{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9166034114833221,
  "eval_steps": 4000,
  "global_step": 32000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 0.3528062401339412,
      "epoch": 0.00028643856608853816,
      "grad_norm": 478.0,
      "learning_rate": 4.5e-07,
      "loss": 7.217581939697266,
      "mean_token_accuracy": 0.4640563763678074,
      "num_tokens": 19473.0,
      "step": 10
    },
    {
      "entropy": 0.461815438978374,
      "epoch": 0.0005728771321770763,
      "grad_norm": 318.0,
      "learning_rate": 9.5e-07,
      "loss": 6.147866058349609,
      "mean_token_accuracy": 0.46377463489770887,
      "num_tokens": 40258.0,
      "step": 20
    },
    {
      "entropy": 0.5792016623541713,
      "epoch": 0.0008593156982656145,
      "grad_norm": 125.5,
      "learning_rate": 1.45e-06,
      "loss": 4.201408767700196,
      "mean_token_accuracy": 0.5143063102848828,
      "num_tokens": 60493.0,
      "step": 30
    },
    {
      "entropy": 0.8141000302508473,
      "epoch": 0.0011457542643541526,
      "grad_norm": 72.0,
      "learning_rate": 1.95e-06,
      "loss": 2.4521657943725588,
      "mean_token_accuracy": 0.6026567563414573,
      "num_tokens": 78942.0,
      "step": 40
    },
    {
      "entropy": 1.61114833727479,
      "epoch": 0.0014321928304426908,
      "grad_norm": 33.25,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 2.2294458389282226,
      "mean_token_accuracy": 0.5755564626306295,
      "num_tokens": 97239.0,
      "step": 50
    },
    {
      "entropy": 1.7803453914821148,
      "epoch": 0.001718631396531229,
      "grad_norm": 34.0,
      "learning_rate": 2.95e-06,
      "loss": 1.895331573486328,
      "mean_token_accuracy": 0.6102932810783386,
      "num_tokens": 116874.0,
      "step": 60
    },
    {
      "entropy": 1.7013822786509991,
      "epoch": 0.0020050699626197673,
      "grad_norm": 39.75,
      "learning_rate": 3.4500000000000004e-06,
      "loss": 1.7933462142944336,
      "mean_token_accuracy": 0.6113637214526534,
      "num_tokens": 137643.0,
      "step": 70
    },
    {
      "entropy": 1.8509704992175102,
      "epoch": 0.0022915085287083053,
      "grad_norm": 32.0,
      "learning_rate": 3.95e-06,
      "loss": 1.9163200378417968,
      "mean_token_accuracy": 0.6079998154193162,
      "num_tokens": 158951.0,
      "step": 80
    },
    {
      "entropy": 1.765385853126645,
      "epoch": 0.0025779470947968433,
      "grad_norm": 36.5,
      "learning_rate": 4.45e-06,
      "loss": 1.8505186080932616,
      "mean_token_accuracy": 0.6070508841425181,
      "num_tokens": 179801.0,
      "step": 90
    },
    {
      "entropy": 1.7926170282065867,
      "epoch": 0.0028643856608853817,
      "grad_norm": 42.75,
      "learning_rate": 4.950000000000001e-06,
      "loss": 1.8546482086181642,
      "mean_token_accuracy": 0.597331989184022,
      "num_tokens": 198667.0,
      "step": 100
    },
    {
      "entropy": 1.8906881004571914,
      "epoch": 0.0031508242269739197,
      "grad_norm": 31.0,
      "learning_rate": 5.45e-06,
      "loss": 1.9200061798095702,
      "mean_token_accuracy": 0.6018830932676792,
      "num_tokens": 217502.0,
      "step": 110
    },
    {
      "entropy": 1.6857450231909752,
      "epoch": 0.003437262793062458,
      "grad_norm": 38.75,
      "learning_rate": 5.95e-06,
      "loss": 1.8439041137695313,
      "mean_token_accuracy": 0.6112990219146013,
      "num_tokens": 237612.0,
      "step": 120
    },
    {
      "entropy": 1.7538917355239392,
      "epoch": 0.003723701359150996,
      "grad_norm": 68.5,
      "learning_rate": 6.45e-06,
      "loss": 1.7615484237670898,
      "mean_token_accuracy": 0.6213274173438549,
      "num_tokens": 257818.0,
      "step": 130
    },
    {
      "entropy": 1.6896496571600437,
      "epoch": 0.0040101399252395345,
      "grad_norm": 42.25,
      "learning_rate": 6.950000000000001e-06,
      "loss": 1.7217609405517578,
      "mean_token_accuracy": 0.6107563082128763,
      "num_tokens": 276933.0,
      "step": 140
    },
    {
      "entropy": 1.7307493343949318,
      "epoch": 0.0042965784913280725,
      "grad_norm": 35.25,
      "learning_rate": 7.45e-06,
      "loss": 1.7894071578979491,
      "mean_token_accuracy": 0.626866776496172,
      "num_tokens": 295470.0,
      "step": 150
    },
    {
      "entropy": 1.7178809575736522,
      "epoch": 0.0045830170574166105,
      "grad_norm": 38.5,
      "learning_rate": 7.95e-06,
      "loss": 1.7587535858154297,
      "mean_token_accuracy": 0.6143172055482864,
      "num_tokens": 315763.0,
      "step": 160
    },
    {
      "entropy": 1.5212275303900242,
      "epoch": 0.0048694556235051485,
      "grad_norm": 29.0,
      "learning_rate": 8.45e-06,
      "loss": 1.5423474311828613,
      "mean_token_accuracy": 0.654015052691102,
      "num_tokens": 336144.0,
      "step": 170
    },
    {
      "entropy": 1.5958088483661412,
      "epoch": 0.0051558941895936865,
      "grad_norm": 28.0,
      "learning_rate": 8.95e-06,
      "loss": 1.5740506172180175,
      "mean_token_accuracy": 0.6468525096774101,
      "num_tokens": 355322.0,
      "step": 180
    },
    {
      "entropy": 1.6313762225210666,
      "epoch": 0.005442332755682225,
      "grad_norm": 45.0,
      "learning_rate": 9.450000000000001e-06,
      "loss": 1.6890562057495118,
      "mean_token_accuracy": 0.6329364471137524,
      "num_tokens": 374375.0,
      "step": 190
    },
    {
      "entropy": 1.6774629056453705,
      "epoch": 0.005728771321770763,
      "grad_norm": 25.875,
      "learning_rate": 9.950000000000001e-06,
      "loss": 1.7464597702026368,
      "mean_token_accuracy": 0.6420648951083422,
      "num_tokens": 395145.0,
      "step": 200
    },
    {
      "entropy": 1.6366954684257506,
      "epoch": 0.006015209887859301,
      "grad_norm": 29.125,
      "learning_rate": 1.045e-05,
      "loss": 1.7245145797729493,
      "mean_token_accuracy": 0.6189848899841308,
      "num_tokens": 414805.0,
      "step": 210
    },
    {
      "entropy": 1.6353202268481255,
      "epoch": 0.006301648453947839,
      "grad_norm": 33.5,
      "learning_rate": 1.095e-05,
      "loss": 1.5854035377502442,
      "mean_token_accuracy": 0.646422104537487,
      "num_tokens": 434175.0,
      "step": 220
    },
    {
      "entropy": 1.531673163920641,
      "epoch": 0.006588087020036377,
      "grad_norm": 29.875,
      "learning_rate": 1.145e-05,
      "loss": 1.7045608520507813,
      "mean_token_accuracy": 0.6308731701225042,
      "num_tokens": 454765.0,
      "step": 230
    },
    {
      "entropy": 1.7325628489255904,
      "epoch": 0.006874525586124916,
      "grad_norm": 29.5,
      "learning_rate": 1.195e-05,
      "loss": 1.7049570083618164,
      "mean_token_accuracy": 0.639626245573163,
      "num_tokens": 474803.0,
      "step": 240
    },
    {
      "entropy": 1.7215023070573807,
      "epoch": 0.007160964152213454,
      "grad_norm": 21.5,
      "learning_rate": 1.2450000000000001e-05,
      "loss": 1.6847394943237304,
      "mean_token_accuracy": 0.6271909900009632,
      "num_tokens": 494924.0,
      "step": 250
    },
    {
      "entropy": 1.4796019680798054,
      "epoch": 0.007447402718301992,
      "grad_norm": 30.0,
      "learning_rate": 1.2950000000000001e-05,
      "loss": 1.612894630432129,
      "mean_token_accuracy": 0.6432699490338564,
      "num_tokens": 514534.0,
      "step": 260
    },
    {
      "entropy": 1.5921406548470258,
      "epoch": 0.00773384128439053,
      "grad_norm": 28.25,
      "learning_rate": 1.3450000000000002e-05,
      "loss": 1.6738925933837892,
      "mean_token_accuracy": 0.6368084598332644,
      "num_tokens": 533648.0,
      "step": 270
    },
    {
      "entropy": 1.6388971492648126,
      "epoch": 0.008020279850479069,
      "grad_norm": 23.125,
      "learning_rate": 1.3950000000000002e-05,
      "loss": 1.5522817611694335,
      "mean_token_accuracy": 0.6521936945617199,
      "num_tokens": 553984.0,
      "step": 280
    },
    {
      "entropy": 1.6042158961296082,
      "epoch": 0.008306718416567606,
      "grad_norm": 34.5,
      "learning_rate": 1.4449999999999999e-05,
      "loss": 1.730119514465332,
      "mean_token_accuracy": 0.636293425410986,
      "num_tokens": 573549.0,
      "step": 290
    },
    {
      "entropy": 1.6883071303367614,
      "epoch": 0.008593156982656145,
      "grad_norm": 30.25,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 1.6591917037963868,
      "mean_token_accuracy": 0.647525903955102,
      "num_tokens": 593360.0,
      "step": 300
    },
    {
      "entropy": 1.481922762095928,
      "epoch": 0.008879595548744682,
      "grad_norm": 21.75,
      "learning_rate": 1.545e-05,
      "loss": 1.528256893157959,
      "mean_token_accuracy": 0.6495554637163877,
      "num_tokens": 614270.0,
      "step": 310
    },
    {
      "entropy": 1.6406896494328975,
      "epoch": 0.009166034114833221,
      "grad_norm": 32.5,
      "learning_rate": 1.595e-05,
      "loss": 1.7234817504882813,
      "mean_token_accuracy": 0.6251039749011398,
      "num_tokens": 634148.0,
      "step": 320
    },
    {
      "entropy": 1.7205604828894139,
      "epoch": 0.00945247268092176,
      "grad_norm": 30.25,
      "learning_rate": 1.645e-05,
      "loss": 1.7638006210327148,
      "mean_token_accuracy": 0.6161015894263983,
      "num_tokens": 654277.0,
      "step": 330
    },
    {
      "entropy": 1.543334063142538,
      "epoch": 0.009738911247010297,
      "grad_norm": 33.0,
      "learning_rate": 1.6950000000000002e-05,
      "loss": 1.5728675842285156,
      "mean_token_accuracy": 0.6445532537996769,
      "num_tokens": 672940.0,
      "step": 340
    },
    {
      "entropy": 1.674956924468279,
      "epoch": 0.010025349813098836,
      "grad_norm": 25.125,
      "learning_rate": 1.745e-05,
      "loss": 1.7410369873046876,
      "mean_token_accuracy": 0.6321388445794582,
      "num_tokens": 693369.0,
      "step": 350
    },
    {
      "entropy": 1.6595345363020897,
      "epoch": 0.010311788379187373,
      "grad_norm": 39.5,
      "learning_rate": 1.795e-05,
      "loss": 1.6917007446289063,
      "mean_token_accuracy": 0.6355412002652884,
      "num_tokens": 713621.0,
      "step": 360
    },
    {
      "entropy": 1.5632868006825447,
      "epoch": 0.010598226945275912,
      "grad_norm": 26.625,
      "learning_rate": 1.845e-05,
      "loss": 1.5559974670410157,
      "mean_token_accuracy": 0.6593380380421877,
      "num_tokens": 733037.0,
      "step": 370
    },
    {
      "entropy": 1.6144609708338975,
      "epoch": 0.01088466551136445,
      "grad_norm": 29.875,
      "learning_rate": 1.895e-05,
      "loss": 1.6511468887329102,
      "mean_token_accuracy": 0.6317056931555272,
      "num_tokens": 752153.0,
      "step": 380
    },
    {
      "entropy": 1.5421662211418152,
      "epoch": 0.011171104077452988,
      "grad_norm": 24.625,
      "learning_rate": 1.9450000000000002e-05,
      "loss": 1.6255882263183594,
      "mean_token_accuracy": 0.6627073056995869,
      "num_tokens": 773098.0,
      "step": 390
    },
    {
      "entropy": 1.7122172061353922,
      "epoch": 0.011457542643541527,
      "grad_norm": 21.75,
      "learning_rate": 1.995e-05,
      "loss": 1.7527050018310546,
      "mean_token_accuracy": 0.6149721326306462,
      "num_tokens": 791765.0,
      "step": 400
    },
    {
      "entropy": 1.7726871855556965,
      "epoch": 0.011743981209630064,
      "grad_norm": 26.0,
      "learning_rate": 2.045e-05,
      "loss": 1.7686229705810548,
      "mean_token_accuracy": 0.6190072495490313,
      "num_tokens": 811401.0,
      "step": 410
    },
    {
      "entropy": 1.5961021460592746,
      "epoch": 0.012030419775718603,
      "grad_norm": 25.875,
      "learning_rate": 2.095e-05,
      "loss": 1.7156158447265626,
      "mean_token_accuracy": 0.6386278152465821,
      "num_tokens": 830948.0,
      "step": 420
    },
    {
      "entropy": 1.6144308768212796,
      "epoch": 0.012316858341807142,
      "grad_norm": 27.875,
      "learning_rate": 2.145e-05,
      "loss": 1.6228755950927733,
      "mean_token_accuracy": 0.634922893717885,
      "num_tokens": 849629.0,
      "step": 430
    },
    {
      "entropy": 1.683132666349411,
      "epoch": 0.012603296907895679,
      "grad_norm": 29.0,
      "learning_rate": 2.195e-05,
      "loss": 1.7449392318725585,
      "mean_token_accuracy": 0.6308464124798775,
      "num_tokens": 868191.0,
      "step": 440
    },
    {
      "entropy": 1.625600465387106,
      "epoch": 0.012889735473984218,
      "grad_norm": 25.25,
      "learning_rate": 2.245e-05,
      "loss": 1.6958326339721679,
      "mean_token_accuracy": 0.6358420550823212,
      "num_tokens": 887523.0,
      "step": 450
    },
    {
      "entropy": 1.7978037655353547,
      "epoch": 0.013176174040072755,
      "grad_norm": 27.0,
      "learning_rate": 2.2950000000000002e-05,
      "loss": 1.83199462890625,
      "mean_token_accuracy": 0.6068651460111141,
      "num_tokens": 908979.0,
      "step": 460
    },
    {
      "entropy": 1.6116399303078652,
      "epoch": 0.013462612606161294,
      "grad_norm": 18.25,
      "learning_rate": 2.345e-05,
      "loss": 1.6807987213134765,
      "mean_token_accuracy": 0.6298470726236701,
      "num_tokens": 929167.0,
      "step": 470
    },
    {
      "entropy": 1.694230755418539,
      "epoch": 0.013749051172249832,
      "grad_norm": 29.125,
      "learning_rate": 2.395e-05,
      "loss": 1.7597953796386718,
      "mean_token_accuracy": 0.6239655200392008,
      "num_tokens": 948118.0,
      "step": 480
    },
    {
      "entropy": 1.6106532506644726,
      "epoch": 0.01403548973833837,
      "grad_norm": 21.0,
      "learning_rate": 2.445e-05,
      "loss": 1.7235946655273438,
      "mean_token_accuracy": 0.6164496175944805,
      "num_tokens": 968805.0,
      "step": 490
    },
    {
      "entropy": 1.8479979574680327,
      "epoch": 0.014321928304426908,
      "grad_norm": 20.625,
      "learning_rate": 2.495e-05,
      "loss": 1.8998939514160156,
      "mean_token_accuracy": 0.6054852087050676,
      "num_tokens": 988060.0,
      "step": 500
    },
    {
      "entropy": 1.8584213972091674,
      "epoch": 0.014608366870515446,
      "grad_norm": 20.375,
      "learning_rate": 2.5450000000000002e-05,
      "loss": 1.8704095840454102,
      "mean_token_accuracy": 0.6105503164231777,
      "num_tokens": 1007701.0,
      "step": 510
    },
    {
      "entropy": 1.6088180273771286,
      "epoch": 0.014894805436603984,
      "grad_norm": 31.0,
      "learning_rate": 2.595e-05,
      "loss": 1.7482704162597655,
      "mean_token_accuracy": 0.6043568558990955,
      "num_tokens": 1026950.0,
      "step": 520
    },
    {
      "entropy": 1.6710659384727478,
      "epoch": 0.015181244002692523,
      "grad_norm": 35.75,
      "learning_rate": 2.6450000000000003e-05,
      "loss": 1.7560743331909179,
      "mean_token_accuracy": 0.6205087993294001,
      "num_tokens": 1048349.0,
      "step": 530
    },
    {
      "entropy": 1.841731084138155,
      "epoch": 0.01546768256878106,
      "grad_norm": 21.25,
      "learning_rate": 2.6950000000000005e-05,
      "loss": 1.9027681350708008,
      "mean_token_accuracy": 0.602517394348979,
      "num_tokens": 1068676.0,
      "step": 540
    },
    {
      "entropy": 1.8934282220900058,
      "epoch": 0.015754121134869598,
      "grad_norm": 36.25,
      "learning_rate": 2.7450000000000003e-05,
      "loss": 1.918873405456543,
      "mean_token_accuracy": 0.5876505618914962,
      "num_tokens": 1090010.0,
      "step": 550
    },
    {
      "entropy": 1.8826362006366253,
      "epoch": 0.016040559700958138,
      "grad_norm": 17.875,
      "learning_rate": 2.7950000000000005e-05,
      "loss": 1.9828632354736329,
      "mean_token_accuracy": 0.5951655959710479,
      "num_tokens": 1109984.0,
      "step": 560
    },
    {
      "entropy": 1.849650678038597,
      "epoch": 0.016326998267046675,
      "grad_norm": 18.5,
      "learning_rate": 2.845e-05,
      "loss": 1.897369384765625,
      "mean_token_accuracy": 0.5904229883104563,
      "num_tokens": 1128870.0,
      "step": 570
    },
    {
      "entropy": 1.6490569829940795,
      "epoch": 0.016613436833135212,
      "grad_norm": 19.5,
      "learning_rate": 2.895e-05,
      "loss": 1.7149398803710938,
      "mean_token_accuracy": 0.6223928950726986,
      "num_tokens": 1149013.0,
      "step": 580
    },
    {
      "entropy": 1.7709774278104304,
      "epoch": 0.016899875399223753,
      "grad_norm": 22.875,
      "learning_rate": 2.945e-05,
      "loss": 1.8379535675048828,
      "mean_token_accuracy": 0.5926799420267344,
      "num_tokens": 1169525.0,
      "step": 590
    },
    {
      "entropy": 1.7111972540616989,
      "epoch": 0.01718631396531229,
      "grad_norm": 24.5,
      "learning_rate": 2.995e-05,
      "loss": 1.7761560440063477,
      "mean_token_accuracy": 0.6236701173707843,
      "num_tokens": 1188120.0,
      "step": 600
    },
    {
      "entropy": 1.823415072262287,
      "epoch": 0.017472752531400827,
      "grad_norm": 24.0,
      "learning_rate": 3.045e-05,
      "loss": 1.8971757888793945,
      "mean_token_accuracy": 0.6017429701983928,
      "num_tokens": 1209105.0,
      "step": 610
    },
    {
      "entropy": 1.7498523406684399,
      "epoch": 0.017759191097489364,
      "grad_norm": 26.75,
      "learning_rate": 3.095e-05,
      "loss": 1.817325973510742,
      "mean_token_accuracy": 0.608129658550024,
      "num_tokens": 1228466.0,
      "step": 620
    },
    {
      "entropy": 1.8355644896626473,
      "epoch": 0.018045629663577905,
      "grad_norm": 27.375,
      "learning_rate": 3.145e-05,
      "loss": 1.9128643035888673,
      "mean_token_accuracy": 0.5950730726122856,
      "num_tokens": 1247797.0,
      "step": 630
    },
    {
      "entropy": 1.8622040383517742,
      "epoch": 0.018332068229666442,
      "grad_norm": 24.875,
      "learning_rate": 3.1950000000000004e-05,
      "loss": 1.9076318740844727,
      "mean_token_accuracy": 0.5939670577645302,
      "num_tokens": 1267311.0,
      "step": 640
    },
    {
      "entropy": 1.9695768848061561,
      "epoch": 0.01861850679575498,
      "grad_norm": 18.125,
      "learning_rate": 3.245e-05,
      "loss": 2.0158262252807617,
      "mean_token_accuracy": 0.5735625583678484,
      "num_tokens": 1287020.0,
      "step": 650
    },
    {
      "entropy": 1.8041729524731636,
      "epoch": 0.01890494536184352,
      "grad_norm": 20.125,
      "learning_rate": 3.295e-05,
      "loss": 1.9124000549316407,
      "mean_token_accuracy": 0.5887357614934444,
      "num_tokens": 1306669.0,
      "step": 660
    },
    {
      "entropy": 1.679208005964756,
      "epoch": 0.019191383927932057,
      "grad_norm": 23.75,
      "learning_rate": 3.345000000000001e-05,
      "loss": 1.755006980895996,
      "mean_token_accuracy": 0.6205299673601985,
      "num_tokens": 1325991.0,
      "step": 670
    },
    {
      "entropy": 1.9598253771662713,
      "epoch": 0.019477822494020594,
      "grad_norm": 29.0,
      "learning_rate": 3.3950000000000005e-05,
      "loss": 2.0003707885742186,
      "mean_token_accuracy": 0.5839478272944689,
      "num_tokens": 1345046.0,
      "step": 680
    },
    {
      "entropy": 1.7889393791556358,
      "epoch": 0.019764261060109135,
      "grad_norm": 24.125,
      "learning_rate": 3.445e-05,
      "loss": 1.8787141799926759,
      "mean_token_accuracy": 0.6056120421737432,
      "num_tokens": 1364092.0,
      "step": 690
    },
    {
      "entropy": 1.8079691872000694,
      "epoch": 0.020050699626197672,
      "grad_norm": 16.625,
      "learning_rate": 3.495e-05,
      "loss": 1.8646484375,
      "mean_token_accuracy": 0.6014414165169001,
      "num_tokens": 1384087.0,
      "step": 700
    },
    {
      "entropy": 1.8085278443992139,
      "epoch": 0.02033713819228621,
      "grad_norm": 18.75,
      "learning_rate": 3.545e-05,
      "loss": 1.9149509429931642,
      "mean_token_accuracy": 0.5895494751632213,
      "num_tokens": 1404794.0,
      "step": 710
    },
    {
      "entropy": 1.9263403855264187,
      "epoch": 0.020623576758374746,
      "grad_norm": 19.625,
      "learning_rate": 3.595e-05,
      "loss": 1.984172248840332,
      "mean_token_accuracy": 0.5828634317964315,
      "num_tokens": 1425881.0,
      "step": 720
    },
    {
      "entropy": 2.035868939012289,
      "epoch": 0.020910015324463287,
      "grad_norm": 22.75,
      "learning_rate": 3.645e-05,
      "loss": 2.1743022918701174,
      "mean_token_accuracy": 0.5523974321782589,
      "num_tokens": 1445642.0,
      "step": 730
    },
    {
      "entropy": 1.9566943377256394,
      "epoch": 0.021196453890551824,
      "grad_norm": 15.1875,
      "learning_rate": 3.6950000000000004e-05,
      "loss": 1.9452693939208985,
      "mean_token_accuracy": 0.59082987960428,
      "num_tokens": 1465223.0,
      "step": 740
    },
    {
      "entropy": 1.7797305561602115,
      "epoch": 0.02148289245664036,
      "grad_norm": 42.25,
      "learning_rate": 3.745e-05,
      "loss": 2.0244304656982424,
      "mean_token_accuracy": 0.5785138387233019,
      "num_tokens": 1485173.0,
      "step": 750
    },
    {
      "entropy": 2.1499163672327994,
      "epoch": 0.0217693310227289,
      "grad_norm": 25.125,
      "learning_rate": 3.795e-05,
      "loss": 2.1835763931274412,
      "mean_token_accuracy": 0.5378579221665859,
      "num_tokens": 1505269.0,
      "step": 760
    },
    {
      "entropy": 1.9885080248117446,
      "epoch": 0.02205576958881744,
      "grad_norm": 23.25,
      "learning_rate": 3.845e-05,
      "loss": 2.0707183837890626,
      "mean_token_accuracy": 0.5810061223804951,
      "num_tokens": 1525423.0,
      "step": 770
    },
    {
      "entropy": 1.8684268310666083,
      "epoch": 0.022342208154905976,
      "grad_norm": 20.0,
      "learning_rate": 3.8950000000000005e-05,
      "loss": 1.964059066772461,
      "mean_token_accuracy": 0.5858997877687215,
      "num_tokens": 1546158.0,
      "step": 780
    },
    {
      "entropy": 1.8782096415758134,
      "epoch": 0.022628646720994516,
      "grad_norm": 17.75,
      "learning_rate": 3.9450000000000003e-05,
      "loss": 1.9113346099853517,
      "mean_token_accuracy": 0.5886166680604219,
      "num_tokens": 1566144.0,
      "step": 790
    },
    {
      "entropy": 1.9770341627299786,
      "epoch": 0.022915085287083053,
      "grad_norm": 17.875,
      "learning_rate": 3.995e-05,
      "loss": 2.07691650390625,
      "mean_token_accuracy": 0.5895924907177686,
      "num_tokens": 1585496.0,
      "step": 800
    },
    {
      "entropy": 1.9110678054392338,
      "epoch": 0.02320152385317159,
      "grad_norm": 20.25,
      "learning_rate": 4.045000000000001e-05,
      "loss": 2.1008880615234373,
      "mean_token_accuracy": 0.572381529957056,
      "num_tokens": 1604848.0,
      "step": 810
    },
    {
      "entropy": 1.9095255173742771,
      "epoch": 0.023487962419260128,
      "grad_norm": 29.0,
      "learning_rate": 4.095e-05,
      "loss": 1.9740182876586914,
      "mean_token_accuracy": 0.5741673868149519,
      "num_tokens": 1624503.0,
      "step": 820
    },
    {
      "entropy": 2.0804278537631036,
      "epoch": 0.02377440098534867,
      "grad_norm": 21.125,
      "learning_rate": 4.145e-05,
      "loss": 2.1728065490722654,
      "mean_token_accuracy": 0.5477346788160503,
      "num_tokens": 1645145.0,
      "step": 830
    },
    {
      "entropy": 1.9922449916601181,
      "epoch": 0.024060839551437205,
      "grad_norm": 18.375,
      "learning_rate": 4.195e-05,
      "loss": 1.9790054321289063,
      "mean_token_accuracy": 0.5938107941299677,
      "num_tokens": 1663819.0,
      "step": 840
    },
    {
      "entropy": 2.0197327487170695,
      "epoch": 0.024347278117525743,
      "grad_norm": 20.375,
      "learning_rate": 4.245e-05,
      "loss": 2.2007863998413084,
      "mean_token_accuracy": 0.5380131371319294,
      "num_tokens": 1683563.0,
      "step": 850
    },
    {
      "entropy": 1.8699860960245132,
      "epoch": 0.024633716683614283,
      "grad_norm": 26.375,
      "learning_rate": 4.295e-05,
      "loss": 1.9295654296875,
      "mean_token_accuracy": 0.58940952681005,
      "num_tokens": 1703387.0,
      "step": 860
    },
    {
      "entropy": 1.9832911752164364,
      "epoch": 0.02492015524970282,
      "grad_norm": 17.25,
      "learning_rate": 4.345e-05,
      "loss": 2.159903717041016,
      "mean_token_accuracy": 0.5491482594981789,
      "num_tokens": 1723297.0,
      "step": 870
    },
    {
      "entropy": 2.1792681008577346,
      "epoch": 0.025206593815791357,
      "grad_norm": 17.625,
      "learning_rate": 4.3950000000000004e-05,
      "loss": 2.219806671142578,
      "mean_token_accuracy": 0.5384906796738506,
      "num_tokens": 1742641.0,
      "step": 880
    },
    {
      "entropy": 1.8695313930511475,
      "epoch": 0.025493032381879898,
      "grad_norm": 18.625,
      "learning_rate": 4.445e-05,
      "loss": 1.9873180389404297,
      "mean_token_accuracy": 0.5735095415264369,
      "num_tokens": 1762901.0,
      "step": 890
    },
    {
      "entropy": 1.8507504656910896,
      "epoch": 0.025779470947968435,
      "grad_norm": 17.25,
      "learning_rate": 4.495e-05,
      "loss": 1.971590805053711,
      "mean_token_accuracy": 0.5827929314225913,
      "num_tokens": 1781407.0,
      "step": 900
    },
    {
      "entropy": 2.032640960812569,
      "epoch": 0.026065909514056972,
      "grad_norm": 26.125,
      "learning_rate": 4.545000000000001e-05,
      "loss": 2.178053855895996,
      "mean_token_accuracy": 0.5478040888905525,
      "num_tokens": 1799836.0,
      "step": 910
    },
    {
      "entropy": 2.1317695669829844,
      "epoch": 0.02635234808014551,
      "grad_norm": 24.75,
      "learning_rate": 4.5950000000000006e-05,
      "loss": 2.202964019775391,
      "mean_token_accuracy": 0.5512260548770428,
      "num_tokens": 1819871.0,
      "step": 920
    },
    {
      "entropy": 2.1580834552645682,
      "epoch": 0.02663878664623405,
      "grad_norm": 25.875,
      "learning_rate": 4.6450000000000004e-05,
      "loss": 2.2967807769775392,
      "mean_token_accuracy": 0.5261422438547015,
      "num_tokens": 1839653.0,
      "step": 930
    },
    {
      "entropy": 1.8693755000829697,
      "epoch": 0.026925225212322587,
      "grad_norm": 26.625,
      "learning_rate": 4.695e-05,
      "loss": 1.920684242248535,
      "mean_token_accuracy": 0.5878301737830043,
      "num_tokens": 1858425.0,
      "step": 940
    },
    {
      "entropy": 1.9177309423685074,
      "epoch": 0.027211663778411124,
      "grad_norm": 25.0,
      "learning_rate": 4.745e-05,
      "loss": 2.0325536727905273,
      "mean_token_accuracy": 0.5708646427839994,
      "num_tokens": 1877432.0,
      "step": 950
    },
    {
      "entropy": 2.0536071196198464,
      "epoch": 0.027498102344499665,
      "grad_norm": 19.75,
      "learning_rate": 4.795e-05,
      "loss": 2.230321502685547,
      "mean_token_accuracy": 0.5467239493504167,
      "num_tokens": 1896411.0,
      "step": 960
    },
    {
      "entropy": 2.0146598778665066,
      "epoch": 0.027784540910588202,
      "grad_norm": 19.25,
      "learning_rate": 4.845e-05,
      "loss": 2.0646324157714844,
      "mean_token_accuracy": 0.5636048406362534,
      "num_tokens": 1916224.0,
      "step": 970
    },
    {
      "entropy": 1.8649965476244688,
      "epoch": 0.02807097947667674,
      "grad_norm": 19.5,
      "learning_rate": 4.8950000000000004e-05,
      "loss": 2.0268020629882812,
      "mean_token_accuracy": 0.5778200808912516,
      "num_tokens": 1935380.0,
      "step": 980
    },
    {
      "entropy": 2.224521817266941,
      "epoch": 0.028357418042765276,
      "grad_norm": 20.25,
      "learning_rate": 4.945e-05,
      "loss": 2.29847412109375,
      "mean_token_accuracy": 0.5251183599233628,
      "num_tokens": 1954696.0,
      "step": 990
    },
    {
      "entropy": 1.9635957852005959,
      "epoch": 0.028643856608853817,
      "grad_norm": 24.125,
      "learning_rate": 4.995e-05,
      "loss": 2.1004880905151366,
      "mean_token_accuracy": 0.5523587435483932,
      "num_tokens": 1975488.0,
      "step": 1000
    },
    {
      "entropy": 2.1570306584239005,
      "epoch": 0.028930295174942354,
      "grad_norm": 21.0,
      "learning_rate": 4.9999997890327946e-05,
      "loss": 2.2056819915771486,
      "mean_token_accuracy": 0.5517607279121876,
      "num_tokens": 1995757.0,
      "step": 1010
    },
    {
      "entropy": 2.2149836353957655,
      "epoch": 0.02921673374103089,
      "grad_norm": 16.25,
      "learning_rate": 4.999999059763485e-05,
      "loss": 2.3002399444580077,
      "mean_token_accuracy": 0.5277919966727496,
      "num_tokens": 2014782.0,
      "step": 1020
    },
    {
      "entropy": 1.9475185506045818,
      "epoch": 0.02950317230711943,
      "grad_norm": 14.5,
      "learning_rate": 4.999997809587693e-05,
      "loss": 2.101923942565918,
      "mean_token_accuracy": 0.562053806334734,
      "num_tokens": 2035343.0,
      "step": 1030
    },
    {
      "entropy": 2.0823429681360723,
      "epoch": 0.02978961087320797,
      "grad_norm": 19.625,
      "learning_rate": 4.9999960385056766e-05,
      "loss": 2.2225202560424804,
      "mean_token_accuracy": 0.5534175606444478,
      "num_tokens": 2055608.0,
      "step": 1040
    },
    {
      "entropy": 2.007365348935127,
      "epoch": 0.030076049439296506,
      "grad_norm": 18.625,
      "learning_rate": 4.999993746517806e-05,
      "loss": 2.1226667404174804,
      "mean_token_accuracy": 0.5498795179650188,
      "num_tokens": 2075593.0,
      "step": 1050
    },
    {
      "entropy": 2.0506470412015916,
      "epoch": 0.030362488005385047,
      "grad_norm": 20.625,
      "learning_rate": 4.9999909336245585e-05,
      "loss": 2.2072338104248046,
      "mean_token_accuracy": 0.5508825996890664,
      "num_tokens": 2095591.0,
      "step": 1060
    },
    {
      "entropy": 2.03544892296195,
      "epoch": 0.030648926571473584,
      "grad_norm": 13.0,
      "learning_rate": 4.999987599826519e-05,
      "loss": 2.0811920166015625,
      "mean_token_accuracy": 0.5673403698951006,
      "num_tokens": 2114586.0,
      "step": 1070
    },
    {
      "entropy": 2.065407223254442,
      "epoch": 0.03093536513756212,
      "grad_norm": 15.8125,
      "learning_rate": 4.999983745124385e-05,
      "loss": 2.3352025985717773,
      "mean_token_accuracy": 0.5290726596489549,
      "num_tokens": 2134892.0,
      "step": 1080
    },
    {
      "entropy": 2.168225467950106,
      "epoch": 0.031221803703650658,
      "grad_norm": 15.3125,
      "learning_rate": 4.999979369518957e-05,
      "loss": 2.169589042663574,
      "mean_token_accuracy": 0.5483372673392296,
      "num_tokens": 2155329.0,
      "step": 1090
    },
    {
      "entropy": 1.9775354914367198,
      "epoch": 0.031508242269739195,
      "grad_norm": 15.1875,
      "learning_rate": 4.9999744730111476e-05,
      "loss": 2.2050197601318358,
      "mean_token_accuracy": 0.5570029642432928,
      "num_tokens": 2174335.0,
      "step": 1100
    },
    {
      "entropy": 2.0876759525388477,
      "epoch": 0.03179468083582774,
      "grad_norm": 16.5,
      "learning_rate": 4.999969055601978e-05,
      "loss": 2.1351606369018556,
      "mean_token_accuracy": 0.5560093311592936,
      "num_tokens": 2193061.0,
      "step": 1110
    },
    {
      "entropy": 1.8840322092175483,
      "epoch": 0.032081119401916276,
      "grad_norm": 21.25,
      "learning_rate": 4.999963117292575e-05,
      "loss": 2.1137130737304686,
      "mean_token_accuracy": 0.5579068209975958,
      "num_tokens": 2211784.0,
      "step": 1120
    },
    {
      "entropy": 2.118617121130228,
      "epoch": 0.03236755796800481,
      "grad_norm": 22.375,
      "learning_rate": 4.999956658084177e-05,
      "loss": 2.180125427246094,
      "mean_token_accuracy": 0.5428189273923636,
      "num_tokens": 2230788.0,
      "step": 1130
    },
    {
      "entropy": 2.233527162671089,
      "epoch": 0.03265399653409335,
      "grad_norm": 15.125,
      "learning_rate": 4.9999496779781306e-05,
      "loss": 2.3609376907348634,
      "mean_token_accuracy": 0.5256578113883734,
      "num_tokens": 2250811.0,
      "step": 1140
    },
    {
      "entropy": 2.031604032218456,
      "epoch": 0.03294043510018189,
      "grad_norm": 15.0625,
      "learning_rate": 4.999942176975889e-05,
      "loss": 2.085622024536133,
      "mean_token_accuracy": 0.5635854002088309,
      "num_tokens": 2269129.0,
      "step": 1150
    },
    {
      "entropy": 2.046806565672159,
      "epoch": 0.033226873666270425,
      "grad_norm": 13.0,
      "learning_rate": 4.999934155079016e-05,
      "loss": 2.2095598220825194,
      "mean_token_accuracy": 0.5431034330278635,
      "num_tokens": 2288066.0,
      "step": 1160
    },
    {
      "entropy": 2.033204253017902,
      "epoch": 0.03351331223235896,
      "grad_norm": 17.875,
      "learning_rate": 4.999925612289182e-05,
      "loss": 2.186530113220215,
      "mean_token_accuracy": 0.5451825274154544,
      "num_tokens": 2309079.0,
      "step": 1170
    },
    {
      "entropy": 2.0099747076630594,
      "epoch": 0.033799750798447506,
      "grad_norm": 12.875,
      "learning_rate": 4.9999165486081686e-05,
      "loss": 2.1805286407470703,
      "mean_token_accuracy": 0.5489942122250795,
      "num_tokens": 2328780.0,
      "step": 1180
    },
    {
      "entropy": 1.8404223743826151,
      "epoch": 0.03408618936453604,
      "grad_norm": 14.125,
      "learning_rate": 4.9999069640378635e-05,
      "loss": 1.974308395385742,
      "mean_token_accuracy": 0.5997153293341398,
      "num_tokens": 2348426.0,
      "step": 1190
    },
    {
      "entropy": 2.000691217184067,
      "epoch": 0.03437262793062458,
      "grad_norm": 22.75,
      "learning_rate": 4.999896858580263e-05,
      "loss": 2.0610605239868165,
      "mean_token_accuracy": 0.5601603195071221,
      "num_tokens": 2368349.0,
      "step": 1200
    },
    {
      "entropy": 2.043005222082138,
      "epoch": 0.03465906649671312,
      "grad_norm": 26.0,
      "learning_rate": 4.999886232237473e-05,
      "loss": 2.1969589233398437,
      "mean_token_accuracy": 0.5667722446843981,
      "num_tokens": 2387562.0,
      "step": 1210
    },
    {
      "entropy": 1.9800917744636535,
      "epoch": 0.034945505062801654,
      "grad_norm": 17.0,
      "learning_rate": 4.9998750850117085e-05,
      "loss": 2.084916687011719,
      "mean_token_accuracy": 0.5710238970816135,
      "num_tokens": 2408208.0,
      "step": 1220
    },
    {
      "entropy": 2.0688542298972608,
      "epoch": 0.03523194362889019,
      "grad_norm": 15.0625,
      "learning_rate": 4.9998634169052915e-05,
      "loss": 2.211404228210449,
      "mean_token_accuracy": 0.5368290698155761,
      "num_tokens": 2428408.0,
      "step": 1230
    },
    {
      "entropy": 2.039776177704334,
      "epoch": 0.03551838219497873,
      "grad_norm": 11.75,
      "learning_rate": 4.999851227920654e-05,
      "loss": 2.1168548583984377,
      "mean_token_accuracy": 0.5706047866493463,
      "num_tokens": 2449498.0,
      "step": 1240
    },
    {
      "entropy": 2.042325101047754,
      "epoch": 0.03580482076106727,
      "grad_norm": 13.0,
      "learning_rate": 4.9998385180603346e-05,
      "loss": 2.1548297882080076,
      "mean_token_accuracy": 0.5527641370892524,
      "num_tokens": 2468777.0,
      "step": 1250
    },
    {
      "entropy": 1.9143792726099491,
      "epoch": 0.03609125932715581,
      "grad_norm": 13.9375,
      "learning_rate": 4.999825287326982e-05,
      "loss": 2.079159736633301,
      "mean_token_accuracy": 0.5709832672029733,
      "num_tokens": 2488258.0,
      "step": 1260
    },
    {
      "entropy": 2.112581159174442,
      "epoch": 0.03637769789324435,
      "grad_norm": 15.8125,
      "learning_rate": 4.9998115357233536e-05,
      "loss": 2.211891746520996,
      "mean_token_accuracy": 0.54996342882514,
      "num_tokens": 2507098.0,
      "step": 1270
    },
    {
      "entropy": 2.0279450356960296,
      "epoch": 0.036664136459332884,
      "grad_norm": 15.625,
      "learning_rate": 4.9997972632523136e-05,
      "loss": 2.1490793228149414,
      "mean_token_accuracy": 0.5442783921957016,
      "num_tokens": 2527918.0,
      "step": 1280
    },
    {
      "entropy": 2.1258341267704965,
      "epoch": 0.03695057502542142,
      "grad_norm": 18.25,
      "learning_rate": 4.999782469916836e-05,
      "loss": 2.2387834548950196,
      "mean_token_accuracy": 0.5409025836735963,
      "num_tokens": 2547346.0,
      "step": 1290
    },
    {
      "entropy": 2.055143154412508,
      "epoch": 0.03723701359150996,
      "grad_norm": 13.625,
      "learning_rate": 4.999767155720004e-05,
      "loss": 2.177887535095215,
      "mean_token_accuracy": 0.5544646181166172,
      "num_tokens": 2566815.0,
      "step": 1300
    },
    {
      "entropy": 1.906372955441475,
      "epoch": 0.0375234521575985,
      "grad_norm": 17.25,
      "learning_rate": 4.999751320665008e-05,
      "loss": 1.993330955505371,
      "mean_token_accuracy": 0.5768570452928543,
      "num_tokens": 2586328.0,
      "step": 1310
    },
    {
      "entropy": 1.9420357584953307,
      "epoch": 0.03780989072368704,
      "grad_norm": 13.0,
      "learning_rate": 4.999734964755148e-05,
      "loss": 2.1593313217163086,
      "mean_token_accuracy": 0.5613118495792151,
      "num_tokens": 2606494.0,
      "step": 1320
    },
    {
      "entropy": 2.064364905655384,
      "epoch": 0.03809632928977558,
      "grad_norm": 12.5,
      "learning_rate": 4.999718087993831e-05,
      "loss": 2.0964239120483397,
      "mean_token_accuracy": 0.564254280552268,
      "num_tokens": 2625743.0,
      "step": 1330
    },
    {
      "entropy": 2.044856058806181,
      "epoch": 0.038382767855864114,
      "grad_norm": 14.75,
      "learning_rate": 4.999700690384574e-05,
      "loss": 2.1857213973999023,
      "mean_token_accuracy": 0.5576331483200192,
      "num_tokens": 2644531.0,
      "step": 1340
    },
    {
      "entropy": 1.991876221448183,
      "epoch": 0.03866920642195265,
      "grad_norm": 13.25,
      "learning_rate": 4.9996827719310024e-05,
      "loss": 2.1038537979125977,
      "mean_token_accuracy": 0.5588052786886692,
      "num_tokens": 2664797.0,
      "step": 1350
    },
    {
      "entropy": 2.0421272471547125,
      "epoch": 0.03895564498804119,
      "grad_norm": 14.1875,
      "learning_rate": 4.9996643326368496e-05,
      "loss": 2.1337377548217775,
      "mean_token_accuracy": 0.5599613673985004,
      "num_tokens": 2685813.0,
      "step": 1360
    },
    {
      "entropy": 2.064424628019333,
      "epoch": 0.039242083554129725,
      "grad_norm": 17.0,
      "learning_rate": 4.9996453725059565e-05,
      "loss": 2.1573770523071287,
      "mean_token_accuracy": 0.5383721258491277,
      "num_tokens": 2705953.0,
      "step": 1370
    },
    {
      "entropy": 1.9960372231900692,
      "epoch": 0.03952852212021827,
      "grad_norm": 18.125,
      "learning_rate": 4.999625891542275e-05,
      "loss": 2.139292335510254,
      "mean_token_accuracy": 0.5599675431847573,
      "num_tokens": 2725694.0,
      "step": 1380
    },
    {
      "entropy": 1.9672050677239894,
      "epoch": 0.039814960686306806,
      "grad_norm": 20.625,
      "learning_rate": 4.999605889749864e-05,
      "loss": 2.080162239074707,
      "mean_token_accuracy": 0.5670911554247141,
      "num_tokens": 2745756.0,
      "step": 1390
    },
    {
      "entropy": 1.8259234957396984,
      "epoch": 0.040101399252395344,
      "grad_norm": 15.0625,
      "learning_rate": 4.999585367132891e-05,
      "loss": 1.9446260452270507,
      "mean_token_accuracy": 0.6032669235020875,
      "num_tokens": 2765605.0,
      "step": 1400
    },
    {
      "entropy": 2.067405442893505,
      "epoch": 0.04038783781848388,
      "grad_norm": 17.5,
      "learning_rate": 4.999564323695631e-05,
      "loss": 2.2064241409301757,
      "mean_token_accuracy": 0.556842989474535,
      "num_tokens": 2784802.0,
      "step": 1410
    },
    {
      "entropy": 2.0353706762194634,
      "epoch": 0.04067427638457242,
      "grad_norm": 12.5,
      "learning_rate": 4.999542759442471e-05,
      "loss": 2.1227603912353517,
      "mean_token_accuracy": 0.5624410983175039,
      "num_tokens": 2804798.0,
      "step": 1420
    },
    {
      "entropy": 1.9824700266122819,
      "epoch": 0.040960714950660955,
      "grad_norm": 14.5625,
      "learning_rate": 4.999520674377902e-05,
      "loss": 2.142858695983887,
      "mean_token_accuracy": 0.5562454015016556,
      "num_tokens": 2824349.0,
      "step": 1430
    },
    {
      "entropy": 1.9648552320897579,
      "epoch": 0.04124715351674949,
      "grad_norm": 13.3125,
      "learning_rate": 4.9994980685065276e-05,
      "loss": 2.0884014129638673,
      "mean_token_accuracy": 0.5572406148537994,
      "num_tokens": 2843817.0,
      "step": 1440
    },
    {
      "entropy": 2.073359905928373,
      "epoch": 0.041533592082838036,
      "grad_norm": 14.5,
      "learning_rate": 4.9994749418330563e-05,
      "loss": 2.236592674255371,
      "mean_token_accuracy": 0.5480291146785021,
      "num_tokens": 2864555.0,
      "step": 1450
    },
    {
      "entropy": 2.026448929309845,
      "epoch": 0.04182003064892657,
      "grad_norm": 12.875,
      "learning_rate": 4.999451294362307e-05,
      "loss": 2.030807685852051,
      "mean_token_accuracy": 0.5731288217008114,
      "num_tokens": 2883361.0,
      "step": 1460
    },
    {
      "entropy": 2.094747570902109,
      "epoch": 0.04210646921501511,
      "grad_norm": 14.0625,
      "learning_rate": 4.9994271260992085e-05,
      "loss": 2.154788017272949,
      "mean_token_accuracy": 0.554672435671091,
      "num_tokens": 2903559.0,
      "step": 1470
    },
    {
      "entropy": 1.834746278077364,
      "epoch": 0.04239290778110365,
      "grad_norm": 15.0,
      "learning_rate": 4.999402437048796e-05,
      "loss": 1.9966672897338866,
      "mean_token_accuracy": 0.5871462974697351,
      "num_tokens": 2922528.0,
      "step": 1480
    },
    {
      "entropy": 1.852889922261238,
      "epoch": 0.042679346347192185,
      "grad_norm": 12.375,
      "learning_rate": 4.999377227216212e-05,
      "loss": 1.992462158203125,
      "mean_token_accuracy": 0.5888149295002222,
      "num_tokens": 2942994.0,
      "step": 1490
    },
    {
      "entropy": 1.9615300245583058,
      "epoch": 0.04296578491328072,
      "grad_norm": 10.125,
      "learning_rate": 4.999351496606711e-05,
      "loss": 2.046071243286133,
      "mean_token_accuracy": 0.5736448410898447,
      "num_tokens": 2963415.0,
      "step": 1500
    },
    {
      "entropy": 2.1007414549589156,
      "epoch": 0.04325222347936926,
      "grad_norm": 15.9375,
      "learning_rate": 4.999325245225654e-05,
      "loss": 2.2497447967529296,
      "mean_token_accuracy": 0.5430479362607002,
      "num_tokens": 2982219.0,
      "step": 1510
    },
    {
      "entropy": 1.9808135472238064,
      "epoch": 0.0435386620454578,
      "grad_norm": 17.375,
      "learning_rate": 4.999298473078512e-05,
      "loss": 2.061616134643555,
      "mean_token_accuracy": 0.5665172085165977,
      "num_tokens": 3002660.0,
      "step": 1520
    },
    {
      "entropy": 2.00408933237195,
      "epoch": 0.04382510061154634,
      "grad_norm": 9.9375,
      "learning_rate": 4.999271180170861e-05,
      "loss": 2.195215034484863,
      "mean_token_accuracy": 0.5668820874765516,
      "num_tokens": 3021865.0,
      "step": 1530
    },
    {
      "entropy": 1.9539212465286255,
      "epoch": 0.04411153917763488,
      "grad_norm": 13.75,
      "learning_rate": 4.999243366508389e-05,
      "loss": 1.9769552230834961,
      "mean_token_accuracy": 0.5828819796442986,
      "num_tokens": 3041788.0,
      "step": 1540
    },
    {
      "entropy": 1.8997029036283493,
      "epoch": 0.044397977743723414,
      "grad_norm": 12.25,
      "learning_rate": 4.9992150320968914e-05,
      "loss": 2.0064029693603516,
      "mean_token_accuracy": 0.5830189131200314,
      "num_tokens": 3061327.0,
      "step": 1550
    },
    {
      "entropy": 1.9043468147516251,
      "epoch": 0.04468441630981195,
      "grad_norm": 11.9375,
      "learning_rate": 4.999186176942271e-05,
      "loss": 2.016738510131836,
      "mean_token_accuracy": 0.5949485957622528,
      "num_tokens": 3080448.0,
      "step": 1560
    },
    {
      "entropy": 2.0285617001354694,
      "epoch": 0.04497085487590049,
      "grad_norm": 14.375,
      "learning_rate": 4.999156801050542e-05,
      "loss": 2.1571699142456056,
      "mean_token_accuracy": 0.5611587069928646,
      "num_tokens": 3099464.0,
      "step": 1570
    },
    {
      "entropy": 2.032850581407547,
      "epoch": 0.04525729344198903,
      "grad_norm": 11.8125,
      "learning_rate": 4.999126904427823e-05,
      "loss": 2.142729568481445,
      "mean_token_accuracy": 0.5683363012969493,
      "num_tokens": 3119445.0,
      "step": 1580
    },
    {
      "entropy": 1.9459272086620332,
      "epoch": 0.04554373200807757,
      "grad_norm": 17.625,
      "learning_rate": 4.999096487080346e-05,
      "loss": 2.0355464935302736,
      "mean_token_accuracy": 0.5715726802125574,
      "num_tokens": 3139277.0,
      "step": 1590
    },
    {
      "entropy": 1.9185783863067627,
      "epoch": 0.04583017057416611,
      "grad_norm": 12.8125,
      "learning_rate": 4.9990655490144464e-05,
      "loss": 2.042800712585449,
      "mean_token_accuracy": 0.5798308115452528,
      "num_tokens": 3158123.0,
      "step": 1600
    },
    {
      "entropy": 2.0223012883216143,
      "epoch": 0.046116609140254644,
      "grad_norm": 12.625,
      "learning_rate": 4.9990340902365724e-05,
      "loss": 2.1582475662231446,
      "mean_token_accuracy": 0.5669877097010613,
      "num_tokens": 3178764.0,
      "step": 1610
    },
    {
      "entropy": 1.984189174324274,
      "epoch": 0.04640304770634318,
      "grad_norm": 15.4375,
      "learning_rate": 4.9990021107532776e-05,
      "loss": 2.1012454986572267,
      "mean_token_accuracy": 0.5656917076557875,
      "num_tokens": 3197906.0,
      "step": 1620
    },
    {
      "entropy": 1.9899665638804436,
      "epoch": 0.04668948627243172,
      "grad_norm": 10.625,
      "learning_rate": 4.9989696105712255e-05,
      "loss": 2.054644966125488,
      "mean_token_accuracy": 0.565249053388834,
      "num_tokens": 3217491.0,
      "step": 1630
    },
    {
      "entropy": 2.173365142196417,
      "epoch": 0.046975924838520255,
      "grad_norm": 28.875,
      "learning_rate": 4.9989365896971886e-05,
      "loss": 2.27383975982666,
      "mean_token_accuracy": 0.5297906056046486,
      "num_tokens": 3237429.0,
      "step": 1640
    },
    {
      "entropy": 2.0662153884768486,
      "epoch": 0.0472623634046088,
      "grad_norm": 20.375,
      "learning_rate": 4.998903048138046e-05,
      "loss": 2.1620344161987304,
      "mean_token_accuracy": 0.5489458622410893,
      "num_tokens": 3256776.0,
      "step": 1650
    },
    {
      "entropy": 1.9505906507372857,
      "epoch": 0.04754880197069734,
      "grad_norm": 13.1875,
      "learning_rate": 4.998868985900788e-05,
      "loss": 2.0178153991699217,
      "mean_token_accuracy": 0.5674532245844602,
      "num_tokens": 3275791.0,
      "step": 1660
    },
    {
      "entropy": 1.966233441978693,
      "epoch": 0.047835240536785874,
      "grad_norm": 10.75,
      "learning_rate": 4.998834402992511e-05,
      "loss": 2.088430404663086,
      "mean_token_accuracy": 0.5653650652617216,
      "num_tokens": 3294501.0,
      "step": 1670
    },
    {
      "entropy": 1.8438799947500228,
      "epoch": 0.04812167910287441,
      "grad_norm": 13.8125,
      "learning_rate": 4.9987992994204205e-05,
      "loss": 1.9424402236938476,
      "mean_token_accuracy": 0.5878816403448581,
      "num_tokens": 3314620.0,
      "step": 1680
    },
    {
      "entropy": 1.8855835229158402,
      "epoch": 0.04840811766896295,
      "grad_norm": 12.25,
      "learning_rate": 4.9987636751918324e-05,
      "loss": 2.058888053894043,
      "mean_token_accuracy": 0.5803608868271113,
      "num_tokens": 3334977.0,
      "step": 1690
    },
    {
      "entropy": 1.9706033617258072,
      "epoch": 0.048694556235051485,
      "grad_norm": 10.4375,
      "learning_rate": 4.998727530314167e-05,
      "loss": 1.9991622924804688,
      "mean_token_accuracy": 0.5765298075973988,
      "num_tokens": 3358046.0,
      "step": 1700
    },
    {
      "entropy": 2.046942899376154,
      "epoch": 0.04898099480114002,
      "grad_norm": 11.5,
      "learning_rate": 4.998690864794957e-05,
      "loss": 2.2661470413208007,
      "mean_token_accuracy": 0.5470717648044229,
      "num_tokens": 3379088.0,
      "step": 1710
    },
    {
      "entropy": 2.120621448010206,
      "epoch": 0.049267433367228566,
      "grad_norm": 11.1875,
      "learning_rate": 4.998653678641843e-05,
      "loss": 2.117679405212402,
      "mean_token_accuracy": 0.5618710162118077,
      "num_tokens": 3398602.0,
      "step": 1720
    },
    {
      "entropy": 1.8022576607763767,
      "epoch": 0.049553871933317103,
      "grad_norm": 11.6875,
      "learning_rate": 4.998615971862571e-05,
      "loss": 1.9418848037719727,
      "mean_token_accuracy": 0.5938253000378608,
      "num_tokens": 3417782.0,
      "step": 1730
    },
    {
      "entropy": 1.7548723980784415,
      "epoch": 0.04984031049940564,
      "grad_norm": 16.125,
      "learning_rate": 4.9985777444649995e-05,
      "loss": 1.989464569091797,
      "mean_token_accuracy": 0.5836741965264082,
      "num_tokens": 3437040.0,
      "step": 1740
    },
    {
      "entropy": 1.889078352600336,
      "epoch": 0.05012674906549418,
      "grad_norm": 12.5625,
      "learning_rate": 4.9985389964570926e-05,
      "loss": 1.8748939514160157,
      "mean_token_accuracy": 0.6075956784188747,
      "num_tokens": 3455855.0,
      "step": 1750
    },
    {
      "entropy": 1.8022679753601551,
      "epoch": 0.050413187631582715,
      "grad_norm": 14.375,
      "learning_rate": 4.998499727846924e-05,
      "loss": 1.9469636917114257,
      "mean_token_accuracy": 0.5890525598078966,
      "num_tokens": 3475979.0,
      "step": 1760
    },
    {
      "entropy": 1.825092475116253,
      "epoch": 0.05069962619767125,
      "grad_norm": 14.875,
      "learning_rate": 4.9984599386426764e-05,
      "loss": 1.9489648818969727,
      "mean_token_accuracy": 0.5951741103082895,
      "num_tokens": 3494650.0,
      "step": 1770
    },
    {
      "entropy": 2.0946372151374817,
      "epoch": 0.050986064763759796,
      "grad_norm": 14.125,
      "learning_rate": 4.998419628852641e-05,
      "loss": 2.222842979431152,
      "mean_token_accuracy": 0.5530176166445017,
      "num_tokens": 3512392.0,
      "step": 1780
    },
    {
      "entropy": 1.986116524785757,
      "epoch": 0.05127250332984833,
      "grad_norm": 16.5,
      "learning_rate": 4.998378798485215e-05,
      "loss": 2.031750297546387,
      "mean_token_accuracy": 0.5753281008452177,
      "num_tokens": 3531841.0,
      "step": 1790
    },
    {
      "entropy": 1.8011541001498699,
      "epoch": 0.05155894189593687,
      "grad_norm": 12.4375,
      "learning_rate": 4.9983374475489076e-05,
      "loss": 1.9964473724365235,
      "mean_token_accuracy": 0.5887253880500793,
      "num_tokens": 3551425.0,
      "step": 1800
    },
    {
      "entropy": 1.83894300237298,
      "epoch": 0.05184538046202541,
      "grad_norm": 12.0625,
      "learning_rate": 4.998295576052334e-05,
      "loss": 1.9472723007202148,
      "mean_token_accuracy": 0.5949041478335857,
      "num_tokens": 3572169.0,
      "step": 1810
    },
    {
      "entropy": 1.820378828048706,
      "epoch": 0.052131819028113945,
      "grad_norm": 12.25,
      "learning_rate": 4.998253184004219e-05,
      "loss": 1.8152381896972656,
      "mean_token_accuracy": 0.6049352586269379,
      "num_tokens": 3591954.0,
      "step": 1820
    },
    {
      "entropy": 1.6917552098631858,
      "epoch": 0.05241825759420248,
      "grad_norm": 13.8125,
      "learning_rate": 4.998210271413395e-05,
      "loss": 1.8625791549682618,
      "mean_token_accuracy": 0.6033434379845858,
      "num_tokens": 3612784.0,
      "step": 1830
    },
    {
      "entropy": 1.9794663526117802,
      "epoch": 0.05270469616029102,
      "grad_norm": 13.25,
      "learning_rate": 4.998166838288803e-05,
      "loss": 2.1264877319335938,
      "mean_token_accuracy": 0.5757445959374309,
      "num_tokens": 3631909.0,
      "step": 1840
    },
    {
      "entropy": 2.042566266655922,
      "epoch": 0.05299113472637956,
      "grad_norm": 14.375,
      "learning_rate": 4.998122884639495e-05,
      "loss": 2.0641660690307617,
      "mean_token_accuracy": 0.5762186653912067,
      "num_tokens": 3651541.0,
      "step": 1850
    },
    {
      "entropy": 1.9482962124049663,
      "epoch": 0.0532775732924681,
      "grad_norm": 9.6875,
      "learning_rate": 4.998078410474626e-05,
      "loss": 2.0754838943481446,
      "mean_token_accuracy": 0.5959542846307159,
      "num_tokens": 3671637.0,
      "step": 1860
    },
    {
      "entropy": 1.772421895712614,
      "epoch": 0.05356401185855664,
      "grad_norm": 10.4375,
      "learning_rate": 4.998033415803466e-05,
      "loss": 1.9321874618530273,
      "mean_token_accuracy": 0.5985866695642471,
      "num_tokens": 3691393.0,
      "step": 1870
    },
    {
      "entropy": 1.9630630321800708,
      "epoch": 0.053850450424645174,
      "grad_norm": 11.3125,
      "learning_rate": 4.997987900635388e-05,
      "loss": 2.095884323120117,
      "mean_token_accuracy": 0.5736457588151097,
      "num_tokens": 3711294.0,
      "step": 1880
    },
    {
      "entropy": 2.044461357593536,
      "epoch": 0.05413688899073371,
      "grad_norm": 13.8125,
      "learning_rate": 4.997941864979877e-05,
      "loss": 2.099529838562012,
      "mean_token_accuracy": 0.5731991751119494,
      "num_tokens": 3731939.0,
      "step": 1890
    },
    {
      "entropy": 1.8967493332922458,
      "epoch": 0.05442332755682225,
      "grad_norm": 21.75,
      "learning_rate": 4.997895308846524e-05,
      "loss": 2.038409614562988,
      "mean_token_accuracy": 0.5800403725355864,
      "num_tokens": 3750220.0,
      "step": 1900
    },
    {
      "entropy": 1.9163446389138699,
      "epoch": 0.054709766122910786,
      "grad_norm": 9.9375,
      "learning_rate": 4.9978482322450305e-05,
      "loss": 1.9895086288452148,
      "mean_token_accuracy": 0.5872535329312086,
      "num_tokens": 3769972.0,
      "step": 1910
    },
    {
      "entropy": 1.8457116827368736,
      "epoch": 0.05499620468899933,
      "grad_norm": 12.3125,
      "learning_rate": 4.997800635185205e-05,
      "loss": 1.910600471496582,
      "mean_token_accuracy": 0.6011968698352576,
      "num_tokens": 3789479.0,
      "step": 1920
    },
    {
      "entropy": 1.8551744908094405,
      "epoch": 0.05528264325508787,
      "grad_norm": 15.0,
      "learning_rate": 4.997752517676965e-05,
      "loss": 1.999060821533203,
      "mean_token_accuracy": 0.5846864935010672,
      "num_tokens": 3808632.0,
      "step": 1930
    },
    {
      "entropy": 1.9533005893230437,
      "epoch": 0.055569081821176404,
      "grad_norm": 11.1875,
      "learning_rate": 4.997703879730337e-05,
      "loss": 1.9968469619750977,
      "mean_token_accuracy": 0.58026770260185,
      "num_tokens": 3827910.0,
      "step": 1940
    },
    {
      "entropy": 1.9085161589086055,
      "epoch": 0.05585552038726494,
      "grad_norm": 8.8125,
      "learning_rate": 4.9976547213554536e-05,
      "loss": 2.019079399108887,
      "mean_token_accuracy": 0.5866671945899725,
      "num_tokens": 3848051.0,
      "step": 1950
    },
    {
      "entropy": 1.8469381153583526,
      "epoch": 0.05614195895335348,
      "grad_norm": 11.625,
      "learning_rate": 4.99760504256256e-05,
      "loss": 1.9547319412231445,
      "mean_token_accuracy": 0.5960608791559935,
      "num_tokens": 3867006.0,
      "step": 1960
    },
    {
      "entropy": 1.9383334703743458,
      "epoch": 0.056428397519442015,
      "grad_norm": 18.5,
      "learning_rate": 4.997554843362006e-05,
      "loss": 2.133906936645508,
      "mean_token_accuracy": 0.5758546967059374,
      "num_tokens": 3886840.0,
      "step": 1970
    },
    {
      "entropy": 1.9013036232441665,
      "epoch": 0.05671483608553055,
      "grad_norm": 10.875,
      "learning_rate": 4.997504123764251e-05,
      "loss": 1.9568716049194337,
      "mean_token_accuracy": 0.5916589047759772,
      "num_tokens": 3907260.0,
      "step": 1980
    },
    {
      "entropy": 1.8071703672409059,
      "epoch": 0.0570012746516191,
      "grad_norm": 8.75,
      "learning_rate": 4.997452883779863e-05,
      "loss": 1.8723196029663085,
      "mean_token_accuracy": 0.6036541979759932,
      "num_tokens": 3926810.0,
      "step": 1990
    },
    {
      "entropy": 1.8475841164588929,
      "epoch": 0.057287713217707634,
      "grad_norm": 12.0625,
      "learning_rate": 4.99740112341952e-05,
      "loss": 2.0067562103271483,
      "mean_token_accuracy": 0.5976184494793415,
      "num_tokens": 3945623.0,
      "step": 2000
    },
    {
      "entropy": 1.7979756787419319,
      "epoch": 0.05757415178379617,
      "grad_norm": 12.4375,
      "learning_rate": 4.997348842694005e-05,
      "loss": 1.8895217895507812,
      "mean_token_accuracy": 0.5975389555096626,
      "num_tokens": 3966359.0,
      "step": 2010
    },
    {
      "entropy": 1.889492254704237,
      "epoch": 0.05786059034988471,
      "grad_norm": 11.875,
      "learning_rate": 4.997296041614213e-05,
      "loss": 1.9238414764404297,
      "mean_token_accuracy": 0.5961540397256613,
      "num_tokens": 3985098.0,
      "step": 2020
    },
    {
      "entropy": 1.7875510536134243,
      "epoch": 0.058147028915973245,
      "grad_norm": 11.0625,
      "learning_rate": 4.9972427201911455e-05,
      "loss": 1.9333005905151368,
      "mean_token_accuracy": 0.5928152399137616,
      "num_tokens": 4004811.0,
      "step": 2030
    },
    {
      "entropy": 1.7978779956698419,
      "epoch": 0.05843346748206178,
      "grad_norm": 10.8125,
      "learning_rate": 4.997188878435911e-05,
      "loss": 1.9755786895751952,
      "mean_token_accuracy": 0.5879885520786047,
      "num_tokens": 4023985.0,
      "step": 2040
    },
    {
      "entropy": 1.8877509631216527,
      "epoch": 0.058719906048150326,
      "grad_norm": 9.6875,
      "learning_rate": 4.99713451635973e-05,
      "loss": 1.9203781127929687,
      "mean_token_accuracy": 0.5888435672968626,
      "num_tokens": 4043964.0,
      "step": 2050
    },
    {
      "entropy": 1.9123047180473804,
      "epoch": 0.05900634461423886,
      "grad_norm": 12.8125,
      "learning_rate": 4.997079633973929e-05,
      "loss": 1.9715444564819335,
      "mean_token_accuracy": 0.5856408469378949,
      "num_tokens": 4063465.0,
      "step": 2060
    },
    {
      "entropy": 1.829633817076683,
      "epoch": 0.0592927831803274,
      "grad_norm": 10.4375,
      "learning_rate": 4.997024231289943e-05,
      "loss": 1.9249038696289062,
      "mean_token_accuracy": 0.5929070865735412,
      "num_tokens": 4082738.0,
      "step": 2070
    },
    {
      "entropy": 2.029878391325474,
      "epoch": 0.05957922174641594,
      "grad_norm": 10.5625,
      "learning_rate": 4.996968308319316e-05,
      "loss": 2.1039791107177734,
      "mean_token_accuracy": 0.5622795660048723,
      "num_tokens": 4101747.0,
      "step": 2080
    },
    {
      "entropy": 1.8725275248289108,
      "epoch": 0.059865660312504475,
      "grad_norm": 11.75,
      "learning_rate": 4.9969118650737004e-05,
      "loss": 1.906928253173828,
      "mean_token_accuracy": 0.6090848710387945,
      "num_tokens": 4121976.0,
      "step": 2090
    },
    {
      "entropy": 1.9075307242572308,
      "epoch": 0.06015209887859301,
      "grad_norm": 11.875,
      "learning_rate": 4.9968549015648576e-05,
      "loss": 2.0786325454711916,
      "mean_token_accuracy": 0.5623055983334779,
      "num_tokens": 4140873.0,
      "step": 2100
    },
    {
      "entropy": 1.7441140189766884,
      "epoch": 0.06043853744468155,
      "grad_norm": 11.625,
      "learning_rate": 4.996797417804656e-05,
      "loss": 1.8118831634521484,
      "mean_token_accuracy": 0.6088214486837387,
      "num_tokens": 4159590.0,
      "step": 2110
    },
    {
      "entropy": 1.8769513357430696,
      "epoch": 0.06072497601077009,
      "grad_norm": 11.25,
      "learning_rate": 4.996739413805072e-05,
      "loss": 1.9912311553955078,
      "mean_token_accuracy": 0.5811572030186654,
      "num_tokens": 4180937.0,
      "step": 2120
    },
    {
      "entropy": 1.9944736510515213,
      "epoch": 0.06101141457685863,
      "grad_norm": 18.625,
      "learning_rate": 4.996680889578193e-05,
      "loss": 2.0236806869506836,
      "mean_token_accuracy": 0.5700880959630013,
      "num_tokens": 4200933.0,
      "step": 2130
    },
    {
      "entropy": 1.6624108001589775,
      "epoch": 0.06129785314294717,
      "grad_norm": 12.875,
      "learning_rate": 4.996621845136213e-05,
      "loss": 1.8008182525634766,
      "mean_token_accuracy": 0.6165046537294984,
      "num_tokens": 4220649.0,
      "step": 2140
    },
    {
      "entropy": 1.9939217180013658,
      "epoch": 0.061584291709035704,
      "grad_norm": 13.75,
      "learning_rate": 4.9965622804914344e-05,
      "loss": 2.1022857666015624,
      "mean_token_accuracy": 0.5655762797221542,
      "num_tokens": 4240827.0,
      "step": 2150
    },
    {
      "entropy": 1.881871872395277,
      "epoch": 0.06187073027512424,
      "grad_norm": 9.375,
      "learning_rate": 4.9965021956562674e-05,
      "loss": 1.9296346664428712,
      "mean_token_accuracy": 0.6011112701147795,
      "num_tokens": 4262156.0,
      "step": 2160
    },
    {
      "entropy": 1.8640183731913567,
      "epoch": 0.06215716884121278,
      "grad_norm": 11.4375,
      "learning_rate": 4.996441590643233e-05,
      "loss": 2.024766540527344,
      "mean_token_accuracy": 0.5818479772657156,
      "num_tokens": 4281067.0,
      "step": 2170
    },
    {
      "entropy": 1.877857768535614,
      "epoch": 0.062443607407301316,
      "grad_norm": 9.3125,
      "learning_rate": 4.996380465464959e-05,
      "loss": 1.973579216003418,
      "mean_token_accuracy": 0.6007902212440968,
      "num_tokens": 4302265.0,
      "step": 2180
    },
    {
      "entropy": 1.8153230141848327,
      "epoch": 0.06273004597338985,
      "grad_norm": 13.1875,
      "learning_rate": 4.9963188201341795e-05,
      "loss": 1.9394966125488282,
      "mean_token_accuracy": 0.6073149384930729,
      "num_tokens": 4320549.0,
      "step": 2190
    },
    {
      "entropy": 1.947351810336113,
      "epoch": 0.06301648453947839,
      "grad_norm": 12.375,
      "learning_rate": 4.996256654663742e-05,
      "loss": 1.9659574508666993,
      "mean_token_accuracy": 0.5926310716196894,
      "num_tokens": 4341293.0,
      "step": 2200
    },
    {
      "entropy": 1.815236010402441,
      "epoch": 0.06330292310556693,
      "grad_norm": 7.96875,
      "learning_rate": 4.996193969066597e-05,
      "loss": 1.9169052124023438,
      "mean_token_accuracy": 0.5960832234472037,
      "num_tokens": 4362117.0,
      "step": 2210
    },
    {
      "entropy": 1.8823579348623753,
      "epoch": 0.06358936167165548,
      "grad_norm": 16.5,
      "learning_rate": 4.996130763355808e-05,
      "loss": 1.961000633239746,
      "mean_token_accuracy": 0.5924124475568533,
      "num_tokens": 4382054.0,
      "step": 2220
    },
    {
      "entropy": 1.7910111367702484,
      "epoch": 0.06387580023774402,
      "grad_norm": 12.8125,
      "learning_rate": 4.996067037544542e-05,
      "loss": 1.8757230758666992,
      "mean_token_accuracy": 0.597107907012105,
      "num_tokens": 4400596.0,
      "step": 2230
    },
    {
      "entropy": 1.8231641046702862,
      "epoch": 0.06416223880383255,
      "grad_norm": 11.6875,
      "learning_rate": 4.9960027916460785e-05,
      "loss": 1.9751651763916016,
      "mean_token_accuracy": 0.601635891199112,
      "num_tokens": 4419627.0,
      "step": 2240
    },
    {
      "entropy": 1.837186548113823,
      "epoch": 0.06444867736992109,
      "grad_norm": 10.5625,
      "learning_rate": 4.995938025673805e-05,
      "loss": 1.8203306198120117,
      "mean_token_accuracy": 0.6274783156812191,
      "num_tokens": 4439004.0,
      "step": 2250
    },
    {
      "entropy": 1.7538263626396655,
      "epoch": 0.06473511593600963,
      "grad_norm": 9.3125,
      "learning_rate": 4.995872739641215e-05,
      "loss": 1.9434621810913086,
      "mean_token_accuracy": 0.5958845272660256,
      "num_tokens": 4460721.0,
      "step": 2260
    },
    {
      "entropy": 1.823099322617054,
      "epoch": 0.06502155450209816,
      "grad_norm": 9.75,
      "learning_rate": 4.995806933561912e-05,
      "loss": 1.834269905090332,
      "mean_token_accuracy": 0.6187158718705177,
      "num_tokens": 4480791.0,
      "step": 2270
    },
    {
      "entropy": 1.8298480186611414,
      "epoch": 0.0653079930681867,
      "grad_norm": 11.375,
      "learning_rate": 4.995740607449608e-05,
      "loss": 1.912055206298828,
      "mean_token_accuracy": 0.5971175793558359,
      "num_tokens": 4500673.0,
      "step": 2280
    },
    {
      "entropy": 1.8175236236304044,
      "epoch": 0.06559443163427524,
      "grad_norm": 10.25,
      "learning_rate": 4.995673761318123e-05,
      "loss": 1.8878864288330077,
      "mean_token_accuracy": 0.6076633669435978,
      "num_tokens": 4519884.0,
      "step": 2290
    },
    {
      "entropy": 1.757424382865429,
      "epoch": 0.06588087020036378,
      "grad_norm": 14.5625,
      "learning_rate": 4.9956063951813836e-05,
      "loss": 1.8500144958496094,
      "mean_token_accuracy": 0.6129124265164136,
      "num_tokens": 4539975.0,
      "step": 2300
    },
    {
      "entropy": 1.7447815477848052,
      "epoch": 0.06616730876645231,
      "grad_norm": 12.6875,
      "learning_rate": 4.995538509053428e-05,
      "loss": 1.9476245880126952,
      "mean_token_accuracy": 0.5854702297598123,
      "num_tokens": 4559907.0,
      "step": 2310
    },
    {
      "entropy": 1.7929293140769005,
      "epoch": 0.06645374733254085,
      "grad_norm": 16.375,
      "learning_rate": 4.9954701029484e-05,
      "loss": 1.8017602920532227,
      "mean_token_accuracy": 0.6247454103082418,
      "num_tokens": 4579395.0,
      "step": 2320
    },
    {
      "entropy": 1.840090711414814,
      "epoch": 0.06674018589862939,
      "grad_norm": 8.6875,
      "learning_rate": 4.9954011768805545e-05,
      "loss": 1.9241155624389648,
      "mean_token_accuracy": 0.5975760428234935,
      "num_tokens": 4598245.0,
      "step": 2330
    },
    {
      "entropy": 1.8547099135816096,
      "epoch": 0.06702662446471792,
      "grad_norm": 8.3125,
      "learning_rate": 4.995331730864251e-05,
      "loss": 2.01320858001709,
      "mean_token_accuracy": 0.5943611957132816,
      "num_tokens": 4618823.0,
      "step": 2340
    },
    {
      "entropy": 2.012850197404623,
      "epoch": 0.06731306303080646,
      "grad_norm": 9.4375,
      "learning_rate": 4.995261764913961e-05,
      "loss": 1.9893373489379882,
      "mean_token_accuracy": 0.5957473564893008,
      "num_tokens": 4637793.0,
      "step": 2350
    },
    {
      "entropy": 1.76784438341856,
      "epoch": 0.06759950159689501,
      "grad_norm": 10.1875,
      "learning_rate": 4.995191279044262e-05,
      "loss": 1.9172800064086915,
      "mean_token_accuracy": 0.5817941173911094,
      "num_tokens": 4657814.0,
      "step": 2360
    },
    {
      "entropy": 1.7323622293770313,
      "epoch": 0.06788594016298355,
      "grad_norm": 14.625,
      "learning_rate": 4.995120273269842e-05,
      "loss": 1.8140583038330078,
      "mean_token_accuracy": 0.6260790122672916,
      "num_tokens": 4676346.0,
      "step": 2370
    },
    {
      "entropy": 1.83290024548769,
      "epoch": 0.06817237872907209,
      "grad_norm": 11.375,
      "learning_rate": 4.995048747605494e-05,
      "loss": 1.9392313003540038,
      "mean_token_accuracy": 0.588176003843546,
      "num_tokens": 4694953.0,
      "step": 2380
    },
    {
      "entropy": 1.8222525171935557,
      "epoch": 0.06845881729516062,
      "grad_norm": 10.9375,
      "learning_rate": 4.994976702066123e-05,
      "loss": 1.8721160888671875,
      "mean_token_accuracy": 0.5986452177166939,
      "num_tokens": 4714501.0,
      "step": 2390
    },
    {
      "entropy": 1.7743729643523694,
      "epoch": 0.06874525586124916,
      "grad_norm": 9.5625,
      "learning_rate": 4.9949041366667394e-05,
      "loss": 1.8765153884887695,
      "mean_token_accuracy": 0.6184959661215543,
      "num_tokens": 4734392.0,
      "step": 2400
    },
    {
      "entropy": 1.8153611175715922,
      "epoch": 0.0690316944273377,
      "grad_norm": 15.125,
      "learning_rate": 4.994831051422464e-05,
      "loss": 1.9120668411254882,
      "mean_token_accuracy": 0.5987484376877547,
      "num_tokens": 4754689.0,
      "step": 2410
    },
    {
      "entropy": 1.846549166738987,
      "epoch": 0.06931813299342623,
      "grad_norm": 12.5,
      "learning_rate": 4.994757446348524e-05,
      "loss": 1.9954065322875976,
      "mean_token_accuracy": 0.5823904737830162,
      "num_tokens": 4774115.0,
      "step": 2420
    },
    {
      "entropy": 1.7396523296833037,
      "epoch": 0.06960457155951477,
      "grad_norm": 8.9375,
      "learning_rate": 4.9946833214602565e-05,
      "loss": 1.7574817657470703,
      "mean_token_accuracy": 0.6228158086538315,
      "num_tokens": 4795914.0,
      "step": 2430
    },
    {
      "entropy": 1.7804276674985886,
      "epoch": 0.06989101012560331,
      "grad_norm": 12.9375,
      "learning_rate": 4.994608676773107e-05,
      "loss": 1.8207117080688477,
      "mean_token_accuracy": 0.617691895738244,
      "num_tokens": 4815550.0,
      "step": 2440
    },
    {
      "entropy": 1.762260028719902,
      "epoch": 0.07017744869169185,
      "grad_norm": 10.875,
      "learning_rate": 4.994533512302627e-05,
      "loss": 1.8566402435302733,
      "mean_token_accuracy": 0.5982673525810241,
      "num_tokens": 4834642.0,
      "step": 2450
    },
    {
      "entropy": 1.671261265128851,
      "epoch": 0.07046388725778038,
      "grad_norm": 10.5,
      "learning_rate": 4.99445782806448e-05,
      "loss": 1.8638725280761719,
      "mean_token_accuracy": 0.6134916298091412,
      "num_tokens": 4854008.0,
      "step": 2460
    },
    {
      "entropy": 1.8840367905795574,
      "epoch": 0.07075032582386892,
      "grad_norm": 9.8125,
      "learning_rate": 4.9943816240744346e-05,
      "loss": 1.9149871826171876,
      "mean_token_accuracy": 0.5981070324778557,
      "num_tokens": 4873434.0,
      "step": 2470
    },
    {
      "entropy": 1.8419511586427688,
      "epoch": 0.07103676438995746,
      "grad_norm": 9.625,
      "learning_rate": 4.994304900348369e-05,
      "loss": 1.837874412536621,
      "mean_token_accuracy": 0.6224506478756666,
      "num_tokens": 4893087.0,
      "step": 2480
    },
    {
      "entropy": 1.7017809920012952,
      "epoch": 0.07132320295604601,
      "grad_norm": 10.4375,
      "learning_rate": 4.994227656902269e-05,
      "loss": 1.838263702392578,
      "mean_token_accuracy": 0.6109026603400707,
      "num_tokens": 4912363.0,
      "step": 2490
    },
    {
      "entropy": 1.828787763416767,
      "epoch": 0.07160964152213455,
      "grad_norm": 9.25,
      "learning_rate": 4.994149893752231e-05,
      "loss": 2.001331329345703,
      "mean_token_accuracy": 0.5902357619255781,
      "num_tokens": 4931432.0,
      "step": 2500
    },
    {
      "entropy": 1.7498065806925296,
      "epoch": 0.07189608008822308,
      "grad_norm": 9.5625,
      "learning_rate": 4.994071610914456e-05,
      "loss": 1.685750389099121,
      "mean_token_accuracy": 0.6296104624867439,
      "num_tokens": 4950916.0,
      "step": 2510
    },
    {
      "entropy": 1.7557595752179622,
      "epoch": 0.07218251865431162,
      "grad_norm": 13.8125,
      "learning_rate": 4.993992808405256e-05,
      "loss": 1.9015703201293945,
      "mean_token_accuracy": 0.6091135945171118,
      "num_tokens": 4970433.0,
      "step": 2520
    },
    {
      "entropy": 1.7408047266304494,
      "epoch": 0.07246895722040016,
      "grad_norm": 11.0,
      "learning_rate": 4.993913486241051e-05,
      "loss": 1.8769393920898438,
      "mean_token_accuracy": 0.6036487333476543,
      "num_tokens": 4990769.0,
      "step": 2530
    },
    {
      "entropy": 1.8650392189621925,
      "epoch": 0.0727553957864887,
      "grad_norm": 12.5,
      "learning_rate": 4.993833644438368e-05,
      "loss": 1.8982149124145509,
      "mean_token_accuracy": 0.6078139618039131,
      "num_tokens": 5011191.0,
      "step": 2540
    },
    {
      "entropy": 1.6541506234556436,
      "epoch": 0.07304183435257723,
      "grad_norm": 10.1875,
      "learning_rate": 4.9937532830138436e-05,
      "loss": 1.6678089141845702,
      "mean_token_accuracy": 0.6447409730404615,
      "num_tokens": 5032349.0,
      "step": 2550
    },
    {
      "entropy": 1.681502054259181,
      "epoch": 0.07332827291866577,
      "grad_norm": 14.625,
      "learning_rate": 4.993672401984222e-05,
      "loss": 1.8666999816894532,
      "mean_token_accuracy": 0.6053219188004733,
      "num_tokens": 5053087.0,
      "step": 2560
    },
    {
      "entropy": 1.5955466687679292,
      "epoch": 0.0736147114847543,
      "grad_norm": 13.125,
      "learning_rate": 4.9935910013663554e-05,
      "loss": 1.631696891784668,
      "mean_token_accuracy": 0.6485393714159727,
      "num_tokens": 5072577.0,
      "step": 2570
    },
    {
      "entropy": 1.7683854140341282,
      "epoch": 0.07390115005084284,
      "grad_norm": 11.75,
      "learning_rate": 4.993509081177205e-05,
      "loss": 1.8348590850830078,
      "mean_token_accuracy": 0.6120606031268835,
      "num_tokens": 5091809.0,
      "step": 2580
    },
    {
      "entropy": 1.64336349517107,
      "epoch": 0.07418758861693138,
      "grad_norm": 11.75,
      "learning_rate": 4.99342664143384e-05,
      "loss": 1.7510248184204102,
      "mean_token_accuracy": 0.6331697985529899,
      "num_tokens": 5111043.0,
      "step": 2590
    },
    {
      "entropy": 1.8232658706605434,
      "epoch": 0.07447402718301992,
      "grad_norm": 8.9375,
      "learning_rate": 4.993343682153437e-05,
      "loss": 1.8883161544799805,
      "mean_token_accuracy": 0.6126664072275162,
      "num_tokens": 5130766.0,
      "step": 2600
    },
    {
      "entropy": 1.713892677426338,
      "epoch": 0.07476046574910845,
      "grad_norm": 17.75,
      "learning_rate": 4.9932602033532824e-05,
      "loss": 1.7505222320556642,
      "mean_token_accuracy": 0.6276527978479862,
      "num_tokens": 5150625.0,
      "step": 2610
    },
    {
      "entropy": 1.637095182389021,
      "epoch": 0.075046904315197,
      "grad_norm": 11.3125,
      "learning_rate": 4.9931762050507705e-05,
      "loss": 1.7237468719482423,
      "mean_token_accuracy": 0.6427792329341173,
      "num_tokens": 5169898.0,
      "step": 2620
    },
    {
      "entropy": 1.7508811518549918,
      "epoch": 0.07533334288128554,
      "grad_norm": 9.3125,
      "learning_rate": 4.993091687263403e-05,
      "loss": 1.907511329650879,
      "mean_token_accuracy": 0.6170182578265667,
      "num_tokens": 5190039.0,
      "step": 2630
    },
    {
      "entropy": 1.8437795139849187,
      "epoch": 0.07561978144737408,
      "grad_norm": 12.3125,
      "learning_rate": 4.993006650008789e-05,
      "loss": 1.8318609237670898,
      "mean_token_accuracy": 0.6023694450035691,
      "num_tokens": 5210247.0,
      "step": 2640
    },
    {
      "entropy": 1.6756767228245735,
      "epoch": 0.07590622001346262,
      "grad_norm": 10.5625,
      "learning_rate": 4.992921093304649e-05,
      "loss": 1.814997100830078,
      "mean_token_accuracy": 0.6139365989714861,
      "num_tokens": 5229421.0,
      "step": 2650
    },
    {
      "entropy": 1.7465621680021286,
      "epoch": 0.07619265857955115,
      "grad_norm": 14.1875,
      "learning_rate": 4.992835017168809e-05,
      "loss": 1.7556392669677734,
      "mean_token_accuracy": 0.6323899663984776,
      "num_tokens": 5248920.0,
      "step": 2660
    },
    {
      "entropy": 1.6148886807262897,
      "epoch": 0.07647909714563969,
      "grad_norm": 12.3125,
      "learning_rate": 4.9927484216192036e-05,
      "loss": 1.7459449768066406,
      "mean_token_accuracy": 0.6273711193352938,
      "num_tokens": 5268690.0,
      "step": 2670
    },
    {
      "entropy": 1.7831590503454209,
      "epoch": 0.07676553571172823,
      "grad_norm": 11.8125,
      "learning_rate": 4.9926613066738755e-05,
      "loss": 1.8846384048461915,
      "mean_token_accuracy": 0.594828499853611,
      "num_tokens": 5288176.0,
      "step": 2680
    },
    {
      "entropy": 1.7635230794548988,
      "epoch": 0.07705197427781676,
      "grad_norm": 12.875,
      "learning_rate": 4.992573672350979e-05,
      "loss": 1.8495376586914063,
      "mean_token_accuracy": 0.6183344263583421,
      "num_tokens": 5308317.0,
      "step": 2690
    },
    {
      "entropy": 1.746929007023573,
      "epoch": 0.0773384128439053,
      "grad_norm": 8.25,
      "learning_rate": 4.9924855186687705e-05,
      "loss": 1.816206169128418,
      "mean_token_accuracy": 0.6279907550662756,
      "num_tokens": 5327759.0,
      "step": 2700
    },
    {
      "entropy": 1.8997233137488365,
      "epoch": 0.07762485140999384,
      "grad_norm": 14.625,
      "learning_rate": 4.992396845645621e-05,
      "loss": 2.018978309631348,
      "mean_token_accuracy": 0.5832574477419257,
      "num_tokens": 5346568.0,
      "step": 2710
    },
    {
      "entropy": 1.7992750965058804,
      "epoch": 0.07791128997608238,
      "grad_norm": 11.3125,
      "learning_rate": 4.992307653300003e-05,
      "loss": 1.8099502563476562,
      "mean_token_accuracy": 0.6049119036644697,
      "num_tokens": 5365573.0,
      "step": 2720
    },
    {
      "entropy": 1.7343350291252135,
      "epoch": 0.07819772854217091,
      "grad_norm": 11.5625,
      "learning_rate": 4.9922179416505044e-05,
      "loss": 1.86175479888916,
      "mean_token_accuracy": 0.6017268471419811,
      "num_tokens": 5385460.0,
      "step": 2730
    },
    {
      "entropy": 1.8600797936320306,
      "epoch": 0.07848416710825945,
      "grad_norm": 10.0625,
      "learning_rate": 4.992127710715817e-05,
      "loss": 1.8717761993408204,
      "mean_token_accuracy": 0.604049276560545,
      "num_tokens": 5406548.0,
      "step": 2740
    },
    {
      "entropy": 1.6443809218704701,
      "epoch": 0.07877060567434799,
      "grad_norm": 9.0625,
      "learning_rate": 4.9920369605147394e-05,
      "loss": 1.7614917755126953,
      "mean_token_accuracy": 0.616263660043478,
      "num_tokens": 5426253.0,
      "step": 2750
    },
    {
      "entropy": 1.617201566696167,
      "epoch": 0.07905704424043654,
      "grad_norm": 9.0,
      "learning_rate": 4.991945691066183e-05,
      "loss": 1.7047906875610352,
      "mean_token_accuracy": 0.632831609621644,
      "num_tokens": 5445019.0,
      "step": 2760
    },
    {
      "entropy": 1.8553661078214645,
      "epoch": 0.07934348280652508,
      "grad_norm": 7.0625,
      "learning_rate": 4.991853902389163e-05,
      "loss": 1.940587043762207,
      "mean_token_accuracy": 0.6004931338131427,
      "num_tokens": 5466361.0,
      "step": 2770
    },
    {
      "entropy": 1.9773833863437176,
      "epoch": 0.07962992137261361,
      "grad_norm": 9.125,
      "learning_rate": 4.9917615945028065e-05,
      "loss": 2.0119298934936523,
      "mean_token_accuracy": 0.5790961310267448,
      "num_tokens": 5486732.0,
      "step": 2780
    },
    {
      "entropy": 1.6732374832034111,
      "epoch": 0.07991635993870215,
      "grad_norm": 11.375,
      "learning_rate": 4.991668767426346e-05,
      "loss": 1.7327560424804687,
      "mean_token_accuracy": 0.6331372715532779,
      "num_tokens": 5505057.0,
      "step": 2790
    },
    {
      "entropy": 1.6872285589575768,
      "epoch": 0.08020279850479069,
      "grad_norm": 9.3125,
      "learning_rate": 4.991575421179123e-05,
      "loss": 1.7661176681518556,
      "mean_token_accuracy": 0.6337466925382614,
      "num_tokens": 5524765.0,
      "step": 2800
    },
    {
      "entropy": 1.7016744412481786,
      "epoch": 0.08048923707087922,
      "grad_norm": 9.1875,
      "learning_rate": 4.991481555780589e-05,
      "loss": 1.7514514923095703,
      "mean_token_accuracy": 0.6232207585126162,
      "num_tokens": 5544812.0,
      "step": 2810
    },
    {
      "entropy": 1.7252783011645079,
      "epoch": 0.08077567563696776,
      "grad_norm": 7.34375,
      "learning_rate": 4.9913871712503004e-05,
      "loss": 1.8788469314575196,
      "mean_token_accuracy": 0.6168043747544288,
      "num_tokens": 5564982.0,
      "step": 2820
    },
    {
      "entropy": 1.627824354916811,
      "epoch": 0.0810621142030563,
      "grad_norm": 12.4375,
      "learning_rate": 4.991292267607923e-05,
      "loss": 1.6839054107666016,
      "mean_token_accuracy": 0.63603176176548,
      "num_tokens": 5584267.0,
      "step": 2830
    },
    {
      "entropy": 1.6540647119283676,
      "epoch": 0.08134855276914484,
      "grad_norm": 11.375,
      "learning_rate": 4.9911968448732324e-05,
      "loss": 1.6764236450195313,
      "mean_token_accuracy": 0.6303022615611553,
      "num_tokens": 5604387.0,
      "step": 2840
    },
    {
      "entropy": 1.6972127817571163,
      "epoch": 0.08163499133523337,
      "grad_norm": 16.375,
      "learning_rate": 4.991100903066112e-05,
      "loss": 1.7772232055664063,
      "mean_token_accuracy": 0.6358709968626499,
      "num_tokens": 5623185.0,
      "step": 2850
    },
    {
      "entropy": 1.7593320779502393,
      "epoch": 0.08192142990132191,
      "grad_norm": 12.75,
      "learning_rate": 4.9910044422065495e-05,
      "loss": 1.807419204711914,
      "mean_token_accuracy": 0.6225467007607222,
      "num_tokens": 5641660.0,
      "step": 2860
    },
    {
      "entropy": 1.6534669509157538,
      "epoch": 0.08220786846741045,
      "grad_norm": 8.8125,
      "learning_rate": 4.9909074623146454e-05,
      "loss": 1.7458194732666015,
      "mean_token_accuracy": 0.6239213116467,
      "num_tokens": 5660939.0,
      "step": 2870
    },
    {
      "entropy": 1.7411190815269948,
      "epoch": 0.08249430703349898,
      "grad_norm": 9.0,
      "learning_rate": 4.9908099634106084e-05,
      "loss": 1.8914546966552734,
      "mean_token_accuracy": 0.6045034378767014,
      "num_tokens": 5680371.0,
      "step": 2880
    },
    {
      "entropy": 1.8807879753410817,
      "epoch": 0.08278074559958754,
      "grad_norm": 13.5,
      "learning_rate": 4.990711945514751e-05,
      "loss": 1.9603790283203124,
      "mean_token_accuracy": 0.59328673183918,
      "num_tokens": 5700817.0,
      "step": 2890
    },
    {
      "entropy": 1.7275540959089994,
      "epoch": 0.08306718416567607,
      "grad_norm": 10.25,
      "learning_rate": 4.9906134086474975e-05,
      "loss": 1.8055536270141601,
      "mean_token_accuracy": 0.6184068739414215,
      "num_tokens": 5722036.0,
      "step": 2900
    },
    {
      "entropy": 1.7929349288344383,
      "epoch": 0.08335362273176461,
      "grad_norm": 8.625,
      "learning_rate": 4.9905143528293794e-05,
      "loss": 1.8911727905273437,
      "mean_token_accuracy": 0.6066431190818549,
      "num_tokens": 5743145.0,
      "step": 2910
    },
    {
      "entropy": 1.6432476229965687,
      "epoch": 0.08364006129785315,
      "grad_norm": 13.5625,
      "learning_rate": 4.990414778081037e-05,
      "loss": 1.5889147758483886,
      "mean_token_accuracy": 0.6647788506001234,
      "num_tokens": 5762275.0,
      "step": 2920
    },
    {
      "entropy": 1.7077577784657478,
      "epoch": 0.08392649986394168,
      "grad_norm": 9.0,
      "learning_rate": 4.990314684423216e-05,
      "loss": 1.9012144088745118,
      "mean_token_accuracy": 0.6113692533224822,
      "num_tokens": 5781647.0,
      "step": 2930
    },
    {
      "entropy": 1.8442125141620636,
      "epoch": 0.08421293843003022,
      "grad_norm": 10.25,
      "learning_rate": 4.990214071876774e-05,
      "loss": 1.9046855926513673,
      "mean_token_accuracy": 0.6001397427171469,
      "num_tokens": 5801299.0,
      "step": 2940
    },
    {
      "entropy": 1.8405231080949307,
      "epoch": 0.08449937699611876,
      "grad_norm": 12.625,
      "learning_rate": 4.990112940462673e-05,
      "loss": 1.8744390487670899,
      "mean_token_accuracy": 0.6011014070361853,
      "num_tokens": 5821261.0,
      "step": 2950
    },
    {
      "entropy": 1.7371204309165478,
      "epoch": 0.0847858155622073,
      "grad_norm": 10.5625,
      "learning_rate": 4.990011290201987e-05,
      "loss": 1.8446439743041991,
      "mean_token_accuracy": 0.6221200305968523,
      "num_tokens": 5839076.0,
      "step": 2960
    },
    {
      "entropy": 1.7352172512561084,
      "epoch": 0.08507225412829583,
      "grad_norm": 13.0625,
      "learning_rate": 4.989909121115896e-05,
      "loss": 1.8494783401489259,
      "mean_token_accuracy": 0.609843698143959,
      "num_tokens": 5859082.0,
      "step": 2970
    },
    {
      "entropy": 1.8165854930877685,
      "epoch": 0.08535869269438437,
      "grad_norm": 9.25,
      "learning_rate": 4.9898064332256863e-05,
      "loss": 1.932807159423828,
      "mean_token_accuracy": 0.5868445619940758,
      "num_tokens": 5879881.0,
      "step": 2980
    },
    {
      "entropy": 1.8055683925747872,
      "epoch": 0.0856451312604729,
      "grad_norm": 10.125,
      "learning_rate": 4.9897032265527564e-05,
      "loss": 1.8408864974975585,
      "mean_token_accuracy": 0.6175807427614928,
      "num_tokens": 5899454.0,
      "step": 2990
    },
    {
      "entropy": 1.8551389299333096,
      "epoch": 0.08593156982656144,
      "grad_norm": 10.9375,
      "learning_rate": 4.989599501118609e-05,
      "loss": 1.9585966110229491,
      "mean_token_accuracy": 0.6069829035550356,
      "num_tokens": 5918680.0,
      "step": 3000
    },
    {
      "entropy": 1.8998742889612914,
      "epoch": 0.08621800839264998,
      "grad_norm": 11.8125,
      "learning_rate": 4.9894952569448586e-05,
      "loss": 1.9223085403442384,
      "mean_token_accuracy": 0.6017805133014917,
      "num_tokens": 5938586.0,
      "step": 3010
    },
    {
      "entropy": 1.7115729860961437,
      "epoch": 0.08650444695873852,
      "grad_norm": 10.125,
      "learning_rate": 4.989390494053223e-05,
      "loss": 1.7946624755859375,
      "mean_token_accuracy": 0.6222159150987864,
      "num_tokens": 5958385.0,
      "step": 3020
    },
    {
      "entropy": 1.7438877299427986,
      "epoch": 0.08679088552482707,
      "grad_norm": 10.9375,
      "learning_rate": 4.989285212465533e-05,
      "loss": 1.8417003631591797,
      "mean_token_accuracy": 0.6071909379214049,
      "num_tokens": 5979957.0,
      "step": 3030
    },
    {
      "entropy": 1.7134101435542106,
      "epoch": 0.0870773240909156,
      "grad_norm": 10.75,
      "learning_rate": 4.989179412203725e-05,
      "loss": 1.8356204986572267,
      "mean_token_accuracy": 0.6159730419516564,
      "num_tokens": 5999711.0,
      "step": 3040
    },
    {
      "entropy": 1.8474023260176182,
      "epoch": 0.08736376265700414,
      "grad_norm": 9.6875,
      "learning_rate": 4.989073093289844e-05,
      "loss": 1.8112276077270508,
      "mean_token_accuracy": 0.6077834688127041,
      "num_tokens": 6020635.0,
      "step": 3050
    },
    {
      "entropy": 1.8343049257993698,
      "epoch": 0.08765020122309268,
      "grad_norm": 10.25,
      "learning_rate": 4.988966255746042e-05,
      "loss": 1.9215509414672851,
      "mean_token_accuracy": 0.599896739795804,
      "num_tokens": 6039040.0,
      "step": 3060
    },
    {
      "entropy": 1.7826895229518414,
      "epoch": 0.08793663978918122,
      "grad_norm": 12.9375,
      "learning_rate": 4.988858899594581e-05,
      "loss": 1.8142948150634766,
      "mean_token_accuracy": 0.6235730387270451,
      "num_tokens": 6057807.0,
      "step": 3070
    },
    {
      "entropy": 1.6476876836270093,
      "epoch": 0.08822307835526975,
      "grad_norm": 8.125,
      "learning_rate": 4.988751024857829e-05,
      "loss": 1.789759635925293,
      "mean_token_accuracy": 0.6290214776992797,
      "num_tokens": 6076905.0,
      "step": 3080
    },
    {
      "entropy": 1.6936001196503638,
      "epoch": 0.08850951692135829,
      "grad_norm": 8.4375,
      "learning_rate": 4.988642631558264e-05,
      "loss": 1.7730207443237305,
      "mean_token_accuracy": 0.6257806573063135,
      "num_tokens": 6095746.0,
      "step": 3090
    },
    {
      "entropy": 1.7270611174404622,
      "epoch": 0.08879595548744683,
      "grad_norm": 9.0,
      "learning_rate": 4.98853371971847e-05,
      "loss": 1.819589614868164,
      "mean_token_accuracy": 0.6188563484698534,
      "num_tokens": 6116052.0,
      "step": 3100
    },
    {
      "entropy": 1.8175524927675724,
      "epoch": 0.08908239405353537,
      "grad_norm": 9.25,
      "learning_rate": 4.988424289361141e-05,
      "loss": 1.8551971435546875,
      "mean_token_accuracy": 0.6094944890588522,
      "num_tokens": 6135285.0,
      "step": 3110
    },
    {
      "entropy": 1.7288947857916355,
      "epoch": 0.0893688326196239,
      "grad_norm": 9.8125,
      "learning_rate": 4.9883143405090784e-05,
      "loss": 1.7936229705810547,
      "mean_token_accuracy": 0.6098865497857332,
      "num_tokens": 6154259.0,
      "step": 3120
    },
    {
      "entropy": 1.611307594552636,
      "epoch": 0.08965527118571244,
      "grad_norm": 11.8125,
      "learning_rate": 4.988203873185191e-05,
      "loss": 1.7600465774536134,
      "mean_token_accuracy": 0.6320841744542122,
      "num_tokens": 6174081.0,
      "step": 3130
    },
    {
      "entropy": 1.6638292074203491,
      "epoch": 0.08994170975180098,
      "grad_norm": 10.5,
      "learning_rate": 4.988092887412495e-05,
      "loss": 1.6762691497802735,
      "mean_token_accuracy": 0.6358845584094525,
      "num_tokens": 6193541.0,
      "step": 3140
    },
    {
      "entropy": 1.5803929798305034,
      "epoch": 0.09022814831788951,
      "grad_norm": 8.5,
      "learning_rate": 4.9879813832141185e-05,
      "loss": 1.674459457397461,
      "mean_token_accuracy": 0.6354668751358986,
      "num_tokens": 6213759.0,
      "step": 3150
    },
    {
      "entropy": 1.7594450511038304,
      "epoch": 0.09051458688397807,
      "grad_norm": 10.1875,
      "learning_rate": 4.987869360613293e-05,
      "loss": 1.871511459350586,
      "mean_token_accuracy": 0.6030887488275767,
      "num_tokens": 6232815.0,
      "step": 3160
    },
    {
      "entropy": 1.6973014652729035,
      "epoch": 0.0908010254500666,
      "grad_norm": 8.3125,
      "learning_rate": 4.987756819633359e-05,
      "loss": 1.7109146118164062,
      "mean_token_accuracy": 0.6301996454596519,
      "num_tokens": 6251993.0,
      "step": 3170
    },
    {
      "entropy": 1.6076964747160674,
      "epoch": 0.09108746401615514,
      "grad_norm": 11.375,
      "learning_rate": 4.987643760297768e-05,
      "loss": 1.641105842590332,
      "mean_token_accuracy": 0.6567545875906944,
      "num_tokens": 6272072.0,
      "step": 3180
    },
    {
      "entropy": 1.595090302079916,
      "epoch": 0.09137390258224368,
      "grad_norm": 14.5,
      "learning_rate": 4.9875301826300755e-05,
      "loss": 1.7908584594726562,
      "mean_token_accuracy": 0.6292324867099524,
      "num_tokens": 6291189.0,
      "step": 3190
    },
    {
      "entropy": 1.8106674984097482,
      "epoch": 0.09166034114833221,
      "grad_norm": 9.5625,
      "learning_rate": 4.9874160866539474e-05,
      "loss": 1.8661542892456056,
      "mean_token_accuracy": 0.6143469832837581,
      "num_tokens": 6310702.0,
      "step": 3200
    },
    {
      "entropy": 1.6575596556067467,
      "epoch": 0.09194677971442075,
      "grad_norm": 9.4375,
      "learning_rate": 4.987301472393157e-05,
      "loss": 1.7238683700561523,
      "mean_token_accuracy": 0.6423762965947389,
      "num_tokens": 6330808.0,
      "step": 3210
    },
    {
      "entropy": 1.6653618112206459,
      "epoch": 0.09223321828050929,
      "grad_norm": 13.3125,
      "learning_rate": 4.987186339871586e-05,
      "loss": 1.7727434158325195,
      "mean_token_accuracy": 0.635314765572548,
      "num_tokens": 6351021.0,
      "step": 3220
    },
    {
      "entropy": 1.610061589628458,
      "epoch": 0.09251965684659783,
      "grad_norm": 9.1875,
      "learning_rate": 4.9870706891132233e-05,
      "loss": 1.670724105834961,
      "mean_token_accuracy": 0.6369951967149973,
      "num_tokens": 6371075.0,
      "step": 3230
    },
    {
      "entropy": 1.7065989926457406,
      "epoch": 0.09280609541268636,
      "grad_norm": 11.375,
      "learning_rate": 4.9869545201421674e-05,
      "loss": 1.7772294998168945,
      "mean_token_accuracy": 0.6219883508980274,
      "num_tokens": 6390917.0,
      "step": 3240
    },
    {
      "entropy": 1.673648745007813,
      "epoch": 0.0930925339787749,
      "grad_norm": 8.9375,
      "learning_rate": 4.986837832982622e-05,
      "loss": 1.7278858184814454,
      "mean_token_accuracy": 0.623109820112586,
      "num_tokens": 6411001.0,
      "step": 3250
    },
    {
      "entropy": 1.7488887436687945,
      "epoch": 0.09337897254486344,
      "grad_norm": 11.9375,
      "learning_rate": 4.9867206276589006e-05,
      "loss": 1.89287109375,
      "mean_token_accuracy": 0.6046258065849542,
      "num_tokens": 6429414.0,
      "step": 3260
    },
    {
      "entropy": 1.7951447792351245,
      "epoch": 0.09366541111095197,
      "grad_norm": 11.875,
      "learning_rate": 4.9866029041954246e-05,
      "loss": 1.825006103515625,
      "mean_token_accuracy": 0.6126693487167358,
      "num_tokens": 6449990.0,
      "step": 3270
    },
    {
      "entropy": 1.7055876299738884,
      "epoch": 0.09395184967704051,
      "grad_norm": 11.3125,
      "learning_rate": 4.986484662616724e-05,
      "loss": 1.7274944305419921,
      "mean_token_accuracy": 0.6351925080642105,
      "num_tokens": 6470278.0,
      "step": 3280
    },
    {
      "entropy": 1.698704645037651,
      "epoch": 0.09423828824312905,
      "grad_norm": 14.0625,
      "learning_rate": 4.9863659029474344e-05,
      "loss": 1.854669952392578,
      "mean_token_accuracy": 0.6096575729548931,
      "num_tokens": 6489941.0,
      "step": 3290
    },
    {
      "entropy": 1.6474793441593647,
      "epoch": 0.0945247268092176,
      "grad_norm": 10.25,
      "learning_rate": 4.986246625212303e-05,
      "loss": 1.720954704284668,
      "mean_token_accuracy": 0.6335517570376397,
      "num_tokens": 6508114.0,
      "step": 3300
    },
    {
      "entropy": 1.6180988121777773,
      "epoch": 0.09481116537530614,
      "grad_norm": 9.0625,
      "learning_rate": 4.986126829436181e-05,
      "loss": 1.6974006652832032,
      "mean_token_accuracy": 0.6366110194474459,
      "num_tokens": 6527533.0,
      "step": 3310
    },
    {
      "entropy": 1.6719712637364865,
      "epoch": 0.09509760394139467,
      "grad_norm": 13.0,
      "learning_rate": 4.9860065156440295e-05,
      "loss": 1.7705236434936524,
      "mean_token_accuracy": 0.6281975015997887,
      "num_tokens": 6546758.0,
      "step": 3320
    },
    {
      "entropy": 1.6859460651874543,
      "epoch": 0.09538404250748321,
      "grad_norm": 9.625,
      "learning_rate": 4.985885683860918e-05,
      "loss": 1.7235284805297852,
      "mean_token_accuracy": 0.6323164828121662,
      "num_tokens": 6566701.0,
      "step": 3330
    },
    {
      "entropy": 1.8225383676588536,
      "epoch": 0.09567048107357175,
      "grad_norm": 10.375,
      "learning_rate": 4.985764334112023e-05,
      "loss": 1.8835548400878905,
      "mean_token_accuracy": 0.6049425881356001,
      "num_tokens": 6586244.0,
      "step": 3340
    },
    {
      "entropy": 1.6037709221243859,
      "epoch": 0.09595691963966028,
      "grad_norm": 10.9375,
      "learning_rate": 4.9856424664226295e-05,
      "loss": 1.6635082244873047,
      "mean_token_accuracy": 0.6288101118057966,
      "num_tokens": 6604512.0,
      "step": 3350
    },
    {
      "entropy": 1.664628365263343,
      "epoch": 0.09624335820574882,
      "grad_norm": 10.5625,
      "learning_rate": 4.9855200808181304e-05,
      "loss": 1.711388397216797,
      "mean_token_accuracy": 0.6356975879520178,
      "num_tokens": 6625014.0,
      "step": 3360
    },
    {
      "entropy": 1.7029705427587032,
      "epoch": 0.09652979677183736,
      "grad_norm": 11.75,
      "learning_rate": 4.985397177324026e-05,
      "loss": 1.8706041336059571,
      "mean_token_accuracy": 0.6043540142476559,
      "num_tokens": 6645301.0,
      "step": 3370
    },
    {
      "entropy": 1.7533822063356639,
      "epoch": 0.0968162353379259,
      "grad_norm": 7.71875,
      "learning_rate": 4.9852737559659245e-05,
      "loss": 1.7002222061157226,
      "mean_token_accuracy": 0.6346283335238695,
      "num_tokens": 6666406.0,
      "step": 3380
    },
    {
      "entropy": 1.5919454231858254,
      "epoch": 0.09710267390401443,
      "grad_norm": 12.0625,
      "learning_rate": 4.985149816769543e-05,
      "loss": 1.713643455505371,
      "mean_token_accuracy": 0.6311788672581315,
      "num_tokens": 6686285.0,
      "step": 3390
    },
    {
      "entropy": 1.5006841901689767,
      "epoch": 0.09738911247010297,
      "grad_norm": 12.1875,
      "learning_rate": 4.985025359760706e-05,
      "loss": 1.5002299308776856,
      "mean_token_accuracy": 0.6706645965576172,
      "num_tokens": 6705878.0,
      "step": 3400
    },
    {
      "entropy": 1.5395627930760383,
      "epoch": 0.09767555103619151,
      "grad_norm": 7.25,
      "learning_rate": 4.9849003849653435e-05,
      "loss": 1.6199533462524414,
      "mean_token_accuracy": 0.650242630019784,
      "num_tokens": 6726940.0,
      "step": 3410
    },
    {
      "entropy": 1.5471817657351494,
      "epoch": 0.09796198960228004,
      "grad_norm": 10.1875,
      "learning_rate": 4.984774892409499e-05,
      "loss": 1.6290157318115235,
      "mean_token_accuracy": 0.6466945555061102,
      "num_tokens": 6745927.0,
      "step": 3420
    },
    {
      "entropy": 1.6916074492037296,
      "epoch": 0.0982484281683686,
      "grad_norm": 10.1875,
      "learning_rate": 4.984648882119317e-05,
      "loss": 1.8082305908203125,
      "mean_token_accuracy": 0.6197715051472187,
      "num_tokens": 6765877.0,
      "step": 3430
    },
    {
      "entropy": 1.7348431102931499,
      "epoch": 0.09853486673445713,
      "grad_norm": 13.4375,
      "learning_rate": 4.984522354121055e-05,
      "loss": 1.7304534912109375,
      "mean_token_accuracy": 0.6302803590893745,
      "num_tokens": 6785495.0,
      "step": 3440
    },
    {
      "entropy": 1.644375092536211,
      "epoch": 0.09882130530054567,
      "grad_norm": 12.6875,
      "learning_rate": 4.9843953084410774e-05,
      "loss": 1.7694244384765625,
      "mean_token_accuracy": 0.6218827169388532,
      "num_tokens": 6804680.0,
      "step": 3450
    },
    {
      "entropy": 1.7094856541603805,
      "epoch": 0.09910774386663421,
      "grad_norm": 12.3125,
      "learning_rate": 4.984267745105855e-05,
      "loss": 1.7599742889404297,
      "mean_token_accuracy": 0.6293886683881282,
      "num_tokens": 6823407.0,
      "step": 3460
    },
    {
      "entropy": 1.5646884053945542,
      "epoch": 0.09939418243272274,
      "grad_norm": 10.5,
      "learning_rate": 4.984139664141967e-05,
      "loss": 1.5917807579040528,
      "mean_token_accuracy": 0.651865516602993,
      "num_tokens": 6843403.0,
      "step": 3470
    },
    {
      "entropy": 1.5828524824231862,
      "epoch": 0.09968062099881128,
      "grad_norm": 8.125,
      "learning_rate": 4.984011065576102e-05,
      "loss": 1.7056592941284179,
      "mean_token_accuracy": 0.6319200482219458,
      "num_tokens": 6863996.0,
      "step": 3480
    },
    {
      "entropy": 1.5854770459234715,
      "epoch": 0.09996705956489982,
      "grad_norm": 10.875,
      "learning_rate": 4.983881949435053e-05,
      "loss": 1.687114143371582,
      "mean_token_accuracy": 0.6348020933568478,
      "num_tokens": 6884193.0,
      "step": 3490
    },
    {
      "entropy": 1.6993172325193882,
      "epoch": 0.10025349813098836,
      "grad_norm": 7.0625,
      "learning_rate": 4.9837523157457246e-05,
      "loss": 1.7156557083129882,
      "mean_token_accuracy": 0.6427715122699738,
      "num_tokens": 6905445.0,
      "step": 3500
    },
    {
      "entropy": 1.6515631839632987,
      "epoch": 0.10053993669707689,
      "grad_norm": 10.0625,
      "learning_rate": 4.983622164535127e-05,
      "loss": 1.6537338256835938,
      "mean_token_accuracy": 0.6338998597115278,
      "num_tokens": 6925008.0,
      "step": 3510
    },
    {
      "entropy": 1.6213870584964751,
      "epoch": 0.10082637526316543,
      "grad_norm": 10.125,
      "learning_rate": 4.9834914958303794e-05,
      "loss": 1.828173828125,
      "mean_token_accuracy": 0.6133347254246473,
      "num_tokens": 6944815.0,
      "step": 3520
    },
    {
      "entropy": 1.7784586071968078,
      "epoch": 0.10111281382925397,
      "grad_norm": 15.5,
      "learning_rate": 4.983360309658708e-05,
      "loss": 1.8722478866577148,
      "mean_token_accuracy": 0.6086947064846754,
      "num_tokens": 6965037.0,
      "step": 3530
    },
    {
      "entropy": 1.8203211426734924,
      "epoch": 0.1013992523953425,
      "grad_norm": 7.90625,
      "learning_rate": 4.983228606047446e-05,
      "loss": 1.8089092254638672,
      "mean_token_accuracy": 0.6207359131425619,
      "num_tokens": 6985131.0,
      "step": 3540
    },
    {
      "entropy": 1.6216672886162997,
      "epoch": 0.10168569096143104,
      "grad_norm": 13.25,
      "learning_rate": 4.983096385024037e-05,
      "loss": 1.6625095367431642,
      "mean_token_accuracy": 0.6502947527915239,
      "num_tokens": 7004418.0,
      "step": 3550
    },
    {
      "entropy": 1.60721864849329,
      "epoch": 0.10197212952751959,
      "grad_norm": 15.75,
      "learning_rate": 4.9829636466160307e-05,
      "loss": 1.688064193725586,
      "mean_token_accuracy": 0.6488478925079108,
      "num_tokens": 7024109.0,
      "step": 3560
    },
    {
      "entropy": 1.5623432353138924,
      "epoch": 0.10225856809360813,
      "grad_norm": 9.375,
      "learning_rate": 4.982830390851084e-05,
      "loss": 1.7058576583862304,
      "mean_token_accuracy": 0.6410759892314672,
      "num_tokens": 7042737.0,
      "step": 3570
    },
    {
      "entropy": 1.7128427192568778,
      "epoch": 0.10254500665969667,
      "grad_norm": 10.75,
      "learning_rate": 4.982696617756963e-05,
      "loss": 1.7900766372680663,
      "mean_token_accuracy": 0.6258037876337766,
      "num_tokens": 7062429.0,
      "step": 3580
    },
    {
      "entropy": 1.745885034650564,
      "epoch": 0.1028314452257852,
      "grad_norm": 11.0,
      "learning_rate": 4.9825623273615416e-05,
      "loss": 1.81469669342041,
      "mean_token_accuracy": 0.6134201630949974,
      "num_tokens": 7081262.0,
      "step": 3590
    },
    {
      "entropy": 1.8495959386229515,
      "epoch": 0.10311788379187374,
      "grad_norm": 9.25,
      "learning_rate": 4.9824275196928006e-05,
      "loss": 1.8182167053222655,
      "mean_token_accuracy": 0.6066920112818479,
      "num_tokens": 7101046.0,
      "step": 3600
    },
    {
      "entropy": 1.5995587468147279,
      "epoch": 0.10340432235796228,
      "grad_norm": 13.625,
      "learning_rate": 4.982292194778828e-05,
      "loss": 1.7035053253173829,
      "mean_token_accuracy": 0.625184166803956,
      "num_tokens": 7121598.0,
      "step": 3610
    },
    {
      "entropy": 1.5674044363200665,
      "epoch": 0.10369076092405081,
      "grad_norm": 13.3125,
      "learning_rate": 4.982156352647821e-05,
      "loss": 1.6441669464111328,
      "mean_token_accuracy": 0.6342104159295558,
      "num_tokens": 7140583.0,
      "step": 3620
    },
    {
      "entropy": 1.7234221808612347,
      "epoch": 0.10397719949013935,
      "grad_norm": 10.25,
      "learning_rate": 4.982019993328084e-05,
      "loss": 1.8390151977539062,
      "mean_token_accuracy": 0.6153890393674374,
      "num_tokens": 7160170.0,
      "step": 3630
    },
    {
      "entropy": 1.7872201681137085,
      "epoch": 0.10426363805622789,
      "grad_norm": 9.1875,
      "learning_rate": 4.98188311684803e-05,
      "loss": 1.74482421875,
      "mean_token_accuracy": 0.6153182607144118,
      "num_tokens": 7180095.0,
      "step": 3640
    },
    {
      "entropy": 1.5389251604676246,
      "epoch": 0.10455007662231643,
      "grad_norm": 10.4375,
      "learning_rate": 4.9817457232361776e-05,
      "loss": 1.6207878112792968,
      "mean_token_accuracy": 0.647048769891262,
      "num_tokens": 7200375.0,
      "step": 3650
    },
    {
      "entropy": 1.5512934401631355,
      "epoch": 0.10483651518840496,
      "grad_norm": 12.5625,
      "learning_rate": 4.981607812521155e-05,
      "loss": 1.6449550628662108,
      "mean_token_accuracy": 0.6466568242758512,
      "num_tokens": 7219626.0,
      "step": 3660
    },
    {
      "entropy": 1.6129456333816052,
      "epoch": 0.1051229537544935,
      "grad_norm": 10.375,
      "learning_rate": 4.981469384731698e-05,
      "loss": 1.7151153564453125,
      "mean_token_accuracy": 0.6369513172656298,
      "num_tokens": 7240379.0,
      "step": 3670
    },
    {
      "entropy": 1.6685446359217166,
      "epoch": 0.10540939232058204,
      "grad_norm": 12.5625,
      "learning_rate": 4.98133043989665e-05,
      "loss": 1.687558937072754,
      "mean_token_accuracy": 0.6465672057121992,
      "num_tokens": 7260083.0,
      "step": 3680
    },
    {
      "entropy": 1.568697489053011,
      "epoch": 0.10569583088667057,
      "grad_norm": 17.375,
      "learning_rate": 4.9811909780449606e-05,
      "loss": 1.6286209106445313,
      "mean_token_accuracy": 0.6435823444277048,
      "num_tokens": 7279720.0,
      "step": 3690
    },
    {
      "entropy": 1.7579448983073234,
      "epoch": 0.10598226945275913,
      "grad_norm": 9.6875,
      "learning_rate": 4.98105099920569e-05,
      "loss": 1.8242202758789063,
      "mean_token_accuracy": 0.622202886082232,
      "num_tokens": 7299830.0,
      "step": 3700
    },
    {
      "entropy": 1.5484590135514735,
      "epoch": 0.10626870801884766,
      "grad_norm": 9.25,
      "learning_rate": 4.980910503408003e-05,
      "loss": 1.5815845489501954,
      "mean_token_accuracy": 0.6635242309421301,
      "num_tokens": 7318924.0,
      "step": 3710
    },
    {
      "entropy": 1.6158774755895138,
      "epoch": 0.1065551465849362,
      "grad_norm": 9.4375,
      "learning_rate": 4.980769490681175e-05,
      "loss": 1.6784629821777344,
      "mean_token_accuracy": 0.6403291534632445,
      "num_tokens": 7339038.0,
      "step": 3720
    },
    {
      "entropy": 1.7512457430362702,
      "epoch": 0.10684158515102474,
      "grad_norm": 8.6875,
      "learning_rate": 4.980627961054588e-05,
      "loss": 1.7499198913574219,
      "mean_token_accuracy": 0.6350882723927498,
      "num_tokens": 7358680.0,
      "step": 3730
    },
    {
      "entropy": 1.6030123636126519,
      "epoch": 0.10712802371711327,
      "grad_norm": 12.6875,
      "learning_rate": 4.9804859145577306e-05,
      "loss": 1.6888309478759767,
      "mean_token_accuracy": 0.64278660453856,
      "num_tokens": 7376809.0,
      "step": 3740
    },
    {
      "entropy": 1.599424796551466,
      "epoch": 0.10741446228320181,
      "grad_norm": 8.5,
      "learning_rate": 4.9803433512202e-05,
      "loss": 1.647484016418457,
      "mean_token_accuracy": 0.654452845454216,
      "num_tokens": 7396027.0,
      "step": 3750
    },
    {
      "entropy": 1.4993393868207932,
      "epoch": 0.10770090084929035,
      "grad_norm": 8.6875,
      "learning_rate": 4.980200271071701e-05,
      "loss": 1.6149499893188477,
      "mean_token_accuracy": 0.6482477340847254,
      "num_tokens": 7415794.0,
      "step": 3760
    },
    {
      "entropy": 1.6712707415223123,
      "epoch": 0.10798733941537889,
      "grad_norm": 8.5625,
      "learning_rate": 4.980056674142047e-05,
      "loss": 1.7163467407226562,
      "mean_token_accuracy": 0.6339017499238253,
      "num_tokens": 7437207.0,
      "step": 3770
    },
    {
      "entropy": 1.6393704406917096,
      "epoch": 0.10827377798146742,
      "grad_norm": 13.625,
      "learning_rate": 4.979912560461157e-05,
      "loss": 1.7169496536254882,
      "mean_token_accuracy": 0.6368111822754144,
      "num_tokens": 7455892.0,
      "step": 3780
    },
    {
      "entropy": 1.7394482620060443,
      "epoch": 0.10856021654755596,
      "grad_norm": 9.25,
      "learning_rate": 4.97976793005906e-05,
      "loss": 1.8061094284057617,
      "mean_token_accuracy": 0.6174036616459488,
      "num_tokens": 7475186.0,
      "step": 3790
    },
    {
      "entropy": 1.552655541524291,
      "epoch": 0.1088466551136445,
      "grad_norm": 9.1875,
      "learning_rate": 4.9796227829658916e-05,
      "loss": 1.611650276184082,
      "mean_token_accuracy": 0.6506694372743368,
      "num_tokens": 7495326.0,
      "step": 3800
    },
    {
      "entropy": 1.6132732205092908,
      "epoch": 0.10913309367973303,
      "grad_norm": 7.5,
      "learning_rate": 4.9794771192118945e-05,
      "loss": 1.6967041015625,
      "mean_token_accuracy": 0.6432500820606947,
      "num_tokens": 7515063.0,
      "step": 3810
    },
    {
      "entropy": 1.7228643730282784,
      "epoch": 0.10941953224582157,
      "grad_norm": 9.625,
      "learning_rate": 4.97933093882742e-05,
      "loss": 1.7208276748657227,
      "mean_token_accuracy": 0.6368235297501087,
      "num_tokens": 7535102.0,
      "step": 3820
    },
    {
      "entropy": 1.5207892529666425,
      "epoch": 0.10970597081191012,
      "grad_norm": 12.625,
      "learning_rate": 4.9791842418429256e-05,
      "loss": 1.5424155235290526,
      "mean_token_accuracy": 0.6668250478804112,
      "num_tokens": 7554024.0,
      "step": 3830
    },
    {
      "entropy": 1.4753357917070389,
      "epoch": 0.10999240937799866,
      "grad_norm": 11.0,
      "learning_rate": 4.979037028288979e-05,
      "loss": 1.5886699676513671,
      "mean_token_accuracy": 0.6615104977041483,
      "num_tokens": 7573906.0,
      "step": 3840
    },
    {
      "entropy": 1.6845372278243302,
      "epoch": 0.1102788479440872,
      "grad_norm": 10.0,
      "learning_rate": 4.978889298196252e-05,
      "loss": 1.769597625732422,
      "mean_token_accuracy": 0.6292838361114264,
      "num_tokens": 7593541.0,
      "step": 3850
    },
    {
      "entropy": 1.5769455425441266,
      "epoch": 0.11056528651017573,
      "grad_norm": 9.125,
      "learning_rate": 4.978741051595528e-05,
      "loss": 1.5808780670166016,
      "mean_token_accuracy": 0.6541871156543493,
      "num_tokens": 7612674.0,
      "step": 3860
    },
    {
      "entropy": 1.6566674526780845,
      "epoch": 0.11085172507626427,
      "grad_norm": 9.25,
      "learning_rate": 4.9785922885176956e-05,
      "loss": 1.7226404190063476,
      "mean_token_accuracy": 0.625541852042079,
      "num_tokens": 7632543.0,
      "step": 3870
    },
    {
      "entropy": 1.6096037060022355,
      "epoch": 0.11113816364235281,
      "grad_norm": 8.6875,
      "learning_rate": 4.9784430089937504e-05,
      "loss": 1.708632469177246,
      "mean_token_accuracy": 0.6383926354348659,
      "num_tokens": 7651652.0,
      "step": 3880
    },
    {
      "entropy": 1.723529090732336,
      "epoch": 0.11142460220844135,
      "grad_norm": 9.75,
      "learning_rate": 4.9782932130547986e-05,
      "loss": 1.759688949584961,
      "mean_token_accuracy": 0.618687403947115,
      "num_tokens": 7671203.0,
      "step": 3890
    },
    {
      "entropy": 1.557471706904471,
      "epoch": 0.11171104077452988,
      "grad_norm": 9.75,
      "learning_rate": 4.978142900732051e-05,
      "loss": 1.5875686645507812,
      "mean_token_accuracy": 0.6499525845050812,
      "num_tokens": 7690516.0,
      "step": 3900
    },
    {
      "entropy": 1.551585179939866,
      "epoch": 0.11199747934061842,
      "grad_norm": 8.8125,
      "learning_rate": 4.977992072056826e-05,
      "loss": 1.650254249572754,
      "mean_token_accuracy": 0.6560598902404309,
      "num_tokens": 7709298.0,
      "step": 3910
    },
    {
      "entropy": 1.5815889041870832,
      "epoch": 0.11228391790670696,
      "grad_norm": 9.5625,
      "learning_rate": 4.977840727060552e-05,
      "loss": 1.593684482574463,
      "mean_token_accuracy": 0.6460284203290939,
      "num_tokens": 7729071.0,
      "step": 3920
    },
    {
      "entropy": 1.5101389355957509,
      "epoch": 0.1125703564727955,
      "grad_norm": 9.6875,
      "learning_rate": 4.977688865774764e-05,
      "loss": 1.6238428115844727,
      "mean_token_accuracy": 0.6475252192467451,
      "num_tokens": 7750776.0,
      "step": 3930
    },
    {
      "entropy": 1.6004488527774812,
      "epoch": 0.11285679503888403,
      "grad_norm": 8.8125,
      "learning_rate": 4.977536488231103e-05,
      "loss": 1.6589231491088867,
      "mean_token_accuracy": 0.6409353025257587,
      "num_tokens": 7771076.0,
      "step": 3940
    },
    {
      "entropy": 1.6314193956553935,
      "epoch": 0.11314323360497257,
      "grad_norm": 9.1875,
      "learning_rate": 4.97738359446132e-05,
      "loss": 1.6660181045532227,
      "mean_token_accuracy": 0.6412204388529062,
      "num_tokens": 7791504.0,
      "step": 3950
    },
    {
      "entropy": 1.6633416850119829,
      "epoch": 0.1134296721710611,
      "grad_norm": 11.25,
      "learning_rate": 4.977230184497271e-05,
      "loss": 1.7595705032348632,
      "mean_token_accuracy": 0.6309101179242134,
      "num_tokens": 7810987.0,
      "step": 3960
    },
    {
      "entropy": 1.5157507963478565,
      "epoch": 0.11371611073714966,
      "grad_norm": 9.8125,
      "learning_rate": 4.977076258370922e-05,
      "loss": 1.607263946533203,
      "mean_token_accuracy": 0.6623361337929964,
      "num_tokens": 7831549.0,
      "step": 3970
    },
    {
      "entropy": 1.6064405851066113,
      "epoch": 0.1140025493032382,
      "grad_norm": 9.0625,
      "learning_rate": 4.976921816114345e-05,
      "loss": 1.6695703506469726,
      "mean_token_accuracy": 0.6595016360282898,
      "num_tokens": 7851226.0,
      "step": 3980
    },
    {
      "entropy": 1.6939660400152206,
      "epoch": 0.11428898786932673,
      "grad_norm": 10.375,
      "learning_rate": 4.97676685775972e-05,
      "loss": 1.7704540252685548,
      "mean_token_accuracy": 0.6205462880432606,
      "num_tokens": 7872015.0,
      "step": 3990
    },
    {
      "epoch": 0.11457542643541527,
      "eval_entropy": 1.6098786599636077,
      "eval_loss": 1.6650747060775757,
      "eval_mean_token_accuracy": 0.6405204639434815,
      "eval_num_tokens": 7891042.0,
      "eval_runtime": 43.7294,
      "eval_samples_per_second": 45.736,
      "eval_steps_per_second": 5.717,
      "step": 4000
    },
    {
      "entropy": 1.6005963683128357,
      "epoch": 0.1148618650015038,
      "grad_norm": 9.625,
      "learning_rate": 4.976455392885585e-05,
      "loss": 1.6549604415893555,
      "mean_token_accuracy": 0.6485565780662,
      "num_tokens": 7911920.0,
      "step": 4010
    },
    {
      "entropy": 1.491986843943596,
      "epoch": 0.11514830356759234,
      "grad_norm": 10.75,
      "learning_rate": 4.9762988864309714e-05,
      "loss": 1.5747516632080079,
      "mean_token_accuracy": 0.6672815855592489,
      "num_tokens": 7932491.0,
      "step": 4020
    },
    {
      "entropy": 1.5924518678337336,
      "epoch": 0.11543474213368088,
      "grad_norm": 9.75,
      "learning_rate": 4.976141864008106e-05,
      "loss": 1.6315956115722656,
      "mean_token_accuracy": 0.6538907889276743,
      "num_tokens": 7952550.0,
      "step": 4030
    },
    {
      "entropy": 1.7031397961080075,
      "epoch": 0.11572118069976942,
      "grad_norm": 10.875,
      "learning_rate": 4.975984325649705e-05,
      "loss": 1.752387237548828,
      "mean_token_accuracy": 0.6324999265372753,
      "num_tokens": 7972914.0,
      "step": 4040
    },
    {
      "entropy": 1.595766806602478,
      "epoch": 0.11600761926585795,
      "grad_norm": 9.3125,
      "learning_rate": 4.975826271388594e-05,
      "loss": 1.5881695747375488,
      "mean_token_accuracy": 0.6493427228182554,
      "num_tokens": 7992312.0,
      "step": 4050
    },
    {
      "entropy": 1.5366006754338741,
      "epoch": 0.11629405783194649,
      "grad_norm": 9.875,
      "learning_rate": 4.975667701257706e-05,
      "loss": 1.6710201263427735,
      "mean_token_accuracy": 0.640213792771101,
      "num_tokens": 8012711.0,
      "step": 4060
    },
    {
      "entropy": 1.5428064178675414,
      "epoch": 0.11658049639803503,
      "grad_norm": 7.90625,
      "learning_rate": 4.9755086152900806e-05,
      "loss": 1.6362485885620117,
      "mean_token_accuracy": 0.6502720098942518,
      "num_tokens": 8033643.0,
      "step": 4070
    },
    {
      "entropy": 1.5757974475622176,
      "epoch": 0.11686693496412356,
      "grad_norm": 11.0625,
      "learning_rate": 4.9753490135188664e-05,
      "loss": 1.5432090759277344,
      "mean_token_accuracy": 0.663056916743517,
      "num_tokens": 8053901.0,
      "step": 4080
    },
    {
      "entropy": 1.5034592628479004,
      "epoch": 0.1171533735302121,
      "grad_norm": 8.5625,
      "learning_rate": 4.9751888959773176e-05,
      "loss": 1.6036109924316406,
      "mean_token_accuracy": 0.6504288144409657,
      "num_tokens": 8074275.0,
      "step": 4090
    },
    {
      "entropy": 1.5517423652112483,
      "epoch": 0.11743981209630065,
      "grad_norm": 11.9375,
      "learning_rate": 4.975028262698795e-05,
      "loss": 1.5871983528137208,
      "mean_token_accuracy": 0.6494060952216387,
      "num_tokens": 8094046.0,
      "step": 4100
    },
    {
      "entropy": 1.5514112140983343,
      "epoch": 0.11772625066238919,
      "grad_norm": 7.5625,
      "learning_rate": 4.9748671137167724e-05,
      "loss": 1.6497369766235352,
      "mean_token_accuracy": 0.6371284432709217,
      "num_tokens": 8114035.0,
      "step": 4110
    },
    {
      "entropy": 1.6656693324446679,
      "epoch": 0.11801268922847773,
      "grad_norm": 11.125,
      "learning_rate": 4.974705449064825e-05,
      "loss": 1.732563591003418,
      "mean_token_accuracy": 0.6293914947658777,
      "num_tokens": 8133291.0,
      "step": 4120
    },
    {
      "entropy": 1.4772680297493934,
      "epoch": 0.11829912779456626,
      "grad_norm": 7.46875,
      "learning_rate": 4.9745432687766366e-05,
      "loss": 1.5282196044921874,
      "mean_token_accuracy": 0.669588616490364,
      "num_tokens": 8152873.0,
      "step": 4130
    },
    {
      "entropy": 1.650682621449232,
      "epoch": 0.1185855663606548,
      "grad_norm": 7.25,
      "learning_rate": 4.9743805728860016e-05,
      "loss": 1.6047649383544922,
      "mean_token_accuracy": 0.6563232995569706,
      "num_tokens": 8172371.0,
      "step": 4140
    },
    {
      "entropy": 1.4914087392389774,
      "epoch": 0.11887200492674334,
      "grad_norm": 9.1875,
      "learning_rate": 4.974217361426819e-05,
      "loss": 1.6290849685668944,
      "mean_token_accuracy": 0.6563458885997534,
      "num_tokens": 8193182.0,
      "step": 4150
    },
    {
      "entropy": 1.53787529990077,
      "epoch": 0.11915844349283188,
      "grad_norm": 10.875,
      "learning_rate": 4.9740536344330956e-05,
      "loss": 1.609867477416992,
      "mean_token_accuracy": 0.6524941785261035,
      "num_tokens": 8213488.0,
      "step": 4160
    },
    {
      "entropy": 1.4693402014672756,
      "epoch": 0.11944488205892041,
      "grad_norm": 9.5,
      "learning_rate": 4.973889391938946e-05,
      "loss": 1.5584410667419433,
      "mean_token_accuracy": 0.6637550175189972,
      "num_tokens": 8233104.0,
      "step": 4170
    },
    {
      "entropy": 1.5683077815920115,
      "epoch": 0.11973132062500895,
      "grad_norm": 8.0,
      "learning_rate": 4.973724633978594e-05,
      "loss": 1.589338207244873,
      "mean_token_accuracy": 0.6586020454764366,
      "num_tokens": 8252390.0,
      "step": 4180
    },
    {
      "entropy": 1.6479635879397392,
      "epoch": 0.12001775919109749,
      "grad_norm": 8.9375,
      "learning_rate": 4.973559360586366e-05,
      "loss": 1.7119781494140625,
      "mean_token_accuracy": 0.63336865529418,
      "num_tokens": 8273234.0,
      "step": 4190
    },
    {
      "entropy": 1.5484042853116988,
      "epoch": 0.12030419775718602,
      "grad_norm": 9.8125,
      "learning_rate": 4.9733935717967004e-05,
      "loss": 1.5573399543762207,
      "mean_token_accuracy": 0.660521986335516,
      "num_tokens": 8292078.0,
      "step": 4200
    },
    {
      "entropy": 1.5313296984881162,
      "epoch": 0.12059063632327456,
      "grad_norm": 7.65625,
      "learning_rate": 4.973227267644142e-05,
      "loss": 1.6131240844726562,
      "mean_token_accuracy": 0.6444024810567498,
      "num_tokens": 8312346.0,
      "step": 4210
    },
    {
      "entropy": 1.5710236757993699,
      "epoch": 0.1208770748893631,
      "grad_norm": 10.75,
      "learning_rate": 4.9730604481633415e-05,
      "loss": 1.6453704833984375,
      "mean_token_accuracy": 0.6383862376213074,
      "num_tokens": 8332138.0,
      "step": 4220
    },
    {
      "entropy": 1.5641222171485425,
      "epoch": 0.12116351345545164,
      "grad_norm": 9.5625,
      "learning_rate": 4.9728931133890575e-05,
      "loss": 1.588129997253418,
      "mean_token_accuracy": 0.6567653119564056,
      "num_tokens": 8350284.0,
      "step": 4230
    },
    {
      "entropy": 1.6658044099807738,
      "epoch": 0.12144995202154019,
      "grad_norm": 8.5,
      "learning_rate": 4.972725263356158e-05,
      "loss": 1.755426025390625,
      "mean_token_accuracy": 0.6316048409789801,
      "num_tokens": 8369770.0,
      "step": 4240
    },
    {
      "entropy": 1.6496659649536014,
      "epoch": 0.12173639058762872,
      "grad_norm": 9.5,
      "learning_rate": 4.9725568980996144e-05,
      "loss": 1.7039173126220704,
      "mean_token_accuracy": 0.6400847017765046,
      "num_tokens": 8388523.0,
      "step": 4250
    },
    {
      "entropy": 1.5752143502235412,
      "epoch": 0.12202282915371726,
      "grad_norm": 9.0625,
      "learning_rate": 4.97238801765451e-05,
      "loss": 1.633380126953125,
      "mean_token_accuracy": 0.6485164921730757,
      "num_tokens": 8407657.0,
      "step": 4260
    },
    {
      "entropy": 1.593596202880144,
      "epoch": 0.1223092677198058,
      "grad_norm": 9.1875,
      "learning_rate": 4.972218622056031e-05,
      "loss": 1.6573898315429687,
      "mean_token_accuracy": 0.6489590808749199,
      "num_tokens": 8426029.0,
      "step": 4270
    },
    {
      "entropy": 1.610462510213256,
      "epoch": 0.12259570628589433,
      "grad_norm": 9.5,
      "learning_rate": 4.9720487113394744e-05,
      "loss": 1.7680988311767578,
      "mean_token_accuracy": 0.6347607377916574,
      "num_tokens": 8446995.0,
      "step": 4280
    },
    {
      "entropy": 1.661719637736678,
      "epoch": 0.12288214485198287,
      "grad_norm": 7.375,
      "learning_rate": 4.971878285540243e-05,
      "loss": 1.6369525909423828,
      "mean_token_accuracy": 0.6406619127839803,
      "num_tokens": 8466787.0,
      "step": 4290
    },
    {
      "entropy": 1.583110580407083,
      "epoch": 0.12316858341807141,
      "grad_norm": 8.8125,
      "learning_rate": 4.971707344693848e-05,
      "loss": 1.6308673858642577,
      "mean_token_accuracy": 0.6463008146733046,
      "num_tokens": 8487763.0,
      "step": 4300
    },
    {
      "entropy": 1.49612433090806,
      "epoch": 0.12345502198415995,
      "grad_norm": 9.0625,
      "learning_rate": 4.9715358888359055e-05,
      "loss": 1.6682989120483398,
      "mean_token_accuracy": 0.6445597290992737,
      "num_tokens": 8507128.0,
      "step": 4310
    },
    {
      "entropy": 1.5271134339272976,
      "epoch": 0.12374146055024848,
      "grad_norm": 10.5,
      "learning_rate": 4.971363918002142e-05,
      "loss": 1.585581398010254,
      "mean_token_accuracy": 0.6652600765228271,
      "num_tokens": 8526072.0,
      "step": 4320
    },
    {
      "entropy": 1.6250931091606617,
      "epoch": 0.12402789911633702,
      "grad_norm": 8.3125,
      "learning_rate": 4.971191432228388e-05,
      "loss": 1.6668222427368165,
      "mean_token_accuracy": 0.6372327048331499,
      "num_tokens": 8546952.0,
      "step": 4330
    },
    {
      "entropy": 1.5376979611814021,
      "epoch": 0.12431433768242556,
      "grad_norm": 9.3125,
      "learning_rate": 4.971018431550585e-05,
      "loss": 1.5736823081970215,
      "mean_token_accuracy": 0.6680327545851469,
      "num_tokens": 8566795.0,
      "step": 4340
    },
    {
      "entropy": 1.5797248736023903,
      "epoch": 0.1246007762485141,
      "grad_norm": 9.6875,
      "learning_rate": 4.970844916004779e-05,
      "loss": 1.5739888191223144,
      "mean_token_accuracy": 0.6595445297658443,
      "num_tokens": 8586617.0,
      "step": 4350
    },
    {
      "entropy": 1.5412679776549338,
      "epoch": 0.12488721481460263,
      "grad_norm": 8.625,
      "learning_rate": 4.970670885627124e-05,
      "loss": 1.6226394653320313,
      "mean_token_accuracy": 0.6362952403724194,
      "num_tokens": 8607139.0,
      "step": 4360
    },
    {
      "entropy": 1.3484292313456536,
      "epoch": 0.12517365338069117,
      "grad_norm": 8.5625,
      "learning_rate": 4.970496340453882e-05,
      "loss": 1.4472694396972656,
      "mean_token_accuracy": 0.6806314267218113,
      "num_tokens": 8627414.0,
      "step": 4370
    },
    {
      "entropy": 1.6275325849652291,
      "epoch": 0.1254600919467797,
      "grad_norm": 9.75,
      "learning_rate": 4.9703212805214214e-05,
      "loss": 1.6965776443481446,
      "mean_token_accuracy": 0.6306652009487153,
      "num_tokens": 8647004.0,
      "step": 4380
    },
    {
      "entropy": 1.5334886122494935,
      "epoch": 0.12574653051286824,
      "grad_norm": 7.375,
      "learning_rate": 4.970145705866218e-05,
      "loss": 1.5669997215270997,
      "mean_token_accuracy": 0.6734387289732695,
      "num_tokens": 8665559.0,
      "step": 4390
    },
    {
      "entropy": 1.6214932195842267,
      "epoch": 0.12603296907895678,
      "grad_norm": 7.4375,
      "learning_rate": 4.9699696165248556e-05,
      "loss": 1.6898490905761718,
      "mean_token_accuracy": 0.6344234116375447,
      "num_tokens": 8685046.0,
      "step": 4400
    },
    {
      "entropy": 1.5889239586889743,
      "epoch": 0.12631940764504532,
      "grad_norm": 6.78125,
      "learning_rate": 4.969793012534023e-05,
      "loss": 1.662093162536621,
      "mean_token_accuracy": 0.6500432893633843,
      "num_tokens": 8703776.0,
      "step": 4410
    },
    {
      "entropy": 1.6307801317423583,
      "epoch": 0.12660584621113385,
      "grad_norm": 12.25,
      "learning_rate": 4.9696158939305205e-05,
      "loss": 1.7354646682739259,
      "mean_token_accuracy": 0.6317333064973354,
      "num_tokens": 8722912.0,
      "step": 4420
    },
    {
      "entropy": 1.5595045857131482,
      "epoch": 0.1268922847772224,
      "grad_norm": 9.25,
      "learning_rate": 4.96943826075125e-05,
      "loss": 1.5597323417663573,
      "mean_token_accuracy": 0.662355512753129,
      "num_tokens": 8741744.0,
      "step": 4430
    },
    {
      "entropy": 1.5082399692386388,
      "epoch": 0.12717872334331096,
      "grad_norm": 9.5,
      "learning_rate": 4.9692601130332264e-05,
      "loss": 1.5893750190734863,
      "mean_token_accuracy": 0.6521870832890272,
      "num_tokens": 8760396.0,
      "step": 4440
    },
    {
      "entropy": 1.6613531835377215,
      "epoch": 0.1274651619093995,
      "grad_norm": 11.1875,
      "learning_rate": 4.9690814508135675e-05,
      "loss": 1.7102399826049806,
      "mean_token_accuracy": 0.6397761654108762,
      "num_tokens": 8778825.0,
      "step": 4450
    },
    {
      "entropy": 1.4875497985631227,
      "epoch": 0.12775160047548803,
      "grad_norm": 9.25,
      "learning_rate": 4.9689022741295e-05,
      "loss": 1.5682934761047362,
      "mean_token_accuracy": 0.6590356089174747,
      "num_tokens": 8797802.0,
      "step": 4460
    },
    {
      "entropy": 1.5024794034659863,
      "epoch": 0.12803803904157657,
      "grad_norm": 9.6875,
      "learning_rate": 4.968722583018358e-05,
      "loss": 1.5422343254089355,
      "mean_token_accuracy": 0.6661927476525307,
      "num_tokens": 8815979.0,
      "step": 4470
    },
    {
      "entropy": 1.6546546783298255,
      "epoch": 0.1283244776076651,
      "grad_norm": 9.4375,
      "learning_rate": 4.968542377517582e-05,
      "loss": 1.6528236389160156,
      "mean_token_accuracy": 0.6444402404129506,
      "num_tokens": 8836753.0,
      "step": 4480
    },
    {
      "entropy": 1.5071230851113797,
      "epoch": 0.12861091617375364,
      "grad_norm": 10.875,
      "learning_rate": 4.968361657664721e-05,
      "loss": 1.4883986473083497,
      "mean_token_accuracy": 0.6741204392164946,
      "num_tokens": 8855993.0,
      "step": 4490
    },
    {
      "entropy": 1.3650911824777723,
      "epoch": 0.12889735473984218,
      "grad_norm": 8.4375,
      "learning_rate": 4.9681804234974285e-05,
      "loss": 1.5139545440673827,
      "mean_token_accuracy": 0.6692131817340851,
      "num_tokens": 8875847.0,
      "step": 4500
    },
    {
      "entropy": 1.5094955375418067,
      "epoch": 0.12918379330593072,
      "grad_norm": 8.5,
      "learning_rate": 4.967998675053469e-05,
      "loss": 1.6065097808837892,
      "mean_token_accuracy": 0.654398288205266,
      "num_tokens": 8895189.0,
      "step": 4510
    },
    {
      "entropy": 1.5785987384617328,
      "epoch": 0.12947023187201925,
      "grad_norm": 8.625,
      "learning_rate": 4.9678164123707105e-05,
      "loss": 1.6469533920288086,
      "mean_token_accuracy": 0.6441236600279808,
      "num_tokens": 8915552.0,
      "step": 4520
    },
    {
      "entropy": 1.5994109526276588,
      "epoch": 0.1297566704381078,
      "grad_norm": 8.4375,
      "learning_rate": 4.9676336354871313e-05,
      "loss": 1.646183967590332,
      "mean_token_accuracy": 0.6489069607108832,
      "num_tokens": 8935329.0,
      "step": 4530
    },
    {
      "entropy": 1.6266275960952044,
      "epoch": 0.13004310900419633,
      "grad_norm": 10.25,
      "learning_rate": 4.967450344440813e-05,
      "loss": 1.6045013427734376,
      "mean_token_accuracy": 0.6422398079186677,
      "num_tokens": 8955819.0,
      "step": 4540
    },
    {
      "entropy": 1.4819649696350097,
      "epoch": 0.13032954757028486,
      "grad_norm": 8.0625,
      "learning_rate": 4.9672665392699484e-05,
      "loss": 1.5252309799194337,
      "mean_token_accuracy": 0.663091566041112,
      "num_tokens": 8975896.0,
      "step": 4550
    },
    {
      "entropy": 1.678335817158222,
      "epoch": 0.1306159861363734,
      "grad_norm": 10.125,
      "learning_rate": 4.967082220012835e-05,
      "loss": 1.7590364456176757,
      "mean_token_accuracy": 0.6407991398125887,
      "num_tokens": 8995151.0,
      "step": 4560
    },
    {
      "entropy": 1.4536094807088376,
      "epoch": 0.13090242470246194,
      "grad_norm": 10.8125,
      "learning_rate": 4.9668973867078796e-05,
      "loss": 1.5351222991943358,
      "mean_token_accuracy": 0.6603596534579992,
      "num_tokens": 9014246.0,
      "step": 4570
    },
    {
      "entropy": 1.5587512321770192,
      "epoch": 0.13118886326855048,
      "grad_norm": 7.34375,
      "learning_rate": 4.966712039393592e-05,
      "loss": 1.5743330001831055,
      "mean_token_accuracy": 0.6638579566031695,
      "num_tokens": 9033312.0,
      "step": 4580
    },
    {
      "entropy": 1.422217233479023,
      "epoch": 0.131475301834639,
      "grad_norm": 8.75,
      "learning_rate": 4.9665261781085924e-05,
      "loss": 1.4170184135437012,
      "mean_token_accuracy": 0.6837819341570139,
      "num_tokens": 9052197.0,
      "step": 4590
    },
    {
      "entropy": 1.4408176511526107,
      "epoch": 0.13176174040072755,
      "grad_norm": 10.4375,
      "learning_rate": 4.966339802891609e-05,
      "loss": 1.5235408782958983,
      "mean_token_accuracy": 0.6589519329369068,
      "num_tokens": 9072361.0,
      "step": 4600
    },
    {
      "entropy": 1.5640794448554516,
      "epoch": 0.1320481789668161,
      "grad_norm": 10.625,
      "learning_rate": 4.966152913781473e-05,
      "loss": 1.6959920883178712,
      "mean_token_accuracy": 0.6450260996818542,
      "num_tokens": 9092265.0,
      "step": 4610
    },
    {
      "entropy": 1.6374080155044795,
      "epoch": 0.13233461753290462,
      "grad_norm": 10.0625,
      "learning_rate": 4.9659655108171276e-05,
      "loss": 1.6527341842651366,
      "mean_token_accuracy": 0.646100414916873,
      "num_tokens": 9111538.0,
      "step": 4620
    },
    {
      "entropy": 1.6406228937208653,
      "epoch": 0.13262105609899316,
      "grad_norm": 8.8125,
      "learning_rate": 4.965777594037619e-05,
      "loss": 1.6782005310058594,
      "mean_token_accuracy": 0.6437648221850395,
      "num_tokens": 9131157.0,
      "step": 4630
    },
    {
      "entropy": 1.5327306509017944,
      "epoch": 0.1329074946650817,
      "grad_norm": 10.375,
      "learning_rate": 4.965589163482102e-05,
      "loss": 1.5165156364440917,
      "mean_token_accuracy": 0.6648626908659935,
      "num_tokens": 9150635.0,
      "step": 4640
    },
    {
      "entropy": 1.4707805193960666,
      "epoch": 0.13319393323117024,
      "grad_norm": 8.4375,
      "learning_rate": 4.96540021918984e-05,
      "loss": 1.5626633644104004,
      "mean_token_accuracy": 0.6587305165827274,
      "num_tokens": 9171534.0,
      "step": 4650
    },
    {
      "entropy": 1.4669506411999464,
      "epoch": 0.13348037179725877,
      "grad_norm": 14.0,
      "learning_rate": 4.965210761200199e-05,
      "loss": 1.6072134017944335,
      "mean_token_accuracy": 0.6649957228451967,
      "num_tokens": 9191107.0,
      "step": 4660
    },
    {
      "entropy": 1.5567944549024104,
      "epoch": 0.1337668103633473,
      "grad_norm": 9.4375,
      "learning_rate": 4.965020789552658e-05,
      "loss": 1.606149673461914,
      "mean_token_accuracy": 0.6558465506881476,
      "num_tokens": 9210231.0,
      "step": 4670
    },
    {
      "entropy": 1.4961183398962021,
      "epoch": 0.13405324892943585,
      "grad_norm": 8.6875,
      "learning_rate": 4.9648303042867984e-05,
      "loss": 1.506659984588623,
      "mean_token_accuracy": 0.6736569181084633,
      "num_tokens": 9229021.0,
      "step": 4680
    },
    {
      "entropy": 1.641214369982481,
      "epoch": 0.13433968749552438,
      "grad_norm": 8.125,
      "learning_rate": 4.9646393054423114e-05,
      "loss": 1.721603775024414,
      "mean_token_accuracy": 0.639400390535593,
      "num_tokens": 9248239.0,
      "step": 4690
    },
    {
      "entropy": 1.6699512496590614,
      "epoch": 0.13462612606161292,
      "grad_norm": 9.6875,
      "learning_rate": 4.9644477930589914e-05,
      "loss": 1.6996458053588868,
      "mean_token_accuracy": 0.6440739132463932,
      "num_tokens": 9266813.0,
      "step": 4700
    },
    {
      "entropy": 1.5381789475679397,
      "epoch": 0.1349125646277015,
      "grad_norm": 9.3125,
      "learning_rate": 4.9642557671767465e-05,
      "loss": 1.5855304718017578,
      "mean_token_accuracy": 0.6539187822490931,
      "num_tokens": 9286202.0,
      "step": 4710
    },
    {
      "entropy": 1.4064394481480123,
      "epoch": 0.13519900319379002,
      "grad_norm": 8.1875,
      "learning_rate": 4.9640632278355844e-05,
      "loss": 1.4934977531433105,
      "mean_token_accuracy": 0.6751538056880235,
      "num_tokens": 9305778.0,
      "step": 4720
    },
    {
      "entropy": 1.5691715724766255,
      "epoch": 0.13548544175987856,
      "grad_norm": 8.875,
      "learning_rate": 4.963870175075624e-05,
      "loss": 1.6240917205810548,
      "mean_token_accuracy": 0.6488645177334547,
      "num_tokens": 9325210.0,
      "step": 4730
    },
    {
      "entropy": 1.497089508920908,
      "epoch": 0.1357718803259671,
      "grad_norm": 9.9375,
      "learning_rate": 4.963676608937091e-05,
      "loss": 1.581840419769287,
      "mean_token_accuracy": 0.6479581639170646,
      "num_tokens": 9345284.0,
      "step": 4740
    },
    {
      "entropy": 1.4659678988158702,
      "epoch": 0.13605831889205564,
      "grad_norm": 10.25,
      "learning_rate": 4.9634825294603165e-05,
      "loss": 1.485072422027588,
      "mean_token_accuracy": 0.6772671807557344,
      "num_tokens": 9366584.0,
      "step": 4750
    },
    {
      "entropy": 1.5600997146219016,
      "epoch": 0.13634475745814417,
      "grad_norm": 11.375,
      "learning_rate": 4.9632879366857396e-05,
      "loss": 1.6657369613647461,
      "mean_token_accuracy": 0.6452657295390963,
      "num_tokens": 9387572.0,
      "step": 4760
    },
    {
      "entropy": 1.5218887124210596,
      "epoch": 0.1366311960242327,
      "grad_norm": 8.0625,
      "learning_rate": 4.963092830653907e-05,
      "loss": 1.5555039405822755,
      "mean_token_accuracy": 0.6518874477595091,
      "num_tokens": 9407440.0,
      "step": 4770
    },
    {
      "entropy": 1.5266014069318772,
      "epoch": 0.13691763459032125,
      "grad_norm": 9.5625,
      "learning_rate": 4.9628972114054706e-05,
      "loss": 1.5886033058166504,
      "mean_token_accuracy": 0.6657060205936431,
      "num_tokens": 9427120.0,
      "step": 4780
    },
    {
      "entropy": 1.504212035238743,
      "epoch": 0.13720407315640978,
      "grad_norm": 8.125,
      "learning_rate": 4.96270107898119e-05,
      "loss": 1.584134578704834,
      "mean_token_accuracy": 0.6573188021779061,
      "num_tokens": 9446852.0,
      "step": 4790
    },
    {
      "entropy": 1.7013119705021382,
      "epoch": 0.13749051172249832,
      "grad_norm": 8.8125,
      "learning_rate": 4.962504433421933e-05,
      "loss": 1.7435585021972657,
      "mean_token_accuracy": 0.6280934553593397,
      "num_tokens": 9466739.0,
      "step": 4800
    },
    {
      "entropy": 1.4621875803917646,
      "epoch": 0.13777695028858686,
      "grad_norm": 8.25,
      "learning_rate": 4.962307274768672e-05,
      "loss": 1.492708969116211,
      "mean_token_accuracy": 0.6729666709899902,
      "num_tokens": 9485680.0,
      "step": 4810
    },
    {
      "entropy": 1.4921036452054977,
      "epoch": 0.1380633888546754,
      "grad_norm": 10.625,
      "learning_rate": 4.9621096030624883e-05,
      "loss": 1.5334936141967774,
      "mean_token_accuracy": 0.6603573184460402,
      "num_tokens": 9506293.0,
      "step": 4820
    },
    {
      "entropy": 1.4879472389817239,
      "epoch": 0.13834982742076393,
      "grad_norm": 13.6875,
      "learning_rate": 4.961911418344569e-05,
      "loss": 1.7079967498779296,
      "mean_token_accuracy": 0.6462376203387976,
      "num_tokens": 9526251.0,
      "step": 4830
    },
    {
      "entropy": 1.5331168532371522,
      "epoch": 0.13863626598685247,
      "grad_norm": 11.75,
      "learning_rate": 4.961712720656208e-05,
      "loss": 1.544976806640625,
      "mean_token_accuracy": 0.6682040221989155,
      "num_tokens": 9545654.0,
      "step": 4840
    },
    {
      "entropy": 1.5735587041825057,
      "epoch": 0.138922704552941,
      "grad_norm": 8.75,
      "learning_rate": 4.961513510038807e-05,
      "loss": 1.6300994873046875,
      "mean_token_accuracy": 0.6456163562834263,
      "num_tokens": 9565414.0,
      "step": 4850
    },
    {
      "entropy": 1.403120832145214,
      "epoch": 0.13920914311902954,
      "grad_norm": 8.25,
      "learning_rate": 4.961313786533874e-05,
      "loss": 1.4635492324829102,
      "mean_token_accuracy": 0.6760737411677837,
      "num_tokens": 9584757.0,
      "step": 4860
    },
    {
      "entropy": 1.519579280912876,
      "epoch": 0.13949558168511808,
      "grad_norm": 9.75,
      "learning_rate": 4.961113550183024e-05,
      "loss": 1.5451878547668456,
      "mean_token_accuracy": 0.6640205007046461,
      "num_tokens": 9604619.0,
      "step": 4870
    },
    {
      "entropy": 1.5056984916329383,
      "epoch": 0.13978202025120662,
      "grad_norm": 8.0625,
      "learning_rate": 4.9609128010279784e-05,
      "loss": 1.5595246315002442,
      "mean_token_accuracy": 0.6584958035498858,
      "num_tokens": 9623688.0,
      "step": 4880
    },
    {
      "entropy": 1.622643904015422,
      "epoch": 0.14006845881729516,
      "grad_norm": 8.6875,
      "learning_rate": 4.960711539110566e-05,
      "loss": 1.6314786911010741,
      "mean_token_accuracy": 0.64293511249125,
      "num_tokens": 9645016.0,
      "step": 4890
    },
    {
      "entropy": 1.4993349730968475,
      "epoch": 0.1403548973833837,
      "grad_norm": 10.0625,
      "learning_rate": 4.960509764472723e-05,
      "loss": 1.5718796730041504,
      "mean_token_accuracy": 0.6560567699372768,
      "num_tokens": 9664646.0,
      "step": 4900
    },
    {
      "entropy": 1.5645937040448188,
      "epoch": 0.14064133594947223,
      "grad_norm": 12.875,
      "learning_rate": 4.960307477156491e-05,
      "loss": 1.5769742012023926,
      "mean_token_accuracy": 0.6528674773871899,
      "num_tokens": 9685110.0,
      "step": 4910
    },
    {
      "entropy": 1.3918135400861502,
      "epoch": 0.14092777451556077,
      "grad_norm": 8.3125,
      "learning_rate": 4.960104677204018e-05,
      "loss": 1.4424909591674804,
      "mean_token_accuracy": 0.6750841304659844,
      "num_tokens": 9704859.0,
      "step": 4920
    },
    {
      "entropy": 1.5218386985361576,
      "epoch": 0.1412142130816493,
      "grad_norm": 6.09375,
      "learning_rate": 4.959901364657562e-05,
      "loss": 1.5807250022888184,
      "mean_token_accuracy": 0.6635899852961302,
      "num_tokens": 9724561.0,
      "step": 4930
    },
    {
      "entropy": 1.5914435897022485,
      "epoch": 0.14150065164773784,
      "grad_norm": 8.3125,
      "learning_rate": 4.9596975395594844e-05,
      "loss": 1.603508186340332,
      "mean_token_accuracy": 0.6510025925934315,
      "num_tokens": 9744644.0,
      "step": 4940
    },
    {
      "entropy": 1.4987780340015888,
      "epoch": 0.14178709021382638,
      "grad_norm": 12.0,
      "learning_rate": 4.9594932019522554e-05,
      "loss": 1.564766025543213,
      "mean_token_accuracy": 0.6559693459421396,
      "num_tokens": 9763835.0,
      "step": 4950
    },
    {
      "entropy": 1.5716330505907536,
      "epoch": 0.14207352877991491,
      "grad_norm": 8.0625,
      "learning_rate": 4.9592883518784517e-05,
      "loss": 1.6398065567016602,
      "mean_token_accuracy": 0.6363564360886812,
      "num_tokens": 9783953.0,
      "step": 4960
    },
    {
      "entropy": 1.5598301596939563,
      "epoch": 0.14235996734600345,
      "grad_norm": 11.4375,
      "learning_rate": 4.959082989380754e-05,
      "loss": 1.623297882080078,
      "mean_token_accuracy": 0.6506641361862421,
      "num_tokens": 9804743.0,
      "step": 4970
    },
    {
      "entropy": 1.4122194364666938,
      "epoch": 0.14264640591209202,
      "grad_norm": 8.25,
      "learning_rate": 4.9588771145019565e-05,
      "loss": 1.4657936096191406,
      "mean_token_accuracy": 0.6727658096700907,
      "num_tokens": 9823929.0,
      "step": 4980
    },
    {
      "entropy": 1.5104284398257732,
      "epoch": 0.14293284447818055,
      "grad_norm": 8.5,
      "learning_rate": 4.958670727284952e-05,
      "loss": 1.5763837814331054,
      "mean_token_accuracy": 0.6548960711807013,
      "num_tokens": 9843241.0,
      "step": 4990
    },
    {
      "entropy": 1.4615675970911979,
      "epoch": 0.1432192830442691,
      "grad_norm": 7.90625,
      "learning_rate": 4.958463827772745e-05,
      "loss": 1.4633464813232422,
      "mean_token_accuracy": 0.6726066563278437,
      "num_tokens": 9864370.0,
      "step": 5000
    },
    {
      "entropy": 1.3798364020884037,
      "epoch": 0.14350572161035763,
      "grad_norm": 10.9375,
      "learning_rate": 4.958256416008447e-05,
      "loss": 1.399475383758545,
      "mean_token_accuracy": 0.6874329604208469,
      "num_tokens": 9885744.0,
      "step": 5010
    },
    {
      "entropy": 1.4340853445231914,
      "epoch": 0.14379216017644617,
      "grad_norm": 8.9375,
      "learning_rate": 4.958048492035272e-05,
      "loss": 1.59286470413208,
      "mean_token_accuracy": 0.6604187369346619,
      "num_tokens": 9906496.0,
      "step": 5020
    },
    {
      "entropy": 1.463975927233696,
      "epoch": 0.1440785987425347,
      "grad_norm": 10.125,
      "learning_rate": 4.9578400558965464e-05,
      "loss": 1.516213607788086,
      "mean_token_accuracy": 0.6712160520255566,
      "num_tokens": 9926960.0,
      "step": 5030
    },
    {
      "entropy": 1.480278616771102,
      "epoch": 0.14436503730862324,
      "grad_norm": 8.875,
      "learning_rate": 4.9576311076356996e-05,
      "loss": 1.5378478050231934,
      "mean_token_accuracy": 0.6715181965380908,
      "num_tokens": 9945994.0,
      "step": 5040
    },
    {
      "entropy": 1.4886713460087777,
      "epoch": 0.14465147587471178,
      "grad_norm": 7.375,
      "learning_rate": 4.957421647296269e-05,
      "loss": 1.5202993392944335,
      "mean_token_accuracy": 0.6785048861056566,
      "num_tokens": 9966115.0,
      "step": 5050
    },
    {
      "entropy": 1.5940536327660084,
      "epoch": 0.14493791444080031,
      "grad_norm": 9.125,
      "learning_rate": 4.957211674921897e-05,
      "loss": 1.6539093017578126,
      "mean_token_accuracy": 0.6506012924015522,
      "num_tokens": 9987080.0,
      "step": 5060
    },
    {
      "entropy": 1.4804301261901855,
      "epoch": 0.14522435300688885,
      "grad_norm": 10.0625,
      "learning_rate": 4.957001190556334e-05,
      "loss": 1.5193848609924316,
      "mean_token_accuracy": 0.6621853016316891,
      "num_tokens": 10007396.0,
      "step": 5070
    },
    {
      "entropy": 1.4721652269363403,
      "epoch": 0.1455107915729774,
      "grad_norm": 11.1875,
      "learning_rate": 4.956790194243439e-05,
      "loss": 1.5131440162658691,
      "mean_token_accuracy": 0.6650624644011259,
      "num_tokens": 10026695.0,
      "step": 5080
    },
    {
      "entropy": 1.45464109107852,
      "epoch": 0.14579723013906593,
      "grad_norm": 10.9375,
      "learning_rate": 4.956578686027175e-05,
      "loss": 1.5026907920837402,
      "mean_token_accuracy": 0.6650346931070089,
      "num_tokens": 10047006.0,
      "step": 5090
    },
    {
      "entropy": 1.491887776553631,
      "epoch": 0.14608366870515446,
      "grad_norm": 8.1875,
      "learning_rate": 4.9563666659516104e-05,
      "loss": 1.6299636840820313,
      "mean_token_accuracy": 0.6577295735478401,
      "num_tokens": 10067044.0,
      "step": 5100
    },
    {
      "entropy": 1.6270928215235472,
      "epoch": 0.146370107271243,
      "grad_norm": 11.6875,
      "learning_rate": 4.956154134060925e-05,
      "loss": 1.6715662002563476,
      "mean_token_accuracy": 0.639185581356287,
      "num_tokens": 10086407.0,
      "step": 5110
    },
    {
      "entropy": 1.532237235084176,
      "epoch": 0.14665654583733154,
      "grad_norm": 9.875,
      "learning_rate": 4.9559410903994015e-05,
      "loss": 1.6115522384643555,
      "mean_token_accuracy": 0.6538224563002586,
      "num_tokens": 10105594.0,
      "step": 5120
    },
    {
      "entropy": 1.469840183854103,
      "epoch": 0.14694298440342007,
      "grad_norm": 7.6875,
      "learning_rate": 4.95572753501143e-05,
      "loss": 1.3854816436767579,
      "mean_token_accuracy": 0.6991850856691599,
      "num_tokens": 10125298.0,
      "step": 5130
    },
    {
      "entropy": 1.4622601751238107,
      "epoch": 0.1472294229695086,
      "grad_norm": 8.25,
      "learning_rate": 4.955513467941507e-05,
      "loss": 1.5373471260070801,
      "mean_token_accuracy": 0.6776952587068081,
      "num_tokens": 10145891.0,
      "step": 5140
    },
    {
      "entropy": 1.3798430625349283,
      "epoch": 0.14751586153559715,
      "grad_norm": 8.0625,
      "learning_rate": 4.955298889234238e-05,
      "loss": 1.4437210083007812,
      "mean_token_accuracy": 0.6797787621617317,
      "num_tokens": 10167404.0,
      "step": 5150
    },
    {
      "entropy": 1.611018529906869,
      "epoch": 0.14780230010168569,
      "grad_norm": 9.8125,
      "learning_rate": 4.955083798934331e-05,
      "loss": 1.7299531936645507,
      "mean_token_accuracy": 0.6256252124905586,
      "num_tokens": 10186809.0,
      "step": 5160
    },
    {
      "entropy": 1.586063664406538,
      "epoch": 0.14808873866777422,
      "grad_norm": 6.9375,
      "learning_rate": 4.954868197086604e-05,
      "loss": 1.6296968460083008,
      "mean_token_accuracy": 0.6475865747779608,
      "num_tokens": 10208143.0,
      "step": 5170
    },
    {
      "entropy": 1.4521633706986905,
      "epoch": 0.14837517723386276,
      "grad_norm": 6.15625,
      "learning_rate": 4.95465208373598e-05,
      "loss": 1.4644355773925781,
      "mean_token_accuracy": 0.6826090008020401,
      "num_tokens": 10227213.0,
      "step": 5180
    },
    {
      "entropy": 1.4087909482419492,
      "epoch": 0.1486616157999513,
      "grad_norm": 8.5,
      "learning_rate": 4.954435458927489e-05,
      "loss": 1.4978672981262207,
      "mean_token_accuracy": 0.682854725793004,
      "num_tokens": 10245827.0,
      "step": 5190
    },
    {
      "entropy": 1.5200366478413343,
      "epoch": 0.14894805436603983,
      "grad_norm": 8.9375,
      "learning_rate": 4.954218322706268e-05,
      "loss": 1.6043817520141601,
      "mean_token_accuracy": 0.6415193796157836,
      "num_tokens": 10266167.0,
      "step": 5200
    },
    {
      "entropy": 1.4410448588430882,
      "epoch": 0.14923449293212837,
      "grad_norm": 7.125,
      "learning_rate": 4.95400067511756e-05,
      "loss": 1.4061705589294433,
      "mean_token_accuracy": 0.6966239595785737,
      "num_tokens": 10284596.0,
      "step": 5210
    },
    {
      "entropy": 1.4189201422035693,
      "epoch": 0.1495209314982169,
      "grad_norm": 8.8125,
      "learning_rate": 4.953782516206713e-05,
      "loss": 1.4706315040588378,
      "mean_token_accuracy": 0.6728576749563218,
      "num_tokens": 10304052.0,
      "step": 5220
    },
    {
      "entropy": 1.474114514142275,
      "epoch": 0.14980737006430545,
      "grad_norm": 10.6875,
      "learning_rate": 4.953563846019185e-05,
      "loss": 1.505852508544922,
      "mean_token_accuracy": 0.6601380426436663,
      "num_tokens": 10323963.0,
      "step": 5230
    },
    {
      "entropy": 1.4801891282200814,
      "epoch": 0.150093808630394,
      "grad_norm": 11.375,
      "learning_rate": 4.953344664600538e-05,
      "loss": 1.5376943588256835,
      "mean_token_accuracy": 0.6654257832095027,
      "num_tokens": 10342977.0,
      "step": 5240
    },
    {
      "entropy": 1.4171194434165955,
      "epoch": 0.15038024719648255,
      "grad_norm": 7.75,
      "learning_rate": 4.9531249719964424e-05,
      "loss": 1.4434309005737305,
      "mean_token_accuracy": 0.688833561912179,
      "num_tokens": 10362549.0,
      "step": 5250
    },
    {
      "entropy": 1.3564347431063652,
      "epoch": 0.15066668576257108,
      "grad_norm": 7.375,
      "learning_rate": 4.9529047682526716e-05,
      "loss": 1.4937264442443847,
      "mean_token_accuracy": 0.6793396513909101,
      "num_tokens": 10380442.0,
      "step": 5260
    },
    {
      "entropy": 1.4401488423347473,
      "epoch": 0.15095312432865962,
      "grad_norm": 7.15625,
      "learning_rate": 4.95268405341511e-05,
      "loss": 1.5216401100158692,
      "mean_token_accuracy": 0.6716520003974438,
      "num_tokens": 10400570.0,
      "step": 5270
    },
    {
      "entropy": 1.449067098274827,
      "epoch": 0.15123956289474816,
      "grad_norm": 8.3125,
      "learning_rate": 4.952462827529745e-05,
      "loss": 1.4734996795654296,
      "mean_token_accuracy": 0.685632624477148,
      "num_tokens": 10419978.0,
      "step": 5280
    },
    {
      "entropy": 1.5881035834550858,
      "epoch": 0.1515260014608367,
      "grad_norm": 8.0,
      "learning_rate": 4.952241090642673e-05,
      "loss": 1.5667628288269042,
      "mean_token_accuracy": 0.6578472722321749,
      "num_tokens": 10439581.0,
      "step": 5290
    },
    {
      "entropy": 1.3544663984328509,
      "epoch": 0.15181244002692523,
      "grad_norm": 8.3125,
      "learning_rate": 4.9520188428000945e-05,
      "loss": 1.4386282920837403,
      "mean_token_accuracy": 0.6912245389074088,
      "num_tokens": 10458254.0,
      "step": 5300
    },
    {
      "entropy": 1.4365281619131565,
      "epoch": 0.15209887859301377,
      "grad_norm": 7.53125,
      "learning_rate": 4.951796084048318e-05,
      "loss": 1.466072940826416,
      "mean_token_accuracy": 0.6794659815728664,
      "num_tokens": 10477263.0,
      "step": 5310
    },
    {
      "entropy": 1.5558462772518395,
      "epoch": 0.1523853171591023,
      "grad_norm": 8.125,
      "learning_rate": 4.9515728144337586e-05,
      "loss": 1.5894271850585937,
      "mean_token_accuracy": 0.6573893807828426,
      "num_tokens": 10498470.0,
      "step": 5320
    },
    {
      "entropy": 1.4525925662368535,
      "epoch": 0.15267175572519084,
      "grad_norm": 10.875,
      "learning_rate": 4.951349034002937e-05,
      "loss": 1.5502291679382325,
      "mean_token_accuracy": 0.665329060703516,
      "num_tokens": 10517680.0,
      "step": 5330
    },
    {
      "entropy": 1.513069298863411,
      "epoch": 0.15295819429127938,
      "grad_norm": 7.59375,
      "learning_rate": 4.9511247428024806e-05,
      "loss": 1.5275460243225099,
      "mean_token_accuracy": 0.6618064153939486,
      "num_tokens": 10537379.0,
      "step": 5340
    },
    {
      "entropy": 1.4216542765498161,
      "epoch": 0.15324463285736792,
      "grad_norm": 11.375,
      "learning_rate": 4.950899940879123e-05,
      "loss": 1.4662826538085938,
      "mean_token_accuracy": 0.6781281795352697,
      "num_tokens": 10556156.0,
      "step": 5350
    },
    {
      "entropy": 1.5469950869679452,
      "epoch": 0.15353107142345646,
      "grad_norm": 10.125,
      "learning_rate": 4.950674628279706e-05,
      "loss": 1.6273674011230468,
      "mean_token_accuracy": 0.6543409667909146,
      "num_tokens": 10576813.0,
      "step": 5360
    },
    {
      "entropy": 1.4365376845002173,
      "epoch": 0.153817509989545,
      "grad_norm": 8.5,
      "learning_rate": 4.9504488050511747e-05,
      "loss": 1.449380588531494,
      "mean_token_accuracy": 0.6771000474691391,
      "num_tokens": 10596672.0,
      "step": 5370
    },
    {
      "entropy": 1.4803869400173426,
      "epoch": 0.15410394855563353,
      "grad_norm": 10.5,
      "learning_rate": 4.950222471240583e-05,
      "loss": 1.6092044830322265,
      "mean_token_accuracy": 0.6441299144178629,
      "num_tokens": 10615364.0,
      "step": 5380
    },
    {
      "entropy": 1.577774827182293,
      "epoch": 0.15439038712172207,
      "grad_norm": 8.8125,
      "learning_rate": 4.94999562689509e-05,
      "loss": 1.5489473342895508,
      "mean_token_accuracy": 0.670935046672821,
      "num_tokens": 10636236.0,
      "step": 5390
    },
    {
      "entropy": 1.5271467573940753,
      "epoch": 0.1546768256878106,
      "grad_norm": 9.125,
      "learning_rate": 4.949768272061962e-05,
      "loss": 1.5601237297058106,
      "mean_token_accuracy": 0.6636587370187044,
      "num_tokens": 10656497.0,
      "step": 5400
    },
    {
      "entropy": 1.4755383789539338,
      "epoch": 0.15496326425389914,
      "grad_norm": 12.375,
      "learning_rate": 4.949540406788571e-05,
      "loss": 1.617512321472168,
      "mean_token_accuracy": 0.6448015566915274,
      "num_tokens": 10676827.0,
      "step": 5410
    },
    {
      "entropy": 1.5340559579432012,
      "epoch": 0.15524970281998768,
      "grad_norm": 10.375,
      "learning_rate": 4.9493120311223964e-05,
      "loss": 1.5395313262939454,
      "mean_token_accuracy": 0.6689881380647421,
      "num_tokens": 10696596.0,
      "step": 5420
    },
    {
      "entropy": 1.4465421587228775,
      "epoch": 0.15553614138607622,
      "grad_norm": 8.3125,
      "learning_rate": 4.949083145111021e-05,
      "loss": 1.4482954025268555,
      "mean_token_accuracy": 0.6931252546608448,
      "num_tokens": 10716557.0,
      "step": 5430
    },
    {
      "entropy": 1.4735890928655864,
      "epoch": 0.15582257995216475,
      "grad_norm": 7.6875,
      "learning_rate": 4.948853748802139e-05,
      "loss": 1.6091737747192383,
      "mean_token_accuracy": 0.6571174290031194,
      "num_tokens": 10736652.0,
      "step": 5440
    },
    {
      "entropy": 1.4969568315893411,
      "epoch": 0.1561090185182533,
      "grad_norm": 10.3125,
      "learning_rate": 4.948623842243546e-05,
      "loss": 1.5009947776794434,
      "mean_token_accuracy": 0.6735967800021172,
      "num_tokens": 10756238.0,
      "step": 5450
    },
    {
      "entropy": 1.3353195700794458,
      "epoch": 0.15639545708434183,
      "grad_norm": 17.0,
      "learning_rate": 4.9483934254831463e-05,
      "loss": 1.3614046096801757,
      "mean_token_accuracy": 0.6964558508247137,
      "num_tokens": 10775753.0,
      "step": 5460
    },
    {
      "entropy": 1.316618399694562,
      "epoch": 0.15668189565043036,
      "grad_norm": 8.5,
      "learning_rate": 4.9481624985689506e-05,
      "loss": 1.4229409217834472,
      "mean_token_accuracy": 0.6905061881989241,
      "num_tokens": 10794651.0,
      "step": 5470
    },
    {
      "entropy": 1.445284367352724,
      "epoch": 0.1569683342165189,
      "grad_norm": 11.5625,
      "learning_rate": 4.947931061549075e-05,
      "loss": 1.5132107734680176,
      "mean_token_accuracy": 0.6724000908434391,
      "num_tokens": 10814677.0,
      "step": 5480
    },
    {
      "entropy": 1.4952793795615436,
      "epoch": 0.15725477278260744,
      "grad_norm": 8.5625,
      "learning_rate": 4.947699114471743e-05,
      "loss": 1.4845052719116212,
      "mean_token_accuracy": 0.6772314850240946,
      "num_tokens": 10833035.0,
      "step": 5490
    },
    {
      "entropy": 1.4620481140911579,
      "epoch": 0.15754121134869598,
      "grad_norm": 10.6875,
      "learning_rate": 4.947466657385284e-05,
      "loss": 1.439906406402588,
      "mean_token_accuracy": 0.679519497230649,
      "num_tokens": 10852990.0,
      "step": 5500
    },
    {
      "entropy": 1.3291530303657055,
      "epoch": 0.15782764991478454,
      "grad_norm": 9.25,
      "learning_rate": 4.947233690338131e-05,
      "loss": 1.3231815338134765,
      "mean_token_accuracy": 0.6966482844203711,
      "num_tokens": 10872874.0,
      "step": 5510
    },
    {
      "entropy": 1.4696352612227201,
      "epoch": 0.15811408848087308,
      "grad_norm": 7.9375,
      "learning_rate": 4.947000213378828e-05,
      "loss": 1.5467846870422364,
      "mean_token_accuracy": 0.6598045524209738,
      "num_tokens": 10892742.0,
      "step": 5520
    },
    {
      "entropy": 1.4622251082211732,
      "epoch": 0.15840052704696161,
      "grad_norm": 9.4375,
      "learning_rate": 4.9467662265560235e-05,
      "loss": 1.5740177154541015,
      "mean_token_accuracy": 0.6724891006946564,
      "num_tokens": 10910712.0,
      "step": 5530
    },
    {
      "entropy": 1.5368123035877943,
      "epoch": 0.15868696561305015,
      "grad_norm": 8.9375,
      "learning_rate": 4.946531729918469e-05,
      "loss": 1.6157594680786134,
      "mean_token_accuracy": 0.6584647875279188,
      "num_tokens": 10930699.0,
      "step": 5540
    },
    {
      "entropy": 1.471268993616104,
      "epoch": 0.1589734041791387,
      "grad_norm": 7.0,
      "learning_rate": 4.946296723515026e-05,
      "loss": 1.5040043830871581,
      "mean_token_accuracy": 0.6715310331434011,
      "num_tokens": 10951002.0,
      "step": 5550
    },
    {
      "entropy": 1.498399916291237,
      "epoch": 0.15925984274522723,
      "grad_norm": 11.5625,
      "learning_rate": 4.946061207394662e-05,
      "loss": 1.5332067489624024,
      "mean_token_accuracy": 0.6573239173740149,
      "num_tokens": 10971945.0,
      "step": 5560
    },
    {
      "entropy": 1.562719462811947,
      "epoch": 0.15954628131131576,
      "grad_norm": 8.9375,
      "learning_rate": 4.945825181606448e-05,
      "loss": 1.6053812026977539,
      "mean_token_accuracy": 0.6590887267142534,
      "num_tokens": 10992197.0,
      "step": 5570
    },
    {
      "entropy": 1.4622655007988214,
      "epoch": 0.1598327198774043,
      "grad_norm": 9.1875,
      "learning_rate": 4.9455886461995646e-05,
      "loss": 1.471405029296875,
      "mean_token_accuracy": 0.6774644616991281,
      "num_tokens": 11012688.0,
      "step": 5580
    },
    {
      "entropy": 1.4587391506880523,
      "epoch": 0.16011915844349284,
      "grad_norm": 8.6875,
      "learning_rate": 4.9453516012232966e-05,
      "loss": 1.51792573928833,
      "mean_token_accuracy": 0.6790623251348734,
      "num_tokens": 11032883.0,
      "step": 5590
    },
    {
      "entropy": 1.3946425825357438,
      "epoch": 0.16040559700958137,
      "grad_norm": 12.75,
      "learning_rate": 4.945114046727035e-05,
      "loss": 1.4670876502990722,
      "mean_token_accuracy": 0.6826338931918144,
      "num_tokens": 11052108.0,
      "step": 5600
    },
    {
      "entropy": 1.4580876618623733,
      "epoch": 0.1606920355756699,
      "grad_norm": 8.3125,
      "learning_rate": 4.9448759827602765e-05,
      "loss": 1.5252602577209473,
      "mean_token_accuracy": 0.675988496094942,
      "num_tokens": 11071728.0,
      "step": 5610
    },
    {
      "entropy": 1.6357592076063157,
      "epoch": 0.16097847414175845,
      "grad_norm": 8.875,
      "learning_rate": 4.9446374093726266e-05,
      "loss": 1.7047426223754882,
      "mean_token_accuracy": 0.633384408429265,
      "num_tokens": 11092369.0,
      "step": 5620
    },
    {
      "entropy": 1.4805690959095954,
      "epoch": 0.16126491270784699,
      "grad_norm": 10.5,
      "learning_rate": 4.944398326613793e-05,
      "loss": 1.5725739479064942,
      "mean_token_accuracy": 0.6648765787482261,
      "num_tokens": 11112188.0,
      "step": 5630
    },
    {
      "entropy": 1.5275758005678655,
      "epoch": 0.16155135127393552,
      "grad_norm": 9.125,
      "learning_rate": 4.944158734533594e-05,
      "loss": 1.470158004760742,
      "mean_token_accuracy": 0.6798445299267769,
      "num_tokens": 11131086.0,
      "step": 5640
    },
    {
      "entropy": 1.3688059270381927,
      "epoch": 0.16183778984002406,
      "grad_norm": 7.15625,
      "learning_rate": 4.9439186331819486e-05,
      "loss": 1.3968552589416503,
      "mean_token_accuracy": 0.6856072615832091,
      "num_tokens": 11151279.0,
      "step": 5650
    },
    {
      "entropy": 1.3473737981170415,
      "epoch": 0.1621242284061126,
      "grad_norm": 9.0,
      "learning_rate": 4.943678022608888e-05,
      "loss": 1.4398218154907227,
      "mean_token_accuracy": 0.6887608714401722,
      "num_tokens": 11171788.0,
      "step": 5660
    },
    {
      "entropy": 1.4796992752701044,
      "epoch": 0.16241066697220113,
      "grad_norm": 10.125,
      "learning_rate": 4.943436902864545e-05,
      "loss": 1.5150390625,
      "mean_token_accuracy": 0.6665746040642262,
      "num_tokens": 11190690.0,
      "step": 5670
    },
    {
      "entropy": 1.4906470637768507,
      "epoch": 0.16269710553828967,
      "grad_norm": 8.25,
      "learning_rate": 4.94319527399916e-05,
      "loss": 1.5597380638122558,
      "mean_token_accuracy": 0.6585033319890499,
      "num_tokens": 11210066.0,
      "step": 5680
    },
    {
      "entropy": 1.445562968403101,
      "epoch": 0.1629835441043782,
      "grad_norm": 6.75,
      "learning_rate": 4.942953136063079e-05,
      "loss": 1.4560264587402343,
      "mean_token_accuracy": 0.6677102066576481,
      "num_tokens": 11229810.0,
      "step": 5690
    },
    {
      "entropy": 1.48755826279521,
      "epoch": 0.16326998267046675,
      "grad_norm": 7.28125,
      "learning_rate": 4.942710489106755e-05,
      "loss": 1.5455761909484864,
      "mean_token_accuracy": 0.6614343244582415,
      "num_tokens": 11251400.0,
      "step": 5700
    },
    {
      "entropy": 1.4296566624194385,
      "epoch": 0.16355642123655528,
      "grad_norm": 8.9375,
      "learning_rate": 4.942467333180748e-05,
      "loss": 1.5093473434448241,
      "mean_token_accuracy": 0.6593443274497985,
      "num_tokens": 11271813.0,
      "step": 5710
    },
    {
      "entropy": 1.5249940477311612,
      "epoch": 0.16384285980264382,
      "grad_norm": 9.6875,
      "learning_rate": 4.94222366833572e-05,
      "loss": 1.5830157279968262,
      "mean_token_accuracy": 0.6588814780116081,
      "num_tokens": 11292623.0,
      "step": 5720
    },
    {
      "entropy": 1.4762375630438327,
      "epoch": 0.16412929836873236,
      "grad_norm": 9.4375,
      "learning_rate": 4.941979494622443e-05,
      "loss": 1.4806387901306153,
      "mean_token_accuracy": 0.6740219008177519,
      "num_tokens": 11313530.0,
      "step": 5730
    },
    {
      "entropy": 1.5683625414967537,
      "epoch": 0.1644157369348209,
      "grad_norm": 8.875,
      "learning_rate": 4.941734812091794e-05,
      "loss": 1.6385662078857421,
      "mean_token_accuracy": 0.6370994074270129,
      "num_tokens": 11332156.0,
      "step": 5740
    },
    {
      "entropy": 1.4443510707467795,
      "epoch": 0.16470217550090943,
      "grad_norm": 8.1875,
      "learning_rate": 4.9414896207947547e-05,
      "loss": 1.4884693145751953,
      "mean_token_accuracy": 0.6757148705422878,
      "num_tokens": 11351052.0,
      "step": 5750
    },
    {
      "entropy": 1.4852353297173977,
      "epoch": 0.16498861406699797,
      "grad_norm": 10.25,
      "learning_rate": 4.941243920782415e-05,
      "loss": 1.5405843734741211,
      "mean_token_accuracy": 0.6612853478640318,
      "num_tokens": 11370918.0,
      "step": 5760
    },
    {
      "entropy": 1.495937005057931,
      "epoch": 0.1652750526330865,
      "grad_norm": 9.75,
      "learning_rate": 4.940997712105969e-05,
      "loss": 1.5172706604003907,
      "mean_token_accuracy": 0.6814956789836287,
      "num_tokens": 11390959.0,
      "step": 5770
    },
    {
      "entropy": 1.4493368919938803,
      "epoch": 0.16556149119917507,
      "grad_norm": 9.3125,
      "learning_rate": 4.940750994816717e-05,
      "loss": 1.5447139739990234,
      "mean_token_accuracy": 0.671611413732171,
      "num_tokens": 11410801.0,
      "step": 5780
    },
    {
      "entropy": 1.4292007364332675,
      "epoch": 0.1658479297652636,
      "grad_norm": 10.625,
      "learning_rate": 4.940503768966066e-05,
      "loss": 1.4733223915100098,
      "mean_token_accuracy": 0.6725025560706854,
      "num_tokens": 11430375.0,
      "step": 5790
    },
    {
      "entropy": 1.5424027353525163,
      "epoch": 0.16613436833135214,
      "grad_norm": 9.1875,
      "learning_rate": 4.9402560346055295e-05,
      "loss": 1.5172248840332032,
      "mean_token_accuracy": 0.6698197159916163,
      "num_tokens": 11450377.0,
      "step": 5800
    },
    {
      "entropy": 1.461921089887619,
      "epoch": 0.16642080689744068,
      "grad_norm": 7.25,
      "learning_rate": 4.940007791786724e-05,
      "loss": 1.5484451293945312,
      "mean_token_accuracy": 0.6609391871839762,
      "num_tokens": 11470439.0,
      "step": 5810
    },
    {
      "entropy": 1.510815527662635,
      "epoch": 0.16670724546352922,
      "grad_norm": 7.9375,
      "learning_rate": 4.939759040561377e-05,
      "loss": 1.5555777549743652,
      "mean_token_accuracy": 0.6504639558494091,
      "num_tokens": 11491305.0,
      "step": 5820
    },
    {
      "entropy": 1.4381535571068524,
      "epoch": 0.16699368402961776,
      "grad_norm": 6.21875,
      "learning_rate": 4.9395097809813163e-05,
      "loss": 1.456609344482422,
      "mean_token_accuracy": 0.6757194273173809,
      "num_tokens": 11511334.0,
      "step": 5830
    },
    {
      "entropy": 1.4057760030031203,
      "epoch": 0.1672801225957063,
      "grad_norm": 7.53125,
      "learning_rate": 4.939260013098479e-05,
      "loss": 1.5399243354797363,
      "mean_token_accuracy": 0.6671665336936712,
      "num_tokens": 11531986.0,
      "step": 5840
    },
    {
      "entropy": 1.4422939270734787,
      "epoch": 0.16756656116179483,
      "grad_norm": 7.34375,
      "learning_rate": 4.939009736964908e-05,
      "loss": 1.3935338973999023,
      "mean_token_accuracy": 0.6853090669959784,
      "num_tokens": 11551264.0,
      "step": 5850
    },
    {
      "entropy": 1.3780545085668563,
      "epoch": 0.16785299972788337,
      "grad_norm": 10.625,
      "learning_rate": 4.9387589526327515e-05,
      "loss": 1.4793409347534179,
      "mean_token_accuracy": 0.6712892156094312,
      "num_tokens": 11571522.0,
      "step": 5860
    },
    {
      "entropy": 1.3406000778079032,
      "epoch": 0.1681394382939719,
      "grad_norm": 7.9375,
      "learning_rate": 4.9385076601542637e-05,
      "loss": 1.394017219543457,
      "mean_token_accuracy": 0.6953121948987245,
      "num_tokens": 11592141.0,
      "step": 5870
    },
    {
      "entropy": 1.3953971412032842,
      "epoch": 0.16842587686006044,
      "grad_norm": 10.1875,
      "learning_rate": 4.938255859581803e-05,
      "loss": 1.4208333015441894,
      "mean_token_accuracy": 0.6905218441039324,
      "num_tokens": 11611976.0,
      "step": 5880
    },
    {
      "entropy": 1.3557428650557994,
      "epoch": 0.16871231542614898,
      "grad_norm": 11.25,
      "learning_rate": 4.938003550967837e-05,
      "loss": 1.417108154296875,
      "mean_token_accuracy": 0.6836075335741043,
      "num_tokens": 11630785.0,
      "step": 5890
    },
    {
      "entropy": 1.3994893398135901,
      "epoch": 0.16899875399223752,
      "grad_norm": 7.0625,
      "learning_rate": 4.937750734364937e-05,
      "loss": 1.4636310577392577,
      "mean_token_accuracy": 0.6778811745345592,
      "num_tokens": 11649442.0,
      "step": 5900
    },
    {
      "entropy": 1.4923362039029597,
      "epoch": 0.16928519255832605,
      "grad_norm": 8.375,
      "learning_rate": 4.9374974098257804e-05,
      "loss": 1.5292404174804688,
      "mean_token_accuracy": 0.6646040573716163,
      "num_tokens": 11670628.0,
      "step": 5910
    },
    {
      "entropy": 1.4678036846220492,
      "epoch": 0.1695716311244146,
      "grad_norm": 10.375,
      "learning_rate": 4.93724357740315e-05,
      "loss": 1.5391175270080566,
      "mean_token_accuracy": 0.6676683530211449,
      "num_tokens": 11690935.0,
      "step": 5920
    },
    {
      "entropy": 1.4951945878565311,
      "epoch": 0.16985806969050313,
      "grad_norm": 9.1875,
      "learning_rate": 4.936989237149936e-05,
      "loss": 1.5292593002319337,
      "mean_token_accuracy": 0.6699582532048225,
      "num_tokens": 11711860.0,
      "step": 5930
    },
    {
      "entropy": 1.5030835304409265,
      "epoch": 0.17014450825659166,
      "grad_norm": 8.8125,
      "learning_rate": 4.936734389119133e-05,
      "loss": 1.6066200256347656,
      "mean_token_accuracy": 0.6637712214142084,
      "num_tokens": 11731687.0,
      "step": 5940
    },
    {
      "entropy": 1.5041240505874156,
      "epoch": 0.1704309468226802,
      "grad_norm": 12.1875,
      "learning_rate": 4.936479033363842e-05,
      "loss": 1.521602725982666,
      "mean_token_accuracy": 0.6634639570489526,
      "num_tokens": 11750091.0,
      "step": 5950
    },
    {
      "entropy": 1.5377890944480896,
      "epoch": 0.17071738538876874,
      "grad_norm": 8.4375,
      "learning_rate": 4.9362231699372687e-05,
      "loss": 1.5145090103149415,
      "mean_token_accuracy": 0.6646853685379028,
      "num_tokens": 11770172.0,
      "step": 5960
    },
    {
      "entropy": 1.4608060277998447,
      "epoch": 0.17100382395485728,
      "grad_norm": 10.125,
      "learning_rate": 4.935966798892727e-05,
      "loss": 1.556221103668213,
      "mean_token_accuracy": 0.6546224039047956,
      "num_tokens": 11788765.0,
      "step": 5970
    },
    {
      "entropy": 1.4546675749123097,
      "epoch": 0.1712902625209458,
      "grad_norm": 7.40625,
      "learning_rate": 4.935709920283633e-05,
      "loss": 1.5009776115417481,
      "mean_token_accuracy": 0.6849946957081556,
      "num_tokens": 11808432.0,
      "step": 5980
    },
    {
      "entropy": 1.4300529669970274,
      "epoch": 0.17157670108703435,
      "grad_norm": 9.25,
      "learning_rate": 4.9354525341635125e-05,
      "loss": 1.4588123321533204,
      "mean_token_accuracy": 0.6799989931285382,
      "num_tokens": 11826682.0,
      "step": 5990
    },
    {
      "entropy": 1.4362173832952976,
      "epoch": 0.1718631396531229,
      "grad_norm": 10.8125,
      "learning_rate": 4.935194640585995e-05,
      "loss": 1.4461870193481445,
      "mean_token_accuracy": 0.6857102800160646,
      "num_tokens": 11846458.0,
      "step": 6000
    },
    {
      "entropy": 1.3662438299506903,
      "epoch": 0.17214957821921142,
      "grad_norm": 8.6875,
      "learning_rate": 4.934936239604815e-05,
      "loss": 1.4425847053527832,
      "mean_token_accuracy": 0.679301156103611,
      "num_tokens": 11866440.0,
      "step": 6010
    },
    {
      "entropy": 1.4570660717785358,
      "epoch": 0.17243601678529996,
      "grad_norm": 8.625,
      "learning_rate": 4.9346773312738136e-05,
      "loss": 1.6374954223632812,
      "mean_token_accuracy": 0.6570733428001404,
      "num_tokens": 11885770.0,
      "step": 6020
    },
    {
      "entropy": 1.4791229892522098,
      "epoch": 0.1727224553513885,
      "grad_norm": 9.6875,
      "learning_rate": 4.934417915646939e-05,
      "loss": 1.3939291954040527,
      "mean_token_accuracy": 0.6806836210191249,
      "num_tokens": 11907690.0,
      "step": 6030
    },
    {
      "entropy": 1.478437875583768,
      "epoch": 0.17300889391747704,
      "grad_norm": 10.875,
      "learning_rate": 4.934157992778242e-05,
      "loss": 1.5354734420776368,
      "mean_token_accuracy": 0.6639037605375051,
      "num_tokens": 11927047.0,
      "step": 6040
    },
    {
      "entropy": 1.3138850256800652,
      "epoch": 0.1732953324835656,
      "grad_norm": 9.875,
      "learning_rate": 4.933897562721882e-05,
      "loss": 1.3663606643676758,
      "mean_token_accuracy": 0.6979944676160812,
      "num_tokens": 11947227.0,
      "step": 6050
    },
    {
      "entropy": 1.3479141156189143,
      "epoch": 0.17358177104965414,
      "grad_norm": 6.96875,
      "learning_rate": 4.933636625532122e-05,
      "loss": 1.4165874481201173,
      "mean_token_accuracy": 0.6908961161971092,
      "num_tokens": 11966928.0,
      "step": 6060
    },
    {
      "entropy": 1.3815735965967177,
      "epoch": 0.17386820961574267,
      "grad_norm": 9.625,
      "learning_rate": 4.9333751812633324e-05,
      "loss": 1.4103723526000977,
      "mean_token_accuracy": 0.6743574034422636,
      "num_tokens": 11985932.0,
      "step": 6070
    },
    {
      "entropy": 1.4296928144991399,
      "epoch": 0.1741546481818312,
      "grad_norm": 10.3125,
      "learning_rate": 4.933113229969988e-05,
      "loss": 1.4588019371032714,
      "mean_token_accuracy": 0.6884045783430338,
      "num_tokens": 12005519.0,
      "step": 6080
    },
    {
      "entropy": 1.405590706691146,
      "epoch": 0.17444108674791975,
      "grad_norm": 7.5,
      "learning_rate": 4.93285077170667e-05,
      "loss": 1.496517276763916,
      "mean_token_accuracy": 0.6813424073159695,
      "num_tokens": 12025617.0,
      "step": 6090
    },
    {
      "entropy": 1.3568504810333253,
      "epoch": 0.1747275253140083,
      "grad_norm": 7.75,
      "learning_rate": 4.932587806528064e-05,
      "loss": 1.3935050010681151,
      "mean_token_accuracy": 0.6781116556376219,
      "num_tokens": 12045534.0,
      "step": 6100
    },
    {
      "entropy": 1.4660755727440118,
      "epoch": 0.17501396388009682,
      "grad_norm": 8.0,
      "learning_rate": 4.932324334488964e-05,
      "loss": 1.5551608085632325,
      "mean_token_accuracy": 0.6606322910636664,
      "num_tokens": 12064832.0,
      "step": 6110
    },
    {
      "entropy": 1.3736946443095803,
      "epoch": 0.17530040244618536,
      "grad_norm": 5.8125,
      "learning_rate": 4.932060355644266e-05,
      "loss": 1.3521848678588868,
      "mean_token_accuracy": 0.705763791501522,
      "num_tokens": 12085203.0,
      "step": 6120
    },
    {
      "entropy": 1.4741830304265022,
      "epoch": 0.1755868410122739,
      "grad_norm": 6.90625,
      "learning_rate": 4.931795870048973e-05,
      "loss": 1.5091776847839355,
      "mean_token_accuracy": 0.6733026508241892,
      "num_tokens": 12103873.0,
      "step": 6130
    },
    {
      "entropy": 1.3984277080744505,
      "epoch": 0.17587327957836243,
      "grad_norm": 7.8125,
      "learning_rate": 4.931530877758196e-05,
      "loss": 1.522230339050293,
      "mean_token_accuracy": 0.6791554145514965,
      "num_tokens": 12124552.0,
      "step": 6140
    },
    {
      "entropy": 1.460199267603457,
      "epoch": 0.17615971814445097,
      "grad_norm": 8.875,
      "learning_rate": 4.931265378827147e-05,
      "loss": 1.515638828277588,
      "mean_token_accuracy": 0.6675545807927847,
      "num_tokens": 12143673.0,
      "step": 6150
    },
    {
      "entropy": 1.4435635447502135,
      "epoch": 0.1764461567105395,
      "grad_norm": 10.75,
      "learning_rate": 4.930999373311149e-05,
      "loss": 1.468454360961914,
      "mean_token_accuracy": 0.67950121127069,
      "num_tokens": 12162407.0,
      "step": 6160
    },
    {
      "entropy": 1.3270676020532846,
      "epoch": 0.17673259527662805,
      "grad_norm": 7.8125,
      "learning_rate": 4.9307328612656245e-05,
      "loss": 1.2954941749572755,
      "mean_token_accuracy": 0.70622633472085,
      "num_tokens": 12181587.0,
      "step": 6170
    },
    {
      "entropy": 1.37058818154037,
      "epoch": 0.17701903384271658,
      "grad_norm": 6.90625,
      "learning_rate": 4.930465842746107e-05,
      "loss": 1.4363130569458007,
      "mean_token_accuracy": 0.6902893602848053,
      "num_tokens": 12201885.0,
      "step": 6180
    },
    {
      "entropy": 1.3929596975445748,
      "epoch": 0.17730547240880512,
      "grad_norm": 8.1875,
      "learning_rate": 4.9301983178082325e-05,
      "loss": 1.4469818115234374,
      "mean_token_accuracy": 0.6855168022215367,
      "num_tokens": 12223135.0,
      "step": 6190
    },
    {
      "entropy": 1.364348041638732,
      "epoch": 0.17759191097489366,
      "grad_norm": 10.9375,
      "learning_rate": 4.929930286507742e-05,
      "loss": 1.437267017364502,
      "mean_token_accuracy": 0.6770233936607838,
      "num_tokens": 12243979.0,
      "step": 6200
    },
    {
      "entropy": 1.31310056373477,
      "epoch": 0.1778783495409822,
      "grad_norm": 8.9375,
      "learning_rate": 4.929661748900485e-05,
      "loss": 1.4153963088989259,
      "mean_token_accuracy": 0.6900998506695032,
      "num_tokens": 12263315.0,
      "step": 6210
    },
    {
      "entropy": 1.4167314041405916,
      "epoch": 0.17816478810707073,
      "grad_norm": 9.0625,
      "learning_rate": 4.929392705042414e-05,
      "loss": 1.394523811340332,
      "mean_token_accuracy": 0.684620825573802,
      "num_tokens": 12285429.0,
      "step": 6220
    },
    {
      "entropy": 1.4154333420097829,
      "epoch": 0.17845122667315927,
      "grad_norm": 8.8125,
      "learning_rate": 4.9291231549895875e-05,
      "loss": 1.4673789978027343,
      "mean_token_accuracy": 0.6823446007445455,
      "num_tokens": 12304339.0,
      "step": 6230
    },
    {
      "entropy": 1.4152748204767704,
      "epoch": 0.1787376652392478,
      "grad_norm": 9.1875,
      "learning_rate": 4.9288530987981706e-05,
      "loss": 1.479453182220459,
      "mean_token_accuracy": 0.6812618706375361,
      "num_tokens": 12324831.0,
      "step": 6240
    },
    {
      "entropy": 1.5209394287317992,
      "epoch": 0.17902410380533634,
      "grad_norm": 8.4375,
      "learning_rate": 4.928582536524432e-05,
      "loss": 1.6179141998291016,
      "mean_token_accuracy": 0.6539077125489712,
      "num_tokens": 12343718.0,
      "step": 6250
    },
    {
      "entropy": 1.4698548503220081,
      "epoch": 0.17931054237142488,
      "grad_norm": 8.75,
      "learning_rate": 4.928311468224746e-05,
      "loss": 1.53178071975708,
      "mean_token_accuracy": 0.6720730647444725,
      "num_tokens": 12363617.0,
      "step": 6260
    },
    {
      "entropy": 1.397044263407588,
      "epoch": 0.17959698093751342,
      "grad_norm": 8.5625,
      "learning_rate": 4.9280398939555946e-05,
      "loss": 1.3592684745788575,
      "mean_token_accuracy": 0.7023213289678096,
      "num_tokens": 12383165.0,
      "step": 6270
    },
    {
      "entropy": 1.4452381137758494,
      "epoch": 0.17988341950360195,
      "grad_norm": 9.5625,
      "learning_rate": 4.927767813773563e-05,
      "loss": 1.466906452178955,
      "mean_token_accuracy": 0.6895480569452047,
      "num_tokens": 12401246.0,
      "step": 6280
    },
    {
      "entropy": 1.5307133223861455,
      "epoch": 0.1801698580696905,
      "grad_norm": 8.5625,
      "learning_rate": 4.927495227735344e-05,
      "loss": 1.5598388671875,
      "mean_token_accuracy": 0.6604364186525344,
      "num_tokens": 12420404.0,
      "step": 6290
    },
    {
      "entropy": 1.348336521908641,
      "epoch": 0.18045629663577903,
      "grad_norm": 8.0,
      "learning_rate": 4.9272221358977316e-05,
      "loss": 1.4303204536437988,
      "mean_token_accuracy": 0.6852028977125884,
      "num_tokens": 12440555.0,
      "step": 6300
    },
    {
      "entropy": 1.5626192063093185,
      "epoch": 0.18074273520186757,
      "grad_norm": 6.65625,
      "learning_rate": 4.92694853831763e-05,
      "loss": 1.6378677368164063,
      "mean_token_accuracy": 0.6500140890479088,
      "num_tokens": 12460358.0,
      "step": 6310
    },
    {
      "entropy": 1.4973537605255842,
      "epoch": 0.18102917376795613,
      "grad_norm": 8.5625,
      "learning_rate": 4.926674435052047e-05,
      "loss": 1.4961321830749512,
      "mean_token_accuracy": 0.6632638920098544,
      "num_tokens": 12479034.0,
      "step": 6320
    },
    {
      "entropy": 1.4329667635262013,
      "epoch": 0.18131561233404467,
      "grad_norm": 9.125,
      "learning_rate": 4.9263998261580934e-05,
      "loss": 1.476184368133545,
      "mean_token_accuracy": 0.6747637540102005,
      "num_tokens": 12498030.0,
      "step": 6330
    },
    {
      "entropy": 1.3940817721188068,
      "epoch": 0.1816020509001332,
      "grad_norm": 10.3125,
      "learning_rate": 4.926124711692989e-05,
      "loss": 1.4757598876953124,
      "mean_token_accuracy": 0.6866230014711618,
      "num_tokens": 12519029.0,
      "step": 6340
    },
    {
      "entropy": 1.5069569904357194,
      "epoch": 0.18188848946622174,
      "grad_norm": 9.3125,
      "learning_rate": 4.925849091714058e-05,
      "loss": 1.5763851165771485,
      "mean_token_accuracy": 0.6652629725635052,
      "num_tokens": 12539227.0,
      "step": 6350
    },
    {
      "entropy": 1.5136225722730159,
      "epoch": 0.18217492803231028,
      "grad_norm": 7.15625,
      "learning_rate": 4.925572966278727e-05,
      "loss": 1.4926812171936035,
      "mean_token_accuracy": 0.6646001763641834,
      "num_tokens": 12558725.0,
      "step": 6360
    },
    {
      "entropy": 1.4106845064088702,
      "epoch": 0.18246136659839882,
      "grad_norm": 7.21875,
      "learning_rate": 4.9252963354445324e-05,
      "loss": 1.405810260772705,
      "mean_token_accuracy": 0.6982049446552991,
      "num_tokens": 12578659.0,
      "step": 6370
    },
    {
      "entropy": 1.4251727391034366,
      "epoch": 0.18274780516448735,
      "grad_norm": 8.8125,
      "learning_rate": 4.925019199269113e-05,
      "loss": 1.4651846885681152,
      "mean_token_accuracy": 0.6776089437305928,
      "num_tokens": 12598023.0,
      "step": 6380
    },
    {
      "entropy": 1.292552138119936,
      "epoch": 0.1830342437305759,
      "grad_norm": 8.5,
      "learning_rate": 4.924741557810214e-05,
      "loss": 1.4057302474975586,
      "mean_token_accuracy": 0.6797358829528093,
      "num_tokens": 12618866.0,
      "step": 6390
    },
    {
      "entropy": 1.5226835690438747,
      "epoch": 0.18332068229666443,
      "grad_norm": 10.1875,
      "learning_rate": 4.924463411125685e-05,
      "loss": 1.57452449798584,
      "mean_token_accuracy": 0.6664018632844091,
      "num_tokens": 12638459.0,
      "step": 6400
    },
    {
      "entropy": 1.393143067136407,
      "epoch": 0.18360712086275296,
      "grad_norm": 8.8125,
      "learning_rate": 4.924184759273481e-05,
      "loss": 1.3292108535766602,
      "mean_token_accuracy": 0.7019755341112613,
      "num_tokens": 12658292.0,
      "step": 6410
    },
    {
      "entropy": 1.367004668340087,
      "epoch": 0.1838935594288415,
      "grad_norm": 7.53125,
      "learning_rate": 4.9239056023116634e-05,
      "loss": 1.4254454612731933,
      "mean_token_accuracy": 0.6795507505536079,
      "num_tokens": 12678574.0,
      "step": 6420
    },
    {
      "entropy": 1.2903104435652495,
      "epoch": 0.18417999799493004,
      "grad_norm": 8.25,
      "learning_rate": 4.923625940298398e-05,
      "loss": 1.36850004196167,
      "mean_token_accuracy": 0.704938055574894,
      "num_tokens": 12697885.0,
      "step": 6430
    },
    {
      "entropy": 1.3884007036685944,
      "epoch": 0.18446643656101858,
      "grad_norm": 8.5,
      "learning_rate": 4.9233457732919555e-05,
      "loss": 1.4910995483398437,
      "mean_token_accuracy": 0.680659681558609,
      "num_tokens": 12717722.0,
      "step": 6440
    },
    {
      "entropy": 1.490414746105671,
      "epoch": 0.1847528751271071,
      "grad_norm": 6.65625,
      "learning_rate": 4.923065101350713e-05,
      "loss": 1.522986888885498,
      "mean_token_accuracy": 0.6765483394265175,
      "num_tokens": 12736489.0,
      "step": 6450
    },
    {
      "entropy": 1.4720280453562737,
      "epoch": 0.18503931369319565,
      "grad_norm": 7.71875,
      "learning_rate": 4.9227839245331516e-05,
      "loss": 1.5006288528442382,
      "mean_token_accuracy": 0.6681873235851526,
      "num_tokens": 12755222.0,
      "step": 6460
    },
    {
      "entropy": 1.4506388876587153,
      "epoch": 0.1853257522592842,
      "grad_norm": 7.9375,
      "learning_rate": 4.922502242897857e-05,
      "loss": 1.413188362121582,
      "mean_token_accuracy": 0.6782549951225519,
      "num_tokens": 12775091.0,
      "step": 6470
    },
    {
      "entropy": 1.366637199744582,
      "epoch": 0.18561219082537272,
      "grad_norm": 8.875,
      "learning_rate": 4.922220056503524e-05,
      "loss": 1.4039117813110351,
      "mean_token_accuracy": 0.6915828939527273,
      "num_tokens": 12795057.0,
      "step": 6480
    },
    {
      "entropy": 1.352109369635582,
      "epoch": 0.18589862939146126,
      "grad_norm": 9.8125,
      "learning_rate": 4.9219373654089465e-05,
      "loss": 1.4554457664489746,
      "mean_token_accuracy": 0.6753610597923398,
      "num_tokens": 12814623.0,
      "step": 6490
    },
    {
      "entropy": 1.3736751217395067,
      "epoch": 0.1861850679575498,
      "grad_norm": 11.0625,
      "learning_rate": 4.921654169673029e-05,
      "loss": 1.454780387878418,
      "mean_token_accuracy": 0.6877236545085907,
      "num_tokens": 12832878.0,
      "step": 6500
    },
    {
      "entropy": 1.4073275100439786,
      "epoch": 0.18647150652363834,
      "grad_norm": 7.125,
      "learning_rate": 4.9213704693547776e-05,
      "loss": 1.3822728157043458,
      "mean_token_accuracy": 0.6995428606867791,
      "num_tokens": 12852910.0,
      "step": 6510
    },
    {
      "entropy": 1.4386454924941063,
      "epoch": 0.18675794508972687,
      "grad_norm": 8.0,
      "learning_rate": 4.921086264513305e-05,
      "loss": 1.5043425559997559,
      "mean_token_accuracy": 0.6692972086369992,
      "num_tokens": 12872515.0,
      "step": 6520
    },
    {
      "entropy": 1.285991306602955,
      "epoch": 0.1870443836558154,
      "grad_norm": 9.375,
      "learning_rate": 4.92080155520783e-05,
      "loss": 1.3467216491699219,
      "mean_token_accuracy": 0.7070346478372812,
      "num_tokens": 12891505.0,
      "step": 6530
    },
    {
      "entropy": 1.432631329074502,
      "epoch": 0.18733082222190395,
      "grad_norm": 9.25,
      "learning_rate": 4.920516341497674e-05,
      "loss": 1.479216957092285,
      "mean_token_accuracy": 0.6747622333467007,
      "num_tokens": 12912536.0,
      "step": 6540
    },
    {
      "entropy": 1.4515864118933677,
      "epoch": 0.18761726078799248,
      "grad_norm": 8.875,
      "learning_rate": 4.9202306234422654e-05,
      "loss": 1.5022727966308593,
      "mean_token_accuracy": 0.6815267663449049,
      "num_tokens": 12933252.0,
      "step": 6550
    },
    {
      "entropy": 1.5399513229727746,
      "epoch": 0.18790369935408102,
      "grad_norm": 7.125,
      "learning_rate": 4.919944401101138e-05,
      "loss": 1.5815183639526367,
      "mean_token_accuracy": 0.6569502755999566,
      "num_tokens": 12952633.0,
      "step": 6560
    },
    {
      "entropy": 1.4467506915330888,
      "epoch": 0.18819013792016956,
      "grad_norm": 8.0625,
      "learning_rate": 4.919657674533929e-05,
      "loss": 1.4737659454345704,
      "mean_token_accuracy": 0.6849691689014434,
      "num_tokens": 12972760.0,
      "step": 6570
    },
    {
      "entropy": 1.3586736030876636,
      "epoch": 0.1884765764862581,
      "grad_norm": 9.9375,
      "learning_rate": 4.919370443800382e-05,
      "loss": 1.406723403930664,
      "mean_token_accuracy": 0.6837186042219401,
      "num_tokens": 12993164.0,
      "step": 6580
    },
    {
      "entropy": 1.3812097579240799,
      "epoch": 0.18876301505234666,
      "grad_norm": 8.25,
      "learning_rate": 4.919082708960344e-05,
      "loss": 1.4273009300231934,
      "mean_token_accuracy": 0.6931082312017679,
      "num_tokens": 13013159.0,
      "step": 6590
    },
    {
      "entropy": 1.2081229861825704,
      "epoch": 0.1890494536184352,
      "grad_norm": 7.65625,
      "learning_rate": 4.918794470073769e-05,
      "loss": 1.1818126678466796,
      "mean_token_accuracy": 0.7259028639644385,
      "num_tokens": 13032257.0,
      "step": 6600
    },
    {
      "entropy": 1.318201196938753,
      "epoch": 0.18933589218452374,
      "grad_norm": 8.3125,
      "learning_rate": 4.9185057272007155e-05,
      "loss": 1.3800811767578125,
      "mean_token_accuracy": 0.6960328120738268,
      "num_tokens": 13052868.0,
      "step": 6610
    },
    {
      "entropy": 1.3484413038939238,
      "epoch": 0.18962233075061227,
      "grad_norm": 6.96875,
      "learning_rate": 4.9182164804013465e-05,
      "loss": 1.353172779083252,
      "mean_token_accuracy": 0.702847919613123,
      "num_tokens": 13074459.0,
      "step": 6620
    },
    {
      "entropy": 1.3870065543800592,
      "epoch": 0.1899087693167008,
      "grad_norm": 8.4375,
      "learning_rate": 4.917926729735931e-05,
      "loss": 1.4827387809753418,
      "mean_token_accuracy": 0.6740652158856392,
      "num_tokens": 13095271.0,
      "step": 6630
    },
    {
      "entropy": 1.4729616351425647,
      "epoch": 0.19019520788278935,
      "grad_norm": 10.125,
      "learning_rate": 4.91763647526484e-05,
      "loss": 1.5072721481323241,
      "mean_token_accuracy": 0.6807315915822982,
      "num_tokens": 13114929.0,
      "step": 6640
    },
    {
      "entropy": 1.4896713271737099,
      "epoch": 0.19048164644887788,
      "grad_norm": 8.25,
      "learning_rate": 4.9173457170485535e-05,
      "loss": 1.5076394081115723,
      "mean_token_accuracy": 0.6796420034021139,
      "num_tokens": 13136852.0,
      "step": 6650
    },
    {
      "entropy": 1.3522751234471797,
      "epoch": 0.19076808501496642,
      "grad_norm": 6.96875,
      "learning_rate": 4.917054455147655e-05,
      "loss": 1.4371529579162599,
      "mean_token_accuracy": 0.6883786235004663,
      "num_tokens": 13157567.0,
      "step": 6660
    },
    {
      "entropy": 1.5284927532076835,
      "epoch": 0.19105452358105496,
      "grad_norm": 9.75,
      "learning_rate": 4.9167626896228314e-05,
      "loss": 1.5774428367614746,
      "mean_token_accuracy": 0.6616813138127327,
      "num_tokens": 13177655.0,
      "step": 6670
    },
    {
      "entropy": 1.4974663585424424,
      "epoch": 0.1913409621471435,
      "grad_norm": 10.0625,
      "learning_rate": 4.9164704205348764e-05,
      "loss": 1.496504783630371,
      "mean_token_accuracy": 0.675908538326621,
      "num_tokens": 13198202.0,
      "step": 6680
    },
    {
      "entropy": 1.5301301073282958,
      "epoch": 0.19162740071323203,
      "grad_norm": 9.4375,
      "learning_rate": 4.9161776479446874e-05,
      "loss": 1.5955947875976562,
      "mean_token_accuracy": 0.6482728697359562,
      "num_tokens": 13218510.0,
      "step": 6690
    },
    {
      "entropy": 1.295935932546854,
      "epoch": 0.19191383927932057,
      "grad_norm": 8.9375,
      "learning_rate": 4.9158843719132685e-05,
      "loss": 1.2973398208618163,
      "mean_token_accuracy": 0.7062223497778177,
      "num_tokens": 13237165.0,
      "step": 6700
    },
    {
      "entropy": 1.4205248415470124,
      "epoch": 0.1922002778454091,
      "grad_norm": 8.25,
      "learning_rate": 4.915590592501727e-05,
      "loss": 1.4779239654541017,
      "mean_token_accuracy": 0.6763545289635658,
      "num_tokens": 13258616.0,
      "step": 6710
    },
    {
      "entropy": 1.3705434907227754,
      "epoch": 0.19248671641149764,
      "grad_norm": 8.125,
      "learning_rate": 4.915296309771273e-05,
      "loss": 1.3864919662475585,
      "mean_token_accuracy": 0.6900668147951364,
      "num_tokens": 13278025.0,
      "step": 6720
    },
    {
      "entropy": 1.3562137771397829,
      "epoch": 0.19277315497758618,
      "grad_norm": 8.75,
      "learning_rate": 4.915001523783228e-05,
      "loss": 1.4141832351684571,
      "mean_token_accuracy": 0.6804671116173268,
      "num_tokens": 13298283.0,
      "step": 6730
    },
    {
      "entropy": 1.4325556550174952,
      "epoch": 0.19305959354367472,
      "grad_norm": 7.0,
      "learning_rate": 4.914706234599012e-05,
      "loss": 1.4590951919555664,
      "mean_token_accuracy": 0.6808883346617222,
      "num_tokens": 13319915.0,
      "step": 6740
    },
    {
      "entropy": 1.3633572649210692,
      "epoch": 0.19334603210976326,
      "grad_norm": 8.375,
      "learning_rate": 4.914410442280154e-05,
      "loss": 1.3709599494934082,
      "mean_token_accuracy": 0.6933745093643665,
      "num_tokens": 13341115.0,
      "step": 6750
    },
    {
      "entropy": 1.436106476932764,
      "epoch": 0.1936324706758518,
      "grad_norm": 10.125,
      "learning_rate": 4.914114146888285e-05,
      "loss": 1.5386438369750977,
      "mean_token_accuracy": 0.6873976532369852,
      "num_tokens": 13361569.0,
      "step": 6760
    },
    {
      "entropy": 1.424841780588031,
      "epoch": 0.19391890924194033,
      "grad_norm": 7.96875,
      "learning_rate": 4.91381734848514e-05,
      "loss": 1.4642927169799804,
      "mean_token_accuracy": 0.68417604342103,
      "num_tokens": 13380885.0,
      "step": 6770
    },
    {
      "entropy": 1.3430028449743987,
      "epoch": 0.19420534780802887,
      "grad_norm": 9.0,
      "learning_rate": 4.913520047132564e-05,
      "loss": 1.372169589996338,
      "mean_token_accuracy": 0.6975071728229523,
      "num_tokens": 13399980.0,
      "step": 6780
    },
    {
      "entropy": 1.371938407421112,
      "epoch": 0.1944917863741174,
      "grad_norm": 9.0625,
      "learning_rate": 4.913222242892502e-05,
      "loss": 1.4246555328369142,
      "mean_token_accuracy": 0.6813033115118742,
      "num_tokens": 13421436.0,
      "step": 6790
    },
    {
      "entropy": 1.340294734016061,
      "epoch": 0.19477822494020594,
      "grad_norm": 7.71875,
      "learning_rate": 4.912923935827005e-05,
      "loss": 1.4265383720397948,
      "mean_token_accuracy": 0.6882929772138595,
      "num_tokens": 13443417.0,
      "step": 6800
    },
    {
      "entropy": 1.2730397436767817,
      "epoch": 0.19506466350629448,
      "grad_norm": 8.625,
      "learning_rate": 4.91262512599823e-05,
      "loss": 1.3082396507263183,
      "mean_token_accuracy": 0.7184704177081584,
      "num_tokens": 13462387.0,
      "step": 6810
    },
    {
      "entropy": 1.3208717335015536,
      "epoch": 0.19535110207238301,
      "grad_norm": 9.125,
      "learning_rate": 4.912325813468437e-05,
      "loss": 1.3207721710205078,
      "mean_token_accuracy": 0.7100768335163593,
      "num_tokens": 13482042.0,
      "step": 6820
    },
    {
      "entropy": 1.4038551222532987,
      "epoch": 0.19563754063847155,
      "grad_norm": 9.0,
      "learning_rate": 4.9120259982999915e-05,
      "loss": 1.5005281448364258,
      "mean_token_accuracy": 0.6722738459706307,
      "num_tokens": 13499911.0,
      "step": 6830
    },
    {
      "entropy": 1.4464069247245788,
      "epoch": 0.1959239792045601,
      "grad_norm": 10.6875,
      "learning_rate": 4.911725680555365e-05,
      "loss": 1.4625755310058595,
      "mean_token_accuracy": 0.6852853037416935,
      "num_tokens": 13518378.0,
      "step": 6840
    },
    {
      "entropy": 1.4794189643114806,
      "epoch": 0.19621041777064863,
      "grad_norm": 7.65625,
      "learning_rate": 4.91142486029713e-05,
      "loss": 1.6115404129028321,
      "mean_token_accuracy": 0.6563941441476345,
      "num_tokens": 13536499.0,
      "step": 6850
    },
    {
      "entropy": 1.4786843556910754,
      "epoch": 0.1964968563367372,
      "grad_norm": 8.125,
      "learning_rate": 4.9111235375879694e-05,
      "loss": 1.4864763259887694,
      "mean_token_accuracy": 0.6904301289469004,
      "num_tokens": 13556247.0,
      "step": 6860
    },
    {
      "entropy": 1.3732381880283355,
      "epoch": 0.19678329490282573,
      "grad_norm": 6.75,
      "learning_rate": 4.9108217124906655e-05,
      "loss": 1.3509425163269042,
      "mean_token_accuracy": 0.7018975906074048,
      "num_tokens": 13576284.0,
      "step": 6870
    },
    {
      "entropy": 1.242267757281661,
      "epoch": 0.19706973346891427,
      "grad_norm": 6.21875,
      "learning_rate": 4.910519385068108e-05,
      "loss": 1.2808285713195802,
      "mean_token_accuracy": 0.7185536555945873,
      "num_tokens": 13597002.0,
      "step": 6880
    },
    {
      "entropy": 1.2604170765727758,
      "epoch": 0.1973561720350028,
      "grad_norm": 7.375,
      "learning_rate": 4.91021655538329e-05,
      "loss": 1.3630292892456055,
      "mean_token_accuracy": 0.6981576796621084,
      "num_tokens": 13616383.0,
      "step": 6890
    },
    {
      "entropy": 1.3864607878029347,
      "epoch": 0.19764261060109134,
      "grad_norm": 8.25,
      "learning_rate": 4.909913223499311e-05,
      "loss": 1.363734245300293,
      "mean_token_accuracy": 0.6959175381809473,
      "num_tokens": 13635954.0,
      "step": 6900
    },
    {
      "entropy": 1.4197487790137528,
      "epoch": 0.19792904916717988,
      "grad_norm": 6.6875,
      "learning_rate": 4.909609389479373e-05,
      "loss": 1.5080803871154784,
      "mean_token_accuracy": 0.6763486627489328,
      "num_tokens": 13656450.0,
      "step": 6910
    },
    {
      "entropy": 1.3377501212060452,
      "epoch": 0.19821548773326841,
      "grad_norm": 8.5,
      "learning_rate": 4.909305053386785e-05,
      "loss": 1.4160242080688477,
      "mean_token_accuracy": 0.7016677893698215,
      "num_tokens": 13675328.0,
      "step": 6920
    },
    {
      "entropy": 1.4381409658119082,
      "epoch": 0.19850192629935695,
      "grad_norm": 8.8125,
      "learning_rate": 4.9090002152849573e-05,
      "loss": 1.3976792335510253,
      "mean_token_accuracy": 0.6895610600709915,
      "num_tokens": 13694442.0,
      "step": 6930
    },
    {
      "entropy": 1.456747367233038,
      "epoch": 0.1987883648654455,
      "grad_norm": 8.875,
      "learning_rate": 4.9086948752374085e-05,
      "loss": 1.504695510864258,
      "mean_token_accuracy": 0.6786894287914038,
      "num_tokens": 13713982.0,
      "step": 6940
    },
    {
      "entropy": 1.2406657040119171,
      "epoch": 0.19907480343153403,
      "grad_norm": 9.5,
      "learning_rate": 4.90838903330776e-05,
      "loss": 1.2401039123535156,
      "mean_token_accuracy": 0.7146953620016575,
      "num_tokens": 13732757.0,
      "step": 6950
    },
    {
      "entropy": 1.3115747354924678,
      "epoch": 0.19936124199762256,
      "grad_norm": 8.875,
      "learning_rate": 4.908082689559736e-05,
      "loss": 1.3519312858581543,
      "mean_token_accuracy": 0.7041184697300196,
      "num_tokens": 13752722.0,
      "step": 6960
    },
    {
      "entropy": 1.3855868946760892,
      "epoch": 0.1996476805637111,
      "grad_norm": 8.75,
      "learning_rate": 4.9077758440571704e-05,
      "loss": 1.5695975303649903,
      "mean_token_accuracy": 0.6662397995591164,
      "num_tokens": 13771213.0,
      "step": 6970
    },
    {
      "entropy": 1.3965993613004684,
      "epoch": 0.19993411912979964,
      "grad_norm": 7.625,
      "learning_rate": 4.907468496863995e-05,
      "loss": 1.3573946952819824,
      "mean_token_accuracy": 0.7015636142343282,
      "num_tokens": 13791964.0,
      "step": 6980
    },
    {
      "entropy": 1.3800709258764983,
      "epoch": 0.20022055769588817,
      "grad_norm": 8.0625,
      "learning_rate": 4.907160648044252e-05,
      "loss": 1.3914112091064452,
      "mean_token_accuracy": 0.7016227278858423,
      "num_tokens": 13812377.0,
      "step": 6990
    },
    {
      "entropy": 1.426254230365157,
      "epoch": 0.2005069962619767,
      "grad_norm": 8.0625,
      "learning_rate": 4.9068522976620836e-05,
      "loss": 1.4344714164733887,
      "mean_token_accuracy": 0.6788370661437512,
      "num_tokens": 13831345.0,
      "step": 7000
    },
    {
      "entropy": 1.3388535548001528,
      "epoch": 0.20079343482806525,
      "grad_norm": 8.75,
      "learning_rate": 4.90654344578174e-05,
      "loss": 1.4119460105895996,
      "mean_token_accuracy": 0.6951319318264723,
      "num_tokens": 13852178.0,
      "step": 7010
    },
    {
      "entropy": 1.445525447279215,
      "epoch": 0.20107987339415379,
      "grad_norm": 6.875,
      "learning_rate": 4.906234092467573e-05,
      "loss": 1.475074577331543,
      "mean_token_accuracy": 0.6768249820917844,
      "num_tokens": 13873038.0,
      "step": 7020
    },
    {
      "entropy": 1.3823361240327359,
      "epoch": 0.20136631196024232,
      "grad_norm": 8.125,
      "learning_rate": 4.9059242377840406e-05,
      "loss": 1.409123706817627,
      "mean_token_accuracy": 0.7008399046957493,
      "num_tokens": 13892299.0,
      "step": 7030
    },
    {
      "entropy": 1.4108676839619876,
      "epoch": 0.20165275052633086,
      "grad_norm": 6.65625,
      "learning_rate": 4.905613881795707e-05,
      "loss": 1.4193554878234864,
      "mean_token_accuracy": 0.6782151397317648,
      "num_tokens": 13913212.0,
      "step": 7040
    },
    {
      "entropy": 1.4194565132260322,
      "epoch": 0.2019391890924194,
      "grad_norm": 12.9375,
      "learning_rate": 4.905303024567236e-05,
      "loss": 1.433009719848633,
      "mean_token_accuracy": 0.6871834032237529,
      "num_tokens": 13932901.0,
      "step": 7050
    },
    {
      "entropy": 1.4155337657779454,
      "epoch": 0.20222562765850793,
      "grad_norm": 9.875,
      "learning_rate": 4.904991666163401e-05,
      "loss": 1.5671911239624023,
      "mean_token_accuracy": 0.6667166270315648,
      "num_tokens": 13951193.0,
      "step": 7060
    },
    {
      "entropy": 1.3376609589904547,
      "epoch": 0.20251206622459647,
      "grad_norm": 9.1875,
      "learning_rate": 4.904679806649075e-05,
      "loss": 1.4647872924804688,
      "mean_token_accuracy": 0.6834654208272696,
      "num_tokens": 13970898.0,
      "step": 7070
    },
    {
      "entropy": 1.346419683471322,
      "epoch": 0.202798504790685,
      "grad_norm": 8.8125,
      "learning_rate": 4.904367446089241e-05,
      "loss": 1.314305591583252,
      "mean_token_accuracy": 0.7175991624593735,
      "num_tokens": 13989894.0,
      "step": 7080
    },
    {
      "entropy": 1.5014225922524929,
      "epoch": 0.20308494335677355,
      "grad_norm": 8.75,
      "learning_rate": 4.90405458454898e-05,
      "loss": 1.5338292121887207,
      "mean_token_accuracy": 0.6804172363132238,
      "num_tokens": 14009862.0,
      "step": 7090
    },
    {
      "entropy": 1.3830394230782985,
      "epoch": 0.20337138192286208,
      "grad_norm": 7.78125,
      "learning_rate": 4.903741222093483e-05,
      "loss": 1.4008390426635742,
      "mean_token_accuracy": 0.6850058656185866,
      "num_tokens": 14030262.0,
      "step": 7100
    },
    {
      "entropy": 1.3387300118803978,
      "epoch": 0.20365782048895062,
      "grad_norm": 5.9375,
      "learning_rate": 4.903427358788042e-05,
      "loss": 1.335759735107422,
      "mean_token_accuracy": 0.6982653558254241,
      "num_tokens": 14049399.0,
      "step": 7110
    },
    {
      "entropy": 1.3435183446854353,
      "epoch": 0.20394425905503918,
      "grad_norm": 8.0,
      "learning_rate": 4.9031129946980536e-05,
      "loss": 1.4409222602844238,
      "mean_token_accuracy": 0.68685202896595,
      "num_tokens": 14068805.0,
      "step": 7120
    },
    {
      "entropy": 1.3806550037115812,
      "epoch": 0.20423069762112772,
      "grad_norm": 7.4375,
      "learning_rate": 4.902798129889021e-05,
      "loss": 1.4160164833068847,
      "mean_token_accuracy": 0.6921298008412122,
      "num_tokens": 14089415.0,
      "step": 7130
    },
    {
      "entropy": 1.3908397756516933,
      "epoch": 0.20451713618721626,
      "grad_norm": 7.75,
      "learning_rate": 4.902482764426551e-05,
      "loss": 1.4265143394470214,
      "mean_token_accuracy": 0.6955434899777174,
      "num_tokens": 14107972.0,
      "step": 7140
    },
    {
      "entropy": 1.4211399715393782,
      "epoch": 0.2048035747533048,
      "grad_norm": 7.875,
      "learning_rate": 4.902166898376352e-05,
      "loss": 1.4096290588378906,
      "mean_token_accuracy": 0.6945870824158191,
      "num_tokens": 14127833.0,
      "step": 7150
    },
    {
      "entropy": 1.4306800190359354,
      "epoch": 0.20509001331939333,
      "grad_norm": 10.3125,
      "learning_rate": 4.90185053180424e-05,
      "loss": 1.5266257286071778,
      "mean_token_accuracy": 0.6688980113714933,
      "num_tokens": 14146431.0,
      "step": 7160
    },
    {
      "entropy": 1.3444069474935532,
      "epoch": 0.20537645188548187,
      "grad_norm": 9.8125,
      "learning_rate": 4.901533664776132e-05,
      "loss": 1.3018610000610351,
      "mean_token_accuracy": 0.7133057370781899,
      "num_tokens": 14166679.0,
      "step": 7170
    },
    {
      "entropy": 1.2997824288904667,
      "epoch": 0.2056628904515704,
      "grad_norm": 8.6875,
      "learning_rate": 4.9012162973580545e-05,
      "loss": 1.3112760543823243,
      "mean_token_accuracy": 0.7090633921325207,
      "num_tokens": 14186681.0,
      "step": 7180
    },
    {
      "entropy": 1.2961572241038084,
      "epoch": 0.20594932901765894,
      "grad_norm": 8.8125,
      "learning_rate": 4.9008984296161324e-05,
      "loss": 1.4168990135192872,
      "mean_token_accuracy": 0.6773347610607743,
      "num_tokens": 14206505.0,
      "step": 7190
    },
    {
      "entropy": 1.3758924208581447,
      "epoch": 0.20623576758374748,
      "grad_norm": 7.9375,
      "learning_rate": 4.9005800616165984e-05,
      "loss": 1.412449264526367,
      "mean_token_accuracy": 0.6859348244965077,
      "num_tokens": 14226510.0,
      "step": 7200
    },
    {
      "entropy": 1.4393426783382892,
      "epoch": 0.20652220614983602,
      "grad_norm": 6.71875,
      "learning_rate": 4.900261193425789e-05,
      "loss": 1.488466739654541,
      "mean_token_accuracy": 0.6797539934515953,
      "num_tokens": 14247291.0,
      "step": 7210
    },
    {
      "entropy": 1.3632361229509116,
      "epoch": 0.20680864471592456,
      "grad_norm": 9.1875,
      "learning_rate": 4.899941825110143e-05,
      "loss": 1.3862874031066894,
      "mean_token_accuracy": 0.6870345346629619,
      "num_tokens": 14267696.0,
      "step": 7220
    },
    {
      "entropy": 1.4363153230398893,
      "epoch": 0.2070950832820131,
      "grad_norm": 8.5,
      "learning_rate": 4.8996219567362064e-05,
      "loss": 1.5279389381408692,
      "mean_token_accuracy": 0.6698644058778882,
      "num_tokens": 14287939.0,
      "step": 7230
    },
    {
      "entropy": 1.447751097753644,
      "epoch": 0.20738152184810163,
      "grad_norm": 7.5,
      "learning_rate": 4.8993015883706266e-05,
      "loss": 1.430112075805664,
      "mean_token_accuracy": 0.688009075075388,
      "num_tokens": 14308839.0,
      "step": 7240
    },
    {
      "entropy": 1.394721020013094,
      "epoch": 0.20766796041419017,
      "grad_norm": 9.1875,
      "learning_rate": 4.8989807200801577e-05,
      "loss": 1.4593260765075684,
      "mean_token_accuracy": 0.6800164137035608,
      "num_tokens": 14328880.0,
      "step": 7250
    },
    {
      "entropy": 1.3601030830293894,
      "epoch": 0.2079543989802787,
      "grad_norm": 7.15625,
      "learning_rate": 4.8986593519316554e-05,
      "loss": 1.3623176574707032,
      "mean_token_accuracy": 0.6869924131780862,
      "num_tokens": 14349528.0,
      "step": 7260
    },
    {
      "entropy": 1.2904255852103232,
      "epoch": 0.20824083754636724,
      "grad_norm": 8.8125,
      "learning_rate": 4.8983374839920816e-05,
      "loss": 1.375534725189209,
      "mean_token_accuracy": 0.7001637656241655,
      "num_tokens": 14369653.0,
      "step": 7270
    },
    {
      "entropy": 1.3482805021107196,
      "epoch": 0.20852727611245578,
      "grad_norm": 10.4375,
      "learning_rate": 4.898015116328501e-05,
      "loss": 1.377899169921875,
      "mean_token_accuracy": 0.6938121780753136,
      "num_tokens": 14388892.0,
      "step": 7280
    },
    {
      "entropy": 1.2915750324726105,
      "epoch": 0.20881371467854432,
      "grad_norm": 6.125,
      "learning_rate": 4.897692249008083e-05,
      "loss": 1.3763263702392579,
      "mean_token_accuracy": 0.6970436487346887,
      "num_tokens": 14408096.0,
      "step": 7290
    },
    {
      "entropy": 1.3968203097581864,
      "epoch": 0.20910015324463285,
      "grad_norm": 11.3125,
      "learning_rate": 4.897368882098102e-05,
      "loss": 1.4194849014282227,
      "mean_token_accuracy": 0.6799007184803486,
      "num_tokens": 14428527.0,
      "step": 7300
    },
    {
      "entropy": 1.390107601508498,
      "epoch": 0.2093865918107214,
      "grad_norm": 8.375,
      "learning_rate": 4.897045015665935e-05,
      "loss": 1.4392899513244628,
      "mean_token_accuracy": 0.6902316275984048,
      "num_tokens": 14447593.0,
      "step": 7310
    },
    {
      "entropy": 1.3729560181498528,
      "epoch": 0.20967303037680993,
      "grad_norm": 9.3125,
      "learning_rate": 4.896720649779064e-05,
      "loss": 1.4080945014953614,
      "mean_token_accuracy": 0.700541740655899,
      "num_tokens": 14467102.0,
      "step": 7320
    },
    {
      "entropy": 1.3027869775891303,
      "epoch": 0.20995946894289846,
      "grad_norm": 6.90625,
      "learning_rate": 4.896395784505074e-05,
      "loss": 1.2823558807373048,
      "mean_token_accuracy": 0.7202371619641781,
      "num_tokens": 14486703.0,
      "step": 7330
    },
    {
      "entropy": 1.3749649554491044,
      "epoch": 0.210245907508987,
      "grad_norm": 8.875,
      "learning_rate": 4.8960704199116546e-05,
      "loss": 1.4123984336853028,
      "mean_token_accuracy": 0.6816035013645887,
      "num_tokens": 14505717.0,
      "step": 7340
    },
    {
      "entropy": 1.356557783856988,
      "epoch": 0.21053234607507554,
      "grad_norm": 11.625,
      "learning_rate": 4.895744556066602e-05,
      "loss": 1.3466686248779296,
      "mean_token_accuracy": 0.7100013606250286,
      "num_tokens": 14524426.0,
      "step": 7350
    },
    {
      "entropy": 1.2298651825636626,
      "epoch": 0.21081878464116408,
      "grad_norm": 8.3125,
      "learning_rate": 4.8954181930378104e-05,
      "loss": 1.3651920318603517,
      "mean_token_accuracy": 0.697707237303257,
      "num_tokens": 14543200.0,
      "step": 7360
    },
    {
      "entropy": 1.3208369448781014,
      "epoch": 0.2111052232072526,
      "grad_norm": 9.5625,
      "learning_rate": 4.8950913308932846e-05,
      "loss": 1.4476146697998047,
      "mean_token_accuracy": 0.6957259006798268,
      "num_tokens": 14562438.0,
      "step": 7370
    },
    {
      "entropy": 1.4516440346837043,
      "epoch": 0.21139166177334115,
      "grad_norm": 9.0,
      "learning_rate": 4.894763969701129e-05,
      "loss": 1.4808592796325684,
      "mean_token_accuracy": 0.6746780332177877,
      "num_tokens": 14581554.0,
      "step": 7380
    },
    {
      "entropy": 1.388659530133009,
      "epoch": 0.21167810033942971,
      "grad_norm": 8.9375,
      "learning_rate": 4.894436109529554e-05,
      "loss": 1.3562264442443848,
      "mean_token_accuracy": 0.7026739425957202,
      "num_tokens": 14601704.0,
      "step": 7390
    },
    {
      "entropy": 1.3755133386701346,
      "epoch": 0.21196453890551825,
      "grad_norm": 7.625,
      "learning_rate": 4.8941077504468736e-05,
      "loss": 1.4169318199157714,
      "mean_token_accuracy": 0.6904120303690433,
      "num_tokens": 14621916.0,
      "step": 7400
    },
    {
      "entropy": 1.293072261661291,
      "epoch": 0.2122509774716068,
      "grad_norm": 9.9375,
      "learning_rate": 4.893778892521505e-05,
      "loss": 1.3679720878601074,
      "mean_token_accuracy": 0.699483815766871,
      "num_tokens": 14641785.0,
      "step": 7410
    },
    {
      "entropy": 1.3480365861207246,
      "epoch": 0.21253741603769533,
      "grad_norm": 8.25,
      "learning_rate": 4.89344953582197e-05,
      "loss": 1.3521265983581543,
      "mean_token_accuracy": 0.689704316854477,
      "num_tokens": 14660100.0,
      "step": 7420
    },
    {
      "entropy": 1.2821091564372182,
      "epoch": 0.21282385460378386,
      "grad_norm": 7.375,
      "learning_rate": 4.893119680416895e-05,
      "loss": 1.2913886070251466,
      "mean_token_accuracy": 0.7034950695931912,
      "num_tokens": 14679788.0,
      "step": 7430
    },
    {
      "entropy": 1.384116856008768,
      "epoch": 0.2131102931698724,
      "grad_norm": 8.5,
      "learning_rate": 4.892789326375009e-05,
      "loss": 1.4870211601257324,
      "mean_token_accuracy": 0.6735535811632871,
      "num_tokens": 14700230.0,
      "step": 7440
    },
    {
      "entropy": 1.4824466433376073,
      "epoch": 0.21339673173596094,
      "grad_norm": 8.4375,
      "learning_rate": 4.8924584737651455e-05,
      "loss": 1.446213436126709,
      "mean_token_accuracy": 0.6871388126164675,
      "num_tokens": 14719768.0,
      "step": 7450
    },
    {
      "entropy": 1.4576281111687421,
      "epoch": 0.21368317030204947,
      "grad_norm": 8.875,
      "learning_rate": 4.892127122656242e-05,
      "loss": 1.4400272369384766,
      "mean_token_accuracy": 0.6881834391504527,
      "num_tokens": 14740810.0,
      "step": 7460
    },
    {
      "entropy": 1.2752697099000216,
      "epoch": 0.213969608868138,
      "grad_norm": 8.375,
      "learning_rate": 4.8917952731173386e-05,
      "loss": 1.3433378219604493,
      "mean_token_accuracy": 0.7034235220402479,
      "num_tokens": 14760696.0,
      "step": 7470
    },
    {
      "entropy": 1.3483004350215197,
      "epoch": 0.21425604743422655,
      "grad_norm": 10.4375,
      "learning_rate": 4.891462925217582e-05,
      "loss": 1.3901077270507813,
      "mean_token_accuracy": 0.682411277294159,
      "num_tokens": 14780311.0,
      "step": 7480
    },
    {
      "entropy": 1.2790289748460055,
      "epoch": 0.21454248600031509,
      "grad_norm": 7.875,
      "learning_rate": 4.89113007902622e-05,
      "loss": 1.3366446495056152,
      "mean_token_accuracy": 0.6983015596866607,
      "num_tokens": 14799917.0,
      "step": 7490
    },
    {
      "entropy": 1.2884593315422534,
      "epoch": 0.21482892456640362,
      "grad_norm": 13.6875,
      "learning_rate": 4.890796734612606e-05,
      "loss": 1.275720977783203,
      "mean_token_accuracy": 0.714143181592226,
      "num_tokens": 14819632.0,
      "step": 7500
    },
    {
      "entropy": 1.3287858117371798,
      "epoch": 0.21511536313249216,
      "grad_norm": 8.4375,
      "learning_rate": 4.890462892046196e-05,
      "loss": 1.3351828575134277,
      "mean_token_accuracy": 0.6934810530394315,
      "num_tokens": 14838534.0,
      "step": 7510
    },
    {
      "entropy": 1.2789661645889283,
      "epoch": 0.2154018016985807,
      "grad_norm": 7.78125,
      "learning_rate": 4.89012855139655e-05,
      "loss": 1.3055315017700195,
      "mean_token_accuracy": 0.7132577586919069,
      "num_tokens": 14857170.0,
      "step": 7520
    },
    {
      "entropy": 1.293819534033537,
      "epoch": 0.21568824026466923,
      "grad_norm": 8.3125,
      "learning_rate": 4.889793712733334e-05,
      "loss": 1.3343281745910645,
      "mean_token_accuracy": 0.7063073042780161,
      "num_tokens": 14877160.0,
      "step": 7530
    },
    {
      "entropy": 1.3489057872444392,
      "epoch": 0.21597467883075777,
      "grad_norm": 10.5625,
      "learning_rate": 4.889458376126313e-05,
      "loss": 1.4079439163208007,
      "mean_token_accuracy": 0.6894423123449087,
      "num_tokens": 14897003.0,
      "step": 7540
    },
    {
      "entropy": 1.4150805279612542,
      "epoch": 0.2162611173968463,
      "grad_norm": 7.375,
      "learning_rate": 4.8891225416453614e-05,
      "loss": 1.4901124954223632,
      "mean_token_accuracy": 0.6762777179479599,
      "num_tokens": 14916059.0,
      "step": 7550
    },
    {
      "entropy": 1.5745364610105752,
      "epoch": 0.21654755596293485,
      "grad_norm": 7.0625,
      "learning_rate": 4.8887862093604526e-05,
      "loss": 1.5903640747070313,
      "mean_token_accuracy": 0.6673523098230362,
      "num_tokens": 14936136.0,
      "step": 7560
    },
    {
      "entropy": 1.2758759081363678,
      "epoch": 0.21683399452902338,
      "grad_norm": 7.90625,
      "learning_rate": 4.8884493793416666e-05,
      "loss": 1.2784101486206054,
      "mean_token_accuracy": 0.7163826014846564,
      "num_tokens": 14954604.0,
      "step": 7570
    },
    {
      "entropy": 1.2785285800695418,
      "epoch": 0.21712043309511192,
      "grad_norm": 10.8125,
      "learning_rate": 4.888112051659186e-05,
      "loss": 1.28880615234375,
      "mean_token_accuracy": 0.6964160978794098,
      "num_tokens": 14974011.0,
      "step": 7580
    },
    {
      "entropy": 1.3938613448292017,
      "epoch": 0.21740687166120046,
      "grad_norm": 7.4375,
      "learning_rate": 4.887774226383297e-05,
      "loss": 1.436103916168213,
      "mean_token_accuracy": 0.6869353208690882,
      "num_tokens": 14994637.0,
      "step": 7590
    },
    {
      "entropy": 1.3140036601573228,
      "epoch": 0.217693310227289,
      "grad_norm": 9.8125,
      "learning_rate": 4.887435903584391e-05,
      "loss": 1.374654483795166,
      "mean_token_accuracy": 0.7032893341034651,
      "num_tokens": 15013632.0,
      "step": 7600
    },
    {
      "entropy": 1.2737130880355836,
      "epoch": 0.21797974879337753,
      "grad_norm": 6.25,
      "learning_rate": 4.88709708333296e-05,
      "loss": 1.4103833198547364,
      "mean_token_accuracy": 0.6863091744482517,
      "num_tokens": 15033378.0,
      "step": 7610
    },
    {
      "entropy": 1.4612667422741652,
      "epoch": 0.21826618735946607,
      "grad_norm": 8.125,
      "learning_rate": 4.8867577656996033e-05,
      "loss": 1.492978286743164,
      "mean_token_accuracy": 0.6864451989531517,
      "num_tokens": 15054848.0,
      "step": 7620
    },
    {
      "entropy": 1.286347733065486,
      "epoch": 0.2185526259255546,
      "grad_norm": 7.1875,
      "learning_rate": 4.886417950755021e-05,
      "loss": 1.2393387794494628,
      "mean_token_accuracy": 0.7158916004002094,
      "num_tokens": 15074781.0,
      "step": 7630
    },
    {
      "entropy": 1.2510924391448497,
      "epoch": 0.21883906449164314,
      "grad_norm": 8.5,
      "learning_rate": 4.886077638570017e-05,
      "loss": 1.2491999626159669,
      "mean_token_accuracy": 0.7227519113570452,
      "num_tokens": 15095337.0,
      "step": 7640
    },
    {
      "entropy": 1.1821011263877153,
      "epoch": 0.21912550305773168,
      "grad_norm": 7.5,
      "learning_rate": 4.8857368292155016e-05,
      "loss": 1.2771282196044922,
      "mean_token_accuracy": 0.7172872584313155,
      "num_tokens": 15114995.0,
      "step": 7650
    },
    {
      "entropy": 1.280007117241621,
      "epoch": 0.21941194162382024,
      "grad_norm": 9.0,
      "learning_rate": 4.885395522762486e-05,
      "loss": 1.4100286483764648,
      "mean_token_accuracy": 0.6866801362484694,
      "num_tokens": 15135519.0,
      "step": 7660
    },
    {
      "entropy": 1.3868811391294003,
      "epoch": 0.21969838018990878,
      "grad_norm": 8.6875,
      "learning_rate": 4.885053719282085e-05,
      "loss": 1.4283332824707031,
      "mean_token_accuracy": 0.6930511899292469,
      "num_tokens": 15156922.0,
      "step": 7670
    },
    {
      "entropy": 1.401895508542657,
      "epoch": 0.21998481875599732,
      "grad_norm": 8.375,
      "learning_rate": 4.884711418845518e-05,
      "loss": 1.3743435859680175,
      "mean_token_accuracy": 0.687605693563819,
      "num_tokens": 15178124.0,
      "step": 7680
    },
    {
      "entropy": 1.271172220259905,
      "epoch": 0.22027125732208586,
      "grad_norm": 5.96875,
      "learning_rate": 4.8843686215241083e-05,
      "loss": 1.2306385040283203,
      "mean_token_accuracy": 0.7218283236026763,
      "num_tokens": 15197504.0,
      "step": 7690
    },
    {
      "entropy": 1.2563523545861244,
      "epoch": 0.2205576958881744,
      "grad_norm": 6.65625,
      "learning_rate": 4.8840253273892825e-05,
      "loss": 1.2723988533020019,
      "mean_token_accuracy": 0.708989767357707,
      "num_tokens": 15216692.0,
      "step": 7700
    },
    {
      "entropy": 1.2661789491772653,
      "epoch": 0.22084413445426293,
      "grad_norm": 7.03125,
      "learning_rate": 4.8836815365125675e-05,
      "loss": 1.3291427612304687,
      "mean_token_accuracy": 0.7063781902194023,
      "num_tokens": 15237696.0,
      "step": 7710
    },
    {
      "entropy": 1.2801757603883743,
      "epoch": 0.22113057302035147,
      "grad_norm": 8.375,
      "learning_rate": 4.8833372489656e-05,
      "loss": 1.3109075546264648,
      "mean_token_accuracy": 0.704571132734418,
      "num_tokens": 15258504.0,
      "step": 7720
    },
    {
      "entropy": 1.284193018451333,
      "epoch": 0.22141701158644,
      "grad_norm": 9.8125,
      "learning_rate": 4.882992464820114e-05,
      "loss": 1.3723615646362304,
      "mean_token_accuracy": 0.7046990979462862,
      "num_tokens": 15277709.0,
      "step": 7730
    },
    {
      "entropy": 1.3690773416310549,
      "epoch": 0.22170345015252854,
      "grad_norm": 9.625,
      "learning_rate": 4.882647184147951e-05,
      "loss": 1.4337735176086426,
      "mean_token_accuracy": 0.6841593950986862,
      "num_tokens": 15296201.0,
      "step": 7740
    },
    {
      "entropy": 1.3079580947756768,
      "epoch": 0.22198988871861708,
      "grad_norm": 9.0625,
      "learning_rate": 4.882301407021055e-05,
      "loss": 1.3392826080322267,
      "mean_token_accuracy": 0.700236376747489,
      "num_tokens": 15316008.0,
      "step": 7750
    },
    {
      "entropy": 1.415876792371273,
      "epoch": 0.22227632728470562,
      "grad_norm": 7.90625,
      "learning_rate": 4.881955133511472e-05,
      "loss": 1.4688376426696776,
      "mean_token_accuracy": 0.6904159473255277,
      "num_tokens": 15335827.0,
      "step": 7760
    },
    {
      "entropy": 1.4446927666664124,
      "epoch": 0.22256276585079415,
      "grad_norm": 7.5625,
      "learning_rate": 4.8816083636913525e-05,
      "loss": 1.538266944885254,
      "mean_token_accuracy": 0.6795007001608611,
      "num_tokens": 15354988.0,
      "step": 7770
    },
    {
      "entropy": 1.4225209131836891,
      "epoch": 0.2228492044168827,
      "grad_norm": 5.78125,
      "learning_rate": 4.8812610976329506e-05,
      "loss": 1.3882936477661132,
      "mean_token_accuracy": 0.6982832193374634,
      "num_tokens": 15375536.0,
      "step": 7780
    },
    {
      "entropy": 1.3170425936579704,
      "epoch": 0.22313564298297123,
      "grad_norm": 8.25,
      "learning_rate": 4.8809133354086244e-05,
      "loss": 1.3739373207092285,
      "mean_token_accuracy": 0.6934373334050179,
      "num_tokens": 15394707.0,
      "step": 7790
    },
    {
      "entropy": 1.2687998063862325,
      "epoch": 0.22342208154905976,
      "grad_norm": 8.375,
      "learning_rate": 4.880565077090834e-05,
      "loss": 1.300158977508545,
      "mean_token_accuracy": 0.7184651874005794,
      "num_tokens": 15414114.0,
      "step": 7800
    },
    {
      "entropy": 1.4427742466330529,
      "epoch": 0.2237085201151483,
      "grad_norm": 8.8125,
      "learning_rate": 4.880216322752142e-05,
      "loss": 1.469937801361084,
      "mean_token_accuracy": 0.6808895222842694,
      "num_tokens": 15434131.0,
      "step": 7810
    },
    {
      "entropy": 1.2740275532007217,
      "epoch": 0.22399495868123684,
      "grad_norm": 7.09375,
      "learning_rate": 4.8798670724652186e-05,
      "loss": 1.317356491088867,
      "mean_token_accuracy": 0.7080950126051903,
      "num_tokens": 15454536.0,
      "step": 7820
    },
    {
      "entropy": 1.3545406349003315,
      "epoch": 0.22428139724732538,
      "grad_norm": 8.875,
      "learning_rate": 4.8795173263028324e-05,
      "loss": 1.4382153511047364,
      "mean_token_accuracy": 0.6884837828576564,
      "num_tokens": 15473937.0,
      "step": 7830
    },
    {
      "entropy": 1.4733360931277275,
      "epoch": 0.2245678358134139,
      "grad_norm": 7.3125,
      "learning_rate": 4.8791670843378576e-05,
      "loss": 1.525193691253662,
      "mean_token_accuracy": 0.6759477905929089,
      "num_tokens": 15493145.0,
      "step": 7840
    },
    {
      "entropy": 1.4691454891115427,
      "epoch": 0.22485427437950245,
      "grad_norm": 9.6875,
      "learning_rate": 4.8788163466432726e-05,
      "loss": 1.5178655624389648,
      "mean_token_accuracy": 0.6732699435204268,
      "num_tokens": 15511641.0,
      "step": 7850
    },
    {
      "entropy": 1.4671073205769063,
      "epoch": 0.225140712945591,
      "grad_norm": 10.0625,
      "learning_rate": 4.878465113292158e-05,
      "loss": 1.5719306945800782,
      "mean_token_accuracy": 0.6798214539885521,
      "num_tokens": 15530670.0,
      "step": 7860
    },
    {
      "entropy": 1.3936611466109752,
      "epoch": 0.22542715151167952,
      "grad_norm": 7.71875,
      "learning_rate": 4.878113384357696e-05,
      "loss": 1.3768365859985352,
      "mean_token_accuracy": 0.7045553438365459,
      "num_tokens": 15549928.0,
      "step": 7870
    },
    {
      "entropy": 1.3084229059517383,
      "epoch": 0.22571359007776806,
      "grad_norm": 9.125,
      "learning_rate": 4.8777611599131756e-05,
      "loss": 1.2989063262939453,
      "mean_token_accuracy": 0.7167639885097742,
      "num_tokens": 15569940.0,
      "step": 7880
    },
    {
      "entropy": 1.3241024676710367,
      "epoch": 0.2260000286438566,
      "grad_norm": 10.9375,
      "learning_rate": 4.877408440031987e-05,
      "loss": 1.3705173492431642,
      "mean_token_accuracy": 0.6987843196839094,
      "num_tokens": 15589786.0,
      "step": 7890
    },
    {
      "entropy": 1.3425384059548378,
      "epoch": 0.22628646720994514,
      "grad_norm": 9.75,
      "learning_rate": 4.877055224787622e-05,
      "loss": 1.3403461456298829,
      "mean_token_accuracy": 0.6985780414193868,
      "num_tokens": 15611059.0,
      "step": 7900
    },
    {
      "entropy": 1.2052909169346093,
      "epoch": 0.22657290577603367,
      "grad_norm": 9.3125,
      "learning_rate": 4.87670151425368e-05,
      "loss": 1.2613117218017578,
      "mean_token_accuracy": 0.7153838749974966,
      "num_tokens": 15630064.0,
      "step": 7910
    },
    {
      "entropy": 1.2749242829158902,
      "epoch": 0.2268593443421222,
      "grad_norm": 6.78125,
      "learning_rate": 4.87634730850386e-05,
      "loss": 1.3177407264709473,
      "mean_token_accuracy": 0.7087857540696859,
      "num_tokens": 15648755.0,
      "step": 7920
    },
    {
      "entropy": 1.356166660040617,
      "epoch": 0.22714578290821077,
      "grad_norm": 8.5625,
      "learning_rate": 4.8759926076119644e-05,
      "loss": 1.3944215774536133,
      "mean_token_accuracy": 0.6909115295857191,
      "num_tokens": 15670035.0,
      "step": 7930
    },
    {
      "entropy": 1.349524899572134,
      "epoch": 0.2274322214742993,
      "grad_norm": 11.0625,
      "learning_rate": 4.875637411651901e-05,
      "loss": 1.3656500816345214,
      "mean_token_accuracy": 0.6972495723515749,
      "num_tokens": 15688307.0,
      "step": 7940
    },
    {
      "entropy": 1.315654394030571,
      "epoch": 0.22771866004038785,
      "grad_norm": 6.71875,
      "learning_rate": 4.8752817206976795e-05,
      "loss": 1.238535213470459,
      "mean_token_accuracy": 0.7184632230550051,
      "num_tokens": 15707217.0,
      "step": 7950
    },
    {
      "entropy": 1.2872909754514694,
      "epoch": 0.2280050986064764,
      "grad_norm": 8.25,
      "learning_rate": 4.874925534823411e-05,
      "loss": 1.372249698638916,
      "mean_token_accuracy": 0.6899086125195026,
      "num_tokens": 15727044.0,
      "step": 7960
    },
    {
      "entropy": 1.32220651358366,
      "epoch": 0.22829153717256492,
      "grad_norm": 7.40625,
      "learning_rate": 4.874568854103313e-05,
      "loss": 1.3752135276794433,
      "mean_token_accuracy": 0.6870768733322621,
      "num_tokens": 15747482.0,
      "step": 7970
    },
    {
      "entropy": 1.2956975664943458,
      "epoch": 0.22857797573865346,
      "grad_norm": 8.375,
      "learning_rate": 4.8742116786117034e-05,
      "loss": 1.3387102127075194,
      "mean_token_accuracy": 0.7106714308261871,
      "num_tokens": 15767769.0,
      "step": 7980
    },
    {
      "entropy": 1.3806438844650983,
      "epoch": 0.228864414304742,
      "grad_norm": 7.46875,
      "learning_rate": 4.8738540084230044e-05,
      "loss": 1.4194136619567872,
      "mean_token_accuracy": 0.6867449875921011,
      "num_tokens": 15787725.0,
      "step": 7990
    },
    {
      "epoch": 0.22915085287083053,
      "eval_entropy": 1.3525832796096802,
      "eval_loss": 1.3715550899505615,
      "eval_mean_token_accuracy": 0.6933415282964707,
      "eval_num_tokens": 15807238.0,
      "eval_runtime": 43.1178,
      "eval_samples_per_second": 46.385,
      "eval_steps_per_second": 5.798,
      "step": 8000
    },
    {
      "entropy": 1.3368042374029756,
      "epoch": 0.22943729143691907,
      "grad_norm": 9.0625,
      "learning_rate": 4.873137184252543e-05,
      "loss": 1.3546357154846191,
      "mean_token_accuracy": 0.6971236057579517,
      "num_tokens": 15827525.0,
      "step": 8010
    },
    {
      "entropy": 1.3948000933974982,
      "epoch": 0.2297237300030076,
      "grad_norm": 6.78125,
      "learning_rate": 4.8727780304201397e-05,
      "loss": 1.4288634300231933,
      "mean_token_accuracy": 0.6948393099009991,
      "num_tokens": 15848229.0,
      "step": 8020
    },
    {
      "entropy": 1.4148659810423851,
      "epoch": 0.23001016856909615,
      "grad_norm": 9.75,
      "learning_rate": 4.872418382189366e-05,
      "loss": 1.4647943496704101,
      "mean_token_accuracy": 0.6829231698065996,
      "num_tokens": 15868056.0,
      "step": 8030
    },
    {
      "entropy": 1.4107302393764258,
      "epoch": 0.23029660713518468,
      "grad_norm": 8.75,
      "learning_rate": 4.872058239635159e-05,
      "loss": 1.4398988723754882,
      "mean_token_accuracy": 0.6907641410827636,
      "num_tokens": 15887526.0,
      "step": 8040
    },
    {
      "entropy": 1.3166353072971106,
      "epoch": 0.23058304570127322,
      "grad_norm": 10.75,
      "learning_rate": 4.871697602832559e-05,
      "loss": 1.274904441833496,
      "mean_token_accuracy": 0.7086723688989878,
      "num_tokens": 15908020.0,
      "step": 8050
    },
    {
      "entropy": 1.215018856152892,
      "epoch": 0.23086948426736176,
      "grad_norm": 7.3125,
      "learning_rate": 4.871336471856709e-05,
      "loss": 1.2969029426574707,
      "mean_token_accuracy": 0.7075026195496321,
      "num_tokens": 15927642.0,
      "step": 8060
    },
    {
      "entropy": 1.245218475162983,
      "epoch": 0.2311559228334503,
      "grad_norm": 9.4375,
      "learning_rate": 4.870974846782856e-05,
      "loss": 1.3223005294799806,
      "mean_token_accuracy": 0.6976545203477145,
      "num_tokens": 15947904.0,
      "step": 8070
    },
    {
      "entropy": 1.314133294299245,
      "epoch": 0.23144236139953883,
      "grad_norm": 8.0625,
      "learning_rate": 4.8706127276863485e-05,
      "loss": 1.379852867126465,
      "mean_token_accuracy": 0.703938664495945,
      "num_tokens": 15968290.0,
      "step": 8080
    },
    {
      "entropy": 1.3094776287674903,
      "epoch": 0.23172879996562737,
      "grad_norm": 10.3125,
      "learning_rate": 4.870250114642639e-05,
      "loss": 1.3534197807312012,
      "mean_token_accuracy": 0.7027083735913038,
      "num_tokens": 15987461.0,
      "step": 8090
    },
    {
      "entropy": 1.3287329260259866,
      "epoch": 0.2320152385317159,
      "grad_norm": 8.1875,
      "learning_rate": 4.8698870077272824e-05,
      "loss": 1.3490395545959473,
      "mean_token_accuracy": 0.7040076270699501,
      "num_tokens": 16007054.0,
      "step": 8100
    },
    {
      "entropy": 1.2291704252362252,
      "epoch": 0.23230167709780444,
      "grad_norm": 8.3125,
      "learning_rate": 4.869523407015936e-05,
      "loss": 1.2816612243652343,
      "mean_token_accuracy": 0.7215011950582266,
      "num_tokens": 16026469.0,
      "step": 8110
    },
    {
      "entropy": 1.305432095564902,
      "epoch": 0.23258811566389298,
      "grad_norm": 12.9375,
      "learning_rate": 4.869159312584362e-05,
      "loss": 1.2637593269348144,
      "mean_token_accuracy": 0.7156236406415701,
      "num_tokens": 16046728.0,
      "step": 8120
    },
    {
      "entropy": 1.3024291686713696,
      "epoch": 0.23287455422998152,
      "grad_norm": 8.3125,
      "learning_rate": 4.8687947245084234e-05,
      "loss": 1.3540659904479981,
      "mean_token_accuracy": 0.6981031056493521,
      "num_tokens": 16065946.0,
      "step": 8130
    },
    {
      "entropy": 1.2802363142371178,
      "epoch": 0.23316099279607005,
      "grad_norm": 6.5,
      "learning_rate": 4.868429642864086e-05,
      "loss": 1.2869328498840331,
      "mean_token_accuracy": 0.7198633503168821,
      "num_tokens": 16085449.0,
      "step": 8140
    },
    {
      "entropy": 1.4078667741268873,
      "epoch": 0.2334474313621586,
      "grad_norm": 7.5,
      "learning_rate": 4.8680640677274194e-05,
      "loss": 1.4981518745422364,
      "mean_token_accuracy": 0.6752590913325548,
      "num_tokens": 16106012.0,
      "step": 8150
    },
    {
      "entropy": 1.3809628207236528,
      "epoch": 0.23373386992824713,
      "grad_norm": 12.6875,
      "learning_rate": 4.8676979991745976e-05,
      "loss": 1.4528388977050781,
      "mean_token_accuracy": 0.6850547280162573,
      "num_tokens": 16126173.0,
      "step": 8160
    },
    {
      "entropy": 1.3444415017962457,
      "epoch": 0.23402030849433567,
      "grad_norm": 8.0625,
      "learning_rate": 4.8673314372818935e-05,
      "loss": 1.4166419982910157,
      "mean_token_accuracy": 0.6987987965345382,
      "num_tokens": 16145614.0,
      "step": 8170
    },
    {
      "entropy": 1.3124527130275965,
      "epoch": 0.2343067470604242,
      "grad_norm": 9.0,
      "learning_rate": 4.866964382125685e-05,
      "loss": 1.336414909362793,
      "mean_token_accuracy": 0.7011750034987927,
      "num_tokens": 16164909.0,
      "step": 8180
    },
    {
      "entropy": 1.4348484814167022,
      "epoch": 0.23459318562651274,
      "grad_norm": 9.25,
      "learning_rate": 4.866596833782454e-05,
      "loss": 1.529961109161377,
      "mean_token_accuracy": 0.6687340304255486,
      "num_tokens": 16185489.0,
      "step": 8190
    },
    {
      "entropy": 1.2911674588918687,
      "epoch": 0.2348796241926013,
      "grad_norm": 6.59375,
      "learning_rate": 4.866228792328783e-05,
      "loss": 1.219797706604004,
      "mean_token_accuracy": 0.7309391036629677,
      "num_tokens": 16204328.0,
      "step": 8200
    },
    {
      "entropy": 1.2788859456777573,
      "epoch": 0.23516606275868984,
      "grad_norm": 11.3125,
      "learning_rate": 4.865860257841359e-05,
      "loss": 1.3128036499023437,
      "mean_token_accuracy": 0.7154462095350027,
      "num_tokens": 16224336.0,
      "step": 8210
    },
    {
      "entropy": 1.2730165421962738,
      "epoch": 0.23545250132477838,
      "grad_norm": 8.0625,
      "learning_rate": 4.86549123039697e-05,
      "loss": 1.354213047027588,
      "mean_token_accuracy": 0.7013130847364664,
      "num_tokens": 16244243.0,
      "step": 8220
    },
    {
      "entropy": 1.3116079408675432,
      "epoch": 0.23573893989086692,
      "grad_norm": 8.625,
      "learning_rate": 4.8651217100725064e-05,
      "loss": 1.4273070335388183,
      "mean_token_accuracy": 0.6908482234925032,
      "num_tokens": 16263926.0,
      "step": 8230
    },
    {
      "entropy": 1.373075483366847,
      "epoch": 0.23602537845695545,
      "grad_norm": 8.875,
      "learning_rate": 4.864751696944965e-05,
      "loss": 1.3720784187316895,
      "mean_token_accuracy": 0.694175549224019,
      "num_tokens": 16284817.0,
      "step": 8240
    },
    {
      "entropy": 1.3922888569533824,
      "epoch": 0.236311817023044,
      "grad_norm": 8.4375,
      "learning_rate": 4.864381191091441e-05,
      "loss": 1.3658974647521973,
      "mean_token_accuracy": 0.6961234033107757,
      "num_tokens": 16304164.0,
      "step": 8250
    },
    {
      "entropy": 1.3457724403589963,
      "epoch": 0.23659825558913253,
      "grad_norm": 5.65625,
      "learning_rate": 4.864010192589135e-05,
      "loss": 1.3793910026550293,
      "mean_token_accuracy": 0.6869689714163542,
      "num_tokens": 16325521.0,
      "step": 8260
    },
    {
      "entropy": 1.3429566673934459,
      "epoch": 0.23688469415522106,
      "grad_norm": 8.9375,
      "learning_rate": 4.863638701515348e-05,
      "loss": 1.3903090476989746,
      "mean_token_accuracy": 0.6891145247966051,
      "num_tokens": 16343427.0,
      "step": 8270
    },
    {
      "entropy": 1.5069112539291383,
      "epoch": 0.2371711327213096,
      "grad_norm": 10.25,
      "learning_rate": 4.863266717947485e-05,
      "loss": 1.5753619194030761,
      "mean_token_accuracy": 0.666107964515686,
      "num_tokens": 16363095.0,
      "step": 8280
    },
    {
      "entropy": 1.3348246820271015,
      "epoch": 0.23745757128739814,
      "grad_norm": 9.625,
      "learning_rate": 4.862894241963055e-05,
      "loss": 1.3638680458068848,
      "mean_token_accuracy": 0.7002866119146347,
      "num_tokens": 16383050.0,
      "step": 8290
    },
    {
      "entropy": 1.3352429240942,
      "epoch": 0.23774400985348668,
      "grad_norm": 6.78125,
      "learning_rate": 4.862521273639668e-05,
      "loss": 1.3364219665527344,
      "mean_token_accuracy": 0.6988799657672644,
      "num_tokens": 16402884.0,
      "step": 8300
    },
    {
      "entropy": 1.3143289241939784,
      "epoch": 0.2380304484195752,
      "grad_norm": 10.5625,
      "learning_rate": 4.8621478130550344e-05,
      "loss": 1.3329084396362305,
      "mean_token_accuracy": 0.7065146721899509,
      "num_tokens": 16422536.0,
      "step": 8310
    },
    {
      "entropy": 1.288917900249362,
      "epoch": 0.23831688698566375,
      "grad_norm": 6.53125,
      "learning_rate": 4.8617738602869714e-05,
      "loss": 1.2743046760559082,
      "mean_token_accuracy": 0.7156042341142893,
      "num_tokens": 16443747.0,
      "step": 8320
    },
    {
      "entropy": 1.311733551695943,
      "epoch": 0.2386033255517523,
      "grad_norm": 11.0,
      "learning_rate": 4.861399415413396e-05,
      "loss": 1.405808162689209,
      "mean_token_accuracy": 0.6953676007688046,
      "num_tokens": 16462398.0,
      "step": 8330
    },
    {
      "entropy": 1.3846714984625579,
      "epoch": 0.23888976411784082,
      "grad_norm": 7.6875,
      "learning_rate": 4.861024478512328e-05,
      "loss": 1.4922773361206054,
      "mean_token_accuracy": 0.6691900324076414,
      "num_tokens": 16481527.0,
      "step": 8340
    },
    {
      "entropy": 1.2748234923928976,
      "epoch": 0.23917620268392936,
      "grad_norm": 9.25,
      "learning_rate": 4.860649049661893e-05,
      "loss": 1.1929106712341309,
      "mean_token_accuracy": 0.7362593207508326,
      "num_tokens": 16501675.0,
      "step": 8350
    },
    {
      "entropy": 1.3361931089311838,
      "epoch": 0.2394626412500179,
      "grad_norm": 7.03125,
      "learning_rate": 4.8602731289403124e-05,
      "loss": 1.3322365760803223,
      "mean_token_accuracy": 0.7035299580544233,
      "num_tokens": 16522516.0,
      "step": 8360
    },
    {
      "entropy": 1.2914295453578233,
      "epoch": 0.23974907981610644,
      "grad_norm": 13.25,
      "learning_rate": 4.8598967164259173e-05,
      "loss": 1.3713109016418457,
      "mean_token_accuracy": 0.6943941611796618,
      "num_tokens": 16542628.0,
      "step": 8370
    },
    {
      "entropy": 1.368678654357791,
      "epoch": 0.24003551838219497,
      "grad_norm": 8.6875,
      "learning_rate": 4.859519812197136e-05,
      "loss": 1.4172140121459962,
      "mean_token_accuracy": 0.6869722083210945,
      "num_tokens": 16561729.0,
      "step": 8380
    },
    {
      "entropy": 1.2999014116823673,
      "epoch": 0.2403219569482835,
      "grad_norm": 8.375,
      "learning_rate": 4.859142416332503e-05,
      "loss": 1.354226016998291,
      "mean_token_accuracy": 0.6968124371021986,
      "num_tokens": 16581204.0,
      "step": 8390
    },
    {
      "entropy": 1.3914289385080338,
      "epoch": 0.24060839551437205,
      "grad_norm": 7.40625,
      "learning_rate": 4.858764528910652e-05,
      "loss": 1.3810704231262207,
      "mean_token_accuracy": 0.6944346055388451,
      "num_tokens": 16600026.0,
      "step": 8400
    },
    {
      "entropy": 1.2593204272910952,
      "epoch": 0.24089483408046058,
      "grad_norm": 7.1875,
      "learning_rate": 4.8583861500103215e-05,
      "loss": 1.2960443496704102,
      "mean_token_accuracy": 0.7045311786234378,
      "num_tokens": 16619606.0,
      "step": 8410
    },
    {
      "entropy": 1.3147992327809335,
      "epoch": 0.24118127264654912,
      "grad_norm": 7.3125,
      "learning_rate": 4.85800727971035e-05,
      "loss": 1.3279443740844727,
      "mean_token_accuracy": 0.7149071618914604,
      "num_tokens": 16638912.0,
      "step": 8420
    },
    {
      "entropy": 1.3592380195856095,
      "epoch": 0.24146771121263766,
      "grad_norm": 10.0,
      "learning_rate": 4.857627918089682e-05,
      "loss": 1.4101136207580567,
      "mean_token_accuracy": 0.6875106453895569,
      "num_tokens": 16660287.0,
      "step": 8430
    },
    {
      "entropy": 1.3409567870199681,
      "epoch": 0.2417541497787262,
      "grad_norm": 6.78125,
      "learning_rate": 4.857248065227362e-05,
      "loss": 1.4069478034973144,
      "mean_token_accuracy": 0.6943238092586398,
      "num_tokens": 16679656.0,
      "step": 8440
    },
    {
      "entropy": 1.3505582470446824,
      "epoch": 0.24204058834481473,
      "grad_norm": 6.09375,
      "learning_rate": 4.856867721202536e-05,
      "loss": 1.396233367919922,
      "mean_token_accuracy": 0.6943115558475256,
      "num_tokens": 16700423.0,
      "step": 8450
    },
    {
      "entropy": 1.2833540592342616,
      "epoch": 0.24232702691090327,
      "grad_norm": 9.6875,
      "learning_rate": 4.856486886094455e-05,
      "loss": 1.2806329727172852,
      "mean_token_accuracy": 0.7195678774267436,
      "num_tokens": 16719942.0,
      "step": 8460
    },
    {
      "entropy": 1.27919750995934,
      "epoch": 0.24261346547699184,
      "grad_norm": 11.625,
      "learning_rate": 4.856105559982468e-05,
      "loss": 1.2620330810546876,
      "mean_token_accuracy": 0.7119680348783731,
      "num_tokens": 16739113.0,
      "step": 8470
    },
    {
      "entropy": 1.2528194408863782,
      "epoch": 0.24289990404308037,
      "grad_norm": 8.0,
      "learning_rate": 4.8557237429460325e-05,
      "loss": 1.3730833053588867,
      "mean_token_accuracy": 0.7003060482442379,
      "num_tokens": 16759328.0,
      "step": 8480
    },
    {
      "entropy": 1.3498518448323011,
      "epoch": 0.2431863426091689,
      "grad_norm": 8.25,
      "learning_rate": 4.855341435064703e-05,
      "loss": 1.426292896270752,
      "mean_token_accuracy": 0.6925671607255935,
      "num_tokens": 16780547.0,
      "step": 8490
    },
    {
      "entropy": 1.3656618740409612,
      "epoch": 0.24347278117525745,
      "grad_norm": 8.4375,
      "learning_rate": 4.854958636418139e-05,
      "loss": 1.3180909156799316,
      "mean_token_accuracy": 0.7094634931534529,
      "num_tokens": 16800575.0,
      "step": 8500
    },
    {
      "entropy": 1.2510054435580968,
      "epoch": 0.24375921974134598,
      "grad_norm": 8.5625,
      "learning_rate": 4.854575347086101e-05,
      "loss": 1.3178614616394042,
      "mean_token_accuracy": 0.7092883057892323,
      "num_tokens": 16820322.0,
      "step": 8510
    },
    {
      "entropy": 1.3439761586487293,
      "epoch": 0.24404565830743452,
      "grad_norm": 8.5,
      "learning_rate": 4.854191567148452e-05,
      "loss": 1.4390742301940918,
      "mean_token_accuracy": 0.6914662979543209,
      "num_tokens": 16841302.0,
      "step": 8520
    },
    {
      "entropy": 1.2478022906929254,
      "epoch": 0.24433209687352306,
      "grad_norm": 7.4375,
      "learning_rate": 4.853807296685158e-05,
      "loss": 1.2608952522277832,
      "mean_token_accuracy": 0.7045908909291029,
      "num_tokens": 16860106.0,
      "step": 8530
    },
    {
      "entropy": 1.2601082272827626,
      "epoch": 0.2446185354396116,
      "grad_norm": 8.1875,
      "learning_rate": 4.8534225357762866e-05,
      "loss": 1.2476003646850586,
      "mean_token_accuracy": 0.723358640819788,
      "num_tokens": 16879829.0,
      "step": 8540
    },
    {
      "entropy": 1.3123316571116448,
      "epoch": 0.24490497400570013,
      "grad_norm": 8.5,
      "learning_rate": 4.853037284502007e-05,
      "loss": 1.3461403846740723,
      "mean_token_accuracy": 0.7071382343769074,
      "num_tokens": 16898616.0,
      "step": 8550
    },
    {
      "entropy": 1.1433360937982797,
      "epoch": 0.24519141257178867,
      "grad_norm": 7.875,
      "learning_rate": 4.852651542942591e-05,
      "loss": 1.1783739089965821,
      "mean_token_accuracy": 0.7301077082753181,
      "num_tokens": 16917587.0,
      "step": 8560
    },
    {
      "entropy": 1.2611917808651925,
      "epoch": 0.2454778511378772,
      "grad_norm": 5.53125,
      "learning_rate": 4.852265311178414e-05,
      "loss": 1.2958187103271483,
      "mean_token_accuracy": 0.7096046347171068,
      "num_tokens": 16936842.0,
      "step": 8570
    },
    {
      "entropy": 1.2707879062741996,
      "epoch": 0.24576428970396574,
      "grad_norm": 7.84375,
      "learning_rate": 4.851878589289951e-05,
      "loss": 1.279293918609619,
      "mean_token_accuracy": 0.7184138230979442,
      "num_tokens": 16956866.0,
      "step": 8580
    },
    {
      "entropy": 1.2512096384540201,
      "epoch": 0.24605072827005428,
      "grad_norm": 6.84375,
      "learning_rate": 4.851491377357781e-05,
      "loss": 1.3372897148132323,
      "mean_token_accuracy": 0.702974797040224,
      "num_tokens": 16975397.0,
      "step": 8590
    },
    {
      "entropy": 1.278340177051723,
      "epoch": 0.24633716683614282,
      "grad_norm": 8.4375,
      "learning_rate": 4.851103675462585e-05,
      "loss": 1.2980100631713867,
      "mean_token_accuracy": 0.7158501110970974,
      "num_tokens": 16996166.0,
      "step": 8600
    },
    {
      "entropy": 1.3300394169986247,
      "epoch": 0.24662360540223136,
      "grad_norm": 6.84375,
      "learning_rate": 4.850715483685145e-05,
      "loss": 1.3693733215332031,
      "mean_token_accuracy": 0.7026569206267596,
      "num_tokens": 17015100.0,
      "step": 8610
    },
    {
      "entropy": 1.4044758338481187,
      "epoch": 0.2469100439683199,
      "grad_norm": 7.34375,
      "learning_rate": 4.850326802106345e-05,
      "loss": 1.4302474021911622,
      "mean_token_accuracy": 0.6916542194783688,
      "num_tokens": 17034058.0,
      "step": 8620
    },
    {
      "entropy": 1.3206113751977682,
      "epoch": 0.24719648253440843,
      "grad_norm": 10.1875,
      "learning_rate": 4.849937630807174e-05,
      "loss": 1.39181489944458,
      "mean_token_accuracy": 0.6876004215329885,
      "num_tokens": 17053089.0,
      "step": 8630
    },
    {
      "entropy": 1.1952310550957919,
      "epoch": 0.24748292110049697,
      "grad_norm": 7.25,
      "learning_rate": 4.849547969868718e-05,
      "loss": 1.2122062683105468,
      "mean_token_accuracy": 0.7337113194167614,
      "num_tokens": 17071991.0,
      "step": 8640
    },
    {
      "entropy": 1.312955129146576,
      "epoch": 0.2477693596665855,
      "grad_norm": 8.125,
      "learning_rate": 4.8491578193721695e-05,
      "loss": 1.3263768196105956,
      "mean_token_accuracy": 0.7012049920856953,
      "num_tokens": 17091919.0,
      "step": 8650
    },
    {
      "entropy": 1.25547207146883,
      "epoch": 0.24805579823267404,
      "grad_norm": 7.8125,
      "learning_rate": 4.848767179398821e-05,
      "loss": 1.2492012977600098,
      "mean_token_accuracy": 0.7070884669199586,
      "num_tokens": 17113562.0,
      "step": 8660
    },
    {
      "entropy": 1.1096773331984877,
      "epoch": 0.24834223679876258,
      "grad_norm": 9.875,
      "learning_rate": 4.848376050030067e-05,
      "loss": 1.1990671157836914,
      "mean_token_accuracy": 0.721477260440588,
      "num_tokens": 17133012.0,
      "step": 8670
    },
    {
      "entropy": 1.2394254699349403,
      "epoch": 0.24862867536485111,
      "grad_norm": 7.4375,
      "learning_rate": 4.847984431347403e-05,
      "loss": 1.3469114303588867,
      "mean_token_accuracy": 0.7041354414075613,
      "num_tokens": 17152334.0,
      "step": 8680
    },
    {
      "entropy": 1.3852784413844348,
      "epoch": 0.24891511393093965,
      "grad_norm": 6.71875,
      "learning_rate": 4.847592323432431e-05,
      "loss": 1.3782036781311036,
      "mean_token_accuracy": 0.6852422401309013,
      "num_tokens": 17171465.0,
      "step": 8690
    },
    {
      "entropy": 1.2349199458956719,
      "epoch": 0.2492015524970282,
      "grad_norm": 7.5625,
      "learning_rate": 4.8471997263668486e-05,
      "loss": 1.226328468322754,
      "mean_token_accuracy": 0.7154191877692938,
      "num_tokens": 17190801.0,
      "step": 8700
    },
    {
      "entropy": 1.251618079841137,
      "epoch": 0.24948799106311673,
      "grad_norm": 7.40625,
      "learning_rate": 4.846806640232461e-05,
      "loss": 1.2519200325012207,
      "mean_token_accuracy": 0.7186239931732417,
      "num_tokens": 17210461.0,
      "step": 8710
    },
    {
      "entropy": 1.2829885203391314,
      "epoch": 0.24977442962920526,
      "grad_norm": 11.375,
      "learning_rate": 4.84641306511117e-05,
      "loss": 1.3736440658569335,
      "mean_token_accuracy": 0.6862790618091822,
      "num_tokens": 17232481.0,
      "step": 8720
    },
    {
      "entropy": 1.2550558131188154,
      "epoch": 0.25006086819529383,
      "grad_norm": 10.0625,
      "learning_rate": 4.8460190010849835e-05,
      "loss": 1.275850486755371,
      "mean_token_accuracy": 0.7059022095054388,
      "num_tokens": 17250556.0,
      "step": 8730
    },
    {
      "entropy": 1.300730186700821,
      "epoch": 0.25034730676138234,
      "grad_norm": 8.6875,
      "learning_rate": 4.845624448236009e-05,
      "loss": 1.3349772453308106,
      "mean_token_accuracy": 0.6941989261657,
      "num_tokens": 17269507.0,
      "step": 8740
    },
    {
      "entropy": 1.2250392898917197,
      "epoch": 0.2506337453274709,
      "grad_norm": 8.875,
      "learning_rate": 4.8452294066464575e-05,
      "loss": 1.2364221572875977,
      "mean_token_accuracy": 0.7252295173704624,
      "num_tokens": 17287983.0,
      "step": 8750
    },
    {
      "entropy": 1.3068331524729728,
      "epoch": 0.2509201838935594,
      "grad_norm": 6.875,
      "learning_rate": 4.84483387639864e-05,
      "loss": 1.4161869049072267,
      "mean_token_accuracy": 0.6999049104750157,
      "num_tokens": 17307814.0,
      "step": 8760
    },
    {
      "entropy": 1.3700547605752944,
      "epoch": 0.251206622459648,
      "grad_norm": 9.75,
      "learning_rate": 4.8444378575749714e-05,
      "loss": 1.4014565467834472,
      "mean_token_accuracy": 0.6874885011464358,
      "num_tokens": 17327057.0,
      "step": 8770
    },
    {
      "entropy": 1.264924530312419,
      "epoch": 0.2514930610257365,
      "grad_norm": 8.9375,
      "learning_rate": 4.844041350257966e-05,
      "loss": 1.2689948081970215,
      "mean_token_accuracy": 0.7169880189001561,
      "num_tokens": 17346271.0,
      "step": 8780
    },
    {
      "entropy": 1.3334347050637008,
      "epoch": 0.25177949959182505,
      "grad_norm": 9.4375,
      "learning_rate": 4.843644354530241e-05,
      "loss": 1.3798280715942384,
      "mean_token_accuracy": 0.6919111352413893,
      "num_tokens": 17366520.0,
      "step": 8790
    },
    {
      "entropy": 1.2361187532544136,
      "epoch": 0.25206593815791356,
      "grad_norm": 7.4375,
      "learning_rate": 4.843246870474517e-05,
      "loss": 1.2816729545593262,
      "mean_token_accuracy": 0.7087344095110893,
      "num_tokens": 17386584.0,
      "step": 8800
    },
    {
      "entropy": 1.2250650390982627,
      "epoch": 0.2523523767240021,
      "grad_norm": 11.25,
      "learning_rate": 4.842848898173614e-05,
      "loss": 1.2804014205932617,
      "mean_token_accuracy": 0.7141382545232773,
      "num_tokens": 17406097.0,
      "step": 8810
    },
    {
      "entropy": 1.2992166604846715,
      "epoch": 0.25263881529009063,
      "grad_norm": 10.25,
      "learning_rate": 4.842450437710454e-05,
      "loss": 1.2975056648254395,
      "mean_token_accuracy": 0.7111137416213751,
      "num_tokens": 17426727.0,
      "step": 8820
    },
    {
      "entropy": 1.318169329687953,
      "epoch": 0.2529252538561792,
      "grad_norm": 7.5625,
      "learning_rate": 4.842051489168062e-05,
      "loss": 1.3865495681762696,
      "mean_token_accuracy": 0.6926764242351056,
      "num_tokens": 17446015.0,
      "step": 8830
    },
    {
      "entropy": 1.274455401301384,
      "epoch": 0.2532116924222677,
      "grad_norm": 7.1875,
      "learning_rate": 4.841652052629564e-05,
      "loss": 1.3350720405578613,
      "mean_token_accuracy": 0.7090015534311533,
      "num_tokens": 17465660.0,
      "step": 8840
    },
    {
      "entropy": 1.3543043985962868,
      "epoch": 0.2534981309883563,
      "grad_norm": 7.8125,
      "learning_rate": 4.8412521281781876e-05,
      "loss": 1.3698068618774415,
      "mean_token_accuracy": 0.6999831791967154,
      "num_tokens": 17485122.0,
      "step": 8850
    },
    {
      "entropy": 1.233489517867565,
      "epoch": 0.2537845695544448,
      "grad_norm": 7.125,
      "learning_rate": 4.840851715897262e-05,
      "loss": 1.2474135398864745,
      "mean_token_accuracy": 0.7246371265500784,
      "num_tokens": 17503707.0,
      "step": 8860
    },
    {
      "entropy": 1.196946768090129,
      "epoch": 0.25407100812053335,
      "grad_norm": 9.0625,
      "learning_rate": 4.8404508158702176e-05,
      "loss": 1.2503541946411132,
      "mean_token_accuracy": 0.7292619161307812,
      "num_tokens": 17522983.0,
      "step": 8870
    },
    {
      "entropy": 1.3353007577359677,
      "epoch": 0.2543574466866219,
      "grad_norm": 10.875,
      "learning_rate": 4.840049428180588e-05,
      "loss": 1.3440793037414551,
      "mean_token_accuracy": 0.7034892737865448,
      "num_tokens": 17541596.0,
      "step": 8880
    },
    {
      "entropy": 1.244178694114089,
      "epoch": 0.2546438852527104,
      "grad_norm": 7.78125,
      "learning_rate": 4.8396475529120064e-05,
      "loss": 1.2328694343566895,
      "mean_token_accuracy": 0.717902934923768,
      "num_tokens": 17562108.0,
      "step": 8890
    },
    {
      "entropy": 1.1850521273910999,
      "epoch": 0.254930323818799,
      "grad_norm": 10.5625,
      "learning_rate": 4.83924519014821e-05,
      "loss": 1.2897089958190917,
      "mean_token_accuracy": 0.7144745960831642,
      "num_tokens": 17581987.0,
      "step": 8900
    },
    {
      "entropy": 1.1607084844261408,
      "epoch": 0.2552167623848875,
      "grad_norm": 8.1875,
      "learning_rate": 4.8388423399730354e-05,
      "loss": 1.1564496040344239,
      "mean_token_accuracy": 0.7309151075780391,
      "num_tokens": 17600491.0,
      "step": 8910
    },
    {
      "entropy": 1.2427813615649939,
      "epoch": 0.25550320095097606,
      "grad_norm": 10.1875,
      "learning_rate": 4.838439002470421e-05,
      "loss": 1.316445255279541,
      "mean_token_accuracy": 0.714518991112709,
      "num_tokens": 17620332.0,
      "step": 8920
    },
    {
      "entropy": 1.3517073668539523,
      "epoch": 0.25578963951706457,
      "grad_norm": 7.125,
      "learning_rate": 4.8380351777244074e-05,
      "loss": 1.3991028785705566,
      "mean_token_accuracy": 0.691733792796731,
      "num_tokens": 17640142.0,
      "step": 8930
    },
    {
      "entropy": 1.3683735538274049,
      "epoch": 0.25607607808315314,
      "grad_norm": 8.625,
      "learning_rate": 4.837630865819137e-05,
      "loss": 1.3935741424560546,
      "mean_token_accuracy": 0.6869533844292164,
      "num_tokens": 17660323.0,
      "step": 8940
    },
    {
      "entropy": 1.2362545177340507,
      "epoch": 0.25636251664924165,
      "grad_norm": 9.6875,
      "learning_rate": 4.8372260668388544e-05,
      "loss": 1.2276459693908692,
      "mean_token_accuracy": 0.7174127500504255,
      "num_tokens": 17679087.0,
      "step": 8950
    },
    {
      "entropy": 1.3572645898908378,
      "epoch": 0.2566489552153302,
      "grad_norm": 6.78125,
      "learning_rate": 4.8368207808679026e-05,
      "loss": 1.4539985656738281,
      "mean_token_accuracy": 0.6940882369875908,
      "num_tokens": 17698229.0,
      "step": 8960
    },
    {
      "entropy": 1.3479806134477257,
      "epoch": 0.2569353937814187,
      "grad_norm": 7.09375,
      "learning_rate": 4.836415007990729e-05,
      "loss": 1.2825389862060548,
      "mean_token_accuracy": 0.7158281572163105,
      "num_tokens": 17717948.0,
      "step": 8970
    },
    {
      "entropy": 1.260199136659503,
      "epoch": 0.2572218323475073,
      "grad_norm": 7.875,
      "learning_rate": 4.8360087482918815e-05,
      "loss": 1.3139952659606933,
      "mean_token_accuracy": 0.7144561346620322,
      "num_tokens": 17737324.0,
      "step": 8980
    },
    {
      "entropy": 1.358280747383833,
      "epoch": 0.2575082709135958,
      "grad_norm": 10.1875,
      "learning_rate": 4.83560200185601e-05,
      "loss": 1.468091106414795,
      "mean_token_accuracy": 0.6907810769975186,
      "num_tokens": 17756566.0,
      "step": 8990
    },
    {
      "entropy": 1.430641485005617,
      "epoch": 0.25779470947968436,
      "grad_norm": 9.5625,
      "learning_rate": 4.835194768767864e-05,
      "loss": 1.5613553047180175,
      "mean_token_accuracy": 0.6674827165901661,
      "num_tokens": 17776048.0,
      "step": 9000
    },
    {
      "entropy": 1.3074854500591755,
      "epoch": 0.25808114804577287,
      "grad_norm": 7.34375,
      "learning_rate": 4.8347870491122974e-05,
      "loss": 1.268750286102295,
      "mean_token_accuracy": 0.7098786775022745,
      "num_tokens": 17795703.0,
      "step": 9010
    },
    {
      "entropy": 1.3482228700071572,
      "epoch": 0.25836758661186143,
      "grad_norm": 8.5,
      "learning_rate": 4.8343788429742616e-05,
      "loss": 1.3972908973693847,
      "mean_token_accuracy": 0.6933056764304638,
      "num_tokens": 17816523.0,
      "step": 9020
    },
    {
      "entropy": 1.2695806331932544,
      "epoch": 0.25865402517794994,
      "grad_norm": 6.90625,
      "learning_rate": 4.833970150438813e-05,
      "loss": 1.2388599395751954,
      "mean_token_accuracy": 0.7200279366225004,
      "num_tokens": 17835831.0,
      "step": 9030
    },
    {
      "entropy": 1.3064831744879484,
      "epoch": 0.2589404637440385,
      "grad_norm": 8.0625,
      "learning_rate": 4.833560971591107e-05,
      "loss": 1.3616947174072265,
      "mean_token_accuracy": 0.6906574167311191,
      "num_tokens": 17854300.0,
      "step": 9040
    },
    {
      "entropy": 1.242756339535117,
      "epoch": 0.259226902310127,
      "grad_norm": 7.5625,
      "learning_rate": 4.8331513065164025e-05,
      "loss": 1.2443609237670898,
      "mean_token_accuracy": 0.721508489176631,
      "num_tokens": 17873915.0,
      "step": 9050
    },
    {
      "entropy": 1.2797224348410965,
      "epoch": 0.2595133408762156,
      "grad_norm": 8.5,
      "learning_rate": 4.832741155300058e-05,
      "loss": 1.3430242538452148,
      "mean_token_accuracy": 0.7018827233463526,
      "num_tokens": 17893109.0,
      "step": 9060
    },
    {
      "entropy": 1.2939703285694122,
      "epoch": 0.2597997794423041,
      "grad_norm": 9.375,
      "learning_rate": 4.832330518027532e-05,
      "loss": 1.3360145568847657,
      "mean_token_accuracy": 0.7011235821992159,
      "num_tokens": 17912712.0,
      "step": 9070
    },
    {
      "entropy": 1.2067293141037225,
      "epoch": 0.26008621800839266,
      "grad_norm": 20.875,
      "learning_rate": 4.8319193947843885e-05,
      "loss": 1.255198860168457,
      "mean_token_accuracy": 0.7131258200854063,
      "num_tokens": 17932165.0,
      "step": 9080
    },
    {
      "entropy": 1.2123133920133113,
      "epoch": 0.26037265657448116,
      "grad_norm": 8.0,
      "learning_rate": 4.8315077856562886e-05,
      "loss": 1.2235565185546875,
      "mean_token_accuracy": 0.7290280099958182,
      "num_tokens": 17952088.0,
      "step": 9090
    },
    {
      "entropy": 1.2446948152035475,
      "epoch": 0.26065909514056973,
      "grad_norm": 10.9375,
      "learning_rate": 4.8310956907289975e-05,
      "loss": 1.2665054321289062,
      "mean_token_accuracy": 0.7128050982952118,
      "num_tokens": 17972537.0,
      "step": 9100
    },
    {
      "entropy": 1.3901444882154466,
      "epoch": 0.26094553370665824,
      "grad_norm": 7.46875,
      "learning_rate": 4.830683110088379e-05,
      "loss": 1.4608612060546875,
      "mean_token_accuracy": 0.678036005795002,
      "num_tokens": 17994472.0,
      "step": 9110
    },
    {
      "entropy": 1.3316185031086207,
      "epoch": 0.2612319722727468,
      "grad_norm": 8.75,
      "learning_rate": 4.8302700438204e-05,
      "loss": 1.3409854888916015,
      "mean_token_accuracy": 0.7039120152592659,
      "num_tokens": 18014318.0,
      "step": 9120
    },
    {
      "entropy": 1.1473009951412678,
      "epoch": 0.2615184108388353,
      "grad_norm": 5.84375,
      "learning_rate": 4.829856492011128e-05,
      "loss": 1.15621976852417,
      "mean_token_accuracy": 0.7365556918084621,
      "num_tokens": 18033977.0,
      "step": 9130
    },
    {
      "entropy": 1.3351036570966244,
      "epoch": 0.2618048494049239,
      "grad_norm": 7.46875,
      "learning_rate": 4.829442454746732e-05,
      "loss": 1.3299053192138672,
      "mean_token_accuracy": 0.7117034964263439,
      "num_tokens": 18053596.0,
      "step": 9140
    },
    {
      "entropy": 1.2761389184743166,
      "epoch": 0.26209128797101244,
      "grad_norm": 6.53125,
      "learning_rate": 4.8290279321134825e-05,
      "loss": 1.3119136810302734,
      "mean_token_accuracy": 0.7001638438552618,
      "num_tokens": 18074125.0,
      "step": 9150
    },
    {
      "entropy": 1.3045391861349345,
      "epoch": 0.26237772653710095,
      "grad_norm": 7.1875,
      "learning_rate": 4.82861292419775e-05,
      "loss": 1.341162872314453,
      "mean_token_accuracy": 0.696571696922183,
      "num_tokens": 18093734.0,
      "step": 9160
    },
    {
      "entropy": 1.225440262258053,
      "epoch": 0.2626641651031895,
      "grad_norm": 8.75,
      "learning_rate": 4.828197431086005e-05,
      "loss": 1.2435415267944336,
      "mean_token_accuracy": 0.7088596798479557,
      "num_tokens": 18113254.0,
      "step": 9170
    },
    {
      "entropy": 1.3132943410426379,
      "epoch": 0.262950603669278,
      "grad_norm": 9.5625,
      "learning_rate": 4.8277814528648243e-05,
      "loss": 1.3694846153259277,
      "mean_token_accuracy": 0.697538585960865,
      "num_tokens": 18132818.0,
      "step": 9180
    },
    {
      "entropy": 1.2257746003568173,
      "epoch": 0.2632370422353666,
      "grad_norm": 8.5625,
      "learning_rate": 4.827364989620879e-05,
      "loss": 1.2318166732788085,
      "mean_token_accuracy": 0.723556337505579,
      "num_tokens": 18153417.0,
      "step": 9190
    },
    {
      "entropy": 1.1300859481096268,
      "epoch": 0.2635234808014551,
      "grad_norm": 8.125,
      "learning_rate": 4.826948041440946e-05,
      "loss": 1.1399371147155761,
      "mean_token_accuracy": 0.7450911700725555,
      "num_tokens": 18172555.0,
      "step": 9200
    },
    {
      "entropy": 1.2252363990992308,
      "epoch": 0.26380991936754367,
      "grad_norm": 9.0,
      "learning_rate": 4.826530608411902e-05,
      "loss": 1.3349379539489745,
      "mean_token_accuracy": 0.7047261077910661,
      "num_tokens": 18190827.0,
      "step": 9210
    },
    {
      "entropy": 1.1883025977760553,
      "epoch": 0.2640963579336322,
      "grad_norm": 7.8125,
      "learning_rate": 4.826112690620724e-05,
      "loss": 1.2126608848571778,
      "mean_token_accuracy": 0.7291012145578861,
      "num_tokens": 18210140.0,
      "step": 9220
    },
    {
      "entropy": 1.2685103014111518,
      "epoch": 0.26438279649972074,
      "grad_norm": 10.8125,
      "learning_rate": 4.825694288154491e-05,
      "loss": 1.2821360588073731,
      "mean_token_accuracy": 0.7151098676025868,
      "num_tokens": 18229663.0,
      "step": 9230
    },
    {
      "entropy": 1.2720585726201534,
      "epoch": 0.26466923506580925,
      "grad_norm": 10.75,
      "learning_rate": 4.8252754011003806e-05,
      "loss": 1.3949750900268554,
      "mean_token_accuracy": 0.6948246248066425,
      "num_tokens": 18249630.0,
      "step": 9240
    },
    {
      "entropy": 1.3435747668147087,
      "epoch": 0.2649556736318978,
      "grad_norm": 10.0,
      "learning_rate": 4.8248560295456755e-05,
      "loss": 1.3088791847229004,
      "mean_token_accuracy": 0.7048242457211018,
      "num_tokens": 18270253.0,
      "step": 9250
    },
    {
      "entropy": 1.307034932821989,
      "epoch": 0.2652421121979863,
      "grad_norm": 6.90625,
      "learning_rate": 4.824436173577756e-05,
      "loss": 1.367745304107666,
      "mean_token_accuracy": 0.6957523223012686,
      "num_tokens": 18290322.0,
      "step": 9260
    },
    {
      "entropy": 1.286383710242808,
      "epoch": 0.2655285507640749,
      "grad_norm": 10.4375,
      "learning_rate": 4.824015833284104e-05,
      "loss": 1.3017475128173828,
      "mean_token_accuracy": 0.7115768007934093,
      "num_tokens": 18310389.0,
      "step": 9270
    },
    {
      "entropy": 1.228379225730896,
      "epoch": 0.2658149893301634,
      "grad_norm": 10.1875,
      "learning_rate": 4.8235950087523034e-05,
      "loss": 1.238663673400879,
      "mean_token_accuracy": 0.7153057418763638,
      "num_tokens": 18329483.0,
      "step": 9280
    },
    {
      "entropy": 1.284581445157528,
      "epoch": 0.26610142789625196,
      "grad_norm": 6.28125,
      "learning_rate": 4.823173700070038e-05,
      "loss": 1.327126407623291,
      "mean_token_accuracy": 0.7110133722424508,
      "num_tokens": 18349027.0,
      "step": 9290
    },
    {
      "entropy": 1.338231517933309,
      "epoch": 0.2663878664623405,
      "grad_norm": 6.34375,
      "learning_rate": 4.822751907325094e-05,
      "loss": 1.3528115272521972,
      "mean_token_accuracy": 0.6897929109632969,
      "num_tokens": 18369125.0,
      "step": 9300
    },
    {
      "entropy": 1.1946829522028566,
      "epoch": 0.26667430502842904,
      "grad_norm": 8.8125,
      "learning_rate": 4.8223296306053555e-05,
      "loss": 1.2565718650817872,
      "mean_token_accuracy": 0.7138553358614445,
      "num_tokens": 18389297.0,
      "step": 9310
    },
    {
      "entropy": 1.2185409259051085,
      "epoch": 0.26696074359451755,
      "grad_norm": 7.78125,
      "learning_rate": 4.82190686999881e-05,
      "loss": 1.2864867210388184,
      "mean_token_accuracy": 0.7176742620766163,
      "num_tokens": 18408737.0,
      "step": 9320
    },
    {
      "entropy": 1.1369651939719916,
      "epoch": 0.2672471821606061,
      "grad_norm": 9.625,
      "learning_rate": 4.8214836255935447e-05,
      "loss": 1.2205231666564942,
      "mean_token_accuracy": 0.7372077208012342,
      "num_tokens": 18428730.0,
      "step": 9330
    },
    {
      "entropy": 1.2750138990581035,
      "epoch": 0.2675336207266946,
      "grad_norm": 7.21875,
      "learning_rate": 4.8210598974777486e-05,
      "loss": 1.2297069549560546,
      "mean_token_accuracy": 0.7142499633133411,
      "num_tokens": 18448023.0,
      "step": 9340
    },
    {
      "entropy": 1.324444257467985,
      "epoch": 0.2678200592927832,
      "grad_norm": 8.3125,
      "learning_rate": 4.82063568573971e-05,
      "loss": 1.433567428588867,
      "mean_token_accuracy": 0.6768547410145402,
      "num_tokens": 18467874.0,
      "step": 9350
    },
    {
      "entropy": 1.2551036842167378,
      "epoch": 0.2681064978588717,
      "grad_norm": 10.625,
      "learning_rate": 4.8202109904678196e-05,
      "loss": 1.227750301361084,
      "mean_token_accuracy": 0.7244132608175278,
      "num_tokens": 18487269.0,
      "step": 9360
    },
    {
      "entropy": 1.225942300632596,
      "epoch": 0.26839293642496026,
      "grad_norm": 10.5,
      "learning_rate": 4.8197858117505675e-05,
      "loss": 1.2767447471618651,
      "mean_token_accuracy": 0.7188018415123224,
      "num_tokens": 18505982.0,
      "step": 9370
    },
    {
      "entropy": 1.2126261051744223,
      "epoch": 0.26867937499104877,
      "grad_norm": 7.625,
      "learning_rate": 4.819360149676545e-05,
      "loss": 1.2597173690795898,
      "mean_token_accuracy": 0.7160643842071295,
      "num_tokens": 18525644.0,
      "step": 9380
    },
    {
      "entropy": 1.225095884874463,
      "epoch": 0.26896581355713733,
      "grad_norm": 9.4375,
      "learning_rate": 4.818934004334445e-05,
      "loss": 1.2749581336975098,
      "mean_token_accuracy": 0.7233197525143623,
      "num_tokens": 18546222.0,
      "step": 9390
    },
    {
      "entropy": 1.3505785837769508,
      "epoch": 0.26925225212322584,
      "grad_norm": 10.0625,
      "learning_rate": 4.818507375813059e-05,
      "loss": 1.3217741966247558,
      "mean_token_accuracy": 0.6942776661366225,
      "num_tokens": 18567731.0,
      "step": 9400
    },
    {
      "entropy": 1.436489612981677,
      "epoch": 0.2695386906893144,
      "grad_norm": 7.96875,
      "learning_rate": 4.818080264201282e-05,
      "loss": 1.451855754852295,
      "mean_token_accuracy": 0.6768117170780897,
      "num_tokens": 18588342.0,
      "step": 9410
    },
    {
      "entropy": 1.245344403758645,
      "epoch": 0.269825129255403,
      "grad_norm": 10.375,
      "learning_rate": 4.817652669588107e-05,
      "loss": 1.3660114288330079,
      "mean_token_accuracy": 0.7104274649173021,
      "num_tokens": 18607445.0,
      "step": 9420
    },
    {
      "entropy": 1.3693070027977228,
      "epoch": 0.2701115678214915,
      "grad_norm": 7.90625,
      "learning_rate": 4.817224592062629e-05,
      "loss": 1.398356056213379,
      "mean_token_accuracy": 0.7086417347192764,
      "num_tokens": 18627526.0,
      "step": 9430
    },
    {
      "entropy": 1.4342213924974203,
      "epoch": 0.27039800638758005,
      "grad_norm": 6.9375,
      "learning_rate": 4.816796031714043e-05,
      "loss": 1.4222000122070313,
      "mean_token_accuracy": 0.6907794957980513,
      "num_tokens": 18647290.0,
      "step": 9440
    },
    {
      "entropy": 1.3352411851286887,
      "epoch": 0.27068444495366856,
      "grad_norm": 11.9375,
      "learning_rate": 4.816366988631646e-05,
      "loss": 1.3825608253479005,
      "mean_token_accuracy": 0.7000203546136617,
      "num_tokens": 18664747.0,
      "step": 9450
    },
    {
      "entropy": 1.3079643428325654,
      "epoch": 0.2709708835197571,
      "grad_norm": 8.5,
      "learning_rate": 4.8159374629048334e-05,
      "loss": 1.2621251106262208,
      "mean_token_accuracy": 0.7136755257844924,
      "num_tokens": 18684534.0,
      "step": 9460
    },
    {
      "entropy": 1.2533658739179372,
      "epoch": 0.27125732208584563,
      "grad_norm": 6.4375,
      "learning_rate": 4.8155074546231034e-05,
      "loss": 1.2466175079345703,
      "mean_token_accuracy": 0.7103608205914498,
      "num_tokens": 18704073.0,
      "step": 9470
    },
    {
      "entropy": 1.2291026808321477,
      "epoch": 0.2715437606519342,
      "grad_norm": 9.0625,
      "learning_rate": 4.815076963876053e-05,
      "loss": 1.2841590881347655,
      "mean_token_accuracy": 0.7114346351474523,
      "num_tokens": 18724562.0,
      "step": 9480
    },
    {
      "entropy": 1.1566391289234161,
      "epoch": 0.2718301992180227,
      "grad_norm": 8.25,
      "learning_rate": 4.8146459907533805e-05,
      "loss": 1.2178935050964355,
      "mean_token_accuracy": 0.7282591745257377,
      "num_tokens": 18743686.0,
      "step": 9490
    },
    {
      "entropy": 1.21599200963974,
      "epoch": 0.27211663778411127,
      "grad_norm": 7.09375,
      "learning_rate": 4.814214535344884e-05,
      "loss": 1.2984207153320313,
      "mean_token_accuracy": 0.713078685849905,
      "num_tokens": 18764538.0,
      "step": 9500
    },
    {
      "entropy": 1.238351733610034,
      "epoch": 0.2724030763501998,
      "grad_norm": 6.75,
      "learning_rate": 4.813782597740465e-05,
      "loss": 1.2873254776000977,
      "mean_token_accuracy": 0.7136316448450089,
      "num_tokens": 18785455.0,
      "step": 9510
    },
    {
      "entropy": 1.3914471410214901,
      "epoch": 0.27268951491628834,
      "grad_norm": 8.125,
      "learning_rate": 4.81335017803012e-05,
      "loss": 1.3496410369873046,
      "mean_token_accuracy": 0.7046843189746141,
      "num_tokens": 18804885.0,
      "step": 9520
    },
    {
      "entropy": 1.22066420763731,
      "epoch": 0.27297595348237685,
      "grad_norm": 5.59375,
      "learning_rate": 4.812917276303952e-05,
      "loss": 1.266666793823242,
      "mean_token_accuracy": 0.7137233637273311,
      "num_tokens": 18824488.0,
      "step": 9530
    },
    {
      "entropy": 1.3131001925095915,
      "epoch": 0.2732623920484654,
      "grad_norm": 9.0625,
      "learning_rate": 4.812483892652159e-05,
      "loss": 1.3390828132629395,
      "mean_token_accuracy": 0.7044373240321875,
      "num_tokens": 18842866.0,
      "step": 9540
    },
    {
      "entropy": 1.1892004791647195,
      "epoch": 0.27354883061455393,
      "grad_norm": 10.5625,
      "learning_rate": 4.812050027165044e-05,
      "loss": 1.2279635429382325,
      "mean_token_accuracy": 0.7233818676322699,
      "num_tokens": 18863108.0,
      "step": 9550
    },
    {
      "entropy": 1.2994504366070032,
      "epoch": 0.2738352691806425,
      "grad_norm": 8.8125,
      "learning_rate": 4.811615679933007e-05,
      "loss": 1.3659720420837402,
      "mean_token_accuracy": 0.7011908940970898,
      "num_tokens": 18883222.0,
      "step": 9560
    },
    {
      "entropy": 1.3316714242100716,
      "epoch": 0.274121707746731,
      "grad_norm": 7.375,
      "learning_rate": 4.811180851046551e-05,
      "loss": 1.3249066352844239,
      "mean_token_accuracy": 0.6979438884183764,
      "num_tokens": 18902895.0,
      "step": 9570
    },
    {
      "entropy": 1.373481733724475,
      "epoch": 0.27440814631281957,
      "grad_norm": 5.78125,
      "learning_rate": 4.810745540596276e-05,
      "loss": 1.434743595123291,
      "mean_token_accuracy": 0.6876088410615921,
      "num_tokens": 18923925.0,
      "step": 9580
    },
    {
      "entropy": 1.281962576508522,
      "epoch": 0.2746945848789081,
      "grad_norm": 9.9375,
      "learning_rate": 4.810309748672888e-05,
      "loss": 1.2731183052062989,
      "mean_token_accuracy": 0.710417527705431,
      "num_tokens": 18943165.0,
      "step": 9590
    },
    {
      "entropy": 1.2833354998379947,
      "epoch": 0.27498102344499664,
      "grad_norm": 8.4375,
      "learning_rate": 4.809873475367186e-05,
      "loss": 1.3538728713989259,
      "mean_token_accuracy": 0.7047929227352142,
      "num_tokens": 18963454.0,
      "step": 9600
    },
    {
      "entropy": 1.2634608965367078,
      "epoch": 0.27526746201108515,
      "grad_norm": 8.0625,
      "learning_rate": 4.8094367207700754e-05,
      "loss": 1.3354544639587402,
      "mean_token_accuracy": 0.7013860233128071,
      "num_tokens": 18983623.0,
      "step": 9610
    },
    {
      "entropy": 1.2145154062658547,
      "epoch": 0.2755539005771737,
      "grad_norm": 9.25,
      "learning_rate": 4.8089994849725596e-05,
      "loss": 1.2072470664978028,
      "mean_token_accuracy": 0.7160839352756738,
      "num_tokens": 19002971.0,
      "step": 9620
    },
    {
      "entropy": 1.3010970495641232,
      "epoch": 0.2758403391432622,
      "grad_norm": 10.625,
      "learning_rate": 4.80856176806574e-05,
      "loss": 1.268671703338623,
      "mean_token_accuracy": 0.7222683660686016,
      "num_tokens": 19021723.0,
      "step": 9630
    },
    {
      "entropy": 1.2018229987472295,
      "epoch": 0.2761267777093508,
      "grad_norm": 7.28125,
      "learning_rate": 4.808123570140823e-05,
      "loss": 1.2600168228149413,
      "mean_token_accuracy": 0.7058022413402796,
      "num_tokens": 19041071.0,
      "step": 9640
    },
    {
      "entropy": 1.252623599767685,
      "epoch": 0.2764132162754393,
      "grad_norm": 7.40625,
      "learning_rate": 4.807684891289111e-05,
      "loss": 1.3150317192077636,
      "mean_token_accuracy": 0.7052531097084284,
      "num_tokens": 19060496.0,
      "step": 9650
    },
    {
      "entropy": 1.3609641756862403,
      "epoch": 0.27669965484152786,
      "grad_norm": 8.125,
      "learning_rate": 4.807245731602009e-05,
      "loss": 1.2961626052856445,
      "mean_token_accuracy": 0.7096053510904312,
      "num_tokens": 19079857.0,
      "step": 9660
    },
    {
      "entropy": 1.1585993632674216,
      "epoch": 0.2769860934076164,
      "grad_norm": 8.0,
      "learning_rate": 4.806806091171022e-05,
      "loss": 1.1728649139404297,
      "mean_token_accuracy": 0.7339285384863615,
      "num_tokens": 19099319.0,
      "step": 9670
    },
    {
      "entropy": 1.1269067496061325,
      "epoch": 0.27727253197370494,
      "grad_norm": 6.28125,
      "learning_rate": 4.806365970087754e-05,
      "loss": 1.1625200271606446,
      "mean_token_accuracy": 0.7388526450842618,
      "num_tokens": 19119539.0,
      "step": 9680
    },
    {
      "entropy": 1.213476287573576,
      "epoch": 0.2775589705397935,
      "grad_norm": 7.46875,
      "learning_rate": 4.80592536844391e-05,
      "loss": 1.238703155517578,
      "mean_token_accuracy": 0.7205547297373414,
      "num_tokens": 19139271.0,
      "step": 9690
    },
    {
      "entropy": 1.1873043950647115,
      "epoch": 0.277845409105882,
      "grad_norm": 8.3125,
      "learning_rate": 4.805484286331294e-05,
      "loss": 1.2164966583251953,
      "mean_token_accuracy": 0.7208237007260323,
      "num_tokens": 19160395.0,
      "step": 9700
    },
    {
      "entropy": 1.1731833301484584,
      "epoch": 0.2781318476719706,
      "grad_norm": 7.90625,
      "learning_rate": 4.805042723841813e-05,
      "loss": 1.187636661529541,
      "mean_token_accuracy": 0.7279136903584004,
      "num_tokens": 19180462.0,
      "step": 9710
    },
    {
      "entropy": 1.1782655270770193,
      "epoch": 0.2784182862380591,
      "grad_norm": 9.375,
      "learning_rate": 4.8046006810674705e-05,
      "loss": 1.2543769836425782,
      "mean_token_accuracy": 0.7170453231781722,
      "num_tokens": 19199875.0,
      "step": 9720
    },
    {
      "entropy": 1.2035142604261637,
      "epoch": 0.27870472480414765,
      "grad_norm": 6.4375,
      "learning_rate": 4.804158158100372e-05,
      "loss": 1.2652222633361816,
      "mean_token_accuracy": 0.7143611006438733,
      "num_tokens": 19220080.0,
      "step": 9730
    },
    {
      "entropy": 1.2720818102359772,
      "epoch": 0.27899116337023616,
      "grad_norm": 7.03125,
      "learning_rate": 4.8037151550327234e-05,
      "loss": 1.2831024169921874,
      "mean_token_accuracy": 0.7145038194954395,
      "num_tokens": 19239545.0,
      "step": 9740
    },
    {
      "entropy": 1.2525265368632972,
      "epoch": 0.2792776019363247,
      "grad_norm": 9.375,
      "learning_rate": 4.80327167195683e-05,
      "loss": 1.2966771125793457,
      "mean_token_accuracy": 0.7163520272821188,
      "num_tokens": 19258693.0,
      "step": 9750
    },
    {
      "entropy": 1.2273115681484341,
      "epoch": 0.27956404050241324,
      "grad_norm": 9.1875,
      "learning_rate": 4.802827708965096e-05,
      "loss": 1.2193683624267577,
      "mean_token_accuracy": 0.7216869089752436,
      "num_tokens": 19279707.0,
      "step": 9760
    },
    {
      "entropy": 1.1521436281502246,
      "epoch": 0.2798504790685018,
      "grad_norm": 6.03125,
      "learning_rate": 4.802383266150028e-05,
      "loss": 1.1548256874084473,
      "mean_token_accuracy": 0.7361404214054346,
      "num_tokens": 19299758.0,
      "step": 9770
    },
    {
      "entropy": 1.2253732860088349,
      "epoch": 0.2801369176345903,
      "grad_norm": 10.0,
      "learning_rate": 4.80193834360423e-05,
      "loss": 1.259799861907959,
      "mean_token_accuracy": 0.717500614747405,
      "num_tokens": 19319031.0,
      "step": 9780
    },
    {
      "entropy": 1.3701102506369351,
      "epoch": 0.2804233562006789,
      "grad_norm": 7.125,
      "learning_rate": 4.8014929414204084e-05,
      "loss": 1.3844647407531738,
      "mean_token_accuracy": 0.6849213771522045,
      "num_tokens": 19339202.0,
      "step": 9790
    },
    {
      "entropy": 1.1747784547507762,
      "epoch": 0.2807097947667674,
      "grad_norm": 10.4375,
      "learning_rate": 4.801047059691368e-05,
      "loss": 1.195345973968506,
      "mean_token_accuracy": 0.7189787935465575,
      "num_tokens": 19358633.0,
      "step": 9800
    },
    {
      "entropy": 1.1420799817889928,
      "epoch": 0.28099623333285595,
      "grad_norm": 7.125,
      "learning_rate": 4.800600698510014e-05,
      "loss": 1.1610635757446288,
      "mean_token_accuracy": 0.7388739325106144,
      "num_tokens": 19378085.0,
      "step": 9810
    },
    {
      "entropy": 1.1826870456337928,
      "epoch": 0.28128267189894446,
      "grad_norm": 5.8125,
      "learning_rate": 4.800153857969351e-05,
      "loss": 1.251106071472168,
      "mean_token_accuracy": 0.718170203641057,
      "num_tokens": 19397525.0,
      "step": 9820
    },
    {
      "entropy": 1.2163832444697618,
      "epoch": 0.281569110465033,
      "grad_norm": 6.125,
      "learning_rate": 4.7997065381624846e-05,
      "loss": 1.2410301208496093,
      "mean_token_accuracy": 0.7348678473383188,
      "num_tokens": 19419740.0,
      "step": 9830
    },
    {
      "entropy": 1.3619404435157776,
      "epoch": 0.28185554903112153,
      "grad_norm": 7.0625,
      "learning_rate": 4.799258739182619e-05,
      "loss": 1.433720874786377,
      "mean_token_accuracy": 0.6913000743836164,
      "num_tokens": 19439721.0,
      "step": 9840
    },
    {
      "entropy": 1.2456842597573996,
      "epoch": 0.2821419875972101,
      "grad_norm": 8.0,
      "learning_rate": 4.798810461123059e-05,
      "loss": 1.2306564331054688,
      "mean_token_accuracy": 0.7235556855797768,
      "num_tokens": 19457940.0,
      "step": 9850
    },
    {
      "entropy": 1.3578756723552943,
      "epoch": 0.2824284261632986,
      "grad_norm": 8.375,
      "learning_rate": 4.7983617040772086e-05,
      "loss": 1.4068643569946289,
      "mean_token_accuracy": 0.6960411306470633,
      "num_tokens": 19477065.0,
      "step": 9860
    },
    {
      "entropy": 1.2337223421782255,
      "epoch": 0.28271486472938717,
      "grad_norm": 8.5625,
      "learning_rate": 4.797912468138572e-05,
      "loss": 1.2814136505126954,
      "mean_token_accuracy": 0.7116973515599966,
      "num_tokens": 19496370.0,
      "step": 9870
    },
    {
      "entropy": 1.158883135765791,
      "epoch": 0.2830013032954757,
      "grad_norm": 5.59375,
      "learning_rate": 4.7974627534007544e-05,
      "loss": 1.1135085105895997,
      "mean_token_accuracy": 0.7410048365592956,
      "num_tokens": 19517537.0,
      "step": 9880
    },
    {
      "entropy": 1.236505088582635,
      "epoch": 0.28328774186156425,
      "grad_norm": 7.46875,
      "learning_rate": 4.797012559957458e-05,
      "loss": 1.3544401168823241,
      "mean_token_accuracy": 0.7023403074592351,
      "num_tokens": 19537138.0,
      "step": 9890
    },
    {
      "entropy": 1.3292220380157231,
      "epoch": 0.28357418042765276,
      "grad_norm": 10.5625,
      "learning_rate": 4.7965618879024876e-05,
      "loss": 1.3328524589538575,
      "mean_token_accuracy": 0.7065465740859509,
      "num_tokens": 19556009.0,
      "step": 9900
    },
    {
      "entropy": 1.284081007540226,
      "epoch": 0.2838606189937413,
      "grad_norm": 9.5625,
      "learning_rate": 4.796110737329745e-05,
      "loss": 1.2665791511535645,
      "mean_token_accuracy": 0.7108983382582664,
      "num_tokens": 19576505.0,
      "step": 9910
    },
    {
      "entropy": 1.2333140067756176,
      "epoch": 0.28414705755982983,
      "grad_norm": 7.5625,
      "learning_rate": 4.795659108333236e-05,
      "loss": 1.1999363899230957,
      "mean_token_accuracy": 0.7268323682248592,
      "num_tokens": 19596879.0,
      "step": 9920
    },
    {
      "entropy": 1.2496559236198663,
      "epoch": 0.2844334961259184,
      "grad_norm": 7.6875,
      "learning_rate": 4.7952070010070594e-05,
      "loss": 1.3359495162963868,
      "mean_token_accuracy": 0.7083397228270769,
      "num_tokens": 19616634.0,
      "step": 9930
    },
    {
      "entropy": 1.237235459499061,
      "epoch": 0.2847199346920069,
      "grad_norm": 9.4375,
      "learning_rate": 4.7947544154454196e-05,
      "loss": 1.234358787536621,
      "mean_token_accuracy": 0.7246622666716576,
      "num_tokens": 19637092.0,
      "step": 9940
    },
    {
      "entropy": 1.1274428755044936,
      "epoch": 0.28500637325809547,
      "grad_norm": 9.9375,
      "learning_rate": 4.7943013517426183e-05,
      "loss": 1.0833609580993653,
      "mean_token_accuracy": 0.747896709293127,
      "num_tokens": 19656900.0,
      "step": 9950
    },
    {
      "entropy": 1.2382439866662025,
      "epoch": 0.28529281182418403,
      "grad_norm": 7.4375,
      "learning_rate": 4.793847809993057e-05,
      "loss": 1.3295849800109862,
      "mean_token_accuracy": 0.7127545036375522,
      "num_tokens": 19675323.0,
      "step": 9960
    },
    {
      "entropy": 1.2497281353920697,
      "epoch": 0.28557925039027254,
      "grad_norm": 8.625,
      "learning_rate": 4.7933937902912375e-05,
      "loss": 1.2507732391357422,
      "mean_token_accuracy": 0.7235575128346682,
      "num_tokens": 19695182.0,
      "step": 9970
    },
    {
      "entropy": 1.1741746734827756,
      "epoch": 0.2858656889563611,
      "grad_norm": 6.90625,
      "learning_rate": 4.79293929273176e-05,
      "loss": 1.233245277404785,
      "mean_token_accuracy": 0.7213643684983253,
      "num_tokens": 19713855.0,
      "step": 9980
    },
    {
      "entropy": 1.2821607805788517,
      "epoch": 0.2861521275224496,
      "grad_norm": 8.375,
      "learning_rate": 4.792484317409324e-05,
      "loss": 1.3093585968017578,
      "mean_token_accuracy": 0.6996051449328661,
      "num_tokens": 19733770.0,
      "step": 9990
    },
    {
      "entropy": 1.2294521793723105,
      "epoch": 0.2864385660885382,
      "grad_norm": 9.875,
      "learning_rate": 4.79202886441873e-05,
      "loss": 1.2560036659240723,
      "mean_token_accuracy": 0.7066733181476593,
      "num_tokens": 19753166.0,
      "step": 10000
    },
    {
      "entropy": 1.28849596939981,
      "epoch": 0.2867250046546267,
      "grad_norm": 10.4375,
      "learning_rate": 4.791572933854879e-05,
      "loss": 1.3138832092285155,
      "mean_token_accuracy": 0.6822977267205715,
      "num_tokens": 19773167.0,
      "step": 10010
    },
    {
      "entropy": 1.2579356957226993,
      "epoch": 0.28701144322071526,
      "grad_norm": 7.15625,
      "learning_rate": 4.791116525812767e-05,
      "loss": 1.328775119781494,
      "mean_token_accuracy": 0.6985458929091692,
      "num_tokens": 19793810.0,
      "step": 10020
    },
    {
      "entropy": 1.2042025054804981,
      "epoch": 0.28729788178680377,
      "grad_norm": 6.0,
      "learning_rate": 4.790659640387494e-05,
      "loss": 1.2034329414367675,
      "mean_token_accuracy": 0.7252053562551737,
      "num_tokens": 19812874.0,
      "step": 10030
    },
    {
      "entropy": 1.27395105548203,
      "epoch": 0.28758432035289233,
      "grad_norm": 7.78125,
      "learning_rate": 4.7902022776742586e-05,
      "loss": 1.304868221282959,
      "mean_token_accuracy": 0.7137073144316674,
      "num_tokens": 19832721.0,
      "step": 10040
    },
    {
      "entropy": 1.158857085183263,
      "epoch": 0.28787075891898084,
      "grad_norm": 7.875,
      "learning_rate": 4.789744437768356e-05,
      "loss": 1.1979079246520996,
      "mean_token_accuracy": 0.7325722940266133,
      "num_tokens": 19853344.0,
      "step": 10050
    },
    {
      "entropy": 1.2469136916100978,
      "epoch": 0.2881571974850694,
      "grad_norm": 7.25,
      "learning_rate": 4.789286120765185e-05,
      "loss": 1.2818082809448241,
      "mean_token_accuracy": 0.7125504605472088,
      "num_tokens": 19874109.0,
      "step": 10060
    },
    {
      "entropy": 1.1972002856433392,
      "epoch": 0.2884436360511579,
      "grad_norm": 8.375,
      "learning_rate": 4.7888273267602414e-05,
      "loss": 1.247776699066162,
      "mean_token_accuracy": 0.720183490216732,
      "num_tokens": 19893140.0,
      "step": 10070
    },
    {
      "entropy": 1.1542341858148575,
      "epoch": 0.2887300746172465,
      "grad_norm": 8.375,
      "learning_rate": 4.7883680558491196e-05,
      "loss": 1.202667808532715,
      "mean_token_accuracy": 0.724473912641406,
      "num_tokens": 19913180.0,
      "step": 10080
    },
    {
      "entropy": 1.2531606178730725,
      "epoch": 0.289016513183335,
      "grad_norm": 8.8125,
      "learning_rate": 4.7879083081275156e-05,
      "loss": 1.212215518951416,
      "mean_token_accuracy": 0.7197392519563437,
      "num_tokens": 19933746.0,
      "step": 10090
    },
    {
      "entropy": 1.2108169622719287,
      "epoch": 0.28930295174942355,
      "grad_norm": 7.4375,
      "learning_rate": 4.787448083691223e-05,
      "loss": 1.2483904838562012,
      "mean_token_accuracy": 0.7140415642410517,
      "num_tokens": 19953380.0,
      "step": 10100
    },
    {
      "entropy": 1.2383288452401757,
      "epoch": 0.28958939031551206,
      "grad_norm": 10.9375,
      "learning_rate": 4.786987382636136e-05,
      "loss": 1.2708030700683595,
      "mean_token_accuracy": 0.7138889364898204,
      "num_tokens": 19973049.0,
      "step": 10110
    },
    {
      "entropy": 1.2284024387598038,
      "epoch": 0.28987582888160063,
      "grad_norm": 8.5625,
      "learning_rate": 4.786526205058246e-05,
      "loss": 1.2730886459350585,
      "mean_token_accuracy": 0.722532881423831,
      "num_tokens": 19992357.0,
      "step": 10120
    },
    {
      "entropy": 1.2366085767745971,
      "epoch": 0.29016226744768914,
      "grad_norm": 9.9375,
      "learning_rate": 4.7860645510536475e-05,
      "loss": 1.2555581092834474,
      "mean_token_accuracy": 0.7109288562089204,
      "num_tokens": 20012033.0,
      "step": 10130
    },
    {
      "entropy": 1.259175629541278,
      "epoch": 0.2904487060137777,
      "grad_norm": 7.15625,
      "learning_rate": 4.78560242071853e-05,
      "loss": 1.2749428749084473,
      "mean_token_accuracy": 0.7174477286636829,
      "num_tokens": 20031070.0,
      "step": 10140
    },
    {
      "entropy": 1.3015303812921046,
      "epoch": 0.2907351445798662,
      "grad_norm": 8.3125,
      "learning_rate": 4.7851398141491855e-05,
      "loss": 1.3474379539489747,
      "mean_token_accuracy": 0.699538454040885,
      "num_tokens": 20050478.0,
      "step": 10150
    },
    {
      "entropy": 1.1652599275112152,
      "epoch": 0.2910215831459548,
      "grad_norm": 8.625,
      "learning_rate": 4.784676731442004e-05,
      "loss": 1.1608721733093261,
      "mean_token_accuracy": 0.7359152492135763,
      "num_tokens": 20070056.0,
      "step": 10160
    },
    {
      "entropy": 1.1339007444679736,
      "epoch": 0.2913080217120433,
      "grad_norm": 9.6875,
      "learning_rate": 4.784213172693473e-05,
      "loss": 1.156212615966797,
      "mean_token_accuracy": 0.7336442414671183,
      "num_tokens": 20088606.0,
      "step": 10170
    },
    {
      "entropy": 1.1140047937631607,
      "epoch": 0.29159446027813185,
      "grad_norm": 8.625,
      "learning_rate": 4.783749138000183e-05,
      "loss": 1.1142316818237306,
      "mean_token_accuracy": 0.7451892249286175,
      "num_tokens": 20107961.0,
      "step": 10180
    },
    {
      "entropy": 1.2975910264998674,
      "epoch": 0.29188089884422036,
      "grad_norm": 7.25,
      "learning_rate": 4.78328462745882e-05,
      "loss": 1.4363594055175781,
      "mean_token_accuracy": 0.694925444573164,
      "num_tokens": 20127462.0,
      "step": 10190
    },
    {
      "entropy": 1.2322942119091749,
      "epoch": 0.2921673374103089,
      "grad_norm": 7.625,
      "learning_rate": 4.7828196411661705e-05,
      "loss": 1.2529497146606445,
      "mean_token_accuracy": 0.7259122528135776,
      "num_tokens": 20146253.0,
      "step": 10200
    },
    {
      "entropy": 1.239514384791255,
      "epoch": 0.2924537759763975,
      "grad_norm": 12.1875,
      "learning_rate": 4.782354179219122e-05,
      "loss": 1.2916205406188965,
      "mean_token_accuracy": 0.6989834289997816,
      "num_tokens": 20165620.0,
      "step": 10210
    },
    {
      "entropy": 1.2953903215005993,
      "epoch": 0.292740214542486,
      "grad_norm": 8.75,
      "learning_rate": 4.7818882417146584e-05,
      "loss": 1.3701395034790038,
      "mean_token_accuracy": 0.7016857746988535,
      "num_tokens": 20185628.0,
      "step": 10220
    },
    {
      "entropy": 1.2485311467200517,
      "epoch": 0.29302665310857456,
      "grad_norm": 7.40625,
      "learning_rate": 4.781421828749863e-05,
      "loss": 1.196932888031006,
      "mean_token_accuracy": 0.7228483323007822,
      "num_tokens": 20204149.0,
      "step": 10230
    },
    {
      "entropy": 1.2259568341076374,
      "epoch": 0.2933130916746631,
      "grad_norm": 9.0625,
      "learning_rate": 4.78095494042192e-05,
      "loss": 1.2339511871337892,
      "mean_token_accuracy": 0.7088088352233172,
      "num_tokens": 20223293.0,
      "step": 10240
    },
    {
      "entropy": 1.2301707200706005,
      "epoch": 0.29359953024075164,
      "grad_norm": 7.46875,
      "learning_rate": 4.78048757682811e-05,
      "loss": 1.263758659362793,
      "mean_token_accuracy": 0.7220470685511827,
      "num_tokens": 20241746.0,
      "step": 10250
    },
    {
      "entropy": 1.1846627168357373,
      "epoch": 0.29388596880684015,
      "grad_norm": 11.8125,
      "learning_rate": 4.7800197380658153e-05,
      "loss": 1.1950181007385254,
      "mean_token_accuracy": 0.728680620715022,
      "num_tokens": 20263134.0,
      "step": 10260
    },
    {
      "entropy": 1.22404460683465,
      "epoch": 0.2941724073729287,
      "grad_norm": 9.0625,
      "learning_rate": 4.779551424232516e-05,
      "loss": 1.3358233451843262,
      "mean_token_accuracy": 0.7044001743197441,
      "num_tokens": 20283085.0,
      "step": 10270
    },
    {
      "entropy": 1.3748682785779238,
      "epoch": 0.2944588459390172,
      "grad_norm": 7.125,
      "learning_rate": 4.779082635425791e-05,
      "loss": 1.3425874710083008,
      "mean_token_accuracy": 0.7075872421264648,
      "num_tokens": 20302780.0,
      "step": 10280
    },
    {
      "entropy": 1.3219470448791981,
      "epoch": 0.2947452845051058,
      "grad_norm": 10.0625,
      "learning_rate": 4.778613371743319e-05,
      "loss": 1.3717927932739258,
      "mean_token_accuracy": 0.6967399388551712,
      "num_tokens": 20322288.0,
      "step": 10290
    },
    {
      "entropy": 1.2808030743151904,
      "epoch": 0.2950317230711943,
      "grad_norm": 8.875,
      "learning_rate": 4.778143633282876e-05,
      "loss": 1.3541928291320802,
      "mean_token_accuracy": 0.6995458628982305,
      "num_tokens": 20342336.0,
      "step": 10300
    },
    {
      "entropy": 1.2334368931129576,
      "epoch": 0.29531816163728286,
      "grad_norm": 8.8125,
      "learning_rate": 4.777673420142338e-05,
      "loss": 1.2389339447021483,
      "mean_token_accuracy": 0.7194286450743675,
      "num_tokens": 20362570.0,
      "step": 10310
    },
    {
      "entropy": 1.3115269932895899,
      "epoch": 0.29560460020337137,
      "grad_norm": 6.625,
      "learning_rate": 4.777202732419681e-05,
      "loss": 1.383128833770752,
      "mean_token_accuracy": 0.6994543489068746,
      "num_tokens": 20383321.0,
      "step": 10320
    },
    {
      "entropy": 1.2894154995679856,
      "epoch": 0.29589103876945994,
      "grad_norm": 8.0625,
      "learning_rate": 4.7767315702129774e-05,
      "loss": 1.2736214637756347,
      "mean_token_accuracy": 0.7091438576579094,
      "num_tokens": 20402286.0,
      "step": 10330
    },
    {
      "entropy": 1.278226450830698,
      "epoch": 0.29617747733554844,
      "grad_norm": 10.6875,
      "learning_rate": 4.7762599336204005e-05,
      "loss": 1.281544303894043,
      "mean_token_accuracy": 0.7182073879987001,
      "num_tokens": 20421520.0,
      "step": 10340
    },
    {
      "entropy": 1.1996090399101376,
      "epoch": 0.296463915901637,
      "grad_norm": 6.53125,
      "learning_rate": 4.775787822740222e-05,
      "loss": 1.2525644302368164,
      "mean_token_accuracy": 0.731365493312478,
      "num_tokens": 20442233.0,
      "step": 10350
    },
    {
      "entropy": 1.1005093213170767,
      "epoch": 0.2967503544677255,
      "grad_norm": 6.90625,
      "learning_rate": 4.775315237670812e-05,
      "loss": 1.1033694267272949,
      "mean_token_accuracy": 0.7439654916524887,
      "num_tokens": 20461079.0,
      "step": 10360
    },
    {
      "entropy": 1.181967235542834,
      "epoch": 0.2970367930338141,
      "grad_norm": 8.375,
      "learning_rate": 4.774842178510639e-05,
      "loss": 1.2287528038024902,
      "mean_token_accuracy": 0.7279417492449284,
      "num_tokens": 20479607.0,
      "step": 10370
    },
    {
      "entropy": 1.2588371988385916,
      "epoch": 0.2973232315999026,
      "grad_norm": 9.5625,
      "learning_rate": 4.774368645358271e-05,
      "loss": 1.2836091995239258,
      "mean_token_accuracy": 0.7170934330672025,
      "num_tokens": 20498614.0,
      "step": 10380
    },
    {
      "entropy": 1.266011572163552,
      "epoch": 0.29760967016599116,
      "grad_norm": 9.3125,
      "learning_rate": 4.773894638312376e-05,
      "loss": 1.3395215034484864,
      "mean_token_accuracy": 0.7086238369345665,
      "num_tokens": 20518950.0,
      "step": 10390
    },
    {
      "entropy": 1.3530218362808228,
      "epoch": 0.29789610873207967,
      "grad_norm": 9.875,
      "learning_rate": 4.7734201574717185e-05,
      "loss": 1.4200379371643066,
      "mean_token_accuracy": 0.6952762430533767,
      "num_tokens": 20538880.0,
      "step": 10400
    },
    {
      "entropy": 1.2130294106900692,
      "epoch": 0.29818254729816823,
      "grad_norm": 7.375,
      "learning_rate": 4.7729452029351616e-05,
      "loss": 1.2266550064086914,
      "mean_token_accuracy": 0.720143285766244,
      "num_tokens": 20559866.0,
      "step": 10410
    },
    {
      "entropy": 1.1608878757804633,
      "epoch": 0.29846898586425674,
      "grad_norm": 5.9375,
      "learning_rate": 4.772469774801669e-05,
      "loss": 1.120636558532715,
      "mean_token_accuracy": 0.7369175810366869,
      "num_tokens": 20579174.0,
      "step": 10420
    },
    {
      "entropy": 1.1649308599531651,
      "epoch": 0.2987554244303453,
      "grad_norm": 7.9375,
      "learning_rate": 4.771993873170303e-05,
      "loss": 1.1998584747314454,
      "mean_token_accuracy": 0.7233378186821937,
      "num_tokens": 20599116.0,
      "step": 10430
    },
    {
      "entropy": 1.206273192167282,
      "epoch": 0.2990418629964338,
      "grad_norm": 6.78125,
      "learning_rate": 4.7715174981402225e-05,
      "loss": 1.258732795715332,
      "mean_token_accuracy": 0.7155660688877106,
      "num_tokens": 20618958.0,
      "step": 10440
    },
    {
      "entropy": 1.209686228260398,
      "epoch": 0.2993283015625224,
      "grad_norm": 9.25,
      "learning_rate": 4.771040649810686e-05,
      "loss": 1.2455236434936523,
      "mean_token_accuracy": 0.7279234286397696,
      "num_tokens": 20639688.0,
      "step": 10450
    },
    {
      "entropy": 1.2300740577280522,
      "epoch": 0.2996147401286109,
      "grad_norm": 12.0,
      "learning_rate": 4.770563328281052e-05,
      "loss": 1.2667814254760743,
      "mean_token_accuracy": 0.714299051836133,
      "num_tokens": 20659168.0,
      "step": 10460
    },
    {
      "entropy": 1.151146410405636,
      "epoch": 0.29990117869469946,
      "grad_norm": 7.25,
      "learning_rate": 4.770085533650776e-05,
      "loss": 1.1962030410766602,
      "mean_token_accuracy": 0.7382386907935142,
      "num_tokens": 20677516.0,
      "step": 10470
    },
    {
      "entropy": 1.2397548377513885,
      "epoch": 0.300187617260788,
      "grad_norm": 8.625,
      "learning_rate": 4.7696072660194124e-05,
      "loss": 1.2270134925842284,
      "mean_token_accuracy": 0.7233962573111057,
      "num_tokens": 20698074.0,
      "step": 10480
    },
    {
      "entropy": 1.241117611527443,
      "epoch": 0.30047405582687653,
      "grad_norm": 8.875,
      "learning_rate": 4.769128525486615e-05,
      "loss": 1.2556010246276856,
      "mean_token_accuracy": 0.7220278877764941,
      "num_tokens": 20717765.0,
      "step": 10490
    },
    {
      "entropy": 1.2247598566114903,
      "epoch": 0.3007604943929651,
      "grad_norm": 7.5,
      "learning_rate": 4.7686493121521336e-05,
      "loss": 1.2303947448730468,
      "mean_token_accuracy": 0.7157254960387945,
      "num_tokens": 20737590.0,
      "step": 10500
    },
    {
      "entropy": 1.1269052591174842,
      "epoch": 0.3010469329590536,
      "grad_norm": 9.0625,
      "learning_rate": 4.768169626115821e-05,
      "loss": 1.1731462478637695,
      "mean_token_accuracy": 0.7308567386120558,
      "num_tokens": 20756224.0,
      "step": 10510
    },
    {
      "entropy": 1.339089971035719,
      "epoch": 0.30133337152514217,
      "grad_norm": 6.65625,
      "learning_rate": 4.767689467477624e-05,
      "loss": 1.3878726959228516,
      "mean_token_accuracy": 0.6907770108431578,
      "num_tokens": 20775368.0,
      "step": 10520
    },
    {
      "entropy": 1.2001933190971612,
      "epoch": 0.3016198100912307,
      "grad_norm": 7.1875,
      "learning_rate": 4.767208836337591e-05,
      "loss": 1.2366168022155761,
      "mean_token_accuracy": 0.7184179678559304,
      "num_tokens": 20795163.0,
      "step": 10530
    },
    {
      "entropy": 1.2310813583433629,
      "epoch": 0.30190624865731924,
      "grad_norm": 9.375,
      "learning_rate": 4.766727732795866e-05,
      "loss": 1.2524584770202636,
      "mean_token_accuracy": 0.7192334160208702,
      "num_tokens": 20814193.0,
      "step": 10540
    },
    {
      "entropy": 1.3205643955618143,
      "epoch": 0.30219268722340775,
      "grad_norm": 7.34375,
      "learning_rate": 4.766246156952695e-05,
      "loss": 1.3525377273559571,
      "mean_token_accuracy": 0.6964046154171228,
      "num_tokens": 20835207.0,
      "step": 10550
    },
    {
      "entropy": 1.3620137978345155,
      "epoch": 0.3024791257894963,
      "grad_norm": 7.125,
      "learning_rate": 4.7657641089084184e-05,
      "loss": 1.3694384574890137,
      "mean_token_accuracy": 0.6839216224849224,
      "num_tokens": 20855984.0,
      "step": 10560
    },
    {
      "entropy": 1.2748428869992494,
      "epoch": 0.3027655643555848,
      "grad_norm": 10.1875,
      "learning_rate": 4.765281588763479e-05,
      "loss": 1.3202522277832032,
      "mean_token_accuracy": 0.7001968938857317,
      "num_tokens": 20876272.0,
      "step": 10570
    },
    {
      "entropy": 1.2867596458643675,
      "epoch": 0.3030520029216734,
      "grad_norm": 8.1875,
      "learning_rate": 4.764798596618414e-05,
      "loss": 1.2923958778381348,
      "mean_token_accuracy": 0.7128220777958632,
      "num_tokens": 20897178.0,
      "step": 10580
    },
    {
      "entropy": 1.2885660350322723,
      "epoch": 0.3033384414877619,
      "grad_norm": 5.84375,
      "learning_rate": 4.764315132573862e-05,
      "loss": 1.3065492630004882,
      "mean_token_accuracy": 0.7092504814267159,
      "num_tokens": 20917917.0,
      "step": 10590
    },
    {
      "entropy": 1.1919785939157008,
      "epoch": 0.30362488005385047,
      "grad_norm": 6.9375,
      "learning_rate": 4.763831196730559e-05,
      "loss": 1.2493278503417968,
      "mean_token_accuracy": 0.7136258292943239,
      "num_tokens": 20936362.0,
      "step": 10600
    },
    {
      "entropy": 1.2055115677416324,
      "epoch": 0.303911318619939,
      "grad_norm": 6.8125,
      "learning_rate": 4.76334678918934e-05,
      "loss": 1.2522308349609375,
      "mean_token_accuracy": 0.7191043566912413,
      "num_tokens": 20954995.0,
      "step": 10610
    },
    {
      "entropy": 1.2108305536210537,
      "epoch": 0.30419775718602754,
      "grad_norm": 11.5,
      "learning_rate": 4.7628619100511354e-05,
      "loss": 1.210112953186035,
      "mean_token_accuracy": 0.722658408433199,
      "num_tokens": 20974260.0,
      "step": 10620
    },
    {
      "entropy": 1.2318328015506268,
      "epoch": 0.30448419575211605,
      "grad_norm": 8.25,
      "learning_rate": 4.762376559416977e-05,
      "loss": 1.285456943511963,
      "mean_token_accuracy": 0.7226238247007132,
      "num_tokens": 20995089.0,
      "step": 10630
    },
    {
      "entropy": 1.2235995415598153,
      "epoch": 0.3047706343182046,
      "grad_norm": 7.9375,
      "learning_rate": 4.761890737387993e-05,
      "loss": 1.2320636749267577,
      "mean_token_accuracy": 0.7227725703269243,
      "num_tokens": 21015865.0,
      "step": 10640
    },
    {
      "entropy": 1.1759483505040407,
      "epoch": 0.3050570728842931,
      "grad_norm": 8.0,
      "learning_rate": 4.7614044440654114e-05,
      "loss": 1.1156691551208495,
      "mean_token_accuracy": 0.7398465003818273,
      "num_tokens": 21035144.0,
      "step": 10650
    },
    {
      "entropy": 1.2341271406039596,
      "epoch": 0.3053435114503817,
      "grad_norm": 7.75,
      "learning_rate": 4.760917679550558e-05,
      "loss": 1.2832679748535156,
      "mean_token_accuracy": 0.7126728795468807,
      "num_tokens": 21055609.0,
      "step": 10660
    },
    {
      "entropy": 1.1682598849758505,
      "epoch": 0.3056299500164702,
      "grad_norm": 9.5625,
      "learning_rate": 4.760430443944855e-05,
      "loss": 1.2545390129089355,
      "mean_token_accuracy": 0.7225143812596798,
      "num_tokens": 21076196.0,
      "step": 10670
    },
    {
      "entropy": 1.2379481434822082,
      "epoch": 0.30591638858255876,
      "grad_norm": 10.1875,
      "learning_rate": 4.759942737349826e-05,
      "loss": 1.236661148071289,
      "mean_token_accuracy": 0.7308149643242359,
      "num_tokens": 21095646.0,
      "step": 10680
    },
    {
      "entropy": 1.2307790230959654,
      "epoch": 0.30620282714864727,
      "grad_norm": 9.1875,
      "learning_rate": 4.759454559867089e-05,
      "loss": 1.2750882148742675,
      "mean_token_accuracy": 0.7240445971488952,
      "num_tokens": 21115304.0,
      "step": 10690
    },
    {
      "entropy": 1.2248421527445317,
      "epoch": 0.30648926571473584,
      "grad_norm": 6.46875,
      "learning_rate": 4.758965911598363e-05,
      "loss": 1.2623680114746094,
      "mean_token_accuracy": 0.7138577774167061,
      "num_tokens": 21135556.0,
      "step": 10700
    },
    {
      "entropy": 1.104908700659871,
      "epoch": 0.30677570428082435,
      "grad_norm": 10.25,
      "learning_rate": 4.7584767926454634e-05,
      "loss": 1.1125308990478515,
      "mean_token_accuracy": 0.751459326967597,
      "num_tokens": 21155140.0,
      "step": 10710
    },
    {
      "entropy": 1.3171832229942084,
      "epoch": 0.3070621428469129,
      "grad_norm": 8.75,
      "learning_rate": 4.757987203110305e-05,
      "loss": 1.400719451904297,
      "mean_token_accuracy": 0.6878932595252991,
      "num_tokens": 21174682.0,
      "step": 10720
    },
    {
      "entropy": 1.3011864509433508,
      "epoch": 0.3073485814130014,
      "grad_norm": 6.84375,
      "learning_rate": 4.7574971430948996e-05,
      "loss": 1.3636573791503905,
      "mean_token_accuracy": 0.6920033153146505,
      "num_tokens": 21194177.0,
      "step": 10730
    },
    {
      "entropy": 1.1832345899194479,
      "epoch": 0.30763501997909,
      "grad_norm": 9.125,
      "learning_rate": 4.757006612701357e-05,
      "loss": 1.1613354682922363,
      "mean_token_accuracy": 0.7329035576432943,
      "num_tokens": 21215496.0,
      "step": 10740
    },
    {
      "entropy": 1.187335892394185,
      "epoch": 0.30792145854517855,
      "grad_norm": 10.3125,
      "learning_rate": 4.7565156120318865e-05,
      "loss": 1.1941236495971679,
      "mean_token_accuracy": 0.7316756926476955,
      "num_tokens": 21235595.0,
      "step": 10750
    },
    {
      "entropy": 1.2101334569975735,
      "epoch": 0.30820789711126706,
      "grad_norm": 9.3125,
      "learning_rate": 4.756024141188793e-05,
      "loss": 1.241830062866211,
      "mean_token_accuracy": 0.7255348522216082,
      "num_tokens": 21255781.0,
      "step": 10760
    },
    {
      "entropy": 1.146967126056552,
      "epoch": 0.3084943356773556,
      "grad_norm": 7.1875,
      "learning_rate": 4.7555322002744816e-05,
      "loss": 1.1321705818176269,
      "mean_token_accuracy": 0.7449606519192458,
      "num_tokens": 21275148.0,
      "step": 10770
    },
    {
      "entropy": 1.1244464922696351,
      "epoch": 0.30878077424344413,
      "grad_norm": 8.0625,
      "learning_rate": 4.755039789391455e-05,
      "loss": 1.1305888175964356,
      "mean_token_accuracy": 0.7334440983831882,
      "num_tokens": 21296414.0,
      "step": 10780
    },
    {
      "entropy": 1.1792588256299497,
      "epoch": 0.3090672128095327,
      "grad_norm": 6.34375,
      "learning_rate": 4.754546908642312e-05,
      "loss": 1.2269912719726563,
      "mean_token_accuracy": 0.7150686468929053,
      "num_tokens": 21315274.0,
      "step": 10790
    },
    {
      "entropy": 1.3087532704696059,
      "epoch": 0.3093536513756212,
      "grad_norm": 8.0625,
      "learning_rate": 4.754053558129751e-05,
      "loss": 1.3459348678588867,
      "mean_token_accuracy": 0.711145069077611,
      "num_tokens": 21334939.0,
      "step": 10800
    },
    {
      "entropy": 1.1619780700653792,
      "epoch": 0.3096400899417098,
      "grad_norm": 9.0,
      "learning_rate": 4.753559737956568e-05,
      "loss": 1.235908317565918,
      "mean_token_accuracy": 0.7183175131678581,
      "num_tokens": 21354540.0,
      "step": 10810
    },
    {
      "entropy": 1.2855201866477728,
      "epoch": 0.3099265285077983,
      "grad_norm": 8.125,
      "learning_rate": 4.753065448225655e-05,
      "loss": 1.3649744033813476,
      "mean_token_accuracy": 0.7008419420570136,
      "num_tokens": 21374023.0,
      "step": 10820
    },
    {
      "entropy": 1.2370940748602153,
      "epoch": 0.31021296707388685,
      "grad_norm": 7.0,
      "learning_rate": 4.7525706890400066e-05,
      "loss": 1.1645638465881347,
      "mean_token_accuracy": 0.7422706112265587,
      "num_tokens": 21394317.0,
      "step": 10830
    },
    {
      "entropy": 1.1378230929374695,
      "epoch": 0.31049940563997536,
      "grad_norm": 8.3125,
      "learning_rate": 4.7520754605027105e-05,
      "loss": 1.1674391746520996,
      "mean_token_accuracy": 0.7362886283546686,
      "num_tokens": 21413188.0,
      "step": 10840
    },
    {
      "entropy": 1.2479592271149158,
      "epoch": 0.3107858442060639,
      "grad_norm": 9.5,
      "learning_rate": 4.751579762716954e-05,
      "loss": 1.2581291198730469,
      "mean_token_accuracy": 0.7095990601927042,
      "num_tokens": 21432740.0,
      "step": 10850
    },
    {
      "entropy": 1.1462712477892638,
      "epoch": 0.31107228277215243,
      "grad_norm": 7.28125,
      "learning_rate": 4.751083595786022e-05,
      "loss": 1.1932430267333984,
      "mean_token_accuracy": 0.7273823566734791,
      "num_tokens": 21452016.0,
      "step": 10860
    },
    {
      "entropy": 1.1188606400042773,
      "epoch": 0.311358721338241,
      "grad_norm": 7.6875,
      "learning_rate": 4.750586959813297e-05,
      "loss": 1.1153950691223145,
      "mean_token_accuracy": 0.7426442261785269,
      "num_tokens": 21470936.0,
      "step": 10870
    },
    {
      "entropy": 1.1895586505532265,
      "epoch": 0.3116451599043295,
      "grad_norm": 8.875,
      "learning_rate": 4.7500898549022585e-05,
      "loss": 1.2804391860961915,
      "mean_token_accuracy": 0.723309475183487,
      "num_tokens": 21491062.0,
      "step": 10880
    },
    {
      "entropy": 1.1611870124936103,
      "epoch": 0.31193159847041807,
      "grad_norm": 9.3125,
      "learning_rate": 4.749592281156487e-05,
      "loss": 1.1890996932983398,
      "mean_token_accuracy": 0.7258432779461146,
      "num_tokens": 21509588.0,
      "step": 10890
    },
    {
      "entropy": 1.1895610868930817,
      "epoch": 0.3122180370365066,
      "grad_norm": 7.4375,
      "learning_rate": 4.749094238679656e-05,
      "loss": 1.249438762664795,
      "mean_token_accuracy": 0.7212285254150629,
      "num_tokens": 21529465.0,
      "step": 10900
    },
    {
      "entropy": 1.1919631615281105,
      "epoch": 0.31250447560259514,
      "grad_norm": 8.1875,
      "learning_rate": 4.748595727575541e-05,
      "loss": 1.1756156921386718,
      "mean_token_accuracy": 0.7407011017203331,
      "num_tokens": 21549110.0,
      "step": 10910
    },
    {
      "entropy": 1.1960717402398586,
      "epoch": 0.31279091416868365,
      "grad_norm": 8.6875,
      "learning_rate": 4.748096747948011e-05,
      "loss": 1.1394862174987792,
      "mean_token_accuracy": 0.7422023169696331,
      "num_tokens": 21569379.0,
      "step": 10920
    },
    {
      "entropy": 1.0542187761515378,
      "epoch": 0.3130773527347722,
      "grad_norm": 6.9375,
      "learning_rate": 4.7475972999010364e-05,
      "loss": 1.1041646957397462,
      "mean_token_accuracy": 0.7483351707458497,
      "num_tokens": 21588808.0,
      "step": 10930
    },
    {
      "entropy": 1.0867573030292987,
      "epoch": 0.31336379130086073,
      "grad_norm": 7.28125,
      "learning_rate": 4.747097383538682e-05,
      "loss": 1.1180338859558105,
      "mean_token_accuracy": 0.7418781992048025,
      "num_tokens": 21608399.0,
      "step": 10940
    },
    {
      "entropy": 1.23358127232641,
      "epoch": 0.3136502298669493,
      "grad_norm": 8.3125,
      "learning_rate": 4.7465969989651135e-05,
      "loss": 1.3244894027709961,
      "mean_token_accuracy": 0.7115334682166576,
      "num_tokens": 21627233.0,
      "step": 10950
    },
    {
      "entropy": 1.2483941223472357,
      "epoch": 0.3139366684330378,
      "grad_norm": 7.84375,
      "learning_rate": 4.746096146284591e-05,
      "loss": 1.3419307708740233,
      "mean_token_accuracy": 0.7024520382285118,
      "num_tokens": 21647844.0,
      "step": 10960
    },
    {
      "entropy": 1.1981158410198987,
      "epoch": 0.31422310699912637,
      "grad_norm": 6.65625,
      "learning_rate": 4.745594825601474e-05,
      "loss": 1.2303057670593263,
      "mean_token_accuracy": 0.7272470030933619,
      "num_tokens": 21667743.0,
      "step": 10970
    },
    {
      "entropy": 1.2136971287429332,
      "epoch": 0.3145095455652149,
      "grad_norm": 7.5,
      "learning_rate": 4.7450930370202186e-05,
      "loss": 1.2649925231933594,
      "mean_token_accuracy": 0.7197300598025322,
      "num_tokens": 21687535.0,
      "step": 10980
    },
    {
      "entropy": 1.3185116354376079,
      "epoch": 0.31479598413130344,
      "grad_norm": 8.5625,
      "learning_rate": 4.744590780645379e-05,
      "loss": 1.2711140632629394,
      "mean_token_accuracy": 0.7174013394862413,
      "num_tokens": 21706072.0,
      "step": 10990
    },
    {
      "entropy": 1.1728610198944807,
      "epoch": 0.31508242269739195,
      "grad_norm": 9.5625,
      "learning_rate": 4.7440880565816085e-05,
      "loss": 1.2049323081970216,
      "mean_token_accuracy": 0.7259841859340668,
      "num_tokens": 21724910.0,
      "step": 11000
    },
    {
      "entropy": 1.1665601275861264,
      "epoch": 0.3153688612634805,
      "grad_norm": 7.84375,
      "learning_rate": 4.743584864933652e-05,
      "loss": 1.2178359031677246,
      "mean_token_accuracy": 0.735757676139474,
      "num_tokens": 21743481.0,
      "step": 11010
    },
    {
      "entropy": 1.2199096255004407,
      "epoch": 0.3156552998295691,
      "grad_norm": 8.125,
      "learning_rate": 4.74308120580636e-05,
      "loss": 1.3107985496520995,
      "mean_token_accuracy": 0.7149065248668194,
      "num_tokens": 21763759.0,
      "step": 11020
    },
    {
      "entropy": 1.205187126994133,
      "epoch": 0.3159417383956576,
      "grad_norm": 9.125,
      "learning_rate": 4.7425770793046734e-05,
      "loss": 1.2412653923034669,
      "mean_token_accuracy": 0.7323855470865965,
      "num_tokens": 21783707.0,
      "step": 11030
    },
    {
      "entropy": 1.1945237398147583,
      "epoch": 0.31622817696174615,
      "grad_norm": 8.25,
      "learning_rate": 4.7420724855336354e-05,
      "loss": 1.1271297454833984,
      "mean_token_accuracy": 0.744668971374631,
      "num_tokens": 21803623.0,
      "step": 11040
    },
    {
      "entropy": 1.118756825476885,
      "epoch": 0.31651461552783466,
      "grad_norm": 6.90625,
      "learning_rate": 4.7415674245983826e-05,
      "loss": 1.1883705139160157,
      "mean_token_accuracy": 0.7386421013623476,
      "num_tokens": 21822695.0,
      "step": 11050
    },
    {
      "entropy": 1.1413742423057556,
      "epoch": 0.31680105409392323,
      "grad_norm": 6.875,
      "learning_rate": 4.741061896604152e-05,
      "loss": 1.0950676918029785,
      "mean_token_accuracy": 0.745844928920269,
      "num_tokens": 21843807.0,
      "step": 11060
    },
    {
      "entropy": 1.048705903813243,
      "epoch": 0.31708749266001174,
      "grad_norm": 6.1875,
      "learning_rate": 4.7405559016562774e-05,
      "loss": 1.0317532539367675,
      "mean_token_accuracy": 0.7623295079916715,
      "num_tokens": 21863669.0,
      "step": 11070
    },
    {
      "entropy": 1.1278911229223012,
      "epoch": 0.3173739312261003,
      "grad_norm": 9.3125,
      "learning_rate": 4.740049439860187e-05,
      "loss": 1.135157871246338,
      "mean_token_accuracy": 0.7335341315716505,
      "num_tokens": 21884077.0,
      "step": 11080
    },
    {
      "entropy": 1.0938856963068246,
      "epoch": 0.3176603697921888,
      "grad_norm": 8.0,
      "learning_rate": 4.739542511321411e-05,
      "loss": 1.146687889099121,
      "mean_token_accuracy": 0.7279066048562527,
      "num_tokens": 21903088.0,
      "step": 11090
    },
    {
      "entropy": 1.0932465061545371,
      "epoch": 0.3179468083582774,
      "grad_norm": 8.5625,
      "learning_rate": 4.7390351161455734e-05,
      "loss": 1.1544628143310547,
      "mean_token_accuracy": 0.7280469022691249,
      "num_tokens": 21921842.0,
      "step": 11100
    },
    {
      "entropy": 1.1031063478440046,
      "epoch": 0.3182332469243659,
      "grad_norm": 7.125,
      "learning_rate": 4.738527254438396e-05,
      "loss": 1.1028409004211426,
      "mean_token_accuracy": 0.7358500190079212,
      "num_tokens": 21941431.0,
      "step": 11110
    },
    {
      "entropy": 1.286591763421893,
      "epoch": 0.31851968549045445,
      "grad_norm": 7.8125,
      "learning_rate": 4.7380189263056986e-05,
      "loss": 1.3475188255310058,
      "mean_token_accuracy": 0.6986013013869524,
      "num_tokens": 21962291.0,
      "step": 11120
    },
    {
      "entropy": 1.13895827755332,
      "epoch": 0.31880612405654296,
      "grad_norm": 7.6875,
      "learning_rate": 4.737510131853398e-05,
      "loss": 1.1346517562866212,
      "mean_token_accuracy": 0.7307674538344144,
      "num_tokens": 21982771.0,
      "step": 11130
    },
    {
      "entropy": 1.1660083957016467,
      "epoch": 0.3190925626226315,
      "grad_norm": 8.9375,
      "learning_rate": 4.737000871187507e-05,
      "loss": 1.2231541633605958,
      "mean_token_accuracy": 0.7201999597251415,
      "num_tokens": 22001315.0,
      "step": 11140
    },
    {
      "entropy": 1.1360034223645925,
      "epoch": 0.31937900118872004,
      "grad_norm": 6.9375,
      "learning_rate": 4.7364911444141375e-05,
      "loss": 1.1206212997436524,
      "mean_token_accuracy": 0.7419365961104631,
      "num_tokens": 22021319.0,
      "step": 11150
    },
    {
      "entropy": 1.2473677471280098,
      "epoch": 0.3196654397548086,
      "grad_norm": 7.21875,
      "learning_rate": 4.735980951639497e-05,
      "loss": 1.3122434616088867,
      "mean_token_accuracy": 0.7226801976561547,
      "num_tokens": 22040644.0,
      "step": 11160
    },
    {
      "entropy": 1.133636212348938,
      "epoch": 0.3199518783208971,
      "grad_norm": 6.78125,
      "learning_rate": 4.735470292969892e-05,
      "loss": 1.1689264297485351,
      "mean_token_accuracy": 0.7301033060997725,
      "num_tokens": 22060516.0,
      "step": 11170
    },
    {
      "entropy": 1.1562007825821639,
      "epoch": 0.3202383168869857,
      "grad_norm": 10.625,
      "learning_rate": 4.7349591685117226e-05,
      "loss": 1.1458330154418945,
      "mean_token_accuracy": 0.742704413831234,
      "num_tokens": 22079717.0,
      "step": 11180
    },
    {
      "entropy": 1.1761312440037728,
      "epoch": 0.3205247554530742,
      "grad_norm": 6.75,
      "learning_rate": 4.7344475783714895e-05,
      "loss": 1.2600318908691406,
      "mean_token_accuracy": 0.7188513174653053,
      "num_tokens": 22100974.0,
      "step": 11190
    },
    {
      "entropy": 1.2415122747421266,
      "epoch": 0.32081119401916275,
      "grad_norm": 11.0,
      "learning_rate": 4.733935522655789e-05,
      "loss": 1.257845401763916,
      "mean_token_accuracy": 0.7230556247755885,
      "num_tokens": 22119661.0,
      "step": 11200
    },
    {
      "entropy": 1.1471884217113257,
      "epoch": 0.32109763258525126,
      "grad_norm": 6.1875,
      "learning_rate": 4.733423001471313e-05,
      "loss": 1.147789192199707,
      "mean_token_accuracy": 0.7372707635164261,
      "num_tokens": 22138558.0,
      "step": 11210
    },
    {
      "entropy": 1.2086162056773901,
      "epoch": 0.3213840711513398,
      "grad_norm": 7.96875,
      "learning_rate": 4.7329100149248534e-05,
      "loss": 1.2477478981018066,
      "mean_token_accuracy": 0.7153390884399414,
      "num_tokens": 22158525.0,
      "step": 11220
    },
    {
      "entropy": 1.2628332007676364,
      "epoch": 0.32167050971742833,
      "grad_norm": 10.9375,
      "learning_rate": 4.732396563123297e-05,
      "loss": 1.2975424766540526,
      "mean_token_accuracy": 0.7086935788393021,
      "num_tokens": 22178480.0,
      "step": 11230
    },
    {
      "entropy": 1.2490848362445832,
      "epoch": 0.3219569482835169,
      "grad_norm": 6.46875,
      "learning_rate": 4.731882646173627e-05,
      "loss": 1.3228703498840333,
      "mean_token_accuracy": 0.7056458298116922,
      "num_tokens": 22199545.0,
      "step": 11240
    },
    {
      "entropy": 1.228704982623458,
      "epoch": 0.3222433868496054,
      "grad_norm": 8.0,
      "learning_rate": 4.731368264182926e-05,
      "loss": 1.1814695358276368,
      "mean_token_accuracy": 0.7292388524860144,
      "num_tokens": 22218935.0,
      "step": 11250
    },
    {
      "entropy": 1.1486282328143715,
      "epoch": 0.32252982541569397,
      "grad_norm": 8.25,
      "learning_rate": 4.7308534172583704e-05,
      "loss": 1.2046530723571778,
      "mean_token_accuracy": 0.7276582639664412,
      "num_tokens": 22239019.0,
      "step": 11260
    },
    {
      "entropy": 1.1681117948144675,
      "epoch": 0.3228162639817825,
      "grad_norm": 9.1875,
      "learning_rate": 4.7303381055072374e-05,
      "loss": 1.189686107635498,
      "mean_token_accuracy": 0.7220390990376473,
      "num_tokens": 22257389.0,
      "step": 11270
    },
    {
      "entropy": 1.1321909628808497,
      "epoch": 0.32310270254787105,
      "grad_norm": 8.75,
      "learning_rate": 4.729822329036896e-05,
      "loss": 1.179995346069336,
      "mean_token_accuracy": 0.7421342566609382,
      "num_tokens": 22277656.0,
      "step": 11280
    },
    {
      "entropy": 1.1787487041205167,
      "epoch": 0.3233891411139596,
      "grad_norm": 6.5625,
      "learning_rate": 4.729306087954817e-05,
      "loss": 1.192238998413086,
      "mean_token_accuracy": 0.7372886165976524,
      "num_tokens": 22296737.0,
      "step": 11290
    },
    {
      "entropy": 1.200702952221036,
      "epoch": 0.3236755796800481,
      "grad_norm": 9.25,
      "learning_rate": 4.728789382368564e-05,
      "loss": 1.206761074066162,
      "mean_token_accuracy": 0.7243234682828188,
      "num_tokens": 22315072.0,
      "step": 11300
    },
    {
      "entropy": 1.150430616736412,
      "epoch": 0.3239620182461367,
      "grad_norm": 6.4375,
      "learning_rate": 4.7282722123858e-05,
      "loss": 1.1563529968261719,
      "mean_token_accuracy": 0.7395915940403939,
      "num_tokens": 22335739.0,
      "step": 11310
    },
    {
      "entropy": 1.1855383176356553,
      "epoch": 0.3242484568122252,
      "grad_norm": 6.78125,
      "learning_rate": 4.727754578114285e-05,
      "loss": 1.2027335166931152,
      "mean_token_accuracy": 0.7233661249279976,
      "num_tokens": 22355845.0,
      "step": 11320
    },
    {
      "entropy": 1.1097164317965507,
      "epoch": 0.32453489537831376,
      "grad_norm": 6.125,
      "learning_rate": 4.727236479661873e-05,
      "loss": 1.1780168533325195,
      "mean_token_accuracy": 0.7526135090738535,
      "num_tokens": 22374465.0,
      "step": 11330
    },
    {
      "entropy": 1.247841464728117,
      "epoch": 0.32482133394440227,
      "grad_norm": 7.03125,
      "learning_rate": 4.726717917136517e-05,
      "loss": 1.2450525283813476,
      "mean_token_accuracy": 0.7134382147341967,
      "num_tokens": 22394949.0,
      "step": 11340
    },
    {
      "entropy": 1.2632878109812737,
      "epoch": 0.32510777251049083,
      "grad_norm": 7.71875,
      "learning_rate": 4.726198890646266e-05,
      "loss": 1.2831339836120605,
      "mean_token_accuracy": 0.7021247051656246,
      "num_tokens": 22415600.0,
      "step": 11350
    },
    {
      "entropy": 1.2132082648575306,
      "epoch": 0.32539421107657934,
      "grad_norm": 8.125,
      "learning_rate": 4.725679400299266e-05,
      "loss": 1.2478737831115723,
      "mean_token_accuracy": 0.7308540228754282,
      "num_tokens": 22435548.0,
      "step": 11360
    },
    {
      "entropy": 1.264779619500041,
      "epoch": 0.3256806496426679,
      "grad_norm": 11.5625,
      "learning_rate": 4.7251594462037584e-05,
      "loss": 1.3077947616577148,
      "mean_token_accuracy": 0.7187631480395794,
      "num_tokens": 22456907.0,
      "step": 11370
    },
    {
      "entropy": 1.156732125580311,
      "epoch": 0.3259670882087564,
      "grad_norm": 14.125,
      "learning_rate": 4.724639028468083e-05,
      "loss": 1.1519679069519042,
      "mean_token_accuracy": 0.7447134781628847,
      "num_tokens": 22476256.0,
      "step": 11380
    },
    {
      "entropy": 1.214553551003337,
      "epoch": 0.326253526774845,
      "grad_norm": 9.8125,
      "learning_rate": 4.724118147200677e-05,
      "loss": 1.241359806060791,
      "mean_token_accuracy": 0.7144936062395573,
      "num_tokens": 22496723.0,
      "step": 11390
    },
    {
      "entropy": 1.1805000428110362,
      "epoch": 0.3265399653409335,
      "grad_norm": 8.25,
      "learning_rate": 4.7235968025100695e-05,
      "loss": 1.1674490928649903,
      "mean_token_accuracy": 0.7388947810977697,
      "num_tokens": 22516967.0,
      "step": 11400
    },
    {
      "entropy": 1.072376536205411,
      "epoch": 0.32682640390702206,
      "grad_norm": 7.5,
      "learning_rate": 4.723074994504891e-05,
      "loss": 1.1109947204589843,
      "mean_token_accuracy": 0.7554388120770454,
      "num_tokens": 22535756.0,
      "step": 11410
    },
    {
      "entropy": 1.1708686981350183,
      "epoch": 0.32711284247311057,
      "grad_norm": 7.1875,
      "learning_rate": 4.722552723293867e-05,
      "loss": 1.2544927597045898,
      "mean_token_accuracy": 0.7224156752228736,
      "num_tokens": 22555473.0,
      "step": 11420
    },
    {
      "entropy": 1.2066254845820368,
      "epoch": 0.32739928103919913,
      "grad_norm": 6.34375,
      "learning_rate": 4.722029988985819e-05,
      "loss": 1.2252647399902343,
      "mean_token_accuracy": 0.7247946180403233,
      "num_tokens": 22576908.0,
      "step": 11430
    },
    {
      "entropy": 1.293650460243225,
      "epoch": 0.32768571960528764,
      "grad_norm": 8.3125,
      "learning_rate": 4.721506791689665e-05,
      "loss": 1.33160343170166,
      "mean_token_accuracy": 0.705178790539503,
      "num_tokens": 22597456.0,
      "step": 11440
    },
    {
      "entropy": 1.3517684258520604,
      "epoch": 0.3279721581713762,
      "grad_norm": 8.1875,
      "learning_rate": 4.7209831315144207e-05,
      "loss": 1.400987434387207,
      "mean_token_accuracy": 0.7010006219148636,
      "num_tokens": 22617341.0,
      "step": 11450
    },
    {
      "entropy": 1.2323766369372606,
      "epoch": 0.3282585967374647,
      "grad_norm": 7.46875,
      "learning_rate": 4.7204590085691956e-05,
      "loss": 1.263054656982422,
      "mean_token_accuracy": 0.7114313792437315,
      "num_tokens": 22637087.0,
      "step": 11460
    },
    {
      "entropy": 1.1799864664673805,
      "epoch": 0.3285450353035533,
      "grad_norm": 7.5,
      "learning_rate": 4.7199344229631995e-05,
      "loss": 1.145082378387451,
      "mean_token_accuracy": 0.7348103608936072,
      "num_tokens": 22657351.0,
      "step": 11470
    },
    {
      "entropy": 1.1265316696837544,
      "epoch": 0.3288314738696418,
      "grad_norm": 5.34375,
      "learning_rate": 4.719409374805735e-05,
      "loss": 1.1641122817993164,
      "mean_token_accuracy": 0.7305260673165321,
      "num_tokens": 22677412.0,
      "step": 11480
    },
    {
      "entropy": 1.124467265419662,
      "epoch": 0.32911791243573035,
      "grad_norm": 6.0625,
      "learning_rate": 4.718883864206203e-05,
      "loss": 1.159708595275879,
      "mean_token_accuracy": 0.7350906219333411,
      "num_tokens": 22697220.0,
      "step": 11490
    },
    {
      "entropy": 1.0693022284656764,
      "epoch": 0.32940435100181886,
      "grad_norm": 7.75,
      "learning_rate": 4.7183578912741e-05,
      "loss": 1.1036873817443849,
      "mean_token_accuracy": 0.7535626616328954,
      "num_tokens": 22717261.0,
      "step": 11500
    },
    {
      "entropy": 1.192902198061347,
      "epoch": 0.3296907895679074,
      "grad_norm": 10.375,
      "learning_rate": 4.7178314561190194e-05,
      "loss": 1.2284892082214356,
      "mean_token_accuracy": 0.7234285466372967,
      "num_tokens": 22735967.0,
      "step": 11510
    },
    {
      "entropy": 1.1275312550365926,
      "epoch": 0.32997722813399594,
      "grad_norm": 7.78125,
      "learning_rate": 4.7173045588506505e-05,
      "loss": 1.129185676574707,
      "mean_token_accuracy": 0.7411923322826623,
      "num_tokens": 22755874.0,
      "step": 11520
    },
    {
      "entropy": 1.1632478546351195,
      "epoch": 0.3302636667000845,
      "grad_norm": 6.0625,
      "learning_rate": 4.7167771995787796e-05,
      "loss": 1.1407917022705079,
      "mean_token_accuracy": 0.7309483349323272,
      "num_tokens": 22775091.0,
      "step": 11530
    },
    {
      "entropy": 1.1731928531080484,
      "epoch": 0.330550105266173,
      "grad_norm": 6.625,
      "learning_rate": 4.716249378413288e-05,
      "loss": 1.1625186920166015,
      "mean_token_accuracy": 0.7388781841844321,
      "num_tokens": 22794778.0,
      "step": 11540
    },
    {
      "entropy": 1.2525504898279904,
      "epoch": 0.3308365438322616,
      "grad_norm": 8.375,
      "learning_rate": 4.715721095464154e-05,
      "loss": 1.323142147064209,
      "mean_token_accuracy": 0.7031543012708426,
      "num_tokens": 22814128.0,
      "step": 11550
    },
    {
      "entropy": 1.2086209027096628,
      "epoch": 0.33112298239835014,
      "grad_norm": 9.4375,
      "learning_rate": 4.715192350841452e-05,
      "loss": 1.3088679313659668,
      "mean_token_accuracy": 0.7132860606536269,
      "num_tokens": 22834576.0,
      "step": 11560
    },
    {
      "entropy": 1.2420202609151603,
      "epoch": 0.33140942096443865,
      "grad_norm": 7.34375,
      "learning_rate": 4.714663144655353e-05,
      "loss": 1.3523476600646973,
      "mean_token_accuracy": 0.7066051442176103,
      "num_tokens": 22854788.0,
      "step": 11570
    },
    {
      "entropy": 1.2554360300302505,
      "epoch": 0.3316958595305272,
      "grad_norm": 9.0,
      "learning_rate": 4.714133477016124e-05,
      "loss": 1.2440761566162108,
      "mean_token_accuracy": 0.7174391876906157,
      "num_tokens": 22872750.0,
      "step": 11580
    },
    {
      "entropy": 1.234518339857459,
      "epoch": 0.3319822980966157,
      "grad_norm": 7.375,
      "learning_rate": 4.7136033480341265e-05,
      "loss": 1.1752326965332032,
      "mean_token_accuracy": 0.7297517832368612,
      "num_tokens": 22893007.0,
      "step": 11590
    },
    {
      "entropy": 1.2630619576200843,
      "epoch": 0.3322687366627043,
      "grad_norm": 8.9375,
      "learning_rate": 4.713072757819822e-05,
      "loss": 1.2878996849060058,
      "mean_token_accuracy": 0.7143452558666468,
      "num_tokens": 22912446.0,
      "step": 11600
    },
    {
      "entropy": 1.1564901519566775,
      "epoch": 0.3325551752287928,
      "grad_norm": 7.59375,
      "learning_rate": 4.712541706483764e-05,
      "loss": 1.1643918037414551,
      "mean_token_accuracy": 0.7464081030339003,
      "num_tokens": 22933465.0,
      "step": 11610
    },
    {
      "entropy": 1.0927067503333092,
      "epoch": 0.33284161379488136,
      "grad_norm": 7.84375,
      "learning_rate": 4.7120101941366043e-05,
      "loss": 1.083414649963379,
      "mean_token_accuracy": 0.7488886006176472,
      "num_tokens": 22952832.0,
      "step": 11620
    },
    {
      "entropy": 1.214715925976634,
      "epoch": 0.3331280523609699,
      "grad_norm": 8.4375,
      "learning_rate": 4.71147822088909e-05,
      "loss": 1.3036578178405762,
      "mean_token_accuracy": 0.7219994474202395,
      "num_tokens": 22973286.0,
      "step": 11630
    },
    {
      "entropy": 1.0849368203431369,
      "epoch": 0.33341449092705844,
      "grad_norm": 6.90625,
      "learning_rate": 4.710945786852064e-05,
      "loss": 1.062319564819336,
      "mean_token_accuracy": 0.7487390957772732,
      "num_tokens": 22992508.0,
      "step": 11640
    },
    {
      "entropy": 1.065082783997059,
      "epoch": 0.33370092949314695,
      "grad_norm": 7.625,
      "learning_rate": 4.710412892136467e-05,
      "loss": 1.1180828094482422,
      "mean_token_accuracy": 0.7427695158869028,
      "num_tokens": 23011084.0,
      "step": 11650
    },
    {
      "entropy": 1.1031024660915136,
      "epoch": 0.3339873680592355,
      "grad_norm": 9.1875,
      "learning_rate": 4.709879536853333e-05,
      "loss": 1.108950710296631,
      "mean_token_accuracy": 0.7525781478732825,
      "num_tokens": 23032469.0,
      "step": 11660
    },
    {
      "entropy": 1.1460798531770706,
      "epoch": 0.334273806625324,
      "grad_norm": 8.875,
      "learning_rate": 4.709345721113794e-05,
      "loss": 1.1499256134033202,
      "mean_token_accuracy": 0.7332233686000109,
      "num_tokens": 23051665.0,
      "step": 11670
    },
    {
      "entropy": 1.1252974119037389,
      "epoch": 0.3345602451914126,
      "grad_norm": 6.6875,
      "learning_rate": 4.708811445029078e-05,
      "loss": 1.1918341636657714,
      "mean_token_accuracy": 0.7303916115313769,
      "num_tokens": 23069938.0,
      "step": 11680
    },
    {
      "entropy": 1.274068509787321,
      "epoch": 0.3348466837575011,
      "grad_norm": 11.4375,
      "learning_rate": 4.708276708710506e-05,
      "loss": 1.3522369384765625,
      "mean_token_accuracy": 0.7006101924926043,
      "num_tokens": 23088053.0,
      "step": 11690
    },
    {
      "entropy": 1.2095244040712714,
      "epoch": 0.33513312232358966,
      "grad_norm": 7.375,
      "learning_rate": 4.7077415122695e-05,
      "loss": 1.2322667121887207,
      "mean_token_accuracy": 0.7209963157773018,
      "num_tokens": 23106406.0,
      "step": 11700
    },
    {
      "entropy": 1.1527428291738033,
      "epoch": 0.33541956088967817,
      "grad_norm": 8.3125,
      "learning_rate": 4.707205855817572e-05,
      "loss": 1.129507064819336,
      "mean_token_accuracy": 0.7344038028270006,
      "num_tokens": 23125319.0,
      "step": 11710
    },
    {
      "entropy": 1.1516671530902385,
      "epoch": 0.33570599945576673,
      "grad_norm": 6.96875,
      "learning_rate": 4.706669739466335e-05,
      "loss": 1.2391933441162108,
      "mean_token_accuracy": 0.7254343092441559,
      "num_tokens": 23145410.0,
      "step": 11720
    },
    {
      "entropy": 1.261005749925971,
      "epoch": 0.33599243802185524,
      "grad_norm": 8.625,
      "learning_rate": 4.7061331633274955e-05,
      "loss": 1.3115338325500487,
      "mean_token_accuracy": 0.7017195284366607,
      "num_tokens": 23164450.0,
      "step": 11730
    },
    {
      "entropy": 1.1775232704356313,
      "epoch": 0.3362788765879438,
      "grad_norm": 6.90625,
      "learning_rate": 4.7055961275128544e-05,
      "loss": 1.1274161338806152,
      "mean_token_accuracy": 0.7441752128303051,
      "num_tokens": 23186475.0,
      "step": 11740
    },
    {
      "entropy": 1.1429383732378482,
      "epoch": 0.3365653151540323,
      "grad_norm": 7.03125,
      "learning_rate": 4.705058632134311e-05,
      "loss": 1.2184623718261718,
      "mean_token_accuracy": 0.7309894770383835,
      "num_tokens": 23207066.0,
      "step": 11750
    },
    {
      "entropy": 1.2083202466368674,
      "epoch": 0.3368517537201209,
      "grad_norm": 10.25,
      "learning_rate": 4.70452067730386e-05,
      "loss": 1.260502243041992,
      "mean_token_accuracy": 0.7165480781346559,
      "num_tokens": 23226649.0,
      "step": 11760
    },
    {
      "entropy": 1.2827447853982448,
      "epoch": 0.3371381922862094,
      "grad_norm": 7.96875,
      "learning_rate": 4.7039822631335895e-05,
      "loss": 1.301980209350586,
      "mean_token_accuracy": 0.7162534985691309,
      "num_tokens": 23245872.0,
      "step": 11770
    },
    {
      "entropy": 1.1434971237555147,
      "epoch": 0.33742463085229796,
      "grad_norm": 6.78125,
      "learning_rate": 4.7034433897356864e-05,
      "loss": 1.178339958190918,
      "mean_token_accuracy": 0.7308500051498413,
      "num_tokens": 23265253.0,
      "step": 11780
    },
    {
      "entropy": 1.2729852348566055,
      "epoch": 0.33771106941838647,
      "grad_norm": 7.53125,
      "learning_rate": 4.7029040572224305e-05,
      "loss": 1.2637948989868164,
      "mean_token_accuracy": 0.7183017022907734,
      "num_tokens": 23286647.0,
      "step": 11790
    },
    {
      "entropy": 1.1382151689380406,
      "epoch": 0.33799750798447503,
      "grad_norm": 7.65625,
      "learning_rate": 4.702364265706199e-05,
      "loss": 1.1338038444519043,
      "mean_token_accuracy": 0.7474264606833458,
      "num_tokens": 23305792.0,
      "step": 11800
    },
    {
      "entropy": 1.2622137024998665,
      "epoch": 0.33828394655056354,
      "grad_norm": 8.375,
      "learning_rate": 4.701824015299464e-05,
      "loss": 1.2841017723083497,
      "mean_token_accuracy": 0.722211503982544,
      "num_tokens": 23325916.0,
      "step": 11810
    },
    {
      "entropy": 1.2117103721946478,
      "epoch": 0.3385703851166521,
      "grad_norm": 7.5625,
      "learning_rate": 4.701283306114795e-05,
      "loss": 1.2188192367553712,
      "mean_token_accuracy": 0.7275662157684565,
      "num_tokens": 23345447.0,
      "step": 11820
    },
    {
      "entropy": 1.1250001799315215,
      "epoch": 0.33885682368274067,
      "grad_norm": 8.625,
      "learning_rate": 4.7007421382648534e-05,
      "loss": 1.144778347015381,
      "mean_token_accuracy": 0.7350890174508095,
      "num_tokens": 23365352.0,
      "step": 11830
    },
    {
      "entropy": 1.0864742234349252,
      "epoch": 0.3391432622488292,
      "grad_norm": 6.9375,
      "learning_rate": 4.7002005118624005e-05,
      "loss": 1.1144563674926757,
      "mean_token_accuracy": 0.7423195265233516,
      "num_tokens": 23385674.0,
      "step": 11840
    },
    {
      "entropy": 1.0908971633762121,
      "epoch": 0.33942970081491775,
      "grad_norm": 7.46875,
      "learning_rate": 4.699658427020289e-05,
      "loss": 1.0974401473999023,
      "mean_token_accuracy": 0.7543679516762495,
      "num_tokens": 23404581.0,
      "step": 11850
    },
    {
      "entropy": 1.1848457358777522,
      "epoch": 0.33971613938100625,
      "grad_norm": 7.9375,
      "learning_rate": 4.699115883851471e-05,
      "loss": 1.2022927284240723,
      "mean_token_accuracy": 0.7327800579369068,
      "num_tokens": 23424920.0,
      "step": 11860
    },
    {
      "entropy": 1.161220794916153,
      "epoch": 0.3400025779470948,
      "grad_norm": 9.375,
      "learning_rate": 4.6985728824689915e-05,
      "loss": 1.2107144355773927,
      "mean_token_accuracy": 0.7207223646342754,
      "num_tokens": 23444101.0,
      "step": 11870
    },
    {
      "entropy": 1.1067387491464615,
      "epoch": 0.34028901651318333,
      "grad_norm": 5.96875,
      "learning_rate": 4.698029422985991e-05,
      "loss": 1.1030991554260254,
      "mean_token_accuracy": 0.7373673833906651,
      "num_tokens": 23463963.0,
      "step": 11880
    },
    {
      "entropy": 1.1548112120479346,
      "epoch": 0.3405754550792719,
      "grad_norm": 7.96875,
      "learning_rate": 4.697485505515707e-05,
      "loss": 1.2031975746154786,
      "mean_token_accuracy": 0.7296447124332189,
      "num_tokens": 23483210.0,
      "step": 11890
    },
    {
      "entropy": 1.12001075707376,
      "epoch": 0.3408618936453604,
      "grad_norm": 10.125,
      "learning_rate": 4.696941130171473e-05,
      "loss": 1.1540513992309571,
      "mean_token_accuracy": 0.7319434903562069,
      "num_tokens": 23502413.0,
      "step": 11900
    },
    {
      "entropy": 1.1788815069943666,
      "epoch": 0.34114833221144897,
      "grad_norm": 7.625,
      "learning_rate": 4.696396297066713e-05,
      "loss": 1.2136086463928222,
      "mean_token_accuracy": 0.7291082628071308,
      "num_tokens": 23522531.0,
      "step": 11910
    },
    {
      "entropy": 1.260196452215314,
      "epoch": 0.3414347707775375,
      "grad_norm": 7.375,
      "learning_rate": 4.695851006314953e-05,
      "loss": 1.2704562187194823,
      "mean_token_accuracy": 0.71053973659873,
      "num_tokens": 23540992.0,
      "step": 11920
    },
    {
      "entropy": 1.2255538059398532,
      "epoch": 0.34172120934362604,
      "grad_norm": 9.0,
      "learning_rate": 4.6953052580298096e-05,
      "loss": 1.2748222351074219,
      "mean_token_accuracy": 0.7164468251168727,
      "num_tokens": 23561046.0,
      "step": 11930
    },
    {
      "entropy": 1.1561801934614777,
      "epoch": 0.34200764790971455,
      "grad_norm": 8.8125,
      "learning_rate": 4.6947590523249965e-05,
      "loss": 1.1780438423156738,
      "mean_token_accuracy": 0.7307541415095329,
      "num_tokens": 23581109.0,
      "step": 11940
    },
    {
      "entropy": 1.2238935600966214,
      "epoch": 0.3422940864758031,
      "grad_norm": 9.875,
      "learning_rate": 4.694212389314323e-05,
      "loss": 1.2320446968078613,
      "mean_token_accuracy": 0.7204213123768568,
      "num_tokens": 23602952.0,
      "step": 11950
    },
    {
      "entropy": 1.1128317618742585,
      "epoch": 0.3425805250418916,
      "grad_norm": 6.375,
      "learning_rate": 4.693665269111693e-05,
      "loss": 1.1011093139648438,
      "mean_token_accuracy": 0.7491160191595554,
      "num_tokens": 23622836.0,
      "step": 11960
    },
    {
      "entropy": 1.1253233417868613,
      "epoch": 0.3428669636079802,
      "grad_norm": 9.5625,
      "learning_rate": 4.6931176918311064e-05,
      "loss": 1.1716981887817384,
      "mean_token_accuracy": 0.7309842467308044,
      "num_tokens": 23642229.0,
      "step": 11970
    },
    {
      "entropy": 1.2138034660369157,
      "epoch": 0.3431534021740687,
      "grad_norm": 7.59375,
      "learning_rate": 4.692569657586657e-05,
      "loss": 1.2460014343261718,
      "mean_token_accuracy": 0.7129536826163531,
      "num_tokens": 23660734.0,
      "step": 11980
    },
    {
      "entropy": 1.1330853898078204,
      "epoch": 0.34343984074015727,
      "grad_norm": 7.5,
      "learning_rate": 4.6920211664925365e-05,
      "loss": 1.1663831710815429,
      "mean_token_accuracy": 0.7308716084808111,
      "num_tokens": 23681258.0,
      "step": 11990
    },
    {
      "epoch": 0.3437262793062458,
      "eval_entropy": 1.2111419036388398,
      "eval_loss": 1.2187772989273071,
      "eval_mean_token_accuracy": 0.7236575634479523,
      "eval_num_tokens": 23701035.0,
      "eval_runtime": 43.1323,
      "eval_samples_per_second": 46.369,
      "eval_steps_per_second": 5.796,
      "step": 12000
    },
    {
      "entropy": 1.1612286842428148,
      "epoch": 0.34401271787233434,
      "grad_norm": 8.5,
      "learning_rate": 4.690922814212512e-05,
      "loss": 1.163602352142334,
      "mean_token_accuracy": 0.7336456025019288,
      "num_tokens": 23720511.0,
      "step": 12010
    },
    {
      "entropy": 1.1802667088806629,
      "epoch": 0.34429915643842285,
      "grad_norm": 7.75,
      "learning_rate": 4.6903729532554654e-05,
      "loss": 1.1653948783874513,
      "mean_token_accuracy": 0.7312414187937974,
      "num_tokens": 23739738.0,
      "step": 12020
    },
    {
      "entropy": 1.1538407292217017,
      "epoch": 0.3445855950045114,
      "grad_norm": 7.53125,
      "learning_rate": 4.6898226359064567e-05,
      "loss": 1.2091781616210937,
      "mean_token_accuracy": 0.7253286480903626,
      "num_tokens": 23759424.0,
      "step": 12030
    },
    {
      "entropy": 1.0579552758485078,
      "epoch": 0.3448720335705999,
      "grad_norm": 8.9375,
      "learning_rate": 4.689271862280152e-05,
      "loss": 1.0912225723266602,
      "mean_token_accuracy": 0.7456759825348854,
      "num_tokens": 23780082.0,
      "step": 12040
    },
    {
      "entropy": 1.0745447628200053,
      "epoch": 0.3451584721366885,
      "grad_norm": 6.28125,
      "learning_rate": 4.6887206324913135e-05,
      "loss": 1.0282931327819824,
      "mean_token_accuracy": 0.7541920978575944,
      "num_tokens": 23799880.0,
      "step": 12050
    },
    {
      "entropy": 1.1610807899385691,
      "epoch": 0.345444910702777,
      "grad_norm": 7.59375,
      "learning_rate": 4.6881689466547955e-05,
      "loss": 1.2484560012817383,
      "mean_token_accuracy": 0.714300836622715,
      "num_tokens": 23818351.0,
      "step": 12060
    },
    {
      "entropy": 1.2199388794600963,
      "epoch": 0.34573134926886556,
      "grad_norm": 7.59375,
      "learning_rate": 4.687616804885549e-05,
      "loss": 1.2164305686950683,
      "mean_token_accuracy": 0.7274310253560543,
      "num_tokens": 23838253.0,
      "step": 12070
    },
    {
      "entropy": 1.2003101717680693,
      "epoch": 0.34601778783495407,
      "grad_norm": 8.1875,
      "learning_rate": 4.687064207298619e-05,
      "loss": 1.2673505783081054,
      "mean_token_accuracy": 0.7238612420856952,
      "num_tokens": 23857884.0,
      "step": 12080
    },
    {
      "entropy": 1.0899714313447475,
      "epoch": 0.34630422640104264,
      "grad_norm": 6.84375,
      "learning_rate": 4.686511154009148e-05,
      "loss": 1.0469305038452148,
      "mean_token_accuracy": 0.7557990480214357,
      "num_tokens": 23877746.0,
      "step": 12090
    },
    {
      "entropy": 1.1794720616191625,
      "epoch": 0.3465906649671312,
      "grad_norm": 6.75,
      "learning_rate": 4.685957645132371e-05,
      "loss": 1.2009960174560548,
      "mean_token_accuracy": 0.7408746611326933,
      "num_tokens": 23897390.0,
      "step": 12100
    },
    {
      "entropy": 1.0683152735233308,
      "epoch": 0.3468771035332197,
      "grad_norm": 10.5625,
      "learning_rate": 4.6854036807836166e-05,
      "loss": 1.0982565879821777,
      "mean_token_accuracy": 0.7555297255516052,
      "num_tokens": 23916665.0,
      "step": 12110
    },
    {
      "entropy": 1.1548657290637494,
      "epoch": 0.3471635420993083,
      "grad_norm": 7.59375,
      "learning_rate": 4.6848492610783134e-05,
      "loss": 1.2286520957946778,
      "mean_token_accuracy": 0.7238879334181547,
      "num_tokens": 23936141.0,
      "step": 12120
    },
    {
      "entropy": 1.194304696470499,
      "epoch": 0.3474499806653968,
      "grad_norm": 8.1875,
      "learning_rate": 4.68429438613198e-05,
      "loss": 1.2142595291137694,
      "mean_token_accuracy": 0.7304021198302507,
      "num_tokens": 23956756.0,
      "step": 12130
    },
    {
      "entropy": 1.1930865988135337,
      "epoch": 0.34773641923148535,
      "grad_norm": 6.71875,
      "learning_rate": 4.683739056060231e-05,
      "loss": 1.1942502975463867,
      "mean_token_accuracy": 0.7369113426655531,
      "num_tokens": 23976231.0,
      "step": 12140
    },
    {
      "entropy": 1.0730652177706361,
      "epoch": 0.34802285779757386,
      "grad_norm": 8.0625,
      "learning_rate": 4.683183270978778e-05,
      "loss": 1.0902474403381348,
      "mean_token_accuracy": 0.7452852111309767,
      "num_tokens": 23996499.0,
      "step": 12150
    },
    {
      "entropy": 1.2924365883693099,
      "epoch": 0.3483092963636624,
      "grad_norm": 8.5625,
      "learning_rate": 4.6826270310034246e-05,
      "loss": 1.3115803718566894,
      "mean_token_accuracy": 0.7057915881276131,
      "num_tokens": 24017347.0,
      "step": 12160
    },
    {
      "entropy": 1.0572031695395707,
      "epoch": 0.34859573492975093,
      "grad_norm": 8.625,
      "learning_rate": 4.6820703362500725e-05,
      "loss": 1.0731576919555663,
      "mean_token_accuracy": 0.7637318138033151,
      "num_tokens": 24036541.0,
      "step": 12170
    },
    {
      "entropy": 1.2365605905652046,
      "epoch": 0.3488821734958395,
      "grad_norm": 6.53125,
      "learning_rate": 4.681513186834714e-05,
      "loss": 1.2278698921203612,
      "mean_token_accuracy": 0.7226788450032473,
      "num_tokens": 24056782.0,
      "step": 12180
    },
    {
      "entropy": 1.1317774463444947,
      "epoch": 0.349168612061928,
      "grad_norm": 8.0625,
      "learning_rate": 4.680955582873438e-05,
      "loss": 1.1577448844909668,
      "mean_token_accuracy": 0.7462615061551332,
      "num_tokens": 24077354.0,
      "step": 12190
    },
    {
      "entropy": 1.0215800609439611,
      "epoch": 0.3494550506280166,
      "grad_norm": 7.3125,
      "learning_rate": 4.68039752448243e-05,
      "loss": 1.0143744468688964,
      "mean_token_accuracy": 0.7628859091550112,
      "num_tokens": 24098397.0,
      "step": 12200
    },
    {
      "entropy": 1.211158404313028,
      "epoch": 0.3497414891941051,
      "grad_norm": 8.5,
      "learning_rate": 4.679839011777968e-05,
      "loss": 1.3002041816711425,
      "mean_token_accuracy": 0.7087782345712185,
      "num_tokens": 24118472.0,
      "step": 12210
    },
    {
      "entropy": 1.0564627762883902,
      "epoch": 0.35002792776019365,
      "grad_norm": 6.875,
      "learning_rate": 4.679280044876425e-05,
      "loss": 1.0647855758666993,
      "mean_token_accuracy": 0.7530502326786518,
      "num_tokens": 24138183.0,
      "step": 12220
    },
    {
      "entropy": 1.1522968597710133,
      "epoch": 0.35031436632628216,
      "grad_norm": 6.75,
      "learning_rate": 4.678720623894269e-05,
      "loss": 1.2218456268310547,
      "mean_token_accuracy": 0.7277620542794466,
      "num_tokens": 24157648.0,
      "step": 12230
    },
    {
      "entropy": 1.1053988685831428,
      "epoch": 0.3506008048923707,
      "grad_norm": 6.71875,
      "learning_rate": 4.678160748948062e-05,
      "loss": 1.1578386306762696,
      "mean_token_accuracy": 0.7310198895633221,
      "num_tokens": 24178043.0,
      "step": 12240
    },
    {
      "entropy": 1.167086797580123,
      "epoch": 0.35088724345845923,
      "grad_norm": 7.53125,
      "learning_rate": 4.677600420154461e-05,
      "loss": 1.1752772331237793,
      "mean_token_accuracy": 0.7255244225263595,
      "num_tokens": 24197375.0,
      "step": 12250
    },
    {
      "entropy": 1.2581705693155527,
      "epoch": 0.3511736820245478,
      "grad_norm": 7.59375,
      "learning_rate": 4.6770396376302184e-05,
      "loss": 1.2895773887634276,
      "mean_token_accuracy": 0.7143769770860672,
      "num_tokens": 24217719.0,
      "step": 12260
    },
    {
      "entropy": 1.1056284755468369,
      "epoch": 0.3514601205906363,
      "grad_norm": 8.375,
      "learning_rate": 4.6764784014921806e-05,
      "loss": 1.1673913955688477,
      "mean_token_accuracy": 0.7443995978683233,
      "num_tokens": 24237291.0,
      "step": 12270
    },
    {
      "entropy": 1.1574806613847612,
      "epoch": 0.35174655915672487,
      "grad_norm": 7.40625,
      "learning_rate": 4.6759167118572863e-05,
      "loss": 1.1845383644104004,
      "mean_token_accuracy": 0.7294176701456309,
      "num_tokens": 24257995.0,
      "step": 12280
    },
    {
      "entropy": 1.0985305689275264,
      "epoch": 0.3520329977228134,
      "grad_norm": 9.0,
      "learning_rate": 4.6753545688425725e-05,
      "loss": 1.138932704925537,
      "mean_token_accuracy": 0.7533436857163907,
      "num_tokens": 24276915.0,
      "step": 12290
    },
    {
      "entropy": 1.1947484020143748,
      "epoch": 0.35231943628890194,
      "grad_norm": 7.375,
      "learning_rate": 4.674791972565168e-05,
      "loss": 1.2274359703063964,
      "mean_token_accuracy": 0.7243363671004772,
      "num_tokens": 24297453.0,
      "step": 12300
    },
    {
      "entropy": 1.1885976530611515,
      "epoch": 0.35260587485499045,
      "grad_norm": 9.3125,
      "learning_rate": 4.674228923142297e-05,
      "loss": 1.2606486320495605,
      "mean_token_accuracy": 0.7255305778235197,
      "num_tokens": 24316690.0,
      "step": 12310
    },
    {
      "entropy": 1.3291325971484185,
      "epoch": 0.352892313421079,
      "grad_norm": 11.5,
      "learning_rate": 4.6736654206912785e-05,
      "loss": 1.3377924919128419,
      "mean_token_accuracy": 0.710524632781744,
      "num_tokens": 24336587.0,
      "step": 12320
    },
    {
      "entropy": 1.1792631223797798,
      "epoch": 0.3531787519871675,
      "grad_norm": 7.59375,
      "learning_rate": 4.6731014653295244e-05,
      "loss": 1.208043384552002,
      "mean_token_accuracy": 0.7287311658263207,
      "num_tokens": 24357904.0,
      "step": 12330
    },
    {
      "entropy": 1.1591805048286914,
      "epoch": 0.3534651905532561,
      "grad_norm": 6.34375,
      "learning_rate": 4.672537057174543e-05,
      "loss": 1.1509671211242676,
      "mean_token_accuracy": 0.7389872211962938,
      "num_tokens": 24378113.0,
      "step": 12340
    },
    {
      "entropy": 1.2786912132054566,
      "epoch": 0.3537516291193446,
      "grad_norm": 7.78125,
      "learning_rate": 4.6719721963439354e-05,
      "loss": 1.3041322708129883,
      "mean_token_accuracy": 0.7151679765433073,
      "num_tokens": 24398619.0,
      "step": 12350
    },
    {
      "entropy": 1.1455071780830621,
      "epoch": 0.35403806768543317,
      "grad_norm": 9.25,
      "learning_rate": 4.671406882955397e-05,
      "loss": 1.1349172592163086,
      "mean_token_accuracy": 0.7472321063280105,
      "num_tokens": 24417762.0,
      "step": 12360
    },
    {
      "entropy": 1.3051155537366868,
      "epoch": 0.35432450625152173,
      "grad_norm": 8.375,
      "learning_rate": 4.670841117126719e-05,
      "loss": 1.3622515678405762,
      "mean_token_accuracy": 0.6951520700007677,
      "num_tokens": 24438024.0,
      "step": 12370
    },
    {
      "entropy": 1.0951440989971162,
      "epoch": 0.35461094481761024,
      "grad_norm": 7.4375,
      "learning_rate": 4.670274898975785e-05,
      "loss": 1.0845849990844727,
      "mean_token_accuracy": 0.7515030659735202,
      "num_tokens": 24456723.0,
      "step": 12380
    },
    {
      "entropy": 1.083009223267436,
      "epoch": 0.3548973833836988,
      "grad_norm": 7.28125,
      "learning_rate": 4.6697082286205746e-05,
      "loss": 1.0777643203735352,
      "mean_token_accuracy": 0.7477686557918787,
      "num_tokens": 24475421.0,
      "step": 12390
    },
    {
      "entropy": 1.1794935889542102,
      "epoch": 0.3551838219497873,
      "grad_norm": 8.125,
      "learning_rate": 4.6691411061791604e-05,
      "loss": 1.2202360153198242,
      "mean_token_accuracy": 0.7202108513563872,
      "num_tokens": 24495907.0,
      "step": 12400
    },
    {
      "entropy": 1.154681123420596,
      "epoch": 0.3554702605158759,
      "grad_norm": 7.34375,
      "learning_rate": 4.668573531769709e-05,
      "loss": 1.1510541915893555,
      "mean_token_accuracy": 0.7336853880435228,
      "num_tokens": 24516539.0,
      "step": 12410
    },
    {
      "entropy": 1.1614432524889708,
      "epoch": 0.3557566990819644,
      "grad_norm": 6.84375,
      "learning_rate": 4.668005505510482e-05,
      "loss": 1.1538786888122559,
      "mean_token_accuracy": 0.7410368528217077,
      "num_tokens": 24536548.0,
      "step": 12420
    },
    {
      "entropy": 1.1056133069097995,
      "epoch": 0.35604313764805295,
      "grad_norm": 7.53125,
      "learning_rate": 4.667437027519835e-05,
      "loss": 1.100418472290039,
      "mean_token_accuracy": 0.7426928225904703,
      "num_tokens": 24556188.0,
      "step": 12430
    },
    {
      "entropy": 1.1577647674828768,
      "epoch": 0.35632957621414146,
      "grad_norm": 8.4375,
      "learning_rate": 4.6668680979162174e-05,
      "loss": 1.2466042518615723,
      "mean_token_accuracy": 0.7233197886496783,
      "num_tokens": 24574767.0,
      "step": 12440
    },
    {
      "entropy": 1.200463857129216,
      "epoch": 0.35661601478023003,
      "grad_norm": 6.28125,
      "learning_rate": 4.6662987168181734e-05,
      "loss": 1.217890167236328,
      "mean_token_accuracy": 0.7174221593886614,
      "num_tokens": 24593741.0,
      "step": 12450
    },
    {
      "entropy": 1.1855518642812968,
      "epoch": 0.35690245334631854,
      "grad_norm": 7.4375,
      "learning_rate": 4.66572888434434e-05,
      "loss": 1.1821393966674805,
      "mean_token_accuracy": 0.7281763609498739,
      "num_tokens": 24614762.0,
      "step": 12460
    },
    {
      "entropy": 1.1904833726584911,
      "epoch": 0.3571888919124071,
      "grad_norm": 7.8125,
      "learning_rate": 4.6651586006134505e-05,
      "loss": 1.1578195571899415,
      "mean_token_accuracy": 0.7455261457711458,
      "num_tokens": 24634841.0,
      "step": 12470
    },
    {
      "entropy": 1.1210447106510402,
      "epoch": 0.3574753304784956,
      "grad_norm": 7.96875,
      "learning_rate": 4.6645878657443284e-05,
      "loss": 1.1873265266418458,
      "mean_token_accuracy": 0.7319249577820301,
      "num_tokens": 24653218.0,
      "step": 12480
    },
    {
      "entropy": 1.1087748092599212,
      "epoch": 0.3577617690445842,
      "grad_norm": 7.5,
      "learning_rate": 4.664016679855896e-05,
      "loss": 1.1220198631286622,
      "mean_token_accuracy": 0.7515978630632162,
      "num_tokens": 24673515.0,
      "step": 12490
    },
    {
      "entropy": 1.1439260589890181,
      "epoch": 0.3580482076106727,
      "grad_norm": 8.625,
      "learning_rate": 4.663445043067166e-05,
      "loss": 1.1369548797607423,
      "mean_token_accuracy": 0.7413583099842072,
      "num_tokens": 24693054.0,
      "step": 12500
    },
    {
      "entropy": 1.142331263795495,
      "epoch": 0.35833464617676125,
      "grad_norm": 9.4375,
      "learning_rate": 4.662872955497245e-05,
      "loss": 1.2061081886291505,
      "mean_token_accuracy": 0.739474031701684,
      "num_tokens": 24713101.0,
      "step": 12510
    },
    {
      "entropy": 1.1761005107313394,
      "epoch": 0.35862108474284976,
      "grad_norm": 5.78125,
      "learning_rate": 4.662300417265337e-05,
      "loss": 1.1623221397399903,
      "mean_token_accuracy": 0.7403987973928452,
      "num_tokens": 24734197.0,
      "step": 12520
    },
    {
      "entropy": 1.164141470566392,
      "epoch": 0.3589075233089383,
      "grad_norm": 8.4375,
      "learning_rate": 4.661727428490736e-05,
      "loss": 1.1945881843566895,
      "mean_token_accuracy": 0.7339757930487394,
      "num_tokens": 24754402.0,
      "step": 12530
    },
    {
      "entropy": 1.2093694768846035,
      "epoch": 0.35919396187502683,
      "grad_norm": 7.25,
      "learning_rate": 4.6611539892928315e-05,
      "loss": 1.2294402122497559,
      "mean_token_accuracy": 0.7220250643789768,
      "num_tokens": 24774423.0,
      "step": 12540
    },
    {
      "entropy": 1.1841968726366758,
      "epoch": 0.3594804004411154,
      "grad_norm": 8.4375,
      "learning_rate": 4.660580099791108e-05,
      "loss": 1.1973770141601563,
      "mean_token_accuracy": 0.7244217835366726,
      "num_tokens": 24794520.0,
      "step": 12550
    },
    {
      "entropy": 1.0361510679125785,
      "epoch": 0.3597668390072039,
      "grad_norm": 7.0625,
      "learning_rate": 4.660005760105141e-05,
      "loss": 1.0199840545654297,
      "mean_token_accuracy": 0.767999405413866,
      "num_tokens": 24812627.0,
      "step": 12560
    },
    {
      "entropy": 1.199740276671946,
      "epoch": 0.3600532775732925,
      "grad_norm": 6.875,
      "learning_rate": 4.659430970354603e-05,
      "loss": 1.182297420501709,
      "mean_token_accuracy": 0.7283029284328222,
      "num_tokens": 24833291.0,
      "step": 12570
    },
    {
      "entropy": 1.1801016733050347,
      "epoch": 0.360339716139381,
      "grad_norm": 5.4375,
      "learning_rate": 4.6588557306592587e-05,
      "loss": 1.3230598449707032,
      "mean_token_accuracy": 0.7075336508452892,
      "num_tokens": 24854305.0,
      "step": 12580
    },
    {
      "entropy": 1.1881453284993768,
      "epoch": 0.36062615470546955,
      "grad_norm": 8.9375,
      "learning_rate": 4.658280041138964e-05,
      "loss": 1.2157601356506347,
      "mean_token_accuracy": 0.7086739990860224,
      "num_tokens": 24873142.0,
      "step": 12590
    },
    {
      "entropy": 1.1658820813521742,
      "epoch": 0.36091259327155806,
      "grad_norm": 9.6875,
      "learning_rate": 4.657703901913675e-05,
      "loss": 1.1919174194335938,
      "mean_token_accuracy": 0.7389684028923511,
      "num_tokens": 24894882.0,
      "step": 12600
    },
    {
      "entropy": 1.197008427977562,
      "epoch": 0.3611990318376466,
      "grad_norm": 5.84375,
      "learning_rate": 4.657127313103434e-05,
      "loss": 1.2174315452575684,
      "mean_token_accuracy": 0.72830794416368,
      "num_tokens": 24914791.0,
      "step": 12610
    },
    {
      "entropy": 1.1865931533277034,
      "epoch": 0.36148547040373513,
      "grad_norm": 7.75,
      "learning_rate": 4.656550274828383e-05,
      "loss": 1.1610807418823241,
      "mean_token_accuracy": 0.7274734944105148,
      "num_tokens": 24934601.0,
      "step": 12620
    },
    {
      "entropy": 1.145196046680212,
      "epoch": 0.3617719089698237,
      "grad_norm": 6.625,
      "learning_rate": 4.6559727872087546e-05,
      "loss": 1.2269505500793456,
      "mean_token_accuracy": 0.7322268161922694,
      "num_tokens": 24955211.0,
      "step": 12630
    },
    {
      "entropy": 1.249273194745183,
      "epoch": 0.36205834753591226,
      "grad_norm": 7.34375,
      "learning_rate": 4.655394850364876e-05,
      "loss": 1.2770082473754882,
      "mean_token_accuracy": 0.7197588823735714,
      "num_tokens": 24973947.0,
      "step": 12640
    },
    {
      "entropy": 1.1048865262418985,
      "epoch": 0.36234478610200077,
      "grad_norm": 6.59375,
      "learning_rate": 4.654816464417166e-05,
      "loss": 1.1112689971923828,
      "mean_token_accuracy": 0.7476239636540413,
      "num_tokens": 24993049.0,
      "step": 12650
    },
    {
      "entropy": 1.182585509121418,
      "epoch": 0.36263122466808934,
      "grad_norm": 9.0,
      "learning_rate": 4.654237629486141e-05,
      "loss": 1.2341876029968262,
      "mean_token_accuracy": 0.7128869269043208,
      "num_tokens": 25012210.0,
      "step": 12660
    },
    {
      "entropy": 1.1946954686194657,
      "epoch": 0.36291766323417785,
      "grad_norm": 9.1875,
      "learning_rate": 4.653658345692407e-05,
      "loss": 1.225621509552002,
      "mean_token_accuracy": 0.7311183448880911,
      "num_tokens": 25030998.0,
      "step": 12670
    },
    {
      "entropy": 1.2007832787930965,
      "epoch": 0.3632041018002664,
      "grad_norm": 10.5625,
      "learning_rate": 4.653078613156665e-05,
      "loss": 1.2481486320495605,
      "mean_token_accuracy": 0.713291997089982,
      "num_tokens": 25051569.0,
      "step": 12680
    },
    {
      "entropy": 1.1712095331400634,
      "epoch": 0.3634905403663549,
      "grad_norm": 8.5,
      "learning_rate": 4.65249843199971e-05,
      "loss": 1.203207302093506,
      "mean_token_accuracy": 0.7292301647365094,
      "num_tokens": 25070888.0,
      "step": 12690
    },
    {
      "entropy": 1.2710014339536428,
      "epoch": 0.3637769789324435,
      "grad_norm": 9.6875,
      "learning_rate": 4.6519178023424317e-05,
      "loss": 1.2887212753295898,
      "mean_token_accuracy": 0.7167570725083351,
      "num_tokens": 25090122.0,
      "step": 12700
    },
    {
      "entropy": 1.2152038557454943,
      "epoch": 0.364063417498532,
      "grad_norm": 7.59375,
      "learning_rate": 4.6513367243058094e-05,
      "loss": 1.1899870872497558,
      "mean_token_accuracy": 0.7301798451691865,
      "num_tokens": 25110748.0,
      "step": 12710
    },
    {
      "entropy": 1.1407973047345876,
      "epoch": 0.36434985606462056,
      "grad_norm": 8.0,
      "learning_rate": 4.650755198010919e-05,
      "loss": 1.1604823112487792,
      "mean_token_accuracy": 0.742802207544446,
      "num_tokens": 25131176.0,
      "step": 12720
    },
    {
      "entropy": 1.1766152661293745,
      "epoch": 0.36463629463070907,
      "grad_norm": 9.1875,
      "learning_rate": 4.650173223578928e-05,
      "loss": 1.195823860168457,
      "mean_token_accuracy": 0.7396425627171993,
      "num_tokens": 25149281.0,
      "step": 12730
    },
    {
      "entropy": 1.0965004667639733,
      "epoch": 0.36492273319679763,
      "grad_norm": 6.84375,
      "learning_rate": 4.6495908011311e-05,
      "loss": 1.1518229484558105,
      "mean_token_accuracy": 0.7351251032203436,
      "num_tokens": 25169373.0,
      "step": 12740
    },
    {
      "entropy": 1.1608255438506603,
      "epoch": 0.36520917176288614,
      "grad_norm": 8.5,
      "learning_rate": 4.6490079307887884e-05,
      "loss": 1.143186092376709,
      "mean_token_accuracy": 0.7366735283285379,
      "num_tokens": 25188956.0,
      "step": 12750
    },
    {
      "entropy": 1.1373560000211,
      "epoch": 0.3654956103289747,
      "grad_norm": 9.875,
      "learning_rate": 4.648424612673443e-05,
      "loss": 1.1254226684570312,
      "mean_token_accuracy": 0.7469326909631491,
      "num_tokens": 25208548.0,
      "step": 12760
    },
    {
      "entropy": 1.20901506729424,
      "epoch": 0.3657820488950632,
      "grad_norm": 8.5,
      "learning_rate": 4.647840846906604e-05,
      "loss": 1.2731991767883302,
      "mean_token_accuracy": 0.7283838141709567,
      "num_tokens": 25230110.0,
      "step": 12770
    },
    {
      "entropy": 1.0768377557396889,
      "epoch": 0.3660684874611518,
      "grad_norm": 10.0625,
      "learning_rate": 4.647256633609907e-05,
      "loss": 1.117074680328369,
      "mean_token_accuracy": 0.7471107337623835,
      "num_tokens": 25249868.0,
      "step": 12780
    },
    {
      "entropy": 1.2449981797486545,
      "epoch": 0.3663549260272403,
      "grad_norm": 8.625,
      "learning_rate": 4.6466719729050804e-05,
      "loss": 1.2835883140563964,
      "mean_token_accuracy": 0.7075161915272474,
      "num_tokens": 25270276.0,
      "step": 12790
    },
    {
      "entropy": 1.1636112209409475,
      "epoch": 0.36664136459332886,
      "grad_norm": 7.875,
      "learning_rate": 4.646086864913947e-05,
      "loss": 1.177051830291748,
      "mean_token_accuracy": 0.7265634182840586,
      "num_tokens": 25289737.0,
      "step": 12800
    },
    {
      "entropy": 1.226197324693203,
      "epoch": 0.36692780315941737,
      "grad_norm": 6.5,
      "learning_rate": 4.645501309758419e-05,
      "loss": 1.2565295219421386,
      "mean_token_accuracy": 0.7182535119354725,
      "num_tokens": 25307889.0,
      "step": 12810
    },
    {
      "entropy": 1.1554649438709022,
      "epoch": 0.36721424172550593,
      "grad_norm": 9.375,
      "learning_rate": 4.644915307560504e-05,
      "loss": 1.112023162841797,
      "mean_token_accuracy": 0.7399850897490978,
      "num_tokens": 25327381.0,
      "step": 12820
    },
    {
      "entropy": 1.178750059939921,
      "epoch": 0.36750068029159444,
      "grad_norm": 7.65625,
      "learning_rate": 4.6443288584423064e-05,
      "loss": 1.1486433029174805,
      "mean_token_accuracy": 0.7444168295711279,
      "num_tokens": 25347214.0,
      "step": 12830
    },
    {
      "entropy": 1.086121116578579,
      "epoch": 0.367787118857683,
      "grad_norm": 9.5625,
      "learning_rate": 4.643741962526018e-05,
      "loss": 1.100342082977295,
      "mean_token_accuracy": 0.7494371876120567,
      "num_tokens": 25367486.0,
      "step": 12840
    },
    {
      "entropy": 1.1032545395195483,
      "epoch": 0.3680735574237715,
      "grad_norm": 8.4375,
      "learning_rate": 4.643154619933926e-05,
      "loss": 1.1286057472229003,
      "mean_token_accuracy": 0.728075797110796,
      "num_tokens": 25388450.0,
      "step": 12850
    },
    {
      "entropy": 1.177094199322164,
      "epoch": 0.3683599959898601,
      "grad_norm": 7.15625,
      "learning_rate": 4.642566830788411e-05,
      "loss": 1.2549120903015136,
      "mean_token_accuracy": 0.7249290011823177,
      "num_tokens": 25409061.0,
      "step": 12860
    },
    {
      "entropy": 1.132564228028059,
      "epoch": 0.3686464345559486,
      "grad_norm": 9.4375,
      "learning_rate": 4.641978595211946e-05,
      "loss": 1.118637466430664,
      "mean_token_accuracy": 0.7513731565326452,
      "num_tokens": 25428750.0,
      "step": 12870
    },
    {
      "entropy": 1.1223754335194827,
      "epoch": 0.36893287312203715,
      "grad_norm": 7.6875,
      "learning_rate": 4.6413899133270975e-05,
      "loss": 1.0885736465454101,
      "mean_token_accuracy": 0.7486430887132883,
      "num_tokens": 25448136.0,
      "step": 12880
    },
    {
      "entropy": 1.0789169408380985,
      "epoch": 0.36921931168812566,
      "grad_norm": 7.0,
      "learning_rate": 4.640800785256525e-05,
      "loss": 1.1167041778564453,
      "mean_token_accuracy": 0.7479406505823135,
      "num_tokens": 25468131.0,
      "step": 12890
    },
    {
      "entropy": 1.137805850803852,
      "epoch": 0.3695057502542142,
      "grad_norm": 8.125,
      "learning_rate": 4.64021121112298e-05,
      "loss": 1.2548883438110352,
      "mean_token_accuracy": 0.7210646469146014,
      "num_tokens": 25487271.0,
      "step": 12900
    },
    {
      "entropy": 1.0110106192529202,
      "epoch": 0.3697921888203028,
      "grad_norm": 8.25,
      "learning_rate": 4.63962119104931e-05,
      "loss": 0.9812099456787109,
      "mean_token_accuracy": 0.7751163944602013,
      "num_tokens": 25506762.0,
      "step": 12910
    },
    {
      "entropy": 1.1466988310217858,
      "epoch": 0.3700786273863913,
      "grad_norm": 8.0625,
      "learning_rate": 4.639030725158451e-05,
      "loss": 1.1649755477905273,
      "mean_token_accuracy": 0.7391448896378279,
      "num_tokens": 25526581.0,
      "step": 12920
    },
    {
      "entropy": 1.2701556280255317,
      "epoch": 0.37036506595247987,
      "grad_norm": 8.5,
      "learning_rate": 4.6384398135734343e-05,
      "loss": 1.2245131492614747,
      "mean_token_accuracy": 0.723613515868783,
      "num_tokens": 25545665.0,
      "step": 12930
    },
    {
      "entropy": 1.1207250935956836,
      "epoch": 0.3706515045185684,
      "grad_norm": 7.71875,
      "learning_rate": 4.637848456417384e-05,
      "loss": 1.1522951126098633,
      "mean_token_accuracy": 0.7404493417590856,
      "num_tokens": 25565966.0,
      "step": 12940
    },
    {
      "entropy": 1.1384451944380998,
      "epoch": 0.37093794308465694,
      "grad_norm": 9.3125,
      "learning_rate": 4.6372566538135174e-05,
      "loss": 1.1789511680603026,
      "mean_token_accuracy": 0.7320579085499048,
      "num_tokens": 25586110.0,
      "step": 12950
    },
    {
      "entropy": 1.2000788673758507,
      "epoch": 0.37122438165074545,
      "grad_norm": 7.59375,
      "learning_rate": 4.6366644058851434e-05,
      "loss": 1.2353042602539062,
      "mean_token_accuracy": 0.7218313202261925,
      "num_tokens": 25605054.0,
      "step": 12960
    },
    {
      "entropy": 1.1470486987382174,
      "epoch": 0.371510820216834,
      "grad_norm": 8.4375,
      "learning_rate": 4.6360717127556655e-05,
      "loss": 1.1547849655151368,
      "mean_token_accuracy": 0.7341628767549991,
      "num_tokens": 25625184.0,
      "step": 12970
    },
    {
      "entropy": 1.1803495505824686,
      "epoch": 0.3717972587829225,
      "grad_norm": 7.0,
      "learning_rate": 4.635478574548577e-05,
      "loss": 1.2092257499694825,
      "mean_token_accuracy": 0.729374673962593,
      "num_tokens": 25647260.0,
      "step": 12980
    },
    {
      "entropy": 1.2424621149897574,
      "epoch": 0.3720836973490111,
      "grad_norm": 7.65625,
      "learning_rate": 4.634884991387467e-05,
      "loss": 1.3212327003479003,
      "mean_token_accuracy": 0.7062042485922575,
      "num_tokens": 25667509.0,
      "step": 12990
    },
    {
      "entropy": 1.148267563059926,
      "epoch": 0.3723701359150996,
      "grad_norm": 9.625,
      "learning_rate": 4.634290963396016e-05,
      "loss": 1.1369263648986816,
      "mean_token_accuracy": 0.7378683008253575,
      "num_tokens": 25686826.0,
      "step": 13000
    },
    {
      "entropy": 1.2015334542840719,
      "epoch": 0.37265657448118816,
      "grad_norm": 8.5625,
      "learning_rate": 4.633696490697997e-05,
      "loss": 1.2385712623596192,
      "mean_token_accuracy": 0.7222250301390887,
      "num_tokens": 25707342.0,
      "step": 13010
    },
    {
      "entropy": 1.2658307656645775,
      "epoch": 0.3729430130472767,
      "grad_norm": 9.0,
      "learning_rate": 4.6331015734172755e-05,
      "loss": 1.3023976325988769,
      "mean_token_accuracy": 0.7146957296878099,
      "num_tokens": 25727302.0,
      "step": 13020
    },
    {
      "entropy": 1.258390760421753,
      "epoch": 0.37322945161336524,
      "grad_norm": 8.5,
      "learning_rate": 4.632506211677812e-05,
      "loss": 1.2614652633666992,
      "mean_token_accuracy": 0.7186996635049582,
      "num_tokens": 25750127.0,
      "step": 13030
    },
    {
      "entropy": 1.1118933875113726,
      "epoch": 0.37351589017945375,
      "grad_norm": 6.375,
      "learning_rate": 4.631910405603654e-05,
      "loss": 1.1046648025512695,
      "mean_token_accuracy": 0.7437384359538555,
      "num_tokens": 25769351.0,
      "step": 13040
    },
    {
      "entropy": 1.0717617638409138,
      "epoch": 0.3738023287455423,
      "grad_norm": 9.6875,
      "learning_rate": 4.6313141553189494e-05,
      "loss": 1.0836981773376464,
      "mean_token_accuracy": 0.750560249760747,
      "num_tokens": 25788352.0,
      "step": 13050
    },
    {
      "entropy": 1.086102511174977,
      "epoch": 0.3740887673116308,
      "grad_norm": 8.3125,
      "learning_rate": 4.630717460947932e-05,
      "loss": 1.1510765075683593,
      "mean_token_accuracy": 0.7447220861911774,
      "num_tokens": 25808519.0,
      "step": 13060
    },
    {
      "entropy": 1.2643359094858169,
      "epoch": 0.3743752058777194,
      "grad_norm": 8.5625,
      "learning_rate": 4.630120322614931e-05,
      "loss": 1.3303447723388673,
      "mean_token_accuracy": 0.710831318795681,
      "num_tokens": 25829264.0,
      "step": 13070
    },
    {
      "entropy": 1.3563353499397635,
      "epoch": 0.3746616444438079,
      "grad_norm": 8.125,
      "learning_rate": 4.629522740444368e-05,
      "loss": 1.344771671295166,
      "mean_token_accuracy": 0.7004264548420907,
      "num_tokens": 25848769.0,
      "step": 13080
    },
    {
      "entropy": 1.1492317229509355,
      "epoch": 0.37494808300989646,
      "grad_norm": 7.3125,
      "learning_rate": 4.6289247145607565e-05,
      "loss": 1.159813404083252,
      "mean_token_accuracy": 0.731747355684638,
      "num_tokens": 25867738.0,
      "step": 13090
    },
    {
      "entropy": 1.0995746402069926,
      "epoch": 0.37523452157598497,
      "grad_norm": 8.125,
      "learning_rate": 4.6283262450887036e-05,
      "loss": 1.0462042808532714,
      "mean_token_accuracy": 0.7593157414346934,
      "num_tokens": 25888821.0,
      "step": 13100
    },
    {
      "entropy": 1.0386188369244338,
      "epoch": 0.37552096014207353,
      "grad_norm": 7.40625,
      "learning_rate": 4.627727332152907e-05,
      "loss": 1.0563633918762207,
      "mean_token_accuracy": 0.747785908356309,
      "num_tokens": 25906907.0,
      "step": 13110
    },
    {
      "entropy": 1.150918895751238,
      "epoch": 0.37580739870816204,
      "grad_norm": 8.0625,
      "learning_rate": 4.627127975878158e-05,
      "loss": 1.1532745361328125,
      "mean_token_accuracy": 0.7451213531196117,
      "num_tokens": 25926223.0,
      "step": 13120
    },
    {
      "entropy": 1.1642468746751546,
      "epoch": 0.3760938372742506,
      "grad_norm": 8.125,
      "learning_rate": 4.626528176389341e-05,
      "loss": 1.1743946075439453,
      "mean_token_accuracy": 0.7289228297770023,
      "num_tokens": 25945788.0,
      "step": 13130
    },
    {
      "entropy": 1.1000314101576805,
      "epoch": 0.3763802758403391,
      "grad_norm": 10.1875,
      "learning_rate": 4.625927933811431e-05,
      "loss": 1.1201741218566894,
      "mean_token_accuracy": 0.7501476533710957,
      "num_tokens": 25965700.0,
      "step": 13140
    },
    {
      "entropy": 1.1132942715659737,
      "epoch": 0.3766667144064277,
      "grad_norm": 9.875,
      "learning_rate": 4.625327248269496e-05,
      "loss": 1.1445460319519043,
      "mean_token_accuracy": 0.7341828580945731,
      "num_tokens": 25984973.0,
      "step": 13150
    },
    {
      "entropy": 1.1168049293570221,
      "epoch": 0.3769531529725162,
      "grad_norm": 8.625,
      "learning_rate": 4.624726119888697e-05,
      "loss": 1.1976551055908202,
      "mean_token_accuracy": 0.734917625784874,
      "num_tokens": 26006644.0,
      "step": 13160
    },
    {
      "entropy": 1.1321493003517389,
      "epoch": 0.37723959153860476,
      "grad_norm": 9.0,
      "learning_rate": 4.624124548794286e-05,
      "loss": 1.162496852874756,
      "mean_token_accuracy": 0.7441964868456126,
      "num_tokens": 26026331.0,
      "step": 13170
    },
    {
      "entropy": 1.1288546834141016,
      "epoch": 0.3775260301046933,
      "grad_norm": 8.1875,
      "learning_rate": 4.6235225351116084e-05,
      "loss": 1.1391674041748048,
      "mean_token_accuracy": 0.7361530654132367,
      "num_tokens": 26046424.0,
      "step": 13180
    },
    {
      "entropy": 1.0502203691750764,
      "epoch": 0.37781246867078183,
      "grad_norm": 6.3125,
      "learning_rate": 4.622920078966102e-05,
      "loss": 1.035305118560791,
      "mean_token_accuracy": 0.7607167676091194,
      "num_tokens": 26066629.0,
      "step": 13190
    },
    {
      "entropy": 1.2320741161704063,
      "epoch": 0.3780989072368704,
      "grad_norm": 7.875,
      "learning_rate": 4.622317180483295e-05,
      "loss": 1.2428441047668457,
      "mean_token_accuracy": 0.7151221863925457,
      "num_tokens": 26087182.0,
      "step": 13200
    },
    {
      "entropy": 1.0963316362351179,
      "epoch": 0.3783853458029589,
      "grad_norm": 12.6875,
      "learning_rate": 4.6217138397888095e-05,
      "loss": 1.1709476470947267,
      "mean_token_accuracy": 0.7431160639971495,
      "num_tokens": 26105706.0,
      "step": 13210
    },
    {
      "entropy": 1.1012681659311057,
      "epoch": 0.37867178436904747,
      "grad_norm": 6.21875,
      "learning_rate": 4.6211100570083594e-05,
      "loss": 1.1351963996887207,
      "mean_token_accuracy": 0.7485731072723866,
      "num_tokens": 26124785.0,
      "step": 13220
    },
    {
      "entropy": 1.1518997572362424,
      "epoch": 0.378958222935136,
      "grad_norm": 6.65625,
      "learning_rate": 4.6205058322677494e-05,
      "loss": 1.1630259513854981,
      "mean_token_accuracy": 0.7407668605446815,
      "num_tokens": 26143407.0,
      "step": 13230
    },
    {
      "entropy": 1.2425869055092336,
      "epoch": 0.37924466150122454,
      "grad_norm": 7.28125,
      "learning_rate": 4.6199011656928795e-05,
      "loss": 1.1965776443481446,
      "mean_token_accuracy": 0.7309584684669972,
      "num_tokens": 26163423.0,
      "step": 13240
    },
    {
      "entropy": 1.1070094164460897,
      "epoch": 0.37953110006731305,
      "grad_norm": 6.71875,
      "learning_rate": 4.619296057409738e-05,
      "loss": 1.1301518440246583,
      "mean_token_accuracy": 0.7570125795900822,
      "num_tokens": 26182911.0,
      "step": 13250
    },
    {
      "entropy": 1.1243059342727064,
      "epoch": 0.3798175386334016,
      "grad_norm": 7.125,
      "learning_rate": 4.618690507544407e-05,
      "loss": 1.1588628768920899,
      "mean_token_accuracy": 0.7309739802032709,
      "num_tokens": 26202205.0,
      "step": 13260
    },
    {
      "entropy": 1.0504294570535422,
      "epoch": 0.38010397719949013,
      "grad_norm": 6.5625,
      "learning_rate": 4.618084516223061e-05,
      "loss": 1.0572033882141114,
      "mean_token_accuracy": 0.760616147890687,
      "num_tokens": 26221281.0,
      "step": 13270
    },
    {
      "entropy": 1.123620861582458,
      "epoch": 0.3803904157655787,
      "grad_norm": 7.28125,
      "learning_rate": 4.617478083571965e-05,
      "loss": 1.1137992858886718,
      "mean_token_accuracy": 0.7539139159023762,
      "num_tokens": 26240713.0,
      "step": 13280
    },
    {
      "entropy": 1.1734141055494547,
      "epoch": 0.3806768543316672,
      "grad_norm": 6.25,
      "learning_rate": 4.616871209717478e-05,
      "loss": 1.1763044357299806,
      "mean_token_accuracy": 0.7280381541699171,
      "num_tokens": 26261462.0,
      "step": 13290
    },
    {
      "entropy": 1.129241468757391,
      "epoch": 0.38096329289775577,
      "grad_norm": 7.78125,
      "learning_rate": 4.6162638947860505e-05,
      "loss": 1.1214021682739257,
      "mean_token_accuracy": 0.7374898888170719,
      "num_tokens": 26281460.0,
      "step": 13300
    },
    {
      "entropy": 1.187988194450736,
      "epoch": 0.3812497314638443,
      "grad_norm": 8.3125,
      "learning_rate": 4.615656138904222e-05,
      "loss": 1.2414039611816405,
      "mean_token_accuracy": 0.7270261164754629,
      "num_tokens": 26301046.0,
      "step": 13310
    },
    {
      "entropy": 1.0679695375263691,
      "epoch": 0.38153617002993284,
      "grad_norm": 8.375,
      "learning_rate": 4.615047942198628e-05,
      "loss": 1.1079164505004884,
      "mean_token_accuracy": 0.7502503082156181,
      "num_tokens": 26320569.0,
      "step": 13320
    },
    {
      "entropy": 1.2352590780705213,
      "epoch": 0.38182260859602135,
      "grad_norm": 8.875,
      "learning_rate": 4.614439304795993e-05,
      "loss": 1.2464200019836427,
      "mean_token_accuracy": 0.7296668808907271,
      "num_tokens": 26340804.0,
      "step": 13330
    },
    {
      "entropy": 1.03939205147326,
      "epoch": 0.3821090471621099,
      "grad_norm": 7.59375,
      "learning_rate": 4.613830226823135e-05,
      "loss": 1.0378402709960937,
      "mean_token_accuracy": 0.7564418643712998,
      "num_tokens": 26360007.0,
      "step": 13340
    },
    {
      "entropy": 1.075854821316898,
      "epoch": 0.3823954857281984,
      "grad_norm": 6.875,
      "learning_rate": 4.6132207084069624e-05,
      "loss": 1.085224437713623,
      "mean_token_accuracy": 0.7537115588784218,
      "num_tokens": 26379789.0,
      "step": 13350
    },
    {
      "entropy": 1.0285083524882794,
      "epoch": 0.382681924294287,
      "grad_norm": 6.125,
      "learning_rate": 4.6126107496744774e-05,
      "loss": 1.0187339782714844,
      "mean_token_accuracy": 0.7502297695726157,
      "num_tokens": 26399546.0,
      "step": 13360
    },
    {
      "entropy": 1.194828474894166,
      "epoch": 0.3829683628603755,
      "grad_norm": 9.0625,
      "learning_rate": 4.612000350752771e-05,
      "loss": 1.1697881698608399,
      "mean_token_accuracy": 0.7338401980698108,
      "num_tokens": 26420157.0,
      "step": 13370
    },
    {
      "entropy": 1.0615633236244322,
      "epoch": 0.38325480142646406,
      "grad_norm": 6.40625,
      "learning_rate": 4.611389511769029e-05,
      "loss": 1.1142437934875489,
      "mean_token_accuracy": 0.7461676564067602,
      "num_tokens": 26440124.0,
      "step": 13380
    },
    {
      "entropy": 1.107361724972725,
      "epoch": 0.3835412399925526,
      "grad_norm": 7.28125,
      "learning_rate": 4.610778232850526e-05,
      "loss": 1.1600430488586426,
      "mean_token_accuracy": 0.7285270459949971,
      "num_tokens": 26459771.0,
      "step": 13390
    },
    {
      "entropy": 1.1385216504335403,
      "epoch": 0.38382767855864114,
      "grad_norm": 7.96875,
      "learning_rate": 4.6101665141246305e-05,
      "loss": 1.2003435134887694,
      "mean_token_accuracy": 0.73530408218503,
      "num_tokens": 26480427.0,
      "step": 13400
    },
    {
      "entropy": 1.0924631152302027,
      "epoch": 0.38411411712472965,
      "grad_norm": 7.90625,
      "learning_rate": 4.609554355718802e-05,
      "loss": 1.080215835571289,
      "mean_token_accuracy": 0.7462619181722403,
      "num_tokens": 26499946.0,
      "step": 13410
    },
    {
      "entropy": 1.1041222266852855,
      "epoch": 0.3844005556908182,
      "grad_norm": 6.46875,
      "learning_rate": 4.608941757760591e-05,
      "loss": 1.091941738128662,
      "mean_token_accuracy": 0.7547808472067118,
      "num_tokens": 26519068.0,
      "step": 13420
    },
    {
      "entropy": 1.111957136541605,
      "epoch": 0.3846869942569067,
      "grad_norm": 6.5625,
      "learning_rate": 4.6083287203776406e-05,
      "loss": 1.1839715957641601,
      "mean_token_accuracy": 0.7361309833824634,
      "num_tokens": 26539661.0,
      "step": 13430
    },
    {
      "entropy": 1.0606804117560387,
      "epoch": 0.3849734328229953,
      "grad_norm": 7.84375,
      "learning_rate": 4.6077152436976846e-05,
      "loss": 1.0993440628051758,
      "mean_token_accuracy": 0.7447210818529129,
      "num_tokens": 26559390.0,
      "step": 13440
    },
    {
      "entropy": 1.140701805986464,
      "epoch": 0.38525987138908385,
      "grad_norm": 7.8125,
      "learning_rate": 4.607101327848549e-05,
      "loss": 1.1306324005126953,
      "mean_token_accuracy": 0.74290308393538,
      "num_tokens": 26578947.0,
      "step": 13450
    },
    {
      "entropy": 1.1164259765297175,
      "epoch": 0.38554630995517236,
      "grad_norm": 9.125,
      "learning_rate": 4.60648697295815e-05,
      "loss": 1.149711799621582,
      "mean_token_accuracy": 0.732585322111845,
      "num_tokens": 26598764.0,
      "step": 13460
    },
    {
      "entropy": 1.1421908893622459,
      "epoch": 0.3858327485212609,
      "grad_norm": 8.1875,
      "learning_rate": 4.605872179154497e-05,
      "loss": 1.1769161224365234,
      "mean_token_accuracy": 0.7379289224743844,
      "num_tokens": 26619657.0,
      "step": 13470
    },
    {
      "entropy": 1.0156867552548647,
      "epoch": 0.38611918708734944,
      "grad_norm": 5.84375,
      "learning_rate": 4.605256946565689e-05,
      "loss": 1.0399087905883788,
      "mean_token_accuracy": 0.7505837459117174,
      "num_tokens": 26638556.0,
      "step": 13480
    },
    {
      "entropy": 1.1651575971394776,
      "epoch": 0.386405625653438,
      "grad_norm": 9.0,
      "learning_rate": 4.60464127531992e-05,
      "loss": 1.1997819900512696,
      "mean_token_accuracy": 0.7357849251478911,
      "num_tokens": 26658281.0,
      "step": 13490
    },
    {
      "entropy": 1.0296811982989311,
      "epoch": 0.3866920642195265,
      "grad_norm": 6.71875,
      "learning_rate": 4.60402516554547e-05,
      "loss": 1.0078508377075195,
      "mean_token_accuracy": 0.7690814465284348,
      "num_tokens": 26678149.0,
      "step": 13500
    },
    {
      "entropy": 1.0274323631078004,
      "epoch": 0.3869785027856151,
      "grad_norm": 7.59375,
      "learning_rate": 4.603408617370716e-05,
      "loss": 1.1229555130004882,
      "mean_token_accuracy": 0.7420459054410458,
      "num_tokens": 26696905.0,
      "step": 13510
    },
    {
      "entropy": 1.0575699789449573,
      "epoch": 0.3872649413517036,
      "grad_norm": 7.0625,
      "learning_rate": 4.6027916309241216e-05,
      "loss": 1.0626741409301759,
      "mean_token_accuracy": 0.7535108409821987,
      "num_tokens": 26716428.0,
      "step": 13520
    },
    {
      "entropy": 1.1197260327637195,
      "epoch": 0.38755137991779215,
      "grad_norm": 7.4375,
      "learning_rate": 4.602174206334243e-05,
      "loss": 1.1604159355163575,
      "mean_token_accuracy": 0.7372713837772608,
      "num_tokens": 26737015.0,
      "step": 13530
    },
    {
      "entropy": 1.11889802031219,
      "epoch": 0.38783781848388066,
      "grad_norm": 8.3125,
      "learning_rate": 4.601556343729731e-05,
      "loss": 1.1389630317687989,
      "mean_token_accuracy": 0.7445575714111328,
      "num_tokens": 26756002.0,
      "step": 13540
    },
    {
      "entropy": 1.1310115091502666,
      "epoch": 0.3881242570499692,
      "grad_norm": 5.875,
      "learning_rate": 4.600938043239324e-05,
      "loss": 1.1146280288696289,
      "mean_token_accuracy": 0.7398996811360121,
      "num_tokens": 26775592.0,
      "step": 13550
    },
    {
      "entropy": 1.2144757624715568,
      "epoch": 0.38841069561605773,
      "grad_norm": 6.25,
      "learning_rate": 4.600319304991853e-05,
      "loss": 1.1686527252197265,
      "mean_token_accuracy": 0.736454214528203,
      "num_tokens": 26796755.0,
      "step": 13560
    },
    {
      "entropy": 1.0207943353801965,
      "epoch": 0.3886971341821463,
      "grad_norm": 9.3125,
      "learning_rate": 4.599700129116239e-05,
      "loss": 1.0400163650512695,
      "mean_token_accuracy": 0.7630011141300201,
      "num_tokens": 26816165.0,
      "step": 13570
    },
    {
      "entropy": 1.1259653121232986,
      "epoch": 0.3889835727482348,
      "grad_norm": 6.625,
      "learning_rate": 4.5990805157414955e-05,
      "loss": 1.2171916961669922,
      "mean_token_accuracy": 0.7357124220579863,
      "num_tokens": 26835292.0,
      "step": 13580
    },
    {
      "entropy": 1.1280373729765416,
      "epoch": 0.38927001131432337,
      "grad_norm": 7.28125,
      "learning_rate": 4.598460464996728e-05,
      "loss": 1.1450432777404784,
      "mean_token_accuracy": 0.7380849156528712,
      "num_tokens": 26855594.0,
      "step": 13590
    },
    {
      "entropy": 1.1650241162627935,
      "epoch": 0.3895564498804119,
      "grad_norm": 8.1875,
      "learning_rate": 4.59783997701113e-05,
      "loss": 1.2044533729553222,
      "mean_token_accuracy": 0.7284088488668203,
      "num_tokens": 26876116.0,
      "step": 13600
    },
    {
      "entropy": 1.207856471836567,
      "epoch": 0.38984288844650045,
      "grad_norm": 12.0625,
      "learning_rate": 4.597219051913989e-05,
      "loss": 1.1620019912719726,
      "mean_token_accuracy": 0.7413180183619261,
      "num_tokens": 26894617.0,
      "step": 13610
    },
    {
      "entropy": 1.1125625487416984,
      "epoch": 0.39012932701258896,
      "grad_norm": 6.40625,
      "learning_rate": 4.596597689834683e-05,
      "loss": 1.1935505867004395,
      "mean_token_accuracy": 0.7307192720472813,
      "num_tokens": 26914385.0,
      "step": 13620
    },
    {
      "entropy": 1.2422377105802298,
      "epoch": 0.3904157655786775,
      "grad_norm": 6.875,
      "learning_rate": 4.5959758909026806e-05,
      "loss": 1.2270179748535157,
      "mean_token_accuracy": 0.717282223328948,
      "num_tokens": 26933008.0,
      "step": 13630
    },
    {
      "entropy": 1.198270209133625,
      "epoch": 0.39070220414476603,
      "grad_norm": 8.9375,
      "learning_rate": 4.595353655247541e-05,
      "loss": 1.234036922454834,
      "mean_token_accuracy": 0.7191448159515857,
      "num_tokens": 26955294.0,
      "step": 13640
    },
    {
      "entropy": 1.0853005960583686,
      "epoch": 0.3909886427108546,
      "grad_norm": 8.125,
      "learning_rate": 4.594730982998915e-05,
      "loss": 1.079219913482666,
      "mean_token_accuracy": 0.7601745601743459,
      "num_tokens": 26974872.0,
      "step": 13650
    },
    {
      "entropy": 1.0133636560291053,
      "epoch": 0.3912750812769431,
      "grad_norm": 9.9375,
      "learning_rate": 4.594107874286543e-05,
      "loss": 1.0367952346801759,
      "mean_token_accuracy": 0.759344007074833,
      "num_tokens": 26994629.0,
      "step": 13660
    },
    {
      "entropy": 0.9951158080250024,
      "epoch": 0.39156151984303167,
      "grad_norm": 7.6875,
      "learning_rate": 4.593484329240261e-05,
      "loss": 0.9945546150207519,
      "mean_token_accuracy": 0.7635503523051739,
      "num_tokens": 27013383.0,
      "step": 13670
    },
    {
      "entropy": 1.1127688560634852,
      "epoch": 0.3918479584091202,
      "grad_norm": 6.25,
      "learning_rate": 4.592860347989989e-05,
      "loss": 1.1570622444152832,
      "mean_token_accuracy": 0.7394256714731455,
      "num_tokens": 27033476.0,
      "step": 13680
    },
    {
      "entropy": 1.1161638729274272,
      "epoch": 0.39213439697520874,
      "grad_norm": 8.375,
      "learning_rate": 4.592235930665743e-05,
      "loss": 1.1405423164367676,
      "mean_token_accuracy": 0.736626398190856,
      "num_tokens": 27053446.0,
      "step": 13690
    },
    {
      "entropy": 1.1689563043415547,
      "epoch": 0.39242083554129725,
      "grad_norm": 7.84375,
      "learning_rate": 4.591611077397629e-05,
      "loss": 1.150797462463379,
      "mean_token_accuracy": 0.7299564730376005,
      "num_tokens": 27073627.0,
      "step": 13700
    },
    {
      "entropy": 1.0925851561129094,
      "epoch": 0.3927072741073858,
      "grad_norm": 7.34375,
      "learning_rate": 4.5909857883158416e-05,
      "loss": 1.131083869934082,
      "mean_token_accuracy": 0.7346471883356571,
      "num_tokens": 27092831.0,
      "step": 13710
    },
    {
      "entropy": 1.2008197639137506,
      "epoch": 0.3929937126734744,
      "grad_norm": 6.4375,
      "learning_rate": 4.5903600635506674e-05,
      "loss": 1.2517218589782715,
      "mean_token_accuracy": 0.7185905132442713,
      "num_tokens": 27112691.0,
      "step": 13720
    },
    {
      "entropy": 1.1140727333724498,
      "epoch": 0.3932801512395629,
      "grad_norm": 11.25,
      "learning_rate": 4.589733903232486e-05,
      "loss": 1.1450485229492187,
      "mean_token_accuracy": 0.7405670329928398,
      "num_tokens": 27132482.0,
      "step": 13730
    },
    {
      "entropy": 1.1029028035700321,
      "epoch": 0.39356658980565146,
      "grad_norm": 7.8125,
      "learning_rate": 4.589107307491764e-05,
      "loss": 1.0774626731872559,
      "mean_token_accuracy": 0.7490176014602185,
      "num_tokens": 27152238.0,
      "step": 13740
    },
    {
      "entropy": 1.0316357742995024,
      "epoch": 0.39385302837173997,
      "grad_norm": 7.46875,
      "learning_rate": 4.588480276459062e-05,
      "loss": 1.0456353187561036,
      "mean_token_accuracy": 0.7569695197045803,
      "num_tokens": 27171086.0,
      "step": 13750
    },
    {
      "entropy": 1.1382686126977206,
      "epoch": 0.39413946693782853,
      "grad_norm": 7.0,
      "learning_rate": 4.587852810265029e-05,
      "loss": 1.183835792541504,
      "mean_token_accuracy": 0.7485539294779301,
      "num_tokens": 27190998.0,
      "step": 13760
    },
    {
      "entropy": 1.101496298611164,
      "epoch": 0.39442590550391704,
      "grad_norm": 9.8125,
      "learning_rate": 4.587224909040406e-05,
      "loss": 1.1658875465393066,
      "mean_token_accuracy": 0.733449823781848,
      "num_tokens": 27209943.0,
      "step": 13770
    },
    {
      "entropy": 1.148446923494339,
      "epoch": 0.3947123440700056,
      "grad_norm": 10.375,
      "learning_rate": 4.586596572916024e-05,
      "loss": 1.1994535446166992,
      "mean_token_accuracy": 0.7350861433893442,
      "num_tokens": 27230491.0,
      "step": 13780
    },
    {
      "entropy": 1.1955406554043293,
      "epoch": 0.3949987826360941,
      "grad_norm": 6.71875,
      "learning_rate": 4.5859678020228045e-05,
      "loss": 1.2236618041992187,
      "mean_token_accuracy": 0.7276370611041785,
      "num_tokens": 27250554.0,
      "step": 13790
    },
    {
      "entropy": 1.1935501739382743,
      "epoch": 0.3952852212021827,
      "grad_norm": 7.75,
      "learning_rate": 4.5853385964917595e-05,
      "loss": 1.166623020172119,
      "mean_token_accuracy": 0.7335661627352238,
      "num_tokens": 27269624.0,
      "step": 13800
    },
    {
      "entropy": 1.1109854746609926,
      "epoch": 0.3955716597682712,
      "grad_norm": 8.0,
      "learning_rate": 4.584708956453994e-05,
      "loss": 1.1576685905456543,
      "mean_token_accuracy": 0.7441776257008315,
      "num_tokens": 27290394.0,
      "step": 13810
    },
    {
      "entropy": 1.1573342241346836,
      "epoch": 0.39585809833435975,
      "grad_norm": 8.625,
      "learning_rate": 4.5840788820406985e-05,
      "loss": 1.212014102935791,
      "mean_token_accuracy": 0.7288065761327743,
      "num_tokens": 27310858.0,
      "step": 13820
    },
    {
      "entropy": 1.0697692852467298,
      "epoch": 0.39614453690044826,
      "grad_norm": 8.375,
      "learning_rate": 4.583448373383159e-05,
      "loss": 1.0947504043579102,
      "mean_token_accuracy": 0.7464196585118771,
      "num_tokens": 27329474.0,
      "step": 13830
    },
    {
      "entropy": 1.0837575756013393,
      "epoch": 0.39643097546653683,
      "grad_norm": 7.5,
      "learning_rate": 4.58281743061275e-05,
      "loss": 1.0747143745422363,
      "mean_token_accuracy": 0.763117165863514,
      "num_tokens": 27347850.0,
      "step": 13840
    },
    {
      "entropy": 1.0639418363571167,
      "epoch": 0.39671741403262534,
      "grad_norm": 8.5,
      "learning_rate": 4.582186053860935e-05,
      "loss": 1.0765370368957519,
      "mean_token_accuracy": 0.7533428136259317,
      "num_tokens": 27366838.0,
      "step": 13850
    },
    {
      "entropy": 1.1670151762664318,
      "epoch": 0.3970038525987139,
      "grad_norm": 11.5,
      "learning_rate": 4.5815542432592706e-05,
      "loss": 1.2374978065490723,
      "mean_token_accuracy": 0.725564244389534,
      "num_tokens": 27385997.0,
      "step": 13860
    },
    {
      "entropy": 1.1199742000550033,
      "epoch": 0.3972902911648024,
      "grad_norm": 8.5625,
      "learning_rate": 4.5809219989394026e-05,
      "loss": 1.1063859939575196,
      "mean_token_accuracy": 0.7471762392669916,
      "num_tokens": 27405043.0,
      "step": 13870
    },
    {
      "entropy": 1.150625254958868,
      "epoch": 0.397576729730891,
      "grad_norm": 8.9375,
      "learning_rate": 4.5802893210330663e-05,
      "loss": 1.1999119758605956,
      "mean_token_accuracy": 0.728228648006916,
      "num_tokens": 27424573.0,
      "step": 13880
    },
    {
      "entropy": 1.1585307117551564,
      "epoch": 0.3978631682969795,
      "grad_norm": 12.5,
      "learning_rate": 4.5796562096720894e-05,
      "loss": 1.1606752395629882,
      "mean_token_accuracy": 0.7357274513691664,
      "num_tokens": 27444115.0,
      "step": 13890
    },
    {
      "entropy": 1.1166127640753984,
      "epoch": 0.39814960686306805,
      "grad_norm": 7.25,
      "learning_rate": 4.579022664988387e-05,
      "loss": 1.1174097061157227,
      "mean_token_accuracy": 0.7531223516911268,
      "num_tokens": 27463671.0,
      "step": 13900
    },
    {
      "entropy": 1.1190488960593938,
      "epoch": 0.39843604542915656,
      "grad_norm": 8.0625,
      "learning_rate": 4.578388687113967e-05,
      "loss": 1.1178046226501466,
      "mean_token_accuracy": 0.7439228083938361,
      "num_tokens": 27483135.0,
      "step": 13910
    },
    {
      "entropy": 1.1120413860306144,
      "epoch": 0.3987224839952451,
      "grad_norm": 11.25,
      "learning_rate": 4.5777542761809264e-05,
      "loss": 1.1160276412963868,
      "mean_token_accuracy": 0.7362407274544239,
      "num_tokens": 27504051.0,
      "step": 13920
    },
    {
      "entropy": 1.1687403585761786,
      "epoch": 0.39900892256133363,
      "grad_norm": 11.3125,
      "learning_rate": 4.577119432321454e-05,
      "loss": 1.2560688972473144,
      "mean_token_accuracy": 0.7249983888119459,
      "num_tokens": 27523505.0,
      "step": 13930
    },
    {
      "entropy": 1.1077825777232646,
      "epoch": 0.3992953611274222,
      "grad_norm": 8.625,
      "learning_rate": 4.5764841556678264e-05,
      "loss": 1.1104825019836426,
      "mean_token_accuracy": 0.7425674747675657,
      "num_tokens": 27543022.0,
      "step": 13940
    },
    {
      "entropy": 1.0267994746565818,
      "epoch": 0.3995817996935107,
      "grad_norm": 6.21875,
      "learning_rate": 4.575848446352412e-05,
      "loss": 1.0367024421691895,
      "mean_token_accuracy": 0.7526102483272552,
      "num_tokens": 27564097.0,
      "step": 13950
    },
    {
      "entropy": 1.0943633997812867,
      "epoch": 0.3998682382595993,
      "grad_norm": 7.15625,
      "learning_rate": 4.5752123045076676e-05,
      "loss": 1.142641830444336,
      "mean_token_accuracy": 0.7388008993119002,
      "num_tokens": 27583009.0,
      "step": 13960
    },
    {
      "entropy": 1.1412461327388883,
      "epoch": 0.40015467682568784,
      "grad_norm": 7.46875,
      "learning_rate": 4.574575730266144e-05,
      "loss": 1.1726662635803222,
      "mean_token_accuracy": 0.7380145356059075,
      "num_tokens": 27602412.0,
      "step": 13970
    },
    {
      "entropy": 1.0107852390035987,
      "epoch": 0.40044111539177635,
      "grad_norm": 7.375,
      "learning_rate": 4.573938723760477e-05,
      "loss": 1.0369004249572753,
      "mean_token_accuracy": 0.7429001055657863,
      "num_tokens": 27623039.0,
      "step": 13980
    },
    {
      "entropy": 1.0852725083008408,
      "epoch": 0.4007275539578649,
      "grad_norm": 9.1875,
      "learning_rate": 4.573301285123397e-05,
      "loss": 1.1279693603515626,
      "mean_token_accuracy": 0.7422417225316167,
      "num_tokens": 27641362.0,
      "step": 13990
    },
    {
      "entropy": 1.1904946736991406,
      "epoch": 0.4010139925239534,
      "grad_norm": 8.8125,
      "learning_rate": 4.57266341448772e-05,
      "loss": 1.1861407279968261,
      "mean_token_accuracy": 0.7369790241122246,
      "num_tokens": 27660854.0,
      "step": 14000
    },
    {
      "entropy": 1.0397769512608648,
      "epoch": 0.401300431090042,
      "grad_norm": 6.8125,
      "learning_rate": 4.572025111986357e-05,
      "loss": 1.0618779182434082,
      "mean_token_accuracy": 0.7596565578132868,
      "num_tokens": 27679227.0,
      "step": 14010
    },
    {
      "entropy": 1.1395037800073624,
      "epoch": 0.4015868696561305,
      "grad_norm": 9.9375,
      "learning_rate": 4.571386377752305e-05,
      "loss": 1.1624042510986328,
      "mean_token_accuracy": 0.7376817010343075,
      "num_tokens": 27698204.0,
      "step": 14020
    },
    {
      "entropy": 1.286335051059723,
      "epoch": 0.40187330822221906,
      "grad_norm": 7.53125,
      "learning_rate": 4.570747211918652e-05,
      "loss": 1.2868769645690918,
      "mean_token_accuracy": 0.7188835613429546,
      "num_tokens": 27718723.0,
      "step": 14030
    },
    {
      "entropy": 1.2616499349474908,
      "epoch": 0.40215974678830757,
      "grad_norm": 8.125,
      "learning_rate": 4.5701076146185774e-05,
      "loss": 1.2330806732177735,
      "mean_token_accuracy": 0.7266953002661467,
      "num_tokens": 27737772.0,
      "step": 14040
    },
    {
      "entropy": 1.1240324717015029,
      "epoch": 0.40244618535439614,
      "grad_norm": 7.15625,
      "learning_rate": 4.569467585985349e-05,
      "loss": 1.1368300437927246,
      "mean_token_accuracy": 0.7383144348859787,
      "num_tokens": 27757844.0,
      "step": 14050
    },
    {
      "entropy": 1.1414099846035242,
      "epoch": 0.40273262392048464,
      "grad_norm": 8.0,
      "learning_rate": 4.568827126152325e-05,
      "loss": 1.1943794250488282,
      "mean_token_accuracy": 0.7240987431257964,
      "num_tokens": 27776796.0,
      "step": 14060
    },
    {
      "entropy": 1.1706598352640867,
      "epoch": 0.4030190624865732,
      "grad_norm": 7.5,
      "learning_rate": 4.5681862352529525e-05,
      "loss": 1.143519973754883,
      "mean_token_accuracy": 0.7322823226451873,
      "num_tokens": 27797353.0,
      "step": 14070
    },
    {
      "entropy": 1.1464680768549442,
      "epoch": 0.4033055010526617,
      "grad_norm": 7.15625,
      "learning_rate": 4.5675449134207695e-05,
      "loss": 1.1763397216796876,
      "mean_token_accuracy": 0.742257084324956,
      "num_tokens": 27816978.0,
      "step": 14080
    },
    {
      "entropy": 1.0622295454144477,
      "epoch": 0.4035919396187503,
      "grad_norm": 6.875,
      "learning_rate": 4.5669031607894056e-05,
      "loss": 0.9949763298034668,
      "mean_token_accuracy": 0.7683422062546015,
      "num_tokens": 27835500.0,
      "step": 14090
    },
    {
      "entropy": 1.1340995959937572,
      "epoch": 0.4038783781848388,
      "grad_norm": 9.5625,
      "learning_rate": 4.566260977492575e-05,
      "loss": 1.1948657989501954,
      "mean_token_accuracy": 0.7219095807522535,
      "num_tokens": 27854983.0,
      "step": 14100
    },
    {
      "entropy": 1.084837320074439,
      "epoch": 0.40416481675092736,
      "grad_norm": 7.25,
      "learning_rate": 4.565618363664086e-05,
      "loss": 1.1277252197265626,
      "mean_token_accuracy": 0.7484170131385326,
      "num_tokens": 27875123.0,
      "step": 14110
    },
    {
      "entropy": 1.1732332149520517,
      "epoch": 0.40445125531701587,
      "grad_norm": 5.84375,
      "learning_rate": 4.5649753194378364e-05,
      "loss": 1.261639404296875,
      "mean_token_accuracy": 0.7230697566643357,
      "num_tokens": 27895586.0,
      "step": 14120
    },
    {
      "entropy": 1.152813283726573,
      "epoch": 0.40473769388310443,
      "grad_norm": 7.40625,
      "learning_rate": 4.5643318449478115e-05,
      "loss": 1.092837429046631,
      "mean_token_accuracy": 0.7515673194080591,
      "num_tokens": 27915034.0,
      "step": 14130
    },
    {
      "entropy": 1.0599627276882528,
      "epoch": 0.40502413244919294,
      "grad_norm": 10.4375,
      "learning_rate": 4.563687940328086e-05,
      "loss": 1.0886774063110352,
      "mean_token_accuracy": 0.752121340110898,
      "num_tokens": 27935056.0,
      "step": 14140
    },
    {
      "entropy": 1.1309829125180841,
      "epoch": 0.4053105710152815,
      "grad_norm": 11.4375,
      "learning_rate": 4.563043605712829e-05,
      "loss": 1.2338820457458497,
      "mean_token_accuracy": 0.7276365384459496,
      "num_tokens": 27956574.0,
      "step": 14150
    },
    {
      "entropy": 1.2226856648921967,
      "epoch": 0.40559700958137,
      "grad_norm": 8.1875,
      "learning_rate": 4.5623988412362936e-05,
      "loss": 1.245493507385254,
      "mean_token_accuracy": 0.7202244870364666,
      "num_tokens": 27977445.0,
      "step": 14160
    },
    {
      "entropy": 1.1402819616720081,
      "epoch": 0.4058834481474586,
      "grad_norm": 7.34375,
      "learning_rate": 4.561753647032824e-05,
      "loss": 1.109068012237549,
      "mean_token_accuracy": 0.756523833796382,
      "num_tokens": 27998152.0,
      "step": 14170
    },
    {
      "entropy": 1.1912747511640192,
      "epoch": 0.4061698867135471,
      "grad_norm": 7.59375,
      "learning_rate": 4.561108023236856e-05,
      "loss": 1.1988101959228517,
      "mean_token_accuracy": 0.7290570970624686,
      "num_tokens": 28018364.0,
      "step": 14180
    },
    {
      "entropy": 1.0169376848265528,
      "epoch": 0.40645632527963566,
      "grad_norm": 7.9375,
      "learning_rate": 4.560461969982913e-05,
      "loss": 0.9945815086364747,
      "mean_token_accuracy": 0.7714498296380043,
      "num_tokens": 28038473.0,
      "step": 14190
    },
    {
      "entropy": 1.1061491932719947,
      "epoch": 0.40674276384572416,
      "grad_norm": 6.0625,
      "learning_rate": 4.5598154874056084e-05,
      "loss": 1.1170211791992188,
      "mean_token_accuracy": 0.7447976555675269,
      "num_tokens": 28058098.0,
      "step": 14200
    },
    {
      "entropy": 1.1419741604477167,
      "epoch": 0.40702920241181273,
      "grad_norm": 8.0,
      "learning_rate": 4.5591685756396456e-05,
      "loss": 1.2156185150146483,
      "mean_token_accuracy": 0.7268393352627754,
      "num_tokens": 28078274.0,
      "step": 14210
    },
    {
      "entropy": 1.153952433448285,
      "epoch": 0.40731564097790124,
      "grad_norm": 10.125,
      "learning_rate": 4.558521234819816e-05,
      "loss": 1.2208589553833007,
      "mean_token_accuracy": 0.7354421325027942,
      "num_tokens": 28097495.0,
      "step": 14220
    },
    {
      "entropy": 1.192527426034212,
      "epoch": 0.4076020795439898,
      "grad_norm": 8.0625,
      "learning_rate": 4.557873465081002e-05,
      "loss": 1.1827711105346679,
      "mean_token_accuracy": 0.7380841430276632,
      "num_tokens": 28116364.0,
      "step": 14230
    },
    {
      "entropy": 1.193804333731532,
      "epoch": 0.40788851811007837,
      "grad_norm": 8.9375,
      "learning_rate": 4.557225266558174e-05,
      "loss": 1.2354117393493653,
      "mean_token_accuracy": 0.7287968829274177,
      "num_tokens": 28135319.0,
      "step": 14240
    },
    {
      "entropy": 1.2366693869233132,
      "epoch": 0.4081749566761669,
      "grad_norm": 8.5,
      "learning_rate": 4.556576639386392e-05,
      "loss": 1.257970428466797,
      "mean_token_accuracy": 0.714319970086217,
      "num_tokens": 28154903.0,
      "step": 14250
    },
    {
      "entropy": 1.1617317341268063,
      "epoch": 0.40846139524225544,
      "grad_norm": 7.5,
      "learning_rate": 4.555927583700806e-05,
      "loss": 1.1627843856811524,
      "mean_token_accuracy": 0.7274182572960853,
      "num_tokens": 28175655.0,
      "step": 14260
    },
    {
      "entropy": 1.1495460934937,
      "epoch": 0.40874783380834395,
      "grad_norm": 7.9375,
      "learning_rate": 4.5552780996366564e-05,
      "loss": 1.1367864608764648,
      "mean_token_accuracy": 0.739438646286726,
      "num_tokens": 28195921.0,
      "step": 14270
    },
    {
      "entropy": 1.0487476052716374,
      "epoch": 0.4090342723744325,
      "grad_norm": 6.34375,
      "learning_rate": 4.554628187329269e-05,
      "loss": 1.050082015991211,
      "mean_token_accuracy": 0.7519128456711769,
      "num_tokens": 28215324.0,
      "step": 14280
    },
    {
      "entropy": 1.2471446372568606,
      "epoch": 0.409320710940521,
      "grad_norm": 10.375,
      "learning_rate": 4.5539778469140637e-05,
      "loss": 1.2902836799621582,
      "mean_token_accuracy": 0.7069186266511679,
      "num_tokens": 28235538.0,
      "step": 14290
    },
    {
      "entropy": 1.0372010607272386,
      "epoch": 0.4096071495066096,
      "grad_norm": 9.0625,
      "learning_rate": 4.553327078526546e-05,
      "loss": 1.0384214401245118,
      "mean_token_accuracy": 0.7600736010819673,
      "num_tokens": 28255648.0,
      "step": 14300
    },
    {
      "entropy": 1.1140098690986633,
      "epoch": 0.4098935880726981,
      "grad_norm": 7.875,
      "learning_rate": 4.5526758823023105e-05,
      "loss": 1.1720968246459962,
      "mean_token_accuracy": 0.739414731413126,
      "num_tokens": 28275709.0,
      "step": 14310
    },
    {
      "entropy": 1.1347706506028772,
      "epoch": 0.41018002663878667,
      "grad_norm": 7.59375,
      "learning_rate": 4.552024258377045e-05,
      "loss": 1.1680035591125488,
      "mean_token_accuracy": 0.7396666649729013,
      "num_tokens": 28295672.0,
      "step": 14320
    },
    {
      "entropy": 1.1031608134508133,
      "epoch": 0.4104664652048752,
      "grad_norm": 6.03125,
      "learning_rate": 4.551372206886522e-05,
      "loss": 1.0621084213256835,
      "mean_token_accuracy": 0.7565102431923151,
      "num_tokens": 28315465.0,
      "step": 14330
    },
    {
      "entropy": 1.1116163678467275,
      "epoch": 0.41075290377096374,
      "grad_norm": 6.625,
      "learning_rate": 4.5507197279666036e-05,
      "loss": 1.1380571365356444,
      "mean_token_accuracy": 0.7422281928360462,
      "num_tokens": 28336517.0,
      "step": 14340
    },
    {
      "entropy": 0.9999041834846139,
      "epoch": 0.41103934233705225,
      "grad_norm": 6.65625,
      "learning_rate": 4.5500668217532435e-05,
      "loss": 1.0338714599609375,
      "mean_token_accuracy": 0.7558140780776739,
      "num_tokens": 28356259.0,
      "step": 14350
    },
    {
      "entropy": 1.2080080317333342,
      "epoch": 0.4113257809031408,
      "grad_norm": 7.71875,
      "learning_rate": 4.5494134883824835e-05,
      "loss": 1.2259339332580566,
      "mean_token_accuracy": 0.7317264955490828,
      "num_tokens": 28376921.0,
      "step": 14360
    },
    {
      "entropy": 1.059236565604806,
      "epoch": 0.4116122194692293,
      "grad_norm": 7.53125,
      "learning_rate": 4.548759727990452e-05,
      "loss": 1.1544683456420899,
      "mean_token_accuracy": 0.7349529288709163,
      "num_tokens": 28395866.0,
      "step": 14370
    },
    {
      "entropy": 1.1545110568404198,
      "epoch": 0.4118986580353179,
      "grad_norm": 7.28125,
      "learning_rate": 4.54810554071337e-05,
      "loss": 1.17848482131958,
      "mean_token_accuracy": 0.7229414105415344,
      "num_tokens": 28416782.0,
      "step": 14380
    },
    {
      "entropy": 1.1624818492680788,
      "epoch": 0.4121850966014064,
      "grad_norm": 9.5,
      "learning_rate": 4.547450926687545e-05,
      "loss": 1.1792043685913085,
      "mean_token_accuracy": 0.7322552625089884,
      "num_tokens": 28436386.0,
      "step": 14390
    },
    {
      "entropy": 1.1268555503338575,
      "epoch": 0.41247153516749496,
      "grad_norm": 8.25,
      "learning_rate": 4.546795886049374e-05,
      "loss": 1.153772735595703,
      "mean_token_accuracy": 0.7431964311748743,
      "num_tokens": 28457064.0,
      "step": 14400
    },
    {
      "entropy": 1.1511641431599855,
      "epoch": 0.41275797373358347,
      "grad_norm": 7.09375,
      "learning_rate": 4.546140418935343e-05,
      "loss": 1.1944644927978516,
      "mean_token_accuracy": 0.734812755137682,
      "num_tokens": 28476986.0,
      "step": 14410
    },
    {
      "entropy": 1.0386159885674715,
      "epoch": 0.41304441229967204,
      "grad_norm": 8.25,
      "learning_rate": 4.5454845254820274e-05,
      "loss": 1.0399702072143555,
      "mean_token_accuracy": 0.7606531362980604,
      "num_tokens": 28495480.0,
      "step": 14420
    },
    {
      "entropy": 1.1865351803600788,
      "epoch": 0.41333085086576055,
      "grad_norm": 9.5,
      "learning_rate": 4.5448282058260905e-05,
      "loss": 1.2295003890991212,
      "mean_token_accuracy": 0.729651914536953,
      "num_tokens": 28514597.0,
      "step": 14430
    },
    {
      "entropy": 1.1568078488111495,
      "epoch": 0.4136172894318491,
      "grad_norm": 6.46875,
      "learning_rate": 4.544171460104285e-05,
      "loss": 1.1574083328247071,
      "mean_token_accuracy": 0.7387333076447249,
      "num_tokens": 28535351.0,
      "step": 14440
    },
    {
      "entropy": 1.1269728902727365,
      "epoch": 0.4139037279979376,
      "grad_norm": 10.3125,
      "learning_rate": 4.5435142884534514e-05,
      "loss": 1.1483331680297852,
      "mean_token_accuracy": 0.7452408958226442,
      "num_tokens": 28554127.0,
      "step": 14450
    },
    {
      "entropy": 1.158478662930429,
      "epoch": 0.4141901665640262,
      "grad_norm": 8.5,
      "learning_rate": 4.542856691010521e-05,
      "loss": 1.132677173614502,
      "mean_token_accuracy": 0.7491106316447258,
      "num_tokens": 28573103.0,
      "step": 14460
    },
    {
      "entropy": 1.161947951465845,
      "epoch": 0.4144766051301147,
      "grad_norm": 8.625,
      "learning_rate": 4.542198667912512e-05,
      "loss": 1.1607860565185546,
      "mean_token_accuracy": 0.7345645684748888,
      "num_tokens": 28592670.0,
      "step": 14470
    },
    {
      "entropy": 1.0801840119063855,
      "epoch": 0.41476304369620326,
      "grad_norm": 6.96875,
      "learning_rate": 4.541540219296532e-05,
      "loss": 1.1090431213378906,
      "mean_token_accuracy": 0.7519734490662813,
      "num_tokens": 28612173.0,
      "step": 14480
    },
    {
      "entropy": 1.1233488850295543,
      "epoch": 0.41504948226229177,
      "grad_norm": 8.0625,
      "learning_rate": 4.540881345299777e-05,
      "loss": 1.1793914794921876,
      "mean_token_accuracy": 0.7267956435680389,
      "num_tokens": 28633160.0,
      "step": 14490
    },
    {
      "entropy": 1.0784925904124976,
      "epoch": 0.41533592082838033,
      "grad_norm": 5.875,
      "learning_rate": 4.5402220460595305e-05,
      "loss": 1.0103962898254395,
      "mean_token_accuracy": 0.7657408118247986,
      "num_tokens": 28652844.0,
      "step": 14500
    },
    {
      "entropy": 1.1368246782571076,
      "epoch": 0.4156223593944689,
      "grad_norm": 8.625,
      "learning_rate": 4.5395623217131684e-05,
      "loss": 1.1693191528320312,
      "mean_token_accuracy": 0.7241356905549765,
      "num_tokens": 28673500.0,
      "step": 14510
    },
    {
      "entropy": 1.0825424958020449,
      "epoch": 0.4159087979605574,
      "grad_norm": 8.4375,
      "learning_rate": 4.538902172398151e-05,
      "loss": 1.0991718292236328,
      "mean_token_accuracy": 0.7465787436813116,
      "num_tokens": 28692481.0,
      "step": 14520
    },
    {
      "entropy": 1.0715698301792145,
      "epoch": 0.416195236526646,
      "grad_norm": 5.375,
      "learning_rate": 4.538241598252029e-05,
      "loss": 1.060258674621582,
      "mean_token_accuracy": 0.7531248986721039,
      "num_tokens": 28713363.0,
      "step": 14530
    },
    {
      "entropy": 1.0631365917623044,
      "epoch": 0.4164816750927345,
      "grad_norm": 7.84375,
      "learning_rate": 4.537580599412441e-05,
      "loss": 1.1247137069702149,
      "mean_token_accuracy": 0.7426707837730646,
      "num_tokens": 28732264.0,
      "step": 14540
    },
    {
      "entropy": 1.0246806401759385,
      "epoch": 0.41676811365882305,
      "grad_norm": 8.4375,
      "learning_rate": 4.536919176017115e-05,
      "loss": 1.046319580078125,
      "mean_token_accuracy": 0.7603113379329443,
      "num_tokens": 28751063.0,
      "step": 14550
    },
    {
      "entropy": 1.1103365659713744,
      "epoch": 0.41705455222491156,
      "grad_norm": 6.875,
      "learning_rate": 4.5362573282038666e-05,
      "loss": 1.1239331245422364,
      "mean_token_accuracy": 0.7415032241493463,
      "num_tokens": 28770650.0,
      "step": 14560
    },
    {
      "entropy": 1.1654425658285619,
      "epoch": 0.4173409907910001,
      "grad_norm": 8.6875,
      "learning_rate": 4.535595056110601e-05,
      "loss": 1.1932677268981933,
      "mean_token_accuracy": 0.7326556798070669,
      "num_tokens": 28790165.0,
      "step": 14570
    },
    {
      "entropy": 1.1261321596801281,
      "epoch": 0.41762742935708863,
      "grad_norm": 8.1875,
      "learning_rate": 4.5349323598753116e-05,
      "loss": 1.155137825012207,
      "mean_token_accuracy": 0.738577152043581,
      "num_tokens": 28809333.0,
      "step": 14580
    },
    {
      "entropy": 1.156882111914456,
      "epoch": 0.4179138679231772,
      "grad_norm": 10.3125,
      "learning_rate": 4.534269239636077e-05,
      "loss": 1.1214960098266602,
      "mean_token_accuracy": 0.7352047309279441,
      "num_tokens": 28829810.0,
      "step": 14590
    },
    {
      "entropy": 1.0867130406200887,
      "epoch": 0.4182003064892657,
      "grad_norm": 6.03125,
      "learning_rate": 4.533605695531068e-05,
      "loss": 1.0952013969421386,
      "mean_token_accuracy": 0.754645861312747,
      "num_tokens": 28851254.0,
      "step": 14600
    },
    {
      "entropy": 1.153727318532765,
      "epoch": 0.41848674505535427,
      "grad_norm": 6.5625,
      "learning_rate": 4.5329417276985425e-05,
      "loss": 1.1840717315673828,
      "mean_token_accuracy": 0.7343263901770115,
      "num_tokens": 28872435.0,
      "step": 14610
    },
    {
      "entropy": 1.0724379697814583,
      "epoch": 0.4187731836214428,
      "grad_norm": 6.625,
      "learning_rate": 4.532277336276847e-05,
      "loss": 1.0653427124023438,
      "mean_token_accuracy": 0.7640966504812241,
      "num_tokens": 28891695.0,
      "step": 14620
    },
    {
      "entropy": 1.0752145254984498,
      "epoch": 0.41905962218753134,
      "grad_norm": 9.3125,
      "learning_rate": 4.531612521404416e-05,
      "loss": 1.0956077575683594,
      "mean_token_accuracy": 0.7625106770545245,
      "num_tokens": 28911596.0,
      "step": 14630
    },
    {
      "entropy": 1.0497026551514863,
      "epoch": 0.41934606075361985,
      "grad_norm": 8.75,
      "learning_rate": 4.530947283219771e-05,
      "loss": 1.1588489532470703,
      "mean_token_accuracy": 0.7413894068449736,
      "num_tokens": 28931752.0,
      "step": 14640
    },
    {
      "entropy": 1.0290535770356655,
      "epoch": 0.4196324993197084,
      "grad_norm": 8.875,
      "learning_rate": 4.530281621861523e-05,
      "loss": 0.9952278137207031,
      "mean_token_accuracy": 0.7689693726599216,
      "num_tokens": 28951601.0,
      "step": 14650
    },
    {
      "entropy": 1.0079265439882874,
      "epoch": 0.41991893788579693,
      "grad_norm": 11.5625,
      "learning_rate": 4.529615537468372e-05,
      "loss": 1.0446131706237793,
      "mean_token_accuracy": 0.766046853736043,
      "num_tokens": 28971185.0,
      "step": 14660
    },
    {
      "entropy": 1.1571363281458615,
      "epoch": 0.4202053764518855,
      "grad_norm": 8.3125,
      "learning_rate": 4.528949030179104e-05,
      "loss": 1.199844741821289,
      "mean_token_accuracy": 0.7387040574103594,
      "num_tokens": 28991216.0,
      "step": 14670
    },
    {
      "entropy": 1.0469587890431284,
      "epoch": 0.420491815017974,
      "grad_norm": 6.28125,
      "learning_rate": 4.5282821001325956e-05,
      "loss": 1.019493579864502,
      "mean_token_accuracy": 0.7620637316256762,
      "num_tokens": 29011357.0,
      "step": 14680
    },
    {
      "entropy": 1.1374478321522474,
      "epoch": 0.42077825358406257,
      "grad_norm": 9.0,
      "learning_rate": 4.527614747467809e-05,
      "loss": 1.1911021232604981,
      "mean_token_accuracy": 0.7331470765173436,
      "num_tokens": 29031085.0,
      "step": 14690
    },
    {
      "entropy": 1.058006624877453,
      "epoch": 0.4210646921501511,
      "grad_norm": 9.625,
      "learning_rate": 4.526946972323795e-05,
      "loss": 1.0812391281127929,
      "mean_token_accuracy": 0.7519583836197853,
      "num_tokens": 29049813.0,
      "step": 14700
    },
    {
      "entropy": 1.1337369695305823,
      "epoch": 0.42135113071623964,
      "grad_norm": 8.6875,
      "learning_rate": 4.5262787748396944e-05,
      "loss": 1.1544131278991698,
      "mean_token_accuracy": 0.7348576031625271,
      "num_tokens": 29069480.0,
      "step": 14710
    },
    {
      "entropy": 1.0956200117245316,
      "epoch": 0.42163756928232815,
      "grad_norm": 7.84375,
      "learning_rate": 4.5256101551547345e-05,
      "loss": 1.134230899810791,
      "mean_token_accuracy": 0.7455510355532169,
      "num_tokens": 29088220.0,
      "step": 14720
    },
    {
      "entropy": 1.2028059044852852,
      "epoch": 0.4219240078484167,
      "grad_norm": 8.25,
      "learning_rate": 4.52494111340823e-05,
      "loss": 1.2144407272338866,
      "mean_token_accuracy": 0.7308700537309051,
      "num_tokens": 29109869.0,
      "step": 14730
    },
    {
      "entropy": 1.1542120784521104,
      "epoch": 0.4222104464145052,
      "grad_norm": 10.6875,
      "learning_rate": 4.524271649739584e-05,
      "loss": 1.1372940063476562,
      "mean_token_accuracy": 0.7420404098927975,
      "num_tokens": 29130188.0,
      "step": 14740
    },
    {
      "entropy": 1.0530661925673486,
      "epoch": 0.4224968849805938,
      "grad_norm": 6.90625,
      "learning_rate": 4.5236017642882894e-05,
      "loss": 1.1334427833557128,
      "mean_token_accuracy": 0.7515544116497039,
      "num_tokens": 29150008.0,
      "step": 14750
    },
    {
      "entropy": 1.1175051979720592,
      "epoch": 0.4227833235466823,
      "grad_norm": 8.375,
      "learning_rate": 4.5229314571939234e-05,
      "loss": 1.108592414855957,
      "mean_token_accuracy": 0.7463943000882864,
      "num_tokens": 29169963.0,
      "step": 14760
    },
    {
      "entropy": 1.0148131728172303,
      "epoch": 0.42306976211277086,
      "grad_norm": 7.09375,
      "learning_rate": 4.522260728596154e-05,
      "loss": 1.062319850921631,
      "mean_token_accuracy": 0.7511447213590146,
      "num_tokens": 29189251.0,
      "step": 14770
    },
    {
      "entropy": 1.0246519651263952,
      "epoch": 0.42335620067885943,
      "grad_norm": 9.375,
      "learning_rate": 4.5215895786347355e-05,
      "loss": 0.9900642395019531,
      "mean_token_accuracy": 0.775403105840087,
      "num_tokens": 29208395.0,
      "step": 14780
    },
    {
      "entropy": 1.0646299868822098,
      "epoch": 0.42364263924494794,
      "grad_norm": 6.59375,
      "learning_rate": 4.520918007449511e-05,
      "loss": 1.092627716064453,
      "mean_token_accuracy": 0.7510326709598303,
      "num_tokens": 29228376.0,
      "step": 14790
    },
    {
      "entropy": 0.9802539642900229,
      "epoch": 0.4239290778110365,
      "grad_norm": 9.3125,
      "learning_rate": 4.52024601518041e-05,
      "loss": 1.0645797729492188,
      "mean_token_accuracy": 0.7547138769179582,
      "num_tokens": 29247899.0,
      "step": 14800
    },
    {
      "entropy": 1.0694452792406082,
      "epoch": 0.424215516377125,
      "grad_norm": 8.9375,
      "learning_rate": 4.5195736019674526e-05,
      "loss": 1.1007006645202637,
      "mean_token_accuracy": 0.7442386582493782,
      "num_tokens": 29267602.0,
      "step": 14810
    },
    {
      "entropy": 1.0794659662991761,
      "epoch": 0.4245019549432136,
      "grad_norm": 7.53125,
      "learning_rate": 4.5189007679507425e-05,
      "loss": 1.1236899375915528,
      "mean_token_accuracy": 0.7348720226436853,
      "num_tokens": 29287942.0,
      "step": 14820
    },
    {
      "entropy": 1.0388847889378667,
      "epoch": 0.4247883935093021,
      "grad_norm": 6.65625,
      "learning_rate": 4.5182275132704734e-05,
      "loss": 1.0027976989746095,
      "mean_token_accuracy": 0.7654800366610288,
      "num_tokens": 29308119.0,
      "step": 14830
    },
    {
      "entropy": 1.0781768810003995,
      "epoch": 0.42507483207539065,
      "grad_norm": 5.9375,
      "learning_rate": 4.517553838066927e-05,
      "loss": 1.086453914642334,
      "mean_token_accuracy": 0.7575242344290019,
      "num_tokens": 29328206.0,
      "step": 14840
    },
    {
      "entropy": 1.143109680339694,
      "epoch": 0.42536127064147916,
      "grad_norm": 7.0,
      "learning_rate": 4.516879742480472e-05,
      "loss": 1.132544994354248,
      "mean_token_accuracy": 0.7425146870315075,
      "num_tokens": 29347227.0,
      "step": 14850
    },
    {
      "entropy": 1.07180531360209,
      "epoch": 0.4256477092075677,
      "grad_norm": 6.96875,
      "learning_rate": 4.516205226651565e-05,
      "loss": 1.074058723449707,
      "mean_token_accuracy": 0.7563571959733963,
      "num_tokens": 29367095.0,
      "step": 14860
    },
    {
      "entropy": 1.032476819679141,
      "epoch": 0.42593414777365624,
      "grad_norm": 7.59375,
      "learning_rate": 4.5155302907207506e-05,
      "loss": 1.0436034202575684,
      "mean_token_accuracy": 0.7617127224802971,
      "num_tokens": 29386991.0,
      "step": 14870
    },
    {
      "entropy": 1.0833059215918184,
      "epoch": 0.4262205863397448,
      "grad_norm": 8.875,
      "learning_rate": 4.514854934828658e-05,
      "loss": 1.1080145835876465,
      "mean_token_accuracy": 0.73960369117558,
      "num_tokens": 29407421.0,
      "step": 14880
    },
    {
      "entropy": 1.1081771060824395,
      "epoch": 0.4265070249058333,
      "grad_norm": 7.25,
      "learning_rate": 4.514179159116009e-05,
      "loss": 1.1386964797973633,
      "mean_token_accuracy": 0.7572002671658993,
      "num_tokens": 29426083.0,
      "step": 14890
    },
    {
      "entropy": 1.1161217886954546,
      "epoch": 0.4267934634719219,
      "grad_norm": 6.40625,
      "learning_rate": 4.513502963723607e-05,
      "loss": 1.1227438926696778,
      "mean_token_accuracy": 0.7454619942232966,
      "num_tokens": 29446027.0,
      "step": 14900
    },
    {
      "entropy": 0.9543903714045883,
      "epoch": 0.4270799020380104,
      "grad_norm": 6.15625,
      "learning_rate": 4.512826348792348e-05,
      "loss": 0.9860871315002442,
      "mean_token_accuracy": 0.769321646168828,
      "num_tokens": 29466195.0,
      "step": 14910
    },
    {
      "entropy": 1.0968077659606934,
      "epoch": 0.42736634060409895,
      "grad_norm": 8.0,
      "learning_rate": 4.5121493144632135e-05,
      "loss": 1.1625382423400878,
      "mean_token_accuracy": 0.73244525231421,
      "num_tokens": 29486788.0,
      "step": 14920
    },
    {
      "entropy": 1.1832200549542904,
      "epoch": 0.42765277917018746,
      "grad_norm": 7.03125,
      "learning_rate": 4.51147186087727e-05,
      "loss": 1.1919404983520507,
      "mean_token_accuracy": 0.728359953686595,
      "num_tokens": 29507523.0,
      "step": 14930
    },
    {
      "entropy": 1.0534551033750177,
      "epoch": 0.427939217736276,
      "grad_norm": 10.3125,
      "learning_rate": 4.510793988175677e-05,
      "loss": 1.0887598991394043,
      "mean_token_accuracy": 0.743910102546215,
      "num_tokens": 29525157.0,
      "step": 14940
    },
    {
      "entropy": 1.0579757260158658,
      "epoch": 0.42822565630236453,
      "grad_norm": 7.875,
      "learning_rate": 4.510115696499674e-05,
      "loss": 1.0850719451904296,
      "mean_token_accuracy": 0.747008890658617,
      "num_tokens": 29546302.0,
      "step": 14950
    },
    {
      "entropy": 1.1311715003103018,
      "epoch": 0.4285120948684531,
      "grad_norm": 10.3125,
      "learning_rate": 4.509436985990594e-05,
      "loss": 1.0996333122253419,
      "mean_token_accuracy": 0.7537644896656275,
      "num_tokens": 29567035.0,
      "step": 14960
    },
    {
      "entropy": 1.215819674730301,
      "epoch": 0.4287985334345416,
      "grad_norm": 7.4375,
      "learning_rate": 4.508757856789855e-05,
      "loss": 1.2629762649536134,
      "mean_token_accuracy": 0.7203603371977806,
      "num_tokens": 29586620.0,
      "step": 14970
    },
    {
      "entropy": 1.1624621389433742,
      "epoch": 0.42908497200063017,
      "grad_norm": 7.40625,
      "learning_rate": 4.508078309038961e-05,
      "loss": 1.1572059631347655,
      "mean_token_accuracy": 0.7274450678378344,
      "num_tokens": 29607864.0,
      "step": 14980
    },
    {
      "entropy": 1.1626730516552926,
      "epoch": 0.4293714105667187,
      "grad_norm": 6.75,
      "learning_rate": 4.507398342879504e-05,
      "loss": 1.2038818359375,
      "mean_token_accuracy": 0.7258980434387923,
      "num_tokens": 29628504.0,
      "step": 14990
    },
    {
      "entropy": 1.0758032567799092,
      "epoch": 0.42965784913280725,
      "grad_norm": 9.8125,
      "learning_rate": 4.506717958453166e-05,
      "loss": 1.1038790702819825,
      "mean_token_accuracy": 0.7464543301612139,
      "num_tokens": 29647000.0,
      "step": 15000
    },
    {
      "entropy": 1.1332037437707185,
      "epoch": 0.42994428769889576,
      "grad_norm": 9.125,
      "learning_rate": 4.5060371559017114e-05,
      "loss": 1.1060691833496095,
      "mean_token_accuracy": 0.7363989561796188,
      "num_tokens": 29667260.0,
      "step": 15010
    },
    {
      "entropy": 1.10912319496274,
      "epoch": 0.4302307262649843,
      "grad_norm": 6.53125,
      "learning_rate": 4.505355935366996e-05,
      "loss": 1.0906852722167968,
      "mean_token_accuracy": 0.7517189957201481,
      "num_tokens": 29686707.0,
      "step": 15020
    },
    {
      "entropy": 1.2027821194380521,
      "epoch": 0.43051716483107283,
      "grad_norm": 9.0,
      "learning_rate": 4.504674296990958e-05,
      "loss": 1.2429282188415527,
      "mean_token_accuracy": 0.7236607603728771,
      "num_tokens": 29707171.0,
      "step": 15030
    },
    {
      "entropy": 1.0220236087217927,
      "epoch": 0.4308036033971614,
      "grad_norm": 9.0,
      "learning_rate": 4.5039922409156286e-05,
      "loss": 1.06076021194458,
      "mean_token_accuracy": 0.7550806157290936,
      "num_tokens": 29726276.0,
      "step": 15040
    },
    {
      "entropy": 1.2244989149272443,
      "epoch": 0.43109004196324996,
      "grad_norm": 9.375,
      "learning_rate": 4.50330976728312e-05,
      "loss": 1.2870461463928222,
      "mean_token_accuracy": 0.7204246643930674,
      "num_tokens": 29744871.0,
      "step": 15050
    },
    {
      "entropy": 1.2958453392609954,
      "epoch": 0.43137648052933847,
      "grad_norm": 7.125,
      "learning_rate": 4.502626876235637e-05,
      "loss": 1.2524999618530273,
      "mean_token_accuracy": 0.7216910842806101,
      "num_tokens": 29764048.0,
      "step": 15060
    },
    {
      "entropy": 1.0974758636206388,
      "epoch": 0.43166291909542703,
      "grad_norm": 7.8125,
      "learning_rate": 4.5019435679154656e-05,
      "loss": 1.1077496528625488,
      "mean_token_accuracy": 0.7481275469064712,
      "num_tokens": 29782919.0,
      "step": 15070
    },
    {
      "entropy": 1.121358971297741,
      "epoch": 0.43194935766151554,
      "grad_norm": 7.65625,
      "learning_rate": 4.501259842464984e-05,
      "loss": 1.1013365745544434,
      "mean_token_accuracy": 0.7497428376227617,
      "num_tokens": 29802801.0,
      "step": 15080
    },
    {
      "entropy": 0.9602189633995295,
      "epoch": 0.4322357962276041,
      "grad_norm": 5.96875,
      "learning_rate": 4.500575700026655e-05,
      "loss": 0.9573726654052734,
      "mean_token_accuracy": 0.7794460028409957,
      "num_tokens": 29822127.0,
      "step": 15090
    },
    {
      "entropy": 1.0065461289137603,
      "epoch": 0.4325222347936926,
      "grad_norm": 10.625,
      "learning_rate": 4.499891140743026e-05,
      "loss": 1.0011902809143067,
      "mean_token_accuracy": 0.7781655717641115,
      "num_tokens": 29841381.0,
      "step": 15100
    },
    {
      "entropy": 1.0659026734530925,
      "epoch": 0.4328086733597812,
      "grad_norm": 7.90625,
      "learning_rate": 4.499206164756736e-05,
      "loss": 1.1443984031677246,
      "mean_token_accuracy": 0.747183459252119,
      "num_tokens": 29861798.0,
      "step": 15110
    },
    {
      "entropy": 1.0250905063003302,
      "epoch": 0.4330951119258697,
      "grad_norm": 7.21875,
      "learning_rate": 4.4985207722105084e-05,
      "loss": 1.0135614395141601,
      "mean_token_accuracy": 0.7757460709661246,
      "num_tokens": 29881747.0,
      "step": 15120
    },
    {
      "entropy": 1.1286430770531297,
      "epoch": 0.43338155049195826,
      "grad_norm": 7.71875,
      "learning_rate": 4.4978349632471516e-05,
      "loss": 1.2127219200134278,
      "mean_token_accuracy": 0.7387678675353527,
      "num_tokens": 29901623.0,
      "step": 15130
    },
    {
      "entropy": 1.1300976606085897,
      "epoch": 0.43366798905804677,
      "grad_norm": 6.34375,
      "learning_rate": 4.497148738009563e-05,
      "loss": 1.188997459411621,
      "mean_token_accuracy": 0.7400190435349941,
      "num_tokens": 29921964.0,
      "step": 15140
    },
    {
      "entropy": 1.1537324644625186,
      "epoch": 0.43395442762413533,
      "grad_norm": 6.15625,
      "learning_rate": 4.496462096640729e-05,
      "loss": 1.1386261940002442,
      "mean_token_accuracy": 0.7381569374352693,
      "num_tokens": 29942664.0,
      "step": 15150
    },
    {
      "entropy": 1.0698983315378428,
      "epoch": 0.43424086619022384,
      "grad_norm": 8.125,
      "learning_rate": 4.495775039283717e-05,
      "loss": 1.1350852012634278,
      "mean_token_accuracy": 0.7508077036589385,
      "num_tokens": 29960934.0,
      "step": 15160
    },
    {
      "entropy": 1.1243393931537866,
      "epoch": 0.4345273047563124,
      "grad_norm": 7.4375,
      "learning_rate": 4.495087566081685e-05,
      "loss": 1.1545008659362792,
      "mean_token_accuracy": 0.7400742474943399,
      "num_tokens": 29980295.0,
      "step": 15170
    },
    {
      "entropy": 1.2673492718487978,
      "epoch": 0.4348137433224009,
      "grad_norm": 8.25,
      "learning_rate": 4.494399677177877e-05,
      "loss": 1.2707715034484863,
      "mean_token_accuracy": 0.7222472470253706,
      "num_tokens": 30000091.0,
      "step": 15180
    },
    {
      "entropy": 1.1011733392253518,
      "epoch": 0.4351001818884895,
      "grad_norm": 8.875,
      "learning_rate": 4.493711372715624e-05,
      "loss": 1.0820725440979004,
      "mean_token_accuracy": 0.7496522646397352,
      "num_tokens": 30019945.0,
      "step": 15190
    },
    {
      "entropy": 1.157320410385728,
      "epoch": 0.435386620454578,
      "grad_norm": 7.5,
      "learning_rate": 4.493022652838341e-05,
      "loss": 1.1891460418701172,
      "mean_token_accuracy": 0.7327429216355086,
      "num_tokens": 30039359.0,
      "step": 15200
    },
    {
      "entropy": 1.1234152719378472,
      "epoch": 0.43567305902066655,
      "grad_norm": 9.3125,
      "learning_rate": 4.492333517689533e-05,
      "loss": 1.1460776329040527,
      "mean_token_accuracy": 0.7314617995172739,
      "num_tokens": 30059358.0,
      "step": 15210
    },
    {
      "entropy": 1.1859248410910368,
      "epoch": 0.43595949758675506,
      "grad_norm": 8.625,
      "learning_rate": 4.491643967412791e-05,
      "loss": 1.2042438507080078,
      "mean_token_accuracy": 0.7284146931022406,
      "num_tokens": 30079940.0,
      "step": 15220
    },
    {
      "entropy": 1.0426985777914524,
      "epoch": 0.4362459361528436,
      "grad_norm": 7.90625,
      "learning_rate": 4.490954002151789e-05,
      "loss": 1.0594447135925293,
      "mean_token_accuracy": 0.7542109236121177,
      "num_tokens": 30100145.0,
      "step": 15230
    },
    {
      "entropy": 1.0687475807964801,
      "epoch": 0.43653237471893214,
      "grad_norm": 10.125,
      "learning_rate": 4.490263622050292e-05,
      "loss": 1.07454252243042,
      "mean_token_accuracy": 0.7627817682921887,
      "num_tokens": 30120956.0,
      "step": 15240
    },
    {
      "entropy": 1.0825959891080856,
      "epoch": 0.4368188132850207,
      "grad_norm": 7.78125,
      "learning_rate": 4.489572827252148e-05,
      "loss": 1.1304000854492187,
      "mean_token_accuracy": 0.744031447917223,
      "num_tokens": 30140803.0,
      "step": 15250
    },
    {
      "entropy": 1.0698766924440861,
      "epoch": 0.4371052518511092,
      "grad_norm": 7.71875,
      "learning_rate": 4.4888816179012946e-05,
      "loss": 1.0154847145080566,
      "mean_token_accuracy": 0.7579271428287029,
      "num_tokens": 30159934.0,
      "step": 15260
    },
    {
      "entropy": 1.0817425351589918,
      "epoch": 0.4373916904171978,
      "grad_norm": 10.625,
      "learning_rate": 4.488189994141753e-05,
      "loss": 1.1247989654541015,
      "mean_token_accuracy": 0.7458157822489738,
      "num_tokens": 30179717.0,
      "step": 15270
    },
    {
      "entropy": 1.1579272992908956,
      "epoch": 0.4376781289832863,
      "grad_norm": 8.75,
      "learning_rate": 4.487497956117631e-05,
      "loss": 1.2569382667541504,
      "mean_token_accuracy": 0.7285999849438667,
      "num_tokens": 30199214.0,
      "step": 15280
    },
    {
      "entropy": 1.2142420245334506,
      "epoch": 0.43796456754937485,
      "grad_norm": 8.9375,
      "learning_rate": 4.486805503973124e-05,
      "loss": 1.3142725944519043,
      "mean_token_accuracy": 0.722496236115694,
      "num_tokens": 30217613.0,
      "step": 15290
    },
    {
      "entropy": 1.1038442328572273,
      "epoch": 0.43825100611546336,
      "grad_norm": 6.75,
      "learning_rate": 4.486112637852514e-05,
      "loss": 1.0534855842590332,
      "mean_token_accuracy": 0.7518042106181383,
      "num_tokens": 30237475.0,
      "step": 15300
    },
    {
      "entropy": 1.0870592888444661,
      "epoch": 0.4385374446815519,
      "grad_norm": 8.5625,
      "learning_rate": 4.4854193579001686e-05,
      "loss": 1.0699615478515625,
      "mean_token_accuracy": 0.7499058477580547,
      "num_tokens": 30258049.0,
      "step": 15310
    },
    {
      "entropy": 1.048794506303966,
      "epoch": 0.4388238832476405,
      "grad_norm": 10.125,
      "learning_rate": 4.48472566426054e-05,
      "loss": 1.0799086570739747,
      "mean_token_accuracy": 0.75180993527174,
      "num_tokens": 30277921.0,
      "step": 15320
    },
    {
      "entropy": 1.0912634551525116,
      "epoch": 0.439110321813729,
      "grad_norm": 7.3125,
      "learning_rate": 4.4840315570781696e-05,
      "loss": 1.1454171180725097,
      "mean_token_accuracy": 0.7444148104637861,
      "num_tokens": 30297088.0,
      "step": 15330
    },
    {
      "entropy": 1.0637216202914714,
      "epoch": 0.43939676037981756,
      "grad_norm": 6.9375,
      "learning_rate": 4.4833370364976815e-05,
      "loss": 1.0256340980529786,
      "mean_token_accuracy": 0.7631965607404709,
      "num_tokens": 30316902.0,
      "step": 15340
    },
    {
      "entropy": 1.0344652894884347,
      "epoch": 0.4396831989459061,
      "grad_norm": 8.0625,
      "learning_rate": 4.48264210266379e-05,
      "loss": 1.055041790008545,
      "mean_token_accuracy": 0.7633745796978474,
      "num_tokens": 30336131.0,
      "step": 15350
    },
    {
      "entropy": 1.0186852632090448,
      "epoch": 0.43996963751199464,
      "grad_norm": 8.4375,
      "learning_rate": 4.481946755721292e-05,
      "loss": 0.9774246215820312,
      "mean_token_accuracy": 0.7767384380102158,
      "num_tokens": 30355607.0,
      "step": 15360
    },
    {
      "entropy": 1.1015155369415879,
      "epoch": 0.44025607607808315,
      "grad_norm": 9.5625,
      "learning_rate": 4.481250995815073e-05,
      "loss": 1.1176934242248535,
      "mean_token_accuracy": 0.7486156545579433,
      "num_tokens": 30375645.0,
      "step": 15370
    },
    {
      "entropy": 1.0753528747707606,
      "epoch": 0.4405425146441717,
      "grad_norm": 7.53125,
      "learning_rate": 4.480554823090102e-05,
      "loss": 1.0784658432006835,
      "mean_token_accuracy": 0.7505337476730347,
      "num_tokens": 30395226.0,
      "step": 15380
    },
    {
      "entropy": 1.0957379538565875,
      "epoch": 0.4408289532102602,
      "grad_norm": 8.625,
      "learning_rate": 4.479858237691435e-05,
      "loss": 1.15682373046875,
      "mean_token_accuracy": 0.7315940871834755,
      "num_tokens": 30414779.0,
      "step": 15390
    },
    {
      "entropy": 1.0562709361314773,
      "epoch": 0.4411153917763488,
      "grad_norm": 7.78125,
      "learning_rate": 4.479161239764217e-05,
      "loss": 1.0142365455627442,
      "mean_token_accuracy": 0.762904430180788,
      "num_tokens": 30433302.0,
      "step": 15400
    },
    {
      "entropy": 1.1158408207818866,
      "epoch": 0.4414018303424373,
      "grad_norm": 7.4375,
      "learning_rate": 4.478463829453674e-05,
      "loss": 1.133906078338623,
      "mean_token_accuracy": 0.7498819269239902,
      "num_tokens": 30452800.0,
      "step": 15410
    },
    {
      "entropy": 1.0397257771342994,
      "epoch": 0.44168826890852586,
      "grad_norm": 7.0,
      "learning_rate": 4.477766006905122e-05,
      "loss": 1.065294075012207,
      "mean_token_accuracy": 0.757225688174367,
      "num_tokens": 30473094.0,
      "step": 15420
    },
    {
      "entropy": 1.1724238276481629,
      "epoch": 0.44197470747461437,
      "grad_norm": 8.0625,
      "learning_rate": 4.47706777226396e-05,
      "loss": 1.1826380729675292,
      "mean_token_accuracy": 0.7338803302496671,
      "num_tokens": 30491937.0,
      "step": 15430
    },
    {
      "entropy": 1.1097569966688752,
      "epoch": 0.44226114604070293,
      "grad_norm": 7.90625,
      "learning_rate": 4.4763691256756735e-05,
      "loss": 1.1232279777526855,
      "mean_token_accuracy": 0.7510500632226467,
      "num_tokens": 30511667.0,
      "step": 15440
    },
    {
      "entropy": 1.0903748935088515,
      "epoch": 0.44254758460679144,
      "grad_norm": 9.125,
      "learning_rate": 4.475670067285835e-05,
      "loss": 1.0526710510253907,
      "mean_token_accuracy": 0.7502595346421004,
      "num_tokens": 30533632.0,
      "step": 15450
    },
    {
      "entropy": 1.0366231199353932,
      "epoch": 0.44283402317288,
      "grad_norm": 8.625,
      "learning_rate": 4.474970597240104e-05,
      "loss": 1.1116233825683595,
      "mean_token_accuracy": 0.7352092389017344,
      "num_tokens": 30552860.0,
      "step": 15460
    },
    {
      "entropy": 0.9916391972452402,
      "epoch": 0.4431204617389685,
      "grad_norm": 8.25,
      "learning_rate": 4.47427071568422e-05,
      "loss": 0.9432309150695801,
      "mean_token_accuracy": 0.7816601600497961,
      "num_tokens": 30572287.0,
      "step": 15470
    },
    {
      "entropy": 1.0725778579711913,
      "epoch": 0.4434069003050571,
      "grad_norm": 7.21875,
      "learning_rate": 4.473570422764016e-05,
      "loss": 1.200136947631836,
      "mean_token_accuracy": 0.7336394723504782,
      "num_tokens": 30590746.0,
      "step": 15480
    },
    {
      "entropy": 1.1194538097828626,
      "epoch": 0.4436933388711456,
      "grad_norm": 7.625,
      "learning_rate": 4.472869718625405e-05,
      "loss": 1.1669986724853516,
      "mean_token_accuracy": 0.7356421247124671,
      "num_tokens": 30611022.0,
      "step": 15490
    },
    {
      "entropy": 1.2160000964999198,
      "epoch": 0.44397977743723416,
      "grad_norm": 8.1875,
      "learning_rate": 4.4721686034143875e-05,
      "loss": 1.2622747421264648,
      "mean_token_accuracy": 0.7339814633131028,
      "num_tokens": 30631123.0,
      "step": 15500
    },
    {
      "entropy": 1.0182258801534771,
      "epoch": 0.44426621600332267,
      "grad_norm": 6.96875,
      "learning_rate": 4.471467077277051e-05,
      "loss": 1.0076227188110352,
      "mean_token_accuracy": 0.7585625186562538,
      "num_tokens": 30650365.0,
      "step": 15510
    },
    {
      "entropy": 1.0845182720571755,
      "epoch": 0.44455265456941123,
      "grad_norm": 7.875,
      "learning_rate": 4.470765140359566e-05,
      "loss": 1.0863000869750976,
      "mean_token_accuracy": 0.7452656861394644,
      "num_tokens": 30670054.0,
      "step": 15520
    },
    {
      "entropy": 1.1078708942979574,
      "epoch": 0.44483909313549974,
      "grad_norm": 7.8125,
      "learning_rate": 4.470062792808191e-05,
      "loss": 1.081822395324707,
      "mean_token_accuracy": 0.7502869699150324,
      "num_tokens": 30689560.0,
      "step": 15530
    },
    {
      "entropy": 1.0245083846151828,
      "epoch": 0.4451255317015883,
      "grad_norm": 9.4375,
      "learning_rate": 4.4693600347692675e-05,
      "loss": 1.0078306198120117,
      "mean_token_accuracy": 0.7596205219626426,
      "num_tokens": 30709145.0,
      "step": 15540
    },
    {
      "entropy": 1.0145925169810652,
      "epoch": 0.4454119702676768,
      "grad_norm": 8.75,
      "learning_rate": 4.468656866389226e-05,
      "loss": 1.0326231002807618,
      "mean_token_accuracy": 0.7633627243340015,
      "num_tokens": 30727833.0,
      "step": 15550
    },
    {
      "entropy": 1.0818233244121074,
      "epoch": 0.4456984088337654,
      "grad_norm": 8.25,
      "learning_rate": 4.4679532878145794e-05,
      "loss": 1.1350409507751464,
      "mean_token_accuracy": 0.7313998382538557,
      "num_tokens": 30749073.0,
      "step": 15560
    },
    {
      "entropy": 1.0278082424774766,
      "epoch": 0.4459848473998539,
      "grad_norm": 9.0,
      "learning_rate": 4.467249299191928e-05,
      "loss": 1.0461588859558106,
      "mean_token_accuracy": 0.7661760602146387,
      "num_tokens": 30769654.0,
      "step": 15570
    },
    {
      "entropy": 1.1154366530478002,
      "epoch": 0.44627128596594245,
      "grad_norm": 6.375,
      "learning_rate": 4.466544900667955e-05,
      "loss": 1.1709959030151367,
      "mean_token_accuracy": 0.7346567582339049,
      "num_tokens": 30789471.0,
      "step": 15580
    },
    {
      "entropy": 0.9800055850297212,
      "epoch": 0.446557724532031,
      "grad_norm": 8.375,
      "learning_rate": 4.465840092389433e-05,
      "loss": 0.9925732612609863,
      "mean_token_accuracy": 0.762231070175767,
      "num_tokens": 30808159.0,
      "step": 15590
    },
    {
      "entropy": 1.1259382989257574,
      "epoch": 0.44684416309811953,
      "grad_norm": 7.71875,
      "learning_rate": 4.4651348745032164e-05,
      "loss": 1.2082551956176757,
      "mean_token_accuracy": 0.7326127052307129,
      "num_tokens": 30827978.0,
      "step": 15600
    },
    {
      "entropy": 1.0815007980912923,
      "epoch": 0.4471306016642081,
      "grad_norm": 7.875,
      "learning_rate": 4.464429247156245e-05,
      "loss": 1.103219985961914,
      "mean_token_accuracy": 0.7470836568623781,
      "num_tokens": 30846309.0,
      "step": 15610
    },
    {
      "entropy": 1.1303607407957315,
      "epoch": 0.4474170402302966,
      "grad_norm": 7.34375,
      "learning_rate": 4.463723210495549e-05,
      "loss": 1.0471221923828125,
      "mean_token_accuracy": 0.7606925543397665,
      "num_tokens": 30866289.0,
      "step": 15620
    },
    {
      "entropy": 1.1047875422984361,
      "epoch": 0.44770347879638517,
      "grad_norm": 7.3125,
      "learning_rate": 4.463016764668236e-05,
      "loss": 1.1855402946472169,
      "mean_token_accuracy": 0.7292618233710527,
      "num_tokens": 30886494.0,
      "step": 15630
    },
    {
      "entropy": 0.9306949958205223,
      "epoch": 0.4479899173624737,
      "grad_norm": 6.625,
      "learning_rate": 4.462309909821506e-05,
      "loss": 0.910073184967041,
      "mean_token_accuracy": 0.7794380374252796,
      "num_tokens": 30905209.0,
      "step": 15640
    },
    {
      "entropy": 1.0533215487375855,
      "epoch": 0.44827635592856224,
      "grad_norm": 9.0,
      "learning_rate": 4.461602646102639e-05,
      "loss": 1.0918192863464355,
      "mean_token_accuracy": 0.7565400037914515,
      "num_tokens": 30925848.0,
      "step": 15650
    },
    {
      "entropy": 1.1468591360375284,
      "epoch": 0.44856279449465075,
      "grad_norm": 6.15625,
      "learning_rate": 4.460894973659004e-05,
      "loss": 1.176361083984375,
      "mean_token_accuracy": 0.728950684517622,
      "num_tokens": 30946468.0,
      "step": 15660
    },
    {
      "entropy": 1.0342023156583309,
      "epoch": 0.4488492330607393,
      "grad_norm": 9.125,
      "learning_rate": 4.460186892638053e-05,
      "loss": 1.0815458297729492,
      "mean_token_accuracy": 0.7629012949764729,
      "num_tokens": 30964684.0,
      "step": 15670
    },
    {
      "entropy": 1.10842342749238,
      "epoch": 0.4491356716268278,
      "grad_norm": 9.9375,
      "learning_rate": 4.459478403187323e-05,
      "loss": 1.0877862930297852,
      "mean_token_accuracy": 0.7531939174979925,
      "num_tokens": 30984101.0,
      "step": 15680
    },
    {
      "entropy": 1.1106369998306036,
      "epoch": 0.4494221101929164,
      "grad_norm": 12.6875,
      "learning_rate": 4.4587695054544375e-05,
      "loss": 1.1306655883789063,
      "mean_token_accuracy": 0.7482433594763279,
      "num_tokens": 31004893.0,
      "step": 15690
    },
    {
      "entropy": 1.1490726996213199,
      "epoch": 0.4497085487590049,
      "grad_norm": 6.625,
      "learning_rate": 4.458060199587104e-05,
      "loss": 1.174320888519287,
      "mean_token_accuracy": 0.730491679534316,
      "num_tokens": 31025858.0,
      "step": 15700
    },
    {
      "entropy": 1.1322781743481756,
      "epoch": 0.44999498732509347,
      "grad_norm": 7.125,
      "learning_rate": 4.4573504857331154e-05,
      "loss": 1.1925125122070312,
      "mean_token_accuracy": 0.7250677168369293,
      "num_tokens": 31045091.0,
      "step": 15710
    },
    {
      "entropy": 1.1189708166755736,
      "epoch": 0.450281425891182,
      "grad_norm": 8.3125,
      "learning_rate": 4.45664036404035e-05,
      "loss": 1.1621577262878418,
      "mean_token_accuracy": 0.7379273366183042,
      "num_tokens": 31065992.0,
      "step": 15720
    },
    {
      "entropy": 1.1584448412060737,
      "epoch": 0.45056786445727054,
      "grad_norm": 7.5625,
      "learning_rate": 4.455929834656769e-05,
      "loss": 1.1830493927001953,
      "mean_token_accuracy": 0.7414366465061903,
      "num_tokens": 31085747.0,
      "step": 15730
    },
    {
      "entropy": 0.9814271181821823,
      "epoch": 0.45085430302335905,
      "grad_norm": 8.5,
      "learning_rate": 4.455218897730422e-05,
      "loss": 0.9263852119445801,
      "mean_token_accuracy": 0.776156510040164,
      "num_tokens": 31104876.0,
      "step": 15740
    },
    {
      "entropy": 1.1353449657559396,
      "epoch": 0.4511407415894476,
      "grad_norm": 8.4375,
      "learning_rate": 4.4545075534094416e-05,
      "loss": 1.1473031044006348,
      "mean_token_accuracy": 0.7438568361103535,
      "num_tokens": 31125953.0,
      "step": 15750
    },
    {
      "entropy": 0.9814729139208793,
      "epoch": 0.4514271801555361,
      "grad_norm": 7.84375,
      "learning_rate": 4.4537958018420444e-05,
      "loss": 1.0090397834777831,
      "mean_token_accuracy": 0.7663807831704617,
      "num_tokens": 31145675.0,
      "step": 15760
    },
    {
      "entropy": 1.1864433780312538,
      "epoch": 0.4517136187216247,
      "grad_norm": 8.375,
      "learning_rate": 4.4530836431765344e-05,
      "loss": 1.1939178466796876,
      "mean_token_accuracy": 0.7385008133947849,
      "num_tokens": 31166114.0,
      "step": 15770
    },
    {
      "entropy": 1.1284055178985,
      "epoch": 0.4520000572877132,
      "grad_norm": 6.9375,
      "learning_rate": 4.452371077561297e-05,
      "loss": 1.1470273017883301,
      "mean_token_accuracy": 0.7356167893856764,
      "num_tokens": 31186158.0,
      "step": 15780
    },
    {
      "entropy": 1.109770084451884,
      "epoch": 0.45228649585380176,
      "grad_norm": 7.8125,
      "learning_rate": 4.451658105144805e-05,
      "loss": 1.1252546310424805,
      "mean_token_accuracy": 0.7287276841700077,
      "num_tokens": 31206809.0,
      "step": 15790
    },
    {
      "entropy": 1.016530017927289,
      "epoch": 0.45257293441989027,
      "grad_norm": 6.78125,
      "learning_rate": 4.450944726075616e-05,
      "loss": 0.9956744194030762,
      "mean_token_accuracy": 0.7646635144948959,
      "num_tokens": 31226739.0,
      "step": 15800
    },
    {
      "entropy": 1.155272276699543,
      "epoch": 0.45285937298597884,
      "grad_norm": 13.125,
      "learning_rate": 4.450230940502371e-05,
      "loss": 1.173872184753418,
      "mean_token_accuracy": 0.7407504335045815,
      "num_tokens": 31247238.0,
      "step": 15810
    },
    {
      "entropy": 1.123815843835473,
      "epoch": 0.45314581155206735,
      "grad_norm": 7.84375,
      "learning_rate": 4.449516748573796e-05,
      "loss": 1.1115382194519043,
      "mean_token_accuracy": 0.7464274741709233,
      "num_tokens": 31268326.0,
      "step": 15820
    },
    {
      "entropy": 1.0370233446359634,
      "epoch": 0.4534322501181559,
      "grad_norm": 8.25,
      "learning_rate": 4.448802150438702e-05,
      "loss": 1.0724099159240723,
      "mean_token_accuracy": 0.7524797070771456,
      "num_tokens": 31288438.0,
      "step": 15830
    },
    {
      "entropy": 1.1587765246629715,
      "epoch": 0.4537186886842444,
      "grad_norm": 9.375,
      "learning_rate": 4.448087146245985e-05,
      "loss": 1.202589225769043,
      "mean_token_accuracy": 0.7304346699267625,
      "num_tokens": 31308882.0,
      "step": 15840
    },
    {
      "entropy": 1.0659090556204318,
      "epoch": 0.454005127250333,
      "grad_norm": 8.25,
      "learning_rate": 4.447371736144625e-05,
      "loss": 1.0883872985839844,
      "mean_token_accuracy": 0.7527126614004374,
      "num_tokens": 31328256.0,
      "step": 15850
    },
    {
      "entropy": 1.036961019784212,
      "epoch": 0.45429156581642155,
      "grad_norm": 7.96875,
      "learning_rate": 4.4466559202836866e-05,
      "loss": 0.9919363021850586,
      "mean_token_accuracy": 0.7785925008356571,
      "num_tokens": 31346630.0,
      "step": 15860
    },
    {
      "entropy": 1.1220204528421163,
      "epoch": 0.45457800438251006,
      "grad_norm": 8.8125,
      "learning_rate": 4.44593969881232e-05,
      "loss": 1.16156005859375,
      "mean_token_accuracy": 0.7372601218521595,
      "num_tokens": 31368320.0,
      "step": 15870
    },
    {
      "entropy": 1.1004175113514065,
      "epoch": 0.4548644429485986,
      "grad_norm": 7.40625,
      "learning_rate": 4.445223071879757e-05,
      "loss": 1.2071547508239746,
      "mean_token_accuracy": 0.7314198646694422,
      "num_tokens": 31389337.0,
      "step": 15880
    },
    {
      "entropy": 1.0097388913854957,
      "epoch": 0.45515088151468713,
      "grad_norm": 6.125,
      "learning_rate": 4.444506039635318e-05,
      "loss": 0.9778777122497558,
      "mean_token_accuracy": 0.7644600432366133,
      "num_tokens": 31409687.0,
      "step": 15890
    },
    {
      "entropy": 1.1191555349156261,
      "epoch": 0.4554373200807757,
      "grad_norm": 10.5,
      "learning_rate": 4.4437886022284046e-05,
      "loss": 1.1028203010559081,
      "mean_token_accuracy": 0.7389698110520839,
      "num_tokens": 31428518.0,
      "step": 15900
    },
    {
      "entropy": 1.1299670485779643,
      "epoch": 0.4557237586468642,
      "grad_norm": 7.9375,
      "learning_rate": 4.4430707598085045e-05,
      "loss": 1.1646889686584472,
      "mean_token_accuracy": 0.7385070286691189,
      "num_tokens": 31447881.0,
      "step": 15910
    },
    {
      "entropy": 1.084318339638412,
      "epoch": 0.4560101972129528,
      "grad_norm": 9.4375,
      "learning_rate": 4.442352512525189e-05,
      "loss": 1.128535556793213,
      "mean_token_accuracy": 0.7514117076992989,
      "num_tokens": 31467296.0,
      "step": 15920
    },
    {
      "entropy": 1.0100587390363216,
      "epoch": 0.4562966357790413,
      "grad_norm": 8.6875,
      "learning_rate": 4.4416338605281126e-05,
      "loss": 1.0301907539367676,
      "mean_token_accuracy": 0.7658644881099462,
      "num_tokens": 31487179.0,
      "step": 15930
    },
    {
      "entropy": 1.0156860490329565,
      "epoch": 0.45658307434512985,
      "grad_norm": 6.21875,
      "learning_rate": 4.440914803967019e-05,
      "loss": 1.0398475646972656,
      "mean_token_accuracy": 0.7669496171176433,
      "num_tokens": 31506289.0,
      "step": 15940
    },
    {
      "entropy": 1.02811189442873,
      "epoch": 0.45686951291121836,
      "grad_norm": 6.21875,
      "learning_rate": 4.4401953429917296e-05,
      "loss": 1.0573338508605956,
      "mean_token_accuracy": 0.7537104487419128,
      "num_tokens": 31525665.0,
      "step": 15950
    },
    {
      "entropy": 1.1318390472792088,
      "epoch": 0.4571559514773069,
      "grad_norm": 8.875,
      "learning_rate": 4.439475477752155e-05,
      "loss": 1.1844322204589843,
      "mean_token_accuracy": 0.7344266023486853,
      "num_tokens": 31544162.0,
      "step": 15960
    },
    {
      "entropy": 1.0630007477477192,
      "epoch": 0.45744239004339543,
      "grad_norm": 8.375,
      "learning_rate": 4.438755208398287e-05,
      "loss": 1.1088850021362304,
      "mean_token_accuracy": 0.753275141865015,
      "num_tokens": 31563714.0,
      "step": 15970
    },
    {
      "entropy": 1.0772043531760573,
      "epoch": 0.457728828609484,
      "grad_norm": 7.65625,
      "learning_rate": 4.4380345350802045e-05,
      "loss": 1.0706336975097657,
      "mean_token_accuracy": 0.7685126382857561,
      "num_tokens": 31582692.0,
      "step": 15980
    },
    {
      "entropy": 1.0212453888729214,
      "epoch": 0.4580152671755725,
      "grad_norm": 7.34375,
      "learning_rate": 4.437313457948067e-05,
      "loss": 1.028015422821045,
      "mean_token_accuracy": 0.76559530608356,
      "num_tokens": 31602040.0,
      "step": 15990
    },
    {
      "epoch": 0.45830170574166107,
      "eval_entropy": 1.095625594496727,
      "eval_loss": 1.1043692827224731,
      "eval_mean_token_accuracy": 0.7455817241668701,
      "eval_num_tokens": 31621760.0,
      "eval_runtime": 43.2042,
      "eval_samples_per_second": 46.292,
      "eval_steps_per_second": 5.786,
      "step": 16000
    },
    {
      "entropy": 0.9860195256769657,
      "epoch": 0.4585881443077496,
      "grad_norm": 8.0,
      "learning_rate": 4.4358700928426965e-05,
      "loss": 0.9887205123901367,
      "mean_token_accuracy": 0.7684080444276333,
      "num_tokens": 31641232.0,
      "step": 16010
    },
    {
      "entropy": 1.155025776475668,
      "epoch": 0.45887458287383814,
      "grad_norm": 7.75,
      "learning_rate": 4.435147805170207e-05,
      "loss": 1.189784622192383,
      "mean_token_accuracy": 0.7314405463635921,
      "num_tokens": 31661901.0,
      "step": 16020
    },
    {
      "entropy": 1.1773955814540387,
      "epoch": 0.45916102143992665,
      "grad_norm": 7.3125,
      "learning_rate": 4.434425114285149e-05,
      "loss": 1.2892212867736816,
      "mean_token_accuracy": 0.712294165790081,
      "num_tokens": 31682431.0,
      "step": 16030
    },
    {
      "entropy": 1.0590716701000928,
      "epoch": 0.4594474600060152,
      "grad_norm": 8.6875,
      "learning_rate": 4.433702020338106e-05,
      "loss": 0.9952661514282226,
      "mean_token_accuracy": 0.773939310759306,
      "num_tokens": 31701945.0,
      "step": 16040
    },
    {
      "entropy": 1.0923929389566183,
      "epoch": 0.4597338985721037,
      "grad_norm": 5.25,
      "learning_rate": 4.432978523479744e-05,
      "loss": 1.0685754776000977,
      "mean_token_accuracy": 0.7541909072548151,
      "num_tokens": 31722788.0,
      "step": 16050
    },
    {
      "entropy": 1.1169374611228704,
      "epoch": 0.4600203371381923,
      "grad_norm": 10.875,
      "learning_rate": 4.432254623860811e-05,
      "loss": 1.1634275436401367,
      "mean_token_accuracy": 0.7388406038284302,
      "num_tokens": 31743735.0,
      "step": 16060
    },
    {
      "entropy": 1.042324186861515,
      "epoch": 0.4603067757042808,
      "grad_norm": 7.625,
      "learning_rate": 4.431530321632143e-05,
      "loss": 1.0316386222839355,
      "mean_token_accuracy": 0.7618239607661963,
      "num_tokens": 31763012.0,
      "step": 16070
    },
    {
      "entropy": 1.0105106366798282,
      "epoch": 0.46059321427036937,
      "grad_norm": 9.3125,
      "learning_rate": 4.4308056169446546e-05,
      "loss": 1.0378657341003419,
      "mean_token_accuracy": 0.759275808185339,
      "num_tokens": 31783147.0,
      "step": 16080
    },
    {
      "entropy": 1.0123815651983024,
      "epoch": 0.4608796528364579,
      "grad_norm": 10.3125,
      "learning_rate": 4.4300805099493494e-05,
      "loss": 1.0375326156616211,
      "mean_token_accuracy": 0.7635608714073896,
      "num_tokens": 31803969.0,
      "step": 16090
    },
    {
      "entropy": 1.0700342457741499,
      "epoch": 0.46116609140254644,
      "grad_norm": 6.6875,
      "learning_rate": 4.4293550007973117e-05,
      "loss": 1.154991626739502,
      "mean_token_accuracy": 0.7375030998140574,
      "num_tokens": 31823862.0,
      "step": 16100
    },
    {
      "entropy": 1.1746778898872434,
      "epoch": 0.46145252996863495,
      "grad_norm": 8.125,
      "learning_rate": 4.428629089639711e-05,
      "loss": 1.2552621841430665,
      "mean_token_accuracy": 0.7300440713763237,
      "num_tokens": 31844345.0,
      "step": 16110
    },
    {
      "entropy": 1.063766848295927,
      "epoch": 0.4617389685347235,
      "grad_norm": 6.3125,
      "learning_rate": 4.4279027766278e-05,
      "loss": 0.9663609504699707,
      "mean_token_accuracy": 0.7593142136931419,
      "num_tokens": 31863625.0,
      "step": 16120
    },
    {
      "entropy": 0.9916869774460793,
      "epoch": 0.4620254071008121,
      "grad_norm": 9.6875,
      "learning_rate": 4.4271760619129154e-05,
      "loss": 1.0022812843322755,
      "mean_token_accuracy": 0.7718964867293835,
      "num_tokens": 31884866.0,
      "step": 16130
    },
    {
      "entropy": 1.1031945832073689,
      "epoch": 0.4623118456669006,
      "grad_norm": 8.75,
      "learning_rate": 4.426448945646477e-05,
      "loss": 1.1073156356811524,
      "mean_token_accuracy": 0.7422095578163862,
      "num_tokens": 31904002.0,
      "step": 16140
    },
    {
      "entropy": 1.060465625859797,
      "epoch": 0.46259828423298915,
      "grad_norm": 13.4375,
      "learning_rate": 4.4257214279799884e-05,
      "loss": 1.0905183792114257,
      "mean_token_accuracy": 0.7510077152401209,
      "num_tokens": 31922522.0,
      "step": 16150
    },
    {
      "entropy": 1.032481526583433,
      "epoch": 0.46288472279907766,
      "grad_norm": 7.78125,
      "learning_rate": 4.424993509065038e-05,
      "loss": 1.056278896331787,
      "mean_token_accuracy": 0.760082607343793,
      "num_tokens": 31944020.0,
      "step": 16160
    },
    {
      "entropy": 1.0038289047777653,
      "epoch": 0.46317116136516623,
      "grad_norm": 7.5625,
      "learning_rate": 4.424265189053296e-05,
      "loss": 0.9994734764099121,
      "mean_token_accuracy": 0.7648904416710138,
      "num_tokens": 31963617.0,
      "step": 16170
    },
    {
      "entropy": 1.1390948429703713,
      "epoch": 0.46345759993125474,
      "grad_norm": 9.0,
      "learning_rate": 4.423536468096519e-05,
      "loss": 1.1763582229614258,
      "mean_token_accuracy": 0.7386130318045616,
      "num_tokens": 31983962.0,
      "step": 16180
    },
    {
      "entropy": 1.1356624189764262,
      "epoch": 0.4637440384973433,
      "grad_norm": 8.75,
      "learning_rate": 4.4228073463465424e-05,
      "loss": 1.138768482208252,
      "mean_token_accuracy": 0.7386961799114943,
      "num_tokens": 32002915.0,
      "step": 16190
    },
    {
      "entropy": 1.1501975405961276,
      "epoch": 0.4640304770634318,
      "grad_norm": 7.5625,
      "learning_rate": 4.422077823955289e-05,
      "loss": 1.1882719039916991,
      "mean_token_accuracy": 0.7252085719257593,
      "num_tokens": 32022696.0,
      "step": 16200
    },
    {
      "entropy": 1.1720309544354677,
      "epoch": 0.4643169156295204,
      "grad_norm": 7.21875,
      "learning_rate": 4.421347901074766e-05,
      "loss": 1.2030263900756837,
      "mean_token_accuracy": 0.7324550963938237,
      "num_tokens": 32042819.0,
      "step": 16210
    },
    {
      "entropy": 1.065559115819633,
      "epoch": 0.4646033541956089,
      "grad_norm": 7.71875,
      "learning_rate": 4.42061757785706e-05,
      "loss": 1.059134578704834,
      "mean_token_accuracy": 0.769792552664876,
      "num_tokens": 32062401.0,
      "step": 16220
    },
    {
      "entropy": 1.0284819804131984,
      "epoch": 0.46488979276169745,
      "grad_norm": 5.875,
      "learning_rate": 4.419886854454343e-05,
      "loss": 0.993977165222168,
      "mean_token_accuracy": 0.7713834125548601,
      "num_tokens": 32080732.0,
      "step": 16230
    },
    {
      "entropy": 1.0163395173847676,
      "epoch": 0.46517623132778596,
      "grad_norm": 8.3125,
      "learning_rate": 4.419155731018872e-05,
      "loss": 1.0098822593688965,
      "mean_token_accuracy": 0.7723608791828156,
      "num_tokens": 32100183.0,
      "step": 16240
    },
    {
      "entropy": 1.016899237036705,
      "epoch": 0.4654626698938745,
      "grad_norm": 9.25,
      "learning_rate": 4.4184242077029845e-05,
      "loss": 1.0311381340026855,
      "mean_token_accuracy": 0.7663547594100237,
      "num_tokens": 32121051.0,
      "step": 16250
    },
    {
      "entropy": 1.0529224690049888,
      "epoch": 0.46574910845996303,
      "grad_norm": 9.5625,
      "learning_rate": 4.417692284659103e-05,
      "loss": 1.10111665725708,
      "mean_token_accuracy": 0.7448539096862078,
      "num_tokens": 32141259.0,
      "step": 16260
    },
    {
      "entropy": 1.0870555348694324,
      "epoch": 0.4660355470260516,
      "grad_norm": 6.4375,
      "learning_rate": 4.416959962039733e-05,
      "loss": 1.112355613708496,
      "mean_token_accuracy": 0.7580804407596589,
      "num_tokens": 32160886.0,
      "step": 16270
    },
    {
      "entropy": 0.9704219080507756,
      "epoch": 0.4663219855921401,
      "grad_norm": 5.90625,
      "learning_rate": 4.4162272399974635e-05,
      "loss": 0.9465706825256348,
      "mean_token_accuracy": 0.7803370431065559,
      "num_tokens": 32181809.0,
      "step": 16280
    },
    {
      "entropy": 1.1362971387803555,
      "epoch": 0.4666084241582287,
      "grad_norm": 8.9375,
      "learning_rate": 4.4154941186849666e-05,
      "loss": 1.2174898147583009,
      "mean_token_accuracy": 0.7339811652898789,
      "num_tokens": 32202112.0,
      "step": 16290
    },
    {
      "entropy": 1.0286120142787696,
      "epoch": 0.4668948627243172,
      "grad_norm": 8.1875,
      "learning_rate": 4.4147605982549956e-05,
      "loss": 1.0287651062011718,
      "mean_token_accuracy": 0.7628949638456106,
      "num_tokens": 32223532.0,
      "step": 16300
    },
    {
      "entropy": 1.048190675675869,
      "epoch": 0.46718130129040575,
      "grad_norm": 9.0625,
      "learning_rate": 4.414026678860391e-05,
      "loss": 1.0524218559265137,
      "mean_token_accuracy": 0.7560220587998628,
      "num_tokens": 32243045.0,
      "step": 16310
    },
    {
      "entropy": 1.019164897315204,
      "epoch": 0.46746773985649426,
      "grad_norm": 8.5625,
      "learning_rate": 4.413292360654073e-05,
      "loss": 1.0262968063354492,
      "mean_token_accuracy": 0.7623854793608189,
      "num_tokens": 32261927.0,
      "step": 16320
    },
    {
      "entropy": 1.037765220925212,
      "epoch": 0.4677541784225828,
      "grad_norm": 6.09375,
      "learning_rate": 4.412557643789047e-05,
      "loss": 1.0248547554016114,
      "mean_token_accuracy": 0.7714346315711736,
      "num_tokens": 32281805.0,
      "step": 16330
    },
    {
      "entropy": 1.1185858346521855,
      "epoch": 0.46804061698867133,
      "grad_norm": 8.8125,
      "learning_rate": 4.4118225284184e-05,
      "loss": 1.1060708999633788,
      "mean_token_accuracy": 0.740351352468133,
      "num_tokens": 32302197.0,
      "step": 16340
    },
    {
      "entropy": 1.0791988208889962,
      "epoch": 0.4683270555547599,
      "grad_norm": 6.65625,
      "learning_rate": 4.411087014695302e-05,
      "loss": 1.1324837684631348,
      "mean_token_accuracy": 0.7491525683552027,
      "num_tokens": 32320842.0,
      "step": 16350
    },
    {
      "entropy": 1.0408789537847043,
      "epoch": 0.4686134941208484,
      "grad_norm": 6.625,
      "learning_rate": 4.410351102773008e-05,
      "loss": 1.0241211891174316,
      "mean_token_accuracy": 0.7648175466805697,
      "num_tokens": 32341103.0,
      "step": 16360
    },
    {
      "entropy": 0.961898205243051,
      "epoch": 0.46889993268693697,
      "grad_norm": 8.875,
      "learning_rate": 4.409614792804853e-05,
      "loss": 1.009656047821045,
      "mean_token_accuracy": 0.7711260125041008,
      "num_tokens": 32360083.0,
      "step": 16370
    },
    {
      "entropy": 0.9625785581767559,
      "epoch": 0.4691863712530255,
      "grad_norm": 5.9375,
      "learning_rate": 4.408878084944258e-05,
      "loss": 0.936217212677002,
      "mean_token_accuracy": 0.7729920975863933,
      "num_tokens": 32380350.0,
      "step": 16380
    },
    {
      "entropy": 0.9434476662427187,
      "epoch": 0.46947280981911405,
      "grad_norm": 8.875,
      "learning_rate": 4.408140979344724e-05,
      "loss": 0.984708023071289,
      "mean_token_accuracy": 0.7744307465851307,
      "num_tokens": 32400028.0,
      "step": 16390
    },
    {
      "entropy": 1.211448898538947,
      "epoch": 0.4697592483852026,
      "grad_norm": 7.375,
      "learning_rate": 4.407403476159838e-05,
      "loss": 1.2483889579772949,
      "mean_token_accuracy": 0.7246389653533697,
      "num_tokens": 32421388.0,
      "step": 16400
    },
    {
      "entropy": 1.0420224644243716,
      "epoch": 0.4700456869512911,
      "grad_norm": 7.59375,
      "learning_rate": 4.406665575543266e-05,
      "loss": 1.0778291702270508,
      "mean_token_accuracy": 0.7510162197053433,
      "num_tokens": 32439293.0,
      "step": 16410
    },
    {
      "entropy": 1.0927003175020218,
      "epoch": 0.4703321255173797,
      "grad_norm": 8.5,
      "learning_rate": 4.4059272776487615e-05,
      "loss": 1.1152701377868652,
      "mean_token_accuracy": 0.7570673380047083,
      "num_tokens": 32458437.0,
      "step": 16420
    },
    {
      "entropy": 1.1206925608217717,
      "epoch": 0.4706185640834682,
      "grad_norm": 7.90625,
      "learning_rate": 4.4051885826301555e-05,
      "loss": 1.083002758026123,
      "mean_token_accuracy": 0.7551890447735786,
      "num_tokens": 32477903.0,
      "step": 16430
    },
    {
      "entropy": 0.9830199084244668,
      "epoch": 0.47090500264955676,
      "grad_norm": 7.6875,
      "learning_rate": 4.404449490641367e-05,
      "loss": 1.006961727142334,
      "mean_token_accuracy": 0.7652568735182286,
      "num_tokens": 32496927.0,
      "step": 16440
    },
    {
      "entropy": 1.0167738554999233,
      "epoch": 0.47119144121564527,
      "grad_norm": 6.46875,
      "learning_rate": 4.403710001836393e-05,
      "loss": 1.0323945999145507,
      "mean_token_accuracy": 0.7582252528518438,
      "num_tokens": 32517507.0,
      "step": 16450
    },
    {
      "entropy": 1.0693297531455754,
      "epoch": 0.47147787978173383,
      "grad_norm": 7.25,
      "learning_rate": 4.402970116369317e-05,
      "loss": 1.0309248924255372,
      "mean_token_accuracy": 0.7591683093458415,
      "num_tokens": 32537100.0,
      "step": 16460
    },
    {
      "entropy": 1.100159036554396,
      "epoch": 0.47176431834782234,
      "grad_norm": 9.8125,
      "learning_rate": 4.402229834394302e-05,
      "loss": 1.0949154853820802,
      "mean_token_accuracy": 0.7535491392016411,
      "num_tokens": 32556238.0,
      "step": 16470
    },
    {
      "entropy": 0.9910267256200314,
      "epoch": 0.4720507569139109,
      "grad_norm": 5.5625,
      "learning_rate": 4.4014891560655966e-05,
      "loss": 1.0019682884216308,
      "mean_token_accuracy": 0.7625079181045293,
      "num_tokens": 32576767.0,
      "step": 16480
    },
    {
      "entropy": 1.0701827768236398,
      "epoch": 0.4723371954799994,
      "grad_norm": 12.875,
      "learning_rate": 4.4007480815375294e-05,
      "loss": 1.0903682708740234,
      "mean_token_accuracy": 0.7553587462753057,
      "num_tokens": 32595311.0,
      "step": 16490
    },
    {
      "entropy": 1.0541426962241531,
      "epoch": 0.472623634046088,
      "grad_norm": 6.0,
      "learning_rate": 4.400006610964514e-05,
      "loss": 1.0682659149169922,
      "mean_token_accuracy": 0.7575784578919411,
      "num_tokens": 32615959.0,
      "step": 16500
    },
    {
      "entropy": 1.0715593788772821,
      "epoch": 0.4729100726121765,
      "grad_norm": 9.6875,
      "learning_rate": 4.3992647445010434e-05,
      "loss": 0.9990396499633789,
      "mean_token_accuracy": 0.7640746839344501,
      "num_tokens": 32636768.0,
      "step": 16510
    },
    {
      "entropy": 0.9712622296065092,
      "epoch": 0.47319651117826506,
      "grad_norm": 5.15625,
      "learning_rate": 4.398522482301697e-05,
      "loss": 0.9884244918823242,
      "mean_token_accuracy": 0.7592273127287626,
      "num_tokens": 32655691.0,
      "step": 16520
    },
    {
      "entropy": 0.9805742347612977,
      "epoch": 0.47348294974435357,
      "grad_norm": 9.25,
      "learning_rate": 4.397779824521132e-05,
      "loss": 1.0709027290344237,
      "mean_token_accuracy": 0.7572538528591395,
      "num_tokens": 32673670.0,
      "step": 16530
    },
    {
      "entropy": 1.035470288991928,
      "epoch": 0.47376938831044213,
      "grad_norm": 8.4375,
      "learning_rate": 4.397036771314092e-05,
      "loss": 1.047757625579834,
      "mean_token_accuracy": 0.7669108416885138,
      "num_tokens": 32695630.0,
      "step": 16540
    },
    {
      "entropy": 0.9915830597281456,
      "epoch": 0.47405582687653064,
      "grad_norm": 9.6875,
      "learning_rate": 4.3962933228354034e-05,
      "loss": 1.0023653030395507,
      "mean_token_accuracy": 0.7671206422150135,
      "num_tokens": 32715304.0,
      "step": 16550
    },
    {
      "entropy": 0.9150912884622813,
      "epoch": 0.4743422654426192,
      "grad_norm": 9.625,
      "learning_rate": 4.39554947923997e-05,
      "loss": 0.8819180488586426,
      "mean_token_accuracy": 0.7916360713541508,
      "num_tokens": 32734383.0,
      "step": 16560
    },
    {
      "entropy": 0.9643123431131244,
      "epoch": 0.4746287040087077,
      "grad_norm": 6.71875,
      "learning_rate": 4.394805240682782e-05,
      "loss": 0.9794898986816406,
      "mean_token_accuracy": 0.7702979065477848,
      "num_tokens": 32753341.0,
      "step": 16570
    },
    {
      "entropy": 1.0536919809877872,
      "epoch": 0.4749151425747963,
      "grad_norm": 6.5625,
      "learning_rate": 4.394060607318911e-05,
      "loss": 1.0409936904907227,
      "mean_token_accuracy": 0.757367143779993,
      "num_tokens": 32773065.0,
      "step": 16580
    },
    {
      "entropy": 1.0067381033673883,
      "epoch": 0.4752015811408848,
      "grad_norm": 8.75,
      "learning_rate": 4.393315579303511e-05,
      "loss": 1.0525212287902832,
      "mean_token_accuracy": 0.7602097991853952,
      "num_tokens": 32793785.0,
      "step": 16590
    },
    {
      "entropy": 1.0108564496040344,
      "epoch": 0.47548801970697335,
      "grad_norm": 7.71875,
      "learning_rate": 4.3925701567918184e-05,
      "loss": 1.0344018936157227,
      "mean_token_accuracy": 0.7682503521442413,
      "num_tokens": 32813155.0,
      "step": 16600
    },
    {
      "entropy": 0.9619921205565334,
      "epoch": 0.47577445827306186,
      "grad_norm": 9.125,
      "learning_rate": 4.391824339939151e-05,
      "loss": 1.00352201461792,
      "mean_token_accuracy": 0.7671362828463316,
      "num_tokens": 32833528.0,
      "step": 16610
    },
    {
      "entropy": 1.0515707001090049,
      "epoch": 0.4760608968391504,
      "grad_norm": 7.6875,
      "learning_rate": 4.391078128900909e-05,
      "loss": 1.0090611457824707,
      "mean_token_accuracy": 0.7649975664913654,
      "num_tokens": 32855153.0,
      "step": 16620
    },
    {
      "entropy": 0.9998664915561676,
      "epoch": 0.47634733540523894,
      "grad_norm": 8.375,
      "learning_rate": 4.3903315238325755e-05,
      "loss": 1.0471282005310059,
      "mean_token_accuracy": 0.770268089696765,
      "num_tokens": 32875204.0,
      "step": 16630
    },
    {
      "entropy": 1.080232111737132,
      "epoch": 0.4766337739713275,
      "grad_norm": 7.65625,
      "learning_rate": 4.3895845248897146e-05,
      "loss": 1.1180587768554688,
      "mean_token_accuracy": 0.7468327790498733,
      "num_tokens": 32893853.0,
      "step": 16640
    },
    {
      "entropy": 1.0761520275846124,
      "epoch": 0.476920212537416,
      "grad_norm": 7.34375,
      "learning_rate": 4.388837132227973e-05,
      "loss": 1.0993581771850587,
      "mean_token_accuracy": 0.7479993399232626,
      "num_tokens": 32913272.0,
      "step": 16650
    },
    {
      "entropy": 0.9716217651963234,
      "epoch": 0.4772066511035046,
      "grad_norm": 9.0625,
      "learning_rate": 4.388089346003079e-05,
      "loss": 0.9898581504821777,
      "mean_token_accuracy": 0.7654787879437208,
      "num_tokens": 32933194.0,
      "step": 16660
    },
    {
      "entropy": 1.0171740382909775,
      "epoch": 0.47749308966959314,
      "grad_norm": 8.5625,
      "learning_rate": 4.387341166370845e-05,
      "loss": 1.0169093132019043,
      "mean_token_accuracy": 0.7633578985929489,
      "num_tokens": 32952794.0,
      "step": 16670
    },
    {
      "entropy": 1.002915321663022,
      "epoch": 0.47777952823568165,
      "grad_norm": 8.25,
      "learning_rate": 4.3865925934871624e-05,
      "loss": 1.0122275352478027,
      "mean_token_accuracy": 0.755439879372716,
      "num_tokens": 32972208.0,
      "step": 16680
    },
    {
      "entropy": 1.0085969856008887,
      "epoch": 0.4780659668017702,
      "grad_norm": 7.875,
      "learning_rate": 4.3858436275080054e-05,
      "loss": 1.041721248626709,
      "mean_token_accuracy": 0.7611064448952675,
      "num_tokens": 32992165.0,
      "step": 16690
    },
    {
      "entropy": 0.9909197997301817,
      "epoch": 0.4783524053678587,
      "grad_norm": 8.25,
      "learning_rate": 4.3850942685894326e-05,
      "loss": 0.9252277374267578,
      "mean_token_accuracy": 0.768148074671626,
      "num_tokens": 33012145.0,
      "step": 16700
    },
    {
      "entropy": 0.9755357466638088,
      "epoch": 0.4786388439339473,
      "grad_norm": 7.375,
      "learning_rate": 4.3843445168875795e-05,
      "loss": 0.9919824600219727,
      "mean_token_accuracy": 0.7660184435546398,
      "num_tokens": 33031648.0,
      "step": 16710
    },
    {
      "entropy": 1.0443235857412219,
      "epoch": 0.4789252825000358,
      "grad_norm": 8.625,
      "learning_rate": 4.383594372558669e-05,
      "loss": 1.0170072555541991,
      "mean_token_accuracy": 0.7575950171798468,
      "num_tokens": 33052487.0,
      "step": 16720
    },
    {
      "entropy": 1.0474604623392225,
      "epoch": 0.47921172106612436,
      "grad_norm": 9.3125,
      "learning_rate": 4.3828438357590015e-05,
      "loss": 1.1549976348876954,
      "mean_token_accuracy": 0.7388071183115243,
      "num_tokens": 33071151.0,
      "step": 16730
    },
    {
      "entropy": 0.9586014881730079,
      "epoch": 0.4794981596322129,
      "grad_norm": 7.84375,
      "learning_rate": 4.382092906644961e-05,
      "loss": 0.988189697265625,
      "mean_token_accuracy": 0.7776207908987999,
      "num_tokens": 33090915.0,
      "step": 16740
    },
    {
      "entropy": 1.1069232895970345,
      "epoch": 0.47978459819830144,
      "grad_norm": 9.0,
      "learning_rate": 4.381341585373014e-05,
      "loss": 1.0873683929443358,
      "mean_token_accuracy": 0.7427266146987677,
      "num_tokens": 33110927.0,
      "step": 16750
    },
    {
      "entropy": 1.1065772384405137,
      "epoch": 0.48007103676438995,
      "grad_norm": 9.5,
      "learning_rate": 4.380589872099708e-05,
      "loss": 1.1303060531616211,
      "mean_token_accuracy": 0.7345760636031627,
      "num_tokens": 33130558.0,
      "step": 16760
    },
    {
      "entropy": 0.9561706442385912,
      "epoch": 0.4803574753304785,
      "grad_norm": 9.25,
      "learning_rate": 4.3798377669816715e-05,
      "loss": 0.9489264488220215,
      "mean_token_accuracy": 0.7789783325046301,
      "num_tokens": 33149793.0,
      "step": 16770
    },
    {
      "entropy": 0.9976607786491514,
      "epoch": 0.480643913896567,
      "grad_norm": 5.53125,
      "learning_rate": 4.379085270175614e-05,
      "loss": 1.0097550392150878,
      "mean_token_accuracy": 0.7661747455596923,
      "num_tokens": 33170110.0,
      "step": 16780
    },
    {
      "entropy": 1.0032922867685556,
      "epoch": 0.4809303524626556,
      "grad_norm": 8.3125,
      "learning_rate": 4.378332381838329e-05,
      "loss": 1.0375161170959473,
      "mean_token_accuracy": 0.7576687451452017,
      "num_tokens": 33189818.0,
      "step": 16790
    },
    {
      "entropy": 1.0627506069839001,
      "epoch": 0.4812167910287441,
      "grad_norm": 8.25,
      "learning_rate": 4.377579102126691e-05,
      "loss": 1.1434754371643066,
      "mean_token_accuracy": 0.7500649422407151,
      "num_tokens": 33209724.0,
      "step": 16800
    },
    {
      "entropy": 0.985107010602951,
      "epoch": 0.48150322959483266,
      "grad_norm": 8.9375,
      "learning_rate": 4.376825431197653e-05,
      "loss": 0.9973355293273926,
      "mean_token_accuracy": 0.7671806927770376,
      "num_tokens": 33229700.0,
      "step": 16810
    },
    {
      "entropy": 0.993674466572702,
      "epoch": 0.48178966816092117,
      "grad_norm": 9.5,
      "learning_rate": 4.376071369208253e-05,
      "loss": 1.018948745727539,
      "mean_token_accuracy": 0.7718550018966198,
      "num_tokens": 33248992.0,
      "step": 16820
    },
    {
      "entropy": 1.0256044575944543,
      "epoch": 0.48207610672700973,
      "grad_norm": 6.59375,
      "learning_rate": 4.375316916315612e-05,
      "loss": 0.980533504486084,
      "mean_token_accuracy": 0.7829587046056986,
      "num_tokens": 33270149.0,
      "step": 16830
    },
    {
      "entropy": 1.1083214286714793,
      "epoch": 0.48236254529309824,
      "grad_norm": 7.25,
      "learning_rate": 4.3745620726769256e-05,
      "loss": 1.1307127952575684,
      "mean_token_accuracy": 0.7472021080553531,
      "num_tokens": 33290780.0,
      "step": 16840
    },
    {
      "entropy": 0.9201938167214394,
      "epoch": 0.4826489838591868,
      "grad_norm": 8.3125,
      "learning_rate": 4.373806838449478e-05,
      "loss": 0.8909637451171875,
      "mean_token_accuracy": 0.785581337660551,
      "num_tokens": 33310849.0,
      "step": 16850
    },
    {
      "entropy": 1.0885490491054952,
      "epoch": 0.4829354224252753,
      "grad_norm": 7.9375,
      "learning_rate": 4.373051213790629e-05,
      "loss": 1.1084395408630372,
      "mean_token_accuracy": 0.7535569231957198,
      "num_tokens": 33330989.0,
      "step": 16860
    },
    {
      "entropy": 1.071072907745838,
      "epoch": 0.4832218609913639,
      "grad_norm": 7.71875,
      "learning_rate": 4.372295198857826e-05,
      "loss": 1.1154266357421876,
      "mean_token_accuracy": 0.7369588062167167,
      "num_tokens": 33352445.0,
      "step": 16870
    },
    {
      "entropy": 1.103079878166318,
      "epoch": 0.4835082995574524,
      "grad_norm": 9.5625,
      "learning_rate": 4.371538793808592e-05,
      "loss": 1.1553272247314452,
      "mean_token_accuracy": 0.7345395524054765,
      "num_tokens": 33372442.0,
      "step": 16880
    },
    {
      "entropy": 0.9715333161875606,
      "epoch": 0.48379473812354096,
      "grad_norm": 8.75,
      "learning_rate": 4.3707819988005346e-05,
      "loss": 0.9788935661315918,
      "mean_token_accuracy": 0.7683930456638336,
      "num_tokens": 33390521.0,
      "step": 16890
    },
    {
      "entropy": 0.9345890924334526,
      "epoch": 0.48408117668962947,
      "grad_norm": 7.71875,
      "learning_rate": 4.370024813991341e-05,
      "loss": 0.926908016204834,
      "mean_token_accuracy": 0.7792803592979908,
      "num_tokens": 33409438.0,
      "step": 16900
    },
    {
      "entropy": 1.0284394092857838,
      "epoch": 0.48436761525571803,
      "grad_norm": 6.78125,
      "learning_rate": 4.36926723953878e-05,
      "loss": 1.0752223014831543,
      "mean_token_accuracy": 0.751723089069128,
      "num_tokens": 33428484.0,
      "step": 16910
    },
    {
      "entropy": 1.0472633289173245,
      "epoch": 0.48465405382180654,
      "grad_norm": 7.5,
      "learning_rate": 4.368509275600703e-05,
      "loss": 1.052652931213379,
      "mean_token_accuracy": 0.7590388901531696,
      "num_tokens": 33448426.0,
      "step": 16920
    },
    {
      "entropy": 1.072967306524515,
      "epoch": 0.4849404923878951,
      "grad_norm": 7.34375,
      "learning_rate": 4.3677509223350406e-05,
      "loss": 1.0929318428039552,
      "mean_token_accuracy": 0.7458045918494463,
      "num_tokens": 33468472.0,
      "step": 16930
    },
    {
      "entropy": 1.0414020497351886,
      "epoch": 0.48522693095398367,
      "grad_norm": 7.46875,
      "learning_rate": 4.366992179899805e-05,
      "loss": 1.0008486747741698,
      "mean_token_accuracy": 0.767263027280569,
      "num_tokens": 33488239.0,
      "step": 16940
    },
    {
      "entropy": 1.0647817399352788,
      "epoch": 0.4855133695200722,
      "grad_norm": 7.75,
      "learning_rate": 4.366233048453091e-05,
      "loss": 1.0782719612121583,
      "mean_token_accuracy": 0.758022814244032,
      "num_tokens": 33508509.0,
      "step": 16950
    },
    {
      "entropy": 0.9815584300085902,
      "epoch": 0.48579980808616074,
      "grad_norm": 7.0,
      "learning_rate": 4.365473528153072e-05,
      "loss": 1.015487003326416,
      "mean_token_accuracy": 0.7724594675004482,
      "num_tokens": 33528425.0,
      "step": 16960
    },
    {
      "entropy": 1.0222778972238302,
      "epoch": 0.48608624665224925,
      "grad_norm": 8.5625,
      "learning_rate": 4.364713619158004e-05,
      "loss": 1.0497114181518554,
      "mean_token_accuracy": 0.7550367487594485,
      "num_tokens": 33549355.0,
      "step": 16970
    },
    {
      "entropy": 1.0406492583453655,
      "epoch": 0.4863726852183378,
      "grad_norm": 6.8125,
      "learning_rate": 4.363953321626224e-05,
      "loss": 1.1019960403442384,
      "mean_token_accuracy": 0.7563046537339687,
      "num_tokens": 33569055.0,
      "step": 16980
    },
    {
      "entropy": 1.0000797294080257,
      "epoch": 0.48665912378442633,
      "grad_norm": 9.0,
      "learning_rate": 4.3631926357161494e-05,
      "loss": 0.9749500274658203,
      "mean_token_accuracy": 0.7699035588651896,
      "num_tokens": 33589169.0,
      "step": 16990
    },
    {
      "entropy": 1.194348792731762,
      "epoch": 0.4869455623505149,
      "grad_norm": 8.875,
      "learning_rate": 4.362431561586279e-05,
      "loss": 1.2443226814270019,
      "mean_token_accuracy": 0.7174331754446029,
      "num_tokens": 33610259.0,
      "step": 17000
    },
    {
      "entropy": 0.924548857472837,
      "epoch": 0.4872320009166034,
      "grad_norm": 6.75,
      "learning_rate": 4.361670099395192e-05,
      "loss": 0.9259262084960938,
      "mean_token_accuracy": 0.7841747581958771,
      "num_tokens": 33629443.0,
      "step": 17010
    },
    {
      "entropy": 1.1185907673090696,
      "epoch": 0.48751843948269197,
      "grad_norm": 10.5,
      "learning_rate": 4.3609082493015486e-05,
      "loss": 1.1625787734985351,
      "mean_token_accuracy": 0.7360367476940155,
      "num_tokens": 33648652.0,
      "step": 17020
    },
    {
      "entropy": 1.0880159426480531,
      "epoch": 0.4878048780487805,
      "grad_norm": 5.65625,
      "learning_rate": 4.3601460114640905e-05,
      "loss": 1.0699710845947266,
      "mean_token_accuracy": 0.7604692164808512,
      "num_tokens": 33667495.0,
      "step": 17030
    },
    {
      "entropy": 1.0287564408034087,
      "epoch": 0.48809131661486904,
      "grad_norm": 6.8125,
      "learning_rate": 4.3593833860416386e-05,
      "loss": 1.0119543075561523,
      "mean_token_accuracy": 0.7653553612530232,
      "num_tokens": 33686875.0,
      "step": 17040
    },
    {
      "entropy": 0.9676519716158509,
      "epoch": 0.48837775518095755,
      "grad_norm": 7.75,
      "learning_rate": 4.358620373193096e-05,
      "loss": 1.002499485015869,
      "mean_token_accuracy": 0.7622711818665266,
      "num_tokens": 33707062.0,
      "step": 17050
    },
    {
      "entropy": 0.9931231487542391,
      "epoch": 0.4886641937470461,
      "grad_norm": 7.25,
      "learning_rate": 4.357856973077447e-05,
      "loss": 1.031269645690918,
      "mean_token_accuracy": 0.7662841275334358,
      "num_tokens": 33727415.0,
      "step": 17060
    },
    {
      "entropy": 0.99451689645648,
      "epoch": 0.4889506323131346,
      "grad_norm": 7.28125,
      "learning_rate": 4.357093185853754e-05,
      "loss": 0.998592472076416,
      "mean_token_accuracy": 0.7713192895054817,
      "num_tokens": 33746744.0,
      "step": 17070
    },
    {
      "entropy": 0.9768349003046751,
      "epoch": 0.4892370708792232,
      "grad_norm": 5.9375,
      "learning_rate": 4.356329011681164e-05,
      "loss": 1.0129617691040038,
      "mean_token_accuracy": 0.7762565024197101,
      "num_tokens": 33766178.0,
      "step": 17080
    },
    {
      "entropy": 1.0487086234614253,
      "epoch": 0.4895235094453117,
      "grad_norm": 8.5625,
      "learning_rate": 4.3555644507189e-05,
      "loss": 1.0835046768188477,
      "mean_token_accuracy": 0.7636639304459095,
      "num_tokens": 33786407.0,
      "step": 17090
    },
    {
      "entropy": 1.1217915868386625,
      "epoch": 0.48980994801140026,
      "grad_norm": 8.625,
      "learning_rate": 4.354799503126269e-05,
      "loss": 1.148108959197998,
      "mean_token_accuracy": 0.7458061631768942,
      "num_tokens": 33806162.0,
      "step": 17100
    },
    {
      "entropy": 0.9219942964613438,
      "epoch": 0.4900963865774888,
      "grad_norm": 7.0,
      "learning_rate": 4.354034169062658e-05,
      "loss": 0.9051288604736328,
      "mean_token_accuracy": 0.788603700697422,
      "num_tokens": 33825708.0,
      "step": 17110
    },
    {
      "entropy": 1.132470526173711,
      "epoch": 0.49038282514357734,
      "grad_norm": 9.1875,
      "learning_rate": 4.353268448687534e-05,
      "loss": 1.1624566078186036,
      "mean_token_accuracy": 0.7464070305228233,
      "num_tokens": 33845744.0,
      "step": 17120
    },
    {
      "entropy": 1.0601427786052227,
      "epoch": 0.49066926370966585,
      "grad_norm": 6.8125,
      "learning_rate": 4.3525023421604426e-05,
      "loss": 1.0351591110229492,
      "mean_token_accuracy": 0.7563232768326997,
      "num_tokens": 33864856.0,
      "step": 17130
    },
    {
      "entropy": 1.0103509912267328,
      "epoch": 0.4909557022757544,
      "grad_norm": 6.59375,
      "learning_rate": 4.3517358496410147e-05,
      "loss": 1.0299256324768067,
      "mean_token_accuracy": 0.7625603768974543,
      "num_tokens": 33884640.0,
      "step": 17140
    },
    {
      "entropy": 1.1366069884970784,
      "epoch": 0.4912421408418429,
      "grad_norm": 7.875,
      "learning_rate": 4.350968971288957e-05,
      "loss": 1.0999591827392579,
      "mean_token_accuracy": 0.7496421072632075,
      "num_tokens": 33904084.0,
      "step": 17150
    },
    {
      "entropy": 1.040277023985982,
      "epoch": 0.4915285794079315,
      "grad_norm": 9.75,
      "learning_rate": 4.350201707264058e-05,
      "loss": 1.0827797889709472,
      "mean_token_accuracy": 0.7595674704760313,
      "num_tokens": 33922737.0,
      "step": 17160
    },
    {
      "entropy": 0.9458992199972271,
      "epoch": 0.49181501797402,
      "grad_norm": 6.96875,
      "learning_rate": 4.349434057726188e-05,
      "loss": 0.9751693725585937,
      "mean_token_accuracy": 0.7761420492082834,
      "num_tokens": 33941159.0,
      "step": 17170
    },
    {
      "entropy": 1.0709429727867246,
      "epoch": 0.49210145654010856,
      "grad_norm": 10.8125,
      "learning_rate": 4.348666022835297e-05,
      "loss": 1.0812703132629395,
      "mean_token_accuracy": 0.7573617920279503,
      "num_tokens": 33960162.0,
      "step": 17180
    },
    {
      "entropy": 1.1367273677140475,
      "epoch": 0.49238789510619707,
      "grad_norm": 10.9375,
      "learning_rate": 4.3478976027514126e-05,
      "loss": 1.1629722595214844,
      "mean_token_accuracy": 0.7320704910904169,
      "num_tokens": 33981917.0,
      "step": 17190
    },
    {
      "entropy": 1.0175492121838032,
      "epoch": 0.49267433367228564,
      "grad_norm": 9.75,
      "learning_rate": 4.3471287976346475e-05,
      "loss": 1.0513174057006835,
      "mean_token_accuracy": 0.7539138633757829,
      "num_tokens": 34000439.0,
      "step": 17200
    },
    {
      "entropy": 1.0400999519973992,
      "epoch": 0.4929607722383742,
      "grad_norm": 8.75,
      "learning_rate": 4.34635960764519e-05,
      "loss": 1.0125757217407227,
      "mean_token_accuracy": 0.7581253346055746,
      "num_tokens": 34020019.0,
      "step": 17210
    },
    {
      "entropy": 1.1167758348397911,
      "epoch": 0.4932472108044627,
      "grad_norm": 7.59375,
      "learning_rate": 4.34559003294331e-05,
      "loss": 1.1307131767272949,
      "mean_token_accuracy": 0.7458459671586752,
      "num_tokens": 34039914.0,
      "step": 17220
    },
    {
      "entropy": 1.0591639783233404,
      "epoch": 0.4935336493705513,
      "grad_norm": 7.6875,
      "learning_rate": 4.344820073689361e-05,
      "loss": 1.1296196937561036,
      "mean_token_accuracy": 0.7466318164020777,
      "num_tokens": 34060611.0,
      "step": 17230
    },
    {
      "entropy": 1.0027164427563549,
      "epoch": 0.4938200879366398,
      "grad_norm": 6.3125,
      "learning_rate": 4.344049730043771e-05,
      "loss": 0.9870219230651855,
      "mean_token_accuracy": 0.7641301296651364,
      "num_tokens": 34079949.0,
      "step": 17240
    },
    {
      "entropy": 1.1063085883855819,
      "epoch": 0.49410652650272835,
      "grad_norm": 10.1875,
      "learning_rate": 4.3432790021670526e-05,
      "loss": 1.0862249374389648,
      "mean_token_accuracy": 0.747665136307478,
      "num_tokens": 34098360.0,
      "step": 17250
    },
    {
      "entropy": 0.9360519904643297,
      "epoch": 0.49439296506881686,
      "grad_norm": 8.5625,
      "learning_rate": 4.342507890219796e-05,
      "loss": 0.9213293075561524,
      "mean_token_accuracy": 0.7864279914647341,
      "num_tokens": 34116974.0,
      "step": 17260
    },
    {
      "entropy": 0.9939828582108021,
      "epoch": 0.4946794036349054,
      "grad_norm": 9.0,
      "learning_rate": 4.341736394362672e-05,
      "loss": 1.0747446060180663,
      "mean_token_accuracy": 0.7556054197251797,
      "num_tokens": 34136181.0,
      "step": 17270
    },
    {
      "entropy": 1.0795869410037995,
      "epoch": 0.49496584220099393,
      "grad_norm": 9.8125,
      "learning_rate": 4.340964514756433e-05,
      "loss": 1.1161347389221192,
      "mean_token_accuracy": 0.7463672518730163,
      "num_tokens": 34155948.0,
      "step": 17280
    },
    {
      "entropy": 0.9304128894582391,
      "epoch": 0.4952522807670825,
      "grad_norm": 6.125,
      "learning_rate": 4.340192251561907e-05,
      "loss": 0.9230742454528809,
      "mean_token_accuracy": 0.7848080396652222,
      "num_tokens": 34175331.0,
      "step": 17290
    },
    {
      "entropy": 0.9416028181090951,
      "epoch": 0.495538719333171,
      "grad_norm": 11.1875,
      "learning_rate": 4.3394196049400074e-05,
      "loss": 0.954836654663086,
      "mean_token_accuracy": 0.7780571576207876,
      "num_tokens": 34195328.0,
      "step": 17300
    },
    {
      "entropy": 1.0265587870031596,
      "epoch": 0.49582515789925957,
      "grad_norm": 7.25,
      "learning_rate": 4.3386465750517236e-05,
      "loss": 0.9978179931640625,
      "mean_token_accuracy": 0.7697482436895371,
      "num_tokens": 34215761.0,
      "step": 17310
    },
    {
      "entropy": 1.0152923617511989,
      "epoch": 0.4961115964653481,
      "grad_norm": 8.125,
      "learning_rate": 4.337873162058127e-05,
      "loss": 1.0142152786254883,
      "mean_token_accuracy": 0.7607778094708919,
      "num_tokens": 34235850.0,
      "step": 17320
    },
    {
      "entropy": 1.0235864974558353,
      "epoch": 0.49639803503143665,
      "grad_norm": 8.375,
      "learning_rate": 4.337099366120367e-05,
      "loss": 1.0215274810791015,
      "mean_token_accuracy": 0.7588078200817108,
      "num_tokens": 34255831.0,
      "step": 17330
    },
    {
      "entropy": 0.9520710717886687,
      "epoch": 0.49668447359752516,
      "grad_norm": 5.6875,
      "learning_rate": 4.336325187399675e-05,
      "loss": 0.9275893211364746,
      "mean_token_accuracy": 0.7845791764557362,
      "num_tokens": 34275446.0,
      "step": 17340
    },
    {
      "entropy": 0.9020056877285242,
      "epoch": 0.4969709121636137,
      "grad_norm": 6.5,
      "learning_rate": 4.33555062605736e-05,
      "loss": 0.9286996841430664,
      "mean_token_accuracy": 0.7855431295931339,
      "num_tokens": 34294828.0,
      "step": 17350
    },
    {
      "entropy": 0.9784899307414889,
      "epoch": 0.49725735072970223,
      "grad_norm": 8.0625,
      "learning_rate": 4.3347756822548116e-05,
      "loss": 1.0615654945373536,
      "mean_token_accuracy": 0.7579973589628934,
      "num_tokens": 34314617.0,
      "step": 17360
    },
    {
      "entropy": 1.0217004781588912,
      "epoch": 0.4975437892957908,
      "grad_norm": 8.0625,
      "learning_rate": 4.3340003561535e-05,
      "loss": 1.105919361114502,
      "mean_token_accuracy": 0.7594532150775194,
      "num_tokens": 34334508.0,
      "step": 17370
    },
    {
      "entropy": 1.0695177033543586,
      "epoch": 0.4978302278618793,
      "grad_norm": 8.375,
      "learning_rate": 4.333224647914974e-05,
      "loss": 1.0597644805908204,
      "mean_token_accuracy": 0.7598655015230179,
      "num_tokens": 34354657.0,
      "step": 17380
    },
    {
      "entropy": 1.0236263621598483,
      "epoch": 0.49811666642796787,
      "grad_norm": 8.8125,
      "learning_rate": 4.332448557700862e-05,
      "loss": 1.0609212875366212,
      "mean_token_accuracy": 0.7695693500339985,
      "num_tokens": 34373246.0,
      "step": 17390
    },
    {
      "entropy": 1.1627912618219853,
      "epoch": 0.4984031049940564,
      "grad_norm": 8.9375,
      "learning_rate": 4.331672085672872e-05,
      "loss": 1.1688899993896484,
      "mean_token_accuracy": 0.7327989466488362,
      "num_tokens": 34393682.0,
      "step": 17400
    },
    {
      "entropy": 1.046683468669653,
      "epoch": 0.49868954356014494,
      "grad_norm": 7.25,
      "learning_rate": 4.330895231992792e-05,
      "loss": 1.0652215957641602,
      "mean_token_accuracy": 0.7527300979942083,
      "num_tokens": 34413753.0,
      "step": 17410
    },
    {
      "entropy": 1.0162945104762913,
      "epoch": 0.49897598212623345,
      "grad_norm": 7.40625,
      "learning_rate": 4.33011799682249e-05,
      "loss": 1.0226430892944336,
      "mean_token_accuracy": 0.7654198616743088,
      "num_tokens": 34434791.0,
      "step": 17420
    },
    {
      "entropy": 0.9305351061746479,
      "epoch": 0.499262420692322,
      "grad_norm": 6.09375,
      "learning_rate": 4.329340380323912e-05,
      "loss": 0.919261360168457,
      "mean_token_accuracy": 0.7803567554801702,
      "num_tokens": 34454707.0,
      "step": 17430
    },
    {
      "entropy": 1.0351398844271897,
      "epoch": 0.4995488592584105,
      "grad_norm": 5.90625,
      "learning_rate": 4.328562382659084e-05,
      "loss": 1.0107476234436035,
      "mean_token_accuracy": 0.7676677696406842,
      "num_tokens": 34474284.0,
      "step": 17440
    },
    {
      "entropy": 0.9402934342622757,
      "epoch": 0.4998352978244991,
      "grad_norm": 6.34375,
      "learning_rate": 4.327784003990111e-05,
      "loss": 0.9410077095031738,
      "mean_token_accuracy": 0.7842558763921261,
      "num_tokens": 34494972.0,
      "step": 17450
    },
    {
      "entropy": 1.0047647509723903,
      "epoch": 0.5001217363905877,
      "grad_norm": 10.4375,
      "learning_rate": 4.3270052444791806e-05,
      "loss": 1.041771411895752,
      "mean_token_accuracy": 0.7594370882958174,
      "num_tokens": 34513960.0,
      "step": 17460
    },
    {
      "entropy": 1.0515913940966128,
      "epoch": 0.5004081749566762,
      "grad_norm": 7.21875,
      "learning_rate": 4.326226104288556e-05,
      "loss": 1.095116424560547,
      "mean_token_accuracy": 0.7581000730395318,
      "num_tokens": 34533912.0,
      "step": 17470
    },
    {
      "entropy": 0.9887978391721844,
      "epoch": 0.5006946135227647,
      "grad_norm": 9.75,
      "learning_rate": 4.325446583580579e-05,
      "loss": 1.016411590576172,
      "mean_token_accuracy": 0.7673349156975746,
      "num_tokens": 34552350.0,
      "step": 17480
    },
    {
      "entropy": 1.0526997491717338,
      "epoch": 0.5009810520888532,
      "grad_norm": 7.0625,
      "learning_rate": 4.324666682517676e-05,
      "loss": 1.1005334854125977,
      "mean_token_accuracy": 0.7466001737862825,
      "num_tokens": 34572192.0,
      "step": 17490
    },
    {
      "entropy": 1.0093298159539699,
      "epoch": 0.5012674906549418,
      "grad_norm": 7.90625,
      "learning_rate": 4.3238864012623466e-05,
      "loss": 0.9956887245178223,
      "mean_token_accuracy": 0.7835841689258813,
      "num_tokens": 34590951.0,
      "step": 17500
    },
    {
      "entropy": 1.0219703314825892,
      "epoch": 0.5015539292210304,
      "grad_norm": 6.5625,
      "learning_rate": 4.323105739977174e-05,
      "loss": 0.9741621017456055,
      "mean_token_accuracy": 0.7731606539338827,
      "num_tokens": 34611442.0,
      "step": 17510
    },
    {
      "entropy": 1.1547549776732922,
      "epoch": 0.5018403677871188,
      "grad_norm": 8.125,
      "learning_rate": 4.322324698824817e-05,
      "loss": 1.1710846900939942,
      "mean_token_accuracy": 0.7453276500105858,
      "num_tokens": 34631767.0,
      "step": 17520
    },
    {
      "entropy": 1.0507111344486475,
      "epoch": 0.5021268063532074,
      "grad_norm": 6.34375,
      "learning_rate": 4.321543277968017e-05,
      "loss": 1.0411754608154298,
      "mean_token_accuracy": 0.7710024759173393,
      "num_tokens": 34652588.0,
      "step": 17530
    },
    {
      "entropy": 1.0625345608219505,
      "epoch": 0.502413244919296,
      "grad_norm": 7.9375,
      "learning_rate": 4.320761477569593e-05,
      "loss": 1.087254524230957,
      "mean_token_accuracy": 0.7561471063643694,
      "num_tokens": 34673214.0,
      "step": 17540
    },
    {
      "entropy": 0.9144243970513344,
      "epoch": 0.5026996834853845,
      "grad_norm": 6.8125,
      "learning_rate": 4.319979297792441e-05,
      "loss": 0.9437449455261231,
      "mean_token_accuracy": 0.7817872405052185,
      "num_tokens": 34691758.0,
      "step": 17550
    },
    {
      "entropy": 1.0869301464408636,
      "epoch": 0.502986122051473,
      "grad_norm": 9.125,
      "learning_rate": 4.3191967387995414e-05,
      "loss": 1.0893980026245118,
      "mean_token_accuracy": 0.7469348337501287,
      "num_tokens": 34711913.0,
      "step": 17560
    },
    {
      "entropy": 1.0717716068029404,
      "epoch": 0.5032725606175615,
      "grad_norm": 7.78125,
      "learning_rate": 4.318413800753948e-05,
      "loss": 1.0915319442749023,
      "mean_token_accuracy": 0.7577225595712662,
      "num_tokens": 34732512.0,
      "step": 17570
    },
    {
      "entropy": 1.0140634212642907,
      "epoch": 0.5035589991836501,
      "grad_norm": 6.84375,
      "learning_rate": 4.317630483818795e-05,
      "loss": 1.0183159828186035,
      "mean_token_accuracy": 0.7603402189910412,
      "num_tokens": 34751350.0,
      "step": 17580
    },
    {
      "entropy": 0.9538967240601778,
      "epoch": 0.5038454377497387,
      "grad_norm": 7.4375,
      "learning_rate": 4.3168467881572994e-05,
      "loss": 0.9289972305297851,
      "mean_token_accuracy": 0.782543795928359,
      "num_tokens": 34769879.0,
      "step": 17590
    },
    {
      "entropy": 0.9489595484919846,
      "epoch": 0.5041318763158271,
      "grad_norm": 7.28125,
      "learning_rate": 4.3160627139327516e-05,
      "loss": 0.9711180686950683,
      "mean_token_accuracy": 0.7739598851650953,
      "num_tokens": 34788562.0,
      "step": 17600
    },
    {
      "entropy": 0.9910988450050354,
      "epoch": 0.5044183148819157,
      "grad_norm": 5.75,
      "learning_rate": 4.315278261308524e-05,
      "loss": 0.9780125617980957,
      "mean_token_accuracy": 0.7693098466843367,
      "num_tokens": 34808318.0,
      "step": 17610
    },
    {
      "entropy": 1.0139494251459837,
      "epoch": 0.5047047534480043,
      "grad_norm": 7.1875,
      "learning_rate": 4.314493430448068e-05,
      "loss": 1.0652266502380372,
      "mean_token_accuracy": 0.7596605405211448,
      "num_tokens": 34827889.0,
      "step": 17620
    },
    {
      "entropy": 0.9520850958302617,
      "epoch": 0.5049911920140928,
      "grad_norm": 6.875,
      "learning_rate": 4.313708221514913e-05,
      "loss": 0.9847989082336426,
      "mean_token_accuracy": 0.758737226948142,
      "num_tokens": 34847778.0,
      "step": 17630
    },
    {
      "entropy": 1.0981547605246305,
      "epoch": 0.5052776305801813,
      "grad_norm": 9.8125,
      "learning_rate": 4.312922634672666e-05,
      "loss": 1.1687585830688476,
      "mean_token_accuracy": 0.7389815121889114,
      "num_tokens": 34868321.0,
      "step": 17640
    },
    {
      "entropy": 1.101518685556948,
      "epoch": 0.5055640691462698,
      "grad_norm": 8.25,
      "learning_rate": 4.312136670085014e-05,
      "loss": 1.1061275482177735,
      "mean_token_accuracy": 0.7636947996914387,
      "num_tokens": 34887196.0,
      "step": 17650
    },
    {
      "entropy": 0.9973408224061131,
      "epoch": 0.5058505077123584,
      "grad_norm": 10.25,
      "learning_rate": 4.3113503279157246e-05,
      "loss": 0.9970702171325684,
      "mean_token_accuracy": 0.7693190291523934,
      "num_tokens": 34906891.0,
      "step": 17660
    },
    {
      "entropy": 0.9564379110932351,
      "epoch": 0.506136946278447,
      "grad_norm": 6.21875,
      "learning_rate": 4.3105636083286406e-05,
      "loss": 0.9570234298706055,
      "mean_token_accuracy": 0.7727764241397381,
      "num_tokens": 34926739.0,
      "step": 17670
    },
    {
      "entropy": 1.0849061015993358,
      "epoch": 0.5064233848445354,
      "grad_norm": 5.65625,
      "learning_rate": 4.309776511487686e-05,
      "loss": 1.1154269218444823,
      "mean_token_accuracy": 0.7550651457160711,
      "num_tokens": 34947443.0,
      "step": 17680
    },
    {
      "entropy": 1.007785512506962,
      "epoch": 0.506709823410624,
      "grad_norm": 6.46875,
      "learning_rate": 4.308989037556861e-05,
      "loss": 1.0166417121887208,
      "mean_token_accuracy": 0.7633139368146657,
      "num_tokens": 34967785.0,
      "step": 17690
    },
    {
      "entropy": 1.004833035171032,
      "epoch": 0.5069962619767125,
      "grad_norm": 8.5,
      "learning_rate": 4.308201186700247e-05,
      "loss": 0.9917290687561036,
      "mean_token_accuracy": 0.7624304544180631,
      "num_tokens": 34987961.0,
      "step": 17700
    },
    {
      "entropy": 0.9888427097350359,
      "epoch": 0.5072827005428011,
      "grad_norm": 10.8125,
      "learning_rate": 4.3074129590820023e-05,
      "loss": 0.9698850631713867,
      "mean_token_accuracy": 0.774132190272212,
      "num_tokens": 35008150.0,
      "step": 17710
    },
    {
      "entropy": 0.959241671487689,
      "epoch": 0.5075691391088896,
      "grad_norm": 7.09375,
      "learning_rate": 4.306624354866363e-05,
      "loss": 0.9847854614257813,
      "mean_token_accuracy": 0.764955461770296,
      "num_tokens": 35027845.0,
      "step": 17720
    },
    {
      "entropy": 0.9634015871211886,
      "epoch": 0.5078555776749781,
      "grad_norm": 7.9375,
      "learning_rate": 4.3058353742176474e-05,
      "loss": 0.9786049842834472,
      "mean_token_accuracy": 0.7781175464391709,
      "num_tokens": 35047439.0,
      "step": 17730
    },
    {
      "entropy": 0.8689752431586385,
      "epoch": 0.5081420162410667,
      "grad_norm": 6.53125,
      "learning_rate": 4.305046017300247e-05,
      "loss": 0.8946043014526367,
      "mean_token_accuracy": 0.7936122681945562,
      "num_tokens": 35066446.0,
      "step": 17740
    },
    {
      "entropy": 1.0056325759738685,
      "epoch": 0.5084284548071553,
      "grad_norm": 6.8125,
      "learning_rate": 4.304256284278636e-05,
      "loss": 0.9554845809936523,
      "mean_token_accuracy": 0.7830309756100178,
      "num_tokens": 35085881.0,
      "step": 17750
    },
    {
      "entropy": 1.0537802072241902,
      "epoch": 0.5087148933732438,
      "grad_norm": 7.1875,
      "learning_rate": 4.303466175317364e-05,
      "loss": 1.0887016296386718,
      "mean_token_accuracy": 0.7583649516105652,
      "num_tokens": 35106967.0,
      "step": 17760
    },
    {
      "entropy": 1.054859310016036,
      "epoch": 0.5090013319393323,
      "grad_norm": 9.3125,
      "learning_rate": 4.3026756905810616e-05,
      "loss": 1.060797119140625,
      "mean_token_accuracy": 0.7545307252556086,
      "num_tokens": 35125277.0,
      "step": 17770
    },
    {
      "entropy": 0.9801409143954516,
      "epoch": 0.5092877705054208,
      "grad_norm": 7.28125,
      "learning_rate": 4.301884830234435e-05,
      "loss": 1.0073978424072265,
      "mean_token_accuracy": 0.7772514838725328,
      "num_tokens": 35143882.0,
      "step": 17780
    },
    {
      "entropy": 0.9958844879642129,
      "epoch": 0.5095742090715094,
      "grad_norm": 8.9375,
      "learning_rate": 4.3010935944422716e-05,
      "loss": 1.0302217483520508,
      "mean_token_accuracy": 0.7519188895821571,
      "num_tokens": 35164005.0,
      "step": 17790
    },
    {
      "entropy": 1.0014954233542084,
      "epoch": 0.509860647637598,
      "grad_norm": 7.53125,
      "learning_rate": 4.300301983369433e-05,
      "loss": 0.9801791191101075,
      "mean_token_accuracy": 0.7741680897772312,
      "num_tokens": 35183950.0,
      "step": 17800
    },
    {
      "entropy": 1.1163689505308867,
      "epoch": 0.5101470862036864,
      "grad_norm": 7.75,
      "learning_rate": 4.2995099971808634e-05,
      "loss": 1.1756239891052247,
      "mean_token_accuracy": 0.7412028018385172,
      "num_tokens": 35205061.0,
      "step": 17810
    },
    {
      "entropy": 0.9667664352804423,
      "epoch": 0.510433524769775,
      "grad_norm": 7.375,
      "learning_rate": 4.298717636041583e-05,
      "loss": 0.9746028900146484,
      "mean_token_accuracy": 0.7753890853375196,
      "num_tokens": 35225069.0,
      "step": 17820
    },
    {
      "entropy": 1.1080086616799236,
      "epoch": 0.5107199633358636,
      "grad_norm": 9.0,
      "learning_rate": 4.297924900116689e-05,
      "loss": 1.1493899345397949,
      "mean_token_accuracy": 0.7486542209982872,
      "num_tokens": 35244995.0,
      "step": 17830
    },
    {
      "entropy": 1.088213350251317,
      "epoch": 0.5110064019019521,
      "grad_norm": 10.0,
      "learning_rate": 4.297131789571359e-05,
      "loss": 1.1243632316589356,
      "mean_token_accuracy": 0.7373803712427616,
      "num_tokens": 35263682.0,
      "step": 17840
    },
    {
      "entropy": 0.9457804903388023,
      "epoch": 0.5112928404680406,
      "grad_norm": 7.9375,
      "learning_rate": 4.296338304570847e-05,
      "loss": 0.911676025390625,
      "mean_token_accuracy": 0.78768520206213,
      "num_tokens": 35282774.0,
      "step": 17850
    },
    {
      "entropy": 0.9700626829639077,
      "epoch": 0.5115792790341291,
      "grad_norm": 8.6875,
      "learning_rate": 4.295544445280487e-05,
      "loss": 0.9432881355285645,
      "mean_token_accuracy": 0.7733259946107864,
      "num_tokens": 35301524.0,
      "step": 17860
    },
    {
      "entropy": 1.0470176629722119,
      "epoch": 0.5118657176002177,
      "grad_norm": 6.71875,
      "learning_rate": 4.294750211865688e-05,
      "loss": 1.0479962348937988,
      "mean_token_accuracy": 0.7583012700080871,
      "num_tokens": 35320391.0,
      "step": 17870
    },
    {
      "entropy": 0.9766532107256353,
      "epoch": 0.5121521561663063,
      "grad_norm": 7.09375,
      "learning_rate": 4.29395560449194e-05,
      "loss": 0.9733241081237793,
      "mean_token_accuracy": 0.7783255320042372,
      "num_tokens": 35340401.0,
      "step": 17880
    },
    {
      "entropy": 0.9712235554121434,
      "epoch": 0.5124385947323947,
      "grad_norm": 7.3125,
      "learning_rate": 4.293160623324809e-05,
      "loss": 0.9993261337280274,
      "mean_token_accuracy": 0.773225960880518,
      "num_tokens": 35360345.0,
      "step": 17890
    },
    {
      "entropy": 0.8779443107545376,
      "epoch": 0.5127250332984833,
      "grad_norm": 6.375,
      "learning_rate": 4.2923652685299384e-05,
      "loss": 0.8596323013305665,
      "mean_token_accuracy": 0.7955453231930733,
      "num_tokens": 35380408.0,
      "step": 17900
    },
    {
      "entropy": 0.974922115728259,
      "epoch": 0.5130114718645719,
      "grad_norm": 5.9375,
      "learning_rate": 4.2915695402730515e-05,
      "loss": 0.9548707962036133,
      "mean_token_accuracy": 0.7789561577141285,
      "num_tokens": 35399306.0,
      "step": 17910
    },
    {
      "entropy": 0.9605494774878025,
      "epoch": 0.5132979104306604,
      "grad_norm": 8.0625,
      "learning_rate": 4.2907734387199484e-05,
      "loss": 0.9792910575866699,
      "mean_token_accuracy": 0.7734184537082911,
      "num_tokens": 35419934.0,
      "step": 17920
    },
    {
      "entropy": 0.9569711744785309,
      "epoch": 0.5135843489967489,
      "grad_norm": 7.28125,
      "learning_rate": 4.289976964036506e-05,
      "loss": 0.9944162368774414,
      "mean_token_accuracy": 0.7711521308869124,
      "num_tokens": 35440015.0,
      "step": 17930
    },
    {
      "entropy": 1.0997381042689085,
      "epoch": 0.5138707875628374,
      "grad_norm": 6.3125,
      "learning_rate": 4.289180116388681e-05,
      "loss": 1.1272847175598144,
      "mean_token_accuracy": 0.7588967274874449,
      "num_tokens": 35460865.0,
      "step": 17940
    },
    {
      "entropy": 1.0393663216382265,
      "epoch": 0.514157226128926,
      "grad_norm": 8.4375,
      "learning_rate": 4.2883828959425066e-05,
      "loss": 1.0994606971740724,
      "mean_token_accuracy": 0.7392833966761827,
      "num_tokens": 35480145.0,
      "step": 17950
    },
    {
      "entropy": 1.120609075948596,
      "epoch": 0.5144436646950146,
      "grad_norm": 8.0625,
      "learning_rate": 4.2875853028640934e-05,
      "loss": 1.1541407585144043,
      "mean_token_accuracy": 0.7382880769670009,
      "num_tokens": 35501089.0,
      "step": 17960
    },
    {
      "entropy": 1.0648599294945598,
      "epoch": 0.514730103261103,
      "grad_norm": 7.28125,
      "learning_rate": 4.28678733731963e-05,
      "loss": 1.002262020111084,
      "mean_token_accuracy": 0.7615043375641107,
      "num_tokens": 35521696.0,
      "step": 17970
    },
    {
      "entropy": 0.9882473286241293,
      "epoch": 0.5150165418271916,
      "grad_norm": 6.5,
      "learning_rate": 4.2859889994753836e-05,
      "loss": 1.0591958045959473,
      "mean_token_accuracy": 0.7614698220044375,
      "num_tokens": 35541679.0,
      "step": 17980
    },
    {
      "entropy": 1.0311295229941606,
      "epoch": 0.5153029803932802,
      "grad_norm": 5.71875,
      "learning_rate": 4.285190289497695e-05,
      "loss": 1.0270458221435548,
      "mean_token_accuracy": 0.7608290709555149,
      "num_tokens": 35561678.0,
      "step": 17990
    },
    {
      "entropy": 0.8813514014706015,
      "epoch": 0.5155894189593687,
      "grad_norm": 7.71875,
      "learning_rate": 4.284391207552989e-05,
      "loss": 0.8853710174560547,
      "mean_token_accuracy": 0.7864216700196266,
      "num_tokens": 35581198.0,
      "step": 18000
    },
    {
      "entropy": 1.0528824353590607,
      "epoch": 0.5158758575254573,
      "grad_norm": 7.34375,
      "learning_rate": 4.283591753807763e-05,
      "loss": 1.0953842163085938,
      "mean_token_accuracy": 0.7572290316224098,
      "num_tokens": 35601760.0,
      "step": 18010
    },
    {
      "entropy": 1.0299434075132012,
      "epoch": 0.5161622960915457,
      "grad_norm": 5.5625,
      "learning_rate": 4.2827919284285944e-05,
      "loss": 1.0040371894836426,
      "mean_token_accuracy": 0.7689940717071295,
      "num_tokens": 35622741.0,
      "step": 18020
    },
    {
      "entropy": 0.948230879381299,
      "epoch": 0.5164487346576343,
      "grad_norm": 10.1875,
      "learning_rate": 4.281991731582134e-05,
      "loss": 0.9695061683654785,
      "mean_token_accuracy": 0.7651624400168657,
      "num_tokens": 35643514.0,
      "step": 18030
    },
    {
      "entropy": 1.044832699932158,
      "epoch": 0.5167351732237229,
      "grad_norm": 5.09375,
      "learning_rate": 4.2811911634351166e-05,
      "loss": 1.1220834732055665,
      "mean_token_accuracy": 0.7488769918680191,
      "num_tokens": 35663795.0,
      "step": 18040
    },
    {
      "entropy": 0.9917106911540031,
      "epoch": 0.5170216117898114,
      "grad_norm": 6.59375,
      "learning_rate": 4.2803902241543484e-05,
      "loss": 0.9732969284057618,
      "mean_token_accuracy": 0.775698683410883,
      "num_tokens": 35684563.0,
      "step": 18050
    },
    {
      "entropy": 0.9947383690625429,
      "epoch": 0.5173080503558999,
      "grad_norm": 8.6875,
      "learning_rate": 4.279588913906716e-05,
      "loss": 1.0314928054809571,
      "mean_token_accuracy": 0.7632889442145825,
      "num_tokens": 35704766.0,
      "step": 18060
    },
    {
      "entropy": 0.9774407394230366,
      "epoch": 0.5175944889219884,
      "grad_norm": 6.28125,
      "learning_rate": 4.278787232859183e-05,
      "loss": 0.9978829383850097,
      "mean_token_accuracy": 0.7766224015504122,
      "num_tokens": 35725558.0,
      "step": 18070
    },
    {
      "entropy": 1.1714583218097687,
      "epoch": 0.517880927488077,
      "grad_norm": 9.0625,
      "learning_rate": 4.277985181178789e-05,
      "loss": 1.1521488189697267,
      "mean_token_accuracy": 0.7412251070141792,
      "num_tokens": 35746801.0,
      "step": 18080
    },
    {
      "entropy": 1.139555537328124,
      "epoch": 0.5181673660541656,
      "grad_norm": 6.65625,
      "learning_rate": 4.277182759032652e-05,
      "loss": 1.160375213623047,
      "mean_token_accuracy": 0.7425269648432732,
      "num_tokens": 35768945.0,
      "step": 18090
    },
    {
      "entropy": 1.0504832845181227,
      "epoch": 0.518453804620254,
      "grad_norm": 8.625,
      "learning_rate": 4.2763799665879653e-05,
      "loss": 1.083017921447754,
      "mean_token_accuracy": 0.7524487499147654,
      "num_tokens": 35787334.0,
      "step": 18100
    },
    {
      "entropy": 0.9651131779886782,
      "epoch": 0.5187402431863426,
      "grad_norm": 7.1875,
      "learning_rate": 4.275576804012004e-05,
      "loss": 1.0056843757629395,
      "mean_token_accuracy": 0.7749624900519848,
      "num_tokens": 35808102.0,
      "step": 18110
    },
    {
      "entropy": 1.0091121995821595,
      "epoch": 0.5190266817524312,
      "grad_norm": 7.8125,
      "learning_rate": 4.274773271472115e-05,
      "loss": 0.9769368171691895,
      "mean_token_accuracy": 0.7774313103407622,
      "num_tokens": 35827210.0,
      "step": 18120
    },
    {
      "entropy": 1.0063993139192462,
      "epoch": 0.5193131203185197,
      "grad_norm": 8.25,
      "learning_rate": 4.273969369135725e-05,
      "loss": 1.0084035873413086,
      "mean_token_accuracy": 0.7613523133099079,
      "num_tokens": 35846125.0,
      "step": 18130
    },
    {
      "entropy": 1.0476998765021563,
      "epoch": 0.5195995588846082,
      "grad_norm": 11.25,
      "learning_rate": 4.2731650971703376e-05,
      "loss": 1.0250536918640136,
      "mean_token_accuracy": 0.7664006385952234,
      "num_tokens": 35867176.0,
      "step": 18140
    },
    {
      "entropy": 1.013476044498384,
      "epoch": 0.5198859974506967,
      "grad_norm": 11.0625,
      "learning_rate": 4.272360455743532e-05,
      "loss": 1.030233669281006,
      "mean_token_accuracy": 0.7586410600692034,
      "num_tokens": 35886917.0,
      "step": 18150
    },
    {
      "entropy": 0.9217176243662835,
      "epoch": 0.5201724360167853,
      "grad_norm": 7.875,
      "learning_rate": 4.2715554450229666e-05,
      "loss": 0.9461106300354004,
      "mean_token_accuracy": 0.7936211973428726,
      "num_tokens": 35907845.0,
      "step": 18160
    },
    {
      "entropy": 0.9715134546160697,
      "epoch": 0.5204588745828739,
      "grad_norm": 9.5625,
      "learning_rate": 4.270750065176374e-05,
      "loss": 1.0033456802368164,
      "mean_token_accuracy": 0.7736696947366,
      "num_tokens": 35928257.0,
      "step": 18170
    },
    {
      "entropy": 0.9426901509985328,
      "epoch": 0.5207453131489623,
      "grad_norm": 4.71875,
      "learning_rate": 4.269944316371568e-05,
      "loss": 0.9362709045410156,
      "mean_token_accuracy": 0.7752047434449196,
      "num_tokens": 35950266.0,
      "step": 18180
    },
    {
      "entropy": 0.9634449645876885,
      "epoch": 0.5210317517150509,
      "grad_norm": 9.1875,
      "learning_rate": 4.2691381987764345e-05,
      "loss": 0.9871819496154786,
      "mean_token_accuracy": 0.7798014666885138,
      "num_tokens": 35969149.0,
      "step": 18190
    },
    {
      "entropy": 0.927098448947072,
      "epoch": 0.5213181902811395,
      "grad_norm": 8.75,
      "learning_rate": 4.2683317125589384e-05,
      "loss": 0.9316041946411133,
      "mean_token_accuracy": 0.7911450002342463,
      "num_tokens": 35988402.0,
      "step": 18200
    },
    {
      "entropy": 0.9820064989849925,
      "epoch": 0.521604628847228,
      "grad_norm": 8.125,
      "learning_rate": 4.267524857887122e-05,
      "loss": 0.9912860870361329,
      "mean_token_accuracy": 0.7589217390865087,
      "num_tokens": 36008375.0,
      "step": 18210
    },
    {
      "entropy": 1.061489969305694,
      "epoch": 0.5218910674133165,
      "grad_norm": 8.875,
      "learning_rate": 4.2667176349291035e-05,
      "loss": 1.0599629402160644,
      "mean_token_accuracy": 0.7619328621774912,
      "num_tokens": 36029129.0,
      "step": 18220
    },
    {
      "entropy": 0.9506415290758014,
      "epoch": 0.522177505979405,
      "grad_norm": 7.8125,
      "learning_rate": 4.2659100438530776e-05,
      "loss": 0.9441619873046875,
      "mean_token_accuracy": 0.7747184585779905,
      "num_tokens": 36049270.0,
      "step": 18230
    },
    {
      "entropy": 0.9565562393516303,
      "epoch": 0.5224639445454936,
      "grad_norm": 7.65625,
      "learning_rate": 4.265102084827316e-05,
      "loss": 0.9955384254455566,
      "mean_token_accuracy": 0.7661261264234781,
      "num_tokens": 36070160.0,
      "step": 18240
    },
    {
      "entropy": 0.9518299348652363,
      "epoch": 0.5227503831115822,
      "grad_norm": 6.28125,
      "learning_rate": 4.264293758020169e-05,
      "loss": 0.9975738525390625,
      "mean_token_accuracy": 0.7728760436177253,
      "num_tokens": 36090296.0,
      "step": 18250
    },
    {
      "entropy": 1.0898528631776572,
      "epoch": 0.5230368216776706,
      "grad_norm": 10.5,
      "learning_rate": 4.26348506360006e-05,
      "loss": 1.0874303817749023,
      "mean_token_accuracy": 0.7389911904931068,
      "num_tokens": 36109871.0,
      "step": 18260
    },
    {
      "entropy": 0.9815294438973069,
      "epoch": 0.5233232602437592,
      "grad_norm": 9.3125,
      "learning_rate": 4.2626760017354906e-05,
      "loss": 0.9133700370788574,
      "mean_token_accuracy": 0.7874503288418054,
      "num_tokens": 36130086.0,
      "step": 18270
    },
    {
      "entropy": 0.9049686649814248,
      "epoch": 0.5236096988098478,
      "grad_norm": 9.25,
      "learning_rate": 4.2618665725950404e-05,
      "loss": 0.9051311492919922,
      "mean_token_accuracy": 0.779395766928792,
      "num_tokens": 36148109.0,
      "step": 18280
    },
    {
      "entropy": 1.0278479263186455,
      "epoch": 0.5238961373759363,
      "grad_norm": 9.5,
      "learning_rate": 4.261056776347363e-05,
      "loss": 1.0871723175048829,
      "mean_token_accuracy": 0.7575385503470897,
      "num_tokens": 36167823.0,
      "step": 18290
    },
    {
      "entropy": 1.035915617272258,
      "epoch": 0.5241825759420249,
      "grad_norm": 6.875,
      "learning_rate": 4.2602466131611896e-05,
      "loss": 1.0602495193481445,
      "mean_token_accuracy": 0.7688746079802513,
      "num_tokens": 36188192.0,
      "step": 18300
    },
    {
      "entropy": 1.0230322202667594,
      "epoch": 0.5244690145081133,
      "grad_norm": 6.59375,
      "learning_rate": 4.2594360832053296e-05,
      "loss": 1.0000368118286134,
      "mean_token_accuracy": 0.7697128526866436,
      "num_tokens": 36207249.0,
      "step": 18310
    },
    {
      "entropy": 0.9453598963096738,
      "epoch": 0.5247554530742019,
      "grad_norm": 8.375,
      "learning_rate": 4.258625186648666e-05,
      "loss": 0.9622668266296387,
      "mean_token_accuracy": 0.7733538698405027,
      "num_tokens": 36227975.0,
      "step": 18320
    },
    {
      "entropy": 0.9114329120144248,
      "epoch": 0.5250418916402905,
      "grad_norm": 7.6875,
      "learning_rate": 4.257813923660159e-05,
      "loss": 0.9374043464660644,
      "mean_token_accuracy": 0.7840469803661108,
      "num_tokens": 36246801.0,
      "step": 18330
    },
    {
      "entropy": 1.0108552426099777,
      "epoch": 0.525328330206379,
      "grad_norm": 9.3125,
      "learning_rate": 4.257002294408846e-05,
      "loss": 1.0148914337158204,
      "mean_token_accuracy": 0.7724619287997484,
      "num_tokens": 36264935.0,
      "step": 18340
    },
    {
      "entropy": 0.9624065946787596,
      "epoch": 0.5256147687724675,
      "grad_norm": 7.59375,
      "learning_rate": 4.25619029906384e-05,
      "loss": 0.9647127151489258,
      "mean_token_accuracy": 0.7750409301370382,
      "num_tokens": 36284127.0,
      "step": 18350
    },
    {
      "entropy": 0.8979532336816192,
      "epoch": 0.525901207338556,
      "grad_norm": 9.25,
      "learning_rate": 4.255377937794332e-05,
      "loss": 0.9485130310058594,
      "mean_token_accuracy": 0.7733627073466778,
      "num_tokens": 36304309.0,
      "step": 18360
    },
    {
      "entropy": 0.924005289003253,
      "epoch": 0.5261876459046446,
      "grad_norm": 9.375,
      "learning_rate": 4.254565210769585e-05,
      "loss": 0.9595027923583984,
      "mean_token_accuracy": 0.7733161136507988,
      "num_tokens": 36323623.0,
      "step": 18370
    },
    {
      "entropy": 1.0779311092570425,
      "epoch": 0.5264740844707332,
      "grad_norm": 7.25,
      "learning_rate": 4.253752118158943e-05,
      "loss": 1.074887752532959,
      "mean_token_accuracy": 0.7523600362241268,
      "num_tokens": 36343963.0,
      "step": 18380
    },
    {
      "entropy": 0.9568657351657748,
      "epoch": 0.5267605230368216,
      "grad_norm": 6.25,
      "learning_rate": 4.2529386601318245e-05,
      "loss": 0.9681062698364258,
      "mean_token_accuracy": 0.7755594749003649,
      "num_tokens": 36363446.0,
      "step": 18390
    },
    {
      "entropy": 0.927198201045394,
      "epoch": 0.5270469616029102,
      "grad_norm": 6.375,
      "learning_rate": 4.2521248368577226e-05,
      "loss": 0.9500542640686035,
      "mean_token_accuracy": 0.7756378091871738,
      "num_tokens": 36383431.0,
      "step": 18400
    },
    {
      "entropy": 1.0249531768262385,
      "epoch": 0.5273334001689988,
      "grad_norm": 9.3125,
      "learning_rate": 4.251310648506208e-05,
      "loss": 1.017384910583496,
      "mean_token_accuracy": 0.7633178159594536,
      "num_tokens": 36402698.0,
      "step": 18410
    },
    {
      "entropy": 0.9823536445386708,
      "epoch": 0.5276198387350873,
      "grad_norm": 5.75,
      "learning_rate": 4.250496095246926e-05,
      "loss": 0.9748650550842285,
      "mean_token_accuracy": 0.7732032962143421,
      "num_tokens": 36423221.0,
      "step": 18420
    },
    {
      "entropy": 0.9957347840070725,
      "epoch": 0.5279062773011758,
      "grad_norm": 8.3125,
      "learning_rate": 4.249681177249602e-05,
      "loss": 1.0128029823303222,
      "mean_token_accuracy": 0.7662801094353199,
      "num_tokens": 36444056.0,
      "step": 18430
    },
    {
      "entropy": 0.9218484982848167,
      "epoch": 0.5281927158672644,
      "grad_norm": 10.0,
      "learning_rate": 4.248865894684032e-05,
      "loss": 0.9279016494750977,
      "mean_token_accuracy": 0.7783108800649643,
      "num_tokens": 36465410.0,
      "step": 18440
    },
    {
      "entropy": 0.8914625532925129,
      "epoch": 0.5284791544333529,
      "grad_norm": 6.1875,
      "learning_rate": 4.248050247720091e-05,
      "loss": 0.9414989471435546,
      "mean_token_accuracy": 0.7907150596380234,
      "num_tokens": 36483713.0,
      "step": 18450
    },
    {
      "entropy": 0.8530762501060962,
      "epoch": 0.5287655929994415,
      "grad_norm": 7.125,
      "learning_rate": 4.2472342365277305e-05,
      "loss": 0.9261279106140137,
      "mean_token_accuracy": 0.7838302284479142,
      "num_tokens": 36503705.0,
      "step": 18460
    },
    {
      "entropy": 0.9298483902588487,
      "epoch": 0.5290520315655299,
      "grad_norm": 6.59375,
      "learning_rate": 4.2464178612769755e-05,
      "loss": 0.9499112129211426,
      "mean_token_accuracy": 0.7738996341824531,
      "num_tokens": 36523486.0,
      "step": 18470
    },
    {
      "entropy": 1.034496882930398,
      "epoch": 0.5293384701316185,
      "grad_norm": 8.9375,
      "learning_rate": 4.245601122137928e-05,
      "loss": 1.0376147270202636,
      "mean_token_accuracy": 0.7546425886452198,
      "num_tokens": 36542986.0,
      "step": 18480
    },
    {
      "entropy": 1.0742289628833532,
      "epoch": 0.5296249086977071,
      "grad_norm": 8.875,
      "learning_rate": 4.244784019280768e-05,
      "loss": 1.068143367767334,
      "mean_token_accuracy": 0.7624827574938535,
      "num_tokens": 36561163.0,
      "step": 18490
    },
    {
      "entropy": 0.9775538988411426,
      "epoch": 0.5299113472637956,
      "grad_norm": 7.28125,
      "learning_rate": 4.2439665528757455e-05,
      "loss": 1.0020852088928223,
      "mean_token_accuracy": 0.768316762894392,
      "num_tokens": 36580890.0,
      "step": 18500
    },
    {
      "entropy": 1.0095560170710087,
      "epoch": 0.5301977858298841,
      "grad_norm": 6.28125,
      "learning_rate": 4.243148723093193e-05,
      "loss": 1.0097763061523437,
      "mean_token_accuracy": 0.7664507023990155,
      "num_tokens": 36600801.0,
      "step": 18510
    },
    {
      "entropy": 0.9847262201830744,
      "epoch": 0.5304842243959726,
      "grad_norm": 8.25,
      "learning_rate": 4.2423305301035155e-05,
      "loss": 1.0043015480041504,
      "mean_token_accuracy": 0.7763553272932768,
      "num_tokens": 36621000.0,
      "step": 18520
    },
    {
      "entropy": 0.89938645940274,
      "epoch": 0.5307706629620612,
      "grad_norm": 8.75,
      "learning_rate": 4.241511974077193e-05,
      "loss": 0.8604474067687988,
      "mean_token_accuracy": 0.7922192800790071,
      "num_tokens": 36641716.0,
      "step": 18530
    },
    {
      "entropy": 1.1025999668985604,
      "epoch": 0.5310571015281498,
      "grad_norm": 8.0625,
      "learning_rate": 4.2406930551847824e-05,
      "loss": 1.141490650177002,
      "mean_token_accuracy": 0.7465562656521797,
      "num_tokens": 36661666.0,
      "step": 18540
    },
    {
      "entropy": 0.9981026839464903,
      "epoch": 0.5313435400942383,
      "grad_norm": 6.3125,
      "learning_rate": 4.239873773596915e-05,
      "loss": 0.9825703620910644,
      "mean_token_accuracy": 0.7621681142598391,
      "num_tokens": 36683091.0,
      "step": 18550
    },
    {
      "entropy": 1.089583594724536,
      "epoch": 0.5316299786603268,
      "grad_norm": 8.6875,
      "learning_rate": 4.2390541294843006e-05,
      "loss": 1.1002711296081542,
      "mean_token_accuracy": 0.7479295961558818,
      "num_tokens": 36703983.0,
      "step": 18560
    },
    {
      "entropy": 1.0272770265117288,
      "epoch": 0.5319164172264154,
      "grad_norm": 8.375,
      "learning_rate": 4.23823412301772e-05,
      "loss": 1.020514965057373,
      "mean_token_accuracy": 0.7728347435593605,
      "num_tokens": 36723679.0,
      "step": 18570
    },
    {
      "entropy": 0.9897090133279562,
      "epoch": 0.5322028557925039,
      "grad_norm": 13.0625,
      "learning_rate": 4.2374137543680334e-05,
      "loss": 1.052346897125244,
      "mean_token_accuracy": 0.7739234142005443,
      "num_tokens": 36743396.0,
      "step": 18580
    },
    {
      "entropy": 0.8769473440945148,
      "epoch": 0.5324892943585925,
      "grad_norm": 8.1875,
      "learning_rate": 4.236593023706175e-05,
      "loss": 0.8517210960388184,
      "mean_token_accuracy": 0.8023064389824868,
      "num_tokens": 36761480.0,
      "step": 18590
    },
    {
      "entropy": 0.9887404832988977,
      "epoch": 0.532775732924681,
      "grad_norm": 9.875,
      "learning_rate": 4.235771931203153e-05,
      "loss": 1.038399887084961,
      "mean_token_accuracy": 0.748694383725524,
      "num_tokens": 36782200.0,
      "step": 18600
    },
    {
      "entropy": 0.9543918523937464,
      "epoch": 0.5330621714907695,
      "grad_norm": 7.8125,
      "learning_rate": 4.234950477030054e-05,
      "loss": 1.0050990104675293,
      "mean_token_accuracy": 0.7775580734014511,
      "num_tokens": 36801432.0,
      "step": 18610
    },
    {
      "entropy": 0.9844637623988092,
      "epoch": 0.5333486100568581,
      "grad_norm": 7.28125,
      "learning_rate": 4.2341286613580376e-05,
      "loss": 0.9462395668029785,
      "mean_token_accuracy": 0.775806350633502,
      "num_tokens": 36820482.0,
      "step": 18620
    },
    {
      "entropy": 0.9975185271352529,
      "epoch": 0.5336350486229466,
      "grad_norm": 8.5,
      "learning_rate": 4.23330648435834e-05,
      "loss": 1.0078046798706055,
      "mean_token_accuracy": 0.7637744177132845,
      "num_tokens": 36841022.0,
      "step": 18630
    },
    {
      "entropy": 0.8633401485159993,
      "epoch": 0.5339214871890351,
      "grad_norm": 7.09375,
      "learning_rate": 4.232483946202271e-05,
      "loss": 0.8383587837219239,
      "mean_token_accuracy": 0.802937027066946,
      "num_tokens": 36859389.0,
      "step": 18640
    },
    {
      "entropy": 0.9264299492351711,
      "epoch": 0.5342079257551237,
      "grad_norm": 9.875,
      "learning_rate": 4.231661047061219e-05,
      "loss": 0.9155117034912109,
      "mean_token_accuracy": 0.7900910649448634,
      "num_tokens": 36878149.0,
      "step": 18650
    },
    {
      "entropy": 1.007754699885845,
      "epoch": 0.5344943643212122,
      "grad_norm": 9.1875,
      "learning_rate": 4.230837787106644e-05,
      "loss": 1.035257911682129,
      "mean_token_accuracy": 0.7732655949890613,
      "num_tokens": 36899380.0,
      "step": 18660
    },
    {
      "entropy": 0.9002570047974586,
      "epoch": 0.5347808028873008,
      "grad_norm": 5.40625,
      "learning_rate": 4.230014166510083e-05,
      "loss": 0.8765363693237305,
      "mean_token_accuracy": 0.8054404400289059,
      "num_tokens": 36919593.0,
      "step": 18670
    },
    {
      "entropy": 0.9912328207865357,
      "epoch": 0.5350672414533892,
      "grad_norm": 9.8125,
      "learning_rate": 4.229190185443148e-05,
      "loss": 1.0469957351684571,
      "mean_token_accuracy": 0.7630418248474597,
      "num_tokens": 36939042.0,
      "step": 18680
    },
    {
      "entropy": 1.110708248242736,
      "epoch": 0.5353536800194778,
      "grad_norm": 9.5,
      "learning_rate": 4.228365844077525e-05,
      "loss": 1.1565019607543945,
      "mean_token_accuracy": 0.7433866772800684,
      "num_tokens": 36958277.0,
      "step": 18690
    },
    {
      "entropy": 1.0424179248511791,
      "epoch": 0.5356401185855664,
      "grad_norm": 7.15625,
      "learning_rate": 4.2275411425849774e-05,
      "loss": 1.0775314331054688,
      "mean_token_accuracy": 0.754136311635375,
      "num_tokens": 36977590.0,
      "step": 18700
    },
    {
      "entropy": 1.0898970341309906,
      "epoch": 0.5359265571516549,
      "grad_norm": 6.5625,
      "learning_rate": 4.226716081137341e-05,
      "loss": 1.0423570632934571,
      "mean_token_accuracy": 0.7628885980695486,
      "num_tokens": 36997352.0,
      "step": 18710
    },
    {
      "entropy": 0.9698012646287679,
      "epoch": 0.5362129957177434,
      "grad_norm": 6.65625,
      "learning_rate": 4.225890659906527e-05,
      "loss": 0.926824951171875,
      "mean_token_accuracy": 0.7787880446761847,
      "num_tokens": 37016913.0,
      "step": 18720
    },
    {
      "entropy": 0.9564575517550111,
      "epoch": 0.536499434283832,
      "grad_norm": 10.25,
      "learning_rate": 4.225064879064525e-05,
      "loss": 0.9458847999572754,
      "mean_token_accuracy": 0.7774242050945759,
      "num_tokens": 37036009.0,
      "step": 18730
    },
    {
      "entropy": 0.9452911943197251,
      "epoch": 0.5367858728499205,
      "grad_norm": 6.15625,
      "learning_rate": 4.224238738783395e-05,
      "loss": 0.9479838371276855,
      "mean_token_accuracy": 0.7813715692609549,
      "num_tokens": 37055949.0,
      "step": 18740
    },
    {
      "entropy": 0.8963361142203212,
      "epoch": 0.5370723114160091,
      "grad_norm": 7.375,
      "learning_rate": 4.223412239235274e-05,
      "loss": 0.9162420272827149,
      "mean_token_accuracy": 0.8018769826740026,
      "num_tokens": 37075289.0,
      "step": 18750
    },
    {
      "entropy": 0.9612562710419297,
      "epoch": 0.5373587499820975,
      "grad_norm": 5.59375,
      "learning_rate": 4.2225853805923745e-05,
      "loss": 1.0085434913635254,
      "mean_token_accuracy": 0.7756607297807931,
      "num_tokens": 37097342.0,
      "step": 18760
    },
    {
      "entropy": 0.9166711486876011,
      "epoch": 0.5376451885481861,
      "grad_norm": 9.125,
      "learning_rate": 4.221758163026981e-05,
      "loss": 0.905830192565918,
      "mean_token_accuracy": 0.7827754568308591,
      "num_tokens": 37115293.0,
      "step": 18770
    },
    {
      "entropy": 0.8870739087462425,
      "epoch": 0.5379316271142747,
      "grad_norm": 7.03125,
      "learning_rate": 4.220930586711458e-05,
      "loss": 0.8888262748718262,
      "mean_token_accuracy": 0.787341770529747,
      "num_tokens": 37134709.0,
      "step": 18780
    },
    {
      "entropy": 0.9866439955309033,
      "epoch": 0.5382180656803632,
      "grad_norm": 5.5,
      "learning_rate": 4.220102651818238e-05,
      "loss": 1.034498405456543,
      "mean_token_accuracy": 0.7656238980591297,
      "num_tokens": 37154779.0,
      "step": 18790
    },
    {
      "entropy": 0.9630633728578687,
      "epoch": 0.5385045042464517,
      "grad_norm": 8.375,
      "learning_rate": 4.2192743585198344e-05,
      "loss": 0.9901782035827636,
      "mean_token_accuracy": 0.7748003490269184,
      "num_tokens": 37174711.0,
      "step": 18800
    },
    {
      "entropy": 0.9653576597571373,
      "epoch": 0.5387909428125403,
      "grad_norm": 6.09375,
      "learning_rate": 4.2184457069888314e-05,
      "loss": 0.9996882438659668,
      "mean_token_accuracy": 0.769518369063735,
      "num_tokens": 37194659.0,
      "step": 18810
    },
    {
      "entropy": 0.968729081004858,
      "epoch": 0.5390773813786288,
      "grad_norm": 8.0,
      "learning_rate": 4.217616697397889e-05,
      "loss": 0.9794042587280274,
      "mean_token_accuracy": 0.7784108947962523,
      "num_tokens": 37213539.0,
      "step": 18820
    },
    {
      "entropy": 0.9478037467226386,
      "epoch": 0.5393638199447174,
      "grad_norm": 7.65625,
      "learning_rate": 4.216787329919743e-05,
      "loss": 0.9019865036010742,
      "mean_token_accuracy": 0.7855732407420873,
      "num_tokens": 37232021.0,
      "step": 18830
    },
    {
      "entropy": 0.9927759733051061,
      "epoch": 0.539650258510806,
      "grad_norm": 8.25,
      "learning_rate": 4.215957604727201e-05,
      "loss": 0.9580434799194336,
      "mean_token_accuracy": 0.7695677071809769,
      "num_tokens": 37251113.0,
      "step": 18840
    },
    {
      "entropy": 0.9095350941643119,
      "epoch": 0.5399366970768944,
      "grad_norm": 6.9375,
      "learning_rate": 4.215127521993147e-05,
      "loss": 0.8837380409240723,
      "mean_token_accuracy": 0.7993823975324631,
      "num_tokens": 37270749.0,
      "step": 18850
    },
    {
      "entropy": 0.8933297876268625,
      "epoch": 0.540223135642983,
      "grad_norm": 8.3125,
      "learning_rate": 4.21429708189054e-05,
      "loss": 0.9428502082824707,
      "mean_token_accuracy": 0.7757887851446867,
      "num_tokens": 37290630.0,
      "step": 18860
    },
    {
      "entropy": 0.9630610842257739,
      "epoch": 0.5405095742090715,
      "grad_norm": 6.34375,
      "learning_rate": 4.2134662845924125e-05,
      "loss": 0.9968518257141114,
      "mean_token_accuracy": 0.7729216940701008,
      "num_tokens": 37309595.0,
      "step": 18870
    },
    {
      "entropy": 0.9585330963134766,
      "epoch": 0.5407960127751601,
      "grad_norm": 6.09375,
      "learning_rate": 4.212635130271872e-05,
      "loss": 0.8919583320617676,
      "mean_token_accuracy": 0.7935737576335669,
      "num_tokens": 37328133.0,
      "step": 18880
    },
    {
      "entropy": 0.9751623921096325,
      "epoch": 0.5410824513412485,
      "grad_norm": 7.78125,
      "learning_rate": 4.211803619102099e-05,
      "loss": 1.0511135101318358,
      "mean_token_accuracy": 0.7501297175884247,
      "num_tokens": 37348991.0,
      "step": 18890
    },
    {
      "entropy": 0.8876290641725063,
      "epoch": 0.5413688899073371,
      "grad_norm": 7.40625,
      "learning_rate": 4.210971751256351e-05,
      "loss": 0.9202899932861328,
      "mean_token_accuracy": 0.78198076300323,
      "num_tokens": 37367649.0,
      "step": 18900
    },
    {
      "entropy": 0.9374773848801852,
      "epoch": 0.5416553284734257,
      "grad_norm": 9.1875,
      "learning_rate": 4.210139526907956e-05,
      "loss": 0.9501521110534668,
      "mean_token_accuracy": 0.7730865444988012,
      "num_tokens": 37386187.0,
      "step": 18910
    },
    {
      "entropy": 1.0318306596949696,
      "epoch": 0.5419417670395142,
      "grad_norm": 10.25,
      "learning_rate": 4.2093069462303204e-05,
      "loss": 1.0163016319274902,
      "mean_token_accuracy": 0.7591164406388998,
      "num_tokens": 37404127.0,
      "step": 18920
    },
    {
      "entropy": 1.013629556633532,
      "epoch": 0.5422282056056027,
      "grad_norm": 9.5625,
      "learning_rate": 4.208474009396922e-05,
      "loss": 1.0392515182495117,
      "mean_token_accuracy": 0.77359305806458,
      "num_tokens": 37423667.0,
      "step": 18930
    },
    {
      "entropy": 0.9632765204645694,
      "epoch": 0.5425146441716913,
      "grad_norm": 6.34375,
      "learning_rate": 4.2076407165813145e-05,
      "loss": 0.9784265518188476,
      "mean_token_accuracy": 0.78217060752213,
      "num_tokens": 37443153.0,
      "step": 18940
    },
    {
      "entropy": 0.9851489506661892,
      "epoch": 0.5428010827377798,
      "grad_norm": 8.1875,
      "learning_rate": 4.2068070679571236e-05,
      "loss": 0.9816110610961915,
      "mean_token_accuracy": 0.7683041960000991,
      "num_tokens": 37462492.0,
      "step": 18950
    },
    {
      "entropy": 0.9793066322803498,
      "epoch": 0.5430875213038684,
      "grad_norm": 6.21875,
      "learning_rate": 4.205973063698053e-05,
      "loss": 1.0025501251220703,
      "mean_token_accuracy": 0.7545711405575275,
      "num_tokens": 37481342.0,
      "step": 18960
    },
    {
      "entropy": 0.9695794058963656,
      "epoch": 0.5433739598699568,
      "grad_norm": 11.3125,
      "learning_rate": 4.205138703977875e-05,
      "loss": 0.9330293655395507,
      "mean_token_accuracy": 0.7800051450729371,
      "num_tokens": 37501601.0,
      "step": 18970
    },
    {
      "entropy": 0.9861360523849726,
      "epoch": 0.5436603984360454,
      "grad_norm": 8.4375,
      "learning_rate": 4.2043039889704404e-05,
      "loss": 1.0017213821411133,
      "mean_token_accuracy": 0.7645394187420607,
      "num_tokens": 37521079.0,
      "step": 18980
    },
    {
      "entropy": 0.9288894783705473,
      "epoch": 0.543946837002134,
      "grad_norm": 12.5625,
      "learning_rate": 4.203468918849674e-05,
      "loss": 0.9886000633239747,
      "mean_token_accuracy": 0.7773781221359968,
      "num_tokens": 37541178.0,
      "step": 18990
    },
    {
      "entropy": 0.9507323738187552,
      "epoch": 0.5442332755682225,
      "grad_norm": 5.875,
      "learning_rate": 4.202633493789572e-05,
      "loss": 0.9986387252807617,
      "mean_token_accuracy": 0.7629520568996668,
      "num_tokens": 37560925.0,
      "step": 19000
    },
    {
      "entropy": 0.9584133286029101,
      "epoch": 0.544519714134311,
      "grad_norm": 7.9375,
      "learning_rate": 4.201797713964205e-05,
      "loss": 0.9792420387268066,
      "mean_token_accuracy": 0.7746828719973564,
      "num_tokens": 37579778.0,
      "step": 19010
    },
    {
      "entropy": 1.126902387291193,
      "epoch": 0.5448061527003996,
      "grad_norm": 8.4375,
      "learning_rate": 4.200961579547719e-05,
      "loss": 1.1250021934509278,
      "mean_token_accuracy": 0.7491728439927101,
      "num_tokens": 37599411.0,
      "step": 19020
    },
    {
      "entropy": 0.9646470233798027,
      "epoch": 0.5450925912664881,
      "grad_norm": 9.1875,
      "learning_rate": 4.200125090714334e-05,
      "loss": 0.9210112571716309,
      "mean_token_accuracy": 0.7871640287339687,
      "num_tokens": 37618961.0,
      "step": 19030
    },
    {
      "entropy": 0.9305791702121496,
      "epoch": 0.5453790298325767,
      "grad_norm": 10.4375,
      "learning_rate": 4.1992882476383425e-05,
      "loss": 0.9539770126342774,
      "mean_token_accuracy": 0.7653025042265653,
      "num_tokens": 37637923.0,
      "step": 19040
    },
    {
      "entropy": 1.0297860287129879,
      "epoch": 0.5456654683986651,
      "grad_norm": 6.8125,
      "learning_rate": 4.198451050494111e-05,
      "loss": 1.0537588119506835,
      "mean_token_accuracy": 0.7532337259501218,
      "num_tokens": 37658554.0,
      "step": 19050
    },
    {
      "entropy": 1.0110898189246655,
      "epoch": 0.5459519069647537,
      "grad_norm": 8.125,
      "learning_rate": 4.1976134994560804e-05,
      "loss": 1.037745761871338,
      "mean_token_accuracy": 0.7760591752827167,
      "num_tokens": 37679574.0,
      "step": 19060
    },
    {
      "entropy": 0.9703520519658924,
      "epoch": 0.5462383455308423,
      "grad_norm": 8.0625,
      "learning_rate": 4.1967755946987645e-05,
      "loss": 0.9784976959228515,
      "mean_token_accuracy": 0.7692656494677067,
      "num_tokens": 37699798.0,
      "step": 19070
    },
    {
      "entropy": 0.8735691886395216,
      "epoch": 0.5465247840969308,
      "grad_norm": 6.28125,
      "learning_rate": 4.1959373363967526e-05,
      "loss": 0.8076004981994629,
      "mean_token_accuracy": 0.8089750308543444,
      "num_tokens": 37718769.0,
      "step": 19080
    },
    {
      "entropy": 0.9451670184731483,
      "epoch": 0.5468112226630194,
      "grad_norm": 9.3125,
      "learning_rate": 4.195098724724705e-05,
      "loss": 0.9979077339172363,
      "mean_token_accuracy": 0.7636483337730169,
      "num_tokens": 37738502.0,
      "step": 19090
    },
    {
      "entropy": 0.9564072618260979,
      "epoch": 0.5470976612291079,
      "grad_norm": 8.3125,
      "learning_rate": 4.1942597598573585e-05,
      "loss": 1.0043194770812989,
      "mean_token_accuracy": 0.7660236690193415,
      "num_tokens": 37756768.0,
      "step": 19100
    },
    {
      "entropy": 1.0382267966866494,
      "epoch": 0.5473840997951964,
      "grad_norm": 7.53125,
      "learning_rate": 4.1934204419695223e-05,
      "loss": 1.1136626243591308,
      "mean_token_accuracy": 0.7422887045890093,
      "num_tokens": 37776520.0,
      "step": 19110
    },
    {
      "entropy": 0.9455624375492334,
      "epoch": 0.547670538361285,
      "grad_norm": 6.15625,
      "learning_rate": 4.192580771236076e-05,
      "loss": 0.9825914382934571,
      "mean_token_accuracy": 0.7808877781033516,
      "num_tokens": 37796170.0,
      "step": 19120
    },
    {
      "entropy": 0.9649914403446018,
      "epoch": 0.5479569769273736,
      "grad_norm": 9.5625,
      "learning_rate": 4.1917407478319795e-05,
      "loss": 0.9664751052856445,
      "mean_token_accuracy": 0.7780234958976507,
      "num_tokens": 37815460.0,
      "step": 19130
    },
    {
      "entropy": 1.0194987587630748,
      "epoch": 0.548243415493462,
      "grad_norm": 8.0,
      "learning_rate": 4.1909003719322594e-05,
      "loss": 1.0122082710266114,
      "mean_token_accuracy": 0.772532419487834,
      "num_tokens": 37834789.0,
      "step": 19140
    },
    {
      "entropy": 0.9845464492216707,
      "epoch": 0.5485298540595506,
      "grad_norm": 8.5625,
      "learning_rate": 4.19005964371202e-05,
      "loss": 0.9775012016296387,
      "mean_token_accuracy": 0.7709212206304074,
      "num_tokens": 37853560.0,
      "step": 19150
    },
    {
      "entropy": 1.0180983047932386,
      "epoch": 0.5488162926256391,
      "grad_norm": 9.9375,
      "learning_rate": 4.1892185633464375e-05,
      "loss": 1.027649211883545,
      "mean_token_accuracy": 0.7632421582937241,
      "num_tokens": 37872336.0,
      "step": 19160
    },
    {
      "entropy": 0.9963744878768921,
      "epoch": 0.5491027311917277,
      "grad_norm": 8.5625,
      "learning_rate": 4.188377131010762e-05,
      "loss": 1.0095620155334473,
      "mean_token_accuracy": 0.7622698973864317,
      "num_tokens": 37890658.0,
      "step": 19170
    },
    {
      "entropy": 1.0037853233516216,
      "epoch": 0.5493891697578162,
      "grad_norm": 6.0,
      "learning_rate": 4.187535346880316e-05,
      "loss": 1.0324936866760255,
      "mean_token_accuracy": 0.767413829267025,
      "num_tokens": 37910269.0,
      "step": 19180
    },
    {
      "entropy": 0.9974068082869053,
      "epoch": 0.5496756083239047,
      "grad_norm": 8.375,
      "learning_rate": 4.186693211130495e-05,
      "loss": 0.9621203422546387,
      "mean_token_accuracy": 0.7779409468173981,
      "num_tokens": 37931342.0,
      "step": 19190
    },
    {
      "entropy": 1.0015677127987146,
      "epoch": 0.5499620468899933,
      "grad_norm": 10.3125,
      "learning_rate": 4.185850723936771e-05,
      "loss": 0.9944249153137207,
      "mean_token_accuracy": 0.7711773902177811,
      "num_tokens": 37950287.0,
      "step": 19200
    },
    {
      "entropy": 0.8902222461998462,
      "epoch": 0.5502484854560818,
      "grad_norm": 7.34375,
      "learning_rate": 4.185007885474685e-05,
      "loss": 0.9176956176757812,
      "mean_token_accuracy": 0.7782740302383899,
      "num_tokens": 37969487.0,
      "step": 19210
    },
    {
      "entropy": 0.9460141375660897,
      "epoch": 0.5505349240221703,
      "grad_norm": 7.125,
      "learning_rate": 4.1841646959198536e-05,
      "loss": 1.009310817718506,
      "mean_token_accuracy": 0.7705666992813349,
      "num_tokens": 37991343.0,
      "step": 19220
    },
    {
      "entropy": 0.9415726711973548,
      "epoch": 0.5508213625882589,
      "grad_norm": 7.28125,
      "learning_rate": 4.1833211554479654e-05,
      "loss": 0.988035774230957,
      "mean_token_accuracy": 0.7750967808067799,
      "num_tokens": 38011386.0,
      "step": 19230
    },
    {
      "entropy": 0.9988466555252671,
      "epoch": 0.5511078011543474,
      "grad_norm": 6.71875,
      "learning_rate": 4.182477264234785e-05,
      "loss": 1.0462284088134766,
      "mean_token_accuracy": 0.7617311853915453,
      "num_tokens": 38030083.0,
      "step": 19240
    },
    {
      "entropy": 0.917192343249917,
      "epoch": 0.551394239720436,
      "grad_norm": 7.0625,
      "learning_rate": 4.1816330224561455e-05,
      "loss": 0.9060014724731446,
      "mean_token_accuracy": 0.7855589151382446,
      "num_tokens": 38049742.0,
      "step": 19250
    },
    {
      "entropy": 1.0207097191363572,
      "epoch": 0.5516806782865245,
      "grad_norm": 8.0625,
      "learning_rate": 4.180788430287956e-05,
      "loss": 0.9629178047180176,
      "mean_token_accuracy": 0.7734496343880892,
      "num_tokens": 38070049.0,
      "step": 19260
    },
    {
      "entropy": 0.9937358295544982,
      "epoch": 0.551967116852613,
      "grad_norm": 5.96875,
      "learning_rate": 4.1799434879061985e-05,
      "loss": 0.980595874786377,
      "mean_token_accuracy": 0.7729320023208857,
      "num_tokens": 38089702.0,
      "step": 19270
    },
    {
      "entropy": 0.9689991388469934,
      "epoch": 0.5522535554187016,
      "grad_norm": 7.25,
      "learning_rate": 4.1790981954869276e-05,
      "loss": 0.9846927642822265,
      "mean_token_accuracy": 0.7769508343189955,
      "num_tokens": 38108614.0,
      "step": 19280
    },
    {
      "entropy": 0.9059847016818822,
      "epoch": 0.5525399939847901,
      "grad_norm": 8.625,
      "learning_rate": 4.1782525532062686e-05,
      "loss": 0.9599862098693848,
      "mean_token_accuracy": 0.7705480575561523,
      "num_tokens": 38127987.0,
      "step": 19290
    },
    {
      "entropy": 1.0685237763449549,
      "epoch": 0.5528264325508786,
      "grad_norm": 7.0625,
      "learning_rate": 4.177406561240425e-05,
      "loss": 1.0888726234436035,
      "mean_token_accuracy": 0.7567665204405785,
      "num_tokens": 38148289.0,
      "step": 19300
    },
    {
      "entropy": 1.058514145296067,
      "epoch": 0.5531128711169672,
      "grad_norm": 11.6875,
      "learning_rate": 4.1765602197656674e-05,
      "loss": 1.0787226676940918,
      "mean_token_accuracy": 0.7461427260190249,
      "num_tokens": 38167344.0,
      "step": 19310
    },
    {
      "entropy": 0.9547070179134607,
      "epoch": 0.5533993096830557,
      "grad_norm": 6.21875,
      "learning_rate": 4.1757135289583424e-05,
      "loss": 0.9950839042663574,
      "mean_token_accuracy": 0.7731689371168613,
      "num_tokens": 38185824.0,
      "step": 19320
    },
    {
      "entropy": 1.0977138634771109,
      "epoch": 0.5536857482491443,
      "grad_norm": 6.4375,
      "learning_rate": 4.17486648899487e-05,
      "loss": 1.0839117050170899,
      "mean_token_accuracy": 0.7628477077931166,
      "num_tokens": 38205884.0,
      "step": 19330
    },
    {
      "entropy": 0.8991302141919733,
      "epoch": 0.5539721868152327,
      "grad_norm": 6.09375,
      "learning_rate": 4.17401910005174e-05,
      "loss": 0.9074352264404297,
      "mean_token_accuracy": 0.7849322475492955,
      "num_tokens": 38225748.0,
      "step": 19340
    },
    {
      "entropy": 0.98412509188056,
      "epoch": 0.5542586253813213,
      "grad_norm": 7.65625,
      "learning_rate": 4.173171362305518e-05,
      "loss": 1.0381229400634766,
      "mean_token_accuracy": 0.7625836465507746,
      "num_tokens": 38243976.0,
      "step": 19350
    },
    {
      "entropy": 0.9593858283013106,
      "epoch": 0.5545450639474099,
      "grad_norm": 8.375,
      "learning_rate": 4.172323275932839e-05,
      "loss": 0.8908781051635742,
      "mean_token_accuracy": 0.7914481859654188,
      "num_tokens": 38264899.0,
      "step": 19360
    },
    {
      "entropy": 0.9918803742155433,
      "epoch": 0.5548315025134984,
      "grad_norm": 6.9375,
      "learning_rate": 4.1714748411104155e-05,
      "loss": 0.9486532211303711,
      "mean_token_accuracy": 0.7768490660935641,
      "num_tokens": 38285152.0,
      "step": 19370
    },
    {
      "entropy": 0.9816761322319507,
      "epoch": 0.555117941079587,
      "grad_norm": 6.46875,
      "learning_rate": 4.170626058015027e-05,
      "loss": 0.9834271430969238,
      "mean_token_accuracy": 0.7713699210435152,
      "num_tokens": 38305085.0,
      "step": 19380
    },
    {
      "entropy": 0.927297318354249,
      "epoch": 0.5554043796456755,
      "grad_norm": 7.84375,
      "learning_rate": 4.169776926823529e-05,
      "loss": 0.9477129936218261,
      "mean_token_accuracy": 0.7822349708527326,
      "num_tokens": 38323771.0,
      "step": 19390
    },
    {
      "entropy": 0.9047140829265118,
      "epoch": 0.555690818211764,
      "grad_norm": 5.8125,
      "learning_rate": 4.168927447712849e-05,
      "loss": 0.8855545043945312,
      "mean_token_accuracy": 0.7883426390588284,
      "num_tokens": 38342737.0,
      "step": 19400
    },
    {
      "entropy": 0.8957334676757455,
      "epoch": 0.5559772567778526,
      "grad_norm": 8.6875,
      "learning_rate": 4.168077620859987e-05,
      "loss": 0.9611346244812011,
      "mean_token_accuracy": 0.7779878914356232,
      "num_tokens": 38361675.0,
      "step": 19410
    },
    {
      "entropy": 1.0248532945290207,
      "epoch": 0.5562636953439412,
      "grad_norm": 8.875,
      "learning_rate": 4.167227446442015e-05,
      "loss": 1.0683626174926757,
      "mean_token_accuracy": 0.7582519229501485,
      "num_tokens": 38382147.0,
      "step": 19420
    },
    {
      "entropy": 1.011183512583375,
      "epoch": 0.5565501339100296,
      "grad_norm": 8.25,
      "learning_rate": 4.1663769246360774e-05,
      "loss": 1.0042107582092286,
      "mean_token_accuracy": 0.7689656566828489,
      "num_tokens": 38401796.0,
      "step": 19430
    },
    {
      "entropy": 1.0000979075208307,
      "epoch": 0.5568365724761182,
      "grad_norm": 10.3125,
      "learning_rate": 4.165526055619391e-05,
      "loss": 0.9932844161987304,
      "mean_token_accuracy": 0.7723197724670172,
      "num_tokens": 38421438.0,
      "step": 19440
    },
    {
      "entropy": 0.86902816593647,
      "epoch": 0.5571230110422067,
      "grad_norm": 6.90625,
      "learning_rate": 4.164674839569245e-05,
      "loss": 0.8739824295043945,
      "mean_token_accuracy": 0.7951953802257776,
      "num_tokens": 38440368.0,
      "step": 19450
    },
    {
      "entropy": 0.8607277456671,
      "epoch": 0.5574094496082953,
      "grad_norm": 7.03125,
      "learning_rate": 4.1638232766630017e-05,
      "loss": 0.8491899490356445,
      "mean_token_accuracy": 0.7993162356317043,
      "num_tokens": 38460559.0,
      "step": 19460
    },
    {
      "entropy": 0.9326961737126112,
      "epoch": 0.5576958881743838,
      "grad_norm": 7.4375,
      "learning_rate": 4.1629713670780944e-05,
      "loss": 0.9336241722106934,
      "mean_token_accuracy": 0.7811349160969258,
      "num_tokens": 38480354.0,
      "step": 19470
    },
    {
      "entropy": 0.8983510158024728,
      "epoch": 0.5579823267404723,
      "grad_norm": 5.8125,
      "learning_rate": 4.16211911099203e-05,
      "loss": 0.914433765411377,
      "mean_token_accuracy": 0.7867476280778647,
      "num_tokens": 38501991.0,
      "step": 19480
    },
    {
      "entropy": 0.970943932980299,
      "epoch": 0.5582687653065609,
      "grad_norm": 7.09375,
      "learning_rate": 4.161266508582388e-05,
      "loss": 1.0002187728881835,
      "mean_token_accuracy": 0.7682420168071985,
      "num_tokens": 38524067.0,
      "step": 19490
    },
    {
      "entropy": 1.0015486091375352,
      "epoch": 0.5585552038726495,
      "grad_norm": 7.0625,
      "learning_rate": 4.160413560026816e-05,
      "loss": 1.0008804321289062,
      "mean_token_accuracy": 0.7752039108425379,
      "num_tokens": 38543929.0,
      "step": 19500
    },
    {
      "entropy": 1.0352138586342334,
      "epoch": 0.5588416424387379,
      "grad_norm": 7.40625,
      "learning_rate": 4.1595602655030376e-05,
      "loss": 1.0092190742492675,
      "mean_token_accuracy": 0.775368720665574,
      "num_tokens": 38564365.0,
      "step": 19510
    },
    {
      "entropy": 0.9257352869957686,
      "epoch": 0.5591280810048265,
      "grad_norm": 6.21875,
      "learning_rate": 4.158706625188848e-05,
      "loss": 0.9338169097900391,
      "mean_token_accuracy": 0.7742497894912959,
      "num_tokens": 38583043.0,
      "step": 19520
    },
    {
      "entropy": 0.9947560198605061,
      "epoch": 0.559414519570915,
      "grad_norm": 7.6875,
      "learning_rate": 4.157852639262114e-05,
      "loss": 1.0305397033691406,
      "mean_token_accuracy": 0.7714460618793965,
      "num_tokens": 38603326.0,
      "step": 19530
    },
    {
      "entropy": 0.9525862235575915,
      "epoch": 0.5597009581370036,
      "grad_norm": 9.375,
      "learning_rate": 4.1569983079007754e-05,
      "loss": 0.9488375663757325,
      "mean_token_accuracy": 0.7794634312391281,
      "num_tokens": 38623176.0,
      "step": 19540
    },
    {
      "entropy": 0.9559039887040853,
      "epoch": 0.559987396703092,
      "grad_norm": 8.0625,
      "learning_rate": 4.1561436312828414e-05,
      "loss": 0.9647384643554687,
      "mean_token_accuracy": 0.7760484226047992,
      "num_tokens": 38642272.0,
      "step": 19550
    },
    {
      "entropy": 0.9793501025997102,
      "epoch": 0.5602738352691806,
      "grad_norm": 6.65625,
      "learning_rate": 4.1552886095863945e-05,
      "loss": 1.0292466163635254,
      "mean_token_accuracy": 0.765915410220623,
      "num_tokens": 38663230.0,
      "step": 19560
    },
    {
      "entropy": 0.8350932704284787,
      "epoch": 0.5605602738352692,
      "grad_norm": 9.0,
      "learning_rate": 4.154433242989591e-05,
      "loss": 0.8241652488708496,
      "mean_token_accuracy": 0.8014059074223041,
      "num_tokens": 38682487.0,
      "step": 19570
    },
    {
      "entropy": 1.108338305912912,
      "epoch": 0.5608467124013577,
      "grad_norm": 7.4375,
      "learning_rate": 4.153577531670656e-05,
      "loss": 1.127332592010498,
      "mean_token_accuracy": 0.7531335406005383,
      "num_tokens": 38702580.0,
      "step": 19580
    },
    {
      "entropy": 0.9807157827541232,
      "epoch": 0.5611331509674462,
      "grad_norm": 7.15625,
      "learning_rate": 4.152721475807888e-05,
      "loss": 0.9722715377807617,
      "mean_token_accuracy": 0.7702742256224155,
      "num_tokens": 38724099.0,
      "step": 19590
    },
    {
      "entropy": 0.9845884306356311,
      "epoch": 0.5614195895335348,
      "grad_norm": 9.5,
      "learning_rate": 4.1518650755796565e-05,
      "loss": 0.9552626609802246,
      "mean_token_accuracy": 0.7801853995770216,
      "num_tokens": 38743454.0,
      "step": 19600
    },
    {
      "entropy": 1.0776905428618193,
      "epoch": 0.5617060280996233,
      "grad_norm": 6.5,
      "learning_rate": 4.1510083311644056e-05,
      "loss": 1.1053731918334961,
      "mean_token_accuracy": 0.7527314476668835,
      "num_tokens": 38762981.0,
      "step": 19610
    },
    {
      "entropy": 0.9004903633147479,
      "epoch": 0.5619924666657119,
      "grad_norm": 6.34375,
      "learning_rate": 4.1501512427406464e-05,
      "loss": 0.9865876197814941,
      "mean_token_accuracy": 0.7680407512933016,
      "num_tokens": 38782250.0,
      "step": 19620
    },
    {
      "entropy": 0.9333377033472061,
      "epoch": 0.5622789052318005,
      "grad_norm": 7.875,
      "learning_rate": 4.149293810486965e-05,
      "loss": 0.8889418601989746,
      "mean_token_accuracy": 0.7849811259657145,
      "num_tokens": 38801274.0,
      "step": 19630
    },
    {
      "entropy": 0.9339837461709977,
      "epoch": 0.5625653437978889,
      "grad_norm": 6.59375,
      "learning_rate": 4.14843603458202e-05,
      "loss": 0.929569149017334,
      "mean_token_accuracy": 0.7795525718480348,
      "num_tokens": 38820569.0,
      "step": 19640
    },
    {
      "entropy": 0.9881141383200884,
      "epoch": 0.5628517823639775,
      "grad_norm": 8.9375,
      "learning_rate": 4.147577915204536e-05,
      "loss": 1.00313081741333,
      "mean_token_accuracy": 0.7617206372320652,
      "num_tokens": 38838828.0,
      "step": 19650
    },
    {
      "entropy": 0.8715533567592502,
      "epoch": 0.563138220930066,
      "grad_norm": 7.15625,
      "learning_rate": 4.1467194525333166e-05,
      "loss": 0.9082240104675293,
      "mean_token_accuracy": 0.7885855607688427,
      "num_tokens": 38858809.0,
      "step": 19660
    },
    {
      "entropy": 1.0498733075335622,
      "epoch": 0.5634246594961546,
      "grad_norm": 6.53125,
      "learning_rate": 4.145860646747232e-05,
      "loss": 1.0786943435668945,
      "mean_token_accuracy": 0.7626278057694436,
      "num_tokens": 38879117.0,
      "step": 19670
    },
    {
      "entropy": 1.0971468076109887,
      "epoch": 0.5637110980622431,
      "grad_norm": 8.875,
      "learning_rate": 4.145001498025224e-05,
      "loss": 1.175865364074707,
      "mean_token_accuracy": 0.7378982320427895,
      "num_tokens": 38897833.0,
      "step": 19680
    },
    {
      "entropy": 0.9721847435459494,
      "epoch": 0.5639975366283316,
      "grad_norm": 7.21875,
      "learning_rate": 4.14414200654631e-05,
      "loss": 0.9975674629211426,
      "mean_token_accuracy": 0.7665717273950576,
      "num_tokens": 38917426.0,
      "step": 19690
    },
    {
      "entropy": 0.9583038538694382,
      "epoch": 0.5642839751944202,
      "grad_norm": 7.84375,
      "learning_rate": 4.1432821724895734e-05,
      "loss": 0.9800829887390137,
      "mean_token_accuracy": 0.7736961282789707,
      "num_tokens": 38937072.0,
      "step": 19700
    },
    {
      "entropy": 0.950890189409256,
      "epoch": 0.5645704137605088,
      "grad_norm": 7.8125,
      "learning_rate": 4.142421996034173e-05,
      "loss": 0.9950137138366699,
      "mean_token_accuracy": 0.7690510574728251,
      "num_tokens": 38955688.0,
      "step": 19710
    },
    {
      "entropy": 1.0428343627601862,
      "epoch": 0.5648568523265972,
      "grad_norm": 6.34375,
      "learning_rate": 4.141561477359337e-05,
      "loss": 1.0521081924438476,
      "mean_token_accuracy": 0.7594671942293644,
      "num_tokens": 38975721.0,
      "step": 19720
    },
    {
      "entropy": 1.0106626216322183,
      "epoch": 0.5651432908926858,
      "grad_norm": 7.8125,
      "learning_rate": 4.1407006166443656e-05,
      "loss": 0.9970620155334473,
      "mean_token_accuracy": 0.7731382846832275,
      "num_tokens": 38994653.0,
      "step": 19730
    },
    {
      "entropy": 1.0779481068253518,
      "epoch": 0.5654297294587743,
      "grad_norm": 9.25,
      "learning_rate": 4.1398394140686294e-05,
      "loss": 1.0601713180541992,
      "mean_token_accuracy": 0.753212433680892,
      "num_tokens": 39014631.0,
      "step": 19740
    },
    {
      "entropy": 0.9348740123212338,
      "epoch": 0.5657161680248629,
      "grad_norm": 11.1875,
      "learning_rate": 4.138977869811571e-05,
      "loss": 0.9546441078186035,
      "mean_token_accuracy": 0.7863631464540959,
      "num_tokens": 39032783.0,
      "step": 19750
    },
    {
      "entropy": 0.9436424765735865,
      "epoch": 0.5660026065909514,
      "grad_norm": 6.0,
      "learning_rate": 4.1381159840527054e-05,
      "loss": 0.9623109817504882,
      "mean_token_accuracy": 0.7768029198050499,
      "num_tokens": 39054086.0,
      "step": 19760
    },
    {
      "entropy": 0.9756453232839704,
      "epoch": 0.5662890451570399,
      "grad_norm": 8.875,
      "learning_rate": 4.1372537569716155e-05,
      "loss": 0.9895077705383301,
      "mean_token_accuracy": 0.7684864412993193,
      "num_tokens": 39073128.0,
      "step": 19770
    },
    {
      "entropy": 0.8633647851645947,
      "epoch": 0.5665754837231285,
      "grad_norm": 6.3125,
      "learning_rate": 4.136391188747958e-05,
      "loss": 0.864451789855957,
      "mean_token_accuracy": 0.806272380799055,
      "num_tokens": 39093150.0,
      "step": 19780
    },
    {
      "entropy": 0.9746080070734024,
      "epoch": 0.5668619222892171,
      "grad_norm": 7.78125,
      "learning_rate": 4.13552827956146e-05,
      "loss": 1.0095254898071289,
      "mean_token_accuracy": 0.7618807692080736,
      "num_tokens": 39113091.0,
      "step": 19790
    },
    {
      "entropy": 1.0073129093274473,
      "epoch": 0.5671483608553055,
      "grad_norm": 10.625,
      "learning_rate": 4.1346650295919205e-05,
      "loss": 1.0164382934570313,
      "mean_token_accuracy": 0.7714897431433201,
      "num_tokens": 39133220.0,
      "step": 19800
    },
    {
      "entropy": 0.8674702906981111,
      "epoch": 0.5674347994213941,
      "grad_norm": 6.40625,
      "learning_rate": 4.133801439019207e-05,
      "loss": 0.8571381568908691,
      "mean_token_accuracy": 0.7922187574207783,
      "num_tokens": 39152441.0,
      "step": 19810
    },
    {
      "entropy": 1.0287092326208949,
      "epoch": 0.5677212379874826,
      "grad_norm": 7.375,
      "learning_rate": 4.1329375080232597e-05,
      "loss": 1.074624729156494,
      "mean_token_accuracy": 0.7560446955263614,
      "num_tokens": 39171863.0,
      "step": 19820
    },
    {
      "entropy": 0.9506982645019889,
      "epoch": 0.5680076765535712,
      "grad_norm": 7.15625,
      "learning_rate": 4.132073236784091e-05,
      "loss": 0.9187474250793457,
      "mean_token_accuracy": 0.7944683220237494,
      "num_tokens": 39190964.0,
      "step": 19830
    },
    {
      "entropy": 1.024363349750638,
      "epoch": 0.5682941151196597,
      "grad_norm": 10.25,
      "learning_rate": 4.13120862548178e-05,
      "loss": 1.0443736076354981,
      "mean_token_accuracy": 0.7614618310704827,
      "num_tokens": 39209836.0,
      "step": 19840
    },
    {
      "entropy": 1.0028297476470471,
      "epoch": 0.5685805536857482,
      "grad_norm": 7.875,
      "learning_rate": 4.130343674296483e-05,
      "loss": 1.0341675758361817,
      "mean_token_accuracy": 0.7620521996170282,
      "num_tokens": 39229932.0,
      "step": 19850
    },
    {
      "entropy": 0.9206191564910113,
      "epoch": 0.5688669922518368,
      "grad_norm": 7.84375,
      "learning_rate": 4.1294783834084216e-05,
      "loss": 0.9203093528747559,
      "mean_token_accuracy": 0.7892377279698849,
      "num_tokens": 39248393.0,
      "step": 19860
    },
    {
      "entropy": 0.9425908589735628,
      "epoch": 0.5691534308179254,
      "grad_norm": 5.5,
      "learning_rate": 4.1286127529978905e-05,
      "loss": 0.9681834220886231,
      "mean_token_accuracy": 0.7748812358826399,
      "num_tokens": 39267957.0,
      "step": 19870
    },
    {
      "entropy": 0.8139951415359974,
      "epoch": 0.5694398693840138,
      "grad_norm": 6.03125,
      "learning_rate": 4.127746783245253e-05,
      "loss": 0.7672500133514404,
      "mean_token_accuracy": 0.8134459607303143,
      "num_tokens": 39287194.0,
      "step": 19880
    },
    {
      "entropy": 1.0402536880224944,
      "epoch": 0.5697263079501024,
      "grad_norm": 9.625,
      "learning_rate": 4.1268804743309485e-05,
      "loss": 1.0744309425354004,
      "mean_token_accuracy": 0.7562359970062971,
      "num_tokens": 39306200.0,
      "step": 19890
    },
    {
      "entropy": 0.8958443019539117,
      "epoch": 0.5700127465161909,
      "grad_norm": 8.375,
      "learning_rate": 4.12601382643548e-05,
      "loss": 0.9081062316894531,
      "mean_token_accuracy": 0.7849580965936184,
      "num_tokens": 39326327.0,
      "step": 19900
    },
    {
      "entropy": 0.969285492040217,
      "epoch": 0.5702991850822795,
      "grad_norm": 10.0625,
      "learning_rate": 4.1251468397394264e-05,
      "loss": 0.9410308837890625,
      "mean_token_accuracy": 0.7844414513558149,
      "num_tokens": 39345585.0,
      "step": 19910
    },
    {
      "entropy": 0.950868820026517,
      "epoch": 0.5705856236483681,
      "grad_norm": 8.4375,
      "learning_rate": 4.1242795144234346e-05,
      "loss": 1.0266042709350587,
      "mean_token_accuracy": 0.7605800732970238,
      "num_tokens": 39364869.0,
      "step": 19920
    },
    {
      "entropy": 0.9508584728464484,
      "epoch": 0.5708720622144565,
      "grad_norm": 8.25,
      "learning_rate": 4.123411850668224e-05,
      "loss": 1.0088253021240234,
      "mean_token_accuracy": 0.7656967118382454,
      "num_tokens": 39385728.0,
      "step": 19930
    },
    {
      "entropy": 0.990834153816104,
      "epoch": 0.5711585007805451,
      "grad_norm": 8.9375,
      "learning_rate": 4.122543848654581e-05,
      "loss": 1.0103014945983886,
      "mean_token_accuracy": 0.7677418265491724,
      "num_tokens": 39405902.0,
      "step": 19940
    },
    {
      "entropy": 0.9897529222071171,
      "epoch": 0.5714449393466337,
      "grad_norm": 10.625,
      "learning_rate": 4.1216755085633674e-05,
      "loss": 0.9950751304626465,
      "mean_token_accuracy": 0.769730270653963,
      "num_tokens": 39426863.0,
      "step": 19950
    },
    {
      "entropy": 0.8844190292060375,
      "epoch": 0.5717313779127222,
      "grad_norm": 8.1875,
      "learning_rate": 4.120806830575512e-05,
      "loss": 0.86932373046875,
      "mean_token_accuracy": 0.7860382415354252,
      "num_tokens": 39445920.0,
      "step": 19960
    },
    {
      "entropy": 0.9261470797471703,
      "epoch": 0.5720178164788107,
      "grad_norm": 7.3125,
      "learning_rate": 4.119937814872014e-05,
      "loss": 0.8737369537353515,
      "mean_token_accuracy": 0.7987537503242492,
      "num_tokens": 39465961.0,
      "step": 19970
    },
    {
      "entropy": 0.9400382541120053,
      "epoch": 0.5723042550448992,
      "grad_norm": 11.1875,
      "learning_rate": 4.1190684616339446e-05,
      "loss": 1.0355422973632813,
      "mean_token_accuracy": 0.7586205005645752,
      "num_tokens": 39485972.0,
      "step": 19980
    },
    {
      "entropy": 0.90003581661731,
      "epoch": 0.5725906936109878,
      "grad_norm": 6.75,
      "learning_rate": 4.118198771042444e-05,
      "loss": 0.872646713256836,
      "mean_token_accuracy": 0.7867083795368671,
      "num_tokens": 39505413.0,
      "step": 19990
    },
    {
      "epoch": 0.5728771321770764,
      "eval_entropy": 0.9877232120037079,
      "eval_loss": 1.013584852218628,
      "eval_mean_token_accuracy": 0.7637251317501068,
      "eval_num_tokens": 39526386.0,
      "eval_runtime": 43.2607,
      "eval_samples_per_second": 46.231,
      "eval_steps_per_second": 5.779,
      "step": 20000
    },
    {
      "entropy": 0.9743144256062806,
      "epoch": 0.5731635707431648,
      "grad_norm": 5.96875,
      "learning_rate": 4.1164583785240656e-05,
      "loss": 0.9881059646606445,
      "mean_token_accuracy": 0.7716233748942614,
      "num_tokens": 39546017.0,
      "step": 20010
    },
    {
      "entropy": 0.9395108044147491,
      "epoch": 0.5734500093092534,
      "grad_norm": 7.28125,
      "learning_rate": 4.11558767695982e-05,
      "loss": 0.9427123069763184,
      "mean_token_accuracy": 0.7766081787645817,
      "num_tokens": 39564709.0,
      "step": 20020
    },
    {
      "entropy": 0.973422285169363,
      "epoch": 0.573736447875342,
      "grad_norm": 7.125,
      "learning_rate": 4.1147166387674097e-05,
      "loss": 1.0254555702209474,
      "mean_token_accuracy": 0.7729189310222864,
      "num_tokens": 39583991.0,
      "step": 20030
    },
    {
      "entropy": 1.015830126963556,
      "epoch": 0.5740228864414305,
      "grad_norm": 6.71875,
      "learning_rate": 4.113845264128326e-05,
      "loss": 1.0577707290649414,
      "mean_token_accuracy": 0.7579512394964695,
      "num_tokens": 39606253.0,
      "step": 20040
    },
    {
      "entropy": 0.9367475468665362,
      "epoch": 0.574309325007519,
      "grad_norm": 6.3125,
      "learning_rate": 4.112973553224131e-05,
      "loss": 0.9405386924743653,
      "mean_token_accuracy": 0.7895747713744641,
      "num_tokens": 39626154.0,
      "step": 20050
    },
    {
      "entropy": 0.9724262777715922,
      "epoch": 0.5745957635736075,
      "grad_norm": 6.59375,
      "learning_rate": 4.112101506236455e-05,
      "loss": 0.9988237380981445,
      "mean_token_accuracy": 0.7739479385316372,
      "num_tokens": 39645381.0,
      "step": 20060
    },
    {
      "entropy": 1.0361915966495872,
      "epoch": 0.5748822021396961,
      "grad_norm": 8.4375,
      "learning_rate": 4.111229123347004e-05,
      "loss": 1.0676725387573243,
      "mean_token_accuracy": 0.7619747675955295,
      "num_tokens": 39665550.0,
      "step": 20070
    },
    {
      "entropy": 1.020539204031229,
      "epoch": 0.5751686407057847,
      "grad_norm": 8.875,
      "learning_rate": 4.1103564047375466e-05,
      "loss": 1.0539648056030273,
      "mean_token_accuracy": 0.7598235569894314,
      "num_tokens": 39685596.0,
      "step": 20080
    },
    {
      "entropy": 1.0001544227823616,
      "epoch": 0.5754550792718731,
      "grad_norm": 8.125,
      "learning_rate": 4.109483350589926e-05,
      "loss": 1.0027913093566894,
      "mean_token_accuracy": 0.7725025493651628,
      "num_tokens": 39704630.0,
      "step": 20090
    },
    {
      "entropy": 0.9718913894146681,
      "epoch": 0.5757415178379617,
      "grad_norm": 10.125,
      "learning_rate": 4.108609961086054e-05,
      "loss": 0.9541418075561523,
      "mean_token_accuracy": 0.7838160447776318,
      "num_tokens": 39724732.0,
      "step": 20100
    },
    {
      "entropy": 0.9927222821861506,
      "epoch": 0.5760279564040502,
      "grad_norm": 7.25,
      "learning_rate": 4.107736236407912e-05,
      "loss": 0.9706440925598144,
      "mean_token_accuracy": 0.7728913251310587,
      "num_tokens": 39745100.0,
      "step": 20110
    },
    {
      "entropy": 0.9994804799556732,
      "epoch": 0.5763143949701388,
      "grad_norm": 11.75,
      "learning_rate": 4.106862176737553e-05,
      "loss": 0.948979377746582,
      "mean_token_accuracy": 0.7827293567359448,
      "num_tokens": 39765468.0,
      "step": 20120
    },
    {
      "entropy": 0.9589021114632488,
      "epoch": 0.5766008335362273,
      "grad_norm": 7.15625,
      "learning_rate": 4.105987782257096e-05,
      "loss": 0.964389705657959,
      "mean_token_accuracy": 0.7803164631128311,
      "num_tokens": 39787139.0,
      "step": 20130
    },
    {
      "entropy": 0.8934176401235163,
      "epoch": 0.5768872721023158,
      "grad_norm": 7.21875,
      "learning_rate": 4.105113053148734e-05,
      "loss": 0.928364372253418,
      "mean_token_accuracy": 0.7907134246081113,
      "num_tokens": 39806794.0,
      "step": 20140
    },
    {
      "entropy": 0.9500906802713871,
      "epoch": 0.5771737106684044,
      "grad_norm": 10.4375,
      "learning_rate": 4.104237989594728e-05,
      "loss": 0.9524734497070313,
      "mean_token_accuracy": 0.7670584224164486,
      "num_tokens": 39825956.0,
      "step": 20150
    },
    {
      "entropy": 0.8618243556469679,
      "epoch": 0.577460149234493,
      "grad_norm": 5.90625,
      "learning_rate": 4.1033625917774064e-05,
      "loss": 0.8813813209533692,
      "mean_token_accuracy": 0.7918251369148492,
      "num_tokens": 39845895.0,
      "step": 20160
    },
    {
      "entropy": 1.0061787150800228,
      "epoch": 0.5777465878005815,
      "grad_norm": 8.9375,
      "learning_rate": 4.102486859879173e-05,
      "loss": 0.9966843605041504,
      "mean_token_accuracy": 0.7624151077121496,
      "num_tokens": 39865281.0,
      "step": 20170
    },
    {
      "entropy": 0.8883185619488358,
      "epoch": 0.57803302636667,
      "grad_norm": 9.4375,
      "learning_rate": 4.101610794082494e-05,
      "loss": 0.8031073570251465,
      "mean_token_accuracy": 0.8050013396888971,
      "num_tokens": 39884844.0,
      "step": 20180
    },
    {
      "entropy": 0.9233494058251381,
      "epoch": 0.5783194649327585,
      "grad_norm": 8.25,
      "learning_rate": 4.100734394569912e-05,
      "loss": 0.9848557472229004,
      "mean_token_accuracy": 0.7704064838588238,
      "num_tokens": 39904345.0,
      "step": 20190
    },
    {
      "entropy": 0.9666221361607313,
      "epoch": 0.5786059034988471,
      "grad_norm": 8.4375,
      "learning_rate": 4.099857661524033e-05,
      "loss": 0.9810226440429688,
      "mean_token_accuracy": 0.7690009452402592,
      "num_tokens": 39924108.0,
      "step": 20200
    },
    {
      "entropy": 0.9859731748700142,
      "epoch": 0.5788923420649357,
      "grad_norm": 8.6875,
      "learning_rate": 4.098980595127537e-05,
      "loss": 1.0264087677001954,
      "mean_token_accuracy": 0.7659872639924288,
      "num_tokens": 39945026.0,
      "step": 20210
    },
    {
      "entropy": 1.0115782588720321,
      "epoch": 0.5791787806310241,
      "grad_norm": 7.875,
      "learning_rate": 4.098103195563173e-05,
      "loss": 0.9613194465637207,
      "mean_token_accuracy": 0.7699773252010346,
      "num_tokens": 39964536.0,
      "step": 20220
    },
    {
      "entropy": 0.8349237475544214,
      "epoch": 0.5794652191971127,
      "grad_norm": 6.875,
      "learning_rate": 4.0972254630137566e-05,
      "loss": 0.8728310585021972,
      "mean_token_accuracy": 0.7912918381392956,
      "num_tokens": 39983862.0,
      "step": 20230
    },
    {
      "entropy": 1.0740262478590012,
      "epoch": 0.5797516577632013,
      "grad_norm": 9.6875,
      "learning_rate": 4.096347397662174e-05,
      "loss": 1.0913339614868165,
      "mean_token_accuracy": 0.7502806585282087,
      "num_tokens": 40003165.0,
      "step": 20240
    },
    {
      "entropy": 0.8802586395293475,
      "epoch": 0.5800380963292898,
      "grad_norm": 11.0625,
      "learning_rate": 4.0954689996913834e-05,
      "loss": 0.8988991737365722,
      "mean_token_accuracy": 0.7896165430545807,
      "num_tokens": 40023456.0,
      "step": 20250
    },
    {
      "entropy": 0.8832138726487756,
      "epoch": 0.5803245348953783,
      "grad_norm": 8.3125,
      "learning_rate": 4.0945902692844087e-05,
      "loss": 0.8604216575622559,
      "mean_token_accuracy": 0.788755489885807,
      "num_tokens": 40043764.0,
      "step": 20260
    },
    {
      "entropy": 0.9689393466338515,
      "epoch": 0.5806109734614668,
      "grad_norm": 6.0,
      "learning_rate": 4.0937112066243436e-05,
      "loss": 0.9914020538330078,
      "mean_token_accuracy": 0.7677198424935341,
      "num_tokens": 40062768.0,
      "step": 20270
    },
    {
      "entropy": 0.8788830837234854,
      "epoch": 0.5808974120275554,
      "grad_norm": 9.9375,
      "learning_rate": 4.092831811894354e-05,
      "loss": 0.9264053344726563,
      "mean_token_accuracy": 0.7914253011345863,
      "num_tokens": 40081780.0,
      "step": 20280
    },
    {
      "entropy": 1.0266006471589209,
      "epoch": 0.581183850593644,
      "grad_norm": 11.75,
      "learning_rate": 4.091952085277673e-05,
      "loss": 1.0371515274047851,
      "mean_token_accuracy": 0.7644747916609049,
      "num_tokens": 40102253.0,
      "step": 20290
    },
    {
      "entropy": 0.8955442847684025,
      "epoch": 0.5814702891597324,
      "grad_norm": 7.65625,
      "learning_rate": 4.0910720269576004e-05,
      "loss": 0.9469646453857422,
      "mean_token_accuracy": 0.7812335304915905,
      "num_tokens": 40122023.0,
      "step": 20300
    },
    {
      "entropy": 0.9341918094083667,
      "epoch": 0.581756727725821,
      "grad_norm": 7.15625,
      "learning_rate": 4.0901916371175095e-05,
      "loss": 0.9658877372741699,
      "mean_token_accuracy": 0.7769235357642174,
      "num_tokens": 40142270.0,
      "step": 20310
    },
    {
      "entropy": 1.0104470368474723,
      "epoch": 0.5820431662919096,
      "grad_norm": 7.28125,
      "learning_rate": 4.0893109159408403e-05,
      "loss": 0.9800692558288574,
      "mean_token_accuracy": 0.7685091432183981,
      "num_tokens": 40162160.0,
      "step": 20320
    },
    {
      "entropy": 0.9051725916564465,
      "epoch": 0.5823296048579981,
      "grad_norm": 9.0,
      "learning_rate": 4.0884298636111015e-05,
      "loss": 0.8798377990722657,
      "mean_token_accuracy": 0.7956476740539074,
      "num_tokens": 40180741.0,
      "step": 20330
    },
    {
      "entropy": 0.9469224305823445,
      "epoch": 0.5826160434240866,
      "grad_norm": 9.5625,
      "learning_rate": 4.087548480311872e-05,
      "loss": 0.9971781730651855,
      "mean_token_accuracy": 0.7667026709765196,
      "num_tokens": 40200449.0,
      "step": 20340
    },
    {
      "entropy": 1.0360435027629138,
      "epoch": 0.5829024819901751,
      "grad_norm": 7.65625,
      "learning_rate": 4.0866667662268e-05,
      "loss": 1.050649070739746,
      "mean_token_accuracy": 0.7645127668976783,
      "num_tokens": 40219897.0,
      "step": 20350
    },
    {
      "entropy": 1.022946797311306,
      "epoch": 0.5831889205562637,
      "grad_norm": 6.25,
      "learning_rate": 4.085784721539601e-05,
      "loss": 1.019237232208252,
      "mean_token_accuracy": 0.7648782677948475,
      "num_tokens": 40239915.0,
      "step": 20360
    },
    {
      "entropy": 0.8659194767475128,
      "epoch": 0.5834753591223523,
      "grad_norm": 6.3125,
      "learning_rate": 4.08490234643406e-05,
      "loss": 0.8371137619018555,
      "mean_token_accuracy": 0.7984774462878704,
      "num_tokens": 40260423.0,
      "step": 20370
    },
    {
      "entropy": 0.8768610382452607,
      "epoch": 0.5837617976884407,
      "grad_norm": 7.0625,
      "learning_rate": 4.084019641094031e-05,
      "loss": 0.9091633796691895,
      "mean_token_accuracy": 0.7855933662503958,
      "num_tokens": 40279822.0,
      "step": 20380
    },
    {
      "entropy": 0.8666309759020805,
      "epoch": 0.5840482362545293,
      "grad_norm": 8.8125,
      "learning_rate": 4.083136605703437e-05,
      "loss": 0.8934616088867188,
      "mean_token_accuracy": 0.7890436481684446,
      "num_tokens": 40299006.0,
      "step": 20390
    },
    {
      "entropy": 0.912430851906538,
      "epoch": 0.5843346748206178,
      "grad_norm": 9.8125,
      "learning_rate": 4.08225324044627e-05,
      "loss": 0.9640564918518066,
      "mean_token_accuracy": 0.7753376845270396,
      "num_tokens": 40317518.0,
      "step": 20400
    },
    {
      "entropy": 0.8935806721448898,
      "epoch": 0.5846211133867064,
      "grad_norm": 10.3125,
      "learning_rate": 4.08136954550659e-05,
      "loss": 0.792830514907837,
      "mean_token_accuracy": 0.8115452572703361,
      "num_tokens": 40336536.0,
      "step": 20410
    },
    {
      "entropy": 0.936208451166749,
      "epoch": 0.584907551952795,
      "grad_norm": 7.8125,
      "learning_rate": 4.0804855210685266e-05,
      "loss": 0.9398514747619628,
      "mean_token_accuracy": 0.774126711115241,
      "num_tokens": 40354968.0,
      "step": 20420
    },
    {
      "entropy": 1.004452271014452,
      "epoch": 0.5851939905188834,
      "grad_norm": 6.875,
      "learning_rate": 4.079601167316276e-05,
      "loss": 1.0822473526000977,
      "mean_token_accuracy": 0.7707275457680225,
      "num_tokens": 40376623.0,
      "step": 20430
    },
    {
      "entropy": 0.8957709370180964,
      "epoch": 0.585480429084972,
      "grad_norm": 8.75,
      "learning_rate": 4.0787164844341066e-05,
      "loss": 0.9186844825744629,
      "mean_token_accuracy": 0.7829184107482433,
      "num_tokens": 40398183.0,
      "step": 20440
    },
    {
      "entropy": 0.906250961497426,
      "epoch": 0.5857668676510606,
      "grad_norm": 9.875,
      "learning_rate": 4.077831472606352e-05,
      "loss": 0.9339326858520508,
      "mean_token_accuracy": 0.7821908906102181,
      "num_tokens": 40417569.0,
      "step": 20450
    },
    {
      "entropy": 0.9379493309184909,
      "epoch": 0.5860533062171491,
      "grad_norm": 8.25,
      "learning_rate": 4.0769461320174154e-05,
      "loss": 0.9542790412902832,
      "mean_token_accuracy": 0.7770782265812158,
      "num_tokens": 40438900.0,
      "step": 20460
    },
    {
      "entropy": 1.097783868573606,
      "epoch": 0.5863397447832376,
      "grad_norm": 11.9375,
      "learning_rate": 4.076060462851769e-05,
      "loss": 1.1625299453735352,
      "mean_token_accuracy": 0.7410434048622847,
      "num_tokens": 40459133.0,
      "step": 20470
    },
    {
      "entropy": 0.9125070013105869,
      "epoch": 0.5866261833493261,
      "grad_norm": 7.46875,
      "learning_rate": 4.075174465293954e-05,
      "loss": 0.8762696266174317,
      "mean_token_accuracy": 0.795248543471098,
      "num_tokens": 40478328.0,
      "step": 20480
    },
    {
      "entropy": 0.936875872220844,
      "epoch": 0.5869126219154147,
      "grad_norm": 5.5,
      "learning_rate": 4.074288139528578e-05,
      "loss": 0.881074333190918,
      "mean_token_accuracy": 0.791917197778821,
      "num_tokens": 40497702.0,
      "step": 20490
    },
    {
      "entropy": 0.8741323390975595,
      "epoch": 0.5871990604815033,
      "grad_norm": 8.9375,
      "learning_rate": 4.073401485740319e-05,
      "loss": 0.9533127784729004,
      "mean_token_accuracy": 0.7899199858307838,
      "num_tokens": 40516319.0,
      "step": 20500
    },
    {
      "entropy": 1.0455211844295262,
      "epoch": 0.5874854990475917,
      "grad_norm": 7.75,
      "learning_rate": 4.072514504113922e-05,
      "loss": 1.048889446258545,
      "mean_token_accuracy": 0.7571695540100336,
      "num_tokens": 40536736.0,
      "step": 20510
    },
    {
      "entropy": 0.9466321220621466,
      "epoch": 0.5877719376136803,
      "grad_norm": 7.46875,
      "learning_rate": 4.071627194834202e-05,
      "loss": 0.9859343528747558,
      "mean_token_accuracy": 0.7682564459741116,
      "num_tokens": 40555745.0,
      "step": 20520
    },
    {
      "entropy": 0.8678402498364448,
      "epoch": 0.5880583761797689,
      "grad_norm": 9.75,
      "learning_rate": 4.0707395580860395e-05,
      "loss": 0.825813102722168,
      "mean_token_accuracy": 0.7993369691073895,
      "num_tokens": 40575010.0,
      "step": 20530
    },
    {
      "entropy": 0.9467619424685836,
      "epoch": 0.5883448147458574,
      "grad_norm": 7.4375,
      "learning_rate": 4.069851594054387e-05,
      "loss": 0.9585908889770508,
      "mean_token_accuracy": 0.7831815924495459,
      "num_tokens": 40595614.0,
      "step": 20540
    },
    {
      "entropy": 0.9116890400648117,
      "epoch": 0.5886312533119459,
      "grad_norm": 7.3125,
      "learning_rate": 4.06896330292426e-05,
      "loss": 0.9282383918762207,
      "mean_token_accuracy": 0.7864695079624653,
      "num_tokens": 40615371.0,
      "step": 20550
    },
    {
      "entropy": 0.893018077686429,
      "epoch": 0.5889176918780344,
      "grad_norm": 10.125,
      "learning_rate": 4.0680746848807486e-05,
      "loss": 0.854468822479248,
      "mean_token_accuracy": 0.7927126739174127,
      "num_tokens": 40634992.0,
      "step": 20560
    },
    {
      "entropy": 0.87540267072618,
      "epoch": 0.589204130444123,
      "grad_norm": 7.0625,
      "learning_rate": 4.067185740109005e-05,
      "loss": 0.8731182098388672,
      "mean_token_accuracy": 0.7959563385695219,
      "num_tokens": 40654685.0,
      "step": 20570
    },
    {
      "entropy": 1.0295328371226788,
      "epoch": 0.5894905690102116,
      "grad_norm": 9.8125,
      "learning_rate": 4.066296468794253e-05,
      "loss": 1.0021032333374023,
      "mean_token_accuracy": 0.7662377797067166,
      "num_tokens": 40674756.0,
      "step": 20580
    },
    {
      "entropy": 1.0410988885909318,
      "epoch": 0.5897770075763,
      "grad_norm": 9.5,
      "learning_rate": 4.065406871121784e-05,
      "loss": 1.0926506042480468,
      "mean_token_accuracy": 0.749431798607111,
      "num_tokens": 40695519.0,
      "step": 20590
    },
    {
      "entropy": 0.8800399286672473,
      "epoch": 0.5900634461423886,
      "grad_norm": 6.40625,
      "learning_rate": 4.064516947276956e-05,
      "loss": 0.9195139884948731,
      "mean_token_accuracy": 0.7888823498040438,
      "num_tokens": 40715652.0,
      "step": 20600
    },
    {
      "entropy": 0.9306248532608151,
      "epoch": 0.5903498847084772,
      "grad_norm": 10.0,
      "learning_rate": 4.0636266974451974e-05,
      "loss": 0.9525252342224121,
      "mean_token_accuracy": 0.7809186842292547,
      "num_tokens": 40733968.0,
      "step": 20610
    },
    {
      "entropy": 1.033230389840901,
      "epoch": 0.5906363232745657,
      "grad_norm": 9.125,
      "learning_rate": 4.062736121812002e-05,
      "loss": 1.0795620918273925,
      "mean_token_accuracy": 0.7626445930451154,
      "num_tokens": 40752228.0,
      "step": 20620
    },
    {
      "entropy": 1.0076969455927611,
      "epoch": 0.5909227618406542,
      "grad_norm": 6.65625,
      "learning_rate": 4.061845220562932e-05,
      "loss": 0.9812061309814453,
      "mean_token_accuracy": 0.7731360297650098,
      "num_tokens": 40770999.0,
      "step": 20630
    },
    {
      "entropy": 0.9555409287102521,
      "epoch": 0.5912092004067427,
      "grad_norm": 7.625,
      "learning_rate": 4.060953993883619e-05,
      "loss": 1.0335517883300782,
      "mean_token_accuracy": 0.7745512427762151,
      "num_tokens": 40791027.0,
      "step": 20640
    },
    {
      "entropy": 0.9497677065432072,
      "epoch": 0.5914956389728313,
      "grad_norm": 6.8125,
      "learning_rate": 4.060062441959761e-05,
      "loss": 0.911768627166748,
      "mean_token_accuracy": 0.783646060153842,
      "num_tokens": 40810600.0,
      "step": 20650
    },
    {
      "entropy": 1.048075314424932,
      "epoch": 0.5917820775389199,
      "grad_norm": 7.25,
      "learning_rate": 4.0591705649771246e-05,
      "loss": 1.029709529876709,
      "mean_token_accuracy": 0.7670080941170454,
      "num_tokens": 40831247.0,
      "step": 20660
    },
    {
      "entropy": 0.8239648763090373,
      "epoch": 0.5920685161050083,
      "grad_norm": 7.6875,
      "learning_rate": 4.0582783631215424e-05,
      "loss": 0.827114200592041,
      "mean_token_accuracy": 0.8017411805689335,
      "num_tokens": 40851361.0,
      "step": 20670
    },
    {
      "entropy": 0.9628026841208339,
      "epoch": 0.5923549546710969,
      "grad_norm": 9.75,
      "learning_rate": 4.057385836578917e-05,
      "loss": 0.9166473388671875,
      "mean_token_accuracy": 0.7875780992209911,
      "num_tokens": 40873110.0,
      "step": 20680
    },
    {
      "entropy": 0.8636716475710273,
      "epoch": 0.5926413932371855,
      "grad_norm": 6.1875,
      "learning_rate": 4.0564929855352174e-05,
      "loss": 0.9192937850952149,
      "mean_token_accuracy": 0.7914801485836506,
      "num_tokens": 40891485.0,
      "step": 20690
    },
    {
      "entropy": 0.8500786259770393,
      "epoch": 0.592927831803274,
      "grad_norm": 7.25,
      "learning_rate": 4.0555998101764805e-05,
      "loss": 0.8672176361083984,
      "mean_token_accuracy": 0.803352776169777,
      "num_tokens": 40912134.0,
      "step": 20700
    },
    {
      "entropy": 1.0164161024615168,
      "epoch": 0.5932142703693626,
      "grad_norm": 7.75,
      "learning_rate": 4.0547063106888106e-05,
      "loss": 1.0464364051818849,
      "mean_token_accuracy": 0.7635302990674973,
      "num_tokens": 40932635.0,
      "step": 20710
    },
    {
      "entropy": 0.9309699583798647,
      "epoch": 0.593500708935451,
      "grad_norm": 7.6875,
      "learning_rate": 4.0538124872583794e-05,
      "loss": 0.9435456275939942,
      "mean_token_accuracy": 0.7785880167037249,
      "num_tokens": 40952026.0,
      "step": 20720
    },
    {
      "entropy": 0.9318991012871265,
      "epoch": 0.5937871475015396,
      "grad_norm": 8.4375,
      "learning_rate": 4.052918340071428e-05,
      "loss": 0.9538392066955567,
      "mean_token_accuracy": 0.7795367680490017,
      "num_tokens": 40970415.0,
      "step": 20730
    },
    {
      "entropy": 0.9007409632205963,
      "epoch": 0.5940735860676282,
      "grad_norm": 7.40625,
      "learning_rate": 4.05202386931426e-05,
      "loss": 0.8817028999328613,
      "mean_token_accuracy": 0.799362127110362,
      "num_tokens": 40991686.0,
      "step": 20740
    },
    {
      "entropy": 1.022671365365386,
      "epoch": 0.5943600246337167,
      "grad_norm": 7.8125,
      "learning_rate": 4.0511290751732524e-05,
      "loss": 1.018362045288086,
      "mean_token_accuracy": 0.7605372924357653,
      "num_tokens": 41011188.0,
      "step": 20750
    },
    {
      "entropy": 0.9238252894952893,
      "epoch": 0.5946464631998052,
      "grad_norm": 7.75,
      "learning_rate": 4.050233957834846e-05,
      "loss": 0.9521265983581543,
      "mean_token_accuracy": 0.7810578010976315,
      "num_tokens": 41030196.0,
      "step": 20760
    },
    {
      "entropy": 0.8961081322282552,
      "epoch": 0.5949329017658938,
      "grad_norm": 9.75,
      "learning_rate": 4.0493385174855505e-05,
      "loss": 0.9701844215393066,
      "mean_token_accuracy": 0.7829433962702751,
      "num_tokens": 41048712.0,
      "step": 20770
    },
    {
      "entropy": 0.907502195239067,
      "epoch": 0.5952193403319823,
      "grad_norm": 9.375,
      "learning_rate": 4.0484427543119405e-05,
      "loss": 0.9321063995361328,
      "mean_token_accuracy": 0.7844234146177769,
      "num_tokens": 41069707.0,
      "step": 20780
    },
    {
      "entropy": 0.9433350671082735,
      "epoch": 0.5955057788980709,
      "grad_norm": 6.21875,
      "learning_rate": 4.047546668500662e-05,
      "loss": 0.9086404800415039,
      "mean_token_accuracy": 0.7825839649885893,
      "num_tokens": 41089278.0,
      "step": 20790
    },
    {
      "entropy": 0.8954040778800845,
      "epoch": 0.5957922174641593,
      "grad_norm": 8.75,
      "learning_rate": 4.046650260238424e-05,
      "loss": 0.8975656509399415,
      "mean_token_accuracy": 0.7964496545493602,
      "num_tokens": 41108864.0,
      "step": 20800
    },
    {
      "entropy": 0.9213673248887062,
      "epoch": 0.5960786560302479,
      "grad_norm": 7.5,
      "learning_rate": 4.0457535297120055e-05,
      "loss": 0.9145282745361328,
      "mean_token_accuracy": 0.7992996439337731,
      "num_tokens": 41127067.0,
      "step": 20810
    },
    {
      "entropy": 0.9741287283599377,
      "epoch": 0.5963650945963365,
      "grad_norm": 8.75,
      "learning_rate": 4.044856477108251e-05,
      "loss": 0.9720040321350097,
      "mean_token_accuracy": 0.7732558086514473,
      "num_tokens": 41147793.0,
      "step": 20820
    },
    {
      "entropy": 0.8446151612326502,
      "epoch": 0.596651533162425,
      "grad_norm": 6.6875,
      "learning_rate": 4.043959102614073e-05,
      "loss": 0.9002360343933106,
      "mean_token_accuracy": 0.7926660932600498,
      "num_tokens": 41167846.0,
      "step": 20830
    },
    {
      "entropy": 0.8957037634216249,
      "epoch": 0.5969379717285135,
      "grad_norm": 7.96875,
      "learning_rate": 4.043061406416452e-05,
      "loss": 0.9112488746643066,
      "mean_token_accuracy": 0.7921467173844576,
      "num_tokens": 41186980.0,
      "step": 20840
    },
    {
      "entropy": 0.9258665155619383,
      "epoch": 0.597224410294602,
      "grad_norm": 7.0625,
      "learning_rate": 4.0421633887024323e-05,
      "loss": 0.9813747406005859,
      "mean_token_accuracy": 0.7790419347584248,
      "num_tokens": 41206120.0,
      "step": 20850
    },
    {
      "entropy": 0.9238900225609541,
      "epoch": 0.5975108488606906,
      "grad_norm": 10.6875,
      "learning_rate": 4.041265049659128e-05,
      "loss": 0.9519866943359375,
      "mean_token_accuracy": 0.779297724366188,
      "num_tokens": 41226601.0,
      "step": 20860
    },
    {
      "entropy": 1.0661541579291225,
      "epoch": 0.5977972874267792,
      "grad_norm": 7.34375,
      "learning_rate": 4.0403663894737205e-05,
      "loss": 1.0956396102905273,
      "mean_token_accuracy": 0.7528447434306145,
      "num_tokens": 41247029.0,
      "step": 20870
    },
    {
      "entropy": 0.8536147611215711,
      "epoch": 0.5980837259928676,
      "grad_norm": 6.4375,
      "learning_rate": 4.039467408333456e-05,
      "loss": 0.8158473968505859,
      "mean_token_accuracy": 0.8012623369693757,
      "num_tokens": 41266124.0,
      "step": 20880
    },
    {
      "entropy": 0.887767580896616,
      "epoch": 0.5983701645589562,
      "grad_norm": 7.28125,
      "learning_rate": 4.0385681064256485e-05,
      "loss": 0.9154246330261231,
      "mean_token_accuracy": 0.7874515231698751,
      "num_tokens": 41285817.0,
      "step": 20890
    },
    {
      "entropy": 0.8830116217955947,
      "epoch": 0.5986566031250448,
      "grad_norm": 8.0625,
      "learning_rate": 4.03766848393768e-05,
      "loss": 0.853181266784668,
      "mean_token_accuracy": 0.7987013965845108,
      "num_tokens": 41304123.0,
      "step": 20900
    },
    {
      "entropy": 0.9142610454931855,
      "epoch": 0.5989430416911333,
      "grad_norm": 7.3125,
      "learning_rate": 4.036768541056997e-05,
      "loss": 0.9103652000427246,
      "mean_token_accuracy": 0.7947945483028889,
      "num_tokens": 41324236.0,
      "step": 20910
    },
    {
      "entropy": 0.984237615764141,
      "epoch": 0.5992294802572218,
      "grad_norm": 6.25,
      "learning_rate": 4.0358682779711156e-05,
      "loss": 0.9487616539001464,
      "mean_token_accuracy": 0.7826851081103087,
      "num_tokens": 41344647.0,
      "step": 20920
    },
    {
      "entropy": 0.8736257541924715,
      "epoch": 0.5995159188233103,
      "grad_norm": 8.9375,
      "learning_rate": 4.034967694867615e-05,
      "loss": 0.8735903739929199,
      "mean_token_accuracy": 0.7949574064463377,
      "num_tokens": 41363345.0,
      "step": 20930
    },
    {
      "entropy": 0.9117791803553701,
      "epoch": 0.5998023573893989,
      "grad_norm": 8.5625,
      "learning_rate": 4.034066791934144e-05,
      "loss": 1.0124193191528321,
      "mean_token_accuracy": 0.7686371643096208,
      "num_tokens": 41383833.0,
      "step": 20940
    },
    {
      "entropy": 0.9353847696445883,
      "epoch": 0.6000887959554875,
      "grad_norm": 6.59375,
      "learning_rate": 4.0331655693584176e-05,
      "loss": 0.9405402183532715,
      "mean_token_accuracy": 0.7820982005447149,
      "num_tokens": 41403314.0,
      "step": 20950
    },
    {
      "entropy": 0.8074826346710324,
      "epoch": 0.600375234521576,
      "grad_norm": 7.5625,
      "learning_rate": 4.0322640273282165e-05,
      "loss": 0.7605173587799072,
      "mean_token_accuracy": 0.8140549000352622,
      "num_tokens": 41423274.0,
      "step": 20960
    },
    {
      "entropy": 0.9272844202816486,
      "epoch": 0.6006616730876645,
      "grad_norm": 9.5625,
      "learning_rate": 4.031362166031388e-05,
      "loss": 0.9665066719055175,
      "mean_token_accuracy": 0.7756766356527806,
      "num_tokens": 41442145.0,
      "step": 20970
    },
    {
      "entropy": 0.9357972528785468,
      "epoch": 0.6009481116537531,
      "grad_norm": 7.96875,
      "learning_rate": 4.030459985655848e-05,
      "loss": 0.9401085853576661,
      "mean_token_accuracy": 0.7800631914287806,
      "num_tokens": 41461342.0,
      "step": 20980
    },
    {
      "entropy": 0.9227657537907362,
      "epoch": 0.6012345502198416,
      "grad_norm": 7.21875,
      "learning_rate": 4.029557486389574e-05,
      "loss": 0.9202791213989258,
      "mean_token_accuracy": 0.7881428830325603,
      "num_tokens": 41480546.0,
      "step": 20990
    },
    {
      "entropy": 0.9187023758888244,
      "epoch": 0.6015209887859302,
      "grad_norm": 7.09375,
      "learning_rate": 4.028654668420617e-05,
      "loss": 0.9150511741638183,
      "mean_token_accuracy": 0.7903459273278713,
      "num_tokens": 41500097.0,
      "step": 21000
    },
    {
      "entropy": 1.0126201277598739,
      "epoch": 0.6018074273520186,
      "grad_norm": 7.09375,
      "learning_rate": 4.027751531937088e-05,
      "loss": 1.0558706283569337,
      "mean_token_accuracy": 0.7638945899903774,
      "num_tokens": 41521353.0,
      "step": 21010
    },
    {
      "entropy": 1.014037550985813,
      "epoch": 0.6020938659181072,
      "grad_norm": 7.28125,
      "learning_rate": 4.0268480771271686e-05,
      "loss": 1.0154370307922362,
      "mean_token_accuracy": 0.7747875913977623,
      "num_tokens": 41541311.0,
      "step": 21020
    },
    {
      "entropy": 0.862037887610495,
      "epoch": 0.6023803044841958,
      "grad_norm": 6.0625,
      "learning_rate": 4.0259443041791034e-05,
      "loss": 0.8692997932434082,
      "mean_token_accuracy": 0.7942831709980964,
      "num_tokens": 41561438.0,
      "step": 21030
    },
    {
      "entropy": 1.0233358446508647,
      "epoch": 0.6026667430502843,
      "grad_norm": 5.625,
      "learning_rate": 4.025040213281206e-05,
      "loss": 0.980075454711914,
      "mean_token_accuracy": 0.7743218224495649,
      "num_tokens": 41581716.0,
      "step": 21040
    },
    {
      "entropy": 0.9603341985493898,
      "epoch": 0.6029531816163728,
      "grad_norm": 7.65625,
      "learning_rate": 4.0241358046218545e-05,
      "loss": 0.967650318145752,
      "mean_token_accuracy": 0.7738530497998,
      "num_tokens": 41600950.0,
      "step": 21050
    },
    {
      "entropy": 0.9888514935970306,
      "epoch": 0.6032396201824614,
      "grad_norm": 8.8125,
      "learning_rate": 4.023231078389495e-05,
      "loss": 1.0712069511413573,
      "mean_token_accuracy": 0.756508257985115,
      "num_tokens": 41622246.0,
      "step": 21060
    },
    {
      "entropy": 0.9421109084039927,
      "epoch": 0.6035260587485499,
      "grad_norm": 7.375,
      "learning_rate": 4.022326034772637e-05,
      "loss": 0.9729910850524902,
      "mean_token_accuracy": 0.7771270956844092,
      "num_tokens": 41641908.0,
      "step": 21070
    },
    {
      "entropy": 0.9741461761295795,
      "epoch": 0.6038124973146385,
      "grad_norm": 7.59375,
      "learning_rate": 4.021420673959859e-05,
      "loss": 1.000966739654541,
      "mean_token_accuracy": 0.7841028429567813,
      "num_tokens": 41660674.0,
      "step": 21080
    },
    {
      "entropy": 0.9843042684718967,
      "epoch": 0.6040989358807269,
      "grad_norm": 8.25,
      "learning_rate": 4.020514996139804e-05,
      "loss": 0.9296276092529296,
      "mean_token_accuracy": 0.7809163060039281,
      "num_tokens": 41679428.0,
      "step": 21090
    },
    {
      "entropy": 0.8387097787111998,
      "epoch": 0.6043853744468155,
      "grad_norm": 5.40625,
      "learning_rate": 4.019609001501182e-05,
      "loss": 0.8288966178894043,
      "mean_token_accuracy": 0.8085266873240471,
      "num_tokens": 41698048.0,
      "step": 21100
    },
    {
      "entropy": 0.9606157869100571,
      "epoch": 0.6046718130129041,
      "grad_norm": 10.4375,
      "learning_rate": 4.018702690232768e-05,
      "loss": 0.983890438079834,
      "mean_token_accuracy": 0.7711589809507131,
      "num_tokens": 41717401.0,
      "step": 21110
    },
    {
      "entropy": 0.918721004575491,
      "epoch": 0.6049582515789926,
      "grad_norm": 7.03125,
      "learning_rate": 4.0177960625234034e-05,
      "loss": 0.9903434753417969,
      "mean_token_accuracy": 0.7762091536074877,
      "num_tokens": 41738753.0,
      "step": 21120
    },
    {
      "entropy": 0.9909901682287454,
      "epoch": 0.6052446901450811,
      "grad_norm": 7.125,
      "learning_rate": 4.016889118561996e-05,
      "loss": 0.9889702796936035,
      "mean_token_accuracy": 0.7745804753154516,
      "num_tokens": 41759950.0,
      "step": 21130
    },
    {
      "entropy": 0.9451931834220886,
      "epoch": 0.6055311287111697,
      "grad_norm": 8.75,
      "learning_rate": 4.0159818585375184e-05,
      "loss": 0.9409404754638672,
      "mean_token_accuracy": 0.7859415173530578,
      "num_tokens": 41778882.0,
      "step": 21140
    },
    {
      "entropy": 0.9710013676434756,
      "epoch": 0.6058175672772582,
      "grad_norm": 7.9375,
      "learning_rate": 4.015074282639011e-05,
      "loss": 1.022063636779785,
      "mean_token_accuracy": 0.7670025512576103,
      "num_tokens": 41798556.0,
      "step": 21150
    },
    {
      "entropy": 0.9538961915299297,
      "epoch": 0.6061040058433468,
      "grad_norm": 6.0,
      "learning_rate": 4.014166391055577e-05,
      "loss": 0.9876542091369629,
      "mean_token_accuracy": 0.7772497501224279,
      "num_tokens": 41818762.0,
      "step": 21160
    },
    {
      "entropy": 0.9731216883286834,
      "epoch": 0.6063904444094352,
      "grad_norm": 6.28125,
      "learning_rate": 4.013258183976388e-05,
      "loss": 0.9990914344787598,
      "mean_token_accuracy": 0.7741889987140894,
      "num_tokens": 41839043.0,
      "step": 21170
    },
    {
      "entropy": 0.9369466709904373,
      "epoch": 0.6066768829755238,
      "grad_norm": 8.3125,
      "learning_rate": 4.01234966159068e-05,
      "loss": 0.9545680046081543,
      "mean_token_accuracy": 0.7743865631520748,
      "num_tokens": 41858436.0,
      "step": 21180
    },
    {
      "entropy": 0.9481218051165342,
      "epoch": 0.6069633215416124,
      "grad_norm": 7.53125,
      "learning_rate": 4.011440824087757e-05,
      "loss": 0.9469070434570312,
      "mean_token_accuracy": 0.7886612102389335,
      "num_tokens": 41878571.0,
      "step": 21190
    },
    {
      "entropy": 0.9387059917673468,
      "epoch": 0.6072497601077009,
      "grad_norm": 10.1875,
      "learning_rate": 4.010531671656984e-05,
      "loss": 0.8841951370239258,
      "mean_token_accuracy": 0.7964214541018009,
      "num_tokens": 41897762.0,
      "step": 21200
    },
    {
      "entropy": 0.9549746591597795,
      "epoch": 0.6075361986737894,
      "grad_norm": 7.8125,
      "learning_rate": 4.009622204487795e-05,
      "loss": 0.9917261123657226,
      "mean_token_accuracy": 0.7753554556518794,
      "num_tokens": 41918338.0,
      "step": 21210
    },
    {
      "entropy": 0.8990516319870949,
      "epoch": 0.607822637239878,
      "grad_norm": 7.4375,
      "learning_rate": 4.0087124227696913e-05,
      "loss": 0.8554092407226562,
      "mean_token_accuracy": 0.7967605404555798,
      "num_tokens": 41938507.0,
      "step": 21220
    },
    {
      "entropy": 0.9326813146471977,
      "epoch": 0.6081090758059665,
      "grad_norm": 7.5625,
      "learning_rate": 4.007802326692236e-05,
      "loss": 0.9167994499206543,
      "mean_token_accuracy": 0.7798323288559914,
      "num_tokens": 41958426.0,
      "step": 21230
    },
    {
      "entropy": 0.8656520204618573,
      "epoch": 0.6083955143720551,
      "grad_norm": 6.8125,
      "learning_rate": 4.006891916445058e-05,
      "loss": 0.8320615768432618,
      "mean_token_accuracy": 0.7998116254806519,
      "num_tokens": 41978095.0,
      "step": 21240
    },
    {
      "entropy": 0.993567768856883,
      "epoch": 0.6086819529381436,
      "grad_norm": 9.8125,
      "learning_rate": 4.005981192217855e-05,
      "loss": 1.0638136863708496,
      "mean_token_accuracy": 0.7658849764615298,
      "num_tokens": 41997571.0,
      "step": 21250
    },
    {
      "entropy": 0.9665811860933899,
      "epoch": 0.6089683915042321,
      "grad_norm": 6.84375,
      "learning_rate": 4.005070154200386e-05,
      "loss": 0.9924683570861816,
      "mean_token_accuracy": 0.7683881558477879,
      "num_tokens": 42019737.0,
      "step": 21260
    },
    {
      "entropy": 0.8849428739398718,
      "epoch": 0.6092548300703207,
      "grad_norm": 8.5,
      "learning_rate": 4.004158802582478e-05,
      "loss": 0.9356993675231934,
      "mean_token_accuracy": 0.7801951166242361,
      "num_tokens": 42037593.0,
      "step": 21270
    },
    {
      "entropy": 0.9559132300317288,
      "epoch": 0.6095412686364092,
      "grad_norm": 8.8125,
      "learning_rate": 4.003247137554024e-05,
      "loss": 0.9920024871826172,
      "mean_token_accuracy": 0.7765748884528876,
      "num_tokens": 42056687.0,
      "step": 21280
    },
    {
      "entropy": 0.9555912531912327,
      "epoch": 0.6098277072024978,
      "grad_norm": 6.875,
      "learning_rate": 4.0023351593049786e-05,
      "loss": 0.9226361274719238,
      "mean_token_accuracy": 0.7857959181070328,
      "num_tokens": 42076948.0,
      "step": 21290
    },
    {
      "entropy": 0.8304474165663123,
      "epoch": 0.6101141457685862,
      "grad_norm": 7.9375,
      "learning_rate": 4.0014228680253655e-05,
      "loss": 0.8037498474121094,
      "mean_token_accuracy": 0.8073724310845136,
      "num_tokens": 42096267.0,
      "step": 21300
    },
    {
      "entropy": 0.91441769618541,
      "epoch": 0.6104005843346748,
      "grad_norm": 8.8125,
      "learning_rate": 4.000510263905273e-05,
      "loss": 0.9193489074707031,
      "mean_token_accuracy": 0.7954089157283306,
      "num_tokens": 42115842.0,
      "step": 21310
    },
    {
      "entropy": 0.8459574146196246,
      "epoch": 0.6106870229007634,
      "grad_norm": 4.875,
      "learning_rate": 3.999597347134851e-05,
      "loss": 0.8010165214538574,
      "mean_token_accuracy": 0.8099937357008458,
      "num_tokens": 42136260.0,
      "step": 21320
    },
    {
      "entropy": 0.8444384815171361,
      "epoch": 0.6109734614668519,
      "grad_norm": 7.59375,
      "learning_rate": 3.9986841179043195e-05,
      "loss": 0.8454648017883301,
      "mean_token_accuracy": 0.7997375346720219,
      "num_tokens": 42155509.0,
      "step": 21330
    },
    {
      "entropy": 1.0567572874948383,
      "epoch": 0.6112599000329404,
      "grad_norm": 7.875,
      "learning_rate": 3.997770576403961e-05,
      "loss": 1.1003796577453613,
      "mean_token_accuracy": 0.7578796736896038,
      "num_tokens": 42175258.0,
      "step": 21340
    },
    {
      "entropy": 0.8765289297327399,
      "epoch": 0.611546338599029,
      "grad_norm": 7.6875,
      "learning_rate": 3.9968567228241225e-05,
      "loss": 0.8882092475891114,
      "mean_token_accuracy": 0.7994453281164169,
      "num_tokens": 42194351.0,
      "step": 21350
    },
    {
      "entropy": 0.7483715382404625,
      "epoch": 0.6118327771651175,
      "grad_norm": 8.25,
      "learning_rate": 3.9959425573552176e-05,
      "loss": 0.7966334342956543,
      "mean_token_accuracy": 0.8065416298806667,
      "num_tokens": 42212865.0,
      "step": 21360
    },
    {
      "entropy": 0.941635518707335,
      "epoch": 0.6121192157312061,
      "grad_norm": 11.0625,
      "learning_rate": 3.995028080187725e-05,
      "loss": 0.9472028732299804,
      "mean_token_accuracy": 0.7869414288550616,
      "num_tokens": 42231316.0,
      "step": 21370
    },
    {
      "entropy": 0.9518864756450057,
      "epoch": 0.6124056542972945,
      "grad_norm": 6.34375,
      "learning_rate": 3.9941132915121864e-05,
      "loss": 1.0018354415893556,
      "mean_token_accuracy": 0.771107779070735,
      "num_tokens": 42251803.0,
      "step": 21380
    },
    {
      "entropy": 0.886208895035088,
      "epoch": 0.6126920928633831,
      "grad_norm": 7.6875,
      "learning_rate": 3.993198191519211e-05,
      "loss": 0.8533462524414063,
      "mean_token_accuracy": 0.7929553411900997,
      "num_tokens": 42271667.0,
      "step": 21390
    },
    {
      "entropy": 0.8870396001264453,
      "epoch": 0.6129785314294717,
      "grad_norm": 8.0625,
      "learning_rate": 3.992282780399469e-05,
      "loss": 0.8210672378540039,
      "mean_token_accuracy": 0.8161398217082023,
      "num_tokens": 42289668.0,
      "step": 21400
    },
    {
      "entropy": 1.0642244413495063,
      "epoch": 0.6132649699955602,
      "grad_norm": 6.46875,
      "learning_rate": 3.9913670583437014e-05,
      "loss": 1.1320651054382325,
      "mean_token_accuracy": 0.7501522053033114,
      "num_tokens": 42309435.0,
      "step": 21410
    },
    {
      "entropy": 0.8919663559645414,
      "epoch": 0.6135514085616487,
      "grad_norm": 6.78125,
      "learning_rate": 3.990451025542707e-05,
      "loss": 0.8975013732910156,
      "mean_token_accuracy": 0.7871919363737107,
      "num_tokens": 42328783.0,
      "step": 21420
    },
    {
      "entropy": 0.8999335646629334,
      "epoch": 0.6138378471277373,
      "grad_norm": 6.375,
      "learning_rate": 3.989534682187356e-05,
      "loss": 0.8726045608520507,
      "mean_token_accuracy": 0.8020846236497163,
      "num_tokens": 42349063.0,
      "step": 21430
    },
    {
      "entropy": 0.949260525777936,
      "epoch": 0.6141242856938258,
      "grad_norm": 8.1875,
      "learning_rate": 3.988618028468578e-05,
      "loss": 1.0122108459472656,
      "mean_token_accuracy": 0.7629811208695173,
      "num_tokens": 42367728.0,
      "step": 21440
    },
    {
      "entropy": 0.9180545607581735,
      "epoch": 0.6144107242599144,
      "grad_norm": 9.4375,
      "learning_rate": 3.9877010645773696e-05,
      "loss": 0.9360054969787598,
      "mean_token_accuracy": 0.788599856942892,
      "num_tokens": 42387145.0,
      "step": 21450
    },
    {
      "entropy": 1.0036912895739079,
      "epoch": 0.6146971628260028,
      "grad_norm": 7.34375,
      "learning_rate": 3.986783790704792e-05,
      "loss": 1.0235818862915038,
      "mean_token_accuracy": 0.7695517284795642,
      "num_tokens": 42407185.0,
      "step": 21460
    },
    {
      "entropy": 0.919987709634006,
      "epoch": 0.6149836013920914,
      "grad_norm": 6.9375,
      "learning_rate": 3.985866207041972e-05,
      "loss": 0.9531692504882813,
      "mean_token_accuracy": 0.7836550898849964,
      "num_tokens": 42425639.0,
      "step": 21470
    },
    {
      "entropy": 1.0659559171646833,
      "epoch": 0.61527003995818,
      "grad_norm": 8.6875,
      "learning_rate": 3.9849483137800984e-05,
      "loss": 1.1032164573669434,
      "mean_token_accuracy": 0.7574559256434441,
      "num_tokens": 42446794.0,
      "step": 21480
    },
    {
      "entropy": 0.9576913375407458,
      "epoch": 0.6155564785242685,
      "grad_norm": 8.25,
      "learning_rate": 3.984030111110426e-05,
      "loss": 0.9815732002258301,
      "mean_token_accuracy": 0.7753238756209612,
      "num_tokens": 42468295.0,
      "step": 21490
    },
    {
      "entropy": 0.9121747886762023,
      "epoch": 0.6158429170903571,
      "grad_norm": 9.1875,
      "learning_rate": 3.983111599224275e-05,
      "loss": 0.9027762413024902,
      "mean_token_accuracy": 0.7837694879621268,
      "num_tokens": 42487238.0,
      "step": 21500
    },
    {
      "entropy": 0.9780509900301695,
      "epoch": 0.6161293556564456,
      "grad_norm": 9.0,
      "learning_rate": 3.982192778313029e-05,
      "loss": 0.9976916313171387,
      "mean_token_accuracy": 0.7688472300767899,
      "num_tokens": 42507164.0,
      "step": 21510
    },
    {
      "entropy": 0.9557992301881313,
      "epoch": 0.6164157942225341,
      "grad_norm": 7.34375,
      "learning_rate": 3.981273648568134e-05,
      "loss": 0.9735465049743652,
      "mean_token_accuracy": 0.7786841005086899,
      "num_tokens": 42527070.0,
      "step": 21520
    },
    {
      "entropy": 0.9095773328095674,
      "epoch": 0.6167022327886227,
      "grad_norm": 8.1875,
      "learning_rate": 3.9803542101811044e-05,
      "loss": 0.9097620010375976,
      "mean_token_accuracy": 0.7947625555098057,
      "num_tokens": 42547018.0,
      "step": 21530
    },
    {
      "entropy": 0.9053039344027638,
      "epoch": 0.6169886713547112,
      "grad_norm": 9.8125,
      "learning_rate": 3.979434463343515e-05,
      "loss": 0.8891491889953613,
      "mean_token_accuracy": 0.8050674170255661,
      "num_tokens": 42566533.0,
      "step": 21540
    },
    {
      "entropy": 0.8114087164402009,
      "epoch": 0.6172751099207997,
      "grad_norm": 8.8125,
      "learning_rate": 3.978514408247009e-05,
      "loss": 0.8709220886230469,
      "mean_token_accuracy": 0.7960758753120899,
      "num_tokens": 42587702.0,
      "step": 21550
    },
    {
      "entropy": 0.7838315076194704,
      "epoch": 0.6175615484868883,
      "grad_norm": 8.125,
      "learning_rate": 3.97759404508329e-05,
      "loss": 0.818719482421875,
      "mean_token_accuracy": 0.8078423563390971,
      "num_tokens": 42607044.0,
      "step": 21560
    },
    {
      "entropy": 0.9934518793597817,
      "epoch": 0.6178479870529768,
      "grad_norm": 10.5,
      "learning_rate": 3.9766733740441274e-05,
      "loss": 0.9469232559204102,
      "mean_token_accuracy": 0.7810003712773324,
      "num_tokens": 42626707.0,
      "step": 21570
    },
    {
      "entropy": 0.885506845265627,
      "epoch": 0.6181344256190654,
      "grad_norm": 7.75,
      "learning_rate": 3.975752395321355e-05,
      "loss": 0.8913700103759765,
      "mean_token_accuracy": 0.7940637603402138,
      "num_tokens": 42645791.0,
      "step": 21580
    },
    {
      "entropy": 0.8530406372621655,
      "epoch": 0.6184208641851539,
      "grad_norm": 7.46875,
      "learning_rate": 3.974831109106871e-05,
      "loss": 0.8667236328125,
      "mean_token_accuracy": 0.7981486685574055,
      "num_tokens": 42663516.0,
      "step": 21590
    },
    {
      "entropy": 1.0198580898344516,
      "epoch": 0.6187073027512424,
      "grad_norm": 5.59375,
      "learning_rate": 3.973909515592636e-05,
      "loss": 1.0000584602355957,
      "mean_token_accuracy": 0.7748429756611586,
      "num_tokens": 42683553.0,
      "step": 21600
    },
    {
      "entropy": 0.8609779054298997,
      "epoch": 0.618993741317331,
      "grad_norm": 9.25,
      "learning_rate": 3.972987614970676e-05,
      "loss": 0.8467598915100097,
      "mean_token_accuracy": 0.8031350560486317,
      "num_tokens": 42703277.0,
      "step": 21610
    },
    {
      "entropy": 0.9730033710598945,
      "epoch": 0.6192801798834195,
      "grad_norm": 9.375,
      "learning_rate": 3.97206540743308e-05,
      "loss": 0.9542336463928223,
      "mean_token_accuracy": 0.7802342090755701,
      "num_tokens": 42723065.0,
      "step": 21620
    },
    {
      "entropy": 0.9092322899028659,
      "epoch": 0.619566618449508,
      "grad_norm": 9.625,
      "learning_rate": 3.971142893172004e-05,
      "loss": 0.9304965972900391,
      "mean_token_accuracy": 0.7851572770625352,
      "num_tokens": 42742077.0,
      "step": 21630
    },
    {
      "entropy": 0.8767265608534217,
      "epoch": 0.6198530570155966,
      "grad_norm": 8.375,
      "learning_rate": 3.970220072379662e-05,
      "loss": 0.908015251159668,
      "mean_token_accuracy": 0.7878688059747219,
      "num_tokens": 42762527.0,
      "step": 21640
    },
    {
      "entropy": 0.8960085434839129,
      "epoch": 0.6201394955816851,
      "grad_norm": 10.125,
      "learning_rate": 3.9692969452483384e-05,
      "loss": 0.882050895690918,
      "mean_token_accuracy": 0.7907858695834875,
      "num_tokens": 42782039.0,
      "step": 21650
    },
    {
      "entropy": 0.8196407787501812,
      "epoch": 0.6204259341477737,
      "grad_norm": 8.25,
      "learning_rate": 3.968373511970378e-05,
      "loss": 0.8495737075805664,
      "mean_token_accuracy": 0.794497049599886,
      "num_tokens": 42801074.0,
      "step": 21660
    },
    {
      "entropy": 0.9734192976728082,
      "epoch": 0.6207123727138621,
      "grad_norm": 10.875,
      "learning_rate": 3.967449772738188e-05,
      "loss": 0.9475679397583008,
      "mean_token_accuracy": 0.7783634439110756,
      "num_tokens": 42821565.0,
      "step": 21670
    },
    {
      "entropy": 0.8653921574354172,
      "epoch": 0.6209988112799507,
      "grad_norm": 7.15625,
      "learning_rate": 3.966525727744243e-05,
      "loss": 0.8672385215759277,
      "mean_token_accuracy": 0.8005895636975765,
      "num_tokens": 42840658.0,
      "step": 21680
    },
    {
      "entropy": 0.9960725169628859,
      "epoch": 0.6212852498460393,
      "grad_norm": 7.4375,
      "learning_rate": 3.965601377181079e-05,
      "loss": 1.001399326324463,
      "mean_token_accuracy": 0.7646365784108639,
      "num_tokens": 42858920.0,
      "step": 21690
    },
    {
      "entropy": 0.976169890165329,
      "epoch": 0.6215716884121278,
      "grad_norm": 8.25,
      "learning_rate": 3.964676721241295e-05,
      "loss": 0.9807143211364746,
      "mean_token_accuracy": 0.7736262317746878,
      "num_tokens": 42878650.0,
      "step": 21700
    },
    {
      "entropy": 0.9904242061078549,
      "epoch": 0.6218581269782163,
      "grad_norm": 8.375,
      "learning_rate": 3.9637517601175564e-05,
      "loss": 1.0749170303344726,
      "mean_token_accuracy": 0.756168657541275,
      "num_tokens": 42898078.0,
      "step": 21710
    },
    {
      "entropy": 0.8459861533716321,
      "epoch": 0.6221445655443049,
      "grad_norm": 7.46875,
      "learning_rate": 3.96282649400259e-05,
      "loss": 0.8309358596801758,
      "mean_token_accuracy": 0.8131593700498343,
      "num_tokens": 42917169.0,
      "step": 21720
    },
    {
      "entropy": 0.8487146643921732,
      "epoch": 0.6224310041103934,
      "grad_norm": 7.9375,
      "learning_rate": 3.961900923089186e-05,
      "loss": 0.8306548118591308,
      "mean_token_accuracy": 0.8059241883456707,
      "num_tokens": 42936288.0,
      "step": 21730
    },
    {
      "entropy": 0.8764183197170496,
      "epoch": 0.622717442676482,
      "grad_norm": 6.28125,
      "learning_rate": 3.9609750475702e-05,
      "loss": 0.8847857475280761,
      "mean_token_accuracy": 0.791563218086958,
      "num_tokens": 42955226.0,
      "step": 21740
    },
    {
      "entropy": 0.9937205128371716,
      "epoch": 0.6230038812425704,
      "grad_norm": 7.15625,
      "learning_rate": 3.9600488676385504e-05,
      "loss": 0.9262320518493652,
      "mean_token_accuracy": 0.7865871626883745,
      "num_tokens": 42974860.0,
      "step": 21750
    },
    {
      "entropy": 0.941679079644382,
      "epoch": 0.623290319808659,
      "grad_norm": 9.75,
      "learning_rate": 3.959122383487216e-05,
      "loss": 0.9909570693969727,
      "mean_token_accuracy": 0.7756388328969479,
      "num_tokens": 42994902.0,
      "step": 21760
    },
    {
      "entropy": 0.9695995369926095,
      "epoch": 0.6235767583747476,
      "grad_norm": 6.375,
      "learning_rate": 3.9581955953092435e-05,
      "loss": 0.9894150733947754,
      "mean_token_accuracy": 0.7699768807739019,
      "num_tokens": 43015736.0,
      "step": 21770
    },
    {
      "entropy": 1.0112689251080156,
      "epoch": 0.6238631969408361,
      "grad_norm": 9.0,
      "learning_rate": 3.9572685032977405e-05,
      "loss": 1.044120979309082,
      "mean_token_accuracy": 0.7665673263370991,
      "num_tokens": 43034517.0,
      "step": 21780
    },
    {
      "entropy": 0.898890046030283,
      "epoch": 0.6241496355069247,
      "grad_norm": 8.6875,
      "learning_rate": 3.956341107645879e-05,
      "loss": 1.0126422882080077,
      "mean_token_accuracy": 0.7743716809898615,
      "num_tokens": 43055043.0,
      "step": 21790
    },
    {
      "entropy": 0.99359982162714,
      "epoch": 0.6244360740730132,
      "grad_norm": 9.5,
      "learning_rate": 3.955413408546892e-05,
      "loss": 1.012258529663086,
      "mean_token_accuracy": 0.7691522307693959,
      "num_tokens": 43075370.0,
      "step": 21800
    },
    {
      "entropy": 0.9251581087708474,
      "epoch": 0.6247225126391017,
      "grad_norm": 8.4375,
      "learning_rate": 3.95448540619408e-05,
      "loss": 0.8824054718017578,
      "mean_token_accuracy": 0.7972671285271644,
      "num_tokens": 43094072.0,
      "step": 21810
    },
    {
      "entropy": 0.857103344425559,
      "epoch": 0.6250089512051903,
      "grad_norm": 6.375,
      "learning_rate": 3.9535571007808006e-05,
      "loss": 0.8577948570251465,
      "mean_token_accuracy": 0.7906800910830498,
      "num_tokens": 43113252.0,
      "step": 21820
    },
    {
      "entropy": 0.9064915968105197,
      "epoch": 0.6252953897712789,
      "grad_norm": 8.375,
      "learning_rate": 3.9526284925004794e-05,
      "loss": 0.8851880073547364,
      "mean_token_accuracy": 0.79220184199512,
      "num_tokens": 43131984.0,
      "step": 21830
    },
    {
      "entropy": 0.9035031069070101,
      "epoch": 0.6255818283373673,
      "grad_norm": 10.4375,
      "learning_rate": 3.951699581546606e-05,
      "loss": 0.8989022254943848,
      "mean_token_accuracy": 0.7871453911066055,
      "num_tokens": 43151513.0,
      "step": 21840
    },
    {
      "entropy": 0.9971307696774602,
      "epoch": 0.6258682669034559,
      "grad_norm": 8.25,
      "learning_rate": 3.9507703681127277e-05,
      "loss": 1.0234750747680663,
      "mean_token_accuracy": 0.7748494818806648,
      "num_tokens": 43172284.0,
      "step": 21850
    },
    {
      "entropy": 0.9900875061750412,
      "epoch": 0.6261547054695444,
      "grad_norm": 8.4375,
      "learning_rate": 3.949840852392459e-05,
      "loss": 1.0449282646179199,
      "mean_token_accuracy": 0.7648820500820875,
      "num_tokens": 43192658.0,
      "step": 21860
    },
    {
      "entropy": 0.913403133675456,
      "epoch": 0.626441144035633,
      "grad_norm": 6.1875,
      "learning_rate": 3.948911034579477e-05,
      "loss": 0.9260476112365723,
      "mean_token_accuracy": 0.788822765275836,
      "num_tokens": 43212271.0,
      "step": 21870
    },
    {
      "entropy": 0.9513774402439594,
      "epoch": 0.6267275826017215,
      "grad_norm": 5.21875,
      "learning_rate": 3.9479809148675206e-05,
      "loss": 0.9205931663513184,
      "mean_token_accuracy": 0.7873339593410492,
      "num_tokens": 43233004.0,
      "step": 21880
    },
    {
      "entropy": 0.9041346196085215,
      "epoch": 0.62701402116781,
      "grad_norm": 8.0625,
      "learning_rate": 3.9470504934503904e-05,
      "loss": 0.9279412269592285,
      "mean_token_accuracy": 0.779343643411994,
      "num_tokens": 43253174.0,
      "step": 21890
    },
    {
      "entropy": 0.9794464088976383,
      "epoch": 0.6273004597338986,
      "grad_norm": 6.78125,
      "learning_rate": 3.9461197705219546e-05,
      "loss": 0.9935440063476563,
      "mean_token_accuracy": 0.7687275119125843,
      "num_tokens": 43272616.0,
      "step": 21900
    },
    {
      "entropy": 0.9813774359878152,
      "epoch": 0.6275868982999872,
      "grad_norm": 8.0625,
      "learning_rate": 3.9451887462761386e-05,
      "loss": 0.9266609191894531,
      "mean_token_accuracy": 0.7832588758319616,
      "num_tokens": 43292168.0,
      "step": 21910
    },
    {
      "entropy": 0.9154514050111174,
      "epoch": 0.6278733368660756,
      "grad_norm": 7.59375,
      "learning_rate": 3.944257420906934e-05,
      "loss": 0.8751792907714844,
      "mean_token_accuracy": 0.7905846066772938,
      "num_tokens": 43313668.0,
      "step": 21920
    },
    {
      "entropy": 0.8968888252973557,
      "epoch": 0.6281597754321642,
      "grad_norm": 8.25,
      "learning_rate": 3.943325794608393e-05,
      "loss": 0.9274250030517578,
      "mean_token_accuracy": 0.7881258871406317,
      "num_tokens": 43332515.0,
      "step": 21930
    },
    {
      "entropy": 0.9620690947398544,
      "epoch": 0.6284462139982527,
      "grad_norm": 8.3125,
      "learning_rate": 3.942393867574635e-05,
      "loss": 1.0148891448974608,
      "mean_token_accuracy": 0.7651188805699348,
      "num_tokens": 43352326.0,
      "step": 21940
    },
    {
      "entropy": 0.9012795368209481,
      "epoch": 0.6287326525643413,
      "grad_norm": 7.25,
      "learning_rate": 3.941461639999835e-05,
      "loss": 0.951259708404541,
      "mean_token_accuracy": 0.781118031218648,
      "num_tokens": 43371666.0,
      "step": 21950
    },
    {
      "entropy": 0.9552686661481857,
      "epoch": 0.6290190911304298,
      "grad_norm": 6.3125,
      "learning_rate": 3.9405291120782374e-05,
      "loss": 0.998451042175293,
      "mean_token_accuracy": 0.7703643348067999,
      "num_tokens": 43392052.0,
      "step": 21960
    },
    {
      "entropy": 0.9365689104422927,
      "epoch": 0.6293055296965183,
      "grad_norm": 7.21875,
      "learning_rate": 3.939596284004143e-05,
      "loss": 0.872525691986084,
      "mean_token_accuracy": 0.7938596583902836,
      "num_tokens": 43412363.0,
      "step": 21970
    },
    {
      "entropy": 0.9144134005531669,
      "epoch": 0.6295919682626069,
      "grad_norm": 8.1875,
      "learning_rate": 3.938663155971922e-05,
      "loss": 0.9601176261901856,
      "mean_token_accuracy": 0.7727459536865353,
      "num_tokens": 43431662.0,
      "step": 21980
    },
    {
      "entropy": 0.8954760676249862,
      "epoch": 0.6298784068286954,
      "grad_norm": 8.375,
      "learning_rate": 3.937729728176002e-05,
      "loss": 0.8879722595214844,
      "mean_token_accuracy": 0.8007520616054535,
      "num_tokens": 43451461.0,
      "step": 21990
    },
    {
      "entropy": 0.9030739717185498,
      "epoch": 0.6301648453947839,
      "grad_norm": 7.09375,
      "learning_rate": 3.936796000810873e-05,
      "loss": 0.9617762565612793,
      "mean_token_accuracy": 0.7725638631731272,
      "num_tokens": 43469531.0,
      "step": 22000
    },
    {
      "entropy": 0.8425525259226561,
      "epoch": 0.6304512839608725,
      "grad_norm": 7.15625,
      "learning_rate": 3.935861974071091e-05,
      "loss": 0.8613410949707031,
      "mean_token_accuracy": 0.8038687307387591,
      "num_tokens": 43488938.0,
      "step": 22010
    },
    {
      "entropy": 0.9750666160136461,
      "epoch": 0.630737722526961,
      "grad_norm": 5.5,
      "learning_rate": 3.934927648151271e-05,
      "loss": 0.994815444946289,
      "mean_token_accuracy": 0.7835616357624531,
      "num_tokens": 43508555.0,
      "step": 22020
    },
    {
      "entropy": 0.9725808409973979,
      "epoch": 0.6310241610930496,
      "grad_norm": 6.5,
      "learning_rate": 3.933993023246093e-05,
      "loss": 0.9908555030822754,
      "mean_token_accuracy": 0.7858356811106205,
      "num_tokens": 43528507.0,
      "step": 22030
    },
    {
      "entropy": 0.8921185329556465,
      "epoch": 0.6313105996591382,
      "grad_norm": 9.1875,
      "learning_rate": 3.933058099550296e-05,
      "loss": 0.8680631637573242,
      "mean_token_accuracy": 0.7986000508069993,
      "num_tokens": 43548330.0,
      "step": 22040
    },
    {
      "entropy": 0.9197222735732794,
      "epoch": 0.6315970382252266,
      "grad_norm": 8.9375,
      "learning_rate": 3.932122877258685e-05,
      "loss": 0.8812421798706055,
      "mean_token_accuracy": 0.7980240326374769,
      "num_tokens": 43568310.0,
      "step": 22050
    },
    {
      "entropy": 0.902125951461494,
      "epoch": 0.6318834767913152,
      "grad_norm": 8.5,
      "learning_rate": 3.9311873565661245e-05,
      "loss": 0.8925121307373047,
      "mean_token_accuracy": 0.7870840545743704,
      "num_tokens": 43587730.0,
      "step": 22060
    },
    {
      "entropy": 0.9313517795875669,
      "epoch": 0.6321699153574037,
      "grad_norm": 9.25,
      "learning_rate": 3.930251537667542e-05,
      "loss": 0.9625690460205079,
      "mean_token_accuracy": 0.7717984311282635,
      "num_tokens": 43609123.0,
      "step": 22070
    },
    {
      "entropy": 0.8662886992096901,
      "epoch": 0.6324563539234923,
      "grad_norm": 9.875,
      "learning_rate": 3.9293154207579274e-05,
      "loss": 0.9740381240844727,
      "mean_token_accuracy": 0.7767984326928854,
      "num_tokens": 43629937.0,
      "step": 22080
    },
    {
      "entropy": 0.8445108029991388,
      "epoch": 0.6327427924895808,
      "grad_norm": 8.625,
      "learning_rate": 3.928379006032333e-05,
      "loss": 0.8383246421813965,
      "mean_token_accuracy": 0.8011367581784725,
      "num_tokens": 43649606.0,
      "step": 22090
    },
    {
      "entropy": 1.0122460559010507,
      "epoch": 0.6330292310556693,
      "grad_norm": 7.34375,
      "learning_rate": 3.927442293685872e-05,
      "loss": 1.0044979095458983,
      "mean_token_accuracy": 0.7732941288501024,
      "num_tokens": 43669662.0,
      "step": 22100
    },
    {
      "entropy": 0.8975360499694943,
      "epoch": 0.6333156696217579,
      "grad_norm": 7.28125,
      "learning_rate": 3.9265052839137204e-05,
      "loss": 0.9004294395446777,
      "mean_token_accuracy": 0.794927317649126,
      "num_tokens": 43688851.0,
      "step": 22110
    },
    {
      "entropy": 0.9065792698413133,
      "epoch": 0.6336021081878465,
      "grad_norm": 7.71875,
      "learning_rate": 3.9255679769111175e-05,
      "loss": 0.9747360229492188,
      "mean_token_accuracy": 0.7772964276373386,
      "num_tokens": 43708800.0,
      "step": 22120
    },
    {
      "entropy": 0.8763492200523615,
      "epoch": 0.6338885467539349,
      "grad_norm": 7.09375,
      "learning_rate": 3.924630372873361e-05,
      "loss": 0.8573235511779785,
      "mean_token_accuracy": 0.802193546667695,
      "num_tokens": 43728763.0,
      "step": 22130
    },
    {
      "entropy": 0.9529889117926359,
      "epoch": 0.6341749853200235,
      "grad_norm": 5.75,
      "learning_rate": 3.923692471995813e-05,
      "loss": 0.9217046737670899,
      "mean_token_accuracy": 0.7940693061798811,
      "num_tokens": 43748054.0,
      "step": 22140
    },
    {
      "entropy": 0.9323940431699157,
      "epoch": 0.634461423886112,
      "grad_norm": 6.875,
      "learning_rate": 3.922754274473898e-05,
      "loss": 0.8833812713623047,
      "mean_token_accuracy": 0.7922771837562322,
      "num_tokens": 43768540.0,
      "step": 22150
    },
    {
      "entropy": 0.8668142009526492,
      "epoch": 0.6347478624522006,
      "grad_norm": 7.8125,
      "learning_rate": 3.921815780503101e-05,
      "loss": 0.916175651550293,
      "mean_token_accuracy": 0.7913786351680756,
      "num_tokens": 43786998.0,
      "step": 22160
    },
    {
      "entropy": 0.9192729352042079,
      "epoch": 0.6350343010182891,
      "grad_norm": 10.0,
      "learning_rate": 3.920876990278968e-05,
      "loss": 0.9265934944152832,
      "mean_token_accuracy": 0.7872365403920412,
      "num_tokens": 43806645.0,
      "step": 22170
    },
    {
      "entropy": 0.889906695485115,
      "epoch": 0.6353207395843776,
      "grad_norm": 6.625,
      "learning_rate": 3.91993790399711e-05,
      "loss": 0.9652253150939941,
      "mean_token_accuracy": 0.7815795511007309,
      "num_tokens": 43826982.0,
      "step": 22180
    },
    {
      "entropy": 0.9756072624586523,
      "epoch": 0.6356071781504662,
      "grad_norm": 9.1875,
      "learning_rate": 3.9189985218531954e-05,
      "loss": 1.0104361534118653,
      "mean_token_accuracy": 0.7678293228149414,
      "num_tokens": 43847019.0,
      "step": 22190
    },
    {
      "entropy": 1.068906707689166,
      "epoch": 0.6358936167165548,
      "grad_norm": 11.8125,
      "learning_rate": 3.918058844042957e-05,
      "loss": 1.1275259971618652,
      "mean_token_accuracy": 0.753492145985365,
      "num_tokens": 43866807.0,
      "step": 22200
    },
    {
      "entropy": 0.8884169787168503,
      "epoch": 0.6361800552826432,
      "grad_norm": 7.09375,
      "learning_rate": 3.91711887076219e-05,
      "loss": 0.9018492698669434,
      "mean_token_accuracy": 0.7966490957885981,
      "num_tokens": 43886949.0,
      "step": 22210
    },
    {
      "entropy": 0.8576098520308733,
      "epoch": 0.6364664938487318,
      "grad_norm": 7.28125,
      "learning_rate": 3.916178602206747e-05,
      "loss": 0.8391011238098145,
      "mean_token_accuracy": 0.8052167434245348,
      "num_tokens": 43906116.0,
      "step": 22220
    },
    {
      "entropy": 0.959751558303833,
      "epoch": 0.6367529324148203,
      "grad_norm": 7.75,
      "learning_rate": 3.9152380385725475e-05,
      "loss": 0.9795754432678223,
      "mean_token_accuracy": 0.772762817516923,
      "num_tokens": 43928132.0,
      "step": 22230
    },
    {
      "entropy": 0.986600223556161,
      "epoch": 0.6370393709809089,
      "grad_norm": 9.5,
      "learning_rate": 3.914297180055569e-05,
      "loss": 1.048435115814209,
      "mean_token_accuracy": 0.7678979497402907,
      "num_tokens": 43948647.0,
      "step": 22240
    },
    {
      "entropy": 0.8919928649440407,
      "epoch": 0.6373258095469974,
      "grad_norm": 16.5,
      "learning_rate": 3.91335602685185e-05,
      "loss": 0.9076590538024902,
      "mean_token_accuracy": 0.7863856438547373,
      "num_tokens": 43969060.0,
      "step": 22250
    },
    {
      "entropy": 0.9068727388978004,
      "epoch": 0.6376122481130859,
      "grad_norm": 7.90625,
      "learning_rate": 3.912414579157494e-05,
      "loss": 0.8976054191589355,
      "mean_token_accuracy": 0.7875074785202741,
      "num_tokens": 43988125.0,
      "step": 22260
    },
    {
      "entropy": 0.9643635045737028,
      "epoch": 0.6378986866791745,
      "grad_norm": 7.0625,
      "learning_rate": 3.911472837168661e-05,
      "loss": 0.9923530578613281,
      "mean_token_accuracy": 0.7769703358411789,
      "num_tokens": 44009432.0,
      "step": 22270
    },
    {
      "entropy": 0.9225421917624772,
      "epoch": 0.638185125245263,
      "grad_norm": 7.25,
      "learning_rate": 3.910530801081578e-05,
      "loss": 0.8976520538330078,
      "mean_token_accuracy": 0.7898676216602325,
      "num_tokens": 44029652.0,
      "step": 22280
    },
    {
      "entropy": 1.0199335306882857,
      "epoch": 0.6384715638113515,
      "grad_norm": 7.6875,
      "learning_rate": 3.9095884710925265e-05,
      "loss": 1.0756010055541991,
      "mean_token_accuracy": 0.757290604710579,
      "num_tokens": 44050761.0,
      "step": 22290
    },
    {
      "entropy": 0.8560690857470036,
      "epoch": 0.6387580023774401,
      "grad_norm": 6.09375,
      "learning_rate": 3.908645847397856e-05,
      "loss": 0.8347340583801269,
      "mean_token_accuracy": 0.796586674451828,
      "num_tokens": 44071014.0,
      "step": 22300
    },
    {
      "entropy": 0.944689245801419,
      "epoch": 0.6390444409435286,
      "grad_norm": 6.5,
      "learning_rate": 3.9077029301939725e-05,
      "loss": 0.9123414993286133,
      "mean_token_accuracy": 0.791766332089901,
      "num_tokens": 44090815.0,
      "step": 22310
    },
    {
      "entropy": 0.8729799523949623,
      "epoch": 0.6393308795096172,
      "grad_norm": 8.3125,
      "learning_rate": 3.9067597196773455e-05,
      "loss": 0.9163877487182617,
      "mean_token_accuracy": 0.7888689719140529,
      "num_tokens": 44110656.0,
      "step": 22320
    },
    {
      "entropy": 0.9006530899554491,
      "epoch": 0.6396173180757058,
      "grad_norm": 7.3125,
      "learning_rate": 3.9058162160445036e-05,
      "loss": 0.9575582504272461,
      "mean_token_accuracy": 0.7837183397263289,
      "num_tokens": 44129545.0,
      "step": 22330
    },
    {
      "entropy": 0.8503710435703397,
      "epoch": 0.6399037566417942,
      "grad_norm": 7.09375,
      "learning_rate": 3.90487241949204e-05,
      "loss": 0.816905403137207,
      "mean_token_accuracy": 0.8006047587841749,
      "num_tokens": 44150330.0,
      "step": 22340
    },
    {
      "entropy": 0.9433940965682268,
      "epoch": 0.6401901952078828,
      "grad_norm": 8.0625,
      "learning_rate": 3.903928330216604e-05,
      "loss": 0.9354496955871582,
      "mean_token_accuracy": 0.78372464440763,
      "num_tokens": 44170369.0,
      "step": 22350
    },
    {
      "entropy": 0.8338274862617254,
      "epoch": 0.6404766337739713,
      "grad_norm": 8.4375,
      "learning_rate": 3.90298394841491e-05,
      "loss": 0.8598258018493652,
      "mean_token_accuracy": 0.79914346113801,
      "num_tokens": 44189857.0,
      "step": 22360
    },
    {
      "entropy": 0.8946568100713194,
      "epoch": 0.6407630723400599,
      "grad_norm": 10.5625,
      "learning_rate": 3.902039274283732e-05,
      "loss": 0.8722133636474609,
      "mean_token_accuracy": 0.794972014427185,
      "num_tokens": 44210535.0,
      "step": 22370
    },
    {
      "entropy": 0.8309621941298246,
      "epoch": 0.6410495109061484,
      "grad_norm": 8.5,
      "learning_rate": 3.9010943080199045e-05,
      "loss": 0.8612496376037597,
      "mean_token_accuracy": 0.7979394897818566,
      "num_tokens": 44230522.0,
      "step": 22380
    },
    {
      "entropy": 1.0275702426210045,
      "epoch": 0.6413359494722369,
      "grad_norm": 7.25,
      "learning_rate": 3.900149049820323e-05,
      "loss": 1.0629887580871582,
      "mean_token_accuracy": 0.7690297964960336,
      "num_tokens": 44249864.0,
      "step": 22390
    },
    {
      "entropy": 0.9767688054591417,
      "epoch": 0.6416223880383255,
      "grad_norm": 9.75,
      "learning_rate": 3.899203499881945e-05,
      "loss": 0.953999137878418,
      "mean_token_accuracy": 0.7919405065476894,
      "num_tokens": 44271131.0,
      "step": 22400
    },
    {
      "entropy": 0.9294959982857108,
      "epoch": 0.6419088266044141,
      "grad_norm": 8.9375,
      "learning_rate": 3.898257658401786e-05,
      "loss": 0.9641952514648438,
      "mean_token_accuracy": 0.7883687581866979,
      "num_tokens": 44291382.0,
      "step": 22410
    },
    {
      "entropy": 0.8459644271060824,
      "epoch": 0.6421952651705025,
      "grad_norm": 9.625,
      "learning_rate": 3.8973115255769265e-05,
      "loss": 0.819830322265625,
      "mean_token_accuracy": 0.8105485625565052,
      "num_tokens": 44310407.0,
      "step": 22420
    },
    {
      "entropy": 0.9576626213267445,
      "epoch": 0.6424817037365911,
      "grad_norm": 8.25,
      "learning_rate": 3.8963651016045034e-05,
      "loss": 0.9917973518371582,
      "mean_token_accuracy": 0.7719242431223392,
      "num_tokens": 44331077.0,
      "step": 22430
    },
    {
      "entropy": 0.9282437870278955,
      "epoch": 0.6427681423026796,
      "grad_norm": 9.8125,
      "learning_rate": 3.895418386681717e-05,
      "loss": 0.9387789726257324,
      "mean_token_accuracy": 0.7888801168650389,
      "num_tokens": 44350166.0,
      "step": 22440
    },
    {
      "entropy": 0.8456738986074924,
      "epoch": 0.6430545808687682,
      "grad_norm": 7.375,
      "learning_rate": 3.894471381005827e-05,
      "loss": 0.8855236053466797,
      "mean_token_accuracy": 0.7942288026213646,
      "num_tokens": 44369510.0,
      "step": 22450
    },
    {
      "entropy": 0.8561990622431039,
      "epoch": 0.6433410194348567,
      "grad_norm": 6.1875,
      "learning_rate": 3.893524084774153e-05,
      "loss": 0.8386794090270996,
      "mean_token_accuracy": 0.7894306804984808,
      "num_tokens": 44390167.0,
      "step": 22460
    },
    {
      "entropy": 0.8619652320630848,
      "epoch": 0.6436274580009452,
      "grad_norm": 6.78125,
      "learning_rate": 3.8925764981840776e-05,
      "loss": 0.8096957206726074,
      "mean_token_accuracy": 0.8049199424684048,
      "num_tokens": 44409600.0,
      "step": 22470
    },
    {
      "entropy": 0.9459259003400803,
      "epoch": 0.6439138965670338,
      "grad_norm": 14.0,
      "learning_rate": 3.8916286214330426e-05,
      "loss": 0.9588747978210449,
      "mean_token_accuracy": 0.7801275923848152,
      "num_tokens": 44428718.0,
      "step": 22480
    },
    {
      "entropy": 0.9236369654536247,
      "epoch": 0.6442003351331224,
      "grad_norm": 6.96875,
      "learning_rate": 3.89068045471855e-05,
      "loss": 0.8956018447875976,
      "mean_token_accuracy": 0.7848766662180424,
      "num_tokens": 44449264.0,
      "step": 22490
    },
    {
      "entropy": 0.8311104310676456,
      "epoch": 0.6444867736992108,
      "grad_norm": 9.0625,
      "learning_rate": 3.8897319982381605e-05,
      "loss": 0.8809353828430175,
      "mean_token_accuracy": 0.7970906995236874,
      "num_tokens": 44469169.0,
      "step": 22500
    },
    {
      "entropy": 0.9523693701252341,
      "epoch": 0.6447732122652994,
      "grad_norm": 7.0625,
      "learning_rate": 3.8887832521895e-05,
      "loss": 0.9376598358154297,
      "mean_token_accuracy": 0.7859144486486912,
      "num_tokens": 44489928.0,
      "step": 22510
    },
    {
      "entropy": 0.9099814087152481,
      "epoch": 0.6450596508313879,
      "grad_norm": 6.9375,
      "learning_rate": 3.88783421677025e-05,
      "loss": 0.9119604110717774,
      "mean_token_accuracy": 0.787454879656434,
      "num_tokens": 44509930.0,
      "step": 22520
    },
    {
      "entropy": 0.8894009451381862,
      "epoch": 0.6453460893974765,
      "grad_norm": 6.75,
      "learning_rate": 3.886884892178154e-05,
      "loss": 0.956300163269043,
      "mean_token_accuracy": 0.7796136941760778,
      "num_tokens": 44529759.0,
      "step": 22530
    },
    {
      "entropy": 0.9388137497007847,
      "epoch": 0.645632527963565,
      "grad_norm": 8.375,
      "learning_rate": 3.885935278611016e-05,
      "loss": 0.9333770751953125,
      "mean_token_accuracy": 0.7783537801355124,
      "num_tokens": 44549383.0,
      "step": 22540
    },
    {
      "entropy": 0.9887114821001888,
      "epoch": 0.6459189665296535,
      "grad_norm": 7.78125,
      "learning_rate": 3.884985376266701e-05,
      "loss": 0.9868999481201172,
      "mean_token_accuracy": 0.7881288662552833,
      "num_tokens": 44569365.0,
      "step": 22550
    },
    {
      "entropy": 0.919053116440773,
      "epoch": 0.6462054050957421,
      "grad_norm": 6.40625,
      "learning_rate": 3.884035185343131e-05,
      "loss": 0.9126174926757813,
      "mean_token_accuracy": 0.7788828633725643,
      "num_tokens": 44587327.0,
      "step": 22560
    },
    {
      "entropy": 1.0299420004710556,
      "epoch": 0.6464918436618307,
      "grad_norm": 9.125,
      "learning_rate": 3.883084706038292e-05,
      "loss": 1.0252874374389649,
      "mean_token_accuracy": 0.7592982098460197,
      "num_tokens": 44606861.0,
      "step": 22570
    },
    {
      "entropy": 0.9452397517859936,
      "epoch": 0.6467782822279192,
      "grad_norm": 10.375,
      "learning_rate": 3.8821339385502286e-05,
      "loss": 0.9575634002685547,
      "mean_token_accuracy": 0.7732752721756697,
      "num_tokens": 44628004.0,
      "step": 22580
    },
    {
      "entropy": 0.9163175793364644,
      "epoch": 0.6470647207940077,
      "grad_norm": 8.5,
      "learning_rate": 3.8811828830770447e-05,
      "loss": 0.8650115966796875,
      "mean_token_accuracy": 0.7982464917004108,
      "num_tokens": 44647277.0,
      "step": 22590
    },
    {
      "entropy": 0.8501980051398277,
      "epoch": 0.6473511593600962,
      "grad_norm": 7.4375,
      "learning_rate": 3.880231539816905e-05,
      "loss": 0.8509060859680175,
      "mean_token_accuracy": 0.7979789350181818,
      "num_tokens": 44666789.0,
      "step": 22600
    },
    {
      "entropy": 0.8882537086494267,
      "epoch": 0.6476375979261848,
      "grad_norm": 7.28125,
      "learning_rate": 3.8792799089680335e-05,
      "loss": 0.8898092269897461,
      "mean_token_accuracy": 0.7982908442616463,
      "num_tokens": 44685735.0,
      "step": 22610
    },
    {
      "entropy": 0.9092659370973706,
      "epoch": 0.6479240364922734,
      "grad_norm": 10.5,
      "learning_rate": 3.878327990728714e-05,
      "loss": 0.9800436973571778,
      "mean_token_accuracy": 0.7881735842674971,
      "num_tokens": 44707240.0,
      "step": 22620
    },
    {
      "entropy": 0.7355032010003925,
      "epoch": 0.6482104750583618,
      "grad_norm": 8.0625,
      "learning_rate": 3.8773757852972925e-05,
      "loss": 0.7902781963348389,
      "mean_token_accuracy": 0.8144808396697044,
      "num_tokens": 44725351.0,
      "step": 22630
    },
    {
      "entropy": 0.9388284519314766,
      "epoch": 0.6484969136244504,
      "grad_norm": 9.25,
      "learning_rate": 3.876423292872171e-05,
      "loss": 1.0065260887145997,
      "mean_token_accuracy": 0.7760294701904058,
      "num_tokens": 44745660.0,
      "step": 22640
    },
    {
      "entropy": 0.9486516360193491,
      "epoch": 0.648783352190539,
      "grad_norm": 9.4375,
      "learning_rate": 3.8754705136518154e-05,
      "loss": 0.9543668746948242,
      "mean_token_accuracy": 0.7778625778853894,
      "num_tokens": 44766653.0,
      "step": 22650
    },
    {
      "entropy": 0.8811884829774499,
      "epoch": 0.6490697907566275,
      "grad_norm": 5.90625,
      "learning_rate": 3.8745174478347476e-05,
      "loss": 0.8522655487060546,
      "mean_token_accuracy": 0.8103316944092512,
      "num_tokens": 44786618.0,
      "step": 22660
    },
    {
      "entropy": 0.8679312387481332,
      "epoch": 0.649356229322716,
      "grad_norm": 8.3125,
      "learning_rate": 3.873564095619552e-05,
      "loss": 0.8800905227661133,
      "mean_token_accuracy": 0.7946135979145765,
      "num_tokens": 44807148.0,
      "step": 22670
    },
    {
      "entropy": 0.9480700666084886,
      "epoch": 0.6496426678888045,
      "grad_norm": 11.4375,
      "learning_rate": 3.872610457204872e-05,
      "loss": 0.9936513900756836,
      "mean_token_accuracy": 0.770840159803629,
      "num_tokens": 44827454.0,
      "step": 22680
    },
    {
      "entropy": 0.8437151649966836,
      "epoch": 0.6499291064548931,
      "grad_norm": 7.625,
      "learning_rate": 3.871656532789409e-05,
      "loss": 0.8436770439147949,
      "mean_token_accuracy": 0.7978602487593889,
      "num_tokens": 44846120.0,
      "step": 22690
    },
    {
      "entropy": 0.9959442108869553,
      "epoch": 0.6502155450209817,
      "grad_norm": 5.90625,
      "learning_rate": 3.8707023225719254e-05,
      "loss": 0.9619386672973633,
      "mean_token_accuracy": 0.792440676316619,
      "num_tokens": 44866513.0,
      "step": 22700
    },
    {
      "entropy": 1.0226761890575289,
      "epoch": 0.6505019835870701,
      "grad_norm": 9.625,
      "learning_rate": 3.869747826751244e-05,
      "loss": 1.027959442138672,
      "mean_token_accuracy": 0.7675781823694706,
      "num_tokens": 44885237.0,
      "step": 22710
    },
    {
      "entropy": 0.9059097156859934,
      "epoch": 0.6507884221531587,
      "grad_norm": 10.75,
      "learning_rate": 3.868793045526245e-05,
      "loss": 0.9633118629455566,
      "mean_token_accuracy": 0.7799510974436998,
      "num_tokens": 44905072.0,
      "step": 22720
    },
    {
      "entropy": 0.9725006885826588,
      "epoch": 0.6510748607192473,
      "grad_norm": 6.65625,
      "learning_rate": 3.86783797909587e-05,
      "loss": 0.9461207389831543,
      "mean_token_accuracy": 0.7843978337943553,
      "num_tokens": 44923643.0,
      "step": 22730
    },
    {
      "entropy": 0.9747046899050474,
      "epoch": 0.6513612992853358,
      "grad_norm": 5.71875,
      "learning_rate": 3.8668826276591185e-05,
      "loss": 0.9596343040466309,
      "mean_token_accuracy": 0.7839418992400169,
      "num_tokens": 44944385.0,
      "step": 22740
    },
    {
      "entropy": 0.8609721411019564,
      "epoch": 0.6516477378514243,
      "grad_norm": 10.1875,
      "learning_rate": 3.865926991415051e-05,
      "loss": 0.8540242195129395,
      "mean_token_accuracy": 0.7934511575847865,
      "num_tokens": 44963887.0,
      "step": 22750
    },
    {
      "entropy": 0.9093576775863766,
      "epoch": 0.6519341764175128,
      "grad_norm": 7.28125,
      "learning_rate": 3.864971070562785e-05,
      "loss": 0.9144698143005371,
      "mean_token_accuracy": 0.7814730063080788,
      "num_tokens": 44983330.0,
      "step": 22760
    },
    {
      "entropy": 0.9059786170721054,
      "epoch": 0.6522206149836014,
      "grad_norm": 7.28125,
      "learning_rate": 3.8640148653014996e-05,
      "loss": 0.9081611633300781,
      "mean_token_accuracy": 0.7908361550420523,
      "num_tokens": 45002457.0,
      "step": 22770
    },
    {
      "entropy": 0.8813687536865473,
      "epoch": 0.65250705354969,
      "grad_norm": 7.875,
      "learning_rate": 3.863058375830433e-05,
      "loss": 0.9298584938049317,
      "mean_token_accuracy": 0.7796319764107466,
      "num_tokens": 45022772.0,
      "step": 22780
    },
    {
      "entropy": 0.7959211783483624,
      "epoch": 0.6527934921157784,
      "grad_norm": 8.9375,
      "learning_rate": 3.8621016023488805e-05,
      "loss": 0.7696258068084717,
      "mean_token_accuracy": 0.8180242717266083,
      "num_tokens": 45042825.0,
      "step": 22790
    },
    {
      "entropy": 0.8323979537934065,
      "epoch": 0.653079930681867,
      "grad_norm": 7.6875,
      "learning_rate": 3.861144545056199e-05,
      "loss": 0.8022558212280273,
      "mean_token_accuracy": 0.8124545853585005,
      "num_tokens": 45062979.0,
      "step": 22800
    },
    {
      "entropy": 0.7982686029747128,
      "epoch": 0.6533663692479555,
      "grad_norm": 8.375,
      "learning_rate": 3.860187204151803e-05,
      "loss": 0.8256909370422363,
      "mean_token_accuracy": 0.8086091376841068,
      "num_tokens": 45082377.0,
      "step": 22810
    },
    {
      "entropy": 0.8759147889912129,
      "epoch": 0.6536528078140441,
      "grad_norm": 7.90625,
      "learning_rate": 3.8592295798351666e-05,
      "loss": 0.9384151458740234,
      "mean_token_accuracy": 0.783957288786769,
      "num_tokens": 45102259.0,
      "step": 22820
    },
    {
      "entropy": 1.0228074008598924,
      "epoch": 0.6539392463801326,
      "grad_norm": 7.03125,
      "learning_rate": 3.858271672305824e-05,
      "loss": 1.0605764389038086,
      "mean_token_accuracy": 0.7615953620523215,
      "num_tokens": 45121661.0,
      "step": 22830
    },
    {
      "entropy": 0.812204186618328,
      "epoch": 0.6542256849462211,
      "grad_norm": 8.625,
      "learning_rate": 3.857313481763365e-05,
      "loss": 0.8009726524353027,
      "mean_token_accuracy": 0.8216852936893702,
      "num_tokens": 45140585.0,
      "step": 22840
    },
    {
      "entropy": 0.7916817590594292,
      "epoch": 0.6545121235123097,
      "grad_norm": 7.46875,
      "learning_rate": 3.856355008407443e-05,
      "loss": 0.8606937408447266,
      "mean_token_accuracy": 0.8061037711799145,
      "num_tokens": 45160519.0,
      "step": 22850
    },
    {
      "entropy": 0.9227229975163936,
      "epoch": 0.6547985620783983,
      "grad_norm": 7.46875,
      "learning_rate": 3.8553962524377674e-05,
      "loss": 0.8921630859375,
      "mean_token_accuracy": 0.7862365450710058,
      "num_tokens": 45179569.0,
      "step": 22860
    },
    {
      "entropy": 0.9050315000116825,
      "epoch": 0.6550850006444868,
      "grad_norm": 7.3125,
      "learning_rate": 3.854437214054107e-05,
      "loss": 0.954566478729248,
      "mean_token_accuracy": 0.7868342835456132,
      "num_tokens": 45198814.0,
      "step": 22870
    },
    {
      "entropy": 0.9461556099355221,
      "epoch": 0.6553714392105753,
      "grad_norm": 9.0,
      "learning_rate": 3.8534778934562895e-05,
      "loss": 0.9370518684387207,
      "mean_token_accuracy": 0.7834421284496784,
      "num_tokens": 45218996.0,
      "step": 22880
    },
    {
      "entropy": 0.9407024182379246,
      "epoch": 0.6556578777766638,
      "grad_norm": 11.3125,
      "learning_rate": 3.852518290844202e-05,
      "loss": 0.9654740333557129,
      "mean_token_accuracy": 0.7831741657108069,
      "num_tokens": 45237838.0,
      "step": 22890
    },
    {
      "entropy": 0.8815357193350792,
      "epoch": 0.6559443163427524,
      "grad_norm": 7.625,
      "learning_rate": 3.851558406417789e-05,
      "loss": 0.8830876350402832,
      "mean_token_accuracy": 0.7963199000805616,
      "num_tokens": 45256805.0,
      "step": 22900
    },
    {
      "entropy": 0.8782296074554324,
      "epoch": 0.656230754908841,
      "grad_norm": 11.0,
      "learning_rate": 3.8505982403770554e-05,
      "loss": 0.9437088966369629,
      "mean_token_accuracy": 0.7836847331374883,
      "num_tokens": 45275512.0,
      "step": 22910
    },
    {
      "entropy": 1.0104555912315845,
      "epoch": 0.6565171934749294,
      "grad_norm": 8.9375,
      "learning_rate": 3.8496377929220625e-05,
      "loss": 1.0315958976745605,
      "mean_token_accuracy": 0.76339814029634,
      "num_tokens": 45295084.0,
      "step": 22920
    },
    {
      "entropy": 0.8609929211437702,
      "epoch": 0.656803632041018,
      "grad_norm": 6.0625,
      "learning_rate": 3.848677064252935e-05,
      "loss": 0.8577310562133789,
      "mean_token_accuracy": 0.8015160631388426,
      "num_tokens": 45314186.0,
      "step": 22930
    },
    {
      "entropy": 0.9470567163079977,
      "epoch": 0.6570900706071066,
      "grad_norm": 11.0625,
      "learning_rate": 3.84771605456985e-05,
      "loss": 0.9081389427185058,
      "mean_token_accuracy": 0.7880242943763733,
      "num_tokens": 45335275.0,
      "step": 22940
    },
    {
      "entropy": 0.8757361777126789,
      "epoch": 0.6573765091731951,
      "grad_norm": 6.46875,
      "learning_rate": 3.8467547640730456e-05,
      "loss": 0.8774962425231934,
      "mean_token_accuracy": 0.7940940015017987,
      "num_tokens": 45355456.0,
      "step": 22950
    },
    {
      "entropy": 0.8433198031038046,
      "epoch": 0.6576629477392836,
      "grad_norm": 11.875,
      "learning_rate": 3.8457931929628215e-05,
      "loss": 0.8380443572998046,
      "mean_token_accuracy": 0.800097682327032,
      "num_tokens": 45375332.0,
      "step": 22960
    },
    {
      "entropy": 0.9152161931619048,
      "epoch": 0.6579493863053721,
      "grad_norm": 7.46875,
      "learning_rate": 3.844831341439531e-05,
      "loss": 0.8949036598205566,
      "mean_token_accuracy": 0.7754564691334963,
      "num_tokens": 45395567.0,
      "step": 22970
    },
    {
      "entropy": 0.7952107956632972,
      "epoch": 0.6582358248714607,
      "grad_norm": 7.375,
      "learning_rate": 3.843869209703589e-05,
      "loss": 0.7261805057525634,
      "mean_token_accuracy": 0.8109205737709999,
      "num_tokens": 45414157.0,
      "step": 22980
    },
    {
      "entropy": 0.8574653865769506,
      "epoch": 0.6585222634375493,
      "grad_norm": 7.65625,
      "learning_rate": 3.842906797955468e-05,
      "loss": 0.8881130218505859,
      "mean_token_accuracy": 0.7971735812723637,
      "num_tokens": 45432740.0,
      "step": 22990
    },
    {
      "entropy": 0.873758516088128,
      "epoch": 0.6588087020036377,
      "grad_norm": 6.25,
      "learning_rate": 3.841944106395698e-05,
      "loss": 0.8962624549865723,
      "mean_token_accuracy": 0.7915884006768465,
      "num_tokens": 45451597.0,
      "step": 23000
    },
    {
      "entropy": 0.8562300365418196,
      "epoch": 0.6590951405697263,
      "grad_norm": 8.4375,
      "learning_rate": 3.840981135224868e-05,
      "loss": 0.8257713317871094,
      "mean_token_accuracy": 0.8008463565260172,
      "num_tokens": 45471432.0,
      "step": 23010
    },
    {
      "entropy": 0.8971815157681704,
      "epoch": 0.6593815791358149,
      "grad_norm": 9.6875,
      "learning_rate": 3.840017884643626e-05,
      "loss": 0.94833402633667,
      "mean_token_accuracy": 0.7707555256783962,
      "num_tokens": 45490699.0,
      "step": 23020
    },
    {
      "entropy": 0.8889431161805987,
      "epoch": 0.6596680177019034,
      "grad_norm": 8.25,
      "learning_rate": 3.839054354852677e-05,
      "loss": 0.9022657394409179,
      "mean_token_accuracy": 0.7944431319832802,
      "num_tokens": 45509939.0,
      "step": 23030
    },
    {
      "entropy": 0.9908782538026571,
      "epoch": 0.6599544562679919,
      "grad_norm": 9.5,
      "learning_rate": 3.838090546052785e-05,
      "loss": 1.0272876739501953,
      "mean_token_accuracy": 0.7685921404510736,
      "num_tokens": 45530238.0,
      "step": 23040
    },
    {
      "entropy": 0.9399600770324469,
      "epoch": 0.6602408948340804,
      "grad_norm": 8.6875,
      "learning_rate": 3.837126458444772e-05,
      "loss": 1.0116050720214844,
      "mean_token_accuracy": 0.7777301128953695,
      "num_tokens": 45550492.0,
      "step": 23050
    },
    {
      "entropy": 0.8885143099352717,
      "epoch": 0.660527333400169,
      "grad_norm": 8.5,
      "learning_rate": 3.836162092229516e-05,
      "loss": 0.8525900840759277,
      "mean_token_accuracy": 0.7940516788512468,
      "num_tokens": 45569578.0,
      "step": 23060
    },
    {
      "entropy": 0.857690074108541,
      "epoch": 0.6608137719662576,
      "grad_norm": 6.625,
      "learning_rate": 3.835197447607957e-05,
      "loss": 0.8312179565429687,
      "mean_token_accuracy": 0.8000508733093739,
      "num_tokens": 45589321.0,
      "step": 23070
    },
    {
      "entropy": 0.7944382214918733,
      "epoch": 0.661100210532346,
      "grad_norm": 10.5,
      "learning_rate": 3.8342325247810895e-05,
      "loss": 0.818697452545166,
      "mean_token_accuracy": 0.8069204747676849,
      "num_tokens": 45609098.0,
      "step": 23080
    },
    {
      "entropy": 0.8796201297082007,
      "epoch": 0.6613866490984346,
      "grad_norm": 8.875,
      "learning_rate": 3.833267323949969e-05,
      "loss": 0.9343830108642578,
      "mean_token_accuracy": 0.7889334291219712,
      "num_tokens": 45628822.0,
      "step": 23090
    },
    {
      "entropy": 0.9392872717231512,
      "epoch": 0.6616730876645232,
      "grad_norm": 7.40625,
      "learning_rate": 3.832301845315706e-05,
      "loss": 0.9675754547119141,
      "mean_token_accuracy": 0.7764622312039137,
      "num_tokens": 45648857.0,
      "step": 23100
    },
    {
      "entropy": 0.9299656821414828,
      "epoch": 0.6619595262306117,
      "grad_norm": 6.1875,
      "learning_rate": 3.831336089079471e-05,
      "loss": 0.9380331039428711,
      "mean_token_accuracy": 0.7860646154731512,
      "num_tokens": 45668057.0,
      "step": 23110
    },
    {
      "entropy": 0.90357981082052,
      "epoch": 0.6622459647967003,
      "grad_norm": 9.6875,
      "learning_rate": 3.8303700554424905e-05,
      "loss": 0.9348695755004883,
      "mean_token_accuracy": 0.788814963772893,
      "num_tokens": 45688533.0,
      "step": 23120
    },
    {
      "entropy": 0.9116713162511587,
      "epoch": 0.6625324033627887,
      "grad_norm": 6.125,
      "learning_rate": 3.8294037446060504e-05,
      "loss": 0.8847149848937989,
      "mean_token_accuracy": 0.7953954737633466,
      "num_tokens": 45708358.0,
      "step": 23130
    },
    {
      "entropy": 0.893862858787179,
      "epoch": 0.6628188419288773,
      "grad_norm": 7.0,
      "learning_rate": 3.828437156771495e-05,
      "loss": 0.8806056976318359,
      "mean_token_accuracy": 0.7973021131008864,
      "num_tokens": 45728216.0,
      "step": 23140
    },
    {
      "entropy": 0.8641118073835969,
      "epoch": 0.6631052804949659,
      "grad_norm": 7.5,
      "learning_rate": 3.827470292140223e-05,
      "loss": 0.8210006713867187,
      "mean_token_accuracy": 0.8014064136892557,
      "num_tokens": 45746818.0,
      "step": 23150
    },
    {
      "entropy": 0.9554972615092993,
      "epoch": 0.6633917190610544,
      "grad_norm": 11.0,
      "learning_rate": 3.826503150913695e-05,
      "loss": 1.0344480514526366,
      "mean_token_accuracy": 0.7642830304801465,
      "num_tokens": 45767389.0,
      "step": 23160
    },
    {
      "entropy": 0.9017923820763827,
      "epoch": 0.6636781576271429,
      "grad_norm": 5.9375,
      "learning_rate": 3.825535733293425e-05,
      "loss": 0.9085247039794921,
      "mean_token_accuracy": 0.7912683442234993,
      "num_tokens": 45786200.0,
      "step": 23170
    },
    {
      "entropy": 0.9547636365517974,
      "epoch": 0.6639645961932314,
      "grad_norm": 7.78125,
      "learning_rate": 3.824568039480988e-05,
      "loss": 0.9796084403991699,
      "mean_token_accuracy": 0.7788782030344009,
      "num_tokens": 45805377.0,
      "step": 23180
    },
    {
      "entropy": 0.9325101274996996,
      "epoch": 0.66425103475932,
      "grad_norm": 8.0,
      "learning_rate": 3.8236000696780156e-05,
      "loss": 0.9932963371276855,
      "mean_token_accuracy": 0.7687444686889648,
      "num_tokens": 45825726.0,
      "step": 23190
    },
    {
      "entropy": 0.8763208875432611,
      "epoch": 0.6645374733254086,
      "grad_norm": 8.8125,
      "learning_rate": 3.8226318240861956e-05,
      "loss": 0.8843619346618652,
      "mean_token_accuracy": 0.794028839096427,
      "num_tokens": 45845252.0,
      "step": 23200
    },
    {
      "entropy": 0.7843074448406696,
      "epoch": 0.664823911891497,
      "grad_norm": 7.6875,
      "learning_rate": 3.821663302907275e-05,
      "loss": 0.7874396800994873,
      "mean_token_accuracy": 0.8177121058106422,
      "num_tokens": 45863676.0,
      "step": 23210
    },
    {
      "entropy": 0.9136452928185463,
      "epoch": 0.6651103504575856,
      "grad_norm": 8.5625,
      "learning_rate": 3.820694506343057e-05,
      "loss": 0.9183270454406738,
      "mean_token_accuracy": 0.8006648808717728,
      "num_tokens": 45883923.0,
      "step": 23220
    },
    {
      "entropy": 0.8632227946072817,
      "epoch": 0.6653967890236742,
      "grad_norm": 9.25,
      "learning_rate": 3.819725434595403e-05,
      "loss": 0.8685300827026368,
      "mean_token_accuracy": 0.8010946236550808,
      "num_tokens": 45902950.0,
      "step": 23230
    },
    {
      "entropy": 0.9303806554526091,
      "epoch": 0.6656832275897627,
      "grad_norm": 8.0,
      "learning_rate": 3.818756087866231e-05,
      "loss": 0.9483536720275879,
      "mean_token_accuracy": 0.7845831368118524,
      "num_tokens": 45923386.0,
      "step": 23240
    },
    {
      "entropy": 0.8897776961326599,
      "epoch": 0.6659696661558512,
      "grad_norm": 7.9375,
      "learning_rate": 3.8177864663575166e-05,
      "loss": 0.8765877723693848,
      "mean_token_accuracy": 0.7978586934506893,
      "num_tokens": 45943512.0,
      "step": 23250
    },
    {
      "entropy": 0.8420126687735319,
      "epoch": 0.6662561047219397,
      "grad_norm": 6.1875,
      "learning_rate": 3.8168165702712935e-05,
      "loss": 0.7989052772521973,
      "mean_token_accuracy": 0.8130182828754187,
      "num_tokens": 45963650.0,
      "step": 23260
    },
    {
      "entropy": 0.9216004284098744,
      "epoch": 0.6665425432880283,
      "grad_norm": 7.25,
      "learning_rate": 3.8158463998096506e-05,
      "loss": 0.9407343864440918,
      "mean_token_accuracy": 0.7956110060214996,
      "num_tokens": 45982428.0,
      "step": 23270
    },
    {
      "entropy": 0.9197284761816263,
      "epoch": 0.6668289818541169,
      "grad_norm": 6.375,
      "learning_rate": 3.814875955174736e-05,
      "loss": 0.9001524925231934,
      "mean_token_accuracy": 0.7977133721113205,
      "num_tokens": 46002699.0,
      "step": 23280
    },
    {
      "entropy": 0.8211668686941266,
      "epoch": 0.6671154204202053,
      "grad_norm": 7.40625,
      "learning_rate": 3.8139052365687546e-05,
      "loss": 0.8624239921569824,
      "mean_token_accuracy": 0.799767604470253,
      "num_tokens": 46022581.0,
      "step": 23290
    },
    {
      "entropy": 0.8306294087320566,
      "epoch": 0.6674018589862939,
      "grad_norm": 6.0,
      "learning_rate": 3.812934244193967e-05,
      "loss": 0.8160804748535156,
      "mean_token_accuracy": 0.8062486603856087,
      "num_tokens": 46042860.0,
      "step": 23300
    },
    {
      "entropy": 0.855907554551959,
      "epoch": 0.6676882975523825,
      "grad_norm": 8.0625,
      "learning_rate": 3.811962978252692e-05,
      "loss": 0.8971040725708008,
      "mean_token_accuracy": 0.7978474866598845,
      "num_tokens": 46062892.0,
      "step": 23310
    },
    {
      "entropy": 0.9086446503177286,
      "epoch": 0.667974736118471,
      "grad_norm": 8.0,
      "learning_rate": 3.810991438947306e-05,
      "loss": 0.9208392143249512,
      "mean_token_accuracy": 0.7840793844312429,
      "num_tokens": 46082210.0,
      "step": 23320
    },
    {
      "entropy": 0.8421589640900493,
      "epoch": 0.6682611746845595,
      "grad_norm": 6.6875,
      "learning_rate": 3.81001962648024e-05,
      "loss": 0.7965210914611817,
      "mean_token_accuracy": 0.8150952901691199,
      "num_tokens": 46100920.0,
      "step": 23330
    },
    {
      "entropy": 0.8577923273667694,
      "epoch": 0.668547613250648,
      "grad_norm": 9.125,
      "learning_rate": 3.809047541053985e-05,
      "loss": 0.9250428199768066,
      "mean_token_accuracy": 0.7848080772906542,
      "num_tokens": 46119557.0,
      "step": 23340
    },
    {
      "entropy": 0.8931930897757411,
      "epoch": 0.6688340518167366,
      "grad_norm": 10.9375,
      "learning_rate": 3.808075182871086e-05,
      "loss": 0.8762700080871582,
      "mean_token_accuracy": 0.7881566148251296,
      "num_tokens": 46139611.0,
      "step": 23350
    },
    {
      "entropy": 0.9251813896000385,
      "epoch": 0.6691204903828252,
      "grad_norm": 6.21875,
      "learning_rate": 3.807102552134148e-05,
      "loss": 0.8911492347717285,
      "mean_token_accuracy": 0.7905508443713188,
      "num_tokens": 46160814.0,
      "step": 23360
    },
    {
      "entropy": 0.8740447130054235,
      "epoch": 0.6694069289489136,
      "grad_norm": 11.1875,
      "learning_rate": 3.8061296490458284e-05,
      "loss": 0.9040107727050781,
      "mean_token_accuracy": 0.7891867198050022,
      "num_tokens": 46180837.0,
      "step": 23370
    },
    {
      "entropy": 0.8201585682108998,
      "epoch": 0.6696933675150022,
      "grad_norm": 6.71875,
      "learning_rate": 3.805156473808845e-05,
      "loss": 0.8851461410522461,
      "mean_token_accuracy": 0.7935717169195413,
      "num_tokens": 46200104.0,
      "step": 23380
    },
    {
      "entropy": 0.9080373041331768,
      "epoch": 0.6699798060810908,
      "grad_norm": 10.0,
      "learning_rate": 3.8041830266259726e-05,
      "loss": 0.9129886627197266,
      "mean_token_accuracy": 0.7867343708872795,
      "num_tokens": 46220555.0,
      "step": 23390
    },
    {
      "entropy": 0.9192170156165957,
      "epoch": 0.6702662446471793,
      "grad_norm": 6.15625,
      "learning_rate": 3.803209307700038e-05,
      "loss": 0.8875781059265136,
      "mean_token_accuracy": 0.7926689025014639,
      "num_tokens": 46239908.0,
      "step": 23400
    },
    {
      "entropy": 0.9543895147740841,
      "epoch": 0.6705526832132679,
      "grad_norm": 6.59375,
      "learning_rate": 3.802235317233932e-05,
      "loss": 0.90999755859375,
      "mean_token_accuracy": 0.7824549444019795,
      "num_tokens": 46260337.0,
      "step": 23410
    },
    {
      "entropy": 0.8238056251779199,
      "epoch": 0.6708391217793563,
      "grad_norm": 6.6875,
      "learning_rate": 3.801261055430595e-05,
      "loss": 0.8561161994934082,
      "mean_token_accuracy": 0.7915153440088034,
      "num_tokens": 46279233.0,
      "step": 23420
    },
    {
      "entropy": 0.8776079719886184,
      "epoch": 0.6711255603454449,
      "grad_norm": 5.53125,
      "learning_rate": 3.8002865224930276e-05,
      "loss": 0.8759838104248047,
      "mean_token_accuracy": 0.7970323484390974,
      "num_tokens": 46300202.0,
      "step": 23430
    },
    {
      "entropy": 0.966572598926723,
      "epoch": 0.6714119989115335,
      "grad_norm": 7.65625,
      "learning_rate": 3.7993117186242854e-05,
      "loss": 0.9568003654479981,
      "mean_token_accuracy": 0.7862712480127811,
      "num_tokens": 46320009.0,
      "step": 23440
    },
    {
      "entropy": 0.7571846072562038,
      "epoch": 0.671698437477622,
      "grad_norm": 6.6875,
      "learning_rate": 3.7983366440274824e-05,
      "loss": 0.7800245761871338,
      "mean_token_accuracy": 0.812897089868784,
      "num_tokens": 46340095.0,
      "step": 23450
    },
    {
      "entropy": 0.8918323492631316,
      "epoch": 0.6719848760437105,
      "grad_norm": 7.5,
      "learning_rate": 3.797361298905787e-05,
      "loss": 0.9016538619995117,
      "mean_token_accuracy": 0.7905817065387964,
      "num_tokens": 46361264.0,
      "step": 23460
    },
    {
      "entropy": 0.8704618336632848,
      "epoch": 0.672271314609799,
      "grad_norm": 7.5,
      "learning_rate": 3.7963856834624244e-05,
      "loss": 0.8952795028686523,
      "mean_token_accuracy": 0.7963431309908628,
      "num_tokens": 46380785.0,
      "step": 23470
    },
    {
      "entropy": 0.7960759287700057,
      "epoch": 0.6725577531758876,
      "grad_norm": 6.96875,
      "learning_rate": 3.7954097979006774e-05,
      "loss": 0.7884669780731202,
      "mean_token_accuracy": 0.8157178360968829,
      "num_tokens": 46399795.0,
      "step": 23480
    },
    {
      "entropy": 0.9695372845977545,
      "epoch": 0.6728441917419762,
      "grad_norm": 6.71875,
      "learning_rate": 3.7944336424238844e-05,
      "loss": 1.0191239356994628,
      "mean_token_accuracy": 0.7655313782393932,
      "num_tokens": 46420465.0,
      "step": 23490
    },
    {
      "entropy": 0.9198993441648782,
      "epoch": 0.6731306303080646,
      "grad_norm": 8.75,
      "learning_rate": 3.793457217235438e-05,
      "loss": 0.8935195922851562,
      "mean_token_accuracy": 0.7888022102415562,
      "num_tokens": 46440636.0,
      "step": 23500
    },
    {
      "entropy": 0.8171261554583907,
      "epoch": 0.6734170688741532,
      "grad_norm": 7.21875,
      "learning_rate": 3.792480522538789e-05,
      "loss": 0.7994399547576905,
      "mean_token_accuracy": 0.8140453927218914,
      "num_tokens": 46461241.0,
      "step": 23510
    },
    {
      "entropy": 0.9676499631255865,
      "epoch": 0.6737035074402418,
      "grad_norm": 8.4375,
      "learning_rate": 3.7915035585374456e-05,
      "loss": 1.0052270889282227,
      "mean_token_accuracy": 0.7816280674189329,
      "num_tokens": 46480362.0,
      "step": 23520
    },
    {
      "entropy": 0.8269085709005595,
      "epoch": 0.6739899460063303,
      "grad_norm": 8.625,
      "learning_rate": 3.79052632543497e-05,
      "loss": 0.842251205444336,
      "mean_token_accuracy": 0.8143334649503231,
      "num_tokens": 46499814.0,
      "step": 23530
    },
    {
      "entropy": 0.9777561783790588,
      "epoch": 0.6742763845724188,
      "grad_norm": 10.0625,
      "learning_rate": 3.789548823434981e-05,
      "loss": 1.0528251647949218,
      "mean_token_accuracy": 0.7656882911920547,
      "num_tokens": 46519519.0,
      "step": 23540
    },
    {
      "entropy": 0.7869767639786005,
      "epoch": 0.6745628231385074,
      "grad_norm": 7.25,
      "learning_rate": 3.788571052741153e-05,
      "loss": 0.7685704231262207,
      "mean_token_accuracy": 0.8163337204605341,
      "num_tokens": 46538019.0,
      "step": 23550
    },
    {
      "entropy": 0.8506517441011965,
      "epoch": 0.6748492617045959,
      "grad_norm": 7.1875,
      "learning_rate": 3.787593013557218e-05,
      "loss": 0.8097248077392578,
      "mean_token_accuracy": 0.8070813931524754,
      "num_tokens": 46557453.0,
      "step": 23560
    },
    {
      "entropy": 0.769198240339756,
      "epoch": 0.6751357002706845,
      "grad_norm": 6.28125,
      "learning_rate": 3.786614706086962e-05,
      "loss": 0.7628098487854004,
      "mean_token_accuracy": 0.8097487449645996,
      "num_tokens": 46575993.0,
      "step": 23570
    },
    {
      "entropy": 0.8417650980874896,
      "epoch": 0.6754221388367729,
      "grad_norm": 6.875,
      "learning_rate": 3.785636130534228e-05,
      "loss": 0.8811795234680175,
      "mean_token_accuracy": 0.789256677776575,
      "num_tokens": 46596226.0,
      "step": 23580
    },
    {
      "entropy": 0.8256034040823579,
      "epoch": 0.6757085774028615,
      "grad_norm": 7.125,
      "learning_rate": 3.784657287102915e-05,
      "loss": 0.8187993049621582,
      "mean_token_accuracy": 0.8100414216518402,
      "num_tokens": 46615578.0,
      "step": 23590
    },
    {
      "entropy": 0.8269222309812904,
      "epoch": 0.6759950159689501,
      "grad_norm": 8.1875,
      "learning_rate": 3.783678175996976e-05,
      "loss": 0.8582427024841308,
      "mean_token_accuracy": 0.8037688598036766,
      "num_tokens": 46635949.0,
      "step": 23600
    },
    {
      "entropy": 0.8698379435576499,
      "epoch": 0.6762814545350386,
      "grad_norm": 7.0,
      "learning_rate": 3.7826987974204245e-05,
      "loss": 0.8518117904663086,
      "mean_token_accuracy": 0.7998022887855768,
      "num_tokens": 46655453.0,
      "step": 23610
    },
    {
      "entropy": 0.9736084995791316,
      "epoch": 0.6765678931011271,
      "grad_norm": 8.25,
      "learning_rate": 3.781719151577322e-05,
      "loss": 1.00388765335083,
      "mean_token_accuracy": 0.774443319439888,
      "num_tokens": 46676981.0,
      "step": 23620
    },
    {
      "entropy": 0.7718381651677191,
      "epoch": 0.6768543316672156,
      "grad_norm": 6.125,
      "learning_rate": 3.7807392386717925e-05,
      "loss": 0.7796341419219971,
      "mean_token_accuracy": 0.8147437456995249,
      "num_tokens": 46696333.0,
      "step": 23630
    },
    {
      "entropy": 0.8309880215674639,
      "epoch": 0.6771407702333042,
      "grad_norm": 8.875,
      "learning_rate": 3.779759058908015e-05,
      "loss": 0.8464741706848145,
      "mean_token_accuracy": 0.8070103503763676,
      "num_tokens": 46715933.0,
      "step": 23640
    },
    {
      "entropy": 0.7961538839153945,
      "epoch": 0.6774272087993928,
      "grad_norm": 8.0625,
      "learning_rate": 3.7787786124902183e-05,
      "loss": 0.7830343246459961,
      "mean_token_accuracy": 0.8080305829644203,
      "num_tokens": 46735029.0,
      "step": 23650
    },
    {
      "entropy": 0.8212189765647053,
      "epoch": 0.6777136473654813,
      "grad_norm": 7.96875,
      "learning_rate": 3.777797899622694e-05,
      "loss": 0.8574483871459961,
      "mean_token_accuracy": 0.804504644498229,
      "num_tokens": 46754781.0,
      "step": 23660
    },
    {
      "entropy": 0.8547359514981508,
      "epoch": 0.6780000859315698,
      "grad_norm": 6.6875,
      "learning_rate": 3.776816920509786e-05,
      "loss": 0.885096549987793,
      "mean_token_accuracy": 0.8049687501043081,
      "num_tokens": 46773648.0,
      "step": 23670
    },
    {
      "entropy": 0.8393470600247384,
      "epoch": 0.6782865244976584,
      "grad_norm": 8.5625,
      "learning_rate": 3.775835675355891e-05,
      "loss": 0.8691526412963867,
      "mean_token_accuracy": 0.7947827219963074,
      "num_tokens": 46792902.0,
      "step": 23680
    },
    {
      "entropy": 0.9281520107761025,
      "epoch": 0.6785729630637469,
      "grad_norm": 7.84375,
      "learning_rate": 3.774854164365467e-05,
      "loss": 0.9006784439086915,
      "mean_token_accuracy": 0.7891056396067142,
      "num_tokens": 46813550.0,
      "step": 23690
    },
    {
      "entropy": 0.9720629405230283,
      "epoch": 0.6788594016298355,
      "grad_norm": 8.4375,
      "learning_rate": 3.7738723877430223e-05,
      "loss": 1.01405668258667,
      "mean_token_accuracy": 0.7691835843026638,
      "num_tokens": 46833224.0,
      "step": 23700
    },
    {
      "entropy": 0.8647079547867179,
      "epoch": 0.6791458401959239,
      "grad_norm": 8.375,
      "learning_rate": 3.772890345693123e-05,
      "loss": 0.883263111114502,
      "mean_token_accuracy": 0.7846177857369184,
      "num_tokens": 46853820.0,
      "step": 23710
    },
    {
      "entropy": 0.9543108440935612,
      "epoch": 0.6794322787620125,
      "grad_norm": 7.78125,
      "learning_rate": 3.77190803842039e-05,
      "loss": 0.9913612365722656,
      "mean_token_accuracy": 0.7813449997454882,
      "num_tokens": 46873954.0,
      "step": 23720
    },
    {
      "entropy": 0.9256950682029128,
      "epoch": 0.6797187173281011,
      "grad_norm": 7.90625,
      "learning_rate": 3.770925466129501e-05,
      "loss": 0.9193596839904785,
      "mean_token_accuracy": 0.7979480311274528,
      "num_tokens": 46893498.0,
      "step": 23730
    },
    {
      "entropy": 0.9301346521824598,
      "epoch": 0.6800051558941896,
      "grad_norm": 7.625,
      "learning_rate": 3.769942629025185e-05,
      "loss": 0.9542093276977539,
      "mean_token_accuracy": 0.7830264322459698,
      "num_tokens": 46914134.0,
      "step": 23740
    },
    {
      "entropy": 0.9466976599767805,
      "epoch": 0.6802915944602781,
      "grad_norm": 6.09375,
      "learning_rate": 3.768959527312229e-05,
      "loss": 0.9109285354614258,
      "mean_token_accuracy": 0.7890682589262724,
      "num_tokens": 46934658.0,
      "step": 23750
    },
    {
      "entropy": 0.9107558296993374,
      "epoch": 0.6805780330263667,
      "grad_norm": 8.25,
      "learning_rate": 3.767976161195475e-05,
      "loss": 0.8836893081665039,
      "mean_token_accuracy": 0.7992097400128841,
      "num_tokens": 46954110.0,
      "step": 23760
    },
    {
      "entropy": 0.9006217638030648,
      "epoch": 0.6808644715924552,
      "grad_norm": 8.1875,
      "learning_rate": 3.766992530879821e-05,
      "loss": 0.9244226455688477,
      "mean_token_accuracy": 0.7843036446720362,
      "num_tokens": 46973567.0,
      "step": 23770
    },
    {
      "entropy": 0.882096703350544,
      "epoch": 0.6811509101585438,
      "grad_norm": 6.0625,
      "learning_rate": 3.766008636570218e-05,
      "loss": 0.8864049911499023,
      "mean_token_accuracy": 0.8032234638929368,
      "num_tokens": 46994552.0,
      "step": 23780
    },
    {
      "entropy": 0.798071656934917,
      "epoch": 0.6814373487246322,
      "grad_norm": 7.5,
      "learning_rate": 3.7650244784716734e-05,
      "loss": 0.7669909954071045,
      "mean_token_accuracy": 0.8172485463321209,
      "num_tokens": 47014602.0,
      "step": 23790
    },
    {
      "entropy": 0.9085443234071136,
      "epoch": 0.6817237872907208,
      "grad_norm": 10.0625,
      "learning_rate": 3.764040056789247e-05,
      "loss": 0.9262447357177734,
      "mean_token_accuracy": 0.776689887791872,
      "num_tokens": 47035523.0,
      "step": 23800
    },
    {
      "entropy": 0.8039026020094753,
      "epoch": 0.6820102258568094,
      "grad_norm": 9.75,
      "learning_rate": 3.763055371728057e-05,
      "loss": 0.7580043315887451,
      "mean_token_accuracy": 0.8162250146269798,
      "num_tokens": 47056567.0,
      "step": 23810
    },
    {
      "entropy": 0.8289847819134593,
      "epoch": 0.6822966644228979,
      "grad_norm": 8.0625,
      "learning_rate": 3.7620704234932755e-05,
      "loss": 0.9210149765014648,
      "mean_token_accuracy": 0.7901448838412761,
      "num_tokens": 47075721.0,
      "step": 23820
    },
    {
      "entropy": 0.786400451976806,
      "epoch": 0.6825831029889864,
      "grad_norm": 7.28125,
      "learning_rate": 3.7610852122901273e-05,
      "loss": 0.7949518680572509,
      "mean_token_accuracy": 0.8030009407550096,
      "num_tokens": 47095868.0,
      "step": 23830
    },
    {
      "entropy": 0.7877665994688868,
      "epoch": 0.682869541555075,
      "grad_norm": 7.0625,
      "learning_rate": 3.760099738323896e-05,
      "loss": 0.7625834941864014,
      "mean_token_accuracy": 0.8196571815758944,
      "num_tokens": 47116376.0,
      "step": 23840
    },
    {
      "entropy": 0.8860090170055628,
      "epoch": 0.6831559801211635,
      "grad_norm": 7.9375,
      "learning_rate": 3.759114001799915e-05,
      "loss": 0.9824397087097168,
      "mean_token_accuracy": 0.7786292754113674,
      "num_tokens": 47135690.0,
      "step": 23850
    },
    {
      "entropy": 0.9434447078034282,
      "epoch": 0.6834424186872521,
      "grad_norm": 7.21875,
      "learning_rate": 3.758128002923578e-05,
      "loss": 0.9440702438354492,
      "mean_token_accuracy": 0.7875627759844065,
      "num_tokens": 47155858.0,
      "step": 23860
    },
    {
      "entropy": 0.9028685019351542,
      "epoch": 0.6837288572533405,
      "grad_norm": 8.0,
      "learning_rate": 3.757141741900327e-05,
      "loss": 0.9039607048034668,
      "mean_token_accuracy": 0.7916537120938301,
      "num_tokens": 47176672.0,
      "step": 23870
    },
    {
      "entropy": 0.9192259043455124,
      "epoch": 0.6840152958194291,
      "grad_norm": 7.34375,
      "learning_rate": 3.756155218935663e-05,
      "loss": 0.9251203536987305,
      "mean_token_accuracy": 0.7870229419320822,
      "num_tokens": 47199714.0,
      "step": 23880
    },
    {
      "entropy": 0.9896009426563979,
      "epoch": 0.6843017343855177,
      "grad_norm": 7.8125,
      "learning_rate": 3.755168434235144e-05,
      "loss": 1.0596629142761231,
      "mean_token_accuracy": 0.7645628433674574,
      "num_tokens": 47219460.0,
      "step": 23890
    },
    {
      "entropy": 0.8386084627360105,
      "epoch": 0.6845881729516062,
      "grad_norm": 6.90625,
      "learning_rate": 3.754181388004374e-05,
      "loss": 0.8011569023132324,
      "mean_token_accuracy": 0.8116822019219398,
      "num_tokens": 47237881.0,
      "step": 23900
    },
    {
      "entropy": 0.8817425843328237,
      "epoch": 0.6848746115176947,
      "grad_norm": 8.625,
      "learning_rate": 3.75319408044902e-05,
      "loss": 0.8680102348327636,
      "mean_token_accuracy": 0.7938362259417773,
      "num_tokens": 47257729.0,
      "step": 23910
    },
    {
      "entropy": 0.9330436183139682,
      "epoch": 0.6851610500837833,
      "grad_norm": 8.375,
      "learning_rate": 3.7522065117747977e-05,
      "loss": 0.9403269767761231,
      "mean_token_accuracy": 0.7798145253211259,
      "num_tokens": 47277172.0,
      "step": 23920
    },
    {
      "entropy": 0.7988518288359046,
      "epoch": 0.6854474886498718,
      "grad_norm": 8.75,
      "learning_rate": 3.7512186821874806e-05,
      "loss": 0.8403150558471679,
      "mean_token_accuracy": 0.8052618935704231,
      "num_tokens": 47295425.0,
      "step": 23930
    },
    {
      "entropy": 0.8242566533386707,
      "epoch": 0.6857339272159604,
      "grad_norm": 9.75,
      "learning_rate": 3.750230591892897e-05,
      "loss": 0.8384486198425293,
      "mean_token_accuracy": 0.8128284107893705,
      "num_tokens": 47314463.0,
      "step": 23940
    },
    {
      "entropy": 0.828976514749229,
      "epoch": 0.686020365782049,
      "grad_norm": 9.9375,
      "learning_rate": 3.749242241096926e-05,
      "loss": 0.7850234985351563,
      "mean_token_accuracy": 0.814599159732461,
      "num_tokens": 47334030.0,
      "step": 23950
    },
    {
      "entropy": 0.8347015677019953,
      "epoch": 0.6863068043481374,
      "grad_norm": 8.0625,
      "learning_rate": 3.7482536300055036e-05,
      "loss": 0.8457798957824707,
      "mean_token_accuracy": 0.8063288155943156,
      "num_tokens": 47354162.0,
      "step": 23960
    },
    {
      "entropy": 0.8873117652721703,
      "epoch": 0.686593242914226,
      "grad_norm": 10.1875,
      "learning_rate": 3.747264758824619e-05,
      "loss": 0.8859106063842773,
      "mean_token_accuracy": 0.7920020740479231,
      "num_tokens": 47374711.0,
      "step": 23970
    },
    {
      "entropy": 0.8185331977438182,
      "epoch": 0.6868796814803145,
      "grad_norm": 8.1875,
      "learning_rate": 3.746275627760318e-05,
      "loss": 0.8844112396240235,
      "mean_token_accuracy": 0.7977825257927179,
      "num_tokens": 47394024.0,
      "step": 23980
    },
    {
      "entropy": 0.9755201451480389,
      "epoch": 0.6871661200464031,
      "grad_norm": 6.03125,
      "learning_rate": 3.745286237018696e-05,
      "loss": 0.9543863296508789,
      "mean_token_accuracy": 0.775180934369564,
      "num_tokens": 47413961.0,
      "step": 23990
    },
    {
      "epoch": 0.6874525586124915,
      "eval_entropy": 0.9170454417467118,
      "eval_loss": 0.9256865978240967,
      "eval_mean_token_accuracy": 0.7836886106729507,
      "eval_num_tokens": 47434849.0,
      "eval_runtime": 43.1079,
      "eval_samples_per_second": 46.395,
      "eval_steps_per_second": 5.799,
      "step": 24000
    },
    {
      "entropy": 0.8320842329412699,
      "epoch": 0.6877389971785801,
      "grad_norm": 8.125,
      "learning_rate": 3.743306677328156e-05,
      "loss": 0.8275287628173829,
      "mean_token_accuracy": 0.8075450304895639,
      "num_tokens": 47453853.0,
      "step": 24010
    },
    {
      "entropy": 0.9436167422682047,
      "epoch": 0.6880254357446687,
      "grad_norm": 11.3125,
      "learning_rate": 3.7423165087917035e-05,
      "loss": 0.9847878456115723,
      "mean_token_accuracy": 0.7843574348837137,
      "num_tokens": 47472536.0,
      "step": 24020
    },
    {
      "entropy": 0.7477353625930846,
      "epoch": 0.6883118743107572,
      "grad_norm": 7.96875,
      "learning_rate": 3.741326081402864e-05,
      "loss": 0.7448188781738281,
      "mean_token_accuracy": 0.8303573679178953,
      "num_tokens": 47492511.0,
      "step": 24030
    },
    {
      "entropy": 0.8753996063023806,
      "epoch": 0.6885983128768457,
      "grad_norm": 10.5625,
      "learning_rate": 3.740335395368005e-05,
      "loss": 0.8863739967346191,
      "mean_token_accuracy": 0.7792684946209192,
      "num_tokens": 47511886.0,
      "step": 24040
    },
    {
      "entropy": 0.9268189895898103,
      "epoch": 0.6888847514429343,
      "grad_norm": 7.78125,
      "learning_rate": 3.739344450893549e-05,
      "loss": 0.9089580535888672,
      "mean_token_accuracy": 0.7892167646437883,
      "num_tokens": 47531608.0,
      "step": 24050
    },
    {
      "entropy": 0.8559127619490028,
      "epoch": 0.6891711900090228,
      "grad_norm": 7.71875,
      "learning_rate": 3.738353248185973e-05,
      "loss": 0.8585469245910644,
      "mean_token_accuracy": 0.7998679392039776,
      "num_tokens": 47551003.0,
      "step": 24060
    },
    {
      "entropy": 0.8106321642175317,
      "epoch": 0.6894576285751114,
      "grad_norm": 8.1875,
      "learning_rate": 3.737361787451804e-05,
      "loss": 0.7850768566131592,
      "mean_token_accuracy": 0.8118389155715704,
      "num_tokens": 47569632.0,
      "step": 24070
    },
    {
      "entropy": 0.8248413396067917,
      "epoch": 0.6897440671411998,
      "grad_norm": 9.0,
      "learning_rate": 3.736370068897626e-05,
      "loss": 0.8475696563720703,
      "mean_token_accuracy": 0.8024980660527945,
      "num_tokens": 47588638.0,
      "step": 24080
    },
    {
      "entropy": 0.7947097034193575,
      "epoch": 0.6900305057072884,
      "grad_norm": 8.875,
      "learning_rate": 3.7353780927300784e-05,
      "loss": 0.8593223571777344,
      "mean_token_accuracy": 0.797731912881136,
      "num_tokens": 47608897.0,
      "step": 24090
    },
    {
      "entropy": 0.8744115469977259,
      "epoch": 0.690316944273377,
      "grad_norm": 6.8125,
      "learning_rate": 3.7343858591558505e-05,
      "loss": 0.9088281631469727,
      "mean_token_accuracy": 0.7928579300642014,
      "num_tokens": 47628310.0,
      "step": 24100
    },
    {
      "entropy": 0.8737337691709399,
      "epoch": 0.6906033828394655,
      "grad_norm": 10.875,
      "learning_rate": 3.7333933683816856e-05,
      "loss": 0.8501747131347657,
      "mean_token_accuracy": 0.8070201613008976,
      "num_tokens": 47647897.0,
      "step": 24110
    },
    {
      "entropy": 0.852601028420031,
      "epoch": 0.690889821405554,
      "grad_norm": 7.0625,
      "learning_rate": 3.732400620614384e-05,
      "loss": 0.9036587715148926,
      "mean_token_accuracy": 0.7916225723922252,
      "num_tokens": 47668014.0,
      "step": 24120
    },
    {
      "entropy": 0.8993831885978579,
      "epoch": 0.6911762599716426,
      "grad_norm": 6.78125,
      "learning_rate": 3.731407616060796e-05,
      "loss": 0.9492385864257813,
      "mean_token_accuracy": 0.7832627769559621,
      "num_tokens": 47688098.0,
      "step": 24130
    },
    {
      "entropy": 0.7997074939310551,
      "epoch": 0.6914626985377311,
      "grad_norm": 9.8125,
      "learning_rate": 3.7304143549278266e-05,
      "loss": 0.8072966575622559,
      "mean_token_accuracy": 0.8164156705141068,
      "num_tokens": 47707525.0,
      "step": 24140
    },
    {
      "entropy": 0.8286505179479718,
      "epoch": 0.6917491371038197,
      "grad_norm": 7.34375,
      "learning_rate": 3.7294208374224335e-05,
      "loss": 0.8787816047668457,
      "mean_token_accuracy": 0.7964305099099874,
      "num_tokens": 47726620.0,
      "step": 24150
    },
    {
      "entropy": 0.8535699050873518,
      "epoch": 0.6920355756699081,
      "grad_norm": 7.5625,
      "learning_rate": 3.728427063751631e-05,
      "loss": 0.8932140350341797,
      "mean_token_accuracy": 0.7885459687560796,
      "num_tokens": 47746580.0,
      "step": 24160
    },
    {
      "entropy": 0.9530906442552805,
      "epoch": 0.6923220142359967,
      "grad_norm": 8.5625,
      "learning_rate": 3.7274330341224836e-05,
      "loss": 0.9543826103210449,
      "mean_token_accuracy": 0.7746529798954725,
      "num_tokens": 47766788.0,
      "step": 24170
    },
    {
      "entropy": 0.8913158476352692,
      "epoch": 0.6926084528020853,
      "grad_norm": 6.5,
      "learning_rate": 3.7264387487421084e-05,
      "loss": 0.8770264625549317,
      "mean_token_accuracy": 0.8090068537741899,
      "num_tokens": 47786360.0,
      "step": 24180
    },
    {
      "entropy": 0.8746037418022752,
      "epoch": 0.6928948913681738,
      "grad_norm": 7.09375,
      "learning_rate": 3.7254442078176786e-05,
      "loss": 0.889888858795166,
      "mean_token_accuracy": 0.8017952561378479,
      "num_tokens": 47805798.0,
      "step": 24190
    },
    {
      "entropy": 0.8507535761222244,
      "epoch": 0.6931813299342624,
      "grad_norm": 8.3125,
      "learning_rate": 3.72444941155642e-05,
      "loss": 0.8407798767089844,
      "mean_token_accuracy": 0.7961721129715442,
      "num_tokens": 47825230.0,
      "step": 24200
    },
    {
      "entropy": 0.8398197388276458,
      "epoch": 0.6934677685003509,
      "grad_norm": 6.96875,
      "learning_rate": 3.723454360165609e-05,
      "loss": 0.8552226066589356,
      "mean_token_accuracy": 0.7996520016342401,
      "num_tokens": 47844135.0,
      "step": 24210
    },
    {
      "entropy": 0.9954616278409958,
      "epoch": 0.6937542070664394,
      "grad_norm": 9.3125,
      "learning_rate": 3.7224590538525796e-05,
      "loss": 0.9790471076965332,
      "mean_token_accuracy": 0.765663568675518,
      "num_tokens": 47864040.0,
      "step": 24220
    },
    {
      "entropy": 0.8452286321669817,
      "epoch": 0.694040645632528,
      "grad_norm": 6.875,
      "learning_rate": 3.721463492824715e-05,
      "loss": 0.8486843109130859,
      "mean_token_accuracy": 0.8110845539718866,
      "num_tokens": 47882612.0,
      "step": 24230
    },
    {
      "entropy": 0.9404670706018805,
      "epoch": 0.6943270841986166,
      "grad_norm": 6.03125,
      "learning_rate": 3.720467677289454e-05,
      "loss": 0.9627102851867676,
      "mean_token_accuracy": 0.7750114072114229,
      "num_tokens": 47902110.0,
      "step": 24240
    },
    {
      "entropy": 0.8143452545627952,
      "epoch": 0.694613522764705,
      "grad_norm": 6.53125,
      "learning_rate": 3.719471607454286e-05,
      "loss": 0.8512958526611328,
      "mean_token_accuracy": 0.8037332255393267,
      "num_tokens": 47922124.0,
      "step": 24250
    },
    {
      "entropy": 0.9068429347127676,
      "epoch": 0.6948999613307936,
      "grad_norm": 7.78125,
      "learning_rate": 3.718475283526755e-05,
      "loss": 0.890805435180664,
      "mean_token_accuracy": 0.7938920769840478,
      "num_tokens": 47941887.0,
      "step": 24260
    },
    {
      "entropy": 0.8921038338914513,
      "epoch": 0.6951863998968821,
      "grad_norm": 8.4375,
      "learning_rate": 3.717478705714459e-05,
      "loss": 0.9193027496337891,
      "mean_token_accuracy": 0.7878501407802105,
      "num_tokens": 47961501.0,
      "step": 24270
    },
    {
      "entropy": 0.8531112568452954,
      "epoch": 0.6954728384629707,
      "grad_norm": 6.6875,
      "learning_rate": 3.716481874225046e-05,
      "loss": 0.8416627883911133,
      "mean_token_accuracy": 0.8026750486344099,
      "num_tokens": 47979434.0,
      "step": 24280
    },
    {
      "entropy": 0.8986908689141273,
      "epoch": 0.6957592770290592,
      "grad_norm": 7.46875,
      "learning_rate": 3.7154847892662206e-05,
      "loss": 0.9239437103271484,
      "mean_token_accuracy": 0.7812710590660572,
      "num_tokens": 47998403.0,
      "step": 24290
    },
    {
      "entropy": 0.7492501908913255,
      "epoch": 0.6960457155951477,
      "grad_norm": 7.40625,
      "learning_rate": 3.714487451045735e-05,
      "loss": 0.7248305320739746,
      "mean_token_accuracy": 0.828965525329113,
      "num_tokens": 48018228.0,
      "step": 24300
    },
    {
      "entropy": 0.900933374837041,
      "epoch": 0.6963321541612363,
      "grad_norm": 6.375,
      "learning_rate": 3.713489859771401e-05,
      "loss": 0.9335248947143555,
      "mean_token_accuracy": 0.7910063769668341,
      "num_tokens": 48037999.0,
      "step": 24310
    },
    {
      "entropy": 0.8868171740323305,
      "epoch": 0.6966185927273248,
      "grad_norm": 7.28125,
      "learning_rate": 3.712492015651076e-05,
      "loss": 0.8473359107971191,
      "mean_token_accuracy": 0.7943813256919384,
      "num_tokens": 48058093.0,
      "step": 24320
    },
    {
      "entropy": 0.8771782496944069,
      "epoch": 0.6969050312934133,
      "grad_norm": 7.875,
      "learning_rate": 3.711493918892675e-05,
      "loss": 0.8803249359130859,
      "mean_token_accuracy": 0.7926054686307907,
      "num_tokens": 48077899.0,
      "step": 24330
    },
    {
      "entropy": 0.7925760008394718,
      "epoch": 0.6971914698595019,
      "grad_norm": 4.40625,
      "learning_rate": 3.710495569704164e-05,
      "loss": 0.791726541519165,
      "mean_token_accuracy": 0.8075418349355459,
      "num_tokens": 48097497.0,
      "step": 24340
    },
    {
      "entropy": 0.8639530636370182,
      "epoch": 0.6974779084255904,
      "grad_norm": 10.5625,
      "learning_rate": 3.709496968293562e-05,
      "loss": 0.8617840766906738,
      "mean_token_accuracy": 0.8056206267327071,
      "num_tokens": 48115580.0,
      "step": 24350
    },
    {
      "entropy": 0.7220375942066312,
      "epoch": 0.697764346991679,
      "grad_norm": 10.6875,
      "learning_rate": 3.708498114868939e-05,
      "loss": 0.7322781085968018,
      "mean_token_accuracy": 0.8130369201302529,
      "num_tokens": 48133298.0,
      "step": 24360
    },
    {
      "entropy": 0.8128467313945293,
      "epoch": 0.6980507855577675,
      "grad_norm": 9.0,
      "learning_rate": 3.707499009638421e-05,
      "loss": 0.8220549583435058,
      "mean_token_accuracy": 0.8082006350159645,
      "num_tokens": 48153122.0,
      "step": 24370
    },
    {
      "entropy": 0.8922281099483371,
      "epoch": 0.698337224123856,
      "grad_norm": 8.0,
      "learning_rate": 3.706499652810183e-05,
      "loss": 0.9855802536010743,
      "mean_token_accuracy": 0.7793151929974556,
      "num_tokens": 48172598.0,
      "step": 24380
    },
    {
      "entropy": 0.8139053754508495,
      "epoch": 0.6986236626899446,
      "grad_norm": 10.125,
      "learning_rate": 3.7055000445924524e-05,
      "loss": 0.7709712028503418,
      "mean_token_accuracy": 0.8123778875917196,
      "num_tokens": 48191728.0,
      "step": 24390
    },
    {
      "entropy": 0.8217745428904891,
      "epoch": 0.6989101012560331,
      "grad_norm": 5.40625,
      "learning_rate": 3.704500185193512e-05,
      "loss": 0.8592321395874023,
      "mean_token_accuracy": 0.8043582398444414,
      "num_tokens": 48211604.0,
      "step": 24400
    },
    {
      "entropy": 0.8804588638246059,
      "epoch": 0.6991965398221216,
      "grad_norm": 7.625,
      "learning_rate": 3.7035000748216946e-05,
      "loss": 0.8346160888671875,
      "mean_token_accuracy": 0.8027471043169498,
      "num_tokens": 48231433.0,
      "step": 24410
    },
    {
      "entropy": 0.9157680001109838,
      "epoch": 0.6994829783882102,
      "grad_norm": 7.71875,
      "learning_rate": 3.7024997136853864e-05,
      "loss": 0.9334989547729492,
      "mean_token_accuracy": 0.7771507669240236,
      "num_tokens": 48251555.0,
      "step": 24420
    },
    {
      "entropy": 0.9577648447826504,
      "epoch": 0.6997694169542987,
      "grad_norm": 7.28125,
      "learning_rate": 3.701499101993024e-05,
      "loss": 1.029730987548828,
      "mean_token_accuracy": 0.7684868574142456,
      "num_tokens": 48273259.0,
      "step": 24430
    },
    {
      "entropy": 0.9802593341097235,
      "epoch": 0.7000558555203873,
      "grad_norm": 8.0,
      "learning_rate": 3.7004982399530994e-05,
      "loss": 0.9533609390258789,
      "mean_token_accuracy": 0.7865672312676907,
      "num_tokens": 48293398.0,
      "step": 24440
    },
    {
      "entropy": 0.9322400186210871,
      "epoch": 0.7003422940864759,
      "grad_norm": 4.90625,
      "learning_rate": 3.699497127774152e-05,
      "loss": 0.8832309722900391,
      "mean_token_accuracy": 0.7979009091854096,
      "num_tokens": 48313745.0,
      "step": 24450
    },
    {
      "entropy": 0.8969527354463935,
      "epoch": 0.7006287326525643,
      "grad_norm": 7.875,
      "learning_rate": 3.698495765664779e-05,
      "loss": 0.9008283615112305,
      "mean_token_accuracy": 0.7838116530328989,
      "num_tokens": 48333126.0,
      "step": 24460
    },
    {
      "entropy": 0.7459261387586593,
      "epoch": 0.7009151712186529,
      "grad_norm": 7.59375,
      "learning_rate": 3.697494153833626e-05,
      "loss": 0.760207986831665,
      "mean_token_accuracy": 0.8132480531930923,
      "num_tokens": 48352446.0,
      "step": 24470
    },
    {
      "entropy": 0.8882219929248094,
      "epoch": 0.7012016097847414,
      "grad_norm": 10.6875,
      "learning_rate": 3.696492292489392e-05,
      "loss": 0.9176007270812988,
      "mean_token_accuracy": 0.7964286029338836,
      "num_tokens": 48371092.0,
      "step": 24480
    },
    {
      "entropy": 0.9542246686294675,
      "epoch": 0.70148804835083,
      "grad_norm": 7.625,
      "learning_rate": 3.695490181840826e-05,
      "loss": 1.001837921142578,
      "mean_token_accuracy": 0.7761807925999165,
      "num_tokens": 48392376.0,
      "step": 24490
    },
    {
      "entropy": 0.8528949799016118,
      "epoch": 0.7017744869169185,
      "grad_norm": 8.0625,
      "learning_rate": 3.6944878220967314e-05,
      "loss": 0.8738615036010742,
      "mean_token_accuracy": 0.7925578620284796,
      "num_tokens": 48412189.0,
      "step": 24500
    },
    {
      "entropy": 0.8872066676616669,
      "epoch": 0.702060925483007,
      "grad_norm": 6.71875,
      "learning_rate": 3.693485213465963e-05,
      "loss": 0.9026021957397461,
      "mean_token_accuracy": 0.7847229167819023,
      "num_tokens": 48433645.0,
      "step": 24510
    },
    {
      "entropy": 0.8101624321192503,
      "epoch": 0.7023473640490956,
      "grad_norm": 12.0,
      "learning_rate": 3.692482356157427e-05,
      "loss": 0.8111007690429688,
      "mean_token_accuracy": 0.8094050262123347,
      "num_tokens": 48453130.0,
      "step": 24520
    },
    {
      "entropy": 0.8125691677443683,
      "epoch": 0.7026338026151842,
      "grad_norm": 8.5,
      "learning_rate": 3.6914792503800797e-05,
      "loss": 0.7810537815093994,
      "mean_token_accuracy": 0.8109431155025959,
      "num_tokens": 48473585.0,
      "step": 24530
    },
    {
      "entropy": 0.7966423815116286,
      "epoch": 0.7029202411812726,
      "grad_norm": 8.875,
      "learning_rate": 3.690475896342933e-05,
      "loss": 0.8331488609313965,
      "mean_token_accuracy": 0.8114167861640453,
      "num_tokens": 48494044.0,
      "step": 24540
    },
    {
      "entropy": 0.8213833974674344,
      "epoch": 0.7032066797473612,
      "grad_norm": 9.4375,
      "learning_rate": 3.689472294255047e-05,
      "loss": 0.8178174018859863,
      "mean_token_accuracy": 0.8069458220154047,
      "num_tokens": 48513800.0,
      "step": 24550
    },
    {
      "entropy": 0.8288965802639723,
      "epoch": 0.7034931183134497,
      "grad_norm": 7.5625,
      "learning_rate": 3.688468444325536e-05,
      "loss": 0.7747990608215332,
      "mean_token_accuracy": 0.8208260878920555,
      "num_tokens": 48533062.0,
      "step": 24560
    },
    {
      "entropy": 0.7590049094520509,
      "epoch": 0.7037795568795383,
      "grad_norm": 6.78125,
      "learning_rate": 3.687464346763565e-05,
      "loss": 0.7685303211212158,
      "mean_token_accuracy": 0.8226388655602932,
      "num_tokens": 48551359.0,
      "step": 24570
    },
    {
      "entropy": 0.7918385278433562,
      "epoch": 0.7040659954456268,
      "grad_norm": 8.125,
      "learning_rate": 3.686460001778349e-05,
      "loss": 0.7780099391937256,
      "mean_token_accuracy": 0.8081594161689282,
      "num_tokens": 48571030.0,
      "step": 24580
    },
    {
      "entropy": 0.9068219905719161,
      "epoch": 0.7043524340117153,
      "grad_norm": 7.5,
      "learning_rate": 3.685455409579157e-05,
      "loss": 0.9807154655456543,
      "mean_token_accuracy": 0.7835599850863219,
      "num_tokens": 48590412.0,
      "step": 24590
    },
    {
      "entropy": 0.8774999534711242,
      "epoch": 0.7046388725778039,
      "grad_norm": 9.9375,
      "learning_rate": 3.6844505703753085e-05,
      "loss": 0.93710355758667,
      "mean_token_accuracy": 0.7930587969720364,
      "num_tokens": 48609509.0,
      "step": 24600
    },
    {
      "entropy": 0.9169191809371113,
      "epoch": 0.7049253111438925,
      "grad_norm": 8.0625,
      "learning_rate": 3.683445484376174e-05,
      "loss": 0.9161656379699707,
      "mean_token_accuracy": 0.7990161798894405,
      "num_tokens": 48628402.0,
      "step": 24610
    },
    {
      "entropy": 0.784486335888505,
      "epoch": 0.7052117497099809,
      "grad_norm": 8.4375,
      "learning_rate": 3.6824401517911765e-05,
      "loss": 0.7554248332977295,
      "mean_token_accuracy": 0.8186745442450046,
      "num_tokens": 48646478.0,
      "step": 24620
    },
    {
      "entropy": 0.917958570830524,
      "epoch": 0.7054981882760695,
      "grad_norm": 8.3125,
      "learning_rate": 3.6814345728297895e-05,
      "loss": 0.8936305046081543,
      "mean_token_accuracy": 0.7865938775241375,
      "num_tokens": 48666432.0,
      "step": 24630
    },
    {
      "entropy": 0.8962295152246952,
      "epoch": 0.705784626842158,
      "grad_norm": 7.78125,
      "learning_rate": 3.680428747701539e-05,
      "loss": 0.9056150436401367,
      "mean_token_accuracy": 0.7982874237000942,
      "num_tokens": 48686377.0,
      "step": 24640
    },
    {
      "entropy": 0.8813980758190155,
      "epoch": 0.7060710654082466,
      "grad_norm": 10.0625,
      "learning_rate": 3.679422676615999e-05,
      "loss": 0.8951904296875,
      "mean_token_accuracy": 0.7825820710510015,
      "num_tokens": 48705216.0,
      "step": 24650
    },
    {
      "entropy": 0.8333178224042058,
      "epoch": 0.706357503974335,
      "grad_norm": 9.1875,
      "learning_rate": 3.678416359782799e-05,
      "loss": 0.7824797630310059,
      "mean_token_accuracy": 0.8113893788307905,
      "num_tokens": 48724770.0,
      "step": 24660
    },
    {
      "entropy": 0.8183223536238075,
      "epoch": 0.7066439425404236,
      "grad_norm": 7.75,
      "learning_rate": 3.677409797411618e-05,
      "loss": 0.8160299301147461,
      "mean_token_accuracy": 0.8100057207047939,
      "num_tokens": 48745284.0,
      "step": 24670
    },
    {
      "entropy": 0.8935088727623224,
      "epoch": 0.7069303811065122,
      "grad_norm": 7.9375,
      "learning_rate": 3.6764029897121856e-05,
      "loss": 0.9133176803588867,
      "mean_token_accuracy": 0.7844197612255812,
      "num_tokens": 48764447.0,
      "step": 24680
    },
    {
      "entropy": 0.8730123307555914,
      "epoch": 0.7072168196726007,
      "grad_norm": 7.375,
      "learning_rate": 3.675395936894283e-05,
      "loss": 0.9146931648254395,
      "mean_token_accuracy": 0.800732908770442,
      "num_tokens": 48784723.0,
      "step": 24690
    },
    {
      "entropy": 0.8042549195699393,
      "epoch": 0.7075032582386892,
      "grad_norm": 7.75,
      "learning_rate": 3.674388639167742e-05,
      "loss": 0.8193799018859863,
      "mean_token_accuracy": 0.8056302301585674,
      "num_tokens": 48803224.0,
      "step": 24700
    },
    {
      "entropy": 0.8660958832129836,
      "epoch": 0.7077896968047778,
      "grad_norm": 9.5,
      "learning_rate": 3.673381096742446e-05,
      "loss": 0.8972262382507324,
      "mean_token_accuracy": 0.7992315500974655,
      "num_tokens": 48822055.0,
      "step": 24710
    },
    {
      "entropy": 0.7201851662248373,
      "epoch": 0.7080761353708663,
      "grad_norm": 8.0,
      "learning_rate": 3.6723733098283305e-05,
      "loss": 0.7442617416381836,
      "mean_token_accuracy": 0.832540974766016,
      "num_tokens": 48840568.0,
      "step": 24720
    },
    {
      "entropy": 0.7807918494567275,
      "epoch": 0.7083625739369549,
      "grad_norm": 8.125,
      "learning_rate": 3.6713652786353786e-05,
      "loss": 0.7479044914245605,
      "mean_token_accuracy": 0.8165827848017215,
      "num_tokens": 48859453.0,
      "step": 24730
    },
    {
      "entropy": 0.8202137500047684,
      "epoch": 0.7086490125030435,
      "grad_norm": 8.5625,
      "learning_rate": 3.6703570033736275e-05,
      "loss": 0.8364398002624511,
      "mean_token_accuracy": 0.8051929485052824,
      "num_tokens": 48880396.0,
      "step": 24740
    },
    {
      "entropy": 0.8442575002089143,
      "epoch": 0.7089354510691319,
      "grad_norm": 8.1875,
      "learning_rate": 3.6693484842531645e-05,
      "loss": 0.8244970321655274,
      "mean_token_accuracy": 0.8019424226135016,
      "num_tokens": 48901228.0,
      "step": 24750
    },
    {
      "entropy": 0.8255954269319773,
      "epoch": 0.7092218896352205,
      "grad_norm": 9.0625,
      "learning_rate": 3.668339721484126e-05,
      "loss": 0.8358854293823242,
      "mean_token_accuracy": 0.8070498321205377,
      "num_tokens": 48920689.0,
      "step": 24760
    },
    {
      "entropy": 0.8563813163898886,
      "epoch": 0.709508328201309,
      "grad_norm": 8.875,
      "learning_rate": 3.667330715276702e-05,
      "loss": 0.8916967391967774,
      "mean_token_accuracy": 0.7933955866843462,
      "num_tokens": 48940730.0,
      "step": 24770
    },
    {
      "entropy": 0.7875993371941149,
      "epoch": 0.7097947667673976,
      "grad_norm": 8.875,
      "learning_rate": 3.6663214658411303e-05,
      "loss": 0.856473731994629,
      "mean_token_accuracy": 0.8085438180714846,
      "num_tokens": 48960175.0,
      "step": 24780
    },
    {
      "entropy": 0.8658637532964348,
      "epoch": 0.7100812053334861,
      "grad_norm": 8.25,
      "learning_rate": 3.665311973387703e-05,
      "loss": 0.8882229804992676,
      "mean_token_accuracy": 0.7904190748929978,
      "num_tokens": 48980088.0,
      "step": 24790
    },
    {
      "entropy": 0.8634634293615818,
      "epoch": 0.7103676438995746,
      "grad_norm": 6.90625,
      "learning_rate": 3.664302238126758e-05,
      "loss": 0.8364635467529297,
      "mean_token_accuracy": 0.7960839185863733,
      "num_tokens": 49000063.0,
      "step": 24800
    },
    {
      "entropy": 0.9048838572576642,
      "epoch": 0.7106540824656632,
      "grad_norm": 9.3125,
      "learning_rate": 3.663292260268688e-05,
      "loss": 0.9128030776977539,
      "mean_token_accuracy": 0.7919073175638914,
      "num_tokens": 49020191.0,
      "step": 24810
    },
    {
      "entropy": 0.9535730510950089,
      "epoch": 0.7109405210317518,
      "grad_norm": 10.375,
      "learning_rate": 3.662282040023936e-05,
      "loss": 0.9490864753723145,
      "mean_token_accuracy": 0.782025120407343,
      "num_tokens": 49039128.0,
      "step": 24820
    },
    {
      "entropy": 0.8247288456186652,
      "epoch": 0.7112269595978402,
      "grad_norm": 8.1875,
      "learning_rate": 3.66127157760299e-05,
      "loss": 0.8833017349243164,
      "mean_token_accuracy": 0.8069334965199232,
      "num_tokens": 49058248.0,
      "step": 24830
    },
    {
      "entropy": 0.9842846591025591,
      "epoch": 0.7115133981639288,
      "grad_norm": 7.65625,
      "learning_rate": 3.660260873216396e-05,
      "loss": 1.0087266921997071,
      "mean_token_accuracy": 0.7835214611142873,
      "num_tokens": 49079877.0,
      "step": 24840
    },
    {
      "entropy": 0.8269510615617037,
      "epoch": 0.7117998367300173,
      "grad_norm": 9.1875,
      "learning_rate": 3.659249927074747e-05,
      "loss": 0.8558285713195801,
      "mean_token_accuracy": 0.8082619827240706,
      "num_tokens": 49098187.0,
      "step": 24850
    },
    {
      "entropy": 0.8203220343217253,
      "epoch": 0.7120862752961059,
      "grad_norm": 6.53125,
      "learning_rate": 3.658238739388685e-05,
      "loss": 0.7810757160186768,
      "mean_token_accuracy": 0.8198346484452486,
      "num_tokens": 49117854.0,
      "step": 24860
    },
    {
      "entropy": 0.8428788097575307,
      "epoch": 0.7123727138621944,
      "grad_norm": 7.9375,
      "learning_rate": 3.6572273103689045e-05,
      "loss": 0.8275889396667481,
      "mean_token_accuracy": 0.8003102771937847,
      "num_tokens": 49137955.0,
      "step": 24870
    },
    {
      "entropy": 0.8262346537783742,
      "epoch": 0.7126591524282829,
      "grad_norm": 6.21875,
      "learning_rate": 3.6562156402261496e-05,
      "loss": 0.8443349838256836,
      "mean_token_accuracy": 0.8036100789904594,
      "num_tokens": 49157801.0,
      "step": 24880
    },
    {
      "entropy": 0.8555899580940605,
      "epoch": 0.7129455909943715,
      "grad_norm": 8.625,
      "learning_rate": 3.655203729171215e-05,
      "loss": 0.862637996673584,
      "mean_token_accuracy": 0.7958963971585036,
      "num_tokens": 49177160.0,
      "step": 24890
    },
    {
      "entropy": 0.7836555119603872,
      "epoch": 0.7132320295604601,
      "grad_norm": 8.9375,
      "learning_rate": 3.654191577414944e-05,
      "loss": 0.7367152690887451,
      "mean_token_accuracy": 0.8232878509908914,
      "num_tokens": 49196941.0,
      "step": 24900
    },
    {
      "entropy": 0.7707573924213648,
      "epoch": 0.7135184681265485,
      "grad_norm": 8.3125,
      "learning_rate": 3.653179185168232e-05,
      "loss": 0.7978598117828369,
      "mean_token_accuracy": 0.813903345912695,
      "num_tokens": 49217342.0,
      "step": 24910
    },
    {
      "entropy": 0.9440126268193125,
      "epoch": 0.7138049066926371,
      "grad_norm": 8.5625,
      "learning_rate": 3.652166552642025e-05,
      "loss": 0.9993322372436524,
      "mean_token_accuracy": 0.7818494781851768,
      "num_tokens": 49239298.0,
      "step": 24920
    },
    {
      "entropy": 0.8518192843534053,
      "epoch": 0.7140913452587256,
      "grad_norm": 6.6875,
      "learning_rate": 3.651153680047315e-05,
      "loss": 0.8596327781677247,
      "mean_token_accuracy": 0.8030738394707442,
      "num_tokens": 49260252.0,
      "step": 24930
    },
    {
      "entropy": 1.0094398139044642,
      "epoch": 0.7143777838248142,
      "grad_norm": 7.5625,
      "learning_rate": 3.6501405675951485e-05,
      "loss": 0.9751731872558593,
      "mean_token_accuracy": 0.7781467054039239,
      "num_tokens": 49280723.0,
      "step": 24940
    },
    {
      "entropy": 0.9012328576296568,
      "epoch": 0.7146642223909027,
      "grad_norm": 7.96875,
      "learning_rate": 3.6491272154966207e-05,
      "loss": 0.8963910102844238,
      "mean_token_accuracy": 0.7931661121547222,
      "num_tokens": 49300898.0,
      "step": 24950
    },
    {
      "entropy": 0.8988064870238304,
      "epoch": 0.7149506609569912,
      "grad_norm": 11.6875,
      "learning_rate": 3.6481136239628755e-05,
      "loss": 0.8269492149353027,
      "mean_token_accuracy": 0.8027182165533304,
      "num_tokens": 49319620.0,
      "step": 24960
    },
    {
      "entropy": 0.8501068165525794,
      "epoch": 0.7152370995230798,
      "grad_norm": 6.78125,
      "learning_rate": 3.647099793205109e-05,
      "loss": 0.8887551307678223,
      "mean_token_accuracy": 0.7910312186926604,
      "num_tokens": 49339040.0,
      "step": 24970
    },
    {
      "entropy": 0.8463203959167004,
      "epoch": 0.7155235380891684,
      "grad_norm": 7.46875,
      "learning_rate": 3.646085723434562e-05,
      "loss": 0.8355653762817383,
      "mean_token_accuracy": 0.7973845578730107,
      "num_tokens": 49358278.0,
      "step": 24980
    },
    {
      "entropy": 0.8013463664799929,
      "epoch": 0.7158099766552569,
      "grad_norm": 7.6875,
      "learning_rate": 3.645071414862534e-05,
      "loss": 0.8089795112609863,
      "mean_token_accuracy": 0.8143985062837601,
      "num_tokens": 49377020.0,
      "step": 24990
    },
    {
      "entropy": 0.811373152025044,
      "epoch": 0.7160964152213454,
      "grad_norm": 9.4375,
      "learning_rate": 3.644056867700365e-05,
      "loss": 0.8276477813720703,
      "mean_token_accuracy": 0.7968491937965154,
      "num_tokens": 49398466.0,
      "step": 25000
    },
    {
      "entropy": 0.8552335403859616,
      "epoch": 0.7163828537874339,
      "grad_norm": 8.875,
      "learning_rate": 3.64304208215945e-05,
      "loss": 0.823689079284668,
      "mean_token_accuracy": 0.802936702221632,
      "num_tokens": 49419793.0,
      "step": 25010
    },
    {
      "entropy": 0.823295722156763,
      "epoch": 0.7166692923535225,
      "grad_norm": 9.375,
      "learning_rate": 3.6420270584512336e-05,
      "loss": 0.9247366905212402,
      "mean_token_accuracy": 0.7920031473040581,
      "num_tokens": 49441303.0,
      "step": 25020
    },
    {
      "entropy": 0.8406313685700297,
      "epoch": 0.7169557309196111,
      "grad_norm": 8.1875,
      "learning_rate": 3.641011796787207e-05,
      "loss": 0.8393570899963378,
      "mean_token_accuracy": 0.8010467164218426,
      "num_tokens": 49459610.0,
      "step": 25030
    },
    {
      "entropy": 0.8841170135885477,
      "epoch": 0.7172421694856995,
      "grad_norm": 6.5,
      "learning_rate": 3.639996297378914e-05,
      "loss": 0.8965392112731934,
      "mean_token_accuracy": 0.7953462149947882,
      "num_tokens": 49480033.0,
      "step": 25040
    },
    {
      "entropy": 0.944437708798796,
      "epoch": 0.7175286080517881,
      "grad_norm": 6.0625,
      "learning_rate": 3.638980560437947e-05,
      "loss": 0.9791392326354981,
      "mean_token_accuracy": 0.7897530149668455,
      "num_tokens": 49500167.0,
      "step": 25050
    },
    {
      "entropy": 0.9637589974328875,
      "epoch": 0.7178150466178767,
      "grad_norm": 8.25,
      "learning_rate": 3.637964586175946e-05,
      "loss": 1.007373046875,
      "mean_token_accuracy": 0.7755039159208537,
      "num_tokens": 49521338.0,
      "step": 25060
    },
    {
      "entropy": 0.8788907270878553,
      "epoch": 0.7181014851839652,
      "grad_norm": 12.1875,
      "learning_rate": 3.636948374804604e-05,
      "loss": 0.8600860595703125,
      "mean_token_accuracy": 0.8039681255817414,
      "num_tokens": 49540880.0,
      "step": 25070
    },
    {
      "entropy": 0.7687335006892682,
      "epoch": 0.7183879237500537,
      "grad_norm": 6.84375,
      "learning_rate": 3.635931926535659e-05,
      "loss": 0.7829152584075928,
      "mean_token_accuracy": 0.818188413232565,
      "num_tokens": 49559596.0,
      "step": 25080
    },
    {
      "entropy": 0.7920591594651342,
      "epoch": 0.7186743623161422,
      "grad_norm": 9.125,
      "learning_rate": 3.634915241580905e-05,
      "loss": 0.786812686920166,
      "mean_token_accuracy": 0.8104493461549283,
      "num_tokens": 49579145.0,
      "step": 25090
    },
    {
      "entropy": 0.8454561771824955,
      "epoch": 0.7189608008822308,
      "grad_norm": 7.9375,
      "learning_rate": 3.633898320152178e-05,
      "loss": 0.8687270164489747,
      "mean_token_accuracy": 0.8007260449230671,
      "num_tokens": 49600230.0,
      "step": 25100
    },
    {
      "entropy": 0.9388252783566713,
      "epoch": 0.7192472394483194,
      "grad_norm": 7.34375,
      "learning_rate": 3.632881162461367e-05,
      "loss": 0.9690038681030273,
      "mean_token_accuracy": 0.7759719740599393,
      "num_tokens": 49620081.0,
      "step": 25110
    },
    {
      "entropy": 0.8162021109834313,
      "epoch": 0.7195336780144078,
      "grad_norm": 10.5,
      "learning_rate": 3.6318637687204094e-05,
      "loss": 0.8754613876342774,
      "mean_token_accuracy": 0.8002492271363735,
      "num_tokens": 49637601.0,
      "step": 25120
    },
    {
      "entropy": 0.8678007081151009,
      "epoch": 0.7198201165804964,
      "grad_norm": 7.78125,
      "learning_rate": 3.6308461391412935e-05,
      "loss": 0.7952215671539307,
      "mean_token_accuracy": 0.8012065790593624,
      "num_tokens": 49657406.0,
      "step": 25130
    },
    {
      "entropy": 0.9015241904184222,
      "epoch": 0.720106555146585,
      "grad_norm": 7.15625,
      "learning_rate": 3.629828273936055e-05,
      "loss": 0.8606724739074707,
      "mean_token_accuracy": 0.8041101556271315,
      "num_tokens": 49677243.0,
      "step": 25140
    },
    {
      "entropy": 0.8616856399923563,
      "epoch": 0.7203929937126735,
      "grad_norm": 6.4375,
      "learning_rate": 3.628810173316777e-05,
      "loss": 0.8727873802185059,
      "mean_token_accuracy": 0.8025168646126986,
      "num_tokens": 49696986.0,
      "step": 25150
    },
    {
      "entropy": 0.7870444439351558,
      "epoch": 0.720679432278762,
      "grad_norm": 7.15625,
      "learning_rate": 3.6277918374955956e-05,
      "loss": 0.7852448463439942,
      "mean_token_accuracy": 0.8184985876083374,
      "num_tokens": 49715352.0,
      "step": 25160
    },
    {
      "entropy": 0.7633501128293574,
      "epoch": 0.7209658708448505,
      "grad_norm": 7.3125,
      "learning_rate": 3.6267732666846934e-05,
      "loss": 0.7551411628723145,
      "mean_token_accuracy": 0.8186328396201134,
      "num_tokens": 49734986.0,
      "step": 25170
    },
    {
      "entropy": 0.85511832581833,
      "epoch": 0.7212523094109391,
      "grad_norm": 6.75,
      "learning_rate": 3.6257544610963025e-05,
      "loss": 0.8977536201477051,
      "mean_token_accuracy": 0.7954110193997621,
      "num_tokens": 49756789.0,
      "step": 25180
    },
    {
      "entropy": 0.8978054605424404,
      "epoch": 0.7215387479770277,
      "grad_norm": 10.875,
      "learning_rate": 3.624735420942705e-05,
      "loss": 0.8928098678588867,
      "mean_token_accuracy": 0.7939706485718488,
      "num_tokens": 49776027.0,
      "step": 25190
    },
    {
      "entropy": 0.9643866701051593,
      "epoch": 0.7218251865431161,
      "grad_norm": 8.5,
      "learning_rate": 3.623716146436229e-05,
      "loss": 0.970761775970459,
      "mean_token_accuracy": 0.779435719549656,
      "num_tokens": 49796575.0,
      "step": 25200
    },
    {
      "entropy": 0.8257108256220818,
      "epoch": 0.7221116251092047,
      "grad_norm": 8.375,
      "learning_rate": 3.6226966377892546e-05,
      "loss": 0.8336959838867187,
      "mean_token_accuracy": 0.8105511389672756,
      "num_tokens": 49814955.0,
      "step": 25210
    },
    {
      "entropy": 0.886202254705131,
      "epoch": 0.7223980636752932,
      "grad_norm": 7.78125,
      "learning_rate": 3.6216768952142093e-05,
      "loss": 0.8790486335754395,
      "mean_token_accuracy": 0.7987617738544941,
      "num_tokens": 49833723.0,
      "step": 25220
    },
    {
      "entropy": 0.8769304728135466,
      "epoch": 0.7226845022413818,
      "grad_norm": 6.59375,
      "learning_rate": 3.620656918923569e-05,
      "loss": 0.8774904251098633,
      "mean_token_accuracy": 0.7949467454105615,
      "num_tokens": 49854822.0,
      "step": 25230
    },
    {
      "entropy": 0.9062212869524956,
      "epoch": 0.7229709408074703,
      "grad_norm": 8.5625,
      "learning_rate": 3.619636709129858e-05,
      "loss": 0.9479777336120605,
      "mean_token_accuracy": 0.7838595557957888,
      "num_tokens": 49873575.0,
      "step": 25240
    },
    {
      "entropy": 0.7590373813640326,
      "epoch": 0.7232573793735588,
      "grad_norm": 7.9375,
      "learning_rate": 3.618616266045651e-05,
      "loss": 0.7561225891113281,
      "mean_token_accuracy": 0.820141177624464,
      "num_tokens": 49893139.0,
      "step": 25250
    },
    {
      "entropy": 0.7808217267505825,
      "epoch": 0.7235438179396474,
      "grad_norm": 8.3125,
      "learning_rate": 3.617595589883569e-05,
      "loss": 0.7352260112762451,
      "mean_token_accuracy": 0.8220109827816486,
      "num_tokens": 49911853.0,
      "step": 25260
    },
    {
      "entropy": 0.8877162696793676,
      "epoch": 0.723830256505736,
      "grad_norm": 4.59375,
      "learning_rate": 3.616574680856285e-05,
      "loss": 0.8666891098022461,
      "mean_token_accuracy": 0.8051755167543888,
      "num_tokens": 49931155.0,
      "step": 25270
    },
    {
      "entropy": 0.8880802664905787,
      "epoch": 0.7241166950718245,
      "grad_norm": 12.1875,
      "learning_rate": 3.6155535391765165e-05,
      "loss": 0.9708821296691894,
      "mean_token_accuracy": 0.7877816524356603,
      "num_tokens": 49950514.0,
      "step": 25280
    },
    {
      "entropy": 0.8563689401373267,
      "epoch": 0.724403133637913,
      "grad_norm": 8.0625,
      "learning_rate": 3.614532165057032e-05,
      "loss": 0.8444599151611328,
      "mean_token_accuracy": 0.7887032944709063,
      "num_tokens": 49970892.0,
      "step": 25290
    },
    {
      "entropy": 0.8647924887016416,
      "epoch": 0.7246895722040015,
      "grad_norm": 9.5,
      "learning_rate": 3.6135105587106474e-05,
      "loss": 0.8575145721435546,
      "mean_token_accuracy": 0.8014380410313606,
      "num_tokens": 49989829.0,
      "step": 25300
    },
    {
      "entropy": 0.8008024420589208,
      "epoch": 0.7249760107700901,
      "grad_norm": 6.34375,
      "learning_rate": 3.6124887203502276e-05,
      "loss": 0.7911601543426514,
      "mean_token_accuracy": 0.8129055164754391,
      "num_tokens": 50009042.0,
      "step": 25310
    },
    {
      "entropy": 0.8116602819412947,
      "epoch": 0.7252624493361787,
      "grad_norm": 14.625,
      "learning_rate": 3.611466650188685e-05,
      "loss": 0.8293948173522949,
      "mean_token_accuracy": 0.8167246378958225,
      "num_tokens": 50027625.0,
      "step": 25320
    },
    {
      "entropy": 0.913325471803546,
      "epoch": 0.7255488879022671,
      "grad_norm": 8.1875,
      "learning_rate": 3.6104443484389815e-05,
      "loss": 0.9089467048645019,
      "mean_token_accuracy": 0.7899933759123087,
      "num_tokens": 50046987.0,
      "step": 25330
    },
    {
      "entropy": 0.7464435068890453,
      "epoch": 0.7258353264683557,
      "grad_norm": 9.25,
      "learning_rate": 3.609421815314127e-05,
      "loss": 0.7980657577514648,
      "mean_token_accuracy": 0.8171733662486076,
      "num_tokens": 50066084.0,
      "step": 25340
    },
    {
      "entropy": 0.9087998055852949,
      "epoch": 0.7261217650344443,
      "grad_norm": 7.90625,
      "learning_rate": 3.6083990510271785e-05,
      "loss": 0.8500869750976563,
      "mean_token_accuracy": 0.8082822248339653,
      "num_tokens": 50085982.0,
      "step": 25350
    },
    {
      "entropy": 0.805213930644095,
      "epoch": 0.7264082036005328,
      "grad_norm": 7.1875,
      "learning_rate": 3.6073760557912424e-05,
      "loss": 0.8128963470458984,
      "mean_token_accuracy": 0.8148401349782943,
      "num_tokens": 50105624.0,
      "step": 25360
    },
    {
      "entropy": 0.857085132598877,
      "epoch": 0.7266946421666213,
      "grad_norm": 7.0625,
      "learning_rate": 3.6063528298194725e-05,
      "loss": 0.883035945892334,
      "mean_token_accuracy": 0.7993238713592291,
      "num_tokens": 50125907.0,
      "step": 25370
    },
    {
      "entropy": 0.8717297909781336,
      "epoch": 0.7269810807327098,
      "grad_norm": 8.375,
      "learning_rate": 3.6053293733250696e-05,
      "loss": 0.9299320220947266,
      "mean_token_accuracy": 0.7942509803920984,
      "num_tokens": 50145674.0,
      "step": 25380
    },
    {
      "entropy": 0.8375752992928028,
      "epoch": 0.7272675192987984,
      "grad_norm": 6.34375,
      "learning_rate": 3.604305686521286e-05,
      "loss": 0.8175417900085449,
      "mean_token_accuracy": 0.800636537373066,
      "num_tokens": 50165434.0,
      "step": 25390
    },
    {
      "entropy": 0.7830001374706625,
      "epoch": 0.727553957864887,
      "grad_norm": 6.78125,
      "learning_rate": 3.6032817696214196e-05,
      "loss": 0.7582241535186768,
      "mean_token_accuracy": 0.8185066506266594,
      "num_tokens": 50184485.0,
      "step": 25400
    },
    {
      "entropy": 0.767955869436264,
      "epoch": 0.7278403964309754,
      "grad_norm": 9.5625,
      "learning_rate": 3.602257622838814e-05,
      "loss": 0.7483684539794921,
      "mean_token_accuracy": 0.8316808439791202,
      "num_tokens": 50202246.0,
      "step": 25410
    },
    {
      "entropy": 0.7491380538791418,
      "epoch": 0.728126834997064,
      "grad_norm": 8.5625,
      "learning_rate": 3.601233246386866e-05,
      "loss": 0.7491294384002686,
      "mean_token_accuracy": 0.8253129716962576,
      "num_tokens": 50221915.0,
      "step": 25420
    },
    {
      "entropy": 0.8463462656363845,
      "epoch": 0.7284132735631526,
      "grad_norm": 8.125,
      "learning_rate": 3.600208640479016e-05,
      "loss": 0.883491039276123,
      "mean_token_accuracy": 0.7949466314166784,
      "num_tokens": 50241010.0,
      "step": 25430
    },
    {
      "entropy": 0.7316294064745307,
      "epoch": 0.7286997121292411,
      "grad_norm": 8.875,
      "learning_rate": 3.599183805328754e-05,
      "loss": 0.7621360301971436,
      "mean_token_accuracy": 0.8111925840377807,
      "num_tokens": 50260881.0,
      "step": 25440
    },
    {
      "entropy": 0.9040399137884378,
      "epoch": 0.7289861506953296,
      "grad_norm": 7.3125,
      "learning_rate": 3.5981587411496174e-05,
      "loss": 0.9092502593994141,
      "mean_token_accuracy": 0.7972023166716099,
      "num_tokens": 50280390.0,
      "step": 25450
    },
    {
      "entropy": 0.8928201645612717,
      "epoch": 0.7292725892614181,
      "grad_norm": 6.21875,
      "learning_rate": 3.597133448155191e-05,
      "loss": 0.9162413597106933,
      "mean_token_accuracy": 0.7866269256919622,
      "num_tokens": 50299917.0,
      "step": 25460
    },
    {
      "entropy": 0.9010098559781909,
      "epoch": 0.7295590278275067,
      "grad_norm": 6.59375,
      "learning_rate": 3.596107926559108e-05,
      "loss": 0.9302038192749024,
      "mean_token_accuracy": 0.7894194804131984,
      "num_tokens": 50319321.0,
      "step": 25470
    },
    {
      "entropy": 0.8775648446753621,
      "epoch": 0.7298454663935953,
      "grad_norm": 10.75,
      "learning_rate": 3.5950821765750485e-05,
      "loss": 0.8193497657775879,
      "mean_token_accuracy": 0.811574999243021,
      "num_tokens": 50338741.0,
      "step": 25480
    },
    {
      "entropy": 0.8680825915187598,
      "epoch": 0.7301319049596837,
      "grad_norm": 8.0625,
      "learning_rate": 3.594056198416741e-05,
      "loss": 0.878885555267334,
      "mean_token_accuracy": 0.7932611789554358,
      "num_tokens": 50358014.0,
      "step": 25490
    },
    {
      "entropy": 0.8629014974460005,
      "epoch": 0.7304183435257723,
      "grad_norm": 7.78125,
      "learning_rate": 3.59302999229796e-05,
      "loss": 0.8506993293762207,
      "mean_token_accuracy": 0.8075351934880018,
      "num_tokens": 50378092.0,
      "step": 25500
    },
    {
      "entropy": 0.8716317000798881,
      "epoch": 0.7307047820918608,
      "grad_norm": 7.75,
      "learning_rate": 3.59200355843253e-05,
      "loss": 0.8653067588806153,
      "mean_token_accuracy": 0.8030672948807478,
      "num_tokens": 50398477.0,
      "step": 25510
    },
    {
      "entropy": 0.8630946153774858,
      "epoch": 0.7309912206579494,
      "grad_norm": 8.125,
      "learning_rate": 3.59097689703432e-05,
      "loss": 0.9012486457824707,
      "mean_token_accuracy": 0.7892118867486715,
      "num_tokens": 50418802.0,
      "step": 25520
    },
    {
      "entropy": 1.0236148610711098,
      "epoch": 0.731277659224038,
      "grad_norm": 9.625,
      "learning_rate": 3.58995000831725e-05,
      "loss": 1.0553603172302246,
      "mean_token_accuracy": 0.7616301469504834,
      "num_tokens": 50440334.0,
      "step": 25530
    },
    {
      "entropy": 0.8734800208359956,
      "epoch": 0.7315640977901264,
      "grad_norm": 9.25,
      "learning_rate": 3.588922892495283e-05,
      "loss": 0.8897005081176758,
      "mean_token_accuracy": 0.7944554537534714,
      "num_tokens": 50460114.0,
      "step": 25540
    },
    {
      "entropy": 0.8958484664559364,
      "epoch": 0.731850536356215,
      "grad_norm": 5.71875,
      "learning_rate": 3.5878955497824326e-05,
      "loss": 0.885152530670166,
      "mean_token_accuracy": 0.793623449653387,
      "num_tokens": 50480277.0,
      "step": 25550
    },
    {
      "entropy": 0.9197828475385904,
      "epoch": 0.7321369749223036,
      "grad_norm": 9.5,
      "learning_rate": 3.586867980392758e-05,
      "loss": 0.9227399826049805,
      "mean_token_accuracy": 0.7779494524002075,
      "num_tokens": 50500179.0,
      "step": 25560
    },
    {
      "entropy": 0.9099521866068244,
      "epoch": 0.7324234134883921,
      "grad_norm": 7.75,
      "learning_rate": 3.585840184540368e-05,
      "loss": 0.9108148574829101,
      "mean_token_accuracy": 0.7942201592028141,
      "num_tokens": 50520406.0,
      "step": 25570
    },
    {
      "entropy": 0.8826533364132046,
      "epoch": 0.7327098520544806,
      "grad_norm": 7.21875,
      "learning_rate": 3.5848121624394156e-05,
      "loss": 0.8318147659301758,
      "mean_token_accuracy": 0.8102368142455816,
      "num_tokens": 50539988.0,
      "step": 25580
    },
    {
      "entropy": 0.8213028468191623,
      "epoch": 0.7329962906205691,
      "grad_norm": 8.625,
      "learning_rate": 3.583783914304102e-05,
      "loss": 0.8443984985351562,
      "mean_token_accuracy": 0.8002722475677728,
      "num_tokens": 50559507.0,
      "step": 25590
    },
    {
      "entropy": 0.8935061763972044,
      "epoch": 0.7332827291866577,
      "grad_norm": 5.96875,
      "learning_rate": 3.582755440348676e-05,
      "loss": 0.9767901420593261,
      "mean_token_accuracy": 0.7856005642563104,
      "num_tokens": 50580096.0,
      "step": 25600
    },
    {
      "entropy": 0.8468574969097972,
      "epoch": 0.7335691677527463,
      "grad_norm": 7.6875,
      "learning_rate": 3.581726740787434e-05,
      "loss": 0.8435724258422852,
      "mean_token_accuracy": 0.803973813354969,
      "num_tokens": 50598573.0,
      "step": 25610
    },
    {
      "entropy": 0.9201878350228071,
      "epoch": 0.7338556063188347,
      "grad_norm": 8.6875,
      "learning_rate": 3.580697815834718e-05,
      "loss": 0.9392012596130371,
      "mean_token_accuracy": 0.7843753419816494,
      "num_tokens": 50619466.0,
      "step": 25620
    },
    {
      "entropy": 0.8217475406825543,
      "epoch": 0.7341420448849233,
      "grad_norm": 8.0625,
      "learning_rate": 3.579668665704917e-05,
      "loss": 0.8090158462524414,
      "mean_token_accuracy": 0.8104457199573517,
      "num_tokens": 50638829.0,
      "step": 25630
    },
    {
      "entropy": 0.8251393364742399,
      "epoch": 0.7344284834510119,
      "grad_norm": 8.9375,
      "learning_rate": 3.5786392906124674e-05,
      "loss": 0.7742052555084229,
      "mean_token_accuracy": 0.8107036590576172,
      "num_tokens": 50657096.0,
      "step": 25640
    },
    {
      "entropy": 0.7661686180159449,
      "epoch": 0.7347149220171004,
      "grad_norm": 9.625,
      "learning_rate": 3.577609690771855e-05,
      "loss": 0.8344218254089355,
      "mean_token_accuracy": 0.8112305730581284,
      "num_tokens": 50675862.0,
      "step": 25650
    },
    {
      "entropy": 0.8311743228696287,
      "epoch": 0.7350013605831889,
      "grad_norm": 8.375,
      "learning_rate": 3.576579866397606e-05,
      "loss": 0.829670524597168,
      "mean_token_accuracy": 0.8036914974451065,
      "num_tokens": 50695614.0,
      "step": 25660
    },
    {
      "entropy": 0.8535969400778413,
      "epoch": 0.7352877991492774,
      "grad_norm": 7.15625,
      "learning_rate": 3.5755498177043e-05,
      "loss": 0.8303332328796387,
      "mean_token_accuracy": 0.8072097986936569,
      "num_tokens": 50715884.0,
      "step": 25670
    },
    {
      "entropy": 0.8131112158298492,
      "epoch": 0.735574237715366,
      "grad_norm": 10.4375,
      "learning_rate": 3.574519544906561e-05,
      "loss": 0.8319113731384278,
      "mean_token_accuracy": 0.7982664216309786,
      "num_tokens": 50735403.0,
      "step": 25680
    },
    {
      "entropy": 0.8632119562476873,
      "epoch": 0.7358606762814546,
      "grad_norm": 6.46875,
      "learning_rate": 3.5734890482190576e-05,
      "loss": 0.8611366271972656,
      "mean_token_accuracy": 0.8116211704909801,
      "num_tokens": 50756019.0,
      "step": 25690
    },
    {
      "entropy": 0.8526527509093285,
      "epoch": 0.736147114847543,
      "grad_norm": 8.3125,
      "learning_rate": 3.572458327856508e-05,
      "loss": 0.8800951957702636,
      "mean_token_accuracy": 0.7935364175587892,
      "num_tokens": 50776095.0,
      "step": 25700
    },
    {
      "entropy": 0.8133847190067172,
      "epoch": 0.7364335534136316,
      "grad_norm": 8.375,
      "learning_rate": 3.5714273840336756e-05,
      "loss": 0.8385985374450684,
      "mean_token_accuracy": 0.8041869215667248,
      "num_tokens": 50795854.0,
      "step": 25710
    },
    {
      "entropy": 0.868499792739749,
      "epoch": 0.7367199919797202,
      "grad_norm": 8.875,
      "learning_rate": 3.570396216965372e-05,
      "loss": 0.8760632514953614,
      "mean_token_accuracy": 0.8023925203830004,
      "num_tokens": 50817022.0,
      "step": 25720
    },
    {
      "entropy": 0.8643961064517498,
      "epoch": 0.7370064305458087,
      "grad_norm": 7.40625,
      "learning_rate": 3.56936482686645e-05,
      "loss": 0.8781773567199707,
      "mean_token_accuracy": 0.8054949671030045,
      "num_tokens": 50835420.0,
      "step": 25730
    },
    {
      "entropy": 0.820056433416903,
      "epoch": 0.7372928691118972,
      "grad_norm": 8.75,
      "learning_rate": 3.568333213951816e-05,
      "loss": 0.8661311149597168,
      "mean_token_accuracy": 0.8077930971980095,
      "num_tokens": 50854304.0,
      "step": 25740
    },
    {
      "entropy": 0.8688059527426958,
      "epoch": 0.7375793076779857,
      "grad_norm": 8.8125,
      "learning_rate": 3.567301378436421e-05,
      "loss": 0.939057445526123,
      "mean_token_accuracy": 0.7895231589674949,
      "num_tokens": 50874263.0,
      "step": 25750
    },
    {
      "entropy": 0.7764919400215149,
      "epoch": 0.7378657462440743,
      "grad_norm": 7.3125,
      "learning_rate": 3.566269320535257e-05,
      "loss": 0.771301555633545,
      "mean_token_accuracy": 0.8211447533220053,
      "num_tokens": 50894763.0,
      "step": 25760
    },
    {
      "entropy": 0.8642727121710777,
      "epoch": 0.7381521848101629,
      "grad_norm": 7.1875,
      "learning_rate": 3.565237040463369e-05,
      "loss": 0.8714384078979492,
      "mean_token_accuracy": 0.8084221445024014,
      "num_tokens": 50913907.0,
      "step": 25770
    },
    {
      "entropy": 0.9059589210897684,
      "epoch": 0.7384386233762513,
      "grad_norm": 8.4375,
      "learning_rate": 3.564204538435845e-05,
      "loss": 0.900432014465332,
      "mean_token_accuracy": 0.7974840853363275,
      "num_tokens": 50933128.0,
      "step": 25780
    },
    {
      "entropy": 0.8344452487304806,
      "epoch": 0.7387250619423399,
      "grad_norm": 11.375,
      "learning_rate": 3.56317181466782e-05,
      "loss": 0.8775764465332031,
      "mean_token_accuracy": 0.8043027073144913,
      "num_tokens": 50951949.0,
      "step": 25790
    },
    {
      "entropy": 0.8202037416398525,
      "epoch": 0.7390115005084285,
      "grad_norm": 8.6875,
      "learning_rate": 3.562138869374474e-05,
      "loss": 0.8269149780273437,
      "mean_token_accuracy": 0.8021810695528984,
      "num_tokens": 50970981.0,
      "step": 25800
    },
    {
      "entropy": 0.9222113093361258,
      "epoch": 0.739297939074517,
      "grad_norm": 8.1875,
      "learning_rate": 3.561105702771036e-05,
      "loss": 0.8812948226928711,
      "mean_token_accuracy": 0.8018766734749079,
      "num_tokens": 50989988.0,
      "step": 25810
    },
    {
      "entropy": 0.7829474653117359,
      "epoch": 0.7395843776406056,
      "grad_norm": 6.5625,
      "learning_rate": 3.560072315072779e-05,
      "loss": 0.7730511665344239,
      "mean_token_accuracy": 0.8189279146492481,
      "num_tokens": 51009472.0,
      "step": 25820
    },
    {
      "entropy": 0.8658244260586798,
      "epoch": 0.739870816206694,
      "grad_norm": 9.375,
      "learning_rate": 3.5590387064950204e-05,
      "loss": 0.8431542396545411,
      "mean_token_accuracy": 0.8013987984508276,
      "num_tokens": 51028566.0,
      "step": 25830
    },
    {
      "entropy": 0.8187919298186899,
      "epoch": 0.7401572547727826,
      "grad_norm": 11.9375,
      "learning_rate": 3.558004877253128e-05,
      "loss": 0.8367950439453125,
      "mean_token_accuracy": 0.8042069599032402,
      "num_tokens": 51049343.0,
      "step": 25840
    },
    {
      "entropy": 0.8362379780039191,
      "epoch": 0.7404436933388712,
      "grad_norm": 8.4375,
      "learning_rate": 3.556970827562513e-05,
      "loss": 0.806118106842041,
      "mean_token_accuracy": 0.8116401378065348,
      "num_tokens": 51068734.0,
      "step": 25850
    },
    {
      "entropy": 0.8945028910413384,
      "epoch": 0.7407301319049597,
      "grad_norm": 7.53125,
      "learning_rate": 3.555936557638632e-05,
      "loss": 0.9049418449401856,
      "mean_token_accuracy": 0.7845442332327366,
      "num_tokens": 51089233.0,
      "step": 25860
    },
    {
      "entropy": 0.8354742009192705,
      "epoch": 0.7410165704710482,
      "grad_norm": 11.5625,
      "learning_rate": 3.5549020676969876e-05,
      "loss": 0.8491250038146972,
      "mean_token_accuracy": 0.8030696041882038,
      "num_tokens": 51108354.0,
      "step": 25870
    },
    {
      "entropy": 0.769727398827672,
      "epoch": 0.7413030090371368,
      "grad_norm": 8.5625,
      "learning_rate": 3.55386735795313e-05,
      "loss": 0.7674683094024658,
      "mean_token_accuracy": 0.8235304746776819,
      "num_tokens": 51127470.0,
      "step": 25880
    },
    {
      "entropy": 0.8559348948299885,
      "epoch": 0.7415894476032253,
      "grad_norm": 6.3125,
      "learning_rate": 3.552832428622653e-05,
      "loss": 0.8524891853332519,
      "mean_token_accuracy": 0.8069794569164515,
      "num_tokens": 51146204.0,
      "step": 25890
    },
    {
      "entropy": 0.7955260834656656,
      "epoch": 0.7418758861693139,
      "grad_norm": 5.78125,
      "learning_rate": 3.5517972799212e-05,
      "loss": 0.788578462600708,
      "mean_token_accuracy": 0.8154920611530543,
      "num_tokens": 51166625.0,
      "step": 25900
    },
    {
      "entropy": 0.8393904096446931,
      "epoch": 0.7421623247354023,
      "grad_norm": 8.75,
      "learning_rate": 3.550761912064454e-05,
      "loss": 0.8390960693359375,
      "mean_token_accuracy": 0.8166884388774633,
      "num_tokens": 51188297.0,
      "step": 25910
    },
    {
      "entropy": 0.8153709162026643,
      "epoch": 0.7424487633014909,
      "grad_norm": 12.6875,
      "learning_rate": 3.549726325268149e-05,
      "loss": 0.828587818145752,
      "mean_token_accuracy": 0.8129624269902707,
      "num_tokens": 51209466.0,
      "step": 25920
    },
    {
      "entropy": 0.9046401615254581,
      "epoch": 0.7427352018675795,
      "grad_norm": 7.3125,
      "learning_rate": 3.548690519748062e-05,
      "loss": 0.977498722076416,
      "mean_token_accuracy": 0.7834621861577034,
      "num_tokens": 51231405.0,
      "step": 25930
    },
    {
      "entropy": 0.7820390421897173,
      "epoch": 0.743021640433668,
      "grad_norm": 7.125,
      "learning_rate": 3.547654495720015e-05,
      "loss": 0.8079113006591797,
      "mean_token_accuracy": 0.8138581451028586,
      "num_tokens": 51250019.0,
      "step": 25940
    },
    {
      "entropy": 1.0124660728499293,
      "epoch": 0.7433080789997565,
      "grad_norm": 10.25,
      "learning_rate": 3.5466182533998796e-05,
      "loss": 1.0319920539855958,
      "mean_token_accuracy": 0.7666232988238335,
      "num_tokens": 51270693.0,
      "step": 25950
    },
    {
      "entropy": 0.9429703708738089,
      "epoch": 0.743594517565845,
      "grad_norm": 9.5,
      "learning_rate": 3.545581793003567e-05,
      "loss": 0.9450200080871582,
      "mean_token_accuracy": 0.7808747459203005,
      "num_tokens": 51292151.0,
      "step": 25960
    },
    {
      "entropy": 0.824187702126801,
      "epoch": 0.7438809561319336,
      "grad_norm": 9.0625,
      "learning_rate": 3.54454511474704e-05,
      "loss": 0.8381131172180176,
      "mean_token_accuracy": 0.8006341129541397,
      "num_tokens": 51311910.0,
      "step": 25970
    },
    {
      "entropy": 0.8678090825676918,
      "epoch": 0.7441673946980222,
      "grad_norm": 8.25,
      "learning_rate": 3.5435082188463e-05,
      "loss": 0.8155076980590821,
      "mean_token_accuracy": 0.8131590910255909,
      "num_tokens": 51330426.0,
      "step": 25980
    },
    {
      "entropy": 0.861419098265469,
      "epoch": 0.7444538332641106,
      "grad_norm": 7.8125,
      "learning_rate": 3.5424711055173995e-05,
      "loss": 0.9492282867431641,
      "mean_token_accuracy": 0.7819139890372753,
      "num_tokens": 51350045.0,
      "step": 25990
    },
    {
      "entropy": 0.8569882918149233,
      "epoch": 0.7447402718301992,
      "grad_norm": 7.59375,
      "learning_rate": 3.541433774976435e-05,
      "loss": 0.8695609092712402,
      "mean_token_accuracy": 0.7995440557599067,
      "num_tokens": 51369696.0,
      "step": 26000
    },
    {
      "entropy": 0.873808422870934,
      "epoch": 0.7450267103962878,
      "grad_norm": 11.125,
      "learning_rate": 3.540396227439546e-05,
      "loss": 0.8560063362121582,
      "mean_token_accuracy": 0.8067202430218459,
      "num_tokens": 51388261.0,
      "step": 26010
    },
    {
      "entropy": 0.8037474295124412,
      "epoch": 0.7453131489623763,
      "grad_norm": 9.0,
      "learning_rate": 3.5393584631229195e-05,
      "loss": 0.7332080364227295,
      "mean_token_accuracy": 0.8144133012741804,
      "num_tokens": 51407096.0,
      "step": 26020
    },
    {
      "entropy": 0.8231912013143301,
      "epoch": 0.7455995875284648,
      "grad_norm": 7.625,
      "learning_rate": 3.538320482242786e-05,
      "loss": 0.8327185630798339,
      "mean_token_accuracy": 0.8151165079325438,
      "num_tokens": 51426906.0,
      "step": 26030
    },
    {
      "entropy": 0.7550816997885704,
      "epoch": 0.7458860260945533,
      "grad_norm": 8.0,
      "learning_rate": 3.537282285015422e-05,
      "loss": 0.8356232643127441,
      "mean_token_accuracy": 0.811156802624464,
      "num_tokens": 51445662.0,
      "step": 26040
    },
    {
      "entropy": 0.8293792847078294,
      "epoch": 0.7461724646606419,
      "grad_norm": 10.0625,
      "learning_rate": 3.5362438716571506e-05,
      "loss": 0.8377252578735351,
      "mean_token_accuracy": 0.8096207238733768,
      "num_tokens": 51465726.0,
      "step": 26050
    },
    {
      "entropy": 0.8844648484140635,
      "epoch": 0.7464589032267305,
      "grad_norm": 8.9375,
      "learning_rate": 3.5352052423843366e-05,
      "loss": 0.8894292831420898,
      "mean_token_accuracy": 0.7933140449225903,
      "num_tokens": 51485857.0,
      "step": 26060
    },
    {
      "entropy": 0.8335542250424623,
      "epoch": 0.746745341792819,
      "grad_norm": 8.8125,
      "learning_rate": 3.534166397413393e-05,
      "loss": 0.8875211715698242,
      "mean_token_accuracy": 0.7995186943560839,
      "num_tokens": 51506422.0,
      "step": 26070
    },
    {
      "entropy": 0.8407507000491024,
      "epoch": 0.7470317803589075,
      "grad_norm": 6.65625,
      "learning_rate": 3.5331273369607744e-05,
      "loss": 0.870060920715332,
      "mean_token_accuracy": 0.7967653557658195,
      "num_tokens": 51525369.0,
      "step": 26080
    },
    {
      "entropy": 0.8357246555387974,
      "epoch": 0.7473182189249961,
      "grad_norm": 7.71875,
      "learning_rate": 3.532088061242983e-05,
      "loss": 0.8532617568969727,
      "mean_token_accuracy": 0.7999270983040333,
      "num_tokens": 51544727.0,
      "step": 26090
    },
    {
      "entropy": 0.7833263717591763,
      "epoch": 0.7476046574910846,
      "grad_norm": 9.625,
      "learning_rate": 3.531048570476567e-05,
      "loss": 0.7686203479766845,
      "mean_token_accuracy": 0.8179000586271286,
      "num_tokens": 51563519.0,
      "step": 26100
    },
    {
      "entropy": 0.8685168832540512,
      "epoch": 0.7478910960571732,
      "grad_norm": 6.96875,
      "learning_rate": 3.530008864878115e-05,
      "loss": 0.8958255767822265,
      "mean_token_accuracy": 0.7918981589376927,
      "num_tokens": 51581897.0,
      "step": 26110
    },
    {
      "entropy": 0.8774122266098857,
      "epoch": 0.7481775346232616,
      "grad_norm": 6.90625,
      "learning_rate": 3.528968944664264e-05,
      "loss": 0.9404688835144043,
      "mean_token_accuracy": 0.7875281944870949,
      "num_tokens": 51600854.0,
      "step": 26120
    },
    {
      "entropy": 0.8331232646480202,
      "epoch": 0.7484639731893502,
      "grad_norm": 6.4375,
      "learning_rate": 3.527928810051695e-05,
      "loss": 0.8630148887634277,
      "mean_token_accuracy": 0.7937082324177027,
      "num_tokens": 51621362.0,
      "step": 26130
    },
    {
      "entropy": 0.833460976742208,
      "epoch": 0.7487504117554388,
      "grad_norm": 5.71875,
      "learning_rate": 3.5268884612571314e-05,
      "loss": 0.7642493724822998,
      "mean_token_accuracy": 0.8224336314946413,
      "num_tokens": 51640498.0,
      "step": 26140
    },
    {
      "entropy": 0.8090186337009072,
      "epoch": 0.7490368503215273,
      "grad_norm": 8.8125,
      "learning_rate": 3.5258478984973456e-05,
      "loss": 0.7941810607910156,
      "mean_token_accuracy": 0.8121600780636072,
      "num_tokens": 51660292.0,
      "step": 26150
    },
    {
      "entropy": 0.879202481219545,
      "epoch": 0.7493232888876158,
      "grad_norm": 5.9375,
      "learning_rate": 3.524807121989149e-05,
      "loss": 0.8528727531433106,
      "mean_token_accuracy": 0.8112926602363586,
      "num_tokens": 51681331.0,
      "step": 26160
    },
    {
      "entropy": 0.7680205410346389,
      "epoch": 0.7496097274537044,
      "grad_norm": 7.71875,
      "learning_rate": 3.523766131949403e-05,
      "loss": 0.76336669921875,
      "mean_token_accuracy": 0.8126701395958662,
      "num_tokens": 51700486.0,
      "step": 26170
    },
    {
      "entropy": 0.8317784878425301,
      "epoch": 0.7498961660197929,
      "grad_norm": 8.4375,
      "learning_rate": 3.52272492859501e-05,
      "loss": 0.7898430824279785,
      "mean_token_accuracy": 0.8107345547527075,
      "num_tokens": 51721362.0,
      "step": 26180
    },
    {
      "entropy": 0.8822249639779329,
      "epoch": 0.7501826045858815,
      "grad_norm": 13.6875,
      "learning_rate": 3.521683512142918e-05,
      "loss": 0.9336329460144043,
      "mean_token_accuracy": 0.7861962988972664,
      "num_tokens": 51740799.0,
      "step": 26190
    },
    {
      "entropy": 0.7635739616118371,
      "epoch": 0.7504690431519699,
      "grad_norm": 5.875,
      "learning_rate": 3.5206418828101197e-05,
      "loss": 0.7606573581695557,
      "mean_token_accuracy": 0.81234714910388,
      "num_tokens": 51759763.0,
      "step": 26200
    },
    {
      "entropy": 0.7891520591452718,
      "epoch": 0.7507554817180585,
      "grad_norm": 7.46875,
      "learning_rate": 3.5196000408136524e-05,
      "loss": 0.8098342895507813,
      "mean_token_accuracy": 0.8095672111958265,
      "num_tokens": 51780483.0,
      "step": 26210
    },
    {
      "entropy": 0.8234855726361274,
      "epoch": 0.7510419202841471,
      "grad_norm": 6.34375,
      "learning_rate": 3.5185579863705944e-05,
      "loss": 0.8616086959838867,
      "mean_token_accuracy": 0.8050942860543728,
      "num_tokens": 51799789.0,
      "step": 26220
    },
    {
      "entropy": 0.8246013527736068,
      "epoch": 0.7513283588502356,
      "grad_norm": 7.71875,
      "learning_rate": 3.517515719698074e-05,
      "loss": 0.8544775962829589,
      "mean_token_accuracy": 0.7987392358481884,
      "num_tokens": 51818793.0,
      "step": 26230
    },
    {
      "entropy": 0.8523023383691907,
      "epoch": 0.7516147974163241,
      "grad_norm": 6.5625,
      "learning_rate": 3.5164732410132594e-05,
      "loss": 0.864194393157959,
      "mean_token_accuracy": 0.7897499937564134,
      "num_tokens": 51839659.0,
      "step": 26240
    },
    {
      "entropy": 0.8035355418920517,
      "epoch": 0.7519012359824127,
      "grad_norm": 7.6875,
      "learning_rate": 3.5154305505333646e-05,
      "loss": 0.7928609848022461,
      "mean_token_accuracy": 0.817879532277584,
      "num_tokens": 51860235.0,
      "step": 26250
    },
    {
      "entropy": 0.8394896944984793,
      "epoch": 0.7521876745485012,
      "grad_norm": 8.625,
      "learning_rate": 3.514387648475646e-05,
      "loss": 0.8573348999023438,
      "mean_token_accuracy": 0.795827804505825,
      "num_tokens": 51879907.0,
      "step": 26260
    },
    {
      "entropy": 0.824612878356129,
      "epoch": 0.7524741131145898,
      "grad_norm": 11.4375,
      "learning_rate": 3.513344535057408e-05,
      "loss": 0.7836009502410889,
      "mean_token_accuracy": 0.8175262290984392,
      "num_tokens": 51899277.0,
      "step": 26270
    },
    {
      "entropy": 0.8277591485530138,
      "epoch": 0.7527605516806782,
      "grad_norm": 5.0,
      "learning_rate": 3.512301210495994e-05,
      "loss": 0.8160225868225097,
      "mean_token_accuracy": 0.8197535883635283,
      "num_tokens": 51917700.0,
      "step": 26280
    },
    {
      "entropy": 0.7451574834994972,
      "epoch": 0.7530469902467668,
      "grad_norm": 7.78125,
      "learning_rate": 3.511257675008795e-05,
      "loss": 0.753144645690918,
      "mean_token_accuracy": 0.8260019727051258,
      "num_tokens": 51935260.0,
      "step": 26290
    },
    {
      "entropy": 0.8926470294594765,
      "epoch": 0.7533334288128554,
      "grad_norm": 10.6875,
      "learning_rate": 3.510213928813245e-05,
      "loss": 0.8888670921325683,
      "mean_token_accuracy": 0.797199783846736,
      "num_tokens": 51954918.0,
      "step": 26300
    },
    {
      "entropy": 0.9743563750758767,
      "epoch": 0.7536198673789439,
      "grad_norm": 8.0,
      "learning_rate": 3.509169972126822e-05,
      "loss": 0.98086576461792,
      "mean_token_accuracy": 0.7941437188535929,
      "num_tokens": 51974465.0,
      "step": 26310
    },
    {
      "entropy": 0.8561498478055001,
      "epoch": 0.7539063059450324,
      "grad_norm": 9.3125,
      "learning_rate": 3.5081258051670464e-05,
      "loss": 0.8359295845031738,
      "mean_token_accuracy": 0.8092184867709875,
      "num_tokens": 51994574.0,
      "step": 26320
    },
    {
      "entropy": 0.7877449616789818,
      "epoch": 0.754192744511121,
      "grad_norm": 8.375,
      "learning_rate": 3.5070814281514844e-05,
      "loss": 0.8276229858398437,
      "mean_token_accuracy": 0.8102397363632917,
      "num_tokens": 52014239.0,
      "step": 26330
    },
    {
      "entropy": 0.8564531905576587,
      "epoch": 0.7544791830772095,
      "grad_norm": 9.8125,
      "learning_rate": 3.506036841297745e-05,
      "loss": 0.8242484092712402,
      "mean_token_accuracy": 0.8077566057443619,
      "num_tokens": 52032999.0,
      "step": 26340
    },
    {
      "entropy": 0.7600436992943287,
      "epoch": 0.7547656216432981,
      "grad_norm": 11.0625,
      "learning_rate": 3.504992044823482e-05,
      "loss": 0.7518444061279297,
      "mean_token_accuracy": 0.8198941953480243,
      "num_tokens": 52052554.0,
      "step": 26350
    },
    {
      "entropy": 0.726802156213671,
      "epoch": 0.7550520602093866,
      "grad_norm": 10.125,
      "learning_rate": 3.50394703894639e-05,
      "loss": 0.7423184394836426,
      "mean_token_accuracy": 0.8298457097262144,
      "num_tokens": 52071545.0,
      "step": 26360
    },
    {
      "entropy": 0.7876180458813906,
      "epoch": 0.7553384987754751,
      "grad_norm": 5.78125,
      "learning_rate": 3.502901823884212e-05,
      "loss": 0.8596455574035644,
      "mean_token_accuracy": 0.7963002383708954,
      "num_tokens": 52090889.0,
      "step": 26370
    },
    {
      "entropy": 0.785641991533339,
      "epoch": 0.7556249373415637,
      "grad_norm": 8.625,
      "learning_rate": 3.5018563998547296e-05,
      "loss": 0.7831881523132325,
      "mean_token_accuracy": 0.8133076779544354,
      "num_tokens": 52109118.0,
      "step": 26380
    },
    {
      "entropy": 0.879093511775136,
      "epoch": 0.7559113759076522,
      "grad_norm": 5.71875,
      "learning_rate": 3.500810767075769e-05,
      "loss": 0.8494023323059082,
      "mean_token_accuracy": 0.8061054971069097,
      "num_tokens": 52127961.0,
      "step": 26390
    },
    {
      "entropy": 0.9178974408656358,
      "epoch": 0.7561978144737408,
      "grad_norm": 9.5625,
      "learning_rate": 3.499764925765206e-05,
      "loss": 0.9461101531982422,
      "mean_token_accuracy": 0.7897113401442766,
      "num_tokens": 52148327.0,
      "step": 26400
    },
    {
      "entropy": 0.8440957156009972,
      "epoch": 0.7564842530398292,
      "grad_norm": 7.28125,
      "learning_rate": 3.4987188761409493e-05,
      "loss": 0.859461784362793,
      "mean_token_accuracy": 0.8005023315548897,
      "num_tokens": 52167939.0,
      "step": 26410
    },
    {
      "entropy": 0.8308511272072792,
      "epoch": 0.7567706916059178,
      "grad_norm": 8.1875,
      "learning_rate": 3.49767261842096e-05,
      "loss": 0.8349628448486328,
      "mean_token_accuracy": 0.7993259865790605,
      "num_tokens": 52186616.0,
      "step": 26420
    },
    {
      "entropy": 0.8234991502016783,
      "epoch": 0.7570571301720064,
      "grad_norm": 9.25,
      "learning_rate": 3.496626152823238e-05,
      "loss": 0.8001429557800293,
      "mean_token_accuracy": 0.8214220400899649,
      "num_tokens": 52205639.0,
      "step": 26430
    },
    {
      "entropy": 0.8433365033939481,
      "epoch": 0.7573435687380949,
      "grad_norm": 10.375,
      "learning_rate": 3.495579479565828e-05,
      "loss": 0.8318280220031739,
      "mean_token_accuracy": 0.8027717597782612,
      "num_tokens": 52225235.0,
      "step": 26440
    },
    {
      "entropy": 0.8551079725846649,
      "epoch": 0.7576300073041834,
      "grad_norm": 10.8125,
      "learning_rate": 3.494532598866817e-05,
      "loss": 0.8449461936950684,
      "mean_token_accuracy": 0.8136380106210709,
      "num_tokens": 52245031.0,
      "step": 26450
    },
    {
      "entropy": 0.768262859620154,
      "epoch": 0.757916445870272,
      "grad_norm": 6.9375,
      "learning_rate": 3.493485510944337e-05,
      "loss": 0.7951266288757324,
      "mean_token_accuracy": 0.8070656560361386,
      "num_tokens": 52265163.0,
      "step": 26460
    },
    {
      "entropy": 0.859142622537911,
      "epoch": 0.7582028844363605,
      "grad_norm": 9.0625,
      "learning_rate": 3.492438216016561e-05,
      "loss": 0.8857991218566894,
      "mean_token_accuracy": 0.7953007224947214,
      "num_tokens": 52285658.0,
      "step": 26470
    },
    {
      "entropy": 0.7794176394119858,
      "epoch": 0.7584893230024491,
      "grad_norm": 8.6875,
      "learning_rate": 3.4913907143017065e-05,
      "loss": 0.7602837562561036,
      "mean_token_accuracy": 0.8252001650631428,
      "num_tokens": 52303837.0,
      "step": 26480
    },
    {
      "entropy": 0.8041056545451284,
      "epoch": 0.7587757615685375,
      "grad_norm": 7.875,
      "learning_rate": 3.490343006018034e-05,
      "loss": 0.8547499656677247,
      "mean_token_accuracy": 0.8084862355142832,
      "num_tokens": 52324898.0,
      "step": 26490
    },
    {
      "entropy": 0.7510609744116664,
      "epoch": 0.7590622001346261,
      "grad_norm": 5.875,
      "learning_rate": 3.4892950913838466e-05,
      "loss": 0.7699031829833984,
      "mean_token_accuracy": 0.8215194433927536,
      "num_tokens": 52343740.0,
      "step": 26500
    },
    {
      "entropy": 0.8653112836182117,
      "epoch": 0.7593486387007147,
      "grad_norm": 8.125,
      "learning_rate": 3.4882469706174904e-05,
      "loss": 0.9233150482177734,
      "mean_token_accuracy": 0.7936650454998017,
      "num_tokens": 52362862.0,
      "step": 26510
    },
    {
      "entropy": 0.8440946815535426,
      "epoch": 0.7596350772668032,
      "grad_norm": 6.75,
      "learning_rate": 3.487198643937354e-05,
      "loss": 0.8550547599792481,
      "mean_token_accuracy": 0.7979623310267925,
      "num_tokens": 52381186.0,
      "step": 26520
    },
    {
      "entropy": 0.8838265772908926,
      "epoch": 0.7599215158328917,
      "grad_norm": 8.875,
      "learning_rate": 3.4861501115618724e-05,
      "loss": 0.9206209182739258,
      "mean_token_accuracy": 0.7834353622049093,
      "num_tokens": 52401555.0,
      "step": 26530
    },
    {
      "entropy": 0.8420044707134366,
      "epoch": 0.7602079543989803,
      "grad_norm": 7.375,
      "learning_rate": 3.485101373709516e-05,
      "loss": 0.8102311134338379,
      "mean_token_accuracy": 0.8176534339785576,
      "num_tokens": 52421708.0,
      "step": 26540
    },
    {
      "entropy": 0.9073421392589808,
      "epoch": 0.7604943929650688,
      "grad_norm": 9.4375,
      "learning_rate": 3.4840524305988076e-05,
      "loss": 0.8838107109069824,
      "mean_token_accuracy": 0.8005119971930981,
      "num_tokens": 52441186.0,
      "step": 26550
    },
    {
      "entropy": 1.0016237566247583,
      "epoch": 0.7607808315311574,
      "grad_norm": 9.6875,
      "learning_rate": 3.483003282448304e-05,
      "loss": 0.9525158882141114,
      "mean_token_accuracy": 0.7797915127128363,
      "num_tokens": 52461889.0,
      "step": 26560
    },
    {
      "entropy": 0.8442256396636367,
      "epoch": 0.7610672700972458,
      "grad_norm": 6.75,
      "learning_rate": 3.4819539294766105e-05,
      "loss": 0.8680953979492188,
      "mean_token_accuracy": 0.7979808766394854,
      "num_tokens": 52482569.0,
      "step": 26570
    },
    {
      "entropy": 0.7813249445520342,
      "epoch": 0.7613537086633344,
      "grad_norm": 9.25,
      "learning_rate": 3.480904371902373e-05,
      "loss": 0.7986842632293701,
      "mean_token_accuracy": 0.8105838857591152,
      "num_tokens": 52502470.0,
      "step": 26580
    },
    {
      "entropy": 0.8098172329366207,
      "epoch": 0.761640147229423,
      "grad_norm": 8.6875,
      "learning_rate": 3.479854609944278e-05,
      "loss": 0.8059099197387696,
      "mean_token_accuracy": 0.8170627236366272,
      "num_tokens": 52521291.0,
      "step": 26590
    },
    {
      "entropy": 0.7610129810869694,
      "epoch": 0.7619265857955115,
      "grad_norm": 9.6875,
      "learning_rate": 3.478804643821059e-05,
      "loss": 0.8094117164611816,
      "mean_token_accuracy": 0.8065736886113882,
      "num_tokens": 52540702.0,
      "step": 26600
    },
    {
      "entropy": 0.769207671098411,
      "epoch": 0.7622130243616001,
      "grad_norm": 8.6875,
      "learning_rate": 3.477754473751489e-05,
      "loss": 0.7706964015960693,
      "mean_token_accuracy": 0.8162071980535984,
      "num_tokens": 52560380.0,
      "step": 26610
    },
    {
      "entropy": 0.8744391903281212,
      "epoch": 0.7624994629276886,
      "grad_norm": 6.9375,
      "learning_rate": 3.476704099954385e-05,
      "loss": 0.8761724472045899,
      "mean_token_accuracy": 0.79612399302423,
      "num_tokens": 52580178.0,
      "step": 26620
    },
    {
      "entropy": 0.8036561131477356,
      "epoch": 0.7627859014937771,
      "grad_norm": 8.375,
      "learning_rate": 3.475653522648604e-05,
      "loss": 0.7980080604553222,
      "mean_token_accuracy": 0.8018644016236067,
      "num_tokens": 52598871.0,
      "step": 26630
    },
    {
      "entropy": 0.8312596496194601,
      "epoch": 0.7630723400598657,
      "grad_norm": 7.5625,
      "learning_rate": 3.474602742053049e-05,
      "loss": 0.855492115020752,
      "mean_token_accuracy": 0.7996783971786499,
      "num_tokens": 52618660.0,
      "step": 26640
    },
    {
      "entropy": 0.7979629758745432,
      "epoch": 0.7633587786259542,
      "grad_norm": 5.34375,
      "learning_rate": 3.473551758386663e-05,
      "loss": 0.8056489944458007,
      "mean_token_accuracy": 0.8090913765132427,
      "num_tokens": 52638310.0,
      "step": 26650
    },
    {
      "entropy": 0.7865894125774503,
      "epoch": 0.7636452171920427,
      "grad_norm": 9.25,
      "learning_rate": 3.47250057186843e-05,
      "loss": 0.7554585933685303,
      "mean_token_accuracy": 0.8267163578420877,
      "num_tokens": 52656953.0,
      "step": 26660
    },
    {
      "entropy": 0.8006116697564721,
      "epoch": 0.7639316557581313,
      "grad_norm": 8.4375,
      "learning_rate": 3.47144918271738e-05,
      "loss": 0.8059024810791016,
      "mean_token_accuracy": 0.8084939509630203,
      "num_tokens": 52677334.0,
      "step": 26670
    },
    {
      "entropy": 0.8334105296060443,
      "epoch": 0.7642180943242198,
      "grad_norm": 8.125,
      "learning_rate": 3.470397591152582e-05,
      "loss": 0.8132050514221192,
      "mean_token_accuracy": 0.8097854886204005,
      "num_tokens": 52697671.0,
      "step": 26680
    },
    {
      "entropy": 0.7812070894986392,
      "epoch": 0.7645045328903084,
      "grad_norm": 8.8125,
      "learning_rate": 3.469345797393149e-05,
      "loss": 0.8130982398986817,
      "mean_token_accuracy": 0.8082083892077208,
      "num_tokens": 52717148.0,
      "step": 26690
    },
    {
      "entropy": 0.7898033026605844,
      "epoch": 0.7647909714563969,
      "grad_norm": 7.4375,
      "learning_rate": 3.468293801658237e-05,
      "loss": 0.7881356716156006,
      "mean_token_accuracy": 0.8161711033433676,
      "num_tokens": 52736480.0,
      "step": 26700
    },
    {
      "entropy": 0.8968030896969139,
      "epoch": 0.7650774100224854,
      "grad_norm": 7.96875,
      "learning_rate": 3.4672416041670396e-05,
      "loss": 0.8516274452209472,
      "mean_token_accuracy": 0.8137021400034428,
      "num_tokens": 52756083.0,
      "step": 26710
    },
    {
      "entropy": 0.7484582712873816,
      "epoch": 0.765363848588574,
      "grad_norm": 9.1875,
      "learning_rate": 3.466189205138798e-05,
      "loss": 0.8066902160644531,
      "mean_token_accuracy": 0.8175411507487297,
      "num_tokens": 52774817.0,
      "step": 26720
    },
    {
      "entropy": 0.8153332147747279,
      "epoch": 0.7656502871546625,
      "grad_norm": 8.1875,
      "learning_rate": 3.4651366047927915e-05,
      "loss": 0.8128476142883301,
      "mean_token_accuracy": 0.8207752674818038,
      "num_tokens": 52794232.0,
      "step": 26730
    },
    {
      "entropy": 0.8779219122603535,
      "epoch": 0.765936725720751,
      "grad_norm": 9.0625,
      "learning_rate": 3.4640838033483435e-05,
      "loss": 0.9276914596557617,
      "mean_token_accuracy": 0.7996852677315474,
      "num_tokens": 52813386.0,
      "step": 26740
    },
    {
      "entropy": 0.9061706382781267,
      "epoch": 0.7662231642868396,
      "grad_norm": 5.59375,
      "learning_rate": 3.4630308010248185e-05,
      "loss": 0.8966855049133301,
      "mean_token_accuracy": 0.7990551006048918,
      "num_tokens": 52832265.0,
      "step": 26750
    },
    {
      "entropy": 0.7319659563712776,
      "epoch": 0.7665096028529281,
      "grad_norm": 6.625,
      "learning_rate": 3.4619775980416217e-05,
      "loss": 0.7310292720794678,
      "mean_token_accuracy": 0.8259186536073685,
      "num_tokens": 52851990.0,
      "step": 26760
    },
    {
      "entropy": 0.7539542730897665,
      "epoch": 0.7667960414190167,
      "grad_norm": 9.5625,
      "learning_rate": 3.460924194618203e-05,
      "loss": 0.7357964515686035,
      "mean_token_accuracy": 0.8321579232811928,
      "num_tokens": 52870064.0,
      "step": 26770
    },
    {
      "entropy": 0.9479546038433909,
      "epoch": 0.7670824799851051,
      "grad_norm": 7.40625,
      "learning_rate": 3.4598705909740516e-05,
      "loss": 0.9756563186645508,
      "mean_token_accuracy": 0.777190325781703,
      "num_tokens": 52890872.0,
      "step": 26780
    },
    {
      "entropy": 0.7963769463822246,
      "epoch": 0.7673689185511937,
      "grad_norm": 9.5625,
      "learning_rate": 3.458816787328698e-05,
      "loss": 0.84909029006958,
      "mean_token_accuracy": 0.7991176892071963,
      "num_tokens": 52909015.0,
      "step": 26790
    },
    {
      "entropy": 0.7373009636998177,
      "epoch": 0.7676553571172823,
      "grad_norm": 8.875,
      "learning_rate": 3.457762783901718e-05,
      "loss": 0.7288544654846192,
      "mean_token_accuracy": 0.8258714508265257,
      "num_tokens": 52927349.0,
      "step": 26800
    },
    {
      "entropy": 0.8150943864136935,
      "epoch": 0.7679417956833708,
      "grad_norm": 8.875,
      "learning_rate": 3.456708580912725e-05,
      "loss": 0.8595602989196778,
      "mean_token_accuracy": 0.797896345332265,
      "num_tokens": 52947370.0,
      "step": 26810
    },
    {
      "entropy": 0.70361780077219,
      "epoch": 0.7682282342494593,
      "grad_norm": 5.9375,
      "learning_rate": 3.455654178581375e-05,
      "loss": 0.6710364818572998,
      "mean_token_accuracy": 0.8373038295656443,
      "num_tokens": 52966777.0,
      "step": 26820
    },
    {
      "entropy": 0.8199217488989234,
      "epoch": 0.7685146728155479,
      "grad_norm": 6.40625,
      "learning_rate": 3.4545995771273674e-05,
      "loss": 0.83297119140625,
      "mean_token_accuracy": 0.8009786643087864,
      "num_tokens": 52985897.0,
      "step": 26830
    },
    {
      "entropy": 0.753619728423655,
      "epoch": 0.7688011113816364,
      "grad_norm": 7.75,
      "learning_rate": 3.453544776770441e-05,
      "loss": 0.7200868606567383,
      "mean_token_accuracy": 0.83602180108428,
      "num_tokens": 53004188.0,
      "step": 26840
    },
    {
      "entropy": 0.933182941749692,
      "epoch": 0.769087549947725,
      "grad_norm": 9.4375,
      "learning_rate": 3.452489777730377e-05,
      "loss": 0.9220870971679688,
      "mean_token_accuracy": 0.7968029022216797,
      "num_tokens": 53023339.0,
      "step": 26850
    },
    {
      "entropy": 0.8836947565898299,
      "epoch": 0.7693739885138134,
      "grad_norm": 9.6875,
      "learning_rate": 3.4514345802269986e-05,
      "loss": 0.8659859657287597,
      "mean_token_accuracy": 0.8050314284861088,
      "num_tokens": 53043522.0,
      "step": 26860
    },
    {
      "entropy": 0.8281948423013091,
      "epoch": 0.769660427079902,
      "grad_norm": 11.125,
      "learning_rate": 3.450379184480168e-05,
      "loss": 0.842377758026123,
      "mean_token_accuracy": 0.7976966060698032,
      "num_tokens": 53063076.0,
      "step": 26870
    },
    {
      "entropy": 0.8752348531968892,
      "epoch": 0.7699468656459906,
      "grad_norm": 8.25,
      "learning_rate": 3.4493235907097916e-05,
      "loss": 0.914576244354248,
      "mean_token_accuracy": 0.7861531101167202,
      "num_tokens": 53083699.0,
      "step": 26880
    },
    {
      "entropy": 0.8393763355910778,
      "epoch": 0.7702333042120791,
      "grad_norm": 7.8125,
      "learning_rate": 3.448267799135815e-05,
      "loss": 0.842592716217041,
      "mean_token_accuracy": 0.8053008638322353,
      "num_tokens": 53103912.0,
      "step": 26890
    },
    {
      "entropy": 0.729134879168123,
      "epoch": 0.7705197427781677,
      "grad_norm": 8.0625,
      "learning_rate": 3.4472118099782264e-05,
      "loss": 0.6856610298156738,
      "mean_token_accuracy": 0.835609806701541,
      "num_tokens": 53122620.0,
      "step": 26900
    },
    {
      "entropy": 0.7276488861069084,
      "epoch": 0.7708061813442562,
      "grad_norm": 9.9375,
      "learning_rate": 3.446155623457054e-05,
      "loss": 0.7291742324829101,
      "mean_token_accuracy": 0.8268118564039468,
      "num_tokens": 53141592.0,
      "step": 26910
    },
    {
      "entropy": 0.8518563400954008,
      "epoch": 0.7710926199103447,
      "grad_norm": 9.0,
      "learning_rate": 3.4450992397923675e-05,
      "loss": 0.8903237342834472,
      "mean_token_accuracy": 0.8002647507935763,
      "num_tokens": 53160523.0,
      "step": 26920
    },
    {
      "entropy": 0.9105667233467102,
      "epoch": 0.7713790584764333,
      "grad_norm": 7.34375,
      "learning_rate": 3.444042659204279e-05,
      "loss": 0.8919118881225586,
      "mean_token_accuracy": 0.7916404575109481,
      "num_tokens": 53181456.0,
      "step": 26930
    },
    {
      "entropy": 0.778540666308254,
      "epoch": 0.7716654970425219,
      "grad_norm": 7.46875,
      "learning_rate": 3.442985881912938e-05,
      "loss": 0.792505931854248,
      "mean_token_accuracy": 0.8024680253118277,
      "num_tokens": 53201986.0,
      "step": 26940
    },
    {
      "entropy": 0.7972658818587661,
      "epoch": 0.7719519356086103,
      "grad_norm": 10.875,
      "learning_rate": 3.44192890813854e-05,
      "loss": 0.8635075569152832,
      "mean_token_accuracy": 0.7937218848615885,
      "num_tokens": 53220745.0,
      "step": 26950
    },
    {
      "entropy": 0.8693443853408098,
      "epoch": 0.7722383741746989,
      "grad_norm": 8.5,
      "learning_rate": 3.4408717381013175e-05,
      "loss": 0.8732491493225097,
      "mean_token_accuracy": 0.7991531960666179,
      "num_tokens": 53241676.0,
      "step": 26960
    },
    {
      "entropy": 0.8666753245517611,
      "epoch": 0.7725248127407874,
      "grad_norm": 6.96875,
      "learning_rate": 3.4398143720215455e-05,
      "loss": 0.850350284576416,
      "mean_token_accuracy": 0.8009880725294352,
      "num_tokens": 53261974.0,
      "step": 26970
    },
    {
      "entropy": 0.806812792737037,
      "epoch": 0.772811251306876,
      "grad_norm": 5.0625,
      "learning_rate": 3.43875681011954e-05,
      "loss": 0.799549913406372,
      "mean_token_accuracy": 0.8158707648515702,
      "num_tokens": 53281481.0,
      "step": 26980
    },
    {
      "entropy": 0.7640191270038486,
      "epoch": 0.7730976898729645,
      "grad_norm": 6.125,
      "learning_rate": 3.437699052615658e-05,
      "loss": 0.7338577270507812,
      "mean_token_accuracy": 0.8351240772753954,
      "num_tokens": 53300775.0,
      "step": 26990
    },
    {
      "entropy": 0.8337939767166972,
      "epoch": 0.773384128439053,
      "grad_norm": 9.8125,
      "learning_rate": 3.436641099730295e-05,
      "loss": 0.840492057800293,
      "mean_token_accuracy": 0.8057096224278212,
      "num_tokens": 53320063.0,
      "step": 27000
    },
    {
      "entropy": 0.8568708445876837,
      "epoch": 0.7736705670051416,
      "grad_norm": 7.3125,
      "learning_rate": 3.43558295168389e-05,
      "loss": 0.9396190643310547,
      "mean_token_accuracy": 0.7830590751022101,
      "num_tokens": 53339413.0,
      "step": 27010
    },
    {
      "entropy": 0.7883619027212262,
      "epoch": 0.7739570055712301,
      "grad_norm": 11.75,
      "learning_rate": 3.434524608696921e-05,
      "loss": 0.7952919960021972,
      "mean_token_accuracy": 0.8152478486299515,
      "num_tokens": 53358055.0,
      "step": 27020
    },
    {
      "entropy": 0.8211036130785943,
      "epoch": 0.7742434441373186,
      "grad_norm": 6.375,
      "learning_rate": 3.433466070989908e-05,
      "loss": 0.8939673423767089,
      "mean_token_accuracy": 0.7951222565025091,
      "num_tokens": 53378796.0,
      "step": 27030
    },
    {
      "entropy": 0.7845817420631647,
      "epoch": 0.7745298827034072,
      "grad_norm": 7.34375,
      "learning_rate": 3.43240733878341e-05,
      "loss": 0.7997756481170655,
      "mean_token_accuracy": 0.8088216118514537,
      "num_tokens": 53399419.0,
      "step": 27040
    },
    {
      "entropy": 0.9287640003487467,
      "epoch": 0.7748163212694957,
      "grad_norm": 8.5,
      "learning_rate": 3.431348412298028e-05,
      "loss": 0.8988548278808594,
      "mean_token_accuracy": 0.79531756862998,
      "num_tokens": 53418434.0,
      "step": 27050
    },
    {
      "entropy": 0.8726784281432629,
      "epoch": 0.7751027598355843,
      "grad_norm": 7.25,
      "learning_rate": 3.430289291754401e-05,
      "loss": 0.8501397132873535,
      "mean_token_accuracy": 0.8100136790424586,
      "num_tokens": 53441720.0,
      "step": 27060
    },
    {
      "entropy": 0.8586845025420189,
      "epoch": 0.7753891984016728,
      "grad_norm": 10.0625,
      "learning_rate": 3.429229977373212e-05,
      "loss": 0.8545910835266113,
      "mean_token_accuracy": 0.8084834858775138,
      "num_tokens": 53462944.0,
      "step": 27070
    },
    {
      "entropy": 0.7835575360804796,
      "epoch": 0.7756756369677613,
      "grad_norm": 7.21875,
      "learning_rate": 3.428170469375183e-05,
      "loss": 0.7763792514801026,
      "mean_token_accuracy": 0.8187863413244486,
      "num_tokens": 53481500.0,
      "step": 27080
    },
    {
      "entropy": 0.862365310639143,
      "epoch": 0.7759620755338499,
      "grad_norm": 9.4375,
      "learning_rate": 3.4271107679810744e-05,
      "loss": 0.9501632690429688,
      "mean_token_accuracy": 0.7812996838241816,
      "num_tokens": 53501424.0,
      "step": 27090
    },
    {
      "entropy": 0.9122387520968914,
      "epoch": 0.7762485140999384,
      "grad_norm": 7.25,
      "learning_rate": 3.4260508734116886e-05,
      "loss": 0.9594810485839844,
      "mean_token_accuracy": 0.7871096946299077,
      "num_tokens": 53521417.0,
      "step": 27100
    },
    {
      "entropy": 0.8092892481014132,
      "epoch": 0.7765349526660269,
      "grad_norm": 9.375,
      "learning_rate": 3.424990785887869e-05,
      "loss": 0.7497461318969727,
      "mean_token_accuracy": 0.8203950241208077,
      "num_tokens": 53540194.0,
      "step": 27110
    },
    {
      "entropy": 0.8705169286578893,
      "epoch": 0.7768213912321155,
      "grad_norm": 8.1875,
      "learning_rate": 3.423930505630497e-05,
      "loss": 0.8663592338562012,
      "mean_token_accuracy": 0.7992919992655516,
      "num_tokens": 53559562.0,
      "step": 27120
    },
    {
      "entropy": 0.8209005341865122,
      "epoch": 0.777107829798204,
      "grad_norm": 10.0,
      "learning_rate": 3.422870032860497e-05,
      "loss": 0.7983068943023681,
      "mean_token_accuracy": 0.8199239693582058,
      "num_tokens": 53580439.0,
      "step": 27130
    },
    {
      "entropy": 0.8295883171260356,
      "epoch": 0.7773942683642926,
      "grad_norm": 7.25,
      "learning_rate": 3.42180936779883e-05,
      "loss": 0.8422317504882812,
      "mean_token_accuracy": 0.8057965029031038,
      "num_tokens": 53598813.0,
      "step": 27140
    },
    {
      "entropy": 0.7256599813699722,
      "epoch": 0.7776807069303812,
      "grad_norm": 10.6875,
      "learning_rate": 3.420748510666501e-05,
      "loss": 0.7299038887023925,
      "mean_token_accuracy": 0.8247861757874488,
      "num_tokens": 53617965.0,
      "step": 27150
    },
    {
      "entropy": 0.8243170566856861,
      "epoch": 0.7779671454964696,
      "grad_norm": 6.9375,
      "learning_rate": 3.4196874616845516e-05,
      "loss": 0.803136157989502,
      "mean_token_accuracy": 0.8106223549693823,
      "num_tokens": 53637152.0,
      "step": 27160
    },
    {
      "entropy": 0.9652814101427794,
      "epoch": 0.7782535840625582,
      "grad_norm": 6.25,
      "learning_rate": 3.418626221074065e-05,
      "loss": 0.9970827102661133,
      "mean_token_accuracy": 0.7834303803741932,
      "num_tokens": 53658367.0,
      "step": 27170
    },
    {
      "entropy": 0.9135570010170341,
      "epoch": 0.7785400226286467,
      "grad_norm": 10.625,
      "learning_rate": 3.417564789056165e-05,
      "loss": 0.9946195602416992,
      "mean_token_accuracy": 0.7765499509871006,
      "num_tokens": 53678214.0,
      "step": 27180
    },
    {
      "entropy": 0.856888217292726,
      "epoch": 0.7788264611947353,
      "grad_norm": 7.34375,
      "learning_rate": 3.4165031658520134e-05,
      "loss": 0.8718758583068847,
      "mean_token_accuracy": 0.803025733679533,
      "num_tokens": 53698503.0,
      "step": 27190
    },
    {
      "entropy": 0.8088695771992207,
      "epoch": 0.7791128997608238,
      "grad_norm": 7.6875,
      "learning_rate": 3.415441351682814e-05,
      "loss": 0.8087964057922363,
      "mean_token_accuracy": 0.811004389077425,
      "num_tokens": 53717688.0,
      "step": 27200
    },
    {
      "entropy": 0.8159357467666268,
      "epoch": 0.7793993383269123,
      "grad_norm": 9.0,
      "learning_rate": 3.414379346769808e-05,
      "loss": 0.8449319839477539,
      "mean_token_accuracy": 0.7994780499488116,
      "num_tokens": 53735991.0,
      "step": 27210
    },
    {
      "entropy": 0.8922705762088299,
      "epoch": 0.7796857768930009,
      "grad_norm": 6.40625,
      "learning_rate": 3.4133171513342774e-05,
      "loss": 0.9272421836853028,
      "mean_token_accuracy": 0.7931429699063302,
      "num_tokens": 53756627.0,
      "step": 27220
    },
    {
      "entropy": 0.7911061668768525,
      "epoch": 0.7799722154590895,
      "grad_norm": 9.75,
      "learning_rate": 3.412254765597547e-05,
      "loss": 0.7809607505798339,
      "mean_token_accuracy": 0.8146204989403486,
      "num_tokens": 53776695.0,
      "step": 27230
    },
    {
      "entropy": 0.9404831908643245,
      "epoch": 0.7802586540251779,
      "grad_norm": 8.5625,
      "learning_rate": 3.411192189780974e-05,
      "loss": 0.9064420700073242,
      "mean_token_accuracy": 0.7950816452503204,
      "num_tokens": 53797100.0,
      "step": 27240
    },
    {
      "entropy": 0.9270672312006354,
      "epoch": 0.7805450925912665,
      "grad_norm": 7.09375,
      "learning_rate": 3.410129424105962e-05,
      "loss": 0.8955494880676269,
      "mean_token_accuracy": 0.7878506675362587,
      "num_tokens": 53818264.0,
      "step": 27250
    },
    {
      "entropy": 0.8404777822084725,
      "epoch": 0.780831531157355,
      "grad_norm": 8.5,
      "learning_rate": 3.409066468793952e-05,
      "loss": 0.8340245246887207,
      "mean_token_accuracy": 0.8061087608337403,
      "num_tokens": 53837796.0,
      "step": 27260
    },
    {
      "entropy": 0.7830475624650717,
      "epoch": 0.7811179697234436,
      "grad_norm": 8.125,
      "learning_rate": 3.408003324066423e-05,
      "loss": 0.7928773403167725,
      "mean_token_accuracy": 0.8180467002093792,
      "num_tokens": 53857690.0,
      "step": 27270
    },
    {
      "entropy": 0.9136176530271769,
      "epoch": 0.7814044082895321,
      "grad_norm": 8.8125,
      "learning_rate": 3.406939990144895e-05,
      "loss": 0.9114476203918457,
      "mean_token_accuracy": 0.7952320639044046,
      "num_tokens": 53878092.0,
      "step": 27280
    },
    {
      "entropy": 0.7446376724168658,
      "epoch": 0.7816908468556206,
      "grad_norm": 7.09375,
      "learning_rate": 3.405876467250928e-05,
      "loss": 0.712208604812622,
      "mean_token_accuracy": 0.8371457189321518,
      "num_tokens": 53897536.0,
      "step": 27290
    },
    {
      "entropy": 0.738610465824604,
      "epoch": 0.7819772854217092,
      "grad_norm": 10.125,
      "learning_rate": 3.4048127556061196e-05,
      "loss": 0.731499195098877,
      "mean_token_accuracy": 0.8251092560589314,
      "num_tokens": 53917562.0,
      "step": 27300
    },
    {
      "entropy": 0.7806694192811847,
      "epoch": 0.7822637239877978,
      "grad_norm": 5.625,
      "learning_rate": 3.4037488554321076e-05,
      "loss": 0.7975717544555664,
      "mean_token_accuracy": 0.8179233971983194,
      "num_tokens": 53937753.0,
      "step": 27310
    },
    {
      "entropy": 0.7662141846492887,
      "epoch": 0.7825501625538862,
      "grad_norm": 7.53125,
      "learning_rate": 3.402684766950569e-05,
      "loss": 0.797464656829834,
      "mean_token_accuracy": 0.8247101880609989,
      "num_tokens": 53957123.0,
      "step": 27320
    },
    {
      "entropy": 0.7789935833774507,
      "epoch": 0.7828366011199748,
      "grad_norm": 7.6875,
      "learning_rate": 3.401620490383222e-05,
      "loss": 0.780854606628418,
      "mean_token_accuracy": 0.8135128449648619,
      "num_tokens": 53976563.0,
      "step": 27330
    },
    {
      "entropy": 0.7429076720029115,
      "epoch": 0.7831230396860633,
      "grad_norm": 7.34375,
      "learning_rate": 3.4005560259518185e-05,
      "loss": 0.815247917175293,
      "mean_token_accuracy": 0.82175083309412,
      "num_tokens": 53995717.0,
      "step": 27340
    },
    {
      "entropy": 0.8562973607331514,
      "epoch": 0.7834094782521519,
      "grad_norm": 6.9375,
      "learning_rate": 3.3994913738781565e-05,
      "loss": 0.9203106880187988,
      "mean_token_accuracy": 0.7865301363170147,
      "num_tokens": 54015424.0,
      "step": 27350
    },
    {
      "entropy": 0.8890067538246512,
      "epoch": 0.7836959168182404,
      "grad_norm": 8.6875,
      "learning_rate": 3.3984265343840675e-05,
      "loss": 0.9090202331542969,
      "mean_token_accuracy": 0.7966403916478157,
      "num_tokens": 54036222.0,
      "step": 27360
    },
    {
      "entropy": 0.8855936180800199,
      "epoch": 0.7839823553843289,
      "grad_norm": 8.5625,
      "learning_rate": 3.3973615076914255e-05,
      "loss": 0.8600604057312011,
      "mean_token_accuracy": 0.8035691384226084,
      "num_tokens": 54055841.0,
      "step": 27370
    },
    {
      "entropy": 0.8426823664456606,
      "epoch": 0.7842687939504175,
      "grad_norm": 6.21875,
      "learning_rate": 3.396296294022142e-05,
      "loss": 0.8297712326049804,
      "mean_token_accuracy": 0.811662421002984,
      "num_tokens": 54073868.0,
      "step": 27380
    },
    {
      "entropy": 0.7921288596466184,
      "epoch": 0.784555232516506,
      "grad_norm": 7.46875,
      "learning_rate": 3.3952308935981685e-05,
      "loss": 0.7447253227233886,
      "mean_token_accuracy": 0.8174099512398243,
      "num_tokens": 54093610.0,
      "step": 27390
    },
    {
      "entropy": 0.7598327457904815,
      "epoch": 0.7848416710825945,
      "grad_norm": 6.59375,
      "learning_rate": 3.394165306641494e-05,
      "loss": 0.7683858871459961,
      "mean_token_accuracy": 0.8222594115883112,
      "num_tokens": 54112996.0,
      "step": 27400
    },
    {
      "entropy": 0.7445010785013437,
      "epoch": 0.7851281096486831,
      "grad_norm": 9.25,
      "learning_rate": 3.3930995333741454e-05,
      "loss": 0.822658634185791,
      "mean_token_accuracy": 0.8040300581604243,
      "num_tokens": 54132530.0,
      "step": 27410
    },
    {
      "entropy": 0.9261618816293776,
      "epoch": 0.7854145482147716,
      "grad_norm": 8.0,
      "learning_rate": 3.392033574018193e-05,
      "loss": 0.930577278137207,
      "mean_token_accuracy": 0.7895480517297984,
      "num_tokens": 54151805.0,
      "step": 27420
    },
    {
      "entropy": 0.7637275498360395,
      "epoch": 0.7857009867808602,
      "grad_norm": 10.875,
      "learning_rate": 3.3909674287957416e-05,
      "loss": 0.7505126476287842,
      "mean_token_accuracy": 0.8293727092444897,
      "num_tokens": 54170983.0,
      "step": 27430
    },
    {
      "entropy": 0.8338512158952653,
      "epoch": 0.7859874253469488,
      "grad_norm": 9.0,
      "learning_rate": 3.389901097928935e-05,
      "loss": 0.7913447380065918,
      "mean_token_accuracy": 0.8192352034151554,
      "num_tokens": 54189833.0,
      "step": 27440
    },
    {
      "entropy": 0.7370092187076807,
      "epoch": 0.7862738639130372,
      "grad_norm": 6.15625,
      "learning_rate": 3.388834581639959e-05,
      "loss": 0.757990026473999,
      "mean_token_accuracy": 0.832134285569191,
      "num_tokens": 54209440.0,
      "step": 27450
    },
    {
      "entropy": 0.8110763914883137,
      "epoch": 0.7865603024791258,
      "grad_norm": 9.9375,
      "learning_rate": 3.387767880151034e-05,
      "loss": 0.8233203887939453,
      "mean_token_accuracy": 0.8048214562237263,
      "num_tokens": 54228539.0,
      "step": 27460
    },
    {
      "entropy": 0.704314601328224,
      "epoch": 0.7868467410452143,
      "grad_norm": 8.1875,
      "learning_rate": 3.386700993684422e-05,
      "loss": 0.7126755237579345,
      "mean_token_accuracy": 0.8277702014893293,
      "num_tokens": 54247802.0,
      "step": 27470
    },
    {
      "entropy": 0.7902232861146331,
      "epoch": 0.7871331796113029,
      "grad_norm": 7.9375,
      "learning_rate": 3.3856339224624216e-05,
      "loss": 0.7953148365020752,
      "mean_token_accuracy": 0.8167730834335089,
      "num_tokens": 54268264.0,
      "step": 27480
    },
    {
      "entropy": 0.8709781529381871,
      "epoch": 0.7874196181773914,
      "grad_norm": 10.1875,
      "learning_rate": 3.3845666667073703e-05,
      "loss": 0.9177001953125,
      "mean_token_accuracy": 0.7905503377318382,
      "num_tokens": 54287998.0,
      "step": 27490
    },
    {
      "entropy": 0.8529580660164356,
      "epoch": 0.7877060567434799,
      "grad_norm": 7.625,
      "learning_rate": 3.3834992266416455e-05,
      "loss": 0.881006908416748,
      "mean_token_accuracy": 0.7997597977519035,
      "num_tokens": 54307952.0,
      "step": 27500
    },
    {
      "entropy": 0.7449952673166991,
      "epoch": 0.7879924953095685,
      "grad_norm": 8.5625,
      "learning_rate": 3.38243160248766e-05,
      "loss": 0.7763334751129151,
      "mean_token_accuracy": 0.819374593347311,
      "num_tokens": 54326697.0,
      "step": 27510
    },
    {
      "entropy": 0.8664587201550603,
      "epoch": 0.7882789338756571,
      "grad_norm": 11.0625,
      "learning_rate": 3.381363794467869e-05,
      "loss": 0.8257038116455078,
      "mean_token_accuracy": 0.8158924922347068,
      "num_tokens": 54346387.0,
      "step": 27520
    },
    {
      "entropy": 0.8010411325842142,
      "epoch": 0.7885653724417455,
      "grad_norm": 7.1875,
      "learning_rate": 3.380295802804762e-05,
      "loss": 0.7511368274688721,
      "mean_token_accuracy": 0.8195327013731003,
      "num_tokens": 54364375.0,
      "step": 27530
    },
    {
      "entropy": 0.7704990653321147,
      "epoch": 0.7888518110078341,
      "grad_norm": 10.125,
      "learning_rate": 3.37922762772087e-05,
      "loss": 0.7807973384857178,
      "mean_token_accuracy": 0.8134928107261657,
      "num_tokens": 54384327.0,
      "step": 27540
    },
    {
      "entropy": 0.7036254572682082,
      "epoch": 0.7891382495739226,
      "grad_norm": 10.9375,
      "learning_rate": 3.3781592694387595e-05,
      "loss": 0.7475922584533692,
      "mean_token_accuracy": 0.8302642039954662,
      "num_tokens": 54402999.0,
      "step": 27550
    },
    {
      "entropy": 0.8672280000522733,
      "epoch": 0.7894246881400112,
      "grad_norm": 7.5,
      "learning_rate": 3.3770907281810374e-05,
      "loss": 0.8906898498535156,
      "mean_token_accuracy": 0.7962109569460154,
      "num_tokens": 54422873.0,
      "step": 27560
    },
    {
      "entropy": 0.886970626655966,
      "epoch": 0.7897111267060997,
      "grad_norm": 9.5625,
      "learning_rate": 3.376022004170348e-05,
      "loss": 0.8768546104431152,
      "mean_token_accuracy": 0.7906805079430341,
      "num_tokens": 54442341.0,
      "step": 27570
    },
    {
      "entropy": 0.7895265596918761,
      "epoch": 0.7899975652721882,
      "grad_norm": 9.125,
      "learning_rate": 3.3749530976293734e-05,
      "loss": 0.8083590507507324,
      "mean_token_accuracy": 0.8171726644039154,
      "num_tokens": 54462517.0,
      "step": 27580
    },
    {
      "entropy": 0.8577082969248295,
      "epoch": 0.7902840038382768,
      "grad_norm": 6.40625,
      "learning_rate": 3.373884008780833e-05,
      "loss": 0.8830826759338379,
      "mean_token_accuracy": 0.7991163253784179,
      "num_tokens": 54482638.0,
      "step": 27590
    },
    {
      "entropy": 0.8720566106960177,
      "epoch": 0.7905704424043654,
      "grad_norm": 12.8125,
      "learning_rate": 3.3728147378474855e-05,
      "loss": 0.8983368873596191,
      "mean_token_accuracy": 0.7954681914299726,
      "num_tokens": 54502593.0,
      "step": 27600
    },
    {
      "entropy": 0.8416969131678342,
      "epoch": 0.7908568809704538,
      "grad_norm": 8.3125,
      "learning_rate": 3.371745285052127e-05,
      "loss": 0.8389544486999512,
      "mean_token_accuracy": 0.8033607788383961,
      "num_tokens": 54521617.0,
      "step": 27610
    },
    {
      "entropy": 0.8012358367443084,
      "epoch": 0.7911433195365424,
      "grad_norm": 7.78125,
      "learning_rate": 3.370675650617592e-05,
      "loss": 0.8166860580444336,
      "mean_token_accuracy": 0.8080071896314621,
      "num_tokens": 54540342.0,
      "step": 27620
    },
    {
      "entropy": 0.8698962925001978,
      "epoch": 0.7914297581026309,
      "grad_norm": 5.75,
      "learning_rate": 3.369605834766752e-05,
      "loss": 0.8347339630126953,
      "mean_token_accuracy": 0.8169273231178522,
      "num_tokens": 54560399.0,
      "step": 27630
    },
    {
      "entropy": 0.7485087238252163,
      "epoch": 0.7917161966687195,
      "grad_norm": 8.4375,
      "learning_rate": 3.368535837722516e-05,
      "loss": 0.7746393203735351,
      "mean_token_accuracy": 0.814680828526616,
      "num_tokens": 54580707.0,
      "step": 27640
    },
    {
      "entropy": 0.8822613493539393,
      "epoch": 0.792002635234808,
      "grad_norm": 9.25,
      "learning_rate": 3.3674656597078326e-05,
      "loss": 0.8859702110290527,
      "mean_token_accuracy": 0.7957746129482984,
      "num_tokens": 54600850.0,
      "step": 27650
    },
    {
      "entropy": 0.8395327577367425,
      "epoch": 0.7922890738008965,
      "grad_norm": 10.5,
      "learning_rate": 3.3663953009456865e-05,
      "loss": 0.8797383308410645,
      "mean_token_accuracy": 0.7987513985484839,
      "num_tokens": 54621396.0,
      "step": 27660
    },
    {
      "entropy": 0.8077081792056561,
      "epoch": 0.7925755123669851,
      "grad_norm": 6.15625,
      "learning_rate": 3.3653247616591014e-05,
      "loss": 0.8178219795227051,
      "mean_token_accuracy": 0.8214574571698904,
      "num_tokens": 54640363.0,
      "step": 27670
    },
    {
      "entropy": 0.6903973253443837,
      "epoch": 0.7928619509330737,
      "grad_norm": 9.375,
      "learning_rate": 3.3642540420711356e-05,
      "loss": 0.6705481052398682,
      "mean_token_accuracy": 0.8393583551049233,
      "num_tokens": 54660367.0,
      "step": 27680
    },
    {
      "entropy": 0.804171965457499,
      "epoch": 0.7931483894991622,
      "grad_norm": 9.8125,
      "learning_rate": 3.363183142404889e-05,
      "loss": 0.8228458404541016,
      "mean_token_accuracy": 0.802550283819437,
      "num_tokens": 54679532.0,
      "step": 27690
    },
    {
      "entropy": 0.8355716333724559,
      "epoch": 0.7934348280652507,
      "grad_norm": 8.1875,
      "learning_rate": 3.362112062883496e-05,
      "loss": 0.888131046295166,
      "mean_token_accuracy": 0.8022970698773861,
      "num_tokens": 54698721.0,
      "step": 27700
    },
    {
      "entropy": 0.7744325898587704,
      "epoch": 0.7937212666313392,
      "grad_norm": 10.0625,
      "learning_rate": 3.36104080373013e-05,
      "loss": 0.7843059062957763,
      "mean_token_accuracy": 0.8119222715497016,
      "num_tokens": 54718188.0,
      "step": 27710
    },
    {
      "entropy": 0.7492793755605817,
      "epoch": 0.7940077051974278,
      "grad_norm": 9.375,
      "learning_rate": 3.3599693651680014e-05,
      "loss": 0.7239820480346679,
      "mean_token_accuracy": 0.8196954429149628,
      "num_tokens": 54736663.0,
      "step": 27720
    },
    {
      "entropy": 0.8001454466953873,
      "epoch": 0.7942941437635164,
      "grad_norm": 8.8125,
      "learning_rate": 3.3588977474203586e-05,
      "loss": 0.7705713748931885,
      "mean_token_accuracy": 0.8149224519729614,
      "num_tokens": 54754746.0,
      "step": 27730
    },
    {
      "entropy": 0.7890005730092525,
      "epoch": 0.7945805823296048,
      "grad_norm": 7.15625,
      "learning_rate": 3.357825950710486e-05,
      "loss": 0.8422120094299317,
      "mean_token_accuracy": 0.8063961088657379,
      "num_tokens": 54774950.0,
      "step": 27740
    },
    {
      "entropy": 0.9306843562051654,
      "epoch": 0.7948670208956934,
      "grad_norm": 9.5625,
      "learning_rate": 3.356753975261706e-05,
      "loss": 0.9213499069213867,
      "mean_token_accuracy": 0.7931857898831367,
      "num_tokens": 54794879.0,
      "step": 27750
    },
    {
      "entropy": 0.8222947584465146,
      "epoch": 0.795153459461782,
      "grad_norm": 8.375,
      "learning_rate": 3.355681821297378e-05,
      "loss": 0.8459925651550293,
      "mean_token_accuracy": 0.797735521197319,
      "num_tokens": 54813878.0,
      "step": 27760
    },
    {
      "entropy": 0.8204174122773111,
      "epoch": 0.7954398980278705,
      "grad_norm": 5.40625,
      "learning_rate": 3.3546094890409e-05,
      "loss": 0.8607306480407715,
      "mean_token_accuracy": 0.8043927855789661,
      "num_tokens": 54832458.0,
      "step": 27770
    },
    {
      "entropy": 0.817804906796664,
      "epoch": 0.795726336593959,
      "grad_norm": 9.5,
      "learning_rate": 3.3535369787157055e-05,
      "loss": 0.803005599975586,
      "mean_token_accuracy": 0.8113078087568283,
      "num_tokens": 54852538.0,
      "step": 27780
    },
    {
      "entropy": 0.8652782214805483,
      "epoch": 0.7960127751600475,
      "grad_norm": 10.625,
      "learning_rate": 3.352464290545265e-05,
      "loss": 0.8507365226745606,
      "mean_token_accuracy": 0.7972301304340362,
      "num_tokens": 54872841.0,
      "step": 27790
    },
    {
      "entropy": 0.916761283390224,
      "epoch": 0.7962992137261361,
      "grad_norm": 7.84375,
      "learning_rate": 3.3513914247530874e-05,
      "loss": 0.9459691047668457,
      "mean_token_accuracy": 0.7870757639408111,
      "num_tokens": 54892453.0,
      "step": 27800
    },
    {
      "entropy": 0.9235548799857497,
      "epoch": 0.7965856522922247,
      "grad_norm": 8.3125,
      "learning_rate": 3.3503183815627176e-05,
      "loss": 0.9205671310424804,
      "mean_token_accuracy": 0.7954581264406443,
      "num_tokens": 54912667.0,
      "step": 27810
    },
    {
      "entropy": 0.8028525048866868,
      "epoch": 0.7968720908583131,
      "grad_norm": 10.25,
      "learning_rate": 3.3492451611977385e-05,
      "loss": 0.7804036140441895,
      "mean_token_accuracy": 0.8128953039646148,
      "num_tokens": 54932708.0,
      "step": 27820
    },
    {
      "entropy": 0.7424164745025337,
      "epoch": 0.7971585294244017,
      "grad_norm": 6.84375,
      "learning_rate": 3.3481717638817675e-05,
      "loss": 0.6822882652282715,
      "mean_token_accuracy": 0.8353868663311005,
      "num_tokens": 54952235.0,
      "step": 27830
    },
    {
      "entropy": 0.7837825223803521,
      "epoch": 0.7974449679904902,
      "grad_norm": 6.59375,
      "learning_rate": 3.347098189838461e-05,
      "loss": 0.7703558444976807,
      "mean_token_accuracy": 0.8114307932555676,
      "num_tokens": 54972200.0,
      "step": 27840
    },
    {
      "entropy": 0.7263157333247363,
      "epoch": 0.7977314065565788,
      "grad_norm": 9.125,
      "learning_rate": 3.3460244392915134e-05,
      "loss": 0.7378087997436523,
      "mean_token_accuracy": 0.8233725275844336,
      "num_tokens": 54993415.0,
      "step": 27850
    },
    {
      "entropy": 0.8423685887828469,
      "epoch": 0.7980178451226673,
      "grad_norm": 7.34375,
      "learning_rate": 3.344950512464653e-05,
      "loss": 0.8709941864013672,
      "mean_token_accuracy": 0.7998374745249748,
      "num_tokens": 55013786.0,
      "step": 27860
    },
    {
      "entropy": 0.7437105781398714,
      "epoch": 0.7983042836887558,
      "grad_norm": 7.84375,
      "learning_rate": 3.343876409581645e-05,
      "loss": 0.7061602592468261,
      "mean_token_accuracy": 0.8246432077139616,
      "num_tokens": 55035007.0,
      "step": 27870
    },
    {
      "entropy": 0.8426749626174569,
      "epoch": 0.7985907222548444,
      "grad_norm": 10.75,
      "learning_rate": 3.342802130866295e-05,
      "loss": 0.9132275581359863,
      "mean_token_accuracy": 0.7971338201314211,
      "num_tokens": 55053796.0,
      "step": 27880
    },
    {
      "entropy": 0.8186741705052555,
      "epoch": 0.798877160820933,
      "grad_norm": 8.125,
      "learning_rate": 3.3417276765424396e-05,
      "loss": 0.7768548488616943,
      "mean_token_accuracy": 0.8114475965499878,
      "num_tokens": 55072392.0,
      "step": 27890
    },
    {
      "entropy": 0.8107778772711753,
      "epoch": 0.7991635993870214,
      "grad_norm": 6.40625,
      "learning_rate": 3.340653046833957e-05,
      "loss": 0.8070590019226074,
      "mean_token_accuracy": 0.8183181062340736,
      "num_tokens": 55091742.0,
      "step": 27900
    },
    {
      "entropy": 0.8377520043402911,
      "epoch": 0.79945003795311,
      "grad_norm": 8.0625,
      "learning_rate": 3.33957824196476e-05,
      "loss": 0.8700351715087891,
      "mean_token_accuracy": 0.8086950473487378,
      "num_tokens": 55111402.0,
      "step": 27910
    },
    {
      "entropy": 0.8931532787159085,
      "epoch": 0.7997364765191985,
      "grad_norm": 7.8125,
      "learning_rate": 3.3385032621587966e-05,
      "loss": 0.8629809379577636,
      "mean_token_accuracy": 0.8035509563982487,
      "num_tokens": 55131816.0,
      "step": 27920
    },
    {
      "entropy": 0.7146860882639885,
      "epoch": 0.8000229150852871,
      "grad_norm": 5.875,
      "learning_rate": 3.3374281076400533e-05,
      "loss": 0.7089365005493165,
      "mean_token_accuracy": 0.8271175030618906,
      "num_tokens": 55151912.0,
      "step": 27930
    },
    {
      "entropy": 0.7818871272727848,
      "epoch": 0.8003093536513757,
      "grad_norm": 9.0625,
      "learning_rate": 3.336352778632551e-05,
      "loss": 0.7777845859527588,
      "mean_token_accuracy": 0.819730905070901,
      "num_tokens": 55172128.0,
      "step": 27940
    },
    {
      "entropy": 0.7809330351650715,
      "epoch": 0.8005957922174641,
      "grad_norm": 9.4375,
      "learning_rate": 3.33527727536035e-05,
      "loss": 0.801824951171875,
      "mean_token_accuracy": 0.8095112022012472,
      "num_tokens": 55193607.0,
      "step": 27950
    },
    {
      "entropy": 0.8180649184621871,
      "epoch": 0.8008822307835527,
      "grad_norm": 10.0,
      "learning_rate": 3.334201598047544e-05,
      "loss": 0.8939565658569336,
      "mean_token_accuracy": 0.8010160747915507,
      "num_tokens": 55213481.0,
      "step": 27960
    },
    {
      "entropy": 0.7968553135171532,
      "epoch": 0.8011686693496413,
      "grad_norm": 9.1875,
      "learning_rate": 3.333125746918263e-05,
      "loss": 0.806216812133789,
      "mean_token_accuracy": 0.8172412820160389,
      "num_tokens": 55234128.0,
      "step": 27970
    },
    {
      "entropy": 0.8370412548072637,
      "epoch": 0.8014551079157298,
      "grad_norm": 7.25,
      "learning_rate": 3.332049722196677e-05,
      "loss": 0.8066755294799804,
      "mean_token_accuracy": 0.7961332734674216,
      "num_tokens": 55253407.0,
      "step": 27980
    },
    {
      "entropy": 0.7825516179203987,
      "epoch": 0.8017415464818183,
      "grad_norm": 9.9375,
      "learning_rate": 3.330973524106986e-05,
      "loss": 0.8366578102111817,
      "mean_token_accuracy": 0.8125187080353499,
      "num_tokens": 55273315.0,
      "step": 27990
    },
    {
      "epoch": 0.8020279850479068,
      "eval_entropy": 0.8579977637529373,
      "eval_loss": 0.8514816761016846,
      "eval_mean_token_accuracy": 0.8004682188034058,
      "eval_num_tokens": 55292588.0,
      "eval_runtime": 42.8981,
      "eval_samples_per_second": 46.622,
      "eval_steps_per_second": 5.828,
      "step": 28000
    },
    {
      "entropy": 0.7965041088871658,
      "epoch": 0.8023144236139954,
      "grad_norm": 5.53125,
      "learning_rate": 3.32882060872029e-05,
      "loss": 0.767036771774292,
      "mean_token_accuracy": 0.8245044101029635,
      "num_tokens": 55312172.0,
      "step": 28010
    },
    {
      "entropy": 0.8278432270511985,
      "epoch": 0.802600862180084,
      "grad_norm": 5.09375,
      "learning_rate": 3.327743891871871e-05,
      "loss": 0.8559127807617187,
      "mean_token_accuracy": 0.798911839723587,
      "num_tokens": 55331720.0,
      "step": 28020
    },
    {
      "entropy": 0.8071053460240364,
      "epoch": 0.8028873007461724,
      "grad_norm": 7.21875,
      "learning_rate": 3.326667002552523e-05,
      "loss": 0.8126046180725097,
      "mean_token_accuracy": 0.8210407923907042,
      "num_tokens": 55350656.0,
      "step": 28030
    },
    {
      "entropy": 0.777292434964329,
      "epoch": 0.803173739312261,
      "grad_norm": 6.78125,
      "learning_rate": 3.32558994098663e-05,
      "loss": 0.8126914978027344,
      "mean_token_accuracy": 0.8060299221426248,
      "num_tokens": 55370524.0,
      "step": 28040
    },
    {
      "entropy": 0.9075377456843853,
      "epoch": 0.8034601778783496,
      "grad_norm": 10.5,
      "learning_rate": 3.32451270739861e-05,
      "loss": 0.9714938163757324,
      "mean_token_accuracy": 0.785105611756444,
      "num_tokens": 55389716.0,
      "step": 28050
    },
    {
      "entropy": 0.9205982012674212,
      "epoch": 0.8037466164444381,
      "grad_norm": 13.1875,
      "learning_rate": 3.32343530201292e-05,
      "loss": 0.9746971130371094,
      "mean_token_accuracy": 0.7889183472841979,
      "num_tokens": 55409068.0,
      "step": 28060
    },
    {
      "entropy": 0.7408014304935933,
      "epoch": 0.8040330550105266,
      "grad_norm": 8.6875,
      "learning_rate": 3.32235772505405e-05,
      "loss": 0.6835726737976074,
      "mean_token_accuracy": 0.8376939497888088,
      "num_tokens": 55428087.0,
      "step": 28070
    },
    {
      "entropy": 0.9731947622261942,
      "epoch": 0.8043194935766151,
      "grad_norm": 7.5,
      "learning_rate": 3.3212799767465266e-05,
      "loss": 1.0123875617980957,
      "mean_token_accuracy": 0.7754117127507925,
      "num_tokens": 55447840.0,
      "step": 28080
    },
    {
      "entropy": 0.8219267277047038,
      "epoch": 0.8046059321427037,
      "grad_norm": 9.5625,
      "learning_rate": 3.320202057314913e-05,
      "loss": 0.8382545471191406,
      "mean_token_accuracy": 0.8078667342662811,
      "num_tokens": 55466296.0,
      "step": 28090
    },
    {
      "entropy": 0.8654251422733068,
      "epoch": 0.8048923707087923,
      "grad_norm": 8.0625,
      "learning_rate": 3.319123966983808e-05,
      "loss": 0.8922965049743652,
      "mean_token_accuracy": 0.7910533580929041,
      "num_tokens": 55485786.0,
      "step": 28100
    },
    {
      "entropy": 0.8985217202454805,
      "epoch": 0.8051788092748807,
      "grad_norm": 6.875,
      "learning_rate": 3.318045705977844e-05,
      "loss": 0.8435938835144043,
      "mean_token_accuracy": 0.8026093490421772,
      "num_tokens": 55507422.0,
      "step": 28110
    },
    {
      "entropy": 0.7964549114927649,
      "epoch": 0.8054652478409693,
      "grad_norm": 10.6875,
      "learning_rate": 3.316967274521691e-05,
      "loss": 0.7600378036499024,
      "mean_token_accuracy": 0.8110054653137923,
      "num_tokens": 55526464.0,
      "step": 28120
    },
    {
      "entropy": 0.7920875212177634,
      "epoch": 0.8057516864070579,
      "grad_norm": 8.75,
      "learning_rate": 3.3158886728400535e-05,
      "loss": 0.7771776199340821,
      "mean_token_accuracy": 0.8169278211891651,
      "num_tokens": 55545723.0,
      "step": 28130
    },
    {
      "entropy": 0.768783807568252,
      "epoch": 0.8060381249731464,
      "grad_norm": 7.1875,
      "learning_rate": 3.3148099011576724e-05,
      "loss": 0.7491528987884521,
      "mean_token_accuracy": 0.8237960509955883,
      "num_tokens": 55564422.0,
      "step": 28140
    },
    {
      "entropy": 0.8112536994740367,
      "epoch": 0.8063245635392349,
      "grad_norm": 7.90625,
      "learning_rate": 3.313730959699322e-05,
      "loss": 0.8170635223388671,
      "mean_token_accuracy": 0.814702083542943,
      "num_tokens": 55583602.0,
      "step": 28150
    },
    {
      "entropy": 0.8673863649368286,
      "epoch": 0.8066110021053234,
      "grad_norm": 6.5625,
      "learning_rate": 3.3126518486898165e-05,
      "loss": 0.8710968017578125,
      "mean_token_accuracy": 0.7991614185273648,
      "num_tokens": 55604271.0,
      "step": 28160
    },
    {
      "entropy": 0.8141280696727335,
      "epoch": 0.806897440671412,
      "grad_norm": 9.125,
      "learning_rate": 3.311572568353999e-05,
      "loss": 0.803193473815918,
      "mean_token_accuracy": 0.8101694494485855,
      "num_tokens": 55623518.0,
      "step": 28170
    },
    {
      "entropy": 0.7534861601889133,
      "epoch": 0.8071838792375006,
      "grad_norm": 8.25,
      "learning_rate": 3.310493118916753e-05,
      "loss": 0.7779221057891845,
      "mean_token_accuracy": 0.8185298666357994,
      "num_tokens": 55643159.0,
      "step": 28180
    },
    {
      "entropy": 0.8317865122109651,
      "epoch": 0.807470317803589,
      "grad_norm": 9.625,
      "learning_rate": 3.309413500602996e-05,
      "loss": 0.8105935096740723,
      "mean_token_accuracy": 0.8151186440140009,
      "num_tokens": 55663553.0,
      "step": 28190
    },
    {
      "entropy": 0.8132535924669355,
      "epoch": 0.8077567563696776,
      "grad_norm": 7.4375,
      "learning_rate": 3.3083337136376774e-05,
      "loss": 0.828223991394043,
      "mean_token_accuracy": 0.8068164076656104,
      "num_tokens": 55683458.0,
      "step": 28200
    },
    {
      "entropy": 0.8282989108934998,
      "epoch": 0.8080431949357662,
      "grad_norm": 17.125,
      "learning_rate": 3.3072537582457876e-05,
      "loss": 0.8760874748229981,
      "mean_token_accuracy": 0.7999740038067102,
      "num_tokens": 55703474.0,
      "step": 28210
    },
    {
      "entropy": 0.6915542869828641,
      "epoch": 0.8083296335018547,
      "grad_norm": 11.8125,
      "learning_rate": 3.306173634652348e-05,
      "loss": 0.7259965419769288,
      "mean_token_accuracy": 0.8261226303875446,
      "num_tokens": 55723444.0,
      "step": 28220
    },
    {
      "entropy": 0.8110984233208001,
      "epoch": 0.8086160720679433,
      "grad_norm": 11.9375,
      "learning_rate": 3.3050933430824155e-05,
      "loss": 0.8824068069458008,
      "mean_token_accuracy": 0.8024394981563091,
      "num_tokens": 55743775.0,
      "step": 28230
    },
    {
      "entropy": 0.7036050744354725,
      "epoch": 0.8089025106340317,
      "grad_norm": 7.0625,
      "learning_rate": 3.304012883761082e-05,
      "loss": 0.7061667919158936,
      "mean_token_accuracy": 0.8298017721623182,
      "num_tokens": 55762481.0,
      "step": 28240
    },
    {
      "entropy": 0.8209268234670162,
      "epoch": 0.8091889492001203,
      "grad_norm": 8.25,
      "learning_rate": 3.3029322569134765e-05,
      "loss": 0.8277984619140625,
      "mean_token_accuracy": 0.8000136941671372,
      "num_tokens": 55782979.0,
      "step": 28250
    },
    {
      "entropy": 0.7996790800243616,
      "epoch": 0.8094753877662089,
      "grad_norm": 7.65625,
      "learning_rate": 3.301851462764761e-05,
      "loss": 0.7634175300598145,
      "mean_token_accuracy": 0.8244042079895735,
      "num_tokens": 55802711.0,
      "step": 28260
    },
    {
      "entropy": 0.7720209307968616,
      "epoch": 0.8097618263322974,
      "grad_norm": 6.625,
      "learning_rate": 3.3007705015401315e-05,
      "loss": 0.7042385101318359,
      "mean_token_accuracy": 0.8331066224724054,
      "num_tokens": 55822042.0,
      "step": 28270
    },
    {
      "entropy": 0.8143428886309266,
      "epoch": 0.8100482648983859,
      "grad_norm": 6.34375,
      "learning_rate": 3.299689373464821e-05,
      "loss": 0.7786048412322998,
      "mean_token_accuracy": 0.8136181354522705,
      "num_tokens": 55842486.0,
      "step": 28280
    },
    {
      "entropy": 0.6399426753632724,
      "epoch": 0.8103347034644744,
      "grad_norm": 9.0625,
      "learning_rate": 3.298608078764096e-05,
      "loss": 0.6788989543914795,
      "mean_token_accuracy": 0.8458281025290489,
      "num_tokens": 55861622.0,
      "step": 28290
    },
    {
      "entropy": 0.7723801395855844,
      "epoch": 0.810621142030563,
      "grad_norm": 7.90625,
      "learning_rate": 3.297526617663257e-05,
      "loss": 0.7573992729187011,
      "mean_token_accuracy": 0.8250520456582308,
      "num_tokens": 55882058.0,
      "step": 28300
    },
    {
      "entropy": 0.8319909907877445,
      "epoch": 0.8109075805966516,
      "grad_norm": 11.1875,
      "learning_rate": 3.296444990387641e-05,
      "loss": 0.8488040924072265,
      "mean_token_accuracy": 0.8045039303600788,
      "num_tokens": 55900741.0,
      "step": 28310
    },
    {
      "entropy": 0.7003204156644642,
      "epoch": 0.81119401916274,
      "grad_norm": 9.875,
      "learning_rate": 3.2953631971626195e-05,
      "loss": 0.7582035541534424,
      "mean_token_accuracy": 0.8310518264770508,
      "num_tokens": 55920154.0,
      "step": 28320
    },
    {
      "entropy": 0.7558060493320227,
      "epoch": 0.8114804577288286,
      "grad_norm": 9.875,
      "learning_rate": 3.294281238213597e-05,
      "loss": 0.8771191596984863,
      "mean_token_accuracy": 0.8069724943488836,
      "num_tokens": 55938835.0,
      "step": 28330
    },
    {
      "entropy": 0.766294585634023,
      "epoch": 0.8117668962949172,
      "grad_norm": 6.59375,
      "learning_rate": 3.293199113766013e-05,
      "loss": 0.8008727073669434,
      "mean_token_accuracy": 0.8161512058228254,
      "num_tokens": 55958557.0,
      "step": 28340
    },
    {
      "entropy": 0.785760188056156,
      "epoch": 0.8120533348610057,
      "grad_norm": 8.5,
      "learning_rate": 3.292116824045343e-05,
      "loss": 0.7491011142730712,
      "mean_token_accuracy": 0.8138927549123764,
      "num_tokens": 55976957.0,
      "step": 28350
    },
    {
      "entropy": 0.7914698297157884,
      "epoch": 0.8123397734270942,
      "grad_norm": 7.03125,
      "learning_rate": 3.291034369277095e-05,
      "loss": 0.7969761848449707,
      "mean_token_accuracy": 0.8066071681678295,
      "num_tokens": 55996228.0,
      "step": 28360
    },
    {
      "entropy": 0.7710186077281833,
      "epoch": 0.8126262119931827,
      "grad_norm": 6.96875,
      "learning_rate": 3.28995174968681e-05,
      "loss": 0.7355557441711426,
      "mean_token_accuracy": 0.8310621995478868,
      "num_tokens": 56016286.0,
      "step": 28370
    },
    {
      "entropy": 0.7336072530597448,
      "epoch": 0.8129126505592713,
      "grad_norm": 6.6875,
      "learning_rate": 3.2888689655000704e-05,
      "loss": 0.7370304107666016,
      "mean_token_accuracy": 0.8258808370679617,
      "num_tokens": 56036154.0,
      "step": 28380
    },
    {
      "entropy": 0.8553747545927763,
      "epoch": 0.8131990891253599,
      "grad_norm": 8.1875,
      "learning_rate": 3.287786016942485e-05,
      "loss": 0.8886061668395996,
      "mean_token_accuracy": 0.7941734835505485,
      "num_tokens": 56057331.0,
      "step": 28390
    },
    {
      "entropy": 0.697972783073783,
      "epoch": 0.8134855276914483,
      "grad_norm": 7.875,
      "learning_rate": 3.286702904239698e-05,
      "loss": 0.6739721775054932,
      "mean_token_accuracy": 0.8350292831659317,
      "num_tokens": 56078279.0,
      "step": 28400
    },
    {
      "entropy": 0.7766121156979352,
      "epoch": 0.8137719662575369,
      "grad_norm": 8.625,
      "learning_rate": 3.285619627617394e-05,
      "loss": 0.8122026443481445,
      "mean_token_accuracy": 0.8055469546467066,
      "num_tokens": 56097224.0,
      "step": 28410
    },
    {
      "entropy": 0.9134762043133378,
      "epoch": 0.8140584048236255,
      "grad_norm": 10.625,
      "learning_rate": 3.284536187301284e-05,
      "loss": 0.9300140380859375,
      "mean_token_accuracy": 0.7876807183027268,
      "num_tokens": 56117879.0,
      "step": 28420
    },
    {
      "entropy": 0.8468256225809455,
      "epoch": 0.814344843389714,
      "grad_norm": 11.4375,
      "learning_rate": 3.2834525835171186e-05,
      "loss": 0.8127585411071777,
      "mean_token_accuracy": 0.8125965811312199,
      "num_tokens": 56137361.0,
      "step": 28430
    },
    {
      "entropy": 0.6850379346869886,
      "epoch": 0.8146312819558025,
      "grad_norm": 9.4375,
      "learning_rate": 3.2823688164906794e-05,
      "loss": 0.7070194244384765,
      "mean_token_accuracy": 0.8307677719742059,
      "num_tokens": 56156527.0,
      "step": 28440
    },
    {
      "entropy": 0.9230156814679503,
      "epoch": 0.814917720521891,
      "grad_norm": 7.5,
      "learning_rate": 3.2812848864477825e-05,
      "loss": 0.959440803527832,
      "mean_token_accuracy": 0.7797123346477747,
      "num_tokens": 56177564.0,
      "step": 28450
    },
    {
      "entropy": 0.7334650885313749,
      "epoch": 0.8152041590879796,
      "grad_norm": 8.75,
      "learning_rate": 3.280200793614279e-05,
      "loss": 0.7139073848724365,
      "mean_token_accuracy": 0.8284463182091713,
      "num_tokens": 56195973.0,
      "step": 28460
    },
    {
      "entropy": 0.7367174927145242,
      "epoch": 0.8154905976540682,
      "grad_norm": 8.5,
      "learning_rate": 3.279116538216054e-05,
      "loss": 0.7559236526489258,
      "mean_token_accuracy": 0.8243048164993525,
      "num_tokens": 56214941.0,
      "step": 28470
    },
    {
      "entropy": 0.8480804964434355,
      "epoch": 0.8157770362201567,
      "grad_norm": 7.0625,
      "learning_rate": 3.278032120479025e-05,
      "loss": 0.8648321151733398,
      "mean_token_accuracy": 0.8149437446147203,
      "num_tokens": 56236347.0,
      "step": 28480
    },
    {
      "entropy": 0.7884205592796206,
      "epoch": 0.8160634747862452,
      "grad_norm": 7.0625,
      "learning_rate": 3.276947540629145e-05,
      "loss": 0.7757551193237304,
      "mean_token_accuracy": 0.8176909003406763,
      "num_tokens": 56256605.0,
      "step": 28490
    },
    {
      "entropy": 0.8537431471049786,
      "epoch": 0.8163499133523338,
      "grad_norm": 8.3125,
      "learning_rate": 3.275862798892399e-05,
      "loss": 0.883885669708252,
      "mean_token_accuracy": 0.7964702393859625,
      "num_tokens": 56277880.0,
      "step": 28500
    },
    {
      "entropy": 0.8472736172378064,
      "epoch": 0.8166363519184223,
      "grad_norm": 6.84375,
      "learning_rate": 3.274777895494808e-05,
      "loss": 0.8621554374694824,
      "mean_token_accuracy": 0.8078082256019116,
      "num_tokens": 56296402.0,
      "step": 28510
    },
    {
      "entropy": 0.8340712578967213,
      "epoch": 0.8169227904845109,
      "grad_norm": 7.28125,
      "learning_rate": 3.273692830662423e-05,
      "loss": 0.8262110710144043,
      "mean_token_accuracy": 0.8100854031741619,
      "num_tokens": 56315837.0,
      "step": 28520
    },
    {
      "entropy": 0.8134298447519541,
      "epoch": 0.8172092290505993,
      "grad_norm": 10.4375,
      "learning_rate": 3.272607604621334e-05,
      "loss": 0.7706833839416504,
      "mean_token_accuracy": 0.82297851331532,
      "num_tokens": 56336066.0,
      "step": 28530
    },
    {
      "entropy": 0.7645992210134864,
      "epoch": 0.8174956676166879,
      "grad_norm": 8.3125,
      "learning_rate": 3.271522217597659e-05,
      "loss": 0.7834047317504883,
      "mean_token_accuracy": 0.8217083159834146,
      "num_tokens": 56356098.0,
      "step": 28540
    },
    {
      "entropy": 0.7113511046394706,
      "epoch": 0.8177821061827765,
      "grad_norm": 9.4375,
      "learning_rate": 3.270436669817554e-05,
      "loss": 0.7099591732025147,
      "mean_token_accuracy": 0.8334423072636128,
      "num_tokens": 56376660.0,
      "step": 28550
    },
    {
      "entropy": 0.7362639323342591,
      "epoch": 0.818068544748865,
      "grad_norm": 9.9375,
      "learning_rate": 3.2693509615072046e-05,
      "loss": 0.7709840297698974,
      "mean_token_accuracy": 0.8163731936365366,
      "num_tokens": 56395724.0,
      "step": 28560
    },
    {
      "entropy": 0.779928307235241,
      "epoch": 0.8183549833149535,
      "grad_norm": 5.0625,
      "learning_rate": 3.268265092892835e-05,
      "loss": 0.8373103141784668,
      "mean_token_accuracy": 0.8215111322700978,
      "num_tokens": 56415559.0,
      "step": 28570
    },
    {
      "entropy": 0.8506896877661347,
      "epoch": 0.818641421881042,
      "grad_norm": 6.1875,
      "learning_rate": 3.267179064200697e-05,
      "loss": 0.888783073425293,
      "mean_token_accuracy": 0.7982708849012852,
      "num_tokens": 56434445.0,
      "step": 28580
    },
    {
      "entropy": 0.8119755172170698,
      "epoch": 0.8189278604471306,
      "grad_norm": 8.1875,
      "learning_rate": 3.2660928756570785e-05,
      "loss": 0.7820671081542969,
      "mean_token_accuracy": 0.8132979936897755,
      "num_tokens": 56455010.0,
      "step": 28590
    },
    {
      "entropy": 0.8307918988168239,
      "epoch": 0.8192142990132192,
      "grad_norm": 7.875,
      "learning_rate": 3.265006527488303e-05,
      "loss": 0.8435712814331054,
      "mean_token_accuracy": 0.810058568790555,
      "num_tokens": 56474307.0,
      "step": 28600
    },
    {
      "entropy": 0.8997833956032991,
      "epoch": 0.8195007375793076,
      "grad_norm": 10.5,
      "learning_rate": 3.263920019920722e-05,
      "loss": 0.93572998046875,
      "mean_token_accuracy": 0.7860208701342344,
      "num_tokens": 56494552.0,
      "step": 28610
    },
    {
      "entropy": 0.8387562816962599,
      "epoch": 0.8197871761453962,
      "grad_norm": 7.25,
      "learning_rate": 3.262833353180725e-05,
      "loss": 0.8587051391601562,
      "mean_token_accuracy": 0.7979085359722375,
      "num_tokens": 56514769.0,
      "step": 28620
    },
    {
      "entropy": 0.8135805578902364,
      "epoch": 0.8200736147114848,
      "grad_norm": 9.0625,
      "learning_rate": 3.261746527494732e-05,
      "loss": 0.7536961555480957,
      "mean_token_accuracy": 0.8163473837077617,
      "num_tokens": 56534548.0,
      "step": 28630
    },
    {
      "entropy": 0.876524878013879,
      "epoch": 0.8203600532775733,
      "grad_norm": 8.875,
      "learning_rate": 3.2606595430891976e-05,
      "loss": 0.8418699264526367,
      "mean_token_accuracy": 0.8009232211858034,
      "num_tokens": 56554726.0,
      "step": 28640
    },
    {
      "entropy": 0.8523045282810926,
      "epoch": 0.8206464918436618,
      "grad_norm": 9.125,
      "learning_rate": 3.259572400190608e-05,
      "loss": 0.8775565147399902,
      "mean_token_accuracy": 0.792010610550642,
      "num_tokens": 56573853.0,
      "step": 28650
    },
    {
      "entropy": 0.7184994790703059,
      "epoch": 0.8209329304097503,
      "grad_norm": 7.46875,
      "learning_rate": 3.2584850990254845e-05,
      "loss": 0.7418111801147461,
      "mean_token_accuracy": 0.8296581000089646,
      "num_tokens": 56592966.0,
      "step": 28660
    },
    {
      "entropy": 0.7740728376433254,
      "epoch": 0.8212193689758389,
      "grad_norm": 8.6875,
      "learning_rate": 3.257397639820377e-05,
      "loss": 0.7192776203155518,
      "mean_token_accuracy": 0.8281273398548364,
      "num_tokens": 56613029.0,
      "step": 28670
    },
    {
      "entropy": 0.7948544580489397,
      "epoch": 0.8215058075419275,
      "grad_norm": 8.8125,
      "learning_rate": 3.2563100228018755e-05,
      "loss": 0.7648997783660889,
      "mean_token_accuracy": 0.8154974408447743,
      "num_tokens": 56633248.0,
      "step": 28680
    },
    {
      "entropy": 0.6921715181320905,
      "epoch": 0.8217922461080159,
      "grad_norm": 5.09375,
      "learning_rate": 3.2552222481965965e-05,
      "loss": 0.7475815296173096,
      "mean_token_accuracy": 0.8150351971387864,
      "num_tokens": 56652176.0,
      "step": 28690
    },
    {
      "entropy": 0.7420700205489993,
      "epoch": 0.8220786846741045,
      "grad_norm": 8.6875,
      "learning_rate": 3.254134316231191e-05,
      "loss": 0.7592799663543701,
      "mean_token_accuracy": 0.8190534114837646,
      "num_tokens": 56670895.0,
      "step": 28700
    },
    {
      "entropy": 0.8127527175471186,
      "epoch": 0.8223651232401931,
      "grad_norm": 9.1875,
      "learning_rate": 3.253046227132345e-05,
      "loss": 0.8312885284423828,
      "mean_token_accuracy": 0.8041517049074173,
      "num_tokens": 56690314.0,
      "step": 28710
    },
    {
      "entropy": 0.6863895329646766,
      "epoch": 0.8226515618062816,
      "grad_norm": 7.78125,
      "learning_rate": 3.2519579811267756e-05,
      "loss": 0.6524003982543946,
      "mean_token_accuracy": 0.8450367726385594,
      "num_tokens": 56709121.0,
      "step": 28720
    },
    {
      "entropy": 0.7860417971387506,
      "epoch": 0.8229380003723701,
      "grad_norm": 9.3125,
      "learning_rate": 3.2508695784412305e-05,
      "loss": 0.7660788059234619,
      "mean_token_accuracy": 0.82461027354002,
      "num_tokens": 56727750.0,
      "step": 28730
    },
    {
      "entropy": 0.7708477799780666,
      "epoch": 0.8232244389384586,
      "grad_norm": 7.96875,
      "learning_rate": 3.2497810193024944e-05,
      "loss": 0.8180028915405273,
      "mean_token_accuracy": 0.8103749591857194,
      "num_tokens": 56746377.0,
      "step": 28740
    },
    {
      "entropy": 0.8290466409176588,
      "epoch": 0.8235108775045472,
      "grad_norm": 9.5625,
      "learning_rate": 3.248692303937382e-05,
      "loss": 0.8325183868408204,
      "mean_token_accuracy": 0.8113024167716503,
      "num_tokens": 56766457.0,
      "step": 28750
    },
    {
      "entropy": 0.8260959193110466,
      "epoch": 0.8237973160706358,
      "grad_norm": 8.125,
      "learning_rate": 3.2476034325727406e-05,
      "loss": 0.8485879898071289,
      "mean_token_accuracy": 0.8011265866458416,
      "num_tokens": 56787596.0,
      "step": 28760
    },
    {
      "entropy": 0.778990913927555,
      "epoch": 0.8240837546367243,
      "grad_norm": 10.125,
      "learning_rate": 3.24651440543545e-05,
      "loss": 0.7551025390625,
      "mean_token_accuracy": 0.8159596487879753,
      "num_tokens": 56807822.0,
      "step": 28770
    },
    {
      "entropy": 0.8110850658267736,
      "epoch": 0.8243701932028128,
      "grad_norm": 11.0,
      "learning_rate": 3.245425222752423e-05,
      "loss": 0.8402592658996582,
      "mean_token_accuracy": 0.8083339963108301,
      "num_tokens": 56828068.0,
      "step": 28780
    },
    {
      "entropy": 0.8704190145246684,
      "epoch": 0.8246566317689014,
      "grad_norm": 10.5625,
      "learning_rate": 3.244335884750606e-05,
      "loss": 0.9087417602539063,
      "mean_token_accuracy": 0.7925322979688645,
      "num_tokens": 56848827.0,
      "step": 28790
    },
    {
      "entropy": 0.7692789008840919,
      "epoch": 0.8249430703349899,
      "grad_norm": 8.3125,
      "learning_rate": 3.243246391656974e-05,
      "loss": 0.7886733055114746,
      "mean_token_accuracy": 0.8189909964799881,
      "num_tokens": 56867883.0,
      "step": 28800
    },
    {
      "entropy": 0.6994083169847727,
      "epoch": 0.8252295089010785,
      "grad_norm": 8.0,
      "learning_rate": 3.242156743698539e-05,
      "loss": 0.6890432834625244,
      "mean_token_accuracy": 0.835236270725727,
      "num_tokens": 56888214.0,
      "step": 28810
    },
    {
      "entropy": 0.7664016184397042,
      "epoch": 0.8255159474671669,
      "grad_norm": 8.625,
      "learning_rate": 3.2410669411023415e-05,
      "loss": 0.764271593093872,
      "mean_token_accuracy": 0.8205571450293064,
      "num_tokens": 56907604.0,
      "step": 28820
    },
    {
      "entropy": 0.7621973350644111,
      "epoch": 0.8258023860332555,
      "grad_norm": 7.5,
      "learning_rate": 3.2399769840954554e-05,
      "loss": 0.7871853828430175,
      "mean_token_accuracy": 0.8190434738993645,
      "num_tokens": 56926689.0,
      "step": 28830
    },
    {
      "entropy": 0.849034524988383,
      "epoch": 0.8260888245993441,
      "grad_norm": 12.1875,
      "learning_rate": 3.238886872904989e-05,
      "loss": 0.8808290481567382,
      "mean_token_accuracy": 0.7923092033714056,
      "num_tokens": 56946105.0,
      "step": 28840
    },
    {
      "entropy": 0.8557514196261764,
      "epoch": 0.8263752631654326,
      "grad_norm": 6.71875,
      "learning_rate": 3.237796607758078e-05,
      "loss": 0.8582379341125488,
      "mean_token_accuracy": 0.7992957204580307,
      "num_tokens": 56965702.0,
      "step": 28850
    },
    {
      "entropy": 0.7085797235369682,
      "epoch": 0.8266617017315211,
      "grad_norm": 5.75,
      "learning_rate": 3.236706188881895e-05,
      "loss": 0.706516170501709,
      "mean_token_accuracy": 0.8275632135570049,
      "num_tokens": 56985004.0,
      "step": 28860
    },
    {
      "entropy": 0.7767271500080823,
      "epoch": 0.8269481402976097,
      "grad_norm": 9.4375,
      "learning_rate": 3.2356156165036414e-05,
      "loss": 0.7479190349578857,
      "mean_token_accuracy": 0.8254701979458332,
      "num_tokens": 57003359.0,
      "step": 28870
    },
    {
      "entropy": 0.8262583189178259,
      "epoch": 0.8272345788636982,
      "grad_norm": 8.25,
      "learning_rate": 3.234524890850553e-05,
      "loss": 0.8163822174072266,
      "mean_token_accuracy": 0.8039794281125069,
      "num_tokens": 57023186.0,
      "step": 28880
    },
    {
      "entropy": 0.824275058042258,
      "epoch": 0.8275210174297868,
      "grad_norm": 9.25,
      "learning_rate": 3.2334340121498964e-05,
      "loss": 0.8308199882507324,
      "mean_token_accuracy": 0.8000341504812241,
      "num_tokens": 57043357.0,
      "step": 28890
    },
    {
      "entropy": 0.7880806102417409,
      "epoch": 0.8278074559958752,
      "grad_norm": 9.5625,
      "learning_rate": 3.232342980628968e-05,
      "loss": 0.8126429557800293,
      "mean_token_accuracy": 0.8105969198048115,
      "num_tokens": 57062215.0,
      "step": 28900
    },
    {
      "entropy": 0.7171701500192285,
      "epoch": 0.8280938945619638,
      "grad_norm": 8.0,
      "learning_rate": 3.2312517965151007e-05,
      "loss": 0.7064039230346679,
      "mean_token_accuracy": 0.8323098070919513,
      "num_tokens": 57083877.0,
      "step": 28910
    },
    {
      "entropy": 0.7919406086206436,
      "epoch": 0.8283803331280524,
      "grad_norm": 7.5625,
      "learning_rate": 3.2301604600356536e-05,
      "loss": 0.8065586090087891,
      "mean_token_accuracy": 0.8159829415380955,
      "num_tokens": 57103525.0,
      "step": 28920
    },
    {
      "entropy": 0.7674940312281251,
      "epoch": 0.8286667716941409,
      "grad_norm": 7.53125,
      "learning_rate": 3.2290689714180236e-05,
      "loss": 0.7834873676300049,
      "mean_token_accuracy": 0.8185291822999716,
      "num_tokens": 57123195.0,
      "step": 28930
    },
    {
      "entropy": 0.6776399201713502,
      "epoch": 0.8289532102602294,
      "grad_norm": 9.375,
      "learning_rate": 3.2279773308896336e-05,
      "loss": 0.6901066780090332,
      "mean_token_accuracy": 0.8391685344278812,
      "num_tokens": 57142673.0,
      "step": 28940
    },
    {
      "entropy": 0.8446305659599602,
      "epoch": 0.829239648826318,
      "grad_norm": 9.5625,
      "learning_rate": 3.2268855386779424e-05,
      "loss": 0.8691059112548828,
      "mean_token_accuracy": 0.804456577450037,
      "num_tokens": 57162096.0,
      "step": 28950
    },
    {
      "entropy": 0.8454586542211473,
      "epoch": 0.8295260873924065,
      "grad_norm": 8.75,
      "learning_rate": 3.225793595010438e-05,
      "loss": 0.9151178359985351,
      "mean_token_accuracy": 0.8009891044348478,
      "num_tokens": 57181544.0,
      "step": 28960
    },
    {
      "entropy": 0.7958985361270606,
      "epoch": 0.8298125259584951,
      "grad_norm": 8.625,
      "learning_rate": 3.224701500114641e-05,
      "loss": 0.8257027626037597,
      "mean_token_accuracy": 0.8114207804203033,
      "num_tokens": 57201976.0,
      "step": 28970
    },
    {
      "entropy": 0.7508324461057783,
      "epoch": 0.8300989645245835,
      "grad_norm": 10.4375,
      "learning_rate": 3.223609254218102e-05,
      "loss": 0.7666289806365967,
      "mean_token_accuracy": 0.8118197880685329,
      "num_tokens": 57222220.0,
      "step": 28980
    },
    {
      "entropy": 0.8517249467782676,
      "epoch": 0.8303854030906721,
      "grad_norm": 5.78125,
      "learning_rate": 3.222516857548407e-05,
      "loss": 0.8572571754455567,
      "mean_token_accuracy": 0.8018763616681099,
      "num_tokens": 57242467.0,
      "step": 28990
    },
    {
      "entropy": 0.8245821138843894,
      "epoch": 0.8306718416567607,
      "grad_norm": 9.0625,
      "learning_rate": 3.2214243103331675e-05,
      "loss": 0.7939984798431396,
      "mean_token_accuracy": 0.818734259530902,
      "num_tokens": 57262247.0,
      "step": 29000
    },
    {
      "entropy": 0.8628938066773116,
      "epoch": 0.8309582802228492,
      "grad_norm": 7.0625,
      "learning_rate": 3.220331612800032e-05,
      "loss": 0.8253108024597168,
      "mean_token_accuracy": 0.8032734628766776,
      "num_tokens": 57283528.0,
      "step": 29010
    },
    {
      "entropy": 0.8129822970367968,
      "epoch": 0.8312447187889378,
      "grad_norm": 7.8125,
      "learning_rate": 3.2192387651766765e-05,
      "loss": 0.7605215549468994,
      "mean_token_accuracy": 0.822435387596488,
      "num_tokens": 57303955.0,
      "step": 29020
    },
    {
      "entropy": 0.8088655080646276,
      "epoch": 0.8315311573550263,
      "grad_norm": 10.4375,
      "learning_rate": 3.21814576769081e-05,
      "loss": 0.8265468597412109,
      "mean_token_accuracy": 0.8128928136080503,
      "num_tokens": 57326556.0,
      "step": 29030
    },
    {
      "entropy": 0.8507853345945477,
      "epoch": 0.8318175959211148,
      "grad_norm": 13.0625,
      "learning_rate": 3.217052620570173e-05,
      "loss": 0.828614616394043,
      "mean_token_accuracy": 0.8098709601908922,
      "num_tokens": 57346665.0,
      "step": 29040
    },
    {
      "entropy": 0.7253714235499501,
      "epoch": 0.8321040344872034,
      "grad_norm": 10.3125,
      "learning_rate": 3.215959324042535e-05,
      "loss": 0.6951502799987793,
      "mean_token_accuracy": 0.828867943584919,
      "num_tokens": 57366497.0,
      "step": 29050
    },
    {
      "entropy": 0.7835259784013033,
      "epoch": 0.832390473053292,
      "grad_norm": 7.5625,
      "learning_rate": 3.214865878335701e-05,
      "loss": 0.7736464023590088,
      "mean_token_accuracy": 0.8104513548314571,
      "num_tokens": 57387538.0,
      "step": 29060
    },
    {
      "entropy": 0.8261747828684747,
      "epoch": 0.8326769116193804,
      "grad_norm": 9.0,
      "learning_rate": 3.213772283677501e-05,
      "loss": 0.8669047355651855,
      "mean_token_accuracy": 0.8016688678413629,
      "num_tokens": 57407327.0,
      "step": 29070
    },
    {
      "entropy": 0.7915076227858663,
      "epoch": 0.832963350185469,
      "grad_norm": 7.625,
      "learning_rate": 3.2126785402958006e-05,
      "loss": 0.788362455368042,
      "mean_token_accuracy": 0.8155329164117575,
      "num_tokens": 57427115.0,
      "step": 29080
    },
    {
      "entropy": 0.6921886822208763,
      "epoch": 0.8332497887515575,
      "grad_norm": 6.71875,
      "learning_rate": 3.2115846484184965e-05,
      "loss": 0.7455945014953613,
      "mean_token_accuracy": 0.8243699949234724,
      "num_tokens": 57447480.0,
      "step": 29090
    },
    {
      "entropy": 0.7629973012953997,
      "epoch": 0.8335362273176461,
      "grad_norm": 9.9375,
      "learning_rate": 3.210490608273513e-05,
      "loss": 0.7807924270629882,
      "mean_token_accuracy": 0.820216416567564,
      "num_tokens": 57467338.0,
      "step": 29100
    },
    {
      "entropy": 0.8359290927648544,
      "epoch": 0.8338226658837345,
      "grad_norm": 9.625,
      "learning_rate": 3.209396420088808e-05,
      "loss": 0.8548377990722656,
      "mean_token_accuracy": 0.8144516542553901,
      "num_tokens": 57486233.0,
      "step": 29110
    },
    {
      "entropy": 0.678508349135518,
      "epoch": 0.8341091044498231,
      "grad_norm": 6.5625,
      "learning_rate": 3.20830208409237e-05,
      "loss": 0.6754230499267578,
      "mean_token_accuracy": 0.84351711794734,
      "num_tokens": 57505667.0,
      "step": 29120
    },
    {
      "entropy": 0.7941045488230885,
      "epoch": 0.8343955430159117,
      "grad_norm": 11.1875,
      "learning_rate": 3.207207600512216e-05,
      "loss": 0.7737654685974121,
      "mean_token_accuracy": 0.8157783303409815,
      "num_tokens": 57524075.0,
      "step": 29130
    },
    {
      "entropy": 0.7165412522852421,
      "epoch": 0.8346819815820002,
      "grad_norm": 8.4375,
      "learning_rate": 3.206112969576397e-05,
      "loss": 0.7099976539611816,
      "mean_token_accuracy": 0.8289723888039588,
      "num_tokens": 57544530.0,
      "step": 29140
    },
    {
      "entropy": 0.7106779512017966,
      "epoch": 0.8349684201480887,
      "grad_norm": 8.4375,
      "learning_rate": 3.205018191512994e-05,
      "loss": 0.7253697872161865,
      "mean_token_accuracy": 0.8257729358971119,
      "num_tokens": 57566056.0,
      "step": 29150
    },
    {
      "entropy": 0.7406747503206134,
      "epoch": 0.8352548587141773,
      "grad_norm": 6.5625,
      "learning_rate": 3.203923266550116e-05,
      "loss": 0.7929279327392578,
      "mean_token_accuracy": 0.8131102744489909,
      "num_tokens": 57586737.0,
      "step": 29160
    },
    {
      "entropy": 0.7373269101604819,
      "epoch": 0.8355412972802658,
      "grad_norm": 8.625,
      "learning_rate": 3.2028281949159053e-05,
      "loss": 0.7766078472137451,
      "mean_token_accuracy": 0.826121573522687,
      "num_tokens": 57605376.0,
      "step": 29170
    },
    {
      "entropy": 0.8228349772747606,
      "epoch": 0.8358277358463544,
      "grad_norm": 8.0625,
      "learning_rate": 3.201732976838534e-05,
      "loss": 0.8680471420288086,
      "mean_token_accuracy": 0.7965134758502245,
      "num_tokens": 57624081.0,
      "step": 29180
    },
    {
      "entropy": 0.9185493508353829,
      "epoch": 0.8361141744124428,
      "grad_norm": 7.3125,
      "learning_rate": 3.200637612546205e-05,
      "loss": 0.9282817840576172,
      "mean_token_accuracy": 0.8032187689095736,
      "num_tokens": 57644531.0,
      "step": 29190
    },
    {
      "entropy": 0.743029922991991,
      "epoch": 0.8364006129785314,
      "grad_norm": 8.375,
      "learning_rate": 3.19954210226715e-05,
      "loss": 0.7601142883300781,
      "mean_token_accuracy": 0.8233107179403305,
      "num_tokens": 57665144.0,
      "step": 29200
    },
    {
      "entropy": 0.8051600575447082,
      "epoch": 0.83668705154462,
      "grad_norm": 9.3125,
      "learning_rate": 3.1984464462296346e-05,
      "loss": 0.7823023796081543,
      "mean_token_accuracy": 0.8154573202133178,
      "num_tokens": 57685220.0,
      "step": 29210
    },
    {
      "entropy": 0.7031362980604172,
      "epoch": 0.8369734901107085,
      "grad_norm": 6.40625,
      "learning_rate": 3.197350644661951e-05,
      "loss": 0.6761599063873291,
      "mean_token_accuracy": 0.8394243247807026,
      "num_tokens": 57704892.0,
      "step": 29220
    },
    {
      "entropy": 0.7564663738943637,
      "epoch": 0.837259928676797,
      "grad_norm": 13.375,
      "learning_rate": 3.1962546977924234e-05,
      "loss": 0.7057957172393798,
      "mean_token_accuracy": 0.8242655277252198,
      "num_tokens": 57724093.0,
      "step": 29230
    },
    {
      "entropy": 0.7181255374103784,
      "epoch": 0.8375463672428856,
      "grad_norm": 6.46875,
      "learning_rate": 3.195158605849407e-05,
      "loss": 0.7168997764587403,
      "mean_token_accuracy": 0.8363456908613444,
      "num_tokens": 57743800.0,
      "step": 29240
    },
    {
      "entropy": 0.8513247050344944,
      "epoch": 0.8378328058089741,
      "grad_norm": 10.375,
      "learning_rate": 3.194062369061286e-05,
      "loss": 0.8606356620788574,
      "mean_token_accuracy": 0.7992973003536463,
      "num_tokens": 57763566.0,
      "step": 29250
    },
    {
      "entropy": 0.7350227138027549,
      "epoch": 0.8381192443750627,
      "grad_norm": 9.0625,
      "learning_rate": 3.192965987656475e-05,
      "loss": 0.7816920757293702,
      "mean_token_accuracy": 0.8240187197923661,
      "num_tokens": 57783694.0,
      "step": 29260
    },
    {
      "entropy": 0.7392491979524494,
      "epoch": 0.8384056829411511,
      "grad_norm": 8.9375,
      "learning_rate": 3.191869461863419e-05,
      "loss": 0.7478960990905762,
      "mean_token_accuracy": 0.826750497892499,
      "num_tokens": 57802249.0,
      "step": 29270
    },
    {
      "entropy": 0.7357187330722809,
      "epoch": 0.8386921215072397,
      "grad_norm": 11.6875,
      "learning_rate": 3.190772791910594e-05,
      "loss": 0.7710955142974854,
      "mean_token_accuracy": 0.8198306538164616,
      "num_tokens": 57822310.0,
      "step": 29280
    },
    {
      "entropy": 0.7296189492568373,
      "epoch": 0.8389785600733283,
      "grad_norm": 7.4375,
      "learning_rate": 3.189675978026504e-05,
      "loss": 0.7288360595703125,
      "mean_token_accuracy": 0.8227283403277397,
      "num_tokens": 57841794.0,
      "step": 29290
    },
    {
      "entropy": 0.7932238206267357,
      "epoch": 0.8392649986394168,
      "grad_norm": 6.9375,
      "learning_rate": 3.188579020439684e-05,
      "loss": 0.771299409866333,
      "mean_token_accuracy": 0.8284369084984065,
      "num_tokens": 57860143.0,
      "step": 29300
    },
    {
      "entropy": 0.8306482901796699,
      "epoch": 0.8395514372055054,
      "grad_norm": 8.3125,
      "learning_rate": 3.1874819193787004e-05,
      "loss": 0.8477312088012695,
      "mean_token_accuracy": 0.8065820928663016,
      "num_tokens": 57879704.0,
      "step": 29310
    },
    {
      "entropy": 0.7541067880578339,
      "epoch": 0.8398378757715939,
      "grad_norm": 6.4375,
      "learning_rate": 3.186384675072146e-05,
      "loss": 0.7423059463500976,
      "mean_token_accuracy": 0.8279709845781327,
      "num_tokens": 57901445.0,
      "step": 29320
    },
    {
      "entropy": 0.766757283359766,
      "epoch": 0.8401243143376824,
      "grad_norm": 8.4375,
      "learning_rate": 3.185287287748647e-05,
      "loss": 0.7305764198303223,
      "mean_token_accuracy": 0.8299424115568399,
      "num_tokens": 57922188.0,
      "step": 29330
    },
    {
      "entropy": 0.7459492382593453,
      "epoch": 0.840410752903771,
      "grad_norm": 5.96875,
      "learning_rate": 3.184189757636859e-05,
      "loss": 0.7391788482666015,
      "mean_token_accuracy": 0.8236037932336331,
      "num_tokens": 57942454.0,
      "step": 29340
    },
    {
      "entropy": 0.7702073644846678,
      "epoch": 0.8406971914698596,
      "grad_norm": 8.9375,
      "learning_rate": 3.183092084965463e-05,
      "loss": 0.7896190166473389,
      "mean_token_accuracy": 0.8088330015540123,
      "num_tokens": 57965346.0,
      "step": 29350
    },
    {
      "entropy": 0.6787860548589378,
      "epoch": 0.840983630035948,
      "grad_norm": 6.71875,
      "learning_rate": 3.1819942699631764e-05,
      "loss": 0.7106056690216065,
      "mean_token_accuracy": 0.8334883838891983,
      "num_tokens": 57985634.0,
      "step": 29360
    },
    {
      "entropy": 0.848620287515223,
      "epoch": 0.8412700686020366,
      "grad_norm": 7.34375,
      "learning_rate": 3.180896312858741e-05,
      "loss": 0.8988020896911622,
      "mean_token_accuracy": 0.8035101570188999,
      "num_tokens": 58005419.0,
      "step": 29370
    },
    {
      "entropy": 0.6616977493977174,
      "epoch": 0.8415565071681251,
      "grad_norm": 7.46875,
      "learning_rate": 3.17979821388093e-05,
      "loss": 0.707094955444336,
      "mean_token_accuracy": 0.8374154612421989,
      "num_tokens": 58023931.0,
      "step": 29380
    },
    {
      "entropy": 0.7452078210189939,
      "epoch": 0.8418429457342137,
      "grad_norm": 8.375,
      "learning_rate": 3.178699973258548e-05,
      "loss": 0.8167258262634277,
      "mean_token_accuracy": 0.8132761023938656,
      "num_tokens": 58043022.0,
      "step": 29390
    },
    {
      "entropy": 0.6937616747803986,
      "epoch": 0.8421293843003022,
      "grad_norm": 9.1875,
      "learning_rate": 3.177601591220425e-05,
      "loss": 0.6656507015228271,
      "mean_token_accuracy": 0.8395255409181118,
      "num_tokens": 58062548.0,
      "step": 29400
    },
    {
      "entropy": 0.7489753557369113,
      "epoch": 0.8424158228663907,
      "grad_norm": 8.8125,
      "learning_rate": 3.176503067995424e-05,
      "loss": 0.8207554817199707,
      "mean_token_accuracy": 0.8116433970630169,
      "num_tokens": 58082839.0,
      "step": 29410
    },
    {
      "entropy": 0.6508860488422215,
      "epoch": 0.8427022614324793,
      "grad_norm": 7.5,
      "learning_rate": 3.175404403812436e-05,
      "loss": 0.6639792919158936,
      "mean_token_accuracy": 0.8384425722062587,
      "num_tokens": 58101865.0,
      "step": 29420
    },
    {
      "entropy": 0.7035850162617863,
      "epoch": 0.8429886999985678,
      "grad_norm": 7.8125,
      "learning_rate": 3.174305598900382e-05,
      "loss": 0.7123931884765625,
      "mean_token_accuracy": 0.8251034196466207,
      "num_tokens": 58120894.0,
      "step": 29430
    },
    {
      "entropy": 0.8038545677438378,
      "epoch": 0.8432751385646563,
      "grad_norm": 7.34375,
      "learning_rate": 3.1732066534882124e-05,
      "loss": 0.7898514747619629,
      "mean_token_accuracy": 0.8154408723115921,
      "num_tokens": 58141245.0,
      "step": 29440
    },
    {
      "entropy": 0.8122020891867578,
      "epoch": 0.8435615771307449,
      "grad_norm": 11.875,
      "learning_rate": 3.1721075678049036e-05,
      "loss": 0.8467793464660645,
      "mean_token_accuracy": 0.8070625845342875,
      "num_tokens": 58161216.0,
      "step": 29450
    },
    {
      "entropy": 0.8285332832485437,
      "epoch": 0.8438480156968334,
      "grad_norm": 10.375,
      "learning_rate": 3.1710083420794686e-05,
      "loss": 0.807187843322754,
      "mean_token_accuracy": 0.8108110353350639,
      "num_tokens": 58180683.0,
      "step": 29460
    },
    {
      "entropy": 0.7554002204909921,
      "epoch": 0.844134454262922,
      "grad_norm": 10.9375,
      "learning_rate": 3.169908976540941e-05,
      "loss": 0.7587087631225586,
      "mean_token_accuracy": 0.8232350915670394,
      "num_tokens": 58199961.0,
      "step": 29470
    },
    {
      "entropy": 0.7055379319004714,
      "epoch": 0.8444208928290104,
      "grad_norm": 9.5,
      "learning_rate": 3.168809471418389e-05,
      "loss": 0.7197497844696045,
      "mean_token_accuracy": 0.8358284704387188,
      "num_tokens": 58219745.0,
      "step": 29480
    },
    {
      "entropy": 0.7964072683826089,
      "epoch": 0.844707331395099,
      "grad_norm": 8.8125,
      "learning_rate": 3.167709826940909e-05,
      "loss": 0.7579833507537842,
      "mean_token_accuracy": 0.8169283546507359,
      "num_tokens": 58239798.0,
      "step": 29490
    },
    {
      "entropy": 0.706321632862091,
      "epoch": 0.8449937699611876,
      "grad_norm": 7.125,
      "learning_rate": 3.166610043337626e-05,
      "loss": 0.6792866230010987,
      "mean_token_accuracy": 0.8281148575246334,
      "num_tokens": 58258662.0,
      "step": 29500
    },
    {
      "entropy": 0.6932786353863776,
      "epoch": 0.8452802085272761,
      "grad_norm": 8.3125,
      "learning_rate": 3.1655101208376926e-05,
      "loss": 0.6874987602233886,
      "mean_token_accuracy": 0.842122595384717,
      "num_tokens": 58277502.0,
      "step": 29510
    },
    {
      "entropy": 0.7157686484977603,
      "epoch": 0.8455666470933646,
      "grad_norm": 7.53125,
      "learning_rate": 3.164410059670292e-05,
      "loss": 0.7199172496795654,
      "mean_token_accuracy": 0.8303503785282373,
      "num_tokens": 58297502.0,
      "step": 29520
    },
    {
      "entropy": 0.8119903335347771,
      "epoch": 0.8458530856594532,
      "grad_norm": 8.0625,
      "learning_rate": 3.163309860064636e-05,
      "loss": 0.8692126274108887,
      "mean_token_accuracy": 0.8046592935919762,
      "num_tokens": 58316476.0,
      "step": 29530
    },
    {
      "entropy": 0.7364863570779562,
      "epoch": 0.8461395242255417,
      "grad_norm": 7.875,
      "learning_rate": 3.162209522249966e-05,
      "loss": 0.728443956375122,
      "mean_token_accuracy": 0.8319874659180642,
      "num_tokens": 58337216.0,
      "step": 29540
    },
    {
      "entropy": 0.7561438208445906,
      "epoch": 0.8464259627916303,
      "grad_norm": 6.125,
      "learning_rate": 3.161109046455549e-05,
      "loss": 0.7627835750579834,
      "mean_token_accuracy": 0.8247303564101458,
      "num_tokens": 58357554.0,
      "step": 29550
    },
    {
      "entropy": 0.7783890915568918,
      "epoch": 0.8467124013577189,
      "grad_norm": 9.625,
      "learning_rate": 3.160008432910686e-05,
      "loss": 0.7792095184326172,
      "mean_token_accuracy": 0.8215398017317057,
      "num_tokens": 58377494.0,
      "step": 29560
    },
    {
      "entropy": 0.8347054060548544,
      "epoch": 0.8469988399238073,
      "grad_norm": 9.4375,
      "learning_rate": 3.158907681844702e-05,
      "loss": 0.8622361183166504,
      "mean_token_accuracy": 0.8008704412728548,
      "num_tokens": 58398122.0,
      "step": 29570
    },
    {
      "entropy": 0.7898128937929869,
      "epoch": 0.8472852784898959,
      "grad_norm": 11.75,
      "learning_rate": 3.1578067934869523e-05,
      "loss": 0.790475606918335,
      "mean_token_accuracy": 0.8117433078587055,
      "num_tokens": 58417631.0,
      "step": 29580
    },
    {
      "entropy": 0.8714401613920927,
      "epoch": 0.8475717170559844,
      "grad_norm": 7.21875,
      "learning_rate": 3.156705768066823e-05,
      "loss": 0.8754071235656739,
      "mean_token_accuracy": 0.7997871626168489,
      "num_tokens": 58436892.0,
      "step": 29590
    },
    {
      "entropy": 0.8440285982564092,
      "epoch": 0.847858155622073,
      "grad_norm": 8.4375,
      "learning_rate": 3.155604605813723e-05,
      "loss": 0.8943504333496094,
      "mean_token_accuracy": 0.8052049677819013,
      "num_tokens": 58456648.0,
      "step": 29600
    },
    {
      "entropy": 0.7448900604620576,
      "epoch": 0.8481445941881615,
      "grad_norm": 7.65625,
      "learning_rate": 3.154503306957097e-05,
      "loss": 0.7124160766601563,
      "mean_token_accuracy": 0.8396284341812134,
      "num_tokens": 58476357.0,
      "step": 29610
    },
    {
      "entropy": 0.873932268191129,
      "epoch": 0.84843103275425,
      "grad_norm": 10.75,
      "learning_rate": 3.1534018717264135e-05,
      "loss": 0.9113627433776855,
      "mean_token_accuracy": 0.7966835845261813,
      "num_tokens": 58496744.0,
      "step": 29620
    },
    {
      "entropy": 0.8021034531295299,
      "epoch": 0.8487174713203386,
      "grad_norm": 6.625,
      "learning_rate": 3.152300300351169e-05,
      "loss": 0.8754084587097168,
      "mean_token_accuracy": 0.8049545519053936,
      "num_tokens": 58516792.0,
      "step": 29630
    },
    {
      "entropy": 0.7801473496481777,
      "epoch": 0.8490039098864272,
      "grad_norm": 8.125,
      "learning_rate": 3.1511985930608914e-05,
      "loss": 0.7542544841766358,
      "mean_token_accuracy": 0.8133180722594261,
      "num_tokens": 58536193.0,
      "step": 29640
    },
    {
      "entropy": 0.7719281597994267,
      "epoch": 0.8492903484525156,
      "grad_norm": 6.6875,
      "learning_rate": 3.150096750085135e-05,
      "loss": 0.7703906059265136,
      "mean_token_accuracy": 0.8207990549504757,
      "num_tokens": 58554859.0,
      "step": 29650
    },
    {
      "entropy": 0.7150281745940447,
      "epoch": 0.8495767870186042,
      "grad_norm": 6.1875,
      "learning_rate": 3.148994771653482e-05,
      "loss": 0.6578533172607421,
      "mean_token_accuracy": 0.8320480920374393,
      "num_tokens": 58575060.0,
      "step": 29660
    },
    {
      "entropy": 0.7639662696048617,
      "epoch": 0.8498632255846927,
      "grad_norm": 9.25,
      "learning_rate": 3.1478926579955455e-05,
      "loss": 0.7797040939331055,
      "mean_token_accuracy": 0.8146282244473696,
      "num_tokens": 58593576.0,
      "step": 29670
    },
    {
      "entropy": 0.8075905228964985,
      "epoch": 0.8501496641507813,
      "grad_norm": 7.875,
      "learning_rate": 3.1467904093409636e-05,
      "loss": 0.803914737701416,
      "mean_token_accuracy": 0.821362116932869,
      "num_tokens": 58613521.0,
      "step": 29680
    },
    {
      "entropy": 0.8702132214792073,
      "epoch": 0.8504361027168698,
      "grad_norm": 7.40625,
      "learning_rate": 3.145688025919403e-05,
      "loss": 0.8660801887512207,
      "mean_token_accuracy": 0.8040414817631245,
      "num_tokens": 58633521.0,
      "step": 29690
    },
    {
      "entropy": 0.8223677053116262,
      "epoch": 0.8507225412829583,
      "grad_norm": 8.5625,
      "learning_rate": 3.14458550796056e-05,
      "loss": 0.8386767387390137,
      "mean_token_accuracy": 0.8053471349179745,
      "num_tokens": 58653572.0,
      "step": 29700
    },
    {
      "entropy": 0.8052819548174739,
      "epoch": 0.8510089798490469,
      "grad_norm": 9.0625,
      "learning_rate": 3.143482855694159e-05,
      "loss": 0.7941282272338868,
      "mean_token_accuracy": 0.817620062455535,
      "num_tokens": 58672872.0,
      "step": 29710
    },
    {
      "entropy": 0.681879812432453,
      "epoch": 0.8512954184151355,
      "grad_norm": 8.125,
      "learning_rate": 3.14238006934995e-05,
      "loss": 0.6477042198181152,
      "mean_token_accuracy": 0.8514641031622887,
      "num_tokens": 58693722.0,
      "step": 29720
    },
    {
      "entropy": 0.7587384727783502,
      "epoch": 0.8515818569812239,
      "grad_norm": 8.125,
      "learning_rate": 3.1412771491577135e-05,
      "loss": 0.7373646259307861,
      "mean_token_accuracy": 0.8188790328800678,
      "num_tokens": 58713507.0,
      "step": 29730
    },
    {
      "entropy": 0.8914771698415279,
      "epoch": 0.8518682955473125,
      "grad_norm": 10.0,
      "learning_rate": 3.1401740953472575e-05,
      "loss": 0.9002939224243164,
      "mean_token_accuracy": 0.8048862464725971,
      "num_tokens": 58733235.0,
      "step": 29740
    },
    {
      "entropy": 0.6263743080198765,
      "epoch": 0.852154734113401,
      "grad_norm": 13.625,
      "learning_rate": 3.139070908148415e-05,
      "loss": 0.634743070602417,
      "mean_token_accuracy": 0.846338263899088,
      "num_tokens": 58751925.0,
      "step": 29750
    },
    {
      "entropy": 0.7055350361391902,
      "epoch": 0.8524411726794896,
      "grad_norm": 4.46875,
      "learning_rate": 3.137967587791052e-05,
      "loss": 0.7316250324249267,
      "mean_token_accuracy": 0.8308370128273964,
      "num_tokens": 58771499.0,
      "step": 29760
    },
    {
      "entropy": 0.7560039680451155,
      "epoch": 0.852727611245578,
      "grad_norm": 8.5625,
      "learning_rate": 3.1368641345050576e-05,
      "loss": 0.7966852188110352,
      "mean_token_accuracy": 0.8252310365438461,
      "num_tokens": 58790374.0,
      "step": 29770
    },
    {
      "entropy": 0.6503319650888443,
      "epoch": 0.8530140498116666,
      "grad_norm": 12.8125,
      "learning_rate": 3.13576054852035e-05,
      "loss": 0.6839728355407715,
      "mean_token_accuracy": 0.8427787020802497,
      "num_tokens": 58808938.0,
      "step": 29780
    },
    {
      "entropy": 0.7183228876441717,
      "epoch": 0.8533004883777552,
      "grad_norm": 8.4375,
      "learning_rate": 3.134656830066875e-05,
      "loss": 0.7589735984802246,
      "mean_token_accuracy": 0.8266226522624492,
      "num_tokens": 58828314.0,
      "step": 29790
    },
    {
      "entropy": 0.7297594799660146,
      "epoch": 0.8535869269438437,
      "grad_norm": 6.6875,
      "learning_rate": 3.1335529793746075e-05,
      "loss": 0.7116528511047363,
      "mean_token_accuracy": 0.8314943522214889,
      "num_tokens": 58847819.0,
      "step": 29800
    },
    {
      "entropy": 0.7228276317939162,
      "epoch": 0.8538733655099322,
      "grad_norm": 7.78125,
      "learning_rate": 3.132448996673549e-05,
      "loss": 0.7128180980682373,
      "mean_token_accuracy": 0.8297586496919394,
      "num_tokens": 58867258.0,
      "step": 29810
    },
    {
      "entropy": 0.7602316838689148,
      "epoch": 0.8541598040760208,
      "grad_norm": 7.96875,
      "learning_rate": 3.1313448821937274e-05,
      "loss": 0.7543773651123047,
      "mean_token_accuracy": 0.820031663402915,
      "num_tokens": 58887613.0,
      "step": 29820
    },
    {
      "entropy": 0.867661448288709,
      "epoch": 0.8544462426421093,
      "grad_norm": 8.25,
      "learning_rate": 3.130240636165199e-05,
      "loss": 0.8908706665039062,
      "mean_token_accuracy": 0.7938743144273758,
      "num_tokens": 58909524.0,
      "step": 29830
    },
    {
      "entropy": 0.8111098986119032,
      "epoch": 0.8547326812081979,
      "grad_norm": 8.625,
      "learning_rate": 3.1291362588180484e-05,
      "loss": 0.7645383358001709,
      "mean_token_accuracy": 0.8261738441884517,
      "num_tokens": 58929868.0,
      "step": 29840
    },
    {
      "entropy": 0.9123094631824642,
      "epoch": 0.8550191197742865,
      "grad_norm": 7.4375,
      "learning_rate": 3.128031750382385e-05,
      "loss": 0.9809228897094726,
      "mean_token_accuracy": 0.7748530812561512,
      "num_tokens": 58950213.0,
      "step": 29850
    },
    {
      "entropy": 0.7807058731094003,
      "epoch": 0.8553055583403749,
      "grad_norm": 7.90625,
      "learning_rate": 3.126927111088349e-05,
      "loss": 0.8220767021179199,
      "mean_token_accuracy": 0.8242660365998745,
      "num_tokens": 58969458.0,
      "step": 29860
    },
    {
      "entropy": 0.8383877983316779,
      "epoch": 0.8555919969064635,
      "grad_norm": 6.53125,
      "learning_rate": 3.125822341166104e-05,
      "loss": 0.8202462196350098,
      "mean_token_accuracy": 0.8145148970186711,
      "num_tokens": 58989053.0,
      "step": 29870
    },
    {
      "entropy": 0.6999437553808093,
      "epoch": 0.855878435472552,
      "grad_norm": 7.53125,
      "learning_rate": 3.1247174408458434e-05,
      "loss": 0.6929960727691651,
      "mean_token_accuracy": 0.8402774814516306,
      "num_tokens": 59008133.0,
      "step": 29880
    },
    {
      "entropy": 0.8392708515748382,
      "epoch": 0.8561648740386406,
      "grad_norm": 9.625,
      "learning_rate": 3.123612410357789e-05,
      "loss": 0.8333162307739258,
      "mean_token_accuracy": 0.8147697720676661,
      "num_tokens": 59029137.0,
      "step": 29890
    },
    {
      "entropy": 0.7976293741725385,
      "epoch": 0.8564513126047291,
      "grad_norm": 10.75,
      "learning_rate": 3.1225072499321854e-05,
      "loss": 0.8260221481323242,
      "mean_token_accuracy": 0.8095488131046296,
      "num_tokens": 59048534.0,
      "step": 29900
    },
    {
      "entropy": 0.8414044523611665,
      "epoch": 0.8567377511708176,
      "grad_norm": 6.65625,
      "learning_rate": 3.1214019597993085e-05,
      "loss": 0.8093011856079102,
      "mean_token_accuracy": 0.8116516247391701,
      "num_tokens": 59067847.0,
      "step": 29910
    },
    {
      "entropy": 0.7112448582425713,
      "epoch": 0.8570241897369062,
      "grad_norm": 10.375,
      "learning_rate": 3.1202965401894586e-05,
      "loss": 0.7427616596221924,
      "mean_token_accuracy": 0.8258505128324032,
      "num_tokens": 59088715.0,
      "step": 29920
    },
    {
      "entropy": 0.7070449813269079,
      "epoch": 0.8573106283029948,
      "grad_norm": 8.5,
      "learning_rate": 3.119190991332964e-05,
      "loss": 0.7183035850524903,
      "mean_token_accuracy": 0.8281891718506813,
      "num_tokens": 59109122.0,
      "step": 29930
    },
    {
      "entropy": 0.7945078454911709,
      "epoch": 0.8575970668690832,
      "grad_norm": 8.75,
      "learning_rate": 3.1180853134601805e-05,
      "loss": 0.8091774940490722,
      "mean_token_accuracy": 0.8144068978726864,
      "num_tokens": 59128079.0,
      "step": 29940
    },
    {
      "entropy": 0.7561862553935498,
      "epoch": 0.8578835054351718,
      "grad_norm": 6.84375,
      "learning_rate": 3.116979506801489e-05,
      "loss": 0.7737064361572266,
      "mean_token_accuracy": 0.8270713735371829,
      "num_tokens": 59147798.0,
      "step": 29950
    },
    {
      "entropy": 0.6933464638888835,
      "epoch": 0.8581699440012603,
      "grad_norm": 9.125,
      "learning_rate": 3.115873571587299e-05,
      "loss": 0.6441201210021973,
      "mean_token_accuracy": 0.8448137640953064,
      "num_tokens": 59167467.0,
      "step": 29960
    },
    {
      "entropy": 0.6791793362237513,
      "epoch": 0.8584563825673489,
      "grad_norm": 9.0,
      "learning_rate": 3.114767508048047e-05,
      "loss": 0.6966997623443604,
      "mean_token_accuracy": 0.8284056231379509,
      "num_tokens": 59187055.0,
      "step": 29970
    },
    {
      "entropy": 0.7402877224609256,
      "epoch": 0.8587428211334374,
      "grad_norm": 9.375,
      "learning_rate": 3.113661316414193e-05,
      "loss": 0.7743849754333496,
      "mean_token_accuracy": 0.8210965856909752,
      "num_tokens": 59206848.0,
      "step": 29980
    },
    {
      "entropy": 0.7097019250504673,
      "epoch": 0.8590292596995259,
      "grad_norm": 7.59375,
      "learning_rate": 3.112554996916229e-05,
      "loss": 0.7464244842529297,
      "mean_token_accuracy": 0.8353237647563219,
      "num_tokens": 59227428.0,
      "step": 29990
    },
    {
      "entropy": 0.8087463481351733,
      "epoch": 0.8593156982656145,
      "grad_norm": 8.4375,
      "learning_rate": 3.111448549784667e-05,
      "loss": 0.8480051994323731,
      "mean_token_accuracy": 0.8011137254536151,
      "num_tokens": 59247267.0,
      "step": 30000
    },
    {
      "entropy": 0.7890509435907006,
      "epoch": 0.8596021368317031,
      "grad_norm": 8.8125,
      "learning_rate": 3.110341975250053e-05,
      "loss": 0.7484670162200928,
      "mean_token_accuracy": 0.8226761814206839,
      "num_tokens": 59265610.0,
      "step": 30010
    },
    {
      "entropy": 0.8186336108483374,
      "epoch": 0.8598885753977915,
      "grad_norm": 7.46875,
      "learning_rate": 3.109235273542954e-05,
      "loss": 0.8160028457641602,
      "mean_token_accuracy": 0.8097330272197724,
      "num_tokens": 59284473.0,
      "step": 30020
    },
    {
      "entropy": 0.7684104894287884,
      "epoch": 0.8601750139638801,
      "grad_norm": 9.5,
      "learning_rate": 3.108128444893965e-05,
      "loss": 0.7655003547668457,
      "mean_token_accuracy": 0.8185867525637149,
      "num_tokens": 59304070.0,
      "step": 30030
    },
    {
      "entropy": 0.6820543326437474,
      "epoch": 0.8604614525299686,
      "grad_norm": 14.0,
      "learning_rate": 3.1070214895337086e-05,
      "loss": 0.6808469772338868,
      "mean_token_accuracy": 0.8350047565996647,
      "num_tokens": 59323070.0,
      "step": 30040
    },
    {
      "entropy": 0.7389034054242074,
      "epoch": 0.8607478910960572,
      "grad_norm": 8.5625,
      "learning_rate": 3.105914407692833e-05,
      "loss": 0.7344375133514405,
      "mean_token_accuracy": 0.8260802514851093,
      "num_tokens": 59343499.0,
      "step": 30050
    },
    {
      "entropy": 0.8010847726836801,
      "epoch": 0.8610343296621457,
      "grad_norm": 11.3125,
      "learning_rate": 3.104807199602011e-05,
      "loss": 0.8188401222229004,
      "mean_token_accuracy": 0.8137032255530358,
      "num_tokens": 59362512.0,
      "step": 30060
    },
    {
      "entropy": 0.7424802869558335,
      "epoch": 0.8613207682282342,
      "grad_norm": 11.1875,
      "learning_rate": 3.103699865491947e-05,
      "loss": 0.7395939826965332,
      "mean_token_accuracy": 0.8249667797237634,
      "num_tokens": 59382459.0,
      "step": 30070
    },
    {
      "entropy": 0.7593066449277103,
      "epoch": 0.8616072067943228,
      "grad_norm": 9.3125,
      "learning_rate": 3.102592405593365e-05,
      "loss": 0.7773125648498536,
      "mean_token_accuracy": 0.8172085732221603,
      "num_tokens": 59403275.0,
      "step": 30080
    },
    {
      "entropy": 0.7122728399001061,
      "epoch": 0.8618936453604114,
      "grad_norm": 7.53125,
      "learning_rate": 3.101484820137019e-05,
      "loss": 0.765667724609375,
      "mean_token_accuracy": 0.8176973961293698,
      "num_tokens": 59422183.0,
      "step": 30090
    },
    {
      "entropy": 0.7277290469035507,
      "epoch": 0.8621800839264999,
      "grad_norm": 8.1875,
      "learning_rate": 3.100377109353688e-05,
      "loss": 0.7209650993347168,
      "mean_token_accuracy": 0.8400656208395958,
      "num_tokens": 59442017.0,
      "step": 30100
    },
    {
      "entropy": 0.7732758129015564,
      "epoch": 0.8624665224925884,
      "grad_norm": 8.625,
      "learning_rate": 3.099269273474179e-05,
      "loss": 0.7924113750457764,
      "mean_token_accuracy": 0.8099720265716315,
      "num_tokens": 59461709.0,
      "step": 30110
    },
    {
      "entropy": 0.8202110813930631,
      "epoch": 0.8627529610586769,
      "grad_norm": 8.5,
      "learning_rate": 3.098161312729324e-05,
      "loss": 0.7806685447692872,
      "mean_token_accuracy": 0.8135755136609077,
      "num_tokens": 59481720.0,
      "step": 30120
    },
    {
      "entropy": 0.7105918015353382,
      "epoch": 0.8630393996247655,
      "grad_norm": 5.84375,
      "learning_rate": 3.097053227349978e-05,
      "loss": 0.7195912837982178,
      "mean_token_accuracy": 0.833269190788269,
      "num_tokens": 59500307.0,
      "step": 30130
    },
    {
      "entropy": 0.6488443666137755,
      "epoch": 0.8633258381908541,
      "grad_norm": 8.375,
      "learning_rate": 3.095945017567027e-05,
      "loss": 0.7129469394683838,
      "mean_token_accuracy": 0.8309390552341938,
      "num_tokens": 59519855.0,
      "step": 30140
    },
    {
      "entropy": 0.6797159533016384,
      "epoch": 0.8636122767569425,
      "grad_norm": 8.25,
      "learning_rate": 3.09483668361138e-05,
      "loss": 0.7184768199920655,
      "mean_token_accuracy": 0.8401269085705281,
      "num_tokens": 59538781.0,
      "step": 30150
    },
    {
      "entropy": 0.7732895592227578,
      "epoch": 0.8638987153230311,
      "grad_norm": 7.71875,
      "learning_rate": 3.093728225713972e-05,
      "loss": 0.7819478988647461,
      "mean_token_accuracy": 0.8132290545850992,
      "num_tokens": 59558165.0,
      "step": 30160
    },
    {
      "entropy": 0.7054218160919845,
      "epoch": 0.8641851538891197,
      "grad_norm": 9.0,
      "learning_rate": 3.092619644105764e-05,
      "loss": 0.7119922637939453,
      "mean_token_accuracy": 0.828713946044445,
      "num_tokens": 59578775.0,
      "step": 30170
    },
    {
      "entropy": 0.7059946920722723,
      "epoch": 0.8644715924552082,
      "grad_norm": 5.90625,
      "learning_rate": 3.091510939017744e-05,
      "loss": 0.6921760082244873,
      "mean_token_accuracy": 0.8393604807555676,
      "num_tokens": 59598523.0,
      "step": 30180
    },
    {
      "entropy": 0.7441066274419427,
      "epoch": 0.8647580310212967,
      "grad_norm": 10.1875,
      "learning_rate": 3.090402110680925e-05,
      "loss": 0.7504639625549316,
      "mean_token_accuracy": 0.8217106312513351,
      "num_tokens": 59618271.0,
      "step": 30190
    },
    {
      "entropy": 0.8230331391096115,
      "epoch": 0.8650444695873852,
      "grad_norm": 7.1875,
      "learning_rate": 3.089293159326343e-05,
      "loss": 0.8328859329223632,
      "mean_token_accuracy": 0.8032637715339661,
      "num_tokens": 59638899.0,
      "step": 30200
    },
    {
      "entropy": 0.6077782147563994,
      "epoch": 0.8653309081534738,
      "grad_norm": 5.5625,
      "learning_rate": 3.088184085185065e-05,
      "loss": 0.5731133937835693,
      "mean_token_accuracy": 0.8662423215806484,
      "num_tokens": 59656804.0,
      "step": 30210
    },
    {
      "entropy": 0.6921277226880193,
      "epoch": 0.8656173467195624,
      "grad_norm": 10.6875,
      "learning_rate": 3.087074888488179e-05,
      "loss": 0.6871782302856445,
      "mean_token_accuracy": 0.830440815538168,
      "num_tokens": 59676124.0,
      "step": 30220
    },
    {
      "entropy": 0.6663014395162463,
      "epoch": 0.8659037852856508,
      "grad_norm": 12.375,
      "learning_rate": 3.0859655694668e-05,
      "loss": 0.6427361011505127,
      "mean_token_accuracy": 0.846822426840663,
      "num_tokens": 59695351.0,
      "step": 30230
    },
    {
      "entropy": 0.7187425933778286,
      "epoch": 0.8661902238517394,
      "grad_norm": 7.5,
      "learning_rate": 3.0848561283520706e-05,
      "loss": 0.7347735404968262,
      "mean_token_accuracy": 0.8389670893549919,
      "num_tokens": 59715467.0,
      "step": 30240
    },
    {
      "entropy": 0.8130124798044562,
      "epoch": 0.866476662417828,
      "grad_norm": 6.3125,
      "learning_rate": 3.083746565375154e-05,
      "loss": 0.7934590816497803,
      "mean_token_accuracy": 0.8166183460503816,
      "num_tokens": 59735247.0,
      "step": 30250
    },
    {
      "entropy": 0.6907448563724756,
      "epoch": 0.8667631009839165,
      "grad_norm": 12.5625,
      "learning_rate": 3.082636880767244e-05,
      "loss": 0.6908100605010986,
      "mean_token_accuracy": 0.8347482241690158,
      "num_tokens": 59754915.0,
      "step": 30260
    },
    {
      "entropy": 0.7391698568128049,
      "epoch": 0.867049539550005,
      "grad_norm": 10.25,
      "learning_rate": 3.081527074759557e-05,
      "loss": 0.7797351837158203,
      "mean_token_accuracy": 0.8219196606427431,
      "num_tokens": 59774510.0,
      "step": 30270
    },
    {
      "entropy": 0.7512252605520189,
      "epoch": 0.8673359781160935,
      "grad_norm": 5.4375,
      "learning_rate": 3.080417147583334e-05,
      "loss": 0.7478962421417237,
      "mean_token_accuracy": 0.830263139307499,
      "num_tokens": 59793510.0,
      "step": 30280
    },
    {
      "entropy": 0.723433973453939,
      "epoch": 0.8676224166821821,
      "grad_norm": 9.875,
      "learning_rate": 3.0793070994698443e-05,
      "loss": 0.7361720085144043,
      "mean_token_accuracy": 0.8414429426193237,
      "num_tokens": 59812650.0,
      "step": 30290
    },
    {
      "entropy": 0.8634285374544561,
      "epoch": 0.8679088552482707,
      "grad_norm": 8.3125,
      "learning_rate": 3.078196930650379e-05,
      "loss": 0.8988590240478516,
      "mean_token_accuracy": 0.7933920502662659,
      "num_tokens": 59831961.0,
      "step": 30300
    },
    {
      "entropy": 0.7539684932678938,
      "epoch": 0.8681952938143591,
      "grad_norm": 9.125,
      "learning_rate": 3.077086641356256e-05,
      "loss": 0.7947973728179931,
      "mean_token_accuracy": 0.8270232316106558,
      "num_tokens": 59852187.0,
      "step": 30310
    },
    {
      "entropy": 0.7588585375808179,
      "epoch": 0.8684817323804477,
      "grad_norm": 7.15625,
      "learning_rate": 3.075976231818819e-05,
      "loss": 0.7555005073547363,
      "mean_token_accuracy": 0.8193250134587288,
      "num_tokens": 59872096.0,
      "step": 30320
    },
    {
      "entropy": 0.7501446263864636,
      "epoch": 0.8687681709465362,
      "grad_norm": 5.71875,
      "learning_rate": 3.074865702269435e-05,
      "loss": 0.720142412185669,
      "mean_token_accuracy": 0.8408188410103321,
      "num_tokens": 59891995.0,
      "step": 30330
    },
    {
      "entropy": 0.7548750305548311,
      "epoch": 0.8690546095126248,
      "grad_norm": 9.125,
      "learning_rate": 3.0737550529394974e-05,
      "loss": 0.7815385341644288,
      "mean_token_accuracy": 0.8103883296251297,
      "num_tokens": 59912277.0,
      "step": 30340
    },
    {
      "entropy": 0.7323271587491036,
      "epoch": 0.8693410480787133,
      "grad_norm": 9.25,
      "learning_rate": 3.0726442840604244e-05,
      "loss": 0.739259910583496,
      "mean_token_accuracy": 0.8307231415063143,
      "num_tokens": 59930995.0,
      "step": 30350
    },
    {
      "entropy": 0.8113714377395809,
      "epoch": 0.8696274866448018,
      "grad_norm": 10.5,
      "learning_rate": 3.071533395863657e-05,
      "loss": 0.8654248237609863,
      "mean_token_accuracy": 0.8046215593814849,
      "num_tokens": 59951011.0,
      "step": 30360
    },
    {
      "entropy": 0.7803494727239013,
      "epoch": 0.8699139252108904,
      "grad_norm": 10.0,
      "learning_rate": 3.070422388580665e-05,
      "loss": 0.8025836944580078,
      "mean_token_accuracy": 0.8116781808435917,
      "num_tokens": 59970388.0,
      "step": 30370
    },
    {
      "entropy": 0.7634242146275938,
      "epoch": 0.870200363776979,
      "grad_norm": 10.5,
      "learning_rate": 3.06931126244294e-05,
      "loss": 0.7575103759765625,
      "mean_token_accuracy": 0.8196574464440346,
      "num_tokens": 59990676.0,
      "step": 30380
    },
    {
      "entropy": 0.8157408440485596,
      "epoch": 0.8704868023430675,
      "grad_norm": 10.875,
      "learning_rate": 3.068200017681999e-05,
      "loss": 0.8240473747253418,
      "mean_token_accuracy": 0.8153008233755827,
      "num_tokens": 60011412.0,
      "step": 30390
    },
    {
      "entropy": 0.7637706554494799,
      "epoch": 0.870773240909156,
      "grad_norm": 6.34375,
      "learning_rate": 3.0670886545293845e-05,
      "loss": 0.7706390857696533,
      "mean_token_accuracy": 0.8275805495679378,
      "num_tokens": 60031496.0,
      "step": 30400
    },
    {
      "entropy": 0.7497460615821183,
      "epoch": 0.8710596794752445,
      "grad_norm": 5.53125,
      "learning_rate": 3.0659771732166614e-05,
      "loss": 0.7561752319335937,
      "mean_token_accuracy": 0.8217258058488369,
      "num_tokens": 60051293.0,
      "step": 30410
    },
    {
      "entropy": 0.7476054855622352,
      "epoch": 0.8713461180413331,
      "grad_norm": 11.125,
      "learning_rate": 3.064865573975423e-05,
      "loss": 0.7450579643249512,
      "mean_token_accuracy": 0.8248162738978863,
      "num_tokens": 60071024.0,
      "step": 30420
    },
    {
      "entropy": 0.7381102896295488,
      "epoch": 0.8716325566074217,
      "grad_norm": 10.875,
      "learning_rate": 3.063753857037285e-05,
      "loss": 0.8093603134155274,
      "mean_token_accuracy": 0.8167049948126077,
      "num_tokens": 60091100.0,
      "step": 30430
    },
    {
      "entropy": 0.7689142506569624,
      "epoch": 0.8719189951735101,
      "grad_norm": 9.375,
      "learning_rate": 3.062642022633886e-05,
      "loss": 0.8225009918212891,
      "mean_token_accuracy": 0.812428206577897,
      "num_tokens": 60112325.0,
      "step": 30440
    },
    {
      "entropy": 0.7268811858259141,
      "epoch": 0.8722054337395987,
      "grad_norm": 9.1875,
      "learning_rate": 3.0615300709968927e-05,
      "loss": 0.7163402557373046,
      "mean_token_accuracy": 0.8231858421117068,
      "num_tokens": 60130899.0,
      "step": 30450
    },
    {
      "entropy": 0.7447432201355696,
      "epoch": 0.8724918723056873,
      "grad_norm": 6.625,
      "learning_rate": 3.0604180023579926e-05,
      "loss": 0.6999099731445313,
      "mean_token_accuracy": 0.8356437928974628,
      "num_tokens": 60150785.0,
      "step": 30460
    },
    {
      "entropy": 0.8216216952539981,
      "epoch": 0.8727783108717758,
      "grad_norm": 11.125,
      "learning_rate": 3.0593058169489004e-05,
      "loss": 0.818881893157959,
      "mean_token_accuracy": 0.8060849711298943,
      "num_tokens": 60169630.0,
      "step": 30470
    },
    {
      "entropy": 0.6953288790769875,
      "epoch": 0.8730647494378643,
      "grad_norm": 8.4375,
      "learning_rate": 3.0581935150013535e-05,
      "loss": 0.6562338829040527,
      "mean_token_accuracy": 0.8441466689109802,
      "num_tokens": 60188783.0,
      "step": 30480
    },
    {
      "entropy": 0.7925171073526144,
      "epoch": 0.8733511880039528,
      "grad_norm": 7.3125,
      "learning_rate": 3.057081096747115e-05,
      "loss": 0.7840872287750245,
      "mean_token_accuracy": 0.8212714832276106,
      "num_tokens": 60209680.0,
      "step": 30490
    },
    {
      "entropy": 0.7325895581394434,
      "epoch": 0.8736376265700414,
      "grad_norm": 9.375,
      "learning_rate": 3.05596856241797e-05,
      "loss": 0.7623010635375976,
      "mean_token_accuracy": 0.8194494169205427,
      "num_tokens": 60228758.0,
      "step": 30500
    },
    {
      "entropy": 0.6919432799564674,
      "epoch": 0.87392406513613,
      "grad_norm": 6.25,
      "learning_rate": 3.05485591224573e-05,
      "loss": 0.6616633892059326,
      "mean_token_accuracy": 0.8463066637516021,
      "num_tokens": 60248417.0,
      "step": 30510
    },
    {
      "entropy": 0.7577609959058463,
      "epoch": 0.8742105037022184,
      "grad_norm": 9.9375,
      "learning_rate": 3.05374314646223e-05,
      "loss": 0.8477311134338379,
      "mean_token_accuracy": 0.8038159959018231,
      "num_tokens": 60266774.0,
      "step": 30520
    },
    {
      "entropy": 0.724104695674032,
      "epoch": 0.874496942268307,
      "grad_norm": 10.6875,
      "learning_rate": 3.0526302652993276e-05,
      "loss": 0.738587474822998,
      "mean_token_accuracy": 0.8188330203294754,
      "num_tokens": 60286500.0,
      "step": 30530
    },
    {
      "entropy": 0.7668314936105162,
      "epoch": 0.8747833808343956,
      "grad_norm": 13.5,
      "learning_rate": 3.051517268988907e-05,
      "loss": 0.8029433250427246,
      "mean_token_accuracy": 0.8261900506913662,
      "num_tokens": 60307656.0,
      "step": 30540
    },
    {
      "entropy": 0.6802383087575435,
      "epoch": 0.8750698194004841,
      "grad_norm": 9.75,
      "learning_rate": 3.0504041577628745e-05,
      "loss": 0.6897707939147949,
      "mean_token_accuracy": 0.8369365632534027,
      "num_tokens": 60326173.0,
      "step": 30550
    },
    {
      "entropy": 0.7740585630759597,
      "epoch": 0.8753562579665726,
      "grad_norm": 6.625,
      "learning_rate": 3.0492909318531604e-05,
      "loss": 0.7605361938476562,
      "mean_token_accuracy": 0.8229627441614866,
      "num_tokens": 60346096.0,
      "step": 30560
    },
    {
      "entropy": 0.7840164438821375,
      "epoch": 0.8756426965326611,
      "grad_norm": 6.96875,
      "learning_rate": 3.048177591491721e-05,
      "loss": 0.784083366394043,
      "mean_token_accuracy": 0.8242418687790632,
      "num_tokens": 60365249.0,
      "step": 30570
    },
    {
      "entropy": 0.8256821922957898,
      "epoch": 0.8759291350987497,
      "grad_norm": 6.78125,
      "learning_rate": 3.0470641369105335e-05,
      "loss": 0.833033275604248,
      "mean_token_accuracy": 0.8103102602064609,
      "num_tokens": 60384833.0,
      "step": 30580
    },
    {
      "entropy": 0.7147727408446372,
      "epoch": 0.8762155736648383,
      "grad_norm": 12.375,
      "learning_rate": 3.0459505683416013e-05,
      "loss": 0.7261588096618652,
      "mean_token_accuracy": 0.8306484211236238,
      "num_tokens": 60404806.0,
      "step": 30590
    },
    {
      "entropy": 0.8441930187866091,
      "epoch": 0.8765020122309267,
      "grad_norm": 7.65625,
      "learning_rate": 3.0448368860169484e-05,
      "loss": 0.9191892623901368,
      "mean_token_accuracy": 0.7935157366096973,
      "num_tokens": 60426421.0,
      "step": 30600
    },
    {
      "entropy": 0.7927142448723317,
      "epoch": 0.8767884507970153,
      "grad_norm": 7.375,
      "learning_rate": 3.0437230901686277e-05,
      "loss": 0.8189248085021973,
      "mean_token_accuracy": 0.8152483072131872,
      "num_tokens": 60446908.0,
      "step": 30610
    },
    {
      "entropy": 0.7185558651573956,
      "epoch": 0.8770748893631038,
      "grad_norm": 15.6875,
      "learning_rate": 3.0426091810287105e-05,
      "loss": 0.7604306221008301,
      "mean_token_accuracy": 0.8188759844750166,
      "num_tokens": 60465624.0,
      "step": 30620
    },
    {
      "entropy": 0.7620188863016665,
      "epoch": 0.8773613279291924,
      "grad_norm": 10.5,
      "learning_rate": 3.041495158829294e-05,
      "loss": 0.7399320602416992,
      "mean_token_accuracy": 0.8275215137749911,
      "num_tokens": 60484169.0,
      "step": 30630
    },
    {
      "entropy": 0.7971155850216747,
      "epoch": 0.877647766495281,
      "grad_norm": 10.8125,
      "learning_rate": 3.0403810238025e-05,
      "loss": 0.7680153846740723,
      "mean_token_accuracy": 0.8237178038805724,
      "num_tokens": 60502227.0,
      "step": 30640
    },
    {
      "entropy": 0.68349031470716,
      "epoch": 0.8779342050613694,
      "grad_norm": 8.6875,
      "learning_rate": 3.039266776180472e-05,
      "loss": 0.6794829368591309,
      "mean_token_accuracy": 0.8362726598978043,
      "num_tokens": 60522403.0,
      "step": 30650
    },
    {
      "entropy": 0.8197264751419425,
      "epoch": 0.878220643627458,
      "grad_norm": 9.1875,
      "learning_rate": 3.0381524161953773e-05,
      "loss": 0.8297025680541992,
      "mean_token_accuracy": 0.8081941585987806,
      "num_tokens": 60541387.0,
      "step": 30660
    },
    {
      "entropy": 0.7062901511788369,
      "epoch": 0.8785070821935466,
      "grad_norm": 8.3125,
      "learning_rate": 3.0370379440794078e-05,
      "loss": 0.6845837116241456,
      "mean_token_accuracy": 0.8361088272184134,
      "num_tokens": 60560624.0,
      "step": 30670
    },
    {
      "entropy": 0.7271107519976795,
      "epoch": 0.8787935207596351,
      "grad_norm": 6.0625,
      "learning_rate": 3.0359233600647762e-05,
      "loss": 0.6891570568084717,
      "mean_token_accuracy": 0.8380659099668264,
      "num_tokens": 60580010.0,
      "step": 30680
    },
    {
      "entropy": 0.8001046588644385,
      "epoch": 0.8790799593257236,
      "grad_norm": 9.25,
      "learning_rate": 3.0348086643837224e-05,
      "loss": 0.7931215286254882,
      "mean_token_accuracy": 0.8099168732762336,
      "num_tokens": 60600544.0,
      "step": 30690
    },
    {
      "entropy": 0.7773623916320502,
      "epoch": 0.8793663978918121,
      "grad_norm": 10.8125,
      "learning_rate": 3.033693857268506e-05,
      "loss": 0.8181222915649414,
      "mean_token_accuracy": 0.8166461087763309,
      "num_tokens": 60620623.0,
      "step": 30700
    },
    {
      "entropy": 0.7564819736406208,
      "epoch": 0.8796528364579007,
      "grad_norm": 8.4375,
      "learning_rate": 3.0325789389514115e-05,
      "loss": 0.7967268943786621,
      "mean_token_accuracy": 0.8130446758121253,
      "num_tokens": 60639611.0,
      "step": 30710
    },
    {
      "entropy": 0.7058803061023354,
      "epoch": 0.8799392750239893,
      "grad_norm": 8.1875,
      "learning_rate": 3.0314639096647458e-05,
      "loss": 0.7387288093566895,
      "mean_token_accuracy": 0.8288683012127877,
      "num_tokens": 60659047.0,
      "step": 30720
    },
    {
      "entropy": 0.6799276730045676,
      "epoch": 0.8802257135900777,
      "grad_norm": 5.4375,
      "learning_rate": 3.0303487696408402e-05,
      "loss": 0.6555741786956787,
      "mean_token_accuracy": 0.8393810819834471,
      "num_tokens": 60677200.0,
      "step": 30730
    },
    {
      "entropy": 0.7212142433971167,
      "epoch": 0.8805121521561663,
      "grad_norm": 5.46875,
      "learning_rate": 3.0292335191120476e-05,
      "loss": 0.7322101593017578,
      "mean_token_accuracy": 0.8280284982174635,
      "num_tokens": 60697507.0,
      "step": 30740
    },
    {
      "entropy": 0.6601237121969461,
      "epoch": 0.8807985907222549,
      "grad_norm": 7.5625,
      "learning_rate": 3.0281181583107454e-05,
      "loss": 0.7039535999298095,
      "mean_token_accuracy": 0.8313065115362406,
      "num_tokens": 60716982.0,
      "step": 30750
    },
    {
      "entropy": 0.6821398260071874,
      "epoch": 0.8810850292883434,
      "grad_norm": 7.03125,
      "learning_rate": 3.0270026874693314e-05,
      "loss": 0.6786174297332763,
      "mean_token_accuracy": 0.8420228850096464,
      "num_tokens": 60736675.0,
      "step": 30760
    },
    {
      "entropy": 0.6836420711129904,
      "epoch": 0.8813714678544319,
      "grad_norm": 6.90625,
      "learning_rate": 3.025887106820231e-05,
      "loss": 0.6533515453338623,
      "mean_token_accuracy": 0.8497208420187234,
      "num_tokens": 60756061.0,
      "step": 30770
    },
    {
      "entropy": 0.764008273743093,
      "epoch": 0.8816579064205204,
      "grad_norm": 7.125,
      "learning_rate": 3.0247714165958867e-05,
      "loss": 0.7639576911926269,
      "mean_token_accuracy": 0.8147179692983627,
      "num_tokens": 60774781.0,
      "step": 30780
    },
    {
      "entropy": 0.7028985355049372,
      "epoch": 0.881944344986609,
      "grad_norm": 6.5,
      "learning_rate": 3.0236556170287684e-05,
      "loss": 0.6570648670196533,
      "mean_token_accuracy": 0.8449258655309677,
      "num_tokens": 60795376.0,
      "step": 30790
    },
    {
      "entropy": 0.7239825608208775,
      "epoch": 0.8822307835526976,
      "grad_norm": 7.5,
      "learning_rate": 3.0225397083513663e-05,
      "loss": 0.7176084518432617,
      "mean_token_accuracy": 0.8264756493270398,
      "num_tokens": 60815773.0,
      "step": 30800
    },
    {
      "entropy": 0.7251630943268538,
      "epoch": 0.882517222118786,
      "grad_norm": 10.25,
      "learning_rate": 3.021423690796194e-05,
      "loss": 0.7179080963134765,
      "mean_token_accuracy": 0.8368282493203878,
      "num_tokens": 60836433.0,
      "step": 30810
    },
    {
      "entropy": 0.7919731775298715,
      "epoch": 0.8828036606848746,
      "grad_norm": 10.6875,
      "learning_rate": 3.020307564595789e-05,
      "loss": 0.8655158042907715,
      "mean_token_accuracy": 0.7968268454074859,
      "num_tokens": 60855765.0,
      "step": 30820
    },
    {
      "entropy": 0.7411508040502668,
      "epoch": 0.8830900992509632,
      "grad_norm": 11.0625,
      "learning_rate": 3.0191913299827092e-05,
      "loss": 0.7353606700897217,
      "mean_token_accuracy": 0.8321244582533837,
      "num_tokens": 60875545.0,
      "step": 30830
    },
    {
      "entropy": 0.6617496353108436,
      "epoch": 0.8833765378170517,
      "grad_norm": 8.25,
      "learning_rate": 3.0180749871895365e-05,
      "loss": 0.6615064144134521,
      "mean_token_accuracy": 0.8370318964123726,
      "num_tokens": 60895462.0,
      "step": 30840
    },
    {
      "entropy": 0.7592549323104322,
      "epoch": 0.8836629763831402,
      "grad_norm": 7.3125,
      "learning_rate": 3.016958536448875e-05,
      "loss": 0.8096492767333985,
      "mean_token_accuracy": 0.8189971137791872,
      "num_tokens": 60915042.0,
      "step": 30850
    },
    {
      "entropy": 0.6978416374884546,
      "epoch": 0.8839494149492287,
      "grad_norm": 8.1875,
      "learning_rate": 3.0158419779933522e-05,
      "loss": 0.7231961250305176,
      "mean_token_accuracy": 0.8337355773895979,
      "num_tokens": 60935163.0,
      "step": 30860
    },
    {
      "entropy": 0.9080134457908571,
      "epoch": 0.8842358535153173,
      "grad_norm": 9.5,
      "learning_rate": 3.0147253120556163e-05,
      "loss": 0.9699234008789063,
      "mean_token_accuracy": 0.7830836530774832,
      "num_tokens": 60954628.0,
      "step": 30870
    },
    {
      "entropy": 0.7694738387130201,
      "epoch": 0.8845222920814059,
      "grad_norm": 9.0,
      "learning_rate": 3.013608538868339e-05,
      "loss": 0.7606583595275879,
      "mean_token_accuracy": 0.8223088577389717,
      "num_tokens": 60975567.0,
      "step": 30880
    },
    {
      "entropy": 0.8216118371114135,
      "epoch": 0.8848087306474943,
      "grad_norm": 8.125,
      "learning_rate": 3.0124916586642138e-05,
      "loss": 0.7614274024963379,
      "mean_token_accuracy": 0.8249473243951797,
      "num_tokens": 60995219.0,
      "step": 30890
    },
    {
      "entropy": 0.7571396732702851,
      "epoch": 0.8850951692135829,
      "grad_norm": 7.5,
      "learning_rate": 3.0113746716759577e-05,
      "loss": 0.7938486576080322,
      "mean_token_accuracy": 0.8243913173675537,
      "num_tokens": 61012933.0,
      "step": 30900
    },
    {
      "entropy": 0.7443052959628403,
      "epoch": 0.8853816077796715,
      "grad_norm": 5.78125,
      "learning_rate": 3.0102575781363086e-05,
      "loss": 0.7210092067718505,
      "mean_token_accuracy": 0.8359295934438705,
      "num_tokens": 61031706.0,
      "step": 30910
    },
    {
      "entropy": 0.7093844223767519,
      "epoch": 0.88566804634576,
      "grad_norm": 9.4375,
      "learning_rate": 3.0091403782780277e-05,
      "loss": 0.7536305904388427,
      "mean_token_accuracy": 0.8251762878149748,
      "num_tokens": 61053386.0,
      "step": 30920
    },
    {
      "entropy": 0.689289870718494,
      "epoch": 0.8859544849118486,
      "grad_norm": 8.8125,
      "learning_rate": 3.008023072333896e-05,
      "loss": 0.6628387928009033,
      "mean_token_accuracy": 0.8332017607986927,
      "num_tokens": 61073416.0,
      "step": 30930
    },
    {
      "entropy": 0.7848109440878034,
      "epoch": 0.886240923477937,
      "grad_norm": 9.5,
      "learning_rate": 3.0069056605367196e-05,
      "loss": 0.7856605052947998,
      "mean_token_accuracy": 0.808146058768034,
      "num_tokens": 61092806.0,
      "step": 30940
    },
    {
      "entropy": 0.7681239143013954,
      "epoch": 0.8865273620440256,
      "grad_norm": 14.125,
      "learning_rate": 3.0057881431193262e-05,
      "loss": 0.7663568019866943,
      "mean_token_accuracy": 0.8171605970710516,
      "num_tokens": 61111766.0,
      "step": 30950
    },
    {
      "entropy": 0.7106232667341829,
      "epoch": 0.8868138006101142,
      "grad_norm": 8.1875,
      "learning_rate": 3.004670520314563e-05,
      "loss": 0.7089918613433838,
      "mean_token_accuracy": 0.827999022975564,
      "num_tokens": 61130136.0,
      "step": 30960
    },
    {
      "entropy": 0.7730832124128938,
      "epoch": 0.8871002391762027,
      "grad_norm": 8.25,
      "learning_rate": 3.003552792355302e-05,
      "loss": 0.7596082210540771,
      "mean_token_accuracy": 0.8352843791246414,
      "num_tokens": 61151084.0,
      "step": 30970
    },
    {
      "entropy": 0.8277015628293156,
      "epoch": 0.8873866777422912,
      "grad_norm": 9.4375,
      "learning_rate": 3.002434959474436e-05,
      "loss": 0.8891929626464844,
      "mean_token_accuracy": 0.7981308862566948,
      "num_tokens": 61171236.0,
      "step": 30980
    },
    {
      "entropy": 0.7458932900801301,
      "epoch": 0.8876731163083798,
      "grad_norm": 13.5,
      "learning_rate": 3.0013170219048785e-05,
      "loss": 0.7512352943420411,
      "mean_token_accuracy": 0.8195790208876133,
      "num_tokens": 61189893.0,
      "step": 30990
    },
    {
      "entropy": 0.6911372106522322,
      "epoch": 0.8879595548744683,
      "grad_norm": 13.8125,
      "learning_rate": 3.0001989798795676e-05,
      "loss": 0.67549147605896,
      "mean_token_accuracy": 0.8400860242545605,
      "num_tokens": 61208585.0,
      "step": 31000
    },
    {
      "entropy": 0.729536060243845,
      "epoch": 0.8882459934405569,
      "grad_norm": 10.75,
      "learning_rate": 2.9990808336314607e-05,
      "loss": 0.7327952861785889,
      "mean_token_accuracy": 0.825945646315813,
      "num_tokens": 61226191.0,
      "step": 31010
    },
    {
      "entropy": 0.7175138656049966,
      "epoch": 0.8885324320066453,
      "grad_norm": 10.6875,
      "learning_rate": 2.997962583393538e-05,
      "loss": 0.7524642944335938,
      "mean_token_accuracy": 0.8236678455024957,
      "num_tokens": 61244689.0,
      "step": 31020
    },
    {
      "entropy": 0.6970237294211984,
      "epoch": 0.8888188705727339,
      "grad_norm": 7.46875,
      "learning_rate": 2.9968442293988003e-05,
      "loss": 0.7233943462371826,
      "mean_token_accuracy": 0.8313529267907143,
      "num_tokens": 61264197.0,
      "step": 31030
    },
    {
      "entropy": 0.7873996950685977,
      "epoch": 0.8891053091388225,
      "grad_norm": 9.9375,
      "learning_rate": 2.9957257718802717e-05,
      "loss": 0.8161042213439942,
      "mean_token_accuracy": 0.8175555322319269,
      "num_tokens": 61283569.0,
      "step": 31040
    },
    {
      "entropy": 0.7179760131984949,
      "epoch": 0.889391747704911,
      "grad_norm": 9.8125,
      "learning_rate": 2.9946072110709965e-05,
      "loss": 0.7448568344116211,
      "mean_token_accuracy": 0.8231851443648338,
      "num_tokens": 61302190.0,
      "step": 31050
    },
    {
      "entropy": 0.7373472161591053,
      "epoch": 0.8896781862709995,
      "grad_norm": 10.0625,
      "learning_rate": 2.9934885472040412e-05,
      "loss": 0.697323226928711,
      "mean_token_accuracy": 0.8315782330930233,
      "num_tokens": 61322250.0,
      "step": 31060
    },
    {
      "entropy": 0.7402709219604731,
      "epoch": 0.889964624837088,
      "grad_norm": 10.125,
      "learning_rate": 2.992369780512494e-05,
      "loss": 0.7481690883636475,
      "mean_token_accuracy": 0.8352062918245793,
      "num_tokens": 61342434.0,
      "step": 31070
    },
    {
      "entropy": 0.6676525418646634,
      "epoch": 0.8902510634031766,
      "grad_norm": 7.09375,
      "learning_rate": 2.9912509112294634e-05,
      "loss": 0.7065566539764404,
      "mean_token_accuracy": 0.8333274267613888,
      "num_tokens": 61362507.0,
      "step": 31080
    },
    {
      "entropy": 0.9020956603810191,
      "epoch": 0.8905375019692652,
      "grad_norm": 11.375,
      "learning_rate": 2.99013193958808e-05,
      "loss": 0.9186498641967773,
      "mean_token_accuracy": 0.7928092129528522,
      "num_tokens": 61381121.0,
      "step": 31090
    },
    {
      "entropy": 0.7983300136402249,
      "epoch": 0.8908239405353536,
      "grad_norm": 9.5,
      "learning_rate": 2.9890128658214965e-05,
      "loss": 0.7900227546691895,
      "mean_token_accuracy": 0.8162098735570907,
      "num_tokens": 61401543.0,
      "step": 31100
    },
    {
      "entropy": 0.8040341572836042,
      "epoch": 0.8911103791014422,
      "grad_norm": 9.5625,
      "learning_rate": 2.9878936901628857e-05,
      "loss": 0.7442851066589355,
      "mean_token_accuracy": 0.8413120318204165,
      "num_tokens": 61420924.0,
      "step": 31110
    },
    {
      "entropy": 0.6577093188650907,
      "epoch": 0.8913968176675308,
      "grad_norm": 10.875,
      "learning_rate": 2.9867744128454422e-05,
      "loss": 0.681004524230957,
      "mean_token_accuracy": 0.8353380359709263,
      "num_tokens": 61442004.0,
      "step": 31120
    },
    {
      "entropy": 0.6750234639272094,
      "epoch": 0.8916832562336193,
      "grad_norm": 11.5,
      "learning_rate": 2.9856550341023804e-05,
      "loss": 0.725086784362793,
      "mean_token_accuracy": 0.8326871432363987,
      "num_tokens": 61462551.0,
      "step": 31130
    },
    {
      "entropy": 0.7027946091257036,
      "epoch": 0.8919696947997078,
      "grad_norm": 8.375,
      "learning_rate": 2.9845355541669388e-05,
      "loss": 0.7278618335723877,
      "mean_token_accuracy": 0.8312160987406969,
      "num_tokens": 61483342.0,
      "step": 31140
    },
    {
      "entropy": 0.7856769408099353,
      "epoch": 0.8922561333657963,
      "grad_norm": 7.625,
      "learning_rate": 2.9834159732723748e-05,
      "loss": 0.7381580352783204,
      "mean_token_accuracy": 0.8243792865425348,
      "num_tokens": 61502653.0,
      "step": 31150
    },
    {
      "entropy": 0.7422053655609488,
      "epoch": 0.8925425719318849,
      "grad_norm": 7.96875,
      "learning_rate": 2.982296291651966e-05,
      "loss": 0.686269474029541,
      "mean_token_accuracy": 0.84371406622231,
      "num_tokens": 61522767.0,
      "step": 31160
    },
    {
      "entropy": 0.6906449371948838,
      "epoch": 0.8928290104979735,
      "grad_norm": 9.625,
      "learning_rate": 2.9811765095390142e-05,
      "loss": 0.7338810443878174,
      "mean_token_accuracy": 0.8254051491618156,
      "num_tokens": 61544979.0,
      "step": 31170
    },
    {
      "entropy": 0.6567263992503285,
      "epoch": 0.893115449064062,
      "grad_norm": 8.0625,
      "learning_rate": 2.9800566271668383e-05,
      "loss": 0.6736366748809814,
      "mean_token_accuracy": 0.8441880621016026,
      "num_tokens": 61564492.0,
      "step": 31180
    },
    {
      "entropy": 0.7523484025150537,
      "epoch": 0.8934018876301505,
      "grad_norm": 7.6875,
      "learning_rate": 2.9789366447687812e-05,
      "loss": 0.7650484085083008,
      "mean_token_accuracy": 0.8245291475206613,
      "num_tokens": 61584968.0,
      "step": 31190
    },
    {
      "entropy": 0.7069922818802297,
      "epoch": 0.8936883261962391,
      "grad_norm": 10.125,
      "learning_rate": 2.9778165625782056e-05,
      "loss": 0.6990158557891846,
      "mean_token_accuracy": 0.838983615860343,
      "num_tokens": 61604066.0,
      "step": 31200
    },
    {
      "entropy": 0.7056529687717557,
      "epoch": 0.8939747647623276,
      "grad_norm": 9.5625,
      "learning_rate": 2.976696380828493e-05,
      "loss": 0.7294295787811279,
      "mean_token_accuracy": 0.8325434073805809,
      "num_tokens": 61622559.0,
      "step": 31210
    },
    {
      "entropy": 0.8315901113674045,
      "epoch": 0.8942612033284162,
      "grad_norm": 8.5,
      "learning_rate": 2.9755760997530503e-05,
      "loss": 0.8582427024841308,
      "mean_token_accuracy": 0.8072742480784655,
      "num_tokens": 61642900.0,
      "step": 31220
    },
    {
      "entropy": 0.8346749696880579,
      "epoch": 0.8945476418945046,
      "grad_norm": 7.5625,
      "learning_rate": 2.9744557195852997e-05,
      "loss": 0.8532216072082519,
      "mean_token_accuracy": 0.808398736268282,
      "num_tokens": 61662931.0,
      "step": 31230
    },
    {
      "entropy": 0.757306386064738,
      "epoch": 0.8948340804605932,
      "grad_norm": 8.375,
      "learning_rate": 2.9733352405586878e-05,
      "loss": 0.7357732772827148,
      "mean_token_accuracy": 0.8290942400693894,
      "num_tokens": 61683000.0,
      "step": 31240
    },
    {
      "entropy": 0.8105206049978733,
      "epoch": 0.8951205190266818,
      "grad_norm": 8.5625,
      "learning_rate": 2.9722146629066806e-05,
      "loss": 0.8476202964782715,
      "mean_token_accuracy": 0.8138906635344029,
      "num_tokens": 61703699.0,
      "step": 31250
    },
    {
      "entropy": 0.7138552552089095,
      "epoch": 0.8954069575927703,
      "grad_norm": 5.9375,
      "learning_rate": 2.9710939868627646e-05,
      "loss": 0.7024792194366455,
      "mean_token_accuracy": 0.8383521988987923,
      "num_tokens": 61724894.0,
      "step": 31260
    },
    {
      "entropy": 0.7535786624066532,
      "epoch": 0.8956933961588588,
      "grad_norm": 8.1875,
      "learning_rate": 2.9699732126604467e-05,
      "loss": 0.7880331993103027,
      "mean_token_accuracy": 0.818932880833745,
      "num_tokens": 61745724.0,
      "step": 31270
    },
    {
      "entropy": 0.7256142221391201,
      "epoch": 0.8959798347249474,
      "grad_norm": 9.5625,
      "learning_rate": 2.968852340533254e-05,
      "loss": 0.7544530868530274,
      "mean_token_accuracy": 0.8239728383719921,
      "num_tokens": 61764695.0,
      "step": 31280
    },
    {
      "entropy": 0.7014750675298274,
      "epoch": 0.8962662732910359,
      "grad_norm": 9.875,
      "learning_rate": 2.9677313707147354e-05,
      "loss": 0.7054186820983886,
      "mean_token_accuracy": 0.8325873002409935,
      "num_tokens": 61785240.0,
      "step": 31290
    },
    {
      "entropy": 0.7169062677770853,
      "epoch": 0.8965527118571245,
      "grad_norm": 8.375,
      "learning_rate": 2.9666103034384586e-05,
      "loss": 0.7243823051452637,
      "mean_token_accuracy": 0.8182522401213645,
      "num_tokens": 61804829.0,
      "step": 31300
    },
    {
      "entropy": 0.7188352025114,
      "epoch": 0.8968391504232129,
      "grad_norm": 9.0,
      "learning_rate": 2.9654891389380112e-05,
      "loss": 0.6789646148681641,
      "mean_token_accuracy": 0.8357134558260441,
      "num_tokens": 61824502.0,
      "step": 31310
    },
    {
      "entropy": 0.7230908527038992,
      "epoch": 0.8971255889893015,
      "grad_norm": 12.0625,
      "learning_rate": 2.9643678774470046e-05,
      "loss": 0.7360464096069336,
      "mean_token_accuracy": 0.8173789419233799,
      "num_tokens": 61843555.0,
      "step": 31320
    },
    {
      "entropy": 0.6575697573833168,
      "epoch": 0.8974120275553901,
      "grad_norm": 7.5,
      "learning_rate": 2.9632465191990655e-05,
      "loss": 0.6717941761016846,
      "mean_token_accuracy": 0.8392052993178367,
      "num_tokens": 61863086.0,
      "step": 31330
    },
    {
      "entropy": 0.7477656358852982,
      "epoch": 0.8976984661214786,
      "grad_norm": 9.8125,
      "learning_rate": 2.962125064427843e-05,
      "loss": 0.8158635139465332,
      "mean_token_accuracy": 0.816642464324832,
      "num_tokens": 61883025.0,
      "step": 31340
    },
    {
      "entropy": 0.7603571466170251,
      "epoch": 0.8979849046875671,
      "grad_norm": 7.5,
      "learning_rate": 2.961003513367009e-05,
      "loss": 0.795188570022583,
      "mean_token_accuracy": 0.8248804248869419,
      "num_tokens": 61903274.0,
      "step": 31350
    },
    {
      "entropy": 0.6878170750103891,
      "epoch": 0.8982713432536557,
      "grad_norm": 9.0625,
      "learning_rate": 2.9598818662502504e-05,
      "loss": 0.7193904399871827,
      "mean_token_accuracy": 0.826022668927908,
      "num_tokens": 61922472.0,
      "step": 31360
    },
    {
      "entropy": 0.7751140899956226,
      "epoch": 0.8985577818197442,
      "grad_norm": 9.25,
      "learning_rate": 2.958760123311277e-05,
      "loss": 0.7936097145080566,
      "mean_token_accuracy": 0.8174173459410667,
      "num_tokens": 61941493.0,
      "step": 31370
    },
    {
      "entropy": 0.8272498942911625,
      "epoch": 0.8988442203858328,
      "grad_norm": 15.9375,
      "learning_rate": 2.9576382847838185e-05,
      "loss": 0.8450757026672363,
      "mean_token_accuracy": 0.8060385465621949,
      "num_tokens": 61961703.0,
      "step": 31380
    },
    {
      "entropy": 0.7364816202782094,
      "epoch": 0.8991306589519212,
      "grad_norm": 10.125,
      "learning_rate": 2.956516350901625e-05,
      "loss": 0.7294492721557617,
      "mean_token_accuracy": 0.8288434680551291,
      "num_tokens": 61981998.0,
      "step": 31390
    },
    {
      "entropy": 0.7489566548727453,
      "epoch": 0.8994170975180098,
      "grad_norm": 7.96875,
      "learning_rate": 2.9553943218984644e-05,
      "loss": 0.6987955093383789,
      "mean_token_accuracy": 0.8359244290739298,
      "num_tokens": 62001904.0,
      "step": 31400
    },
    {
      "entropy": 0.7545577180571854,
      "epoch": 0.8997035360840984,
      "grad_norm": 9.5,
      "learning_rate": 2.954272198008125e-05,
      "loss": 0.7620506763458252,
      "mean_token_accuracy": 0.8204831626266241,
      "num_tokens": 62022370.0,
      "step": 31410
    },
    {
      "entropy": 0.6748882193118334,
      "epoch": 0.8999899746501869,
      "grad_norm": 11.3125,
      "learning_rate": 2.9531499794644184e-05,
      "loss": 0.7021085262298584,
      "mean_token_accuracy": 0.8351902060210705,
      "num_tokens": 62040834.0,
      "step": 31420
    },
    {
      "entropy": 0.6967566643841565,
      "epoch": 0.9002764132162754,
      "grad_norm": 7.59375,
      "learning_rate": 2.9520276665011703e-05,
      "loss": 0.6857142448425293,
      "mean_token_accuracy": 0.830024653673172,
      "num_tokens": 62059972.0,
      "step": 31430
    },
    {
      "entropy": 0.7148443370126187,
      "epoch": 0.900562851782364,
      "grad_norm": 14.0625,
      "learning_rate": 2.9509052593522295e-05,
      "loss": 0.8035076141357422,
      "mean_token_accuracy": 0.8185323245823384,
      "num_tokens": 62079715.0,
      "step": 31440
    },
    {
      "entropy": 0.7593570794910193,
      "epoch": 0.9008492903484525,
      "grad_norm": 10.6875,
      "learning_rate": 2.9497827582514652e-05,
      "loss": 0.8087222099304199,
      "mean_token_accuracy": 0.8143355693668127,
      "num_tokens": 62098451.0,
      "step": 31450
    },
    {
      "entropy": 0.8112756595015526,
      "epoch": 0.9011357289145411,
      "grad_norm": 11.1875,
      "learning_rate": 2.9486601634327625e-05,
      "loss": 0.7825881481170655,
      "mean_token_accuracy": 0.8222827546298503,
      "num_tokens": 62118079.0,
      "step": 31460
    },
    {
      "entropy": 0.7795058527030051,
      "epoch": 0.9014221674806296,
      "grad_norm": 15.8125,
      "learning_rate": 2.94753747513003e-05,
      "loss": 0.7744061470031738,
      "mean_token_accuracy": 0.8158979788422585,
      "num_tokens": 62137035.0,
      "step": 31470
    },
    {
      "entropy": 0.6742089619860053,
      "epoch": 0.9017086060467181,
      "grad_norm": 10.8125,
      "learning_rate": 2.946414693577193e-05,
      "loss": 0.6621050357818603,
      "mean_token_accuracy": 0.8385760270059108,
      "num_tokens": 62156036.0,
      "step": 31480
    },
    {
      "entropy": 0.703850641194731,
      "epoch": 0.9019950446128067,
      "grad_norm": 9.875,
      "learning_rate": 2.9452918190081972e-05,
      "loss": 0.7163612365722656,
      "mean_token_accuracy": 0.8330206520855427,
      "num_tokens": 62174362.0,
      "step": 31490
    },
    {
      "entropy": 0.8593509015627205,
      "epoch": 0.9022814831788952,
      "grad_norm": 8.8125,
      "learning_rate": 2.9441688516570094e-05,
      "loss": 0.801654052734375,
      "mean_token_accuracy": 0.8163538090884686,
      "num_tokens": 62194863.0,
      "step": 31500
    },
    {
      "entropy": 0.8585990556515753,
      "epoch": 0.9025679217449838,
      "grad_norm": 11.25,
      "learning_rate": 2.9430457917576122e-05,
      "loss": 0.8224228858947754,
      "mean_token_accuracy": 0.806947735697031,
      "num_tokens": 62215015.0,
      "step": 31510
    },
    {
      "entropy": 0.7357140269130469,
      "epoch": 0.9028543603110722,
      "grad_norm": 10.6875,
      "learning_rate": 2.9419226395440102e-05,
      "loss": 0.7166444301605225,
      "mean_token_accuracy": 0.827725674957037,
      "num_tokens": 62234649.0,
      "step": 31520
    },
    {
      "entropy": 0.6936469420790672,
      "epoch": 0.9031407988771608,
      "grad_norm": 7.90625,
      "learning_rate": 2.9407993952502255e-05,
      "loss": 0.6898537635803222,
      "mean_token_accuracy": 0.8411212660372257,
      "num_tokens": 62254305.0,
      "step": 31530
    },
    {
      "entropy": 0.7241666647605598,
      "epoch": 0.9034272374432494,
      "grad_norm": 7.03125,
      "learning_rate": 2.9396760591103016e-05,
      "loss": 0.7040422439575196,
      "mean_token_accuracy": 0.8353231672197581,
      "num_tokens": 62273565.0,
      "step": 31540
    },
    {
      "entropy": 0.7190627657808364,
      "epoch": 0.9037136760093379,
      "grad_norm": 11.125,
      "learning_rate": 2.9385526313582994e-05,
      "loss": 0.7573830127716065,
      "mean_token_accuracy": 0.8212734516710043,
      "num_tokens": 62293974.0,
      "step": 31550
    },
    {
      "entropy": 0.7020899516530335,
      "epoch": 0.9040001145754264,
      "grad_norm": 6.84375,
      "learning_rate": 2.9374291122282982e-05,
      "loss": 0.7332519054412842,
      "mean_token_accuracy": 0.8295746177434922,
      "num_tokens": 62313750.0,
      "step": 31560
    },
    {
      "entropy": 0.6634957105852664,
      "epoch": 0.904286553141515,
      "grad_norm": 10.5,
      "learning_rate": 2.9363055019543995e-05,
      "loss": 0.6631300449371338,
      "mean_token_accuracy": 0.8447445970028639,
      "num_tokens": 62333375.0,
      "step": 31570
    },
    {
      "entropy": 0.6728171294555068,
      "epoch": 0.9045729917076035,
      "grad_norm": 10.8125,
      "learning_rate": 2.93518180077072e-05,
      "loss": 0.6319359302520752,
      "mean_token_accuracy": 0.8445135407149792,
      "num_tokens": 62351824.0,
      "step": 31580
    },
    {
      "entropy": 0.7362899030558765,
      "epoch": 0.9048594302736921,
      "grad_norm": 13.0625,
      "learning_rate": 2.934058008911398e-05,
      "loss": 0.7283154964447022,
      "mean_token_accuracy": 0.8316422224044799,
      "num_tokens": 62372580.0,
      "step": 31590
    },
    {
      "entropy": 0.7187744650989771,
      "epoch": 0.9051458688397805,
      "grad_norm": 8.0,
      "learning_rate": 2.9329341266105896e-05,
      "loss": 0.7219864368438721,
      "mean_token_accuracy": 0.825536771863699,
      "num_tokens": 62392614.0,
      "step": 31600
    },
    {
      "entropy": 0.6740025511477142,
      "epoch": 0.9054323074058691,
      "grad_norm": 7.34375,
      "learning_rate": 2.931810154102469e-05,
      "loss": 0.6770771026611329,
      "mean_token_accuracy": 0.8338985428214073,
      "num_tokens": 62412714.0,
      "step": 31610
    },
    {
      "entropy": 0.6248881016857922,
      "epoch": 0.9057187459719577,
      "grad_norm": 11.3125,
      "learning_rate": 2.9306860916212313e-05,
      "loss": 0.6387589931488037,
      "mean_token_accuracy": 0.8355521783232689,
      "num_tokens": 62431619.0,
      "step": 31620
    },
    {
      "entropy": 0.7290095265954732,
      "epoch": 0.9060051845380462,
      "grad_norm": 10.8125,
      "learning_rate": 2.9295619394010894e-05,
      "loss": 0.731112289428711,
      "mean_token_accuracy": 0.8288287714123725,
      "num_tokens": 62452247.0,
      "step": 31630
    },
    {
      "entropy": 0.8544218683615327,
      "epoch": 0.9062916231041347,
      "grad_norm": 9.4375,
      "learning_rate": 2.928437697676273e-05,
      "loss": 0.9018502235412598,
      "mean_token_accuracy": 0.8107961840927601,
      "num_tokens": 62471562.0,
      "step": 31640
    },
    {
      "entropy": 0.7390581543557346,
      "epoch": 0.9065780616702233,
      "grad_norm": 9.9375,
      "learning_rate": 2.9273133666810336e-05,
      "loss": 0.7546993255615234,
      "mean_token_accuracy": 0.8403464127331972,
      "num_tokens": 62491570.0,
      "step": 31650
    },
    {
      "entropy": 0.7329729869961739,
      "epoch": 0.9068645002363118,
      "grad_norm": 8.4375,
      "learning_rate": 2.926188946649639e-05,
      "loss": 0.7525569438934326,
      "mean_token_accuracy": 0.8365613598376512,
      "num_tokens": 62510125.0,
      "step": 31660
    },
    {
      "entropy": 0.7492935438640416,
      "epoch": 0.9071509388024004,
      "grad_norm": 10.5625,
      "learning_rate": 2.9250644378163776e-05,
      "loss": 0.7845086097717285,
      "mean_token_accuracy": 0.8255662865936756,
      "num_tokens": 62528954.0,
      "step": 31670
    },
    {
      "entropy": 0.7449580421671271,
      "epoch": 0.9074373773684888,
      "grad_norm": 7.8125,
      "learning_rate": 2.923939840415553e-05,
      "loss": 0.7695770740509034,
      "mean_token_accuracy": 0.8159236744046211,
      "num_tokens": 62550239.0,
      "step": 31680
    },
    {
      "entropy": 0.7755580667406321,
      "epoch": 0.9077238159345774,
      "grad_norm": 10.5,
      "learning_rate": 2.9228151546814914e-05,
      "loss": 0.767540168762207,
      "mean_token_accuracy": 0.820448062941432,
      "num_tokens": 62570696.0,
      "step": 31690
    },
    {
      "entropy": 0.6946647071279586,
      "epoch": 0.908010254500666,
      "grad_norm": 7.6875,
      "learning_rate": 2.9216903808485345e-05,
      "loss": 0.6396018505096436,
      "mean_token_accuracy": 0.8424449384212493,
      "num_tokens": 62590084.0,
      "step": 31700
    },
    {
      "entropy": 0.7388099811971187,
      "epoch": 0.9082966930667545,
      "grad_norm": 8.375,
      "learning_rate": 2.9205655191510423e-05,
      "loss": 0.7431666851043701,
      "mean_token_accuracy": 0.8243351325392723,
      "num_tokens": 62609231.0,
      "step": 31710
    },
    {
      "entropy": 0.8131321831606328,
      "epoch": 0.9085831316328431,
      "grad_norm": 10.875,
      "learning_rate": 2.9194405698233958e-05,
      "loss": 0.8745104789733886,
      "mean_token_accuracy": 0.801719157025218,
      "num_tokens": 62629374.0,
      "step": 31720
    },
    {
      "entropy": 0.7507283895276486,
      "epoch": 0.9088695701989316,
      "grad_norm": 7.15625,
      "learning_rate": 2.9183155330999905e-05,
      "loss": 0.6649708271026611,
      "mean_token_accuracy": 0.8437069479376078,
      "num_tokens": 62648655.0,
      "step": 31730
    },
    {
      "entropy": 0.7885823886841535,
      "epoch": 0.9091560087650201,
      "grad_norm": 11.0625,
      "learning_rate": 2.917190409215243e-05,
      "loss": 0.7782921314239502,
      "mean_token_accuracy": 0.8211049899458885,
      "num_tokens": 62668855.0,
      "step": 31740
    },
    {
      "entropy": 0.7401616541668773,
      "epoch": 0.9094424473311087,
      "grad_norm": 7.375,
      "learning_rate": 2.916065198403588e-05,
      "loss": 0.7428979873657227,
      "mean_token_accuracy": 0.8281892757862807,
      "num_tokens": 62688241.0,
      "step": 31750
    },
    {
      "entropy": 0.7733067704364658,
      "epoch": 0.9097288858971972,
      "grad_norm": 7.625,
      "learning_rate": 2.914939900899476e-05,
      "loss": 0.8419400215148926,
      "mean_token_accuracy": 0.8085472885519266,
      "num_tokens": 62708061.0,
      "step": 31760
    },
    {
      "entropy": 0.6649927786551416,
      "epoch": 0.9100153244632857,
      "grad_norm": 7.96875,
      "learning_rate": 2.9138145169373777e-05,
      "loss": 0.6369128227233887,
      "mean_token_accuracy": 0.8477082684636116,
      "num_tokens": 62729123.0,
      "step": 31770
    },
    {
      "entropy": 0.7138575472868979,
      "epoch": 0.9103017630293743,
      "grad_norm": 9.8125,
      "learning_rate": 2.9126890467517802e-05,
      "loss": 0.721143388748169,
      "mean_token_accuracy": 0.8366635706275701,
      "num_tokens": 62748957.0,
      "step": 31780
    },
    {
      "entropy": 0.7686808630824089,
      "epoch": 0.9105882015954628,
      "grad_norm": 9.1875,
      "learning_rate": 2.911563490577191e-05,
      "loss": 0.7773112773895263,
      "mean_token_accuracy": 0.819014685228467,
      "num_tokens": 62769270.0,
      "step": 31790
    },
    {
      "entropy": 0.7524145107716322,
      "epoch": 0.9108746401615514,
      "grad_norm": 8.0625,
      "learning_rate": 2.910437848648133e-05,
      "loss": 0.710072660446167,
      "mean_token_accuracy": 0.8368499759584666,
      "num_tokens": 62789066.0,
      "step": 31800
    },
    {
      "entropy": 0.7506163151934743,
      "epoch": 0.9111610787276399,
      "grad_norm": 10.6875,
      "learning_rate": 2.909312121199148e-05,
      "loss": 0.7587294101715087,
      "mean_token_accuracy": 0.8208259314298629,
      "num_tokens": 62811309.0,
      "step": 31810
    },
    {
      "entropy": 0.7362167067825794,
      "epoch": 0.9114475172937284,
      "grad_norm": 8.6875,
      "learning_rate": 2.9081863084647965e-05,
      "loss": 0.7316089630126953,
      "mean_token_accuracy": 0.8291648272424936,
      "num_tokens": 62832870.0,
      "step": 31820
    },
    {
      "entropy": 0.728352018352598,
      "epoch": 0.911733955859817,
      "grad_norm": 6.9375,
      "learning_rate": 2.9070604106796547e-05,
      "loss": 0.7544605731964111,
      "mean_token_accuracy": 0.8219658058136702,
      "num_tokens": 62852188.0,
      "step": 31830
    },
    {
      "entropy": 0.8075040376745164,
      "epoch": 0.9120203944259055,
      "grad_norm": 10.8125,
      "learning_rate": 2.9059344280783174e-05,
      "loss": 0.8330258369445801,
      "mean_token_accuracy": 0.8064204879105091,
      "num_tokens": 62872733.0,
      "step": 31840
    },
    {
      "entropy": 0.7481285523623228,
      "epoch": 0.912306832991994,
      "grad_norm": 7.78125,
      "learning_rate": 2.9048083608953992e-05,
      "loss": 0.7964035511016846,
      "mean_token_accuracy": 0.8239479623734951,
      "num_tokens": 62892601.0,
      "step": 31850
    },
    {
      "entropy": 0.6737873202189804,
      "epoch": 0.9125932715580826,
      "grad_norm": 7.53125,
      "learning_rate": 2.903682209365528e-05,
      "loss": 0.6516993999481201,
      "mean_token_accuracy": 0.8448345594108104,
      "num_tokens": 62912315.0,
      "step": 31860
    },
    {
      "entropy": 0.7904157625511289,
      "epoch": 0.9128797101241711,
      "grad_norm": 7.8125,
      "learning_rate": 2.9025559737233537e-05,
      "loss": 0.8032742500305176,
      "mean_token_accuracy": 0.813806938752532,
      "num_tokens": 62930955.0,
      "step": 31870
    },
    {
      "entropy": 0.8429348040372133,
      "epoch": 0.9131661486902597,
      "grad_norm": 9.5625,
      "learning_rate": 2.9014296542035414e-05,
      "loss": 0.8911533355712891,
      "mean_token_accuracy": 0.8017095036804676,
      "num_tokens": 62952002.0,
      "step": 31880
    },
    {
      "entropy": 0.7410570094361901,
      "epoch": 0.9134525872563481,
      "grad_norm": 9.4375,
      "learning_rate": 2.900303251040773e-05,
      "loss": 0.7683367252349853,
      "mean_token_accuracy": 0.8241291139274836,
      "num_tokens": 62971662.0,
      "step": 31890
    },
    {
      "entropy": 0.7138321973383427,
      "epoch": 0.9137390258224367,
      "grad_norm": 7.0,
      "learning_rate": 2.8991767644697497e-05,
      "loss": 0.7363371849060059,
      "mean_token_accuracy": 0.8282822303473949,
      "num_tokens": 62991283.0,
      "step": 31900
    },
    {
      "entropy": 0.8175505900755524,
      "epoch": 0.9140254643885253,
      "grad_norm": 7.78125,
      "learning_rate": 2.8980501947251898e-05,
      "loss": 0.8047837257385254,
      "mean_token_accuracy": 0.8185700699687004,
      "num_tokens": 63012485.0,
      "step": 31910
    },
    {
      "entropy": 0.7987545650452376,
      "epoch": 0.9143119029546138,
      "grad_norm": 9.625,
      "learning_rate": 2.8969235420418274e-05,
      "loss": 0.7709282875061035,
      "mean_token_accuracy": 0.8309549584984779,
      "num_tokens": 63032969.0,
      "step": 31920
    },
    {
      "entropy": 0.741315673943609,
      "epoch": 0.9145983415207023,
      "grad_norm": 10.1875,
      "learning_rate": 2.895796806654415e-05,
      "loss": 0.754824161529541,
      "mean_token_accuracy": 0.8188746023923159,
      "num_tokens": 63052590.0,
      "step": 31930
    },
    {
      "entropy": 0.7676701046526432,
      "epoch": 0.9148847800867909,
      "grad_norm": 8.3125,
      "learning_rate": 2.894669988797723e-05,
      "loss": 0.7613906383514404,
      "mean_token_accuracy": 0.830718194320798,
      "num_tokens": 63072058.0,
      "step": 31940
    },
    {
      "entropy": 0.6324413171969354,
      "epoch": 0.9151712186528794,
      "grad_norm": 9.0625,
      "learning_rate": 2.8935430887065374e-05,
      "loss": 0.6272721767425538,
      "mean_token_accuracy": 0.8514070764183999,
      "num_tokens": 63092101.0,
      "step": 31950
    },
    {
      "entropy": 0.7214945896528662,
      "epoch": 0.915457657218968,
      "grad_norm": 5.875,
      "learning_rate": 2.8924161066156614e-05,
      "loss": 0.6717380046844482,
      "mean_token_accuracy": 0.8345632407814264,
      "num_tokens": 63113831.0,
      "step": 31960
    },
    {
      "entropy": 0.6731991032138467,
      "epoch": 0.9157440957850566,
      "grad_norm": 10.625,
      "learning_rate": 2.8912890427599177e-05,
      "loss": 0.6964722156524659,
      "mean_token_accuracy": 0.830223273485899,
      "num_tokens": 63133580.0,
      "step": 31970
    },
    {
      "entropy": 0.6781891839578748,
      "epoch": 0.916030534351145,
      "grad_norm": 8.5,
      "learning_rate": 2.8901618973741433e-05,
      "loss": 0.6699615478515625,
      "mean_token_accuracy": 0.8291253335773945,
      "num_tokens": 63153098.0,
      "step": 31980
    },
    {
      "entropy": 0.7845070753246546,
      "epoch": 0.9163169729172336,
      "grad_norm": 9.0,
      "learning_rate": 2.8890346706931935e-05,
      "loss": 0.8258766174316406,
      "mean_token_accuracy": 0.8216951984912158,
      "num_tokens": 63172249.0,
      "step": 31990
    },
    {
      "epoch": 0.9166034114833221,
      "eval_entropy": 0.772518795311451,
      "eval_loss": 0.7784592509269714,
      "eval_mean_token_accuracy": 0.8175579745769501,
      "eval_num_tokens": 63190998.0,
      "eval_runtime": 43.0079,
      "eval_samples_per_second": 46.503,
      "eval_steps_per_second": 5.813,
      "step": 32000
    }
  ],
  "logging_steps": 10,
  "max_steps": 69824,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 4000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.10716081203948e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}