch": 2.852412917271408,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041911929036139457,
      "loss": 5.0198,
      "mean_token_accuracy": 0.20056467354297638,
      "num_tokens": 65262695.0,
      "step": 31445
    },
    {
      "entropy": 5.388340997695923,
      "epoch": 2.852866473149492,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041909444130880553,
      "loss": 5.0843,
      "mean_token_accuracy": 0.20154549181461334,
      "num_tokens": 65274131.0,
      "step": 31450
    },
    {
      "entropy": 5.41184515953064,
      "epoch": 2.853320029027576,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041906958927630673,
      "loss": 5.0324,
      "mean_token_accuracy": 0.1991874098777771,
      "num_tokens": 65285080.0,
      "step": 31455
    },
    {
      "entropy": 5.455717515945435,
      "epoch": 2.8537735849056602,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004190447342644122,
      "loss": 5.2139,
      "mean_token_accuracy": 0.19967146962881088,
      "num_tokens": 65296512.0,
      "step": 31460
    },
    {
      "entropy": 5.323091173171997,
      "epoch": 2.8542271407837445,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004190198762736358,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21402711123228074,
      "num_tokens": 65306054.0,
      "step": 31465
    },
    {
      "entropy": 5.293573379516602,
      "epoch": 2.8546806966618288,
      "grad_norm": 1.0,
      "learning_rate": 0.0004189950153044917,
      "loss": 5.0114,
      "mean_token_accuracy": 0.2015373021364212,
      "num_tokens": 65315046.0,
      "step": 31470
    },
    {
      "entropy": 5.3876536846160885,
      "epoch": 2.855134252539913,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004189701513574939,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20435848385095595,
      "num_tokens": 65325482.0,
      "step": 31475
    },
    {
      "entropy": 5.335081481933594,
      "epoch": 2.8555878084179973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004189452844331567,
      "loss": 4.9304,
      "mean_token_accuracy": 0.20633240193128585,
      "num_tokens": 65335518.0,
      "step": 31480
    },
    {
      "entropy": 5.350926399230957,
      "epoch": 2.856041364296081,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041892041453199443,
      "loss": 5.0733,
      "mean_token_accuracy": 0.2017892614006996,
      "num_tokens": 65346079.0,
      "step": 31485
    },
    {
      "entropy": 5.337039279937744,
      "epoch": 2.8564949201741654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004188955416545212,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20483266115188598,
      "num_tokens": 65356719.0,
      "step": 31490
    },
    {
      "entropy": 5.398135995864868,
      "epoch": 2.8569484760522497,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004188706658012516,
      "loss": 5.0447,
      "mean_token_accuracy": 0.19827880710363388,
      "num_tokens": 65367877.0,
      "step": 31495
    },
    {
      "entropy": 5.416129493713379,
      "epoch": 2.857402031930334,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041884578697269994,
      "loss": 5.0869,
      "mean_token_accuracy": 0.1956907704472542,
      "num_tokens": 65379624.0,
      "step": 31500
    },
    {
      "entropy": 5.433998107910156,
      "epoch": 2.8578555878084178,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004188209051693807,
      "loss": 5.0754,
      "mean_token_accuracy": 0.20035671591758727,
      "num_tokens": 65390417.0,
      "step": 31505
    },
    {
      "entropy": 5.230963182449341,
      "epoch": 2.858309143686502,
      "grad_norm": 1.0,
      "learning_rate": 0.00041879602039180856,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2171856313943863,
      "num_tokens": 65401681.0,
      "step": 31510
    },
    {
      "entropy": 5.31591763496399,
      "epoch": 2.8587626995645863,
      "grad_norm": 3.921875,
      "learning_rate": 0.00041877113264049805,
      "loss": 4.9742,
      "mean_token_accuracy": 0.20785432159900666,
      "num_tokens": 65412520.0,
      "step": 31515
    },
    {
      "entropy": 5.352530193328858,
      "epoch": 2.8592162554426706,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041874624191596383,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20420881509780883,
      "num_tokens": 65423201.0,
      "step": 31520
    },
    {
      "entropy": 5.3253851413726805,
      "epoch": 2.859669811320755,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004187213482187207,
      "loss": 5.025,
      "mean_token_accuracy": 0.20134246051311494,
      "num_tokens": 65432778.0,
      "step": 31525
    },
    {
      "entropy": 5.446673107147217,
      "epoch": 2.860123367198839,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004186964515492835,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20461848825216294,
      "num_tokens": 65442323.0,
      "step": 31530
    },
    {
      "entropy": 5.364483261108399,
      "epoch": 2.8605769230769234,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004186715519081671,
      "loss": 4.9998,
      "mean_token_accuracy": 0.20361677408218384,
      "num_tokens": 65451980.0,
      "step": 31535
    },
    {
      "entropy": 5.233925247192383,
      "epoch": 2.861030478955007,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004186466492958864,
      "loss": 4.9432,
      "mean_token_accuracy": 0.20644866824150085,
      "num_tokens": 65463359.0,
      "step": 31540
    },
    {
      "entropy": 5.320037221908569,
      "epoch": 2.8614840348330914,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004186217437129562,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2038339152932167,
      "num_tokens": 65474155.0,
      "step": 31545
    },
    {
      "entropy": 5.350466060638428,
      "epoch": 2.8619375907111757,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041859683515989196,
      "loss": 4.953,
      "mean_token_accuracy": 0.21124183237552643,
      "num_tokens": 65484884.0,
      "step": 31550
    },
    {
      "entropy": 5.369892120361328,
      "epoch": 2.8623911465892595,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041857192363720845,
      "loss": 5.0318,
      "mean_token_accuracy": 0.2065276399254799,
      "num_tokens": 65494380.0,
      "step": 31555
    },
    {
      "entropy": 5.338099336624145,
      "epoch": 2.862844702467344,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004185470091454211,
      "loss": 5.0069,
      "mean_token_accuracy": 0.2076554074883461,
      "num_tokens": 65505735.0,
      "step": 31560
    },
    {
      "entropy": 5.3931629180908205,
      "epoch": 2.863298258345428,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004185220916850449,
      "loss": 4.9632,
      "mean_token_accuracy": 0.20647501200437546,
      "num_tokens": 65515956.0,
      "step": 31565
    },
    {
      "entropy": 5.307736158370972,
      "epoch": 2.8637518142235123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041849717125659537,
      "loss": 4.9378,
      "mean_token_accuracy": 0.20722079575061797,
      "num_tokens": 65526369.0,
      "step": 31570
    },
    {
      "entropy": 5.297730016708374,
      "epoch": 2.8642053701015966,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004184722478605877,
      "loss": 4.978,
      "mean_token_accuracy": 0.21061307489871978,
      "num_tokens": 65536032.0,
      "step": 31575
    },
    {
      "entropy": 5.330682420730591,
      "epoch": 2.864658925979681,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004184473214975374,
      "loss": 5.0175,
      "mean_token_accuracy": 0.2025287017226219,
      "num_tokens": 65546121.0,
      "step": 31580
    },
    {
      "entropy": 5.373110914230347,
      "epoch": 2.865112481857765,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041842239216795994,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20936665385961534,
      "num_tokens": 65555915.0,
      "step": 31585
    },
    {
      "entropy": 5.363604307174683,
      "epoch": 2.865566037735849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041839745987237086,
      "loss": 5.0277,
      "mean_token_accuracy": 0.1946626275777817,
      "num_tokens": 65566267.0,
      "step": 31590
    },
    {
      "entropy": 5.377901268005371,
      "epoch": 2.866019593613933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004183725246112857,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2097027450799942,
      "num_tokens": 65577126.0,
      "step": 31595
    },
    {
      "entropy": 5.389923524856568,
      "epoch": 2.8664731494920175,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041834758638522034,
      "loss": 5.117,
      "mean_token_accuracy": 0.1937570959329605,
      "num_tokens": 65587158.0,
      "step": 31600
    },
    {
      "entropy": 5.353647518157959,
      "epoch": 2.8669267053701017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041832264519469026,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20050076246261597,
      "num_tokens": 65597720.0,
      "step": 31605
    },
    {
      "entropy": 5.406346940994263,
      "epoch": 2.8673802612481856,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041829770104021143,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20422150641679765,
      "num_tokens": 65607409.0,
      "step": 31610
    },
    {
      "entropy": 5.379460191726684,
      "epoch": 2.86783381712627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041827275392229956,
      "loss": 4.963,
      "mean_token_accuracy": 0.21618593037128447,
      "num_tokens": 65616785.0,
      "step": 31615
    },
    {
      "entropy": 5.340894746780395,
      "epoch": 2.868287373004354,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004182478038414706,
      "loss": 5.0751,
      "mean_token_accuracy": 0.1995897725224495,
      "num_tokens": 65625794.0,
      "step": 31620
    },
    {
      "entropy": 5.25696587562561,
      "epoch": 2.8687409288824384,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041822285079824056,
      "loss": 4.8892,
      "mean_token_accuracy": 0.21320734322071075,
      "num_tokens": 65635789.0,
      "step": 31625
    },
    {
      "entropy": 5.406510972976685,
      "epoch": 2.8691944847605226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004181978947931256,
      "loss": 5.1245,
      "mean_token_accuracy": 0.20105577856302262,
      "num_tokens": 65646266.0,
      "step": 31630
    },
    {
      "entropy": 5.377872514724731,
      "epoch": 2.869648040638607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004181729358266415,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20329605787992477,
      "num_tokens": 65656538.0,
      "step": 31635
    },
    {
      "entropy": 5.451664113998413,
      "epoch": 2.8701015965166907,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041814797389930474,
      "loss": 5.0178,
      "mean_token_accuracy": 0.20649702697992325,
      "num_tokens": 65666926.0,
      "step": 31640
    },
    {
      "entropy": 5.357976627349854,
      "epoch": 2.870555152394775,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004181230090116313,
      "loss": 5.0016,
      "mean_token_accuracy": 0.20185298174619676,
      "num_tokens": 65677219.0,
      "step": 31645
    },
    {
      "entropy": 5.3431542873382565,
      "epoch": 2.8710087082728593,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041809804116413763,
      "loss": 5.0659,
      "mean_token_accuracy": 0.19386952072381974,
      "num_tokens": 65688281.0,
      "step": 31650
    },
    {
      "entropy": 5.4860954761505125,
      "epoch": 2.8714622641509435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041807307035734,
      "loss": 5.1133,
      "mean_token_accuracy": 0.19550684690475464,
      "num_tokens": 65698815.0,
      "step": 31655
    },
    {
      "entropy": 5.414642477035523,
      "epoch": 2.8719158200290273,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004180480965917548,
      "loss": 5.0599,
      "mean_token_accuracy": 0.20057271420955658,
      "num_tokens": 65709481.0,
      "step": 31660
    },
    {
      "entropy": 5.419437408447266,
      "epoch": 2.8723693759071116,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00041802311986789854,
      "loss": 5.0036,
      "mean_token_accuracy": 0.19606589823961257,
      "num_tokens": 65720561.0,
      "step": 31665
    },
    {
      "entropy": 5.391331958770752,
      "epoch": 2.872822931785196,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041799814018628765,
      "loss": 5.0532,
      "mean_token_accuracy": 0.19782052487134932,
      "num_tokens": 65730384.0,
      "step": 31670
    },
    {
      "entropy": 5.366039800643921,
      "epoch": 2.87327648766328,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041797315754743883,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20884816348552704,
      "num_tokens": 65740711.0,
      "step": 31675
    },
    {
      "entropy": 5.48991584777832,
      "epoch": 2.8737300435413644,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041794817195186867,
      "loss": 5.137,
      "mean_token_accuracy": 0.1909068703651428,
      "num_tokens": 65751033.0,
      "step": 31680
    },
    {
      "entropy": 5.430130815505981,
      "epoch": 2.8741835994194487,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004179231834000939,
      "loss": 5.1054,
      "mean_token_accuracy": 0.19714543521404265,
      "num_tokens": 65761500.0,
      "step": 31685
    },
    {
      "entropy": 5.27670841217041,
      "epoch": 2.8746371552975325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041789819189263117,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2111854672431946,
      "num_tokens": 65771095.0,
      "step": 31690
    },
    {
      "entropy": 5.340111303329468,
      "epoch": 2.8750907111756168,
      "grad_norm": 1.0,
      "learning_rate": 0.00041787319742999756,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20568643361330033,
      "num_tokens": 65782058.0,
      "step": 31695
    },
    {
      "entropy": 5.387290620803833,
      "epoch": 2.875544267053701,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004178482000127098,
      "loss": 5.0174,
      "mean_token_accuracy": 0.209634692966938,
      "num_tokens": 65791710.0,
      "step": 31700
    },
    {
      "entropy": 5.3925989151000975,
      "epoch": 2.8759978229317853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004178231996412848,
      "loss": 5.0824,
      "mean_token_accuracy": 0.1950618252158165,
      "num_tokens": 65802180.0,
      "step": 31705
    },
    {
      "entropy": 5.325191450119019,
      "epoch": 2.876451378809869,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004177981963162396,
      "loss": 4.958,
      "mean_token_accuracy": 0.21527987718582153,
      "num_tokens": 65812998.0,
      "step": 31710
    },
    {
      "entropy": 5.403501796722412,
      "epoch": 2.8769049346879534,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004177731900380914,
      "loss": 5.0381,
      "mean_token_accuracy": 0.20321642756462097,
      "num_tokens": 65824210.0,
      "step": 31715
    },
    {
      "entropy": 5.36378173828125,
      "epoch": 2.8773584905660377,
      "grad_norm": 1.125,
      "learning_rate": 0.0004177481808073572,
      "loss": 4.9871,
      "mean_token_accuracy": 0.2081717312335968,
      "num_tokens": 65834322.0,
      "step": 31720
    },
    {
      "entropy": 5.3815442562103275,
      "epoch": 2.877812046444122,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041772316862455426,
      "loss": 4.9424,
      "mean_token_accuracy": 0.211194272339344,
      "num_tokens": 65845347.0,
      "step": 31725
    },
    {
      "entropy": 5.398548316955567,
      "epoch": 2.878265602322206,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041769815349019986,
      "loss": 5.0399,
      "mean_token_accuracy": 0.20264209508895875,
      "num_tokens": 65855616.0,
      "step": 31730
    },
    {
      "entropy": 5.363584709167481,
      "epoch": 2.8787191582002905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041767313540481125,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20226377099752427,
      "num_tokens": 65865542.0,
      "step": 31735
    },
    {
      "entropy": 5.324860239028931,
      "epoch": 2.8791727140783747,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004176481143689058,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2024829313158989,
      "num_tokens": 65874478.0,
      "step": 31740
    },
    {
      "entropy": 5.435469245910644,
      "epoch": 2.8796262699564585,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004176230903830011,
      "loss": 5.0517,
      "mean_token_accuracy": 0.20232833325862884,
      "num_tokens": 65884609.0,
      "step": 31745
    },
    {
      "entropy": 5.342041110992431,
      "epoch": 2.880079825834543,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004175980634476145,
      "loss": 5.0054,
      "mean_token_accuracy": 0.20707691460847855,
      "num_tokens": 65894486.0,
      "step": 31750
    },
    {
      "entropy": 5.26611704826355,
      "epoch": 2.880533381712627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004175730335632635,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2079906567931175,
      "num_tokens": 65904643.0,
      "step": 31755
    },
    {
      "entropy": 5.26160888671875,
      "epoch": 2.8809869375907113,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041754800073046593,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21648788154125215,
      "num_tokens": 65914107.0,
      "step": 31760
    },
    {
      "entropy": 5.355012893676758,
      "epoch": 2.881440493468795,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004175229649497394,
      "loss": 4.9346,
      "mean_token_accuracy": 0.20578435063362122,
      "num_tokens": 65925112.0,
      "step": 31765
    },
    {
      "entropy": 5.39633150100708,
      "epoch": 2.8818940493468794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004174979262216016,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20145976841449736,
      "num_tokens": 65934624.0,
      "step": 31770
    },
    {
      "entropy": 5.388868093490601,
      "epoch": 2.8823476052249637,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004174728845465704,
      "loss": 5.0173,
      "mean_token_accuracy": 0.19742471873760223,
      "num_tokens": 65945951.0,
      "step": 31775
    },
    {
      "entropy": 5.298048591613769,
      "epoch": 2.882801161103048,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004174478399251636,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21097800135612488,
      "num_tokens": 65956871.0,
      "step": 31780
    },
    {
      "entropy": 5.4150299549102785,
      "epoch": 2.8832547169811322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004174227923578992,
      "loss": 5.1352,
      "mean_token_accuracy": 0.19260210990905763,
      "num_tokens": 65967814.0,
      "step": 31785
    },
    {
      "entropy": 5.359264945983886,
      "epoch": 2.8837082728592165,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004173977418452952,
      "loss": 5.0091,
      "mean_token_accuracy": 0.200546433031559,
      "num_tokens": 65978118.0,
      "step": 31790
    },
    {
      "entropy": 5.336725521087646,
      "epoch": 2.8841618287373003,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041737268838786955,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20841561406850814,
      "num_tokens": 65988240.0,
      "step": 31795
    },
    {
      "entropy": 5.385180234909058,
      "epoch": 2.8846153846153846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004173476319861404,
      "loss": 4.9546,
      "mean_token_accuracy": 0.20272916704416274,
      "num_tokens": 65998126.0,
      "step": 31800
    },
    {
      "entropy": 5.352530145645142,
      "epoch": 2.885068940493469,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004173225726406261,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20645343214273454,
      "num_tokens": 66009118.0,
      "step": 31805
    },
    {
      "entropy": 5.373167037963867,
      "epoch": 2.885522496371553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041729751035184457,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19794801771640777,
      "num_tokens": 66019340.0,
      "step": 31810
    },
    {
      "entropy": 5.324721622467041,
      "epoch": 2.885976052249637,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041727244512031425,
      "loss": 4.9695,
      "mean_token_accuracy": 0.2062823086977005,
      "num_tokens": 66029784.0,
      "step": 31815
    },
    {
      "entropy": 5.3806404113769535,
      "epoch": 2.886429608127721,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004172473769465536,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19053297936916352,
      "num_tokens": 66039987.0,
      "step": 31820
    },
    {
      "entropy": 5.3813286304473875,
      "epoch": 2.8868831640058055,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041722230583108097,
      "loss": 5.052,
      "mean_token_accuracy": 0.20510578155517578,
      "num_tokens": 66049150.0,
      "step": 31825
    },
    {
      "entropy": 5.392625570297241,
      "epoch": 2.8873367198838897,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041719723177441467,
      "loss": 5.0599,
      "mean_token_accuracy": 0.20349549055099486,
      "num_tokens": 66060719.0,
      "step": 31830
    },
    {
      "entropy": 5.400241184234619,
      "epoch": 2.887790275761974,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004171721547770735,
      "loss": 5.0359,
      "mean_token_accuracy": 0.1996895492076874,
      "num_tokens": 66071126.0,
      "step": 31835
    },
    {
      "entropy": 5.342827033996582,
      "epoch": 2.8882438316400583,
      "grad_norm": 1.125,
      "learning_rate": 0.0004171470748395759,
      "loss": 4.9547,
      "mean_token_accuracy": 0.20711992383003236,
      "num_tokens": 66080928.0,
      "step": 31840
    },
    {
      "entropy": 5.401762866973877,
      "epoch": 2.888697387518142,
      "grad_norm": 1.0,
      "learning_rate": 0.0004171219919624406,
      "loss": 5.024,
      "mean_token_accuracy": 0.19675727933645248,
      "num_tokens": 66090900.0,
      "step": 31845
    },
    {
      "entropy": 5.3553056716918945,
      "epoch": 2.8891509433962264,
      "grad_norm": 1.25,
      "learning_rate": 0.00041709690614618624,
      "loss": 5.0208,
      "mean_token_accuracy": 0.19977837502956391,
      "num_tokens": 66100720.0,
      "step": 31850
    },
    {
      "entropy": 5.350343227386475,
      "epoch": 2.8896044992743106,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041707181739133163,
      "loss": 5.0394,
      "mean_token_accuracy": 0.19504331797361374,
      "num_tokens": 66110780.0,
      "step": 31855
    },
    {
      "entropy": 5.3716367244720455,
      "epoch": 2.890058055152395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004170467256983957,
      "loss": 4.9919,
      "mean_token_accuracy": 0.21246565133333206,
      "num_tokens": 66120976.0,
      "step": 31860
    },
    {
      "entropy": 5.367414331436157,
      "epoch": 2.8905116110304787,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004170216310678973,
      "loss": 5.0196,
      "mean_token_accuracy": 0.20645262897014618,
      "num_tokens": 66130221.0,
      "step": 31865
    },
    {
      "entropy": 5.326510810852051,
      "epoch": 2.890965166908563,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041699653350035527,
      "loss": 4.9435,
      "mean_token_accuracy": 0.2150791347026825,
      "num_tokens": 66140060.0,
      "step": 31870
    },
    {
      "entropy": 5.329116153717041,
      "epoch": 2.8914187227866472,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041697143299628886,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20797624737024306,
      "num_tokens": 66151640.0,
      "step": 31875
    },
    {
      "entropy": 5.315364265441895,
      "epoch": 2.8918722786647315,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041694632955621697,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21243895143270491,
      "num_tokens": 66162923.0,
      "step": 31880
    },
    {
      "entropy": 5.388528633117676,
      "epoch": 2.8923258345428158,
      "grad_norm": 1.125,
      "learning_rate": 0.00041692122318065885,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20276117324829102,
      "num_tokens": 66173108.0,
      "step": 31885
    },
    {
      "entropy": 5.356553745269776,
      "epoch": 2.8927793904209,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041689611387013363,
      "loss": 4.9513,
      "mean_token_accuracy": 0.20713712722063066,
      "num_tokens": 66185280.0,
      "step": 31890
    },
    {
      "entropy": 5.3344158172607425,
      "epoch": 2.8932329462989843,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004168710016251606,
      "loss": 4.972,
      "mean_token_accuracy": 0.20237459540367125,
      "num_tokens": 66196020.0,
      "step": 31895
    },
    {
      "entropy": 5.275509166717529,
      "epoch": 2.893686502177068,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004168458864462592,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2050391986966133,
      "num_tokens": 66205753.0,
      "step": 31900
    },
    {
      "entropy": 5.365378713607788,
      "epoch": 2.8941400580551524,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041682076833394864,
      "loss": 5.0904,
      "mean_token_accuracy": 0.20152926146984101,
      "num_tokens": 66218415.0,
      "step": 31905
    },
    {
      "entropy": 5.385711717605591,
      "epoch": 2.8945936139332367,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041679564728874845,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20199463069438933,
      "num_tokens": 66228090.0,
      "step": 31910
    },
    {
      "entropy": 5.342091512680054,
      "epoch": 2.8950471698113205,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004167705233111782,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2085132494568825,
      "num_tokens": 66237587.0,
      "step": 31915
    },
    {
      "entropy": 5.359386205673218,
      "epoch": 2.8955007256894048,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041674539640175733,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20804242491722108,
      "num_tokens": 66248162.0,
      "step": 31920
    },
    {
      "entropy": 5.322322845458984,
      "epoch": 2.895954281567489,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004167202665610055,
      "loss": 4.9525,
      "mean_token_accuracy": 0.20880566090345382,
      "num_tokens": 66258370.0,
      "step": 31925
    },
    {
      "entropy": 5.345619583129883,
      "epoch": 2.8964078374455733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041669513378944254,
      "loss": 5.0285,
      "mean_token_accuracy": 0.19817960113286973,
      "num_tokens": 66269267.0,
      "step": 31930
    },
    {
      "entropy": 5.43332200050354,
      "epoch": 2.8968613933236576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041666999808758815,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20356244295835496,
      "num_tokens": 66280077.0,
      "step": 31935
    },
    {
      "entropy": 5.341119718551636,
      "epoch": 2.897314949201742,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000416644859455962,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2063751459121704,
      "num_tokens": 66291678.0,
      "step": 31940
    },
    {
      "entropy": 5.328046655654907,
      "epoch": 2.897768505079826,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004166197178950841,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20132209360599518,
      "num_tokens": 66302046.0,
      "step": 31945
    },
    {
      "entropy": 5.331369781494141,
      "epoch": 2.89822206095791,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041659457340547436,
      "loss": 4.9984,
      "mean_token_accuracy": 0.20747543424367904,
      "num_tokens": 66311888.0,
      "step": 31950
    },
    {
      "entropy": 5.33133749961853,
      "epoch": 2.898675616835994,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041656942598765275,
      "loss": 4.9791,
      "mean_token_accuracy": 0.212832410633564,
      "num_tokens": 66321507.0,
      "step": 31955
    },
    {
      "entropy": 5.286359882354736,
      "epoch": 2.8991291727140784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041654427564213924,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21190388649702072,
      "num_tokens": 66331458.0,
      "step": 31960
    },
    {
      "entropy": 5.332270956039428,
      "epoch": 2.8995827285921627,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041651912236945416,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20865980833768843,
      "num_tokens": 66340875.0,
      "step": 31965
    },
    {
      "entropy": 5.353117561340332,
      "epoch": 2.9000362844702465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041649396617011753,
      "loss": 5.0682,
      "mean_token_accuracy": 0.19800521731376647,
      "num_tokens": 66350784.0,
      "step": 31970
    },
    {
      "entropy": 5.364161729812622,
      "epoch": 2.900489840348331,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004164688070446496,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19805718213319778,
      "num_tokens": 66361591.0,
      "step": 31975
    },
    {
      "entropy": 5.378185224533081,
      "epoch": 2.900943396226415,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004164436449935707,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2088303118944168,
      "num_tokens": 66371540.0,
      "step": 31980
    },
    {
      "entropy": 5.362077283859253,
      "epoch": 2.9013969521044993,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041641848001740126,
      "loss": 5.0157,
      "mean_token_accuracy": 0.2065187633037567,
      "num_tokens": 66382053.0,
      "step": 31985
    },
    {
      "entropy": 5.325444030761719,
      "epoch": 2.9018505079825836,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004163933121166615,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21078934669494628,
      "num_tokens": 66392563.0,
      "step": 31990
    },
    {
      "entropy": 5.326361036300659,
      "epoch": 2.902304063860668,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000416368141291872,
      "loss": 4.9332,
      "mean_token_accuracy": 0.20743950605392455,
      "num_tokens": 66403582.0,
      "step": 31995
    },
    {
      "entropy": 5.290842342376709,
      "epoch": 2.9027576197387517,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041634296754355346,
      "loss": 4.9172,
      "mean_token_accuracy": 0.20775769799947738,
      "num_tokens": 66413713.0,
      "step": 32000
    },
    {
      "entropy": 5.359246253967285,
      "epoch": 2.903211175616836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004163177908722262,
      "loss": 4.9941,
      "mean_token_accuracy": 0.20469143390655517,
      "num_tokens": 66424802.0,
      "step": 32005
    },
    {
      "entropy": 5.311826419830322,
      "epoch": 2.90366473149492,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041629261127841105,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21013171076774598,
      "num_tokens": 66434716.0,
      "step": 32010
    },
    {
      "entropy": 5.2865136623382565,
      "epoch": 2.9041182873730045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041626742876262876,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20352882742881775,
      "num_tokens": 66445725.0,
      "step": 32015
    },
    {
      "entropy": 5.401998281478882,
      "epoch": 2.9045718432510883,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041624224332539997,
      "loss": 5.0516,
      "mean_token_accuracy": 0.19550885558128356,
      "num_tokens": 66457492.0,
      "step": 32020
    },
    {
      "entropy": 5.379173851013183,
      "epoch": 2.9050253991291726,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004162170549672455,
      "loss": 4.962,
      "mean_token_accuracy": 0.2088708385825157,
      "num_tokens": 66467490.0,
      "step": 32025
    },
    {
      "entropy": 5.39162049293518,
      "epoch": 2.905478955007257,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004161918636886865,
      "loss": 5.0972,
      "mean_token_accuracy": 0.1958886370062828,
      "num_tokens": 66478160.0,
      "step": 32030
    },
    {
      "entropy": 5.390858697891235,
      "epoch": 2.905932510885341,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004161666694902437,
      "loss": 4.998,
      "mean_token_accuracy": 0.19981719851493834,
      "num_tokens": 66488181.0,
      "step": 32035
    },
    {
      "entropy": 5.363636064529419,
      "epoch": 2.9063860667634254,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004161414723724383,
      "loss": 5.0544,
      "mean_token_accuracy": 0.20503250360488892,
      "num_tokens": 66498507.0,
      "step": 32040
    },
    {
      "entropy": 5.284357929229737,
      "epoch": 2.9068396226415096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004161162723357912,
      "loss": 4.9148,
      "mean_token_accuracy": 0.20977286398410797,
      "num_tokens": 66507752.0,
      "step": 32045
    },
    {
      "entropy": 5.303616046905518,
      "epoch": 2.9072931785195935,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004160910693808236,
      "loss": 5.004,
      "mean_token_accuracy": 0.2017525091767311,
      "num_tokens": 66519360.0,
      "step": 32050
    },
    {
      "entropy": 5.43273606300354,
      "epoch": 2.9077467343976777,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041606586350805686,
      "loss": 5.078,
      "mean_token_accuracy": 0.20391293317079545,
      "num_tokens": 66529790.0,
      "step": 32055
    },
    {
      "entropy": 5.365130186080933,
      "epoch": 2.908200290275762,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000416040654718012,
      "loss": 4.9405,
      "mean_token_accuracy": 0.21414923965930938,
      "num_tokens": 66539617.0,
      "step": 32060
    },
    {
      "entropy": 5.3080915927886965,
      "epoch": 2.9086538461538463,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041601544301121045,
      "loss": 4.9675,
      "mean_token_accuracy": 0.20629144906997682,
      "num_tokens": 66549861.0,
      "step": 32065
    },
    {
      "entropy": 5.406466627120972,
      "epoch": 2.90910740203193,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004159902283881737,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20067891776561736,
      "num_tokens": 66560388.0,
      "step": 32070
    },
    {
      "entropy": 5.4347076416015625,
      "epoch": 2.9095609579100143,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000415965010849423,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19858078211545943,
      "num_tokens": 66570795.0,
      "step": 32075
    },
    {
      "entropy": 5.386423110961914,
      "epoch": 2.9100145137880986,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041593979039548,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20216501802206038,
      "num_tokens": 66580268.0,
      "step": 32080
    },
    {
      "entropy": 5.395645809173584,
      "epoch": 2.910468069666183,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004159145670268662,
      "loss": 5.0588,
      "mean_token_accuracy": 0.2030414506793022,
      "num_tokens": 66591192.0,
      "step": 32085
    },
    {
      "entropy": 5.3955755710601805,
      "epoch": 2.910921625544267,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004158893407441033,
      "loss": 5.0767,
      "mean_token_accuracy": 0.20822654217481612,
      "num_tokens": 66600768.0,
      "step": 32090
    },
    {
      "entropy": 5.465906620025635,
      "epoch": 2.9113751814223514,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041586411154771285,
      "loss": 5.049,
      "mean_token_accuracy": 0.1993163898587227,
      "num_tokens": 66612207.0,
      "step": 32095
    },
    {
      "entropy": 5.281161212921143,
      "epoch": 2.9118287373004357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041583887943821667,
      "loss": 4.9353,
      "mean_token_accuracy": 0.20763570368289946,
      "num_tokens": 66622175.0,
      "step": 32100
    },
    {
      "entropy": 5.320460605621338,
      "epoch": 2.9122822931785195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004158136444161367,
      "loss": 4.9863,
      "mean_token_accuracy": 0.200809808075428,
      "num_tokens": 66633936.0,
      "step": 32105
    },
    {
      "entropy": 5.411285924911499,
      "epoch": 2.9127358490566038,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041578840648199455,
      "loss": 5.0457,
      "mean_token_accuracy": 0.2005571946501732,
      "num_tokens": 66643502.0,
      "step": 32110
    },
    {
      "entropy": 5.366814374923706,
      "epoch": 2.913189404934688,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004157631656363124,
      "loss": 5.0134,
      "mean_token_accuracy": 0.20601053088903426,
      "num_tokens": 66654890.0,
      "step": 32115
    },
    {
      "entropy": 5.401779365539551,
      "epoch": 2.9136429608127723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000415737921879612,
      "loss": 5.0962,
      "mean_token_accuracy": 0.1926824390888214,
      "num_tokens": 66667037.0,
      "step": 32120
    },
    {
      "entropy": 5.34242091178894,
      "epoch": 2.914096516690856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041571267521241554,
      "loss": 4.9745,
      "mean_token_accuracy": 0.20403825044631957,
      "num_tokens": 66677291.0,
      "step": 32125
    },
    {
      "entropy": 5.31213231086731,
      "epoch": 2.9145500725689404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004156874256352451,
      "loss": 4.9524,
      "mean_token_accuracy": 0.20055051296949386,
      "num_tokens": 66687421.0,
      "step": 32130
    },
    {
      "entropy": 5.476664781570435,
      "epoch": 2.9150036284470247,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004156621731486228,
      "loss": 5.0736,
      "mean_token_accuracy": 0.19713356345891953,
      "num_tokens": 66697190.0,
      "step": 32135
    },
    {
      "entropy": 5.390998888015747,
      "epoch": 2.915457184325109,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000415636917753071,
      "loss": 5.0681,
      "mean_token_accuracy": 0.20008155405521394,
      "num_tokens": 66707255.0,
      "step": 32140
    },
    {
      "entropy": 5.266979598999024,
      "epoch": 2.915910740203193,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004156116594491119,
      "loss": 4.9128,
      "mean_token_accuracy": 0.20420518517494202,
      "num_tokens": 66716451.0,
      "step": 32145
    },
    {
      "entropy": 5.326523971557617,
      "epoch": 2.9163642960812775,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004155863982372678,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20645487010478974,
      "num_tokens": 66727157.0,
      "step": 32150
    },
    {
      "entropy": 5.4475545406341555,
      "epoch": 2.9168178519593613,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041556113411806117,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20377058833837508,
      "num_tokens": 66737523.0,
      "step": 32155
    },
    {
      "entropy": 5.381815433502197,
      "epoch": 2.9172714078374455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041553586709201453,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2043487772345543,
      "num_tokens": 66748183.0,
      "step": 32160
    },
    {
      "entropy": 5.355172634124756,
      "epoch": 2.91772496371553,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004155105971596503,
      "loss": 4.9757,
      "mean_token_accuracy": 0.20705778449773787,
      "num_tokens": 66758794.0,
      "step": 32165
    },
    {
      "entropy": 5.423040676116943,
      "epoch": 2.918178519593614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041548532432149105,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20810611844062804,
      "num_tokens": 66769916.0,
      "step": 32170
    },
    {
      "entropy": 5.36120457649231,
      "epoch": 2.918632075471698,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004154600485780595,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2101347416639328,
      "num_tokens": 66779658.0,
      "step": 32175
    },
    {
      "entropy": 5.301400995254516,
      "epoch": 2.919085631349782,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041543476992987843,
      "loss": 4.951,
      "mean_token_accuracy": 0.20982602685689927,
      "num_tokens": 66788909.0,
      "step": 32180
    },
    {
      "entropy": 5.4446608543396,
      "epoch": 2.9195391872278664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041540948837747056,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19978921264410018,
      "num_tokens": 66799519.0,
      "step": 32185
    },
    {
      "entropy": 5.328560733795166,
      "epoch": 2.9199927431059507,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004153842039213586,
      "loss": 4.9321,
      "mean_token_accuracy": 0.20892735421657563,
      "num_tokens": 66808915.0,
      "step": 32190
    },
    {
      "entropy": 5.369980573654175,
      "epoch": 2.920446298984035,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004153589165620656,
      "loss": 5.0553,
      "mean_token_accuracy": 0.19760142862796784,
      "num_tokens": 66818819.0,
      "step": 32195
    },
    {
      "entropy": 5.243846321105957,
      "epoch": 2.9208998548621192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041533362630011435,
      "loss": 4.8697,
      "mean_token_accuracy": 0.219241102039814,
      "num_tokens": 66829274.0,
      "step": 32200
    },
    {
      "entropy": 5.3757514476776125,
      "epoch": 2.921353410740203,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000415308333136028,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20549867302179337,
      "num_tokens": 66839659.0,
      "step": 32205
    },
    {
      "entropy": 5.3885290145874025,
      "epoch": 2.9218069666182873,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041528303707032957,
      "loss": 5.0731,
      "mean_token_accuracy": 0.19927890449762345,
      "num_tokens": 66850415.0,
      "step": 32210
    },
    {
      "entropy": 5.351044511795044,
      "epoch": 2.9222605224963716,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004152577381035422,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22051679193973542,
      "num_tokens": 66860663.0,
      "step": 32215
    },
    {
      "entropy": 5.407913637161255,
      "epoch": 2.922714078374456,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041523243623618907,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20212013572454451,
      "num_tokens": 66871473.0,
      "step": 32220
    },
    {
      "entropy": 5.292897176742554,
      "epoch": 2.9231676342525397,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041520713146879333,
      "loss": 4.9546,
      "mean_token_accuracy": 0.20864801853895187,
      "num_tokens": 66881609.0,
      "step": 32225
    },
    {
      "entropy": 5.391390419006347,
      "epoch": 2.923621190130624,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041518182380187853,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2038584902882576,
      "num_tokens": 66891780.0,
      "step": 32230
    },
    {
      "entropy": 5.407641506195068,
      "epoch": 2.924074746008708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041515651323596774,
      "loss": 5.1111,
      "mean_token_accuracy": 0.1992737516760826,
      "num_tokens": 66902251.0,
      "step": 32235
    },
    {
      "entropy": 5.380971670150757,
      "epoch": 2.9245283018867925,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041513119977158464,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20232230573892593,
      "num_tokens": 66912538.0,
      "step": 32240
    },
    {
      "entropy": 5.372594213485717,
      "epoch": 2.9249818577648767,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004151058834092526,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20932898670434952,
      "num_tokens": 66922648.0,
      "step": 32245
    },
    {
      "entropy": 5.328006410598755,
      "epoch": 2.925435413642961,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041508056414949515,
      "loss": 4.9204,
      "mean_token_accuracy": 0.20556726753711702,
      "num_tokens": 66933115.0,
      "step": 32250
    },
    {
      "entropy": 5.429669618606567,
      "epoch": 2.9258889695210453,
      "grad_norm": 1.015625,
      "learning_rate": 0.000415055241992836,
      "loss": 5.1466,
      "mean_token_accuracy": 0.1928657129406929,
      "num_tokens": 66943727.0,
      "step": 32255
    },
    {
      "entropy": 5.3638664245605465,
      "epoch": 2.926342525399129,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004150299169397987,
      "loss": 4.9571,
      "mean_token_accuracy": 0.20725325793027877,
      "num_tokens": 66953488.0,
      "step": 32260
    },
    {
      "entropy": 5.410525798797607,
      "epoch": 2.9267960812772134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004150045889909071,
      "loss": 5.1157,
      "mean_token_accuracy": 0.19255953133106232,
      "num_tokens": 66964586.0,
      "step": 32265
    },
    {
      "entropy": 5.3645953178405765,
      "epoch": 2.9272496371552976,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004149792581466848,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20147883743047715,
      "num_tokens": 66975032.0,
      "step": 32270
    },
    {
      "entropy": 5.377267980575562,
      "epoch": 2.9277031930333814,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041495392440765586,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20676131844520568,
      "num_tokens": 66984910.0,
      "step": 32275
    },
    {
      "entropy": 5.302928447723389,
      "epoch": 2.9281567489114657,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041492858777434407,
      "loss": 4.82,
      "mean_token_accuracy": 0.2146179899573326,
      "num_tokens": 66995238.0,
      "step": 32280
    },
    {
      "entropy": 5.32975926399231,
      "epoch": 2.92861030478955,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004149032482472734,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2140320897102356,
      "num_tokens": 67005596.0,
      "step": 32285
    },
    {
      "entropy": 5.411073923110962,
      "epoch": 2.9290638606676342,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041487790582696803,
      "loss": 5.1629,
      "mean_token_accuracy": 0.19110160171985627,
      "num_tokens": 67017312.0,
      "step": 32290
    },
    {
      "entropy": 5.346672630310058,
      "epoch": 2.9295174165457185,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004148525605139518,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21267084926366805,
      "num_tokens": 67026692.0,
      "step": 32295
    },
    {
      "entropy": 5.278961992263794,
      "epoch": 2.9299709724238028,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041482721230874897,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20721954256296157,
      "num_tokens": 67036801.0,
      "step": 32300
    },
    {
      "entropy": 5.286829614639283,
      "epoch": 2.930424528301887,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041480186121188383,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21115856617689133,
      "num_tokens": 67047262.0,
      "step": 32305
    },
    {
      "entropy": 5.415988731384277,
      "epoch": 2.930878084179971,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004147765072238806,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20223256796598435,
      "num_tokens": 67058019.0,
      "step": 32310
    },
    {
      "entropy": 5.434537649154663,
      "epoch": 2.931331640058055,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004147511503452635,
      "loss": 5.0767,
      "mean_token_accuracy": 0.20228389501571656,
      "num_tokens": 67068844.0,
      "step": 32315
    },
    {
      "entropy": 5.3639901638031,
      "epoch": 2.9317851959361394,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041472579057655703,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21286086440086366,
      "num_tokens": 67077400.0,
      "step": 32320
    },
    {
      "entropy": 5.3391368865966795,
      "epoch": 2.9322387518142237,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004147004279182856,
      "loss": 5.1365,
      "mean_token_accuracy": 0.19398838579654692,
      "num_tokens": 67088352.0,
      "step": 32325
    },
    {
      "entropy": 5.415532445907592,
      "epoch": 2.9326923076923075,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004146750623709737,
      "loss": 5.0423,
      "mean_token_accuracy": 0.2022481679916382,
      "num_tokens": 67097879.0,
      "step": 32330
    },
    {
      "entropy": 5.376211309432984,
      "epoch": 2.9331458635703918,
      "grad_norm": 1.25,
      "learning_rate": 0.00041464969393514595,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2096438929438591,
      "num_tokens": 67108355.0,
      "step": 32335
    },
    {
      "entropy": 5.322367811203003,
      "epoch": 2.933599419448476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004146243226113269,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20070355385541916,
      "num_tokens": 67118348.0,
      "step": 32340
    },
    {
      "entropy": 5.24345760345459,
      "epoch": 2.9340529753265603,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041459894840004135,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2120596006512642,
      "num_tokens": 67128944.0,
      "step": 32345
    },
    {
      "entropy": 5.390466642379761,
      "epoch": 2.9345065312046446,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041457357130181396,
      "loss": 5.0611,
      "mean_token_accuracy": 0.19882676899433135,
      "num_tokens": 67139365.0,
      "step": 32350
    },
    {
      "entropy": 5.3851478099823,
      "epoch": 2.934960087082729,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041454819131716945,
      "loss": 5.0004,
      "mean_token_accuracy": 0.19818630963563919,
      "num_tokens": 67150406.0,
      "step": 32355
    },
    {
      "entropy": 5.4284039497375485,
      "epoch": 2.9354136429608126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041452280844663295,
      "loss": 5.1679,
      "mean_token_accuracy": 0.19812532663345336,
      "num_tokens": 67161631.0,
      "step": 32360
    },
    {
      "entropy": 5.303620481491089,
      "epoch": 2.935867198838897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004144974226907291,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20953986197710037,
      "num_tokens": 67171735.0,
      "step": 32365
    },
    {
      "entropy": 5.373064565658569,
      "epoch": 2.936320754716981,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000414472034049983,
      "loss": 5.0371,
      "mean_token_accuracy": 0.1990324780344963,
      "num_tokens": 67181616.0,
      "step": 32370
    },
    {
      "entropy": 5.402244520187378,
      "epoch": 2.9367743105950654,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004144466425249197,
      "loss": 4.9846,
      "mean_token_accuracy": 0.2060014620423317,
      "num_tokens": 67192519.0,
      "step": 32375
    },
    {
      "entropy": 5.371507263183593,
      "epoch": 2.9372278664731493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004144212481160643,
      "loss": 4.9993,
      "mean_token_accuracy": 0.20236235558986665,
      "num_tokens": 67202805.0,
      "step": 32380
    },
    {
      "entropy": 5.367984247207642,
      "epoch": 2.9376814223512335,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004143958508239419,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20757365077733994,
      "num_tokens": 67212778.0,
      "step": 32385
    },
    {
      "entropy": 5.333669805526734,
      "epoch": 2.938134978229318,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004143704506490779,
      "loss": 5.045,
      "mean_token_accuracy": 0.20445097386837005,
      "num_tokens": 67222588.0,
      "step": 32390
    },
    {
      "entropy": 5.423421812057495,
      "epoch": 2.938588534107402,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004143450475919975,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20247101336717604,
      "num_tokens": 67233022.0,
      "step": 32395
    },
    {
      "entropy": 5.3568565368652346,
      "epoch": 2.9390420899854863,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004143196416532259,
      "loss": 4.9905,
      "mean_token_accuracy": 0.20844801217317582,
      "num_tokens": 67242631.0,
      "step": 32400
    },
    {
      "entropy": 5.285913610458374,
      "epoch": 2.9394956458635706,
      "grad_norm": 1.0,
      "learning_rate": 0.0004142942328332887,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2096813678741455,
      "num_tokens": 67252376.0,
      "step": 32405
    },
    {
      "entropy": 5.328869104385376,
      "epoch": 2.9399492017416544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004142688211327111,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2032950758934021,
      "num_tokens": 67262209.0,
      "step": 32410
    },
    {
      "entropy": 5.417429351806641,
      "epoch": 2.9404027576197387,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004142434065520189,
      "loss": 5.0227,
      "mean_token_accuracy": 0.1974325105547905,
      "num_tokens": 67272893.0,
      "step": 32415
    },
    {
      "entropy": 5.3539618968963625,
      "epoch": 2.940856313497823,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004142179890917376,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20224272310733796,
      "num_tokens": 67284017.0,
      "step": 32420
    },
    {
      "entropy": 5.279170656204224,
      "epoch": 2.941309869375907,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041419256875239275,
      "loss": 4.9705,
      "mean_token_accuracy": 0.2097451478242874,
      "num_tokens": 67293674.0,
      "step": 32425
    },
    {
      "entropy": 5.324466800689697,
      "epoch": 2.941763425253991,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004141671455345101,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20652943700551987,
      "num_tokens": 67303759.0,
      "step": 32430
    },
    {
      "entropy": 5.430917692184448,
      "epoch": 2.9422169811320753,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041414171943861543,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2019360437989235,
      "num_tokens": 67314962.0,
      "step": 32435
    },
    {
      "entropy": 5.428606367111206,
      "epoch": 2.9426705370101596,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004141162904652346,
      "loss": 5.1107,
      "mean_token_accuracy": 0.19670023024082184,
      "num_tokens": 67325210.0,
      "step": 32440
    },
    {
      "entropy": 5.349099826812744,
      "epoch": 2.943124092888244,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004140908586148934,
      "loss": 5.0429,
      "mean_token_accuracy": 0.19892754405736923,
      "num_tokens": 67335110.0,
      "step": 32445
    },
    {
      "entropy": 5.33529839515686,
      "epoch": 2.943577648766328,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004140654238881178,
      "loss": 5.1074,
      "mean_token_accuracy": 0.19610674679279327,
      "num_tokens": 67345872.0,
      "step": 32450
    },
    {
      "entropy": 5.341188240051269,
      "epoch": 2.9440312046444124,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041403998628543375,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2018139809370041,
      "num_tokens": 67356532.0,
      "step": 32455
    },
    {
      "entropy": 5.340752029418946,
      "epoch": 2.9444847605224966,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041401454580736744,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21190954744815826,
      "num_tokens": 67366603.0,
      "step": 32460
    },
    {
      "entropy": 5.4689093112945555,
      "epoch": 2.9449383164005805,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004139891024544448,
      "loss": 5.1496,
      "mean_token_accuracy": 0.19797108173370362,
      "num_tokens": 67377487.0,
      "step": 32465
    },
    {
      "entropy": 5.415403509140015,
      "epoch": 2.9453918722786647,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004139636562271921,
      "loss": 5.0904,
      "mean_token_accuracy": 0.20175853967666627,
      "num_tokens": 67387704.0,
      "step": 32470
    },
    {
      "entropy": 5.35092601776123,
      "epoch": 2.945845428156749,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041393820712613567,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20259959846735,
      "num_tokens": 67398409.0,
      "step": 32475
    },
    {
      "entropy": 5.377112627029419,
      "epoch": 2.9462989840348333,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041391275515180166,
      "loss": 4.941,
      "mean_token_accuracy": 0.20536339282989502,
      "num_tokens": 67408905.0,
      "step": 32480
    },
    {
      "entropy": 5.330749988555908,
      "epoch": 2.946752539912917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004138873003047165,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21897065341472627,
      "num_tokens": 67418466.0,
      "step": 32485
    },
    {
      "entropy": 5.269274091720581,
      "epoch": 2.9472060957910013,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004138618425854066,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21039409637451173,
      "num_tokens": 67428597.0,
      "step": 32490
    },
    {
      "entropy": 5.384949350357056,
      "epoch": 2.9476596516690856,
      "grad_norm": 1.125,
      "learning_rate": 0.00041383638199439824,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2058815985918045,
      "num_tokens": 67439158.0,
      "step": 32495
    },
    {
      "entropy": 5.425584030151367,
      "epoch": 2.94811320754717,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041381091853221835,
      "loss": 5.0166,
      "mean_token_accuracy": 0.19851868748664855,
      "num_tokens": 67449540.0,
      "step": 32500
    },
    {
      "entropy": 5.3440886497497555,
      "epoch": 2.948566763425254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004137854521993932,
      "loss": 4.9693,
      "mean_token_accuracy": 0.20833881795406342,
      "num_tokens": 67459385.0,
      "step": 32505
    },
    {
      "entropy": 5.364635419845581,
      "epoch": 2.9490203193033384,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041375998299644946,
      "loss": 5.0487,
      "mean_token_accuracy": 0.2049234226346016,
      "num_tokens": 67469597.0,
      "step": 32510
    },
    {
      "entropy": 5.301775169372559,
      "epoch": 2.9494738751814222,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000413734510923914,
      "loss": 4.9308,
      "mean_token_accuracy": 0.20617831796407698,
      "num_tokens": 67479271.0,
      "step": 32515
    },
    {
      "entropy": 5.272887372970581,
      "epoch": 2.9499274310595065,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004137090359823135,
      "loss": 4.936,
      "mean_token_accuracy": 0.21072113513946533,
      "num_tokens": 67488997.0,
      "step": 32520
    },
    {
      "entropy": 5.342120265960693,
      "epoch": 2.9503809869375908,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041368355817217474,
      "loss": 4.9773,
      "mean_token_accuracy": 0.207565139234066,
      "num_tokens": 67499507.0,
      "step": 32525
    },
    {
      "entropy": 5.36089596748352,
      "epoch": 2.950834542815675,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041365807749402467,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20658674389123916,
      "num_tokens": 67509916.0,
      "step": 32530
    },
    {
      "entropy": 5.339598941802978,
      "epoch": 2.951288098693759,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004136325939483902,
      "loss": 4.9603,
      "mean_token_accuracy": 0.1983932748436928,
      "num_tokens": 67519759.0,
      "step": 32535
    },
    {
      "entropy": 5.300631999969482,
      "epoch": 2.951741654571843,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004136071075357984,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21075653433799743,
      "num_tokens": 67529261.0,
      "step": 32540
    },
    {
      "entropy": 5.307101249694824,
      "epoch": 2.9521952104499274,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004135816182567763,
      "loss": 4.9631,
      "mean_token_accuracy": 0.20697320401668548,
      "num_tokens": 67539084.0,
      "step": 32545
    },
    {
      "entropy": 5.384823513031006,
      "epoch": 2.9526487663280117,
      "grad_norm": 1.03125,
      "learning_rate": 0.000413556126111851,
      "loss": 5.0421,
      "mean_token_accuracy": 0.20501959472894668,
      "num_tokens": 67548551.0,
      "step": 32550
    },
    {
      "entropy": 5.373802661895752,
      "epoch": 2.953102322206096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004135306311015497,
      "loss": 4.9811,
      "mean_token_accuracy": 0.201568304002285,
      "num_tokens": 67558364.0,
      "step": 32555
    },
    {
      "entropy": 5.374468803405762,
      "epoch": 2.95355587808418,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004135051332263996,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20935927480459213,
      "num_tokens": 67569257.0,
      "step": 32560
    },
    {
      "entropy": 5.340837574005127,
      "epoch": 2.954009433962264,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041347963248692814,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20091799199581145,
      "num_tokens": 67580135.0,
      "step": 32565
    },
    {
      "entropy": 5.3731366157531735,
      "epoch": 2.9544629898403483,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004134541288836626,
      "loss": 4.9958,
      "mean_token_accuracy": 0.2078840896487236,
      "num_tokens": 67589705.0,
      "step": 32570
    },
    {
      "entropy": 5.361728954315185,
      "epoch": 2.9549165457184325,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041342862241713034,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2047303467988968,
      "num_tokens": 67599648.0,
      "step": 32575
    },
    {
      "entropy": 5.368472480773926,
      "epoch": 2.955370101596517,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041340311308785886,
      "loss": 4.9682,
      "mean_token_accuracy": 0.20415790230035782,
      "num_tokens": 67609624.0,
      "step": 32580
    },
    {
      "entropy": 5.279648637771606,
      "epoch": 2.9558236574746006,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041337760089637577,
      "loss": 4.989,
      "mean_token_accuracy": 0.20785954743623733,
      "num_tokens": 67619596.0,
      "step": 32585
    },
    {
      "entropy": 5.329959917068481,
      "epoch": 2.956277213352685,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041335208584320865,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20121295154094695,
      "num_tokens": 67629667.0,
      "step": 32590
    },
    {
      "entropy": 5.388042402267456,
      "epoch": 2.956730769230769,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004133265679288851,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20246201604604722,
      "num_tokens": 67640048.0,
      "step": 32595
    },
    {
      "entropy": 5.374866342544555,
      "epoch": 2.9571843251088534,
      "grad_norm": 1.0,
      "learning_rate": 0.0004133010471539329,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20159481316804886,
      "num_tokens": 67650731.0,
      "step": 32600
    },
    {
      "entropy": 5.358546590805053,
      "epoch": 2.9576378809869377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041327552351887983,
      "loss": 4.9978,
      "mean_token_accuracy": 0.20390666127204896,
      "num_tokens": 67661696.0,
      "step": 32605
    },
    {
      "entropy": 5.365850591659546,
      "epoch": 2.958091436865022,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041324999702425366,
      "loss": 5.0153,
      "mean_token_accuracy": 0.2004029080271721,
      "num_tokens": 67671758.0,
      "step": 32610
    },
    {
      "entropy": 5.337644815444946,
      "epoch": 2.9585449927431062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004132244676705823,
      "loss": 4.9672,
      "mean_token_accuracy": 0.20783085823059083,
      "num_tokens": 67682488.0,
      "step": 32615
    },
    {
      "entropy": 5.399378395080566,
      "epoch": 2.95899854862119,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004131989354583938,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20507284849882126,
      "num_tokens": 67692945.0,
      "step": 32620
    },
    {
      "entropy": 5.354639768600464,
      "epoch": 2.9594521044992743,
      "grad_norm": 1.03125,
      "learning_rate": 0.000413173400388216,
      "loss": 4.921,
      "mean_token_accuracy": 0.20879835337400438,
      "num_tokens": 67702924.0,
      "step": 32625
    },
    {
      "entropy": 5.338720417022705,
      "epoch": 2.9599056603773586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004131478624605771,
      "loss": 5.0158,
      "mean_token_accuracy": 0.20126423090696335,
      "num_tokens": 67714284.0,
      "step": 32630
    },
    {
      "entropy": 5.363140392303467,
      "epoch": 2.9603592162554424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004131223216760052,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19547925889492035,
      "num_tokens": 67725902.0,
      "step": 32635
    },
    {
      "entropy": 5.3852311134338375,
      "epoch": 2.9608127721335267,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004130967780350286,
      "loss": 4.9899,
      "mean_token_accuracy": 0.19851239770650864,
      "num_tokens": 67736268.0,
      "step": 32640
    },
    {
      "entropy": 5.448613023757934,
      "epoch": 2.961266328011611,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004130712315381753,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20815595984458923,
      "num_tokens": 67745709.0,
      "step": 32645
    },
    {
      "entropy": 5.410993528366089,
      "epoch": 2.961719883889695,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041304568218597385,
      "loss": 5.0019,
      "mean_token_accuracy": 0.2000914618372917,
      "num_tokens": 67756092.0,
      "step": 32650
    },
    {
      "entropy": 5.382706022262573,
      "epoch": 2.9621734397677795,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041302012997895246,
      "loss": 5.009,
      "mean_token_accuracy": 0.20134355872869492,
      "num_tokens": 67766224.0,
      "step": 32655
    },
    {
      "entropy": 5.4464154720306395,
      "epoch": 2.9626269956458637,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004129945749176395,
      "loss": 5.1505,
      "mean_token_accuracy": 0.19994537830352782,
      "num_tokens": 67776787.0,
      "step": 32660
    },
    {
      "entropy": 5.351326751708984,
      "epoch": 2.963080551523948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041296901700256375,
      "loss": 4.9854,
      "mean_token_accuracy": 0.2062183141708374,
      "num_tokens": 67787924.0,
      "step": 32665
    },
    {
      "entropy": 5.365162134170532,
      "epoch": 2.963534107402032,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004129434562342535,
      "loss": 4.9977,
      "mean_token_accuracy": 0.19683745354413987,
      "num_tokens": 67798544.0,
      "step": 32670
    },
    {
      "entropy": 5.334612560272217,
      "epoch": 2.963987663280116,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041291789261323734,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2041091501712799,
      "num_tokens": 67809528.0,
      "step": 32675
    },
    {
      "entropy": 5.25989260673523,
      "epoch": 2.9644412191582004,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041289232614004416,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21629827320575715,
      "num_tokens": 67820263.0,
      "step": 32680
    },
    {
      "entropy": 5.442530870437622,
      "epoch": 2.9648947750362846,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004128667568152024,
      "loss": 5.061,
      "mean_token_accuracy": 0.20147672891616822,
      "num_tokens": 67830687.0,
      "step": 32685
    },
    {
      "entropy": 5.438401460647583,
      "epoch": 2.9653483309143684,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041284118463924107,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19971842020750047,
      "num_tokens": 67841714.0,
      "step": 32690
    },
    {
      "entropy": 5.316351842880249,
      "epoch": 2.9658018867924527,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004128156096126889,
      "loss": 4.9882,
      "mean_token_accuracy": 0.20308611392974854,
      "num_tokens": 67851952.0,
      "step": 32695
    },
    {
      "entropy": 5.255450057983398,
      "epoch": 2.966255442670537,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041279003173607477,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2066622331738472,
      "num_tokens": 67861958.0,
      "step": 32700
    },
    {
      "entropy": 5.325444889068604,
      "epoch": 2.9667089985486212,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004127644510099278,
      "loss": 4.8994,
      "mean_token_accuracy": 0.21302953958511353,
      "num_tokens": 67871654.0,
      "step": 32705
    },
    {
      "entropy": 5.291216087341309,
      "epoch": 2.9671625544267055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004127388674347767,
      "loss": 4.962,
      "mean_token_accuracy": 0.21156464964151384,
      "num_tokens": 67881678.0,
      "step": 32710
    },
    {
      "entropy": 5.333769083023071,
      "epoch": 2.96761611030479,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041271328101115075,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20253877937793732,
      "num_tokens": 67892663.0,
      "step": 32715
    },
    {
      "entropy": 5.506520557403564,
      "epoch": 2.9680696661828736,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041268769173957915,
      "loss": 5.1834,
      "mean_token_accuracy": 0.19482302367687226,
      "num_tokens": 67903308.0,
      "step": 32720
    },
    {
      "entropy": 5.394654369354248,
      "epoch": 2.968523222060958,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041266209962059097,
      "loss": 4.9866,
      "mean_token_accuracy": 0.20763224363327026,
      "num_tokens": 67913092.0,
      "step": 32725
    },
    {
      "entropy": 5.330311727523804,
      "epoch": 2.968976777939042,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004126365046547154,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2118850141763687,
      "num_tokens": 67923708.0,
      "step": 32730
    },
    {
      "entropy": 5.2754779815673825,
      "epoch": 2.9694303338171264,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004126109068424819,
      "loss": 4.9758,
      "mean_token_accuracy": 0.2086584076285362,
      "num_tokens": 67934164.0,
      "step": 32735
    },
    {
      "entropy": 5.4248240947723385,
      "epoch": 2.96988388969521,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004125853061844198,
      "loss": 5.088,
      "mean_token_accuracy": 0.1991724908351898,
      "num_tokens": 67943162.0,
      "step": 32740
    },
    {
      "entropy": 5.429990243911743,
      "epoch": 2.9703374455732945,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004125597026810585,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20408604443073272,
      "num_tokens": 67953943.0,
      "step": 32745
    },
    {
      "entropy": 5.289671945571899,
      "epoch": 2.9707910014513788,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041253409633292747,
      "loss": 4.9684,
      "mean_token_accuracy": 0.20733024179935455,
      "num_tokens": 67964603.0,
      "step": 32750
    },
    {
      "entropy": 5.40630350112915,
      "epoch": 2.971244557329463,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041250848714055626,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20573512762784957,
      "num_tokens": 67974541.0,
      "step": 32755
    },
    {
      "entropy": 5.391997003555298,
      "epoch": 2.9716981132075473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004124828751044746,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2027280420064926,
      "num_tokens": 67984968.0,
      "step": 32760
    },
    {
      "entropy": 5.324978685379028,
      "epoch": 2.9721516690856316,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004124572602252119,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20620399415493013,
      "num_tokens": 67995567.0,
      "step": 32765
    },
    {
      "entropy": 5.414917230606079,
      "epoch": 2.9726052249637154,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004124316425032981,
      "loss": 5.1444,
      "mean_token_accuracy": 0.19053545892238616,
      "num_tokens": 68008530.0,
      "step": 32770
    },
    {
      "entropy": 5.385799741744995,
      "epoch": 2.9730587808417996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041240602193926294,
      "loss": 5.0817,
      "mean_token_accuracy": 0.20200604796409607,
      "num_tokens": 68018393.0,
      "step": 32775
    },
    {
      "entropy": 5.405347681045532,
      "epoch": 2.973512336719884,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004123803985336362,
      "loss": 5.0173,
      "mean_token_accuracy": 0.19906455278396606,
      "num_tokens": 68027677.0,
      "step": 32780
    },
    {
      "entropy": 5.312229824066162,
      "epoch": 2.973965892597968,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041235477228694786,
      "loss": 4.9027,
      "mean_token_accuracy": 0.21300780773162842,
      "num_tokens": 68037157.0,
      "step": 32785
    },
    {
      "entropy": 5.306584978103638,
      "epoch": 2.974419448476052,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041232914319972775,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20802386552095414,
      "num_tokens": 68047793.0,
      "step": 32790
    },
    {
      "entropy": 5.350949668884278,
      "epoch": 2.9748730043541363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041230351127250603,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2056042343378067,
      "num_tokens": 68058155.0,
      "step": 32795
    },
    {
      "entropy": 5.372919607162475,
      "epoch": 2.9753265602322205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041227787650581266,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2111235588788986,
      "num_tokens": 68067440.0,
      "step": 32800
    },
    {
      "entropy": 5.360297632217407,
      "epoch": 2.975780116110305,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041225223890017785,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2054919943213463,
      "num_tokens": 68077815.0,
      "step": 32805
    },
    {
      "entropy": 5.345359563827515,
      "epoch": 2.976233671988389,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041222659845613173,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2085588052868843,
      "num_tokens": 68087692.0,
      "step": 32810
    },
    {
      "entropy": 5.312851095199585,
      "epoch": 2.9766872278664733,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041220095517420464,
      "loss": 4.9545,
      "mean_token_accuracy": 0.20774792283773422,
      "num_tokens": 68097774.0,
      "step": 32815
    },
    {
      "entropy": 5.330124235153198,
      "epoch": 2.9771407837445576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004121753090549268,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20497495383024217,
      "num_tokens": 68107992.0,
      "step": 32820
    },
    {
      "entropy": 5.3058990955352785,
      "epoch": 2.9775943396226414,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041214966009882857,
      "loss": 4.936,
      "mean_token_accuracy": 0.21283112168312074,
      "num_tokens": 68117103.0,
      "step": 32825
    },
    {
      "entropy": 5.45619764328003,
      "epoch": 2.9780478955007257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004121240083064405,
      "loss": 5.054,
      "mean_token_accuracy": 0.2036134794354439,
      "num_tokens": 68127951.0,
      "step": 32830
    },
    {
      "entropy": 5.400400018692016,
      "epoch": 2.97850145137881,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041209835367829293,
      "loss": 5.0594,
      "mean_token_accuracy": 0.20701372176408767,
      "num_tokens": 68138229.0,
      "step": 32835
    },
    {
      "entropy": 5.214537048339844,
      "epoch": 2.9789550072568938,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041207269621491654,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21013036370277405,
      "num_tokens": 68148019.0,
      "step": 32840
    },
    {
      "entropy": 5.2869316101074215,
      "epoch": 2.979408563134978,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004120470359168418,
      "loss": 5.0088,
      "mean_token_accuracy": 0.19820925146341323,
      "num_tokens": 68158704.0,
      "step": 32845
    },
    {
      "entropy": 5.429519605636597,
      "epoch": 2.9798621190130623,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041202137278459946,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2098425954580307,
      "num_tokens": 68169105.0,
      "step": 32850
    },
    {
      "entropy": 5.378216028213501,
      "epoch": 2.9803156748911466,
      "grad_norm": 1.0,
      "learning_rate": 0.0004119957068187202,
      "loss": 4.9416,
      "mean_token_accuracy": 0.20615691542625428,
      "num_tokens": 68179778.0,
      "step": 32855
    },
    {
      "entropy": 5.318905067443848,
      "epoch": 2.980769230769231,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004119700380197348,
      "loss": 4.9805,
      "mean_token_accuracy": 0.20121067017316818,
      "num_tokens": 68189405.0,
      "step": 32860
    },
    {
      "entropy": 5.2420539379119875,
      "epoch": 2.981222786647315,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041194436638817417,
      "loss": 4.9172,
      "mean_token_accuracy": 0.2153043583035469,
      "num_tokens": 68199351.0,
      "step": 32865
    },
    {
      "entropy": 5.389995288848877,
      "epoch": 2.9816763425253994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041191869192456907,
      "loss": 5.0807,
      "mean_token_accuracy": 0.20200488865375518,
      "num_tokens": 68209904.0,
      "step": 32870
    },
    {
      "entropy": 5.419693183898926,
      "epoch": 2.982129898403483,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004118930146294506,
      "loss": 5.0753,
      "mean_token_accuracy": 0.20023217648267747,
      "num_tokens": 68221754.0,
      "step": 32875
    },
    {
      "entropy": 5.386533594131469,
      "epoch": 2.9825834542815675,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004118673345033496,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20043370574712754,
      "num_tokens": 68232328.0,
      "step": 32880
    },
    {
      "entropy": 5.2903179168701175,
      "epoch": 2.9830370101596517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004118416515467974,
      "loss": 4.9348,
      "mean_token_accuracy": 0.20257487148046494,
      "num_tokens": 68242812.0,
      "step": 32885
    },
    {
      "entropy": 5.412081623077393,
      "epoch": 2.983490566037736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004118159657603248,
      "loss": 5.0742,
      "mean_token_accuracy": 0.20124855935573577,
      "num_tokens": 68253888.0,
      "step": 32890
    },
    {
      "entropy": 5.399179983139038,
      "epoch": 2.98394412191582,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004117902771444633,
      "loss": 5.0106,
      "mean_token_accuracy": 0.2083141729235649,
      "num_tokens": 68264877.0,
      "step": 32895
    },
    {
      "entropy": 5.450057029724121,
      "epoch": 2.984397677793904,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000411764585699744,
      "loss": 5.0942,
      "mean_token_accuracy": 0.19438871443271638,
      "num_tokens": 68275863.0,
      "step": 32900
    },
    {
      "entropy": 5.380949783325195,
      "epoch": 2.9848512336719883,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041173889142669806,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20144784450531006,
      "num_tokens": 68286768.0,
      "step": 32905
    },
    {
      "entropy": 5.317904424667359,
      "epoch": 2.9853047895500726,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004117131943258572,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20985837429761886,
      "num_tokens": 68296990.0,
      "step": 32910
    },
    {
      "entropy": 5.242599773406982,
      "epoch": 2.985758345428157,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041168749439775255,
      "loss": 4.9238,
      "mean_token_accuracy": 0.20848282873630525,
      "num_tokens": 68307646.0,
      "step": 32915
    },
    {
      "entropy": 5.397363185882568,
      "epoch": 2.986211901306241,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004116617916429156,
      "loss": 5.0835,
      "mean_token_accuracy": 0.19075114279985428,
      "num_tokens": 68319586.0,
      "step": 32920
    },
    {
      "entropy": 5.467837810516357,
      "epoch": 2.986665457184325,
      "grad_norm": 1.125,
      "learning_rate": 0.0004116360860618781,
      "loss": 5.0745,
      "mean_token_accuracy": 0.19810774475336074,
      "num_tokens": 68329692.0,
      "step": 32925
    },
    {
      "entropy": 5.437750911712646,
      "epoch": 2.9871190130624092,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004116103776551715,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20448826253414154,
      "num_tokens": 68339589.0,
      "step": 32930
    },
    {
      "entropy": 5.361123704910279,
      "epoch": 2.9875725689404935,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041158466642332743,
      "loss": 5.0573,
      "mean_token_accuracy": 0.19319397062063218,
      "num_tokens": 68349485.0,
      "step": 32935
    },
    {
      "entropy": 5.356835126876831,
      "epoch": 2.9880261248185778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004115589523668777,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2107324942946434,
      "num_tokens": 68359695.0,
      "step": 32940
    },
    {
      "entropy": 5.249476337432862,
      "epoch": 2.9884796806966616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000411533235486354,
      "loss": 4.87,
      "mean_token_accuracy": 0.21146491467952727,
      "num_tokens": 68369591.0,
      "step": 32945
    },
    {
      "entropy": 5.338753700256348,
      "epoch": 2.988933236574746,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004115075157822883,
      "loss": 5.0268,
      "mean_token_accuracy": 0.20901493579149247,
      "num_tokens": 68380382.0,
      "step": 32950
    },
    {
      "entropy": 5.212205934524536,
      "epoch": 2.98938679245283,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004114817932552122,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2099349692463875,
      "num_tokens": 68389865.0,
      "step": 32955
    },
    {
      "entropy": 5.293766784667969,
      "epoch": 2.9898403483309144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041145606790565797,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2158318802714348,
      "num_tokens": 68399267.0,
      "step": 32960
    },
    {
      "entropy": 5.362417507171631,
      "epoch": 2.9902939042089987,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004114303397341575,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21219673603773118,
      "num_tokens": 68409480.0,
      "step": 32965
    },
    {
      "entropy": 5.457457113265991,
      "epoch": 2.990747460087083,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004114046087412427,
      "loss": 5.1138,
      "mean_token_accuracy": 0.20461657345294954,
      "num_tokens": 68420019.0,
      "step": 32970
    },
    {
      "entropy": 5.293764638900757,
      "epoch": 2.9912010159651667,
      "grad_norm": 1.171875,
      "learning_rate": 0.000411378874927446,
      "loss": 4.9524,
      "mean_token_accuracy": 0.20919755846261978,
      "num_tokens": 68429967.0,
      "step": 32975
    },
    {
      "entropy": 5.298889112472534,
      "epoch": 2.991654571843251,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004113531382932993,
      "loss": 4.9597,
      "mean_token_accuracy": 0.20333252251148223,
      "num_tokens": 68441639.0,
      "step": 32980
    },
    {
      "entropy": 5.394142818450928,
      "epoch": 2.9921081277213353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000411327398839335,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20134075582027436,
      "num_tokens": 68452781.0,
      "step": 32985
    },
    {
      "entropy": 5.335870027542114,
      "epoch": 2.9925616835994195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004113016565660854,
      "loss": 4.9401,
      "mean_token_accuracy": 0.20975549519062042,
      "num_tokens": 68463353.0,
      "step": 32990
    },
    {
      "entropy": 5.286372947692871,
      "epoch": 2.9930152394775034,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041127591147408274,
      "loss": 4.9643,
      "mean_token_accuracy": 0.20140987187623977,
      "num_tokens": 68473507.0,
      "step": 32995
    },
    {
      "entropy": 5.353664970397949,
      "epoch": 2.9934687953555876,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041125016356385945,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20939079523086548,
      "num_tokens": 68483737.0,
      "step": 33000
    },
    {
      "epoch": 2.9934687953555876,
      "eval_entropy": 5.1660794934536,
      "eval_loss": 5.062716484069824,
      "eval_mean_token_accuracy": 0.21104055240632316,
      "eval_num_tokens": 68483737.0,
      "eval_runtime": 26.6199,
      "eval_samples_per_second": 1328.331,
      "eval_steps_per_second": 166.041,
      "step": 33000
    },
    {
      "entropy": 5.385732698440552,
      "epoch": 2.993922351233672,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041122441283594817,
      "loss": 5.0007,
      "mean_token_accuracy": 0.20872917473316194,
      "num_tokens": 68494013.0,
      "step": 33005
    },
    {
      "entropy": 5.307618808746338,
      "epoch": 2.994375907111756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004111986592908813,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21309976428747177,
      "num_tokens": 68504361.0,
      "step": 33010
    },
    {
      "entropy": 5.294649648666382,
      "epoch": 2.9948294629898404,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041117290292919134,
      "loss": 4.9264,
      "mean_token_accuracy": 0.20708640813827514,
      "num_tokens": 68514825.0,
      "step": 33015
    },
    {
      "entropy": 5.323504829406739,
      "epoch": 2.9952830188679247,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004111471437514111,
      "loss": 4.9391,
      "mean_token_accuracy": 0.2031451314687729,
      "num_tokens": 68524858.0,
      "step": 33020
    },
    {
      "entropy": 5.2922710418701175,
      "epoch": 2.995736574746009,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004111213817580733,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21943661272525788,
      "num_tokens": 68534481.0,
      "step": 33025
    },
    {
      "entropy": 5.277309560775757,
      "epoch": 2.996190130624093,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004110956169497106,
      "loss": 4.9861,
      "mean_token_accuracy": 0.2062538281083107,
      "num_tokens": 68545436.0,
      "step": 33030
    },
    {
      "entropy": 5.32977180480957,
      "epoch": 2.996643686502177,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004110698493268558,
      "loss": 4.985,
      "mean_token_accuracy": 0.2086052790284157,
      "num_tokens": 68555434.0,
      "step": 33035
    },
    {
      "entropy": 5.365571117401123,
      "epoch": 2.9970972423802613,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004110440788900419,
      "loss": 5.0979,
      "mean_token_accuracy": 0.20006747543811798,
      "num_tokens": 68566295.0,
      "step": 33040
    },
    {
      "entropy": 5.290982580184936,
      "epoch": 2.9975507982583456,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041101830563980174,
      "loss": 4.9079,
      "mean_token_accuracy": 0.21765578389167786,
      "num_tokens": 68575927.0,
      "step": 33045
    },
    {
      "entropy": 5.400971937179565,
      "epoch": 2.9980043541364294,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041099252957666843,
      "loss": 5.008,
      "mean_token_accuracy": 0.2076717346906662,
      "num_tokens": 68586254.0,
      "step": 33050
    },
    {
      "entropy": 5.324668931961059,
      "epoch": 2.9984579100145137,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041096675070117496,
      "loss": 5.01,
      "mean_token_accuracy": 0.20441973060369492,
      "num_tokens": 68596881.0,
      "step": 33055
    },
    {
      "entropy": 5.240698289871216,
      "epoch": 2.998911465892598,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041094096901385434,
      "loss": 4.883,
      "mean_token_accuracy": 0.2121098592877388,
      "num_tokens": 68606875.0,
      "step": 33060
    },
    {
      "entropy": 5.366143608093262,
      "epoch": 2.999365021770682,
      "grad_norm": 1.125,
      "learning_rate": 0.0004109151845152399,
      "loss": 5.0206,
      "mean_token_accuracy": 0.20344352275133132,
      "num_tokens": 68617011.0,
      "step": 33065
    },
    {
      "entropy": 5.402824926376343,
      "epoch": 2.9998185776487665,
      "grad_norm": 1.125,
      "learning_rate": 0.0004108893972058649,
      "loss": 5.0402,
      "mean_token_accuracy": 0.19922117441892623,
      "num_tokens": 68627764.0,
      "step": 33070
    },
    {
      "entropy": 5.366007137298584,
      "epoch": 3.0002721335268503,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041086360708626236,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2132004350423813,
      "num_tokens": 68637272.0,
      "step": 33075
    },
    {
      "entropy": 5.268539619445801,
      "epoch": 3.0007256894049346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041083781415696585,
      "loss": 4.92,
      "mean_token_accuracy": 0.20663753747940064,
      "num_tokens": 68648058.0,
      "step": 33080
    },
    {
      "entropy": 5.275871324539184,
      "epoch": 3.001179245283019,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004108120184185088,
      "loss": 4.8494,
      "mean_token_accuracy": 0.21280302703380585,
      "num_tokens": 68658174.0,
      "step": 33085
    },
    {
      "entropy": 5.347722101211548,
      "epoch": 3.001632801161103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041078621987142457,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2174378365278244,
      "num_tokens": 68669281.0,
      "step": 33090
    },
    {
      "entropy": 5.292038345336914,
      "epoch": 3.0020863570391874,
      "grad_norm": 1.125,
      "learning_rate": 0.0004107604185162467,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2097453787922859,
      "num_tokens": 68679879.0,
      "step": 33095
    },
    {
      "entropy": 5.26264762878418,
      "epoch": 3.0025399129172716,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041073461435350877,
      "loss": 4.8288,
      "mean_token_accuracy": 0.21347414702177048,
      "num_tokens": 68689852.0,
      "step": 33100
    },
    {
      "entropy": 5.250535917282105,
      "epoch": 3.0029934687953554,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041070880738374447,
      "loss": 4.9128,
      "mean_token_accuracy": 0.20845771282911302,
      "num_tokens": 68700149.0,
      "step": 33105
    },
    {
      "entropy": 5.321685457229615,
      "epoch": 3.0034470246734397,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041068299760748734,
      "loss": 4.916,
      "mean_token_accuracy": 0.2112745985388756,
      "num_tokens": 68710837.0,
      "step": 33110
    },
    {
      "entropy": 5.3516845703125,
      "epoch": 3.003900580551524,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041065718502527134,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2089512214064598,
      "num_tokens": 68721451.0,
      "step": 33115
    },
    {
      "entropy": 5.27822642326355,
      "epoch": 3.0043541364296082,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041063136963763013,
      "loss": 4.8897,
      "mean_token_accuracy": 0.2127857968211174,
      "num_tokens": 68732911.0,
      "step": 33120
    },
    {
      "entropy": 5.247108268737793,
      "epoch": 3.0048076923076925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004106055514450977,
      "loss": 4.8643,
      "mean_token_accuracy": 0.2140848457813263,
      "num_tokens": 68742650.0,
      "step": 33125
    },
    {
      "entropy": 5.259944105148316,
      "epoch": 3.0052612481857763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004105797304482079,
      "loss": 4.8684,
      "mean_token_accuracy": 0.214817014336586,
      "num_tokens": 68753660.0,
      "step": 33130
    },
    {
      "entropy": 5.280070066452026,
      "epoch": 3.0057148040638606,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004105539066474947,
      "loss": 4.8005,
      "mean_token_accuracy": 0.21791994869709014,
      "num_tokens": 68763868.0,
      "step": 33135
    },
    {
      "entropy": 5.295200634002685,
      "epoch": 3.006168359941945,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041052808004349214,
      "loss": 4.9115,
      "mean_token_accuracy": 0.20896798968315125,
      "num_tokens": 68773911.0,
      "step": 33140
    },
    {
      "entropy": 5.336948490142822,
      "epoch": 3.006621915820029,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004105022506367344,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20191432535648346,
      "num_tokens": 68786423.0,
      "step": 33145
    },
    {
      "entropy": 5.348840761184692,
      "epoch": 3.0070754716981134,
      "grad_norm": 1.125,
      "learning_rate": 0.0004104764184277555,
      "loss": 4.9721,
      "mean_token_accuracy": 0.20155239403247832,
      "num_tokens": 68797329.0,
      "step": 33150
    },
    {
      "entropy": 5.4015655517578125,
      "epoch": 3.0075290275761972,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004104505834170898,
      "loss": 4.9974,
      "mean_token_accuracy": 0.2051040142774582,
      "num_tokens": 68808447.0,
      "step": 33155
    },
    {
      "entropy": 5.365656995773316,
      "epoch": 3.0079825834542815,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041042474560527153,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21102637499570848,
      "num_tokens": 68819381.0,
      "step": 33160
    },
    {
      "entropy": 5.26922492980957,
      "epoch": 3.0084361393323658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000410398904992835,
      "loss": 4.8711,
      "mean_token_accuracy": 0.2146022766828537,
      "num_tokens": 68829319.0,
      "step": 33165
    },
    {
      "entropy": 5.288395166397095,
      "epoch": 3.00888969521045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041037306158031464,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2122725561261177,
      "num_tokens": 68838640.0,
      "step": 33170
    },
    {
      "entropy": 5.385033655166626,
      "epoch": 3.0093432510885343,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004103472153682448,
      "loss": 4.9092,
      "mean_token_accuracy": 0.20761121958494186,
      "num_tokens": 68847960.0,
      "step": 33175
    },
    {
      "entropy": 5.303008413314819,
      "epoch": 3.009796806966618,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041032136635716016,
      "loss": 4.92,
      "mean_token_accuracy": 0.20787417590618135,
      "num_tokens": 68858851.0,
      "step": 33180
    },
    {
      "entropy": 5.253830289840698,
      "epoch": 3.0102503628447024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004102955145475951,
      "loss": 4.8898,
      "mean_token_accuracy": 0.2056427463889122,
      "num_tokens": 68869006.0,
      "step": 33185
    },
    {
      "entropy": 5.440251398086548,
      "epoch": 3.0107039187227866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041026965994008435,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20459049195051193,
      "num_tokens": 68879606.0,
      "step": 33190
    },
    {
      "entropy": 5.298396253585816,
      "epoch": 3.011157474600871,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041024380253516253,
      "loss": 4.8663,
      "mean_token_accuracy": 0.21971344500780104,
      "num_tokens": 68889862.0,
      "step": 33195
    },
    {
      "entropy": 5.398309564590454,
      "epoch": 3.011611030478955,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041021794233336444,
      "loss": 4.9745,
      "mean_token_accuracy": 0.2117897853255272,
      "num_tokens": 68900512.0,
      "step": 33200
    },
    {
      "entropy": 5.32092809677124,
      "epoch": 3.012064586357039,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004101920793352249,
      "loss": 4.9331,
      "mean_token_accuracy": 0.20672511011362077,
      "num_tokens": 68910475.0,
      "step": 33205
    },
    {
      "entropy": 5.33672399520874,
      "epoch": 3.0125181422351233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004101662135412786,
      "loss": 4.95,
      "mean_token_accuracy": 0.21183894574642181,
      "num_tokens": 68920583.0,
      "step": 33210
    },
    {
      "entropy": 5.231509351730347,
      "epoch": 3.0129716981132075,
      "grad_norm": 1.125,
      "learning_rate": 0.0004101403449520606,
      "loss": 4.8551,
      "mean_token_accuracy": 0.20743171870708466,
      "num_tokens": 68931016.0,
      "step": 33215
    },
    {
      "entropy": 5.3316926002502445,
      "epoch": 3.013425253991292,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041011447356810584,
      "loss": 4.888,
      "mean_token_accuracy": 0.21115488409996033,
      "num_tokens": 68941489.0,
      "step": 33220
    },
    {
      "entropy": 5.391120338439942,
      "epoch": 3.013878809869376,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004100885993899493,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2027078926563263,
      "num_tokens": 68952322.0,
      "step": 33225
    },
    {
      "entropy": 5.312001705169678,
      "epoch": 3.01433236574746,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004100627224181261,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20956186056137086,
      "num_tokens": 68962570.0,
      "step": 33230
    },
    {
      "entropy": 5.330828905105591,
      "epoch": 3.014785921625544,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041003684265317143,
      "loss": 4.8935,
      "mean_token_accuracy": 0.20868588835000992,
      "num_tokens": 68973733.0,
      "step": 33235
    },
    {
      "entropy": 5.338713502883911,
      "epoch": 3.0152394775036284,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041001096009562045,
      "loss": 4.9217,
      "mean_token_accuracy": 0.21358666121959685,
      "num_tokens": 68983526.0,
      "step": 33240
    },
    {
      "entropy": 5.377814483642578,
      "epoch": 3.0156930333817127,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004099850747460084,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2103472888469696,
      "num_tokens": 68993968.0,
      "step": 33245
    },
    {
      "entropy": 5.303847599029541,
      "epoch": 3.016146589259797,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040995918660487055,
      "loss": 4.8916,
      "mean_token_accuracy": 0.21986321955919266,
      "num_tokens": 69003599.0,
      "step": 33250
    },
    {
      "entropy": 5.295163345336914,
      "epoch": 3.0166001451378808,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004099332956727424,
      "loss": 4.897,
      "mean_token_accuracy": 0.2128329485654831,
      "num_tokens": 69013686.0,
      "step": 33255
    },
    {
      "entropy": 5.306044483184815,
      "epoch": 3.017053701015965,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004099074019501593,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2079255148768425,
      "num_tokens": 69023803.0,
      "step": 33260
    },
    {
      "entropy": 5.273927879333496,
      "epoch": 3.0175072568940493,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004098815054376567,
      "loss": 4.8201,
      "mean_token_accuracy": 0.2145162582397461,
      "num_tokens": 69033885.0,
      "step": 33265
    },
    {
      "entropy": 5.31258864402771,
      "epoch": 3.0179608127721336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004098556061357702,
      "loss": 4.885,
      "mean_token_accuracy": 0.21365167945623398,
      "num_tokens": 69044471.0,
      "step": 33270
    },
    {
      "entropy": 5.268884611129761,
      "epoch": 3.018414368650218,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040982970404503544,
      "loss": 4.7944,
      "mean_token_accuracy": 0.2163750261068344,
      "num_tokens": 69055109.0,
      "step": 33275
    },
    {
      "entropy": 5.26045413017273,
      "epoch": 3.018867924528302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040980379916598803,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21301300972700118,
      "num_tokens": 69064692.0,
      "step": 33280
    },
    {
      "entropy": 5.315602779388428,
      "epoch": 3.019321480406386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004097778914991637,
      "loss": 4.9132,
      "mean_token_accuracy": 0.20833991169929506,
      "num_tokens": 69075331.0,
      "step": 33285
    },
    {
      "entropy": 5.320168924331665,
      "epoch": 3.01977503628447,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040975198104509827,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21403435319662095,
      "num_tokens": 69085017.0,
      "step": 33290
    },
    {
      "entropy": 5.254235410690308,
      "epoch": 3.0202285921625545,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004097260678043275,
      "loss": 4.8969,
      "mean_token_accuracy": 0.20588230788707734,
      "num_tokens": 69096128.0,
      "step": 33295
    },
    {
      "entropy": 5.358696365356446,
      "epoch": 3.0206821480406387,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040970015177738725,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2052266076207161,
      "num_tokens": 69106479.0,
      "step": 33300
    },
    {
      "entropy": 5.425902318954468,
      "epoch": 3.021135703918723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004096742329648136,
      "loss": 4.9404,
      "mean_token_accuracy": 0.20876569598913192,
      "num_tokens": 69116346.0,
      "step": 33305
    },
    {
      "entropy": 5.3005493640899655,
      "epoch": 3.021589259796807,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004096483113671425,
      "loss": 4.9672,
      "mean_token_accuracy": 0.20265064835548402,
      "num_tokens": 69126489.0,
      "step": 33310
    },
    {
      "entropy": 5.329137372970581,
      "epoch": 3.022042815674891,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040962238698490994,
      "loss": 4.8957,
      "mean_token_accuracy": 0.20997516959905624,
      "num_tokens": 69136618.0,
      "step": 33315
    },
    {
      "entropy": 5.397400808334351,
      "epoch": 3.0224963715529753,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004095964598186522,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2080133005976677,
      "num_tokens": 69146242.0,
      "step": 33320
    },
    {
      "entropy": 5.323878431320191,
      "epoch": 3.0229499274310596,
      "grad_norm": 1.0,
      "learning_rate": 0.0004095705298689053,
      "loss": 4.8982,
      "mean_token_accuracy": 0.2126981571316719,
      "num_tokens": 69157269.0,
      "step": 33325
    },
    {
      "entropy": 5.424605846405029,
      "epoch": 3.023403483309144,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040954459713620563,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20688370913267135,
      "num_tokens": 69169260.0,
      "step": 33330
    },
    {
      "entropy": 5.429667901992798,
      "epoch": 3.0238570391872277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004095186616210893,
      "loss": 5.04,
      "mean_token_accuracy": 0.19681639820337296,
      "num_tokens": 69179815.0,
      "step": 33335
    },
    {
      "entropy": 5.3755255222320555,
      "epoch": 3.024310595065312,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004094927233240928,
      "loss": 5.008,
      "mean_token_accuracy": 0.20745415091514588,
      "num_tokens": 69189959.0,
      "step": 33340
    },
    {
      "entropy": 5.36087818145752,
      "epoch": 3.0247641509433962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040946678224575264,
      "loss": 4.8952,
      "mean_token_accuracy": 0.2112766370177269,
      "num_tokens": 69199485.0,
      "step": 33345
    },
    {
      "entropy": 5.24544734954834,
      "epoch": 3.0252177068214805,
      "grad_norm": 1.125,
      "learning_rate": 0.000409440838386605,
      "loss": 4.9146,
      "mean_token_accuracy": 0.21304986923933028,
      "num_tokens": 69209180.0,
      "step": 33350
    },
    {
      "entropy": 5.317651081085205,
      "epoch": 3.0256712626995648,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004094148917471866,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2077927365899086,
      "num_tokens": 69220648.0,
      "step": 33355
    },
    {
      "entropy": 5.298368358612061,
      "epoch": 3.0261248185776486,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000409388942328034,
      "loss": 4.7978,
      "mean_token_accuracy": 0.2137380689382553,
      "num_tokens": 69230685.0,
      "step": 33360
    },
    {
      "entropy": 5.287924575805664,
      "epoch": 3.026578374455733,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004093629901296838,
      "loss": 4.9509,
      "mean_token_accuracy": 0.20590994209051133,
      "num_tokens": 69240407.0,
      "step": 33365
    },
    {
      "entropy": 5.439332056045532,
      "epoch": 3.027031930333817,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004093370351526727,
      "loss": 4.9894,
      "mean_token_accuracy": 0.2033974125981331,
      "num_tokens": 69250624.0,
      "step": 33370
    },
    {
      "entropy": 5.45145902633667,
      "epoch": 3.0274854862119014,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004093110773975376,
      "loss": 5.0115,
      "mean_token_accuracy": 0.19849569201469422,
      "num_tokens": 69261131.0,
      "step": 33375
    },
    {
      "entropy": 5.299784564971924,
      "epoch": 3.0279390420899857,
      "grad_norm": 1.125,
      "learning_rate": 0.0004092851168648152,
      "loss": 4.9497,
      "mean_token_accuracy": 0.20209716707468034,
      "num_tokens": 69270668.0,
      "step": 33380
    },
    {
      "entropy": 5.308194923400879,
      "epoch": 3.0283925979680695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004092591535550422,
      "loss": 4.8555,
      "mean_token_accuracy": 0.21897859275341033,
      "num_tokens": 69280528.0,
      "step": 33385
    },
    {
      "entropy": 5.283936405181885,
      "epoch": 3.0288461538461537,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004092331874687558,
      "loss": 4.8973,
      "mean_token_accuracy": 0.21412773728370665,
      "num_tokens": 69290376.0,
      "step": 33390
    },
    {
      "entropy": 5.3415940284729,
      "epoch": 3.029299709724238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040920721860649285,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21649360060691833,
      "num_tokens": 69301766.0,
      "step": 33395
    },
    {
      "entropy": 5.365674638748169,
      "epoch": 3.0297532656023223,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004091812469687904,
      "loss": 4.8845,
      "mean_token_accuracy": 0.2059804305434227,
      "num_tokens": 69312668.0,
      "step": 33400
    },
    {
      "entropy": 5.316924953460694,
      "epoch": 3.0302068214804065,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004091552725561856,
      "loss": 4.953,
      "mean_token_accuracy": 0.20256489664316177,
      "num_tokens": 69323195.0,
      "step": 33405
    },
    {
      "entropy": 5.278154087066651,
      "epoch": 3.0306603773584904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004091292953692156,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21026352494955064,
      "num_tokens": 69333812.0,
      "step": 33410
    },
    {
      "entropy": 5.33851113319397,
      "epoch": 3.0311139332365746,
      "grad_norm": 1.0,
      "learning_rate": 0.00040910331540841755,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2153009667992592,
      "num_tokens": 69344378.0,
      "step": 33415
    },
    {
      "entropy": 5.276911115646362,
      "epoch": 3.031567489114659,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040907733267432877,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21275716722011567,
      "num_tokens": 69354592.0,
      "step": 33420
    },
    {
      "entropy": 5.353150320053101,
      "epoch": 3.032021044992743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040905134716748656,
      "loss": 4.9167,
      "mean_token_accuracy": 0.20919893831014633,
      "num_tokens": 69365011.0,
      "step": 33425
    },
    {
      "entropy": 5.296702909469604,
      "epoch": 3.0324746008708274,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004090253588884284,
      "loss": 4.9031,
      "mean_token_accuracy": 0.2065386265516281,
      "num_tokens": 69377010.0,
      "step": 33430
    },
    {
      "entropy": 5.348358058929444,
      "epoch": 3.0329281567489113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040899936783769157,
      "loss": 4.9371,
      "mean_token_accuracy": 0.2112121820449829,
      "num_tokens": 69387263.0,
      "step": 33435
    },
    {
      "entropy": 5.345802068710327,
      "epoch": 3.0333817126269955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040897337401581374,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21098800450563432,
      "num_tokens": 69399188.0,
      "step": 33440
    },
    {
      "entropy": 5.30116457939148,
      "epoch": 3.03383526850508,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040894737742333244,
      "loss": 4.9232,
      "mean_token_accuracy": 0.20289838761091233,
      "num_tokens": 69408271.0,
      "step": 33445
    },
    {
      "entropy": 5.385930013656616,
      "epoch": 3.034288824383164,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004089213780607851,
      "loss": 5.0297,
      "mean_token_accuracy": 0.1971869021654129,
      "num_tokens": 69418424.0,
      "step": 33450
    },
    {
      "entropy": 5.3674641132354735,
      "epoch": 3.0347423802612483,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004088953759287096,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21111880242824554,
      "num_tokens": 69427802.0,
      "step": 33455
    },
    {
      "entropy": 5.317855501174927,
      "epoch": 3.0351959361393326,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004088693710276436,
      "loss": 4.9524,
      "mean_token_accuracy": 0.21167428195476531,
      "num_tokens": 69437581.0,
      "step": 33460
    },
    {
      "entropy": 5.354666423797608,
      "epoch": 3.0356494920174164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040884336335812494,
      "loss": 4.881,
      "mean_token_accuracy": 0.21361382752656938,
      "num_tokens": 69447711.0,
      "step": 33465
    },
    {
      "entropy": 5.365551519393921,
      "epoch": 3.0361030478955007,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004088173529206914,
      "loss": 4.8586,
      "mean_token_accuracy": 0.21098092347383499,
      "num_tokens": 69457692.0,
      "step": 33470
    },
    {
      "entropy": 5.292859649658203,
      "epoch": 3.036556603773585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004087913397158809,
      "loss": 4.8543,
      "mean_token_accuracy": 0.21131246536970139,
      "num_tokens": 69468034.0,
      "step": 33475
    },
    {
      "entropy": 5.267245578765869,
      "epoch": 3.037010159651669,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040876532374423135,
      "loss": 4.8941,
      "mean_token_accuracy": 0.21294918060302734,
      "num_tokens": 69477719.0,
      "step": 33480
    },
    {
      "entropy": 5.317737102508545,
      "epoch": 3.0374637155297535,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004087393050062809,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2102628618478775,
      "num_tokens": 69487261.0,
      "step": 33485
    },
    {
      "entropy": 5.3674804210662845,
      "epoch": 3.0379172714078373,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040871328350256745,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2111978366971016,
      "num_tokens": 69496658.0,
      "step": 33490
    },
    {
      "entropy": 5.400088214874268,
      "epoch": 3.0383708272859216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040868725923362927,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19932497143745423,
      "num_tokens": 69507197.0,
      "step": 33495
    },
    {
      "entropy": 5.317652606964112,
      "epoch": 3.038824383164006,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040866123220000455,
      "loss": 4.9654,
      "mean_token_accuracy": 0.216985259950161,
      "num_tokens": 69516905.0,
      "step": 33500
    },
    {
      "entropy": 5.390602397918701,
      "epoch": 3.03927793904209,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040863520240223144,
      "loss": 4.8766,
      "mean_token_accuracy": 0.21431356370449067,
      "num_tokens": 69526869.0,
      "step": 33505
    },
    {
      "entropy": 5.35844087600708,
      "epoch": 3.0397314949201744,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004086091698408483,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2053476244211197,
      "num_tokens": 69537058.0,
      "step": 33510
    },
    {
      "entropy": 5.367183828353882,
      "epoch": 3.040185050798258,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040858313451639343,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20035415291786193,
      "num_tokens": 69547617.0,
      "step": 33515
    },
    {
      "entropy": 5.35741400718689,
      "epoch": 3.0406386066763424,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004085570964294053,
      "loss": 4.9757,
      "mean_token_accuracy": 0.20049203783273697,
      "num_tokens": 69558733.0,
      "step": 33520
    },
    {
      "entropy": 5.425289583206177,
      "epoch": 3.0410921625544267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004085310555804225,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20332031995058059,
      "num_tokens": 69570011.0,
      "step": 33525
    },
    {
      "entropy": 5.396490526199341,
      "epoch": 3.041545718432511,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040850501196998336,
      "loss": 5.0398,
      "mean_token_accuracy": 0.19671372324228287,
      "num_tokens": 69580894.0,
      "step": 33530
    },
    {
      "entropy": 5.329995012283325,
      "epoch": 3.0419992743105952,
      "grad_norm": 1.0,
      "learning_rate": 0.00040847896559862654,
      "loss": 4.9447,
      "mean_token_accuracy": 0.20690785348415375,
      "num_tokens": 69591438.0,
      "step": 33535
    },
    {
      "entropy": 5.336126089096069,
      "epoch": 3.042452830188679,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004084529164668907,
      "loss": 4.9624,
      "mean_token_accuracy": 0.20556763261556626,
      "num_tokens": 69601745.0,
      "step": 33540
    },
    {
      "entropy": 5.348664045333862,
      "epoch": 3.0429063860667633,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040842686457531453,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21058484315872192,
      "num_tokens": 69611872.0,
      "step": 33545
    },
    {
      "entropy": 5.365072345733642,
      "epoch": 3.0433599419448476,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004084008099244368,
      "loss": 4.9008,
      "mean_token_accuracy": 0.20909989774227142,
      "num_tokens": 69621673.0,
      "step": 33550
    },
    {
      "entropy": 5.27770848274231,
      "epoch": 3.043813497822932,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040837475251479633,
      "loss": 4.8611,
      "mean_token_accuracy": 0.2129132091999054,
      "num_tokens": 69632090.0,
      "step": 33555
    },
    {
      "entropy": 5.282564973831176,
      "epoch": 3.044267053701016,
      "grad_norm": 1.015625,
      "learning_rate": 0.000408348692346932,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2136421397328377,
      "num_tokens": 69642920.0,
      "step": 33560
    },
    {
      "entropy": 5.357941389083862,
      "epoch": 3.0447206095791,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040832262942138277,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20795419663190842,
      "num_tokens": 69653936.0,
      "step": 33565
    },
    {
      "entropy": 5.278037595748901,
      "epoch": 3.0451741654571842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004082965637386875,
      "loss": 4.8617,
      "mean_token_accuracy": 0.21192155480384828,
      "num_tokens": 69663362.0,
      "step": 33570
    },
    {
      "entropy": 5.2803748607635494,
      "epoch": 3.0456277213352685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004082704952993854,
      "loss": 4.8814,
      "mean_token_accuracy": 0.2160523682832718,
      "num_tokens": 69673178.0,
      "step": 33575
    },
    {
      "entropy": 5.40604567527771,
      "epoch": 3.0460812772133528,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004082444241040155,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20479342937469483,
      "num_tokens": 69683627.0,
      "step": 33580
    },
    {
      "entropy": 5.347581005096435,
      "epoch": 3.046534833091437,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004082183501531169,
      "loss": 4.943,
      "mean_token_accuracy": 0.2089016169309616,
      "num_tokens": 69694010.0,
      "step": 33585
    },
    {
      "entropy": 5.307797145843506,
      "epoch": 3.046988388969521,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040819227344722893,
      "loss": 4.9501,
      "mean_token_accuracy": 0.20085043907165528,
      "num_tokens": 69704888.0,
      "step": 33590
    },
    {
      "entropy": 5.2927985191345215,
      "epoch": 3.047441944847605,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040816619398689074,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21326832920312883,
      "num_tokens": 69714718.0,
      "step": 33595
    },
    {
      "entropy": 5.29482479095459,
      "epoch": 3.0478955007256894,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040814011177264174,
      "loss": 4.9023,
      "mean_token_accuracy": 0.21220126897096633,
      "num_tokens": 69726076.0,
      "step": 33600
    },
    {
      "entropy": 5.301858377456665,
      "epoch": 3.0483490566037736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040811402680502137,
      "loss": 4.8676,
      "mean_token_accuracy": 0.21320598423480988,
      "num_tokens": 69736507.0,
      "step": 33605
    },
    {
      "entropy": 5.2634984970092775,
      "epoch": 3.048802612481858,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040808793908456885,
      "loss": 4.8814,
      "mean_token_accuracy": 0.21204362511634828,
      "num_tokens": 69746061.0,
      "step": 33610
    },
    {
      "entropy": 5.315767240524292,
      "epoch": 3.0492561683599417,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004080618486118241,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2058211460709572,
      "num_tokens": 69756493.0,
      "step": 33615
    },
    {
      "entropy": 5.3454673290252686,
      "epoch": 3.049709724238026,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004080357553873261,
      "loss": 4.8773,
      "mean_token_accuracy": 0.21234503984451295,
      "num_tokens": 69766937.0,
      "step": 33620
    },
    {
      "entropy": 5.294748115539551,
      "epoch": 3.0501632801161103,
      "grad_norm": 1.0625,
      "learning_rate": 0.000408009659411615,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21593217551708221,
      "num_tokens": 69776872.0,
      "step": 33625
    },
    {
      "entropy": 5.331532335281372,
      "epoch": 3.0506168359941945,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004079835606852301,
      "loss": 4.913,
      "mean_token_accuracy": 0.20819045156240462,
      "num_tokens": 69788145.0,
      "step": 33630
    },
    {
      "entropy": 5.328314113616943,
      "epoch": 3.051070391872279,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004079574592087114,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21159184128046035,
      "num_tokens": 69797353.0,
      "step": 33635
    },
    {
      "entropy": 5.292540407180786,
      "epoch": 3.051523947750363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040793135498259854,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21647855639457703,
      "num_tokens": 69806915.0,
      "step": 33640
    },
    {
      "entropy": 5.29088978767395,
      "epoch": 3.051977503628447,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040790524800743134,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21175978183746338,
      "num_tokens": 69817695.0,
      "step": 33645
    },
    {
      "entropy": 5.256285715103149,
      "epoch": 3.052431059506531,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004078791382837498,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22576122879981994,
      "num_tokens": 69827451.0,
      "step": 33650
    },
    {
      "entropy": 5.335959768295288,
      "epoch": 3.0528846153846154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040785302581209374,
      "loss": 4.9365,
      "mean_token_accuracy": 0.21460291296243666,
      "num_tokens": 69837897.0,
      "step": 33655
    },
    {
      "entropy": 5.331196355819702,
      "epoch": 3.0533381712626997,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040782691059300326,
      "loss": 4.8994,
      "mean_token_accuracy": 0.21169035285711288,
      "num_tokens": 69847378.0,
      "step": 33660
    },
    {
      "entropy": 5.336082887649536,
      "epoch": 3.053791727140784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004078007926270185,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21011639386415482,
      "num_tokens": 69857522.0,
      "step": 33665
    },
    {
      "entropy": 5.326633071899414,
      "epoch": 3.0542452830188678,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040777467191467946,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2122240424156189,
      "num_tokens": 69868384.0,
      "step": 33670
    },
    {
      "entropy": 5.333763217926025,
      "epoch": 3.054698838896952,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004077485484565264,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21221586763858796,
      "num_tokens": 69878761.0,
      "step": 33675
    },
    {
      "entropy": 5.3623510837554935,
      "epoch": 3.0551523947750363,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004077224222530994,
      "loss": 4.895,
      "mean_token_accuracy": 0.2087968409061432,
      "num_tokens": 69888366.0,
      "step": 33680
    },
    {
      "entropy": 5.304198169708252,
      "epoch": 3.0556059506531206,
      "grad_norm": 0.984375,
      "learning_rate": 0.000407696293304939,
      "loss": 4.921,
      "mean_token_accuracy": 0.20184600204229355,
      "num_tokens": 69898489.0,
      "step": 33685
    },
    {
      "entropy": 5.3021729469299315,
      "epoch": 3.056059506531205,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004076701616125854,
      "loss": 4.8942,
      "mean_token_accuracy": 0.20803744792938234,
      "num_tokens": 69908425.0,
      "step": 33690
    },
    {
      "entropy": 5.3524984359741214,
      "epoch": 3.0565130624092887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004076440271765791,
      "loss": 4.9537,
      "mean_token_accuracy": 0.20990763157606124,
      "num_tokens": 69918488.0,
      "step": 33695
    },
    {
      "entropy": 5.253677654266357,
      "epoch": 3.056966618287373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004076178899974605,
      "loss": 4.8222,
      "mean_token_accuracy": 0.2189752236008644,
      "num_tokens": 69928831.0,
      "step": 33700
    },
    {
      "entropy": 5.375267219543457,
      "epoch": 3.057420174165457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040759175007577017,
      "loss": 4.9335,
      "mean_token_accuracy": 0.20594750940799714,
      "num_tokens": 69938974.0,
      "step": 33705
    },
    {
      "entropy": 5.309829807281494,
      "epoch": 3.0578737300435415,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004075656074120486,
      "loss": 4.8486,
      "mean_token_accuracy": 0.21355448067188262,
      "num_tokens": 69948804.0,
      "step": 33710
    },
    {
      "entropy": 5.310379552841186,
      "epoch": 3.0583272859216257,
      "grad_norm": 1.0,
      "learning_rate": 0.0004075394620068365,
      "loss": 4.9155,
      "mean_token_accuracy": 0.20916415452957154,
      "num_tokens": 69958945.0,
      "step": 33715
    },
    {
      "entropy": 5.282806396484375,
      "epoch": 3.0587808417997095,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040751331386067456,
      "loss": 4.8658,
      "mean_token_accuracy": 0.21313996315002443,
      "num_tokens": 69968826.0,
      "step": 33720
    },
    {
      "entropy": 5.348487186431885,
      "epoch": 3.059234397677794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004074871629741035,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21026697456836702,
      "num_tokens": 69978961.0,
      "step": 33725
    },
    {
      "entropy": 5.251104164123535,
      "epoch": 3.059687953555878,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004074610093476641,
      "loss": 4.9533,
      "mean_token_accuracy": 0.2082328662276268,
      "num_tokens": 69989848.0,
      "step": 33730
    },
    {
      "entropy": 5.300694799423217,
      "epoch": 3.0601415094339623,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040743485298189744,
      "loss": 4.8963,
      "mean_token_accuracy": 0.20889753550291063,
      "num_tokens": 69999607.0,
      "step": 33735
    },
    {
      "entropy": 5.362535285949707,
      "epoch": 3.0605950653120466,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040740869387734415,
      "loss": 4.9346,
      "mean_token_accuracy": 0.20807442665100098,
      "num_tokens": 70010446.0,
      "step": 33740
    },
    {
      "entropy": 5.269730091094971,
      "epoch": 3.0610486211901304,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004073825320345453,
      "loss": 4.8477,
      "mean_token_accuracy": 0.20850915908813478,
      "num_tokens": 70021148.0,
      "step": 33745
    },
    {
      "entropy": 5.35215015411377,
      "epoch": 3.0615021770682147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040735636745404196,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2020315110683441,
      "num_tokens": 70030765.0,
      "step": 33750
    },
    {
      "entropy": 5.3964190006256105,
      "epoch": 3.061955732946299,
      "grad_norm": 1.125,
      "learning_rate": 0.00040733020013637526,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2097747728228569,
      "num_tokens": 70041717.0,
      "step": 33755
    },
    {
      "entropy": 5.379013967514038,
      "epoch": 3.0624092888243832,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040730403008208633,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20471912622451782,
      "num_tokens": 70052178.0,
      "step": 33760
    },
    {
      "entropy": 5.329321098327637,
      "epoch": 3.0628628447024675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004072778572917163,
      "loss": 4.9439,
      "mean_token_accuracy": 0.20557831972837448,
      "num_tokens": 70062819.0,
      "step": 33765
    },
    {
      "entropy": 5.260271835327148,
      "epoch": 3.0633164005805513,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004072516817658065,
      "loss": 4.941,
      "mean_token_accuracy": 0.2082672581076622,
      "num_tokens": 70073966.0,
      "step": 33770
    },
    {
      "entropy": 5.338323402404785,
      "epoch": 3.0637699564586356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040722550350489825,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2152749314904213,
      "num_tokens": 70083884.0,
      "step": 33775
    },
    {
      "entropy": 5.389253616333008,
      "epoch": 3.06422351233672,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004071993225095327,
      "loss": 4.9,
      "mean_token_accuracy": 0.2122363939881325,
      "num_tokens": 70094910.0,
      "step": 33780
    },
    {
      "entropy": 5.256891441345215,
      "epoch": 3.064677068214804,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004071731387802517,
      "loss": 4.7572,
      "mean_token_accuracy": 0.22478001564741135,
      "num_tokens": 70104480.0,
      "step": 33785
    },
    {
      "entropy": 5.234063529968262,
      "epoch": 3.0651306240928884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004071469523175964,
      "loss": 4.8914,
      "mean_token_accuracy": 0.21364732831716537,
      "num_tokens": 70114003.0,
      "step": 33790
    },
    {
      "entropy": 5.328470897674561,
      "epoch": 3.065584179970972,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040712076312210844,
      "loss": 4.9551,
      "mean_token_accuracy": 0.20867010951042175,
      "num_tokens": 70123604.0,
      "step": 33795
    },
    {
      "entropy": 5.36540961265564,
      "epoch": 3.0660377358490565,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040709457119432945,
      "loss": 4.958,
      "mean_token_accuracy": 0.2070180892944336,
      "num_tokens": 70133895.0,
      "step": 33800
    },
    {
      "entropy": 5.355367803573609,
      "epoch": 3.0664912917271407,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040706837653480113,
      "loss": 4.8841,
      "mean_token_accuracy": 0.21631893515586853,
      "num_tokens": 70145063.0,
      "step": 33805
    },
    {
      "entropy": 5.309192752838134,
      "epoch": 3.066944847605225,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000407042179144065,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20614121854305267,
      "num_tokens": 70155456.0,
      "step": 33810
    },
    {
      "entropy": 5.254522609710693,
      "epoch": 3.0673984034833093,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040701597902266306,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2166297882795334,
      "num_tokens": 70165427.0,
      "step": 33815
    },
    {
      "entropy": 5.412307214736939,
      "epoch": 3.067851959361393,
      "grad_norm": 1.03125,
      "learning_rate": 0.000406989776171137,
      "loss": 4.9582,
      "mean_token_accuracy": 0.20921160727739335,
      "num_tokens": 70175161.0,
      "step": 33820
    },
    {
      "entropy": 5.31557183265686,
      "epoch": 3.0683055152394774,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040696357059002877,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21380755603313445,
      "num_tokens": 70185026.0,
      "step": 33825
    },
    {
      "entropy": 5.340367174148559,
      "epoch": 3.0687590711175616,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004069373622798802,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21851409375667571,
      "num_tokens": 70195645.0,
      "step": 33830
    },
    {
      "entropy": 5.353116226196289,
      "epoch": 3.069212626995646,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004069111512412334,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21000400483608245,
      "num_tokens": 70206591.0,
      "step": 33835
    },
    {
      "entropy": 5.356537485122681,
      "epoch": 3.06966618287373,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040688493747463036,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21386778056621553,
      "num_tokens": 70216945.0,
      "step": 33840
    },
    {
      "entropy": 5.279181051254272,
      "epoch": 3.0701197387518144,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004068587209806132,
      "loss": 4.8265,
      "mean_token_accuracy": 0.2198666512966156,
      "num_tokens": 70226983.0,
      "step": 33845
    },
    {
      "entropy": 5.275662088394165,
      "epoch": 3.0705732946298983,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040683250175972414,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2078466981649399,
      "num_tokens": 70237730.0,
      "step": 33850
    },
    {
      "entropy": 5.305922174453736,
      "epoch": 3.0710268505079825,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004068062798125053,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20518492609262468,
      "num_tokens": 70248499.0,
      "step": 33855
    },
    {
      "entropy": 5.319008016586304,
      "epoch": 3.071480406386067,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000406780055139499,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2125723272562027,
      "num_tokens": 70259294.0,
      "step": 33860
    },
    {
      "entropy": 5.377072668075561,
      "epoch": 3.071933962264151,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004067538277412476,
      "loss": 4.969,
      "mean_token_accuracy": 0.20720113515853883,
      "num_tokens": 70269400.0,
      "step": 33865
    },
    {
      "entropy": 5.402789449691772,
      "epoch": 3.0723875181422353,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004067275976182935,
      "loss": 5.0229,
      "mean_token_accuracy": 0.20589520782232285,
      "num_tokens": 70279380.0,
      "step": 33870
    },
    {
      "entropy": 5.349596929550171,
      "epoch": 3.072841074020319,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040670136477117906,
      "loss": 4.9303,
      "mean_token_accuracy": 0.21070094555616378,
      "num_tokens": 70289919.0,
      "step": 33875
    },
    {
      "entropy": 5.36038122177124,
      "epoch": 3.0732946298984034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040667512920044684,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20994904935359954,
      "num_tokens": 70300136.0,
      "step": 33880
    },
    {
      "entropy": 5.235158109664917,
      "epoch": 3.0737481857764877,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040664889090663944,
      "loss": 4.8206,
      "mean_token_accuracy": 0.22305165529251098,
      "num_tokens": 70310072.0,
      "step": 33885
    },
    {
      "entropy": 5.351038217544556,
      "epoch": 3.074201741654572,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040662264989029933,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21088896095752716,
      "num_tokens": 70320185.0,
      "step": 33890
    },
    {
      "entropy": 5.312657356262207,
      "epoch": 3.074655297532656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004065964061519694,
      "loss": 4.8988,
      "mean_token_accuracy": 0.21698197424411775,
      "num_tokens": 70330211.0,
      "step": 33895
    },
    {
      "entropy": 5.349937105178833,
      "epoch": 3.07510885341074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004065701596921922,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20607902407646178,
      "num_tokens": 70341044.0,
      "step": 33900
    },
    {
      "entropy": 5.370581197738647,
      "epoch": 3.0755624092888243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004065439105115106,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21453426033258438,
      "num_tokens": 70351320.0,
      "step": 33905
    },
    {
      "entropy": 5.301221561431885,
      "epoch": 3.0760159651669086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040651765861046734,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22207313925027847,
      "num_tokens": 70361214.0,
      "step": 33910
    },
    {
      "entropy": 5.359299945831299,
      "epoch": 3.076469521044993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040649140398960537,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21233001351356506,
      "num_tokens": 70371083.0,
      "step": 33915
    },
    {
      "entropy": 5.387025547027588,
      "epoch": 3.076923076923077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040646514664946775,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20489591509103774,
      "num_tokens": 70381987.0,
      "step": 33920
    },
    {
      "entropy": 5.3229875564575195,
      "epoch": 3.077376632801161,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004064388865905974,
      "loss": 4.8798,
      "mean_token_accuracy": 0.21026701182127,
      "num_tokens": 70393462.0,
      "step": 33925
    },
    {
      "entropy": 5.281838655471802,
      "epoch": 3.077830188679245,
      "grad_norm": 1.0,
      "learning_rate": 0.0004064126238135373,
      "loss": 4.7821,
      "mean_token_accuracy": 0.22336730360984802,
      "num_tokens": 70403664.0,
      "step": 33930
    },
    {
      "entropy": 5.261631917953491,
      "epoch": 3.0782837445573294,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004063863583188308,
      "loss": 4.9356,
      "mean_token_accuracy": 0.20607129782438277,
      "num_tokens": 70414588.0,
      "step": 33935
    },
    {
      "entropy": 5.285341691970825,
      "epoch": 3.0787373004354137,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004063600901070207,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20983688533306122,
      "num_tokens": 70425572.0,
      "step": 33940
    },
    {
      "entropy": 5.3295135498046875,
      "epoch": 3.079190856313498,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040633381917865057,
      "loss": 4.8896,
      "mean_token_accuracy": 0.20872991681098937,
      "num_tokens": 70436140.0,
      "step": 33945
    },
    {
      "entropy": 5.329586935043335,
      "epoch": 3.079644412191582,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004063075455342636,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21371787339448928,
      "num_tokens": 70447117.0,
      "step": 33950
    },
    {
      "entropy": 5.333913040161133,
      "epoch": 3.080097968069666,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004062812691744032,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2110637456178665,
      "num_tokens": 70457756.0,
      "step": 33955
    },
    {
      "entropy": 5.310097932815552,
      "epoch": 3.0805515239477503,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004062549900996126,
      "loss": 4.96,
      "mean_token_accuracy": 0.20958720594644548,
      "num_tokens": 70468671.0,
      "step": 33960
    },
    {
      "entropy": 5.350635814666748,
      "epoch": 3.0810050798258346,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004062287083104354,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2108735755085945,
      "num_tokens": 70478212.0,
      "step": 33965
    },
    {
      "entropy": 5.2974481105804445,
      "epoch": 3.081458635703919,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040620242380741504,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21027336418628692,
      "num_tokens": 70489062.0,
      "step": 33970
    },
    {
      "entropy": 5.351611852645874,
      "epoch": 3.0819121915820027,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004061761365910951,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2116449937224388,
      "num_tokens": 70499732.0,
      "step": 33975
    },
    {
      "entropy": 5.3803187847137455,
      "epoch": 3.082365747460087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040614984666201927,
      "loss": 4.9184,
      "mean_token_accuracy": 0.20672208219766616,
      "num_tokens": 70509777.0,
      "step": 33980
    },
    {
      "entropy": 5.368807029724121,
      "epoch": 3.0828193033381712,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004061235540207312,
      "loss": 4.9524,
      "mean_token_accuracy": 0.20214006453752517,
      "num_tokens": 70520674.0,
      "step": 33985
    },
    {
      "entropy": 5.357945489883423,
      "epoch": 3.0832728592162555,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004060972586677746,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2082164466381073,
      "num_tokens": 70531115.0,
      "step": 33990
    },
    {
      "entropy": 5.2646355628967285,
      "epoch": 3.0837264150943398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004060709606036933,
      "loss": 4.8977,
      "mean_token_accuracy": 0.213665147125721,
      "num_tokens": 70540726.0,
      "step": 33995
    },
    {
      "entropy": 5.304157972335815,
      "epoch": 3.084179970972424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004060446598290311,
      "loss": 4.8775,
      "mean_token_accuracy": 0.2168179377913475,
      "num_tokens": 70551227.0,
      "step": 34000
    },
    {
      "entropy": 5.263565349578857,
      "epoch": 3.084633526850508,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000406018356344332,
      "loss": 4.9004,
      "mean_token_accuracy": 0.20720234513282776,
      "num_tokens": 70560815.0,
      "step": 34005
    },
    {
      "entropy": 5.390627431869507,
      "epoch": 3.085087082728592,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004059920501501398,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20568181425333024,
      "num_tokens": 70570247.0,
      "step": 34010
    },
    {
      "entropy": 5.352170467376709,
      "epoch": 3.0855406386066764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004059657412469987,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2070194214582443,
      "num_tokens": 70581497.0,
      "step": 34015
    },
    {
      "entropy": 5.276773881912232,
      "epoch": 3.0859941944847606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004059394296354527,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21062894016504288,
      "num_tokens": 70593133.0,
      "step": 34020
    },
    {
      "entropy": 5.323035478591919,
      "epoch": 3.086447750362845,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004059131153160459,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21356357932090758,
      "num_tokens": 70603953.0,
      "step": 34025
    },
    {
      "entropy": 5.345710325241089,
      "epoch": 3.0869013062409287,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004058867982893225,
      "loss": 4.8728,
      "mean_token_accuracy": 0.2174101948738098,
      "num_tokens": 70615292.0,
      "step": 34030
    },
    {
      "entropy": 5.378701210021973,
      "epoch": 3.087354862119013,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004058604785558268,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2030142977833748,
      "num_tokens": 70626681.0,
      "step": 34035
    },
    {
      "entropy": 5.367637062072754,
      "epoch": 3.0878084179970973,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040583415611610305,
      "loss": 4.9582,
      "mean_token_accuracy": 0.20659077763557435,
      "num_tokens": 70637339.0,
      "step": 34040
    },
    {
      "entropy": 5.386714458465576,
      "epoch": 3.0882619738751815,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004058078309706956,
      "loss": 4.8998,
      "mean_token_accuracy": 0.21472346931695938,
      "num_tokens": 70647980.0,
      "step": 34045
    },
    {
      "entropy": 5.238881301879883,
      "epoch": 3.088715529753266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004057815031201488,
      "loss": 4.7523,
      "mean_token_accuracy": 0.22548708617687224,
      "num_tokens": 70657042.0,
      "step": 34050
    },
    {
      "entropy": 5.287211751937866,
      "epoch": 3.0891690856313496,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004057551725650073,
      "loss": 4.9809,
      "mean_token_accuracy": 0.20288518965244293,
      "num_tokens": 70668542.0,
      "step": 34055
    },
    {
      "entropy": 5.35205979347229,
      "epoch": 3.089622641509434,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040572883930581544,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20739299207925796,
      "num_tokens": 70679332.0,
      "step": 34060
    },
    {
      "entropy": 5.453405284881592,
      "epoch": 3.090076197387518,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004057025033431179,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20404285490512847,
      "num_tokens": 70689897.0,
      "step": 34065
    },
    {
      "entropy": 5.321261596679688,
      "epoch": 3.0905297532656024,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040567616467745933,
      "loss": 4.8782,
      "mean_token_accuracy": 0.21727245599031447,
      "num_tokens": 70699965.0,
      "step": 34070
    },
    {
      "entropy": 5.353748178482055,
      "epoch": 3.0909833091436867,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004056498233093843,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20825624018907546,
      "num_tokens": 70710583.0,
      "step": 34075
    },
    {
      "entropy": 5.317381572723389,
      "epoch": 3.0914368650217705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040562347923943766,
      "loss": 4.881,
      "mean_token_accuracy": 0.22130576074123381,
      "num_tokens": 70721800.0,
      "step": 34080
    },
    {
      "entropy": 5.3168388366699215,
      "epoch": 3.0918904208998548,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004055971324681642,
      "loss": 4.9323,
      "mean_token_accuracy": 0.20572961568832399,
      "num_tokens": 70731527.0,
      "step": 34085
    },
    {
      "entropy": 5.349139261245727,
      "epoch": 3.092343976777939,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004055707829961087,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2085531011223793,
      "num_tokens": 70740821.0,
      "step": 34090
    },
    {
      "entropy": 5.275716400146484,
      "epoch": 3.0927975326560233,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040554443082381627,
      "loss": 4.8819,
      "mean_token_accuracy": 0.21117914319038392,
      "num_tokens": 70750506.0,
      "step": 34095
    },
    {
      "entropy": 5.283831930160522,
      "epoch": 3.0932510885341076,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004055180759518315,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2129928722977638,
      "num_tokens": 70762009.0,
      "step": 34100
    },
    {
      "entropy": 5.341449356079101,
      "epoch": 3.0937046444121914,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004054917183806999,
      "loss": 4.8873,
      "mean_token_accuracy": 0.21395602226257324,
      "num_tokens": 70772030.0,
      "step": 34105
    },
    {
      "entropy": 5.362677431106567,
      "epoch": 3.0941582002902757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040546535811096626,
      "loss": 4.9447,
      "mean_token_accuracy": 0.2070620536804199,
      "num_tokens": 70782539.0,
      "step": 34110
    },
    {
      "entropy": 5.43872709274292,
      "epoch": 3.09461175616836,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040543899514317564,
      "loss": 5.0813,
      "mean_token_accuracy": 0.19735892564058305,
      "num_tokens": 70793433.0,
      "step": 34115
    },
    {
      "entropy": 5.340062952041626,
      "epoch": 3.095065312046444,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004054126294778735,
      "loss": 4.9601,
      "mean_token_accuracy": 0.20477636456489562,
      "num_tokens": 70802919.0,
      "step": 34120
    },
    {
      "entropy": 5.342052221298218,
      "epoch": 3.0955188679245285,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040538626111560487,
      "loss": 4.8725,
      "mean_token_accuracy": 0.2130136176943779,
      "num_tokens": 70813827.0,
      "step": 34125
    },
    {
      "entropy": 5.374582147598266,
      "epoch": 3.0959724238026123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040535989005691515,
      "loss": 4.99,
      "mean_token_accuracy": 0.21107760667800904,
      "num_tokens": 70824117.0,
      "step": 34130
    },
    {
      "entropy": 5.249261951446533,
      "epoch": 3.0964259796806965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040533351630234985,
      "loss": 4.8537,
      "mean_token_accuracy": 0.21756429523229598,
      "num_tokens": 70833387.0,
      "step": 34135
    },
    {
      "entropy": 5.334499406814575,
      "epoch": 3.096879535558781,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000405307139852454,
      "loss": 4.8891,
      "mean_token_accuracy": 0.21065168529748918,
      "num_tokens": 70844363.0,
      "step": 34140
    },
    {
      "entropy": 5.346499109268189,
      "epoch": 3.097333091436865,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040528076070777335,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21014628410339356,
      "num_tokens": 70854759.0,
      "step": 34145
    },
    {
      "entropy": 5.2437115669250485,
      "epoch": 3.0977866473149493,
      "grad_norm": 1.0,
      "learning_rate": 0.0004052543788688534,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21214858144521714,
      "num_tokens": 70865682.0,
      "step": 34150
    },
    {
      "entropy": 5.4096386432647705,
      "epoch": 3.098240203193033,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004052279943362396,
      "loss": 4.9817,
      "mean_token_accuracy": 0.2081989362835884,
      "num_tokens": 70875449.0,
      "step": 34155
    },
    {
      "entropy": 5.369579696655274,
      "epoch": 3.0986937590711174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004052016071104779,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21194473057985305,
      "num_tokens": 70885885.0,
      "step": 34160
    },
    {
      "entropy": 5.338026714324951,
      "epoch": 3.0991473149492017,
      "grad_norm": 1.125,
      "learning_rate": 0.0004051752171921136,
      "loss": 4.8974,
      "mean_token_accuracy": 0.20905450731515884,
      "num_tokens": 70896320.0,
      "step": 34165
    },
    {
      "entropy": 5.277422380447388,
      "epoch": 3.099600870827286,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040514882458169267,
      "loss": 4.8353,
      "mean_token_accuracy": 0.21556879729032516,
      "num_tokens": 70907074.0,
      "step": 34170
    },
    {
      "entropy": 5.282642221450805,
      "epoch": 3.1000544267053702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004051224292797609,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2083863139152527,
      "num_tokens": 70916567.0,
      "step": 34175
    },
    {
      "entropy": 5.313330841064453,
      "epoch": 3.100507982583454,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040509603128686413,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2190946340560913,
      "num_tokens": 70926163.0,
      "step": 34180
    },
    {
      "entropy": 5.369273996353149,
      "epoch": 3.1009615384615383,
      "grad_norm": 1.0,
      "learning_rate": 0.0004050696306035483,
      "loss": 4.9314,
      "mean_token_accuracy": 0.20535815209150315,
      "num_tokens": 70937358.0,
      "step": 34185
    },
    {
      "entropy": 5.3005656719207765,
      "epoch": 3.1014150943396226,
      "grad_norm": 1.0,
      "learning_rate": 0.00040504322723035937,
      "loss": 4.8971,
      "mean_token_accuracy": 0.21034999787807465,
      "num_tokens": 70949305.0,
      "step": 34190
    },
    {
      "entropy": 5.262331247329712,
      "epoch": 3.101868650217707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040501682116784326,
      "loss": 4.8837,
      "mean_token_accuracy": 0.21318188160657883,
      "num_tokens": 70959146.0,
      "step": 34195
    },
    {
      "entropy": 5.259044647216797,
      "epoch": 3.102322206095791,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040499041241654626,
      "loss": 4.9252,
      "mean_token_accuracy": 0.21121613979339598,
      "num_tokens": 70968828.0,
      "step": 34200
    },
    {
      "entropy": 5.295136785507202,
      "epoch": 3.1027757619738754,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004049640009770144,
      "loss": 4.9046,
      "mean_token_accuracy": 0.2177492320537567,
      "num_tokens": 70979733.0,
      "step": 34205
    },
    {
      "entropy": 5.374333381652832,
      "epoch": 3.103229317851959,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004049375868497939,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21004872471094133,
      "num_tokens": 70989726.0,
      "step": 34210
    },
    {
      "entropy": 5.375807094573974,
      "epoch": 3.1036828737300435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040491117003543087,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2092895895242691,
      "num_tokens": 71000715.0,
      "step": 34215
    },
    {
      "entropy": 5.24205846786499,
      "epoch": 3.1041364296081277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040488475053447186,
      "loss": 4.7545,
      "mean_token_accuracy": 0.22518436312675477,
      "num_tokens": 71011453.0,
      "step": 34220
    },
    {
      "entropy": 5.314016389846802,
      "epoch": 3.104589985486212,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040485832834746306,
      "loss": 4.9319,
      "mean_token_accuracy": 0.20980067551136017,
      "num_tokens": 71021477.0,
      "step": 34225
    },
    {
      "entropy": 5.296921014785767,
      "epoch": 3.1050435413642963,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040483190347495097,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2086577832698822,
      "num_tokens": 71031706.0,
      "step": 34230
    },
    {
      "entropy": 5.3510795593261715,
      "epoch": 3.10549709724238,
      "grad_norm": 1.109375,
      "learning_rate": 0.000404805475917482,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2124472066760063,
      "num_tokens": 71042882.0,
      "step": 34235
    },
    {
      "entropy": 5.291550731658935,
      "epoch": 3.1059506531204644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040477904567560275,
      "loss": 4.892,
      "mean_token_accuracy": 0.21894123256206513,
      "num_tokens": 71052804.0,
      "step": 34240
    },
    {
      "entropy": 5.358235025405884,
      "epoch": 3.1064042089985486,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004047526127498597,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20537713766098023,
      "num_tokens": 71064349.0,
      "step": 34245
    },
    {
      "entropy": 5.318432426452636,
      "epoch": 3.106857764876633,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040472617714079964,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21259081363677979,
      "num_tokens": 71075754.0,
      "step": 34250
    },
    {
      "entropy": 5.330706453323364,
      "epoch": 3.107311320754717,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004046997388489691,
      "loss": 4.8648,
      "mean_token_accuracy": 0.21359627991914748,
      "num_tokens": 71085338.0,
      "step": 34255
    },
    {
      "entropy": 5.222161293029785,
      "epoch": 3.107764876632801,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004046732978749149,
      "loss": 4.8547,
      "mean_token_accuracy": 0.21300897151231765,
      "num_tokens": 71097089.0,
      "step": 34260
    },
    {
      "entropy": 5.251526689529419,
      "epoch": 3.1082184325108853,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000404646854219184,
      "loss": 4.8359,
      "mean_token_accuracy": 0.2206099435687065,
      "num_tokens": 71106990.0,
      "step": 34265
    },
    {
      "entropy": 5.293855285644531,
      "epoch": 3.1086719883889695,
      "grad_norm": 0.9375,
      "learning_rate": 0.000404620407882323,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2128816306591034,
      "num_tokens": 71117634.0,
      "step": 34270
    },
    {
      "entropy": 5.245510005950928,
      "epoch": 3.109125544267054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040459395886487885,
      "loss": 4.8217,
      "mean_token_accuracy": 0.21085273921489717,
      "num_tokens": 71127717.0,
      "step": 34275
    },
    {
      "entropy": 5.265050840377808,
      "epoch": 3.109579100145138,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004045675071673988,
      "loss": 4.8541,
      "mean_token_accuracy": 0.2188464730978012,
      "num_tokens": 71137172.0,
      "step": 34280
    },
    {
      "entropy": 5.369022607803345,
      "epoch": 3.110032656023222,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004045410527904296,
      "loss": 4.9484,
      "mean_token_accuracy": 0.20683628916740418,
      "num_tokens": 71146843.0,
      "step": 34285
    },
    {
      "entropy": 5.262702417373657,
      "epoch": 3.110486211901306,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004045145957345184,
      "loss": 4.8283,
      "mean_token_accuracy": 0.22046493738889694,
      "num_tokens": 71156619.0,
      "step": 34290
    },
    {
      "entropy": 5.30780553817749,
      "epoch": 3.1109397677793904,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004044881360002123,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21425212174654007,
      "num_tokens": 71167234.0,
      "step": 34295
    },
    {
      "entropy": 5.268366861343384,
      "epoch": 3.1113933236574747,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004044616735880586,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21645151376724242,
      "num_tokens": 71178049.0,
      "step": 34300
    },
    {
      "entropy": 5.2862176418304445,
      "epoch": 3.111846879535559,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040443520849860453,
      "loss": 4.823,
      "mean_token_accuracy": 0.21109345257282258,
      "num_tokens": 71188568.0,
      "step": 34305
    },
    {
      "entropy": 5.374424648284912,
      "epoch": 3.1123004354136428,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040440874073239733,
      "loss": 4.9229,
      "mean_token_accuracy": 0.20889021456241608,
      "num_tokens": 71199004.0,
      "step": 34310
    },
    {
      "entropy": 5.2841521263122555,
      "epoch": 3.112753991291727,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040438227028998436,
      "loss": 4.9006,
      "mean_token_accuracy": 0.21164056211709975,
      "num_tokens": 71209239.0,
      "step": 34315
    },
    {
      "entropy": 5.292957305908203,
      "epoch": 3.1132075471698113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040435579717191313,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20960231870412827,
      "num_tokens": 71219469.0,
      "step": 34320
    },
    {
      "entropy": 5.355325078964233,
      "epoch": 3.1136611030478956,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040432932137873107,
      "loss": 4.9747,
      "mean_token_accuracy": 0.20236138850450516,
      "num_tokens": 71230670.0,
      "step": 34325
    },
    {
      "entropy": 5.387889289855957,
      "epoch": 3.11411465892598,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004043028429109857,
      "loss": 5.0348,
      "mean_token_accuracy": 0.2056300938129425,
      "num_tokens": 71241512.0,
      "step": 34330
    },
    {
      "entropy": 5.293631315231323,
      "epoch": 3.1145682148040637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004042763617692245,
      "loss": 4.8571,
      "mean_token_accuracy": 0.21545981615781784,
      "num_tokens": 71250869.0,
      "step": 34335
    },
    {
      "entropy": 5.269321060180664,
      "epoch": 3.115021770682148,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040424987795399523,
      "loss": 4.8534,
      "mean_token_accuracy": 0.21567787528038024,
      "num_tokens": 71260577.0,
      "step": 34340
    },
    {
      "entropy": 5.329599285125733,
      "epoch": 3.115475326560232,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004042233914658455,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21119116693735124,
      "num_tokens": 71270185.0,
      "step": 34345
    },
    {
      "entropy": 5.28538670539856,
      "epoch": 3.1159288824383164,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040419690230532324,
      "loss": 4.8593,
      "mean_token_accuracy": 0.22054924070835114,
      "num_tokens": 71279427.0,
      "step": 34350
    },
    {
      "entropy": 5.288170003890992,
      "epoch": 3.1163824383164007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040417041047297606,
      "loss": 4.8486,
      "mean_token_accuracy": 0.21136364191770554,
      "num_tokens": 71290857.0,
      "step": 34355
    },
    {
      "entropy": 5.327934789657593,
      "epoch": 3.116835994194485,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004041439159693518,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21535975933074952,
      "num_tokens": 71300971.0,
      "step": 34360
    },
    {
      "entropy": 5.378671264648437,
      "epoch": 3.117289550072569,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004041174187949985,
      "loss": 5.0388,
      "mean_token_accuracy": 0.20008520036935806,
      "num_tokens": 71311017.0,
      "step": 34365
    },
    {
      "entropy": 5.3916455745697025,
      "epoch": 3.117743105950653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040409091895046413,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2013016939163208,
      "num_tokens": 71321546.0,
      "step": 34370
    },
    {
      "entropy": 5.3474798679351805,
      "epoch": 3.1181966618287373,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040406441643629656,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2099734902381897,
      "num_tokens": 71332345.0,
      "step": 34375
    },
    {
      "entropy": 5.404249000549316,
      "epoch": 3.1186502177068216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040403791125304405,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20711077004671097,
      "num_tokens": 71342670.0,
      "step": 34380
    },
    {
      "entropy": 5.326215744018555,
      "epoch": 3.119103773584906,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004040114034012546,
      "loss": 4.8402,
      "mean_token_accuracy": 0.21430433839559554,
      "num_tokens": 71353440.0,
      "step": 34385
    },
    {
      "entropy": 5.308976030349731,
      "epoch": 3.1195573294629897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040398489288147646,
      "loss": 4.9818,
      "mean_token_accuracy": 0.2141313672065735,
      "num_tokens": 71363199.0,
      "step": 34390
    },
    {
      "entropy": 5.349179315567016,
      "epoch": 3.120010885341074,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040395837969425786,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21302036941051483,
      "num_tokens": 71372925.0,
      "step": 34395
    },
    {
      "entropy": 5.368973588943481,
      "epoch": 3.1204644412191582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004039318638401471,
      "loss": 4.9283,
      "mean_token_accuracy": 0.20836957693099975,
      "num_tokens": 71383350.0,
      "step": 34400
    },
    {
      "entropy": 5.27810378074646,
      "epoch": 3.1209179970972425,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004039053453196926,
      "loss": 4.8663,
      "mean_token_accuracy": 0.2189398378133774,
      "num_tokens": 71394006.0,
      "step": 34405
    },
    {
      "entropy": 5.278778553009033,
      "epoch": 3.1213715529753268,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004038788241334427,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21342604905366896,
      "num_tokens": 71403945.0,
      "step": 34410
    },
    {
      "entropy": 5.35325026512146,
      "epoch": 3.1218251088534106,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004038523002819458,
      "loss": 4.9313,
      "mean_token_accuracy": 0.20747089385986328,
      "num_tokens": 71414238.0,
      "step": 34415
    },
    {
      "entropy": 5.288902711868286,
      "epoch": 3.122278664731495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040382577376575055,
      "loss": 4.9307,
      "mean_token_accuracy": 0.20681874454021454,
      "num_tokens": 71423586.0,
      "step": 34420
    },
    {
      "entropy": 5.326851177215576,
      "epoch": 3.122732220609579,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004037992445854054,
      "loss": 4.8786,
      "mean_token_accuracy": 0.21989890038967133,
      "num_tokens": 71434261.0,
      "step": 34425
    },
    {
      "entropy": 5.324724245071411,
      "epoch": 3.1231857764876634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040377271274145916,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21293159574270248,
      "num_tokens": 71443952.0,
      "step": 34430
    },
    {
      "entropy": 5.352297592163086,
      "epoch": 3.1236393323657476,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004037461782344603,
      "loss": 4.8831,
      "mean_token_accuracy": 0.21124214082956314,
      "num_tokens": 71454051.0,
      "step": 34435
    },
    {
      "entropy": 5.290301418304443,
      "epoch": 3.1240928882438315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040371964106495777,
      "loss": 4.8956,
      "mean_token_accuracy": 0.21713093668222427,
      "num_tokens": 71464579.0,
      "step": 34440
    },
    {
      "entropy": 5.332940673828125,
      "epoch": 3.1245464441219157,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040369310123350015,
      "loss": 4.9087,
      "mean_token_accuracy": 0.21426438689231872,
      "num_tokens": 71474631.0,
      "step": 34445
    },
    {
      "entropy": 5.401003122329712,
      "epoch": 3.125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040366655874063645,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21480478197336197,
      "num_tokens": 71485266.0,
      "step": 34450
    },
    {
      "entropy": 5.271424245834351,
      "epoch": 3.1254535558780843,
      "grad_norm": 1.0,
      "learning_rate": 0.0004036400135869156,
      "loss": 4.7679,
      "mean_token_accuracy": 0.22598811984062195,
      "num_tokens": 71495248.0,
      "step": 34455
    },
    {
      "entropy": 5.276937341690063,
      "epoch": 3.1259071117561685,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040361346577288645,
      "loss": 4.8801,
      "mean_token_accuracy": 0.21456499248743058,
      "num_tokens": 71505600.0,
      "step": 34460
    },
    {
      "entropy": 5.312733745574951,
      "epoch": 3.1263606676342524,
      "grad_norm": 1.0,
      "learning_rate": 0.00040358691529909796,
      "loss": 4.9339,
      "mean_token_accuracy": 0.20895766317844391,
      "num_tokens": 71516634.0,
      "step": 34465
    },
    {
      "entropy": 5.359246158599854,
      "epoch": 3.1268142235123366,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040356036216609943,
      "loss": 4.8996,
      "mean_token_accuracy": 0.20912552922964095,
      "num_tokens": 71526745.0,
      "step": 34470
    },
    {
      "entropy": 5.332573461532593,
      "epoch": 3.127267779390421,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040353380637443974,
      "loss": 5.0378,
      "mean_token_accuracy": 0.19783654659986497,
      "num_tokens": 71537015.0,
      "step": 34475
    },
    {
      "entropy": 5.286272668838501,
      "epoch": 3.127721335268505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040350724792466826,
      "loss": 4.7746,
      "mean_token_accuracy": 0.2186601683497429,
      "num_tokens": 71546660.0,
      "step": 34480
    },
    {
      "entropy": 5.3265793800354,
      "epoch": 3.1281748911465894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004034806868173341,
      "loss": 4.9387,
      "mean_token_accuracy": 0.21322117745876312,
      "num_tokens": 71557309.0,
      "step": 34485
    },
    {
      "entropy": 5.281906938552856,
      "epoch": 3.1286284470246732,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004034541230529866,
      "loss": 4.8124,
      "mean_token_accuracy": 0.21497984677553178,
      "num_tokens": 71567651.0,
      "step": 34490
    },
    {
      "entropy": 5.271552562713623,
      "epoch": 3.1290820029027575,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040342755663217507,
      "loss": 4.8463,
      "mean_token_accuracy": 0.21691277325153352,
      "num_tokens": 71577538.0,
      "step": 34495
    },
    {
      "entropy": 5.268912744522095,
      "epoch": 3.1295355587808418,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000403400987555449,
      "loss": 4.8681,
      "mean_token_accuracy": 0.21860136687755585,
      "num_tokens": 71588299.0,
      "step": 34500
    },
    {
      "entropy": 5.330367469787598,
      "epoch": 3.129989114658926,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004033744158233577,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21227826178073883,
      "num_tokens": 71598989.0,
      "step": 34505
    },
    {
      "entropy": 5.3390038967132565,
      "epoch": 3.1304426705370103,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040334784143645087,
      "loss": 4.8973,
      "mean_token_accuracy": 0.208171109855175,
      "num_tokens": 71608596.0,
      "step": 34510
    },
    {
      "entropy": 5.220809602737427,
      "epoch": 3.1308962264150946,
      "grad_norm": 1.125,
      "learning_rate": 0.0004033212643952779,
      "loss": 4.8636,
      "mean_token_accuracy": 0.21748322546482085,
      "num_tokens": 71619182.0,
      "step": 34515
    },
    {
      "entropy": 5.280950307846069,
      "epoch": 3.1313497822931784,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040329468470038857,
      "loss": 4.9296,
      "mean_token_accuracy": 0.2089809149503708,
      "num_tokens": 71629526.0,
      "step": 34520
    },
    {
      "entropy": 5.374242925643921,
      "epoch": 3.1318033381712627,
      "grad_norm": 1.125,
      "learning_rate": 0.0004032681023523324,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2124768689274788,
      "num_tokens": 71640146.0,
      "step": 34525
    },
    {
      "entropy": 5.306454133987427,
      "epoch": 3.132256894049347,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004032415173516592,
      "loss": 4.8086,
      "mean_token_accuracy": 0.21620590537786483,
      "num_tokens": 71649652.0,
      "step": 34530
    },
    {
      "entropy": 5.302905893325805,
      "epoch": 3.132710449927431,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004032149296989187,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21732944548130034,
      "num_tokens": 71659264.0,
      "step": 34535
    },
    {
      "entropy": 5.289631795883179,
      "epoch": 3.133164005805515,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040318833939466083,
      "loss": 4.9143,
      "mean_token_accuracy": 0.20601845681667327,
      "num_tokens": 71669175.0,
      "step": 34540
    },
    {
      "entropy": 5.311059427261353,
      "epoch": 3.1336175616835993,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004031617464394354,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2077139750123024,
      "num_tokens": 71679537.0,
      "step": 34545
    },
    {
      "entropy": 5.338895845413208,
      "epoch": 3.1340711175616836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004031351508337925,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21211262345314025,
      "num_tokens": 71690024.0,
      "step": 34550
    },
    {
      "entropy": 5.3877839088439945,
      "epoch": 3.134524673439768,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004031085525782819,
      "loss": 4.9218,
      "mean_token_accuracy": 0.21461703777313232,
      "num_tokens": 71700019.0,
      "step": 34555
    },
    {
      "entropy": 5.36411828994751,
      "epoch": 3.134978229317852,
      "grad_norm": 1.0,
      "learning_rate": 0.0004030819516734539,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20359236747026443,
      "num_tokens": 71710570.0,
      "step": 34560
    },
    {
      "entropy": 5.293434572219849,
      "epoch": 3.1354317851959363,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040305534811985844,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21742439419031143,
      "num_tokens": 71720619.0,
      "step": 34565
    },
    {
      "entropy": 5.300870752334594,
      "epoch": 3.13588534107402,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004030287419180457,
      "loss": 4.8825,
      "mean_token_accuracy": 0.21737452894449233,
      "num_tokens": 71730201.0,
      "step": 34570
    },
    {
      "entropy": 5.3427647113800045,
      "epoch": 3.1363388969521044,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040300213306856606,
      "loss": 4.9378,
      "mean_token_accuracy": 0.214421084523201,
      "num_tokens": 71741368.0,
      "step": 34575
    },
    {
      "entropy": 5.264199304580688,
      "epoch": 3.1367924528301887,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004029755215719696,
      "loss": 4.8431,
      "mean_token_accuracy": 0.21674838662147522,
      "num_tokens": 71751361.0,
      "step": 34580
    },
    {
      "entropy": 5.319463014602661,
      "epoch": 3.137246008708273,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040294890742880677,
      "loss": 4.8866,
      "mean_token_accuracy": 0.20875578075647355,
      "num_tokens": 71761164.0,
      "step": 34585
    },
    {
      "entropy": 5.282718992233276,
      "epoch": 3.1376995645863572,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040292229063962787,
      "loss": 4.8851,
      "mean_token_accuracy": 0.21948029696941376,
      "num_tokens": 71770726.0,
      "step": 34590
    },
    {
      "entropy": 5.356213331222534,
      "epoch": 3.138153120464441,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004028956712049834,
      "loss": 4.9595,
      "mean_token_accuracy": 0.21462424993515014,
      "num_tokens": 71781124.0,
      "step": 34595
    },
    {
      "entropy": 5.284512090682983,
      "epoch": 3.1386066763425253,
      "grad_norm": 1.0,
      "learning_rate": 0.000402869049125424,
      "loss": 4.8767,
      "mean_token_accuracy": 0.20794234424829483,
      "num_tokens": 71791218.0,
      "step": 34600
    },
    {
      "entropy": 5.33690333366394,
      "epoch": 3.1390602322206096,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004028424244015,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2132733017206192,
      "num_tokens": 71801013.0,
      "step": 34605
    },
    {
      "entropy": 5.324379110336304,
      "epoch": 3.139513788098694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000402815797033762,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2028840497136116,
      "num_tokens": 71811742.0,
      "step": 34610
    },
    {
      "entropy": 5.237501096725464,
      "epoch": 3.139967343976778,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004027891670227609,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2078912854194641,
      "num_tokens": 71822265.0,
      "step": 34615
    },
    {
      "entropy": 5.3271852970123295,
      "epoch": 3.140420899854862,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004027625343690472,
      "loss": 4.8738,
      "mean_token_accuracy": 0.21308585554361342,
      "num_tokens": 71832090.0,
      "step": 34620
    },
    {
      "entropy": 5.315815830230713,
      "epoch": 3.140874455732946,
      "grad_norm": 1.0,
      "learning_rate": 0.0004027358990731717,
      "loss": 4.8788,
      "mean_token_accuracy": 0.21446592658758162,
      "num_tokens": 71841381.0,
      "step": 34625
    },
    {
      "entropy": 5.29408745765686,
      "epoch": 3.1413280116110305,
      "grad_norm": 1.0,
      "learning_rate": 0.00040270926113568524,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2126562252640724,
      "num_tokens": 71851383.0,
      "step": 34630
    },
    {
      "entropy": 5.258868837356568,
      "epoch": 3.1417815674891147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040268262055713873,
      "loss": 4.8773,
      "mean_token_accuracy": 0.21313113272190093,
      "num_tokens": 71862588.0,
      "step": 34635
    },
    {
      "entropy": 5.265064430236817,
      "epoch": 3.142235123367199,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004026559773380831,
      "loss": 4.8766,
      "mean_token_accuracy": 0.21203845292329787,
      "num_tokens": 71873975.0,
      "step": 34640
    },
    {
      "entropy": 5.35780611038208,
      "epoch": 3.142688679245283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004026293314790694,
      "loss": 4.9634,
      "mean_token_accuracy": 0.19934522062540055,
      "num_tokens": 71884084.0,
      "step": 34645
    },
    {
      "entropy": 5.29548134803772,
      "epoch": 3.143142235123367,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004026026829806485,
      "loss": 4.9045,
      "mean_token_accuracy": 0.20822277963161467,
      "num_tokens": 71894643.0,
      "step": 34650
    },
    {
      "entropy": 5.238706016540528,
      "epoch": 3.1435957910014514,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040257603184337167,
      "loss": 4.8509,
      "mean_token_accuracy": 0.21365768015384673,
      "num_tokens": 71905245.0,
      "step": 34655
    },
    {
      "entropy": 5.364257383346557,
      "epoch": 3.1440493468795356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040254937806778995,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20491572618484497,
      "num_tokens": 71916473.0,
      "step": 34660
    },
    {
      "entropy": 5.2871551513671875,
      "epoch": 3.14450290275762,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040252272165445456,
      "loss": 4.9063,
      "mean_token_accuracy": 0.2113608717918396,
      "num_tokens": 71927465.0,
      "step": 34665
    },
    {
      "entropy": 5.383847999572754,
      "epoch": 3.1449564586357037,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004024960626039168,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2091105028986931,
      "num_tokens": 71937380.0,
      "step": 34670
    },
    {
      "entropy": 5.336333847045898,
      "epoch": 3.145410014513788,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040246940091672806,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21523355096578597,
      "num_tokens": 71948088.0,
      "step": 34675
    },
    {
      "entropy": 5.269011497497559,
      "epoch": 3.1458635703918723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004024427365934395,
      "loss": 4.857,
      "mean_token_accuracy": 0.22031983435153962,
      "num_tokens": 71958284.0,
      "step": 34680
    },
    {
      "entropy": 5.242222595214844,
      "epoch": 3.1463171262699565,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040241606963460283,
      "loss": 4.9792,
      "mean_token_accuracy": 0.20530966967344283,
      "num_tokens": 71968138.0,
      "step": 34685
    },
    {
      "entropy": 5.356296443939209,
      "epoch": 3.146770682148041,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040238940004076924,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21094761341810225,
      "num_tokens": 71978275.0,
      "step": 34690
    },
    {
      "entropy": 5.309158039093018,
      "epoch": 3.1472242380261246,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040236272781249047,
      "loss": 4.9272,
      "mean_token_accuracy": 0.2090173214673996,
      "num_tokens": 71988875.0,
      "step": 34695
    },
    {
      "entropy": 5.305198049545288,
      "epoch": 3.147677793904209,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040233605295031795,
      "loss": 4.8832,
      "mean_token_accuracy": 0.21325127482414247,
      "num_tokens": 71999560.0,
      "step": 34700
    },
    {
      "entropy": 5.284794521331787,
      "epoch": 3.148131349782293,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004023093754548034,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21716900765895844,
      "num_tokens": 72009970.0,
      "step": 34705
    },
    {
      "entropy": 5.296391534805298,
      "epoch": 3.1485849056603774,
      "grad_norm": 1.0,
      "learning_rate": 0.0004022826953264986,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22175343185663224,
      "num_tokens": 72020196.0,
      "step": 34710
    },
    {
      "entropy": 5.299097776412964,
      "epoch": 3.1490384615384617,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040225601256595525,
      "loss": 4.8623,
      "mean_token_accuracy": 0.21055195927619935,
      "num_tokens": 72031456.0,
      "step": 34715
    },
    {
      "entropy": 5.330206346511841,
      "epoch": 3.149492017416546,
      "grad_norm": 1.046875,
      "learning_rate": 0.000402229327173725,
      "loss": 4.9639,
      "mean_token_accuracy": 0.20690548568964004,
      "num_tokens": 72042260.0,
      "step": 34720
    },
    {
      "entropy": 5.422563600540161,
      "epoch": 3.1499455732946298,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040220263915035997,
      "loss": 5.0431,
      "mean_token_accuracy": 0.19109285473823548,
      "num_tokens": 72053463.0,
      "step": 34725
    },
    {
      "entropy": 5.425201559066773,
      "epoch": 3.150399129172714,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040217594849641195,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20082563161849976,
      "num_tokens": 72064508.0,
      "step": 34730
    },
    {
      "entropy": 5.2701709270477295,
      "epoch": 3.1508526850507983,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040214925521243287,
      "loss": 4.9195,
      "mean_token_accuracy": 0.20554297119379045,
      "num_tokens": 72074758.0,
      "step": 34735
    },
    {
      "entropy": 5.278521871566772,
      "epoch": 3.1513062409288826,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040212255929897474,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2194508582353592,
      "num_tokens": 72085548.0,
      "step": 34740
    },
    {
      "entropy": 5.278566122055054,
      "epoch": 3.1517597968069664,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004020958607565897,
      "loss": 4.9988,
      "mean_token_accuracy": 0.20651766508817673,
      "num_tokens": 72096773.0,
      "step": 34745
    },
    {
      "entropy": 5.321801233291626,
      "epoch": 3.1522133526850507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004020691595858299,
      "loss": 4.9745,
      "mean_token_accuracy": 0.21180784106254577,
      "num_tokens": 72106832.0,
      "step": 34750
    },
    {
      "entropy": 5.345190572738647,
      "epoch": 3.152666908563135,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040204245578724754,
      "loss": 4.9531,
      "mean_token_accuracy": 0.21203841865062714,
      "num_tokens": 72117758.0,
      "step": 34755
    },
    {
      "entropy": 5.345208835601807,
      "epoch": 3.153120464441219,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040201574936139475,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2112305596470833,
      "num_tokens": 72127522.0,
      "step": 34760
    },
    {
      "entropy": 5.239628601074219,
      "epoch": 3.1535740203193035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004019890403088239,
      "loss": 4.9414,
      "mean_token_accuracy": 0.214928737282753,
      "num_tokens": 72138310.0,
      "step": 34765
    },
    {
      "entropy": 5.293467092514038,
      "epoch": 3.1540275761973877,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040196232863008735,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2188929408788681,
      "num_tokens": 72148958.0,
      "step": 34770
    },
    {
      "entropy": 5.373406076431275,
      "epoch": 3.1544811320754715,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040193561432573754,
      "loss": 4.9002,
      "mean_token_accuracy": 0.21552331298589705,
      "num_tokens": 72158594.0,
      "step": 34775
    },
    {
      "entropy": 5.391397285461426,
      "epoch": 3.154934687953556,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004019088973963268,
      "loss": 4.9937,
      "mean_token_accuracy": 0.21007902175188065,
      "num_tokens": 72168552.0,
      "step": 34780
    },
    {
      "entropy": 5.33831934928894,
      "epoch": 3.15538824383164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040188217784240785,
      "loss": 4.8655,
      "mean_token_accuracy": 0.2145914077758789,
      "num_tokens": 72178584.0,
      "step": 34785
    },
    {
      "entropy": 5.246712017059326,
      "epoch": 3.1558417997097243,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000401855455664533,
      "loss": 4.8877,
      "mean_token_accuracy": 0.2183474063873291,
      "num_tokens": 72189144.0,
      "step": 34790
    },
    {
      "entropy": 5.348818588256836,
      "epoch": 3.1562953555878086,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004018287308632551,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21322552859783173,
      "num_tokens": 72198326.0,
      "step": 34795
    },
    {
      "entropy": 5.350150108337402,
      "epoch": 3.1567489114658924,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004018020034391266,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21354753524065018,
      "num_tokens": 72207916.0,
      "step": 34800
    },
    {
      "entropy": 5.35081672668457,
      "epoch": 3.1572024673439767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004017752733927004,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21852322816848754,
      "num_tokens": 72219111.0,
      "step": 34805
    },
    {
      "entropy": 5.300081491470337,
      "epoch": 3.157656023222061,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004017485407245293,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2169164925813675,
      "num_tokens": 72230081.0,
      "step": 34810
    },
    {
      "entropy": 5.2609930515289305,
      "epoch": 3.1581095791001452,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000401721805435166,
      "loss": 4.8765,
      "mean_token_accuracy": 0.21538702696561812,
      "num_tokens": 72240693.0,
      "step": 34815
    },
    {
      "entropy": 5.277742624282837,
      "epoch": 3.1585631349782295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040169506752516356,
      "loss": 4.8378,
      "mean_token_accuracy": 0.22587065994739533,
      "num_tokens": 72250925.0,
      "step": 34820
    },
    {
      "entropy": 5.262257099151611,
      "epoch": 3.1590166908563133,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004016683269950748,
      "loss": 4.8251,
      "mean_token_accuracy": 0.22468801140785216,
      "num_tokens": 72261027.0,
      "step": 34825
    },
    {
      "entropy": 5.289924573898316,
      "epoch": 3.1594702467343976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040164158384545267,
      "loss": 4.7906,
      "mean_token_accuracy": 0.22900804728269578,
      "num_tokens": 72269430.0,
      "step": 34830
    },
    {
      "entropy": 5.211703109741211,
      "epoch": 3.159923802612482,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004016148380768504,
      "loss": 4.8453,
      "mean_token_accuracy": 0.21914753317832947,
      "num_tokens": 72280207.0,
      "step": 34835
    },
    {
      "entropy": 5.282384252548217,
      "epoch": 3.160377358490566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040158808968982103,
      "loss": 5.0118,
      "mean_token_accuracy": 0.2111773356795311,
      "num_tokens": 72290093.0,
      "step": 34840
    },
    {
      "entropy": 5.290272045135498,
      "epoch": 3.1608309143686504,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040156133868491766,
      "loss": 4.827,
      "mean_token_accuracy": 0.2172797992825508,
      "num_tokens": 72300203.0,
      "step": 34845
    },
    {
      "entropy": 5.314862585067749,
      "epoch": 3.161284470246734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040153458506269365,
      "loss": 4.9069,
      "mean_token_accuracy": 0.212433061003685,
      "num_tokens": 72311778.0,
      "step": 34850
    },
    {
      "entropy": 5.329032516479492,
      "epoch": 3.1617380261248185,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000401507828823702,
      "loss": 4.8849,
      "mean_token_accuracy": 0.20768565982580184,
      "num_tokens": 72322142.0,
      "step": 34855
    },
    {
      "entropy": 5.334596252441406,
      "epoch": 3.1621915820029027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040148106996849636,
      "loss": 4.9711,
      "mean_token_accuracy": 0.20560424774885178,
      "num_tokens": 72332312.0,
      "step": 34860
    },
    {
      "entropy": 5.231681394577026,
      "epoch": 3.162645137880987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040145430849762993,
      "loss": 4.849,
      "mean_token_accuracy": 0.21421671360731126,
      "num_tokens": 72342157.0,
      "step": 34865
    },
    {
      "entropy": 5.336580419540406,
      "epoch": 3.1630986937590713,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040142754441165615,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2158343091607094,
      "num_tokens": 72353417.0,
      "step": 34870
    },
    {
      "entropy": 5.3428511142730715,
      "epoch": 3.1635522496371555,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004014007777111285,
      "loss": 4.8785,
      "mean_token_accuracy": 0.21811963468790055,
      "num_tokens": 72363744.0,
      "step": 34875
    },
    {
      "entropy": 5.300018405914306,
      "epoch": 3.1640058055152394,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040137400839660066,
      "loss": 4.951,
      "mean_token_accuracy": 0.21057909578084946,
      "num_tokens": 72375241.0,
      "step": 34880
    },
    {
      "entropy": 5.32844762802124,
      "epoch": 3.1644593613933236,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040134723646862607,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21187514513731004,
      "num_tokens": 72385475.0,
      "step": 34885
    },
    {
      "entropy": 5.3244706153869625,
      "epoch": 3.164912917271408,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040132046192775835,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20686640441417695,
      "num_tokens": 72396417.0,
      "step": 34890
    },
    {
      "entropy": 5.351113367080688,
      "epoch": 3.165366473149492,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004012936847745514,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21169223189353942,
      "num_tokens": 72406169.0,
      "step": 34895
    },
    {
      "entropy": 5.338552856445313,
      "epoch": 3.165820029027576,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004012669050095589,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20712537914514542,
      "num_tokens": 72416522.0,
      "step": 34900
    },
    {
      "entropy": 5.36074275970459,
      "epoch": 3.1662735849056602,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004012401226333345,
      "loss": 4.9586,
      "mean_token_accuracy": 0.21003906726837157,
      "num_tokens": 72426000.0,
      "step": 34905
    },
    {
      "entropy": 5.356430912017823,
      "epoch": 3.1667271407837445,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004012133376464322,
      "loss": 5.0016,
      "mean_token_accuracy": 0.20603262931108474,
      "num_tokens": 72435952.0,
      "step": 34910
    },
    {
      "entropy": 5.352180767059326,
      "epoch": 3.1671806966618288,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040118655004940597,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21070233285427092,
      "num_tokens": 72449957.0,
      "step": 34915
    },
    {
      "entropy": 5.353117227554321,
      "epoch": 3.167634252539913,
      "grad_norm": 1.0,
      "learning_rate": 0.0004011597598428097,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21236525177955629,
      "num_tokens": 72459956.0,
      "step": 34920
    },
    {
      "entropy": 5.2927769184112545,
      "epoch": 3.1680878084179973,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004011329670271975,
      "loss": 4.8532,
      "mean_token_accuracy": 0.21736321598291397,
      "num_tokens": 72471107.0,
      "step": 34925
    },
    {
      "entropy": 5.25229697227478,
      "epoch": 3.168541364296081,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040110617160312334,
      "loss": 4.8598,
      "mean_token_accuracy": 0.222350512444973,
      "num_tokens": 72481428.0,
      "step": 34930
    },
    {
      "entropy": 5.357256603240967,
      "epoch": 3.1689949201741654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004010793735711415,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20183736085891724,
      "num_tokens": 72491133.0,
      "step": 34935
    },
    {
      "entropy": 5.261280965805054,
      "epoch": 3.1694484760522497,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040105257293180605,
      "loss": 4.8363,
      "mean_token_accuracy": 0.2190167561173439,
      "num_tokens": 72502047.0,
      "step": 34940
    },
    {
      "entropy": 5.318171834945678,
      "epoch": 3.169902031930334,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004010257696856712,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2086109295487404,
      "num_tokens": 72513120.0,
      "step": 34945
    },
    {
      "entropy": 5.357873868942261,
      "epoch": 3.170355587808418,
      "grad_norm": 1.125,
      "learning_rate": 0.00040099896383329135,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2146776631474495,
      "num_tokens": 72523118.0,
      "step": 34950
    },
    {
      "entropy": 5.38518443107605,
      "epoch": 3.170809143686502,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040097215537522086,
      "loss": 5.0466,
      "mean_token_accuracy": 0.21111540496349335,
      "num_tokens": 72533697.0,
      "step": 34955
    },
    {
      "entropy": 5.319959878921509,
      "epoch": 3.1712626995645863,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040094534431201406,
      "loss": 4.8561,
      "mean_token_accuracy": 0.21973746567964553,
      "num_tokens": 72543661.0,
      "step": 34960
    },
    {
      "entropy": 5.334146738052368,
      "epoch": 3.1717162554426706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004009185306442255,
      "loss": 5.0072,
      "mean_token_accuracy": 0.20775910913944245,
      "num_tokens": 72554551.0,
      "step": 34965
    },
    {
      "entropy": 5.331138658523559,
      "epoch": 3.172169811320755,
      "grad_norm": 1.0,
      "learning_rate": 0.00040089171437240954,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20243676155805587,
      "num_tokens": 72565134.0,
      "step": 34970
    },
    {
      "entropy": 5.3115190982818605,
      "epoch": 3.172623367198839,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040086489549712097,
      "loss": 4.9205,
      "mean_token_accuracy": 0.21778281927108764,
      "num_tokens": 72574981.0,
      "step": 34975
    },
    {
      "entropy": 5.356239271163941,
      "epoch": 3.173076923076923,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004008380740189142,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21235227286815644,
      "num_tokens": 72585073.0,
      "step": 34980
    },
    {
      "entropy": 5.349426507949829,
      "epoch": 3.173530478955007,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000400811249938344,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21505210995674134,
      "num_tokens": 72595719.0,
      "step": 34985
    },
    {
      "entropy": 5.339492464065552,
      "epoch": 3.1739840348330914,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004007844232559652,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21098743081092836,
      "num_tokens": 72605276.0,
      "step": 34990
    },
    {
      "entropy": 5.368939208984375,
      "epoch": 3.1744375907111757,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004007575939723324,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20770449191331863,
      "num_tokens": 72614964.0,
      "step": 34995
    },
    {
      "entropy": 5.3612548828125,
      "epoch": 3.17489114658926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004007307620880004,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2118418961763382,
      "num_tokens": 72625461.0,
      "step": 35000
    },
    {
      "entropy": 5.304065799713134,
      "epoch": 3.175344702467344,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040070392760352424,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21295981705188752,
      "num_tokens": 72636192.0,
      "step": 35005
    },
    {
      "entropy": 5.326938438415527,
      "epoch": 3.175798258345428,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040067709051945895,
      "loss": 4.9742,
      "mean_token_accuracy": 0.20930993556976318,
      "num_tokens": 72648432.0,
      "step": 35010
    },
    {
      "entropy": 5.316373062133789,
      "epoch": 3.1762518142235123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004006502508363593,
      "loss": 4.884,
      "mean_token_accuracy": 0.21020959317684174,
      "num_tokens": 72658678.0,
      "step": 35015
    },
    {
      "entropy": 5.418822336196899,
      "epoch": 3.1767053701015966,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040062340855478046,
      "loss": 5.0961,
      "mean_token_accuracy": 0.19492300301790239,
      "num_tokens": 72669191.0,
      "step": 35020
    },
    {
      "entropy": 5.329587745666504,
      "epoch": 3.177158925979681,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040059656367527754,
      "loss": 4.8623,
      "mean_token_accuracy": 0.21416086554527283,
      "num_tokens": 72679327.0,
      "step": 35025
    },
    {
      "entropy": 5.315107250213623,
      "epoch": 3.1776124818577647,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004005697161984056,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20362989753484725,
      "num_tokens": 72689673.0,
      "step": 35030
    },
    {
      "entropy": 5.274407625198364,
      "epoch": 3.178066037735849,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040054286612472006,
      "loss": 4.9689,
      "mean_token_accuracy": 0.20006797462701797,
      "num_tokens": 72700606.0,
      "step": 35035
    },
    {
      "entropy": 5.388038778305054,
      "epoch": 3.178519593613933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040051601345477593,
      "loss": 4.9048,
      "mean_token_accuracy": 0.2165857419371605,
      "num_tokens": 72711195.0,
      "step": 35040
    },
    {
      "entropy": 5.334883499145508,
      "epoch": 3.1789731494920175,
      "grad_norm": 1.0,
      "learning_rate": 0.00040048915818912875,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2130228191614151,
      "num_tokens": 72721160.0,
      "step": 35045
    },
    {
      "entropy": 5.252718162536621,
      "epoch": 3.1794267053701017,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040046230032833374,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21008228957653047,
      "num_tokens": 72730867.0,
      "step": 35050
    },
    {
      "entropy": 5.282723665237427,
      "epoch": 3.1798802612481856,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004004354398729464,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21612727344036103,
      "num_tokens": 72740453.0,
      "step": 35055
    },
    {
      "entropy": 5.413291931152344,
      "epoch": 3.18033381712627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004004085768235222,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20462477952241898,
      "num_tokens": 72751395.0,
      "step": 35060
    },
    {
      "entropy": 5.351336574554443,
      "epoch": 3.180787373004354,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040038171118061673,
      "loss": 4.904,
      "mean_token_accuracy": 0.21483373194932937,
      "num_tokens": 72761300.0,
      "step": 35065
    },
    {
      "entropy": 5.262792921066284,
      "epoch": 3.1812409288824384,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004003548429447855,
      "loss": 4.8877,
      "mean_token_accuracy": 0.2215481400489807,
      "num_tokens": 72772011.0,
      "step": 35070
    },
    {
      "entropy": 5.3054252624511715,
      "epoch": 3.1816944847605226,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004003279721165841,
      "loss": 4.9346,
      "mean_token_accuracy": 0.20597139298915862,
      "num_tokens": 72782521.0,
      "step": 35075
    },
    {
      "entropy": 5.279986381530762,
      "epoch": 3.182148040638607,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004003010986965683,
      "loss": 4.8136,
      "mean_token_accuracy": 0.21831827610731125,
      "num_tokens": 72793125.0,
      "step": 35080
    },
    {
      "entropy": 5.366335535049439,
      "epoch": 3.1826015965166907,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000400274222685294,
      "loss": 4.966,
      "mean_token_accuracy": 0.21092172265052794,
      "num_tokens": 72804541.0,
      "step": 35085
    },
    {
      "entropy": 5.4593555450439455,
      "epoch": 3.183055152394775,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040024734408331664,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2104245901107788,
      "num_tokens": 72814447.0,
      "step": 35090
    },
    {
      "entropy": 5.342739582061768,
      "epoch": 3.1835087082728593,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040022046289119246,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2193668767809868,
      "num_tokens": 72823545.0,
      "step": 35095
    },
    {
      "entropy": 5.2658203125,
      "epoch": 3.1839622641509435,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040019357910947714,
      "loss": 4.8199,
      "mean_token_accuracy": 0.21497847586870195,
      "num_tokens": 72833803.0,
      "step": 35100
    },
    {
      "entropy": 5.3417370319366455,
      "epoch": 3.1844158200290273,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040016669273872666,
      "loss": 4.943,
      "mean_token_accuracy": 0.21218933612108232,
      "num_tokens": 72843383.0,
      "step": 35105
    },
    {
      "entropy": 5.356571817398072,
      "epoch": 3.1848693759071116,
      "grad_norm": 1.0,
      "learning_rate": 0.00040013980377949716,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20874018967151642,
      "num_tokens": 72853285.0,
      "step": 35110
    },
    {
      "entropy": 5.334869623184204,
      "epoch": 3.185322931785196,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040011291223234456,
      "loss": 4.9729,
      "mean_token_accuracy": 0.2100227490067482,
      "num_tokens": 72863556.0,
      "step": 35115
    },
    {
      "entropy": 5.328460788726806,
      "epoch": 3.18577648766328,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040008601809782507,
      "loss": 4.8552,
      "mean_token_accuracy": 0.2144877627491951,
      "num_tokens": 72873296.0,
      "step": 35120
    },
    {
      "entropy": 5.341270685195923,
      "epoch": 3.1862300435413644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004000591213764948,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2001987800002098,
      "num_tokens": 72884738.0,
      "step": 35125
    },
    {
      "entropy": 5.380631780624389,
      "epoch": 3.1866835994194487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040003222206891017,
      "loss": 4.9595,
      "mean_token_accuracy": 0.20488062500953674,
      "num_tokens": 72894573.0,
      "step": 35130
    },
    {
      "entropy": 5.270681715011596,
      "epoch": 3.1871371552975325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004000053201756272,
      "loss": 4.8819,
      "mean_token_accuracy": 0.21309182941913604,
      "num_tokens": 72904976.0,
      "step": 35135
    },
    {
      "entropy": 5.30842957496643,
      "epoch": 3.1875907111756168,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003999784156972024,
      "loss": 4.9291,
      "mean_token_accuracy": 0.20580876767635345,
      "num_tokens": 72915152.0,
      "step": 35140
    },
    {
      "entropy": 5.360043048858643,
      "epoch": 3.188044267053701,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003999515086341921,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2080184191465378,
      "num_tokens": 72925450.0,
      "step": 35145
    },
    {
      "entropy": 5.309264612197876,
      "epoch": 3.1884978229317853,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003999245989871527,
      "loss": 4.8659,
      "mean_token_accuracy": 0.21094968318939208,
      "num_tokens": 72935990.0,
      "step": 35150
    },
    {
      "entropy": 5.275072240829468,
      "epoch": 3.1889513788098696,
      "grad_norm": 1.25,
      "learning_rate": 0.00039989768675664084,
      "loss": 4.8028,
      "mean_token_accuracy": 0.21563132405281066,
      "num_tokens": 72945662.0,
      "step": 35155
    },
    {
      "entropy": 5.213641166687012,
      "epoch": 3.1894049346879534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039987077194321305,
      "loss": 4.7361,
      "mean_token_accuracy": 0.22245804518461226,
      "num_tokens": 72955313.0,
      "step": 35160
    },
    {
      "entropy": 5.348676586151123,
      "epoch": 3.1898584905660377,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039984385454742576,
      "loss": 5.0402,
      "mean_token_accuracy": 0.20266651064157487,
      "num_tokens": 72965347.0,
      "step": 35165
    },
    {
      "entropy": 5.311732578277588,
      "epoch": 3.190312046444122,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039981693456983574,
      "loss": 4.8442,
      "mean_token_accuracy": 0.21723738610744475,
      "num_tokens": 72976794.0,
      "step": 35170
    },
    {
      "entropy": 5.2801947593688965,
      "epoch": 3.190765602322206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003997900120109998,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21509758830070497,
      "num_tokens": 72986837.0,
      "step": 35175
    },
    {
      "entropy": 5.308062934875489,
      "epoch": 3.1912191582002905,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003997630868714745,
      "loss": 4.8895,
      "mean_token_accuracy": 0.21268744319677352,
      "num_tokens": 72996819.0,
      "step": 35180
    },
    {
      "entropy": 5.252859354019165,
      "epoch": 3.1916727140783743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039973615915181687,
      "loss": 4.7841,
      "mean_token_accuracy": 0.22289938628673553,
      "num_tokens": 73005963.0,
      "step": 35185
    },
    {
      "entropy": 5.287258291244507,
      "epoch": 3.1921262699564585,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039970922885258363,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2064031779766083,
      "num_tokens": 73015912.0,
      "step": 35190
    },
    {
      "entropy": 5.2830119132995605,
      "epoch": 3.192579825834543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003996822959743318,
      "loss": 4.8206,
      "mean_token_accuracy": 0.22417337596416473,
      "num_tokens": 73026528.0,
      "step": 35195
    },
    {
      "entropy": 5.323281478881836,
      "epoch": 3.193033381712627,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003996553605176183,
      "loss": 4.9567,
      "mean_token_accuracy": 0.20249081701040267,
      "num_tokens": 73036255.0,
      "step": 35200
    },
    {
      "entropy": 5.301980447769165,
      "epoch": 3.1934869375907113,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003996284224830002,
      "loss": 5.0383,
      "mean_token_accuracy": 0.20897114127874375,
      "num_tokens": 73046686.0,
      "step": 35205
    },
    {
      "entropy": 5.2988037109375,
      "epoch": 3.193940493468795,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039960148187103453,
      "loss": 4.8382,
      "mean_token_accuracy": 0.21798473596572876,
      "num_tokens": 73056070.0,
      "step": 35210
    },
    {
      "entropy": 5.38137583732605,
      "epoch": 3.1943940493468794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039957453868227845,
      "loss": 4.952,
      "mean_token_accuracy": 0.2168365642428398,
      "num_tokens": 73066288.0,
      "step": 35215
    },
    {
      "entropy": 5.305613040924072,
      "epoch": 3.1948476052249637,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039954759291728915,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21141639202833176,
      "num_tokens": 73076297.0,
      "step": 35220
    },
    {
      "entropy": 5.309349822998047,
      "epoch": 3.195301161103048,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039952064457662395,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21016620844602585,
      "num_tokens": 73086249.0,
      "step": 35225
    },
    {
      "entropy": 5.294515800476074,
      "epoch": 3.1957547169811322,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039949369366084,
      "loss": 4.8749,
      "mean_token_accuracy": 0.2172505959868431,
      "num_tokens": 73096358.0,
      "step": 35230
    },
    {
      "entropy": 5.268304491043091,
      "epoch": 3.196208272859216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039946674017049484,
      "loss": 4.8528,
      "mean_token_accuracy": 0.20997247099876404,
      "num_tokens": 73106891.0,
      "step": 35235
    },
    {
      "entropy": 5.223990154266358,
      "epoch": 3.1966618287373003,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039943978410614565,
      "loss": 4.8715,
      "mean_token_accuracy": 0.21123632341623305,
      "num_tokens": 73117106.0,
      "step": 35240
    },
    {
      "entropy": 5.290913438796997,
      "epoch": 3.1971153846153846,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039941282546835017,
      "loss": 4.9322,
      "mean_token_accuracy": 0.2104165866971016,
      "num_tokens": 73127112.0,
      "step": 35245
    },
    {
      "entropy": 5.314920377731323,
      "epoch": 3.197568940493469,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039938586425766564,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21020931750535965,
      "num_tokens": 73137751.0,
      "step": 35250
    },
    {
      "entropy": 5.346518659591675,
      "epoch": 3.198022496371553,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003993589004746498,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2151741236448288,
      "num_tokens": 73149303.0,
      "step": 35255
    },
    {
      "entropy": 5.26245813369751,
      "epoch": 3.198476052249637,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039933193411986024,
      "loss": 4.8904,
      "mean_token_accuracy": 0.21251840740442277,
      "num_tokens": 73160025.0,
      "step": 35260
    },
    {
      "entropy": 5.305142641067505,
      "epoch": 3.198929608127721,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039930496519385454,
      "loss": 4.8877,
      "mean_token_accuracy": 0.21790559142827987,
      "num_tokens": 73170721.0,
      "step": 35265
    },
    {
      "entropy": 5.335173416137695,
      "epoch": 3.1993831640058055,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039927799369719047,
      "loss": 4.9337,
      "mean_token_accuracy": 0.20681352019309998,
      "num_tokens": 73180902.0,
      "step": 35270
    },
    {
      "entropy": 5.184840726852417,
      "epoch": 3.1998367198838897,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003992510196304258,
      "loss": 4.7739,
      "mean_token_accuracy": 0.219899919629097,
      "num_tokens": 73191259.0,
      "step": 35275
    },
    {
      "entropy": 5.300670146942139,
      "epoch": 3.200290275761974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039922404299411846,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2218126580119133,
      "num_tokens": 73201654.0,
      "step": 35280
    },
    {
      "entropy": 5.32168288230896,
      "epoch": 3.2007438316400583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003991970637888262,
      "loss": 4.9306,
      "mean_token_accuracy": 0.20303671061992645,
      "num_tokens": 73212658.0,
      "step": 35285
    },
    {
      "entropy": 5.370536804199219,
      "epoch": 3.201197387518142,
      "grad_norm": 1.046875,
      "learning_rate": 0.000399170082015107,
      "loss": 4.937,
      "mean_token_accuracy": 0.20677734762430192,
      "num_tokens": 73222883.0,
      "step": 35290
    },
    {
      "entropy": 5.299330806732177,
      "epoch": 3.2016509433962264,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000399143097673519,
      "loss": 4.8632,
      "mean_token_accuracy": 0.21483744978904723,
      "num_tokens": 73233731.0,
      "step": 35295
    },
    {
      "entropy": 5.274930620193482,
      "epoch": 3.2021044992743106,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039911611076461995,
      "loss": 4.8306,
      "mean_token_accuracy": 0.2216074749827385,
      "num_tokens": 73243794.0,
      "step": 35300
    },
    {
      "entropy": 5.279996156692505,
      "epoch": 3.202558055152395,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039908912128896817,
      "loss": 4.8791,
      "mean_token_accuracy": 0.20272012501955033,
      "num_tokens": 73254708.0,
      "step": 35305
    },
    {
      "entropy": 5.25786509513855,
      "epoch": 3.203011611030479,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003990621292471217,
      "loss": 4.8759,
      "mean_token_accuracy": 0.20942484587430954,
      "num_tokens": 73265311.0,
      "step": 35310
    },
    {
      "entropy": 5.417667865753174,
      "epoch": 3.203465166908563,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039903513463963874,
      "loss": 5.0738,
      "mean_token_accuracy": 0.20525032430887222,
      "num_tokens": 73275886.0,
      "step": 35315
    },
    {
      "entropy": 5.390987682342529,
      "epoch": 3.2039187227866472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003990081374670777,
      "loss": 4.945,
      "mean_token_accuracy": 0.21305500566959382,
      "num_tokens": 73285903.0,
      "step": 35320
    },
    {
      "entropy": 5.27974681854248,
      "epoch": 3.2043722786647315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039898113772999665,
      "loss": 4.8931,
      "mean_token_accuracy": 0.21555203050374985,
      "num_tokens": 73296571.0,
      "step": 35325
    },
    {
      "entropy": 5.308307218551636,
      "epoch": 3.2048258345428158,
      "grad_norm": 1.0,
      "learning_rate": 0.00039895413542895427,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20329193621873856,
      "num_tokens": 73307105.0,
      "step": 35330
    },
    {
      "entropy": 5.2608240127563475,
      "epoch": 3.2052793904209,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003989271305645086,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2172118380665779,
      "num_tokens": 73316664.0,
      "step": 35335
    },
    {
      "entropy": 5.292282295227051,
      "epoch": 3.205732946298984,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039890012313721835,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21577272862195968,
      "num_tokens": 73327438.0,
      "step": 35340
    },
    {
      "entropy": 5.315598630905152,
      "epoch": 3.206186502177068,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000398873113147642,
      "loss": 4.921,
      "mean_token_accuracy": 0.21555407345294952,
      "num_tokens": 73337635.0,
      "step": 35345
    },
    {
      "entropy": 5.2754240989685055,
      "epoch": 3.2066400580551524,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039884610059633807,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21736757606267929,
      "num_tokens": 73347032.0,
      "step": 35350
    },
    {
      "entropy": 5.369875288009643,
      "epoch": 3.2070936139332367,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003988190854838652,
      "loss": 4.9908,
      "mean_token_accuracy": 0.20806473791599273,
      "num_tokens": 73357780.0,
      "step": 35355
    },
    {
      "entropy": 5.355873250961304,
      "epoch": 3.207547169811321,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039879206781078214,
      "loss": 4.9487,
      "mean_token_accuracy": 0.21119759529829024,
      "num_tokens": 73367831.0,
      "step": 35360
    },
    {
      "entropy": 5.333029127120971,
      "epoch": 3.2080007256894048,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003987650475776476,
      "loss": 4.9093,
      "mean_token_accuracy": 0.21198583245277405,
      "num_tokens": 73378588.0,
      "step": 35365
    },
    {
      "entropy": 5.269574737548828,
      "epoch": 3.208454281567489,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039873802478502027,
      "loss": 4.8949,
      "mean_token_accuracy": 0.21539209634065629,
      "num_tokens": 73388414.0,
      "step": 35370
    },
    {
      "entropy": 5.28154444694519,
      "epoch": 3.2089078374455733,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003987109994334591,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21122357845306397,
      "num_tokens": 73398891.0,
      "step": 35375
    },
    {
      "entropy": 5.295464706420899,
      "epoch": 3.2093613933236576,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003986839715235229,
      "loss": 4.9455,
      "mean_token_accuracy": 0.20879193544387817,
      "num_tokens": 73409605.0,
      "step": 35380
    },
    {
      "entropy": 5.3677895069122314,
      "epoch": 3.209814949201742,
      "grad_norm": 1.125,
      "learning_rate": 0.00039865694105577063,
      "loss": 4.9833,
      "mean_token_accuracy": 0.20525723099708557,
      "num_tokens": 73419103.0,
      "step": 35385
    },
    {
      "entropy": 5.246015453338623,
      "epoch": 3.2102685050798256,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003986299080307614,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21601397395133973,
      "num_tokens": 73429134.0,
      "step": 35390
    },
    {
      "entropy": 5.358482789993286,
      "epoch": 3.21072206095791,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003986028724490541,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20782105922698973,
      "num_tokens": 73440382.0,
      "step": 35395
    },
    {
      "entropy": 5.346194791793823,
      "epoch": 3.211175616835994,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003985758343112079,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2166496753692627,
      "num_tokens": 73451300.0,
      "step": 35400
    },
    {
      "entropy": 5.242894649505615,
      "epoch": 3.2116291727140784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039854879361778193,
      "loss": 4.8102,
      "mean_token_accuracy": 0.2172458991408348,
      "num_tokens": 73461339.0,
      "step": 35405
    },
    {
      "entropy": 5.2782573223114015,
      "epoch": 3.2120827285921627,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039852175036933545,
      "loss": 4.9106,
      "mean_token_accuracy": 0.21642778664827347,
      "num_tokens": 73470961.0,
      "step": 35410
    },
    {
      "entropy": 5.373208284378052,
      "epoch": 3.2125362844702465,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039849470456642777,
      "loss": 4.999,
      "mean_token_accuracy": 0.20344596058130265,
      "num_tokens": 73481492.0,
      "step": 35415
    },
    {
      "entropy": 5.285233688354492,
      "epoch": 3.212989840348331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039846765620961806,
      "loss": 4.8522,
      "mean_token_accuracy": 0.2133144721388817,
      "num_tokens": 73492351.0,
      "step": 35420
    },
    {
      "entropy": 5.349156045913697,
      "epoch": 3.213443396226415,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003984406052994658,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20852123945951462,
      "num_tokens": 73503265.0,
      "step": 35425
    },
    {
      "entropy": 5.334804630279541,
      "epoch": 3.2138969521044993,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003984135518365303,
      "loss": 4.9342,
      "mean_token_accuracy": 0.20437347888946533,
      "num_tokens": 73513804.0,
      "step": 35430
    },
    {
      "entropy": 5.3465433597564695,
      "epoch": 3.2143505079825836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039838649582137113,
      "loss": 4.9524,
      "mean_token_accuracy": 0.21072442680597306,
      "num_tokens": 73523453.0,
      "step": 35435
    },
    {
      "entropy": 5.403426647186279,
      "epoch": 3.214804063860668,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039835943725454777,
      "loss": 5.0384,
      "mean_token_accuracy": 0.20116156339645386,
      "num_tokens": 73533931.0,
      "step": 35440
    },
    {
      "entropy": 5.385284328460694,
      "epoch": 3.2152576197387517,
      "grad_norm": 1.0,
      "learning_rate": 0.00039833237613661984,
      "loss": 4.9738,
      "mean_token_accuracy": 0.20942142158746718,
      "num_tokens": 73544562.0,
      "step": 35445
    },
    {
      "entropy": 5.3122742652893065,
      "epoch": 3.215711175616836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003983053124681469,
      "loss": 4.9153,
      "mean_token_accuracy": 0.2178727939724922,
      "num_tokens": 73555149.0,
      "step": 35450
    },
    {
      "entropy": 5.236322593688965,
      "epoch": 3.21616473149492,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039827824624968873,
      "loss": 4.8376,
      "mean_token_accuracy": 0.2108605369925499,
      "num_tokens": 73565131.0,
      "step": 35455
    },
    {
      "entropy": 5.295153856277466,
      "epoch": 3.2166182873730045,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039825117748180493,
      "loss": 4.8852,
      "mean_token_accuracy": 0.21318047493696213,
      "num_tokens": 73574185.0,
      "step": 35460
    },
    {
      "entropy": 5.356266355514526,
      "epoch": 3.2170718432510883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039822410616505545,
      "loss": 4.9997,
      "mean_token_accuracy": 0.20503681451082229,
      "num_tokens": 73585733.0,
      "step": 35465
    },
    {
      "entropy": 5.392999649047852,
      "epoch": 3.2175253991291726,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039819703230000006,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21064327657222748,
      "num_tokens": 73597077.0,
      "step": 35470
    },
    {
      "entropy": 5.288382053375244,
      "epoch": 3.217978955007257,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003981699558871986,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21104532480239868,
      "num_tokens": 73607885.0,
      "step": 35475
    },
    {
      "entropy": 5.3407745361328125,
      "epoch": 3.218432510885341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000398142876927211,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21061483025550842,
      "num_tokens": 73618769.0,
      "step": 35480
    },
    {
      "entropy": 5.373265552520752,
      "epoch": 3.2188860667634254,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039811579542059744,
      "loss": 5.0115,
      "mean_token_accuracy": 0.20298121571540834,
      "num_tokens": 73629345.0,
      "step": 35485
    },
    {
      "entropy": 5.305987453460693,
      "epoch": 3.2193396226415096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039808871136791786,
      "loss": 4.8798,
      "mean_token_accuracy": 0.21472629457712172,
      "num_tokens": 73639698.0,
      "step": 35490
    },
    {
      "entropy": 5.2571244716644285,
      "epoch": 3.2197931785195935,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039806162476973227,
      "loss": 4.9003,
      "mean_token_accuracy": 0.21138675808906554,
      "num_tokens": 73650954.0,
      "step": 35495
    },
    {
      "entropy": 5.346823740005493,
      "epoch": 3.2202467343976777,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000398034535626601,
      "loss": 5.0156,
      "mean_token_accuracy": 0.20288006365299224,
      "num_tokens": 73661195.0,
      "step": 35500
    },
    {
      "entropy": 5.432377290725708,
      "epoch": 3.220700290275762,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039800744393908414,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20584979504346848,
      "num_tokens": 73670512.0,
      "step": 35505
    },
    {
      "entropy": 5.373293113708496,
      "epoch": 3.2211538461538463,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000397980349707742,
      "loss": 4.962,
      "mean_token_accuracy": 0.20632748454809188,
      "num_tokens": 73681927.0,
      "step": 35510
    },
    {
      "entropy": 5.374514865875244,
      "epoch": 3.2216074020319305,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039795325293313493,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20488712042570115,
      "num_tokens": 73692025.0,
      "step": 35515
    },
    {
      "entropy": 5.300686168670654,
      "epoch": 3.2220609579100143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039792615361582324,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2076975464820862,
      "num_tokens": 73702278.0,
      "step": 35520
    },
    {
      "entropy": 5.410167980194092,
      "epoch": 3.2225145137880986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039789905175636726,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20016473084688186,
      "num_tokens": 73713511.0,
      "step": 35525
    },
    {
      "entropy": 5.2529504776000975,
      "epoch": 3.222968069666183,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039787194735532765,
      "loss": 4.819,
      "mean_token_accuracy": 0.22111839503049852,
      "num_tokens": 73723683.0,
      "step": 35530
    },
    {
      "entropy": 5.306318473815918,
      "epoch": 3.223421625544267,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003978448404132649,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2169344812631607,
      "num_tokens": 73734191.0,
      "step": 35535
    },
    {
      "entropy": 5.300194549560547,
      "epoch": 3.2238751814223514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039781773093073956,
      "loss": 4.934,
      "mean_token_accuracy": 0.20787128061056137,
      "num_tokens": 73743747.0,
      "step": 35540
    },
    {
      "entropy": 5.301626205444336,
      "epoch": 3.2243287373004352,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003977906189083122,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21311426162719727,
      "num_tokens": 73754466.0,
      "step": 35545
    },
    {
      "entropy": 5.3463890075683596,
      "epoch": 3.2247822931785195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003977635043465435,
      "loss": 4.8837,
      "mean_token_accuracy": 0.21561094224452973,
      "num_tokens": 73765948.0,
      "step": 35550
    },
    {
      "entropy": 5.222392225265503,
      "epoch": 3.2252358490566038,
      "grad_norm": 1.0,
      "learning_rate": 0.0003977363872459943,
      "loss": 4.857,
      "mean_token_accuracy": 0.21234640777111052,
      "num_tokens": 73776003.0,
      "step": 35555
    },
    {
      "entropy": 5.354861736297607,
      "epoch": 3.225689404934688,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003977092676072254,
      "loss": 5.0852,
      "mean_token_accuracy": 0.2052811414003372,
      "num_tokens": 73787821.0,
      "step": 35560
    },
    {
      "entropy": 5.334349060058594,
      "epoch": 3.2261429608127723,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039768214543079746,
      "loss": 4.922,
      "mean_token_accuracy": 0.20764459669589996,
      "num_tokens": 73799207.0,
      "step": 35565
    },
    {
      "entropy": 5.342871475219726,
      "epoch": 3.226596516690856,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003976550207172716,
      "loss": 4.8704,
      "mean_token_accuracy": 0.21767805963754655,
      "num_tokens": 73809987.0,
      "step": 35570
    },
    {
      "entropy": 5.264537715911866,
      "epoch": 3.2270500725689404,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039762789346720854,
      "loss": 4.8718,
      "mean_token_accuracy": 0.20877341330051422,
      "num_tokens": 73819075.0,
      "step": 35575
    },
    {
      "entropy": 5.330611801147461,
      "epoch": 3.2275036284470247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003976007636811694,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2186291992664337,
      "num_tokens": 73829831.0,
      "step": 35580
    },
    {
      "entropy": 5.3404154777526855,
      "epoch": 3.227957184325109,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003975736313597153,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21637640595436097,
      "num_tokens": 73839826.0,
      "step": 35585
    },
    {
      "entropy": 5.295796489715576,
      "epoch": 3.228410740203193,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003975464965034072,
      "loss": 4.8465,
      "mean_token_accuracy": 0.21565226316452027,
      "num_tokens": 73850006.0,
      "step": 35590
    },
    {
      "entropy": 5.313666343688965,
      "epoch": 3.228864296081277,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003975193591128064,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21258568465709687,
      "num_tokens": 73860309.0,
      "step": 35595
    },
    {
      "entropy": 5.3189167976379395,
      "epoch": 3.2293178519593613,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039749221918847406,
      "loss": 5.0132,
      "mean_token_accuracy": 0.19955051839351653,
      "num_tokens": 73871613.0,
      "step": 35600
    },
    {
      "entropy": 5.22365517616272,
      "epoch": 3.2297714078374455,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003974650767309713,
      "loss": 4.8419,
      "mean_token_accuracy": 0.2173735350370407,
      "num_tokens": 73882303.0,
      "step": 35605
    },
    {
      "entropy": 5.371206331253052,
      "epoch": 3.23022496371553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039743793174085965,
      "loss": 4.9595,
      "mean_token_accuracy": 0.20349878519773484,
      "num_tokens": 73893548.0,
      "step": 35610
    },
    {
      "entropy": 5.363835382461548,
      "epoch": 3.230678519593614,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039741078421870024,
      "loss": 4.9133,
      "mean_token_accuracy": 0.20813069492578506,
      "num_tokens": 73903934.0,
      "step": 35615
    },
    {
      "entropy": 5.355837440490722,
      "epoch": 3.231132075471698,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039738363416505474,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20468242913484574,
      "num_tokens": 73913813.0,
      "step": 35620
    },
    {
      "entropy": 5.350599145889282,
      "epoch": 3.231585631349782,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039735648158048444,
      "loss": 4.967,
      "mean_token_accuracy": 0.206342676281929,
      "num_tokens": 73924339.0,
      "step": 35625
    },
    {
      "entropy": 5.309059381484985,
      "epoch": 3.2320391872278664,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003973293264655509,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2070791244506836,
      "num_tokens": 73935318.0,
      "step": 35630
    },
    {
      "entropy": 5.30779275894165,
      "epoch": 3.2324927431059507,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003973021688208158,
      "loss": 4.8806,
      "mean_token_accuracy": 0.2144397586584091,
      "num_tokens": 73945547.0,
      "step": 35635
    },
    {
      "entropy": 5.329086542129517,
      "epoch": 3.232946298984035,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003972750086468406,
      "loss": 4.9051,
      "mean_token_accuracy": 0.20819886773824692,
      "num_tokens": 73956195.0,
      "step": 35640
    },
    {
      "entropy": 5.302552700042725,
      "epoch": 3.2333998548621192,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039724784594418714,
      "loss": 4.8471,
      "mean_token_accuracy": 0.22272312194108962,
      "num_tokens": 73966843.0,
      "step": 35645
    },
    {
      "entropy": 5.295849895477295,
      "epoch": 3.233853410740203,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000397220680713417,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21279617846012117,
      "num_tokens": 73978516.0,
      "step": 35650
    },
    {
      "entropy": 5.2434937953948975,
      "epoch": 3.2343069666182873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039719351295509203,
      "loss": 4.8565,
      "mean_token_accuracy": 0.21988393664360045,
      "num_tokens": 73988376.0,
      "step": 35655
    },
    {
      "entropy": 5.308851861953736,
      "epoch": 3.2347605224963716,
      "grad_norm": 1.0,
      "learning_rate": 0.00039716634266977413,
      "loss": 4.8755,
      "mean_token_accuracy": 0.21757373511791228,
      "num_tokens": 73998232.0,
      "step": 35660
    },
    {
      "entropy": 5.281655979156494,
      "epoch": 3.235214078374456,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003971391698580252,
      "loss": 4.9021,
      "mean_token_accuracy": 0.21138708293437958,
      "num_tokens": 74008726.0,
      "step": 35665
    },
    {
      "entropy": 5.282895421981811,
      "epoch": 3.23566763425254,
      "grad_norm": 1.0,
      "learning_rate": 0.00039711199452040696,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21503802686929702,
      "num_tokens": 74018746.0,
      "step": 35670
    },
    {
      "entropy": 5.271306848526001,
      "epoch": 3.236121190130624,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039708481665748157,
      "loss": 4.8719,
      "mean_token_accuracy": 0.21611329764127732,
      "num_tokens": 74028615.0,
      "step": 35675
    },
    {
      "entropy": 5.305068063735962,
      "epoch": 3.236574746008708,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003970576362698111,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2090702623128891,
      "num_tokens": 74039219.0,
      "step": 35680
    },
    {
      "entropy": 5.336165714263916,
      "epoch": 3.2370283018867925,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039703045335795766,
      "loss": 4.892,
      "mean_token_accuracy": 0.21111079156398774,
      "num_tokens": 74049385.0,
      "step": 35685
    },
    {
      "entropy": 5.375534296035767,
      "epoch": 3.2374818577648767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003970032679224833,
      "loss": 5.0601,
      "mean_token_accuracy": 0.19503388404846192,
      "num_tokens": 74059437.0,
      "step": 35690
    },
    {
      "entropy": 5.371767902374268,
      "epoch": 3.237935413642961,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039697607996395023,
      "loss": 4.9745,
      "mean_token_accuracy": 0.20288293510675431,
      "num_tokens": 74068834.0,
      "step": 35695
    },
    {
      "entropy": 5.364496088027954,
      "epoch": 3.238388969521045,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003969488894829208,
      "loss": 4.9493,
      "mean_token_accuracy": 0.20485356599092483,
      "num_tokens": 74079757.0,
      "step": 35700
    },
    {
      "entropy": 5.328310203552246,
      "epoch": 3.238842525399129,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003969216964799573,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21514159440994263,
      "num_tokens": 74091092.0,
      "step": 35705
    },
    {
      "entropy": 5.278910827636719,
      "epoch": 3.2392960812772134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003968945009556219,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2136056214570999,
      "num_tokens": 74102272.0,
      "step": 35710
    },
    {
      "entropy": 5.268464994430542,
      "epoch": 3.2397496371552976,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003968673029104773,
      "loss": 4.8798,
      "mean_token_accuracy": 0.21763618886470795,
      "num_tokens": 74113101.0,
      "step": 35715
    },
    {
      "entropy": 5.356098890304565,
      "epoch": 3.240203193033382,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003968401023450858,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21519833505153657,
      "num_tokens": 74123180.0,
      "step": 35720
    },
    {
      "entropy": 5.27186393737793,
      "epoch": 3.2406567489114657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039681289926000985,
      "loss": 4.8892,
      "mean_token_accuracy": 0.21323482245206832,
      "num_tokens": 74133675.0,
      "step": 35725
    },
    {
      "entropy": 5.208748245239258,
      "epoch": 3.24111030478955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003967856936558122,
      "loss": 4.8434,
      "mean_token_accuracy": 0.21340498328208923,
      "num_tokens": 74143472.0,
      "step": 35730
    },
    {
      "entropy": 5.324640464782715,
      "epoch": 3.2415638606676342,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039675848553305526,
      "loss": 4.8824,
      "mean_token_accuracy": 0.20890802145004272,
      "num_tokens": 74154682.0,
      "step": 35735
    },
    {
      "entropy": 5.334448528289795,
      "epoch": 3.2420174165457185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003967312748923019,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21251148730516434,
      "num_tokens": 74164745.0,
      "step": 35740
    },
    {
      "entropy": 5.232695484161377,
      "epoch": 3.2424709724238028,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039670406173411473,
      "loss": 4.824,
      "mean_token_accuracy": 0.21831464767456055,
      "num_tokens": 74175388.0,
      "step": 35745
    },
    {
      "entropy": 5.273793029785156,
      "epoch": 3.2429245283018866,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003966768460590566,
      "loss": 4.8633,
      "mean_token_accuracy": 0.21310745030641556,
      "num_tokens": 74185929.0,
      "step": 35750
    },
    {
      "entropy": 5.303574657440185,
      "epoch": 3.243378084179971,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039664962786769013,
      "loss": 4.879,
      "mean_token_accuracy": 0.21542396247386933,
      "num_tokens": 74196165.0,
      "step": 35755
    },
    {
      "entropy": 5.30879054069519,
      "epoch": 3.243831640058055,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039662240716057854,
      "loss": 4.886,
      "mean_token_accuracy": 0.21513912826776505,
      "num_tokens": 74206843.0,
      "step": 35760
    },
    {
      "entropy": 5.295144414901733,
      "epoch": 3.2442851959361394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003965951839382845,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21232873052358628,
      "num_tokens": 74216854.0,
      "step": 35765
    },
    {
      "entropy": 5.238753938674927,
      "epoch": 3.2447387518142237,
      "grad_norm": 0.984375,
      "learning_rate": 0.000396567958201371,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2117325931787491,
      "num_tokens": 74226158.0,
      "step": 35770
    },
    {
      "entropy": 5.295272159576416,
      "epoch": 3.2451923076923075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003965407299504013,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2111606240272522,
      "num_tokens": 74236860.0,
      "step": 35775
    },
    {
      "entropy": 5.3288352489471436,
      "epoch": 3.2456458635703918,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003965134991859382,
      "loss": 4.8955,
      "mean_token_accuracy": 0.2158100500702858,
      "num_tokens": 74247706.0,
      "step": 35780
    },
    {
      "entropy": 5.361194610595703,
      "epoch": 3.246099419448476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000396486265908545,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21005010157823562,
      "num_tokens": 74259479.0,
      "step": 35785
    },
    {
      "entropy": 5.281999301910401,
      "epoch": 3.2465529753265603,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003964590301187849,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21290305405855178,
      "num_tokens": 74270371.0,
      "step": 35790
    },
    {
      "entropy": 5.280937814712525,
      "epoch": 3.2470065312046446,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003964317918172211,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22170497328042985,
      "num_tokens": 74279637.0,
      "step": 35795
    },
    {
      "entropy": 5.331215190887451,
      "epoch": 3.247460087082729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039640455100441684,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2078032150864601,
      "num_tokens": 74289889.0,
      "step": 35800
    },
    {
      "entropy": 5.298671007156372,
      "epoch": 3.2479136429608126,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039637730768093564,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21866786926984788,
      "num_tokens": 74300949.0,
      "step": 35805
    },
    {
      "entropy": 5.310999774932862,
      "epoch": 3.248367198838897,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003963500618473407,
      "loss": 4.8674,
      "mean_token_accuracy": 0.21987191438674927,
      "num_tokens": 74312077.0,
      "step": 35810
    },
    {
      "entropy": 5.302640390396118,
      "epoch": 3.248820754716981,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003963228135041956,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20796080082654952,
      "num_tokens": 74322610.0,
      "step": 35815
    },
    {
      "entropy": 5.250725984573364,
      "epoch": 3.2492743105950654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039629556265206373,
      "loss": 4.8043,
      "mean_token_accuracy": 0.2146112635731697,
      "num_tokens": 74331970.0,
      "step": 35820
    },
    {
      "entropy": 5.294125366210937,
      "epoch": 3.2497278664731493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003962683092915088,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22088005244731904,
      "num_tokens": 74342812.0,
      "step": 35825
    },
    {
      "entropy": 5.331630659103394,
      "epoch": 3.2501814223512335,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003962410534230943,
      "loss": 4.8346,
      "mean_token_accuracy": 0.21918106079101562,
      "num_tokens": 74353786.0,
      "step": 35830
    },
    {
      "entropy": 5.336825656890869,
      "epoch": 3.250634978229318,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039621379504738385,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21341964304447175,
      "num_tokens": 74363732.0,
      "step": 35835
    },
    {
      "entropy": 5.315958786010742,
      "epoch": 3.251088534107402,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039618653416494137,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2091994985938072,
      "num_tokens": 74374077.0,
      "step": 35840
    },
    {
      "entropy": 5.307712984085083,
      "epoch": 3.2515420899854863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039615927077633043,
      "loss": 4.9579,
      "mean_token_accuracy": 0.20807545483112336,
      "num_tokens": 74384758.0,
      "step": 35845
    },
    {
      "entropy": 5.405876016616821,
      "epoch": 3.2519956458635706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039613200488211483,
      "loss": 4.9666,
      "mean_token_accuracy": 0.20926786959171295,
      "num_tokens": 74395811.0,
      "step": 35850
    },
    {
      "entropy": 5.338453388214111,
      "epoch": 3.2524492017416544,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039610473648285857,
      "loss": 4.9533,
      "mean_token_accuracy": 0.2076203316450119,
      "num_tokens": 74407830.0,
      "step": 35855
    },
    {
      "entropy": 5.257277154922486,
      "epoch": 3.2529027576197387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003960774655791255,
      "loss": 4.8683,
      "mean_token_accuracy": 0.211029452085495,
      "num_tokens": 74417565.0,
      "step": 35860
    },
    {
      "entropy": 5.36267352104187,
      "epoch": 3.253356313497823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039605019217147947,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2116367131471634,
      "num_tokens": 74428015.0,
      "step": 35865
    },
    {
      "entropy": 5.268441534042358,
      "epoch": 3.253809869375907,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039602291626048475,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21670102030038835,
      "num_tokens": 74439523.0,
      "step": 35870
    },
    {
      "entropy": 5.365372991561889,
      "epoch": 3.2542634252539915,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039599563784670526,
      "loss": 5.0052,
      "mean_token_accuracy": 0.2003256633877754,
      "num_tokens": 74449897.0,
      "step": 35875
    },
    {
      "entropy": 5.281784963607788,
      "epoch": 3.2547169811320753,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003959683569307051,
      "loss": 4.8257,
      "mean_token_accuracy": 0.22016068249940873,
      "num_tokens": 74459888.0,
      "step": 35880
    },
    {
      "entropy": 5.2460307598114015,
      "epoch": 3.2551705370101596,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003959410735130485,
      "loss": 4.7965,
      "mean_token_accuracy": 0.21237492859363555,
      "num_tokens": 74470894.0,
      "step": 35885
    },
    {
      "entropy": 5.275335741043091,
      "epoch": 3.255624092888244,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039591378759429974,
      "loss": 4.821,
      "mean_token_accuracy": 0.22151262760162355,
      "num_tokens": 74481306.0,
      "step": 35890
    },
    {
      "entropy": 5.29795093536377,
      "epoch": 3.256077648766328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000395886499175023,
      "loss": 4.8985,
      "mean_token_accuracy": 0.21219355016946792,
      "num_tokens": 74491048.0,
      "step": 35895
    },
    {
      "entropy": 5.225656032562256,
      "epoch": 3.2565312046444124,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00039585920825578266,
      "loss": 4.8321,
      "mean_token_accuracy": 0.2230220079421997,
      "num_tokens": 74501656.0,
      "step": 35900
    },
    {
      "entropy": 5.329667949676514,
      "epoch": 3.256984760522496,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003958319148371431,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21127210706472396,
      "num_tokens": 74513280.0,
      "step": 35905
    },
    {
      "entropy": 5.289567375183106,
      "epoch": 3.2574383164005805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003958046189196687,
      "loss": 4.9313,
      "mean_token_accuracy": 0.21268393844366074,
      "num_tokens": 74522898.0,
      "step": 35910
    },
    {
      "entropy": 5.400722980499268,
      "epoch": 3.2578918722786647,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039577732050392395,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2102117583155632,
      "num_tokens": 74532551.0,
      "step": 35915
    },
    {
      "entropy": 5.323628759384155,
      "epoch": 3.258345428156749,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003957500195904735,
      "loss": 4.9119,
      "mean_token_accuracy": 0.21967433094978334,
      "num_tokens": 74542285.0,
      "step": 35920
    },
    {
      "entropy": 5.333619117736816,
      "epoch": 3.2587989840348333,
      "grad_norm": 1.078125,
      "learning_rate": 0.000395722716179882,
      "loss": 4.9595,
      "mean_token_accuracy": 0.2149643197655678,
      "num_tokens": 74551982.0,
      "step": 35925
    },
    {
      "entropy": 5.338657426834106,
      "epoch": 3.259252539912917,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003956954102727138,
      "loss": 4.9154,
      "mean_token_accuracy": 0.21350584477186202,
      "num_tokens": 74562190.0,
      "step": 35930
    },
    {
      "entropy": 5.2751349925994875,
      "epoch": 3.2597060957910013,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039566810186953376,
      "loss": 4.8334,
      "mean_token_accuracy": 0.21779631376266478,
      "num_tokens": 74572673.0,
      "step": 35935
    },
    {
      "entropy": 5.229164171218872,
      "epoch": 3.2601596516690856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003956407909709068,
      "loss": 4.8531,
      "mean_token_accuracy": 0.22008637636899947,
      "num_tokens": 74582686.0,
      "step": 35940
    },
    {
      "entropy": 5.357076168060303,
      "epoch": 3.26061320754717,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039561347757739734,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20921580046415328,
      "num_tokens": 74593658.0,
      "step": 35945
    },
    {
      "entropy": 5.414747476577759,
      "epoch": 3.261066763425254,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039558616168957044,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21360286474227905,
      "num_tokens": 74603585.0,
      "step": 35950
    },
    {
      "entropy": 5.305263710021973,
      "epoch": 3.2615203193033384,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003955588433079911,
      "loss": 4.8734,
      "mean_token_accuracy": 0.21629781275987625,
      "num_tokens": 74612806.0,
      "step": 35955
    },
    {
      "entropy": 5.327981567382812,
      "epoch": 3.2619738751814222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039553152243322414,
      "loss": 4.924,
      "mean_token_accuracy": 0.21959902793169023,
      "num_tokens": 74623252.0,
      "step": 35960
    },
    {
      "entropy": 5.31094331741333,
      "epoch": 3.2624274310595065,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003955041990658346,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22054146677255632,
      "num_tokens": 74633555.0,
      "step": 35965
    },
    {
      "entropy": 5.299252605438232,
      "epoch": 3.2628809869375908,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003954768732063874,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2139931470155716,
      "num_tokens": 74643322.0,
      "step": 35970
    },
    {
      "entropy": 5.2803479671478275,
      "epoch": 3.263334542815675,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003954495448554478,
      "loss": 4.8751,
      "mean_token_accuracy": 0.21790193170309066,
      "num_tokens": 74653461.0,
      "step": 35975
    },
    {
      "entropy": 5.2948614120483395,
      "epoch": 3.263788098693759,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039542221401358104,
      "loss": 4.9323,
      "mean_token_accuracy": 0.20781135261058808,
      "num_tokens": 74663470.0,
      "step": 35980
    },
    {
      "entropy": 5.285750007629394,
      "epoch": 3.264241654571843,
      "grad_norm": 0.984375,
      "learning_rate": 0.000395394880681352,
      "loss": 4.8245,
      "mean_token_accuracy": 0.22153884917497635,
      "num_tokens": 74674074.0,
      "step": 35985
    },
    {
      "entropy": 5.394092178344726,
      "epoch": 3.2646952104499274,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039536754485932626,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2027967870235443,
      "num_tokens": 74684968.0,
      "step": 35990
    },
    {
      "entropy": 5.27443790435791,
      "epoch": 3.2651487663280117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039534020654806906,
      "loss": 4.8361,
      "mean_token_accuracy": 0.22656063586473466,
      "num_tokens": 74694776.0,
      "step": 35995
    },
    {
      "entropy": 5.246951675415039,
      "epoch": 3.265602322206096,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039531286574814563,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2132621467113495,
      "num_tokens": 74705506.0,
      "step": 36000
    },
    {
      "epoch": 3.265602322206096,
      "eval_entropy": 5.054796942326818,
      "eval_loss": 5.020824432373047,
      "eval_mean_token_accuracy": 0.21490875097977766,
      "eval_num_tokens": 74705506.0,
      "eval_runtime": 26.5659,
      "eval_samples_per_second": 1331.027,
      "eval_steps_per_second": 166.378,
      "step": 36000
    }
  ],
  "logging_steps": 5,
  "max_steps": 110230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.09311181193088e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}