{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.898403483309144,
  "eval_steps": 3000,
  "global_step": 54000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.902761459350586,
      "epoch": 0.00045355587808418,
      "grad_norm": 15.0625,
      "learning_rate": 2e-06,
      "loss": 14.2424,
      "mean_token_accuracy": 6.570302066393197e-05,
      "num_tokens": 10827.0,
      "step": 5
    },
    {
      "entropy": 4.906017255783081,
      "epoch": 0.00090711175616836,
      "grad_norm": 15.0625,
      "learning_rate": 4.5e-06,
      "loss": 14.1679,
      "mean_token_accuracy": 0.00036321740481071174,
      "num_tokens": 21025.0,
      "step": 10
    },
    {
      "entropy": 4.919331836700439,
      "epoch": 0.00136066763425254,
      "grad_norm": 16.25,
      "learning_rate": 7e-06,
      "loss": 14.1379,
      "mean_token_accuracy": 0.00018726952257566153,
      "num_tokens": 31678.0,
      "step": 15
    },
    {
      "entropy": 4.969941806793213,
      "epoch": 0.00181422351233672,
      "grad_norm": 17.75,
      "learning_rate": 9.5e-06,
      "loss": 13.9183,
      "mean_token_accuracy": 0.0,
      "num_tokens": 41025.0,
      "step": 20
    },
    {
      "entropy": 5.119146680831909,
      "epoch": 0.0022677793904208998,
      "grad_norm": 21.875,
      "learning_rate": 1.2e-05,
      "loss": 13.5438,
      "mean_token_accuracy": 0.0,
      "num_tokens": 51270.0,
      "step": 25
    },
    {
      "entropy": 5.423425340652466,
      "epoch": 0.00272133526850508,
      "grad_norm": 26.625,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.082,
      "mean_token_accuracy": 0.0,
      "num_tokens": 62142.0,
      "step": 30
    },
    {
      "entropy": 6.89132776260376,
      "epoch": 0.00317489114658926,
      "grad_norm": 25.625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.989,
      "mean_token_accuracy": 0.0,
      "num_tokens": 72709.0,
      "step": 35
    },
    {
      "entropy": 10.029912281036378,
      "epoch": 0.00362844702467344,
      "grad_norm": 3.671875,
      "learning_rate": 1.95e-05,
      "loss": 10.9002,
      "mean_token_accuracy": 0.0,
      "num_tokens": 82313.0,
      "step": 40
    },
    {
      "entropy": 10.710572910308837,
      "epoch": 0.00408200290275762,
      "grad_norm": 3.59375,
      "learning_rate": 2.2e-05,
      "loss": 10.6789,
      "mean_token_accuracy": 0.001249837997602299,
      "num_tokens": 92450.0,
      "step": 45
    },
    {
      "entropy": 10.739872074127197,
      "epoch": 0.0045355587808417995,
      "grad_norm": 3.03125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.5409,
      "mean_token_accuracy": 0.010005939414259046,
      "num_tokens": 103978.0,
      "step": 50
    },
    {
      "entropy": 10.732357597351074,
      "epoch": 0.00498911465892598,
      "grad_norm": 2.140625,
      "learning_rate": 2.7e-05,
      "loss": 10.2901,
      "mean_token_accuracy": 0.0165672049857676,
      "num_tokens": 114134.0,
      "step": 55
    },
    {
      "entropy": 10.688269710540771,
      "epoch": 0.00544267053701016,
      "grad_norm": 2.296875,
      "learning_rate": 2.95e-05,
      "loss": 10.1413,
      "mean_token_accuracy": 0.015207256283611059,
      "num_tokens": 125037.0,
      "step": 60
    },
    {
      "entropy": 10.711676120758057,
      "epoch": 0.00589622641509434,
      "grad_norm": 1.90625,
      "learning_rate": 3.2e-05,
      "loss": 9.9497,
      "mean_token_accuracy": 0.017865559458732604,
      "num_tokens": 134721.0,
      "step": 65
    },
    {
      "entropy": 10.7262864112854,
      "epoch": 0.00634978229317852,
      "grad_norm": 1.9765625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8497,
      "mean_token_accuracy": 0.03912395965307951,
      "num_tokens": 144448.0,
      "step": 70
    },
    {
      "entropy": 10.72291088104248,
      "epoch": 0.006803338171262699,
      "grad_norm": 1.9609375,
      "learning_rate": 3.7e-05,
      "loss": 9.7736,
      "mean_token_accuracy": 0.0496896930038929,
      "num_tokens": 154343.0,
      "step": 75
    },
    {
      "entropy": 10.716559219360352,
      "epoch": 0.00725689404934688,
      "grad_norm": 1.953125,
      "learning_rate": 3.95e-05,
      "loss": 9.7443,
      "mean_token_accuracy": 0.046880177780985834,
      "num_tokens": 165650.0,
      "step": 80
    },
    {
      "entropy": 10.707624149322509,
      "epoch": 0.007710449927431059,
      "grad_norm": 1.7734375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.711,
      "mean_token_accuracy": 0.04450351819396019,
      "num_tokens": 177098.0,
      "step": 85
    },
    {
      "entropy": 10.704974555969239,
      "epoch": 0.00816400580551524,
      "grad_norm": 1.96875,
      "learning_rate": 4.45e-05,
      "loss": 9.4919,
      "mean_token_accuracy": 0.05480348989367485,
      "num_tokens": 187903.0,
      "step": 90
    },
    {
      "entropy": 10.689633083343505,
      "epoch": 0.00861756168359942,
      "grad_norm": 1.8671875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.4033,
      "mean_token_accuracy": 0.052090702950954436,
      "num_tokens": 198077.0,
      "step": 95
    },
    {
      "entropy": 10.674059963226318,
      "epoch": 0.009071117561683599,
      "grad_norm": 1.8046875,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.3565,
      "mean_token_accuracy": 0.05150817483663559,
      "num_tokens": 208641.0,
      "step": 100
    },
    {
      "entropy": 10.649502754211426,
      "epoch": 0.009524673439767779,
      "grad_norm": 1.84375,
      "learning_rate": 5.2e-05,
      "loss": 9.1936,
      "mean_token_accuracy": 0.05499474816024304,
      "num_tokens": 218398.0,
      "step": 105
    },
    {
      "entropy": 10.621120071411132,
      "epoch": 0.00997822931785196,
      "grad_norm": 1.6796875,
      "learning_rate": 5.45e-05,
      "loss": 9.066,
      "mean_token_accuracy": 0.05329935699701309,
      "num_tokens": 229059.0,
      "step": 110
    },
    {
      "entropy": 10.593021678924561,
      "epoch": 0.01043178519593614,
      "grad_norm": 1.7421875,
      "learning_rate": 5.7e-05,
      "loss": 8.9831,
      "mean_token_accuracy": 0.05141277946531773,
      "num_tokens": 239775.0,
      "step": 115
    },
    {
      "entropy": 10.551456928253174,
      "epoch": 0.01088534107402032,
      "grad_norm": 1.8515625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.842,
      "mean_token_accuracy": 0.05272237621247768,
      "num_tokens": 249954.0,
      "step": 120
    },
    {
      "entropy": 10.502390956878662,
      "epoch": 0.011338896952104499,
      "grad_norm": 1.65625,
      "learning_rate": 6.2e-05,
      "loss": 8.7334,
      "mean_token_accuracy": 0.05634862557053566,
      "num_tokens": 260056.0,
      "step": 125
    },
    {
      "entropy": 10.426200675964356,
      "epoch": 0.01179245283018868,
      "grad_norm": 1.5078125,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.6347,
      "mean_token_accuracy": 0.05641283132135868,
      "num_tokens": 269904.0,
      "step": 130
    },
    {
      "entropy": 10.35210132598877,
      "epoch": 0.01224600870827286,
      "grad_norm": 1.65625,
      "learning_rate": 6.7e-05,
      "loss": 8.5138,
      "mean_token_accuracy": 0.05234182849526405,
      "num_tokens": 280900.0,
      "step": 135
    },
    {
      "entropy": 10.270804977416992,
      "epoch": 0.01269956458635704,
      "grad_norm": 1.5546875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.406,
      "mean_token_accuracy": 0.05256233848631382,
      "num_tokens": 290844.0,
      "step": 140
    },
    {
      "entropy": 10.143964767456055,
      "epoch": 0.013153120464441219,
      "grad_norm": 1.2578125,
      "learning_rate": 7.2e-05,
      "loss": 8.2834,
      "mean_token_accuracy": 0.056692027300596234,
      "num_tokens": 301345.0,
      "step": 145
    },
    {
      "entropy": 10.030607223510742,
      "epoch": 0.013606676342525399,
      "grad_norm": 1.40625,
      "learning_rate": 7.45e-05,
      "loss": 8.1592,
      "mean_token_accuracy": 0.05243662744760513,
      "num_tokens": 312681.0,
      "step": 150
    },
    {
      "entropy": 9.876211452484132,
      "epoch": 0.01406023222060958,
      "grad_norm": 1.203125,
      "learning_rate": 7.7e-05,
      "loss": 8.0357,
      "mean_token_accuracy": 0.05974825695157051,
      "num_tokens": 323492.0,
      "step": 155
    },
    {
      "entropy": 9.703791522979737,
      "epoch": 0.01451378809869376,
      "grad_norm": 1.1015625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.9778,
      "mean_token_accuracy": 0.05336504280567169,
      "num_tokens": 334521.0,
      "step": 160
    },
    {
      "entropy": 9.493086814880371,
      "epoch": 0.014967343976777939,
      "grad_norm": 0.99609375,
      "learning_rate": 8.2e-05,
      "loss": 7.9038,
      "mean_token_accuracy": 0.05283032245934009,
      "num_tokens": 345468.0,
      "step": 165
    },
    {
      "entropy": 9.287709426879882,
      "epoch": 0.015420899854862119,
      "grad_norm": 1.0625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.7788,
      "mean_token_accuracy": 0.055317680910229684,
      "num_tokens": 355911.0,
      "step": 170
    },
    {
      "entropy": 9.030819416046143,
      "epoch": 0.0158744557329463,
      "grad_norm": 0.87890625,
      "learning_rate": 8.7e-05,
      "loss": 7.7522,
      "mean_token_accuracy": 0.05364949442446232,
      "num_tokens": 367537.0,
      "step": 175
    },
    {
      "entropy": 8.806273078918457,
      "epoch": 0.01632801161103048,
      "grad_norm": 0.890625,
      "learning_rate": 8.95e-05,
      "loss": 7.8067,
      "mean_token_accuracy": 0.05446445494890213,
      "num_tokens": 377918.0,
      "step": 180
    },
    {
      "entropy": 8.59828290939331,
      "epoch": 0.01678156748911466,
      "grad_norm": 0.703125,
      "learning_rate": 9.2e-05,
      "loss": 7.6514,
      "mean_token_accuracy": 0.055352770164608954,
      "num_tokens": 388461.0,
      "step": 185
    },
    {
      "entropy": 8.414749145507812,
      "epoch": 0.01723512336719884,
      "grad_norm": 0.9140625,
      "learning_rate": 9.45e-05,
      "loss": 7.6092,
      "mean_token_accuracy": 0.06770730689167977,
      "num_tokens": 399771.0,
      "step": 190
    },
    {
      "entropy": 8.264602565765381,
      "epoch": 0.01768867924528302,
      "grad_norm": 0.6953125,
      "learning_rate": 9.7e-05,
      "loss": 7.4766,
      "mean_token_accuracy": 0.07282306849956513,
      "num_tokens": 410254.0,
      "step": 195
    },
    {
      "entropy": 8.190978908538819,
      "epoch": 0.018142235123367198,
      "grad_norm": 1.25,
      "learning_rate": 9.95e-05,
      "loss": 7.4886,
      "mean_token_accuracy": 0.06817287020385265,
      "num_tokens": 419792.0,
      "step": 200
    },
    {
      "entropy": 8.113365936279298,
      "epoch": 0.018595791001451378,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000102,
      "loss": 7.4782,
      "mean_token_accuracy": 0.07438576593995094,
      "num_tokens": 429264.0,
      "step": 205
    },
    {
      "entropy": 8.046414375305176,
      "epoch": 0.019049346879535557,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.4294,
      "mean_token_accuracy": 0.0784988671541214,
      "num_tokens": 439624.0,
      "step": 210
    },
    {
      "entropy": 8.01142177581787,
      "epoch": 0.01950290275761974,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000107,
      "loss": 7.4206,
      "mean_token_accuracy": 0.07665222510695457,
      "num_tokens": 449556.0,
      "step": 215
    },
    {
      "entropy": 7.982991123199463,
      "epoch": 0.01995645863570392,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001095,
      "loss": 7.4192,
      "mean_token_accuracy": 0.07819079160690308,
      "num_tokens": 460243.0,
      "step": 220
    },
    {
      "entropy": 7.913471603393555,
      "epoch": 0.0204100145137881,
      "grad_norm": 1.015625,
      "learning_rate": 0.000112,
      "loss": 7.4682,
      "mean_token_accuracy": 0.07565705217421055,
      "num_tokens": 469767.0,
      "step": 225
    },
    {
      "entropy": 7.974834537506103,
      "epoch": 0.02086357039187228,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0001145,
      "loss": 7.3948,
      "mean_token_accuracy": 0.07845633998513221,
      "num_tokens": 480329.0,
      "step": 230
    },
    {
      "entropy": 7.921621561050415,
      "epoch": 0.02131712626995646,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.3597,
      "mean_token_accuracy": 0.07565821185708047,
      "num_tokens": 490096.0,
      "step": 235
    },
    {
      "entropy": 7.9462236881256105,
      "epoch": 0.02177068214804064,
      "grad_norm": 1.328125,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.404,
      "mean_token_accuracy": 0.07110006958246232,
      "num_tokens": 501231.0,
      "step": 240
    },
    {
      "entropy": 7.90762996673584,
      "epoch": 0.022224238026124818,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000122,
      "loss": 7.3003,
      "mean_token_accuracy": 0.07855217903852463,
      "num_tokens": 510722.0,
      "step": 245
    },
    {
      "entropy": 7.791062259674073,
      "epoch": 0.022677793904208998,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0001245,
      "loss": 7.342,
      "mean_token_accuracy": 0.0785796582698822,
      "num_tokens": 521452.0,
      "step": 250
    },
    {
      "entropy": 7.835426616668701,
      "epoch": 0.023131349782293177,
      "grad_norm": 1.0,
      "learning_rate": 0.000127,
      "loss": 7.3275,
      "mean_token_accuracy": 0.07848411723971367,
      "num_tokens": 532397.0,
      "step": 255
    },
    {
      "entropy": 7.834844970703125,
      "epoch": 0.02358490566037736,
      "grad_norm": 0.9375,
      "learning_rate": 0.0001295,
      "loss": 7.3126,
      "mean_token_accuracy": 0.08216355144977569,
      "num_tokens": 542271.0,
      "step": 260
    },
    {
      "entropy": 7.783105802536011,
      "epoch": 0.02403846153846154,
      "grad_norm": 1.046875,
      "learning_rate": 0.000132,
      "loss": 7.2852,
      "mean_token_accuracy": 0.08326507434248924,
      "num_tokens": 552411.0,
      "step": 265
    },
    {
      "entropy": 7.789214849472046,
      "epoch": 0.02449201741654572,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.2853,
      "mean_token_accuracy": 0.07970908731222152,
      "num_tokens": 561518.0,
      "step": 270
    },
    {
      "entropy": 7.8078306198120115,
      "epoch": 0.0249455732946299,
      "grad_norm": 0.859375,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.3048,
      "mean_token_accuracy": 0.08076792433857918,
      "num_tokens": 571361.0,
      "step": 275
    },
    {
      "entropy": 7.704005384445191,
      "epoch": 0.02539912917271408,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0001395,
      "loss": 7.2079,
      "mean_token_accuracy": 0.08890472128987312,
      "num_tokens": 580926.0,
      "step": 280
    },
    {
      "entropy": 7.777441549301147,
      "epoch": 0.025852685050798258,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.305,
      "mean_token_accuracy": 0.07825151681900025,
      "num_tokens": 592282.0,
      "step": 285
    },
    {
      "entropy": 7.756225681304931,
      "epoch": 0.026306240928882438,
      "grad_norm": 1.25,
      "learning_rate": 0.0001445,
      "loss": 7.3123,
      "mean_token_accuracy": 0.08080485239624977,
      "num_tokens": 602767.0,
      "step": 290
    },
    {
      "entropy": 7.716958808898926,
      "epoch": 0.026759796806966617,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000147,
      "loss": 7.2177,
      "mean_token_accuracy": 0.08552644699811936,
      "num_tokens": 612763.0,
      "step": 295
    },
    {
      "entropy": 7.706277990341187,
      "epoch": 0.027213352685050797,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0001495,
      "loss": 7.1737,
      "mean_token_accuracy": 0.08685687705874442,
      "num_tokens": 622796.0,
      "step": 300
    },
    {
      "entropy": 7.668240880966186,
      "epoch": 0.027666908563134977,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000152,
      "loss": 7.1609,
      "mean_token_accuracy": 0.08536986410617828,
      "num_tokens": 633136.0,
      "step": 305
    },
    {
      "entropy": 7.649664688110351,
      "epoch": 0.02812046444121916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.143,
      "mean_token_accuracy": 0.08583157956600189,
      "num_tokens": 643347.0,
      "step": 310
    },
    {
      "entropy": 7.697743082046509,
      "epoch": 0.02857402031930334,
      "grad_norm": 0.859375,
      "learning_rate": 0.000157,
      "loss": 7.1414,
      "mean_token_accuracy": 0.09304612576961517,
      "num_tokens": 653096.0,
      "step": 315
    },
    {
      "entropy": 7.587596750259399,
      "epoch": 0.02902757619738752,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0001595,
      "loss": 7.1706,
      "mean_token_accuracy": 0.08879687860608101,
      "num_tokens": 663557.0,
      "step": 320
    },
    {
      "entropy": 7.612160730361938,
      "epoch": 0.0294811320754717,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000162,
      "loss": 7.2254,
      "mean_token_accuracy": 0.08831041827797889,
      "num_tokens": 674395.0,
      "step": 325
    },
    {
      "entropy": 7.695938348770142,
      "epoch": 0.029934687953555878,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.1265,
      "mean_token_accuracy": 0.09078574851155281,
      "num_tokens": 684776.0,
      "step": 330
    },
    {
      "entropy": 7.576247215270996,
      "epoch": 0.030388243831640058,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.109,
      "mean_token_accuracy": 0.08843011409044266,
      "num_tokens": 694886.0,
      "step": 335
    },
    {
      "entropy": 7.556908130645752,
      "epoch": 0.030841799709724237,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.171,
      "mean_token_accuracy": 0.09018958061933517,
      "num_tokens": 705426.0,
      "step": 340
    },
    {
      "entropy": 7.578840160369873,
      "epoch": 0.03129535558780842,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.1594,
      "mean_token_accuracy": 0.08782409057021141,
      "num_tokens": 716271.0,
      "step": 345
    },
    {
      "entropy": 7.638495922088623,
      "epoch": 0.0317489114658926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.1606,
      "mean_token_accuracy": 0.09265835508704186,
      "num_tokens": 726875.0,
      "step": 350
    },
    {
      "entropy": 7.563690900802612,
      "epoch": 0.032202467343976776,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000177,
      "loss": 7.0773,
      "mean_token_accuracy": 0.08827165514230728,
      "num_tokens": 737825.0,
      "step": 355
    },
    {
      "entropy": 7.58997654914856,
      "epoch": 0.03265602322206096,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0001795,
      "loss": 7.089,
      "mean_token_accuracy": 0.0890809178352356,
      "num_tokens": 747424.0,
      "step": 360
    },
    {
      "entropy": 7.506178569793701,
      "epoch": 0.033109579100145135,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000182,
      "loss": 7.0901,
      "mean_token_accuracy": 0.09158140644431115,
      "num_tokens": 757649.0,
      "step": 365
    },
    {
      "entropy": 7.567615652084351,
      "epoch": 0.03356313497822932,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001845,
      "loss": 7.0671,
      "mean_token_accuracy": 0.0931968331336975,
      "num_tokens": 768556.0,
      "step": 370
    },
    {
      "entropy": 7.400697326660156,
      "epoch": 0.034016690856313495,
      "grad_norm": 1.140625,
      "learning_rate": 0.000187,
      "loss": 7.0591,
      "mean_token_accuracy": 0.09333095103502273,
      "num_tokens": 778331.0,
      "step": 375
    },
    {
      "entropy": 7.539828157424926,
      "epoch": 0.03447024673439768,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001895,
      "loss": 7.0387,
      "mean_token_accuracy": 0.09365462064743042,
      "num_tokens": 788065.0,
      "step": 380
    },
    {
      "entropy": 7.495446348190308,
      "epoch": 0.03492380261248186,
      "grad_norm": 1.0,
      "learning_rate": 0.000192,
      "loss": 7.0671,
      "mean_token_accuracy": 0.09628497660160065,
      "num_tokens": 798912.0,
      "step": 385
    },
    {
      "entropy": 7.502418661117554,
      "epoch": 0.03537735849056604,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0001945,
      "loss": 7.0166,
      "mean_token_accuracy": 0.09867717772722244,
      "num_tokens": 809580.0,
      "step": 390
    },
    {
      "entropy": 7.49851360321045,
      "epoch": 0.03583091436865022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.1108,
      "mean_token_accuracy": 0.09084957540035248,
      "num_tokens": 820207.0,
      "step": 395
    },
    {
      "entropy": 7.348110675811768,
      "epoch": 0.036284470246734396,
      "grad_norm": 1.015625,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9336,
      "mean_token_accuracy": 0.08944878205657006,
      "num_tokens": 831038.0,
      "step": 400
    },
    {
      "entropy": 7.545411586761475,
      "epoch": 0.03673802612481858,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000202,
      "loss": 7.0186,
      "mean_token_accuracy": 0.09372441694140435,
      "num_tokens": 842144.0,
      "step": 405
    },
    {
      "entropy": 7.47612156867981,
      "epoch": 0.037191582002902755,
      "grad_norm": 1.609375,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.9989,
      "mean_token_accuracy": 0.09804348424077033,
      "num_tokens": 851315.0,
      "step": 410
    },
    {
      "entropy": 7.322836017608642,
      "epoch": 0.03764513788098694,
      "grad_norm": 1.0625,
      "learning_rate": 0.000207,
      "loss": 7.0063,
      "mean_token_accuracy": 0.09803526178002357,
      "num_tokens": 861233.0,
      "step": 415
    },
    {
      "entropy": 7.406570482254028,
      "epoch": 0.038098693759071114,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002095,
      "loss": 6.9994,
      "mean_token_accuracy": 0.0946230135858059,
      "num_tokens": 871131.0,
      "step": 420
    },
    {
      "entropy": 7.391973686218262,
      "epoch": 0.0385522496371553,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000212,
      "loss": 6.975,
      "mean_token_accuracy": 0.09175262376666068,
      "num_tokens": 882195.0,
      "step": 425
    },
    {
      "entropy": 7.452529048919677,
      "epoch": 0.03900580551523948,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002145,
      "loss": 6.9467,
      "mean_token_accuracy": 0.09589996412396432,
      "num_tokens": 891806.0,
      "step": 430
    },
    {
      "entropy": 7.431856489181518,
      "epoch": 0.03945936139332366,
      "grad_norm": 0.9375,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.0087,
      "mean_token_accuracy": 0.09551806971430779,
      "num_tokens": 901350.0,
      "step": 435
    },
    {
      "entropy": 7.414364910125732,
      "epoch": 0.03991291727140784,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002195,
      "loss": 6.9182,
      "mean_token_accuracy": 0.09558806866407395,
      "num_tokens": 912061.0,
      "step": 440
    },
    {
      "entropy": 7.305682420730591,
      "epoch": 0.040366473149492016,
      "grad_norm": 1.765625,
      "learning_rate": 0.000222,
      "loss": 6.89,
      "mean_token_accuracy": 0.09567090198397636,
      "num_tokens": 922220.0,
      "step": 445
    },
    {
      "entropy": 7.367461013793945,
      "epoch": 0.0408200290275762,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002245,
      "loss": 6.8992,
      "mean_token_accuracy": 0.09762913659214974,
      "num_tokens": 932319.0,
      "step": 450
    },
    {
      "entropy": 7.292191886901856,
      "epoch": 0.041273584905660375,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.9231,
      "mean_token_accuracy": 0.09886682778596878,
      "num_tokens": 943254.0,
      "step": 455
    },
    {
      "entropy": 7.387064361572266,
      "epoch": 0.04172714078374456,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.939,
      "mean_token_accuracy": 0.09336472377181053,
      "num_tokens": 953921.0,
      "step": 460
    },
    {
      "entropy": 7.296272230148316,
      "epoch": 0.042180696661828734,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.8498,
      "mean_token_accuracy": 0.10429160594940186,
      "num_tokens": 963832.0,
      "step": 465
    },
    {
      "entropy": 7.304851102828979,
      "epoch": 0.04263425253991292,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.9459,
      "mean_token_accuracy": 0.09589096009731293,
      "num_tokens": 974066.0,
      "step": 470
    },
    {
      "entropy": 7.324310684204102,
      "epoch": 0.0430878084179971,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000237,
      "loss": 6.9353,
      "mean_token_accuracy": 0.09945596382021904,
      "num_tokens": 984402.0,
      "step": 475
    },
    {
      "entropy": 7.312732458114624,
      "epoch": 0.04354136429608128,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002395,
      "loss": 6.8591,
      "mean_token_accuracy": 0.09938167333602906,
      "num_tokens": 993555.0,
      "step": 480
    },
    {
      "entropy": 7.334518146514893,
      "epoch": 0.04399492017416546,
      "grad_norm": 1.171875,
      "learning_rate": 0.000242,
      "loss": 6.8404,
      "mean_token_accuracy": 0.10574882701039315,
      "num_tokens": 1002920.0,
      "step": 485
    },
    {
      "entropy": 7.289060401916504,
      "epoch": 0.044448476052249636,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002445,
      "loss": 6.907,
      "mean_token_accuracy": 0.09448245465755463,
      "num_tokens": 1013611.0,
      "step": 490
    },
    {
      "entropy": 7.166425371170044,
      "epoch": 0.04490203193033382,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000247,
      "loss": 6.8012,
      "mean_token_accuracy": 0.0957680583000183,
      "num_tokens": 1024121.0,
      "step": 495
    },
    {
      "entropy": 7.363502073287964,
      "epoch": 0.045355587808417995,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002495,
      "loss": 7.032,
      "mean_token_accuracy": 0.1009724922478199,
      "num_tokens": 1033845.0,
      "step": 500
    },
    {
      "entropy": 7.24678144454956,
      "epoch": 0.04580914368650218,
      "grad_norm": 1.0625,
      "learning_rate": 0.000252,
      "loss": 6.831,
      "mean_token_accuracy": 0.09820592999458314,
      "num_tokens": 1044642.0,
      "step": 505
    },
    {
      "entropy": 7.280577898025513,
      "epoch": 0.046262699564586354,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002545,
      "loss": 6.8867,
      "mean_token_accuracy": 0.09730055406689644,
      "num_tokens": 1054728.0,
      "step": 510
    },
    {
      "entropy": 7.20595383644104,
      "epoch": 0.04671625544267054,
      "grad_norm": 1.015625,
      "learning_rate": 0.000257,
      "loss": 6.8504,
      "mean_token_accuracy": 0.0942535676062107,
      "num_tokens": 1065743.0,
      "step": 515
    },
    {
      "entropy": 7.1944211483001705,
      "epoch": 0.04716981132075472,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002595,
      "loss": 6.7719,
      "mean_token_accuracy": 0.09884541779756546,
      "num_tokens": 1076798.0,
      "step": 520
    },
    {
      "entropy": 7.200849342346191,
      "epoch": 0.047623367198838897,
      "grad_norm": 1.078125,
      "learning_rate": 0.000262,
      "loss": 6.8921,
      "mean_token_accuracy": 0.09789391681551933,
      "num_tokens": 1087445.0,
      "step": 525
    },
    {
      "entropy": 7.216012907028198,
      "epoch": 0.04807692307692308,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7985,
      "mean_token_accuracy": 0.10271445363759994,
      "num_tokens": 1098379.0,
      "step": 530
    },
    {
      "entropy": 7.228269386291504,
      "epoch": 0.048530478955007256,
      "grad_norm": 1.171875,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.8866,
      "mean_token_accuracy": 0.09828322008252144,
      "num_tokens": 1108016.0,
      "step": 535
    },
    {
      "entropy": 7.184875774383545,
      "epoch": 0.04898403483309144,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7837,
      "mean_token_accuracy": 0.09823340401053429,
      "num_tokens": 1117887.0,
      "step": 540
    },
    {
      "entropy": 7.186046075820923,
      "epoch": 0.049437590711175615,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.7773,
      "mean_token_accuracy": 0.0981587328016758,
      "num_tokens": 1128497.0,
      "step": 545
    },
    {
      "entropy": 7.1298119068145756,
      "epoch": 0.0498911465892598,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002745,
      "loss": 6.8017,
      "mean_token_accuracy": 0.10637210011482238,
      "num_tokens": 1138168.0,
      "step": 550
    },
    {
      "entropy": 7.136086797714233,
      "epoch": 0.050344702467343974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000277,
      "loss": 6.7618,
      "mean_token_accuracy": 0.10049154385924339,
      "num_tokens": 1148357.0,
      "step": 555
    },
    {
      "entropy": 7.1673754215240475,
      "epoch": 0.05079825834542816,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002795,
      "loss": 6.7907,
      "mean_token_accuracy": 0.10017752796411514,
      "num_tokens": 1159964.0,
      "step": 560
    },
    {
      "entropy": 7.23973536491394,
      "epoch": 0.05125181422351233,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.8508,
      "mean_token_accuracy": 0.09828138649463654,
      "num_tokens": 1170683.0,
      "step": 565
    },
    {
      "entropy": 7.192407703399658,
      "epoch": 0.051705370101596516,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002845,
      "loss": 6.8052,
      "mean_token_accuracy": 0.1006213940680027,
      "num_tokens": 1181680.0,
      "step": 570
    },
    {
      "entropy": 7.093324756622314,
      "epoch": 0.0521589259796807,
      "grad_norm": 1.34375,
      "learning_rate": 0.000287,
      "loss": 6.6997,
      "mean_token_accuracy": 0.10036976039409637,
      "num_tokens": 1191619.0,
      "step": 575
    },
    {
      "entropy": 7.011592292785645,
      "epoch": 0.052612481857764876,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002895,
      "loss": 6.6916,
      "mean_token_accuracy": 0.10292767509818077,
      "num_tokens": 1202264.0,
      "step": 580
    },
    {
      "entropy": 7.043795013427735,
      "epoch": 0.05306603773584906,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000292,
      "loss": 6.6999,
      "mean_token_accuracy": 0.1045819990336895,
      "num_tokens": 1212119.0,
      "step": 585
    },
    {
      "entropy": 7.129189682006836,
      "epoch": 0.053519593613933235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002945,
      "loss": 6.8212,
      "mean_token_accuracy": 0.102519890666008,
      "num_tokens": 1222317.0,
      "step": 590
    },
    {
      "entropy": 7.168576860427857,
      "epoch": 0.05397314949201742,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000297,
      "loss": 6.7716,
      "mean_token_accuracy": 0.10512009784579276,
      "num_tokens": 1232914.0,
      "step": 595
    },
    {
      "entropy": 6.968721580505371,
      "epoch": 0.054426705370101594,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002995,
      "loss": 6.6181,
      "mean_token_accuracy": 0.10760264396667481,
      "num_tokens": 1244214.0,
      "step": 600
    },
    {
      "entropy": 7.105316591262818,
      "epoch": 0.05488026124818578,
      "grad_norm": 1.1875,
      "learning_rate": 0.000302,
      "loss": 6.7149,
      "mean_token_accuracy": 0.10045301765203477,
      "num_tokens": 1254626.0,
      "step": 605
    },
    {
      "entropy": 6.98553991317749,
      "epoch": 0.05533381712626995,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003045,
      "loss": 6.8113,
      "mean_token_accuracy": 0.10129074826836586,
      "num_tokens": 1265461.0,
      "step": 610
    },
    {
      "entropy": 7.082133436203003,
      "epoch": 0.055787373004354136,
      "grad_norm": 1.0625,
      "learning_rate": 0.000307,
      "loss": 6.6817,
      "mean_token_accuracy": 0.11003390699625015,
      "num_tokens": 1275810.0,
      "step": 615
    },
    {
      "entropy": 6.964151763916016,
      "epoch": 0.05624092888243832,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003095,
      "loss": 6.6878,
      "mean_token_accuracy": 0.11000289767980576,
      "num_tokens": 1285170.0,
      "step": 620
    },
    {
      "entropy": 7.021602058410645,
      "epoch": 0.056694484760522496,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000312,
      "loss": 6.6361,
      "mean_token_accuracy": 0.10547553822398185,
      "num_tokens": 1295460.0,
      "step": 625
    },
    {
      "entropy": 7.010881614685059,
      "epoch": 0.05714804063860668,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003145,
      "loss": 6.7523,
      "mean_token_accuracy": 0.10284495502710342,
      "num_tokens": 1306656.0,
      "step": 630
    },
    {
      "entropy": 7.070465517044068,
      "epoch": 0.057601596516690855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000317,
      "loss": 6.7338,
      "mean_token_accuracy": 0.1056830108165741,
      "num_tokens": 1318019.0,
      "step": 635
    },
    {
      "entropy": 6.88353590965271,
      "epoch": 0.05805515239477504,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003195,
      "loss": 6.6511,
      "mean_token_accuracy": 0.10918820276856422,
      "num_tokens": 1328120.0,
      "step": 640
    },
    {
      "entropy": 7.09418797492981,
      "epoch": 0.058508708272859214,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000322,
      "loss": 6.7483,
      "mean_token_accuracy": 0.10312345400452613,
      "num_tokens": 1338354.0,
      "step": 645
    },
    {
      "entropy": 7.009733867645264,
      "epoch": 0.0589622641509434,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.7237,
      "mean_token_accuracy": 0.1024971716105938,
      "num_tokens": 1348666.0,
      "step": 650
    },
    {
      "entropy": 7.047331762313843,
      "epoch": 0.05941582002902757,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.7554,
      "mean_token_accuracy": 0.10805487632751465,
      "num_tokens": 1359713.0,
      "step": 655
    },
    {
      "entropy": 6.9947686195373535,
      "epoch": 0.059869375907111756,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.6124,
      "mean_token_accuracy": 0.10892275646328926,
      "num_tokens": 1369507.0,
      "step": 660
    },
    {
      "entropy": 6.998069620132446,
      "epoch": 0.06032293178519594,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.768,
      "mean_token_accuracy": 0.10034214109182357,
      "num_tokens": 1380726.0,
      "step": 665
    },
    {
      "entropy": 7.112451839447021,
      "epoch": 0.060776487663280115,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.6933,
      "mean_token_accuracy": 0.10579130649566651,
      "num_tokens": 1391423.0,
      "step": 670
    },
    {
      "entropy": 7.052730226516724,
      "epoch": 0.0612300435413643,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000337,
      "loss": 6.6678,
      "mean_token_accuracy": 0.11295373067259788,
      "num_tokens": 1402492.0,
      "step": 675
    },
    {
      "entropy": 6.922665405273437,
      "epoch": 0.061683599419448475,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003395,
      "loss": 6.7161,
      "mean_token_accuracy": 0.10829840004444122,
      "num_tokens": 1412497.0,
      "step": 680
    },
    {
      "entropy": 7.03723840713501,
      "epoch": 0.06213715529753266,
      "grad_norm": 1.171875,
      "learning_rate": 0.000342,
      "loss": 6.6772,
      "mean_token_accuracy": 0.11061608865857124,
      "num_tokens": 1423214.0,
      "step": 685
    },
    {
      "entropy": 6.935256671905518,
      "epoch": 0.06259071117561683,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.6745,
      "mean_token_accuracy": 0.10797292739152908,
      "num_tokens": 1434044.0,
      "step": 690
    },
    {
      "entropy": 6.918494129180909,
      "epoch": 0.06304426705370102,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000347,
      "loss": 6.6361,
      "mean_token_accuracy": 0.10982377529144287,
      "num_tokens": 1444387.0,
      "step": 695
    },
    {
      "entropy": 6.957173490524292,
      "epoch": 0.0634978229317852,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003495,
      "loss": 6.5862,
      "mean_token_accuracy": 0.10744345635175705,
      "num_tokens": 1455157.0,
      "step": 700
    },
    {
      "entropy": 6.938698959350586,
      "epoch": 0.06395137880986937,
      "grad_norm": 1.125,
      "learning_rate": 0.000352,
      "loss": 6.6693,
      "mean_token_accuracy": 0.11478404849767684,
      "num_tokens": 1466145.0,
      "step": 705
    },
    {
      "entropy": 6.962664556503296,
      "epoch": 0.06440493468795355,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003545,
      "loss": 6.667,
      "mean_token_accuracy": 0.10709880217909813,
      "num_tokens": 1476713.0,
      "step": 710
    },
    {
      "entropy": 6.92603645324707,
      "epoch": 0.06485849056603774,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000357,
      "loss": 6.671,
      "mean_token_accuracy": 0.11409479826688766,
      "num_tokens": 1486908.0,
      "step": 715
    },
    {
      "entropy": 6.899812459945679,
      "epoch": 0.06531204644412192,
      "grad_norm": 1.375,
      "learning_rate": 0.0003595,
      "loss": 6.5941,
      "mean_token_accuracy": 0.11084994897246361,
      "num_tokens": 1498185.0,
      "step": 720
    },
    {
      "entropy": 6.901278781890869,
      "epoch": 0.0657656023222061,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000362,
      "loss": 6.6413,
      "mean_token_accuracy": 0.1077586367726326,
      "num_tokens": 1508376.0,
      "step": 725
    },
    {
      "entropy": 6.9933178424835205,
      "epoch": 0.06621915820029027,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003645,
      "loss": 6.7199,
      "mean_token_accuracy": 0.10835012272000313,
      "num_tokens": 1519953.0,
      "step": 730
    },
    {
      "entropy": 6.853632211685181,
      "epoch": 0.06667271407837445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000367,
      "loss": 6.592,
      "mean_token_accuracy": 0.11243074238300324,
      "num_tokens": 1529793.0,
      "step": 735
    },
    {
      "entropy": 6.829979610443115,
      "epoch": 0.06712626995645864,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003695,
      "loss": 6.5882,
      "mean_token_accuracy": 0.11395004838705063,
      "num_tokens": 1540859.0,
      "step": 740
    },
    {
      "entropy": 6.908363008499146,
      "epoch": 0.06757982583454282,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000372,
      "loss": 6.5104,
      "mean_token_accuracy": 0.11262088865041733,
      "num_tokens": 1549783.0,
      "step": 745
    },
    {
      "entropy": 6.84120545387268,
      "epoch": 0.06803338171262699,
      "grad_norm": 1.125,
      "learning_rate": 0.0003745,
      "loss": 6.5803,
      "mean_token_accuracy": 0.10933975651860237,
      "num_tokens": 1560522.0,
      "step": 750
    },
    {
      "entropy": 6.878204107284546,
      "epoch": 0.06848693759071117,
      "grad_norm": 1.421875,
      "learning_rate": 0.000377,
      "loss": 6.553,
      "mean_token_accuracy": 0.11197188422083855,
      "num_tokens": 1570668.0,
      "step": 755
    },
    {
      "entropy": 6.856545925140381,
      "epoch": 0.06894049346879536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003795,
      "loss": 6.669,
      "mean_token_accuracy": 0.11744550019502639,
      "num_tokens": 1581523.0,
      "step": 760
    },
    {
      "entropy": 6.84909029006958,
      "epoch": 0.06939404934687954,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000382,
      "loss": 6.5917,
      "mean_token_accuracy": 0.1113414227962494,
      "num_tokens": 1591356.0,
      "step": 765
    },
    {
      "entropy": 6.83960976600647,
      "epoch": 0.06984760522496372,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003845,
      "loss": 6.5574,
      "mean_token_accuracy": 0.11633669957518578,
      "num_tokens": 1601779.0,
      "step": 770
    },
    {
      "entropy": 6.888131904602051,
      "epoch": 0.07030116110304789,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.5633,
      "mean_token_accuracy": 0.1150720700621605,
      "num_tokens": 1611995.0,
      "step": 775
    },
    {
      "entropy": 6.856516313552857,
      "epoch": 0.07075471698113207,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.586,
      "mean_token_accuracy": 0.11050437688827515,
      "num_tokens": 1621834.0,
      "step": 780
    },
    {
      "entropy": 6.778967618942261,
      "epoch": 0.07120827285921626,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.6288,
      "mean_token_accuracy": 0.10744166076183319,
      "num_tokens": 1632881.0,
      "step": 785
    },
    {
      "entropy": 6.7607427597045895,
      "epoch": 0.07166182873730044,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.6028,
      "mean_token_accuracy": 0.10911152362823487,
      "num_tokens": 1642870.0,
      "step": 790
    },
    {
      "entropy": 6.8341546058654785,
      "epoch": 0.07211538461538461,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.484,
      "mean_token_accuracy": 0.11779236048460007,
      "num_tokens": 1653390.0,
      "step": 795
    },
    {
      "entropy": 6.802682399749756,
      "epoch": 0.07256894049346879,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003995,
      "loss": 6.6078,
      "mean_token_accuracy": 0.11580143421888352,
      "num_tokens": 1663382.0,
      "step": 800
    },
    {
      "entropy": 6.889751863479614,
      "epoch": 0.07302249637155298,
      "grad_norm": 1.03125,
      "learning_rate": 0.000402,
      "loss": 6.6263,
      "mean_token_accuracy": 0.10960384383797646,
      "num_tokens": 1673601.0,
      "step": 805
    },
    {
      "entropy": 6.744623231887817,
      "epoch": 0.07347605224963716,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004045,
      "loss": 6.5301,
      "mean_token_accuracy": 0.1178050920367241,
      "num_tokens": 1683565.0,
      "step": 810
    },
    {
      "entropy": 6.720245599746704,
      "epoch": 0.07392960812772134,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4975,
      "mean_token_accuracy": 0.1157220259308815,
      "num_tokens": 1693636.0,
      "step": 815
    },
    {
      "entropy": 6.888108444213867,
      "epoch": 0.07438316400580551,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004095,
      "loss": 6.6028,
      "mean_token_accuracy": 0.11490728184580803,
      "num_tokens": 1703415.0,
      "step": 820
    },
    {
      "entropy": 6.8595898151397705,
      "epoch": 0.0748367198838897,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000412,
      "loss": 6.5759,
      "mean_token_accuracy": 0.11677221283316612,
      "num_tokens": 1713579.0,
      "step": 825
    },
    {
      "entropy": 6.7610352516174315,
      "epoch": 0.07529027576197388,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004145,
      "loss": 6.5869,
      "mean_token_accuracy": 0.11174456626176835,
      "num_tokens": 1724583.0,
      "step": 830
    },
    {
      "entropy": 6.793128204345703,
      "epoch": 0.07574383164005806,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000417,
      "loss": 6.4998,
      "mean_token_accuracy": 0.12149128615856171,
      "num_tokens": 1735293.0,
      "step": 835
    },
    {
      "entropy": 6.784815979003906,
      "epoch": 0.07619738751814223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004195,
      "loss": 6.5285,
      "mean_token_accuracy": 0.11545130759477615,
      "num_tokens": 1745623.0,
      "step": 840
    },
    {
      "entropy": 6.785441732406616,
      "epoch": 0.07665094339622641,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000422,
      "loss": 6.5589,
      "mean_token_accuracy": 0.11196033582091332,
      "num_tokens": 1755713.0,
      "step": 845
    },
    {
      "entropy": 6.844113492965699,
      "epoch": 0.0771044992743106,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004245,
      "loss": 6.5448,
      "mean_token_accuracy": 0.11228399202227593,
      "num_tokens": 1765264.0,
      "step": 850
    },
    {
      "entropy": 6.778715133666992,
      "epoch": 0.07755805515239478,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000427,
      "loss": 6.5551,
      "mean_token_accuracy": 0.11371968314051628,
      "num_tokens": 1775769.0,
      "step": 855
    },
    {
      "entropy": 6.709914636611939,
      "epoch": 0.07801161103047896,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004295,
      "loss": 6.4821,
      "mean_token_accuracy": 0.11689155176281929,
      "num_tokens": 1786245.0,
      "step": 860
    },
    {
      "entropy": 6.716907548904419,
      "epoch": 0.07846516690856313,
      "grad_norm": 1.21875,
      "learning_rate": 0.000432,
      "loss": 6.5182,
      "mean_token_accuracy": 0.12198375314474105,
      "num_tokens": 1796040.0,
      "step": 865
    },
    {
      "entropy": 6.72428708076477,
      "epoch": 0.07891872278664731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004345,
      "loss": 6.5268,
      "mean_token_accuracy": 0.11597820520401,
      "num_tokens": 1806007.0,
      "step": 870
    },
    {
      "entropy": 6.862047719955444,
      "epoch": 0.0793722786647315,
      "grad_norm": 1.109375,
      "learning_rate": 0.000437,
      "loss": 6.6034,
      "mean_token_accuracy": 0.11016635000705718,
      "num_tokens": 1816935.0,
      "step": 875
    },
    {
      "entropy": 6.7226348400115965,
      "epoch": 0.07982583454281568,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004395,
      "loss": 6.4647,
      "mean_token_accuracy": 0.11464838534593583,
      "num_tokens": 1827419.0,
      "step": 880
    },
    {
      "entropy": 6.80533652305603,
      "epoch": 0.08027939042089985,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000442,
      "loss": 6.5762,
      "mean_token_accuracy": 0.11279575228691101,
      "num_tokens": 1837933.0,
      "step": 885
    },
    {
      "entropy": 6.692126846313476,
      "epoch": 0.08073294629898403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004445,
      "loss": 6.6181,
      "mean_token_accuracy": 0.11968202516436577,
      "num_tokens": 1848551.0,
      "step": 890
    },
    {
      "entropy": 6.773914003372193,
      "epoch": 0.08118650217706821,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000447,
      "loss": 6.4874,
      "mean_token_accuracy": 0.11803188994526863,
      "num_tokens": 1858380.0,
      "step": 895
    },
    {
      "entropy": 6.64626522064209,
      "epoch": 0.0816400580551524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.4255,
      "mean_token_accuracy": 0.11616697832942009,
      "num_tokens": 1868291.0,
      "step": 900
    },
    {
      "entropy": 6.739772891998291,
      "epoch": 0.08209361393323658,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.5222,
      "mean_token_accuracy": 0.11596911773085594,
      "num_tokens": 1879219.0,
      "step": 905
    },
    {
      "entropy": 6.701965379714966,
      "epoch": 0.08254716981132075,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.5064,
      "mean_token_accuracy": 0.1169955737888813,
      "num_tokens": 1889809.0,
      "step": 910
    },
    {
      "entropy": 6.736312532424927,
      "epoch": 0.08300072568940493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4667,
      "mean_token_accuracy": 0.11598963215947151,
      "num_tokens": 1899986.0,
      "step": 915
    },
    {
      "entropy": 6.627064943313599,
      "epoch": 0.08345428156748912,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4631,
      "mean_token_accuracy": 0.11359920203685761,
      "num_tokens": 1909314.0,
      "step": 920
    },
    {
      "entropy": 6.714141511917115,
      "epoch": 0.0839078374455733,
      "grad_norm": 1.078125,
      "learning_rate": 0.000462,
      "loss": 6.4954,
      "mean_token_accuracy": 0.1206825278699398,
      "num_tokens": 1918443.0,
      "step": 925
    },
    {
      "entropy": 6.711059856414795,
      "epoch": 0.08436139332365747,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004645,
      "loss": 6.4599,
      "mean_token_accuracy": 0.11760578602552414,
      "num_tokens": 1928683.0,
      "step": 930
    },
    {
      "entropy": 6.563105583190918,
      "epoch": 0.08481494920174165,
      "grad_norm": 1.109375,
      "learning_rate": 0.000467,
      "loss": 6.3733,
      "mean_token_accuracy": 0.11783806160092354,
      "num_tokens": 1938293.0,
      "step": 935
    },
    {
      "entropy": 6.621684551239014,
      "epoch": 0.08526850507982583,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004695,
      "loss": 6.4649,
      "mean_token_accuracy": 0.1159488506615162,
      "num_tokens": 1948233.0,
      "step": 940
    },
    {
      "entropy": 6.703399705886841,
      "epoch": 0.08572206095791002,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000472,
      "loss": 6.3795,
      "mean_token_accuracy": 0.12225733622908593,
      "num_tokens": 1958611.0,
      "step": 945
    },
    {
      "entropy": 6.67765588760376,
      "epoch": 0.0861756168359942,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004745,
      "loss": 6.5279,
      "mean_token_accuracy": 0.12170163914561272,
      "num_tokens": 1968957.0,
      "step": 950
    },
    {
      "entropy": 6.721297121047973,
      "epoch": 0.08662917271407837,
      "grad_norm": 1.203125,
      "learning_rate": 0.000477,
      "loss": 6.4022,
      "mean_token_accuracy": 0.12271194607019424,
      "num_tokens": 1978503.0,
      "step": 955
    },
    {
      "entropy": 6.51479959487915,
      "epoch": 0.08708272859216255,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004795,
      "loss": 6.3838,
      "mean_token_accuracy": 0.12056222185492516,
      "num_tokens": 1988458.0,
      "step": 960
    },
    {
      "entropy": 6.663824939727784,
      "epoch": 0.08753628447024674,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000482,
      "loss": 6.5235,
      "mean_token_accuracy": 0.1133016251027584,
      "num_tokens": 1999984.0,
      "step": 965
    },
    {
      "entropy": 6.534238243103028,
      "epoch": 0.08798984034833092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004845,
      "loss": 6.38,
      "mean_token_accuracy": 0.12614745199680327,
      "num_tokens": 2010645.0,
      "step": 970
    },
    {
      "entropy": 6.69314603805542,
      "epoch": 0.08844339622641509,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000487,
      "loss": 6.4879,
      "mean_token_accuracy": 0.11517181396484374,
      "num_tokens": 2020196.0,
      "step": 975
    },
    {
      "entropy": 6.586405944824219,
      "epoch": 0.08889695210449927,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004895,
      "loss": 6.3749,
      "mean_token_accuracy": 0.12108265981078148,
      "num_tokens": 2030101.0,
      "step": 980
    },
    {
      "entropy": 6.620171546936035,
      "epoch": 0.08935050798258345,
      "grad_norm": 1.03125,
      "learning_rate": 0.000492,
      "loss": 6.4594,
      "mean_token_accuracy": 0.1197955846786499,
      "num_tokens": 2040454.0,
      "step": 985
    },
    {
      "entropy": 6.581764078140258,
      "epoch": 0.08980406386066764,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004945,
      "loss": 6.402,
      "mean_token_accuracy": 0.1192368857562542,
      "num_tokens": 2050378.0,
      "step": 990
    },
    {
      "entropy": 6.651663303375244,
      "epoch": 0.09025761973875182,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000497,
      "loss": 6.4549,
      "mean_token_accuracy": 0.11869885623455048,
      "num_tokens": 2060282.0,
      "step": 995
    },
    {
      "entropy": 6.589950847625732,
      "epoch": 0.09071117561683599,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004995,
      "loss": 6.4433,
      "mean_token_accuracy": 0.12011284530162811,
      "num_tokens": 2070563.0,
      "step": 1000
    },
    {
      "entropy": 6.661682939529419,
      "epoch": 0.09116473149492017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999985110217,
      "loss": 6.5128,
      "mean_token_accuracy": 0.11984694451093673,
      "num_tokens": 2080153.0,
      "step": 1005
    },
    {
      "entropy": 6.546395254135132,
      "epoch": 0.09161828737300436,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999999924620471,
      "loss": 6.3602,
      "mean_token_accuracy": 0.12557145729660987,
      "num_tokens": 2090590.0,
      "step": 1010
    },
    {
      "entropy": 6.653158855438233,
      "epoch": 0.09207184325108854,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999999817600155,
      "loss": 6.4314,
      "mean_token_accuracy": 0.11608105525374413,
      "num_tokens": 2101945.0,
      "step": 1015
    },
    {
      "entropy": 6.635874509811401,
      "epoch": 0.09252539912917271,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999999664049268,
      "loss": 6.4312,
      "mean_token_accuracy": 0.12001314610242844,
      "num_tokens": 2112514.0,
      "step": 1020
    },
    {
      "entropy": 6.546846055984497,
      "epoch": 0.09297895500725689,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004999999463967816,
      "loss": 6.3852,
      "mean_token_accuracy": 0.11658050194382667,
      "num_tokens": 2122296.0,
      "step": 1025
    },
    {
      "entropy": 6.713669490814209,
      "epoch": 0.09343251088534107,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049999992173558,
      "loss": 6.5311,
      "mean_token_accuracy": 0.10938856825232506,
      "num_tokens": 2134512.0,
      "step": 1030
    },
    {
      "entropy": 6.649443817138672,
      "epoch": 0.09388606676342526,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999998924213228,
      "loss": 6.4339,
      "mean_token_accuracy": 0.12238849774003029,
      "num_tokens": 2143934.0,
      "step": 1035
    },
    {
      "entropy": 6.488268947601318,
      "epoch": 0.09433962264150944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999998584540105,
      "loss": 6.3351,
      "mean_token_accuracy": 0.11762306764721871,
      "num_tokens": 2155730.0,
      "step": 1040
    },
    {
      "entropy": 6.570074510574341,
      "epoch": 0.09479317851959361,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999998198336437,
      "loss": 6.3681,
      "mean_token_accuracy": 0.11555026769638062,
      "num_tokens": 2166422.0,
      "step": 1045
    },
    {
      "entropy": 6.646327686309815,
      "epoch": 0.09524673439767779,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999997765602233,
      "loss": 6.469,
      "mean_token_accuracy": 0.11361956149339676,
      "num_tokens": 2177842.0,
      "step": 1050
    },
    {
      "entropy": 6.617121315002441,
      "epoch": 0.09570029027576198,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999997286337502,
      "loss": 6.5398,
      "mean_token_accuracy": 0.11744040250778198,
      "num_tokens": 2188614.0,
      "step": 1055
    },
    {
      "entropy": 6.604910278320313,
      "epoch": 0.09615384615384616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999996760542254,
      "loss": 6.3246,
      "mean_token_accuracy": 0.12567146718502045,
      "num_tokens": 2197886.0,
      "step": 1060
    },
    {
      "entropy": 6.525793075561523,
      "epoch": 0.09660740203193033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049999961882165,
      "loss": 6.3772,
      "mean_token_accuracy": 0.12232817932963372,
      "num_tokens": 2209930.0,
      "step": 1065
    },
    {
      "entropy": 6.564913749694824,
      "epoch": 0.09706095791001451,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999995569360251,
      "loss": 6.3959,
      "mean_token_accuracy": 0.11825868040323258,
      "num_tokens": 2219767.0,
      "step": 1070
    },
    {
      "entropy": 6.555662298202515,
      "epoch": 0.0975145137880987,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499999490397352,
      "loss": 6.3192,
      "mean_token_accuracy": 0.121416836977005,
      "num_tokens": 2230082.0,
      "step": 1075
    },
    {
      "entropy": 6.495458173751831,
      "epoch": 0.09796806966618288,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999994192056321,
      "loss": 6.3673,
      "mean_token_accuracy": 0.12748734578490256,
      "num_tokens": 2240066.0,
      "step": 1080
    },
    {
      "entropy": 6.524483871459961,
      "epoch": 0.09842162554426705,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999993433608669,
      "loss": 6.3942,
      "mean_token_accuracy": 0.11839633882045746,
      "num_tokens": 2250732.0,
      "step": 1085
    },
    {
      "entropy": 6.593288707733154,
      "epoch": 0.09887518142235123,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999992628630579,
      "loss": 6.2797,
      "mean_token_accuracy": 0.12327790930867195,
      "num_tokens": 2261452.0,
      "step": 1090
    },
    {
      "entropy": 6.485133695602417,
      "epoch": 0.09932873730043541,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999991777122069,
      "loss": 6.3782,
      "mean_token_accuracy": 0.12279095351696015,
      "num_tokens": 2271929.0,
      "step": 1095
    },
    {
      "entropy": 6.531335401535034,
      "epoch": 0.0997822931785196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999990879083156,
      "loss": 6.3604,
      "mean_token_accuracy": 0.12319791615009308,
      "num_tokens": 2282987.0,
      "step": 1100
    },
    {
      "entropy": 6.444648933410645,
      "epoch": 0.10023584905660378,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999989934513857,
      "loss": 6.3295,
      "mean_token_accuracy": 0.12565099447965622,
      "num_tokens": 2292796.0,
      "step": 1105
    },
    {
      "entropy": 6.598123359680176,
      "epoch": 0.10068940493468795,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999988943414193,
      "loss": 6.3766,
      "mean_token_accuracy": 0.12076297253370286,
      "num_tokens": 2303057.0,
      "step": 1110
    },
    {
      "entropy": 6.4050017356872555,
      "epoch": 0.10114296081277213,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999987905784184,
      "loss": 6.3638,
      "mean_token_accuracy": 0.12473835423588753,
      "num_tokens": 2314168.0,
      "step": 1115
    },
    {
      "entropy": 6.548490381240844,
      "epoch": 0.10159651669085631,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999986821623853,
      "loss": 6.335,
      "mean_token_accuracy": 0.12088372111320496,
      "num_tokens": 2325241.0,
      "step": 1120
    },
    {
      "entropy": 6.579931640625,
      "epoch": 0.1020500725689405,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999985690933219,
      "loss": 6.3584,
      "mean_token_accuracy": 0.12107386961579322,
      "num_tokens": 2334866.0,
      "step": 1125
    },
    {
      "entropy": 6.416501808166504,
      "epoch": 0.10250362844702467,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999984513712311,
      "loss": 6.366,
      "mean_token_accuracy": 0.12216197848320007,
      "num_tokens": 2345422.0,
      "step": 1130
    },
    {
      "entropy": 6.573203802108765,
      "epoch": 0.10295718432510885,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999983289961146,
      "loss": 6.3667,
      "mean_token_accuracy": 0.12462597936391831,
      "num_tokens": 2356285.0,
      "step": 1135
    },
    {
      "entropy": 6.414275217056274,
      "epoch": 0.10341074020319303,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999982019679755,
      "loss": 6.2962,
      "mean_token_accuracy": 0.12918335497379302,
      "num_tokens": 2366861.0,
      "step": 1140
    },
    {
      "entropy": 6.546162033081055,
      "epoch": 0.10386429608127722,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999980702868164,
      "loss": 6.3186,
      "mean_token_accuracy": 0.12778222039341927,
      "num_tokens": 2376851.0,
      "step": 1145
    },
    {
      "entropy": 6.468019723892212,
      "epoch": 0.1043178519593614,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999979339526396,
      "loss": 6.3967,
      "mean_token_accuracy": 0.12601374089717865,
      "num_tokens": 2386857.0,
      "step": 1150
    },
    {
      "entropy": 6.512000989913941,
      "epoch": 0.10477140783744557,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999977929654484,
      "loss": 6.3197,
      "mean_token_accuracy": 0.12597798183560371,
      "num_tokens": 2397035.0,
      "step": 1155
    },
    {
      "entropy": 6.556376361846924,
      "epoch": 0.10522496371552975,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004999976473252453,
      "loss": 6.3936,
      "mean_token_accuracy": 0.12374778017401696,
      "num_tokens": 2408177.0,
      "step": 1160
    },
    {
      "entropy": 6.569843435287476,
      "epoch": 0.10567851959361393,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999974970320338,
      "loss": 6.342,
      "mean_token_accuracy": 0.12336181774735451,
      "num_tokens": 2418445.0,
      "step": 1165
    },
    {
      "entropy": 6.446996784210205,
      "epoch": 0.10613207547169812,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999973420858165,
      "loss": 6.3375,
      "mean_token_accuracy": 0.12320749461650848,
      "num_tokens": 2428990.0,
      "step": 1170
    },
    {
      "entropy": 6.456551933288575,
      "epoch": 0.10658563134978229,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999971824865968,
      "loss": 6.3706,
      "mean_token_accuracy": 0.11951386407017708,
      "num_tokens": 2439643.0,
      "step": 1175
    },
    {
      "entropy": 6.636844635009766,
      "epoch": 0.10703918722786647,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999970182343782,
      "loss": 6.4152,
      "mean_token_accuracy": 0.12312159314751625,
      "num_tokens": 2450691.0,
      "step": 1180
    },
    {
      "entropy": 6.513713598251343,
      "epoch": 0.10749274310595065,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999968493291638,
      "loss": 6.3519,
      "mean_token_accuracy": 0.12444976568222046,
      "num_tokens": 2460706.0,
      "step": 1185
    },
    {
      "entropy": 6.332309865951538,
      "epoch": 0.10794629898403484,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999966757709573,
      "loss": 6.2426,
      "mean_token_accuracy": 0.13491440415382386,
      "num_tokens": 2470947.0,
      "step": 1190
    },
    {
      "entropy": 6.500934410095215,
      "epoch": 0.10839985486211902,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999964975597622,
      "loss": 6.2077,
      "mean_token_accuracy": 0.12923589274287223,
      "num_tokens": 2481132.0,
      "step": 1195
    },
    {
      "entropy": 6.5095055103302,
      "epoch": 0.10885341074020319,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999963146955821,
      "loss": 6.2492,
      "mean_token_accuracy": 0.12882558181881903,
      "num_tokens": 2491701.0,
      "step": 1200
    },
    {
      "entropy": 6.308234691619873,
      "epoch": 0.10930696661828737,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499996127178421,
      "loss": 6.3173,
      "mean_token_accuracy": 0.1222999058663845,
      "num_tokens": 2502360.0,
      "step": 1205
    },
    {
      "entropy": 6.526607275009155,
      "epoch": 0.10976052249637155,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999959350082825,
      "loss": 6.2211,
      "mean_token_accuracy": 0.1294765926897526,
      "num_tokens": 2512140.0,
      "step": 1210
    },
    {
      "entropy": 6.4609130859375,
      "epoch": 0.11021407837445574,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999957381851709,
      "loss": 6.3314,
      "mean_token_accuracy": 0.12224320769309997,
      "num_tokens": 2522721.0,
      "step": 1215
    },
    {
      "entropy": 6.3936292171478275,
      "epoch": 0.1106676342525399,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999955367090901,
      "loss": 6.1999,
      "mean_token_accuracy": 0.1303488314151764,
      "num_tokens": 2531786.0,
      "step": 1220
    },
    {
      "entropy": 6.481017780303955,
      "epoch": 0.11112119013062409,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999953305800441,
      "loss": 6.3062,
      "mean_token_accuracy": 0.12003768384456634,
      "num_tokens": 2542895.0,
      "step": 1225
    },
    {
      "entropy": 6.442375755310058,
      "epoch": 0.11157474600870827,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999951197980374,
      "loss": 6.305,
      "mean_token_accuracy": 0.12665605992078782,
      "num_tokens": 2554120.0,
      "step": 1230
    },
    {
      "entropy": 6.489529991149903,
      "epoch": 0.11202830188679246,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999949043630744,
      "loss": 6.3339,
      "mean_token_accuracy": 0.12409112006425857,
      "num_tokens": 2563935.0,
      "step": 1235
    },
    {
      "entropy": 6.3626786231994625,
      "epoch": 0.11248185776487664,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999946842751593,
      "loss": 6.2477,
      "mean_token_accuracy": 0.1251929335296154,
      "num_tokens": 2574465.0,
      "step": 1240
    },
    {
      "entropy": 6.431154251098633,
      "epoch": 0.11293541364296081,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999944595342968,
      "loss": 6.2144,
      "mean_token_accuracy": 0.12585682272911072,
      "num_tokens": 2584694.0,
      "step": 1245
    },
    {
      "entropy": 6.402558374404907,
      "epoch": 0.11338896952104499,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999942301404916,
      "loss": 6.1836,
      "mean_token_accuracy": 0.1319870851933956,
      "num_tokens": 2594540.0,
      "step": 1250
    },
    {
      "entropy": 6.334234380722046,
      "epoch": 0.11384252539912917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999939960937484,
      "loss": 6.1343,
      "mean_token_accuracy": 0.1355540633201599,
      "num_tokens": 2604896.0,
      "step": 1255
    },
    {
      "entropy": 6.494117069244385,
      "epoch": 0.11429608127721336,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999937573940721,
      "loss": 6.2886,
      "mean_token_accuracy": 0.1345356211066246,
      "num_tokens": 2614143.0,
      "step": 1260
    },
    {
      "entropy": 6.274349880218506,
      "epoch": 0.11474963715529753,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999935140414674,
      "loss": 6.1951,
      "mean_token_accuracy": 0.12857473343610765,
      "num_tokens": 2623999.0,
      "step": 1265
    },
    {
      "entropy": 6.511336469650269,
      "epoch": 0.11520319303338171,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999932660359395,
      "loss": 6.4052,
      "mean_token_accuracy": 0.12467710450291633,
      "num_tokens": 2636208.0,
      "step": 1270
    },
    {
      "entropy": 6.449693775177002,
      "epoch": 0.11565674891146589,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999930133774936,
      "loss": 6.2795,
      "mean_token_accuracy": 0.13579537123441696,
      "num_tokens": 2645854.0,
      "step": 1275
    },
    {
      "entropy": 6.469302797317505,
      "epoch": 0.11611030478955008,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999927560661348,
      "loss": 6.2593,
      "mean_token_accuracy": 0.1359974443912506,
      "num_tokens": 2655877.0,
      "step": 1280
    },
    {
      "entropy": 6.474381637573242,
      "epoch": 0.11656386066763426,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999924941018685,
      "loss": 6.2797,
      "mean_token_accuracy": 0.1305759347975254,
      "num_tokens": 2666157.0,
      "step": 1285
    },
    {
      "entropy": 6.2998865127563475,
      "epoch": 0.11701741654571843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999922274847,
      "loss": 6.1961,
      "mean_token_accuracy": 0.13332010209560394,
      "num_tokens": 2676647.0,
      "step": 1290
    },
    {
      "entropy": 6.388902854919434,
      "epoch": 0.11747097242380261,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499991956214635,
      "loss": 6.2308,
      "mean_token_accuracy": 0.14106824845075608,
      "num_tokens": 2686417.0,
      "step": 1295
    },
    {
      "entropy": 6.417198610305786,
      "epoch": 0.1179245283018868,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499991680291679,
      "loss": 6.2281,
      "mean_token_accuracy": 0.12554471418261529,
      "num_tokens": 2697540.0,
      "step": 1300
    },
    {
      "entropy": 6.371242427825928,
      "epoch": 0.11837808417997098,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999913997158378,
      "loss": 6.2766,
      "mean_token_accuracy": 0.12081151455640793,
      "num_tokens": 2709397.0,
      "step": 1305
    },
    {
      "entropy": 6.381704521179199,
      "epoch": 0.11883164005805515,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999911144871169,
      "loss": 6.2198,
      "mean_token_accuracy": 0.12881449088454247,
      "num_tokens": 2720430.0,
      "step": 1310
    },
    {
      "entropy": 6.3669020652771,
      "epoch": 0.11928519593613933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999908246055226,
      "loss": 6.1855,
      "mean_token_accuracy": 0.1260046534240246,
      "num_tokens": 2730648.0,
      "step": 1315
    },
    {
      "entropy": 6.3640461444854735,
      "epoch": 0.11973875181422351,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999905300710607,
      "loss": 6.1767,
      "mean_token_accuracy": 0.12971787378191948,
      "num_tokens": 2739979.0,
      "step": 1320
    },
    {
      "entropy": 6.467068099975586,
      "epoch": 0.1201923076923077,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999902308837373,
      "loss": 6.3039,
      "mean_token_accuracy": 0.12591902986168862,
      "num_tokens": 2750329.0,
      "step": 1325
    },
    {
      "entropy": 6.297173976898193,
      "epoch": 0.12064586357039188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999899270435584,
      "loss": 6.1752,
      "mean_token_accuracy": 0.1377357542514801,
      "num_tokens": 2760535.0,
      "step": 1330
    },
    {
      "entropy": 6.402283620834351,
      "epoch": 0.12109941944847605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999896185505307,
      "loss": 6.2099,
      "mean_token_accuracy": 0.1282433770596981,
      "num_tokens": 2770470.0,
      "step": 1335
    },
    {
      "entropy": 6.357119464874268,
      "epoch": 0.12155297532656023,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999893054046603,
      "loss": 6.2395,
      "mean_token_accuracy": 0.12727272063493728,
      "num_tokens": 2781506.0,
      "step": 1340
    },
    {
      "entropy": 6.281704378128052,
      "epoch": 0.12200653120464441,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999889876059537,
      "loss": 6.0945,
      "mean_token_accuracy": 0.1367440052330494,
      "num_tokens": 2791245.0,
      "step": 1345
    },
    {
      "entropy": 6.390601301193238,
      "epoch": 0.1224600870827286,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999886651544176,
      "loss": 6.1272,
      "mean_token_accuracy": 0.13428937569260596,
      "num_tokens": 2801680.0,
      "step": 1350
    },
    {
      "entropy": 6.298570585250855,
      "epoch": 0.12291364296081277,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999883380500584,
      "loss": 6.2077,
      "mean_token_accuracy": 0.13555841743946076,
      "num_tokens": 2812805.0,
      "step": 1355
    },
    {
      "entropy": 6.452785491943359,
      "epoch": 0.12336719883889695,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999880062928831,
      "loss": 6.2401,
      "mean_token_accuracy": 0.12772336453199387,
      "num_tokens": 2822834.0,
      "step": 1360
    },
    {
      "entropy": 6.322773456573486,
      "epoch": 0.12382075471698113,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999876698828985,
      "loss": 6.1829,
      "mean_token_accuracy": 0.1308009184896946,
      "num_tokens": 2832312.0,
      "step": 1365
    },
    {
      "entropy": 6.387462949752807,
      "epoch": 0.12427431059506532,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999873288201116,
      "loss": 6.1993,
      "mean_token_accuracy": 0.13682973608374596,
      "num_tokens": 2842998.0,
      "step": 1370
    },
    {
      "entropy": 6.381214332580567,
      "epoch": 0.1247278664731495,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999869831045294,
      "loss": 6.2182,
      "mean_token_accuracy": 0.1281896151602268,
      "num_tokens": 2854218.0,
      "step": 1375
    },
    {
      "entropy": 6.274281024932861,
      "epoch": 0.12518142235123367,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999866327361589,
      "loss": 6.1073,
      "mean_token_accuracy": 0.13887619599699974,
      "num_tokens": 2864484.0,
      "step": 1380
    },
    {
      "entropy": 6.392649173736572,
      "epoch": 0.12563497822931785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999862777150077,
      "loss": 6.2187,
      "mean_token_accuracy": 0.13294630274176597,
      "num_tokens": 2873800.0,
      "step": 1385
    },
    {
      "entropy": 6.378524923324585,
      "epoch": 0.12608853410740203,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999859180410829,
      "loss": 6.2196,
      "mean_token_accuracy": 0.12941672652959824,
      "num_tokens": 2884247.0,
      "step": 1390
    },
    {
      "entropy": 6.3022987842559814,
      "epoch": 0.12654208998548622,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999855537143919,
      "loss": 6.1277,
      "mean_token_accuracy": 0.13321435153484346,
      "num_tokens": 2894739.0,
      "step": 1395
    },
    {
      "entropy": 6.330821084976196,
      "epoch": 0.1269956458635704,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999851847349425,
      "loss": 6.2018,
      "mean_token_accuracy": 0.13309190422296524,
      "num_tokens": 2904376.0,
      "step": 1400
    },
    {
      "entropy": 6.41134204864502,
      "epoch": 0.12744920174165458,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999848111027419,
      "loss": 6.2438,
      "mean_token_accuracy": 0.12967129945755004,
      "num_tokens": 2914304.0,
      "step": 1405
    },
    {
      "entropy": 6.346340560913086,
      "epoch": 0.12790275761973874,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999844328177983,
      "loss": 6.1142,
      "mean_token_accuracy": 0.13302754908800124,
      "num_tokens": 2924400.0,
      "step": 1410
    },
    {
      "entropy": 6.468166971206665,
      "epoch": 0.12835631349782292,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999840498801191,
      "loss": 6.3637,
      "mean_token_accuracy": 0.13450959250330924,
      "num_tokens": 2935040.0,
      "step": 1415
    },
    {
      "entropy": 6.250949144363403,
      "epoch": 0.1288098693759071,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999836622897126,
      "loss": 6.139,
      "mean_token_accuracy": 0.13291264548897744,
      "num_tokens": 2945536.0,
      "step": 1420
    },
    {
      "entropy": 6.240394401550293,
      "epoch": 0.1292634252539913,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999832700465865,
      "loss": 6.0631,
      "mean_token_accuracy": 0.13600366339087486,
      "num_tokens": 2954969.0,
      "step": 1425
    },
    {
      "entropy": 6.372559261322022,
      "epoch": 0.12971698113207547,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999828731507491,
      "loss": 6.1846,
      "mean_token_accuracy": 0.13019020035862922,
      "num_tokens": 2966218.0,
      "step": 1430
    },
    {
      "entropy": 6.348944854736328,
      "epoch": 0.13017053701015965,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999824716022085,
      "loss": 6.2124,
      "mean_token_accuracy": 0.12568282783031465,
      "num_tokens": 2976640.0,
      "step": 1435
    },
    {
      "entropy": 6.317825126647949,
      "epoch": 0.13062409288824384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999820654009731,
      "loss": 6.1754,
      "mean_token_accuracy": 0.13333930373191832,
      "num_tokens": 2985883.0,
      "step": 1440
    },
    {
      "entropy": 6.33205885887146,
      "epoch": 0.13107764876632802,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999816545470513,
      "loss": 6.1738,
      "mean_token_accuracy": 0.1286979652941227,
      "num_tokens": 2996695.0,
      "step": 1445
    },
    {
      "entropy": 6.338008069992066,
      "epoch": 0.1315312046444122,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999812390404514,
      "loss": 6.1786,
      "mean_token_accuracy": 0.13488187193870543,
      "num_tokens": 3007500.0,
      "step": 1450
    },
    {
      "entropy": 6.171080303192139,
      "epoch": 0.13198476052249636,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999808188811823,
      "loss": 6.0426,
      "mean_token_accuracy": 0.13422525823116302,
      "num_tokens": 3017299.0,
      "step": 1455
    },
    {
      "entropy": 6.42709584236145,
      "epoch": 0.13243831640058054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999803940692524,
      "loss": 6.3248,
      "mean_token_accuracy": 0.12495478615164757,
      "num_tokens": 3028389.0,
      "step": 1460
    },
    {
      "entropy": 6.34055233001709,
      "epoch": 0.13289187227866472,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999799646046707,
      "loss": 6.1313,
      "mean_token_accuracy": 0.13551339507102966,
      "num_tokens": 3038883.0,
      "step": 1465
    },
    {
      "entropy": 6.360403776168823,
      "epoch": 0.1333454281567489,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999795304874461,
      "loss": 6.2634,
      "mean_token_accuracy": 0.13484074100852012,
      "num_tokens": 3049904.0,
      "step": 1470
    },
    {
      "entropy": 6.292270040512085,
      "epoch": 0.1337989840348331,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999790917175873,
      "loss": 6.0954,
      "mean_token_accuracy": 0.13213692530989646,
      "num_tokens": 3059788.0,
      "step": 1475
    },
    {
      "entropy": 6.332358312606812,
      "epoch": 0.13425253991291727,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999786482951036,
      "loss": 6.0914,
      "mean_token_accuracy": 0.13121994063258172,
      "num_tokens": 3070413.0,
      "step": 1480
    },
    {
      "entropy": 6.2656394958496096,
      "epoch": 0.13470609579100146,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999782002200041,
      "loss": 6.2047,
      "mean_token_accuracy": 0.1304186299443245,
      "num_tokens": 3081093.0,
      "step": 1485
    },
    {
      "entropy": 6.334182834625244,
      "epoch": 0.13515965166908564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999777474922982,
      "loss": 6.1612,
      "mean_token_accuracy": 0.12822958827018738,
      "num_tokens": 3090594.0,
      "step": 1490
    },
    {
      "entropy": 6.352655935287475,
      "epoch": 0.13561320754716982,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999772901119951,
      "loss": 6.2409,
      "mean_token_accuracy": 0.13285785242915155,
      "num_tokens": 3101061.0,
      "step": 1495
    },
    {
      "entropy": 6.28689432144165,
      "epoch": 0.13606676342525398,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999768280791043,
      "loss": 6.1627,
      "mean_token_accuracy": 0.13513913974165917,
      "num_tokens": 3111715.0,
      "step": 1500
    },
    {
      "entropy": 6.282320547103882,
      "epoch": 0.13652031930333816,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999763613936352,
      "loss": 6.0892,
      "mean_token_accuracy": 0.13638533875346184,
      "num_tokens": 3121378.0,
      "step": 1505
    },
    {
      "entropy": 6.252392864227295,
      "epoch": 0.13697387518142234,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999758900555979,
      "loss": 6.0183,
      "mean_token_accuracy": 0.13862146139144899,
      "num_tokens": 3131229.0,
      "step": 1510
    },
    {
      "entropy": 6.241233825683594,
      "epoch": 0.13742743105950653,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999754140650017,
      "loss": 6.1025,
      "mean_token_accuracy": 0.13484096452593802,
      "num_tokens": 3140859.0,
      "step": 1515
    },
    {
      "entropy": 6.310092401504517,
      "epoch": 0.1378809869375907,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999749334218567,
      "loss": 6.1048,
      "mean_token_accuracy": 0.13604142144322395,
      "num_tokens": 3151692.0,
      "step": 1520
    },
    {
      "entropy": 6.2527360916137695,
      "epoch": 0.1383345428156749,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999744481261726,
      "loss": 6.1801,
      "mean_token_accuracy": 0.13370710536837577,
      "num_tokens": 3161879.0,
      "step": 1525
    },
    {
      "entropy": 6.248764514923096,
      "epoch": 0.13878809869375908,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999739581779597,
      "loss": 6.0456,
      "mean_token_accuracy": 0.1446085900068283,
      "num_tokens": 3171282.0,
      "step": 1530
    },
    {
      "entropy": 6.270263290405273,
      "epoch": 0.13924165457184326,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499973463577228,
      "loss": 6.1068,
      "mean_token_accuracy": 0.1418723337352276,
      "num_tokens": 3181398.0,
      "step": 1535
    },
    {
      "entropy": 6.2894738674163815,
      "epoch": 0.13969521044992744,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999729643239877,
      "loss": 6.2273,
      "mean_token_accuracy": 0.1322890356183052,
      "num_tokens": 3191431.0,
      "step": 1540
    },
    {
      "entropy": 6.347840642929077,
      "epoch": 0.1401487663280116,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999724604182492,
      "loss": 6.2145,
      "mean_token_accuracy": 0.12926804050803184,
      "num_tokens": 3202147.0,
      "step": 1545
    },
    {
      "entropy": 6.323010873794556,
      "epoch": 0.14060232220609578,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999719518600229,
      "loss": 6.0864,
      "mean_token_accuracy": 0.13912193179130555,
      "num_tokens": 3212428.0,
      "step": 1550
    },
    {
      "entropy": 6.246621894836426,
      "epoch": 0.14105587808417996,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999714386493192,
      "loss": 6.1409,
      "mean_token_accuracy": 0.1369505174458027,
      "num_tokens": 3222099.0,
      "step": 1555
    },
    {
      "entropy": 6.316340351104737,
      "epoch": 0.14150943396226415,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999709207861489,
      "loss": 6.0828,
      "mean_token_accuracy": 0.13749494701623916,
      "num_tokens": 3232079.0,
      "step": 1560
    },
    {
      "entropy": 6.204212045669555,
      "epoch": 0.14196298984034833,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999703982705227,
      "loss": 6.005,
      "mean_token_accuracy": 0.14854954034090043,
      "num_tokens": 3241762.0,
      "step": 1565
    },
    {
      "entropy": 6.180720520019531,
      "epoch": 0.1424165457184325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999698711024513,
      "loss": 6.0869,
      "mean_token_accuracy": 0.13718333020806311,
      "num_tokens": 3252571.0,
      "step": 1570
    },
    {
      "entropy": 6.293680715560913,
      "epoch": 0.1428701015965167,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999693392819456,
      "loss": 6.1328,
      "mean_token_accuracy": 0.1329519271850586,
      "num_tokens": 3263542.0,
      "step": 1575
    },
    {
      "entropy": 6.287594938278199,
      "epoch": 0.14332365747460088,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999688028090166,
      "loss": 6.1936,
      "mean_token_accuracy": 0.13597823828458785,
      "num_tokens": 3273295.0,
      "step": 1580
    },
    {
      "entropy": 6.440543842315674,
      "epoch": 0.14377721335268506,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999682616836755,
      "loss": 6.2261,
      "mean_token_accuracy": 0.1362229973077774,
      "num_tokens": 3283605.0,
      "step": 1585
    },
    {
      "entropy": 6.197280740737915,
      "epoch": 0.14423076923076922,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999677159059334,
      "loss": 6.0532,
      "mean_token_accuracy": 0.14048462361097336,
      "num_tokens": 3293837.0,
      "step": 1590
    },
    {
      "entropy": 6.210299730300903,
      "epoch": 0.1446843251088534,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999671654758016,
      "loss": 6.0692,
      "mean_token_accuracy": 0.14433105364441873,
      "num_tokens": 3304131.0,
      "step": 1595
    },
    {
      "entropy": 6.238543367385864,
      "epoch": 0.14513788098693758,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999666103932915,
      "loss": 6.1103,
      "mean_token_accuracy": 0.13127385154366494,
      "num_tokens": 3313543.0,
      "step": 1600
    },
    {
      "entropy": 6.23192195892334,
      "epoch": 0.14559143686502177,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999660506584145,
      "loss": 6.0745,
      "mean_token_accuracy": 0.1385864198207855,
      "num_tokens": 3324487.0,
      "step": 1605
    },
    {
      "entropy": 6.296097135543823,
      "epoch": 0.14604499274310595,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999654862711823,
      "loss": 6.1001,
      "mean_token_accuracy": 0.13781747072935105,
      "num_tokens": 3335094.0,
      "step": 1610
    },
    {
      "entropy": 6.208214139938354,
      "epoch": 0.14649854862119013,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999649172316065,
      "loss": 6.04,
      "mean_token_accuracy": 0.13336168527603148,
      "num_tokens": 3344252.0,
      "step": 1615
    },
    {
      "entropy": 6.273281383514404,
      "epoch": 0.14695210449927432,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499964343539699,
      "loss": 6.1072,
      "mean_token_accuracy": 0.1359778583049774,
      "num_tokens": 3354265.0,
      "step": 1620
    },
    {
      "entropy": 6.312763309478759,
      "epoch": 0.1474056603773585,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999637651954714,
      "loss": 6.2133,
      "mean_token_accuracy": 0.13838177993893624,
      "num_tokens": 3364616.0,
      "step": 1625
    },
    {
      "entropy": 6.300515222549438,
      "epoch": 0.14785921625544268,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999631821989358,
      "loss": 6.1601,
      "mean_token_accuracy": 0.13766503930091858,
      "num_tokens": 3373650.0,
      "step": 1630
    },
    {
      "entropy": 6.221302127838134,
      "epoch": 0.14831277213352684,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999625945501043,
      "loss": 6.0573,
      "mean_token_accuracy": 0.1438056230545044,
      "num_tokens": 3384590.0,
      "step": 1635
    },
    {
      "entropy": 6.271721410751343,
      "epoch": 0.14876632801161102,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499962002248989,
      "loss": 6.0793,
      "mean_token_accuracy": 0.13807034716010094,
      "num_tokens": 3395616.0,
      "step": 1640
    },
    {
      "entropy": 6.2336437702178955,
      "epoch": 0.1492198838896952,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999614052956024,
      "loss": 6.0885,
      "mean_token_accuracy": 0.13720295056700707,
      "num_tokens": 3405810.0,
      "step": 1645
    },
    {
      "entropy": 6.2871723651885985,
      "epoch": 0.1496734397677794,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999608036899563,
      "loss": 6.1248,
      "mean_token_accuracy": 0.13411571234464645,
      "num_tokens": 3415932.0,
      "step": 1650
    },
    {
      "entropy": 6.227649116516114,
      "epoch": 0.15012699564586357,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999601974320636,
      "loss": 6.0773,
      "mean_token_accuracy": 0.13326597362756729,
      "num_tokens": 3425626.0,
      "step": 1655
    },
    {
      "entropy": 6.198207044601441,
      "epoch": 0.15058055152394775,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999595865219367,
      "loss": 6.098,
      "mean_token_accuracy": 0.13101402595639228,
      "num_tokens": 3434837.0,
      "step": 1660
    },
    {
      "entropy": 6.252737283706665,
      "epoch": 0.15103410740203194,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999589709595881,
      "loss": 6.0898,
      "mean_token_accuracy": 0.14350827634334565,
      "num_tokens": 3445238.0,
      "step": 1665
    },
    {
      "entropy": 6.187345504760742,
      "epoch": 0.15148766328011612,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999583507450308,
      "loss": 6.0692,
      "mean_token_accuracy": 0.13529428243637084,
      "num_tokens": 3456738.0,
      "step": 1670
    },
    {
      "entropy": 6.133850383758545,
      "epoch": 0.1519412191582003,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999577258782774,
      "loss": 5.9664,
      "mean_token_accuracy": 0.13891834691166877,
      "num_tokens": 3466296.0,
      "step": 1675
    },
    {
      "entropy": 6.328598546981811,
      "epoch": 0.15239477503628446,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499957096359341,
      "loss": 6.1502,
      "mean_token_accuracy": 0.13001871183514596,
      "num_tokens": 3476989.0,
      "step": 1680
    },
    {
      "entropy": 6.21191725730896,
      "epoch": 0.15284833091436864,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999564621882343,
      "loss": 5.9815,
      "mean_token_accuracy": 0.14194255620241164,
      "num_tokens": 3487050.0,
      "step": 1685
    },
    {
      "entropy": 6.256222677230835,
      "epoch": 0.15330188679245282,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999558233649708,
      "loss": 6.2052,
      "mean_token_accuracy": 0.1338474340736866,
      "num_tokens": 3497777.0,
      "step": 1690
    },
    {
      "entropy": 6.257563257217408,
      "epoch": 0.153755442670537,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999551798895636,
      "loss": 6.1012,
      "mean_token_accuracy": 0.13192513063549996,
      "num_tokens": 3509066.0,
      "step": 1695
    },
    {
      "entropy": 6.171849775314331,
      "epoch": 0.1542089985486212,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999545317620258,
      "loss": 6.0377,
      "mean_token_accuracy": 0.13509414941072465,
      "num_tokens": 3519567.0,
      "step": 1700
    },
    {
      "entropy": 6.180337810516358,
      "epoch": 0.15466255442670537,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999538789823711,
      "loss": 6.0677,
      "mean_token_accuracy": 0.1376472532749176,
      "num_tokens": 3529402.0,
      "step": 1705
    },
    {
      "entropy": 6.29026083946228,
      "epoch": 0.15511611030478956,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999532215506127,
      "loss": 6.0768,
      "mean_token_accuracy": 0.14445007145404815,
      "num_tokens": 3539805.0,
      "step": 1710
    },
    {
      "entropy": 6.146165990829468,
      "epoch": 0.15556966618287374,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999525594667645,
      "loss": 5.9557,
      "mean_token_accuracy": 0.1396943673491478,
      "num_tokens": 3550477.0,
      "step": 1715
    },
    {
      "entropy": 6.103784704208374,
      "epoch": 0.15602322206095792,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049995189273084,
      "loss": 6.0572,
      "mean_token_accuracy": 0.13744304925203324,
      "num_tokens": 3560729.0,
      "step": 1720
    },
    {
      "entropy": 6.271568822860718,
      "epoch": 0.15647677793904208,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499951221342853,
      "loss": 6.1321,
      "mean_token_accuracy": 0.1289873942732811,
      "num_tokens": 3571538.0,
      "step": 1725
    },
    {
      "entropy": 6.223130464553833,
      "epoch": 0.15693033381712626,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999505453028174,
      "loss": 6.1212,
      "mean_token_accuracy": 0.1351299576461315,
      "num_tokens": 3583128.0,
      "step": 1730
    },
    {
      "entropy": 6.263720464706421,
      "epoch": 0.15738388969521044,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999498646107472,
      "loss": 6.0628,
      "mean_token_accuracy": 0.13980068638920784,
      "num_tokens": 3594358.0,
      "step": 1735
    },
    {
      "entropy": 6.227974891662598,
      "epoch": 0.15783744557329463,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999491792666566,
      "loss": 5.9844,
      "mean_token_accuracy": 0.14432255774736405,
      "num_tokens": 3604345.0,
      "step": 1740
    },
    {
      "entropy": 6.122195100784301,
      "epoch": 0.1582910014513788,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999484892705597,
      "loss": 6.0287,
      "mean_token_accuracy": 0.13807731494307518,
      "num_tokens": 3616129.0,
      "step": 1745
    },
    {
      "entropy": 6.275023746490478,
      "epoch": 0.158744557329463,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999477946224705,
      "loss": 6.0836,
      "mean_token_accuracy": 0.14040916562080383,
      "num_tokens": 3627458.0,
      "step": 1750
    },
    {
      "entropy": 6.181599187850952,
      "epoch": 0.15919811320754718,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999470953224037,
      "loss": 5.9586,
      "mean_token_accuracy": 0.1387786313891411,
      "num_tokens": 3636047.0,
      "step": 1755
    },
    {
      "entropy": 6.1431786060333256,
      "epoch": 0.15965166908563136,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999463913703735,
      "loss": 5.9542,
      "mean_token_accuracy": 0.13981650844216348,
      "num_tokens": 3646930.0,
      "step": 1760
    },
    {
      "entropy": 6.141822052001953,
      "epoch": 0.16010522496371554,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999456827663948,
      "loss": 6.0495,
      "mean_token_accuracy": 0.1432150684297085,
      "num_tokens": 3657279.0,
      "step": 1765
    },
    {
      "entropy": 6.173039388656616,
      "epoch": 0.1605587808417997,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499944969510482,
      "loss": 5.9656,
      "mean_token_accuracy": 0.141604046523571,
      "num_tokens": 3667070.0,
      "step": 1770
    },
    {
      "entropy": 6.148836374282837,
      "epoch": 0.16101233671988388,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999442516026498,
      "loss": 5.9909,
      "mean_token_accuracy": 0.13877027854323387,
      "num_tokens": 3677560.0,
      "step": 1775
    },
    {
      "entropy": 6.135176277160644,
      "epoch": 0.16146589259796806,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999435290429133,
      "loss": 5.951,
      "mean_token_accuracy": 0.14585819393396376,
      "num_tokens": 3686701.0,
      "step": 1780
    },
    {
      "entropy": 6.194845485687256,
      "epoch": 0.16191944847605225,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999428018312872,
      "loss": 6.0453,
      "mean_token_accuracy": 0.14197150394320487,
      "num_tokens": 3696338.0,
      "step": 1785
    },
    {
      "entropy": 6.213848018646241,
      "epoch": 0.16237300435413643,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999420699677867,
      "loss": 6.0178,
      "mean_token_accuracy": 0.13948137164115906,
      "num_tokens": 3705752.0,
      "step": 1790
    },
    {
      "entropy": 6.173350286483765,
      "epoch": 0.1628265602322206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999413334524269,
      "loss": 6.1086,
      "mean_token_accuracy": 0.13366367369890214,
      "num_tokens": 3716225.0,
      "step": 1795
    },
    {
      "entropy": 6.1766557693481445,
      "epoch": 0.1632801161103048,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999405922852229,
      "loss": 5.9812,
      "mean_token_accuracy": 0.1403820961713791,
      "num_tokens": 3726593.0,
      "step": 1800
    },
    {
      "entropy": 6.1498106002807615,
      "epoch": 0.16373367198838898,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999398464661901,
      "loss": 6.0292,
      "mean_token_accuracy": 0.13524944633245467,
      "num_tokens": 3737563.0,
      "step": 1805
    },
    {
      "entropy": 6.278463888168335,
      "epoch": 0.16418722786647316,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999390959953441,
      "loss": 6.0868,
      "mean_token_accuracy": 0.13811569064855575,
      "num_tokens": 3747752.0,
      "step": 1810
    },
    {
      "entropy": 6.124238109588623,
      "epoch": 0.16464078374455732,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999383408727002,
      "loss": 6.0109,
      "mean_token_accuracy": 0.13542224168777467,
      "num_tokens": 3758795.0,
      "step": 1815
    },
    {
      "entropy": 6.21006441116333,
      "epoch": 0.1650943396226415,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999375810982741,
      "loss": 6.0051,
      "mean_token_accuracy": 0.14309126287698745,
      "num_tokens": 3768690.0,
      "step": 1820
    },
    {
      "entropy": 6.147620582580567,
      "epoch": 0.16554789550072568,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999368166720815,
      "loss": 6.0588,
      "mean_token_accuracy": 0.137413576990366,
      "num_tokens": 3779018.0,
      "step": 1825
    },
    {
      "entropy": 6.241029739379883,
      "epoch": 0.16600145137880987,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999360475941382,
      "loss": 5.9743,
      "mean_token_accuracy": 0.14609253406524658,
      "num_tokens": 3788985.0,
      "step": 1830
    },
    {
      "entropy": 6.181500005722046,
      "epoch": 0.16645500725689405,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049993527386446,
      "loss": 6.0916,
      "mean_token_accuracy": 0.13334065973758696,
      "num_tokens": 3800787.0,
      "step": 1835
    },
    {
      "entropy": 6.164537858963013,
      "epoch": 0.16690856313497823,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999344954830633,
      "loss": 5.9801,
      "mean_token_accuracy": 0.1378617025911808,
      "num_tokens": 3811381.0,
      "step": 1840
    },
    {
      "entropy": 6.188645315170288,
      "epoch": 0.16736211901306242,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999337124499638,
      "loss": 6.0844,
      "mean_token_accuracy": 0.14289560168981552,
      "num_tokens": 3822055.0,
      "step": 1845
    },
    {
      "entropy": 6.152506732940674,
      "epoch": 0.1678156748911466,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999329247651777,
      "loss": 6.0558,
      "mean_token_accuracy": 0.13555024489760398,
      "num_tokens": 3831900.0,
      "step": 1850
    },
    {
      "entropy": 6.184814882278443,
      "epoch": 0.16826923076923078,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999321324287213,
      "loss": 6.0046,
      "mean_token_accuracy": 0.1400987111032009,
      "num_tokens": 3842702.0,
      "step": 1855
    },
    {
      "entropy": 6.251664876937866,
      "epoch": 0.16872278664731494,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999313354406114,
      "loss": 5.986,
      "mean_token_accuracy": 0.1382150486111641,
      "num_tokens": 3852690.0,
      "step": 1860
    },
    {
      "entropy": 6.014293146133423,
      "epoch": 0.16917634252539912,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499930533800864,
      "loss": 5.9598,
      "mean_token_accuracy": 0.15227705091238022,
      "num_tokens": 3863040.0,
      "step": 1865
    },
    {
      "entropy": 6.209018230438232,
      "epoch": 0.1696298984034833,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999297275094958,
      "loss": 6.065,
      "mean_token_accuracy": 0.13639552444219588,
      "num_tokens": 3873613.0,
      "step": 1870
    },
    {
      "entropy": 6.046637344360351,
      "epoch": 0.1700834542815675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999289165665235,
      "loss": 5.8751,
      "mean_token_accuracy": 0.14939435720443725,
      "num_tokens": 3883936.0,
      "step": 1875
    },
    {
      "entropy": 6.090919876098633,
      "epoch": 0.17053701015965167,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999281009719639,
      "loss": 5.8273,
      "mean_token_accuracy": 0.15036647394299507,
      "num_tokens": 3893788.0,
      "step": 1880
    },
    {
      "entropy": 6.0542162418365475,
      "epoch": 0.17099056603773585,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999272807258339,
      "loss": 5.912,
      "mean_token_accuracy": 0.14638229012489318,
      "num_tokens": 3903434.0,
      "step": 1885
    },
    {
      "entropy": 6.142886686325073,
      "epoch": 0.17144412191582004,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999264558281504,
      "loss": 5.9101,
      "mean_token_accuracy": 0.14875417798757554,
      "num_tokens": 3913476.0,
      "step": 1890
    },
    {
      "entropy": 5.993232488632202,
      "epoch": 0.17189767779390422,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999256262789305,
      "loss": 5.8601,
      "mean_token_accuracy": 0.14939367920160293,
      "num_tokens": 3923301.0,
      "step": 1895
    },
    {
      "entropy": 6.073164653778076,
      "epoch": 0.1723512336719884,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999247920781913,
      "loss": 6.0334,
      "mean_token_accuracy": 0.13975203409790993,
      "num_tokens": 3933335.0,
      "step": 1900
    },
    {
      "entropy": 6.233634042739868,
      "epoch": 0.17280478955007256,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999239532259502,
      "loss": 6.037,
      "mean_token_accuracy": 0.13394607827067376,
      "num_tokens": 3944171.0,
      "step": 1905
    },
    {
      "entropy": 6.214393186569214,
      "epoch": 0.17325834542815674,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999231097222244,
      "loss": 6.1071,
      "mean_token_accuracy": 0.14265160858631135,
      "num_tokens": 3954160.0,
      "step": 1910
    },
    {
      "entropy": 6.173836088180542,
      "epoch": 0.17371190130624092,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999222615670312,
      "loss": 5.9094,
      "mean_token_accuracy": 0.14953405708074569,
      "num_tokens": 3964110.0,
      "step": 1915
    },
    {
      "entropy": 6.1780534267425535,
      "epoch": 0.1741654571843251,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999214087603885,
      "loss": 6.0034,
      "mean_token_accuracy": 0.14108952656388282,
      "num_tokens": 3974293.0,
      "step": 1920
    },
    {
      "entropy": 6.1007123470306395,
      "epoch": 0.1746190130624093,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999205513023135,
      "loss": 5.9491,
      "mean_token_accuracy": 0.1505696728825569,
      "num_tokens": 3984346.0,
      "step": 1925
    },
    {
      "entropy": 6.014599132537842,
      "epoch": 0.17507256894049347,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999196891928245,
      "loss": 5.8944,
      "mean_token_accuracy": 0.1506063535809517,
      "num_tokens": 3992897.0,
      "step": 1930
    },
    {
      "entropy": 6.1384530544281,
      "epoch": 0.17552612481857766,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999188224319388,
      "loss": 5.929,
      "mean_token_accuracy": 0.1495840311050415,
      "num_tokens": 4002575.0,
      "step": 1935
    },
    {
      "entropy": 6.0449896335601805,
      "epoch": 0.17597968069666184,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999179510196746,
      "loss": 6.0023,
      "mean_token_accuracy": 0.1404266342520714,
      "num_tokens": 4013752.0,
      "step": 1940
    },
    {
      "entropy": 6.181995582580567,
      "epoch": 0.17643323657474602,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999170749560498,
      "loss": 5.8891,
      "mean_token_accuracy": 0.14904519468545913,
      "num_tokens": 4023718.0,
      "step": 1945
    },
    {
      "entropy": 6.080067110061646,
      "epoch": 0.17688679245283018,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999161942410827,
      "loss": 5.9877,
      "mean_token_accuracy": 0.13862392008304597,
      "num_tokens": 4034210.0,
      "step": 1950
    },
    {
      "entropy": 6.194681644439697,
      "epoch": 0.17734034833091436,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999153088747913,
      "loss": 5.9752,
      "mean_token_accuracy": 0.14118978679180144,
      "num_tokens": 4044868.0,
      "step": 1955
    },
    {
      "entropy": 6.018691873550415,
      "epoch": 0.17779390420899854,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499914418857194,
      "loss": 5.9759,
      "mean_token_accuracy": 0.14467088282108306,
      "num_tokens": 4055049.0,
      "step": 1960
    },
    {
      "entropy": 6.195721387863159,
      "epoch": 0.17824746008708273,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999135241883092,
      "loss": 6.0328,
      "mean_token_accuracy": 0.13953902348876,
      "num_tokens": 4065889.0,
      "step": 1965
    },
    {
      "entropy": 6.124079751968384,
      "epoch": 0.1787010159651669,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999126248681555,
      "loss": 6.0438,
      "mean_token_accuracy": 0.13967278376221656,
      "num_tokens": 4077288.0,
      "step": 1970
    },
    {
      "entropy": 6.1740796089172365,
      "epoch": 0.1791545718432511,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999117208967513,
      "loss": 5.9939,
      "mean_token_accuracy": 0.14228827133774757,
      "num_tokens": 4087402.0,
      "step": 1975
    },
    {
      "entropy": 6.136812305450439,
      "epoch": 0.17960812772133528,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999108122741154,
      "loss": 6.0412,
      "mean_token_accuracy": 0.13958762884140014,
      "num_tokens": 4098150.0,
      "step": 1980
    },
    {
      "entropy": 6.085280323028565,
      "epoch": 0.18006168359941946,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999098990002666,
      "loss": 5.9088,
      "mean_token_accuracy": 0.14206465110182762,
      "num_tokens": 4108134.0,
      "step": 1985
    },
    {
      "entropy": 6.158121109008789,
      "epoch": 0.18051523947750364,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999089810752237,
      "loss": 6.0448,
      "mean_token_accuracy": 0.13666467741131783,
      "num_tokens": 4118251.0,
      "step": 1990
    },
    {
      "entropy": 6.078621101379395,
      "epoch": 0.1809687953555878,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499908058499006,
      "loss": 5.8876,
      "mean_token_accuracy": 0.14455655217170715,
      "num_tokens": 4129228.0,
      "step": 1995
    },
    {
      "entropy": 6.047148942947388,
      "epoch": 0.18142235123367198,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999071312716321,
      "loss": 5.9426,
      "mean_token_accuracy": 0.15172967463731765,
      "num_tokens": 4139391.0,
      "step": 2000
    },
    {
      "entropy": 6.093998432159424,
      "epoch": 0.18187590711175616,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999061993931215,
      "loss": 5.88,
      "mean_token_accuracy": 0.15255183577537537,
      "num_tokens": 4148891.0,
      "step": 2005
    },
    {
      "entropy": 6.305850887298584,
      "epoch": 0.18232946298984035,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999052628634934,
      "loss": 6.2083,
      "mean_token_accuracy": 0.13769319280982018,
      "num_tokens": 4159551.0,
      "step": 2010
    },
    {
      "entropy": 6.18409833908081,
      "epoch": 0.18278301886792453,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999043216827671,
      "loss": 6.0219,
      "mean_token_accuracy": 0.1419105350971222,
      "num_tokens": 4170055.0,
      "step": 2015
    },
    {
      "entropy": 6.019263553619385,
      "epoch": 0.1832365747460087,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999033758509621,
      "loss": 5.8725,
      "mean_token_accuracy": 0.142510387301445,
      "num_tokens": 4180767.0,
      "step": 2020
    },
    {
      "entropy": 6.1171001434326175,
      "epoch": 0.1836901306240929,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499902425368098,
      "loss": 5.99,
      "mean_token_accuracy": 0.1484679900109768,
      "num_tokens": 4190654.0,
      "step": 2025
    },
    {
      "entropy": 6.084097576141358,
      "epoch": 0.18414368650217708,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999014702341944,
      "loss": 5.9217,
      "mean_token_accuracy": 0.14448753595352173,
      "num_tokens": 4200753.0,
      "step": 2030
    },
    {
      "entropy": 6.135722398757935,
      "epoch": 0.18459724238026126,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999005104492712,
      "loss": 5.9985,
      "mean_token_accuracy": 0.14054833576083184,
      "num_tokens": 4210994.0,
      "step": 2035
    },
    {
      "entropy": 6.120487928390503,
      "epoch": 0.18505079825834542,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998995460133481,
      "loss": 5.9412,
      "mean_token_accuracy": 0.14108474180102348,
      "num_tokens": 4220883.0,
      "step": 2040
    },
    {
      "entropy": 6.127297973632812,
      "epoch": 0.1855043541364296,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998985769264451,
      "loss": 5.977,
      "mean_token_accuracy": 0.13780092522501947,
      "num_tokens": 4231269.0,
      "step": 2045
    },
    {
      "entropy": 6.101497316360474,
      "epoch": 0.18595791001451378,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998976031885821,
      "loss": 5.9776,
      "mean_token_accuracy": 0.1418766401708126,
      "num_tokens": 4241559.0,
      "step": 2050
    },
    {
      "entropy": 6.121247720718384,
      "epoch": 0.18641146589259797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998966247997794,
      "loss": 5.8733,
      "mean_token_accuracy": 0.14405595809221267,
      "num_tokens": 4250870.0,
      "step": 2055
    },
    {
      "entropy": 6.015826034545898,
      "epoch": 0.18686502177068215,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998956417600572,
      "loss": 5.8246,
      "mean_token_accuracy": 0.15058073103427888,
      "num_tokens": 4260848.0,
      "step": 2060
    },
    {
      "entropy": 6.08300142288208,
      "epoch": 0.18731857764876633,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998946540694358,
      "loss": 5.9464,
      "mean_token_accuracy": 0.1485067129135132,
      "num_tokens": 4270973.0,
      "step": 2065
    },
    {
      "entropy": 6.14107756614685,
      "epoch": 0.18777213352685052,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998936617279357,
      "loss": 6.0361,
      "mean_token_accuracy": 0.14370361119508743,
      "num_tokens": 4281211.0,
      "step": 2070
    },
    {
      "entropy": 6.045505952835083,
      "epoch": 0.1882256894049347,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998926647355773,
      "loss": 5.8954,
      "mean_token_accuracy": 0.14735701233148574,
      "num_tokens": 4291425.0,
      "step": 2075
    },
    {
      "entropy": 6.1746431350708,
      "epoch": 0.18867924528301888,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998916630923813,
      "loss": 5.9993,
      "mean_token_accuracy": 0.14234295561909677,
      "num_tokens": 4301079.0,
      "step": 2080
    },
    {
      "entropy": 6.0917683124542235,
      "epoch": 0.18913280116110304,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998906567983685,
      "loss": 5.9599,
      "mean_token_accuracy": 0.14310015216469765,
      "num_tokens": 4312374.0,
      "step": 2085
    },
    {
      "entropy": 6.10637731552124,
      "epoch": 0.18958635703918722,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998896458535595,
      "loss": 5.8884,
      "mean_token_accuracy": 0.15468773394823074,
      "num_tokens": 4322810.0,
      "step": 2090
    },
    {
      "entropy": 6.0409440994262695,
      "epoch": 0.1900399129172714,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998886302579752,
      "loss": 5.888,
      "mean_token_accuracy": 0.14332466721534728,
      "num_tokens": 4332082.0,
      "step": 2095
    },
    {
      "entropy": 6.115694570541382,
      "epoch": 0.19049346879535559,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499887610011637,
      "loss": 5.9771,
      "mean_token_accuracy": 0.14411457404494285,
      "num_tokens": 4342632.0,
      "step": 2100
    },
    {
      "entropy": 6.025098752975464,
      "epoch": 0.19094702467343977,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998865851145656,
      "loss": 5.9079,
      "mean_token_accuracy": 0.14562978297472,
      "num_tokens": 4353544.0,
      "step": 2105
    },
    {
      "entropy": 6.080524492263794,
      "epoch": 0.19140058055152395,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998855555667822,
      "loss": 6.0337,
      "mean_token_accuracy": 0.13668391555547715,
      "num_tokens": 4363636.0,
      "step": 2110
    },
    {
      "entropy": 6.174575281143189,
      "epoch": 0.19185413642960814,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998845213683084,
      "loss": 5.91,
      "mean_token_accuracy": 0.14217197224497796,
      "num_tokens": 4373672.0,
      "step": 2115
    },
    {
      "entropy": 5.952403831481933,
      "epoch": 0.19230769230769232,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998834825191652,
      "loss": 5.8609,
      "mean_token_accuracy": 0.14928537905216216,
      "num_tokens": 4383199.0,
      "step": 2120
    },
    {
      "entropy": 6.0944489479064945,
      "epoch": 0.1927612481857765,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998824390193743,
      "loss": 5.9295,
      "mean_token_accuracy": 0.14216511845588684,
      "num_tokens": 4394446.0,
      "step": 2125
    },
    {
      "entropy": 6.156260538101196,
      "epoch": 0.19321480406386066,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998813908689573,
      "loss": 5.9814,
      "mean_token_accuracy": 0.14446035400032997,
      "num_tokens": 4405882.0,
      "step": 2130
    },
    {
      "entropy": 6.021085166931153,
      "epoch": 0.19366835994194484,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998803380679359,
      "loss": 5.8395,
      "mean_token_accuracy": 0.15445481687784196,
      "num_tokens": 4415280.0,
      "step": 2135
    },
    {
      "entropy": 6.026923942565918,
      "epoch": 0.19412191582002902,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998792806163317,
      "loss": 6.0418,
      "mean_token_accuracy": 0.1423596292734146,
      "num_tokens": 4425755.0,
      "step": 2140
    },
    {
      "entropy": 6.0691611766815186,
      "epoch": 0.1945754716981132,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998782185141667,
      "loss": 5.899,
      "mean_token_accuracy": 0.1436906062066555,
      "num_tokens": 4436259.0,
      "step": 2145
    },
    {
      "entropy": 6.100642156600952,
      "epoch": 0.1950290275761974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998771517614629,
      "loss": 5.9354,
      "mean_token_accuracy": 0.14032572209835054,
      "num_tokens": 4446423.0,
      "step": 2150
    },
    {
      "entropy": 6.079388666152954,
      "epoch": 0.19548258345428157,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998760803582422,
      "loss": 5.8765,
      "mean_token_accuracy": 0.15558486431837082,
      "num_tokens": 4456678.0,
      "step": 2155
    },
    {
      "entropy": 6.038434076309204,
      "epoch": 0.19593613933236576,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998750043045269,
      "loss": 5.88,
      "mean_token_accuracy": 0.1474220186471939,
      "num_tokens": 4467031.0,
      "step": 2160
    },
    {
      "entropy": 6.084740686416626,
      "epoch": 0.19638969521044994,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998739236003392,
      "loss": 5.9457,
      "mean_token_accuracy": 0.13701039925217628,
      "num_tokens": 4477771.0,
      "step": 2165
    },
    {
      "entropy": 6.058000755310059,
      "epoch": 0.1968432510885341,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998728382457014,
      "loss": 5.9724,
      "mean_token_accuracy": 0.14374302327632904,
      "num_tokens": 4488426.0,
      "step": 2170
    },
    {
      "entropy": 6.157823610305786,
      "epoch": 0.19729680696661828,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998717482406362,
      "loss": 5.9512,
      "mean_token_accuracy": 0.14008950740098952,
      "num_tokens": 4499735.0,
      "step": 2175
    },
    {
      "entropy": 6.083208417892456,
      "epoch": 0.19775036284470246,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998706535851657,
      "loss": 5.9067,
      "mean_token_accuracy": 0.1461469426751137,
      "num_tokens": 4509500.0,
      "step": 2180
    },
    {
      "entropy": 6.037715578079224,
      "epoch": 0.19820391872278664,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998695542793129,
      "loss": 5.9079,
      "mean_token_accuracy": 0.14525417983531952,
      "num_tokens": 4520431.0,
      "step": 2185
    },
    {
      "entropy": 6.084038925170899,
      "epoch": 0.19865747460087083,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998684503231003,
      "loss": 5.9846,
      "mean_token_accuracy": 0.14418850764632224,
      "num_tokens": 4532294.0,
      "step": 2190
    },
    {
      "entropy": 6.027761554718017,
      "epoch": 0.199111030478955,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499867341716551,
      "loss": 5.8985,
      "mean_token_accuracy": 0.1443702645599842,
      "num_tokens": 4542488.0,
      "step": 2195
    },
    {
      "entropy": 6.043829488754272,
      "epoch": 0.1995645863570392,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998662284596878,
      "loss": 5.8528,
      "mean_token_accuracy": 0.1418704494833946,
      "num_tokens": 4552933.0,
      "step": 2200
    },
    {
      "entropy": 6.060906505584716,
      "epoch": 0.20001814223512338,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998651105525335,
      "loss": 5.8336,
      "mean_token_accuracy": 0.15473553836345671,
      "num_tokens": 4562373.0,
      "step": 2205
    },
    {
      "entropy": 5.9967522621154785,
      "epoch": 0.20047169811320756,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998639879951115,
      "loss": 5.9557,
      "mean_token_accuracy": 0.14503294080495835,
      "num_tokens": 4572715.0,
      "step": 2210
    },
    {
      "entropy": 6.131484937667847,
      "epoch": 0.2009252539912917,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499862860787445,
      "loss": 5.8461,
      "mean_token_accuracy": 0.1420556351542473,
      "num_tokens": 4582782.0,
      "step": 2215
    },
    {
      "entropy": 6.036686944961548,
      "epoch": 0.2013788098693759,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998617289295571,
      "loss": 5.9439,
      "mean_token_accuracy": 0.14234896376729012,
      "num_tokens": 4594108.0,
      "step": 2220
    },
    {
      "entropy": 6.102933645248413,
      "epoch": 0.20183236574746008,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998605924214714,
      "loss": 5.9103,
      "mean_token_accuracy": 0.14780050218105317,
      "num_tokens": 4604732.0,
      "step": 2225
    },
    {
      "entropy": 6.049504566192627,
      "epoch": 0.20228592162554426,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998594512632114,
      "loss": 6.0456,
      "mean_token_accuracy": 0.13591379299759865,
      "num_tokens": 4614998.0,
      "step": 2230
    },
    {
      "entropy": 6.057260751724243,
      "epoch": 0.20273947750362845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998583054548006,
      "loss": 5.8648,
      "mean_token_accuracy": 0.14422834739089013,
      "num_tokens": 4624665.0,
      "step": 2235
    },
    {
      "entropy": 6.085105228424072,
      "epoch": 0.20319303338171263,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998571549962628,
      "loss": 5.8612,
      "mean_token_accuracy": 0.15319595485925674,
      "num_tokens": 4633731.0,
      "step": 2240
    },
    {
      "entropy": 6.046474456787109,
      "epoch": 0.2036465892597968,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998559998876216,
      "loss": 5.8897,
      "mean_token_accuracy": 0.14516430646181105,
      "num_tokens": 4643421.0,
      "step": 2245
    },
    {
      "entropy": 6.105594730377197,
      "epoch": 0.204100145137881,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998548401289012,
      "loss": 5.9729,
      "mean_token_accuracy": 0.14359490424394608,
      "num_tokens": 4653795.0,
      "step": 2250
    },
    {
      "entropy": 6.111778020858765,
      "epoch": 0.20455370101596518,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998536757201253,
      "loss": 5.9528,
      "mean_token_accuracy": 0.14500485807657243,
      "num_tokens": 4664840.0,
      "step": 2255
    },
    {
      "entropy": 6.058202457427979,
      "epoch": 0.20500725689404933,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998525066613181,
      "loss": 5.9537,
      "mean_token_accuracy": 0.14941285997629167,
      "num_tokens": 4675508.0,
      "step": 2260
    },
    {
      "entropy": 6.049920320510864,
      "epoch": 0.20546081277213352,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998513329525038,
      "loss": 5.8734,
      "mean_token_accuracy": 0.15032386034727097,
      "num_tokens": 4685995.0,
      "step": 2265
    },
    {
      "entropy": 6.082958173751831,
      "epoch": 0.2059143686502177,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998501545937066,
      "loss": 5.9052,
      "mean_token_accuracy": 0.14433911442756653,
      "num_tokens": 4696659.0,
      "step": 2270
    },
    {
      "entropy": 6.0467150688171385,
      "epoch": 0.20636792452830188,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499848971584951,
      "loss": 5.8595,
      "mean_token_accuracy": 0.15614939630031585,
      "num_tokens": 4706438.0,
      "step": 2275
    },
    {
      "entropy": 5.930180215835572,
      "epoch": 0.20682148040638607,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998477839262611,
      "loss": 5.8678,
      "mean_token_accuracy": 0.1469499036669731,
      "num_tokens": 4716402.0,
      "step": 2280
    },
    {
      "entropy": 6.050282192230225,
      "epoch": 0.20727503628447025,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499846591617662,
      "loss": 5.9457,
      "mean_token_accuracy": 0.1466635584831238,
      "num_tokens": 4728488.0,
      "step": 2285
    },
    {
      "entropy": 6.057934761047363,
      "epoch": 0.20772859216255443,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998453946591779,
      "loss": 5.976,
      "mean_token_accuracy": 0.14741047471761703,
      "num_tokens": 4738693.0,
      "step": 2290
    },
    {
      "entropy": 6.034505462646484,
      "epoch": 0.20818214804063861,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998441930508339,
      "loss": 5.8617,
      "mean_token_accuracy": 0.149900358915329,
      "num_tokens": 4748895.0,
      "step": 2295
    },
    {
      "entropy": 6.12192120552063,
      "epoch": 0.2086357039187228,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998429867926547,
      "loss": 5.9731,
      "mean_token_accuracy": 0.1417195826768875,
      "num_tokens": 4760135.0,
      "step": 2300
    },
    {
      "entropy": 6.1114363193511965,
      "epoch": 0.20908925979680695,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998417758846651,
      "loss": 6.0106,
      "mean_token_accuracy": 0.1492314174771309,
      "num_tokens": 4769935.0,
      "step": 2305
    },
    {
      "entropy": 6.163050842285156,
      "epoch": 0.20954281567489114,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998405603268903,
      "loss": 5.9454,
      "mean_token_accuracy": 0.14845321476459503,
      "num_tokens": 4779520.0,
      "step": 2310
    },
    {
      "entropy": 5.973990440368652,
      "epoch": 0.20999637155297532,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998393401193554,
      "loss": 5.8416,
      "mean_token_accuracy": 0.15259672403335572,
      "num_tokens": 4791256.0,
      "step": 2315
    },
    {
      "entropy": 5.966838359832764,
      "epoch": 0.2104499274310595,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998381152620857,
      "loss": 5.811,
      "mean_token_accuracy": 0.15300733894109725,
      "num_tokens": 4801334.0,
      "step": 2320
    },
    {
      "entropy": 5.96552734375,
      "epoch": 0.21090348330914369,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998368857551067,
      "loss": 5.7882,
      "mean_token_accuracy": 0.15469710677862167,
      "num_tokens": 4811361.0,
      "step": 2325
    },
    {
      "entropy": 6.016611623764038,
      "epoch": 0.21135703918722787,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998356515984433,
      "loss": 5.85,
      "mean_token_accuracy": 0.1501396507024765,
      "num_tokens": 4821190.0,
      "step": 2330
    },
    {
      "entropy": 6.041166496276856,
      "epoch": 0.21181059506531205,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998344127921213,
      "loss": 5.8821,
      "mean_token_accuracy": 0.1504140943288803,
      "num_tokens": 4830694.0,
      "step": 2335
    },
    {
      "entropy": 5.997649621963501,
      "epoch": 0.21226415094339623,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998331693361666,
      "loss": 5.8687,
      "mean_token_accuracy": 0.1483940526843071,
      "num_tokens": 4841344.0,
      "step": 2340
    },
    {
      "entropy": 6.003773546218872,
      "epoch": 0.21271770682148042,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998319212306046,
      "loss": 5.8398,
      "mean_token_accuracy": 0.14764055758714675,
      "num_tokens": 4851161.0,
      "step": 2345
    },
    {
      "entropy": 5.979840612411499,
      "epoch": 0.21317126269956457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499830668475461,
      "loss": 5.831,
      "mean_token_accuracy": 0.1507522940635681,
      "num_tokens": 4860740.0,
      "step": 2350
    },
    {
      "entropy": 6.029822587966919,
      "epoch": 0.21362481857764876,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499829411070762,
      "loss": 5.856,
      "mean_token_accuracy": 0.147916804254055,
      "num_tokens": 4871733.0,
      "step": 2355
    },
    {
      "entropy": 5.999858570098877,
      "epoch": 0.21407837445573294,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998281490165335,
      "loss": 5.9117,
      "mean_token_accuracy": 0.14421676099300385,
      "num_tokens": 4883617.0,
      "step": 2360
    },
    {
      "entropy": 6.043055486679077,
      "epoch": 0.21453193033381712,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998268823128016,
      "loss": 5.8607,
      "mean_token_accuracy": 0.14836216866970062,
      "num_tokens": 4894519.0,
      "step": 2365
    },
    {
      "entropy": 6.004064750671387,
      "epoch": 0.2149854862119013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998256109595923,
      "loss": 5.8134,
      "mean_token_accuracy": 0.14691278859972953,
      "num_tokens": 4904965.0,
      "step": 2370
    },
    {
      "entropy": 5.955627918243408,
      "epoch": 0.2154390420899855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998243349569323,
      "loss": 5.7279,
      "mean_token_accuracy": 0.157587631046772,
      "num_tokens": 4914618.0,
      "step": 2375
    },
    {
      "entropy": 5.92179274559021,
      "epoch": 0.21589259796806967,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998230543048477,
      "loss": 5.7896,
      "mean_token_accuracy": 0.15410214364528657,
      "num_tokens": 4924341.0,
      "step": 2380
    },
    {
      "entropy": 5.918932485580444,
      "epoch": 0.21634615384615385,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499821769003365,
      "loss": 5.7633,
      "mean_token_accuracy": 0.15452138409018518,
      "num_tokens": 4934231.0,
      "step": 2385
    },
    {
      "entropy": 5.887126874923706,
      "epoch": 0.21679970972423804,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998204790525109,
      "loss": 5.7905,
      "mean_token_accuracy": 0.1511450305581093,
      "num_tokens": 4944303.0,
      "step": 2390
    },
    {
      "entropy": 6.001142930984497,
      "epoch": 0.2172532656023222,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998191844523119,
      "loss": 5.8693,
      "mean_token_accuracy": 0.15268588662147523,
      "num_tokens": 4954571.0,
      "step": 2395
    },
    {
      "entropy": 5.987411212921143,
      "epoch": 0.21770682148040638,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499817885202795,
      "loss": 5.848,
      "mean_token_accuracy": 0.14666955098509787,
      "num_tokens": 4964418.0,
      "step": 2400
    },
    {
      "entropy": 6.03399133682251,
      "epoch": 0.21816037735849056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998165813039869,
      "loss": 5.8291,
      "mean_token_accuracy": 0.1442165307700634,
      "num_tokens": 4975036.0,
      "step": 2405
    },
    {
      "entropy": 5.994223356246948,
      "epoch": 0.21861393323657474,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998152727559145,
      "loss": 5.9302,
      "mean_token_accuracy": 0.14429473355412484,
      "num_tokens": 4985272.0,
      "step": 2410
    },
    {
      "entropy": 6.091732931137085,
      "epoch": 0.21906748911465893,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499813959558605,
      "loss": 5.8548,
      "mean_token_accuracy": 0.15049488544464112,
      "num_tokens": 4995187.0,
      "step": 2415
    },
    {
      "entropy": 6.026802444458008,
      "epoch": 0.2195210449927431,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998126417120856,
      "loss": 5.9175,
      "mean_token_accuracy": 0.15209949463605882,
      "num_tokens": 5005013.0,
      "step": 2420
    },
    {
      "entropy": 6.094258975982666,
      "epoch": 0.2199746008708273,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998113192163835,
      "loss": 5.9259,
      "mean_token_accuracy": 0.14370185285806655,
      "num_tokens": 5016789.0,
      "step": 2425
    },
    {
      "entropy": 5.985634708404541,
      "epoch": 0.22042815674891147,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499809992071526,
      "loss": 5.8749,
      "mean_token_accuracy": 0.15012458711862564,
      "num_tokens": 5026913.0,
      "step": 2430
    },
    {
      "entropy": 6.017763757705689,
      "epoch": 0.22088171262699566,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998086602775406,
      "loss": 5.9623,
      "mean_token_accuracy": 0.1502877563238144,
      "num_tokens": 5037501.0,
      "step": 2435
    },
    {
      "entropy": 5.990037059783935,
      "epoch": 0.2213352685050798,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998073238344547,
      "loss": 5.8105,
      "mean_token_accuracy": 0.144505974650383,
      "num_tokens": 5048478.0,
      "step": 2440
    },
    {
      "entropy": 6.0197502136230465,
      "epoch": 0.221788824383164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998059827422962,
      "loss": 5.8716,
      "mean_token_accuracy": 0.13962874338030815,
      "num_tokens": 5059118.0,
      "step": 2445
    },
    {
      "entropy": 6.010491847991943,
      "epoch": 0.22224238026124818,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998046370010926,
      "loss": 5.8321,
      "mean_token_accuracy": 0.15523343831300734,
      "num_tokens": 5070180.0,
      "step": 2450
    },
    {
      "entropy": 6.002876043319702,
      "epoch": 0.22269593613933236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998032866108718,
      "loss": 5.8404,
      "mean_token_accuracy": 0.15497579723596572,
      "num_tokens": 5079912.0,
      "step": 2455
    },
    {
      "entropy": 6.074828386306763,
      "epoch": 0.22314949201741655,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998019315716618,
      "loss": 5.9112,
      "mean_token_accuracy": 0.15170105695724487,
      "num_tokens": 5090549.0,
      "step": 2460
    },
    {
      "entropy": 5.936306381225586,
      "epoch": 0.22360304789550073,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998005718834905,
      "loss": 5.767,
      "mean_token_accuracy": 0.15424081236124038,
      "num_tokens": 5100880.0,
      "step": 2465
    },
    {
      "entropy": 5.983072996139526,
      "epoch": 0.2240566037735849,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997992075463862,
      "loss": 5.8865,
      "mean_token_accuracy": 0.1511937439441681,
      "num_tokens": 5111796.0,
      "step": 2470
    },
    {
      "entropy": 6.071778011322022,
      "epoch": 0.2245101596516691,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997978385603769,
      "loss": 5.8722,
      "mean_token_accuracy": 0.14768328070640563,
      "num_tokens": 5123468.0,
      "step": 2475
    },
    {
      "entropy": 5.959640073776245,
      "epoch": 0.22496371552975328,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997964649254911,
      "loss": 5.8051,
      "mean_token_accuracy": 0.15414749532938005,
      "num_tokens": 5132925.0,
      "step": 2480
    },
    {
      "entropy": 5.989591789245606,
      "epoch": 0.22541727140783743,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997950866417572,
      "loss": 5.8777,
      "mean_token_accuracy": 0.1485254153609276,
      "num_tokens": 5143445.0,
      "step": 2485
    },
    {
      "entropy": 5.993931770324707,
      "epoch": 0.22587082728592162,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997937037092034,
      "loss": 5.8058,
      "mean_token_accuracy": 0.14833374321460724,
      "num_tokens": 5153176.0,
      "step": 2490
    },
    {
      "entropy": 6.067162322998047,
      "epoch": 0.2263243831640058,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997923161278587,
      "loss": 5.8445,
      "mean_token_accuracy": 0.14823048263788224,
      "num_tokens": 5162765.0,
      "step": 2495
    },
    {
      "entropy": 5.973460483551025,
      "epoch": 0.22677793904208998,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997909238977514,
      "loss": 5.8977,
      "mean_token_accuracy": 0.150830414891243,
      "num_tokens": 5173390.0,
      "step": 2500
    },
    {
      "entropy": 5.978850793838501,
      "epoch": 0.22723149492017417,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997895270189108,
      "loss": 5.8062,
      "mean_token_accuracy": 0.1472632072865963,
      "num_tokens": 5184212.0,
      "step": 2505
    },
    {
      "entropy": 5.894932794570923,
      "epoch": 0.22768505079825835,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997881254913653,
      "loss": 5.7866,
      "mean_token_accuracy": 0.14934516549110413,
      "num_tokens": 5194311.0,
      "step": 2510
    },
    {
      "entropy": 5.951768064498902,
      "epoch": 0.22813860667634253,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997867193151441,
      "loss": 5.7758,
      "mean_token_accuracy": 0.1520687684416771,
      "num_tokens": 5204708.0,
      "step": 2515
    },
    {
      "entropy": 5.913540315628052,
      "epoch": 0.22859216255442671,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997853084902763,
      "loss": 5.7608,
      "mean_token_accuracy": 0.15067852139472962,
      "num_tokens": 5214566.0,
      "step": 2520
    },
    {
      "entropy": 5.952923536300659,
      "epoch": 0.2290457184325109,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997838930167911,
      "loss": 5.8035,
      "mean_token_accuracy": 0.15393685698509216,
      "num_tokens": 5224762.0,
      "step": 2525
    },
    {
      "entropy": 6.007725048065185,
      "epoch": 0.22949927431059505,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997824728947176,
      "loss": 5.8629,
      "mean_token_accuracy": 0.14823582395911217,
      "num_tokens": 5235146.0,
      "step": 2530
    },
    {
      "entropy": 6.064925336837769,
      "epoch": 0.22995283018867924,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997810481240853,
      "loss": 5.8276,
      "mean_token_accuracy": 0.1468263030052185,
      "num_tokens": 5243867.0,
      "step": 2535
    },
    {
      "entropy": 5.916410827636719,
      "epoch": 0.23040638606676342,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997796187049238,
      "loss": 5.7346,
      "mean_token_accuracy": 0.1514027461409569,
      "num_tokens": 5254279.0,
      "step": 2540
    },
    {
      "entropy": 6.060354948043823,
      "epoch": 0.2308599419448476,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997781846372623,
      "loss": 5.8574,
      "mean_token_accuracy": 0.14669395685195924,
      "num_tokens": 5264769.0,
      "step": 2545
    },
    {
      "entropy": 5.961148881912232,
      "epoch": 0.23131349782293179,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997767459211307,
      "loss": 5.8023,
      "mean_token_accuracy": 0.14596094489097594,
      "num_tokens": 5274291.0,
      "step": 2550
    },
    {
      "entropy": 5.973608541488647,
      "epoch": 0.23176705370101597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997753025565587,
      "loss": 5.751,
      "mean_token_accuracy": 0.15511572808027269,
      "num_tokens": 5284366.0,
      "step": 2555
    },
    {
      "entropy": 5.983015012741089,
      "epoch": 0.23222060957910015,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997738545435763,
      "loss": 5.8651,
      "mean_token_accuracy": 0.14840636104345323,
      "num_tokens": 5294494.0,
      "step": 2560
    },
    {
      "entropy": 5.955273532867432,
      "epoch": 0.23267416545718433,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997724018822133,
      "loss": 5.8663,
      "mean_token_accuracy": 0.14921268746256827,
      "num_tokens": 5305606.0,
      "step": 2565
    },
    {
      "entropy": 5.970898771286011,
      "epoch": 0.23312772133526852,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997709445724996,
      "loss": 5.856,
      "mean_token_accuracy": 0.1487169645726681,
      "num_tokens": 5315677.0,
      "step": 2570
    },
    {
      "entropy": 6.037318992614746,
      "epoch": 0.23358127721335267,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997694826144656,
      "loss": 5.9139,
      "mean_token_accuracy": 0.14366162866353988,
      "num_tokens": 5326453.0,
      "step": 2575
    },
    {
      "entropy": 5.969986534118652,
      "epoch": 0.23403483309143686,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004997680160081414,
      "loss": 5.805,
      "mean_token_accuracy": 0.15401693433523178,
      "num_tokens": 5337591.0,
      "step": 2580
    },
    {
      "entropy": 5.923111820220948,
      "epoch": 0.23448838896952104,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997665447535573,
      "loss": 5.798,
      "mean_token_accuracy": 0.15265310406684876,
      "num_tokens": 5349400.0,
      "step": 2585
    },
    {
      "entropy": 6.048951911926269,
      "epoch": 0.23494194484760522,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997650688507438,
      "loss": 5.7909,
      "mean_token_accuracy": 0.1504625990986824,
      "num_tokens": 5359984.0,
      "step": 2590
    },
    {
      "entropy": 5.95233383178711,
      "epoch": 0.2353955007256894,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997635882997314,
      "loss": 5.8509,
      "mean_token_accuracy": 0.15640779882669448,
      "num_tokens": 5369254.0,
      "step": 2595
    },
    {
      "entropy": 5.96616530418396,
      "epoch": 0.2358490566037736,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997621031005507,
      "loss": 5.7868,
      "mean_token_accuracy": 0.15061139017343522,
      "num_tokens": 5379648.0,
      "step": 2600
    },
    {
      "entropy": 5.9050251007080075,
      "epoch": 0.23630261248185777,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997606132532325,
      "loss": 5.7438,
      "mean_token_accuracy": 0.1579679898917675,
      "num_tokens": 5390321.0,
      "step": 2605
    },
    {
      "entropy": 5.978159952163696,
      "epoch": 0.23675616835994195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997591187578075,
      "loss": 5.7975,
      "mean_token_accuracy": 0.15755213797092438,
      "num_tokens": 5401078.0,
      "step": 2610
    },
    {
      "entropy": 5.922601938247681,
      "epoch": 0.23720972423802614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997576196143067,
      "loss": 5.7705,
      "mean_token_accuracy": 0.15626103281974793,
      "num_tokens": 5411977.0,
      "step": 2615
    },
    {
      "entropy": 5.919599151611328,
      "epoch": 0.2376632801161103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499756115822761,
      "loss": 5.7429,
      "mean_token_accuracy": 0.14914836585521699,
      "num_tokens": 5422967.0,
      "step": 2620
    },
    {
      "entropy": 6.0454573154449465,
      "epoch": 0.23811683599419448,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997546073832015,
      "loss": 5.7987,
      "mean_token_accuracy": 0.16146147549152373,
      "num_tokens": 5432260.0,
      "step": 2625
    },
    {
      "entropy": 5.874481344223023,
      "epoch": 0.23857039187227866,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997530942956595,
      "loss": 5.7685,
      "mean_token_accuracy": 0.15337550789117813,
      "num_tokens": 5441706.0,
      "step": 2630
    },
    {
      "entropy": 5.954086065292358,
      "epoch": 0.23902394775036284,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997515765601662,
      "loss": 5.7734,
      "mean_token_accuracy": 0.16176194995641707,
      "num_tokens": 5452397.0,
      "step": 2635
    },
    {
      "entropy": 5.8905235767364506,
      "epoch": 0.23947750362844702,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004997500541767531,
      "loss": 5.732,
      "mean_token_accuracy": 0.15274471640586854,
      "num_tokens": 5463762.0,
      "step": 2640
    },
    {
      "entropy": 5.96301908493042,
      "epoch": 0.2399310595065312,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997485271454516,
      "loss": 5.8042,
      "mean_token_accuracy": 0.14668141603469848,
      "num_tokens": 5474622.0,
      "step": 2645
    },
    {
      "entropy": 5.9286675453186035,
      "epoch": 0.2403846153846154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997469954662933,
      "loss": 5.7846,
      "mean_token_accuracy": 0.1540173590183258,
      "num_tokens": 5484697.0,
      "step": 2650
    },
    {
      "entropy": 5.917514657974243,
      "epoch": 0.24083817126269957,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997454591393097,
      "loss": 5.7995,
      "mean_token_accuracy": 0.15156363248825072,
      "num_tokens": 5496433.0,
      "step": 2655
    },
    {
      "entropy": 6.03740963935852,
      "epoch": 0.24129172714078376,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997439181645329,
      "loss": 5.9087,
      "mean_token_accuracy": 0.14503674134612082,
      "num_tokens": 5507917.0,
      "step": 2660
    },
    {
      "entropy": 6.060143232345581,
      "epoch": 0.2417452830188679,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997423725419945,
      "loss": 5.8745,
      "mean_token_accuracy": 0.14898139089345933,
      "num_tokens": 5519159.0,
      "step": 2665
    },
    {
      "entropy": 5.990707302093506,
      "epoch": 0.2421988388969521,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997408222717267,
      "loss": 5.8503,
      "mean_token_accuracy": 0.15473444908857345,
      "num_tokens": 5529180.0,
      "step": 2670
    },
    {
      "entropy": 5.912744617462158,
      "epoch": 0.24265239477503628,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997392673537612,
      "loss": 5.7518,
      "mean_token_accuracy": 0.1590703994035721,
      "num_tokens": 5539363.0,
      "step": 2675
    },
    {
      "entropy": 5.91375207901001,
      "epoch": 0.24310595065312046,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997377077881306,
      "loss": 5.8534,
      "mean_token_accuracy": 0.15084453225135802,
      "num_tokens": 5549841.0,
      "step": 2680
    },
    {
      "entropy": 5.941425514221192,
      "epoch": 0.24355950653120464,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997361435748668,
      "loss": 5.7527,
      "mean_token_accuracy": 0.1558105766773224,
      "num_tokens": 5560299.0,
      "step": 2685
    },
    {
      "entropy": 5.983996868133545,
      "epoch": 0.24401306240928883,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997345747140024,
      "loss": 5.818,
      "mean_token_accuracy": 0.1520366132259369,
      "num_tokens": 5571081.0,
      "step": 2690
    },
    {
      "entropy": 5.893987655639648,
      "epoch": 0.244466618287373,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997330012055695,
      "loss": 5.7623,
      "mean_token_accuracy": 0.1579376369714737,
      "num_tokens": 5581389.0,
      "step": 2695
    },
    {
      "entropy": 5.951496028900147,
      "epoch": 0.2449201741654572,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997314230496011,
      "loss": 5.7861,
      "mean_token_accuracy": 0.14917654246091844,
      "num_tokens": 5592347.0,
      "step": 2700
    },
    {
      "entropy": 5.928359746932983,
      "epoch": 0.24537373004354138,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997298402461296,
      "loss": 5.7622,
      "mean_token_accuracy": 0.1548761934041977,
      "num_tokens": 5603411.0,
      "step": 2705
    },
    {
      "entropy": 5.997703504562378,
      "epoch": 0.24582728592162553,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997282527951876,
      "loss": 5.7484,
      "mean_token_accuracy": 0.1618112340569496,
      "num_tokens": 5614615.0,
      "step": 2710
    },
    {
      "entropy": 5.855966997146607,
      "epoch": 0.24628084179970972,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997266606968081,
      "loss": 5.7219,
      "mean_token_accuracy": 0.1555338978767395,
      "num_tokens": 5624582.0,
      "step": 2715
    },
    {
      "entropy": 5.882028532028198,
      "epoch": 0.2467343976777939,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997250639510241,
      "loss": 5.7521,
      "mean_token_accuracy": 0.1485395088791847,
      "num_tokens": 5634310.0,
      "step": 2720
    },
    {
      "entropy": 5.959417104721069,
      "epoch": 0.24718795355587808,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997234625578684,
      "loss": 5.8284,
      "mean_token_accuracy": 0.1493539556860924,
      "num_tokens": 5644018.0,
      "step": 2725
    },
    {
      "entropy": 5.905500411987305,
      "epoch": 0.24764150943396226,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997218565173742,
      "loss": 5.7685,
      "mean_token_accuracy": 0.15543410778045655,
      "num_tokens": 5653592.0,
      "step": 2730
    },
    {
      "entropy": 5.994117975234985,
      "epoch": 0.24809506531204645,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997202458295748,
      "loss": 5.829,
      "mean_token_accuracy": 0.15618719160556793,
      "num_tokens": 5662812.0,
      "step": 2735
    },
    {
      "entropy": 5.947774648666382,
      "epoch": 0.24854862119013063,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997186304945034,
      "loss": 5.7997,
      "mean_token_accuracy": 0.15028751492500306,
      "num_tokens": 5672939.0,
      "step": 2740
    },
    {
      "entropy": 5.8947399139404295,
      "epoch": 0.24900217706821481,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997170105121935,
      "loss": 5.7789,
      "mean_token_accuracy": 0.15528277158737183,
      "num_tokens": 5683350.0,
      "step": 2745
    },
    {
      "entropy": 6.005437898635864,
      "epoch": 0.249455732946299,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997153858826786,
      "loss": 5.8528,
      "mean_token_accuracy": 0.1439196802675724,
      "num_tokens": 5694516.0,
      "step": 2750
    },
    {
      "entropy": 5.938060188293457,
      "epoch": 0.24990928882438315,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997137566059922,
      "loss": 5.6847,
      "mean_token_accuracy": 0.15682510733604432,
      "num_tokens": 5705443.0,
      "step": 2755
    },
    {
      "entropy": 5.872813892364502,
      "epoch": 0.25036284470246734,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997121226821681,
      "loss": 5.7821,
      "mean_token_accuracy": 0.15022583901882172,
      "num_tokens": 5715933.0,
      "step": 2760
    },
    {
      "entropy": 5.941016101837159,
      "epoch": 0.25081640058055155,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049971048411124,
      "loss": 5.8439,
      "mean_token_accuracy": 0.14352803230285643,
      "num_tokens": 5726630.0,
      "step": 2765
    },
    {
      "entropy": 6.003946161270141,
      "epoch": 0.2512699564586357,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997088408932418,
      "loss": 5.7792,
      "mean_token_accuracy": 0.1525881379842758,
      "num_tokens": 5736890.0,
      "step": 2770
    },
    {
      "entropy": 5.804270029067993,
      "epoch": 0.25172351233671986,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997071930282076,
      "loss": 5.6826,
      "mean_token_accuracy": 0.16186858713626862,
      "num_tokens": 5747042.0,
      "step": 2775
    },
    {
      "entropy": 5.935064744949341,
      "epoch": 0.25217706821480407,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997055405161712,
      "loss": 5.7996,
      "mean_token_accuracy": 0.1525806337594986,
      "num_tokens": 5757367.0,
      "step": 2780
    },
    {
      "entropy": 5.97401123046875,
      "epoch": 0.2526306240928882,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997038833571671,
      "loss": 5.8662,
      "mean_token_accuracy": 0.1486934617161751,
      "num_tokens": 5769557.0,
      "step": 2785
    },
    {
      "entropy": 5.932326173782348,
      "epoch": 0.25308417997097243,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997022215512295,
      "loss": 5.7594,
      "mean_token_accuracy": 0.1602458581328392,
      "num_tokens": 5780615.0,
      "step": 2790
    },
    {
      "entropy": 5.959806776046753,
      "epoch": 0.2535377358490566,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997005550983925,
      "loss": 5.737,
      "mean_token_accuracy": 0.15268604904413224,
      "num_tokens": 5791271.0,
      "step": 2795
    },
    {
      "entropy": 5.863793230056762,
      "epoch": 0.2539912917271408,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499698883998691,
      "loss": 5.6886,
      "mean_token_accuracy": 0.15530947148799895,
      "num_tokens": 5801117.0,
      "step": 2800
    },
    {
      "entropy": 5.88543930053711,
      "epoch": 0.25444484760522496,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996972082521591,
      "loss": 5.7122,
      "mean_token_accuracy": 0.1563066676259041,
      "num_tokens": 5811696.0,
      "step": 2805
    },
    {
      "entropy": 5.911672639846802,
      "epoch": 0.25489840348330917,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996955278588319,
      "loss": 5.738,
      "mean_token_accuracy": 0.16307492852210997,
      "num_tokens": 5822207.0,
      "step": 2810
    },
    {
      "entropy": 5.838271713256836,
      "epoch": 0.2553519593613933,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996938428187437,
      "loss": 5.7025,
      "mean_token_accuracy": 0.15967562049627304,
      "num_tokens": 5831927.0,
      "step": 2815
    },
    {
      "entropy": 5.887790107727051,
      "epoch": 0.2558055152394775,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996921531319297,
      "loss": 5.7578,
      "mean_token_accuracy": 0.15641293302178383,
      "num_tokens": 5842093.0,
      "step": 2820
    },
    {
      "entropy": 5.887936401367187,
      "epoch": 0.2562590711175617,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996904587984246,
      "loss": 5.704,
      "mean_token_accuracy": 0.16297227442264556,
      "num_tokens": 5850551.0,
      "step": 2825
    },
    {
      "entropy": 5.81576018333435,
      "epoch": 0.25671262699564584,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996887598182636,
      "loss": 5.6901,
      "mean_token_accuracy": 0.15636582523584366,
      "num_tokens": 5860651.0,
      "step": 2830
    },
    {
      "entropy": 5.827975130081176,
      "epoch": 0.25716618287373005,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996870561914819,
      "loss": 5.6698,
      "mean_token_accuracy": 0.15702708512544633,
      "num_tokens": 5871500.0,
      "step": 2835
    },
    {
      "entropy": 5.840231275558471,
      "epoch": 0.2576197387518142,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996853479181145,
      "loss": 5.6725,
      "mean_token_accuracy": 0.15341195613145828,
      "num_tokens": 5881523.0,
      "step": 2840
    },
    {
      "entropy": 5.9528480052948,
      "epoch": 0.2580732946298984,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996836349981969,
      "loss": 5.8174,
      "mean_token_accuracy": 0.15260155200958253,
      "num_tokens": 5891162.0,
      "step": 2845
    },
    {
      "entropy": 5.902073764801026,
      "epoch": 0.2585268505079826,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996819174317645,
      "loss": 5.7173,
      "mean_token_accuracy": 0.1552115336060524,
      "num_tokens": 5901818.0,
      "step": 2850
    },
    {
      "entropy": 5.874346494674683,
      "epoch": 0.2589804063860668,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996801952188528,
      "loss": 5.79,
      "mean_token_accuracy": 0.15259438008069992,
      "num_tokens": 5913889.0,
      "step": 2855
    },
    {
      "entropy": 5.9258890628814695,
      "epoch": 0.25943396226415094,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996784683594973,
      "loss": 5.7571,
      "mean_token_accuracy": 0.156342051923275,
      "num_tokens": 5923429.0,
      "step": 2860
    },
    {
      "entropy": 5.998136568069458,
      "epoch": 0.2598875181422351,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996767368537338,
      "loss": 5.7779,
      "mean_token_accuracy": 0.15545209050178527,
      "num_tokens": 5933847.0,
      "step": 2865
    },
    {
      "entropy": 5.854821443557739,
      "epoch": 0.2603410740203193,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996750007015983,
      "loss": 5.6677,
      "mean_token_accuracy": 0.15953482687473297,
      "num_tokens": 5943874.0,
      "step": 2870
    },
    {
      "entropy": 5.930334520339966,
      "epoch": 0.26079462989840346,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996732599031264,
      "loss": 5.7357,
      "mean_token_accuracy": 0.1538098156452179,
      "num_tokens": 5953653.0,
      "step": 2875
    },
    {
      "entropy": 5.918991279602051,
      "epoch": 0.2612481857764877,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996715144583543,
      "loss": 5.6866,
      "mean_token_accuracy": 0.15937823802232742,
      "num_tokens": 5964039.0,
      "step": 2880
    },
    {
      "entropy": 5.940640115737915,
      "epoch": 0.26170174165457183,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499669764367318,
      "loss": 5.8108,
      "mean_token_accuracy": 0.15168801695108414,
      "num_tokens": 5974083.0,
      "step": 2885
    },
    {
      "entropy": 5.8325118064880375,
      "epoch": 0.26215529753265604,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996680096300537,
      "loss": 5.7435,
      "mean_token_accuracy": 0.15367064327001573,
      "num_tokens": 5984629.0,
      "step": 2890
    },
    {
      "entropy": 5.8481368064880375,
      "epoch": 0.2626088534107402,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996662502465978,
      "loss": 5.6702,
      "mean_token_accuracy": 0.1644247755408287,
      "num_tokens": 5994295.0,
      "step": 2895
    },
    {
      "entropy": 5.873128223419189,
      "epoch": 0.2630624092888244,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996644862169864,
      "loss": 5.7046,
      "mean_token_accuracy": 0.15299894511699677,
      "num_tokens": 6004156.0,
      "step": 2900
    },
    {
      "entropy": 5.876518249511719,
      "epoch": 0.26351596516690856,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996627175412563,
      "loss": 5.7805,
      "mean_token_accuracy": 0.15570748448371888,
      "num_tokens": 6014188.0,
      "step": 2905
    },
    {
      "entropy": 5.96129150390625,
      "epoch": 0.2639695210449927,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499660944219444,
      "loss": 5.8435,
      "mean_token_accuracy": 0.1530625268816948,
      "num_tokens": 6024879.0,
      "step": 2910
    },
    {
      "entropy": 5.983526945114136,
      "epoch": 0.2644230769230769,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004996591662515861,
      "loss": 5.8974,
      "mean_token_accuracy": 0.15041209682822226,
      "num_tokens": 6036374.0,
      "step": 2915
    },
    {
      "entropy": 5.880353546142578,
      "epoch": 0.2648766328011611,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996573836377193,
      "loss": 5.7655,
      "mean_token_accuracy": 0.1547638565301895,
      "num_tokens": 6046432.0,
      "step": 2920
    },
    {
      "entropy": 5.962761545181275,
      "epoch": 0.2653301886792453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996555963778806,
      "loss": 5.763,
      "mean_token_accuracy": 0.15179126113653182,
      "num_tokens": 6056864.0,
      "step": 2925
    },
    {
      "entropy": 6.006943416595459,
      "epoch": 0.26578374455732945,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499653804472107,
      "loss": 5.7827,
      "mean_token_accuracy": 0.15777789205312728,
      "num_tokens": 6067294.0,
      "step": 2930
    },
    {
      "entropy": 5.932409477233887,
      "epoch": 0.26623730043541366,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996520079204353,
      "loss": 5.7948,
      "mean_token_accuracy": 0.15130407512187957,
      "num_tokens": 6077587.0,
      "step": 2935
    },
    {
      "entropy": 5.947808265686035,
      "epoch": 0.2666908563134978,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499650206722903,
      "loss": 5.7315,
      "mean_token_accuracy": 0.1543162226676941,
      "num_tokens": 6089306.0,
      "step": 2940
    },
    {
      "entropy": 5.866966199874878,
      "epoch": 0.267144412191582,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996484008795471,
      "loss": 5.6891,
      "mean_token_accuracy": 0.15265745371580125,
      "num_tokens": 6099651.0,
      "step": 2945
    },
    {
      "entropy": 5.864347743988037,
      "epoch": 0.2675979680696662,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499646590390405,
      "loss": 5.688,
      "mean_token_accuracy": 0.1571669578552246,
      "num_tokens": 6110063.0,
      "step": 2950
    },
    {
      "entropy": 5.873627424240112,
      "epoch": 0.26805152394775034,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996447752555142,
      "loss": 5.7144,
      "mean_token_accuracy": 0.15664682239294053,
      "num_tokens": 6120031.0,
      "step": 2955
    },
    {
      "entropy": 5.848915719985962,
      "epoch": 0.26850507982583455,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996429554749122,
      "loss": 5.7616,
      "mean_token_accuracy": 0.15996424555778505,
      "num_tokens": 6131274.0,
      "step": 2960
    },
    {
      "entropy": 5.9580456733703615,
      "epoch": 0.2689586357039187,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996411310486367,
      "loss": 5.8021,
      "mean_token_accuracy": 0.1580547109246254,
      "num_tokens": 6140878.0,
      "step": 2965
    },
    {
      "entropy": 5.900578355789184,
      "epoch": 0.2694121915820029,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004996393019767253,
      "loss": 5.7986,
      "mean_token_accuracy": 0.15211150646209717,
      "num_tokens": 6152229.0,
      "step": 2970
    },
    {
      "entropy": 5.871568775177002,
      "epoch": 0.26986574746008707,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996374682592158,
      "loss": 5.6994,
      "mean_token_accuracy": 0.15565146058797835,
      "num_tokens": 6162311.0,
      "step": 2975
    },
    {
      "entropy": 5.939528369903565,
      "epoch": 0.2703193033381713,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996356298961463,
      "loss": 5.8077,
      "mean_token_accuracy": 0.15412216633558273,
      "num_tokens": 6172806.0,
      "step": 2980
    },
    {
      "entropy": 5.916967105865479,
      "epoch": 0.27077285921625543,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996337868875549,
      "loss": 5.774,
      "mean_token_accuracy": 0.15645444989204407,
      "num_tokens": 6183479.0,
      "step": 2985
    },
    {
      "entropy": 5.928597736358642,
      "epoch": 0.27122641509433965,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996319392334792,
      "loss": 5.7363,
      "mean_token_accuracy": 0.1543424740433693,
      "num_tokens": 6194201.0,
      "step": 2990
    },
    {
      "entropy": 5.853014802932739,
      "epoch": 0.2716799709724238,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499630086933958,
      "loss": 5.6793,
      "mean_token_accuracy": 0.15483027845621108,
      "num_tokens": 6204774.0,
      "step": 2995
    },
    {
      "entropy": 5.876689767837524,
      "epoch": 0.27213352685050796,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996282299890292,
      "loss": 5.6781,
      "mean_token_accuracy": 0.15287237316370011,
      "num_tokens": 6215056.0,
      "step": 3000
    },
    {
      "epoch": 0.27213352685050796,
      "eval_entropy": 5.712364828856282,
      "eval_loss": 5.759267330169678,
      "eval_mean_token_accuracy": 0.1617233954799863,
      "eval_num_tokens": 6215056.0,
      "eval_runtime": 21.065,
      "eval_samples_per_second": 1678.618,
      "eval_steps_per_second": 209.827,
      "step": 3000
    },
    {
      "entropy": 5.969257974624634,
      "epoch": 0.27258708272859217,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004996263683987316,
      "loss": 5.7935,
      "mean_token_accuracy": 0.15573310405015944,
      "num_tokens": 6226264.0,
      "step": 3005
    },
    {
      "entropy": 5.891745710372925,
      "epoch": 0.2730406386066763,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996245021631033,
      "loss": 5.7187,
      "mean_token_accuracy": 0.16169505268335344,
      "num_tokens": 6235137.0,
      "step": 3010
    },
    {
      "entropy": 5.822071361541748,
      "epoch": 0.27349419448476053,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499622631282183,
      "loss": 5.7275,
      "mean_token_accuracy": 0.1584876984357834,
      "num_tokens": 6245411.0,
      "step": 3015
    },
    {
      "entropy": 5.966612482070923,
      "epoch": 0.2739477503628447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996207557560095,
      "loss": 5.7953,
      "mean_token_accuracy": 0.152059169113636,
      "num_tokens": 6255633.0,
      "step": 3020
    },
    {
      "entropy": 5.933960390090943,
      "epoch": 0.2744013062409289,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996188755846216,
      "loss": 5.7423,
      "mean_token_accuracy": 0.15434571653604506,
      "num_tokens": 6265109.0,
      "step": 3025
    },
    {
      "entropy": 5.862824583053589,
      "epoch": 0.27485486211901305,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499616990768058,
      "loss": 5.7933,
      "mean_token_accuracy": 0.14787728786468507,
      "num_tokens": 6275241.0,
      "step": 3030
    },
    {
      "entropy": 5.949290132522583,
      "epoch": 0.27530841799709727,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996151013063579,
      "loss": 5.8252,
      "mean_token_accuracy": 0.14647485911846161,
      "num_tokens": 6286603.0,
      "step": 3035
    },
    {
      "entropy": 5.912906169891357,
      "epoch": 0.2757619738751814,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996132071995601,
      "loss": 5.6852,
      "mean_token_accuracy": 0.16648433357477188,
      "num_tokens": 6297219.0,
      "step": 3040
    },
    {
      "entropy": 5.883165311813355,
      "epoch": 0.2762155297532656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996113084477042,
      "loss": 5.719,
      "mean_token_accuracy": 0.16386881619691848,
      "num_tokens": 6307694.0,
      "step": 3045
    },
    {
      "entropy": 5.807434129714966,
      "epoch": 0.2766690856313498,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996094050508291,
      "loss": 5.6527,
      "mean_token_accuracy": 0.16033376455307008,
      "num_tokens": 6317384.0,
      "step": 3050
    },
    {
      "entropy": 5.947029972076416,
      "epoch": 0.27712264150943394,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996074970089741,
      "loss": 5.7779,
      "mean_token_accuracy": 0.15559548288583755,
      "num_tokens": 6327558.0,
      "step": 3055
    },
    {
      "entropy": 5.87425684928894,
      "epoch": 0.27757619738751815,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996055843221789,
      "loss": 5.6409,
      "mean_token_accuracy": 0.1627122566103935,
      "num_tokens": 6337828.0,
      "step": 3060
    },
    {
      "entropy": 5.822861051559448,
      "epoch": 0.2780297532656023,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996036669904831,
      "loss": 5.7058,
      "mean_token_accuracy": 0.16010485142469405,
      "num_tokens": 6348678.0,
      "step": 3065
    },
    {
      "entropy": 5.901842355728149,
      "epoch": 0.2784833091436865,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996017450139261,
      "loss": 5.7969,
      "mean_token_accuracy": 0.1491397202014923,
      "num_tokens": 6359670.0,
      "step": 3070
    },
    {
      "entropy": 5.95522518157959,
      "epoch": 0.2789368650217707,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995998183925478,
      "loss": 5.7541,
      "mean_token_accuracy": 0.15521771609783172,
      "num_tokens": 6371034.0,
      "step": 3075
    },
    {
      "entropy": 5.83247013092041,
      "epoch": 0.2793904208998549,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995978871263881,
      "loss": 5.7014,
      "mean_token_accuracy": 0.15728666633367538,
      "num_tokens": 6380672.0,
      "step": 3080
    },
    {
      "entropy": 5.874285125732422,
      "epoch": 0.27984397677793904,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995959512154868,
      "loss": 5.8151,
      "mean_token_accuracy": 0.14317371249198912,
      "num_tokens": 6390799.0,
      "step": 3085
    },
    {
      "entropy": 5.8934972286224365,
      "epoch": 0.2802975326560232,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499594010659884,
      "loss": 5.6884,
      "mean_token_accuracy": 0.16007015258073806,
      "num_tokens": 6400965.0,
      "step": 3090
    },
    {
      "entropy": 5.899712657928466,
      "epoch": 0.2807510885341074,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995920654596199,
      "loss": 5.7049,
      "mean_token_accuracy": 0.1535528853535652,
      "num_tokens": 6411567.0,
      "step": 3095
    },
    {
      "entropy": 5.8801034450531,
      "epoch": 0.28120464441219156,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995901156147345,
      "loss": 5.7613,
      "mean_token_accuracy": 0.1587476283311844,
      "num_tokens": 6421193.0,
      "step": 3100
    },
    {
      "entropy": 5.9482683658599855,
      "epoch": 0.2816582002902758,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995881611252684,
      "loss": 5.7637,
      "mean_token_accuracy": 0.164001926779747,
      "num_tokens": 6430281.0,
      "step": 3105
    },
    {
      "entropy": 5.899417304992676,
      "epoch": 0.28211175616835993,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995862019912619,
      "loss": 5.756,
      "mean_token_accuracy": 0.15074933171272278,
      "num_tokens": 6441055.0,
      "step": 3110
    },
    {
      "entropy": 5.860140228271485,
      "epoch": 0.28256531204644414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995842382127555,
      "loss": 5.6706,
      "mean_token_accuracy": 0.15401192158460617,
      "num_tokens": 6451082.0,
      "step": 3115
    },
    {
      "entropy": 5.84559817314148,
      "epoch": 0.2830188679245283,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995822697897897,
      "loss": 5.7639,
      "mean_token_accuracy": 0.1507149502635002,
      "num_tokens": 6461617.0,
      "step": 3120
    },
    {
      "entropy": 5.967565059661865,
      "epoch": 0.2834724238026125,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995802967224056,
      "loss": 5.7685,
      "mean_token_accuracy": 0.15601267218589782,
      "num_tokens": 6472780.0,
      "step": 3125
    },
    {
      "entropy": 5.77694001197815,
      "epoch": 0.28392597968069666,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995783190106434,
      "loss": 5.6677,
      "mean_token_accuracy": 0.15176742225885392,
      "num_tokens": 6484078.0,
      "step": 3130
    },
    {
      "entropy": 5.900559663772583,
      "epoch": 0.2843795355587808,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995763366545445,
      "loss": 5.7619,
      "mean_token_accuracy": 0.15375943332910538,
      "num_tokens": 6494366.0,
      "step": 3135
    },
    {
      "entropy": 5.787074327468872,
      "epoch": 0.284833091436865,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995743496541497,
      "loss": 5.6902,
      "mean_token_accuracy": 0.16150298565626145,
      "num_tokens": 6505461.0,
      "step": 3140
    },
    {
      "entropy": 5.881974554061889,
      "epoch": 0.2852866473149492,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995723580095001,
      "loss": 5.6923,
      "mean_token_accuracy": 0.15881099700927734,
      "num_tokens": 6516141.0,
      "step": 3145
    },
    {
      "entropy": 5.84017424583435,
      "epoch": 0.2857402031930334,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499570361720637,
      "loss": 5.7058,
      "mean_token_accuracy": 0.1565474107861519,
      "num_tokens": 6525536.0,
      "step": 3150
    },
    {
      "entropy": 5.8899266719818115,
      "epoch": 0.28619375907111755,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995683607876015,
      "loss": 5.7579,
      "mean_token_accuracy": 0.15209461599588395,
      "num_tokens": 6535852.0,
      "step": 3155
    },
    {
      "entropy": 5.924435234069824,
      "epoch": 0.28664731494920176,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995663552104352,
      "loss": 5.7252,
      "mean_token_accuracy": 0.16235753893852234,
      "num_tokens": 6546353.0,
      "step": 3160
    },
    {
      "entropy": 5.779373979568481,
      "epoch": 0.2871008708272859,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995643449891793,
      "loss": 5.6751,
      "mean_token_accuracy": 0.1516367569565773,
      "num_tokens": 6557782.0,
      "step": 3165
    },
    {
      "entropy": 5.916032218933106,
      "epoch": 0.2875544267053701,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995623301238757,
      "loss": 5.6995,
      "mean_token_accuracy": 0.15973981767892836,
      "num_tokens": 6568100.0,
      "step": 3170
    },
    {
      "entropy": 5.865595197677612,
      "epoch": 0.2880079825834543,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995603106145657,
      "loss": 5.7505,
      "mean_token_accuracy": 0.15267078280448915,
      "num_tokens": 6578698.0,
      "step": 3175
    },
    {
      "entropy": 5.876104164123535,
      "epoch": 0.28846153846153844,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995582864612913,
      "loss": 5.7308,
      "mean_token_accuracy": 0.1522951140999794,
      "num_tokens": 6588908.0,
      "step": 3180
    },
    {
      "entropy": 5.780710887908936,
      "epoch": 0.28891509433962265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995562576640943,
      "loss": 5.589,
      "mean_token_accuracy": 0.15807687640190124,
      "num_tokens": 6599445.0,
      "step": 3185
    },
    {
      "entropy": 5.823026847839356,
      "epoch": 0.2893686502177068,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995542242230167,
      "loss": 5.6838,
      "mean_token_accuracy": 0.16184857338666916,
      "num_tokens": 6610246.0,
      "step": 3190
    },
    {
      "entropy": 5.840323448181152,
      "epoch": 0.289822206095791,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995521861381006,
      "loss": 5.6896,
      "mean_token_accuracy": 0.16104330569505693,
      "num_tokens": 6620501.0,
      "step": 3195
    },
    {
      "entropy": 5.950109958648682,
      "epoch": 0.29027576197387517,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499550143409388,
      "loss": 5.7599,
      "mean_token_accuracy": 0.1512766182422638,
      "num_tokens": 6631206.0,
      "step": 3200
    },
    {
      "entropy": 5.829269504547119,
      "epoch": 0.2907293178519594,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995480960369211,
      "loss": 5.7188,
      "mean_token_accuracy": 0.1548565149307251,
      "num_tokens": 6641600.0,
      "step": 3205
    },
    {
      "entropy": 5.827369594573975,
      "epoch": 0.29118287373004353,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995460440207425,
      "loss": 5.7759,
      "mean_token_accuracy": 0.15423371642827988,
      "num_tokens": 6652319.0,
      "step": 3210
    },
    {
      "entropy": 5.945125961303711,
      "epoch": 0.29163642960812775,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995439873608945,
      "loss": 5.712,
      "mean_token_accuracy": 0.14905817657709122,
      "num_tokens": 6663702.0,
      "step": 3215
    },
    {
      "entropy": 5.812268924713135,
      "epoch": 0.2920899854862119,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995419260574196,
      "loss": 5.7027,
      "mean_token_accuracy": 0.1491316169500351,
      "num_tokens": 6673908.0,
      "step": 3220
    },
    {
      "entropy": 5.916238164901733,
      "epoch": 0.29254354136429606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995398601103603,
      "loss": 5.727,
      "mean_token_accuracy": 0.15841997563838958,
      "num_tokens": 6683237.0,
      "step": 3225
    },
    {
      "entropy": 5.824212837219238,
      "epoch": 0.29299709724238027,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995377895197597,
      "loss": 5.6527,
      "mean_token_accuracy": 0.15979209095239638,
      "num_tokens": 6693451.0,
      "step": 3230
    },
    {
      "entropy": 5.761458158493042,
      "epoch": 0.2934506531204644,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995357142856602,
      "loss": 5.6124,
      "mean_token_accuracy": 0.15973278135061264,
      "num_tokens": 6703479.0,
      "step": 3235
    },
    {
      "entropy": 5.817833805084229,
      "epoch": 0.29390420899854863,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499533634408105,
      "loss": 5.619,
      "mean_token_accuracy": 0.15864339768886565,
      "num_tokens": 6715321.0,
      "step": 3240
    },
    {
      "entropy": 5.808075428009033,
      "epoch": 0.2943577648766328,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499531549887137,
      "loss": 5.6835,
      "mean_token_accuracy": 0.15339466109871863,
      "num_tokens": 6725541.0,
      "step": 3245
    },
    {
      "entropy": 5.856081151962281,
      "epoch": 0.294811320754717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995294607227993,
      "loss": 5.7241,
      "mean_token_accuracy": 0.16161034405231475,
      "num_tokens": 6735165.0,
      "step": 3250
    },
    {
      "entropy": 5.899437427520752,
      "epoch": 0.29526487663280115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995273669151351,
      "loss": 5.7105,
      "mean_token_accuracy": 0.16171984821558,
      "num_tokens": 6744245.0,
      "step": 3255
    },
    {
      "entropy": 5.866788148880005,
      "epoch": 0.29571843251088537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499525268464188,
      "loss": 5.6832,
      "mean_token_accuracy": 0.16346385776996614,
      "num_tokens": 6754894.0,
      "step": 3260
    },
    {
      "entropy": 5.863985061645508,
      "epoch": 0.2961719883889695,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995231653700009,
      "loss": 5.7333,
      "mean_token_accuracy": 0.15298420935869217,
      "num_tokens": 6764442.0,
      "step": 3265
    },
    {
      "entropy": 5.820025396347046,
      "epoch": 0.2966255442670537,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995210576326176,
      "loss": 5.6891,
      "mean_token_accuracy": 0.15496449172496796,
      "num_tokens": 6775054.0,
      "step": 3270
    },
    {
      "entropy": 5.862255716323853,
      "epoch": 0.2970791001451379,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995189452520816,
      "loss": 5.6861,
      "mean_token_accuracy": 0.16984087377786636,
      "num_tokens": 6784825.0,
      "step": 3275
    },
    {
      "entropy": 5.884560394287109,
      "epoch": 0.29753265602322204,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995168282284366,
      "loss": 5.7215,
      "mean_token_accuracy": 0.160723876953125,
      "num_tokens": 6795928.0,
      "step": 3280
    },
    {
      "entropy": 5.787168312072754,
      "epoch": 0.29798621190130625,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995147065617263,
      "loss": 5.6523,
      "mean_token_accuracy": 0.15839823037385942,
      "num_tokens": 6805456.0,
      "step": 3285
    },
    {
      "entropy": 5.925667858123779,
      "epoch": 0.2984397677793904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995125802519946,
      "loss": 5.7188,
      "mean_token_accuracy": 0.15525027811527253,
      "num_tokens": 6816215.0,
      "step": 3290
    },
    {
      "entropy": 5.892514085769653,
      "epoch": 0.2988933236574746,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995104492992857,
      "loss": 5.7558,
      "mean_token_accuracy": 0.15865902602672577,
      "num_tokens": 6826545.0,
      "step": 3295
    },
    {
      "entropy": 5.798425722122192,
      "epoch": 0.2993468795355588,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995083137036434,
      "loss": 5.6637,
      "mean_token_accuracy": 0.1575020283460617,
      "num_tokens": 6836184.0,
      "step": 3300
    },
    {
      "entropy": 5.924666786193848,
      "epoch": 0.299800435413643,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995061734651121,
      "loss": 5.7419,
      "mean_token_accuracy": 0.15869775861501695,
      "num_tokens": 6846646.0,
      "step": 3305
    },
    {
      "entropy": 5.840148878097534,
      "epoch": 0.30025399129172714,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995040285837358,
      "loss": 5.6502,
      "mean_token_accuracy": 0.16409194618463516,
      "num_tokens": 6857597.0,
      "step": 3310
    },
    {
      "entropy": 5.853920841217041,
      "epoch": 0.3007075471698113,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995018790595589,
      "loss": 5.7448,
      "mean_token_accuracy": 0.15398296862840652,
      "num_tokens": 6867523.0,
      "step": 3315
    },
    {
      "entropy": 5.901988887786866,
      "epoch": 0.3011611030478955,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994997248926261,
      "loss": 5.8385,
      "mean_token_accuracy": 0.15451575368642806,
      "num_tokens": 6878190.0,
      "step": 3320
    },
    {
      "entropy": 5.849097967147827,
      "epoch": 0.30161465892597966,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994975660829817,
      "loss": 5.6408,
      "mean_token_accuracy": 0.16445966064929962,
      "num_tokens": 6888483.0,
      "step": 3325
    },
    {
      "entropy": 5.7745849132537845,
      "epoch": 0.3020682148040639,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994954026306705,
      "loss": 5.6223,
      "mean_token_accuracy": 0.15969739556312562,
      "num_tokens": 6898337.0,
      "step": 3330
    },
    {
      "entropy": 5.876712703704834,
      "epoch": 0.30252177068214803,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994932345357371,
      "loss": 5.6971,
      "mean_token_accuracy": 0.15568175464868544,
      "num_tokens": 6909131.0,
      "step": 3335
    },
    {
      "entropy": 5.822457551956177,
      "epoch": 0.30297532656023224,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994910617982264,
      "loss": 5.6823,
      "mean_token_accuracy": 0.15465617924928665,
      "num_tokens": 6919790.0,
      "step": 3340
    },
    {
      "entropy": 5.841094827651977,
      "epoch": 0.3034288824383164,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994888844181833,
      "loss": 5.6997,
      "mean_token_accuracy": 0.16144252866506575,
      "num_tokens": 6930072.0,
      "step": 3345
    },
    {
      "entropy": 5.7801776885986325,
      "epoch": 0.3038824383164006,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994867023956529,
      "loss": 5.5965,
      "mean_token_accuracy": 0.15485866814851762,
      "num_tokens": 6940646.0,
      "step": 3350
    },
    {
      "entropy": 5.740756988525391,
      "epoch": 0.30433599419448476,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994845157306803,
      "loss": 5.5767,
      "mean_token_accuracy": 0.16163998693227768,
      "num_tokens": 6951645.0,
      "step": 3355
    },
    {
      "entropy": 5.876171684265136,
      "epoch": 0.3047895500725689,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994823244233107,
      "loss": 5.7416,
      "mean_token_accuracy": 0.1575745850801468,
      "num_tokens": 6963020.0,
      "step": 3360
    },
    {
      "entropy": 5.882272481918335,
      "epoch": 0.3052431059506531,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994801284735895,
      "loss": 5.6823,
      "mean_token_accuracy": 0.1615156725049019,
      "num_tokens": 6972895.0,
      "step": 3365
    },
    {
      "entropy": 5.877938604354858,
      "epoch": 0.3056966618287373,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994779278815619,
      "loss": 5.8195,
      "mean_token_accuracy": 0.1476362816989422,
      "num_tokens": 6985363.0,
      "step": 3370
    },
    {
      "entropy": 5.906639528274536,
      "epoch": 0.3061502177068215,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994757226472736,
      "loss": 5.7638,
      "mean_token_accuracy": 0.15788324922323227,
      "num_tokens": 6997040.0,
      "step": 3375
    },
    {
      "entropy": 5.863248682022094,
      "epoch": 0.30660377358490565,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994735127707702,
      "loss": 5.6891,
      "mean_token_accuracy": 0.15913514643907548,
      "num_tokens": 7007591.0,
      "step": 3380
    },
    {
      "entropy": 5.887237691879273,
      "epoch": 0.30705732946298986,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994712982520972,
      "loss": 5.6886,
      "mean_token_accuracy": 0.1630169630050659,
      "num_tokens": 7017858.0,
      "step": 3385
    },
    {
      "entropy": 5.760348892211914,
      "epoch": 0.307510885341074,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994690790913008,
      "loss": 5.6526,
      "mean_token_accuracy": 0.15853054076433182,
      "num_tokens": 7027365.0,
      "step": 3390
    },
    {
      "entropy": 5.775932741165161,
      "epoch": 0.3079644412191582,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994668552884265,
      "loss": 5.6393,
      "mean_token_accuracy": 0.16287402510643006,
      "num_tokens": 7038728.0,
      "step": 3395
    },
    {
      "entropy": 5.94070987701416,
      "epoch": 0.3084179970972424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994646268435204,
      "loss": 5.5879,
      "mean_token_accuracy": 0.16382095515727996,
      "num_tokens": 7048736.0,
      "step": 3400
    },
    {
      "entropy": 5.77776608467102,
      "epoch": 0.30887155297532654,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994623937566286,
      "loss": 5.7134,
      "mean_token_accuracy": 0.16350821107625962,
      "num_tokens": 7059205.0,
      "step": 3405
    },
    {
      "entropy": 5.802759027481079,
      "epoch": 0.30932510885341075,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994601560277974,
      "loss": 5.7095,
      "mean_token_accuracy": 0.15878489315509797,
      "num_tokens": 7069513.0,
      "step": 3410
    },
    {
      "entropy": 5.834039211273193,
      "epoch": 0.3097786647314949,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499457913657073,
      "loss": 5.6357,
      "mean_token_accuracy": 0.1570906937122345,
      "num_tokens": 7079940.0,
      "step": 3415
    },
    {
      "entropy": 5.844655847549438,
      "epoch": 0.3102322206095791,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994556666445016,
      "loss": 5.6323,
      "mean_token_accuracy": 0.1617613285779953,
      "num_tokens": 7090792.0,
      "step": 3420
    },
    {
      "entropy": 5.776029062271118,
      "epoch": 0.31068577648766327,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994534149901299,
      "loss": 5.583,
      "mean_token_accuracy": 0.16228142231702805,
      "num_tokens": 7101964.0,
      "step": 3425
    },
    {
      "entropy": 5.8066362857818605,
      "epoch": 0.3111393323657475,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994511586940044,
      "loss": 5.6499,
      "mean_token_accuracy": 0.15664990097284318,
      "num_tokens": 7113332.0,
      "step": 3430
    },
    {
      "entropy": 5.942155599594116,
      "epoch": 0.31159288824383163,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994488977561717,
      "loss": 5.8081,
      "mean_token_accuracy": 0.15569447129964828,
      "num_tokens": 7123481.0,
      "step": 3435
    },
    {
      "entropy": 5.719643592834473,
      "epoch": 0.31204644412191584,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994466321766786,
      "loss": 5.5667,
      "mean_token_accuracy": 0.16554133594036102,
      "num_tokens": 7133694.0,
      "step": 3440
    },
    {
      "entropy": 5.836681747436524,
      "epoch": 0.3125,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499444361955572,
      "loss": 5.6901,
      "mean_token_accuracy": 0.16068354696035386,
      "num_tokens": 7144491.0,
      "step": 3445
    },
    {
      "entropy": 5.970473051071167,
      "epoch": 0.31295355587808416,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994420870928988,
      "loss": 5.7727,
      "mean_token_accuracy": 0.1537349671125412,
      "num_tokens": 7154716.0,
      "step": 3450
    },
    {
      "entropy": 5.861110353469849,
      "epoch": 0.31340711175616837,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994398075887059,
      "loss": 5.7654,
      "mean_token_accuracy": 0.15032783523201942,
      "num_tokens": 7164258.0,
      "step": 3455
    },
    {
      "entropy": 5.91595458984375,
      "epoch": 0.3138606676342525,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994375234430407,
      "loss": 5.6193,
      "mean_token_accuracy": 0.15683988332748414,
      "num_tokens": 7175098.0,
      "step": 3460
    },
    {
      "entropy": 5.792357158660889,
      "epoch": 0.31431422351233673,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994352346559504,
      "loss": 5.7077,
      "mean_token_accuracy": 0.15706134140491484,
      "num_tokens": 7186480.0,
      "step": 3465
    },
    {
      "entropy": 5.765750789642334,
      "epoch": 0.3147677793904209,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994329412274821,
      "loss": 5.6699,
      "mean_token_accuracy": 0.16495281904935838,
      "num_tokens": 7197056.0,
      "step": 3470
    },
    {
      "entropy": 5.9141796112060545,
      "epoch": 0.3152213352685051,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994306431576835,
      "loss": 5.8035,
      "mean_token_accuracy": 0.1533171132206917,
      "num_tokens": 7206040.0,
      "step": 3475
    },
    {
      "entropy": 5.938131999969483,
      "epoch": 0.31567489114658925,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499428340446602,
      "loss": 5.6422,
      "mean_token_accuracy": 0.16303373575210572,
      "num_tokens": 7216344.0,
      "step": 3480
    },
    {
      "entropy": 5.862201881408692,
      "epoch": 0.31612844702467346,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994260330942851,
      "loss": 5.7168,
      "mean_token_accuracy": 0.1532277211546898,
      "num_tokens": 7226729.0,
      "step": 3485
    },
    {
      "entropy": 5.746544456481933,
      "epoch": 0.3165820029027576,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994237211007808,
      "loss": 5.6695,
      "mean_token_accuracy": 0.15737406313419341,
      "num_tokens": 7237616.0,
      "step": 3490
    },
    {
      "entropy": 5.85200867652893,
      "epoch": 0.3170355587808418,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994214044661367,
      "loss": 5.5787,
      "mean_token_accuracy": 0.16485872715711594,
      "num_tokens": 7247857.0,
      "step": 3495
    },
    {
      "entropy": 5.749833250045777,
      "epoch": 0.317489114658926,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994190831904007,
      "loss": 5.5717,
      "mean_token_accuracy": 0.16816714257001877,
      "num_tokens": 7257858.0,
      "step": 3500
    },
    {
      "entropy": 5.678181076049805,
      "epoch": 0.31794267053701014,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499416757273621,
      "loss": 5.588,
      "mean_token_accuracy": 0.1626364231109619,
      "num_tokens": 7268440.0,
      "step": 3505
    },
    {
      "entropy": 5.893909215927124,
      "epoch": 0.31839622641509435,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994144267158454,
      "loss": 5.6453,
      "mean_token_accuracy": 0.1645614832639694,
      "num_tokens": 7278560.0,
      "step": 3510
    },
    {
      "entropy": 5.826103019714355,
      "epoch": 0.3188497822931785,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994120915171224,
      "loss": 5.6564,
      "mean_token_accuracy": 0.16230306029319763,
      "num_tokens": 7288422.0,
      "step": 3515
    },
    {
      "entropy": 5.89535002708435,
      "epoch": 0.3193033381712627,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994097516775002,
      "loss": 5.7283,
      "mean_token_accuracy": 0.15988563597202302,
      "num_tokens": 7298613.0,
      "step": 3520
    },
    {
      "entropy": 5.8722318649292,
      "epoch": 0.3197568940493469,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994074071970271,
      "loss": 5.7538,
      "mean_token_accuracy": 0.15464695021510125,
      "num_tokens": 7309509.0,
      "step": 3525
    },
    {
      "entropy": 5.882004261016846,
      "epoch": 0.3202104499274311,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994050580757515,
      "loss": 5.6563,
      "mean_token_accuracy": 0.15909888744354247,
      "num_tokens": 7319015.0,
      "step": 3530
    },
    {
      "entropy": 5.759677457809448,
      "epoch": 0.32066400580551524,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994027043137223,
      "loss": 5.6356,
      "mean_token_accuracy": 0.1678672254085541,
      "num_tokens": 7329026.0,
      "step": 3535
    },
    {
      "entropy": 5.873507308959961,
      "epoch": 0.3211175616835994,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994003459109879,
      "loss": 5.7247,
      "mean_token_accuracy": 0.1562524437904358,
      "num_tokens": 7339638.0,
      "step": 3540
    },
    {
      "entropy": 5.856282138824463,
      "epoch": 0.3215711175616836,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993979828675972,
      "loss": 5.6467,
      "mean_token_accuracy": 0.1605513334274292,
      "num_tokens": 7349979.0,
      "step": 3545
    },
    {
      "entropy": 5.731431865692139,
      "epoch": 0.32202467343976776,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499395615183599,
      "loss": 5.5759,
      "mean_token_accuracy": 0.1669188141822815,
      "num_tokens": 7359398.0,
      "step": 3550
    },
    {
      "entropy": 5.755496263504028,
      "epoch": 0.32247822931785197,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993932428590422,
      "loss": 5.6178,
      "mean_token_accuracy": 0.16151758432388305,
      "num_tokens": 7370418.0,
      "step": 3555
    },
    {
      "entropy": 5.8418971538543705,
      "epoch": 0.3229317851959361,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993908658939762,
      "loss": 5.643,
      "mean_token_accuracy": 0.15817937552928923,
      "num_tokens": 7380461.0,
      "step": 3560
    },
    {
      "entropy": 5.819117832183838,
      "epoch": 0.32338534107402034,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993884842884496,
      "loss": 5.656,
      "mean_token_accuracy": 0.16232847571372985,
      "num_tokens": 7390646.0,
      "step": 3565
    },
    {
      "entropy": 5.789481115341187,
      "epoch": 0.3238388969521045,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993860980425121,
      "loss": 5.7402,
      "mean_token_accuracy": 0.16244688481092454,
      "num_tokens": 7402106.0,
      "step": 3570
    },
    {
      "entropy": 5.829900217056275,
      "epoch": 0.3242924528301887,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993837071562131,
      "loss": 5.7286,
      "mean_token_accuracy": 0.15734863579273223,
      "num_tokens": 7412936.0,
      "step": 3575
    },
    {
      "entropy": 5.834792232513427,
      "epoch": 0.32474600870827286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993813116296017,
      "loss": 5.5932,
      "mean_token_accuracy": 0.1646380230784416,
      "num_tokens": 7422303.0,
      "step": 3580
    },
    {
      "entropy": 5.787908029556275,
      "epoch": 0.325199564586357,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993789114627275,
      "loss": 5.6734,
      "mean_token_accuracy": 0.1588895797729492,
      "num_tokens": 7432829.0,
      "step": 3585
    },
    {
      "entropy": 5.904963636398316,
      "epoch": 0.3256531204644412,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993765066556403,
      "loss": 5.6859,
      "mean_token_accuracy": 0.15745299607515334,
      "num_tokens": 7443218.0,
      "step": 3590
    },
    {
      "entropy": 5.824350261688233,
      "epoch": 0.3261066763425254,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993740972083899,
      "loss": 5.6363,
      "mean_token_accuracy": 0.16018294245004655,
      "num_tokens": 7454531.0,
      "step": 3595
    },
    {
      "entropy": 5.835407304763794,
      "epoch": 0.3265602322206096,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993716831210259,
      "loss": 5.7004,
      "mean_token_accuracy": 0.16122788041830063,
      "num_tokens": 7464993.0,
      "step": 3600
    },
    {
      "entropy": 5.760484457015991,
      "epoch": 0.32701378809869375,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993692643935983,
      "loss": 5.7176,
      "mean_token_accuracy": 0.15775671675801278,
      "num_tokens": 7475755.0,
      "step": 3605
    },
    {
      "entropy": 5.829902124404907,
      "epoch": 0.32746734397677796,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993668410261571,
      "loss": 5.7147,
      "mean_token_accuracy": 0.16331727802753448,
      "num_tokens": 7486778.0,
      "step": 3610
    },
    {
      "entropy": 5.919011974334717,
      "epoch": 0.3279208998548621,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993644130187525,
      "loss": 5.6816,
      "mean_token_accuracy": 0.16154048442840577,
      "num_tokens": 7497439.0,
      "step": 3615
    },
    {
      "entropy": 5.813072729110718,
      "epoch": 0.3283744557329463,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993619803714347,
      "loss": 5.7209,
      "mean_token_accuracy": 0.15986181199550628,
      "num_tokens": 7508274.0,
      "step": 3620
    },
    {
      "entropy": 5.799656772613526,
      "epoch": 0.3288280116110305,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993595430842539,
      "loss": 5.6925,
      "mean_token_accuracy": 0.15679641515016557,
      "num_tokens": 7519315.0,
      "step": 3625
    },
    {
      "entropy": 5.926063013076782,
      "epoch": 0.32928156748911463,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993571011572606,
      "loss": 5.6989,
      "mean_token_accuracy": 0.16154777109622956,
      "num_tokens": 7529910.0,
      "step": 3630
    },
    {
      "entropy": 5.83719687461853,
      "epoch": 0.32973512336719885,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993546545905053,
      "loss": 5.6263,
      "mean_token_accuracy": 0.16304632276296616,
      "num_tokens": 7538966.0,
      "step": 3635
    },
    {
      "entropy": 5.795618629455566,
      "epoch": 0.330188679245283,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993522033840386,
      "loss": 5.6618,
      "mean_token_accuracy": 0.1593761846423149,
      "num_tokens": 7549683.0,
      "step": 3640
    },
    {
      "entropy": 5.746255350112915,
      "epoch": 0.3306422351233672,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499349747537911,
      "loss": 5.6375,
      "mean_token_accuracy": 0.1627708211541176,
      "num_tokens": 7560629.0,
      "step": 3645
    },
    {
      "entropy": 5.824515104293823,
      "epoch": 0.33109579100145137,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993472870521736,
      "loss": 5.5937,
      "mean_token_accuracy": 0.16678274869918824,
      "num_tokens": 7570613.0,
      "step": 3650
    },
    {
      "entropy": 5.780102157592774,
      "epoch": 0.3315493468795356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993448219268771,
      "loss": 5.5616,
      "mean_token_accuracy": 0.16153461784124373,
      "num_tokens": 7580640.0,
      "step": 3655
    },
    {
      "entropy": 5.7738007545471195,
      "epoch": 0.33200290275761973,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993423521620726,
      "loss": 5.6566,
      "mean_token_accuracy": 0.16400932371616364,
      "num_tokens": 7590164.0,
      "step": 3660
    },
    {
      "entropy": 5.878917837142945,
      "epoch": 0.33245645863570394,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993398777578109,
      "loss": 5.7057,
      "mean_token_accuracy": 0.15703917518258095,
      "num_tokens": 7600173.0,
      "step": 3665
    },
    {
      "entropy": 5.718817949295044,
      "epoch": 0.3329100145137881,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993373987141435,
      "loss": 5.5171,
      "mean_token_accuracy": 0.16777506917715074,
      "num_tokens": 7609958.0,
      "step": 3670
    },
    {
      "entropy": 5.847012901306153,
      "epoch": 0.33336357039187225,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993349150311214,
      "loss": 5.6535,
      "mean_token_accuracy": 0.15843262523412704,
      "num_tokens": 7620237.0,
      "step": 3675
    },
    {
      "entropy": 5.759278964996338,
      "epoch": 0.33381712626995647,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993324267087962,
      "loss": 5.5564,
      "mean_token_accuracy": 0.16717470437288284,
      "num_tokens": 7629886.0,
      "step": 3680
    },
    {
      "entropy": 5.757597494125366,
      "epoch": 0.3342706821480406,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993299337472192,
      "loss": 5.6311,
      "mean_token_accuracy": 0.16425763219594955,
      "num_tokens": 7640057.0,
      "step": 3685
    },
    {
      "entropy": 5.725239229202271,
      "epoch": 0.33472423802612483,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499327436146442,
      "loss": 5.5767,
      "mean_token_accuracy": 0.16163882464170456,
      "num_tokens": 7649766.0,
      "step": 3690
    },
    {
      "entropy": 5.778099012374878,
      "epoch": 0.335177793904209,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993249339065162,
      "loss": 5.6323,
      "mean_token_accuracy": 0.16390450447797775,
      "num_tokens": 7660890.0,
      "step": 3695
    },
    {
      "entropy": 5.83032078742981,
      "epoch": 0.3356313497822932,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993224270274937,
      "loss": 5.6562,
      "mean_token_accuracy": 0.16339348554611205,
      "num_tokens": 7670737.0,
      "step": 3700
    },
    {
      "entropy": 5.714012575149536,
      "epoch": 0.33608490566037735,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993199155094263,
      "loss": 5.6157,
      "mean_token_accuracy": 0.16412198096513747,
      "num_tokens": 7681407.0,
      "step": 3705
    },
    {
      "entropy": 5.795091533660889,
      "epoch": 0.33653846153846156,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993173993523658,
      "loss": 5.5506,
      "mean_token_accuracy": 0.1710594192147255,
      "num_tokens": 7691758.0,
      "step": 3710
    },
    {
      "entropy": 5.689993810653687,
      "epoch": 0.3369920174165457,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993148785563643,
      "loss": 5.5826,
      "mean_token_accuracy": 0.1621573567390442,
      "num_tokens": 7702934.0,
      "step": 3715
    },
    {
      "entropy": 5.717300510406494,
      "epoch": 0.3374455732946299,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499312353121474,
      "loss": 5.5771,
      "mean_token_accuracy": 0.1630354642868042,
      "num_tokens": 7713001.0,
      "step": 3720
    },
    {
      "entropy": 5.836460018157959,
      "epoch": 0.3378991291727141,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993098230477471,
      "loss": 5.6475,
      "mean_token_accuracy": 0.15897637307643891,
      "num_tokens": 7724005.0,
      "step": 3725
    },
    {
      "entropy": 5.8066576480865475,
      "epoch": 0.33835268505079824,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993072883352358,
      "loss": 5.6038,
      "mean_token_accuracy": 0.1632534846663475,
      "num_tokens": 7733743.0,
      "step": 3730
    },
    {
      "entropy": 5.836714839935302,
      "epoch": 0.33880624092888245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993047489839927,
      "loss": 5.7837,
      "mean_token_accuracy": 0.15582592636346818,
      "num_tokens": 7743762.0,
      "step": 3735
    },
    {
      "entropy": 5.886009645462036,
      "epoch": 0.3392597968069666,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993022049940702,
      "loss": 5.666,
      "mean_token_accuracy": 0.15896076560020447,
      "num_tokens": 7753758.0,
      "step": 3740
    },
    {
      "entropy": 5.778275203704834,
      "epoch": 0.3397133526850508,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992996563655209,
      "loss": 5.6211,
      "mean_token_accuracy": 0.16299140006303786,
      "num_tokens": 7764692.0,
      "step": 3745
    },
    {
      "entropy": 5.749185180664062,
      "epoch": 0.340166908563135,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992971030983975,
      "loss": 5.6453,
      "mean_token_accuracy": 0.16236765682697296,
      "num_tokens": 7776892.0,
      "step": 3750
    },
    {
      "entropy": 5.860137844085694,
      "epoch": 0.3406204644412192,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992945451927531,
      "loss": 5.7017,
      "mean_token_accuracy": 0.15955468267202377,
      "num_tokens": 7787143.0,
      "step": 3755
    },
    {
      "entropy": 5.734303283691406,
      "epoch": 0.34107402031930334,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992919826486402,
      "loss": 5.6101,
      "mean_token_accuracy": 0.16399556547403335,
      "num_tokens": 7798537.0,
      "step": 3760
    },
    {
      "entropy": 5.734641885757446,
      "epoch": 0.3415275761973875,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992894154661118,
      "loss": 5.5799,
      "mean_token_accuracy": 0.16684457361698152,
      "num_tokens": 7808434.0,
      "step": 3765
    },
    {
      "entropy": 5.806569623947143,
      "epoch": 0.3419811320754717,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992868436452213,
      "loss": 5.6955,
      "mean_token_accuracy": 0.15859481543302537,
      "num_tokens": 7819592.0,
      "step": 3770
    },
    {
      "entropy": 5.850445795059204,
      "epoch": 0.34243468795355586,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992842671860217,
      "loss": 5.5788,
      "mean_token_accuracy": 0.16591217517852783,
      "num_tokens": 7830602.0,
      "step": 3775
    },
    {
      "entropy": 5.835497093200684,
      "epoch": 0.34288824383164007,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992816860885662,
      "loss": 5.735,
      "mean_token_accuracy": 0.15822505801916123,
      "num_tokens": 7840850.0,
      "step": 3780
    },
    {
      "entropy": 5.917944717407226,
      "epoch": 0.3433417997097242,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992791003529084,
      "loss": 5.6846,
      "mean_token_accuracy": 0.15368564128875734,
      "num_tokens": 7852371.0,
      "step": 3785
    },
    {
      "entropy": 5.72100830078125,
      "epoch": 0.34379535558780844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992765099791017,
      "loss": 5.6964,
      "mean_token_accuracy": 0.15722172409296037,
      "num_tokens": 7863290.0,
      "step": 3790
    },
    {
      "entropy": 5.7610937595367435,
      "epoch": 0.3442489114658926,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992739149671995,
      "loss": 5.5333,
      "mean_token_accuracy": 0.1643339917063713,
      "num_tokens": 7873927.0,
      "step": 3795
    },
    {
      "entropy": 5.7489160060882565,
      "epoch": 0.3447024673439768,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992713153172556,
      "loss": 5.5691,
      "mean_token_accuracy": 0.1632430613040924,
      "num_tokens": 7884977.0,
      "step": 3800
    },
    {
      "entropy": 5.745196342468262,
      "epoch": 0.34515602322206096,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992687110293238,
      "loss": 5.6531,
      "mean_token_accuracy": 0.16766562685370445,
      "num_tokens": 7894620.0,
      "step": 3805
    },
    {
      "entropy": 5.78018045425415,
      "epoch": 0.3456095791001451,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992661021034579,
      "loss": 5.5848,
      "mean_token_accuracy": 0.16409669667482377,
      "num_tokens": 7904541.0,
      "step": 3810
    },
    {
      "entropy": 5.75669412612915,
      "epoch": 0.3460631349782293,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992634885397118,
      "loss": 5.6036,
      "mean_token_accuracy": 0.1641060397028923,
      "num_tokens": 7915435.0,
      "step": 3815
    },
    {
      "entropy": 5.751991319656372,
      "epoch": 0.3465166908563135,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992608703381396,
      "loss": 5.5681,
      "mean_token_accuracy": 0.1672964349389076,
      "num_tokens": 7925870.0,
      "step": 3820
    },
    {
      "entropy": 5.787299251556396,
      "epoch": 0.3469702467343977,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992582474987955,
      "loss": 5.5792,
      "mean_token_accuracy": 0.16019600182771682,
      "num_tokens": 7937046.0,
      "step": 3825
    },
    {
      "entropy": 5.747065591812134,
      "epoch": 0.34742380261248185,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992556200217337,
      "loss": 5.5366,
      "mean_token_accuracy": 0.1621662676334381,
      "num_tokens": 7947150.0,
      "step": 3830
    },
    {
      "entropy": 5.782345485687256,
      "epoch": 0.34787735849056606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992529879070085,
      "loss": 5.6463,
      "mean_token_accuracy": 0.16273225396871566,
      "num_tokens": 7956815.0,
      "step": 3835
    },
    {
      "entropy": 5.83603720664978,
      "epoch": 0.3483309143686502,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992503511546745,
      "loss": 5.6669,
      "mean_token_accuracy": 0.16219530552625655,
      "num_tokens": 7968114.0,
      "step": 3840
    },
    {
      "entropy": 5.786204385757446,
      "epoch": 0.3487844702467344,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992477097647859,
      "loss": 5.6001,
      "mean_token_accuracy": 0.15802547633647918,
      "num_tokens": 7977840.0,
      "step": 3845
    },
    {
      "entropy": 5.839146471023559,
      "epoch": 0.3492380261248186,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992450637373976,
      "loss": 5.7159,
      "mean_token_accuracy": 0.16593586057424545,
      "num_tokens": 7988663.0,
      "step": 3850
    },
    {
      "entropy": 5.7088793277740475,
      "epoch": 0.34969158200290273,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992424130725642,
      "loss": 5.5105,
      "mean_token_accuracy": 0.1633152462542057,
      "num_tokens": 7999693.0,
      "step": 3855
    },
    {
      "entropy": 5.813177061080933,
      "epoch": 0.35014513788098695,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992397577703406,
      "loss": 5.7066,
      "mean_token_accuracy": 0.15328351110219957,
      "num_tokens": 8010154.0,
      "step": 3860
    },
    {
      "entropy": 5.895718193054199,
      "epoch": 0.3505986937590711,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992370978307816,
      "loss": 5.6764,
      "mean_token_accuracy": 0.15621304959058763,
      "num_tokens": 8020921.0,
      "step": 3865
    },
    {
      "entropy": 5.780423545837403,
      "epoch": 0.3510522496371553,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992344332539424,
      "loss": 5.6006,
      "mean_token_accuracy": 0.16626923233270646,
      "num_tokens": 8031404.0,
      "step": 3870
    },
    {
      "entropy": 5.6891073226928714,
      "epoch": 0.35150580551523947,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992317640398779,
      "loss": 5.6086,
      "mean_token_accuracy": 0.16590815931558608,
      "num_tokens": 8041310.0,
      "step": 3875
    },
    {
      "entropy": 5.7839179039001465,
      "epoch": 0.3519593613933237,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992290901886434,
      "loss": 5.5748,
      "mean_token_accuracy": 0.16641812920570373,
      "num_tokens": 8051054.0,
      "step": 3880
    },
    {
      "entropy": 5.776237964630127,
      "epoch": 0.35241291727140783,
      "grad_norm": 1.0,
      "learning_rate": 0.000499226411700294,
      "loss": 5.577,
      "mean_token_accuracy": 0.16576359570026397,
      "num_tokens": 8061063.0,
      "step": 3885
    },
    {
      "entropy": 5.721409654617309,
      "epoch": 0.35286647314949204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992237285748855,
      "loss": 5.6016,
      "mean_token_accuracy": 0.16601926535367967,
      "num_tokens": 8071715.0,
      "step": 3890
    },
    {
      "entropy": 5.865892457962036,
      "epoch": 0.3533200290275762,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992210408124731,
      "loss": 5.6575,
      "mean_token_accuracy": 0.1572970688343048,
      "num_tokens": 8082679.0,
      "step": 3895
    },
    {
      "entropy": 5.7754655361175535,
      "epoch": 0.35377358490566035,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992183484131123,
      "loss": 5.6963,
      "mean_token_accuracy": 0.15824397653341293,
      "num_tokens": 8093785.0,
      "step": 3900
    },
    {
      "entropy": 5.733349323272705,
      "epoch": 0.35422714078374457,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992156513768591,
      "loss": 5.5862,
      "mean_token_accuracy": 0.16918893307447433,
      "num_tokens": 8105822.0,
      "step": 3905
    },
    {
      "entropy": 5.746071147918701,
      "epoch": 0.3546806966618287,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499212949703769,
      "loss": 5.5009,
      "mean_token_accuracy": 0.16611845195293426,
      "num_tokens": 8115521.0,
      "step": 3910
    },
    {
      "entropy": 5.764930868148804,
      "epoch": 0.35513425253991293,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992102433938981,
      "loss": 5.5949,
      "mean_token_accuracy": 0.1668561689555645,
      "num_tokens": 8125524.0,
      "step": 3915
    },
    {
      "entropy": 5.734392976760864,
      "epoch": 0.3555878084179971,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992075324473021,
      "loss": 5.5131,
      "mean_token_accuracy": 0.17593824863433838,
      "num_tokens": 8135552.0,
      "step": 3920
    },
    {
      "entropy": 5.831735563278198,
      "epoch": 0.3560413642960813,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992048168640371,
      "loss": 5.6308,
      "mean_token_accuracy": 0.1653963729739189,
      "num_tokens": 8145617.0,
      "step": 3925
    },
    {
      "entropy": 5.727599000930786,
      "epoch": 0.35649492017416545,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992020966441595,
      "loss": 5.6406,
      "mean_token_accuracy": 0.1605748012661934,
      "num_tokens": 8156729.0,
      "step": 3930
    },
    {
      "entropy": 5.772192573547363,
      "epoch": 0.35694847605224966,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991993717877254,
      "loss": 5.6209,
      "mean_token_accuracy": 0.1616794392466545,
      "num_tokens": 8167637.0,
      "step": 3935
    },
    {
      "entropy": 5.763149976730347,
      "epoch": 0.3574020319303338,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991966422947911,
      "loss": 5.6039,
      "mean_token_accuracy": 0.1692925289273262,
      "num_tokens": 8177354.0,
      "step": 3940
    },
    {
      "entropy": 5.721889448165894,
      "epoch": 0.357855587808418,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991939081654132,
      "loss": 5.5608,
      "mean_token_accuracy": 0.17194263488054276,
      "num_tokens": 8187580.0,
      "step": 3945
    },
    {
      "entropy": 5.723434352874756,
      "epoch": 0.3583091436865022,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499191169399648,
      "loss": 5.6598,
      "mean_token_accuracy": 0.16233123391866683,
      "num_tokens": 8199382.0,
      "step": 3950
    },
    {
      "entropy": 5.781085586547851,
      "epoch": 0.35876269956458634,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991884259975525,
      "loss": 5.5557,
      "mean_token_accuracy": 0.16617808789014815,
      "num_tokens": 8208484.0,
      "step": 3955
    },
    {
      "entropy": 5.797117471694946,
      "epoch": 0.35921625544267055,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499185677959183,
      "loss": 5.6052,
      "mean_token_accuracy": 0.15923126488924028,
      "num_tokens": 8219056.0,
      "step": 3960
    },
    {
      "entropy": 5.777331686019897,
      "epoch": 0.3596698113207547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991829252845966,
      "loss": 5.6232,
      "mean_token_accuracy": 0.1578235447406769,
      "num_tokens": 8230149.0,
      "step": 3965
    },
    {
      "entropy": 5.772022676467896,
      "epoch": 0.3601233671988389,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991801679738502,
      "loss": 5.6806,
      "mean_token_accuracy": 0.15451672971248626,
      "num_tokens": 8240803.0,
      "step": 3970
    },
    {
      "entropy": 5.766995000839233,
      "epoch": 0.3605769230769231,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991774060270008,
      "loss": 5.632,
      "mean_token_accuracy": 0.1565280094742775,
      "num_tokens": 8251563.0,
      "step": 3975
    },
    {
      "entropy": 5.758525943756103,
      "epoch": 0.3610304789550073,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991746394441055,
      "loss": 5.5818,
      "mean_token_accuracy": 0.15869568139314652,
      "num_tokens": 8261342.0,
      "step": 3980
    },
    {
      "entropy": 5.76960334777832,
      "epoch": 0.36148403483309144,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991718682252216,
      "loss": 5.6749,
      "mean_token_accuracy": 0.15879870802164078,
      "num_tokens": 8271540.0,
      "step": 3985
    },
    {
      "entropy": 5.807124662399292,
      "epoch": 0.3619375907111756,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991690923704063,
      "loss": 5.6018,
      "mean_token_accuracy": 0.16289105713367463,
      "num_tokens": 8280593.0,
      "step": 3990
    },
    {
      "entropy": 5.7926068782806395,
      "epoch": 0.3623911465892598,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991663118797171,
      "loss": 5.6189,
      "mean_token_accuracy": 0.16317871809005738,
      "num_tokens": 8290126.0,
      "step": 3995
    },
    {
      "entropy": 5.760343790054321,
      "epoch": 0.36284470246734396,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991635267532113,
      "loss": 5.542,
      "mean_token_accuracy": 0.16164259910583495,
      "num_tokens": 8301416.0,
      "step": 4000
    },
    {
      "entropy": 5.794639825820923,
      "epoch": 0.36329825834542817,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991607369909468,
      "loss": 5.6311,
      "mean_token_accuracy": 0.15973753184080125,
      "num_tokens": 8311034.0,
      "step": 4005
    },
    {
      "entropy": 5.767417287826538,
      "epoch": 0.3637518142235123,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991579425929811,
      "loss": 5.6506,
      "mean_token_accuracy": 0.16089941710233688,
      "num_tokens": 8322203.0,
      "step": 4010
    },
    {
      "entropy": 5.765074586868286,
      "epoch": 0.36420537010159654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499155143559372,
      "loss": 5.5372,
      "mean_token_accuracy": 0.1657981812953949,
      "num_tokens": 8330981.0,
      "step": 4015
    },
    {
      "entropy": 5.65937442779541,
      "epoch": 0.3646589259796807,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004991523398901773,
      "loss": 5.5613,
      "mean_token_accuracy": 0.15952345728874207,
      "num_tokens": 8341294.0,
      "step": 4020
    },
    {
      "entropy": 5.869492626190185,
      "epoch": 0.3651124818577649,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991495315854553,
      "loss": 5.7122,
      "mean_token_accuracy": 0.15741421282291412,
      "num_tokens": 8351824.0,
      "step": 4025
    },
    {
      "entropy": 5.71394624710083,
      "epoch": 0.36556603773584906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991467186452637,
      "loss": 5.5716,
      "mean_token_accuracy": 0.16435191184282302,
      "num_tokens": 8361975.0,
      "step": 4030
    },
    {
      "entropy": 5.745710849761963,
      "epoch": 0.3660195936139332,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991439010696609,
      "loss": 5.5203,
      "mean_token_accuracy": 0.16883532404899598,
      "num_tokens": 8372145.0,
      "step": 4035
    },
    {
      "entropy": 5.778649616241455,
      "epoch": 0.3664731494920174,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991410788587051,
      "loss": 5.6151,
      "mean_token_accuracy": 0.159750834107399,
      "num_tokens": 8382542.0,
      "step": 4040
    },
    {
      "entropy": 5.749556159973144,
      "epoch": 0.3669267053701016,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991382520124547,
      "loss": 5.6097,
      "mean_token_accuracy": 0.16143932342529296,
      "num_tokens": 8393450.0,
      "step": 4045
    },
    {
      "entropy": 5.771503353118897,
      "epoch": 0.3673802612481858,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991354205309681,
      "loss": 5.5472,
      "mean_token_accuracy": 0.16442203521728516,
      "num_tokens": 8404929.0,
      "step": 4050
    },
    {
      "entropy": 5.770533847808838,
      "epoch": 0.36783381712626995,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499132584414304,
      "loss": 5.5662,
      "mean_token_accuracy": 0.1639400228857994,
      "num_tokens": 8415703.0,
      "step": 4055
    },
    {
      "entropy": 5.767830038070679,
      "epoch": 0.36828737300435416,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991297436625207,
      "loss": 5.6591,
      "mean_token_accuracy": 0.1587003231048584,
      "num_tokens": 8426220.0,
      "step": 4060
    },
    {
      "entropy": 5.680564117431641,
      "epoch": 0.3687409288824383,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991268982756774,
      "loss": 5.5295,
      "mean_token_accuracy": 0.1664980798959732,
      "num_tokens": 8436220.0,
      "step": 4065
    },
    {
      "entropy": 5.850871753692627,
      "epoch": 0.3691944847605225,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991240482538327,
      "loss": 5.6776,
      "mean_token_accuracy": 0.15412717759609224,
      "num_tokens": 8446724.0,
      "step": 4070
    },
    {
      "entropy": 5.742097663879394,
      "epoch": 0.3696480406386067,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991211935970455,
      "loss": 5.5408,
      "mean_token_accuracy": 0.16396154165267945,
      "num_tokens": 8457270.0,
      "step": 4075
    },
    {
      "entropy": 5.729263162612915,
      "epoch": 0.37010159651669083,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991183343053748,
      "loss": 5.6115,
      "mean_token_accuracy": 0.15851412415504457,
      "num_tokens": 8468137.0,
      "step": 4080
    },
    {
      "entropy": 5.771952438354492,
      "epoch": 0.37055515239477504,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049911547037888,
      "loss": 5.5452,
      "mean_token_accuracy": 0.1679990693926811,
      "num_tokens": 8477574.0,
      "step": 4085
    },
    {
      "entropy": 5.742566680908203,
      "epoch": 0.3710087082728592,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991126018176202,
      "loss": 5.6413,
      "mean_token_accuracy": 0.1597900688648224,
      "num_tokens": 8487268.0,
      "step": 4090
    },
    {
      "entropy": 5.877599859237671,
      "epoch": 0.3714622641509434,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991097286216545,
      "loss": 5.7002,
      "mean_token_accuracy": 0.1654977947473526,
      "num_tokens": 8497785.0,
      "step": 4095
    },
    {
      "entropy": 5.80135407447815,
      "epoch": 0.37191582002902757,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991068507910427,
      "loss": 5.5666,
      "mean_token_accuracy": 0.16508724838495253,
      "num_tokens": 8508169.0,
      "step": 4100
    },
    {
      "entropy": 5.709864664077759,
      "epoch": 0.3723693759071118,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499103968325844,
      "loss": 5.5366,
      "mean_token_accuracy": 0.16589654237031937,
      "num_tokens": 8519201.0,
      "step": 4105
    },
    {
      "entropy": 5.795285272598266,
      "epoch": 0.37282293178519593,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991010812261181,
      "loss": 5.6332,
      "mean_token_accuracy": 0.1585421934723854,
      "num_tokens": 8531093.0,
      "step": 4110
    },
    {
      "entropy": 5.659151411056518,
      "epoch": 0.37327648766328014,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990981894919248,
      "loss": 5.5081,
      "mean_token_accuracy": 0.16795153617858888,
      "num_tokens": 8541730.0,
      "step": 4115
    },
    {
      "entropy": 5.7505875587463375,
      "epoch": 0.3737300435413643,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990952931233238,
      "loss": 5.588,
      "mean_token_accuracy": 0.16281794607639313,
      "num_tokens": 8552570.0,
      "step": 4120
    },
    {
      "entropy": 5.799954271316528,
      "epoch": 0.37418359941944845,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499092392120375,
      "loss": 5.4892,
      "mean_token_accuracy": 0.1670244187116623,
      "num_tokens": 8562340.0,
      "step": 4125
    },
    {
      "entropy": 5.737644052505493,
      "epoch": 0.37463715529753266,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990894864831385,
      "loss": 5.6899,
      "mean_token_accuracy": 0.15905696153640747,
      "num_tokens": 8573240.0,
      "step": 4130
    },
    {
      "entropy": 5.830232810974121,
      "epoch": 0.3750907111756168,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990865762116743,
      "loss": 5.5424,
      "mean_token_accuracy": 0.16745613217353822,
      "num_tokens": 8583194.0,
      "step": 4135
    },
    {
      "entropy": 5.733665180206299,
      "epoch": 0.37554426705370103,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990836613060426,
      "loss": 5.6071,
      "mean_token_accuracy": 0.16095204204320906,
      "num_tokens": 8593603.0,
      "step": 4140
    },
    {
      "entropy": 5.6964035987854,
      "epoch": 0.3759978229317852,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990807417663036,
      "loss": 5.6133,
      "mean_token_accuracy": 0.1622975468635559,
      "num_tokens": 8603176.0,
      "step": 4145
    },
    {
      "entropy": 5.8237403392791744,
      "epoch": 0.3764513788098694,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990778175925179,
      "loss": 5.6325,
      "mean_token_accuracy": 0.16655509024858475,
      "num_tokens": 8614801.0,
      "step": 4150
    },
    {
      "entropy": 5.793196535110473,
      "epoch": 0.37690493468795355,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990748887847457,
      "loss": 5.6557,
      "mean_token_accuracy": 0.1599831074476242,
      "num_tokens": 8624999.0,
      "step": 4155
    },
    {
      "entropy": 5.700607967376709,
      "epoch": 0.37735849056603776,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990719553430477,
      "loss": 5.507,
      "mean_token_accuracy": 0.17496906369924545,
      "num_tokens": 8635358.0,
      "step": 4160
    },
    {
      "entropy": 5.739441013336181,
      "epoch": 0.3778120464441219,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990690172674847,
      "loss": 5.5865,
      "mean_token_accuracy": 0.16173218041658402,
      "num_tokens": 8645430.0,
      "step": 4165
    },
    {
      "entropy": 5.801822137832642,
      "epoch": 0.3782656023222061,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990660745581172,
      "loss": 5.5416,
      "mean_token_accuracy": 0.1737535923719406,
      "num_tokens": 8655115.0,
      "step": 4170
    },
    {
      "entropy": 5.693522024154663,
      "epoch": 0.3787191582002903,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990631272150062,
      "loss": 5.5394,
      "mean_token_accuracy": 0.16801448911428452,
      "num_tokens": 8665391.0,
      "step": 4175
    },
    {
      "entropy": 5.767375373840332,
      "epoch": 0.37917271407837444,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990601752382127,
      "loss": 5.632,
      "mean_token_accuracy": 0.1633898064494133,
      "num_tokens": 8676819.0,
      "step": 4180
    },
    {
      "entropy": 5.8388364791870115,
      "epoch": 0.37962626995645865,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990572186277975,
      "loss": 5.5607,
      "mean_token_accuracy": 0.16711350232362748,
      "num_tokens": 8687052.0,
      "step": 4185
    },
    {
      "entropy": 5.7258948802948,
      "epoch": 0.3800798258345428,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990542573838221,
      "loss": 5.6467,
      "mean_token_accuracy": 0.1595759868621826,
      "num_tokens": 8697756.0,
      "step": 4190
    },
    {
      "entropy": 5.6451514720916744,
      "epoch": 0.380533381712627,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990512915063475,
      "loss": 5.5475,
      "mean_token_accuracy": 0.17015636712312698,
      "num_tokens": 8708388.0,
      "step": 4195
    },
    {
      "entropy": 5.812650442123413,
      "epoch": 0.38098693759071117,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990483209954352,
      "loss": 5.6359,
      "mean_token_accuracy": 0.1576849937438965,
      "num_tokens": 8719475.0,
      "step": 4200
    },
    {
      "entropy": 5.7326311588287355,
      "epoch": 0.3814404934687954,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990453458511464,
      "loss": 5.4986,
      "mean_token_accuracy": 0.17547574192285537,
      "num_tokens": 8729395.0,
      "step": 4205
    },
    {
      "entropy": 5.758198976516724,
      "epoch": 0.38189404934687954,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990423660735428,
      "loss": 5.5483,
      "mean_token_accuracy": 0.16976142972707747,
      "num_tokens": 8740351.0,
      "step": 4210
    },
    {
      "entropy": 5.7952666759490965,
      "epoch": 0.3823476052249637,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499039381662686,
      "loss": 5.5672,
      "mean_token_accuracy": 0.16848132163286209,
      "num_tokens": 8749260.0,
      "step": 4215
    },
    {
      "entropy": 5.78771185874939,
      "epoch": 0.3828011611030479,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990363926186377,
      "loss": 5.666,
      "mean_token_accuracy": 0.15561919510364533,
      "num_tokens": 8759604.0,
      "step": 4220
    },
    {
      "entropy": 5.8073570251464846,
      "epoch": 0.38325471698113206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990333989414597,
      "loss": 5.6851,
      "mean_token_accuracy": 0.155071659386158,
      "num_tokens": 8769815.0,
      "step": 4225
    },
    {
      "entropy": 5.687266063690186,
      "epoch": 0.38370827285921627,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990304006312138,
      "loss": 5.6015,
      "mean_token_accuracy": 0.16185775846242906,
      "num_tokens": 8780557.0,
      "step": 4230
    },
    {
      "entropy": 5.777766180038452,
      "epoch": 0.3841618287373004,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990273976879622,
      "loss": 5.5629,
      "mean_token_accuracy": 0.17335893511772155,
      "num_tokens": 8790993.0,
      "step": 4235
    },
    {
      "entropy": 5.768019533157348,
      "epoch": 0.38461538461538464,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990243901117669,
      "loss": 5.6234,
      "mean_token_accuracy": 0.16034475564956666,
      "num_tokens": 8800657.0,
      "step": 4240
    },
    {
      "entropy": 5.67515926361084,
      "epoch": 0.3850689404934688,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990213779026903,
      "loss": 5.6138,
      "mean_token_accuracy": 0.15936714857816697,
      "num_tokens": 8812983.0,
      "step": 4245
    },
    {
      "entropy": 5.800050735473633,
      "epoch": 0.385522496371553,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990183610607944,
      "loss": 5.5895,
      "mean_token_accuracy": 0.16557528376579284,
      "num_tokens": 8822980.0,
      "step": 4250
    },
    {
      "entropy": 5.733969593048096,
      "epoch": 0.38597605224963716,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990153395861415,
      "loss": 5.5572,
      "mean_token_accuracy": 0.16649172306060792,
      "num_tokens": 8833821.0,
      "step": 4255
    },
    {
      "entropy": 5.672335815429688,
      "epoch": 0.3864296081277213,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990123134787946,
      "loss": 5.509,
      "mean_token_accuracy": 0.16983023583889006,
      "num_tokens": 8843929.0,
      "step": 4260
    },
    {
      "entropy": 5.780575275421143,
      "epoch": 0.3868831640058055,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990092827388159,
      "loss": 5.493,
      "mean_token_accuracy": 0.16734486669301987,
      "num_tokens": 8854180.0,
      "step": 4265
    },
    {
      "entropy": 5.6601338386535645,
      "epoch": 0.3873367198838897,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990062473662681,
      "loss": 5.5037,
      "mean_token_accuracy": 0.17173054963350295,
      "num_tokens": 8865018.0,
      "step": 4270
    },
    {
      "entropy": 5.781656074523926,
      "epoch": 0.3877902757619739,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499003207361214,
      "loss": 5.6868,
      "mean_token_accuracy": 0.1598363995552063,
      "num_tokens": 8875423.0,
      "step": 4275
    },
    {
      "entropy": 5.845520830154419,
      "epoch": 0.38824383164005805,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990001627237165,
      "loss": 5.6193,
      "mean_token_accuracy": 0.1609180137515068,
      "num_tokens": 8887850.0,
      "step": 4280
    },
    {
      "entropy": 5.6677868366241455,
      "epoch": 0.38869738751814226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989971134538386,
      "loss": 5.5183,
      "mean_token_accuracy": 0.16817747205495834,
      "num_tokens": 8897429.0,
      "step": 4285
    },
    {
      "entropy": 5.684417486190796,
      "epoch": 0.3891509433962264,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989940595516432,
      "loss": 5.5406,
      "mean_token_accuracy": 0.1706473186612129,
      "num_tokens": 8907048.0,
      "step": 4290
    },
    {
      "entropy": 5.759358978271484,
      "epoch": 0.3896044992743106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989910010171938,
      "loss": 5.6105,
      "mean_token_accuracy": 0.16018391102552415,
      "num_tokens": 8917987.0,
      "step": 4295
    },
    {
      "entropy": 5.771234941482544,
      "epoch": 0.3900580551523948,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004989879378505532,
      "loss": 5.4998,
      "mean_token_accuracy": 0.1641297698020935,
      "num_tokens": 8929339.0,
      "step": 4300
    },
    {
      "entropy": 5.629133081436157,
      "epoch": 0.39051161103047893,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989848700517849,
      "loss": 5.5118,
      "mean_token_accuracy": 0.1678486481308937,
      "num_tokens": 8939473.0,
      "step": 4305
    },
    {
      "entropy": 5.693001365661621,
      "epoch": 0.39096516690856314,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989817976209526,
      "loss": 5.542,
      "mean_token_accuracy": 0.16416483521461486,
      "num_tokens": 8949173.0,
      "step": 4310
    },
    {
      "entropy": 5.759200525283814,
      "epoch": 0.3914187227866473,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989787205581196,
      "loss": 5.5195,
      "mean_token_accuracy": 0.17088344991207122,
      "num_tokens": 8959879.0,
      "step": 4315
    },
    {
      "entropy": 5.704134798049926,
      "epoch": 0.3918722786647315,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989756388633496,
      "loss": 5.6041,
      "mean_token_accuracy": 0.1627824455499649,
      "num_tokens": 8970643.0,
      "step": 4320
    },
    {
      "entropy": 5.710903739929199,
      "epoch": 0.39232583454281567,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989725525367063,
      "loss": 5.4835,
      "mean_token_accuracy": 0.17103770971298218,
      "num_tokens": 8979882.0,
      "step": 4325
    },
    {
      "entropy": 5.7799781322479244,
      "epoch": 0.3927793904208999,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989694615782534,
      "loss": 5.5981,
      "mean_token_accuracy": 0.15929665714502333,
      "num_tokens": 8990925.0,
      "step": 4330
    },
    {
      "entropy": 5.762790632247925,
      "epoch": 0.39323294629898403,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989663659880552,
      "loss": 5.6236,
      "mean_token_accuracy": 0.16410700231790543,
      "num_tokens": 9001365.0,
      "step": 4335
    },
    {
      "entropy": 5.808406639099121,
      "epoch": 0.3936865021770682,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989632657661754,
      "loss": 5.6151,
      "mean_token_accuracy": 0.16846253722906113,
      "num_tokens": 9011942.0,
      "step": 4340
    },
    {
      "entropy": 5.748916530609131,
      "epoch": 0.3941400580551524,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498960160912678,
      "loss": 5.665,
      "mean_token_accuracy": 0.16369327455759047,
      "num_tokens": 9022610.0,
      "step": 4345
    },
    {
      "entropy": 5.734598922729492,
      "epoch": 0.39459361393323655,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989570514276276,
      "loss": 5.5523,
      "mean_token_accuracy": 0.16294773668050766,
      "num_tokens": 9033034.0,
      "step": 4350
    },
    {
      "entropy": 5.767314720153808,
      "epoch": 0.39504716981132076,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989539373110883,
      "loss": 5.5898,
      "mean_token_accuracy": 0.160574010014534,
      "num_tokens": 9043418.0,
      "step": 4355
    },
    {
      "entropy": 5.782724094390869,
      "epoch": 0.3955007256894049,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989508185631245,
      "loss": 5.6765,
      "mean_token_accuracy": 0.15920473337173463,
      "num_tokens": 9052708.0,
      "step": 4360
    },
    {
      "entropy": 5.673357582092285,
      "epoch": 0.39595428156748913,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989476951838006,
      "loss": 5.5266,
      "mean_token_accuracy": 0.1611687034368515,
      "num_tokens": 9062883.0,
      "step": 4365
    },
    {
      "entropy": 5.750123310089111,
      "epoch": 0.3964078374455733,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989445671731814,
      "loss": 5.4954,
      "mean_token_accuracy": 0.17254894524812697,
      "num_tokens": 9072830.0,
      "step": 4370
    },
    {
      "entropy": 5.673724603652954,
      "epoch": 0.3968613933236575,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989414345313315,
      "loss": 5.4632,
      "mean_token_accuracy": 0.16166009455919267,
      "num_tokens": 9082749.0,
      "step": 4375
    },
    {
      "entropy": 5.779737043380737,
      "epoch": 0.39731494920174165,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989382972583156,
      "loss": 5.5433,
      "mean_token_accuracy": 0.1635667696595192,
      "num_tokens": 9094188.0,
      "step": 4380
    },
    {
      "entropy": 5.701415920257569,
      "epoch": 0.3977685050798258,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989351553541986,
      "loss": 5.5767,
      "mean_token_accuracy": 0.15773947685956954,
      "num_tokens": 9105023.0,
      "step": 4385
    },
    {
      "entropy": 5.63522777557373,
      "epoch": 0.39822206095791,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989320088190456,
      "loss": 5.5136,
      "mean_token_accuracy": 0.16800401508808135,
      "num_tokens": 9115563.0,
      "step": 4390
    },
    {
      "entropy": 5.710983753204346,
      "epoch": 0.3986756168359942,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989288576529216,
      "loss": 5.5451,
      "mean_token_accuracy": 0.16579101979732513,
      "num_tokens": 9125394.0,
      "step": 4395
    },
    {
      "entropy": 5.7169290542602536,
      "epoch": 0.3991291727140784,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989257018558917,
      "loss": 5.5305,
      "mean_token_accuracy": 0.1630067154765129,
      "num_tokens": 9137061.0,
      "step": 4400
    },
    {
      "entropy": 5.747994899749756,
      "epoch": 0.39958272859216254,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989225414280213,
      "loss": 5.5093,
      "mean_token_accuracy": 0.16748605370521547,
      "num_tokens": 9148196.0,
      "step": 4405
    },
    {
      "entropy": 5.673751354217529,
      "epoch": 0.40003628447024675,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989193763693756,
      "loss": 5.5305,
      "mean_token_accuracy": 0.17181627005338668,
      "num_tokens": 9158013.0,
      "step": 4410
    },
    {
      "entropy": 5.692364835739136,
      "epoch": 0.4004898403483309,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989162066800202,
      "loss": 5.4596,
      "mean_token_accuracy": 0.17420485466718674,
      "num_tokens": 9167813.0,
      "step": 4415
    },
    {
      "entropy": 5.68905611038208,
      "epoch": 0.4009433962264151,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989130323600205,
      "loss": 5.5125,
      "mean_token_accuracy": 0.16890402287244796,
      "num_tokens": 9177783.0,
      "step": 4420
    },
    {
      "entropy": 5.698638820648194,
      "epoch": 0.40139695210449927,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989098534094424,
      "loss": 5.5409,
      "mean_token_accuracy": 0.16684163510799407,
      "num_tokens": 9187556.0,
      "step": 4425
    },
    {
      "entropy": 5.702458286285401,
      "epoch": 0.4018505079825834,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989066698283512,
      "loss": 5.6142,
      "mean_token_accuracy": 0.16072579473257065,
      "num_tokens": 9198659.0,
      "step": 4430
    },
    {
      "entropy": 5.740787506103516,
      "epoch": 0.40230406386066764,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989034816168132,
      "loss": 5.5916,
      "mean_token_accuracy": 0.16459962129592895,
      "num_tokens": 9209546.0,
      "step": 4435
    },
    {
      "entropy": 5.698708724975586,
      "epoch": 0.4027576197387518,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989002887748941,
      "loss": 5.5123,
      "mean_token_accuracy": 0.16735931783914565,
      "num_tokens": 9219339.0,
      "step": 4440
    },
    {
      "entropy": 5.7313940048217775,
      "epoch": 0.403211175616836,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988970913026601,
      "loss": 5.6901,
      "mean_token_accuracy": 0.15663133114576339,
      "num_tokens": 9230364.0,
      "step": 4445
    },
    {
      "entropy": 5.746043014526367,
      "epoch": 0.40366473149492016,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988938892001771,
      "loss": 5.5391,
      "mean_token_accuracy": 0.16293537616729736,
      "num_tokens": 9240732.0,
      "step": 4450
    },
    {
      "entropy": 5.651700019836426,
      "epoch": 0.40411828737300437,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988906824675113,
      "loss": 5.4611,
      "mean_token_accuracy": 0.16927590519189833,
      "num_tokens": 9251049.0,
      "step": 4455
    },
    {
      "entropy": 5.702632808685303,
      "epoch": 0.4045718432510885,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988874711047293,
      "loss": 5.5209,
      "mean_token_accuracy": 0.1687337875366211,
      "num_tokens": 9261660.0,
      "step": 4460
    },
    {
      "entropy": 5.723224210739136,
      "epoch": 0.40502539912917274,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988842551118972,
      "loss": 5.5738,
      "mean_token_accuracy": 0.1679667592048645,
      "num_tokens": 9272114.0,
      "step": 4465
    },
    {
      "entropy": 5.7394585609436035,
      "epoch": 0.4054789550072569,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988810344890818,
      "loss": 5.5537,
      "mean_token_accuracy": 0.1705055981874466,
      "num_tokens": 9282125.0,
      "step": 4470
    },
    {
      "entropy": 5.77500205039978,
      "epoch": 0.40593251088534105,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988778092363494,
      "loss": 5.5773,
      "mean_token_accuracy": 0.16275593191385268,
      "num_tokens": 9294078.0,
      "step": 4475
    },
    {
      "entropy": 5.723077011108399,
      "epoch": 0.40638606676342526,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988745793537671,
      "loss": 5.5452,
      "mean_token_accuracy": 0.16284733265638351,
      "num_tokens": 9304026.0,
      "step": 4480
    },
    {
      "entropy": 5.714406442642212,
      "epoch": 0.4068396226415094,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988713448414011,
      "loss": 5.5858,
      "mean_token_accuracy": 0.17267868667840958,
      "num_tokens": 9313933.0,
      "step": 4485
    },
    {
      "entropy": 5.722243356704712,
      "epoch": 0.4072931785195936,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498868105699319,
      "loss": 5.6213,
      "mean_token_accuracy": 0.16226827800273896,
      "num_tokens": 9323932.0,
      "step": 4490
    },
    {
      "entropy": 5.6608726978302,
      "epoch": 0.4077467343976778,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988648619275872,
      "loss": 5.5272,
      "mean_token_accuracy": 0.17585674673318863,
      "num_tokens": 9333467.0,
      "step": 4495
    },
    {
      "entropy": 5.7471529006958,
      "epoch": 0.408200290275762,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988616135262729,
      "loss": 5.5802,
      "mean_token_accuracy": 0.16007377952337265,
      "num_tokens": 9344909.0,
      "step": 4500
    },
    {
      "entropy": 5.8212768077850345,
      "epoch": 0.40865384615384615,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988583604954436,
      "loss": 5.6265,
      "mean_token_accuracy": 0.1633604183793068,
      "num_tokens": 9355630.0,
      "step": 4505
    },
    {
      "entropy": 5.634971952438354,
      "epoch": 0.40910740203193036,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988551028351663,
      "loss": 5.5539,
      "mean_token_accuracy": 0.16576218158006667,
      "num_tokens": 9365700.0,
      "step": 4510
    },
    {
      "entropy": 5.6949104309082035,
      "epoch": 0.4095609579100145,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988518405455084,
      "loss": 5.6367,
      "mean_token_accuracy": 0.15977054387331008,
      "num_tokens": 9377088.0,
      "step": 4515
    },
    {
      "entropy": 5.749986362457276,
      "epoch": 0.41001451378809867,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988485736265374,
      "loss": 5.4339,
      "mean_token_accuracy": 0.1753544732928276,
      "num_tokens": 9386855.0,
      "step": 4520
    },
    {
      "entropy": 5.69621696472168,
      "epoch": 0.4104680696661829,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498845302078321,
      "loss": 5.5548,
      "mean_token_accuracy": 0.16240098476409912,
      "num_tokens": 9398661.0,
      "step": 4525
    },
    {
      "entropy": 5.738919639587403,
      "epoch": 0.41092162554426703,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988420259009266,
      "loss": 5.5603,
      "mean_token_accuracy": 0.17131945639848709,
      "num_tokens": 9408670.0,
      "step": 4530
    },
    {
      "entropy": 5.6992714405059814,
      "epoch": 0.41137518142235124,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000498838745094422,
      "loss": 5.5582,
      "mean_token_accuracy": 0.16838853061199188,
      "num_tokens": 9418740.0,
      "step": 4535
    },
    {
      "entropy": 5.6787707805633545,
      "epoch": 0.4118287373004354,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988354596588751,
      "loss": 5.5485,
      "mean_token_accuracy": 0.16721292436122895,
      "num_tokens": 9429339.0,
      "step": 4540
    },
    {
      "entropy": 5.7141166687011715,
      "epoch": 0.4122822931785196,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988321695943539,
      "loss": 5.4913,
      "mean_token_accuracy": 0.17288169413805007,
      "num_tokens": 9438996.0,
      "step": 4545
    },
    {
      "entropy": 5.6817183017730715,
      "epoch": 0.41273584905660377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988288749009264,
      "loss": 5.5176,
      "mean_token_accuracy": 0.16724181026220322,
      "num_tokens": 9448669.0,
      "step": 4550
    },
    {
      "entropy": 5.677099323272705,
      "epoch": 0.413189404934688,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988255755786608,
      "loss": 5.489,
      "mean_token_accuracy": 0.17341368198394774,
      "num_tokens": 9458857.0,
      "step": 4555
    },
    {
      "entropy": 5.6843749523162845,
      "epoch": 0.41364296081277213,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988222716276251,
      "loss": 5.4927,
      "mean_token_accuracy": 0.16991052776575089,
      "num_tokens": 9470241.0,
      "step": 4560
    },
    {
      "entropy": 5.75704345703125,
      "epoch": 0.4140965166908563,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498818963047888,
      "loss": 5.6476,
      "mean_token_accuracy": 0.16511964052915573,
      "num_tokens": 9480528.0,
      "step": 4565
    },
    {
      "entropy": 5.7459704875946045,
      "epoch": 0.4145500725689405,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988156498395176,
      "loss": 5.5663,
      "mean_token_accuracy": 0.16632131338119507,
      "num_tokens": 9491000.0,
      "step": 4570
    },
    {
      "entropy": 5.728471374511718,
      "epoch": 0.41500362844702465,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004988123320025825,
      "loss": 5.5504,
      "mean_token_accuracy": 0.1686439707875252,
      "num_tokens": 9501612.0,
      "step": 4575
    },
    {
      "entropy": 5.721394157409668,
      "epoch": 0.41545718432510886,
      "grad_norm": 0.875,
      "learning_rate": 0.0004988090095371514,
      "loss": 5.6569,
      "mean_token_accuracy": 0.16329480931162835,
      "num_tokens": 9513999.0,
      "step": 4580
    },
    {
      "entropy": 5.723319435119629,
      "epoch": 0.415910740203193,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988056824432928,
      "loss": 5.5296,
      "mean_token_accuracy": 0.1624149963259697,
      "num_tokens": 9524392.0,
      "step": 4585
    },
    {
      "entropy": 5.747483777999878,
      "epoch": 0.41636429608127723,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988023507210758,
      "loss": 5.5387,
      "mean_token_accuracy": 0.16955992728471755,
      "num_tokens": 9534227.0,
      "step": 4590
    },
    {
      "entropy": 5.693861675262451,
      "epoch": 0.4168178519593614,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987990143705692,
      "loss": 5.5163,
      "mean_token_accuracy": 0.1732896625995636,
      "num_tokens": 9544242.0,
      "step": 4595
    },
    {
      "entropy": 5.671651411056518,
      "epoch": 0.4172714078374456,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987956733918418,
      "loss": 5.4526,
      "mean_token_accuracy": 0.16845765262842177,
      "num_tokens": 9553921.0,
      "step": 4600
    },
    {
      "entropy": 5.6189477920532225,
      "epoch": 0.41772496371552975,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004987923277849629,
      "loss": 5.5082,
      "mean_token_accuracy": 0.16487912237644195,
      "num_tokens": 9564768.0,
      "step": 4605
    },
    {
      "entropy": 5.716268110275268,
      "epoch": 0.4181785195936139,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987889775500017,
      "loss": 5.5388,
      "mean_token_accuracy": 0.16002229750156402,
      "num_tokens": 9576360.0,
      "step": 4610
    },
    {
      "entropy": 5.702955532073974,
      "epoch": 0.4186320754716981,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004987856226870272,
      "loss": 5.5542,
      "mean_token_accuracy": 0.16782574057579042,
      "num_tokens": 9587623.0,
      "step": 4615
    },
    {
      "entropy": 5.687364339828491,
      "epoch": 0.4190856313497823,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987822631961092,
      "loss": 5.5773,
      "mean_token_accuracy": 0.16299049258232118,
      "num_tokens": 9597837.0,
      "step": 4620
    },
    {
      "entropy": 5.740718221664428,
      "epoch": 0.4195391872278665,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498778899077317,
      "loss": 5.5345,
      "mean_token_accuracy": 0.1651468500494957,
      "num_tokens": 9609509.0,
      "step": 4625
    },
    {
      "entropy": 5.688720798492431,
      "epoch": 0.41999274310595064,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049877553033072,
      "loss": 5.5257,
      "mean_token_accuracy": 0.1684991091489792,
      "num_tokens": 9619954.0,
      "step": 4630
    },
    {
      "entropy": 5.647857522964477,
      "epoch": 0.42044629898403485,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987721569563882,
      "loss": 5.5167,
      "mean_token_accuracy": 0.1662772074341774,
      "num_tokens": 9629742.0,
      "step": 4635
    },
    {
      "entropy": 5.702399444580078,
      "epoch": 0.420899854862119,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498768778954391,
      "loss": 5.5327,
      "mean_token_accuracy": 0.16915884912014006,
      "num_tokens": 9641378.0,
      "step": 4640
    },
    {
      "entropy": 5.798724412918091,
      "epoch": 0.4213534107402032,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987653963247985,
      "loss": 5.5901,
      "mean_token_accuracy": 0.16833552569150925,
      "num_tokens": 9652279.0,
      "step": 4645
    },
    {
      "entropy": 5.7283261775970455,
      "epoch": 0.42180696661828737,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004987620090676805,
      "loss": 5.6067,
      "mean_token_accuracy": 0.16763487905263902,
      "num_tokens": 9663416.0,
      "step": 4650
    },
    {
      "entropy": 5.70868330001831,
      "epoch": 0.4222605224963715,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987586171831072,
      "loss": 5.5433,
      "mean_token_accuracy": 0.16505215167999268,
      "num_tokens": 9674647.0,
      "step": 4655
    },
    {
      "entropy": 5.6503729820251465,
      "epoch": 0.42271407837445574,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987552206711487,
      "loss": 5.4025,
      "mean_token_accuracy": 0.17310979068279267,
      "num_tokens": 9685084.0,
      "step": 4660
    },
    {
      "entropy": 5.647482967376709,
      "epoch": 0.4231676342525399,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987518195318752,
      "loss": 5.4772,
      "mean_token_accuracy": 0.16532573699951172,
      "num_tokens": 9695661.0,
      "step": 4665
    },
    {
      "entropy": 5.748704338073731,
      "epoch": 0.4236211901306241,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498748413765357,
      "loss": 5.5315,
      "mean_token_accuracy": 0.1663697361946106,
      "num_tokens": 9705773.0,
      "step": 4670
    },
    {
      "entropy": 5.718989038467408,
      "epoch": 0.42407474600870826,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987450033716646,
      "loss": 5.5801,
      "mean_token_accuracy": 0.16458225697278978,
      "num_tokens": 9715799.0,
      "step": 4675
    },
    {
      "entropy": 5.71646819114685,
      "epoch": 0.42452830188679247,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987415883508686,
      "loss": 5.5205,
      "mean_token_accuracy": 0.1714474231004715,
      "num_tokens": 9727012.0,
      "step": 4680
    },
    {
      "entropy": 5.704070425033569,
      "epoch": 0.4249818577648766,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987381687030394,
      "loss": 5.5332,
      "mean_token_accuracy": 0.16558614522218704,
      "num_tokens": 9737044.0,
      "step": 4685
    },
    {
      "entropy": 5.787160110473633,
      "epoch": 0.42543541364296084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987347444282479,
      "loss": 5.6288,
      "mean_token_accuracy": 0.157316155731678,
      "num_tokens": 9747648.0,
      "step": 4690
    },
    {
      "entropy": 5.674534463882447,
      "epoch": 0.425888969521045,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987313155265649,
      "loss": 5.5383,
      "mean_token_accuracy": 0.16501632183790207,
      "num_tokens": 9757812.0,
      "step": 4695
    },
    {
      "entropy": 5.7545582294464115,
      "epoch": 0.42634252539912915,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987278819980613,
      "loss": 5.5327,
      "mean_token_accuracy": 0.17061222344636917,
      "num_tokens": 9768337.0,
      "step": 4700
    },
    {
      "entropy": 5.7120167255401615,
      "epoch": 0.42679608127721336,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987244438428081,
      "loss": 5.5817,
      "mean_token_accuracy": 0.16203325092792512,
      "num_tokens": 9779433.0,
      "step": 4705
    },
    {
      "entropy": 5.6578521728515625,
      "epoch": 0.4272496371552975,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987210010608764,
      "loss": 5.4676,
      "mean_token_accuracy": 0.16993781328201293,
      "num_tokens": 9790069.0,
      "step": 4710
    },
    {
      "entropy": 5.673728132247925,
      "epoch": 0.4277031930333817,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987175536523373,
      "loss": 5.4468,
      "mean_token_accuracy": 0.17233940958976746,
      "num_tokens": 9799669.0,
      "step": 4715
    },
    {
      "entropy": 5.647679948806763,
      "epoch": 0.4281567489114659,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987141016172622,
      "loss": 5.49,
      "mean_token_accuracy": 0.17301111817359924,
      "num_tokens": 9809615.0,
      "step": 4720
    },
    {
      "entropy": 5.668295860290527,
      "epoch": 0.4286103047895501,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987106449557225,
      "loss": 5.4647,
      "mean_token_accuracy": 0.17097459584474564,
      "num_tokens": 9819666.0,
      "step": 4725
    },
    {
      "entropy": 5.747781991958618,
      "epoch": 0.42906386066763424,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987071836677896,
      "loss": 5.6161,
      "mean_token_accuracy": 0.16620567291975022,
      "num_tokens": 9831015.0,
      "step": 4730
    },
    {
      "entropy": 5.681993722915649,
      "epoch": 0.42951741654571846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987037177535353,
      "loss": 5.4838,
      "mean_token_accuracy": 0.16538906544446946,
      "num_tokens": 9841354.0,
      "step": 4735
    },
    {
      "entropy": 5.61437578201294,
      "epoch": 0.4299709724238026,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987002472130309,
      "loss": 5.4796,
      "mean_token_accuracy": 0.1654519483447075,
      "num_tokens": 9850467.0,
      "step": 4740
    },
    {
      "entropy": 5.749066305160523,
      "epoch": 0.43042452830188677,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986967720463485,
      "loss": 5.5535,
      "mean_token_accuracy": 0.1634979486465454,
      "num_tokens": 9861115.0,
      "step": 4745
    },
    {
      "entropy": 5.75732159614563,
      "epoch": 0.430878084179971,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986932922535598,
      "loss": 5.5587,
      "mean_token_accuracy": 0.16462093889713286,
      "num_tokens": 9871172.0,
      "step": 4750
    },
    {
      "entropy": 5.690928268432617,
      "epoch": 0.43133164005805513,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986898078347368,
      "loss": 5.4576,
      "mean_token_accuracy": 0.171144238114357,
      "num_tokens": 9881997.0,
      "step": 4755
    },
    {
      "entropy": 5.67418532371521,
      "epoch": 0.43178519593613934,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986863187899516,
      "loss": 5.4497,
      "mean_token_accuracy": 0.17098109871149064,
      "num_tokens": 9892310.0,
      "step": 4760
    },
    {
      "entropy": 5.6178083419799805,
      "epoch": 0.4322387518142235,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986828251192764,
      "loss": 5.4971,
      "mean_token_accuracy": 0.16893473714590074,
      "num_tokens": 9902664.0,
      "step": 4765
    },
    {
      "entropy": 5.691957664489746,
      "epoch": 0.4326923076923077,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986793268227833,
      "loss": 5.4794,
      "mean_token_accuracy": 0.17000094205141067,
      "num_tokens": 9912491.0,
      "step": 4770
    },
    {
      "entropy": 5.6628247737884525,
      "epoch": 0.43314586357039186,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986758239005448,
      "loss": 5.4546,
      "mean_token_accuracy": 0.1711384743452072,
      "num_tokens": 9922057.0,
      "step": 4775
    },
    {
      "entropy": 5.629938793182373,
      "epoch": 0.4335994194484761,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004986723163526331,
      "loss": 5.4915,
      "mean_token_accuracy": 0.16978790760040283,
      "num_tokens": 9933661.0,
      "step": 4780
    },
    {
      "entropy": 5.644828224182129,
      "epoch": 0.43405297532656023,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986688041791209,
      "loss": 5.4595,
      "mean_token_accuracy": 0.16317514777183534,
      "num_tokens": 9945177.0,
      "step": 4785
    },
    {
      "entropy": 5.686159992218018,
      "epoch": 0.4345065312046444,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498665287380081,
      "loss": 5.5007,
      "mean_token_accuracy": 0.1679864302277565,
      "num_tokens": 9956226.0,
      "step": 4790
    },
    {
      "entropy": 5.728806781768799,
      "epoch": 0.4349600870827286,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986617659555859,
      "loss": 5.566,
      "mean_token_accuracy": 0.1681407630443573,
      "num_tokens": 9965877.0,
      "step": 4795
    },
    {
      "entropy": 5.666096782684326,
      "epoch": 0.43541364296081275,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986582399057085,
      "loss": 5.5077,
      "mean_token_accuracy": 0.1807382047176361,
      "num_tokens": 9975592.0,
      "step": 4800
    },
    {
      "entropy": 5.740148448944092,
      "epoch": 0.43586719883889696,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986547092305216,
      "loss": 5.5476,
      "mean_token_accuracy": 0.16806693226099015,
      "num_tokens": 9986094.0,
      "step": 4805
    },
    {
      "entropy": 5.738724565505981,
      "epoch": 0.4363207547169811,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986511739300984,
      "loss": 5.5802,
      "mean_token_accuracy": 0.16459282785654067,
      "num_tokens": 9997241.0,
      "step": 4810
    },
    {
      "entropy": 5.62156834602356,
      "epoch": 0.43677431059506533,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986476340045119,
      "loss": 5.4877,
      "mean_token_accuracy": 0.1692790299654007,
      "num_tokens": 10008091.0,
      "step": 4815
    },
    {
      "entropy": 5.640922546386719,
      "epoch": 0.4372278664731495,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986440894538354,
      "loss": 5.5512,
      "mean_token_accuracy": 0.16694139540195466,
      "num_tokens": 10018794.0,
      "step": 4820
    },
    {
      "entropy": 5.696912527084351,
      "epoch": 0.4376814223512337,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986405402781421,
      "loss": 5.477,
      "mean_token_accuracy": 0.17061326652765274,
      "num_tokens": 10028414.0,
      "step": 4825
    },
    {
      "entropy": 5.734801959991455,
      "epoch": 0.43813497822931785,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986369864775054,
      "loss": 5.5394,
      "mean_token_accuracy": 0.1690880686044693,
      "num_tokens": 10038749.0,
      "step": 4830
    },
    {
      "entropy": 5.673574686050415,
      "epoch": 0.438588534107402,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986334280519989,
      "loss": 5.4989,
      "mean_token_accuracy": 0.16510852426290512,
      "num_tokens": 10049866.0,
      "step": 4835
    },
    {
      "entropy": 5.658323574066162,
      "epoch": 0.4390420899854862,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498629865001696,
      "loss": 5.5213,
      "mean_token_accuracy": 0.173689503967762,
      "num_tokens": 10060680.0,
      "step": 4840
    },
    {
      "entropy": 5.773933458328247,
      "epoch": 0.43949564586357037,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986262973266706,
      "loss": 5.6142,
      "mean_token_accuracy": 0.1640276402235031,
      "num_tokens": 10070931.0,
      "step": 4845
    },
    {
      "entropy": 5.665406703948975,
      "epoch": 0.4399492017416546,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986227250269964,
      "loss": 5.5471,
      "mean_token_accuracy": 0.1679753065109253,
      "num_tokens": 10080189.0,
      "step": 4850
    },
    {
      "entropy": 5.692187070846558,
      "epoch": 0.44040275761973874,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986191481027472,
      "loss": 5.4578,
      "mean_token_accuracy": 0.17034661769866943,
      "num_tokens": 10088813.0,
      "step": 4855
    },
    {
      "entropy": 5.625132513046265,
      "epoch": 0.44085631349782295,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986155665539971,
      "loss": 5.4795,
      "mean_token_accuracy": 0.16770929843187332,
      "num_tokens": 10098125.0,
      "step": 4860
    },
    {
      "entropy": 5.675574922561646,
      "epoch": 0.4413098693759071,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049861198038082,
      "loss": 5.5161,
      "mean_token_accuracy": 0.1693737342953682,
      "num_tokens": 10109125.0,
      "step": 4865
    },
    {
      "entropy": 5.682723617553711,
      "epoch": 0.4417634252539913,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986083895832903,
      "loss": 5.4948,
      "mean_token_accuracy": 0.16549968123435974,
      "num_tokens": 10120182.0,
      "step": 4870
    },
    {
      "entropy": 5.645736122131348,
      "epoch": 0.44221698113207547,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986047941614821,
      "loss": 5.561,
      "mean_token_accuracy": 0.1642124354839325,
      "num_tokens": 10130941.0,
      "step": 4875
    },
    {
      "entropy": 5.717188692092895,
      "epoch": 0.4426705370101596,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986011941154696,
      "loss": 5.5279,
      "mean_token_accuracy": 0.1694575533270836,
      "num_tokens": 10141708.0,
      "step": 4880
    },
    {
      "entropy": 5.697762584686279,
      "epoch": 0.44312409288824384,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985975894453275,
      "loss": 5.4623,
      "mean_token_accuracy": 0.1702752724289894,
      "num_tokens": 10151816.0,
      "step": 4885
    },
    {
      "entropy": 5.738364410400391,
      "epoch": 0.443577648766328,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985939801511304,
      "loss": 5.5813,
      "mean_token_accuracy": 0.163410022854805,
      "num_tokens": 10162470.0,
      "step": 4890
    },
    {
      "entropy": 5.693057489395142,
      "epoch": 0.4440312046444122,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985903662329527,
      "loss": 5.4715,
      "mean_token_accuracy": 0.17365510314702987,
      "num_tokens": 10173171.0,
      "step": 4895
    },
    {
      "entropy": 5.6850179672241214,
      "epoch": 0.44448476052249636,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985867476908693,
      "loss": 5.5027,
      "mean_token_accuracy": 0.17133045196533203,
      "num_tokens": 10184185.0,
      "step": 4900
    },
    {
      "entropy": 5.753661489486694,
      "epoch": 0.44493831640058057,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985831245249549,
      "loss": 5.5589,
      "mean_token_accuracy": 0.17009343653917314,
      "num_tokens": 10195516.0,
      "step": 4905
    },
    {
      "entropy": 5.651303815841675,
      "epoch": 0.4453918722786647,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985794967352846,
      "loss": 5.4905,
      "mean_token_accuracy": 0.17255396246910096,
      "num_tokens": 10204936.0,
      "step": 4910
    },
    {
      "entropy": 5.680718088150025,
      "epoch": 0.44584542815674894,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004985758643219334,
      "loss": 5.5193,
      "mean_token_accuracy": 0.1686270222067833,
      "num_tokens": 10216218.0,
      "step": 4915
    },
    {
      "entropy": 5.769064855575562,
      "epoch": 0.4462989840348331,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985722272849762,
      "loss": 5.5302,
      "mean_token_accuracy": 0.1625892087817192,
      "num_tokens": 10226370.0,
      "step": 4920
    },
    {
      "entropy": 5.608448934555054,
      "epoch": 0.44675253991291725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985685856244884,
      "loss": 5.4482,
      "mean_token_accuracy": 0.17939605712890624,
      "num_tokens": 10236170.0,
      "step": 4925
    },
    {
      "entropy": 5.718491077423096,
      "epoch": 0.44720609579100146,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985649393405453,
      "loss": 5.5439,
      "mean_token_accuracy": 0.16827276796102525,
      "num_tokens": 10246119.0,
      "step": 4930
    },
    {
      "entropy": 5.671393918991089,
      "epoch": 0.4476596516690856,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985612884332223,
      "loss": 5.5383,
      "mean_token_accuracy": 0.16240749210119249,
      "num_tokens": 10256762.0,
      "step": 4935
    },
    {
      "entropy": 5.633547115325928,
      "epoch": 0.4481132075471698,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985576329025949,
      "loss": 5.4727,
      "mean_token_accuracy": 0.17245685309171677,
      "num_tokens": 10267317.0,
      "step": 4940
    },
    {
      "entropy": 5.726488828659058,
      "epoch": 0.448566763425254,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985539727487385,
      "loss": 5.6217,
      "mean_token_accuracy": 0.15205846652388572,
      "num_tokens": 10279111.0,
      "step": 4945
    },
    {
      "entropy": 5.708870220184326,
      "epoch": 0.4490203193033382,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985503079717292,
      "loss": 5.4791,
      "mean_token_accuracy": 0.16987659484148027,
      "num_tokens": 10289383.0,
      "step": 4950
    },
    {
      "entropy": 5.69928617477417,
      "epoch": 0.44947387518142234,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985466385716425,
      "loss": 5.49,
      "mean_token_accuracy": 0.17275461703538894,
      "num_tokens": 10299806.0,
      "step": 4955
    },
    {
      "entropy": 5.615327405929565,
      "epoch": 0.44992743105950656,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985429645485543,
      "loss": 5.5166,
      "mean_token_accuracy": 0.16595803648233415,
      "num_tokens": 10310348.0,
      "step": 4960
    },
    {
      "entropy": 5.738679695129394,
      "epoch": 0.4503809869375907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985392859025407,
      "loss": 5.5192,
      "mean_token_accuracy": 0.1689467638731003,
      "num_tokens": 10320354.0,
      "step": 4965
    },
    {
      "entropy": 5.647083950042725,
      "epoch": 0.45083454281567487,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985356026336776,
      "loss": 5.4311,
      "mean_token_accuracy": 0.1705056294798851,
      "num_tokens": 10330292.0,
      "step": 4970
    },
    {
      "entropy": 5.634524917602539,
      "epoch": 0.4512880986937591,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985319147420414,
      "loss": 5.4075,
      "mean_token_accuracy": 0.17841226756572723,
      "num_tokens": 10340449.0,
      "step": 4975
    },
    {
      "entropy": 5.688153219223023,
      "epoch": 0.45174165457184323,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985282222277082,
      "loss": 5.4677,
      "mean_token_accuracy": 0.1669263258576393,
      "num_tokens": 10350637.0,
      "step": 4980
    },
    {
      "entropy": 5.597547960281372,
      "epoch": 0.45219521044992744,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985245250907544,
      "loss": 5.4141,
      "mean_token_accuracy": 0.17463358342647553,
      "num_tokens": 10361548.0,
      "step": 4985
    },
    {
      "entropy": 5.60842866897583,
      "epoch": 0.4526487663280116,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985208233312564,
      "loss": 5.4248,
      "mean_token_accuracy": 0.1648386925458908,
      "num_tokens": 10371765.0,
      "step": 4990
    },
    {
      "entropy": 5.716899347305298,
      "epoch": 0.4531023222060958,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985171169492908,
      "loss": 5.5395,
      "mean_token_accuracy": 0.1704647123813629,
      "num_tokens": 10382108.0,
      "step": 4995
    },
    {
      "entropy": 5.6541526317596436,
      "epoch": 0.45355587808417996,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985134059449344,
      "loss": 5.4909,
      "mean_token_accuracy": 0.1751002311706543,
      "num_tokens": 10392390.0,
      "step": 5000
    },
    {
      "entropy": 5.6331511497497555,
      "epoch": 0.4540094339622642,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985096903182637,
      "loss": 5.5349,
      "mean_token_accuracy": 0.1757347598671913,
      "num_tokens": 10402094.0,
      "step": 5005
    },
    {
      "entropy": 5.689352416992188,
      "epoch": 0.45446298984034833,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985059700693557,
      "loss": 5.4904,
      "mean_token_accuracy": 0.17052320688962935,
      "num_tokens": 10413350.0,
      "step": 5010
    },
    {
      "entropy": 5.726061534881592,
      "epoch": 0.4549165457184325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985022451982872,
      "loss": 5.6072,
      "mean_token_accuracy": 0.16370004415512085,
      "num_tokens": 10424292.0,
      "step": 5015
    },
    {
      "entropy": 5.757494688034058,
      "epoch": 0.4553701015965167,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984985157051354,
      "loss": 5.4711,
      "mean_token_accuracy": 0.17044610232114793,
      "num_tokens": 10433724.0,
      "step": 5020
    },
    {
      "entropy": 5.700105333328247,
      "epoch": 0.45582365747460085,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984947815899774,
      "loss": 5.5438,
      "mean_token_accuracy": 0.1677020400762558,
      "num_tokens": 10444724.0,
      "step": 5025
    },
    {
      "entropy": 5.677012205123901,
      "epoch": 0.45627721335268506,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984910428528902,
      "loss": 5.4799,
      "mean_token_accuracy": 0.1747812494635582,
      "num_tokens": 10455086.0,
      "step": 5030
    },
    {
      "entropy": 5.592632293701172,
      "epoch": 0.4567307692307692,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984872994939514,
      "loss": 5.4252,
      "mean_token_accuracy": 0.16925584375858307,
      "num_tokens": 10465076.0,
      "step": 5035
    },
    {
      "entropy": 5.68324499130249,
      "epoch": 0.45718432510885343,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984835515132382,
      "loss": 5.4663,
      "mean_token_accuracy": 0.17937684208154678,
      "num_tokens": 10474817.0,
      "step": 5040
    },
    {
      "entropy": 5.662622547149658,
      "epoch": 0.4576378809869376,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984797989108282,
      "loss": 5.4851,
      "mean_token_accuracy": 0.16741085201501846,
      "num_tokens": 10485735.0,
      "step": 5045
    },
    {
      "entropy": 5.636821985244751,
      "epoch": 0.4580914368650218,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984760416867991,
      "loss": 5.5723,
      "mean_token_accuracy": 0.16558835059404373,
      "num_tokens": 10496731.0,
      "step": 5050
    },
    {
      "entropy": 5.630299758911133,
      "epoch": 0.45854499274310595,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984722798412286,
      "loss": 5.4242,
      "mean_token_accuracy": 0.1714159443974495,
      "num_tokens": 10506787.0,
      "step": 5055
    },
    {
      "entropy": 5.7004170417785645,
      "epoch": 0.4589985486211901,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984685133741941,
      "loss": 5.5117,
      "mean_token_accuracy": 0.16912039518356323,
      "num_tokens": 10517170.0,
      "step": 5060
    },
    {
      "entropy": 5.6526471138000485,
      "epoch": 0.4594521044992743,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498464742285774,
      "loss": 5.4944,
      "mean_token_accuracy": 0.16828947961330415,
      "num_tokens": 10527342.0,
      "step": 5065
    },
    {
      "entropy": 5.692163467407227,
      "epoch": 0.45990566037735847,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498460966576046,
      "loss": 5.4743,
      "mean_token_accuracy": 0.17233156114816667,
      "num_tokens": 10537253.0,
      "step": 5070
    },
    {
      "entropy": 5.7053141593933105,
      "epoch": 0.4603592162554427,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004984571862450884,
      "loss": 5.4828,
      "mean_token_accuracy": 0.1697017401456833,
      "num_tokens": 10547249.0,
      "step": 5075
    },
    {
      "entropy": 5.753048896789551,
      "epoch": 0.46081277213352684,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984534012929791,
      "loss": 5.5371,
      "mean_token_accuracy": 0.17120666801929474,
      "num_tokens": 10558174.0,
      "step": 5080
    },
    {
      "entropy": 5.728115892410278,
      "epoch": 0.46126632801161105,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984496117197967,
      "loss": 5.4753,
      "mean_token_accuracy": 0.17589109838008882,
      "num_tokens": 10567610.0,
      "step": 5085
    },
    {
      "entropy": 5.601778030395508,
      "epoch": 0.4617198838896952,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984458175256193,
      "loss": 5.4797,
      "mean_token_accuracy": 0.16605047583580018,
      "num_tokens": 10578494.0,
      "step": 5090
    },
    {
      "entropy": 5.7371217727661135,
      "epoch": 0.4621734397677794,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984420187105253,
      "loss": 5.586,
      "mean_token_accuracy": 0.17094989866018295,
      "num_tokens": 10589482.0,
      "step": 5095
    },
    {
      "entropy": 5.755274152755737,
      "epoch": 0.46262699564586357,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984382152745936,
      "loss": 5.5401,
      "mean_token_accuracy": 0.16257912814617156,
      "num_tokens": 10599965.0,
      "step": 5100
    },
    {
      "entropy": 5.66287579536438,
      "epoch": 0.4630805515239477,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004984344072179026,
      "loss": 5.5603,
      "mean_token_accuracy": 0.17168972343206407,
      "num_tokens": 10610500.0,
      "step": 5105
    },
    {
      "entropy": 5.742537546157837,
      "epoch": 0.46353410740203194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984305945405312,
      "loss": 5.5188,
      "mean_token_accuracy": 0.17038158923387528,
      "num_tokens": 10619843.0,
      "step": 5110
    },
    {
      "entropy": 5.658458423614502,
      "epoch": 0.4639876632801161,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498426777242558,
      "loss": 5.4436,
      "mean_token_accuracy": 0.17783846333622932,
      "num_tokens": 10631026.0,
      "step": 5115
    },
    {
      "entropy": 5.538843965530395,
      "epoch": 0.4644412191582003,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984229553240623,
      "loss": 5.4496,
      "mean_token_accuracy": 0.17878477275371552,
      "num_tokens": 10641130.0,
      "step": 5120
    },
    {
      "entropy": 5.746715879440307,
      "epoch": 0.46489477503628446,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984191287851228,
      "loss": 5.5067,
      "mean_token_accuracy": 0.17264201045036315,
      "num_tokens": 10652305.0,
      "step": 5125
    },
    {
      "entropy": 5.73653655052185,
      "epoch": 0.46534833091436867,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984152976258188,
      "loss": 5.5573,
      "mean_token_accuracy": 0.1695479616522789,
      "num_tokens": 10663083.0,
      "step": 5130
    },
    {
      "entropy": 5.683012914657593,
      "epoch": 0.4658018867924528,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984114618462296,
      "loss": 5.572,
      "mean_token_accuracy": 0.1649271011352539,
      "num_tokens": 10673602.0,
      "step": 5135
    },
    {
      "entropy": 5.6376933574676515,
      "epoch": 0.46625544267053703,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984076214464343,
      "loss": 5.392,
      "mean_token_accuracy": 0.1789440020918846,
      "num_tokens": 10684622.0,
      "step": 5140
    },
    {
      "entropy": 5.665349912643433,
      "epoch": 0.4667089985486212,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984037764265126,
      "loss": 5.5034,
      "mean_token_accuracy": 0.1705288678407669,
      "num_tokens": 10695129.0,
      "step": 5145
    },
    {
      "entropy": 5.6029088497161865,
      "epoch": 0.46716255442670535,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983999267865438,
      "loss": 5.3525,
      "mean_token_accuracy": 0.18145803660154342,
      "num_tokens": 10705236.0,
      "step": 5150
    },
    {
      "entropy": 5.638808965682983,
      "epoch": 0.46761611030478956,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983960725266078,
      "loss": 5.4762,
      "mean_token_accuracy": 0.17050447314977646,
      "num_tokens": 10715748.0,
      "step": 5155
    },
    {
      "entropy": 5.656491661071778,
      "epoch": 0.4680696661828737,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983922136467838,
      "loss": 5.3904,
      "mean_token_accuracy": 0.17664045691490174,
      "num_tokens": 10725799.0,
      "step": 5160
    },
    {
      "entropy": 5.659603071212769,
      "epoch": 0.4685232220609579,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498388350147152,
      "loss": 5.5361,
      "mean_token_accuracy": 0.1669347897171974,
      "num_tokens": 10735668.0,
      "step": 5165
    },
    {
      "entropy": 5.695495986938477,
      "epoch": 0.4689767779390421,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983844820277922,
      "loss": 5.6331,
      "mean_token_accuracy": 0.16362748593091964,
      "num_tokens": 10746466.0,
      "step": 5170
    },
    {
      "entropy": 5.791701602935791,
      "epoch": 0.4694303338171263,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983806092887843,
      "loss": 5.5238,
      "mean_token_accuracy": 0.16722518354654312,
      "num_tokens": 10756724.0,
      "step": 5175
    },
    {
      "entropy": 5.698781108856201,
      "epoch": 0.46988388969521044,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983767319302086,
      "loss": 5.554,
      "mean_token_accuracy": 0.17089424878358841,
      "num_tokens": 10766919.0,
      "step": 5180
    },
    {
      "entropy": 5.584889268875122,
      "epoch": 0.47033744557329465,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983728499521451,
      "loss": 5.3443,
      "mean_token_accuracy": 0.18152898848056792,
      "num_tokens": 10777550.0,
      "step": 5185
    },
    {
      "entropy": 5.667958736419678,
      "epoch": 0.4707910014513788,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983689633546741,
      "loss": 5.4103,
      "mean_token_accuracy": 0.16942705661058427,
      "num_tokens": 10788456.0,
      "step": 5190
    },
    {
      "entropy": 5.64146933555603,
      "epoch": 0.47124455732946297,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983650721378761,
      "loss": 5.449,
      "mean_token_accuracy": 0.17425691485404968,
      "num_tokens": 10800049.0,
      "step": 5195
    },
    {
      "entropy": 5.6370322704315186,
      "epoch": 0.4716981132075472,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983611763018314,
      "loss": 5.5038,
      "mean_token_accuracy": 0.16732366234064103,
      "num_tokens": 10810743.0,
      "step": 5200
    },
    {
      "entropy": 5.801402473449707,
      "epoch": 0.47215166908563133,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983572758466208,
      "loss": 5.6878,
      "mean_token_accuracy": 0.16397374272346496,
      "num_tokens": 10821897.0,
      "step": 5205
    },
    {
      "entropy": 5.7822551250457765,
      "epoch": 0.47260522496371554,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983533707723247,
      "loss": 5.5451,
      "mean_token_accuracy": 0.17132560312747955,
      "num_tokens": 10831589.0,
      "step": 5210
    },
    {
      "entropy": 5.684686899185181,
      "epoch": 0.4730587808417997,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983494610790241,
      "loss": 5.4469,
      "mean_token_accuracy": 0.16281144618988036,
      "num_tokens": 10841404.0,
      "step": 5215
    },
    {
      "entropy": 5.6485227108001705,
      "epoch": 0.4735123367198839,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983455467667996,
      "loss": 5.5001,
      "mean_token_accuracy": 0.16492253094911574,
      "num_tokens": 10851577.0,
      "step": 5220
    },
    {
      "entropy": 5.6523786067962645,
      "epoch": 0.47396589259796806,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983416278357322,
      "loss": 5.4073,
      "mean_token_accuracy": 0.17646127343177795,
      "num_tokens": 10861403.0,
      "step": 5225
    },
    {
      "entropy": 5.64488115310669,
      "epoch": 0.4744194484760523,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983377042859032,
      "loss": 5.4546,
      "mean_token_accuracy": 0.16768310219049454,
      "num_tokens": 10871933.0,
      "step": 5230
    },
    {
      "entropy": 5.648560285568237,
      "epoch": 0.47487300435413643,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983337761173936,
      "loss": 5.5315,
      "mean_token_accuracy": 0.16874219328165055,
      "num_tokens": 10881800.0,
      "step": 5235
    },
    {
      "entropy": 5.737661933898925,
      "epoch": 0.4753265602322206,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983298433302843,
      "loss": 5.4753,
      "mean_token_accuracy": 0.1727318361401558,
      "num_tokens": 10891830.0,
      "step": 5240
    },
    {
      "entropy": 5.6505979061126705,
      "epoch": 0.4757801161103048,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983259059246572,
      "loss": 5.403,
      "mean_token_accuracy": 0.17735837250947953,
      "num_tokens": 10901557.0,
      "step": 5245
    },
    {
      "entropy": 5.579076051712036,
      "epoch": 0.47623367198838895,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983219639005934,
      "loss": 5.5252,
      "mean_token_accuracy": 0.16280789524316788,
      "num_tokens": 10912155.0,
      "step": 5250
    },
    {
      "entropy": 5.642884588241577,
      "epoch": 0.47668722786647316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983180172581745,
      "loss": 5.4296,
      "mean_token_accuracy": 0.17732747346162797,
      "num_tokens": 10922291.0,
      "step": 5255
    },
    {
      "entropy": 5.7082366943359375,
      "epoch": 0.4771407837445573,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983140659974819,
      "loss": 5.5024,
      "mean_token_accuracy": 0.1679125025868416,
      "num_tokens": 10933176.0,
      "step": 5260
    },
    {
      "entropy": 5.698912668228149,
      "epoch": 0.47759433962264153,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983101101185977,
      "loss": 5.5828,
      "mean_token_accuracy": 0.16323981285095215,
      "num_tokens": 10944049.0,
      "step": 5265
    },
    {
      "entropy": 5.633695650100708,
      "epoch": 0.4780478955007257,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983061496216035,
      "loss": 5.4238,
      "mean_token_accuracy": 0.17720418572425842,
      "num_tokens": 10954347.0,
      "step": 5270
    },
    {
      "entropy": 5.6351593971252445,
      "epoch": 0.4785014513788099,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983021845065812,
      "loss": 5.4756,
      "mean_token_accuracy": 0.175036196410656,
      "num_tokens": 10965123.0,
      "step": 5275
    },
    {
      "entropy": 5.597521591186523,
      "epoch": 0.47895500725689405,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982982147736128,
      "loss": 5.4642,
      "mean_token_accuracy": 0.17434579282999038,
      "num_tokens": 10976395.0,
      "step": 5280
    },
    {
      "entropy": 5.691222858428955,
      "epoch": 0.4794085631349782,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982942404227805,
      "loss": 5.4586,
      "mean_token_accuracy": 0.16569167971611024,
      "num_tokens": 10986797.0,
      "step": 5285
    },
    {
      "entropy": 5.708698081970215,
      "epoch": 0.4798621190130624,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982902614541662,
      "loss": 5.5117,
      "mean_token_accuracy": 0.17047134935855865,
      "num_tokens": 10997229.0,
      "step": 5290
    },
    {
      "entropy": 5.614826726913452,
      "epoch": 0.48031567489114657,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982862778678526,
      "loss": 5.4412,
      "mean_token_accuracy": 0.17899638414382935,
      "num_tokens": 11008180.0,
      "step": 5295
    },
    {
      "entropy": 5.683491897583008,
      "epoch": 0.4807692307692308,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982822896639218,
      "loss": 5.4705,
      "mean_token_accuracy": 0.17429982572793962,
      "num_tokens": 11019248.0,
      "step": 5300
    },
    {
      "entropy": 5.617243051528931,
      "epoch": 0.48122278664731494,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982782968424563,
      "loss": 5.3761,
      "mean_token_accuracy": 0.1775210216641426,
      "num_tokens": 11029760.0,
      "step": 5305
    },
    {
      "entropy": 5.588413190841675,
      "epoch": 0.48167634252539915,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982742994035388,
      "loss": 5.4821,
      "mean_token_accuracy": 0.17242670059204102,
      "num_tokens": 11039211.0,
      "step": 5310
    },
    {
      "entropy": 5.662622356414795,
      "epoch": 0.4821298984034833,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982702973472518,
      "loss": 5.454,
      "mean_token_accuracy": 0.17342988401651382,
      "num_tokens": 11049892.0,
      "step": 5315
    },
    {
      "entropy": 5.746756601333618,
      "epoch": 0.4825834542815675,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982662906736783,
      "loss": 5.5041,
      "mean_token_accuracy": 0.17146852016448974,
      "num_tokens": 11060823.0,
      "step": 5320
    },
    {
      "entropy": 5.609480905532837,
      "epoch": 0.48303701015965167,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004982622793829009,
      "loss": 5.4768,
      "mean_token_accuracy": 0.17211514413356782,
      "num_tokens": 11072353.0,
      "step": 5325
    },
    {
      "entropy": 5.6528795719146725,
      "epoch": 0.4834905660377358,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982582634750027,
      "loss": 5.491,
      "mean_token_accuracy": 0.1721246212720871,
      "num_tokens": 11082655.0,
      "step": 5330
    },
    {
      "entropy": 5.676597452163696,
      "epoch": 0.48394412191582004,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982542429500667,
      "loss": 5.5053,
      "mean_token_accuracy": 0.17382133454084397,
      "num_tokens": 11093315.0,
      "step": 5335
    },
    {
      "entropy": 5.524261379241944,
      "epoch": 0.4843976777939042,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982502178081761,
      "loss": 5.4042,
      "mean_token_accuracy": 0.17310083359479905,
      "num_tokens": 11103074.0,
      "step": 5340
    },
    {
      "entropy": 5.666402530670166,
      "epoch": 0.4848512336719884,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982461880494141,
      "loss": 5.4279,
      "mean_token_accuracy": 0.17441916912794114,
      "num_tokens": 11113636.0,
      "step": 5345
    },
    {
      "entropy": 5.656398725509644,
      "epoch": 0.48530478955007256,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498242153673864,
      "loss": 5.4039,
      "mean_token_accuracy": 0.17608933597803117,
      "num_tokens": 11123509.0,
      "step": 5350
    },
    {
      "entropy": 5.613732099533081,
      "epoch": 0.48575834542815677,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982381146816094,
      "loss": 5.4686,
      "mean_token_accuracy": 0.16522315591573716,
      "num_tokens": 11133405.0,
      "step": 5355
    },
    {
      "entropy": 5.73980131149292,
      "epoch": 0.4862119013062409,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004982340710727336,
      "loss": 5.533,
      "mean_token_accuracy": 0.16312984079122544,
      "num_tokens": 11144725.0,
      "step": 5360
    },
    {
      "entropy": 5.67049765586853,
      "epoch": 0.48666545718432513,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982300228473203,
      "loss": 5.4184,
      "mean_token_accuracy": 0.16829311102628708,
      "num_tokens": 11155446.0,
      "step": 5365
    },
    {
      "entropy": 5.636474370956421,
      "epoch": 0.4871190130624093,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982259700054533,
      "loss": 5.4818,
      "mean_token_accuracy": 0.17630883306264877,
      "num_tokens": 11165380.0,
      "step": 5370
    },
    {
      "entropy": 5.647053050994873,
      "epoch": 0.48757256894049344,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982219125472163,
      "loss": 5.5238,
      "mean_token_accuracy": 0.1691724866628647,
      "num_tokens": 11175641.0,
      "step": 5375
    },
    {
      "entropy": 5.799345302581787,
      "epoch": 0.48802612481857766,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982178504726933,
      "loss": 5.6734,
      "mean_token_accuracy": 0.17320263534784316,
      "num_tokens": 11186257.0,
      "step": 5380
    },
    {
      "entropy": 5.727895784378052,
      "epoch": 0.4884796806966618,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982137837819682,
      "loss": 5.4966,
      "mean_token_accuracy": 0.16289989352226258,
      "num_tokens": 11197284.0,
      "step": 5385
    },
    {
      "entropy": 5.688380861282349,
      "epoch": 0.488933236574746,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982097124751252,
      "loss": 5.4741,
      "mean_token_accuracy": 0.17410095632076264,
      "num_tokens": 11208413.0,
      "step": 5390
    },
    {
      "entropy": 5.629041337966919,
      "epoch": 0.4893867924528302,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982056365522486,
      "loss": 5.4667,
      "mean_token_accuracy": 0.17381504625082017,
      "num_tokens": 11219016.0,
      "step": 5395
    },
    {
      "entropy": 5.625274276733398,
      "epoch": 0.4898403483309144,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982015560134225,
      "loss": 5.4451,
      "mean_token_accuracy": 0.1747204378247261,
      "num_tokens": 11228934.0,
      "step": 5400
    },
    {
      "entropy": 5.639862775802612,
      "epoch": 0.49029390420899854,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004981974708587312,
      "loss": 5.3724,
      "mean_token_accuracy": 0.17818964570760726,
      "num_tokens": 11239596.0,
      "step": 5405
    },
    {
      "entropy": 5.603957653045654,
      "epoch": 0.49074746008708275,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981933810882594,
      "loss": 5.3515,
      "mean_token_accuracy": 0.17084526121616364,
      "num_tokens": 11249560.0,
      "step": 5410
    },
    {
      "entropy": 5.707994842529297,
      "epoch": 0.4912010159651669,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981892867020917,
      "loss": 5.5617,
      "mean_token_accuracy": 0.17124133110046386,
      "num_tokens": 11259787.0,
      "step": 5415
    },
    {
      "entropy": 5.722805786132812,
      "epoch": 0.49165457184325106,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981851877003125,
      "loss": 5.5011,
      "mean_token_accuracy": 0.1731458768248558,
      "num_tokens": 11269717.0,
      "step": 5420
    },
    {
      "entropy": 5.625773620605469,
      "epoch": 0.4921081277213353,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981810840830068,
      "loss": 5.474,
      "mean_token_accuracy": 0.18215908706188202,
      "num_tokens": 11279863.0,
      "step": 5425
    },
    {
      "entropy": 5.666403102874756,
      "epoch": 0.49256168359941943,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981769758502596,
      "loss": 5.5915,
      "mean_token_accuracy": 0.1605565443634987,
      "num_tokens": 11291161.0,
      "step": 5430
    },
    {
      "entropy": 5.609889507293701,
      "epoch": 0.49301523947750364,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981728630021555,
      "loss": 5.4073,
      "mean_token_accuracy": 0.17431867867708206,
      "num_tokens": 11301151.0,
      "step": 5435
    },
    {
      "entropy": 5.583493232727051,
      "epoch": 0.4934687953555878,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981687455387797,
      "loss": 5.4231,
      "mean_token_accuracy": 0.1704528123140335,
      "num_tokens": 11311466.0,
      "step": 5440
    },
    {
      "entropy": 5.651216983795166,
      "epoch": 0.493922351233672,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981646234602173,
      "loss": 5.3677,
      "mean_token_accuracy": 0.17543147057294844,
      "num_tokens": 11321522.0,
      "step": 5445
    },
    {
      "entropy": 5.70766978263855,
      "epoch": 0.49437590711175616,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981604967665539,
      "loss": 5.4904,
      "mean_token_accuracy": 0.17429745495319365,
      "num_tokens": 11332317.0,
      "step": 5450
    },
    {
      "entropy": 5.638960981369019,
      "epoch": 0.4948294629898404,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981563654578743,
      "loss": 5.4383,
      "mean_token_accuracy": 0.1693056493997574,
      "num_tokens": 11343024.0,
      "step": 5455
    },
    {
      "entropy": 5.579675197601318,
      "epoch": 0.49528301886792453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981522295342643,
      "loss": 5.392,
      "mean_token_accuracy": 0.17219439148902893,
      "num_tokens": 11353114.0,
      "step": 5460
    },
    {
      "entropy": 5.644372177124024,
      "epoch": 0.4957365747460087,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004981480889958092,
      "loss": 5.4615,
      "mean_token_accuracy": 0.17464147955179216,
      "num_tokens": 11363446.0,
      "step": 5465
    },
    {
      "entropy": 5.612289381027222,
      "epoch": 0.4961901306240929,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981439438425949,
      "loss": 5.4536,
      "mean_token_accuracy": 0.1771087482571602,
      "num_tokens": 11373413.0,
      "step": 5470
    },
    {
      "entropy": 5.613510656356811,
      "epoch": 0.49664368650217705,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498139794074707,
      "loss": 5.3264,
      "mean_token_accuracy": 0.17951995581388475,
      "num_tokens": 11383142.0,
      "step": 5475
    },
    {
      "entropy": 5.666054725646973,
      "epoch": 0.49709724238026126,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981356396922313,
      "loss": 5.4847,
      "mean_token_accuracy": 0.17279811352491378,
      "num_tokens": 11394635.0,
      "step": 5480
    },
    {
      "entropy": 5.604884958267212,
      "epoch": 0.4975507982583454,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981314806952535,
      "loss": 5.3892,
      "mean_token_accuracy": 0.17838599234819413,
      "num_tokens": 11404442.0,
      "step": 5485
    },
    {
      "entropy": 5.557573413848877,
      "epoch": 0.49800435413642963,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049812731708386,
      "loss": 5.3433,
      "mean_token_accuracy": 0.17274410128593445,
      "num_tokens": 11414217.0,
      "step": 5490
    },
    {
      "entropy": 5.678751802444458,
      "epoch": 0.4984579100145138,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981231488581367,
      "loss": 5.4886,
      "mean_token_accuracy": 0.16931686848402022,
      "num_tokens": 11423834.0,
      "step": 5495
    },
    {
      "entropy": 5.5513814926147464,
      "epoch": 0.498911465892598,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981189760181699,
      "loss": 5.358,
      "mean_token_accuracy": 0.17370127141475677,
      "num_tokens": 11433613.0,
      "step": 5500
    },
    {
      "entropy": 5.579940223693848,
      "epoch": 0.49936502177068215,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981147985640456,
      "loss": 5.4257,
      "mean_token_accuracy": 0.17723614424467088,
      "num_tokens": 11443915.0,
      "step": 5505
    },
    {
      "entropy": 5.704259824752808,
      "epoch": 0.4998185776487663,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981106164958506,
      "loss": 5.4561,
      "mean_token_accuracy": 0.1709003448486328,
      "num_tokens": 11454588.0,
      "step": 5510
    },
    {
      "entropy": 5.648460054397583,
      "epoch": 0.5002721335268505,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981064298136712,
      "loss": 5.5067,
      "mean_token_accuracy": 0.16957272291183473,
      "num_tokens": 11465533.0,
      "step": 5515
    },
    {
      "entropy": 5.622145414352417,
      "epoch": 0.5007256894049347,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981022385175939,
      "loss": 5.4586,
      "mean_token_accuracy": 0.16735394597053527,
      "num_tokens": 11475776.0,
      "step": 5520
    },
    {
      "entropy": 5.620416879653931,
      "epoch": 0.5011792452830188,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980980426077054,
      "loss": 5.4385,
      "mean_token_accuracy": 0.18021399825811385,
      "num_tokens": 11486200.0,
      "step": 5525
    },
    {
      "entropy": 5.633817100524903,
      "epoch": 0.5016328011611031,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980938420840926,
      "loss": 5.4609,
      "mean_token_accuracy": 0.16805864125490189,
      "num_tokens": 11496768.0,
      "step": 5530
    },
    {
      "entropy": 5.64311728477478,
      "epoch": 0.5020863570391872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980896369468423,
      "loss": 5.438,
      "mean_token_accuracy": 0.17364895343780518,
      "num_tokens": 11507096.0,
      "step": 5535
    },
    {
      "entropy": 5.617955064773559,
      "epoch": 0.5025399129172714,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980854271960415,
      "loss": 5.4858,
      "mean_token_accuracy": 0.17146076411008834,
      "num_tokens": 11517301.0,
      "step": 5540
    },
    {
      "entropy": 5.691181278228759,
      "epoch": 0.5029934687953556,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980812128317772,
      "loss": 5.5381,
      "mean_token_accuracy": 0.17311448901891707,
      "num_tokens": 11528616.0,
      "step": 5545
    },
    {
      "entropy": 5.710403871536255,
      "epoch": 0.5034470246734397,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004980769938541364,
      "loss": 5.4596,
      "mean_token_accuracy": 0.17918587774038314,
      "num_tokens": 11539344.0,
      "step": 5550
    },
    {
      "entropy": 5.611640930175781,
      "epoch": 0.503900580551524,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980727702632066,
      "loss": 5.5158,
      "mean_token_accuracy": 0.1702592894434929,
      "num_tokens": 11549511.0,
      "step": 5555
    },
    {
      "entropy": 5.654125452041626,
      "epoch": 0.5043541364296081,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980685420590752,
      "loss": 5.4425,
      "mean_token_accuracy": 0.17124981731176375,
      "num_tokens": 11559425.0,
      "step": 5560
    },
    {
      "entropy": 5.7445450782775875,
      "epoch": 0.5048076923076923,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980643092418294,
      "loss": 5.564,
      "mean_token_accuracy": 0.16161303967237473,
      "num_tokens": 11570406.0,
      "step": 5565
    },
    {
      "entropy": 5.652337265014649,
      "epoch": 0.5052612481857764,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980600718115568,
      "loss": 5.4435,
      "mean_token_accuracy": 0.17120484560728072,
      "num_tokens": 11580509.0,
      "step": 5570
    },
    {
      "entropy": 5.657799530029297,
      "epoch": 0.5057148040638607,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004980558297683452,
      "loss": 5.4658,
      "mean_token_accuracy": 0.17254962921142578,
      "num_tokens": 11590860.0,
      "step": 5575
    },
    {
      "entropy": 5.69189043045044,
      "epoch": 0.5061683599419449,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980515831122822,
      "loss": 5.4658,
      "mean_token_accuracy": 0.16972111016511918,
      "num_tokens": 11601954.0,
      "step": 5580
    },
    {
      "entropy": 5.641917419433594,
      "epoch": 0.506621915820029,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980473318434555,
      "loss": 5.4643,
      "mean_token_accuracy": 0.17524617165327072,
      "num_tokens": 11612581.0,
      "step": 5585
    },
    {
      "entropy": 5.54950852394104,
      "epoch": 0.5070754716981132,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004980430759619532,
      "loss": 5.3893,
      "mean_token_accuracy": 0.17424534559249877,
      "num_tokens": 11622599.0,
      "step": 5590
    },
    {
      "entropy": 5.572847175598144,
      "epoch": 0.5075290275761973,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980388154678633,
      "loss": 5.4041,
      "mean_token_accuracy": 0.1734546795487404,
      "num_tokens": 11633611.0,
      "step": 5595
    },
    {
      "entropy": 5.59787654876709,
      "epoch": 0.5079825834542816,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004980345503612738,
      "loss": 5.3836,
      "mean_token_accuracy": 0.1732645958662033,
      "num_tokens": 11643159.0,
      "step": 5600
    },
    {
      "entropy": 5.615271997451782,
      "epoch": 0.5084361393323658,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980302806422729,
      "loss": 5.3753,
      "mean_token_accuracy": 0.1775027260184288,
      "num_tokens": 11653174.0,
      "step": 5605
    },
    {
      "entropy": 5.674450540542603,
      "epoch": 0.5088896952104499,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980260063109491,
      "loss": 5.4227,
      "mean_token_accuracy": 0.17737311273813247,
      "num_tokens": 11662621.0,
      "step": 5610
    },
    {
      "entropy": 5.543775224685669,
      "epoch": 0.5093432510885341,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980217273673906,
      "loss": 5.3793,
      "mean_token_accuracy": 0.17156035602092742,
      "num_tokens": 11673538.0,
      "step": 5615
    },
    {
      "entropy": 5.653875684738159,
      "epoch": 0.5097968069666183,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980174438116859,
      "loss": 5.3914,
      "mean_token_accuracy": 0.17867197394371032,
      "num_tokens": 11683203.0,
      "step": 5620
    },
    {
      "entropy": 5.552722072601318,
      "epoch": 0.5102503628447025,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980131556439236,
      "loss": 5.3829,
      "mean_token_accuracy": 0.180565182864666,
      "num_tokens": 11693313.0,
      "step": 5625
    },
    {
      "entropy": 5.640235567092896,
      "epoch": 0.5107039187227866,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980088628641925,
      "loss": 5.4783,
      "mean_token_accuracy": 0.17274700552225114,
      "num_tokens": 11704194.0,
      "step": 5630
    },
    {
      "entropy": 5.642412567138672,
      "epoch": 0.5111574746008708,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980045654725812,
      "loss": 5.336,
      "mean_token_accuracy": 0.17914541512727739,
      "num_tokens": 11715109.0,
      "step": 5635
    },
    {
      "entropy": 5.5468549728393555,
      "epoch": 0.511611030478955,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980002634691787,
      "loss": 5.3796,
      "mean_token_accuracy": 0.17843705415725708,
      "num_tokens": 11724961.0,
      "step": 5640
    },
    {
      "entropy": 5.552511072158813,
      "epoch": 0.5120645863570392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497995956854074,
      "loss": 5.3914,
      "mean_token_accuracy": 0.17831180095672608,
      "num_tokens": 11735334.0,
      "step": 5645
    },
    {
      "entropy": 5.550053977966309,
      "epoch": 0.5125181422351234,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979916456273561,
      "loss": 5.3935,
      "mean_token_accuracy": 0.1712997004389763,
      "num_tokens": 11746858.0,
      "step": 5650
    },
    {
      "entropy": 5.689655733108521,
      "epoch": 0.5129716981132075,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979873297891141,
      "loss": 5.4965,
      "mean_token_accuracy": 0.16972004026174545,
      "num_tokens": 11757819.0,
      "step": 5655
    },
    {
      "entropy": 5.650430011749267,
      "epoch": 0.5134252539912917,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979830093394372,
      "loss": 5.4349,
      "mean_token_accuracy": 0.17406515330076217,
      "num_tokens": 11767470.0,
      "step": 5660
    },
    {
      "entropy": 5.574281358718872,
      "epoch": 0.513878809869376,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979786842784149,
      "loss": 5.3549,
      "mean_token_accuracy": 0.1825924515724182,
      "num_tokens": 11777060.0,
      "step": 5665
    },
    {
      "entropy": 5.566286706924439,
      "epoch": 0.5143323657474601,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979743546061366,
      "loss": 5.4107,
      "mean_token_accuracy": 0.17651544213294984,
      "num_tokens": 11787176.0,
      "step": 5670
    },
    {
      "entropy": 5.5616296291351315,
      "epoch": 0.5147859216255443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979700203226918,
      "loss": 5.2849,
      "mean_token_accuracy": 0.17933404445648193,
      "num_tokens": 11796419.0,
      "step": 5675
    },
    {
      "entropy": 5.682236576080323,
      "epoch": 0.5152394775036284,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979656814281702,
      "loss": 5.5213,
      "mean_token_accuracy": 0.1706225335597992,
      "num_tokens": 11807279.0,
      "step": 5680
    },
    {
      "entropy": 5.606097984313965,
      "epoch": 0.5156930333817126,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979613379226615,
      "loss": 5.4088,
      "mean_token_accuracy": 0.1719408765435219,
      "num_tokens": 11817669.0,
      "step": 5685
    },
    {
      "entropy": 5.544135332107544,
      "epoch": 0.5161465892597968,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979569898062554,
      "loss": 5.415,
      "mean_token_accuracy": 0.179472017288208,
      "num_tokens": 11827247.0,
      "step": 5690
    },
    {
      "entropy": 5.619077444076538,
      "epoch": 0.516600145137881,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497952637079042,
      "loss": 5.402,
      "mean_token_accuracy": 0.17077631801366805,
      "num_tokens": 11838201.0,
      "step": 5695
    },
    {
      "entropy": 5.608509302139282,
      "epoch": 0.5170537010159652,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979482797411114,
      "loss": 5.3538,
      "mean_token_accuracy": 0.18213389366865157,
      "num_tokens": 11847561.0,
      "step": 5700
    },
    {
      "entropy": 5.63592209815979,
      "epoch": 0.5175072568940493,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979439177925534,
      "loss": 5.4601,
      "mean_token_accuracy": 0.17978523373603822,
      "num_tokens": 11858272.0,
      "step": 5705
    },
    {
      "entropy": 5.664514636993408,
      "epoch": 0.5179608127721336,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979395512334584,
      "loss": 5.4897,
      "mean_token_accuracy": 0.1710379973053932,
      "num_tokens": 11869660.0,
      "step": 5710
    },
    {
      "entropy": 5.657863283157349,
      "epoch": 0.5184143686502177,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979351800639166,
      "loss": 5.522,
      "mean_token_accuracy": 0.16624149084091186,
      "num_tokens": 11880239.0,
      "step": 5715
    },
    {
      "entropy": 5.723998975753784,
      "epoch": 0.5188679245283019,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979308042840186,
      "loss": 5.5019,
      "mean_token_accuracy": 0.16436144411563874,
      "num_tokens": 11890630.0,
      "step": 5720
    },
    {
      "entropy": 5.692220687866211,
      "epoch": 0.519321480406386,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979264238938547,
      "loss": 5.5387,
      "mean_token_accuracy": 0.16605993658304213,
      "num_tokens": 11900948.0,
      "step": 5725
    },
    {
      "entropy": 5.5566572666168215,
      "epoch": 0.5197750362844702,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979220388935155,
      "loss": 5.4197,
      "mean_token_accuracy": 0.1711593523621559,
      "num_tokens": 11912528.0,
      "step": 5730
    },
    {
      "entropy": 5.620328950881958,
      "epoch": 0.5202285921625545,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979176492830918,
      "loss": 5.36,
      "mean_token_accuracy": 0.1769677922129631,
      "num_tokens": 11922211.0,
      "step": 5735
    },
    {
      "entropy": 5.6201256275177,
      "epoch": 0.5206821480406386,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979132550626742,
      "loss": 5.4102,
      "mean_token_accuracy": 0.1742950886487961,
      "num_tokens": 11932520.0,
      "step": 5740
    },
    {
      "entropy": 5.587960624694825,
      "epoch": 0.5211357039187228,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979088562323538,
      "loss": 5.3434,
      "mean_token_accuracy": 0.17956214398145676,
      "num_tokens": 11943312.0,
      "step": 5745
    },
    {
      "entropy": 5.564781379699707,
      "epoch": 0.5215892597968069,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979044527922214,
      "loss": 5.4926,
      "mean_token_accuracy": 0.1721508026123047,
      "num_tokens": 11954455.0,
      "step": 5750
    },
    {
      "entropy": 5.6627356052398685,
      "epoch": 0.5220428156748912,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979000447423681,
      "loss": 5.3863,
      "mean_token_accuracy": 0.17926096618175508,
      "num_tokens": 11964650.0,
      "step": 5755
    },
    {
      "entropy": 5.575634002685547,
      "epoch": 0.5224963715529753,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978956320828851,
      "loss": 5.4195,
      "mean_token_accuracy": 0.1762705847620964,
      "num_tokens": 11975134.0,
      "step": 5760
    },
    {
      "entropy": 5.606097459793091,
      "epoch": 0.5229499274310595,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978912148138637,
      "loss": 5.4532,
      "mean_token_accuracy": 0.17324555963277816,
      "num_tokens": 11986486.0,
      "step": 5765
    },
    {
      "entropy": 5.678210592269897,
      "epoch": 0.5234034833091437,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497886792935395,
      "loss": 5.4704,
      "mean_token_accuracy": 0.1706312969326973,
      "num_tokens": 11998045.0,
      "step": 5770
    },
    {
      "entropy": 5.717112493515015,
      "epoch": 0.5238570391872278,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978823664475707,
      "loss": 5.4668,
      "mean_token_accuracy": 0.17400625944137574,
      "num_tokens": 12008130.0,
      "step": 5775
    },
    {
      "entropy": 5.6386171817779545,
      "epoch": 0.5243105950653121,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978779353504824,
      "loss": 5.5292,
      "mean_token_accuracy": 0.1656391978263855,
      "num_tokens": 12018753.0,
      "step": 5780
    },
    {
      "entropy": 5.631770372390747,
      "epoch": 0.5247641509433962,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978734996442214,
      "loss": 5.4048,
      "mean_token_accuracy": 0.17983568012714385,
      "num_tokens": 12028202.0,
      "step": 5785
    },
    {
      "entropy": 5.612694168090821,
      "epoch": 0.5252177068214804,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978690593288798,
      "loss": 5.4843,
      "mean_token_accuracy": 0.17659876495599747,
      "num_tokens": 12039652.0,
      "step": 5790
    },
    {
      "entropy": 5.678992748260498,
      "epoch": 0.5256712626995645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978646144045491,
      "loss": 5.3938,
      "mean_token_accuracy": 0.16736219227313995,
      "num_tokens": 12051269.0,
      "step": 5795
    },
    {
      "entropy": 5.46131854057312,
      "epoch": 0.5261248185776488,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978601648713215,
      "loss": 5.291,
      "mean_token_accuracy": 0.17717837989330293,
      "num_tokens": 12060647.0,
      "step": 5800
    },
    {
      "entropy": 5.585661935806274,
      "epoch": 0.526578374455733,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978557107292889,
      "loss": 5.4256,
      "mean_token_accuracy": 0.17812251448631286,
      "num_tokens": 12071669.0,
      "step": 5805
    },
    {
      "entropy": 5.561780834197998,
      "epoch": 0.5270319303338171,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978512519785433,
      "loss": 5.3141,
      "mean_token_accuracy": 0.1775515452027321,
      "num_tokens": 12081971.0,
      "step": 5810
    },
    {
      "entropy": 5.567921876907349,
      "epoch": 0.5274854862119013,
      "grad_norm": 0.875,
      "learning_rate": 0.000497846788619177,
      "loss": 5.4346,
      "mean_token_accuracy": 0.18071742057800294,
      "num_tokens": 12092465.0,
      "step": 5815
    },
    {
      "entropy": 5.65427074432373,
      "epoch": 0.5279390420899854,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978423206512825,
      "loss": 5.3935,
      "mean_token_accuracy": 0.17432788610458375,
      "num_tokens": 12101752.0,
      "step": 5820
    },
    {
      "entropy": 5.673962211608886,
      "epoch": 0.5283925979680697,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497837848074952,
      "loss": 5.4643,
      "mean_token_accuracy": 0.17431971430778503,
      "num_tokens": 12111805.0,
      "step": 5825
    },
    {
      "entropy": 5.694564437866211,
      "epoch": 0.5288461538461539,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497833370890278,
      "loss": 5.4918,
      "mean_token_accuracy": 0.17020126730203627,
      "num_tokens": 12122511.0,
      "step": 5830
    },
    {
      "entropy": 5.607981729507446,
      "epoch": 0.529299709724238,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978288890973529,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1757800757884979,
      "num_tokens": 12132510.0,
      "step": 5835
    },
    {
      "entropy": 5.561420106887818,
      "epoch": 0.5297532656023222,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978244026962698,
      "loss": 5.3859,
      "mean_token_accuracy": 0.17332411408424378,
      "num_tokens": 12142419.0,
      "step": 5840
    },
    {
      "entropy": 5.6124217987060545,
      "epoch": 0.5302068214804064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978199116871212,
      "loss": 5.4348,
      "mean_token_accuracy": 0.17327335476875305,
      "num_tokens": 12152759.0,
      "step": 5845
    },
    {
      "entropy": 5.605248260498047,
      "epoch": 0.5306603773584906,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00049781541607,
      "loss": 5.3847,
      "mean_token_accuracy": 0.17561568468809127,
      "num_tokens": 12162758.0,
      "step": 5850
    },
    {
      "entropy": 5.59306902885437,
      "epoch": 0.5311139332365747,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978109158449993,
      "loss": 5.444,
      "mean_token_accuracy": 0.17255426645278932,
      "num_tokens": 12172664.0,
      "step": 5855
    },
    {
      "entropy": 5.647549057006836,
      "epoch": 0.5315674891146589,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978064110122121,
      "loss": 5.402,
      "mean_token_accuracy": 0.17630172818899154,
      "num_tokens": 12182028.0,
      "step": 5860
    },
    {
      "entropy": 5.641605806350708,
      "epoch": 0.532021044992743,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978019015717314,
      "loss": 5.4504,
      "mean_token_accuracy": 0.17413344085216523,
      "num_tokens": 12192168.0,
      "step": 5865
    },
    {
      "entropy": 5.551989269256592,
      "epoch": 0.5324746008708273,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977973875236508,
      "loss": 5.3544,
      "mean_token_accuracy": 0.1735835164785385,
      "num_tokens": 12202889.0,
      "step": 5870
    },
    {
      "entropy": 5.55653338432312,
      "epoch": 0.5329281567489115,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977928688680634,
      "loss": 5.3953,
      "mean_token_accuracy": 0.17994537502527236,
      "num_tokens": 12212514.0,
      "step": 5875
    },
    {
      "entropy": 5.547123765945434,
      "epoch": 0.5333817126269956,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977883456050626,
      "loss": 5.2653,
      "mean_token_accuracy": 0.18208599984645843,
      "num_tokens": 12222353.0,
      "step": 5880
    },
    {
      "entropy": 5.6921580791473385,
      "epoch": 0.5338352685050798,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497783817734742,
      "loss": 5.5429,
      "mean_token_accuracy": 0.16759776920080185,
      "num_tokens": 12232845.0,
      "step": 5885
    },
    {
      "entropy": 5.606061315536499,
      "epoch": 0.534288824383164,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977792852571954,
      "loss": 5.4554,
      "mean_token_accuracy": 0.16563831120729447,
      "num_tokens": 12242950.0,
      "step": 5890
    },
    {
      "entropy": 5.595444107055664,
      "epoch": 0.5347423802612482,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977747481725164,
      "loss": 5.4387,
      "mean_token_accuracy": 0.17201048582792283,
      "num_tokens": 12253132.0,
      "step": 5895
    },
    {
      "entropy": 5.632798957824707,
      "epoch": 0.5351959361393324,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977702064807987,
      "loss": 5.395,
      "mean_token_accuracy": 0.17704534381628037,
      "num_tokens": 12263214.0,
      "step": 5900
    },
    {
      "entropy": 5.541617584228516,
      "epoch": 0.5356494920174165,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977656601821366,
      "loss": 5.3208,
      "mean_token_accuracy": 0.1798873171210289,
      "num_tokens": 12273250.0,
      "step": 5905
    },
    {
      "entropy": 5.7590607643127445,
      "epoch": 0.5361030478955007,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977611092766238,
      "loss": 5.5101,
      "mean_token_accuracy": 0.1787897080183029,
      "num_tokens": 12283767.0,
      "step": 5910
    },
    {
      "entropy": 5.534255886077881,
      "epoch": 0.5365566037735849,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004977565537643544,
      "loss": 5.3581,
      "mean_token_accuracy": 0.17475240379571916,
      "num_tokens": 12294612.0,
      "step": 5915
    },
    {
      "entropy": 5.534429550170898,
      "epoch": 0.5370101596516691,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977519936454227,
      "loss": 5.4397,
      "mean_token_accuracy": 0.17676630169153212,
      "num_tokens": 12305614.0,
      "step": 5920
    },
    {
      "entropy": 5.567504644393921,
      "epoch": 0.5374637155297532,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977474289199231,
      "loss": 5.3132,
      "mean_token_accuracy": 0.1735585629940033,
      "num_tokens": 12316152.0,
      "step": 5925
    },
    {
      "entropy": 5.551148366928101,
      "epoch": 0.5379172714078374,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977428595879499,
      "loss": 5.3905,
      "mean_token_accuracy": 0.178743177652359,
      "num_tokens": 12325940.0,
      "step": 5930
    },
    {
      "entropy": 5.566306829452515,
      "epoch": 0.5383708272859217,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004977382856495975,
      "loss": 5.39,
      "mean_token_accuracy": 0.17976804822683334,
      "num_tokens": 12335482.0,
      "step": 5935
    },
    {
      "entropy": 5.655665397644043,
      "epoch": 0.5388243831640058,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977337071049606,
      "loss": 5.4884,
      "mean_token_accuracy": 0.1725362166762352,
      "num_tokens": 12346386.0,
      "step": 5940
    },
    {
      "entropy": 5.696004199981689,
      "epoch": 0.53927793904209,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497729123954134,
      "loss": 5.4451,
      "mean_token_accuracy": 0.17383963018655776,
      "num_tokens": 12356282.0,
      "step": 5945
    },
    {
      "entropy": 5.601561498641968,
      "epoch": 0.5397314949201741,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977245361972123,
      "loss": 5.4141,
      "mean_token_accuracy": 0.17678642570972442,
      "num_tokens": 12366994.0,
      "step": 5950
    },
    {
      "entropy": 5.545989513397217,
      "epoch": 0.5401850507982583,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977199438342904,
      "loss": 5.3017,
      "mean_token_accuracy": 0.18050941973924636,
      "num_tokens": 12377524.0,
      "step": 5955
    },
    {
      "entropy": 5.595731210708618,
      "epoch": 0.5406386066763426,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977153468654634,
      "loss": 5.4655,
      "mean_token_accuracy": 0.17503209561109542,
      "num_tokens": 12387593.0,
      "step": 5960
    },
    {
      "entropy": 5.697256326675415,
      "epoch": 0.5410921625544267,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977107452908262,
      "loss": 5.4273,
      "mean_token_accuracy": 0.1787885919213295,
      "num_tokens": 12398732.0,
      "step": 5965
    },
    {
      "entropy": 5.61314206123352,
      "epoch": 0.5415457184325109,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497706139110474,
      "loss": 5.4252,
      "mean_token_accuracy": 0.18124329149723054,
      "num_tokens": 12409821.0,
      "step": 5970
    },
    {
      "entropy": 5.593740558624267,
      "epoch": 0.541999274310595,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977015283245021,
      "loss": 5.4482,
      "mean_token_accuracy": 0.17267510890960694,
      "num_tokens": 12420044.0,
      "step": 5975
    },
    {
      "entropy": 5.538191413879394,
      "epoch": 0.5424528301886793,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976969129330059,
      "loss": 5.3615,
      "mean_token_accuracy": 0.17179904878139496,
      "num_tokens": 12430980.0,
      "step": 5980
    },
    {
      "entropy": 5.598214149475098,
      "epoch": 0.5429063860667634,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976922929360808,
      "loss": 5.4067,
      "mean_token_accuracy": 0.1799796149134636,
      "num_tokens": 12441999.0,
      "step": 5985
    },
    {
      "entropy": 5.655162572860718,
      "epoch": 0.5433599419448476,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976876683338222,
      "loss": 5.4457,
      "mean_token_accuracy": 0.1742375373840332,
      "num_tokens": 12452837.0,
      "step": 5990
    },
    {
      "entropy": 5.573116827011108,
      "epoch": 0.5438134978229318,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497683039126326,
      "loss": 5.3746,
      "mean_token_accuracy": 0.17719283998012542,
      "num_tokens": 12462142.0,
      "step": 5995
    },
    {
      "entropy": 5.566186046600341,
      "epoch": 0.5442670537010159,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976784053136878,
      "loss": 5.4094,
      "mean_token_accuracy": 0.1800991877913475,
      "num_tokens": 12472756.0,
      "step": 6000
    },
    {
      "epoch": 0.5442670537010159,
      "eval_entropy": 5.364191224024846,
      "eval_loss": 5.434150695800781,
      "eval_mean_token_accuracy": 0.18153274842238737,
      "eval_num_tokens": 12472756.0,
      "eval_runtime": 20.5722,
      "eval_samples_per_second": 1718.825,
      "eval_steps_per_second": 214.853,
      "step": 6000
    },
    {
      "entropy": 5.625031852722168,
      "epoch": 0.5447206095791002,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004976737668960034,
      "loss": 5.5318,
      "mean_token_accuracy": 0.1683395102620125,
      "num_tokens": 12483115.0,
      "step": 6005
    },
    {
      "entropy": 5.665145683288574,
      "epoch": 0.5451741654571843,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004976691238733688,
      "loss": 5.3863,
      "mean_token_accuracy": 0.17331260144710542,
      "num_tokens": 12493280.0,
      "step": 6010
    },
    {
      "entropy": 5.628579378128052,
      "epoch": 0.5456277213352685,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049766447624588,
      "loss": 5.5164,
      "mean_token_accuracy": 0.17446643561124803,
      "num_tokens": 12503665.0,
      "step": 6015
    },
    {
      "entropy": 5.585593891143799,
      "epoch": 0.5460812772133526,
      "grad_norm": 1.0,
      "learning_rate": 0.000497659824013633,
      "loss": 5.4308,
      "mean_token_accuracy": 0.17111265659332275,
      "num_tokens": 12513381.0,
      "step": 6020
    },
    {
      "entropy": 5.681701898574829,
      "epoch": 0.5465348330914369,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976551671767241,
      "loss": 5.4617,
      "mean_token_accuracy": 0.16975319087505342,
      "num_tokens": 12524121.0,
      "step": 6025
    },
    {
      "entropy": 5.595137882232666,
      "epoch": 0.5469883889695211,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004976505057352496,
      "loss": 5.4145,
      "mean_token_accuracy": 0.17792607247829437,
      "num_tokens": 12533972.0,
      "step": 6030
    },
    {
      "entropy": 5.633500862121582,
      "epoch": 0.5474419448476052,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976458396893059,
      "loss": 5.5324,
      "mean_token_accuracy": 0.1650451496243477,
      "num_tokens": 12544110.0,
      "step": 6035
    },
    {
      "entropy": 5.641628408432007,
      "epoch": 0.5478955007256894,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976411690389896,
      "loss": 5.3603,
      "mean_token_accuracy": 0.17129435241222382,
      "num_tokens": 12554273.0,
      "step": 6040
    },
    {
      "entropy": 5.655753946304321,
      "epoch": 0.5483490566037735,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497636493784397,
      "loss": 5.4287,
      "mean_token_accuracy": 0.17427540570497513,
      "num_tokens": 12564633.0,
      "step": 6045
    },
    {
      "entropy": 5.596440839767456,
      "epoch": 0.5488026124818578,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976318139256251,
      "loss": 5.3635,
      "mean_token_accuracy": 0.1775827169418335,
      "num_tokens": 12573790.0,
      "step": 6050
    },
    {
      "entropy": 5.63446683883667,
      "epoch": 0.549256168359942,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004976271294627706,
      "loss": 5.441,
      "mean_token_accuracy": 0.17123547047376633,
      "num_tokens": 12584265.0,
      "step": 6055
    },
    {
      "entropy": 5.5846747875213625,
      "epoch": 0.5497097242380261,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976224403959303,
      "loss": 5.3174,
      "mean_token_accuracy": 0.17618338912725448,
      "num_tokens": 12594413.0,
      "step": 6060
    },
    {
      "entropy": 5.568248081207275,
      "epoch": 0.5501632801161103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976177467252012,
      "loss": 5.3923,
      "mean_token_accuracy": 0.1851392775774002,
      "num_tokens": 12604637.0,
      "step": 6065
    },
    {
      "entropy": 5.624586772918701,
      "epoch": 0.5506168359941945,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976130484506804,
      "loss": 5.4582,
      "mean_token_accuracy": 0.17681263536214828,
      "num_tokens": 12614873.0,
      "step": 6070
    },
    {
      "entropy": 5.648846769332886,
      "epoch": 0.5510703918722787,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976083455724649,
      "loss": 5.3823,
      "mean_token_accuracy": 0.17441451996564866,
      "num_tokens": 12626182.0,
      "step": 6075
    },
    {
      "entropy": 5.535418891906739,
      "epoch": 0.5515239477503628,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004976036380906523,
      "loss": 5.3149,
      "mean_token_accuracy": 0.17663441896438598,
      "num_tokens": 12635418.0,
      "step": 6080
    },
    {
      "entropy": 5.582258033752441,
      "epoch": 0.551977503628447,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975989260053395,
      "loss": 5.4765,
      "mean_token_accuracy": 0.16567691415548325,
      "num_tokens": 12645942.0,
      "step": 6085
    },
    {
      "entropy": 5.617289209365845,
      "epoch": 0.5524310595065312,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004975942093166242,
      "loss": 5.3982,
      "mean_token_accuracy": 0.17294339835643768,
      "num_tokens": 12656460.0,
      "step": 6090
    },
    {
      "entropy": 5.636151075363159,
      "epoch": 0.5528846153846154,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975894880246041,
      "loss": 5.3228,
      "mean_token_accuracy": 0.1788621500134468,
      "num_tokens": 12666916.0,
      "step": 6095
    },
    {
      "entropy": 5.526174592971802,
      "epoch": 0.5533381712626996,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975847621293766,
      "loss": 5.3853,
      "mean_token_accuracy": 0.18073209524154663,
      "num_tokens": 12677562.0,
      "step": 6100
    },
    {
      "entropy": 5.550758695602417,
      "epoch": 0.5537917271407837,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975800316310394,
      "loss": 5.3618,
      "mean_token_accuracy": 0.1773673951625824,
      "num_tokens": 12687349.0,
      "step": 6105
    },
    {
      "entropy": 5.643798589706421,
      "epoch": 0.5542452830188679,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004975752965296906,
      "loss": 5.4098,
      "mean_token_accuracy": 0.17631264328956603,
      "num_tokens": 12697850.0,
      "step": 6110
    },
    {
      "entropy": 5.540865325927735,
      "epoch": 0.5546988388969522,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004975705568254278,
      "loss": 5.2774,
      "mean_token_accuracy": 0.17912191450595855,
      "num_tokens": 12708283.0,
      "step": 6115
    },
    {
      "entropy": 5.57166051864624,
      "epoch": 0.5551523947750363,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975658125183493,
      "loss": 5.3814,
      "mean_token_accuracy": 0.17749570310115814,
      "num_tokens": 12718201.0,
      "step": 6120
    },
    {
      "entropy": 5.551891422271728,
      "epoch": 0.5556059506531205,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497561063608553,
      "loss": 5.3217,
      "mean_token_accuracy": 0.18132489323616027,
      "num_tokens": 12728860.0,
      "step": 6125
    },
    {
      "entropy": 5.574612474441528,
      "epoch": 0.5560595065312046,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004975563100961372,
      "loss": 5.3979,
      "mean_token_accuracy": 0.17559077590703964,
      "num_tokens": 12739007.0,
      "step": 6130
    },
    {
      "entropy": 5.578898191452026,
      "epoch": 0.5565130624092888,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975515519812002,
      "loss": 5.2461,
      "mean_token_accuracy": 0.1819448158144951,
      "num_tokens": 12749004.0,
      "step": 6135
    },
    {
      "entropy": 5.543379449844361,
      "epoch": 0.556966618287373,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975467892638402,
      "loss": 5.3959,
      "mean_token_accuracy": 0.17813464105129242,
      "num_tokens": 12760085.0,
      "step": 6140
    },
    {
      "entropy": 5.59183669090271,
      "epoch": 0.5574201741654572,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004975420219441561,
      "loss": 5.3429,
      "mean_token_accuracy": 0.18399171233177186,
      "num_tokens": 12770303.0,
      "step": 6145
    },
    {
      "entropy": 5.58007698059082,
      "epoch": 0.5578737300435413,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975372500222462,
      "loss": 5.3724,
      "mean_token_accuracy": 0.17818562537431717,
      "num_tokens": 12780334.0,
      "step": 6150
    },
    {
      "entropy": 5.587021446228027,
      "epoch": 0.5583272859216255,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975324734982094,
      "loss": 5.4218,
      "mean_token_accuracy": 0.17108138799667358,
      "num_tokens": 12790980.0,
      "step": 6155
    },
    {
      "entropy": 5.621369981765747,
      "epoch": 0.5587808417997098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004975276923721442,
      "loss": 5.4611,
      "mean_token_accuracy": 0.17504812330007552,
      "num_tokens": 12801713.0,
      "step": 6160
    },
    {
      "entropy": 5.644250059127808,
      "epoch": 0.5592343976777939,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004975229066441496,
      "loss": 5.4677,
      "mean_token_accuracy": 0.1687561482191086,
      "num_tokens": 12813040.0,
      "step": 6165
    },
    {
      "entropy": 5.577824926376342,
      "epoch": 0.5596879535558781,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975181163143247,
      "loss": 5.4127,
      "mean_token_accuracy": 0.17071571797132493,
      "num_tokens": 12824048.0,
      "step": 6170
    },
    {
      "entropy": 5.64132513999939,
      "epoch": 0.5601415094339622,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975133213827684,
      "loss": 5.3687,
      "mean_token_accuracy": 0.17483995854854584,
      "num_tokens": 12834297.0,
      "step": 6175
    },
    {
      "entropy": 5.66928539276123,
      "epoch": 0.5605950653120464,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975085218495798,
      "loss": 5.4364,
      "mean_token_accuracy": 0.17078946828842162,
      "num_tokens": 12844574.0,
      "step": 6180
    },
    {
      "entropy": 5.536151313781739,
      "epoch": 0.5610486211901307,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975037177148585,
      "loss": 5.3661,
      "mean_token_accuracy": 0.177015121281147,
      "num_tokens": 12854670.0,
      "step": 6185
    },
    {
      "entropy": 5.536833953857422,
      "epoch": 0.5615021770682148,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974989089787035,
      "loss": 5.3828,
      "mean_token_accuracy": 0.17939384132623673,
      "num_tokens": 12865281.0,
      "step": 6190
    },
    {
      "entropy": 5.683811044692993,
      "epoch": 0.561955732946299,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974940956412143,
      "loss": 5.3903,
      "mean_token_accuracy": 0.1783251017332077,
      "num_tokens": 12875222.0,
      "step": 6195
    },
    {
      "entropy": 5.7399190902709964,
      "epoch": 0.5624092888243831,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974892777024905,
      "loss": 5.4623,
      "mean_token_accuracy": 0.16634879112243653,
      "num_tokens": 12885079.0,
      "step": 6200
    },
    {
      "entropy": 5.523637962341309,
      "epoch": 0.5628628447024674,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004974844551626318,
      "loss": 5.3998,
      "mean_token_accuracy": 0.1781408593058586,
      "num_tokens": 12896402.0,
      "step": 6205
    },
    {
      "entropy": 5.585960006713867,
      "epoch": 0.5633164005805515,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974796280217379,
      "loss": 5.3653,
      "mean_token_accuracy": 0.18042189478874207,
      "num_tokens": 12906403.0,
      "step": 6210
    },
    {
      "entropy": 5.643087482452392,
      "epoch": 0.5637699564586357,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974747962799086,
      "loss": 5.3895,
      "mean_token_accuracy": 0.17042631208896636,
      "num_tokens": 12917061.0,
      "step": 6215
    },
    {
      "entropy": 5.540905570983886,
      "epoch": 0.5642235123367199,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974699599372437,
      "loss": 5.427,
      "mean_token_accuracy": 0.1783065214753151,
      "num_tokens": 12927182.0,
      "step": 6220
    },
    {
      "entropy": 5.561410760879516,
      "epoch": 0.564677068214804,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004974651189938434,
      "loss": 5.3857,
      "mean_token_accuracy": 0.1792475163936615,
      "num_tokens": 12937371.0,
      "step": 6225
    },
    {
      "entropy": 5.633272504806518,
      "epoch": 0.5651306240928883,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974602734498079,
      "loss": 5.4352,
      "mean_token_accuracy": 0.1712818294763565,
      "num_tokens": 12948569.0,
      "step": 6230
    },
    {
      "entropy": 5.621938037872314,
      "epoch": 0.5655841799709724,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000497455423305237,
      "loss": 5.3768,
      "mean_token_accuracy": 0.17150397896766661,
      "num_tokens": 12958503.0,
      "step": 6235
    },
    {
      "entropy": 5.601767444610596,
      "epoch": 0.5660377358490566,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974505685602314,
      "loss": 5.3988,
      "mean_token_accuracy": 0.17525841742753984,
      "num_tokens": 12968714.0,
      "step": 6240
    },
    {
      "entropy": 5.590086603164673,
      "epoch": 0.5664912917271407,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974457092148912,
      "loss": 5.4284,
      "mean_token_accuracy": 0.1741039887070656,
      "num_tokens": 12978857.0,
      "step": 6245
    },
    {
      "entropy": 5.720824956893921,
      "epoch": 0.566944847605225,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974408452693172,
      "loss": 5.5322,
      "mean_token_accuracy": 0.16714059710502624,
      "num_tokens": 12989863.0,
      "step": 6250
    },
    {
      "entropy": 5.586817884445191,
      "epoch": 0.5673984034833092,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974359767236099,
      "loss": 5.3248,
      "mean_token_accuracy": 0.17573193907737733,
      "num_tokens": 13000307.0,
      "step": 6255
    },
    {
      "entropy": 5.545666122436524,
      "epoch": 0.5678519593613933,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004974311035778698,
      "loss": 5.411,
      "mean_token_accuracy": 0.18314283788204194,
      "num_tokens": 13009053.0,
      "step": 6260
    },
    {
      "entropy": 5.5987457752227785,
      "epoch": 0.5683055152394775,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974262258321979,
      "loss": 5.4269,
      "mean_token_accuracy": 0.1716617226600647,
      "num_tokens": 13019136.0,
      "step": 6265
    },
    {
      "entropy": 5.655656051635742,
      "epoch": 0.5687590711175616,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004974213434866949,
      "loss": 5.4723,
      "mean_token_accuracy": 0.16619940251111984,
      "num_tokens": 13030356.0,
      "step": 6270
    },
    {
      "entropy": 5.637882804870605,
      "epoch": 0.5692126269956459,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974164565414617,
      "loss": 5.4883,
      "mean_token_accuracy": 0.1748768150806427,
      "num_tokens": 13040477.0,
      "step": 6275
    },
    {
      "entropy": 5.596619510650635,
      "epoch": 0.56966618287373,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004974115649965998,
      "loss": 5.3524,
      "mean_token_accuracy": 0.18151768296957016,
      "num_tokens": 13051755.0,
      "step": 6280
    },
    {
      "entropy": 5.604301977157593,
      "epoch": 0.5701197387518142,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974066688522098,
      "loss": 5.3913,
      "mean_token_accuracy": 0.18713292777538298,
      "num_tokens": 13061670.0,
      "step": 6285
    },
    {
      "entropy": 5.60211033821106,
      "epoch": 0.5705732946298984,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974017681083934,
      "loss": 5.3646,
      "mean_token_accuracy": 0.1776226207613945,
      "num_tokens": 13070917.0,
      "step": 6290
    },
    {
      "entropy": 5.585918712615967,
      "epoch": 0.5710268505079826,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004973968627652516,
      "loss": 5.4146,
      "mean_token_accuracy": 0.17641005665063858,
      "num_tokens": 13080577.0,
      "step": 6295
    },
    {
      "entropy": 5.6024633884429935,
      "epoch": 0.5714804063860668,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497391952822886,
      "loss": 5.4371,
      "mean_token_accuracy": 0.16923494786024093,
      "num_tokens": 13091446.0,
      "step": 6300
    },
    {
      "entropy": 5.649172449111939,
      "epoch": 0.5719339622641509,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973870382813983,
      "loss": 5.4614,
      "mean_token_accuracy": 0.17273851484060287,
      "num_tokens": 13102752.0,
      "step": 6305
    },
    {
      "entropy": 5.702681827545166,
      "epoch": 0.5723875181422351,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973821191408899,
      "loss": 5.4819,
      "mean_token_accuracy": 0.17141962498426438,
      "num_tokens": 13113944.0,
      "step": 6310
    },
    {
      "entropy": 5.622271108627319,
      "epoch": 0.5728410740203193,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973771954014626,
      "loss": 5.4073,
      "mean_token_accuracy": 0.17560603171586991,
      "num_tokens": 13124316.0,
      "step": 6315
    },
    {
      "entropy": 5.6871038436889645,
      "epoch": 0.5732946298984035,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004973722670632182,
      "loss": 5.5038,
      "mean_token_accuracy": 0.16931897699832915,
      "num_tokens": 13133567.0,
      "step": 6320
    },
    {
      "entropy": 5.574843692779541,
      "epoch": 0.5737481857764877,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973673341262587,
      "loss": 5.4081,
      "mean_token_accuracy": 0.17334946990013123,
      "num_tokens": 13144279.0,
      "step": 6325
    },
    {
      "entropy": 5.626540994644165,
      "epoch": 0.5742017416545718,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497362396590686,
      "loss": 5.4835,
      "mean_token_accuracy": 0.16946994364261628,
      "num_tokens": 13154463.0,
      "step": 6330
    },
    {
      "entropy": 5.65912938117981,
      "epoch": 0.574655297532656,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004973574544566024,
      "loss": 5.4457,
      "mean_token_accuracy": 0.16746120303869247,
      "num_tokens": 13165565.0,
      "step": 6335
    },
    {
      "entropy": 5.559062385559082,
      "epoch": 0.5751088534107403,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973525077241098,
      "loss": 5.3533,
      "mean_token_accuracy": 0.18145330846309662,
      "num_tokens": 13175839.0,
      "step": 6340
    },
    {
      "entropy": 5.5439225196838375,
      "epoch": 0.5755624092888244,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004973475563933109,
      "loss": 5.3421,
      "mean_token_accuracy": 0.17647033035755158,
      "num_tokens": 13186372.0,
      "step": 6345
    },
    {
      "entropy": 5.629715633392334,
      "epoch": 0.5760159651669086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973426004643077,
      "loss": 5.4044,
      "mean_token_accuracy": 0.18002848476171493,
      "num_tokens": 13196259.0,
      "step": 6350
    },
    {
      "entropy": 5.643421077728272,
      "epoch": 0.5764695210449927,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004973376399372029,
      "loss": 5.3786,
      "mean_token_accuracy": 0.1787116289138794,
      "num_tokens": 13206497.0,
      "step": 6355
    },
    {
      "entropy": 5.605941486358643,
      "epoch": 0.5769230769230769,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973326748120991,
      "loss": 5.4201,
      "mean_token_accuracy": 0.17948209047317504,
      "num_tokens": 13217188.0,
      "step": 6360
    },
    {
      "entropy": 5.575176668167114,
      "epoch": 0.5773766328011611,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973277050890988,
      "loss": 5.4289,
      "mean_token_accuracy": 0.17635478228330612,
      "num_tokens": 13226544.0,
      "step": 6365
    },
    {
      "entropy": 5.662482213973999,
      "epoch": 0.5778301886792453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973227307683049,
      "loss": 5.4469,
      "mean_token_accuracy": 0.17289083003997802,
      "num_tokens": 13236572.0,
      "step": 6370
    },
    {
      "entropy": 5.716796112060547,
      "epoch": 0.5782837445573294,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973177518498204,
      "loss": 5.4467,
      "mean_token_accuracy": 0.1772434666752815,
      "num_tokens": 13247958.0,
      "step": 6375
    },
    {
      "entropy": 5.598628664016724,
      "epoch": 0.5787373004354136,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973127683337482,
      "loss": 5.3668,
      "mean_token_accuracy": 0.17798337638378142,
      "num_tokens": 13258255.0,
      "step": 6380
    },
    {
      "entropy": 5.528913688659668,
      "epoch": 0.5791908563134979,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973077802201911,
      "loss": 5.3936,
      "mean_token_accuracy": 0.17277796119451522,
      "num_tokens": 13268416.0,
      "step": 6385
    },
    {
      "entropy": 5.572034215927124,
      "epoch": 0.579644412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973027875092526,
      "loss": 5.3724,
      "mean_token_accuracy": 0.17329427152872084,
      "num_tokens": 13278087.0,
      "step": 6390
    },
    {
      "entropy": 5.559347438812256,
      "epoch": 0.5800979680696662,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972977902010357,
      "loss": 5.2809,
      "mean_token_accuracy": 0.18577326983213424,
      "num_tokens": 13288121.0,
      "step": 6395
    },
    {
      "entropy": 5.696083450317383,
      "epoch": 0.5805515239477503,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972927882956439,
      "loss": 5.4854,
      "mean_token_accuracy": 0.16661122292280198,
      "num_tokens": 13298620.0,
      "step": 6400
    },
    {
      "entropy": 5.555794334411621,
      "epoch": 0.5810050798258345,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004972877817931807,
      "loss": 5.3054,
      "mean_token_accuracy": 0.17329084128141403,
      "num_tokens": 13308413.0,
      "step": 6405
    },
    {
      "entropy": 5.545690584182739,
      "epoch": 0.5814586357039188,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972827706937493,
      "loss": 5.3926,
      "mean_token_accuracy": 0.17548721879720688,
      "num_tokens": 13319387.0,
      "step": 6410
    },
    {
      "entropy": 5.60416431427002,
      "epoch": 0.5819121915820029,
      "grad_norm": 0.875,
      "learning_rate": 0.0004972777549974537,
      "loss": 5.3385,
      "mean_token_accuracy": 0.17935979068279267,
      "num_tokens": 13330494.0,
      "step": 6415
    },
    {
      "entropy": 5.593369865417481,
      "epoch": 0.5823657474600871,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972727347043975,
      "loss": 5.3587,
      "mean_token_accuracy": 0.17618109732866288,
      "num_tokens": 13341009.0,
      "step": 6420
    },
    {
      "entropy": 5.541242694854736,
      "epoch": 0.5828193033381712,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972677098146845,
      "loss": 5.3054,
      "mean_token_accuracy": 0.17940136939287185,
      "num_tokens": 13351269.0,
      "step": 6425
    },
    {
      "entropy": 5.521207714080811,
      "epoch": 0.5832728592162555,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972626803284185,
      "loss": 5.4117,
      "mean_token_accuracy": 0.17470855563879012,
      "num_tokens": 13362160.0,
      "step": 6430
    },
    {
      "entropy": 5.640540981292725,
      "epoch": 0.5837264150943396,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972576462457038,
      "loss": 5.4703,
      "mean_token_accuracy": 0.17759566456079484,
      "num_tokens": 13372497.0,
      "step": 6435
    },
    {
      "entropy": 5.697097206115723,
      "epoch": 0.5841799709724238,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972526075666442,
      "loss": 5.4133,
      "mean_token_accuracy": 0.18020133376121522,
      "num_tokens": 13381665.0,
      "step": 6440
    },
    {
      "entropy": 5.543485450744629,
      "epoch": 0.584633526850508,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972475642913442,
      "loss": 5.3197,
      "mean_token_accuracy": 0.18023221790790558,
      "num_tokens": 13392507.0,
      "step": 6445
    },
    {
      "entropy": 5.505417203903198,
      "epoch": 0.5850870827285921,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972425164199078,
      "loss": 5.3031,
      "mean_token_accuracy": 0.18267336785793303,
      "num_tokens": 13402171.0,
      "step": 6450
    },
    {
      "entropy": 5.613723850250244,
      "epoch": 0.5855406386066764,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004972374639524396,
      "loss": 5.525,
      "mean_token_accuracy": 0.16901201158761978,
      "num_tokens": 13412933.0,
      "step": 6455
    },
    {
      "entropy": 5.6370524883270265,
      "epoch": 0.5859941944847605,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000497232406889044,
      "loss": 5.3751,
      "mean_token_accuracy": 0.16950386315584182,
      "num_tokens": 13423563.0,
      "step": 6460
    },
    {
      "entropy": 5.6557940483093265,
      "epoch": 0.5864477503628447,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972273452298255,
      "loss": 5.3847,
      "mean_token_accuracy": 0.1834123909473419,
      "num_tokens": 13433998.0,
      "step": 6465
    },
    {
      "entropy": 5.559069585800171,
      "epoch": 0.5869013062409288,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972222789748888,
      "loss": 5.3315,
      "mean_token_accuracy": 0.18007088154554368,
      "num_tokens": 13443878.0,
      "step": 6470
    },
    {
      "entropy": 5.556202745437622,
      "epoch": 0.5873548621190131,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004972172081243389,
      "loss": 5.4097,
      "mean_token_accuracy": 0.1707281395792961,
      "num_tokens": 13456051.0,
      "step": 6475
    },
    {
      "entropy": 5.606028318405151,
      "epoch": 0.5878084179970973,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004972121326782806,
      "loss": 5.3974,
      "mean_token_accuracy": 0.17891442328691481,
      "num_tokens": 13467276.0,
      "step": 6480
    },
    {
      "entropy": 5.6323326587677,
      "epoch": 0.5882619738751814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972070526368186,
      "loss": 5.3717,
      "mean_token_accuracy": 0.17427799254655837,
      "num_tokens": 13478331.0,
      "step": 6485
    },
    {
      "entropy": 5.632275676727295,
      "epoch": 0.5887155297532656,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972019680000582,
      "loss": 5.3205,
      "mean_token_accuracy": 0.17633114755153656,
      "num_tokens": 13489879.0,
      "step": 6490
    },
    {
      "entropy": 5.563889217376709,
      "epoch": 0.5891690856313497,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004971968787681044,
      "loss": 5.4351,
      "mean_token_accuracy": 0.18205238431692122,
      "num_tokens": 13500879.0,
      "step": 6495
    },
    {
      "entropy": 5.674321556091309,
      "epoch": 0.589622641509434,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971917849410627,
      "loss": 5.4231,
      "mean_token_accuracy": 0.16968515366315842,
      "num_tokens": 13511719.0,
      "step": 6500
    },
    {
      "entropy": 5.5601965427398685,
      "epoch": 0.5900761973875182,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971866865190381,
      "loss": 5.3502,
      "mean_token_accuracy": 0.17365157157182692,
      "num_tokens": 13521850.0,
      "step": 6505
    },
    {
      "entropy": 5.614747524261475,
      "epoch": 0.5905297532656023,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971815835021362,
      "loss": 5.3834,
      "mean_token_accuracy": 0.18184393644332886,
      "num_tokens": 13532095.0,
      "step": 6510
    },
    {
      "entropy": 5.575075674057007,
      "epoch": 0.5909833091436865,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971764758904626,
      "loss": 5.3719,
      "mean_token_accuracy": 0.17141948640346527,
      "num_tokens": 13543133.0,
      "step": 6515
    },
    {
      "entropy": 5.587063646316528,
      "epoch": 0.5914368650217707,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004971713636841229,
      "loss": 5.438,
      "mean_token_accuracy": 0.17123900204896927,
      "num_tokens": 13553771.0,
      "step": 6520
    },
    {
      "entropy": 5.544393587112427,
      "epoch": 0.5918904208998549,
      "grad_norm": 0.875,
      "learning_rate": 0.0004971662468832228,
      "loss": 5.3763,
      "mean_token_accuracy": 0.1815660074353218,
      "num_tokens": 13564548.0,
      "step": 6525
    },
    {
      "entropy": 5.5182641506195065,
      "epoch": 0.592343976777939,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497161125487868,
      "loss": 5.3729,
      "mean_token_accuracy": 0.17794521152973175,
      "num_tokens": 13574567.0,
      "step": 6530
    },
    {
      "entropy": 5.551377677917481,
      "epoch": 0.5927975326560232,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971559994981645,
      "loss": 5.3972,
      "mean_token_accuracy": 0.17787606418132781,
      "num_tokens": 13585302.0,
      "step": 6535
    },
    {
      "entropy": 5.653764629364014,
      "epoch": 0.5932510885341074,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004971508689142184,
      "loss": 5.428,
      "mean_token_accuracy": 0.17052206993103028,
      "num_tokens": 13594754.0,
      "step": 6540
    },
    {
      "entropy": 5.5617927551269535,
      "epoch": 0.5937046444121916,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971457337361356,
      "loss": 5.3676,
      "mean_token_accuracy": 0.17397885471582414,
      "num_tokens": 13604691.0,
      "step": 6545
    },
    {
      "entropy": 5.476242637634277,
      "epoch": 0.5941582002902758,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971405939640225,
      "loss": 5.2587,
      "mean_token_accuracy": 0.17924531549215317,
      "num_tokens": 13613911.0,
      "step": 6550
    },
    {
      "entropy": 5.727000761032104,
      "epoch": 0.5946117561683599,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971354495979854,
      "loss": 5.4647,
      "mean_token_accuracy": 0.17712009847164153,
      "num_tokens": 13624285.0,
      "step": 6555
    },
    {
      "entropy": 5.584723997116089,
      "epoch": 0.5950653120464441,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971303006381305,
      "loss": 5.3066,
      "mean_token_accuracy": 0.18122798055410386,
      "num_tokens": 13635434.0,
      "step": 6560
    },
    {
      "entropy": 5.511235284805298,
      "epoch": 0.5955188679245284,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971251470845643,
      "loss": 5.3757,
      "mean_token_accuracy": 0.17572649121284484,
      "num_tokens": 13645582.0,
      "step": 6565
    },
    {
      "entropy": 5.580762100219727,
      "epoch": 0.5959724238026125,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004971199889373935,
      "loss": 5.3365,
      "mean_token_accuracy": 0.179083950817585,
      "num_tokens": 13655029.0,
      "step": 6570
    },
    {
      "entropy": 5.564962148666382,
      "epoch": 0.5964259796806967,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971148261967249,
      "loss": 5.3325,
      "mean_token_accuracy": 0.17644450217485427,
      "num_tokens": 13665204.0,
      "step": 6575
    },
    {
      "entropy": 5.571329927444458,
      "epoch": 0.5968795355587808,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004971096588626649,
      "loss": 5.4285,
      "mean_token_accuracy": 0.174003441631794,
      "num_tokens": 13676281.0,
      "step": 6580
    },
    {
      "entropy": 5.577637481689453,
      "epoch": 0.597333091436865,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004971044869353206,
      "loss": 5.3917,
      "mean_token_accuracy": 0.18069213032722473,
      "num_tokens": 13687144.0,
      "step": 6585
    },
    {
      "entropy": 5.600950384140015,
      "epoch": 0.5977866473149492,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497099310414799,
      "loss": 5.3665,
      "mean_token_accuracy": 0.1816061720252037,
      "num_tokens": 13696371.0,
      "step": 6590
    },
    {
      "entropy": 5.622022247314453,
      "epoch": 0.5982402031930334,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970941293012069,
      "loss": 5.3139,
      "mean_token_accuracy": 0.17685348838567733,
      "num_tokens": 13706900.0,
      "step": 6595
    },
    {
      "entropy": 5.555879068374634,
      "epoch": 0.5986937590711175,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004970889435946518,
      "loss": 5.3977,
      "mean_token_accuracy": 0.18049706071615218,
      "num_tokens": 13717874.0,
      "step": 6600
    },
    {
      "entropy": 5.588134765625,
      "epoch": 0.5991473149492017,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970837532952406,
      "loss": 5.3653,
      "mean_token_accuracy": 0.1767413467168808,
      "num_tokens": 13728020.0,
      "step": 6605
    },
    {
      "entropy": 5.60174593925476,
      "epoch": 0.599600870827286,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970785584030808,
      "loss": 5.3963,
      "mean_token_accuracy": 0.1784649208188057,
      "num_tokens": 13738007.0,
      "step": 6610
    },
    {
      "entropy": 5.601901006698609,
      "epoch": 0.6000544267053701,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970733589182799,
      "loss": 5.4084,
      "mean_token_accuracy": 0.1740790531039238,
      "num_tokens": 13749429.0,
      "step": 6615
    },
    {
      "entropy": 5.5603643417358395,
      "epoch": 0.6005079825834543,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004970681548409452,
      "loss": 5.2764,
      "mean_token_accuracy": 0.18648878186941148,
      "num_tokens": 13760279.0,
      "step": 6620
    },
    {
      "entropy": 5.591698074340821,
      "epoch": 0.6009615384615384,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970629461711846,
      "loss": 5.4861,
      "mean_token_accuracy": 0.1701574757695198,
      "num_tokens": 13770575.0,
      "step": 6625
    },
    {
      "entropy": 5.536564922332763,
      "epoch": 0.6014150943396226,
      "grad_norm": 0.875,
      "learning_rate": 0.0004970577329091057,
      "loss": 5.326,
      "mean_token_accuracy": 0.18177662342786788,
      "num_tokens": 13781068.0,
      "step": 6630
    },
    {
      "entropy": 5.604051685333252,
      "epoch": 0.6018686502177069,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970525150548162,
      "loss": 5.4189,
      "mean_token_accuracy": 0.17805569767951965,
      "num_tokens": 13790745.0,
      "step": 6635
    },
    {
      "entropy": 5.590482139587403,
      "epoch": 0.602322206095791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004970472926084242,
      "loss": 5.3506,
      "mean_token_accuracy": 0.17441321313381195,
      "num_tokens": 13800196.0,
      "step": 6640
    },
    {
      "entropy": 5.5683002948760985,
      "epoch": 0.6027757619738752,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970420655700375,
      "loss": 5.2585,
      "mean_token_accuracy": 0.18444105684757234,
      "num_tokens": 13809954.0,
      "step": 6645
    },
    {
      "entropy": 5.542122173309326,
      "epoch": 0.6032293178519593,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970368339397643,
      "loss": 5.436,
      "mean_token_accuracy": 0.17155808210372925,
      "num_tokens": 13820281.0,
      "step": 6650
    },
    {
      "entropy": 5.6209453582763675,
      "epoch": 0.6036828737300436,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004970315977177129,
      "loss": 5.3867,
      "mean_token_accuracy": 0.178961344063282,
      "num_tokens": 13830614.0,
      "step": 6655
    },
    {
      "entropy": 5.566452884674073,
      "epoch": 0.6041364296081277,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970263569039914,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1751037433743477,
      "num_tokens": 13840551.0,
      "step": 6660
    },
    {
      "entropy": 5.64999303817749,
      "epoch": 0.6045899854862119,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004970211114987083,
      "loss": 5.5084,
      "mean_token_accuracy": 0.17182745039463043,
      "num_tokens": 13851830.0,
      "step": 6665
    },
    {
      "entropy": 5.598976469039917,
      "epoch": 0.6050435413642961,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004970158615019719,
      "loss": 5.3166,
      "mean_token_accuracy": 0.17749295830726625,
      "num_tokens": 13862299.0,
      "step": 6670
    },
    {
      "entropy": 5.650000429153442,
      "epoch": 0.6054970972423802,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970106069138912,
      "loss": 5.4553,
      "mean_token_accuracy": 0.16745540499687195,
      "num_tokens": 13872852.0,
      "step": 6675
    },
    {
      "entropy": 5.541075944900513,
      "epoch": 0.6059506531204645,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004970053477345743,
      "loss": 5.3897,
      "mean_token_accuracy": 0.18472966998815538,
      "num_tokens": 13883601.0,
      "step": 6680
    },
    {
      "entropy": 5.573489952087402,
      "epoch": 0.6064042089985486,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970000839641304,
      "loss": 5.3771,
      "mean_token_accuracy": 0.1835797294974327,
      "num_tokens": 13894246.0,
      "step": 6685
    },
    {
      "entropy": 5.5467630386352536,
      "epoch": 0.6068577648766328,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969948156026681,
      "loss": 5.4114,
      "mean_token_accuracy": 0.1743486538529396,
      "num_tokens": 13904823.0,
      "step": 6690
    },
    {
      "entropy": 5.654760503768921,
      "epoch": 0.6073113207547169,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969895426502965,
      "loss": 5.4214,
      "mean_token_accuracy": 0.17350844889879227,
      "num_tokens": 13915338.0,
      "step": 6695
    },
    {
      "entropy": 5.578554439544678,
      "epoch": 0.6077648766328012,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969842651071246,
      "loss": 5.3676,
      "mean_token_accuracy": 0.18019723445177077,
      "num_tokens": 13925838.0,
      "step": 6700
    },
    {
      "entropy": 5.63786473274231,
      "epoch": 0.6082184325108854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969789829732615,
      "loss": 5.3876,
      "mean_token_accuracy": 0.1755361571907997,
      "num_tokens": 13936341.0,
      "step": 6705
    },
    {
      "entropy": 5.589788913726807,
      "epoch": 0.6086719883889695,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969736962488164,
      "loss": 5.3615,
      "mean_token_accuracy": 0.17749459445476531,
      "num_tokens": 13945438.0,
      "step": 6710
    },
    {
      "entropy": 5.552593803405761,
      "epoch": 0.6091255442670537,
      "grad_norm": 0.875,
      "learning_rate": 0.0004969684049338987,
      "loss": 5.27,
      "mean_token_accuracy": 0.1792834594845772,
      "num_tokens": 13956029.0,
      "step": 6715
    },
    {
      "entropy": 5.6268645286560055,
      "epoch": 0.6095791001451378,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496963109028618,
      "loss": 5.3858,
      "mean_token_accuracy": 0.1727551519870758,
      "num_tokens": 13966498.0,
      "step": 6720
    },
    {
      "entropy": 5.575082731246948,
      "epoch": 0.6100326560232221,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969578085330834,
      "loss": 5.3385,
      "mean_token_accuracy": 0.17911534309387206,
      "num_tokens": 13976997.0,
      "step": 6725
    },
    {
      "entropy": 5.628406524658203,
      "epoch": 0.6104862119013063,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969525034474049,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1798671528697014,
      "num_tokens": 13986664.0,
      "step": 6730
    },
    {
      "entropy": 5.485960388183594,
      "epoch": 0.6109397677793904,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496947193771692,
      "loss": 5.3458,
      "mean_token_accuracy": 0.18172517567873,
      "num_tokens": 13997828.0,
      "step": 6735
    },
    {
      "entropy": 5.5296248435974125,
      "epoch": 0.6113933236574746,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969418795060548,
      "loss": 5.2938,
      "mean_token_accuracy": 0.17550018876791001,
      "num_tokens": 14008293.0,
      "step": 6740
    },
    {
      "entropy": 5.513893079757691,
      "epoch": 0.6118468795355588,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969365606506027,
      "loss": 5.2821,
      "mean_token_accuracy": 0.1838718071579933,
      "num_tokens": 14018031.0,
      "step": 6745
    },
    {
      "entropy": 5.516136503219604,
      "epoch": 0.612300435413643,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969312372054461,
      "loss": 5.2563,
      "mean_token_accuracy": 0.18612198829650878,
      "num_tokens": 14028013.0,
      "step": 6750
    },
    {
      "entropy": 5.543424034118653,
      "epoch": 0.6127539912917271,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004969259091706948,
      "loss": 5.3988,
      "mean_token_accuracy": 0.17460741549730302,
      "num_tokens": 14039787.0,
      "step": 6755
    },
    {
      "entropy": 5.577391767501831,
      "epoch": 0.6132075471698113,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004969205765464593,
      "loss": 5.3011,
      "mean_token_accuracy": 0.17705341130495073,
      "num_tokens": 14049223.0,
      "step": 6760
    },
    {
      "entropy": 5.564587926864624,
      "epoch": 0.6136611030478955,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969152393328497,
      "loss": 5.3386,
      "mean_token_accuracy": 0.1833565875887871,
      "num_tokens": 14059004.0,
      "step": 6765
    },
    {
      "entropy": 5.589889240264893,
      "epoch": 0.6141146589259797,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969098975299764,
      "loss": 5.433,
      "mean_token_accuracy": 0.17130700796842574,
      "num_tokens": 14069615.0,
      "step": 6770
    },
    {
      "entropy": 5.584448528289795,
      "epoch": 0.6145682148040639,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969045511379499,
      "loss": 5.3674,
      "mean_token_accuracy": 0.17663028091192245,
      "num_tokens": 14078886.0,
      "step": 6775
    },
    {
      "entropy": 5.650965642929077,
      "epoch": 0.615021770682148,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968992001568806,
      "loss": 5.4354,
      "mean_token_accuracy": 0.17254979014396668,
      "num_tokens": 14089021.0,
      "step": 6780
    },
    {
      "entropy": 5.535510587692261,
      "epoch": 0.6154753265602322,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968938445868795,
      "loss": 5.2588,
      "mean_token_accuracy": 0.18511583507061005,
      "num_tokens": 14098316.0,
      "step": 6785
    },
    {
      "entropy": 5.4947816848754885,
      "epoch": 0.6159288824383164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968884844280569,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1732075110077858,
      "num_tokens": 14108874.0,
      "step": 6790
    },
    {
      "entropy": 5.591892051696777,
      "epoch": 0.6163824383164006,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968831196805242,
      "loss": 5.3912,
      "mean_token_accuracy": 0.17782093733549117,
      "num_tokens": 14119613.0,
      "step": 6795
    },
    {
      "entropy": 5.60273003578186,
      "epoch": 0.6168359941944848,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004968777503443918,
      "loss": 5.3499,
      "mean_token_accuracy": 0.17755615413188935,
      "num_tokens": 14129085.0,
      "step": 6800
    },
    {
      "entropy": 5.454257392883301,
      "epoch": 0.6172895500725689,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004968723764197711,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1916737213730812,
      "num_tokens": 14138461.0,
      "step": 6805
    },
    {
      "entropy": 5.528428459167481,
      "epoch": 0.6177431059506531,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496866997906773,
      "loss": 5.3161,
      "mean_token_accuracy": 0.18484087139368058,
      "num_tokens": 14149126.0,
      "step": 6810
    },
    {
      "entropy": 5.5417094230651855,
      "epoch": 0.6181966618287373,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968616148055089,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18229269832372666,
      "num_tokens": 14160068.0,
      "step": 6815
    },
    {
      "entropy": 5.5963634014129635,
      "epoch": 0.6186502177068215,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968562271160901,
      "loss": 5.4037,
      "mean_token_accuracy": 0.1814342111349106,
      "num_tokens": 14170288.0,
      "step": 6820
    },
    {
      "entropy": 5.6165848731994625,
      "epoch": 0.6191037735849056,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496850834838628,
      "loss": 5.4088,
      "mean_token_accuracy": 0.1776348978281021,
      "num_tokens": 14181231.0,
      "step": 6825
    },
    {
      "entropy": 5.5837572574615475,
      "epoch": 0.6195573294629898,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004968454379732342,
      "loss": 5.3227,
      "mean_token_accuracy": 0.18313741385936738,
      "num_tokens": 14191715.0,
      "step": 6830
    },
    {
      "entropy": 5.502680969238281,
      "epoch": 0.6200108853410741,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049684003652002,
      "loss": 5.31,
      "mean_token_accuracy": 0.17466335743665695,
      "num_tokens": 14203416.0,
      "step": 6835
    },
    {
      "entropy": 5.496374702453613,
      "epoch": 0.6204644412191582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968346304790973,
      "loss": 5.2663,
      "mean_token_accuracy": 0.18976710140705108,
      "num_tokens": 14213658.0,
      "step": 6840
    },
    {
      "entropy": 5.611422061920166,
      "epoch": 0.6209179970972424,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968292198505781,
      "loss": 5.449,
      "mean_token_accuracy": 0.17389458268880845,
      "num_tokens": 14223956.0,
      "step": 6845
    },
    {
      "entropy": 5.6339301586151125,
      "epoch": 0.6213715529753265,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000496823804634574,
      "loss": 5.4206,
      "mean_token_accuracy": 0.17541918456554412,
      "num_tokens": 14236801.0,
      "step": 6850
    },
    {
      "entropy": 5.559810400009155,
      "epoch": 0.6218251088534107,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004968183848311972,
      "loss": 5.241,
      "mean_token_accuracy": 0.1860935151576996,
      "num_tokens": 14247382.0,
      "step": 6855
    },
    {
      "entropy": 5.591649961471558,
      "epoch": 0.622278664731495,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968129604405596,
      "loss": 5.3308,
      "mean_token_accuracy": 0.18287506848573684,
      "num_tokens": 14258627.0,
      "step": 6860
    },
    {
      "entropy": 5.489435148239136,
      "epoch": 0.6227322206095791,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968075314627734,
      "loss": 5.3319,
      "mean_token_accuracy": 0.18305183500051497,
      "num_tokens": 14269013.0,
      "step": 6865
    },
    {
      "entropy": 5.5313599586486815,
      "epoch": 0.6231857764876633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496802097897951,
      "loss": 5.3212,
      "mean_token_accuracy": 0.18085480630397796,
      "num_tokens": 14279620.0,
      "step": 6870
    },
    {
      "entropy": 5.593470907211303,
      "epoch": 0.6236393323657474,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004967966597462046,
      "loss": 5.3233,
      "mean_token_accuracy": 0.18621707558631898,
      "num_tokens": 14289448.0,
      "step": 6875
    },
    {
      "entropy": 5.573109865188599,
      "epoch": 0.6240928882438317,
      "grad_norm": 0.875,
      "learning_rate": 0.0004967912170076469,
      "loss": 5.3223,
      "mean_token_accuracy": 0.17920063734054564,
      "num_tokens": 14299999.0,
      "step": 6880
    },
    {
      "entropy": 5.4895679473876955,
      "epoch": 0.6245464441219158,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967857696823901,
      "loss": 5.2935,
      "mean_token_accuracy": 0.18258029073476792,
      "num_tokens": 14310384.0,
      "step": 6885
    },
    {
      "entropy": 5.6066499710083,
      "epoch": 0.625,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967803177705472,
      "loss": 5.421,
      "mean_token_accuracy": 0.17535581439733505,
      "num_tokens": 14320644.0,
      "step": 6890
    },
    {
      "entropy": 5.613523244857788,
      "epoch": 0.6254535558780842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004967748612722308,
      "loss": 5.3771,
      "mean_token_accuracy": 0.1768083542585373,
      "num_tokens": 14330077.0,
      "step": 6895
    },
    {
      "entropy": 5.6285642147064205,
      "epoch": 0.6259071117561683,
      "grad_norm": 0.875,
      "learning_rate": 0.0004967694001875538,
      "loss": 5.4657,
      "mean_token_accuracy": 0.17162772566080092,
      "num_tokens": 14341549.0,
      "step": 6900
    },
    {
      "entropy": 5.5867125511169435,
      "epoch": 0.6263606676342526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496763934516629,
      "loss": 5.3252,
      "mean_token_accuracy": 0.18651908189058303,
      "num_tokens": 14352146.0,
      "step": 6905
    },
    {
      "entropy": 5.49106731414795,
      "epoch": 0.6268142235123367,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967584642595696,
      "loss": 5.3502,
      "mean_token_accuracy": 0.1742198124527931,
      "num_tokens": 14362521.0,
      "step": 6910
    },
    {
      "entropy": 5.613452339172364,
      "epoch": 0.6272677793904209,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967529894164887,
      "loss": 5.3842,
      "mean_token_accuracy": 0.1770589143037796,
      "num_tokens": 14373216.0,
      "step": 6915
    },
    {
      "entropy": 5.554586124420166,
      "epoch": 0.627721335268505,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967475099874992,
      "loss": 5.2975,
      "mean_token_accuracy": 0.18457451462745667,
      "num_tokens": 14383104.0,
      "step": 6920
    },
    {
      "entropy": 5.516722202301025,
      "epoch": 0.6281748911465893,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496742025972715,
      "loss": 5.3284,
      "mean_token_accuracy": 0.19065247923135759,
      "num_tokens": 14393854.0,
      "step": 6925
    },
    {
      "entropy": 5.571130990982056,
      "epoch": 0.6286284470246735,
      "grad_norm": 1.0625,
      "learning_rate": 0.000496736537372249,
      "loss": 5.2746,
      "mean_token_accuracy": 0.18790257424116136,
      "num_tokens": 14403213.0,
      "step": 6930
    },
    {
      "entropy": 5.6476743698120115,
      "epoch": 0.6290820029027576,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967310441862151,
      "loss": 5.4889,
      "mean_token_accuracy": 0.17305789291858673,
      "num_tokens": 14412792.0,
      "step": 6935
    },
    {
      "entropy": 5.5903888702392575,
      "epoch": 0.6295355587808418,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004967255464147265,
      "loss": 5.3434,
      "mean_token_accuracy": 0.18606058061122893,
      "num_tokens": 14423061.0,
      "step": 6940
    },
    {
      "entropy": 5.547424602508545,
      "epoch": 0.6299891146589259,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004967200440578972,
      "loss": 5.2801,
      "mean_token_accuracy": 0.1876331478357315,
      "num_tokens": 14433338.0,
      "step": 6945
    },
    {
      "entropy": 5.512516736984253,
      "epoch": 0.6304426705370102,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967145371158408,
      "loss": 5.282,
      "mean_token_accuracy": 0.1900091901421547,
      "num_tokens": 14444124.0,
      "step": 6950
    },
    {
      "entropy": 5.559875965118408,
      "epoch": 0.6308962264150944,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004967090255886714,
      "loss": 5.4059,
      "mean_token_accuracy": 0.17928589284420013,
      "num_tokens": 14453868.0,
      "step": 6955
    },
    {
      "entropy": 5.547641086578369,
      "epoch": 0.6313497822931785,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967035094765028,
      "loss": 5.3001,
      "mean_token_accuracy": 0.17467721551656723,
      "num_tokens": 14465015.0,
      "step": 6960
    },
    {
      "entropy": 5.53034987449646,
      "epoch": 0.6318033381712627,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966979887794491,
      "loss": 5.2563,
      "mean_token_accuracy": 0.18848009556531906,
      "num_tokens": 14475523.0,
      "step": 6965
    },
    {
      "entropy": 5.5402673244476315,
      "epoch": 0.6322568940493469,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966924634976245,
      "loss": 5.3295,
      "mean_token_accuracy": 0.18160206973552703,
      "num_tokens": 14485523.0,
      "step": 6970
    },
    {
      "entropy": 5.61271300315857,
      "epoch": 0.6327104499274311,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966869336311433,
      "loss": 5.4094,
      "mean_token_accuracy": 0.17774551808834077,
      "num_tokens": 14495237.0,
      "step": 6975
    },
    {
      "entropy": 5.5918501853942875,
      "epoch": 0.6331640058055152,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966813991801197,
      "loss": 5.5141,
      "mean_token_accuracy": 0.17085262537002563,
      "num_tokens": 14506431.0,
      "step": 6980
    },
    {
      "entropy": 5.62278151512146,
      "epoch": 0.6336175616835994,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966758601446684,
      "loss": 5.4451,
      "mean_token_accuracy": 0.17433076500892639,
      "num_tokens": 14517445.0,
      "step": 6985
    },
    {
      "entropy": 5.636280870437622,
      "epoch": 0.6340711175616836,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966703165249038,
      "loss": 5.3707,
      "mean_token_accuracy": 0.1714874342083931,
      "num_tokens": 14527818.0,
      "step": 6990
    },
    {
      "entropy": 5.606985330581665,
      "epoch": 0.6345246734397678,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966647683209406,
      "loss": 5.4616,
      "mean_token_accuracy": 0.16598857343196868,
      "num_tokens": 14538235.0,
      "step": 6995
    },
    {
      "entropy": 5.69085431098938,
      "epoch": 0.634978229317852,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966592155328935,
      "loss": 5.4577,
      "mean_token_accuracy": 0.17015783339738846,
      "num_tokens": 14549784.0,
      "step": 7000
    },
    {
      "entropy": 5.577909851074219,
      "epoch": 0.6354317851959361,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004966536581608771,
      "loss": 5.4601,
      "mean_token_accuracy": 0.16502557843923568,
      "num_tokens": 14561057.0,
      "step": 7005
    },
    {
      "entropy": 5.6043314933776855,
      "epoch": 0.6358853410740203,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004966480962050069,
      "loss": 5.3663,
      "mean_token_accuracy": 0.17885040044784545,
      "num_tokens": 14570859.0,
      "step": 7010
    },
    {
      "entropy": 5.596745586395263,
      "epoch": 0.6363388969521045,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966425296653974,
      "loss": 5.3727,
      "mean_token_accuracy": 0.18546929955482483,
      "num_tokens": 14581066.0,
      "step": 7015
    },
    {
      "entropy": 5.507768726348877,
      "epoch": 0.6367924528301887,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496636958542164,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17755770236253737,
      "num_tokens": 14590709.0,
      "step": 7020
    },
    {
      "entropy": 5.553393363952637,
      "epoch": 0.6372460087082729,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004966313828354218,
      "loss": 5.3268,
      "mean_token_accuracy": 0.184244504570961,
      "num_tokens": 14600849.0,
      "step": 7025
    },
    {
      "entropy": 5.520700263977051,
      "epoch": 0.637699564586357,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496625802545286,
      "loss": 5.2116,
      "mean_token_accuracy": 0.1859947547316551,
      "num_tokens": 14611427.0,
      "step": 7030
    },
    {
      "entropy": 5.530115747451783,
      "epoch": 0.6381531204644412,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004966202176718723,
      "loss": 5.3635,
      "mean_token_accuracy": 0.182247893512249,
      "num_tokens": 14623860.0,
      "step": 7035
    },
    {
      "entropy": 5.564819049835205,
      "epoch": 0.6386066763425254,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966146282152959,
      "loss": 5.2998,
      "mean_token_accuracy": 0.1783895254135132,
      "num_tokens": 14633481.0,
      "step": 7040
    },
    {
      "entropy": 5.564633512496949,
      "epoch": 0.6390602322206096,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966090341756726,
      "loss": 5.2811,
      "mean_token_accuracy": 0.18869292736053467,
      "num_tokens": 14643519.0,
      "step": 7045
    },
    {
      "entropy": 5.450642681121826,
      "epoch": 0.6395137880986937,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496603435553118,
      "loss": 5.2722,
      "mean_token_accuracy": 0.18621875643730162,
      "num_tokens": 14654570.0,
      "step": 7050
    },
    {
      "entropy": 5.501469135284424,
      "epoch": 0.6399673439767779,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965978323477478,
      "loss": 5.3098,
      "mean_token_accuracy": 0.17667924016714096,
      "num_tokens": 14665596.0,
      "step": 7055
    },
    {
      "entropy": 5.580057382583618,
      "epoch": 0.6404208998548622,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496592224559678,
      "loss": 5.2669,
      "mean_token_accuracy": 0.18460346013307571,
      "num_tokens": 14675660.0,
      "step": 7060
    },
    {
      "entropy": 5.580855798721314,
      "epoch": 0.6408744557329463,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004965866121890245,
      "loss": 5.4209,
      "mean_token_accuracy": 0.17853408306837082,
      "num_tokens": 14685229.0,
      "step": 7065
    },
    {
      "entropy": 5.619711685180664,
      "epoch": 0.6413280116110305,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965809952359034,
      "loss": 5.3312,
      "mean_token_accuracy": 0.17109032422304155,
      "num_tokens": 14695672.0,
      "step": 7070
    },
    {
      "entropy": 5.563270711898804,
      "epoch": 0.6417815674891146,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965753737004309,
      "loss": 5.2857,
      "mean_token_accuracy": 0.18523964136838914,
      "num_tokens": 14705160.0,
      "step": 7075
    },
    {
      "entropy": 5.51725435256958,
      "epoch": 0.6422351233671988,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965697475827231,
      "loss": 5.3204,
      "mean_token_accuracy": 0.1778504505753517,
      "num_tokens": 14715487.0,
      "step": 7080
    },
    {
      "entropy": 5.580355167388916,
      "epoch": 0.6426886792452831,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965641168828965,
      "loss": 5.3743,
      "mean_token_accuracy": 0.18160596191883088,
      "num_tokens": 14726987.0,
      "step": 7085
    },
    {
      "entropy": 5.494480085372925,
      "epoch": 0.6431422351233672,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965584816010676,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1852257654070854,
      "num_tokens": 14737810.0,
      "step": 7090
    },
    {
      "entropy": 5.610119819641113,
      "epoch": 0.6435957910014514,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965528417373528,
      "loss": 5.3285,
      "mean_token_accuracy": 0.17879804968833923,
      "num_tokens": 14748997.0,
      "step": 7095
    },
    {
      "entropy": 5.566685581207276,
      "epoch": 0.6440493468795355,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004965471972918688,
      "loss": 5.3404,
      "mean_token_accuracy": 0.1756565198302269,
      "num_tokens": 14760439.0,
      "step": 7100
    },
    {
      "entropy": 5.460477256774903,
      "epoch": 0.6445029027576198,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004965415482647322,
      "loss": 5.2521,
      "mean_token_accuracy": 0.18281868398189544,
      "num_tokens": 14770828.0,
      "step": 7105
    },
    {
      "entropy": 5.489846992492676,
      "epoch": 0.6449564586357039,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049653589465606,
      "loss": 5.3957,
      "mean_token_accuracy": 0.18351997137069703,
      "num_tokens": 14781753.0,
      "step": 7110
    },
    {
      "entropy": 5.6520702838897705,
      "epoch": 0.6454100145137881,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496530236465969,
      "loss": 5.4102,
      "mean_token_accuracy": 0.17198631465435027,
      "num_tokens": 14792241.0,
      "step": 7115
    },
    {
      "entropy": 5.531635856628418,
      "epoch": 0.6458635703918723,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965245736945763,
      "loss": 5.2285,
      "mean_token_accuracy": 0.1889130190014839,
      "num_tokens": 14802580.0,
      "step": 7120
    },
    {
      "entropy": 5.519139194488526,
      "epoch": 0.6463171262699564,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004965189063419989,
      "loss": 5.442,
      "mean_token_accuracy": 0.17884785532951356,
      "num_tokens": 14813103.0,
      "step": 7125
    },
    {
      "entropy": 5.568569326400757,
      "epoch": 0.6467706821480407,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965132344083541,
      "loss": 5.3276,
      "mean_token_accuracy": 0.17959365546703338,
      "num_tokens": 14823334.0,
      "step": 7130
    },
    {
      "entropy": 5.568736124038696,
      "epoch": 0.6472242380261248,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965075578937593,
      "loss": 5.3492,
      "mean_token_accuracy": 0.17829880565404893,
      "num_tokens": 14834850.0,
      "step": 7135
    },
    {
      "entropy": 5.513080644607544,
      "epoch": 0.647677793904209,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965018767983316,
      "loss": 5.3575,
      "mean_token_accuracy": 0.1796330839395523,
      "num_tokens": 14844276.0,
      "step": 7140
    },
    {
      "entropy": 5.4893872261047365,
      "epoch": 0.6481313497822931,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964961911221886,
      "loss": 5.1964,
      "mean_token_accuracy": 0.1915486693382263,
      "num_tokens": 14854352.0,
      "step": 7145
    },
    {
      "entropy": 5.584615421295166,
      "epoch": 0.6485849056603774,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496490500865448,
      "loss": 5.378,
      "mean_token_accuracy": 0.1796365812420845,
      "num_tokens": 14864395.0,
      "step": 7150
    },
    {
      "entropy": 5.5620684146881105,
      "epoch": 0.6490384615384616,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964848060282275,
      "loss": 5.2599,
      "mean_token_accuracy": 0.18747452944517135,
      "num_tokens": 14874373.0,
      "step": 7155
    },
    {
      "entropy": 5.53101077079773,
      "epoch": 0.6494920174165457,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964791066106447,
      "loss": 5.2642,
      "mean_token_accuracy": 0.18159855753183365,
      "num_tokens": 14884932.0,
      "step": 7160
    },
    {
      "entropy": 5.5659397602081295,
      "epoch": 0.6499455732946299,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964734026128175,
      "loss": 5.4253,
      "mean_token_accuracy": 0.17503994554281235,
      "num_tokens": 14896254.0,
      "step": 7165
    },
    {
      "entropy": 5.623956346511841,
      "epoch": 0.650399129172714,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964676940348638,
      "loss": 5.3453,
      "mean_token_accuracy": 0.17856290340423583,
      "num_tokens": 14907392.0,
      "step": 7170
    },
    {
      "entropy": 5.5940752029418945,
      "epoch": 0.6508526850507983,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496461980876902,
      "loss": 5.4614,
      "mean_token_accuracy": 0.17526094764471054,
      "num_tokens": 14917676.0,
      "step": 7175
    },
    {
      "entropy": 5.5578306198120115,
      "epoch": 0.6513062409288825,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004964562631390498,
      "loss": 5.3008,
      "mean_token_accuracy": 0.18212647140026092,
      "num_tokens": 14928500.0,
      "step": 7180
    },
    {
      "entropy": 5.551118040084839,
      "epoch": 0.6517597968069666,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964505408214257,
      "loss": 5.2703,
      "mean_token_accuracy": 0.18634399622678757,
      "num_tokens": 14938970.0,
      "step": 7185
    },
    {
      "entropy": 5.526061344146728,
      "epoch": 0.6522133526850508,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496444813924148,
      "loss": 5.3255,
      "mean_token_accuracy": 0.18173780143260956,
      "num_tokens": 14949017.0,
      "step": 7190
    },
    {
      "entropy": 5.52117953300476,
      "epoch": 0.652666908563135,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964390824473351,
      "loss": 5.2443,
      "mean_token_accuracy": 0.19174836874008178,
      "num_tokens": 14958529.0,
      "step": 7195
    },
    {
      "entropy": 5.527310991287232,
      "epoch": 0.6531204644412192,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004964333463911056,
      "loss": 5.3604,
      "mean_token_accuracy": 0.17524895071983337,
      "num_tokens": 14968044.0,
      "step": 7200
    },
    {
      "entropy": 5.596441411972046,
      "epoch": 0.6535740203193033,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004964276057555779,
      "loss": 5.3771,
      "mean_token_accuracy": 0.17824496030807496,
      "num_tokens": 14977867.0,
      "step": 7205
    },
    {
      "entropy": 5.6655659675598145,
      "epoch": 0.6540275761973875,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496421860540871,
      "loss": 5.3958,
      "mean_token_accuracy": 0.1837482750415802,
      "num_tokens": 14986936.0,
      "step": 7210
    },
    {
      "entropy": 5.506159448623658,
      "epoch": 0.6544811320754716,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964161107471036,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18265933394432068,
      "num_tokens": 14996681.0,
      "step": 7215
    },
    {
      "entropy": 5.5519446849823,
      "epoch": 0.6549346879535559,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964103563743946,
      "loss": 5.3364,
      "mean_token_accuracy": 0.18156073987483978,
      "num_tokens": 15007100.0,
      "step": 7220
    },
    {
      "entropy": 5.5976849555969235,
      "epoch": 0.6553882438316401,
      "grad_norm": 0.984375,
      "learning_rate": 0.000496404597422863,
      "loss": 5.4102,
      "mean_token_accuracy": 0.18126775324344635,
      "num_tokens": 15017452.0,
      "step": 7225
    },
    {
      "entropy": 5.539013767242432,
      "epoch": 0.6558417997097242,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496398833892628,
      "loss": 5.254,
      "mean_token_accuracy": 0.18298355340957642,
      "num_tokens": 15028362.0,
      "step": 7230
    },
    {
      "entropy": 5.589620161056518,
      "epoch": 0.6562953555878084,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963930657838086,
      "loss": 5.3474,
      "mean_token_accuracy": 0.17923036962747574,
      "num_tokens": 15038691.0,
      "step": 7235
    },
    {
      "entropy": 5.547629928588867,
      "epoch": 0.6567489114658926,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963872930965242,
      "loss": 5.3283,
      "mean_token_accuracy": 0.1816704109311104,
      "num_tokens": 15050037.0,
      "step": 7240
    },
    {
      "entropy": 5.559830236434936,
      "epoch": 0.6572024673439768,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004963815158308941,
      "loss": 5.3057,
      "mean_token_accuracy": 0.18500083684921265,
      "num_tokens": 15059793.0,
      "step": 7245
    },
    {
      "entropy": 5.537407398223877,
      "epoch": 0.657656023222061,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963757339870379,
      "loss": 5.3896,
      "mean_token_accuracy": 0.17124081701040267,
      "num_tokens": 15070500.0,
      "step": 7250
    },
    {
      "entropy": 5.612113428115845,
      "epoch": 0.6581095791001451,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963699475650751,
      "loss": 5.4467,
      "mean_token_accuracy": 0.17665672302246094,
      "num_tokens": 15081109.0,
      "step": 7255
    },
    {
      "entropy": 5.5193017482757565,
      "epoch": 0.6585631349782293,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963641565651255,
      "loss": 5.323,
      "mean_token_accuracy": 0.18454677164554595,
      "num_tokens": 15092315.0,
      "step": 7260
    },
    {
      "entropy": 5.568742418289185,
      "epoch": 0.6590166908563135,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963583609873084,
      "loss": 5.3849,
      "mean_token_accuracy": 0.1800144374370575,
      "num_tokens": 15103957.0,
      "step": 7265
    },
    {
      "entropy": 5.6056946277618405,
      "epoch": 0.6594702467343977,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004963525608317443,
      "loss": 5.3721,
      "mean_token_accuracy": 0.1770993635058403,
      "num_tokens": 15114996.0,
      "step": 7270
    },
    {
      "entropy": 5.574876546859741,
      "epoch": 0.6599238026124818,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004963467560985527,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1811333030462265,
      "num_tokens": 15125098.0,
      "step": 7275
    },
    {
      "entropy": 5.508492565155029,
      "epoch": 0.660377358490566,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004963409467878538,
      "loss": 5.3165,
      "mean_token_accuracy": 0.17881869971752168,
      "num_tokens": 15134746.0,
      "step": 7280
    },
    {
      "entropy": 5.631923818588257,
      "epoch": 0.6608309143686503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963351328997678,
      "loss": 5.3384,
      "mean_token_accuracy": 0.17652247250080108,
      "num_tokens": 15145698.0,
      "step": 7285
    },
    {
      "entropy": 5.570781326293945,
      "epoch": 0.6612844702467344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963293144344148,
      "loss": 5.3904,
      "mean_token_accuracy": 0.17904452234506607,
      "num_tokens": 15156026.0,
      "step": 7290
    },
    {
      "entropy": 5.555085945129394,
      "epoch": 0.6617380261248186,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963234913919151,
      "loss": 5.3322,
      "mean_token_accuracy": 0.17398147583007811,
      "num_tokens": 15165900.0,
      "step": 7295
    },
    {
      "entropy": 5.565875434875489,
      "epoch": 0.6621915820029027,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963176637723893,
      "loss": 5.3794,
      "mean_token_accuracy": 0.17941929996013642,
      "num_tokens": 15176693.0,
      "step": 7300
    },
    {
      "entropy": 5.58468017578125,
      "epoch": 0.6626451378809869,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963118315759578,
      "loss": 5.4007,
      "mean_token_accuracy": 0.16982991248369217,
      "num_tokens": 15186535.0,
      "step": 7305
    },
    {
      "entropy": 5.587566995620728,
      "epoch": 0.6630986937590712,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963059948027411,
      "loss": 5.3162,
      "mean_token_accuracy": 0.1796189233660698,
      "num_tokens": 15196019.0,
      "step": 7310
    },
    {
      "entropy": 5.465688180923462,
      "epoch": 0.6635522496371553,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963001534528601,
      "loss": 5.1812,
      "mean_token_accuracy": 0.19228140413761138,
      "num_tokens": 15205273.0,
      "step": 7315
    },
    {
      "entropy": 5.560149812698365,
      "epoch": 0.6640058055152395,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004962943075264355,
      "loss": 5.3427,
      "mean_token_accuracy": 0.1783326655626297,
      "num_tokens": 15216716.0,
      "step": 7320
    },
    {
      "entropy": 5.571094942092896,
      "epoch": 0.6644593613933236,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962884570235884,
      "loss": 5.3928,
      "mean_token_accuracy": 0.17626622021198274,
      "num_tokens": 15227306.0,
      "step": 7325
    },
    {
      "entropy": 5.510771894454956,
      "epoch": 0.6649129172714079,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004962826019444394,
      "loss": 5.2786,
      "mean_token_accuracy": 0.18220029771327972,
      "num_tokens": 15237302.0,
      "step": 7330
    },
    {
      "entropy": 5.515999460220337,
      "epoch": 0.665366473149492,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004962767422891099,
      "loss": 5.2859,
      "mean_token_accuracy": 0.1832537442445755,
      "num_tokens": 15247064.0,
      "step": 7335
    },
    {
      "entropy": 5.559939670562744,
      "epoch": 0.6658200290275762,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496270878057721,
      "loss": 5.3565,
      "mean_token_accuracy": 0.18223199248313904,
      "num_tokens": 15257006.0,
      "step": 7340
    },
    {
      "entropy": 5.5652186393737795,
      "epoch": 0.6662735849056604,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962650092503939,
      "loss": 5.3333,
      "mean_token_accuracy": 0.1857373908162117,
      "num_tokens": 15266720.0,
      "step": 7345
    },
    {
      "entropy": 5.540846633911133,
      "epoch": 0.6667271407837445,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962591358672502,
      "loss": 5.3651,
      "mean_token_accuracy": 0.17481739670038224,
      "num_tokens": 15277330.0,
      "step": 7350
    },
    {
      "entropy": 5.548562622070312,
      "epoch": 0.6671806966618288,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496253257908411,
      "loss": 5.2984,
      "mean_token_accuracy": 0.1819293111562729,
      "num_tokens": 15287401.0,
      "step": 7355
    },
    {
      "entropy": 5.485146522521973,
      "epoch": 0.6676342525399129,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004962473753739981,
      "loss": 5.2411,
      "mean_token_accuracy": 0.18803683817386627,
      "num_tokens": 15297426.0,
      "step": 7360
    },
    {
      "entropy": 5.565172481536865,
      "epoch": 0.6680878084179971,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496241488264133,
      "loss": 5.2356,
      "mean_token_accuracy": 0.1826209083199501,
      "num_tokens": 15308412.0,
      "step": 7365
    },
    {
      "entropy": 5.527874851226807,
      "epoch": 0.6685413642960812,
      "grad_norm": 1.0,
      "learning_rate": 0.0004962355965789377,
      "loss": 5.2316,
      "mean_token_accuracy": 0.17745898216962813,
      "num_tokens": 15318821.0,
      "step": 7370
    },
    {
      "entropy": 5.4749537944793705,
      "epoch": 0.6689949201741655,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962297003185338,
      "loss": 5.3039,
      "mean_token_accuracy": 0.18267119228839873,
      "num_tokens": 15328737.0,
      "step": 7375
    },
    {
      "entropy": 5.598972654342651,
      "epoch": 0.6694484760522497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962237994830433,
      "loss": 5.411,
      "mean_token_accuracy": 0.17150809168815612,
      "num_tokens": 15339695.0,
      "step": 7380
    },
    {
      "entropy": 5.633320331573486,
      "epoch": 0.6699020319303338,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004962178940725883,
      "loss": 5.3667,
      "mean_token_accuracy": 0.17928486466407775,
      "num_tokens": 15350270.0,
      "step": 7385
    },
    {
      "entropy": 5.533000087738037,
      "epoch": 0.670355587808418,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004962119840872908,
      "loss": 5.367,
      "mean_token_accuracy": 0.18240504264831542,
      "num_tokens": 15360986.0,
      "step": 7390
    },
    {
      "entropy": 5.531470489501953,
      "epoch": 0.6708091436865021,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962060695272733,
      "loss": 5.3501,
      "mean_token_accuracy": 0.18166522681713104,
      "num_tokens": 15371786.0,
      "step": 7395
    },
    {
      "entropy": 5.559743118286133,
      "epoch": 0.6712626995645864,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962001503926577,
      "loss": 5.309,
      "mean_token_accuracy": 0.18303101807832717,
      "num_tokens": 15382960.0,
      "step": 7400
    },
    {
      "entropy": 5.551979637145996,
      "epoch": 0.6717162554426706,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004961942266835667,
      "loss": 5.3077,
      "mean_token_accuracy": 0.18894487917423247,
      "num_tokens": 15392807.0,
      "step": 7405
    },
    {
      "entropy": 5.554458045959473,
      "epoch": 0.6721698113207547,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004961882984001227,
      "loss": 5.3527,
      "mean_token_accuracy": 0.1719239681959152,
      "num_tokens": 15403712.0,
      "step": 7410
    },
    {
      "entropy": 5.548364591598511,
      "epoch": 0.6726233671988389,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961823655424484,
      "loss": 5.4027,
      "mean_token_accuracy": 0.1698218837380409,
      "num_tokens": 15414839.0,
      "step": 7415
    },
    {
      "entropy": 5.563115072250366,
      "epoch": 0.6730769230769231,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961764281106664,
      "loss": 5.293,
      "mean_token_accuracy": 0.18119843751192094,
      "num_tokens": 15425320.0,
      "step": 7420
    },
    {
      "entropy": 5.544220876693726,
      "epoch": 0.6735304789550073,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961704861048995,
      "loss": 5.288,
      "mean_token_accuracy": 0.1794519156217575,
      "num_tokens": 15435476.0,
      "step": 7425
    },
    {
      "entropy": 5.5448249816894535,
      "epoch": 0.6739840348330914,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004961645395252705,
      "loss": 5.4071,
      "mean_token_accuracy": 0.1752161681652069,
      "num_tokens": 15445786.0,
      "step": 7430
    },
    {
      "entropy": 5.524516010284424,
      "epoch": 0.6744375907111756,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961585883719025,
      "loss": 5.3816,
      "mean_token_accuracy": 0.1858324810862541,
      "num_tokens": 15455915.0,
      "step": 7435
    },
    {
      "entropy": 5.600856065750122,
      "epoch": 0.6748911465892597,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004961526326449186,
      "loss": 5.3712,
      "mean_token_accuracy": 0.17852281779050827,
      "num_tokens": 15466523.0,
      "step": 7440
    },
    {
      "entropy": 5.600769948959351,
      "epoch": 0.675344702467344,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004961466723444419,
      "loss": 5.3662,
      "mean_token_accuracy": 0.17874794453382492,
      "num_tokens": 15476821.0,
      "step": 7445
    },
    {
      "entropy": 5.514429330825806,
      "epoch": 0.6757982583454282,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961407074705955,
      "loss": 5.2923,
      "mean_token_accuracy": 0.17971670627593994,
      "num_tokens": 15487182.0,
      "step": 7450
    },
    {
      "entropy": 5.549853372573852,
      "epoch": 0.6762518142235123,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961347380235031,
      "loss": 5.2785,
      "mean_token_accuracy": 0.18686819225549697,
      "num_tokens": 15497600.0,
      "step": 7455
    },
    {
      "entropy": 5.616761255264282,
      "epoch": 0.6767053701015965,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961287640032879,
      "loss": 5.4268,
      "mean_token_accuracy": 0.17058165371418,
      "num_tokens": 15508278.0,
      "step": 7460
    },
    {
      "entropy": 5.64322600364685,
      "epoch": 0.6771589259796807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961227854100734,
      "loss": 5.4467,
      "mean_token_accuracy": 0.17577812671661378,
      "num_tokens": 15518458.0,
      "step": 7465
    },
    {
      "entropy": 5.549437522888184,
      "epoch": 0.6776124818577649,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004961168022439835,
      "loss": 5.2543,
      "mean_token_accuracy": 0.1862138107419014,
      "num_tokens": 15528738.0,
      "step": 7470
    },
    {
      "entropy": 5.498688793182373,
      "epoch": 0.6780660377358491,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961108145051417,
      "loss": 5.3016,
      "mean_token_accuracy": 0.1857891008257866,
      "num_tokens": 15538656.0,
      "step": 7475
    },
    {
      "entropy": 5.4875250339508055,
      "epoch": 0.6785195936139332,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496104822193672,
      "loss": 5.3009,
      "mean_token_accuracy": 0.18368335962295532,
      "num_tokens": 15549285.0,
      "step": 7480
    },
    {
      "entropy": 5.609712409973144,
      "epoch": 0.6789731494920174,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960988253096982,
      "loss": 5.3807,
      "mean_token_accuracy": 0.1867237627506256,
      "num_tokens": 15559766.0,
      "step": 7485
    },
    {
      "entropy": 5.533083581924439,
      "epoch": 0.6794267053701016,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004960928238533443,
      "loss": 5.2478,
      "mean_token_accuracy": 0.18986229747533798,
      "num_tokens": 15570873.0,
      "step": 7490
    },
    {
      "entropy": 5.5704052448272705,
      "epoch": 0.6798802612481858,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960868178247344,
      "loss": 5.4891,
      "mean_token_accuracy": 0.17788162529468537,
      "num_tokens": 15581229.0,
      "step": 7495
    },
    {
      "entropy": 5.540935468673706,
      "epoch": 0.68033381712627,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960808072239929,
      "loss": 5.3557,
      "mean_token_accuracy": 0.17538125514984132,
      "num_tokens": 15591686.0,
      "step": 7500
    },
    {
      "entropy": 5.620772171020508,
      "epoch": 0.6807873730043541,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960747920512439,
      "loss": 5.3755,
      "mean_token_accuracy": 0.17827560752630234,
      "num_tokens": 15601869.0,
      "step": 7505
    },
    {
      "entropy": 5.558009386062622,
      "epoch": 0.6812409288824384,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000496068772306612,
      "loss": 5.3513,
      "mean_token_accuracy": 0.17730835527181626,
      "num_tokens": 15612946.0,
      "step": 7510
    },
    {
      "entropy": 5.558612203598022,
      "epoch": 0.6816944847605225,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960627479902214,
      "loss": 5.3976,
      "mean_token_accuracy": 0.1726248562335968,
      "num_tokens": 15623761.0,
      "step": 7515
    },
    {
      "entropy": 5.559710741043091,
      "epoch": 0.6821480406386067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496056719102197,
      "loss": 5.3926,
      "mean_token_accuracy": 0.17348738610744477,
      "num_tokens": 15634500.0,
      "step": 7520
    },
    {
      "entropy": 5.487872171401977,
      "epoch": 0.6826015965166908,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004960506856426631,
      "loss": 5.2369,
      "mean_token_accuracy": 0.18374142199754714,
      "num_tokens": 15644896.0,
      "step": 7525
    },
    {
      "entropy": 5.5576379776000975,
      "epoch": 0.683055152394775,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960446476117448,
      "loss": 5.3577,
      "mean_token_accuracy": 0.1801358714699745,
      "num_tokens": 15655008.0,
      "step": 7530
    },
    {
      "entropy": 5.5583741664886475,
      "epoch": 0.6835087082728593,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960386050095668,
      "loss": 5.2713,
      "mean_token_accuracy": 0.17511853575706482,
      "num_tokens": 15665944.0,
      "step": 7535
    },
    {
      "entropy": 5.568644237518311,
      "epoch": 0.6839622641509434,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960325578362543,
      "loss": 5.3761,
      "mean_token_accuracy": 0.1809030920267105,
      "num_tokens": 15677099.0,
      "step": 7540
    },
    {
      "entropy": 5.601432275772095,
      "epoch": 0.6844158200290276,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496026506091932,
      "loss": 5.3376,
      "mean_token_accuracy": 0.1795408010482788,
      "num_tokens": 15687531.0,
      "step": 7545
    },
    {
      "entropy": 5.5705641746521,
      "epoch": 0.6848693759071117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960204497767253,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18146915584802628,
      "num_tokens": 15697002.0,
      "step": 7550
    },
    {
      "entropy": 5.470752716064453,
      "epoch": 0.685322931785196,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960143888907594,
      "loss": 5.2251,
      "mean_token_accuracy": 0.1819280982017517,
      "num_tokens": 15707249.0,
      "step": 7555
    },
    {
      "entropy": 5.5658492088317875,
      "epoch": 0.6857764876632801,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004960083234341595,
      "loss": 5.3005,
      "mean_token_accuracy": 0.18364818543195724,
      "num_tokens": 15718764.0,
      "step": 7560
    },
    {
      "entropy": 5.512791872024536,
      "epoch": 0.6862300435413643,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960022534070513,
      "loss": 5.2503,
      "mean_token_accuracy": 0.18830282539129256,
      "num_tokens": 15728854.0,
      "step": 7565
    },
    {
      "entropy": 5.458533000946045,
      "epoch": 0.6866835994194485,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049599617880956,
      "loss": 5.2675,
      "mean_token_accuracy": 0.18406953662633896,
      "num_tokens": 15739682.0,
      "step": 7570
    },
    {
      "entropy": 5.505916738510132,
      "epoch": 0.6871371552975326,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004959900996418115,
      "loss": 5.2657,
      "mean_token_accuracy": 0.1874595806002617,
      "num_tokens": 15749628.0,
      "step": 7575
    },
    {
      "entropy": 5.530863189697266,
      "epoch": 0.6875907111756169,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959840159039315,
      "loss": 5.3281,
      "mean_token_accuracy": 0.1795841246843338,
      "num_tokens": 15760137.0,
      "step": 7580
    },
    {
      "entropy": 5.592649602890015,
      "epoch": 0.688044267053701,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959779275960456,
      "loss": 5.2881,
      "mean_token_accuracy": 0.18079988062381744,
      "num_tokens": 15769889.0,
      "step": 7585
    },
    {
      "entropy": 5.537004804611206,
      "epoch": 0.6884978229317852,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00049597183471828,
      "loss": 5.3876,
      "mean_token_accuracy": 0.17896100878715515,
      "num_tokens": 15780912.0,
      "step": 7590
    },
    {
      "entropy": 5.553856658935547,
      "epoch": 0.6889513788098693,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959657372707604,
      "loss": 5.2688,
      "mean_token_accuracy": 0.18778923004865647,
      "num_tokens": 15791719.0,
      "step": 7595
    },
    {
      "entropy": 5.547277927398682,
      "epoch": 0.6894049346879536,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959596352536132,
      "loss": 5.3197,
      "mean_token_accuracy": 0.18113720566034316,
      "num_tokens": 15802551.0,
      "step": 7600
    },
    {
      "entropy": 5.58339672088623,
      "epoch": 0.6898584905660378,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004959535286669643,
      "loss": 5.3653,
      "mean_token_accuracy": 0.17841881066560744,
      "num_tokens": 15813551.0,
      "step": 7605
    },
    {
      "entropy": 5.554733753204346,
      "epoch": 0.6903120464441219,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004959474175109402,
      "loss": 5.3373,
      "mean_token_accuracy": 0.18408777713775634,
      "num_tokens": 15824262.0,
      "step": 7610
    },
    {
      "entropy": 5.616608047485352,
      "epoch": 0.6907656023222061,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004959413017856672,
      "loss": 5.3833,
      "mean_token_accuracy": 0.18150201439857483,
      "num_tokens": 15835493.0,
      "step": 7615
    },
    {
      "entropy": 5.446601104736328,
      "epoch": 0.6912191582002902,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004959351814912716,
      "loss": 5.2793,
      "mean_token_accuracy": 0.1799856498837471,
      "num_tokens": 15846224.0,
      "step": 7620
    },
    {
      "entropy": 5.5849120140075685,
      "epoch": 0.6916727140783745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959290566278803,
      "loss": 5.3063,
      "mean_token_accuracy": 0.1804905876517296,
      "num_tokens": 15855622.0,
      "step": 7625
    },
    {
      "entropy": 5.562844371795654,
      "epoch": 0.6921262699564587,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004959229271956199,
      "loss": 5.3789,
      "mean_token_accuracy": 0.17874677032232283,
      "num_tokens": 15865835.0,
      "step": 7630
    },
    {
      "entropy": 5.54128589630127,
      "epoch": 0.6925798258345428,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959167931946169,
      "loss": 5.3154,
      "mean_token_accuracy": 0.18270861357450485,
      "num_tokens": 15875915.0,
      "step": 7635
    },
    {
      "entropy": 5.585084104537964,
      "epoch": 0.693033381712627,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004959106546249984,
      "loss": 5.3078,
      "mean_token_accuracy": 0.18133632242679595,
      "num_tokens": 15887111.0,
      "step": 7640
    },
    {
      "entropy": 5.5735945224761965,
      "epoch": 0.6934869375907112,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004959045114868913,
      "loss": 5.4066,
      "mean_token_accuracy": 0.17114051580429077,
      "num_tokens": 15899528.0,
      "step": 7645
    },
    {
      "entropy": 5.567831516265869,
      "epoch": 0.6939404934687954,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004958983637804225,
      "loss": 5.3237,
      "mean_token_accuracy": 0.18822288066148757,
      "num_tokens": 15910881.0,
      "step": 7650
    },
    {
      "entropy": 5.634607219696045,
      "epoch": 0.6943940493468795,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958922115057194,
      "loss": 5.3678,
      "mean_token_accuracy": 0.17664799243211746,
      "num_tokens": 15921189.0,
      "step": 7655
    },
    {
      "entropy": 5.580757188796997,
      "epoch": 0.6948476052249637,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004958860546629089,
      "loss": 5.3637,
      "mean_token_accuracy": 0.18410303443670273,
      "num_tokens": 15931276.0,
      "step": 7660
    },
    {
      "entropy": 5.585839080810547,
      "epoch": 0.6953011611030478,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958798932521186,
      "loss": 5.442,
      "mean_token_accuracy": 0.17055700719356537,
      "num_tokens": 15941506.0,
      "step": 7665
    },
    {
      "entropy": 5.496153450012207,
      "epoch": 0.6957547169811321,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958737272734759,
      "loss": 5.2914,
      "mean_token_accuracy": 0.18633339554071426,
      "num_tokens": 15951158.0,
      "step": 7670
    },
    {
      "entropy": 5.613697624206543,
      "epoch": 0.6962082728592163,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004958675567271081,
      "loss": 5.3059,
      "mean_token_accuracy": 0.18212190866470337,
      "num_tokens": 15960852.0,
      "step": 7675
    },
    {
      "entropy": 5.579945516586304,
      "epoch": 0.6966618287373004,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495861381613143,
      "loss": 5.3139,
      "mean_token_accuracy": 0.1766521543264389,
      "num_tokens": 15971291.0,
      "step": 7680
    },
    {
      "entropy": 5.51633129119873,
      "epoch": 0.6971153846153846,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004958552019317082,
      "loss": 5.377,
      "mean_token_accuracy": 0.17974497824907304,
      "num_tokens": 15982180.0,
      "step": 7685
    },
    {
      "entropy": 5.4834810256958,
      "epoch": 0.6975689404934688,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958490176829316,
      "loss": 5.2641,
      "mean_token_accuracy": 0.18528165817260742,
      "num_tokens": 15992363.0,
      "step": 7690
    },
    {
      "entropy": 5.580470180511474,
      "epoch": 0.698022496371553,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495842828866941,
      "loss": 5.3566,
      "mean_token_accuracy": 0.17209097146987914,
      "num_tokens": 16003477.0,
      "step": 7695
    },
    {
      "entropy": 5.598789501190185,
      "epoch": 0.6984760522496372,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004958366354838645,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18081470876932143,
      "num_tokens": 16013747.0,
      "step": 7700
    },
    {
      "entropy": 5.516934061050415,
      "epoch": 0.6989296081277213,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049583043753383,
      "loss": 5.3303,
      "mean_token_accuracy": 0.18622443526983262,
      "num_tokens": 16025662.0,
      "step": 7705
    },
    {
      "entropy": 5.5949502944946286,
      "epoch": 0.6993831640058055,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004958242350169658,
      "loss": 5.407,
      "mean_token_accuracy": 0.18321597278118135,
      "num_tokens": 16037315.0,
      "step": 7710
    },
    {
      "entropy": 5.47448148727417,
      "epoch": 0.6998367198838897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004958180279334001,
      "loss": 5.2332,
      "mean_token_accuracy": 0.18107860833406447,
      "num_tokens": 16047297.0,
      "step": 7715
    },
    {
      "entropy": 5.4877598762512205,
      "epoch": 0.7002902757619739,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958118162832614,
      "loss": 5.2136,
      "mean_token_accuracy": 0.18834895491600037,
      "num_tokens": 16057573.0,
      "step": 7720
    },
    {
      "entropy": 5.5364970684051515,
      "epoch": 0.700743831640058,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004958056000666781,
      "loss": 5.3848,
      "mean_token_accuracy": 0.17313725650310516,
      "num_tokens": 16067926.0,
      "step": 7725
    },
    {
      "entropy": 5.547545146942139,
      "epoch": 0.7011973875181422,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957993792837786,
      "loss": 5.3135,
      "mean_token_accuracy": 0.18305060714483262,
      "num_tokens": 16077938.0,
      "step": 7730
    },
    {
      "entropy": 5.518269681930542,
      "epoch": 0.7016509433962265,
      "grad_norm": 1.0,
      "learning_rate": 0.0004957931539346917,
      "loss": 5.2126,
      "mean_token_accuracy": 0.183458811044693,
      "num_tokens": 16088163.0,
      "step": 7735
    },
    {
      "entropy": 5.6031519889831545,
      "epoch": 0.7021044992743106,
      "grad_norm": 0.875,
      "learning_rate": 0.0004957869240195462,
      "loss": 5.4064,
      "mean_token_accuracy": 0.1752445250749588,
      "num_tokens": 16100640.0,
      "step": 7740
    },
    {
      "entropy": 5.54415545463562,
      "epoch": 0.7025580551523948,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004957806895384707,
      "loss": 5.2409,
      "mean_token_accuracy": 0.18716129958629607,
      "num_tokens": 16111018.0,
      "step": 7745
    },
    {
      "entropy": 5.503624105453492,
      "epoch": 0.7030116110304789,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957744504915944,
      "loss": 5.2766,
      "mean_token_accuracy": 0.17956742346286775,
      "num_tokens": 16120873.0,
      "step": 7750
    },
    {
      "entropy": 5.564220190048218,
      "epoch": 0.7034651669085631,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957682068790463,
      "loss": 5.2998,
      "mean_token_accuracy": 0.17889865934848787,
      "num_tokens": 16131178.0,
      "step": 7755
    },
    {
      "entropy": 5.516436815261841,
      "epoch": 0.7039187227866474,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957619587009553,
      "loss": 5.3,
      "mean_token_accuracy": 0.18320292234420776,
      "num_tokens": 16141156.0,
      "step": 7760
    },
    {
      "entropy": 5.544557237625122,
      "epoch": 0.7043722786647315,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957557059574507,
      "loss": 5.405,
      "mean_token_accuracy": 0.17417299449443818,
      "num_tokens": 16150077.0,
      "step": 7765
    },
    {
      "entropy": 5.541102647781372,
      "epoch": 0.7048258345428157,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957494486486618,
      "loss": 5.3246,
      "mean_token_accuracy": 0.17782566845417022,
      "num_tokens": 16160514.0,
      "step": 7770
    },
    {
      "entropy": 5.551746559143067,
      "epoch": 0.7052793904208998,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957431867747182,
      "loss": 5.3671,
      "mean_token_accuracy": 0.18455520123243332,
      "num_tokens": 16170811.0,
      "step": 7775
    },
    {
      "entropy": 5.478965806961059,
      "epoch": 0.7057329462989841,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957369203357492,
      "loss": 5.2566,
      "mean_token_accuracy": 0.1847606912255287,
      "num_tokens": 16181553.0,
      "step": 7780
    },
    {
      "entropy": 5.55667085647583,
      "epoch": 0.7061865021770682,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004957306493318844,
      "loss": 5.2794,
      "mean_token_accuracy": 0.1900545060634613,
      "num_tokens": 16191308.0,
      "step": 7785
    },
    {
      "entropy": 5.5251086235046385,
      "epoch": 0.7066400580551524,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957243737632535,
      "loss": 5.2563,
      "mean_token_accuracy": 0.18701473474502564,
      "num_tokens": 16201311.0,
      "step": 7790
    },
    {
      "entropy": 5.555533313751221,
      "epoch": 0.7070936139332366,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957180936299863,
      "loss": 5.3075,
      "mean_token_accuracy": 0.1831939309835434,
      "num_tokens": 16211105.0,
      "step": 7795
    },
    {
      "entropy": 5.557322025299072,
      "epoch": 0.7075471698113207,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957118089322127,
      "loss": 5.328,
      "mean_token_accuracy": 0.17431456744670867,
      "num_tokens": 16220836.0,
      "step": 7800
    },
    {
      "entropy": 5.510331630706787,
      "epoch": 0.708000725689405,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004957055196700628,
      "loss": 5.2236,
      "mean_token_accuracy": 0.1855901911854744,
      "num_tokens": 16231630.0,
      "step": 7805
    },
    {
      "entropy": 5.560601472854614,
      "epoch": 0.7084542815674891,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956992258436663,
      "loss": 5.3236,
      "mean_token_accuracy": 0.18146177083253862,
      "num_tokens": 16242272.0,
      "step": 7810
    },
    {
      "entropy": 5.4510585308074955,
      "epoch": 0.7089078374455733,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956929274531536,
      "loss": 5.235,
      "mean_token_accuracy": 0.1826574593782425,
      "num_tokens": 16252555.0,
      "step": 7815
    },
    {
      "entropy": 5.519144773483276,
      "epoch": 0.7093613933236574,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000495686624498655,
      "loss": 5.3509,
      "mean_token_accuracy": 0.18719311952590942,
      "num_tokens": 16261964.0,
      "step": 7820
    },
    {
      "entropy": 5.61521577835083,
      "epoch": 0.7098149492017417,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004956803169803007,
      "loss": 5.3896,
      "mean_token_accuracy": 0.17692660838365554,
      "num_tokens": 16274543.0,
      "step": 7825
    },
    {
      "entropy": 5.509320163726807,
      "epoch": 0.7102685050798259,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004956740048982212,
      "loss": 5.3055,
      "mean_token_accuracy": 0.17120635658502578,
      "num_tokens": 16285124.0,
      "step": 7830
    },
    {
      "entropy": 5.523042774200439,
      "epoch": 0.71072206095791,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004956676882525472,
      "loss": 5.3757,
      "mean_token_accuracy": 0.17950959950685502,
      "num_tokens": 16295727.0,
      "step": 7835
    },
    {
      "entropy": 5.622578430175781,
      "epoch": 0.7111756168359942,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495661367043409,
      "loss": 5.2772,
      "mean_token_accuracy": 0.19084594547748565,
      "num_tokens": 16305496.0,
      "step": 7840
    },
    {
      "entropy": 5.515881633758545,
      "epoch": 0.7116291727140783,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956550412709376,
      "loss": 5.2645,
      "mean_token_accuracy": 0.18688923120498657,
      "num_tokens": 16315112.0,
      "step": 7845
    },
    {
      "entropy": 5.537652444839478,
      "epoch": 0.7120827285921626,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956487109352637,
      "loss": 5.304,
      "mean_token_accuracy": 0.18639636486768724,
      "num_tokens": 16325613.0,
      "step": 7850
    },
    {
      "entropy": 5.522394752502441,
      "epoch": 0.7125362844702468,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956423760365181,
      "loss": 5.2433,
      "mean_token_accuracy": 0.18728332221508026,
      "num_tokens": 16335317.0,
      "step": 7855
    },
    {
      "entropy": 5.5846155166625975,
      "epoch": 0.7129898403483309,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956360365748321,
      "loss": 5.3926,
      "mean_token_accuracy": 0.17209500968456268,
      "num_tokens": 16345003.0,
      "step": 7860
    },
    {
      "entropy": 5.571717071533203,
      "epoch": 0.7134433962264151,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004956296925503367,
      "loss": 5.379,
      "mean_token_accuracy": 0.17814605087041854,
      "num_tokens": 16356108.0,
      "step": 7865
    },
    {
      "entropy": 5.506393384933472,
      "epoch": 0.7138969521044993,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956233439631628,
      "loss": 5.1874,
      "mean_token_accuracy": 0.18501874059438705,
      "num_tokens": 16364733.0,
      "step": 7870
    },
    {
      "entropy": 5.588735723495484,
      "epoch": 0.7143505079825835,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956169908134421,
      "loss": 5.2879,
      "mean_token_accuracy": 0.18495833426713942,
      "num_tokens": 16375229.0,
      "step": 7875
    },
    {
      "entropy": 5.532235622406006,
      "epoch": 0.7148040638606676,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004956106331013057,
      "loss": 5.3668,
      "mean_token_accuracy": 0.18426259309053422,
      "num_tokens": 16386366.0,
      "step": 7880
    },
    {
      "entropy": 5.4934896469116214,
      "epoch": 0.7152576197387518,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004956042708268853,
      "loss": 5.2602,
      "mean_token_accuracy": 0.1904539719223976,
      "num_tokens": 16397067.0,
      "step": 7885
    },
    {
      "entropy": 5.464821481704712,
      "epoch": 0.715711175616836,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004955979039903123,
      "loss": 5.2136,
      "mean_token_accuracy": 0.1833643287420273,
      "num_tokens": 16407256.0,
      "step": 7890
    },
    {
      "entropy": 5.49287257194519,
      "epoch": 0.7161647314949202,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955915325917183,
      "loss": 5.2733,
      "mean_token_accuracy": 0.1792703688144684,
      "num_tokens": 16418701.0,
      "step": 7895
    },
    {
      "entropy": 5.476460409164429,
      "epoch": 0.7166182873730044,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955851566312354,
      "loss": 5.1877,
      "mean_token_accuracy": 0.1884955197572708,
      "num_tokens": 16428019.0,
      "step": 7900
    },
    {
      "entropy": 5.452414226531983,
      "epoch": 0.7170718432510885,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955787761089952,
      "loss": 5.2773,
      "mean_token_accuracy": 0.18074552565813065,
      "num_tokens": 16440122.0,
      "step": 7905
    },
    {
      "entropy": 5.62509765625,
      "epoch": 0.7175253991291727,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955723910251296,
      "loss": 5.3613,
      "mean_token_accuracy": 0.17514893114566804,
      "num_tokens": 16449801.0,
      "step": 7910
    },
    {
      "entropy": 5.495181322097778,
      "epoch": 0.717978955007257,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955660013797709,
      "loss": 5.2102,
      "mean_token_accuracy": 0.18785545378923416,
      "num_tokens": 16460155.0,
      "step": 7915
    },
    {
      "entropy": 5.575676202774048,
      "epoch": 0.7184325108853411,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495559607173051,
      "loss": 5.4192,
      "mean_token_accuracy": 0.17114308923482896,
      "num_tokens": 16470656.0,
      "step": 7920
    },
    {
      "entropy": 5.56002631187439,
      "epoch": 0.7188860667634253,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004955532084051022,
      "loss": 5.3067,
      "mean_token_accuracy": 0.18146676570177078,
      "num_tokens": 16481449.0,
      "step": 7925
    },
    {
      "entropy": 5.530605125427246,
      "epoch": 0.7193396226415094,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955468050760568,
      "loss": 5.3295,
      "mean_token_accuracy": 0.1831614077091217,
      "num_tokens": 16492032.0,
      "step": 7930
    },
    {
      "entropy": 5.593453073501587,
      "epoch": 0.7197931785195936,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004955403971860474,
      "loss": 5.2551,
      "mean_token_accuracy": 0.18039434105157853,
      "num_tokens": 16502221.0,
      "step": 7935
    },
    {
      "entropy": 5.539740943908692,
      "epoch": 0.7202467343976778,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955339847352063,
      "loss": 5.2528,
      "mean_token_accuracy": 0.18457072079181672,
      "num_tokens": 16512021.0,
      "step": 7940
    },
    {
      "entropy": 5.499135541915893,
      "epoch": 0.720700290275762,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955275677236662,
      "loss": 5.3026,
      "mean_token_accuracy": 0.1828033670783043,
      "num_tokens": 16522060.0,
      "step": 7945
    },
    {
      "entropy": 5.528159189224243,
      "epoch": 0.7211538461538461,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955211461515599,
      "loss": 5.2898,
      "mean_token_accuracy": 0.1860054388642311,
      "num_tokens": 16532827.0,
      "step": 7950
    },
    {
      "entropy": 5.551040935516357,
      "epoch": 0.7216074020319303,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049551472001902,
      "loss": 5.3431,
      "mean_token_accuracy": 0.18500829637050628,
      "num_tokens": 16542838.0,
      "step": 7955
    },
    {
      "entropy": 5.509801578521729,
      "epoch": 0.7220609579100146,
      "grad_norm": 0.875,
      "learning_rate": 0.0004955082893261796,
      "loss": 5.2815,
      "mean_token_accuracy": 0.18686290830373764,
      "num_tokens": 16552615.0,
      "step": 7960
    },
    {
      "entropy": 5.583999919891357,
      "epoch": 0.7225145137880987,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004955018540731714,
      "loss": 5.3489,
      "mean_token_accuracy": 0.18032243102788925,
      "num_tokens": 16561777.0,
      "step": 7965
    },
    {
      "entropy": 5.553142452239991,
      "epoch": 0.7229680696661829,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004954954142601288,
      "loss": 5.3082,
      "mean_token_accuracy": 0.18332842886447906,
      "num_tokens": 16572111.0,
      "step": 7970
    },
    {
      "entropy": 5.5559593677520756,
      "epoch": 0.723421625544267,
      "grad_norm": 1.0,
      "learning_rate": 0.0004954889698871849,
      "loss": 5.3549,
      "mean_token_accuracy": 0.18638369292020798,
      "num_tokens": 16581674.0,
      "step": 7975
    },
    {
      "entropy": 5.570011758804322,
      "epoch": 0.7238751814223512,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004954825209544728,
      "loss": 5.335,
      "mean_token_accuracy": 0.19101521968841553,
      "num_tokens": 16591802.0,
      "step": 7980
    },
    {
      "entropy": 5.542544221878051,
      "epoch": 0.7243287373004355,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495476067462126,
      "loss": 5.2403,
      "mean_token_accuracy": 0.1896628573536873,
      "num_tokens": 16602772.0,
      "step": 7985
    },
    {
      "entropy": 5.562042999267578,
      "epoch": 0.7247822931785196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495469609410278,
      "loss": 5.2866,
      "mean_token_accuracy": 0.18497096598148347,
      "num_tokens": 16613526.0,
      "step": 7990
    },
    {
      "entropy": 5.4583722114562985,
      "epoch": 0.7252358490566038,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004954631467990622,
      "loss": 5.2867,
      "mean_token_accuracy": 0.18482371866703035,
      "num_tokens": 16623583.0,
      "step": 7995
    },
    {
      "entropy": 5.544550371170044,
      "epoch": 0.7256894049346879,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954566796286124,
      "loss": 5.346,
      "mean_token_accuracy": 0.17807176262140273,
      "num_tokens": 16635068.0,
      "step": 8000
    },
    {
      "entropy": 5.589382553100586,
      "epoch": 0.7261429608127722,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954502078990623,
      "loss": 5.2463,
      "mean_token_accuracy": 0.18150110691785812,
      "num_tokens": 16645021.0,
      "step": 8005
    },
    {
      "entropy": 5.525119733810425,
      "epoch": 0.7265965166908563,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004954437316105457,
      "loss": 5.333,
      "mean_token_accuracy": 0.1788739249110222,
      "num_tokens": 16656021.0,
      "step": 8010
    },
    {
      "entropy": 5.494053888320923,
      "epoch": 0.7270500725689405,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954372507631965,
      "loss": 5.2201,
      "mean_token_accuracy": 0.19220224618911744,
      "num_tokens": 16665236.0,
      "step": 8015
    },
    {
      "entropy": 5.558600378036499,
      "epoch": 0.7275036284470247,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954307653571489,
      "loss": 5.4189,
      "mean_token_accuracy": 0.17521561533212662,
      "num_tokens": 16676690.0,
      "step": 8020
    },
    {
      "entropy": 5.557137393951416,
      "epoch": 0.7279571843251088,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954242753925367,
      "loss": 5.3067,
      "mean_token_accuracy": 0.17874979972839355,
      "num_tokens": 16687288.0,
      "step": 8025
    },
    {
      "entropy": 5.520791292190552,
      "epoch": 0.7284107402031931,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954177808694946,
      "loss": 5.3288,
      "mean_token_accuracy": 0.17826912701129913,
      "num_tokens": 16697873.0,
      "step": 8030
    },
    {
      "entropy": 5.5447221279144285,
      "epoch": 0.7288642960812772,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954112817881565,
      "loss": 5.3312,
      "mean_token_accuracy": 0.17438385188579558,
      "num_tokens": 16709002.0,
      "step": 8035
    },
    {
      "entropy": 5.5151759624481205,
      "epoch": 0.7293178519593614,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954047781486568,
      "loss": 5.2998,
      "mean_token_accuracy": 0.1815633311867714,
      "num_tokens": 16718989.0,
      "step": 8040
    },
    {
      "entropy": 5.594701957702637,
      "epoch": 0.7297714078374455,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004953982699511303,
      "loss": 5.3723,
      "mean_token_accuracy": 0.1834247812628746,
      "num_tokens": 16730425.0,
      "step": 8045
    },
    {
      "entropy": 5.574563789367676,
      "epoch": 0.7302249637155298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953917571957114,
      "loss": 5.2844,
      "mean_token_accuracy": 0.18223824948072434,
      "num_tokens": 16740216.0,
      "step": 8050
    },
    {
      "entropy": 5.513731193542481,
      "epoch": 0.730678519593614,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953852398825346,
      "loss": 5.2599,
      "mean_token_accuracy": 0.1835854694247246,
      "num_tokens": 16752436.0,
      "step": 8055
    },
    {
      "entropy": 5.4929563999176025,
      "epoch": 0.7311320754716981,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495378718011735,
      "loss": 5.2833,
      "mean_token_accuracy": 0.18290273547172547,
      "num_tokens": 16762631.0,
      "step": 8060
    },
    {
      "entropy": 5.570620965957642,
      "epoch": 0.7315856313497823,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004953721915834473,
      "loss": 5.3062,
      "mean_token_accuracy": 0.18442205041646959,
      "num_tokens": 16772613.0,
      "step": 8065
    },
    {
      "entropy": 5.580399417877198,
      "epoch": 0.7320391872278664,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953656605978067,
      "loss": 5.3126,
      "mean_token_accuracy": 0.17444958239793779,
      "num_tokens": 16782963.0,
      "step": 8070
    },
    {
      "entropy": 5.502667856216431,
      "epoch": 0.7324927431059507,
      "grad_norm": 0.875,
      "learning_rate": 0.000495359125054948,
      "loss": 5.2461,
      "mean_token_accuracy": 0.1868554711341858,
      "num_tokens": 16793892.0,
      "step": 8075
    },
    {
      "entropy": 5.453513860702515,
      "epoch": 0.7329462989840348,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004953525849550063,
      "loss": 5.2373,
      "mean_token_accuracy": 0.17641406506299973,
      "num_tokens": 16803388.0,
      "step": 8080
    },
    {
      "entropy": 5.5092305660247805,
      "epoch": 0.733399854862119,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004953460402981172,
      "loss": 5.2064,
      "mean_token_accuracy": 0.18928418755531312,
      "num_tokens": 16813201.0,
      "step": 8085
    },
    {
      "entropy": 5.587248516082764,
      "epoch": 0.7338534107402032,
      "grad_norm": 0.875,
      "learning_rate": 0.0004953394910844157,
      "loss": 5.4576,
      "mean_token_accuracy": 0.17749661803245545,
      "num_tokens": 16823265.0,
      "step": 8090
    },
    {
      "entropy": 5.514876079559326,
      "epoch": 0.7343069666182874,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953329373140374,
      "loss": 5.2368,
      "mean_token_accuracy": 0.18455445468425752,
      "num_tokens": 16833057.0,
      "step": 8095
    },
    {
      "entropy": 5.529699420928955,
      "epoch": 0.7347605224963716,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004953263789871179,
      "loss": 5.2033,
      "mean_token_accuracy": 0.18069398403167725,
      "num_tokens": 16843488.0,
      "step": 8100
    },
    {
      "entropy": 5.451847314834595,
      "epoch": 0.7352140783744557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953198161037925,
      "loss": 5.2423,
      "mean_token_accuracy": 0.1972143605351448,
      "num_tokens": 16852448.0,
      "step": 8105
    },
    {
      "entropy": 5.398800182342529,
      "epoch": 0.7356676342525399,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953132486641975,
      "loss": 5.2678,
      "mean_token_accuracy": 0.182267963886261,
      "num_tokens": 16863944.0,
      "step": 8110
    },
    {
      "entropy": 5.561413669586182,
      "epoch": 0.736121190130624,
      "grad_norm": 1.078125,
      "learning_rate": 0.000495306676668468,
      "loss": 5.1982,
      "mean_token_accuracy": 0.18368888199329375,
      "num_tokens": 16873923.0,
      "step": 8115
    },
    {
      "entropy": 5.5462486743927,
      "epoch": 0.7365747460087083,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004953001001167406,
      "loss": 5.3103,
      "mean_token_accuracy": 0.17963693737983705,
      "num_tokens": 16884373.0,
      "step": 8120
    },
    {
      "entropy": 5.5512104511260985,
      "epoch": 0.7370283018867925,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952935190091508,
      "loss": 5.4311,
      "mean_token_accuracy": 0.17719766050577163,
      "num_tokens": 16895869.0,
      "step": 8125
    },
    {
      "entropy": 5.482508754730224,
      "epoch": 0.7374818577648766,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952869333458349,
      "loss": 5.2713,
      "mean_token_accuracy": 0.1831250846385956,
      "num_tokens": 16906737.0,
      "step": 8130
    },
    {
      "entropy": 5.466742324829101,
      "epoch": 0.7379354136429608,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004952803431269292,
      "loss": 5.2061,
      "mean_token_accuracy": 0.18994556814432145,
      "num_tokens": 16916083.0,
      "step": 8135
    },
    {
      "entropy": 5.5881688594818115,
      "epoch": 0.738388969521045,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004952737483525699,
      "loss": 5.3836,
      "mean_token_accuracy": 0.182338385283947,
      "num_tokens": 16926419.0,
      "step": 8140
    },
    {
      "entropy": 5.624546480178833,
      "epoch": 0.7388425253991292,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004952671490228932,
      "loss": 5.3094,
      "mean_token_accuracy": 0.18363170474767684,
      "num_tokens": 16936067.0,
      "step": 8145
    },
    {
      "entropy": 5.578293991088867,
      "epoch": 0.7392960812772134,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004952605451380357,
      "loss": 5.253,
      "mean_token_accuracy": 0.18858592212200165,
      "num_tokens": 16946571.0,
      "step": 8150
    },
    {
      "entropy": 5.479947805404663,
      "epoch": 0.7397496371552975,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004952539366981342,
      "loss": 5.3013,
      "mean_token_accuracy": 0.1838264137506485,
      "num_tokens": 16956764.0,
      "step": 8155
    },
    {
      "entropy": 5.496211385726928,
      "epoch": 0.7402031930333817,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000495247323703325,
      "loss": 5.2438,
      "mean_token_accuracy": 0.192402121424675,
      "num_tokens": 16966002.0,
      "step": 8160
    },
    {
      "entropy": 5.586853837966919,
      "epoch": 0.7406567489114659,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000495240706153745,
      "loss": 5.341,
      "mean_token_accuracy": 0.1828116625547409,
      "num_tokens": 16975705.0,
      "step": 8165
    },
    {
      "entropy": 5.531956195831299,
      "epoch": 0.7411103047895501,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952340840495312,
      "loss": 5.24,
      "mean_token_accuracy": 0.18215106874704362,
      "num_tokens": 16985945.0,
      "step": 8170
    },
    {
      "entropy": 5.4779589653015135,
      "epoch": 0.7415638606676342,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952274573908203,
      "loss": 5.2237,
      "mean_token_accuracy": 0.1873907685279846,
      "num_tokens": 16995475.0,
      "step": 8175
    },
    {
      "entropy": 5.511494302749634,
      "epoch": 0.7420174165457184,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952208261777496,
      "loss": 5.2523,
      "mean_token_accuracy": 0.1826030820608139,
      "num_tokens": 17005028.0,
      "step": 8180
    },
    {
      "entropy": 5.54114727973938,
      "epoch": 0.7424709724238027,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000495214190410456,
      "loss": 5.309,
      "mean_token_accuracy": 0.18562476485967636,
      "num_tokens": 17014786.0,
      "step": 8185
    },
    {
      "entropy": 5.539929485321045,
      "epoch": 0.7429245283018868,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952075500890768,
      "loss": 5.3124,
      "mean_token_accuracy": 0.18346714675426484,
      "num_tokens": 17024431.0,
      "step": 8190
    },
    {
      "entropy": 5.508410024642944,
      "epoch": 0.743378084179971,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952009052137493,
      "loss": 5.2929,
      "mean_token_accuracy": 0.18714380860328675,
      "num_tokens": 17034368.0,
      "step": 8195
    },
    {
      "entropy": 5.503003263473511,
      "epoch": 0.7438316400580551,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951942557846112,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18206837624311448,
      "num_tokens": 17043818.0,
      "step": 8200
    },
    {
      "entropy": 5.5616518497467045,
      "epoch": 0.7442851959361393,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004951876018017997,
      "loss": 5.3205,
      "mean_token_accuracy": 0.18540932089090348,
      "num_tokens": 17054843.0,
      "step": 8205
    },
    {
      "entropy": 5.535712146759034,
      "epoch": 0.7447387518142236,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004951809432654524,
      "loss": 5.3265,
      "mean_token_accuracy": 0.182991661131382,
      "num_tokens": 17065539.0,
      "step": 8210
    },
    {
      "entropy": 5.500705003738403,
      "epoch": 0.7451923076923077,
      "grad_norm": 0.96875,
      "learning_rate": 0.000495174280175707,
      "loss": 5.2422,
      "mean_token_accuracy": 0.18335951268672943,
      "num_tokens": 17075443.0,
      "step": 8215
    },
    {
      "entropy": 5.61527829170227,
      "epoch": 0.7456458635703919,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004951676125327015,
      "loss": 5.3673,
      "mean_token_accuracy": 0.1847109615802765,
      "num_tokens": 17085490.0,
      "step": 8220
    },
    {
      "entropy": 5.5724077224731445,
      "epoch": 0.746099419448476,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004951609403365737,
      "loss": 5.3522,
      "mean_token_accuracy": 0.1829810321331024,
      "num_tokens": 17094803.0,
      "step": 8225
    },
    {
      "entropy": 5.525753211975098,
      "epoch": 0.7465529753265603,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004951542635874615,
      "loss": 5.2136,
      "mean_token_accuracy": 0.18302233666181564,
      "num_tokens": 17104373.0,
      "step": 8230
    },
    {
      "entropy": 5.557919216156006,
      "epoch": 0.7470065312046444,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004951475822855029,
      "loss": 5.2884,
      "mean_token_accuracy": 0.18372104465961456,
      "num_tokens": 17115469.0,
      "step": 8235
    },
    {
      "entropy": 5.532179355621338,
      "epoch": 0.7474600870827286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951408964308364,
      "loss": 5.2668,
      "mean_token_accuracy": 0.18367121368646622,
      "num_tokens": 17125443.0,
      "step": 8240
    },
    {
      "entropy": 5.463788461685181,
      "epoch": 0.7479136429608128,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951342060236,
      "loss": 5.2343,
      "mean_token_accuracy": 0.18693330585956575,
      "num_tokens": 17136253.0,
      "step": 8245
    },
    {
      "entropy": 5.573167133331299,
      "epoch": 0.7483671988388969,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004951275110639321,
      "loss": 5.2911,
      "mean_token_accuracy": 0.18700899928808212,
      "num_tokens": 17145544.0,
      "step": 8250
    },
    {
      "entropy": 5.55022964477539,
      "epoch": 0.7488207547169812,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004951208115519712,
      "loss": 5.2952,
      "mean_token_accuracy": 0.18260560929775238,
      "num_tokens": 17155043.0,
      "step": 8255
    },
    {
      "entropy": 5.5254528522491455,
      "epoch": 0.7492743105950653,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951141074878558,
      "loss": 5.3393,
      "mean_token_accuracy": 0.1820871725678444,
      "num_tokens": 17164885.0,
      "step": 8260
    },
    {
      "entropy": 5.631131505966186,
      "epoch": 0.7497278664731495,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951073988717246,
      "loss": 5.3404,
      "mean_token_accuracy": 0.1810414507985115,
      "num_tokens": 17175681.0,
      "step": 8265
    },
    {
      "entropy": 5.473759317398072,
      "epoch": 0.7501814223512336,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004951006857037163,
      "loss": 5.213,
      "mean_token_accuracy": 0.1888574704527855,
      "num_tokens": 17186002.0,
      "step": 8270
    },
    {
      "entropy": 5.524263381958008,
      "epoch": 0.7506349782293179,
      "grad_norm": 1.0,
      "learning_rate": 0.0004950939679839696,
      "loss": 5.2646,
      "mean_token_accuracy": 0.18346552699804305,
      "num_tokens": 17195499.0,
      "step": 8275
    },
    {
      "entropy": 5.533853340148926,
      "epoch": 0.7510885341074021,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004950872457126237,
      "loss": 5.2716,
      "mean_token_accuracy": 0.18572352528572084,
      "num_tokens": 17204684.0,
      "step": 8280
    },
    {
      "entropy": 5.538997459411621,
      "epoch": 0.7515420899854862,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950805188898175,
      "loss": 5.3085,
      "mean_token_accuracy": 0.17798569202423095,
      "num_tokens": 17213987.0,
      "step": 8285
    },
    {
      "entropy": 5.532698488235473,
      "epoch": 0.7519956458635704,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049507378751569,
      "loss": 5.3141,
      "mean_token_accuracy": 0.17800866514444352,
      "num_tokens": 17224512.0,
      "step": 8290
    },
    {
      "entropy": 5.616262674331665,
      "epoch": 0.7524492017416545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004950670515903805,
      "loss": 5.3375,
      "mean_token_accuracy": 0.18572027534246444,
      "num_tokens": 17234362.0,
      "step": 8295
    },
    {
      "entropy": 5.554524564743042,
      "epoch": 0.7529027576197388,
      "grad_norm": 1.0,
      "learning_rate": 0.0004950603111140283,
      "loss": 5.3147,
      "mean_token_accuracy": 0.18971499800682068,
      "num_tokens": 17244241.0,
      "step": 8300
    },
    {
      "entropy": 5.4980474472045895,
      "epoch": 0.753356313497823,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004950535660867728,
      "loss": 5.2489,
      "mean_token_accuracy": 0.18815081119537352,
      "num_tokens": 17254689.0,
      "step": 8305
    },
    {
      "entropy": 5.564259672164917,
      "epoch": 0.7538098693759071,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004950468165087535,
      "loss": 5.3434,
      "mean_token_accuracy": 0.17959407716989517,
      "num_tokens": 17265506.0,
      "step": 8310
    },
    {
      "entropy": 5.543921327590942,
      "epoch": 0.7542634252539913,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004950400623801098,
      "loss": 5.4162,
      "mean_token_accuracy": 0.18204887360334396,
      "num_tokens": 17276376.0,
      "step": 8315
    },
    {
      "entropy": 5.59347243309021,
      "epoch": 0.7547169811320755,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950333037009818,
      "loss": 5.3005,
      "mean_token_accuracy": 0.17268142551183702,
      "num_tokens": 17287005.0,
      "step": 8320
    },
    {
      "entropy": 5.582093858718872,
      "epoch": 0.7551705370101597,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004950265404715089,
      "loss": 5.3053,
      "mean_token_accuracy": 0.17970646917819977,
      "num_tokens": 17297595.0,
      "step": 8325
    },
    {
      "entropy": 5.458274936676025,
      "epoch": 0.7556240928882438,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000495019772691831,
      "loss": 5.175,
      "mean_token_accuracy": 0.1917903169989586,
      "num_tokens": 17307319.0,
      "step": 8330
    },
    {
      "entropy": 5.572049331665039,
      "epoch": 0.756077648766328,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950130003620882,
      "loss": 5.4002,
      "mean_token_accuracy": 0.1848435267806053,
      "num_tokens": 17317978.0,
      "step": 8335
    },
    {
      "entropy": 5.51245903968811,
      "epoch": 0.7565312046444121,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004950062234824204,
      "loss": 5.2928,
      "mean_token_accuracy": 0.18260230571031572,
      "num_tokens": 17328339.0,
      "step": 8340
    },
    {
      "entropy": 5.52795376777649,
      "epoch": 0.7569847605224964,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949994420529679,
      "loss": 5.3443,
      "mean_token_accuracy": 0.18249064534902573,
      "num_tokens": 17338668.0,
      "step": 8345
    },
    {
      "entropy": 5.498517084121704,
      "epoch": 0.7574383164005806,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004949926560738709,
      "loss": 5.2917,
      "mean_token_accuracy": 0.177150858938694,
      "num_tokens": 17348609.0,
      "step": 8350
    },
    {
      "entropy": 5.578651428222656,
      "epoch": 0.7578918722786647,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004949858655452697,
      "loss": 5.2962,
      "mean_token_accuracy": 0.18345605581998825,
      "num_tokens": 17358014.0,
      "step": 8355
    },
    {
      "entropy": 5.5646138191223145,
      "epoch": 0.7583454281567489,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004949790704673048,
      "loss": 5.296,
      "mean_token_accuracy": 0.186521714925766,
      "num_tokens": 17367975.0,
      "step": 8360
    },
    {
      "entropy": 5.550845050811768,
      "epoch": 0.7587989840348331,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004949722708401166,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1873399719595909,
      "num_tokens": 17377960.0,
      "step": 8365
    },
    {
      "entropy": 5.570762348175049,
      "epoch": 0.7592525399129173,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004949654666638457,
      "loss": 5.3189,
      "mean_token_accuracy": 0.18686124682426453,
      "num_tokens": 17387334.0,
      "step": 8370
    },
    {
      "entropy": 5.525769948959351,
      "epoch": 0.7597060957910015,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494958657938633,
      "loss": 5.2811,
      "mean_token_accuracy": 0.1862280398607254,
      "num_tokens": 17396521.0,
      "step": 8375
    },
    {
      "entropy": 5.485896825790405,
      "epoch": 0.7601596516690856,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004949518446646192,
      "loss": 5.2506,
      "mean_token_accuracy": 0.18615950345993043,
      "num_tokens": 17407217.0,
      "step": 8380
    },
    {
      "entropy": 5.662817144393921,
      "epoch": 0.7606132075471698,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949450268419451,
      "loss": 5.3806,
      "mean_token_accuracy": 0.17769204825162888,
      "num_tokens": 17416982.0,
      "step": 8385
    },
    {
      "entropy": 5.483257293701172,
      "epoch": 0.761066763425254,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494938204470752,
      "loss": 5.2216,
      "mean_token_accuracy": 0.18211308419704436,
      "num_tokens": 17427636.0,
      "step": 8390
    },
    {
      "entropy": 5.508321857452392,
      "epoch": 0.7615203193033382,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004949313775511806,
      "loss": 5.3514,
      "mean_token_accuracy": 0.17798120081424712,
      "num_tokens": 17437488.0,
      "step": 8395
    },
    {
      "entropy": 5.528272867202759,
      "epoch": 0.7619738751814223,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004949245460833723,
      "loss": 5.1376,
      "mean_token_accuracy": 0.19336626231670379,
      "num_tokens": 17447838.0,
      "step": 8400
    },
    {
      "entropy": 5.554759931564331,
      "epoch": 0.7624274310595065,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949177100674684,
      "loss": 5.3676,
      "mean_token_accuracy": 0.18535151779651643,
      "num_tokens": 17457257.0,
      "step": 8405
    },
    {
      "entropy": 5.545297145843506,
      "epoch": 0.7628809869375908,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004949108695036102,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18526664525270461,
      "num_tokens": 17467729.0,
      "step": 8410
    },
    {
      "entropy": 5.5293680191040036,
      "epoch": 0.7633345428156749,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004949040243919392,
      "loss": 5.3014,
      "mean_token_accuracy": 0.17469221204519272,
      "num_tokens": 17477692.0,
      "step": 8415
    },
    {
      "entropy": 5.5139758586883545,
      "epoch": 0.7637880986937591,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948971747325968,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18337151855230333,
      "num_tokens": 17488730.0,
      "step": 8420
    },
    {
      "entropy": 5.5645856857299805,
      "epoch": 0.7642416545718432,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000494890320525725,
      "loss": 5.3605,
      "mean_token_accuracy": 0.18363076001405715,
      "num_tokens": 17498374.0,
      "step": 8425
    },
    {
      "entropy": 5.513505125045777,
      "epoch": 0.7646952104499274,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004948834617714652,
      "loss": 5.2246,
      "mean_token_accuracy": 0.19082147926092147,
      "num_tokens": 17508601.0,
      "step": 8430
    },
    {
      "entropy": 5.500728511810303,
      "epoch": 0.7651487663280117,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004948765984699595,
      "loss": 5.2605,
      "mean_token_accuracy": 0.1817544087767601,
      "num_tokens": 17518164.0,
      "step": 8435
    },
    {
      "entropy": 5.560157108306885,
      "epoch": 0.7656023222060958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004948697306213494,
      "loss": 5.2384,
      "mean_token_accuracy": 0.19223212003707885,
      "num_tokens": 17527244.0,
      "step": 8440
    },
    {
      "entropy": 5.5458314418792725,
      "epoch": 0.76605587808418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004948628582257775,
      "loss": 5.2992,
      "mean_token_accuracy": 0.18211422115564346,
      "num_tokens": 17538045.0,
      "step": 8445
    },
    {
      "entropy": 5.500934314727783,
      "epoch": 0.7665094339622641,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004948559812833857,
      "loss": 5.2567,
      "mean_token_accuracy": 0.18168097883462905,
      "num_tokens": 17547584.0,
      "step": 8450
    },
    {
      "entropy": 5.520407390594483,
      "epoch": 0.7669629898403484,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004948490997943162,
      "loss": 5.3063,
      "mean_token_accuracy": 0.17883863002061845,
      "num_tokens": 17558925.0,
      "step": 8455
    },
    {
      "entropy": 5.576011705398559,
      "epoch": 0.7674165457184325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004948422137587111,
      "loss": 5.3824,
      "mean_token_accuracy": 0.17435939311981202,
      "num_tokens": 17569355.0,
      "step": 8460
    },
    {
      "entropy": 5.576209831237793,
      "epoch": 0.7678701015965167,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004948353231767131,
      "loss": 5.2571,
      "mean_token_accuracy": 0.1864034727215767,
      "num_tokens": 17579782.0,
      "step": 8465
    },
    {
      "entropy": 5.518272352218628,
      "epoch": 0.7683236574746009,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004948284280484645,
      "loss": 5.301,
      "mean_token_accuracy": 0.19304556250572205,
      "num_tokens": 17589772.0,
      "step": 8470
    },
    {
      "entropy": 5.47354040145874,
      "epoch": 0.768777213352685,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004948215283741081,
      "loss": 5.2239,
      "mean_token_accuracy": 0.1867895171046257,
      "num_tokens": 17599613.0,
      "step": 8475
    },
    {
      "entropy": 5.533515357971192,
      "epoch": 0.7692307692307693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004948146241537866,
      "loss": 5.3134,
      "mean_token_accuracy": 0.17987559139728546,
      "num_tokens": 17609056.0,
      "step": 8480
    },
    {
      "entropy": 5.526946115493774,
      "epoch": 0.7696843251088534,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948077153876425,
      "loss": 5.2421,
      "mean_token_accuracy": 0.1833679959177971,
      "num_tokens": 17618620.0,
      "step": 8485
    },
    {
      "entropy": 5.498790073394775,
      "epoch": 0.7701378809869376,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948008020758188,
      "loss": 5.2163,
      "mean_token_accuracy": 0.190741603076458,
      "num_tokens": 17627630.0,
      "step": 8490
    },
    {
      "entropy": 5.430834245681763,
      "epoch": 0.7705914368650217,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004947938842184585,
      "loss": 5.1991,
      "mean_token_accuracy": 0.1896286591887474,
      "num_tokens": 17638903.0,
      "step": 8495
    },
    {
      "entropy": 5.4956201076507565,
      "epoch": 0.771044992743106,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947869618157048,
      "loss": 5.2584,
      "mean_token_accuracy": 0.18483953028917313,
      "num_tokens": 17649257.0,
      "step": 8500
    },
    {
      "entropy": 5.515977144241333,
      "epoch": 0.7714985486211902,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004947800348677006,
      "loss": 5.2287,
      "mean_token_accuracy": 0.189926940202713,
      "num_tokens": 17659495.0,
      "step": 8505
    },
    {
      "entropy": 5.451489019393921,
      "epoch": 0.7719521044992743,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004947731033745893,
      "loss": 5.1979,
      "mean_token_accuracy": 0.1842445269227028,
      "num_tokens": 17669671.0,
      "step": 8510
    },
    {
      "entropy": 5.574065208435059,
      "epoch": 0.7724056603773585,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004947661673365142,
      "loss": 5.3786,
      "mean_token_accuracy": 0.1808546781539917,
      "num_tokens": 17680927.0,
      "step": 8515
    },
    {
      "entropy": 5.518579769134521,
      "epoch": 0.7728592162554426,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947592267536188,
      "loss": 5.2422,
      "mean_token_accuracy": 0.18695141077041627,
      "num_tokens": 17691378.0,
      "step": 8520
    },
    {
      "entropy": 5.5204674243927006,
      "epoch": 0.7733127721335269,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004947522816260466,
      "loss": 5.327,
      "mean_token_accuracy": 0.18174126297235488,
      "num_tokens": 17700942.0,
      "step": 8525
    },
    {
      "entropy": 5.538609600067138,
      "epoch": 0.773766328011611,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947453319539411,
      "loss": 5.2441,
      "mean_token_accuracy": 0.18784799724817275,
      "num_tokens": 17710076.0,
      "step": 8530
    },
    {
      "entropy": 5.509739780426026,
      "epoch": 0.7742198838896952,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004947383777374462,
      "loss": 5.2607,
      "mean_token_accuracy": 0.18655650168657303,
      "num_tokens": 17721009.0,
      "step": 8535
    },
    {
      "entropy": 5.4949547290802006,
      "epoch": 0.7746734397677794,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004947314189767058,
      "loss": 5.2825,
      "mean_token_accuracy": 0.18507248163223267,
      "num_tokens": 17731821.0,
      "step": 8540
    },
    {
      "entropy": 5.481418752670288,
      "epoch": 0.7751269956458636,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004947244556718635,
      "loss": 5.1695,
      "mean_token_accuracy": 0.19244110435247422,
      "num_tokens": 17741425.0,
      "step": 8545
    },
    {
      "entropy": 5.49365496635437,
      "epoch": 0.7755805515239478,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004947174878230635,
      "loss": 5.2346,
      "mean_token_accuracy": 0.18750640451908113,
      "num_tokens": 17752381.0,
      "step": 8550
    },
    {
      "entropy": 5.427494478225708,
      "epoch": 0.7760341074020319,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947105154304498,
      "loss": 5.2263,
      "mean_token_accuracy": 0.18348337858915328,
      "num_tokens": 17762768.0,
      "step": 8555
    },
    {
      "entropy": 5.489666557312011,
      "epoch": 0.7764876632801161,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004947035384941667,
      "loss": 5.2553,
      "mean_token_accuracy": 0.1805388867855072,
      "num_tokens": 17774032.0,
      "step": 8560
    },
    {
      "entropy": 5.582749938964843,
      "epoch": 0.7769412191582002,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004946965570143584,
      "loss": 5.3812,
      "mean_token_accuracy": 0.17563943713903427,
      "num_tokens": 17785098.0,
      "step": 8565
    },
    {
      "entropy": 5.570175743103027,
      "epoch": 0.7773947750362845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946895709911695,
      "loss": 5.2531,
      "mean_token_accuracy": 0.1870201513171196,
      "num_tokens": 17795487.0,
      "step": 8570
    },
    {
      "entropy": 5.5367103099823,
      "epoch": 0.7778483309143687,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946825804247442,
      "loss": 5.3738,
      "mean_token_accuracy": 0.17667229026556014,
      "num_tokens": 17806045.0,
      "step": 8575
    },
    {
      "entropy": 5.615284633636475,
      "epoch": 0.7783018867924528,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000494675585315227,
      "loss": 5.3441,
      "mean_token_accuracy": 0.17770788669586182,
      "num_tokens": 17816948.0,
      "step": 8580
    },
    {
      "entropy": 5.56788387298584,
      "epoch": 0.778755442670537,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004946685856627628,
      "loss": 5.2718,
      "mean_token_accuracy": 0.18811210244894028,
      "num_tokens": 17826929.0,
      "step": 8585
    },
    {
      "entropy": 5.522318029403687,
      "epoch": 0.7792089985486212,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004946615814674963,
      "loss": 5.3625,
      "mean_token_accuracy": 0.18369622081518172,
      "num_tokens": 17837780.0,
      "step": 8590
    },
    {
      "entropy": 5.443318319320679,
      "epoch": 0.7796625544267054,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004946545727295724,
      "loss": 5.1325,
      "mean_token_accuracy": 0.1949272483587265,
      "num_tokens": 17847835.0,
      "step": 8595
    },
    {
      "entropy": 5.502077627182007,
      "epoch": 0.7801161103047896,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004946475594491359,
      "loss": 5.2175,
      "mean_token_accuracy": 0.19205821305513382,
      "num_tokens": 17858324.0,
      "step": 8600
    },
    {
      "entropy": 5.500710916519165,
      "epoch": 0.7805696661828737,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946405416263319,
      "loss": 5.2842,
      "mean_token_accuracy": 0.17803624123334885,
      "num_tokens": 17868455.0,
      "step": 8605
    },
    {
      "entropy": 5.545377159118653,
      "epoch": 0.7810232220609579,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946335192613054,
      "loss": 5.2662,
      "mean_token_accuracy": 0.18817005157470704,
      "num_tokens": 17879038.0,
      "step": 8610
    },
    {
      "entropy": 5.456969118118286,
      "epoch": 0.7814767779390421,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004946264923542019,
      "loss": 5.1368,
      "mean_token_accuracy": 0.19142608642578124,
      "num_tokens": 17888153.0,
      "step": 8615
    },
    {
      "entropy": 5.476630306243896,
      "epoch": 0.7819303338171263,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004946194609051664,
      "loss": 5.262,
      "mean_token_accuracy": 0.18682017773389817,
      "num_tokens": 17898289.0,
      "step": 8620
    },
    {
      "entropy": 5.496731328964233,
      "epoch": 0.7823838896952104,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004946124249143447,
      "loss": 5.2527,
      "mean_token_accuracy": 0.18772162050008773,
      "num_tokens": 17909192.0,
      "step": 8625
    },
    {
      "entropy": 5.486784315109253,
      "epoch": 0.7828374455732946,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000494605384381882,
      "loss": 5.2113,
      "mean_token_accuracy": 0.18046634793281555,
      "num_tokens": 17918937.0,
      "step": 8630
    },
    {
      "entropy": 5.44954481124878,
      "epoch": 0.7832910014513788,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945983393079239,
      "loss": 5.2437,
      "mean_token_accuracy": 0.19104671776294707,
      "num_tokens": 17929346.0,
      "step": 8635
    },
    {
      "entropy": 5.5133326053619385,
      "epoch": 0.783744557329463,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945912896926163,
      "loss": 5.2583,
      "mean_token_accuracy": 0.18662541210651398,
      "num_tokens": 17939479.0,
      "step": 8640
    },
    {
      "entropy": 5.582485771179199,
      "epoch": 0.7841981132075472,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945842355361049,
      "loss": 5.3415,
      "mean_token_accuracy": 0.1828889235854149,
      "num_tokens": 17949841.0,
      "step": 8645
    },
    {
      "entropy": 5.457309579849243,
      "epoch": 0.7846516690856313,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004945771768385355,
      "loss": 5.2602,
      "mean_token_accuracy": 0.19243107438087464,
      "num_tokens": 17960334.0,
      "step": 8650
    },
    {
      "entropy": 5.485492467880249,
      "epoch": 0.7851052249637155,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494570113600054,
      "loss": 5.2071,
      "mean_token_accuracy": 0.19041529893875123,
      "num_tokens": 17970595.0,
      "step": 8655
    },
    {
      "entropy": 5.548058986663818,
      "epoch": 0.7855587808417998,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004945630458208068,
      "loss": 5.3477,
      "mean_token_accuracy": 0.17438185960054398,
      "num_tokens": 17981301.0,
      "step": 8660
    },
    {
      "entropy": 5.470705938339234,
      "epoch": 0.7860123367198839,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945559735009397,
      "loss": 5.1765,
      "mean_token_accuracy": 0.19010596424341203,
      "num_tokens": 17991262.0,
      "step": 8665
    },
    {
      "entropy": 5.533977603912353,
      "epoch": 0.7864658925979681,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945488966405993,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18723469227552414,
      "num_tokens": 18001212.0,
      "step": 8670
    },
    {
      "entropy": 5.502318048477173,
      "epoch": 0.7869194484760522,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945418152399317,
      "loss": 5.2378,
      "mean_token_accuracy": 0.1913598209619522,
      "num_tokens": 18010432.0,
      "step": 8675
    },
    {
      "entropy": 5.518057823181152,
      "epoch": 0.7873730043541364,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945347292990833,
      "loss": 5.2782,
      "mean_token_accuracy": 0.18441058844327926,
      "num_tokens": 18020902.0,
      "step": 8680
    },
    {
      "entropy": 5.491929054260254,
      "epoch": 0.7878265602322206,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004945276388182008,
      "loss": 5.1539,
      "mean_token_accuracy": 0.1898062765598297,
      "num_tokens": 18031133.0,
      "step": 8685
    },
    {
      "entropy": 5.53517780303955,
      "epoch": 0.7882801161103048,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945205437974309,
      "loss": 5.255,
      "mean_token_accuracy": 0.18925394713878632,
      "num_tokens": 18042050.0,
      "step": 8690
    },
    {
      "entropy": 5.424103832244873,
      "epoch": 0.788733671988389,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004945134442369201,
      "loss": 5.2306,
      "mean_token_accuracy": 0.1864193484187126,
      "num_tokens": 18051642.0,
      "step": 8695
    },
    {
      "entropy": 5.54458270072937,
      "epoch": 0.7891872278664731,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004945063401368154,
      "loss": 5.2821,
      "mean_token_accuracy": 0.17556448727846147,
      "num_tokens": 18062180.0,
      "step": 8700
    },
    {
      "entropy": 5.519570016860962,
      "epoch": 0.7896407837445574,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004944992314972637,
      "loss": 5.2274,
      "mean_token_accuracy": 0.18374626189470292,
      "num_tokens": 18072302.0,
      "step": 8705
    },
    {
      "entropy": 5.469928026199341,
      "epoch": 0.7900943396226415,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004944921183184118,
      "loss": 5.2358,
      "mean_token_accuracy": 0.19111284017562866,
      "num_tokens": 18081824.0,
      "step": 8710
    },
    {
      "entropy": 5.518294858932495,
      "epoch": 0.7905478955007257,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494485000600407,
      "loss": 5.2924,
      "mean_token_accuracy": 0.188755401968956,
      "num_tokens": 18092696.0,
      "step": 8715
    },
    {
      "entropy": 5.5013707160949705,
      "epoch": 0.7910014513788098,
      "grad_norm": 1.0,
      "learning_rate": 0.0004944778783433965,
      "loss": 5.1994,
      "mean_token_accuracy": 0.19762906432151794,
      "num_tokens": 18101790.0,
      "step": 8720
    },
    {
      "entropy": 5.563456010818482,
      "epoch": 0.791455007256894,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004944707515475276,
      "loss": 5.3227,
      "mean_token_accuracy": 0.18671682476997375,
      "num_tokens": 18111572.0,
      "step": 8725
    },
    {
      "entropy": 5.548861169815064,
      "epoch": 0.7919085631349783,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944636202129475,
      "loss": 5.3125,
      "mean_token_accuracy": 0.1888110116124153,
      "num_tokens": 18120650.0,
      "step": 8730
    },
    {
      "entropy": 5.486426162719726,
      "epoch": 0.7923621190130624,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944564843398039,
      "loss": 5.1831,
      "mean_token_accuracy": 0.1958876892924309,
      "num_tokens": 18131055.0,
      "step": 8735
    },
    {
      "entropy": 5.5568522930145265,
      "epoch": 0.7928156748911466,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004944493439282443,
      "loss": 5.292,
      "mean_token_accuracy": 0.18856184631586076,
      "num_tokens": 18141423.0,
      "step": 8740
    },
    {
      "entropy": 5.456881761550903,
      "epoch": 0.7932692307692307,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004944421989784162,
      "loss": 5.2098,
      "mean_token_accuracy": 0.18828487694263457,
      "num_tokens": 18152015.0,
      "step": 8745
    },
    {
      "entropy": 5.548474550247192,
      "epoch": 0.793722786647315,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944350494904676,
      "loss": 5.3002,
      "mean_token_accuracy": 0.18480850607156754,
      "num_tokens": 18161934.0,
      "step": 8750
    },
    {
      "entropy": 5.479960775375366,
      "epoch": 0.7941763425253991,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944278954645463,
      "loss": 5.2243,
      "mean_token_accuracy": 0.18462625294923782,
      "num_tokens": 18172287.0,
      "step": 8755
    },
    {
      "entropy": 5.498562955856324,
      "epoch": 0.7946298984034833,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004944207369008002,
      "loss": 5.2699,
      "mean_token_accuracy": 0.1924784377217293,
      "num_tokens": 18181015.0,
      "step": 8760
    },
    {
      "entropy": 5.600859260559082,
      "epoch": 0.7950834542815675,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004944135737993773,
      "loss": 5.3619,
      "mean_token_accuracy": 0.1822371244430542,
      "num_tokens": 18192382.0,
      "step": 8765
    },
    {
      "entropy": 5.50996584892273,
      "epoch": 0.7955370101596516,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004944064061604258,
      "loss": 5.2123,
      "mean_token_accuracy": 0.19568186849355698,
      "num_tokens": 18202597.0,
      "step": 8770
    },
    {
      "entropy": 5.5370594501495365,
      "epoch": 0.7959905660377359,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943992339840939,
      "loss": 5.2795,
      "mean_token_accuracy": 0.1780181646347046,
      "num_tokens": 18213756.0,
      "step": 8775
    },
    {
      "entropy": 5.419239091873169,
      "epoch": 0.79644412191582,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00049439205727053,
      "loss": 5.2022,
      "mean_token_accuracy": 0.18264902830123902,
      "num_tokens": 18223669.0,
      "step": 8780
    },
    {
      "entropy": 5.498378324508667,
      "epoch": 0.7968976777939042,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004943848760198823,
      "loss": 5.2325,
      "mean_token_accuracy": 0.19068809002637863,
      "num_tokens": 18234884.0,
      "step": 8785
    },
    {
      "entropy": 5.486072826385498,
      "epoch": 0.7973512336719883,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943776902322995,
      "loss": 5.1603,
      "mean_token_accuracy": 0.19515904933214187,
      "num_tokens": 18245155.0,
      "step": 8790
    },
    {
      "entropy": 5.39517731666565,
      "epoch": 0.7978047895500726,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943704999079302,
      "loss": 5.1019,
      "mean_token_accuracy": 0.2020408481359482,
      "num_tokens": 18254686.0,
      "step": 8795
    },
    {
      "entropy": 5.470184278488159,
      "epoch": 0.7982583454281568,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004943633050469231,
      "loss": 5.1363,
      "mean_token_accuracy": 0.1927419751882553,
      "num_tokens": 18265473.0,
      "step": 8800
    },
    {
      "entropy": 5.523016786575317,
      "epoch": 0.7987119013062409,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004943561056494269,
      "loss": 5.3438,
      "mean_token_accuracy": 0.18321252465248108,
      "num_tokens": 18276024.0,
      "step": 8805
    },
    {
      "entropy": 5.513629722595215,
      "epoch": 0.7991654571843251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004943489017155906,
      "loss": 5.3079,
      "mean_token_accuracy": 0.17381245493888856,
      "num_tokens": 18287261.0,
      "step": 8810
    },
    {
      "entropy": 5.530607414245606,
      "epoch": 0.7996190130624092,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943416932455631,
      "loss": 5.3088,
      "mean_token_accuracy": 0.186485455930233,
      "num_tokens": 18299374.0,
      "step": 8815
    },
    {
      "entropy": 5.456578254699707,
      "epoch": 0.8000725689404935,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004943344802394935,
      "loss": 5.2843,
      "mean_token_accuracy": 0.1819461315870285,
      "num_tokens": 18309344.0,
      "step": 8820
    },
    {
      "entropy": 5.578178882598877,
      "epoch": 0.8005261248185777,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943272626975309,
      "loss": 5.2664,
      "mean_token_accuracy": 0.18679665625095368,
      "num_tokens": 18319778.0,
      "step": 8825
    },
    {
      "entropy": 5.534620904922486,
      "epoch": 0.8009796806966618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943200406198246,
      "loss": 5.2527,
      "mean_token_accuracy": 0.18746354728937148,
      "num_tokens": 18329787.0,
      "step": 8830
    },
    {
      "entropy": 5.45460352897644,
      "epoch": 0.801433236574746,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494312814006524,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19326054751873017,
      "num_tokens": 18340670.0,
      "step": 8835
    },
    {
      "entropy": 5.486133813858032,
      "epoch": 0.8018867924528302,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004943055828577784,
      "loss": 5.2071,
      "mean_token_accuracy": 0.18043867200613023,
      "num_tokens": 18352484.0,
      "step": 8840
    },
    {
      "entropy": 5.4944739818573,
      "epoch": 0.8023403483309144,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942983471737376,
      "loss": 5.1898,
      "mean_token_accuracy": 0.1878673404455185,
      "num_tokens": 18363103.0,
      "step": 8845
    },
    {
      "entropy": 5.427543020248413,
      "epoch": 0.8027939042089985,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942911069545511,
      "loss": 5.2345,
      "mean_token_accuracy": 0.1885596677660942,
      "num_tokens": 18373925.0,
      "step": 8850
    },
    {
      "entropy": 5.503770256042481,
      "epoch": 0.8032474600870827,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004942838622003686,
      "loss": 5.3038,
      "mean_token_accuracy": 0.18221967965364455,
      "num_tokens": 18384081.0,
      "step": 8855
    },
    {
      "entropy": 5.5707417011260985,
      "epoch": 0.8037010159651669,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942766129113399,
      "loss": 5.1835,
      "mean_token_accuracy": 0.18571112006902696,
      "num_tokens": 18392510.0,
      "step": 8860
    },
    {
      "entropy": 5.6038836479187015,
      "epoch": 0.8041545718432511,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004942693590876151,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18732109665870667,
      "num_tokens": 18402809.0,
      "step": 8865
    },
    {
      "entropy": 5.571398639678955,
      "epoch": 0.8046081277213353,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000494262100729344,
      "loss": 5.3388,
      "mean_token_accuracy": 0.18695981055498123,
      "num_tokens": 18413342.0,
      "step": 8870
    },
    {
      "entropy": 5.530296659469604,
      "epoch": 0.8050616835994194,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004942548378366768,
      "loss": 5.3305,
      "mean_token_accuracy": 0.18433057516813278,
      "num_tokens": 18424735.0,
      "step": 8875
    },
    {
      "entropy": 5.518608522415161,
      "epoch": 0.8055152394775036,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942475704097635,
      "loss": 5.2385,
      "mean_token_accuracy": 0.18847633004188538,
      "num_tokens": 18434797.0,
      "step": 8880
    },
    {
      "entropy": 5.474301147460937,
      "epoch": 0.8059687953555879,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004942402984487548,
      "loss": 5.2676,
      "mean_token_accuracy": 0.1808571606874466,
      "num_tokens": 18445456.0,
      "step": 8885
    },
    {
      "entropy": 5.495883464813232,
      "epoch": 0.806422351233672,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942330219538008,
      "loss": 5.2401,
      "mean_token_accuracy": 0.18492464423179628,
      "num_tokens": 18456728.0,
      "step": 8890
    },
    {
      "entropy": 5.701735019683838,
      "epoch": 0.8068759071117562,
      "grad_norm": 1.109375,
      "learning_rate": 0.000494225740925052,
      "loss": 5.3843,
      "mean_token_accuracy": 0.18247078955173493,
      "num_tokens": 18468813.0,
      "step": 8895
    },
    {
      "entropy": 5.576275253295899,
      "epoch": 0.8073294629898403,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942184553626591,
      "loss": 5.2811,
      "mean_token_accuracy": 0.18496292531490327,
      "num_tokens": 18477912.0,
      "step": 8900
    },
    {
      "entropy": 5.4743218421936035,
      "epoch": 0.8077830188679245,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942111652667726,
      "loss": 5.2282,
      "mean_token_accuracy": 0.19448551237583162,
      "num_tokens": 18488115.0,
      "step": 8905
    },
    {
      "entropy": 5.5077948570251465,
      "epoch": 0.8082365747460087,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004942038706375433,
      "loss": 5.2667,
      "mean_token_accuracy": 0.18090982586145402,
      "num_tokens": 18498853.0,
      "step": 8910
    },
    {
      "entropy": 5.5216956615448,
      "epoch": 0.8086901306240929,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004941965714751222,
      "loss": 5.3132,
      "mean_token_accuracy": 0.18184340447187425,
      "num_tokens": 18510047.0,
      "step": 8915
    },
    {
      "entropy": 5.518053579330444,
      "epoch": 0.809143686502177,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941892677796601,
      "loss": 5.2792,
      "mean_token_accuracy": 0.18840029388666152,
      "num_tokens": 18519673.0,
      "step": 8920
    },
    {
      "entropy": 5.470839023590088,
      "epoch": 0.8095972423802612,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941819595513081,
      "loss": 5.2919,
      "mean_token_accuracy": 0.18582451194524766,
      "num_tokens": 18530090.0,
      "step": 8925
    },
    {
      "entropy": 5.61807107925415,
      "epoch": 0.8100507982583455,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004941746467902173,
      "loss": 5.3426,
      "mean_token_accuracy": 0.18165055215358733,
      "num_tokens": 18541862.0,
      "step": 8930
    },
    {
      "entropy": 5.5928370475769045,
      "epoch": 0.8105043541364296,
      "grad_norm": 0.921875,
      "learning_rate": 0.000494167329496539,
      "loss": 5.336,
      "mean_token_accuracy": 0.17665406912565232,
      "num_tokens": 18551624.0,
      "step": 8935
    },
    {
      "entropy": 5.498425483703613,
      "epoch": 0.8109579100145138,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004941600076704244,
      "loss": 5.2709,
      "mean_token_accuracy": 0.18561748266220093,
      "num_tokens": 18561784.0,
      "step": 8940
    },
    {
      "entropy": 5.549371147155762,
      "epoch": 0.8114114658925979,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941526813120251,
      "loss": 5.3371,
      "mean_token_accuracy": 0.18125243335962296,
      "num_tokens": 18572110.0,
      "step": 8945
    },
    {
      "entropy": 5.547827100753784,
      "epoch": 0.8118650217706821,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004941453504214924,
      "loss": 5.2792,
      "mean_token_accuracy": 0.17551750242710112,
      "num_tokens": 18582012.0,
      "step": 8950
    },
    {
      "entropy": 5.443738269805908,
      "epoch": 0.8123185776487664,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004941380149989781,
      "loss": 5.1279,
      "mean_token_accuracy": 0.1954687461256981,
      "num_tokens": 18592850.0,
      "step": 8955
    },
    {
      "entropy": 5.469958543777466,
      "epoch": 0.8127721335268505,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004941306750446336,
      "loss": 5.3026,
      "mean_token_accuracy": 0.18596144616603852,
      "num_tokens": 18602590.0,
      "step": 8960
    },
    {
      "entropy": 5.473424339294434,
      "epoch": 0.8132256894049347,
      "grad_norm": 1.0,
      "learning_rate": 0.000494123330558611,
      "loss": 5.1992,
      "mean_token_accuracy": 0.1891501232981682,
      "num_tokens": 18612607.0,
      "step": 8965
    },
    {
      "entropy": 5.497609806060791,
      "epoch": 0.8136792452830188,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004941159815410622,
      "loss": 5.1835,
      "mean_token_accuracy": 0.19193311184644699,
      "num_tokens": 18622936.0,
      "step": 8970
    },
    {
      "entropy": 5.518314170837402,
      "epoch": 0.8141328011611031,
      "grad_norm": 1.046875,
      "learning_rate": 0.000494108627992139,
      "loss": 5.3462,
      "mean_token_accuracy": 0.17709500193595887,
      "num_tokens": 18633997.0,
      "step": 8975
    },
    {
      "entropy": 5.47372260093689,
      "epoch": 0.8145863570391872,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004941012699119936,
      "loss": 5.183,
      "mean_token_accuracy": 0.19488824605941774,
      "num_tokens": 18644347.0,
      "step": 8980
    },
    {
      "entropy": 5.532529067993164,
      "epoch": 0.8150399129172714,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940939073007779,
      "loss": 5.2676,
      "mean_token_accuracy": 0.18627604395151137,
      "num_tokens": 18654989.0,
      "step": 8985
    },
    {
      "entropy": 5.508123207092285,
      "epoch": 0.8154934687953556,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004940865401586446,
      "loss": 5.2336,
      "mean_token_accuracy": 0.1885010004043579,
      "num_tokens": 18665526.0,
      "step": 8990
    },
    {
      "entropy": 5.4809952735900875,
      "epoch": 0.8159470246734397,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940791684857458,
      "loss": 5.2181,
      "mean_token_accuracy": 0.1932695686817169,
      "num_tokens": 18676226.0,
      "step": 8995
    },
    {
      "entropy": 5.489676904678345,
      "epoch": 0.816400580551524,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940717922822339,
      "loss": 5.1614,
      "mean_token_accuracy": 0.1885027691721916,
      "num_tokens": 18686499.0,
      "step": 9000
    },
    {
      "epoch": 0.816400580551524,
      "eval_entropy": 5.369530144104591,
      "eval_loss": 5.262591361999512,
      "eval_mean_token_accuracy": 0.19346466002421142,
      "eval_num_tokens": 18686499.0,
      "eval_runtime": 20.8639,
      "eval_samples_per_second": 1694.795,
      "eval_steps_per_second": 211.849,
      "step": 9000
    },
    {
      "entropy": 5.573095512390137,
      "epoch": 0.8168541364296081,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004940644115482615,
      "loss": 5.2266,
      "mean_token_accuracy": 0.18318518102169037,
      "num_tokens": 18696543.0,
      "step": 9005
    },
    {
      "entropy": 5.455431079864502,
      "epoch": 0.8173076923076923,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004940570262839812,
      "loss": 5.2858,
      "mean_token_accuracy": 0.18343722075223923,
      "num_tokens": 18707029.0,
      "step": 9010
    },
    {
      "entropy": 5.486490535736084,
      "epoch": 0.8177612481857764,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004940496364895457,
      "loss": 5.2753,
      "mean_token_accuracy": 0.18756048530340194,
      "num_tokens": 18716937.0,
      "step": 9015
    },
    {
      "entropy": 5.61928539276123,
      "epoch": 0.8182148040638607,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940422421651082,
      "loss": 5.2611,
      "mean_token_accuracy": 0.18611169457435608,
      "num_tokens": 18727478.0,
      "step": 9020
    },
    {
      "entropy": 5.478516626358032,
      "epoch": 0.8186683599419449,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004940348433108211,
      "loss": 5.177,
      "mean_token_accuracy": 0.18611731976270676,
      "num_tokens": 18737339.0,
      "step": 9025
    },
    {
      "entropy": 5.460969638824463,
      "epoch": 0.819121915820029,
      "grad_norm": 0.875,
      "learning_rate": 0.0004940274399268377,
      "loss": 5.3027,
      "mean_token_accuracy": 0.18202684968709945,
      "num_tokens": 18747810.0,
      "step": 9030
    },
    {
      "entropy": 5.50574049949646,
      "epoch": 0.8195754716981132,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004940200320133109,
      "loss": 5.2689,
      "mean_token_accuracy": 0.18433540016412736,
      "num_tokens": 18757564.0,
      "step": 9035
    },
    {
      "entropy": 5.58858289718628,
      "epoch": 0.8200290275761973,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004940126195703941,
      "loss": 5.2907,
      "mean_token_accuracy": 0.17852483838796615,
      "num_tokens": 18768196.0,
      "step": 9040
    },
    {
      "entropy": 5.563227224349975,
      "epoch": 0.8204825834542816,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940052025982405,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19672567546367645,
      "num_tokens": 18778350.0,
      "step": 9045
    },
    {
      "entropy": 5.480757808685302,
      "epoch": 0.8209361393323658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004939977810970035,
      "loss": 5.2175,
      "mean_token_accuracy": 0.18037040233612062,
      "num_tokens": 18789825.0,
      "step": 9050
    },
    {
      "entropy": 5.474230337142944,
      "epoch": 0.8213896952104499,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939903550668366,
      "loss": 5.2226,
      "mean_token_accuracy": 0.18781944513320922,
      "num_tokens": 18799793.0,
      "step": 9055
    },
    {
      "entropy": 5.498127698898315,
      "epoch": 0.8218432510885341,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939829245078934,
      "loss": 5.2642,
      "mean_token_accuracy": 0.1901239588856697,
      "num_tokens": 18809771.0,
      "step": 9060
    },
    {
      "entropy": 5.501215839385987,
      "epoch": 0.8222968069666183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004939754894203275,
      "loss": 5.2714,
      "mean_token_accuracy": 0.1833363428711891,
      "num_tokens": 18820529.0,
      "step": 9065
    },
    {
      "entropy": 5.4365602970123295,
      "epoch": 0.8227503628447025,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004939680498042926,
      "loss": 5.1891,
      "mean_token_accuracy": 0.19763432294130326,
      "num_tokens": 18830797.0,
      "step": 9070
    },
    {
      "entropy": 5.490671968460083,
      "epoch": 0.8232039187227866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004939606056599428,
      "loss": 5.1978,
      "mean_token_accuracy": 0.18469800502061845,
      "num_tokens": 18840526.0,
      "step": 9075
    },
    {
      "entropy": 5.4513916015625,
      "epoch": 0.8236574746008708,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004939531569874316,
      "loss": 5.2314,
      "mean_token_accuracy": 0.1814821556210518,
      "num_tokens": 18851201.0,
      "step": 9080
    },
    {
      "entropy": 5.403475999832153,
      "epoch": 0.824111030478955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939457037869135,
      "loss": 5.0451,
      "mean_token_accuracy": 0.207870976626873,
      "num_tokens": 18861676.0,
      "step": 9085
    },
    {
      "entropy": 5.449726533889771,
      "epoch": 0.8245645863570392,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004939382460585424,
      "loss": 5.1931,
      "mean_token_accuracy": 0.18134815692901612,
      "num_tokens": 18872669.0,
      "step": 9090
    },
    {
      "entropy": 5.451654291152954,
      "epoch": 0.8250181422351234,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004939307838024726,
      "loss": 5.2454,
      "mean_token_accuracy": 0.18319953382015228,
      "num_tokens": 18882130.0,
      "step": 9095
    },
    {
      "entropy": 5.457068681716919,
      "epoch": 0.8254716981132075,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004939233170188585,
      "loss": 5.2186,
      "mean_token_accuracy": 0.18625770807266234,
      "num_tokens": 18892457.0,
      "step": 9100
    },
    {
      "entropy": 5.556639909744263,
      "epoch": 0.8259252539912917,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004939158457078543,
      "loss": 5.2405,
      "mean_token_accuracy": 0.18714095652103424,
      "num_tokens": 18903152.0,
      "step": 9105
    },
    {
      "entropy": 5.496951627731323,
      "epoch": 0.826378809869376,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004939083698696146,
      "loss": 5.208,
      "mean_token_accuracy": 0.18759516775608062,
      "num_tokens": 18913740.0,
      "step": 9110
    },
    {
      "entropy": 5.478869867324829,
      "epoch": 0.8268323657474601,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493900889504294,
      "loss": 5.2528,
      "mean_token_accuracy": 0.1898530900478363,
      "num_tokens": 18923501.0,
      "step": 9115
    },
    {
      "entropy": 5.482793998718262,
      "epoch": 0.8272859216255443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004938934046120473,
      "loss": 5.238,
      "mean_token_accuracy": 0.18646990209817887,
      "num_tokens": 18933255.0,
      "step": 9120
    },
    {
      "entropy": 5.451048803329468,
      "epoch": 0.8277394775036284,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004938859151930293,
      "loss": 5.2061,
      "mean_token_accuracy": 0.18621120303869249,
      "num_tokens": 18943649.0,
      "step": 9125
    },
    {
      "entropy": 5.4907585144042965,
      "epoch": 0.8281930333817126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004938784212473947,
      "loss": 5.1236,
      "mean_token_accuracy": 0.1988166391849518,
      "num_tokens": 18953187.0,
      "step": 9130
    },
    {
      "entropy": 5.570322036743164,
      "epoch": 0.8286465892597968,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004938709227752985,
      "loss": 5.2262,
      "mean_token_accuracy": 0.19010987728834153,
      "num_tokens": 18963015.0,
      "step": 9135
    },
    {
      "entropy": 5.494019985198975,
      "epoch": 0.829100145137881,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493863419776896,
      "loss": 5.2711,
      "mean_token_accuracy": 0.18064223527908324,
      "num_tokens": 18974184.0,
      "step": 9140
    },
    {
      "entropy": 5.450629234313965,
      "epoch": 0.8295537010159652,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938559122523421,
      "loss": 5.2186,
      "mean_token_accuracy": 0.1903683364391327,
      "num_tokens": 18985036.0,
      "step": 9145
    },
    {
      "entropy": 5.573251056671142,
      "epoch": 0.8300072568940493,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938484002017921,
      "loss": 5.3258,
      "mean_token_accuracy": 0.18096655309200288,
      "num_tokens": 18995237.0,
      "step": 9150
    },
    {
      "entropy": 5.539033031463623,
      "epoch": 0.8304608127721336,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004938408836254015,
      "loss": 5.3335,
      "mean_token_accuracy": 0.17874883860349655,
      "num_tokens": 19006272.0,
      "step": 9155
    },
    {
      "entropy": 5.459480285644531,
      "epoch": 0.8309143686502177,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938333625233258,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18578039556741716,
      "num_tokens": 19016728.0,
      "step": 9160
    },
    {
      "entropy": 5.545877456665039,
      "epoch": 0.8313679245283019,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004938258368957202,
      "loss": 5.2922,
      "mean_token_accuracy": 0.17642530798912048,
      "num_tokens": 19028024.0,
      "step": 9165
    },
    {
      "entropy": 5.562115573883057,
      "epoch": 0.831821480406386,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938183067427406,
      "loss": 5.2373,
      "mean_token_accuracy": 0.1918642923235893,
      "num_tokens": 19038218.0,
      "step": 9170
    },
    {
      "entropy": 5.454094696044922,
      "epoch": 0.8322750362844702,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004938107720645426,
      "loss": 5.2595,
      "mean_token_accuracy": 0.18231881111860276,
      "num_tokens": 19048574.0,
      "step": 9175
    },
    {
      "entropy": 5.495149421691894,
      "epoch": 0.8327285921625545,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004938032328612821,
      "loss": 5.2385,
      "mean_token_accuracy": 0.1852733686566353,
      "num_tokens": 19057721.0,
      "step": 9180
    },
    {
      "entropy": 5.491014194488526,
      "epoch": 0.8331821480406386,
      "grad_norm": 0.984375,
      "learning_rate": 0.000493795689133115,
      "loss": 5.2183,
      "mean_token_accuracy": 0.17995734810829161,
      "num_tokens": 19067582.0,
      "step": 9185
    },
    {
      "entropy": 5.468677616119384,
      "epoch": 0.8336357039187228,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004937881408801974,
      "loss": 5.1722,
      "mean_token_accuracy": 0.1918642297387123,
      "num_tokens": 19077896.0,
      "step": 9190
    },
    {
      "entropy": 5.4998781204223635,
      "epoch": 0.8340892597968069,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493780588102685,
      "loss": 5.2799,
      "mean_token_accuracy": 0.19266057163476943,
      "num_tokens": 19088082.0,
      "step": 9195
    },
    {
      "entropy": 5.577761936187744,
      "epoch": 0.8345428156748912,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004937730308007346,
      "loss": 5.2969,
      "mean_token_accuracy": 0.18470134884119033,
      "num_tokens": 19097176.0,
      "step": 9200
    },
    {
      "entropy": 5.499684047698975,
      "epoch": 0.8349963715529753,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004937654689745021,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19089347273111343,
      "num_tokens": 19107835.0,
      "step": 9205
    },
    {
      "entropy": 5.430692720413208,
      "epoch": 0.8354499274310595,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004937579026241439,
      "loss": 5.2017,
      "mean_token_accuracy": 0.1925992026925087,
      "num_tokens": 19118028.0,
      "step": 9210
    },
    {
      "entropy": 5.486698341369629,
      "epoch": 0.8359034833091437,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004937503317498165,
      "loss": 5.2276,
      "mean_token_accuracy": 0.19192424416542053,
      "num_tokens": 19129223.0,
      "step": 9215
    },
    {
      "entropy": 5.497008609771728,
      "epoch": 0.8363570391872278,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937427563516765,
      "loss": 5.3335,
      "mean_token_accuracy": 0.18186921924352645,
      "num_tokens": 19140048.0,
      "step": 9220
    },
    {
      "entropy": 5.568466663360596,
      "epoch": 0.8368105950653121,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004937351764298805,
      "loss": 5.2985,
      "mean_token_accuracy": 0.1766539365053177,
      "num_tokens": 19150853.0,
      "step": 9225
    },
    {
      "entropy": 5.551784181594849,
      "epoch": 0.8372641509433962,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004937275919845855,
      "loss": 5.2839,
      "mean_token_accuracy": 0.18331123888492584,
      "num_tokens": 19161645.0,
      "step": 9230
    },
    {
      "entropy": 5.518892812728882,
      "epoch": 0.8377177068214804,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004937200030159481,
      "loss": 5.3715,
      "mean_token_accuracy": 0.17670943588018417,
      "num_tokens": 19173276.0,
      "step": 9235
    },
    {
      "entropy": 5.517861652374267,
      "epoch": 0.8381712626995645,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004937124095241252,
      "loss": 5.248,
      "mean_token_accuracy": 0.19307708144187927,
      "num_tokens": 19182949.0,
      "step": 9240
    },
    {
      "entropy": 5.553634691238403,
      "epoch": 0.8386248185776488,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937048115092739,
      "loss": 5.2414,
      "mean_token_accuracy": 0.18353976160287858,
      "num_tokens": 19193504.0,
      "step": 9245
    },
    {
      "entropy": 5.5102275848388675,
      "epoch": 0.839078374455733,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004936972089715515,
      "loss": 5.3207,
      "mean_token_accuracy": 0.18096430003643035,
      "num_tokens": 19203406.0,
      "step": 9250
    },
    {
      "entropy": 5.4884428024292,
      "epoch": 0.8395319303338171,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493689601911115,
      "loss": 5.2644,
      "mean_token_accuracy": 0.18698281645774842,
      "num_tokens": 19214325.0,
      "step": 9255
    },
    {
      "entropy": 5.499520349502563,
      "epoch": 0.8399854862119013,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936819903281219,
      "loss": 5.3032,
      "mean_token_accuracy": 0.17399130165576934,
      "num_tokens": 19226466.0,
      "step": 9260
    },
    {
      "entropy": 5.601823616027832,
      "epoch": 0.8404390420899854,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936743742227295,
      "loss": 5.3281,
      "mean_token_accuracy": 0.1854452520608902,
      "num_tokens": 19237032.0,
      "step": 9265
    },
    {
      "entropy": 5.55287766456604,
      "epoch": 0.8408925979680697,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936667535950952,
      "loss": 5.3213,
      "mean_token_accuracy": 0.1790757581591606,
      "num_tokens": 19248009.0,
      "step": 9270
    },
    {
      "entropy": 5.548722505569458,
      "epoch": 0.8413461538461539,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004936591284453768,
      "loss": 5.372,
      "mean_token_accuracy": 0.1818295806646347,
      "num_tokens": 19259363.0,
      "step": 9275
    },
    {
      "entropy": 5.5150541305542,
      "epoch": 0.841799709724238,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004936514987737319,
      "loss": 5.2678,
      "mean_token_accuracy": 0.18624502122402192,
      "num_tokens": 19269191.0,
      "step": 9280
    },
    {
      "entropy": 5.576925992965698,
      "epoch": 0.8422532656023222,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936438645803182,
      "loss": 5.2344,
      "mean_token_accuracy": 0.18304483443498612,
      "num_tokens": 19280278.0,
      "step": 9285
    },
    {
      "entropy": 5.511117076873779,
      "epoch": 0.8427068214804064,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004936362258652938,
      "loss": 5.2203,
      "mean_token_accuracy": 0.19137237817049027,
      "num_tokens": 19290130.0,
      "step": 9290
    },
    {
      "entropy": 5.563754463195801,
      "epoch": 0.8431603773584906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004936285826288165,
      "loss": 5.3249,
      "mean_token_accuracy": 0.18893066346645354,
      "num_tokens": 19301669.0,
      "step": 9295
    },
    {
      "entropy": 5.46899380683899,
      "epoch": 0.8436139332365747,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004936209348710444,
      "loss": 5.2003,
      "mean_token_accuracy": 0.18793248534202575,
      "num_tokens": 19312490.0,
      "step": 9300
    },
    {
      "entropy": 5.502799987792969,
      "epoch": 0.8440674891146589,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004936132825921355,
      "loss": 5.2355,
      "mean_token_accuracy": 0.18560146540403366,
      "num_tokens": 19321743.0,
      "step": 9305
    },
    {
      "entropy": 5.462376737594605,
      "epoch": 0.844521044992743,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004936056257922484,
      "loss": 5.1537,
      "mean_token_accuracy": 0.18894000202417374,
      "num_tokens": 19332149.0,
      "step": 9310
    },
    {
      "entropy": 5.499111986160278,
      "epoch": 0.8449746008708273,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935979644715412,
      "loss": 5.259,
      "mean_token_accuracy": 0.1819048583507538,
      "num_tokens": 19341655.0,
      "step": 9315
    },
    {
      "entropy": 5.5421395778656,
      "epoch": 0.8454281567489115,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935902986301724,
      "loss": 5.2892,
      "mean_token_accuracy": 0.17833438664674758,
      "num_tokens": 19351404.0,
      "step": 9320
    },
    {
      "entropy": 5.5445465564727785,
      "epoch": 0.8458817126269956,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935826282683005,
      "loss": 5.3296,
      "mean_token_accuracy": 0.17884413450956343,
      "num_tokens": 19361788.0,
      "step": 9325
    },
    {
      "entropy": 5.52082576751709,
      "epoch": 0.8463352685050798,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004935749533860842,
      "loss": 5.2941,
      "mean_token_accuracy": 0.1879544198513031,
      "num_tokens": 19372902.0,
      "step": 9330
    },
    {
      "entropy": 5.456793355941772,
      "epoch": 0.846788824383164,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493567273983682,
      "loss": 5.164,
      "mean_token_accuracy": 0.18452823758125306,
      "num_tokens": 19382181.0,
      "step": 9335
    },
    {
      "entropy": 5.480916547775268,
      "epoch": 0.8472423802612482,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000493559590061253,
      "loss": 5.2346,
      "mean_token_accuracy": 0.18342400044202806,
      "num_tokens": 19393692.0,
      "step": 9340
    },
    {
      "entropy": 5.526205110549927,
      "epoch": 0.8476959361393324,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004935519016189559,
      "loss": 5.2629,
      "mean_token_accuracy": 0.19290686994791031,
      "num_tokens": 19404587.0,
      "step": 9345
    },
    {
      "entropy": 5.486667442321777,
      "epoch": 0.8481494920174165,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004935442086569498,
      "loss": 5.235,
      "mean_token_accuracy": 0.1877764046192169,
      "num_tokens": 19415296.0,
      "step": 9350
    },
    {
      "entropy": 5.559003782272339,
      "epoch": 0.8486030478955007,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004935365111753939,
      "loss": 5.3125,
      "mean_token_accuracy": 0.1837704747915268,
      "num_tokens": 19426119.0,
      "step": 9355
    },
    {
      "entropy": 5.567704677581787,
      "epoch": 0.8490566037735849,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935288091744472,
      "loss": 5.2818,
      "mean_token_accuracy": 0.17886171042919158,
      "num_tokens": 19436828.0,
      "step": 9360
    },
    {
      "entropy": 5.552606248855591,
      "epoch": 0.8495101596516691,
      "grad_norm": 1.015625,
      "learning_rate": 0.000493521102654269,
      "loss": 5.2655,
      "mean_token_accuracy": 0.18383633941411973,
      "num_tokens": 19446583.0,
      "step": 9365
    },
    {
      "entropy": 5.441853189468384,
      "epoch": 0.8499637155297532,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004935133916150187,
      "loss": 5.1426,
      "mean_token_accuracy": 0.19354938864707946,
      "num_tokens": 19456757.0,
      "step": 9370
    },
    {
      "entropy": 5.4375901222229,
      "epoch": 0.8504172714078374,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004935056760568558,
      "loss": 5.2203,
      "mean_token_accuracy": 0.18387676924467086,
      "num_tokens": 19467624.0,
      "step": 9375
    },
    {
      "entropy": 5.507636260986328,
      "epoch": 0.8508708272859217,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934979559799399,
      "loss": 5.1784,
      "mean_token_accuracy": 0.19261115789413452,
      "num_tokens": 19477759.0,
      "step": 9380
    },
    {
      "entropy": 5.50710015296936,
      "epoch": 0.8513243831640058,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934902313844306,
      "loss": 5.1971,
      "mean_token_accuracy": 0.1891529604792595,
      "num_tokens": 19488877.0,
      "step": 9385
    },
    {
      "entropy": 5.452629899978637,
      "epoch": 0.85177793904209,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004934825022704875,
      "loss": 5.1932,
      "mean_token_accuracy": 0.18864367604255677,
      "num_tokens": 19499170.0,
      "step": 9390
    },
    {
      "entropy": 5.495832824707032,
      "epoch": 0.8522314949201741,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004934747686382707,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1803721621632576,
      "num_tokens": 19509944.0,
      "step": 9395
    },
    {
      "entropy": 5.506520843505859,
      "epoch": 0.8526850507982583,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934670304879398,
      "loss": 5.1906,
      "mean_token_accuracy": 0.18765731602907182,
      "num_tokens": 19519014.0,
      "step": 9400
    },
    {
      "entropy": 5.4330939769744875,
      "epoch": 0.8531386066763426,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934592878196553,
      "loss": 5.2022,
      "mean_token_accuracy": 0.18718813359737396,
      "num_tokens": 19528586.0,
      "step": 9405
    },
    {
      "entropy": 5.527695608139038,
      "epoch": 0.8535921625544267,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493451540633577,
      "loss": 5.3372,
      "mean_token_accuracy": 0.18059038519859313,
      "num_tokens": 19539912.0,
      "step": 9410
    },
    {
      "entropy": 5.507913064956665,
      "epoch": 0.8540457184325109,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004934437889298651,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1787128046154976,
      "num_tokens": 19549860.0,
      "step": 9415
    },
    {
      "entropy": 5.537213087081909,
      "epoch": 0.854499274310595,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004934360327086799,
      "loss": 5.2581,
      "mean_token_accuracy": 0.1831861764192581,
      "num_tokens": 19559446.0,
      "step": 9420
    },
    {
      "entropy": 5.57012186050415,
      "epoch": 0.8549528301886793,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493428271970182,
      "loss": 5.2586,
      "mean_token_accuracy": 0.19021736234426498,
      "num_tokens": 19570282.0,
      "step": 9425
    },
    {
      "entropy": 5.526591825485229,
      "epoch": 0.8554063860667634,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934205067145318,
      "loss": 5.1816,
      "mean_token_accuracy": 0.18626246899366378,
      "num_tokens": 19581087.0,
      "step": 9430
    },
    {
      "entropy": 5.495096015930176,
      "epoch": 0.8558599419448476,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004934127369418898,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19055798500776291,
      "num_tokens": 19592100.0,
      "step": 9435
    },
    {
      "entropy": 5.49104151725769,
      "epoch": 0.8563134978229318,
      "grad_norm": 1.0,
      "learning_rate": 0.0004934049626524168,
      "loss": 5.2558,
      "mean_token_accuracy": 0.18784750998020172,
      "num_tokens": 19602557.0,
      "step": 9440
    },
    {
      "entropy": 5.539940404891968,
      "epoch": 0.8567670537010159,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933971838462734,
      "loss": 5.2965,
      "mean_token_accuracy": 0.17805867195129393,
      "num_tokens": 19613643.0,
      "step": 9445
    },
    {
      "entropy": 5.516980266571045,
      "epoch": 0.8572206095791002,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933894005236207,
      "loss": 5.212,
      "mean_token_accuracy": 0.1881819874048233,
      "num_tokens": 19624180.0,
      "step": 9450
    },
    {
      "entropy": 5.535392236709595,
      "epoch": 0.8576741654571843,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933816126846195,
      "loss": 5.2281,
      "mean_token_accuracy": 0.18857050836086273,
      "num_tokens": 19634799.0,
      "step": 9455
    },
    {
      "entropy": 5.4610528469085695,
      "epoch": 0.8581277213352685,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933738203294309,
      "loss": 5.2042,
      "mean_token_accuracy": 0.1919635698199272,
      "num_tokens": 19644618.0,
      "step": 9460
    },
    {
      "entropy": 5.458015203475952,
      "epoch": 0.8585812772133526,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000493366023458216,
      "loss": 5.2114,
      "mean_token_accuracy": 0.18096943944692612,
      "num_tokens": 19655550.0,
      "step": 9465
    },
    {
      "entropy": 5.538266515731811,
      "epoch": 0.8590348330914369,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933582220711362,
      "loss": 5.2769,
      "mean_token_accuracy": 0.19205545336008073,
      "num_tokens": 19665734.0,
      "step": 9470
    },
    {
      "entropy": 5.527783441543579,
      "epoch": 0.8594883889695211,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933504161683526,
      "loss": 5.2234,
      "mean_token_accuracy": 0.1832062631845474,
      "num_tokens": 19677130.0,
      "step": 9475
    },
    {
      "entropy": 5.452576208114624,
      "epoch": 0.8599419448476052,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004933426057500269,
      "loss": 5.1462,
      "mean_token_accuracy": 0.19649119079113006,
      "num_tokens": 19687374.0,
      "step": 9480
    },
    {
      "entropy": 5.5538092136383055,
      "epoch": 0.8603955007256894,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004933347908163203,
      "loss": 5.2585,
      "mean_token_accuracy": 0.1857781171798706,
      "num_tokens": 19698182.0,
      "step": 9485
    },
    {
      "entropy": 5.515981197357178,
      "epoch": 0.8608490566037735,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004933269713673947,
      "loss": 5.2181,
      "mean_token_accuracy": 0.18882352858781815,
      "num_tokens": 19707664.0,
      "step": 9490
    },
    {
      "entropy": 5.435214948654175,
      "epoch": 0.8613026124818578,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004933191474034116,
      "loss": 5.2636,
      "mean_token_accuracy": 0.18053202778100969,
      "num_tokens": 19717647.0,
      "step": 9495
    },
    {
      "entropy": 5.499503469467163,
      "epoch": 0.861756168359942,
      "grad_norm": 0.9375,
      "learning_rate": 0.000493311318924533,
      "loss": 5.2917,
      "mean_token_accuracy": 0.18424864560365678,
      "num_tokens": 19727444.0,
      "step": 9500
    },
    {
      "entropy": 5.5348368167877195,
      "epoch": 0.8622097242380261,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933034859309208,
      "loss": 5.3074,
      "mean_token_accuracy": 0.17919572144746782,
      "num_tokens": 19738719.0,
      "step": 9505
    },
    {
      "entropy": 5.5768194675445555,
      "epoch": 0.8626632801161103,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004932956484227367,
      "loss": 5.2775,
      "mean_token_accuracy": 0.19209030717611314,
      "num_tokens": 19749786.0,
      "step": 9510
    },
    {
      "entropy": 5.507620334625244,
      "epoch": 0.8631168359941945,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493287806400143,
      "loss": 5.2268,
      "mean_token_accuracy": 0.19589438587427138,
      "num_tokens": 19759623.0,
      "step": 9515
    },
    {
      "entropy": 5.516129493713379,
      "epoch": 0.8635703918722787,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004932799598633018,
      "loss": 5.2349,
      "mean_token_accuracy": 0.1868974268436432,
      "num_tokens": 19770117.0,
      "step": 9520
    },
    {
      "entropy": 5.536385726928711,
      "epoch": 0.8640239477503628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932721088123753,
      "loss": 5.3231,
      "mean_token_accuracy": 0.18734574615955352,
      "num_tokens": 19780748.0,
      "step": 9525
    },
    {
      "entropy": 5.513698625564575,
      "epoch": 0.864477503628447,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004932642532475262,
      "loss": 5.2632,
      "mean_token_accuracy": 0.18844823986291886,
      "num_tokens": 19791816.0,
      "step": 9530
    },
    {
      "entropy": 5.5973162174224855,
      "epoch": 0.8649310595065312,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004932563931689168,
      "loss": 5.3348,
      "mean_token_accuracy": 0.18349990099668503,
      "num_tokens": 19801096.0,
      "step": 9535
    },
    {
      "entropy": 5.519701194763184,
      "epoch": 0.8653846153846154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932485285767093,
      "loss": 5.1903,
      "mean_token_accuracy": 0.1904284492135048,
      "num_tokens": 19811258.0,
      "step": 9540
    },
    {
      "entropy": 5.541377305984497,
      "epoch": 0.8658381712626996,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932406594710667,
      "loss": 5.2816,
      "mean_token_accuracy": 0.19132487922906877,
      "num_tokens": 19821765.0,
      "step": 9545
    },
    {
      "entropy": 5.5216514587402346,
      "epoch": 0.8662917271407837,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004932327858521517,
      "loss": 5.1445,
      "mean_token_accuracy": 0.19365739524364473,
      "num_tokens": 19831537.0,
      "step": 9550
    },
    {
      "entropy": 5.463690662384034,
      "epoch": 0.8667452830188679,
      "grad_norm": 0.96875,
      "learning_rate": 0.000493224907720127,
      "loss": 5.1567,
      "mean_token_accuracy": 0.18910293579101561,
      "num_tokens": 19841830.0,
      "step": 9555
    },
    {
      "entropy": 5.486150312423706,
      "epoch": 0.8671988388969522,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932170250751556,
      "loss": 5.2631,
      "mean_token_accuracy": 0.18096060454845428,
      "num_tokens": 19851896.0,
      "step": 9560
    },
    {
      "entropy": 5.550594329833984,
      "epoch": 0.8676523947750363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932091379174006,
      "loss": 5.298,
      "mean_token_accuracy": 0.17869172990322113,
      "num_tokens": 19861735.0,
      "step": 9565
    },
    {
      "entropy": 5.523482179641723,
      "epoch": 0.8681059506531205,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493201246247025,
      "loss": 5.1781,
      "mean_token_accuracy": 0.19788408875465394,
      "num_tokens": 19872333.0,
      "step": 9570
    },
    {
      "entropy": 5.48471360206604,
      "epoch": 0.8685595065312046,
      "grad_norm": 0.890625,
      "learning_rate": 0.000493193350064192,
      "loss": 5.1956,
      "mean_token_accuracy": 0.1868165984749794,
      "num_tokens": 19882896.0,
      "step": 9575
    },
    {
      "entropy": 5.470310354232788,
      "epoch": 0.8690130624092888,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004931854493690649,
      "loss": 5.1672,
      "mean_token_accuracy": 0.18842218518257142,
      "num_tokens": 19893767.0,
      "step": 9580
    },
    {
      "entropy": 5.4305318832397464,
      "epoch": 0.869466618287373,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004931775441618071,
      "loss": 5.1424,
      "mean_token_accuracy": 0.19409455507993698,
      "num_tokens": 19904242.0,
      "step": 9585
    },
    {
      "entropy": 5.474945592880249,
      "epoch": 0.8699201741654572,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004931696344425821,
      "loss": 5.2879,
      "mean_token_accuracy": 0.18741385638713837,
      "num_tokens": 19913898.0,
      "step": 9590
    },
    {
      "entropy": 5.589535808563232,
      "epoch": 0.8703737300435413,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004931617202115535,
      "loss": 5.2727,
      "mean_token_accuracy": 0.18905438631772994,
      "num_tokens": 19923756.0,
      "step": 9595
    },
    {
      "entropy": 5.507710361480713,
      "epoch": 0.8708272859216255,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004931538014688849,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19067411869764328,
      "num_tokens": 19932720.0,
      "step": 9600
    },
    {
      "entropy": 5.565124797821045,
      "epoch": 0.8712808417997098,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004931458782147401,
      "loss": 5.4096,
      "mean_token_accuracy": 0.17565523087978363,
      "num_tokens": 19942962.0,
      "step": 9605
    },
    {
      "entropy": 5.584523773193359,
      "epoch": 0.8717343976777939,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493137950449283,
      "loss": 5.2822,
      "mean_token_accuracy": 0.19391026049852372,
      "num_tokens": 19953735.0,
      "step": 9610
    },
    {
      "entropy": 5.485487079620361,
      "epoch": 0.8721879535558781,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004931300181726775,
      "loss": 5.2473,
      "mean_token_accuracy": 0.18759722262620926,
      "num_tokens": 19964416.0,
      "step": 9615
    },
    {
      "entropy": 5.493893098831177,
      "epoch": 0.8726415094339622,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004931220813850875,
      "loss": 5.2659,
      "mean_token_accuracy": 0.18439658731222153,
      "num_tokens": 19974531.0,
      "step": 9620
    },
    {
      "entropy": 5.551490545272827,
      "epoch": 0.8730950653120464,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004931141400866773,
      "loss": 5.2484,
      "mean_token_accuracy": 0.18889303654432296,
      "num_tokens": 19984964.0,
      "step": 9625
    },
    {
      "entropy": 5.4911480903625485,
      "epoch": 0.8735486211901307,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931061942776112,
      "loss": 5.186,
      "mean_token_accuracy": 0.19389327317476274,
      "num_tokens": 19995568.0,
      "step": 9630
    },
    {
      "entropy": 5.588620948791504,
      "epoch": 0.8740021770682148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930982439580534,
      "loss": 5.3836,
      "mean_token_accuracy": 0.17700846791267394,
      "num_tokens": 20005856.0,
      "step": 9635
    },
    {
      "entropy": 5.57611494064331,
      "epoch": 0.874455732946299,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004930902891281683,
      "loss": 5.283,
      "mean_token_accuracy": 0.18396832197904586,
      "num_tokens": 20015683.0,
      "step": 9640
    },
    {
      "entropy": 5.41850471496582,
      "epoch": 0.8749092888243831,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004930823297881205,
      "loss": 5.1013,
      "mean_token_accuracy": 0.19724885672330855,
      "num_tokens": 20025998.0,
      "step": 9645
    },
    {
      "entropy": 5.539435148239136,
      "epoch": 0.8753628447024674,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004930743659380745,
      "loss": 5.2599,
      "mean_token_accuracy": 0.18563385009765626,
      "num_tokens": 20036486.0,
      "step": 9650
    },
    {
      "entropy": 5.479112577438355,
      "epoch": 0.8758164005805515,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493066397578195,
      "loss": 5.1674,
      "mean_token_accuracy": 0.19481005817651748,
      "num_tokens": 20046784.0,
      "step": 9655
    },
    {
      "entropy": 5.480585336685181,
      "epoch": 0.8762699564586357,
      "grad_norm": 1.0,
      "learning_rate": 0.0004930584247086468,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1877565011382103,
      "num_tokens": 20056999.0,
      "step": 9660
    },
    {
      "entropy": 5.465516996383667,
      "epoch": 0.8767235123367199,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000493050447329595,
      "loss": 5.122,
      "mean_token_accuracy": 0.19252620935440062,
      "num_tokens": 20067612.0,
      "step": 9665
    },
    {
      "entropy": 5.5493165969848635,
      "epoch": 0.877177068214804,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004930424654412043,
      "loss": 5.275,
      "mean_token_accuracy": 0.1878029227256775,
      "num_tokens": 20077330.0,
      "step": 9670
    },
    {
      "entropy": 5.546081256866455,
      "epoch": 0.8776306240928883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004930344790436398,
      "loss": 5.2745,
      "mean_token_accuracy": 0.17999620139598846,
      "num_tokens": 20086801.0,
      "step": 9675
    },
    {
      "entropy": 5.52485613822937,
      "epoch": 0.8780841799709724,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930264881370666,
      "loss": 5.2166,
      "mean_token_accuracy": 0.18604013770818711,
      "num_tokens": 20097047.0,
      "step": 9680
    },
    {
      "entropy": 5.471764802932739,
      "epoch": 0.8785377358490566,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004930184927216504,
      "loss": 5.2546,
      "mean_token_accuracy": 0.18681764751672744,
      "num_tokens": 20107185.0,
      "step": 9685
    },
    {
      "entropy": 5.53357629776001,
      "epoch": 0.8789912917271407,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004930104927975559,
      "loss": 5.1947,
      "mean_token_accuracy": 0.19048551768064498,
      "num_tokens": 20118493.0,
      "step": 9690
    },
    {
      "entropy": 5.484797573089599,
      "epoch": 0.879444847605225,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004930024883649489,
      "loss": 5.2524,
      "mean_token_accuracy": 0.18750473260879516,
      "num_tokens": 20129294.0,
      "step": 9695
    },
    {
      "entropy": 5.442465543746948,
      "epoch": 0.8798984034833092,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492994479423995,
      "loss": 5.2124,
      "mean_token_accuracy": 0.19394612908363343,
      "num_tokens": 20139194.0,
      "step": 9700
    },
    {
      "entropy": 5.467197561264038,
      "epoch": 0.8803519593613933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004929864659748597,
      "loss": 5.2374,
      "mean_token_accuracy": 0.18322389721870422,
      "num_tokens": 20149681.0,
      "step": 9705
    },
    {
      "entropy": 5.549167013168335,
      "epoch": 0.8808055152394775,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004929784480177087,
      "loss": 5.3222,
      "mean_token_accuracy": 0.18286841213703156,
      "num_tokens": 20160557.0,
      "step": 9710
    },
    {
      "entropy": 5.519404029846191,
      "epoch": 0.8812590711175616,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004929704255527078,
      "loss": 5.1887,
      "mean_token_accuracy": 0.187822188436985,
      "num_tokens": 20171047.0,
      "step": 9715
    },
    {
      "entropy": 5.523070621490478,
      "epoch": 0.8817126269956459,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000492962398580023,
      "loss": 5.3707,
      "mean_token_accuracy": 0.17758570462465287,
      "num_tokens": 20182148.0,
      "step": 9720
    },
    {
      "entropy": 5.421023654937744,
      "epoch": 0.88216618287373,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004929543670998203,
      "loss": 5.1276,
      "mean_token_accuracy": 0.19135259091854095,
      "num_tokens": 20192131.0,
      "step": 9725
    },
    {
      "entropy": 5.531173181533814,
      "epoch": 0.8826197387518142,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004929463311122657,
      "loss": 5.2426,
      "mean_token_accuracy": 0.1861494779586792,
      "num_tokens": 20202805.0,
      "step": 9730
    },
    {
      "entropy": 5.539679002761841,
      "epoch": 0.8830732946298984,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004929382906175253,
      "loss": 5.1717,
      "mean_token_accuracy": 0.19339725226163865,
      "num_tokens": 20212702.0,
      "step": 9735
    },
    {
      "entropy": 5.504897594451904,
      "epoch": 0.8835268505079826,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004929302456157657,
      "loss": 5.2068,
      "mean_token_accuracy": 0.19339656680822373,
      "num_tokens": 20222661.0,
      "step": 9740
    },
    {
      "entropy": 5.490061807632446,
      "epoch": 0.8839804063860668,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004929221961071531,
      "loss": 5.2559,
      "mean_token_accuracy": 0.18547530472278595,
      "num_tokens": 20233012.0,
      "step": 9745
    },
    {
      "entropy": 5.521569061279297,
      "epoch": 0.8844339622641509,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929141420918538,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19188135862350464,
      "num_tokens": 20242918.0,
      "step": 9750
    },
    {
      "entropy": 5.4918468475341795,
      "epoch": 0.8848875181422351,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929060835700347,
      "loss": 5.191,
      "mean_token_accuracy": 0.19425636231899263,
      "num_tokens": 20253254.0,
      "step": 9755
    },
    {
      "entropy": 5.500729370117187,
      "epoch": 0.8853410740203193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492898020541862,
      "loss": 5.2469,
      "mean_token_accuracy": 0.18291963785886764,
      "num_tokens": 20263355.0,
      "step": 9760
    },
    {
      "entropy": 5.450434637069702,
      "epoch": 0.8857946298984035,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928899530075029,
      "loss": 5.1715,
      "mean_token_accuracy": 0.1866050571203232,
      "num_tokens": 20274716.0,
      "step": 9765
    },
    {
      "entropy": 5.4292707443237305,
      "epoch": 0.8862481857764877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928818809671241,
      "loss": 5.1607,
      "mean_token_accuracy": 0.18965840190649033,
      "num_tokens": 20285631.0,
      "step": 9770
    },
    {
      "entropy": 5.558332729339599,
      "epoch": 0.8867017416545718,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928738044208923,
      "loss": 5.2411,
      "mean_token_accuracy": 0.1915654197335243,
      "num_tokens": 20295717.0,
      "step": 9775
    },
    {
      "entropy": 5.459675216674805,
      "epoch": 0.887155297532656,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004928657233689748,
      "loss": 5.2183,
      "mean_token_accuracy": 0.18936942964792253,
      "num_tokens": 20306637.0,
      "step": 9780
    },
    {
      "entropy": 5.490768623352051,
      "epoch": 0.8876088534107403,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004928576378115387,
      "loss": 5.1846,
      "mean_token_accuracy": 0.1875806614756584,
      "num_tokens": 20316629.0,
      "step": 9785
    },
    {
      "entropy": 5.597564649581909,
      "epoch": 0.8880624092888244,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004928495477487509,
      "loss": 5.3573,
      "mean_token_accuracy": 0.17805679738521576,
      "num_tokens": 20327935.0,
      "step": 9790
    },
    {
      "entropy": 5.491835212707519,
      "epoch": 0.8885159651669086,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004928414531807792,
      "loss": 5.1762,
      "mean_token_accuracy": 0.1839013561606407,
      "num_tokens": 20339142.0,
      "step": 9795
    },
    {
      "entropy": 5.464007425308227,
      "epoch": 0.8889695210449927,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004928333541077906,
      "loss": 5.1896,
      "mean_token_accuracy": 0.20151338577270508,
      "num_tokens": 20348474.0,
      "step": 9800
    },
    {
      "entropy": 5.500573539733887,
      "epoch": 0.8894230769230769,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004928252505299527,
      "loss": 5.2498,
      "mean_token_accuracy": 0.1855317085981369,
      "num_tokens": 20359440.0,
      "step": 9805
    },
    {
      "entropy": 5.579951906204224,
      "epoch": 0.8898766328011611,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004928171424474332,
      "loss": 5.2963,
      "mean_token_accuracy": 0.18381822854280472,
      "num_tokens": 20370355.0,
      "step": 9810
    },
    {
      "entropy": 5.492842674255371,
      "epoch": 0.8903301886792453,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004928090298603996,
      "loss": 5.1795,
      "mean_token_accuracy": 0.18734454959630967,
      "num_tokens": 20380214.0,
      "step": 9815
    },
    {
      "entropy": 5.527317905426026,
      "epoch": 0.8907837445573294,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004928009127690199,
      "loss": 5.1963,
      "mean_token_accuracy": 0.18887149095535277,
      "num_tokens": 20389190.0,
      "step": 9820
    },
    {
      "entropy": 5.4723756313323975,
      "epoch": 0.8912373004354136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004927927911734617,
      "loss": 5.238,
      "mean_token_accuracy": 0.18437635898590088,
      "num_tokens": 20399531.0,
      "step": 9825
    },
    {
      "entropy": 5.463613748550415,
      "epoch": 0.8916908563134979,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004927846650738931,
      "loss": 5.2246,
      "mean_token_accuracy": 0.18858011811971664,
      "num_tokens": 20410201.0,
      "step": 9830
    },
    {
      "entropy": 5.494712591171265,
      "epoch": 0.892144412191582,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004927765344704821,
      "loss": 5.1865,
      "mean_token_accuracy": 0.19300392419099807,
      "num_tokens": 20421579.0,
      "step": 9835
    },
    {
      "entropy": 5.510917329788208,
      "epoch": 0.8925979680696662,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004927683993633969,
      "loss": 5.2522,
      "mean_token_accuracy": 0.18945046067237853,
      "num_tokens": 20432048.0,
      "step": 9840
    },
    {
      "entropy": 5.559658622741699,
      "epoch": 0.8930515239477503,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004927602597528057,
      "loss": 5.2763,
      "mean_token_accuracy": 0.18216387331485748,
      "num_tokens": 20441752.0,
      "step": 9845
    },
    {
      "entropy": 5.437015390396118,
      "epoch": 0.8935050798258345,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004927521156388769,
      "loss": 5.1443,
      "mean_token_accuracy": 0.1886148139834404,
      "num_tokens": 20451693.0,
      "step": 9850
    },
    {
      "entropy": 5.512848711013794,
      "epoch": 0.8939586357039188,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004927439670217788,
      "loss": 5.2784,
      "mean_token_accuracy": 0.18574748635292054,
      "num_tokens": 20462312.0,
      "step": 9855
    },
    {
      "entropy": 5.51384711265564,
      "epoch": 0.8944121915820029,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049273581390168,
      "loss": 5.1842,
      "mean_token_accuracy": 0.18325104117393493,
      "num_tokens": 20472546.0,
      "step": 9860
    },
    {
      "entropy": 5.482762384414673,
      "epoch": 0.8948657474600871,
      "grad_norm": 0.875,
      "learning_rate": 0.000492727656278749,
      "loss": 5.2265,
      "mean_token_accuracy": 0.189279143512249,
      "num_tokens": 20483266.0,
      "step": 9865
    },
    {
      "entropy": 5.43660979270935,
      "epoch": 0.8953193033381712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004927194941531547,
      "loss": 5.1597,
      "mean_token_accuracy": 0.18750689327716827,
      "num_tokens": 20492490.0,
      "step": 9870
    },
    {
      "entropy": 5.482614469528198,
      "epoch": 0.8957728592162555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004927113275250657,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19175224751234055,
      "num_tokens": 20503077.0,
      "step": 9875
    },
    {
      "entropy": 5.448323917388916,
      "epoch": 0.8962264150943396,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004927031563946511,
      "loss": 5.2231,
      "mean_token_accuracy": 0.18685709685087204,
      "num_tokens": 20513771.0,
      "step": 9880
    },
    {
      "entropy": 5.513888883590698,
      "epoch": 0.8966799709724238,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004926949807620796,
      "loss": 5.3107,
      "mean_token_accuracy": 0.1844270795583725,
      "num_tokens": 20523616.0,
      "step": 9885
    },
    {
      "entropy": 5.525784826278686,
      "epoch": 0.897133526850508,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004926868006275204,
      "loss": 5.0908,
      "mean_token_accuracy": 0.19574417024850846,
      "num_tokens": 20533613.0,
      "step": 9890
    },
    {
      "entropy": 5.473994970321655,
      "epoch": 0.8975870827285921,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926786159911427,
      "loss": 5.2053,
      "mean_token_accuracy": 0.18760427534580232,
      "num_tokens": 20543812.0,
      "step": 9895
    },
    {
      "entropy": 5.449514675140381,
      "epoch": 0.8980406386066764,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492670426853116,
      "loss": 5.1502,
      "mean_token_accuracy": 0.19117735475301742,
      "num_tokens": 20553785.0,
      "step": 9900
    },
    {
      "entropy": 5.578982305526734,
      "epoch": 0.8984941944847605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926622332136092,
      "loss": 5.2671,
      "mean_token_accuracy": 0.1900632604956627,
      "num_tokens": 20563326.0,
      "step": 9905
    },
    {
      "entropy": 5.484088993072509,
      "epoch": 0.8989477503628447,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004926540350727918,
      "loss": 5.1148,
      "mean_token_accuracy": 0.19918729811906816,
      "num_tokens": 20573643.0,
      "step": 9910
    },
    {
      "entropy": 5.387427949905396,
      "epoch": 0.8994013062409288,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926458324308337,
      "loss": 5.1496,
      "mean_token_accuracy": 0.1918700322508812,
      "num_tokens": 20583916.0,
      "step": 9915
    },
    {
      "entropy": 5.506546497344971,
      "epoch": 0.8998548621190131,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004926376252879043,
      "loss": 5.3024,
      "mean_token_accuracy": 0.18872980624437333,
      "num_tokens": 20593984.0,
      "step": 9920
    },
    {
      "entropy": 5.531512498855591,
      "epoch": 0.9003084179970973,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004926294136441733,
      "loss": 5.141,
      "mean_token_accuracy": 0.19161704182624817,
      "num_tokens": 20603927.0,
      "step": 9925
    },
    {
      "entropy": 5.491614151000976,
      "epoch": 0.9007619738751814,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926211974998106,
      "loss": 5.2693,
      "mean_token_accuracy": 0.18857511729002,
      "num_tokens": 20614147.0,
      "step": 9930
    },
    {
      "entropy": 5.524622821807862,
      "epoch": 0.9012155297532656,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492612976854986,
      "loss": 5.2403,
      "mean_token_accuracy": 0.18621373921632767,
      "num_tokens": 20623631.0,
      "step": 9935
    },
    {
      "entropy": 5.438816690444947,
      "epoch": 0.9016690856313497,
      "grad_norm": 1.0,
      "learning_rate": 0.0004926047517098697,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19730497598648072,
      "num_tokens": 20632566.0,
      "step": 9940
    },
    {
      "entropy": 5.500843334197998,
      "epoch": 0.902122641509434,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004925965220646315,
      "loss": 5.233,
      "mean_token_accuracy": 0.18363679200410843,
      "num_tokens": 20643246.0,
      "step": 9945
    },
    {
      "entropy": 5.507595014572144,
      "epoch": 0.9025761973875182,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925882879194419,
      "loss": 5.296,
      "mean_token_accuracy": 0.18696873188018798,
      "num_tokens": 20653372.0,
      "step": 9950
    },
    {
      "entropy": 5.515101146697998,
      "epoch": 0.9030297532656023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004925800492744712,
      "loss": 5.2455,
      "mean_token_accuracy": 0.18492357879877092,
      "num_tokens": 20662687.0,
      "step": 9955
    },
    {
      "entropy": 5.465108442306518,
      "epoch": 0.9034833091436865,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004925718061298895,
      "loss": 5.1691,
      "mean_token_accuracy": 0.19537947028875352,
      "num_tokens": 20673168.0,
      "step": 9960
    },
    {
      "entropy": 5.555444622039795,
      "epoch": 0.9039368650217707,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004925635584858675,
      "loss": 5.2785,
      "mean_token_accuracy": 0.18311350643634797,
      "num_tokens": 20683100.0,
      "step": 9965
    },
    {
      "entropy": 5.589123487472534,
      "epoch": 0.9043904208998549,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004925553063425755,
      "loss": 5.2853,
      "mean_token_accuracy": 0.18646423369646073,
      "num_tokens": 20694097.0,
      "step": 9970
    },
    {
      "entropy": 5.501810359954834,
      "epoch": 0.904843976777939,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925470497001844,
      "loss": 5.2467,
      "mean_token_accuracy": 0.18978470414876938,
      "num_tokens": 20704640.0,
      "step": 9975
    },
    {
      "entropy": 5.486598873138428,
      "epoch": 0.9052975326560232,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925387885588648,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19009471386671067,
      "num_tokens": 20714419.0,
      "step": 9980
    },
    {
      "entropy": 5.574535608291626,
      "epoch": 0.9057510885341074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925305229187877,
      "loss": 5.2849,
      "mean_token_accuracy": 0.1879521057009697,
      "num_tokens": 20724705.0,
      "step": 9985
    },
    {
      "entropy": 5.561985349655151,
      "epoch": 0.9062046444121916,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492522252780124,
      "loss": 5.3439,
      "mean_token_accuracy": 0.18506076335906982,
      "num_tokens": 20734753.0,
      "step": 9990
    },
    {
      "entropy": 5.5091822147369385,
      "epoch": 0.9066582002902758,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004925139781430446,
      "loss": 5.1525,
      "mean_token_accuracy": 0.1960643097758293,
      "num_tokens": 20745389.0,
      "step": 9995
    },
    {
      "entropy": 5.480530834197998,
      "epoch": 0.9071117561683599,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004925056990077207,
      "loss": 5.2167,
      "mean_token_accuracy": 0.19059233814477922,
      "num_tokens": 20755702.0,
      "step": 10000
    },
    {
      "entropy": 5.462615489959717,
      "epoch": 0.9075653120464441,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924974153743235,
      "loss": 5.1057,
      "mean_token_accuracy": 0.1962900161743164,
      "num_tokens": 20766046.0,
      "step": 10005
    },
    {
      "entropy": 5.492237567901611,
      "epoch": 0.9080188679245284,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004924891272430244,
      "loss": 5.2233,
      "mean_token_accuracy": 0.19067611545324326,
      "num_tokens": 20776654.0,
      "step": 10010
    },
    {
      "entropy": 5.538366174697876,
      "epoch": 0.9084724238026125,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004924808346139948,
      "loss": 5.2699,
      "mean_token_accuracy": 0.19199298322200775,
      "num_tokens": 20787066.0,
      "step": 10015
    },
    {
      "entropy": 5.483625793457032,
      "epoch": 0.9089259796806967,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004924725374874059,
      "loss": 5.1807,
      "mean_token_accuracy": 0.18984902501106263,
      "num_tokens": 20796822.0,
      "step": 10020
    },
    {
      "entropy": 5.530815505981446,
      "epoch": 0.9093795355587808,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004924642358634297,
      "loss": 5.2346,
      "mean_token_accuracy": 0.18641554266214372,
      "num_tokens": 20807602.0,
      "step": 10025
    },
    {
      "entropy": 5.51546082496643,
      "epoch": 0.909833091436865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924559297422376,
      "loss": 5.193,
      "mean_token_accuracy": 0.19486915022134782,
      "num_tokens": 20819046.0,
      "step": 10030
    },
    {
      "entropy": 5.520950222015381,
      "epoch": 0.9102866473149492,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004924476191240014,
      "loss": 5.2297,
      "mean_token_accuracy": 0.18605352193117142,
      "num_tokens": 20828989.0,
      "step": 10035
    },
    {
      "entropy": 5.47696099281311,
      "epoch": 0.9107402031930334,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492439304008893,
      "loss": 5.2303,
      "mean_token_accuracy": 0.18616906255483628,
      "num_tokens": 20839366.0,
      "step": 10040
    },
    {
      "entropy": 5.507669305801391,
      "epoch": 0.9111937590711175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004924309843970847,
      "loss": 5.2253,
      "mean_token_accuracy": 0.18729967921972274,
      "num_tokens": 20849513.0,
      "step": 10045
    },
    {
      "entropy": 5.5649864196777346,
      "epoch": 0.9116473149492017,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004924226602887479,
      "loss": 5.2685,
      "mean_token_accuracy": 0.18344479352235793,
      "num_tokens": 20861406.0,
      "step": 10050
    },
    {
      "entropy": 5.495461368560791,
      "epoch": 0.912100870827286,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924143316840552,
      "loss": 5.1824,
      "mean_token_accuracy": 0.19835592210292816,
      "num_tokens": 20871190.0,
      "step": 10055
    },
    {
      "entropy": 5.505907154083252,
      "epoch": 0.9125544267053701,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924059985831787,
      "loss": 5.3313,
      "mean_token_accuracy": 0.18035977929830552,
      "num_tokens": 20881589.0,
      "step": 10060
    },
    {
      "entropy": 5.508808231353759,
      "epoch": 0.9130079825834543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923976609862908,
      "loss": 5.26,
      "mean_token_accuracy": 0.1787053018808365,
      "num_tokens": 20892072.0,
      "step": 10065
    },
    {
      "entropy": 5.547299718856811,
      "epoch": 0.9134615384615384,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923893188935638,
      "loss": 5.2435,
      "mean_token_accuracy": 0.18379553705453872,
      "num_tokens": 20902616.0,
      "step": 10070
    },
    {
      "entropy": 5.545641326904297,
      "epoch": 0.9139150943396226,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004923809723051702,
      "loss": 5.2337,
      "mean_token_accuracy": 0.18797029107809066,
      "num_tokens": 20913832.0,
      "step": 10075
    },
    {
      "entropy": 5.4296571731567385,
      "epoch": 0.9143686502177069,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923726212212829,
      "loss": 5.1311,
      "mean_token_accuracy": 0.1946079269051552,
      "num_tokens": 20923322.0,
      "step": 10080
    },
    {
      "entropy": 5.484694910049439,
      "epoch": 0.914822206095791,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923642656420742,
      "loss": 5.2231,
      "mean_token_accuracy": 0.18851592838764192,
      "num_tokens": 20933334.0,
      "step": 10085
    },
    {
      "entropy": 5.500257778167724,
      "epoch": 0.9152757619738752,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923559055677173,
      "loss": 5.1922,
      "mean_token_accuracy": 0.1872135117650032,
      "num_tokens": 20944171.0,
      "step": 10090
    },
    {
      "entropy": 5.50403790473938,
      "epoch": 0.9157293178519593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923475409983848,
      "loss": 5.2372,
      "mean_token_accuracy": 0.1875006526708603,
      "num_tokens": 20954200.0,
      "step": 10095
    },
    {
      "entropy": 5.483087778091431,
      "epoch": 0.9161828737300436,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004923391719342497,
      "loss": 5.2503,
      "mean_token_accuracy": 0.18236380070447922,
      "num_tokens": 20964935.0,
      "step": 10100
    },
    {
      "entropy": 5.502628421783447,
      "epoch": 0.9166364296081277,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004923307983754852,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19101871997117997,
      "num_tokens": 20975499.0,
      "step": 10105
    },
    {
      "entropy": 5.433144903182983,
      "epoch": 0.9170899854862119,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004923224203222645,
      "loss": 5.1678,
      "mean_token_accuracy": 0.19008324295282364,
      "num_tokens": 20986199.0,
      "step": 10110
    },
    {
      "entropy": 5.4952733516693115,
      "epoch": 0.9175435413642961,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923140377747606,
      "loss": 5.2464,
      "mean_token_accuracy": 0.18770486861467361,
      "num_tokens": 20996313.0,
      "step": 10115
    },
    {
      "entropy": 5.533897113800049,
      "epoch": 0.9179970972423802,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923056507331471,
      "loss": 5.2228,
      "mean_token_accuracy": 0.18642332404851913,
      "num_tokens": 21006133.0,
      "step": 10120
    },
    {
      "entropy": 5.531605768203735,
      "epoch": 0.9184506531204645,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004922972591975974,
      "loss": 5.192,
      "mean_token_accuracy": 0.18831069022417068,
      "num_tokens": 21016448.0,
      "step": 10125
    },
    {
      "entropy": 5.451457118988037,
      "epoch": 0.9189042089985486,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922888631682849,
      "loss": 5.1271,
      "mean_token_accuracy": 0.18411844819784165,
      "num_tokens": 21026458.0,
      "step": 10130
    },
    {
      "entropy": 5.449035692214966,
      "epoch": 0.9193577648766328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004922804626453834,
      "loss": 5.1663,
      "mean_token_accuracy": 0.192974516749382,
      "num_tokens": 21037111.0,
      "step": 10135
    },
    {
      "entropy": 5.457956552505493,
      "epoch": 0.9198113207547169,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004922720576290665,
      "loss": 5.2224,
      "mean_token_accuracy": 0.19142512679100038,
      "num_tokens": 21048041.0,
      "step": 10140
    },
    {
      "entropy": 5.433384132385254,
      "epoch": 0.9202648766328012,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922636481195082,
      "loss": 5.0919,
      "mean_token_accuracy": 0.20250018388032914,
      "num_tokens": 21058099.0,
      "step": 10145
    },
    {
      "entropy": 5.501229667663575,
      "epoch": 0.9207184325108854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004922552341168823,
      "loss": 5.2367,
      "mean_token_accuracy": 0.18169017881155014,
      "num_tokens": 21068137.0,
      "step": 10150
    },
    {
      "entropy": 5.41462550163269,
      "epoch": 0.9211719883889695,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922468156213627,
      "loss": 5.1188,
      "mean_token_accuracy": 0.19230714738368987,
      "num_tokens": 21077556.0,
      "step": 10155
    },
    {
      "entropy": 5.499750471115112,
      "epoch": 0.9216255442670537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004922383926331236,
      "loss": 5.1527,
      "mean_token_accuracy": 0.19040969759225845,
      "num_tokens": 21087524.0,
      "step": 10160
    },
    {
      "entropy": 5.438167142868042,
      "epoch": 0.9220791001451378,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004922299651523392,
      "loss": 5.2295,
      "mean_token_accuracy": 0.1860123857855797,
      "num_tokens": 21098446.0,
      "step": 10165
    },
    {
      "entropy": 5.48976616859436,
      "epoch": 0.9225326560232221,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922215331791837,
      "loss": 5.2352,
      "mean_token_accuracy": 0.18725680112838744,
      "num_tokens": 21109291.0,
      "step": 10170
    },
    {
      "entropy": 5.514573240280152,
      "epoch": 0.9229862119013063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922130967138316,
      "loss": 5.1208,
      "mean_token_accuracy": 0.1932806611061096,
      "num_tokens": 21120043.0,
      "step": 10175
    },
    {
      "entropy": 5.518204164505005,
      "epoch": 0.9234397677793904,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004922046557564574,
      "loss": 5.2357,
      "mean_token_accuracy": 0.18277695327997207,
      "num_tokens": 21130339.0,
      "step": 10180
    },
    {
      "entropy": 5.500779438018799,
      "epoch": 0.9238933236574746,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921962103072355,
      "loss": 5.3023,
      "mean_token_accuracy": 0.18717591762542723,
      "num_tokens": 21142307.0,
      "step": 10185
    },
    {
      "entropy": 5.526569557189942,
      "epoch": 0.9243468795355588,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004921877603663406,
      "loss": 5.2036,
      "mean_token_accuracy": 0.19003356993198395,
      "num_tokens": 21151843.0,
      "step": 10190
    },
    {
      "entropy": 5.502495908737183,
      "epoch": 0.924800435413643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004921793059339474,
      "loss": 5.2119,
      "mean_token_accuracy": 0.19041070491075515,
      "num_tokens": 21162574.0,
      "step": 10195
    },
    {
      "entropy": 5.510891437530518,
      "epoch": 0.9252539912917271,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921708470102309,
      "loss": 5.16,
      "mean_token_accuracy": 0.19265214800834657,
      "num_tokens": 21172637.0,
      "step": 10200
    },
    {
      "entropy": 5.473992538452149,
      "epoch": 0.9257075471698113,
      "grad_norm": 0.953125,
      "learning_rate": 0.000492162383595366,
      "loss": 5.168,
      "mean_token_accuracy": 0.19391350150108339,
      "num_tokens": 21182272.0,
      "step": 10205
    },
    {
      "entropy": 5.540949773788452,
      "epoch": 0.9261611030478955,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921539156895275,
      "loss": 5.3006,
      "mean_token_accuracy": 0.18155533522367479,
      "num_tokens": 21193712.0,
      "step": 10210
    },
    {
      "entropy": 5.566745138168335,
      "epoch": 0.9266146589259797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004921454432928908,
      "loss": 5.303,
      "mean_token_accuracy": 0.18065586537122727,
      "num_tokens": 21205104.0,
      "step": 10215
    },
    {
      "entropy": 5.536470413208008,
      "epoch": 0.9270682148040639,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000492136966405631,
      "loss": 5.1908,
      "mean_token_accuracy": 0.18957612067461013,
      "num_tokens": 21215495.0,
      "step": 10220
    },
    {
      "entropy": 5.49644718170166,
      "epoch": 0.927521770682148,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004921284850279233,
      "loss": 5.2023,
      "mean_token_accuracy": 0.1892085149884224,
      "num_tokens": 21226047.0,
      "step": 10225
    },
    {
      "entropy": 5.50606107711792,
      "epoch": 0.9279753265602322,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921199991599433,
      "loss": 5.2756,
      "mean_token_accuracy": 0.18349545150995256,
      "num_tokens": 21235784.0,
      "step": 10230
    },
    {
      "entropy": 5.548442935943603,
      "epoch": 0.9284288824383164,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004921115088018665,
      "loss": 5.2245,
      "mean_token_accuracy": 0.18641836047172547,
      "num_tokens": 21245493.0,
      "step": 10235
    },
    {
      "entropy": 5.592808103561401,
      "epoch": 0.9288824383164006,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492103013953868,
      "loss": 5.3075,
      "mean_token_accuracy": 0.18445847779512406,
      "num_tokens": 21255541.0,
      "step": 10240
    },
    {
      "entropy": 5.46482138633728,
      "epoch": 0.9293359941944848,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920945146161242,
      "loss": 5.16,
      "mean_token_accuracy": 0.19516148865222932,
      "num_tokens": 21265427.0,
      "step": 10245
    },
    {
      "entropy": 5.470294094085693,
      "epoch": 0.9297895500725689,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004920860107888105,
      "loss": 5.164,
      "mean_token_accuracy": 0.19581332057714462,
      "num_tokens": 21276597.0,
      "step": 10250
    },
    {
      "entropy": 5.450829267501831,
      "epoch": 0.9302431059506531,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004920775024721026,
      "loss": 5.1916,
      "mean_token_accuracy": 0.18773079216480254,
      "num_tokens": 21286416.0,
      "step": 10255
    },
    {
      "entropy": 5.497037076950074,
      "epoch": 0.9306966618287373,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004920689896661767,
      "loss": 5.2054,
      "mean_token_accuracy": 0.1940332069993019,
      "num_tokens": 21296064.0,
      "step": 10260
    },
    {
      "entropy": 5.476668024063111,
      "epoch": 0.9311502177068215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004920604723712088,
      "loss": 5.214,
      "mean_token_accuracy": 0.19466549903154373,
      "num_tokens": 21306836.0,
      "step": 10265
    },
    {
      "entropy": 5.485483455657959,
      "epoch": 0.9316037735849056,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920519505873751,
      "loss": 5.1735,
      "mean_token_accuracy": 0.19273321628570556,
      "num_tokens": 21316936.0,
      "step": 10270
    },
    {
      "entropy": 5.449879693984985,
      "epoch": 0.9320573294629898,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920434243148515,
      "loss": 5.1367,
      "mean_token_accuracy": 0.19561291486024857,
      "num_tokens": 21326892.0,
      "step": 10275
    },
    {
      "entropy": 5.469815301895141,
      "epoch": 0.9325108853410741,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920348935538147,
      "loss": 5.1665,
      "mean_token_accuracy": 0.18587524741888045,
      "num_tokens": 21337270.0,
      "step": 10280
    },
    {
      "entropy": 5.525975322723388,
      "epoch": 0.9329644412191582,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000492026358304441,
      "loss": 5.2366,
      "mean_token_accuracy": 0.1867948666214943,
      "num_tokens": 21347968.0,
      "step": 10285
    },
    {
      "entropy": 5.483292770385742,
      "epoch": 0.9334179970972424,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492017818566907,
      "loss": 5.1862,
      "mean_token_accuracy": 0.1958588868379593,
      "num_tokens": 21358312.0,
      "step": 10290
    },
    {
      "entropy": 5.477849340438842,
      "epoch": 0.9338715529753265,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004920092743413892,
      "loss": 5.1513,
      "mean_token_accuracy": 0.19156461209058762,
      "num_tokens": 21368794.0,
      "step": 10295
    },
    {
      "entropy": 5.515837144851685,
      "epoch": 0.9343251088534107,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004920007256280642,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19680030047893524,
      "num_tokens": 21380080.0,
      "step": 10300
    },
    {
      "entropy": 5.507870388031006,
      "epoch": 0.934778664731495,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000491992172427109,
      "loss": 5.1906,
      "mean_token_accuracy": 0.19244442880153656,
      "num_tokens": 21389961.0,
      "step": 10305
    },
    {
      "entropy": 5.452802419662476,
      "epoch": 0.9352322206095791,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004919836147387002,
      "loss": 5.2319,
      "mean_token_accuracy": 0.18641486167907714,
      "num_tokens": 21399869.0,
      "step": 10310
    },
    {
      "entropy": 5.499537134170533,
      "epoch": 0.9356857764876633,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919750525630151,
      "loss": 5.2171,
      "mean_token_accuracy": 0.19221128672361373,
      "num_tokens": 21410343.0,
      "step": 10315
    },
    {
      "entropy": 5.527597951889038,
      "epoch": 0.9361393323657474,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919664859002306,
      "loss": 5.1953,
      "mean_token_accuracy": 0.18969294428825378,
      "num_tokens": 21420915.0,
      "step": 10320
    },
    {
      "entropy": 5.487546443939209,
      "epoch": 0.9365928882438317,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919579147505239,
      "loss": 5.1638,
      "mean_token_accuracy": 0.19285682290792466,
      "num_tokens": 21430879.0,
      "step": 10325
    },
    {
      "entropy": 5.531203317642212,
      "epoch": 0.9370464441219158,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004919493391140721,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18837825804948807,
      "num_tokens": 21440843.0,
      "step": 10330
    },
    {
      "entropy": 5.592268323898315,
      "epoch": 0.9375,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919407589910529,
      "loss": 5.4086,
      "mean_token_accuracy": 0.1807285413146019,
      "num_tokens": 21452132.0,
      "step": 10335
    },
    {
      "entropy": 5.493322324752808,
      "epoch": 0.9379535558780842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004919321743816434,
      "loss": 5.194,
      "mean_token_accuracy": 0.18563437461853027,
      "num_tokens": 21461861.0,
      "step": 10340
    },
    {
      "entropy": 5.499019622802734,
      "epoch": 0.9384071117561683,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004919235852860213,
      "loss": 5.1976,
      "mean_token_accuracy": 0.18736050128936768,
      "num_tokens": 21472244.0,
      "step": 10345
    },
    {
      "entropy": 5.545416593551636,
      "epoch": 0.9388606676342526,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004919149917043641,
      "loss": 5.2515,
      "mean_token_accuracy": 0.1836113914847374,
      "num_tokens": 21482394.0,
      "step": 10350
    },
    {
      "entropy": 5.498862457275391,
      "epoch": 0.9393142235123367,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004919063936368496,
      "loss": 5.276,
      "mean_token_accuracy": 0.18350204527378083,
      "num_tokens": 21493334.0,
      "step": 10355
    },
    {
      "entropy": 5.553494691848755,
      "epoch": 0.9397677793904209,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004918977910836557,
      "loss": 5.2386,
      "mean_token_accuracy": 0.1844620883464813,
      "num_tokens": 21504443.0,
      "step": 10360
    },
    {
      "entropy": 5.539630651473999,
      "epoch": 0.940221335268505,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918891840449601,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18865265250205993,
      "num_tokens": 21513908.0,
      "step": 10365
    },
    {
      "entropy": 5.566462993621826,
      "epoch": 0.9406748911465893,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000491880572520941,
      "loss": 5.2809,
      "mean_token_accuracy": 0.17776426523923874,
      "num_tokens": 21524349.0,
      "step": 10370
    },
    {
      "entropy": 5.651615619659424,
      "epoch": 0.9411284470246735,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918719565117763,
      "loss": 5.3383,
      "mean_token_accuracy": 0.18980624675750732,
      "num_tokens": 21535003.0,
      "step": 10375
    },
    {
      "entropy": 5.603833770751953,
      "epoch": 0.9415820029027576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004918633360176444,
      "loss": 5.3558,
      "mean_token_accuracy": 0.17872778922319413,
      "num_tokens": 21545045.0,
      "step": 10380
    },
    {
      "entropy": 5.461150074005127,
      "epoch": 0.9420355587808418,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918547110387234,
      "loss": 5.1816,
      "mean_token_accuracy": 0.19119360893964768,
      "num_tokens": 21554830.0,
      "step": 10385
    },
    {
      "entropy": 5.5173086643219,
      "epoch": 0.9424891146589259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004918460815751916,
      "loss": 5.1546,
      "mean_token_accuracy": 0.19252003729343414,
      "num_tokens": 21564693.0,
      "step": 10390
    },
    {
      "entropy": 5.464719247817993,
      "epoch": 0.9429426705370102,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004918374476272277,
      "loss": 5.1676,
      "mean_token_accuracy": 0.19065018594264985,
      "num_tokens": 21575475.0,
      "step": 10395
    },
    {
      "entropy": 5.483359146118164,
      "epoch": 0.9433962264150944,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004918288091950101,
      "loss": 5.1769,
      "mean_token_accuracy": 0.188684843480587,
      "num_tokens": 21586944.0,
      "step": 10400
    },
    {
      "entropy": 5.520866012573242,
      "epoch": 0.9438497822931785,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918201662787174,
      "loss": 5.2096,
      "mean_token_accuracy": 0.18385448306798935,
      "num_tokens": 21598170.0,
      "step": 10405
    },
    {
      "entropy": 5.47841067314148,
      "epoch": 0.9443033381712627,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004918115188785285,
      "loss": 5.2258,
      "mean_token_accuracy": 0.1873992457985878,
      "num_tokens": 21609170.0,
      "step": 10410
    },
    {
      "entropy": 5.554736661911011,
      "epoch": 0.9447568940493469,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004918028669946221,
      "loss": 5.2796,
      "mean_token_accuracy": 0.18662892132997513,
      "num_tokens": 21620171.0,
      "step": 10415
    },
    {
      "entropy": 5.514125776290894,
      "epoch": 0.9452104499274311,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004917942106271771,
      "loss": 5.2234,
      "mean_token_accuracy": 0.19277503490447997,
      "num_tokens": 21631072.0,
      "step": 10420
    },
    {
      "entropy": 5.437767839431762,
      "epoch": 0.9456640058055152,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917855497763727,
      "loss": 5.0861,
      "mean_token_accuracy": 0.19857796430587768,
      "num_tokens": 21641543.0,
      "step": 10425
    },
    {
      "entropy": 5.445878791809082,
      "epoch": 0.9461175616835994,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004917768844423878,
      "loss": 5.2319,
      "mean_token_accuracy": 0.19050681293010713,
      "num_tokens": 21651342.0,
      "step": 10430
    },
    {
      "entropy": 5.514336013793946,
      "epoch": 0.9465711175616836,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004917682146254017,
      "loss": 5.1868,
      "mean_token_accuracy": 0.19425029307603836,
      "num_tokens": 21661043.0,
      "step": 10435
    },
    {
      "entropy": 5.445535373687744,
      "epoch": 0.9470246734397678,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004917595403255938,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20240829586982728,
      "num_tokens": 21671197.0,
      "step": 10440
    },
    {
      "entropy": 5.40161247253418,
      "epoch": 0.947478229317852,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917508615431431,
      "loss": 5.1289,
      "mean_token_accuracy": 0.19438163042068482,
      "num_tokens": 21682066.0,
      "step": 10445
    },
    {
      "entropy": 5.521121835708618,
      "epoch": 0.9479317851959361,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004917421782782296,
      "loss": 5.193,
      "mean_token_accuracy": 0.19061861038208008,
      "num_tokens": 21692117.0,
      "step": 10450
    },
    {
      "entropy": 5.518373441696167,
      "epoch": 0.9483853410740203,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004917334905310325,
      "loss": 5.302,
      "mean_token_accuracy": 0.183126300573349,
      "num_tokens": 21703544.0,
      "step": 10455
    },
    {
      "entropy": 5.4479773998260494,
      "epoch": 0.9488388969521045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004917247983017317,
      "loss": 5.1334,
      "mean_token_accuracy": 0.19209074079990388,
      "num_tokens": 21712850.0,
      "step": 10460
    },
    {
      "entropy": 5.4908843517303465,
      "epoch": 0.9492924528301887,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917161015905067,
      "loss": 5.2731,
      "mean_token_accuracy": 0.1897297501564026,
      "num_tokens": 21723015.0,
      "step": 10465
    },
    {
      "entropy": 5.532942295074463,
      "epoch": 0.9497460087082729,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004917074003975376,
      "loss": 5.2384,
      "mean_token_accuracy": 0.18643336445093156,
      "num_tokens": 21732453.0,
      "step": 10470
    },
    {
      "entropy": 5.5696944236755375,
      "epoch": 0.950199564586357,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004916986947230042,
      "loss": 5.3059,
      "mean_token_accuracy": 0.18505144864320755,
      "num_tokens": 21743251.0,
      "step": 10475
    },
    {
      "entropy": 5.516680526733398,
      "epoch": 0.9506531204644412,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004916899845670866,
      "loss": 5.2131,
      "mean_token_accuracy": 0.18668761998414993,
      "num_tokens": 21752891.0,
      "step": 10480
    },
    {
      "entropy": 5.535280227661133,
      "epoch": 0.9511066763425254,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004916812699299649,
      "loss": 5.2144,
      "mean_token_accuracy": 0.19159982055425645,
      "num_tokens": 21763230.0,
      "step": 10485
    },
    {
      "entropy": 5.482826519012451,
      "epoch": 0.9515602322206096,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004916725508118193,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19009662866592408,
      "num_tokens": 21773663.0,
      "step": 10490
    },
    {
      "entropy": 5.464633321762085,
      "epoch": 0.9520137880986937,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004916638272128301,
      "loss": 5.2258,
      "mean_token_accuracy": 0.1891879215836525,
      "num_tokens": 21784265.0,
      "step": 10495
    },
    {
      "entropy": 5.4609263896942135,
      "epoch": 0.9524673439767779,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916550991331777,
      "loss": 5.1746,
      "mean_token_accuracy": 0.19152846932411194,
      "num_tokens": 21795056.0,
      "step": 10500
    },
    {
      "entropy": 5.498981094360351,
      "epoch": 0.9529208998548622,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004916463665730427,
      "loss": 5.1927,
      "mean_token_accuracy": 0.19459545016288757,
      "num_tokens": 21806204.0,
      "step": 10505
    },
    {
      "entropy": 5.49833664894104,
      "epoch": 0.9533744557329463,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916376295326057,
      "loss": 5.2152,
      "mean_token_accuracy": 0.1886645570397377,
      "num_tokens": 21816993.0,
      "step": 10510
    },
    {
      "entropy": 5.439201021194458,
      "epoch": 0.9538280116110305,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004916288880120472,
      "loss": 5.0649,
      "mean_token_accuracy": 0.2021304950118065,
      "num_tokens": 21827457.0,
      "step": 10515
    },
    {
      "entropy": 5.504736804962159,
      "epoch": 0.9542815674891146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004916201420115482,
      "loss": 5.1988,
      "mean_token_accuracy": 0.18791094422340393,
      "num_tokens": 21837555.0,
      "step": 10520
    },
    {
      "entropy": 5.517563915252685,
      "epoch": 0.9547351233671988,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916113915312895,
      "loss": 5.2174,
      "mean_token_accuracy": 0.19141177833080292,
      "num_tokens": 21847074.0,
      "step": 10525
    },
    {
      "entropy": 5.554729223251343,
      "epoch": 0.9551886792452831,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004916026365714519,
      "loss": 5.198,
      "mean_token_accuracy": 0.18883113265037538,
      "num_tokens": 21858333.0,
      "step": 10530
    },
    {
      "entropy": 5.469289255142212,
      "epoch": 0.9556422351233672,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004915938771322165,
      "loss": 5.2141,
      "mean_token_accuracy": 0.19106341451406478,
      "num_tokens": 21869132.0,
      "step": 10535
    },
    {
      "entropy": 5.455795764923096,
      "epoch": 0.9560957910014514,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915851132137646,
      "loss": 5.2446,
      "mean_token_accuracy": 0.1872527241706848,
      "num_tokens": 21879153.0,
      "step": 10540
    },
    {
      "entropy": 5.502502346038819,
      "epoch": 0.9565493468795355,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004915763448162774,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19310393184423447,
      "num_tokens": 21889063.0,
      "step": 10545
    },
    {
      "entropy": 5.538162755966186,
      "epoch": 0.9570029027576198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004915675719399362,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18670725971460342,
      "num_tokens": 21899005.0,
      "step": 10550
    },
    {
      "entropy": 5.509344387054443,
      "epoch": 0.9574564586357039,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004915587945849225,
      "loss": 5.1119,
      "mean_token_accuracy": 0.19825160056352614,
      "num_tokens": 21908391.0,
      "step": 10555
    },
    {
      "entropy": 5.460941791534424,
      "epoch": 0.9579100145137881,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004915500127514176,
      "loss": 5.193,
      "mean_token_accuracy": 0.19470654428005219,
      "num_tokens": 21917655.0,
      "step": 10560
    },
    {
      "entropy": 5.555629873275757,
      "epoch": 0.9583635703918723,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004915412264396034,
      "loss": 5.2493,
      "mean_token_accuracy": 0.19042410403490068,
      "num_tokens": 21928038.0,
      "step": 10565
    },
    {
      "entropy": 5.521947193145752,
      "epoch": 0.9588171262699564,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915324356496613,
      "loss": 5.14,
      "mean_token_accuracy": 0.19917869120836257,
      "num_tokens": 21938106.0,
      "step": 10570
    },
    {
      "entropy": 5.487814474105835,
      "epoch": 0.9592706821480407,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004915236403817733,
      "loss": 5.1474,
      "mean_token_accuracy": 0.19220392405986786,
      "num_tokens": 21947484.0,
      "step": 10575
    },
    {
      "entropy": 5.473394870758057,
      "epoch": 0.9597242380261248,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004915148406361213,
      "loss": 5.2315,
      "mean_token_accuracy": 0.18975483775138854,
      "num_tokens": 21957139.0,
      "step": 10580
    },
    {
      "entropy": 5.487856483459472,
      "epoch": 0.960177793904209,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004915060364128872,
      "loss": 5.2299,
      "mean_token_accuracy": 0.185565547645092,
      "num_tokens": 21967030.0,
      "step": 10585
    },
    {
      "entropy": 5.448148107528686,
      "epoch": 0.9606313497822931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914972277122532,
      "loss": 5.169,
      "mean_token_accuracy": 0.1888902634382248,
      "num_tokens": 21976418.0,
      "step": 10590
    },
    {
      "entropy": 5.49601378440857,
      "epoch": 0.9610849056603774,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004914884145344012,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19263349920511247,
      "num_tokens": 21987284.0,
      "step": 10595
    },
    {
      "entropy": 5.42640495300293,
      "epoch": 0.9615384615384616,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914795968795137,
      "loss": 5.103,
      "mean_token_accuracy": 0.1972813591361046,
      "num_tokens": 21997106.0,
      "step": 10600
    },
    {
      "entropy": 5.467606544494629,
      "epoch": 0.9619920174165457,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004914707747477728,
      "loss": 5.0823,
      "mean_token_accuracy": 0.192961286008358,
      "num_tokens": 22007589.0,
      "step": 10605
    },
    {
      "entropy": 5.48398494720459,
      "epoch": 0.9624455732946299,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914619481393613,
      "loss": 5.2221,
      "mean_token_accuracy": 0.189901639521122,
      "num_tokens": 22017669.0,
      "step": 10610
    },
    {
      "entropy": 5.617108011245728,
      "epoch": 0.962899129172714,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914531170544615,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18461449742317199,
      "num_tokens": 22029330.0,
      "step": 10615
    },
    {
      "entropy": 5.473918676376343,
      "epoch": 0.9633526850507983,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914442814932561,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19325099438428878,
      "num_tokens": 22039877.0,
      "step": 10620
    },
    {
      "entropy": 5.480970811843872,
      "epoch": 0.9638062409288825,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914354414559278,
      "loss": 5.2025,
      "mean_token_accuracy": 0.19430873394012452,
      "num_tokens": 22049579.0,
      "step": 10625
    },
    {
      "entropy": 5.378883743286133,
      "epoch": 0.9642597968069666,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004914265969426594,
      "loss": 5.0784,
      "mean_token_accuracy": 0.20659508407115937,
      "num_tokens": 22059608.0,
      "step": 10630
    },
    {
      "entropy": 5.469490385055542,
      "epoch": 0.9647133526850508,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004914177479536338,
      "loss": 5.2372,
      "mean_token_accuracy": 0.186066897213459,
      "num_tokens": 22070036.0,
      "step": 10635
    },
    {
      "entropy": 5.514468145370484,
      "epoch": 0.965166908563135,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491408894489034,
      "loss": 5.2266,
      "mean_token_accuracy": 0.18473343849182128,
      "num_tokens": 22080679.0,
      "step": 10640
    },
    {
      "entropy": 5.485489368438721,
      "epoch": 0.9656204644412192,
      "grad_norm": 0.875,
      "learning_rate": 0.0004914000365490432,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20180499255657197,
      "num_tokens": 22092192.0,
      "step": 10645
    },
    {
      "entropy": 5.480154228210449,
      "epoch": 0.9660740203193033,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004913911741338445,
      "loss": 5.2587,
      "mean_token_accuracy": 0.18562012463808059,
      "num_tokens": 22102722.0,
      "step": 10650
    },
    {
      "entropy": 5.419040966033935,
      "epoch": 0.9665275761973875,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004913823072436211,
      "loss": 5.1247,
      "mean_token_accuracy": 0.2006438970565796,
      "num_tokens": 22113521.0,
      "step": 10655
    },
    {
      "entropy": 5.514348745346069,
      "epoch": 0.9669811320754716,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913734358785565,
      "loss": 5.2083,
      "mean_token_accuracy": 0.19085652083158494,
      "num_tokens": 22124891.0,
      "step": 10660
    },
    {
      "entropy": 5.4864380836486815,
      "epoch": 0.9674346879535559,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004913645600388341,
      "loss": 5.2233,
      "mean_token_accuracy": 0.18323416858911515,
      "num_tokens": 22135972.0,
      "step": 10665
    },
    {
      "entropy": 5.536486482620239,
      "epoch": 0.9678882438316401,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004913556797246374,
      "loss": 5.2795,
      "mean_token_accuracy": 0.17618962824344636,
      "num_tokens": 22146623.0,
      "step": 10670
    },
    {
      "entropy": 5.481680917739868,
      "epoch": 0.9683417997097242,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913467949361503,
      "loss": 5.2002,
      "mean_token_accuracy": 0.18713169544935226,
      "num_tokens": 22157619.0,
      "step": 10675
    },
    {
      "entropy": 5.543104553222657,
      "epoch": 0.9687953555878084,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004913379056735561,
      "loss": 5.1873,
      "mean_token_accuracy": 0.1843522697687149,
      "num_tokens": 22168655.0,
      "step": 10680
    },
    {
      "entropy": 5.477829074859619,
      "epoch": 0.9692489114658926,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000491329011937039,
      "loss": 5.2341,
      "mean_token_accuracy": 0.18528518527746202,
      "num_tokens": 22178500.0,
      "step": 10685
    },
    {
      "entropy": 5.516538763046265,
      "epoch": 0.9697024673439768,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004913201137267828,
      "loss": 5.2278,
      "mean_token_accuracy": 0.18400785326957703,
      "num_tokens": 22188873.0,
      "step": 10690
    },
    {
      "entropy": 5.441410827636719,
      "epoch": 0.970156023222061,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004913112110429715,
      "loss": 5.1433,
      "mean_token_accuracy": 0.19197122305631636,
      "num_tokens": 22199743.0,
      "step": 10695
    },
    {
      "entropy": 5.446821880340576,
      "epoch": 0.9706095791001451,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004913023038857892,
      "loss": 5.096,
      "mean_token_accuracy": 0.1934520870447159,
      "num_tokens": 22210301.0,
      "step": 10700
    },
    {
      "entropy": 5.454444456100464,
      "epoch": 0.9710631349782293,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912933922554203,
      "loss": 5.1086,
      "mean_token_accuracy": 0.19499189853668214,
      "num_tokens": 22220771.0,
      "step": 10705
    },
    {
      "entropy": 5.497599124908447,
      "epoch": 0.9715166908563135,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912844761520487,
      "loss": 5.2156,
      "mean_token_accuracy": 0.18877754956483841,
      "num_tokens": 22231100.0,
      "step": 10710
    },
    {
      "entropy": 5.513539361953735,
      "epoch": 0.9719702467343977,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004912755555758592,
      "loss": 5.1916,
      "mean_token_accuracy": 0.19085400849580764,
      "num_tokens": 22240997.0,
      "step": 10715
    },
    {
      "entropy": 5.515014171600342,
      "epoch": 0.9724238026124818,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491266630527036,
      "loss": 5.2256,
      "mean_token_accuracy": 0.18837058991193772,
      "num_tokens": 22251795.0,
      "step": 10720
    },
    {
      "entropy": 5.54486312866211,
      "epoch": 0.972877358490566,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004912577010057638,
      "loss": 5.2592,
      "mean_token_accuracy": 0.1850780189037323,
      "num_tokens": 22262465.0,
      "step": 10725
    },
    {
      "entropy": 5.606755971908569,
      "epoch": 0.9733309143686503,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004912487670122273,
      "loss": 5.3558,
      "mean_token_accuracy": 0.18126363158226014,
      "num_tokens": 22272399.0,
      "step": 10730
    },
    {
      "entropy": 5.5661256313323975,
      "epoch": 0.9737844702467344,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912398285466111,
      "loss": 5.2756,
      "mean_token_accuracy": 0.18773142844438553,
      "num_tokens": 22282545.0,
      "step": 10735
    },
    {
      "entropy": 5.54059739112854,
      "epoch": 0.9742380261248186,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912308856091002,
      "loss": 5.2244,
      "mean_token_accuracy": 0.19068962037563325,
      "num_tokens": 22293154.0,
      "step": 10740
    },
    {
      "entropy": 5.5106744289398195,
      "epoch": 0.9746915820029027,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004912219381998795,
      "loss": 5.1466,
      "mean_token_accuracy": 0.19592952430248262,
      "num_tokens": 22302309.0,
      "step": 10745
    },
    {
      "entropy": 5.534269857406616,
      "epoch": 0.9751451378809869,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912129863191339,
      "loss": 5.2587,
      "mean_token_accuracy": 0.17782827466726303,
      "num_tokens": 22312316.0,
      "step": 10750
    },
    {
      "entropy": 5.525972032546997,
      "epoch": 0.9755986937590712,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912040299670487,
      "loss": 5.2329,
      "mean_token_accuracy": 0.19108207672834396,
      "num_tokens": 22322387.0,
      "step": 10755
    },
    {
      "entropy": 5.502649593353271,
      "epoch": 0.9760522496371553,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911950691438091,
      "loss": 5.2336,
      "mean_token_accuracy": 0.19430004805326462,
      "num_tokens": 22332759.0,
      "step": 10760
    },
    {
      "entropy": 5.5075767993927,
      "epoch": 0.9765058055152395,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004911861038496004,
      "loss": 5.2656,
      "mean_token_accuracy": 0.18402133882045746,
      "num_tokens": 22344056.0,
      "step": 10765
    },
    {
      "entropy": 5.455955648422242,
      "epoch": 0.9769593613933236,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911771340846079,
      "loss": 5.2084,
      "mean_token_accuracy": 0.18565851747989653,
      "num_tokens": 22353877.0,
      "step": 10770
    },
    {
      "entropy": 5.542308616638183,
      "epoch": 0.9774129172714079,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004911681598490172,
      "loss": 5.2525,
      "mean_token_accuracy": 0.1882826343178749,
      "num_tokens": 22363549.0,
      "step": 10775
    },
    {
      "entropy": 5.567686128616333,
      "epoch": 0.977866473149492,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004911591811430139,
      "loss": 5.2817,
      "mean_token_accuracy": 0.1853038787841797,
      "num_tokens": 22374040.0,
      "step": 10780
    },
    {
      "entropy": 5.568281555175782,
      "epoch": 0.9783200290275762,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004911501979667836,
      "loss": 5.3363,
      "mean_token_accuracy": 0.18623031675815582,
      "num_tokens": 22384819.0,
      "step": 10785
    },
    {
      "entropy": 5.527229642868042,
      "epoch": 0.9787735849056604,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004911412103205122,
      "loss": 5.199,
      "mean_token_accuracy": 0.19130196273326874,
      "num_tokens": 22395130.0,
      "step": 10790
    },
    {
      "entropy": 5.504364633560181,
      "epoch": 0.9792271407837445,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911322182043855,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18604805320501328,
      "num_tokens": 22405093.0,
      "step": 10795
    },
    {
      "entropy": 5.493117141723633,
      "epoch": 0.9796806966618288,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004911232216185893,
      "loss": 5.2306,
      "mean_token_accuracy": 0.18620621860027314,
      "num_tokens": 22415008.0,
      "step": 10800
    },
    {
      "entropy": 5.524082040786743,
      "epoch": 0.9801342525399129,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004911142205633099,
      "loss": 5.2778,
      "mean_token_accuracy": 0.1835907757282257,
      "num_tokens": 22426486.0,
      "step": 10805
    },
    {
      "entropy": 5.492581939697265,
      "epoch": 0.9805878084179971,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004911052150387334,
      "loss": 5.1811,
      "mean_token_accuracy": 0.18908338397741317,
      "num_tokens": 22436844.0,
      "step": 10810
    },
    {
      "entropy": 5.553999710083008,
      "epoch": 0.9810413642960812,
      "grad_norm": 1.125,
      "learning_rate": 0.0004910962050450458,
      "loss": 5.3059,
      "mean_token_accuracy": 0.18974106162786483,
      "num_tokens": 22446630.0,
      "step": 10815
    },
    {
      "entropy": 5.542135238647461,
      "epoch": 0.9814949201741655,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004910871905824338,
      "loss": 5.1387,
      "mean_token_accuracy": 0.1928548112511635,
      "num_tokens": 22456406.0,
      "step": 10820
    },
    {
      "entropy": 5.45408296585083,
      "epoch": 0.9819484760522497,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004910781716510837,
      "loss": 5.138,
      "mean_token_accuracy": 0.1933705762028694,
      "num_tokens": 22466210.0,
      "step": 10825
    },
    {
      "entropy": 5.4947069644927975,
      "epoch": 0.9824020319303338,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910691482511817,
      "loss": 5.2356,
      "mean_token_accuracy": 0.1944747418165207,
      "num_tokens": 22476978.0,
      "step": 10830
    },
    {
      "entropy": 5.59715723991394,
      "epoch": 0.982855587808418,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004910601203829148,
      "loss": 5.194,
      "mean_token_accuracy": 0.18800618201494218,
      "num_tokens": 22488613.0,
      "step": 10835
    },
    {
      "entropy": 5.467583513259887,
      "epoch": 0.9833091436865021,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004910510880464696,
      "loss": 5.1093,
      "mean_token_accuracy": 0.20096988826990128,
      "num_tokens": 22498312.0,
      "step": 10840
    },
    {
      "entropy": 5.482868099212647,
      "epoch": 0.9837626995645864,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910420512420327,
      "loss": 5.2111,
      "mean_token_accuracy": 0.1898895427584648,
      "num_tokens": 22508131.0,
      "step": 10845
    },
    {
      "entropy": 5.539039945602417,
      "epoch": 0.9842162554426706,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004910330099697913,
      "loss": 5.2132,
      "mean_token_accuracy": 0.1903369903564453,
      "num_tokens": 22517842.0,
      "step": 10850
    },
    {
      "entropy": 5.527409839630127,
      "epoch": 0.9846698113207547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910239642299321,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19023292511701584,
      "num_tokens": 22528239.0,
      "step": 10855
    },
    {
      "entropy": 5.466523361206055,
      "epoch": 0.9851233671988389,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004910149140226422,
      "loss": 5.1724,
      "mean_token_accuracy": 0.1914309471845627,
      "num_tokens": 22539734.0,
      "step": 10860
    },
    {
      "entropy": 5.503134536743164,
      "epoch": 0.9855769230769231,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004910058593481089,
      "loss": 5.1701,
      "mean_token_accuracy": 0.19570856541395187,
      "num_tokens": 22549796.0,
      "step": 10865
    },
    {
      "entropy": 5.4575072765350345,
      "epoch": 0.9860304789550073,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909968002065195,
      "loss": 5.158,
      "mean_token_accuracy": 0.1910772830247879,
      "num_tokens": 22559934.0,
      "step": 10870
    },
    {
      "entropy": 5.6052713871002195,
      "epoch": 0.9864840348330914,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000490987736598061,
      "loss": 5.2598,
      "mean_token_accuracy": 0.1797322601079941,
      "num_tokens": 22569864.0,
      "step": 10875
    },
    {
      "entropy": 5.504677820205688,
      "epoch": 0.9869375907111756,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004909786685229211,
      "loss": 5.1964,
      "mean_token_accuracy": 0.19451596289873124,
      "num_tokens": 22580687.0,
      "step": 10880
    },
    {
      "entropy": 5.538554286956787,
      "epoch": 0.9873911465892597,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004909695959812875,
      "loss": 5.2494,
      "mean_token_accuracy": 0.18825656920671463,
      "num_tokens": 22590507.0,
      "step": 10885
    },
    {
      "entropy": 5.55866584777832,
      "epoch": 0.987844702467344,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909605189733474,
      "loss": 5.2668,
      "mean_token_accuracy": 0.18947904407978058,
      "num_tokens": 22601901.0,
      "step": 10890
    },
    {
      "entropy": 5.501526641845703,
      "epoch": 0.9882982583454282,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909514374992889,
      "loss": 5.2155,
      "mean_token_accuracy": 0.18769705891609192,
      "num_tokens": 22613003.0,
      "step": 10895
    },
    {
      "entropy": 5.447288465499878,
      "epoch": 0.9887518142235123,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909423515592996,
      "loss": 5.0577,
      "mean_token_accuracy": 0.1979197680950165,
      "num_tokens": 22623401.0,
      "step": 10900
    },
    {
      "entropy": 5.519672107696533,
      "epoch": 0.9892053701015965,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909332611535675,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1883891209959984,
      "num_tokens": 22633147.0,
      "step": 10905
    },
    {
      "entropy": 5.47981538772583,
      "epoch": 0.9896589259796807,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004909241662822804,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20440202653408052,
      "num_tokens": 22642990.0,
      "step": 10910
    },
    {
      "entropy": 5.50399341583252,
      "epoch": 0.9901124818577649,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909150669456265,
      "loss": 5.2827,
      "mean_token_accuracy": 0.18473678827285767,
      "num_tokens": 22652734.0,
      "step": 10915
    },
    {
      "entropy": 5.488545656204224,
      "epoch": 0.9905660377358491,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000490905963143794,
      "loss": 5.2849,
      "mean_token_accuracy": 0.18099266439676284,
      "num_tokens": 22662968.0,
      "step": 10920
    },
    {
      "entropy": 5.44248948097229,
      "epoch": 0.9910195936139332,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004908968548769713,
      "loss": 5.1281,
      "mean_token_accuracy": 0.20352931469678878,
      "num_tokens": 22673157.0,
      "step": 10925
    },
    {
      "entropy": 5.515015172958374,
      "epoch": 0.9914731494920174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908877421453466,
      "loss": 5.2599,
      "mean_token_accuracy": 0.18747966289520263,
      "num_tokens": 22683105.0,
      "step": 10930
    },
    {
      "entropy": 5.4896814823150635,
      "epoch": 0.9919267053701016,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004908786249491082,
      "loss": 5.2022,
      "mean_token_accuracy": 0.1996642142534256,
      "num_tokens": 22693793.0,
      "step": 10935
    },
    {
      "entropy": 5.618080711364746,
      "epoch": 0.9923802612481858,
      "grad_norm": 1.0,
      "learning_rate": 0.000490869503288445,
      "loss": 5.252,
      "mean_token_accuracy": 0.18813893795013428,
      "num_tokens": 22703884.0,
      "step": 10940
    },
    {
      "entropy": 5.548908472061157,
      "epoch": 0.99283381712627,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004908603771635454,
      "loss": 5.2763,
      "mean_token_accuracy": 0.18646926432847977,
      "num_tokens": 22713699.0,
      "step": 10945
    },
    {
      "entropy": 5.472669267654419,
      "epoch": 0.9932873730043541,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908512465745982,
      "loss": 5.1256,
      "mean_token_accuracy": 0.19451946765184402,
      "num_tokens": 22722871.0,
      "step": 10950
    },
    {
      "entropy": 5.479227352142334,
      "epoch": 0.9937409288824384,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004908421115217922,
      "loss": 5.1383,
      "mean_token_accuracy": 0.1876304790377617,
      "num_tokens": 22733314.0,
      "step": 10955
    },
    {
      "entropy": 5.432501792907715,
      "epoch": 0.9941944847605225,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004908329720053164,
      "loss": 5.1323,
      "mean_token_accuracy": 0.19701928347349168,
      "num_tokens": 22743255.0,
      "step": 10960
    },
    {
      "entropy": 5.412756109237671,
      "epoch": 0.9946480406386067,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004908238280253597,
      "loss": 5.1195,
      "mean_token_accuracy": 0.19759708195924758,
      "num_tokens": 22753857.0,
      "step": 10965
    },
    {
      "entropy": 5.531611490249634,
      "epoch": 0.9951015965166908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908146795821112,
      "loss": 5.1134,
      "mean_token_accuracy": 0.19179392755031585,
      "num_tokens": 22764326.0,
      "step": 10970
    },
    {
      "entropy": 5.5185198307037355,
      "epoch": 0.995555152394775,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004908055266757602,
      "loss": 5.1934,
      "mean_token_accuracy": 0.19486576914787293,
      "num_tokens": 22774958.0,
      "step": 10975
    },
    {
      "entropy": 5.425844955444336,
      "epoch": 0.9960087082728593,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004907963693064958,
      "loss": 5.1128,
      "mean_token_accuracy": 0.1925733730196953,
      "num_tokens": 22785616.0,
      "step": 10980
    },
    {
      "entropy": 5.565258312225342,
      "epoch": 0.9964622641509434,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004907872074745075,
      "loss": 5.2837,
      "mean_token_accuracy": 0.1845133572816849,
      "num_tokens": 22795967.0,
      "step": 10985
    },
    {
      "entropy": 5.490162181854248,
      "epoch": 0.9969158200290276,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907780411799849,
      "loss": 5.2794,
      "mean_token_accuracy": 0.18557634502649306,
      "num_tokens": 22806335.0,
      "step": 10990
    },
    {
      "entropy": 5.548031806945801,
      "epoch": 0.9973693759071117,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907688704231173,
      "loss": 5.2144,
      "mean_token_accuracy": 0.19313637167215347,
      "num_tokens": 22816390.0,
      "step": 10995
    },
    {
      "entropy": 5.6186151027679445,
      "epoch": 0.997822931785196,
      "grad_norm": 1.0,
      "learning_rate": 0.0004907596952040945,
      "loss": 5.2726,
      "mean_token_accuracy": 0.18671630173921586,
      "num_tokens": 22826395.0,
      "step": 11000
    },
    {
      "entropy": 5.547494173049927,
      "epoch": 0.9982764876632801,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004907505155231063,
      "loss": 5.2584,
      "mean_token_accuracy": 0.18838359266519547,
      "num_tokens": 22836790.0,
      "step": 11005
    },
    {
      "entropy": 5.488029050827026,
      "epoch": 0.9987300435413643,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004907413313803424,
      "loss": 5.1407,
      "mean_token_accuracy": 0.19004008322954177,
      "num_tokens": 22846687.0,
      "step": 11010
    },
    {
      "entropy": 5.53559718132019,
      "epoch": 0.9991835994194485,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004907321427759928,
      "loss": 5.168,
      "mean_token_accuracy": 0.19271109402179717,
      "num_tokens": 22857208.0,
      "step": 11015
    },
    {
      "entropy": 5.513091611862182,
      "epoch": 0.9996371552975326,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004907229497102475,
      "loss": 5.2714,
      "mean_token_accuracy": 0.18417642563581466,
      "num_tokens": 22868556.0,
      "step": 11020
    },
    {
      "entropy": 5.550210618972779,
      "epoch": 1.0000907111756168,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004907137521832966,
      "loss": 5.2531,
      "mean_token_accuracy": 0.18888288289308547,
      "num_tokens": 22880081.0,
      "step": 11025
    },
    {
      "entropy": 5.494132804870605,
      "epoch": 1.000544267053701,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004907045501953304,
      "loss": 5.0935,
      "mean_token_accuracy": 0.19637981653213502,
      "num_tokens": 22892141.0,
      "step": 11030
    },
    {
      "entropy": 5.521714591979981,
      "epoch": 1.0009978229317853,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490695343746539,
      "loss": 5.1043,
      "mean_token_accuracy": 0.19702826887369157,
      "num_tokens": 22902688.0,
      "step": 11035
    },
    {
      "entropy": 5.4263575077056885,
      "epoch": 1.0014513788098693,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004906861328371131,
      "loss": 5.0424,
      "mean_token_accuracy": 0.1953244224190712,
      "num_tokens": 22912453.0,
      "step": 11040
    },
    {
      "entropy": 5.511580848693848,
      "epoch": 1.0019049346879536,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004906769174672429,
      "loss": 5.2007,
      "mean_token_accuracy": 0.18563229739665985,
      "num_tokens": 22923167.0,
      "step": 11045
    },
    {
      "entropy": 5.539322376251221,
      "epoch": 1.0023584905660377,
      "grad_norm": 0.953125,
      "learning_rate": 0.000490667697637119,
      "loss": 5.1607,
      "mean_token_accuracy": 0.18884927779436111,
      "num_tokens": 22934753.0,
      "step": 11050
    },
    {
      "entropy": 5.468543291091919,
      "epoch": 1.002812046444122,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004906584733469324,
      "loss": 5.0864,
      "mean_token_accuracy": 0.20028799921274185,
      "num_tokens": 22945277.0,
      "step": 11055
    },
    {
      "entropy": 5.522582197189331,
      "epoch": 1.0032656023222062,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004906492445968735,
      "loss": 5.153,
      "mean_token_accuracy": 0.18677352368831635,
      "num_tokens": 22955797.0,
      "step": 11060
    },
    {
      "entropy": 5.507099533081055,
      "epoch": 1.0037191582002902,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004906400113871332,
      "loss": 5.1494,
      "mean_token_accuracy": 0.19461202323436738,
      "num_tokens": 22965969.0,
      "step": 11065
    },
    {
      "entropy": 5.490888786315918,
      "epoch": 1.0041727140783745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004906307737179026,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20535642355680467,
      "num_tokens": 22975323.0,
      "step": 11070
    },
    {
      "entropy": 5.465540266036987,
      "epoch": 1.0046262699564585,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004906215315893726,
      "loss": 5.1059,
      "mean_token_accuracy": 0.19620248824357986,
      "num_tokens": 22985063.0,
      "step": 11075
    },
    {
      "entropy": 5.544905090332032,
      "epoch": 1.0050798258345428,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004906122850017344,
      "loss": 5.2382,
      "mean_token_accuracy": 0.1847388505935669,
      "num_tokens": 22996001.0,
      "step": 11080
    },
    {
      "entropy": 5.487858152389526,
      "epoch": 1.005533381712627,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004906030339551792,
      "loss": 5.1552,
      "mean_token_accuracy": 0.1921267867088318,
      "num_tokens": 23006896.0,
      "step": 11085
    },
    {
      "entropy": 5.534958028793335,
      "epoch": 1.0059869375907111,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004905937784498983,
      "loss": 5.1214,
      "mean_token_accuracy": 0.19611595422029496,
      "num_tokens": 23017080.0,
      "step": 11090
    },
    {
      "entropy": 5.504563283920288,
      "epoch": 1.0064404934687954,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004905845184860831,
      "loss": 5.0746,
      "mean_token_accuracy": 0.19814439564943315,
      "num_tokens": 23027168.0,
      "step": 11095
    },
    {
      "entropy": 5.45788164138794,
      "epoch": 1.0068940493468794,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004905752540639252,
      "loss": 5.1253,
      "mean_token_accuracy": 0.1916827917098999,
      "num_tokens": 23037364.0,
      "step": 11100
    },
    {
      "entropy": 5.4299335956573485,
      "epoch": 1.0073476052249637,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004905659851836161,
      "loss": 5.0356,
      "mean_token_accuracy": 0.19612774699926377,
      "num_tokens": 23047609.0,
      "step": 11105
    },
    {
      "entropy": 5.502022409439087,
      "epoch": 1.007801161103048,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004905567118453475,
      "loss": 5.1641,
      "mean_token_accuracy": 0.19047075808048247,
      "num_tokens": 23057942.0,
      "step": 11110
    },
    {
      "entropy": 5.470209121704102,
      "epoch": 1.008254716981132,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004905474340493111,
      "loss": 5.0727,
      "mean_token_accuracy": 0.19394411146640778,
      "num_tokens": 23069119.0,
      "step": 11115
    },
    {
      "entropy": 5.443198490142822,
      "epoch": 1.0087082728592163,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004905381517956988,
      "loss": 5.0687,
      "mean_token_accuracy": 0.19524529427289963,
      "num_tokens": 23080060.0,
      "step": 11120
    },
    {
      "entropy": 5.4462425231933596,
      "epoch": 1.0091618287373005,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004905288650847027,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1914682611823082,
      "num_tokens": 23090645.0,
      "step": 11125
    },
    {
      "entropy": 5.510693311691284,
      "epoch": 1.0096153846153846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004905195739165146,
      "loss": 5.1079,
      "mean_token_accuracy": 0.19879512786865233,
      "num_tokens": 23100449.0,
      "step": 11130
    },
    {
      "entropy": 5.496145629882813,
      "epoch": 1.0100689404934688,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490510278291327,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19123156666755675,
      "num_tokens": 23110415.0,
      "step": 11135
    },
    {
      "entropy": 5.447995948791504,
      "epoch": 1.010522496371553,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004905009782093318,
      "loss": 5.097,
      "mean_token_accuracy": 0.19483419209718705,
      "num_tokens": 23120336.0,
      "step": 11140
    },
    {
      "entropy": 5.425385141372681,
      "epoch": 1.0109760522496372,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004904916736707214,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1853318989276886,
      "num_tokens": 23130568.0,
      "step": 11145
    },
    {
      "entropy": 5.5349946975708,
      "epoch": 1.0114296081277214,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004904823646756885,
      "loss": 5.1248,
      "mean_token_accuracy": 0.19526209384202958,
      "num_tokens": 23140080.0,
      "step": 11150
    },
    {
      "entropy": 5.4925604343414305,
      "epoch": 1.0118831640058055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004904730512244252,
      "loss": 5.0562,
      "mean_token_accuracy": 0.1917285680770874,
      "num_tokens": 23150360.0,
      "step": 11155
    },
    {
      "entropy": 5.417743730545044,
      "epoch": 1.0123367198838897,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004904637333171243,
      "loss": 5.0548,
      "mean_token_accuracy": 0.18818178176879882,
      "num_tokens": 23160620.0,
      "step": 11160
    },
    {
      "entropy": 5.523185205459595,
      "epoch": 1.0127902757619738,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904544109539785,
      "loss": 5.1555,
      "mean_token_accuracy": 0.19521294981241227,
      "num_tokens": 23171920.0,
      "step": 11165
    },
    {
      "entropy": 5.492150783538818,
      "epoch": 1.013243831640058,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004904450841351805,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19893324375152588,
      "num_tokens": 23182929.0,
      "step": 11170
    },
    {
      "entropy": 5.490244626998901,
      "epoch": 1.0136973875181423,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004904357528609236,
      "loss": 5.0738,
      "mean_token_accuracy": 0.18984648883342742,
      "num_tokens": 23193020.0,
      "step": 11175
    },
    {
      "entropy": 5.427248907089234,
      "epoch": 1.0141509433962264,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004904264171314,
      "loss": 5.0854,
      "mean_token_accuracy": 0.19872664362192155,
      "num_tokens": 23204073.0,
      "step": 11180
    },
    {
      "entropy": 5.426294994354248,
      "epoch": 1.0146044992743106,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004904170769468035,
      "loss": 4.9959,
      "mean_token_accuracy": 0.1919043093919754,
      "num_tokens": 23214376.0,
      "step": 11185
    },
    {
      "entropy": 5.511206245422363,
      "epoch": 1.0150580551523947,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004904077323073268,
      "loss": 5.1612,
      "mean_token_accuracy": 0.19419621229171752,
      "num_tokens": 23224304.0,
      "step": 11190
    },
    {
      "entropy": 5.498494100570679,
      "epoch": 1.015511611030479,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004903983832131633,
      "loss": 5.1779,
      "mean_token_accuracy": 0.1909933015704155,
      "num_tokens": 23235324.0,
      "step": 11195
    },
    {
      "entropy": 5.499328231811523,
      "epoch": 1.0159651669085632,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903890296645063,
      "loss": 5.1352,
      "mean_token_accuracy": 0.19310501515865325,
      "num_tokens": 23245774.0,
      "step": 11200
    },
    {
      "entropy": 5.526724481582642,
      "epoch": 1.0164187227866472,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004903796716615495,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19474828839302064,
      "num_tokens": 23256274.0,
      "step": 11205
    },
    {
      "entropy": 5.4702582359313965,
      "epoch": 1.0168722786647315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903703092044861,
      "loss": 5.107,
      "mean_token_accuracy": 0.19234081357717514,
      "num_tokens": 23266304.0,
      "step": 11210
    },
    {
      "entropy": 5.389174270629883,
      "epoch": 1.0173258345428158,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004903609422935097,
      "loss": 5.0099,
      "mean_token_accuracy": 0.19811663776636124,
      "num_tokens": 23276552.0,
      "step": 11215
    },
    {
      "entropy": 5.4639805316925045,
      "epoch": 1.0177793904208998,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004903515709288143,
      "loss": 5.1256,
      "mean_token_accuracy": 0.1883719265460968,
      "num_tokens": 23287799.0,
      "step": 11220
    },
    {
      "entropy": 5.401679563522339,
      "epoch": 1.018232946298984,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004903421951105933,
      "loss": 4.9809,
      "mean_token_accuracy": 0.20238814353942872,
      "num_tokens": 23297958.0,
      "step": 11225
    },
    {
      "entropy": 5.5005217552185055,
      "epoch": 1.0186865021770681,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004903328148390411,
      "loss": 5.1624,
      "mean_token_accuracy": 0.18677162379026413,
      "num_tokens": 23307523.0,
      "step": 11230
    },
    {
      "entropy": 5.501028251647949,
      "epoch": 1.0191400580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004903234301143513,
      "loss": 5.1515,
      "mean_token_accuracy": 0.19437760561704637,
      "num_tokens": 23317970.0,
      "step": 11235
    },
    {
      "entropy": 5.460223770141601,
      "epoch": 1.0195936139332367,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004903140409367181,
      "loss": 5.1074,
      "mean_token_accuracy": 0.19670169800519943,
      "num_tokens": 23328612.0,
      "step": 11240
    },
    {
      "entropy": 5.452767467498779,
      "epoch": 1.0200471698113207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004903046473063357,
      "loss": 5.0766,
      "mean_token_accuracy": 0.1902572751045227,
      "num_tokens": 23338104.0,
      "step": 11245
    },
    {
      "entropy": 5.446892786026001,
      "epoch": 1.020500725689405,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004902952492233983,
      "loss": 5.0129,
      "mean_token_accuracy": 0.2074433147907257,
      "num_tokens": 23348105.0,
      "step": 11250
    },
    {
      "entropy": 5.42758994102478,
      "epoch": 1.020954281567489,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004902858466881002,
      "loss": 4.9701,
      "mean_token_accuracy": 0.19822793155908586,
      "num_tokens": 23357150.0,
      "step": 11255
    },
    {
      "entropy": 5.470262479782105,
      "epoch": 1.0214078374455733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490276439700636,
      "loss": 5.182,
      "mean_token_accuracy": 0.18963402211666108,
      "num_tokens": 23367223.0,
      "step": 11260
    },
    {
      "entropy": 5.476350355148315,
      "epoch": 1.0218613933236576,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004902670282612002,
      "loss": 5.1177,
      "mean_token_accuracy": 0.19103692173957826,
      "num_tokens": 23377664.0,
      "step": 11265
    },
    {
      "entropy": 5.442233085632324,
      "epoch": 1.0223149492017416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004902576123699873,
      "loss": 5.0096,
      "mean_token_accuracy": 0.19532963782548904,
      "num_tokens": 23388057.0,
      "step": 11270
    },
    {
      "entropy": 5.381910991668701,
      "epoch": 1.0227685050798259,
      "grad_norm": 1.0,
      "learning_rate": 0.0004902481920271922,
      "loss": 4.9596,
      "mean_token_accuracy": 0.19629869908094405,
      "num_tokens": 23397808.0,
      "step": 11275
    },
    {
      "entropy": 5.395374584197998,
      "epoch": 1.02322206095791,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004902387672330097,
      "loss": 5.0443,
      "mean_token_accuracy": 0.19195702075958251,
      "num_tokens": 23406979.0,
      "step": 11280
    },
    {
      "entropy": 5.450476551055909,
      "epoch": 1.0236756168359942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004902293379876345,
      "loss": 5.0932,
      "mean_token_accuracy": 0.19491908848285674,
      "num_tokens": 23416545.0,
      "step": 11285
    },
    {
      "entropy": 5.444452714920044,
      "epoch": 1.0241291727140784,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004902199042912618,
      "loss": 5.0985,
      "mean_token_accuracy": 0.18906430751085282,
      "num_tokens": 23427464.0,
      "step": 11290
    },
    {
      "entropy": 5.502170610427856,
      "epoch": 1.0245827285921625,
      "grad_norm": 1.0,
      "learning_rate": 0.0004902104661440868,
      "loss": 5.1325,
      "mean_token_accuracy": 0.19716407209634781,
      "num_tokens": 23436690.0,
      "step": 11295
    },
    {
      "entropy": 5.468004512786865,
      "epoch": 1.0250362844702468,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004902010235463042,
      "loss": 5.0506,
      "mean_token_accuracy": 0.19812326580286027,
      "num_tokens": 23446701.0,
      "step": 11300
    },
    {
      "entropy": 5.535991668701172,
      "epoch": 1.025489840348331,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901915764981098,
      "loss": 5.1445,
      "mean_token_accuracy": 0.19191909879446029,
      "num_tokens": 23456966.0,
      "step": 11305
    },
    {
      "entropy": 5.470449829101563,
      "epoch": 1.025943396226415,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004901821249996988,
      "loss": 5.1001,
      "mean_token_accuracy": 0.1941937953233719,
      "num_tokens": 23467300.0,
      "step": 11310
    },
    {
      "entropy": 5.5109635353088375,
      "epoch": 1.0263969521044993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004901726690512665,
      "loss": 5.162,
      "mean_token_accuracy": 0.18955274671316147,
      "num_tokens": 23478457.0,
      "step": 11315
    },
    {
      "entropy": 5.4429868221282955,
      "epoch": 1.0268505079825834,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004901632086530087,
      "loss": 5.044,
      "mean_token_accuracy": 0.2002337396144867,
      "num_tokens": 23488066.0,
      "step": 11320
    },
    {
      "entropy": 5.4729485511779785,
      "epoch": 1.0273040638606676,
      "grad_norm": 1.0,
      "learning_rate": 0.0004901537438051207,
      "loss": 5.1232,
      "mean_token_accuracy": 0.18899547904729844,
      "num_tokens": 23498856.0,
      "step": 11325
    },
    {
      "entropy": 5.491955137252807,
      "epoch": 1.027757619738752,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901442745077986,
      "loss": 5.0879,
      "mean_token_accuracy": 0.1923009291291237,
      "num_tokens": 23509845.0,
      "step": 11330
    },
    {
      "entropy": 5.527870321273804,
      "epoch": 1.028211175616836,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901348007612381,
      "loss": 5.2004,
      "mean_token_accuracy": 0.1924690917134285,
      "num_tokens": 23520322.0,
      "step": 11335
    },
    {
      "entropy": 5.480651807785034,
      "epoch": 1.0286647314949202,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004901253225656351,
      "loss": 5.1352,
      "mean_token_accuracy": 0.19338575154542922,
      "num_tokens": 23530030.0,
      "step": 11340
    },
    {
      "entropy": 5.565394401550293,
      "epoch": 1.0291182873730043,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004901158399211855,
      "loss": 5.1925,
      "mean_token_accuracy": 0.19459724128246308,
      "num_tokens": 23541063.0,
      "step": 11345
    },
    {
      "entropy": 5.423709774017334,
      "epoch": 1.0295718432510885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004901063528280856,
      "loss": 5.0923,
      "mean_token_accuracy": 0.19012126177549363,
      "num_tokens": 23551693.0,
      "step": 11350
    },
    {
      "entropy": 5.477392244338989,
      "epoch": 1.0300253991291728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900968612865314,
      "loss": 5.1251,
      "mean_token_accuracy": 0.1971671923995018,
      "num_tokens": 23561758.0,
      "step": 11355
    },
    {
      "entropy": 5.418482398986816,
      "epoch": 1.0304789550072568,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004900873652967195,
      "loss": 5.0701,
      "mean_token_accuracy": 0.19031274020671846,
      "num_tokens": 23571912.0,
      "step": 11360
    },
    {
      "entropy": 5.442516422271728,
      "epoch": 1.030932510885341,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004900778648588459,
      "loss": 5.035,
      "mean_token_accuracy": 0.19667790979146957,
      "num_tokens": 23582600.0,
      "step": 11365
    },
    {
      "entropy": 5.440606641769409,
      "epoch": 1.0313860667634251,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900683599731073,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20124102234840394,
      "num_tokens": 23592761.0,
      "step": 11370
    },
    {
      "entropy": 5.49815354347229,
      "epoch": 1.0318396226415094,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900588506397002,
      "loss": 5.1829,
      "mean_token_accuracy": 0.1900082364678383,
      "num_tokens": 23603119.0,
      "step": 11375
    },
    {
      "entropy": 5.384500169754029,
      "epoch": 1.0322931785195937,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004900493368588213,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20479914397001267,
      "num_tokens": 23613916.0,
      "step": 11380
    },
    {
      "entropy": 5.526654195785523,
      "epoch": 1.0327467343976777,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004900398186306672,
      "loss": 5.2217,
      "mean_token_accuracy": 0.1877501666545868,
      "num_tokens": 23623956.0,
      "step": 11385
    },
    {
      "entropy": 5.456167316436767,
      "epoch": 1.033200290275762,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000490030295955435,
      "loss": 5.1032,
      "mean_token_accuracy": 0.19367883652448653,
      "num_tokens": 23634666.0,
      "step": 11390
    },
    {
      "entropy": 5.424622917175293,
      "epoch": 1.0336538461538463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900207688333215,
      "loss": 5.0608,
      "mean_token_accuracy": 0.18913532048463821,
      "num_tokens": 23645601.0,
      "step": 11395
    },
    {
      "entropy": 5.510989427566528,
      "epoch": 1.0341074020319303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900112372645236,
      "loss": 5.1375,
      "mean_token_accuracy": 0.19037441164255142,
      "num_tokens": 23656339.0,
      "step": 11400
    },
    {
      "entropy": 5.461063098907471,
      "epoch": 1.0345609579100146,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004900017012492386,
      "loss": 5.1066,
      "mean_token_accuracy": 0.19083172529935838,
      "num_tokens": 23667433.0,
      "step": 11405
    },
    {
      "entropy": 5.502894353866577,
      "epoch": 1.0350145137880986,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004899921607876634,
      "loss": 5.1234,
      "mean_token_accuracy": 0.19392693638801575,
      "num_tokens": 23678186.0,
      "step": 11410
    },
    {
      "entropy": 5.466448020935059,
      "epoch": 1.0354680696661829,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004899826158799958,
      "loss": 5.0781,
      "mean_token_accuracy": 0.19278447926044465,
      "num_tokens": 23688317.0,
      "step": 11415
    },
    {
      "entropy": 5.432892513275147,
      "epoch": 1.0359216255442671,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899730665264328,
      "loss": 5.0298,
      "mean_token_accuracy": 0.19775377660989762,
      "num_tokens": 23697752.0,
      "step": 11420
    },
    {
      "entropy": 5.389712142944336,
      "epoch": 1.0363751814223512,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004899635127271721,
      "loss": 5.007,
      "mean_token_accuracy": 0.20241805762052537,
      "num_tokens": 23707803.0,
      "step": 11425
    },
    {
      "entropy": 5.372399187088012,
      "epoch": 1.0368287373004355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899539544824111,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2028212293982506,
      "num_tokens": 23717630.0,
      "step": 11430
    },
    {
      "entropy": 5.464516067504883,
      "epoch": 1.0372822931785195,
      "grad_norm": 1.0,
      "learning_rate": 0.0004899443917923476,
      "loss": 5.1428,
      "mean_token_accuracy": 0.1955379366874695,
      "num_tokens": 23728836.0,
      "step": 11435
    },
    {
      "entropy": 5.42895884513855,
      "epoch": 1.0377358490566038,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004899348246571792,
      "loss": 5.1186,
      "mean_token_accuracy": 0.18906114101409913,
      "num_tokens": 23740145.0,
      "step": 11440
    },
    {
      "entropy": 5.456284904479981,
      "epoch": 1.038189404934688,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000489925253077104,
      "loss": 5.1029,
      "mean_token_accuracy": 0.19079043120145797,
      "num_tokens": 23751162.0,
      "step": 11445
    },
    {
      "entropy": 5.491913223266602,
      "epoch": 1.038642960812772,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004899156770523196,
      "loss": 5.1081,
      "mean_token_accuracy": 0.19240654408931732,
      "num_tokens": 23761634.0,
      "step": 11450
    },
    {
      "entropy": 5.470863246917725,
      "epoch": 1.0390965166908563,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004899060965830244,
      "loss": 5.0733,
      "mean_token_accuracy": 0.198917418718338,
      "num_tokens": 23772483.0,
      "step": 11455
    },
    {
      "entropy": 5.45157265663147,
      "epoch": 1.0395500725689404,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004898965116694162,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1927415356040001,
      "num_tokens": 23782298.0,
      "step": 11460
    },
    {
      "entropy": 5.468653631210327,
      "epoch": 1.0400036284470247,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898869223116935,
      "loss": 5.1224,
      "mean_token_accuracy": 0.19079854786396028,
      "num_tokens": 23792492.0,
      "step": 11465
    },
    {
      "entropy": 5.510800075531006,
      "epoch": 1.040457184325109,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004898773285100546,
      "loss": 5.2672,
      "mean_token_accuracy": 0.184591606259346,
      "num_tokens": 23802999.0,
      "step": 11470
    },
    {
      "entropy": 5.473024654388428,
      "epoch": 1.040910740203193,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004898677302646975,
      "loss": 5.0636,
      "mean_token_accuracy": 0.19724741876125335,
      "num_tokens": 23812541.0,
      "step": 11475
    },
    {
      "entropy": 5.47519736289978,
      "epoch": 1.0413642960812772,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004898581275758211,
      "loss": 5.1705,
      "mean_token_accuracy": 0.18411127626895904,
      "num_tokens": 23823244.0,
      "step": 11480
    },
    {
      "entropy": 5.45018162727356,
      "epoch": 1.0418178519593615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898485204436239,
      "loss": 5.0092,
      "mean_token_accuracy": 0.2057377353310585,
      "num_tokens": 23832346.0,
      "step": 11485
    },
    {
      "entropy": 5.522273445129395,
      "epoch": 1.0422714078374455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004898389088683047,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20061220675706865,
      "num_tokens": 23842554.0,
      "step": 11490
    },
    {
      "entropy": 5.405847120285034,
      "epoch": 1.0427249637155298,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004898292928500619,
      "loss": 4.9989,
      "mean_token_accuracy": 0.1994738131761551,
      "num_tokens": 23852938.0,
      "step": 11495
    },
    {
      "entropy": 5.394134998321533,
      "epoch": 1.0431785195936139,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898196723890947,
      "loss": 5.0153,
      "mean_token_accuracy": 0.1991426929831505,
      "num_tokens": 23862367.0,
      "step": 11500
    },
    {
      "entropy": 5.448208570480347,
      "epoch": 1.0436320754716981,
      "grad_norm": 1.0,
      "learning_rate": 0.0004898100474856019,
      "loss": 5.1276,
      "mean_token_accuracy": 0.19397516846656798,
      "num_tokens": 23873249.0,
      "step": 11505
    },
    {
      "entropy": 5.48231987953186,
      "epoch": 1.0440856313497824,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004898004181397824,
      "loss": 5.1428,
      "mean_token_accuracy": 0.18477273732423782,
      "num_tokens": 23883830.0,
      "step": 11510
    },
    {
      "entropy": 5.4982703685760494,
      "epoch": 1.0445391872278664,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004897907843518358,
      "loss": 5.0533,
      "mean_token_accuracy": 0.19381560236215592,
      "num_tokens": 23894359.0,
      "step": 11515
    },
    {
      "entropy": 5.442965316772461,
      "epoch": 1.0449927431059507,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897811461219609,
      "loss": 5.0583,
      "mean_token_accuracy": 0.19391554594039917,
      "num_tokens": 23903779.0,
      "step": 11520
    },
    {
      "entropy": 5.438404083251953,
      "epoch": 1.0454462989840347,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004897715034503571,
      "loss": 5.1322,
      "mean_token_accuracy": 0.19373399317264556,
      "num_tokens": 23914523.0,
      "step": 11525
    },
    {
      "entropy": 5.456286334991455,
      "epoch": 1.045899854862119,
      "grad_norm": 0.96875,
      "learning_rate": 0.000489761856337224,
      "loss": 5.0995,
      "mean_token_accuracy": 0.19320344924926758,
      "num_tokens": 23924996.0,
      "step": 11530
    },
    {
      "entropy": 5.477241516113281,
      "epoch": 1.0463534107402033,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004897522047827609,
      "loss": 5.0655,
      "mean_token_accuracy": 0.19749463647603988,
      "num_tokens": 23935843.0,
      "step": 11535
    },
    {
      "entropy": 5.444448661804199,
      "epoch": 1.0468069666182873,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004897425487871676,
      "loss": 5.0376,
      "mean_token_accuracy": 0.19542952328920365,
      "num_tokens": 23945044.0,
      "step": 11540
    },
    {
      "entropy": 5.400130081176758,
      "epoch": 1.0472605224963716,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004897328883506436,
      "loss": 5.0997,
      "mean_token_accuracy": 0.19553857147693635,
      "num_tokens": 23955933.0,
      "step": 11545
    },
    {
      "entropy": 5.461530494689941,
      "epoch": 1.0477140783744556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897232234733887,
      "loss": 5.1683,
      "mean_token_accuracy": 0.18461367785930632,
      "num_tokens": 23966320.0,
      "step": 11550
    },
    {
      "entropy": 5.461652421951294,
      "epoch": 1.04816763425254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004897135541556028,
      "loss": 5.061,
      "mean_token_accuracy": 0.19855878949165345,
      "num_tokens": 23976233.0,
      "step": 11555
    },
    {
      "entropy": 5.471328020095825,
      "epoch": 1.0486211901306242,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000489703880397486,
      "loss": 5.1227,
      "mean_token_accuracy": 0.1904810532927513,
      "num_tokens": 23987854.0,
      "step": 11560
    },
    {
      "entropy": 5.4512513160705565,
      "epoch": 1.0490747460087082,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004896942021992382,
      "loss": 5.1113,
      "mean_token_accuracy": 0.1881568431854248,
      "num_tokens": 23997973.0,
      "step": 11565
    },
    {
      "entropy": 5.459875631332397,
      "epoch": 1.0495283018867925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004896845195610595,
      "loss": 5.1236,
      "mean_token_accuracy": 0.1908821791410446,
      "num_tokens": 24007313.0,
      "step": 11570
    },
    {
      "entropy": 5.449109792709351,
      "epoch": 1.0499818577648767,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004896748324831504,
      "loss": 5.1376,
      "mean_token_accuracy": 0.18894692361354828,
      "num_tokens": 24018254.0,
      "step": 11575
    },
    {
      "entropy": 5.488360786437989,
      "epoch": 1.0504354136429608,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004896651409657109,
      "loss": 5.1272,
      "mean_token_accuracy": 0.19029901772737504,
      "num_tokens": 24029799.0,
      "step": 11580
    },
    {
      "entropy": 5.407639503479004,
      "epoch": 1.050888969521045,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896554450089417,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20030184984207153,
      "num_tokens": 24040972.0,
      "step": 11585
    },
    {
      "entropy": 5.443329524993897,
      "epoch": 1.051342525399129,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896457446130431,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20220689922571183,
      "num_tokens": 24050781.0,
      "step": 11590
    },
    {
      "entropy": 5.429394340515136,
      "epoch": 1.0517960812772134,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004896360397782159,
      "loss": 5.0991,
      "mean_token_accuracy": 0.19227786213159562,
      "num_tokens": 24060184.0,
      "step": 11595
    },
    {
      "entropy": 5.489581871032715,
      "epoch": 1.0522496371552976,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896263305046607,
      "loss": 5.1549,
      "mean_token_accuracy": 0.19510595351457596,
      "num_tokens": 24070777.0,
      "step": 11600
    },
    {
      "entropy": 5.507211112976075,
      "epoch": 1.0527031930333817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004896166167925782,
      "loss": 5.1887,
      "mean_token_accuracy": 0.1927416890859604,
      "num_tokens": 24080878.0,
      "step": 11605
    },
    {
      "entropy": 5.492761325836182,
      "epoch": 1.053156748911466,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004896068986421693,
      "loss": 5.137,
      "mean_token_accuracy": 0.20071599036455154,
      "num_tokens": 24090038.0,
      "step": 11610
    },
    {
      "entropy": 5.528735876083374,
      "epoch": 1.05361030478955,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004895971760536351,
      "loss": 5.1714,
      "mean_token_accuracy": 0.19025161862373352,
      "num_tokens": 24100425.0,
      "step": 11615
    },
    {
      "entropy": 5.519040775299072,
      "epoch": 1.0540638606676342,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004895874490271768,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18544750064611434,
      "num_tokens": 24112182.0,
      "step": 11620
    },
    {
      "entropy": 5.479070711135864,
      "epoch": 1.0545174165457185,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895777175629952,
      "loss": 5.0768,
      "mean_token_accuracy": 0.19402081668376922,
      "num_tokens": 24122182.0,
      "step": 11625
    },
    {
      "entropy": 5.512710905075073,
      "epoch": 1.0549709724238026,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004895679816612917,
      "loss": 5.0755,
      "mean_token_accuracy": 0.1985362336039543,
      "num_tokens": 24131930.0,
      "step": 11630
    },
    {
      "entropy": 5.499684190750122,
      "epoch": 1.0554245283018868,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004895582413222678,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20388585031032563,
      "num_tokens": 24141575.0,
      "step": 11635
    },
    {
      "entropy": 5.482347249984741,
      "epoch": 1.0558780841799709,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895484965461247,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1858080580830574,
      "num_tokens": 24151720.0,
      "step": 11640
    },
    {
      "entropy": 5.427149391174316,
      "epoch": 1.0563316400580551,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000489538747333064,
      "loss": 5.1075,
      "mean_token_accuracy": 0.19254984259605407,
      "num_tokens": 24161881.0,
      "step": 11645
    },
    {
      "entropy": 5.449149322509766,
      "epoch": 1.0567851959361394,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004895289936832874,
      "loss": 5.0637,
      "mean_token_accuracy": 0.20020053237676622,
      "num_tokens": 24171474.0,
      "step": 11650
    },
    {
      "entropy": 5.552807998657227,
      "epoch": 1.0572387518142234,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004895192355969966,
      "loss": 5.1718,
      "mean_token_accuracy": 0.19173015356063844,
      "num_tokens": 24182186.0,
      "step": 11655
    },
    {
      "entropy": 5.49303035736084,
      "epoch": 1.0576923076923077,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004895094730743932,
      "loss": 5.0972,
      "mean_token_accuracy": 0.20089954733848572,
      "num_tokens": 24192318.0,
      "step": 11660
    },
    {
      "entropy": 5.480742073059082,
      "epoch": 1.058145863570392,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004894997061156793,
      "loss": 5.1583,
      "mean_token_accuracy": 0.19043267518281937,
      "num_tokens": 24202918.0,
      "step": 11665
    },
    {
      "entropy": 5.426549005508423,
      "epoch": 1.058599419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894899347210567,
      "loss": 5.0545,
      "mean_token_accuracy": 0.19537004977464675,
      "num_tokens": 24213218.0,
      "step": 11670
    },
    {
      "entropy": 5.448081922531128,
      "epoch": 1.0590529753265603,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004894801588907278,
      "loss": 5.0414,
      "mean_token_accuracy": 0.19819533377885817,
      "num_tokens": 24223229.0,
      "step": 11675
    },
    {
      "entropy": 5.508687114715576,
      "epoch": 1.0595065312046443,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004894703786248944,
      "loss": 5.1189,
      "mean_token_accuracy": 0.19095162451267242,
      "num_tokens": 24233637.0,
      "step": 11680
    },
    {
      "entropy": 5.4921238899230955,
      "epoch": 1.0599600870827286,
      "grad_norm": 1.03125,
      "learning_rate": 0.000489460593923759,
      "loss": 5.149,
      "mean_token_accuracy": 0.19814313352108,
      "num_tokens": 24244251.0,
      "step": 11685
    },
    {
      "entropy": 5.471730422973633,
      "epoch": 1.0604136429608129,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004894508047875237,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19785570055246354,
      "num_tokens": 24254380.0,
      "step": 11690
    },
    {
      "entropy": 5.340200662612915,
      "epoch": 1.060867198838897,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004894410112163913,
      "loss": 4.9937,
      "mean_token_accuracy": 0.19913735389709472,
      "num_tokens": 24264841.0,
      "step": 11695
    },
    {
      "entropy": 5.391559076309204,
      "epoch": 1.0613207547169812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004894312132105641,
      "loss": 5.0239,
      "mean_token_accuracy": 0.19785647094249725,
      "num_tokens": 24274383.0,
      "step": 11700
    },
    {
      "entropy": 5.534787845611572,
      "epoch": 1.0617743105950652,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004894214107702447,
      "loss": 5.11,
      "mean_token_accuracy": 0.19086095243692397,
      "num_tokens": 24284140.0,
      "step": 11705
    },
    {
      "entropy": 5.560810375213623,
      "epoch": 1.0622278664731495,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000489411603895636,
      "loss": 5.1878,
      "mean_token_accuracy": 0.1921544551849365,
      "num_tokens": 24295044.0,
      "step": 11710
    },
    {
      "entropy": 5.451190757751465,
      "epoch": 1.0626814223512338,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894017925869406,
      "loss": 5.0975,
      "mean_token_accuracy": 0.19553182721138002,
      "num_tokens": 24305080.0,
      "step": 11715
    },
    {
      "entropy": 5.466040515899659,
      "epoch": 1.0631349782293178,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893919768443615,
      "loss": 5.1308,
      "mean_token_accuracy": 0.19310673475265502,
      "num_tokens": 24316306.0,
      "step": 11720
    },
    {
      "entropy": 5.445856237411499,
      "epoch": 1.063588534107402,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004893821566681017,
      "loss": 5.1232,
      "mean_token_accuracy": 0.19693950265645982,
      "num_tokens": 24326740.0,
      "step": 11725
    },
    {
      "entropy": 5.489151525497436,
      "epoch": 1.0640420899854863,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004893723320583644,
      "loss": 5.162,
      "mean_token_accuracy": 0.188465815782547,
      "num_tokens": 24337621.0,
      "step": 11730
    },
    {
      "entropy": 5.386200571060181,
      "epoch": 1.0644956458635704,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004893625030153525,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20463671237230302,
      "num_tokens": 24347732.0,
      "step": 11735
    },
    {
      "entropy": 5.40708589553833,
      "epoch": 1.0649492017416546,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893526695392694,
      "loss": 5.0112,
      "mean_token_accuracy": 0.20490798354148865,
      "num_tokens": 24359173.0,
      "step": 11740
    },
    {
      "entropy": 5.431822299957275,
      "epoch": 1.0654027576197387,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004893428316303185,
      "loss": 5.0139,
      "mean_token_accuracy": 0.20014695078134537,
      "num_tokens": 24369465.0,
      "step": 11745
    },
    {
      "entropy": 5.387581729888916,
      "epoch": 1.065856313497823,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004893329892887031,
      "loss": 5.0756,
      "mean_token_accuracy": 0.19761465936899186,
      "num_tokens": 24378899.0,
      "step": 11750
    },
    {
      "entropy": 5.4371435165405275,
      "epoch": 1.0663098693759072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893231425146271,
      "loss": 5.1867,
      "mean_token_accuracy": 0.18819210082292556,
      "num_tokens": 24388919.0,
      "step": 11755
    },
    {
      "entropy": 5.407600069046021,
      "epoch": 1.0667634252539913,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004893132913082937,
      "loss": 5.0453,
      "mean_token_accuracy": 0.19777143746614456,
      "num_tokens": 24399632.0,
      "step": 11760
    },
    {
      "entropy": 5.466056203842163,
      "epoch": 1.0672169811320755,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004893034356699069,
      "loss": 5.0721,
      "mean_token_accuracy": 0.19814738631248474,
      "num_tokens": 24410817.0,
      "step": 11765
    },
    {
      "entropy": 5.454527950286865,
      "epoch": 1.0676705370101596,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004892935755996705,
      "loss": 5.0955,
      "mean_token_accuracy": 0.19848891198635102,
      "num_tokens": 24421058.0,
      "step": 11770
    },
    {
      "entropy": 5.412892913818359,
      "epoch": 1.0681240928882438,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892837110977883,
      "loss": 5.0184,
      "mean_token_accuracy": 0.1985347628593445,
      "num_tokens": 24431522.0,
      "step": 11775
    },
    {
      "entropy": 5.420146083831787,
      "epoch": 1.068577648766328,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004892738421644644,
      "loss": 5.0563,
      "mean_token_accuracy": 0.19299584180116652,
      "num_tokens": 24441746.0,
      "step": 11780
    },
    {
      "entropy": 5.373950052261352,
      "epoch": 1.0690312046444121,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892639687999027,
      "loss": 4.9627,
      "mean_token_accuracy": 0.206170853972435,
      "num_tokens": 24451916.0,
      "step": 11785
    },
    {
      "entropy": 5.455459642410278,
      "epoch": 1.0694847605224964,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004892540910043076,
      "loss": 5.0879,
      "mean_token_accuracy": 0.19918798059225082,
      "num_tokens": 24461935.0,
      "step": 11790
    },
    {
      "entropy": 5.435303163528443,
      "epoch": 1.0699383164005805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892442087778833,
      "loss": 5.0002,
      "mean_token_accuracy": 0.19787367433309555,
      "num_tokens": 24472437.0,
      "step": 11795
    },
    {
      "entropy": 5.431952667236328,
      "epoch": 1.0703918722786647,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004892343221208342,
      "loss": 5.1029,
      "mean_token_accuracy": 0.19617521166801452,
      "num_tokens": 24481726.0,
      "step": 11800
    },
    {
      "entropy": 5.498169612884522,
      "epoch": 1.070845428156749,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004892244310333646,
      "loss": 5.184,
      "mean_token_accuracy": 0.19032939225435258,
      "num_tokens": 24492340.0,
      "step": 11805
    },
    {
      "entropy": 5.476718950271606,
      "epoch": 1.071298984034833,
      "grad_norm": 1.0,
      "learning_rate": 0.0004892145355156793,
      "loss": 5.05,
      "mean_token_accuracy": 0.19708126336336135,
      "num_tokens": 24502138.0,
      "step": 11810
    },
    {
      "entropy": 5.4611780643463135,
      "epoch": 1.0717525399129173,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004892046355679829,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19246915280818938,
      "num_tokens": 24512723.0,
      "step": 11815
    },
    {
      "entropy": 5.510536956787109,
      "epoch": 1.0722060957910013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004891947311904799,
      "loss": 5.1444,
      "mean_token_accuracy": 0.1910214379429817,
      "num_tokens": 24523766.0,
      "step": 11820
    },
    {
      "entropy": 5.4847740650177,
      "epoch": 1.0726596516690856,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004891848223833753,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19305452108383178,
      "num_tokens": 24534571.0,
      "step": 11825
    },
    {
      "entropy": 5.456302833557129,
      "epoch": 1.0731132075471699,
      "grad_norm": 1.015625,
      "learning_rate": 0.000489174909146874,
      "loss": 5.1088,
      "mean_token_accuracy": 0.19871936440467836,
      "num_tokens": 24545265.0,
      "step": 11830
    },
    {
      "entropy": 5.524204874038697,
      "epoch": 1.073566763425254,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004891649914811811,
      "loss": 5.1979,
      "mean_token_accuracy": 0.18889098316431047,
      "num_tokens": 24555524.0,
      "step": 11835
    },
    {
      "entropy": 5.410844326019287,
      "epoch": 1.0740203193033382,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891550693865014,
      "loss": 5.052,
      "mean_token_accuracy": 0.20847347378730774,
      "num_tokens": 24565420.0,
      "step": 11840
    },
    {
      "entropy": 5.468236207962036,
      "epoch": 1.0744738751814222,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004891451428630405,
      "loss": 5.0912,
      "mean_token_accuracy": 0.19379322081804276,
      "num_tokens": 24576794.0,
      "step": 11845
    },
    {
      "entropy": 5.4929125785827635,
      "epoch": 1.0749274310595065,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004891352119110033,
      "loss": 5.1068,
      "mean_token_accuracy": 0.19399791359901428,
      "num_tokens": 24586320.0,
      "step": 11850
    },
    {
      "entropy": 5.448437881469727,
      "epoch": 1.0753809869375908,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891252765305954,
      "loss": 5.0875,
      "mean_token_accuracy": 0.20086531192064286,
      "num_tokens": 24596524.0,
      "step": 11855
    },
    {
      "entropy": 5.4746064186096195,
      "epoch": 1.0758345428156748,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004891153367220222,
      "loss": 5.0173,
      "mean_token_accuracy": 0.2089245080947876,
      "num_tokens": 24605877.0,
      "step": 11860
    },
    {
      "entropy": 5.502732610702514,
      "epoch": 1.076288098693759,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004891053924854894,
      "loss": 5.2305,
      "mean_token_accuracy": 0.18587556779384612,
      "num_tokens": 24616719.0,
      "step": 11865
    },
    {
      "entropy": 5.43260178565979,
      "epoch": 1.0767416545718433,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890954438212024,
      "loss": 4.9843,
      "mean_token_accuracy": 0.20658824294805528,
      "num_tokens": 24626882.0,
      "step": 11870
    },
    {
      "entropy": 5.486688423156738,
      "epoch": 1.0771952104499274,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004890854907293671,
      "loss": 5.0936,
      "mean_token_accuracy": 0.1938313826918602,
      "num_tokens": 24637884.0,
      "step": 11875
    },
    {
      "entropy": 5.4502464771270756,
      "epoch": 1.0776487663280117,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004890755332101893,
      "loss": 5.0895,
      "mean_token_accuracy": 0.19480091631412505,
      "num_tokens": 24647389.0,
      "step": 11880
    },
    {
      "entropy": 5.448260498046875,
      "epoch": 1.0781023222060957,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004890655712638749,
      "loss": 5.0976,
      "mean_token_accuracy": 0.19201536923646928,
      "num_tokens": 24657595.0,
      "step": 11885
    },
    {
      "entropy": 5.448187303543091,
      "epoch": 1.07855587808418,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048905560489063,
      "loss": 5.015,
      "mean_token_accuracy": 0.19760590493679048,
      "num_tokens": 24667683.0,
      "step": 11890
    },
    {
      "entropy": 5.492201042175293,
      "epoch": 1.0790094339622642,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004890456340906605,
      "loss": 5.1742,
      "mean_token_accuracy": 0.1939292684197426,
      "num_tokens": 24678041.0,
      "step": 11895
    },
    {
      "entropy": 5.450703811645508,
      "epoch": 1.0794629898403483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890356588641729,
      "loss": 5.1517,
      "mean_token_accuracy": 0.19592384546995162,
      "num_tokens": 24688569.0,
      "step": 11900
    },
    {
      "entropy": 5.494032478332519,
      "epoch": 1.0799165457184325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004890256792113733,
      "loss": 5.1069,
      "mean_token_accuracy": 0.19837442487478257,
      "num_tokens": 24699913.0,
      "step": 11905
    },
    {
      "entropy": 5.5294829368591305,
      "epoch": 1.0803701015965168,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004890156951324681,
      "loss": 5.1323,
      "mean_token_accuracy": 0.1929474338889122,
      "num_tokens": 24710682.0,
      "step": 11910
    },
    {
      "entropy": 5.526803922653198,
      "epoch": 1.0808236574746009,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004890057066276636,
      "loss": 5.2058,
      "mean_token_accuracy": 0.18813426196575164,
      "num_tokens": 24721087.0,
      "step": 11915
    },
    {
      "entropy": 5.43285641670227,
      "epoch": 1.0812772133526851,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004889957136971668,
      "loss": 5.0328,
      "mean_token_accuracy": 0.20273849219083787,
      "num_tokens": 24731323.0,
      "step": 11920
    },
    {
      "entropy": 5.442595529556274,
      "epoch": 1.0817307692307692,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488985716341184,
      "loss": 5.097,
      "mean_token_accuracy": 0.19515804946422577,
      "num_tokens": 24742559.0,
      "step": 11925
    },
    {
      "entropy": 5.497393083572388,
      "epoch": 1.0821843251088534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004889757145599222,
      "loss": 5.1055,
      "mean_token_accuracy": 0.19816650450229645,
      "num_tokens": 24752307.0,
      "step": 11930
    },
    {
      "entropy": 5.361609315872192,
      "epoch": 1.0826378809869377,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000488965708353588,
      "loss": 5.0406,
      "mean_token_accuracy": 0.1924867585301399,
      "num_tokens": 24762792.0,
      "step": 11935
    },
    {
      "entropy": 5.508419227600098,
      "epoch": 1.0830914368650217,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004889556977223883,
      "loss": 5.1613,
      "mean_token_accuracy": 0.19042742848396302,
      "num_tokens": 24774377.0,
      "step": 11940
    },
    {
      "entropy": 5.484889316558838,
      "epoch": 1.083544992743106,
      "grad_norm": 0.875,
      "learning_rate": 0.0004889456826665304,
      "loss": 5.0786,
      "mean_token_accuracy": 0.195466311275959,
      "num_tokens": 24785919.0,
      "step": 11945
    },
    {
      "entropy": 5.465774250030518,
      "epoch": 1.08399854862119,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889356631862211,
      "loss": 5.1187,
      "mean_token_accuracy": 0.18686763644218446,
      "num_tokens": 24796594.0,
      "step": 11950
    },
    {
      "entropy": 5.453439331054687,
      "epoch": 1.0844521044992743,
      "grad_norm": 0.9375,
      "learning_rate": 0.000488925639281668,
      "loss": 5.1122,
      "mean_token_accuracy": 0.1970463752746582,
      "num_tokens": 24807322.0,
      "step": 11955
    },
    {
      "entropy": 5.438632583618164,
      "epoch": 1.0849056603773586,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004889156109530781,
      "loss": 4.999,
      "mean_token_accuracy": 0.1978117287158966,
      "num_tokens": 24817593.0,
      "step": 11960
    },
    {
      "entropy": 5.442756128311157,
      "epoch": 1.0853592162554426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889055782006588,
      "loss": 5.0073,
      "mean_token_accuracy": 0.2037275940179825,
      "num_tokens": 24827577.0,
      "step": 11965
    },
    {
      "entropy": 5.4089408874511715,
      "epoch": 1.085812772133527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004888955410246176,
      "loss": 5.0577,
      "mean_token_accuracy": 0.19720772951841353,
      "num_tokens": 24838148.0,
      "step": 11970
    },
    {
      "entropy": 5.425878286361694,
      "epoch": 1.086266328011611,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004888854994251621,
      "loss": 5.0117,
      "mean_token_accuracy": 0.19913591742515563,
      "num_tokens": 24848451.0,
      "step": 11975
    },
    {
      "entropy": 5.492869567871094,
      "epoch": 1.0867198838896952,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004888754534025,
      "loss": 5.1374,
      "mean_token_accuracy": 0.18826611191034318,
      "num_tokens": 24859619.0,
      "step": 11980
    },
    {
      "entropy": 5.474278116226197,
      "epoch": 1.0871734397677795,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004888654029568391,
      "loss": 5.1732,
      "mean_token_accuracy": 0.18208261132240294,
      "num_tokens": 24869863.0,
      "step": 11985
    },
    {
      "entropy": 5.450355195999146,
      "epoch": 1.0876269956458635,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004888553480883871,
      "loss": 5.1074,
      "mean_token_accuracy": 0.1931120127439499,
      "num_tokens": 24880524.0,
      "step": 11990
    },
    {
      "entropy": 5.3892419815063475,
      "epoch": 1.0880805515239478,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000488845288797352,
      "loss": 5.076,
      "mean_token_accuracy": 0.19475484490394593,
      "num_tokens": 24891514.0,
      "step": 11995
    },
    {
      "entropy": 5.428224277496338,
      "epoch": 1.0885341074020318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004888352250839418,
      "loss": 5.0443,
      "mean_token_accuracy": 0.1945518046617508,
      "num_tokens": 24900842.0,
      "step": 12000
    },
    {
      "epoch": 1.0885341074020318,
      "eval_entropy": 5.255034273674045,
      "eval_loss": 5.174998760223389,
      "eval_mean_token_accuracy": 0.20042603093608202,
      "eval_num_tokens": 24900842.0,
      "eval_runtime": 20.8388,
      "eval_samples_per_second": 1696.832,
      "eval_steps_per_second": 212.104,
      "step": 12000
    },
    {
      "entropy": 5.4489906311035154,
      "epoch": 1.088987663280116,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004888251569483647,
      "loss": 5.025,
      "mean_token_accuracy": 0.19985171556472778,
      "num_tokens": 24911845.0,
      "step": 12005
    },
    {
      "entropy": 5.368015575408935,
      "epoch": 1.0894412191582004,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004888150843908288,
      "loss": 5.0431,
      "mean_token_accuracy": 0.20289771109819413,
      "num_tokens": 24922211.0,
      "step": 12010
    },
    {
      "entropy": 5.479585027694702,
      "epoch": 1.0898947750362844,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004888050074115426,
      "loss": 5.1556,
      "mean_token_accuracy": 0.19275439232587815,
      "num_tokens": 24932129.0,
      "step": 12015
    },
    {
      "entropy": 5.521604299545288,
      "epoch": 1.0903483309143687,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004887949260107143,
      "loss": 5.1387,
      "mean_token_accuracy": 0.18956592082977294,
      "num_tokens": 24942869.0,
      "step": 12020
    },
    {
      "entropy": 5.414851093292237,
      "epoch": 1.0908018867924527,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004887848401885524,
      "loss": 5.08,
      "mean_token_accuracy": 0.19793568402528763,
      "num_tokens": 24952621.0,
      "step": 12025
    },
    {
      "entropy": 5.561692905426026,
      "epoch": 1.091255442670537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004887747499452656,
      "loss": 5.1966,
      "mean_token_accuracy": 0.18556403815746308,
      "num_tokens": 24963587.0,
      "step": 12030
    },
    {
      "entropy": 5.484190034866333,
      "epoch": 1.0917089985486212,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004887646552810624,
      "loss": 5.046,
      "mean_token_accuracy": 0.19763047248125076,
      "num_tokens": 24974122.0,
      "step": 12035
    },
    {
      "entropy": 5.438176679611206,
      "epoch": 1.0921625544267053,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004887545561961517,
      "loss": 5.1421,
      "mean_token_accuracy": 0.18937054723501207,
      "num_tokens": 24985145.0,
      "step": 12040
    },
    {
      "entropy": 5.410459756851196,
      "epoch": 1.0926161103047896,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004887444526907424,
      "loss": 5.0487,
      "mean_token_accuracy": 0.19937893897294998,
      "num_tokens": 24995511.0,
      "step": 12045
    },
    {
      "entropy": 5.481537866592407,
      "epoch": 1.0930696661828738,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004887343447650432,
      "loss": 5.0793,
      "mean_token_accuracy": 0.19209745675325393,
      "num_tokens": 25006193.0,
      "step": 12050
    },
    {
      "entropy": 5.504720258712768,
      "epoch": 1.0935232220609579,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004887242324192634,
      "loss": 5.1121,
      "mean_token_accuracy": 0.19623176604509354,
      "num_tokens": 25016444.0,
      "step": 12055
    },
    {
      "entropy": 5.4893440246582035,
      "epoch": 1.0939767779390421,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004887141156536119,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19245172739028932,
      "num_tokens": 25027898.0,
      "step": 12060
    },
    {
      "entropy": 5.439944267272949,
      "epoch": 1.0944303338171262,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004887039944682981,
      "loss": 5.0109,
      "mean_token_accuracy": 0.19628615528345109,
      "num_tokens": 25038201.0,
      "step": 12065
    },
    {
      "entropy": 5.535740566253662,
      "epoch": 1.0948838896952104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886938688635313,
      "loss": 5.2025,
      "mean_token_accuracy": 0.1854793146252632,
      "num_tokens": 25048466.0,
      "step": 12070
    },
    {
      "entropy": 5.569825315475464,
      "epoch": 1.0953374455732947,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004886837388395207,
      "loss": 5.1386,
      "mean_token_accuracy": 0.19498365968465806,
      "num_tokens": 25059064.0,
      "step": 12075
    },
    {
      "entropy": 5.544608736038208,
      "epoch": 1.0957910014513788,
      "grad_norm": 1.0625,
      "learning_rate": 0.000488673604396476,
      "loss": 5.229,
      "mean_token_accuracy": 0.18380790054798127,
      "num_tokens": 25068834.0,
      "step": 12080
    },
    {
      "entropy": 5.470860290527344,
      "epoch": 1.096244557329463,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886634655346067,
      "loss": 5.1643,
      "mean_token_accuracy": 0.18487847447395325,
      "num_tokens": 25080058.0,
      "step": 12085
    },
    {
      "entropy": 5.452997398376465,
      "epoch": 1.0966981132075473,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886533222541225,
      "loss": 5.0314,
      "mean_token_accuracy": 0.1977093443274498,
      "num_tokens": 25089751.0,
      "step": 12090
    },
    {
      "entropy": 5.490142726898194,
      "epoch": 1.0971516690856313,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004886431745552331,
      "loss": 5.1177,
      "mean_token_accuracy": 0.1926141932606697,
      "num_tokens": 25099055.0,
      "step": 12095
    },
    {
      "entropy": 5.540484094619751,
      "epoch": 1.0976052249637156,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004886330224381485,
      "loss": 5.1985,
      "mean_token_accuracy": 0.18552614599466324,
      "num_tokens": 25109793.0,
      "step": 12100
    },
    {
      "entropy": 5.512198829650879,
      "epoch": 1.0980587808417996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004886228659030785,
      "loss": 5.1145,
      "mean_token_accuracy": 0.19541315287351607,
      "num_tokens": 25119814.0,
      "step": 12105
    },
    {
      "entropy": 5.5101316452026365,
      "epoch": 1.098512336719884,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004886127049502331,
      "loss": 5.1705,
      "mean_token_accuracy": 0.19278408586978912,
      "num_tokens": 25130747.0,
      "step": 12110
    },
    {
      "entropy": 5.426028537750244,
      "epoch": 1.0989658925979682,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886025395798228,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19459046125411988,
      "num_tokens": 25142235.0,
      "step": 12115
    },
    {
      "entropy": 5.361623859405517,
      "epoch": 1.0994194484760522,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004885923697920573,
      "loss": 5.0037,
      "mean_token_accuracy": 0.20498927682638168,
      "num_tokens": 25152047.0,
      "step": 12120
    },
    {
      "entropy": 5.411035346984863,
      "epoch": 1.0998730043541365,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004885821955871472,
      "loss": 5.058,
      "mean_token_accuracy": 0.2009945407509804,
      "num_tokens": 25162678.0,
      "step": 12125
    },
    {
      "entropy": 5.496362113952637,
      "epoch": 1.1003265602322205,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885720169653028,
      "loss": 5.0533,
      "mean_token_accuracy": 0.19901553094387053,
      "num_tokens": 25172973.0,
      "step": 12130
    },
    {
      "entropy": 5.540875244140625,
      "epoch": 1.1007801161103048,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004885618339267348,
      "loss": 5.1815,
      "mean_token_accuracy": 0.18809981942176818,
      "num_tokens": 25183458.0,
      "step": 12135
    },
    {
      "entropy": 5.438296604156494,
      "epoch": 1.101233671988389,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885516464716537,
      "loss": 5.061,
      "mean_token_accuracy": 0.198656103014946,
      "num_tokens": 25193827.0,
      "step": 12140
    },
    {
      "entropy": 5.515053462982178,
      "epoch": 1.101687227866473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048854145460027,
      "loss": 5.0917,
      "mean_token_accuracy": 0.18708029985427857,
      "num_tokens": 25204448.0,
      "step": 12145
    },
    {
      "entropy": 5.481131553649902,
      "epoch": 1.1021407837445574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885312583127946,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2026984691619873,
      "num_tokens": 25215041.0,
      "step": 12150
    },
    {
      "entropy": 5.445676517486572,
      "epoch": 1.1025943396226414,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004885210576094385,
      "loss": 5.0978,
      "mean_token_accuracy": 0.19386008232831956,
      "num_tokens": 25225937.0,
      "step": 12155
    },
    {
      "entropy": 5.3993080139160154,
      "epoch": 1.1030478955007257,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885108524904124,
      "loss": 5.012,
      "mean_token_accuracy": 0.19879624843597413,
      "num_tokens": 25235270.0,
      "step": 12160
    },
    {
      "entropy": 5.419803476333618,
      "epoch": 1.10350145137881,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004885006429559275,
      "loss": 5.1018,
      "mean_token_accuracy": 0.18878085315227508,
      "num_tokens": 25247080.0,
      "step": 12165
    },
    {
      "entropy": 5.491727638244629,
      "epoch": 1.103955007256894,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004884904290061949,
      "loss": 5.0817,
      "mean_token_accuracy": 0.18722966760396959,
      "num_tokens": 25256718.0,
      "step": 12170
    },
    {
      "entropy": 5.499802446365356,
      "epoch": 1.1044085631349783,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004884802106414258,
      "loss": 5.2064,
      "mean_token_accuracy": 0.1875415787100792,
      "num_tokens": 25266632.0,
      "step": 12175
    },
    {
      "entropy": 5.468026447296142,
      "epoch": 1.1048621190130623,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004884699878618316,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20060146898031234,
      "num_tokens": 25275598.0,
      "step": 12180
    },
    {
      "entropy": 5.5242133140563965,
      "epoch": 1.1053156748911466,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004884597606676236,
      "loss": 5.1664,
      "mean_token_accuracy": 0.19410443753004075,
      "num_tokens": 25286598.0,
      "step": 12185
    },
    {
      "entropy": 5.47405595779419,
      "epoch": 1.1057692307692308,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004884495290590134,
      "loss": 5.0676,
      "mean_token_accuracy": 0.199165478348732,
      "num_tokens": 25296730.0,
      "step": 12190
    },
    {
      "entropy": 5.415151071548462,
      "epoch": 1.1062227866473149,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004884392930362124,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20322936922311782,
      "num_tokens": 25306326.0,
      "step": 12195
    },
    {
      "entropy": 5.477290248870849,
      "epoch": 1.1066763425253991,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004884290525994326,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19518473744392395,
      "num_tokens": 25316342.0,
      "step": 12200
    },
    {
      "entropy": 5.539191722869873,
      "epoch": 1.1071298984034832,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004884188077488855,
      "loss": 5.1426,
      "mean_token_accuracy": 0.18671423494815825,
      "num_tokens": 25326129.0,
      "step": 12205
    },
    {
      "entropy": 5.463911628723144,
      "epoch": 1.1075834542815675,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004884085584847831,
      "loss": 5.1031,
      "mean_token_accuracy": 0.1914559096097946,
      "num_tokens": 25337156.0,
      "step": 12210
    },
    {
      "entropy": 5.540527200698852,
      "epoch": 1.1080370101596517,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004883983048073375,
      "loss": 5.2233,
      "mean_token_accuracy": 0.18347180634737015,
      "num_tokens": 25347632.0,
      "step": 12215
    },
    {
      "entropy": 5.4638416290283205,
      "epoch": 1.1084905660377358,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004883880467167603,
      "loss": 4.9907,
      "mean_token_accuracy": 0.19963539987802506,
      "num_tokens": 25359186.0,
      "step": 12220
    },
    {
      "entropy": 5.502979707717896,
      "epoch": 1.10894412191582,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000488377784213264,
      "loss": 5.1043,
      "mean_token_accuracy": 0.19738257825374603,
      "num_tokens": 25368993.0,
      "step": 12225
    },
    {
      "entropy": 5.3920735836029055,
      "epoch": 1.1093976777939043,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004883675172970608,
      "loss": 5.0669,
      "mean_token_accuracy": 0.19466077238321305,
      "num_tokens": 25379880.0,
      "step": 12230
    },
    {
      "entropy": 5.495429229736328,
      "epoch": 1.1098512336719883,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000488357245968363,
      "loss": 5.2538,
      "mean_token_accuracy": 0.1847075939178467,
      "num_tokens": 25389942.0,
      "step": 12235
    },
    {
      "entropy": 5.468625593185425,
      "epoch": 1.1103047895500726,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004883469702273829,
      "loss": 5.0429,
      "mean_token_accuracy": 0.19933029413223266,
      "num_tokens": 25400836.0,
      "step": 12240
    },
    {
      "entropy": 5.4450860023498535,
      "epoch": 1.1107583454281567,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004883366900743331,
      "loss": 5.062,
      "mean_token_accuracy": 0.20043474137783052,
      "num_tokens": 25410502.0,
      "step": 12245
    },
    {
      "entropy": 5.436722755432129,
      "epoch": 1.111211901306241,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004883264055094262,
      "loss": 5.0162,
      "mean_token_accuracy": 0.2023913726210594,
      "num_tokens": 25420301.0,
      "step": 12250
    },
    {
      "entropy": 5.462346839904785,
      "epoch": 1.1116654571843252,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004883161165328749,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19105413258075715,
      "num_tokens": 25431107.0,
      "step": 12255
    },
    {
      "entropy": 5.468860864639282,
      "epoch": 1.1121190130624092,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488305823144892,
      "loss": 5.109,
      "mean_token_accuracy": 0.1985253646969795,
      "num_tokens": 25441256.0,
      "step": 12260
    },
    {
      "entropy": 5.512408590316772,
      "epoch": 1.1125725689404935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004882955253456902,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20018399506807327,
      "num_tokens": 25451627.0,
      "step": 12265
    },
    {
      "entropy": 5.519826793670655,
      "epoch": 1.1130261248185778,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004882852231354826,
      "loss": 5.1596,
      "mean_token_accuracy": 0.1922303855419159,
      "num_tokens": 25461273.0,
      "step": 12270
    },
    {
      "entropy": 5.394473886489868,
      "epoch": 1.1134796806966618,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004882749165144822,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1974688798189163,
      "num_tokens": 25471949.0,
      "step": 12275
    },
    {
      "entropy": 5.462115383148193,
      "epoch": 1.113933236574746,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004882646054829023,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19317834824323654,
      "num_tokens": 25482175.0,
      "step": 12280
    },
    {
      "entropy": 5.498352575302124,
      "epoch": 1.1143867924528301,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048825429004095597,
      "loss": 5.1434,
      "mean_token_accuracy": 0.19709102362394332,
      "num_tokens": 25491786.0,
      "step": 12285
    },
    {
      "entropy": 5.473578929901123,
      "epoch": 1.1148403483309144,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004882439701888566,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20382800549268723,
      "num_tokens": 25502103.0,
      "step": 12290
    },
    {
      "entropy": 5.438637828826904,
      "epoch": 1.1152939042089987,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048823364592681754,
      "loss": 5.0393,
      "mean_token_accuracy": 0.1983577087521553,
      "num_tokens": 25512530.0,
      "step": 12295
    },
    {
      "entropy": 5.4492512226104735,
      "epoch": 1.1157474600870827,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048822331725505234,
      "loss": 5.0464,
      "mean_token_accuracy": 0.19616839587688445,
      "num_tokens": 25522905.0,
      "step": 12300
    },
    {
      "entropy": 5.4275946617126465,
      "epoch": 1.116201015965167,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048821298417377456,
      "loss": 5.0121,
      "mean_token_accuracy": 0.2009132131934166,
      "num_tokens": 25532963.0,
      "step": 12305
    },
    {
      "entropy": 5.365885400772095,
      "epoch": 1.116654571843251,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488202646683198,
      "loss": 5.093,
      "mean_token_accuracy": 0.19705985486507416,
      "num_tokens": 25542583.0,
      "step": 12310
    },
    {
      "entropy": 5.445810461044312,
      "epoch": 1.1171081277213353,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004881923047835364,
      "loss": 5.0706,
      "mean_token_accuracy": 0.19825900644063948,
      "num_tokens": 25553367.0,
      "step": 12315
    },
    {
      "entropy": 5.599606609344482,
      "epoch": 1.1175616835994195,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004881819584750036,
      "loss": 5.1615,
      "mean_token_accuracy": 0.18648526966571807,
      "num_tokens": 25564823.0,
      "step": 12320
    },
    {
      "entropy": 5.49623966217041,
      "epoch": 1.1180152394775036,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048817160775781356,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19345640689134597,
      "num_tokens": 25574946.0,
      "step": 12325
    },
    {
      "entropy": 5.480826711654663,
      "epoch": 1.1184687953555879,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048816125263218023,
      "loss": 5.123,
      "mean_token_accuracy": 0.19235989600419998,
      "num_tokens": 25584779.0,
      "step": 12330
    },
    {
      "entropy": 5.5709977626800535,
      "epoch": 1.118922351233672,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048815089309831803,
      "loss": 5.1842,
      "mean_token_accuracy": 0.1899712488055229,
      "num_tokens": 25594725.0,
      "step": 12335
    },
    {
      "entropy": 5.492442321777344,
      "epoch": 1.1193759071117562,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000488140529156441,
      "loss": 5.1615,
      "mean_token_accuracy": 0.19133716225624084,
      "num_tokens": 25605499.0,
      "step": 12340
    },
    {
      "entropy": 5.456181573867798,
      "epoch": 1.1198294629898404,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004881301608067635,
      "loss": 5.1027,
      "mean_token_accuracy": 0.19733177721500397,
      "num_tokens": 25616443.0,
      "step": 12345
    },
    {
      "entropy": 5.474880027770996,
      "epoch": 1.1202830188679245,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048811978804949997,
      "loss": 5.0835,
      "mean_token_accuracy": 0.19235514104366302,
      "num_tokens": 25627762.0,
      "step": 12350
    },
    {
      "entropy": 5.548942899703979,
      "epoch": 1.1207365747460087,
      "grad_norm": 1.078125,
      "learning_rate": 0.000488109410884865,
      "loss": 5.1536,
      "mean_token_accuracy": 0.1956962913274765,
      "num_tokens": 25637904.0,
      "step": 12355
    },
    {
      "entropy": 5.486640548706054,
      "epoch": 1.1211901306240928,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048809902931307294,
      "loss": 5.0252,
      "mean_token_accuracy": 0.19654428511857985,
      "num_tokens": 25649058.0,
      "step": 12360
    },
    {
      "entropy": 5.459781503677368,
      "epoch": 1.121643686502177,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004880886433343388,
      "loss": 5.0976,
      "mean_token_accuracy": 0.1945563241839409,
      "num_tokens": 25659161.0,
      "step": 12365
    },
    {
      "entropy": 5.438920974731445,
      "epoch": 1.1220972423802613,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004880782529488771,
      "loss": 4.9989,
      "mean_token_accuracy": 0.20202931463718415,
      "num_tokens": 25669336.0,
      "step": 12370
    },
    {
      "entropy": 5.476479721069336,
      "epoch": 1.1225507982583454,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004880678581569029,
      "loss": 5.1366,
      "mean_token_accuracy": 0.19516809284687042,
      "num_tokens": 25680443.0,
      "step": 12375
    },
    {
      "entropy": 5.456211900711059,
      "epoch": 1.1230043541364296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048805745895863105,
      "loss": 5.0602,
      "mean_token_accuracy": 0.19789458215236663,
      "num_tokens": 25690202.0,
      "step": 12380
    },
    {
      "entropy": 5.44689302444458,
      "epoch": 1.1234579100145137,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004880470553542767,
      "loss": 5.117,
      "mean_token_accuracy": 0.19373457729816437,
      "num_tokens": 25699255.0,
      "step": 12385
    },
    {
      "entropy": 5.493829441070557,
      "epoch": 1.123911465892598,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048803664734405495,
      "loss": 5.1017,
      "mean_token_accuracy": 0.19037623703479767,
      "num_tokens": 25710714.0,
      "step": 12390
    },
    {
      "entropy": 5.485815620422363,
      "epoch": 1.1243650217706822,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000488026234928181,
      "loss": 5.0729,
      "mean_token_accuracy": 0.1951461464166641,
      "num_tokens": 25721722.0,
      "step": 12395
    },
    {
      "entropy": 5.467492341995239,
      "epoch": 1.1248185776487662,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004880158181068703,
      "loss": 5.0577,
      "mean_token_accuracy": 0.206204853951931,
      "num_tokens": 25732096.0,
      "step": 12400
    },
    {
      "entropy": 5.469112682342529,
      "epoch": 1.1252721335268505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048800539688033807,
      "loss": 5.1189,
      "mean_token_accuracy": 0.19479695409536363,
      "num_tokens": 25741499.0,
      "step": 12405
    },
    {
      "entropy": 5.50004940032959,
      "epoch": 1.1257256894049348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004879949712488001,
      "loss": 5.1899,
      "mean_token_accuracy": 0.19636364728212358,
      "num_tokens": 25751913.0,
      "step": 12410
    },
    {
      "entropy": 5.4780515193939205,
      "epoch": 1.1261792452830188,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048798454121247177,
      "loss": 5.1325,
      "mean_token_accuracy": 0.19104093164205552,
      "num_tokens": 25762602.0,
      "step": 12415
    },
    {
      "entropy": 5.4434812545776365,
      "epoch": 1.126632801161103,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004879741067715688,
      "loss": 5.0821,
      "mean_token_accuracy": 0.19571268558502197,
      "num_tokens": 25773872.0,
      "step": 12420
    },
    {
      "entropy": 5.521503114700318,
      "epoch": 1.1270863570391871,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048796366792630715,
      "loss": 5.0933,
      "mean_token_accuracy": 0.19417802542448043,
      "num_tokens": 25784723.0,
      "step": 12425
    },
    {
      "entropy": 5.4688825607299805,
      "epoch": 1.1275399129172714,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004879532246769025,
      "loss": 5.1012,
      "mean_token_accuracy": 0.19289073944091797,
      "num_tokens": 25794897.0,
      "step": 12430
    },
    {
      "entropy": 5.425449371337891,
      "epoch": 1.1279934687953557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004879427770235709,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20814288258552552,
      "num_tokens": 25804445.0,
      "step": 12435
    },
    {
      "entropy": 5.499363708496094,
      "epoch": 1.1284470246734397,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048793232496652846,
      "loss": 5.1645,
      "mean_token_accuracy": 0.18461741954088212,
      "num_tokens": 25815495.0,
      "step": 12440
    },
    {
      "entropy": 5.4840599536895756,
      "epoch": 1.128900580551524,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048792186850599124,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19891237020492553,
      "num_tokens": 25826120.0,
      "step": 12445
    },
    {
      "entropy": 5.423397397994995,
      "epoch": 1.1293541364296082,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004879114076421755,
      "loss": 5.0847,
      "mean_token_accuracy": 0.19845402240753174,
      "num_tokens": 25835926.0,
      "step": 12450
    },
    {
      "entropy": 5.442421865463257,
      "epoch": 1.1298076923076923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048790094237529763,
      "loss": 5.1257,
      "mean_token_accuracy": 0.19075507521629334,
      "num_tokens": 25846095.0,
      "step": 12455
    },
    {
      "entropy": 5.547850894927978,
      "epoch": 1.1302612481857766,
      "grad_norm": 1.03125,
      "learning_rate": 0.000487890472705574,
      "loss": 5.1221,
      "mean_token_accuracy": 0.1937321200966835,
      "num_tokens": 25858541.0,
      "step": 12460
    },
    {
      "entropy": 5.51123275756836,
      "epoch": 1.1307148040638606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048787999863322113,
      "loss": 5.0326,
      "mean_token_accuracy": 0.1969616174697876,
      "num_tokens": 25867664.0,
      "step": 12465
    },
    {
      "entropy": 5.458842515945435,
      "epoch": 1.1311683599419449,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004878695201584557,
      "loss": 5.1305,
      "mean_token_accuracy": 0.19804387539625168,
      "num_tokens": 25877300.0,
      "step": 12470
    },
    {
      "entropy": 5.440077924728394,
      "epoch": 1.1316219158200291,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004878590372814943,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1993800789117813,
      "num_tokens": 25888009.0,
      "step": 12475
    },
    {
      "entropy": 5.5107615947723385,
      "epoch": 1.1320754716981132,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048784855000255374,
      "loss": 5.139,
      "mean_token_accuracy": 0.18950151801109313,
      "num_tokens": 25897285.0,
      "step": 12480
    },
    {
      "entropy": 5.44799313545227,
      "epoch": 1.1325290275761974,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048783805832185097,
      "loss": 5.0295,
      "mean_token_accuracy": 0.1994224727153778,
      "num_tokens": 25907355.0,
      "step": 12485
    },
    {
      "entropy": 5.481715393066406,
      "epoch": 1.1329825834542815,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004878275622396029,
      "loss": 5.0931,
      "mean_token_accuracy": 0.20050271451473237,
      "num_tokens": 25918566.0,
      "step": 12490
    },
    {
      "entropy": 5.4791121006011965,
      "epoch": 1.1334361393323658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048781706175602665,
      "loss": 5.1519,
      "mean_token_accuracy": 0.1925746187567711,
      "num_tokens": 25929145.0,
      "step": 12495
    },
    {
      "entropy": 5.451447343826294,
      "epoch": 1.13388969521045,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048780655687133927,
      "loss": 5.0148,
      "mean_token_accuracy": 0.19808944463729858,
      "num_tokens": 25938163.0,
      "step": 12500
    },
    {
      "entropy": 5.449240875244141,
      "epoch": 1.134343251088534,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048779604758575806,
      "loss": 5.0501,
      "mean_token_accuracy": 0.19204022139310836,
      "num_tokens": 25948962.0,
      "step": 12505
    },
    {
      "entropy": 5.41950626373291,
      "epoch": 1.1347968069666183,
      "grad_norm": 1.0,
      "learning_rate": 0.0004877855338995004,
      "loss": 5.0298,
      "mean_token_accuracy": 0.195741006731987,
      "num_tokens": 25959653.0,
      "step": 12510
    },
    {
      "entropy": 5.537498331069946,
      "epoch": 1.1352503628447024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048777501581278367,
      "loss": 5.1117,
      "mean_token_accuracy": 0.1958885371685028,
      "num_tokens": 25970268.0,
      "step": 12515
    },
    {
      "entropy": 5.4535798072814945,
      "epoch": 1.1357039187227866,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004877644933258254,
      "loss": 5.0691,
      "mean_token_accuracy": 0.19766748696565628,
      "num_tokens": 25980321.0,
      "step": 12520
    },
    {
      "entropy": 5.394514036178589,
      "epoch": 1.136157474600871,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004877539664388432,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19097693562507628,
      "num_tokens": 25992466.0,
      "step": 12525
    },
    {
      "entropy": 5.459747552871704,
      "epoch": 1.136611030478955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048774343515205464,
      "loss": 5.054,
      "mean_token_accuracy": 0.2004119947552681,
      "num_tokens": 26003553.0,
      "step": 12530
    },
    {
      "entropy": 5.461809778213501,
      "epoch": 1.1370645863570392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004877328994656778,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19754336774349213,
      "num_tokens": 26013299.0,
      "step": 12535
    },
    {
      "entropy": 5.421274185180664,
      "epoch": 1.1375181422351233,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004877223593799303,
      "loss": 5.0922,
      "mean_token_accuracy": 0.1976127028465271,
      "num_tokens": 26022913.0,
      "step": 12540
    },
    {
      "entropy": 5.381703662872314,
      "epoch": 1.1379716981132075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004877118148950302,
      "loss": 5.0667,
      "mean_token_accuracy": 0.2003282517194748,
      "num_tokens": 26033191.0,
      "step": 12545
    },
    {
      "entropy": 5.607171106338501,
      "epoch": 1.1384252539912918,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048770126601119555,
      "loss": 5.2287,
      "mean_token_accuracy": 0.18697791695594787,
      "num_tokens": 26043517.0,
      "step": 12550
    },
    {
      "entropy": 5.564448261260987,
      "epoch": 1.1388788098693758,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048769071272864454,
      "loss": 5.1749,
      "mean_token_accuracy": 0.18821585327386856,
      "num_tokens": 26054173.0,
      "step": 12555
    },
    {
      "entropy": 5.46494140625,
      "epoch": 1.13933236574746,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048768015504759536,
      "loss": 5.1396,
      "mean_token_accuracy": 0.1895514503121376,
      "num_tokens": 26064929.0,
      "step": 12560
    },
    {
      "entropy": 5.430631017684936,
      "epoch": 1.1397859216255442,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004876695929682665,
      "loss": 5.051,
      "mean_token_accuracy": 0.20218482315540315,
      "num_tokens": 26074872.0,
      "step": 12565
    },
    {
      "entropy": 5.438392305374146,
      "epoch": 1.1402394775036284,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048765902649087605,
      "loss": 4.9689,
      "mean_token_accuracy": 0.20143624842166902,
      "num_tokens": 26086328.0,
      "step": 12570
    },
    {
      "entropy": 5.489910888671875,
      "epoch": 1.1406930333817127,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048764845561564285,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19734783172607423,
      "num_tokens": 26096477.0,
      "step": 12575
    },
    {
      "entropy": 5.443499040603638,
      "epoch": 1.1411465892597967,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048763788034278545,
      "loss": 5.0228,
      "mean_token_accuracy": 0.2042930543422699,
      "num_tokens": 26106183.0,
      "step": 12580
    },
    {
      "entropy": 5.452051162719727,
      "epoch": 1.141600145137881,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004876273006725225,
      "loss": 5.0863,
      "mean_token_accuracy": 0.1972722053527832,
      "num_tokens": 26117932.0,
      "step": 12585
    },
    {
      "entropy": 5.421517372131348,
      "epoch": 1.1420537010159653,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048761671660507266,
      "loss": 5.0487,
      "mean_token_accuracy": 0.20064830929040908,
      "num_tokens": 26128616.0,
      "step": 12590
    },
    {
      "entropy": 5.465033340454101,
      "epoch": 1.1425072568940493,
      "grad_norm": 0.9375,
      "learning_rate": 0.000487606128140655,
      "loss": 5.0636,
      "mean_token_accuracy": 0.194732204079628,
      "num_tokens": 26138758.0,
      "step": 12595
    },
    {
      "entropy": 5.431522226333618,
      "epoch": 1.1429608127721336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004875955352794885,
      "loss": 5.0686,
      "mean_token_accuracy": 0.19568051546812057,
      "num_tokens": 26149199.0,
      "step": 12600
    },
    {
      "entropy": 5.385068941116333,
      "epoch": 1.1434143686502176,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004875849380217921,
      "loss": 5.1498,
      "mean_token_accuracy": 0.19365084618330003,
      "num_tokens": 26160276.0,
      "step": 12605
    },
    {
      "entropy": 5.453707456588745,
      "epoch": 1.1438679245283019,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000487574336367785,
      "loss": 5.0343,
      "mean_token_accuracy": 0.20267068594694138,
      "num_tokens": 26169564.0,
      "step": 12610
    },
    {
      "entropy": 5.410078430175782,
      "epoch": 1.1443214804063861,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004875637303176865,
      "loss": 5.0018,
      "mean_token_accuracy": 0.20318748354911803,
      "num_tokens": 26179135.0,
      "step": 12615
    },
    {
      "entropy": 5.486770486831665,
      "epoch": 1.1447750362844702,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048755311987171585,
      "loss": 5.1419,
      "mean_token_accuracy": 0.19420150220394133,
      "num_tokens": 26190116.0,
      "step": 12620
    },
    {
      "entropy": 5.481070184707642,
      "epoch": 1.1452285921625545,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048754250503009244,
      "loss": 5.2052,
      "mean_token_accuracy": 0.19010931998491287,
      "num_tokens": 26200847.0,
      "step": 12625
    },
    {
      "entropy": 5.567831230163574,
      "epoch": 1.1456821480406387,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048753188579303596,
      "loss": 5.1633,
      "mean_token_accuracy": 0.18948815912008285,
      "num_tokens": 26211151.0,
      "step": 12630
    },
    {
      "entropy": 5.521000099182129,
      "epoch": 1.1461357039187228,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004875212621607659,
      "loss": 5.0759,
      "mean_token_accuracy": 0.2008240759372711,
      "num_tokens": 26221273.0,
      "step": 12635
    },
    {
      "entropy": 5.4748125076293945,
      "epoch": 1.146589259796807,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048751063413350203,
      "loss": 5.1833,
      "mean_token_accuracy": 0.19045755565166472,
      "num_tokens": 26231770.0,
      "step": 12640
    },
    {
      "entropy": 5.499720907211303,
      "epoch": 1.147042815674891,
      "grad_norm": 1.0625,
      "learning_rate": 0.000487500001711464,
      "loss": 5.1391,
      "mean_token_accuracy": 0.18866709470748902,
      "num_tokens": 26241386.0,
      "step": 12645
    },
    {
      "entropy": 5.602534341812134,
      "epoch": 1.1474963715529753,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004874893648948719,
      "loss": 5.2091,
      "mean_token_accuracy": 0.18955794870853424,
      "num_tokens": 26251991.0,
      "step": 12650
    },
    {
      "entropy": 5.6351649284362795,
      "epoch": 1.1479499274310596,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048747872368394546,
      "loss": 5.2513,
      "mean_token_accuracy": 0.19125227779150009,
      "num_tokens": 26262331.0,
      "step": 12655
    },
    {
      "entropy": 5.529968738555908,
      "epoch": 1.1484034833091437,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048746807807890496,
      "loss": 5.1308,
      "mean_token_accuracy": 0.19330211877822875,
      "num_tokens": 26272939.0,
      "step": 12660
    },
    {
      "entropy": 5.46607928276062,
      "epoch": 1.148857039187228,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048745742807997046,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2012448266148567,
      "num_tokens": 26282849.0,
      "step": 12665
    },
    {
      "entropy": 5.438662672042847,
      "epoch": 1.149310595065312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048744677368736216,
      "loss": 5.0912,
      "mean_token_accuracy": 0.20526786148548126,
      "num_tokens": 26292449.0,
      "step": 12670
    },
    {
      "entropy": 5.380973482131958,
      "epoch": 1.1497641509433962,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048743611490130044,
      "loss": 5.0434,
      "mean_token_accuracy": 0.19729205071926117,
      "num_tokens": 26302506.0,
      "step": 12675
    },
    {
      "entropy": 5.474449586868286,
      "epoch": 1.1502177068214805,
      "grad_norm": 1.0,
      "learning_rate": 0.00048742545172200576,
      "loss": 5.1151,
      "mean_token_accuracy": 0.20258864015340805,
      "num_tokens": 26312346.0,
      "step": 12680
    },
    {
      "entropy": 5.537198209762574,
      "epoch": 1.1506712626995645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004874147841496986,
      "loss": 5.108,
      "mean_token_accuracy": 0.19865725040435792,
      "num_tokens": 26323173.0,
      "step": 12685
    },
    {
      "entropy": 5.49872899055481,
      "epoch": 1.1511248185776488,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004874041121845996,
      "loss": 5.1664,
      "mean_token_accuracy": 0.18717677444219588,
      "num_tokens": 26334371.0,
      "step": 12690
    },
    {
      "entropy": 5.454615879058838,
      "epoch": 1.1515783744557329,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004873934358269294,
      "loss": 5.1039,
      "mean_token_accuracy": 0.19639908522367477,
      "num_tokens": 26344506.0,
      "step": 12695
    },
    {
      "entropy": 5.584876155853271,
      "epoch": 1.1520319303338171,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004873827550769088,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19069299846887589,
      "num_tokens": 26354534.0,
      "step": 12700
    },
    {
      "entropy": 5.472890901565552,
      "epoch": 1.1524854862119014,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004873720699347587,
      "loss": 5.0802,
      "mean_token_accuracy": 0.1938564732670784,
      "num_tokens": 26365682.0,
      "step": 12705
    },
    {
      "entropy": 5.532165241241455,
      "epoch": 1.1529390420899854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004873613804007001,
      "loss": 5.1995,
      "mean_token_accuracy": 0.19634334295988082,
      "num_tokens": 26376702.0,
      "step": 12710
    },
    {
      "entropy": 5.466576480865479,
      "epoch": 1.1533925979680697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004873506864749541,
      "loss": 5.0072,
      "mean_token_accuracy": 0.19673150926828384,
      "num_tokens": 26387154.0,
      "step": 12715
    },
    {
      "entropy": 5.5687531471252445,
      "epoch": 1.1538461538461537,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004873399881577417,
      "loss": 5.1791,
      "mean_token_accuracy": 0.18616852909326553,
      "num_tokens": 26397883.0,
      "step": 12720
    },
    {
      "entropy": 5.3821008682250975,
      "epoch": 1.154299709724238,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004873292854492842,
      "loss": 5.016,
      "mean_token_accuracy": 0.20302650928497315,
      "num_tokens": 26407613.0,
      "step": 12725
    },
    {
      "entropy": 5.341051197052002,
      "epoch": 1.1547532656023223,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004873185783498031,
      "loss": 5.0092,
      "mean_token_accuracy": 0.19885518699884414,
      "num_tokens": 26417221.0,
      "step": 12730
    },
    {
      "entropy": 5.447017526626587,
      "epoch": 1.1552068214804063,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004873078668595197,
      "loss": 5.0995,
      "mean_token_accuracy": 0.19121682792901992,
      "num_tokens": 26427976.0,
      "step": 12735
    },
    {
      "entropy": 5.496642923355102,
      "epoch": 1.1556603773584906,
      "grad_norm": 1.0,
      "learning_rate": 0.00048729715097865534,
      "loss": 5.1308,
      "mean_token_accuracy": 0.19671087712049484,
      "num_tokens": 26439213.0,
      "step": 12740
    },
    {
      "entropy": 5.446049642562866,
      "epoch": 1.1561139332365746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004872864307074319,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20202318876981734,
      "num_tokens": 26449763.0,
      "step": 12745
    },
    {
      "entropy": 5.519284391403199,
      "epoch": 1.156567489114659,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048727570604607096,
      "loss": 5.1338,
      "mean_token_accuracy": 0.19712906181812287,
      "num_tokens": 26460133.0,
      "step": 12750
    },
    {
      "entropy": 5.514515590667725,
      "epoch": 1.1570210449927432,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048726497699479437,
      "loss": 5.095,
      "mean_token_accuracy": 0.19766767770051957,
      "num_tokens": 26470164.0,
      "step": 12755
    },
    {
      "entropy": 5.455454349517822,
      "epoch": 1.1574746008708272,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000487254243553824,
      "loss": 5.1113,
      "mean_token_accuracy": 0.19574321955442428,
      "num_tokens": 26480314.0,
      "step": 12760
    },
    {
      "entropy": 5.411750841140747,
      "epoch": 1.1579281567489115,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004872435057233817,
      "loss": 5.0544,
      "mean_token_accuracy": 0.198037189245224,
      "num_tokens": 26491184.0,
      "step": 12765
    },
    {
      "entropy": 5.450663471221924,
      "epoch": 1.1583817126269957,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048723276350368957,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20026253461837767,
      "num_tokens": 26501432.0,
      "step": 12770
    },
    {
      "entropy": 5.466466188430786,
      "epoch": 1.1588352685050798,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048722201689496984,
      "loss": 5.0924,
      "mean_token_accuracy": 0.1962462067604065,
      "num_tokens": 26510379.0,
      "step": 12775
    },
    {
      "entropy": 5.437568092346192,
      "epoch": 1.159288824383164,
      "grad_norm": 1.0,
      "learning_rate": 0.0004872112658974447,
      "loss": 5.0774,
      "mean_token_accuracy": 0.20696530789136885,
      "num_tokens": 26519454.0,
      "step": 12780
    },
    {
      "entropy": 5.414034652709961,
      "epoch": 1.159742380261248,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048720051051133657,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20221693068742752,
      "num_tokens": 26531325.0,
      "step": 12785
    },
    {
      "entropy": 5.438589096069336,
      "epoch": 1.1601959361393324,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004871897507368678,
      "loss": 5.0816,
      "mean_token_accuracy": 0.1968888595700264,
      "num_tokens": 26541510.0,
      "step": 12790
    },
    {
      "entropy": 5.505600023269653,
      "epoch": 1.1606494920174166,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048717898657426093,
      "loss": 5.276,
      "mean_token_accuracy": 0.18606994152069092,
      "num_tokens": 26553400.0,
      "step": 12795
    },
    {
      "entropy": 5.515153980255127,
      "epoch": 1.1611030478955007,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048716821802373856,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19739335477352143,
      "num_tokens": 26564340.0,
      "step": 12800
    },
    {
      "entropy": 5.503932476043701,
      "epoch": 1.161556603773585,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048715744508552334,
      "loss": 5.0786,
      "mean_token_accuracy": 0.1971112832427025,
      "num_tokens": 26573615.0,
      "step": 12805
    },
    {
      "entropy": 5.499038791656494,
      "epoch": 1.1620101596516692,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004871466677598381,
      "loss": 5.2099,
      "mean_token_accuracy": 0.18656449317932128,
      "num_tokens": 26584149.0,
      "step": 12810
    },
    {
      "entropy": 5.446244955062866,
      "epoch": 1.1624637155297532,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004871358860469057,
      "loss": 5.132,
      "mean_token_accuracy": 0.18631608486175538,
      "num_tokens": 26595064.0,
      "step": 12815
    },
    {
      "entropy": 5.452750778198242,
      "epoch": 1.1629172714078375,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048712509994694903,
      "loss": 5.1086,
      "mean_token_accuracy": 0.19741549342870712,
      "num_tokens": 26605636.0,
      "step": 12820
    },
    {
      "entropy": 5.553749752044678,
      "epoch": 1.1633708272859216,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004871143094601914,
      "loss": 5.2182,
      "mean_token_accuracy": 0.18952305167913436,
      "num_tokens": 26616065.0,
      "step": 12825
    },
    {
      "entropy": 5.554434728622437,
      "epoch": 1.1638243831640058,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004871035145868557,
      "loss": 5.1269,
      "mean_token_accuracy": 0.19301836639642717,
      "num_tokens": 26627246.0,
      "step": 12830
    },
    {
      "entropy": 5.463024234771728,
      "epoch": 1.16427793904209,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048709271532716524,
      "loss": 5.1245,
      "mean_token_accuracy": 0.19043943732976915,
      "num_tokens": 26637630.0,
      "step": 12835
    },
    {
      "entropy": 5.514437246322632,
      "epoch": 1.1647314949201741,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004870819116813434,
      "loss": 5.163,
      "mean_token_accuracy": 0.1924004301428795,
      "num_tokens": 26648683.0,
      "step": 12840
    },
    {
      "entropy": 5.532897567749023,
      "epoch": 1.1651850507982584,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048707110364961363,
      "loss": 5.1197,
      "mean_token_accuracy": 0.19806863516569137,
      "num_tokens": 26659573.0,
      "step": 12845
    },
    {
      "entropy": 5.439857530593872,
      "epoch": 1.1656386066763424,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048706029123219937,
      "loss": 5.0343,
      "mean_token_accuracy": 0.20120469629764556,
      "num_tokens": 26669384.0,
      "step": 12850
    },
    {
      "entropy": 5.511091804504394,
      "epoch": 1.1660921625544267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048704947442932426,
      "loss": 5.1868,
      "mean_token_accuracy": 0.1870729684829712,
      "num_tokens": 26679721.0,
      "step": 12855
    },
    {
      "entropy": 5.5536027431488035,
      "epoch": 1.166545718432511,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048703865324121196,
      "loss": 5.0409,
      "mean_token_accuracy": 0.2025377035140991,
      "num_tokens": 26689923.0,
      "step": 12860
    },
    {
      "entropy": 5.392845726013183,
      "epoch": 1.166999274310595,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048702782766808637,
      "loss": 5.059,
      "mean_token_accuracy": 0.19589492827653884,
      "num_tokens": 26699953.0,
      "step": 12865
    },
    {
      "entropy": 5.492673873901367,
      "epoch": 1.1674528301886793,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004870169977101711,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18201562464237214,
      "num_tokens": 26712099.0,
      "step": 12870
    },
    {
      "entropy": 5.562848711013794,
      "epoch": 1.1679063860667633,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048700616336769037,
      "loss": 5.1057,
      "mean_token_accuracy": 0.19736160188913346,
      "num_tokens": 26722334.0,
      "step": 12875
    },
    {
      "entropy": 5.480432987213135,
      "epoch": 1.1683599419448476,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048699532464086817,
      "loss": 5.0638,
      "mean_token_accuracy": 0.19508631229400636,
      "num_tokens": 26732470.0,
      "step": 12880
    },
    {
      "entropy": 5.472351932525635,
      "epoch": 1.1688134978229319,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048698448152992865,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19666123986244202,
      "num_tokens": 26742015.0,
      "step": 12885
    },
    {
      "entropy": 5.415881586074829,
      "epoch": 1.169267053701016,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048697363403509605,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19333956092596055,
      "num_tokens": 26753231.0,
      "step": 12890
    },
    {
      "entropy": 5.477305698394775,
      "epoch": 1.1697206095791002,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004869627821565946,
      "loss": 5.0836,
      "mean_token_accuracy": 0.19506437778472902,
      "num_tokens": 26763895.0,
      "step": 12895
    },
    {
      "entropy": 5.524114513397217,
      "epoch": 1.1701741654571842,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004869519258946489,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19316886812448503,
      "num_tokens": 26774208.0,
      "step": 12900
    },
    {
      "entropy": 5.472482967376709,
      "epoch": 1.1706277213352685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048694106524948324,
      "loss": 5.1097,
      "mean_token_accuracy": 0.19007119983434678,
      "num_tokens": 26785554.0,
      "step": 12905
    },
    {
      "entropy": 5.507984304428101,
      "epoch": 1.1710812772133528,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048693020022132243,
      "loss": 5.0287,
      "mean_token_accuracy": 0.19974969327449799,
      "num_tokens": 26795488.0,
      "step": 12910
    },
    {
      "entropy": 5.523132848739624,
      "epoch": 1.1715348330914368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000486919330810391,
      "loss": 5.1786,
      "mean_token_accuracy": 0.18989856988191606,
      "num_tokens": 26806801.0,
      "step": 12915
    },
    {
      "entropy": 5.499798250198364,
      "epoch": 1.171988388969521,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004869084570169139,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1985375016927719,
      "num_tokens": 26816869.0,
      "step": 12920
    },
    {
      "entropy": 5.437505578994751,
      "epoch": 1.172441944847605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048689757884111585,
      "loss": 5.0449,
      "mean_token_accuracy": 0.19810945689678192,
      "num_tokens": 26827003.0,
      "step": 12925
    },
    {
      "entropy": 5.412903690338135,
      "epoch": 1.1728955007256894,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048688669628322185,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20033528953790664,
      "num_tokens": 26837065.0,
      "step": 12930
    },
    {
      "entropy": 5.415971422195435,
      "epoch": 1.1733490566037736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000486875809343457,
      "loss": 4.9935,
      "mean_token_accuracy": 0.19860540628433226,
      "num_tokens": 26847487.0,
      "step": 12935
    },
    {
      "entropy": 5.4496697902679445,
      "epoch": 1.1738026124818577,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048686491802204645,
      "loss": 5.0589,
      "mean_token_accuracy": 0.2041519746184349,
      "num_tokens": 26858407.0,
      "step": 12940
    },
    {
      "entropy": 5.50268669128418,
      "epoch": 1.174256168359942,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004868540223192154,
      "loss": 5.1209,
      "mean_token_accuracy": 0.19359382688999177,
      "num_tokens": 26868469.0,
      "step": 12945
    },
    {
      "entropy": 5.496904468536377,
      "epoch": 1.1747097242380262,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004868431222351892,
      "loss": 5.1311,
      "mean_token_accuracy": 0.19695413410663604,
      "num_tokens": 26879580.0,
      "step": 12950
    },
    {
      "entropy": 5.48253755569458,
      "epoch": 1.1751632801161103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048683221777019313,
      "loss": 5.0919,
      "mean_token_accuracy": 0.20048645436763762,
      "num_tokens": 26889513.0,
      "step": 12955
    },
    {
      "entropy": 5.4880575180053714,
      "epoch": 1.1756168359941945,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048682130892445293,
      "loss": 5.1826,
      "mean_token_accuracy": 0.1907653406262398,
      "num_tokens": 26900792.0,
      "step": 12960
    },
    {
      "entropy": 5.44629693031311,
      "epoch": 1.1760703918722786,
      "grad_norm": 1.03125,
      "learning_rate": 0.000486810395698194,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20455744862556458,
      "num_tokens": 26911298.0,
      "step": 12965
    },
    {
      "entropy": 5.521710538864136,
      "epoch": 1.1765239477503628,
      "grad_norm": 1.0,
      "learning_rate": 0.00048679947809164223,
      "loss": 5.1252,
      "mean_token_accuracy": 0.19028752148151398,
      "num_tokens": 26921545.0,
      "step": 12970
    },
    {
      "entropy": 5.504205179214478,
      "epoch": 1.176977503628447,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004867885561050232,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20825183987617493,
      "num_tokens": 26932140.0,
      "step": 12975
    },
    {
      "entropy": 5.494733381271362,
      "epoch": 1.1774310595065312,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004867776297385629,
      "loss": 5.1252,
      "mean_token_accuracy": 0.1926521986722946,
      "num_tokens": 26942634.0,
      "step": 12980
    },
    {
      "entropy": 5.4838728427886965,
      "epoch": 1.1778846153846154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048676669899248723,
      "loss": 5.0897,
      "mean_token_accuracy": 0.1977209985256195,
      "num_tokens": 26952758.0,
      "step": 12985
    },
    {
      "entropy": 5.42335000038147,
      "epoch": 1.1783381712626997,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004867557638670222,
      "loss": 5.0317,
      "mean_token_accuracy": 0.20029369443655015,
      "num_tokens": 26963439.0,
      "step": 12990
    },
    {
      "entropy": 5.463872289657592,
      "epoch": 1.1787917271407837,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004867448243623941,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19276442527770996,
      "num_tokens": 26974331.0,
      "step": 12995
    },
    {
      "entropy": 5.4251199722290036,
      "epoch": 1.179245283018868,
      "grad_norm": 1.078125,
      "learning_rate": 0.000486733880478829,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20203347504138947,
      "num_tokens": 26983416.0,
      "step": 13000
    },
    {
      "entropy": 5.454279279708862,
      "epoch": 1.179698838896952,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048672293221655333,
      "loss": 5.0946,
      "mean_token_accuracy": 0.1995358645915985,
      "num_tokens": 26994176.0,
      "step": 13005
    },
    {
      "entropy": 5.443188714981079,
      "epoch": 1.1801523947750363,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867119795757935,
      "loss": 5.093,
      "mean_token_accuracy": 0.19391274601221084,
      "num_tokens": 27003295.0,
      "step": 13010
    },
    {
      "entropy": 5.487621116638183,
      "epoch": 1.1806059506531206,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000486701022556776,
      "loss": 5.0536,
      "mean_token_accuracy": 0.19664328843355178,
      "num_tokens": 27014018.0,
      "step": 13015
    },
    {
      "entropy": 5.482475519180298,
      "epoch": 1.1810595065312046,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004866900611597273,
      "loss": 5.0813,
      "mean_token_accuracy": 0.19794545322656631,
      "num_tokens": 27024706.0,
      "step": 13020
    },
    {
      "entropy": 5.491993808746338,
      "epoch": 1.1815130624092889,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048667909538487427,
      "loss": 5.0877,
      "mean_token_accuracy": 0.20044560730457306,
      "num_tokens": 27035801.0,
      "step": 13025
    },
    {
      "entropy": 5.386587238311767,
      "epoch": 1.181966618287373,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004866681252324436,
      "loss": 4.969,
      "mean_token_accuracy": 0.20623714625835418,
      "num_tokens": 27046596.0,
      "step": 13030
    },
    {
      "entropy": 5.445072841644287,
      "epoch": 1.1824201741654572,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048665715070266214,
      "loss": 5.0696,
      "mean_token_accuracy": 0.1931762620806694,
      "num_tokens": 27057057.0,
      "step": 13035
    },
    {
      "entropy": 5.394736051559448,
      "epoch": 1.1828737300435415,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048664617179575685,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20542708784341812,
      "num_tokens": 27067947.0,
      "step": 13040
    },
    {
      "entropy": 5.499419689178467,
      "epoch": 1.1833272859216255,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004866351885119548,
      "loss": 5.0745,
      "mean_token_accuracy": 0.19452230632305145,
      "num_tokens": 27077503.0,
      "step": 13045
    },
    {
      "entropy": 5.47869291305542,
      "epoch": 1.1837808417997098,
      "grad_norm": 1.0,
      "learning_rate": 0.0004866242008514832,
      "loss": 5.1851,
      "mean_token_accuracy": 0.1935773551464081,
      "num_tokens": 27088298.0,
      "step": 13050
    },
    {
      "entropy": 5.44662823677063,
      "epoch": 1.1842343976777938,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048661320881456904,
      "loss": 5.0567,
      "mean_token_accuracy": 0.19659170955419542,
      "num_tokens": 27097841.0,
      "step": 13055
    },
    {
      "entropy": 5.3786825180053714,
      "epoch": 1.184687953555878,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004866022124014399,
      "loss": 5.0059,
      "mean_token_accuracy": 0.2037520781159401,
      "num_tokens": 27108263.0,
      "step": 13060
    },
    {
      "entropy": 5.43059778213501,
      "epoch": 1.1851415094339623,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048659121161232313,
      "loss": 5.0636,
      "mean_token_accuracy": 0.1974889948964119,
      "num_tokens": 27119339.0,
      "step": 13065
    },
    {
      "entropy": 5.508020401000977,
      "epoch": 1.1855950653120464,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004865802064474461,
      "loss": 5.1138,
      "mean_token_accuracy": 0.19474893063306808,
      "num_tokens": 27130393.0,
      "step": 13070
    },
    {
      "entropy": 5.490651798248291,
      "epoch": 1.1860486211901307,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048656919690703653,
      "loss": 5.108,
      "mean_token_accuracy": 0.18990262150764464,
      "num_tokens": 27141156.0,
      "step": 13075
    },
    {
      "entropy": 5.511962604522705,
      "epoch": 1.1865021770682147,
      "grad_norm": 1.0,
      "learning_rate": 0.000486558182991322,
      "loss": 5.1341,
      "mean_token_accuracy": 0.19704943150281906,
      "num_tokens": 27151630.0,
      "step": 13080
    },
    {
      "entropy": 5.4343561172485355,
      "epoch": 1.186955732946299,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004865471647005304,
      "loss": 5.0101,
      "mean_token_accuracy": 0.19913599342107774,
      "num_tokens": 27160734.0,
      "step": 13085
    },
    {
      "entropy": 5.443252897262573,
      "epoch": 1.1874092888243832,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048653614203488947,
      "loss": 5.1014,
      "mean_token_accuracy": 0.19857311248779297,
      "num_tokens": 27171584.0,
      "step": 13090
    },
    {
      "entropy": 5.445315742492676,
      "epoch": 1.1878628447024673,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004865251149946273,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20458347499370574,
      "num_tokens": 27180817.0,
      "step": 13095
    },
    {
      "entropy": 5.498758268356323,
      "epoch": 1.1883164005805515,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048651408357997173,
      "loss": 5.0997,
      "mean_token_accuracy": 0.19201551973819733,
      "num_tokens": 27190363.0,
      "step": 13100
    },
    {
      "entropy": 5.456967306137085,
      "epoch": 1.1887699564586356,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048650304779115114,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20326564013957976,
      "num_tokens": 27202181.0,
      "step": 13105
    },
    {
      "entropy": 5.41337947845459,
      "epoch": 1.1892235123367199,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004864920076283935,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19548821747303008,
      "num_tokens": 27212514.0,
      "step": 13110
    },
    {
      "entropy": 5.52031717300415,
      "epoch": 1.1896770682148041,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004864809630919273,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20517010539770125,
      "num_tokens": 27223352.0,
      "step": 13115
    },
    {
      "entropy": 5.459183692932129,
      "epoch": 1.1901306240928882,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004864699141819809,
      "loss": 5.0186,
      "mean_token_accuracy": 0.20343892574310302,
      "num_tokens": 27232643.0,
      "step": 13120
    },
    {
      "entropy": 5.368879985809326,
      "epoch": 1.1905841799709724,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048645886089878276,
      "loss": 5.0647,
      "mean_token_accuracy": 0.19952337443828583,
      "num_tokens": 27243793.0,
      "step": 13125
    },
    {
      "entropy": 5.4699231624603275,
      "epoch": 1.1910377358490567,
      "grad_norm": 1.0,
      "learning_rate": 0.00048644780324256153,
      "loss": 5.0822,
      "mean_token_accuracy": 0.19614837914705277,
      "num_tokens": 27254554.0,
      "step": 13130
    },
    {
      "entropy": 5.496552801132202,
      "epoch": 1.1914912917271407,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004864367412135458,
      "loss": 5.1303,
      "mean_token_accuracy": 0.19582335501909257,
      "num_tokens": 27265842.0,
      "step": 13135
    },
    {
      "entropy": 5.496447563171387,
      "epoch": 1.191944847605225,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048642567481196446,
      "loss": 5.1131,
      "mean_token_accuracy": 0.19201270192861558,
      "num_tokens": 27277617.0,
      "step": 13140
    },
    {
      "entropy": 5.461776828765869,
      "epoch": 1.192398403483309,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004864146040380462,
      "loss": 5.0622,
      "mean_token_accuracy": 0.1964070349931717,
      "num_tokens": 27287450.0,
      "step": 13145
    },
    {
      "entropy": 5.460700607299804,
      "epoch": 1.1928519593613933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004864035288920202,
      "loss": 5.0682,
      "mean_token_accuracy": 0.20016777515411377,
      "num_tokens": 27297010.0,
      "step": 13150
    },
    {
      "entropy": 5.455751180648804,
      "epoch": 1.1933055152394776,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048639244937411516,
      "loss": 5.1094,
      "mean_token_accuracy": 0.19363460689783096,
      "num_tokens": 27306874.0,
      "step": 13155
    },
    {
      "entropy": 5.536878776550293,
      "epoch": 1.1937590711175616,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048638136548456045,
      "loss": 5.1071,
      "mean_token_accuracy": 0.1993100017309189,
      "num_tokens": 27318762.0,
      "step": 13160
    },
    {
      "entropy": 5.464996767044068,
      "epoch": 1.194212626995646,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048637027722358536,
      "loss": 4.9826,
      "mean_token_accuracy": 0.20638133138418197,
      "num_tokens": 27327231.0,
      "step": 13165
    },
    {
      "entropy": 5.395813655853272,
      "epoch": 1.1946661828737302,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048635918459141893,
      "loss": 5.0149,
      "mean_token_accuracy": 0.19718607515096664,
      "num_tokens": 27338517.0,
      "step": 13170
    },
    {
      "entropy": 5.485543155670166,
      "epoch": 1.1951197387518142,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048634808758829073,
      "loss": 5.0998,
      "mean_token_accuracy": 0.19311325401067733,
      "num_tokens": 27348861.0,
      "step": 13175
    },
    {
      "entropy": 5.473777770996094,
      "epoch": 1.1955732946298985,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048633698621443024,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20883545130491257,
      "num_tokens": 27358610.0,
      "step": 13180
    },
    {
      "entropy": 5.458179235458374,
      "epoch": 1.1960268505079825,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000486325880470067,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19589586257934571,
      "num_tokens": 27369707.0,
      "step": 13185
    },
    {
      "entropy": 5.470612907409668,
      "epoch": 1.1964804063860668,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004863147703554307,
      "loss": 5.0763,
      "mean_token_accuracy": 0.19433997869491576,
      "num_tokens": 27381030.0,
      "step": 13190
    },
    {
      "entropy": 5.499072742462158,
      "epoch": 1.196933962264151,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048630365587075115,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19215388000011444,
      "num_tokens": 27390866.0,
      "step": 13195
    },
    {
      "entropy": 5.472330141067505,
      "epoch": 1.197387518142235,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048629253701625816,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20051812678575515,
      "num_tokens": 27400724.0,
      "step": 13200
    },
    {
      "entropy": 5.4769697189331055,
      "epoch": 1.1978410740203194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004862814137921816,
      "loss": 5.0573,
      "mean_token_accuracy": 0.19676778018474578,
      "num_tokens": 27411212.0,
      "step": 13205
    },
    {
      "entropy": 5.438337230682373,
      "epoch": 1.1982946298984034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048627028619875157,
      "loss": 5.0807,
      "mean_token_accuracy": 0.19674912691116334,
      "num_tokens": 27421279.0,
      "step": 13210
    },
    {
      "entropy": 5.466053295135498,
      "epoch": 1.1987481857764877,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004862591542361982,
      "loss": 5.0637,
      "mean_token_accuracy": 0.19462242722511292,
      "num_tokens": 27432520.0,
      "step": 13215
    },
    {
      "entropy": 5.410159397125244,
      "epoch": 1.199201741654572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004862480179047516,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20343201607465744,
      "num_tokens": 27442246.0,
      "step": 13220
    },
    {
      "entropy": 5.463499593734741,
      "epoch": 1.199655297532656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004862368772046423,
      "loss": 5.0684,
      "mean_token_accuracy": 0.195850670337677,
      "num_tokens": 27452134.0,
      "step": 13225
    },
    {
      "entropy": 5.512796545028687,
      "epoch": 1.2001088534107403,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048622573213610057,
      "loss": 5.1497,
      "mean_token_accuracy": 0.1871631845831871,
      "num_tokens": 27462253.0,
      "step": 13230
    },
    {
      "entropy": 5.477445459365844,
      "epoch": 1.2005624092888243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048621458269935676,
      "loss": 5.047,
      "mean_token_accuracy": 0.1999499797821045,
      "num_tokens": 27473131.0,
      "step": 13235
    },
    {
      "entropy": 5.465921449661255,
      "epoch": 1.2010159651669086,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048620342889464153,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20435721725225447,
      "num_tokens": 27483683.0,
      "step": 13240
    },
    {
      "entropy": 5.383798551559448,
      "epoch": 1.2014695210449928,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048619227072218565,
      "loss": 5.055,
      "mean_token_accuracy": 0.19200326055288314,
      "num_tokens": 27493679.0,
      "step": 13245
    },
    {
      "entropy": 5.422492218017578,
      "epoch": 1.2019230769230769,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004861811081822198,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20524911731481552,
      "num_tokens": 27503897.0,
      "step": 13250
    },
    {
      "entropy": 5.520388746261597,
      "epoch": 1.2023766328011611,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004861699412749748,
      "loss": 5.1464,
      "mean_token_accuracy": 0.19123760461807252,
      "num_tokens": 27514560.0,
      "step": 13255
    },
    {
      "entropy": 5.461540365219117,
      "epoch": 1.2028301886792452,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004861587700006816,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1977246880531311,
      "num_tokens": 27525594.0,
      "step": 13260
    },
    {
      "entropy": 5.387940454483032,
      "epoch": 1.2032837445573294,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048614759435957114,
      "loss": 4.991,
      "mean_token_accuracy": 0.2023061916232109,
      "num_tokens": 27535676.0,
      "step": 13265
    },
    {
      "entropy": 5.392179822921753,
      "epoch": 1.2037373004354137,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004861364143518747,
      "loss": 5.0217,
      "mean_token_accuracy": 0.19669488519430162,
      "num_tokens": 27545520.0,
      "step": 13270
    },
    {
      "entropy": 5.527988004684448,
      "epoch": 1.2041908563134978,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004861252299778234,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19062118828296662,
      "num_tokens": 27556866.0,
      "step": 13275
    },
    {
      "entropy": 5.429691362380981,
      "epoch": 1.204644412191582,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048611404123764855,
      "loss": 5.0426,
      "mean_token_accuracy": 0.20097893178462983,
      "num_tokens": 27566951.0,
      "step": 13280
    },
    {
      "entropy": 5.440868282318116,
      "epoch": 1.205097968069666,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048610284813158154,
      "loss": 5.1474,
      "mean_token_accuracy": 0.1993878111243248,
      "num_tokens": 27576796.0,
      "step": 13285
    },
    {
      "entropy": 5.375292730331421,
      "epoch": 1.2055515239477503,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004860916506598538,
      "loss": 5.0122,
      "mean_token_accuracy": 0.1960039883852005,
      "num_tokens": 27586902.0,
      "step": 13290
    },
    {
      "entropy": 5.583034420013428,
      "epoch": 1.2060050798258346,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048608044882269695,
      "loss": 5.1944,
      "mean_token_accuracy": 0.19058731198310852,
      "num_tokens": 27599232.0,
      "step": 13295
    },
    {
      "entropy": 5.469849109649658,
      "epoch": 1.2064586357039186,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004860692426203427,
      "loss": 4.9867,
      "mean_token_accuracy": 0.1980736255645752,
      "num_tokens": 27609022.0,
      "step": 13300
    },
    {
      "entropy": 5.465440654754639,
      "epoch": 1.206912191582003,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004860580320530226,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19946297705173494,
      "num_tokens": 27619829.0,
      "step": 13305
    },
    {
      "entropy": 5.488517141342163,
      "epoch": 1.2073657474600872,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004860468171209687,
      "loss": 5.1323,
      "mean_token_accuracy": 0.1990528702735901,
      "num_tokens": 27629337.0,
      "step": 13310
    },
    {
      "entropy": 5.55136661529541,
      "epoch": 1.2078193033381712,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004860355978244128,
      "loss": 5.1626,
      "mean_token_accuracy": 0.19753118902444838,
      "num_tokens": 27639377.0,
      "step": 13315
    },
    {
      "entropy": 5.485705947875976,
      "epoch": 1.2082728592162555,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000486024374163587,
      "loss": 5.0487,
      "mean_token_accuracy": 0.19785596281290055,
      "num_tokens": 27649989.0,
      "step": 13320
    },
    {
      "entropy": 5.465881109237671,
      "epoch": 1.2087264150943395,
      "grad_norm": 1.0,
      "learning_rate": 0.00048601314613872334,
      "loss": 5.1105,
      "mean_token_accuracy": 0.1890886515378952,
      "num_tokens": 27659967.0,
      "step": 13325
    },
    {
      "entropy": 5.469597816467285,
      "epoch": 1.2091799709724238,
      "grad_norm": 1.0625,
      "learning_rate": 0.000486001913750054,
      "loss": 5.0743,
      "mean_token_accuracy": 0.19993726462125777,
      "num_tokens": 27670099.0,
      "step": 13330
    },
    {
      "entropy": 5.449666976928711,
      "epoch": 1.209633526850508,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004859906769978113,
      "loss": 5.1182,
      "mean_token_accuracy": 0.1906730830669403,
      "num_tokens": 27681497.0,
      "step": 13335
    },
    {
      "entropy": 5.556787586212158,
      "epoch": 1.2100870827285921,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048597943588222776,
      "loss": 5.2207,
      "mean_token_accuracy": 0.19126688838005065,
      "num_tokens": 27691828.0,
      "step": 13340
    },
    {
      "entropy": 5.517446708679199,
      "epoch": 1.2105406386066764,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048596819040353564,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20397276282310486,
      "num_tokens": 27701426.0,
      "step": 13345
    },
    {
      "entropy": 5.471580123901367,
      "epoch": 1.2109941944847606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859569405619676,
      "loss": 5.02,
      "mean_token_accuracy": 0.2044997438788414,
      "num_tokens": 27711112.0,
      "step": 13350
    },
    {
      "entropy": 5.45523157119751,
      "epoch": 1.2114477503628447,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004859456863577562,
      "loss": 5.1359,
      "mean_token_accuracy": 0.194906784594059,
      "num_tokens": 27723160.0,
      "step": 13355
    },
    {
      "entropy": 5.506615352630615,
      "epoch": 1.211901306240929,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048593442779113437,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19654038101434707,
      "num_tokens": 27733143.0,
      "step": 13360
    },
    {
      "entropy": 5.465633964538574,
      "epoch": 1.212354862119013,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004859231648623348,
      "loss": 5.0949,
      "mean_token_accuracy": 0.19797311574220658,
      "num_tokens": 27743956.0,
      "step": 13365
    },
    {
      "entropy": 5.477122449874878,
      "epoch": 1.2128084179970973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048591189757159035,
      "loss": 5.0989,
      "mean_token_accuracy": 0.20149367451667785,
      "num_tokens": 27754192.0,
      "step": 13370
    },
    {
      "entropy": 5.450489521026611,
      "epoch": 1.2132619738751815,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004859006259191342,
      "loss": 5.1346,
      "mean_token_accuracy": 0.1882784515619278,
      "num_tokens": 27763939.0,
      "step": 13375
    },
    {
      "entropy": 5.462380647659302,
      "epoch": 1.2137155297532656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004858893499051993,
      "loss": 5.0858,
      "mean_token_accuracy": 0.20263192504644395,
      "num_tokens": 27774743.0,
      "step": 13380
    },
    {
      "entropy": 5.48222074508667,
      "epoch": 1.2141690856313498,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004858780695300189,
      "loss": 5.0647,
      "mean_token_accuracy": 0.1981904089450836,
      "num_tokens": 27785344.0,
      "step": 13385
    },
    {
      "entropy": 5.501078081130982,
      "epoch": 1.2146226415094339,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004858667847938263,
      "loss": 5.1402,
      "mean_token_accuracy": 0.19375266283750534,
      "num_tokens": 27797101.0,
      "step": 13390
    },
    {
      "entropy": 5.436707925796509,
      "epoch": 1.2150761973875182,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048585549569685486,
      "loss": 5.0428,
      "mean_token_accuracy": 0.20690445750951766,
      "num_tokens": 27807581.0,
      "step": 13395
    },
    {
      "entropy": 5.416332244873047,
      "epoch": 1.2155297532656024,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048584420223933805,
      "loss": 5.0154,
      "mean_token_accuracy": 0.19844310879707336,
      "num_tokens": 27818170.0,
      "step": 13400
    },
    {
      "entropy": 5.414180326461792,
      "epoch": 1.2159833091436865,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004858329044215094,
      "loss": 5.0384,
      "mean_token_accuracy": 0.19891184717416763,
      "num_tokens": 27828153.0,
      "step": 13405
    },
    {
      "entropy": 5.476301193237305,
      "epoch": 1.2164368650217707,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048582160224360253,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20489715337753295,
      "num_tokens": 27837681.0,
      "step": 13410
    },
    {
      "entropy": 5.449605989456177,
      "epoch": 1.2168904208998548,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004858102957058513,
      "loss": 5.1101,
      "mean_token_accuracy": 0.1980344220995903,
      "num_tokens": 27847447.0,
      "step": 13415
    },
    {
      "entropy": 5.381835842132569,
      "epoch": 1.217343976777939,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048579898480848933,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20579961985349654,
      "num_tokens": 27858586.0,
      "step": 13420
    },
    {
      "entropy": 5.446962022781372,
      "epoch": 1.2177975326560233,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004857876695517507,
      "loss": 5.102,
      "mean_token_accuracy": 0.19679246246814727,
      "num_tokens": 27868832.0,
      "step": 13425
    },
    {
      "entropy": 5.4643807888031,
      "epoch": 1.2182510885341074,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048577634993586923,
      "loss": 5.0896,
      "mean_token_accuracy": 0.19575538486242294,
      "num_tokens": 27879339.0,
      "step": 13430
    },
    {
      "entropy": 5.522566127777099,
      "epoch": 1.2187046444121916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048576502596107924,
      "loss": 5.094,
      "mean_token_accuracy": 0.19788532853126525,
      "num_tokens": 27889964.0,
      "step": 13435
    },
    {
      "entropy": 5.506246519088745,
      "epoch": 1.2191582002902757,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004857536976276148,
      "loss": 5.1457,
      "mean_token_accuracy": 0.1924672469496727,
      "num_tokens": 27901004.0,
      "step": 13440
    },
    {
      "entropy": 5.431326627731323,
      "epoch": 1.21961175616836,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004857423649357102,
      "loss": 5.0627,
      "mean_token_accuracy": 0.19715383052825927,
      "num_tokens": 27910788.0,
      "step": 13445
    },
    {
      "entropy": 5.488765573501587,
      "epoch": 1.2200653120464442,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048573102788559974,
      "loss": 5.0954,
      "mean_token_accuracy": 0.19984215795993804,
      "num_tokens": 27921360.0,
      "step": 13450
    },
    {
      "entropy": 5.448400926589966,
      "epoch": 1.2205188679245282,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004857196864775179,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20081372559070587,
      "num_tokens": 27931889.0,
      "step": 13455
    },
    {
      "entropy": 5.435266065597534,
      "epoch": 1.2209724238026125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048570834071169926,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20292796045541764,
      "num_tokens": 27942049.0,
      "step": 13460
    },
    {
      "entropy": 5.521530485153198,
      "epoch": 1.2214259796806965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048569699058837856,
      "loss": 5.1713,
      "mean_token_accuracy": 0.19670335203409195,
      "num_tokens": 27952416.0,
      "step": 13465
    },
    {
      "entropy": 5.450357818603516,
      "epoch": 1.2218795355587808,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004856856361077903,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20239394903182983,
      "num_tokens": 27962009.0,
      "step": 13470
    },
    {
      "entropy": 5.495120096206665,
      "epoch": 1.222333091436865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004856742772701694,
      "loss": 5.0895,
      "mean_token_accuracy": 0.19622547924518585,
      "num_tokens": 27972175.0,
      "step": 13475
    },
    {
      "entropy": 5.491621303558349,
      "epoch": 1.2227866473149491,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048566291407575073,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19951627105474473,
      "num_tokens": 27982899.0,
      "step": 13480
    },
    {
      "entropy": 5.510865592956543,
      "epoch": 1.2232402031930334,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048565154652476937,
      "loss": 5.0314,
      "mean_token_accuracy": 0.1987655982375145,
      "num_tokens": 27993231.0,
      "step": 13485
    },
    {
      "entropy": 5.397731161117553,
      "epoch": 1.2236937590711174,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004856401746174603,
      "loss": 5.0049,
      "mean_token_accuracy": 0.20129185318946838,
      "num_tokens": 28003623.0,
      "step": 13490
    },
    {
      "entropy": 5.455231428146362,
      "epoch": 1.2241473149492017,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048562879835405874,
      "loss": 5.0118,
      "mean_token_accuracy": 0.19701914489269257,
      "num_tokens": 28013878.0,
      "step": 13495
    },
    {
      "entropy": 5.433118629455566,
      "epoch": 1.224600870827286,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004856174177348,
      "loss": 5.1104,
      "mean_token_accuracy": 0.19880663901567458,
      "num_tokens": 28024224.0,
      "step": 13500
    },
    {
      "entropy": 5.461540269851684,
      "epoch": 1.22505442670537,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004856060327599194,
      "loss": 5.0032,
      "mean_token_accuracy": 0.19849352687597274,
      "num_tokens": 28034819.0,
      "step": 13505
    },
    {
      "entropy": 5.429188776016235,
      "epoch": 1.2255079825834543,
      "grad_norm": 1.0,
      "learning_rate": 0.00048559464342965233,
      "loss": 5.0944,
      "mean_token_accuracy": 0.19491631388664246,
      "num_tokens": 28044240.0,
      "step": 13510
    },
    {
      "entropy": 5.544237375259399,
      "epoch": 1.2259615384615385,
      "grad_norm": 1.0,
      "learning_rate": 0.0004855832497442344,
      "loss": 5.2122,
      "mean_token_accuracy": 0.189100544154644,
      "num_tokens": 28055046.0,
      "step": 13515
    },
    {
      "entropy": 5.425240516662598,
      "epoch": 1.2264150943396226,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004855718517039012,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20204996764659883,
      "num_tokens": 28065272.0,
      "step": 13520
    },
    {
      "entropy": 5.490952062606811,
      "epoch": 1.2268686502177069,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048556044930888844,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20701412707567216,
      "num_tokens": 28076255.0,
      "step": 13525
    },
    {
      "entropy": 5.460171794891357,
      "epoch": 1.2273222060957911,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004855490425594319,
      "loss": 5.1362,
      "mean_token_accuracy": 0.1976022332906723,
      "num_tokens": 28086258.0,
      "step": 13530
    },
    {
      "entropy": 5.450643873214721,
      "epoch": 1.2277757619738752,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004855376314557675,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20115755647420883,
      "num_tokens": 28096497.0,
      "step": 13535
    },
    {
      "entropy": 5.500379276275635,
      "epoch": 1.2282293178519594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004855262159981313,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20335399657487868,
      "num_tokens": 28106282.0,
      "step": 13540
    },
    {
      "entropy": 5.439828586578369,
      "epoch": 1.2286828737300435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048551479618675925,
      "loss": 5.1141,
      "mean_token_accuracy": 0.19755692631006241,
      "num_tokens": 28115771.0,
      "step": 13545
    },
    {
      "entropy": 5.497058439254761,
      "epoch": 1.2291364296081277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048550337202188756,
      "loss": 5.0382,
      "mean_token_accuracy": 0.20264212042093277,
      "num_tokens": 28125477.0,
      "step": 13550
    },
    {
      "entropy": 5.520259809494019,
      "epoch": 1.229589985486212,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004854919435037525,
      "loss": 5.0868,
      "mean_token_accuracy": 0.19803554713726043,
      "num_tokens": 28135800.0,
      "step": 13555
    },
    {
      "entropy": 5.379249429702758,
      "epoch": 1.230043541364296,
      "grad_norm": 1.0,
      "learning_rate": 0.0004854805106325904,
      "loss": 4.9601,
      "mean_token_accuracy": 0.1981172114610672,
      "num_tokens": 28146779.0,
      "step": 13560
    },
    {
      "entropy": 5.455027389526367,
      "epoch": 1.2304970972423803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004854690734086378,
      "loss": 5.1082,
      "mean_token_accuracy": 0.1966257244348526,
      "num_tokens": 28156819.0,
      "step": 13565
    },
    {
      "entropy": 5.474935817718506,
      "epoch": 1.2309506531204644,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004854576318321311,
      "loss": 5.0927,
      "mean_token_accuracy": 0.20191271901130675,
      "num_tokens": 28168553.0,
      "step": 13570
    },
    {
      "entropy": 5.49287052154541,
      "epoch": 1.2314042089985486,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048544618590330685,
      "loss": 5.0638,
      "mean_token_accuracy": 0.1998132899403572,
      "num_tokens": 28178064.0,
      "step": 13575
    },
    {
      "entropy": 5.531149291992188,
      "epoch": 1.231857764876633,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048543473562240194,
      "loss": 5.1918,
      "mean_token_accuracy": 0.20091952234506608,
      "num_tokens": 28188580.0,
      "step": 13580
    },
    {
      "entropy": 5.513899660110473,
      "epoch": 1.232311320754717,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000485423280989653,
      "loss": 5.1592,
      "mean_token_accuracy": 0.19042342901229858,
      "num_tokens": 28199955.0,
      "step": 13585
    },
    {
      "entropy": 5.512563610076905,
      "epoch": 1.2327648766328012,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000485411822005297,
      "loss": 5.1129,
      "mean_token_accuracy": 0.1964597672224045,
      "num_tokens": 28210506.0,
      "step": 13590
    },
    {
      "entropy": 5.463253450393677,
      "epoch": 1.2332184325108853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048540035866957097,
      "loss": 5.0449,
      "mean_token_accuracy": 0.2079138681292534,
      "num_tokens": 28220486.0,
      "step": 13595
    },
    {
      "entropy": 5.510498094558716,
      "epoch": 1.2336719883889695,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004853888909827118,
      "loss": 5.1748,
      "mean_token_accuracy": 0.18756349235773087,
      "num_tokens": 28232668.0,
      "step": 13600
    },
    {
      "entropy": 5.4791748046875,
      "epoch": 1.2341255442670538,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853774189449568,
      "loss": 5.1066,
      "mean_token_accuracy": 0.19442913085222244,
      "num_tokens": 28243733.0,
      "step": 13605
    },
    {
      "entropy": 5.5049889087677,
      "epoch": 1.2345791001451378,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048536594255654315,
      "loss": 5.1473,
      "mean_token_accuracy": 0.1973620057106018,
      "num_tokens": 28254953.0,
      "step": 13610
    },
    {
      "entropy": 5.4978475093841555,
      "epoch": 1.235032656023222,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853544618177083,
      "loss": 5.0958,
      "mean_token_accuracy": 0.1997261881828308,
      "num_tokens": 28264959.0,
      "step": 13615
    },
    {
      "entropy": 5.510729646682739,
      "epoch": 1.2354862119013061,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048534297672868945,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19888156205415725,
      "num_tokens": 28275509.0,
      "step": 13620
    },
    {
      "entropy": 5.475466060638428,
      "epoch": 1.2359397677793904,
      "grad_norm": 1.0,
      "learning_rate": 0.00048533148728972434,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20334992408752442,
      "num_tokens": 28285983.0,
      "step": 13625
    },
    {
      "entropy": 5.499315881729126,
      "epoch": 1.2363933236574747,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048531999350105037,
      "loss": 5.1161,
      "mean_token_accuracy": 0.19479625076055526,
      "num_tokens": 28296178.0,
      "step": 13630
    },
    {
      "entropy": 5.468041181564331,
      "epoch": 1.2368468795355587,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048530849536290537,
      "loss": 5.0874,
      "mean_token_accuracy": 0.18888186365365983,
      "num_tokens": 28305741.0,
      "step": 13635
    },
    {
      "entropy": 5.425489664077759,
      "epoch": 1.237300435413643,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004852969928755271,
      "loss": 5.0267,
      "mean_token_accuracy": 0.20155393332242966,
      "num_tokens": 28315607.0,
      "step": 13640
    },
    {
      "entropy": 5.400540924072265,
      "epoch": 1.237753991291727,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048528548603915344,
      "loss": 5.0544,
      "mean_token_accuracy": 0.20124746710062028,
      "num_tokens": 28327349.0,
      "step": 13645
    },
    {
      "entropy": 5.484639644622803,
      "epoch": 1.2382075471698113,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048527397485402236,
      "loss": 5.0484,
      "mean_token_accuracy": 0.19532727748155593,
      "num_tokens": 28339949.0,
      "step": 13650
    },
    {
      "entropy": 5.502997541427613,
      "epoch": 1.2386611030478956,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048526245932037186,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20198170691728592,
      "num_tokens": 28351340.0,
      "step": 13655
    },
    {
      "entropy": 5.357645320892334,
      "epoch": 1.2391146589259796,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048525093943844013,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2061908334493637,
      "num_tokens": 28360929.0,
      "step": 13660
    },
    {
      "entropy": 5.406160879135132,
      "epoch": 1.2395682148040639,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004852394152084654,
      "loss": 5.045,
      "mean_token_accuracy": 0.19744858294725418,
      "num_tokens": 28370477.0,
      "step": 13665
    },
    {
      "entropy": 5.410417938232422,
      "epoch": 1.240021770682148,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000485227886630686,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2147770643234253,
      "num_tokens": 28380876.0,
      "step": 13670
    },
    {
      "entropy": 5.360565090179444,
      "epoch": 1.2404753265602322,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004852163537053403,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20302784889936448,
      "num_tokens": 28390685.0,
      "step": 13675
    },
    {
      "entropy": 5.4372467517852785,
      "epoch": 1.2409288824383164,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004852048164326669,
      "loss": 5.0647,
      "mean_token_accuracy": 0.19382647573947906,
      "num_tokens": 28400236.0,
      "step": 13680
    },
    {
      "entropy": 5.514097785949707,
      "epoch": 1.2413824383164005,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004851932748129043,
      "loss": 5.1066,
      "mean_token_accuracy": 0.20565356761217118,
      "num_tokens": 28410353.0,
      "step": 13685
    },
    {
      "entropy": 5.482953500747681,
      "epoch": 1.2418359941944848,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004851817288462912,
      "loss": 5.023,
      "mean_token_accuracy": 0.1984424039721489,
      "num_tokens": 28420241.0,
      "step": 13690
    },
    {
      "entropy": 5.446218776702881,
      "epoch": 1.242289550072569,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004851701785330663,
      "loss": 5.099,
      "mean_token_accuracy": 0.19731494635343552,
      "num_tokens": 28430610.0,
      "step": 13695
    },
    {
      "entropy": 5.443164396286011,
      "epoch": 1.242743105950653,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004851586238734687,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20113435238599778,
      "num_tokens": 28440026.0,
      "step": 13700
    },
    {
      "entropy": 5.466651964187622,
      "epoch": 1.2431966618287373,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004851470648677371,
      "loss": 5.0943,
      "mean_token_accuracy": 0.18652191311120986,
      "num_tokens": 28450707.0,
      "step": 13705
    },
    {
      "entropy": 5.482295846939087,
      "epoch": 1.2436502177068216,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004851355015161107,
      "loss": 5.0342,
      "mean_token_accuracy": 0.19864806681871414,
      "num_tokens": 28462378.0,
      "step": 13710
    },
    {
      "entropy": 5.530763387680054,
      "epoch": 1.2441037735849056,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004851239338188286,
      "loss": 5.1238,
      "mean_token_accuracy": 0.1938784420490265,
      "num_tokens": 28473084.0,
      "step": 13715
    },
    {
      "entropy": 5.45649676322937,
      "epoch": 1.24455732946299,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048511236177612994,
      "loss": 5.0648,
      "mean_token_accuracy": 0.19571323394775392,
      "num_tokens": 28482767.0,
      "step": 13720
    },
    {
      "entropy": 5.450520944595337,
      "epoch": 1.245010885341074,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048510078538825413,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20403538793325424,
      "num_tokens": 28493274.0,
      "step": 13725
    },
    {
      "entropy": 5.501022720336914,
      "epoch": 1.2454644412191582,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048508920465544056,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20316553562879563,
      "num_tokens": 28503823.0,
      "step": 13730
    },
    {
      "entropy": 5.4013471603393555,
      "epoch": 1.2459179970972425,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004850776195779286,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20056904554367067,
      "num_tokens": 28514209.0,
      "step": 13735
    },
    {
      "entropy": 5.504916524887085,
      "epoch": 1.2463715529753265,
      "grad_norm": 0.984375,
      "learning_rate": 0.000485066030155958,
      "loss": 5.0889,
      "mean_token_accuracy": 0.19162350594997407,
      "num_tokens": 28524665.0,
      "step": 13740
    },
    {
      "entropy": 5.453751277923584,
      "epoch": 1.2468251088534108,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004850544363897684,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20215146988630295,
      "num_tokens": 28533639.0,
      "step": 13745
    },
    {
      "entropy": 5.350737571716309,
      "epoch": 1.2472786647314948,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048504283827959947,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20369759500026702,
      "num_tokens": 28544469.0,
      "step": 13750
    },
    {
      "entropy": 5.499988698959351,
      "epoch": 1.2477322206095791,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004850312358256912,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2021592527627945,
      "num_tokens": 28554637.0,
      "step": 13755
    },
    {
      "entropy": 5.422092723846435,
      "epoch": 1.2481857764876634,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004850196290282834,
      "loss": 5.0486,
      "mean_token_accuracy": 0.20041492730379104,
      "num_tokens": 28564752.0,
      "step": 13760
    },
    {
      "entropy": 5.465345001220703,
      "epoch": 1.2486393323657474,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048500801788761613,
      "loss": 5.1718,
      "mean_token_accuracy": 0.1989479646086693,
      "num_tokens": 28575007.0,
      "step": 13765
    },
    {
      "entropy": 5.534001684188842,
      "epoch": 1.2490928882438317,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004849964024039296,
      "loss": 5.0591,
      "mean_token_accuracy": 0.19341023862361909,
      "num_tokens": 28584907.0,
      "step": 13770
    },
    {
      "entropy": 5.562993860244751,
      "epoch": 1.2495464441219157,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048498478257746387,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20470116883516312,
      "num_tokens": 28594244.0,
      "step": 13775
    },
    {
      "entropy": 5.505571889877319,
      "epoch": 1.25,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048497315840845934,
      "loss": 5.0859,
      "mean_token_accuracy": 0.20225802212953567,
      "num_tokens": 28603816.0,
      "step": 13780
    },
    {
      "entropy": 5.5165565490722654,
      "epoch": 1.2504535558780843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004849615298971565,
      "loss": 5.1478,
      "mean_token_accuracy": 0.19335326850414275,
      "num_tokens": 28613067.0,
      "step": 13785
    },
    {
      "entropy": 5.4746994972229,
      "epoch": 1.2509071117561683,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048494989704379556,
      "loss": 5.0639,
      "mean_token_accuracy": 0.2013370230793953,
      "num_tokens": 28623752.0,
      "step": 13790
    },
    {
      "entropy": 5.424532461166382,
      "epoch": 1.2513606676342526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004849382598486173,
      "loss": 5.036,
      "mean_token_accuracy": 0.20286421626806259,
      "num_tokens": 28633190.0,
      "step": 13795
    },
    {
      "entropy": 5.5794517993927,
      "epoch": 1.2518142235123366,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048492661831186243,
      "loss": 5.1173,
      "mean_token_accuracy": 0.1939726009964943,
      "num_tokens": 28644567.0,
      "step": 13800
    },
    {
      "entropy": 5.531249856948852,
      "epoch": 1.2522677793904209,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004849149724337715,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19003446996212006,
      "num_tokens": 28654785.0,
      "step": 13805
    },
    {
      "entropy": 5.394008159637451,
      "epoch": 1.2527213352685052,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048490332221458543,
      "loss": 5.0068,
      "mean_token_accuracy": 0.19606393128633498,
      "num_tokens": 28665286.0,
      "step": 13810
    },
    {
      "entropy": 5.487374210357666,
      "epoch": 1.2531748911465892,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004848916676545452,
      "loss": 5.1148,
      "mean_token_accuracy": 0.19793294072151185,
      "num_tokens": 28675830.0,
      "step": 13815
    },
    {
      "entropy": 5.577816677093506,
      "epoch": 1.2536284470246735,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048488000875389174,
      "loss": 5.1786,
      "mean_token_accuracy": 0.19012221544981003,
      "num_tokens": 28686640.0,
      "step": 13820
    },
    {
      "entropy": 5.465760231018066,
      "epoch": 1.2540820029027575,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004848683455128663,
      "loss": 5.056,
      "mean_token_accuracy": 0.19991220235824586,
      "num_tokens": 28697677.0,
      "step": 13825
    },
    {
      "entropy": 5.452733993530273,
      "epoch": 1.2545355587808418,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048485667793170994,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19883179068565368,
      "num_tokens": 28708831.0,
      "step": 13830
    },
    {
      "entropy": 5.497370481491089,
      "epoch": 1.254989114658926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000484845006010664,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19644265174865722,
      "num_tokens": 28718959.0,
      "step": 13835
    },
    {
      "entropy": 5.454852151870727,
      "epoch": 1.25544267053701,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004848333297499699,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20070620626211166,
      "num_tokens": 28728719.0,
      "step": 13840
    },
    {
      "entropy": 5.512126350402832,
      "epoch": 1.2558962264150944,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048482164914986903,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1850140228867531,
      "num_tokens": 28740032.0,
      "step": 13845
    },
    {
      "entropy": 5.433183240890503,
      "epoch": 1.2563497822931784,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048480996421060307,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20678891539573668,
      "num_tokens": 28749374.0,
      "step": 13850
    },
    {
      "entropy": 5.4531410217285154,
      "epoch": 1.2568033381712627,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048479827493241347,
      "loss": 5.0522,
      "mean_token_accuracy": 0.19824664890766144,
      "num_tokens": 28760038.0,
      "step": 13855
    },
    {
      "entropy": 5.492013168334961,
      "epoch": 1.257256894049347,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048478658131554216,
      "loss": 5.083,
      "mean_token_accuracy": 0.20080812126398087,
      "num_tokens": 28770247.0,
      "step": 13860
    },
    {
      "entropy": 5.422021341323853,
      "epoch": 1.2577104499274312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004847748833602308,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2001783072948456,
      "num_tokens": 28780251.0,
      "step": 13865
    },
    {
      "entropy": 5.410889577865601,
      "epoch": 1.2581640058055152,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004847631810667215,
      "loss": 5.0378,
      "mean_token_accuracy": 0.20500718355178832,
      "num_tokens": 28790829.0,
      "step": 13870
    },
    {
      "entropy": 5.42143063545227,
      "epoch": 1.2586175616835993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004847514744352561,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20584760904312133,
      "num_tokens": 28801036.0,
      "step": 13875
    },
    {
      "entropy": 5.376169872283936,
      "epoch": 1.2590711175616836,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004847397634660768,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20472907721996308,
      "num_tokens": 28811013.0,
      "step": 13880
    },
    {
      "entropy": 5.407306241989136,
      "epoch": 1.2595246734397678,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048472804815942575,
      "loss": 5.026,
      "mean_token_accuracy": 0.1978759750723839,
      "num_tokens": 28821367.0,
      "step": 13885
    },
    {
      "entropy": 5.475953722000122,
      "epoch": 1.259978229317852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048471632851554517,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20478258281946182,
      "num_tokens": 28831077.0,
      "step": 13890
    },
    {
      "entropy": 5.479902696609497,
      "epoch": 1.2604317851959361,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004847046045346775,
      "loss": 5.0887,
      "mean_token_accuracy": 0.19575609862804413,
      "num_tokens": 28841490.0,
      "step": 13895
    },
    {
      "entropy": 5.496106100082398,
      "epoch": 1.2608853410740204,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048469287621706516,
      "loss": 5.0975,
      "mean_token_accuracy": 0.19364373236894608,
      "num_tokens": 28852510.0,
      "step": 13900
    },
    {
      "entropy": 5.453580331802368,
      "epoch": 1.2613388969521044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004846811435629507,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20205463320016862,
      "num_tokens": 28863388.0,
      "step": 13905
    },
    {
      "entropy": 5.371431636810303,
      "epoch": 1.2617924528301887,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004846694065725768,
      "loss": 5.0237,
      "mean_token_accuracy": 0.1987258344888687,
      "num_tokens": 28873798.0,
      "step": 13910
    },
    {
      "entropy": 5.477045679092408,
      "epoch": 1.262246008708273,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004846576652461861,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19760629385709763,
      "num_tokens": 28883482.0,
      "step": 13915
    },
    {
      "entropy": 5.460442686080933,
      "epoch": 1.262699564586357,
      "grad_norm": 1.0,
      "learning_rate": 0.00048464591958402147,
      "loss": 5.002,
      "mean_token_accuracy": 0.2040782615542412,
      "num_tokens": 28893573.0,
      "step": 13920
    },
    {
      "entropy": 5.488413143157959,
      "epoch": 1.2631531204644413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048463416958632573,
      "loss": 5.1558,
      "mean_token_accuracy": 0.19369145929813386,
      "num_tokens": 28904064.0,
      "step": 13925
    },
    {
      "entropy": 5.496292495727539,
      "epoch": 1.2636066763425253,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048462241525334206,
      "loss": 5.0952,
      "mean_token_accuracy": 0.19503602236509324,
      "num_tokens": 28915329.0,
      "step": 13930
    },
    {
      "entropy": 5.488839101791382,
      "epoch": 1.2640602322206096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004846106565853133,
      "loss": 5.1786,
      "mean_token_accuracy": 0.1940974101424217,
      "num_tokens": 28925228.0,
      "step": 13935
    },
    {
      "entropy": 5.454708576202393,
      "epoch": 1.2645137880986939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048459889358248284,
      "loss": 5.0234,
      "mean_token_accuracy": 0.1984314426779747,
      "num_tokens": 28935891.0,
      "step": 13940
    },
    {
      "entropy": 5.43781156539917,
      "epoch": 1.264967343976778,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048458712624509383,
      "loss": 4.9921,
      "mean_token_accuracy": 0.20576515346765517,
      "num_tokens": 28946372.0,
      "step": 13945
    },
    {
      "entropy": 5.475255060195923,
      "epoch": 1.2654208998548622,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048457535457338964,
      "loss": 5.0767,
      "mean_token_accuracy": 0.19604025781154633,
      "num_tokens": 28957874.0,
      "step": 13950
    },
    {
      "entropy": 5.479547214508057,
      "epoch": 1.2658744557329462,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004845635785676137,
      "loss": 5.0597,
      "mean_token_accuracy": 0.18991512060165405,
      "num_tokens": 28968023.0,
      "step": 13955
    },
    {
      "entropy": 5.42999267578125,
      "epoch": 1.2663280116110305,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048455179822800957,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20645622462034224,
      "num_tokens": 28977744.0,
      "step": 13960
    },
    {
      "entropy": 5.52282338142395,
      "epoch": 1.2667815674891147,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004845400135548208,
      "loss": 5.2031,
      "mean_token_accuracy": 0.18940825760364532,
      "num_tokens": 28988875.0,
      "step": 13965
    },
    {
      "entropy": 5.4079639434814455,
      "epoch": 1.2672351233671988,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004845282245482912,
      "loss": 5.011,
      "mean_token_accuracy": 0.2026742398738861,
      "num_tokens": 28998562.0,
      "step": 13970
    },
    {
      "entropy": 5.474730110168457,
      "epoch": 1.267688679245283,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845164312086646,
      "loss": 5.0452,
      "mean_token_accuracy": 0.2058258980512619,
      "num_tokens": 29008783.0,
      "step": 13975
    },
    {
      "entropy": 5.420112085342407,
      "epoch": 1.268142235123367,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004845046335361847,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2004618838429451,
      "num_tokens": 29019303.0,
      "step": 13980
    },
    {
      "entropy": 5.500355291366577,
      "epoch": 1.2685957910014514,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048449283153109566,
      "loss": 5.1978,
      "mean_token_accuracy": 0.18973972648382187,
      "num_tokens": 29029843.0,
      "step": 13985
    },
    {
      "entropy": 5.511347150802612,
      "epoch": 1.2690493468795356,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004844810251936413,
      "loss": 5.1089,
      "mean_token_accuracy": 0.19228664934635162,
      "num_tokens": 29040603.0,
      "step": 13990
    },
    {
      "entropy": 5.46467547416687,
      "epoch": 1.2695029027576197,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004844692145240661,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20179881155490875,
      "num_tokens": 29050599.0,
      "step": 13995
    },
    {
      "entropy": 5.44791259765625,
      "epoch": 1.269956458635704,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004844573995226142,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19354299306869507,
      "num_tokens": 29061292.0,
      "step": 14000
    },
    {
      "entropy": 5.39538836479187,
      "epoch": 1.270410014513788,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048444558018952984,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20973615050315858,
      "num_tokens": 29071779.0,
      "step": 14005
    },
    {
      "entropy": 5.394731140136718,
      "epoch": 1.2708635703918723,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048443375652505754,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2039197489619255,
      "num_tokens": 29082073.0,
      "step": 14010
    },
    {
      "entropy": 5.456092166900635,
      "epoch": 1.2713171262699565,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004844219285294418,
      "loss": 4.9651,
      "mean_token_accuracy": 0.20332453697919844,
      "num_tokens": 29092032.0,
      "step": 14015
    },
    {
      "entropy": 5.456031847000122,
      "epoch": 1.2717706821480406,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048441009620292714,
      "loss": 5.012,
      "mean_token_accuracy": 0.20634298771619797,
      "num_tokens": 29102936.0,
      "step": 14020
    },
    {
      "entropy": 5.451512145996094,
      "epoch": 1.2722242380261248,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004843982595457584,
      "loss": 5.05,
      "mean_token_accuracy": 0.20064093321561813,
      "num_tokens": 29113385.0,
      "step": 14025
    },
    {
      "entropy": 5.43087363243103,
      "epoch": 1.2726777939042089,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048438641855818026,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1958359807729721,
      "num_tokens": 29125075.0,
      "step": 14030
    },
    {
      "entropy": 5.471334028244018,
      "epoch": 1.2731313497822931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004843745732404376,
      "loss": 5.0238,
      "mean_token_accuracy": 0.19629547894001007,
      "num_tokens": 29135098.0,
      "step": 14035
    },
    {
      "entropy": 5.53118257522583,
      "epoch": 1.2735849056603774,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004843627235927755,
      "loss": 5.1128,
      "mean_token_accuracy": 0.19498565495014192,
      "num_tokens": 29145577.0,
      "step": 14040
    },
    {
      "entropy": 5.4487803936004635,
      "epoch": 1.2740384615384617,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004843508696154388,
      "loss": 5.0529,
      "mean_token_accuracy": 0.2005828231573105,
      "num_tokens": 29155905.0,
      "step": 14045
    },
    {
      "entropy": 5.411189699172974,
      "epoch": 1.2744920174165457,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048433901130867283,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20621789246797562,
      "num_tokens": 29164926.0,
      "step": 14050
    },
    {
      "entropy": 5.406766605377197,
      "epoch": 1.2749455732946298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004843271486727228,
      "loss": 5.0156,
      "mean_token_accuracy": 0.19711293280124664,
      "num_tokens": 29175050.0,
      "step": 14055
    },
    {
      "entropy": 5.4486559391021725,
      "epoch": 1.275399129172714,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048431528170783395,
      "loss": 5.0923,
      "mean_token_accuracy": 0.19830595701932907,
      "num_tokens": 29186255.0,
      "step": 14060
    },
    {
      "entropy": 5.471863603591919,
      "epoch": 1.2758526850507983,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004843034104142517,
      "loss": 5.0657,
      "mean_token_accuracy": 0.20228296816349028,
      "num_tokens": 29195639.0,
      "step": 14065
    },
    {
      "entropy": 5.458148193359375,
      "epoch": 1.2763062409288826,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004842915347922216,
      "loss": 5.0802,
      "mean_token_accuracy": 0.1935840979218483,
      "num_tokens": 29206482.0,
      "step": 14070
    },
    {
      "entropy": 5.49206395149231,
      "epoch": 1.2767597968069666,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004842796548419893,
      "loss": 5.0558,
      "mean_token_accuracy": 0.19867073446512223,
      "num_tokens": 29217433.0,
      "step": 14075
    },
    {
      "entropy": 5.525887107849121,
      "epoch": 1.2772133526850509,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048426777056380035,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19111961275339126,
      "num_tokens": 29226193.0,
      "step": 14080
    },
    {
      "entropy": 5.457399654388428,
      "epoch": 1.277666908563135,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004842558819579006,
      "loss": 5.0642,
      "mean_token_accuracy": 0.2021912306547165,
      "num_tokens": 29237061.0,
      "step": 14085
    },
    {
      "entropy": 5.421337223052978,
      "epoch": 1.2781204644412192,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004842439890245358,
      "loss": 5.022,
      "mean_token_accuracy": 0.20495831370353698,
      "num_tokens": 29247689.0,
      "step": 14090
    },
    {
      "entropy": 5.431916189193726,
      "epoch": 1.2785740203193035,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048423209176395204,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2016739085316658,
      "num_tokens": 29258235.0,
      "step": 14095
    },
    {
      "entropy": 5.448566150665283,
      "epoch": 1.2790275761973875,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004842201901763954,
      "loss": 4.9887,
      "mean_token_accuracy": 0.2021269142627716,
      "num_tokens": 29268884.0,
      "step": 14100
    },
    {
      "entropy": 5.418848896026612,
      "epoch": 1.2794811320754718,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004842082842621118,
      "loss": 5.1334,
      "mean_token_accuracy": 0.19570833295583726,
      "num_tokens": 29280484.0,
      "step": 14105
    },
    {
      "entropy": 5.488956546783447,
      "epoch": 1.2799346879535558,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004841963740213476,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19842521697282792,
      "num_tokens": 29291199.0,
      "step": 14110
    },
    {
      "entropy": 5.534154844284058,
      "epoch": 1.28038824383164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004841844594543491,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19891676008701326,
      "num_tokens": 29302007.0,
      "step": 14115
    },
    {
      "entropy": 5.4513023853302,
      "epoch": 1.2808417997097243,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841725405613626,
      "loss": 5.0398,
      "mean_token_accuracy": 0.20795077979564666,
      "num_tokens": 29312127.0,
      "step": 14120
    },
    {
      "entropy": 5.4811718463897705,
      "epoch": 1.2812953555878084,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004841606173426347,
      "loss": 5.1642,
      "mean_token_accuracy": 0.19157685488462448,
      "num_tokens": 29323691.0,
      "step": 14125
    },
    {
      "entropy": 5.46334490776062,
      "epoch": 1.2817489114658926,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048414868979841197,
      "loss": 5.0798,
      "mean_token_accuracy": 0.1970996528863907,
      "num_tokens": 29335203.0,
      "step": 14130
    },
    {
      "entropy": 5.507711458206177,
      "epoch": 1.2822024673439767,
      "grad_norm": 1.0,
      "learning_rate": 0.00048413675792894105,
      "loss": 5.095,
      "mean_token_accuracy": 0.19961245954036713,
      "num_tokens": 29346422.0,
      "step": 14135
    },
    {
      "entropy": 5.438190031051636,
      "epoch": 1.282656023222061,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841248217344686,
      "loss": 5.0282,
      "mean_token_accuracy": 0.20192458480596542,
      "num_tokens": 29356084.0,
      "step": 14140
    },
    {
      "entropy": 5.429281902313233,
      "epoch": 1.2831095791001452,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004841128812152415,
      "loss": 5.0308,
      "mean_token_accuracy": 0.19918602108955383,
      "num_tokens": 29366093.0,
      "step": 14145
    },
    {
      "entropy": 5.484508800506592,
      "epoch": 1.2835631349782293,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004841009363715068,
      "loss": 5.0892,
      "mean_token_accuracy": 0.19784653931856155,
      "num_tokens": 29376621.0,
      "step": 14150
    },
    {
      "entropy": 5.510742902755737,
      "epoch": 1.2840166908563135,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004840889872035115,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19336931705474852,
      "num_tokens": 29387119.0,
      "step": 14155
    },
    {
      "entropy": 5.46896710395813,
      "epoch": 1.2844702467343976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004840770337115026,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20229030400514603,
      "num_tokens": 29396981.0,
      "step": 14160
    },
    {
      "entropy": 5.504418849945068,
      "epoch": 1.2849238026124818,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048406507589572747,
      "loss": 5.0347,
      "mean_token_accuracy": 0.200796540081501,
      "num_tokens": 29406737.0,
      "step": 14165
    },
    {
      "entropy": 5.473976898193359,
      "epoch": 1.2853773584905661,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004840531137564331,
      "loss": 5.0579,
      "mean_token_accuracy": 0.19947080463171005,
      "num_tokens": 29417300.0,
      "step": 14170
    },
    {
      "entropy": 5.47886095046997,
      "epoch": 1.2858309143686502,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004840411472938672,
      "loss": 4.9791,
      "mean_token_accuracy": 0.20759643763303756,
      "num_tokens": 29426749.0,
      "step": 14175
    },
    {
      "entropy": 5.409400892257691,
      "epoch": 1.2862844702467344,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004840291765082771,
      "loss": 5.0038,
      "mean_token_accuracy": 0.19947194159030915,
      "num_tokens": 29437264.0,
      "step": 14180
    },
    {
      "entropy": 5.426572322845459,
      "epoch": 1.2867380261248185,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048401720139991036,
      "loss": 5.0046,
      "mean_token_accuracy": 0.19459227323532105,
      "num_tokens": 29448192.0,
      "step": 14185
    },
    {
      "entropy": 5.482336568832397,
      "epoch": 1.2871915820029027,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004840052219690146,
      "loss": 4.911,
      "mean_token_accuracy": 0.21285396963357925,
      "num_tokens": 29457923.0,
      "step": 14190
    },
    {
      "entropy": 5.365271377563476,
      "epoch": 1.287645137880987,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004839932382158376,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2121126487851143,
      "num_tokens": 29466914.0,
      "step": 14195
    },
    {
      "entropy": 5.429365825653076,
      "epoch": 1.288098693759071,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004839812501406272,
      "loss": 5.0911,
      "mean_token_accuracy": 0.19476546943187714,
      "num_tokens": 29477648.0,
      "step": 14200
    },
    {
      "entropy": 5.48526816368103,
      "epoch": 1.2885522496371553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004839692577436313,
      "loss": 5.0693,
      "mean_token_accuracy": 0.2014537572860718,
      "num_tokens": 29487947.0,
      "step": 14205
    },
    {
      "entropy": 5.572406339645386,
      "epoch": 1.2890058055152394,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004839572610250978,
      "loss": 5.162,
      "mean_token_accuracy": 0.18987073600292206,
      "num_tokens": 29497667.0,
      "step": 14210
    },
    {
      "entropy": 5.466194772720337,
      "epoch": 1.2894593613933236,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048394525998527496,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19715804904699324,
      "num_tokens": 29507065.0,
      "step": 14215
    },
    {
      "entropy": 5.515470218658447,
      "epoch": 1.2899129172714079,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004839332546244109,
      "loss": 5.1207,
      "mean_token_accuracy": 0.19964057207107544,
      "num_tokens": 29517752.0,
      "step": 14220
    },
    {
      "entropy": 5.451497173309326,
      "epoch": 1.2903664731494922,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048392124494275385,
      "loss": 5.0396,
      "mean_token_accuracy": 0.19701832681894302,
      "num_tokens": 29527894.0,
      "step": 14225
    },
    {
      "entropy": 5.436139106750488,
      "epoch": 1.2908200290275762,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004839092309405523,
      "loss": 5.0849,
      "mean_token_accuracy": 0.1972740739583969,
      "num_tokens": 29538136.0,
      "step": 14230
    },
    {
      "entropy": 5.476589345932007,
      "epoch": 1.2912735849056602,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048389721261805457,
      "loss": 5.06,
      "mean_token_accuracy": 0.19925848245620728,
      "num_tokens": 29549193.0,
      "step": 14235
    },
    {
      "entropy": 5.418052005767822,
      "epoch": 1.2917271407837445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048388518997550926,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20766545832157135,
      "num_tokens": 29558889.0,
      "step": 14240
    },
    {
      "entropy": 5.454746294021606,
      "epoch": 1.2921806966618288,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048387316301316493,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19525161534547805,
      "num_tokens": 29569691.0,
      "step": 14245
    },
    {
      "entropy": 5.488633298873902,
      "epoch": 1.292634252539913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004838611317312705,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19238779544830323,
      "num_tokens": 29580459.0,
      "step": 14250
    },
    {
      "entropy": 5.471126317977905,
      "epoch": 1.293087808417997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004838490961300745,
      "loss": 5.0597,
      "mean_token_accuracy": 0.20004819929599763,
      "num_tokens": 29590652.0,
      "step": 14255
    },
    {
      "entropy": 5.503458690643311,
      "epoch": 1.2935413642960814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048383705620982606,
      "loss": 5.0768,
      "mean_token_accuracy": 0.20538977682590484,
      "num_tokens": 29600353.0,
      "step": 14260
    },
    {
      "entropy": 5.4731116771698,
      "epoch": 1.2939949201741654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048382501197077405,
      "loss": 5.0315,
      "mean_token_accuracy": 0.21094868183135987,
      "num_tokens": 29610561.0,
      "step": 14265
    },
    {
      "entropy": 5.405670833587647,
      "epoch": 1.2944484760522497,
      "grad_norm": 0.875,
      "learning_rate": 0.0004838129634131675,
      "loss": 5.0372,
      "mean_token_accuracy": 0.1995544448494911,
      "num_tokens": 29621741.0,
      "step": 14270
    },
    {
      "entropy": 5.4140002727508545,
      "epoch": 1.294902031930334,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004838009105372558,
      "loss": 4.9694,
      "mean_token_accuracy": 0.1997040554881096,
      "num_tokens": 29633097.0,
      "step": 14275
    },
    {
      "entropy": 5.444863653182983,
      "epoch": 1.295355587808418,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048378885334328793,
      "loss": 5.0716,
      "mean_token_accuracy": 0.19167536050081252,
      "num_tokens": 29644290.0,
      "step": 14280
    },
    {
      "entropy": 5.548151206970215,
      "epoch": 1.2958091436865022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004837767918315135,
      "loss": 5.1291,
      "mean_token_accuracy": 0.19501840472221374,
      "num_tokens": 29654977.0,
      "step": 14285
    },
    {
      "entropy": 5.546808052062988,
      "epoch": 1.2962626995645863,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048376472600218174,
      "loss": 5.1217,
      "mean_token_accuracy": 0.1964360550045967,
      "num_tokens": 29665020.0,
      "step": 14290
    },
    {
      "entropy": 5.469972610473633,
      "epoch": 1.2967162554426706,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004837526558555422,
      "loss": 5.0756,
      "mean_token_accuracy": 0.20108735859394072,
      "num_tokens": 29675098.0,
      "step": 14295
    },
    {
      "entropy": 5.390252161026001,
      "epoch": 1.2971698113207548,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004837405813918446,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20647544860839845,
      "num_tokens": 29685450.0,
      "step": 14300
    },
    {
      "entropy": 5.44470853805542,
      "epoch": 1.2976233671988389,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048372850261133856,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20059045404195786,
      "num_tokens": 29694818.0,
      "step": 14305
    },
    {
      "entropy": 5.487842845916748,
      "epoch": 1.2980769230769231,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048371641951427393,
      "loss": 5.1113,
      "mean_token_accuracy": 0.1938343659043312,
      "num_tokens": 29705573.0,
      "step": 14310
    },
    {
      "entropy": 5.478750896453858,
      "epoch": 1.2985304789550072,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048370433210090055,
      "loss": 5.014,
      "mean_token_accuracy": 0.19965904355049133,
      "num_tokens": 29716366.0,
      "step": 14315
    },
    {
      "entropy": 5.4626648902893065,
      "epoch": 1.2989840348330914,
      "grad_norm": 1.0,
      "learning_rate": 0.0004836922403714683,
      "loss": 5.0361,
      "mean_token_accuracy": 0.19631998986005783,
      "num_tokens": 29727181.0,
      "step": 14320
    },
    {
      "entropy": 5.407274341583252,
      "epoch": 1.2994375907111757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048368014432622746,
      "loss": 5.0567,
      "mean_token_accuracy": 0.2011177882552147,
      "num_tokens": 29736960.0,
      "step": 14325
    },
    {
      "entropy": 5.408423233032226,
      "epoch": 1.2998911465892597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000483668043965428,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2119763821363449,
      "num_tokens": 29746354.0,
      "step": 14330
    },
    {
      "entropy": 5.4474845886230465,
      "epoch": 1.300344702467344,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004836559392893203,
      "loss": 5.0611,
      "mean_token_accuracy": 0.19734853953123094,
      "num_tokens": 29756627.0,
      "step": 14335
    },
    {
      "entropy": 5.443008661270142,
      "epoch": 1.300798258345428,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004836438302981545,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20591117888689042,
      "num_tokens": 29766684.0,
      "step": 14340
    },
    {
      "entropy": 5.4376171112060545,
      "epoch": 1.3012518142235123,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004836317169921811,
      "loss": 5.062,
      "mean_token_accuracy": 0.1998087629675865,
      "num_tokens": 29776556.0,
      "step": 14345
    },
    {
      "entropy": 5.5232834815979,
      "epoch": 1.3017053701015966,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048361959937165075,
      "loss": 5.1949,
      "mean_token_accuracy": 0.1920908659696579,
      "num_tokens": 29786896.0,
      "step": 14350
    },
    {
      "entropy": 5.478964185714721,
      "epoch": 1.3021589259796806,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004836074774368138,
      "loss": 5.0017,
      "mean_token_accuracy": 0.20372447669506072,
      "num_tokens": 29797332.0,
      "step": 14355
    },
    {
      "entropy": 5.4851744174957275,
      "epoch": 1.302612481857765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048359535118792116,
      "loss": 5.0952,
      "mean_token_accuracy": 0.20031577795743943,
      "num_tokens": 29806428.0,
      "step": 14360
    },
    {
      "entropy": 5.419641160964966,
      "epoch": 1.303066037735849,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004835832206252235,
      "loss": 5.0647,
      "mean_token_accuracy": 0.1950739622116089,
      "num_tokens": 29816794.0,
      "step": 14365
    },
    {
      "entropy": 5.536451816558838,
      "epoch": 1.3035195936139332,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004835710857489716,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20131175667047502,
      "num_tokens": 29827077.0,
      "step": 14370
    },
    {
      "entropy": 5.44500036239624,
      "epoch": 1.3039731494920175,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004835589465594165,
      "loss": 5.0549,
      "mean_token_accuracy": 0.195481513440609,
      "num_tokens": 29837868.0,
      "step": 14375
    },
    {
      "entropy": 5.45031886100769,
      "epoch": 1.3044267053701015,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048354680305680926,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20101405531167985,
      "num_tokens": 29846929.0,
      "step": 14380
    },
    {
      "entropy": 5.445803642272949,
      "epoch": 1.3048802612481858,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048353465524140105,
      "loss": 5.0217,
      "mean_token_accuracy": 0.2037921890616417,
      "num_tokens": 29857905.0,
      "step": 14385
    },
    {
      "entropy": 5.464574575424194,
      "epoch": 1.3053338171262698,
      "grad_norm": 1.0625,
      "learning_rate": 0.000483522503113443,
      "loss": 5.0777,
      "mean_token_accuracy": 0.1997160106897354,
      "num_tokens": 29868230.0,
      "step": 14390
    },
    {
      "entropy": 5.444516229629516,
      "epoch": 1.305787373004354,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048351034667318634,
      "loss": 5.0378,
      "mean_token_accuracy": 0.2014074981212616,
      "num_tokens": 29879224.0,
      "step": 14395
    },
    {
      "entropy": 5.525046682357788,
      "epoch": 1.3062409288824384,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004834981859208827,
      "loss": 5.1799,
      "mean_token_accuracy": 0.19462355971336365,
      "num_tokens": 29890416.0,
      "step": 14400
    },
    {
      "entropy": 5.508916282653809,
      "epoch": 1.3066944847605226,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048348602085678346,
      "loss": 5.1018,
      "mean_token_accuracy": 0.19744647145271302,
      "num_tokens": 29901064.0,
      "step": 14405
    },
    {
      "entropy": 5.538735198974609,
      "epoch": 1.3071480406386067,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048347385148114004,
      "loss": 5.1547,
      "mean_token_accuracy": 0.19499599188566208,
      "num_tokens": 29911084.0,
      "step": 14410
    },
    {
      "entropy": 5.428773021697998,
      "epoch": 1.3076015965166907,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048346167779420436,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20560617744922638,
      "num_tokens": 29921104.0,
      "step": 14415
    },
    {
      "entropy": 5.46851372718811,
      "epoch": 1.308055152394775,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048344949979622806,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2017678901553154,
      "num_tokens": 29932014.0,
      "step": 14420
    },
    {
      "entropy": 5.4437024116516115,
      "epoch": 1.3085087082728593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048343731748746294,
      "loss": 5.0517,
      "mean_token_accuracy": 0.1981023967266083,
      "num_tokens": 29941707.0,
      "step": 14425
    },
    {
      "entropy": 5.467644309997558,
      "epoch": 1.3089622641509435,
      "grad_norm": 1.09375,
      "learning_rate": 0.000483425130868161,
      "loss": 5.0408,
      "mean_token_accuracy": 0.19608643800020217,
      "num_tokens": 29952082.0,
      "step": 14430
    },
    {
      "entropy": 5.40396409034729,
      "epoch": 1.3094158200290276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004834129399385742,
      "loss": 4.8721,
      "mean_token_accuracy": 0.21680345386266708,
      "num_tokens": 29961744.0,
      "step": 14435
    },
    {
      "entropy": 5.432842874526978,
      "epoch": 1.3098693759071118,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004834007446989548,
      "loss": 5.025,
      "mean_token_accuracy": 0.21272454410791397,
      "num_tokens": 29972611.0,
      "step": 14440
    },
    {
      "entropy": 5.401928472518921,
      "epoch": 1.3103229317851959,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048338854514955474,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2062120631337166,
      "num_tokens": 29982660.0,
      "step": 14445
    },
    {
      "entropy": 5.446466875076294,
      "epoch": 1.3107764876632801,
      "grad_norm": 1.0,
      "learning_rate": 0.00048337634129062655,
      "loss": 5.0776,
      "mean_token_accuracy": 0.18785275816917418,
      "num_tokens": 29993349.0,
      "step": 14450
    },
    {
      "entropy": 5.570342636108398,
      "epoch": 1.3112300435413644,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048336413312242255,
      "loss": 5.1498,
      "mean_token_accuracy": 0.19252850264310836,
      "num_tokens": 30004474.0,
      "step": 14455
    },
    {
      "entropy": 5.484516429901123,
      "epoch": 1.3116835994194485,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048335192064519514,
      "loss": 5.0997,
      "mean_token_accuracy": 0.19876566678285598,
      "num_tokens": 30015189.0,
      "step": 14460
    },
    {
      "entropy": 5.453644609451294,
      "epoch": 1.3121371552975327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048333970385919695,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20059075802564622,
      "num_tokens": 30025491.0,
      "step": 14465
    },
    {
      "entropy": 5.501976776123047,
      "epoch": 1.3125907111756168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048332748276468053,
      "loss": 5.0647,
      "mean_token_accuracy": 0.20270003527402877,
      "num_tokens": 30034683.0,
      "step": 14470
    },
    {
      "entropy": 5.436651277542114,
      "epoch": 1.313044267053701,
      "grad_norm": 1.0,
      "learning_rate": 0.0004833152573618987,
      "loss": 5.0282,
      "mean_token_accuracy": 0.20032430738210677,
      "num_tokens": 30044135.0,
      "step": 14475
    },
    {
      "entropy": 5.495923280715942,
      "epoch": 1.3134978229317853,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048330302765110436,
      "loss": 5.111,
      "mean_token_accuracy": 0.19856578707695008,
      "num_tokens": 30054520.0,
      "step": 14480
    },
    {
      "entropy": 5.477340984344482,
      "epoch": 1.3139513788098693,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048329079363255024,
      "loss": 5.0921,
      "mean_token_accuracy": 0.19396299123764038,
      "num_tokens": 30065439.0,
      "step": 14485
    },
    {
      "entropy": 5.484131002426148,
      "epoch": 1.3144049346879536,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048327855530648934,
      "loss": 5.0769,
      "mean_token_accuracy": 0.1978706508874893,
      "num_tokens": 30075362.0,
      "step": 14490
    },
    {
      "entropy": 5.46552734375,
      "epoch": 1.3148584905660377,
      "grad_norm": 1.046875,
      "learning_rate": 0.000483266312673175,
      "loss": 5.0492,
      "mean_token_accuracy": 0.1971236765384674,
      "num_tokens": 30084913.0,
      "step": 14495
    },
    {
      "entropy": 5.467570734024048,
      "epoch": 1.315312046444122,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004832540657328602,
      "loss": 5.0579,
      "mean_token_accuracy": 0.19743605107069015,
      "num_tokens": 30095704.0,
      "step": 14500
    },
    {
      "entropy": 5.436670017242432,
      "epoch": 1.3157656023222062,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004832418144857982,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20036741942167283,
      "num_tokens": 30105981.0,
      "step": 14505
    },
    {
      "entropy": 5.463035917282104,
      "epoch": 1.3162191582002902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048322955893224245,
      "loss": 5.0763,
      "mean_token_accuracy": 0.18886080384254456,
      "num_tokens": 30115772.0,
      "step": 14510
    },
    {
      "entropy": 5.453473901748657,
      "epoch": 1.3166727140783745,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048321729907244626,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20168087184429168,
      "num_tokens": 30127050.0,
      "step": 14515
    },
    {
      "entropy": 5.467558574676514,
      "epoch": 1.3171262699564585,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004832050349066634,
      "loss": 5.0707,
      "mean_token_accuracy": 0.1975627213716507,
      "num_tokens": 30138221.0,
      "step": 14520
    },
    {
      "entropy": 5.512048435211182,
      "epoch": 1.3175798258345428,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004831927664351473,
      "loss": 5.1326,
      "mean_token_accuracy": 0.19621213227510453,
      "num_tokens": 30150181.0,
      "step": 14525
    },
    {
      "entropy": 5.4882982730865475,
      "epoch": 1.318033381712627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004831804936581518,
      "loss": 5.0917,
      "mean_token_accuracy": 0.20324126482009888,
      "num_tokens": 30160521.0,
      "step": 14530
    },
    {
      "entropy": 5.497390937805176,
      "epoch": 1.3184869375907111,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004831682165759305,
      "loss": 5.0713,
      "mean_token_accuracy": 0.19804712533950805,
      "num_tokens": 30170611.0,
      "step": 14535
    },
    {
      "entropy": 5.481413221359253,
      "epoch": 1.3189404934687954,
      "grad_norm": 1.0,
      "learning_rate": 0.00048315593518873757,
      "loss": 5.0389,
      "mean_token_accuracy": 0.21308245956897737,
      "num_tokens": 30179914.0,
      "step": 14540
    },
    {
      "entropy": 5.426105308532715,
      "epoch": 1.3193940493468794,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004831436494968268,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20314780473709107,
      "num_tokens": 30191426.0,
      "step": 14545
    },
    {
      "entropy": 5.449762868881225,
      "epoch": 1.3198476052249637,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048313135950045234,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2015809252858162,
      "num_tokens": 30202813.0,
      "step": 14550
    },
    {
      "entropy": 5.46356725692749,
      "epoch": 1.320301161103048,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048311906519986824,
      "loss": 4.9521,
      "mean_token_accuracy": 0.20454393625259398,
      "num_tokens": 30213044.0,
      "step": 14555
    },
    {
      "entropy": 5.469812726974487,
      "epoch": 1.320754716981132,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000483106766595329,
      "loss": 5.0443,
      "mean_token_accuracy": 0.19463009834289552,
      "num_tokens": 30223419.0,
      "step": 14560
    },
    {
      "entropy": 5.502302455902099,
      "epoch": 1.3212082728592163,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004830944636870887,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1953321412205696,
      "num_tokens": 30233203.0,
      "step": 14565
    },
    {
      "entropy": 5.46925950050354,
      "epoch": 1.3216618287373003,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048308215647540183,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19541971534490585,
      "num_tokens": 30244604.0,
      "step": 14570
    },
    {
      "entropy": 5.537690687179565,
      "epoch": 1.3221153846153846,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048306984496052297,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19922395199537277,
      "num_tokens": 30255355.0,
      "step": 14575
    },
    {
      "entropy": 5.56852240562439,
      "epoch": 1.3225689404934688,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048305752914270665,
      "loss": 5.1509,
      "mean_token_accuracy": 0.19310900270938874,
      "num_tokens": 30265665.0,
      "step": 14580
    },
    {
      "entropy": 5.525383424758911,
      "epoch": 1.3230224963715531,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048304520902220767,
      "loss": 5.1272,
      "mean_token_accuracy": 0.20262507051229478,
      "num_tokens": 30274347.0,
      "step": 14585
    },
    {
      "entropy": 5.494300079345703,
      "epoch": 1.3234760522496372,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004830328845992808,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2104866087436676,
      "num_tokens": 30284243.0,
      "step": 14590
    },
    {
      "entropy": 5.389914608001709,
      "epoch": 1.3239296081277212,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048302055587418074,
      "loss": 4.9932,
      "mean_token_accuracy": 0.20106371939182283,
      "num_tokens": 30294327.0,
      "step": 14595
    },
    {
      "entropy": 5.408010959625244,
      "epoch": 1.3243831640058055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004830082228471626,
      "loss": 5.0871,
      "mean_token_accuracy": 0.19902536869049073,
      "num_tokens": 30304188.0,
      "step": 14600
    },
    {
      "entropy": 5.46526951789856,
      "epoch": 1.3248367198838897,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048299588551848146,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2069094717502594,
      "num_tokens": 30314917.0,
      "step": 14605
    },
    {
      "entropy": 5.475058126449585,
      "epoch": 1.325290275761974,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048298354388839237,
      "loss": 5.001,
      "mean_token_accuracy": 0.20261586606502532,
      "num_tokens": 30325221.0,
      "step": 14610
    },
    {
      "entropy": 5.51237211227417,
      "epoch": 1.325743831640058,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004829711979571506,
      "loss": 5.1072,
      "mean_token_accuracy": 0.19983938336372375,
      "num_tokens": 30336196.0,
      "step": 14615
    },
    {
      "entropy": 5.4390116214752195,
      "epoch": 1.3261973875181423,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004829588477250115,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2147581547498703,
      "num_tokens": 30345546.0,
      "step": 14620
    },
    {
      "entropy": 5.432073783874512,
      "epoch": 1.3266509433962264,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004829464931922304,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1990799367427826,
      "num_tokens": 30355672.0,
      "step": 14625
    },
    {
      "entropy": 5.4688026905059814,
      "epoch": 1.3271044992743106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048293413435906285,
      "loss": 5.0597,
      "mean_token_accuracy": 0.19611307680606843,
      "num_tokens": 30366233.0,
      "step": 14630
    },
    {
      "entropy": 5.497386026382446,
      "epoch": 1.327558055152395,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004829217712257644,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20204354375600814,
      "num_tokens": 30377366.0,
      "step": 14635
    },
    {
      "entropy": 5.476496028900146,
      "epoch": 1.328011611030479,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048290940379259073,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20131379812955857,
      "num_tokens": 30388022.0,
      "step": 14640
    },
    {
      "entropy": 5.456486654281616,
      "epoch": 1.3284651669085632,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048289703205979756,
      "loss": 4.9573,
      "mean_token_accuracy": 0.19921093732118605,
      "num_tokens": 30399235.0,
      "step": 14645
    },
    {
      "entropy": 5.572162103652954,
      "epoch": 1.3289187227866472,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048288465602764084,
      "loss": 5.1251,
      "mean_token_accuracy": 0.197314490377903,
      "num_tokens": 30409112.0,
      "step": 14650
    },
    {
      "entropy": 5.440092134475708,
      "epoch": 1.3293722786647315,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048287227569637645,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20172401070594786,
      "num_tokens": 30420202.0,
      "step": 14655
    },
    {
      "entropy": 5.446317291259765,
      "epoch": 1.3298258345428158,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048285989106626036,
      "loss": 5.0917,
      "mean_token_accuracy": 0.202370548248291,
      "num_tokens": 30431210.0,
      "step": 14660
    },
    {
      "entropy": 5.453888940811157,
      "epoch": 1.3302793904208998,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004828475021375489,
      "loss": 5.0018,
      "mean_token_accuracy": 0.20676329433918,
      "num_tokens": 30441291.0,
      "step": 14665
    },
    {
      "entropy": 5.445957326889038,
      "epoch": 1.330732946298984,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004828351089104981,
      "loss": 4.9628,
      "mean_token_accuracy": 0.20900101065635682,
      "num_tokens": 30450285.0,
      "step": 14670
    },
    {
      "entropy": 5.398617792129516,
      "epoch": 1.3311865021770681,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048282271138536424,
      "loss": 5.0678,
      "mean_token_accuracy": 0.2006645008921623,
      "num_tokens": 30460311.0,
      "step": 14675
    },
    {
      "entropy": 5.515023851394654,
      "epoch": 1.3316400580551524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004828103095624037,
      "loss": 5.08,
      "mean_token_accuracy": 0.19864129275083542,
      "num_tokens": 30470188.0,
      "step": 14680
    },
    {
      "entropy": 5.424178123474121,
      "epoch": 1.3320936139332367,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000482797903441873,
      "loss": 5.0092,
      "mean_token_accuracy": 0.19827580600976943,
      "num_tokens": 30481406.0,
      "step": 14685
    },
    {
      "entropy": 5.3669273853302,
      "epoch": 1.3325471698113207,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004827854930240288,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20122469812631608,
      "num_tokens": 30492314.0,
      "step": 14690
    },
    {
      "entropy": 5.5031452655792235,
      "epoch": 1.333000725689405,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004827730783091276,
      "loss": 5.1494,
      "mean_token_accuracy": 0.19222668260335923,
      "num_tokens": 30503273.0,
      "step": 14695
    },
    {
      "entropy": 5.5146866798400875,
      "epoch": 1.333454281567489,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048276065929742616,
      "loss": 5.0528,
      "mean_token_accuracy": 0.19601842612028123,
      "num_tokens": 30513582.0,
      "step": 14700
    },
    {
      "entropy": 5.419421625137329,
      "epoch": 1.3339078374455733,
      "grad_norm": 1.125,
      "learning_rate": 0.0004827482359891814,
      "loss": 4.9681,
      "mean_token_accuracy": 0.20890097320079803,
      "num_tokens": 30523261.0,
      "step": 14705
    },
    {
      "entropy": 5.488310194015503,
      "epoch": 1.3343613933236576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048273580838465016,
      "loss": 5.0861,
      "mean_token_accuracy": 0.20133254528045655,
      "num_tokens": 30533774.0,
      "step": 14710
    },
    {
      "entropy": 5.472191381454468,
      "epoch": 1.3348149492017416,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004827233764840894,
      "loss": 5.1304,
      "mean_token_accuracy": 0.19558546394109727,
      "num_tokens": 30544580.0,
      "step": 14715
    },
    {
      "entropy": 5.479812335968018,
      "epoch": 1.3352685050798259,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004827109402877564,
      "loss": 4.977,
      "mean_token_accuracy": 0.20922414660453797,
      "num_tokens": 30555105.0,
      "step": 14720
    },
    {
      "entropy": 5.459916496276856,
      "epoch": 1.33572206095791,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004826984997959082,
      "loss": 5.12,
      "mean_token_accuracy": 0.18927535861730577,
      "num_tokens": 30564912.0,
      "step": 14725
    },
    {
      "entropy": 5.4101158618927006,
      "epoch": 1.3361756168359942,
      "grad_norm": 0.953125,
      "learning_rate": 0.000482686055008802,
      "loss": 5.045,
      "mean_token_accuracy": 0.19525108784437178,
      "num_tokens": 30575862.0,
      "step": 14730
    },
    {
      "entropy": 5.481841659545898,
      "epoch": 1.3366291727140784,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048267360592669535,
      "loss": 5.1607,
      "mean_token_accuracy": 0.19661880582571029,
      "num_tokens": 30586951.0,
      "step": 14735
    },
    {
      "entropy": 5.499304676055909,
      "epoch": 1.3370827285921625,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004826611525498456,
      "loss": 5.0775,
      "mean_token_accuracy": 0.1934316173195839,
      "num_tokens": 30597604.0,
      "step": 14740
    },
    {
      "entropy": 5.521953535079956,
      "epoch": 1.3375362844702468,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004826486948785103,
      "loss": 5.1222,
      "mean_token_accuracy": 0.1946622535586357,
      "num_tokens": 30608157.0,
      "step": 14745
    },
    {
      "entropy": 5.504887104034424,
      "epoch": 1.3379898403483308,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048263623291294706,
      "loss": 5.0639,
      "mean_token_accuracy": 0.20173074901103974,
      "num_tokens": 30618527.0,
      "step": 14750
    },
    {
      "entropy": 5.509787225723267,
      "epoch": 1.338443396226415,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048262376665341355,
      "loss": 5.1546,
      "mean_token_accuracy": 0.1950371593236923,
      "num_tokens": 30629253.0,
      "step": 14755
    },
    {
      "entropy": 5.474182891845703,
      "epoch": 1.3388969521044993,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004826112961001677,
      "loss": 5.1213,
      "mean_token_accuracy": 0.18329809308052064,
      "num_tokens": 30639015.0,
      "step": 14760
    },
    {
      "entropy": 5.5530884742736815,
      "epoch": 1.3393505079825834,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048259882125346735,
      "loss": 5.1857,
      "mean_token_accuracy": 0.18424032181501387,
      "num_tokens": 30650035.0,
      "step": 14765
    },
    {
      "entropy": 5.540981578826904,
      "epoch": 1.3398040638606676,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048258634211357045,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19299462735652922,
      "num_tokens": 30659813.0,
      "step": 14770
    },
    {
      "entropy": 5.463839292526245,
      "epoch": 1.3402576197387517,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004825738586807351,
      "loss": 5.052,
      "mean_token_accuracy": 0.19634215533733368,
      "num_tokens": 30670243.0,
      "step": 14775
    },
    {
      "entropy": 5.429623126983643,
      "epoch": 1.340711175616836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048256137095521945,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20084659308195113,
      "num_tokens": 30680056.0,
      "step": 14780
    },
    {
      "entropy": 5.449593257904053,
      "epoch": 1.3411647314949202,
      "grad_norm": 1.125,
      "learning_rate": 0.00048254887893728175,
      "loss": 4.9581,
      "mean_token_accuracy": 0.20344403088092805,
      "num_tokens": 30690536.0,
      "step": 14785
    },
    {
      "entropy": 5.412839269638061,
      "epoch": 1.3416182873730045,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004825363826271804,
      "loss": 5.0553,
      "mean_token_accuracy": 0.1934947982430458,
      "num_tokens": 30700200.0,
      "step": 14790
    },
    {
      "entropy": 5.435486841201782,
      "epoch": 1.3420718432510885,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004825238820251737,
      "loss": 4.955,
      "mean_token_accuracy": 0.201063634455204,
      "num_tokens": 30710576.0,
      "step": 14795
    },
    {
      "entropy": 5.445123243331909,
      "epoch": 1.3425253991291728,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048251137713152027,
      "loss": 5.0726,
      "mean_token_accuracy": 0.1973447948694229,
      "num_tokens": 30721495.0,
      "step": 14800
    },
    {
      "entropy": 5.471863698959351,
      "epoch": 1.3429789550072568,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004824988679464787,
      "loss": 5.0947,
      "mean_token_accuracy": 0.19781219959259033,
      "num_tokens": 30731700.0,
      "step": 14805
    },
    {
      "entropy": 5.519728946685791,
      "epoch": 1.343432510885341,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048248635447030767,
      "loss": 5.0695,
      "mean_token_accuracy": 0.19798239916563035,
      "num_tokens": 30743421.0,
      "step": 14810
    },
    {
      "entropy": 5.450216913223267,
      "epoch": 1.3438860667634254,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004824738367032659,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21160862743854522,
      "num_tokens": 30754238.0,
      "step": 14815
    },
    {
      "entropy": 5.4210090160369875,
      "epoch": 1.3443396226415094,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004824613146456123,
      "loss": 5.0741,
      "mean_token_accuracy": 0.19306592792272567,
      "num_tokens": 30763938.0,
      "step": 14820
    },
    {
      "entropy": 5.5478850364685055,
      "epoch": 1.3447931785195937,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004824487882976059,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19809957444667817,
      "num_tokens": 30773654.0,
      "step": 14825
    },
    {
      "entropy": 5.549471998214722,
      "epoch": 1.3452467343976777,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048243625765950564,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19516264647245407,
      "num_tokens": 30784659.0,
      "step": 14830
    },
    {
      "entropy": 5.42782187461853,
      "epoch": 1.345700290275762,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004824237227315708,
      "loss": 4.9903,
      "mean_token_accuracy": 0.19554276317358016,
      "num_tokens": 30795194.0,
      "step": 14835
    },
    {
      "entropy": 5.377277421951294,
      "epoch": 1.3461538461538463,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004824111835140604,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21078951060771942,
      "num_tokens": 30805623.0,
      "step": 14840
    },
    {
      "entropy": 5.444870948791504,
      "epoch": 1.3466074020319303,
      "grad_norm": 0.984375,
      "learning_rate": 0.000482398640007234,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20046668946743013,
      "num_tokens": 30814899.0,
      "step": 14845
    },
    {
      "entropy": 5.45310230255127,
      "epoch": 1.3470609579100146,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048238609221135076,
      "loss": 5.1044,
      "mean_token_accuracy": 0.19705725312232972,
      "num_tokens": 30825546.0,
      "step": 14850
    },
    {
      "entropy": 5.390522193908692,
      "epoch": 1.3475145137880986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004823735401266704,
      "loss": 5.0051,
      "mean_token_accuracy": 0.20084163993597032,
      "num_tokens": 30836406.0,
      "step": 14855
    },
    {
      "entropy": 5.426594924926758,
      "epoch": 1.3479680696661829,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048236098375345227,
      "loss": 5.0773,
      "mean_token_accuracy": 0.19645799398422242,
      "num_tokens": 30847966.0,
      "step": 14860
    },
    {
      "entropy": 5.549108076095581,
      "epoch": 1.3484216255442671,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004823484230919563,
      "loss": 5.1136,
      "mean_token_accuracy": 0.19642357230186464,
      "num_tokens": 30856831.0,
      "step": 14865
    },
    {
      "entropy": 5.419999265670777,
      "epoch": 1.3488751814223512,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048233585814244195,
      "loss": 4.9012,
      "mean_token_accuracy": 0.21475642621517183,
      "num_tokens": 30866076.0,
      "step": 14870
    },
    {
      "entropy": 5.456047296524048,
      "epoch": 1.3493287373004355,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004823232889051693,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20405160784721374,
      "num_tokens": 30876845.0,
      "step": 14875
    },
    {
      "entropy": 5.412720108032227,
      "epoch": 1.3497822931785195,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048231071538039816,
      "loss": 5.0493,
      "mean_token_accuracy": 0.19774454087018967,
      "num_tokens": 30887448.0,
      "step": 14880
    },
    {
      "entropy": 5.48894739151001,
      "epoch": 1.3502358490566038,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048229813756838863,
      "loss": 5.0432,
      "mean_token_accuracy": 0.19980101734399797,
      "num_tokens": 30898414.0,
      "step": 14885
    },
    {
      "entropy": 5.447532844543457,
      "epoch": 1.350689404934688,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004822855554694007,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20601948350667953,
      "num_tokens": 30908559.0,
      "step": 14890
    },
    {
      "entropy": 5.48134036064148,
      "epoch": 1.351142960812772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048227296908369475,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20251586586236953,
      "num_tokens": 30918655.0,
      "step": 14895
    },
    {
      "entropy": 5.429618644714355,
      "epoch": 1.3515965166908563,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048226037841153104,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20045506954193115,
      "num_tokens": 30929188.0,
      "step": 14900
    },
    {
      "entropy": 5.479700326919556,
      "epoch": 1.3520500725689404,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004822477834531698,
      "loss": 5.1245,
      "mean_token_accuracy": 0.19964405596256257,
      "num_tokens": 30939902.0,
      "step": 14905
    },
    {
      "entropy": 5.485004711151123,
      "epoch": 1.3525036284470247,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004822351842088716,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20128275007009505,
      "num_tokens": 30951289.0,
      "step": 14910
    },
    {
      "entropy": 5.554483938217163,
      "epoch": 1.352957184325109,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004822225806788969,
      "loss": 5.1911,
      "mean_token_accuracy": 0.18752760589122772,
      "num_tokens": 30961233.0,
      "step": 14915
    },
    {
      "entropy": 5.470905351638794,
      "epoch": 1.353410740203193,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048220997286350654,
      "loss": 5.0064,
      "mean_token_accuracy": 0.20632905662059783,
      "num_tokens": 30971922.0,
      "step": 14920
    },
    {
      "entropy": 5.432500791549683,
      "epoch": 1.3538642960812772,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048219736076296114,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20198279917240142,
      "num_tokens": 30981358.0,
      "step": 14925
    },
    {
      "entropy": 5.508542013168335,
      "epoch": 1.3543178519593613,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004821847443775214,
      "loss": 5.1255,
      "mean_token_accuracy": 0.19548910856246948,
      "num_tokens": 30991984.0,
      "step": 14930
    },
    {
      "entropy": 5.48363094329834,
      "epoch": 1.3547714078374455,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048217212370744853,
      "loss": 5.012,
      "mean_token_accuracy": 0.2035112589597702,
      "num_tokens": 31002790.0,
      "step": 14935
    },
    {
      "entropy": 5.354962682723999,
      "epoch": 1.3552249637155298,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048215949875300327,
      "loss": 4.901,
      "mean_token_accuracy": 0.21152547597885132,
      "num_tokens": 31012827.0,
      "step": 14940
    },
    {
      "entropy": 5.364318990707398,
      "epoch": 1.3556785195936139,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004821468695144467,
      "loss": 5.0066,
      "mean_token_accuracy": 0.19802123308181763,
      "num_tokens": 31023075.0,
      "step": 14945
    },
    {
      "entropy": 5.5149751663208,
      "epoch": 1.3561320754716981,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048213423599204015,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20247676521539687,
      "num_tokens": 31032454.0,
      "step": 14950
    },
    {
      "entropy": 5.6186676025390625,
      "epoch": 1.3565856313497822,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048212159818604477,
      "loss": 5.1925,
      "mean_token_accuracy": 0.1810927450656891,
      "num_tokens": 31042498.0,
      "step": 14955
    },
    {
      "entropy": 5.4338109493255615,
      "epoch": 1.3570391872278664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000482108956096722,
      "loss": 5.0364,
      "mean_token_accuracy": 0.1993908926844597,
      "num_tokens": 31053188.0,
      "step": 14960
    },
    {
      "entropy": 5.442950296401977,
      "epoch": 1.3574927431059507,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048209630972433326,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20058060586452484,
      "num_tokens": 31062840.0,
      "step": 14965
    },
    {
      "entropy": 5.524459314346314,
      "epoch": 1.357946298984035,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048208365906914,
      "loss": 5.168,
      "mean_token_accuracy": 0.19421842992305755,
      "num_tokens": 31072646.0,
      "step": 14970
    },
    {
      "entropy": 5.450298213958741,
      "epoch": 1.358399854862119,
      "grad_norm": 1.125,
      "learning_rate": 0.00048207100413140397,
      "loss": 4.9871,
      "mean_token_accuracy": 0.19912175089120865,
      "num_tokens": 31082700.0,
      "step": 14975
    },
    {
      "entropy": 5.545132780075074,
      "epoch": 1.3588534107402033,
      "grad_norm": 1.0,
      "learning_rate": 0.0004820583449113868,
      "loss": 5.0835,
      "mean_token_accuracy": 0.19726261049509047,
      "num_tokens": 31093743.0,
      "step": 14980
    },
    {
      "entropy": 5.483267974853516,
      "epoch": 1.3593069666182873,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048204568140935023,
      "loss": 5.0115,
      "mean_token_accuracy": 0.20355513095855712,
      "num_tokens": 31104293.0,
      "step": 14985
    },
    {
      "entropy": 5.392659854888916,
      "epoch": 1.3597605224963716,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004820330136255563,
      "loss": 5.0478,
      "mean_token_accuracy": 0.20134875327348709,
      "num_tokens": 31114244.0,
      "step": 14990
    },
    {
      "entropy": 5.4197766304016115,
      "epoch": 1.3602140783744558,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048202034156026683,
      "loss": 5.0153,
      "mean_token_accuracy": 0.19811320751905442,
      "num_tokens": 31123596.0,
      "step": 14995
    },
    {
      "entropy": 5.513525199890137,
      "epoch": 1.36066763425254,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048200766521374396,
      "loss": 5.0825,
      "mean_token_accuracy": 0.18835832476615905,
      "num_tokens": 31134258.0,
      "step": 15000
    },
    {
      "epoch": 1.36066763425254,
      "eval_entropy": 5.275724767451912,
      "eval_loss": 5.111834526062012,
      "eval_mean_token_accuracy": 0.20582324026902607,
      "eval_num_tokens": 31134258.0,
      "eval_runtime": 20.6029,
      "eval_samples_per_second": 1716.262,
      "eval_steps_per_second": 214.533,
      "step": 15000
    },
    {
      "entropy": 5.456406593322754,
      "epoch": 1.3611211901306242,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004819949845862498,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20120365917682648,
      "num_tokens": 31144806.0,
      "step": 15005
    },
    {
      "entropy": 5.505934858322144,
      "epoch": 1.3615747460087082,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004819822996780466,
      "loss": 5.1438,
      "mean_token_accuracy": 0.19504766017198563,
      "num_tokens": 31155642.0,
      "step": 15010
    },
    {
      "entropy": 5.507110118865967,
      "epoch": 1.3620283018867925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004819696104893967,
      "loss": 4.9994,
      "mean_token_accuracy": 0.20176680386066437,
      "num_tokens": 31165034.0,
      "step": 15015
    },
    {
      "entropy": 5.468253564834595,
      "epoch": 1.3624818577648767,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004819569170205625,
      "loss": 5.0514,
      "mean_token_accuracy": 0.19909994155168534,
      "num_tokens": 31176491.0,
      "step": 15020
    },
    {
      "entropy": 5.488339900970459,
      "epoch": 1.3629354136429608,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048194421927180654,
      "loss": 5.068,
      "mean_token_accuracy": 0.19485954642295839,
      "num_tokens": 31186198.0,
      "step": 15025
    },
    {
      "entropy": 5.518307781219482,
      "epoch": 1.363388969521045,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004819315172433914,
      "loss": 5.0467,
      "mean_token_accuracy": 0.20029103010892868,
      "num_tokens": 31197094.0,
      "step": 15030
    },
    {
      "entropy": 5.494000768661499,
      "epoch": 1.363842525399129,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048191881093557977,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19154375940561294,
      "num_tokens": 31207505.0,
      "step": 15035
    },
    {
      "entropy": 5.5333664894104,
      "epoch": 1.3642960812772134,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048190610034863434,
      "loss": 5.1862,
      "mean_token_accuracy": 0.19173948913812638,
      "num_tokens": 31218809.0,
      "step": 15040
    },
    {
      "entropy": 5.508054113388061,
      "epoch": 1.3647496371552976,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004818933854828181,
      "loss": 5.1105,
      "mean_token_accuracy": 0.1877943217754364,
      "num_tokens": 31230281.0,
      "step": 15045
    },
    {
      "entropy": 5.43084864616394,
      "epoch": 1.3652031930333817,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004818806663383939,
      "loss": 4.9882,
      "mean_token_accuracy": 0.20633529424667357,
      "num_tokens": 31239573.0,
      "step": 15050
    },
    {
      "entropy": 5.463028526306152,
      "epoch": 1.365656748911466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004818679429156248,
      "loss": 5.028,
      "mean_token_accuracy": 0.19652057439088821,
      "num_tokens": 31249671.0,
      "step": 15055
    },
    {
      "entropy": 5.4452272891998295,
      "epoch": 1.36611030478955,
      "grad_norm": 1.0,
      "learning_rate": 0.00048185521521477385,
      "loss": 5.0508,
      "mean_token_accuracy": 0.19330381453037263,
      "num_tokens": 31259601.0,
      "step": 15060
    },
    {
      "entropy": 5.439665651321411,
      "epoch": 1.3665638606676342,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004818424832361044,
      "loss": 5.016,
      "mean_token_accuracy": 0.20150518268346787,
      "num_tokens": 31269085.0,
      "step": 15065
    },
    {
      "entropy": 5.419748544692993,
      "epoch": 1.3670174165457185,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004818297469798797,
      "loss": 5.0592,
      "mean_token_accuracy": 0.19574157893657684,
      "num_tokens": 31280177.0,
      "step": 15070
    },
    {
      "entropy": 5.5028900623321535,
      "epoch": 1.3674709724238026,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048181700644636314,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2023068070411682,
      "num_tokens": 31290545.0,
      "step": 15075
    },
    {
      "entropy": 5.394643783569336,
      "epoch": 1.3679245283018868,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048180426163581816,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21144856065511702,
      "num_tokens": 31301143.0,
      "step": 15080
    },
    {
      "entropy": 5.434935140609741,
      "epoch": 1.3683780841799709,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004817915125485083,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20205068588256836,
      "num_tokens": 31311071.0,
      "step": 15085
    },
    {
      "entropy": 5.510846424102783,
      "epoch": 1.3688316400580551,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048177875918469735,
      "loss": 5.113,
      "mean_token_accuracy": 0.19890975058078766,
      "num_tokens": 31322263.0,
      "step": 15090
    },
    {
      "entropy": 5.495527935028076,
      "epoch": 1.3692851959361394,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048176600154464897,
      "loss": 5.0074,
      "mean_token_accuracy": 0.201681686937809,
      "num_tokens": 31332975.0,
      "step": 15095
    },
    {
      "entropy": 5.548653268814087,
      "epoch": 1.3697387518142234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004817532396286269,
      "loss": 5.1204,
      "mean_token_accuracy": 0.1935277834534645,
      "num_tokens": 31345252.0,
      "step": 15100
    },
    {
      "entropy": 5.489936208724975,
      "epoch": 1.3701923076923077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004817404734368952,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21002684384584427,
      "num_tokens": 31355479.0,
      "step": 15105
    },
    {
      "entropy": 5.509968948364258,
      "epoch": 1.3706458635703918,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048172770296971785,
      "loss": 5.124,
      "mean_token_accuracy": 0.19348738938570023,
      "num_tokens": 31366020.0,
      "step": 15110
    },
    {
      "entropy": 5.490027999877929,
      "epoch": 1.371099419448476,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004817149282273589,
      "loss": 5.0569,
      "mean_token_accuracy": 0.20011508613824844,
      "num_tokens": 31376055.0,
      "step": 15115
    },
    {
      "entropy": 5.490016269683838,
      "epoch": 1.3715529753265603,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004817021492100825,
      "loss": 5.1669,
      "mean_token_accuracy": 0.1912306636571884,
      "num_tokens": 31386179.0,
      "step": 15120
    },
    {
      "entropy": 5.437982177734375,
      "epoch": 1.3720065312046443,
      "grad_norm": 1.015625,
      "learning_rate": 0.000481689365918153,
      "loss": 4.9713,
      "mean_token_accuracy": 0.20575360506772994,
      "num_tokens": 31395850.0,
      "step": 15125
    },
    {
      "entropy": 5.469276380538941,
      "epoch": 1.3724600870827286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004816765783518348,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2126341849565506,
      "num_tokens": 31405053.0,
      "step": 15130
    },
    {
      "entropy": 5.424975299835205,
      "epoch": 1.3729136429608126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004816637865113923,
      "loss": 4.9868,
      "mean_token_accuracy": 0.2035997673869133,
      "num_tokens": 31415153.0,
      "step": 15135
    },
    {
      "entropy": 5.450738382339478,
      "epoch": 1.373367198838897,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048165099039708994,
      "loss": 5.0448,
      "mean_token_accuracy": 0.1943662479519844,
      "num_tokens": 31425451.0,
      "step": 15140
    },
    {
      "entropy": 5.429855871200561,
      "epoch": 1.3738207547169812,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004816381900091925,
      "loss": 5.0263,
      "mean_token_accuracy": 0.19575783163309096,
      "num_tokens": 31436298.0,
      "step": 15145
    },
    {
      "entropy": 5.482528924942017,
      "epoch": 1.3742743105950654,
      "grad_norm": 1.0,
      "learning_rate": 0.00048162538534796465,
      "loss": 5.0497,
      "mean_token_accuracy": 0.2034153923392296,
      "num_tokens": 31446932.0,
      "step": 15150
    },
    {
      "entropy": 5.5162004947662355,
      "epoch": 1.3747278664731495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004816125764136711,
      "loss": 5.0866,
      "mean_token_accuracy": 0.18720396608114243,
      "num_tokens": 31457582.0,
      "step": 15155
    },
    {
      "entropy": 5.542390441894531,
      "epoch": 1.3751814223512338,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004815997632065769,
      "loss": 5.0513,
      "mean_token_accuracy": 0.19751511067152022,
      "num_tokens": 31468168.0,
      "step": 15160
    },
    {
      "entropy": 5.435163164138794,
      "epoch": 1.3756349782293178,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048158694572694694,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20532604455947875,
      "num_tokens": 31478524.0,
      "step": 15165
    },
    {
      "entropy": 5.4072364330291744,
      "epoch": 1.376088534107402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048157412397504623,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20544958263635635,
      "num_tokens": 31488608.0,
      "step": 15170
    },
    {
      "entropy": 5.451086902618409,
      "epoch": 1.3765420899854863,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004815612979511401,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20353641360998154,
      "num_tokens": 31498984.0,
      "step": 15175
    },
    {
      "entropy": 5.523633098602295,
      "epoch": 1.3769956458635704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048154846765549353,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20317785888910295,
      "num_tokens": 31509110.0,
      "step": 15180
    },
    {
      "entropy": 5.476151084899902,
      "epoch": 1.3774492017416546,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048153563308837215,
      "loss": 5.0368,
      "mean_token_accuracy": 0.20274375528097152,
      "num_tokens": 31519266.0,
      "step": 15185
    },
    {
      "entropy": 5.4552610397338865,
      "epoch": 1.3779027576197387,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048152279425004124,
      "loss": 5.0538,
      "mean_token_accuracy": 0.20143594443798066,
      "num_tokens": 31529043.0,
      "step": 15190
    },
    {
      "entropy": 5.483608388900757,
      "epoch": 1.378356313497823,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004815099511407663,
      "loss": 5.0393,
      "mean_token_accuracy": 0.2035144805908203,
      "num_tokens": 31538650.0,
      "step": 15195
    },
    {
      "entropy": 5.400378751754761,
      "epoch": 1.3788098693759072,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000481497103760813,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20239569693803788,
      "num_tokens": 31549392.0,
      "step": 15200
    },
    {
      "entropy": 5.51238226890564,
      "epoch": 1.3792634252539913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048148425211044684,
      "loss": 5.1609,
      "mean_token_accuracy": 0.18882282227277755,
      "num_tokens": 31559999.0,
      "step": 15205
    },
    {
      "entropy": 5.533956813812256,
      "epoch": 1.3797169811320755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004814713961899338,
      "loss": 5.0553,
      "mean_token_accuracy": 0.19863239973783492,
      "num_tokens": 31570108.0,
      "step": 15210
    },
    {
      "entropy": 5.471551513671875,
      "epoch": 1.3801705370101596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048145853599953966,
      "loss": 5.0164,
      "mean_token_accuracy": 0.1964831069111824,
      "num_tokens": 31579698.0,
      "step": 15215
    },
    {
      "entropy": 5.4328672885894775,
      "epoch": 1.3806240928882438,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048144567153953036,
      "loss": 5.0713,
      "mean_token_accuracy": 0.19548798054456712,
      "num_tokens": 31589292.0,
      "step": 15220
    },
    {
      "entropy": 5.529256439208984,
      "epoch": 1.381077648766328,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048143280281017197,
      "loss": 5.1392,
      "mean_token_accuracy": 0.194683501124382,
      "num_tokens": 31599848.0,
      "step": 15225
    },
    {
      "entropy": 5.525971078872681,
      "epoch": 1.3815312046444121,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004814199298117307,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20013927072286605,
      "num_tokens": 31611109.0,
      "step": 15230
    },
    {
      "entropy": 5.412837266921997,
      "epoch": 1.3819847605224964,
      "grad_norm": 1.0,
      "learning_rate": 0.00048140705254447256,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20205961763858796,
      "num_tokens": 31621140.0,
      "step": 15235
    },
    {
      "entropy": 5.52912654876709,
      "epoch": 1.3824383164005805,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048139417100866405,
      "loss": 5.0911,
      "mean_token_accuracy": 0.1939746081829071,
      "num_tokens": 31631268.0,
      "step": 15240
    },
    {
      "entropy": 5.528049659729004,
      "epoch": 1.3828918722786647,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048138128520457146,
      "loss": 5.0372,
      "mean_token_accuracy": 0.19948768615722656,
      "num_tokens": 31640758.0,
      "step": 15245
    },
    {
      "entropy": 5.4472252368927006,
      "epoch": 1.383345428156749,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004813683951324613,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19628065526485444,
      "num_tokens": 31651060.0,
      "step": 15250
    },
    {
      "entropy": 5.495986890792847,
      "epoch": 1.383798984034833,
      "grad_norm": 1.0,
      "learning_rate": 0.00048135550079260014,
      "loss": 5.1384,
      "mean_token_accuracy": 0.19810788631439208,
      "num_tokens": 31662282.0,
      "step": 15255
    },
    {
      "entropy": 5.462776374816895,
      "epoch": 1.3842525399129173,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004813426021852547,
      "loss": 5.009,
      "mean_token_accuracy": 0.19480444192886354,
      "num_tokens": 31672364.0,
      "step": 15260
    },
    {
      "entropy": 5.481860637664795,
      "epoch": 1.3847060957910013,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048132969931069156,
      "loss": 5.0597,
      "mean_token_accuracy": 0.19527684450149535,
      "num_tokens": 31683094.0,
      "step": 15265
    },
    {
      "entropy": 5.486013317108155,
      "epoch": 1.3851596516690856,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004813167921691777,
      "loss": 5.035,
      "mean_token_accuracy": 0.20312470048666,
      "num_tokens": 31693087.0,
      "step": 15270
    },
    {
      "entropy": 5.480851173400879,
      "epoch": 1.3856132075471699,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048130388076098003,
      "loss": 5.0259,
      "mean_token_accuracy": 0.19591172337532042,
      "num_tokens": 31703708.0,
      "step": 15275
    },
    {
      "entropy": 5.413898563385009,
      "epoch": 1.386066763425254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048129096508636547,
      "loss": 4.9689,
      "mean_token_accuracy": 0.20569125115871428,
      "num_tokens": 31713640.0,
      "step": 15280
    },
    {
      "entropy": 5.476882982254028,
      "epoch": 1.3865203193033382,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048127804514560126,
      "loss": 5.0702,
      "mean_token_accuracy": 0.19898393005132675,
      "num_tokens": 31723526.0,
      "step": 15285
    },
    {
      "entropy": 5.417530298233032,
      "epoch": 1.3869738751814222,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048126512093895447,
      "loss": 4.9925,
      "mean_token_accuracy": 0.20380101948976517,
      "num_tokens": 31733265.0,
      "step": 15290
    },
    {
      "entropy": 5.482056903839111,
      "epoch": 1.3874274310595065,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004812521924666925,
      "loss": 4.9968,
      "mean_token_accuracy": 0.20224116146564483,
      "num_tokens": 31743832.0,
      "step": 15295
    },
    {
      "entropy": 5.508058977127075,
      "epoch": 1.3878809869375908,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004812392597290825,
      "loss": 5.0582,
      "mean_token_accuracy": 0.21130288690328597,
      "num_tokens": 31754443.0,
      "step": 15300
    },
    {
      "entropy": 5.400997304916382,
      "epoch": 1.3883345428156748,
      "grad_norm": 1.0,
      "learning_rate": 0.0004812263227263921,
      "loss": 5.0241,
      "mean_token_accuracy": 0.19950023740530015,
      "num_tokens": 31765304.0,
      "step": 15305
    },
    {
      "entropy": 5.395258903503418,
      "epoch": 1.388788098693759,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004812133814588888,
      "loss": 5.0624,
      "mean_token_accuracy": 0.19518595784902573,
      "num_tokens": 31775809.0,
      "step": 15310
    },
    {
      "entropy": 5.463767671585083,
      "epoch": 1.3892416545718431,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004812004359268402,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20550617277622224,
      "num_tokens": 31786271.0,
      "step": 15315
    },
    {
      "entropy": 5.524199151992798,
      "epoch": 1.3896952104499274,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004811874861305141,
      "loss": 5.0928,
      "mean_token_accuracy": 0.20250213742256165,
      "num_tokens": 31797358.0,
      "step": 15320
    },
    {
      "entropy": 5.477554082870483,
      "epoch": 1.3901487663280117,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004811745320701782,
      "loss": 5.005,
      "mean_token_accuracy": 0.20689014941453934,
      "num_tokens": 31808303.0,
      "step": 15325
    },
    {
      "entropy": 5.406842136383057,
      "epoch": 1.390602322206096,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004811615737461004,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20240322798490523,
      "num_tokens": 31818837.0,
      "step": 15330
    },
    {
      "entropy": 5.503112363815307,
      "epoch": 1.39105587808418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048114861115854877,
      "loss": 5.0376,
      "mean_token_accuracy": 0.19811082929372786,
      "num_tokens": 31829729.0,
      "step": 15335
    },
    {
      "entropy": 5.46115345954895,
      "epoch": 1.3915094339622642,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048113564430779136,
      "loss": 4.9801,
      "mean_token_accuracy": 0.19977615475654603,
      "num_tokens": 31839996.0,
      "step": 15340
    },
    {
      "entropy": 5.373750066757202,
      "epoch": 1.3919629898403483,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004811226731940962,
      "loss": 4.9753,
      "mean_token_accuracy": 0.20393562465906143,
      "num_tokens": 31850298.0,
      "step": 15345
    },
    {
      "entropy": 5.4442449569702145,
      "epoch": 1.3924165457184325,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048110969781773173,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2088144063949585,
      "num_tokens": 31862170.0,
      "step": 15350
    },
    {
      "entropy": 5.551015758514405,
      "epoch": 1.3928701015965168,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048109671817896614,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19858555495738983,
      "num_tokens": 31872081.0,
      "step": 15355
    },
    {
      "entropy": 5.462356090545654,
      "epoch": 1.3933236574746009,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048108373427806797,
      "loss": 5.043,
      "mean_token_accuracy": 0.2021666869521141,
      "num_tokens": 31882715.0,
      "step": 15360
    },
    {
      "entropy": 5.470950031280518,
      "epoch": 1.3937772133526851,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048107074611530555,
      "loss": 5.0286,
      "mean_token_accuracy": 0.2057737648487091,
      "num_tokens": 31891925.0,
      "step": 15365
    },
    {
      "entropy": 5.505783176422119,
      "epoch": 1.3942307692307692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048105775369094756,
      "loss": 5.0476,
      "mean_token_accuracy": 0.19917611926794052,
      "num_tokens": 31902320.0,
      "step": 15370
    },
    {
      "entropy": 5.4439263343811035,
      "epoch": 1.3946843251088534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004810447570052628,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2073088452219963,
      "num_tokens": 31912726.0,
      "step": 15375
    },
    {
      "entropy": 5.4545204639434814,
      "epoch": 1.3951378809869377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048103175605851997,
      "loss": 5.0646,
      "mean_token_accuracy": 0.1981646478176117,
      "num_tokens": 31924039.0,
      "step": 15380
    },
    {
      "entropy": 5.380268859863281,
      "epoch": 1.3955914368650217,
      "grad_norm": 1.0,
      "learning_rate": 0.0004810187508509879,
      "loss": 4.9447,
      "mean_token_accuracy": 0.21212603002786637,
      "num_tokens": 31933280.0,
      "step": 15385
    },
    {
      "entropy": 5.469443225860596,
      "epoch": 1.396044992743106,
      "grad_norm": 1.0,
      "learning_rate": 0.0004810057413829355,
      "loss": 5.1515,
      "mean_token_accuracy": 0.19671646058559417,
      "num_tokens": 31943182.0,
      "step": 15390
    },
    {
      "entropy": 5.516176080703735,
      "epoch": 1.39649854862119,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048099272765463193,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20332831591367723,
      "num_tokens": 31953256.0,
      "step": 15395
    },
    {
      "entropy": 5.561001348495483,
      "epoch": 1.3969521044992743,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004809797096663463,
      "loss": 5.1809,
      "mean_token_accuracy": 0.19255884289741515,
      "num_tokens": 31964231.0,
      "step": 15400
    },
    {
      "entropy": 5.494784498214722,
      "epoch": 1.3974056603773586,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004809666874183477,
      "loss": 5.0164,
      "mean_token_accuracy": 0.1996073842048645,
      "num_tokens": 31974322.0,
      "step": 15405
    },
    {
      "entropy": 5.449486875534058,
      "epoch": 1.3978592162554426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004809536609109056,
      "loss": 5.0634,
      "mean_token_accuracy": 0.20428977459669112,
      "num_tokens": 31984239.0,
      "step": 15410
    },
    {
      "entropy": 5.438951206207276,
      "epoch": 1.398312772133527,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048094063014428925,
      "loss": 5.0689,
      "mean_token_accuracy": 0.19528097063302993,
      "num_tokens": 31994366.0,
      "step": 15415
    },
    {
      "entropy": 5.548461151123047,
      "epoch": 1.398766328011611,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004809275951187682,
      "loss": 5.1392,
      "mean_token_accuracy": 0.19679973125457764,
      "num_tokens": 32004921.0,
      "step": 15420
    },
    {
      "entropy": 5.449678993225097,
      "epoch": 1.3992198838896952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048091455583461204,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20041013807058333,
      "num_tokens": 32016385.0,
      "step": 15425
    },
    {
      "entropy": 5.460604858398438,
      "epoch": 1.3996734397677795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048090151229209036,
      "loss": 5.0646,
      "mean_token_accuracy": 0.19743129163980483,
      "num_tokens": 32027398.0,
      "step": 15430
    },
    {
      "entropy": 5.5013045310974125,
      "epoch": 1.4001269956458635,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004808884644914729,
      "loss": 5.0638,
      "mean_token_accuracy": 0.20003216117620468,
      "num_tokens": 32038326.0,
      "step": 15435
    },
    {
      "entropy": 5.484676790237427,
      "epoch": 1.4005805515239478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004808754124330296,
      "loss": 5.1221,
      "mean_token_accuracy": 0.19728430211544037,
      "num_tokens": 32048744.0,
      "step": 15440
    },
    {
      "entropy": 5.480916595458984,
      "epoch": 1.4010341074020318,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004808623561170302,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2117033764719963,
      "num_tokens": 32059133.0,
      "step": 15445
    },
    {
      "entropy": 5.453863954544067,
      "epoch": 1.401487663280116,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004808492955437449,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20180729478597642,
      "num_tokens": 32069277.0,
      "step": 15450
    },
    {
      "entropy": 5.47017912864685,
      "epoch": 1.4019412191582004,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048083623071344374,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20217232257127762,
      "num_tokens": 32079802.0,
      "step": 15455
    },
    {
      "entropy": 5.472327995300293,
      "epoch": 1.4023947750362844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048082316162639685,
      "loss": 5.1019,
      "mean_token_accuracy": 0.19852525889873504,
      "num_tokens": 32090962.0,
      "step": 15460
    },
    {
      "entropy": 5.458762836456299,
      "epoch": 1.4028483309143687,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048081008828287447,
      "loss": 5.0714,
      "mean_token_accuracy": 0.19673923403024673,
      "num_tokens": 32101058.0,
      "step": 15465
    },
    {
      "entropy": 5.529480838775635,
      "epoch": 1.4033018867924527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048079701068314706,
      "loss": 5.067,
      "mean_token_accuracy": 0.20346747487783431,
      "num_tokens": 32111763.0,
      "step": 15470
    },
    {
      "entropy": 5.504124450683594,
      "epoch": 1.403755442670537,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048078392882748507,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19799258857965468,
      "num_tokens": 32123556.0,
      "step": 15475
    },
    {
      "entropy": 5.4855447769165036,
      "epoch": 1.4042089985486212,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048077084271615895,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20316202342510223,
      "num_tokens": 32134012.0,
      "step": 15480
    },
    {
      "entropy": 5.389006567001343,
      "epoch": 1.4046625544267053,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004807577523494394,
      "loss": 4.9911,
      "mean_token_accuracy": 0.2130587175488472,
      "num_tokens": 32144540.0,
      "step": 15485
    },
    {
      "entropy": 5.4568949222564695,
      "epoch": 1.4051161103047896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048074465772759707,
      "loss": 5.1083,
      "mean_token_accuracy": 0.19622644633054734,
      "num_tokens": 32154655.0,
      "step": 15490
    },
    {
      "entropy": 5.44367094039917,
      "epoch": 1.4055696661828736,
      "grad_norm": 1.0,
      "learning_rate": 0.00048073155885090276,
      "loss": 4.9845,
      "mean_token_accuracy": 0.20090605169534684,
      "num_tokens": 32165191.0,
      "step": 15495
    },
    {
      "entropy": 5.490029430389404,
      "epoch": 1.4060232220609579,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048071845571962733,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19338328689336776,
      "num_tokens": 32175961.0,
      "step": 15500
    },
    {
      "entropy": 5.523004674911499,
      "epoch": 1.4064767779390421,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004807053483340419,
      "loss": 5.0543,
      "mean_token_accuracy": 0.20064833015203476,
      "num_tokens": 32187073.0,
      "step": 15505
    },
    {
      "entropy": 5.489319515228272,
      "epoch": 1.4069303338171264,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004806922366944175,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20276428312063216,
      "num_tokens": 32197426.0,
      "step": 15510
    },
    {
      "entropy": 5.542129039764404,
      "epoch": 1.4073838896952104,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048067912080102513,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20442990511655806,
      "num_tokens": 32206847.0,
      "step": 15515
    },
    {
      "entropy": 5.4296956062316895,
      "epoch": 1.4078374455732947,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004806660006541362,
      "loss": 5.0558,
      "mean_token_accuracy": 0.19724748134613038,
      "num_tokens": 32217647.0,
      "step": 15520
    },
    {
      "entropy": 5.371573066711425,
      "epoch": 1.4082910014513788,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004806528762540219,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21202264577150345,
      "num_tokens": 32227133.0,
      "step": 15525
    },
    {
      "entropy": 5.496533393859863,
      "epoch": 1.408744557329463,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048063974760095374,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20068910866975784,
      "num_tokens": 32237378.0,
      "step": 15530
    },
    {
      "entropy": 5.445556259155273,
      "epoch": 1.4091981132075473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048062661469520327,
      "loss": 5.003,
      "mean_token_accuracy": 0.2052423760294914,
      "num_tokens": 32247213.0,
      "step": 15535
    },
    {
      "entropy": 5.518263244628907,
      "epoch": 1.4096516690856313,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048061347753704193,
      "loss": 5.0724,
      "mean_token_accuracy": 0.2015862599015236,
      "num_tokens": 32257100.0,
      "step": 15540
    },
    {
      "entropy": 5.474034023284912,
      "epoch": 1.4101052249637156,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004806003361267415,
      "loss": 5.0426,
      "mean_token_accuracy": 0.19485013782978058,
      "num_tokens": 32267456.0,
      "step": 15545
    },
    {
      "entropy": 5.427051591873169,
      "epoch": 1.4105587808417996,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004805871904645737,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20818628072738649,
      "num_tokens": 32278220.0,
      "step": 15550
    },
    {
      "entropy": 5.4042943000793455,
      "epoch": 1.411012336719884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004805740405508103,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20270225405693054,
      "num_tokens": 32289036.0,
      "step": 15555
    },
    {
      "entropy": 5.419226264953613,
      "epoch": 1.4114658925979682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004805608863857236,
      "loss": 4.9856,
      "mean_token_accuracy": 0.1986078605055809,
      "num_tokens": 32298909.0,
      "step": 15560
    },
    {
      "entropy": 5.476052188873291,
      "epoch": 1.4119194484760522,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048054772796958517,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20129195749759674,
      "num_tokens": 32309851.0,
      "step": 15565
    },
    {
      "entropy": 5.433191394805908,
      "epoch": 1.4123730043541365,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004805345653026675,
      "loss": 4.989,
      "mean_token_accuracy": 0.20712781250476836,
      "num_tokens": 32319336.0,
      "step": 15570
    },
    {
      "entropy": 5.4757801532745365,
      "epoch": 1.4128265602322205,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048052139838524263,
      "loss": 5.034,
      "mean_token_accuracy": 0.20390822142362594,
      "num_tokens": 32329898.0,
      "step": 15575
    },
    {
      "entropy": 5.436186265945435,
      "epoch": 1.4132801161103048,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048050822721758274,
      "loss": 4.927,
      "mean_token_accuracy": 0.1992727115750313,
      "num_tokens": 32340100.0,
      "step": 15580
    },
    {
      "entropy": 5.454240322113037,
      "epoch": 1.413733671988389,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004804950517999604,
      "loss": 5.0396,
      "mean_token_accuracy": 0.1967187151312828,
      "num_tokens": 32349663.0,
      "step": 15585
    },
    {
      "entropy": 5.503654623031617,
      "epoch": 1.414187227866473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004804818721326481,
      "loss": 5.069,
      "mean_token_accuracy": 0.2061235189437866,
      "num_tokens": 32360461.0,
      "step": 15590
    },
    {
      "entropy": 5.490416049957275,
      "epoch": 1.4146407837445574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004804686882159183,
      "loss": 5.0824,
      "mean_token_accuracy": 0.19882456213235855,
      "num_tokens": 32370910.0,
      "step": 15595
    },
    {
      "entropy": 5.500588846206665,
      "epoch": 1.4150943396226414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048045550005004355,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20531621873378753,
      "num_tokens": 32381669.0,
      "step": 15600
    },
    {
      "entropy": 5.435694789886474,
      "epoch": 1.4155478955007257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004804423076352968,
      "loss": 5.0115,
      "mean_token_accuracy": 0.20064838975667953,
      "num_tokens": 32392024.0,
      "step": 15605
    },
    {
      "entropy": 5.399255847930908,
      "epoch": 1.41600145137881,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048042911097195075,
      "loss": 5.0247,
      "mean_token_accuracy": 0.20648539066314697,
      "num_tokens": 32401872.0,
      "step": 15610
    },
    {
      "entropy": 5.384214258193969,
      "epoch": 1.416455007256894,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004804159100602784,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2077912986278534,
      "num_tokens": 32411749.0,
      "step": 15615
    },
    {
      "entropy": 5.557152605056762,
      "epoch": 1.4169085631349783,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048040270490055264,
      "loss": 5.1548,
      "mean_token_accuracy": 0.19015388786792756,
      "num_tokens": 32421358.0,
      "step": 15620
    },
    {
      "entropy": 5.519060230255127,
      "epoch": 1.4173621190130623,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004803894954930465,
      "loss": 5.0508,
      "mean_token_accuracy": 0.19911952018737794,
      "num_tokens": 32432322.0,
      "step": 15625
    },
    {
      "entropy": 5.472004652023315,
      "epoch": 1.4178156748911466,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004803762818380334,
      "loss": 5.0694,
      "mean_token_accuracy": 0.19749792218208312,
      "num_tokens": 32440930.0,
      "step": 15630
    },
    {
      "entropy": 5.486641693115234,
      "epoch": 1.4182692307692308,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048036306393578644,
      "loss": 5.0702,
      "mean_token_accuracy": 0.19684210419654846,
      "num_tokens": 32451124.0,
      "step": 15635
    },
    {
      "entropy": 5.513130187988281,
      "epoch": 1.4187227866473149,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004803498417865789,
      "loss": 5.0444,
      "mean_token_accuracy": 0.20012172907590867,
      "num_tokens": 32460636.0,
      "step": 15640
    },
    {
      "entropy": 5.513883113861084,
      "epoch": 1.4191763425253991,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004803366153906844,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1964539885520935,
      "num_tokens": 32470501.0,
      "step": 15645
    },
    {
      "entropy": 5.506065940856933,
      "epoch": 1.4196298984034832,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004803233847483763,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19621914476156235,
      "num_tokens": 32481559.0,
      "step": 15650
    },
    {
      "entropy": 5.58727011680603,
      "epoch": 1.4200834542815675,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048031014985992825,
      "loss": 5.2193,
      "mean_token_accuracy": 0.1937747061252594,
      "num_tokens": 32493000.0,
      "step": 15655
    },
    {
      "entropy": 5.50272855758667,
      "epoch": 1.4205370101596517,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000480296910725614,
      "loss": 5.016,
      "mean_token_accuracy": 0.2094208136200905,
      "num_tokens": 32502773.0,
      "step": 15660
    },
    {
      "entropy": 5.4724994659423825,
      "epoch": 1.4209905660377358,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004802836673457074,
      "loss": 5.114,
      "mean_token_accuracy": 0.1955271691083908,
      "num_tokens": 32513430.0,
      "step": 15665
    },
    {
      "entropy": 5.415779685974121,
      "epoch": 1.42144412191582,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048027041972048225,
      "loss": 5.014,
      "mean_token_accuracy": 0.20310366600751878,
      "num_tokens": 32524409.0,
      "step": 15670
    },
    {
      "entropy": 5.523543453216552,
      "epoch": 1.421897677793904,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004802571678502124,
      "loss": 5.0637,
      "mean_token_accuracy": 0.19521717131137847,
      "num_tokens": 32534696.0,
      "step": 15675
    },
    {
      "entropy": 5.557691240310669,
      "epoch": 1.4223512336719883,
      "grad_norm": 0.921875,
      "learning_rate": 0.000480243911735172,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20486830770969391,
      "num_tokens": 32544782.0,
      "step": 15680
    },
    {
      "entropy": 5.4342268943786625,
      "epoch": 1.4228047895500726,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048023065137563534,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1975138857960701,
      "num_tokens": 32554818.0,
      "step": 15685
    },
    {
      "entropy": 5.446495103836059,
      "epoch": 1.4232583454281569,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004802173867718764,
      "loss": 5.0916,
      "mean_token_accuracy": 0.19939793050289153,
      "num_tokens": 32565007.0,
      "step": 15690
    },
    {
      "entropy": 5.508305311203003,
      "epoch": 1.423711901306241,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004802041179241697,
      "loss": 5.0222,
      "mean_token_accuracy": 0.20239437520503997,
      "num_tokens": 32575657.0,
      "step": 15695
    },
    {
      "entropy": 5.482709550857544,
      "epoch": 1.4241654571843252,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004801908448327895,
      "loss": 5.0802,
      "mean_token_accuracy": 0.19548770189285278,
      "num_tokens": 32586676.0,
      "step": 15700
    },
    {
      "entropy": 5.481083011627197,
      "epoch": 1.4246190130624092,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004801775674980103,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2019490569829941,
      "num_tokens": 32596191.0,
      "step": 15705
    },
    {
      "entropy": 5.517948341369629,
      "epoch": 1.4250725689404935,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048016428592010676,
      "loss": 5.0849,
      "mean_token_accuracy": 0.1982520654797554,
      "num_tokens": 32607178.0,
      "step": 15710
    },
    {
      "entropy": 5.50854721069336,
      "epoch": 1.4255261248185778,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048015100009935347,
      "loss": 5.019,
      "mean_token_accuracy": 0.20194409042596817,
      "num_tokens": 32617301.0,
      "step": 15715
    },
    {
      "entropy": 5.397899341583252,
      "epoch": 1.4259796806966618,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004801377100360253,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20444879978895186,
      "num_tokens": 32627395.0,
      "step": 15720
    },
    {
      "entropy": 5.425991058349609,
      "epoch": 1.426433236574746,
      "grad_norm": 0.953125,
      "learning_rate": 0.000480124415730397,
      "loss": 5.0305,
      "mean_token_accuracy": 0.20138730853796005,
      "num_tokens": 32638460.0,
      "step": 15725
    },
    {
      "entropy": 5.538776016235351,
      "epoch": 1.4268867924528301,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004801111171827434,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19654250293970107,
      "num_tokens": 32649319.0,
      "step": 15730
    },
    {
      "entropy": 5.501091003417969,
      "epoch": 1.4273403483309144,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004800978143933398,
      "loss": 5.0367,
      "mean_token_accuracy": 0.20251665711402894,
      "num_tokens": 32659976.0,
      "step": 15735
    },
    {
      "entropy": 5.428359079360962,
      "epoch": 1.4277939042089987,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048008450736246094,
      "loss": 4.9659,
      "mean_token_accuracy": 0.19979297667741774,
      "num_tokens": 32670772.0,
      "step": 15740
    },
    {
      "entropy": 5.4992910861969,
      "epoch": 1.4282474600870827,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048007119609038237,
      "loss": 5.1025,
      "mean_token_accuracy": 0.19040610939264296,
      "num_tokens": 32682368.0,
      "step": 15745
    },
    {
      "entropy": 5.393808746337891,
      "epoch": 1.428701015965167,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048005788057737907,
      "loss": 4.9348,
      "mean_token_accuracy": 0.20572381764650344,
      "num_tokens": 32692649.0,
      "step": 15750
    },
    {
      "entropy": 5.491480875015259,
      "epoch": 1.429154571843251,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004800445608237266,
      "loss": 5.1442,
      "mean_token_accuracy": 0.20096337646245957,
      "num_tokens": 32704075.0,
      "step": 15755
    },
    {
      "entropy": 5.547732925415039,
      "epoch": 1.4296081277213353,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048003123682970045,
      "loss": 5.1663,
      "mean_token_accuracy": 0.1921308755874634,
      "num_tokens": 32714670.0,
      "step": 15760
    },
    {
      "entropy": 5.468905401229859,
      "epoch": 1.4300616835994195,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048001790859557596,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21046035885810851,
      "num_tokens": 32724601.0,
      "step": 15765
    },
    {
      "entropy": 5.418109798431397,
      "epoch": 1.4305152394775036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004800045761216289,
      "loss": 4.9591,
      "mean_token_accuracy": 0.20541216880083085,
      "num_tokens": 32735015.0,
      "step": 15770
    },
    {
      "entropy": 5.4703583240509035,
      "epoch": 1.4309687953555879,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000479991239408135,
      "loss": 5.045,
      "mean_token_accuracy": 0.20821180194616318,
      "num_tokens": 32745899.0,
      "step": 15775
    },
    {
      "entropy": 5.461346435546875,
      "epoch": 1.431422351233672,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047997789845537003,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20634830445051194,
      "num_tokens": 32756052.0,
      "step": 15780
    },
    {
      "entropy": 5.3901145458221436,
      "epoch": 1.4318759071117562,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004799645532636098,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20348911136388778,
      "num_tokens": 32767118.0,
      "step": 15785
    },
    {
      "entropy": 5.40638279914856,
      "epoch": 1.4323294629898404,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047995120383313044,
      "loss": 4.8862,
      "mean_token_accuracy": 0.21329449415206908,
      "num_tokens": 32777069.0,
      "step": 15790
    },
    {
      "entropy": 5.486865758895874,
      "epoch": 1.4327830188679245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047993785016420796,
      "loss": 5.1306,
      "mean_token_accuracy": 0.1897103577852249,
      "num_tokens": 32786967.0,
      "step": 15795
    },
    {
      "entropy": 5.473412895202637,
      "epoch": 1.4332365747460087,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047992449225711854,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20178884714841844,
      "num_tokens": 32797563.0,
      "step": 15800
    },
    {
      "entropy": 5.470533037185669,
      "epoch": 1.4336901306240928,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047991113011213837,
      "loss": 4.9736,
      "mean_token_accuracy": 0.20526070892810822,
      "num_tokens": 32808730.0,
      "step": 15805
    },
    {
      "entropy": 5.418741655349732,
      "epoch": 1.434143686502177,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004798977637295438,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20455852299928665,
      "num_tokens": 32818050.0,
      "step": 15810
    },
    {
      "entropy": 5.436024475097656,
      "epoch": 1.4345972423802613,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004798843931096113,
      "loss": 5.0085,
      "mean_token_accuracy": 0.2018474042415619,
      "num_tokens": 32828993.0,
      "step": 15815
    },
    {
      "entropy": 5.470671367645264,
      "epoch": 1.4350507982583454,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004798710182526173,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20728410184383392,
      "num_tokens": 32839331.0,
      "step": 15820
    },
    {
      "entropy": 5.470010042190552,
      "epoch": 1.4355043541364296,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047985763915883847,
      "loss": 5.0273,
      "mean_token_accuracy": 0.2018204912543297,
      "num_tokens": 32849946.0,
      "step": 15825
    },
    {
      "entropy": 5.388349962234497,
      "epoch": 1.4359579100145137,
      "grad_norm": 1.0,
      "learning_rate": 0.0004798442558285514,
      "loss": 4.9572,
      "mean_token_accuracy": 0.20606671422719955,
      "num_tokens": 32859960.0,
      "step": 15830
    },
    {
      "entropy": 5.4434051513671875,
      "epoch": 1.436411465892598,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004798308682620329,
      "loss": 4.9816,
      "mean_token_accuracy": 0.1954537570476532,
      "num_tokens": 32870317.0,
      "step": 15835
    },
    {
      "entropy": 5.541861629486084,
      "epoch": 1.4368650217706822,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047981747645955996,
      "loss": 5.1581,
      "mean_token_accuracy": 0.1936565160751343,
      "num_tokens": 32880906.0,
      "step": 15840
    },
    {
      "entropy": 5.533978796005249,
      "epoch": 1.4373185776487662,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004798040804214094,
      "loss": 5.1911,
      "mean_token_accuracy": 0.19131509214639664,
      "num_tokens": 32891038.0,
      "step": 15845
    },
    {
      "entropy": 5.570695543289185,
      "epoch": 1.4377721335268505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797906801478582,
      "loss": 5.1348,
      "mean_token_accuracy": 0.20294843167066573,
      "num_tokens": 32901013.0,
      "step": 15850
    },
    {
      "entropy": 5.543267345428466,
      "epoch": 1.4382256894049346,
      "grad_norm": 1.0,
      "learning_rate": 0.0004797772756391836,
      "loss": 5.0557,
      "mean_token_accuracy": 0.20270682871341705,
      "num_tokens": 32911592.0,
      "step": 15855
    },
    {
      "entropy": 5.4332098960876465,
      "epoch": 1.4386792452830188,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004797638668956627,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20300047993659973,
      "num_tokens": 32922087.0,
      "step": 15860
    },
    {
      "entropy": 5.471498680114746,
      "epoch": 1.439132801161103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004797504539175729,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20456573367118835,
      "num_tokens": 32932039.0,
      "step": 15865
    },
    {
      "entropy": 5.462139129638672,
      "epoch": 1.4395863570391874,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004797370367051915,
      "loss": 5.0036,
      "mean_token_accuracy": 0.200211963057518,
      "num_tokens": 32941896.0,
      "step": 15870
    },
    {
      "entropy": 5.44089994430542,
      "epoch": 1.4400399129172714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000479723615258796,
      "loss": 5.0184,
      "mean_token_accuracy": 0.19286007285118104,
      "num_tokens": 32951675.0,
      "step": 15875
    },
    {
      "entropy": 5.474519538879394,
      "epoch": 1.4404934687953557,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047971018957866396,
      "loss": 5.1058,
      "mean_token_accuracy": 0.20704136788845062,
      "num_tokens": 32961378.0,
      "step": 15880
    },
    {
      "entropy": 5.483541822433471,
      "epoch": 1.4409470246734397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047969675966507314,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2028401717543602,
      "num_tokens": 32971770.0,
      "step": 15885
    },
    {
      "entropy": 5.458818769454956,
      "epoch": 1.441400580551524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000479683325518301,
      "loss": 5.027,
      "mean_token_accuracy": 0.20152291655540466,
      "num_tokens": 32982934.0,
      "step": 15890
    },
    {
      "entropy": 5.436052989959717,
      "epoch": 1.4418541364296082,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004796698871386257,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20192381739616394,
      "num_tokens": 32993252.0,
      "step": 15895
    },
    {
      "entropy": 5.496966934204101,
      "epoch": 1.4423076923076923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004796564445263249,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20016196519136428,
      "num_tokens": 33003610.0,
      "step": 15900
    },
    {
      "entropy": 5.482189559936524,
      "epoch": 1.4427612481857766,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047964299768167665,
      "loss": 4.9995,
      "mean_token_accuracy": 0.20372257828712464,
      "num_tokens": 33014037.0,
      "step": 15905
    },
    {
      "entropy": 5.478550577163697,
      "epoch": 1.4432148040638606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004796295466049591,
      "loss": 5.0386,
      "mean_token_accuracy": 0.19845939427614212,
      "num_tokens": 33023299.0,
      "step": 15910
    },
    {
      "entropy": 5.499065399169922,
      "epoch": 1.4436683599419449,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047961609129645036,
      "loss": 5.0815,
      "mean_token_accuracy": 0.1942521184682846,
      "num_tokens": 33034194.0,
      "step": 15915
    },
    {
      "entropy": 5.490591764450073,
      "epoch": 1.4441219158200291,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047960263175642867,
      "loss": 4.9545,
      "mean_token_accuracy": 0.2109111577272415,
      "num_tokens": 33044715.0,
      "step": 15920
    },
    {
      "entropy": 5.407504224777222,
      "epoch": 1.4445754716981132,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004795891679851724,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21129163801670076,
      "num_tokens": 33055214.0,
      "step": 15925
    },
    {
      "entropy": 5.442496252059937,
      "epoch": 1.4450290275761974,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004795756999829601,
      "loss": 5.0054,
      "mean_token_accuracy": 0.20301516205072404,
      "num_tokens": 33065026.0,
      "step": 15930
    },
    {
      "entropy": 5.5316237449646,
      "epoch": 1.4454825834542815,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047956222775007016,
      "loss": 5.0594,
      "mean_token_accuracy": 0.1945800945162773,
      "num_tokens": 33075138.0,
      "step": 15935
    },
    {
      "entropy": 5.504463338851929,
      "epoch": 1.4459361393323658,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004795487512867812,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20541611164808274,
      "num_tokens": 33086625.0,
      "step": 15940
    },
    {
      "entropy": 5.5062093257904055,
      "epoch": 1.44638969521045,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047953527059337194,
      "loss": 5.082,
      "mean_token_accuracy": 0.19432028532028198,
      "num_tokens": 33097076.0,
      "step": 15945
    },
    {
      "entropy": 5.4663145542144775,
      "epoch": 1.446843251088534,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004795217856701212,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20231714248657226,
      "num_tokens": 33107168.0,
      "step": 15950
    },
    {
      "entropy": 5.481114721298217,
      "epoch": 1.4472968069666183,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047950829651730774,
      "loss": 5.056,
      "mean_token_accuracy": 0.20539132058620452,
      "num_tokens": 33117862.0,
      "step": 15955
    },
    {
      "entropy": 5.437439918518066,
      "epoch": 1.4477503628447024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004794948031352107,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2031979814171791,
      "num_tokens": 33127144.0,
      "step": 15960
    },
    {
      "entropy": 5.5240884780883786,
      "epoch": 1.4482039187227866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047948130552410897,
      "loss": 5.1119,
      "mean_token_accuracy": 0.1925661489367485,
      "num_tokens": 33139180.0,
      "step": 15965
    },
    {
      "entropy": 5.532641744613647,
      "epoch": 1.448657474600871,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047946780368428175,
      "loss": 5.0333,
      "mean_token_accuracy": 0.19493882209062577,
      "num_tokens": 33149467.0,
      "step": 15970
    },
    {
      "entropy": 5.422984313964844,
      "epoch": 1.449111030478955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004794542976160081,
      "loss": 4.8927,
      "mean_token_accuracy": 0.21469856351613997,
      "num_tokens": 33158752.0,
      "step": 15975
    },
    {
      "entropy": 5.424119520187378,
      "epoch": 1.4495645863570392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047944078731956765,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20479470044374465,
      "num_tokens": 33168242.0,
      "step": 15980
    },
    {
      "entropy": 5.406200122833252,
      "epoch": 1.4500181422351233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047942727279523957,
      "loss": 4.9954,
      "mean_token_accuracy": 0.20532746464014054,
      "num_tokens": 33179134.0,
      "step": 15985
    },
    {
      "entropy": 5.453504800796509,
      "epoch": 1.4504716981132075,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004794137540433034,
      "loss": 5.1209,
      "mean_token_accuracy": 0.1939221903681755,
      "num_tokens": 33188388.0,
      "step": 15990
    },
    {
      "entropy": 5.495200109481812,
      "epoch": 1.4509252539912918,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004794002310640386,
      "loss": 5.021,
      "mean_token_accuracy": 0.20037586092948914,
      "num_tokens": 33198750.0,
      "step": 15995
    },
    {
      "entropy": 5.575225591659546,
      "epoch": 1.4513788098693758,
      "grad_norm": 0.875,
      "learning_rate": 0.00047938670385772503,
      "loss": 5.1325,
      "mean_token_accuracy": 0.18529079705476761,
      "num_tokens": 33209532.0,
      "step": 16000
    },
    {
      "entropy": 5.478796434402466,
      "epoch": 1.45183236574746,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047937317242464244,
      "loss": 4.9824,
      "mean_token_accuracy": 0.206252583861351,
      "num_tokens": 33219675.0,
      "step": 16005
    },
    {
      "entropy": 5.494966888427735,
      "epoch": 1.4522859216255442,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004793596367650704,
      "loss": 5.0944,
      "mean_token_accuracy": 0.1961601510643959,
      "num_tokens": 33230919.0,
      "step": 16010
    },
    {
      "entropy": 5.469520187377929,
      "epoch": 1.4527394775036284,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004793460968792891,
      "loss": 4.9976,
      "mean_token_accuracy": 0.20618475079536439,
      "num_tokens": 33241146.0,
      "step": 16015
    },
    {
      "entropy": 5.479263734817505,
      "epoch": 1.4531930333817127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047933255276757836,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19950176626443863,
      "num_tokens": 33251434.0,
      "step": 16020
    },
    {
      "entropy": 5.528160095214844,
      "epoch": 1.4536465892597967,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004793190044302184,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19096487164497375,
      "num_tokens": 33261868.0,
      "step": 16025
    },
    {
      "entropy": 5.474983215332031,
      "epoch": 1.454100145137881,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047930545186748936,
      "loss": 4.9595,
      "mean_token_accuracy": 0.2037203401327133,
      "num_tokens": 33271798.0,
      "step": 16030
    },
    {
      "entropy": 5.460812616348266,
      "epoch": 1.454553701015965,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047929189507967153,
      "loss": 5.0707,
      "mean_token_accuracy": 0.19978414475917816,
      "num_tokens": 33282614.0,
      "step": 16035
    },
    {
      "entropy": 5.400892162322998,
      "epoch": 1.4550072568940493,
      "grad_norm": 1.0,
      "learning_rate": 0.00047927833406704516,
      "loss": 4.9878,
      "mean_token_accuracy": 0.19926196336746216,
      "num_tokens": 33292906.0,
      "step": 16040
    },
    {
      "entropy": 5.449161148071289,
      "epoch": 1.4554608127721336,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004792647688298909,
      "loss": 4.9946,
      "mean_token_accuracy": 0.20244804173707961,
      "num_tokens": 33302720.0,
      "step": 16045
    },
    {
      "entropy": 5.430754852294922,
      "epoch": 1.4559143686502178,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004792511993684891,
      "loss": 4.9956,
      "mean_token_accuracy": 0.20477315187454223,
      "num_tokens": 33312959.0,
      "step": 16050
    },
    {
      "entropy": 5.42355523109436,
      "epoch": 1.4563679245283019,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047923762568312053,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21084294021129607,
      "num_tokens": 33322737.0,
      "step": 16055
    },
    {
      "entropy": 5.408132886886596,
      "epoch": 1.4568214804063861,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004792240477740657,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2052319809794426,
      "num_tokens": 33333185.0,
      "step": 16060
    },
    {
      "entropy": 5.496552848815918,
      "epoch": 1.4572750362844702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004792104656416056,
      "loss": 5.0933,
      "mean_token_accuracy": 0.19294783622026443,
      "num_tokens": 33343414.0,
      "step": 16065
    },
    {
      "entropy": 5.4850352764129635,
      "epoch": 1.4577285921625545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047919687928602104,
      "loss": 5.01,
      "mean_token_accuracy": 0.20524217933416367,
      "num_tokens": 33352527.0,
      "step": 16070
    },
    {
      "entropy": 5.477297830581665,
      "epoch": 1.4581821480406387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047918328870759293,
      "loss": 5.0688,
      "mean_token_accuracy": 0.19643971771001817,
      "num_tokens": 33363526.0,
      "step": 16075
    },
    {
      "entropy": 5.457568788528443,
      "epoch": 1.4586357039187228,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791696939066024,
      "loss": 5.0815,
      "mean_token_accuracy": 0.2038888618350029,
      "num_tokens": 33374302.0,
      "step": 16080
    },
    {
      "entropy": 5.488372564315796,
      "epoch": 1.459089259796807,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004791560948833306,
      "loss": 5.0624,
      "mean_token_accuracy": 0.20285871326923371,
      "num_tokens": 33384786.0,
      "step": 16085
    },
    {
      "entropy": 5.478956651687622,
      "epoch": 1.459542815674891,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791424916380586,
      "loss": 5.0481,
      "mean_token_accuracy": 0.20411778539419173,
      "num_tokens": 33394464.0,
      "step": 16090
    },
    {
      "entropy": 5.479393053054809,
      "epoch": 1.4599963715529753,
      "grad_norm": 0.984375,
      "learning_rate": 0.000479128884171068,
      "loss": 5.0644,
      "mean_token_accuracy": 0.19856854975223542,
      "num_tokens": 33405476.0,
      "step": 16095
    },
    {
      "entropy": 5.509451675415039,
      "epoch": 1.4604499274310596,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047911527248264003,
      "loss": 5.088,
      "mean_token_accuracy": 0.2023351863026619,
      "num_tokens": 33415758.0,
      "step": 16100
    },
    {
      "entropy": 5.450877618789673,
      "epoch": 1.4609034833091437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047910165657305613,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20088934898376465,
      "num_tokens": 33424845.0,
      "step": 16105
    },
    {
      "entropy": 5.4813680171966555,
      "epoch": 1.461357039187228,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047908803644259804,
      "loss": 5.089,
      "mean_token_accuracy": 0.1970730498433113,
      "num_tokens": 33435928.0,
      "step": 16110
    },
    {
      "entropy": 5.485598707199097,
      "epoch": 1.461810595065312,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047907441209154726,
      "loss": 4.987,
      "mean_token_accuracy": 0.20164019763469695,
      "num_tokens": 33447307.0,
      "step": 16115
    },
    {
      "entropy": 5.460571575164795,
      "epoch": 1.4622641509433962,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047906078352018575,
      "loss": 5.0579,
      "mean_token_accuracy": 0.19293845891952516,
      "num_tokens": 33458116.0,
      "step": 16120
    },
    {
      "entropy": 5.335807943344117,
      "epoch": 1.4627177068214805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004790471507287952,
      "loss": 4.8939,
      "mean_token_accuracy": 0.21040795743465424,
      "num_tokens": 33469065.0,
      "step": 16125
    },
    {
      "entropy": 5.503554630279541,
      "epoch": 1.4631712626995645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047903351371765757,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20361997336149215,
      "num_tokens": 33479118.0,
      "step": 16130
    },
    {
      "entropy": 5.522790050506591,
      "epoch": 1.4636248185776488,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047901987248705484,
      "loss": 5.0594,
      "mean_token_accuracy": 0.19907349944114686,
      "num_tokens": 33489642.0,
      "step": 16135
    },
    {
      "entropy": 5.454414033889771,
      "epoch": 1.4640783744557329,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004790062270372693,
      "loss": 5.0972,
      "mean_token_accuracy": 0.1919420763850212,
      "num_tokens": 33499771.0,
      "step": 16140
    },
    {
      "entropy": 5.458485412597656,
      "epoch": 1.4645319303338171,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004789925773685828,
      "loss": 5.0324,
      "mean_token_accuracy": 0.19968261271715165,
      "num_tokens": 33509975.0,
      "step": 16145
    },
    {
      "entropy": 5.477542972564697,
      "epoch": 1.4649854862119014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789789234812779,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19430531114339827,
      "num_tokens": 33520017.0,
      "step": 16150
    },
    {
      "entropy": 5.483196592330932,
      "epoch": 1.4654390420899854,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047896526537563686,
      "loss": 5.1284,
      "mean_token_accuracy": 0.1917436122894287,
      "num_tokens": 33531977.0,
      "step": 16155
    },
    {
      "entropy": 5.468639326095581,
      "epoch": 1.4658925979680697,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047895160305194223,
      "loss": 5.0825,
      "mean_token_accuracy": 0.1999267041683197,
      "num_tokens": 33543660.0,
      "step": 16160
    },
    {
      "entropy": 5.474434423446655,
      "epoch": 1.4663461538461537,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004789379365104764,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2019614279270172,
      "num_tokens": 33553814.0,
      "step": 16165
    },
    {
      "entropy": 5.493988704681397,
      "epoch": 1.466799709724238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004789242657515221,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20582034438848495,
      "num_tokens": 33564250.0,
      "step": 16170
    },
    {
      "entropy": 5.461010551452636,
      "epoch": 1.4672532656023223,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789105907753621,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20410160422325135,
      "num_tokens": 33574755.0,
      "step": 16175
    },
    {
      "entropy": 5.439325046539307,
      "epoch": 1.4677068214804063,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000478896911582279,
      "loss": 4.9751,
      "mean_token_accuracy": 0.2002486065030098,
      "num_tokens": 33585268.0,
      "step": 16180
    },
    {
      "entropy": 5.457125616073609,
      "epoch": 1.4681603773584906,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004788832281725558,
      "loss": 5.0848,
      "mean_token_accuracy": 0.20506412535905838,
      "num_tokens": 33595753.0,
      "step": 16185
    },
    {
      "entropy": 5.452261877059937,
      "epoch": 1.4686139332365746,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004788695405464755,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2044246807694435,
      "num_tokens": 33605270.0,
      "step": 16190
    },
    {
      "entropy": 5.448578643798828,
      "epoch": 1.469067489114659,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047885584870432127,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20005607455968857,
      "num_tokens": 33615096.0,
      "step": 16195
    },
    {
      "entropy": 5.453908920288086,
      "epoch": 1.4695210449927432,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047884215264637606,
      "loss": 4.9705,
      "mean_token_accuracy": 0.20482762157917023,
      "num_tokens": 33624770.0,
      "step": 16200
    },
    {
      "entropy": 5.514669418334961,
      "epoch": 1.4699746008708272,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004788284523729232,
      "loss": 5.1089,
      "mean_token_accuracy": 0.19363716393709182,
      "num_tokens": 33634545.0,
      "step": 16205
    },
    {
      "entropy": 5.529506874084473,
      "epoch": 1.4704281567489115,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047881474788424595,
      "loss": 5.0652,
      "mean_token_accuracy": 0.20077794492244722,
      "num_tokens": 33644950.0,
      "step": 16210
    },
    {
      "entropy": 5.479498720169067,
      "epoch": 1.4708817126269955,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047880103918062776,
      "loss": 5.0076,
      "mean_token_accuracy": 0.2026103839278221,
      "num_tokens": 33654369.0,
      "step": 16215
    },
    {
      "entropy": 5.368625020980835,
      "epoch": 1.4713352685050798,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047878732626235225,
      "loss": 5.0446,
      "mean_token_accuracy": 0.19709902852773667,
      "num_tokens": 33664687.0,
      "step": 16220
    },
    {
      "entropy": 5.434810829162598,
      "epoch": 1.471788824383164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004787736091297028,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21163554936647416,
      "num_tokens": 33675832.0,
      "step": 16225
    },
    {
      "entropy": 5.503649663925171,
      "epoch": 1.4722423802612483,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004787598877829633,
      "loss": 5.074,
      "mean_token_accuracy": 0.19952421486377717,
      "num_tokens": 33686788.0,
      "step": 16230
    },
    {
      "entropy": 5.438729190826416,
      "epoch": 1.4726959361393324,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004787461622224173,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21022354513406755,
      "num_tokens": 33696384.0,
      "step": 16235
    },
    {
      "entropy": 5.4412675380706785,
      "epoch": 1.4731494920174166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004787324324483488,
      "loss": 5.0081,
      "mean_token_accuracy": 0.19890953004360198,
      "num_tokens": 33707758.0,
      "step": 16240
    },
    {
      "entropy": 5.462802696228027,
      "epoch": 1.4736030478955007,
      "grad_norm": 1.0,
      "learning_rate": 0.0004787186984610416,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2026902347803116,
      "num_tokens": 33717747.0,
      "step": 16245
    },
    {
      "entropy": 5.427540922164917,
      "epoch": 1.474056603773585,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047870496026077984,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20165301412343978,
      "num_tokens": 33727960.0,
      "step": 16250
    },
    {
      "entropy": 5.405560445785523,
      "epoch": 1.4745101596516692,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004786912178478476,
      "loss": 4.9615,
      "mean_token_accuracy": 0.20851432383060456,
      "num_tokens": 33738874.0,
      "step": 16255
    },
    {
      "entropy": 5.426895332336426,
      "epoch": 1.4749637155297532,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004786774712225291,
      "loss": 4.9643,
      "mean_token_accuracy": 0.21164844632148744,
      "num_tokens": 33749071.0,
      "step": 16260
    },
    {
      "entropy": 5.4466687679290775,
      "epoch": 1.4754172714078375,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004786637203851086,
      "loss": 4.9916,
      "mean_token_accuracy": 0.2025420770049095,
      "num_tokens": 33759509.0,
      "step": 16265
    },
    {
      "entropy": 5.507283735275268,
      "epoch": 1.4758708272859216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004786499653358705,
      "loss": 5.0361,
      "mean_token_accuracy": 0.20561971068382262,
      "num_tokens": 33769974.0,
      "step": 16270
    },
    {
      "entropy": 5.548575830459595,
      "epoch": 1.4763243831640058,
      "grad_norm": 1.0,
      "learning_rate": 0.0004786362060750991,
      "loss": 5.0762,
      "mean_token_accuracy": 0.20296125039458274,
      "num_tokens": 33780403.0,
      "step": 16275
    },
    {
      "entropy": 5.459582138061523,
      "epoch": 1.47677793904209,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047862244260307913,
      "loss": 5.0763,
      "mean_token_accuracy": 0.20544939637184143,
      "num_tokens": 33791364.0,
      "step": 16280
    },
    {
      "entropy": 5.447279024124145,
      "epoch": 1.4772314949201741,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004786086749200951,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20414622575044633,
      "num_tokens": 33801964.0,
      "step": 16285
    },
    {
      "entropy": 5.504152631759643,
      "epoch": 1.4776850507982584,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004785949030264319,
      "loss": 5.0246,
      "mean_token_accuracy": 0.20249088406562804,
      "num_tokens": 33813132.0,
      "step": 16290
    },
    {
      "entropy": 5.44755973815918,
      "epoch": 1.4781386066763424,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004785811269223741,
      "loss": 5.0101,
      "mean_token_accuracy": 0.2099921613931656,
      "num_tokens": 33823069.0,
      "step": 16295
    },
    {
      "entropy": 5.50404543876648,
      "epoch": 1.4785921625544267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004785673466082068,
      "loss": 5.0149,
      "mean_token_accuracy": 0.19754787534475327,
      "num_tokens": 33832626.0,
      "step": 16300
    },
    {
      "entropy": 5.426759576797485,
      "epoch": 1.479045718432511,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047855356208421484,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20693277418613434,
      "num_tokens": 33843384.0,
      "step": 16305
    },
    {
      "entropy": 5.3980459690094,
      "epoch": 1.479499274310595,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004785397733506834,
      "loss": 4.9885,
      "mean_token_accuracy": 0.2088317394256592,
      "num_tokens": 33854286.0,
      "step": 16310
    },
    {
      "entropy": 5.406148290634155,
      "epoch": 1.4799528301886793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004785259804078976,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20360536724328995,
      "num_tokens": 33863379.0,
      "step": 16315
    },
    {
      "entropy": 5.4936699867248535,
      "epoch": 1.4804063860667633,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004785121832561426,
      "loss": 5.184,
      "mean_token_accuracy": 0.19662310928106308,
      "num_tokens": 33875641.0,
      "step": 16320
    },
    {
      "entropy": 5.4985864639282225,
      "epoch": 1.4808599419448476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047849838189570375,
      "loss": 4.9796,
      "mean_token_accuracy": 0.20845424234867097,
      "num_tokens": 33885023.0,
      "step": 16325
    },
    {
      "entropy": 5.5839954853057865,
      "epoch": 1.4813134978229319,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004784845763268666,
      "loss": 5.121,
      "mean_token_accuracy": 0.1973605066537857,
      "num_tokens": 33895562.0,
      "step": 16330
    },
    {
      "entropy": 5.433398008346558,
      "epoch": 1.481767053701016,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004784707665499165,
      "loss": 5.0334,
      "mean_token_accuracy": 0.20686419904232026,
      "num_tokens": 33907095.0,
      "step": 16335
    },
    {
      "entropy": 5.3774275302886965,
      "epoch": 1.4822206095791002,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004784569525651391,
      "loss": 4.8602,
      "mean_token_accuracy": 0.2152920514345169,
      "num_tokens": 33917475.0,
      "step": 16340
    },
    {
      "entropy": 5.42539005279541,
      "epoch": 1.4826741654571842,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047844313437282005,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2041505143046379,
      "num_tokens": 33927740.0,
      "step": 16345
    },
    {
      "entropy": 5.410151386260987,
      "epoch": 1.4831277213352685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047842931197324525,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20373953580856324,
      "num_tokens": 33937416.0,
      "step": 16350
    },
    {
      "entropy": 5.409642457962036,
      "epoch": 1.4835812772133528,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047841548536670033,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21611718833446503,
      "num_tokens": 33946968.0,
      "step": 16355
    },
    {
      "entropy": 5.516166925430298,
      "epoch": 1.4840348330914368,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047840165455347136,
      "loss": 5.0801,
      "mean_token_accuracy": 0.19407518357038497,
      "num_tokens": 33957877.0,
      "step": 16360
    },
    {
      "entropy": 5.499650478363037,
      "epoch": 1.484488388969521,
      "grad_norm": 1.0,
      "learning_rate": 0.00047838781953384436,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1990896999835968,
      "num_tokens": 33967890.0,
      "step": 16365
    },
    {
      "entropy": 5.427718257904052,
      "epoch": 1.484941944847605,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004783739803081054,
      "loss": 5.0084,
      "mean_token_accuracy": 0.205517914891243,
      "num_tokens": 33977117.0,
      "step": 16370
    },
    {
      "entropy": 5.469697761535644,
      "epoch": 1.4853955007256894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047836013687654075,
      "loss": 5.1103,
      "mean_token_accuracy": 0.1963505521416664,
      "num_tokens": 33986943.0,
      "step": 16375
    },
    {
      "entropy": 5.468562316894531,
      "epoch": 1.4858490566037736,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047834628923943654,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20292943418025972,
      "num_tokens": 33996499.0,
      "step": 16380
    },
    {
      "entropy": 5.511895990371704,
      "epoch": 1.4863026124818577,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004783324373970793,
      "loss": 5.1124,
      "mean_token_accuracy": 0.1989882156252861,
      "num_tokens": 34006749.0,
      "step": 16385
    },
    {
      "entropy": 5.5462761402130125,
      "epoch": 1.486756168359942,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004783185813497555,
      "loss": 5.1142,
      "mean_token_accuracy": 0.19380537718534468,
      "num_tokens": 34017194.0,
      "step": 16390
    },
    {
      "entropy": 5.528011274337769,
      "epoch": 1.487209724238026,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004783047210977516,
      "loss": 5.0467,
      "mean_token_accuracy": 0.20265622884035112,
      "num_tokens": 34027386.0,
      "step": 16395
    },
    {
      "entropy": 5.534073352813721,
      "epoch": 1.4876632801161103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047829085664135424,
      "loss": 5.067,
      "mean_token_accuracy": 0.1954236224293709,
      "num_tokens": 34037540.0,
      "step": 16400
    },
    {
      "entropy": 5.428753852844238,
      "epoch": 1.4881168359941945,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047827698798085015,
      "loss": 5.0028,
      "mean_token_accuracy": 0.20653520077466964,
      "num_tokens": 34048008.0,
      "step": 16405
    },
    {
      "entropy": 5.503815984725952,
      "epoch": 1.4885703918722788,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004782631151165262,
      "loss": 5.0187,
      "mean_token_accuracy": 0.2068993330001831,
      "num_tokens": 34058009.0,
      "step": 16410
    },
    {
      "entropy": 5.5186210632324215,
      "epoch": 1.4890239477503628,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047824923804866914,
      "loss": 4.971,
      "mean_token_accuracy": 0.20440451204776763,
      "num_tokens": 34067625.0,
      "step": 16415
    },
    {
      "entropy": 5.44329948425293,
      "epoch": 1.4894775036284469,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047823535677756607,
      "loss": 5.0274,
      "mean_token_accuracy": 0.19958255141973497,
      "num_tokens": 34077349.0,
      "step": 16420
    },
    {
      "entropy": 5.520338439941407,
      "epoch": 1.4899310595065312,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004782214713035041,
      "loss": 5.1387,
      "mean_token_accuracy": 0.19416306912899017,
      "num_tokens": 34088636.0,
      "step": 16425
    },
    {
      "entropy": 5.518615245819092,
      "epoch": 1.4903846153846154,
      "grad_norm": 1.0,
      "learning_rate": 0.00047820758162677026,
      "loss": 5.0519,
      "mean_token_accuracy": 0.20491172522306442,
      "num_tokens": 34098806.0,
      "step": 16430
    },
    {
      "entropy": 5.448275423049926,
      "epoch": 1.4908381712626997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047819368774765187,
      "loss": 5.039,
      "mean_token_accuracy": 0.20588411688804625,
      "num_tokens": 34109713.0,
      "step": 16435
    },
    {
      "entropy": 5.424658107757568,
      "epoch": 1.4912917271407837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047817978966643626,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2082641527056694,
      "num_tokens": 34119585.0,
      "step": 16440
    },
    {
      "entropy": 5.534990835189819,
      "epoch": 1.491745283018868,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004781658873834108,
      "loss": 5.0459,
      "mean_token_accuracy": 0.1953447639942169,
      "num_tokens": 34131245.0,
      "step": 16445
    },
    {
      "entropy": 5.513223123550415,
      "epoch": 1.492198838896952,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047815198089886296,
      "loss": 5.0228,
      "mean_token_accuracy": 0.2032491758465767,
      "num_tokens": 34141229.0,
      "step": 16450
    },
    {
      "entropy": 5.458254146575928,
      "epoch": 1.4926523947750363,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004781380702130805,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19513534903526306,
      "num_tokens": 34151781.0,
      "step": 16455
    },
    {
      "entropy": 5.426235008239746,
      "epoch": 1.4931059506531206,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047812415532635087,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20614495277404785,
      "num_tokens": 34162305.0,
      "step": 16460
    },
    {
      "entropy": 5.492601633071899,
      "epoch": 1.4935595065312046,
      "grad_norm": 0.9375,
      "learning_rate": 0.000478110236238962,
      "loss": 5.0856,
      "mean_token_accuracy": 0.19664422869682313,
      "num_tokens": 34173613.0,
      "step": 16465
    },
    {
      "entropy": 5.444189500808716,
      "epoch": 1.4940130624092889,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047809631295120164,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20410215258598327,
      "num_tokens": 34183540.0,
      "step": 16470
    },
    {
      "entropy": 5.437660074234008,
      "epoch": 1.494466618287373,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047808238546335786,
      "loss": 5.033,
      "mean_token_accuracy": 0.20073717087507248,
      "num_tokens": 34193292.0,
      "step": 16475
    },
    {
      "entropy": 5.490480756759643,
      "epoch": 1.4949201741654572,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047806845377571856,
      "loss": 5.0788,
      "mean_token_accuracy": 0.2080323338508606,
      "num_tokens": 34202697.0,
      "step": 16480
    },
    {
      "entropy": 5.4534783363342285,
      "epoch": 1.4953737300435415,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047805451788857187,
      "loss": 4.986,
      "mean_token_accuracy": 0.20318156331777573,
      "num_tokens": 34213396.0,
      "step": 16485
    },
    {
      "entropy": 5.480810022354126,
      "epoch": 1.4958272859216255,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000478040577802206,
      "loss": 4.9772,
      "mean_token_accuracy": 0.20488856732845306,
      "num_tokens": 34223020.0,
      "step": 16490
    },
    {
      "entropy": 5.404968214035034,
      "epoch": 1.4962808417997098,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004780266335169093,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20515892803668975,
      "num_tokens": 34233249.0,
      "step": 16495
    },
    {
      "entropy": 5.5380853652954105,
      "epoch": 1.4967343976777938,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047801268503297,
      "loss": 5.0762,
      "mean_token_accuracy": 0.1973375201225281,
      "num_tokens": 34244091.0,
      "step": 16500
    },
    {
      "entropy": 5.49814920425415,
      "epoch": 1.497187953555878,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047799873235067667,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20048414617776872,
      "num_tokens": 34255060.0,
      "step": 16505
    },
    {
      "entropy": 5.422490882873535,
      "epoch": 1.4976415094339623,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004779847754703178,
      "loss": 4.9489,
      "mean_token_accuracy": 0.206888285279274,
      "num_tokens": 34264361.0,
      "step": 16510
    },
    {
      "entropy": 5.423403739929199,
      "epoch": 1.4980950653120464,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004779708143921821,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2039859652519226,
      "num_tokens": 34274480.0,
      "step": 16515
    },
    {
      "entropy": 5.452390336990357,
      "epoch": 1.4985486211901307,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047795684911655824,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20332777947187425,
      "num_tokens": 34285951.0,
      "step": 16520
    },
    {
      "entropy": 5.48722357749939,
      "epoch": 1.4990021770682147,
      "grad_norm": 1.0,
      "learning_rate": 0.000477942879643735,
      "loss": 5.071,
      "mean_token_accuracy": 0.19658953696489334,
      "num_tokens": 34296384.0,
      "step": 16525
    },
    {
      "entropy": 5.462945890426636,
      "epoch": 1.499455732946299,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047792890597400125,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20717779994010926,
      "num_tokens": 34306776.0,
      "step": 16530
    },
    {
      "entropy": 5.502072858810425,
      "epoch": 1.4999092888243832,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047791492810764604,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20408144295215608,
      "num_tokens": 34317047.0,
      "step": 16535
    },
    {
      "entropy": 5.450231695175171,
      "epoch": 1.5003628447024675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004779009460449584,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20671842098236085,
      "num_tokens": 34326392.0,
      "step": 16540
    },
    {
      "entropy": 5.4880836486816404,
      "epoch": 1.5008164005805515,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047788695978622756,
      "loss": 5.0363,
      "mean_token_accuracy": 0.209957118332386,
      "num_tokens": 34336580.0,
      "step": 16545
    },
    {
      "entropy": 5.4223161220550535,
      "epoch": 1.5012699564586356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047787296933174265,
      "loss": 5.0481,
      "mean_token_accuracy": 0.2016463026404381,
      "num_tokens": 34346792.0,
      "step": 16550
    },
    {
      "entropy": 5.499814701080322,
      "epoch": 1.5017235123367199,
      "grad_norm": 1.015625,
      "learning_rate": 0.000477858974681793,
      "loss": 5.0654,
      "mean_token_accuracy": 0.1987570568919182,
      "num_tokens": 34356675.0,
      "step": 16555
    },
    {
      "entropy": 5.512963771820068,
      "epoch": 1.5021770682148041,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004778449758366681,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21385493725538254,
      "num_tokens": 34366357.0,
      "step": 16560
    },
    {
      "entropy": 5.437784957885742,
      "epoch": 1.5026306240928884,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004778309727966574,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20755893886089324,
      "num_tokens": 34377229.0,
      "step": 16565
    },
    {
      "entropy": 5.405068588256836,
      "epoch": 1.5030841799709724,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047781696556205045,
      "loss": 4.9645,
      "mean_token_accuracy": 0.20646806359291076,
      "num_tokens": 34387873.0,
      "step": 16570
    },
    {
      "entropy": 5.459356737136841,
      "epoch": 1.5035377358490565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047780295413313696,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21081265956163406,
      "num_tokens": 34398889.0,
      "step": 16575
    },
    {
      "entropy": 5.429965829849243,
      "epoch": 1.5039912917271407,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047778893851020676,
      "loss": 4.9218,
      "mean_token_accuracy": 0.20821678191423415,
      "num_tokens": 34408508.0,
      "step": 16580
    },
    {
      "entropy": 5.436028480529785,
      "epoch": 1.504444847605225,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004777749186935496,
      "loss": 5.0968,
      "mean_token_accuracy": 0.20087021440267563,
      "num_tokens": 34418928.0,
      "step": 16585
    },
    {
      "entropy": 5.492291975021362,
      "epoch": 1.5048984034833093,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004777608946834555,
      "loss": 5.0675,
      "mean_token_accuracy": 0.19984024167060851,
      "num_tokens": 34429789.0,
      "step": 16590
    },
    {
      "entropy": 5.525197219848633,
      "epoch": 1.5053519593613933,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004777468664802144,
      "loss": 5.1595,
      "mean_token_accuracy": 0.19593982249498368,
      "num_tokens": 34440795.0,
      "step": 16595
    },
    {
      "entropy": 5.5145854473114015,
      "epoch": 1.5058055152394774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047773283408411636,
      "loss": 5.0635,
      "mean_token_accuracy": 0.19572877287864685,
      "num_tokens": 34450772.0,
      "step": 16600
    },
    {
      "entropy": 5.445775747299194,
      "epoch": 1.5062590711175616,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047771879749545175,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20167359709739685,
      "num_tokens": 34461273.0,
      "step": 16605
    },
    {
      "entropy": 5.469457149505615,
      "epoch": 1.506712626995646,
      "grad_norm": 1.0,
      "learning_rate": 0.00047770475671451073,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19966941922903061,
      "num_tokens": 34470942.0,
      "step": 16610
    },
    {
      "entropy": 5.509217023849487,
      "epoch": 1.5071661828737302,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004776907117415837,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2073441728949547,
      "num_tokens": 34481117.0,
      "step": 16615
    },
    {
      "entropy": 5.429892539978027,
      "epoch": 1.5076197387518142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047767666257696103,
      "loss": 4.9633,
      "mean_token_accuracy": 0.20793781578540801,
      "num_tokens": 34490953.0,
      "step": 16620
    },
    {
      "entropy": 5.421083641052246,
      "epoch": 1.5080732946298983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047766260922093337,
      "loss": 4.9287,
      "mean_token_accuracy": 0.20670849829912186,
      "num_tokens": 34502765.0,
      "step": 16625
    },
    {
      "entropy": 5.52827205657959,
      "epoch": 1.5085268505079825,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004776485516737913,
      "loss": 5.0962,
      "mean_token_accuracy": 0.1967792272567749,
      "num_tokens": 34514473.0,
      "step": 16630
    },
    {
      "entropy": 5.533112478256226,
      "epoch": 1.5089804063860668,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004776344899358256,
      "loss": 5.1332,
      "mean_token_accuracy": 0.19769856035709382,
      "num_tokens": 34524218.0,
      "step": 16635
    },
    {
      "entropy": 5.490613555908203,
      "epoch": 1.509433962264151,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004776204240073269,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20425005704164506,
      "num_tokens": 34535697.0,
      "step": 16640
    },
    {
      "entropy": 5.524890518188476,
      "epoch": 1.509887518142235,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047760635388858624,
      "loss": 5.055,
      "mean_token_accuracy": 0.20605057328939438,
      "num_tokens": 34545130.0,
      "step": 16645
    },
    {
      "entropy": 5.4361841678619385,
      "epoch": 1.5103410740203191,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047759227957989465,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20409290939569474,
      "num_tokens": 34554974.0,
      "step": 16650
    },
    {
      "entropy": 5.47480525970459,
      "epoch": 1.5107946298984034,
      "grad_norm": 1.03125,
      "learning_rate": 0.000477578201081543,
      "loss": 5.0313,
      "mean_token_accuracy": 0.2005136340856552,
      "num_tokens": 34565660.0,
      "step": 16655
    },
    {
      "entropy": 5.52646484375,
      "epoch": 1.5112481857764877,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004775641183938225,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20296634435653688,
      "num_tokens": 34576447.0,
      "step": 16660
    },
    {
      "entropy": 5.542837333679199,
      "epoch": 1.511701741654572,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047755003151702443,
      "loss": 5.1749,
      "mean_token_accuracy": 0.19983410090208054,
      "num_tokens": 34588333.0,
      "step": 16665
    },
    {
      "entropy": 5.417311525344848,
      "epoch": 1.512155297532656,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047753594045144015,
      "loss": 4.9536,
      "mean_token_accuracy": 0.20125583112239837,
      "num_tokens": 34598166.0,
      "step": 16670
    },
    {
      "entropy": 5.449014949798584,
      "epoch": 1.5126088534107403,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004775218451973609,
      "loss": 4.9596,
      "mean_token_accuracy": 0.20207023173570632,
      "num_tokens": 34608615.0,
      "step": 16675
    },
    {
      "entropy": 5.45111665725708,
      "epoch": 1.5130624092888243,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004775077457550784,
      "loss": 5.0394,
      "mean_token_accuracy": 0.2035181000828743,
      "num_tokens": 34618618.0,
      "step": 16680
    },
    {
      "entropy": 5.498604202270508,
      "epoch": 1.5135159651669086,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047749364212488403,
      "loss": 5.0443,
      "mean_token_accuracy": 0.2016269311308861,
      "num_tokens": 34629412.0,
      "step": 16685
    },
    {
      "entropy": 5.554146385192871,
      "epoch": 1.5139695210449928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004774795343070696,
      "loss": 5.123,
      "mean_token_accuracy": 0.19738198965787887,
      "num_tokens": 34640487.0,
      "step": 16690
    },
    {
      "entropy": 5.4859092235565186,
      "epoch": 1.5144230769230769,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004774654223019268,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20505344718694687,
      "num_tokens": 34650126.0,
      "step": 16695
    },
    {
      "entropy": 5.395961618423462,
      "epoch": 1.5148766328011611,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004774513061097474,
      "loss": 4.8903,
      "mean_token_accuracy": 0.20872655212879182,
      "num_tokens": 34659721.0,
      "step": 16700
    },
    {
      "entropy": 5.442564296722412,
      "epoch": 1.5153301886792452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004774371857308235,
      "loss": 5.014,
      "mean_token_accuracy": 0.20492298454046248,
      "num_tokens": 34670611.0,
      "step": 16705
    },
    {
      "entropy": 5.48886137008667,
      "epoch": 1.5157837445573294,
      "grad_norm": 1.0,
      "learning_rate": 0.00047742306116544693,
      "loss": 5.0092,
      "mean_token_accuracy": 0.20673186480998992,
      "num_tokens": 34680173.0,
      "step": 16710
    },
    {
      "entropy": 5.466881942749024,
      "epoch": 1.5162373004354137,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047740893241390994,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20070289373397826,
      "num_tokens": 34690176.0,
      "step": 16715
    },
    {
      "entropy": 5.518379592895508,
      "epoch": 1.516690856313498,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004773947994765046,
      "loss": 4.983,
      "mean_token_accuracy": 0.21208746135234832,
      "num_tokens": 34700481.0,
      "step": 16720
    },
    {
      "entropy": 5.449109649658203,
      "epoch": 1.517144412191582,
      "grad_norm": 1.0,
      "learning_rate": 0.0004773806623535232,
      "loss": 5.1467,
      "mean_token_accuracy": 0.2002828910946846,
      "num_tokens": 34710643.0,
      "step": 16725
    },
    {
      "entropy": 5.492352247238159,
      "epoch": 1.517597968069666,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047736652104525816,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20047015994787215,
      "num_tokens": 34720998.0,
      "step": 16730
    },
    {
      "entropy": 5.532505941390991,
      "epoch": 1.5180515239477503,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004773523755520019,
      "loss": 5.0728,
      "mean_token_accuracy": 0.1998874217271805,
      "num_tokens": 34731071.0,
      "step": 16735
    },
    {
      "entropy": 5.438583946228027,
      "epoch": 1.5185050798258346,
      "grad_norm": 1.046875,
      "learning_rate": 0.000477338225874047,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20761779695749283,
      "num_tokens": 34741346.0,
      "step": 16740
    },
    {
      "entropy": 5.498762369155884,
      "epoch": 1.5189586357039189,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047732407201168595,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20281408131122589,
      "num_tokens": 34751092.0,
      "step": 16745
    },
    {
      "entropy": 5.544218111038208,
      "epoch": 1.519412191582003,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004773099139652116,
      "loss": 5.0918,
      "mean_token_accuracy": 0.19871319681406022,
      "num_tokens": 34761477.0,
      "step": 16750
    },
    {
      "entropy": 5.425150299072266,
      "epoch": 1.519865747460087,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004772957517349167,
      "loss": 5.0129,
      "mean_token_accuracy": 0.2022230938076973,
      "num_tokens": 34772491.0,
      "step": 16755
    },
    {
      "entropy": 5.4748804569244385,
      "epoch": 1.5203193033381712,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004772815853210939,
      "loss": 5.0592,
      "mean_token_accuracy": 0.19688044041395186,
      "num_tokens": 34783192.0,
      "step": 16760
    },
    {
      "entropy": 5.477023792266846,
      "epoch": 1.5207728592162555,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047726741472403655,
      "loss": 5.026,
      "mean_token_accuracy": 0.20596080124378205,
      "num_tokens": 34794164.0,
      "step": 16765
    },
    {
      "entropy": 5.490565490722656,
      "epoch": 1.5212264150943398,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047725323994403753,
      "loss": 5.0809,
      "mean_token_accuracy": 0.19330866187810897,
      "num_tokens": 34804390.0,
      "step": 16770
    },
    {
      "entropy": 5.591202640533448,
      "epoch": 1.5216799709724238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047723906098138993,
      "loss": 5.1577,
      "mean_token_accuracy": 0.19566252678632737,
      "num_tokens": 34814751.0,
      "step": 16775
    },
    {
      "entropy": 5.498318529129028,
      "epoch": 1.5221335268505078,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047722487783638697,
      "loss": 4.995,
      "mean_token_accuracy": 0.20829154253005983,
      "num_tokens": 34824977.0,
      "step": 16780
    },
    {
      "entropy": 5.4572632789611815,
      "epoch": 1.5225870827285921,
      "grad_norm": 1.09375,
      "learning_rate": 0.000477210690509322,
      "loss": 5.0428,
      "mean_token_accuracy": 0.20372531265020372,
      "num_tokens": 34835540.0,
      "step": 16785
    },
    {
      "entropy": 5.523160743713379,
      "epoch": 1.5230406386066764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047719649900048843,
      "loss": 5.0357,
      "mean_token_accuracy": 0.19952683597803117,
      "num_tokens": 34845233.0,
      "step": 16790
    },
    {
      "entropy": 5.537006473541259,
      "epoch": 1.5234941944847606,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047718230331017975,
      "loss": 5.1319,
      "mean_token_accuracy": 0.1903398737311363,
      "num_tokens": 34855756.0,
      "step": 16795
    },
    {
      "entropy": 5.468630743026734,
      "epoch": 1.5239477503628447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047716810343868956,
      "loss": 5.037,
      "mean_token_accuracy": 0.1923644244670868,
      "num_tokens": 34867869.0,
      "step": 16800
    },
    {
      "entropy": 5.585168933868408,
      "epoch": 1.5244013062409287,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047715389938631143,
      "loss": 5.088,
      "mean_token_accuracy": 0.19788610190153122,
      "num_tokens": 34878625.0,
      "step": 16805
    },
    {
      "entropy": 5.5586803436279295,
      "epoch": 1.524854862119013,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047713969115333916,
      "loss": 5.0944,
      "mean_token_accuracy": 0.1903443843126297,
      "num_tokens": 34889408.0,
      "step": 16810
    },
    {
      "entropy": 5.450815200805664,
      "epoch": 1.5253084179970973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004771254787400666,
      "loss": 5.1132,
      "mean_token_accuracy": 0.1963221773505211,
      "num_tokens": 34899602.0,
      "step": 16815
    },
    {
      "entropy": 5.44207181930542,
      "epoch": 1.5257619738751815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004771112621467876,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2069842994213104,
      "num_tokens": 34910215.0,
      "step": 16820
    },
    {
      "entropy": 5.456229877471924,
      "epoch": 1.5262155297532656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004770970413737961,
      "loss": 5.0213,
      "mean_token_accuracy": 0.20532291382551193,
      "num_tokens": 34919641.0,
      "step": 16825
    },
    {
      "entropy": 5.471002292633057,
      "epoch": 1.5266690856313496,
      "grad_norm": 1.0,
      "learning_rate": 0.0004770828164213865,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21224741041660308,
      "num_tokens": 34930243.0,
      "step": 16830
    },
    {
      "entropy": 5.45788197517395,
      "epoch": 1.5271226415094339,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004770685872898525,
      "loss": 5.0918,
      "mean_token_accuracy": 0.20198608040809632,
      "num_tokens": 34941057.0,
      "step": 16835
    },
    {
      "entropy": 5.445715427398682,
      "epoch": 1.5275761973875182,
      "grad_norm": 0.875,
      "learning_rate": 0.00047705435397948876,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21087784171104432,
      "num_tokens": 34951150.0,
      "step": 16840
    },
    {
      "entropy": 5.500549840927124,
      "epoch": 1.5280297532656024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004770401164905895,
      "loss": 5.0544,
      "mean_token_accuracy": 0.20310222804546357,
      "num_tokens": 34961648.0,
      "step": 16845
    },
    {
      "entropy": 5.493280649185181,
      "epoch": 1.5284833091436865,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047702587482344915,
      "loss": 5.0518,
      "mean_token_accuracy": 0.19782250225543976,
      "num_tokens": 34971573.0,
      "step": 16850
    },
    {
      "entropy": 5.533852434158325,
      "epoch": 1.5289368650217707,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004770116289783622,
      "loss": 5.0572,
      "mean_token_accuracy": 0.19783697724342347,
      "num_tokens": 34981984.0,
      "step": 16855
    },
    {
      "entropy": 5.477599287033081,
      "epoch": 1.5293904208998548,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004769973789556233,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19427067041397095,
      "num_tokens": 34990861.0,
      "step": 16860
    },
    {
      "entropy": 5.468864870071411,
      "epoch": 1.529843976777939,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047698312475552714,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19865902960300447,
      "num_tokens": 35001504.0,
      "step": 16865
    },
    {
      "entropy": 5.513449048995971,
      "epoch": 1.5302975326560233,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004769688663783684,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21178119778633117,
      "num_tokens": 35012425.0,
      "step": 16870
    },
    {
      "entropy": 5.4784149646759035,
      "epoch": 1.5307510885341074,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004769546038244421,
      "loss": 5.0175,
      "mean_token_accuracy": 0.2098090261220932,
      "num_tokens": 35022854.0,
      "step": 16875
    },
    {
      "entropy": 5.484833526611328,
      "epoch": 1.5312046444121916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047694033709404313,
      "loss": 5.0873,
      "mean_token_accuracy": 0.19617604464292526,
      "num_tokens": 35032694.0,
      "step": 16880
    },
    {
      "entropy": 5.469423055648804,
      "epoch": 1.5316582002902757,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047692606618746655,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20585087388753892,
      "num_tokens": 35042217.0,
      "step": 16885
    },
    {
      "entropy": 5.493360424041748,
      "epoch": 1.53211175616836,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047691179110500735,
      "loss": 5.0947,
      "mean_token_accuracy": 0.19408292472362518,
      "num_tokens": 35052831.0,
      "step": 16890
    },
    {
      "entropy": 5.550473403930664,
      "epoch": 1.5325653120464442,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047689751184696097,
      "loss": 5.1233,
      "mean_token_accuracy": 0.1965404063463211,
      "num_tokens": 35061955.0,
      "step": 16895
    },
    {
      "entropy": 5.496463680267334,
      "epoch": 1.5330188679245285,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004768832284136226,
      "loss": 5.0052,
      "mean_token_accuracy": 0.2008708745241165,
      "num_tokens": 35072526.0,
      "step": 16900
    },
    {
      "entropy": 5.461774492263794,
      "epoch": 1.5334724238026125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004768689408052876,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20676737129688263,
      "num_tokens": 35082415.0,
      "step": 16905
    },
    {
      "entropy": 5.41850266456604,
      "epoch": 1.5339259796806965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047685464902225144,
      "loss": 4.9846,
      "mean_token_accuracy": 0.20743297040462494,
      "num_tokens": 35092805.0,
      "step": 16910
    },
    {
      "entropy": 5.480488967895508,
      "epoch": 1.5343795355587808,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004768403530648097,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20416479408740998,
      "num_tokens": 35102963.0,
      "step": 16915
    },
    {
      "entropy": 5.512208795547485,
      "epoch": 1.534833091436865,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047682605293325806,
      "loss": 5.0476,
      "mean_token_accuracy": 0.2029525101184845,
      "num_tokens": 35114140.0,
      "step": 16920
    },
    {
      "entropy": 5.556297397613525,
      "epoch": 1.5352866473149493,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047681174862789215,
      "loss": 5.1231,
      "mean_token_accuracy": 0.20083469599485398,
      "num_tokens": 35125314.0,
      "step": 16925
    },
    {
      "entropy": 5.4553551197052,
      "epoch": 1.5357402031930334,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047679744014900795,
      "loss": 5.031,
      "mean_token_accuracy": 0.20120671689510344,
      "num_tokens": 35134934.0,
      "step": 16930
    },
    {
      "entropy": 5.445842266082764,
      "epoch": 1.5361937590711174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004767831274969011,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20646168738603593,
      "num_tokens": 35145341.0,
      "step": 16935
    },
    {
      "entropy": 5.405188322067261,
      "epoch": 1.5366473149492017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047676881067186793,
      "loss": 4.9621,
      "mean_token_accuracy": 0.20318075865507126,
      "num_tokens": 35156046.0,
      "step": 16940
    },
    {
      "entropy": 5.509601879119873,
      "epoch": 1.537100870827286,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047675448967420426,
      "loss": 5.1078,
      "mean_token_accuracy": 0.19604229927062988,
      "num_tokens": 35168232.0,
      "step": 16945
    },
    {
      "entropy": 5.554637670516968,
      "epoch": 1.5375544267053702,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004767401645042064,
      "loss": 5.0774,
      "mean_token_accuracy": 0.19910754561424254,
      "num_tokens": 35178585.0,
      "step": 16950
    },
    {
      "entropy": 5.462687826156616,
      "epoch": 1.5380079825834543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047672583516217043,
      "loss": 4.925,
      "mean_token_accuracy": 0.20886939018964767,
      "num_tokens": 35189417.0,
      "step": 16955
    },
    {
      "entropy": 5.446977043151856,
      "epoch": 1.5384615384615383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004767115016483928,
      "loss": 4.9911,
      "mean_token_accuracy": 0.20271214097738266,
      "num_tokens": 35200167.0,
      "step": 16960
    },
    {
      "entropy": 5.450547170639038,
      "epoch": 1.5389150943396226,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004766971639631699,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2063390478491783,
      "num_tokens": 35210711.0,
      "step": 16965
    },
    {
      "entropy": 5.46052041053772,
      "epoch": 1.5393686502177069,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047668282210679825,
      "loss": 5.027,
      "mean_token_accuracy": 0.19931799173355103,
      "num_tokens": 35220866.0,
      "step": 16970
    },
    {
      "entropy": 5.4065755844116214,
      "epoch": 1.5398222060957911,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004766684760795745,
      "loss": 4.991,
      "mean_token_accuracy": 0.20433398187160492,
      "num_tokens": 35231478.0,
      "step": 16975
    },
    {
      "entropy": 5.4228180885314945,
      "epoch": 1.5402757619738752,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047665412588179524,
      "loss": 4.9414,
      "mean_token_accuracy": 0.21080203652381896,
      "num_tokens": 35242233.0,
      "step": 16980
    },
    {
      "entropy": 5.461465692520141,
      "epoch": 1.5407293178519592,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004766397715137573,
      "loss": 4.9863,
      "mean_token_accuracy": 0.20382720977067947,
      "num_tokens": 35253097.0,
      "step": 16985
    },
    {
      "entropy": 5.542710876464843,
      "epoch": 1.5411828737300435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047662541297575743,
      "loss": 5.1513,
      "mean_token_accuracy": 0.19889177531003951,
      "num_tokens": 35264258.0,
      "step": 16990
    },
    {
      "entropy": 5.590462064743042,
      "epoch": 1.5416364296081277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047661105026809274,
      "loss": 5.1201,
      "mean_token_accuracy": 0.19689531922340392,
      "num_tokens": 35273501.0,
      "step": 16995
    },
    {
      "entropy": 5.468330144882202,
      "epoch": 1.542089985486212,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004765966833910601,
      "loss": 5.0759,
      "mean_token_accuracy": 0.19669248908758163,
      "num_tokens": 35284125.0,
      "step": 17000
    },
    {
      "entropy": 5.526789712905884,
      "epoch": 1.542543541364296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004765823123449567,
      "loss": 5.0628,
      "mean_token_accuracy": 0.1985797628760338,
      "num_tokens": 35296039.0,
      "step": 17005
    },
    {
      "entropy": 5.508655261993408,
      "epoch": 1.54299709724238,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047656793713007965,
      "loss": 5.026,
      "mean_token_accuracy": 0.19966490119695662,
      "num_tokens": 35306178.0,
      "step": 17010
    },
    {
      "entropy": 5.454088258743286,
      "epoch": 1.5434506531204644,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047655355774672636,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20606312453746795,
      "num_tokens": 35316445.0,
      "step": 17015
    },
    {
      "entropy": 5.4946277141571045,
      "epoch": 1.5439042089985486,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047653917419519413,
      "loss": 5.0172,
      "mean_token_accuracy": 0.19846705198287964,
      "num_tokens": 35327274.0,
      "step": 17020
    },
    {
      "entropy": 5.464788627624512,
      "epoch": 1.544357764876633,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004765247864757804,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2081199824810028,
      "num_tokens": 35339007.0,
      "step": 17025
    },
    {
      "entropy": 5.501817512512207,
      "epoch": 1.544811320754717,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004765103945887827,
      "loss": 5.1151,
      "mean_token_accuracy": 0.1939708948135376,
      "num_tokens": 35349303.0,
      "step": 17030
    },
    {
      "entropy": 5.495471715927124,
      "epoch": 1.5452648766328012,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004764959985344987,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2076071947813034,
      "num_tokens": 35359556.0,
      "step": 17035
    },
    {
      "entropy": 5.4351763248443605,
      "epoch": 1.5457184325108853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004764815983132261,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2113174244761467,
      "num_tokens": 35370043.0,
      "step": 17040
    },
    {
      "entropy": 5.492875957489014,
      "epoch": 1.5461719883889695,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004764671939252628,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20762020200490952,
      "num_tokens": 35380031.0,
      "step": 17045
    },
    {
      "entropy": 5.480923652648926,
      "epoch": 1.5466255442670538,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004764527853709064,
      "loss": 5.0822,
      "mean_token_accuracy": 0.19705075174570083,
      "num_tokens": 35390481.0,
      "step": 17050
    },
    {
      "entropy": 5.484901428222656,
      "epoch": 1.5470791001451378,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004764383726504552,
      "loss": 5.0492,
      "mean_token_accuracy": 0.2008197546005249,
      "num_tokens": 35401756.0,
      "step": 17055
    },
    {
      "entropy": 5.496542692184448,
      "epoch": 1.547532656023222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004764239557642071,
      "loss": 4.9519,
      "mean_token_accuracy": 0.210030297935009,
      "num_tokens": 35411045.0,
      "step": 17060
    },
    {
      "entropy": 5.452884578704834,
      "epoch": 1.5479862119013061,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004764095347124603,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19790094047784806,
      "num_tokens": 35421256.0,
      "step": 17065
    },
    {
      "entropy": 5.477028656005859,
      "epoch": 1.5484397677793904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004763951094955129,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20634722858667373,
      "num_tokens": 35430798.0,
      "step": 17070
    },
    {
      "entropy": 5.459696531295776,
      "epoch": 1.5488933236574747,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004763806801136634,
      "loss": 4.98,
      "mean_token_accuracy": 0.20889404863119126,
      "num_tokens": 35440459.0,
      "step": 17075
    },
    {
      "entropy": 5.47511854171753,
      "epoch": 1.549346879535559,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004763662465672101,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20117164999246598,
      "num_tokens": 35451586.0,
      "step": 17080
    },
    {
      "entropy": 5.461205959320068,
      "epoch": 1.549800435413643,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047635180885645146,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20177337080240249,
      "num_tokens": 35462197.0,
      "step": 17085
    },
    {
      "entropy": 5.4906360626220705,
      "epoch": 1.550253991291727,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004763373669816862,
      "loss": 4.976,
      "mean_token_accuracy": 0.20594133883714677,
      "num_tokens": 35471737.0,
      "step": 17090
    },
    {
      "entropy": 5.531971597671509,
      "epoch": 1.5507075471698113,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004763229209432128,
      "loss": 5.0663,
      "mean_token_accuracy": 0.19618499726057054,
      "num_tokens": 35482016.0,
      "step": 17095
    },
    {
      "entropy": 5.5213315963745115,
      "epoch": 1.5511611030478956,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004763084707413301,
      "loss": 5.1459,
      "mean_token_accuracy": 0.19054511934518814,
      "num_tokens": 35492622.0,
      "step": 17100
    },
    {
      "entropy": 5.563022708892822,
      "epoch": 1.5516146589259798,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047629401637633696,
      "loss": 5.1641,
      "mean_token_accuracy": 0.19278692454099655,
      "num_tokens": 35504468.0,
      "step": 17105
    },
    {
      "entropy": 5.48358039855957,
      "epoch": 1.5520682148040639,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047627955784853226,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20332055240869523,
      "num_tokens": 35514272.0,
      "step": 17110
    },
    {
      "entropy": 5.466148757934571,
      "epoch": 1.552521770682148,
      "grad_norm": 1.046875,
      "learning_rate": 0.000476265095158215,
      "loss": 5.0114,
      "mean_token_accuracy": 0.1989905506372452,
      "num_tokens": 35524110.0,
      "step": 17115
    },
    {
      "entropy": 5.470710515975952,
      "epoch": 1.5529753265602322,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004762506283056843,
      "loss": 4.9318,
      "mean_token_accuracy": 0.20816477686166762,
      "num_tokens": 35535504.0,
      "step": 17120
    },
    {
      "entropy": 5.559068012237549,
      "epoch": 1.5534288824383164,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004762361572912393,
      "loss": 5.1155,
      "mean_token_accuracy": 0.19728371649980544,
      "num_tokens": 35544885.0,
      "step": 17125
    },
    {
      "entropy": 5.5046227931976315,
      "epoch": 1.5538824383164007,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047622168211517926,
      "loss": 5.1221,
      "mean_token_accuracy": 0.20438388139009475,
      "num_tokens": 35555485.0,
      "step": 17130
    },
    {
      "entropy": 5.402403116226196,
      "epoch": 1.5543359941944848,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047620720277780353,
      "loss": 4.9691,
      "mean_token_accuracy": 0.20560143738985062,
      "num_tokens": 35567197.0,
      "step": 17135
    },
    {
      "entropy": 5.494960880279541,
      "epoch": 1.5547895500725688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047619271927941165,
      "loss": 4.9806,
      "mean_token_accuracy": 0.20391893535852432,
      "num_tokens": 35577576.0,
      "step": 17140
    },
    {
      "entropy": 5.511598587036133,
      "epoch": 1.555243105950653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000476178231620303,
      "loss": 5.0432,
      "mean_token_accuracy": 0.20576838552951812,
      "num_tokens": 35588012.0,
      "step": 17145
    },
    {
      "entropy": 5.445739316940307,
      "epoch": 1.5556966618287373,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004761637398007774,
      "loss": 5.041,
      "mean_token_accuracy": 0.19566175490617752,
      "num_tokens": 35598076.0,
      "step": 17150
    },
    {
      "entropy": 5.369255447387696,
      "epoch": 1.5561502177068216,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004761492438211342,
      "loss": 4.8647,
      "mean_token_accuracy": 0.21140484064817427,
      "num_tokens": 35607170.0,
      "step": 17155
    },
    {
      "entropy": 5.459671115875244,
      "epoch": 1.5566037735849056,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047613474368167355,
      "loss": 4.9952,
      "mean_token_accuracy": 0.1980486512184143,
      "num_tokens": 35617865.0,
      "step": 17160
    },
    {
      "entropy": 5.46323184967041,
      "epoch": 1.5570573294629897,
      "grad_norm": 1.078125,
      "learning_rate": 0.000476120239382695,
      "loss": 5.052,
      "mean_token_accuracy": 0.2007809489965439,
      "num_tokens": 35629329.0,
      "step": 17165
    },
    {
      "entropy": 5.436124324798584,
      "epoch": 1.557510885341074,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047610573092449877,
      "loss": 5.0538,
      "mean_token_accuracy": 0.2006940186023712,
      "num_tokens": 35640044.0,
      "step": 17170
    },
    {
      "entropy": 5.494205141067505,
      "epoch": 1.5579644412191582,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004760912183073847,
      "loss": 5.0762,
      "mean_token_accuracy": 0.20123564600944518,
      "num_tokens": 35650435.0,
      "step": 17175
    },
    {
      "entropy": 5.4975340366363525,
      "epoch": 1.5584179970972425,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000476076701531653,
      "loss": 4.9882,
      "mean_token_accuracy": 0.19984416663646698,
      "num_tokens": 35659613.0,
      "step": 17180
    },
    {
      "entropy": 5.459752035140991,
      "epoch": 1.5588715529753265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004760621805976039,
      "loss": 4.9859,
      "mean_token_accuracy": 0.20718168765306472,
      "num_tokens": 35670987.0,
      "step": 17185
    },
    {
      "entropy": 5.444336748123169,
      "epoch": 1.5593251088534106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047604765550553766,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2051151990890503,
      "num_tokens": 35682505.0,
      "step": 17190
    },
    {
      "entropy": 5.491911554336548,
      "epoch": 1.5597786647314948,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047603312625575466,
      "loss": 4.9619,
      "mean_token_accuracy": 0.201934751868248,
      "num_tokens": 35693056.0,
      "step": 17195
    },
    {
      "entropy": 5.482431507110595,
      "epoch": 1.5602322206095791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047601859284855544,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2076956808567047,
      "num_tokens": 35703659.0,
      "step": 17200
    },
    {
      "entropy": 5.4399041652679445,
      "epoch": 1.5606857764876634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004760040552842403,
      "loss": 5.0202,
      "mean_token_accuracy": 0.20117031484842302,
      "num_tokens": 35714367.0,
      "step": 17205
    },
    {
      "entropy": 5.427831983566284,
      "epoch": 1.5611393323657474,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047598951356311035,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2012833297252655,
      "num_tokens": 35724773.0,
      "step": 17210
    },
    {
      "entropy": 5.503310632705689,
      "epoch": 1.5615928882438317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004759749676854658,
      "loss": 5.1229,
      "mean_token_accuracy": 0.20039318054914473,
      "num_tokens": 35735622.0,
      "step": 17215
    },
    {
      "entropy": 5.483645105361939,
      "epoch": 1.5620464441219157,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004759604176516079,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20171262621879577,
      "num_tokens": 35744967.0,
      "step": 17220
    },
    {
      "entropy": 5.514835691452026,
      "epoch": 1.5625,
      "grad_norm": 1.0,
      "learning_rate": 0.0004759458634618372,
      "loss": 5.038,
      "mean_token_accuracy": 0.20375445038080214,
      "num_tokens": 35755647.0,
      "step": 17225
    },
    {
      "entropy": 5.5152352809906,
      "epoch": 1.5629535558780843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004759313051164549,
      "loss": 5.0657,
      "mean_token_accuracy": 0.19798210710287095,
      "num_tokens": 35765595.0,
      "step": 17230
    },
    {
      "entropy": 5.488061761856079,
      "epoch": 1.5634071117561683,
      "grad_norm": 0.96875,
      "learning_rate": 0.000475916742615762,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2076081156730652,
      "num_tokens": 35775804.0,
      "step": 17235
    },
    {
      "entropy": 5.439621925354004,
      "epoch": 1.5638606676342526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004759021759600597,
      "loss": 4.9551,
      "mean_token_accuracy": 0.20495265871286392,
      "num_tokens": 35786227.0,
      "step": 17240
    },
    {
      "entropy": 5.534026145935059,
      "epoch": 1.5643142235123366,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004758876051496492,
      "loss": 5.0943,
      "mean_token_accuracy": 0.20041575878858567,
      "num_tokens": 35797511.0,
      "step": 17245
    },
    {
      "entropy": 5.522281551361084,
      "epoch": 1.5647677793904209,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047587303018483177,
      "loss": 5.0393,
      "mean_token_accuracy": 0.1994968682527542,
      "num_tokens": 35807248.0,
      "step": 17250
    },
    {
      "entropy": 5.456273937225342,
      "epoch": 1.5652213352685052,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047585845106590897,
      "loss": 5.0502,
      "mean_token_accuracy": 0.19496332108974457,
      "num_tokens": 35818164.0,
      "step": 17255
    },
    {
      "entropy": 5.452196455001831,
      "epoch": 1.5656748911465894,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047584386779318214,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2102777451276779,
      "num_tokens": 35828746.0,
      "step": 17260
    },
    {
      "entropy": 5.505143165588379,
      "epoch": 1.5661284470246735,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000475829280366953,
      "loss": 5.0681,
      "mean_token_accuracy": 0.19939089864492415,
      "num_tokens": 35839325.0,
      "step": 17265
    },
    {
      "entropy": 5.5186365127563475,
      "epoch": 1.5665820029027575,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047581468878752313,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2021453469991684,
      "num_tokens": 35849406.0,
      "step": 17270
    },
    {
      "entropy": 5.4193699836730955,
      "epoch": 1.5670355587808418,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047580009305519427,
      "loss": 4.9274,
      "mean_token_accuracy": 0.20751346796751022,
      "num_tokens": 35858715.0,
      "step": 17275
    },
    {
      "entropy": 5.423898935317993,
      "epoch": 1.567489114658926,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004757854931702684,
      "loss": 5.0165,
      "mean_token_accuracy": 0.19944613724946975,
      "num_tokens": 35869063.0,
      "step": 17280
    },
    {
      "entropy": 5.410208749771118,
      "epoch": 1.5679426705370103,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047577088913304733,
      "loss": 4.8774,
      "mean_token_accuracy": 0.211069056391716,
      "num_tokens": 35878428.0,
      "step": 17285
    },
    {
      "entropy": 5.506030941009522,
      "epoch": 1.5683962264150944,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047575628094383317,
      "loss": 5.0343,
      "mean_token_accuracy": 0.20377001017332078,
      "num_tokens": 35888801.0,
      "step": 17290
    },
    {
      "entropy": 5.4538154125213625,
      "epoch": 1.5688497822931784,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004757416686029279,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19881603121757507,
      "num_tokens": 35899381.0,
      "step": 17295
    },
    {
      "entropy": 5.532286357879639,
      "epoch": 1.5693033381712627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004757270521106338,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20354195833206176,
      "num_tokens": 35909546.0,
      "step": 17300
    },
    {
      "entropy": 5.494843816757202,
      "epoch": 1.569756894049347,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004757124314672531,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20009820461273192,
      "num_tokens": 35919301.0,
      "step": 17305
    },
    {
      "entropy": 5.554700803756714,
      "epoch": 1.5702104499274312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047569780667308816,
      "loss": 5.0887,
      "mean_token_accuracy": 0.19441407918930054,
      "num_tokens": 35928603.0,
      "step": 17310
    },
    {
      "entropy": 5.518079233169556,
      "epoch": 1.5706640058055152,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047568317772844147,
      "loss": 5.0644,
      "mean_token_accuracy": 0.20570893883705138,
      "num_tokens": 35939005.0,
      "step": 17315
    },
    {
      "entropy": 5.520797824859619,
      "epoch": 1.5711175616835993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047566854463361544,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21340527683496474,
      "num_tokens": 35948471.0,
      "step": 17320
    },
    {
      "entropy": 5.453506898880005,
      "epoch": 1.5715711175616836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004756539073889128,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20369320809841157,
      "num_tokens": 35958488.0,
      "step": 17325
    },
    {
      "entropy": 5.469555282592774,
      "epoch": 1.5720246734397678,
      "grad_norm": 1.125,
      "learning_rate": 0.00047563926599463623,
      "loss": 5.0951,
      "mean_token_accuracy": 0.19554762095212935,
      "num_tokens": 35968291.0,
      "step": 17330
    },
    {
      "entropy": 5.52074933052063,
      "epoch": 1.572478229317852,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004756246204510886,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20523636490106584,
      "num_tokens": 35978565.0,
      "step": 17335
    },
    {
      "entropy": 5.518031311035156,
      "epoch": 1.5729317851959361,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047560997075857253,
      "loss": 4.9782,
      "mean_token_accuracy": 0.2091346338391304,
      "num_tokens": 35987483.0,
      "step": 17340
    },
    {
      "entropy": 5.463924551010132,
      "epoch": 1.5733853410740202,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004755953169173913,
      "loss": 5.05,
      "mean_token_accuracy": 0.2015296921133995,
      "num_tokens": 35998355.0,
      "step": 17345
    },
    {
      "entropy": 5.430660200119019,
      "epoch": 1.5738388969521044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004755806589278477,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20899967551231385,
      "num_tokens": 36008365.0,
      "step": 17350
    },
    {
      "entropy": 5.555283308029175,
      "epoch": 1.5742924528301887,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047556599679024497,
      "loss": 5.0972,
      "mean_token_accuracy": 0.1958293840289116,
      "num_tokens": 36018880.0,
      "step": 17355
    },
    {
      "entropy": 5.51357364654541,
      "epoch": 1.574746008708273,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047555133050488637,
      "loss": 5.0895,
      "mean_token_accuracy": 0.1966924175620079,
      "num_tokens": 36029335.0,
      "step": 17360
    },
    {
      "entropy": 5.453938341140747,
      "epoch": 1.575199564586357,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047553666007207516,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20600392371416093,
      "num_tokens": 36039774.0,
      "step": 17365
    },
    {
      "entropy": 5.546820974349975,
      "epoch": 1.575653120464441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047552198549211463,
      "loss": 5.1191,
      "mean_token_accuracy": 0.2065078005194664,
      "num_tokens": 36049864.0,
      "step": 17370
    },
    {
      "entropy": 5.493776798248291,
      "epoch": 1.5761066763425253,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004755073067653084,
      "loss": 5.0559,
      "mean_token_accuracy": 0.201724611222744,
      "num_tokens": 36059224.0,
      "step": 17375
    },
    {
      "entropy": 5.396344470977783,
      "epoch": 1.5765602322206096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004754926238919599,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2090330973267555,
      "num_tokens": 36069243.0,
      "step": 17380
    },
    {
      "entropy": 5.4583470821380615,
      "epoch": 1.5770137880986939,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004754779368723729,
      "loss": 5.0194,
      "mean_token_accuracy": 0.19869593530893326,
      "num_tokens": 36079833.0,
      "step": 17385
    },
    {
      "entropy": 5.4581746578216555,
      "epoch": 1.577467343976778,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004754632457068511,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20282623767852784,
      "num_tokens": 36090673.0,
      "step": 17390
    },
    {
      "entropy": 5.523070287704468,
      "epoch": 1.5779208998548622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004754485503956982,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20229505598545075,
      "num_tokens": 36101719.0,
      "step": 17395
    },
    {
      "entropy": 5.474701833724976,
      "epoch": 1.5783744557329462,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004754338509392183,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20093010663986205,
      "num_tokens": 36112352.0,
      "step": 17400
    },
    {
      "entropy": 5.390445518493652,
      "epoch": 1.5788280116110305,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004754191473377153,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21064423024654388,
      "num_tokens": 36121797.0,
      "step": 17405
    },
    {
      "entropy": 5.509427547454834,
      "epoch": 1.5792815674891147,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047540443959149316,
      "loss": 5.0366,
      "mean_token_accuracy": 0.2050464019179344,
      "num_tokens": 36132094.0,
      "step": 17410
    },
    {
      "entropy": 5.456129741668701,
      "epoch": 1.5797351233671988,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004753897277008562,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20656346082687377,
      "num_tokens": 36142860.0,
      "step": 17415
    },
    {
      "entropy": 5.454633522033691,
      "epoch": 1.580188679245283,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004753750116661086,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2003056824207306,
      "num_tokens": 36154555.0,
      "step": 17420
    },
    {
      "entropy": 5.552618885040284,
      "epoch": 1.580642235123367,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047536029148755477,
      "loss": 5.1381,
      "mean_token_accuracy": 0.19052140563726425,
      "num_tokens": 36165479.0,
      "step": 17425
    },
    {
      "entropy": 5.520594263076783,
      "epoch": 1.5810957910014514,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047534556716549893,
      "loss": 5.0481,
      "mean_token_accuracy": 0.19806084036827087,
      "num_tokens": 36175873.0,
      "step": 17430
    },
    {
      "entropy": 5.451658058166504,
      "epoch": 1.5815493468795356,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047533083870024575,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20834531933069228,
      "num_tokens": 36186792.0,
      "step": 17435
    },
    {
      "entropy": 5.58074460029602,
      "epoch": 1.58200290275762,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004753161060920998,
      "loss": 5.1355,
      "mean_token_accuracy": 0.19570015370845795,
      "num_tokens": 36196691.0,
      "step": 17440
    },
    {
      "entropy": 5.543552112579346,
      "epoch": 1.582456458635704,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004753013693413657,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20646632611751556,
      "num_tokens": 36207173.0,
      "step": 17445
    },
    {
      "entropy": 5.421489238739014,
      "epoch": 1.582910014513788,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047528662844834827,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2075946345925331,
      "num_tokens": 36217912.0,
      "step": 17450
    },
    {
      "entropy": 5.398891305923462,
      "epoch": 1.5833635703918723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047527188341335233,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21199720352888107,
      "num_tokens": 36227714.0,
      "step": 17455
    },
    {
      "entropy": 5.445325946807861,
      "epoch": 1.5838171262699565,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004752571342366828,
      "loss": 5.0147,
      "mean_token_accuracy": 0.20611771643161775,
      "num_tokens": 36239114.0,
      "step": 17460
    },
    {
      "entropy": 5.477443838119507,
      "epoch": 1.5842706821480408,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004752423809186446,
      "loss": 5.0398,
      "mean_token_accuracy": 0.1966667056083679,
      "num_tokens": 36249436.0,
      "step": 17465
    },
    {
      "entropy": 5.484793281555175,
      "epoch": 1.5847242380261248,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047522762345954305,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20474329739809036,
      "num_tokens": 36259541.0,
      "step": 17470
    },
    {
      "entropy": 5.50005259513855,
      "epoch": 1.5851777939042089,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004752128618596832,
      "loss": 5.0283,
      "mean_token_accuracy": 0.20062110126018523,
      "num_tokens": 36270689.0,
      "step": 17475
    },
    {
      "entropy": 5.544193458557129,
      "epoch": 1.5856313497822931,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047519809611937036,
      "loss": 5.0876,
      "mean_token_accuracy": 0.20171278715133667,
      "num_tokens": 36281191.0,
      "step": 17480
    },
    {
      "entropy": 5.404801845550537,
      "epoch": 1.5860849056603774,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004751833262389098,
      "loss": 4.977,
      "mean_token_accuracy": 0.20927282869815828,
      "num_tokens": 36291022.0,
      "step": 17485
    },
    {
      "entropy": 5.480946731567383,
      "epoch": 1.5865384615384617,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047516855221860715,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20702706277370453,
      "num_tokens": 36301237.0,
      "step": 17490
    },
    {
      "entropy": 5.524582624435425,
      "epoch": 1.5869920174165457,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047515377405876774,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2019112601876259,
      "num_tokens": 36311792.0,
      "step": 17495
    },
    {
      "entropy": 5.429810190200806,
      "epoch": 1.5874455732946298,
      "grad_norm": 1.0,
      "learning_rate": 0.0004751389917596973,
      "loss": 5.002,
      "mean_token_accuracy": 0.21173593252897263,
      "num_tokens": 36321847.0,
      "step": 17500
    },
    {
      "entropy": 5.47958607673645,
      "epoch": 1.587899129172714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047512420532170154,
      "loss": 5.0899,
      "mean_token_accuracy": 0.19647079110145568,
      "num_tokens": 36332735.0,
      "step": 17505
    },
    {
      "entropy": 5.551001119613647,
      "epoch": 1.5883526850507983,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751094147450862,
      "loss": 5.0935,
      "mean_token_accuracy": 0.2003138318657875,
      "num_tokens": 36343149.0,
      "step": 17510
    },
    {
      "entropy": 5.507878112792969,
      "epoch": 1.5888062409288826,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004750946200301571,
      "loss": 4.993,
      "mean_token_accuracy": 0.20172460824251176,
      "num_tokens": 36353081.0,
      "step": 17515
    },
    {
      "entropy": 5.495229005813599,
      "epoch": 1.5892597968069666,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047507982117722035,
      "loss": 5.0591,
      "mean_token_accuracy": 0.19884295612573624,
      "num_tokens": 36363624.0,
      "step": 17520
    },
    {
      "entropy": 5.502510929107666,
      "epoch": 1.5897133526850507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000475065018186582,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20070028454065322,
      "num_tokens": 36374081.0,
      "step": 17525
    },
    {
      "entropy": 5.531485462188721,
      "epoch": 1.590166908563135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047505021105854794,
      "loss": 5.152,
      "mean_token_accuracy": 0.19617709517478943,
      "num_tokens": 36384360.0,
      "step": 17530
    },
    {
      "entropy": 5.496904516220093,
      "epoch": 1.5906204644412192,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047503539979342456,
      "loss": 5.0613,
      "mean_token_accuracy": 0.2044987738132477,
      "num_tokens": 36395924.0,
      "step": 17535
    },
    {
      "entropy": 5.415366888046265,
      "epoch": 1.5910740203193035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004750205843915182,
      "loss": 4.9455,
      "mean_token_accuracy": 0.20962637960910796,
      "num_tokens": 36405246.0,
      "step": 17540
    },
    {
      "entropy": 5.430236911773681,
      "epoch": 1.5915275761973875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004750057648531351,
      "loss": 4.9812,
      "mean_token_accuracy": 0.2136492133140564,
      "num_tokens": 36414964.0,
      "step": 17545
    },
    {
      "entropy": 5.446323585510254,
      "epoch": 1.5919811320754715,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004749909411785819,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20096035599708556,
      "num_tokens": 36425011.0,
      "step": 17550
    },
    {
      "entropy": 5.405447196960449,
      "epoch": 1.5924346879535558,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000474976113368165,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21336206048727036,
      "num_tokens": 36434734.0,
      "step": 17555
    },
    {
      "entropy": 5.499809455871582,
      "epoch": 1.59288824383164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004749612814221912,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20356599539518355,
      "num_tokens": 36444748.0,
      "step": 17560
    },
    {
      "entropy": 5.572061347961426,
      "epoch": 1.5933417997097243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047494644534096703,
      "loss": 5.0557,
      "mean_token_accuracy": 0.19582525193691253,
      "num_tokens": 36455076.0,
      "step": 17565
    },
    {
      "entropy": 5.4799888134002686,
      "epoch": 1.5937953555878084,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004749316051247995,
      "loss": 4.9621,
      "mean_token_accuracy": 0.20652810633182525,
      "num_tokens": 36465302.0,
      "step": 17570
    },
    {
      "entropy": 5.472163486480713,
      "epoch": 1.5942489114658926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047491676077399543,
      "loss": 5.082,
      "mean_token_accuracy": 0.1985471338033676,
      "num_tokens": 36476666.0,
      "step": 17575
    },
    {
      "entropy": 5.448248291015625,
      "epoch": 1.5947024673439767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004749019122888618,
      "loss": 4.9232,
      "mean_token_accuracy": 0.20897077918052673,
      "num_tokens": 36487379.0,
      "step": 17580
    },
    {
      "entropy": 5.476648759841919,
      "epoch": 1.595156023222061,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047488705966970565,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2087975949048996,
      "num_tokens": 36497897.0,
      "step": 17585
    },
    {
      "entropy": 5.47394323348999,
      "epoch": 1.5956095791001452,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004748722029168342,
      "loss": 5.0716,
      "mean_token_accuracy": 0.21110234707593917,
      "num_tokens": 36508528.0,
      "step": 17590
    },
    {
      "entropy": 5.546425628662109,
      "epoch": 1.5960631349782293,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004748573420305546,
      "loss": 5.0717,
      "mean_token_accuracy": 0.19552092254161835,
      "num_tokens": 36519196.0,
      "step": 17595
    },
    {
      "entropy": 5.486793088912964,
      "epoch": 1.5965166908563135,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047484247701117427,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20387091487646103,
      "num_tokens": 36529871.0,
      "step": 17600
    },
    {
      "entropy": 5.559619188308716,
      "epoch": 1.5969702467343976,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004748276078590006,
      "loss": 5.1168,
      "mean_token_accuracy": 0.1933345079421997,
      "num_tokens": 36541188.0,
      "step": 17605
    },
    {
      "entropy": 5.570701646804809,
      "epoch": 1.5974238026124818,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004748127345743411,
      "loss": 5.0872,
      "mean_token_accuracy": 0.19958331286907197,
      "num_tokens": 36551968.0,
      "step": 17610
    },
    {
      "entropy": 5.499534845352173,
      "epoch": 1.5978773584905661,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004747978571575033,
      "loss": 5.0261,
      "mean_token_accuracy": 0.20338878333568572,
      "num_tokens": 36562870.0,
      "step": 17615
    },
    {
      "entropy": 5.434991502761841,
      "epoch": 1.5983309143686504,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004747829756087949,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20664169192314147,
      "num_tokens": 36573295.0,
      "step": 17620
    },
    {
      "entropy": 5.422763395309448,
      "epoch": 1.5987844702467344,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047476808992852364,
      "loss": 4.9687,
      "mean_token_accuracy": 0.20708438158035278,
      "num_tokens": 36583830.0,
      "step": 17625
    },
    {
      "entropy": 5.492862510681152,
      "epoch": 1.5992380261248185,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004747532001169974,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20989180207252503,
      "num_tokens": 36594052.0,
      "step": 17630
    },
    {
      "entropy": 5.492591047286988,
      "epoch": 1.5996915820029027,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047473830617452415,
      "loss": 5.0543,
      "mean_token_accuracy": 0.2044448286294937,
      "num_tokens": 36604378.0,
      "step": 17635
    },
    {
      "entropy": 5.549684762954712,
      "epoch": 1.600145137880987,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047472340810141163,
      "loss": 5.1528,
      "mean_token_accuracy": 0.1934247389435768,
      "num_tokens": 36615588.0,
      "step": 17640
    },
    {
      "entropy": 5.529037714004517,
      "epoch": 1.6005986937590713,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047470850589796833,
      "loss": 5.0221,
      "mean_token_accuracy": 0.2031117007136345,
      "num_tokens": 36626369.0,
      "step": 17645
    },
    {
      "entropy": 5.557148599624634,
      "epoch": 1.6010522496371553,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047469359956450215,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20016029179096223,
      "num_tokens": 36636565.0,
      "step": 17650
    },
    {
      "entropy": 5.576334810256958,
      "epoch": 1.6015058055152394,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047467868910132144,
      "loss": 5.143,
      "mean_token_accuracy": 0.195332433283329,
      "num_tokens": 36647721.0,
      "step": 17655
    },
    {
      "entropy": 5.43845763206482,
      "epoch": 1.6019593613933236,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004746637745087346,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21079145073890687,
      "num_tokens": 36657887.0,
      "step": 17660
    },
    {
      "entropy": 5.54886908531189,
      "epoch": 1.6024129172714079,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047464885578705,
      "loss": 5.0325,
      "mean_token_accuracy": 0.19845932424068452,
      "num_tokens": 36667918.0,
      "step": 17665
    },
    {
      "entropy": 5.412156295776367,
      "epoch": 1.6028664731494922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047463393293657616,
      "loss": 4.997,
      "mean_token_accuracy": 0.20855554342269897,
      "num_tokens": 36678491.0,
      "step": 17670
    },
    {
      "entropy": 5.461192035675049,
      "epoch": 1.6033200290275762,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004746190059576217,
      "loss": 5.0682,
      "mean_token_accuracy": 0.20264270305633544,
      "num_tokens": 36688638.0,
      "step": 17675
    },
    {
      "entropy": 5.509416723251343,
      "epoch": 1.6037735849056602,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004746040748504954,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20608117133378984,
      "num_tokens": 36699344.0,
      "step": 17680
    },
    {
      "entropy": 5.536354207992554,
      "epoch": 1.6042271407837445,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047458913961550593,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20165850818157197,
      "num_tokens": 36709606.0,
      "step": 17685
    },
    {
      "entropy": 5.404119205474854,
      "epoch": 1.6046806966618288,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004745742002529622,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20458258390426637,
      "num_tokens": 36718740.0,
      "step": 17690
    },
    {
      "entropy": 5.48478217124939,
      "epoch": 1.605134252539913,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004745592567631731,
      "loss": 5.0489,
      "mean_token_accuracy": 0.19957852363586426,
      "num_tokens": 36729960.0,
      "step": 17695
    },
    {
      "entropy": 5.447217512130737,
      "epoch": 1.605587808417997,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047454430914644773,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2100689634680748,
      "num_tokens": 36740221.0,
      "step": 17700
    },
    {
      "entropy": 5.379448938369751,
      "epoch": 1.6060413642960811,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004745293574030952,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21736128628253937,
      "num_tokens": 36750841.0,
      "step": 17705
    },
    {
      "entropy": 5.435540294647216,
      "epoch": 1.6064949201741654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047451440153342477,
      "loss": 5.0262,
      "mean_token_accuracy": 0.196967151761055,
      "num_tokens": 36760944.0,
      "step": 17710
    },
    {
      "entropy": 5.5174620151519775,
      "epoch": 1.6069484760522497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004744994415377457,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20317866504192353,
      "num_tokens": 36770922.0,
      "step": 17715
    },
    {
      "entropy": 5.5168322086334225,
      "epoch": 1.607402031930334,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047448447741636724,
      "loss": 5.0333,
      "mean_token_accuracy": 0.19724370688199996,
      "num_tokens": 36780702.0,
      "step": 17720
    },
    {
      "entropy": 5.512592697143555,
      "epoch": 1.607855587808418,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047446950916959905,
      "loss": 5.0667,
      "mean_token_accuracy": 0.1964994862675667,
      "num_tokens": 36791868.0,
      "step": 17725
    },
    {
      "entropy": 5.527849912643433,
      "epoch": 1.608309143686502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004744545367977505,
      "loss": 5.0488,
      "mean_token_accuracy": 0.20466661155223848,
      "num_tokens": 36802330.0,
      "step": 17730
    },
    {
      "entropy": 5.483583068847656,
      "epoch": 1.6087626995645863,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047443956030113134,
      "loss": 5.0719,
      "mean_token_accuracy": 0.19980328381061555,
      "num_tokens": 36812936.0,
      "step": 17735
    },
    {
      "entropy": 5.525945472717285,
      "epoch": 1.6092162554426706,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047442457968005123,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20068635195493698,
      "num_tokens": 36823123.0,
      "step": 17740
    },
    {
      "entropy": 5.415410566329956,
      "epoch": 1.6096698113207548,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047440959493481996,
      "loss": 4.907,
      "mean_token_accuracy": 0.21537703424692153,
      "num_tokens": 36833050.0,
      "step": 17745
    },
    {
      "entropy": 5.484205007553101,
      "epoch": 1.6101233671988389,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047439460606574745,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20483941286802293,
      "num_tokens": 36842520.0,
      "step": 17750
    },
    {
      "entropy": 5.514116859436035,
      "epoch": 1.6105769230769231,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004743796130731437,
      "loss": 5.0126,
      "mean_token_accuracy": 0.20166711509227753,
      "num_tokens": 36853024.0,
      "step": 17755
    },
    {
      "entropy": 5.528896522521973,
      "epoch": 1.6110304789550072,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004743646159573188,
      "loss": 4.9752,
      "mean_token_accuracy": 0.20968101918697357,
      "num_tokens": 36862534.0,
      "step": 17760
    },
    {
      "entropy": 5.441176795959473,
      "epoch": 1.6114840348330914,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004743496147185828,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2123408406972885,
      "num_tokens": 36872356.0,
      "step": 17765
    },
    {
      "entropy": 5.438937854766846,
      "epoch": 1.6119375907111757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047433460935724595,
      "loss": 5.0246,
      "mean_token_accuracy": 0.20223264694213866,
      "num_tokens": 36882643.0,
      "step": 17770
    },
    {
      "entropy": 5.512832260131836,
      "epoch": 1.6123911465892597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004743195998736186,
      "loss": 5.0085,
      "mean_token_accuracy": 0.2032598689198494,
      "num_tokens": 36893349.0,
      "step": 17775
    },
    {
      "entropy": 5.539227628707886,
      "epoch": 1.612844702467344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047430458626801107,
      "loss": 5.02,
      "mean_token_accuracy": 0.2073807030916214,
      "num_tokens": 36903713.0,
      "step": 17780
    },
    {
      "entropy": 5.456194686889648,
      "epoch": 1.613298258345428,
      "grad_norm": 1.109375,
      "learning_rate": 0.000474289568540734,
      "loss": 4.935,
      "mean_token_accuracy": 0.21000806093215943,
      "num_tokens": 36912999.0,
      "step": 17785
    },
    {
      "entropy": 5.504201412200928,
      "epoch": 1.6137518142235123,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047427454669209785,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20465178042650223,
      "num_tokens": 36923598.0,
      "step": 17790
    },
    {
      "entropy": 5.464404344558716,
      "epoch": 1.6142053701015966,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047425952072241326,
      "loss": 4.9777,
      "mean_token_accuracy": 0.20436245054006577,
      "num_tokens": 36933036.0,
      "step": 17795
    },
    {
      "entropy": 5.493456077575684,
      "epoch": 1.6146589259796809,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000474244490631991,
      "loss": 5.0794,
      "mean_token_accuracy": 0.19855207800865174,
      "num_tokens": 36943137.0,
      "step": 17800
    },
    {
      "entropy": 5.6092346668243405,
      "epoch": 1.615112481857765,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004742294564211419,
      "loss": 5.1892,
      "mean_token_accuracy": 0.19220228791236876,
      "num_tokens": 36955619.0,
      "step": 17805
    },
    {
      "entropy": 5.4836140155792235,
      "epoch": 1.615566037735849,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004742144180901769,
      "loss": 4.9685,
      "mean_token_accuracy": 0.21661929041147232,
      "num_tokens": 36966377.0,
      "step": 17810
    },
    {
      "entropy": 5.483592319488525,
      "epoch": 1.6160195936139332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047419937563940697,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21046488732099533,
      "num_tokens": 36977100.0,
      "step": 17815
    },
    {
      "entropy": 5.4998737335205075,
      "epoch": 1.6164731494920175,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047418432906914314,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19394870847463608,
      "num_tokens": 36987120.0,
      "step": 17820
    },
    {
      "entropy": 5.5570456981658936,
      "epoch": 1.6169267053701017,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004741692783796967,
      "loss": 5.0517,
      "mean_token_accuracy": 0.20102904438972474,
      "num_tokens": 36996885.0,
      "step": 17825
    },
    {
      "entropy": 5.4943199634552,
      "epoch": 1.6173802612481858,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047415422357137876,
      "loss": 4.981,
      "mean_token_accuracy": 0.21046871095895767,
      "num_tokens": 37006570.0,
      "step": 17830
    },
    {
      "entropy": 5.528774309158325,
      "epoch": 1.6178338171262698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004741391646445008,
      "loss": 5.0817,
      "mean_token_accuracy": 0.1988207072019577,
      "num_tokens": 37017185.0,
      "step": 17835
    },
    {
      "entropy": 5.501001405715942,
      "epoch": 1.618287373004354,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004741241015993741,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20650382190942765,
      "num_tokens": 37028386.0,
      "step": 17840
    },
    {
      "entropy": 5.441536998748779,
      "epoch": 1.6187409288824384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047410903443631026,
      "loss": 4.9536,
      "mean_token_accuracy": 0.20377426147460936,
      "num_tokens": 37038326.0,
      "step": 17845
    },
    {
      "entropy": 5.528948402404785,
      "epoch": 1.6191944847605226,
      "grad_norm": 1.0,
      "learning_rate": 0.0004740939631556209,
      "loss": 5.0402,
      "mean_token_accuracy": 0.2045020267367363,
      "num_tokens": 37048570.0,
      "step": 17850
    },
    {
      "entropy": 5.433118009567261,
      "epoch": 1.6196480406386067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004740788877576176,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2163080468773842,
      "num_tokens": 37058909.0,
      "step": 17855
    },
    {
      "entropy": 5.483098125457763,
      "epoch": 1.6201015965166907,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004740638082426121,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2163893461227417,
      "num_tokens": 37068957.0,
      "step": 17860
    },
    {
      "entropy": 5.472447204589844,
      "epoch": 1.620555152394775,
      "grad_norm": 1.0,
      "learning_rate": 0.00047404872461091643,
      "loss": 4.9738,
      "mean_token_accuracy": 0.20412914603948593,
      "num_tokens": 37080187.0,
      "step": 17865
    },
    {
      "entropy": 5.518881130218506,
      "epoch": 1.6210087082728593,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047403363686284236,
      "loss": 5.0354,
      "mean_token_accuracy": 0.20058336704969407,
      "num_tokens": 37090314.0,
      "step": 17870
    },
    {
      "entropy": 5.55664210319519,
      "epoch": 1.6214622641509435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047401854499870197,
      "loss": 5.0831,
      "mean_token_accuracy": 0.2025315895676613,
      "num_tokens": 37099907.0,
      "step": 17875
    },
    {
      "entropy": 5.510245418548584,
      "epoch": 1.6219158200290276,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047400344901880735,
      "loss": 5.0881,
      "mean_token_accuracy": 0.19591856598854065,
      "num_tokens": 37110495.0,
      "step": 17880
    },
    {
      "entropy": 5.491468000411987,
      "epoch": 1.6223693759071116,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004739883489234707,
      "loss": 5.0595,
      "mean_token_accuracy": 0.20072028636932374,
      "num_tokens": 37121957.0,
      "step": 17885
    },
    {
      "entropy": 5.482203578948974,
      "epoch": 1.6228229317851959,
      "grad_norm": 1.0,
      "learning_rate": 0.00047397324471300424,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20953701436519623,
      "num_tokens": 37132743.0,
      "step": 17890
    },
    {
      "entropy": 5.511427783966065,
      "epoch": 1.6232764876632801,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047395813638772047,
      "loss": 5.0635,
      "mean_token_accuracy": 0.2001722663640976,
      "num_tokens": 37143947.0,
      "step": 17895
    },
    {
      "entropy": 5.624027299880981,
      "epoch": 1.6237300435413644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047394302394793164,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19035278409719467,
      "num_tokens": 37154316.0,
      "step": 17900
    },
    {
      "entropy": 5.544618606567383,
      "epoch": 1.6241835994194485,
      "grad_norm": 1.0,
      "learning_rate": 0.0004739279073939504,
      "loss": 5.0725,
      "mean_token_accuracy": 0.2005895271897316,
      "num_tokens": 37164999.0,
      "step": 17905
    },
    {
      "entropy": 5.517786598205566,
      "epoch": 1.6246371552975325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004739127867260894,
      "loss": 5.0635,
      "mean_token_accuracy": 0.20029191970825194,
      "num_tokens": 37174392.0,
      "step": 17910
    },
    {
      "entropy": 5.431928443908691,
      "epoch": 1.6250907111756168,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004738976619446613,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20038048475980758,
      "num_tokens": 37185341.0,
      "step": 17915
    },
    {
      "entropy": 5.522448110580444,
      "epoch": 1.625544267053701,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047388253304997874,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19450522512197493,
      "num_tokens": 37196725.0,
      "step": 17920
    },
    {
      "entropy": 5.511097574234009,
      "epoch": 1.6259978229317853,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004738674000423548,
      "loss": 4.9513,
      "mean_token_accuracy": 0.21172694861888885,
      "num_tokens": 37206877.0,
      "step": 17925
    },
    {
      "entropy": 5.4013652324676515,
      "epoch": 1.6264513788098693,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004738522629221023,
      "loss": 4.9395,
      "mean_token_accuracy": 0.20670881420373916,
      "num_tokens": 37217277.0,
      "step": 17930
    },
    {
      "entropy": 5.424076795578003,
      "epoch": 1.6269049346879536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004738371216895344,
      "loss": 4.9386,
      "mean_token_accuracy": 0.20832673758268355,
      "num_tokens": 37227716.0,
      "step": 17935
    },
    {
      "entropy": 5.435279035568238,
      "epoch": 1.6273584905660377,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004738219763449641,
      "loss": 5.0389,
      "mean_token_accuracy": 0.19950771182775498,
      "num_tokens": 37238178.0,
      "step": 17940
    },
    {
      "entropy": 5.485125970840454,
      "epoch": 1.627812046444122,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004738068268887047,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20506935119628905,
      "num_tokens": 37248408.0,
      "step": 17945
    },
    {
      "entropy": 5.490612411499024,
      "epoch": 1.6282656023222062,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004737916733210694,
      "loss": 5.1667,
      "mean_token_accuracy": 0.19898591786623002,
      "num_tokens": 37259176.0,
      "step": 17950
    },
    {
      "entropy": 5.517525768280029,
      "epoch": 1.6287191582002902,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047377651564237165,
      "loss": 4.931,
      "mean_token_accuracy": 0.20828703343868255,
      "num_tokens": 37269205.0,
      "step": 17955
    },
    {
      "entropy": 5.534256649017334,
      "epoch": 1.6291727140783745,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047376135385292494,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20858412683010102,
      "num_tokens": 37278841.0,
      "step": 17960
    },
    {
      "entropy": 5.498790836334228,
      "epoch": 1.6296262699564585,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004737461879530428,
      "loss": 5.0783,
      "mean_token_accuracy": 0.1986847698688507,
      "num_tokens": 37288487.0,
      "step": 17965
    },
    {
      "entropy": 5.504429960250855,
      "epoch": 1.6300798258345428,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047373101794303884,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20283679962158202,
      "num_tokens": 37300086.0,
      "step": 17970
    },
    {
      "entropy": 5.500997257232666,
      "epoch": 1.630533381712627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004737158438232267,
      "loss": 5.0817,
      "mean_token_accuracy": 0.1948433056473732,
      "num_tokens": 37310072.0,
      "step": 17975
    },
    {
      "entropy": 5.528009462356567,
      "epoch": 1.6309869375907113,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047370066559392033,
      "loss": 5.0087,
      "mean_token_accuracy": 0.213375024497509,
      "num_tokens": 37320278.0,
      "step": 17980
    },
    {
      "entropy": 5.514726829528809,
      "epoch": 1.6314404934687954,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047368548325543366,
      "loss": 4.9738,
      "mean_token_accuracy": 0.19879747331142425,
      "num_tokens": 37330741.0,
      "step": 17985
    },
    {
      "entropy": 5.4943592071533205,
      "epoch": 1.6318940493468794,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004736702968080804,
      "loss": 5.0616,
      "mean_token_accuracy": 0.19932493865489959,
      "num_tokens": 37341213.0,
      "step": 17990
    },
    {
      "entropy": 5.419654178619385,
      "epoch": 1.6323476052249637,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047365510625217483,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2059243679046631,
      "num_tokens": 37350528.0,
      "step": 17995
    },
    {
      "entropy": 5.4861565113067625,
      "epoch": 1.632801161103048,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047363991158803117,
      "loss": 5.0122,
      "mean_token_accuracy": 0.203314208984375,
      "num_tokens": 37359938.0,
      "step": 18000
    },
    {
      "epoch": 1.632801161103048,
      "eval_entropy": 5.260263482585752,
      "eval_loss": 5.059215068817139,
      "eval_mean_token_accuracy": 0.21069170047963098,
      "eval_num_tokens": 37359938.0,
      "eval_runtime": 20.762,
      "eval_samples_per_second": 1703.112,
      "eval_steps_per_second": 212.889,
      "step": 18000
    },
    {
      "entropy": 5.504788255691528,
      "epoch": 1.6332547169811322,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047362471281596334,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2054093912243843,
      "num_tokens": 37369744.0,
      "step": 18005
    },
    {
      "entropy": 5.53725905418396,
      "epoch": 1.6337082728592163,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047360950993628585,
      "loss": 5.1098,
      "mean_token_accuracy": 0.19825389236211777,
      "num_tokens": 37380578.0,
      "step": 18010
    },
    {
      "entropy": 5.488230991363525,
      "epoch": 1.6341618287373003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004735943029493132,
      "loss": 5.0264,
      "mean_token_accuracy": 0.19835548251867294,
      "num_tokens": 37390237.0,
      "step": 18015
    },
    {
      "entropy": 5.49710602760315,
      "epoch": 1.6346153846153846,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047357909185535965,
      "loss": 5.0715,
      "mean_token_accuracy": 0.1987991601228714,
      "num_tokens": 37400261.0,
      "step": 18020
    },
    {
      "entropy": 5.4864672183990475,
      "epoch": 1.6350689404934688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004735638766547399,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20591530203819275,
      "num_tokens": 37410012.0,
      "step": 18025
    },
    {
      "entropy": 5.531466436386109,
      "epoch": 1.6355224963715531,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047354865734776857,
      "loss": 5.1278,
      "mean_token_accuracy": 0.18817878067493438,
      "num_tokens": 37420530.0,
      "step": 18030
    },
    {
      "entropy": 5.503242444992066,
      "epoch": 1.6359760522496372,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004735334339347605,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21525170058012008,
      "num_tokens": 37431146.0,
      "step": 18035
    },
    {
      "entropy": 5.489639949798584,
      "epoch": 1.6364296081277212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047351820641603035,
      "loss": 4.9571,
      "mean_token_accuracy": 0.20719221830368043,
      "num_tokens": 37441253.0,
      "step": 18040
    },
    {
      "entropy": 5.406484794616699,
      "epoch": 1.6368831640058055,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004735029747918931,
      "loss": 5.0266,
      "mean_token_accuracy": 0.19968244284391404,
      "num_tokens": 37451326.0,
      "step": 18045
    },
    {
      "entropy": 5.427188110351563,
      "epoch": 1.6373367198838897,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047348773906266374,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2051646888256073,
      "num_tokens": 37462462.0,
      "step": 18050
    },
    {
      "entropy": 5.4523402690887455,
      "epoch": 1.637790275761974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004734724992286574,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20920010507106782,
      "num_tokens": 37473016.0,
      "step": 18055
    },
    {
      "entropy": 5.520586776733398,
      "epoch": 1.638243831640058,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004734572552901892,
      "loss": 5.1017,
      "mean_token_accuracy": 0.20339616537094116,
      "num_tokens": 37483737.0,
      "step": 18060
    },
    {
      "entropy": 5.467749977111817,
      "epoch": 1.638697387518142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004734420072475743,
      "loss": 4.9777,
      "mean_token_accuracy": 0.2044924795627594,
      "num_tokens": 37493070.0,
      "step": 18065
    },
    {
      "entropy": 5.478689813613892,
      "epoch": 1.6391509433962264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004734267551011282,
      "loss": 4.9813,
      "mean_token_accuracy": 0.2088460236787796,
      "num_tokens": 37503696.0,
      "step": 18070
    },
    {
      "entropy": 5.438482141494751,
      "epoch": 1.6396044992743106,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004734114988511662,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20678571611642838,
      "num_tokens": 37513468.0,
      "step": 18075
    },
    {
      "entropy": 5.3958323955535885,
      "epoch": 1.640058055152395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004733962384980039,
      "loss": 4.8767,
      "mean_token_accuracy": 0.21020763814449311,
      "num_tokens": 37523631.0,
      "step": 18080
    },
    {
      "entropy": 5.505127477645874,
      "epoch": 1.640511611030479,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004733809740419568,
      "loss": 5.0625,
      "mean_token_accuracy": 0.2120346814393997,
      "num_tokens": 37534602.0,
      "step": 18085
    },
    {
      "entropy": 5.4306511878967285,
      "epoch": 1.640965166908563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047336570548334053,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2057059645652771,
      "num_tokens": 37544547.0,
      "step": 18090
    },
    {
      "entropy": 5.508368301391601,
      "epoch": 1.6414187227866472,
      "grad_norm": 1.0,
      "learning_rate": 0.00047335043282247105,
      "loss": 5.0177,
      "mean_token_accuracy": 0.2000522419810295,
      "num_tokens": 37554903.0,
      "step": 18095
    },
    {
      "entropy": 5.484733724594117,
      "epoch": 1.6418722786647315,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047333515605966396,
      "loss": 4.9458,
      "mean_token_accuracy": 0.20402987599372863,
      "num_tokens": 37565586.0,
      "step": 18100
    },
    {
      "entropy": 5.493042945861816,
      "epoch": 1.6423258345428158,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047331987519523533,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2090039610862732,
      "num_tokens": 37575996.0,
      "step": 18105
    },
    {
      "entropy": 5.419713401794434,
      "epoch": 1.6427793904208998,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004733045902295012,
      "loss": 4.9442,
      "mean_token_accuracy": 0.1986428067088127,
      "num_tokens": 37586441.0,
      "step": 18110
    },
    {
      "entropy": 5.495657444000244,
      "epoch": 1.643232946298984,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047328930116277755,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20472251176834105,
      "num_tokens": 37598226.0,
      "step": 18115
    },
    {
      "entropy": 5.479895353317261,
      "epoch": 1.6436865021770681,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004732740079953806,
      "loss": 5.0536,
      "mean_token_accuracy": 0.2034666508436203,
      "num_tokens": 37609102.0,
      "step": 18120
    },
    {
      "entropy": 5.482530164718628,
      "epoch": 1.6441400580551524,
      "grad_norm": 1.0,
      "learning_rate": 0.0004732587107276267,
      "loss": 5.0476,
      "mean_token_accuracy": 0.20244789868593216,
      "num_tokens": 37619349.0,
      "step": 18125
    },
    {
      "entropy": 5.483021640777588,
      "epoch": 1.6445936139332367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047324340935983225,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21043339371681213,
      "num_tokens": 37629844.0,
      "step": 18130
    },
    {
      "entropy": 5.4109281539917,
      "epoch": 1.6450471698113207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047322810389231333,
      "loss": 4.945,
      "mean_token_accuracy": 0.2070995345711708,
      "num_tokens": 37639339.0,
      "step": 18135
    },
    {
      "entropy": 5.4147608280181885,
      "epoch": 1.645500725689405,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004732127943253868,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21008183807134628,
      "num_tokens": 37648836.0,
      "step": 18140
    },
    {
      "entropy": 5.49785008430481,
      "epoch": 1.645954281567489,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004731974806593693,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20498585402965547,
      "num_tokens": 37658424.0,
      "step": 18145
    },
    {
      "entropy": 5.483016633987427,
      "epoch": 1.6464078374455733,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047318216289457727,
      "loss": 4.99,
      "mean_token_accuracy": 0.20634360313415528,
      "num_tokens": 37669849.0,
      "step": 18150
    },
    {
      "entropy": 5.461455488204956,
      "epoch": 1.6468613933236576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047316684103132766,
      "loss": 4.999,
      "mean_token_accuracy": 0.2050923764705658,
      "num_tokens": 37678987.0,
      "step": 18155
    },
    {
      "entropy": 5.512372159957886,
      "epoch": 1.6473149492017418,
      "grad_norm": 1.0,
      "learning_rate": 0.0004731515150699372,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20386629551649094,
      "num_tokens": 37689692.0,
      "step": 18160
    },
    {
      "entropy": 5.5446165084838865,
      "epoch": 1.6477685050798259,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047313618501072303,
      "loss": 5.1572,
      "mean_token_accuracy": 0.19308646470308305,
      "num_tokens": 37700092.0,
      "step": 18165
    },
    {
      "entropy": 5.49167947769165,
      "epoch": 1.64822206095791,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047312085085400194,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20333592146635054,
      "num_tokens": 37710305.0,
      "step": 18170
    },
    {
      "entropy": 5.602780675888061,
      "epoch": 1.6486756168359942,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047310551260009126,
      "loss": 5.1349,
      "mean_token_accuracy": 0.19278147369623183,
      "num_tokens": 37720451.0,
      "step": 18175
    },
    {
      "entropy": 5.512181091308594,
      "epoch": 1.6491291727140784,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000473090170249308,
      "loss": 5.0226,
      "mean_token_accuracy": 0.20533677637577058,
      "num_tokens": 37731952.0,
      "step": 18180
    },
    {
      "entropy": 5.4516700267791744,
      "epoch": 1.6495827285921627,
      "grad_norm": 1.0,
      "learning_rate": 0.0004730748238019696,
      "loss": 4.9713,
      "mean_token_accuracy": 0.203942608833313,
      "num_tokens": 37742201.0,
      "step": 18185
    },
    {
      "entropy": 5.4428267002105715,
      "epoch": 1.6500362844702468,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047305947325839337,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20398137122392654,
      "num_tokens": 37753317.0,
      "step": 18190
    },
    {
      "entropy": 5.5094935417175295,
      "epoch": 1.6504898403483308,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004730441186188968,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20391041785478592,
      "num_tokens": 37763452.0,
      "step": 18195
    },
    {
      "entropy": 5.481544160842896,
      "epoch": 1.650943396226415,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047302875988379734,
      "loss": 4.9627,
      "mean_token_accuracy": 0.20198788344860077,
      "num_tokens": 37772804.0,
      "step": 18200
    },
    {
      "entropy": 5.504287528991699,
      "epoch": 1.6513969521044993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047301339705341266,
      "loss": 5.0393,
      "mean_token_accuracy": 0.19558282643556596,
      "num_tokens": 37784441.0,
      "step": 18205
    },
    {
      "entropy": 5.483072805404663,
      "epoch": 1.6518505079825836,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047299803012806056,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20530251711606978,
      "num_tokens": 37795103.0,
      "step": 18210
    },
    {
      "entropy": 5.471498250961304,
      "epoch": 1.6523040638606676,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004729826591080586,
      "loss": 5.0386,
      "mean_token_accuracy": 0.20357936173677443,
      "num_tokens": 37806654.0,
      "step": 18215
    },
    {
      "entropy": 5.503791809082031,
      "epoch": 1.6527576197387517,
      "grad_norm": 1.125,
      "learning_rate": 0.00047296728399372483,
      "loss": 5.013,
      "mean_token_accuracy": 0.20491640120744706,
      "num_tokens": 37816673.0,
      "step": 18220
    },
    {
      "entropy": 5.537406301498413,
      "epoch": 1.653211175616836,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047295190478537725,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19826985746622086,
      "num_tokens": 37827389.0,
      "step": 18225
    },
    {
      "entropy": 5.5817930698394775,
      "epoch": 1.6536647314949202,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047293652148333376,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20327634662389754,
      "num_tokens": 37838087.0,
      "step": 18230
    },
    {
      "entropy": 5.4010388374328615,
      "epoch": 1.6541182873730045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047292113408791266,
      "loss": 4.9429,
      "mean_token_accuracy": 0.20644395798444748,
      "num_tokens": 37848630.0,
      "step": 18235
    },
    {
      "entropy": 5.408254671096802,
      "epoch": 1.6545718432510885,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047290574259943195,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2214161679148674,
      "num_tokens": 37858689.0,
      "step": 18240
    },
    {
      "entropy": 5.506332588195801,
      "epoch": 1.6550253991291726,
      "grad_norm": 1.0,
      "learning_rate": 0.00047289034701821013,
      "loss": 5.0772,
      "mean_token_accuracy": 0.2026723563671112,
      "num_tokens": 37869049.0,
      "step": 18245
    },
    {
      "entropy": 5.449204921722412,
      "epoch": 1.6554789550072568,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047287494734456545,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2054959073662758,
      "num_tokens": 37879495.0,
      "step": 18250
    },
    {
      "entropy": 5.492386531829834,
      "epoch": 1.655932510885341,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047285954357881646,
      "loss": 5.0258,
      "mean_token_accuracy": 0.2014095425605774,
      "num_tokens": 37889442.0,
      "step": 18255
    },
    {
      "entropy": 5.556931495666504,
      "epoch": 1.6563860667634254,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004728441357212817,
      "loss": 5.1634,
      "mean_token_accuracy": 0.18957946300506592,
      "num_tokens": 37900147.0,
      "step": 18260
    },
    {
      "entropy": 5.518980598449707,
      "epoch": 1.6568396226415094,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004728287237722798,
      "loss": 4.987,
      "mean_token_accuracy": 0.1999994158744812,
      "num_tokens": 37910307.0,
      "step": 18265
    },
    {
      "entropy": 5.425207281112671,
      "epoch": 1.6572931785195935,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004728133077321295,
      "loss": 4.9517,
      "mean_token_accuracy": 0.2038370817899704,
      "num_tokens": 37920838.0,
      "step": 18270
    },
    {
      "entropy": 5.534366941452026,
      "epoch": 1.6577467343976777,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047279788760114954,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20199616104364396,
      "num_tokens": 37930717.0,
      "step": 18275
    },
    {
      "entropy": 5.535144472122193,
      "epoch": 1.658200290275762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004727824633796589,
      "loss": 5.0346,
      "mean_token_accuracy": 0.20121545791625978,
      "num_tokens": 37941254.0,
      "step": 18280
    },
    {
      "entropy": 5.4626060962677006,
      "epoch": 1.6586538461538463,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004727670350679765,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20613199472427368,
      "num_tokens": 37952132.0,
      "step": 18285
    },
    {
      "entropy": 5.4736772060394285,
      "epoch": 1.6591074020319303,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004727516026664214,
      "loss": 5.0412,
      "mean_token_accuracy": 0.2066334918141365,
      "num_tokens": 37962026.0,
      "step": 18290
    },
    {
      "entropy": 5.497957229614258,
      "epoch": 1.6595609579100146,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004727361661753128,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2055540695786476,
      "num_tokens": 37972911.0,
      "step": 18295
    },
    {
      "entropy": 5.507284355163574,
      "epoch": 1.6600145137880986,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004727207255949699,
      "loss": 4.983,
      "mean_token_accuracy": 0.20701647102832793,
      "num_tokens": 37983577.0,
      "step": 18300
    },
    {
      "entropy": 5.512321901321411,
      "epoch": 1.6604680696661829,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047270528092571197,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20108760744333268,
      "num_tokens": 37994294.0,
      "step": 18305
    },
    {
      "entropy": 5.492896032333374,
      "epoch": 1.6609216255442671,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726898321678584,
      "loss": 5.0247,
      "mean_token_accuracy": 0.20610663294792175,
      "num_tokens": 38005144.0,
      "step": 18310
    },
    {
      "entropy": 5.460435152053833,
      "epoch": 1.6613751814223512,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004726743793217288,
      "loss": 5.048,
      "mean_token_accuracy": 0.1973502218723297,
      "num_tokens": 38015277.0,
      "step": 18315
    },
    {
      "entropy": 5.4694592475891115,
      "epoch": 1.6618287373004355,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047265892238764265,
      "loss": 4.965,
      "mean_token_accuracy": 0.20670148283243178,
      "num_tokens": 38027530.0,
      "step": 18320
    },
    {
      "entropy": 5.5590980529785154,
      "epoch": 1.6622822931785195,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004726434613659196,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20439765751361846,
      "num_tokens": 38036895.0,
      "step": 18325
    },
    {
      "entropy": 5.5358936309814455,
      "epoch": 1.6627358490566038,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004726279962568794,
      "loss": 5.0965,
      "mean_token_accuracy": 0.2030717298388481,
      "num_tokens": 38047001.0,
      "step": 18330
    },
    {
      "entropy": 5.482818841934204,
      "epoch": 1.663189404934688,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004726125270608419,
      "loss": 4.9701,
      "mean_token_accuracy": 0.20969488322734833,
      "num_tokens": 38056815.0,
      "step": 18335
    },
    {
      "entropy": 5.534871244430542,
      "epoch": 1.6636429608127723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047259705377812703,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20374379605054854,
      "num_tokens": 38067139.0,
      "step": 18340
    },
    {
      "entropy": 5.5016638278961185,
      "epoch": 1.6640965166908563,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004725815764090546,
      "loss": 4.987,
      "mean_token_accuracy": 0.2108481466770172,
      "num_tokens": 38076824.0,
      "step": 18345
    },
    {
      "entropy": 5.484319543838501,
      "epoch": 1.6645500725689404,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047256609495394494,
      "loss": 5.0305,
      "mean_token_accuracy": 0.20714865922927855,
      "num_tokens": 38087517.0,
      "step": 18350
    },
    {
      "entropy": 5.517207193374634,
      "epoch": 1.6650036284470247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000472550609413118,
      "loss": 5.0212,
      "mean_token_accuracy": 0.20530910193920135,
      "num_tokens": 38098499.0,
      "step": 18355
    },
    {
      "entropy": 5.451664876937866,
      "epoch": 1.665457184325109,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004725351197868942,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2072666198015213,
      "num_tokens": 38108327.0,
      "step": 18360
    },
    {
      "entropy": 5.5305112361907955,
      "epoch": 1.6659107402031932,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047251962607559373,
      "loss": 5.0264,
      "mean_token_accuracy": 0.19983526468276977,
      "num_tokens": 38118667.0,
      "step": 18365
    },
    {
      "entropy": 5.434758472442627,
      "epoch": 1.6663642960812772,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047250412827953717,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20379660427570342,
      "num_tokens": 38128571.0,
      "step": 18370
    },
    {
      "entropy": 5.534202909469604,
      "epoch": 1.6668178519593613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004724886263990448,
      "loss": 5.1095,
      "mean_token_accuracy": 0.19703283607959748,
      "num_tokens": 38138921.0,
      "step": 18375
    },
    {
      "entropy": 5.555098199844361,
      "epoch": 1.6672714078374455,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004724731204344372,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20281906872987748,
      "num_tokens": 38148858.0,
      "step": 18380
    },
    {
      "entropy": 5.404624509811401,
      "epoch": 1.6677249637155298,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047245761038603536,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2107952728867531,
      "num_tokens": 38159107.0,
      "step": 18385
    },
    {
      "entropy": 5.535781669616699,
      "epoch": 1.668178519593614,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047244209625415975,
      "loss": 5.0863,
      "mean_token_accuracy": 0.19797467142343522,
      "num_tokens": 38169228.0,
      "step": 18390
    },
    {
      "entropy": 5.540907859802246,
      "epoch": 1.6686320754716981,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004724265780391312,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2063402235507965,
      "num_tokens": 38179175.0,
      "step": 18395
    },
    {
      "entropy": 5.48925633430481,
      "epoch": 1.6690856313497822,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004724110557412708,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20149583369493484,
      "num_tokens": 38188537.0,
      "step": 18400
    },
    {
      "entropy": 5.4690220832824705,
      "epoch": 1.6695391872278664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047239552936089943,
      "loss": 5.0554,
      "mean_token_accuracy": 0.19593087881803511,
      "num_tokens": 38198972.0,
      "step": 18405
    },
    {
      "entropy": 5.479906463623047,
      "epoch": 1.6699927431059507,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047237999889833816,
      "loss": 4.9343,
      "mean_token_accuracy": 0.20837448686361312,
      "num_tokens": 38209472.0,
      "step": 18410
    },
    {
      "entropy": 5.374909400939941,
      "epoch": 1.670446298984035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047236446435390826,
      "loss": 4.9252,
      "mean_token_accuracy": 0.2070350855588913,
      "num_tokens": 38219977.0,
      "step": 18415
    },
    {
      "entropy": 5.455624008178711,
      "epoch": 1.670899854862119,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047234892572793093,
      "loss": 5.0477,
      "mean_token_accuracy": 0.20311958193778992,
      "num_tokens": 38230492.0,
      "step": 18420
    },
    {
      "entropy": 5.491666412353515,
      "epoch": 1.671353410740203,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004723333830207275,
      "loss": 5.006,
      "mean_token_accuracy": 0.19737563282251358,
      "num_tokens": 38240799.0,
      "step": 18425
    },
    {
      "entropy": 5.4675342559814455,
      "epoch": 1.6718069666182873,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047231783623261954,
      "loss": 4.967,
      "mean_token_accuracy": 0.20395370274782182,
      "num_tokens": 38251439.0,
      "step": 18430
    },
    {
      "entropy": 5.52753529548645,
      "epoch": 1.6722605224963716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004723022853639283,
      "loss": 5.1033,
      "mean_token_accuracy": 0.2033165216445923,
      "num_tokens": 38261795.0,
      "step": 18435
    },
    {
      "entropy": 5.436008405685425,
      "epoch": 1.6727140783744558,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004722867304149756,
      "loss": 4.9758,
      "mean_token_accuracy": 0.2048972725868225,
      "num_tokens": 38272172.0,
      "step": 18440
    },
    {
      "entropy": 5.515102338790894,
      "epoch": 1.67316763425254,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004722711713860831,
      "loss": 5.0766,
      "mean_token_accuracy": 0.1952103555202484,
      "num_tokens": 38282829.0,
      "step": 18445
    },
    {
      "entropy": 5.513536834716797,
      "epoch": 1.673621190130624,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004722556082775724,
      "loss": 5.0701,
      "mean_token_accuracy": 0.20707326233386994,
      "num_tokens": 38293793.0,
      "step": 18450
    },
    {
      "entropy": 5.5105146884918215,
      "epoch": 1.6740747460087082,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004722400410897655,
      "loss": 5.0751,
      "mean_token_accuracy": 0.2046295389533043,
      "num_tokens": 38302577.0,
      "step": 18455
    },
    {
      "entropy": 5.515171194076538,
      "epoch": 1.6745283018867925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004722244698229842,
      "loss": 5.0152,
      "mean_token_accuracy": 0.20464314967393876,
      "num_tokens": 38312571.0,
      "step": 18460
    },
    {
      "entropy": 5.473025941848755,
      "epoch": 1.6749818577648767,
      "grad_norm": 1.0,
      "learning_rate": 0.00047220889447755073,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2091301903128624,
      "num_tokens": 38323544.0,
      "step": 18465
    },
    {
      "entropy": 5.518221282958985,
      "epoch": 1.6754354136429608,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004721933150537869,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20110619217157363,
      "num_tokens": 38334509.0,
      "step": 18470
    },
    {
      "entropy": 5.412545156478882,
      "epoch": 1.675888969521045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047217773155201526,
      "loss": 4.9571,
      "mean_token_accuracy": 0.203582501411438,
      "num_tokens": 38344747.0,
      "step": 18475
    },
    {
      "entropy": 5.507736158370972,
      "epoch": 1.676342525399129,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047216214397255776,
      "loss": 5.0614,
      "mean_token_accuracy": 0.19674084335565567,
      "num_tokens": 38355671.0,
      "step": 18480
    },
    {
      "entropy": 5.5069090843200685,
      "epoch": 1.6767960812772134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047214655231573693,
      "loss": 5.035,
      "mean_token_accuracy": 0.20265609323978423,
      "num_tokens": 38366368.0,
      "step": 18485
    },
    {
      "entropy": 5.499286317825318,
      "epoch": 1.6772496371552976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004721309565818752,
      "loss": 5.0981,
      "mean_token_accuracy": 0.20174516886472701,
      "num_tokens": 38377322.0,
      "step": 18490
    },
    {
      "entropy": 5.563813877105713,
      "epoch": 1.6777031930333817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047211535677129496,
      "loss": 5.0962,
      "mean_token_accuracy": 0.19940244406461716,
      "num_tokens": 38389204.0,
      "step": 18495
    },
    {
      "entropy": 5.508306884765625,
      "epoch": 1.678156748911466,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047209975288431893,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20390467196702958,
      "num_tokens": 38399117.0,
      "step": 18500
    },
    {
      "entropy": 5.4711321830749515,
      "epoch": 1.67861030478955,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047208414492126986,
      "loss": 5.0082,
      "mean_token_accuracy": 0.2012804701924324,
      "num_tokens": 38410341.0,
      "step": 18505
    },
    {
      "entropy": 5.547240686416626,
      "epoch": 1.6790638606676342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047206853288247043,
      "loss": 5.0798,
      "mean_token_accuracy": 0.1975940242409706,
      "num_tokens": 38422742.0,
      "step": 18510
    },
    {
      "entropy": 5.499547147750855,
      "epoch": 1.6795174165457185,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047205291676824345,
      "loss": 5.0729,
      "mean_token_accuracy": 0.1988615036010742,
      "num_tokens": 38433022.0,
      "step": 18515
    },
    {
      "entropy": 5.520501708984375,
      "epoch": 1.6799709724238028,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000472037296578912,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20151687860488893,
      "num_tokens": 38444325.0,
      "step": 18520
    },
    {
      "entropy": 5.492207193374634,
      "epoch": 1.6804245283018868,
      "grad_norm": 1.125,
      "learning_rate": 0.000472021672314799,
      "loss": 4.9626,
      "mean_token_accuracy": 0.20981447249650956,
      "num_tokens": 38454576.0,
      "step": 18525
    },
    {
      "entropy": 5.473684406280517,
      "epoch": 1.6808780841799709,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047200604397622763,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20934527814388276,
      "num_tokens": 38465090.0,
      "step": 18530
    },
    {
      "entropy": 5.548080396652222,
      "epoch": 1.6813316400580551,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004719904115635211,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19276446849107742,
      "num_tokens": 38477253.0,
      "step": 18535
    },
    {
      "entropy": 5.548889064788819,
      "epoch": 1.6817851959361394,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004719747750770026,
      "loss": 5.108,
      "mean_token_accuracy": 0.19910923689603804,
      "num_tokens": 38488040.0,
      "step": 18540
    },
    {
      "entropy": 5.48395676612854,
      "epoch": 1.6822387518142237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004719591345169956,
      "loss": 5.0274,
      "mean_token_accuracy": 0.20620906800031663,
      "num_tokens": 38498016.0,
      "step": 18545
    },
    {
      "entropy": 5.555508136749268,
      "epoch": 1.6826923076923077,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047194348988382345,
      "loss": 5.1085,
      "mean_token_accuracy": 0.19993061870336531,
      "num_tokens": 38508996.0,
      "step": 18550
    },
    {
      "entropy": 5.479267978668213,
      "epoch": 1.6831458635703918,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004719278411778097,
      "loss": 5.0547,
      "mean_token_accuracy": 0.2078235015273094,
      "num_tokens": 38519608.0,
      "step": 18555
    },
    {
      "entropy": 5.482069873809815,
      "epoch": 1.683599419448476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047191218839927817,
      "loss": 5.0009,
      "mean_token_accuracy": 0.2001231387257576,
      "num_tokens": 38530737.0,
      "step": 18560
    },
    {
      "entropy": 5.477081966400147,
      "epoch": 1.6840529753265603,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004718965315485523,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20080161988735198,
      "num_tokens": 38540925.0,
      "step": 18565
    },
    {
      "entropy": 5.496979665756226,
      "epoch": 1.6845065312046446,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004718808706259561,
      "loss": 5.057,
      "mean_token_accuracy": 0.19720547795295715,
      "num_tokens": 38553176.0,
      "step": 18570
    },
    {
      "entropy": 5.550701713562011,
      "epoch": 1.6849600870827286,
      "grad_norm": 1.0,
      "learning_rate": 0.00047186520563181325,
      "loss": 5.0858,
      "mean_token_accuracy": 0.1965736374258995,
      "num_tokens": 38563394.0,
      "step": 18575
    },
    {
      "entropy": 5.578579998016357,
      "epoch": 1.6854136429608126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004718495365664477,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20721545964479446,
      "num_tokens": 38573243.0,
      "step": 18580
    },
    {
      "entropy": 5.509956169128418,
      "epoch": 1.685867198838897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047183386343018364,
      "loss": 5.0474,
      "mean_token_accuracy": 0.19944120943546295,
      "num_tokens": 38584014.0,
      "step": 18585
    },
    {
      "entropy": 5.4849968433380125,
      "epoch": 1.6863207547169812,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004718181862233451,
      "loss": 4.986,
      "mean_token_accuracy": 0.20743076503276825,
      "num_tokens": 38594467.0,
      "step": 18590
    },
    {
      "entropy": 5.546944284439087,
      "epoch": 1.6867743105950654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004718025049462563,
      "loss": 5.0573,
      "mean_token_accuracy": 0.19999323934316635,
      "num_tokens": 38604407.0,
      "step": 18595
    },
    {
      "entropy": 5.553864574432373,
      "epoch": 1.6872278664731495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004717868195992416,
      "loss": 5.0564,
      "mean_token_accuracy": 0.2070892110466957,
      "num_tokens": 38614508.0,
      "step": 18600
    },
    {
      "entropy": 5.563734436035157,
      "epoch": 1.6876814223512335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004717711301826253,
      "loss": 5.0579,
      "mean_token_accuracy": 0.2006372556090355,
      "num_tokens": 38626244.0,
      "step": 18605
    },
    {
      "entropy": 5.454728603363037,
      "epoch": 1.6881349782293178,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047175543669673186,
      "loss": 5.0024,
      "mean_token_accuracy": 0.20256820917129517,
      "num_tokens": 38636694.0,
      "step": 18610
    },
    {
      "entropy": 5.588367605209351,
      "epoch": 1.688588534107402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047173973914188587,
      "loss": 5.1174,
      "mean_token_accuracy": 0.19444027543067932,
      "num_tokens": 38647414.0,
      "step": 18615
    },
    {
      "entropy": 5.50858736038208,
      "epoch": 1.6890420899854863,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047172403751841196,
      "loss": 5.0247,
      "mean_token_accuracy": 0.20138438045978546,
      "num_tokens": 38657425.0,
      "step": 18620
    },
    {
      "entropy": 5.52455883026123,
      "epoch": 1.6894956458635704,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047170833182663485,
      "loss": 5.0397,
      "mean_token_accuracy": 0.2063281774520874,
      "num_tokens": 38667751.0,
      "step": 18625
    },
    {
      "entropy": 5.431008005142212,
      "epoch": 1.6899492017416544,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004716926220668793,
      "loss": 4.9392,
      "mean_token_accuracy": 0.2107343167066574,
      "num_tokens": 38676918.0,
      "step": 18630
    },
    {
      "entropy": 5.420585203170776,
      "epoch": 1.6904027576197387,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047167690823947015,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20803968012332916,
      "num_tokens": 38685887.0,
      "step": 18635
    },
    {
      "entropy": 5.460736083984375,
      "epoch": 1.690856313497823,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004716611903447325,
      "loss": 5.0114,
      "mean_token_accuracy": 0.21264311224222182,
      "num_tokens": 38696179.0,
      "step": 18640
    },
    {
      "entropy": 5.507325601577759,
      "epoch": 1.6913098693759072,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004716454683829912,
      "loss": 5.001,
      "mean_token_accuracy": 0.20156118422746658,
      "num_tokens": 38705777.0,
      "step": 18645
    },
    {
      "entropy": 5.476040840148926,
      "epoch": 1.6917634252539913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047162974235457163,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20877761244773865,
      "num_tokens": 38717709.0,
      "step": 18650
    },
    {
      "entropy": 5.500707626342773,
      "epoch": 1.6922169811320755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004716140122597988,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20815141648054122,
      "num_tokens": 38729517.0,
      "step": 18655
    },
    {
      "entropy": 5.505517816543579,
      "epoch": 1.6926705370101596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004715982780989981,
      "loss": 5.0166,
      "mean_token_accuracy": 0.204551038146019,
      "num_tokens": 38740878.0,
      "step": 18660
    },
    {
      "entropy": 5.4452697277069095,
      "epoch": 1.6931240928882438,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047158253987249487,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2077188938856125,
      "num_tokens": 38751291.0,
      "step": 18665
    },
    {
      "entropy": 5.434791088104248,
      "epoch": 1.693577648766328,
      "grad_norm": 1.125,
      "learning_rate": 0.00047156679758061465,
      "loss": 4.962,
      "mean_token_accuracy": 0.2114056169986725,
      "num_tokens": 38760940.0,
      "step": 18670
    },
    {
      "entropy": 5.487574815750122,
      "epoch": 1.6940312046444121,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047155105122368297,
      "loss": 4.9557,
      "mean_token_accuracy": 0.20877281725406646,
      "num_tokens": 38770240.0,
      "step": 18675
    },
    {
      "entropy": 5.405946826934814,
      "epoch": 1.6944847605224964,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004715353008020255,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20899415612220765,
      "num_tokens": 38780452.0,
      "step": 18680
    },
    {
      "entropy": 5.4908122539520265,
      "epoch": 1.6949383164005805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047151954631596794,
      "loss": 5.0653,
      "mean_token_accuracy": 0.2054619684815407,
      "num_tokens": 38790763.0,
      "step": 18685
    },
    {
      "entropy": 5.4548359394073485,
      "epoch": 1.6953918722786647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047150378776583603,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2129041150212288,
      "num_tokens": 38802164.0,
      "step": 18690
    },
    {
      "entropy": 5.485100030899048,
      "epoch": 1.695845428156749,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004714880251519557,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2020278677344322,
      "num_tokens": 38812901.0,
      "step": 18695
    },
    {
      "entropy": 5.45844874382019,
      "epoch": 1.6962989840348333,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000471472258474653,
      "loss": 4.972,
      "mean_token_accuracy": 0.20479967594146728,
      "num_tokens": 38823305.0,
      "step": 18700
    },
    {
      "entropy": 5.481187152862549,
      "epoch": 1.6967525399129173,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047145648773425384,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20233571976423265,
      "num_tokens": 38833602.0,
      "step": 18705
    },
    {
      "entropy": 5.509334135055542,
      "epoch": 1.6972060957910013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047144071293108456,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20224757939577104,
      "num_tokens": 38843950.0,
      "step": 18710
    },
    {
      "entropy": 5.469134950637818,
      "epoch": 1.6976596516690856,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004714249340654712,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2077180936932564,
      "num_tokens": 38853753.0,
      "step": 18715
    },
    {
      "entropy": 5.431052494049072,
      "epoch": 1.6981132075471699,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004714091511377402,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2057625710964203,
      "num_tokens": 38863394.0,
      "step": 18720
    },
    {
      "entropy": 5.4556787490844725,
      "epoch": 1.6985667634252541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047139336414821784,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20279431343078613,
      "num_tokens": 38872972.0,
      "step": 18725
    },
    {
      "entropy": 5.4457580089569095,
      "epoch": 1.6990203193033382,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047137757309723073,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21284911036491394,
      "num_tokens": 38883251.0,
      "step": 18730
    },
    {
      "entropy": 5.4460916996002195,
      "epoch": 1.6994738751814222,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004713617779851053,
      "loss": 5.0124,
      "mean_token_accuracy": 0.2037664383649826,
      "num_tokens": 38893274.0,
      "step": 18735
    },
    {
      "entropy": 5.554559230804443,
      "epoch": 1.6999274310595065,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004713459788121683,
      "loss": 5.1013,
      "mean_token_accuracy": 0.19608234465122223,
      "num_tokens": 38903870.0,
      "step": 18740
    },
    {
      "entropy": 5.526161575317383,
      "epoch": 1.7003809869375908,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004713301755787464,
      "loss": 5.0392,
      "mean_token_accuracy": 0.203370663523674,
      "num_tokens": 38914921.0,
      "step": 18745
    },
    {
      "entropy": 5.492236137390137,
      "epoch": 1.700834542815675,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004713143682851665,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21087078750133514,
      "num_tokens": 38924630.0,
      "step": 18750
    },
    {
      "entropy": 5.429457378387451,
      "epoch": 1.701288098693759,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004712985569317553,
      "loss": 4.985,
      "mean_token_accuracy": 0.20480854213237762,
      "num_tokens": 38934926.0,
      "step": 18755
    },
    {
      "entropy": 5.541030025482177,
      "epoch": 1.7017416545718431,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004712827415188401,
      "loss": 5.1384,
      "mean_token_accuracy": 0.19464268088340758,
      "num_tokens": 38944513.0,
      "step": 18760
    },
    {
      "entropy": 5.5305335521698,
      "epoch": 1.7021952104499274,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004712669220467476,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20468886196613312,
      "num_tokens": 38955918.0,
      "step": 18765
    },
    {
      "entropy": 5.5056816101074215,
      "epoch": 1.7026487663280117,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047125109851580533,
      "loss": 5.0207,
      "mean_token_accuracy": 0.20369864255189896,
      "num_tokens": 38966221.0,
      "step": 18770
    },
    {
      "entropy": 5.468465137481689,
      "epoch": 1.703102322206096,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047123527092634025,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20372759103775023,
      "num_tokens": 38976463.0,
      "step": 18775
    },
    {
      "entropy": 5.415204572677612,
      "epoch": 1.70355587808418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004712194392786797,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2066280022263527,
      "num_tokens": 38986375.0,
      "step": 18780
    },
    {
      "entropy": 5.568876600265503,
      "epoch": 1.704009433962264,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047120360357315127,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20506335645914078,
      "num_tokens": 38996116.0,
      "step": 18785
    },
    {
      "entropy": 5.44083342552185,
      "epoch": 1.7044629898403483,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047118776381008216,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20426292568445206,
      "num_tokens": 39005846.0,
      "step": 18790
    },
    {
      "entropy": 5.50305027961731,
      "epoch": 1.7049165457184325,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047117191998980026,
      "loss": 5.0209,
      "mean_token_accuracy": 0.1990242898464203,
      "num_tokens": 39016744.0,
      "step": 18795
    },
    {
      "entropy": 5.503367137908936,
      "epoch": 1.7053701015965168,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047115607211263295,
      "loss": 5.0696,
      "mean_token_accuracy": 0.19460780322551727,
      "num_tokens": 39026725.0,
      "step": 18800
    },
    {
      "entropy": 5.5182078838348385,
      "epoch": 1.7058236574746009,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047114022017890814,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20314220190048218,
      "num_tokens": 39036851.0,
      "step": 18805
    },
    {
      "entropy": 5.460090637207031,
      "epoch": 1.706277213352685,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004711243641889536,
      "loss": 4.964,
      "mean_token_accuracy": 0.20226655155420303,
      "num_tokens": 39047179.0,
      "step": 18810
    },
    {
      "entropy": 5.423959970474243,
      "epoch": 1.7067307692307692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047110850414309714,
      "loss": 4.8583,
      "mean_token_accuracy": 0.21936832517385482,
      "num_tokens": 39057758.0,
      "step": 18815
    },
    {
      "entropy": 5.481882333755493,
      "epoch": 1.7071843251088534,
      "grad_norm": 1.125,
      "learning_rate": 0.00047109264004166696,
      "loss": 5.0756,
      "mean_token_accuracy": 0.19900205433368684,
      "num_tokens": 39068142.0,
      "step": 18820
    },
    {
      "entropy": 5.466638708114624,
      "epoch": 1.7076378809869377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004710767718849909,
      "loss": 4.9717,
      "mean_token_accuracy": 0.19948351234197617,
      "num_tokens": 39078157.0,
      "step": 18825
    },
    {
      "entropy": 5.439548063278198,
      "epoch": 1.7080914368650217,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004710608996733973,
      "loss": 5.009,
      "mean_token_accuracy": 0.20742802023887635,
      "num_tokens": 39089090.0,
      "step": 18830
    },
    {
      "entropy": 5.492271900177002,
      "epoch": 1.708544992743106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047104502340721433,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19618065804243087,
      "num_tokens": 39099079.0,
      "step": 18835
    },
    {
      "entropy": 5.467269468307495,
      "epoch": 1.70899854862119,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047102914308677033,
      "loss": 4.9891,
      "mean_token_accuracy": 0.20357683300971985,
      "num_tokens": 39109927.0,
      "step": 18840
    },
    {
      "entropy": 5.506478881835937,
      "epoch": 1.7094521044992743,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047101325871239363,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20357514470815657,
      "num_tokens": 39119806.0,
      "step": 18845
    },
    {
      "entropy": 5.539082431793213,
      "epoch": 1.7099056603773586,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004709973702844128,
      "loss": 5.093,
      "mean_token_accuracy": 0.19915326982736586,
      "num_tokens": 39130147.0,
      "step": 18850
    },
    {
      "entropy": 5.424869155883789,
      "epoch": 1.7103592162554426,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047098147780315645,
      "loss": 4.8969,
      "mean_token_accuracy": 0.21334052979946136,
      "num_tokens": 39139937.0,
      "step": 18855
    },
    {
      "entropy": 5.456284618377685,
      "epoch": 1.710812772133527,
      "grad_norm": 1.625,
      "learning_rate": 0.0004709655812689532,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2109014242887497,
      "num_tokens": 39149977.0,
      "step": 18860
    },
    {
      "entropy": 5.443518829345703,
      "epoch": 1.711266328011611,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004709496806821318,
      "loss": 5.0368,
      "mean_token_accuracy": 0.20697655975818635,
      "num_tokens": 39161076.0,
      "step": 18865
    },
    {
      "entropy": 5.534329795837403,
      "epoch": 1.7117198838896952,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047093377604302106,
      "loss": 5.0853,
      "mean_token_accuracy": 0.195909982919693,
      "num_tokens": 39170776.0,
      "step": 18870
    },
    {
      "entropy": 5.534866428375244,
      "epoch": 1.7121734397677795,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004709178673519499,
      "loss": 5.0304,
      "mean_token_accuracy": 0.1954835370182991,
      "num_tokens": 39181134.0,
      "step": 18875
    },
    {
      "entropy": 5.490071582794189,
      "epoch": 1.7126269956458637,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004709019546092473,
      "loss": 5.114,
      "mean_token_accuracy": 0.19726359695196152,
      "num_tokens": 39193331.0,
      "step": 18880
    },
    {
      "entropy": 5.5813719749450685,
      "epoch": 1.7130805515239478,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047088603781524237,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19530190974473954,
      "num_tokens": 39204372.0,
      "step": 18885
    },
    {
      "entropy": 5.506999826431274,
      "epoch": 1.7135341074020318,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004708701169702642,
      "loss": 4.9908,
      "mean_token_accuracy": 0.20710833072662355,
      "num_tokens": 39214725.0,
      "step": 18890
    },
    {
      "entropy": 5.408149671554566,
      "epoch": 1.713987663280116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004708541920746422,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20177210867404938,
      "num_tokens": 39225119.0,
      "step": 18895
    },
    {
      "entropy": 5.452188968658447,
      "epoch": 1.7144412191582004,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004708382631287055,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20001497566699983,
      "num_tokens": 39235021.0,
      "step": 18900
    },
    {
      "entropy": 5.496608829498291,
      "epoch": 1.7148947750362846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004708223301327837,
      "loss": 5.1335,
      "mean_token_accuracy": 0.19637949019670486,
      "num_tokens": 39245830.0,
      "step": 18905
    },
    {
      "entropy": 5.543840312957764,
      "epoch": 1.7153483309143687,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047080639308720617,
      "loss": 5.1015,
      "mean_token_accuracy": 0.20476004779338836,
      "num_tokens": 39255264.0,
      "step": 18910
    },
    {
      "entropy": 5.470672464370727,
      "epoch": 1.7158018867924527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047079045199230254,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20544988065958023,
      "num_tokens": 39264966.0,
      "step": 18915
    },
    {
      "entropy": 5.477617597579956,
      "epoch": 1.716255442670537,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047077450684840247,
      "loss": 5.0934,
      "mean_token_accuracy": 0.1990952104330063,
      "num_tokens": 39275082.0,
      "step": 18920
    },
    {
      "entropy": 5.565166902542114,
      "epoch": 1.7167089985486212,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004707585576558357,
      "loss": 5.1601,
      "mean_token_accuracy": 0.1952443391084671,
      "num_tokens": 39285426.0,
      "step": 18925
    },
    {
      "entropy": 5.57421875,
      "epoch": 1.7171625544267055,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004707426044149321,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19887081980705262,
      "num_tokens": 39296633.0,
      "step": 18930
    },
    {
      "entropy": 5.487073659896851,
      "epoch": 1.7176161103047896,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047072664712602153,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21383173018693924,
      "num_tokens": 39306775.0,
      "step": 18935
    },
    {
      "entropy": 5.49683313369751,
      "epoch": 1.7180696661828736,
      "grad_norm": 1.046875,
      "learning_rate": 0.000470710685789434,
      "loss": 5.0301,
      "mean_token_accuracy": 0.1975339636206627,
      "num_tokens": 39317917.0,
      "step": 18940
    },
    {
      "entropy": 5.5069968700408936,
      "epoch": 1.7185232220609579,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047069472040549965,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20173940807580948,
      "num_tokens": 39328387.0,
      "step": 18945
    },
    {
      "entropy": 5.403943729400635,
      "epoch": 1.7189767779390421,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004706787509745486,
      "loss": 4.8847,
      "mean_token_accuracy": 0.21390847563743592,
      "num_tokens": 39337639.0,
      "step": 18950
    },
    {
      "entropy": 5.490783739089966,
      "epoch": 1.7194303338171264,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047066277749691115,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20759454667568206,
      "num_tokens": 39347971.0,
      "step": 18955
    },
    {
      "entropy": 5.536470937728882,
      "epoch": 1.7198838896952104,
      "grad_norm": 1.0,
      "learning_rate": 0.0004706467999729175,
      "loss": 5.0615,
      "mean_token_accuracy": 0.19836702644824983,
      "num_tokens": 39358038.0,
      "step": 18960
    },
    {
      "entropy": 5.528147506713867,
      "epoch": 1.7203374455732945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004706308184028983,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20155510157346726,
      "num_tokens": 39367834.0,
      "step": 18965
    },
    {
      "entropy": 5.522714042663575,
      "epoch": 1.7207910014513788,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004706148327871839,
      "loss": 5.041,
      "mean_token_accuracy": 0.2054267257452011,
      "num_tokens": 39378924.0,
      "step": 18970
    },
    {
      "entropy": 5.569248008728027,
      "epoch": 1.721244557329463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047059884312610487,
      "loss": 5.1299,
      "mean_token_accuracy": 0.19488806575536727,
      "num_tokens": 39390082.0,
      "step": 18975
    },
    {
      "entropy": 5.506864213943482,
      "epoch": 1.7216981132075473,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047058284941999185,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20530381947755813,
      "num_tokens": 39400485.0,
      "step": 18980
    },
    {
      "entropy": 5.4896307468414305,
      "epoch": 1.7221516690856313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004705668516691758,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20410714745521547,
      "num_tokens": 39411601.0,
      "step": 18985
    },
    {
      "entropy": 5.483820247650146,
      "epoch": 1.7226052249637154,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047055084987398735,
      "loss": 4.9675,
      "mean_token_accuracy": 0.20924535095691682,
      "num_tokens": 39421625.0,
      "step": 18990
    },
    {
      "entropy": 5.473049306869507,
      "epoch": 1.7230587808417996,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004705348440347575,
      "loss": 5.006,
      "mean_token_accuracy": 0.20160515159368514,
      "num_tokens": 39432033.0,
      "step": 18995
    },
    {
      "entropy": 5.4737790584564205,
      "epoch": 1.723512336719884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047051883415181724,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20564049929380418,
      "num_tokens": 39442514.0,
      "step": 19000
    },
    {
      "entropy": 5.48325548171997,
      "epoch": 1.7239658925979682,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004705028202254977,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21157418042421341,
      "num_tokens": 39452497.0,
      "step": 19005
    },
    {
      "entropy": 5.4872393131256105,
      "epoch": 1.7244194484760522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047048680225612996,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20488055944442748,
      "num_tokens": 39462173.0,
      "step": 19010
    },
    {
      "entropy": 5.5290426254272464,
      "epoch": 1.7248730043541363,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004704707802440454,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20708537846803665,
      "num_tokens": 39473123.0,
      "step": 19015
    },
    {
      "entropy": 5.503283500671387,
      "epoch": 1.7253265602322205,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047045475418957524,
      "loss": 5.0346,
      "mean_token_accuracy": 0.19464755058288574,
      "num_tokens": 39484142.0,
      "step": 19020
    },
    {
      "entropy": 5.49665675163269,
      "epoch": 1.7257801161103048,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004704387240930509,
      "loss": 5.0475,
      "mean_token_accuracy": 0.1961239457130432,
      "num_tokens": 39494926.0,
      "step": 19025
    },
    {
      "entropy": 5.444309186935425,
      "epoch": 1.726233671988389,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004704226899548041,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20878380239009858,
      "num_tokens": 39505153.0,
      "step": 19030
    },
    {
      "entropy": 5.456864356994629,
      "epoch": 1.726687227866473,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047040665177516616,
      "loss": 4.9609,
      "mean_token_accuracy": 0.20865223854780196,
      "num_tokens": 39515905.0,
      "step": 19035
    },
    {
      "entropy": 5.4507979393005375,
      "epoch": 1.7271407837445574,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047039060955446886,
      "loss": 4.9777,
      "mean_token_accuracy": 0.20940704494714737,
      "num_tokens": 39526762.0,
      "step": 19040
    },
    {
      "entropy": 5.45876727104187,
      "epoch": 1.7275943396226414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000470374563293044,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2033142626285553,
      "num_tokens": 39537340.0,
      "step": 19045
    },
    {
      "entropy": 5.497812652587891,
      "epoch": 1.7280478955007257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004703585129912234,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20277515202760696,
      "num_tokens": 39547144.0,
      "step": 19050
    },
    {
      "entropy": 5.5948323726654055,
      "epoch": 1.72850145137881,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047034245864933894,
      "loss": 5.1148,
      "mean_token_accuracy": 0.19469091147184373,
      "num_tokens": 39557707.0,
      "step": 19055
    },
    {
      "entropy": 5.551272392272949,
      "epoch": 1.728955007256894,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047032640026772274,
      "loss": 5.0938,
      "mean_token_accuracy": 0.1990850269794464,
      "num_tokens": 39567629.0,
      "step": 19060
    },
    {
      "entropy": 5.494695329666138,
      "epoch": 1.7294085631349783,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004703103378467067,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2044454872608185,
      "num_tokens": 39577918.0,
      "step": 19065
    },
    {
      "entropy": 5.516372489929199,
      "epoch": 1.7298621190130623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047029427138662314,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21252963989973067,
      "num_tokens": 39587838.0,
      "step": 19070
    },
    {
      "entropy": 5.459414339065551,
      "epoch": 1.7303156748911466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004702782008878042,
      "loss": 4.9262,
      "mean_token_accuracy": 0.20903893411159516,
      "num_tokens": 39597353.0,
      "step": 19075
    },
    {
      "entropy": 5.454975748062134,
      "epoch": 1.7307692307692308,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004702621263505824,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20501814484596254,
      "num_tokens": 39607349.0,
      "step": 19080
    },
    {
      "entropy": 5.472273063659668,
      "epoch": 1.731222786647315,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004702460477752901,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20289291739463805,
      "num_tokens": 39617810.0,
      "step": 19085
    },
    {
      "entropy": 5.476509857177734,
      "epoch": 1.7316763425253991,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004702299651622597,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21759312748908996,
      "num_tokens": 39626805.0,
      "step": 19090
    },
    {
      "entropy": 5.442428302764893,
      "epoch": 1.7321298984034832,
      "grad_norm": 1.0,
      "learning_rate": 0.0004702138785118239,
      "loss": 4.9584,
      "mean_token_accuracy": 0.21143175065517425,
      "num_tokens": 39637839.0,
      "step": 19095
    },
    {
      "entropy": 5.429023170471192,
      "epoch": 1.7325834542815675,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004701977878243153,
      "loss": 4.953,
      "mean_token_accuracy": 0.20404670387506485,
      "num_tokens": 39648697.0,
      "step": 19100
    },
    {
      "entropy": 5.453158378601074,
      "epoch": 1.7330370101596517,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004701816931000668,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2053990915417671,
      "num_tokens": 39658154.0,
      "step": 19105
    },
    {
      "entropy": 5.4667986869812015,
      "epoch": 1.733490566037736,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047016559433941106,
      "loss": 4.9967,
      "mean_token_accuracy": 0.2058219015598297,
      "num_tokens": 39667939.0,
      "step": 19110
    },
    {
      "entropy": 5.437241506576538,
      "epoch": 1.73394412191582,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004701494915426812,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2034673810005188,
      "num_tokens": 39678289.0,
      "step": 19115
    },
    {
      "entropy": 5.50600380897522,
      "epoch": 1.734397677793904,
      "grad_norm": 1.125,
      "learning_rate": 0.00047013338471021,
      "loss": 4.9929,
      "mean_token_accuracy": 0.2037924885749817,
      "num_tokens": 39689901.0,
      "step": 19120
    },
    {
      "entropy": 5.5021851539611815,
      "epoch": 1.7348512336719883,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047011727384233074,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2013510659337044,
      "num_tokens": 39700831.0,
      "step": 19125
    },
    {
      "entropy": 5.5257683277130125,
      "epoch": 1.7353047895500726,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004701011589393765,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20323415845632553,
      "num_tokens": 39711794.0,
      "step": 19130
    },
    {
      "entropy": 5.475721979141236,
      "epoch": 1.7357583454281569,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047008504000168054,
      "loss": 4.9858,
      "mean_token_accuracy": 0.2066682130098343,
      "num_tokens": 39721889.0,
      "step": 19135
    },
    {
      "entropy": 5.442572212219238,
      "epoch": 1.736211901306241,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004700689170295763,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20893426686525346,
      "num_tokens": 39732562.0,
      "step": 19140
    },
    {
      "entropy": 5.567757034301758,
      "epoch": 1.736665457184325,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047005279002339717,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2066911593079567,
      "num_tokens": 39742286.0,
      "step": 19145
    },
    {
      "entropy": 5.522202491760254,
      "epoch": 1.7371190130624092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047003665898347663,
      "loss": 5.0416,
      "mean_token_accuracy": 0.20541482120752336,
      "num_tokens": 39754336.0,
      "step": 19150
    },
    {
      "entropy": 5.449558210372925,
      "epoch": 1.7375725689404935,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004700205239101483,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21176816672086715,
      "num_tokens": 39763545.0,
      "step": 19155
    },
    {
      "entropy": 5.518767642974853,
      "epoch": 1.7380261248185778,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004700043848037458,
      "loss": 5.083,
      "mean_token_accuracy": 0.19528307467699052,
      "num_tokens": 39774536.0,
      "step": 19160
    },
    {
      "entropy": 5.4985637187957765,
      "epoch": 1.7384796806966618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004699882416646029,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20992634445428848,
      "num_tokens": 39785377.0,
      "step": 19165
    },
    {
      "entropy": 5.498961877822876,
      "epoch": 1.7389332365747459,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004699720944930535,
      "loss": 5.1028,
      "mean_token_accuracy": 0.18861977607011796,
      "num_tokens": 39796692.0,
      "step": 19170
    },
    {
      "entropy": 5.429786586761475,
      "epoch": 1.7393867924528301,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046995594328943147,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20851065516471862,
      "num_tokens": 39807927.0,
      "step": 19175
    },
    {
      "entropy": 5.600195407867432,
      "epoch": 1.7398403483309144,
      "grad_norm": 1.125,
      "learning_rate": 0.0004699397880540709,
      "loss": 5.1774,
      "mean_token_accuracy": 0.19841932952404023,
      "num_tokens": 39817617.0,
      "step": 19180
    },
    {
      "entropy": 5.512768173217774,
      "epoch": 1.7402939042089987,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004699236287873058,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2104815885424614,
      "num_tokens": 39827092.0,
      "step": 19185
    },
    {
      "entropy": 5.498181104660034,
      "epoch": 1.7407474600870827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004699074654894705,
      "loss": 4.966,
      "mean_token_accuracy": 0.20659948736429215,
      "num_tokens": 39837378.0,
      "step": 19190
    },
    {
      "entropy": 5.481163215637207,
      "epoch": 1.7412010159651667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000469891298160899,
      "loss": 4.972,
      "mean_token_accuracy": 0.20108917057514192,
      "num_tokens": 39848191.0,
      "step": 19195
    },
    {
      "entropy": 5.426167154312134,
      "epoch": 1.741654571843251,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046987512680192586,
      "loss": 4.9915,
      "mean_token_accuracy": 0.2045610874891281,
      "num_tokens": 39857886.0,
      "step": 19200
    },
    {
      "entropy": 5.5520631790161135,
      "epoch": 1.7421081277213353,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046985895141288546,
      "loss": 5.0735,
      "mean_token_accuracy": 0.19812504202127457,
      "num_tokens": 39868348.0,
      "step": 19205
    },
    {
      "entropy": 5.5643822193145756,
      "epoch": 1.7425616835994195,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046984277199411224,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20363999903202057,
      "num_tokens": 39879129.0,
      "step": 19210
    },
    {
      "entropy": 5.445794534683228,
      "epoch": 1.7430152394775036,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046982658854594083,
      "loss": 4.9688,
      "mean_token_accuracy": 0.20444275438785553,
      "num_tokens": 39890239.0,
      "step": 19215
    },
    {
      "entropy": 5.40272331237793,
      "epoch": 1.7434687953555879,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046981040106870595,
      "loss": 4.895,
      "mean_token_accuracy": 0.2079245164990425,
      "num_tokens": 39900907.0,
      "step": 19220
    },
    {
      "entropy": 5.508737373352051,
      "epoch": 1.743922351233672,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004697942095627423,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20607868880033492,
      "num_tokens": 39911564.0,
      "step": 19225
    },
    {
      "entropy": 5.492405939102173,
      "epoch": 1.7443759071117562,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004697780140283848,
      "loss": 4.9498,
      "mean_token_accuracy": 0.2060716688632965,
      "num_tokens": 39922168.0,
      "step": 19230
    },
    {
      "entropy": 5.473089551925659,
      "epoch": 1.7448294629898404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004697618144659683,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20918959975242615,
      "num_tokens": 39932690.0,
      "step": 19235
    },
    {
      "entropy": 5.578668594360352,
      "epoch": 1.7452830188679245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004697456108758278,
      "loss": 5.1499,
      "mean_token_accuracy": 0.19803129583597184,
      "num_tokens": 39942970.0,
      "step": 19240
    },
    {
      "entropy": 5.495582628250122,
      "epoch": 1.7457365747460087,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004697294032582986,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20389143228530884,
      "num_tokens": 39953255.0,
      "step": 19245
    },
    {
      "entropy": 5.463984870910645,
      "epoch": 1.7461901306240928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046971319161371564,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21246978640556335,
      "num_tokens": 39963521.0,
      "step": 19250
    },
    {
      "entropy": 5.493888282775879,
      "epoch": 1.746643686502177,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046969697594241426,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20456201434135438,
      "num_tokens": 39973494.0,
      "step": 19255
    },
    {
      "entropy": 5.478393030166626,
      "epoch": 1.7470972423802613,
      "grad_norm": 1.0,
      "learning_rate": 0.0004696807562447298,
      "loss": 5.1048,
      "mean_token_accuracy": 0.19585141241550447,
      "num_tokens": 39984652.0,
      "step": 19260
    },
    {
      "entropy": 5.50634355545044,
      "epoch": 1.7475507982583456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046966453252099763,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20787999927997589,
      "num_tokens": 39995490.0,
      "step": 19265
    },
    {
      "entropy": 5.401324892044068,
      "epoch": 1.7480043541364296,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046964830477155335,
      "loss": 4.8095,
      "mean_token_accuracy": 0.22924857586622238,
      "num_tokens": 40006447.0,
      "step": 19270
    },
    {
      "entropy": 5.537764167785644,
      "epoch": 1.7484579100145137,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046963207299673257,
      "loss": 5.0852,
      "mean_token_accuracy": 0.1998353958129883,
      "num_tokens": 40015733.0,
      "step": 19275
    },
    {
      "entropy": 5.560251331329345,
      "epoch": 1.748911465892598,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046961583719687085,
      "loss": 4.9493,
      "mean_token_accuracy": 0.20954487472772598,
      "num_tokens": 40025844.0,
      "step": 19280
    },
    {
      "entropy": 5.46172685623169,
      "epoch": 1.7493650217706822,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046959959737230414,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21447892785072326,
      "num_tokens": 40035942.0,
      "step": 19285
    },
    {
      "entropy": 5.392023992538452,
      "epoch": 1.7498185776487665,
      "grad_norm": 1.125,
      "learning_rate": 0.00046958335352336806,
      "loss": 4.8873,
      "mean_token_accuracy": 0.21472759693861007,
      "num_tokens": 40045551.0,
      "step": 19290
    },
    {
      "entropy": 5.505699729919433,
      "epoch": 1.7502721335268505,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004695671056503987,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20877962857484816,
      "num_tokens": 40055803.0,
      "step": 19295
    },
    {
      "entropy": 5.466961812973023,
      "epoch": 1.7507256894049346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000469550853753732,
      "loss": 5.0047,
      "mean_token_accuracy": 0.1995977506041527,
      "num_tokens": 40066004.0,
      "step": 19300
    },
    {
      "entropy": 5.479906797409058,
      "epoch": 1.7511792452830188,
      "grad_norm": 1.109375,
      "learning_rate": 0.000469534597833704,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21072949320077897,
      "num_tokens": 40075522.0,
      "step": 19305
    },
    {
      "entropy": 5.466987228393554,
      "epoch": 1.751632801161103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000469518337890651,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21147502809762955,
      "num_tokens": 40084839.0,
      "step": 19310
    },
    {
      "entropy": 5.492431545257569,
      "epoch": 1.7520863570391874,
      "grad_norm": 1.0,
      "learning_rate": 0.00046950207392490917,
      "loss": 5.0308,
      "mean_token_accuracy": 0.2053804352879524,
      "num_tokens": 40094392.0,
      "step": 19315
    },
    {
      "entropy": 5.445017576217651,
      "epoch": 1.7525399129172714,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046948580593681496,
      "loss": 4.9614,
      "mean_token_accuracy": 0.20911876112222672,
      "num_tokens": 40105990.0,
      "step": 19320
    },
    {
      "entropy": 5.430374050140381,
      "epoch": 1.7529934687953554,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004694695339267047,
      "loss": 4.8983,
      "mean_token_accuracy": 0.21301756352186202,
      "num_tokens": 40116143.0,
      "step": 19325
    },
    {
      "entropy": 5.574617528915406,
      "epoch": 1.7534470246734397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046945325789491483,
      "loss": 5.0767,
      "mean_token_accuracy": 0.20233528912067414,
      "num_tokens": 40126849.0,
      "step": 19330
    },
    {
      "entropy": 5.478389644622803,
      "epoch": 1.753900580551524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004694369778417822,
      "loss": 5.0199,
      "mean_token_accuracy": 0.2064966917037964,
      "num_tokens": 40137259.0,
      "step": 19335
    },
    {
      "entropy": 5.394656181335449,
      "epoch": 1.7543541364296082,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004694206937676431,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21250836700201034,
      "num_tokens": 40148578.0,
      "step": 19340
    },
    {
      "entropy": 5.519765138626099,
      "epoch": 1.7548076923076923,
      "grad_norm": 1.0,
      "learning_rate": 0.0004694044056728347,
      "loss": 5.0723,
      "mean_token_accuracy": 0.19564414769411087,
      "num_tokens": 40159754.0,
      "step": 19345
    },
    {
      "entropy": 5.521185493469238,
      "epoch": 1.7552612481857763,
      "grad_norm": 1.125,
      "learning_rate": 0.00046938811355769363,
      "loss": 5.005,
      "mean_token_accuracy": 0.20478369146585465,
      "num_tokens": 40169388.0,
      "step": 19350
    },
    {
      "entropy": 5.477996444702148,
      "epoch": 1.7557148040638606,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046937181742255673,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2044983148574829,
      "num_tokens": 40179271.0,
      "step": 19355
    },
    {
      "entropy": 5.5135838985443115,
      "epoch": 1.7561683599419449,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004693555172677612,
      "loss": 5.0557,
      "mean_token_accuracy": 0.2031428873538971,
      "num_tokens": 40189292.0,
      "step": 19360
    },
    {
      "entropy": 5.485254383087158,
      "epoch": 1.7566219158200291,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000469339213093644,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20674334615468978,
      "num_tokens": 40199271.0,
      "step": 19365
    },
    {
      "entropy": 5.485046863555908,
      "epoch": 1.7570754716981132,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046932290490054235,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20664295554161072,
      "num_tokens": 40210386.0,
      "step": 19370
    },
    {
      "entropy": 5.514654541015625,
      "epoch": 1.7575290275761972,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004693065926887936,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20283467024564744,
      "num_tokens": 40220892.0,
      "step": 19375
    },
    {
      "entropy": 5.534471464157105,
      "epoch": 1.7579825834542815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046929027645873484,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2017950177192688,
      "num_tokens": 40231737.0,
      "step": 19380
    },
    {
      "entropy": 5.502135562896728,
      "epoch": 1.7584361393323658,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004692739562107038,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20254693329334258,
      "num_tokens": 40242218.0,
      "step": 19385
    },
    {
      "entropy": 5.468260288238525,
      "epoch": 1.75888969521045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046925763194503775,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20756390541791916,
      "num_tokens": 40253659.0,
      "step": 19390
    },
    {
      "entropy": 5.397907686233521,
      "epoch": 1.759343251088534,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004692413036620744,
      "loss": 4.8436,
      "mean_token_accuracy": 0.21793943494558335,
      "num_tokens": 40264328.0,
      "step": 19395
    },
    {
      "entropy": 5.451465272903443,
      "epoch": 1.7597968069666183,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004692249713621514,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20580762326717378,
      "num_tokens": 40275188.0,
      "step": 19400
    },
    {
      "entropy": 5.558916187286377,
      "epoch": 1.7602503628447024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046920863504560647,
      "loss": 5.1234,
      "mean_token_accuracy": 0.19936256259679794,
      "num_tokens": 40287094.0,
      "step": 19405
    },
    {
      "entropy": 5.499798583984375,
      "epoch": 1.7607039187227866,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004691922947127775,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20331257134675979,
      "num_tokens": 40296551.0,
      "step": 19410
    },
    {
      "entropy": 5.467429494857788,
      "epoch": 1.761157474600871,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004691759503640024,
      "loss": 4.9796,
      "mean_token_accuracy": 0.2025754451751709,
      "num_tokens": 40306920.0,
      "step": 19415
    },
    {
      "entropy": 5.441870784759521,
      "epoch": 1.761611030478955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004691596019996191,
      "loss": 4.8822,
      "mean_token_accuracy": 0.21416640728712083,
      "num_tokens": 40316797.0,
      "step": 19420
    },
    {
      "entropy": 5.396556568145752,
      "epoch": 1.7620645863570392,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004691432496199658,
      "loss": 4.9538,
      "mean_token_accuracy": 0.20391108095645905,
      "num_tokens": 40327002.0,
      "step": 19425
    },
    {
      "entropy": 5.573301506042481,
      "epoch": 1.7625181422351233,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046912689322538064,
      "loss": 5.0905,
      "mean_token_accuracy": 0.2061540275812149,
      "num_tokens": 40337247.0,
      "step": 19430
    },
    {
      "entropy": 5.5015565872192385,
      "epoch": 1.7629716981132075,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046911053281620187,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20459424555301667,
      "num_tokens": 40346960.0,
      "step": 19435
    },
    {
      "entropy": 5.539232969284058,
      "epoch": 1.7634252539912918,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046909416839276775,
      "loss": 5.0489,
      "mean_token_accuracy": 0.2052585870027542,
      "num_tokens": 40357169.0,
      "step": 19440
    },
    {
      "entropy": 5.4448387145996096,
      "epoch": 1.763878809869376,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004690777999554168,
      "loss": 4.8883,
      "mean_token_accuracy": 0.21226165145635606,
      "num_tokens": 40367660.0,
      "step": 19445
    },
    {
      "entropy": 5.447745370864868,
      "epoch": 1.76433236574746,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004690614275044875,
      "loss": 4.9031,
      "mean_token_accuracy": 0.20838938057422637,
      "num_tokens": 40378217.0,
      "step": 19450
    },
    {
      "entropy": 5.483260631561279,
      "epoch": 1.7647859216255442,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004690450510403184,
      "loss": 5.002,
      "mean_token_accuracy": 0.20388685613870622,
      "num_tokens": 40388249.0,
      "step": 19455
    },
    {
      "entropy": 5.528651094436645,
      "epoch": 1.7652394775036284,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004690286705632483,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20168517678976058,
      "num_tokens": 40398239.0,
      "step": 19460
    },
    {
      "entropy": 5.510349464416504,
      "epoch": 1.7656930333817127,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004690122860736157,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20132900178432464,
      "num_tokens": 40408473.0,
      "step": 19465
    },
    {
      "entropy": 5.616847658157349,
      "epoch": 1.766146589259797,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046899589757175964,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19569293558597564,
      "num_tokens": 40418674.0,
      "step": 19470
    },
    {
      "entropy": 5.47779483795166,
      "epoch": 1.766600145137881,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000468979505058019,
      "loss": 4.9657,
      "mean_token_accuracy": 0.20367812365293503,
      "num_tokens": 40429815.0,
      "step": 19475
    },
    {
      "entropy": 5.557933712005616,
      "epoch": 1.767053701015965,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046896310853273274,
      "loss": 5.083,
      "mean_token_accuracy": 0.19921138435602187,
      "num_tokens": 40439809.0,
      "step": 19480
    },
    {
      "entropy": 5.580274629592895,
      "epoch": 1.7675072568940493,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046894670799624,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20024756342172623,
      "num_tokens": 40449503.0,
      "step": 19485
    },
    {
      "entropy": 5.49308066368103,
      "epoch": 1.7679608127721336,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046893030344887995,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19226822406053543,
      "num_tokens": 40459794.0,
      "step": 19490
    },
    {
      "entropy": 5.464216518402099,
      "epoch": 1.7684143686502178,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046891389489099166,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21299748867750168,
      "num_tokens": 40470044.0,
      "step": 19495
    },
    {
      "entropy": 5.464532041549683,
      "epoch": 1.7688679245283019,
      "grad_norm": 1.0,
      "learning_rate": 0.00046889748232291476,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20078731328248978,
      "num_tokens": 40480709.0,
      "step": 19500
    },
    {
      "entropy": 5.430858850479126,
      "epoch": 1.769321480406386,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046888106574498843,
      "loss": 4.8269,
      "mean_token_accuracy": 0.21487310528755188,
      "num_tokens": 40490906.0,
      "step": 19505
    },
    {
      "entropy": 5.425414514541626,
      "epoch": 1.7697750362844702,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004688646451575522,
      "loss": 4.942,
      "mean_token_accuracy": 0.21000200062990187,
      "num_tokens": 40501971.0,
      "step": 19510
    },
    {
      "entropy": 5.456477117538452,
      "epoch": 1.7702285921625545,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046884822056094587,
      "loss": 5.0428,
      "mean_token_accuracy": 0.19533804655075074,
      "num_tokens": 40512685.0,
      "step": 19515
    },
    {
      "entropy": 5.475785541534424,
      "epoch": 1.7706821480406387,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046883179195550883,
      "loss": 4.9737,
      "mean_token_accuracy": 0.20894474387168885,
      "num_tokens": 40523460.0,
      "step": 19520
    },
    {
      "entropy": 5.521895551681519,
      "epoch": 1.7711357039187228,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046881535934158094,
      "loss": 4.964,
      "mean_token_accuracy": 0.2081453248858452,
      "num_tokens": 40533594.0,
      "step": 19525
    },
    {
      "entropy": 5.472539567947388,
      "epoch": 1.7715892597968068,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004687989227195021,
      "loss": 4.9983,
      "mean_token_accuracy": 0.2078337699174881,
      "num_tokens": 40543504.0,
      "step": 19530
    },
    {
      "entropy": 5.493492698669433,
      "epoch": 1.772042815674891,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046878248208961204,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20091839283704757,
      "num_tokens": 40553711.0,
      "step": 19535
    },
    {
      "entropy": 5.486490678787232,
      "epoch": 1.7724963715529753,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000468766037452251,
      "loss": 5.0194,
      "mean_token_accuracy": 0.20321821868419648,
      "num_tokens": 40563731.0,
      "step": 19540
    },
    {
      "entropy": 5.515470361709594,
      "epoch": 1.7729499274310596,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046874958880775886,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20378756523132324,
      "num_tokens": 40574235.0,
      "step": 19545
    },
    {
      "entropy": 5.476559495925903,
      "epoch": 1.7734034833091437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004687331361564758,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21053395420312881,
      "num_tokens": 40584459.0,
      "step": 19550
    },
    {
      "entropy": 5.4867815494537355,
      "epoch": 1.7738570391872277,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004687166794987422,
      "loss": 4.9979,
      "mean_token_accuracy": 0.20661632269620894,
      "num_tokens": 40594268.0,
      "step": 19555
    },
    {
      "entropy": 5.46446123123169,
      "epoch": 1.774310595065312,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004687002188348983,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21058797538280488,
      "num_tokens": 40604673.0,
      "step": 19560
    },
    {
      "entropy": 5.406919240951538,
      "epoch": 1.7747641509433962,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046868375416528443,
      "loss": 4.9581,
      "mean_token_accuracy": 0.20515688955783845,
      "num_tokens": 40614998.0,
      "step": 19565
    },
    {
      "entropy": 5.477487850189209,
      "epoch": 1.7752177068214805,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046866728549024123,
      "loss": 5.0315,
      "mean_token_accuracy": 0.20351595878601075,
      "num_tokens": 40624729.0,
      "step": 19570
    },
    {
      "entropy": 5.4780138492584225,
      "epoch": 1.7756712626995645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004686508128101092,
      "loss": 5.0915,
      "mean_token_accuracy": 0.2033676564693451,
      "num_tokens": 40635673.0,
      "step": 19575
    },
    {
      "entropy": 5.537872552871704,
      "epoch": 1.7761248185776488,
      "grad_norm": 1.015625,
      "learning_rate": 0.000468634336125229,
      "loss": 5.0543,
      "mean_token_accuracy": 0.19503015875816346,
      "num_tokens": 40645990.0,
      "step": 19580
    },
    {
      "entropy": 5.535602378845215,
      "epoch": 1.7765783744557329,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046861785543594143,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2150138199329376,
      "num_tokens": 40656339.0,
      "step": 19585
    },
    {
      "entropy": 5.473038530349731,
      "epoch": 1.7770319303338171,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004686013707425873,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2127785563468933,
      "num_tokens": 40666371.0,
      "step": 19590
    },
    {
      "entropy": 5.507061529159546,
      "epoch": 1.7774854862119014,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046858488204550735,
      "loss": 4.9721,
      "mean_token_accuracy": 0.20457061380147934,
      "num_tokens": 40676423.0,
      "step": 19595
    },
    {
      "entropy": 5.474527072906494,
      "epoch": 1.7779390420899854,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004685683893450428,
      "loss": 4.9323,
      "mean_token_accuracy": 0.21180423796176912,
      "num_tokens": 40686331.0,
      "step": 19600
    },
    {
      "entropy": 5.409183311462402,
      "epoch": 1.7783925979680697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004685518926415346,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21339729726314544,
      "num_tokens": 40697408.0,
      "step": 19605
    },
    {
      "entropy": 5.38602614402771,
      "epoch": 1.7788461538461537,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000468535391935324,
      "loss": 4.8721,
      "mean_token_accuracy": 0.21699881553649902,
      "num_tokens": 40707078.0,
      "step": 19610
    },
    {
      "entropy": 5.482337093353271,
      "epoch": 1.779299709724238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685188872267521,
      "loss": 4.9779,
      "mean_token_accuracy": 0.20937298089265824,
      "num_tokens": 40718116.0,
      "step": 19615
    },
    {
      "entropy": 5.490728092193604,
      "epoch": 1.7797532656023223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004685023785161603,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20560814142227174,
      "num_tokens": 40727737.0,
      "step": 19620
    },
    {
      "entropy": 5.514803743362426,
      "epoch": 1.7802068214804065,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004684858658038901,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20210180282592774,
      "num_tokens": 40737455.0,
      "step": 19625
    },
    {
      "entropy": 5.570596647262573,
      "epoch": 1.7806603773584906,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004684693490902828,
      "loss": 5.0073,
      "mean_token_accuracy": 0.20343903005123137,
      "num_tokens": 40746715.0,
      "step": 19630
    },
    {
      "entropy": 5.435371255874633,
      "epoch": 1.7811139332365746,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046845282837568,
      "loss": 4.8913,
      "mean_token_accuracy": 0.21110435873270034,
      "num_tokens": 40756432.0,
      "step": 19635
    },
    {
      "entropy": 5.3773823261260985,
      "epoch": 1.781567489114659,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046843630366042355,
      "loss": 4.9445,
      "mean_token_accuracy": 0.20851500928401948,
      "num_tokens": 40766629.0,
      "step": 19640
    },
    {
      "entropy": 5.5339432716369625,
      "epoch": 1.7820210449927432,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046841977494485506,
      "loss": 5.0804,
      "mean_token_accuracy": 0.19790629148483277,
      "num_tokens": 40776650.0,
      "step": 19645
    },
    {
      "entropy": 5.5536092758178714,
      "epoch": 1.7824746008708274,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004684032422293162,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2108376607298851,
      "num_tokens": 40786722.0,
      "step": 19650
    },
    {
      "entropy": 5.382116651535034,
      "epoch": 1.7829281567489115,
      "grad_norm": 1.015625,
      "learning_rate": 0.000468386705514149,
      "loss": 4.8357,
      "mean_token_accuracy": 0.22145829796791078,
      "num_tokens": 40797058.0,
      "step": 19655
    },
    {
      "entropy": 5.413093757629395,
      "epoch": 1.7833817126269955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004683701647996956,
      "loss": 5.0216,
      "mean_token_accuracy": 0.20678603053092956,
      "num_tokens": 40808207.0,
      "step": 19660
    },
    {
      "entropy": 5.547549057006836,
      "epoch": 1.7838352685050798,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004683536200862978,
      "loss": 5.0834,
      "mean_token_accuracy": 0.20411568135023117,
      "num_tokens": 40819655.0,
      "step": 19665
    },
    {
      "entropy": 5.5205058574676515,
      "epoch": 1.784288824383164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004683370713742979,
      "loss": 5.0702,
      "mean_token_accuracy": 0.202297905087471,
      "num_tokens": 40831015.0,
      "step": 19670
    },
    {
      "entropy": 5.456383419036865,
      "epoch": 1.7847423802612483,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046832051866403814,
      "loss": 4.9442,
      "mean_token_accuracy": 0.2121114194393158,
      "num_tokens": 40841763.0,
      "step": 19675
    },
    {
      "entropy": 5.440851402282715,
      "epoch": 1.7851959361393324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004683039619558607,
      "loss": 4.978,
      "mean_token_accuracy": 0.2044297695159912,
      "num_tokens": 40852479.0,
      "step": 19680
    },
    {
      "entropy": 5.504888010025025,
      "epoch": 1.7856494920174164,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046828740125010813,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20793140083551406,
      "num_tokens": 40862486.0,
      "step": 19685
    },
    {
      "entropy": 5.558384370803833,
      "epoch": 1.7861030478955007,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004682708365471228,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20163077712059022,
      "num_tokens": 40872182.0,
      "step": 19690
    },
    {
      "entropy": 5.506349802017212,
      "epoch": 1.786556603773585,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046825426784724734,
      "loss": 5.0457,
      "mean_token_accuracy": 0.20396464169025422,
      "num_tokens": 40883036.0,
      "step": 19695
    },
    {
      "entropy": 5.561575031280517,
      "epoch": 1.7870101596516692,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004682376951508244,
      "loss": 5.091,
      "mean_token_accuracy": 0.19055148363113403,
      "num_tokens": 40893684.0,
      "step": 19700
    },
    {
      "entropy": 5.541323900222778,
      "epoch": 1.7874637155297532,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046822111845819665,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20494856238365172,
      "num_tokens": 40904540.0,
      "step": 19705
    },
    {
      "entropy": 5.537597799301148,
      "epoch": 1.7879172714078373,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046820453776970697,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20565734505653382,
      "num_tokens": 40915269.0,
      "step": 19710
    },
    {
      "entropy": 5.5129152774810795,
      "epoch": 1.7883708272859216,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046818795308569815,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20370717644691466,
      "num_tokens": 40925987.0,
      "step": 19715
    },
    {
      "entropy": 5.433760643005371,
      "epoch": 1.7888243831640058,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004681713644065133,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21342145949602126,
      "num_tokens": 40936130.0,
      "step": 19720
    },
    {
      "entropy": 5.4213385581970215,
      "epoch": 1.78927793904209,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004681547717324954,
      "loss": 4.8764,
      "mean_token_accuracy": 0.21682085692882538,
      "num_tokens": 40945476.0,
      "step": 19725
    },
    {
      "entropy": 5.392783832550049,
      "epoch": 1.7897314949201741,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046813817506398755,
      "loss": 4.8584,
      "mean_token_accuracy": 0.21738901883363723,
      "num_tokens": 40956492.0,
      "step": 19730
    },
    {
      "entropy": 5.5081888198852536,
      "epoch": 1.7901850507982582,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046812157440133303,
      "loss": 5.0794,
      "mean_token_accuracy": 0.20247095227241516,
      "num_tokens": 40966990.0,
      "step": 19735
    },
    {
      "entropy": 5.509311485290527,
      "epoch": 1.7906386066763424,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004681049697448752,
      "loss": 4.9557,
      "mean_token_accuracy": 0.20979665368795394,
      "num_tokens": 40977065.0,
      "step": 19740
    },
    {
      "entropy": 5.445501327514648,
      "epoch": 1.7910921625544267,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046808836109495734,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20353686213493347,
      "num_tokens": 40987207.0,
      "step": 19745
    },
    {
      "entropy": 5.463134384155273,
      "epoch": 1.791545718432511,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004680717484519229,
      "loss": 4.9515,
      "mean_token_accuracy": 0.20822951644659043,
      "num_tokens": 40997541.0,
      "step": 19750
    },
    {
      "entropy": 5.452787780761719,
      "epoch": 1.791999274310595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004680551318161156,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2096974954009056,
      "num_tokens": 41009044.0,
      "step": 19755
    },
    {
      "entropy": 5.4872293949127195,
      "epoch": 1.7924528301886793,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046803851118787897,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21013010740280152,
      "num_tokens": 41018165.0,
      "step": 19760
    },
    {
      "entropy": 5.418614387512207,
      "epoch": 1.7929063860667633,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004680218865675567,
      "loss": 4.9464,
      "mean_token_accuracy": 0.20754242390394212,
      "num_tokens": 41028972.0,
      "step": 19765
    },
    {
      "entropy": 5.456475925445557,
      "epoch": 1.7933599419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004680052579554926,
      "loss": 4.9764,
      "mean_token_accuracy": 0.2124166816473007,
      "num_tokens": 41039552.0,
      "step": 19770
    },
    {
      "entropy": 5.578098201751709,
      "epoch": 1.7938134978229319,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004679886253520307,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19782216101884842,
      "num_tokens": 41049808.0,
      "step": 19775
    },
    {
      "entropy": 5.502855920791626,
      "epoch": 1.794267053701016,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046797198875751474,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20240353643894196,
      "num_tokens": 41060574.0,
      "step": 19780
    },
    {
      "entropy": 5.557843923568726,
      "epoch": 1.7947206095791002,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004679553481722889,
      "loss": 5.114,
      "mean_token_accuracy": 0.19920523911714555,
      "num_tokens": 41071239.0,
      "step": 19785
    },
    {
      "entropy": 5.560058546066284,
      "epoch": 1.7951741654571842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046793870359669733,
      "loss": 5.1328,
      "mean_token_accuracy": 0.2041477620601654,
      "num_tokens": 41080819.0,
      "step": 19790
    },
    {
      "entropy": 5.527128267288208,
      "epoch": 1.7956277213352685,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004679220550310842,
      "loss": 4.927,
      "mean_token_accuracy": 0.21851669251918793,
      "num_tokens": 41090457.0,
      "step": 19795
    },
    {
      "entropy": 5.4643487453460695,
      "epoch": 1.7960812772133528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004679054024757937,
      "loss": 4.9849,
      "mean_token_accuracy": 0.2091774135828018,
      "num_tokens": 41101210.0,
      "step": 19800
    },
    {
      "entropy": 5.431383275985718,
      "epoch": 1.796534833091437,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004678887459311704,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2089541047811508,
      "num_tokens": 41110566.0,
      "step": 19805
    },
    {
      "entropy": 5.413889169692993,
      "epoch": 1.796988388969521,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004678720853975587,
      "loss": 4.9528,
      "mean_token_accuracy": 0.21217461824417114,
      "num_tokens": 41120000.0,
      "step": 19810
    },
    {
      "entropy": 5.414611577987671,
      "epoch": 1.797441944847605,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046785542087530307,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2142721250653267,
      "num_tokens": 41129944.0,
      "step": 19815
    },
    {
      "entropy": 5.445405101776123,
      "epoch": 1.7978955007256894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004678387523647483,
      "loss": 4.9775,
      "mean_token_accuracy": 0.2055494248867035,
      "num_tokens": 41138803.0,
      "step": 19820
    },
    {
      "entropy": 5.493655824661255,
      "epoch": 1.7983490566037736,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046782207986623885,
      "loss": 5.0722,
      "mean_token_accuracy": 0.1955258533358574,
      "num_tokens": 41150365.0,
      "step": 19825
    },
    {
      "entropy": 5.62580189704895,
      "epoch": 1.798802612481858,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046780540338011974,
      "loss": 5.0287,
      "mean_token_accuracy": 0.20109088867902755,
      "num_tokens": 41160310.0,
      "step": 19830
    },
    {
      "entropy": 5.464039134979248,
      "epoch": 1.799256168359942,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004677887229067358,
      "loss": 4.894,
      "mean_token_accuracy": 0.21150053292512894,
      "num_tokens": 41170506.0,
      "step": 19835
    },
    {
      "entropy": 5.487150049209594,
      "epoch": 1.799709724238026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046777203844643184,
      "loss": 5.0238,
      "mean_token_accuracy": 0.19993437677621842,
      "num_tokens": 41181293.0,
      "step": 19840
    },
    {
      "entropy": 5.455536556243897,
      "epoch": 1.8001632801161103,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046775534999955303,
      "loss": 4.9545,
      "mean_token_accuracy": 0.20341767817735673,
      "num_tokens": 41191637.0,
      "step": 19845
    },
    {
      "entropy": 5.460039043426514,
      "epoch": 1.8006168359941945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004677386575664445,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2183227390050888,
      "num_tokens": 41201833.0,
      "step": 19850
    },
    {
      "entropy": 5.522752523422241,
      "epoch": 1.8010703918722788,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046772196114745133,
      "loss": 5.0139,
      "mean_token_accuracy": 0.20963264256715775,
      "num_tokens": 41211243.0,
      "step": 19855
    },
    {
      "entropy": 5.501462697982788,
      "epoch": 1.8015239477503628,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004677052607429189,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2089851349592209,
      "num_tokens": 41221111.0,
      "step": 19860
    },
    {
      "entropy": 5.443460083007812,
      "epoch": 1.8019775036284469,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004676885563531927,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20403359532356263,
      "num_tokens": 41231578.0,
      "step": 19865
    },
    {
      "entropy": 5.4814146041870115,
      "epoch": 1.8024310595065312,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046767184797861795,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20797489285469056,
      "num_tokens": 41242130.0,
      "step": 19870
    },
    {
      "entropy": 5.517671585083008,
      "epoch": 1.8028846153846154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046765513561954026,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20290981829166413,
      "num_tokens": 41252885.0,
      "step": 19875
    },
    {
      "entropy": 5.534820604324341,
      "epoch": 1.8033381712626997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046763841927630534,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20083272457122803,
      "num_tokens": 41263532.0,
      "step": 19880
    },
    {
      "entropy": 5.542397499084473,
      "epoch": 1.8037917271407837,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046762169894925876,
      "loss": 4.9931,
      "mean_token_accuracy": 0.20428968071937562,
      "num_tokens": 41273469.0,
      "step": 19885
    },
    {
      "entropy": 5.500296211242675,
      "epoch": 1.8042452830188678,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004676049746387464,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20310233831405639,
      "num_tokens": 41283663.0,
      "step": 19890
    },
    {
      "entropy": 5.447289371490479,
      "epoch": 1.804698838896952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004675882463451141,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21334871649742126,
      "num_tokens": 41294657.0,
      "step": 19895
    },
    {
      "entropy": 5.526976108551025,
      "epoch": 1.8051523947750363,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004675715140687078,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20500462055206298,
      "num_tokens": 41305124.0,
      "step": 19900
    },
    {
      "entropy": 5.407881498336792,
      "epoch": 1.8056059506531206,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046755477780987344,
      "loss": 4.9383,
      "mean_token_accuracy": 0.20866432934999465,
      "num_tokens": 41315839.0,
      "step": 19905
    },
    {
      "entropy": 5.51832537651062,
      "epoch": 1.8060595065312046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046753803756895726,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20364923328161239,
      "num_tokens": 41326388.0,
      "step": 19910
    },
    {
      "entropy": 5.486484050750732,
      "epoch": 1.8065130624092887,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004675212933463054,
      "loss": 5.0667,
      "mean_token_accuracy": 0.1960567370057106,
      "num_tokens": 41337333.0,
      "step": 19915
    },
    {
      "entropy": 5.580039548873901,
      "epoch": 1.806966618287373,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004675045451422641,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19424721002578735,
      "num_tokens": 41348353.0,
      "step": 19920
    },
    {
      "entropy": 5.54701042175293,
      "epoch": 1.8074201741654572,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046748779295717973,
      "loss": 5.0226,
      "mean_token_accuracy": 0.21309632658958436,
      "num_tokens": 41358797.0,
      "step": 19925
    },
    {
      "entropy": 5.512184047698975,
      "epoch": 1.8078737300435415,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004674710367913989,
      "loss": 5.0224,
      "mean_token_accuracy": 0.20770838856697083,
      "num_tokens": 41368818.0,
      "step": 19930
    },
    {
      "entropy": 5.526298379898071,
      "epoch": 1.8083272859216255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004674542766452678,
      "loss": 5.0055,
      "mean_token_accuracy": 0.20612242072820663,
      "num_tokens": 41378949.0,
      "step": 19935
    },
    {
      "entropy": 5.501234817504883,
      "epoch": 1.8087808417997098,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046743751251913327,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2014146015048027,
      "num_tokens": 41388892.0,
      "step": 19940
    },
    {
      "entropy": 5.545752954483032,
      "epoch": 1.8092343976777938,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046742074441334197,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20861671566963197,
      "num_tokens": 41399197.0,
      "step": 19945
    },
    {
      "entropy": 5.510498189926148,
      "epoch": 1.809687953555878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004674039723282406,
      "loss": 5.0272,
      "mean_token_accuracy": 0.19636441618204117,
      "num_tokens": 41409551.0,
      "step": 19950
    },
    {
      "entropy": 5.468385839462281,
      "epoch": 1.8101415094339623,
      "grad_norm": 1.125,
      "learning_rate": 0.00046738719626417607,
      "loss": 4.9337,
      "mean_token_accuracy": 0.20791163742542268,
      "num_tokens": 41418900.0,
      "step": 19955
    },
    {
      "entropy": 5.435397815704346,
      "epoch": 1.8105950653120464,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046737041622149536,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2077833577990532,
      "num_tokens": 41428204.0,
      "step": 19960
    },
    {
      "entropy": 5.475276470184326,
      "epoch": 1.8110486211901307,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004673536322005454,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20147773921489714,
      "num_tokens": 41437922.0,
      "step": 19965
    },
    {
      "entropy": 5.540088272094726,
      "epoch": 1.8115021770682147,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004673368442016732,
      "loss": 5.0115,
      "mean_token_accuracy": 0.19927300810813903,
      "num_tokens": 41447411.0,
      "step": 19970
    },
    {
      "entropy": 5.578990268707275,
      "epoch": 1.811955732946299,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004673200522252261,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20416804701089858,
      "num_tokens": 41458756.0,
      "step": 19975
    },
    {
      "entropy": 5.508602094650269,
      "epoch": 1.8124092888243832,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004673032562715514,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20580196678638457,
      "num_tokens": 41468610.0,
      "step": 19980
    },
    {
      "entropy": 5.503843450546265,
      "epoch": 1.8128628447024675,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046728645634099627,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20053563714027406,
      "num_tokens": 41479849.0,
      "step": 19985
    },
    {
      "entropy": 5.4787674903869625,
      "epoch": 1.8133164005805515,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046726965243390825,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2042739525437355,
      "num_tokens": 41490945.0,
      "step": 19990
    },
    {
      "entropy": 5.451112937927246,
      "epoch": 1.8137699564586356,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004672528445506348,
      "loss": 4.9564,
      "mean_token_accuracy": 0.19990914165973664,
      "num_tokens": 41501479.0,
      "step": 19995
    },
    {
      "entropy": 5.494516277313233,
      "epoch": 1.8142235123367199,
      "grad_norm": 1.125,
      "learning_rate": 0.0004672360326915236,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2147782877087593,
      "num_tokens": 41512447.0,
      "step": 20000
    },
    {
      "entropy": 5.467796421051025,
      "epoch": 1.8146770682148041,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004672192168569222,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2156030833721161,
      "num_tokens": 41521905.0,
      "step": 20005
    },
    {
      "entropy": 5.481906700134277,
      "epoch": 1.8151306240928884,
      "grad_norm": 1.0,
      "learning_rate": 0.00046720239704717845,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20792933702468872,
      "num_tokens": 41533143.0,
      "step": 20010
    },
    {
      "entropy": 5.4936775207519535,
      "epoch": 1.8155841799709724,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046718557326264,
      "loss": 4.956,
      "mean_token_accuracy": 0.20631704181432725,
      "num_tokens": 41542234.0,
      "step": 20015
    },
    {
      "entropy": 5.532553005218506,
      "epoch": 1.8160377358490565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004671687455036551,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20308319926261903,
      "num_tokens": 41552632.0,
      "step": 20020
    },
    {
      "entropy": 5.479432058334351,
      "epoch": 1.8164912917271407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004671519137705715,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20832333713769913,
      "num_tokens": 41563723.0,
      "step": 20025
    },
    {
      "entropy": 5.475570821762085,
      "epoch": 1.816944847605225,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046713507806373736,
      "loss": 4.9629,
      "mean_token_accuracy": 0.20661877393722533,
      "num_tokens": 41573680.0,
      "step": 20030
    },
    {
      "entropy": 5.518227291107178,
      "epoch": 1.8173984034833093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046711823838350086,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20710206627845765,
      "num_tokens": 41584722.0,
      "step": 20035
    },
    {
      "entropy": 5.508468818664551,
      "epoch": 1.8178519593613933,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046710139473021025,
      "loss": 5.0779,
      "mean_token_accuracy": 0.1963120073080063,
      "num_tokens": 41595785.0,
      "step": 20040
    },
    {
      "entropy": 5.507241678237915,
      "epoch": 1.8183055152394774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046708454710421376,
      "loss": 5.0907,
      "mean_token_accuracy": 0.19826340824365615,
      "num_tokens": 41606023.0,
      "step": 20045
    },
    {
      "entropy": 5.491629981994629,
      "epoch": 1.8187590711175616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046706769550586,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20706136375665665,
      "num_tokens": 41616010.0,
      "step": 20050
    },
    {
      "entropy": 5.489751577377319,
      "epoch": 1.819212626995646,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046705083993549723,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20925389379262924,
      "num_tokens": 41625105.0,
      "step": 20055
    },
    {
      "entropy": 5.483782625198364,
      "epoch": 1.8196661828737302,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004670339803934743,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20866930335760117,
      "num_tokens": 41634989.0,
      "step": 20060
    },
    {
      "entropy": 5.477025508880615,
      "epoch": 1.8201197387518142,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004670171168801396,
      "loss": 4.9939,
      "mean_token_accuracy": 0.20496300458908082,
      "num_tokens": 41645736.0,
      "step": 20065
    },
    {
      "entropy": 5.5498247146606445,
      "epoch": 1.8205732946298983,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000467000249395842,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2136788547039032,
      "num_tokens": 41655505.0,
      "step": 20070
    },
    {
      "entropy": 5.474360561370849,
      "epoch": 1.8210268505079825,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004669833779409303,
      "loss": 4.9804,
      "mean_token_accuracy": 0.21159180998802185,
      "num_tokens": 41665732.0,
      "step": 20075
    },
    {
      "entropy": 5.469687223434448,
      "epoch": 1.8214804063860668,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004669665025157535,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2019777148962021,
      "num_tokens": 41676050.0,
      "step": 20080
    },
    {
      "entropy": 5.546350574493408,
      "epoch": 1.821933962264151,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046694962312066046,
      "loss": 5.1338,
      "mean_token_accuracy": 0.19883247166872026,
      "num_tokens": 41686787.0,
      "step": 20085
    },
    {
      "entropy": 5.478995037078858,
      "epoch": 1.822387518142235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046693273975600036,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21162950843572617,
      "num_tokens": 41697168.0,
      "step": 20090
    },
    {
      "entropy": 5.469726848602295,
      "epoch": 1.8228410740203191,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004669158524221222,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20439180731773376,
      "num_tokens": 41707524.0,
      "step": 20095
    },
    {
      "entropy": 5.531174993515014,
      "epoch": 1.8232946298984034,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046689896111937537,
      "loss": 5.0566,
      "mean_token_accuracy": 0.2075119212269783,
      "num_tokens": 41718109.0,
      "step": 20100
    },
    {
      "entropy": 5.569089317321778,
      "epoch": 1.8237481857764877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004668820658481091,
      "loss": 5.0191,
      "mean_token_accuracy": 0.2051020994782448,
      "num_tokens": 41728862.0,
      "step": 20105
    },
    {
      "entropy": 5.577348279953003,
      "epoch": 1.824201741654572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046686516660867283,
      "loss": 5.0825,
      "mean_token_accuracy": 0.20027903616428375,
      "num_tokens": 41739024.0,
      "step": 20110
    },
    {
      "entropy": 5.505842971801758,
      "epoch": 1.824655297532656,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004668482634014159,
      "loss": 5.0253,
      "mean_token_accuracy": 0.20351107120513917,
      "num_tokens": 41748979.0,
      "step": 20115
    },
    {
      "entropy": 5.486580085754395,
      "epoch": 1.8251088534107403,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004668313562266882,
      "loss": 5.0337,
      "mean_token_accuracy": 0.20621970891952515,
      "num_tokens": 41759972.0,
      "step": 20120
    },
    {
      "entropy": 5.456143379211426,
      "epoch": 1.8255624092888243,
      "grad_norm": 1.078125,
      "learning_rate": 0.000466814445084839,
      "loss": 4.8901,
      "mean_token_accuracy": 0.21461684107780457,
      "num_tokens": 41770150.0,
      "step": 20125
    },
    {
      "entropy": 5.478838014602661,
      "epoch": 1.8260159651669086,
      "grad_norm": 1.125,
      "learning_rate": 0.0004667975299762183,
      "loss": 4.9686,
      "mean_token_accuracy": 0.20933082401752473,
      "num_tokens": 41780405.0,
      "step": 20130
    },
    {
      "entropy": 5.472921085357666,
      "epoch": 1.8264695210449928,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046678061090117576,
      "loss": 4.897,
      "mean_token_accuracy": 0.21493014693260193,
      "num_tokens": 41791269.0,
      "step": 20135
    },
    {
      "entropy": 5.416600275039673,
      "epoch": 1.8269230769230769,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004667636878600613,
      "loss": 4.91,
      "mean_token_accuracy": 0.20953293442726134,
      "num_tokens": 41802513.0,
      "step": 20140
    },
    {
      "entropy": 5.494254446029663,
      "epoch": 1.8273766328011611,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004667467608532249,
      "loss": 5.0374,
      "mean_token_accuracy": 0.19939105957746506,
      "num_tokens": 41814149.0,
      "step": 20145
    },
    {
      "entropy": 5.5285478115081785,
      "epoch": 1.8278301886792452,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004667298298810167,
      "loss": 5.0231,
      "mean_token_accuracy": 0.20312486886978148,
      "num_tokens": 41825763.0,
      "step": 20150
    },
    {
      "entropy": 5.472637367248535,
      "epoch": 1.8282837445573294,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004667128949437867,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20898271799087526,
      "num_tokens": 41834543.0,
      "step": 20155
    },
    {
      "entropy": 5.44488296508789,
      "epoch": 1.8287373004354137,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004666959560418852,
      "loss": 4.9,
      "mean_token_accuracy": 0.2235007792711258,
      "num_tokens": 41844940.0,
      "step": 20160
    },
    {
      "entropy": 5.422686767578125,
      "epoch": 1.829190856313498,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046667901317566245,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21362340897321702,
      "num_tokens": 41855532.0,
      "step": 20165
    },
    {
      "entropy": 5.5042558193206785,
      "epoch": 1.829644412191582,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004666620663454689,
      "loss": 5.0277,
      "mean_token_accuracy": 0.20293494015932084,
      "num_tokens": 41865661.0,
      "step": 20170
    },
    {
      "entropy": 5.498259353637695,
      "epoch": 1.830097968069666,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046664511555165496,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20515236705541612,
      "num_tokens": 41875662.0,
      "step": 20175
    },
    {
      "entropy": 5.469655990600586,
      "epoch": 1.8305515239477503,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046662816079457114,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21497792154550552,
      "num_tokens": 41886211.0,
      "step": 20180
    },
    {
      "entropy": 5.566941928863526,
      "epoch": 1.8310050798258346,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004666112020745682,
      "loss": 5.0905,
      "mean_token_accuracy": 0.1984022229909897,
      "num_tokens": 41896737.0,
      "step": 20185
    },
    {
      "entropy": 5.490166807174683,
      "epoch": 1.8314586357039189,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004665942393919967,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20530310571193694,
      "num_tokens": 41907768.0,
      "step": 20190
    },
    {
      "entropy": 5.489934015274048,
      "epoch": 1.831912191582003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046657727274720755,
      "loss": 4.9329,
      "mean_token_accuracy": 0.20314387530088424,
      "num_tokens": 41917587.0,
      "step": 20195
    },
    {
      "entropy": 5.4147562980651855,
      "epoch": 1.832365747460087,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004665603021405516,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21483613103628157,
      "num_tokens": 41927842.0,
      "step": 20200
    },
    {
      "entropy": 5.438058376312256,
      "epoch": 1.8328193033381712,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004665433275723798,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2155444696545601,
      "num_tokens": 41939899.0,
      "step": 20205
    },
    {
      "entropy": 5.460914945602417,
      "epoch": 1.8332728592162555,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004665263490430431,
      "loss": 4.9553,
      "mean_token_accuracy": 0.20326370745897293,
      "num_tokens": 41950932.0,
      "step": 20210
    },
    {
      "entropy": 5.532335186004639,
      "epoch": 1.8337264150943398,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004665093665528928,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20834872424602507,
      "num_tokens": 41963512.0,
      "step": 20215
    },
    {
      "entropy": 5.545543670654297,
      "epoch": 1.8341799709724238,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046649238010227983,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2091602176427841,
      "num_tokens": 41974677.0,
      "step": 20220
    },
    {
      "entropy": 5.390093803405762,
      "epoch": 1.8346335268505078,
      "grad_norm": 1.0,
      "learning_rate": 0.0004664753896915558,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2121974930167198,
      "num_tokens": 41984582.0,
      "step": 20225
    },
    {
      "entropy": 5.461968660354614,
      "epoch": 1.8350870827285921,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004664583953210719,
      "loss": 4.943,
      "mean_token_accuracy": 0.20632002502679825,
      "num_tokens": 41994624.0,
      "step": 20230
    },
    {
      "entropy": 5.4864002704620365,
      "epoch": 1.8355406386066764,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004664413969911795,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21286633610725403,
      "num_tokens": 42004835.0,
      "step": 20235
    },
    {
      "entropy": 5.426284551620483,
      "epoch": 1.8359941944847606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004664243947022303,
      "loss": 4.9597,
      "mean_token_accuracy": 0.2085785835981369,
      "num_tokens": 42014727.0,
      "step": 20240
    },
    {
      "entropy": 5.475789403915405,
      "epoch": 1.8364477503628447,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046640738845457586,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20792288333177567,
      "num_tokens": 42023628.0,
      "step": 20245
    },
    {
      "entropy": 5.553183269500733,
      "epoch": 1.8369013062409287,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004663903782485678,
      "loss": 5.0523,
      "mean_token_accuracy": 0.19967232644557953,
      "num_tokens": 42033587.0,
      "step": 20250
    },
    {
      "entropy": 5.560832595825195,
      "epoch": 1.837354862119013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046637336408455793,
      "loss": 5.021,
      "mean_token_accuracy": 0.20333033055067062,
      "num_tokens": 42043765.0,
      "step": 20255
    },
    {
      "entropy": 5.563081312179565,
      "epoch": 1.8378084179970973,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046635634596289816,
      "loss": 5.0832,
      "mean_token_accuracy": 0.2054321736097336,
      "num_tokens": 42054205.0,
      "step": 20260
    },
    {
      "entropy": 5.401016807556152,
      "epoch": 1.8382619738751815,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004663393238839404,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2077532708644867,
      "num_tokens": 42064512.0,
      "step": 20265
    },
    {
      "entropy": 5.4527510643005375,
      "epoch": 1.8387155297532656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004663222978480366,
      "loss": 4.9961,
      "mean_token_accuracy": 0.202800253033638,
      "num_tokens": 42074954.0,
      "step": 20270
    },
    {
      "entropy": 5.539501094818116,
      "epoch": 1.8391690856313496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000466305267855539,
      "loss": 5.0011,
      "mean_token_accuracy": 0.20585793405771255,
      "num_tokens": 42084914.0,
      "step": 20275
    },
    {
      "entropy": 5.524646329879761,
      "epoch": 1.8396226415094339,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004662882339067996,
      "loss": 5.0194,
      "mean_token_accuracy": 0.20999543070793153,
      "num_tokens": 42095454.0,
      "step": 20280
    },
    {
      "entropy": 5.478907346725464,
      "epoch": 1.8400761973875182,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004662711960021709,
      "loss": 5.0565,
      "mean_token_accuracy": 0.20169686377048493,
      "num_tokens": 42106960.0,
      "step": 20285
    },
    {
      "entropy": 5.497449636459351,
      "epoch": 1.8405297532656024,
      "grad_norm": 1.125,
      "learning_rate": 0.00046625415414200513,
      "loss": 4.9615,
      "mean_token_accuracy": 0.20768564641475679,
      "num_tokens": 42117009.0,
      "step": 20290
    },
    {
      "entropy": 5.526269197463989,
      "epoch": 1.8409833091436865,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004662371083266546,
      "loss": 4.9976,
      "mean_token_accuracy": 0.211554653942585,
      "num_tokens": 42126371.0,
      "step": 20295
    },
    {
      "entropy": 5.51113486289978,
      "epoch": 1.8414368650217707,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000466220058556472,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20377402901649475,
      "num_tokens": 42135890.0,
      "step": 20300
    },
    {
      "entropy": 5.552067518234253,
      "epoch": 1.8418904208998548,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046620300483180986,
      "loss": 5.1245,
      "mean_token_accuracy": 0.19851610958576202,
      "num_tokens": 42145696.0,
      "step": 20305
    },
    {
      "entropy": 5.514868354797363,
      "epoch": 1.842343976777939,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004661859471530209,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21122143566608428,
      "num_tokens": 42156236.0,
      "step": 20310
    },
    {
      "entropy": 5.5421325206756595,
      "epoch": 1.8427975326560233,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004661688855204578,
      "loss": 5.041,
      "mean_token_accuracy": 0.20424984842538835,
      "num_tokens": 42166179.0,
      "step": 20315
    },
    {
      "entropy": 5.45050687789917,
      "epoch": 1.8432510885341074,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046615181993447344,
      "loss": 4.9217,
      "mean_token_accuracy": 0.20572050362825395,
      "num_tokens": 42176643.0,
      "step": 20320
    },
    {
      "entropy": 5.550262832641602,
      "epoch": 1.8437046444121916,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004661347503954207,
      "loss": 5.19,
      "mean_token_accuracy": 0.19296130239963533,
      "num_tokens": 42188328.0,
      "step": 20325
    },
    {
      "entropy": 5.56198468208313,
      "epoch": 1.8441582002902757,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046611767690365265,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20894695967435836,
      "num_tokens": 42198110.0,
      "step": 20330
    },
    {
      "entropy": 5.44454779624939,
      "epoch": 1.84461175616836,
      "grad_norm": 1.125,
      "learning_rate": 0.0004661005994595223,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20088055431842805,
      "num_tokens": 42208438.0,
      "step": 20335
    },
    {
      "entropy": 5.473340320587158,
      "epoch": 1.8450653120464442,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004660835180633829,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2040846347808838,
      "num_tokens": 42220052.0,
      "step": 20340
    },
    {
      "entropy": 5.47958402633667,
      "epoch": 1.8455188679245285,
      "grad_norm": 1.125,
      "learning_rate": 0.0004660664327155876,
      "loss": 4.967,
      "mean_token_accuracy": 0.21261545717716218,
      "num_tokens": 42229497.0,
      "step": 20345
    },
    {
      "entropy": 5.46050238609314,
      "epoch": 1.8459724238026125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004660493434164898,
      "loss": 5.0286,
      "mean_token_accuracy": 0.1993781164288521,
      "num_tokens": 42240725.0,
      "step": 20350
    },
    {
      "entropy": 5.543264579772949,
      "epoch": 1.8464259796806965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004660322501664429,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20247761607170106,
      "num_tokens": 42251114.0,
      "step": 20355
    },
    {
      "entropy": 5.580515146255493,
      "epoch": 1.8468795355587808,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004660151529658004,
      "loss": 5.1806,
      "mean_token_accuracy": 0.19662023484706878,
      "num_tokens": 42262825.0,
      "step": 20360
    },
    {
      "entropy": 5.54121413230896,
      "epoch": 1.847333091436865,
      "grad_norm": 1.125,
      "learning_rate": 0.0004659980518149158,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21283550858497619,
      "num_tokens": 42272176.0,
      "step": 20365
    },
    {
      "entropy": 5.5341148853302,
      "epoch": 1.8477866473149493,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046598094671414285,
      "loss": 5.0617,
      "mean_token_accuracy": 0.20483849495649337,
      "num_tokens": 42281934.0,
      "step": 20370
    },
    {
      "entropy": 5.507667827606201,
      "epoch": 1.8482402031930334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046596383766383533,
      "loss": 4.948,
      "mean_token_accuracy": 0.20928735733032228,
      "num_tokens": 42291927.0,
      "step": 20375
    },
    {
      "entropy": 5.46729621887207,
      "epoch": 1.8486937590711174,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004659467246643469,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2095313012599945,
      "num_tokens": 42301407.0,
      "step": 20380
    },
    {
      "entropy": 5.39543194770813,
      "epoch": 1.8491473149492017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046592960771603154,
      "loss": 4.8996,
      "mean_token_accuracy": 0.21575131714344026,
      "num_tokens": 42311199.0,
      "step": 20385
    },
    {
      "entropy": 5.490356159210205,
      "epoch": 1.849600870827286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004659124868192433,
      "loss": 4.9938,
      "mean_token_accuracy": 0.2058865636587143,
      "num_tokens": 42320725.0,
      "step": 20390
    },
    {
      "entropy": 5.550732135772705,
      "epoch": 1.8500544267053702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004658953619743361,
      "loss": 5.0822,
      "mean_token_accuracy": 0.20801912099123002,
      "num_tokens": 42330857.0,
      "step": 20395
    },
    {
      "entropy": 5.472535848617554,
      "epoch": 1.8505079825834543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046587823318166427,
      "loss": 4.8952,
      "mean_token_accuracy": 0.20971542596817017,
      "num_tokens": 42340241.0,
      "step": 20400
    },
    {
      "entropy": 5.531248378753662,
      "epoch": 1.8509615384615383,
      "grad_norm": 1.0,
      "learning_rate": 0.00046586110044158184,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19616152048110963,
      "num_tokens": 42351756.0,
      "step": 20405
    },
    {
      "entropy": 5.493772363662719,
      "epoch": 1.8514150943396226,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046584396375444335,
      "loss": 4.946,
      "mean_token_accuracy": 0.21230319291353225,
      "num_tokens": 42361649.0,
      "step": 20410
    },
    {
      "entropy": 5.492153215408325,
      "epoch": 1.8518686502177069,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046582682312060295,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2071332037448883,
      "num_tokens": 42371273.0,
      "step": 20415
    },
    {
      "entropy": 5.422970247268677,
      "epoch": 1.8523222060957911,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004658096785404152,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20888074934482576,
      "num_tokens": 42382275.0,
      "step": 20420
    },
    {
      "entropy": 5.555678844451904,
      "epoch": 1.8527757619738752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046579253001423473,
      "loss": 5.0086,
      "mean_token_accuracy": 0.2013235032558441,
      "num_tokens": 42392360.0,
      "step": 20425
    },
    {
      "entropy": 5.509761190414428,
      "epoch": 1.8532293178519592,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046577537754241604,
      "loss": 5.0019,
      "mean_token_accuracy": 0.1991646781563759,
      "num_tokens": 42403695.0,
      "step": 20430
    },
    {
      "entropy": 5.48563666343689,
      "epoch": 1.8536828737300435,
      "grad_norm": 0.96875,
      "learning_rate": 0.000465758221125314,
      "loss": 4.941,
      "mean_token_accuracy": 0.20885043293237687,
      "num_tokens": 42414199.0,
      "step": 20435
    },
    {
      "entropy": 5.505152606964112,
      "epoch": 1.8541364296081277,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046574106076328333,
      "loss": 5.0589,
      "mean_token_accuracy": 0.20741629600524902,
      "num_tokens": 42423655.0,
      "step": 20440
    },
    {
      "entropy": 5.484323310852051,
      "epoch": 1.854589985486212,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004657238964566789,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21008907109498978,
      "num_tokens": 42434235.0,
      "step": 20445
    },
    {
      "entropy": 5.441372680664062,
      "epoch": 1.855043541364296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004657067282058557,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20948562026023865,
      "num_tokens": 42444701.0,
      "step": 20450
    },
    {
      "entropy": 5.419203615188598,
      "epoch": 1.85549709724238,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004656895560111687,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20637976378202438,
      "num_tokens": 42454499.0,
      "step": 20455
    },
    {
      "entropy": 5.532791233062744,
      "epoch": 1.8559506531204644,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046567237987297323,
      "loss": 4.9561,
      "mean_token_accuracy": 0.19987907409667968,
      "num_tokens": 42465391.0,
      "step": 20460
    },
    {
      "entropy": 5.55144419670105,
      "epoch": 1.8564042089985486,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004656551997916243,
      "loss": 4.961,
      "mean_token_accuracy": 0.19955171942710875,
      "num_tokens": 42476126.0,
      "step": 20465
    },
    {
      "entropy": 5.52060604095459,
      "epoch": 1.856857764876633,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046563801576747716,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20867771059274673,
      "num_tokens": 42486833.0,
      "step": 20470
    },
    {
      "entropy": 5.593228769302368,
      "epoch": 1.857311320754717,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046562082780088733,
      "loss": 5.0775,
      "mean_token_accuracy": 0.2004850447177887,
      "num_tokens": 42496986.0,
      "step": 20475
    },
    {
      "entropy": 5.513063526153564,
      "epoch": 1.8577648766328012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004656036358922103,
      "loss": 4.9818,
      "mean_token_accuracy": 0.2030161887407303,
      "num_tokens": 42507352.0,
      "step": 20480
    },
    {
      "entropy": 5.512900876998901,
      "epoch": 1.8582184325108853,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004655864400418014,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21155350506305695,
      "num_tokens": 42517655.0,
      "step": 20485
    },
    {
      "entropy": 5.454225349426269,
      "epoch": 1.8586719883889695,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004655692402500163,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21709479838609697,
      "num_tokens": 42527502.0,
      "step": 20490
    },
    {
      "entropy": 5.460311651229858,
      "epoch": 1.8591255442670538,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046555203651721085,
      "loss": 4.9621,
      "mean_token_accuracy": 0.20758305490016937,
      "num_tokens": 42538893.0,
      "step": 20495
    },
    {
      "entropy": 5.373780870437622,
      "epoch": 1.8595791001451378,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004655348288437407,
      "loss": 4.8608,
      "mean_token_accuracy": 0.21704194992780684,
      "num_tokens": 42550372.0,
      "step": 20500
    },
    {
      "entropy": 5.526129817962646,
      "epoch": 1.860032656023222,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004655176172299618,
      "loss": 5.0732,
      "mean_token_accuracy": 0.20077297538518907,
      "num_tokens": 42560814.0,
      "step": 20505
    },
    {
      "entropy": 5.5809577941894535,
      "epoch": 1.8604862119013061,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046550040167622986,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1971066802740097,
      "num_tokens": 42570993.0,
      "step": 20510
    },
    {
      "entropy": 5.517111206054688,
      "epoch": 1.8609397677793904,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046548318218290124,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21109461635351182,
      "num_tokens": 42581357.0,
      "step": 20515
    },
    {
      "entropy": 5.476054048538208,
      "epoch": 1.8613933236574747,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004654659587503318,
      "loss": 5.0456,
      "mean_token_accuracy": 0.19604293555021285,
      "num_tokens": 42591653.0,
      "step": 20520
    },
    {
      "entropy": 5.563223028182984,
      "epoch": 1.861846879535559,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004654487313788778,
      "loss": 5.0968,
      "mean_token_accuracy": 0.19883365482091903,
      "num_tokens": 42602232.0,
      "step": 20525
    },
    {
      "entropy": 5.5313554286956785,
      "epoch": 1.862300435413643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046543150006889546,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21429141610860825,
      "num_tokens": 42613283.0,
      "step": 20530
    },
    {
      "entropy": 5.474745416641236,
      "epoch": 1.862753991291727,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004654142648207412,
      "loss": 4.9336,
      "mean_token_accuracy": 0.209904220700264,
      "num_tokens": 42623594.0,
      "step": 20535
    },
    {
      "entropy": 5.5373499393463135,
      "epoch": 1.8632075471698113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004653970256347715,
      "loss": 5.0483,
      "mean_token_accuracy": 0.2047143191099167,
      "num_tokens": 42633587.0,
      "step": 20540
    },
    {
      "entropy": 5.542779874801636,
      "epoch": 1.8636611030478956,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046537978251134265,
      "loss": 4.9779,
      "mean_token_accuracy": 0.20232055485248565,
      "num_tokens": 42643426.0,
      "step": 20545
    },
    {
      "entropy": 5.42120099067688,
      "epoch": 1.8641146589259798,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004653625354508115,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20807553082704544,
      "num_tokens": 42653521.0,
      "step": 20550
    },
    {
      "entropy": 5.505074691772461,
      "epoch": 1.8645682148040639,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004653452844535346,
      "loss": 5.025,
      "mean_token_accuracy": 0.20993955433368683,
      "num_tokens": 42664818.0,
      "step": 20555
    },
    {
      "entropy": 5.59228253364563,
      "epoch": 1.865021770682148,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046532802951986863,
      "loss": 5.0577,
      "mean_token_accuracy": 0.2088144451379776,
      "num_tokens": 42674738.0,
      "step": 20560
    },
    {
      "entropy": 5.500059461593628,
      "epoch": 1.8654753265602322,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004653107706501705,
      "loss": 4.9561,
      "mean_token_accuracy": 0.2192043334245682,
      "num_tokens": 42685003.0,
      "step": 20565
    },
    {
      "entropy": 5.44640154838562,
      "epoch": 1.8659288824383164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004652935078447973,
      "loss": 5.0064,
      "mean_token_accuracy": 0.20773745477199554,
      "num_tokens": 42695428.0,
      "step": 20570
    },
    {
      "entropy": 5.500756025314331,
      "epoch": 1.8663824383164007,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046527624110410574,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21193516999483109,
      "num_tokens": 42704901.0,
      "step": 20575
    },
    {
      "entropy": 5.521413898468017,
      "epoch": 1.8668359941944848,
      "grad_norm": 1.109375,
      "learning_rate": 0.000465258970428453,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20478675365447999,
      "num_tokens": 42714379.0,
      "step": 20580
    },
    {
      "entropy": 5.577829313278198,
      "epoch": 1.8672895500725688,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046524169581819633,
      "loss": 5.1601,
      "mean_token_accuracy": 0.1995236471295357,
      "num_tokens": 42724813.0,
      "step": 20585
    },
    {
      "entropy": 5.455184555053711,
      "epoch": 1.867743105950653,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004652244172736929,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2048965722322464,
      "num_tokens": 42736590.0,
      "step": 20590
    },
    {
      "entropy": 5.5260978698730465,
      "epoch": 1.8681966618287373,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004652071347953,
      "loss": 5.0807,
      "mean_token_accuracy": 0.1991588518023491,
      "num_tokens": 42746718.0,
      "step": 20595
    },
    {
      "entropy": 5.511317920684815,
      "epoch": 1.8686502177068216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046518984838337513,
      "loss": 4.9682,
      "mean_token_accuracy": 0.2041142016649246,
      "num_tokens": 42756484.0,
      "step": 20600
    },
    {
      "entropy": 5.532030200958252,
      "epoch": 1.8691037735849056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004651725580382757,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20200948566198348,
      "num_tokens": 42768512.0,
      "step": 20605
    },
    {
      "entropy": 5.473254299163818,
      "epoch": 1.8695573294629897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046515526376035934,
      "loss": 4.9198,
      "mean_token_accuracy": 0.20792983025312423,
      "num_tokens": 42779812.0,
      "step": 20610
    },
    {
      "entropy": 5.382527160644531,
      "epoch": 1.870010885341074,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004651379655499837,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2151414394378662,
      "num_tokens": 42791163.0,
      "step": 20615
    },
    {
      "entropy": 5.455286312103271,
      "epoch": 1.8704644412191582,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046512066340750636,
      "loss": 4.9883,
      "mean_token_accuracy": 0.20826934427022933,
      "num_tokens": 42801810.0,
      "step": 20620
    },
    {
      "entropy": 5.532750940322876,
      "epoch": 1.8709179970972425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004651033573332853,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2087985247373581,
      "num_tokens": 42813358.0,
      "step": 20625
    },
    {
      "entropy": 5.396786737442016,
      "epoch": 1.8713715529753265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004650860473276784,
      "loss": 4.8341,
      "mean_token_accuracy": 0.21987762600183486,
      "num_tokens": 42823519.0,
      "step": 20630
    },
    {
      "entropy": 5.506504392623901,
      "epoch": 1.8718251088534106,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004650687333910436,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20566814988851548,
      "num_tokens": 42833257.0,
      "step": 20635
    },
    {
      "entropy": 5.558252477645874,
      "epoch": 1.8722786647314948,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046505141552373897,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21207219064235688,
      "num_tokens": 42843536.0,
      "step": 20640
    },
    {
      "entropy": 5.460723066329956,
      "epoch": 1.8727322206095791,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004650340937261226,
      "loss": 4.9834,
      "mean_token_accuracy": 0.204876309633255,
      "num_tokens": 42853911.0,
      "step": 20645
    },
    {
      "entropy": 5.533601236343384,
      "epoch": 1.8731857764876634,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004650167679985528,
      "loss": 5.1035,
      "mean_token_accuracy": 0.19457658678293227,
      "num_tokens": 42864173.0,
      "step": 20650
    },
    {
      "entropy": 5.497031116485596,
      "epoch": 1.8736393323657474,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004649994383413877,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2092764064669609,
      "num_tokens": 42873470.0,
      "step": 20655
    },
    {
      "entropy": 5.494451093673706,
      "epoch": 1.8740928882438317,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046498210475498595,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21581810265779494,
      "num_tokens": 42882496.0,
      "step": 20660
    },
    {
      "entropy": 5.450921154022216,
      "epoch": 1.8745464441219157,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046496476723970574,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20707394033670426,
      "num_tokens": 42893163.0,
      "step": 20665
    },
    {
      "entropy": 5.463796806335449,
      "epoch": 1.875,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004649474257959058,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20770224928855896,
      "num_tokens": 42903406.0,
      "step": 20670
    },
    {
      "entropy": 5.494471883773803,
      "epoch": 1.8754535558780843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004649300804239447,
      "loss": 4.9655,
      "mean_token_accuracy": 0.20818703174591063,
      "num_tokens": 42913702.0,
      "step": 20675
    },
    {
      "entropy": 5.424079513549804,
      "epoch": 1.8759071117561683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046491273112418105,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21260336339473723,
      "num_tokens": 42923877.0,
      "step": 20680
    },
    {
      "entropy": 5.4752942562103275,
      "epoch": 1.8763606676342526,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046489537789697386,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20910740345716478,
      "num_tokens": 42934393.0,
      "step": 20685
    },
    {
      "entropy": 5.485442686080932,
      "epoch": 1.8768142235123366,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046487802074268177,
      "loss": 5.0226,
      "mean_token_accuracy": 0.21294792741537094,
      "num_tokens": 42945803.0,
      "step": 20690
    },
    {
      "entropy": 5.455105447769165,
      "epoch": 1.8772677793904209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046486065966166377,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21094850897789003,
      "num_tokens": 42955829.0,
      "step": 20695
    },
    {
      "entropy": 5.510855436325073,
      "epoch": 1.8777213352685052,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004648432946542791,
      "loss": 5.0054,
      "mean_token_accuracy": 0.201736980676651,
      "num_tokens": 42967080.0,
      "step": 20700
    },
    {
      "entropy": 5.45847520828247,
      "epoch": 1.8781748911465894,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046482592572088654,
      "loss": 4.9678,
      "mean_token_accuracy": 0.19793493598699569,
      "num_tokens": 42978082.0,
      "step": 20705
    },
    {
      "entropy": 5.5167255878448485,
      "epoch": 1.8786284470246735,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046480855286184565,
      "loss": 5.0411,
      "mean_token_accuracy": 0.20442744046449662,
      "num_tokens": 42988982.0,
      "step": 20710
    },
    {
      "entropy": 5.52120566368103,
      "epoch": 1.8790820029027575,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046479117607751535,
      "loss": 4.9687,
      "mean_token_accuracy": 0.19724340736865997,
      "num_tokens": 42998982.0,
      "step": 20715
    },
    {
      "entropy": 5.482936239242553,
      "epoch": 1.8795355587808418,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046477379536825534,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21133214384317398,
      "num_tokens": 43009574.0,
      "step": 20720
    },
    {
      "entropy": 5.489485073089599,
      "epoch": 1.879989114658926,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004647564107344247,
      "loss": 4.9815,
      "mean_token_accuracy": 0.20238498002290725,
      "num_tokens": 43019335.0,
      "step": 20725
    },
    {
      "entropy": 5.413117980957031,
      "epoch": 1.8804426705370103,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046473902217638327,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2092140480875969,
      "num_tokens": 43029264.0,
      "step": 20730
    },
    {
      "entropy": 5.45079345703125,
      "epoch": 1.8808962264150944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046472162969449037,
      "loss": 4.976,
      "mean_token_accuracy": 0.21093496084213256,
      "num_tokens": 43040519.0,
      "step": 20735
    },
    {
      "entropy": 5.5293262004852295,
      "epoch": 1.8813497822931784,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046470423328910594,
      "loss": 4.9915,
      "mean_token_accuracy": 0.21045207381248474,
      "num_tokens": 43051146.0,
      "step": 20740
    },
    {
      "entropy": 5.5510368824005125,
      "epoch": 1.8818033381712627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046468683296058957,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20232977569103242,
      "num_tokens": 43061339.0,
      "step": 20745
    },
    {
      "entropy": 5.464311504364014,
      "epoch": 1.882256894049347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046466942870930126,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21198181957006454,
      "num_tokens": 43072624.0,
      "step": 20750
    },
    {
      "entropy": 5.481714487075806,
      "epoch": 1.8827104499274312,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004646520205356007,
      "loss": 4.9341,
      "mean_token_accuracy": 0.21154951304197311,
      "num_tokens": 43082120.0,
      "step": 20755
    },
    {
      "entropy": 5.555873250961303,
      "epoch": 1.8831640058055152,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046463460843984805,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20004474520683288,
      "num_tokens": 43093202.0,
      "step": 20760
    },
    {
      "entropy": 5.653390550613404,
      "epoch": 1.8836175616835993,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004646171924224034,
      "loss": 5.158,
      "mean_token_accuracy": 0.20311066061258315,
      "num_tokens": 43104801.0,
      "step": 20765
    },
    {
      "entropy": 5.528464365005493,
      "epoch": 1.8840711175616836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046459977248362687,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21161752939224243,
      "num_tokens": 43114855.0,
      "step": 20770
    },
    {
      "entropy": 5.470214557647705,
      "epoch": 1.8845246734397678,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004645823486238788,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20377101004123688,
      "num_tokens": 43124759.0,
      "step": 20775
    },
    {
      "entropy": 5.412673711776733,
      "epoch": 1.884978229317852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004645649208435193,
      "loss": 4.954,
      "mean_token_accuracy": 0.21378117203712463,
      "num_tokens": 43135526.0,
      "step": 20780
    },
    {
      "entropy": 5.500769424438476,
      "epoch": 1.8854317851959361,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046454748914290906,
      "loss": 5.0654,
      "mean_token_accuracy": 0.20297730416059495,
      "num_tokens": 43145605.0,
      "step": 20785
    },
    {
      "entropy": 5.567600202560425,
      "epoch": 1.8858853410740202,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046453005352240835,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20047670155763625,
      "num_tokens": 43156578.0,
      "step": 20790
    },
    {
      "entropy": 5.536592960357666,
      "epoch": 1.8863388969521044,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000464512613982378,
      "loss": 4.9965,
      "mean_token_accuracy": 0.20926790535449982,
      "num_tokens": 43166870.0,
      "step": 20795
    },
    {
      "entropy": 5.484249591827393,
      "epoch": 1.8867924528301887,
      "grad_norm": 1.125,
      "learning_rate": 0.0004644951705231783,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21397003382444382,
      "num_tokens": 43176406.0,
      "step": 20800
    },
    {
      "entropy": 5.597718524932861,
      "epoch": 1.887246008708273,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046447772314517024,
      "loss": 5.1387,
      "mean_token_accuracy": 0.20166807174682616,
      "num_tokens": 43187502.0,
      "step": 20805
    },
    {
      "entropy": 5.442931222915649,
      "epoch": 1.887699564586357,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046446027184871464,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2146605759859085,
      "num_tokens": 43197928.0,
      "step": 20810
    },
    {
      "entropy": 5.443324756622315,
      "epoch": 1.888153120464441,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046444281663417236,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21431098133325577,
      "num_tokens": 43209083.0,
      "step": 20815
    },
    {
      "entropy": 5.493828392028808,
      "epoch": 1.8886066763425253,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004644253575019043,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2034241959452629,
      "num_tokens": 43219089.0,
      "step": 20820
    },
    {
      "entropy": 5.556142091751099,
      "epoch": 1.8890602322206096,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004644078944522716,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20326305776834488,
      "num_tokens": 43229440.0,
      "step": 20825
    },
    {
      "entropy": 5.511632108688355,
      "epoch": 1.8895137880986939,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004643904274856353,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20982567220926285,
      "num_tokens": 43239995.0,
      "step": 20830
    },
    {
      "entropy": 5.459226465225219,
      "epoch": 1.889967343976778,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046437295660235674,
      "loss": 4.95,
      "mean_token_accuracy": 0.21392186135053634,
      "num_tokens": 43249722.0,
      "step": 20835
    },
    {
      "entropy": 5.435919952392578,
      "epoch": 1.8904208998548622,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004643554818027972,
      "loss": 5.0206,
      "mean_token_accuracy": 0.20488353222608566,
      "num_tokens": 43260029.0,
      "step": 20840
    },
    {
      "entropy": 5.513094043731689,
      "epoch": 1.8908744557329462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004643380030873181,
      "loss": 4.9711,
      "mean_token_accuracy": 0.21063131392002105,
      "num_tokens": 43270571.0,
      "step": 20845
    },
    {
      "entropy": 5.51750373840332,
      "epoch": 1.8913280116110305,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004643205204562808,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21767070293426513,
      "num_tokens": 43281310.0,
      "step": 20850
    },
    {
      "entropy": 5.454575872421264,
      "epoch": 1.8917815674891147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046430303391004687,
      "loss": 4.958,
      "mean_token_accuracy": 0.21215643137693405,
      "num_tokens": 43292904.0,
      "step": 20855
    },
    {
      "entropy": 5.480486917495727,
      "epoch": 1.8922351233671988,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004642855434489779,
      "loss": 4.9809,
      "mean_token_accuracy": 0.2019377127289772,
      "num_tokens": 43303745.0,
      "step": 20860
    },
    {
      "entropy": 5.498012447357178,
      "epoch": 1.892688679245283,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046426804907343574,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20734751522541045,
      "num_tokens": 43314652.0,
      "step": 20865
    },
    {
      "entropy": 5.460072231292725,
      "epoch": 1.893142235123367,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046425055078378206,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2202502280473709,
      "num_tokens": 43324242.0,
      "step": 20870
    },
    {
      "entropy": 5.444532680511474,
      "epoch": 1.8935957910014514,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046423304858037885,
      "loss": 4.9844,
      "mean_token_accuracy": 0.2136056199669838,
      "num_tokens": 43334054.0,
      "step": 20875
    },
    {
      "entropy": 5.396278333663941,
      "epoch": 1.8940493468795356,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004642155424635878,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22408942878246307,
      "num_tokens": 43343389.0,
      "step": 20880
    },
    {
      "entropy": 5.434442472457886,
      "epoch": 1.89450290275762,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046419803243377125,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21852685362100602,
      "num_tokens": 43353892.0,
      "step": 20885
    },
    {
      "entropy": 5.411701440811157,
      "epoch": 1.894956458635704,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046418051849129115,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21896405071020125,
      "num_tokens": 43363912.0,
      "step": 20890
    },
    {
      "entropy": 5.470510387420655,
      "epoch": 1.895410014513788,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004641630006365097,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20144977122545243,
      "num_tokens": 43374719.0,
      "step": 20895
    },
    {
      "entropy": 5.526678800582886,
      "epoch": 1.8958635703918723,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046414547886978926,
      "loss": 5.0059,
      "mean_token_accuracy": 0.2012788951396942,
      "num_tokens": 43385549.0,
      "step": 20900
    },
    {
      "entropy": 5.502257776260376,
      "epoch": 1.8963171262699565,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046412795319149204,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21317504495382308,
      "num_tokens": 43395503.0,
      "step": 20905
    },
    {
      "entropy": 5.458832597732544,
      "epoch": 1.8967706821480408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046411042360198053,
      "loss": 5.0212,
      "mean_token_accuracy": 0.20107796043157578,
      "num_tokens": 43406913.0,
      "step": 20910
    },
    {
      "entropy": 5.510247039794922,
      "epoch": 1.8972242380261248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004640928901016173,
      "loss": 4.9305,
      "mean_token_accuracy": 0.20687611997127534,
      "num_tokens": 43416459.0,
      "step": 20915
    },
    {
      "entropy": 5.472636413574219,
      "epoch": 1.8976777939042089,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046407535269076493,
      "loss": 5.0021,
      "mean_token_accuracy": 0.20746291279792786,
      "num_tokens": 43426280.0,
      "step": 20920
    },
    {
      "entropy": 5.510209083557129,
      "epoch": 1.8981313497822931,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046405781136978606,
      "loss": 5.0168,
      "mean_token_accuracy": 0.2058311492204666,
      "num_tokens": 43436676.0,
      "step": 20925
    },
    {
      "entropy": 5.591404819488526,
      "epoch": 1.8985849056603774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046404026613904357,
      "loss": 5.0245,
      "mean_token_accuracy": 0.19956643730401993,
      "num_tokens": 43447019.0,
      "step": 20930
    },
    {
      "entropy": 5.5263426303863525,
      "epoch": 1.8990384615384617,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004640227169989001,
      "loss": 4.9968,
      "mean_token_accuracy": 0.20376960039138795,
      "num_tokens": 43456609.0,
      "step": 20935
    },
    {
      "entropy": 5.4787603378295895,
      "epoch": 1.8994920174165457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004640051639497187,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20301377028226852,
      "num_tokens": 43467616.0,
      "step": 20940
    },
    {
      "entropy": 5.422423601150513,
      "epoch": 1.8999455732946298,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004639876069918624,
      "loss": 5.0073,
      "mean_token_accuracy": 0.20585259795188904,
      "num_tokens": 43478209.0,
      "step": 20945
    },
    {
      "entropy": 5.449855709075928,
      "epoch": 1.900399129172714,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046397004612569416,
      "loss": 4.9031,
      "mean_token_accuracy": 0.20877286344766616,
      "num_tokens": 43487278.0,
      "step": 20950
    },
    {
      "entropy": 5.452475738525391,
      "epoch": 1.9008526850507983,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046395248135157715,
      "loss": 4.928,
      "mean_token_accuracy": 0.21707647144794465,
      "num_tokens": 43497457.0,
      "step": 20955
    },
    {
      "entropy": 5.5127417087554935,
      "epoch": 1.9013062409288826,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046393491266987484,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20186402797698974,
      "num_tokens": 43507761.0,
      "step": 20960
    },
    {
      "entropy": 5.423038196563721,
      "epoch": 1.9017597968069666,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004639173400809502,
      "loss": 4.962,
      "mean_token_accuracy": 0.20437643080949783,
      "num_tokens": 43517198.0,
      "step": 20965
    },
    {
      "entropy": 5.498367214202881,
      "epoch": 1.9022133526850507,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004638997635851669,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20256541669368744,
      "num_tokens": 43527429.0,
      "step": 20970
    },
    {
      "entropy": 5.557362747192383,
      "epoch": 1.902666908563135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046388218318288844,
      "loss": 5.0469,
      "mean_token_accuracy": 0.21011619716882707,
      "num_tokens": 43537867.0,
      "step": 20975
    },
    {
      "entropy": 5.4825639724731445,
      "epoch": 1.9031204644412192,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638645988744782,
      "loss": 4.9794,
      "mean_token_accuracy": 0.207144296169281,
      "num_tokens": 43547284.0,
      "step": 20980
    },
    {
      "entropy": 5.484060525894165,
      "epoch": 1.9035740203193035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004638470106603,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20703898668289183,
      "num_tokens": 43557693.0,
      "step": 20985
    },
    {
      "entropy": 5.538303470611572,
      "epoch": 1.9040275761973875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004638294185407174,
      "loss": 4.9851,
      "mean_token_accuracy": 0.21754462271928787,
      "num_tokens": 43566909.0,
      "step": 20990
    },
    {
      "entropy": 5.51477017402649,
      "epoch": 1.9044811320754715,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638118225160944,
      "loss": 5.0466,
      "mean_token_accuracy": 0.212554831802845,
      "num_tokens": 43576963.0,
      "step": 20995
    },
    {
      "entropy": 5.396061134338379,
      "epoch": 1.9049346879535558,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046379422258679477,
      "loss": 4.9639,
      "mean_token_accuracy": 0.20660515278577804,
      "num_tokens": 43586649.0,
      "step": 21000
    },
    {
      "epoch": 1.9049346879535558,
      "eval_entropy": 5.208128798493433,
      "eval_loss": 5.0293426513671875,
      "eval_mean_token_accuracy": 0.2137755580752144,
      "eval_num_tokens": 43586649.0,
      "eval_runtime": 20.7081,
      "eval_samples_per_second": 1707.542,
      "eval_steps_per_second": 213.443,
      "step": 21000
    },
    {
      "entropy": 5.429493808746338,
      "epoch": 1.90538824383164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004637766187531825,
      "loss": 4.9498,
      "mean_token_accuracy": 0.208621047437191,
      "num_tokens": 43596506.0,
      "step": 21005
    },
    {
      "entropy": 5.4850757122039795,
      "epoch": 1.9058417997097243,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046375901101562163,
      "loss": 5.0317,
      "mean_token_accuracy": 0.20423005819320678,
      "num_tokens": 43607188.0,
      "step": 21010
    },
    {
      "entropy": 5.525533723831177,
      "epoch": 1.9062953555878084,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004637413993744763,
      "loss": 4.9415,
      "mean_token_accuracy": 0.21625768393278122,
      "num_tokens": 43617465.0,
      "step": 21015
    },
    {
      "entropy": 5.480540561676025,
      "epoch": 1.9067489114658926,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046372378383011077,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21178802102804184,
      "num_tokens": 43627692.0,
      "step": 21020
    },
    {
      "entropy": 5.497624731063842,
      "epoch": 1.9072024673439767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046370616438288933,
      "loss": 5.0238,
      "mean_token_accuracy": 0.2034093677997589,
      "num_tokens": 43637350.0,
      "step": 21025
    },
    {
      "entropy": 5.536014795303345,
      "epoch": 1.907656023222061,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046368854103317624,
      "loss": 5.0111,
      "mean_token_accuracy": 0.200029593706131,
      "num_tokens": 43647092.0,
      "step": 21030
    },
    {
      "entropy": 5.49554295539856,
      "epoch": 1.9081095791001452,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046367091378133615,
      "loss": 4.9774,
      "mean_token_accuracy": 0.21101588904857635,
      "num_tokens": 43657237.0,
      "step": 21035
    },
    {
      "entropy": 5.50356969833374,
      "epoch": 1.9085631349782293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004636532826277333,
      "loss": 4.9698,
      "mean_token_accuracy": 0.2119905561208725,
      "num_tokens": 43667966.0,
      "step": 21040
    },
    {
      "entropy": 5.468063116073608,
      "epoch": 1.9090166908563135,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046363564757273264,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20919253230094909,
      "num_tokens": 43678059.0,
      "step": 21045
    },
    {
      "entropy": 5.43682074546814,
      "epoch": 1.9094702467343976,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046361800861669867,
      "loss": 4.985,
      "mean_token_accuracy": 0.21779242306947708,
      "num_tokens": 43687556.0,
      "step": 21050
    },
    {
      "entropy": 5.530478382110596,
      "epoch": 1.9099238026124818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046360036575999623,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20452617555856706,
      "num_tokens": 43697791.0,
      "step": 21055
    },
    {
      "entropy": 5.473403692245483,
      "epoch": 1.9103773584905661,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004635827190029902,
      "loss": 4.9316,
      "mean_token_accuracy": 0.20062480568885804,
      "num_tokens": 43707390.0,
      "step": 21060
    },
    {
      "entropy": 5.557409620285034,
      "epoch": 1.9108309143686504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004635650683460454,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20512668639421464,
      "num_tokens": 43716842.0,
      "step": 21065
    },
    {
      "entropy": 5.451348209381104,
      "epoch": 1.9112844702467344,
      "grad_norm": 1.0625,
      "learning_rate": 0.000463547413789527,
      "loss": 4.97,
      "mean_token_accuracy": 0.20878010243177414,
      "num_tokens": 43726967.0,
      "step": 21070
    },
    {
      "entropy": 5.519096517562867,
      "epoch": 1.9117380261248185,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046352975533379997,
      "loss": 4.9736,
      "mean_token_accuracy": 0.20838417559862138,
      "num_tokens": 43737425.0,
      "step": 21075
    },
    {
      "entropy": 5.388634204864502,
      "epoch": 1.9121915820029027,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046351209297922964,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21248041540384294,
      "num_tokens": 43747395.0,
      "step": 21080
    },
    {
      "entropy": 5.479843807220459,
      "epoch": 1.912645137880987,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046349442672618114,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20361318290233613,
      "num_tokens": 43756348.0,
      "step": 21085
    },
    {
      "entropy": 5.524621629714966,
      "epoch": 1.9130986937590713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046347675657501986,
      "loss": 4.967,
      "mean_token_accuracy": 0.20842225849628448,
      "num_tokens": 43767629.0,
      "step": 21090
    },
    {
      "entropy": 5.49835844039917,
      "epoch": 1.9135522496371553,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004634590825261113,
      "loss": 4.8908,
      "mean_token_accuracy": 0.21746014207601547,
      "num_tokens": 43777398.0,
      "step": 21095
    },
    {
      "entropy": 5.4525140762329105,
      "epoch": 1.9140058055152394,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046344140457982075,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20511661767959594,
      "num_tokens": 43787605.0,
      "step": 21100
    },
    {
      "entropy": 5.451720666885376,
      "epoch": 1.9144593613933236,
      "grad_norm": 1.09375,
      "learning_rate": 0.000463423722736514,
      "loss": 4.9996,
      "mean_token_accuracy": 0.20990871042013168,
      "num_tokens": 43796989.0,
      "step": 21105
    },
    {
      "entropy": 5.515697860717774,
      "epoch": 1.9149129172714079,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004634060369965567,
      "loss": 5.001,
      "mean_token_accuracy": 0.21095823645591735,
      "num_tokens": 43806653.0,
      "step": 21110
    },
    {
      "entropy": 5.456710863113403,
      "epoch": 1.9153664731494922,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046338834736031446,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2085887983441353,
      "num_tokens": 43817152.0,
      "step": 21115
    },
    {
      "entropy": 5.451695442199707,
      "epoch": 1.9158200290275762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046337065382815324,
      "loss": 4.9466,
      "mean_token_accuracy": 0.20445048809051514,
      "num_tokens": 43827341.0,
      "step": 21120
    },
    {
      "entropy": 5.422119998931885,
      "epoch": 1.9162735849056602,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046335295640043887,
      "loss": 4.9415,
      "mean_token_accuracy": 0.21808240562677383,
      "num_tokens": 43837195.0,
      "step": 21125
    },
    {
      "entropy": 5.4731855392456055,
      "epoch": 1.9167271407837445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046333525507753743,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21081597805023194,
      "num_tokens": 43847473.0,
      "step": 21130
    },
    {
      "entropy": 5.4429951190948485,
      "epoch": 1.9171806966618288,
      "grad_norm": 1.125,
      "learning_rate": 0.0004633175498598148,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21239457428455352,
      "num_tokens": 43857590.0,
      "step": 21135
    },
    {
      "entropy": 5.478535318374634,
      "epoch": 1.917634252539913,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046329984074763733,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19819005131721495,
      "num_tokens": 43867085.0,
      "step": 21140
    },
    {
      "entropy": 5.523321866989136,
      "epoch": 1.918087808417997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004632821277413711,
      "loss": 4.9392,
      "mean_token_accuracy": 0.20900174081325532,
      "num_tokens": 43877153.0,
      "step": 21145
    },
    {
      "entropy": 5.550936269760132,
      "epoch": 1.9185413642960811,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004632644108413825,
      "loss": 5.0638,
      "mean_token_accuracy": 0.18979740887880325,
      "num_tokens": 43887977.0,
      "step": 21150
    },
    {
      "entropy": 5.410569953918457,
      "epoch": 1.9189949201741654,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046324669004803803,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2081144392490387,
      "num_tokens": 43898649.0,
      "step": 21155
    },
    {
      "entropy": 5.577789545059204,
      "epoch": 1.9194484760522497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000463228965361704,
      "loss": 5.0785,
      "mean_token_accuracy": 0.20410786122083663,
      "num_tokens": 43909369.0,
      "step": 21160
    },
    {
      "entropy": 5.526481056213379,
      "epoch": 1.919902031930334,
      "grad_norm": 1.125,
      "learning_rate": 0.0004632112367827469,
      "loss": 4.9789,
      "mean_token_accuracy": 0.20510622560977937,
      "num_tokens": 43919941.0,
      "step": 21165
    },
    {
      "entropy": 5.548752641677856,
      "epoch": 1.920355587808418,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004631935043115335,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20475824177265167,
      "num_tokens": 43929710.0,
      "step": 21170
    },
    {
      "entropy": 5.493607759475708,
      "epoch": 1.920809143686502,
      "grad_norm": 1.125,
      "learning_rate": 0.00046317576794843056,
      "loss": 4.9805,
      "mean_token_accuracy": 0.20738983750343323,
      "num_tokens": 43939850.0,
      "step": 21175
    },
    {
      "entropy": 5.476680755615234,
      "epoch": 1.9212626995645863,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046315802769380474,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21504528671503068,
      "num_tokens": 43950067.0,
      "step": 21180
    },
    {
      "entropy": 5.48453049659729,
      "epoch": 1.9217162554426706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004631402835480229,
      "loss": 4.9713,
      "mean_token_accuracy": 0.211235174536705,
      "num_tokens": 43960029.0,
      "step": 21185
    },
    {
      "entropy": 5.503409719467163,
      "epoch": 1.9221698113207548,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046312253551145213,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20770203322172165,
      "num_tokens": 43971221.0,
      "step": 21190
    },
    {
      "entropy": 5.50986795425415,
      "epoch": 1.9226233671988389,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046310478358445935,
      "loss": 5.0254,
      "mean_token_accuracy": 0.2000241219997406,
      "num_tokens": 43981579.0,
      "step": 21195
    },
    {
      "entropy": 5.547817897796631,
      "epoch": 1.9230769230769231,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004630870277674117,
      "loss": 5.1223,
      "mean_token_accuracy": 0.19511819779872894,
      "num_tokens": 43991132.0,
      "step": 21200
    },
    {
      "entropy": 5.463128709793091,
      "epoch": 1.9235304789550072,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004630692680606764,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21112724393606186,
      "num_tokens": 44001255.0,
      "step": 21205
    },
    {
      "entropy": 5.431452322006225,
      "epoch": 1.9239840348330914,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004630515044646207,
      "loss": 4.8268,
      "mean_token_accuracy": 0.21773632168769835,
      "num_tokens": 44011275.0,
      "step": 21210
    },
    {
      "entropy": 5.476786661148071,
      "epoch": 1.9244375907111757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000463033736979612,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20560188591480255,
      "num_tokens": 44022524.0,
      "step": 21215
    },
    {
      "entropy": 5.462142992019653,
      "epoch": 1.9248911465892597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046301596560601767,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2084485411643982,
      "num_tokens": 44033955.0,
      "step": 21220
    },
    {
      "entropy": 5.490705776214599,
      "epoch": 1.925344702467344,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004629981903442052,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20409136563539504,
      "num_tokens": 44045149.0,
      "step": 21225
    },
    {
      "entropy": 5.542683744430542,
      "epoch": 1.925798258345428,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004629804111945424,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21039204746484758,
      "num_tokens": 44055280.0,
      "step": 21230
    },
    {
      "entropy": 5.478211784362793,
      "epoch": 1.9262518142235123,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004629626281573966,
      "loss": 4.948,
      "mean_token_accuracy": 0.21064088344573975,
      "num_tokens": 44065578.0,
      "step": 21235
    },
    {
      "entropy": 5.4070117473602295,
      "epoch": 1.9267053701015966,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004629448412331359,
      "loss": 4.9108,
      "mean_token_accuracy": 0.21519255191087722,
      "num_tokens": 44075454.0,
      "step": 21240
    },
    {
      "entropy": 5.463044929504394,
      "epoch": 1.9271589259796809,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004629270504221279,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20468932688236235,
      "num_tokens": 44085680.0,
      "step": 21245
    },
    {
      "entropy": 5.4293310165405275,
      "epoch": 1.927612481857765,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004629092557247407,
      "loss": 4.93,
      "mean_token_accuracy": 0.20808823704719542,
      "num_tokens": 44097177.0,
      "step": 21250
    },
    {
      "entropy": 5.508694648742676,
      "epoch": 1.928066037735849,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004628914571413422,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2045628771185875,
      "num_tokens": 44107251.0,
      "step": 21255
    },
    {
      "entropy": 5.473277282714844,
      "epoch": 1.9285195936139332,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046287365467230035,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20199953764677048,
      "num_tokens": 44117943.0,
      "step": 21260
    },
    {
      "entropy": 5.545036268234253,
      "epoch": 1.9289731494920175,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046285584831798354,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20755217522382735,
      "num_tokens": 44128608.0,
      "step": 21265
    },
    {
      "entropy": 5.515242004394532,
      "epoch": 1.9294267053701017,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046283803807875996,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2142454758286476,
      "num_tokens": 44138809.0,
      "step": 21270
    },
    {
      "entropy": 5.485807943344116,
      "epoch": 1.9298802612481858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046282022395499785,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20848530381917954,
      "num_tokens": 44148532.0,
      "step": 21275
    },
    {
      "entropy": 5.50192813873291,
      "epoch": 1.9303338171262698,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004628024059470656,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20679432451725005,
      "num_tokens": 44158628.0,
      "step": 21280
    },
    {
      "entropy": 5.485290431976319,
      "epoch": 1.930787373004354,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004627845840553318,
      "loss": 4.8865,
      "mean_token_accuracy": 0.21345238983631135,
      "num_tokens": 44168574.0,
      "step": 21285
    },
    {
      "entropy": 5.54254903793335,
      "epoch": 1.9312409288824384,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046276675828016493,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20906085819005965,
      "num_tokens": 44180207.0,
      "step": 21290
    },
    {
      "entropy": 5.53311505317688,
      "epoch": 1.9316944847605226,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004627489286219336,
      "loss": 5.0867,
      "mean_token_accuracy": 0.19624409228563308,
      "num_tokens": 44190473.0,
      "step": 21295
    },
    {
      "entropy": 5.569922637939453,
      "epoch": 1.9321480406386067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004627310950810066,
      "loss": 5.0354,
      "mean_token_accuracy": 0.20721171349287032,
      "num_tokens": 44201150.0,
      "step": 21300
    },
    {
      "entropy": 5.487321233749389,
      "epoch": 1.9326015965166907,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004627132576577528,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20068553239107131,
      "num_tokens": 44211272.0,
      "step": 21305
    },
    {
      "entropy": 5.535678052902222,
      "epoch": 1.933055152394775,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000462695416352541,
      "loss": 5.0627,
      "mean_token_accuracy": 0.20610422343015672,
      "num_tokens": 44223556.0,
      "step": 21310
    },
    {
      "entropy": 5.493580770492554,
      "epoch": 1.9335087082728593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046267757116574,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21254609525203705,
      "num_tokens": 44233911.0,
      "step": 21315
    },
    {
      "entropy": 5.538025426864624,
      "epoch": 1.9339622641509435,
      "grad_norm": 1.125,
      "learning_rate": 0.0004626597220977191,
      "loss": 5.0341,
      "mean_token_accuracy": 0.2088310554623604,
      "num_tokens": 44244108.0,
      "step": 21320
    },
    {
      "entropy": 5.451167440414428,
      "epoch": 1.9344158200290276,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046264186914884737,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21035722494125367,
      "num_tokens": 44253604.0,
      "step": 21325
    },
    {
      "entropy": 5.480927276611328,
      "epoch": 1.9348693759071116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046262401231949386,
      "loss": 5.0467,
      "mean_token_accuracy": 0.1987640842795372,
      "num_tokens": 44263573.0,
      "step": 21330
    },
    {
      "entropy": 5.4717604637146,
      "epoch": 1.9353229317851959,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046260615161002803,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2070356383919716,
      "num_tokens": 44274071.0,
      "step": 21335
    },
    {
      "entropy": 5.581668424606323,
      "epoch": 1.9357764876632801,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004625882870208192,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2075563758611679,
      "num_tokens": 44283997.0,
      "step": 21340
    },
    {
      "entropy": 5.538790607452393,
      "epoch": 1.9362300435413644,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046257041855223674,
      "loss": 4.983,
      "mean_token_accuracy": 0.20960690528154374,
      "num_tokens": 44295024.0,
      "step": 21345
    },
    {
      "entropy": 5.435490417480469,
      "epoch": 1.9366835994194485,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046255254620465033,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21106469631195068,
      "num_tokens": 44305476.0,
      "step": 21350
    },
    {
      "entropy": 5.450983095169067,
      "epoch": 1.9371371552975325,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046253466997842936,
      "loss": 4.9067,
      "mean_token_accuracy": 0.21374634206295012,
      "num_tokens": 44316101.0,
      "step": 21355
    },
    {
      "entropy": 5.451597499847412,
      "epoch": 1.9375907111756168,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004625167898739436,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2147537052631378,
      "num_tokens": 44325058.0,
      "step": 21360
    },
    {
      "entropy": 5.616130304336548,
      "epoch": 1.938044267053701,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000462498905891563,
      "loss": 5.1526,
      "mean_token_accuracy": 0.19316584020853042,
      "num_tokens": 44336585.0,
      "step": 21365
    },
    {
      "entropy": 5.499198532104492,
      "epoch": 1.9384978229317853,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004624810180316571,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20698606818914414,
      "num_tokens": 44346911.0,
      "step": 21370
    },
    {
      "entropy": 5.42497410774231,
      "epoch": 1.9389513788098693,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046246312629459596,
      "loss": 4.9466,
      "mean_token_accuracy": 0.20535250157117843,
      "num_tokens": 44358040.0,
      "step": 21375
    },
    {
      "entropy": 5.487753295898438,
      "epoch": 1.9394049346879536,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004624452306807497,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2171310856938362,
      "num_tokens": 44368781.0,
      "step": 21380
    },
    {
      "entropy": 5.430509090423584,
      "epoch": 1.9398584905660377,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004624273311904882,
      "loss": 4.9113,
      "mean_token_accuracy": 0.20898612439632416,
      "num_tokens": 44378809.0,
      "step": 21385
    },
    {
      "entropy": 5.545567846298217,
      "epoch": 1.940312046444122,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046240942782418184,
      "loss": 5.0658,
      "mean_token_accuracy": 0.2038144513964653,
      "num_tokens": 44388517.0,
      "step": 21390
    },
    {
      "entropy": 5.586041116714478,
      "epoch": 1.9407656023222062,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004623915205822007,
      "loss": 5.0184,
      "mean_token_accuracy": 0.20314415246248246,
      "num_tokens": 44399364.0,
      "step": 21395
    },
    {
      "entropy": 5.482648944854736,
      "epoch": 1.9412191582002902,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046237360946491514,
      "loss": 4.9268,
      "mean_token_accuracy": 0.20687646716833114,
      "num_tokens": 44409506.0,
      "step": 21400
    },
    {
      "entropy": 5.529890346527099,
      "epoch": 1.9416727140783745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004623556944726956,
      "loss": 5.0125,
      "mean_token_accuracy": 0.207782182097435,
      "num_tokens": 44419446.0,
      "step": 21405
    },
    {
      "entropy": 5.464230966567993,
      "epoch": 1.9421262699564585,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046233777560591256,
      "loss": 4.9596,
      "mean_token_accuracy": 0.2046886771917343,
      "num_tokens": 44429560.0,
      "step": 21410
    },
    {
      "entropy": 5.484513521194458,
      "epoch": 1.9425798258345428,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046231985286493654,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2046727254986763,
      "num_tokens": 44439640.0,
      "step": 21415
    },
    {
      "entropy": 5.5007791996002195,
      "epoch": 1.943033381712627,
      "grad_norm": 1.125,
      "learning_rate": 0.00046230192625013836,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20616423338651657,
      "num_tokens": 44450480.0,
      "step": 21420
    },
    {
      "entropy": 5.431567001342773,
      "epoch": 1.9434869375907113,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046228399576188844,
      "loss": 4.8867,
      "mean_token_accuracy": 0.21243804693222046,
      "num_tokens": 44460633.0,
      "step": 21425
    },
    {
      "entropy": 5.489944505691528,
      "epoch": 1.9439404934687954,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046226606140055784,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2186029702425003,
      "num_tokens": 44470917.0,
      "step": 21430
    },
    {
      "entropy": 5.513890886306763,
      "epoch": 1.9443940493468794,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046224812316651736,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2036394014954567,
      "num_tokens": 44482538.0,
      "step": 21435
    },
    {
      "entropy": 5.5470575332641605,
      "epoch": 1.9448476052249637,
      "grad_norm": 1.03125,
      "learning_rate": 0.000462230181060138,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20244336426258086,
      "num_tokens": 44493518.0,
      "step": 21440
    },
    {
      "entropy": 5.505383920669556,
      "epoch": 1.945301161103048,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046221223508179077,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20661171525716782,
      "num_tokens": 44503685.0,
      "step": 21445
    },
    {
      "entropy": 5.43728928565979,
      "epoch": 1.9457547169811322,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046219428523184685,
      "loss": 4.959,
      "mean_token_accuracy": 0.21058472990989685,
      "num_tokens": 44514015.0,
      "step": 21450
    },
    {
      "entropy": 5.464501571655274,
      "epoch": 1.9462082728592163,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004621763315106774,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20697275698184966,
      "num_tokens": 44525187.0,
      "step": 21455
    },
    {
      "entropy": 5.418845224380493,
      "epoch": 1.9466618287373003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046215837391865364,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21209617406129838,
      "num_tokens": 44535964.0,
      "step": 21460
    },
    {
      "entropy": 5.44310040473938,
      "epoch": 1.9471153846153846,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046214041245614713,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2061736151576042,
      "num_tokens": 44546985.0,
      "step": 21465
    },
    {
      "entropy": 5.521988582611084,
      "epoch": 1.9475689404934688,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004621224471235291,
      "loss": 5.0181,
      "mean_token_accuracy": 0.19978773444890977,
      "num_tokens": 44556888.0,
      "step": 21470
    },
    {
      "entropy": 5.499261426925659,
      "epoch": 1.9480224963715531,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046210447792117114,
      "loss": 4.9447,
      "mean_token_accuracy": 0.2085475966334343,
      "num_tokens": 44566857.0,
      "step": 21475
    },
    {
      "entropy": 5.510692119598389,
      "epoch": 1.9484760522496372,
      "grad_norm": 1.0,
      "learning_rate": 0.000462086504849445,
      "loss": 4.9893,
      "mean_token_accuracy": 0.2052849531173706,
      "num_tokens": 44578152.0,
      "step": 21480
    },
    {
      "entropy": 5.487776136398315,
      "epoch": 1.9489296081277212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004620685279087222,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21359857469797133,
      "num_tokens": 44588814.0,
      "step": 21485
    },
    {
      "entropy": 5.470897579193116,
      "epoch": 1.9493831640058055,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046205054709937474,
      "loss": 4.9479,
      "mean_token_accuracy": 0.20818614214658737,
      "num_tokens": 44599366.0,
      "step": 21490
    },
    {
      "entropy": 5.586637926101685,
      "epoch": 1.9498367198838897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004620325624217741,
      "loss": 5.067,
      "mean_token_accuracy": 0.19677798599004745,
      "num_tokens": 44609962.0,
      "step": 21495
    },
    {
      "entropy": 5.561524391174316,
      "epoch": 1.950290275761974,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004620145738762925,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20589086413383484,
      "num_tokens": 44619917.0,
      "step": 21500
    },
    {
      "entropy": 5.476059675216675,
      "epoch": 1.950743831640058,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004619965814633019,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20783425718545914,
      "num_tokens": 44629723.0,
      "step": 21505
    },
    {
      "entropy": 5.4737201690673825,
      "epoch": 1.951197387518142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004619785851831742,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20820776671171187,
      "num_tokens": 44641011.0,
      "step": 21510
    },
    {
      "entropy": 5.519546842575073,
      "epoch": 1.9516509433962264,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046196058503628176,
      "loss": 4.9442,
      "mean_token_accuracy": 0.20904525220394135,
      "num_tokens": 44651154.0,
      "step": 21515
    },
    {
      "entropy": 5.518677663803101,
      "epoch": 1.9521044992743106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004619425810229968,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20177183747291566,
      "num_tokens": 44661672.0,
      "step": 21520
    },
    {
      "entropy": 5.45524230003357,
      "epoch": 1.952558055152395,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004619245731436917,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21084539890289306,
      "num_tokens": 44671449.0,
      "step": 21525
    },
    {
      "entropy": 5.505606937408447,
      "epoch": 1.953011611030479,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046190656139873863,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2141725167632103,
      "num_tokens": 44681977.0,
      "step": 21530
    },
    {
      "entropy": 5.494058895111084,
      "epoch": 1.953465166908563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004618885457885103,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20989644527435303,
      "num_tokens": 44691946.0,
      "step": 21535
    },
    {
      "entropy": 5.48291506767273,
      "epoch": 1.9539187227866472,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004618705263133793,
      "loss": 4.9977,
      "mean_token_accuracy": 0.2076904892921448,
      "num_tokens": 44702012.0,
      "step": 21540
    },
    {
      "entropy": 5.4927314758300785,
      "epoch": 1.9543722786647315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004618525029737182,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21195246875286103,
      "num_tokens": 44711605.0,
      "step": 21545
    },
    {
      "entropy": 5.545970249176025,
      "epoch": 1.9548258345428158,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046183447576989966,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20986457616090776,
      "num_tokens": 44722765.0,
      "step": 21550
    },
    {
      "entropy": 5.507702732086182,
      "epoch": 1.9552793904208998,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046181644470229656,
      "loss": 4.96,
      "mean_token_accuracy": 0.20797914117574692,
      "num_tokens": 44732082.0,
      "step": 21555
    },
    {
      "entropy": 5.42052583694458,
      "epoch": 1.955732946298984,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004617984097712817,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21292117089033127,
      "num_tokens": 44742054.0,
      "step": 21560
    },
    {
      "entropy": 5.437080812454224,
      "epoch": 1.9561865021770681,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004617803709772282,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21103788465261458,
      "num_tokens": 44752954.0,
      "step": 21565
    },
    {
      "entropy": 5.504434204101562,
      "epoch": 1.9566400580551524,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004617623283205091,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20853299796581268,
      "num_tokens": 44764277.0,
      "step": 21570
    },
    {
      "entropy": 5.542857456207275,
      "epoch": 1.9570936139332367,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046174428180149733,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20178077220916749,
      "num_tokens": 44775429.0,
      "step": 21575
    },
    {
      "entropy": 5.529280710220337,
      "epoch": 1.9575471698113207,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004617262314205663,
      "loss": 5.0213,
      "mean_token_accuracy": 0.20454297661781312,
      "num_tokens": 44786390.0,
      "step": 21580
    },
    {
      "entropy": 5.524914407730103,
      "epoch": 1.958000725689405,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004617081771780893,
      "loss": 4.9919,
      "mean_token_accuracy": 0.2072996199131012,
      "num_tokens": 44796473.0,
      "step": 21585
    },
    {
      "entropy": 5.467043685913086,
      "epoch": 1.958454281567489,
      "grad_norm": 1.0,
      "learning_rate": 0.00046169011907443944,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2105545297265053,
      "num_tokens": 44806379.0,
      "step": 21590
    },
    {
      "entropy": 5.480042934417725,
      "epoch": 1.9589078374455733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004616720571099904,
      "loss": 5.007,
      "mean_token_accuracy": 0.20666434466838837,
      "num_tokens": 44817659.0,
      "step": 21595
    },
    {
      "entropy": 5.5198516845703125,
      "epoch": 1.9593613933236576,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004616539912851157,
      "loss": 4.9348,
      "mean_token_accuracy": 0.2071309357881546,
      "num_tokens": 44827073.0,
      "step": 21600
    },
    {
      "entropy": 5.482797718048095,
      "epoch": 1.9598149492017418,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046163592160018886,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21293228268623351,
      "num_tokens": 44837019.0,
      "step": 21605
    },
    {
      "entropy": 5.511962890625,
      "epoch": 1.9602685050798259,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004616178480555836,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2088865652680397,
      "num_tokens": 44846670.0,
      "step": 21610
    },
    {
      "entropy": 5.403671073913574,
      "epoch": 1.96072206095791,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004615997706516737,
      "loss": 4.8893,
      "mean_token_accuracy": 0.21173563748598098,
      "num_tokens": 44857016.0,
      "step": 21615
    },
    {
      "entropy": 5.539744758605957,
      "epoch": 1.9611756168359942,
      "grad_norm": 1.078125,
      "learning_rate": 0.000461581689388833,
      "loss": 5.0615,
      "mean_token_accuracy": 0.19902532547712326,
      "num_tokens": 44868180.0,
      "step": 21620
    },
    {
      "entropy": 5.4574614524841305,
      "epoch": 1.9616291727140784,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046156360426743536,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21507873386144638,
      "num_tokens": 44877864.0,
      "step": 21625
    },
    {
      "entropy": 5.506433629989624,
      "epoch": 1.9620827285921627,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004615455152878549,
      "loss": 4.9831,
      "mean_token_accuracy": 0.217892561852932,
      "num_tokens": 44887737.0,
      "step": 21630
    },
    {
      "entropy": 5.44438533782959,
      "epoch": 1.9625362844702468,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004615274224504657,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2048384889960289,
      "num_tokens": 44898669.0,
      "step": 21635
    },
    {
      "entropy": 5.5493402004241945,
      "epoch": 1.9629898403483308,
      "grad_norm": 1.125,
      "learning_rate": 0.00046150932575564173,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20997308194637299,
      "num_tokens": 44909074.0,
      "step": 21640
    },
    {
      "entropy": 5.482325506210327,
      "epoch": 1.963443396226415,
      "grad_norm": 1.125,
      "learning_rate": 0.00046149122520375746,
      "loss": 4.945,
      "mean_token_accuracy": 0.20922146886587142,
      "num_tokens": 44919468.0,
      "step": 21645
    },
    {
      "entropy": 5.524408769607544,
      "epoch": 1.9638969521044993,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004614731207951871,
      "loss": 5.052,
      "mean_token_accuracy": 0.19886288195848464,
      "num_tokens": 44930470.0,
      "step": 21650
    },
    {
      "entropy": 5.482134580612183,
      "epoch": 1.9643505079825836,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004614550125303051,
      "loss": 5.0491,
      "mean_token_accuracy": 0.20468457341194152,
      "num_tokens": 44940611.0,
      "step": 21655
    },
    {
      "entropy": 5.507540035247803,
      "epoch": 1.9648040638606676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004614369004094859,
      "loss": 4.9692,
      "mean_token_accuracy": 0.21159935146570205,
      "num_tokens": 44950985.0,
      "step": 21660
    },
    {
      "entropy": 5.530377340316773,
      "epoch": 1.9652576197387517,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046141878443310415,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20684940814971925,
      "num_tokens": 44961858.0,
      "step": 21665
    },
    {
      "entropy": 5.516536521911621,
      "epoch": 1.965711175616836,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046140066460153444,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21126515418291092,
      "num_tokens": 44972048.0,
      "step": 21670
    },
    {
      "entropy": 5.464878225326538,
      "epoch": 1.9661647314949202,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046138254091515136,
      "loss": 4.913,
      "mean_token_accuracy": 0.21505399644374848,
      "num_tokens": 44981835.0,
      "step": 21675
    },
    {
      "entropy": 5.454610919952392,
      "epoch": 1.9666182873730045,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046136441337432996,
      "loss": 4.8908,
      "mean_token_accuracy": 0.21620130985975267,
      "num_tokens": 44992108.0,
      "step": 21680
    },
    {
      "entropy": 5.577081823348999,
      "epoch": 1.9670718432510885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046134628197944496,
      "loss": 5.0521,
      "mean_token_accuracy": 0.19921357035636902,
      "num_tokens": 45002105.0,
      "step": 21685
    },
    {
      "entropy": 5.489145088195801,
      "epoch": 1.9675253991291726,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046132814673087143,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20368688106536864,
      "num_tokens": 45013934.0,
      "step": 21690
    },
    {
      "entropy": 5.4539268016815186,
      "epoch": 1.9679789550072568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004613100076289842,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20767173618078233,
      "num_tokens": 45023851.0,
      "step": 21695
    },
    {
      "entropy": 5.436827516555786,
      "epoch": 1.968432510885341,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004612918646741586,
      "loss": 4.9038,
      "mean_token_accuracy": 0.21096698343753814,
      "num_tokens": 45034140.0,
      "step": 21700
    },
    {
      "entropy": 5.495441913604736,
      "epoch": 1.9688860667634254,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004612737178667698,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20554138571023942,
      "num_tokens": 45044452.0,
      "step": 21705
    },
    {
      "entropy": 5.447922229766846,
      "epoch": 1.9693396226415094,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046125556720719304,
      "loss": 4.993,
      "mean_token_accuracy": 0.20312803685665132,
      "num_tokens": 45053845.0,
      "step": 21710
    },
    {
      "entropy": 5.443632173538208,
      "epoch": 1.9697931785195935,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004612374126958037,
      "loss": 4.9225,
      "mean_token_accuracy": 0.20919681638479232,
      "num_tokens": 45063973.0,
      "step": 21715
    },
    {
      "entropy": 5.529424810409546,
      "epoch": 1.9702467343976777,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004612192543329771,
      "loss": 5.0763,
      "mean_token_accuracy": 0.20683616846799852,
      "num_tokens": 45074879.0,
      "step": 21720
    },
    {
      "entropy": 5.473941850662231,
      "epoch": 1.970700290275762,
      "grad_norm": 1.0625,
      "learning_rate": 0.000461201092119089,
      "loss": 4.9159,
      "mean_token_accuracy": 0.21797876358032225,
      "num_tokens": 45085161.0,
      "step": 21725
    },
    {
      "entropy": 5.530436086654663,
      "epoch": 1.9711538461538463,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004611829260545148,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19993252158164979,
      "num_tokens": 45095862.0,
      "step": 21730
    },
    {
      "entropy": 5.536198997497559,
      "epoch": 1.9716074020319303,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046116475613963026,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20735560208559037,
      "num_tokens": 45106396.0,
      "step": 21735
    },
    {
      "entropy": 5.4965776920318605,
      "epoch": 1.9720609579100146,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046114658237481114,
      "loss": 4.8973,
      "mean_token_accuracy": 0.21421259194612502,
      "num_tokens": 45117178.0,
      "step": 21740
    },
    {
      "entropy": 5.40320782661438,
      "epoch": 1.9725145137880986,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004611284047604332,
      "loss": 4.9178,
      "mean_token_accuracy": 0.21373512893915175,
      "num_tokens": 45126571.0,
      "step": 21745
    },
    {
      "entropy": 5.46913743019104,
      "epoch": 1.9729680696661829,
      "grad_norm": 1.125,
      "learning_rate": 0.00046111022329687246,
      "loss": 5.0184,
      "mean_token_accuracy": 0.20797273367643357,
      "num_tokens": 45137247.0,
      "step": 21750
    },
    {
      "entropy": 5.562497091293335,
      "epoch": 1.9734216255442671,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046109203798450483,
      "loss": 4.978,
      "mean_token_accuracy": 0.20356351286172866,
      "num_tokens": 45147531.0,
      "step": 21755
    },
    {
      "entropy": 5.500815486907959,
      "epoch": 1.9738751814223512,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004610738488237065,
      "loss": 4.9566,
      "mean_token_accuracy": 0.20879297703504562,
      "num_tokens": 45158312.0,
      "step": 21760
    },
    {
      "entropy": 5.43378963470459,
      "epoch": 1.9743287373004355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004610556558148535,
      "loss": 4.9926,
      "mean_token_accuracy": 0.209259033203125,
      "num_tokens": 45169754.0,
      "step": 21765
    },
    {
      "entropy": 5.516285419464111,
      "epoch": 1.9747822931785195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004610374589583222,
      "loss": 5.0213,
      "mean_token_accuracy": 0.2030855596065521,
      "num_tokens": 45180428.0,
      "step": 21770
    },
    {
      "entropy": 5.519256258010865,
      "epoch": 1.9752358490566038,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004610192582544888,
      "loss": 4.9254,
      "mean_token_accuracy": 0.20888968408107758,
      "num_tokens": 45191558.0,
      "step": 21775
    },
    {
      "entropy": 5.449000215530395,
      "epoch": 1.975689404934688,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004610010537037297,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21123051643371582,
      "num_tokens": 45201567.0,
      "step": 21780
    },
    {
      "entropy": 5.528047323226929,
      "epoch": 1.9761429608127723,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004609828453064214,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19533353298902512,
      "num_tokens": 45211359.0,
      "step": 21785
    },
    {
      "entropy": 5.532426500320435,
      "epoch": 1.9765965166908563,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046096463306294057,
      "loss": 4.9687,
      "mean_token_accuracy": 0.20437736958265304,
      "num_tokens": 45222106.0,
      "step": 21790
    },
    {
      "entropy": 5.440474176406861,
      "epoch": 1.9770500725689404,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046094641697366366,
      "loss": 4.9912,
      "mean_token_accuracy": 0.2081102818250656,
      "num_tokens": 45233169.0,
      "step": 21795
    },
    {
      "entropy": 5.549408197402954,
      "epoch": 1.9775036284470247,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004609281970389675,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19811852574348449,
      "num_tokens": 45243627.0,
      "step": 21800
    },
    {
      "entropy": 5.5237767696380615,
      "epoch": 1.977957184325109,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046090997325922887,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21338146477937697,
      "num_tokens": 45253730.0,
      "step": 21805
    },
    {
      "entropy": 5.460017251968384,
      "epoch": 1.9784107402031932,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046089174563482456,
      "loss": 5.036,
      "mean_token_accuracy": 0.2059587851166725,
      "num_tokens": 45263443.0,
      "step": 21810
    },
    {
      "entropy": 5.438510465621948,
      "epoch": 1.9788642960812772,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004608735141661316,
      "loss": 4.9348,
      "mean_token_accuracy": 0.21124192029237748,
      "num_tokens": 45275298.0,
      "step": 21815
    },
    {
      "entropy": 5.509208106994629,
      "epoch": 1.9793178519593613,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046085527885352694,
      "loss": 5.0271,
      "mean_token_accuracy": 0.206145940721035,
      "num_tokens": 45285628.0,
      "step": 21820
    },
    {
      "entropy": 5.471783781051636,
      "epoch": 1.9797714078374455,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004608370396973878,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2130176916718483,
      "num_tokens": 45295559.0,
      "step": 21825
    },
    {
      "entropy": 5.472705936431884,
      "epoch": 1.9802249637155298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046081879669809133,
      "loss": 4.922,
      "mean_token_accuracy": 0.2102159634232521,
      "num_tokens": 45305395.0,
      "step": 21830
    },
    {
      "entropy": 5.47149510383606,
      "epoch": 1.980678519593614,
      "grad_norm": 1.0,
      "learning_rate": 0.0004608005498560148,
      "loss": 4.9451,
      "mean_token_accuracy": 0.21422380656003953,
      "num_tokens": 45316763.0,
      "step": 21835
    },
    {
      "entropy": 5.531671667098999,
      "epoch": 1.9811320754716981,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004607822991715356,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20736559629440307,
      "num_tokens": 45326484.0,
      "step": 21840
    },
    {
      "entropy": 5.558336400985718,
      "epoch": 1.9815856313497822,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000460764044645031,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20397043079137803,
      "num_tokens": 45335792.0,
      "step": 21845
    },
    {
      "entropy": 5.484403228759765,
      "epoch": 1.9820391872278664,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004607457862768787,
      "loss": 4.9441,
      "mean_token_accuracy": 0.20361879467964172,
      "num_tokens": 45347247.0,
      "step": 21850
    },
    {
      "entropy": 5.518598508834839,
      "epoch": 1.9824927431059507,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004607275240674562,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20043198615312577,
      "num_tokens": 45357994.0,
      "step": 21855
    },
    {
      "entropy": 5.414259338378907,
      "epoch": 1.982946298984035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004607092580171411,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20804253220558167,
      "num_tokens": 45368170.0,
      "step": 21860
    },
    {
      "entropy": 5.5120526313781735,
      "epoch": 1.983399854862119,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004606909881263113,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20648445785045624,
      "num_tokens": 45378696.0,
      "step": 21865
    },
    {
      "entropy": 5.560987710952759,
      "epoch": 1.983853410740203,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046067271439534445,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20440033376216887,
      "num_tokens": 45388578.0,
      "step": 21870
    },
    {
      "entropy": 5.5059057712554935,
      "epoch": 1.9843069666182873,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046065443682461864,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20822376161813735,
      "num_tokens": 45398835.0,
      "step": 21875
    },
    {
      "entropy": 5.504858016967773,
      "epoch": 1.9847605224963716,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046063615541451177,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20658082515001297,
      "num_tokens": 45409032.0,
      "step": 21880
    },
    {
      "entropy": 5.561472606658936,
      "epoch": 1.9852140783744558,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046061787016540184,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19895123094320297,
      "num_tokens": 45420661.0,
      "step": 21885
    },
    {
      "entropy": 5.490915203094483,
      "epoch": 1.98566763425254,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046059958107766706,
      "loss": 4.8524,
      "mean_token_accuracy": 0.2225732073187828,
      "num_tokens": 45429751.0,
      "step": 21890
    },
    {
      "entropy": 5.520289754867553,
      "epoch": 1.986121190130624,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004605812881516857,
      "loss": 4.987,
      "mean_token_accuracy": 0.2120147854089737,
      "num_tokens": 45439551.0,
      "step": 21895
    },
    {
      "entropy": 5.385705280303955,
      "epoch": 1.9865747460087082,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046056299138783595,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22473577111959459,
      "num_tokens": 45449543.0,
      "step": 21900
    },
    {
      "entropy": 5.496645212173462,
      "epoch": 1.9870283018867925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004605446907864963,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21451092809438704,
      "num_tokens": 45459069.0,
      "step": 21905
    },
    {
      "entropy": 5.5340653419494625,
      "epoch": 1.9874818577648767,
      "grad_norm": 1.0,
      "learning_rate": 0.0004605263863480452,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19922302067279815,
      "num_tokens": 45468987.0,
      "step": 21910
    },
    {
      "entropy": 5.538180208206176,
      "epoch": 1.9879354136429608,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046050807807286105,
      "loss": 4.9851,
      "mean_token_accuracy": 0.2133519634604454,
      "num_tokens": 45479781.0,
      "step": 21915
    },
    {
      "entropy": 5.585193681716919,
      "epoch": 1.988388969521045,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046048976596132265,
      "loss": 5.1169,
      "mean_token_accuracy": 0.19086560755968093,
      "num_tokens": 45490318.0,
      "step": 21920
    },
    {
      "entropy": 5.483713531494141,
      "epoch": 1.988842525399129,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046047145001380856,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2105340212583542,
      "num_tokens": 45501514.0,
      "step": 21925
    },
    {
      "entropy": 5.4879233837127686,
      "epoch": 1.9892960812772134,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046045313023069764,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21038736253976822,
      "num_tokens": 45511727.0,
      "step": 21930
    },
    {
      "entropy": 5.528943061828613,
      "epoch": 1.9897496371552976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004604348066123688,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20235438793897628,
      "num_tokens": 45521807.0,
      "step": 21935
    },
    {
      "entropy": 5.58302583694458,
      "epoch": 1.9902031930333817,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004604164791592008,
      "loss": 5.0968,
      "mean_token_accuracy": 0.19826555103063584,
      "num_tokens": 45532033.0,
      "step": 21940
    },
    {
      "entropy": 5.517005348205567,
      "epoch": 1.990656748911466,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046039814787157283,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2091621235013008,
      "num_tokens": 45542806.0,
      "step": 21945
    },
    {
      "entropy": 5.467431116104126,
      "epoch": 1.99111030478955,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004603798127498639,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22371553480625153,
      "num_tokens": 45551982.0,
      "step": 21950
    },
    {
      "entropy": 5.500916004180908,
      "epoch": 1.9915638606676342,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004603614737944532,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20723975896835328,
      "num_tokens": 45561998.0,
      "step": 21955
    },
    {
      "entropy": 5.4483404636383055,
      "epoch": 1.9920174165457185,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046034313100571994,
      "loss": 4.9793,
      "mean_token_accuracy": 0.20845555067062377,
      "num_tokens": 45572246.0,
      "step": 21960
    },
    {
      "entropy": 5.557849979400634,
      "epoch": 1.9924709724238028,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004603247843840436,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20728835314512253,
      "num_tokens": 45582251.0,
      "step": 21965
    },
    {
      "entropy": 5.5004191398620605,
      "epoch": 1.9929245283018868,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004603064339298034,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2093706488609314,
      "num_tokens": 45592749.0,
      "step": 21970
    },
    {
      "entropy": 5.464179182052613,
      "epoch": 1.9933780841799709,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004602880796433789,
      "loss": 4.9358,
      "mean_token_accuracy": 0.21098134964704512,
      "num_tokens": 45602412.0,
      "step": 21975
    },
    {
      "entropy": 5.523681259155273,
      "epoch": 1.9938316400580551,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046026972152514976,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20651887357234955,
      "num_tokens": 45612736.0,
      "step": 21980
    },
    {
      "entropy": 5.508471536636352,
      "epoch": 1.9942851959361394,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004602513595754956,
      "loss": 4.995,
      "mean_token_accuracy": 0.20767152905464173,
      "num_tokens": 45622952.0,
      "step": 21985
    },
    {
      "entropy": 5.549241304397583,
      "epoch": 1.9947387518142237,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046023299379479606,
      "loss": 4.9995,
      "mean_token_accuracy": 0.20532257854938507,
      "num_tokens": 45633569.0,
      "step": 21990
    },
    {
      "entropy": 5.458420610427856,
      "epoch": 1.9951923076923077,
      "grad_norm": 1.125,
      "learning_rate": 0.00046021462418343104,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21177952736616135,
      "num_tokens": 45644347.0,
      "step": 21995
    },
    {
      "entropy": 5.515032339096069,
      "epoch": 1.9956458635703918,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004601962507417804,
      "loss": 5.018,
      "mean_token_accuracy": 0.20814317166805268,
      "num_tokens": 45655268.0,
      "step": 22000
    },
    {
      "entropy": 5.488854789733887,
      "epoch": 1.996099419448476,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046017787347022403,
      "loss": 4.9444,
      "mean_token_accuracy": 0.20772619992494584,
      "num_tokens": 45664900.0,
      "step": 22005
    },
    {
      "entropy": 5.526731061935425,
      "epoch": 1.9965529753265603,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046015949236914217,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21357100754976271,
      "num_tokens": 45675252.0,
      "step": 22010
    },
    {
      "entropy": 5.5692308902740475,
      "epoch": 1.9970065312046446,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004601411074389147,
      "loss": 4.9376,
      "mean_token_accuracy": 0.21583616733551025,
      "num_tokens": 45686027.0,
      "step": 22015
    },
    {
      "entropy": 5.4841132164001465,
      "epoch": 1.9974600870827286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000460122718679922,
      "loss": 4.9521,
      "mean_token_accuracy": 0.20540669709444045,
      "num_tokens": 45696436.0,
      "step": 22020
    },
    {
      "entropy": 5.511106491088867,
      "epoch": 1.9979136429608126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004601043260925443,
      "loss": 5.0534,
      "mean_token_accuracy": 0.21106892377138137,
      "num_tokens": 45706509.0,
      "step": 22025
    },
    {
      "entropy": 5.539281511306763,
      "epoch": 1.998367198838897,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046008592967716197,
      "loss": 5.0826,
      "mean_token_accuracy": 0.19876661598682405,
      "num_tokens": 45717933.0,
      "step": 22030
    },
    {
      "entropy": 5.515984582901001,
      "epoch": 1.9988207547169812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004600675294341554,
      "loss": 5.0323,
      "mean_token_accuracy": 0.19881228655576705,
      "num_tokens": 45727966.0,
      "step": 22035
    },
    {
      "entropy": 5.528542995452881,
      "epoch": 1.9992743105950654,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046004912536390526,
      "loss": 5.0017,
      "mean_token_accuracy": 0.20533232837915422,
      "num_tokens": 45737649.0,
      "step": 22040
    },
    {
      "entropy": 5.514850234985351,
      "epoch": 1.9997278664731495,
      "grad_norm": 1.125,
      "learning_rate": 0.000460030717466792,
      "loss": 4.9446,
      "mean_token_accuracy": 0.20659542232751846,
      "num_tokens": 45747561.0,
      "step": 22045
    },
    {
      "entropy": 5.46459813117981,
      "epoch": 2.0001814223512335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004600123057431963,
      "loss": 4.9053,
      "mean_token_accuracy": 0.21400094628334046,
      "num_tokens": 45758183.0,
      "step": 22050
    },
    {
      "entropy": 5.480481052398682,
      "epoch": 2.000634978229318,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000459993890193499,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22108574956655502,
      "num_tokens": 45768006.0,
      "step": 22055
    },
    {
      "entropy": 5.55034327507019,
      "epoch": 2.001088534107402,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004599754708180809,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20299693793058396,
      "num_tokens": 45777981.0,
      "step": 22060
    },
    {
      "entropy": 5.518570852279663,
      "epoch": 2.0015420899854863,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000459957047617323,
      "loss": 4.9216,
      "mean_token_accuracy": 0.20742516666650773,
      "num_tokens": 45789083.0,
      "step": 22065
    },
    {
      "entropy": 5.509594202041626,
      "epoch": 2.0019956458635706,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045993862059160617,
      "loss": 4.8743,
      "mean_token_accuracy": 0.21316053420305253,
      "num_tokens": 45799750.0,
      "step": 22070
    },
    {
      "entropy": 5.464544439315796,
      "epoch": 2.0024492017416544,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004599201897413116,
      "loss": 4.881,
      "mean_token_accuracy": 0.21451300084590913,
      "num_tokens": 45810321.0,
      "step": 22075
    },
    {
      "entropy": 5.412771844863892,
      "epoch": 2.0029027576197387,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045990175506682024,
      "loss": 4.8571,
      "mean_token_accuracy": 0.2219959482550621,
      "num_tokens": 45820581.0,
      "step": 22080
    },
    {
      "entropy": 5.459510803222656,
      "epoch": 2.003356313497823,
      "grad_norm": 1.0,
      "learning_rate": 0.00045988331656851355,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21032775193452835,
      "num_tokens": 45831504.0,
      "step": 22085
    },
    {
      "entropy": 5.509075069427491,
      "epoch": 2.003809869375907,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045986487424677284,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20583331286907197,
      "num_tokens": 45841100.0,
      "step": 22090
    },
    {
      "entropy": 5.528689241409301,
      "epoch": 2.0042634252539915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004598464281019793,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21508843302726746,
      "num_tokens": 45851647.0,
      "step": 22095
    },
    {
      "entropy": 5.456178092956543,
      "epoch": 2.0047169811320753,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045982797813451463,
      "loss": 4.8579,
      "mean_token_accuracy": 0.21951217353343963,
      "num_tokens": 45862995.0,
      "step": 22100
    },
    {
      "entropy": 5.411767911911011,
      "epoch": 2.0051705370101596,
      "grad_norm": 1.125,
      "learning_rate": 0.00045980952434476014,
      "loss": 4.8822,
      "mean_token_accuracy": 0.20999489426612855,
      "num_tokens": 45872360.0,
      "step": 22105
    },
    {
      "entropy": 5.462488794326783,
      "epoch": 2.005624092888244,
      "grad_norm": 1.0,
      "learning_rate": 0.00045979106673309765,
      "loss": 4.959,
      "mean_token_accuracy": 0.20704466104507446,
      "num_tokens": 45883440.0,
      "step": 22110
    },
    {
      "entropy": 5.45153169631958,
      "epoch": 2.006077648766328,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045977260529990883,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2062060907483101,
      "num_tokens": 45894323.0,
      "step": 22115
    },
    {
      "entropy": 5.480880451202393,
      "epoch": 2.0065312046444124,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004597541400455755,
      "loss": 4.9627,
      "mean_token_accuracy": 0.20593496710062026,
      "num_tokens": 45904544.0,
      "step": 22120
    },
    {
      "entropy": 5.526825284957885,
      "epoch": 2.006984760522496,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045973567097047933,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23071294724941255,
      "num_tokens": 45913225.0,
      "step": 22125
    },
    {
      "entropy": 5.4859724044799805,
      "epoch": 2.0074383164005805,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045971719807500253,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21513185799121856,
      "num_tokens": 45923306.0,
      "step": 22130
    },
    {
      "entropy": 5.442527914047242,
      "epoch": 2.0078918722786647,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000459698721359527,
      "loss": 4.9092,
      "mean_token_accuracy": 0.21128465235233307,
      "num_tokens": 45933243.0,
      "step": 22135
    },
    {
      "entropy": 5.437540817260742,
      "epoch": 2.008345428156749,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004596802408244348,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21018754690885544,
      "num_tokens": 45944498.0,
      "step": 22140
    },
    {
      "entropy": 5.4187340259552,
      "epoch": 2.0087989840348333,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004596617564701081,
      "loss": 4.805,
      "mean_token_accuracy": 0.21613098531961442,
      "num_tokens": 45954550.0,
      "step": 22145
    },
    {
      "entropy": 5.540858745574951,
      "epoch": 2.009252539912917,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045964326829692925,
      "loss": 4.9747,
      "mean_token_accuracy": 0.20606791526079177,
      "num_tokens": 45965425.0,
      "step": 22150
    },
    {
      "entropy": 5.443374872207642,
      "epoch": 2.0097060957910013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045962477630528064,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22071807086467743,
      "num_tokens": 45975637.0,
      "step": 22155
    },
    {
      "entropy": 5.432153749465942,
      "epoch": 2.0101596516690856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045960628049554453,
      "loss": 4.9597,
      "mean_token_accuracy": 0.20494724810123444,
      "num_tokens": 45985268.0,
      "step": 22160
    },
    {
      "entropy": 5.440721940994263,
      "epoch": 2.01061320754717,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045958778086810354,
      "loss": 4.8465,
      "mean_token_accuracy": 0.21323048025369645,
      "num_tokens": 45995335.0,
      "step": 22165
    },
    {
      "entropy": 5.484790372848511,
      "epoch": 2.011066763425254,
      "grad_norm": 1.125,
      "learning_rate": 0.0004595692774233402,
      "loss": 4.9267,
      "mean_token_accuracy": 0.20849782526493071,
      "num_tokens": 46006037.0,
      "step": 22170
    },
    {
      "entropy": 5.51012864112854,
      "epoch": 2.011520319303338,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045955077016163713,
      "loss": 4.8565,
      "mean_token_accuracy": 0.21440623104572296,
      "num_tokens": 46017016.0,
      "step": 22175
    },
    {
      "entropy": 5.553492593765259,
      "epoch": 2.0119738751814222,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045953225908337717,
      "loss": 5.0193,
      "mean_token_accuracy": 0.20529192835092544,
      "num_tokens": 46026797.0,
      "step": 22180
    },
    {
      "entropy": 5.427034521102906,
      "epoch": 2.0124274310595065,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045951374418894304,
      "loss": 4.8309,
      "mean_token_accuracy": 0.2171259731054306,
      "num_tokens": 46037799.0,
      "step": 22185
    },
    {
      "entropy": 5.457459783554077,
      "epoch": 2.0128809869375908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045949522547871765,
      "loss": 4.8577,
      "mean_token_accuracy": 0.21180443912744523,
      "num_tokens": 46048617.0,
      "step": 22190
    },
    {
      "entropy": 5.501497077941894,
      "epoch": 2.013334542815675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000459476702953084,
      "loss": 4.921,
      "mean_token_accuracy": 0.21469594538211823,
      "num_tokens": 46060122.0,
      "step": 22195
    },
    {
      "entropy": 5.472378444671631,
      "epoch": 2.013788098693759,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000459458176612425,
      "loss": 4.928,
      "mean_token_accuracy": 0.2141127973794937,
      "num_tokens": 46071488.0,
      "step": 22200
    },
    {
      "entropy": 5.553925323486328,
      "epoch": 2.014241654571843,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045943964645712407,
      "loss": 4.9665,
      "mean_token_accuracy": 0.20472838580608368,
      "num_tokens": 46081954.0,
      "step": 22205
    },
    {
      "entropy": 5.5036547660827635,
      "epoch": 2.0146952104499274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004594211124875642,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2147393763065338,
      "num_tokens": 46092488.0,
      "step": 22210
    },
    {
      "entropy": 5.413253545761108,
      "epoch": 2.0151487663280117,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004594025747041287,
      "loss": 4.8861,
      "mean_token_accuracy": 0.21815194934606552,
      "num_tokens": 46102451.0,
      "step": 22215
    },
    {
      "entropy": 5.441384744644165,
      "epoch": 2.015602322206096,
      "grad_norm": 1.0625,
      "learning_rate": 0.000459384033107201,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21631981134414674,
      "num_tokens": 46113862.0,
      "step": 22220
    },
    {
      "entropy": 5.5733143329620365,
      "epoch": 2.01605587808418,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004593654876971645,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20906862616539001,
      "num_tokens": 46124517.0,
      "step": 22225
    },
    {
      "entropy": 5.4340015888214115,
      "epoch": 2.016509433962264,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004593469384744027,
      "loss": 4.912,
      "mean_token_accuracy": 0.21289127171039582,
      "num_tokens": 46134221.0,
      "step": 22230
    },
    {
      "entropy": 5.480803680419922,
      "epoch": 2.0169629898403483,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045932838543929924,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20713029205799102,
      "num_tokens": 46144512.0,
      "step": 22235
    },
    {
      "entropy": 5.501719570159912,
      "epoch": 2.0174165457184325,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004593098285922378,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2086881324648857,
      "num_tokens": 46154373.0,
      "step": 22240
    },
    {
      "entropy": 5.558524036407471,
      "epoch": 2.017870101596517,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004592912679336021,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2064037799835205,
      "num_tokens": 46164834.0,
      "step": 22245
    },
    {
      "entropy": 5.475487947463989,
      "epoch": 2.018323657474601,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004592727034637761,
      "loss": 4.8636,
      "mean_token_accuracy": 0.21568524986505508,
      "num_tokens": 46174815.0,
      "step": 22250
    },
    {
      "entropy": 5.454462671279908,
      "epoch": 2.018777213352685,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045925413518314363,
      "loss": 4.9069,
      "mean_token_accuracy": 0.21128300875425338,
      "num_tokens": 46185155.0,
      "step": 22255
    },
    {
      "entropy": 5.4956889152526855,
      "epoch": 2.019230769230769,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004592355630920886,
      "loss": 4.94,
      "mean_token_accuracy": 0.2127099871635437,
      "num_tokens": 46194537.0,
      "step": 22260
    },
    {
      "entropy": 5.5741795063018795,
      "epoch": 2.0196843251088534,
      "grad_norm": 1.125,
      "learning_rate": 0.00045921698719099527,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21196166425943375,
      "num_tokens": 46205494.0,
      "step": 22265
    },
    {
      "entropy": 5.429159784317017,
      "epoch": 2.0201378809869377,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045919840748024764,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2178618535399437,
      "num_tokens": 46216399.0,
      "step": 22270
    },
    {
      "entropy": 5.426583003997803,
      "epoch": 2.020591436865022,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045917982396023,
      "loss": 4.8725,
      "mean_token_accuracy": 0.21391150802373887,
      "num_tokens": 46226515.0,
      "step": 22275
    },
    {
      "entropy": 5.542899942398071,
      "epoch": 2.021044992743106,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045916123663132676,
      "loss": 4.964,
      "mean_token_accuracy": 0.20689981877803804,
      "num_tokens": 46236688.0,
      "step": 22280
    },
    {
      "entropy": 5.470803308486938,
      "epoch": 2.02149854862119,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045914264549392215,
      "loss": 4.8932,
      "mean_token_accuracy": 0.21992812007665635,
      "num_tokens": 46247166.0,
      "step": 22285
    },
    {
      "entropy": 5.492167997360229,
      "epoch": 2.0219521044992743,
      "grad_norm": 1.125,
      "learning_rate": 0.0004591240505484007,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21752907186746598,
      "num_tokens": 46258902.0,
      "step": 22290
    },
    {
      "entropy": 5.525707244873047,
      "epoch": 2.0224056603773586,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045910545179514694,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20512161701917647,
      "num_tokens": 46269819.0,
      "step": 22295
    },
    {
      "entropy": 5.5949952602386475,
      "epoch": 2.022859216255443,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004590868492345455,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21553308814764022,
      "num_tokens": 46280715.0,
      "step": 22300
    },
    {
      "entropy": 5.533560085296631,
      "epoch": 2.0233127721335267,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004590682428669813,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20288393795490264,
      "num_tokens": 46291079.0,
      "step": 22305
    },
    {
      "entropy": 5.500043725967407,
      "epoch": 2.023766328011611,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004590496326928387,
      "loss": 4.9173,
      "mean_token_accuracy": 0.213050240278244,
      "num_tokens": 46301362.0,
      "step": 22310
    },
    {
      "entropy": 5.493936061859131,
      "epoch": 2.024219883889695,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045903101871250295,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21889467239379884,
      "num_tokens": 46311820.0,
      "step": 22315
    },
    {
      "entropy": 5.557138156890869,
      "epoch": 2.0246734397677795,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004590124009263587,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21402305066585542,
      "num_tokens": 46321992.0,
      "step": 22320
    },
    {
      "entropy": 5.493464040756225,
      "epoch": 2.0251269956458637,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004589937793347912,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21023651510477065,
      "num_tokens": 46331537.0,
      "step": 22325
    },
    {
      "entropy": 5.493724393844604,
      "epoch": 2.0255805515239476,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045897515393818547,
      "loss": 4.9146,
      "mean_token_accuracy": 0.20679419338703156,
      "num_tokens": 46342214.0,
      "step": 22330
    },
    {
      "entropy": 5.468299293518067,
      "epoch": 2.026034107402032,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004589565247369266,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2111467257142067,
      "num_tokens": 46352471.0,
      "step": 22335
    },
    {
      "entropy": 5.460948801040649,
      "epoch": 2.026487663280116,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004589378917314,
      "loss": 4.94,
      "mean_token_accuracy": 0.21209260672330857,
      "num_tokens": 46363221.0,
      "step": 22340
    },
    {
      "entropy": 5.4629148006439205,
      "epoch": 2.0269412191582004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045891925492199087,
      "loss": 4.8072,
      "mean_token_accuracy": 0.22391250729560852,
      "num_tokens": 46373454.0,
      "step": 22345
    },
    {
      "entropy": 5.595673894882202,
      "epoch": 2.0273947750362846,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004589006143090847,
      "loss": 5.0733,
      "mean_token_accuracy": 0.20614026188850404,
      "num_tokens": 46384403.0,
      "step": 22350
    },
    {
      "entropy": 5.432787895202637,
      "epoch": 2.0278483309143684,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045888196989306697,
      "loss": 4.7967,
      "mean_token_accuracy": 0.22252882421016693,
      "num_tokens": 46393362.0,
      "step": 22355
    },
    {
      "entropy": 5.538713645935059,
      "epoch": 2.0283018867924527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004588633216743232,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2108989030122757,
      "num_tokens": 46405100.0,
      "step": 22360
    },
    {
      "entropy": 5.455027770996094,
      "epoch": 2.028755442670537,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045884466965323914,
      "loss": 4.8502,
      "mean_token_accuracy": 0.21202486008405685,
      "num_tokens": 46415058.0,
      "step": 22365
    },
    {
      "entropy": 5.486629438400269,
      "epoch": 2.0292089985486212,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045882601383020045,
      "loss": 4.9178,
      "mean_token_accuracy": 0.21323165893554688,
      "num_tokens": 46425863.0,
      "step": 22370
    },
    {
      "entropy": 5.498825359344482,
      "epoch": 2.0296625544267055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004588073542055929,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21583594381809235,
      "num_tokens": 46436172.0,
      "step": 22375
    },
    {
      "entropy": 5.526539707183838,
      "epoch": 2.0301161103047893,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045878869077980246,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2084222838282585,
      "num_tokens": 46446717.0,
      "step": 22380
    },
    {
      "entropy": 5.527638721466064,
      "epoch": 2.0305696661828736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045877002355321504,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20316054821014404,
      "num_tokens": 46456944.0,
      "step": 22385
    },
    {
      "entropy": 5.521604347229004,
      "epoch": 2.031023222060958,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004587513525262168,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2060095727443695,
      "num_tokens": 46467231.0,
      "step": 22390
    },
    {
      "entropy": 5.458830881118774,
      "epoch": 2.031476777939042,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045873267769919364,
      "loss": 4.8622,
      "mean_token_accuracy": 0.21016845852136612,
      "num_tokens": 46477107.0,
      "step": 22395
    },
    {
      "entropy": 5.474247550964355,
      "epoch": 2.0319303338171264,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004587139990725319,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21607415080070497,
      "num_tokens": 46487351.0,
      "step": 22400
    },
    {
      "entropy": 5.502768325805664,
      "epoch": 2.0323838896952107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004586953166466179,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21330981850624084,
      "num_tokens": 46497685.0,
      "step": 22405
    },
    {
      "entropy": 5.487029075622559,
      "epoch": 2.0328374455732945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004586766304218379,
      "loss": 4.9355,
      "mean_token_accuracy": 0.20457724779844283,
      "num_tokens": 46507866.0,
      "step": 22410
    },
    {
      "entropy": 5.401696968078613,
      "epoch": 2.0332910014513788,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045865794039857846,
      "loss": 4.8282,
      "mean_token_accuracy": 0.21892595440149307,
      "num_tokens": 46518797.0,
      "step": 22415
    },
    {
      "entropy": 5.48150782585144,
      "epoch": 2.033744557329463,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004586392465772259,
      "loss": 4.9002,
      "mean_token_accuracy": 0.21555282175540924,
      "num_tokens": 46528759.0,
      "step": 22420
    },
    {
      "entropy": 5.560601282119751,
      "epoch": 2.0341981132075473,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045862054895816703,
      "loss": 5.0347,
      "mean_token_accuracy": 0.19790460765361786,
      "num_tokens": 46539014.0,
      "step": 22425
    },
    {
      "entropy": 5.483238792419433,
      "epoch": 2.0346516690856316,
      "grad_norm": 1.125,
      "learning_rate": 0.0004586018475417884,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2139499917626381,
      "num_tokens": 46548856.0,
      "step": 22430
    },
    {
      "entropy": 5.473536729812622,
      "epoch": 2.0351052249637154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004585831423284766,
      "loss": 4.8672,
      "mean_token_accuracy": 0.21415794640779495,
      "num_tokens": 46559164.0,
      "step": 22435
    },
    {
      "entropy": 5.486605215072632,
      "epoch": 2.0355587808417996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004585644333186188,
      "loss": 4.9777,
      "mean_token_accuracy": 0.2078063130378723,
      "num_tokens": 46569793.0,
      "step": 22440
    },
    {
      "entropy": 5.477545547485351,
      "epoch": 2.036012336719884,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045854572051260166,
      "loss": 4.9349,
      "mean_token_accuracy": 0.20827910751104356,
      "num_tokens": 46580257.0,
      "step": 22445
    },
    {
      "entropy": 5.468678188323975,
      "epoch": 2.036465892597968,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045852700391081234,
      "loss": 4.9183,
      "mean_token_accuracy": 0.20818666964769364,
      "num_tokens": 46590297.0,
      "step": 22450
    },
    {
      "entropy": 5.461681509017945,
      "epoch": 2.0369194484760524,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004585082835136378,
      "loss": 4.8845,
      "mean_token_accuracy": 0.2167753607034683,
      "num_tokens": 46600647.0,
      "step": 22455
    },
    {
      "entropy": 5.49242467880249,
      "epoch": 2.0373730043541363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004584895593214652,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21101376116275788,
      "num_tokens": 46610478.0,
      "step": 22460
    },
    {
      "entropy": 5.492039585113526,
      "epoch": 2.0378265602322205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045847083133468164,
      "loss": 4.964,
      "mean_token_accuracy": 0.21080789268016814,
      "num_tokens": 46620159.0,
      "step": 22465
    },
    {
      "entropy": 5.493364334106445,
      "epoch": 2.038280116110305,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045845209955367464,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2189805641770363,
      "num_tokens": 46630555.0,
      "step": 22470
    },
    {
      "entropy": 5.516729116439819,
      "epoch": 2.038733671988389,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045843336397883146,
      "loss": 4.9849,
      "mean_token_accuracy": 0.20775813162326812,
      "num_tokens": 46640188.0,
      "step": 22475
    },
    {
      "entropy": 5.479578590393066,
      "epoch": 2.0391872278664733,
      "grad_norm": 1.125,
      "learning_rate": 0.0004584146246105396,
      "loss": 4.8869,
      "mean_token_accuracy": 0.21095136553049088,
      "num_tokens": 46650284.0,
      "step": 22480
    },
    {
      "entropy": 5.526302719116211,
      "epoch": 2.039640783744557,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045839588144918643,
      "loss": 4.9168,
      "mean_token_accuracy": 0.21497301161289215,
      "num_tokens": 46661466.0,
      "step": 22485
    },
    {
      "entropy": 5.508177757263184,
      "epoch": 2.0400943396226414,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004583771344951598,
      "loss": 4.9175,
      "mean_token_accuracy": 0.20957376956939697,
      "num_tokens": 46671683.0,
      "step": 22490
    },
    {
      "entropy": 5.469556903839111,
      "epoch": 2.0405478955007257,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045835838374884733,
      "loss": 4.9621,
      "mean_token_accuracy": 0.20684552937746048,
      "num_tokens": 46681787.0,
      "step": 22495
    },
    {
      "entropy": 5.48986234664917,
      "epoch": 2.04100145137881,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045833962921063665,
      "loss": 4.8977,
      "mean_token_accuracy": 0.2110169693827629,
      "num_tokens": 46692752.0,
      "step": 22500
    },
    {
      "entropy": 5.506281328201294,
      "epoch": 2.041455007256894,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045832087088091574,
      "loss": 4.9304,
      "mean_token_accuracy": 0.2128895863890648,
      "num_tokens": 46702980.0,
      "step": 22505
    },
    {
      "entropy": 5.4788877964019775,
      "epoch": 2.041908563134978,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045830210876007257,
      "loss": 4.8461,
      "mean_token_accuracy": 0.2124609351158142,
      "num_tokens": 46712281.0,
      "step": 22510
    },
    {
      "entropy": 5.463829183578492,
      "epoch": 2.0423621190130623,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000458283342848495,
      "loss": 4.8865,
      "mean_token_accuracy": 0.21514485031366348,
      "num_tokens": 46722097.0,
      "step": 22515
    },
    {
      "entropy": 5.388540363311767,
      "epoch": 2.0428156748911466,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004582645731465713,
      "loss": 4.8145,
      "mean_token_accuracy": 0.22377265691757203,
      "num_tokens": 46733056.0,
      "step": 22520
    },
    {
      "entropy": 5.39781084060669,
      "epoch": 2.043269230769231,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045824579965468944,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2160139337182045,
      "num_tokens": 46742608.0,
      "step": 22525
    },
    {
      "entropy": 5.436618804931641,
      "epoch": 2.043722786647315,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045822702237323777,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22209740430116653,
      "num_tokens": 46751762.0,
      "step": 22530
    },
    {
      "entropy": 5.444215440750122,
      "epoch": 2.044176342525399,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00045820824130260457,
      "loss": 4.8622,
      "mean_token_accuracy": 0.2185160905122757,
      "num_tokens": 46762252.0,
      "step": 22535
    },
    {
      "entropy": 5.471076488494873,
      "epoch": 2.044629898403483,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004581894564431783,
      "loss": 4.9378,
      "mean_token_accuracy": 0.20716733634471893,
      "num_tokens": 46772172.0,
      "step": 22540
    },
    {
      "entropy": 5.489015817642212,
      "epoch": 2.0450834542815675,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004581706677953473,
      "loss": 4.923,
      "mean_token_accuracy": 0.21070101708173752,
      "num_tokens": 46782894.0,
      "step": 22545
    },
    {
      "entropy": 5.472652006149292,
      "epoch": 2.0455370101596517,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045815187535950034,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21281129419803618,
      "num_tokens": 46793097.0,
      "step": 22550
    },
    {
      "entropy": 5.5373883724212645,
      "epoch": 2.045990566037736,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045813307913602586,
      "loss": 5.027,
      "mean_token_accuracy": 0.1990057036280632,
      "num_tokens": 46804196.0,
      "step": 22555
    },
    {
      "entropy": 5.524840545654297,
      "epoch": 2.04644412191582,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004581142791253126,
      "loss": 4.9132,
      "mean_token_accuracy": 0.20842732787132262,
      "num_tokens": 46813985.0,
      "step": 22560
    },
    {
      "entropy": 5.425367116928101,
      "epoch": 2.046897677793904,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045809547532774943,
      "loss": 4.8283,
      "mean_token_accuracy": 0.20991829484701158,
      "num_tokens": 46824125.0,
      "step": 22565
    },
    {
      "entropy": 5.442494964599609,
      "epoch": 2.0473512336719883,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045807666774372514,
      "loss": 4.8712,
      "mean_token_accuracy": 0.21538679748773576,
      "num_tokens": 46834982.0,
      "step": 22570
    },
    {
      "entropy": 5.47568154335022,
      "epoch": 2.0478047895500726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045805785637362874,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22492958307266236,
      "num_tokens": 46844669.0,
      "step": 22575
    },
    {
      "entropy": 5.484006261825561,
      "epoch": 2.048258345428157,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004580390412178492,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21772030889987945,
      "num_tokens": 46855108.0,
      "step": 22580
    },
    {
      "entropy": 5.465361642837524,
      "epoch": 2.048711901306241,
      "grad_norm": 1.125,
      "learning_rate": 0.00045802022227677566,
      "loss": 4.8518,
      "mean_token_accuracy": 0.21798664778470994,
      "num_tokens": 46864878.0,
      "step": 22585
    },
    {
      "entropy": 5.4708630561828615,
      "epoch": 2.049165457184325,
      "grad_norm": 1.125,
      "learning_rate": 0.00045800139955079727,
      "loss": 4.942,
      "mean_token_accuracy": 0.21026579439640045,
      "num_tokens": 46874916.0,
      "step": 22590
    },
    {
      "entropy": 5.517731666564941,
      "epoch": 2.0496190130624092,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004579825730403033,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20373817533254623,
      "num_tokens": 46885167.0,
      "step": 22595
    },
    {
      "entropy": 5.548223352432251,
      "epoch": 2.0500725689404935,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004579637427456831,
      "loss": 4.9352,
      "mean_token_accuracy": 0.20912650376558303,
      "num_tokens": 46895526.0,
      "step": 22600
    },
    {
      "entropy": 5.44613356590271,
      "epoch": 2.0505261248185778,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045794490866732607,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2139369800686836,
      "num_tokens": 46905426.0,
      "step": 22605
    },
    {
      "entropy": 5.452028369903564,
      "epoch": 2.050979680696662,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004579260708056217,
      "loss": 4.9353,
      "mean_token_accuracy": 0.20866733342409133,
      "num_tokens": 46915728.0,
      "step": 22610
    },
    {
      "entropy": 5.445578145980835,
      "epoch": 2.051433236574746,
      "grad_norm": 1.125,
      "learning_rate": 0.0004579072291609596,
      "loss": 4.8802,
      "mean_token_accuracy": 0.21575091928243637,
      "num_tokens": 46925972.0,
      "step": 22615
    },
    {
      "entropy": 5.48175277709961,
      "epoch": 2.05188679245283,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045788838373372935,
      "loss": 4.9045,
      "mean_token_accuracy": 0.21737487614154816,
      "num_tokens": 46937108.0,
      "step": 22620
    },
    {
      "entropy": 5.448115539550781,
      "epoch": 2.0523403483309144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045786953452432076,
      "loss": 4.8713,
      "mean_token_accuracy": 0.2131539911031723,
      "num_tokens": 46947684.0,
      "step": 22625
    },
    {
      "entropy": 5.543088340759278,
      "epoch": 2.0527939042089987,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004578506815331236,
      "loss": 4.9803,
      "mean_token_accuracy": 0.208010733127594,
      "num_tokens": 46958089.0,
      "step": 22630
    },
    {
      "entropy": 5.48493013381958,
      "epoch": 2.053247460087083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004578318247605277,
      "loss": 4.8793,
      "mean_token_accuracy": 0.2187783345580101,
      "num_tokens": 46968618.0,
      "step": 22635
    },
    {
      "entropy": 5.49419641494751,
      "epoch": 2.0537010159651667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004578129642069231,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21490394771099092,
      "num_tokens": 46978791.0,
      "step": 22640
    },
    {
      "entropy": 5.460851430892944,
      "epoch": 2.054154571843251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004577940998726998,
      "loss": 4.8519,
      "mean_token_accuracy": 0.21549999266862868,
      "num_tokens": 46988706.0,
      "step": 22645
    },
    {
      "entropy": 5.566716861724854,
      "epoch": 2.0546081277213353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045777523175824796,
      "loss": 5.0376,
      "mean_token_accuracy": 0.19761423468589784,
      "num_tokens": 47000140.0,
      "step": 22650
    },
    {
      "entropy": 5.49652214050293,
      "epoch": 2.0550616835994195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004577563598639577,
      "loss": 4.8768,
      "mean_token_accuracy": 0.21478489637374878,
      "num_tokens": 47009773.0,
      "step": 22655
    },
    {
      "entropy": 5.541280889511109,
      "epoch": 2.055515239477504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045773748419021944,
      "loss": 4.9528,
      "mean_token_accuracy": 0.20578388422727584,
      "num_tokens": 47020326.0,
      "step": 22660
    },
    {
      "entropy": 5.4121160984039305,
      "epoch": 2.0559687953555876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045771860473742347,
      "loss": 4.8971,
      "mean_token_accuracy": 0.20739168226718901,
      "num_tokens": 47030410.0,
      "step": 22665
    },
    {
      "entropy": 5.486596632003784,
      "epoch": 2.056422351233672,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045769972150596006,
      "loss": 4.9247,
      "mean_token_accuracy": 0.21051902920007706,
      "num_tokens": 47040546.0,
      "step": 22670
    },
    {
      "entropy": 5.490625810623169,
      "epoch": 2.056875907111756,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045768083449621987,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2249297946691513,
      "num_tokens": 47050210.0,
      "step": 22675
    },
    {
      "entropy": 5.571322011947632,
      "epoch": 2.0573294629898404,
      "grad_norm": 1.125,
      "learning_rate": 0.00045766194370859347,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20076712518930434,
      "num_tokens": 47060641.0,
      "step": 22680
    },
    {
      "entropy": 5.506687164306641,
      "epoch": 2.0577830188679247,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004576430491434716,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21023054271936417,
      "num_tokens": 47070737.0,
      "step": 22685
    },
    {
      "entropy": 5.535109186172486,
      "epoch": 2.0582365747460085,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045762415080124487,
      "loss": 4.9395,
      "mean_token_accuracy": 0.211958584189415,
      "num_tokens": 47080884.0,
      "step": 22690
    },
    {
      "entropy": 5.453709077835083,
      "epoch": 2.058690130624093,
      "grad_norm": 1.125,
      "learning_rate": 0.00045760524868230417,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2075933963060379,
      "num_tokens": 47090865.0,
      "step": 22695
    },
    {
      "entropy": 5.488585186004639,
      "epoch": 2.059143686502177,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004575863427870404,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21432511061429976,
      "num_tokens": 47100570.0,
      "step": 22700
    },
    {
      "entropy": 5.477503395080566,
      "epoch": 2.0595972423802613,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004575674331158446,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2190001592040062,
      "num_tokens": 47111199.0,
      "step": 22705
    },
    {
      "entropy": 5.463322114944458,
      "epoch": 2.0600507982583456,
      "grad_norm": 1.125,
      "learning_rate": 0.0004575485196691076,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21376549303531647,
      "num_tokens": 47121191.0,
      "step": 22710
    },
    {
      "entropy": 5.524882221221924,
      "epoch": 2.0605043541364294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004575296024472209,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21664262413978577,
      "num_tokens": 47131608.0,
      "step": 22715
    },
    {
      "entropy": 5.566353893280029,
      "epoch": 2.0609579100145137,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004575106814505753,
      "loss": 4.9454,
      "mean_token_accuracy": 0.20247257351875306,
      "num_tokens": 47141492.0,
      "step": 22720
    },
    {
      "entropy": 5.465687274932861,
      "epoch": 2.061411465892598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004574917566795624,
      "loss": 4.8077,
      "mean_token_accuracy": 0.2244683638215065,
      "num_tokens": 47152576.0,
      "step": 22725
    },
    {
      "entropy": 5.508010625839233,
      "epoch": 2.061865021770682,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045747282813457345,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2109052613377571,
      "num_tokens": 47163863.0,
      "step": 22730
    },
    {
      "entropy": 5.472030591964722,
      "epoch": 2.0623185776487665,
      "grad_norm": 1.171875,
      "learning_rate": 0.000457453895816,
      "loss": 4.8826,
      "mean_token_accuracy": 0.21333328485488892,
      "num_tokens": 47174140.0,
      "step": 22735
    },
    {
      "entropy": 5.50460844039917,
      "epoch": 2.0627721335268503,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045743495972423346,
      "loss": 4.9525,
      "mean_token_accuracy": 0.209660504758358,
      "num_tokens": 47184708.0,
      "step": 22740
    },
    {
      "entropy": 5.4794152736663815,
      "epoch": 2.0632256894049346,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004574160198596653,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21145080626010895,
      "num_tokens": 47196286.0,
      "step": 22745
    },
    {
      "entropy": 5.394988393783569,
      "epoch": 2.063679245283019,
      "grad_norm": 1.125,
      "learning_rate": 0.00045739707622268757,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2230606883764267,
      "num_tokens": 47206774.0,
      "step": 22750
    },
    {
      "entropy": 5.435655879974365,
      "epoch": 2.064132801161103,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004573781288136917,
      "loss": 4.7897,
      "mean_token_accuracy": 0.22175561040639877,
      "num_tokens": 47217009.0,
      "step": 22755
    },
    {
      "entropy": 5.491229295730591,
      "epoch": 2.0645863570391874,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045735917763306967,
      "loss": 4.9302,
      "mean_token_accuracy": 0.20813517570495604,
      "num_tokens": 47226718.0,
      "step": 22760
    },
    {
      "entropy": 5.442124891281128,
      "epoch": 2.065039912917271,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004573402226812134,
      "loss": 4.8108,
      "mean_token_accuracy": 0.21806750148534776,
      "num_tokens": 47237666.0,
      "step": 22765
    },
    {
      "entropy": 5.477184247970581,
      "epoch": 2.0654934687953554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045732126395851483,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20914516746997833,
      "num_tokens": 47249230.0,
      "step": 22770
    },
    {
      "entropy": 5.457957029342651,
      "epoch": 2.0659470246734397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004573023014653661,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22038664072751998,
      "num_tokens": 47259232.0,
      "step": 22775
    },
    {
      "entropy": 5.473471641540527,
      "epoch": 2.066400580551524,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045728333520215933,
      "loss": 4.9232,
      "mean_token_accuracy": 0.20911433398723603,
      "num_tokens": 47269006.0,
      "step": 22780
    },
    {
      "entropy": 5.531424236297608,
      "epoch": 2.0668541364296082,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004572643651692867,
      "loss": 5.016,
      "mean_token_accuracy": 0.20064657181501389,
      "num_tokens": 47280048.0,
      "step": 22785
    },
    {
      "entropy": 5.4775042057037355,
      "epoch": 2.0673076923076925,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045724539136714063,
      "loss": 4.8474,
      "mean_token_accuracy": 0.21167343258857726,
      "num_tokens": 47290019.0,
      "step": 22790
    },
    {
      "entropy": 5.484529304504394,
      "epoch": 2.0677612481857763,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004572264137961133,
      "loss": 4.8878,
      "mean_token_accuracy": 0.21879397034645082,
      "num_tokens": 47300573.0,
      "step": 22795
    },
    {
      "entropy": 5.461075973510742,
      "epoch": 2.0682148040638606,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045720743245659734,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21219272762537003,
      "num_tokens": 47311795.0,
      "step": 22800
    },
    {
      "entropy": 5.426529455184936,
      "epoch": 2.068668359941945,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004571884473489853,
      "loss": 4.8025,
      "mean_token_accuracy": 0.22498297840356826,
      "num_tokens": 47321768.0,
      "step": 22805
    },
    {
      "entropy": 5.491577863693237,
      "epoch": 2.069121915820029,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004571694584736696,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2167012944817543,
      "num_tokens": 47332662.0,
      "step": 22810
    },
    {
      "entropy": 5.483455753326416,
      "epoch": 2.0695754716981134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004571504658310432,
      "loss": 4.8846,
      "mean_token_accuracy": 0.21651652455329895,
      "num_tokens": 47344355.0,
      "step": 22815
    },
    {
      "entropy": 5.48982720375061,
      "epoch": 2.0700290275761972,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004571314694214987,
      "loss": 4.8651,
      "mean_token_accuracy": 0.21297242790460585,
      "num_tokens": 47354228.0,
      "step": 22820
    },
    {
      "entropy": 5.463857889175415,
      "epoch": 2.0704825834542815,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004571124692454291,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22278206348419188,
      "num_tokens": 47364477.0,
      "step": 22825
    },
    {
      "entropy": 5.5134130954742435,
      "epoch": 2.0709361393323658,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045709346530322715,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2120780035853386,
      "num_tokens": 47375745.0,
      "step": 22830
    },
    {
      "entropy": 5.497640466690063,
      "epoch": 2.07138969521045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004570744575952859,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2155940890312195,
      "num_tokens": 47385341.0,
      "step": 22835
    },
    {
      "entropy": 5.474624919891357,
      "epoch": 2.0718432510885343,
      "grad_norm": 1.125,
      "learning_rate": 0.0004570554461219985,
      "loss": 4.905,
      "mean_token_accuracy": 0.21767914444208145,
      "num_tokens": 47396266.0,
      "step": 22840
    },
    {
      "entropy": 5.512789726257324,
      "epoch": 2.072296806966618,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045703643088375806,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21131039559841155,
      "num_tokens": 47406772.0,
      "step": 22845
    },
    {
      "entropy": 5.516601085662842,
      "epoch": 2.0727503628447024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045701741188095793,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20497196167707443,
      "num_tokens": 47416833.0,
      "step": 22850
    },
    {
      "entropy": 5.445116949081421,
      "epoch": 2.0732039187227866,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004569983891139912,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2162475824356079,
      "num_tokens": 47426954.0,
      "step": 22855
    },
    {
      "entropy": 5.447731828689575,
      "epoch": 2.073657474600871,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045697936258325155,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22146638333797455,
      "num_tokens": 47437520.0,
      "step": 22860
    },
    {
      "entropy": 5.453663635253906,
      "epoch": 2.074111030478955,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045696033228913226,
      "loss": 4.8731,
      "mean_token_accuracy": 0.2159669116139412,
      "num_tokens": 47447977.0,
      "step": 22865
    },
    {
      "entropy": 5.5334813594818115,
      "epoch": 2.074564586357039,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045694129823202684,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2063072517514229,
      "num_tokens": 47458458.0,
      "step": 22870
    },
    {
      "entropy": 5.5341509819030765,
      "epoch": 2.0750181422351233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004569222604123291,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21012024879455565,
      "num_tokens": 47469048.0,
      "step": 22875
    },
    {
      "entropy": 5.516000175476075,
      "epoch": 2.0754716981132075,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004569032188304326,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20815815329551696,
      "num_tokens": 47480353.0,
      "step": 22880
    },
    {
      "entropy": 5.423048639297486,
      "epoch": 2.075925253991292,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004568841734867313,
      "loss": 4.8067,
      "mean_token_accuracy": 0.21712733060121536,
      "num_tokens": 47489951.0,
      "step": 22885
    },
    {
      "entropy": 5.431924533843994,
      "epoch": 2.076378809869376,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004568651243816188,
      "loss": 4.8809,
      "mean_token_accuracy": 0.21959687322378157,
      "num_tokens": 47499015.0,
      "step": 22890
    },
    {
      "entropy": 5.487962532043457,
      "epoch": 2.07683236574746,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045684607151548923,
      "loss": 4.9124,
      "mean_token_accuracy": 0.21256985366344452,
      "num_tokens": 47508694.0,
      "step": 22895
    },
    {
      "entropy": 5.645472764968872,
      "epoch": 2.077285921625544,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045682701488873655,
      "loss": 5.0518,
      "mean_token_accuracy": 0.20268099755048752,
      "num_tokens": 47521336.0,
      "step": 22900
    },
    {
      "entropy": 5.54237585067749,
      "epoch": 2.0777394775036284,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004568079545017549,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20089402943849563,
      "num_tokens": 47531857.0,
      "step": 22905
    },
    {
      "entropy": 5.5057131290435795,
      "epoch": 2.0781930333817127,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004567888903549385,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2058774784207344,
      "num_tokens": 47542224.0,
      "step": 22910
    },
    {
      "entropy": 5.49770302772522,
      "epoch": 2.078646589259797,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004567698224486813,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21274778246879578,
      "num_tokens": 47552370.0,
      "step": 22915
    },
    {
      "entropy": 5.597073554992676,
      "epoch": 2.0791001451378808,
      "grad_norm": 1.125,
      "learning_rate": 0.00045675075078337803,
      "loss": 4.9093,
      "mean_token_accuracy": 0.21762989908456803,
      "num_tokens": 47561997.0,
      "step": 22920
    },
    {
      "entropy": 5.521228837966919,
      "epoch": 2.079553701015965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004567316753594229,
      "loss": 4.9305,
      "mean_token_accuracy": 0.21692138016223908,
      "num_tokens": 47571968.0,
      "step": 22925
    },
    {
      "entropy": 5.503274202346802,
      "epoch": 2.0800072568940493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045671259617721037,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20989677608013152,
      "num_tokens": 47583459.0,
      "step": 22930
    },
    {
      "entropy": 5.4500175476074215,
      "epoch": 2.0804608127721336,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000456693513237135,
      "loss": 4.8654,
      "mean_token_accuracy": 0.21740245819091797,
      "num_tokens": 47594026.0,
      "step": 22935
    },
    {
      "entropy": 5.546015691757202,
      "epoch": 2.080914368650218,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004566744265395916,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2143869146704674,
      "num_tokens": 47605246.0,
      "step": 22940
    },
    {
      "entropy": 5.513857746124268,
      "epoch": 2.081367924528302,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045665533608497467,
      "loss": 4.9908,
      "mean_token_accuracy": 0.20220822542905809,
      "num_tokens": 47616702.0,
      "step": 22945
    },
    {
      "entropy": 5.541510677337646,
      "epoch": 2.081821480406386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045663624187367914,
      "loss": 4.9729,
      "mean_token_accuracy": 0.2058124527335167,
      "num_tokens": 47627616.0,
      "step": 22950
    },
    {
      "entropy": 5.491442155838013,
      "epoch": 2.08227503628447,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004566171439060998,
      "loss": 4.8934,
      "mean_token_accuracy": 0.21114227920770645,
      "num_tokens": 47637287.0,
      "step": 22955
    },
    {
      "entropy": 5.446555089950562,
      "epoch": 2.0827285921625545,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004565980421826317,
      "loss": 4.9271,
      "mean_token_accuracy": 0.20941709578037263,
      "num_tokens": 47647474.0,
      "step": 22960
    },
    {
      "entropy": 5.5255317211151125,
      "epoch": 2.0831821480406387,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004565789367036697,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2142787531018257,
      "num_tokens": 47657036.0,
      "step": 22965
    },
    {
      "entropy": 5.480075216293335,
      "epoch": 2.083635703918723,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004565598274696091,
      "loss": 4.8425,
      "mean_token_accuracy": 0.21526095867156983,
      "num_tokens": 47668026.0,
      "step": 22970
    },
    {
      "entropy": 5.564390563964844,
      "epoch": 2.084089259796807,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045654071448084506,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20536720901727676,
      "num_tokens": 47679508.0,
      "step": 22975
    },
    {
      "entropy": 5.525553560256958,
      "epoch": 2.084542815674891,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045652159773777273,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21098029166460036,
      "num_tokens": 47688474.0,
      "step": 22980
    },
    {
      "entropy": 5.596520280838012,
      "epoch": 2.0849963715529753,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004565024772407875,
      "loss": 5.0933,
      "mean_token_accuracy": 0.19464746564626695,
      "num_tokens": 47698675.0,
      "step": 22985
    },
    {
      "entropy": 5.4251062870025635,
      "epoch": 2.0854499274310596,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004564833529902848,
      "loss": 4.8456,
      "mean_token_accuracy": 0.22778284996747972,
      "num_tokens": 47709214.0,
      "step": 22990
    },
    {
      "entropy": 5.52123646736145,
      "epoch": 2.085903483309144,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045646422498666,
      "loss": 4.9863,
      "mean_token_accuracy": 0.2125869482755661,
      "num_tokens": 47721000.0,
      "step": 22995
    },
    {
      "entropy": 5.458207416534424,
      "epoch": 2.0863570391872277,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045644509323030893,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2120954141020775,
      "num_tokens": 47731218.0,
      "step": 23000
    },
    {
      "entropy": 5.515310430526734,
      "epoch": 2.086810595065312,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045642595772162704,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2121683418750763,
      "num_tokens": 47742339.0,
      "step": 23005
    },
    {
      "entropy": 5.528716516494751,
      "epoch": 2.0872641509433962,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045640681846101014,
      "loss": 4.9717,
      "mean_token_accuracy": 0.213791923224926,
      "num_tokens": 47752858.0,
      "step": 23010
    },
    {
      "entropy": 5.393567991256714,
      "epoch": 2.0877177068214805,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004563876754488539,
      "loss": 4.8539,
      "mean_token_accuracy": 0.21912543922662736,
      "num_tokens": 47763303.0,
      "step": 23015
    },
    {
      "entropy": 5.465265083312988,
      "epoch": 2.0881712626995648,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004563685286855544,
      "loss": 4.8967,
      "mean_token_accuracy": 0.2120954841375351,
      "num_tokens": 47773212.0,
      "step": 23020
    },
    {
      "entropy": 5.570278692245483,
      "epoch": 2.0886248185776486,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004563493781715075,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21023815274238586,
      "num_tokens": 47783376.0,
      "step": 23025
    },
    {
      "entropy": 5.452899408340454,
      "epoch": 2.089078374455733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004563302239071093,
      "loss": 4.9216,
      "mean_token_accuracy": 0.214186891913414,
      "num_tokens": 47794859.0,
      "step": 23030
    },
    {
      "entropy": 5.523024034500122,
      "epoch": 2.089531930333817,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045631106589275576,
      "loss": 4.9326,
      "mean_token_accuracy": 0.2159760430455208,
      "num_tokens": 47804721.0,
      "step": 23035
    },
    {
      "entropy": 5.471039724349976,
      "epoch": 2.0899854862119014,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045629190412884324,
      "loss": 4.8192,
      "mean_token_accuracy": 0.21751789152622222,
      "num_tokens": 47814701.0,
      "step": 23040
    },
    {
      "entropy": 5.470632553100586,
      "epoch": 2.0904390420899857,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045627273861576786,
      "loss": 4.8759,
      "mean_token_accuracy": 0.21244453638792038,
      "num_tokens": 47823941.0,
      "step": 23045
    },
    {
      "entropy": 5.482703065872192,
      "epoch": 2.0908925979680695,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045625356935392614,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21367649734020233,
      "num_tokens": 47835793.0,
      "step": 23050
    },
    {
      "entropy": 5.536644554138183,
      "epoch": 2.0913461538461537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004562343963437144,
      "loss": 4.975,
      "mean_token_accuracy": 0.2123349905014038,
      "num_tokens": 47845677.0,
      "step": 23055
    },
    {
      "entropy": 5.510829067230224,
      "epoch": 2.091799709724238,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004562152195855291,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22265082895755767,
      "num_tokens": 47855611.0,
      "step": 23060
    },
    {
      "entropy": 5.4408656597137455,
      "epoch": 2.0922532656023223,
      "grad_norm": 1.125,
      "learning_rate": 0.0004561960390797669,
      "loss": 4.9331,
      "mean_token_accuracy": 0.21240527480840682,
      "num_tokens": 47865115.0,
      "step": 23065
    },
    {
      "entropy": 5.5532386302948,
      "epoch": 2.0927068214804065,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045617685482682454,
      "loss": 4.9937,
      "mean_token_accuracy": 0.21412605941295623,
      "num_tokens": 47876690.0,
      "step": 23070
    },
    {
      "entropy": 5.487460374832153,
      "epoch": 2.0931603773584904,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004561576668270986,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21145294606685638,
      "num_tokens": 47887709.0,
      "step": 23075
    },
    {
      "entropy": 5.485498762130737,
      "epoch": 2.0936139332365746,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045613847508098587,
      "loss": 4.9158,
      "mean_token_accuracy": 0.21514301300048827,
      "num_tokens": 47898209.0,
      "step": 23080
    },
    {
      "entropy": 5.472861003875733,
      "epoch": 2.094067489114659,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045611927958888335,
      "loss": 4.8214,
      "mean_token_accuracy": 0.22030145972967147,
      "num_tokens": 47907494.0,
      "step": 23085
    },
    {
      "entropy": 5.521265316009521,
      "epoch": 2.094521044992743,
      "grad_norm": 1.125,
      "learning_rate": 0.00045610008035118797,
      "loss": 4.9642,
      "mean_token_accuracy": 0.21299418807029724,
      "num_tokens": 47916923.0,
      "step": 23090
    },
    {
      "entropy": 5.413331365585327,
      "epoch": 2.0949746008708274,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045608087736829687,
      "loss": 4.8406,
      "mean_token_accuracy": 0.22052141726017,
      "num_tokens": 47927043.0,
      "step": 23095
    },
    {
      "entropy": 5.485216951370239,
      "epoch": 2.0954281567489113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000456061670640607,
      "loss": 4.8981,
      "mean_token_accuracy": 0.21329971253871918,
      "num_tokens": 47937296.0,
      "step": 23100
    },
    {
      "entropy": 5.4986968517303465,
      "epoch": 2.0958817126269955,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004560424601685157,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2136102080345154,
      "num_tokens": 47948268.0,
      "step": 23105
    },
    {
      "entropy": 5.514401912689209,
      "epoch": 2.09633526850508,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004560232459524201,
      "loss": 4.8707,
      "mean_token_accuracy": 0.20965292304754257,
      "num_tokens": 47958375.0,
      "step": 23110
    },
    {
      "entropy": 5.521043300628662,
      "epoch": 2.096788824383164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045600402799271766,
      "loss": 4.9751,
      "mean_token_accuracy": 0.2117014080286026,
      "num_tokens": 47968469.0,
      "step": 23115
    },
    {
      "entropy": 5.463395023345948,
      "epoch": 2.0972423802612483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004559848062898058,
      "loss": 4.8447,
      "mean_token_accuracy": 0.22461419105529784,
      "num_tokens": 47979630.0,
      "step": 23120
    },
    {
      "entropy": 5.465185546875,
      "epoch": 2.097695936139332,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004559655808440821,
      "loss": 4.9238,
      "mean_token_accuracy": 0.2127090126276016,
      "num_tokens": 47989828.0,
      "step": 23125
    },
    {
      "entropy": 5.54462947845459,
      "epoch": 2.0981494920174164,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045594635165594404,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2026979297399521,
      "num_tokens": 47999219.0,
      "step": 23130
    },
    {
      "entropy": 5.493430423736572,
      "epoch": 2.0986030478955007,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045592711872578925,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2234855994582176,
      "num_tokens": 48009244.0,
      "step": 23135
    },
    {
      "entropy": 5.51678056716919,
      "epoch": 2.099056603773585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045590788205401557,
      "loss": 4.8705,
      "mean_token_accuracy": 0.21752927750349044,
      "num_tokens": 48019066.0,
      "step": 23140
    },
    {
      "entropy": 5.433136987686157,
      "epoch": 2.099510159651669,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045588864164102086,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2143240123987198,
      "num_tokens": 48028704.0,
      "step": 23145
    },
    {
      "entropy": 5.447130155563355,
      "epoch": 2.0999637155297535,
      "grad_norm": 1.125,
      "learning_rate": 0.00045586939748720284,
      "loss": 4.8949,
      "mean_token_accuracy": 0.21494856029748916,
      "num_tokens": 48039210.0,
      "step": 23150
    },
    {
      "entropy": 5.559927654266358,
      "epoch": 2.1004172714078373,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045585014959295965,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20766939222812653,
      "num_tokens": 48048446.0,
      "step": 23155
    },
    {
      "entropy": 5.490615320205689,
      "epoch": 2.1008708272859216,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045583089795868934,
      "loss": 4.9195,
      "mean_token_accuracy": 0.2178313046693802,
      "num_tokens": 48058880.0,
      "step": 23160
    },
    {
      "entropy": 5.503872108459473,
      "epoch": 2.101324383164006,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004558116425847899,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21973822861909867,
      "num_tokens": 48070224.0,
      "step": 23165
    },
    {
      "entropy": 5.501051902770996,
      "epoch": 2.10177793904209,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004557923834716596,
      "loss": 4.9698,
      "mean_token_accuracy": 0.20269555747509002,
      "num_tokens": 48080914.0,
      "step": 23170
    },
    {
      "entropy": 5.48310923576355,
      "epoch": 2.1022314949201744,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045577312061969677,
      "loss": 4.9457,
      "mean_token_accuracy": 0.2059015601873398,
      "num_tokens": 48090873.0,
      "step": 23175
    },
    {
      "entropy": 5.587416172027588,
      "epoch": 2.102685050798258,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004557538540292997,
      "loss": 4.9521,
      "mean_token_accuracy": 0.21020243167877198,
      "num_tokens": 48101885.0,
      "step": 23180
    },
    {
      "entropy": 5.5352270126342775,
      "epoch": 2.1031386066763424,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004557345837008669,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2115281343460083,
      "num_tokens": 48111472.0,
      "step": 23185
    },
    {
      "entropy": 5.417670631408692,
      "epoch": 2.1035921625544267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045571530963479684,
      "loss": 4.8359,
      "mean_token_accuracy": 0.21877902746200562,
      "num_tokens": 48121351.0,
      "step": 23190
    },
    {
      "entropy": 5.520017051696778,
      "epoch": 2.104045718432511,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045569603183148814,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2263432264328003,
      "num_tokens": 48131104.0,
      "step": 23195
    },
    {
      "entropy": 5.542947053909302,
      "epoch": 2.1044992743105952,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004556767502913395,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2107452467083931,
      "num_tokens": 48142895.0,
      "step": 23200
    },
    {
      "entropy": 5.437056636810302,
      "epoch": 2.104952830188679,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045565746501474946,
      "loss": 4.7981,
      "mean_token_accuracy": 0.2176461786031723,
      "num_tokens": 48153294.0,
      "step": 23205
    },
    {
      "entropy": 5.514990711212159,
      "epoch": 2.1054063860667633,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045563817600211714,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20512441992759706,
      "num_tokens": 48164319.0,
      "step": 23210
    },
    {
      "entropy": 5.490831184387207,
      "epoch": 2.1058599419448476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045561888325384124,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21474379450082778,
      "num_tokens": 48173918.0,
      "step": 23215
    },
    {
      "entropy": 5.557566690444946,
      "epoch": 2.106313497822932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004555995867703209,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2102363407611847,
      "num_tokens": 48184127.0,
      "step": 23220
    },
    {
      "entropy": 5.445851564407349,
      "epoch": 2.106767053701016,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045558028655195504,
      "loss": 4.8803,
      "mean_token_accuracy": 0.21301581859588622,
      "num_tokens": 48194198.0,
      "step": 23225
    },
    {
      "entropy": 5.419101428985596,
      "epoch": 2.1072206095791,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045556098259914293,
      "loss": 4.8899,
      "mean_token_accuracy": 0.21379010528326034,
      "num_tokens": 48204654.0,
      "step": 23230
    },
    {
      "entropy": 5.495003175735474,
      "epoch": 2.1076741654571842,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045554167491228357,
      "loss": 5.0373,
      "mean_token_accuracy": 0.19996090978384018,
      "num_tokens": 48215322.0,
      "step": 23235
    },
    {
      "entropy": 5.508116340637207,
      "epoch": 2.1081277213352685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045552236349177637,
      "loss": 4.8195,
      "mean_token_accuracy": 0.2218370571732521,
      "num_tokens": 48226142.0,
      "step": 23240
    },
    {
      "entropy": 5.443790626525879,
      "epoch": 2.1085812772133528,
      "grad_norm": 1.125,
      "learning_rate": 0.00045550304833802066,
      "loss": 4.8548,
      "mean_token_accuracy": 0.21408893316984176,
      "num_tokens": 48236896.0,
      "step": 23245
    },
    {
      "entropy": 5.401785564422608,
      "epoch": 2.109034833091437,
      "grad_norm": 1.078125,
      "learning_rate": 0.000455483729451416,
      "loss": 4.8876,
      "mean_token_accuracy": 0.21324791759252548,
      "num_tokens": 48247425.0,
      "step": 23250
    },
    {
      "entropy": 5.5107598304748535,
      "epoch": 2.109488388969521,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045546440683236173,
      "loss": 4.9849,
      "mean_token_accuracy": 0.2117615222930908,
      "num_tokens": 48258560.0,
      "step": 23255
    },
    {
      "entropy": 5.616119909286499,
      "epoch": 2.109941944847605,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045544508048125763,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2058547705411911,
      "num_tokens": 48269469.0,
      "step": 23260
    },
    {
      "entropy": 5.5094382762908936,
      "epoch": 2.1103955007256894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004554257503985032,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21174388527870178,
      "num_tokens": 48278847.0,
      "step": 23265
    },
    {
      "entropy": 5.389681243896485,
      "epoch": 2.1108490566037736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045540641658449834,
      "loss": 4.858,
      "mean_token_accuracy": 0.21684678196907042,
      "num_tokens": 48288091.0,
      "step": 23270
    },
    {
      "entropy": 5.444777774810791,
      "epoch": 2.111302612481858,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045538707903964275,
      "loss": 4.8556,
      "mean_token_accuracy": 0.20940592736005784,
      "num_tokens": 48297946.0,
      "step": 23275
    },
    {
      "entropy": 5.469247627258301,
      "epoch": 2.1117561683599417,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045536773776433647,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2152735933661461,
      "num_tokens": 48307650.0,
      "step": 23280
    },
    {
      "entropy": 5.50627703666687,
      "epoch": 2.112209724238026,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045534839275897937,
      "loss": 5.0169,
      "mean_token_accuracy": 0.19798472821712493,
      "num_tokens": 48318824.0,
      "step": 23285
    },
    {
      "entropy": 5.483741569519043,
      "epoch": 2.1126632801161103,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045532904402397157,
      "loss": 4.8767,
      "mean_token_accuracy": 0.21293772011995316,
      "num_tokens": 48329909.0,
      "step": 23290
    },
    {
      "entropy": 5.510102319717407,
      "epoch": 2.1131168359941945,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004553096915597132,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20591017305850984,
      "num_tokens": 48340142.0,
      "step": 23295
    },
    {
      "entropy": 5.480522680282593,
      "epoch": 2.113570391872279,
      "grad_norm": 1.125,
      "learning_rate": 0.0004552903353666044,
      "loss": 4.8929,
      "mean_token_accuracy": 0.21731432378292084,
      "num_tokens": 48350540.0,
      "step": 23300
    },
    {
      "entropy": 5.4842475891113285,
      "epoch": 2.114023947750363,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045527097544504556,
      "loss": 4.8671,
      "mean_token_accuracy": 0.21836074441671371,
      "num_tokens": 48360153.0,
      "step": 23305
    },
    {
      "entropy": 5.451916790008545,
      "epoch": 2.114477503628447,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000455251611795437,
      "loss": 4.9128,
      "mean_token_accuracy": 0.21206988990306855,
      "num_tokens": 48370088.0,
      "step": 23310
    },
    {
      "entropy": 5.500127172470092,
      "epoch": 2.114931059506531,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004552322444181792,
      "loss": 4.9351,
      "mean_token_accuracy": 0.20827945470809936,
      "num_tokens": 48381325.0,
      "step": 23315
    },
    {
      "entropy": 5.499481821060181,
      "epoch": 2.1153846153846154,
      "grad_norm": 1.0,
      "learning_rate": 0.0004552128733136726,
      "loss": 4.9151,
      "mean_token_accuracy": 0.20841754078865052,
      "num_tokens": 48393077.0,
      "step": 23320
    },
    {
      "entropy": 5.478001642227173,
      "epoch": 2.1158381712626997,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004551934984823179,
      "loss": 4.9106,
      "mean_token_accuracy": 0.2194019839167595,
      "num_tokens": 48403260.0,
      "step": 23325
    },
    {
      "entropy": 5.448865509033203,
      "epoch": 2.116291727140784,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004551741199245157,
      "loss": 4.8686,
      "mean_token_accuracy": 0.2162272408604622,
      "num_tokens": 48413616.0,
      "step": 23330
    },
    {
      "entropy": 5.490435171127319,
      "epoch": 2.1167452830188678,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045515473764066685,
      "loss": 4.916,
      "mean_token_accuracy": 0.214474119246006,
      "num_tokens": 48424189.0,
      "step": 23335
    },
    {
      "entropy": 5.494106197357178,
      "epoch": 2.117198838896952,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045513535163117205,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2117234781384468,
      "num_tokens": 48433603.0,
      "step": 23340
    },
    {
      "entropy": 5.529455280303955,
      "epoch": 2.1176523947750363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004551159618964323,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21157184541225432,
      "num_tokens": 48443043.0,
      "step": 23345
    },
    {
      "entropy": 5.436767148971557,
      "epoch": 2.1181059506531206,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004550965684368486,
      "loss": 4.8117,
      "mean_token_accuracy": 0.22824996560811997,
      "num_tokens": 48452765.0,
      "step": 23350
    },
    {
      "entropy": 5.461091947555542,
      "epoch": 2.118559506531205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004550771712528219,
      "loss": 4.8515,
      "mean_token_accuracy": 0.21886809915304184,
      "num_tokens": 48463191.0,
      "step": 23355
    },
    {
      "entropy": 5.473910236358643,
      "epoch": 2.1190130624092887,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004550577703447534,
      "loss": 4.8738,
      "mean_token_accuracy": 0.21538019329309463,
      "num_tokens": 48473280.0,
      "step": 23360
    },
    {
      "entropy": 5.454883003234864,
      "epoch": 2.119466618287373,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045503836571304443,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22145962417125703,
      "num_tokens": 48483465.0,
      "step": 23365
    },
    {
      "entropy": 5.428310871124268,
      "epoch": 2.119920174165457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004550189573580961,
      "loss": 4.8365,
      "mean_token_accuracy": 0.2230508640408516,
      "num_tokens": 48494497.0,
      "step": 23370
    },
    {
      "entropy": 5.466896724700928,
      "epoch": 2.1203737300435415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004549995452803098,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21349977850914,
      "num_tokens": 48504106.0,
      "step": 23375
    },
    {
      "entropy": 5.521496438980103,
      "epoch": 2.1208272859216257,
      "grad_norm": 1.125,
      "learning_rate": 0.0004549801294800872,
      "loss": 4.933,
      "mean_token_accuracy": 0.2103397622704506,
      "num_tokens": 48515729.0,
      "step": 23380
    },
    {
      "entropy": 5.541421270370483,
      "epoch": 2.1212808417997095,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004549607099578295,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20378561019897462,
      "num_tokens": 48526611.0,
      "step": 23385
    },
    {
      "entropy": 5.455163621902466,
      "epoch": 2.121734397677794,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004549412867139386,
      "loss": 4.8612,
      "mean_token_accuracy": 0.2138544961810112,
      "num_tokens": 48537016.0,
      "step": 23390
    },
    {
      "entropy": 5.508710718154907,
      "epoch": 2.122187953555878,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045492185974881587,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21076148003339767,
      "num_tokens": 48546786.0,
      "step": 23395
    },
    {
      "entropy": 5.521090030670166,
      "epoch": 2.1226415094339623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004549024290628634,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21057303696870805,
      "num_tokens": 48556306.0,
      "step": 23400
    },
    {
      "entropy": 5.51818265914917,
      "epoch": 2.1230950653120466,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004548829946564827,
      "loss": 4.9609,
      "mean_token_accuracy": 0.214289091527462,
      "num_tokens": 48566303.0,
      "step": 23405
    },
    {
      "entropy": 5.510607194900513,
      "epoch": 2.1235486211901304,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004548635565300759,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21444592922925948,
      "num_tokens": 48576313.0,
      "step": 23410
    },
    {
      "entropy": 5.446835279464722,
      "epoch": 2.1240021770682147,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004548441146840449,
      "loss": 4.9222,
      "mean_token_accuracy": 0.21519480645656586,
      "num_tokens": 48587133.0,
      "step": 23415
    },
    {
      "entropy": 5.5302201271057125,
      "epoch": 2.124455732946299,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045482466911879184,
      "loss": 5.0401,
      "mean_token_accuracy": 0.207182577252388,
      "num_tokens": 48598524.0,
      "step": 23420
    },
    {
      "entropy": 5.471316528320313,
      "epoch": 2.1249092888243832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045480521983471875,
      "loss": 4.8855,
      "mean_token_accuracy": 0.21422095894813536,
      "num_tokens": 48609870.0,
      "step": 23425
    },
    {
      "entropy": 5.4870758056640625,
      "epoch": 2.1253628447024675,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004547857668322279,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21868045181035994,
      "num_tokens": 48619828.0,
      "step": 23430
    },
    {
      "entropy": 5.471749591827392,
      "epoch": 2.1258164005805513,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004547663101117215,
      "loss": 4.8741,
      "mean_token_accuracy": 0.21767976582050325,
      "num_tokens": 48630361.0,
      "step": 23435
    },
    {
      "entropy": 5.477697134017944,
      "epoch": 2.1262699564586356,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045474684967360216,
      "loss": 4.9188,
      "mean_token_accuracy": 0.20345781445503236,
      "num_tokens": 48641113.0,
      "step": 23440
    },
    {
      "entropy": 5.525275135040284,
      "epoch": 2.12672351233672,
      "grad_norm": 1.0625,
      "learning_rate": 0.000454727385518272,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20239971280097963,
      "num_tokens": 48651777.0,
      "step": 23445
    },
    {
      "entropy": 5.577066802978516,
      "epoch": 2.127177068214804,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004547079176461338,
      "loss": 5.0018,
      "mean_token_accuracy": 0.20499918311834336,
      "num_tokens": 48663237.0,
      "step": 23450
    },
    {
      "entropy": 5.474946737289429,
      "epoch": 2.1276306240928884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004546884460575901,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2172031208872795,
      "num_tokens": 48673932.0,
      "step": 23455
    },
    {
      "entropy": 5.437037086486816,
      "epoch": 2.1280841799709727,
      "grad_norm": 1.125,
      "learning_rate": 0.0004546689707530434,
      "loss": 4.853,
      "mean_token_accuracy": 0.21925485283136367,
      "num_tokens": 48684304.0,
      "step": 23460
    },
    {
      "entropy": 5.439531660079956,
      "epoch": 2.1285377358490565,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045464949173289676,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2139076918363571,
      "num_tokens": 48695569.0,
      "step": 23465
    },
    {
      "entropy": 5.412794542312622,
      "epoch": 2.1289912917271407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004546300089975528,
      "loss": 4.8958,
      "mean_token_accuracy": 0.21566626578569412,
      "num_tokens": 48706550.0,
      "step": 23470
    },
    {
      "entropy": 5.524947643280029,
      "epoch": 2.129444847605225,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045461052254741445,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20545850098133087,
      "num_tokens": 48716852.0,
      "step": 23475
    },
    {
      "entropy": 5.450227499008179,
      "epoch": 2.1298984034833093,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045459103238288474,
      "loss": 4.8514,
      "mean_token_accuracy": 0.21342543959617616,
      "num_tokens": 48726607.0,
      "step": 23480
    },
    {
      "entropy": 5.509646558761597,
      "epoch": 2.130351959361393,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004545715385043667,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20278929620981218,
      "num_tokens": 48738011.0,
      "step": 23485
    },
    {
      "entropy": 5.488582706451416,
      "epoch": 2.1308055152394774,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004545520409122635,
      "loss": 4.9358,
      "mean_token_accuracy": 0.20685756504535674,
      "num_tokens": 48747899.0,
      "step": 23490
    },
    {
      "entropy": 5.549203586578369,
      "epoch": 2.1312590711175616,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004545325396069783,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21067034006118773,
      "num_tokens": 48758719.0,
      "step": 23495
    },
    {
      "entropy": 5.53995418548584,
      "epoch": 2.131712626995646,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004545130345889145,
      "loss": 4.969,
      "mean_token_accuracy": 0.21649544388055803,
      "num_tokens": 48769677.0,
      "step": 23500
    },
    {
      "entropy": 5.541886806488037,
      "epoch": 2.13216618287373,
      "grad_norm": 1.375,
      "learning_rate": 0.0004544935258584754,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21052988171577453,
      "num_tokens": 48778440.0,
      "step": 23505
    },
    {
      "entropy": 5.552027034759521,
      "epoch": 2.1326197387518144,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004544740134160644,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20672980844974517,
      "num_tokens": 48788328.0,
      "step": 23510
    },
    {
      "entropy": 5.54856309890747,
      "epoch": 2.1330732946298983,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004544544972620851,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2093269035220146,
      "num_tokens": 48799218.0,
      "step": 23515
    },
    {
      "entropy": 5.554470872879028,
      "epoch": 2.1335268505079825,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004544349773969411,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2045355036854744,
      "num_tokens": 48809648.0,
      "step": 23520
    },
    {
      "entropy": 5.530418586730957,
      "epoch": 2.133980406386067,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004544154538210359,
      "loss": 4.9471,
      "mean_token_accuracy": 0.20711568892002105,
      "num_tokens": 48819965.0,
      "step": 23525
    },
    {
      "entropy": 5.586858081817627,
      "epoch": 2.134433962264151,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004543959265347735,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21295797675848008,
      "num_tokens": 48829862.0,
      "step": 23530
    },
    {
      "entropy": 5.5078051567077635,
      "epoch": 2.1348875181422353,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045437639553855757,
      "loss": 4.9089,
      "mean_token_accuracy": 0.21560557335615158,
      "num_tokens": 48840551.0,
      "step": 23535
    },
    {
      "entropy": 5.44795503616333,
      "epoch": 2.135341074020319,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004543568608327921,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22059373408555985,
      "num_tokens": 48851032.0,
      "step": 23540
    },
    {
      "entropy": 5.473871088027954,
      "epoch": 2.1357946298984034,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045433732241788105,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20635011792182922,
      "num_tokens": 48861773.0,
      "step": 23545
    },
    {
      "entropy": 5.550565385818482,
      "epoch": 2.1362481857764877,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045431778029422846,
      "loss": 5.04,
      "mean_token_accuracy": 0.20750613361597062,
      "num_tokens": 48872430.0,
      "step": 23550
    },
    {
      "entropy": 5.51609959602356,
      "epoch": 2.136701741654572,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004542982344622385,
      "loss": 4.8288,
      "mean_token_accuracy": 0.21848976463079453,
      "num_tokens": 48882575.0,
      "step": 23555
    },
    {
      "entropy": 5.476479244232178,
      "epoch": 2.137155297532656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045427868492231533,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2100328490138054,
      "num_tokens": 48893337.0,
      "step": 23560
    },
    {
      "entropy": 5.574572038650513,
      "epoch": 2.13760885341074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045425913167486333,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20403328388929368,
      "num_tokens": 48904675.0,
      "step": 23565
    },
    {
      "entropy": 5.464079999923706,
      "epoch": 2.1380624092888243,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045423957472028677,
      "loss": 4.8678,
      "mean_token_accuracy": 0.22486539483070372,
      "num_tokens": 48914831.0,
      "step": 23570
    },
    {
      "entropy": 5.417195558547974,
      "epoch": 2.1385159651669086,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004542200140589901,
      "loss": 4.8339,
      "mean_token_accuracy": 0.21685407757759095,
      "num_tokens": 48924574.0,
      "step": 23575
    },
    {
      "entropy": 5.524708843231201,
      "epoch": 2.138969521044993,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045420044969137787,
      "loss": 4.9849,
      "mean_token_accuracy": 0.2117667317390442,
      "num_tokens": 48933921.0,
      "step": 23580
    },
    {
      "entropy": 5.47751293182373,
      "epoch": 2.139423076923077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004541808816178547,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21422084122896196,
      "num_tokens": 48944571.0,
      "step": 23585
    },
    {
      "entropy": 5.484706926345825,
      "epoch": 2.139876632801161,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045416130983882524,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21103066205978394,
      "num_tokens": 48954569.0,
      "step": 23590
    },
    {
      "entropy": 5.497363519668579,
      "epoch": 2.140330188679245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045414173435469427,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21008778512477874,
      "num_tokens": 48965886.0,
      "step": 23595
    },
    {
      "entropy": 5.53784966468811,
      "epoch": 2.1407837445573294,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045412215516586654,
      "loss": 4.9377,
      "mean_token_accuracy": 0.20866110622882844,
      "num_tokens": 48976219.0,
      "step": 23600
    },
    {
      "entropy": 5.45023102760315,
      "epoch": 2.1412373004354137,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000454102572272747,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2103630483150482,
      "num_tokens": 48986322.0,
      "step": 23605
    },
    {
      "entropy": 5.492270517349243,
      "epoch": 2.141690856313498,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004540829856757406,
      "loss": 4.8627,
      "mean_token_accuracy": 0.21788825392723082,
      "num_tokens": 48996297.0,
      "step": 23610
    },
    {
      "entropy": 5.483014345169067,
      "epoch": 2.142144412191582,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004540633953752525,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21615786999464034,
      "num_tokens": 49006165.0,
      "step": 23615
    },
    {
      "entropy": 5.469926595687866,
      "epoch": 2.142597968069666,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004540438013716876,
      "loss": 4.8874,
      "mean_token_accuracy": 0.20500117987394334,
      "num_tokens": 49017278.0,
      "step": 23620
    },
    {
      "entropy": 5.42211012840271,
      "epoch": 2.1430515239477503,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004540242036654514,
      "loss": 4.847,
      "mean_token_accuracy": 0.21552079766988755,
      "num_tokens": 49027088.0,
      "step": 23625
    },
    {
      "entropy": 5.4985424995422365,
      "epoch": 2.1435050798258346,
      "grad_norm": 1.0,
      "learning_rate": 0.000454004602256949,
      "loss": 4.8575,
      "mean_token_accuracy": 0.21263304799795152,
      "num_tokens": 49038156.0,
      "step": 23630
    },
    {
      "entropy": 5.556272077560425,
      "epoch": 2.143958635703919,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004539849971465858,
      "loss": 4.9598,
      "mean_token_accuracy": 0.20618132054805755,
      "num_tokens": 49047480.0,
      "step": 23635
    },
    {
      "entropy": 5.541137933731079,
      "epoch": 2.1444121915820027,
      "grad_norm": 1.125,
      "learning_rate": 0.00045396538833476725,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2112905576825142,
      "num_tokens": 49057373.0,
      "step": 23640
    },
    {
      "entropy": 5.4651641845703125,
      "epoch": 2.144865747460087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004539457758218989,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2085328996181488,
      "num_tokens": 49068380.0,
      "step": 23645
    },
    {
      "entropy": 5.452646493911743,
      "epoch": 2.1453193033381712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004539261596083863,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21304250061511992,
      "num_tokens": 49078646.0,
      "step": 23650
    },
    {
      "entropy": 5.5052141666412355,
      "epoch": 2.1457728592162555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004539065396946351,
      "loss": 4.8538,
      "mean_token_accuracy": 0.22275823056697847,
      "num_tokens": 49089281.0,
      "step": 23655
    },
    {
      "entropy": 5.523444175720215,
      "epoch": 2.1462264150943398,
      "grad_norm": 1.125,
      "learning_rate": 0.0004538869160810511,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2068428725004196,
      "num_tokens": 49099514.0,
      "step": 23660
    },
    {
      "entropy": 5.516108751296997,
      "epoch": 2.146679970972424,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045386728876804,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20855603218078614,
      "num_tokens": 49110247.0,
      "step": 23665
    },
    {
      "entropy": 5.522738456726074,
      "epoch": 2.147133526850508,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045384765775600787,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20599158853292465,
      "num_tokens": 49122492.0,
      "step": 23670
    },
    {
      "entropy": 5.504331350326538,
      "epoch": 2.147587082728592,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004538280230453606,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21072427928447723,
      "num_tokens": 49132395.0,
      "step": 23675
    },
    {
      "entropy": 5.546470785140992,
      "epoch": 2.1480406386066764,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004538083846365042,
      "loss": 4.9703,
      "mean_token_accuracy": 0.21081093102693557,
      "num_tokens": 49142909.0,
      "step": 23680
    },
    {
      "entropy": 5.46927399635315,
      "epoch": 2.1484941944847606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004537887425298448,
      "loss": 4.9156,
      "mean_token_accuracy": 0.20832701176404952,
      "num_tokens": 49152933.0,
      "step": 23685
    },
    {
      "entropy": 5.455355453491211,
      "epoch": 2.1489477503628445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045376909672578875,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21210006177425383,
      "num_tokens": 49163595.0,
      "step": 23690
    },
    {
      "entropy": 5.4438024997711185,
      "epoch": 2.1494013062409287,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004537494472247422,
      "loss": 4.8838,
      "mean_token_accuracy": 0.21796485781669617,
      "num_tokens": 49174664.0,
      "step": 23695
    },
    {
      "entropy": 5.539899635314941,
      "epoch": 2.149854862119013,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004537297940271115,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21209616959095,
      "num_tokens": 49185105.0,
      "step": 23700
    },
    {
      "entropy": 5.454533100128174,
      "epoch": 2.1503084179970973,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004537101371333031,
      "loss": 4.8728,
      "mean_token_accuracy": 0.21711899489164352,
      "num_tokens": 49194679.0,
      "step": 23705
    },
    {
      "entropy": 5.44823055267334,
      "epoch": 2.1507619738751815,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045369047654372356,
      "loss": 4.8317,
      "mean_token_accuracy": 0.21542900204658508,
      "num_tokens": 49204563.0,
      "step": 23710
    },
    {
      "entropy": 5.5081446170806885,
      "epoch": 2.151215529753266,
      "grad_norm": 1.125,
      "learning_rate": 0.0004536708122587795,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21585543751716613,
      "num_tokens": 49213759.0,
      "step": 23715
    },
    {
      "entropy": 5.450832748413086,
      "epoch": 2.1516690856313496,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004536511442788774,
      "loss": 4.8247,
      "mean_token_accuracy": 0.2188010811805725,
      "num_tokens": 49223604.0,
      "step": 23720
    },
    {
      "entropy": 5.444315719604492,
      "epoch": 2.152122641509434,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004536314726044241,
      "loss": 4.9208,
      "mean_token_accuracy": 0.21494246274232864,
      "num_tokens": 49234000.0,
      "step": 23725
    },
    {
      "entropy": 5.480324411392212,
      "epoch": 2.152576197387518,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045361179723582654,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21288597881793975,
      "num_tokens": 49244874.0,
      "step": 23730
    },
    {
      "entropy": 5.606625080108643,
      "epoch": 2.1530297532656024,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004535921181734914,
      "loss": 5.078,
      "mean_token_accuracy": 0.20926969796419143,
      "num_tokens": 49255495.0,
      "step": 23735
    },
    {
      "entropy": 5.508075857162476,
      "epoch": 2.1534833091436867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004535724354178257,
      "loss": 4.8806,
      "mean_token_accuracy": 0.20972933918237685,
      "num_tokens": 49265357.0,
      "step": 23740
    },
    {
      "entropy": 5.516286039352417,
      "epoch": 2.1539368650217705,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045355274896923664,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2041096121072769,
      "num_tokens": 49276396.0,
      "step": 23745
    },
    {
      "entropy": 5.494228839874268,
      "epoch": 2.1543904208998548,
      "grad_norm": 1.125,
      "learning_rate": 0.0004535330588281312,
      "loss": 4.8502,
      "mean_token_accuracy": 0.21652241349220275,
      "num_tokens": 49285925.0,
      "step": 23750
    },
    {
      "entropy": 5.450982713699341,
      "epoch": 2.154843976777939,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004535133649949166,
      "loss": 4.8866,
      "mean_token_accuracy": 0.21966630816459656,
      "num_tokens": 49296493.0,
      "step": 23755
    },
    {
      "entropy": 5.451837348937988,
      "epoch": 2.1552975326560233,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004534936674700001,
      "loss": 4.8677,
      "mean_token_accuracy": 0.21596649587154387,
      "num_tokens": 49307273.0,
      "step": 23760
    },
    {
      "entropy": 5.4932938575744625,
      "epoch": 2.1557510885341076,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045347396625378914,
      "loss": 4.8981,
      "mean_token_accuracy": 0.21853408664464952,
      "num_tokens": 49317055.0,
      "step": 23765
    },
    {
      "entropy": 5.5065773010253904,
      "epoch": 2.1562046444121914,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045345426134669103,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2129323661327362,
      "num_tokens": 49327420.0,
      "step": 23770
    },
    {
      "entropy": 5.512909412384033,
      "epoch": 2.1566582002902757,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045343455274911325,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2095880091190338,
      "num_tokens": 49338323.0,
      "step": 23775
    },
    {
      "entropy": 5.509288549423218,
      "epoch": 2.15711175616836,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004534148404614635,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21000067591667176,
      "num_tokens": 49348364.0,
      "step": 23780
    },
    {
      "entropy": 5.561091899871826,
      "epoch": 2.157565312046444,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045339512448414943,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2130022242665291,
      "num_tokens": 49358274.0,
      "step": 23785
    },
    {
      "entropy": 5.4732283592224125,
      "epoch": 2.1580188679245285,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004533754048175787,
      "loss": 4.85,
      "mean_token_accuracy": 0.20857401341199874,
      "num_tokens": 49368651.0,
      "step": 23790
    },
    {
      "entropy": 5.449948358535766,
      "epoch": 2.1584724238026123,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004533556814621591,
      "loss": 4.8791,
      "mean_token_accuracy": 0.20844231992959977,
      "num_tokens": 49378363.0,
      "step": 23795
    },
    {
      "entropy": 5.492123746871949,
      "epoch": 2.1589259796806965,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045333595441829865,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21369218677282334,
      "num_tokens": 49388241.0,
      "step": 23800
    },
    {
      "entropy": 5.482124328613281,
      "epoch": 2.159379535558781,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004533162236864051,
      "loss": 4.9129,
      "mean_token_accuracy": 0.208723221719265,
      "num_tokens": 49397678.0,
      "step": 23805
    },
    {
      "entropy": 5.523954916000366,
      "epoch": 2.159833091436865,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004532964892668867,
      "loss": 4.9854,
      "mean_token_accuracy": 0.2093636840581894,
      "num_tokens": 49407368.0,
      "step": 23810
    },
    {
      "entropy": 5.46703314781189,
      "epoch": 2.1602866473149493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045327675116015146,
      "loss": 4.8889,
      "mean_token_accuracy": 0.21758810430765152,
      "num_tokens": 49417926.0,
      "step": 23815
    },
    {
      "entropy": 5.590454483032227,
      "epoch": 2.1607402031930336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004532570093666075,
      "loss": 5.0371,
      "mean_token_accuracy": 0.19904615134000778,
      "num_tokens": 49427899.0,
      "step": 23820
    },
    {
      "entropy": 5.500293254852295,
      "epoch": 2.1611937590711174,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004532372638866633,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22029965817928315,
      "num_tokens": 49439102.0,
      "step": 23825
    },
    {
      "entropy": 5.4722662448883055,
      "epoch": 2.1616473149492017,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045321751472072693,
      "loss": 4.8218,
      "mean_token_accuracy": 0.22426599264144897,
      "num_tokens": 49448013.0,
      "step": 23830
    },
    {
      "entropy": 5.390433073043823,
      "epoch": 2.162100870827286,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045319776186920695,
      "loss": 4.8873,
      "mean_token_accuracy": 0.2208450898528099,
      "num_tokens": 49457620.0,
      "step": 23835
    },
    {
      "entropy": 5.416310024261475,
      "epoch": 2.1625544267053702,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045317800533251193,
      "loss": 4.9335,
      "mean_token_accuracy": 0.2055414006114006,
      "num_tokens": 49468038.0,
      "step": 23840
    },
    {
      "entropy": 5.451368904113769,
      "epoch": 2.163007982583454,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045315824511105027,
      "loss": 4.8597,
      "mean_token_accuracy": 0.21456691473722458,
      "num_tokens": 49477714.0,
      "step": 23845
    },
    {
      "entropy": 5.534938669204712,
      "epoch": 2.1634615384615383,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045313848120523073,
      "loss": 4.847,
      "mean_token_accuracy": 0.21914579570293427,
      "num_tokens": 49488214.0,
      "step": 23850
    },
    {
      "entropy": 5.507592058181762,
      "epoch": 2.1639150943396226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045311871361546205,
      "loss": 4.977,
      "mean_token_accuracy": 0.2114535838365555,
      "num_tokens": 49498444.0,
      "step": 23855
    },
    {
      "entropy": 5.389393854141235,
      "epoch": 2.164368650217707,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004530989423421529,
      "loss": 4.8297,
      "mean_token_accuracy": 0.22005860656499862,
      "num_tokens": 49508759.0,
      "step": 23860
    },
    {
      "entropy": 5.376272201538086,
      "epoch": 2.164822206095791,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045307916738571223,
      "loss": 4.7955,
      "mean_token_accuracy": 0.22844140976667404,
      "num_tokens": 49518787.0,
      "step": 23865
    },
    {
      "entropy": 5.463872623443604,
      "epoch": 2.1652757619738754,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004530593887465491,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2188235953450203,
      "num_tokens": 49528130.0,
      "step": 23870
    },
    {
      "entropy": 5.561067152023315,
      "epoch": 2.165729317851959,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045303960642507235,
      "loss": 5.0574,
      "mean_token_accuracy": 0.20429391413927078,
      "num_tokens": 49539128.0,
      "step": 23875
    },
    {
      "entropy": 5.427201843261718,
      "epoch": 2.1661828737300435,
      "grad_norm": 1.125,
      "learning_rate": 0.00045301982042169123,
      "loss": 4.8802,
      "mean_token_accuracy": 0.21698668450117112,
      "num_tokens": 49550571.0,
      "step": 23880
    },
    {
      "entropy": 5.470988368988037,
      "epoch": 2.1666364296081277,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045300003073681485,
      "loss": 4.917,
      "mean_token_accuracy": 0.2148282289505005,
      "num_tokens": 49561106.0,
      "step": 23885
    },
    {
      "entropy": 5.430626773834229,
      "epoch": 2.167089985486212,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045298023737085243,
      "loss": 4.8116,
      "mean_token_accuracy": 0.2152087152004242,
      "num_tokens": 49571389.0,
      "step": 23890
    },
    {
      "entropy": 5.473476266860962,
      "epoch": 2.1675435413642963,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004529604403242134,
      "loss": 4.9365,
      "mean_token_accuracy": 0.21314407587051393,
      "num_tokens": 49582224.0,
      "step": 23895
    },
    {
      "entropy": 5.4875846862792965,
      "epoch": 2.16799709724238,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045294063959730707,
      "loss": 4.932,
      "mean_token_accuracy": 0.2137991487979889,
      "num_tokens": 49592413.0,
      "step": 23900
    },
    {
      "entropy": 5.529356527328491,
      "epoch": 2.1684506531204644,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045292083519054297,
      "loss": 4.9991,
      "mean_token_accuracy": 0.19885564893484114,
      "num_tokens": 49602796.0,
      "step": 23905
    },
    {
      "entropy": 5.490967655181885,
      "epoch": 2.1689042089985486,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004529010271043307,
      "loss": 4.8815,
      "mean_token_accuracy": 0.21977989077568055,
      "num_tokens": 49613625.0,
      "step": 23910
    },
    {
      "entropy": 5.615268516540527,
      "epoch": 2.169357764876633,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045288121533907977,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21041471213102342,
      "num_tokens": 49624719.0,
      "step": 23915
    },
    {
      "entropy": 5.540744781494141,
      "epoch": 2.169811320754717,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045286139989520006,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20469337552785874,
      "num_tokens": 49634916.0,
      "step": 23920
    },
    {
      "entropy": 5.4966377258300785,
      "epoch": 2.170264876632801,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004528415807731012,
      "loss": 4.8628,
      "mean_token_accuracy": 0.21104287207126618,
      "num_tokens": 49645215.0,
      "step": 23925
    },
    {
      "entropy": 5.473695898056031,
      "epoch": 2.1707184325108853,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045282175797319317,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21484462171792984,
      "num_tokens": 49655805.0,
      "step": 23930
    },
    {
      "entropy": 5.523554372787475,
      "epoch": 2.1711719883889695,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004528019314958859,
      "loss": 4.9164,
      "mean_token_accuracy": 0.21340756714344025,
      "num_tokens": 49665264.0,
      "step": 23935
    },
    {
      "entropy": 5.509702348709107,
      "epoch": 2.171625544267054,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045278210134158935,
      "loss": 4.9769,
      "mean_token_accuracy": 0.20975067764520644,
      "num_tokens": 49676152.0,
      "step": 23940
    },
    {
      "entropy": 5.537830018997193,
      "epoch": 2.172079100145138,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004527622675107137,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2114069014787674,
      "num_tokens": 49686762.0,
      "step": 23945
    },
    {
      "entropy": 5.544857978820801,
      "epoch": 2.172532656023222,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045274243000366897,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21516836285591126,
      "num_tokens": 49697504.0,
      "step": 23950
    },
    {
      "entropy": 5.501814126968384,
      "epoch": 2.172986211901306,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004527225888208655,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20459358394145966,
      "num_tokens": 49709197.0,
      "step": 23955
    },
    {
      "entropy": 5.536350059509277,
      "epoch": 2.1734397677793904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004527027439627137,
      "loss": 4.9377,
      "mean_token_accuracy": 0.20753052830696106,
      "num_tokens": 49719764.0,
      "step": 23960
    },
    {
      "entropy": 5.493772077560425,
      "epoch": 2.1738933236574747,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045268289542962377,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21154548674821855,
      "num_tokens": 49730363.0,
      "step": 23965
    },
    {
      "entropy": 5.5016883373260494,
      "epoch": 2.174346879535559,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004526630432220063,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20137335658073424,
      "num_tokens": 49741037.0,
      "step": 23970
    },
    {
      "entropy": 5.447806072235108,
      "epoch": 2.1748004354136428,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004526431873402719,
      "loss": 4.8481,
      "mean_token_accuracy": 0.21980459094047547,
      "num_tokens": 49750127.0,
      "step": 23975
    },
    {
      "entropy": 5.549685001373291,
      "epoch": 2.175253991291727,
      "grad_norm": 1.078125,
      "learning_rate": 0.000452623327784831,
      "loss": 5.0268,
      "mean_token_accuracy": 0.20665594637393953,
      "num_tokens": 49760298.0,
      "step": 23980
    },
    {
      "entropy": 5.512126779556274,
      "epoch": 2.1757075471698113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004526034645560945,
      "loss": 4.9396,
      "mean_token_accuracy": 0.20661046355962753,
      "num_tokens": 49770841.0,
      "step": 23985
    },
    {
      "entropy": 5.448244667053222,
      "epoch": 2.1761611030478956,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045258359765447306,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21660372018814086,
      "num_tokens": 49781000.0,
      "step": 23990
    },
    {
      "entropy": 5.474337244033814,
      "epoch": 2.17661465892598,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004525637270803776,
      "loss": 4.8965,
      "mean_token_accuracy": 0.21679802983999252,
      "num_tokens": 49790886.0,
      "step": 23995
    },
    {
      "entropy": 5.470546007156372,
      "epoch": 2.1770682148040637,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045254385283421896,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22061644047498702,
      "num_tokens": 49801477.0,
      "step": 24000
    },
    {
      "epoch": 2.1770682148040637,
      "eval_entropy": 5.246431207225333,
      "eval_loss": 5.00707483291626,
      "eval_mean_token_accuracy": 0.21752887504402868,
      "eval_num_tokens": 49801477.0,
      "eval_runtime": 20.8461,
      "eval_samples_per_second": 1696.242,
      "eval_steps_per_second": 212.03,
      "step": 24000
    },
    {
      "entropy": 5.5039448738098145,
      "epoch": 2.177521770682148,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004525239749164082,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21380907148122788,
      "num_tokens": 49812044.0,
      "step": 24005
    },
    {
      "entropy": 5.444004058837891,
      "epoch": 2.177975326560232,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004525040933273564,
      "loss": 4.8487,
      "mean_token_accuracy": 0.21202043443918228,
      "num_tokens": 49821559.0,
      "step": 24010
    },
    {
      "entropy": 5.545518302917481,
      "epoch": 2.1784288824383164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045248420806747483,
      "loss": 5.0237,
      "mean_token_accuracy": 0.1998417243361473,
      "num_tokens": 49832937.0,
      "step": 24015
    },
    {
      "entropy": 5.560429906845092,
      "epoch": 2.1788824383164007,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004524643191371745,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21355916261672975,
      "num_tokens": 49843176.0,
      "step": 24020
    },
    {
      "entropy": 5.4942545890808105,
      "epoch": 2.179335994194485,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045244442653686684,
      "loss": 4.8584,
      "mean_token_accuracy": 0.2214878559112549,
      "num_tokens": 49852522.0,
      "step": 24025
    },
    {
      "entropy": 5.402928876876831,
      "epoch": 2.179789550072569,
      "grad_norm": 1.125,
      "learning_rate": 0.00045242453026696325,
      "loss": 4.8522,
      "mean_token_accuracy": 0.2213190272450447,
      "num_tokens": 49863143.0,
      "step": 24030
    },
    {
      "entropy": 5.514883184432984,
      "epoch": 2.180243105950653,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004524046303278752,
      "loss": 4.959,
      "mean_token_accuracy": 0.21164660155773163,
      "num_tokens": 49873465.0,
      "step": 24035
    },
    {
      "entropy": 5.591395378112793,
      "epoch": 2.1806966618287373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004523847267200141,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19992619007825851,
      "num_tokens": 49883690.0,
      "step": 24040
    },
    {
      "entropy": 5.557153129577637,
      "epoch": 2.1811502177068216,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045236481944379165,
      "loss": 4.9954,
      "mean_token_accuracy": 0.20721260458230972,
      "num_tokens": 49894418.0,
      "step": 24045
    },
    {
      "entropy": 5.529432725906372,
      "epoch": 2.1816037735849054,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004523449084996196,
      "loss": 4.8747,
      "mean_token_accuracy": 0.21668112576007842,
      "num_tokens": 49903927.0,
      "step": 24050
    },
    {
      "entropy": 5.5112556457519535,
      "epoch": 2.1820573294629897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004523249938879096,
      "loss": 4.933,
      "mean_token_accuracy": 0.2119065046310425,
      "num_tokens": 49914203.0,
      "step": 24055
    },
    {
      "entropy": 5.489736843109131,
      "epoch": 2.182510885341074,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045230507560907354,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21616462022066116,
      "num_tokens": 49924320.0,
      "step": 24060
    },
    {
      "entropy": 5.453796148300171,
      "epoch": 2.1829644412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045228515366352334,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21154101938009262,
      "num_tokens": 49934634.0,
      "step": 24065
    },
    {
      "entropy": 5.563674783706665,
      "epoch": 2.1834179970972425,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045226522805167095,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21521165519952773,
      "num_tokens": 49944586.0,
      "step": 24070
    },
    {
      "entropy": 5.5198675155639645,
      "epoch": 2.1838715529753268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004522452987739286,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21574558466672897,
      "num_tokens": 49955441.0,
      "step": 24075
    },
    {
      "entropy": 5.568985033035278,
      "epoch": 2.1843251088534106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045222536583070823,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21131619364023208,
      "num_tokens": 49966718.0,
      "step": 24080
    },
    {
      "entropy": 5.581003475189209,
      "epoch": 2.184778664731495,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045220542922242205,
      "loss": 5.0085,
      "mean_token_accuracy": 0.206375452876091,
      "num_tokens": 49979483.0,
      "step": 24085
    },
    {
      "entropy": 5.492413759231567,
      "epoch": 2.185232220609579,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045218548894948254,
      "loss": 4.907,
      "mean_token_accuracy": 0.21802786439657212,
      "num_tokens": 49989976.0,
      "step": 24090
    },
    {
      "entropy": 5.46004090309143,
      "epoch": 2.1856857764876634,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045216554501230187,
      "loss": 4.7954,
      "mean_token_accuracy": 0.21937553137540816,
      "num_tokens": 50000069.0,
      "step": 24095
    },
    {
      "entropy": 5.5020547866821286,
      "epoch": 2.1861393323657476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004521455974112927,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21807566285133362,
      "num_tokens": 50009970.0,
      "step": 24100
    },
    {
      "entropy": 5.520513486862183,
      "epoch": 2.1865928882438315,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045212564614686735,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2135014295578003,
      "num_tokens": 50019590.0,
      "step": 24105
    },
    {
      "entropy": 5.483197832107544,
      "epoch": 2.1870464441219157,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004521056912194386,
      "loss": 4.9384,
      "mean_token_accuracy": 0.20277114510536193,
      "num_tokens": 50030092.0,
      "step": 24110
    },
    {
      "entropy": 5.46104679107666,
      "epoch": 2.1875,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004520857326294189,
      "loss": 4.9222,
      "mean_token_accuracy": 0.20964971631765367,
      "num_tokens": 50040318.0,
      "step": 24115
    },
    {
      "entropy": 5.556581783294678,
      "epoch": 2.1879535558780843,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045206577037722114,
      "loss": 5.0575,
      "mean_token_accuracy": 0.19703535437583924,
      "num_tokens": 50050507.0,
      "step": 24120
    },
    {
      "entropy": 5.642433214187622,
      "epoch": 2.1884071117561685,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004520458044632582,
      "loss": 5.003,
      "mean_token_accuracy": 0.20688743889331818,
      "num_tokens": 50059727.0,
      "step": 24125
    },
    {
      "entropy": 5.518909168243408,
      "epoch": 2.1888606676342524,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045202583488794285,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21375075578689576,
      "num_tokens": 50070812.0,
      "step": 24130
    },
    {
      "entropy": 5.5386049270629885,
      "epoch": 2.1893142235123366,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004520058616516881,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20394109636545182,
      "num_tokens": 50081175.0,
      "step": 24135
    },
    {
      "entropy": 5.521824359893799,
      "epoch": 2.189767779390421,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045198588475490716,
      "loss": 4.9229,
      "mean_token_accuracy": 0.21007878631353377,
      "num_tokens": 50092246.0,
      "step": 24140
    },
    {
      "entropy": 5.560021018981933,
      "epoch": 2.190221335268505,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004519659041980129,
      "loss": 4.9657,
      "mean_token_accuracy": 0.21317261159420015,
      "num_tokens": 50101934.0,
      "step": 24145
    },
    {
      "entropy": 5.504631280899048,
      "epoch": 2.1906748911465894,
      "grad_norm": 1.25,
      "learning_rate": 0.0004519459199814187,
      "loss": 4.9851,
      "mean_token_accuracy": 0.21087743639945983,
      "num_tokens": 50112159.0,
      "step": 24150
    },
    {
      "entropy": 5.471140956878662,
      "epoch": 2.1911284470246732,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004519259321055377,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20728392601013185,
      "num_tokens": 50123079.0,
      "step": 24155
    },
    {
      "entropy": 5.505477809906006,
      "epoch": 2.1915820029027575,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004519059405707834,
      "loss": 4.9275,
      "mean_token_accuracy": 0.20931319892406464,
      "num_tokens": 50133097.0,
      "step": 24160
    },
    {
      "entropy": 5.577991390228272,
      "epoch": 2.1920355587808418,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045188594537756916,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2101932629942894,
      "num_tokens": 50143062.0,
      "step": 24165
    },
    {
      "entropy": 5.566372299194336,
      "epoch": 2.192489114658926,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045186594652630853,
      "loss": 4.9257,
      "mean_token_accuracy": 0.21754238605499268,
      "num_tokens": 50154482.0,
      "step": 24170
    },
    {
      "entropy": 5.528503942489624,
      "epoch": 2.1929426705370103,
      "grad_norm": 1.171875,
      "learning_rate": 0.000451845944017415,
      "loss": 4.8899,
      "mean_token_accuracy": 0.21995092779397965,
      "num_tokens": 50164880.0,
      "step": 24175
    },
    {
      "entropy": 5.466913318634033,
      "epoch": 2.1933962264150946,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045182593785130225,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2187932923436165,
      "num_tokens": 50173896.0,
      "step": 24180
    },
    {
      "entropy": 5.504962921142578,
      "epoch": 2.1938497822931784,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004518059280283842,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2107956662774086,
      "num_tokens": 50185078.0,
      "step": 24185
    },
    {
      "entropy": 5.547479677200317,
      "epoch": 2.1943033381712627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045178591454907434,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21438332051038742,
      "num_tokens": 50195500.0,
      "step": 24190
    },
    {
      "entropy": 5.469525337219238,
      "epoch": 2.194756894049347,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045176589741378675,
      "loss": 4.8764,
      "mean_token_accuracy": 0.21044387370347978,
      "num_tokens": 50206103.0,
      "step": 24195
    },
    {
      "entropy": 5.429343175888062,
      "epoch": 2.195210449927431,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004517458766229354,
      "loss": 4.8764,
      "mean_token_accuracy": 0.21101847887039185,
      "num_tokens": 50216522.0,
      "step": 24200
    },
    {
      "entropy": 5.541996335983276,
      "epoch": 2.195664005805515,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045172585217693426,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2189615085721016,
      "num_tokens": 50226721.0,
      "step": 24205
    },
    {
      "entropy": 5.591645336151123,
      "epoch": 2.1961175616835993,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045170582407619747,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20701381117105483,
      "num_tokens": 50236327.0,
      "step": 24210
    },
    {
      "entropy": 5.44909987449646,
      "epoch": 2.1965711175616836,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004516857923211392,
      "loss": 4.8717,
      "mean_token_accuracy": 0.2169783368706703,
      "num_tokens": 50245729.0,
      "step": 24215
    },
    {
      "entropy": 5.515542888641358,
      "epoch": 2.197024673439768,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045166575691217375,
      "loss": 5.0042,
      "mean_token_accuracy": 0.2025711104273796,
      "num_tokens": 50257024.0,
      "step": 24220
    },
    {
      "entropy": 5.526003837585449,
      "epoch": 2.197478229317852,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045164571784971536,
      "loss": 4.8705,
      "mean_token_accuracy": 0.217717806994915,
      "num_tokens": 50266910.0,
      "step": 24225
    },
    {
      "entropy": 5.501551961898803,
      "epoch": 2.1979317851959363,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045162567513417864,
      "loss": 4.9089,
      "mean_token_accuracy": 0.21626132875680923,
      "num_tokens": 50278233.0,
      "step": 24230
    },
    {
      "entropy": 5.4796960830688475,
      "epoch": 2.19838534107402,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045160562876597776,
      "loss": 4.8978,
      "mean_token_accuracy": 0.20843525528907775,
      "num_tokens": 50289276.0,
      "step": 24235
    },
    {
      "entropy": 5.493786001205445,
      "epoch": 2.1988388969521044,
      "grad_norm": 1.125,
      "learning_rate": 0.0004515855787455276,
      "loss": 4.921,
      "mean_token_accuracy": 0.2138884648680687,
      "num_tokens": 50299336.0,
      "step": 24240
    },
    {
      "entropy": 5.5388110160827635,
      "epoch": 2.1992924528301887,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004515655250732426,
      "loss": 4.8757,
      "mean_token_accuracy": 0.21270934790372847,
      "num_tokens": 50308308.0,
      "step": 24245
    },
    {
      "entropy": 5.471859073638916,
      "epoch": 2.199746008708273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045154546774953764,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21223640888929368,
      "num_tokens": 50318295.0,
      "step": 24250
    },
    {
      "entropy": 5.425398683547973,
      "epoch": 2.2001995645863572,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004515254067748274,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21996940523386002,
      "num_tokens": 50329296.0,
      "step": 24255
    },
    {
      "entropy": 5.465482568740844,
      "epoch": 2.200653120464441,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004515053421495267,
      "loss": 4.987,
      "mean_token_accuracy": 0.20182658731937408,
      "num_tokens": 50339389.0,
      "step": 24260
    },
    {
      "entropy": 5.567392778396607,
      "epoch": 2.2011066763425253,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045148527387405056,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20201555639505386,
      "num_tokens": 50349647.0,
      "step": 24265
    },
    {
      "entropy": 5.580852079391479,
      "epoch": 2.2015602322206096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004514652019488141,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20866041481494904,
      "num_tokens": 50361292.0,
      "step": 24270
    },
    {
      "entropy": 5.500916814804077,
      "epoch": 2.202013788098694,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004514451263742321,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2093900680541992,
      "num_tokens": 50372293.0,
      "step": 24275
    },
    {
      "entropy": 5.445493888854981,
      "epoch": 2.202467343976778,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004514250471507201,
      "loss": 4.8766,
      "mean_token_accuracy": 0.2177124574780464,
      "num_tokens": 50382462.0,
      "step": 24280
    },
    {
      "entropy": 5.515185117721558,
      "epoch": 2.202920899854862,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045140496427869306,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22146715372800826,
      "num_tokens": 50393157.0,
      "step": 24285
    },
    {
      "entropy": 5.430197286605835,
      "epoch": 2.203374455732946,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004513848777585664,
      "loss": 4.84,
      "mean_token_accuracy": 0.2179591417312622,
      "num_tokens": 50403474.0,
      "step": 24290
    },
    {
      "entropy": 5.483253717422485,
      "epoch": 2.2038280116110305,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004513647875907556,
      "loss": 4.918,
      "mean_token_accuracy": 0.21016072779893874,
      "num_tokens": 50414337.0,
      "step": 24295
    },
    {
      "entropy": 5.552049350738526,
      "epoch": 2.2042815674891147,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045134469377567597,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20477916300296783,
      "num_tokens": 50424863.0,
      "step": 24300
    },
    {
      "entropy": 5.54852032661438,
      "epoch": 2.204735123367199,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004513245963137432,
      "loss": 5.0038,
      "mean_token_accuracy": 0.2095029190182686,
      "num_tokens": 50435982.0,
      "step": 24305
    },
    {
      "entropy": 5.505387687683106,
      "epoch": 2.205188679245283,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004513044952053728,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21179662346839906,
      "num_tokens": 50445112.0,
      "step": 24310
    },
    {
      "entropy": 5.416015768051148,
      "epoch": 2.205642235123367,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045128439045098045,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22095187455415727,
      "num_tokens": 50455266.0,
      "step": 24315
    },
    {
      "entropy": 5.5569178581237795,
      "epoch": 2.2060957910014514,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045126428205098207,
      "loss": 5.0117,
      "mean_token_accuracy": 0.20119641721248627,
      "num_tokens": 50466930.0,
      "step": 24320
    },
    {
      "entropy": 5.521805667877198,
      "epoch": 2.2065493468795356,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004512441700057934,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21509936451911926,
      "num_tokens": 50477534.0,
      "step": 24325
    },
    {
      "entropy": 5.479548025131225,
      "epoch": 2.20700290275762,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004512240543158305,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21623065322637558,
      "num_tokens": 50487352.0,
      "step": 24330
    },
    {
      "entropy": 5.440809631347657,
      "epoch": 2.2074564586357037,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045120393498150905,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2224089980125427,
      "num_tokens": 50496227.0,
      "step": 24335
    },
    {
      "entropy": 5.518359899520874,
      "epoch": 2.207910014513788,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045118381200324537,
      "loss": 4.9586,
      "mean_token_accuracy": 0.20798122137784958,
      "num_tokens": 50507536.0,
      "step": 24340
    },
    {
      "entropy": 5.553762292861938,
      "epoch": 2.2083635703918723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004511636853814556,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21207121163606643,
      "num_tokens": 50517659.0,
      "step": 24345
    },
    {
      "entropy": 5.500620412826538,
      "epoch": 2.2088171262699565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045114355511655595,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2070940151810646,
      "num_tokens": 50528717.0,
      "step": 24350
    },
    {
      "entropy": 5.451986169815063,
      "epoch": 2.209270682148041,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045112342120896267,
      "loss": 4.8326,
      "mean_token_accuracy": 0.21786334961652756,
      "num_tokens": 50538449.0,
      "step": 24355
    },
    {
      "entropy": 5.561678552627564,
      "epoch": 2.2097242380261246,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045110328365909215,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2076120674610138,
      "num_tokens": 50549098.0,
      "step": 24360
    },
    {
      "entropy": 5.5467907905578615,
      "epoch": 2.210177793904209,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004510831424673609,
      "loss": 4.9513,
      "mean_token_accuracy": 0.21326931565999985,
      "num_tokens": 50560047.0,
      "step": 24365
    },
    {
      "entropy": 5.519158411026001,
      "epoch": 2.210631349782293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004510629976341853,
      "loss": 4.9082,
      "mean_token_accuracy": 0.20999544262886047,
      "num_tokens": 50569704.0,
      "step": 24370
    },
    {
      "entropy": 5.488677644729615,
      "epoch": 2.2110849056603774,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045104284915998215,
      "loss": 4.9697,
      "mean_token_accuracy": 0.20823805332183837,
      "num_tokens": 50580006.0,
      "step": 24375
    },
    {
      "entropy": 5.445297813415527,
      "epoch": 2.2115384615384617,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045102269704516793,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2198157787322998,
      "num_tokens": 50589593.0,
      "step": 24380
    },
    {
      "entropy": 5.513132762908936,
      "epoch": 2.211992017416546,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004510025412901595,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21415512561798095,
      "num_tokens": 50598774.0,
      "step": 24385
    },
    {
      "entropy": 5.439147520065307,
      "epoch": 2.2124455732946298,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045098238189537367,
      "loss": 4.8599,
      "mean_token_accuracy": 0.21891965121030807,
      "num_tokens": 50608815.0,
      "step": 24390
    },
    {
      "entropy": 5.4105188846588135,
      "epoch": 2.212899129172714,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045096221886122733,
      "loss": 5.0031,
      "mean_token_accuracy": 0.22026191800832748,
      "num_tokens": 50619255.0,
      "step": 24395
    },
    {
      "entropy": 5.497635412216186,
      "epoch": 2.2133526850507983,
      "grad_norm": 1.125,
      "learning_rate": 0.0004509420521881375,
      "loss": 4.9136,
      "mean_token_accuracy": 0.21755017936229706,
      "num_tokens": 50629483.0,
      "step": 24400
    },
    {
      "entropy": 5.462251853942871,
      "epoch": 2.2138062409288826,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004509218818765211,
      "loss": 4.867,
      "mean_token_accuracy": 0.2167743816971779,
      "num_tokens": 50638552.0,
      "step": 24405
    },
    {
      "entropy": 5.530922937393188,
      "epoch": 2.2142597968069664,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045090170792679544,
      "loss": 4.8606,
      "mean_token_accuracy": 0.2127256065607071,
      "num_tokens": 50648752.0,
      "step": 24410
    },
    {
      "entropy": 5.538662338256836,
      "epoch": 2.2147133526850507,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045088153033937765,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21609770506620407,
      "num_tokens": 50658373.0,
      "step": 24415
    },
    {
      "entropy": 5.476626062393189,
      "epoch": 2.215166908563135,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045086134911468496,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21640825569629668,
      "num_tokens": 50668529.0,
      "step": 24420
    },
    {
      "entropy": 5.530965518951416,
      "epoch": 2.215620464441219,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004508411642531347,
      "loss": 4.8712,
      "mean_token_accuracy": 0.21209428906440736,
      "num_tokens": 50678038.0,
      "step": 24425
    },
    {
      "entropy": 5.491864633560181,
      "epoch": 2.2160740203193035,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004508209757551445,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21791955679655076,
      "num_tokens": 50687788.0,
      "step": 24430
    },
    {
      "entropy": 5.521287965774536,
      "epoch": 2.2165275761973877,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045080078362113153,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21067391335964203,
      "num_tokens": 50697748.0,
      "step": 24435
    },
    {
      "entropy": 5.426697444915772,
      "epoch": 2.2169811320754715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045078058785151363,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22197608649730682,
      "num_tokens": 50707776.0,
      "step": 24440
    },
    {
      "entropy": 5.461878871917724,
      "epoch": 2.217434687953556,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004507603884467085,
      "loss": 4.9745,
      "mean_token_accuracy": 0.20863485783338548,
      "num_tokens": 50717547.0,
      "step": 24445
    },
    {
      "entropy": 5.44483323097229,
      "epoch": 2.21788824383164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004507401854071336,
      "loss": 4.8648,
      "mean_token_accuracy": 0.21784850358963012,
      "num_tokens": 50728446.0,
      "step": 24450
    },
    {
      "entropy": 5.480695533752441,
      "epoch": 2.2183417997097243,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045071997873320687,
      "loss": 4.8573,
      "mean_token_accuracy": 0.2089684024453163,
      "num_tokens": 50737909.0,
      "step": 24455
    },
    {
      "entropy": 5.455913972854614,
      "epoch": 2.2187953555878086,
      "grad_norm": 1.0,
      "learning_rate": 0.00045069976842534635,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21954180002212526,
      "num_tokens": 50749745.0,
      "step": 24460
    },
    {
      "entropy": 5.5276398181915285,
      "epoch": 2.2192489114658924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004506795544839697,
      "loss": 4.9117,
      "mean_token_accuracy": 0.20980933755636216,
      "num_tokens": 50760073.0,
      "step": 24465
    },
    {
      "entropy": 5.540490531921387,
      "epoch": 2.2197024673439767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004506593369094951,
      "loss": 4.8976,
      "mean_token_accuracy": 0.20876749753952026,
      "num_tokens": 50769261.0,
      "step": 24470
    },
    {
      "entropy": 5.577752923965454,
      "epoch": 2.220156023222061,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045063911570234074,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20549732595682144,
      "num_tokens": 50779572.0,
      "step": 24475
    },
    {
      "entropy": 5.443750715255737,
      "epoch": 2.2206095791001452,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004506188908629247,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2161690041422844,
      "num_tokens": 50789893.0,
      "step": 24480
    },
    {
      "entropy": 5.41756763458252,
      "epoch": 2.2210631349782295,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004505986623916652,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2138019859790802,
      "num_tokens": 50799177.0,
      "step": 24485
    },
    {
      "entropy": 5.441324996948242,
      "epoch": 2.2215166908563133,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004505784302889806,
      "loss": 4.887,
      "mean_token_accuracy": 0.21745399832725526,
      "num_tokens": 50810093.0,
      "step": 24490
    },
    {
      "entropy": 5.57575159072876,
      "epoch": 2.2219702467343976,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045055819455528935,
      "loss": 5.04,
      "mean_token_accuracy": 0.20586762428283692,
      "num_tokens": 50820272.0,
      "step": 24495
    },
    {
      "entropy": 5.526523685455322,
      "epoch": 2.222423802612482,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004505379551910099,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20946702659130095,
      "num_tokens": 50830301.0,
      "step": 24500
    },
    {
      "entropy": 5.592215633392334,
      "epoch": 2.222877358490566,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004505177121965608,
      "loss": 4.9769,
      "mean_token_accuracy": 0.2144727036356926,
      "num_tokens": 50839937.0,
      "step": 24505
    },
    {
      "entropy": 5.4974335670471195,
      "epoch": 2.2233309143686504,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045049746557236067,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22041078358888627,
      "num_tokens": 50850295.0,
      "step": 24510
    },
    {
      "entropy": 5.534443330764771,
      "epoch": 2.223784470246734,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045047721531882827,
      "loss": 4.9692,
      "mean_token_accuracy": 0.20637685507535936,
      "num_tokens": 50861889.0,
      "step": 24515
    },
    {
      "entropy": 5.501532459259034,
      "epoch": 2.2242380261248185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004504569614363823,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21105297207832335,
      "num_tokens": 50872011.0,
      "step": 24520
    },
    {
      "entropy": 5.554976177215576,
      "epoch": 2.2246915820029027,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004504367039254416,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20835031270980836,
      "num_tokens": 50882559.0,
      "step": 24525
    },
    {
      "entropy": 5.508613061904907,
      "epoch": 2.225145137880987,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045041644278642523,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20145211070775987,
      "num_tokens": 50894132.0,
      "step": 24530
    },
    {
      "entropy": 5.544841957092285,
      "epoch": 2.2255986937590713,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045039617801975214,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21191783845424653,
      "num_tokens": 50904337.0,
      "step": 24535
    },
    {
      "entropy": 5.54058198928833,
      "epoch": 2.2260522496371555,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045037590962584143,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20757041722536088,
      "num_tokens": 50914961.0,
      "step": 24540
    },
    {
      "entropy": 5.549249887466431,
      "epoch": 2.2265058055152394,
      "grad_norm": 1.125,
      "learning_rate": 0.0004503556376051121,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21164951473474503,
      "num_tokens": 50925682.0,
      "step": 24545
    },
    {
      "entropy": 5.475205898284912,
      "epoch": 2.2269593613933236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045033536195798355,
      "loss": 4.8447,
      "mean_token_accuracy": 0.2260774403810501,
      "num_tokens": 50935809.0,
      "step": 24550
    },
    {
      "entropy": 5.5669139385223385,
      "epoch": 2.227412917271408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000450315082684875,
      "loss": 5.035,
      "mean_token_accuracy": 0.20063673108816146,
      "num_tokens": 50946098.0,
      "step": 24555
    },
    {
      "entropy": 5.542860651016236,
      "epoch": 2.227866473149492,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004502947997862059,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20911372303962708,
      "num_tokens": 50956240.0,
      "step": 24560
    },
    {
      "entropy": 5.523819303512573,
      "epoch": 2.228320029027576,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004502745132623957,
      "loss": 4.896,
      "mean_token_accuracy": 0.2186339095234871,
      "num_tokens": 50966870.0,
      "step": 24565
    },
    {
      "entropy": 5.498482418060303,
      "epoch": 2.2287735849056602,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004502542231138639,
      "loss": 4.8986,
      "mean_token_accuracy": 0.20982664078474045,
      "num_tokens": 50977528.0,
      "step": 24570
    },
    {
      "entropy": 5.484511232376098,
      "epoch": 2.2292271407837445,
      "grad_norm": 1.0,
      "learning_rate": 0.00045023392934103005,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2067478433251381,
      "num_tokens": 50988565.0,
      "step": 24575
    },
    {
      "entropy": 5.493417549133301,
      "epoch": 2.2296806966618288,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045021363194431397,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21599013358354568,
      "num_tokens": 50998410.0,
      "step": 24580
    },
    {
      "entropy": 5.561251401901245,
      "epoch": 2.230134252539913,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004501933309241352,
      "loss": 4.9639,
      "mean_token_accuracy": 0.20599304735660554,
      "num_tokens": 51008890.0,
      "step": 24585
    },
    {
      "entropy": 5.499842500686645,
      "epoch": 2.2305878084179973,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004501730262809138,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21749539524316788,
      "num_tokens": 51020264.0,
      "step": 24590
    },
    {
      "entropy": 5.457573461532593,
      "epoch": 2.231041364296081,
      "grad_norm": 1.125,
      "learning_rate": 0.0004501527180150696,
      "loss": 4.8393,
      "mean_token_accuracy": 0.21981181353330612,
      "num_tokens": 51029892.0,
      "step": 24595
    },
    {
      "entropy": 5.578875160217285,
      "epoch": 2.2314949201741654,
      "grad_norm": 1.125,
      "learning_rate": 0.0004501324061270224,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20302218496799468,
      "num_tokens": 51040499.0,
      "step": 24600
    },
    {
      "entropy": 5.398529386520385,
      "epoch": 2.2319484760522497,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004501120906171926,
      "loss": 4.8409,
      "mean_token_accuracy": 0.2168241873383522,
      "num_tokens": 51050395.0,
      "step": 24605
    },
    {
      "entropy": 5.443792819976807,
      "epoch": 2.232402031930334,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045009177148600004,
      "loss": 4.8417,
      "mean_token_accuracy": 0.22068117260932923,
      "num_tokens": 51060316.0,
      "step": 24610
    },
    {
      "entropy": 5.413550853729248,
      "epoch": 2.232855587808418,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045007144873386514,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21476766020059584,
      "num_tokens": 51070032.0,
      "step": 24615
    },
    {
      "entropy": 5.576509952545166,
      "epoch": 2.233309143686502,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045005112236120795,
      "loss": 5.0515,
      "mean_token_accuracy": 0.20426271706819535,
      "num_tokens": 51080738.0,
      "step": 24620
    },
    {
      "entropy": 5.522678804397583,
      "epoch": 2.2337626995645863,
      "grad_norm": 1.21875,
      "learning_rate": 0.000450030792368449,
      "loss": 4.8765,
      "mean_token_accuracy": 0.21783071607351304,
      "num_tokens": 51090287.0,
      "step": 24625
    },
    {
      "entropy": 5.470934820175171,
      "epoch": 2.2342162554426706,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004500104587560086,
      "loss": 4.9395,
      "mean_token_accuracy": 0.20700586438179017,
      "num_tokens": 51101188.0,
      "step": 24630
    },
    {
      "entropy": 5.459479570388794,
      "epoch": 2.234669811320755,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004499901215243074,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2162790447473526,
      "num_tokens": 51112310.0,
      "step": 24635
    },
    {
      "entropy": 5.501789855957031,
      "epoch": 2.235123367198839,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044996978067376575,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2125258967280388,
      "num_tokens": 51122481.0,
      "step": 24640
    },
    {
      "entropy": 5.466570663452148,
      "epoch": 2.235576923076923,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044994943620480453,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2072527751326561,
      "num_tokens": 51131918.0,
      "step": 24645
    },
    {
      "entropy": 5.517190742492676,
      "epoch": 2.236030478955007,
      "grad_norm": 1.0,
      "learning_rate": 0.00044992908811784446,
      "loss": 5.0,
      "mean_token_accuracy": 0.20055712312459945,
      "num_tokens": 51142861.0,
      "step": 24650
    },
    {
      "entropy": 5.535826683044434,
      "epoch": 2.2364840348330914,
      "grad_norm": 1.125,
      "learning_rate": 0.00044990873641330615,
      "loss": 4.9718,
      "mean_token_accuracy": 0.21103595048189164,
      "num_tokens": 51153956.0,
      "step": 24655
    },
    {
      "entropy": 5.535042858123779,
      "epoch": 2.2369375907111757,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004498883810916107,
      "loss": 4.8427,
      "mean_token_accuracy": 0.21812597513198853,
      "num_tokens": 51164499.0,
      "step": 24660
    },
    {
      "entropy": 5.539503908157348,
      "epoch": 2.23739114658926,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004498680221531789,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21072584837675096,
      "num_tokens": 51175722.0,
      "step": 24665
    },
    {
      "entropy": 5.486478090286255,
      "epoch": 2.237844702467344,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004498476595984319,
      "loss": 4.905,
      "mean_token_accuracy": 0.21377230137586595,
      "num_tokens": 51184597.0,
      "step": 24670
    },
    {
      "entropy": 5.4386683940887455,
      "epoch": 2.238298258345428,
      "grad_norm": 1.125,
      "learning_rate": 0.0004498272934277906,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21161817759275436,
      "num_tokens": 51194751.0,
      "step": 24675
    },
    {
      "entropy": 5.496107482910157,
      "epoch": 2.2387518142235123,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044980692364167646,
      "loss": 4.8969,
      "mean_token_accuracy": 0.2190478965640068,
      "num_tokens": 51204184.0,
      "step": 24680
    },
    {
      "entropy": 5.507309103012085,
      "epoch": 2.2392053701015966,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004497865502405105,
      "loss": 4.9703,
      "mean_token_accuracy": 0.21511485278606415,
      "num_tokens": 51213761.0,
      "step": 24685
    },
    {
      "entropy": 5.513066959381104,
      "epoch": 2.239658925979681,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044976617322471424,
      "loss": 4.9026,
      "mean_token_accuracy": 0.21550100892782212,
      "num_tokens": 51224150.0,
      "step": 24690
    },
    {
      "entropy": 5.535574197769165,
      "epoch": 2.2401124818577647,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000449745792594709,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21105946898460387,
      "num_tokens": 51235133.0,
      "step": 24695
    },
    {
      "entropy": 5.533110189437866,
      "epoch": 2.240566037735849,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044972540835091617,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21311247795820237,
      "num_tokens": 51245632.0,
      "step": 24700
    },
    {
      "entropy": 5.535616540908814,
      "epoch": 2.241019593613933,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044970502049375736,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21472794711589813,
      "num_tokens": 51255747.0,
      "step": 24705
    },
    {
      "entropy": 5.491992998123169,
      "epoch": 2.2414731494920175,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004496846290236542,
      "loss": 4.8889,
      "mean_token_accuracy": 0.21574348211288452,
      "num_tokens": 51265531.0,
      "step": 24710
    },
    {
      "entropy": 5.459112119674683,
      "epoch": 2.2419267053701017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044966423394102845,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21495272368192672,
      "num_tokens": 51275319.0,
      "step": 24715
    },
    {
      "entropy": 5.506104612350464,
      "epoch": 2.2423802612481856,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044964383524630176,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2196805328130722,
      "num_tokens": 51285655.0,
      "step": 24720
    },
    {
      "entropy": 5.446424865722657,
      "epoch": 2.24283381712627,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004496234329398961,
      "loss": 4.8898,
      "mean_token_accuracy": 0.20930022299289702,
      "num_tokens": 51295938.0,
      "step": 24725
    },
    {
      "entropy": 5.499974632263184,
      "epoch": 2.243287373004354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004496030270222333,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21732360869646072,
      "num_tokens": 51306483.0,
      "step": 24730
    },
    {
      "entropy": 5.501000499725341,
      "epoch": 2.2437409288824384,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044958261749373547,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2183854639530182,
      "num_tokens": 51316446.0,
      "step": 24735
    },
    {
      "entropy": 5.460595941543579,
      "epoch": 2.2441944847605226,
      "grad_norm": 1.40625,
      "learning_rate": 0.00044956220435482465,
      "loss": 4.9399,
      "mean_token_accuracy": 0.212315334379673,
      "num_tokens": 51326840.0,
      "step": 24740
    },
    {
      "entropy": 5.4905448913574215,
      "epoch": 2.244648040638607,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044954178760592283,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20576706528663635,
      "num_tokens": 51337573.0,
      "step": 24745
    },
    {
      "entropy": 5.548055505752563,
      "epoch": 2.2451015965166907,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044952136724745254,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21130934804677964,
      "num_tokens": 51347827.0,
      "step": 24750
    },
    {
      "entropy": 5.493374490737915,
      "epoch": 2.245555152394775,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004495009432798358,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21528324037790297,
      "num_tokens": 51358396.0,
      "step": 24755
    },
    {
      "entropy": 5.529035186767578,
      "epoch": 2.2460087082728593,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044948051570349503,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20652328729629515,
      "num_tokens": 51368882.0,
      "step": 24760
    },
    {
      "entropy": 5.496187686920166,
      "epoch": 2.2464622641509435,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004494600845188528,
      "loss": 4.983,
      "mean_token_accuracy": 0.21016175150871277,
      "num_tokens": 51378717.0,
      "step": 24765
    },
    {
      "entropy": 5.5063189506530765,
      "epoch": 2.2469158200290273,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044943964972633154,
      "loss": 4.9309,
      "mean_token_accuracy": 0.20500435829162597,
      "num_tokens": 51389270.0,
      "step": 24770
    },
    {
      "entropy": 5.497609996795655,
      "epoch": 2.2473693759071116,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004494192113263539,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2126495897769928,
      "num_tokens": 51400025.0,
      "step": 24775
    },
    {
      "entropy": 5.520210933685303,
      "epoch": 2.247822931785196,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044939876931934255,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21629325896501542,
      "num_tokens": 51410903.0,
      "step": 24780
    },
    {
      "entropy": 5.47692551612854,
      "epoch": 2.24827648766328,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004493783237057201,
      "loss": 4.8886,
      "mean_token_accuracy": 0.2160784959793091,
      "num_tokens": 51420246.0,
      "step": 24785
    },
    {
      "entropy": 5.5707673072814945,
      "epoch": 2.2487300435413644,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004493578744859095,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20975974947214127,
      "num_tokens": 51431309.0,
      "step": 24790
    },
    {
      "entropy": 5.544101285934448,
      "epoch": 2.2491835994194487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493374216603338,
      "loss": 4.9816,
      "mean_token_accuracy": 0.20526008754968644,
      "num_tokens": 51443142.0,
      "step": 24795
    },
    {
      "entropy": 5.637153005599975,
      "epoch": 2.2496371552975325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004493169652294156,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20157742798328399,
      "num_tokens": 51454499.0,
      "step": 24800
    },
    {
      "entropy": 5.506328010559082,
      "epoch": 2.2500907111756168,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044929650519357823,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21498187780380248,
      "num_tokens": 51464667.0,
      "step": 24805
    },
    {
      "entropy": 5.545187854766846,
      "epoch": 2.250544267053701,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004492760415532447,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2146904483437538,
      "num_tokens": 51473627.0,
      "step": 24810
    },
    {
      "entropy": 5.484700441360474,
      "epoch": 2.2509978229317853,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004492555743088382,
      "loss": 4.8097,
      "mean_token_accuracy": 0.2185356765985489,
      "num_tokens": 51483052.0,
      "step": 24815
    },
    {
      "entropy": 5.46072359085083,
      "epoch": 2.2514513788098696,
      "grad_norm": 1.140625,
      "learning_rate": 0.000449235103460782,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21084153056144714,
      "num_tokens": 51492422.0,
      "step": 24820
    },
    {
      "entropy": 5.419351291656494,
      "epoch": 2.2519049346879534,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004492146290094995,
      "loss": 4.8775,
      "mean_token_accuracy": 0.21834704428911209,
      "num_tokens": 51503566.0,
      "step": 24825
    },
    {
      "entropy": 5.483908700942993,
      "epoch": 2.2523584905660377,
      "grad_norm": 1.125,
      "learning_rate": 0.00044919415095541406,
      "loss": 4.9057,
      "mean_token_accuracy": 0.21446417272090912,
      "num_tokens": 51513699.0,
      "step": 24830
    },
    {
      "entropy": 5.525264263153076,
      "epoch": 2.252812046444122,
      "grad_norm": 1.125,
      "learning_rate": 0.00044917366929894914,
      "loss": 4.924,
      "mean_token_accuracy": 0.21129962801933289,
      "num_tokens": 51524745.0,
      "step": 24835
    },
    {
      "entropy": 5.497190380096436,
      "epoch": 2.253265602322206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044915318404052846,
      "loss": 4.8706,
      "mean_token_accuracy": 0.20967897027730942,
      "num_tokens": 51536516.0,
      "step": 24840
    },
    {
      "entropy": 5.5260673522949215,
      "epoch": 2.2537191582002905,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044913269518057547,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2036701887845993,
      "num_tokens": 51547063.0,
      "step": 24845
    },
    {
      "entropy": 5.485346078872681,
      "epoch": 2.2541727140783743,
      "grad_norm": 1.125,
      "learning_rate": 0.000449112202719514,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21127251535654068,
      "num_tokens": 51557244.0,
      "step": 24850
    },
    {
      "entropy": 5.544744873046875,
      "epoch": 2.2546262699564585,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004490917066577678,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2155696615576744,
      "num_tokens": 51567630.0,
      "step": 24855
    },
    {
      "entropy": 5.4772899627685545,
      "epoch": 2.255079825834543,
      "grad_norm": 1.125,
      "learning_rate": 0.00044907120699576075,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21144208461046218,
      "num_tokens": 51577951.0,
      "step": 24860
    },
    {
      "entropy": 5.516797971725464,
      "epoch": 2.255533381712627,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044905070373391684,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2137904241681099,
      "num_tokens": 51587533.0,
      "step": 24865
    },
    {
      "entropy": 5.581253719329834,
      "epoch": 2.2559869375907113,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004490301968726599,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2248891219496727,
      "num_tokens": 51597212.0,
      "step": 24870
    },
    {
      "entropy": 5.520354604721069,
      "epoch": 2.256440493468795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044900968641241426,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22100985795259476,
      "num_tokens": 51608520.0,
      "step": 24875
    },
    {
      "entropy": 5.514130020141602,
      "epoch": 2.2568940493468794,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004489891723536039,
      "loss": 5.012,
      "mean_token_accuracy": 0.20341854840517043,
      "num_tokens": 51619651.0,
      "step": 24880
    },
    {
      "entropy": 5.442519950866699,
      "epoch": 2.2573476052249637,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004489686546966531,
      "loss": 4.838,
      "mean_token_accuracy": 0.21501216739416124,
      "num_tokens": 51630184.0,
      "step": 24885
    },
    {
      "entropy": 5.414009380340576,
      "epoch": 2.257801161103048,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044894813344198626,
      "loss": 4.8403,
      "mean_token_accuracy": 0.22066620141267776,
      "num_tokens": 51641293.0,
      "step": 24890
    },
    {
      "entropy": 5.4918395519256595,
      "epoch": 2.2582547169811322,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004489276085900276,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2189863309264183,
      "num_tokens": 51652002.0,
      "step": 24895
    },
    {
      "entropy": 5.407748174667359,
      "epoch": 2.2587082728592165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044890708014120174,
      "loss": 4.8033,
      "mean_token_accuracy": 0.2253187268972397,
      "num_tokens": 51661668.0,
      "step": 24900
    },
    {
      "entropy": 5.58123369216919,
      "epoch": 2.2591618287373003,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004488865480959331,
      "loss": 4.9786,
      "mean_token_accuracy": 0.20617177039384843,
      "num_tokens": 51672879.0,
      "step": 24905
    },
    {
      "entropy": 5.50647087097168,
      "epoch": 2.2596153846153846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004488660124546464,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2115379422903061,
      "num_tokens": 51683878.0,
      "step": 24910
    },
    {
      "entropy": 5.456657075881958,
      "epoch": 2.260068940493469,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004488454732177662,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21216744929552078,
      "num_tokens": 51693971.0,
      "step": 24915
    },
    {
      "entropy": 5.485080003738403,
      "epoch": 2.260522496371553,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044882493038571737,
      "loss": 4.8495,
      "mean_token_accuracy": 0.22081977277994155,
      "num_tokens": 51703685.0,
      "step": 24920
    },
    {
      "entropy": 5.5374064445495605,
      "epoch": 2.260976052249637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044880438395892457,
      "loss": 4.9533,
      "mean_token_accuracy": 0.20925583839416503,
      "num_tokens": 51714469.0,
      "step": 24925
    },
    {
      "entropy": 5.521102619171143,
      "epoch": 2.261429608127721,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004487838339378129,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20159583836793898,
      "num_tokens": 51725283.0,
      "step": 24930
    },
    {
      "entropy": 5.502027416229248,
      "epoch": 2.2618831640058055,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004487632803228072,
      "loss": 4.9348,
      "mean_token_accuracy": 0.2101256638765335,
      "num_tokens": 51735628.0,
      "step": 24935
    },
    {
      "entropy": 5.439960670471192,
      "epoch": 2.2623367198838897,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044874272311433265,
      "loss": 4.798,
      "mean_token_accuracy": 0.22405184656381608,
      "num_tokens": 51745991.0,
      "step": 24940
    },
    {
      "entropy": 5.469972610473633,
      "epoch": 2.262790275761974,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004487221623128143,
      "loss": 4.8452,
      "mean_token_accuracy": 0.22149375677108765,
      "num_tokens": 51756306.0,
      "step": 24945
    },
    {
      "entropy": 5.490349435806275,
      "epoch": 2.2632438316400583,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044870159791867743,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21372815519571303,
      "num_tokens": 51765466.0,
      "step": 24950
    },
    {
      "entropy": 5.513730669021607,
      "epoch": 2.263697387518142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044868102993234716,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2100026860833168,
      "num_tokens": 51775526.0,
      "step": 24955
    },
    {
      "entropy": 5.528044748306274,
      "epoch": 2.2641509433962264,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044866045835424896,
      "loss": 4.9259,
      "mean_token_accuracy": 0.20677538514137267,
      "num_tokens": 51785434.0,
      "step": 24960
    },
    {
      "entropy": 5.518482732772827,
      "epoch": 2.2646044992743106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004486398831848082,
      "loss": 4.9966,
      "mean_token_accuracy": 0.20300792455673217,
      "num_tokens": 51796047.0,
      "step": 24965
    },
    {
      "entropy": 5.512976408004761,
      "epoch": 2.265058055152395,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044861930442445037,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2072262153029442,
      "num_tokens": 51806173.0,
      "step": 24970
    },
    {
      "entropy": 5.594843912124634,
      "epoch": 2.2655116110304787,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004485987220736012,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20474183261394502,
      "num_tokens": 51817644.0,
      "step": 24975
    },
    {
      "entropy": 5.5645793914794925,
      "epoch": 2.265965166908563,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004485781361326861,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20522074550390243,
      "num_tokens": 51829746.0,
      "step": 24980
    },
    {
      "entropy": 5.516353368759155,
      "epoch": 2.2664187227866472,
      "grad_norm": 1.15625,
      "learning_rate": 0.000448557546602131,
      "loss": 4.9911,
      "mean_token_accuracy": 0.20501539409160613,
      "num_tokens": 51840868.0,
      "step": 24985
    },
    {
      "entropy": 5.501433801651001,
      "epoch": 2.2668722786647315,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044853695348236156,
      "loss": 4.8673,
      "mean_token_accuracy": 0.21613098978996276,
      "num_tokens": 51851385.0,
      "step": 24990
    },
    {
      "entropy": 5.5069276809692385,
      "epoch": 2.2673258345428158,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044851635677380374,
      "loss": 4.9106,
      "mean_token_accuracy": 0.20622318536043166,
      "num_tokens": 51861021.0,
      "step": 24995
    },
    {
      "entropy": 5.499903392791748,
      "epoch": 2.2677793904209,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004484957564768834,
      "loss": 4.9941,
      "mean_token_accuracy": 0.20895010679960252,
      "num_tokens": 51870688.0,
      "step": 25000
    },
    {
      "entropy": 5.483315944671631,
      "epoch": 2.268232946298984,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044847515259202657,
      "loss": 4.9425,
      "mean_token_accuracy": 0.20906522125005722,
      "num_tokens": 51880439.0,
      "step": 25005
    },
    {
      "entropy": 5.443620491027832,
      "epoch": 2.268686502177068,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044845454511965947,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21530711203813552,
      "num_tokens": 51890861.0,
      "step": 25010
    },
    {
      "entropy": 5.5371160984039305,
      "epoch": 2.2691400580551524,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044843393406020823,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21241849660873413,
      "num_tokens": 51900687.0,
      "step": 25015
    },
    {
      "entropy": 5.5482300281524655,
      "epoch": 2.2695936139332367,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044841331941409884,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19882063269615174,
      "num_tokens": 51911805.0,
      "step": 25020
    },
    {
      "entropy": 5.477043485641479,
      "epoch": 2.270047169811321,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044839270118175793,
      "loss": 4.8435,
      "mean_token_accuracy": 0.2224956512451172,
      "num_tokens": 51922520.0,
      "step": 25025
    },
    {
      "entropy": 5.568369674682617,
      "epoch": 2.2705007256894048,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004483720793636118,
      "loss": 4.9998,
      "mean_token_accuracy": 0.20784735381603242,
      "num_tokens": 51934056.0,
      "step": 25030
    },
    {
      "entropy": 5.528074073791504,
      "epoch": 2.270954281567489,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044835145396008685,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2076761394739151,
      "num_tokens": 51944696.0,
      "step": 25035
    },
    {
      "entropy": 5.551304531097412,
      "epoch": 2.2714078374455733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004483308249716097,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21023217290639878,
      "num_tokens": 51955111.0,
      "step": 25040
    },
    {
      "entropy": 5.491563367843628,
      "epoch": 2.2718613933236576,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004483101923986068,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21420104503631593,
      "num_tokens": 51966036.0,
      "step": 25045
    },
    {
      "entropy": 5.456214714050293,
      "epoch": 2.272314949201742,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044828955624150505,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21493476033210754,
      "num_tokens": 51976060.0,
      "step": 25050
    },
    {
      "entropy": 5.522627496719361,
      "epoch": 2.2727685050798256,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004482689165007311,
      "loss": 4.9288,
      "mean_token_accuracy": 0.20994595885276796,
      "num_tokens": 51986016.0,
      "step": 25055
    },
    {
      "entropy": 5.475983476638794,
      "epoch": 2.27322206095791,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044824827317671177,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22039448767900466,
      "num_tokens": 51994794.0,
      "step": 25060
    },
    {
      "entropy": 5.390714311599732,
      "epoch": 2.273675616835994,
      "grad_norm": 1.140625,
      "learning_rate": 0.000448227626269874,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2154348686337471,
      "num_tokens": 52004976.0,
      "step": 25065
    },
    {
      "entropy": 5.523627901077271,
      "epoch": 2.2741291727140784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044820697578064485,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20627030432224275,
      "num_tokens": 52016380.0,
      "step": 25070
    },
    {
      "entropy": 5.553480815887451,
      "epoch": 2.2745827285921627,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004481863217094513,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2135339841246605,
      "num_tokens": 52026439.0,
      "step": 25075
    },
    {
      "entropy": 5.48488039970398,
      "epoch": 2.2750362844702465,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004481656640567204,
      "loss": 4.8799,
      "mean_token_accuracy": 0.21348977833986282,
      "num_tokens": 52037111.0,
      "step": 25080
    },
    {
      "entropy": 5.458954763412476,
      "epoch": 2.275489840348331,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004481450028228795,
      "loss": 4.9367,
      "mean_token_accuracy": 0.20922750979661942,
      "num_tokens": 52047463.0,
      "step": 25085
    },
    {
      "entropy": 5.503176927566528,
      "epoch": 2.275943396226415,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004481243380083558,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2129725322127342,
      "num_tokens": 52058081.0,
      "step": 25090
    },
    {
      "entropy": 5.500329685211182,
      "epoch": 2.2763969521044993,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004481036696135767,
      "loss": 4.9487,
      "mean_token_accuracy": 0.21019009202718736,
      "num_tokens": 52069367.0,
      "step": 25095
    },
    {
      "entropy": 5.553674745559692,
      "epoch": 2.2768505079825836,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004480829976389695,
      "loss": 4.9019,
      "mean_token_accuracy": 0.21269881278276442,
      "num_tokens": 52080023.0,
      "step": 25100
    },
    {
      "entropy": 5.464703989028931,
      "epoch": 2.277304063860668,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044806232208496194,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2161496415734291,
      "num_tokens": 52089184.0,
      "step": 25105
    },
    {
      "entropy": 5.532656240463257,
      "epoch": 2.2777576197387517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004480416429519814,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21149049550294877,
      "num_tokens": 52098711.0,
      "step": 25110
    },
    {
      "entropy": 5.541463422775268,
      "epoch": 2.278211175616836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004480209602404556,
      "loss": 4.911,
      "mean_token_accuracy": 0.22044273167848588,
      "num_tokens": 52108826.0,
      "step": 25115
    },
    {
      "entropy": 5.508780145645142,
      "epoch": 2.27866473149492,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004480002739508122,
      "loss": 4.8892,
      "mean_token_accuracy": 0.21802321821451187,
      "num_tokens": 52118934.0,
      "step": 25120
    },
    {
      "entropy": 5.449331474304199,
      "epoch": 2.2791182873730045,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004479795840834791,
      "loss": 4.8689,
      "mean_token_accuracy": 0.2189303532242775,
      "num_tokens": 52129803.0,
      "step": 25125
    },
    {
      "entropy": 5.408039951324463,
      "epoch": 2.2795718432510883,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044795889063888413,
      "loss": 4.8086,
      "mean_token_accuracy": 0.21691848635673522,
      "num_tokens": 52140070.0,
      "step": 25130
    },
    {
      "entropy": 5.638384771347046,
      "epoch": 2.2800253991291726,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004479381936174552,
      "loss": 5.0906,
      "mean_token_accuracy": 0.2016774982213974,
      "num_tokens": 52150462.0,
      "step": 25135
    },
    {
      "entropy": 5.5687517642974855,
      "epoch": 2.280478955007257,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004479174930196204,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2137855738401413,
      "num_tokens": 52160552.0,
      "step": 25140
    },
    {
      "entropy": 5.416904401779175,
      "epoch": 2.280932510885341,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044789678884580767,
      "loss": 4.8516,
      "mean_token_accuracy": 0.219556987285614,
      "num_tokens": 52169780.0,
      "step": 25145
    },
    {
      "entropy": 5.4925432205200195,
      "epoch": 2.2813860667634254,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044787608109644536,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21881770640611647,
      "num_tokens": 52180055.0,
      "step": 25150
    },
    {
      "entropy": 5.473763942718506,
      "epoch": 2.2818396226415096,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004478553697719616,
      "loss": 4.8813,
      "mean_token_accuracy": 0.21632434725761412,
      "num_tokens": 52189772.0,
      "step": 25155
    },
    {
      "entropy": 5.50630750656128,
      "epoch": 2.2822931785195935,
      "grad_norm": 1.125,
      "learning_rate": 0.0004478346548727848,
      "loss": 4.8524,
      "mean_token_accuracy": 0.22450515031814575,
      "num_tokens": 52199608.0,
      "step": 25160
    },
    {
      "entropy": 5.4512614727020265,
      "epoch": 2.2827467343976777,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004478139363993433,
      "loss": 4.9469,
      "mean_token_accuracy": 0.20535383224487305,
      "num_tokens": 52208236.0,
      "step": 25165
    },
    {
      "entropy": 5.508632230758667,
      "epoch": 2.283200290275762,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004477932143520655,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21591737121343613,
      "num_tokens": 52218403.0,
      "step": 25170
    },
    {
      "entropy": 5.532778644561768,
      "epoch": 2.2836538461538463,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044777248873138005,
      "loss": 4.927,
      "mean_token_accuracy": 0.21506611555814742,
      "num_tokens": 52228255.0,
      "step": 25175
    },
    {
      "entropy": 5.5219035148620605,
      "epoch": 2.2841074020319305,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044775175953771547,
      "loss": 4.9505,
      "mean_token_accuracy": 0.20973069220781326,
      "num_tokens": 52237732.0,
      "step": 25180
    },
    {
      "entropy": 5.5970922946929935,
      "epoch": 2.2845609579100143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004477310267715005,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20052250623703002,
      "num_tokens": 52249226.0,
      "step": 25185
    },
    {
      "entropy": 5.538551330566406,
      "epoch": 2.2850145137880986,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004477102904331639,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21238505840301514,
      "num_tokens": 52259550.0,
      "step": 25190
    },
    {
      "entropy": 5.562503480911255,
      "epoch": 2.285468069666183,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044768955052313443,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21429784148931502,
      "num_tokens": 52269670.0,
      "step": 25195
    },
    {
      "entropy": 5.541666793823242,
      "epoch": 2.285921625544267,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044766880704184104,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2136317938566208,
      "num_tokens": 52280135.0,
      "step": 25200
    },
    {
      "entropy": 5.486477565765381,
      "epoch": 2.2863751814223514,
      "grad_norm": 1.125,
      "learning_rate": 0.0004476480599897128,
      "loss": 4.8544,
      "mean_token_accuracy": 0.2182181343436241,
      "num_tokens": 52291806.0,
      "step": 25205
    },
    {
      "entropy": 5.415923261642456,
      "epoch": 2.2868287373004352,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004476273093671786,
      "loss": 4.8887,
      "mean_token_accuracy": 0.21271032840013504,
      "num_tokens": 52302118.0,
      "step": 25210
    },
    {
      "entropy": 5.515542936325073,
      "epoch": 2.2872822931785195,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044760655517466766,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21006736159324646,
      "num_tokens": 52312762.0,
      "step": 25215
    },
    {
      "entropy": 5.562449169158936,
      "epoch": 2.2877358490566038,
      "grad_norm": 1.125,
      "learning_rate": 0.00044758579741260916,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2087487295269966,
      "num_tokens": 52323367.0,
      "step": 25220
    },
    {
      "entropy": 5.58119568824768,
      "epoch": 2.288189404934688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044756503608143247,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20538643300533294,
      "num_tokens": 52335231.0,
      "step": 25225
    },
    {
      "entropy": 5.571872615814209,
      "epoch": 2.2886429608127723,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044754427118156675,
      "loss": 4.8911,
      "mean_token_accuracy": 0.21371162235736846,
      "num_tokens": 52345066.0,
      "step": 25230
    },
    {
      "entropy": 5.50447678565979,
      "epoch": 2.289096516690856,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004475235027134416,
      "loss": 4.944,
      "mean_token_accuracy": 0.21927476525306702,
      "num_tokens": 52356383.0,
      "step": 25235
    },
    {
      "entropy": 5.5222252368927,
      "epoch": 2.2895500725689404,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004475027306774865,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21130027621984482,
      "num_tokens": 52366013.0,
      "step": 25240
    },
    {
      "entropy": 5.538517093658447,
      "epoch": 2.2900036284470247,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044748195507413095,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21573576182127,
      "num_tokens": 52376298.0,
      "step": 25245
    },
    {
      "entropy": 5.42292046546936,
      "epoch": 2.290457184325109,
      "grad_norm": 1.25,
      "learning_rate": 0.00044746117590380456,
      "loss": 4.8472,
      "mean_token_accuracy": 0.21487735509872435,
      "num_tokens": 52387134.0,
      "step": 25250
    },
    {
      "entropy": 5.483355140686035,
      "epoch": 2.290910740203193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004474403931669371,
      "loss": 4.8337,
      "mean_token_accuracy": 0.21606738716363907,
      "num_tokens": 52397389.0,
      "step": 25255
    },
    {
      "entropy": 5.48673095703125,
      "epoch": 2.2913642960812775,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004474196068639584,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22078489065170287,
      "num_tokens": 52407456.0,
      "step": 25260
    },
    {
      "entropy": 5.45056037902832,
      "epoch": 2.2918178519593613,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004473988169952983,
      "loss": 4.9459,
      "mean_token_accuracy": 0.20922325998544694,
      "num_tokens": 52417967.0,
      "step": 25265
    },
    {
      "entropy": 5.555845069885254,
      "epoch": 2.2922714078374455,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044737802356138674,
      "loss": 5.01,
      "mean_token_accuracy": 0.21173913776874542,
      "num_tokens": 52428757.0,
      "step": 25270
    },
    {
      "entropy": 5.496116065979004,
      "epoch": 2.29272496371553,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004473572265626537,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22192395925521852,
      "num_tokens": 52438973.0,
      "step": 25275
    },
    {
      "entropy": 5.474850368499756,
      "epoch": 2.293178519593614,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004473364259995293,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2105103150010109,
      "num_tokens": 52449323.0,
      "step": 25280
    },
    {
      "entropy": 5.522763013839722,
      "epoch": 2.293632075471698,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004473156218724436,
      "loss": 4.8962,
      "mean_token_accuracy": 0.21868317276239396,
      "num_tokens": 52459299.0,
      "step": 25285
    },
    {
      "entropy": 5.440842151641846,
      "epoch": 2.294085631349782,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044729481418182704,
      "loss": 4.8834,
      "mean_token_accuracy": 0.20932069122791291,
      "num_tokens": 52470026.0,
      "step": 25290
    },
    {
      "entropy": 5.507685327529908,
      "epoch": 2.2945391872278664,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044727400292810985,
      "loss": 4.872,
      "mean_token_accuracy": 0.2175827980041504,
      "num_tokens": 52479975.0,
      "step": 25295
    },
    {
      "entropy": 5.565283632278442,
      "epoch": 2.2949927431059507,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044725318811172227,
      "loss": 5.009,
      "mean_token_accuracy": 0.2099424719810486,
      "num_tokens": 52491067.0,
      "step": 25300
    },
    {
      "entropy": 5.48869833946228,
      "epoch": 2.295446298984035,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000447232369733095,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21529011130332948,
      "num_tokens": 52500671.0,
      "step": 25305
    },
    {
      "entropy": 5.4945282459259035,
      "epoch": 2.2958998548621192,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004472115477926583,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2124891072511673,
      "num_tokens": 52509910.0,
      "step": 25310
    },
    {
      "entropy": 5.475480651855468,
      "epoch": 2.296353410740203,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044719072229084295,
      "loss": 4.8956,
      "mean_token_accuracy": 0.21980800479650497,
      "num_tokens": 52520517.0,
      "step": 25315
    },
    {
      "entropy": 5.5273809909820555,
      "epoch": 2.2968069666182873,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004471698932280796,
      "loss": 4.9742,
      "mean_token_accuracy": 0.217047218978405,
      "num_tokens": 52530760.0,
      "step": 25320
    },
    {
      "entropy": 5.581479692459107,
      "epoch": 2.2972605224963716,
      "grad_norm": 1.15625,
      "learning_rate": 0.000447149060604799,
      "loss": 5.0084,
      "mean_token_accuracy": 0.206503826379776,
      "num_tokens": 52541791.0,
      "step": 25325
    },
    {
      "entropy": 5.587798547744751,
      "epoch": 2.297714078374456,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044712822442143194,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21387242674827575,
      "num_tokens": 52551885.0,
      "step": 25330
    },
    {
      "entropy": 5.514059209823609,
      "epoch": 2.2981676342525397,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044710738467840927,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2184450551867485,
      "num_tokens": 52563080.0,
      "step": 25335
    },
    {
      "entropy": 5.476502323150635,
      "epoch": 2.298621190130624,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004470865413761621,
      "loss": 4.8514,
      "mean_token_accuracy": 0.2222488522529602,
      "num_tokens": 52573564.0,
      "step": 25340
    },
    {
      "entropy": 5.49130220413208,
      "epoch": 2.299074746008708,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044706569451512143,
      "loss": 4.9086,
      "mean_token_accuracy": 0.2145822301506996,
      "num_tokens": 52583801.0,
      "step": 25345
    },
    {
      "entropy": 5.506036710739136,
      "epoch": 2.2995283018867925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044704484409571825,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21144819706678392,
      "num_tokens": 52594908.0,
      "step": 25350
    },
    {
      "entropy": 5.4511350154876705,
      "epoch": 2.2999818577648767,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044702399011838393,
      "loss": 4.8744,
      "mean_token_accuracy": 0.20743487179279327,
      "num_tokens": 52604861.0,
      "step": 25355
    },
    {
      "entropy": 5.510947561264038,
      "epoch": 2.300435413642961,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004470031325835497,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2156983032822609,
      "num_tokens": 52616203.0,
      "step": 25360
    },
    {
      "entropy": 5.50053768157959,
      "epoch": 2.300888969521045,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004469822714916468,
      "loss": 4.8596,
      "mean_token_accuracy": 0.2153087228536606,
      "num_tokens": 52625838.0,
      "step": 25365
    },
    {
      "entropy": 5.4293008804321286,
      "epoch": 2.301342525399129,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004469614068431067,
      "loss": 4.9076,
      "mean_token_accuracy": 0.21574092507362366,
      "num_tokens": 52635741.0,
      "step": 25370
    },
    {
      "entropy": 5.4517176151275635,
      "epoch": 2.3017960812772134,
      "grad_norm": 1.125,
      "learning_rate": 0.0004469405386383608,
      "loss": 4.8454,
      "mean_token_accuracy": 0.21863141804933547,
      "num_tokens": 52645948.0,
      "step": 25375
    },
    {
      "entropy": 5.529945135116577,
      "epoch": 2.3022496371552976,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044691966687784087,
      "loss": 4.9136,
      "mean_token_accuracy": 0.2179325446486473,
      "num_tokens": 52655073.0,
      "step": 25380
    },
    {
      "entropy": 5.45610933303833,
      "epoch": 2.302703193033382,
      "grad_norm": 1.125,
      "learning_rate": 0.00044689879156197833,
      "loss": 4.8825,
      "mean_token_accuracy": 0.2225338786840439,
      "num_tokens": 52664429.0,
      "step": 25385
    },
    {
      "entropy": 5.548095703125,
      "epoch": 2.3031567489114657,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044687791269120504,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20901824980974198,
      "num_tokens": 52674890.0,
      "step": 25390
    },
    {
      "entropy": 5.589496660232544,
      "epoch": 2.30361030478955,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044685703026595266,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21246989965438842,
      "num_tokens": 52685673.0,
      "step": 25395
    },
    {
      "entropy": 5.60549578666687,
      "epoch": 2.3040638606676342,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044683614428665314,
      "loss": 5.022,
      "mean_token_accuracy": 0.2052040383219719,
      "num_tokens": 52696696.0,
      "step": 25400
    },
    {
      "entropy": 5.513788414001465,
      "epoch": 2.3045174165457185,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044681525475373843,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21050163805484773,
      "num_tokens": 52707662.0,
      "step": 25405
    },
    {
      "entropy": 5.476322889328003,
      "epoch": 2.3049709724238028,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004467943616676403,
      "loss": 4.9014,
      "mean_token_accuracy": 0.21685788333415984,
      "num_tokens": 52718438.0,
      "step": 25410
    },
    {
      "entropy": 5.4688409805297855,
      "epoch": 2.3054245283018866,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004467734650287911,
      "loss": 4.8951,
      "mean_token_accuracy": 0.2187250018119812,
      "num_tokens": 52729267.0,
      "step": 25415
    },
    {
      "entropy": 5.442692613601684,
      "epoch": 2.305878084179971,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004467525648376229,
      "loss": 4.8686,
      "mean_token_accuracy": 0.22430212795734406,
      "num_tokens": 52738754.0,
      "step": 25420
    },
    {
      "entropy": 5.446636390686035,
      "epoch": 2.306331640058055,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004467316610945679,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22109164148569108,
      "num_tokens": 52749086.0,
      "step": 25425
    },
    {
      "entropy": 5.554483032226562,
      "epoch": 2.3067851959361394,
      "grad_norm": 1.40625,
      "learning_rate": 0.00044671075380005825,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21061239242553711,
      "num_tokens": 52758961.0,
      "step": 25430
    },
    {
      "entropy": 5.557551670074463,
      "epoch": 2.3072387518142237,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004466898429545266,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21503866463899612,
      "num_tokens": 52769447.0,
      "step": 25435
    },
    {
      "entropy": 5.56520643234253,
      "epoch": 2.3076923076923075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044666892855840513,
      "loss": 4.9989,
      "mean_token_accuracy": 0.20663607865571976,
      "num_tokens": 52780306.0,
      "step": 25440
    },
    {
      "entropy": 5.495461368560791,
      "epoch": 2.3081458635703918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004466480106121266,
      "loss": 4.8985,
      "mean_token_accuracy": 0.21171589344739913,
      "num_tokens": 52791263.0,
      "step": 25445
    },
    {
      "entropy": 5.50752534866333,
      "epoch": 2.308599419448476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004466270891161234,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2074984833598137,
      "num_tokens": 52802194.0,
      "step": 25450
    },
    {
      "entropy": 5.508196926116943,
      "epoch": 2.3090529753265603,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044660616407082827,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22165672481060028,
      "num_tokens": 52812391.0,
      "step": 25455
    },
    {
      "entropy": 5.5353086471557615,
      "epoch": 2.3095065312046446,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044658523547667385,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21094203740358353,
      "num_tokens": 52823540.0,
      "step": 25460
    },
    {
      "entropy": 5.435025548934936,
      "epoch": 2.309960087082729,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004465643033340931,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22150187492370604,
      "num_tokens": 52833238.0,
      "step": 25465
    },
    {
      "entropy": 5.5223917961120605,
      "epoch": 2.3104136429608126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044654336764351886,
      "loss": 4.9945,
      "mean_token_accuracy": 0.20679476410150527,
      "num_tokens": 52844510.0,
      "step": 25470
    },
    {
      "entropy": 5.524321269989014,
      "epoch": 2.310867198838897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044652242840538406,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21345377564430237,
      "num_tokens": 52856311.0,
      "step": 25475
    },
    {
      "entropy": 5.480737590789795,
      "epoch": 2.311320754716981,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004465014856201218,
      "loss": 4.8707,
      "mean_token_accuracy": 0.21930532306432723,
      "num_tokens": 52866295.0,
      "step": 25480
    },
    {
      "entropy": 5.487839794158935,
      "epoch": 2.3117743105950654,
      "grad_norm": 1.15625,
      "learning_rate": 0.000446480539288165,
      "loss": 4.95,
      "mean_token_accuracy": 0.21395142525434493,
      "num_tokens": 52876818.0,
      "step": 25485
    },
    {
      "entropy": 5.490958166122437,
      "epoch": 2.3122278664731493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004464595894099469,
      "loss": 4.8855,
      "mean_token_accuracy": 0.21433917433023453,
      "num_tokens": 52886173.0,
      "step": 25490
    },
    {
      "entropy": 5.469081974029541,
      "epoch": 2.3126814223512335,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044643863598590094,
      "loss": 4.8973,
      "mean_token_accuracy": 0.20922717601060867,
      "num_tokens": 52895875.0,
      "step": 25495
    },
    {
      "entropy": 5.533479261398315,
      "epoch": 2.313134978229318,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044641767901646025,
      "loss": 5.0104,
      "mean_token_accuracy": 0.20614977180957794,
      "num_tokens": 52906465.0,
      "step": 25500
    },
    {
      "entropy": 5.549294328689575,
      "epoch": 2.313588534107402,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004463967185020583,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2115151360630989,
      "num_tokens": 52916378.0,
      "step": 25505
    },
    {
      "entropy": 5.508175182342529,
      "epoch": 2.3140420899854863,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004463757544431285,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21898386925458907,
      "num_tokens": 52926698.0,
      "step": 25510
    },
    {
      "entropy": 5.548330211639405,
      "epoch": 2.3144956458635706,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004463547868401044,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2118153303861618,
      "num_tokens": 52936469.0,
      "step": 25515
    },
    {
      "entropy": 5.530008459091187,
      "epoch": 2.3149492017416544,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044633381569341964,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21313218623399735,
      "num_tokens": 52947006.0,
      "step": 25520
    },
    {
      "entropy": 5.379259014129639,
      "epoch": 2.3154027576197387,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004463128410035079,
      "loss": 4.8118,
      "mean_token_accuracy": 0.2164311572909355,
      "num_tokens": 52956663.0,
      "step": 25525
    },
    {
      "entropy": 5.456977510452271,
      "epoch": 2.315856313497823,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044629186277080307,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2170025020837784,
      "num_tokens": 52967383.0,
      "step": 25530
    },
    {
      "entropy": 5.479689693450927,
      "epoch": 2.316309869375907,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044627088099573884,
      "loss": 4.8514,
      "mean_token_accuracy": 0.22086726874113083,
      "num_tokens": 52976567.0,
      "step": 25535
    },
    {
      "entropy": 5.462152719497681,
      "epoch": 2.3167634252539915,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044624989567874907,
      "loss": 4.9266,
      "mean_token_accuracy": 0.21869005858898163,
      "num_tokens": 52986159.0,
      "step": 25540
    },
    {
      "entropy": 5.507683944702149,
      "epoch": 2.3172169811320753,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044622890682026787,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20400695949792863,
      "num_tokens": 52996028.0,
      "step": 25545
    },
    {
      "entropy": 5.556287670135498,
      "epoch": 2.3176705370101596,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004462079144207292,
      "loss": 4.9328,
      "mean_token_accuracy": 0.2133473575115204,
      "num_tokens": 53006446.0,
      "step": 25550
    },
    {
      "entropy": 5.5622576713562015,
      "epoch": 2.318124092888244,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044618691848056726,
      "loss": 4.8883,
      "mean_token_accuracy": 0.21131509095430373,
      "num_tokens": 53016497.0,
      "step": 25555
    },
    {
      "entropy": 5.541997241973877,
      "epoch": 2.318577648766328,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004461659190002162,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22148124426603316,
      "num_tokens": 53027811.0,
      "step": 25560
    },
    {
      "entropy": 5.512529754638672,
      "epoch": 2.3190312046444124,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044614491598011027,
      "loss": 4.9356,
      "mean_token_accuracy": 0.20826315879821777,
      "num_tokens": 53038206.0,
      "step": 25565
    },
    {
      "entropy": 5.546404886245727,
      "epoch": 2.319484760522496,
      "grad_norm": 1.125,
      "learning_rate": 0.0004461239094206839,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21190930157899857,
      "num_tokens": 53047974.0,
      "step": 25570
    },
    {
      "entropy": 5.54844446182251,
      "epoch": 2.3199383164005805,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044610289932237154,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2132696807384491,
      "num_tokens": 53057904.0,
      "step": 25575
    },
    {
      "entropy": 5.4844156265258786,
      "epoch": 2.3203918722786647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044608188568560756,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2118034243583679,
      "num_tokens": 53068447.0,
      "step": 25580
    },
    {
      "entropy": 5.484172201156616,
      "epoch": 2.320845428156749,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004460608685108266,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20456577688455582,
      "num_tokens": 53079369.0,
      "step": 25585
    },
    {
      "entropy": 5.568174362182617,
      "epoch": 2.3212989840348333,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044603984779846334,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20579463690519334,
      "num_tokens": 53090794.0,
      "step": 25590
    },
    {
      "entropy": 5.534527587890625,
      "epoch": 2.321752539912917,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004460188235489524,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2139367565512657,
      "num_tokens": 53100802.0,
      "step": 25595
    },
    {
      "entropy": 5.459441089630127,
      "epoch": 2.3222060957910013,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004459977957627285,
      "loss": 4.946,
      "mean_token_accuracy": 0.21548793464899063,
      "num_tokens": 53111950.0,
      "step": 25600
    },
    {
      "entropy": 5.396631717681885,
      "epoch": 2.3226596516690856,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044597676444022683,
      "loss": 4.7942,
      "mean_token_accuracy": 0.21895018368959426,
      "num_tokens": 53121908.0,
      "step": 25605
    },
    {
      "entropy": 5.481389951705933,
      "epoch": 2.32311320754717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044595572958188194,
      "loss": 4.9351,
      "mean_token_accuracy": 0.20947088152170182,
      "num_tokens": 53133065.0,
      "step": 25610
    },
    {
      "entropy": 5.4266551494598385,
      "epoch": 2.323566763425254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004459346911881291,
      "loss": 4.8774,
      "mean_token_accuracy": 0.2153947174549103,
      "num_tokens": 53144135.0,
      "step": 25615
    },
    {
      "entropy": 5.5237603187561035,
      "epoch": 2.3240203193033384,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004459136492594032,
      "loss": 4.8597,
      "mean_token_accuracy": 0.22346723824739456,
      "num_tokens": 53153916.0,
      "step": 25620
    },
    {
      "entropy": 5.4423905372619625,
      "epoch": 2.3244738751814222,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004458926037961394,
      "loss": 4.8389,
      "mean_token_accuracy": 0.2224319487810135,
      "num_tokens": 53163937.0,
      "step": 25625
    },
    {
      "entropy": 5.533673858642578,
      "epoch": 2.3249274310595065,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004458715547987732,
      "loss": 4.9998,
      "mean_token_accuracy": 0.20808079242706298,
      "num_tokens": 53174327.0,
      "step": 25630
    },
    {
      "entropy": 5.5080187797546385,
      "epoch": 2.3253809869375908,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004458505022677396,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21214041560888292,
      "num_tokens": 53185946.0,
      "step": 25635
    },
    {
      "entropy": 5.513505268096924,
      "epoch": 2.325834542815675,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004458294462034741,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2171740263700485,
      "num_tokens": 53195944.0,
      "step": 25640
    },
    {
      "entropy": 5.467567300796508,
      "epoch": 2.326288098693759,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004458083866064121,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21866397559642792,
      "num_tokens": 53206054.0,
      "step": 25645
    },
    {
      "entropy": 5.481159400939942,
      "epoch": 2.326741654571843,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004457873234769891,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2143828585743904,
      "num_tokens": 53216923.0,
      "step": 25650
    },
    {
      "entropy": 5.468651008605957,
      "epoch": 2.3271952104499274,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044576625681564086,
      "loss": 4.9121,
      "mean_token_accuracy": 0.21062431782484053,
      "num_tokens": 53226788.0,
      "step": 25655
    },
    {
      "entropy": 5.522043228149414,
      "epoch": 2.3276487663280117,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044574518662280276,
      "loss": 4.9657,
      "mean_token_accuracy": 0.21107705533504487,
      "num_tokens": 53237573.0,
      "step": 25660
    },
    {
      "entropy": 5.550842905044556,
      "epoch": 2.328102322206096,
      "grad_norm": 1.125,
      "learning_rate": 0.0004457241128989108,
      "loss": 4.974,
      "mean_token_accuracy": 0.20487329214811326,
      "num_tokens": 53247724.0,
      "step": 25665
    },
    {
      "entropy": 5.532074975967407,
      "epoch": 2.32855587808418,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004457030356444006,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21214183270931244,
      "num_tokens": 53258774.0,
      "step": 25670
    },
    {
      "entropy": 5.471618604660034,
      "epoch": 2.329009433962264,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004456819548597082,
      "loss": 4.8936,
      "mean_token_accuracy": 0.21844785660505295,
      "num_tokens": 53268550.0,
      "step": 25675
    },
    {
      "entropy": 5.393625116348266,
      "epoch": 2.3294629898403483,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044566087054526946,
      "loss": 4.8667,
      "mean_token_accuracy": 0.21955114752054214,
      "num_tokens": 53277975.0,
      "step": 25680
    },
    {
      "entropy": 5.568359804153443,
      "epoch": 2.3299165457184325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004456397827015204,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21184577643871308,
      "num_tokens": 53287939.0,
      "step": 25685
    },
    {
      "entropy": 5.560759496688843,
      "epoch": 2.330370101596517,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004456186913288971,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21400628089904786,
      "num_tokens": 53298355.0,
      "step": 25690
    },
    {
      "entropy": 5.583919286727905,
      "epoch": 2.3308236574746006,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004455975964278359,
      "loss": 5.0282,
      "mean_token_accuracy": 0.21015584617853164,
      "num_tokens": 53308513.0,
      "step": 25695
    },
    {
      "entropy": 5.55291781425476,
      "epoch": 2.331277213352685,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044557649799877285,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2102041095495224,
      "num_tokens": 53318249.0,
      "step": 25700
    },
    {
      "entropy": 5.467944049835205,
      "epoch": 2.331730769230769,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004455553960421443,
      "loss": 4.8472,
      "mean_token_accuracy": 0.21671323627233505,
      "num_tokens": 53328460.0,
      "step": 25705
    },
    {
      "entropy": 5.483862495422363,
      "epoch": 2.3321843251088534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044553429055838685,
      "loss": 4.9131,
      "mean_token_accuracy": 0.2191723644733429,
      "num_tokens": 53338984.0,
      "step": 25710
    },
    {
      "entropy": 5.550682020187378,
      "epoch": 2.3326378809869377,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044551318154793676,
      "loss": 4.972,
      "mean_token_accuracy": 0.2095512792468071,
      "num_tokens": 53348092.0,
      "step": 25715
    },
    {
      "entropy": 5.47348198890686,
      "epoch": 2.333091436865022,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044549206901123056,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2167353004217148,
      "num_tokens": 53357343.0,
      "step": 25720
    },
    {
      "entropy": 5.468302297592163,
      "epoch": 2.333544992743106,
      "grad_norm": 1.0625,
      "learning_rate": 0.000445470952948705,
      "loss": 4.8351,
      "mean_token_accuracy": 0.21819607317447662,
      "num_tokens": 53367343.0,
      "step": 25725
    },
    {
      "entropy": 5.569767570495605,
      "epoch": 2.33399854862119,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044544983336079665,
      "loss": 5.0161,
      "mean_token_accuracy": 0.2124939501285553,
      "num_tokens": 53379093.0,
      "step": 25730
    },
    {
      "entropy": 5.490252065658569,
      "epoch": 2.3344521044992743,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004454287102479423,
      "loss": 4.8819,
      "mean_token_accuracy": 0.21499982327222825,
      "num_tokens": 53388883.0,
      "step": 25735
    },
    {
      "entropy": 5.523770952224732,
      "epoch": 2.3349056603773586,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004454075836105788,
      "loss": 4.898,
      "mean_token_accuracy": 0.2194881409406662,
      "num_tokens": 53399393.0,
      "step": 25740
    },
    {
      "entropy": 5.4959920883178714,
      "epoch": 2.335359216255443,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044538645344914307,
      "loss": 4.855,
      "mean_token_accuracy": 0.21944254487752915,
      "num_tokens": 53408545.0,
      "step": 25745
    },
    {
      "entropy": 5.484006595611572,
      "epoch": 2.3358127721335267,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004453653197640721,
      "loss": 4.9554,
      "mean_token_accuracy": 0.20721954256296157,
      "num_tokens": 53418325.0,
      "step": 25750
    },
    {
      "entropy": 5.512723398208618,
      "epoch": 2.336266328011611,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044534418255580286,
      "loss": 4.937,
      "mean_token_accuracy": 0.20314341336488723,
      "num_tokens": 53427984.0,
      "step": 25755
    },
    {
      "entropy": 5.544159555435181,
      "epoch": 2.336719883889695,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044532304182477245,
      "loss": 5.0251,
      "mean_token_accuracy": 0.20258364975452423,
      "num_tokens": 53438765.0,
      "step": 25760
    },
    {
      "entropy": 5.535161638259888,
      "epoch": 2.3371734397677795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004453018975714182,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21545352190732955,
      "num_tokens": 53448792.0,
      "step": 25765
    },
    {
      "entropy": 5.568411731719971,
      "epoch": 2.3376269956458637,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004452807497961773,
      "loss": 4.988,
      "mean_token_accuracy": 0.21059752851724625,
      "num_tokens": 53459334.0,
      "step": 25770
    },
    {
      "entropy": 5.5368695735931395,
      "epoch": 2.3380805515239476,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044525959849948704,
      "loss": 4.991,
      "mean_token_accuracy": 0.21514967530965806,
      "num_tokens": 53469841.0,
      "step": 25775
    },
    {
      "entropy": 5.489812278747559,
      "epoch": 2.338534107402032,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044523844368178503,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2160777449607849,
      "num_tokens": 53480660.0,
      "step": 25780
    },
    {
      "entropy": 5.5110344886779785,
      "epoch": 2.338987663280116,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004452172853435085,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2163378193974495,
      "num_tokens": 53490276.0,
      "step": 25785
    },
    {
      "entropy": 5.404148864746094,
      "epoch": 2.3394412191582004,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004451961234850953,
      "loss": 4.8419,
      "mean_token_accuracy": 0.2290509432554245,
      "num_tokens": 53501100.0,
      "step": 25790
    },
    {
      "entropy": 5.450394058227539,
      "epoch": 2.3398947750362846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004451749581069827,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21793881058692932,
      "num_tokens": 53511680.0,
      "step": 25795
    },
    {
      "entropy": 5.532208204269409,
      "epoch": 2.3403483309143684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004451537892096087,
      "loss": 4.995,
      "mean_token_accuracy": 0.2163803219795227,
      "num_tokens": 53521298.0,
      "step": 25800
    },
    {
      "entropy": 5.488109350204468,
      "epoch": 2.3408018867924527,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044513261679341094,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22513474076986312,
      "num_tokens": 53530413.0,
      "step": 25805
    },
    {
      "entropy": 5.457996082305908,
      "epoch": 2.341255442670537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004451114408588273,
      "loss": 4.9363,
      "mean_token_accuracy": 0.21366408765316008,
      "num_tokens": 53540959.0,
      "step": 25810
    },
    {
      "entropy": 5.519728374481201,
      "epoch": 2.3417089985486212,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044509026140629584,
      "loss": 4.924,
      "mean_token_accuracy": 0.21457249969244002,
      "num_tokens": 53551265.0,
      "step": 25815
    },
    {
      "entropy": 5.5442688941955565,
      "epoch": 2.3421625544267055,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044506907843625427,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21055022180080413,
      "num_tokens": 53562666.0,
      "step": 25820
    },
    {
      "entropy": 5.478188943862915,
      "epoch": 2.34261611030479,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044504789194914086,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20967674404382705,
      "num_tokens": 53573688.0,
      "step": 25825
    },
    {
      "entropy": 5.565692472457886,
      "epoch": 2.3430696661828736,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044502670194539374,
      "loss": 4.8968,
      "mean_token_accuracy": 0.21761015355587005,
      "num_tokens": 53584055.0,
      "step": 25830
    },
    {
      "entropy": 5.444281530380249,
      "epoch": 2.343523222060958,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044500550842545104,
      "loss": 4.7926,
      "mean_token_accuracy": 0.22537012845277787,
      "num_tokens": 53593765.0,
      "step": 25835
    },
    {
      "entropy": 5.417661714553833,
      "epoch": 2.343976777939042,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004449843113897512,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22143977731466294,
      "num_tokens": 53603480.0,
      "step": 25840
    },
    {
      "entropy": 5.477065086364746,
      "epoch": 2.3444303338171264,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004449631108387325,
      "loss": 4.9025,
      "mean_token_accuracy": 0.22046411037445068,
      "num_tokens": 53613797.0,
      "step": 25845
    },
    {
      "entropy": 5.445242309570313,
      "epoch": 2.34488388969521,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004449419067728333,
      "loss": 4.8732,
      "mean_token_accuracy": 0.21652236133813857,
      "num_tokens": 53623646.0,
      "step": 25850
    },
    {
      "entropy": 5.459939336776733,
      "epoch": 2.3453374455732945,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044492069919249206,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21198905408382415,
      "num_tokens": 53634118.0,
      "step": 25855
    },
    {
      "entropy": 5.520653486251831,
      "epoch": 2.3457910014513788,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004448994880981476,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21317032873630523,
      "num_tokens": 53644370.0,
      "step": 25860
    },
    {
      "entropy": 5.529612970352173,
      "epoch": 2.346244557329463,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044487827349023833,
      "loss": 4.9651,
      "mean_token_accuracy": 0.20310250967741011,
      "num_tokens": 53655560.0,
      "step": 25865
    },
    {
      "entropy": 5.515697336196899,
      "epoch": 2.3466981132075473,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044485705536920315,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2099432587623596,
      "num_tokens": 53666401.0,
      "step": 25870
    },
    {
      "entropy": 5.562315988540649,
      "epoch": 2.3471516690856316,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044483583373548075,
      "loss": 4.9629,
      "mean_token_accuracy": 0.22292011529207229,
      "num_tokens": 53676133.0,
      "step": 25875
    },
    {
      "entropy": 5.482073259353638,
      "epoch": 2.3476052249637154,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004448146085895101,
      "loss": 4.933,
      "mean_token_accuracy": 0.20581683963537217,
      "num_tokens": 53687683.0,
      "step": 25880
    },
    {
      "entropy": 5.464105224609375,
      "epoch": 2.3480587808417996,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004447933799317299,
      "loss": 4.9946,
      "mean_token_accuracy": 0.21131130307912827,
      "num_tokens": 53698343.0,
      "step": 25885
    },
    {
      "entropy": 5.527396249771118,
      "epoch": 2.348512336719884,
      "grad_norm": 1.125,
      "learning_rate": 0.0004447721477625795,
      "loss": 4.9,
      "mean_token_accuracy": 0.21536218971014023,
      "num_tokens": 53709014.0,
      "step": 25890
    },
    {
      "entropy": 5.485438013076783,
      "epoch": 2.348965892597968,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044475091208249774,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2127739116549492,
      "num_tokens": 53718838.0,
      "step": 25895
    },
    {
      "entropy": 5.433662223815918,
      "epoch": 2.3494194484760524,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004447296728919239,
      "loss": 4.8893,
      "mean_token_accuracy": 0.21522442400455474,
      "num_tokens": 53728526.0,
      "step": 25900
    },
    {
      "entropy": 5.528424024581909,
      "epoch": 2.3498730043541363,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004447084301912972,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21140301525592803,
      "num_tokens": 53739261.0,
      "step": 25905
    },
    {
      "entropy": 5.5276045322418215,
      "epoch": 2.3503265602322205,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004446871839810568,
      "loss": 4.9733,
      "mean_token_accuracy": 0.2137870192527771,
      "num_tokens": 53749599.0,
      "step": 25910
    },
    {
      "entropy": 5.527643632888794,
      "epoch": 2.350780116110305,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044466593426164224,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2095325067639351,
      "num_tokens": 53759898.0,
      "step": 25915
    },
    {
      "entropy": 5.498171854019165,
      "epoch": 2.351233671988389,
      "grad_norm": 1.046875,
      "learning_rate": 0.000444644681033493,
      "loss": 4.8974,
      "mean_token_accuracy": 0.20869439095258713,
      "num_tokens": 53770561.0,
      "step": 25920
    },
    {
      "entropy": 5.4753681182861325,
      "epoch": 2.3516872278664733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044462342429704845,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21487085074186324,
      "num_tokens": 53780408.0,
      "step": 25925
    },
    {
      "entropy": 5.407979679107666,
      "epoch": 2.352140783744557,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044460216405274827,
      "loss": 4.8391,
      "mean_token_accuracy": 0.22644920349121095,
      "num_tokens": 53789125.0,
      "step": 25930
    },
    {
      "entropy": 5.458460998535156,
      "epoch": 2.3525943396226414,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044458090030103216,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22322857975959778,
      "num_tokens": 53799164.0,
      "step": 25935
    },
    {
      "entropy": 5.493204164505005,
      "epoch": 2.3530478955007257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044455963304233976,
      "loss": 4.8525,
      "mean_token_accuracy": 0.22303333431482314,
      "num_tokens": 53810710.0,
      "step": 25940
    },
    {
      "entropy": 5.420898866653443,
      "epoch": 2.35350145137881,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000444538362277111,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2077305406332016,
      "num_tokens": 53820692.0,
      "step": 25945
    },
    {
      "entropy": 5.5188360691070555,
      "epoch": 2.353955007256894,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004445170880057857,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21033201664686202,
      "num_tokens": 53831082.0,
      "step": 25950
    },
    {
      "entropy": 5.5739295959472654,
      "epoch": 2.354408563134978,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044449581022880375,
      "loss": 5.0053,
      "mean_token_accuracy": 0.19920249432325363,
      "num_tokens": 53842602.0,
      "step": 25955
    },
    {
      "entropy": 5.499769926071167,
      "epoch": 2.3548621190130623,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004444745289466053,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21018957048654557,
      "num_tokens": 53853784.0,
      "step": 25960
    },
    {
      "entropy": 5.416641759872436,
      "epoch": 2.3553156748911466,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044445324415963043,
      "loss": 4.8122,
      "mean_token_accuracy": 0.22084206938743592,
      "num_tokens": 53863526.0,
      "step": 25965
    },
    {
      "entropy": 5.424200487136841,
      "epoch": 2.355769230769231,
      "grad_norm": 1.125,
      "learning_rate": 0.0004444319558683192,
      "loss": 4.9397,
      "mean_token_accuracy": 0.21259866505861283,
      "num_tokens": 53874594.0,
      "step": 25970
    },
    {
      "entropy": 5.5079224586486815,
      "epoch": 2.356222786647315,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044441066407311205,
      "loss": 4.9069,
      "mean_token_accuracy": 0.21742993742227554,
      "num_tokens": 53884771.0,
      "step": 25975
    },
    {
      "entropy": 5.52966456413269,
      "epoch": 2.3566763425253994,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004443893687744492,
      "loss": 4.9162,
      "mean_token_accuracy": 0.2155672937631607,
      "num_tokens": 53895181.0,
      "step": 25980
    },
    {
      "entropy": 5.432828760147094,
      "epoch": 2.357129898403483,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044436806997277096,
      "loss": 4.7982,
      "mean_token_accuracy": 0.22504699379205703,
      "num_tokens": 53904153.0,
      "step": 25985
    },
    {
      "entropy": 5.425666093826294,
      "epoch": 2.3575834542815675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044434676766851787,
      "loss": 4.8836,
      "mean_token_accuracy": 0.21994280815124512,
      "num_tokens": 53914064.0,
      "step": 25990
    },
    {
      "entropy": 5.499133539199829,
      "epoch": 2.3580370101596517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044432546186213054,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21021739542484283,
      "num_tokens": 53924240.0,
      "step": 25995
    },
    {
      "entropy": 5.551114273071289,
      "epoch": 2.358490566037736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044430415255404944,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21483293026685715,
      "num_tokens": 53935395.0,
      "step": 26000
    },
    {
      "entropy": 5.522525310516357,
      "epoch": 2.35894412191582,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044428283974471533,
      "loss": 4.9026,
      "mean_token_accuracy": 0.21348591297864913,
      "num_tokens": 53945768.0,
      "step": 26005
    },
    {
      "entropy": 5.429514694213867,
      "epoch": 2.359397677793904,
      "grad_norm": 1.15625,
      "learning_rate": 0.000444261523434569,
      "loss": 4.8906,
      "mean_token_accuracy": 0.21914070546627046,
      "num_tokens": 53956467.0,
      "step": 26010
    },
    {
      "entropy": 5.420666313171386,
      "epoch": 2.3598512336719883,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044424020362405126,
      "loss": 4.8448,
      "mean_token_accuracy": 0.21922084242105483,
      "num_tokens": 53967587.0,
      "step": 26015
    },
    {
      "entropy": 5.4528296947479244,
      "epoch": 2.3603047895500726,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004442188803136029,
      "loss": 4.8843,
      "mean_token_accuracy": 0.21346233785152435,
      "num_tokens": 53977691.0,
      "step": 26020
    },
    {
      "entropy": 5.474106359481811,
      "epoch": 2.360758345428157,
      "grad_norm": 1.1875,
      "learning_rate": 0.000444197553503665,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20730575323104858,
      "num_tokens": 53987608.0,
      "step": 26025
    },
    {
      "entropy": 5.5305156230926515,
      "epoch": 2.361211901306241,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044417622319467857,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20671143680810927,
      "num_tokens": 53998010.0,
      "step": 26030
    },
    {
      "entropy": 5.461687660217285,
      "epoch": 2.361665457184325,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044415488938708475,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21553837209939958,
      "num_tokens": 54009852.0,
      "step": 26035
    },
    {
      "entropy": 5.487236881256104,
      "epoch": 2.3621190130624092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004441335520813247,
      "loss": 4.8689,
      "mean_token_accuracy": 0.21619959622621537,
      "num_tokens": 54019635.0,
      "step": 26040
    },
    {
      "entropy": 5.490652942657471,
      "epoch": 2.3625725689404935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044411221127783966,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2171410545706749,
      "num_tokens": 54030046.0,
      "step": 26045
    },
    {
      "entropy": 5.493072986602783,
      "epoch": 2.3630261248185778,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004440908669770711,
      "loss": 4.9006,
      "mean_token_accuracy": 0.2116263523697853,
      "num_tokens": 54041197.0,
      "step": 26050
    },
    {
      "entropy": 5.496097707748413,
      "epoch": 2.3634796806966616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004440695191794602,
      "loss": 4.892,
      "mean_token_accuracy": 0.2177186354994774,
      "num_tokens": 54051592.0,
      "step": 26055
    },
    {
      "entropy": 5.5228636264801025,
      "epoch": 2.363933236574746,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004440481678854487,
      "loss": 4.971,
      "mean_token_accuracy": 0.20590161830186843,
      "num_tokens": 54062893.0,
      "step": 26060
    },
    {
      "entropy": 5.540108299255371,
      "epoch": 2.36438679245283,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044402681309547786,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20974503606557846,
      "num_tokens": 54073651.0,
      "step": 26065
    },
    {
      "entropy": 5.472658491134643,
      "epoch": 2.3648403483309144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044400545480998956,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21841016560792922,
      "num_tokens": 54084889.0,
      "step": 26070
    },
    {
      "entropy": 5.446832990646362,
      "epoch": 2.3652939042089987,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044398409302942533,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2129442349076271,
      "num_tokens": 54095724.0,
      "step": 26075
    },
    {
      "entropy": 5.469037294387817,
      "epoch": 2.365747460087083,
      "grad_norm": 1.171875,
      "learning_rate": 0.000443962727754227,
      "loss": 4.8798,
      "mean_token_accuracy": 0.2137051820755005,
      "num_tokens": 54106581.0,
      "step": 26080
    },
    {
      "entropy": 5.522040557861328,
      "epoch": 2.3662010159651667,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004439413589848364,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21595065742731095,
      "num_tokens": 54116074.0,
      "step": 26085
    },
    {
      "entropy": 5.498589754104614,
      "epoch": 2.366654571843251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004439199867216955,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21261702626943588,
      "num_tokens": 54126902.0,
      "step": 26090
    },
    {
      "entropy": 5.495638275146485,
      "epoch": 2.3671081277213353,
      "grad_norm": 1.125,
      "learning_rate": 0.0004438986109652462,
      "loss": 4.8815,
      "mean_token_accuracy": 0.2148958757519722,
      "num_tokens": 54137554.0,
      "step": 26095
    },
    {
      "entropy": 5.486440753936767,
      "epoch": 2.3675616835994195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004438772317159306,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2148678183555603,
      "num_tokens": 54148035.0,
      "step": 26100
    },
    {
      "entropy": 5.456903886795044,
      "epoch": 2.368015239477504,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044385584897419075,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22034476548433304,
      "num_tokens": 54157417.0,
      "step": 26105
    },
    {
      "entropy": 5.511321973800659,
      "epoch": 2.3684687953555876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000443834462740469,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2103291317820549,
      "num_tokens": 54168353.0,
      "step": 26110
    },
    {
      "entropy": 5.471779298782349,
      "epoch": 2.368922351233672,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004438130730152074,
      "loss": 4.9854,
      "mean_token_accuracy": 0.2033338651061058,
      "num_tokens": 54180182.0,
      "step": 26115
    },
    {
      "entropy": 5.530212879180908,
      "epoch": 2.369375907111756,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044379167979884847,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21144528388977052,
      "num_tokens": 54190973.0,
      "step": 26120
    },
    {
      "entropy": 5.562690353393554,
      "epoch": 2.3698294629898404,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004437702830918347,
      "loss": 5.03,
      "mean_token_accuracy": 0.2066943496465683,
      "num_tokens": 54202077.0,
      "step": 26125
    },
    {
      "entropy": 5.502487277984619,
      "epoch": 2.3702830188679247,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004437488828946084,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21228874623775482,
      "num_tokens": 54211778.0,
      "step": 26130
    },
    {
      "entropy": 5.44524655342102,
      "epoch": 2.3707365747460085,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004437274792076122,
      "loss": 4.7989,
      "mean_token_accuracy": 0.22863289564847947,
      "num_tokens": 54220941.0,
      "step": 26135
    },
    {
      "entropy": 5.482705020904541,
      "epoch": 2.371190130624093,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044370607203128874,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21129117906093597,
      "num_tokens": 54231602.0,
      "step": 26140
    },
    {
      "entropy": 5.491693687438965,
      "epoch": 2.371643686502177,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004436846613660807,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21241433024406434,
      "num_tokens": 54241797.0,
      "step": 26145
    },
    {
      "entropy": 5.547004127502442,
      "epoch": 2.3720972423802613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004436632472124308,
      "loss": 4.9727,
      "mean_token_accuracy": 0.20779201984405518,
      "num_tokens": 54253085.0,
      "step": 26150
    },
    {
      "entropy": 5.567381143569946,
      "epoch": 2.3725507982583456,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004436418295707821,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21601478159427642,
      "num_tokens": 54262164.0,
      "step": 26155
    },
    {
      "entropy": 5.498871231079102,
      "epoch": 2.3730043541364294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004436204084415774,
      "loss": 4.9499,
      "mean_token_accuracy": 0.20918332785367966,
      "num_tokens": 54271922.0,
      "step": 26160
    },
    {
      "entropy": 5.469789838790893,
      "epoch": 2.3734579100145137,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004435989838252596,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22025145292282106,
      "num_tokens": 54281738.0,
      "step": 26165
    },
    {
      "entropy": 5.4567787647247314,
      "epoch": 2.373911465892598,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004435775557222719,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2132453814148903,
      "num_tokens": 54292252.0,
      "step": 26170
    },
    {
      "entropy": 5.496927547454834,
      "epoch": 2.374365021770682,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044355612413305736,
      "loss": 4.965,
      "mean_token_accuracy": 0.21742047667503356,
      "num_tokens": 54302518.0,
      "step": 26175
    },
    {
      "entropy": 5.512274074554443,
      "epoch": 2.3748185776487665,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004435346890580593,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21677685528993607,
      "num_tokens": 54312917.0,
      "step": 26180
    },
    {
      "entropy": 5.516729784011841,
      "epoch": 2.3752721335268507,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044351325049772085,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2186713546514511,
      "num_tokens": 54322552.0,
      "step": 26185
    },
    {
      "entropy": 5.515956211090088,
      "epoch": 2.3757256894049346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044349180845248536,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2173202946782112,
      "num_tokens": 54332917.0,
      "step": 26190
    },
    {
      "entropy": 5.437029409408569,
      "epoch": 2.376179245283019,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004434703629227964,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22058886289596558,
      "num_tokens": 54342791.0,
      "step": 26195
    },
    {
      "entropy": 5.387643671035766,
      "epoch": 2.376632801161103,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004434489139090975,
      "loss": 4.7979,
      "mean_token_accuracy": 0.22771116346120834,
      "num_tokens": 54352821.0,
      "step": 26200
    },
    {
      "entropy": 5.5476642608642575,
      "epoch": 2.3770863570391874,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044342746141183206,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20798380225896834,
      "num_tokens": 54363095.0,
      "step": 26205
    },
    {
      "entropy": 5.477079010009765,
      "epoch": 2.377539912917271,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004434060054314438,
      "loss": 4.8901,
      "mean_token_accuracy": 0.20930180102586746,
      "num_tokens": 54373841.0,
      "step": 26210
    },
    {
      "entropy": 5.434181213378906,
      "epoch": 2.3779934687953554,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044338454596837637,
      "loss": 4.8723,
      "mean_token_accuracy": 0.2124147817492485,
      "num_tokens": 54385012.0,
      "step": 26215
    },
    {
      "entropy": 5.492944765090942,
      "epoch": 2.3784470246734397,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004433630830230737,
      "loss": 4.9198,
      "mean_token_accuracy": 0.21250682771205903,
      "num_tokens": 54395298.0,
      "step": 26220
    },
    {
      "entropy": 5.4894873142242435,
      "epoch": 2.378900580551524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004433416165959795,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2137940987944603,
      "num_tokens": 54405752.0,
      "step": 26225
    },
    {
      "entropy": 5.495602798461914,
      "epoch": 2.3793541364296082,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004433201466875378,
      "loss": 4.9264,
      "mean_token_accuracy": 0.2129967987537384,
      "num_tokens": 54416607.0,
      "step": 26230
    },
    {
      "entropy": 5.474749183654785,
      "epoch": 2.3798076923076925,
      "grad_norm": 1.125,
      "learning_rate": 0.0004432986732981926,
      "loss": 4.895,
      "mean_token_accuracy": 0.22310136258602142,
      "num_tokens": 54426873.0,
      "step": 26235
    },
    {
      "entropy": 5.434404277801514,
      "epoch": 2.3802612481857763,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000443277196428388,
      "loss": 4.8331,
      "mean_token_accuracy": 0.2220876157283783,
      "num_tokens": 54437611.0,
      "step": 26240
    },
    {
      "entropy": 5.542353963851928,
      "epoch": 2.3807148040638606,
      "grad_norm": 1.078125,
      "learning_rate": 0.000443255716078568,
      "loss": 4.9139,
      "mean_token_accuracy": 0.2206832766532898,
      "num_tokens": 54448992.0,
      "step": 26245
    },
    {
      "entropy": 5.47555890083313,
      "epoch": 2.381168359941945,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004432342322491769,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21989913582801818,
      "num_tokens": 54459172.0,
      "step": 26250
    },
    {
      "entropy": 5.4420390129089355,
      "epoch": 2.381621915820029,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044321274494065903,
      "loss": 4.8739,
      "mean_token_accuracy": 0.21507968306541442,
      "num_tokens": 54469660.0,
      "step": 26255
    },
    {
      "entropy": 5.536232662200928,
      "epoch": 2.3820754716981134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044319125415345886,
      "loss": 5.0204,
      "mean_token_accuracy": 0.2102372020483017,
      "num_tokens": 54479567.0,
      "step": 26260
    },
    {
      "entropy": 5.486782455444336,
      "epoch": 2.3825290275761972,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004431697598880205,
      "loss": 4.963,
      "mean_token_accuracy": 0.21284139454364776,
      "num_tokens": 54489690.0,
      "step": 26265
    },
    {
      "entropy": 5.489284896850586,
      "epoch": 2.3829825834542815,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044314826214478877,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21616387367248535,
      "num_tokens": 54500876.0,
      "step": 26270
    },
    {
      "entropy": 5.502010679244995,
      "epoch": 2.3834361393323658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044312676092420804,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2057827368378639,
      "num_tokens": 54510988.0,
      "step": 26275
    },
    {
      "entropy": 5.509961318969727,
      "epoch": 2.38388969521045,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004431052562267231,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22014069557189941,
      "num_tokens": 54522149.0,
      "step": 26280
    },
    {
      "entropy": 5.582751083374023,
      "epoch": 2.3843432510885343,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044308374805277867,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20298152565956115,
      "num_tokens": 54533482.0,
      "step": 26285
    },
    {
      "entropy": 5.472028589248657,
      "epoch": 2.384796806966618,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004430622364028194,
      "loss": 4.98,
      "mean_token_accuracy": 0.20509568005800247,
      "num_tokens": 54543058.0,
      "step": 26290
    },
    {
      "entropy": 5.510845375061035,
      "epoch": 2.3852503628447024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044304072127729035,
      "loss": 4.9237,
      "mean_token_accuracy": 0.21259500980377197,
      "num_tokens": 54552655.0,
      "step": 26295
    },
    {
      "entropy": 5.4802491664886475,
      "epoch": 2.3857039187227866,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044301920267663633,
      "loss": 4.8688,
      "mean_token_accuracy": 0.20867816507816314,
      "num_tokens": 54562696.0,
      "step": 26300
    },
    {
      "entropy": 5.503185701370239,
      "epoch": 2.386157474600871,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004429976806013024,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2215511053800583,
      "num_tokens": 54572803.0,
      "step": 26305
    },
    {
      "entropy": 5.469847917556763,
      "epoch": 2.386611030478955,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004429761550517336,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21451132893562316,
      "num_tokens": 54582450.0,
      "step": 26310
    },
    {
      "entropy": 5.40881462097168,
      "epoch": 2.387064586357039,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044295462602837504,
      "loss": 4.8847,
      "mean_token_accuracy": 0.21612351536750793,
      "num_tokens": 54591868.0,
      "step": 26315
    },
    {
      "entropy": 5.483367490768432,
      "epoch": 2.3875181422351233,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004429330935316721,
      "loss": 4.9014,
      "mean_token_accuracy": 0.21119055896997452,
      "num_tokens": 54601858.0,
      "step": 26320
    },
    {
      "entropy": 5.493379211425781,
      "epoch": 2.3879716981132075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044291155756206996,
      "loss": 4.945,
      "mean_token_accuracy": 0.2118200033903122,
      "num_tokens": 54612243.0,
      "step": 26325
    },
    {
      "entropy": 5.489482116699219,
      "epoch": 2.388425253991292,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044289001812001404,
      "loss": 4.8372,
      "mean_token_accuracy": 0.21867071837186813,
      "num_tokens": 54622365.0,
      "step": 26330
    },
    {
      "entropy": 5.478691244125367,
      "epoch": 2.388878809869376,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004428684752059497,
      "loss": 4.8365,
      "mean_token_accuracy": 0.2181360512971878,
      "num_tokens": 54632356.0,
      "step": 26335
    },
    {
      "entropy": 5.5544685363769535,
      "epoch": 2.3893323657474603,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004428469288203226,
      "loss": 4.9463,
      "mean_token_accuracy": 0.20535167008638383,
      "num_tokens": 54642976.0,
      "step": 26340
    },
    {
      "entropy": 5.578728580474854,
      "epoch": 2.389785921625544,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004428253789635783,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21630270928144454,
      "num_tokens": 54652597.0,
      "step": 26345
    },
    {
      "entropy": 5.464432954788208,
      "epoch": 2.3902394775036284,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004428038256361622,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20704686939716338,
      "num_tokens": 54662930.0,
      "step": 26350
    },
    {
      "entropy": 5.591292905807495,
      "epoch": 2.3906930333817127,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004427822688385203,
      "loss": 5.1322,
      "mean_token_accuracy": 0.2049167960882187,
      "num_tokens": 54674562.0,
      "step": 26355
    },
    {
      "entropy": 5.583617067337036,
      "epoch": 2.391146589259797,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004427607085710984,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21285902559757233,
      "num_tokens": 54684432.0,
      "step": 26360
    },
    {
      "entropy": 5.574669933319091,
      "epoch": 2.3916001451378808,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044273914483434226,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21097928732633592,
      "num_tokens": 54694060.0,
      "step": 26365
    },
    {
      "entropy": 5.462815141677856,
      "epoch": 2.392053701015965,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044271757762869777,
      "loss": 4.9426,
      "mean_token_accuracy": 0.20573419481515884,
      "num_tokens": 54704411.0,
      "step": 26370
    },
    {
      "entropy": 5.470274353027344,
      "epoch": 2.3925072568940493,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004426960069546112,
      "loss": 4.9595,
      "mean_token_accuracy": 0.21424695253372192,
      "num_tokens": 54715185.0,
      "step": 26375
    },
    {
      "entropy": 5.51666488647461,
      "epoch": 2.3929608127721336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004426744328125284,
      "loss": 4.8388,
      "mean_token_accuracy": 0.22204079627990722,
      "num_tokens": 54725529.0,
      "step": 26380
    },
    {
      "entropy": 5.599893188476562,
      "epoch": 2.393414368650218,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004426528552028955,
      "loss": 4.9884,
      "mean_token_accuracy": 0.21393145322799684,
      "num_tokens": 54736992.0,
      "step": 26385
    },
    {
      "entropy": 5.482192087173462,
      "epoch": 2.393867924528302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044263127412615897,
      "loss": 4.8802,
      "mean_token_accuracy": 0.21235992312431334,
      "num_tokens": 54746951.0,
      "step": 26390
    },
    {
      "entropy": 5.426538944244385,
      "epoch": 2.394321480406386,
      "grad_norm": 1.25,
      "learning_rate": 0.0004426096895827649,
      "loss": 4.8282,
      "mean_token_accuracy": 0.22736789733171464,
      "num_tokens": 54757272.0,
      "step": 26395
    },
    {
      "entropy": 5.49282431602478,
      "epoch": 2.39477503628447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044258810157315977,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20758346021175383,
      "num_tokens": 54767835.0,
      "step": 26400
    },
    {
      "entropy": 5.5019268035888675,
      "epoch": 2.3952285921625545,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044256651009779,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21451252698898315,
      "num_tokens": 54778432.0,
      "step": 26405
    },
    {
      "entropy": 5.519278287887573,
      "epoch": 2.3956821480406387,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044254491515710213,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20820215940475464,
      "num_tokens": 54788582.0,
      "step": 26410
    },
    {
      "entropy": 5.551551818847656,
      "epoch": 2.3961357039187225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004425233167515426,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21647589653730392,
      "num_tokens": 54798848.0,
      "step": 26415
    },
    {
      "entropy": 5.463963603973388,
      "epoch": 2.396589259796807,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004425017148815583,
      "loss": 4.8417,
      "mean_token_accuracy": 0.21646227240562438,
      "num_tokens": 54809867.0,
      "step": 26420
    },
    {
      "entropy": 5.475587987899781,
      "epoch": 2.397042815674891,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044248010954759575,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21429212540388107,
      "num_tokens": 54819905.0,
      "step": 26425
    },
    {
      "entropy": 5.525354862213135,
      "epoch": 2.3974963715529753,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044245850075010193,
      "loss": 4.9353,
      "mean_token_accuracy": 0.2138149157166481,
      "num_tokens": 54829723.0,
      "step": 26430
    },
    {
      "entropy": 5.503822326660156,
      "epoch": 2.3979499274310596,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044243688848952364,
      "loss": 4.8599,
      "mean_token_accuracy": 0.21812826097011567,
      "num_tokens": 54840008.0,
      "step": 26435
    },
    {
      "entropy": 5.5065418720245365,
      "epoch": 2.398403483309144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004424152727663078,
      "loss": 4.9255,
      "mean_token_accuracy": 0.20818552374839783,
      "num_tokens": 54851693.0,
      "step": 26440
    },
    {
      "entropy": 5.524015617370606,
      "epoch": 2.3988570391872277,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004423936535809015,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2128281906247139,
      "num_tokens": 54862086.0,
      "step": 26445
    },
    {
      "entropy": 5.583481884002685,
      "epoch": 2.399310595065312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004423720309337517,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2111051380634308,
      "num_tokens": 54872883.0,
      "step": 26450
    },
    {
      "entropy": 5.533581829071045,
      "epoch": 2.3997641509433962,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044235040482530576,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21077869534492494,
      "num_tokens": 54882815.0,
      "step": 26455
    },
    {
      "entropy": 5.363004207611084,
      "epoch": 2.4002177068214805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004423287752560108,
      "loss": 4.7454,
      "mean_token_accuracy": 0.2299979418516159,
      "num_tokens": 54892480.0,
      "step": 26460
    },
    {
      "entropy": 5.398974895477295,
      "epoch": 2.4006712626995648,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044230714222631397,
      "loss": 4.8198,
      "mean_token_accuracy": 0.22443083971738814,
      "num_tokens": 54901607.0,
      "step": 26465
    },
    {
      "entropy": 5.588579559326172,
      "epoch": 2.4011248185776486,
      "grad_norm": 1.0,
      "learning_rate": 0.00044228550573666294,
      "loss": 5.0479,
      "mean_token_accuracy": 0.21278466135263444,
      "num_tokens": 54913545.0,
      "step": 26470
    },
    {
      "entropy": 5.575187492370605,
      "epoch": 2.401578374455733,
      "grad_norm": 1.1875,
      "learning_rate": 0.000442263865787505,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21193353682756425,
      "num_tokens": 54924392.0,
      "step": 26475
    },
    {
      "entropy": 5.533935022354126,
      "epoch": 2.402031930333817,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004422422223792877,
      "loss": 4.9065,
      "mean_token_accuracy": 0.21484020948410035,
      "num_tokens": 54934114.0,
      "step": 26480
    },
    {
      "entropy": 5.5698761463165285,
      "epoch": 2.4024854862119014,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004422205755124586,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2120340183377266,
      "num_tokens": 54944268.0,
      "step": 26485
    },
    {
      "entropy": 5.532761526107788,
      "epoch": 2.4029390420899857,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044219892518746536,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2159234493970871,
      "num_tokens": 54954877.0,
      "step": 26490
    },
    {
      "entropy": 5.4670099258422855,
      "epoch": 2.4033925979680695,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044217727140475585,
      "loss": 4.8846,
      "mean_token_accuracy": 0.21790484786033631,
      "num_tokens": 54964745.0,
      "step": 26495
    },
    {
      "entropy": 5.4789206981658936,
      "epoch": 2.4038461538461537,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044215561416477764,
      "loss": 4.8278,
      "mean_token_accuracy": 0.21640582829713823,
      "num_tokens": 54975256.0,
      "step": 26500
    },
    {
      "entropy": 5.545788240432739,
      "epoch": 2.404299709724238,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004421339534679788,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20957930088043214,
      "num_tokens": 54986226.0,
      "step": 26505
    },
    {
      "entropy": 5.529408550262451,
      "epoch": 2.4047532656023223,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004421122893148072,
      "loss": 4.9515,
      "mean_token_accuracy": 0.20250663310289382,
      "num_tokens": 54997350.0,
      "step": 26510
    },
    {
      "entropy": 5.539572477340698,
      "epoch": 2.4052068214804065,
      "grad_norm": 1.25,
      "learning_rate": 0.00044209062170571086,
      "loss": 4.9048,
      "mean_token_accuracy": 0.21496990621089934,
      "num_tokens": 55007947.0,
      "step": 26515
    },
    {
      "entropy": 5.522122144699097,
      "epoch": 2.4056603773584904,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004420689506411379,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2112934932112694,
      "num_tokens": 55018789.0,
      "step": 26520
    },
    {
      "entropy": 5.48811354637146,
      "epoch": 2.4061139332365746,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044204727612153643,
      "loss": 4.8648,
      "mean_token_accuracy": 0.21057552248239517,
      "num_tokens": 55028934.0,
      "step": 26525
    },
    {
      "entropy": 5.485390281677246,
      "epoch": 2.406567489114659,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044202559814735475,
      "loss": 4.9003,
      "mean_token_accuracy": 0.22090499848127365,
      "num_tokens": 55039395.0,
      "step": 26530
    },
    {
      "entropy": 5.465528583526611,
      "epoch": 2.407021044992743,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004420039167190411,
      "loss": 4.8381,
      "mean_token_accuracy": 0.21799101084470748,
      "num_tokens": 55049932.0,
      "step": 26535
    },
    {
      "entropy": 5.523106670379638,
      "epoch": 2.4074746008708274,
      "grad_norm": 1.296875,
      "learning_rate": 0.000441982231837044,
      "loss": 4.953,
      "mean_token_accuracy": 0.20853873044252397,
      "num_tokens": 55061532.0,
      "step": 26540
    },
    {
      "entropy": 5.525784063339233,
      "epoch": 2.4079281567489117,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004419605435018117,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2074953332543373,
      "num_tokens": 55072368.0,
      "step": 26545
    },
    {
      "entropy": 5.515232896804809,
      "epoch": 2.4083817126269955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004419388517137928,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21289550215005876,
      "num_tokens": 55083128.0,
      "step": 26550
    },
    {
      "entropy": 5.528514385223389,
      "epoch": 2.40883526850508,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044191715647343594,
      "loss": 4.92,
      "mean_token_accuracy": 0.21328637152910232,
      "num_tokens": 55094071.0,
      "step": 26555
    },
    {
      "entropy": 5.516706895828247,
      "epoch": 2.409288824383164,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044189545778118973,
      "loss": 4.8906,
      "mean_token_accuracy": 0.2104053720831871,
      "num_tokens": 55104504.0,
      "step": 26560
    },
    {
      "entropy": 5.514849138259888,
      "epoch": 2.4097423802612483,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000441873755637503,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21149232983589172,
      "num_tokens": 55115722.0,
      "step": 26565
    },
    {
      "entropy": 5.457224941253662,
      "epoch": 2.410195936139332,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004418520500428244,
      "loss": 4.876,
      "mean_token_accuracy": 0.21758212447166442,
      "num_tokens": 55126628.0,
      "step": 26570
    },
    {
      "entropy": 5.486545515060425,
      "epoch": 2.4106494920174164,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044183034099760287,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2114276796579361,
      "num_tokens": 55137352.0,
      "step": 26575
    },
    {
      "entropy": 5.530210208892822,
      "epoch": 2.4111030478955007,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004418086285022874,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22247123569250107,
      "num_tokens": 55147145.0,
      "step": 26580
    },
    {
      "entropy": 5.5359681129455565,
      "epoch": 2.411556603773585,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044178691255732707,
      "loss": 4.9348,
      "mean_token_accuracy": 0.20887413918972014,
      "num_tokens": 55157386.0,
      "step": 26585
    },
    {
      "entropy": 5.522435903549194,
      "epoch": 2.412010159651669,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004417651931631708,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20828535556793212,
      "num_tokens": 55168969.0,
      "step": 26590
    },
    {
      "entropy": 5.475583791732788,
      "epoch": 2.4124637155297535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004417434703202678,
      "loss": 4.9251,
      "mean_token_accuracy": 0.20538465529680253,
      "num_tokens": 55180123.0,
      "step": 26595
    },
    {
      "entropy": 5.499003887176514,
      "epoch": 2.4129172714078373,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004417217440290673,
      "loss": 4.8898,
      "mean_token_accuracy": 0.21401241421699524,
      "num_tokens": 55190668.0,
      "step": 26600
    },
    {
      "entropy": 5.464096736907959,
      "epoch": 2.4133708272859216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004417000142900188,
      "loss": 4.8517,
      "mean_token_accuracy": 0.21448955982923507,
      "num_tokens": 55200852.0,
      "step": 26605
    },
    {
      "entropy": 5.594669008255005,
      "epoch": 2.413824383164006,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004416782811035715,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20347885936498641,
      "num_tokens": 55211174.0,
      "step": 26610
    },
    {
      "entropy": 5.439728689193726,
      "epoch": 2.41427793904209,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044165654447017485,
      "loss": 4.8107,
      "mean_token_accuracy": 0.22405715733766557,
      "num_tokens": 55222042.0,
      "step": 26615
    },
    {
      "entropy": 5.491124248504638,
      "epoch": 2.4147314949201744,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044163480439027845,
      "loss": 4.9459,
      "mean_token_accuracy": 0.20825361013412474,
      "num_tokens": 55231688.0,
      "step": 26620
    },
    {
      "entropy": 5.498802423477173,
      "epoch": 2.415185050798258,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004416130608643317,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21603261679410934,
      "num_tokens": 55241089.0,
      "step": 26625
    },
    {
      "entropy": 5.518965387344361,
      "epoch": 2.4156386066763424,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004415913138927844,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21926067322492598,
      "num_tokens": 55251511.0,
      "step": 26630
    },
    {
      "entropy": 5.457004117965698,
      "epoch": 2.4160921625544267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004415695634760864,
      "loss": 4.8998,
      "mean_token_accuracy": 0.2099840298295021,
      "num_tokens": 55262855.0,
      "step": 26635
    },
    {
      "entropy": 5.436475849151611,
      "epoch": 2.416545718432511,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004415478096146873,
      "loss": 4.8114,
      "mean_token_accuracy": 0.21919535398483275,
      "num_tokens": 55273035.0,
      "step": 26640
    },
    {
      "entropy": 5.490250062942505,
      "epoch": 2.4169992743105952,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000441526052309037,
      "loss": 4.8971,
      "mean_token_accuracy": 0.22140019834041597,
      "num_tokens": 55284033.0,
      "step": 26645
    },
    {
      "entropy": 5.4840551853179935,
      "epoch": 2.417452830188679,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004415042915595856,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22105709612369537,
      "num_tokens": 55294290.0,
      "step": 26650
    },
    {
      "entropy": 5.53508939743042,
      "epoch": 2.4179063860667633,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000441482527366783,
      "loss": 4.9284,
      "mean_token_accuracy": 0.21372416615486145,
      "num_tokens": 55305681.0,
      "step": 26655
    },
    {
      "entropy": 5.456181335449219,
      "epoch": 2.4183599419448476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004414607597310793,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21563009321689605,
      "num_tokens": 55316647.0,
      "step": 26660
    },
    {
      "entropy": 5.550612688064575,
      "epoch": 2.418813497822932,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004414389886529247,
      "loss": 5.0257,
      "mean_token_accuracy": 0.20627919286489488,
      "num_tokens": 55328318.0,
      "step": 26665
    },
    {
      "entropy": 5.540642976760864,
      "epoch": 2.419267053701016,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004414172141327693,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21341045796871186,
      "num_tokens": 55338483.0,
      "step": 26670
    },
    {
      "entropy": 5.528587532043457,
      "epoch": 2.4197206095791,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004413954361710636,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20577431619167327,
      "num_tokens": 55348648.0,
      "step": 26675
    },
    {
      "entropy": 5.47016658782959,
      "epoch": 2.4201741654571842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044137365476825785,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21380711644887923,
      "num_tokens": 55358599.0,
      "step": 26680
    },
    {
      "entropy": 5.572705888748169,
      "epoch": 2.4206277213352685,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044135186992480253,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21307087540626526,
      "num_tokens": 55369847.0,
      "step": 26685
    },
    {
      "entropy": 5.479419469833374,
      "epoch": 2.4210812772133528,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044133008164114814,
      "loss": 4.9208,
      "mean_token_accuracy": 0.21522193998098374,
      "num_tokens": 55381235.0,
      "step": 26690
    },
    {
      "entropy": 5.490272092819214,
      "epoch": 2.421534833091437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044130828991774526,
      "loss": 4.9629,
      "mean_token_accuracy": 0.2074080303311348,
      "num_tokens": 55391527.0,
      "step": 26695
    },
    {
      "entropy": 5.56069974899292,
      "epoch": 2.4219883889695213,
      "grad_norm": 1.125,
      "learning_rate": 0.00044128649475504454,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21089034229516984,
      "num_tokens": 55401528.0,
      "step": 26700
    },
    {
      "entropy": 5.573121118545532,
      "epoch": 2.422441944847605,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004412646961534968,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20303410291671753,
      "num_tokens": 55411411.0,
      "step": 26705
    },
    {
      "entropy": 5.424116611480713,
      "epoch": 2.4228955007256894,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004412428941135528,
      "loss": 4.7837,
      "mean_token_accuracy": 0.22403921484947203,
      "num_tokens": 55423164.0,
      "step": 26710
    },
    {
      "entropy": 5.495169687271118,
      "epoch": 2.4233490566037736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004412210886356633,
      "loss": 4.8546,
      "mean_token_accuracy": 0.2217934936285019,
      "num_tokens": 55433397.0,
      "step": 26715
    },
    {
      "entropy": 5.469728660583496,
      "epoch": 2.423802612481858,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044119927972027936,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2103663921356201,
      "num_tokens": 55442898.0,
      "step": 26720
    },
    {
      "entropy": 5.484694719314575,
      "epoch": 2.4242561683599417,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044117746736785195,
      "loss": 4.916,
      "mean_token_accuracy": 0.22058435082435607,
      "num_tokens": 55452226.0,
      "step": 26725
    },
    {
      "entropy": 5.530117416381836,
      "epoch": 2.424709724238026,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004411556515788322,
      "loss": 4.9204,
      "mean_token_accuracy": 0.21022847294807434,
      "num_tokens": 55462503.0,
      "step": 26730
    },
    {
      "entropy": 5.561010503768921,
      "epoch": 2.4251632801161103,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004411338323536712,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20880862474441528,
      "num_tokens": 55472789.0,
      "step": 26735
    },
    {
      "entropy": 5.434799861907959,
      "epoch": 2.4256168359941945,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004411120096928203,
      "loss": 4.8447,
      "mean_token_accuracy": 0.21778091490268708,
      "num_tokens": 55482572.0,
      "step": 26740
    },
    {
      "entropy": 5.52170844078064,
      "epoch": 2.426070391872279,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004410901835967306,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22017523795366287,
      "num_tokens": 55493033.0,
      "step": 26745
    },
    {
      "entropy": 5.529430389404297,
      "epoch": 2.426523947750363,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044106835406585354,
      "loss": 4.8716,
      "mean_token_accuracy": 0.21206064373254777,
      "num_tokens": 55503574.0,
      "step": 26750
    },
    {
      "entropy": 5.544883966445923,
      "epoch": 2.426977503628447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044104652110064067,
      "loss": 4.952,
      "mean_token_accuracy": 0.21000666171312332,
      "num_tokens": 55514119.0,
      "step": 26755
    },
    {
      "entropy": 5.496326208114624,
      "epoch": 2.427431059506531,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004410246847015435,
      "loss": 4.8759,
      "mean_token_accuracy": 0.21246764957904815,
      "num_tokens": 55524973.0,
      "step": 26760
    },
    {
      "entropy": 5.4857885360717775,
      "epoch": 2.4278846153846154,
      "grad_norm": 1.125,
      "learning_rate": 0.0004410028448690134,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2148415982723236,
      "num_tokens": 55534847.0,
      "step": 26765
    },
    {
      "entropy": 5.536700296401977,
      "epoch": 2.4283381712626997,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004409810016035023,
      "loss": 4.9402,
      "mean_token_accuracy": 0.20440245270729065,
      "num_tokens": 55544886.0,
      "step": 26770
    },
    {
      "entropy": 5.509374284744263,
      "epoch": 2.4287917271407835,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004409591549054617,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20709252059459687,
      "num_tokens": 55554534.0,
      "step": 26775
    },
    {
      "entropy": 5.527661991119385,
      "epoch": 2.4292452830188678,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044093730477534355,
      "loss": 4.9647,
      "mean_token_accuracy": 0.20093503445386887,
      "num_tokens": 55564302.0,
      "step": 26780
    },
    {
      "entropy": 5.4597742557525635,
      "epoch": 2.429698838896952,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004409154512135996,
      "loss": 4.8863,
      "mean_token_accuracy": 0.21747547686100005,
      "num_tokens": 55574543.0,
      "step": 26785
    },
    {
      "entropy": 5.46315336227417,
      "epoch": 2.4301523947750363,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004408935942206818,
      "loss": 4.884,
      "mean_token_accuracy": 0.21548329740762712,
      "num_tokens": 55583849.0,
      "step": 26790
    },
    {
      "entropy": 5.461973476409912,
      "epoch": 2.4306059506531206,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004408717337970423,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21405625641345977,
      "num_tokens": 55594294.0,
      "step": 26795
    },
    {
      "entropy": 5.5222148418426515,
      "epoch": 2.431059506531205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044084986994313304,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20487088561058045,
      "num_tokens": 55604185.0,
      "step": 26800
    },
    {
      "entropy": 5.4942230701446535,
      "epoch": 2.4315130624092887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004408280026594062,
      "loss": 4.8899,
      "mean_token_accuracy": 0.2136186718940735,
      "num_tokens": 55615203.0,
      "step": 26805
    },
    {
      "entropy": 5.454693746566773,
      "epoch": 2.431966618287373,
      "grad_norm": 1.203125,
      "learning_rate": 0.000440806131946314,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21074686646461488,
      "num_tokens": 55626157.0,
      "step": 26810
    },
    {
      "entropy": 5.570779085159302,
      "epoch": 2.432420174165457,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044078425780430877,
      "loss": 4.9412,
      "mean_token_accuracy": 0.20626144260168075,
      "num_tokens": 55636862.0,
      "step": 26815
    },
    {
      "entropy": 5.513393831253052,
      "epoch": 2.4328737300435415,
      "grad_norm": 1.125,
      "learning_rate": 0.0004407623802338428,
      "loss": 4.8414,
      "mean_token_accuracy": 0.21836733520030976,
      "num_tokens": 55647646.0,
      "step": 26820
    },
    {
      "entropy": 5.483344316482544,
      "epoch": 2.4333272859216257,
      "grad_norm": 1.125,
      "learning_rate": 0.00044074049923536855,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21991498917341232,
      "num_tokens": 55659847.0,
      "step": 26825
    },
    {
      "entropy": 5.445178079605102,
      "epoch": 2.4337808417997095,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004407186148093386,
      "loss": 4.9205,
      "mean_token_accuracy": 0.21364904344081878,
      "num_tokens": 55670626.0,
      "step": 26830
    },
    {
      "entropy": 5.512595319747925,
      "epoch": 2.434234397677794,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044069672695620543,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20941813439130783,
      "num_tokens": 55681368.0,
      "step": 26835
    },
    {
      "entropy": 5.479383230209351,
      "epoch": 2.434687953555878,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004406748356764217,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2141915038228035,
      "num_tokens": 55691606.0,
      "step": 26840
    },
    {
      "entropy": 5.528767919540405,
      "epoch": 2.4351415094339623,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044065294097044017,
      "loss": 5.0295,
      "mean_token_accuracy": 0.20866839438676835,
      "num_tokens": 55702223.0,
      "step": 26845
    },
    {
      "entropy": 5.440175533294678,
      "epoch": 2.4355950653120466,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004406310428387136,
      "loss": 4.8277,
      "mean_token_accuracy": 0.21594141274690629,
      "num_tokens": 55712729.0,
      "step": 26850
    },
    {
      "entropy": 5.532445621490479,
      "epoch": 2.4360486211901304,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004406091412816949,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21529546231031418,
      "num_tokens": 55722863.0,
      "step": 26855
    },
    {
      "entropy": 5.549607372283935,
      "epoch": 2.4365021770682147,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044058723629983694,
      "loss": 5.0032,
      "mean_token_accuracy": 0.21473173648118973,
      "num_tokens": 55733124.0,
      "step": 26860
    },
    {
      "entropy": 5.547212982177735,
      "epoch": 2.436955732946299,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004405653278935927,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2171642452478409,
      "num_tokens": 55743438.0,
      "step": 26865
    },
    {
      "entropy": 5.443336105346679,
      "epoch": 2.4374092888243832,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044054341606341533,
      "loss": 4.8177,
      "mean_token_accuracy": 0.2146530881524086,
      "num_tokens": 55753602.0,
      "step": 26870
    },
    {
      "entropy": 5.43633017539978,
      "epoch": 2.4378628447024675,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004405215008097579,
      "loss": 4.8711,
      "mean_token_accuracy": 0.22178202122449875,
      "num_tokens": 55763473.0,
      "step": 26875
    },
    {
      "entropy": 5.462363290786743,
      "epoch": 2.4383164005805513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044049958213307364,
      "loss": 4.8827,
      "mean_token_accuracy": 0.21548015773296356,
      "num_tokens": 55772848.0,
      "step": 26880
    },
    {
      "entropy": 5.459492158889771,
      "epoch": 2.4387699564586356,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044047766003381584,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22162742614746095,
      "num_tokens": 55783468.0,
      "step": 26885
    },
    {
      "entropy": 5.486462831497192,
      "epoch": 2.43922351233672,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044045573451243793,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21840264052152633,
      "num_tokens": 55793619.0,
      "step": 26890
    },
    {
      "entropy": 5.479871797561645,
      "epoch": 2.439677068214804,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044043380556939326,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2143921285867691,
      "num_tokens": 55803869.0,
      "step": 26895
    },
    {
      "entropy": 5.542277097702026,
      "epoch": 2.4401306240928884,
      "grad_norm": 1.125,
      "learning_rate": 0.0004404118732051353,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21084595769643782,
      "num_tokens": 55814754.0,
      "step": 26900
    },
    {
      "entropy": 5.479099321365356,
      "epoch": 2.4405841799709727,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004403899374201176,
      "loss": 4.871,
      "mean_token_accuracy": 0.21173604875802993,
      "num_tokens": 55824862.0,
      "step": 26905
    },
    {
      "entropy": 5.440395927429199,
      "epoch": 2.4410377358490565,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044036799821479397,
      "loss": 4.8796,
      "mean_token_accuracy": 0.21630153506994249,
      "num_tokens": 55835506.0,
      "step": 26910
    },
    {
      "entropy": 5.518917751312256,
      "epoch": 2.4414912917271407,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044034605558961786,
      "loss": 4.9654,
      "mean_token_accuracy": 0.2110930159687996,
      "num_tokens": 55846244.0,
      "step": 26915
    },
    {
      "entropy": 5.440672254562378,
      "epoch": 2.441944847605225,
      "grad_norm": 1.125,
      "learning_rate": 0.00044032410954504325,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22054927796125412,
      "num_tokens": 55856192.0,
      "step": 26920
    },
    {
      "entropy": 5.488173961639404,
      "epoch": 2.4423984034833093,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044030216008152394,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21053328961133957,
      "num_tokens": 55866509.0,
      "step": 26925
    },
    {
      "entropy": 5.522663545608521,
      "epoch": 2.442851959361393,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044028020719951376,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21562650799751282,
      "num_tokens": 55877321.0,
      "step": 26930
    },
    {
      "entropy": 5.503424739837646,
      "epoch": 2.4433055152394774,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004402582508994669,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2158783942461014,
      "num_tokens": 55887933.0,
      "step": 26935
    },
    {
      "entropy": 5.439522743225098,
      "epoch": 2.4437590711175616,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044023629118183726,
      "loss": 4.8613,
      "mean_token_accuracy": 0.21243632286787034,
      "num_tokens": 55897274.0,
      "step": 26940
    },
    {
      "entropy": 5.525461149215698,
      "epoch": 2.444212626995646,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044021432804707895,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20866023898124694,
      "num_tokens": 55907393.0,
      "step": 26945
    },
    {
      "entropy": 5.460586357116699,
      "epoch": 2.44466618287373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004401923614956463,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2113513767719269,
      "num_tokens": 55918429.0,
      "step": 26950
    },
    {
      "entropy": 5.490865707397461,
      "epoch": 2.4451197387518144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004401703915279935,
      "loss": 4.874,
      "mean_token_accuracy": 0.2138999193906784,
      "num_tokens": 55928733.0,
      "step": 26955
    },
    {
      "entropy": 5.535603857040405,
      "epoch": 2.4455732946298983,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044014841814457486,
      "loss": 4.8843,
      "mean_token_accuracy": 0.21638813316822053,
      "num_tokens": 55939128.0,
      "step": 26960
    },
    {
      "entropy": 5.55042519569397,
      "epoch": 2.4460268505079825,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004401264413458449,
      "loss": 5.0305,
      "mean_token_accuracy": 0.20636268258094786,
      "num_tokens": 55950237.0,
      "step": 26965
    },
    {
      "entropy": 5.504288530349731,
      "epoch": 2.446480406386067,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004401044611322581,
      "loss": 4.8384,
      "mean_token_accuracy": 0.22138212323188783,
      "num_tokens": 55960436.0,
      "step": 26970
    },
    {
      "entropy": 5.4609678268432615,
      "epoch": 2.446933962264151,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004400824775042689,
      "loss": 4.8277,
      "mean_token_accuracy": 0.21638987064361573,
      "num_tokens": 55970446.0,
      "step": 26975
    },
    {
      "entropy": 5.505931377410889,
      "epoch": 2.447387518142235,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000440060490462332,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21207243502140044,
      "num_tokens": 55980706.0,
      "step": 26980
    },
    {
      "entropy": 5.510345697402954,
      "epoch": 2.447841074020319,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004400385000069021,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20478834807872773,
      "num_tokens": 55990669.0,
      "step": 26985
    },
    {
      "entropy": 5.585621404647827,
      "epoch": 2.4482946298984034,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044001650613843395,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20946276932954788,
      "num_tokens": 56000709.0,
      "step": 26990
    },
    {
      "entropy": 5.5095092296600345,
      "epoch": 2.4487481857764877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004399945088573824,
      "loss": 4.8999,
      "mean_token_accuracy": 0.2201507106423378,
      "num_tokens": 56011435.0,
      "step": 26995
    },
    {
      "entropy": 5.482499456405639,
      "epoch": 2.449201741654572,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043997250816420246,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21885482519865035,
      "num_tokens": 56021264.0,
      "step": 27000
    },
    {
      "epoch": 2.449201741654572,
      "eval_entropy": 5.2357296030985285,
      "eval_loss": 4.991609573364258,
      "eval_mean_token_accuracy": 0.21966902533908506,
      "eval_num_tokens": 56021264.0,
      "eval_runtime": 20.6142,
      "eval_samples_per_second": 1715.321,
      "eval_steps_per_second": 214.415,
      "step": 27000
    },
    {
      "entropy": 5.494916391372681,
      "epoch": 2.449655297532656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004399505040593489,
      "loss": 4.9391,
      "mean_token_accuracy": 0.20938896685838698,
      "num_tokens": 56031463.0,
      "step": 27005
    },
    {
      "entropy": 5.469224977493286,
      "epoch": 2.45010885341074,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043992849654327695,
      "loss": 4.8225,
      "mean_token_accuracy": 0.21603455990552903,
      "num_tokens": 56041195.0,
      "step": 27010
    },
    {
      "entropy": 5.4915258407592775,
      "epoch": 2.4505624092888243,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004399064856164416,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21389532536268235,
      "num_tokens": 56051840.0,
      "step": 27015
    },
    {
      "entropy": 5.520238780975342,
      "epoch": 2.4510159651669086,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043988447127929817,
      "loss": 4.971,
      "mean_token_accuracy": 0.21247733235359192,
      "num_tokens": 56062535.0,
      "step": 27020
    },
    {
      "entropy": 5.504625940322876,
      "epoch": 2.451469521044993,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043986245353230173,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2248425841331482,
      "num_tokens": 56072177.0,
      "step": 27025
    },
    {
      "entropy": 5.4725605010986325,
      "epoch": 2.451923076923077,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043984043237590785,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2161737784743309,
      "num_tokens": 56081607.0,
      "step": 27030
    },
    {
      "entropy": 5.478176212310791,
      "epoch": 2.452376632801161,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043981840781057185,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21207477599382402,
      "num_tokens": 56093371.0,
      "step": 27035
    },
    {
      "entropy": 5.503431224822998,
      "epoch": 2.452830188679245,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043979637983674913,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2126280650496483,
      "num_tokens": 56103973.0,
      "step": 27040
    },
    {
      "entropy": 5.501952362060547,
      "epoch": 2.4532837445573294,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004397743484548953,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21286799609661103,
      "num_tokens": 56113906.0,
      "step": 27045
    },
    {
      "entropy": 5.48510513305664,
      "epoch": 2.4537373004354137,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004397523136654659,
      "loss": 4.8889,
      "mean_token_accuracy": 0.21415451616048814,
      "num_tokens": 56124172.0,
      "step": 27050
    },
    {
      "entropy": 5.464991283416748,
      "epoch": 2.454190856313498,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043973027546891677,
      "loss": 4.7674,
      "mean_token_accuracy": 0.2282601222395897,
      "num_tokens": 56133088.0,
      "step": 27055
    },
    {
      "entropy": 5.464124155044556,
      "epoch": 2.4546444121915822,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004397082338657034,
      "loss": 4.8948,
      "mean_token_accuracy": 0.21559770852327348,
      "num_tokens": 56143098.0,
      "step": 27060
    },
    {
      "entropy": 5.47809739112854,
      "epoch": 2.455097968069666,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043968618885628185,
      "loss": 4.8908,
      "mean_token_accuracy": 0.21945605874061586,
      "num_tokens": 56153889.0,
      "step": 27065
    },
    {
      "entropy": 5.458990097045898,
      "epoch": 2.4555515239477503,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043966414044110805,
      "loss": 4.8367,
      "mean_token_accuracy": 0.21992984116077424,
      "num_tokens": 56164164.0,
      "step": 27070
    },
    {
      "entropy": 5.522607040405274,
      "epoch": 2.4560050798258346,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043964208862063775,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21104868799448012,
      "num_tokens": 56173172.0,
      "step": 27075
    },
    {
      "entropy": 5.5256640911102295,
      "epoch": 2.456458635703919,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043962003339532714,
      "loss": 4.9052,
      "mean_token_accuracy": 0.2197902038693428,
      "num_tokens": 56183127.0,
      "step": 27080
    },
    {
      "entropy": 5.526291561126709,
      "epoch": 2.4569121915820027,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043959797476563226,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21260267347097397,
      "num_tokens": 56192528.0,
      "step": 27085
    },
    {
      "entropy": 5.483093118667602,
      "epoch": 2.457365747460087,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043957591273200943,
      "loss": 4.8952,
      "mean_token_accuracy": 0.21251945197582245,
      "num_tokens": 56203239.0,
      "step": 27090
    },
    {
      "entropy": 5.524399089813232,
      "epoch": 2.4578193033381712,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004395538472949146,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21127965450286865,
      "num_tokens": 56214068.0,
      "step": 27095
    },
    {
      "entropy": 5.557310247421265,
      "epoch": 2.4582728592162555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004395317784548045,
      "loss": 4.969,
      "mean_token_accuracy": 0.202183036506176,
      "num_tokens": 56224829.0,
      "step": 27100
    },
    {
      "entropy": 5.593783712387085,
      "epoch": 2.4587264150943398,
      "grad_norm": 1.125,
      "learning_rate": 0.0004395097062121351,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21175989508628845,
      "num_tokens": 56235412.0,
      "step": 27105
    },
    {
      "entropy": 5.5662078857421875,
      "epoch": 2.459179970972424,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043948763056736315,
      "loss": 4.9991,
      "mean_token_accuracy": 0.20703294575214387,
      "num_tokens": 56246785.0,
      "step": 27110
    },
    {
      "entropy": 5.538149070739746,
      "epoch": 2.459633526850508,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000439465551520945,
      "loss": 4.9795,
      "mean_token_accuracy": 0.2101506069302559,
      "num_tokens": 56256701.0,
      "step": 27115
    },
    {
      "entropy": 5.479132223129272,
      "epoch": 2.460087082728592,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043944346907333744,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2113966926932335,
      "num_tokens": 56267418.0,
      "step": 27120
    },
    {
      "entropy": 5.443803930282593,
      "epoch": 2.4605406386066764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000439421383224997,
      "loss": 4.8529,
      "mean_token_accuracy": 0.21567052602767944,
      "num_tokens": 56278232.0,
      "step": 27125
    },
    {
      "entropy": 5.552145719528198,
      "epoch": 2.4609941944847606,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004393992939763804,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2112806424498558,
      "num_tokens": 56289866.0,
      "step": 27130
    },
    {
      "entropy": 5.510740137100219,
      "epoch": 2.4614477503628445,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043937720132794457,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2172268807888031,
      "num_tokens": 56299595.0,
      "step": 27135
    },
    {
      "entropy": 5.548333692550659,
      "epoch": 2.4619013062409287,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043935510528014625,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2175906166434288,
      "num_tokens": 56310152.0,
      "step": 27140
    },
    {
      "entropy": 5.519309711456299,
      "epoch": 2.462354862119013,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004393330058334426,
      "loss": 5.0309,
      "mean_token_accuracy": 0.20516949892044067,
      "num_tokens": 56320288.0,
      "step": 27145
    },
    {
      "entropy": 5.547877645492553,
      "epoch": 2.4628084179970973,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043931090298829047,
      "loss": 4.9333,
      "mean_token_accuracy": 0.20865002274513245,
      "num_tokens": 56330345.0,
      "step": 27150
    },
    {
      "entropy": 5.614681196212769,
      "epoch": 2.4632619738751815,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043928879674514694,
      "loss": 5.0675,
      "mean_token_accuracy": 0.2119154930114746,
      "num_tokens": 56340996.0,
      "step": 27155
    },
    {
      "entropy": 5.434743213653564,
      "epoch": 2.463715529753266,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004392666871044692,
      "loss": 4.7862,
      "mean_token_accuracy": 0.22223087400197983,
      "num_tokens": 56350326.0,
      "step": 27160
    },
    {
      "entropy": 5.47396445274353,
      "epoch": 2.4641690856313496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043924457406671456,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2186198577284813,
      "num_tokens": 56360556.0,
      "step": 27165
    },
    {
      "entropy": 5.541845321655273,
      "epoch": 2.464622641509434,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004392224576323403,
      "loss": 4.8881,
      "mean_token_accuracy": 0.21551176756620408,
      "num_tokens": 56369980.0,
      "step": 27170
    },
    {
      "entropy": 5.525144863128662,
      "epoch": 2.465076197387518,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043920033780180367,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2072213813662529,
      "num_tokens": 56379974.0,
      "step": 27175
    },
    {
      "entropy": 5.481790161132812,
      "epoch": 2.4655297532656024,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004391782145755623,
      "loss": 4.8552,
      "mean_token_accuracy": 0.21741439700126647,
      "num_tokens": 56390542.0,
      "step": 27180
    },
    {
      "entropy": 5.570679426193237,
      "epoch": 2.4659833091436867,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004391560879540736,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20456839799880983,
      "num_tokens": 56401755.0,
      "step": 27185
    },
    {
      "entropy": 5.500103330612182,
      "epoch": 2.4664368650217705,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004391339579377952,
      "loss": 4.9238,
      "mean_token_accuracy": 0.2165648266673088,
      "num_tokens": 56412223.0,
      "step": 27190
    },
    {
      "entropy": 5.539244604110718,
      "epoch": 2.4668904208998548,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043911182452718476,
      "loss": 4.9602,
      "mean_token_accuracy": 0.2088968962430954,
      "num_tokens": 56422947.0,
      "step": 27195
    },
    {
      "entropy": 5.489877223968506,
      "epoch": 2.467343976777939,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043908968772269984,
      "loss": 4.9132,
      "mean_token_accuracy": 0.21561992019414902,
      "num_tokens": 56433189.0,
      "step": 27200
    },
    {
      "entropy": 5.527774047851563,
      "epoch": 2.4677975326560233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043906754752479836,
      "loss": 4.934,
      "mean_token_accuracy": 0.2143586054444313,
      "num_tokens": 56444460.0,
      "step": 27205
    },
    {
      "entropy": 5.54619402885437,
      "epoch": 2.4682510885341076,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004390454039339383,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20837389826774597,
      "num_tokens": 56454368.0,
      "step": 27210
    },
    {
      "entropy": 5.451146364212036,
      "epoch": 2.4687046444121914,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043902325695057743,
      "loss": 4.8696,
      "mean_token_accuracy": 0.2145874246954918,
      "num_tokens": 56464153.0,
      "step": 27215
    },
    {
      "entropy": 5.484639883041382,
      "epoch": 2.4691582002902757,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043900110657517377,
      "loss": 4.8779,
      "mean_token_accuracy": 0.21641258150339127,
      "num_tokens": 56473627.0,
      "step": 27220
    },
    {
      "entropy": 5.479943943023682,
      "epoch": 2.46961175616836,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004389789528081855,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21771203875541686,
      "num_tokens": 56483660.0,
      "step": 27225
    },
    {
      "entropy": 5.5281702995300295,
      "epoch": 2.470065312046444,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004389567956500707,
      "loss": 4.8954,
      "mean_token_accuracy": 0.21584168076515198,
      "num_tokens": 56492653.0,
      "step": 27230
    },
    {
      "entropy": 5.578667068481446,
      "epoch": 2.4705188679245285,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004389346351012876,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20452366322278975,
      "num_tokens": 56503289.0,
      "step": 27235
    },
    {
      "entropy": 5.570003795623779,
      "epoch": 2.4709724238026123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043891247116229436,
      "loss": 4.9983,
      "mean_token_accuracy": 0.2053408905863762,
      "num_tokens": 56513843.0,
      "step": 27240
    },
    {
      "entropy": 5.558351325988769,
      "epoch": 2.4714259796806965,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043889030383354957,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2121433585882187,
      "num_tokens": 56524175.0,
      "step": 27245
    },
    {
      "entropy": 5.480007219314575,
      "epoch": 2.471879535558781,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004388681331155114,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2238924190402031,
      "num_tokens": 56534976.0,
      "step": 27250
    },
    {
      "entropy": 5.557298278808593,
      "epoch": 2.472333091436865,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004388459590086386,
      "loss": 4.9308,
      "mean_token_accuracy": 0.21302677690982819,
      "num_tokens": 56545967.0,
      "step": 27255
    },
    {
      "entropy": 5.5358631134033205,
      "epoch": 2.4727866473149493,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043882378151338956,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20955596268177032,
      "num_tokens": 56555527.0,
      "step": 27260
    },
    {
      "entropy": 5.535481262207031,
      "epoch": 2.4732402031930336,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000438801600630223,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20499177724123002,
      "num_tokens": 56565874.0,
      "step": 27265
    },
    {
      "entropy": 5.496514415740966,
      "epoch": 2.4736937590711174,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043877941635959755,
      "loss": 4.8711,
      "mean_token_accuracy": 0.2151979088783264,
      "num_tokens": 56575452.0,
      "step": 27270
    },
    {
      "entropy": 5.407858562469483,
      "epoch": 2.4741473149492017,
      "grad_norm": 1.125,
      "learning_rate": 0.00043875722870197214,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22353904992341994,
      "num_tokens": 56586176.0,
      "step": 27275
    },
    {
      "entropy": 5.517267560958862,
      "epoch": 2.474600870827286,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004387350376578054,
      "loss": 4.9049,
      "mean_token_accuracy": 0.21410378515720369,
      "num_tokens": 56595473.0,
      "step": 27280
    },
    {
      "entropy": 5.5282081127166744,
      "epoch": 2.4750544267053702,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004387128432275563,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21571965664625167,
      "num_tokens": 56606016.0,
      "step": 27285
    },
    {
      "entropy": 5.48876314163208,
      "epoch": 2.475507982583454,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000438690645411684,
      "loss": 4.9268,
      "mean_token_accuracy": 0.21178403496742249,
      "num_tokens": 56617674.0,
      "step": 27290
    },
    {
      "entropy": 5.536155033111572,
      "epoch": 2.4759615384615383,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043866844421064745,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21504662036895753,
      "num_tokens": 56628077.0,
      "step": 27295
    },
    {
      "entropy": 5.541977024078369,
      "epoch": 2.4764150943396226,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043864623962490575,
      "loss": 4.8735,
      "mean_token_accuracy": 0.2176972433924675,
      "num_tokens": 56638691.0,
      "step": 27300
    },
    {
      "entropy": 5.518357276916504,
      "epoch": 2.476868650217707,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004386240316549181,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21259113103151323,
      "num_tokens": 56650338.0,
      "step": 27305
    },
    {
      "entropy": 5.478986406326294,
      "epoch": 2.477322206095791,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043860182030114374,
      "loss": 4.8501,
      "mean_token_accuracy": 0.2234350174665451,
      "num_tokens": 56660328.0,
      "step": 27310
    },
    {
      "entropy": 5.573250818252563,
      "epoch": 2.4777757619738754,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043857960556404214,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21366402953863145,
      "num_tokens": 56670674.0,
      "step": 27315
    },
    {
      "entropy": 5.506168985366822,
      "epoch": 2.478229317851959,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004385573874440725,
      "loss": 4.917,
      "mean_token_accuracy": 0.21611077785491944,
      "num_tokens": 56680656.0,
      "step": 27320
    },
    {
      "entropy": 5.4909961223602295,
      "epoch": 2.4786828737300435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043853516594169446,
      "loss": 4.8412,
      "mean_token_accuracy": 0.22142103910446168,
      "num_tokens": 56691039.0,
      "step": 27325
    },
    {
      "entropy": 5.502261638641357,
      "epoch": 2.4791364296081277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004385129410573675,
      "loss": 4.9012,
      "mean_token_accuracy": 0.21016360223293304,
      "num_tokens": 56703101.0,
      "step": 27330
    },
    {
      "entropy": 5.517206430435181,
      "epoch": 2.479589985486212,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043849071279155124,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22036022543907166,
      "num_tokens": 56712351.0,
      "step": 27335
    },
    {
      "entropy": 5.495806884765625,
      "epoch": 2.480043541364296,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043846848114470537,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21987007558345795,
      "num_tokens": 56722836.0,
      "step": 27340
    },
    {
      "entropy": 5.527901601791382,
      "epoch": 2.48049709724238,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043844624611728974,
      "loss": 4.9156,
      "mean_token_accuracy": 0.21343635469675065,
      "num_tokens": 56734032.0,
      "step": 27345
    },
    {
      "entropy": 5.525054550170898,
      "epoch": 2.4809506531204644,
      "grad_norm": 1.125,
      "learning_rate": 0.00043842400770976406,
      "loss": 4.8958,
      "mean_token_accuracy": 0.21624885499477386,
      "num_tokens": 56743536.0,
      "step": 27350
    },
    {
      "entropy": 5.476931762695313,
      "epoch": 2.4814042089985486,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004384017659225882,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21204424202442168,
      "num_tokens": 56753475.0,
      "step": 27355
    },
    {
      "entropy": 5.458152723312378,
      "epoch": 2.481857764876633,
      "grad_norm": 1.125,
      "learning_rate": 0.00043837952075622224,
      "loss": 4.8401,
      "mean_token_accuracy": 0.22411585301160813,
      "num_tokens": 56763602.0,
      "step": 27360
    },
    {
      "entropy": 5.517909812927246,
      "epoch": 2.482311320754717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004383572722111261,
      "loss": 4.8494,
      "mean_token_accuracy": 0.2242525562644005,
      "num_tokens": 56773859.0,
      "step": 27365
    },
    {
      "entropy": 5.498459386825561,
      "epoch": 2.482764876632801,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043833502028775995,
      "loss": 4.9059,
      "mean_token_accuracy": 0.21762668043375016,
      "num_tokens": 56783810.0,
      "step": 27370
    },
    {
      "entropy": 5.510944843292236,
      "epoch": 2.4832184325108853,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043831276498658395,
      "loss": 5.0046,
      "mean_token_accuracy": 0.21070886254310608,
      "num_tokens": 56794088.0,
      "step": 27375
    },
    {
      "entropy": 5.437960433959961,
      "epoch": 2.4836719883889695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004382905063080584,
      "loss": 4.8507,
      "mean_token_accuracy": 0.21804012060165406,
      "num_tokens": 56804623.0,
      "step": 27380
    },
    {
      "entropy": 5.47583794593811,
      "epoch": 2.484125544267054,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043826824425264357,
      "loss": 4.9306,
      "mean_token_accuracy": 0.21052938997745513,
      "num_tokens": 56813842.0,
      "step": 27385
    },
    {
      "entropy": 5.538039636611939,
      "epoch": 2.484579100145138,
      "grad_norm": 1.125,
      "learning_rate": 0.0004382459788207997,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21324970722198486,
      "num_tokens": 56823501.0,
      "step": 27390
    },
    {
      "entropy": 5.545355939865113,
      "epoch": 2.485032656023222,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043822371001298754,
      "loss": 4.9427,
      "mean_token_accuracy": 0.211967134475708,
      "num_tokens": 56834601.0,
      "step": 27395
    },
    {
      "entropy": 5.460286712646484,
      "epoch": 2.485486211901306,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004382014378296674,
      "loss": 4.9414,
      "mean_token_accuracy": 0.20564347356557847,
      "num_tokens": 56844701.0,
      "step": 27400
    },
    {
      "entropy": 5.466500902175904,
      "epoch": 2.4859397677793904,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004381791622713,
      "loss": 4.9124,
      "mean_token_accuracy": 0.21440732032060622,
      "num_tokens": 56854867.0,
      "step": 27405
    },
    {
      "entropy": 5.556684875488282,
      "epoch": 2.4863933236574747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000438156883338346,
      "loss": 4.944,
      "mean_token_accuracy": 0.21684509813785552,
      "num_tokens": 56865688.0,
      "step": 27410
    },
    {
      "entropy": 5.5753742218017575,
      "epoch": 2.486846879535559,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043813460103126595,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21565346121788026,
      "num_tokens": 56876504.0,
      "step": 27415
    },
    {
      "entropy": 5.50859580039978,
      "epoch": 2.487300435413643,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043811231535052076,
      "loss": 4.8916,
      "mean_token_accuracy": 0.21681600362062453,
      "num_tokens": 56886083.0,
      "step": 27420
    },
    {
      "entropy": 5.466057634353637,
      "epoch": 2.487753991291727,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004380900262965714,
      "loss": 4.949,
      "mean_token_accuracy": 0.21614607572555541,
      "num_tokens": 56897079.0,
      "step": 27425
    },
    {
      "entropy": 5.48646650314331,
      "epoch": 2.4882075471698113,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004380677338698787,
      "loss": 4.8877,
      "mean_token_accuracy": 0.21964190006256104,
      "num_tokens": 56906962.0,
      "step": 27430
    },
    {
      "entropy": 5.566875267028808,
      "epoch": 2.4886611030478956,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043804543807090374,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20811552107334136,
      "num_tokens": 56918156.0,
      "step": 27435
    },
    {
      "entropy": 5.483881998062134,
      "epoch": 2.48911465892598,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043802313890010757,
      "loss": 4.9143,
      "mean_token_accuracy": 0.2126583307981491,
      "num_tokens": 56928888.0,
      "step": 27440
    },
    {
      "entropy": 5.518128824234009,
      "epoch": 2.4895682148040637,
      "grad_norm": 1.125,
      "learning_rate": 0.0004380008363579513,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21166997402906418,
      "num_tokens": 56939929.0,
      "step": 27445
    },
    {
      "entropy": 5.539379644393921,
      "epoch": 2.490021770682148,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043797853044489623,
      "loss": 4.8384,
      "mean_token_accuracy": 0.21962590962648393,
      "num_tokens": 56950992.0,
      "step": 27450
    },
    {
      "entropy": 5.56898717880249,
      "epoch": 2.490475326560232,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004379562211614036,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21231268495321273,
      "num_tokens": 56962118.0,
      "step": 27455
    },
    {
      "entropy": 5.524201726913452,
      "epoch": 2.4909288824383164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379339085079348,
      "loss": 4.9232,
      "mean_token_accuracy": 0.21393794566392899,
      "num_tokens": 56972837.0,
      "step": 27460
    },
    {
      "entropy": 5.532465982437134,
      "epoch": 2.4913824383164007,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004379115924849512,
      "loss": 4.9239,
      "mean_token_accuracy": 0.2131011039018631,
      "num_tokens": 56982794.0,
      "step": 27465
    },
    {
      "entropy": 5.596230411529541,
      "epoch": 2.491835994194485,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004378892730929145,
      "loss": 5.0567,
      "mean_token_accuracy": 0.215960992872715,
      "num_tokens": 56993979.0,
      "step": 27470
    },
    {
      "entropy": 5.52348952293396,
      "epoch": 2.492289550072569,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043786695033228604,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21323892921209336,
      "num_tokens": 57004505.0,
      "step": 27475
    },
    {
      "entropy": 5.532655858993531,
      "epoch": 2.492743105950653,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043784462420352747,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21141747534275054,
      "num_tokens": 57014121.0,
      "step": 27480
    },
    {
      "entropy": 5.569469881057739,
      "epoch": 2.4931966618287373,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043782229470710067,
      "loss": 4.9324,
      "mean_token_accuracy": 0.2121428593993187,
      "num_tokens": 57023748.0,
      "step": 27485
    },
    {
      "entropy": 5.478572797775269,
      "epoch": 2.4936502177068216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004377999618434673,
      "loss": 4.9023,
      "mean_token_accuracy": 0.21529392749071122,
      "num_tokens": 57034782.0,
      "step": 27490
    },
    {
      "entropy": 5.451092100143432,
      "epoch": 2.4941037735849054,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004377776256130892,
      "loss": 4.848,
      "mean_token_accuracy": 0.21864885240793228,
      "num_tokens": 57045136.0,
      "step": 27495
    },
    {
      "entropy": 5.526750230789185,
      "epoch": 2.4945573294629897,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043775528601642833,
      "loss": 5.0655,
      "mean_token_accuracy": 0.2005676433444023,
      "num_tokens": 57055893.0,
      "step": 27500
    },
    {
      "entropy": 5.523357105255127,
      "epoch": 2.495010885341074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043773294305394665,
      "loss": 4.9415,
      "mean_token_accuracy": 0.21413020193576812,
      "num_tokens": 57067189.0,
      "step": 27505
    },
    {
      "entropy": 5.554200935363769,
      "epoch": 2.4954644412191582,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043771059672610636,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2193602055311203,
      "num_tokens": 57077374.0,
      "step": 27510
    },
    {
      "entropy": 5.50458345413208,
      "epoch": 2.4959179970972425,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043768824703336933,
      "loss": 4.8874,
      "mean_token_accuracy": 0.21179157197475434,
      "num_tokens": 57088168.0,
      "step": 27515
    },
    {
      "entropy": 5.497486877441406,
      "epoch": 2.4963715529753268,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004376658939761979,
      "loss": 4.8765,
      "mean_token_accuracy": 0.2154663622379303,
      "num_tokens": 57098743.0,
      "step": 27520
    },
    {
      "entropy": 5.512390184402466,
      "epoch": 2.4968251088534106,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004376435375550544,
      "loss": 5.0045,
      "mean_token_accuracy": 0.20618267208337784,
      "num_tokens": 57109854.0,
      "step": 27525
    },
    {
      "entropy": 5.468565130233765,
      "epoch": 2.497278664731495,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043762117777040106,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22250816524028777,
      "num_tokens": 57120296.0,
      "step": 27530
    },
    {
      "entropy": 5.480441570281982,
      "epoch": 2.497732220609579,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004375988146227003,
      "loss": 4.8894,
      "mean_token_accuracy": 0.21591870933771135,
      "num_tokens": 57131295.0,
      "step": 27535
    },
    {
      "entropy": 5.5250892162323,
      "epoch": 2.4981857764876634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004375764481124147,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21266426891088486,
      "num_tokens": 57141936.0,
      "step": 27540
    },
    {
      "entropy": 5.472104263305664,
      "epoch": 2.4986393323657476,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043755407824000664,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22408146113157273,
      "num_tokens": 57152510.0,
      "step": 27545
    },
    {
      "entropy": 5.488446235656738,
      "epoch": 2.4990928882438315,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043753170500593884,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2145903527736664,
      "num_tokens": 57162546.0,
      "step": 27550
    },
    {
      "entropy": 5.431600761413574,
      "epoch": 2.4995464441219157,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043750932841067397,
      "loss": 4.822,
      "mean_token_accuracy": 0.2214560925960541,
      "num_tokens": 57174658.0,
      "step": 27555
    },
    {
      "entropy": 5.585208129882813,
      "epoch": 2.5,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043748694845467483,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20471915155649184,
      "num_tokens": 57185157.0,
      "step": 27560
    },
    {
      "entropy": 5.506630182266235,
      "epoch": 2.5004535558780843,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043746456513840405,
      "loss": 4.8495,
      "mean_token_accuracy": 0.2161380410194397,
      "num_tokens": 57195277.0,
      "step": 27565
    },
    {
      "entropy": 5.546444845199585,
      "epoch": 2.5009071117561685,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004374421784623248,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21481461822986603,
      "num_tokens": 57205374.0,
      "step": 27570
    },
    {
      "entropy": 5.48497462272644,
      "epoch": 2.501360667634253,
      "grad_norm": 1.25,
      "learning_rate": 0.00043741978842689987,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2142408162355423,
      "num_tokens": 57215700.0,
      "step": 27575
    },
    {
      "entropy": 5.502138757705689,
      "epoch": 2.5018142235123366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004373973950325923,
      "loss": 4.939,
      "mean_token_accuracy": 0.21474092006683348,
      "num_tokens": 57225765.0,
      "step": 27580
    },
    {
      "entropy": 5.515675735473633,
      "epoch": 2.502267779390421,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043737499827986527,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21452175974845886,
      "num_tokens": 57235913.0,
      "step": 27585
    },
    {
      "entropy": 5.475171518325806,
      "epoch": 2.502721335268505,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043735259816918184,
      "loss": 4.8576,
      "mean_token_accuracy": 0.21677391976118088,
      "num_tokens": 57246204.0,
      "step": 27590
    },
    {
      "entropy": 5.480260229110717,
      "epoch": 2.5031748911465894,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004373301947010053,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20743590593338013,
      "num_tokens": 57256167.0,
      "step": 27595
    },
    {
      "entropy": 5.44982876777649,
      "epoch": 2.5036284470246732,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043730778787579895,
      "loss": 4.8889,
      "mean_token_accuracy": 0.21928549706935882,
      "num_tokens": 57266621.0,
      "step": 27600
    },
    {
      "entropy": 5.481981897354126,
      "epoch": 2.5040820029027575,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043728537769402627,
      "loss": 4.9095,
      "mean_token_accuracy": 0.21795809417963027,
      "num_tokens": 57277282.0,
      "step": 27605
    },
    {
      "entropy": 5.542119979858398,
      "epoch": 2.5045355587808418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004372629641561506,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21075396090745926,
      "num_tokens": 57287541.0,
      "step": 27610
    },
    {
      "entropy": 5.5087456703186035,
      "epoch": 2.504989114658926,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043724054726263547,
      "loss": 4.9292,
      "mean_token_accuracy": 0.21412270218133928,
      "num_tokens": 57298158.0,
      "step": 27615
    },
    {
      "entropy": 5.558322334289551,
      "epoch": 2.5054426705370103,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004372181270139444,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2086131066083908,
      "num_tokens": 57308356.0,
      "step": 27620
    },
    {
      "entropy": 5.528935384750366,
      "epoch": 2.5058962264150946,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043719570341054127,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2186117261648178,
      "num_tokens": 57317905.0,
      "step": 27625
    },
    {
      "entropy": 5.524879407882691,
      "epoch": 2.5063497822931784,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043717327645288956,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21665917634963988,
      "num_tokens": 57328277.0,
      "step": 27630
    },
    {
      "entropy": 5.56379246711731,
      "epoch": 2.5068033381712627,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043715084614145316,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20803385972976685,
      "num_tokens": 57338975.0,
      "step": 27635
    },
    {
      "entropy": 5.518955421447754,
      "epoch": 2.507256894049347,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000437128412476696,
      "loss": 4.9151,
      "mean_token_accuracy": 0.21916110217571258,
      "num_tokens": 57349439.0,
      "step": 27640
    },
    {
      "entropy": 5.530729007720947,
      "epoch": 2.507710449927431,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004371059754590819,
      "loss": 4.9675,
      "mean_token_accuracy": 0.20922602862119674,
      "num_tokens": 57359414.0,
      "step": 27645
    },
    {
      "entropy": 5.573251533508301,
      "epoch": 2.508164005805515,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043708353508907494,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2157943442463875,
      "num_tokens": 57369711.0,
      "step": 27650
    },
    {
      "entropy": 5.512748908996582,
      "epoch": 2.5086175616835993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004370610913671392,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22621740847826005,
      "num_tokens": 57380001.0,
      "step": 27655
    },
    {
      "entropy": 5.548326921463013,
      "epoch": 2.5090711175616836,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043703864429373875,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20096748769283296,
      "num_tokens": 57390794.0,
      "step": 27660
    },
    {
      "entropy": 5.525460958480835,
      "epoch": 2.509524673439768,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043701619386933793,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2023957073688507,
      "num_tokens": 57401975.0,
      "step": 27665
    },
    {
      "entropy": 5.447001123428345,
      "epoch": 2.509978229317852,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043699374009440085,
      "loss": 4.8221,
      "mean_token_accuracy": 0.2262996554374695,
      "num_tokens": 57411249.0,
      "step": 27670
    },
    {
      "entropy": 5.442929029464722,
      "epoch": 2.5104317851959363,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004369712829693919,
      "loss": 4.8587,
      "mean_token_accuracy": 0.22130753546953202,
      "num_tokens": 57422679.0,
      "step": 27675
    },
    {
      "entropy": 5.503858947753907,
      "epoch": 2.51088534107402,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004369488224947757,
      "loss": 4.9302,
      "mean_token_accuracy": 0.21813174039125444,
      "num_tokens": 57433265.0,
      "step": 27680
    },
    {
      "entropy": 5.5370386123657225,
      "epoch": 2.5113388969521044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004369263586710164,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2111544355750084,
      "num_tokens": 57444095.0,
      "step": 27685
    },
    {
      "entropy": 5.5152997970581055,
      "epoch": 2.5117924528301887,
      "grad_norm": 1.25,
      "learning_rate": 0.0004369038914985789,
      "loss": 4.903,
      "mean_token_accuracy": 0.21714168339967727,
      "num_tokens": 57454059.0,
      "step": 27690
    },
    {
      "entropy": 5.508581829071045,
      "epoch": 2.512246008708273,
      "grad_norm": 1.125,
      "learning_rate": 0.0004368814209779276,
      "loss": 4.8739,
      "mean_token_accuracy": 0.21691862493753433,
      "num_tokens": 57463940.0,
      "step": 27695
    },
    {
      "entropy": 5.464889240264893,
      "epoch": 2.512699564586357,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043685894710952724,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21298775374889373,
      "num_tokens": 57474731.0,
      "step": 27700
    },
    {
      "entropy": 5.495354223251343,
      "epoch": 2.513153120464441,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004368364698938426,
      "loss": 4.8644,
      "mean_token_accuracy": 0.2191845789551735,
      "num_tokens": 57485096.0,
      "step": 27705
    },
    {
      "entropy": 5.496347284317016,
      "epoch": 2.5136066763425253,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004368139893313385,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21534126698970796,
      "num_tokens": 57496371.0,
      "step": 27710
    },
    {
      "entropy": 5.520041561126709,
      "epoch": 2.5140602322206096,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043679150542248,
      "loss": 4.9044,
      "mean_token_accuracy": 0.21714208275079727,
      "num_tokens": 57506012.0,
      "step": 27715
    },
    {
      "entropy": 5.515847158432007,
      "epoch": 2.514513788098694,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043676901816773174,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20679448992013932,
      "num_tokens": 57516536.0,
      "step": 27720
    },
    {
      "entropy": 5.480475521087646,
      "epoch": 2.514967343976778,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004367465275675591,
      "loss": 4.848,
      "mean_token_accuracy": 0.2189696326851845,
      "num_tokens": 57527450.0,
      "step": 27725
    },
    {
      "entropy": 5.533561182022095,
      "epoch": 2.5154208998548624,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043672403362242704,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21798517256975175,
      "num_tokens": 57538234.0,
      "step": 27730
    },
    {
      "entropy": 5.49214882850647,
      "epoch": 2.515874455732946,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043670153633280066,
      "loss": 4.8941,
      "mean_token_accuracy": 0.2198481634259224,
      "num_tokens": 57547938.0,
      "step": 27735
    },
    {
      "entropy": 5.52827820777893,
      "epoch": 2.5163280116110305,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004366790356991454,
      "loss": 4.926,
      "mean_token_accuracy": 0.21719639599323273,
      "num_tokens": 57560553.0,
      "step": 27740
    },
    {
      "entropy": 5.481798982620239,
      "epoch": 2.5167815674891147,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004366565317219265,
      "loss": 4.8893,
      "mean_token_accuracy": 0.21737712174654006,
      "num_tokens": 57570670.0,
      "step": 27745
    },
    {
      "entropy": 5.592238998413086,
      "epoch": 2.5172351233671986,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043663402440160924,
      "loss": 5.0984,
      "mean_token_accuracy": 0.2019535407423973,
      "num_tokens": 57582377.0,
      "step": 27750
    },
    {
      "entropy": 5.483578014373779,
      "epoch": 2.517688679245283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004366115137386592,
      "loss": 4.8157,
      "mean_token_accuracy": 0.2220057025551796,
      "num_tokens": 57592764.0,
      "step": 27755
    },
    {
      "entropy": 5.473769426345825,
      "epoch": 2.518142235123367,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004365889997335419,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2066652238368988,
      "num_tokens": 57603768.0,
      "step": 27760
    },
    {
      "entropy": 5.475136041641235,
      "epoch": 2.5185957910014514,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043656648238672285,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21326394081115724,
      "num_tokens": 57614346.0,
      "step": 27765
    },
    {
      "entropy": 5.622374725341797,
      "epoch": 2.5190493468795356,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004365439616986678,
      "loss": 5.0318,
      "mean_token_accuracy": 0.2063527747988701,
      "num_tokens": 57625015.0,
      "step": 27770
    },
    {
      "entropy": 5.543213605880737,
      "epoch": 2.51950290275762,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043652143766984247,
      "loss": 4.9906,
      "mean_token_accuracy": 0.2083511084318161,
      "num_tokens": 57635476.0,
      "step": 27775
    },
    {
      "entropy": 5.449545574188233,
      "epoch": 2.519956458635704,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043649891030071264,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21324318945407866,
      "num_tokens": 57645512.0,
      "step": 27780
    },
    {
      "entropy": 5.486173915863037,
      "epoch": 2.520410014513788,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043647637959174425,
      "loss": 4.8922,
      "mean_token_accuracy": 0.21642659604549408,
      "num_tokens": 57655766.0,
      "step": 27785
    },
    {
      "entropy": 5.546034097671509,
      "epoch": 2.5208635703918723,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043645384554340315,
      "loss": 4.8871,
      "mean_token_accuracy": 0.21995265036821365,
      "num_tokens": 57667325.0,
      "step": 27790
    },
    {
      "entropy": 5.471993350982666,
      "epoch": 2.5213171262699565,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043643130815615536,
      "loss": 4.9151,
      "mean_token_accuracy": 0.20932620018720627,
      "num_tokens": 57677884.0,
      "step": 27795
    },
    {
      "entropy": 5.523683404922485,
      "epoch": 2.521770682148041,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043640876743046694,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2210479885339737,
      "num_tokens": 57687765.0,
      "step": 27800
    },
    {
      "entropy": 5.517579984664917,
      "epoch": 2.5222242380261246,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004363862233668041,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2118662565946579,
      "num_tokens": 57697777.0,
      "step": 27805
    },
    {
      "entropy": 5.482229995727539,
      "epoch": 2.522677793904209,
      "grad_norm": 1.078125,
      "learning_rate": 0.000436363675965633,
      "loss": 4.9183,
      "mean_token_accuracy": 0.21222163438796998,
      "num_tokens": 57708765.0,
      "step": 27810
    },
    {
      "entropy": 5.537875556945801,
      "epoch": 2.523131349782293,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004363411252274201,
      "loss": 4.9353,
      "mean_token_accuracy": 0.21945043951272963,
      "num_tokens": 57719658.0,
      "step": 27815
    },
    {
      "entropy": 5.5319188117980955,
      "epoch": 2.5235849056603774,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043631857115263153,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2217099353671074,
      "num_tokens": 57730018.0,
      "step": 27820
    },
    {
      "entropy": 5.43439998626709,
      "epoch": 2.5240384615384617,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004362960137417338,
      "loss": 4.8598,
      "mean_token_accuracy": 0.2194720461964607,
      "num_tokens": 57741251.0,
      "step": 27825
    },
    {
      "entropy": 5.484037065505982,
      "epoch": 2.524492017416546,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004362734529951933,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21405988782644272,
      "num_tokens": 57751020.0,
      "step": 27830
    },
    {
      "entropy": 5.521639633178711,
      "epoch": 2.5249455732946298,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043625088891347685,
      "loss": 4.8932,
      "mean_token_accuracy": 0.21852710098028183,
      "num_tokens": 57760989.0,
      "step": 27835
    },
    {
      "entropy": 5.58126745223999,
      "epoch": 2.525399129172714,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043622832149705086,
      "loss": 4.973,
      "mean_token_accuracy": 0.21619820296764375,
      "num_tokens": 57770882.0,
      "step": 27840
    },
    {
      "entropy": 5.571942377090454,
      "epoch": 2.5258526850507983,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043620575074638206,
      "loss": 4.9857,
      "mean_token_accuracy": 0.21061291098594664,
      "num_tokens": 57781812.0,
      "step": 27845
    },
    {
      "entropy": 5.471639490127563,
      "epoch": 2.5263062409288826,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043618317666193734,
      "loss": 4.8762,
      "mean_token_accuracy": 0.2247692108154297,
      "num_tokens": 57792009.0,
      "step": 27850
    },
    {
      "entropy": 5.486639881134034,
      "epoch": 2.5267597968069664,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043616059924418346,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2197083905339241,
      "num_tokens": 57802109.0,
      "step": 27855
    },
    {
      "entropy": 5.5288773536682125,
      "epoch": 2.5272133526850507,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004361380184935872,
      "loss": 4.9418,
      "mean_token_accuracy": 0.2158348485827446,
      "num_tokens": 57812374.0,
      "step": 27860
    },
    {
      "entropy": 5.480235528945923,
      "epoch": 2.527666908563135,
      "grad_norm": 1.125,
      "learning_rate": 0.0004361154344106158,
      "loss": 4.8488,
      "mean_token_accuracy": 0.21825022101402283,
      "num_tokens": 57822936.0,
      "step": 27865
    },
    {
      "entropy": 5.5933677673339846,
      "epoch": 2.528120464441219,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043609284699573604,
      "loss": 5.0821,
      "mean_token_accuracy": 0.2073484629392624,
      "num_tokens": 57832693.0,
      "step": 27870
    },
    {
      "entropy": 5.578198051452636,
      "epoch": 2.5285740203193035,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043607025624941523,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21290663331747056,
      "num_tokens": 57843161.0,
      "step": 27875
    },
    {
      "entropy": 5.5319836139678955,
      "epoch": 2.5290275761973877,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004360476621721204,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2210705816745758,
      "num_tokens": 57853027.0,
      "step": 27880
    },
    {
      "entropy": 5.534374761581421,
      "epoch": 2.5294811320754715,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004360250647643189,
      "loss": 4.92,
      "mean_token_accuracy": 0.2234542638063431,
      "num_tokens": 57862821.0,
      "step": 27885
    },
    {
      "entropy": 5.509299898147583,
      "epoch": 2.529934687953556,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043600246402647805,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2223489135503769,
      "num_tokens": 57874065.0,
      "step": 27890
    },
    {
      "entropy": 5.509321451187134,
      "epoch": 2.53038824383164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004359798599590652,
      "loss": 4.9543,
      "mean_token_accuracy": 0.20906148701906205,
      "num_tokens": 57885190.0,
      "step": 27895
    },
    {
      "entropy": 5.490346431732178,
      "epoch": 2.5308417997097243,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043595725256254786,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2137491524219513,
      "num_tokens": 57894234.0,
      "step": 27900
    },
    {
      "entropy": 5.556329917907715,
      "epoch": 2.531295355587808,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004359346418373935,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20328215807676314,
      "num_tokens": 57905401.0,
      "step": 27905
    },
    {
      "entropy": 5.490631580352783,
      "epoch": 2.5317489114658924,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043591202778406975,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21316498816013335,
      "num_tokens": 57916097.0,
      "step": 27910
    },
    {
      "entropy": 5.432488012313843,
      "epoch": 2.5322024673439767,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004358894104030442,
      "loss": 4.8073,
      "mean_token_accuracy": 0.21938321590423585,
      "num_tokens": 57926871.0,
      "step": 27915
    },
    {
      "entropy": 5.482897758483887,
      "epoch": 2.532656023222061,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043586678969478464,
      "loss": 4.8584,
      "mean_token_accuracy": 0.2240841031074524,
      "num_tokens": 57937344.0,
      "step": 27920
    },
    {
      "entropy": 5.560228490829468,
      "epoch": 2.5331095791001452,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000435844165659759,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21373773068189622,
      "num_tokens": 57947633.0,
      "step": 27925
    },
    {
      "entropy": 5.477146863937378,
      "epoch": 2.5335631349782295,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043582153829843493,
      "loss": 4.8545,
      "mean_token_accuracy": 0.2168970748782158,
      "num_tokens": 57957948.0,
      "step": 27930
    },
    {
      "entropy": 5.527875328063965,
      "epoch": 2.5340166908563138,
      "grad_norm": 1.125,
      "learning_rate": 0.0004357989076112805,
      "loss": 4.8931,
      "mean_token_accuracy": 0.21876860707998275,
      "num_tokens": 57967719.0,
      "step": 27935
    },
    {
      "entropy": 5.519929981231689,
      "epoch": 2.5344702467343976,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004357762735987637,
      "loss": 4.9713,
      "mean_token_accuracy": 0.20958924293518066,
      "num_tokens": 57978218.0,
      "step": 27940
    },
    {
      "entropy": 5.518215227127075,
      "epoch": 2.534923802612482,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004357536362613525,
      "loss": 4.8456,
      "mean_token_accuracy": 0.21913884729146957,
      "num_tokens": 57987860.0,
      "step": 27945
    },
    {
      "entropy": 5.4863457679748535,
      "epoch": 2.535377358490566,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043573099559951516,
      "loss": 4.9124,
      "mean_token_accuracy": 0.21714988797903062,
      "num_tokens": 57997872.0,
      "step": 27950
    },
    {
      "entropy": 5.555253553390503,
      "epoch": 2.5358309143686504,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004357083516137199,
      "loss": 4.994,
      "mean_token_accuracy": 0.2131021276116371,
      "num_tokens": 58007771.0,
      "step": 27955
    },
    {
      "entropy": 5.534782934188843,
      "epoch": 2.536284470246734,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000435685704304435,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21169153451919556,
      "num_tokens": 58018972.0,
      "step": 27960
    },
    {
      "entropy": 5.490043210983276,
      "epoch": 2.5367380261248185,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043566305367212884,
      "loss": 4.8119,
      "mean_token_accuracy": 0.2191553846001625,
      "num_tokens": 58028581.0,
      "step": 27965
    },
    {
      "entropy": 5.490304803848266,
      "epoch": 2.5371915820029027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043564039971726964,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2148470625281334,
      "num_tokens": 58038877.0,
      "step": 27970
    },
    {
      "entropy": 5.476082897186279,
      "epoch": 2.537645137880987,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043561774244032614,
      "loss": 4.8941,
      "mean_token_accuracy": 0.220500348508358,
      "num_tokens": 58048777.0,
      "step": 27975
    },
    {
      "entropy": 5.540262508392334,
      "epoch": 2.5380986937590713,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043559508184176675,
      "loss": 4.981,
      "mean_token_accuracy": 0.21174225062131882,
      "num_tokens": 58060270.0,
      "step": 27980
    },
    {
      "entropy": 5.496583461761475,
      "epoch": 2.5385522496371555,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043557241792206014,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21867234706878663,
      "num_tokens": 58070838.0,
      "step": 27985
    },
    {
      "entropy": 5.565356206893921,
      "epoch": 2.5390058055152394,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043554975068167505,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21263106614351274,
      "num_tokens": 58079933.0,
      "step": 27990
    },
    {
      "entropy": 5.501317167282105,
      "epoch": 2.5394593613933236,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004355270801210802,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21623146831989287,
      "num_tokens": 58090673.0,
      "step": 27995
    },
    {
      "entropy": 5.448718786239624,
      "epoch": 2.539912917271408,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004355044062407444,
      "loss": 4.7646,
      "mean_token_accuracy": 0.23167373687028886,
      "num_tokens": 58100172.0,
      "step": 28000
    },
    {
      "entropy": 5.487764596939087,
      "epoch": 2.540366473149492,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004354817290411366,
      "loss": 4.9239,
      "mean_token_accuracy": 0.21400413513183594,
      "num_tokens": 58108965.0,
      "step": 28005
    },
    {
      "entropy": 5.495748853683471,
      "epoch": 2.540820029027576,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043545904852272573,
      "loss": 4.8271,
      "mean_token_accuracy": 0.22209217250347138,
      "num_tokens": 58120094.0,
      "step": 28010
    },
    {
      "entropy": 5.5234373092651365,
      "epoch": 2.5412735849056602,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043543636468598083,
      "loss": 4.9642,
      "mean_token_accuracy": 0.20915620774030685,
      "num_tokens": 58131525.0,
      "step": 28015
    },
    {
      "entropy": 5.510516691207886,
      "epoch": 2.5417271407837445,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043541367753137107,
      "loss": 4.8726,
      "mean_token_accuracy": 0.21721156388521196,
      "num_tokens": 58141986.0,
      "step": 28020
    },
    {
      "entropy": 5.5301285743713375,
      "epoch": 2.5421806966618288,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004353909870593656,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21358744651079178,
      "num_tokens": 58154122.0,
      "step": 28025
    },
    {
      "entropy": 5.53275842666626,
      "epoch": 2.542634252539913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004353682932704336,
      "loss": 4.8572,
      "mean_token_accuracy": 0.23040631413459778,
      "num_tokens": 58164148.0,
      "step": 28030
    },
    {
      "entropy": 5.479786968231201,
      "epoch": 2.5430878084179973,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004353455961650444,
      "loss": 4.8578,
      "mean_token_accuracy": 0.21293506771326065,
      "num_tokens": 58173836.0,
      "step": 28035
    },
    {
      "entropy": 5.471020174026489,
      "epoch": 2.543541364296081,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004353228957436675,
      "loss": 4.9046,
      "mean_token_accuracy": 0.21567854583263396,
      "num_tokens": 58184598.0,
      "step": 28040
    },
    {
      "entropy": 5.457410430908203,
      "epoch": 2.5439949201741654,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043530019200677215,
      "loss": 4.8787,
      "mean_token_accuracy": 0.21554286479949952,
      "num_tokens": 58194820.0,
      "step": 28045
    },
    {
      "entropy": 5.505194187164307,
      "epoch": 2.5444484760522497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043527748495482806,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2107910394668579,
      "num_tokens": 58206589.0,
      "step": 28050
    },
    {
      "entropy": 5.587746953964233,
      "epoch": 2.544902031930334,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043525477458830473,
      "loss": 4.928,
      "mean_token_accuracy": 0.21810096651315689,
      "num_tokens": 58216750.0,
      "step": 28055
    },
    {
      "entropy": 5.465854978561401,
      "epoch": 2.5453555878084178,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004352320609076718,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21865989714860917,
      "num_tokens": 58228496.0,
      "step": 28060
    },
    {
      "entropy": 5.479026365280151,
      "epoch": 2.545809143686502,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043520934391339894,
      "loss": 4.8729,
      "mean_token_accuracy": 0.21455207616090774,
      "num_tokens": 58237844.0,
      "step": 28065
    },
    {
      "entropy": 5.464645624160767,
      "epoch": 2.5462626995645863,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004351866236059561,
      "loss": 4.8585,
      "mean_token_accuracy": 0.21462422609329224,
      "num_tokens": 58248522.0,
      "step": 28070
    },
    {
      "entropy": 5.499445104598999,
      "epoch": 2.5467162554426706,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043516389998581305,
      "loss": 4.8719,
      "mean_token_accuracy": 0.21963100582361222,
      "num_tokens": 58258492.0,
      "step": 28075
    },
    {
      "entropy": 5.4501711368560795,
      "epoch": 2.547169811320755,
      "grad_norm": 1.125,
      "learning_rate": 0.0004351411730534398,
      "loss": 4.8933,
      "mean_token_accuracy": 0.21421114653348922,
      "num_tokens": 58268683.0,
      "step": 28080
    },
    {
      "entropy": 5.5106712818145756,
      "epoch": 2.547623367198839,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004351184428093062,
      "loss": 4.957,
      "mean_token_accuracy": 0.20706234276294708,
      "num_tokens": 58279479.0,
      "step": 28085
    },
    {
      "entropy": 5.5375994682312015,
      "epoch": 2.5480769230769234,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004350957092538825,
      "loss": 4.985,
      "mean_token_accuracy": 0.21224521547555925,
      "num_tokens": 58290751.0,
      "step": 28090
    },
    {
      "entropy": 5.559394073486328,
      "epoch": 2.548530478955007,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004350729723876386,
      "loss": 4.9948,
      "mean_token_accuracy": 0.20879129320383072,
      "num_tokens": 58302163.0,
      "step": 28095
    },
    {
      "entropy": 5.499516201019287,
      "epoch": 2.5489840348330914,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043505023221104497,
      "loss": 4.9194,
      "mean_token_accuracy": 0.2061060681939125,
      "num_tokens": 58311488.0,
      "step": 28100
    },
    {
      "entropy": 5.487242889404297,
      "epoch": 2.5494375907111757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043502748872457166,
      "loss": 4.8962,
      "mean_token_accuracy": 0.21689498722553252,
      "num_tokens": 58321599.0,
      "step": 28105
    },
    {
      "entropy": 5.509772253036499,
      "epoch": 2.5498911465892595,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004350047419286892,
      "loss": 4.8838,
      "mean_token_accuracy": 0.21568981409072877,
      "num_tokens": 58330696.0,
      "step": 28110
    },
    {
      "entropy": 5.4617715835571286,
      "epoch": 2.550344702467344,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043498199182386775,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2220580592751503,
      "num_tokens": 58341793.0,
      "step": 28115
    },
    {
      "entropy": 5.507754516601563,
      "epoch": 2.550798258345428,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043495923841057807,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21068783104419708,
      "num_tokens": 58352260.0,
      "step": 28120
    },
    {
      "entropy": 5.614972496032715,
      "epoch": 2.5512518142235123,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043493648168929053,
      "loss": 5.0498,
      "mean_token_accuracy": 0.2003721445798874,
      "num_tokens": 58363052.0,
      "step": 28125
    },
    {
      "entropy": 5.57716965675354,
      "epoch": 2.5517053701015966,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004349137216604758,
      "loss": 4.9365,
      "mean_token_accuracy": 0.21506949365139008,
      "num_tokens": 58374039.0,
      "step": 28130
    },
    {
      "entropy": 5.4511034965515135,
      "epoch": 2.552158925979681,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004348909583246045,
      "loss": 4.7831,
      "mean_token_accuracy": 0.22838286757469178,
      "num_tokens": 58383441.0,
      "step": 28135
    },
    {
      "entropy": 5.450941038131714,
      "epoch": 2.552612481857765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004348681916821475,
      "loss": 4.898,
      "mean_token_accuracy": 0.21414742171764373,
      "num_tokens": 58394164.0,
      "step": 28140
    },
    {
      "entropy": 5.461052894592285,
      "epoch": 2.553066037735849,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004348454217335755,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2184153378009796,
      "num_tokens": 58404610.0,
      "step": 28145
    },
    {
      "entropy": 5.495025110244751,
      "epoch": 2.553519593613933,
      "grad_norm": 1.125,
      "learning_rate": 0.00043482264847935953,
      "loss": 4.8591,
      "mean_token_accuracy": 0.21553471088409423,
      "num_tokens": 58414512.0,
      "step": 28150
    },
    {
      "entropy": 5.539588260650635,
      "epoch": 2.5539731494920175,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043479987191997045,
      "loss": 5.0046,
      "mean_token_accuracy": 0.21369443237781524,
      "num_tokens": 58424669.0,
      "step": 28155
    },
    {
      "entropy": 5.4342506408691404,
      "epoch": 2.5544267053701017,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004347770920558792,
      "loss": 4.8092,
      "mean_token_accuracy": 0.2217569351196289,
      "num_tokens": 58435348.0,
      "step": 28160
    },
    {
      "entropy": 5.531671857833862,
      "epoch": 2.5548802612481856,
      "grad_norm": 1.125,
      "learning_rate": 0.00043475430888755706,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21354020833969117,
      "num_tokens": 58445534.0,
      "step": 28165
    },
    {
      "entropy": 5.513222885131836,
      "epoch": 2.55533381712627,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004347315224154751,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2151782065629959,
      "num_tokens": 58456718.0,
      "step": 28170
    },
    {
      "entropy": 5.596724510192871,
      "epoch": 2.555787373004354,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004347087326401045,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2181641563773155,
      "num_tokens": 58467176.0,
      "step": 28175
    },
    {
      "entropy": 5.493879985809326,
      "epoch": 2.5562409288824384,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043468593956191663,
      "loss": 4.8567,
      "mean_token_accuracy": 0.21785433143377303,
      "num_tokens": 58477642.0,
      "step": 28180
    },
    {
      "entropy": 5.4853198528289795,
      "epoch": 2.5566944847605226,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004346631431813829,
      "loss": 4.9142,
      "mean_token_accuracy": 0.20857097059488297,
      "num_tokens": 58489055.0,
      "step": 28185
    },
    {
      "entropy": 5.551945543289184,
      "epoch": 2.557148040638607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004346403434989746,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21246396750211716,
      "num_tokens": 58499062.0,
      "step": 28190
    },
    {
      "entropy": 5.521152067184448,
      "epoch": 2.5576015965166907,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004346175405151633,
      "loss": 4.9346,
      "mean_token_accuracy": 0.20794905126094818,
      "num_tokens": 58509154.0,
      "step": 28195
    },
    {
      "entropy": 5.479930782318116,
      "epoch": 2.558055152394775,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004345947342304207,
      "loss": 4.832,
      "mean_token_accuracy": 0.22104021906852722,
      "num_tokens": 58520056.0,
      "step": 28200
    },
    {
      "entropy": 5.508643674850464,
      "epoch": 2.5585087082728593,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043457192464521827,
      "loss": 4.9144,
      "mean_token_accuracy": 0.21632963865995408,
      "num_tokens": 58529934.0,
      "step": 28205
    },
    {
      "entropy": 5.437611389160156,
      "epoch": 2.5589622641509435,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004345491117600278,
      "loss": 4.8315,
      "mean_token_accuracy": 0.21674349904060364,
      "num_tokens": 58540013.0,
      "step": 28210
    },
    {
      "entropy": 5.574136686325073,
      "epoch": 2.5594158200290273,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043452629557532097,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21669500917196274,
      "num_tokens": 58550282.0,
      "step": 28215
    },
    {
      "entropy": 5.574855661392212,
      "epoch": 2.5598693759071116,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043450347609156974,
      "loss": 4.9497,
      "mean_token_accuracy": 0.20531440675258636,
      "num_tokens": 58560490.0,
      "step": 28220
    },
    {
      "entropy": 5.465995025634766,
      "epoch": 2.560322931785196,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043448065330924595,
      "loss": 4.9178,
      "mean_token_accuracy": 0.213985837996006,
      "num_tokens": 58570371.0,
      "step": 28225
    },
    {
      "entropy": 5.53932752609253,
      "epoch": 2.56077648766328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004344578272288216,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2195175141096115,
      "num_tokens": 58580701.0,
      "step": 28230
    },
    {
      "entropy": 5.580049800872803,
      "epoch": 2.5612300435413644,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043443499785076885,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21302091777324678,
      "num_tokens": 58590903.0,
      "step": 28235
    },
    {
      "entropy": 5.543482112884521,
      "epoch": 2.5616835994194487,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004344121651755596,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21193549782037735,
      "num_tokens": 58600924.0,
      "step": 28240
    },
    {
      "entropy": 5.542697143554688,
      "epoch": 2.5621371552975325,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043438932920366615,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21586293131113052,
      "num_tokens": 58611369.0,
      "step": 28245
    },
    {
      "entropy": 5.568830966949463,
      "epoch": 2.5625907111756168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004343664899355608,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21262579113245011,
      "num_tokens": 58622725.0,
      "step": 28250
    },
    {
      "entropy": 5.5199089527130125,
      "epoch": 2.563044267053701,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043434364737171587,
      "loss": 4.9003,
      "mean_token_accuracy": 0.21693652421236037,
      "num_tokens": 58634157.0,
      "step": 28255
    },
    {
      "entropy": 5.5146430969238285,
      "epoch": 2.5634978229317853,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043432080151260366,
      "loss": 4.8353,
      "mean_token_accuracy": 0.21184777468442917,
      "num_tokens": 58643384.0,
      "step": 28260
    },
    {
      "entropy": 5.504140567779541,
      "epoch": 2.563951378809869,
      "grad_norm": 1.125,
      "learning_rate": 0.00043429795235869667,
      "loss": 4.8782,
      "mean_token_accuracy": 0.21168771833181382,
      "num_tokens": 58653748.0,
      "step": 28265
    },
    {
      "entropy": 5.506859397888183,
      "epoch": 2.5644049346879534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043427509991046746,
      "loss": 4.9218,
      "mean_token_accuracy": 0.20965004116296768,
      "num_tokens": 58664619.0,
      "step": 28270
    },
    {
      "entropy": 5.469032430648804,
      "epoch": 2.5648584905660377,
      "grad_norm": 1.125,
      "learning_rate": 0.00043425224416838857,
      "loss": 4.8366,
      "mean_token_accuracy": 0.22153815478086472,
      "num_tokens": 58674506.0,
      "step": 28275
    },
    {
      "entropy": 5.53009443283081,
      "epoch": 2.565312046444122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004342293851329327,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21152455657720565,
      "num_tokens": 58686252.0,
      "step": 28280
    },
    {
      "entropy": 5.523956346511841,
      "epoch": 2.565765602322206,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043420652280457255,
      "loss": 4.905,
      "mean_token_accuracy": 0.21134098321199418,
      "num_tokens": 58696477.0,
      "step": 28285
    },
    {
      "entropy": 5.4551633358001705,
      "epoch": 2.5662191582002905,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043418365718378096,
      "loss": 4.7977,
      "mean_token_accuracy": 0.22520421743392943,
      "num_tokens": 58706664.0,
      "step": 28290
    },
    {
      "entropy": 5.430073881149292,
      "epoch": 2.5666727140783747,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004341607882710307,
      "loss": 4.8879,
      "mean_token_accuracy": 0.21757084578275682,
      "num_tokens": 58716950.0,
      "step": 28295
    },
    {
      "entropy": 5.501533651351929,
      "epoch": 2.5671262699564585,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004341379160667949,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21849264800548554,
      "num_tokens": 58727723.0,
      "step": 28300
    },
    {
      "entropy": 5.520059823989868,
      "epoch": 2.567579825834543,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004341150405715464,
      "loss": 4.9848,
      "mean_token_accuracy": 0.20906808823347092,
      "num_tokens": 58737956.0,
      "step": 28305
    },
    {
      "entropy": 5.521510410308838,
      "epoch": 2.568033381712627,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043409216178575826,
      "loss": 4.843,
      "mean_token_accuracy": 0.22448870837688445,
      "num_tokens": 58748476.0,
      "step": 28310
    },
    {
      "entropy": 5.542687273025512,
      "epoch": 2.5684869375907113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004340692797099037,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2177031382918358,
      "num_tokens": 58759604.0,
      "step": 28315
    },
    {
      "entropy": 5.513464832305909,
      "epoch": 2.568940493468795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004340463943444559,
      "loss": 4.89,
      "mean_token_accuracy": 0.21834378838539123,
      "num_tokens": 58769781.0,
      "step": 28320
    },
    {
      "entropy": 5.491867876052856,
      "epoch": 2.5693940493468794,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004340235056898882,
      "loss": 4.9115,
      "mean_token_accuracy": 0.2135869473218918,
      "num_tokens": 58779384.0,
      "step": 28325
    },
    {
      "entropy": 5.597129678726196,
      "epoch": 2.5698476052249637,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043400061374667374,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20939935892820358,
      "num_tokens": 58791953.0,
      "step": 28330
    },
    {
      "entropy": 5.466710376739502,
      "epoch": 2.570301161103048,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004339777185152862,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22059821784496308,
      "num_tokens": 58802415.0,
      "step": 28335
    },
    {
      "entropy": 5.564668750762939,
      "epoch": 2.5707547169811322,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004339548199961989,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21577564179897307,
      "num_tokens": 58812582.0,
      "step": 28340
    },
    {
      "entropy": 5.516291713714599,
      "epoch": 2.5712082728592165,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043393191818988543,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21318424195051194,
      "num_tokens": 58822987.0,
      "step": 28345
    },
    {
      "entropy": 5.498049926757813,
      "epoch": 2.5716618287373003,
      "grad_norm": 1.25,
      "learning_rate": 0.0004339090130968193,
      "loss": 4.8697,
      "mean_token_accuracy": 0.2186885729432106,
      "num_tokens": 58832651.0,
      "step": 28350
    },
    {
      "entropy": 5.527301597595215,
      "epoch": 2.5721153846153846,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004338861047174743,
      "loss": 4.928,
      "mean_token_accuracy": 0.21318465918302537,
      "num_tokens": 58842493.0,
      "step": 28355
    },
    {
      "entropy": 5.474476671218872,
      "epoch": 2.572568940493469,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004338631930523242,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22572716027498246,
      "num_tokens": 58852428.0,
      "step": 28360
    },
    {
      "entropy": 5.450385618209839,
      "epoch": 2.573022496371553,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004338402781018429,
      "loss": 4.8373,
      "mean_token_accuracy": 0.22232268899679183,
      "num_tokens": 58863234.0,
      "step": 28365
    },
    {
      "entropy": 5.455707788467407,
      "epoch": 2.573476052249637,
      "grad_norm": 1.21875,
      "learning_rate": 0.000433817359866504,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21828245669603347,
      "num_tokens": 58872632.0,
      "step": 28370
    },
    {
      "entropy": 5.614319562911987,
      "epoch": 2.573929608127721,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004337944383467817,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20388540923595427,
      "num_tokens": 58882774.0,
      "step": 28375
    },
    {
      "entropy": 5.455606698989868,
      "epoch": 2.5743831640058055,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043377151354314996,
      "loss": 4.849,
      "mean_token_accuracy": 0.22130952775478363,
      "num_tokens": 58892380.0,
      "step": 28380
    },
    {
      "entropy": 5.553823232650757,
      "epoch": 2.5748367198838897,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004337485854560829,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21346257627010345,
      "num_tokens": 58902659.0,
      "step": 28385
    },
    {
      "entropy": 5.475580501556396,
      "epoch": 2.575290275761974,
      "grad_norm": 1.125,
      "learning_rate": 0.0004337256540860545,
      "loss": 4.8858,
      "mean_token_accuracy": 0.2171442523598671,
      "num_tokens": 58913979.0,
      "step": 28390
    },
    {
      "entropy": 5.527932119369507,
      "epoch": 2.5757438316400583,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004337027194335393,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22058646380901337,
      "num_tokens": 58924481.0,
      "step": 28395
    },
    {
      "entropy": 5.511540269851684,
      "epoch": 2.576197387518142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043367978149901125,
      "loss": 4.9423,
      "mean_token_accuracy": 0.20822691172361374,
      "num_tokens": 58936287.0,
      "step": 28400
    },
    {
      "entropy": 5.56310338973999,
      "epoch": 2.5766509433962264,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043365684028294493,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21415733844041823,
      "num_tokens": 58946070.0,
      "step": 28405
    },
    {
      "entropy": 5.59272027015686,
      "epoch": 2.5771044992743106,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043363389578581483,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2040593296289444,
      "num_tokens": 58956083.0,
      "step": 28410
    },
    {
      "entropy": 5.539585447311401,
      "epoch": 2.577558055152395,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004336109480080953,
      "loss": 4.9119,
      "mean_token_accuracy": 0.2091962590813637,
      "num_tokens": 58967557.0,
      "step": 28415
    },
    {
      "entropy": 5.497257471084595,
      "epoch": 2.5780116110304787,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004335879969502608,
      "loss": 4.8803,
      "mean_token_accuracy": 0.21549337208271027,
      "num_tokens": 58977509.0,
      "step": 28420
    },
    {
      "entropy": 5.416629695892334,
      "epoch": 2.578465166908563,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043356504261278625,
      "loss": 4.7521,
      "mean_token_accuracy": 0.23429533541202546,
      "num_tokens": 58987498.0,
      "step": 28425
    },
    {
      "entropy": 5.561904621124268,
      "epoch": 2.5789187227866472,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004335420849961462,
      "loss": 5.0268,
      "mean_token_accuracy": 0.20248255729675294,
      "num_tokens": 58998503.0,
      "step": 28430
    },
    {
      "entropy": 5.493903875350952,
      "epoch": 2.5793722786647315,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004335191241008154,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2124171495437622,
      "num_tokens": 59007384.0,
      "step": 28435
    },
    {
      "entropy": 5.5975525856018065,
      "epoch": 2.5798258345428158,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004334961599272687,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21256424188613893,
      "num_tokens": 59017739.0,
      "step": 28440
    },
    {
      "entropy": 5.573173475265503,
      "epoch": 2.5802793904209,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004334731924759811,
      "loss": 4.9247,
      "mean_token_accuracy": 0.21728047877550125,
      "num_tokens": 59028439.0,
      "step": 28445
    },
    {
      "entropy": 5.498218822479248,
      "epoch": 2.5807329462989843,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004334502217474274,
      "loss": 4.8891,
      "mean_token_accuracy": 0.21522742360830308,
      "num_tokens": 59039250.0,
      "step": 28450
    },
    {
      "entropy": 5.546133613586425,
      "epoch": 2.581186502177068,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043342724774208277,
      "loss": 5.0014,
      "mean_token_accuracy": 0.216753788292408,
      "num_tokens": 59049375.0,
      "step": 28455
    },
    {
      "entropy": 5.590890216827392,
      "epoch": 2.5816400580551524,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043340427046042237,
      "loss": 5.007,
      "mean_token_accuracy": 0.21226588040590286,
      "num_tokens": 59059873.0,
      "step": 28460
    },
    {
      "entropy": 5.528668165206909,
      "epoch": 2.5820936139332367,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004333812899029212,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21904878318309784,
      "num_tokens": 59069288.0,
      "step": 28465
    },
    {
      "entropy": 5.511293172836304,
      "epoch": 2.5825471698113205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043335830607005456,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21105093210935594,
      "num_tokens": 59079339.0,
      "step": 28470
    },
    {
      "entropy": 5.527543449401856,
      "epoch": 2.5830007256894048,
      "grad_norm": 1.125,
      "learning_rate": 0.0004333353189622979,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2143719896674156,
      "num_tokens": 59089883.0,
      "step": 28475
    },
    {
      "entropy": 5.524476003646851,
      "epoch": 2.583454281567489,
      "grad_norm": 2.75,
      "learning_rate": 0.0004333123285801264,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21947799623012543,
      "num_tokens": 59100532.0,
      "step": 28480
    },
    {
      "entropy": 5.518876600265503,
      "epoch": 2.5839078374455733,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004332893349240156,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21807062029838561,
      "num_tokens": 59111295.0,
      "step": 28485
    },
    {
      "entropy": 5.551098585128784,
      "epoch": 2.5843613933236576,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004332663379944411,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21073273718357086,
      "num_tokens": 59120808.0,
      "step": 28490
    },
    {
      "entropy": 5.574889707565307,
      "epoch": 2.584814949201742,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043324333779187844,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21958013474941254,
      "num_tokens": 59131438.0,
      "step": 28495
    },
    {
      "entropy": 5.459694766998291,
      "epoch": 2.585268505079826,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043322033431680314,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22098694890737533,
      "num_tokens": 59141701.0,
      "step": 28500
    },
    {
      "entropy": 5.474941873550415,
      "epoch": 2.58572206095791,
      "grad_norm": 1.1875,
      "learning_rate": 0.000433197327569691,
      "loss": 4.8847,
      "mean_token_accuracy": 0.2176327422261238,
      "num_tokens": 59152184.0,
      "step": 28505
    },
    {
      "entropy": 5.511813402175903,
      "epoch": 2.586175616835994,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043317431755101785,
      "loss": 4.8497,
      "mean_token_accuracy": 0.22375211864709854,
      "num_tokens": 59162177.0,
      "step": 28510
    },
    {
      "entropy": 5.584166860580444,
      "epoch": 2.5866291727140784,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004331513042612594,
      "loss": 5.0085,
      "mean_token_accuracy": 0.21243423968553543,
      "num_tokens": 59171964.0,
      "step": 28515
    },
    {
      "entropy": 5.511908102035522,
      "epoch": 2.5870827285921627,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004331282877008918,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22004627734422683,
      "num_tokens": 59182224.0,
      "step": 28520
    },
    {
      "entropy": 5.521733045578003,
      "epoch": 2.5875362844702465,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004331052678703909,
      "loss": 4.913,
      "mean_token_accuracy": 0.21449595540761948,
      "num_tokens": 59191912.0,
      "step": 28525
    },
    {
      "entropy": 5.507656049728394,
      "epoch": 2.587989840348331,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004330822447702327,
      "loss": 4.8947,
      "mean_token_accuracy": 0.20993456691503526,
      "num_tokens": 59200984.0,
      "step": 28530
    },
    {
      "entropy": 5.555482196807861,
      "epoch": 2.588443396226415,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043305921840089347,
      "loss": 4.95,
      "mean_token_accuracy": 0.21745003759860992,
      "num_tokens": 59210672.0,
      "step": 28535
    },
    {
      "entropy": 5.530939960479737,
      "epoch": 2.5888969521044993,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043303618876284935,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2157348409295082,
      "num_tokens": 59220230.0,
      "step": 28540
    },
    {
      "entropy": 5.4644629001617435,
      "epoch": 2.5893505079825836,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043301315585657645,
      "loss": 4.8597,
      "mean_token_accuracy": 0.21800754368305206,
      "num_tokens": 59230615.0,
      "step": 28545
    },
    {
      "entropy": 5.546888399124145,
      "epoch": 2.589804063860668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004329901196825513,
      "loss": 4.9692,
      "mean_token_accuracy": 0.21738160848617555,
      "num_tokens": 59241717.0,
      "step": 28550
    },
    {
      "entropy": 5.584340286254883,
      "epoch": 2.5902576197387517,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043296708024125025,
      "loss": 4.9836,
      "mean_token_accuracy": 0.2095291867852211,
      "num_tokens": 59251791.0,
      "step": 28555
    },
    {
      "entropy": 5.485751914978027,
      "epoch": 2.590711175616836,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043294403753314965,
      "loss": 4.8092,
      "mean_token_accuracy": 0.22592369914054872,
      "num_tokens": 59262434.0,
      "step": 28560
    },
    {
      "entropy": 5.503426218032837,
      "epoch": 2.59116473149492,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004329209915587261,
      "loss": 4.988,
      "mean_token_accuracy": 0.20915014594793319,
      "num_tokens": 59272138.0,
      "step": 28565
    },
    {
      "entropy": 5.486407661437989,
      "epoch": 2.5916182873730045,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043289794231845624,
      "loss": 4.9365,
      "mean_token_accuracy": 0.2142576038837433,
      "num_tokens": 59283470.0,
      "step": 28570
    },
    {
      "entropy": 5.468969535827637,
      "epoch": 2.5920718432510883,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004328748898128168,
      "loss": 4.9353,
      "mean_token_accuracy": 0.21285001188516617,
      "num_tokens": 59293670.0,
      "step": 28575
    },
    {
      "entropy": 5.54750247001648,
      "epoch": 2.5925253991291726,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004328518340422842,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21523593217134476,
      "num_tokens": 59303728.0,
      "step": 28580
    },
    {
      "entropy": 5.519813013076782,
      "epoch": 2.592978955007257,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043282877500733555,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2209109589457512,
      "num_tokens": 59314271.0,
      "step": 28585
    },
    {
      "entropy": 5.518555116653443,
      "epoch": 2.593432510885341,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004328057127084476,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21364410519599913,
      "num_tokens": 59324581.0,
      "step": 28590
    },
    {
      "entropy": 5.5035045623779295,
      "epoch": 2.5938860667634254,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043278264714609733,
      "loss": 4.9121,
      "mean_token_accuracy": 0.21637875139713286,
      "num_tokens": 59333912.0,
      "step": 28595
    },
    {
      "entropy": 5.4700884342193605,
      "epoch": 2.5943396226415096,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043275957832076173,
      "loss": 4.8647,
      "mean_token_accuracy": 0.21697210371494294,
      "num_tokens": 59344245.0,
      "step": 28600
    },
    {
      "entropy": 5.518507385253907,
      "epoch": 2.5947931785195935,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043273650623291775,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21529100984334945,
      "num_tokens": 59353825.0,
      "step": 28605
    },
    {
      "entropy": 5.515340566635132,
      "epoch": 2.5952467343976777,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004327134308830427,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21119250506162643,
      "num_tokens": 59364901.0,
      "step": 28610
    },
    {
      "entropy": 5.496701049804687,
      "epoch": 2.595700290275762,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004326903522716137,
      "loss": 4.8937,
      "mean_token_accuracy": 0.21021356880664827,
      "num_tokens": 59374772.0,
      "step": 28615
    },
    {
      "entropy": 5.450021409988404,
      "epoch": 2.5961538461538463,
      "grad_norm": 1.234375,
      "learning_rate": 0.000432667270399108,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22559859454631806,
      "num_tokens": 59383640.0,
      "step": 28620
    },
    {
      "entropy": 5.523759126663208,
      "epoch": 2.59660740203193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043264418526600305,
      "loss": 4.8811,
      "mean_token_accuracy": 0.21806788295507432,
      "num_tokens": 59392928.0,
      "step": 28625
    },
    {
      "entropy": 5.565279388427735,
      "epoch": 2.5970609579100143,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004326210968727761,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2071376398205757,
      "num_tokens": 59402969.0,
      "step": 28630
    },
    {
      "entropy": 5.5832356929779055,
      "epoch": 2.5975145137880986,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004325980052199047,
      "loss": 4.99,
      "mean_token_accuracy": 0.20268357396125794,
      "num_tokens": 59413439.0,
      "step": 28635
    },
    {
      "entropy": 5.58742847442627,
      "epoch": 2.597968069666183,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043257491030786644,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21679740846157075,
      "num_tokens": 59423403.0,
      "step": 28640
    },
    {
      "entropy": 5.525634813308716,
      "epoch": 2.598421625544267,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043255181213713883,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20762742608785628,
      "num_tokens": 59434519.0,
      "step": 28645
    },
    {
      "entropy": 5.53591136932373,
      "epoch": 2.5988751814223514,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004325287107081996,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2096611425280571,
      "num_tokens": 59445056.0,
      "step": 28650
    },
    {
      "entropy": 5.508650445938111,
      "epoch": 2.5993287373004357,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004325056060215265,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2145665094256401,
      "num_tokens": 59455037.0,
      "step": 28655
    },
    {
      "entropy": 5.50084171295166,
      "epoch": 2.5997822931785195,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004324824980775973,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21044404655694962,
      "num_tokens": 59465093.0,
      "step": 28660
    },
    {
      "entropy": 5.503364515304566,
      "epoch": 2.6002358490566038,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043245938687688995,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21246374398469925,
      "num_tokens": 59475449.0,
      "step": 28665
    },
    {
      "entropy": 5.560793256759643,
      "epoch": 2.600689404934688,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043243627241988235,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2096658915281296,
      "num_tokens": 59484990.0,
      "step": 28670
    },
    {
      "entropy": 5.5800196647644045,
      "epoch": 2.6011429608127723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043241315470705254,
      "loss": 4.9304,
      "mean_token_accuracy": 0.2114672601222992,
      "num_tokens": 59495739.0,
      "step": 28675
    },
    {
      "entropy": 5.531115531921387,
      "epoch": 2.601596516690856,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043239003373887855,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21694314032793044,
      "num_tokens": 59506432.0,
      "step": 28680
    },
    {
      "entropy": 5.530005359649659,
      "epoch": 2.6020500725689404,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004323669095158385,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21335145831108093,
      "num_tokens": 59516585.0,
      "step": 28685
    },
    {
      "entropy": 5.497039127349853,
      "epoch": 2.6025036284470247,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004323437820384107,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21011703163385392,
      "num_tokens": 59526824.0,
      "step": 28690
    },
    {
      "entropy": 5.561870002746582,
      "epoch": 2.602957184325109,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004323206513070734,
      "loss": 4.9573,
      "mean_token_accuracy": 0.2124274656176567,
      "num_tokens": 59537749.0,
      "step": 28695
    },
    {
      "entropy": 5.579768800735474,
      "epoch": 2.603410740203193,
      "grad_norm": 1.171875,
      "learning_rate": 0.000432297517322305,
      "loss": 4.919,
      "mean_token_accuracy": 0.2060007095336914,
      "num_tokens": 59548310.0,
      "step": 28700
    },
    {
      "entropy": 5.543799543380738,
      "epoch": 2.6038642960812775,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004322743800845838,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21550447791814803,
      "num_tokens": 59558191.0,
      "step": 28705
    },
    {
      "entropy": 5.5542418479919435,
      "epoch": 2.6043178519593613,
      "grad_norm": 1.125,
      "learning_rate": 0.00043225123959438836,
      "loss": 4.9889,
      "mean_token_accuracy": 0.19904627799987792,
      "num_tokens": 59568961.0,
      "step": 28710
    },
    {
      "entropy": 5.537586832046509,
      "epoch": 2.6047714078374455,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043222809585219716,
      "loss": 4.9231,
      "mean_token_accuracy": 0.21050258874893188,
      "num_tokens": 59579137.0,
      "step": 28715
    },
    {
      "entropy": 5.388093757629394,
      "epoch": 2.60522496371553,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043220494885848896,
      "loss": 4.727,
      "mean_token_accuracy": 0.22887020707130432,
      "num_tokens": 59589612.0,
      "step": 28720
    },
    {
      "entropy": 5.545928144454956,
      "epoch": 2.605678519593614,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004321817986137423,
      "loss": 4.9611,
      "mean_token_accuracy": 0.2099051907658577,
      "num_tokens": 59600332.0,
      "step": 28725
    },
    {
      "entropy": 5.445740222930908,
      "epoch": 2.606132075471698,
      "grad_norm": 1.21875,
      "learning_rate": 0.000432158645118436,
      "loss": 4.8353,
      "mean_token_accuracy": 0.2302090272307396,
      "num_tokens": 59610491.0,
      "step": 28730
    },
    {
      "entropy": 5.63829550743103,
      "epoch": 2.606585631349782,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004321354883730488,
      "loss": 5.0801,
      "mean_token_accuracy": 0.20283650159835814,
      "num_tokens": 59620729.0,
      "step": 28735
    },
    {
      "entropy": 5.502729272842407,
      "epoch": 2.6070391872278664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043211232837805973,
      "loss": 4.8741,
      "mean_token_accuracy": 0.22019591480493544,
      "num_tokens": 59631243.0,
      "step": 28740
    },
    {
      "entropy": 5.529031705856323,
      "epoch": 2.6074927431059507,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043208916513394767,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22065686881542207,
      "num_tokens": 59642055.0,
      "step": 28745
    },
    {
      "entropy": 5.562448740005493,
      "epoch": 2.607946298984035,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043206599864119164,
      "loss": 4.9884,
      "mean_token_accuracy": 0.2103422462940216,
      "num_tokens": 59652144.0,
      "step": 28750
    },
    {
      "entropy": 5.508689260482788,
      "epoch": 2.6083998548621192,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043204282890027074,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22429406493902207,
      "num_tokens": 59662367.0,
      "step": 28755
    },
    {
      "entropy": 5.552458477020264,
      "epoch": 2.608853410740203,
      "grad_norm": 1.125,
      "learning_rate": 0.00043201965591166413,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21458108425140382,
      "num_tokens": 59672729.0,
      "step": 28760
    },
    {
      "entropy": 5.578047609329223,
      "epoch": 2.6093069666182873,
      "grad_norm": 1.140625,
      "learning_rate": 0.000431996479675851,
      "loss": 4.9779,
      "mean_token_accuracy": 0.22175480425357819,
      "num_tokens": 59683282.0,
      "step": 28765
    },
    {
      "entropy": 5.585192537307739,
      "epoch": 2.6097605224963716,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043197330019331074,
      "loss": 5.0277,
      "mean_token_accuracy": 0.2032526522874832,
      "num_tokens": 59694297.0,
      "step": 28770
    },
    {
      "entropy": 5.460672426223755,
      "epoch": 2.610214078374456,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043195011746452244,
      "loss": 4.8568,
      "mean_token_accuracy": 0.21983457654714583,
      "num_tokens": 59704149.0,
      "step": 28775
    },
    {
      "entropy": 5.5555281162261965,
      "epoch": 2.6106676342525397,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043192693148996585,
      "loss": 4.9667,
      "mean_token_accuracy": 0.2182912364602089,
      "num_tokens": 59714811.0,
      "step": 28780
    },
    {
      "entropy": 5.49904613494873,
      "epoch": 2.611121190130624,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004319037422701203,
      "loss": 4.8845,
      "mean_token_accuracy": 0.22046362310647966,
      "num_tokens": 59724950.0,
      "step": 28785
    },
    {
      "entropy": 5.549964046478271,
      "epoch": 2.611574746008708,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004318805498054654,
      "loss": 5.0348,
      "mean_token_accuracy": 0.19872065633535385,
      "num_tokens": 59735905.0,
      "step": 28790
    },
    {
      "entropy": 5.6039002418518065,
      "epoch": 2.6120283018867925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043185735409648077,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21075978577136995,
      "num_tokens": 59746743.0,
      "step": 28795
    },
    {
      "entropy": 5.56124210357666,
      "epoch": 2.6124818577648767,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043183415514364606,
      "loss": 4.9184,
      "mean_token_accuracy": 0.21381641924381256,
      "num_tokens": 59757205.0,
      "step": 28800
    },
    {
      "entropy": 5.454240703582764,
      "epoch": 2.612935413642961,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004318109529474411,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21552783101797104,
      "num_tokens": 59767691.0,
      "step": 28805
    },
    {
      "entropy": 5.476266956329345,
      "epoch": 2.6133889695210453,
      "grad_norm": 1.125,
      "learning_rate": 0.0004317877475083456,
      "loss": 4.942,
      "mean_token_accuracy": 0.21620311439037324,
      "num_tokens": 59778618.0,
      "step": 28810
    },
    {
      "entropy": 5.560245561599731,
      "epoch": 2.613842525399129,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043176453882683955,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20953458994627,
      "num_tokens": 59789423.0,
      "step": 28815
    },
    {
      "entropy": 5.594815969467163,
      "epoch": 2.6142960812772134,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043174132690340287,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20201612561941146,
      "num_tokens": 59800021.0,
      "step": 28820
    },
    {
      "entropy": 5.4969874858856205,
      "epoch": 2.6147496371552976,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004317181117385156,
      "loss": 4.8919,
      "mean_token_accuracy": 0.20912107974290847,
      "num_tokens": 59810006.0,
      "step": 28825
    },
    {
      "entropy": 5.534179019927978,
      "epoch": 2.6152031930333814,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043169489333265795,
      "loss": 4.951,
      "mean_token_accuracy": 0.2081662893295288,
      "num_tokens": 59819505.0,
      "step": 28830
    },
    {
      "entropy": 5.555849361419678,
      "epoch": 2.6156567489114657,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043167167168630977,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2156184807419777,
      "num_tokens": 59830439.0,
      "step": 28835
    },
    {
      "entropy": 5.525593614578247,
      "epoch": 2.61611030478955,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004316484467999516,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21081707626581192,
      "num_tokens": 59841281.0,
      "step": 28840
    },
    {
      "entropy": 5.518883752822876,
      "epoch": 2.6165638606676342,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043162521867406376,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20398472100496293,
      "num_tokens": 59853581.0,
      "step": 28845
    },
    {
      "entropy": 5.605485153198242,
      "epoch": 2.6170174165457185,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004316019873091263,
      "loss": 4.9973,
      "mean_token_accuracy": 0.21135033071041107,
      "num_tokens": 59864787.0,
      "step": 28850
    },
    {
      "entropy": 5.576481533050537,
      "epoch": 2.6174709724238028,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004315787527056199,
      "loss": 4.951,
      "mean_token_accuracy": 0.21697171330451964,
      "num_tokens": 59875244.0,
      "step": 28855
    },
    {
      "entropy": 5.497420310974121,
      "epoch": 2.617924528301887,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004315555148640251,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2159500852227211,
      "num_tokens": 59886535.0,
      "step": 28860
    },
    {
      "entropy": 5.5382331848144535,
      "epoch": 2.618378084179971,
      "grad_norm": 1.125,
      "learning_rate": 0.0004315322737848222,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2105987548828125,
      "num_tokens": 59896298.0,
      "step": 28865
    },
    {
      "entropy": 5.440474510192871,
      "epoch": 2.618831640058055,
      "grad_norm": 1.1875,
      "learning_rate": 0.000431509029468492,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22039612680673598,
      "num_tokens": 59907010.0,
      "step": 28870
    },
    {
      "entropy": 5.559852600097656,
      "epoch": 2.6192851959361394,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004314857819155153,
      "loss": 4.9013,
      "mean_token_accuracy": 0.21951362490653992,
      "num_tokens": 59916247.0,
      "step": 28875
    },
    {
      "entropy": 5.45871376991272,
      "epoch": 2.6197387518142237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043146253112637267,
      "loss": 4.8314,
      "mean_token_accuracy": 0.22198484390974044,
      "num_tokens": 59926680.0,
      "step": 28880
    },
    {
      "entropy": 5.4570149898529055,
      "epoch": 2.6201923076923075,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043143927710154504,
      "loss": 4.8521,
      "mean_token_accuracy": 0.22045730501413346,
      "num_tokens": 59936346.0,
      "step": 28885
    },
    {
      "entropy": 5.505706834793091,
      "epoch": 2.6206458635703918,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004314160198415133,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22199222296476365,
      "num_tokens": 59947007.0,
      "step": 28890
    },
    {
      "entropy": 5.543805456161499,
      "epoch": 2.621099419448476,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043139275934675844,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21432872712612153,
      "num_tokens": 59957123.0,
      "step": 28895
    },
    {
      "entropy": 5.51826434135437,
      "epoch": 2.6215529753265603,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004313694956177613,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21661525517702102,
      "num_tokens": 59966566.0,
      "step": 28900
    },
    {
      "entropy": 5.534532976150513,
      "epoch": 2.6220065312046446,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043134622865500334,
      "loss": 4.9244,
      "mean_token_accuracy": 0.20877003967761992,
      "num_tokens": 59976065.0,
      "step": 28905
    },
    {
      "entropy": 5.585506200790405,
      "epoch": 2.622460087082729,
      "grad_norm": 1.25,
      "learning_rate": 0.0004313229584589655,
      "loss": 4.933,
      "mean_token_accuracy": 0.2137759655714035,
      "num_tokens": 59986201.0,
      "step": 28910
    },
    {
      "entropy": 5.505525302886963,
      "epoch": 2.6229136429608126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000431299685030129,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21248521506786347,
      "num_tokens": 59996574.0,
      "step": 28915
    },
    {
      "entropy": 5.512580919265747,
      "epoch": 2.623367198838897,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043127640836897516,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21798425763845444,
      "num_tokens": 60006082.0,
      "step": 28920
    },
    {
      "entropy": 5.528341341018677,
      "epoch": 2.623820754716981,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004312531284759854,
      "loss": 4.9324,
      "mean_token_accuracy": 0.20959401428699492,
      "num_tokens": 60015483.0,
      "step": 28925
    },
    {
      "entropy": 5.512495851516723,
      "epoch": 2.6242743105950654,
      "grad_norm": 1.421875,
      "learning_rate": 0.00043122984535164116,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22911366820335388,
      "num_tokens": 60025984.0,
      "step": 28930
    },
    {
      "entropy": 5.55941071510315,
      "epoch": 2.6247278664731493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043120655899642386,
      "loss": 4.93,
      "mean_token_accuracy": 0.2076868385076523,
      "num_tokens": 60036109.0,
      "step": 28935
    },
    {
      "entropy": 5.615510129928589,
      "epoch": 2.6251814223512335,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043118326941081513,
      "loss": 4.9935,
      "mean_token_accuracy": 0.20689695328474045,
      "num_tokens": 60046546.0,
      "step": 28940
    },
    {
      "entropy": 5.583363437652588,
      "epoch": 2.625634978229318,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004311599765952966,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2163356751203537,
      "num_tokens": 60056768.0,
      "step": 28945
    },
    {
      "entropy": 5.552965116500855,
      "epoch": 2.626088534107402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043113668055034995,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20677520781755448,
      "num_tokens": 60066962.0,
      "step": 28950
    },
    {
      "entropy": 5.526785373687744,
      "epoch": 2.6265420899854863,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004311133812764569,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21789804995059966,
      "num_tokens": 60076477.0,
      "step": 28955
    },
    {
      "entropy": 5.4595733165740965,
      "epoch": 2.6269956458635706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043109007877409946,
      "loss": 4.8094,
      "mean_token_accuracy": 0.22258938997983932,
      "num_tokens": 60088060.0,
      "step": 28960
    },
    {
      "entropy": 5.510951566696167,
      "epoch": 2.6274492017416544,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043106677304375935,
      "loss": 4.909,
      "mean_token_accuracy": 0.2175913155078888,
      "num_tokens": 60098677.0,
      "step": 28965
    },
    {
      "entropy": 5.5853455543518065,
      "epoch": 2.6279027576197387,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004310434640859186,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20754190981388093,
      "num_tokens": 60109054.0,
      "step": 28970
    },
    {
      "entropy": 5.515770101547242,
      "epoch": 2.628356313497823,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004310201519010592,
      "loss": 4.8409,
      "mean_token_accuracy": 0.22024760991334916,
      "num_tokens": 60119951.0,
      "step": 28975
    },
    {
      "entropy": 5.432485294342041,
      "epoch": 2.628809869375907,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043099683648966335,
      "loss": 4.8251,
      "mean_token_accuracy": 0.21306003630161285,
      "num_tokens": 60130026.0,
      "step": 28980
    },
    {
      "entropy": 5.4275596141815186,
      "epoch": 2.629263425253991,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004309735178522132,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22112511247396469,
      "num_tokens": 60140748.0,
      "step": 28985
    },
    {
      "entropy": 5.504708671569825,
      "epoch": 2.6297169811320753,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004309501959891909,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21946708112955093,
      "num_tokens": 60151042.0,
      "step": 28990
    },
    {
      "entropy": 5.561709356307984,
      "epoch": 2.6301705370101596,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043092687090107883,
      "loss": 4.93,
      "mean_token_accuracy": 0.20676448494195937,
      "num_tokens": 60160758.0,
      "step": 28995
    },
    {
      "entropy": 5.490038824081421,
      "epoch": 2.630624092888244,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043090354258835934,
      "loss": 4.934,
      "mean_token_accuracy": 0.21050809174776078,
      "num_tokens": 60170249.0,
      "step": 29000
    },
    {
      "entropy": 5.459096765518188,
      "epoch": 2.631077648766328,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043088021105151483,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21335221976041793,
      "num_tokens": 60181216.0,
      "step": 29005
    },
    {
      "entropy": 5.498613595962524,
      "epoch": 2.6315312046444124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004308568762910279,
      "loss": 4.8848,
      "mean_token_accuracy": 0.21757752150297166,
      "num_tokens": 60190559.0,
      "step": 29010
    },
    {
      "entropy": 5.61933217048645,
      "epoch": 2.6319847605224966,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043083353830738097,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2060284808278084,
      "num_tokens": 60201268.0,
      "step": 29015
    },
    {
      "entropy": 5.4622515678405765,
      "epoch": 2.6324383164005805,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043081019710105683,
      "loss": 4.7889,
      "mean_token_accuracy": 0.21759428679943085,
      "num_tokens": 60210907.0,
      "step": 29020
    },
    {
      "entropy": 5.4503264904022215,
      "epoch": 2.6328918722786647,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004307868526725381,
      "loss": 4.8364,
      "mean_token_accuracy": 0.22276439815759658,
      "num_tokens": 60220298.0,
      "step": 29025
    },
    {
      "entropy": 5.434282636642456,
      "epoch": 2.633345428156749,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004307635050223075,
      "loss": 4.8711,
      "mean_token_accuracy": 0.22295601814985275,
      "num_tokens": 60230811.0,
      "step": 29030
    },
    {
      "entropy": 5.509660768508911,
      "epoch": 2.6337989840348333,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000430740154150848,
      "loss": 4.8397,
      "mean_token_accuracy": 0.21989746689796447,
      "num_tokens": 60241211.0,
      "step": 29035
    },
    {
      "entropy": 5.471844291687011,
      "epoch": 2.634252539912917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004307168000586423,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22211599498987197,
      "num_tokens": 60251506.0,
      "step": 29040
    },
    {
      "entropy": 5.451631021499634,
      "epoch": 2.6347060957910013,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004306934427461736,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23554385900497438,
      "num_tokens": 60261416.0,
      "step": 29045
    },
    {
      "entropy": 5.5329022884368895,
      "epoch": 2.6351596516690856,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004306700822139248,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21408605128526687,
      "num_tokens": 60272315.0,
      "step": 29050
    },
    {
      "entropy": 5.51502103805542,
      "epoch": 2.63561320754717,
      "grad_norm": 1.15625,
      "learning_rate": 0.000430646718462379,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21138989329338073,
      "num_tokens": 60282722.0,
      "step": 29055
    },
    {
      "entropy": 5.495158720016479,
      "epoch": 2.636066763425254,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004306233514920194,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2150113493204117,
      "num_tokens": 60292636.0,
      "step": 29060
    },
    {
      "entropy": 5.510214757919312,
      "epoch": 2.6365203193033384,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043059998130332933,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2177960082888603,
      "num_tokens": 60302979.0,
      "step": 29065
    },
    {
      "entropy": 5.518082761764527,
      "epoch": 2.6369738751814222,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043057660789679194,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21595758944749832,
      "num_tokens": 60313181.0,
      "step": 29070
    },
    {
      "entropy": 5.572794961929321,
      "epoch": 2.6374274310595065,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004305532312728906,
      "loss": 4.893,
      "mean_token_accuracy": 0.21098823547363282,
      "num_tokens": 60323401.0,
      "step": 29075
    },
    {
      "entropy": 5.5677135467529295,
      "epoch": 2.6378809869375908,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004305298514321088,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20664920061826705,
      "num_tokens": 60333478.0,
      "step": 29080
    },
    {
      "entropy": 5.496002149581909,
      "epoch": 2.638334542815675,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043050646837493014,
      "loss": 4.876,
      "mean_token_accuracy": 0.2160879999399185,
      "num_tokens": 60343980.0,
      "step": 29085
    },
    {
      "entropy": 5.5128497123718265,
      "epoch": 2.638788098693759,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004304830821018381,
      "loss": 4.888,
      "mean_token_accuracy": 0.2214886099100113,
      "num_tokens": 60354262.0,
      "step": 29090
    },
    {
      "entropy": 5.551802730560302,
      "epoch": 2.639241654571843,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004304596926133162,
      "loss": 4.9821,
      "mean_token_accuracy": 0.213053897023201,
      "num_tokens": 60364137.0,
      "step": 29095
    },
    {
      "entropy": 5.5155517578125,
      "epoch": 2.6396952104499274,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043043629990984823,
      "loss": 4.9561,
      "mean_token_accuracy": 0.20881974697113037,
      "num_tokens": 60374528.0,
      "step": 29100
    },
    {
      "entropy": 5.484981679916382,
      "epoch": 2.6401487663280117,
      "grad_norm": 1.25,
      "learning_rate": 0.00043041290399191797,
      "loss": 4.945,
      "mean_token_accuracy": 0.2179246127605438,
      "num_tokens": 60384880.0,
      "step": 29105
    },
    {
      "entropy": 5.503172445297241,
      "epoch": 2.640602322206096,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004303895048600093,
      "loss": 4.901,
      "mean_token_accuracy": 0.21249927282333375,
      "num_tokens": 60395543.0,
      "step": 29110
    },
    {
      "entropy": 5.564700317382813,
      "epoch": 2.64105587808418,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004303661025146061,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22175834476947784,
      "num_tokens": 60405432.0,
      "step": 29115
    },
    {
      "entropy": 5.581679821014404,
      "epoch": 2.641509433962264,
      "grad_norm": 1.125,
      "learning_rate": 0.0004303426969561922,
      "loss": 4.976,
      "mean_token_accuracy": 0.20390943735837935,
      "num_tokens": 60416360.0,
      "step": 29120
    },
    {
      "entropy": 5.552780103683472,
      "epoch": 2.6419629898403483,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004303192881852518,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21306895911693574,
      "num_tokens": 60426371.0,
      "step": 29125
    },
    {
      "entropy": 5.555319452285767,
      "epoch": 2.6424165457184325,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043029587620226897,
      "loss": 4.933,
      "mean_token_accuracy": 0.2140775963664055,
      "num_tokens": 60437505.0,
      "step": 29130
    },
    {
      "entropy": 5.538500690460205,
      "epoch": 2.642870101596517,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004302724610077278,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21363855004310608,
      "num_tokens": 60448170.0,
      "step": 29135
    },
    {
      "entropy": 5.577473974227905,
      "epoch": 2.6433236574746006,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004302490426021126,
      "loss": 5.003,
      "mean_token_accuracy": 0.20730844289064407,
      "num_tokens": 60458776.0,
      "step": 29140
    },
    {
      "entropy": 5.532087993621826,
      "epoch": 2.643777213352685,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004302256209859076,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2262054592370987,
      "num_tokens": 60468279.0,
      "step": 29145
    },
    {
      "entropy": 5.532051610946655,
      "epoch": 2.644230769230769,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043020219615959727,
      "loss": 4.984,
      "mean_token_accuracy": 0.20770479142665862,
      "num_tokens": 60478507.0,
      "step": 29150
    },
    {
      "entropy": 5.578500699996948,
      "epoch": 2.6446843251088534,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004301787681236659,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20529094487428665,
      "num_tokens": 60489016.0,
      "step": 29155
    },
    {
      "entropy": 5.592386770248413,
      "epoch": 2.6451378809869377,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004301553368785981,
      "loss": 4.9833,
      "mean_token_accuracy": 0.2058970332145691,
      "num_tokens": 60500169.0,
      "step": 29160
    },
    {
      "entropy": 5.533572435379028,
      "epoch": 2.645591436865022,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004301319024248784,
      "loss": 4.883,
      "mean_token_accuracy": 0.21767321676015855,
      "num_tokens": 60511109.0,
      "step": 29165
    },
    {
      "entropy": 5.523146057128907,
      "epoch": 2.6460449927431062,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043010846476299145,
      "loss": 4.8955,
      "mean_token_accuracy": 0.22105156779289245,
      "num_tokens": 60521735.0,
      "step": 29170
    },
    {
      "entropy": 5.569445180892944,
      "epoch": 2.64649854862119,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004300850238934219,
      "loss": 4.9631,
      "mean_token_accuracy": 0.20607843846082688,
      "num_tokens": 60532071.0,
      "step": 29175
    },
    {
      "entropy": 5.5547194480896,
      "epoch": 2.6469521044992743,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043006157981665446,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21811448633670807,
      "num_tokens": 60541909.0,
      "step": 29180
    },
    {
      "entropy": 5.469224452972412,
      "epoch": 2.6474056603773586,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043003813253317415,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22079289555549622,
      "num_tokens": 60551361.0,
      "step": 29185
    },
    {
      "entropy": 5.48556661605835,
      "epoch": 2.6478592162554424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043001468204346564,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21266020685434342,
      "num_tokens": 60562112.0,
      "step": 29190
    },
    {
      "entropy": 5.556151866912842,
      "epoch": 2.6483127721335267,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042999122834801406,
      "loss": 4.9618,
      "mean_token_accuracy": 0.20834096819162368,
      "num_tokens": 60572827.0,
      "step": 29195
    },
    {
      "entropy": 5.462076997756958,
      "epoch": 2.648766328011611,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004299677714473044,
      "loss": 4.8733,
      "mean_token_accuracy": 0.22373484820127487,
      "num_tokens": 60582222.0,
      "step": 29200
    },
    {
      "entropy": 5.522726345062256,
      "epoch": 2.649219883889695,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004299443113418217,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21997703164815902,
      "num_tokens": 60592943.0,
      "step": 29205
    },
    {
      "entropy": 5.53337926864624,
      "epoch": 2.6496734397677795,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042992084803205124,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21421058177948,
      "num_tokens": 60603206.0,
      "step": 29210
    },
    {
      "entropy": 5.5014848709106445,
      "epoch": 2.6501269956458637,
      "grad_norm": 1.234375,
      "learning_rate": 0.000429897381518478,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22122532576322557,
      "num_tokens": 60614776.0,
      "step": 29215
    },
    {
      "entropy": 5.5081761360168455,
      "epoch": 2.650580551523948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042987391180158757,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2124914437532425,
      "num_tokens": 60625914.0,
      "step": 29220
    },
    {
      "entropy": 5.51166844367981,
      "epoch": 2.651034107402032,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042985043888186516,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22042528688907623,
      "num_tokens": 60636934.0,
      "step": 29225
    },
    {
      "entropy": 5.494295215606689,
      "epoch": 2.651487663280116,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004298269627597962,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21049118787050247,
      "num_tokens": 60647707.0,
      "step": 29230
    },
    {
      "entropy": 5.462206268310547,
      "epoch": 2.6519412191582004,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042980348343586624,
      "loss": 4.8639,
      "mean_token_accuracy": 0.21600887179374695,
      "num_tokens": 60658691.0,
      "step": 29235
    },
    {
      "entropy": 5.520254993438721,
      "epoch": 2.6523947750362846,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042978000091056074,
      "loss": 4.9658,
      "mean_token_accuracy": 0.2106821730732918,
      "num_tokens": 60668479.0,
      "step": 29240
    },
    {
      "entropy": 5.505654907226562,
      "epoch": 2.6528483309143684,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004297565151843653,
      "loss": 4.8278,
      "mean_token_accuracy": 0.21815417110919952,
      "num_tokens": 60679158.0,
      "step": 29245
    },
    {
      "entropy": 5.558917665481568,
      "epoch": 2.6533018867924527,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004297330262577659,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20906615108251572,
      "num_tokens": 60689363.0,
      "step": 29250
    },
    {
      "entropy": 5.475180435180664,
      "epoch": 2.653755442670537,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004297095341312479,
      "loss": 4.8121,
      "mean_token_accuracy": 0.22131693065166474,
      "num_tokens": 60699650.0,
      "step": 29255
    },
    {
      "entropy": 5.496839427947998,
      "epoch": 2.6542089985486212,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004296860388052974,
      "loss": 4.8889,
      "mean_token_accuracy": 0.21212936639785768,
      "num_tokens": 60709921.0,
      "step": 29260
    },
    {
      "entropy": 5.528440237045288,
      "epoch": 2.6546625544267055,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042966254028040015,
      "loss": 4.8758,
      "mean_token_accuracy": 0.22641344219446183,
      "num_tokens": 60720112.0,
      "step": 29265
    },
    {
      "entropy": 5.483203601837158,
      "epoch": 2.65511611030479,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042963903855704216,
      "loss": 4.8446,
      "mean_token_accuracy": 0.211634162068367,
      "num_tokens": 60730604.0,
      "step": 29270
    },
    {
      "entropy": 5.45366358757019,
      "epoch": 2.6555696661828736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042961553363570946,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22004604935646058,
      "num_tokens": 60740896.0,
      "step": 29275
    },
    {
      "entropy": 5.551656723022461,
      "epoch": 2.656023222060958,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004295920255168881,
      "loss": 4.883,
      "mean_token_accuracy": 0.21988172382116317,
      "num_tokens": 60751771.0,
      "step": 29280
    },
    {
      "entropy": 5.566930103302002,
      "epoch": 2.656476777939042,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042956851420106423,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2093845561146736,
      "num_tokens": 60762802.0,
      "step": 29285
    },
    {
      "entropy": 5.439630126953125,
      "epoch": 2.6569303338171264,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004295449996887242,
      "loss": 4.8455,
      "mean_token_accuracy": 0.21708481758832932,
      "num_tokens": 60774102.0,
      "step": 29290
    },
    {
      "entropy": 5.499790906906128,
      "epoch": 2.65738388969521,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004295214819803541,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2099471166729927,
      "num_tokens": 60785383.0,
      "step": 29295
    },
    {
      "entropy": 5.494332122802734,
      "epoch": 2.6578374455732945,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042949796107644053,
      "loss": 4.914,
      "mean_token_accuracy": 0.22100272476673127,
      "num_tokens": 60795173.0,
      "step": 29300
    },
    {
      "entropy": 5.522036218643189,
      "epoch": 2.6582910014513788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042947443697746956,
      "loss": 4.894,
      "mean_token_accuracy": 0.21763916462659835,
      "num_tokens": 60805621.0,
      "step": 29305
    },
    {
      "entropy": 5.574655294418335,
      "epoch": 2.658744557329463,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004294509096839279,
      "loss": 5.0274,
      "mean_token_accuracy": 0.20500217974185944,
      "num_tokens": 60816799.0,
      "step": 29310
    },
    {
      "entropy": 5.445596599578858,
      "epoch": 2.6591981132075473,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042942737919630215,
      "loss": 4.8604,
      "mean_token_accuracy": 0.217450875043869,
      "num_tokens": 60826827.0,
      "step": 29315
    },
    {
      "entropy": 5.509045076370239,
      "epoch": 2.6596516690856316,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042940384551507876,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21547484397888184,
      "num_tokens": 60836567.0,
      "step": 29320
    },
    {
      "entropy": 5.538186740875244,
      "epoch": 2.6601052249637154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004293803086407447,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21494536697864533,
      "num_tokens": 60846723.0,
      "step": 29325
    },
    {
      "entropy": 5.5920497417449955,
      "epoch": 2.6605587808417996,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004293567685737863,
      "loss": 5.018,
      "mean_token_accuracy": 0.2082287684082985,
      "num_tokens": 60857434.0,
      "step": 29330
    },
    {
      "entropy": 5.564563655853272,
      "epoch": 2.661012336719884,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042933322531469074,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21093194633722306,
      "num_tokens": 60868780.0,
      "step": 29335
    },
    {
      "entropy": 5.602622222900391,
      "epoch": 2.661465892597968,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042930967886394456,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21860561966896058,
      "num_tokens": 60879037.0,
      "step": 29340
    },
    {
      "entropy": 5.450182771682739,
      "epoch": 2.661919448476052,
      "grad_norm": 1.046875,
      "learning_rate": 0.000429286129222035,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21577067524194718,
      "num_tokens": 60889543.0,
      "step": 29345
    },
    {
      "entropy": 5.4811704635620115,
      "epoch": 2.6623730043541363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000429262576389449,
      "loss": 4.904,
      "mean_token_accuracy": 0.21057627052068711,
      "num_tokens": 60900141.0,
      "step": 29350
    },
    {
      "entropy": 5.557174205780029,
      "epoch": 2.6628265602322205,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004292390203666736,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2155096486210823,
      "num_tokens": 60910244.0,
      "step": 29355
    },
    {
      "entropy": 5.484345149993897,
      "epoch": 2.663280116110305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000429215461154196,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2236874520778656,
      "num_tokens": 60919908.0,
      "step": 29360
    },
    {
      "entropy": 5.5377521991729735,
      "epoch": 2.663733671988389,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042919189875250324,
      "loss": 4.916,
      "mean_token_accuracy": 0.21647603064775467,
      "num_tokens": 60929096.0,
      "step": 29365
    },
    {
      "entropy": 5.517966842651367,
      "epoch": 2.6641872278664733,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042916833316208284,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21016127467155457,
      "num_tokens": 60939346.0,
      "step": 29370
    },
    {
      "entropy": 5.524883985519409,
      "epoch": 2.6646407837445576,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004291447643834219,
      "loss": 4.927,
      "mean_token_accuracy": 0.20598089843988418,
      "num_tokens": 60950364.0,
      "step": 29375
    },
    {
      "entropy": 5.523609447479248,
      "epoch": 2.6650943396226414,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000429121192417008,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22504774183034898,
      "num_tokens": 60960905.0,
      "step": 29380
    },
    {
      "entropy": 5.568932151794433,
      "epoch": 2.6655478955007257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042909761726332864,
      "loss": 4.9353,
      "mean_token_accuracy": 0.2208971157670021,
      "num_tokens": 60971281.0,
      "step": 29385
    },
    {
      "entropy": 5.504688310623169,
      "epoch": 2.66600145137881,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004290740389228712,
      "loss": 4.8912,
      "mean_token_accuracy": 0.21884311884641647,
      "num_tokens": 60982522.0,
      "step": 29390
    },
    {
      "entropy": 5.47869553565979,
      "epoch": 2.6664550072568938,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042905045739612347,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21246335953474044,
      "num_tokens": 60992738.0,
      "step": 29395
    },
    {
      "entropy": 5.577544403076172,
      "epoch": 2.666908563134978,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042902687268357297,
      "loss": 5.0463,
      "mean_token_accuracy": 0.21005794554948806,
      "num_tokens": 61003665.0,
      "step": 29400
    },
    {
      "entropy": 5.6628200054168705,
      "epoch": 2.6673621190130623,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042900328478570745,
      "loss": 5.0274,
      "mean_token_accuracy": 0.2140689492225647,
      "num_tokens": 61014663.0,
      "step": 29405
    },
    {
      "entropy": 5.556534337997436,
      "epoch": 2.6678156748911466,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004289796937030148,
      "loss": 4.8529,
      "mean_token_accuracy": 0.21854888051748275,
      "num_tokens": 61024916.0,
      "step": 29410
    },
    {
      "entropy": 5.451074457168579,
      "epoch": 2.668269230769231,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042895609943598286,
      "loss": 4.8574,
      "mean_token_accuracy": 0.21569318622350692,
      "num_tokens": 61035043.0,
      "step": 29415
    },
    {
      "entropy": 5.4901340961456295,
      "epoch": 2.668722786647315,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004289325019850995,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21336960047483444,
      "num_tokens": 61046153.0,
      "step": 29420
    },
    {
      "entropy": 5.576411485671997,
      "epoch": 2.6691763425253994,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004289089013508528,
      "loss": 4.8944,
      "mean_token_accuracy": 0.2169823244214058,
      "num_tokens": 61057009.0,
      "step": 29425
    },
    {
      "entropy": 5.614254379272461,
      "epoch": 2.669629898403483,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004288852975337308,
      "loss": 5.0404,
      "mean_token_accuracy": 0.2043972358107567,
      "num_tokens": 61068514.0,
      "step": 29430
    },
    {
      "entropy": 5.562333679199218,
      "epoch": 2.6700834542815675,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042886169053422165,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21562980115413666,
      "num_tokens": 61078902.0,
      "step": 29435
    },
    {
      "entropy": 5.517271566390991,
      "epoch": 2.6705370101596517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004288380803528135,
      "loss": 4.9556,
      "mean_token_accuracy": 0.21599119454622268,
      "num_tokens": 61089216.0,
      "step": 29440
    },
    {
      "entropy": 5.464359855651855,
      "epoch": 2.670990566037736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042881446698999465,
      "loss": 4.8516,
      "mean_token_accuracy": 0.2214545950293541,
      "num_tokens": 61098470.0,
      "step": 29445
    },
    {
      "entropy": 5.635874605178833,
      "epoch": 2.67144412191582,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004287908504462534,
      "loss": 5.0225,
      "mean_token_accuracy": 0.22053878307342528,
      "num_tokens": 61108694.0,
      "step": 29450
    },
    {
      "entropy": 5.4864819049835205,
      "epoch": 2.671897677793904,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042876723072207825,
      "loss": 4.9128,
      "mean_token_accuracy": 0.21818741112947465,
      "num_tokens": 61119389.0,
      "step": 29455
    },
    {
      "entropy": 5.5896800518035885,
      "epoch": 2.6723512336719883,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042874360781795757,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20870788842439653,
      "num_tokens": 61130491.0,
      "step": 29460
    },
    {
      "entropy": 5.5717939853668215,
      "epoch": 2.6728047895500726,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042871998173437987,
      "loss": 4.9344,
      "mean_token_accuracy": 0.218288154900074,
      "num_tokens": 61140404.0,
      "step": 29465
    },
    {
      "entropy": 5.59447603225708,
      "epoch": 2.673258345428157,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042869635247183373,
      "loss": 4.9971,
      "mean_token_accuracy": 0.21020450294017792,
      "num_tokens": 61150777.0,
      "step": 29470
    },
    {
      "entropy": 5.499708938598633,
      "epoch": 2.673711901306241,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042867272003080793,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21499894857406615,
      "num_tokens": 61160685.0,
      "step": 29475
    },
    {
      "entropy": 5.554017543792725,
      "epoch": 2.674165457184325,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004286490844117911,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21183238476514815,
      "num_tokens": 61170553.0,
      "step": 29480
    },
    {
      "entropy": 5.596896743774414,
      "epoch": 2.6746190130624092,
      "grad_norm": 1.125,
      "learning_rate": 0.00042862544561527203,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21244041323661805,
      "num_tokens": 61181396.0,
      "step": 29485
    },
    {
      "entropy": 5.547948932647705,
      "epoch": 2.6750725689404935,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004286018036417396,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2158227324485779,
      "num_tokens": 61192315.0,
      "step": 29490
    },
    {
      "entropy": 5.5333428382873535,
      "epoch": 2.6755261248185778,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042857815849168277,
      "loss": 4.9123,
      "mean_token_accuracy": 0.2198034092783928,
      "num_tokens": 61202719.0,
      "step": 29495
    },
    {
      "entropy": 5.429728698730469,
      "epoch": 2.6759796806966616,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004285545101655904,
      "loss": 4.8207,
      "mean_token_accuracy": 0.22307713627815245,
      "num_tokens": 61212703.0,
      "step": 29500
    },
    {
      "entropy": 5.4103169441223145,
      "epoch": 2.676433236574746,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004285308586639517,
      "loss": 4.7825,
      "mean_token_accuracy": 0.2283257484436035,
      "num_tokens": 61222406.0,
      "step": 29505
    },
    {
      "entropy": 5.518270969390869,
      "epoch": 2.67688679245283,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042850720398725575,
      "loss": 4.995,
      "mean_token_accuracy": 0.2073739618062973,
      "num_tokens": 61232758.0,
      "step": 29510
    },
    {
      "entropy": 5.518177556991577,
      "epoch": 2.6773403483309144,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042848354613599164,
      "loss": 4.8975,
      "mean_token_accuracy": 0.21801682710647582,
      "num_tokens": 61243715.0,
      "step": 29515
    },
    {
      "entropy": 5.573863220214844,
      "epoch": 2.6777939042089987,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042845988511064863,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2211415484547615,
      "num_tokens": 61254669.0,
      "step": 29520
    },
    {
      "entropy": 5.512311601638794,
      "epoch": 2.678247460087083,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042843622091171625,
      "loss": 4.9131,
      "mean_token_accuracy": 0.21606892943382264,
      "num_tokens": 61264377.0,
      "step": 29525
    },
    {
      "entropy": 5.44507794380188,
      "epoch": 2.6787010159651667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004284125535396836,
      "loss": 4.9129,
      "mean_token_accuracy": 0.21446050703525543,
      "num_tokens": 61275348.0,
      "step": 29530
    },
    {
      "entropy": 5.472241497039795,
      "epoch": 2.679154571843251,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042838888299504024,
      "loss": 4.8036,
      "mean_token_accuracy": 0.2257521390914917,
      "num_tokens": 61285579.0,
      "step": 29535
    },
    {
      "entropy": 5.507660722732544,
      "epoch": 2.6796081277213353,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042836520927827575,
      "loss": 4.8648,
      "mean_token_accuracy": 0.21811010986566542,
      "num_tokens": 61296037.0,
      "step": 29540
    },
    {
      "entropy": 5.498992300033569,
      "epoch": 2.6800616835994195,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004283415323898797,
      "loss": 4.8609,
      "mean_token_accuracy": 0.22554197311401367,
      "num_tokens": 61305882.0,
      "step": 29545
    },
    {
      "entropy": 5.4819655418396,
      "epoch": 2.6805152394775034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042831785233034156,
      "loss": 4.8475,
      "mean_token_accuracy": 0.2201354831457138,
      "num_tokens": 61317117.0,
      "step": 29550
    },
    {
      "entropy": 5.56939115524292,
      "epoch": 2.6809687953555876,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042829416910015116,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21144878417253493,
      "num_tokens": 61328190.0,
      "step": 29555
    },
    {
      "entropy": 5.476047325134277,
      "epoch": 2.681422351233672,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042827048269979834,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21556256860494613,
      "num_tokens": 61337978.0,
      "step": 29560
    },
    {
      "entropy": 5.57666425704956,
      "epoch": 2.681875907111756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004282467931297729,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21204071193933488,
      "num_tokens": 61348523.0,
      "step": 29565
    },
    {
      "entropy": 5.495347881317139,
      "epoch": 2.6823294629898404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004282231003905647,
      "loss": 4.8087,
      "mean_token_accuracy": 0.22539813816547394,
      "num_tokens": 61359250.0,
      "step": 29570
    },
    {
      "entropy": 5.457917070388794,
      "epoch": 2.6827830188679247,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042819940448266376,
      "loss": 4.8283,
      "mean_token_accuracy": 0.22017841637134553,
      "num_tokens": 61369999.0,
      "step": 29575
    },
    {
      "entropy": 5.571237277984619,
      "epoch": 2.683236574746009,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042817570540656003,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2068251058459282,
      "num_tokens": 61380279.0,
      "step": 29580
    },
    {
      "entropy": 5.486316204071045,
      "epoch": 2.683690130624093,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004281520031627438,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22008431553840638,
      "num_tokens": 61389794.0,
      "step": 29585
    },
    {
      "entropy": 5.570743894577026,
      "epoch": 2.684143686502177,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042812829775170497,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2086791440844536,
      "num_tokens": 61401136.0,
      "step": 29590
    },
    {
      "entropy": 5.533440780639649,
      "epoch": 2.6845972423802613,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042810458917393404,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2123529940843582,
      "num_tokens": 61412830.0,
      "step": 29595
    },
    {
      "entropy": 5.526773595809937,
      "epoch": 2.6850507982583456,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004280808774299211,
      "loss": 4.8528,
      "mean_token_accuracy": 0.22239815294742585,
      "num_tokens": 61422168.0,
      "step": 29600
    },
    {
      "entropy": 5.497328233718872,
      "epoch": 2.6855043541364294,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042805716252015665,
      "loss": 4.8812,
      "mean_token_accuracy": 0.21743890643119812,
      "num_tokens": 61432471.0,
      "step": 29605
    },
    {
      "entropy": 5.522314262390137,
      "epoch": 2.6859579100145137,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042803344444513106,
      "loss": 4.9081,
      "mean_token_accuracy": 0.21788983941078185,
      "num_tokens": 61443402.0,
      "step": 29610
    },
    {
      "entropy": 5.573214197158814,
      "epoch": 2.686411465892598,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004280097232053349,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21274835020303726,
      "num_tokens": 61454465.0,
      "step": 29615
    },
    {
      "entropy": 5.5291448593139645,
      "epoch": 2.686865021770682,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004279859988012586,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21383485645055772,
      "num_tokens": 61464671.0,
      "step": 29620
    },
    {
      "entropy": 5.489496088027954,
      "epoch": 2.6873185776487665,
      "grad_norm": 1.125,
      "learning_rate": 0.0004279622712333929,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22458026856184005,
      "num_tokens": 61474940.0,
      "step": 29625
    },
    {
      "entropy": 5.547333765029907,
      "epoch": 2.6877721335268507,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004279385405022284,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21095149964094162,
      "num_tokens": 61485240.0,
      "step": 29630
    },
    {
      "entropy": 5.52120008468628,
      "epoch": 2.6882256894049346,
      "grad_norm": 1.109375,
      "learning_rate": 0.000427914806608256,
      "loss": 4.8878,
      "mean_token_accuracy": 0.21645450443029404,
      "num_tokens": 61495382.0,
      "step": 29635
    },
    {
      "entropy": 5.500570821762085,
      "epoch": 2.688679245283019,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004278910695519663,
      "loss": 4.8472,
      "mean_token_accuracy": 0.22481829822063445,
      "num_tokens": 61505372.0,
      "step": 29640
    },
    {
      "entropy": 5.522217464447022,
      "epoch": 2.689132801161103,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042786732933385046,
      "loss": 4.9048,
      "mean_token_accuracy": 0.2189964845776558,
      "num_tokens": 61516039.0,
      "step": 29645
    },
    {
      "entropy": 5.5330604076385494,
      "epoch": 2.6895863570391874,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004278435859543992,
      "loss": 4.9809,
      "mean_token_accuracy": 0.21166449785232544,
      "num_tokens": 61525435.0,
      "step": 29650
    },
    {
      "entropy": 5.574708366394043,
      "epoch": 2.690039912917271,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004278198394141037,
      "loss": 4.9164,
      "mean_token_accuracy": 0.20987374633550643,
      "num_tokens": 61534714.0,
      "step": 29655
    },
    {
      "entropy": 5.593967437744141,
      "epoch": 2.6904934687953554,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042779608971345496,
      "loss": 5.0514,
      "mean_token_accuracy": 0.20717464089393617,
      "num_tokens": 61545841.0,
      "step": 29660
    },
    {
      "entropy": 5.476044416427612,
      "epoch": 2.6909470246734397,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042777233685294407,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22667205035686494,
      "num_tokens": 61555355.0,
      "step": 29665
    },
    {
      "entropy": 5.532976293563843,
      "epoch": 2.691400580551524,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042774858083306246,
      "loss": 4.9394,
      "mean_token_accuracy": 0.21693466305732728,
      "num_tokens": 61565474.0,
      "step": 29670
    },
    {
      "entropy": 5.542480993270874,
      "epoch": 2.6918541364296082,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042772482165430117,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22625344693660737,
      "num_tokens": 61575631.0,
      "step": 29675
    },
    {
      "entropy": 5.493725061416626,
      "epoch": 2.6923076923076925,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042770105931715167,
      "loss": 4.9065,
      "mean_token_accuracy": 0.216996268928051,
      "num_tokens": 61586398.0,
      "step": 29680
    },
    {
      "entropy": 5.501855659484863,
      "epoch": 2.6927612481857763,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004276772938221054,
      "loss": 4.8021,
      "mean_token_accuracy": 0.22439393103122712,
      "num_tokens": 61596630.0,
      "step": 29685
    },
    {
      "entropy": 5.536890649795533,
      "epoch": 2.6932148040638606,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004276535251696537,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21917167603969573,
      "num_tokens": 61605965.0,
      "step": 29690
    },
    {
      "entropy": 5.5594219207763675,
      "epoch": 2.693668359941945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004276297533602883,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21620827466249465,
      "num_tokens": 61616973.0,
      "step": 29695
    },
    {
      "entropy": 5.590553903579712,
      "epoch": 2.694121915820029,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042760597839450057,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21248731911182403,
      "num_tokens": 61627420.0,
      "step": 29700
    },
    {
      "entropy": 5.542947006225586,
      "epoch": 2.694575471698113,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004275822002727824,
      "loss": 4.9014,
      "mean_token_accuracy": 0.2199798196554184,
      "num_tokens": 61638315.0,
      "step": 29705
    },
    {
      "entropy": 5.486486625671387,
      "epoch": 2.6950290275761972,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042755841899562544,
      "loss": 4.919,
      "mean_token_accuracy": 0.2210564434528351,
      "num_tokens": 61649294.0,
      "step": 29710
    },
    {
      "entropy": 5.570138597488404,
      "epoch": 2.6954825834542815,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042753463456352155,
      "loss": 4.9577,
      "mean_token_accuracy": 0.20820011645555497,
      "num_tokens": 61660312.0,
      "step": 29715
    },
    {
      "entropy": 5.529924964904785,
      "epoch": 2.6959361393323658,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004275108469769625,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21851568520069123,
      "num_tokens": 61670539.0,
      "step": 29720
    },
    {
      "entropy": 5.526054048538208,
      "epoch": 2.69638969521045,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042748705623644025,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22114203870296478,
      "num_tokens": 61679613.0,
      "step": 29725
    },
    {
      "entropy": 5.582476472854614,
      "epoch": 2.6968432510885343,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004274632623424468,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21281811445951462,
      "num_tokens": 61689714.0,
      "step": 29730
    },
    {
      "entropy": 5.610582447052002,
      "epoch": 2.6972968069666186,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042743946529547435,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21272632777690886,
      "num_tokens": 61700701.0,
      "step": 29735
    },
    {
      "entropy": 5.568483972549439,
      "epoch": 2.6977503628447024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004274156650960148,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21539850383996964,
      "num_tokens": 61711403.0,
      "step": 29740
    },
    {
      "entropy": 5.4400615215301515,
      "epoch": 2.6982039187227866,
      "grad_norm": 1.125,
      "learning_rate": 0.0004273918617445605,
      "loss": 4.7753,
      "mean_token_accuracy": 0.21845744401216508,
      "num_tokens": 61721787.0,
      "step": 29745
    },
    {
      "entropy": 5.622704935073853,
      "epoch": 2.698657474600871,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042736805524160363,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21351802498102188,
      "num_tokens": 61732702.0,
      "step": 29750
    },
    {
      "entropy": 5.507934045791626,
      "epoch": 2.6991110304789547,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042734424558763656,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21935366243124008,
      "num_tokens": 61743477.0,
      "step": 29755
    },
    {
      "entropy": 5.561422109603882,
      "epoch": 2.699564586357039,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004273204327831516,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21048598736524582,
      "num_tokens": 61753620.0,
      "step": 29760
    },
    {
      "entropy": 5.554521656036377,
      "epoch": 2.7000181422351233,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004272966168286414,
      "loss": 4.8618,
      "mean_token_accuracy": 0.22496049851179123,
      "num_tokens": 61763678.0,
      "step": 29765
    },
    {
      "entropy": 5.558667373657227,
      "epoch": 2.7004716981132075,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042727279772459837,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22454155385494232,
      "num_tokens": 61774419.0,
      "step": 29770
    },
    {
      "entropy": 5.567608070373535,
      "epoch": 2.700925253991292,
      "grad_norm": 1.21875,
      "learning_rate": 0.000427248975471515,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20972986817359923,
      "num_tokens": 61785546.0,
      "step": 29775
    },
    {
      "entropy": 5.5971606254577635,
      "epoch": 2.701378809869376,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042722515006988397,
      "loss": 4.8781,
      "mean_token_accuracy": 0.21897144764661788,
      "num_tokens": 61795846.0,
      "step": 29780
    },
    {
      "entropy": 5.560446834564209,
      "epoch": 2.7018323657474603,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004272013215201982,
      "loss": 4.9513,
      "mean_token_accuracy": 0.21278668493032454,
      "num_tokens": 61806440.0,
      "step": 29785
    },
    {
      "entropy": 5.4758460521698,
      "epoch": 2.702285921625544,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042717748982295015,
      "loss": 4.9121,
      "mean_token_accuracy": 0.21631584465503692,
      "num_tokens": 61815371.0,
      "step": 29790
    },
    {
      "entropy": 5.554646062850952,
      "epoch": 2.7027394775036284,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004271536549786329,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21848005950450897,
      "num_tokens": 61825225.0,
      "step": 29795
    },
    {
      "entropy": 5.545880651473999,
      "epoch": 2.7031930333817127,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004271298169877393,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21982549279928207,
      "num_tokens": 61836705.0,
      "step": 29800
    },
    {
      "entropy": 5.5258073806762695,
      "epoch": 2.703646589259797,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004271059758507624,
      "loss": 4.8926,
      "mean_token_accuracy": 0.21772266626358033,
      "num_tokens": 61847082.0,
      "step": 29805
    },
    {
      "entropy": 5.519984912872315,
      "epoch": 2.7041001451378808,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042708213156819505,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21363881081342698,
      "num_tokens": 61857130.0,
      "step": 29810
    },
    {
      "entropy": 5.591157913208008,
      "epoch": 2.704553701015965,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004270582841405305,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21570713967084884,
      "num_tokens": 61866796.0,
      "step": 29815
    },
    {
      "entropy": 5.580596446990967,
      "epoch": 2.7050072568940493,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004270344335682618,
      "loss": 4.9438,
      "mean_token_accuracy": 0.20958711504936217,
      "num_tokens": 61876356.0,
      "step": 29820
    },
    {
      "entropy": 5.5217150211334225,
      "epoch": 2.7054608127721336,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042701057985188235,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20681913793087006,
      "num_tokens": 61887374.0,
      "step": 29825
    },
    {
      "entropy": 5.577119207382202,
      "epoch": 2.705914368650218,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004269867229918854,
      "loss": 4.9864,
      "mean_token_accuracy": 0.21167127788066864,
      "num_tokens": 61897376.0,
      "step": 29830
    },
    {
      "entropy": 5.587528562545776,
      "epoch": 2.706367924528302,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042696286298876424,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20952281653881072,
      "num_tokens": 61908212.0,
      "step": 29835
    },
    {
      "entropy": 5.569792318344116,
      "epoch": 2.706821480406386,
      "grad_norm": 1.125,
      "learning_rate": 0.0004269389998430124,
      "loss": 5.0325,
      "mean_token_accuracy": 0.19804124683141708,
      "num_tokens": 61920699.0,
      "step": 29840
    },
    {
      "entropy": 5.557897853851318,
      "epoch": 2.70727503628447,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004269151335551233,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2117043063044548,
      "num_tokens": 61931556.0,
      "step": 29845
    },
    {
      "entropy": 5.653575086593628,
      "epoch": 2.7077285921625545,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004268912641255905,
      "loss": 5.0665,
      "mean_token_accuracy": 0.21260046660900117,
      "num_tokens": 61941483.0,
      "step": 29850
    },
    {
      "entropy": 5.564683485031128,
      "epoch": 2.7081821480406387,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042686739155490763,
      "loss": 4.8947,
      "mean_token_accuracy": 0.21681239902973176,
      "num_tokens": 61951262.0,
      "step": 29855
    },
    {
      "entropy": 5.5906154155731205,
      "epoch": 2.7086357039187225,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004268435158435684,
      "loss": 4.978,
      "mean_token_accuracy": 0.2153024196624756,
      "num_tokens": 61962423.0,
      "step": 29860
    },
    {
      "entropy": 5.631505870819092,
      "epoch": 2.709089259796807,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042681963699206667,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2059436097741127,
      "num_tokens": 61973276.0,
      "step": 29865
    },
    {
      "entropy": 5.515093469619751,
      "epoch": 2.709542815674891,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042679575500089605,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2180198147892952,
      "num_tokens": 61984369.0,
      "step": 29870
    },
    {
      "entropy": 5.506880426406861,
      "epoch": 2.7099963715529753,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004267718698705506,
      "loss": 4.8558,
      "mean_token_accuracy": 0.21804390102624893,
      "num_tokens": 61994224.0,
      "step": 29875
    },
    {
      "entropy": 5.507638835906983,
      "epoch": 2.7104499274310596,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004267479816015242,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2174692288041115,
      "num_tokens": 62004546.0,
      "step": 29880
    },
    {
      "entropy": 5.556619548797608,
      "epoch": 2.710903483309144,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004267240901943109,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2190815657377243,
      "num_tokens": 62013924.0,
      "step": 29885
    },
    {
      "entropy": 5.539492416381836,
      "epoch": 2.7113570391872277,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042670019564940474,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2149542674422264,
      "num_tokens": 62026150.0,
      "step": 29890
    },
    {
      "entropy": 5.576632261276245,
      "epoch": 2.711810595065312,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004266762979672998,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2127609819173813,
      "num_tokens": 62036203.0,
      "step": 29895
    },
    {
      "entropy": 5.55012149810791,
      "epoch": 2.7122641509433962,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042665239714849044,
      "loss": 4.9283,
      "mean_token_accuracy": 0.21250315010547638,
      "num_tokens": 62046961.0,
      "step": 29900
    },
    {
      "entropy": 5.476844024658203,
      "epoch": 2.7127177068214805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042662849319347073,
      "loss": 4.8786,
      "mean_token_accuracy": 0.21491149961948394,
      "num_tokens": 62057365.0,
      "step": 29905
    },
    {
      "entropy": 5.493049144744873,
      "epoch": 2.7131712626995643,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042660458610273527,
      "loss": 4.888,
      "mean_token_accuracy": 0.21077376008033752,
      "num_tokens": 62067934.0,
      "step": 29910
    },
    {
      "entropy": 5.611263656616211,
      "epoch": 2.7136248185776486,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004265806758767783,
      "loss": 4.9911,
      "mean_token_accuracy": 0.20873421281576157,
      "num_tokens": 62077203.0,
      "step": 29915
    },
    {
      "entropy": 5.59975037574768,
      "epoch": 2.714078374455733,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042655676251609435,
      "loss": 4.924,
      "mean_token_accuracy": 0.21966552436351777,
      "num_tokens": 62087262.0,
      "step": 29920
    },
    {
      "entropy": 5.53824143409729,
      "epoch": 2.714531930333817,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042653284602117795,
      "loss": 4.9138,
      "mean_token_accuracy": 0.21824792325496672,
      "num_tokens": 62097931.0,
      "step": 29925
    },
    {
      "entropy": 5.53220419883728,
      "epoch": 2.7149854862119014,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004265089263925236,
      "loss": 4.9908,
      "mean_token_accuracy": 0.21161436885595322,
      "num_tokens": 62109478.0,
      "step": 29930
    },
    {
      "entropy": 5.589300346374512,
      "epoch": 2.7154390420899857,
      "grad_norm": 1.078125,
      "learning_rate": 0.000426485003630626,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21019331067800523,
      "num_tokens": 62120735.0,
      "step": 29935
    },
    {
      "entropy": 5.574778127670288,
      "epoch": 2.71589259796807,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004264610777359801,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2108905553817749,
      "num_tokens": 62130950.0,
      "step": 29940
    },
    {
      "entropy": 5.575260782241822,
      "epoch": 2.7163461538461537,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004264371487090804,
      "loss": 4.9283,
      "mean_token_accuracy": 0.21393319964408875,
      "num_tokens": 62141351.0,
      "step": 29945
    },
    {
      "entropy": 5.44889874458313,
      "epoch": 2.716799709724238,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042641321655042185,
      "loss": 4.8725,
      "mean_token_accuracy": 0.21684699207544328,
      "num_tokens": 62151388.0,
      "step": 29950
    },
    {
      "entropy": 5.568221807479858,
      "epoch": 2.7172532656023223,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004263892812604995,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20715204775333404,
      "num_tokens": 62162809.0,
      "step": 29955
    },
    {
      "entropy": 5.513254261016845,
      "epoch": 2.7177068214804065,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042636534283980814,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21289895176887513,
      "num_tokens": 62172738.0,
      "step": 29960
    },
    {
      "entropy": 5.632068824768067,
      "epoch": 2.7181603773584904,
      "grad_norm": 1.171875,
      "learning_rate": 0.000426341401288843,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20962415337562562,
      "num_tokens": 62182508.0,
      "step": 29965
    },
    {
      "entropy": 5.57133412361145,
      "epoch": 2.7186139332365746,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042631745660809905,
      "loss": 5.0319,
      "mean_token_accuracy": 0.20589052587747575,
      "num_tokens": 62192873.0,
      "step": 29970
    },
    {
      "entropy": 5.5889201164245605,
      "epoch": 2.719067489114659,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042629350879807155,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2085902824997902,
      "num_tokens": 62204727.0,
      "step": 29975
    },
    {
      "entropy": 5.56460976600647,
      "epoch": 2.719521044992743,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042626955785925576,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21326752454042436,
      "num_tokens": 62214268.0,
      "step": 29980
    },
    {
      "entropy": 5.594280147552491,
      "epoch": 2.7199746008708274,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042624560379214693,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20895089358091354,
      "num_tokens": 62224756.0,
      "step": 29985
    },
    {
      "entropy": 5.602207279205322,
      "epoch": 2.7204281567489117,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004262216465972405,
      "loss": 4.9199,
      "mean_token_accuracy": 0.2133926346898079,
      "num_tokens": 62234065.0,
      "step": 29990
    },
    {
      "entropy": 5.558912038803101,
      "epoch": 2.7208817126269955,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004261976862750318,
      "loss": 4.929,
      "mean_token_accuracy": 0.22078927904367446,
      "num_tokens": 62243732.0,
      "step": 29995
    },
    {
      "entropy": 5.564157962799072,
      "epoch": 2.72133526850508,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004261737228260165,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2240482360124588,
      "num_tokens": 62253461.0,
      "step": 30000
    },
    {
      "epoch": 2.72133526850508,
      "eval_entropy": 5.240016961205599,
      "eval_loss": 4.978549957275391,
      "eval_mean_token_accuracy": 0.22184979436219548,
      "eval_num_tokens": 62253461.0,
      "eval_runtime": 20.303,
      "eval_samples_per_second": 1741.615,
      "eval_steps_per_second": 217.702,
      "step": 30000
    },
    {
      "entropy": 5.566419315338135,
      "epoch": 2.721788824383164,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004261497562506901,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21412499994039536,
      "num_tokens": 62263830.0,
      "step": 30005
    },
    {
      "entropy": 5.478595304489136,
      "epoch": 2.7222423802612483,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004261257865495483,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2191891446709633,
      "num_tokens": 62274383.0,
      "step": 30010
    },
    {
      "entropy": 5.643680810928345,
      "epoch": 2.722695936139332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004261018137230865,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19756740033626558,
      "num_tokens": 62286476.0,
      "step": 30015
    },
    {
      "entropy": 5.665280103683472,
      "epoch": 2.7231494920174164,
      "grad_norm": 1.125,
      "learning_rate": 0.00042607783777180085,
      "loss": 4.9884,
      "mean_token_accuracy": 0.2095344066619873,
      "num_tokens": 62296192.0,
      "step": 30020
    },
    {
      "entropy": 5.520650100708008,
      "epoch": 2.7236030478955007,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042605385869618693,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22464792877435685,
      "num_tokens": 62305300.0,
      "step": 30025
    },
    {
      "entropy": 5.546018743515015,
      "epoch": 2.724056603773585,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004260298764967408,
      "loss": 4.903,
      "mean_token_accuracy": 0.21652857214212418,
      "num_tokens": 62316802.0,
      "step": 30030
    },
    {
      "entropy": 5.467512941360473,
      "epoch": 2.724510159651669,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004260058911739583,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21453212946653366,
      "num_tokens": 62326444.0,
      "step": 30035
    },
    {
      "entropy": 5.532042646408081,
      "epoch": 2.7249637155297535,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042598190272833547,
      "loss": 4.8548,
      "mean_token_accuracy": 0.21847527623176574,
      "num_tokens": 62336872.0,
      "step": 30040
    },
    {
      "entropy": 5.478587579727173,
      "epoch": 2.7254172714078373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004259579111603685,
      "loss": 4.8263,
      "mean_token_accuracy": 0.21862645894289018,
      "num_tokens": 62346771.0,
      "step": 30045
    },
    {
      "entropy": 5.543378257751465,
      "epoch": 2.7258708272859216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042593391647055334,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21706598103046418,
      "num_tokens": 62357503.0,
      "step": 30050
    },
    {
      "entropy": 5.442430257797241,
      "epoch": 2.726324383164006,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004259099186593864,
      "loss": 4.8347,
      "mean_token_accuracy": 0.22069754898548127,
      "num_tokens": 62367664.0,
      "step": 30055
    },
    {
      "entropy": 5.570179510116577,
      "epoch": 2.72677793904209,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004258859177273639,
      "loss": 4.9074,
      "mean_token_accuracy": 0.21304928362369538,
      "num_tokens": 62376924.0,
      "step": 30060
    },
    {
      "entropy": 5.5092486381530765,
      "epoch": 2.727231494920174,
      "grad_norm": 1.125,
      "learning_rate": 0.0004258619136749821,
      "loss": 4.825,
      "mean_token_accuracy": 0.22234923988580704,
      "num_tokens": 62387529.0,
      "step": 30065
    },
    {
      "entropy": 5.494977331161499,
      "epoch": 2.727685050798258,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004258379065027375,
      "loss": 4.8933,
      "mean_token_accuracy": 0.2205989897251129,
      "num_tokens": 62397833.0,
      "step": 30070
    },
    {
      "entropy": 5.570089197158813,
      "epoch": 2.7281386066763424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004258138962111266,
      "loss": 4.9613,
      "mean_token_accuracy": 0.20982608199119568,
      "num_tokens": 62409022.0,
      "step": 30075
    },
    {
      "entropy": 5.552255058288575,
      "epoch": 2.7285921625544267,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004257898828006459,
      "loss": 4.9545,
      "mean_token_accuracy": 0.2039222925901413,
      "num_tokens": 62419105.0,
      "step": 30080
    },
    {
      "entropy": 5.5453815937042235,
      "epoch": 2.729045718432511,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004257658662717919,
      "loss": 4.924,
      "mean_token_accuracy": 0.22040488123893737,
      "num_tokens": 62428371.0,
      "step": 30085
    },
    {
      "entropy": 5.533915662765503,
      "epoch": 2.7294992743105952,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004257418466250615,
      "loss": 4.8964,
      "mean_token_accuracy": 0.21438532173633576,
      "num_tokens": 62439301.0,
      "step": 30090
    },
    {
      "entropy": 5.5338358879089355,
      "epoch": 2.7299528301886795,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042571782386095124,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21243853271007537,
      "num_tokens": 62449450.0,
      "step": 30095
    },
    {
      "entropy": 5.5330808639526365,
      "epoch": 2.7304063860667633,
      "grad_norm": 1.421875,
      "learning_rate": 0.00042569379797995793,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21238136291503906,
      "num_tokens": 62461051.0,
      "step": 30100
    },
    {
      "entropy": 5.5157959938049315,
      "epoch": 2.7308599419448476,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004256697689825785,
      "loss": 4.8721,
      "mean_token_accuracy": 0.21956984251737593,
      "num_tokens": 62471755.0,
      "step": 30105
    },
    {
      "entropy": 5.516527462005615,
      "epoch": 2.731313497822932,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042564573686930987,
      "loss": 4.8948,
      "mean_token_accuracy": 0.217100889980793,
      "num_tokens": 62482060.0,
      "step": 30110
    },
    {
      "entropy": 5.555571222305298,
      "epoch": 2.7317670537010157,
      "grad_norm": 1.09375,
      "learning_rate": 0.000425621701640649,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21580011546611785,
      "num_tokens": 62493735.0,
      "step": 30115
    },
    {
      "entropy": 5.553857326507568,
      "epoch": 2.7322206095791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000425597663297093,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21320728808641434,
      "num_tokens": 62504582.0,
      "step": 30120
    },
    {
      "entropy": 5.464041900634766,
      "epoch": 2.7326741654571842,
      "grad_norm": 1.125,
      "learning_rate": 0.0004255736218391389,
      "loss": 4.886,
      "mean_token_accuracy": 0.21035362780094147,
      "num_tokens": 62516665.0,
      "step": 30125
    },
    {
      "entropy": 5.546331739425659,
      "epoch": 2.7331277213352685,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042554957726728394,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21485030353069307,
      "num_tokens": 62527745.0,
      "step": 30130
    },
    {
      "entropy": 5.546920204162598,
      "epoch": 2.7335812772133528,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004255255295820253,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21797107309103012,
      "num_tokens": 62538622.0,
      "step": 30135
    },
    {
      "entropy": 5.524048519134522,
      "epoch": 2.734034833091437,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042550147878386044,
      "loss": 4.9073,
      "mean_token_accuracy": 0.217749984562397,
      "num_tokens": 62548302.0,
      "step": 30140
    },
    {
      "entropy": 5.467053365707398,
      "epoch": 2.7344883889695213,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004254774248732867,
      "loss": 4.8315,
      "mean_token_accuracy": 0.2245108738541603,
      "num_tokens": 62557150.0,
      "step": 30145
    },
    {
      "entropy": 5.684099578857422,
      "epoch": 2.734941944847605,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042545336785080126,
      "loss": 5.1213,
      "mean_token_accuracy": 0.20072893053293228,
      "num_tokens": 62568291.0,
      "step": 30150
    },
    {
      "entropy": 5.550385618209839,
      "epoch": 2.7353955007256894,
      "grad_norm": 1.34375,
      "learning_rate": 0.000425429307716902,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21782101988792418,
      "num_tokens": 62579351.0,
      "step": 30155
    },
    {
      "entropy": 5.454301166534424,
      "epoch": 2.7358490566037736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042540524447208625,
      "loss": 4.7901,
      "mean_token_accuracy": 0.22473580837249757,
      "num_tokens": 62588614.0,
      "step": 30160
    },
    {
      "entropy": 5.551528406143189,
      "epoch": 2.736302612481858,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042538117811685174,
      "loss": 4.9173,
      "mean_token_accuracy": 0.21639949083328247,
      "num_tokens": 62598844.0,
      "step": 30165
    },
    {
      "entropy": 5.562780809402466,
      "epoch": 2.7367561683599417,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042535710865169616,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20663230270147323,
      "num_tokens": 62608722.0,
      "step": 30170
    },
    {
      "entropy": 5.493654251098633,
      "epoch": 2.737209724238026,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004253330360771173,
      "loss": 4.8539,
      "mean_token_accuracy": 0.21182840168476105,
      "num_tokens": 62618712.0,
      "step": 30175
    },
    {
      "entropy": 5.565095281600952,
      "epoch": 2.7376632801161103,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042530896039361284,
      "loss": 5.0466,
      "mean_token_accuracy": 0.20055664628744124,
      "num_tokens": 62629329.0,
      "step": 30180
    },
    {
      "entropy": 5.5933506965637205,
      "epoch": 2.7381168359941945,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042528488160168077,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20959201008081435,
      "num_tokens": 62639644.0,
      "step": 30185
    },
    {
      "entropy": 5.575213050842285,
      "epoch": 2.738570391872279,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042526079970181915,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2056760534644127,
      "num_tokens": 62650979.0,
      "step": 30190
    },
    {
      "entropy": 5.571151208877564,
      "epoch": 2.739023947750363,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004252367146945259,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20874431729316711,
      "num_tokens": 62661163.0,
      "step": 30195
    },
    {
      "entropy": 5.550344657897949,
      "epoch": 2.739477503628447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042521262658029905,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20526358485221863,
      "num_tokens": 62672371.0,
      "step": 30200
    },
    {
      "entropy": 5.553168153762817,
      "epoch": 2.739931059506531,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042518853535963673,
      "loss": 4.88,
      "mean_token_accuracy": 0.21575011759996415,
      "num_tokens": 62682521.0,
      "step": 30205
    },
    {
      "entropy": 5.525324964523316,
      "epoch": 2.7403846153846154,
      "grad_norm": 2.140625,
      "learning_rate": 0.00042516444103303736,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22690353095531463,
      "num_tokens": 62692420.0,
      "step": 30210
    },
    {
      "entropy": 5.4934889793396,
      "epoch": 2.7408381712626997,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042514034360099907,
      "loss": 4.897,
      "mean_token_accuracy": 0.21617562919855118,
      "num_tokens": 62702686.0,
      "step": 30215
    },
    {
      "entropy": 5.556487083435059,
      "epoch": 2.7412917271407835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042511624306402015,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21402305215597153,
      "num_tokens": 62713262.0,
      "step": 30220
    },
    {
      "entropy": 5.552816247940063,
      "epoch": 2.7417452830188678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004250921394225992,
      "loss": 4.8883,
      "mean_token_accuracy": 0.21969353705644606,
      "num_tokens": 62724460.0,
      "step": 30225
    },
    {
      "entropy": 5.53007698059082,
      "epoch": 2.742198838896952,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004250680326772343,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20912925153970718,
      "num_tokens": 62734190.0,
      "step": 30230
    },
    {
      "entropy": 5.541458940505981,
      "epoch": 2.7426523947750363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004250439228284244,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21260523200035095,
      "num_tokens": 62744323.0,
      "step": 30235
    },
    {
      "entropy": 5.5656445026397705,
      "epoch": 2.7431059506531206,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042501980987666807,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2125800520181656,
      "num_tokens": 62753868.0,
      "step": 30240
    },
    {
      "entropy": 5.563860511779785,
      "epoch": 2.743559506531205,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004249956938224636,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21010760962963104,
      "num_tokens": 62763954.0,
      "step": 30245
    },
    {
      "entropy": 5.526519346237182,
      "epoch": 2.7440130624092887,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042497157466631015,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22240310907363892,
      "num_tokens": 62774116.0,
      "step": 30250
    },
    {
      "entropy": 5.610444450378418,
      "epoch": 2.744466618287373,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042494745240870626,
      "loss": 4.9849,
      "mean_token_accuracy": 0.204432912170887,
      "num_tokens": 62784939.0,
      "step": 30255
    },
    {
      "entropy": 5.535037660598755,
      "epoch": 2.744920174165457,
      "grad_norm": 1.125,
      "learning_rate": 0.00042492332705015075,
      "loss": 4.9394,
      "mean_token_accuracy": 0.21138351261615754,
      "num_tokens": 62794988.0,
      "step": 30260
    },
    {
      "entropy": 5.56043815612793,
      "epoch": 2.7453737300435415,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042489919859114273,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21653874963521957,
      "num_tokens": 62804984.0,
      "step": 30265
    },
    {
      "entropy": 5.571105337142944,
      "epoch": 2.7458272859216253,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042487506703218104,
      "loss": 4.895,
      "mean_token_accuracy": 0.21171075403690337,
      "num_tokens": 62816401.0,
      "step": 30270
    },
    {
      "entropy": 5.511363458633423,
      "epoch": 2.7462808417997095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004248509323737648,
      "loss": 4.8601,
      "mean_token_accuracy": 0.21492574214935303,
      "num_tokens": 62826411.0,
      "step": 30275
    },
    {
      "entropy": 5.572367000579834,
      "epoch": 2.746734397677794,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004248267946163931,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21042945235967636,
      "num_tokens": 62837228.0,
      "step": 30280
    },
    {
      "entropy": 5.502644634246826,
      "epoch": 2.747187953555878,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042480265376056506,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22120752930641174,
      "num_tokens": 62847852.0,
      "step": 30285
    },
    {
      "entropy": 5.526437520980835,
      "epoch": 2.7476415094339623,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042477850980678,
      "loss": 4.972,
      "mean_token_accuracy": 0.20401092767715454,
      "num_tokens": 62859516.0,
      "step": 30290
    },
    {
      "entropy": 5.54732666015625,
      "epoch": 2.7480950653120466,
      "grad_norm": 1.125,
      "learning_rate": 0.0004247543627555372,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21135311722755432,
      "num_tokens": 62869757.0,
      "step": 30295
    },
    {
      "entropy": 5.539496231079101,
      "epoch": 2.748548621190131,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000424730212607336,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21248133927583696,
      "num_tokens": 62879618.0,
      "step": 30300
    },
    {
      "entropy": 5.541743183135987,
      "epoch": 2.7490021770682147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004247060593626758,
      "loss": 4.952,
      "mean_token_accuracy": 0.2070973351597786,
      "num_tokens": 62890491.0,
      "step": 30305
    },
    {
      "entropy": 5.618363523483277,
      "epoch": 2.749455732946299,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004246819030220562,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22238344252109526,
      "num_tokens": 62899954.0,
      "step": 30310
    },
    {
      "entropy": 5.5683800220489506,
      "epoch": 2.7499092888243832,
      "grad_norm": 1.125,
      "learning_rate": 0.0004246577435859767,
      "loss": 4.9654,
      "mean_token_accuracy": 0.2128121241927147,
      "num_tokens": 62911857.0,
      "step": 30315
    },
    {
      "entropy": 5.523028326034546,
      "epoch": 2.7503628447024675,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004246335810549369,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21678196787834167,
      "num_tokens": 62922023.0,
      "step": 30320
    },
    {
      "entropy": 5.562381362915039,
      "epoch": 2.7508164005805513,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042460941542943656,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2209171712398529,
      "num_tokens": 62932845.0,
      "step": 30325
    },
    {
      "entropy": 5.541231632232666,
      "epoch": 2.7512699564586356,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004245852467099753,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2103466272354126,
      "num_tokens": 62942881.0,
      "step": 30330
    },
    {
      "entropy": 5.500460386276245,
      "epoch": 2.75172351233672,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004245610748970531,
      "loss": 4.8643,
      "mean_token_accuracy": 0.21816278994083405,
      "num_tokens": 62953957.0,
      "step": 30335
    },
    {
      "entropy": 5.473409223556518,
      "epoch": 2.752177068214804,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042453689999116974,
      "loss": 4.8863,
      "mean_token_accuracy": 0.21897784769535064,
      "num_tokens": 62964803.0,
      "step": 30340
    },
    {
      "entropy": 5.517470407485962,
      "epoch": 2.7526306240928884,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004245127219928252,
      "loss": 4.8847,
      "mean_token_accuracy": 0.21829183846712114,
      "num_tokens": 62975477.0,
      "step": 30345
    },
    {
      "entropy": 5.5250951766967775,
      "epoch": 2.7530841799709727,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042448854090251944,
      "loss": 4.9413,
      "mean_token_accuracy": 0.2134728580713272,
      "num_tokens": 62985430.0,
      "step": 30350
    },
    {
      "entropy": 5.511751508712768,
      "epoch": 2.7535377358490565,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042446435672075254,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21664171665906906,
      "num_tokens": 62997493.0,
      "step": 30355
    },
    {
      "entropy": 5.600302267074585,
      "epoch": 2.7539912917271407,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00042444016944802467,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21506956368684768,
      "num_tokens": 63007573.0,
      "step": 30360
    },
    {
      "entropy": 5.571696424484253,
      "epoch": 2.754444847605225,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042441597908483603,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21918165385723115,
      "num_tokens": 63018372.0,
      "step": 30365
    },
    {
      "entropy": 5.536999750137329,
      "epoch": 2.7548984034833093,
      "grad_norm": 1.125,
      "learning_rate": 0.0004243917856316869,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21993626654148102,
      "num_tokens": 63028632.0,
      "step": 30370
    },
    {
      "entropy": 5.540465831756592,
      "epoch": 2.755351959361393,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042436758908907747,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2103507936000824,
      "num_tokens": 63038510.0,
      "step": 30375
    },
    {
      "entropy": 5.650352716445923,
      "epoch": 2.7558055152394774,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042434338945750833,
      "loss": 5.0875,
      "mean_token_accuracy": 0.19803033471107484,
      "num_tokens": 63049375.0,
      "step": 30380
    },
    {
      "entropy": 5.579443025588989,
      "epoch": 2.7562590711175616,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042431918673747974,
      "loss": 4.921,
      "mean_token_accuracy": 0.2144318550825119,
      "num_tokens": 63059422.0,
      "step": 30385
    },
    {
      "entropy": 5.586562728881836,
      "epoch": 2.756712626995646,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042429498092949245,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20266233384609222,
      "num_tokens": 63071101.0,
      "step": 30390
    },
    {
      "entropy": 5.507417058944702,
      "epoch": 2.75716618287373,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004242707720340468,
      "loss": 4.861,
      "mean_token_accuracy": 0.2204086571931839,
      "num_tokens": 63081104.0,
      "step": 30395
    },
    {
      "entropy": 5.5523826599121096,
      "epoch": 2.7576197387518144,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004242465600516436,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21363636702299119,
      "num_tokens": 63091123.0,
      "step": 30400
    },
    {
      "entropy": 5.492327117919922,
      "epoch": 2.7580732946298983,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042422234498278353,
      "loss": 4.9163,
      "mean_token_accuracy": 0.219136281311512,
      "num_tokens": 63101541.0,
      "step": 30405
    },
    {
      "entropy": 5.575545358657837,
      "epoch": 2.7585268505079825,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042419812682796727,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20531836450099944,
      "num_tokens": 63112012.0,
      "step": 30410
    },
    {
      "entropy": 5.580181217193603,
      "epoch": 2.758980406386067,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004241739055876957,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2149147555232048,
      "num_tokens": 63122021.0,
      "step": 30415
    },
    {
      "entropy": 5.637083959579468,
      "epoch": 2.759433962264151,
      "grad_norm": 1.125,
      "learning_rate": 0.00042414968126246985,
      "loss": 5.0196,
      "mean_token_accuracy": 0.21200834661722184,
      "num_tokens": 63132286.0,
      "step": 30420
    },
    {
      "entropy": 5.521076107025147,
      "epoch": 2.759887518142235,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042412545385279055,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22026446610689163,
      "num_tokens": 63142159.0,
      "step": 30425
    },
    {
      "entropy": 5.490342807769776,
      "epoch": 2.760341074020319,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042410122335915883,
      "loss": 4.8572,
      "mean_token_accuracy": 0.21545362770557402,
      "num_tokens": 63152225.0,
      "step": 30430
    },
    {
      "entropy": 5.536431932449341,
      "epoch": 2.7607946298984034,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004240769897820758,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21610554903745652,
      "num_tokens": 63162522.0,
      "step": 30435
    },
    {
      "entropy": 5.565923070907592,
      "epoch": 2.7612481857764877,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042405275312204266,
      "loss": 4.8671,
      "mean_token_accuracy": 0.22013120204210282,
      "num_tokens": 63172094.0,
      "step": 30440
    },
    {
      "entropy": 5.556920385360717,
      "epoch": 2.761701741654572,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042402851337956066,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21307509541511535,
      "num_tokens": 63182122.0,
      "step": 30445
    },
    {
      "entropy": 5.498685550689697,
      "epoch": 2.762155297532656,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042400427055513094,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2125438541173935,
      "num_tokens": 63193174.0,
      "step": 30450
    },
    {
      "entropy": 5.532286310195923,
      "epoch": 2.7626088534107405,
      "grad_norm": 1.140625,
      "learning_rate": 0.000423980024649255,
      "loss": 4.87,
      "mean_token_accuracy": 0.2257883921265602,
      "num_tokens": 63203698.0,
      "step": 30455
    },
    {
      "entropy": 5.466397762298584,
      "epoch": 2.7630624092888243,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004239557756624342,
      "loss": 4.8277,
      "mean_token_accuracy": 0.21761790663003922,
      "num_tokens": 63213343.0,
      "step": 30460
    },
    {
      "entropy": 5.603932666778564,
      "epoch": 2.7635159651669086,
      "grad_norm": 1.125,
      "learning_rate": 0.00042393152359516994,
      "loss": 5.0038,
      "mean_token_accuracy": 0.2097679555416107,
      "num_tokens": 63223962.0,
      "step": 30465
    },
    {
      "entropy": 5.526310825347901,
      "epoch": 2.763969521044993,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042390726844796384,
      "loss": 4.8926,
      "mean_token_accuracy": 0.21778165251016618,
      "num_tokens": 63234279.0,
      "step": 30470
    },
    {
      "entropy": 5.592938423156738,
      "epoch": 2.7644230769230766,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004238830102213174,
      "loss": 4.9794,
      "mean_token_accuracy": 0.21780745238065718,
      "num_tokens": 63245567.0,
      "step": 30475
    },
    {
      "entropy": 5.525066709518432,
      "epoch": 2.764876632801161,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004238587489157325,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2200300484895706,
      "num_tokens": 63256227.0,
      "step": 30480
    },
    {
      "entropy": 5.509989643096924,
      "epoch": 2.765330188679245,
      "grad_norm": 1.125,
      "learning_rate": 0.0004238344845317107,
      "loss": 4.8257,
      "mean_token_accuracy": 0.22090769708156585,
      "num_tokens": 63266618.0,
      "step": 30485
    },
    {
      "entropy": 5.498007917404175,
      "epoch": 2.7657837445573294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004238102170697539,
      "loss": 4.8653,
      "mean_token_accuracy": 0.22233632653951646,
      "num_tokens": 63277130.0,
      "step": 30490
    },
    {
      "entropy": 5.460305118560791,
      "epoch": 2.7662373004354137,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004237859465303637,
      "loss": 4.8029,
      "mean_token_accuracy": 0.2217716932296753,
      "num_tokens": 63287351.0,
      "step": 30495
    },
    {
      "entropy": 5.460341453552246,
      "epoch": 2.766690856313498,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042376167291404234,
      "loss": 4.8828,
      "mean_token_accuracy": 0.2162327215075493,
      "num_tokens": 63298573.0,
      "step": 30500
    },
    {
      "entropy": 5.550261068344116,
      "epoch": 2.7671444121915822,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042373739622129166,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21402664333581925,
      "num_tokens": 63310740.0,
      "step": 30505
    },
    {
      "entropy": 5.530867242813111,
      "epoch": 2.767597968069666,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042371311645261376,
      "loss": 4.922,
      "mean_token_accuracy": 0.2176116332411766,
      "num_tokens": 63321323.0,
      "step": 30510
    },
    {
      "entropy": 5.544354677200317,
      "epoch": 2.7680515239477503,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004236888336085107,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21989640295505525,
      "num_tokens": 63332322.0,
      "step": 30515
    },
    {
      "entropy": 5.523030090332031,
      "epoch": 2.7685050798258346,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004236645476894846,
      "loss": 4.8245,
      "mean_token_accuracy": 0.222117879986763,
      "num_tokens": 63342204.0,
      "step": 30520
    },
    {
      "entropy": 5.444663763046265,
      "epoch": 2.768958635703919,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004236402586960378,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22342048734426498,
      "num_tokens": 63352460.0,
      "step": 30525
    },
    {
      "entropy": 5.4670610427856445,
      "epoch": 2.7694121915820027,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004236159666286726,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22196444720029831,
      "num_tokens": 63362474.0,
      "step": 30530
    },
    {
      "entropy": 5.596309423446655,
      "epoch": 2.769865747460087,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042359167148789135,
      "loss": 4.8982,
      "mean_token_accuracy": 0.22139718979597092,
      "num_tokens": 63372537.0,
      "step": 30535
    },
    {
      "entropy": 5.506354713439942,
      "epoch": 2.7703193033381712,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004235673732741964,
      "loss": 4.8238,
      "mean_token_accuracy": 0.22366830855607986,
      "num_tokens": 63382985.0,
      "step": 30540
    },
    {
      "entropy": 5.484205627441407,
      "epoch": 2.7707728592162555,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042354307198809037,
      "loss": 4.8834,
      "mean_token_accuracy": 0.21712336093187332,
      "num_tokens": 63393424.0,
      "step": 30545
    },
    {
      "entropy": 5.4567066669464115,
      "epoch": 2.7712264150943398,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004235187676300757,
      "loss": 4.8305,
      "mean_token_accuracy": 0.2220056802034378,
      "num_tokens": 63403160.0,
      "step": 30550
    },
    {
      "entropy": 5.5553553104400635,
      "epoch": 2.771679970972424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004234944602006552,
      "loss": 4.9437,
      "mean_token_accuracy": 0.21894798725843428,
      "num_tokens": 63413001.0,
      "step": 30555
    },
    {
      "entropy": 5.561584758758545,
      "epoch": 2.772133526850508,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042347014970033124,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22200512886047363,
      "num_tokens": 63422735.0,
      "step": 30560
    },
    {
      "entropy": 5.469807195663452,
      "epoch": 2.772587082728592,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004234458361296069,
      "loss": 4.8894,
      "mean_token_accuracy": 0.21574612855911254,
      "num_tokens": 63432566.0,
      "step": 30565
    },
    {
      "entropy": 5.5209135055542,
      "epoch": 2.7730406386066764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042342151948898476,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21468929350376129,
      "num_tokens": 63442412.0,
      "step": 30570
    },
    {
      "entropy": 5.517923212051391,
      "epoch": 2.7734941944847606,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042339719977896784,
      "loss": 4.9106,
      "mean_token_accuracy": 0.21989350616931916,
      "num_tokens": 63453018.0,
      "step": 30575
    },
    {
      "entropy": 5.594983959197998,
      "epoch": 2.7739477503628445,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000423372877000059,
      "loss": 4.9594,
      "mean_token_accuracy": 0.20720903873443602,
      "num_tokens": 63464286.0,
      "step": 30580
    },
    {
      "entropy": 5.527478504180908,
      "epoch": 2.7744013062409287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042334855115276117,
      "loss": 4.9229,
      "mean_token_accuracy": 0.21502056270837783,
      "num_tokens": 63475138.0,
      "step": 30585
    },
    {
      "entropy": 5.595297527313233,
      "epoch": 2.774854862119013,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042332422223757756,
      "loss": 4.9881,
      "mean_token_accuracy": 0.21430517137050628,
      "num_tokens": 63486104.0,
      "step": 30590
    },
    {
      "entropy": 5.47593502998352,
      "epoch": 2.7753084179970973,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042329989025501123,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2147979035973549,
      "num_tokens": 63496490.0,
      "step": 30595
    },
    {
      "entropy": 5.491402626037598,
      "epoch": 2.7757619738751815,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042327555520556536,
      "loss": 4.8727,
      "mean_token_accuracy": 0.21748384684324265,
      "num_tokens": 63506562.0,
      "step": 30600
    },
    {
      "entropy": 5.5572162628173825,
      "epoch": 2.776215529753266,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042325121708974324,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22070915549993514,
      "num_tokens": 63516641.0,
      "step": 30605
    },
    {
      "entropy": 5.567569398880005,
      "epoch": 2.7766690856313496,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004232268759080482,
      "loss": 4.9105,
      "mean_token_accuracy": 0.21740418374538423,
      "num_tokens": 63526864.0,
      "step": 30610
    },
    {
      "entropy": 5.531572675704956,
      "epoch": 2.777122641509434,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042320253166098354,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2093336135149002,
      "num_tokens": 63537160.0,
      "step": 30615
    },
    {
      "entropy": 5.572648048400879,
      "epoch": 2.777576197387518,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042317818434905275,
      "loss": 4.8826,
      "mean_token_accuracy": 0.21674850583076477,
      "num_tokens": 63547485.0,
      "step": 30620
    },
    {
      "entropy": 5.555351829528808,
      "epoch": 2.7780297532656024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004231538339727594,
      "loss": 4.8848,
      "mean_token_accuracy": 0.21273261606693267,
      "num_tokens": 63559369.0,
      "step": 30625
    },
    {
      "entropy": 5.628942346572876,
      "epoch": 2.7784833091436862,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042312948053260704,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20401397198438645,
      "num_tokens": 63570033.0,
      "step": 30630
    },
    {
      "entropy": 5.533664608001709,
      "epoch": 2.7789368650217705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004231051240290992,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2176375150680542,
      "num_tokens": 63581274.0,
      "step": 30635
    },
    {
      "entropy": 5.575471591949463,
      "epoch": 2.7793904208998548,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042308076446273963,
      "loss": 4.926,
      "mean_token_accuracy": 0.21599977016448973,
      "num_tokens": 63590903.0,
      "step": 30640
    },
    {
      "entropy": 5.510876941680908,
      "epoch": 2.779843976777939,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042305640183403214,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2141154080629349,
      "num_tokens": 63600982.0,
      "step": 30645
    },
    {
      "entropy": 5.452576398849487,
      "epoch": 2.7802975326560233,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042303203614348057,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22087979465723037,
      "num_tokens": 63612411.0,
      "step": 30650
    },
    {
      "entropy": 5.523834943771362,
      "epoch": 2.7807510885341076,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004230076673915888,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22421470582485198,
      "num_tokens": 63622128.0,
      "step": 30655
    },
    {
      "entropy": 5.5620626449584964,
      "epoch": 2.781204644412192,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042298329557886074,
      "loss": 4.9394,
      "mean_token_accuracy": 0.21709243059158326,
      "num_tokens": 63632047.0,
      "step": 30660
    },
    {
      "entropy": 5.557314729690551,
      "epoch": 2.7816582002902757,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004229589207058003,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21210708767175673,
      "num_tokens": 63642490.0,
      "step": 30665
    },
    {
      "entropy": 5.504116725921631,
      "epoch": 2.78211175616836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042293454277291177,
      "loss": 4.8996,
      "mean_token_accuracy": 0.21591146290302277,
      "num_tokens": 63653206.0,
      "step": 30670
    },
    {
      "entropy": 5.4893557071685795,
      "epoch": 2.782565312046444,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042291016178069916,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2182347446680069,
      "num_tokens": 63663427.0,
      "step": 30675
    },
    {
      "entropy": 5.458497762680054,
      "epoch": 2.7830188679245285,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004228857777296668,
      "loss": 4.8822,
      "mean_token_accuracy": 0.219146491587162,
      "num_tokens": 63674269.0,
      "step": 30680
    },
    {
      "entropy": 5.535872936248779,
      "epoch": 2.7834724238026123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004228613906203188,
      "loss": 4.8846,
      "mean_token_accuracy": 0.21309252232313156,
      "num_tokens": 63683989.0,
      "step": 30685
    },
    {
      "entropy": 5.527919578552246,
      "epoch": 2.7839259796806965,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042283700045315955,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2108805075287819,
      "num_tokens": 63694540.0,
      "step": 30690
    },
    {
      "entropy": 5.531034660339356,
      "epoch": 2.784379535558781,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042281260722869355,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2139385774731636,
      "num_tokens": 63705149.0,
      "step": 30695
    },
    {
      "entropy": 5.521725273132324,
      "epoch": 2.784833091436865,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042278821094742495,
      "loss": 4.8575,
      "mean_token_accuracy": 0.21866624653339387,
      "num_tokens": 63714061.0,
      "step": 30700
    },
    {
      "entropy": 5.552300548553466,
      "epoch": 2.7852866473149493,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042276381160985874,
      "loss": 4.9523,
      "mean_token_accuracy": 0.21544505655765533,
      "num_tokens": 63724732.0,
      "step": 30705
    },
    {
      "entropy": 5.532246065139771,
      "epoch": 2.7857402031930336,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004227394092164991,
      "loss": 4.946,
      "mean_token_accuracy": 0.21359337568283082,
      "num_tokens": 63734942.0,
      "step": 30710
    },
    {
      "entropy": 5.607596731185913,
      "epoch": 2.7861937590711174,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004227150037678509,
      "loss": 5.026,
      "mean_token_accuracy": 0.21817888766527177,
      "num_tokens": 63745326.0,
      "step": 30715
    },
    {
      "entropy": 5.539146471023559,
      "epoch": 2.7866473149492017,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004226905952644188,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21137509793043135,
      "num_tokens": 63755355.0,
      "step": 30720
    },
    {
      "entropy": 5.524612092971802,
      "epoch": 2.787100870827286,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042266618370670743,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2129056289792061,
      "num_tokens": 63765112.0,
      "step": 30725
    },
    {
      "entropy": 5.506501150131226,
      "epoch": 2.7875544267053702,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042264176909522183,
      "loss": 4.856,
      "mean_token_accuracy": 0.22156573534011842,
      "num_tokens": 63774680.0,
      "step": 30730
    },
    {
      "entropy": 5.574916410446167,
      "epoch": 2.788007982583454,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004226173514304669,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21419061422348024,
      "num_tokens": 63784939.0,
      "step": 30735
    },
    {
      "entropy": 5.5736040592193605,
      "epoch": 2.7884615384615383,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004225929307129474,
      "loss": 4.997,
      "mean_token_accuracy": 0.2051139935851097,
      "num_tokens": 63795836.0,
      "step": 30740
    },
    {
      "entropy": 5.524158906936646,
      "epoch": 2.7889150943396226,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042256850694316844,
      "loss": 4.8616,
      "mean_token_accuracy": 0.21171969324350357,
      "num_tokens": 63805607.0,
      "step": 30745
    },
    {
      "entropy": 5.496328830718994,
      "epoch": 2.789368650217707,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004225440801216353,
      "loss": 4.8775,
      "mean_token_accuracy": 0.21394830495119094,
      "num_tokens": 63815081.0,
      "step": 30750
    },
    {
      "entropy": 5.4668434143066404,
      "epoch": 2.789822206095791,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004225196502488529,
      "loss": 4.8864,
      "mean_token_accuracy": 0.21572887152433395,
      "num_tokens": 63825932.0,
      "step": 30755
    },
    {
      "entropy": 5.4969518184661865,
      "epoch": 2.7902757619738754,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042249521732532657,
      "loss": 4.844,
      "mean_token_accuracy": 0.219173763692379,
      "num_tokens": 63835653.0,
      "step": 30760
    },
    {
      "entropy": 5.633457279205322,
      "epoch": 2.790729317851959,
      "grad_norm": 1.125,
      "learning_rate": 0.0004224707813515615,
      "loss": 5.0419,
      "mean_token_accuracy": 0.20953703671693802,
      "num_tokens": 63847657.0,
      "step": 30765
    },
    {
      "entropy": 5.6358260154724125,
      "epoch": 2.7911828737300435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042244634232806324,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2163883164525032,
      "num_tokens": 63858476.0,
      "step": 30770
    },
    {
      "entropy": 5.488044548034668,
      "epoch": 2.7916364296081277,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042242190025533694,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2192759931087494,
      "num_tokens": 63867970.0,
      "step": 30775
    },
    {
      "entropy": 5.542297458648681,
      "epoch": 2.792089985486212,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004223974551338881,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22458372563123702,
      "num_tokens": 63878294.0,
      "step": 30780
    },
    {
      "entropy": 5.5060694217681885,
      "epoch": 2.792543541364296,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004223730069642224,
      "loss": 4.8991,
      "mean_token_accuracy": 0.2111222416162491,
      "num_tokens": 63889315.0,
      "step": 30785
    },
    {
      "entropy": 5.470957612991333,
      "epoch": 2.79299709724238,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004223485557468454,
      "loss": 4.8696,
      "mean_token_accuracy": 0.21330437809228897,
      "num_tokens": 63899923.0,
      "step": 30790
    },
    {
      "entropy": 5.484261560440063,
      "epoch": 2.7934506531204644,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042232410148226277,
      "loss": 4.868,
      "mean_token_accuracy": 0.2193808600306511,
      "num_tokens": 63909813.0,
      "step": 30795
    },
    {
      "entropy": 5.547342586517334,
      "epoch": 2.7939042089985486,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042229964417098007,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21508964598178865,
      "num_tokens": 63920931.0,
      "step": 30800
    },
    {
      "entropy": 5.509487819671631,
      "epoch": 2.794357764876633,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042227518381350316,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22297311425209046,
      "num_tokens": 63930799.0,
      "step": 30805
    },
    {
      "entropy": 5.620544481277466,
      "epoch": 2.794811320754717,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042225072041033804,
      "loss": 5.019,
      "mean_token_accuracy": 0.2113138258457184,
      "num_tokens": 63941763.0,
      "step": 30810
    },
    {
      "entropy": 5.603292083740234,
      "epoch": 2.7952648766328014,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042222625396199043,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20460831820964814,
      "num_tokens": 63952775.0,
      "step": 30815
    },
    {
      "entropy": 5.559855699539185,
      "epoch": 2.7957184325108853,
      "grad_norm": 1.125,
      "learning_rate": 0.0004222017844689663,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21407727152109146,
      "num_tokens": 63964223.0,
      "step": 30820
    },
    {
      "entropy": 5.576049184799194,
      "epoch": 2.7961719883889695,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004221773119317719,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21223429590463638,
      "num_tokens": 63976413.0,
      "step": 30825
    },
    {
      "entropy": 5.638586139678955,
      "epoch": 2.796625544267054,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042215283635091313,
      "loss": 4.955,
      "mean_token_accuracy": 0.21725727915763854,
      "num_tokens": 63986578.0,
      "step": 30830
    },
    {
      "entropy": 5.567459535598755,
      "epoch": 2.7970791001451376,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004221283577268962,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21995437890291214,
      "num_tokens": 63996856.0,
      "step": 30835
    },
    {
      "entropy": 5.484815454483032,
      "epoch": 2.797532656023222,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042210387606022724,
      "loss": 4.874,
      "mean_token_accuracy": 0.22053915858268738,
      "num_tokens": 64007700.0,
      "step": 30840
    },
    {
      "entropy": 5.593621444702149,
      "epoch": 2.797986211901306,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004220793913514128,
      "loss": 5.0274,
      "mean_token_accuracy": 0.20663743019104003,
      "num_tokens": 64018771.0,
      "step": 30845
    },
    {
      "entropy": 5.562261533737183,
      "epoch": 2.7984397677793904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004220549036009589,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2145187094807625,
      "num_tokens": 64029381.0,
      "step": 30850
    },
    {
      "entropy": 5.455021715164184,
      "epoch": 2.7988933236574747,
      "grad_norm": 1.25,
      "learning_rate": 0.00042203041280937224,
      "loss": 4.8575,
      "mean_token_accuracy": 0.2154773622751236,
      "num_tokens": 64039596.0,
      "step": 30855
    },
    {
      "entropy": 5.583695697784424,
      "epoch": 2.799346879535559,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004220059189771592,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22600321918725969,
      "num_tokens": 64049123.0,
      "step": 30860
    },
    {
      "entropy": 5.539047765731811,
      "epoch": 2.799800435413643,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042198142210482623,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2102591410279274,
      "num_tokens": 64059426.0,
      "step": 30865
    },
    {
      "entropy": 5.506822872161865,
      "epoch": 2.800253991291727,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042195692219287997,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22850680947303773,
      "num_tokens": 64069227.0,
      "step": 30870
    },
    {
      "entropy": 5.550188207626343,
      "epoch": 2.8007075471698113,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004219324192418272,
      "loss": 4.9217,
      "mean_token_accuracy": 0.21839002519845963,
      "num_tokens": 64078482.0,
      "step": 30875
    },
    {
      "entropy": 5.536701250076294,
      "epoch": 2.8011611030478956,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042190791325217447,
      "loss": 4.8882,
      "mean_token_accuracy": 0.2170430138707161,
      "num_tokens": 64088664.0,
      "step": 30880
    },
    {
      "entropy": 5.508675050735474,
      "epoch": 2.80161465892598,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004218834042244287,
      "loss": 4.8394,
      "mean_token_accuracy": 0.221296264231205,
      "num_tokens": 64099316.0,
      "step": 30885
    },
    {
      "entropy": 5.482932806015015,
      "epoch": 2.8020682148040637,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042185889215909665,
      "loss": 4.8846,
      "mean_token_accuracy": 0.21833451837301254,
      "num_tokens": 64109786.0,
      "step": 30890
    },
    {
      "entropy": 5.540563488006592,
      "epoch": 2.802521770682148,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042183437705668535,
      "loss": 4.9045,
      "mean_token_accuracy": 0.21909187585115433,
      "num_tokens": 64120771.0,
      "step": 30895
    },
    {
      "entropy": 5.581007051467895,
      "epoch": 2.802975326560232,
      "grad_norm": 1.125,
      "learning_rate": 0.00042180985891770176,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21043097525835036,
      "num_tokens": 64130833.0,
      "step": 30900
    },
    {
      "entropy": 5.505196952819825,
      "epoch": 2.8034288824383164,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004217853377426528,
      "loss": 4.8772,
      "mean_token_accuracy": 0.21740821301937102,
      "num_tokens": 64139672.0,
      "step": 30905
    },
    {
      "entropy": 5.619898319244385,
      "epoch": 2.8038824383164007,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004217608135320456,
      "loss": 5.0441,
      "mean_token_accuracy": 0.21198188960552217,
      "num_tokens": 64150383.0,
      "step": 30910
    },
    {
      "entropy": 5.488873100280761,
      "epoch": 2.804335994194485,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004217362862863875,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22111288607120513,
      "num_tokens": 64160786.0,
      "step": 30915
    },
    {
      "entropy": 5.589868783950806,
      "epoch": 2.804789550072569,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042171175600618546,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2092306911945343,
      "num_tokens": 64172057.0,
      "step": 30920
    },
    {
      "entropy": 5.543269205093384,
      "epoch": 2.805243105950653,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042168722269194704,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2217864692211151,
      "num_tokens": 64182599.0,
      "step": 30925
    },
    {
      "entropy": 5.5318528175354,
      "epoch": 2.8056966618287373,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004216626863441794,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21121252030134202,
      "num_tokens": 64193356.0,
      "step": 30930
    },
    {
      "entropy": 5.586373376846313,
      "epoch": 2.8061502177068216,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042163814696339014,
      "loss": 5.0334,
      "mean_token_accuracy": 0.20548325926065444,
      "num_tokens": 64203625.0,
      "step": 30935
    },
    {
      "entropy": 5.563460111618042,
      "epoch": 2.8066037735849054,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004216136045500866,
      "loss": 5.0292,
      "mean_token_accuracy": 0.21040269136428832,
      "num_tokens": 64214742.0,
      "step": 30940
    },
    {
      "entropy": 5.611554336547852,
      "epoch": 2.8070573294629897,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004215890591047763,
      "loss": 4.934,
      "mean_token_accuracy": 0.21399957835674285,
      "num_tokens": 64223811.0,
      "step": 30945
    },
    {
      "entropy": 5.5755163669586185,
      "epoch": 2.807510885341074,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042156451062796686,
      "loss": 4.8368,
      "mean_token_accuracy": 0.225464129447937,
      "num_tokens": 64234854.0,
      "step": 30950
    },
    {
      "entropy": 5.530154371261597,
      "epoch": 2.8079644412191582,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042153995912016606,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21348231732845308,
      "num_tokens": 64245914.0,
      "step": 30955
    },
    {
      "entropy": 5.537575960159302,
      "epoch": 2.8084179970972425,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004215154045818816,
      "loss": 4.9862,
      "mean_token_accuracy": 0.2114098533987999,
      "num_tokens": 64256217.0,
      "step": 30960
    },
    {
      "entropy": 5.575512027740478,
      "epoch": 2.8088715529753268,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004214908470136212,
      "loss": 4.9862,
      "mean_token_accuracy": 0.2037487134337425,
      "num_tokens": 64267211.0,
      "step": 30965
    },
    {
      "entropy": 5.61586627960205,
      "epoch": 2.8093251088534106,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004214662864158927,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2164837673306465,
      "num_tokens": 64277770.0,
      "step": 30970
    },
    {
      "entropy": 5.51845965385437,
      "epoch": 2.809778664731495,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042144172278920413,
      "loss": 4.8413,
      "mean_token_accuracy": 0.22367362529039383,
      "num_tokens": 64287973.0,
      "step": 30975
    },
    {
      "entropy": 5.490904235839844,
      "epoch": 2.810232220609579,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004214171561340634,
      "loss": 4.9305,
      "mean_token_accuracy": 0.20926944613456727,
      "num_tokens": 64297898.0,
      "step": 30980
    },
    {
      "entropy": 5.496457910537719,
      "epoch": 2.8106857764876634,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042139258645097863,
      "loss": 4.8051,
      "mean_token_accuracy": 0.22897038161754607,
      "num_tokens": 64308114.0,
      "step": 30985
    },
    {
      "entropy": 5.531622076034546,
      "epoch": 2.811139332365747,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042136801374045775,
      "loss": 4.8895,
      "mean_token_accuracy": 0.21641914546489716,
      "num_tokens": 64319586.0,
      "step": 30990
    },
    {
      "entropy": 5.556856679916382,
      "epoch": 2.8115928882438315,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004213434380030091,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2129749149084091,
      "num_tokens": 64330612.0,
      "step": 30995
    },
    {
      "entropy": 5.581273460388184,
      "epoch": 2.8120464441219157,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004213188592391409,
      "loss": 5.0095,
      "mean_token_accuracy": 0.21192465126514434,
      "num_tokens": 64341934.0,
      "step": 31000
    },
    {
      "entropy": 5.557264375686645,
      "epoch": 2.8125,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004212942774493614,
      "loss": 4.9065,
      "mean_token_accuracy": 0.212347112596035,
      "num_tokens": 64352166.0,
      "step": 31005
    },
    {
      "entropy": 5.558771371841431,
      "epoch": 2.8129535558780843,
      "grad_norm": 1.125,
      "learning_rate": 0.00042126969263417885,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21027209162712096,
      "num_tokens": 64361995.0,
      "step": 31010
    },
    {
      "entropy": 5.511719560623169,
      "epoch": 2.8134071117561685,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042124510479410183,
      "loss": 4.8937,
      "mean_token_accuracy": 0.20847588330507277,
      "num_tokens": 64372336.0,
      "step": 31015
    },
    {
      "entropy": 5.5432158470153805,
      "epoch": 2.813860667634253,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042122051392963885,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22187695056200027,
      "num_tokens": 64382309.0,
      "step": 31020
    },
    {
      "entropy": 5.555031156539917,
      "epoch": 2.8143142235123366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004211959200412984,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21324883550405502,
      "num_tokens": 64393205.0,
      "step": 31025
    },
    {
      "entropy": 5.50804386138916,
      "epoch": 2.814767779390421,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004211713231295889,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21478293985128402,
      "num_tokens": 64403068.0,
      "step": 31030
    },
    {
      "entropy": 5.566445684432983,
      "epoch": 2.815221335268505,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004211467231950194,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2172972872853279,
      "num_tokens": 64414296.0,
      "step": 31035
    },
    {
      "entropy": 5.527777481079101,
      "epoch": 2.8156748911465894,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042112212023809824,
      "loss": 4.9439,
      "mean_token_accuracy": 0.20978185534477234,
      "num_tokens": 64424748.0,
      "step": 31040
    },
    {
      "entropy": 5.530280590057373,
      "epoch": 2.8161284470246732,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004210975142593345,
      "loss": 4.88,
      "mean_token_accuracy": 0.2162256196141243,
      "num_tokens": 64434723.0,
      "step": 31045
    },
    {
      "entropy": 5.6028824806213375,
      "epoch": 2.8165820029027575,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004210729052592369,
      "loss": 5.0032,
      "mean_token_accuracy": 0.21000772565603257,
      "num_tokens": 64445864.0,
      "step": 31050
    },
    {
      "entropy": 5.511378288269043,
      "epoch": 2.8170355587808418,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004210482932383144,
      "loss": 4.8566,
      "mean_token_accuracy": 0.21516052931547164,
      "num_tokens": 64456704.0,
      "step": 31055
    },
    {
      "entropy": 5.458221054077148,
      "epoch": 2.817489114658926,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000421023678197076,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22078704535961152,
      "num_tokens": 64466870.0,
      "step": 31060
    },
    {
      "entropy": 5.554971504211426,
      "epoch": 2.8179426705370103,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004209990601360307,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21220726221799852,
      "num_tokens": 64477069.0,
      "step": 31065
    },
    {
      "entropy": 5.601056718826294,
      "epoch": 2.8183962264150946,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042097443905568765,
      "loss": 4.9435,
      "mean_token_accuracy": 0.20832460820674897,
      "num_tokens": 64487165.0,
      "step": 31070
    },
    {
      "entropy": 5.500105142593384,
      "epoch": 2.8188497822931784,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042094981495655596,
      "loss": 4.8116,
      "mean_token_accuracy": 0.2192513808608055,
      "num_tokens": 64497977.0,
      "step": 31075
    },
    {
      "entropy": 5.550768661499023,
      "epoch": 2.8193033381712627,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000420925187839145,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21169591397047044,
      "num_tokens": 64508767.0,
      "step": 31080
    },
    {
      "entropy": 5.496765518188477,
      "epoch": 2.819756894049347,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042090055770396383,
      "loss": 4.8408,
      "mean_token_accuracy": 0.2166010171175003,
      "num_tokens": 64519040.0,
      "step": 31085
    },
    {
      "entropy": 5.523709487915039,
      "epoch": 2.820210449927431,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042087592455152206,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22410670518875123,
      "num_tokens": 64529721.0,
      "step": 31090
    },
    {
      "entropy": 5.542969274520874,
      "epoch": 2.820664005805515,
      "grad_norm": 1.046875,
      "learning_rate": 0.000420851288382329,
      "loss": 5.0251,
      "mean_token_accuracy": 0.19784930050373079,
      "num_tokens": 64541299.0,
      "step": 31095
    },
    {
      "entropy": 5.567285108566284,
      "epoch": 2.8211175616835993,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042082664919689406,
      "loss": 4.8815,
      "mean_token_accuracy": 0.21351644843816758,
      "num_tokens": 64550831.0,
      "step": 31100
    },
    {
      "entropy": 5.544571781158448,
      "epoch": 2.8215711175616836,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004208020069957269,
      "loss": 4.9118,
      "mean_token_accuracy": 0.21320996731519698,
      "num_tokens": 64562076.0,
      "step": 31105
    },
    {
      "entropy": 5.4988298416137695,
      "epoch": 2.822024673439768,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000420777361779337,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2141510710120201,
      "num_tokens": 64572955.0,
      "step": 31110
    },
    {
      "entropy": 5.577213478088379,
      "epoch": 2.822478229317852,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042075271354823427,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22119008898735046,
      "num_tokens": 64584376.0,
      "step": 31115
    },
    {
      "entropy": 5.61672568321228,
      "epoch": 2.8229317851959363,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004207280623029281,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2101093515753746,
      "num_tokens": 64594522.0,
      "step": 31120
    },
    {
      "entropy": 5.5586175441741945,
      "epoch": 2.82338534107402,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042070340804392853,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21815845966339112,
      "num_tokens": 64604625.0,
      "step": 31125
    },
    {
      "entropy": 5.548998212814331,
      "epoch": 2.8238388969521044,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004206787507717454,
      "loss": 4.927,
      "mean_token_accuracy": 0.21258337795734406,
      "num_tokens": 64614184.0,
      "step": 31130
    },
    {
      "entropy": 5.633258056640625,
      "epoch": 2.8242924528301887,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042065409048688856,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20751901119947433,
      "num_tokens": 64623019.0,
      "step": 31135
    },
    {
      "entropy": 5.550436067581177,
      "epoch": 2.824746008708273,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000420629427189868,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2100003182888031,
      "num_tokens": 64634017.0,
      "step": 31140
    },
    {
      "entropy": 5.516027450561523,
      "epoch": 2.825199564586357,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042060476088119376,
      "loss": 4.9095,
      "mean_token_accuracy": 0.223791466653347,
      "num_tokens": 64645007.0,
      "step": 31145
    },
    {
      "entropy": 5.621215581893921,
      "epoch": 2.825653120464441,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000420580091561376,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20933672934770584,
      "num_tokens": 64655705.0,
      "step": 31150
    },
    {
      "entropy": 5.634806108474732,
      "epoch": 2.8261066763425253,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042055541923092485,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21371605694293977,
      "num_tokens": 64666094.0,
      "step": 31155
    },
    {
      "entropy": 5.590093755722046,
      "epoch": 2.8265602322206096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004205307438903505,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2107161909341812,
      "num_tokens": 64676107.0,
      "step": 31160
    },
    {
      "entropy": 5.536475038528442,
      "epoch": 2.827013788098694,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004205060655401633,
      "loss": 4.8826,
      "mean_token_accuracy": 0.21870268136262894,
      "num_tokens": 64685846.0,
      "step": 31165
    },
    {
      "entropy": 5.527658843994141,
      "epoch": 2.827467343976778,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004204813841808736,
      "loss": 4.8739,
      "mean_token_accuracy": 0.2178611069917679,
      "num_tokens": 64695959.0,
      "step": 31170
    },
    {
      "entropy": 5.539642810821533,
      "epoch": 2.8279208998548624,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004204566998129917,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2217714875936508,
      "num_tokens": 64705594.0,
      "step": 31175
    },
    {
      "entropy": 5.5490350246429445,
      "epoch": 2.828374455732946,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042043201243702834,
      "loss": 4.9405,
      "mean_token_accuracy": 0.21713307201862336,
      "num_tokens": 64715974.0,
      "step": 31180
    },
    {
      "entropy": 5.484439325332642,
      "epoch": 2.8288280116110305,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004204073220534938,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21724633872509003,
      "num_tokens": 64725846.0,
      "step": 31185
    },
    {
      "entropy": 5.513561248779297,
      "epoch": 2.8292815674891147,
      "grad_norm": 1.125,
      "learning_rate": 0.0004203826286628988,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22389216125011444,
      "num_tokens": 64735789.0,
      "step": 31190
    },
    {
      "entropy": 5.513552951812744,
      "epoch": 2.8297351233671986,
      "grad_norm": 1.1875,
      "learning_rate": 0.000420357932265754,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22930199801921844,
      "num_tokens": 64745738.0,
      "step": 31195
    },
    {
      "entropy": 5.5790260314941404,
      "epoch": 2.830188679245283,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004203332328625701,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21140057891607283,
      "num_tokens": 64756523.0,
      "step": 31200
    },
    {
      "entropy": 5.615804433822632,
      "epoch": 2.830642235123367,
      "grad_norm": 1.125,
      "learning_rate": 0.00042030853045385796,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20886973887681962,
      "num_tokens": 64767367.0,
      "step": 31205
    },
    {
      "entropy": 5.559985256195068,
      "epoch": 2.8310957910014514,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042028382504012837,
      "loss": 4.9365,
      "mean_token_accuracy": 0.2171519175171852,
      "num_tokens": 64777729.0,
      "step": 31210
    },
    {
      "entropy": 5.618748760223388,
      "epoch": 2.8315493468795356,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042025911662189224,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21584361046552658,
      "num_tokens": 64788223.0,
      "step": 31215
    },
    {
      "entropy": 5.524864530563354,
      "epoch": 2.83200290275762,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042023440519966057,
      "loss": 4.931,
      "mean_token_accuracy": 0.21378059685230255,
      "num_tokens": 64797367.0,
      "step": 31220
    },
    {
      "entropy": 5.568392181396485,
      "epoch": 2.832456458635704,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042020969077394434,
      "loss": 4.932,
      "mean_token_accuracy": 0.2151729241013527,
      "num_tokens": 64807179.0,
      "step": 31225
    },
    {
      "entropy": 5.597186040878296,
      "epoch": 2.832910014513788,
      "grad_norm": 1.125,
      "learning_rate": 0.00042018497334525475,
      "loss": 4.8408,
      "mean_token_accuracy": 0.21051837801933287,
      "num_tokens": 64817276.0,
      "step": 31230
    },
    {
      "entropy": 5.534233093261719,
      "epoch": 2.8333635703918723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004201602529141029,
      "loss": 4.928,
      "mean_token_accuracy": 0.21452396661043166,
      "num_tokens": 64827509.0,
      "step": 31235
    },
    {
      "entropy": 5.53713207244873,
      "epoch": 2.8338171262699565,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042013552948100003,
      "loss": 4.9159,
      "mean_token_accuracy": 0.21845897287130356,
      "num_tokens": 64836578.0,
      "step": 31240
    },
    {
      "entropy": 5.46690001487732,
      "epoch": 2.834270682148041,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004201108030464574,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2260923132300377,
      "num_tokens": 64846518.0,
      "step": 31245
    },
    {
      "entropy": 5.5815574645996096,
      "epoch": 2.8347242380261246,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042008607361098646,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2170808047056198,
      "num_tokens": 64856304.0,
      "step": 31250
    },
    {
      "entropy": 5.541314220428466,
      "epoch": 2.835177793904209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004200613411750985,
      "loss": 4.8353,
      "mean_token_accuracy": 0.2260214149951935,
      "num_tokens": 64867381.0,
      "step": 31255
    },
    {
      "entropy": 5.531509590148926,
      "epoch": 2.835631349782293,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000420036605739305,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21288643628358841,
      "num_tokens": 64877306.0,
      "step": 31260
    },
    {
      "entropy": 5.440706348419189,
      "epoch": 2.8360849056603774,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004200118673041176,
      "loss": 4.8773,
      "mean_token_accuracy": 0.21655137687921525,
      "num_tokens": 64886932.0,
      "step": 31265
    },
    {
      "entropy": 5.577686405181884,
      "epoch": 2.8365384615384617,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041998712587004776,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21200074553489684,
      "num_tokens": 64897387.0,
      "step": 31270
    },
    {
      "entropy": 5.5929523468017575,
      "epoch": 2.836992017416546,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041996238143760726,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21789601892232896,
      "num_tokens": 64907251.0,
      "step": 31275
    },
    {
      "entropy": 5.517402791976929,
      "epoch": 2.8374455732946298,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041993763400730776,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2156141385436058,
      "num_tokens": 64917300.0,
      "step": 31280
    },
    {
      "entropy": 5.493494844436645,
      "epoch": 2.837899129172714,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004199128835796611,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22205340266227722,
      "num_tokens": 64927133.0,
      "step": 31285
    },
    {
      "entropy": 5.525535583496094,
      "epoch": 2.8383526850507983,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000419888130155179,
      "loss": 4.9064,
      "mean_token_accuracy": 0.21540738195180892,
      "num_tokens": 64937789.0,
      "step": 31290
    },
    {
      "entropy": 5.617279815673828,
      "epoch": 2.8388062409288826,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004198633737343735,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2179100066423416,
      "num_tokens": 64948127.0,
      "step": 31295
    },
    {
      "entropy": 5.6134850025177006,
      "epoch": 2.8392597968069664,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004198386143177565,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2190896451473236,
      "num_tokens": 64959632.0,
      "step": 31300
    },
    {
      "entropy": 5.402682638168335,
      "epoch": 2.8397133526850507,
      "grad_norm": 1.125,
      "learning_rate": 0.00041981385190584003,
      "loss": 4.7589,
      "mean_token_accuracy": 0.22537483721971513,
      "num_tokens": 64970411.0,
      "step": 31305
    },
    {
      "entropy": 5.4788289070129395,
      "epoch": 2.840166908563135,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004197890864991362,
      "loss": 4.8839,
      "mean_token_accuracy": 0.21666204631328584,
      "num_tokens": 64980748.0,
      "step": 31310
    },
    {
      "entropy": 5.50677080154419,
      "epoch": 2.840620464441219,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004197643180981572,
      "loss": 4.8483,
      "mean_token_accuracy": 0.21664761900901794,
      "num_tokens": 64990818.0,
      "step": 31315
    },
    {
      "entropy": 5.566329097747802,
      "epoch": 2.8410740203193035,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004197395467034152,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22178769409656524,
      "num_tokens": 65001993.0,
      "step": 31320
    },
    {
      "entropy": 5.486306428909302,
      "epoch": 2.8415275761973877,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004197147723154225,
      "loss": 4.8763,
      "mean_token_accuracy": 0.221916626393795,
      "num_tokens": 65013007.0,
      "step": 31325
    },
    {
      "entropy": 5.563535594940186,
      "epoch": 2.8419811320754715,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004196899949346914,
      "loss": 4.8785,
      "mean_token_accuracy": 0.2166494458913803,
      "num_tokens": 65023595.0,
      "step": 31330
    },
    {
      "entropy": 5.521514558792115,
      "epoch": 2.842434687953556,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004196652145617343,
      "loss": 4.9025,
      "mean_token_accuracy": 0.21616182178258897,
      "num_tokens": 65033631.0,
      "step": 31335
    },
    {
      "entropy": 5.590752363204956,
      "epoch": 2.84288824383164,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004196404311970638,
      "loss": 4.934,
      "mean_token_accuracy": 0.21498224139213562,
      "num_tokens": 65044059.0,
      "step": 31340
    },
    {
      "entropy": 5.550260591506958,
      "epoch": 2.8433417997097243,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004196156448411922,
      "loss": 4.9167,
      "mean_token_accuracy": 0.21222257763147354,
      "num_tokens": 65053931.0,
      "step": 31345
    },
    {
      "entropy": 5.5581341743469235,
      "epoch": 2.843795355587808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041959085549463234,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21883310675621032,
      "num_tokens": 65064011.0,
      "step": 31350
    },
    {
      "entropy": 5.600523853302002,
      "epoch": 2.8442489114658924,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004195660631578966,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21627790182828904,
      "num_tokens": 65074417.0,
      "step": 31355
    },
    {
      "entropy": 5.527333450317383,
      "epoch": 2.8447024673439767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000419541267831498,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2167319431900978,
      "num_tokens": 65084593.0,
      "step": 31360
    },
    {
      "entropy": 5.513705587387085,
      "epoch": 2.845156023222061,
      "grad_norm": 1.171875,
      "learning_rate": 0.000419516469515949,
      "loss": 4.896,
      "mean_token_accuracy": 0.21882117092609404,
      "num_tokens": 65095047.0,
      "step": 31365
    },
    {
      "entropy": 5.58816123008728,
      "epoch": 2.8456095791001452,
      "grad_norm": 1.25,
      "learning_rate": 0.00041949166821176264,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21852320581674575,
      "num_tokens": 65104434.0,
      "step": 31370
    },
    {
      "entropy": 5.598996543884278,
      "epoch": 2.8460631349782295,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041946686391945183,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21633891761302948,
      "num_tokens": 65114039.0,
      "step": 31375
    },
    {
      "entropy": 5.646165800094605,
      "epoch": 2.8465166908563138,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004194420566395294,
      "loss": 5.036,
      "mean_token_accuracy": 0.2072843462228775,
      "num_tokens": 65125387.0,
      "step": 31380
    },
    {
      "entropy": 5.587533044815063,
      "epoch": 2.8469702467343976,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004194172463725084,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21542913615703582,
      "num_tokens": 65136351.0,
      "step": 31385
    },
    {
      "entropy": 5.528819561004639,
      "epoch": 2.847423802612482,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041939243311890195,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2140688717365265,
      "num_tokens": 65147343.0,
      "step": 31390
    },
    {
      "entropy": 5.468143606185913,
      "epoch": 2.847877358490566,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041936761687922324,
      "loss": 4.8693,
      "mean_token_accuracy": 0.22103351056575776,
      "num_tokens": 65157097.0,
      "step": 31395
    },
    {
      "entropy": 5.5240765571594235,
      "epoch": 2.8483309143686504,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041934279765398535,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21291904598474504,
      "num_tokens": 65167546.0,
      "step": 31400
    },
    {
      "entropy": 5.59366250038147,
      "epoch": 2.848784470246734,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004193179754437017,
      "loss": 4.9427,
      "mean_token_accuracy": 0.20671755075454712,
      "num_tokens": 65178497.0,
      "step": 31405
    },
    {
      "entropy": 5.527767992019653,
      "epoch": 2.8492380261248185,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004192931502488855,
      "loss": 4.841,
      "mean_token_accuracy": 0.22376588135957717,
      "num_tokens": 65189530.0,
      "step": 31410
    },
    {
      "entropy": 5.551055192947388,
      "epoch": 2.8496915820029027,
      "grad_norm": 1.125,
      "learning_rate": 0.00041926832207005025,
      "loss": 4.9056,
      "mean_token_accuracy": 0.21482572555541993,
      "num_tokens": 65199836.0,
      "step": 31415
    },
    {
      "entropy": 5.455656909942627,
      "epoch": 2.850145137880987,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004192434909077092,
      "loss": 4.7617,
      "mean_token_accuracy": 0.2233502060174942,
      "num_tokens": 65210208.0,
      "step": 31420
    },
    {
      "entropy": 5.549488115310669,
      "epoch": 2.8505986937590713,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041921865676237605,
      "loss": 4.913,
      "mean_token_accuracy": 0.21836962252855302,
      "num_tokens": 65220954.0,
      "step": 31425
    },
    {
      "entropy": 5.588891124725341,
      "epoch": 2.8510522496371555,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041919381963456437,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21905021965503693,
      "num_tokens": 65230870.0,
      "step": 31430
    },
    {
      "entropy": 5.54392557144165,
      "epoch": 2.8515058055152394,
      "grad_norm": 1.25,
      "learning_rate": 0.0004191689795247877,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22279935479164123,
      "num_tokens": 65241913.0,
      "step": 31435
    },
    {
      "entropy": 5.588877773284912,
      "epoch": 2.8519593613933236,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004191441364335598,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2176218494772911,
      "num_tokens": 65252513.0,
      "step": 31440
    },
    {
      "entropy": 5.489598035812378,
      "epoch": 2.852412917271408,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041911929036139457,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2195203870534897,
      "num_tokens": 65262695.0,
      "step": 31445
    },
    {
      "entropy": 5.581487417221069,
      "epoch": 2.852866473149492,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041909444130880553,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21721937358379365,
      "num_tokens": 65274131.0,
      "step": 31450
    },
    {
      "entropy": 5.552785587310791,
      "epoch": 2.853320029027576,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041906958927630673,
      "loss": 4.9178,
      "mean_token_accuracy": 0.21341437697410584,
      "num_tokens": 65285080.0,
      "step": 31455
    },
    {
      "entropy": 5.67452974319458,
      "epoch": 2.8537735849056602,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004190447342644122,
      "loss": 5.0955,
      "mean_token_accuracy": 0.21516249179840088,
      "num_tokens": 65296512.0,
      "step": 31460
    },
    {
      "entropy": 5.5065295696258545,
      "epoch": 2.8542271407837445,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004190198762736358,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22480050176382066,
      "num_tokens": 65306054.0,
      "step": 31465
    },
    {
      "entropy": 5.436642837524414,
      "epoch": 2.8546806966618288,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004189950153044917,
      "loss": 4.9136,
      "mean_token_accuracy": 0.21854021549224853,
      "num_tokens": 65315046.0,
      "step": 31470
    },
    {
      "entropy": 5.498241710662842,
      "epoch": 2.855134252539913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004189701513574939,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2167905956506729,
      "num_tokens": 65325482.0,
      "step": 31475
    },
    {
      "entropy": 5.51598162651062,
      "epoch": 2.8555878084179973,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004189452844331567,
      "loss": 4.8527,
      "mean_token_accuracy": 0.21807391345500945,
      "num_tokens": 65335518.0,
      "step": 31480
    },
    {
      "entropy": 5.544434928894043,
      "epoch": 2.856041364296081,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041892041453199443,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21861279010772705,
      "num_tokens": 65346079.0,
      "step": 31485
    },
    {
      "entropy": 5.504207706451416,
      "epoch": 2.8564949201741654,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004188955416545212,
      "loss": 4.8813,
      "mean_token_accuracy": 0.21876951307058334,
      "num_tokens": 65356719.0,
      "step": 31490
    },
    {
      "entropy": 5.554988718032837,
      "epoch": 2.8569484760522497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004188706658012516,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21187306195497513,
      "num_tokens": 65367877.0,
      "step": 31495
    },
    {
      "entropy": 5.57737922668457,
      "epoch": 2.857402031930334,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041884578697269994,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20789577662944794,
      "num_tokens": 65379624.0,
      "step": 31500
    },
    {
      "entropy": 5.576460647583008,
      "epoch": 2.8578555878084178,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004188209051693807,
      "loss": 4.9871,
      "mean_token_accuracy": 0.21411170363426207,
      "num_tokens": 65390417.0,
      "step": 31505
    },
    {
      "entropy": 5.459558582305908,
      "epoch": 2.858309143686502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041879602039180856,
      "loss": 4.751,
      "mean_token_accuracy": 0.22666174471378325,
      "num_tokens": 65401681.0,
      "step": 31510
    },
    {
      "entropy": 5.515451097488404,
      "epoch": 2.8587626995645863,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041877113264049805,
      "loss": 4.8612,
      "mean_token_accuracy": 0.22111100405454637,
      "num_tokens": 65412520.0,
      "step": 31515
    },
    {
      "entropy": 5.550540208816528,
      "epoch": 2.8592162554426706,
      "grad_norm": 1.25,
      "learning_rate": 0.00041874624191596383,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21934369802474976,
      "num_tokens": 65423201.0,
      "step": 31520
    },
    {
      "entropy": 5.50769362449646,
      "epoch": 2.859669811320755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004187213482187207,
      "loss": 4.934,
      "mean_token_accuracy": 0.2097564771771431,
      "num_tokens": 65432778.0,
      "step": 31525
    },
    {
      "entropy": 5.60222544670105,
      "epoch": 2.860123367198839,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004186964515492835,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2192388266324997,
      "num_tokens": 65442323.0,
      "step": 31530
    },
    {
      "entropy": 5.540174341201782,
      "epoch": 2.8605769230769234,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004186715519081671,
      "loss": 4.927,
      "mean_token_accuracy": 0.2159638896584511,
      "num_tokens": 65451980.0,
      "step": 31535
    },
    {
      "entropy": 5.441647148132324,
      "epoch": 2.861030478955007,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004186466492958864,
      "loss": 4.867,
      "mean_token_accuracy": 0.2177277013659477,
      "num_tokens": 65463359.0,
      "step": 31540
    },
    {
      "entropy": 5.552570915222168,
      "epoch": 2.8614840348330914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004186217437129562,
      "loss": 4.9074,
      "mean_token_accuracy": 0.21424112766981124,
      "num_tokens": 65474155.0,
      "step": 31545
    },
    {
      "entropy": 5.585643911361695,
      "epoch": 2.8619375907111757,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041859683515989196,
      "loss": 4.8557,
      "mean_token_accuracy": 0.21864478588104247,
      "num_tokens": 65484884.0,
      "step": 31550
    },
    {
      "entropy": 5.567428016662598,
      "epoch": 2.8623911465892595,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041857192363720845,
      "loss": 4.9404,
      "mean_token_accuracy": 0.2179499313235283,
      "num_tokens": 65494380.0,
      "step": 31555
    },
    {
      "entropy": 5.518616390228272,
      "epoch": 2.862844702467344,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004185470091454211,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21639350354671477,
      "num_tokens": 65505735.0,
      "step": 31560
    },
    {
      "entropy": 5.567360496520996,
      "epoch": 2.863298258345428,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004185220916850449,
      "loss": 4.8684,
      "mean_token_accuracy": 0.21369979977607728,
      "num_tokens": 65515956.0,
      "step": 31565
    },
    {
      "entropy": 5.5191755294799805,
      "epoch": 2.8637518142235123,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041849717125659537,
      "loss": 4.8524,
      "mean_token_accuracy": 0.2151741310954094,
      "num_tokens": 65526369.0,
      "step": 31570
    },
    {
      "entropy": 5.5277210712432865,
      "epoch": 2.8642053701015966,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004184722478605877,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22305546402931214,
      "num_tokens": 65536032.0,
      "step": 31575
    },
    {
      "entropy": 5.495082521438599,
      "epoch": 2.864658925979681,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004184473214975374,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2094807043671608,
      "num_tokens": 65546121.0,
      "step": 31580
    },
    {
      "entropy": 5.507816171646118,
      "epoch": 2.865112481857765,
      "grad_norm": 1.125,
      "learning_rate": 0.00041842239216795994,
      "loss": 4.8926,
      "mean_token_accuracy": 0.2115117311477661,
      "num_tokens": 65555915.0,
      "step": 31585
    },
    {
      "entropy": 5.507788801193238,
      "epoch": 2.865566037735849,
      "grad_norm": 1.125,
      "learning_rate": 0.00041839745987237086,
      "loss": 4.9149,
      "mean_token_accuracy": 0.20928386896848677,
      "num_tokens": 65566267.0,
      "step": 31590
    },
    {
      "entropy": 5.553118371963501,
      "epoch": 2.866019593613933,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004183725246112857,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2217654287815094,
      "num_tokens": 65577126.0,
      "step": 31595
    },
    {
      "entropy": 5.595052194595337,
      "epoch": 2.8664731494920175,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041834758638522034,
      "loss": 5.0191,
      "mean_token_accuracy": 0.20554105341434478,
      "num_tokens": 65587158.0,
      "step": 31600
    },
    {
      "entropy": 5.544302606582642,
      "epoch": 2.8669267053701017,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041832264519469026,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2119647666811943,
      "num_tokens": 65597720.0,
      "step": 31605
    },
    {
      "entropy": 5.612184286117554,
      "epoch": 2.8673802612481856,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041829770104021143,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21834521889686584,
      "num_tokens": 65607409.0,
      "step": 31610
    },
    {
      "entropy": 5.575307130813599,
      "epoch": 2.86783381712627,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041827275392229956,
      "loss": 4.871,
      "mean_token_accuracy": 0.22842340767383576,
      "num_tokens": 65616785.0,
      "step": 31615
    },
    {
      "entropy": 5.544488906860352,
      "epoch": 2.868287373004354,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004182478038414706,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21097181737422943,
      "num_tokens": 65625794.0,
      "step": 31620
    },
    {
      "entropy": 5.447157526016236,
      "epoch": 2.8687409288824384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041822285079824056,
      "loss": 4.7927,
      "mean_token_accuracy": 0.2222556620836258,
      "num_tokens": 65635789.0,
      "step": 31625
    },
    {
      "entropy": 5.579034233093262,
      "epoch": 2.8691944847605226,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004181978947931256,
      "loss": 5.026,
      "mean_token_accuracy": 0.21067075580358505,
      "num_tokens": 65646266.0,
      "step": 31630
    },
    {
      "entropy": 5.585314702987671,
      "epoch": 2.869648040638607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004181729358266415,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21557358354330064,
      "num_tokens": 65656538.0,
      "step": 31635
    },
    {
      "entropy": 5.596712923049926,
      "epoch": 2.8701015965166907,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041814797389930474,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21700470745563508,
      "num_tokens": 65666926.0,
      "step": 31640
    },
    {
      "entropy": 5.551916790008545,
      "epoch": 2.870555152394775,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004181230090116313,
      "loss": 4.9065,
      "mean_token_accuracy": 0.21022246927022933,
      "num_tokens": 65677219.0,
      "step": 31645
    },
    {
      "entropy": 5.5526855945587155,
      "epoch": 2.8710087082728593,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041809804116413763,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21329140663146973,
      "num_tokens": 65688281.0,
      "step": 31650
    },
    {
      "entropy": 5.661717510223388,
      "epoch": 2.8714622641509435,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041807307035734,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20881217569112778,
      "num_tokens": 65698815.0,
      "step": 31655
    },
    {
      "entropy": 5.567882919311524,
      "epoch": 2.8719158200290273,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004180480965917548,
      "loss": 4.9382,
      "mean_token_accuracy": 0.217138934135437,
      "num_tokens": 65709481.0,
      "step": 31660
    },
    {
      "entropy": 5.608654451370239,
      "epoch": 2.8723693759071116,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041802311986789854,
      "loss": 4.9175,
      "mean_token_accuracy": 0.20534139275550842,
      "num_tokens": 65720561.0,
      "step": 31665
    },
    {
      "entropy": 5.581717586517334,
      "epoch": 2.872822931785196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041799814018628765,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20782429426908494,
      "num_tokens": 65730384.0,
      "step": 31670
    },
    {
      "entropy": 5.556689929962158,
      "epoch": 2.87327648766328,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041797315754743883,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21964192986488343,
      "num_tokens": 65740711.0,
      "step": 31675
    },
    {
      "entropy": 5.659464502334595,
      "epoch": 2.8737300435413644,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041794817195186867,
      "loss": 5.0292,
      "mean_token_accuracy": 0.21126116961240768,
      "num_tokens": 65751033.0,
      "step": 31680
    },
    {
      "entropy": 5.611872816085816,
      "epoch": 2.8741835994194487,
      "grad_norm": 1.25,
      "learning_rate": 0.0004179231834000939,
      "loss": 5.0021,
      "mean_token_accuracy": 0.20993344038724898,
      "num_tokens": 65761500.0,
      "step": 31685
    },
    {
      "entropy": 5.458570289611816,
      "epoch": 2.8746371552975325,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041789819189263117,
      "loss": 4.8487,
      "mean_token_accuracy": 0.21971217691898345,
      "num_tokens": 65771095.0,
      "step": 31690
    },
    {
      "entropy": 5.523027086257935,
      "epoch": 2.8750907111756168,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041787319742999756,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21993601620197295,
      "num_tokens": 65782058.0,
      "step": 31695
    },
    {
      "entropy": 5.558678436279297,
      "epoch": 2.875544267053701,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004178482000127098,
      "loss": 4.9462,
      "mean_token_accuracy": 0.21851180493831635,
      "num_tokens": 65791710.0,
      "step": 31700
    },
    {
      "entropy": 5.586021614074707,
      "epoch": 2.8759978229317853,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004178231996412848,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20874281823635102,
      "num_tokens": 65802180.0,
      "step": 31705
    },
    {
      "entropy": 5.5061994075775145,
      "epoch": 2.876451378809869,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004177981963162396,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2269972786307335,
      "num_tokens": 65812998.0,
      "step": 31710
    },
    {
      "entropy": 5.558642053604126,
      "epoch": 2.8769049346879534,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004177731900380914,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2155836045742035,
      "num_tokens": 65824210.0,
      "step": 31715
    },
    {
      "entropy": 5.501532411575317,
      "epoch": 2.8773584905660377,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004177481808073572,
      "loss": 4.8796,
      "mean_token_accuracy": 0.21981393694877624,
      "num_tokens": 65834322.0,
      "step": 31720
    },
    {
      "entropy": 5.526997041702271,
      "epoch": 2.877812046444122,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041772316862455426,
      "loss": 4.8358,
      "mean_token_accuracy": 0.22797203958034515,
      "num_tokens": 65845347.0,
      "step": 31725
    },
    {
      "entropy": 5.578409194946289,
      "epoch": 2.878265602322206,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041769815349019986,
      "loss": 4.9498,
      "mean_token_accuracy": 0.2114505410194397,
      "num_tokens": 65855616.0,
      "step": 31730
    },
    {
      "entropy": 5.575336074829101,
      "epoch": 2.8787191582002905,
      "grad_norm": 1.59375,
      "learning_rate": 0.00041767313540481125,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21302270293235778,
      "num_tokens": 65865542.0,
      "step": 31735
    },
    {
      "entropy": 5.498986530303955,
      "epoch": 2.8791727140783747,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004176481143689058,
      "loss": 4.8995,
      "mean_token_accuracy": 0.21235269755125047,
      "num_tokens": 65874478.0,
      "step": 31740
    },
    {
      "entropy": 5.580822515487671,
      "epoch": 2.8796262699564585,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004176230903830011,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2134009853005409,
      "num_tokens": 65884609.0,
      "step": 31745
    },
    {
      "entropy": 5.552547597885132,
      "epoch": 2.880079825834543,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004175980634476145,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2203726813197136,
      "num_tokens": 65894486.0,
      "step": 31750
    },
    {
      "entropy": 5.498010540008545,
      "epoch": 2.880533381712627,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004175730335632635,
      "loss": 4.8733,
      "mean_token_accuracy": 0.2206169530749321,
      "num_tokens": 65904643.0,
      "step": 31755
    },
    {
      "entropy": 5.41986837387085,
      "epoch": 2.8809869375907113,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041754800073046593,
      "loss": 4.882,
      "mean_token_accuracy": 0.22199921011924745,
      "num_tokens": 65914107.0,
      "step": 31760
    },
    {
      "entropy": 5.489398670196533,
      "epoch": 2.881440493468795,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004175229649497394,
      "loss": 4.8405,
      "mean_token_accuracy": 0.21779906004667282,
      "num_tokens": 65925112.0,
      "step": 31765
    },
    {
      "entropy": 5.589184904098511,
      "epoch": 2.8818940493468794,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004174979262216016,
      "loss": 4.9311,
      "mean_token_accuracy": 0.21531084328889846,
      "num_tokens": 65934624.0,
      "step": 31770
    },
    {
      "entropy": 5.605479621887207,
      "epoch": 2.8823476052249637,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004174728845465704,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21430217921733857,
      "num_tokens": 65945951.0,
      "step": 31775
    },
    {
      "entropy": 5.473323726654053,
      "epoch": 2.882801161103048,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004174478399251636,
      "loss": 4.8261,
      "mean_token_accuracy": 0.22606817930936812,
      "num_tokens": 65956871.0,
      "step": 31780
    },
    {
      "entropy": 5.61882586479187,
      "epoch": 2.8832547169811322,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004174227923578992,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2055545151233673,
      "num_tokens": 65967814.0,
      "step": 31785
    },
    {
      "entropy": 5.548902416229248,
      "epoch": 2.8837082728592165,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004173977418452952,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21030566245317459,
      "num_tokens": 65978118.0,
      "step": 31790
    },
    {
      "entropy": 5.522923994064331,
      "epoch": 2.8841618287373003,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041737268838786955,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22173561304807662,
      "num_tokens": 65988240.0,
      "step": 31795
    },
    {
      "entropy": 5.540399980545044,
      "epoch": 2.8846153846153846,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004173476319861404,
      "loss": 4.8775,
      "mean_token_accuracy": 0.2168329820036888,
      "num_tokens": 65998126.0,
      "step": 31800
    },
    {
      "entropy": 5.5352075576782225,
      "epoch": 2.885068940493469,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004173225726406261,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22301522344350816,
      "num_tokens": 66009118.0,
      "step": 31805
    },
    {
      "entropy": 5.594109153747558,
      "epoch": 2.885522496371553,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041729751035184457,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21014766842126847,
      "num_tokens": 66019340.0,
      "step": 31810
    },
    {
      "entropy": 5.542562007904053,
      "epoch": 2.885976052249637,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041727244512031425,
      "loss": 4.8907,
      "mean_token_accuracy": 0.21946422308683394,
      "num_tokens": 66029784.0,
      "step": 31815
    },
    {
      "entropy": 5.606982946395874,
      "epoch": 2.886429608127721,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004172473769465536,
      "loss": 5.0363,
      "mean_token_accuracy": 0.2029428318142891,
      "num_tokens": 66039987.0,
      "step": 31820
    },
    {
      "entropy": 5.562028884887695,
      "epoch": 2.8868831640058055,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041722230583108097,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21664694994688033,
      "num_tokens": 66049150.0,
      "step": 31825
    },
    {
      "entropy": 5.608209753036499,
      "epoch": 2.8873367198838897,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041719723177441467,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21542547941207885,
      "num_tokens": 66060719.0,
      "step": 31830
    },
    {
      "entropy": 5.620509481430053,
      "epoch": 2.887790275761974,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004171721547770735,
      "loss": 4.96,
      "mean_token_accuracy": 0.2112122505903244,
      "num_tokens": 66071126.0,
      "step": 31835
    },
    {
      "entropy": 5.571649789810181,
      "epoch": 2.8882438316400583,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004171470748395759,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22009020149707795,
      "num_tokens": 66080928.0,
      "step": 31840
    },
    {
      "entropy": 5.571517753601074,
      "epoch": 2.888697387518142,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004171219919624406,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2110880881547928,
      "num_tokens": 66090900.0,
      "step": 31845
    },
    {
      "entropy": 5.545731449127198,
      "epoch": 2.8891509433962264,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041709690614618624,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21194626092910768,
      "num_tokens": 66100720.0,
      "step": 31850
    },
    {
      "entropy": 5.527884149551392,
      "epoch": 2.8896044992743106,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041707181739133163,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21062547862529754,
      "num_tokens": 66110780.0,
      "step": 31855
    },
    {
      "entropy": 5.550134372711182,
      "epoch": 2.890058055152395,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004170467256983957,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22698114067316055,
      "num_tokens": 66120976.0,
      "step": 31860
    },
    {
      "entropy": 5.587691593170166,
      "epoch": 2.8905116110304787,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004170216310678973,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2190371111035347,
      "num_tokens": 66130221.0,
      "step": 31865
    },
    {
      "entropy": 5.518986940383911,
      "epoch": 2.890965166908563,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041699653350035527,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22471992820501327,
      "num_tokens": 66140060.0,
      "step": 31870
    },
    {
      "entropy": 5.514130544662476,
      "epoch": 2.8914187227866472,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041697143299628886,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22354376763105394,
      "num_tokens": 66151640.0,
      "step": 31875
    },
    {
      "entropy": 5.499039268493652,
      "epoch": 2.8918722786647315,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041694632955621697,
      "loss": 4.8781,
      "mean_token_accuracy": 0.2253771498799324,
      "num_tokens": 66162923.0,
      "step": 31880
    },
    {
      "entropy": 5.570623445510864,
      "epoch": 2.8923258345428158,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041692122318065885,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21463344246149063,
      "num_tokens": 66173108.0,
      "step": 31885
    },
    {
      "entropy": 5.540338802337646,
      "epoch": 2.8927793904209,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041689611387013363,
      "loss": 4.8565,
      "mean_token_accuracy": 0.2213662877678871,
      "num_tokens": 66185280.0,
      "step": 31890
    },
    {
      "entropy": 5.526933431625366,
      "epoch": 2.8932329462989843,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004168710016251606,
      "loss": 4.883,
      "mean_token_accuracy": 0.21428462117910385,
      "num_tokens": 66196020.0,
      "step": 31895
    },
    {
      "entropy": 5.51864185333252,
      "epoch": 2.893686502177068,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004168458864462592,
      "loss": 4.8344,
      "mean_token_accuracy": 0.2194029688835144,
      "num_tokens": 66205753.0,
      "step": 31900
    },
    {
      "entropy": 5.576111841201782,
      "epoch": 2.8941400580551524,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041682076833394864,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21293427795171738,
      "num_tokens": 66218415.0,
      "step": 31905
    },
    {
      "entropy": 5.567892026901245,
      "epoch": 2.8945936139332367,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041679564728874845,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21444406658411025,
      "num_tokens": 66228090.0,
      "step": 31910
    },
    {
      "entropy": 5.486314344406128,
      "epoch": 2.8950471698113205,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004167705233111782,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22523657977581024,
      "num_tokens": 66237587.0,
      "step": 31915
    },
    {
      "entropy": 5.532775974273681,
      "epoch": 2.8955007256894048,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041674539640175733,
      "loss": 4.8715,
      "mean_token_accuracy": 0.21704545468091965,
      "num_tokens": 66248162.0,
      "step": 31920
    },
    {
      "entropy": 5.501201152801514,
      "epoch": 2.895954281567489,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004167202665610055,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22793399542570114,
      "num_tokens": 66258370.0,
      "step": 31925
    },
    {
      "entropy": 5.524153852462769,
      "epoch": 2.8964078374455733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041669513378944254,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2114126369357109,
      "num_tokens": 66269267.0,
      "step": 31930
    },
    {
      "entropy": 5.5975542068481445,
      "epoch": 2.8968613933236576,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041666999808758815,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21983028948307037,
      "num_tokens": 66280077.0,
      "step": 31935
    },
    {
      "entropy": 5.549991130828857,
      "epoch": 2.897314949201742,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000416644859455962,
      "loss": 4.8562,
      "mean_token_accuracy": 0.21780994087457656,
      "num_tokens": 66291678.0,
      "step": 31940
    },
    {
      "entropy": 5.516115808486939,
      "epoch": 2.897768505079826,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004166197178950841,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21561197340488433,
      "num_tokens": 66302046.0,
      "step": 31945
    },
    {
      "entropy": 5.521388673782349,
      "epoch": 2.89822206095791,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041659457340547436,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2168835759162903,
      "num_tokens": 66311888.0,
      "step": 31950
    },
    {
      "entropy": 5.487433338165284,
      "epoch": 2.898675616835994,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041656942598765275,
      "loss": 4.8676,
      "mean_token_accuracy": 0.21865642815828323,
      "num_tokens": 66321507.0,
      "step": 31955
    },
    {
      "entropy": 5.490264463424682,
      "epoch": 2.8991291727140784,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041654427564213924,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22566469460725785,
      "num_tokens": 66331458.0,
      "step": 31960
    },
    {
      "entropy": 5.539827585220337,
      "epoch": 2.8995827285921627,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041651912236945416,
      "loss": 4.8912,
      "mean_token_accuracy": 0.21899575144052505,
      "num_tokens": 66340875.0,
      "step": 31965
    },
    {
      "entropy": 5.559274101257325,
      "epoch": 2.9000362844702465,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041649396617011753,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2099599704146385,
      "num_tokens": 66350784.0,
      "step": 31970
    },
    {
      "entropy": 5.576534700393677,
      "epoch": 2.900489840348331,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004164688070446496,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21176678091287612,
      "num_tokens": 66361591.0,
      "step": 31975
    },
    {
      "entropy": 5.533900165557862,
      "epoch": 2.900943396226415,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004164436449935707,
      "loss": 4.8646,
      "mean_token_accuracy": 0.21987906396389006,
      "num_tokens": 66371540.0,
      "step": 31980
    },
    {
      "entropy": 5.542740297317505,
      "epoch": 2.9013969521044993,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041641848001740126,
      "loss": 4.9201,
      "mean_token_accuracy": 0.21813347637653352,
      "num_tokens": 66382053.0,
      "step": 31985
    },
    {
      "entropy": 5.54427695274353,
      "epoch": 2.9018505079825836,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004163933121166615,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22122047394514083,
      "num_tokens": 66392563.0,
      "step": 31990
    },
    {
      "entropy": 5.543047714233398,
      "epoch": 2.902304063860668,
      "grad_norm": 1.078125,
      "learning_rate": 0.000416368141291872,
      "loss": 4.85,
      "mean_token_accuracy": 0.22260381579399108,
      "num_tokens": 66403582.0,
      "step": 31995
    },
    {
      "entropy": 5.493227434158325,
      "epoch": 2.9027576197387517,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041634296754355346,
      "loss": 4.8329,
      "mean_token_accuracy": 0.21896252930164337,
      "num_tokens": 66413713.0,
      "step": 32000
    },
    {
      "entropy": 5.5138264179229735,
      "epoch": 2.903211175616836,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004163177908722262,
      "loss": 4.9026,
      "mean_token_accuracy": 0.2192167341709137,
      "num_tokens": 66424802.0,
      "step": 32005
    },
    {
      "entropy": 5.52631368637085,
      "epoch": 2.90366473149492,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041629261127841105,
      "loss": 4.8615,
      "mean_token_accuracy": 0.2189914897084236,
      "num_tokens": 66434716.0,
      "step": 32010
    },
    {
      "entropy": 5.531491470336914,
      "epoch": 2.9041182873730045,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041626742876262876,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21337605267763138,
      "num_tokens": 66445725.0,
      "step": 32015
    },
    {
      "entropy": 5.556710815429687,
      "epoch": 2.9045718432510883,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041624224332539997,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2088825985789299,
      "num_tokens": 66457492.0,
      "step": 32020
    },
    {
      "entropy": 5.558711385726928,
      "epoch": 2.9050253991291726,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004162170549672455,
      "loss": 4.8769,
      "mean_token_accuracy": 0.21862435042858125,
      "num_tokens": 66467490.0,
      "step": 32025
    },
    {
      "entropy": 5.57815260887146,
      "epoch": 2.905478955007257,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004161918636886865,
      "loss": 4.987,
      "mean_token_accuracy": 0.21289435178041458,
      "num_tokens": 66478160.0,
      "step": 32030
    },
    {
      "entropy": 5.559884071350098,
      "epoch": 2.905932510885341,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004161666694902437,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21486809551715852,
      "num_tokens": 66488181.0,
      "step": 32035
    },
    {
      "entropy": 5.556135940551758,
      "epoch": 2.9063860667634254,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004161414723724383,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21877287179231644,
      "num_tokens": 66498507.0,
      "step": 32040
    },
    {
      "entropy": 5.476687574386597,
      "epoch": 2.9068396226415096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004161162723357912,
      "loss": 4.8398,
      "mean_token_accuracy": 0.22045327425003053,
      "num_tokens": 66507752.0,
      "step": 32045
    },
    {
      "entropy": 5.497104358673096,
      "epoch": 2.9072931785195935,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004160910693808236,
      "loss": 4.9184,
      "mean_token_accuracy": 0.208297298848629,
      "num_tokens": 66519360.0,
      "step": 32050
    },
    {
      "entropy": 5.579237079620361,
      "epoch": 2.9077467343976777,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041606586350805686,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21351252645254135,
      "num_tokens": 66529790.0,
      "step": 32055
    },
    {
      "entropy": 5.530319452285767,
      "epoch": 2.908200290275762,
      "grad_norm": 1.15625,
      "learning_rate": 0.000416040654718012,
      "loss": 4.8531,
      "mean_token_accuracy": 0.2273558884859085,
      "num_tokens": 66539617.0,
      "step": 32060
    },
    {
      "entropy": 5.5499505519866945,
      "epoch": 2.9086538461538463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041601544301121045,
      "loss": 4.8714,
      "mean_token_accuracy": 0.22175043672323227,
      "num_tokens": 66549861.0,
      "step": 32065
    },
    {
      "entropy": 5.614412021636963,
      "epoch": 2.90910740203193,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004159902283881737,
      "loss": 5.0051,
      "mean_token_accuracy": 0.21870703995227814,
      "num_tokens": 66560388.0,
      "step": 32070
    },
    {
      "entropy": 5.601671934127808,
      "epoch": 2.9095609579100143,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000415965010849423,
      "loss": 4.97,
      "mean_token_accuracy": 0.2153163507580757,
      "num_tokens": 66570795.0,
      "step": 32075
    },
    {
      "entropy": 5.594749069213867,
      "epoch": 2.9100145137880986,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041593979039548,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21092817485332488,
      "num_tokens": 66580268.0,
      "step": 32080
    },
    {
      "entropy": 5.58156909942627,
      "epoch": 2.910468069666183,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004159145670268662,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21238549798727036,
      "num_tokens": 66591192.0,
      "step": 32085
    },
    {
      "entropy": 5.568034076690674,
      "epoch": 2.910921625544267,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004158893407441033,
      "loss": 4.9856,
      "mean_token_accuracy": 0.2187513068318367,
      "num_tokens": 66600768.0,
      "step": 32090
    },
    {
      "entropy": 5.647961902618408,
      "epoch": 2.9113751814223514,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041586411154771285,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21644292175769805,
      "num_tokens": 66612207.0,
      "step": 32095
    },
    {
      "entropy": 5.521655416488647,
      "epoch": 2.9118287373004357,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041583887943821667,
      "loss": 4.8482,
      "mean_token_accuracy": 0.2201209157705307,
      "num_tokens": 66622175.0,
      "step": 32100
    },
    {
      "entropy": 5.505579662322998,
      "epoch": 2.9122822931785195,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004158136444161367,
      "loss": 4.8862,
      "mean_token_accuracy": 0.21468835473060607,
      "num_tokens": 66633936.0,
      "step": 32105
    },
    {
      "entropy": 5.569595241546631,
      "epoch": 2.9127358490566038,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041578840648199455,
      "loss": 4.9568,
      "mean_token_accuracy": 0.20726978331804274,
      "num_tokens": 66643502.0,
      "step": 32110
    },
    {
      "entropy": 5.544066095352173,
      "epoch": 2.913189404934688,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004157631656363124,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21949195265769958,
      "num_tokens": 66654890.0,
      "step": 32115
    },
    {
      "entropy": 5.589392137527466,
      "epoch": 2.9136429608127723,
      "grad_norm": 1.0625,
      "learning_rate": 0.000415737921879612,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20125688314437867,
      "num_tokens": 66667037.0,
      "step": 32120
    },
    {
      "entropy": 5.5436938285827635,
      "epoch": 2.914096516690856,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041571267521241554,
      "loss": 4.879,
      "mean_token_accuracy": 0.2174460083246231,
      "num_tokens": 66677291.0,
      "step": 32125
    },
    {
      "entropy": 5.486691331863403,
      "epoch": 2.9145500725689404,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004156874256352451,
      "loss": 4.8622,
      "mean_token_accuracy": 0.21483613252639772,
      "num_tokens": 66687421.0,
      "step": 32130
    },
    {
      "entropy": 5.600220966339111,
      "epoch": 2.9150036284470247,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004156621731486228,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2091139778494835,
      "num_tokens": 66697190.0,
      "step": 32135
    },
    {
      "entropy": 5.586568164825439,
      "epoch": 2.915457184325109,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000415636917753071,
      "loss": 4.983,
      "mean_token_accuracy": 0.21069155186414718,
      "num_tokens": 66707255.0,
      "step": 32140
    },
    {
      "entropy": 5.488911867141724,
      "epoch": 2.915910740203193,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004156116594491119,
      "loss": 4.8321,
      "mean_token_accuracy": 0.21835922300815583,
      "num_tokens": 66716451.0,
      "step": 32145
    },
    {
      "entropy": 5.501500654220581,
      "epoch": 2.9163642960812775,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004155863982372678,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22173197120428084,
      "num_tokens": 66727157.0,
      "step": 32150
    },
    {
      "entropy": 5.598024225234985,
      "epoch": 2.9168178519593613,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041556113411806117,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2129424676299095,
      "num_tokens": 66737523.0,
      "step": 32155
    },
    {
      "entropy": 5.591394519805908,
      "epoch": 2.9172714078374455,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041553586709201453,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2162012815475464,
      "num_tokens": 66748183.0,
      "step": 32160
    },
    {
      "entropy": 5.570503282546997,
      "epoch": 2.91772496371553,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004155105971596503,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22067442536354065,
      "num_tokens": 66758794.0,
      "step": 32165
    },
    {
      "entropy": 5.592117547988892,
      "epoch": 2.918178519593614,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041548532432149105,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21960476487874986,
      "num_tokens": 66769916.0,
      "step": 32170
    },
    {
      "entropy": 5.57194037437439,
      "epoch": 2.918632075471698,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004154600485780595,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22378370761871338,
      "num_tokens": 66779658.0,
      "step": 32175
    },
    {
      "entropy": 5.501144027709961,
      "epoch": 2.919085631349782,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041543476992987843,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22131458371877671,
      "num_tokens": 66788909.0,
      "step": 32180
    },
    {
      "entropy": 5.590155696868896,
      "epoch": 2.9195391872278664,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041540948837747056,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21467978060245513,
      "num_tokens": 66799519.0,
      "step": 32185
    },
    {
      "entropy": 5.501236915588379,
      "epoch": 2.9199927431059507,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004153842039213586,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22042181938886643,
      "num_tokens": 66808915.0,
      "step": 32190
    },
    {
      "entropy": 5.564164686203003,
      "epoch": 2.920446298984035,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004153589165620656,
      "loss": 4.967,
      "mean_token_accuracy": 0.21259564757347107,
      "num_tokens": 66818819.0,
      "step": 32195
    },
    {
      "entropy": 5.4631490230560305,
      "epoch": 2.9208998548621192,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041533362630011435,
      "loss": 4.8019,
      "mean_token_accuracy": 0.22978342175483704,
      "num_tokens": 66829274.0,
      "step": 32200
    },
    {
      "entropy": 5.618209600448608,
      "epoch": 2.921353410740203,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000415308333136028,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22303779125213624,
      "num_tokens": 66839659.0,
      "step": 32205
    },
    {
      "entropy": 5.574804258346558,
      "epoch": 2.9218069666182873,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00041528303707032957,
      "loss": 4.9702,
      "mean_token_accuracy": 0.21683358550071716,
      "num_tokens": 66850415.0,
      "step": 32210
    },
    {
      "entropy": 5.541540813446045,
      "epoch": 2.9222605224963716,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004152577381035422,
      "loss": 4.8689,
      "mean_token_accuracy": 0.2276713564991951,
      "num_tokens": 66860663.0,
      "step": 32215
    },
    {
      "entropy": 5.609417819976807,
      "epoch": 2.922714078374456,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041523243623618907,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2171732172369957,
      "num_tokens": 66871473.0,
      "step": 32220
    },
    {
      "entropy": 5.50028657913208,
      "epoch": 2.9231676342525397,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041520713146879333,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2147811993956566,
      "num_tokens": 66881609.0,
      "step": 32225
    },
    {
      "entropy": 5.546068620681763,
      "epoch": 2.923621190130624,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041518182380187853,
      "loss": 4.9226,
      "mean_token_accuracy": 0.2135738417506218,
      "num_tokens": 66891780.0,
      "step": 32230
    },
    {
      "entropy": 5.606244850158691,
      "epoch": 2.924074746008708,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041515651323596774,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2069510981440544,
      "num_tokens": 66902251.0,
      "step": 32235
    },
    {
      "entropy": 5.585474920272827,
      "epoch": 2.9245283018867925,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041513119977158464,
      "loss": 4.9414,
      "mean_token_accuracy": 0.2141524687409401,
      "num_tokens": 66912538.0,
      "step": 32240
    },
    {
      "entropy": 5.544709396362305,
      "epoch": 2.9249818577648767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004151058834092526,
      "loss": 4.9026,
      "mean_token_accuracy": 0.2185381069779396,
      "num_tokens": 66922648.0,
      "step": 32245
    },
    {
      "entropy": 5.512761020660401,
      "epoch": 2.925435413642961,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041508056414949515,
      "loss": 4.8318,
      "mean_token_accuracy": 0.2172267198562622,
      "num_tokens": 66933115.0,
      "step": 32250
    },
    {
      "entropy": 5.582364511489868,
      "epoch": 2.9258889695210453,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000415055241992836,
      "loss": 5.0544,
      "mean_token_accuracy": 0.20653746724128724,
      "num_tokens": 66943727.0,
      "step": 32255
    },
    {
      "entropy": 5.557615661621094,
      "epoch": 2.926342525399129,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004150299169397987,
      "loss": 4.8749,
      "mean_token_accuracy": 0.21730950027704238,
      "num_tokens": 66953488.0,
      "step": 32260
    },
    {
      "entropy": 5.624796056747437,
      "epoch": 2.9267960812772134,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004150045889909071,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2051413595676422,
      "num_tokens": 66964586.0,
      "step": 32265
    },
    {
      "entropy": 5.552080202102661,
      "epoch": 2.9272496371552976,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004149792581466848,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20885419845581055,
      "num_tokens": 66975032.0,
      "step": 32270
    },
    {
      "entropy": 5.574852657318115,
      "epoch": 2.9277031930333814,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041495392440765586,
      "loss": 4.8919,
      "mean_token_accuracy": 0.2216252401471138,
      "num_tokens": 66984910.0,
      "step": 32275
    },
    {
      "entropy": 5.541440820693969,
      "epoch": 2.9281567489114657,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041492858777434407,
      "loss": 4.7116,
      "mean_token_accuracy": 0.23320505321025847,
      "num_tokens": 66995238.0,
      "step": 32280
    },
    {
      "entropy": 5.520256280899048,
      "epoch": 2.92861030478955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004149032482472734,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22672948837280274,
      "num_tokens": 67005596.0,
      "step": 32285
    },
    {
      "entropy": 5.632413101196289,
      "epoch": 2.9290638606676342,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041487790582696803,
      "loss": 5.0736,
      "mean_token_accuracy": 0.20305684208869934,
      "num_tokens": 67017312.0,
      "step": 32290
    },
    {
      "entropy": 5.5625011920928955,
      "epoch": 2.9295174165457185,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004148525605139518,
      "loss": 4.8707,
      "mean_token_accuracy": 0.2237891972064972,
      "num_tokens": 67026692.0,
      "step": 32295
    },
    {
      "entropy": 5.503364038467407,
      "epoch": 2.9299709724238028,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041482721230874897,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2196035847067833,
      "num_tokens": 67036801.0,
      "step": 32300
    },
    {
      "entropy": 5.498670291900635,
      "epoch": 2.930424528301887,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041480186121188383,
      "loss": 4.8665,
      "mean_token_accuracy": 0.22340067923069,
      "num_tokens": 67047262.0,
      "step": 32305
    },
    {
      "entropy": 5.555596113204956,
      "epoch": 2.930878084179971,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004147765072238806,
      "loss": 4.9604,
      "mean_token_accuracy": 0.20992930382490158,
      "num_tokens": 67058019.0,
      "step": 32310
    },
    {
      "entropy": 5.608548498153686,
      "epoch": 2.931331640058055,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004147511503452635,
      "loss": 4.9919,
      "mean_token_accuracy": 0.21123094558715821,
      "num_tokens": 67068844.0,
      "step": 32315
    },
    {
      "entropy": 5.578943872451783,
      "epoch": 2.9317851959361394,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041472579057655703,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22444686740636827,
      "num_tokens": 67077400.0,
      "step": 32320
    },
    {
      "entropy": 5.554438591003418,
      "epoch": 2.9322387518142237,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004147004279182856,
      "loss": 5.033,
      "mean_token_accuracy": 0.20300233364105225,
      "num_tokens": 67088352.0,
      "step": 32325
    },
    {
      "entropy": 5.617301321029663,
      "epoch": 2.9326923076923075,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004146750623709737,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21767656952142717,
      "num_tokens": 67097879.0,
      "step": 32330
    },
    {
      "entropy": 5.561146640777588,
      "epoch": 2.9331458635703918,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041464969393514595,
      "loss": 4.879,
      "mean_token_accuracy": 0.2222677767276764,
      "num_tokens": 67108355.0,
      "step": 32335
    },
    {
      "entropy": 5.536419105529785,
      "epoch": 2.933599419448476,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004146243226113269,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21575187593698503,
      "num_tokens": 67118348.0,
      "step": 32340
    },
    {
      "entropy": 5.469606494903564,
      "epoch": 2.9340529753265603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041459894840004135,
      "loss": 4.8433,
      "mean_token_accuracy": 0.226780304312706,
      "num_tokens": 67128944.0,
      "step": 32345
    },
    {
      "entropy": 5.578501033782959,
      "epoch": 2.9345065312046446,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041457357130181396,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21428166031837464,
      "num_tokens": 67139365.0,
      "step": 32350
    },
    {
      "entropy": 5.5610785484313965,
      "epoch": 2.934960087082729,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041454819131716945,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21414373070001602,
      "num_tokens": 67150406.0,
      "step": 32355
    },
    {
      "entropy": 5.639899253845215,
      "epoch": 2.9354136429608126,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041452280844663295,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21201469302177428,
      "num_tokens": 67161631.0,
      "step": 32360
    },
    {
      "entropy": 5.535846757888794,
      "epoch": 2.935867198838897,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004144974226907291,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2199413761496544,
      "num_tokens": 67171735.0,
      "step": 32365
    },
    {
      "entropy": 5.564001512527466,
      "epoch": 2.936320754716981,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000414472034049983,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2133524790406227,
      "num_tokens": 67181616.0,
      "step": 32370
    },
    {
      "entropy": 5.559957885742188,
      "epoch": 2.9367743105950654,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004144466425249197,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2192874878644943,
      "num_tokens": 67192519.0,
      "step": 32375
    },
    {
      "entropy": 5.579954433441162,
      "epoch": 2.9372278664731493,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004144212481160643,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2143664538860321,
      "num_tokens": 67202805.0,
      "step": 32380
    },
    {
      "entropy": 5.551706457138062,
      "epoch": 2.9376814223512335,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004143958508239419,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21774810552597046,
      "num_tokens": 67212778.0,
      "step": 32385
    },
    {
      "entropy": 5.511287355422974,
      "epoch": 2.938134978229318,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004143704506490779,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21707876175642013,
      "num_tokens": 67222588.0,
      "step": 32390
    },
    {
      "entropy": 5.605085372924805,
      "epoch": 2.938588534107402,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004143450475919975,
      "loss": 4.918,
      "mean_token_accuracy": 0.21651542633771897,
      "num_tokens": 67233022.0,
      "step": 32395
    },
    {
      "entropy": 5.538739109039307,
      "epoch": 2.9390420899854863,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004143196416532259,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21880135536193848,
      "num_tokens": 67242631.0,
      "step": 32400
    },
    {
      "entropy": 5.507452583312988,
      "epoch": 2.9394956458635706,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004142942328332887,
      "loss": 4.8688,
      "mean_token_accuracy": 0.2221100449562073,
      "num_tokens": 67252376.0,
      "step": 32405
    },
    {
      "entropy": 5.54774432182312,
      "epoch": 2.9399492017416544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004142688211327111,
      "loss": 4.8567,
      "mean_token_accuracy": 0.21777940392494202,
      "num_tokens": 67262209.0,
      "step": 32410
    },
    {
      "entropy": 5.587184095382691,
      "epoch": 2.9404027576197387,
      "grad_norm": 1.125,
      "learning_rate": 0.0004142434065520189,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21069523096084594,
      "num_tokens": 67272893.0,
      "step": 32415
    },
    {
      "entropy": 5.536428737640381,
      "epoch": 2.940856313497823,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004142179890917376,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2123550593852997,
      "num_tokens": 67284017.0,
      "step": 32420
    },
    {
      "entropy": 5.4997986316680905,
      "epoch": 2.941309869375907,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041419256875239275,
      "loss": 4.8936,
      "mean_token_accuracy": 0.21920786052942276,
      "num_tokens": 67293674.0,
      "step": 32425
    },
    {
      "entropy": 5.479498195648193,
      "epoch": 2.941763425253991,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004141671455345101,
      "loss": 4.9311,
      "mean_token_accuracy": 0.21710473001003266,
      "num_tokens": 67303759.0,
      "step": 32430
    },
    {
      "entropy": 5.592838859558105,
      "epoch": 2.9422169811320753,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041414171943861543,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21648956835269928,
      "num_tokens": 67314962.0,
      "step": 32435
    },
    {
      "entropy": 5.604931354522705,
      "epoch": 2.9426705370101596,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004141162904652346,
      "loss": 5.0519,
      "mean_token_accuracy": 0.207221382856369,
      "num_tokens": 67325210.0,
      "step": 32440
    },
    {
      "entropy": 5.567899179458618,
      "epoch": 2.943124092888244,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004140908586148934,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2108098417520523,
      "num_tokens": 67335110.0,
      "step": 32445
    },
    {
      "entropy": 5.571517467498779,
      "epoch": 2.943577648766328,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004140654238881178,
      "loss": 5.0166,
      "mean_token_accuracy": 0.21228591054677964,
      "num_tokens": 67345872.0,
      "step": 32450
    },
    {
      "entropy": 5.55539755821228,
      "epoch": 2.9440312046444124,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041403998628543375,
      "loss": 4.9271,
      "mean_token_accuracy": 0.210849791765213,
      "num_tokens": 67356532.0,
      "step": 32455
    },
    {
      "entropy": 5.534859561920166,
      "epoch": 2.9444847605224966,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041401454580736744,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21846393197774888,
      "num_tokens": 67366603.0,
      "step": 32460
    },
    {
      "entropy": 5.650631237030029,
      "epoch": 2.9449383164005805,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004139891024544448,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2128206416964531,
      "num_tokens": 67377487.0,
      "step": 32465
    },
    {
      "entropy": 5.631402111053466,
      "epoch": 2.9453918722786647,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004139636562271921,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21029378920793534,
      "num_tokens": 67387704.0,
      "step": 32470
    },
    {
      "entropy": 5.592033004760742,
      "epoch": 2.945845428156749,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041393820712613567,
      "loss": 4.9925,
      "mean_token_accuracy": 0.2059996634721756,
      "num_tokens": 67398409.0,
      "step": 32475
    },
    {
      "entropy": 5.581689071655274,
      "epoch": 2.9462989840348333,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041391275515180166,
      "loss": 4.8787,
      "mean_token_accuracy": 0.21427714675664902,
      "num_tokens": 67408905.0,
      "step": 32480
    },
    {
      "entropy": 5.567756795883179,
      "epoch": 2.946752539912917,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004138873003047165,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2292168289422989,
      "num_tokens": 67418466.0,
      "step": 32485
    },
    {
      "entropy": 5.459932041168213,
      "epoch": 2.9472060957910013,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004138618425854066,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2191898614168167,
      "num_tokens": 67428597.0,
      "step": 32490
    },
    {
      "entropy": 5.536923360824585,
      "epoch": 2.9476596516690856,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041383638199439824,
      "loss": 4.8899,
      "mean_token_accuracy": 0.21513280868530274,
      "num_tokens": 67439158.0,
      "step": 32495
    },
    {
      "entropy": 5.606556940078735,
      "epoch": 2.94811320754717,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041381091853221835,
      "loss": 4.9369,
      "mean_token_accuracy": 0.212184838950634,
      "num_tokens": 67449540.0,
      "step": 32500
    },
    {
      "entropy": 5.567067193984985,
      "epoch": 2.948566763425254,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004137854521993932,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21786169558763505,
      "num_tokens": 67459385.0,
      "step": 32505
    },
    {
      "entropy": 5.557390785217285,
      "epoch": 2.9490203193033384,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041375998299644946,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2185183957219124,
      "num_tokens": 67469597.0,
      "step": 32510
    },
    {
      "entropy": 5.5290497779846195,
      "epoch": 2.9494738751814222,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000413734510923914,
      "loss": 4.8605,
      "mean_token_accuracy": 0.21801808774471282,
      "num_tokens": 67479271.0,
      "step": 32515
    },
    {
      "entropy": 5.51042914390564,
      "epoch": 2.9499274310595065,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004137090359823135,
      "loss": 4.859,
      "mean_token_accuracy": 0.22150608897209167,
      "num_tokens": 67488997.0,
      "step": 32520
    },
    {
      "entropy": 5.508020210266113,
      "epoch": 2.9503809869375908,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041368355817217474,
      "loss": 4.901,
      "mean_token_accuracy": 0.21757570952177047,
      "num_tokens": 67499507.0,
      "step": 32525
    },
    {
      "entropy": 5.546919870376587,
      "epoch": 2.950834542815675,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041365807749402467,
      "loss": 4.8665,
      "mean_token_accuracy": 0.21749072074890136,
      "num_tokens": 67509916.0,
      "step": 32530
    },
    {
      "entropy": 5.537541437149048,
      "epoch": 2.951288098693759,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004136325939483902,
      "loss": 4.8968,
      "mean_token_accuracy": 0.21004502475261688,
      "num_tokens": 67519759.0,
      "step": 32535
    },
    {
      "entropy": 5.524202632904053,
      "epoch": 2.951741654571843,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004136071075357984,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21901537626981735,
      "num_tokens": 67529261.0,
      "step": 32540
    },
    {
      "entropy": 5.548251295089722,
      "epoch": 2.9521952104499274,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004135816182567763,
      "loss": 4.8944,
      "mean_token_accuracy": 0.21697842478752136,
      "num_tokens": 67539084.0,
      "step": 32545
    },
    {
      "entropy": 5.592967367172241,
      "epoch": 2.9526487663280117,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000413556126111851,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21511635780334473,
      "num_tokens": 67548551.0,
      "step": 32550
    },
    {
      "entropy": 5.560702085494995,
      "epoch": 2.953102322206096,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004135306311015497,
      "loss": 4.8968,
      "mean_token_accuracy": 0.21317071318626404,
      "num_tokens": 67558364.0,
      "step": 32555
    },
    {
      "entropy": 5.577459716796875,
      "epoch": 2.95355587808418,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004135051332263996,
      "loss": 4.9263,
      "mean_token_accuracy": 0.22179224640130996,
      "num_tokens": 67569257.0,
      "step": 32560
    },
    {
      "entropy": 5.543733167648315,
      "epoch": 2.954009433962264,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041347963248692814,
      "loss": 4.9982,
      "mean_token_accuracy": 0.21047332286834716,
      "num_tokens": 67580135.0,
      "step": 32565
    },
    {
      "entropy": 5.509096574783325,
      "epoch": 2.9544629898403483,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004134541288836626,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21457071453332902,
      "num_tokens": 67589705.0,
      "step": 32570
    },
    {
      "entropy": 5.546480274200439,
      "epoch": 2.9549165457184325,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041342862241713034,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2144412264227867,
      "num_tokens": 67599648.0,
      "step": 32575
    },
    {
      "entropy": 5.5888511657714846,
      "epoch": 2.955370101596517,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041340311308785886,
      "loss": 4.882,
      "mean_token_accuracy": 0.21569282561540604,
      "num_tokens": 67609624.0,
      "step": 32580
    },
    {
      "entropy": 5.5008931159973145,
      "epoch": 2.9558236574746006,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041337760089637577,
      "loss": 4.9101,
      "mean_token_accuracy": 0.2176475018262863,
      "num_tokens": 67619596.0,
      "step": 32585
    },
    {
      "entropy": 5.543699264526367,
      "epoch": 2.956277213352685,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041335208584320865,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2097705915570259,
      "num_tokens": 67629667.0,
      "step": 32590
    },
    {
      "entropy": 5.6078942775726315,
      "epoch": 2.956730769230769,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004133265679288851,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2139340043067932,
      "num_tokens": 67640048.0,
      "step": 32595
    },
    {
      "entropy": 5.614735889434814,
      "epoch": 2.9571843251088534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004133010471539329,
      "loss": 4.9786,
      "mean_token_accuracy": 0.20843395292758943,
      "num_tokens": 67650731.0,
      "step": 32600
    },
    {
      "entropy": 5.519395875930786,
      "epoch": 2.9576378809869377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041327552351887983,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21610843986272812,
      "num_tokens": 67661696.0,
      "step": 32605
    },
    {
      "entropy": 5.5737104415893555,
      "epoch": 2.958091436865022,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041324999702425366,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21364403516054153,
      "num_tokens": 67671758.0,
      "step": 32610
    },
    {
      "entropy": 5.528751516342163,
      "epoch": 2.9585449927431062,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004132244676705823,
      "loss": 4.916,
      "mean_token_accuracy": 0.21375818699598312,
      "num_tokens": 67682488.0,
      "step": 32615
    },
    {
      "entropy": 5.581802225112915,
      "epoch": 2.95899854862119,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004131989354583938,
      "loss": 4.9272,
      "mean_token_accuracy": 0.21794793605804444,
      "num_tokens": 67692945.0,
      "step": 32620
    },
    {
      "entropy": 5.5473089694976805,
      "epoch": 2.9594521044992743,
      "grad_norm": 1.15625,
      "learning_rate": 0.000413173400388216,
      "loss": 4.853,
      "mean_token_accuracy": 0.21994850039482117,
      "num_tokens": 67702924.0,
      "step": 32625
    },
    {
      "entropy": 5.5517168045043945,
      "epoch": 2.9599056603773586,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004131478624605771,
      "loss": 4.9406,
      "mean_token_accuracy": 0.216270712018013,
      "num_tokens": 67714284.0,
      "step": 32630
    },
    {
      "entropy": 5.5863282680511475,
      "epoch": 2.9603592162554424,
      "grad_norm": 1.125,
      "learning_rate": 0.0004131223216760052,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20805287063121797,
      "num_tokens": 67725902.0,
      "step": 32635
    },
    {
      "entropy": 5.571265602111817,
      "epoch": 2.9608127721335267,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004130967780350286,
      "loss": 4.901,
      "mean_token_accuracy": 0.21466223895549774,
      "num_tokens": 67736268.0,
      "step": 32640
    },
    {
      "entropy": 5.614508199691772,
      "epoch": 2.961266328011611,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004130712315381753,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22287954688072203,
      "num_tokens": 67745709.0,
      "step": 32645
    },
    {
      "entropy": 5.5743358612060545,
      "epoch": 2.961719883889695,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041304568218597385,
      "loss": 4.9124,
      "mean_token_accuracy": 0.21217063516378404,
      "num_tokens": 67756092.0,
      "step": 32650
    },
    {
      "entropy": 5.591142416000366,
      "epoch": 2.9621734397677795,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041302012997895246,
      "loss": 4.954,
      "mean_token_accuracy": 0.2143070727586746,
      "num_tokens": 67766224.0,
      "step": 32655
    },
    {
      "entropy": 5.677934980392456,
      "epoch": 2.9626269956458637,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004129945749176395,
      "loss": 5.0637,
      "mean_token_accuracy": 0.2124010443687439,
      "num_tokens": 67776787.0,
      "step": 32660
    },
    {
      "entropy": 5.552021551132202,
      "epoch": 2.963080551523948,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041296901700256375,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2166893571615219,
      "num_tokens": 67787924.0,
      "step": 32665
    },
    {
      "entropy": 5.553544044494629,
      "epoch": 2.963534107402032,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004129434562342535,
      "loss": 4.9214,
      "mean_token_accuracy": 0.20735210031270981,
      "num_tokens": 67798544.0,
      "step": 32670
    },
    {
      "entropy": 5.530894565582275,
      "epoch": 2.963987663280116,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041291789261323734,
      "loss": 4.9064,
      "mean_token_accuracy": 0.21655816733837127,
      "num_tokens": 67809528.0,
      "step": 32675
    },
    {
      "entropy": 5.459666967391968,
      "epoch": 2.9644412191582004,
      "grad_norm": 1.125,
      "learning_rate": 0.00041289232614004416,
      "loss": 4.7963,
      "mean_token_accuracy": 0.23257045298814774,
      "num_tokens": 67820263.0,
      "step": 32680
    },
    {
      "entropy": 5.628181838989258,
      "epoch": 2.9648947750362846,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004128667568152024,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21517910063266754,
      "num_tokens": 67830687.0,
      "step": 32685
    },
    {
      "entropy": 5.644575548171997,
      "epoch": 2.9653483309143684,
      "grad_norm": 1.125,
      "learning_rate": 0.00041284118463924107,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21090296506881714,
      "num_tokens": 67841714.0,
      "step": 32690
    },
    {
      "entropy": 5.523320627212525,
      "epoch": 2.9658018867924527,
      "grad_norm": 1.25,
      "learning_rate": 0.0004128156096126889,
      "loss": 4.9081,
      "mean_token_accuracy": 0.21795988082885742,
      "num_tokens": 67851952.0,
      "step": 32695
    },
    {
      "entropy": 5.476687383651734,
      "epoch": 2.966255442670537,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041279003173607477,
      "loss": 4.8065,
      "mean_token_accuracy": 0.22004805654287338,
      "num_tokens": 67861958.0,
      "step": 32700
    },
    {
      "entropy": 5.494622421264649,
      "epoch": 2.9667089985486212,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004127644510099278,
      "loss": 4.7933,
      "mean_token_accuracy": 0.2230198323726654,
      "num_tokens": 67871654.0,
      "step": 32705
    },
    {
      "entropy": 5.495678520202636,
      "epoch": 2.9671625544267055,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004127388674347767,
      "loss": 4.8901,
      "mean_token_accuracy": 0.22344845831394194,
      "num_tokens": 67881678.0,
      "step": 32710
    },
    {
      "entropy": 5.525390672683716,
      "epoch": 2.96761611030479,
      "grad_norm": 1.125,
      "learning_rate": 0.00041271328101115075,
      "loss": 4.9323,
      "mean_token_accuracy": 0.2161878004670143,
      "num_tokens": 67892663.0,
      "step": 32715
    },
    {
      "entropy": 5.667909574508667,
      "epoch": 2.9680696661828736,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041268769173957915,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20655827969312668,
      "num_tokens": 67903308.0,
      "step": 32720
    },
    {
      "entropy": 5.570732927322387,
      "epoch": 2.968523222060958,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041266209962059097,
      "loss": 4.9148,
      "mean_token_accuracy": 0.2176555275917053,
      "num_tokens": 67913092.0,
      "step": 32725
    },
    {
      "entropy": 5.540024709701538,
      "epoch": 2.968976777939042,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004126365046547154,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22821951061487197,
      "num_tokens": 67923708.0,
      "step": 32730
    },
    {
      "entropy": 5.4940650939941404,
      "epoch": 2.9694303338171264,
      "grad_norm": 1.125,
      "learning_rate": 0.0004126109068424819,
      "loss": 4.9046,
      "mean_token_accuracy": 0.21351844668388367,
      "num_tokens": 67934164.0,
      "step": 32735
    },
    {
      "entropy": 5.624765014648437,
      "epoch": 2.96988388969521,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004125853061844198,
      "loss": 5.0018,
      "mean_token_accuracy": 0.20874401777982712,
      "num_tokens": 67943162.0,
      "step": 32740
    },
    {
      "entropy": 5.597575759887695,
      "epoch": 2.9703374455732945,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004125597026810585,
      "loss": 4.984,
      "mean_token_accuracy": 0.2133081629872322,
      "num_tokens": 67953943.0,
      "step": 32745
    },
    {
      "entropy": 5.508371829986572,
      "epoch": 2.9707910014513788,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041253409633292747,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2165821373462677,
      "num_tokens": 67964603.0,
      "step": 32750
    },
    {
      "entropy": 5.573017215728759,
      "epoch": 2.971244557329463,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041250848714055626,
      "loss": 4.9704,
      "mean_token_accuracy": 0.21357410699129104,
      "num_tokens": 67974541.0,
      "step": 32755
    },
    {
      "entropy": 5.597278022766114,
      "epoch": 2.9716981132075473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004124828751044746,
      "loss": 4.9376,
      "mean_token_accuracy": 0.20984164625406265,
      "num_tokens": 67984968.0,
      "step": 32760
    },
    {
      "entropy": 5.54750804901123,
      "epoch": 2.9721516690856316,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004124572602252119,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22026091665029526,
      "num_tokens": 67995567.0,
      "step": 32765
    },
    {
      "entropy": 5.620554685592651,
      "epoch": 2.9726052249637154,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004124316425032981,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20670343041419983,
      "num_tokens": 68008530.0,
      "step": 32770
    },
    {
      "entropy": 5.640254974365234,
      "epoch": 2.9730587808417996,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041240602193926294,
      "loss": 5.0069,
      "mean_token_accuracy": 0.2080129712820053,
      "num_tokens": 68018393.0,
      "step": 32775
    },
    {
      "entropy": 5.5888418674469,
      "epoch": 2.973512336719884,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004123803985336362,
      "loss": 4.941,
      "mean_token_accuracy": 0.21129584014415742,
      "num_tokens": 68027677.0,
      "step": 32780
    },
    {
      "entropy": 5.524590635299683,
      "epoch": 2.973965892597968,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041235477228694786,
      "loss": 4.8169,
      "mean_token_accuracy": 0.227382592856884,
      "num_tokens": 68037157.0,
      "step": 32785
    },
    {
      "entropy": 5.547086095809936,
      "epoch": 2.974419448476052,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041232914319972775,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22174557745456697,
      "num_tokens": 68047793.0,
      "step": 32790
    },
    {
      "entropy": 5.549725818634033,
      "epoch": 2.9748730043541363,
      "grad_norm": 1.25,
      "learning_rate": 0.00041230351127250603,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2223878026008606,
      "num_tokens": 68058155.0,
      "step": 32795
    },
    {
      "entropy": 5.617917060852051,
      "epoch": 2.9753265602322205,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041227787650581266,
      "loss": 4.8619,
      "mean_token_accuracy": 0.21902658194303512,
      "num_tokens": 68067440.0,
      "step": 32800
    },
    {
      "entropy": 5.5729207515716555,
      "epoch": 2.975780116110305,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041225223890017785,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2187219515442848,
      "num_tokens": 68077815.0,
      "step": 32805
    },
    {
      "entropy": 5.536134243011475,
      "epoch": 2.976233671988389,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041222659845613173,
      "loss": 4.8792,
      "mean_token_accuracy": 0.2223426803946495,
      "num_tokens": 68087692.0,
      "step": 32810
    },
    {
      "entropy": 5.492402791976929,
      "epoch": 2.9766872278664733,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041220095517420464,
      "loss": 4.891,
      "mean_token_accuracy": 0.2175753191113472,
      "num_tokens": 68097774.0,
      "step": 32815
    },
    {
      "entropy": 5.550098276138305,
      "epoch": 2.9771407837445576,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004121753090549268,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21566751450300217,
      "num_tokens": 68107992.0,
      "step": 32820
    },
    {
      "entropy": 5.521971845626831,
      "epoch": 2.9775943396226414,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041214966009882857,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22559628635644913,
      "num_tokens": 68117103.0,
      "step": 32825
    },
    {
      "entropy": 5.597184371948242,
      "epoch": 2.9780478955007257,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004121240083064405,
      "loss": 4.9801,
      "mean_token_accuracy": 0.21236259043216704,
      "num_tokens": 68127951.0,
      "step": 32830
    },
    {
      "entropy": 5.610168504714966,
      "epoch": 2.97850145137881,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041209835367829293,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2209950089454651,
      "num_tokens": 68138229.0,
      "step": 32835
    },
    {
      "entropy": 5.468213129043579,
      "epoch": 2.9789550072568938,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041207269621491654,
      "loss": 4.8873,
      "mean_token_accuracy": 0.22095452547073363,
      "num_tokens": 68148019.0,
      "step": 32840
    },
    {
      "entropy": 5.4690577507019045,
      "epoch": 2.979408563134978,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004120470359168418,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2088165193796158,
      "num_tokens": 68158704.0,
      "step": 32845
    },
    {
      "entropy": 5.591777992248535,
      "epoch": 2.9798621190130623,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041202137278459946,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21985854655504228,
      "num_tokens": 68169105.0,
      "step": 32850
    },
    {
      "entropy": 5.586925745010376,
      "epoch": 2.9803156748911466,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004119957068187202,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22090157270431518,
      "num_tokens": 68179778.0,
      "step": 32855
    },
    {
      "entropy": 5.5218699932098385,
      "epoch": 2.980769230769231,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004119700380197348,
      "loss": 4.8926,
      "mean_token_accuracy": 0.21476766616106033,
      "num_tokens": 68189405.0,
      "step": 32860
    },
    {
      "entropy": 5.4587029933929445,
      "epoch": 2.981222786647315,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041194436638817417,
      "loss": 4.8441,
      "mean_token_accuracy": 0.22606901079416275,
      "num_tokens": 68199351.0,
      "step": 32865
    },
    {
      "entropy": 5.598010873794555,
      "epoch": 2.9816763425253994,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041191869192456907,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21060204654932022,
      "num_tokens": 68209904.0,
      "step": 32870
    },
    {
      "entropy": 5.615426874160766,
      "epoch": 2.982129898403483,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004118930146294506,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20826841443777083,
      "num_tokens": 68221754.0,
      "step": 32875
    },
    {
      "entropy": 5.564703559875488,
      "epoch": 2.9825834542815675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004118673345033496,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2094820261001587,
      "num_tokens": 68232328.0,
      "step": 32880
    },
    {
      "entropy": 5.492561292648316,
      "epoch": 2.9830370101596517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004118416515467974,
      "loss": 4.8533,
      "mean_token_accuracy": 0.217547769844532,
      "num_tokens": 68242812.0,
      "step": 32885
    },
    {
      "entropy": 5.636599349975586,
      "epoch": 2.983490566037736,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004118159657603248,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2141905352473259,
      "num_tokens": 68253888.0,
      "step": 32890
    },
    {
      "entropy": 5.6142597675323485,
      "epoch": 2.98394412191582,
      "grad_norm": 1.25,
      "learning_rate": 0.0004117902771444633,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2176228553056717,
      "num_tokens": 68264877.0,
      "step": 32895
    },
    {
      "entropy": 5.657246208190918,
      "epoch": 2.984397677793904,
      "grad_norm": 1.125,
      "learning_rate": 0.000411764585699744,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20741144567728043,
      "num_tokens": 68275863.0,
      "step": 32900
    },
    {
      "entropy": 5.573932266235351,
      "epoch": 2.9848512336719883,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041173889142669806,
      "loss": 4.93,
      "mean_token_accuracy": 0.21535502523183822,
      "num_tokens": 68286768.0,
      "step": 32905
    },
    {
      "entropy": 5.54905219078064,
      "epoch": 2.9853047895500726,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004117131943258572,
      "loss": 4.8689,
      "mean_token_accuracy": 0.2205004498362541,
      "num_tokens": 68296990.0,
      "step": 32910
    },
    {
      "entropy": 5.480344486236572,
      "epoch": 2.985758345428157,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041168749439775255,
      "loss": 4.8718,
      "mean_token_accuracy": 0.2165363386273384,
      "num_tokens": 68307646.0,
      "step": 32915
    },
    {
      "entropy": 5.596737194061279,
      "epoch": 2.986211901306241,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004116617916429156,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20310408025979995,
      "num_tokens": 68319586.0,
      "step": 32920
    },
    {
      "entropy": 5.645182704925537,
      "epoch": 2.986665457184325,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004116360860618781,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21313963681459427,
      "num_tokens": 68329692.0,
      "step": 32925
    },
    {
      "entropy": 5.6401551246643065,
      "epoch": 2.9871190130624092,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004116103776551715,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21233301907777785,
      "num_tokens": 68339589.0,
      "step": 32930
    },
    {
      "entropy": 5.595186614990235,
      "epoch": 2.9875725689404935,
      "grad_norm": 1.125,
      "learning_rate": 0.00041158466642332743,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2034110650420189,
      "num_tokens": 68349485.0,
      "step": 32935
    },
    {
      "entropy": 5.589399147033691,
      "epoch": 2.9880261248185778,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004115589523668777,
      "loss": 4.9,
      "mean_token_accuracy": 0.22183861583471298,
      "num_tokens": 68359695.0,
      "step": 32940
    },
    {
      "entropy": 5.5047607421875,
      "epoch": 2.9884796806966616,
      "grad_norm": 1.140625,
      "learning_rate": 0.000411533235486354,
      "loss": 4.8119,
      "mean_token_accuracy": 0.22178723961114882,
      "num_tokens": 68369591.0,
      "step": 32945
    },
    {
      "entropy": 5.586577844619751,
      "epoch": 2.988933236574746,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004115075157822883,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22116850614547728,
      "num_tokens": 68380382.0,
      "step": 32950
    },
    {
      "entropy": 5.453122091293335,
      "epoch": 2.98938679245283,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004114817932552122,
      "loss": 4.8156,
      "mean_token_accuracy": 0.21976580321788788,
      "num_tokens": 68389865.0,
      "step": 32955
    },
    {
      "entropy": 5.498558187484742,
      "epoch": 2.9898403483309144,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041145606790565797,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2309402585029602,
      "num_tokens": 68399267.0,
      "step": 32960
    },
    {
      "entropy": 5.5747607231140135,
      "epoch": 2.9902939042089987,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004114303397341575,
      "loss": 4.942,
      "mean_token_accuracy": 0.22618331760168076,
      "num_tokens": 68409480.0,
      "step": 32965
    },
    {
      "entropy": 5.6772459030151365,
      "epoch": 2.990747460087083,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004114046087412427,
      "loss": 5.0389,
      "mean_token_accuracy": 0.2100787028670311,
      "num_tokens": 68420019.0,
      "step": 32970
    },
    {
      "entropy": 5.490229701995849,
      "epoch": 2.9912010159651667,
      "grad_norm": 1.234375,
      "learning_rate": 0.000411378874927446,
      "loss": 4.882,
      "mean_token_accuracy": 0.2197356641292572,
      "num_tokens": 68429967.0,
      "step": 32975
    },
    {
      "entropy": 5.4569525718688965,
      "epoch": 2.991654571843251,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004113531382932993,
      "loss": 4.882,
      "mean_token_accuracy": 0.21734683513641356,
      "num_tokens": 68441639.0,
      "step": 32980
    },
    {
      "entropy": 5.564846992492676,
      "epoch": 2.9921081277213353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000411327398839335,
      "loss": 4.94,
      "mean_token_accuracy": 0.21031366735696794,
      "num_tokens": 68452781.0,
      "step": 32985
    },
    {
      "entropy": 5.561367416381836,
      "epoch": 2.9925616835994195,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004113016565660854,
      "loss": 4.8775,
      "mean_token_accuracy": 0.21896420568227767,
      "num_tokens": 68463353.0,
      "step": 32990
    },
    {
      "entropy": 5.539358901977539,
      "epoch": 2.9930152394775034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041127591147408274,
      "loss": 4.8932,
      "mean_token_accuracy": 0.20935145020484924,
      "num_tokens": 68473507.0,
      "step": 32995
    },
    {
      "entropy": 5.554093742370606,
      "epoch": 2.9934687953555876,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041125016356385945,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22207578420639038,
      "num_tokens": 68483737.0,
      "step": 33000
    },
    {
      "epoch": 2.9934687953555876,
      "eval_entropy": 5.290873151343333,
      "eval_loss": 4.967285633087158,
      "eval_mean_token_accuracy": 0.22357556195550374,
      "eval_num_tokens": 68483737.0,
      "eval_runtime": 20.5868,
      "eval_samples_per_second": 1717.608,
      "eval_steps_per_second": 214.701,
      "step": 33000
    },
    {
      "entropy": 5.558495664596558,
      "epoch": 2.993922351233672,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041122441283594817,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21845910847187042,
      "num_tokens": 68494013.0,
      "step": 33005
    },
    {
      "entropy": 5.490463590621948,
      "epoch": 2.994375907111756,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004111986592908813,
      "loss": 4.8585,
      "mean_token_accuracy": 0.22673794031143188,
      "num_tokens": 68504361.0,
      "step": 33010
    },
    {
      "entropy": 5.502665758132935,
      "epoch": 2.9948294629898404,
      "grad_norm": 1.375,
      "learning_rate": 0.00041117290292919134,
      "loss": 4.8446,
      "mean_token_accuracy": 0.2210651308298111,
      "num_tokens": 68514825.0,
      "step": 33015
    },
    {
      "entropy": 5.548485660552979,
      "epoch": 2.9952830188679247,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004111471437514111,
      "loss": 4.8597,
      "mean_token_accuracy": 0.21454495191574097,
      "num_tokens": 68524858.0,
      "step": 33020
    },
    {
      "entropy": 5.535737037658691,
      "epoch": 2.995736574746009,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004111213817580733,
      "loss": 4.8347,
      "mean_token_accuracy": 0.2262229725718498,
      "num_tokens": 68534481.0,
      "step": 33025
    },
    {
      "entropy": 5.508427286148072,
      "epoch": 2.996190130624093,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004110956169497106,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21640910804271699,
      "num_tokens": 68545436.0,
      "step": 33030
    },
    {
      "entropy": 5.53284215927124,
      "epoch": 2.996643686502177,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004110698493268558,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22356501817703248,
      "num_tokens": 68555434.0,
      "step": 33035
    },
    {
      "entropy": 5.564860868453979,
      "epoch": 2.9970972423802613,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004110440788900419,
      "loss": 5.0303,
      "mean_token_accuracy": 0.21122870445251465,
      "num_tokens": 68566295.0,
      "step": 33040
    },
    {
      "entropy": 5.526291751861573,
      "epoch": 2.9975507982583456,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041101830563980174,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22296057045459747,
      "num_tokens": 68575927.0,
      "step": 33045
    },
    {
      "entropy": 5.6168183326721195,
      "epoch": 2.9980043541364294,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041099252957666843,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2188567265868187,
      "num_tokens": 68586254.0,
      "step": 33050
    },
    {
      "entropy": 5.502985429763794,
      "epoch": 2.9984579100145137,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041096675070117496,
      "loss": 4.9314,
      "mean_token_accuracy": 0.21814779937267303,
      "num_tokens": 68596881.0,
      "step": 33055
    },
    {
      "entropy": 5.475627183914185,
      "epoch": 2.998911465892598,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041094096901385434,
      "loss": 4.8276,
      "mean_token_accuracy": 0.2229520007967949,
      "num_tokens": 68606875.0,
      "step": 33060
    },
    {
      "entropy": 5.601109743118286,
      "epoch": 2.999365021770682,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004109151845152399,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21973366290330887,
      "num_tokens": 68617011.0,
      "step": 33065
    },
    {
      "entropy": 5.608200120925903,
      "epoch": 2.9998185776487665,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004108893972058649,
      "loss": 4.9551,
      "mean_token_accuracy": 0.21498948335647583,
      "num_tokens": 68627764.0,
      "step": 33070
    },
    {
      "entropy": 5.589737606048584,
      "epoch": 3.0002721335268503,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041086360708626236,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21905701756477355,
      "num_tokens": 68637272.0,
      "step": 33075
    },
    {
      "entropy": 5.4978930950164795,
      "epoch": 3.0007256894049346,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041083781415696585,
      "loss": 4.8717,
      "mean_token_accuracy": 0.21449463069438934,
      "num_tokens": 68648058.0,
      "step": 33080
    },
    {
      "entropy": 5.4735297679901125,
      "epoch": 3.001179245283019,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004108120184185088,
      "loss": 4.7903,
      "mean_token_accuracy": 0.22719339579343795,
      "num_tokens": 68658174.0,
      "step": 33085
    },
    {
      "entropy": 5.5593860149383545,
      "epoch": 3.001632801161103,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041078621987142457,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23413125574588775,
      "num_tokens": 68669281.0,
      "step": 33090
    },
    {
      "entropy": 5.534533882141114,
      "epoch": 3.0020863570391874,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004107604185162467,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21341838091611862,
      "num_tokens": 68679879.0,
      "step": 33095
    },
    {
      "entropy": 5.478369951248169,
      "epoch": 3.0025399129172716,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041073461435350877,
      "loss": 4.7921,
      "mean_token_accuracy": 0.22448095083236694,
      "num_tokens": 68689852.0,
      "step": 33100
    },
    {
      "entropy": 5.472542095184326,
      "epoch": 3.0029934687953554,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041070880738374447,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22312019318342208,
      "num_tokens": 68700149.0,
      "step": 33105
    },
    {
      "entropy": 5.509248447418213,
      "epoch": 3.0034470246734397,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041068299760748734,
      "loss": 4.8517,
      "mean_token_accuracy": 0.2193642243742943,
      "num_tokens": 68710837.0,
      "step": 33110
    },
    {
      "entropy": 5.552023124694824,
      "epoch": 3.003900580551524,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041065718502527134,
      "loss": 4.9038,
      "mean_token_accuracy": 0.21803360432386398,
      "num_tokens": 68721451.0,
      "step": 33115
    },
    {
      "entropy": 5.506927442550659,
      "epoch": 3.0043541364296082,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041063136963763013,
      "loss": 4.8469,
      "mean_token_accuracy": 0.22254530042409898,
      "num_tokens": 68732911.0,
      "step": 33120
    },
    {
      "entropy": 5.520870733261108,
      "epoch": 3.0048076923076925,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004106055514450977,
      "loss": 4.8136,
      "mean_token_accuracy": 0.22911326885223388,
      "num_tokens": 68742650.0,
      "step": 33125
    },
    {
      "entropy": 5.496462965011597,
      "epoch": 3.0052612481857763,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004105797304482079,
      "loss": 4.8263,
      "mean_token_accuracy": 0.22387445569038392,
      "num_tokens": 68753660.0,
      "step": 33130
    },
    {
      "entropy": 5.467442417144776,
      "epoch": 3.0057148040638606,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004105539066474947,
      "loss": 4.7386,
      "mean_token_accuracy": 0.22541164308786393,
      "num_tokens": 68763868.0,
      "step": 33135
    },
    {
      "entropy": 5.5195940971374515,
      "epoch": 3.006168359941945,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041052808004349214,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2229118600487709,
      "num_tokens": 68773911.0,
      "step": 33140
    },
    {
      "entropy": 5.550774717330933,
      "epoch": 3.006621915820029,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004105022506367344,
      "loss": 4.9398,
      "mean_token_accuracy": 0.21408320367336273,
      "num_tokens": 68786423.0,
      "step": 33145
    },
    {
      "entropy": 5.523630094528198,
      "epoch": 3.0070754716981134,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004104764184277555,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21312060505151748,
      "num_tokens": 68797329.0,
      "step": 33150
    },
    {
      "entropy": 5.59370265007019,
      "epoch": 3.0075290275761972,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004104505834170898,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2170417219400406,
      "num_tokens": 68808447.0,
      "step": 33155
    },
    {
      "entropy": 5.565026903152466,
      "epoch": 3.0079825834542815,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041042474560527153,
      "loss": 4.869,
      "mean_token_accuracy": 0.22275021374225618,
      "num_tokens": 68819381.0,
      "step": 33160
    },
    {
      "entropy": 5.467403173446655,
      "epoch": 3.0084361393323658,
      "grad_norm": 1.15625,
      "learning_rate": 0.000410398904992835,
      "loss": 4.8084,
      "mean_token_accuracy": 0.22707327157258989,
      "num_tokens": 68829319.0,
      "step": 33165
    },
    {
      "entropy": 5.497539520263672,
      "epoch": 3.00888969521045,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041037306158031464,
      "loss": 4.9171,
      "mean_token_accuracy": 0.2244374379515648,
      "num_tokens": 68838640.0,
      "step": 33170
    },
    {
      "entropy": 5.5500000476837155,
      "epoch": 3.0093432510885343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004103472153682448,
      "loss": 4.8494,
      "mean_token_accuracy": 0.21816093772649764,
      "num_tokens": 68847960.0,
      "step": 33175
    },
    {
      "entropy": 5.543204116821289,
      "epoch": 3.009796806966618,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041032136635716016,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22023076564073563,
      "num_tokens": 68858851.0,
      "step": 33180
    },
    {
      "entropy": 5.494747877120972,
      "epoch": 3.0102503628447024,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004102955145475951,
      "loss": 4.8383,
      "mean_token_accuracy": 0.21889970153570176,
      "num_tokens": 68869006.0,
      "step": 33185
    },
    {
      "entropy": 5.614867734909057,
      "epoch": 3.0107039187227866,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041026965994008435,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21492185145616532,
      "num_tokens": 68879606.0,
      "step": 33190
    },
    {
      "entropy": 5.490984010696411,
      "epoch": 3.011157474600871,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041024380253516253,
      "loss": 4.8224,
      "mean_token_accuracy": 0.22771223187446593,
      "num_tokens": 68889862.0,
      "step": 33195
    },
    {
      "entropy": 5.624411869049072,
      "epoch": 3.011611030478955,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041021794233336444,
      "loss": 4.9458,
      "mean_token_accuracy": 0.2189653217792511,
      "num_tokens": 68900512.0,
      "step": 33200
    },
    {
      "entropy": 5.538083600997925,
      "epoch": 3.012064586357039,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004101920793352249,
      "loss": 4.884,
      "mean_token_accuracy": 0.2161460280418396,
      "num_tokens": 68910475.0,
      "step": 33205
    },
    {
      "entropy": 5.578966331481934,
      "epoch": 3.0125181422351233,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004101662135412786,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21984424591064453,
      "num_tokens": 68920583.0,
      "step": 33210
    },
    {
      "entropy": 5.450491619110108,
      "epoch": 3.0129716981132075,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004101403449520606,
      "loss": 4.8085,
      "mean_token_accuracy": 0.21742058545351028,
      "num_tokens": 68931016.0,
      "step": 33215
    },
    {
      "entropy": 5.503797245025635,
      "epoch": 3.013425253991292,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041011447356810584,
      "loss": 4.8251,
      "mean_token_accuracy": 0.22179421037435532,
      "num_tokens": 68941489.0,
      "step": 33220
    },
    {
      "entropy": 5.570745182037354,
      "epoch": 3.013878809869376,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004100885993899493,
      "loss": 4.906,
      "mean_token_accuracy": 0.21582674980163574,
      "num_tokens": 68952322.0,
      "step": 33225
    },
    {
      "entropy": 5.567753028869629,
      "epoch": 3.01433236574746,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004100627224181261,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22080958038568496,
      "num_tokens": 68962570.0,
      "step": 33230
    },
    {
      "entropy": 5.582212638854981,
      "epoch": 3.014785921625544,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041003684265317143,
      "loss": 4.8265,
      "mean_token_accuracy": 0.22261277437210084,
      "num_tokens": 68973733.0,
      "step": 33235
    },
    {
      "entropy": 5.52776026725769,
      "epoch": 3.0152394775036284,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041001096009562045,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22449675053358079,
      "num_tokens": 68983526.0,
      "step": 33240
    },
    {
      "entropy": 5.612430381774902,
      "epoch": 3.0156930333817127,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004099850747460084,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22062678933143615,
      "num_tokens": 68993968.0,
      "step": 33245
    },
    {
      "entropy": 5.551955318450927,
      "epoch": 3.016146589259797,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040995918660487055,
      "loss": 4.833,
      "mean_token_accuracy": 0.2303554356098175,
      "num_tokens": 69003599.0,
      "step": 33250
    },
    {
      "entropy": 5.521548748016357,
      "epoch": 3.0166001451378808,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004099332956727424,
      "loss": 4.8481,
      "mean_token_accuracy": 0.21906791031360626,
      "num_tokens": 69013686.0,
      "step": 33255
    },
    {
      "entropy": 5.543167400360107,
      "epoch": 3.017053701015965,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004099074019501593,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22182500660419463,
      "num_tokens": 69023803.0,
      "step": 33260
    },
    {
      "entropy": 5.49210033416748,
      "epoch": 3.0175072568940493,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004098815054376567,
      "loss": 4.776,
      "mean_token_accuracy": 0.2271982342004776,
      "num_tokens": 69033885.0,
      "step": 33265
    },
    {
      "entropy": 5.513218784332276,
      "epoch": 3.0179608127721336,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004098556061357702,
      "loss": 4.8308,
      "mean_token_accuracy": 0.22322162687778474,
      "num_tokens": 69044471.0,
      "step": 33270
    },
    {
      "entropy": 5.462755012512207,
      "epoch": 3.018414368650218,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040982970404503544,
      "loss": 4.7563,
      "mean_token_accuracy": 0.2294916346669197,
      "num_tokens": 69055109.0,
      "step": 33275
    },
    {
      "entropy": 5.488037633895874,
      "epoch": 3.018867924528302,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040980379916598803,
      "loss": 4.8628,
      "mean_token_accuracy": 0.22305349707603456,
      "num_tokens": 69064692.0,
      "step": 33280
    },
    {
      "entropy": 5.543799877166748,
      "epoch": 3.019321480406386,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004097778914991637,
      "loss": 4.8356,
      "mean_token_accuracy": 0.22339656054973603,
      "num_tokens": 69075331.0,
      "step": 33285
    },
    {
      "entropy": 5.517863368988037,
      "epoch": 3.01977503628447,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040975198104509827,
      "loss": 4.8312,
      "mean_token_accuracy": 0.22567791789770125,
      "num_tokens": 69085017.0,
      "step": 33290
    },
    {
      "entropy": 5.476214218139648,
      "epoch": 3.0202285921625545,
      "grad_norm": 1.25,
      "learning_rate": 0.0004097260678043275,
      "loss": 4.8498,
      "mean_token_accuracy": 0.21534609347581862,
      "num_tokens": 69096128.0,
      "step": 33295
    },
    {
      "entropy": 5.567341756820679,
      "epoch": 3.0206821480406387,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040970015177738725,
      "loss": 4.96,
      "mean_token_accuracy": 0.21352556198835373,
      "num_tokens": 69106479.0,
      "step": 33300
    },
    {
      "entropy": 5.613250112533569,
      "epoch": 3.021135703918723,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004096742329648136,
      "loss": 4.8939,
      "mean_token_accuracy": 0.21715326011180877,
      "num_tokens": 69116346.0,
      "step": 33305
    },
    {
      "entropy": 5.544456720352173,
      "epoch": 3.021589259796807,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004096483113671425,
      "loss": 4.9034,
      "mean_token_accuracy": 0.21691830903291703,
      "num_tokens": 69126489.0,
      "step": 33310
    },
    {
      "entropy": 5.523610544204712,
      "epoch": 3.022042815674891,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040962238698490994,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22266105115413665,
      "num_tokens": 69136618.0,
      "step": 33315
    },
    {
      "entropy": 5.571689081192017,
      "epoch": 3.0224963715529753,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004095964598186522,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21185413002967834,
      "num_tokens": 69146242.0,
      "step": 33320
    },
    {
      "entropy": 5.530975580215454,
      "epoch": 3.0229499274310596,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004095705298689053,
      "loss": 4.8318,
      "mean_token_accuracy": 0.22369010895490646,
      "num_tokens": 69157269.0,
      "step": 33325
    },
    {
      "entropy": 5.621033000946045,
      "epoch": 3.023403483309144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040954459713620563,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2155551165342331,
      "num_tokens": 69169260.0,
      "step": 33330
    },
    {
      "entropy": 5.642344331741333,
      "epoch": 3.0238570391872277,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004095186616210893,
      "loss": 4.9905,
      "mean_token_accuracy": 0.20207439363002777,
      "num_tokens": 69179815.0,
      "step": 33335
    },
    {
      "entropy": 5.601299142837524,
      "epoch": 3.024310595065312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004094927233240928,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2140293002128601,
      "num_tokens": 69189959.0,
      "step": 33340
    },
    {
      "entropy": 5.512681531906128,
      "epoch": 3.0247641509433962,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040946678224575264,
      "loss": 4.8532,
      "mean_token_accuracy": 0.22089072614908217,
      "num_tokens": 69199485.0,
      "step": 33345
    },
    {
      "entropy": 5.4578361988067625,
      "epoch": 3.0252177068214805,
      "grad_norm": 1.296875,
      "learning_rate": 0.000409440838386605,
      "loss": 4.8716,
      "mean_token_accuracy": 0.21979523599147796,
      "num_tokens": 69209180.0,
      "step": 33350
    },
    {
      "entropy": 5.5685145378112795,
      "epoch": 3.0256712626995648,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004094148917471866,
      "loss": 4.9108,
      "mean_token_accuracy": 0.217694553732872,
      "num_tokens": 69220648.0,
      "step": 33355
    },
    {
      "entropy": 5.500479030609131,
      "epoch": 3.0261248185776486,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000409388942328034,
      "loss": 4.7516,
      "mean_token_accuracy": 0.22733724117279053,
      "num_tokens": 69230685.0,
      "step": 33360
    },
    {
      "entropy": 5.499433183670044,
      "epoch": 3.026578374455733,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004093629901296838,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21350348442792894,
      "num_tokens": 69240407.0,
      "step": 33365
    },
    {
      "entropy": 5.614395618438721,
      "epoch": 3.027031930333817,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004093370351526727,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2155456319451332,
      "num_tokens": 69250624.0,
      "step": 33370
    },
    {
      "entropy": 5.603523063659668,
      "epoch": 3.0274854862119014,
      "grad_norm": 1.5,
      "learning_rate": 0.0004093110773975376,
      "loss": 4.9408,
      "mean_token_accuracy": 0.20913257598876953,
      "num_tokens": 69261131.0,
      "step": 33375
    },
    {
      "entropy": 5.503702163696289,
      "epoch": 3.0279390420899857,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004092851168648152,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2144733726978302,
      "num_tokens": 69270668.0,
      "step": 33380
    },
    {
      "entropy": 5.528714561462403,
      "epoch": 3.0283925979680695,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004092591535550422,
      "loss": 4.824,
      "mean_token_accuracy": 0.22638099938631057,
      "num_tokens": 69280528.0,
      "step": 33385
    },
    {
      "entropy": 5.513356113433838,
      "epoch": 3.0288461538461537,
      "grad_norm": 1.25,
      "learning_rate": 0.0004092331874687558,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22600357234477997,
      "num_tokens": 69290376.0,
      "step": 33390
    },
    {
      "entropy": 5.539975214004516,
      "epoch": 3.029299709724238,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040920721860649285,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22962289154529572,
      "num_tokens": 69301766.0,
      "step": 33395
    },
    {
      "entropy": 5.576062345504761,
      "epoch": 3.0297532656023223,
      "grad_norm": 1.125,
      "learning_rate": 0.0004091812469687904,
      "loss": 4.8445,
      "mean_token_accuracy": 0.217105495929718,
      "num_tokens": 69312668.0,
      "step": 33400
    },
    {
      "entropy": 5.562944650650024,
      "epoch": 3.0302068214804065,
      "grad_norm": 1.125,
      "learning_rate": 0.0004091552725561856,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21654330790042878,
      "num_tokens": 69323195.0,
      "step": 33405
    },
    {
      "entropy": 5.488876152038574,
      "epoch": 3.0306603773584904,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004091292953692156,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22053744196891784,
      "num_tokens": 69333812.0,
      "step": 33410
    },
    {
      "entropy": 5.54521164894104,
      "epoch": 3.0311139332365746,
      "grad_norm": 1.125,
      "learning_rate": 0.00040910331540841755,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2262769639492035,
      "num_tokens": 69344378.0,
      "step": 33415
    },
    {
      "entropy": 5.513668727874756,
      "epoch": 3.031567489114659,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040907733267432877,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22027383297681807,
      "num_tokens": 69354592.0,
      "step": 33420
    },
    {
      "entropy": 5.550172901153564,
      "epoch": 3.032021044992743,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040905134716748656,
      "loss": 4.8693,
      "mean_token_accuracy": 0.21954271346330642,
      "num_tokens": 69365011.0,
      "step": 33425
    },
    {
      "entropy": 5.505629968643189,
      "epoch": 3.0324746008708274,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004090253588884284,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2127205953001976,
      "num_tokens": 69377010.0,
      "step": 33430
    },
    {
      "entropy": 5.55498194694519,
      "epoch": 3.0329281567489113,
      "grad_norm": 1.25,
      "learning_rate": 0.00040899936783769157,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22480493634939194,
      "num_tokens": 69387263.0,
      "step": 33435
    },
    {
      "entropy": 5.562593650817871,
      "epoch": 3.0333817126269955,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040897337401581374,
      "loss": 4.8923,
      "mean_token_accuracy": 0.2265462100505829,
      "num_tokens": 69399188.0,
      "step": 33440
    },
    {
      "entropy": 5.525684547424317,
      "epoch": 3.03383526850508,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040894737742333244,
      "loss": 4.8749,
      "mean_token_accuracy": 0.21651416420936584,
      "num_tokens": 69408271.0,
      "step": 33445
    },
    {
      "entropy": 5.5857994556427,
      "epoch": 3.034288824383164,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004089213780607851,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20722945630550385,
      "num_tokens": 69418424.0,
      "step": 33450
    },
    {
      "entropy": 5.592473268508911,
      "epoch": 3.0347423802612483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004088953759287096,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22269641906023024,
      "num_tokens": 69427802.0,
      "step": 33455
    },
    {
      "entropy": 5.588563442230225,
      "epoch": 3.0351959361393326,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004088693710276436,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22252792119979858,
      "num_tokens": 69437581.0,
      "step": 33460
    },
    {
      "entropy": 5.546655941009521,
      "epoch": 3.0356494920174164,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040884336335812494,
      "loss": 4.8271,
      "mean_token_accuracy": 0.22554265558719636,
      "num_tokens": 69447711.0,
      "step": 33465
    },
    {
      "entropy": 5.590954446792603,
      "epoch": 3.0361030478955007,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004088173529206914,
      "loss": 4.8159,
      "mean_token_accuracy": 0.2266578882932663,
      "num_tokens": 69457692.0,
      "step": 33470
    },
    {
      "entropy": 5.548431253433227,
      "epoch": 3.036556603773585,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004087913397158809,
      "loss": 4.8145,
      "mean_token_accuracy": 0.21738813817501068,
      "num_tokens": 69468034.0,
      "step": 33475
    },
    {
      "entropy": 5.490089559555054,
      "epoch": 3.037010159651669,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040876532374423135,
      "loss": 4.8477,
      "mean_token_accuracy": 0.22133027613162995,
      "num_tokens": 69477719.0,
      "step": 33480
    },
    {
      "entropy": 5.546320056915283,
      "epoch": 3.0374637155297535,
      "grad_norm": 1.25,
      "learning_rate": 0.0004087393050062809,
      "loss": 4.871,
      "mean_token_accuracy": 0.21676785349845887,
      "num_tokens": 69487261.0,
      "step": 33485
    },
    {
      "entropy": 5.570474863052368,
      "epoch": 3.0379172714078373,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040871328350256745,
      "loss": 4.8471,
      "mean_token_accuracy": 0.2211972251534462,
      "num_tokens": 69496658.0,
      "step": 33490
    },
    {
      "entropy": 5.6306154251098635,
      "epoch": 3.0383708272859216,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040868725923362927,
      "loss": 5.0096,
      "mean_token_accuracy": 0.21236396580934525,
      "num_tokens": 69507197.0,
      "step": 33495
    },
    {
      "entropy": 5.542728614807129,
      "epoch": 3.038824383164006,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040866123220000455,
      "loss": 4.9201,
      "mean_token_accuracy": 0.21712354123592376,
      "num_tokens": 69516905.0,
      "step": 33500
    },
    {
      "entropy": 5.579092884063721,
      "epoch": 3.03927793904209,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040863520240223144,
      "loss": 4.8165,
      "mean_token_accuracy": 0.22702620178461075,
      "num_tokens": 69526869.0,
      "step": 33505
    },
    {
      "entropy": 5.5965564250946045,
      "epoch": 3.0397314949201744,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004086091698408483,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21850751489400863,
      "num_tokens": 69537058.0,
      "step": 33510
    },
    {
      "entropy": 5.616084337234497,
      "epoch": 3.040185050798258,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040858313451639343,
      "loss": 4.9736,
      "mean_token_accuracy": 0.20910998731851577,
      "num_tokens": 69547617.0,
      "step": 33515
    },
    {
      "entropy": 5.594964599609375,
      "epoch": 3.0406386066763424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004085570964294053,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2093049630522728,
      "num_tokens": 69558733.0,
      "step": 33520
    },
    {
      "entropy": 5.641829538345337,
      "epoch": 3.0410921625544267,
      "grad_norm": 1.125,
      "learning_rate": 0.0004085310555804225,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21158185601234436,
      "num_tokens": 69570011.0,
      "step": 33525
    },
    {
      "entropy": 5.6617076873779295,
      "epoch": 3.041545718432511,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040850501196998336,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20798504650592803,
      "num_tokens": 69580894.0,
      "step": 33530
    },
    {
      "entropy": 5.587724399566651,
      "epoch": 3.0419992743105952,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040847896559862654,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2174672544002533,
      "num_tokens": 69591438.0,
      "step": 33535
    },
    {
      "entropy": 5.569055795669556,
      "epoch": 3.042452830188679,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004084529164668907,
      "loss": 4.9076,
      "mean_token_accuracy": 0.2176385700702667,
      "num_tokens": 69601745.0,
      "step": 33540
    },
    {
      "entropy": 5.589884185791016,
      "epoch": 3.0429063860667633,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040842686457531453,
      "loss": 4.873,
      "mean_token_accuracy": 0.21928049474954606,
      "num_tokens": 69611872.0,
      "step": 33545
    },
    {
      "entropy": 5.5757533550262455,
      "epoch": 3.0433599419448476,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004084008099244368,
      "loss": 4.8527,
      "mean_token_accuracy": 0.21931985020637512,
      "num_tokens": 69621673.0,
      "step": 33550
    },
    {
      "entropy": 5.508589172363282,
      "epoch": 3.043813497822932,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040837475251479633,
      "loss": 4.8094,
      "mean_token_accuracy": 0.21971536576747894,
      "num_tokens": 69632090.0,
      "step": 33555
    },
    {
      "entropy": 5.5114587306976315,
      "epoch": 3.044267053701016,
      "grad_norm": 1.171875,
      "learning_rate": 0.000408348692346932,
      "loss": 4.8425,
      "mean_token_accuracy": 0.22412486523389816,
      "num_tokens": 69642920.0,
      "step": 33560
    },
    {
      "entropy": 5.563113641738892,
      "epoch": 3.0447206095791,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040832262942138277,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2107337847352028,
      "num_tokens": 69653936.0,
      "step": 33565
    },
    {
      "entropy": 5.521078824996948,
      "epoch": 3.0451741654571842,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004082965637386875,
      "loss": 4.8154,
      "mean_token_accuracy": 0.22167031317949296,
      "num_tokens": 69663362.0,
      "step": 33570
    },
    {
      "entropy": 5.5225889682769775,
      "epoch": 3.0456277213352685,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004082704952993854,
      "loss": 4.8278,
      "mean_token_accuracy": 0.22145363092422485,
      "num_tokens": 69673178.0,
      "step": 33575
    },
    {
      "entropy": 5.606073045730591,
      "epoch": 3.0460812772133528,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004082444241040155,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22147749960422516,
      "num_tokens": 69683627.0,
      "step": 33580
    },
    {
      "entropy": 5.534637928009033,
      "epoch": 3.046534833091437,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004082183501531169,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21796853691339493,
      "num_tokens": 69694010.0,
      "step": 33585
    },
    {
      "entropy": 5.551808309555054,
      "epoch": 3.046988388969521,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040819227344722893,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21642842292785644,
      "num_tokens": 69704888.0,
      "step": 33590
    },
    {
      "entropy": 5.510912322998047,
      "epoch": 3.047441944847605,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040816619398689074,
      "loss": 4.847,
      "mean_token_accuracy": 0.22189804762601853,
      "num_tokens": 69714718.0,
      "step": 33595
    },
    {
      "entropy": 5.535586261749268,
      "epoch": 3.0478955007256894,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040814011177264174,
      "loss": 4.8509,
      "mean_token_accuracy": 0.21921432465314866,
      "num_tokens": 69726076.0,
      "step": 33600
    },
    {
      "entropy": 5.514611434936524,
      "epoch": 3.0483490566037736,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040811402680502137,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22057845145463945,
      "num_tokens": 69736507.0,
      "step": 33605
    },
    {
      "entropy": 5.470878219604492,
      "epoch": 3.048802612481858,
      "grad_norm": 1.390625,
      "learning_rate": 0.00040808793908456885,
      "loss": 4.8419,
      "mean_token_accuracy": 0.22356429547071457,
      "num_tokens": 69746061.0,
      "step": 33610
    },
    {
      "entropy": 5.528071117401123,
      "epoch": 3.0492561683599417,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004080618486118241,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2178589478135109,
      "num_tokens": 69756493.0,
      "step": 33615
    },
    {
      "entropy": 5.564393997192383,
      "epoch": 3.049709724238026,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004080357553873261,
      "loss": 4.8423,
      "mean_token_accuracy": 0.2219926416873932,
      "num_tokens": 69766937.0,
      "step": 33620
    },
    {
      "entropy": 5.519056034088135,
      "epoch": 3.0501632801161103,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000408009659411615,
      "loss": 4.8591,
      "mean_token_accuracy": 0.21903112828731536,
      "num_tokens": 69776872.0,
      "step": 33625
    },
    {
      "entropy": 5.541478109359741,
      "epoch": 3.0506168359941945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004079835606852301,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2190685823559761,
      "num_tokens": 69788145.0,
      "step": 33630
    },
    {
      "entropy": 5.541377258300781,
      "epoch": 3.051070391872279,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004079574592087114,
      "loss": 4.8723,
      "mean_token_accuracy": 0.21633816212415696,
      "num_tokens": 69797353.0,
      "step": 33635
    },
    {
      "entropy": 5.518158769607544,
      "epoch": 3.051523947750363,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040793135498259854,
      "loss": 4.8569,
      "mean_token_accuracy": 0.22336736917495728,
      "num_tokens": 69806915.0,
      "step": 33640
    },
    {
      "entropy": 5.510580539703369,
      "epoch": 3.051977503628447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040790524800743134,
      "loss": 4.866,
      "mean_token_accuracy": 0.21869227439165115,
      "num_tokens": 69817695.0,
      "step": 33645
    },
    {
      "entropy": 5.464036750793457,
      "epoch": 3.052431059506531,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004078791382837498,
      "loss": 4.7868,
      "mean_token_accuracy": 0.237777242064476,
      "num_tokens": 69827451.0,
      "step": 33650
    },
    {
      "entropy": 5.584479284286499,
      "epoch": 3.0528846153846154,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040785302581209374,
      "loss": 4.8741,
      "mean_token_accuracy": 0.21909368634223939,
      "num_tokens": 69837897.0,
      "step": 33655
    },
    {
      "entropy": 5.543436717987061,
      "epoch": 3.0533381712626997,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040782691059300326,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22179660201072693,
      "num_tokens": 69847378.0,
      "step": 33660
    },
    {
      "entropy": 5.549053001403808,
      "epoch": 3.053791727140784,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004078007926270185,
      "loss": 4.9173,
      "mean_token_accuracy": 0.21969237327575683,
      "num_tokens": 69857522.0,
      "step": 33665
    },
    {
      "entropy": 5.581828260421753,
      "epoch": 3.0542452830188678,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040777467191467946,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22366366982460023,
      "num_tokens": 69868384.0,
      "step": 33670
    },
    {
      "entropy": 5.543659687042236,
      "epoch": 3.054698838896952,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004077485484565264,
      "loss": 4.8812,
      "mean_token_accuracy": 0.2202191635966301,
      "num_tokens": 69878761.0,
      "step": 33675
    },
    {
      "entropy": 5.592347192764282,
      "epoch": 3.0551523947750363,
      "grad_norm": 1.25,
      "learning_rate": 0.0004077224222530994,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22071305960416793,
      "num_tokens": 69888366.0,
      "step": 33680
    },
    {
      "entropy": 5.5675581932067875,
      "epoch": 3.0556059506531206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000407696293304939,
      "loss": 4.8891,
      "mean_token_accuracy": 0.21480489671230316,
      "num_tokens": 69898489.0,
      "step": 33685
    },
    {
      "entropy": 5.512990140914917,
      "epoch": 3.056059506531205,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004076701616125854,
      "loss": 4.8463,
      "mean_token_accuracy": 0.21698013246059417,
      "num_tokens": 69908425.0,
      "step": 33690
    },
    {
      "entropy": 5.601546669006348,
      "epoch": 3.0565130624092887,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004076440271765791,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22019850462675095,
      "num_tokens": 69918488.0,
      "step": 33695
    },
    {
      "entropy": 5.483079195022583,
      "epoch": 3.056966618287373,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004076178899974605,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23036219477653502,
      "num_tokens": 69928831.0,
      "step": 33700
    },
    {
      "entropy": 5.548114585876465,
      "epoch": 3.057420174165457,
      "grad_norm": 1.125,
      "learning_rate": 0.00040759175007577017,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2161843404173851,
      "num_tokens": 69938974.0,
      "step": 33705
    },
    {
      "entropy": 5.537838315963745,
      "epoch": 3.0578737300435415,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004075656074120486,
      "loss": 4.8125,
      "mean_token_accuracy": 0.22158703953027725,
      "num_tokens": 69948804.0,
      "step": 33710
    },
    {
      "entropy": 5.578821754455566,
      "epoch": 3.0583272859216257,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004075394620068365,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2204371601343155,
      "num_tokens": 69958945.0,
      "step": 33715
    },
    {
      "entropy": 5.530725574493408,
      "epoch": 3.0587808417997095,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040751331386067456,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2199044868350029,
      "num_tokens": 69968826.0,
      "step": 33720
    },
    {
      "entropy": 5.576429843902588,
      "epoch": 3.059234397677794,
      "grad_norm": 1.25,
      "learning_rate": 0.0004074871629741035,
      "loss": 4.9131,
      "mean_token_accuracy": 0.21878617107868195,
      "num_tokens": 69978961.0,
      "step": 33725
    },
    {
      "entropy": 5.5038923740386965,
      "epoch": 3.059687953555878,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004074610093476641,
      "loss": 4.888,
      "mean_token_accuracy": 0.22178491055965424,
      "num_tokens": 69989848.0,
      "step": 33730
    },
    {
      "entropy": 5.532172203063965,
      "epoch": 3.0601415094339623,
      "grad_norm": 1.25,
      "learning_rate": 0.00040743485298189744,
      "loss": 4.8434,
      "mean_token_accuracy": 0.22129712402820587,
      "num_tokens": 69999607.0,
      "step": 33735
    },
    {
      "entropy": 5.554972457885742,
      "epoch": 3.0605950653120466,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040740869387734415,
      "loss": 4.8962,
      "mean_token_accuracy": 0.21320026516914367,
      "num_tokens": 70010446.0,
      "step": 33740
    },
    {
      "entropy": 5.503174209594727,
      "epoch": 3.0610486211901304,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004073825320345453,
      "loss": 4.8083,
      "mean_token_accuracy": 0.2243623897433281,
      "num_tokens": 70021148.0,
      "step": 33745
    },
    {
      "entropy": 5.595740747451782,
      "epoch": 3.0615021770682147,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040735636745404196,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2131958231329918,
      "num_tokens": 70030765.0,
      "step": 33750
    },
    {
      "entropy": 5.578330039978027,
      "epoch": 3.061955732946299,
      "grad_norm": 1.25,
      "learning_rate": 0.00040733020013637526,
      "loss": 4.8914,
      "mean_token_accuracy": 0.21820108592510223,
      "num_tokens": 70041717.0,
      "step": 33755
    },
    {
      "entropy": 5.592785024642945,
      "epoch": 3.0624092888243832,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040730403008208633,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2181514471769333,
      "num_tokens": 70052178.0,
      "step": 33760
    },
    {
      "entropy": 5.539941358566284,
      "epoch": 3.0628628447024675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004072778572917163,
      "loss": 4.8887,
      "mean_token_accuracy": 0.21673464179039,
      "num_tokens": 70062819.0,
      "step": 33765
    },
    {
      "entropy": 5.485319137573242,
      "epoch": 3.0633164005805513,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004072516817658065,
      "loss": 4.88,
      "mean_token_accuracy": 0.21750646084547043,
      "num_tokens": 70073966.0,
      "step": 33770
    },
    {
      "entropy": 5.564143800735474,
      "epoch": 3.0637699564586356,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040722550350489825,
      "loss": 4.874,
      "mean_token_accuracy": 0.21885902285575867,
      "num_tokens": 70083884.0,
      "step": 33775
    },
    {
      "entropy": 5.605109310150146,
      "epoch": 3.06422351233672,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004071993225095327,
      "loss": 4.8672,
      "mean_token_accuracy": 0.21771135032176972,
      "num_tokens": 70094910.0,
      "step": 33780
    },
    {
      "entropy": 5.50017614364624,
      "epoch": 3.064677068214804,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004071731387802517,
      "loss": 4.7133,
      "mean_token_accuracy": 0.237326018512249,
      "num_tokens": 70104480.0,
      "step": 33785
    },
    {
      "entropy": 5.501518249511719,
      "epoch": 3.0651306240928884,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004071469523175964,
      "loss": 4.8429,
      "mean_token_accuracy": 0.2214748591184616,
      "num_tokens": 70114003.0,
      "step": 33790
    },
    {
      "entropy": 5.564014291763305,
      "epoch": 3.065584179970972,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040712076312210844,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2124596655368805,
      "num_tokens": 70123604.0,
      "step": 33795
    },
    {
      "entropy": 5.549854183197022,
      "epoch": 3.0660377358490565,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040709457119432945,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21756191700696945,
      "num_tokens": 70133895.0,
      "step": 33800
    },
    {
      "entropy": 5.56729941368103,
      "epoch": 3.0664912917271407,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040706837653480113,
      "loss": 4.853,
      "mean_token_accuracy": 0.22682829052209855,
      "num_tokens": 70145063.0,
      "step": 33805
    },
    {
      "entropy": 5.543490362167359,
      "epoch": 3.066944847605225,
      "grad_norm": 1.296875,
      "learning_rate": 0.000407042179144065,
      "loss": 4.941,
      "mean_token_accuracy": 0.2149817779660225,
      "num_tokens": 70155456.0,
      "step": 33810
    },
    {
      "entropy": 5.449727344512939,
      "epoch": 3.0673984034833093,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040701597902266306,
      "loss": 4.8017,
      "mean_token_accuracy": 0.22480495423078536,
      "num_tokens": 70165427.0,
      "step": 33815
    },
    {
      "entropy": 5.6175901889801025,
      "epoch": 3.067851959361393,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000406989776171137,
      "loss": 4.94,
      "mean_token_accuracy": 0.21473718732595443,
      "num_tokens": 70175161.0,
      "step": 33820
    },
    {
      "entropy": 5.545640325546264,
      "epoch": 3.0683055152394774,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040696357059002877,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22670613527297973,
      "num_tokens": 70185026.0,
      "step": 33825
    },
    {
      "entropy": 5.554097318649292,
      "epoch": 3.0687590711175616,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004069373622798802,
      "loss": 4.856,
      "mean_token_accuracy": 0.23433369398117065,
      "num_tokens": 70195645.0,
      "step": 33830
    },
    {
      "entropy": 5.573609066009522,
      "epoch": 3.069212626995646,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004069111512412334,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21601267606019975,
      "num_tokens": 70206591.0,
      "step": 33835
    },
    {
      "entropy": 5.611148357391357,
      "epoch": 3.06966618287373,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040688493747463036,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22039080113172532,
      "num_tokens": 70216945.0,
      "step": 33840
    },
    {
      "entropy": 5.5124369144439695,
      "epoch": 3.0701197387518144,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004068587209806132,
      "loss": 4.7727,
      "mean_token_accuracy": 0.23457358479499818,
      "num_tokens": 70226983.0,
      "step": 33845
    },
    {
      "entropy": 5.505220508575439,
      "epoch": 3.0705732946298983,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040683250175972414,
      "loss": 4.8513,
      "mean_token_accuracy": 0.21699624508619308,
      "num_tokens": 70237730.0,
      "step": 33850
    },
    {
      "entropy": 5.559257316589355,
      "epoch": 3.0710268505079825,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004068062798125053,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21530958712100984,
      "num_tokens": 70248499.0,
      "step": 33855
    },
    {
      "entropy": 5.529851388931275,
      "epoch": 3.071480406386067,
      "grad_norm": 1.234375,
      "learning_rate": 0.000406780055139499,
      "loss": 4.8599,
      "mean_token_accuracy": 0.2240528181195259,
      "num_tokens": 70259294.0,
      "step": 33860
    },
    {
      "entropy": 5.590238332748413,
      "epoch": 3.071933962264151,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004067538277412476,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21715408861637114,
      "num_tokens": 70269400.0,
      "step": 33865
    },
    {
      "entropy": 5.624511766433716,
      "epoch": 3.0723875181422353,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004067275976182935,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2177477076649666,
      "num_tokens": 70279380.0,
      "step": 33870
    },
    {
      "entropy": 5.561013269424438,
      "epoch": 3.072841074020319,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040670136477117906,
      "loss": 4.872,
      "mean_token_accuracy": 0.2224866196513176,
      "num_tokens": 70289919.0,
      "step": 33875
    },
    {
      "entropy": 5.613761615753174,
      "epoch": 3.0732946298984034,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040667512920044684,
      "loss": 4.9686,
      "mean_token_accuracy": 0.21821146905422212,
      "num_tokens": 70300136.0,
      "step": 33880
    },
    {
      "entropy": 5.47821159362793,
      "epoch": 3.0737481857764877,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040664889090663944,
      "loss": 4.7717,
      "mean_token_accuracy": 0.23293721973896026,
      "num_tokens": 70310072.0,
      "step": 33885
    },
    {
      "entropy": 5.581581115722656,
      "epoch": 3.074201741654572,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040662264989029933,
      "loss": 4.877,
      "mean_token_accuracy": 0.22536811530590056,
      "num_tokens": 70320185.0,
      "step": 33890
    },
    {
      "entropy": 5.562211799621582,
      "epoch": 3.074655297532656,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004065964061519694,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22142807245254517,
      "num_tokens": 70330211.0,
      "step": 33895
    },
    {
      "entropy": 5.570418024063111,
      "epoch": 3.07510885341074,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004065701596921922,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2103681430220604,
      "num_tokens": 70341044.0,
      "step": 33900
    },
    {
      "entropy": 5.60465636253357,
      "epoch": 3.0755624092888243,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004065439105115106,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2235433742403984,
      "num_tokens": 70351320.0,
      "step": 33905
    },
    {
      "entropy": 5.533235883712768,
      "epoch": 3.0760159651669086,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040651765861046734,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23035095036029815,
      "num_tokens": 70361214.0,
      "step": 33910
    },
    {
      "entropy": 5.600605106353759,
      "epoch": 3.076469521044993,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040649140398960537,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22376378774642944,
      "num_tokens": 70371083.0,
      "step": 33915
    },
    {
      "entropy": 5.5534913539886475,
      "epoch": 3.076923076923077,
      "grad_norm": 1.25,
      "learning_rate": 0.00040646514664946775,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2149829313158989,
      "num_tokens": 70381987.0,
      "step": 33920
    },
    {
      "entropy": 5.514731121063233,
      "epoch": 3.077376632801161,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004064388865905974,
      "loss": 4.8234,
      "mean_token_accuracy": 0.22246827632188798,
      "num_tokens": 70393462.0,
      "step": 33925
    },
    {
      "entropy": 5.523762893676758,
      "epoch": 3.077830188679245,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004064126238135373,
      "loss": 4.7414,
      "mean_token_accuracy": 0.23506968915462495,
      "num_tokens": 70403664.0,
      "step": 33930
    },
    {
      "entropy": 5.530210018157959,
      "epoch": 3.0782837445573294,
      "grad_norm": 1.125,
      "learning_rate": 0.0004063863583188308,
      "loss": 4.8762,
      "mean_token_accuracy": 0.2177021622657776,
      "num_tokens": 70414588.0,
      "step": 33935
    },
    {
      "entropy": 5.513446283340454,
      "epoch": 3.0787373004354137,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004063600901070207,
      "loss": 4.928,
      "mean_token_accuracy": 0.21794008761644362,
      "num_tokens": 70425572.0,
      "step": 33940
    },
    {
      "entropy": 5.526793909072876,
      "epoch": 3.079190856313498,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040633381917865057,
      "loss": 4.846,
      "mean_token_accuracy": 0.20999166667461394,
      "num_tokens": 70436140.0,
      "step": 33945
    },
    {
      "entropy": 5.526464128494263,
      "epoch": 3.079644412191582,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004063075455342636,
      "loss": 4.841,
      "mean_token_accuracy": 0.22553852796554566,
      "num_tokens": 70447117.0,
      "step": 33950
    },
    {
      "entropy": 5.563627099990844,
      "epoch": 3.080097968069666,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004062812691744032,
      "loss": 4.8647,
      "mean_token_accuracy": 0.22334660440683365,
      "num_tokens": 70457756.0,
      "step": 33955
    },
    {
      "entropy": 5.555367088317871,
      "epoch": 3.0805515239477503,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004062549900996126,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21666637659072877,
      "num_tokens": 70468671.0,
      "step": 33960
    },
    {
      "entropy": 5.560975027084351,
      "epoch": 3.0810050798258346,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004062287083104354,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22531879991292952,
      "num_tokens": 70478212.0,
      "step": 33965
    },
    {
      "entropy": 5.5024810314178465,
      "epoch": 3.081458635703919,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040620242380741504,
      "loss": 4.8461,
      "mean_token_accuracy": 0.2202053412795067,
      "num_tokens": 70489062.0,
      "step": 33970
    },
    {
      "entropy": 5.564344358444214,
      "epoch": 3.0819121915820027,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004061761365910951,
      "loss": 4.912,
      "mean_token_accuracy": 0.2230479747056961,
      "num_tokens": 70499732.0,
      "step": 33975
    },
    {
      "entropy": 5.574536561965942,
      "epoch": 3.082365747460087,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040614984666201927,
      "loss": 4.8858,
      "mean_token_accuracy": 0.21846262514591216,
      "num_tokens": 70509777.0,
      "step": 33980
    },
    {
      "entropy": 5.5796490669250485,
      "epoch": 3.0828193033381712,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004061235540207312,
      "loss": 4.8998,
      "mean_token_accuracy": 0.21855698823928832,
      "num_tokens": 70520674.0,
      "step": 33985
    },
    {
      "entropy": 5.545018815994263,
      "epoch": 3.0832728592162555,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004060972586677746,
      "loss": 4.8683,
      "mean_token_accuracy": 0.21895248293876649,
      "num_tokens": 70531115.0,
      "step": 33990
    },
    {
      "entropy": 5.510321140289307,
      "epoch": 3.0837264150943398,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004060709606036933,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22225042879581453,
      "num_tokens": 70540726.0,
      "step": 33995
    },
    {
      "entropy": 5.555085849761963,
      "epoch": 3.084179970972424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004060446598290311,
      "loss": 4.8387,
      "mean_token_accuracy": 0.22498957216739654,
      "num_tokens": 70551227.0,
      "step": 34000
    },
    {
      "entropy": 5.488412189483642,
      "epoch": 3.084633526850508,
      "grad_norm": 1.21875,
      "learning_rate": 0.000406018356344332,
      "loss": 4.8546,
      "mean_token_accuracy": 0.21756804138422012,
      "num_tokens": 70560815.0,
      "step": 34005
    },
    {
      "entropy": 5.596951484680176,
      "epoch": 3.085087082728592,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004059920501501398,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2134745627641678,
      "num_tokens": 70570247.0,
      "step": 34010
    },
    {
      "entropy": 5.570509958267212,
      "epoch": 3.0855406386066764,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004059657412469987,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21469805091619493,
      "num_tokens": 70581497.0,
      "step": 34015
    },
    {
      "entropy": 5.533399534225464,
      "epoch": 3.0859941944847606,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004059394296354527,
      "loss": 4.8813,
      "mean_token_accuracy": 0.213567852973938,
      "num_tokens": 70593133.0,
      "step": 34020
    },
    {
      "entropy": 5.558902072906494,
      "epoch": 3.086447750362845,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004059131153160459,
      "loss": 4.8732,
      "mean_token_accuracy": 0.223296095430851,
      "num_tokens": 70603953.0,
      "step": 34025
    },
    {
      "entropy": 5.495404624938965,
      "epoch": 3.0869013062409287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004058867982893225,
      "loss": 4.8344,
      "mean_token_accuracy": 0.22647749185562133,
      "num_tokens": 70615292.0,
      "step": 34030
    },
    {
      "entropy": 5.614220380783081,
      "epoch": 3.087354862119013,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004058604785558268,
      "loss": 5.0003,
      "mean_token_accuracy": 0.20786291360855103,
      "num_tokens": 70626681.0,
      "step": 34035
    },
    {
      "entropy": 5.607170057296753,
      "epoch": 3.0878084179970973,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040583415611610305,
      "loss": 4.9138,
      "mean_token_accuracy": 0.21122521609067918,
      "num_tokens": 70637339.0,
      "step": 34040
    },
    {
      "entropy": 5.624288415908813,
      "epoch": 3.0882619738751815,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004058078309706956,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22347125560045242,
      "num_tokens": 70647980.0,
      "step": 34045
    },
    {
      "entropy": 5.523255395889282,
      "epoch": 3.088715529753266,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004057815031201488,
      "loss": 4.7241,
      "mean_token_accuracy": 0.23689903020858766,
      "num_tokens": 70657042.0,
      "step": 34050
    },
    {
      "entropy": 5.545471906661987,
      "epoch": 3.0891690856313496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004057551725650073,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2176677480340004,
      "num_tokens": 70668542.0,
      "step": 34055
    },
    {
      "entropy": 5.547145795822144,
      "epoch": 3.089622641509434,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040572883930581544,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21912151277065278,
      "num_tokens": 70679332.0,
      "step": 34060
    },
    {
      "entropy": 5.667421293258667,
      "epoch": 3.090076197387518,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004057025033431179,
      "loss": 5.0171,
      "mean_token_accuracy": 0.21360383033752442,
      "num_tokens": 70689897.0,
      "step": 34065
    },
    {
      "entropy": 5.547329998016357,
      "epoch": 3.0905297532656024,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040567616467745933,
      "loss": 4.8228,
      "mean_token_accuracy": 0.22579584419727325,
      "num_tokens": 70699965.0,
      "step": 34070
    },
    {
      "entropy": 5.591562366485595,
      "epoch": 3.0909833091436867,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004056498233093843,
      "loss": 4.943,
      "mean_token_accuracy": 0.22123480290174485,
      "num_tokens": 70710583.0,
      "step": 34075
    },
    {
      "entropy": 5.529441928863525,
      "epoch": 3.0914368650217705,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040562347923943766,
      "loss": 4.8096,
      "mean_token_accuracy": 0.2367551863193512,
      "num_tokens": 70721800.0,
      "step": 34080
    },
    {
      "entropy": 5.539307308197022,
      "epoch": 3.0918904208998548,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004055971324681642,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2146783247590065,
      "num_tokens": 70731527.0,
      "step": 34085
    },
    {
      "entropy": 5.589014148712158,
      "epoch": 3.092343976777939,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004055707829961087,
      "loss": 4.8882,
      "mean_token_accuracy": 0.21945282816886902,
      "num_tokens": 70740821.0,
      "step": 34090
    },
    {
      "entropy": 5.517031908035278,
      "epoch": 3.0927975326560233,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040554443082381627,
      "loss": 4.8406,
      "mean_token_accuracy": 0.22421352714300155,
      "num_tokens": 70750506.0,
      "step": 34095
    },
    {
      "entropy": 5.529907512664795,
      "epoch": 3.0932510885341076,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004055180759518315,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22267231792211534,
      "num_tokens": 70762009.0,
      "step": 34100
    },
    {
      "entropy": 5.501133632659912,
      "epoch": 3.0937046444121914,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004054917183806999,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22369867712259292,
      "num_tokens": 70772030.0,
      "step": 34105
    },
    {
      "entropy": 5.596992444992066,
      "epoch": 3.0941582002902757,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040546535811096626,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21585000604391097,
      "num_tokens": 70782539.0,
      "step": 34110
    },
    {
      "entropy": 5.709330654144287,
      "epoch": 3.09461175616836,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040543899514317564,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20582645535469055,
      "num_tokens": 70793433.0,
      "step": 34115
    },
    {
      "entropy": 5.561357498168945,
      "epoch": 3.095065312046444,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004054126294778735,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21783615052700042,
      "num_tokens": 70802919.0,
      "step": 34120
    },
    {
      "entropy": 5.498158979415893,
      "epoch": 3.0955188679245285,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040538626111560487,
      "loss": 4.838,
      "mean_token_accuracy": 0.22393683940172196,
      "num_tokens": 70813827.0,
      "step": 34125
    },
    {
      "entropy": 5.588648128509521,
      "epoch": 3.0959724238026123,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040535989005691515,
      "loss": 4.928,
      "mean_token_accuracy": 0.21952465921640396,
      "num_tokens": 70824117.0,
      "step": 34130
    },
    {
      "entropy": 5.460154390335083,
      "epoch": 3.0964259796806965,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040533351630234985,
      "loss": 4.818,
      "mean_token_accuracy": 0.22205373644828796,
      "num_tokens": 70833387.0,
      "step": 34135
    },
    {
      "entropy": 5.523236274719238,
      "epoch": 3.096879535558781,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000405307139852454,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22127053439617156,
      "num_tokens": 70844363.0,
      "step": 34140
    },
    {
      "entropy": 5.598881673812866,
      "epoch": 3.097333091436865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040528076070777335,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2141459196805954,
      "num_tokens": 70854759.0,
      "step": 34145
    },
    {
      "entropy": 5.48958101272583,
      "epoch": 3.0977866473149493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004052543788688534,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22366365492343904,
      "num_tokens": 70865682.0,
      "step": 34150
    },
    {
      "entropy": 5.626471710205078,
      "epoch": 3.098240203193033,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004052279943362396,
      "loss": 4.9161,
      "mean_token_accuracy": 0.217058165371418,
      "num_tokens": 70875449.0,
      "step": 34155
    },
    {
      "entropy": 5.600069332122803,
      "epoch": 3.0986937590711174,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004052016071104779,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22467945665121078,
      "num_tokens": 70885885.0,
      "step": 34160
    },
    {
      "entropy": 5.552690601348877,
      "epoch": 3.0991473149492017,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004051752171921136,
      "loss": 4.8404,
      "mean_token_accuracy": 0.2204853817820549,
      "num_tokens": 70896320.0,
      "step": 34165
    },
    {
      "entropy": 5.489721441268921,
      "epoch": 3.099600870827286,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040514882458169267,
      "loss": 4.7955,
      "mean_token_accuracy": 0.22706378847360612,
      "num_tokens": 70907074.0,
      "step": 34170
    },
    {
      "entropy": 5.521933650970459,
      "epoch": 3.1000544267053702,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004051224292797609,
      "loss": 4.9012,
      "mean_token_accuracy": 0.21267931759357453,
      "num_tokens": 70916567.0,
      "step": 34175
    },
    {
      "entropy": 5.559085655212402,
      "epoch": 3.100507982583454,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040509603128686413,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22927618473768235,
      "num_tokens": 70926163.0,
      "step": 34180
    },
    {
      "entropy": 5.604204940795898,
      "epoch": 3.1009615384615383,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004050696306035483,
      "loss": 4.8676,
      "mean_token_accuracy": 0.21646308451890944,
      "num_tokens": 70937358.0,
      "step": 34185
    },
    {
      "entropy": 5.502066898345947,
      "epoch": 3.1014150943396226,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040504322723035937,
      "loss": 4.8571,
      "mean_token_accuracy": 0.21675165891647338,
      "num_tokens": 70949305.0,
      "step": 34190
    },
    {
      "entropy": 5.477740621566772,
      "epoch": 3.101868650217707,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040501682116784326,
      "loss": 4.8337,
      "mean_token_accuracy": 0.2275699719786644,
      "num_tokens": 70959146.0,
      "step": 34195
    },
    {
      "entropy": 5.481984090805054,
      "epoch": 3.102322206095791,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040499041241654626,
      "loss": 4.903,
      "mean_token_accuracy": 0.21833328902721405,
      "num_tokens": 70968828.0,
      "step": 34200
    },
    {
      "entropy": 5.5220703125,
      "epoch": 3.1027757619738754,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004049640009770144,
      "loss": 4.8589,
      "mean_token_accuracy": 0.22461532503366471,
      "num_tokens": 70979733.0,
      "step": 34205
    },
    {
      "entropy": 5.6102495193481445,
      "epoch": 3.103229317851959,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004049375868497939,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22263315469026565,
      "num_tokens": 70989726.0,
      "step": 34210
    },
    {
      "entropy": 5.564761304855347,
      "epoch": 3.1036828737300435,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040491117003543087,
      "loss": 4.9395,
      "mean_token_accuracy": 0.21389199495315553,
      "num_tokens": 71000715.0,
      "step": 34215
    },
    {
      "entropy": 5.49772539138794,
      "epoch": 3.1041364296081277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040488475053447186,
      "loss": 4.7253,
      "mean_token_accuracy": 0.23314786702394485,
      "num_tokens": 71011453.0,
      "step": 34220
    },
    {
      "entropy": 5.569862747192383,
      "epoch": 3.104589985486212,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040485832834746306,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21896491795778275,
      "num_tokens": 71021477.0,
      "step": 34225
    },
    {
      "entropy": 5.496838235855103,
      "epoch": 3.1050435413642963,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040483190347495097,
      "loss": 4.8439,
      "mean_token_accuracy": 0.21613687574863433,
      "num_tokens": 71031706.0,
      "step": 34230
    },
    {
      "entropy": 5.5622227668762205,
      "epoch": 3.10549709724238,
      "grad_norm": 1.21875,
      "learning_rate": 0.000404805475917482,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22462908327579498,
      "num_tokens": 71042882.0,
      "step": 34235
    },
    {
      "entropy": 5.550917625427246,
      "epoch": 3.1059506531204644,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040477904567560275,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22683164328336716,
      "num_tokens": 71052804.0,
      "step": 34240
    },
    {
      "entropy": 5.613945436477661,
      "epoch": 3.1064042089985486,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004047526127498597,
      "loss": 4.9809,
      "mean_token_accuracy": 0.21108725070953369,
      "num_tokens": 71064349.0,
      "step": 34245
    },
    {
      "entropy": 5.547394847869873,
      "epoch": 3.106857764876633,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040472617714079964,
      "loss": 4.8962,
      "mean_token_accuracy": 0.21982925683259963,
      "num_tokens": 71075754.0,
      "step": 34250
    },
    {
      "entropy": 5.5435456275939945,
      "epoch": 3.107311320754717,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004046997388489691,
      "loss": 4.8344,
      "mean_token_accuracy": 0.2221750020980835,
      "num_tokens": 71085338.0,
      "step": 34255
    },
    {
      "entropy": 5.462207984924317,
      "epoch": 3.107764876632801,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004046732978749149,
      "loss": 4.827,
      "mean_token_accuracy": 0.22284100353717803,
      "num_tokens": 71097089.0,
      "step": 34260
    },
    {
      "entropy": 5.477288770675659,
      "epoch": 3.1082184325108853,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000404646854219184,
      "loss": 4.7953,
      "mean_token_accuracy": 0.22552372366189957,
      "num_tokens": 71106990.0,
      "step": 34265
    },
    {
      "entropy": 5.5535900592803955,
      "epoch": 3.1086719883889695,
      "grad_norm": 1.125,
      "learning_rate": 0.000404620407882323,
      "loss": 4.8319,
      "mean_token_accuracy": 0.2222897782921791,
      "num_tokens": 71117634.0,
      "step": 34270
    },
    {
      "entropy": 5.500090551376343,
      "epoch": 3.109125544267054,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040459395886487885,
      "loss": 4.7861,
      "mean_token_accuracy": 0.22173462957143783,
      "num_tokens": 71127717.0,
      "step": 34275
    },
    {
      "entropy": 5.535745096206665,
      "epoch": 3.109579100145138,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004045675071673988,
      "loss": 4.8015,
      "mean_token_accuracy": 0.2289023369550705,
      "num_tokens": 71137172.0,
      "step": 34280
    },
    {
      "entropy": 5.596367311477661,
      "epoch": 3.110032656023222,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004045410527904296,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21055590063333512,
      "num_tokens": 71146843.0,
      "step": 34285
    },
    {
      "entropy": 5.488558721542359,
      "epoch": 3.110486211901306,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004045145957345184,
      "loss": 4.7596,
      "mean_token_accuracy": 0.23404090106487274,
      "num_tokens": 71156619.0,
      "step": 34290
    },
    {
      "entropy": 5.551185750961304,
      "epoch": 3.1109397677793904,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004044881360002123,
      "loss": 4.8667,
      "mean_token_accuracy": 0.21848648637533188,
      "num_tokens": 71167234.0,
      "step": 34295
    },
    {
      "entropy": 5.5674694061279295,
      "epoch": 3.1113933236574747,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004044616735880586,
      "loss": 4.8526,
      "mean_token_accuracy": 0.22306906878948213,
      "num_tokens": 71178049.0,
      "step": 34300
    },
    {
      "entropy": 5.526227235794067,
      "epoch": 3.111846879535559,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040443520849860453,
      "loss": 4.7868,
      "mean_token_accuracy": 0.2195261836051941,
      "num_tokens": 71188568.0,
      "step": 34305
    },
    {
      "entropy": 5.550675106048584,
      "epoch": 3.1123004354136428,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040440874073239733,
      "loss": 4.8907,
      "mean_token_accuracy": 0.21608961671590804,
      "num_tokens": 71199004.0,
      "step": 34310
    },
    {
      "entropy": 5.489264965057373,
      "epoch": 3.112753991291727,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040438227028998436,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22214228808879852,
      "num_tokens": 71209239.0,
      "step": 34315
    },
    {
      "entropy": 5.546603393554688,
      "epoch": 3.1132075471698113,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040435579717191313,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22109529823064805,
      "num_tokens": 71219469.0,
      "step": 34320
    },
    {
      "entropy": 5.577359628677368,
      "epoch": 3.1136611030478956,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040432932137873107,
      "loss": 4.9351,
      "mean_token_accuracy": 0.2169351115822792,
      "num_tokens": 71230670.0,
      "step": 34325
    },
    {
      "entropy": 5.632251405715943,
      "epoch": 3.11411465892598,
      "grad_norm": 1.25,
      "learning_rate": 0.0004043028429109857,
      "loss": 4.968,
      "mean_token_accuracy": 0.21826846450567244,
      "num_tokens": 71241512.0,
      "step": 34330
    },
    {
      "entropy": 5.568650150299073,
      "epoch": 3.1145682148040637,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004042763617692245,
      "loss": 4.8253,
      "mean_token_accuracy": 0.2204554110765457,
      "num_tokens": 71250869.0,
      "step": 34335
    },
    {
      "entropy": 5.532284212112427,
      "epoch": 3.115021770682148,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040424987795399523,
      "loss": 4.8231,
      "mean_token_accuracy": 0.2191571369767189,
      "num_tokens": 71260577.0,
      "step": 34340
    },
    {
      "entropy": 5.553238725662231,
      "epoch": 3.115475326560232,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004042233914658455,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2218807727098465,
      "num_tokens": 71270185.0,
      "step": 34345
    },
    {
      "entropy": 5.505624008178711,
      "epoch": 3.1159288824383164,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040419690230532324,
      "loss": 4.828,
      "mean_token_accuracy": 0.2262366756796837,
      "num_tokens": 71279427.0,
      "step": 34350
    },
    {
      "entropy": 5.522997760772705,
      "epoch": 3.1163824383164007,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040417041047297606,
      "loss": 4.8188,
      "mean_token_accuracy": 0.22212269455194472,
      "num_tokens": 71290857.0,
      "step": 34355
    },
    {
      "entropy": 5.599454879760742,
      "epoch": 3.116835994194485,
      "grad_norm": 1.25,
      "learning_rate": 0.0004041439159693518,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2227695569396019,
      "num_tokens": 71300971.0,
      "step": 34360
    },
    {
      "entropy": 5.582221794128418,
      "epoch": 3.117289550072569,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004041174187949985,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2091573342680931,
      "num_tokens": 71311017.0,
      "step": 34365
    },
    {
      "entropy": 5.604055595397949,
      "epoch": 3.117743105950653,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040409091895046413,
      "loss": 4.959,
      "mean_token_accuracy": 0.21323770582675933,
      "num_tokens": 71321546.0,
      "step": 34370
    },
    {
      "entropy": 5.541482162475586,
      "epoch": 3.1181966618287373,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040406441643629656,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21633637249469756,
      "num_tokens": 71332345.0,
      "step": 34375
    },
    {
      "entropy": 5.609480524063111,
      "epoch": 3.1186502177068216,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040403791125304405,
      "loss": 4.9658,
      "mean_token_accuracy": 0.2216763287782669,
      "num_tokens": 71342670.0,
      "step": 34380
    },
    {
      "entropy": 5.539908599853516,
      "epoch": 3.119103773584906,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004040114034012546,
      "loss": 4.807,
      "mean_token_accuracy": 0.22060341387987137,
      "num_tokens": 71353440.0,
      "step": 34385
    },
    {
      "entropy": 5.549414873123169,
      "epoch": 3.1195573294629897,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040398489288147646,
      "loss": 4.933,
      "mean_token_accuracy": 0.22389867901802063,
      "num_tokens": 71363199.0,
      "step": 34390
    },
    {
      "entropy": 5.554641771316528,
      "epoch": 3.120010885341074,
      "grad_norm": 1.25,
      "learning_rate": 0.00040395837969425786,
      "loss": 4.8838,
      "mean_token_accuracy": 0.21970954686403274,
      "num_tokens": 71372925.0,
      "step": 34395
    },
    {
      "entropy": 5.581045722961425,
      "epoch": 3.1204644412191582,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004039318638401471,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21976799964904786,
      "num_tokens": 71383350.0,
      "step": 34400
    },
    {
      "entropy": 5.5178351402282715,
      "epoch": 3.1209179970972425,
      "grad_norm": 1.25,
      "learning_rate": 0.0004039053453196926,
      "loss": 4.8109,
      "mean_token_accuracy": 0.22881397604942322,
      "num_tokens": 71394006.0,
      "step": 34405
    },
    {
      "entropy": 5.496097803115845,
      "epoch": 3.1213715529753268,
      "grad_norm": 1.375,
      "learning_rate": 0.0004038788241334427,
      "loss": 4.8837,
      "mean_token_accuracy": 0.21768177598714827,
      "num_tokens": 71403945.0,
      "step": 34410
    },
    {
      "entropy": 5.594575262069702,
      "epoch": 3.1218251088534106,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004038523002819458,
      "loss": 4.9009,
      "mean_token_accuracy": 0.21371961683034896,
      "num_tokens": 71414238.0,
      "step": 34415
    },
    {
      "entropy": 5.528382253646851,
      "epoch": 3.122278664731495,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040382577376575055,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21522002220153807,
      "num_tokens": 71423586.0,
      "step": 34420
    },
    {
      "entropy": 5.527049160003662,
      "epoch": 3.122732220609579,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004037992445854054,
      "loss": 4.847,
      "mean_token_accuracy": 0.22468358725309373,
      "num_tokens": 71434261.0,
      "step": 34425
    },
    {
      "entropy": 5.5256880760192875,
      "epoch": 3.1231857764876634,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040377271274145916,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22078035324811934,
      "num_tokens": 71443952.0,
      "step": 34430
    },
    {
      "entropy": 5.575451517105103,
      "epoch": 3.1236393323657476,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004037461782344603,
      "loss": 4.8606,
      "mean_token_accuracy": 0.2240452364087105,
      "num_tokens": 71454051.0,
      "step": 34435
    },
    {
      "entropy": 5.538706588745117,
      "epoch": 3.1240928882438315,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040371964106495777,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22862416952848436,
      "num_tokens": 71464579.0,
      "step": 34440
    },
    {
      "entropy": 5.533867692947387,
      "epoch": 3.1245464441219157,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040369310123350015,
      "loss": 4.8741,
      "mean_token_accuracy": 0.219801427423954,
      "num_tokens": 71474631.0,
      "step": 34445
    },
    {
      "entropy": 5.634460496902466,
      "epoch": 3.125,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040366655874063645,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22250818461179733,
      "num_tokens": 71485266.0,
      "step": 34450
    },
    {
      "entropy": 5.495659637451172,
      "epoch": 3.1254535558780843,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004036400135869156,
      "loss": 4.7511,
      "mean_token_accuracy": 0.22743380814790726,
      "num_tokens": 71495248.0,
      "step": 34455
    },
    {
      "entropy": 5.519757986068726,
      "epoch": 3.1259071117561685,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040361346577288645,
      "loss": 4.8453,
      "mean_token_accuracy": 0.22273631989955903,
      "num_tokens": 71505600.0,
      "step": 34460
    },
    {
      "entropy": 5.555629396438599,
      "epoch": 3.1263606676342524,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040358691529909796,
      "loss": 4.8689,
      "mean_token_accuracy": 0.22039875090122224,
      "num_tokens": 71516634.0,
      "step": 34465
    },
    {
      "entropy": 5.611446332931519,
      "epoch": 3.1268142235123366,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040356036216609943,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2178143084049225,
      "num_tokens": 71526745.0,
      "step": 34470
    },
    {
      "entropy": 5.596765422821045,
      "epoch": 3.127267779390421,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040353380637443974,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20675043016672134,
      "num_tokens": 71537015.0,
      "step": 34475
    },
    {
      "entropy": 5.488035058975219,
      "epoch": 3.127721335268505,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040350724792466826,
      "loss": 4.7438,
      "mean_token_accuracy": 0.22461557388305664,
      "num_tokens": 71546660.0,
      "step": 34480
    },
    {
      "entropy": 5.542403507232666,
      "epoch": 3.1281748911465894,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004034806868173341,
      "loss": 4.895,
      "mean_token_accuracy": 0.21752571761608125,
      "num_tokens": 71557309.0,
      "step": 34485
    },
    {
      "entropy": 5.522272300720215,
      "epoch": 3.1286284470246732,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004034541230529866,
      "loss": 4.7718,
      "mean_token_accuracy": 0.2207999050617218,
      "num_tokens": 71567651.0,
      "step": 34490
    },
    {
      "entropy": 5.510285711288452,
      "epoch": 3.1290820029027575,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040342755663217507,
      "loss": 4.8118,
      "mean_token_accuracy": 0.23012871146202088,
      "num_tokens": 71577538.0,
      "step": 34495
    },
    {
      "entropy": 5.4960369110107425,
      "epoch": 3.1295355587808418,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000403400987555449,
      "loss": 4.8214,
      "mean_token_accuracy": 0.2278378963470459,
      "num_tokens": 71588299.0,
      "step": 34500
    },
    {
      "entropy": 5.524139738082885,
      "epoch": 3.129989114658926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004033744158233577,
      "loss": 4.9203,
      "mean_token_accuracy": 0.21831177324056625,
      "num_tokens": 71598989.0,
      "step": 34505
    },
    {
      "entropy": 5.564099311828613,
      "epoch": 3.1304426705370103,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040334784143645087,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2162409245967865,
      "num_tokens": 71608596.0,
      "step": 34510
    },
    {
      "entropy": 5.475565099716187,
      "epoch": 3.1308962264150946,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004033212643952779,
      "loss": 4.8158,
      "mean_token_accuracy": 0.22638071924448014,
      "num_tokens": 71619182.0,
      "step": 34515
    },
    {
      "entropy": 5.53093695640564,
      "epoch": 3.1313497822931784,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040329468470038857,
      "loss": 4.895,
      "mean_token_accuracy": 0.2178827315568924,
      "num_tokens": 71629526.0,
      "step": 34520
    },
    {
      "entropy": 5.583569431304932,
      "epoch": 3.1318033381712627,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004032681023523324,
      "loss": 4.94,
      "mean_token_accuracy": 0.2171468123793602,
      "num_tokens": 71640146.0,
      "step": 34525
    },
    {
      "entropy": 5.541940546035766,
      "epoch": 3.132256894049347,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004032415173516592,
      "loss": 4.7691,
      "mean_token_accuracy": 0.22338294088840485,
      "num_tokens": 71649652.0,
      "step": 34530
    },
    {
      "entropy": 5.536274814605713,
      "epoch": 3.132710449927431,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004032149296989187,
      "loss": 4.8643,
      "mean_token_accuracy": 0.22438554465770721,
      "num_tokens": 71659264.0,
      "step": 34535
    },
    {
      "entropy": 5.5584269046783445,
      "epoch": 3.133164005805515,
      "grad_norm": 1.25,
      "learning_rate": 0.00040318833939466083,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22244865149259568,
      "num_tokens": 71669175.0,
      "step": 34540
    },
    {
      "entropy": 5.520556020736694,
      "epoch": 3.1336175616835993,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004031617464394354,
      "loss": 4.8807,
      "mean_token_accuracy": 0.21735884249210358,
      "num_tokens": 71679537.0,
      "step": 34545
    },
    {
      "entropy": 5.568161821365356,
      "epoch": 3.1340711175616836,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004031351508337925,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22201237380504607,
      "num_tokens": 71690024.0,
      "step": 34550
    },
    {
      "entropy": 5.609154605865479,
      "epoch": 3.134524673439768,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004031085525782819,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2237461119890213,
      "num_tokens": 71700019.0,
      "step": 34555
    },
    {
      "entropy": 5.6059965133667,
      "epoch": 3.134978229317852,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004030819516734539,
      "loss": 4.9135,
      "mean_token_accuracy": 0.2089861661195755,
      "num_tokens": 71710570.0,
      "step": 34560
    },
    {
      "entropy": 5.558532810211181,
      "epoch": 3.1354317851959363,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040305534811985844,
      "loss": 4.9119,
      "mean_token_accuracy": 0.2248402014374733,
      "num_tokens": 71720619.0,
      "step": 34565
    },
    {
      "entropy": 5.512317943572998,
      "epoch": 3.13588534107402,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004030287419180457,
      "loss": 4.8488,
      "mean_token_accuracy": 0.2248748481273651,
      "num_tokens": 71730201.0,
      "step": 34570
    },
    {
      "entropy": 5.557100057601929,
      "epoch": 3.1363388969521044,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040300213306856606,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22717063128948212,
      "num_tokens": 71741368.0,
      "step": 34575
    },
    {
      "entropy": 5.5056891441345215,
      "epoch": 3.1367924528301887,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004029755215719696,
      "loss": 4.802,
      "mean_token_accuracy": 0.22738597840070723,
      "num_tokens": 71751361.0,
      "step": 34580
    },
    {
      "entropy": 5.585212469100952,
      "epoch": 3.137246008708273,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040294890742880677,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2245437756180763,
      "num_tokens": 71761164.0,
      "step": 34585
    },
    {
      "entropy": 5.521195125579834,
      "epoch": 3.1376995645863572,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040292229063962787,
      "loss": 4.8433,
      "mean_token_accuracy": 0.2285145953297615,
      "num_tokens": 71770726.0,
      "step": 34590
    },
    {
      "entropy": 5.59365439414978,
      "epoch": 3.138153120464441,
      "grad_norm": 1.25,
      "learning_rate": 0.0004028956712049834,
      "loss": 4.9192,
      "mean_token_accuracy": 0.2242444321513176,
      "num_tokens": 71781124.0,
      "step": 34595
    },
    {
      "entropy": 5.529942369461059,
      "epoch": 3.1386066763425253,
      "grad_norm": 1.1875,
      "learning_rate": 0.000402869049125424,
      "loss": 4.8583,
      "mean_token_accuracy": 0.21993466317653657,
      "num_tokens": 71791218.0,
      "step": 34600
    },
    {
      "entropy": 5.575375318527222,
      "epoch": 3.1390602322206096,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004028424244015,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22249774038791656,
      "num_tokens": 71801013.0,
      "step": 34605
    },
    {
      "entropy": 5.585590505599976,
      "epoch": 3.139513788098694,
      "grad_norm": 1.171875,
      "learning_rate": 0.000402815797033762,
      "loss": 4.9447,
      "mean_token_accuracy": 0.2149223878979683,
      "num_tokens": 71811742.0,
      "step": 34610
    },
    {
      "entropy": 5.510783624649048,
      "epoch": 3.139967343976778,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004027891670227609,
      "loss": 4.853,
      "mean_token_accuracy": 0.21937250941991807,
      "num_tokens": 71822265.0,
      "step": 34615
    },
    {
      "entropy": 5.537871313095093,
      "epoch": 3.140420899854862,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004027625343690472,
      "loss": 4.8268,
      "mean_token_accuracy": 0.2208012416958809,
      "num_tokens": 71832090.0,
      "step": 34620
    },
    {
      "entropy": 5.5265788555145265,
      "epoch": 3.140874455732946,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004027358990731717,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22246159017086028,
      "num_tokens": 71841381.0,
      "step": 34625
    },
    {
      "entropy": 5.5233588218688965,
      "epoch": 3.1413280116110305,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040270926113568524,
      "loss": 4.8743,
      "mean_token_accuracy": 0.21706269979476928,
      "num_tokens": 71851383.0,
      "step": 34630
    },
    {
      "entropy": 5.51249680519104,
      "epoch": 3.1417815674891147,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040268262055713873,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2236299678683281,
      "num_tokens": 71862588.0,
      "step": 34635
    },
    {
      "entropy": 5.523455667495727,
      "epoch": 3.142235123367199,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004026559773380831,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22294923067092895,
      "num_tokens": 71873975.0,
      "step": 34640
    },
    {
      "entropy": 5.612429761886597,
      "epoch": 3.142688679245283,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004026293314790694,
      "loss": 4.9239,
      "mean_token_accuracy": 0.21310394257307053,
      "num_tokens": 71884084.0,
      "step": 34645
    },
    {
      "entropy": 5.52652587890625,
      "epoch": 3.143142235123367,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004026026829806485,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2140487551689148,
      "num_tokens": 71894643.0,
      "step": 34650
    },
    {
      "entropy": 5.4806227684021,
      "epoch": 3.1435957910014514,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040257603184337167,
      "loss": 4.8008,
      "mean_token_accuracy": 0.22399496287107468,
      "num_tokens": 71905245.0,
      "step": 34655
    },
    {
      "entropy": 5.57176513671875,
      "epoch": 3.1440493468795356,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040254937806778995,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21435303688049318,
      "num_tokens": 71916473.0,
      "step": 34660
    },
    {
      "entropy": 5.535465908050537,
      "epoch": 3.14450290275762,
      "grad_norm": 1.125,
      "learning_rate": 0.00040252272165445456,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2214602917432785,
      "num_tokens": 71927465.0,
      "step": 34665
    },
    {
      "entropy": 5.556954669952392,
      "epoch": 3.1449564586357037,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004024960626039168,
      "loss": 4.8793,
      "mean_token_accuracy": 0.2214064672589302,
      "num_tokens": 71937380.0,
      "step": 34670
    },
    {
      "entropy": 5.568004941940307,
      "epoch": 3.145410014513788,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040246940091672806,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22360762357711791,
      "num_tokens": 71948088.0,
      "step": 34675
    },
    {
      "entropy": 5.552810096740723,
      "epoch": 3.1458635703918723,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004024427365934395,
      "loss": 4.8308,
      "mean_token_accuracy": 0.22851835787296296,
      "num_tokens": 71958284.0,
      "step": 34680
    },
    {
      "entropy": 5.54061541557312,
      "epoch": 3.1463171262699565,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040241606963460283,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21367644071578978,
      "num_tokens": 71968138.0,
      "step": 34685
    },
    {
      "entropy": 5.5956042289733885,
      "epoch": 3.146770682148041,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040238940004076924,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22157119810581208,
      "num_tokens": 71978275.0,
      "step": 34690
    },
    {
      "entropy": 5.510547256469726,
      "epoch": 3.1472242380261246,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040236272781249047,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22388822585344315,
      "num_tokens": 71988875.0,
      "step": 34695
    },
    {
      "entropy": 5.541162109375,
      "epoch": 3.147677793904209,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040233605295031795,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22453829199075698,
      "num_tokens": 71999560.0,
      "step": 34700
    },
    {
      "entropy": 5.56273627281189,
      "epoch": 3.148131349782293,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004023093754548034,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22605786770582198,
      "num_tokens": 72009970.0,
      "step": 34705
    },
    {
      "entropy": 5.564571189880371,
      "epoch": 3.1485849056603774,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004022826953264986,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2246354714035988,
      "num_tokens": 72020196.0,
      "step": 34710
    },
    {
      "entropy": 5.526821184158325,
      "epoch": 3.1490384615384617,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040225601256595525,
      "loss": 4.7992,
      "mean_token_accuracy": 0.21854398995637894,
      "num_tokens": 72031456.0,
      "step": 34715
    },
    {
      "entropy": 5.581139945983887,
      "epoch": 3.149492017416546,
      "grad_norm": 1.21875,
      "learning_rate": 0.000402229327173725,
      "loss": 4.9331,
      "mean_token_accuracy": 0.21744608581066133,
      "num_tokens": 72042260.0,
      "step": 34720
    },
    {
      "entropy": 5.618552589416504,
      "epoch": 3.1499455732946298,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040220263915035997,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20093594789505004,
      "num_tokens": 72053463.0,
      "step": 34725
    },
    {
      "entropy": 5.6715727806091305,
      "epoch": 3.150399129172714,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040217594849641195,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2126060262322426,
      "num_tokens": 72064508.0,
      "step": 34730
    },
    {
      "entropy": 5.497507047653198,
      "epoch": 3.1508526850507983,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040214925521243287,
      "loss": 4.883,
      "mean_token_accuracy": 0.213919235765934,
      "num_tokens": 72074758.0,
      "step": 34735
    },
    {
      "entropy": 5.538395071029663,
      "epoch": 3.1513062409288826,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040212255929897474,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22325537204742432,
      "num_tokens": 72085548.0,
      "step": 34740
    },
    {
      "entropy": 5.58392562866211,
      "epoch": 3.1517597968069664,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004020958607565897,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2131912738084793,
      "num_tokens": 72096773.0,
      "step": 34745
    },
    {
      "entropy": 5.565107345581055,
      "epoch": 3.1522133526850507,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004020691595858299,
      "loss": 4.9248,
      "mean_token_accuracy": 0.218213951587677,
      "num_tokens": 72106832.0,
      "step": 34750
    },
    {
      "entropy": 5.568673610687256,
      "epoch": 3.152666908563135,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040204245578724754,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2204669386148453,
      "num_tokens": 72117758.0,
      "step": 34755
    },
    {
      "entropy": 5.5883142948150635,
      "epoch": 3.153120464441219,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040201574936139475,
      "loss": 4.8347,
      "mean_token_accuracy": 0.21859075129032135,
      "num_tokens": 72127522.0,
      "step": 34760
    },
    {
      "entropy": 5.510472536087036,
      "epoch": 3.1535740203193035,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004019890403088239,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22430847436189652,
      "num_tokens": 72138310.0,
      "step": 34765
    },
    {
      "entropy": 5.5198845863342285,
      "epoch": 3.1540275761973877,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040196232863008735,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22783053815364837,
      "num_tokens": 72148958.0,
      "step": 34770
    },
    {
      "entropy": 5.590095138549804,
      "epoch": 3.1544811320754715,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040193561432573754,
      "loss": 4.8642,
      "mean_token_accuracy": 0.2260219097137451,
      "num_tokens": 72158594.0,
      "step": 34775
    },
    {
      "entropy": 5.635335397720337,
      "epoch": 3.154934687953556,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004019088973963268,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21859782487154006,
      "num_tokens": 72168552.0,
      "step": 34780
    },
    {
      "entropy": 5.549925327301025,
      "epoch": 3.15538824383164,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040188217784240785,
      "loss": 4.8395,
      "mean_token_accuracy": 0.22059770673513412,
      "num_tokens": 72178584.0,
      "step": 34785
    },
    {
      "entropy": 5.5287374496459964,
      "epoch": 3.1558417997097243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000401855455664533,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22517748773097992,
      "num_tokens": 72189144.0,
      "step": 34790
    },
    {
      "entropy": 5.5797717571258545,
      "epoch": 3.1562953555878086,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004018287308632551,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22065042853355407,
      "num_tokens": 72198326.0,
      "step": 34795
    },
    {
      "entropy": 5.556631326675415,
      "epoch": 3.1567489114658924,
      "grad_norm": 1.375,
      "learning_rate": 0.0004018020034391266,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2223059594631195,
      "num_tokens": 72207916.0,
      "step": 34800
    },
    {
      "entropy": 5.608988618850708,
      "epoch": 3.1572024673439767,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004017752733927004,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22504230886697768,
      "num_tokens": 72219111.0,
      "step": 34805
    },
    {
      "entropy": 5.552172231674194,
      "epoch": 3.157656023222061,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004017485407245293,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22499859929084778,
      "num_tokens": 72230081.0,
      "step": 34810
    },
    {
      "entropy": 5.468620777130127,
      "epoch": 3.1581095791001452,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000401721805435166,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22468551397323608,
      "num_tokens": 72240693.0,
      "step": 34815
    },
    {
      "entropy": 5.481899642944336,
      "epoch": 3.1585631349782295,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040169506752516356,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23367743641138078,
      "num_tokens": 72250925.0,
      "step": 34820
    },
    {
      "entropy": 5.511834764480591,
      "epoch": 3.1590166908563133,
      "grad_norm": 1.25,
      "learning_rate": 0.0004016683269950748,
      "loss": 4.8046,
      "mean_token_accuracy": 0.2326040670275688,
      "num_tokens": 72261027.0,
      "step": 34825
    },
    {
      "entropy": 5.576647710800171,
      "epoch": 3.1594702467343976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040164158384545267,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23506312668323517,
      "num_tokens": 72269430.0,
      "step": 34830
    },
    {
      "entropy": 5.461353731155396,
      "epoch": 3.159923802612482,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004016148380768504,
      "loss": 4.8047,
      "mean_token_accuracy": 0.22523862719535828,
      "num_tokens": 72280207.0,
      "step": 34835
    },
    {
      "entropy": 5.5288159370422365,
      "epoch": 3.160377358490566,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040158808968982103,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2219698116183281,
      "num_tokens": 72290093.0,
      "step": 34840
    },
    {
      "entropy": 5.525186681747437,
      "epoch": 3.1608309143686504,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040156133868491766,
      "loss": 4.8053,
      "mean_token_accuracy": 0.22647631168365479,
      "num_tokens": 72300203.0,
      "step": 34845
    },
    {
      "entropy": 5.545291566848755,
      "epoch": 3.161284470246734,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040153458506269365,
      "loss": 4.8663,
      "mean_token_accuracy": 0.21986546218395234,
      "num_tokens": 72311778.0,
      "step": 34850
    },
    {
      "entropy": 5.5552538394927975,
      "epoch": 3.1617380261248185,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000401507828823702,
      "loss": 4.8572,
      "mean_token_accuracy": 0.2222330629825592,
      "num_tokens": 72322142.0,
      "step": 34855
    },
    {
      "entropy": 5.615859699249268,
      "epoch": 3.1621915820029027,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040148106996849636,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21920069307088852,
      "num_tokens": 72332312.0,
      "step": 34860
    },
    {
      "entropy": 5.464946699142456,
      "epoch": 3.162645137880987,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040145430849762993,
      "loss": 4.8077,
      "mean_token_accuracy": 0.2248086154460907,
      "num_tokens": 72342157.0,
      "step": 34865
    },
    {
      "entropy": 5.584286880493164,
      "epoch": 3.1630986937590713,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040142754441165615,
      "loss": 4.889,
      "mean_token_accuracy": 0.22141101956367493,
      "num_tokens": 72353417.0,
      "step": 34870
    },
    {
      "entropy": 5.596046209335327,
      "epoch": 3.1635522496371555,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004014007777111285,
      "loss": 4.8509,
      "mean_token_accuracy": 0.22268968522548677,
      "num_tokens": 72363744.0,
      "step": 34875
    },
    {
      "entropy": 5.555856084823608,
      "epoch": 3.1640058055152394,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040137400839660066,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21781760305166245,
      "num_tokens": 72375241.0,
      "step": 34880
    },
    {
      "entropy": 5.559301280975342,
      "epoch": 3.1644593613933236,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040134723646862607,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21870714873075486,
      "num_tokens": 72385475.0,
      "step": 34885
    },
    {
      "entropy": 5.555531167984009,
      "epoch": 3.164912917271408,
      "grad_norm": 1.125,
      "learning_rate": 0.00040132046192775835,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21874299943447112,
      "num_tokens": 72396417.0,
      "step": 34890
    },
    {
      "entropy": 5.577684831619263,
      "epoch": 3.165366473149492,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004012936847745514,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22032735049724578,
      "num_tokens": 72406169.0,
      "step": 34895
    },
    {
      "entropy": 5.558085775375366,
      "epoch": 3.165820029027576,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004012669050095589,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2150781363248825,
      "num_tokens": 72416522.0,
      "step": 34900
    },
    {
      "entropy": 5.5857851028442385,
      "epoch": 3.1662735849056602,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004012401226333345,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21964346170425414,
      "num_tokens": 72426000.0,
      "step": 34905
    },
    {
      "entropy": 5.6045068264007565,
      "epoch": 3.1667271407837445,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004012133376464322,
      "loss": 4.9705,
      "mean_token_accuracy": 0.2172339752316475,
      "num_tokens": 72435952.0,
      "step": 34910
    },
    {
      "entropy": 5.589573383331299,
      "epoch": 3.1671806966618288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040118655004940597,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22168498188257219,
      "num_tokens": 72449957.0,
      "step": 34915
    },
    {
      "entropy": 5.557134580612183,
      "epoch": 3.167634252539913,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004011597598428097,
      "loss": 4.889,
      "mean_token_accuracy": 0.22602726072072982,
      "num_tokens": 72459956.0,
      "step": 34920
    },
    {
      "entropy": 5.533215761184692,
      "epoch": 3.1680878084179973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004011329670271975,
      "loss": 4.8189,
      "mean_token_accuracy": 0.22595339119434357,
      "num_tokens": 72471107.0,
      "step": 34925
    },
    {
      "entropy": 5.502209091186524,
      "epoch": 3.168541364296081,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040110617160312334,
      "loss": 4.8421,
      "mean_token_accuracy": 0.22599032670259475,
      "num_tokens": 72481428.0,
      "step": 34930
    },
    {
      "entropy": 5.5806890487670895,
      "epoch": 3.1689949201741654,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004010793735711415,
      "loss": 4.9226,
      "mean_token_accuracy": 0.2110712170600891,
      "num_tokens": 72491133.0,
      "step": 34935
    },
    {
      "entropy": 5.500141620635986,
      "epoch": 3.1694484760522497,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040105257293180605,
      "loss": 4.8131,
      "mean_token_accuracy": 0.22647165805101394,
      "num_tokens": 72502047.0,
      "step": 34940
    },
    {
      "entropy": 5.593183851242065,
      "epoch": 3.169902031930334,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004010257696856712,
      "loss": 4.867,
      "mean_token_accuracy": 0.21808237284421922,
      "num_tokens": 72513120.0,
      "step": 34945
    },
    {
      "entropy": 5.627342224121094,
      "epoch": 3.170355587808418,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040099896383329135,
      "loss": 4.971,
      "mean_token_accuracy": 0.22320060580968856,
      "num_tokens": 72523118.0,
      "step": 34950
    },
    {
      "entropy": 5.602311086654663,
      "epoch": 3.170809143686502,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040097215537522086,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2184840112924576,
      "num_tokens": 72533697.0,
      "step": 34955
    },
    {
      "entropy": 5.526277256011963,
      "epoch": 3.1712626995645863,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040094534431201406,
      "loss": 4.8297,
      "mean_token_accuracy": 0.2328079089522362,
      "num_tokens": 72543661.0,
      "step": 34960
    },
    {
      "entropy": 5.588747167587281,
      "epoch": 3.1717162554426706,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004009185306442255,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2202364131808281,
      "num_tokens": 72554551.0,
      "step": 34965
    },
    {
      "entropy": 5.574983739852906,
      "epoch": 3.172169811320755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040089171437240954,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21238282471895217,
      "num_tokens": 72565134.0,
      "step": 34970
    },
    {
      "entropy": 5.552337646484375,
      "epoch": 3.172623367198839,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040086489549712097,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22483900785446168,
      "num_tokens": 72574981.0,
      "step": 34975
    },
    {
      "entropy": 5.595092535018921,
      "epoch": 3.173076923076923,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004008380740189142,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2197675108909607,
      "num_tokens": 72585073.0,
      "step": 34980
    },
    {
      "entropy": 5.586976146697998,
      "epoch": 3.173530478955007,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000400811249938344,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22212421596050264,
      "num_tokens": 72595719.0,
      "step": 34985
    },
    {
      "entropy": 5.599682950973511,
      "epoch": 3.1739840348330914,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004007844232559652,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22111351191997528,
      "num_tokens": 72605276.0,
      "step": 34990
    },
    {
      "entropy": 5.581099367141723,
      "epoch": 3.1744375907111757,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004007575939723324,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2199219286441803,
      "num_tokens": 72614964.0,
      "step": 34995
    },
    {
      "entropy": 5.5865983963012695,
      "epoch": 3.17489114658926,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004007307620880004,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22133212089538573,
      "num_tokens": 72625461.0,
      "step": 35000
    },
    {
      "entropy": 5.574169397354126,
      "epoch": 3.175344702467344,
      "grad_norm": 1.125,
      "learning_rate": 0.00040070392760352424,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22211068272590637,
      "num_tokens": 72636192.0,
      "step": 35005
    },
    {
      "entropy": 5.597622108459473,
      "epoch": 3.175798258345428,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040067709051945895,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22018223404884338,
      "num_tokens": 72648432.0,
      "step": 35010
    },
    {
      "entropy": 5.597772455215454,
      "epoch": 3.1762518142235123,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004006502508363593,
      "loss": 4.8458,
      "mean_token_accuracy": 0.2180676653981209,
      "num_tokens": 72658678.0,
      "step": 35015
    },
    {
      "entropy": 5.661026191711426,
      "epoch": 3.1767053701015966,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040062340855478046,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21105266511440277,
      "num_tokens": 72669191.0,
      "step": 35020
    },
    {
      "entropy": 5.5895562171936035,
      "epoch": 3.177158925979681,
      "grad_norm": 1.25,
      "learning_rate": 0.00040059656367527754,
      "loss": 4.8501,
      "mean_token_accuracy": 0.22149389684200288,
      "num_tokens": 72679327.0,
      "step": 35025
    },
    {
      "entropy": 5.593973016738891,
      "epoch": 3.1776124818577647,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004005697161984056,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22123364061117173,
      "num_tokens": 72689673.0,
      "step": 35030
    },
    {
      "entropy": 5.556650829315186,
      "epoch": 3.178066037735849,
      "grad_norm": 1.25,
      "learning_rate": 0.00040054286612472006,
      "loss": 4.9445,
      "mean_token_accuracy": 0.20845353752374648,
      "num_tokens": 72700606.0,
      "step": 35035
    },
    {
      "entropy": 5.61268835067749,
      "epoch": 3.178519593613933,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040051601345477593,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2196554571390152,
      "num_tokens": 72711195.0,
      "step": 35040
    },
    {
      "entropy": 5.605026912689209,
      "epoch": 3.1789731494920175,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040048915818912875,
      "loss": 4.9081,
      "mean_token_accuracy": 0.2199632927775383,
      "num_tokens": 72721160.0,
      "step": 35045
    },
    {
      "entropy": 5.535356855392456,
      "epoch": 3.1794267053701017,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040046230032833374,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21814224869012833,
      "num_tokens": 72730867.0,
      "step": 35050
    },
    {
      "entropy": 5.536412715911865,
      "epoch": 3.1798802612481856,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004004354398729464,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21642815321683884,
      "num_tokens": 72740453.0,
      "step": 35055
    },
    {
      "entropy": 5.616608333587647,
      "epoch": 3.18033381712627,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004004085768235222,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2130058228969574,
      "num_tokens": 72751395.0,
      "step": 35060
    },
    {
      "entropy": 5.584973859786987,
      "epoch": 3.180787373004354,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040038171118061673,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22164758592844008,
      "num_tokens": 72761300.0,
      "step": 35065
    },
    {
      "entropy": 5.557423400878906,
      "epoch": 3.1812409288824384,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004003548429447855,
      "loss": 4.8692,
      "mean_token_accuracy": 0.2269958645105362,
      "num_tokens": 72772011.0,
      "step": 35070
    },
    {
      "entropy": 5.5699926853179935,
      "epoch": 3.1816944847605226,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004003279721165841,
      "loss": 4.9029,
      "mean_token_accuracy": 0.21509736329317092,
      "num_tokens": 72782521.0,
      "step": 35075
    },
    {
      "entropy": 5.5035923480987545,
      "epoch": 3.182148040638607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004003010986965683,
      "loss": 4.7867,
      "mean_token_accuracy": 0.22660263478755951,
      "num_tokens": 72793125.0,
      "step": 35080
    },
    {
      "entropy": 5.62787766456604,
      "epoch": 3.1826015965166907,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000400274222685294,
      "loss": 4.9194,
      "mean_token_accuracy": 0.2203624278306961,
      "num_tokens": 72804541.0,
      "step": 35085
    },
    {
      "entropy": 5.698417901992798,
      "epoch": 3.183055152394775,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040024734408331664,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21887696385383607,
      "num_tokens": 72814447.0,
      "step": 35090
    },
    {
      "entropy": 5.594095611572266,
      "epoch": 3.1835087082728593,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040022046289119246,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22660548239946365,
      "num_tokens": 72823545.0,
      "step": 35095
    },
    {
      "entropy": 5.524208736419678,
      "epoch": 3.1839622641509435,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040019357910947714,
      "loss": 4.7707,
      "mean_token_accuracy": 0.22666800320148467,
      "num_tokens": 72833803.0,
      "step": 35100
    },
    {
      "entropy": 5.617096900939941,
      "epoch": 3.1844158200290273,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040016669273872666,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22041612565517427,
      "num_tokens": 72843383.0,
      "step": 35105
    },
    {
      "entropy": 5.631731224060059,
      "epoch": 3.1848693759071116,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040013980377949716,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21513321101665497,
      "num_tokens": 72853285.0,
      "step": 35110
    },
    {
      "entropy": 5.569305658340454,
      "epoch": 3.185322931785196,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040011291223234456,
      "loss": 4.947,
      "mean_token_accuracy": 0.21698566675186157,
      "num_tokens": 72863556.0,
      "step": 35115
    },
    {
      "entropy": 5.59058723449707,
      "epoch": 3.18577648766328,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040008601809782507,
      "loss": 4.8243,
      "mean_token_accuracy": 0.2241978883743286,
      "num_tokens": 72873296.0,
      "step": 35120
    },
    {
      "entropy": 5.611737680435181,
      "epoch": 3.1862300435413644,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004000591213764948,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2104458764195442,
      "num_tokens": 72884738.0,
      "step": 35125
    },
    {
      "entropy": 5.646515226364135,
      "epoch": 3.1866835994194487,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040003222206891017,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2151230528950691,
      "num_tokens": 72894573.0,
      "step": 35130
    },
    {
      "entropy": 5.506661128997803,
      "epoch": 3.1871371552975325,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004000053201756272,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2242366909980774,
      "num_tokens": 72904976.0,
      "step": 35135
    },
    {
      "entropy": 5.580720853805542,
      "epoch": 3.1875907111756168,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003999784156972024,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21811815053224565,
      "num_tokens": 72915152.0,
      "step": 35140
    },
    {
      "entropy": 5.604629564285278,
      "epoch": 3.188044267053701,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003999515086341921,
      "loss": 4.98,
      "mean_token_accuracy": 0.20945134311914443,
      "num_tokens": 72925450.0,
      "step": 35145
    },
    {
      "entropy": 5.526330614089966,
      "epoch": 3.1884978229317853,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003999245989871527,
      "loss": 4.8354,
      "mean_token_accuracy": 0.21800946295261384,
      "num_tokens": 72935990.0,
      "step": 35150
    },
    {
      "entropy": 5.523207330703736,
      "epoch": 3.1889513788098696,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039989768675664084,
      "loss": 4.7583,
      "mean_token_accuracy": 0.2233793780207634,
      "num_tokens": 72945662.0,
      "step": 35155
    },
    {
      "entropy": 5.510714197158814,
      "epoch": 3.1894049346879534,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039987077194321305,
      "loss": 4.7422,
      "mean_token_accuracy": 0.2282448634505272,
      "num_tokens": 72955313.0,
      "step": 35160
    },
    {
      "entropy": 5.627778434753418,
      "epoch": 3.1898584905660377,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039984385454742576,
      "loss": 4.995,
      "mean_token_accuracy": 0.21152457743883132,
      "num_tokens": 72965347.0,
      "step": 35165
    },
    {
      "entropy": 5.538826370239258,
      "epoch": 3.190312046444122,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039981693456983574,
      "loss": 4.8136,
      "mean_token_accuracy": 0.2261049672961235,
      "num_tokens": 72976794.0,
      "step": 35170
    },
    {
      "entropy": 5.541289377212524,
      "epoch": 3.190765602322206,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003997900120109998,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2250302866101265,
      "num_tokens": 72986837.0,
      "step": 35175
    },
    {
      "entropy": 5.561410236358642,
      "epoch": 3.1912191582002905,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003997630868714745,
      "loss": 4.8838,
      "mean_token_accuracy": 0.21482582688331603,
      "num_tokens": 72996819.0,
      "step": 35180
    },
    {
      "entropy": 5.507845735549926,
      "epoch": 3.1916727140783743,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039973615915181687,
      "loss": 4.7727,
      "mean_token_accuracy": 0.2298409655690193,
      "num_tokens": 73005963.0,
      "step": 35185
    },
    {
      "entropy": 5.546430444717407,
      "epoch": 3.1921262699564585,
      "grad_norm": 1.25,
      "learning_rate": 0.00039970922885258363,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2144476979970932,
      "num_tokens": 73015912.0,
      "step": 35190
    },
    {
      "entropy": 5.512864255905152,
      "epoch": 3.192579825834543,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003996822959743318,
      "loss": 4.7937,
      "mean_token_accuracy": 0.2349949970841408,
      "num_tokens": 73026528.0,
      "step": 35195
    },
    {
      "entropy": 5.582260179519653,
      "epoch": 3.193033381712627,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003996553605176183,
      "loss": 4.9326,
      "mean_token_accuracy": 0.2152012676000595,
      "num_tokens": 73036255.0,
      "step": 35200
    },
    {
      "entropy": 5.585186386108399,
      "epoch": 3.1934869375907113,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003996284224830002,
      "loss": 4.9827,
      "mean_token_accuracy": 0.22125737369060516,
      "num_tokens": 73046686.0,
      "step": 35205
    },
    {
      "entropy": 5.540912342071533,
      "epoch": 3.193940493468795,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039960148187103453,
      "loss": 4.793,
      "mean_token_accuracy": 0.22285069078207015,
      "num_tokens": 73056070.0,
      "step": 35210
    },
    {
      "entropy": 5.616141605377197,
      "epoch": 3.1943940493468794,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039957453868227845,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22896254807710648,
      "num_tokens": 73066288.0,
      "step": 35215
    },
    {
      "entropy": 5.580283546447754,
      "epoch": 3.1948476052249637,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039954759291728915,
      "loss": 4.887,
      "mean_token_accuracy": 0.22154993265867234,
      "num_tokens": 73076297.0,
      "step": 35220
    },
    {
      "entropy": 5.597363519668579,
      "epoch": 3.195301161103048,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039952064457662395,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21727415919303894,
      "num_tokens": 73086249.0,
      "step": 35225
    },
    {
      "entropy": 5.557369899749756,
      "epoch": 3.1957547169811322,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039949369366084,
      "loss": 4.85,
      "mean_token_accuracy": 0.22280970215797424,
      "num_tokens": 73096358.0,
      "step": 35230
    },
    {
      "entropy": 5.503918266296386,
      "epoch": 3.196208272859216,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039946674017049484,
      "loss": 4.8134,
      "mean_token_accuracy": 0.2199138045310974,
      "num_tokens": 73106891.0,
      "step": 35235
    },
    {
      "entropy": 5.46235466003418,
      "epoch": 3.1966618287373003,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039943978410614565,
      "loss": 4.8454,
      "mean_token_accuracy": 0.2196485459804535,
      "num_tokens": 73117106.0,
      "step": 35240
    },
    {
      "entropy": 5.563179969787598,
      "epoch": 3.1971153846153846,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039941282546835017,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21892719566822053,
      "num_tokens": 73127112.0,
      "step": 35245
    },
    {
      "entropy": 5.5766366481781,
      "epoch": 3.197568940493469,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039938586425766564,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22278990149497985,
      "num_tokens": 73137751.0,
      "step": 35250
    },
    {
      "entropy": 5.601244354248047,
      "epoch": 3.198022496371553,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003993589004746498,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22690341174602507,
      "num_tokens": 73149303.0,
      "step": 35255
    },
    {
      "entropy": 5.522311544418335,
      "epoch": 3.198476052249637,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039933193411986024,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22230055034160615,
      "num_tokens": 73160025.0,
      "step": 35260
    },
    {
      "entropy": 5.535379838943482,
      "epoch": 3.198929608127721,
      "grad_norm": 1.125,
      "learning_rate": 0.00039930496519385454,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22209324687719345,
      "num_tokens": 73170721.0,
      "step": 35265
    },
    {
      "entropy": 5.594349145889282,
      "epoch": 3.1993831640058055,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039927799369719047,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21740833222866057,
      "num_tokens": 73180902.0,
      "step": 35270
    },
    {
      "entropy": 5.490474081039428,
      "epoch": 3.1998367198838897,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003992510196304258,
      "loss": 4.746,
      "mean_token_accuracy": 0.22814437597990037,
      "num_tokens": 73191259.0,
      "step": 35275
    },
    {
      "entropy": 5.5359045505523685,
      "epoch": 3.200290275761974,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039922404299411846,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22672757655382156,
      "num_tokens": 73201654.0,
      "step": 35280
    },
    {
      "entropy": 5.5473448753356935,
      "epoch": 3.2007438316400583,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003991970637888262,
      "loss": 4.906,
      "mean_token_accuracy": 0.21040115654468536,
      "num_tokens": 73212658.0,
      "step": 35285
    },
    {
      "entropy": 5.606217527389527,
      "epoch": 3.201197387518142,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000399170082015107,
      "loss": 4.896,
      "mean_token_accuracy": 0.21643820106983186,
      "num_tokens": 73222883.0,
      "step": 35290
    },
    {
      "entropy": 5.54066801071167,
      "epoch": 3.2016509433962264,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000399143097673519,
      "loss": 4.8511,
      "mean_token_accuracy": 0.21891450583934785,
      "num_tokens": 73233731.0,
      "step": 35295
    },
    {
      "entropy": 5.5221171379089355,
      "epoch": 3.2021044992743106,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039911611076461995,
      "loss": 4.7939,
      "mean_token_accuracy": 0.2306176319718361,
      "num_tokens": 73243794.0,
      "step": 35300
    },
    {
      "entropy": 5.560508871078492,
      "epoch": 3.202558055152395,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039908912128896817,
      "loss": 4.8502,
      "mean_token_accuracy": 0.21862610131502153,
      "num_tokens": 73254708.0,
      "step": 35305
    },
    {
      "entropy": 5.500275564193726,
      "epoch": 3.203011611030479,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003990621292471217,
      "loss": 4.8713,
      "mean_token_accuracy": 0.21953899413347244,
      "num_tokens": 73265311.0,
      "step": 35310
    },
    {
      "entropy": 5.660466861724854,
      "epoch": 3.203465166908563,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039903513463963874,
      "loss": 5.0493,
      "mean_token_accuracy": 0.21423365026712418,
      "num_tokens": 73275886.0,
      "step": 35315
    },
    {
      "entropy": 5.6322455406188965,
      "epoch": 3.2039187227866472,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003990081374670777,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22120142430067063,
      "num_tokens": 73285903.0,
      "step": 35320
    },
    {
      "entropy": 5.5270287036895756,
      "epoch": 3.2043722786647315,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039898113772999665,
      "loss": 4.8642,
      "mean_token_accuracy": 0.21886866092681884,
      "num_tokens": 73296571.0,
      "step": 35325
    },
    {
      "entropy": 5.545783853530883,
      "epoch": 3.2048258345428158,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039895413542895427,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2103846400976181,
      "num_tokens": 73307105.0,
      "step": 35330
    },
    {
      "entropy": 5.523897981643676,
      "epoch": 3.2052793904209,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003989271305645086,
      "loss": 4.8316,
      "mean_token_accuracy": 0.22511658817529678,
      "num_tokens": 73316664.0,
      "step": 35335
    },
    {
      "entropy": 5.579960346221924,
      "epoch": 3.205732946298984,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039890012313721835,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22622746527194976,
      "num_tokens": 73327438.0,
      "step": 35340
    },
    {
      "entropy": 5.581797552108765,
      "epoch": 3.206186502177068,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000398873113147642,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22193843573331834,
      "num_tokens": 73337635.0,
      "step": 35345
    },
    {
      "entropy": 5.532924604415894,
      "epoch": 3.2066400580551524,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039884610059633807,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22390449792146683,
      "num_tokens": 73347032.0,
      "step": 35350
    },
    {
      "entropy": 5.635136127471924,
      "epoch": 3.2070936139332367,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003988190854838652,
      "loss": 4.9681,
      "mean_token_accuracy": 0.21327707320451736,
      "num_tokens": 73357780.0,
      "step": 35355
    },
    {
      "entropy": 5.6298424243927006,
      "epoch": 3.207547169811321,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039879206781078214,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21806975901126863,
      "num_tokens": 73367831.0,
      "step": 35360
    },
    {
      "entropy": 5.5899327278137205,
      "epoch": 3.2080007256894048,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003987650475776476,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22172315269708634,
      "num_tokens": 73378588.0,
      "step": 35365
    },
    {
      "entropy": 5.545698404312134,
      "epoch": 3.208454281567489,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039873802478502027,
      "loss": 4.8692,
      "mean_token_accuracy": 0.22284936904907227,
      "num_tokens": 73388414.0,
      "step": 35370
    },
    {
      "entropy": 5.557383966445923,
      "epoch": 3.2089078374455733,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003987109994334591,
      "loss": 4.862,
      "mean_token_accuracy": 0.22220928519964217,
      "num_tokens": 73398891.0,
      "step": 35375
    },
    {
      "entropy": 5.593102025985718,
      "epoch": 3.2093613933236576,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003986839715235229,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2166339486837387,
      "num_tokens": 73409605.0,
      "step": 35380
    },
    {
      "entropy": 5.636483192443848,
      "epoch": 3.209814949201742,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039865694105577063,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21568632423877715,
      "num_tokens": 73419103.0,
      "step": 35385
    },
    {
      "entropy": 5.499610090255738,
      "epoch": 3.2102685050798256,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003986299080307614,
      "loss": 4.873,
      "mean_token_accuracy": 0.21650802195072175,
      "num_tokens": 73429134.0,
      "step": 35390
    },
    {
      "entropy": 5.554768323898315,
      "epoch": 3.21072206095791,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003986028724490541,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21035676151514054,
      "num_tokens": 73440382.0,
      "step": 35395
    },
    {
      "entropy": 5.615599536895752,
      "epoch": 3.211175616835994,
      "grad_norm": 1.25,
      "learning_rate": 0.0003985758343112079,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22766050994396209,
      "num_tokens": 73451300.0,
      "step": 35400
    },
    {
      "entropy": 5.53894453048706,
      "epoch": 3.2116291727140784,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039854879361778193,
      "loss": 4.7933,
      "mean_token_accuracy": 0.21921161711215972,
      "num_tokens": 73461339.0,
      "step": 35405
    },
    {
      "entropy": 5.56909213066101,
      "epoch": 3.2120827285921627,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039852175036933545,
      "loss": 4.899,
      "mean_token_accuracy": 0.2208245202898979,
      "num_tokens": 73470961.0,
      "step": 35410
    },
    {
      "entropy": 5.6495246410369875,
      "epoch": 3.2125362844702465,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039849470456642777,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21246253103017806,
      "num_tokens": 73481492.0,
      "step": 35415
    },
    {
      "entropy": 5.529250001907348,
      "epoch": 3.212989840348331,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039846765620961806,
      "loss": 4.8382,
      "mean_token_accuracy": 0.21976940035820008,
      "num_tokens": 73492351.0,
      "step": 35420
    },
    {
      "entropy": 5.572080278396607,
      "epoch": 3.213443396226415,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003984406052994658,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21864311546087264,
      "num_tokens": 73503265.0,
      "step": 35425
    },
    {
      "entropy": 5.5797278881073,
      "epoch": 3.2138969521044993,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003984135518365303,
      "loss": 4.9154,
      "mean_token_accuracy": 0.21512652486562728,
      "num_tokens": 73513804.0,
      "step": 35430
    },
    {
      "entropy": 5.639270925521851,
      "epoch": 3.2143505079825836,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039838649582137113,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21635469943284988,
      "num_tokens": 73523453.0,
      "step": 35435
    },
    {
      "entropy": 5.644436550140381,
      "epoch": 3.214804063860668,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039835943725454777,
      "loss": 5.0051,
      "mean_token_accuracy": 0.20617130547761917,
      "num_tokens": 73533931.0,
      "step": 35440
    },
    {
      "entropy": 5.628381061553955,
      "epoch": 3.2152576197387517,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039833237613661984,
      "loss": 4.951,
      "mean_token_accuracy": 0.21855593621730804,
      "num_tokens": 73544562.0,
      "step": 35445
    },
    {
      "entropy": 5.595402526855469,
      "epoch": 3.215711175616836,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003983053124681469,
      "loss": 4.8933,
      "mean_token_accuracy": 0.2220338687300682,
      "num_tokens": 73555149.0,
      "step": 35450
    },
    {
      "entropy": 5.505956172943115,
      "epoch": 3.21616473149492,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039827824624968873,
      "loss": 4.8236,
      "mean_token_accuracy": 0.21488001346588134,
      "num_tokens": 73565131.0,
      "step": 35455
    },
    {
      "entropy": 5.5565910816192625,
      "epoch": 3.2166182873730045,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039825117748180493,
      "loss": 4.8568,
      "mean_token_accuracy": 0.2261882930994034,
      "num_tokens": 73574185.0,
      "step": 35460
    },
    {
      "entropy": 5.602627468109131,
      "epoch": 3.2170718432510883,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039822410616505545,
      "loss": 4.958,
      "mean_token_accuracy": 0.21362265795469285,
      "num_tokens": 73585733.0,
      "step": 35465
    },
    {
      "entropy": 5.66994161605835,
      "epoch": 3.2175253991291726,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039819703230000006,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21711804419755937,
      "num_tokens": 73597077.0,
      "step": 35470
    },
    {
      "entropy": 5.538753366470337,
      "epoch": 3.217978955007257,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003981699558871986,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22189857065677643,
      "num_tokens": 73607885.0,
      "step": 35475
    },
    {
      "entropy": 5.584167861938477,
      "epoch": 3.218432510885341,
      "grad_norm": 1.296875,
      "learning_rate": 0.000398142876927211,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21857452094554902,
      "num_tokens": 73618769.0,
      "step": 35480
    },
    {
      "entropy": 5.65290904045105,
      "epoch": 3.2188860667634254,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039811579542059744,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21279487609863282,
      "num_tokens": 73629345.0,
      "step": 35485
    },
    {
      "entropy": 5.549328279495239,
      "epoch": 3.2193396226415096,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039808871136791786,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2175550267100334,
      "num_tokens": 73639698.0,
      "step": 35490
    },
    {
      "entropy": 5.526072835922241,
      "epoch": 3.2197931785195935,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039806162476973227,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22099649757146836,
      "num_tokens": 73650954.0,
      "step": 35495
    },
    {
      "entropy": 5.646353721618652,
      "epoch": 3.2202467343976777,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000398034535626601,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2157284840941429,
      "num_tokens": 73661195.0,
      "step": 35500
    },
    {
      "entropy": 5.664518737792969,
      "epoch": 3.220700290275762,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039800744393908414,
      "loss": 5.0066,
      "mean_token_accuracy": 0.21799689084291457,
      "num_tokens": 73670512.0,
      "step": 35505
    },
    {
      "entropy": 5.603823232650757,
      "epoch": 3.2211538461538463,
      "grad_norm": 1.171875,
      "learning_rate": 0.000397980349707742,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21542716771364212,
      "num_tokens": 73681927.0,
      "step": 35510
    },
    {
      "entropy": 5.610978698730468,
      "epoch": 3.2216074020319305,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039795325293313493,
      "loss": 4.9399,
      "mean_token_accuracy": 0.21248573511838914,
      "num_tokens": 73692025.0,
      "step": 35515
    },
    {
      "entropy": 5.571760845184326,
      "epoch": 3.2220609579100143,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039792615361582324,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2161061093211174,
      "num_tokens": 73702278.0,
      "step": 35520
    },
    {
      "entropy": 5.660702657699585,
      "epoch": 3.2225145137880986,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039789905175636726,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20843842774629592,
      "num_tokens": 73713511.0,
      "step": 35525
    },
    {
      "entropy": 5.494663047790527,
      "epoch": 3.222968069666183,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039787194735532765,
      "loss": 4.7952,
      "mean_token_accuracy": 0.22784166038036346,
      "num_tokens": 73723683.0,
      "step": 35530
    },
    {
      "entropy": 5.586947727203369,
      "epoch": 3.223421625544267,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003978448404132649,
      "loss": 4.859,
      "mean_token_accuracy": 0.22581427842378615,
      "num_tokens": 73734191.0,
      "step": 35535
    },
    {
      "entropy": 5.602861738204956,
      "epoch": 3.2238751814223514,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039781773093073956,
      "loss": 4.8958,
      "mean_token_accuracy": 0.21986245214939118,
      "num_tokens": 73743747.0,
      "step": 35540
    },
    {
      "entropy": 5.507540369033814,
      "epoch": 3.2243287373004352,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003977906189083122,
      "loss": 4.8599,
      "mean_token_accuracy": 0.2210827574133873,
      "num_tokens": 73754466.0,
      "step": 35545
    },
    {
      "entropy": 5.550479555130005,
      "epoch": 3.2247822931785195,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003977635043465435,
      "loss": 4.8453,
      "mean_token_accuracy": 0.2201464742422104,
      "num_tokens": 73765948.0,
      "step": 35550
    },
    {
      "entropy": 5.456330347061157,
      "epoch": 3.2252358490566038,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003977363872459943,
      "loss": 4.8352,
      "mean_token_accuracy": 0.22234871089458466,
      "num_tokens": 73776003.0,
      "step": 35555
    },
    {
      "entropy": 5.62430591583252,
      "epoch": 3.225689404934688,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003977092676072254,
      "loss": 5.0418,
      "mean_token_accuracy": 0.21233248114585876,
      "num_tokens": 73787821.0,
      "step": 35560
    },
    {
      "entropy": 5.596877241134644,
      "epoch": 3.2261429608127723,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039768214543079746,
      "loss": 4.8876,
      "mean_token_accuracy": 0.21422988474369048,
      "num_tokens": 73799207.0,
      "step": 35565
    },
    {
      "entropy": 5.585151243209839,
      "epoch": 3.226596516690856,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003976550207172716,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22435288578271867,
      "num_tokens": 73809987.0,
      "step": 35570
    },
    {
      "entropy": 5.5061040878295895,
      "epoch": 3.2270500725689404,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039762789346720854,
      "loss": 4.8628,
      "mean_token_accuracy": 0.21346703618764878,
      "num_tokens": 73819075.0,
      "step": 35575
    },
    {
      "entropy": 5.555589962005615,
      "epoch": 3.2275036284470247,
      "grad_norm": 1.25,
      "learning_rate": 0.0003976007636811694,
      "loss": 4.9096,
      "mean_token_accuracy": 0.2242543578147888,
      "num_tokens": 73829831.0,
      "step": 35580
    },
    {
      "entropy": 5.586147880554199,
      "epoch": 3.227957184325109,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003975736313597153,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2226834088563919,
      "num_tokens": 73839826.0,
      "step": 35585
    },
    {
      "entropy": 5.557337903976441,
      "epoch": 3.228410740203193,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003975464965034072,
      "loss": 4.8167,
      "mean_token_accuracy": 0.2251250058412552,
      "num_tokens": 73850006.0,
      "step": 35590
    },
    {
      "entropy": 5.589664888381958,
      "epoch": 3.228864296081277,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003975193591128064,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2181267336010933,
      "num_tokens": 73860309.0,
      "step": 35595
    },
    {
      "entropy": 5.587058258056641,
      "epoch": 3.2293178519593613,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039749221918847406,
      "loss": 4.9745,
      "mean_token_accuracy": 0.2111610248684883,
      "num_tokens": 73871613.0,
      "step": 35600
    },
    {
      "entropy": 5.483116292953492,
      "epoch": 3.2297714078374455,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003974650767309713,
      "loss": 4.7968,
      "mean_token_accuracy": 0.2235073983669281,
      "num_tokens": 73882303.0,
      "step": 35605
    },
    {
      "entropy": 5.627234029769897,
      "epoch": 3.23022496371553,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039743793174085965,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2055797502398491,
      "num_tokens": 73893548.0,
      "step": 35610
    },
    {
      "entropy": 5.621247720718384,
      "epoch": 3.230678519593614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039741078421870024,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2164752170443535,
      "num_tokens": 73903934.0,
      "step": 35615
    },
    {
      "entropy": 5.6132568359375,
      "epoch": 3.231132075471698,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039738363416505474,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2117013528943062,
      "num_tokens": 73913813.0,
      "step": 35620
    },
    {
      "entropy": 5.598922395706177,
      "epoch": 3.231585631349782,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039735648158048444,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21685894429683686,
      "num_tokens": 73924339.0,
      "step": 35625
    },
    {
      "entropy": 5.602565336227417,
      "epoch": 3.2320391872278664,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003973293264655509,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21107783168554306,
      "num_tokens": 73935318.0,
      "step": 35630
    },
    {
      "entropy": 5.571333742141723,
      "epoch": 3.2324927431059507,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003973021688208158,
      "loss": 4.8664,
      "mean_token_accuracy": 0.22020641416311265,
      "num_tokens": 73945547.0,
      "step": 35635
    },
    {
      "entropy": 5.570557308197022,
      "epoch": 3.232946298984035,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003972750086468406,
      "loss": 4.902,
      "mean_token_accuracy": 0.2132236674427986,
      "num_tokens": 73956195.0,
      "step": 35640
    },
    {
      "entropy": 5.591267347335815,
      "epoch": 3.2333998548621192,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039724784594418714,
      "loss": 4.8283,
      "mean_token_accuracy": 0.23204536885023117,
      "num_tokens": 73966843.0,
      "step": 35645
    },
    {
      "entropy": 5.584470987319946,
      "epoch": 3.233853410740203,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000397220680713417,
      "loss": 4.8666,
      "mean_token_accuracy": 0.2210089534521103,
      "num_tokens": 73978516.0,
      "step": 35650
    },
    {
      "entropy": 5.505542135238647,
      "epoch": 3.2343069666182873,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039719351295509203,
      "loss": 4.8304,
      "mean_token_accuracy": 0.22288220077753068,
      "num_tokens": 73988376.0,
      "step": 35655
    },
    {
      "entropy": 5.547857475280762,
      "epoch": 3.2347605224963716,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039716634266977413,
      "loss": 4.8515,
      "mean_token_accuracy": 0.2306395560503006,
      "num_tokens": 73998232.0,
      "step": 35660
    },
    {
      "entropy": 5.528950357437134,
      "epoch": 3.235214078374456,
      "grad_norm": 1.125,
      "learning_rate": 0.0003971391698580252,
      "loss": 4.8965,
      "mean_token_accuracy": 0.2167160004377365,
      "num_tokens": 74008726.0,
      "step": 35665
    },
    {
      "entropy": 5.565345001220703,
      "epoch": 3.23566763425254,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039711199452040696,
      "loss": 4.8768,
      "mean_token_accuracy": 0.2212584510445595,
      "num_tokens": 74018746.0,
      "step": 35670
    },
    {
      "entropy": 5.586586141586304,
      "epoch": 3.236121190130624,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039708481665748157,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2201661378145218,
      "num_tokens": 74028615.0,
      "step": 35675
    },
    {
      "entropy": 5.5900920867919925,
      "epoch": 3.236574746008708,
      "grad_norm": 1.125,
      "learning_rate": 0.0003970576362698111,
      "loss": 4.9435,
      "mean_token_accuracy": 0.2127767339348793,
      "num_tokens": 74039219.0,
      "step": 35680
    },
    {
      "entropy": 5.579528951644898,
      "epoch": 3.2370283018867925,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039703045335795766,
      "loss": 4.8545,
      "mean_token_accuracy": 0.22033316344022752,
      "num_tokens": 74049385.0,
      "step": 35685
    },
    {
      "entropy": 5.642141103744507,
      "epoch": 3.2374818577648767,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003970032679224833,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20535451024770737,
      "num_tokens": 74059437.0,
      "step": 35690
    },
    {
      "entropy": 5.620270919799805,
      "epoch": 3.237935413642961,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039697607996395023,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21353611946105958,
      "num_tokens": 74068834.0,
      "step": 35695
    },
    {
      "entropy": 5.640044689178467,
      "epoch": 3.238388969521045,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003969488894829208,
      "loss": 4.926,
      "mean_token_accuracy": 0.2143849790096283,
      "num_tokens": 74079757.0,
      "step": 35700
    },
    {
      "entropy": 5.599082136154175,
      "epoch": 3.238842525399129,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003969216964799573,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22597562074661254,
      "num_tokens": 74091092.0,
      "step": 35705
    },
    {
      "entropy": 5.517797565460205,
      "epoch": 3.2392960812772134,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003968945009556219,
      "loss": 4.8526,
      "mean_token_accuracy": 0.22344509661197662,
      "num_tokens": 74102272.0,
      "step": 35710
    },
    {
      "entropy": 5.51051435470581,
      "epoch": 3.2397496371552976,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003968673029104773,
      "loss": 4.8363,
      "mean_token_accuracy": 0.22607310861349106,
      "num_tokens": 74113101.0,
      "step": 35715
    },
    {
      "entropy": 5.623663711547851,
      "epoch": 3.240203193033382,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003968401023450858,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22196249812841415,
      "num_tokens": 74123180.0,
      "step": 35720
    },
    {
      "entropy": 5.5582293510437015,
      "epoch": 3.2406567489114657,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039681289926000985,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22209576070308684,
      "num_tokens": 74133675.0,
      "step": 35725
    },
    {
      "entropy": 5.498506546020508,
      "epoch": 3.24111030478955,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003967856936558122,
      "loss": 4.811,
      "mean_token_accuracy": 0.22030034065246581,
      "num_tokens": 74143472.0,
      "step": 35730
    },
    {
      "entropy": 5.574520301818848,
      "epoch": 3.2415638606676342,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039675848553305526,
      "loss": 4.8702,
      "mean_token_accuracy": 0.2159616157412529,
      "num_tokens": 74154682.0,
      "step": 35735
    },
    {
      "entropy": 5.553822183609009,
      "epoch": 3.2420174165457185,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003967312748923019,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22248150110244752,
      "num_tokens": 74164745.0,
      "step": 35740
    },
    {
      "entropy": 5.538847160339356,
      "epoch": 3.2424709724238028,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039670406173411473,
      "loss": 4.8044,
      "mean_token_accuracy": 0.22437722384929656,
      "num_tokens": 74175388.0,
      "step": 35745
    },
    {
      "entropy": 5.553724193572998,
      "epoch": 3.2429245283018866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003966768460590566,
      "loss": 4.8448,
      "mean_token_accuracy": 0.21892396807670594,
      "num_tokens": 74185929.0,
      "step": 35750
    },
    {
      "entropy": 5.583537960052491,
      "epoch": 3.243378084179971,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039664962786769013,
      "loss": 4.8554,
      "mean_token_accuracy": 0.22650326788425446,
      "num_tokens": 74196165.0,
      "step": 35755
    },
    {
      "entropy": 5.591218519210815,
      "epoch": 3.243831640058055,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039662240716057854,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22238103151321412,
      "num_tokens": 74206843.0,
      "step": 35760
    },
    {
      "entropy": 5.578534984588623,
      "epoch": 3.2442851959361394,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003965951839382845,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22483187317848205,
      "num_tokens": 74216854.0,
      "step": 35765
    },
    {
      "entropy": 5.526900148391723,
      "epoch": 3.2447387518142237,
      "grad_norm": 1.1875,
      "learning_rate": 0.000396567958201371,
      "loss": 4.8563,
      "mean_token_accuracy": 0.22132735103368759,
      "num_tokens": 74226158.0,
      "step": 35770
    },
    {
      "entropy": 5.557786369323731,
      "epoch": 3.2451923076923075,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003965407299504013,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2221299722790718,
      "num_tokens": 74236860.0,
      "step": 35775
    },
    {
      "entropy": 5.575841856002808,
      "epoch": 3.2456458635703918,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003965134991859382,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22420749366283416,
      "num_tokens": 74247706.0,
      "step": 35780
    },
    {
      "entropy": 5.614956045150757,
      "epoch": 3.246099419448476,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000396486265908545,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21840111613273622,
      "num_tokens": 74259479.0,
      "step": 35785
    },
    {
      "entropy": 5.5454388618469235,
      "epoch": 3.2465529753265603,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003964590301187849,
      "loss": 4.926,
      "mean_token_accuracy": 0.21754302680492402,
      "num_tokens": 74270371.0,
      "step": 35790
    },
    {
      "entropy": 5.5508137226104735,
      "epoch": 3.2470065312046446,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003964317918172211,
      "loss": 4.8443,
      "mean_token_accuracy": 0.23269229978322983,
      "num_tokens": 74279637.0,
      "step": 35795
    },
    {
      "entropy": 5.610334205627441,
      "epoch": 3.247460087082729,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039640455100441684,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21622756272554397,
      "num_tokens": 74289889.0,
      "step": 35800
    },
    {
      "entropy": 5.619816017150879,
      "epoch": 3.2479136429608126,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039637730768093564,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22813587486743928,
      "num_tokens": 74300949.0,
      "step": 35805
    },
    {
      "entropy": 5.589548063278198,
      "epoch": 3.248367198838897,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003963500618473407,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2223416820168495,
      "num_tokens": 74312077.0,
      "step": 35810
    },
    {
      "entropy": 5.525066328048706,
      "epoch": 3.248820754716981,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003963228135041956,
      "loss": 4.9324,
      "mean_token_accuracy": 0.21726930439472197,
      "num_tokens": 74322610.0,
      "step": 35815
    },
    {
      "entropy": 5.506282472610474,
      "epoch": 3.2492743105950654,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039629556265206373,
      "loss": 4.8052,
      "mean_token_accuracy": 0.219085294008255,
      "num_tokens": 74331970.0,
      "step": 35820
    },
    {
      "entropy": 5.566981792449951,
      "epoch": 3.2497278664731493,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003962683092915088,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2264290913939476,
      "num_tokens": 74342812.0,
      "step": 35825
    },
    {
      "entropy": 5.552651643753052,
      "epoch": 3.2501814223512335,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003962410534230943,
      "loss": 4.8182,
      "mean_token_accuracy": 0.22734687179327012,
      "num_tokens": 74353786.0,
      "step": 35830
    },
    {
      "entropy": 5.610492944717407,
      "epoch": 3.250634978229318,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039621379504738385,
      "loss": 4.9284,
      "mean_token_accuracy": 0.21962619870901107,
      "num_tokens": 74363732.0,
      "step": 35835
    },
    {
      "entropy": 5.588639497756958,
      "epoch": 3.251088534107402,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039618653416494137,
      "loss": 4.9121,
      "mean_token_accuracy": 0.21881858110427857,
      "num_tokens": 74374077.0,
      "step": 35840
    },
    {
      "entropy": 5.581950712203979,
      "epoch": 3.2515420899854863,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039615927077633043,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2169348731637001,
      "num_tokens": 74384758.0,
      "step": 35845
    },
    {
      "entropy": 5.629990768432617,
      "epoch": 3.2519956458635706,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039613200488211483,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2218780294060707,
      "num_tokens": 74395811.0,
      "step": 35850
    },
    {
      "entropy": 5.605717182159424,
      "epoch": 3.2524492017416544,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039610473648285857,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2174245223402977,
      "num_tokens": 74407830.0,
      "step": 35855
    },
    {
      "entropy": 5.543435859680176,
      "epoch": 3.2529027576197387,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003960774655791255,
      "loss": 4.8405,
      "mean_token_accuracy": 0.2171020582318306,
      "num_tokens": 74417565.0,
      "step": 35860
    },
    {
      "entropy": 5.59379734992981,
      "epoch": 3.253356313497823,
      "grad_norm": 1.25,
      "learning_rate": 0.00039605019217147947,
      "loss": 4.907,
      "mean_token_accuracy": 0.2193870797753334,
      "num_tokens": 74428015.0,
      "step": 35865
    },
    {
      "entropy": 5.521070432662964,
      "epoch": 3.253809869375907,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039602291626048475,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22335321456193924,
      "num_tokens": 74439523.0,
      "step": 35870
    },
    {
      "entropy": 5.624947452545166,
      "epoch": 3.2542634252539915,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039599563784670526,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21045320481061935,
      "num_tokens": 74449897.0,
      "step": 35875
    },
    {
      "entropy": 5.538494396209717,
      "epoch": 3.2547169811320753,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003959683569307051,
      "loss": 4.8152,
      "mean_token_accuracy": 0.2272212862968445,
      "num_tokens": 74459888.0,
      "step": 35880
    },
    {
      "entropy": 5.527754449844361,
      "epoch": 3.2551705370101596,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003959410735130485,
      "loss": 4.7919,
      "mean_token_accuracy": 0.22100142985582352,
      "num_tokens": 74470894.0,
      "step": 35885
    },
    {
      "entropy": 5.5353187084197994,
      "epoch": 3.255624092888244,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039591378759429974,
      "loss": 4.7955,
      "mean_token_accuracy": 0.23311396241188048,
      "num_tokens": 74481306.0,
      "step": 35890
    },
    {
      "entropy": 5.606129121780396,
      "epoch": 3.256077648766328,
      "grad_norm": 1.21875,
      "learning_rate": 0.000395886499175023,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22075065970420837,
      "num_tokens": 74491048.0,
      "step": 35895
    },
    {
      "entropy": 5.539161014556885,
      "epoch": 3.2565312046444124,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039585920825578266,
      "loss": 4.8311,
      "mean_token_accuracy": 0.22583355754613876,
      "num_tokens": 74501656.0,
      "step": 35900
    },
    {
      "entropy": 5.595508670806884,
      "epoch": 3.256984760522496,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003958319148371431,
      "loss": 4.9119,
      "mean_token_accuracy": 0.21966097503900528,
      "num_tokens": 74513280.0,
      "step": 35905
    },
    {
      "entropy": 5.542800855636597,
      "epoch": 3.2574383164005805,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003958046189196687,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2159629926085472,
      "num_tokens": 74522898.0,
      "step": 35910
    },
    {
      "entropy": 5.654973268508911,
      "epoch": 3.2578918722786647,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039577732050392395,
      "loss": 4.9306,
      "mean_token_accuracy": 0.21430440098047257,
      "num_tokens": 74532551.0,
      "step": 35915
    },
    {
      "entropy": 5.633441543579101,
      "epoch": 3.258345428156749,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003957500195904735,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22329531461000443,
      "num_tokens": 74542285.0,
      "step": 35920
    },
    {
      "entropy": 5.632100439071655,
      "epoch": 3.2587989840348333,
      "grad_norm": 1.296875,
      "learning_rate": 0.000395722716179882,
      "loss": 4.9426,
      "mean_token_accuracy": 0.22196954637765884,
      "num_tokens": 74551982.0,
      "step": 35925
    },
    {
      "entropy": 5.589123392105103,
      "epoch": 3.259252539912917,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003956954102727138,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22074819952249528,
      "num_tokens": 74562190.0,
      "step": 35930
    },
    {
      "entropy": 5.50792236328125,
      "epoch": 3.2597060957910013,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039566810186953376,
      "loss": 4.8201,
      "mean_token_accuracy": 0.22470529675483703,
      "num_tokens": 74572673.0,
      "step": 35935
    },
    {
      "entropy": 5.481866073608399,
      "epoch": 3.2601596516690856,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003956407909709068,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22701329439878465,
      "num_tokens": 74582686.0,
      "step": 35940
    },
    {
      "entropy": 5.598630380630493,
      "epoch": 3.26061320754717,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039561347757739734,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21836612671613692,
      "num_tokens": 74593658.0,
      "step": 35945
    },
    {
      "entropy": 5.663361930847168,
      "epoch": 3.261066763425254,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039558616168957044,
      "loss": 4.9697,
      "mean_token_accuracy": 0.222406043112278,
      "num_tokens": 74603585.0,
      "step": 35950
    },
    {
      "entropy": 5.566291379928589,
      "epoch": 3.2615203193033384,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003955588433079911,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22361500263214112,
      "num_tokens": 74612806.0,
      "step": 35955
    },
    {
      "entropy": 5.583341789245606,
      "epoch": 3.2619738751814222,
      "grad_norm": 1.375,
      "learning_rate": 0.00039553152243322414,
      "loss": 4.884,
      "mean_token_accuracy": 0.22757286280393602,
      "num_tokens": 74623252.0,
      "step": 35960
    },
    {
      "entropy": 5.5419103622436525,
      "epoch": 3.2624274310595065,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003955041990658346,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22339923828840255,
      "num_tokens": 74633555.0,
      "step": 35965
    },
    {
      "entropy": 5.554964160919189,
      "epoch": 3.2628809869375908,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003954768732063874,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22323688715696335,
      "num_tokens": 74643322.0,
      "step": 35970
    },
    {
      "entropy": 5.564716768264771,
      "epoch": 3.263334542815675,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003954495448554478,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22901166677474977,
      "num_tokens": 74653461.0,
      "step": 35975
    },
    {
      "entropy": 5.599018526077271,
      "epoch": 3.263788098693759,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039542221401358104,
      "loss": 4.9067,
      "mean_token_accuracy": 0.21747326850891113,
      "num_tokens": 74663470.0,
      "step": 35980
    },
    {
      "entropy": 5.541120004653931,
      "epoch": 3.264241654571843,
      "grad_norm": 1.21875,
      "learning_rate": 0.000395394880681352,
      "loss": 4.8306,
      "mean_token_accuracy": 0.23152374625205993,
      "num_tokens": 74674074.0,
      "step": 35985
    },
    {
      "entropy": 5.634343576431275,
      "epoch": 3.2646952104499274,
      "grad_norm": 1.25,
      "learning_rate": 0.00039536754485932626,
      "loss": 4.9848,
      "mean_token_accuracy": 0.21333698481321334,
      "num_tokens": 74684968.0,
      "step": 35990
    },
    {
      "entropy": 5.5366740226745605,
      "epoch": 3.2651487663280117,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039534020654806906,
      "loss": 4.7995,
      "mean_token_accuracy": 0.23547699451446533,
      "num_tokens": 74694776.0,
      "step": 35995
    },
    {
      "entropy": 5.548335313796997,
      "epoch": 3.265602322206096,
      "grad_norm": 1.25,
      "learning_rate": 0.00039531286574814563,
      "loss": 4.8296,
      "mean_token_accuracy": 0.21822982877492905,
      "num_tokens": 74705506.0,
      "step": 36000
    },
    {
      "epoch": 3.265602322206096,
      "eval_entropy": 5.294058892198278,
      "eval_loss": 4.964892387390137,
      "eval_mean_token_accuracy": 0.2248716133347464,
      "eval_num_tokens": 74705506.0,
      "eval_runtime": 20.5595,
      "eval_samples_per_second": 1719.884,
      "eval_steps_per_second": 214.985,
      "step": 36000
    },
    {
      "entropy": 5.545662355422974,
      "epoch": 3.26605587808418,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003952855224601215,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2197272762656212,
      "num_tokens": 74715677.0,
      "step": 36005
    },
    {
      "entropy": 5.605515623092652,
      "epoch": 3.266509433962264,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003952581766845622,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22720492631196976,
      "num_tokens": 74725296.0,
      "step": 36010
    },
    {
      "entropy": 5.533355236053467,
      "epoch": 3.2669629898403483,
      "grad_norm": 1.25,
      "learning_rate": 0.0003952308284220331,
      "loss": 4.8335,
      "mean_token_accuracy": 0.22886460721492768,
      "num_tokens": 74736124.0,
      "step": 36015
    },
    {
      "entropy": 5.4804911613464355,
      "epoch": 3.2674165457184325,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039520347767309975,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21759670376777648,
      "num_tokens": 74746447.0,
      "step": 36020
    },
    {
      "entropy": 5.561615467071533,
      "epoch": 3.267870101596517,
      "grad_norm": 1.21875,
      "learning_rate": 0.000395176124438328,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2165571227669716,
      "num_tokens": 74756947.0,
      "step": 36025
    },
    {
      "entropy": 5.584126806259155,
      "epoch": 3.2683236574746006,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039514876871828326,
      "loss": 4.8082,
      "mean_token_accuracy": 0.23073782920837402,
      "num_tokens": 74767856.0,
      "step": 36030
    },
    {
      "entropy": 5.533388328552246,
      "epoch": 3.268777213352685,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039512141051353144,
      "loss": 4.857,
      "mean_token_accuracy": 0.22511068284511565,
      "num_tokens": 74778335.0,
      "step": 36035
    },
    {
      "entropy": 5.578894472122192,
      "epoch": 3.269230769230769,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003950940498246382,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22317657470703126,
      "num_tokens": 74788430.0,
      "step": 36040
    },
    {
      "entropy": 5.61752667427063,
      "epoch": 3.2696843251088534,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039506668665216944,
      "loss": 4.9738,
      "mean_token_accuracy": 0.21870671659708024,
      "num_tokens": 74800193.0,
      "step": 36045
    },
    {
      "entropy": 5.6183687210083,
      "epoch": 3.2701378809869377,
      "grad_norm": 1.25,
      "learning_rate": 0.000395039320996691,
      "loss": 4.922,
      "mean_token_accuracy": 0.2128383621573448,
      "num_tokens": 74810368.0,
      "step": 36050
    },
    {
      "entropy": 5.616834545135498,
      "epoch": 3.270591436865022,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039501195285876874,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22767363339662552,
      "num_tokens": 74820687.0,
      "step": 36055
    },
    {
      "entropy": 5.53968620300293,
      "epoch": 3.271044992743106,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003949845822389688,
      "loss": 4.8249,
      "mean_token_accuracy": 0.23022043704986572,
      "num_tokens": 74830489.0,
      "step": 36060
    },
    {
      "entropy": 5.484873628616333,
      "epoch": 3.27149854862119,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000394957209137857,
      "loss": 4.8091,
      "mean_token_accuracy": 0.2222230151295662,
      "num_tokens": 74840638.0,
      "step": 36065
    },
    {
      "entropy": 5.628452253341675,
      "epoch": 3.2719521044992743,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003949298335559997,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20446978360414506,
      "num_tokens": 74850600.0,
      "step": 36070
    },
    {
      "entropy": 5.601524400711059,
      "epoch": 3.2724056603773586,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003949024554939628,
      "loss": 4.889,
      "mean_token_accuracy": 0.22400035709142685,
      "num_tokens": 74861177.0,
      "step": 36075
    },
    {
      "entropy": 5.618128061294556,
      "epoch": 3.272859216255443,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039487507495231263,
      "loss": 4.9276,
      "mean_token_accuracy": 0.21720543503761292,
      "num_tokens": 74871747.0,
      "step": 36080
    },
    {
      "entropy": 5.529277706146241,
      "epoch": 3.2733127721335267,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003948476919316153,
      "loss": 4.9058,
      "mean_token_accuracy": 0.21479418575763704,
      "num_tokens": 74882796.0,
      "step": 36085
    },
    {
      "entropy": 5.612152099609375,
      "epoch": 3.273766328011611,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003948203064324371,
      "loss": 4.8905,
      "mean_token_accuracy": 0.21775318384170533,
      "num_tokens": 74894123.0,
      "step": 36090
    },
    {
      "entropy": 5.611866807937622,
      "epoch": 3.274219883889695,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039479291845534453,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21352228969335557,
      "num_tokens": 74904787.0,
      "step": 36095
    },
    {
      "entropy": 5.617761182785034,
      "epoch": 3.2746734397677795,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039476552800090384,
      "loss": 4.9578,
      "mean_token_accuracy": 0.21021807938814163,
      "num_tokens": 74915041.0,
      "step": 36100
    },
    {
      "entropy": 5.598668193817138,
      "epoch": 3.2751269956458637,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039473813506968146,
      "loss": 4.8506,
      "mean_token_accuracy": 0.21639005690813065,
      "num_tokens": 74926217.0,
      "step": 36105
    },
    {
      "entropy": 5.6326018333435055,
      "epoch": 3.2755805515239476,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039471073966224404,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21554133296012878,
      "num_tokens": 74936616.0,
      "step": 36110
    },
    {
      "entropy": 5.6091225147247314,
      "epoch": 3.276034107402032,
      "grad_norm": 1.171875,
      "learning_rate": 0.000394683341779158,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2122535675764084,
      "num_tokens": 74947443.0,
      "step": 36115
    },
    {
      "entropy": 5.56631326675415,
      "epoch": 3.276487663280116,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003946559414209899,
      "loss": 4.8646,
      "mean_token_accuracy": 0.2218962073326111,
      "num_tokens": 74957618.0,
      "step": 36120
    },
    {
      "entropy": 5.572535037994385,
      "epoch": 3.2769412191582004,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003946285385883065,
      "loss": 4.8142,
      "mean_token_accuracy": 0.22155697792768478,
      "num_tokens": 74967982.0,
      "step": 36125
    },
    {
      "entropy": 5.526893091201782,
      "epoch": 3.2773947750362846,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039460113328167434,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22548264712095262,
      "num_tokens": 74978801.0,
      "step": 36130
    },
    {
      "entropy": 5.519194650650024,
      "epoch": 3.2778483309143684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003945737255016603,
      "loss": 4.8255,
      "mean_token_accuracy": 0.22515552788972854,
      "num_tokens": 74989000.0,
      "step": 36135
    },
    {
      "entropy": 5.594220447540283,
      "epoch": 3.2783018867924527,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039454631524883123,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21769806295633315,
      "num_tokens": 74999664.0,
      "step": 36140
    },
    {
      "entropy": 5.603321361541748,
      "epoch": 3.278755442670537,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003945189025237538,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22431514263153077,
      "num_tokens": 75010187.0,
      "step": 36145
    },
    {
      "entropy": 5.657714939117431,
      "epoch": 3.2792089985486212,
      "grad_norm": 1.15625,
      "learning_rate": 0.000394491487326995,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21483942866325378,
      "num_tokens": 75021222.0,
      "step": 36150
    },
    {
      "entropy": 5.521349811553955,
      "epoch": 3.2796625544267055,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003944640696591219,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22486232817173005,
      "num_tokens": 75031136.0,
      "step": 36155
    },
    {
      "entropy": 5.492633533477783,
      "epoch": 3.28011611030479,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039443664952070135,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22387594729661942,
      "num_tokens": 75041545.0,
      "step": 36160
    },
    {
      "entropy": 5.558061599731445,
      "epoch": 3.2805696661828736,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039440922691230044,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22227349877357483,
      "num_tokens": 75051805.0,
      "step": 36165
    },
    {
      "entropy": 5.565986251831054,
      "epoch": 3.281023222060958,
      "grad_norm": 1.125,
      "learning_rate": 0.00039438180183448625,
      "loss": 4.863,
      "mean_token_accuracy": 0.2212325632572174,
      "num_tokens": 75062551.0,
      "step": 36170
    },
    {
      "entropy": 5.546606826782226,
      "epoch": 3.281476777939042,
      "grad_norm": 1.21875,
      "learning_rate": 0.000394354374287826,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22117944061756134,
      "num_tokens": 75073040.0,
      "step": 36175
    },
    {
      "entropy": 5.6007702350616455,
      "epoch": 3.2819303338171264,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039432694427288695,
      "loss": 4.8692,
      "mean_token_accuracy": 0.22687319964170455,
      "num_tokens": 75082543.0,
      "step": 36180
    },
    {
      "entropy": 5.632669925689697,
      "epoch": 3.28238388969521,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039429951179023614,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21692252904176712,
      "num_tokens": 75092480.0,
      "step": 36185
    },
    {
      "entropy": 5.522711324691772,
      "epoch": 3.2828374455732945,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039427207684044116,
      "loss": 4.7865,
      "mean_token_accuracy": 0.2230077087879181,
      "num_tokens": 75102445.0,
      "step": 36190
    },
    {
      "entropy": 5.566614389419556,
      "epoch": 3.2832910014513788,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039424463942406917,
      "loss": 4.96,
      "mean_token_accuracy": 0.21516534090042114,
      "num_tokens": 75112793.0,
      "step": 36195
    },
    {
      "entropy": 5.653137159347534,
      "epoch": 3.283744557329463,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003942171995416877,
      "loss": 4.95,
      "mean_token_accuracy": 0.21202856600284575,
      "num_tokens": 75123312.0,
      "step": 36200
    },
    {
      "entropy": 5.573168420791626,
      "epoch": 3.2841981132075473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039418975719386406,
      "loss": 4.8442,
      "mean_token_accuracy": 0.2201594054698944,
      "num_tokens": 75133584.0,
      "step": 36205
    },
    {
      "entropy": 5.559657096862793,
      "epoch": 3.2846516690856316,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003941623123811659,
      "loss": 4.876,
      "mean_token_accuracy": 0.21634067595005035,
      "num_tokens": 75144344.0,
      "step": 36210
    },
    {
      "entropy": 5.645954036712647,
      "epoch": 3.2851052249637154,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039413486510416087,
      "loss": 4.9081,
      "mean_token_accuracy": 0.21361046135425568,
      "num_tokens": 75155086.0,
      "step": 36215
    },
    {
      "entropy": 5.597145652770996,
      "epoch": 3.2855587808417996,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039410741536341634,
      "loss": 4.8496,
      "mean_token_accuracy": 0.2237755089998245,
      "num_tokens": 75166046.0,
      "step": 36220
    },
    {
      "entropy": 5.528163146972656,
      "epoch": 3.286012336719884,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039407996315950016,
      "loss": 4.8248,
      "mean_token_accuracy": 0.2245148479938507,
      "num_tokens": 75175388.0,
      "step": 36225
    },
    {
      "entropy": 5.547383975982666,
      "epoch": 3.286465892597968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039405250849297997,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2212709292769432,
      "num_tokens": 75186213.0,
      "step": 36230
    },
    {
      "entropy": 5.5306977272033695,
      "epoch": 3.2869194484760524,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003940250513644236,
      "loss": 4.851,
      "mean_token_accuracy": 0.2243834227323532,
      "num_tokens": 75196141.0,
      "step": 36235
    },
    {
      "entropy": 5.492659187316894,
      "epoch": 3.2873730043541363,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003939975917743988,
      "loss": 4.7385,
      "mean_token_accuracy": 0.23586864322423934,
      "num_tokens": 75205414.0,
      "step": 36240
    },
    {
      "entropy": 5.611056852340698,
      "epoch": 3.2878265602322205,
      "grad_norm": 1.25,
      "learning_rate": 0.0003939701297234735,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21699569076299668,
      "num_tokens": 75215860.0,
      "step": 36245
    },
    {
      "entropy": 5.58050799369812,
      "epoch": 3.288280116110305,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039394266521221565,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2166503205895424,
      "num_tokens": 75226358.0,
      "step": 36250
    },
    {
      "entropy": 5.482949304580688,
      "epoch": 3.288733671988389,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039391519824119315,
      "loss": 4.748,
      "mean_token_accuracy": 0.23686134517192842,
      "num_tokens": 75235779.0,
      "step": 36255
    },
    {
      "entropy": 5.596523284912109,
      "epoch": 3.2891872278664733,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039388772881097396,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2118769958615303,
      "num_tokens": 75246417.0,
      "step": 36260
    },
    {
      "entropy": 5.567182731628418,
      "epoch": 3.289640783744557,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039386025692212637,
      "loss": 4.843,
      "mean_token_accuracy": 0.22138625234365464,
      "num_tokens": 75257094.0,
      "step": 36265
    },
    {
      "entropy": 5.5510725498199465,
      "epoch": 3.2900943396226414,
      "grad_norm": 1.125,
      "learning_rate": 0.0003938327825752183,
      "loss": 4.8466,
      "mean_token_accuracy": 0.22223186641931533,
      "num_tokens": 75267374.0,
      "step": 36270
    },
    {
      "entropy": 5.526826810836792,
      "epoch": 3.2905478955007257,
      "grad_norm": 1.328125,
      "learning_rate": 0.000393805305770818,
      "loss": 4.875,
      "mean_token_accuracy": 0.2252560630440712,
      "num_tokens": 75276936.0,
      "step": 36275
    },
    {
      "entropy": 5.513179969787598,
      "epoch": 3.29100145137881,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003937778265094937,
      "loss": 4.8341,
      "mean_token_accuracy": 0.22409375607967377,
      "num_tokens": 75286403.0,
      "step": 36280
    },
    {
      "entropy": 5.567946243286133,
      "epoch": 3.291455007256894,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003937503447918137,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21889721900224685,
      "num_tokens": 75297421.0,
      "step": 36285
    },
    {
      "entropy": 5.6119222164154055,
      "epoch": 3.291908563134978,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039372286061834633,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2186017408967018,
      "num_tokens": 75307955.0,
      "step": 36290
    },
    {
      "entropy": 5.609750032424927,
      "epoch": 3.2923621190130623,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039369537398966,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2186901390552521,
      "num_tokens": 75318739.0,
      "step": 36295
    },
    {
      "entropy": 5.505655431747437,
      "epoch": 3.2928156748911466,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039366788490632286,
      "loss": 4.7663,
      "mean_token_accuracy": 0.2291013315320015,
      "num_tokens": 75328629.0,
      "step": 36300
    },
    {
      "entropy": 5.512588357925415,
      "epoch": 3.293269230769231,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039364039336890386,
      "loss": 4.8365,
      "mean_token_accuracy": 0.2279593378305435,
      "num_tokens": 75337986.0,
      "step": 36305
    },
    {
      "entropy": 5.513869857788086,
      "epoch": 3.293722786647315,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003936128993779712,
      "loss": 4.8078,
      "mean_token_accuracy": 0.22542369812726976,
      "num_tokens": 75346169.0,
      "step": 36310
    },
    {
      "entropy": 5.599983024597168,
      "epoch": 3.2941763425253994,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003935854029340936,
      "loss": 4.8721,
      "mean_token_accuracy": 0.2276986613869667,
      "num_tokens": 75356111.0,
      "step": 36315
    },
    {
      "entropy": 5.5595966339111325,
      "epoch": 3.294629898403483,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039355790403783954,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22865329384803773,
      "num_tokens": 75366679.0,
      "step": 36320
    },
    {
      "entropy": 5.476230144500732,
      "epoch": 3.2950834542815675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039353040268977785,
      "loss": 4.7493,
      "mean_token_accuracy": 0.2314259961247444,
      "num_tokens": 75377500.0,
      "step": 36325
    },
    {
      "entropy": 5.534752273559571,
      "epoch": 3.2955370101596517,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003935028988904773,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22389221787452698,
      "num_tokens": 75387115.0,
      "step": 36330
    },
    {
      "entropy": 5.563125705718994,
      "epoch": 3.295990566037736,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039347539264050654,
      "loss": 4.9465,
      "mean_token_accuracy": 0.216111621260643,
      "num_tokens": 75396330.0,
      "step": 36335
    },
    {
      "entropy": 5.541569328308105,
      "epoch": 3.29644412191582,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003934478839404345,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22661789655685424,
      "num_tokens": 75407168.0,
      "step": 36340
    },
    {
      "entropy": 5.620918655395508,
      "epoch": 3.296897677793904,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039342037279083006,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21865038573741913,
      "num_tokens": 75418117.0,
      "step": 36345
    },
    {
      "entropy": 5.650157403945923,
      "epoch": 3.2973512336719883,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003933928591922621,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21132252216339112,
      "num_tokens": 75429026.0,
      "step": 36350
    },
    {
      "entropy": 5.584028482437134,
      "epoch": 3.2978047895500726,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003933653431452996,
      "loss": 4.8288,
      "mean_token_accuracy": 0.22415409237146378,
      "num_tokens": 75439710.0,
      "step": 36355
    },
    {
      "entropy": 5.567065095901489,
      "epoch": 3.298258345428157,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039333782465051173,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22537643760442733,
      "num_tokens": 75450086.0,
      "step": 36360
    },
    {
      "entropy": 5.56113109588623,
      "epoch": 3.298711901306241,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003933103037084674,
      "loss": 4.8647,
      "mean_token_accuracy": 0.2129943624138832,
      "num_tokens": 75460988.0,
      "step": 36365
    },
    {
      "entropy": 5.6151158809661865,
      "epoch": 3.299165457184325,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039328278031973586,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21281445771455765,
      "num_tokens": 75471560.0,
      "step": 36370
    },
    {
      "entropy": 5.579126453399658,
      "epoch": 3.2996190130624092,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003932552544848863,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2168205514550209,
      "num_tokens": 75482124.0,
      "step": 36375
    },
    {
      "entropy": 5.561543703079224,
      "epoch": 3.3000725689404935,
      "grad_norm": 1.328125,
      "learning_rate": 0.000393227726204488,
      "loss": 4.8703,
      "mean_token_accuracy": 0.222262841463089,
      "num_tokens": 75492351.0,
      "step": 36380
    },
    {
      "entropy": 5.553583574295044,
      "epoch": 3.3005261248185778,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039320019547911013,
      "loss": 4.7943,
      "mean_token_accuracy": 0.22528759837150575,
      "num_tokens": 75502475.0,
      "step": 36385
    },
    {
      "entropy": 5.509480381011963,
      "epoch": 3.3009796806966616,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003931726623093221,
      "loss": 4.8013,
      "mean_token_accuracy": 0.23014139086008073,
      "num_tokens": 75512345.0,
      "step": 36390
    },
    {
      "entropy": 5.509067153930664,
      "epoch": 3.301433236574746,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039314512669569323,
      "loss": 4.7646,
      "mean_token_accuracy": 0.23132060170173646,
      "num_tokens": 75523011.0,
      "step": 36395
    },
    {
      "entropy": 5.576844024658203,
      "epoch": 3.30188679245283,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039311758863879316,
      "loss": 4.8676,
      "mean_token_accuracy": 0.2184252232313156,
      "num_tokens": 75533775.0,
      "step": 36400
    },
    {
      "entropy": 5.577095985412598,
      "epoch": 3.3023403483309144,
      "grad_norm": 1.25,
      "learning_rate": 0.00039309004813919117,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21013036817312242,
      "num_tokens": 75544558.0,
      "step": 36405
    },
    {
      "entropy": 5.653550481796264,
      "epoch": 3.3027939042089987,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039306250519745696,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21368105262517928,
      "num_tokens": 75555599.0,
      "step": 36410
    },
    {
      "entropy": 5.562288999557495,
      "epoch": 3.303247460087083,
      "grad_norm": 1.390625,
      "learning_rate": 0.00039303495981416007,
      "loss": 4.8679,
      "mean_token_accuracy": 0.2204214245080948,
      "num_tokens": 75566320.0,
      "step": 36415
    },
    {
      "entropy": 5.582990884780884,
      "epoch": 3.3037010159651667,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003930074119898701,
      "loss": 4.896,
      "mean_token_accuracy": 0.2257215768098831,
      "num_tokens": 75576069.0,
      "step": 36420
    },
    {
      "entropy": 5.572000122070312,
      "epoch": 3.304154571843251,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003929798617251567,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2184012070298195,
      "num_tokens": 75586012.0,
      "step": 36425
    },
    {
      "entropy": 5.585925436019897,
      "epoch": 3.3046081277213353,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003929523090205898,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2137787252664566,
      "num_tokens": 75596499.0,
      "step": 36430
    },
    {
      "entropy": 5.538427686691284,
      "epoch": 3.3050616835994195,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000392924753876739,
      "loss": 4.8078,
      "mean_token_accuracy": 0.22474002987146377,
      "num_tokens": 75606329.0,
      "step": 36435
    },
    {
      "entropy": 5.582177686691284,
      "epoch": 3.305515239477504,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003928971962941743,
      "loss": 4.884,
      "mean_token_accuracy": 0.21986208707094193,
      "num_tokens": 75617371.0,
      "step": 36440
    },
    {
      "entropy": 5.6060436248779295,
      "epoch": 3.3059687953555876,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003928696362734655,
      "loss": 4.8683,
      "mean_token_accuracy": 0.21940613240003587,
      "num_tokens": 75627949.0,
      "step": 36445
    },
    {
      "entropy": 5.578892755508423,
      "epoch": 3.306422351233672,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039284207381518266,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2178596407175064,
      "num_tokens": 75638642.0,
      "step": 36450
    },
    {
      "entropy": 5.501861190795898,
      "epoch": 3.306875907111756,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039281450891989555,
      "loss": 4.872,
      "mean_token_accuracy": 0.22131170481443405,
      "num_tokens": 75649997.0,
      "step": 36455
    },
    {
      "entropy": 5.522936153411865,
      "epoch": 3.3073294629898404,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003927869415881745,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22792258560657502,
      "num_tokens": 75660223.0,
      "step": 36460
    },
    {
      "entropy": 5.566356515884399,
      "epoch": 3.3077830188679247,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003927593718205895,
      "loss": 4.824,
      "mean_token_accuracy": 0.2315408930182457,
      "num_tokens": 75670550.0,
      "step": 36465
    },
    {
      "entropy": 5.618841648101807,
      "epoch": 3.3082365747460085,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039273179961771065,
      "loss": 4.9905,
      "mean_token_accuracy": 0.20904747098684312,
      "num_tokens": 75681631.0,
      "step": 36470
    },
    {
      "entropy": 5.553884983062744,
      "epoch": 3.308690130624093,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003927042249801082,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2253678634762764,
      "num_tokens": 75691561.0,
      "step": 36475
    },
    {
      "entropy": 5.57857346534729,
      "epoch": 3.309143686502177,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039267664790835235,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22244116812944412,
      "num_tokens": 75701639.0,
      "step": 36480
    },
    {
      "entropy": 5.582505178451538,
      "epoch": 3.3095972423802613,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003926490684030134,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22268700152635573,
      "num_tokens": 75712823.0,
      "step": 36485
    },
    {
      "entropy": 5.554870986938477,
      "epoch": 3.3100507982583456,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039262148646466185,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21693439781665802,
      "num_tokens": 75723221.0,
      "step": 36490
    },
    {
      "entropy": 5.559944295883179,
      "epoch": 3.3105043541364294,
      "grad_norm": 1.125,
      "learning_rate": 0.0003925939020938679,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22327907532453536,
      "num_tokens": 75733565.0,
      "step": 36495
    },
    {
      "entropy": 5.62131290435791,
      "epoch": 3.3109579100145137,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039256631529120206,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22839851081371307,
      "num_tokens": 75743511.0,
      "step": 36500
    },
    {
      "entropy": 5.661493110656738,
      "epoch": 3.311411465892598,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003925387260572349,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22204354554414749,
      "num_tokens": 75754593.0,
      "step": 36505
    },
    {
      "entropy": 5.475900268554687,
      "epoch": 3.311865021770682,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039251113439253693,
      "loss": 4.8162,
      "mean_token_accuracy": 0.2310926303267479,
      "num_tokens": 75764534.0,
      "step": 36510
    },
    {
      "entropy": 5.506560754776001,
      "epoch": 3.3123185776487665,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003924835402976788,
      "loss": 4.8197,
      "mean_token_accuracy": 0.22540393173694612,
      "num_tokens": 75774707.0,
      "step": 36515
    },
    {
      "entropy": 5.496322822570801,
      "epoch": 3.3127721335268507,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039245594377323106,
      "loss": 4.7811,
      "mean_token_accuracy": 0.22616538554430007,
      "num_tokens": 75784900.0,
      "step": 36520
    },
    {
      "entropy": 5.635550165176392,
      "epoch": 3.3132256894049346,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003924283448197645,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21307288706302643,
      "num_tokens": 75795711.0,
      "step": 36525
    },
    {
      "entropy": 5.5833546161651615,
      "epoch": 3.313679245283019,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003924007434378498,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22391290217638016,
      "num_tokens": 75805951.0,
      "step": 36530
    },
    {
      "entropy": 5.555324459075928,
      "epoch": 3.314132801161103,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003923731396280579,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22122474610805512,
      "num_tokens": 75816023.0,
      "step": 36535
    },
    {
      "entropy": 5.584045553207398,
      "epoch": 3.3145863570391874,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039234553339095944,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21870590448379518,
      "num_tokens": 75827395.0,
      "step": 36540
    },
    {
      "entropy": 5.626761245727539,
      "epoch": 3.315039912917271,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039231792472712554,
      "loss": 4.8523,
      "mean_token_accuracy": 0.2258223608136177,
      "num_tokens": 75836663.0,
      "step": 36545
    },
    {
      "entropy": 5.5829822540283205,
      "epoch": 3.3154934687953554,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000392290313637127,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22135494500398636,
      "num_tokens": 75846686.0,
      "step": 36550
    },
    {
      "entropy": 5.507472801208496,
      "epoch": 3.3159470246734397,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039226270012153485,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22080935388803483,
      "num_tokens": 75857386.0,
      "step": 36555
    },
    {
      "entropy": 5.544924020767212,
      "epoch": 3.316400580551524,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003922350841809203,
      "loss": 4.8964,
      "mean_token_accuracy": 0.21730443388223647,
      "num_tokens": 75867003.0,
      "step": 36560
    },
    {
      "entropy": 5.565072917938233,
      "epoch": 3.3168541364296082,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003922074658158543,
      "loss": 4.9367,
      "mean_token_accuracy": 0.22167457193136214,
      "num_tokens": 75878155.0,
      "step": 36565
    },
    {
      "entropy": 5.529049491882324,
      "epoch": 3.3173076923076925,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039217984502690797,
      "loss": 4.884,
      "mean_token_accuracy": 0.22216019332408904,
      "num_tokens": 75888478.0,
      "step": 36570
    },
    {
      "entropy": 5.522053623199463,
      "epoch": 3.3177612481857763,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039215222181465263,
      "loss": 4.8204,
      "mean_token_accuracy": 0.22702420949935914,
      "num_tokens": 75899541.0,
      "step": 36575
    },
    {
      "entropy": 5.570736217498779,
      "epoch": 3.3182148040638606,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039212459617965956,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22341932952404023,
      "num_tokens": 75910175.0,
      "step": 36580
    },
    {
      "entropy": 5.52685399055481,
      "epoch": 3.318668359941945,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003920969681224999,
      "loss": 4.8792,
      "mean_token_accuracy": 0.22080364376306533,
      "num_tokens": 75919713.0,
      "step": 36585
    },
    {
      "entropy": 5.561062049865723,
      "epoch": 3.319121915820029,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039206933764374515,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21347711235284805,
      "num_tokens": 75929366.0,
      "step": 36590
    },
    {
      "entropy": 5.670263338088989,
      "epoch": 3.3195754716981134,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039204170474396656,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21077465862035752,
      "num_tokens": 75940284.0,
      "step": 36595
    },
    {
      "entropy": 5.561238574981689,
      "epoch": 3.3200290275761972,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003920140694237358,
      "loss": 4.8437,
      "mean_token_accuracy": 0.22407121658325196,
      "num_tokens": 75950752.0,
      "step": 36600
    },
    {
      "entropy": 5.625689601898193,
      "epoch": 3.3204825834542815,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003919864316836242,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20774998664855956,
      "num_tokens": 75961308.0,
      "step": 36605
    },
    {
      "entropy": 5.598316478729248,
      "epoch": 3.3209361393323658,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003919587915242034,
      "loss": 4.8676,
      "mean_token_accuracy": 0.2274664595723152,
      "num_tokens": 75972317.0,
      "step": 36610
    },
    {
      "entropy": 5.554518938064575,
      "epoch": 3.32138969521045,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039193114894604504,
      "loss": 4.8625,
      "mean_token_accuracy": 0.23019461780786515,
      "num_tokens": 75982769.0,
      "step": 36615
    },
    {
      "entropy": 5.525543165206909,
      "epoch": 3.3218432510885343,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039190350394972074,
      "loss": 4.8459,
      "mean_token_accuracy": 0.22585543543100356,
      "num_tokens": 75992209.0,
      "step": 36620
    },
    {
      "entropy": 5.609144878387451,
      "epoch": 3.322296806966618,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003918758565358021,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2239365667104721,
      "num_tokens": 76001749.0,
      "step": 36625
    },
    {
      "entropy": 5.636905336380005,
      "epoch": 3.3227503628447024,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000391848206704861,
      "loss": 4.8643,
      "mean_token_accuracy": 0.2198459357023239,
      "num_tokens": 76011326.0,
      "step": 36630
    },
    {
      "entropy": 5.518182706832886,
      "epoch": 3.3232039187227866,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039182055445746917,
      "loss": 4.8097,
      "mean_token_accuracy": 0.2273189216852188,
      "num_tokens": 76021527.0,
      "step": 36635
    },
    {
      "entropy": 5.537480068206787,
      "epoch": 3.323657474600871,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039179289979419853,
      "loss": 4.7498,
      "mean_token_accuracy": 0.22769413143396378,
      "num_tokens": 76031217.0,
      "step": 36640
    },
    {
      "entropy": 5.601564073562622,
      "epoch": 3.324111030478955,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039176524271562096,
      "loss": 4.9934,
      "mean_token_accuracy": 0.21185964047908784,
      "num_tokens": 76041936.0,
      "step": 36645
    },
    {
      "entropy": 5.560208988189697,
      "epoch": 3.324564586357039,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039173758322230847,
      "loss": 4.9384,
      "mean_token_accuracy": 0.21843719631433486,
      "num_tokens": 76052431.0,
      "step": 36650
    },
    {
      "entropy": 5.569587516784668,
      "epoch": 3.3250181422351233,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003917099213148329,
      "loss": 4.7963,
      "mean_token_accuracy": 0.23374972343444825,
      "num_tokens": 76061589.0,
      "step": 36655
    },
    {
      "entropy": 5.638384580612183,
      "epoch": 3.3254716981132075,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039168225699376646,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22585769444704057,
      "num_tokens": 76071642.0,
      "step": 36660
    },
    {
      "entropy": 5.544366264343262,
      "epoch": 3.325925253991292,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039165459025968126,
      "loss": 4.8263,
      "mean_token_accuracy": 0.22551517188549042,
      "num_tokens": 76081084.0,
      "step": 36665
    },
    {
      "entropy": 5.567027473449707,
      "epoch": 3.326378809869376,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003916269211131494,
      "loss": 4.8668,
      "mean_token_accuracy": 0.21986715644598007,
      "num_tokens": 76091125.0,
      "step": 36670
    },
    {
      "entropy": 5.6285316944122314,
      "epoch": 3.3268323657474603,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003915992495547431,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2209387257695198,
      "num_tokens": 76101452.0,
      "step": 36675
    },
    {
      "entropy": 5.531765937805176,
      "epoch": 3.327285921625544,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039157157558503456,
      "loss": 4.8898,
      "mean_token_accuracy": 0.21396030187606813,
      "num_tokens": 76112034.0,
      "step": 36680
    },
    {
      "entropy": 5.558709287643433,
      "epoch": 3.3277394775036284,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039154389920459615,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2231830358505249,
      "num_tokens": 76122395.0,
      "step": 36685
    },
    {
      "entropy": 5.626301193237305,
      "epoch": 3.3281930333817127,
      "grad_norm": 1.25,
      "learning_rate": 0.0003915162204140002,
      "loss": 4.8669,
      "mean_token_accuracy": 0.22527687549591063,
      "num_tokens": 76133622.0,
      "step": 36690
    },
    {
      "entropy": 5.6073386669158936,
      "epoch": 3.328646589259797,
      "grad_norm": 1.25,
      "learning_rate": 0.00039148853921381916,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2241377428174019,
      "num_tokens": 76143757.0,
      "step": 36695
    },
    {
      "entropy": 5.725376176834106,
      "epoch": 3.3291001451378808,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003914608556046254,
      "loss": 5.1208,
      "mean_token_accuracy": 0.20432931929826736,
      "num_tokens": 76154943.0,
      "step": 36700
    },
    {
      "entropy": 5.549917221069336,
      "epoch": 3.329553701015965,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003914331695869914,
      "loss": 4.8892,
      "mean_token_accuracy": 0.21629741042852402,
      "num_tokens": 76165227.0,
      "step": 36705
    },
    {
      "entropy": 5.599585247039795,
      "epoch": 3.3300072568940493,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003914054811614899,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2204759508371353,
      "num_tokens": 76174127.0,
      "step": 36710
    },
    {
      "entropy": 5.654601526260376,
      "epoch": 3.3304608127721336,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003913777903286934,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21576366871595382,
      "num_tokens": 76183518.0,
      "step": 36715
    },
    {
      "entropy": 5.593705701828003,
      "epoch": 3.330914368650218,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003913500970891745,
      "loss": 4.918,
      "mean_token_accuracy": 0.21149157285690307,
      "num_tokens": 76193865.0,
      "step": 36720
    },
    {
      "entropy": 5.627123641967773,
      "epoch": 3.331367924528302,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003913224014435058,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21997116953134538,
      "num_tokens": 76204273.0,
      "step": 36725
    },
    {
      "entropy": 5.528572845458984,
      "epoch": 3.331821480406386,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039129470339226044,
      "loss": 4.793,
      "mean_token_accuracy": 0.22748868614435197,
      "num_tokens": 76213706.0,
      "step": 36730
    },
    {
      "entropy": 5.590824222564697,
      "epoch": 3.33227503628447,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039126700293601086,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22295429706573486,
      "num_tokens": 76224484.0,
      "step": 36735
    },
    {
      "entropy": 5.584521627426147,
      "epoch": 3.3327285921625545,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003912393000753299,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2123736023902893,
      "num_tokens": 76236004.0,
      "step": 36740
    },
    {
      "entropy": 5.634949445724487,
      "epoch": 3.3331821480406387,
      "grad_norm": 1.375,
      "learning_rate": 0.00039121159481079077,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2249806135892868,
      "num_tokens": 76245109.0,
      "step": 36745
    },
    {
      "entropy": 5.6107385635375975,
      "epoch": 3.3336357039187225,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039118388714296626,
      "loss": 4.8649,
      "mean_token_accuracy": 0.22179746478796006,
      "num_tokens": 76255154.0,
      "step": 36750
    },
    {
      "entropy": 5.5614673614501955,
      "epoch": 3.334089259796807,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003911561770724292,
      "loss": 4.8241,
      "mean_token_accuracy": 0.22113037109375,
      "num_tokens": 76266014.0,
      "step": 36755
    },
    {
      "entropy": 5.4840186595916744,
      "epoch": 3.334542815674891,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000391128464599753,
      "loss": 4.8195,
      "mean_token_accuracy": 0.22680775672197342,
      "num_tokens": 76277593.0,
      "step": 36760
    },
    {
      "entropy": 5.496492004394531,
      "epoch": 3.3349963715529753,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003911007497255104,
      "loss": 4.8004,
      "mean_token_accuracy": 0.23136518001556397,
      "num_tokens": 76287816.0,
      "step": 36765
    },
    {
      "entropy": 5.591173219680786,
      "epoch": 3.3354499274310596,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003910730324502748,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21267748028039932,
      "num_tokens": 76298299.0,
      "step": 36770
    },
    {
      "entropy": 5.588947868347168,
      "epoch": 3.335903483309144,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003910453127746192,
      "loss": 4.923,
      "mean_token_accuracy": 0.21363819241523743,
      "num_tokens": 76308747.0,
      "step": 36775
    },
    {
      "entropy": 5.561166143417358,
      "epoch": 3.3363570391872277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003910175906991171,
      "loss": 4.828,
      "mean_token_accuracy": 0.22078896164894105,
      "num_tokens": 76318923.0,
      "step": 36780
    },
    {
      "entropy": 5.6138872623443605,
      "epoch": 3.336810595065312,
      "grad_norm": 1.25,
      "learning_rate": 0.0003909898662243415,
      "loss": 4.9305,
      "mean_token_accuracy": 0.21744290143251419,
      "num_tokens": 76327985.0,
      "step": 36785
    },
    {
      "entropy": 5.609305334091187,
      "epoch": 3.3372641509433962,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039096213935086604,
      "loss": 4.9899,
      "mean_token_accuracy": 0.211505426466465,
      "num_tokens": 76338933.0,
      "step": 36790
    },
    {
      "entropy": 5.700739622116089,
      "epoch": 3.3377177068214805,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039093441007926396,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2164185866713524,
      "num_tokens": 76349601.0,
      "step": 36795
    },
    {
      "entropy": 5.583732032775879,
      "epoch": 3.3381712626995648,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039090667841010874,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2239080548286438,
      "num_tokens": 76359968.0,
      "step": 36800
    },
    {
      "entropy": 5.485671520233154,
      "epoch": 3.3386248185776486,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003908789443439737,
      "loss": 4.7716,
      "mean_token_accuracy": 0.229628586769104,
      "num_tokens": 76369864.0,
      "step": 36805
    },
    {
      "entropy": 5.517950010299683,
      "epoch": 3.339078374455733,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003908512078814329,
      "loss": 4.7962,
      "mean_token_accuracy": 0.2274636834859848,
      "num_tokens": 76379158.0,
      "step": 36810
    },
    {
      "entropy": 5.566267681121826,
      "epoch": 3.339531930333817,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039082346902305937,
      "loss": 4.8349,
      "mean_token_accuracy": 0.22249929010868072,
      "num_tokens": 76388798.0,
      "step": 36815
    },
    {
      "entropy": 5.519220685958862,
      "epoch": 3.3399854862119014,
      "grad_norm": 1.15625,
      "learning_rate": 0.000390795727769427,
      "loss": 4.8324,
      "mean_token_accuracy": 0.22755888998508453,
      "num_tokens": 76399315.0,
      "step": 36820
    },
    {
      "entropy": 5.56700029373169,
      "epoch": 3.3404390420899857,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003907679841211095,
      "loss": 4.8919,
      "mean_token_accuracy": 0.219604754447937,
      "num_tokens": 76408841.0,
      "step": 36825
    },
    {
      "entropy": 5.614740324020386,
      "epoch": 3.3408925979680695,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003907402380786806,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2144710823893547,
      "num_tokens": 76420055.0,
      "step": 36830
    },
    {
      "entropy": 5.553021955490112,
      "epoch": 3.3413461538461537,
      "grad_norm": 1.125,
      "learning_rate": 0.0003907124896427141,
      "loss": 4.825,
      "mean_token_accuracy": 0.22464940696954727,
      "num_tokens": 76430921.0,
      "step": 36835
    },
    {
      "entropy": 5.572552585601807,
      "epoch": 3.341799709724238,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003906847388137837,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22144886255264282,
      "num_tokens": 76441216.0,
      "step": 36840
    },
    {
      "entropy": 5.55699462890625,
      "epoch": 3.3422532656023223,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003906569855924636,
      "loss": 4.9042,
      "mean_token_accuracy": 0.2220016449689865,
      "num_tokens": 76451940.0,
      "step": 36845
    },
    {
      "entropy": 5.588142538070679,
      "epoch": 3.3427068214804065,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003906292299793275,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22202715575695037,
      "num_tokens": 76461701.0,
      "step": 36850
    },
    {
      "entropy": 5.617266893386841,
      "epoch": 3.3431603773584904,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039060147197494943,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21108526289463042,
      "num_tokens": 76473658.0,
      "step": 36855
    },
    {
      "entropy": 5.567918682098389,
      "epoch": 3.3436139332365746,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039057371157990347,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22024325728416444,
      "num_tokens": 76483821.0,
      "step": 36860
    },
    {
      "entropy": 5.5942771434783936,
      "epoch": 3.344067489114659,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039054594879476374,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21405478417873383,
      "num_tokens": 76494699.0,
      "step": 36865
    },
    {
      "entropy": 5.587513637542725,
      "epoch": 3.344521044992743,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039051818362010433,
      "loss": 4.8677,
      "mean_token_accuracy": 0.22719480991363525,
      "num_tokens": 76504299.0,
      "step": 36870
    },
    {
      "entropy": 5.610410261154175,
      "epoch": 3.3449746008708274,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003904904160564994,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2180797502398491,
      "num_tokens": 76514469.0,
      "step": 36875
    },
    {
      "entropy": 5.584064865112305,
      "epoch": 3.3454281567489117,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003904626461045232,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21947478801012038,
      "num_tokens": 76524666.0,
      "step": 36880
    },
    {
      "entropy": 5.519516897201538,
      "epoch": 3.3458817126269955,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039043487376475017,
      "loss": 4.8577,
      "mean_token_accuracy": 0.21699734032154083,
      "num_tokens": 76534787.0,
      "step": 36885
    },
    {
      "entropy": 5.585465717315674,
      "epoch": 3.34633526850508,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00039040709903775446,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21751937568187713,
      "num_tokens": 76545055.0,
      "step": 36890
    },
    {
      "entropy": 5.514423894882202,
      "epoch": 3.346788824383164,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003903793219241106,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22558341473340987,
      "num_tokens": 76555458.0,
      "step": 36895
    },
    {
      "entropy": 5.516722965240478,
      "epoch": 3.3472423802612483,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039035154242439294,
      "loss": 4.8329,
      "mean_token_accuracy": 0.22858362048864364,
      "num_tokens": 76566272.0,
      "step": 36900
    },
    {
      "entropy": 5.630861759185791,
      "epoch": 3.347695936139332,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003903237605391759,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2212006688117981,
      "num_tokens": 76575656.0,
      "step": 36905
    },
    {
      "entropy": 5.677237319946289,
      "epoch": 3.3481494920174164,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003902959762690342,
      "loss": 4.8406,
      "mean_token_accuracy": 0.22116517871618271,
      "num_tokens": 76586481.0,
      "step": 36910
    },
    {
      "entropy": 5.547810173034668,
      "epoch": 3.3486030478955007,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003902681896145423,
      "loss": 4.827,
      "mean_token_accuracy": 0.23021179884672166,
      "num_tokens": 76596023.0,
      "step": 36915
    },
    {
      "entropy": 5.462574100494384,
      "epoch": 3.349056603773585,
      "grad_norm": 1.125,
      "learning_rate": 0.00039024040057627487,
      "loss": 4.783,
      "mean_token_accuracy": 0.2278587743639946,
      "num_tokens": 76605887.0,
      "step": 36920
    },
    {
      "entropy": 5.5292778491973875,
      "epoch": 3.349510159651669,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003902126091548066,
      "loss": 4.7922,
      "mean_token_accuracy": 0.232846137881279,
      "num_tokens": 76616007.0,
      "step": 36925
    },
    {
      "entropy": 5.620429754257202,
      "epoch": 3.3499637155297535,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039018481535071215,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21589034795761108,
      "num_tokens": 76626253.0,
      "step": 36930
    },
    {
      "entropy": 5.577888822555542,
      "epoch": 3.3504172714078373,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003901570191645665,
      "loss": 4.82,
      "mean_token_accuracy": 0.22529965490102768,
      "num_tokens": 76636341.0,
      "step": 36935
    },
    {
      "entropy": 5.571004104614258,
      "epoch": 3.3508708272859216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039012922059694416,
      "loss": 4.8375,
      "mean_token_accuracy": 0.22479562759399413,
      "num_tokens": 76647616.0,
      "step": 36940
    },
    {
      "entropy": 5.559922218322754,
      "epoch": 3.351324383164006,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003901014196484203,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22256214320659637,
      "num_tokens": 76658261.0,
      "step": 36945
    },
    {
      "entropy": 5.547712659835815,
      "epoch": 3.35177793904209,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039007361631956976,
      "loss": 4.8784,
      "mean_token_accuracy": 0.216556616127491,
      "num_tokens": 76667985.0,
      "step": 36950
    },
    {
      "entropy": 5.540558481216431,
      "epoch": 3.3522314949201744,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039004581061096745,
      "loss": 4.8361,
      "mean_token_accuracy": 0.22946188896894454,
      "num_tokens": 76678538.0,
      "step": 36955
    },
    {
      "entropy": 5.572626066207886,
      "epoch": 3.352685050798258,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003900180025231886,
      "loss": 4.7921,
      "mean_token_accuracy": 0.2263064280152321,
      "num_tokens": 76688052.0,
      "step": 36960
    },
    {
      "entropy": 5.527972936630249,
      "epoch": 3.3531386066763424,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038999019205680796,
      "loss": 4.8317,
      "mean_token_accuracy": 0.22590949088335038,
      "num_tokens": 76699241.0,
      "step": 36965
    },
    {
      "entropy": 5.5455437183380125,
      "epoch": 3.3535921625544267,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000389962379212401,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22348733097314835,
      "num_tokens": 76708979.0,
      "step": 36970
    },
    {
      "entropy": 5.581981229782104,
      "epoch": 3.354045718432511,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003899345639905426,
      "loss": 4.8388,
      "mean_token_accuracy": 0.22403321266174317,
      "num_tokens": 76719330.0,
      "step": 36975
    },
    {
      "entropy": 5.609098482131958,
      "epoch": 3.3544992743105952,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003899067463918083,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21030414700508118,
      "num_tokens": 76730210.0,
      "step": 36980
    },
    {
      "entropy": 5.671187496185302,
      "epoch": 3.354952830188679,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038987892641677307,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21665871888399124,
      "num_tokens": 76739977.0,
      "step": 36985
    },
    {
      "entropy": 5.564933967590332,
      "epoch": 3.3554063860667633,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003898511040660124,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22152408957481384,
      "num_tokens": 76750200.0,
      "step": 36990
    },
    {
      "entropy": 5.633210897445679,
      "epoch": 3.3558599419448476,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003898232793401017,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21016151756048201,
      "num_tokens": 76761420.0,
      "step": 36995
    },
    {
      "entropy": 5.542958927154541,
      "epoch": 3.356313497822932,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038979545223961627,
      "loss": 4.8384,
      "mean_token_accuracy": 0.2248327448964119,
      "num_tokens": 76771872.0,
      "step": 37000
    },
    {
      "entropy": 5.495049667358399,
      "epoch": 3.356767053701016,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038976762276513163,
      "loss": 4.8188,
      "mean_token_accuracy": 0.22955177426338197,
      "num_tokens": 76782303.0,
      "step": 37005
    },
    {
      "entropy": 5.605542516708374,
      "epoch": 3.3572206095791,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003897397909172232,
      "loss": 4.9562,
      "mean_token_accuracy": 0.2155901610851288,
      "num_tokens": 76792595.0,
      "step": 37010
    },
    {
      "entropy": 5.610218715667725,
      "epoch": 3.3576741654571842,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003897119566964669,
      "loss": 4.9095,
      "mean_token_accuracy": 0.22005278021097183,
      "num_tokens": 76802577.0,
      "step": 37015
    },
    {
      "entropy": 5.588788795471191,
      "epoch": 3.3581277213352685,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003896841201034379,
      "loss": 4.8325,
      "mean_token_accuracy": 0.23271469920873641,
      "num_tokens": 76812686.0,
      "step": 37020
    },
    {
      "entropy": 5.4986647605896,
      "epoch": 3.3585812772133528,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003896562811387121,
      "loss": 4.7644,
      "mean_token_accuracy": 0.23151354640722274,
      "num_tokens": 76823641.0,
      "step": 37025
    },
    {
      "entropy": 5.588286972045898,
      "epoch": 3.359034833091437,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038962843980286523,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2219184711575508,
      "num_tokens": 76834060.0,
      "step": 37030
    },
    {
      "entropy": 5.636357259750366,
      "epoch": 3.3594883889695213,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038960059609647306,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22237439602613449,
      "num_tokens": 76845078.0,
      "step": 37035
    },
    {
      "entropy": 5.627853488922119,
      "epoch": 3.359941944847605,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003895727500201113,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21625707447528839,
      "num_tokens": 76856539.0,
      "step": 37040
    },
    {
      "entropy": 5.647389984130859,
      "epoch": 3.3603955007256894,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038954490157435583,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2166483446955681,
      "num_tokens": 76866941.0,
      "step": 37045
    },
    {
      "entropy": 5.57813606262207,
      "epoch": 3.3608490566037736,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038951705075978264,
      "loss": 4.8758,
      "mean_token_accuracy": 0.21708226799964905,
      "num_tokens": 76878205.0,
      "step": 37050
    },
    {
      "entropy": 5.622052574157715,
      "epoch": 3.361302612481858,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003894891975769677,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2097334548830986,
      "num_tokens": 76888609.0,
      "step": 37055
    },
    {
      "entropy": 5.550989961624145,
      "epoch": 3.3617561683599417,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003894613420264869,
      "loss": 4.8841,
      "mean_token_accuracy": 0.2170402243733406,
      "num_tokens": 76899637.0,
      "step": 37060
    },
    {
      "entropy": 5.566029644012451,
      "epoch": 3.362209724238026,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038943348410891643,
      "loss": 4.903,
      "mean_token_accuracy": 0.21379348188638686,
      "num_tokens": 76911186.0,
      "step": 37065
    },
    {
      "entropy": 5.566127109527588,
      "epoch": 3.3626632801161103,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003894056238248324,
      "loss": 4.8604,
      "mean_token_accuracy": 0.22173467725515367,
      "num_tokens": 76921704.0,
      "step": 37070
    },
    {
      "entropy": 5.513340711593628,
      "epoch": 3.3631168359941945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003893777611748108,
      "loss": 4.8511,
      "mean_token_accuracy": 0.2271147832274437,
      "num_tokens": 76932362.0,
      "step": 37075
    },
    {
      "entropy": 5.588864374160766,
      "epoch": 3.363570391872279,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038934989615942807,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21515003591775894,
      "num_tokens": 76942581.0,
      "step": 37080
    },
    {
      "entropy": 5.573402023315429,
      "epoch": 3.364023947750363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003893220287792603,
      "loss": 4.873,
      "mean_token_accuracy": 0.21924083828926086,
      "num_tokens": 76952912.0,
      "step": 37085
    },
    {
      "entropy": 5.587262439727783,
      "epoch": 3.364477503628447,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038929415903488383,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22035903930664064,
      "num_tokens": 76963320.0,
      "step": 37090
    },
    {
      "entropy": 5.525803327560425,
      "epoch": 3.364931059506531,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000389266286926875,
      "loss": 4.7735,
      "mean_token_accuracy": 0.22972386330366135,
      "num_tokens": 76973448.0,
      "step": 37095
    },
    {
      "entropy": 5.508778810501099,
      "epoch": 3.3653846153846154,
      "grad_norm": 1.25,
      "learning_rate": 0.00038923841245581026,
      "loss": 4.8388,
      "mean_token_accuracy": 0.22881875783205033,
      "num_tokens": 76983671.0,
      "step": 37100
    },
    {
      "entropy": 5.608328771591187,
      "epoch": 3.3658381712626997,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038921053562226607,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2089029684662819,
      "num_tokens": 76994345.0,
      "step": 37105
    },
    {
      "entropy": 5.556074380874634,
      "epoch": 3.3662917271407835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038918265642681884,
      "loss": 4.7902,
      "mean_token_accuracy": 0.2277903288602829,
      "num_tokens": 77004809.0,
      "step": 37110
    },
    {
      "entropy": 5.539338636398315,
      "epoch": 3.3667452830188678,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038915477487004525,
      "loss": 4.8746,
      "mean_token_accuracy": 0.22206738740205764,
      "num_tokens": 77015505.0,
      "step": 37115
    },
    {
      "entropy": 5.460027122497559,
      "epoch": 3.367198838896952,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038912689095252183,
      "loss": 4.7075,
      "mean_token_accuracy": 0.2344900131225586,
      "num_tokens": 77025572.0,
      "step": 37120
    },
    {
      "entropy": 5.62530665397644,
      "epoch": 3.3676523947750363,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003890990046748251,
      "loss": 5.0211,
      "mean_token_accuracy": 0.20564277172088624,
      "num_tokens": 77037273.0,
      "step": 37125
    },
    {
      "entropy": 5.658636903762817,
      "epoch": 3.3681059506531206,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038907111603753196,
      "loss": 4.944,
      "mean_token_accuracy": 0.21606577783823014,
      "num_tokens": 77047724.0,
      "step": 37130
    },
    {
      "entropy": 5.591211891174316,
      "epoch": 3.368559506531205,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038904322504121916,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22384998053312302,
      "num_tokens": 77057839.0,
      "step": 37135
    },
    {
      "entropy": 5.544419145584106,
      "epoch": 3.3690130624092887,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003890153316864632,
      "loss": 4.8383,
      "mean_token_accuracy": 0.215443979203701,
      "num_tokens": 77068505.0,
      "step": 37140
    },
    {
      "entropy": 5.627410793304444,
      "epoch": 3.369466618287373,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038898743597384126,
      "loss": 4.966,
      "mean_token_accuracy": 0.21652162373065947,
      "num_tokens": 77078126.0,
      "step": 37145
    },
    {
      "entropy": 5.56817855834961,
      "epoch": 3.369920174165457,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038895953790393014,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22385476231575013,
      "num_tokens": 77087802.0,
      "step": 37150
    },
    {
      "entropy": 5.575368213653564,
      "epoch": 3.3703737300435415,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038893163747730664,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22743118554353714,
      "num_tokens": 77097901.0,
      "step": 37155
    },
    {
      "entropy": 5.5749739646911625,
      "epoch": 3.3708272859216257,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038890373469454794,
      "loss": 4.8361,
      "mean_token_accuracy": 0.22008615732192993,
      "num_tokens": 77107604.0,
      "step": 37160
    },
    {
      "entropy": 5.567868566513061,
      "epoch": 3.3712808417997095,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038887582955623084,
      "loss": 4.8187,
      "mean_token_accuracy": 0.2270792990922928,
      "num_tokens": 77117410.0,
      "step": 37165
    },
    {
      "entropy": 5.61644606590271,
      "epoch": 3.371734397677794,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003888479220629326,
      "loss": 4.8779,
      "mean_token_accuracy": 0.21819025576114653,
      "num_tokens": 77129546.0,
      "step": 37170
    },
    {
      "entropy": 5.567762422561645,
      "epoch": 3.372187953555878,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003888200122152304,
      "loss": 4.8269,
      "mean_token_accuracy": 0.22551273554563522,
      "num_tokens": 77139217.0,
      "step": 37175
    },
    {
      "entropy": 5.653349304199219,
      "epoch": 3.3726415094339623,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003887921000137013,
      "loss": 4.9838,
      "mean_token_accuracy": 0.21518248915672303,
      "num_tokens": 77149694.0,
      "step": 37180
    },
    {
      "entropy": 5.521298360824585,
      "epoch": 3.3730950653120466,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003887641854589226,
      "loss": 4.8627,
      "mean_token_accuracy": 0.2236031711101532,
      "num_tokens": 77159291.0,
      "step": 37185
    },
    {
      "entropy": 5.524296236038208,
      "epoch": 3.3735486211901304,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003887362685514716,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22009043246507645,
      "num_tokens": 77170507.0,
      "step": 37190
    },
    {
      "entropy": 5.586553239822388,
      "epoch": 3.3740021770682147,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003887083492919254,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22325950413942336,
      "num_tokens": 77179673.0,
      "step": 37195
    },
    {
      "entropy": 5.651239681243896,
      "epoch": 3.374455732946299,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003886804276808617,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22717688083648682,
      "num_tokens": 77190163.0,
      "step": 37200
    },
    {
      "entropy": 5.6128607273101805,
      "epoch": 3.3749092888243832,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003886525037188578,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22132845371961593,
      "num_tokens": 77201498.0,
      "step": 37205
    },
    {
      "entropy": 5.54303674697876,
      "epoch": 3.3753628447024675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003886245774064911,
      "loss": 4.8319,
      "mean_token_accuracy": 0.22729023098945617,
      "num_tokens": 77211475.0,
      "step": 37210
    },
    {
      "entropy": 5.559172677993774,
      "epoch": 3.3758164005805513,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038859664874433923,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22628084868192672,
      "num_tokens": 77221972.0,
      "step": 37215
    },
    {
      "entropy": 5.567298364639282,
      "epoch": 3.3762699564586356,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038856871773297965,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2166341558098793,
      "num_tokens": 77232701.0,
      "step": 37220
    },
    {
      "entropy": 5.588164043426514,
      "epoch": 3.37672351233672,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038854078437299005,
      "loss": 4.8478,
      "mean_token_accuracy": 0.2275938242673874,
      "num_tokens": 77243242.0,
      "step": 37225
    },
    {
      "entropy": 5.547331428527832,
      "epoch": 3.377177068214804,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003885128486649482,
      "loss": 4.8232,
      "mean_token_accuracy": 0.22338929027318954,
      "num_tokens": 77254279.0,
      "step": 37230
    },
    {
      "entropy": 5.550972175598145,
      "epoch": 3.3776306240928884,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038848491060943164,
      "loss": 4.8353,
      "mean_token_accuracy": 0.21936096101999283,
      "num_tokens": 77265483.0,
      "step": 37235
    },
    {
      "entropy": 5.509790325164795,
      "epoch": 3.3780841799709727,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003884569702070182,
      "loss": 4.8214,
      "mean_token_accuracy": 0.21751652359962464,
      "num_tokens": 77275848.0,
      "step": 37240
    },
    {
      "entropy": 5.527715969085693,
      "epoch": 3.3785377358490565,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038842902745828573,
      "loss": 4.8527,
      "mean_token_accuracy": 0.21773028522729873,
      "num_tokens": 77285943.0,
      "step": 37245
    },
    {
      "entropy": 5.537000751495361,
      "epoch": 3.3789912917271407,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003884010823638121,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22167652547359468,
      "num_tokens": 77295276.0,
      "step": 37250
    },
    {
      "entropy": 5.603711748123169,
      "epoch": 3.379444847605225,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003883731349241751,
      "loss": 4.8635,
      "mean_token_accuracy": 0.21951636672019958,
      "num_tokens": 77305713.0,
      "step": 37255
    },
    {
      "entropy": 5.581877946853638,
      "epoch": 3.3798984034833093,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003883451851399528,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22515012621879577,
      "num_tokens": 77316717.0,
      "step": 37260
    },
    {
      "entropy": 5.529958248138428,
      "epoch": 3.380351959361393,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038831723301172325,
      "loss": 4.8095,
      "mean_token_accuracy": 0.22012078911066055,
      "num_tokens": 77327763.0,
      "step": 37265
    },
    {
      "entropy": 5.617612075805664,
      "epoch": 3.3808055152394774,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003882892785400644,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2180537611246109,
      "num_tokens": 77337495.0,
      "step": 37270
    },
    {
      "entropy": 5.636267423629761,
      "epoch": 3.3812590711175616,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038826132172555444,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21607993096113204,
      "num_tokens": 77348849.0,
      "step": 37275
    },
    {
      "entropy": 5.568083429336548,
      "epoch": 3.381712626995646,
      "grad_norm": 1.25,
      "learning_rate": 0.0003882333625687714,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2179555982351303,
      "num_tokens": 77359634.0,
      "step": 37280
    },
    {
      "entropy": 5.5771341800689695,
      "epoch": 3.38216618287373,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038820540107029365,
      "loss": 4.8534,
      "mean_token_accuracy": 0.2242253914475441,
      "num_tokens": 77369342.0,
      "step": 37285
    },
    {
      "entropy": 5.594110345840454,
      "epoch": 3.3826197387518144,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003881774372306993,
      "loss": 4.8826,
      "mean_token_accuracy": 0.22842645794153213,
      "num_tokens": 77379963.0,
      "step": 37290
    },
    {
      "entropy": 5.521760320663452,
      "epoch": 3.3830732946298983,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003881494710505667,
      "loss": 4.8219,
      "mean_token_accuracy": 0.22620476186275482,
      "num_tokens": 77390135.0,
      "step": 37295
    },
    {
      "entropy": 5.61454586982727,
      "epoch": 3.3835268505079825,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038812150253047427,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22026821821928025,
      "num_tokens": 77400151.0,
      "step": 37300
    },
    {
      "entropy": 5.536823892593384,
      "epoch": 3.383980406386067,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003880935316710003,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23001347184181214,
      "num_tokens": 77410723.0,
      "step": 37305
    },
    {
      "entropy": 5.5244384765625,
      "epoch": 3.384433962264151,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003880655584727233,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22677410691976546,
      "num_tokens": 77421213.0,
      "step": 37310
    },
    {
      "entropy": 5.598486852645874,
      "epoch": 3.384887518142235,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003880375829362217,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22285457849502563,
      "num_tokens": 77431705.0,
      "step": 37315
    },
    {
      "entropy": 5.552673292160034,
      "epoch": 3.385341074020319,
      "grad_norm": 1.25,
      "learning_rate": 0.0003880096050620741,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22337566912174225,
      "num_tokens": 77442120.0,
      "step": 37320
    },
    {
      "entropy": 5.578946208953857,
      "epoch": 3.3857946298984034,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003879816248508591,
      "loss": 4.8739,
      "mean_token_accuracy": 0.21787858307361602,
      "num_tokens": 77451992.0,
      "step": 37325
    },
    {
      "entropy": 5.626970291137695,
      "epoch": 3.3862481857764877,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038795364230315523,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20458158999681472,
      "num_tokens": 77462830.0,
      "step": 37330
    },
    {
      "entropy": 5.621602964401245,
      "epoch": 3.386701741654572,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003879256574195414,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2266684129834175,
      "num_tokens": 77472617.0,
      "step": 37335
    },
    {
      "entropy": 5.530554246902466,
      "epoch": 3.387155297532656,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038789767020059614,
      "loss": 4.8031,
      "mean_token_accuracy": 0.2304278790950775,
      "num_tokens": 77483430.0,
      "step": 37340
    },
    {
      "entropy": 5.4940722465515135,
      "epoch": 3.38760885341074,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038786968064689816,
      "loss": 4.8007,
      "mean_token_accuracy": 0.22678707987070085,
      "num_tokens": 77493038.0,
      "step": 37345
    },
    {
      "entropy": 5.574372911453247,
      "epoch": 3.3880624092888243,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003878416887590265,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21327053904533386,
      "num_tokens": 77502816.0,
      "step": 37350
    },
    {
      "entropy": 5.506357955932617,
      "epoch": 3.3885159651669086,
      "grad_norm": 1.25,
      "learning_rate": 0.00038781369453756004,
      "loss": 4.8206,
      "mean_token_accuracy": 0.2308778613805771,
      "num_tokens": 77513026.0,
      "step": 37355
    },
    {
      "entropy": 5.593189525604248,
      "epoch": 3.388969521044993,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003877856979830775,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2154422014951706,
      "num_tokens": 77523217.0,
      "step": 37360
    },
    {
      "entropy": 5.553214979171753,
      "epoch": 3.389423076923077,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038775769909615805,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22305671274662017,
      "num_tokens": 77533339.0,
      "step": 37365
    },
    {
      "entropy": 5.570053577423096,
      "epoch": 3.389876632801161,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038772969787738074,
      "loss": 4.8561,
      "mean_token_accuracy": 0.22445495277643204,
      "num_tokens": 77543438.0,
      "step": 37370
    },
    {
      "entropy": 5.593005704879761,
      "epoch": 3.390330188679245,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003877016943273244,
      "loss": 4.9398,
      "mean_token_accuracy": 0.21611054837703705,
      "num_tokens": 77552825.0,
      "step": 37375
    },
    {
      "entropy": 5.574637222290039,
      "epoch": 3.3907837445573294,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038767368844656836,
      "loss": 4.9159,
      "mean_token_accuracy": 0.21429501920938493,
      "num_tokens": 77563107.0,
      "step": 37380
    },
    {
      "entropy": 5.631849670410157,
      "epoch": 3.3912373004354137,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038764568023569177,
      "loss": 4.9427,
      "mean_token_accuracy": 0.20987144708633423,
      "num_tokens": 77573312.0,
      "step": 37385
    },
    {
      "entropy": 5.559608697891235,
      "epoch": 3.391690856313498,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003876176696952737,
      "loss": 4.8929,
      "mean_token_accuracy": 0.214712230861187,
      "num_tokens": 77583742.0,
      "step": 37390
    },
    {
      "entropy": 5.6415328025817875,
      "epoch": 3.3921444121915822,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038758965682589357,
      "loss": 4.919,
      "mean_token_accuracy": 0.21015006750822068,
      "num_tokens": 77593667.0,
      "step": 37395
    },
    {
      "entropy": 5.563626670837403,
      "epoch": 3.392597968069666,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003875616416281307,
      "loss": 4.8212,
      "mean_token_accuracy": 0.22304535061120986,
      "num_tokens": 77604249.0,
      "step": 37400
    },
    {
      "entropy": 5.562035799026489,
      "epoch": 3.3930515239477503,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038753362410256435,
      "loss": 4.848,
      "mean_token_accuracy": 0.2229387417435646,
      "num_tokens": 77615095.0,
      "step": 37405
    },
    {
      "entropy": 5.54753680229187,
      "epoch": 3.3935050798258346,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003875056042497739,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22631120979785918,
      "num_tokens": 77625182.0,
      "step": 37410
    },
    {
      "entropy": 5.575972557067871,
      "epoch": 3.393958635703919,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038747758207033903,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22275305837392806,
      "num_tokens": 77635423.0,
      "step": 37415
    },
    {
      "entropy": 5.482124757766724,
      "epoch": 3.3944121915820027,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003874495575648391,
      "loss": 4.7737,
      "mean_token_accuracy": 0.22512744069099427,
      "num_tokens": 77645262.0,
      "step": 37420
    },
    {
      "entropy": 5.535907411575318,
      "epoch": 3.394865747460087,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038742153073385356,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21343840211629866,
      "num_tokens": 77657647.0,
      "step": 37425
    },
    {
      "entropy": 5.5897904396057125,
      "epoch": 3.3953193033381712,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038739350157796214,
      "loss": 4.8759,
      "mean_token_accuracy": 0.2206784799695015,
      "num_tokens": 77668074.0,
      "step": 37430
    },
    {
      "entropy": 5.605467653274536,
      "epoch": 3.3957728592162555,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038736547009774445,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22134194672107696,
      "num_tokens": 77678661.0,
      "step": 37435
    },
    {
      "entropy": 5.581776714324951,
      "epoch": 3.3962264150943398,
      "grad_norm": 1.25,
      "learning_rate": 0.0003873374362937803,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2159433051943779,
      "num_tokens": 77689206.0,
      "step": 37440
    },
    {
      "entropy": 5.591617918014526,
      "epoch": 3.396679970972424,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003873094001666493,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2202355980873108,
      "num_tokens": 77699254.0,
      "step": 37445
    },
    {
      "entropy": 5.5721673488616945,
      "epoch": 3.397133526850508,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003872813617169312,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22925124317407608,
      "num_tokens": 77709595.0,
      "step": 37450
    },
    {
      "entropy": 5.51563835144043,
      "epoch": 3.397587082728592,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038725332094520607,
      "loss": 4.8167,
      "mean_token_accuracy": 0.22312337756156922,
      "num_tokens": 77719446.0,
      "step": 37455
    },
    {
      "entropy": 5.5081305503845215,
      "epoch": 3.3980406386066764,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003872252778520536,
      "loss": 4.8306,
      "mean_token_accuracy": 0.232090000808239,
      "num_tokens": 77729723.0,
      "step": 37460
    },
    {
      "entropy": 5.552455759048462,
      "epoch": 3.3984941944847606,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003871972324380537,
      "loss": 4.8464,
      "mean_token_accuracy": 0.21916398853063584,
      "num_tokens": 77739380.0,
      "step": 37465
    },
    {
      "entropy": 5.579446315765381,
      "epoch": 3.3989477503628445,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003871691847037865,
      "loss": 4.8909,
      "mean_token_accuracy": 0.2260361507534981,
      "num_tokens": 77750683.0,
      "step": 37470
    },
    {
      "entropy": 5.56347074508667,
      "epoch": 3.3994013062409287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038714113464983205,
      "loss": 4.9467,
      "mean_token_accuracy": 0.20855045765638353,
      "num_tokens": 77761951.0,
      "step": 37475
    },
    {
      "entropy": 5.617277908325195,
      "epoch": 3.399854862119013,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003871130822767702,
      "loss": 4.9206,
      "mean_token_accuracy": 0.220148266851902,
      "num_tokens": 77772165.0,
      "step": 37480
    },
    {
      "entropy": 5.624896669387818,
      "epoch": 3.4003084179970973,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003870850275851814,
      "loss": 4.9578,
      "mean_token_accuracy": 0.21392548829317093,
      "num_tokens": 77781916.0,
      "step": 37485
    },
    {
      "entropy": 5.565075254440307,
      "epoch": 3.4007619738751815,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038705697057564556,
      "loss": 4.8643,
      "mean_token_accuracy": 0.21626005470752716,
      "num_tokens": 77791397.0,
      "step": 37490
    },
    {
      "entropy": 5.505886840820312,
      "epoch": 3.401215529753266,
      "grad_norm": 1.265625,
      "learning_rate": 0.000387028911248743,
      "loss": 4.8249,
      "mean_token_accuracy": 0.2207539975643158,
      "num_tokens": 77801694.0,
      "step": 37495
    },
    {
      "entropy": 5.598957157135009,
      "epoch": 3.4016690856313496,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038700084960505414,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2225702553987503,
      "num_tokens": 77813199.0,
      "step": 37500
    },
    {
      "entropy": 5.617013311386108,
      "epoch": 3.402122641509434,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003869727856451591,
      "loss": 4.9948,
      "mean_token_accuracy": 0.2159516230225563,
      "num_tokens": 77823689.0,
      "step": 37505
    },
    {
      "entropy": 5.574537229537964,
      "epoch": 3.402576197387518,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038694471936963825,
      "loss": 4.8556,
      "mean_token_accuracy": 0.22915137112140654,
      "num_tokens": 77834331.0,
      "step": 37510
    },
    {
      "entropy": 5.563747930526733,
      "epoch": 3.4030297532656024,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038691665077907213,
      "loss": 4.8125,
      "mean_token_accuracy": 0.2187774121761322,
      "num_tokens": 77843988.0,
      "step": 37515
    },
    {
      "entropy": 5.526642894744873,
      "epoch": 3.4034833091436867,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003868885798740412,
      "loss": 4.8554,
      "mean_token_accuracy": 0.22213855087757112,
      "num_tokens": 77854553.0,
      "step": 37520
    },
    {
      "entropy": 5.5393883228302006,
      "epoch": 3.4039368650217705,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038686050665512583,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22345863580703734,
      "num_tokens": 77864629.0,
      "step": 37525
    },
    {
      "entropy": 5.581824684143067,
      "epoch": 3.4043904208998548,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038683243112290673,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21039967387914657,
      "num_tokens": 77875465.0,
      "step": 37530
    },
    {
      "entropy": 5.6146400451660154,
      "epoch": 3.404843976777939,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003868043532779645,
      "loss": 4.8976,
      "mean_token_accuracy": 0.21607093662023544,
      "num_tokens": 77885337.0,
      "step": 37535
    },
    {
      "entropy": 5.581432151794433,
      "epoch": 3.4052975326560233,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038677627312087967,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22215999215841292,
      "num_tokens": 77896465.0,
      "step": 37540
    },
    {
      "entropy": 5.558106184005737,
      "epoch": 3.4057510885341076,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003867481906522331,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21435817033052446,
      "num_tokens": 77906214.0,
      "step": 37545
    },
    {
      "entropy": 5.594953823089599,
      "epoch": 3.4062046444121914,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003867201058726054,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20812302827835083,
      "num_tokens": 77917466.0,
      "step": 37550
    },
    {
      "entropy": 5.659019613265992,
      "epoch": 3.4066582002902757,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003866920187825775,
      "loss": 4.9348,
      "mean_token_accuracy": 0.22146466225385666,
      "num_tokens": 77927648.0,
      "step": 37555
    },
    {
      "entropy": 5.524556493759155,
      "epoch": 3.40711175616836,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038666392938273015,
      "loss": 4.798,
      "mean_token_accuracy": 0.22683977335691452,
      "num_tokens": 77938028.0,
      "step": 37560
    },
    {
      "entropy": 5.536692428588867,
      "epoch": 3.407565312046444,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003866358376736443,
      "loss": 4.8452,
      "mean_token_accuracy": 0.22351489663124086,
      "num_tokens": 77947276.0,
      "step": 37565
    },
    {
      "entropy": 5.610159635543823,
      "epoch": 3.4080188679245285,
      "grad_norm": 1.25,
      "learning_rate": 0.0003866077436559009,
      "loss": 4.9076,
      "mean_token_accuracy": 0.21851395964622497,
      "num_tokens": 77956791.0,
      "step": 37570
    },
    {
      "entropy": 5.611198234558105,
      "epoch": 3.4084724238026123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003865796473300809,
      "loss": 4.899,
      "mean_token_accuracy": 0.21521719694137573,
      "num_tokens": 77967402.0,
      "step": 37575
    },
    {
      "entropy": 5.666193580627441,
      "epoch": 3.4089259796806965,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003865515486967653,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21833535730838777,
      "num_tokens": 77978000.0,
      "step": 37580
    },
    {
      "entropy": 5.547773170471191,
      "epoch": 3.409379535558781,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038652344775653535,
      "loss": 4.8153,
      "mean_token_accuracy": 0.22897100299596787,
      "num_tokens": 77987959.0,
      "step": 37585
    },
    {
      "entropy": 5.555764675140381,
      "epoch": 3.409833091436865,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003864953445099721,
      "loss": 4.8469,
      "mean_token_accuracy": 0.22847384512424468,
      "num_tokens": 77997990.0,
      "step": 37590
    },
    {
      "entropy": 5.566091299057007,
      "epoch": 3.4102866473149493,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038646723895765654,
      "loss": 4.8923,
      "mean_token_accuracy": 0.21740319281816484,
      "num_tokens": 78008831.0,
      "step": 37595
    },
    {
      "entropy": 5.644970703125,
      "epoch": 3.4107402031930336,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003864391311001702,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21596316248178482,
      "num_tokens": 78019338.0,
      "step": 37600
    },
    {
      "entropy": 5.652904987335205,
      "epoch": 3.4111937590711174,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003864110209380942,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2117679625749588,
      "num_tokens": 78030700.0,
      "step": 37605
    },
    {
      "entropy": 5.594487237930298,
      "epoch": 3.4116473149492017,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038638290847200997,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22827038317918777,
      "num_tokens": 78040190.0,
      "step": 37610
    },
    {
      "entropy": 5.549529981613159,
      "epoch": 3.412100870827286,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003863547937024987,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22790148705244065,
      "num_tokens": 78050847.0,
      "step": 37615
    },
    {
      "entropy": 5.634043979644775,
      "epoch": 3.4125544267053702,
      "grad_norm": 1.390625,
      "learning_rate": 0.000386326676630142,
      "loss": 4.8807,
      "mean_token_accuracy": 0.2265731394290924,
      "num_tokens": 78060584.0,
      "step": 37620
    },
    {
      "entropy": 5.517734384536743,
      "epoch": 3.413007982583454,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003862985572555212,
      "loss": 4.7832,
      "mean_token_accuracy": 0.22339124232530594,
      "num_tokens": 78071189.0,
      "step": 37625
    },
    {
      "entropy": 5.563355159759522,
      "epoch": 3.4134615384615383,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038627043557921785,
      "loss": 4.8695,
      "mean_token_accuracy": 0.22120943665504456,
      "num_tokens": 78082054.0,
      "step": 37630
    },
    {
      "entropy": 5.517155981063842,
      "epoch": 3.4139150943396226,
      "grad_norm": 1.25,
      "learning_rate": 0.0003862423116018136,
      "loss": 4.836,
      "mean_token_accuracy": 0.21968642473220826,
      "num_tokens": 78094032.0,
      "step": 37635
    },
    {
      "entropy": 5.575125360488892,
      "epoch": 3.414368650217707,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038621418532389004,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2173152193427086,
      "num_tokens": 78104724.0,
      "step": 37640
    },
    {
      "entropy": 5.623721122741699,
      "epoch": 3.414822206095791,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038618605674602867,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21335508078336715,
      "num_tokens": 78115927.0,
      "step": 37645
    },
    {
      "entropy": 5.621487474441528,
      "epoch": 3.4152757619738754,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038615792586881145,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22204871028661727,
      "num_tokens": 78126744.0,
      "step": 37650
    },
    {
      "entropy": 5.572717380523682,
      "epoch": 3.415729317851959,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003861297926928199,
      "loss": 4.8883,
      "mean_token_accuracy": 0.21774442195892335,
      "num_tokens": 78137625.0,
      "step": 37655
    },
    {
      "entropy": 5.48847131729126,
      "epoch": 3.4161828737300435,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003861016572186359,
      "loss": 4.8222,
      "mean_token_accuracy": 0.22572675794363023,
      "num_tokens": 78147709.0,
      "step": 37660
    },
    {
      "entropy": 5.632107448577881,
      "epoch": 3.4166364296081277,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038607351944684135,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21770864874124526,
      "num_tokens": 78158634.0,
      "step": 37665
    },
    {
      "entropy": 5.679557657241821,
      "epoch": 3.417089985486212,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003860453793780181,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2137502446770668,
      "num_tokens": 78170279.0,
      "step": 37670
    },
    {
      "entropy": 5.640170001983643,
      "epoch": 3.417543541364296,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00038601723701274813,
      "loss": 4.8931,
      "mean_token_accuracy": 0.22085128873586654,
      "num_tokens": 78180209.0,
      "step": 37675
    },
    {
      "entropy": 5.586086893081665,
      "epoch": 3.41799709724238,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003859890923516134,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21939034909009933,
      "num_tokens": 78189937.0,
      "step": 37680
    },
    {
      "entropy": 5.58044810295105,
      "epoch": 3.4184506531204644,
      "grad_norm": 1.3125,
      "learning_rate": 0.000385960945395196,
      "loss": 4.8735,
      "mean_token_accuracy": 0.221746926009655,
      "num_tokens": 78199224.0,
      "step": 37685
    },
    {
      "entropy": 5.585460042953491,
      "epoch": 3.4189042089985486,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003859327961440779,
      "loss": 4.8215,
      "mean_token_accuracy": 0.22606613785028457,
      "num_tokens": 78209912.0,
      "step": 37690
    },
    {
      "entropy": 5.590497446060181,
      "epoch": 3.419357764876633,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038590464459884133,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2193215548992157,
      "num_tokens": 78220388.0,
      "step": 37695
    },
    {
      "entropy": 5.560060644149781,
      "epoch": 3.419811320754717,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038587649076006844,
      "loss": 5.011,
      "mean_token_accuracy": 0.21287312507629394,
      "num_tokens": 78231094.0,
      "step": 37700
    },
    {
      "entropy": 5.577611637115479,
      "epoch": 3.420264876632801,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003858483346283415,
      "loss": 4.8665,
      "mean_token_accuracy": 0.21670253574848175,
      "num_tokens": 78241111.0,
      "step": 37705
    },
    {
      "entropy": 5.659032154083252,
      "epoch": 3.4207184325108853,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003858201762042428,
      "loss": 4.8699,
      "mean_token_accuracy": 0.22290827333927155,
      "num_tokens": 78251036.0,
      "step": 37710
    },
    {
      "entropy": 5.571799230575562,
      "epoch": 3.4211719883889695,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003857920154883545,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21621798127889633,
      "num_tokens": 78261840.0,
      "step": 37715
    },
    {
      "entropy": 5.578745698928833,
      "epoch": 3.421625544267054,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003857638524812591,
      "loss": 4.8518,
      "mean_token_accuracy": 0.21958670914173126,
      "num_tokens": 78272539.0,
      "step": 37720
    },
    {
      "entropy": 5.627433252334595,
      "epoch": 3.422079100145138,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038573568718353914,
      "loss": 4.8677,
      "mean_token_accuracy": 0.21954717636108398,
      "num_tokens": 78283849.0,
      "step": 37725
    },
    {
      "entropy": 5.612842702865601,
      "epoch": 3.422532656023222,
      "grad_norm": 1.25,
      "learning_rate": 0.0003857075195957768,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22082866877317428,
      "num_tokens": 78293640.0,
      "step": 37730
    },
    {
      "entropy": 5.553917503356933,
      "epoch": 3.422986211901306,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003856793497185548,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22568700313568116,
      "num_tokens": 78303980.0,
      "step": 37735
    },
    {
      "entropy": 5.579537773132325,
      "epoch": 3.4234397677793904,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003856511775524556,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22009177207946778,
      "num_tokens": 78314233.0,
      "step": 37740
    },
    {
      "entropy": 5.616269063949585,
      "epoch": 3.4238933236574747,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038562300309806193,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22699591666460037,
      "num_tokens": 78324449.0,
      "step": 37745
    },
    {
      "entropy": 5.626227521896363,
      "epoch": 3.424346879535559,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003855948263559563,
      "loss": 4.873,
      "mean_token_accuracy": 0.22185973972082138,
      "num_tokens": 78334357.0,
      "step": 37750
    },
    {
      "entropy": 5.58059606552124,
      "epoch": 3.424800435413643,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038556664732672147,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22304437309503555,
      "num_tokens": 78344221.0,
      "step": 37755
    },
    {
      "entropy": 5.553372812271118,
      "epoch": 3.425253991291727,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038553846601094033,
      "loss": 4.8233,
      "mean_token_accuracy": 0.2300659239292145,
      "num_tokens": 78354097.0,
      "step": 37760
    },
    {
      "entropy": 5.706850385665893,
      "epoch": 3.4257075471698113,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003855102824091954,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21903660297393798,
      "num_tokens": 78363797.0,
      "step": 37765
    },
    {
      "entropy": 5.607125377655029,
      "epoch": 3.4261611030478956,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003854820965220697,
      "loss": 4.871,
      "mean_token_accuracy": 0.21720145791769027,
      "num_tokens": 78374084.0,
      "step": 37770
    },
    {
      "entropy": 5.6068727493286135,
      "epoch": 3.42661465892598,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038545390835014614,
      "loss": 4.8663,
      "mean_token_accuracy": 0.21819179654121398,
      "num_tokens": 78385346.0,
      "step": 37775
    },
    {
      "entropy": 5.547117233276367,
      "epoch": 3.4270682148040637,
      "grad_norm": 1.25,
      "learning_rate": 0.00038542571789400754,
      "loss": 4.8541,
      "mean_token_accuracy": 0.22410322427749635,
      "num_tokens": 78395498.0,
      "step": 37780
    },
    {
      "entropy": 5.533360290527344,
      "epoch": 3.427521770682148,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038539752515423704,
      "loss": 4.8444,
      "mean_token_accuracy": 0.2157760202884674,
      "num_tokens": 78405792.0,
      "step": 37785
    },
    {
      "entropy": 5.624130773544311,
      "epoch": 3.427975326560232,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003853693301314175,
      "loss": 4.928,
      "mean_token_accuracy": 0.21824511140584946,
      "num_tokens": 78415672.0,
      "step": 37790
    },
    {
      "entropy": 5.5702355861663815,
      "epoch": 3.4284288824383164,
      "grad_norm": 1.25,
      "learning_rate": 0.0003853411328261321,
      "loss": 4.8087,
      "mean_token_accuracy": 0.22396182864904404,
      "num_tokens": 78426937.0,
      "step": 37795
    },
    {
      "entropy": 5.566282939910889,
      "epoch": 3.4288824383164007,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000385312933238964,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22220931351184844,
      "num_tokens": 78436817.0,
      "step": 37800
    },
    {
      "entropy": 5.6230556011199955,
      "epoch": 3.429335994194485,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003852847313704963,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22057377249002458,
      "num_tokens": 78447348.0,
      "step": 37805
    },
    {
      "entropy": 5.6190252780914305,
      "epoch": 3.429789550072569,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003852565272213123,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22141849100589753,
      "num_tokens": 78457631.0,
      "step": 37810
    },
    {
      "entropy": 5.530671262741089,
      "epoch": 3.430243105950653,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003852283207919951,
      "loss": 4.8168,
      "mean_token_accuracy": 0.23200076669454575,
      "num_tokens": 78468542.0,
      "step": 37815
    },
    {
      "entropy": 5.620871829986572,
      "epoch": 3.4306966618287373,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003852001120831282,
      "loss": 4.918,
      "mean_token_accuracy": 0.22305482029914855,
      "num_tokens": 78478580.0,
      "step": 37820
    },
    {
      "entropy": 5.631638479232788,
      "epoch": 3.4311502177068216,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038517190109529497,
      "loss": 4.8825,
      "mean_token_accuracy": 0.21819714605808258,
      "num_tokens": 78488477.0,
      "step": 37825
    },
    {
      "entropy": 5.624925756454468,
      "epoch": 3.4316037735849054,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038514368782907864,
      "loss": 4.917,
      "mean_token_accuracy": 0.221673646569252,
      "num_tokens": 78498887.0,
      "step": 37830
    },
    {
      "entropy": 5.56370325088501,
      "epoch": 3.4320573294629897,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038511547228506284,
      "loss": 4.8358,
      "mean_token_accuracy": 0.2232913166284561,
      "num_tokens": 78509030.0,
      "step": 37835
    },
    {
      "entropy": 5.461496353149414,
      "epoch": 3.432510885341074,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038508725446383097,
      "loss": 4.775,
      "mean_token_accuracy": 0.23104398101568221,
      "num_tokens": 78518254.0,
      "step": 37840
    },
    {
      "entropy": 5.59017596244812,
      "epoch": 3.4329644412191582,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003850590343659667,
      "loss": 4.8637,
      "mean_token_accuracy": 0.22207069844007493,
      "num_tokens": 78527966.0,
      "step": 37845
    },
    {
      "entropy": 5.630570125579834,
      "epoch": 3.4334179970972425,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038503081199205353,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21948488056659698,
      "num_tokens": 78537932.0,
      "step": 37850
    },
    {
      "entropy": 5.642708730697632,
      "epoch": 3.4338715529753268,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038500258734267505,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21760938465595245,
      "num_tokens": 78549213.0,
      "step": 37855
    },
    {
      "entropy": 5.543390798568725,
      "epoch": 3.4343251088534106,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038497436041841517,
      "loss": 4.8209,
      "mean_token_accuracy": 0.21856172382831573,
      "num_tokens": 78558990.0,
      "step": 37860
    },
    {
      "entropy": 5.5828773975372314,
      "epoch": 3.434778664731495,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038494613121985735,
      "loss": 4.9247,
      "mean_token_accuracy": 0.21493295431137086,
      "num_tokens": 78568496.0,
      "step": 37865
    },
    {
      "entropy": 5.588915109634399,
      "epoch": 3.435232220609579,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003849178997475856,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23063721060752868,
      "num_tokens": 78578386.0,
      "step": 37870
    },
    {
      "entropy": 5.509966039657593,
      "epoch": 3.4356857764876634,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038488966600218365,
      "loss": 4.8125,
      "mean_token_accuracy": 0.22878154516220092,
      "num_tokens": 78588855.0,
      "step": 37875
    },
    {
      "entropy": 5.495119667053222,
      "epoch": 3.4361393323657476,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038486142998423535,
      "loss": 4.7574,
      "mean_token_accuracy": 0.2312979981303215,
      "num_tokens": 78599573.0,
      "step": 37880
    },
    {
      "entropy": 5.617792463302612,
      "epoch": 3.4365928882438315,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003848331916943248,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22138137370347977,
      "num_tokens": 78609793.0,
      "step": 37885
    },
    {
      "entropy": 5.5148101329803465,
      "epoch": 3.4370464441219157,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003848049511330358,
      "loss": 4.8346,
      "mean_token_accuracy": 0.22339835911989211,
      "num_tokens": 78619637.0,
      "step": 37890
    },
    {
      "entropy": 5.638696765899658,
      "epoch": 3.4375,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003847767083009525,
      "loss": 4.907,
      "mean_token_accuracy": 0.229443359375,
      "num_tokens": 78629889.0,
      "step": 37895
    },
    {
      "entropy": 5.546358728408814,
      "epoch": 3.4379535558780843,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038474846319865876,
      "loss": 4.8367,
      "mean_token_accuracy": 0.22491746693849562,
      "num_tokens": 78639853.0,
      "step": 37900
    },
    {
      "entropy": 5.579529285430908,
      "epoch": 3.4384071117561685,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003847202158267389,
      "loss": 4.8172,
      "mean_token_accuracy": 0.22443247586488724,
      "num_tokens": 78649732.0,
      "step": 37905
    },
    {
      "entropy": 5.473654317855835,
      "epoch": 3.4388606676342524,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038469196618577694,
      "loss": 4.7827,
      "mean_token_accuracy": 0.22584399282932283,
      "num_tokens": 78659794.0,
      "step": 37910
    },
    {
      "entropy": 5.644205713272095,
      "epoch": 3.4393142235123366,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003846637142763573,
      "loss": 4.9685,
      "mean_token_accuracy": 0.21565909683704376,
      "num_tokens": 78670526.0,
      "step": 37915
    },
    {
      "entropy": 5.602915811538696,
      "epoch": 3.439767779390421,
      "grad_norm": 1.21875,
      "learning_rate": 0.000384635460099064,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22212835103273393,
      "num_tokens": 78680177.0,
      "step": 37920
    },
    {
      "entropy": 5.634299993515015,
      "epoch": 3.440221335268505,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038460720365448154,
      "loss": 5.029,
      "mean_token_accuracy": 0.20938150733709335,
      "num_tokens": 78690958.0,
      "step": 37925
    },
    {
      "entropy": 5.625418996810913,
      "epoch": 3.4406748911465894,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003845789449431941,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2142394870519638,
      "num_tokens": 78700904.0,
      "step": 37930
    },
    {
      "entropy": 5.6607897758483885,
      "epoch": 3.4411284470246732,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003845506839657862,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21944146156311034,
      "num_tokens": 78712085.0,
      "step": 37935
    },
    {
      "entropy": 5.528128385543823,
      "epoch": 3.4415820029027575,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003845224207228422,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22611829787492752,
      "num_tokens": 78722161.0,
      "step": 37940
    },
    {
      "entropy": 5.650383949279785,
      "epoch": 3.4420355587808418,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038449415521494667,
      "loss": 4.8764,
      "mean_token_accuracy": 0.21643019914627076,
      "num_tokens": 78732485.0,
      "step": 37945
    },
    {
      "entropy": 5.557985591888428,
      "epoch": 3.442489114658926,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003844658874426841,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2184637576341629,
      "num_tokens": 78744564.0,
      "step": 37950
    },
    {
      "entropy": 5.566000080108642,
      "epoch": 3.4429426705370103,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000384437617406639,
      "loss": 4.8755,
      "mean_token_accuracy": 0.2239830657839775,
      "num_tokens": 78755227.0,
      "step": 37955
    },
    {
      "entropy": 5.573590230941773,
      "epoch": 3.4433962264150946,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003844093451073962,
      "loss": 4.8653,
      "mean_token_accuracy": 0.21312370002269745,
      "num_tokens": 78765393.0,
      "step": 37960
    },
    {
      "entropy": 5.600601005554199,
      "epoch": 3.4438497822931784,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038438107054554027,
      "loss": 4.8505,
      "mean_token_accuracy": 0.22546310871839523,
      "num_tokens": 78775587.0,
      "step": 37965
    },
    {
      "entropy": 5.567040491104126,
      "epoch": 3.4443033381712627,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003843527937216558,
      "loss": 4.7975,
      "mean_token_accuracy": 0.23344036787748337,
      "num_tokens": 78785098.0,
      "step": 37970
    },
    {
      "entropy": 5.537916278839111,
      "epoch": 3.444756894049347,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003843245146363278,
      "loss": 4.8274,
      "mean_token_accuracy": 0.22525506168603898,
      "num_tokens": 78795348.0,
      "step": 37975
    },
    {
      "entropy": 5.6354875564575195,
      "epoch": 3.445210449927431,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038429623329014096,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2171437367796898,
      "num_tokens": 78805348.0,
      "step": 37980
    },
    {
      "entropy": 5.49457483291626,
      "epoch": 3.445664005805515,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038426794968368015,
      "loss": 4.8069,
      "mean_token_accuracy": 0.2231268599629402,
      "num_tokens": 78814765.0,
      "step": 37985
    },
    {
      "entropy": 5.604942321777344,
      "epoch": 3.4461175616835993,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003842396638175303,
      "loss": 4.8483,
      "mean_token_accuracy": 0.2237356498837471,
      "num_tokens": 78824745.0,
      "step": 37990
    },
    {
      "entropy": 5.5691405773162845,
      "epoch": 3.4465711175616836,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003842113756922763,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2210797533392906,
      "num_tokens": 78835101.0,
      "step": 37995
    },
    {
      "entropy": 5.642512321472168,
      "epoch": 3.447024673439768,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038418308530850333,
      "loss": 4.963,
      "mean_token_accuracy": 0.2100715458393097,
      "num_tokens": 78845881.0,
      "step": 38000
    },
    {
      "entropy": 5.605302476882935,
      "epoch": 3.447478229317852,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038415479266679624,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21887559890747071,
      "num_tokens": 78856536.0,
      "step": 38005
    },
    {
      "entropy": 5.593984651565552,
      "epoch": 3.4479317851959363,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003841264977677403,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22260956913232804,
      "num_tokens": 78866887.0,
      "step": 38010
    },
    {
      "entropy": 5.566642379760742,
      "epoch": 3.44838534107402,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038409820061192045,
      "loss": 4.894,
      "mean_token_accuracy": 0.22510103285312652,
      "num_tokens": 78877032.0,
      "step": 38015
    },
    {
      "entropy": 5.542991304397583,
      "epoch": 3.4488388969521044,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038406990119992215,
      "loss": 4.8431,
      "mean_token_accuracy": 0.23045724034309387,
      "num_tokens": 78887237.0,
      "step": 38020
    },
    {
      "entropy": 5.474837970733643,
      "epoch": 3.4492924528301887,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038404159953233047,
      "loss": 4.7955,
      "mean_token_accuracy": 0.23510887622833251,
      "num_tokens": 78897436.0,
      "step": 38025
    },
    {
      "entropy": 5.587875747680664,
      "epoch": 3.449746008708273,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003840132956097307,
      "loss": 4.8777,
      "mean_token_accuracy": 0.21943197250366211,
      "num_tokens": 78907691.0,
      "step": 38030
    },
    {
      "entropy": 5.586595821380615,
      "epoch": 3.450199564586357,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038398498943270816,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22054661363363265,
      "num_tokens": 78917517.0,
      "step": 38035
    },
    {
      "entropy": 5.662408018112183,
      "epoch": 3.450653120464441,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003839566810018483,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22155400663614272,
      "num_tokens": 78927835.0,
      "step": 38040
    },
    {
      "entropy": 5.560951900482178,
      "epoch": 3.4511066763425253,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038392837031773656,
      "loss": 4.7987,
      "mean_token_accuracy": 0.225715933740139,
      "num_tokens": 78938853.0,
      "step": 38045
    },
    {
      "entropy": 5.555013132095337,
      "epoch": 3.4515602322206096,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003839000573809583,
      "loss": 4.8486,
      "mean_token_accuracy": 0.233382685482502,
      "num_tokens": 78948508.0,
      "step": 38050
    },
    {
      "entropy": 5.595494604110717,
      "epoch": 3.452013788098694,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038387174219209916,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22643874287605287,
      "num_tokens": 78959002.0,
      "step": 38055
    },
    {
      "entropy": 5.587902450561524,
      "epoch": 3.452467343976778,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003838434247517446,
      "loss": 4.8572,
      "mean_token_accuracy": 0.2246174231171608,
      "num_tokens": 78970358.0,
      "step": 38060
    },
    {
      "entropy": 5.578626871109009,
      "epoch": 3.452920899854862,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003838151050604803,
      "loss": 4.8033,
      "mean_token_accuracy": 0.2319348856806755,
      "num_tokens": 78979408.0,
      "step": 38065
    },
    {
      "entropy": 5.477584505081177,
      "epoch": 3.453374455732946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038378678311889184,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2304510295391083,
      "num_tokens": 78990411.0,
      "step": 38070
    },
    {
      "entropy": 5.5676275253295895,
      "epoch": 3.4538280116110305,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000383758458927565,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22198909223079683,
      "num_tokens": 79001734.0,
      "step": 38075
    },
    {
      "entropy": 5.6270896911621096,
      "epoch": 3.4542815674891147,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003837301324870856,
      "loss": 4.8034,
      "mean_token_accuracy": 0.22631442546844482,
      "num_tokens": 79011381.0,
      "step": 38080
    },
    {
      "entropy": 5.635945129394531,
      "epoch": 3.454735123367199,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038370180379803927,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22044396996498108,
      "num_tokens": 79021124.0,
      "step": 38085
    },
    {
      "entropy": 5.6172760963439945,
      "epoch": 3.455188679245283,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038367347286101197,
      "loss": 4.9936,
      "mean_token_accuracy": 0.20363245159387589,
      "num_tokens": 79033105.0,
      "step": 38090
    },
    {
      "entropy": 5.578976345062256,
      "epoch": 3.455642235123367,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003836451396765896,
      "loss": 4.909,
      "mean_token_accuracy": 0.21669618487358094,
      "num_tokens": 79043458.0,
      "step": 38095
    },
    {
      "entropy": 5.59036922454834,
      "epoch": 3.4560957910014514,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038361680424535795,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22429827153682708,
      "num_tokens": 79055186.0,
      "step": 38100
    },
    {
      "entropy": 5.606384181976319,
      "epoch": 3.4565493468795356,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003835884665679032,
      "loss": 4.9104,
      "mean_token_accuracy": 0.2200383499264717,
      "num_tokens": 79065743.0,
      "step": 38105
    },
    {
      "entropy": 5.603261709213257,
      "epoch": 3.45700290275762,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003835601266448112,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2131814643740654,
      "num_tokens": 79077285.0,
      "step": 38110
    },
    {
      "entropy": 5.606313705444336,
      "epoch": 3.457456458635704,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003835317844766682,
      "loss": 4.8867,
      "mean_token_accuracy": 0.22218118011951446,
      "num_tokens": 79087885.0,
      "step": 38115
    },
    {
      "entropy": 5.627710962295533,
      "epoch": 3.457910014513788,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038350344006406016,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22139535546302797,
      "num_tokens": 79097907.0,
      "step": 38120
    },
    {
      "entropy": 5.654132223129272,
      "epoch": 3.4583635703918723,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038347509340757343,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2136502295732498,
      "num_tokens": 79108478.0,
      "step": 38125
    },
    {
      "entropy": 5.628736686706543,
      "epoch": 3.4588171262699565,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003834467445077941,
      "loss": 4.9057,
      "mean_token_accuracy": 0.21378462463617326,
      "num_tokens": 79117988.0,
      "step": 38130
    },
    {
      "entropy": 5.669226789474488,
      "epoch": 3.459270682148041,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003834183933653084,
      "loss": 4.916,
      "mean_token_accuracy": 0.22349207997322082,
      "num_tokens": 79127786.0,
      "step": 38135
    },
    {
      "entropy": 5.593431615829468,
      "epoch": 3.4597242380261246,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003833900399807027,
      "loss": 4.842,
      "mean_token_accuracy": 0.22651919573545456,
      "num_tokens": 79137775.0,
      "step": 38140
    },
    {
      "entropy": 5.505549335479737,
      "epoch": 3.460177793904209,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00038336168435456345,
      "loss": 4.8109,
      "mean_token_accuracy": 0.22458068430423736,
      "num_tokens": 79147803.0,
      "step": 38145
    },
    {
      "entropy": 5.47860369682312,
      "epoch": 3.460631349782293,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038333332648747684,
      "loss": 4.8012,
      "mean_token_accuracy": 0.22418748885393142,
      "num_tokens": 79158377.0,
      "step": 38150
    },
    {
      "entropy": 5.548859930038452,
      "epoch": 3.4610849056603774,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003833049663800295,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22145052552223204,
      "num_tokens": 79168514.0,
      "step": 38155
    },
    {
      "entropy": 5.617284965515137,
      "epoch": 3.4615384615384617,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003832766040328078,
      "loss": 4.985,
      "mean_token_accuracy": 0.21048517376184464,
      "num_tokens": 79177588.0,
      "step": 38160
    },
    {
      "entropy": 5.56250376701355,
      "epoch": 3.461992017416546,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003832482394463984,
      "loss": 4.8482,
      "mean_token_accuracy": 0.22323570251464844,
      "num_tokens": 79188128.0,
      "step": 38165
    },
    {
      "entropy": 5.518023681640625,
      "epoch": 3.4624455732946298,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003832198726213878,
      "loss": 4.8279,
      "mean_token_accuracy": 0.2263498768210411,
      "num_tokens": 79198928.0,
      "step": 38170
    },
    {
      "entropy": 5.503010988235474,
      "epoch": 3.462899129172714,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038319150355836266,
      "loss": 4.8024,
      "mean_token_accuracy": 0.23240560442209243,
      "num_tokens": 79209751.0,
      "step": 38175
    },
    {
      "entropy": 5.583473825454712,
      "epoch": 3.4633526850507983,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038316313225790965,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21401432901620865,
      "num_tokens": 79220210.0,
      "step": 38180
    },
    {
      "entropy": 5.5621545791625975,
      "epoch": 3.4638062409288826,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003831347587206155,
      "loss": 4.909,
      "mean_token_accuracy": 0.22388985455036164,
      "num_tokens": 79231530.0,
      "step": 38185
    },
    {
      "entropy": 5.545995473861694,
      "epoch": 3.4642597968069664,
      "grad_norm": 1.1875,
      "learning_rate": 0.000383106382947067,
      "loss": 4.8364,
      "mean_token_accuracy": 0.22009781002998352,
      "num_tokens": 79242771.0,
      "step": 38190
    },
    {
      "entropy": 5.57250075340271,
      "epoch": 3.4647133526850507,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038307800493785093,
      "loss": 4.8263,
      "mean_token_accuracy": 0.22681186348199844,
      "num_tokens": 79253713.0,
      "step": 38195
    },
    {
      "entropy": 5.59170560836792,
      "epoch": 3.465166908563135,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038304962469355415,
      "loss": 4.9206,
      "mean_token_accuracy": 0.2186089351773262,
      "num_tokens": 79264087.0,
      "step": 38200
    },
    {
      "entropy": 5.5950604438781735,
      "epoch": 3.465620464441219,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003830212422147637,
      "loss": 4.8867,
      "mean_token_accuracy": 0.22162627726793288,
      "num_tokens": 79274859.0,
      "step": 38205
    },
    {
      "entropy": 5.571768474578858,
      "epoch": 3.4660740203193035,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003829928575020663,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22423165142536164,
      "num_tokens": 79284778.0,
      "step": 38210
    },
    {
      "entropy": 5.580922985076905,
      "epoch": 3.4665275761973877,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038296447055604916,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21261829733848572,
      "num_tokens": 79295034.0,
      "step": 38215
    },
    {
      "entropy": 5.540123224258423,
      "epoch": 3.4669811320754715,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003829360813772992,
      "loss": 4.8184,
      "mean_token_accuracy": 0.22912162095308303,
      "num_tokens": 79306013.0,
      "step": 38220
    },
    {
      "entropy": 5.597008752822876,
      "epoch": 3.467434687953556,
      "grad_norm": 1.125,
      "learning_rate": 0.00038290768996640366,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21307953745126723,
      "num_tokens": 79317590.0,
      "step": 38225
    },
    {
      "entropy": 5.631895065307617,
      "epoch": 3.46788824383164,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038287929632394957,
      "loss": 4.9136,
      "mean_token_accuracy": 0.21645509749650954,
      "num_tokens": 79328669.0,
      "step": 38230
    },
    {
      "entropy": 5.48932991027832,
      "epoch": 3.4683417997097243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003828509004505241,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23281917572021485,
      "num_tokens": 79339336.0,
      "step": 38235
    },
    {
      "entropy": 5.617661046981811,
      "epoch": 3.4687953555878086,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003828225023467146,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2166643723845482,
      "num_tokens": 79350199.0,
      "step": 38240
    },
    {
      "entropy": 5.589135026931762,
      "epoch": 3.4692489114658924,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003827941020131082,
      "loss": 4.8541,
      "mean_token_accuracy": 0.2314881831407547,
      "num_tokens": 79359906.0,
      "step": 38245
    },
    {
      "entropy": 5.531719541549682,
      "epoch": 3.4697024673439767,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038276569945029225,
      "loss": 4.8051,
      "mean_token_accuracy": 0.22069131433963776,
      "num_tokens": 79371395.0,
      "step": 38250
    },
    {
      "entropy": 5.5622820377349855,
      "epoch": 3.470156023222061,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003827372946588543,
      "loss": 4.8714,
      "mean_token_accuracy": 0.22540852427482605,
      "num_tokens": 79381808.0,
      "step": 38255
    },
    {
      "entropy": 5.547688817977905,
      "epoch": 3.4706095791001452,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003827088876393816,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22239598482847214,
      "num_tokens": 79391906.0,
      "step": 38260
    },
    {
      "entropy": 5.544122982025146,
      "epoch": 3.4710631349782295,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038268047839246165,
      "loss": 4.8222,
      "mean_token_accuracy": 0.22709712386131287,
      "num_tokens": 79402522.0,
      "step": 38265
    },
    {
      "entropy": 5.492143249511718,
      "epoch": 3.4715166908563133,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038265206691868195,
      "loss": 4.7942,
      "mean_token_accuracy": 0.22038724273443222,
      "num_tokens": 79412751.0,
      "step": 38270
    },
    {
      "entropy": 5.6408251285552975,
      "epoch": 3.4719702467343976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003826236532186301,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2184956356883049,
      "num_tokens": 79424512.0,
      "step": 38275
    },
    {
      "entropy": 5.740165424346924,
      "epoch": 3.472423802612482,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003825952372928936,
      "loss": 5.0491,
      "mean_token_accuracy": 0.2086956486105919,
      "num_tokens": 79434082.0,
      "step": 38280
    },
    {
      "entropy": 5.620924043655395,
      "epoch": 3.472877358490566,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038256681914206026,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22047431915998458,
      "num_tokens": 79444469.0,
      "step": 38285
    },
    {
      "entropy": 5.5147374153137205,
      "epoch": 3.4733309143686504,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038253839876671765,
      "loss": 4.8797,
      "mean_token_accuracy": 0.2194928213953972,
      "num_tokens": 79455036.0,
      "step": 38290
    },
    {
      "entropy": 5.583739233016968,
      "epoch": 3.473784470246734,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003825099761674535,
      "loss": 4.8637,
      "mean_token_accuracy": 0.21832606494426726,
      "num_tokens": 79464600.0,
      "step": 38295
    },
    {
      "entropy": 5.601495456695557,
      "epoch": 3.4742380261248185,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003824815513448558,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22018514424562455,
      "num_tokens": 79474912.0,
      "step": 38300
    },
    {
      "entropy": 5.488470649719238,
      "epoch": 3.4746915820029027,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038245312429951214,
      "loss": 4.7555,
      "mean_token_accuracy": 0.22767754793167114,
      "num_tokens": 79485283.0,
      "step": 38305
    },
    {
      "entropy": 5.462492227554321,
      "epoch": 3.475145137880987,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003824246950320105,
      "loss": 4.7493,
      "mean_token_accuracy": 0.2249643862247467,
      "num_tokens": 79495323.0,
      "step": 38310
    },
    {
      "entropy": 5.563406753540039,
      "epoch": 3.4755986937590713,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003823962635429388,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22347124814987182,
      "num_tokens": 79505144.0,
      "step": 38315
    },
    {
      "entropy": 5.606898546218872,
      "epoch": 3.4760522496371555,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003823678298328849,
      "loss": 4.934,
      "mean_token_accuracy": 0.21643825769424438,
      "num_tokens": 79515402.0,
      "step": 38320
    },
    {
      "entropy": 5.572846460342407,
      "epoch": 3.4765058055152394,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038233939390243697,
      "loss": 4.8099,
      "mean_token_accuracy": 0.23493856489658355,
      "num_tokens": 79525528.0,
      "step": 38325
    },
    {
      "entropy": 5.606199645996094,
      "epoch": 3.4769593613933236,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038231095575218303,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22247189283370972,
      "num_tokens": 79535991.0,
      "step": 38330
    },
    {
      "entropy": 5.621145486831665,
      "epoch": 3.477412917271408,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003822825153827112,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2143198609352112,
      "num_tokens": 79545890.0,
      "step": 38335
    },
    {
      "entropy": 5.595956087112427,
      "epoch": 3.477866473149492,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038225407279460964,
      "loss": 4.9213,
      "mean_token_accuracy": 0.22014188021421432,
      "num_tokens": 79556363.0,
      "step": 38340
    },
    {
      "entropy": 5.583060693740845,
      "epoch": 3.478320029027576,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003822256279884665,
      "loss": 4.888,
      "mean_token_accuracy": 0.2177642896771431,
      "num_tokens": 79566924.0,
      "step": 38345
    },
    {
      "entropy": 5.548189449310303,
      "epoch": 3.4787735849056602,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038219718096487007,
      "loss": 4.8731,
      "mean_token_accuracy": 0.21998017728328706,
      "num_tokens": 79577322.0,
      "step": 38350
    },
    {
      "entropy": 5.674130153656006,
      "epoch": 3.4792271407837445,
      "grad_norm": 1.25,
      "learning_rate": 0.00038216873172440855,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2106102079153061,
      "num_tokens": 79588436.0,
      "step": 38355
    },
    {
      "entropy": 5.647344350814819,
      "epoch": 3.4796806966618288,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003821402802676704,
      "loss": 4.9099,
      "mean_token_accuracy": 0.2200780153274536,
      "num_tokens": 79599015.0,
      "step": 38360
    },
    {
      "entropy": 5.555490970611572,
      "epoch": 3.480134252539913,
      "grad_norm": 1.375,
      "learning_rate": 0.0003821118265952439,
      "loss": 4.8513,
      "mean_token_accuracy": 0.2311486467719078,
      "num_tokens": 79608443.0,
      "step": 38365
    },
    {
      "entropy": 5.504388904571533,
      "epoch": 3.4805878084179973,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003820833707077176,
      "loss": 4.856,
      "mean_token_accuracy": 0.218905708193779,
      "num_tokens": 79619010.0,
      "step": 38370
    },
    {
      "entropy": 5.624173259735107,
      "epoch": 3.481041364296081,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003820549126056799,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21930173933506011,
      "num_tokens": 79629665.0,
      "step": 38375
    },
    {
      "entropy": 5.5009589195251465,
      "epoch": 3.4814949201741654,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003820264522897193,
      "loss": 4.7606,
      "mean_token_accuracy": 0.23094146847724914,
      "num_tokens": 79640578.0,
      "step": 38380
    },
    {
      "entropy": 5.60428524017334,
      "epoch": 3.4819484760522497,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003819979897604243,
      "loss": 4.8633,
      "mean_token_accuracy": 0.21523242592811584,
      "num_tokens": 79649576.0,
      "step": 38385
    },
    {
      "entropy": 5.584999370574951,
      "epoch": 3.482402031930334,
      "grad_norm": 1.25,
      "learning_rate": 0.00038196952501838376,
      "loss": 4.898,
      "mean_token_accuracy": 0.21202096790075303,
      "num_tokens": 79659926.0,
      "step": 38390
    },
    {
      "entropy": 5.650516319274902,
      "epoch": 3.4828555878084178,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00038194105806418617,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21183135658502578,
      "num_tokens": 79669969.0,
      "step": 38395
    },
    {
      "entropy": 5.572910785675049,
      "epoch": 3.483309143686502,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003819125888984202,
      "loss": 4.82,
      "mean_token_accuracy": 0.22524684220552443,
      "num_tokens": 79679706.0,
      "step": 38400
    },
    {
      "entropy": 5.535404014587402,
      "epoch": 3.4837626995645863,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038188411752167466,
      "loss": 4.8578,
      "mean_token_accuracy": 0.22157871276140212,
      "num_tokens": 79691543.0,
      "step": 38405
    },
    {
      "entropy": 5.579240036010742,
      "epoch": 3.4842162554426706,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003818556439345383,
      "loss": 4.8904,
      "mean_token_accuracy": 0.22293858379125595,
      "num_tokens": 79701657.0,
      "step": 38410
    },
    {
      "entropy": 5.557514190673828,
      "epoch": 3.484669811320755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038182716813760004,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22366421520709992,
      "num_tokens": 79711615.0,
      "step": 38415
    },
    {
      "entropy": 5.608257389068603,
      "epoch": 3.485123367198839,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003817986901314487,
      "loss": 4.9095,
      "mean_token_accuracy": 0.22644211798906327,
      "num_tokens": 79721427.0,
      "step": 38420
    },
    {
      "entropy": 5.589441585540771,
      "epoch": 3.485576923076923,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003817702099166732,
      "loss": 4.9085,
      "mean_token_accuracy": 0.22616522908210754,
      "num_tokens": 79732565.0,
      "step": 38425
    },
    {
      "entropy": 5.515077304840088,
      "epoch": 3.486030478955007,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003817417274938626,
      "loss": 4.7445,
      "mean_token_accuracy": 0.23268060982227326,
      "num_tokens": 79744496.0,
      "step": 38430
    },
    {
      "entropy": 5.572480154037476,
      "epoch": 3.4864840348330914,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003817132428636058,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21835279315710068,
      "num_tokens": 79755626.0,
      "step": 38435
    },
    {
      "entropy": 5.575641202926636,
      "epoch": 3.4869375907111757,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038168475602649203,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22059637755155564,
      "num_tokens": 79765447.0,
      "step": 38440
    },
    {
      "entropy": 5.58689250946045,
      "epoch": 3.48739114658926,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003816562669831103,
      "loss": 4.881,
      "mean_token_accuracy": 0.21754232048988342,
      "num_tokens": 79776047.0,
      "step": 38445
    },
    {
      "entropy": 5.567230367660523,
      "epoch": 3.487844702467344,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038162777573404975,
      "loss": 4.849,
      "mean_token_accuracy": 0.22663322687149048,
      "num_tokens": 79786061.0,
      "step": 38450
    },
    {
      "entropy": 5.538536500930786,
      "epoch": 3.488298258345428,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038159928227989964,
      "loss": 4.9018,
      "mean_token_accuracy": 0.21942832171916962,
      "num_tokens": 79796844.0,
      "step": 38455
    },
    {
      "entropy": 5.562500858306885,
      "epoch": 3.4887518142235123,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003815707866212493,
      "loss": 4.854,
      "mean_token_accuracy": 0.2302628129720688,
      "num_tokens": 79808120.0,
      "step": 38460
    },
    {
      "entropy": 5.59022798538208,
      "epoch": 3.4892053701015966,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003815422887586878,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22848547846078873,
      "num_tokens": 79818901.0,
      "step": 38465
    },
    {
      "entropy": 5.639407777786255,
      "epoch": 3.489658925979681,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003815137886928047,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21581187844276428,
      "num_tokens": 79829090.0,
      "step": 38470
    },
    {
      "entropy": 5.669987392425537,
      "epoch": 3.490112481857765,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003814852864241892,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22077712863683702,
      "num_tokens": 79838751.0,
      "step": 38475
    },
    {
      "entropy": 5.528832674026489,
      "epoch": 3.490566037735849,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038145678195343095,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2150067061185837,
      "num_tokens": 79849410.0,
      "step": 38480
    },
    {
      "entropy": 5.652285194396972,
      "epoch": 3.491019593613933,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003814282752811193,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21702227145433425,
      "num_tokens": 79859279.0,
      "step": 38485
    },
    {
      "entropy": 5.651129150390625,
      "epoch": 3.4914731494920175,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003813997664078438,
      "loss": 4.8424,
      "mean_token_accuracy": 0.22598513662815095,
      "num_tokens": 79869535.0,
      "step": 38490
    },
    {
      "entropy": 5.6297143459320065,
      "epoch": 3.4919267053701017,
      "grad_norm": 1.25,
      "learning_rate": 0.00038137125533419403,
      "loss": 5.0174,
      "mean_token_accuracy": 0.20284934639930724,
      "num_tokens": 79881396.0,
      "step": 38495
    },
    {
      "entropy": 5.556525993347168,
      "epoch": 3.4923802612481856,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003813427420607596,
      "loss": 4.836,
      "mean_token_accuracy": 0.21439954936504363,
      "num_tokens": 79891250.0,
      "step": 38500
    },
    {
      "entropy": 5.5284528732299805,
      "epoch": 3.49283381712627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038131422658813004,
      "loss": 4.8212,
      "mean_token_accuracy": 0.22004528045654298,
      "num_tokens": 79901982.0,
      "step": 38505
    },
    {
      "entropy": 5.554597616195679,
      "epoch": 3.493287373004354,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038128570891689527,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21771818548440933,
      "num_tokens": 79912480.0,
      "step": 38510
    },
    {
      "entropy": 5.586081266403198,
      "epoch": 3.4937409288824384,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000381257189047645,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22716558426618577,
      "num_tokens": 79923793.0,
      "step": 38515
    },
    {
      "entropy": 5.605240392684936,
      "epoch": 3.4941944847605226,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038122866698096894,
      "loss": 4.903,
      "mean_token_accuracy": 0.21741262972354888,
      "num_tokens": 79933858.0,
      "step": 38520
    },
    {
      "entropy": 5.649842691421509,
      "epoch": 3.494648040638607,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038120014271745693,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22017600536346435,
      "num_tokens": 79944093.0,
      "step": 38525
    },
    {
      "entropy": 5.610756826400757,
      "epoch": 3.4951015965166907,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000381171616257699,
      "loss": 4.9093,
      "mean_token_accuracy": 0.21904992163181305,
      "num_tokens": 79955072.0,
      "step": 38530
    },
    {
      "entropy": 5.607961654663086,
      "epoch": 3.495555152394775,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038114308760228493,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2190464735031128,
      "num_tokens": 79966949.0,
      "step": 38535
    },
    {
      "entropy": 5.5506988048553465,
      "epoch": 3.4960087082728593,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003811145567518048,
      "loss": 4.8524,
      "mean_token_accuracy": 0.2186318576335907,
      "num_tokens": 79977109.0,
      "step": 38540
    },
    {
      "entropy": 5.638629341125489,
      "epoch": 3.4964622641509435,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038108602370684854,
      "loss": 4.913,
      "mean_token_accuracy": 0.21931980401277543,
      "num_tokens": 79987094.0,
      "step": 38545
    },
    {
      "entropy": 5.606905603408814,
      "epoch": 3.4969158200290273,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003810574884680063,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2213740736246109,
      "num_tokens": 79997503.0,
      "step": 38550
    },
    {
      "entropy": 5.5754317283630375,
      "epoch": 3.4973693759071116,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038102895103586817,
      "loss": 4.8975,
      "mean_token_accuracy": 0.21907483041286469,
      "num_tokens": 80007745.0,
      "step": 38555
    },
    {
      "entropy": 5.6276535987854,
      "epoch": 3.497822931785196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003810004114110243,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21386051028966904,
      "num_tokens": 80019330.0,
      "step": 38560
    },
    {
      "entropy": 5.585355663299561,
      "epoch": 3.49827648766328,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003809718695940649,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2204200565814972,
      "num_tokens": 80029732.0,
      "step": 38565
    },
    {
      "entropy": 5.547657489776611,
      "epoch": 3.4987300435413644,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003809433255855803,
      "loss": 4.79,
      "mean_token_accuracy": 0.22486642450094224,
      "num_tokens": 80041052.0,
      "step": 38570
    },
    {
      "entropy": 5.590043783187866,
      "epoch": 3.4991835994194487,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038091477938616075,
      "loss": 4.9117,
      "mean_token_accuracy": 0.222893887758255,
      "num_tokens": 80051929.0,
      "step": 38575
    },
    {
      "entropy": 5.625316953659057,
      "epoch": 3.4996371552975325,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038088623099639653,
      "loss": 4.96,
      "mean_token_accuracy": 0.21340042352676392,
      "num_tokens": 80061153.0,
      "step": 38580
    },
    {
      "entropy": 5.570801448822022,
      "epoch": 3.5000907111756168,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000380857680416878,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22102087140083312,
      "num_tokens": 80070560.0,
      "step": 38585
    },
    {
      "entropy": 5.713388442993164,
      "epoch": 3.500544267053701,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003808291276481957,
      "loss": 4.9223,
      "mean_token_accuracy": 0.21835293024778366,
      "num_tokens": 80081141.0,
      "step": 38590
    },
    {
      "entropy": 5.5361510753631595,
      "epoch": 3.5009978229317853,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003808005726909401,
      "loss": 4.8051,
      "mean_token_accuracy": 0.2279229462146759,
      "num_tokens": 80092114.0,
      "step": 38595
    },
    {
      "entropy": 5.521596431732178,
      "epoch": 3.501451378809869,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038077201554570175,
      "loss": 4.8306,
      "mean_token_accuracy": 0.23021965771913527,
      "num_tokens": 80101966.0,
      "step": 38600
    },
    {
      "entropy": 5.650166177749634,
      "epoch": 3.5019049346879534,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003807434562130711,
      "loss": 4.984,
      "mean_token_accuracy": 0.2238600254058838,
      "num_tokens": 80113420.0,
      "step": 38605
    },
    {
      "entropy": 5.573134088516236,
      "epoch": 3.5023584905660377,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003807148946936387,
      "loss": 4.911,
      "mean_token_accuracy": 0.21818079203367233,
      "num_tokens": 80123398.0,
      "step": 38610
    },
    {
      "entropy": 5.580202293395996,
      "epoch": 3.502812046444122,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003806863309879956,
      "loss": 4.8495,
      "mean_token_accuracy": 0.22924806773662568,
      "num_tokens": 80133885.0,
      "step": 38615
    },
    {
      "entropy": 5.535071659088135,
      "epoch": 3.503265602322206,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003806577650967321,
      "loss": 4.7868,
      "mean_token_accuracy": 0.24213505983352662,
      "num_tokens": 80144168.0,
      "step": 38620
    },
    {
      "entropy": 5.5498871326446535,
      "epoch": 3.5037191582002905,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003806291970204391,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22394380271434783,
      "num_tokens": 80153956.0,
      "step": 38625
    },
    {
      "entropy": 5.630527353286743,
      "epoch": 3.5041727140783747,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038060062675970743,
      "loss": 4.926,
      "mean_token_accuracy": 0.21100822538137437,
      "num_tokens": 80163617.0,
      "step": 38630
    },
    {
      "entropy": 5.567703485488892,
      "epoch": 3.5046262699564585,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003805720543151278,
      "loss": 4.823,
      "mean_token_accuracy": 0.22542985081672667,
      "num_tokens": 80174049.0,
      "step": 38635
    },
    {
      "entropy": 5.629847574234009,
      "epoch": 3.505079825834543,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003805434796872912,
      "loss": 4.9133,
      "mean_token_accuracy": 0.219144606590271,
      "num_tokens": 80185043.0,
      "step": 38640
    },
    {
      "entropy": 5.529052495956421,
      "epoch": 3.505533381712627,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038051490287678856,
      "loss": 4.8618,
      "mean_token_accuracy": 0.22410847395658492,
      "num_tokens": 80194810.0,
      "step": 38645
    },
    {
      "entropy": 5.533047246932983,
      "epoch": 3.5059869375907113,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003804863238842109,
      "loss": 4.854,
      "mean_token_accuracy": 0.227314855158329,
      "num_tokens": 80204861.0,
      "step": 38650
    },
    {
      "entropy": 5.582651567459107,
      "epoch": 3.506440493468795,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003804577427101491,
      "loss": 4.7835,
      "mean_token_accuracy": 0.22975051999092103,
      "num_tokens": 80214447.0,
      "step": 38655
    },
    {
      "entropy": 5.545746326446533,
      "epoch": 3.5068940493468794,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003804291593551944,
      "loss": 4.8125,
      "mean_token_accuracy": 0.2260776773095131,
      "num_tokens": 80224213.0,
      "step": 38660
    },
    {
      "entropy": 5.594546556472778,
      "epoch": 3.5073476052249637,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003804005738199377,
      "loss": 4.8837,
      "mean_token_accuracy": 0.22270173728466033,
      "num_tokens": 80234161.0,
      "step": 38665
    },
    {
      "entropy": 5.5904776096344,
      "epoch": 3.507801161103048,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00038037198610497037,
      "loss": 4.8638,
      "mean_token_accuracy": 0.21857569068670274,
      "num_tokens": 80243788.0,
      "step": 38670
    },
    {
      "entropy": 5.599396705627441,
      "epoch": 3.5082547169811322,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003803433962108835,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22369408905506133,
      "num_tokens": 80254658.0,
      "step": 38675
    },
    {
      "entropy": 5.618504190444947,
      "epoch": 3.5087082728592165,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003803148041382683,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22115891426801682,
      "num_tokens": 80264820.0,
      "step": 38680
    },
    {
      "entropy": 5.49139552116394,
      "epoch": 3.5091618287373003,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038028620988771617,
      "loss": 4.7995,
      "mean_token_accuracy": 0.22857493460178374,
      "num_tokens": 80274639.0,
      "step": 38685
    },
    {
      "entropy": 5.528353595733643,
      "epoch": 3.5096153846153846,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003802576134598183,
      "loss": 4.843,
      "mean_token_accuracy": 0.22858888506889344,
      "num_tokens": 80284542.0,
      "step": 38690
    },
    {
      "entropy": 5.539374780654907,
      "epoch": 3.510068940493469,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038022901485516625,
      "loss": 4.8213,
      "mean_token_accuracy": 0.22647523880004883,
      "num_tokens": 80294405.0,
      "step": 38695
    },
    {
      "entropy": 5.583394145965576,
      "epoch": 3.510522496371553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038020041407435133,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22299689203500747,
      "num_tokens": 80305326.0,
      "step": 38700
    },
    {
      "entropy": 5.544467449188232,
      "epoch": 3.510976052249637,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038017181111796506,
      "loss": 4.874,
      "mean_token_accuracy": 0.22173336148262024,
      "num_tokens": 80315423.0,
      "step": 38705
    },
    {
      "entropy": 5.5622800350189205,
      "epoch": 3.511429608127721,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003801432059865989,
      "loss": 4.9065,
      "mean_token_accuracy": 0.21991646587848662,
      "num_tokens": 80325372.0,
      "step": 38710
    },
    {
      "entropy": 5.579424285888672,
      "epoch": 3.5118831640058055,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038011459868084446,
      "loss": 4.7864,
      "mean_token_accuracy": 0.22271990329027175,
      "num_tokens": 80335467.0,
      "step": 38715
    },
    {
      "entropy": 5.5865997791290285,
      "epoch": 3.5123367198838897,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003800859892012933,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21876744776964188,
      "num_tokens": 80345420.0,
      "step": 38720
    },
    {
      "entropy": 5.557285737991333,
      "epoch": 3.512790275761974,
      "grad_norm": 1.375,
      "learning_rate": 0.0003800573775485372,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22895897030830384,
      "num_tokens": 80354369.0,
      "step": 38725
    },
    {
      "entropy": 5.59940357208252,
      "epoch": 3.5132438316400583,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00038002876372316774,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21917159706354142,
      "num_tokens": 80365292.0,
      "step": 38730
    },
    {
      "entropy": 5.601279973983765,
      "epoch": 3.513697387518142,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038000014772577657,
      "loss": 4.891,
      "mean_token_accuracy": 0.22118531316518783,
      "num_tokens": 80376340.0,
      "step": 38735
    },
    {
      "entropy": 5.519328165054321,
      "epoch": 3.5141509433962264,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003799715295569557,
      "loss": 4.8404,
      "mean_token_accuracy": 0.22184036076068878,
      "num_tokens": 80386193.0,
      "step": 38740
    },
    {
      "entropy": 5.70469913482666,
      "epoch": 3.5146044992743106,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003799429092172968,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21426361948251724,
      "num_tokens": 80397055.0,
      "step": 38745
    },
    {
      "entropy": 5.626441287994385,
      "epoch": 3.515058055152395,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003799142867073918,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2166975885629654,
      "num_tokens": 80407375.0,
      "step": 38750
    },
    {
      "entropy": 5.624079942703247,
      "epoch": 3.5155116110304787,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037988566202783256,
      "loss": 4.9447,
      "mean_token_accuracy": 0.21558694094419478,
      "num_tokens": 80417942.0,
      "step": 38755
    },
    {
      "entropy": 5.529418563842773,
      "epoch": 3.515965166908563,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003798570351792112,
      "loss": 4.786,
      "mean_token_accuracy": 0.22493006587028502,
      "num_tokens": 80428716.0,
      "step": 38760
    },
    {
      "entropy": 5.569956970214844,
      "epoch": 3.5164187227866472,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037982840616211964,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22145590782165528,
      "num_tokens": 80439688.0,
      "step": 38765
    },
    {
      "entropy": 5.647992324829102,
      "epoch": 3.5168722786647315,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037979977497714984,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21624843180179595,
      "num_tokens": 80450434.0,
      "step": 38770
    },
    {
      "entropy": 5.605155372619629,
      "epoch": 3.5173258345428158,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037977114162489404,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21331779211759566,
      "num_tokens": 80462187.0,
      "step": 38775
    },
    {
      "entropy": 5.56873083114624,
      "epoch": 3.5177793904209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003797425061059443,
      "loss": 4.8403,
      "mean_token_accuracy": 0.22241857796907424,
      "num_tokens": 80473049.0,
      "step": 38780
    },
    {
      "entropy": 5.542993831634521,
      "epoch": 3.5182329462989843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037971386842089287,
      "loss": 4.8375,
      "mean_token_accuracy": 0.2269691303372383,
      "num_tokens": 80484119.0,
      "step": 38785
    },
    {
      "entropy": 5.531101179122925,
      "epoch": 3.518686502177068,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003796852285703319,
      "loss": 4.7749,
      "mean_token_accuracy": 0.2306852385401726,
      "num_tokens": 80494882.0,
      "step": 38790
    },
    {
      "entropy": 5.594499921798706,
      "epoch": 3.5191400580551524,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003796565865548538,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22860032469034194,
      "num_tokens": 80505362.0,
      "step": 38795
    },
    {
      "entropy": 5.595897674560547,
      "epoch": 3.5195936139332367,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003796279423750508,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2137245550751686,
      "num_tokens": 80516027.0,
      "step": 38800
    },
    {
      "entropy": 5.645077514648437,
      "epoch": 3.5200471698113205,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003795992960315153,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21933257281780244,
      "num_tokens": 80527045.0,
      "step": 38805
    },
    {
      "entropy": 5.566857433319091,
      "epoch": 3.5205007256894048,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037957064752483966,
      "loss": 4.7917,
      "mean_token_accuracy": 0.23683036267757415,
      "num_tokens": 80537373.0,
      "step": 38810
    },
    {
      "entropy": 5.517697381973266,
      "epoch": 3.520954281567489,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003795419968556164,
      "loss": 4.8649,
      "mean_token_accuracy": 0.21873768419027328,
      "num_tokens": 80546632.0,
      "step": 38815
    },
    {
      "entropy": 5.542700242996216,
      "epoch": 3.5214078374455733,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003795133440244381,
      "loss": 4.8273,
      "mean_token_accuracy": 0.22546139508485794,
      "num_tokens": 80556694.0,
      "step": 38820
    },
    {
      "entropy": 5.609101724624634,
      "epoch": 3.5218613933236576,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003794846890318971,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21590420454740525,
      "num_tokens": 80567834.0,
      "step": 38825
    },
    {
      "entropy": 5.606833124160767,
      "epoch": 3.522314949201742,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003794560318785861,
      "loss": 4.8702,
      "mean_token_accuracy": 0.22537576854228974,
      "num_tokens": 80578299.0,
      "step": 38830
    },
    {
      "entropy": 5.5998069763183596,
      "epoch": 3.522768505079826,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037942737256509786,
      "loss": 4.9018,
      "mean_token_accuracy": 0.21864100247621537,
      "num_tokens": 80588620.0,
      "step": 38835
    },
    {
      "entropy": 5.705990600585937,
      "epoch": 3.52322206095791,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003793987110920248,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21080234199762343,
      "num_tokens": 80599076.0,
      "step": 38840
    },
    {
      "entropy": 5.649797105789185,
      "epoch": 3.523675616835994,
      "grad_norm": 1.25,
      "learning_rate": 0.00037937004745995987,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21303995996713637,
      "num_tokens": 80609440.0,
      "step": 38845
    },
    {
      "entropy": 5.705781888961792,
      "epoch": 3.5241291727140784,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003793413816694957,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21472978889942168,
      "num_tokens": 80619169.0,
      "step": 38850
    },
    {
      "entropy": 5.56747407913208,
      "epoch": 3.5245827285921627,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003793127137212252,
      "loss": 4.8225,
      "mean_token_accuracy": 0.22456487119197846,
      "num_tokens": 80628719.0,
      "step": 38855
    },
    {
      "entropy": 5.6235559463500975,
      "epoch": 3.5250362844702465,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037928404361574125,
      "loss": 4.862,
      "mean_token_accuracy": 0.2197541505098343,
      "num_tokens": 80638391.0,
      "step": 38860
    },
    {
      "entropy": 5.602475166320801,
      "epoch": 3.525489840348331,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003792553713536367,
      "loss": 4.941,
      "mean_token_accuracy": 0.21766994744539261,
      "num_tokens": 80647720.0,
      "step": 38865
    },
    {
      "entropy": 5.518101644515991,
      "epoch": 3.525943396226415,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003792266969355044,
      "loss": 4.8493,
      "mean_token_accuracy": 0.22654558718204498,
      "num_tokens": 80658841.0,
      "step": 38870
    },
    {
      "entropy": 5.581154870986938,
      "epoch": 3.5263969521044993,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037919802036193753,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22155964225530625,
      "num_tokens": 80668735.0,
      "step": 38875
    },
    {
      "entropy": 5.609885358810425,
      "epoch": 3.5268505079825836,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037916934163352903,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22482971101999283,
      "num_tokens": 80680757.0,
      "step": 38880
    },
    {
      "entropy": 5.559836530685425,
      "epoch": 3.527304063860668,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000379140660750872,
      "loss": 4.8316,
      "mean_token_accuracy": 0.224076047539711,
      "num_tokens": 80689833.0,
      "step": 38885
    },
    {
      "entropy": 5.536722183227539,
      "epoch": 3.5277576197387517,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003791119777145595,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22633997052907945,
      "num_tokens": 80700347.0,
      "step": 38890
    },
    {
      "entropy": 5.498111248016357,
      "epoch": 3.528211175616836,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003790832925251849,
      "loss": 4.7287,
      "mean_token_accuracy": 0.23608938604593277,
      "num_tokens": 80710559.0,
      "step": 38895
    },
    {
      "entropy": 5.591253995895386,
      "epoch": 3.52866473149492,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003790546051833412,
      "loss": 4.9263,
      "mean_token_accuracy": 0.22009024471044542,
      "num_tokens": 80720805.0,
      "step": 38900
    },
    {
      "entropy": 5.608805704116821,
      "epoch": 3.5291182873730045,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003790259156896217,
      "loss": 4.9426,
      "mean_token_accuracy": 0.2055571362376213,
      "num_tokens": 80730937.0,
      "step": 38905
    },
    {
      "entropy": 5.579081678390503,
      "epoch": 3.5295718432510883,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037899722404461983,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22347204983234406,
      "num_tokens": 80741597.0,
      "step": 38910
    },
    {
      "entropy": 5.600455713272095,
      "epoch": 3.5300253991291726,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003789685302489289,
      "loss": 4.894,
      "mean_token_accuracy": 0.21452593952417373,
      "num_tokens": 80752496.0,
      "step": 38915
    },
    {
      "entropy": 5.629484367370606,
      "epoch": 3.530478955007257,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037893983430314226,
      "loss": 4.868,
      "mean_token_accuracy": 0.21937715113162995,
      "num_tokens": 80762329.0,
      "step": 38920
    },
    {
      "entropy": 5.5722349166870115,
      "epoch": 3.530932510885341,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037891113620785327,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21556684374809265,
      "num_tokens": 80773152.0,
      "step": 38925
    },
    {
      "entropy": 5.5604596614837645,
      "epoch": 3.5313860667634254,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037888243596365557,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22231341004371644,
      "num_tokens": 80783502.0,
      "step": 38930
    },
    {
      "entropy": 5.610735130310059,
      "epoch": 3.5318396226415096,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037885373357114265,
      "loss": 4.8062,
      "mean_token_accuracy": 0.2228667914867401,
      "num_tokens": 80792644.0,
      "step": 38935
    },
    {
      "entropy": 5.5927187442779545,
      "epoch": 3.5322931785195935,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000378825029030908,
      "loss": 4.9348,
      "mean_token_accuracy": 0.2176352322101593,
      "num_tokens": 80803184.0,
      "step": 38940
    },
    {
      "entropy": 5.563448762893676,
      "epoch": 3.5327467343976777,
      "grad_norm": 1.25,
      "learning_rate": 0.0003787963223435453,
      "loss": 4.8809,
      "mean_token_accuracy": 0.21825526654720306,
      "num_tokens": 80813353.0,
      "step": 38945
    },
    {
      "entropy": 5.592575979232788,
      "epoch": 3.533200290275762,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003787676135096482,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21718582212924958,
      "num_tokens": 80824663.0,
      "step": 38950
    },
    {
      "entropy": 5.6945765018463135,
      "epoch": 3.5336538461538463,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003787389025298104,
      "loss": 5.0028,
      "mean_token_accuracy": 0.212713360786438,
      "num_tokens": 80835507.0,
      "step": 38955
    },
    {
      "entropy": 5.622861766815186,
      "epoch": 3.53410740203193,
      "grad_norm": 1.25,
      "learning_rate": 0.0003787101894046257,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23465158492326738,
      "num_tokens": 80845035.0,
      "step": 38960
    },
    {
      "entropy": 5.579658031463623,
      "epoch": 3.5345609579100143,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037868147413468785,
      "loss": 4.8532,
      "mean_token_accuracy": 0.21917715966701506,
      "num_tokens": 80854992.0,
      "step": 38965
    },
    {
      "entropy": 5.630315256118775,
      "epoch": 3.5350145137880986,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003786527567205907,
      "loss": 4.9847,
      "mean_token_accuracy": 0.20633552521467208,
      "num_tokens": 80864658.0,
      "step": 38970
    },
    {
      "entropy": 5.642623662948608,
      "epoch": 3.535468069666183,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037862403716292805,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22229874581098558,
      "num_tokens": 80873964.0,
      "step": 38975
    },
    {
      "entropy": 5.61938099861145,
      "epoch": 3.535921625544267,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037859531546229405,
      "loss": 4.9224,
      "mean_token_accuracy": 0.21652499437332154,
      "num_tokens": 80883752.0,
      "step": 38980
    },
    {
      "entropy": 5.620621061325073,
      "epoch": 3.5363751814223514,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003785665916192824,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2172365143895149,
      "num_tokens": 80894265.0,
      "step": 38985
    },
    {
      "entropy": 5.532026481628418,
      "epoch": 3.5368287373004357,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037853786563448724,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22571807652711867,
      "num_tokens": 80904246.0,
      "step": 38990
    },
    {
      "entropy": 5.590527009963989,
      "epoch": 3.5372822931785195,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003785091375085027,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2155577853322029,
      "num_tokens": 80913595.0,
      "step": 38995
    },
    {
      "entropy": 5.562059497833252,
      "epoch": 3.5377358490566038,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003784804072419228,
      "loss": 4.876,
      "mean_token_accuracy": 0.2178777351975441,
      "num_tokens": 80924754.0,
      "step": 39000
    },
    {
      "epoch": 3.5377358490566038,
      "eval_entropy": 5.314401951958152,
      "eval_loss": 4.957967281341553,
      "eval_mean_token_accuracy": 0.22618731720819732,
      "eval_num_tokens": 80924754.0,
      "eval_runtime": 20.8055,
      "eval_samples_per_second": 1699.549,
      "eval_steps_per_second": 212.444,
      "step": 39000
    },
    {
      "entropy": 5.548458003997803,
      "epoch": 3.538189404934688,
      "grad_norm": 1.25,
      "learning_rate": 0.0003784516748353417,
      "loss": 4.8232,
      "mean_token_accuracy": 0.22716688066720964,
      "num_tokens": 80935318.0,
      "step": 39005
    },
    {
      "entropy": 5.582900905609131,
      "epoch": 3.5386429608127723,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003784229402893536,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21798910349607467,
      "num_tokens": 80944886.0,
      "step": 39010
    },
    {
      "entropy": 5.59848313331604,
      "epoch": 3.539096516690856,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003783942036045528,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2172340616583824,
      "num_tokens": 80955566.0,
      "step": 39015
    },
    {
      "entropy": 5.625179147720337,
      "epoch": 3.5395500725689404,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037836546478153344,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21408702731132506,
      "num_tokens": 80965285.0,
      "step": 39020
    },
    {
      "entropy": 5.51876015663147,
      "epoch": 3.5400036284470247,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003783367238208899,
      "loss": 4.8526,
      "mean_token_accuracy": 0.2283048927783966,
      "num_tokens": 80976402.0,
      "step": 39025
    },
    {
      "entropy": 5.615394449234008,
      "epoch": 3.540457184325109,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037830798072321665,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21808362007141113,
      "num_tokens": 80988408.0,
      "step": 39030
    },
    {
      "entropy": 5.641329431533814,
      "epoch": 3.540910740203193,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037827923548910806,
      "loss": 4.8834,
      "mean_token_accuracy": 0.2174405038356781,
      "num_tokens": 80999099.0,
      "step": 39035
    },
    {
      "entropy": 5.579897785186768,
      "epoch": 3.5413642960812775,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037825048811915854,
      "loss": 4.783,
      "mean_token_accuracy": 0.23029969185590743,
      "num_tokens": 81009350.0,
      "step": 39040
    },
    {
      "entropy": 5.514993667602539,
      "epoch": 3.5418178519593613,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037822173861396256,
      "loss": 4.783,
      "mean_token_accuracy": 0.22419554591178895,
      "num_tokens": 81019851.0,
      "step": 39045
    },
    {
      "entropy": 5.558784103393554,
      "epoch": 3.5422714078374455,
      "grad_norm": 1.25,
      "learning_rate": 0.00037819298697411475,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21869858205318451,
      "num_tokens": 81028987.0,
      "step": 39050
    },
    {
      "entropy": 5.5637351989746096,
      "epoch": 3.54272496371553,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037816423320020964,
      "loss": 4.8324,
      "mean_token_accuracy": 0.22129010707139968,
      "num_tokens": 81038081.0,
      "step": 39055
    },
    {
      "entropy": 5.602155065536499,
      "epoch": 3.543178519593614,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037813547729284195,
      "loss": 4.8551,
      "mean_token_accuracy": 0.21958280801773072,
      "num_tokens": 81048066.0,
      "step": 39060
    },
    {
      "entropy": 5.452717924118042,
      "epoch": 3.543632075471698,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037810671925260633,
      "loss": 4.7193,
      "mean_token_accuracy": 0.23591026812791824,
      "num_tokens": 81057421.0,
      "step": 39065
    },
    {
      "entropy": 5.512541723251343,
      "epoch": 3.544085631349782,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037807795908009745,
      "loss": 4.8785,
      "mean_token_accuracy": 0.21922834664583207,
      "num_tokens": 81068435.0,
      "step": 39070
    },
    {
      "entropy": 5.596990013122559,
      "epoch": 3.5445391872278664,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037804919677591006,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22014454305171965,
      "num_tokens": 81079378.0,
      "step": 39075
    },
    {
      "entropy": 5.563273239135742,
      "epoch": 3.5449927431059507,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003780204323406391,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2193951666355133,
      "num_tokens": 81089872.0,
      "step": 39080
    },
    {
      "entropy": 5.548710298538208,
      "epoch": 3.545446298984035,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003779916657748793,
      "loss": 4.8121,
      "mean_token_accuracy": 0.22848846465349198,
      "num_tokens": 81099653.0,
      "step": 39085
    },
    {
      "entropy": 5.5385668754577635,
      "epoch": 3.5458998548621192,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003779628970792256,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22012288123369217,
      "num_tokens": 81109840.0,
      "step": 39090
    },
    {
      "entropy": 5.520224618911743,
      "epoch": 3.546353410740203,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000377934126254273,
      "loss": 4.8102,
      "mean_token_accuracy": 0.22792165726423264,
      "num_tokens": 81120991.0,
      "step": 39095
    },
    {
      "entropy": 5.637347269058227,
      "epoch": 3.5468069666182873,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037790535330061635,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2206135407090187,
      "num_tokens": 81129831.0,
      "step": 39100
    },
    {
      "entropy": 5.6217021465301515,
      "epoch": 3.5472605224963716,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003778765782188508,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22815175354480743,
      "num_tokens": 81140604.0,
      "step": 39105
    },
    {
      "entropy": 5.580640268325806,
      "epoch": 3.547714078374456,
      "grad_norm": 1.375,
      "learning_rate": 0.00037784780100957145,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22785049229860305,
      "num_tokens": 81150591.0,
      "step": 39110
    },
    {
      "entropy": 5.631751489639282,
      "epoch": 3.5481676342525397,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003778190216733733,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21697317510843278,
      "num_tokens": 81161444.0,
      "step": 39115
    },
    {
      "entropy": 5.582563257217407,
      "epoch": 3.548621190130624,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037779024021085167,
      "loss": 4.8206,
      "mean_token_accuracy": 0.23244941532611846,
      "num_tokens": 81171216.0,
      "step": 39120
    },
    {
      "entropy": 5.561651229858398,
      "epoch": 3.549074746008708,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003777614566226016,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22428617626428604,
      "num_tokens": 81181665.0,
      "step": 39125
    },
    {
      "entropy": 5.477673816680908,
      "epoch": 3.5495283018867925,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003777326709092184,
      "loss": 4.8342,
      "mean_token_accuracy": 0.22457398623228073,
      "num_tokens": 81191683.0,
      "step": 39130
    },
    {
      "entropy": 5.566465044021607,
      "epoch": 3.5499818577648767,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037770388307129747,
      "loss": 4.8333,
      "mean_token_accuracy": 0.22929604798555375,
      "num_tokens": 81201211.0,
      "step": 39135
    },
    {
      "entropy": 5.5779046535491945,
      "epoch": 3.550435413642961,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037767509310943407,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22412375509738922,
      "num_tokens": 81210668.0,
      "step": 39140
    },
    {
      "entropy": 5.538126468658447,
      "epoch": 3.5508889695210453,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003776463010242235,
      "loss": 4.7378,
      "mean_token_accuracy": 0.23208339065313338,
      "num_tokens": 81221073.0,
      "step": 39145
    },
    {
      "entropy": 5.56144871711731,
      "epoch": 3.551342525399129,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037761750681626134,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21407207995653152,
      "num_tokens": 81231687.0,
      "step": 39150
    },
    {
      "entropy": 5.628942441940308,
      "epoch": 3.5517960812772134,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037758871048614295,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21539255678653718,
      "num_tokens": 81242323.0,
      "step": 39155
    },
    {
      "entropy": 5.550937652587891,
      "epoch": 3.5522496371552976,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003775599120344638,
      "loss": 4.7329,
      "mean_token_accuracy": 0.23486093282699586,
      "num_tokens": 81252543.0,
      "step": 39160
    },
    {
      "entropy": 5.575171566009521,
      "epoch": 3.5527031930333814,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037753111146181966,
      "loss": 4.8686,
      "mean_token_accuracy": 0.21733078956604004,
      "num_tokens": 81262180.0,
      "step": 39165
    },
    {
      "entropy": 5.691906547546386,
      "epoch": 3.5531567489114657,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000377502308768806,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20363983660936355,
      "num_tokens": 81272088.0,
      "step": 39170
    },
    {
      "entropy": 5.593766403198242,
      "epoch": 3.55361030478955,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003774735039560184,
      "loss": 4.8664,
      "mean_token_accuracy": 0.2227915495634079,
      "num_tokens": 81281723.0,
      "step": 39175
    },
    {
      "entropy": 5.628231525421143,
      "epoch": 3.5540638606676342,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037744469702405273,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2187478795647621,
      "num_tokens": 81292030.0,
      "step": 39180
    },
    {
      "entropy": 5.539047050476074,
      "epoch": 3.5545174165457185,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003774158879735045,
      "loss": 4.84,
      "mean_token_accuracy": 0.22979723960161208,
      "num_tokens": 81302283.0,
      "step": 39185
    },
    {
      "entropy": 5.619292736053467,
      "epoch": 3.5549709724238028,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003773870768049697,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21409715563058854,
      "num_tokens": 81313058.0,
      "step": 39190
    },
    {
      "entropy": 5.59077296257019,
      "epoch": 3.555424528301887,
      "grad_norm": 1.234375,
      "learning_rate": 0.000377358263519044,
      "loss": 4.8665,
      "mean_token_accuracy": 0.22295138984918594,
      "num_tokens": 81323827.0,
      "step": 39195
    },
    {
      "entropy": 5.624802875518799,
      "epoch": 3.555878084179971,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003773294481163234,
      "loss": 4.9146,
      "mean_token_accuracy": 0.21668390035629273,
      "num_tokens": 81333106.0,
      "step": 39200
    },
    {
      "entropy": 5.565656471252441,
      "epoch": 3.556331640058055,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003773006305974037,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22297347337007523,
      "num_tokens": 81343245.0,
      "step": 39205
    },
    {
      "entropy": 5.5489387035369875,
      "epoch": 3.5567851959361394,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037727181096288077,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22889523357152938,
      "num_tokens": 81352917.0,
      "step": 39210
    },
    {
      "entropy": 5.502739810943604,
      "epoch": 3.5572387518142237,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003772429892133509,
      "loss": 4.818,
      "mean_token_accuracy": 0.22434785068035126,
      "num_tokens": 81362838.0,
      "step": 39215
    },
    {
      "entropy": 5.548039102554322,
      "epoch": 3.5576923076923075,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003772141653494099,
      "loss": 4.799,
      "mean_token_accuracy": 0.2289262130856514,
      "num_tokens": 81373395.0,
      "step": 39220
    },
    {
      "entropy": 5.548663806915283,
      "epoch": 3.5581458635703918,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003771853393716538,
      "loss": 4.8526,
      "mean_token_accuracy": 0.22017838358879088,
      "num_tokens": 81383976.0,
      "step": 39225
    },
    {
      "entropy": 5.5360664367675785,
      "epoch": 3.558599419448476,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000377156511280679,
      "loss": 4.8609,
      "mean_token_accuracy": 0.21764481961727142,
      "num_tokens": 81394328.0,
      "step": 39230
    },
    {
      "entropy": 5.569854307174682,
      "epoch": 3.5590529753265603,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003771276810770815,
      "loss": 4.8911,
      "mean_token_accuracy": 0.2207305207848549,
      "num_tokens": 81405451.0,
      "step": 39235
    },
    {
      "entropy": 5.616385126113892,
      "epoch": 3.5595065312046446,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003770988487614575,
      "loss": 4.9704,
      "mean_token_accuracy": 0.21415517926216127,
      "num_tokens": 81416589.0,
      "step": 39240
    },
    {
      "entropy": 5.525298595428467,
      "epoch": 3.559960087082729,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003770700143344033,
      "loss": 4.8298,
      "mean_token_accuracy": 0.22174085527658463,
      "num_tokens": 81428553.0,
      "step": 39245
    },
    {
      "entropy": 5.68926887512207,
      "epoch": 3.5604136429608126,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037704117779651514,
      "loss": 4.9722,
      "mean_token_accuracy": 0.2174268126487732,
      "num_tokens": 81438720.0,
      "step": 39250
    },
    {
      "entropy": 5.543157339096069,
      "epoch": 3.560867198838897,
      "grad_norm": 1.546875,
      "learning_rate": 0.00037701233914838947,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21926826536655425,
      "num_tokens": 81448142.0,
      "step": 39255
    },
    {
      "entropy": 5.590852546691894,
      "epoch": 3.561320754716981,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003769834983906226,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2185932844877243,
      "num_tokens": 81459448.0,
      "step": 39260
    },
    {
      "entropy": 5.584807968139648,
      "epoch": 3.5617743105950654,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037695465552381105,
      "loss": 4.8553,
      "mean_token_accuracy": 0.22540136575698852,
      "num_tokens": 81470356.0,
      "step": 39265
    },
    {
      "entropy": 5.567627000808716,
      "epoch": 3.5622278664731493,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037692581054855125,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2222718060016632,
      "num_tokens": 81480472.0,
      "step": 39270
    },
    {
      "entropy": 5.5348879337310795,
      "epoch": 3.5626814223512335,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037689696346543964,
      "loss": 4.8256,
      "mean_token_accuracy": 0.22938759624958038,
      "num_tokens": 81489949.0,
      "step": 39275
    },
    {
      "entropy": 5.500655794143677,
      "epoch": 3.563134978229318,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037686811427507284,
      "loss": 4.7786,
      "mean_token_accuracy": 0.22332041561603547,
      "num_tokens": 81500276.0,
      "step": 39280
    },
    {
      "entropy": 5.55662841796875,
      "epoch": 3.563588534107402,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037683926297804756,
      "loss": 4.7778,
      "mean_token_accuracy": 0.22417154163122177,
      "num_tokens": 81510431.0,
      "step": 39285
    },
    {
      "entropy": 5.545302104949951,
      "epoch": 3.5640420899854863,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003768104095749603,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2211029350757599,
      "num_tokens": 81520718.0,
      "step": 39290
    },
    {
      "entropy": 5.5862468719482425,
      "epoch": 3.5644956458635706,
      "grad_norm": 1.25,
      "learning_rate": 0.0003767815540664079,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21822448968887329,
      "num_tokens": 81531904.0,
      "step": 39295
    },
    {
      "entropy": 5.572169351577759,
      "epoch": 3.5649492017416544,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037675269645298697,
      "loss": 4.908,
      "mean_token_accuracy": 0.21582047045230865,
      "num_tokens": 81542093.0,
      "step": 39300
    },
    {
      "entropy": 5.575176668167114,
      "epoch": 3.5654027576197387,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003767238367352944,
      "loss": 4.8093,
      "mean_token_accuracy": 0.228555391728878,
      "num_tokens": 81552020.0,
      "step": 39305
    },
    {
      "entropy": 5.54360032081604,
      "epoch": 3.565856313497823,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037669497491392683,
      "loss": 4.773,
      "mean_token_accuracy": 0.231776687502861,
      "num_tokens": 81561644.0,
      "step": 39310
    },
    {
      "entropy": 5.584216547012329,
      "epoch": 3.566309869375907,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037666611098948144,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22431233078241347,
      "num_tokens": 81572158.0,
      "step": 39315
    },
    {
      "entropy": 5.6988341331481935,
      "epoch": 3.566763425253991,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003766372449625549,
      "loss": 4.97,
      "mean_token_accuracy": 0.21029850095510483,
      "num_tokens": 81582335.0,
      "step": 39320
    },
    {
      "entropy": 5.681170415878296,
      "epoch": 3.5672169811320753,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003766083768337442,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21388877034187317,
      "num_tokens": 81593004.0,
      "step": 39325
    },
    {
      "entropy": 5.610628080368042,
      "epoch": 3.5676705370101596,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003765795066036464,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2195033147931099,
      "num_tokens": 81603656.0,
      "step": 39330
    },
    {
      "entropy": 5.582458925247193,
      "epoch": 3.568124092888244,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037655063427285854,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22062468975782396,
      "num_tokens": 81613039.0,
      "step": 39335
    },
    {
      "entropy": 5.538067817687988,
      "epoch": 3.568577648766328,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037652175984197764,
      "loss": 4.8417,
      "mean_token_accuracy": 0.2230832263827324,
      "num_tokens": 81624410.0,
      "step": 39340
    },
    {
      "entropy": 5.597209882736206,
      "epoch": 3.5690312046444124,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037649288331160086,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22319785356521607,
      "num_tokens": 81635160.0,
      "step": 39345
    },
    {
      "entropy": 5.5896398544311525,
      "epoch": 3.5694847605224966,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037646400468232546,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22187502384185792,
      "num_tokens": 81645686.0,
      "step": 39350
    },
    {
      "entropy": 5.574426174163818,
      "epoch": 3.5699383164005805,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037643512395474853,
      "loss": 4.81,
      "mean_token_accuracy": 0.22558621019124986,
      "num_tokens": 81656330.0,
      "step": 39355
    },
    {
      "entropy": 5.534422111511231,
      "epoch": 3.5703918722786647,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003764062411294674,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2156221479177475,
      "num_tokens": 81665895.0,
      "step": 39360
    },
    {
      "entropy": 5.633473873138428,
      "epoch": 3.570845428156749,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037637735620707935,
      "loss": 4.9182,
      "mean_token_accuracy": 0.2181766837835312,
      "num_tokens": 81676616.0,
      "step": 39365
    },
    {
      "entropy": 5.615085554122925,
      "epoch": 3.5712989840348333,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003763484691881818,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2162467807531357,
      "num_tokens": 81687169.0,
      "step": 39370
    },
    {
      "entropy": 5.609271001815796,
      "epoch": 3.571752539912917,
      "grad_norm": 1.25,
      "learning_rate": 0.00037631958007337195,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22232787311077118,
      "num_tokens": 81698291.0,
      "step": 39375
    },
    {
      "entropy": 5.593199348449707,
      "epoch": 3.5722060957910013,
      "grad_norm": 1.25,
      "learning_rate": 0.0003762906888632475,
      "loss": 4.8852,
      "mean_token_accuracy": 0.21934384107589722,
      "num_tokens": 81708750.0,
      "step": 39380
    },
    {
      "entropy": 5.5114984035491945,
      "epoch": 3.5726596516690856,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003762617955584058,
      "loss": 4.8642,
      "mean_token_accuracy": 0.22185833752155304,
      "num_tokens": 81719474.0,
      "step": 39385
    },
    {
      "entropy": 5.670875549316406,
      "epoch": 3.57311320754717,
      "grad_norm": 1.25,
      "learning_rate": 0.0003762329001594443,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21744270771741867,
      "num_tokens": 81730092.0,
      "step": 39390
    },
    {
      "entropy": 5.620110034942627,
      "epoch": 3.573566763425254,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003762040026669606,
      "loss": 4.8762,
      "mean_token_accuracy": 0.21583929657936096,
      "num_tokens": 81741235.0,
      "step": 39395
    },
    {
      "entropy": 5.542365312576294,
      "epoch": 3.5740203193033384,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003761751030815524,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22471666634082793,
      "num_tokens": 81751434.0,
      "step": 39400
    },
    {
      "entropy": 5.568959188461304,
      "epoch": 3.5744738751814222,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003761462014038173,
      "loss": 4.885,
      "mean_token_accuracy": 0.21340702772140502,
      "num_tokens": 81763935.0,
      "step": 39405
    },
    {
      "entropy": 5.547994232177734,
      "epoch": 3.5749274310595065,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003761172976343528,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22052109986543655,
      "num_tokens": 81775007.0,
      "step": 39410
    },
    {
      "entropy": 5.599076938629151,
      "epoch": 3.5753809869375908,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037608839177375705,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22316237092018126,
      "num_tokens": 81785690.0,
      "step": 39415
    },
    {
      "entropy": 5.551195669174194,
      "epoch": 3.575834542815675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003760594838226275,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22254292964935302,
      "num_tokens": 81796143.0,
      "step": 39420
    },
    {
      "entropy": 5.677768850326538,
      "epoch": 3.576288098693759,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003760305737815621,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21999324560165406,
      "num_tokens": 81806607.0,
      "step": 39425
    },
    {
      "entropy": 5.5444282531738285,
      "epoch": 3.576741654571843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003760016616511586,
      "loss": 4.811,
      "mean_token_accuracy": 0.22482334226369857,
      "num_tokens": 81817216.0,
      "step": 39430
    },
    {
      "entropy": 5.5515465259552,
      "epoch": 3.5771952104499274,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037597274743201507,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22070506513118743,
      "num_tokens": 81827211.0,
      "step": 39435
    },
    {
      "entropy": 5.575213432312012,
      "epoch": 3.5776487663280117,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003759438311247294,
      "loss": 4.8157,
      "mean_token_accuracy": 0.2306778147816658,
      "num_tokens": 81837365.0,
      "step": 39440
    },
    {
      "entropy": 5.557666778564453,
      "epoch": 3.578102322206096,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003759149127298995,
      "loss": 4.8297,
      "mean_token_accuracy": 0.2288535326719284,
      "num_tokens": 81847803.0,
      "step": 39445
    },
    {
      "entropy": 5.528546380996704,
      "epoch": 3.57855587808418,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037588599224812363,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22581283301115035,
      "num_tokens": 81858123.0,
      "step": 39450
    },
    {
      "entropy": 5.501288938522339,
      "epoch": 3.579009433962264,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003758570696799997,
      "loss": 4.8253,
      "mean_token_accuracy": 0.23119283765554427,
      "num_tokens": 81868668.0,
      "step": 39455
    },
    {
      "entropy": 5.546344947814942,
      "epoch": 3.5794629898403483,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037582814502612575,
      "loss": 4.7989,
      "mean_token_accuracy": 0.2365462988615036,
      "num_tokens": 81877482.0,
      "step": 39460
    },
    {
      "entropy": 5.493146228790283,
      "epoch": 3.5799165457184325,
      "grad_norm": 1.375,
      "learning_rate": 0.0003757992182871001,
      "loss": 4.7216,
      "mean_token_accuracy": 0.23373111486434936,
      "num_tokens": 81886700.0,
      "step": 39465
    },
    {
      "entropy": 5.5622992515563965,
      "epoch": 3.580370101596517,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037577028946352107,
      "loss": 4.8467,
      "mean_token_accuracy": 0.22651808708906174,
      "num_tokens": 81896446.0,
      "step": 39470
    },
    {
      "entropy": 5.595756244659424,
      "epoch": 3.5808236574746006,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003757413585559866,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2276629701256752,
      "num_tokens": 81907418.0,
      "step": 39475
    },
    {
      "entropy": 5.608978509902954,
      "epoch": 3.581277213352685,
      "grad_norm": 1.25,
      "learning_rate": 0.00037571242556509526,
      "loss": 4.8391,
      "mean_token_accuracy": 0.225987708568573,
      "num_tokens": 81917940.0,
      "step": 39480
    },
    {
      "entropy": 5.521082782745362,
      "epoch": 3.581730769230769,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003756834904914453,
      "loss": 4.8314,
      "mean_token_accuracy": 0.223852701485157,
      "num_tokens": 81928158.0,
      "step": 39485
    },
    {
      "entropy": 5.614514350891113,
      "epoch": 3.5821843251088534,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003756545533356351,
      "loss": 4.9022,
      "mean_token_accuracy": 0.2230327919125557,
      "num_tokens": 81938111.0,
      "step": 39490
    },
    {
      "entropy": 5.6086221694946286,
      "epoch": 3.5826378809869377,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000375625614098263,
      "loss": 4.8501,
      "mean_token_accuracy": 0.22814274132251738,
      "num_tokens": 81948798.0,
      "step": 39495
    },
    {
      "entropy": 5.605922365188599,
      "epoch": 3.583091436865022,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037559667277992765,
      "loss": 4.8941,
      "mean_token_accuracy": 0.2272637203335762,
      "num_tokens": 81959240.0,
      "step": 39500
    },
    {
      "entropy": 5.561976337432862,
      "epoch": 3.5835449927431062,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003755677293812274,
      "loss": 4.8505,
      "mean_token_accuracy": 0.22572755962610244,
      "num_tokens": 81969929.0,
      "step": 39505
    },
    {
      "entropy": 5.638457345962524,
      "epoch": 3.58399854862119,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003755387839027609,
      "loss": 4.932,
      "mean_token_accuracy": 0.21088100522756575,
      "num_tokens": 81981007.0,
      "step": 39510
    },
    {
      "entropy": 5.674823331832886,
      "epoch": 3.5844521044992743,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037550983634512683,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21432940363883973,
      "num_tokens": 81991172.0,
      "step": 39515
    },
    {
      "entropy": 5.5497047901153564,
      "epoch": 3.5849056603773586,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003754808867089236,
      "loss": 4.7852,
      "mean_token_accuracy": 0.23025105744600297,
      "num_tokens": 82001282.0,
      "step": 39520
    },
    {
      "entropy": 5.567294120788574,
      "epoch": 3.5853592162554424,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037545193499475003,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22077575027942659,
      "num_tokens": 82011929.0,
      "step": 39525
    },
    {
      "entropy": 5.652157878875732,
      "epoch": 3.5858127721335267,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003754229812032049,
      "loss": 4.9082,
      "mean_token_accuracy": 0.22406138628721237,
      "num_tokens": 82021663.0,
      "step": 39530
    },
    {
      "entropy": 5.6033374786376955,
      "epoch": 3.586266328011611,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037539402533488694,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21336558759212493,
      "num_tokens": 82031547.0,
      "step": 39535
    },
    {
      "entropy": 5.64754090309143,
      "epoch": 3.586719883889695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003753650673903949,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21823500841856003,
      "num_tokens": 82041722.0,
      "step": 39540
    },
    {
      "entropy": 5.59665813446045,
      "epoch": 3.5871734397677795,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037533610737032773,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22547902315855026,
      "num_tokens": 82052715.0,
      "step": 39545
    },
    {
      "entropy": 5.549384260177613,
      "epoch": 3.5876269956458637,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037530714527528433,
      "loss": 4.8398,
      "mean_token_accuracy": 0.2286929801106453,
      "num_tokens": 82062709.0,
      "step": 39550
    },
    {
      "entropy": 5.596023178100586,
      "epoch": 3.588080551523948,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037527818110586354,
      "loss": 4.9022,
      "mean_token_accuracy": 0.2219044268131256,
      "num_tokens": 82072731.0,
      "step": 39555
    },
    {
      "entropy": 5.595376825332641,
      "epoch": 3.588534107402032,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003752492148626644,
      "loss": 4.8066,
      "mean_token_accuracy": 0.22683290690183638,
      "num_tokens": 82083216.0,
      "step": 39560
    },
    {
      "entropy": 5.594009208679199,
      "epoch": 3.588987663280116,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000375220246546286,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21765778958797455,
      "num_tokens": 82094482.0,
      "step": 39565
    },
    {
      "entropy": 5.583010053634643,
      "epoch": 3.5894412191582004,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003751912761573273,
      "loss": 4.8378,
      "mean_token_accuracy": 0.22457533180713654,
      "num_tokens": 82104988.0,
      "step": 39570
    },
    {
      "entropy": 5.603489589691162,
      "epoch": 3.5898947750362846,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037516230369638755,
      "loss": 4.8995,
      "mean_token_accuracy": 0.21480146050453186,
      "num_tokens": 82114455.0,
      "step": 39575
    },
    {
      "entropy": 5.600613832473755,
      "epoch": 3.5903483309143684,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003751333291640658,
      "loss": 4.8843,
      "mean_token_accuracy": 0.22257035672664643,
      "num_tokens": 82125001.0,
      "step": 39580
    },
    {
      "entropy": 5.5865860939025875,
      "epoch": 3.5908018867924527,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037510435256096124,
      "loss": 4.8542,
      "mean_token_accuracy": 0.2213310644030571,
      "num_tokens": 82134586.0,
      "step": 39585
    },
    {
      "entropy": 5.537910461425781,
      "epoch": 3.591255442670537,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037507537388767326,
      "loss": 4.8309,
      "mean_token_accuracy": 0.22830667495727539,
      "num_tokens": 82144612.0,
      "step": 39590
    },
    {
      "entropy": 5.5490772247314455,
      "epoch": 3.5917089985486212,
      "grad_norm": 1.3125,
      "learning_rate": 0.000375046393144801,
      "loss": 4.8456,
      "mean_token_accuracy": 0.22967538088560105,
      "num_tokens": 82154913.0,
      "step": 39595
    },
    {
      "entropy": 5.601358556747437,
      "epoch": 3.5921625544267055,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003750174103329438,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21824212074279786,
      "num_tokens": 82164469.0,
      "step": 39600
    },
    {
      "entropy": 5.621843481063843,
      "epoch": 3.59261611030479,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037498842545270105,
      "loss": 4.897,
      "mean_token_accuracy": 0.22346892207860947,
      "num_tokens": 82174575.0,
      "step": 39605
    },
    {
      "entropy": 5.627682876586914,
      "epoch": 3.5930696661828736,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037495943850467226,
      "loss": 4.8686,
      "mean_token_accuracy": 0.2256598725914955,
      "num_tokens": 82184661.0,
      "step": 39610
    },
    {
      "entropy": 5.659826612472534,
      "epoch": 3.593523222060958,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037493044948945677,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2209211990237236,
      "num_tokens": 82194877.0,
      "step": 39615
    },
    {
      "entropy": 5.604482173919678,
      "epoch": 3.593976777939042,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037490145840765414,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22413596659898757,
      "num_tokens": 82206278.0,
      "step": 39620
    },
    {
      "entropy": 5.592150545120239,
      "epoch": 3.5944303338171264,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037487246525986384,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2219835937023163,
      "num_tokens": 82216140.0,
      "step": 39625
    },
    {
      "entropy": 5.6267670631408695,
      "epoch": 3.59488388969521,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003748434700466856,
      "loss": 4.9245,
      "mean_token_accuracy": 0.22556490004062651,
      "num_tokens": 82226239.0,
      "step": 39630
    },
    {
      "entropy": 5.581395816802979,
      "epoch": 3.5953374455732945,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003748144727687188,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22574800699949266,
      "num_tokens": 82235795.0,
      "step": 39635
    },
    {
      "entropy": 5.524588584899902,
      "epoch": 3.5957910014513788,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037478547342656345,
      "loss": 4.811,
      "mean_token_accuracy": 0.22947650402784348,
      "num_tokens": 82245644.0,
      "step": 39640
    },
    {
      "entropy": 5.562024545669556,
      "epoch": 3.596244557329463,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000374756472020819,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22304305285215378,
      "num_tokens": 82255608.0,
      "step": 39645
    },
    {
      "entropy": 5.5828674793243405,
      "epoch": 3.5966981132075473,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037472746855208525,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2244326055049896,
      "num_tokens": 82265488.0,
      "step": 39650
    },
    {
      "entropy": 5.608457946777344,
      "epoch": 3.5971516690856316,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003746984630209621,
      "loss": 4.8877,
      "mean_token_accuracy": 0.2146323651075363,
      "num_tokens": 82276866.0,
      "step": 39655
    },
    {
      "entropy": 5.587748098373413,
      "epoch": 3.5976052249637154,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037466945542804933,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22724784761667252,
      "num_tokens": 82287125.0,
      "step": 39660
    },
    {
      "entropy": 5.620104074478149,
      "epoch": 3.5980587808417996,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037464044577394676,
      "loss": 4.908,
      "mean_token_accuracy": 0.2170126736164093,
      "num_tokens": 82296887.0,
      "step": 39665
    },
    {
      "entropy": 5.6313426971435545,
      "epoch": 3.598512336719884,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003746114340592544,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21381559371948242,
      "num_tokens": 82306928.0,
      "step": 39670
    },
    {
      "entropy": 5.5836591720581055,
      "epoch": 3.598965892597968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037458242028457216,
      "loss": 4.9002,
      "mean_token_accuracy": 0.2217973828315735,
      "num_tokens": 82317873.0,
      "step": 39675
    },
    {
      "entropy": 5.5234252452850345,
      "epoch": 3.599419448476052,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037455340445050023,
      "loss": 4.7642,
      "mean_token_accuracy": 0.22885881811380387,
      "num_tokens": 82328184.0,
      "step": 39680
    },
    {
      "entropy": 5.523097705841065,
      "epoch": 3.5998730043541363,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037452438655763843,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22360922247171403,
      "num_tokens": 82338191.0,
      "step": 39685
    },
    {
      "entropy": 5.5666886329650875,
      "epoch": 3.6003265602322205,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037449536660658696,
      "loss": 4.9551,
      "mean_token_accuracy": 0.21854311972856522,
      "num_tokens": 82348095.0,
      "step": 39690
    },
    {
      "entropy": 5.575230693817138,
      "epoch": 3.600780116110305,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000374466344597946,
      "loss": 4.7863,
      "mean_token_accuracy": 0.2260938763618469,
      "num_tokens": 82358216.0,
      "step": 39695
    },
    {
      "entropy": 5.542170906066895,
      "epoch": 3.601233671988389,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003744373205323157,
      "loss": 4.8326,
      "mean_token_accuracy": 0.2243800014257431,
      "num_tokens": 82368581.0,
      "step": 39700
    },
    {
      "entropy": 5.551475381851196,
      "epoch": 3.6016872278664733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003744082944102962,
      "loss": 4.8803,
      "mean_token_accuracy": 0.21467635184526443,
      "num_tokens": 82379609.0,
      "step": 39705
    },
    {
      "entropy": 5.628689384460449,
      "epoch": 3.6021407837445576,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003743792662324878,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22144703567028046,
      "num_tokens": 82390589.0,
      "step": 39710
    },
    {
      "entropy": 5.552912616729737,
      "epoch": 3.6025943396226414,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037435023599949086,
      "loss": 4.8554,
      "mean_token_accuracy": 0.22040871828794478,
      "num_tokens": 82400915.0,
      "step": 39715
    },
    {
      "entropy": 5.563165950775146,
      "epoch": 3.6030478955007257,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037432120371190585,
      "loss": 4.8561,
      "mean_token_accuracy": 0.22860354781150818,
      "num_tokens": 82411544.0,
      "step": 39720
    },
    {
      "entropy": 5.55045337677002,
      "epoch": 3.60350145137881,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037429216937033294,
      "loss": 4.8228,
      "mean_token_accuracy": 0.228674952685833,
      "num_tokens": 82422391.0,
      "step": 39725
    },
    {
      "entropy": 5.536394834518433,
      "epoch": 3.6039550072568938,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037426313297537263,
      "loss": 4.861,
      "mean_token_accuracy": 0.224294613301754,
      "num_tokens": 82433564.0,
      "step": 39730
    },
    {
      "entropy": 5.5612335205078125,
      "epoch": 3.604408563134978,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003742340945276255,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22343081682920457,
      "num_tokens": 82445005.0,
      "step": 39735
    },
    {
      "entropy": 5.566359996795654,
      "epoch": 3.6048621190130623,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000374205054027692,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21974200904369354,
      "num_tokens": 82455998.0,
      "step": 39740
    },
    {
      "entropy": 5.563248348236084,
      "epoch": 3.6053156748911466,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037417601147617265,
      "loss": 4.8323,
      "mean_token_accuracy": 0.22561975121498107,
      "num_tokens": 82466789.0,
      "step": 39745
    },
    {
      "entropy": 5.630987739562988,
      "epoch": 3.605769230769231,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037414696687366816,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21442769169807435,
      "num_tokens": 82478181.0,
      "step": 39750
    },
    {
      "entropy": 5.594155979156494,
      "epoch": 3.606222786647315,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037411792022077917,
      "loss": 4.8245,
      "mean_token_accuracy": 0.2188940703868866,
      "num_tokens": 82488452.0,
      "step": 39755
    },
    {
      "entropy": 5.612184762954712,
      "epoch": 3.6066763425253994,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003740888715181063,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2234674036502838,
      "num_tokens": 82498179.0,
      "step": 39760
    },
    {
      "entropy": 5.556216192245484,
      "epoch": 3.607129898403483,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037405982076625026,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22625814378261566,
      "num_tokens": 82508242.0,
      "step": 39765
    },
    {
      "entropy": 5.6044919967651365,
      "epoch": 3.6075834542815675,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037403076796581195,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21506549119949342,
      "num_tokens": 82519363.0,
      "step": 39770
    },
    {
      "entropy": 5.6317511081695555,
      "epoch": 3.6080370101596517,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003740017131173921,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2176603466272354,
      "num_tokens": 82529745.0,
      "step": 39775
    },
    {
      "entropy": 5.652537202835083,
      "epoch": 3.608490566037736,
      "grad_norm": 1.25,
      "learning_rate": 0.0003739726562215916,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21786196827888488,
      "num_tokens": 82539573.0,
      "step": 39780
    },
    {
      "entropy": 5.634916973114014,
      "epoch": 3.60894412191582,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037394359727901136,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22218762338161469,
      "num_tokens": 82551313.0,
      "step": 39785
    },
    {
      "entropy": 5.560735130310059,
      "epoch": 3.609397677793904,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003739145362902523,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22180914282798767,
      "num_tokens": 82561283.0,
      "step": 39790
    },
    {
      "entropy": 5.529203462600708,
      "epoch": 3.6098512336719883,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003738854732559153,
      "loss": 4.7904,
      "mean_token_accuracy": 0.22714564055204392,
      "num_tokens": 82571118.0,
      "step": 39795
    },
    {
      "entropy": 5.5521503448486325,
      "epoch": 3.6103047895500726,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003738564081766017,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22489626407623292,
      "num_tokens": 82581663.0,
      "step": 39800
    },
    {
      "entropy": 5.528533458709717,
      "epoch": 3.610758345428157,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037382734105291234,
      "loss": 4.8367,
      "mean_token_accuracy": 0.22135457247495652,
      "num_tokens": 82592593.0,
      "step": 39805
    },
    {
      "entropy": 5.568611001968383,
      "epoch": 3.611211901306241,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037379827188544835,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21753670126199723,
      "num_tokens": 82603047.0,
      "step": 39810
    },
    {
      "entropy": 5.647225999832154,
      "epoch": 3.611665457184325,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037376920067481095,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22242050021886825,
      "num_tokens": 82612789.0,
      "step": 39815
    },
    {
      "entropy": 5.579960584640503,
      "epoch": 3.6121190130624092,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003737401274216012,
      "loss": 4.8454,
      "mean_token_accuracy": 0.22030610889196395,
      "num_tokens": 82622881.0,
      "step": 39820
    },
    {
      "entropy": 5.5353028774261475,
      "epoch": 3.6125725689404935,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003737110521264206,
      "loss": 4.8037,
      "mean_token_accuracy": 0.22181320935487747,
      "num_tokens": 82633933.0,
      "step": 39825
    },
    {
      "entropy": 5.5855536460876465,
      "epoch": 3.6130261248185778,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003736819747898702,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2231051057577133,
      "num_tokens": 82645069.0,
      "step": 39830
    },
    {
      "entropy": 5.669990015029907,
      "epoch": 3.6134796806966616,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037365289541255154,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22316997200250627,
      "num_tokens": 82655152.0,
      "step": 39835
    },
    {
      "entropy": 5.54179892539978,
      "epoch": 3.613933236574746,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037362381399506575,
      "loss": 4.8441,
      "mean_token_accuracy": 0.2198281854391098,
      "num_tokens": 82665260.0,
      "step": 39840
    },
    {
      "entropy": 5.551840686798096,
      "epoch": 3.61438679245283,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037359473053801435,
      "loss": 4.8463,
      "mean_token_accuracy": 0.2278337746858597,
      "num_tokens": 82676287.0,
      "step": 39845
    },
    {
      "entropy": 5.582655715942383,
      "epoch": 3.6148403483309144,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003735656450419988,
      "loss": 4.9597,
      "mean_token_accuracy": 0.21073424965143203,
      "num_tokens": 82686958.0,
      "step": 39850
    },
    {
      "entropy": 5.616000747680664,
      "epoch": 3.6152939042089987,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037353655750762063,
      "loss": 5.004,
      "mean_token_accuracy": 0.20761482566595077,
      "num_tokens": 82698623.0,
      "step": 39855
    },
    {
      "entropy": 5.675088262557983,
      "epoch": 3.615747460087083,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037350746793548135,
      "loss": 4.9531,
      "mean_token_accuracy": 0.21714994609355925,
      "num_tokens": 82709146.0,
      "step": 39860
    },
    {
      "entropy": 5.605067300796509,
      "epoch": 3.6162010159651667,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037347837632618246,
      "loss": 4.8258,
      "mean_token_accuracy": 0.22284420430660248,
      "num_tokens": 82719323.0,
      "step": 39865
    },
    {
      "entropy": 5.6138804912567135,
      "epoch": 3.616654571843251,
      "grad_norm": 1.421875,
      "learning_rate": 0.00037344928268032574,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21643898636102676,
      "num_tokens": 82730068.0,
      "step": 39870
    },
    {
      "entropy": 5.568813514709473,
      "epoch": 3.6171081277213353,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003734201869985127,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2208503380417824,
      "num_tokens": 82740460.0,
      "step": 39875
    },
    {
      "entropy": 5.5625104904174805,
      "epoch": 3.6175616835994195,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003733910892813451,
      "loss": 4.8382,
      "mean_token_accuracy": 0.21515371948480605,
      "num_tokens": 82751488.0,
      "step": 39880
    },
    {
      "entropy": 5.494701814651489,
      "epoch": 3.6180152394775034,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037336198952942477,
      "loss": 4.768,
      "mean_token_accuracy": 0.23381744027137757,
      "num_tokens": 82762061.0,
      "step": 39885
    },
    {
      "entropy": 5.543667030334473,
      "epoch": 3.6184687953555876,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003733328877433534,
      "loss": 4.8239,
      "mean_token_accuracy": 0.22706289291381837,
      "num_tokens": 82773421.0,
      "step": 39890
    },
    {
      "entropy": 5.5062695980072025,
      "epoch": 3.618922351233672,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037330378392373285,
      "loss": 4.7367,
      "mean_token_accuracy": 0.22476134449243546,
      "num_tokens": 82783342.0,
      "step": 39895
    },
    {
      "entropy": 5.564427900314331,
      "epoch": 3.619375907111756,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037327467807116507,
      "loss": 4.8863,
      "mean_token_accuracy": 0.21438806056976317,
      "num_tokens": 82793885.0,
      "step": 39900
    },
    {
      "entropy": 5.630543661117554,
      "epoch": 3.6198294629898404,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037324557018625186,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21958522051572799,
      "num_tokens": 82803746.0,
      "step": 39905
    },
    {
      "entropy": 5.594546794891357,
      "epoch": 3.6202830188679247,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003732164602695952,
      "loss": 4.864,
      "mean_token_accuracy": 0.22522586435079575,
      "num_tokens": 82813858.0,
      "step": 39910
    },
    {
      "entropy": 5.6029878616333,
      "epoch": 3.620736574746009,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037318734832179705,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22916320860385894,
      "num_tokens": 82824358.0,
      "step": 39915
    },
    {
      "entropy": 5.536659240722656,
      "epoch": 3.621190130624093,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037315823434345963,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2316565543413162,
      "num_tokens": 82835801.0,
      "step": 39920
    },
    {
      "entropy": 5.493958568572998,
      "epoch": 3.621643686502177,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003731291183351848,
      "loss": 4.7965,
      "mean_token_accuracy": 0.23135312646627426,
      "num_tokens": 82845818.0,
      "step": 39925
    },
    {
      "entropy": 5.5462545394897464,
      "epoch": 3.6220972423802613,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037310000029757486,
      "loss": 4.847,
      "mean_token_accuracy": 0.2225813940167427,
      "num_tokens": 82856920.0,
      "step": 39930
    },
    {
      "entropy": 5.649670124053955,
      "epoch": 3.6225507982583456,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003730708802312319,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21842000186443328,
      "num_tokens": 82867220.0,
      "step": 39935
    },
    {
      "entropy": 5.564120817184448,
      "epoch": 3.6230043541364294,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003730417581367582,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2275705099105835,
      "num_tokens": 82877541.0,
      "step": 39940
    },
    {
      "entropy": 5.602983903884888,
      "epoch": 3.6234579100145137,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003730126340147559,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22266925722360612,
      "num_tokens": 82888346.0,
      "step": 39945
    },
    {
      "entropy": 5.655581283569336,
      "epoch": 3.623911465892598,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003729835078658273,
      "loss": 4.9603,
      "mean_token_accuracy": 0.22218163758516313,
      "num_tokens": 82899721.0,
      "step": 39950
    },
    {
      "entropy": 5.587097883224487,
      "epoch": 3.624365021770682,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037295437969057493,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22589590698480605,
      "num_tokens": 82910149.0,
      "step": 39955
    },
    {
      "entropy": 5.598569631576538,
      "epoch": 3.6248185776487665,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037292524948960093,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22186598926782608,
      "num_tokens": 82920883.0,
      "step": 39960
    },
    {
      "entropy": 5.59524245262146,
      "epoch": 3.6252721335268507,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003728961172635078,
      "loss": 4.923,
      "mean_token_accuracy": 0.22082389891147614,
      "num_tokens": 82931091.0,
      "step": 39965
    },
    {
      "entropy": 5.566177225112915,
      "epoch": 3.6257256894049346,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037286698301289804,
      "loss": 4.7764,
      "mean_token_accuracy": 0.2263589769601822,
      "num_tokens": 82941108.0,
      "step": 39970
    },
    {
      "entropy": 5.556240177154541,
      "epoch": 3.626179245283019,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003728378467383742,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22205670028924943,
      "num_tokens": 82951943.0,
      "step": 39975
    },
    {
      "entropy": 5.650640916824341,
      "epoch": 3.626632801161103,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003728087084405387,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22331616878509522,
      "num_tokens": 82961495.0,
      "step": 39980
    },
    {
      "entropy": 5.672296333312988,
      "epoch": 3.6270863570391874,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037277956811999414,
      "loss": 4.8963,
      "mean_token_accuracy": 0.22200254797935487,
      "num_tokens": 82971858.0,
      "step": 39985
    },
    {
      "entropy": 5.616363096237182,
      "epoch": 3.627539912917271,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003727504257773433,
      "loss": 4.8718,
      "mean_token_accuracy": 0.2186845749616623,
      "num_tokens": 82982805.0,
      "step": 39990
    },
    {
      "entropy": 5.5951322555542,
      "epoch": 3.6279934687953554,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037272128141318866,
      "loss": 4.9172,
      "mean_token_accuracy": 0.2165425941348076,
      "num_tokens": 82993446.0,
      "step": 39995
    },
    {
      "entropy": 5.533211183547974,
      "epoch": 3.6284470246734397,
      "grad_norm": 1.34375,
      "learning_rate": 0.000372692135028133,
      "loss": 4.8052,
      "mean_token_accuracy": 0.22380539029836655,
      "num_tokens": 83003378.0,
      "step": 40000
    },
    {
      "entropy": 5.586108732223511,
      "epoch": 3.628900580551524,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037266298662277914,
      "loss": 4.8209,
      "mean_token_accuracy": 0.22107890248298645,
      "num_tokens": 83014223.0,
      "step": 40005
    },
    {
      "entropy": 5.576086044311523,
      "epoch": 3.6293541364296082,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037263383619772986,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21756452620029448,
      "num_tokens": 83024691.0,
      "step": 40010
    },
    {
      "entropy": 5.608897304534912,
      "epoch": 3.6298076923076925,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037260468375358794,
      "loss": 4.881,
      "mean_token_accuracy": 0.21757373511791228,
      "num_tokens": 83035704.0,
      "step": 40015
    },
    {
      "entropy": 5.638431787490845,
      "epoch": 3.6302612481857763,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003725755292909562,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2153886079788208,
      "num_tokens": 83047173.0,
      "step": 40020
    },
    {
      "entropy": 5.641260099411011,
      "epoch": 3.6307148040638606,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003725463728104378,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21805346012115479,
      "num_tokens": 83056399.0,
      "step": 40025
    },
    {
      "entropy": 5.571602010726929,
      "epoch": 3.631168359941945,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003725172143126354,
      "loss": 4.8305,
      "mean_token_accuracy": 0.21817951500415803,
      "num_tokens": 83066413.0,
      "step": 40030
    },
    {
      "entropy": 5.582342195510864,
      "epoch": 3.631621915820029,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037248805379815227,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21812659054994582,
      "num_tokens": 83077843.0,
      "step": 40035
    },
    {
      "entropy": 5.628216361999511,
      "epoch": 3.632075471698113,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003724588912675913,
      "loss": 4.9704,
      "mean_token_accuracy": 0.21749621033668518,
      "num_tokens": 83088585.0,
      "step": 40040
    },
    {
      "entropy": 5.670231628417969,
      "epoch": 3.6325290275761972,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003724297267215557,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21903275847434997,
      "num_tokens": 83099295.0,
      "step": 40045
    },
    {
      "entropy": 5.562008905410766,
      "epoch": 3.6329825834542815,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037240056016064844,
      "loss": 4.7939,
      "mean_token_accuracy": 0.23146501928567886,
      "num_tokens": 83108530.0,
      "step": 40050
    },
    {
      "entropy": 5.563482046127319,
      "epoch": 3.6334361393323658,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003723713915854728,
      "loss": 4.8824,
      "mean_token_accuracy": 0.22078048139810563,
      "num_tokens": 83119703.0,
      "step": 40055
    },
    {
      "entropy": 5.606525039672851,
      "epoch": 3.63388969521045,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003723422209966319,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22562302201986312,
      "num_tokens": 83129323.0,
      "step": 40060
    },
    {
      "entropy": 5.58606071472168,
      "epoch": 3.6343432510885343,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003723130483947292,
      "loss": 4.8128,
      "mean_token_accuracy": 0.22730310559272765,
      "num_tokens": 83139339.0,
      "step": 40065
    },
    {
      "entropy": 5.546669578552246,
      "epoch": 3.6347968069666186,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003722838737803678,
      "loss": 4.863,
      "mean_token_accuracy": 0.22367920428514482,
      "num_tokens": 83149845.0,
      "step": 40070
    },
    {
      "entropy": 5.603301191329956,
      "epoch": 3.6352503628447024,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037225469715415106,
      "loss": 4.8964,
      "mean_token_accuracy": 0.21535771042108537,
      "num_tokens": 83160332.0,
      "step": 40075
    },
    {
      "entropy": 5.610947322845459,
      "epoch": 3.6357039187227866,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037222551851668245,
      "loss": 4.902,
      "mean_token_accuracy": 0.2161208689212799,
      "num_tokens": 83170892.0,
      "step": 40080
    },
    {
      "entropy": 5.567786836624146,
      "epoch": 3.636157474600871,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003721963378685652,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22651458382606507,
      "num_tokens": 83181035.0,
      "step": 40085
    },
    {
      "entropy": 5.546350383758545,
      "epoch": 3.6366110304789547,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037216715521040314,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23031052500009536,
      "num_tokens": 83191886.0,
      "step": 40090
    },
    {
      "entropy": 5.5786596775054935,
      "epoch": 3.637064586357039,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003721379705427994,
      "loss": 4.8555,
      "mean_token_accuracy": 0.2228209063410759,
      "num_tokens": 83202356.0,
      "step": 40095
    },
    {
      "entropy": 5.588511323928833,
      "epoch": 3.6375181422351233,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037210878386635774,
      "loss": 4.9022,
      "mean_token_accuracy": 0.2290998339653015,
      "num_tokens": 83212909.0,
      "step": 40100
    },
    {
      "entropy": 5.614213991165161,
      "epoch": 3.6379716981132075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037207959518168165,
      "loss": 4.8945,
      "mean_token_accuracy": 0.21607406586408615,
      "num_tokens": 83223523.0,
      "step": 40105
    },
    {
      "entropy": 5.720131683349609,
      "epoch": 3.638425253991292,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037205040448937485,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2137068212032318,
      "num_tokens": 83233665.0,
      "step": 40110
    },
    {
      "entropy": 5.582072639465332,
      "epoch": 3.638878809869376,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003720212117900409,
      "loss": 4.7939,
      "mean_token_accuracy": 0.23158206194639205,
      "num_tokens": 83243563.0,
      "step": 40115
    },
    {
      "entropy": 5.591418981552124,
      "epoch": 3.6393323657474603,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037199201708428356,
      "loss": 4.924,
      "mean_token_accuracy": 0.2201562389731407,
      "num_tokens": 83254246.0,
      "step": 40120
    },
    {
      "entropy": 5.602092361450195,
      "epoch": 3.639785921625544,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003719628203727066,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22000578045845032,
      "num_tokens": 83264361.0,
      "step": 40125
    },
    {
      "entropy": 5.625526285171508,
      "epoch": 3.6402394775036284,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003719336216559138,
      "loss": 4.9022,
      "mean_token_accuracy": 0.22219188064336776,
      "num_tokens": 83274193.0,
      "step": 40130
    },
    {
      "entropy": 5.590416049957275,
      "epoch": 3.6406930333817127,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000371904420934509,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22752234786748887,
      "num_tokens": 83284593.0,
      "step": 40135
    },
    {
      "entropy": 5.535084199905396,
      "epoch": 3.641146589259797,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003718752182090961,
      "loss": 4.8392,
      "mean_token_accuracy": 0.22235280573368071,
      "num_tokens": 83294602.0,
      "step": 40140
    },
    {
      "entropy": 5.563156509399414,
      "epoch": 3.6416001451378808,
      "grad_norm": 1.234375,
      "learning_rate": 0.000371846013480279,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22184473574161528,
      "num_tokens": 83305399.0,
      "step": 40145
    },
    {
      "entropy": 5.512012386322022,
      "epoch": 3.642053701015965,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037181680674866165,
      "loss": 4.8322,
      "mean_token_accuracy": 0.21899324059486389,
      "num_tokens": 83315696.0,
      "step": 40150
    },
    {
      "entropy": 5.659463214874267,
      "epoch": 3.6425072568940493,
      "grad_norm": 1.25,
      "learning_rate": 0.0003717875980148481,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21547501236200334,
      "num_tokens": 83326686.0,
      "step": 40155
    },
    {
      "entropy": 5.587709951400757,
      "epoch": 3.6429608127721336,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003717583872794423,
      "loss": 4.8518,
      "mean_token_accuracy": 0.23250453621149064,
      "num_tokens": 83336902.0,
      "step": 40160
    },
    {
      "entropy": 5.576013231277466,
      "epoch": 3.643414368650218,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003717291745430484,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22342123836278915,
      "num_tokens": 83347324.0,
      "step": 40165
    },
    {
      "entropy": 5.516481637954712,
      "epoch": 3.643867924528302,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037169995980627056,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22125705778598787,
      "num_tokens": 83358458.0,
      "step": 40170
    },
    {
      "entropy": 5.533251857757568,
      "epoch": 3.644321480406386,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037167074306971285,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2192295253276825,
      "num_tokens": 83369047.0,
      "step": 40175
    },
    {
      "entropy": 5.598499536514282,
      "epoch": 3.64477503628447,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037164152433397964,
      "loss": 4.8663,
      "mean_token_accuracy": 0.21933342069387435,
      "num_tokens": 83378891.0,
      "step": 40180
    },
    {
      "entropy": 5.582867956161499,
      "epoch": 3.6452285921625545,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000371612303599675,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22062188982963563,
      "num_tokens": 83389289.0,
      "step": 40185
    },
    {
      "entropy": 5.671322393417358,
      "epoch": 3.6456821480406387,
      "grad_norm": 1.25,
      "learning_rate": 0.00037158308086740336,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21880175769329072,
      "num_tokens": 83399316.0,
      "step": 40190
    },
    {
      "entropy": 5.586907243728637,
      "epoch": 3.6461357039187225,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003715538561377689,
      "loss": 4.8887,
      "mean_token_accuracy": 0.22841109782457353,
      "num_tokens": 83409507.0,
      "step": 40195
    },
    {
      "entropy": 5.6554450511932375,
      "epoch": 3.646589259796807,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003715246294113762,
      "loss": 4.9862,
      "mean_token_accuracy": 0.2105707585811615,
      "num_tokens": 83419687.0,
      "step": 40200
    },
    {
      "entropy": 5.688238096237183,
      "epoch": 3.647042815674891,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037149540068882945,
      "loss": 4.9758,
      "mean_token_accuracy": 0.21756171584129333,
      "num_tokens": 83429711.0,
      "step": 40205
    },
    {
      "entropy": 5.570241928100586,
      "epoch": 3.6474963715529753,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037146616997073327,
      "loss": 4.8434,
      "mean_token_accuracy": 0.21197809725999833,
      "num_tokens": 83440116.0,
      "step": 40210
    },
    {
      "entropy": 5.57893443107605,
      "epoch": 3.6479499274310596,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037143693725769214,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22238032817840575,
      "num_tokens": 83450877.0,
      "step": 40215
    },
    {
      "entropy": 5.524405193328858,
      "epoch": 3.648403483309144,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037140770255031066,
      "loss": 4.77,
      "mean_token_accuracy": 0.2300071597099304,
      "num_tokens": 83460080.0,
      "step": 40220
    },
    {
      "entropy": 5.5808837890625,
      "epoch": 3.6488570391872277,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037137846584919316,
      "loss": 4.934,
      "mean_token_accuracy": 0.2192599281668663,
      "num_tokens": 83469933.0,
      "step": 40225
    },
    {
      "entropy": 5.552589797973633,
      "epoch": 3.649310595065312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003713492271549445,
      "loss": 4.8278,
      "mean_token_accuracy": 0.22599529922008516,
      "num_tokens": 83480477.0,
      "step": 40230
    },
    {
      "entropy": 5.562912702560425,
      "epoch": 3.6497641509433962,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003713199864681693,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22306572645902634,
      "num_tokens": 83491022.0,
      "step": 40235
    },
    {
      "entropy": 5.633314943313598,
      "epoch": 3.6502177068214805,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037129074378947216,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21957073658704757,
      "num_tokens": 83501643.0,
      "step": 40240
    },
    {
      "entropy": 5.642423486709594,
      "epoch": 3.6506712626995643,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037126149911945796,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21100935637950896,
      "num_tokens": 83511948.0,
      "step": 40245
    },
    {
      "entropy": 5.5929368019104,
      "epoch": 3.6511248185776486,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037123225245873137,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22216069996356963,
      "num_tokens": 83521939.0,
      "step": 40250
    },
    {
      "entropy": 5.677875185012818,
      "epoch": 3.651578374455733,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037120300380789727,
      "loss": 4.9161,
      "mean_token_accuracy": 0.221902734041214,
      "num_tokens": 83532098.0,
      "step": 40255
    },
    {
      "entropy": 5.617684268951416,
      "epoch": 3.652031930333817,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037117375316756063,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2231852427124977,
      "num_tokens": 83543238.0,
      "step": 40260
    },
    {
      "entropy": 5.612978982925415,
      "epoch": 3.6524854862119014,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003711445005383263,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22354606091976165,
      "num_tokens": 83552969.0,
      "step": 40265
    },
    {
      "entropy": 5.633970308303833,
      "epoch": 3.6529390420899857,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037111524592079896,
      "loss": 4.9434,
      "mean_token_accuracy": 0.20890241712331772,
      "num_tokens": 83564569.0,
      "step": 40270
    },
    {
      "entropy": 5.618559503555298,
      "epoch": 3.65339259796807,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000371085989315584,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22019869089126587,
      "num_tokens": 83575613.0,
      "step": 40275
    },
    {
      "entropy": 5.6460196018219,
      "epoch": 3.6538461538461537,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003710567307232864,
      "loss": 4.927,
      "mean_token_accuracy": 0.21827946603298187,
      "num_tokens": 83586311.0,
      "step": 40280
    },
    {
      "entropy": 5.640784692764282,
      "epoch": 3.654299709724238,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037102747014451095,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21777148246765138,
      "num_tokens": 83596660.0,
      "step": 40285
    },
    {
      "entropy": 5.63956937789917,
      "epoch": 3.6547532656023223,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000370998207579863,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22158859074115753,
      "num_tokens": 83606882.0,
      "step": 40290
    },
    {
      "entropy": 5.6697368144989015,
      "epoch": 3.6552068214804065,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003709689430299477,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22426167130470276,
      "num_tokens": 83617380.0,
      "step": 40295
    },
    {
      "entropy": 5.635519790649414,
      "epoch": 3.6556603773584904,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003709396764953701,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22047493755817413,
      "num_tokens": 83627961.0,
      "step": 40300
    },
    {
      "entropy": 5.460635232925415,
      "epoch": 3.6561139332365746,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003709104079767357,
      "loss": 4.7874,
      "mean_token_accuracy": 0.23177144974470137,
      "num_tokens": 83637646.0,
      "step": 40305
    },
    {
      "entropy": 5.571348476409912,
      "epoch": 3.656567489114659,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037088113747464955,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22469684183597566,
      "num_tokens": 83647807.0,
      "step": 40310
    },
    {
      "entropy": 5.568463754653931,
      "epoch": 3.657021044992743,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000370851864989717,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21802911311388015,
      "num_tokens": 83658002.0,
      "step": 40315
    },
    {
      "entropy": 5.535338544845581,
      "epoch": 3.6574746008708274,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003708225905225435,
      "loss": 4.856,
      "mean_token_accuracy": 0.2304181396961212,
      "num_tokens": 83667875.0,
      "step": 40320
    },
    {
      "entropy": 5.571807622909546,
      "epoch": 3.6579281567489117,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003707933140737345,
      "loss": 4.8686,
      "mean_token_accuracy": 0.2227008119225502,
      "num_tokens": 83678376.0,
      "step": 40325
    },
    {
      "entropy": 5.62206802368164,
      "epoch": 3.6583817126269955,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037076403564389516,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20957673490047454,
      "num_tokens": 83688318.0,
      "step": 40330
    },
    {
      "entropy": 5.606731748580932,
      "epoch": 3.65883526850508,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003707347552336313,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21604602932929992,
      "num_tokens": 83698935.0,
      "step": 40335
    },
    {
      "entropy": 5.6294807434082035,
      "epoch": 3.659288824383164,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003707054728435482,
      "loss": 4.9682,
      "mean_token_accuracy": 0.22047323882579803,
      "num_tokens": 83709382.0,
      "step": 40340
    },
    {
      "entropy": 5.588862323760987,
      "epoch": 3.6597423802612483,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003706761884742516,
      "loss": 4.7945,
      "mean_token_accuracy": 0.2298911526799202,
      "num_tokens": 83719661.0,
      "step": 40345
    },
    {
      "entropy": 5.668063068389893,
      "epoch": 3.660195936139332,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037064690212634707,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2175290808081627,
      "num_tokens": 83730679.0,
      "step": 40350
    },
    {
      "entropy": 5.542167043685913,
      "epoch": 3.6606494920174164,
      "grad_norm": 1.25,
      "learning_rate": 0.0003706176138004401,
      "loss": 4.8039,
      "mean_token_accuracy": 0.22326629906892775,
      "num_tokens": 83740595.0,
      "step": 40355
    },
    {
      "entropy": 5.584343957901001,
      "epoch": 3.6611030478955007,
      "grad_norm": 1.25,
      "learning_rate": 0.00037058832349713667,
      "loss": 4.92,
      "mean_token_accuracy": 0.22128765732049943,
      "num_tokens": 83751605.0,
      "step": 40360
    },
    {
      "entropy": 5.651593971252441,
      "epoch": 3.661556603773585,
      "grad_norm": 1.375,
      "learning_rate": 0.0003705590312170422,
      "loss": 4.9971,
      "mean_token_accuracy": 0.21514691263437272,
      "num_tokens": 83763780.0,
      "step": 40365
    },
    {
      "entropy": 5.585455083847046,
      "epoch": 3.662010159651669,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037052973696076274,
      "loss": 4.8323,
      "mean_token_accuracy": 0.23000090271234513,
      "num_tokens": 83773800.0,
      "step": 40370
    },
    {
      "entropy": 5.657545375823974,
      "epoch": 3.6624637155297535,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037050044072890395,
      "loss": 4.9408,
      "mean_token_accuracy": 0.20326773226261138,
      "num_tokens": 83784402.0,
      "step": 40375
    },
    {
      "entropy": 5.556073427200317,
      "epoch": 3.6629172714078373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003704711425220717,
      "loss": 4.8505,
      "mean_token_accuracy": 0.21633521169424058,
      "num_tokens": 83794852.0,
      "step": 40380
    },
    {
      "entropy": 5.607731485366822,
      "epoch": 3.6633708272859216,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003704418423408718,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2201412469148636,
      "num_tokens": 83805673.0,
      "step": 40385
    },
    {
      "entropy": 5.591914367675781,
      "epoch": 3.663824383164006,
      "grad_norm": 1.375,
      "learning_rate": 0.0003704125401859104,
      "loss": 4.8923,
      "mean_token_accuracy": 0.21845129430294036,
      "num_tokens": 83815779.0,
      "step": 40390
    },
    {
      "entropy": 5.634136152267456,
      "epoch": 3.66427793904209,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037038323605779333,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2250106915831566,
      "num_tokens": 83826804.0,
      "step": 40395
    },
    {
      "entropy": 5.6084047794342045,
      "epoch": 3.664731494920174,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003703539299571266,
      "loss": 4.8932,
      "mean_token_accuracy": 0.21886310875415801,
      "num_tokens": 83836500.0,
      "step": 40400
    },
    {
      "entropy": 5.610776138305664,
      "epoch": 3.665185050798258,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003703246218845163,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22471242249011994,
      "num_tokens": 83846647.0,
      "step": 40405
    },
    {
      "entropy": 5.513819313049316,
      "epoch": 3.6656386066763424,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003702953118405686,
      "loss": 4.8431,
      "mean_token_accuracy": 0.22977578192949294,
      "num_tokens": 83858088.0,
      "step": 40410
    },
    {
      "entropy": 5.617538833618164,
      "epoch": 3.6660921625544267,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003702659998258895,
      "loss": 4.887,
      "mean_token_accuracy": 0.21979725807905198,
      "num_tokens": 83868936.0,
      "step": 40415
    },
    {
      "entropy": 5.683864593505859,
      "epoch": 3.666545718432511,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037023668584108523,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21251712441444398,
      "num_tokens": 83879813.0,
      "step": 40420
    },
    {
      "entropy": 5.568056678771972,
      "epoch": 3.6669992743105952,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037020736988676206,
      "loss": 4.8516,
      "mean_token_accuracy": 0.2191893130540848,
      "num_tokens": 83891619.0,
      "step": 40425
    },
    {
      "entropy": 5.643178462982178,
      "epoch": 3.6674528301886795,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037017805196352623,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2209170177578926,
      "num_tokens": 83901905.0,
      "step": 40430
    },
    {
      "entropy": 5.541421842575073,
      "epoch": 3.6679063860667633,
      "grad_norm": 1.3125,
      "learning_rate": 0.000370148732071984,
      "loss": 4.8154,
      "mean_token_accuracy": 0.22036763280630112,
      "num_tokens": 83912827.0,
      "step": 40435
    },
    {
      "entropy": 5.536515235900879,
      "epoch": 3.6683599419448476,
      "grad_norm": 1.25,
      "learning_rate": 0.0003701194102127418,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23038166910409927,
      "num_tokens": 83923719.0,
      "step": 40440
    },
    {
      "entropy": 5.499744510650634,
      "epoch": 3.668813497822932,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003700900863864059,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22733352929353715,
      "num_tokens": 83933715.0,
      "step": 40445
    },
    {
      "entropy": 5.632646131515503,
      "epoch": 3.6692670537010157,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003700607605935828,
      "loss": 4.9387,
      "mean_token_accuracy": 0.2187739446759224,
      "num_tokens": 83943946.0,
      "step": 40450
    },
    {
      "entropy": 5.627836227416992,
      "epoch": 3.6697206095791,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003700314328348789,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22024703472852708,
      "num_tokens": 83953414.0,
      "step": 40455
    },
    {
      "entropy": 5.579219913482666,
      "epoch": 3.6701741654571842,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003700021031109008,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22096900939941405,
      "num_tokens": 83963717.0,
      "step": 40460
    },
    {
      "entropy": 5.569381856918335,
      "epoch": 3.6706277213352685,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036997277142225506,
      "loss": 4.8739,
      "mean_token_accuracy": 0.2230770096182823,
      "num_tokens": 83973467.0,
      "step": 40465
    },
    {
      "entropy": 5.573579883575439,
      "epoch": 3.6710812772133528,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036994343776954806,
      "loss": 4.9128,
      "mean_token_accuracy": 0.21792598068714142,
      "num_tokens": 83984013.0,
      "step": 40470
    },
    {
      "entropy": 5.6065747261047365,
      "epoch": 3.671534833091437,
      "grad_norm": 1.25,
      "learning_rate": 0.0003699141021533867,
      "loss": 4.791,
      "mean_token_accuracy": 0.2270532637834549,
      "num_tokens": 83994349.0,
      "step": 40475
    },
    {
      "entropy": 5.579814481735229,
      "epoch": 3.6719883889695213,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003698847645743775,
      "loss": 4.7921,
      "mean_token_accuracy": 0.22049401104450225,
      "num_tokens": 84004525.0,
      "step": 40480
    },
    {
      "entropy": 5.567108869552612,
      "epoch": 3.672441944847605,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003698554250331271,
      "loss": 4.886,
      "mean_token_accuracy": 0.2230028837919235,
      "num_tokens": 84014286.0,
      "step": 40485
    },
    {
      "entropy": 5.548601055145264,
      "epoch": 3.6728955007256894,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003698260835302424,
      "loss": 4.8475,
      "mean_token_accuracy": 0.22993536442518234,
      "num_tokens": 84024826.0,
      "step": 40490
    },
    {
      "entropy": 5.613683223724365,
      "epoch": 3.6733490566037736,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003697967400663302,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21963182538747789,
      "num_tokens": 84035726.0,
      "step": 40495
    },
    {
      "entropy": 5.558452701568603,
      "epoch": 3.673802612481858,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036976739464199715,
      "loss": 4.7885,
      "mean_token_accuracy": 0.22456915080547332,
      "num_tokens": 84045573.0,
      "step": 40500
    },
    {
      "entropy": 5.625650644302368,
      "epoch": 3.6742561683599417,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036973804725785024,
      "loss": 4.8926,
      "mean_token_accuracy": 0.21655002236366272,
      "num_tokens": 84055983.0,
      "step": 40505
    },
    {
      "entropy": 5.670609855651856,
      "epoch": 3.674709724238026,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036970869791449644,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21881693750619888,
      "num_tokens": 84066905.0,
      "step": 40510
    },
    {
      "entropy": 5.586445426940918,
      "epoch": 3.6751632801161103,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036967934661254263,
      "loss": 4.7938,
      "mean_token_accuracy": 0.23213589936494827,
      "num_tokens": 84076861.0,
      "step": 40515
    },
    {
      "entropy": 5.621592712402344,
      "epoch": 3.6756168359941945,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003696499933525957,
      "loss": 4.8709,
      "mean_token_accuracy": 0.22316233217716216,
      "num_tokens": 84086167.0,
      "step": 40520
    },
    {
      "entropy": 5.5445457935333256,
      "epoch": 3.676070391872279,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036962063813526287,
      "loss": 4.86,
      "mean_token_accuracy": 0.22470114082098008,
      "num_tokens": 84095306.0,
      "step": 40525
    },
    {
      "entropy": 5.540723180770874,
      "epoch": 3.676523947750363,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003695912809611511,
      "loss": 4.885,
      "mean_token_accuracy": 0.21967659145593643,
      "num_tokens": 84106505.0,
      "step": 40530
    },
    {
      "entropy": 5.639750051498413,
      "epoch": 3.676977503628447,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003695619218308676,
      "loss": 4.8344,
      "mean_token_accuracy": 0.22541401386260987,
      "num_tokens": 84116393.0,
      "step": 40535
    },
    {
      "entropy": 5.6461762428283695,
      "epoch": 3.677431059506531,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036953256074501937,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22074188143014908,
      "num_tokens": 84125890.0,
      "step": 40540
    },
    {
      "entropy": 5.642495012283325,
      "epoch": 3.6778846153846154,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003695031977042138,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2238151878118515,
      "num_tokens": 84137736.0,
      "step": 40545
    },
    {
      "entropy": 5.588791608810425,
      "epoch": 3.6783381712626997,
      "grad_norm": 1.375,
      "learning_rate": 0.00036947383270905795,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22469405084848404,
      "num_tokens": 84148974.0,
      "step": 40550
    },
    {
      "entropy": 5.591322612762451,
      "epoch": 3.6787917271407835,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036944446576015914,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22222004234790801,
      "num_tokens": 84159041.0,
      "step": 40555
    },
    {
      "entropy": 5.5700091361999515,
      "epoch": 3.6792452830188678,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036941509685812476,
      "loss": 4.7992,
      "mean_token_accuracy": 0.22876355350017546,
      "num_tokens": 84169432.0,
      "step": 40560
    },
    {
      "entropy": 5.579265022277832,
      "epoch": 3.679698838896952,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036938572600356213,
      "loss": 4.8512,
      "mean_token_accuracy": 0.2202232375741005,
      "num_tokens": 84179964.0,
      "step": 40565
    },
    {
      "entropy": 5.5722352981567385,
      "epoch": 3.6801523947750363,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036935635319707856,
      "loss": 4.9168,
      "mean_token_accuracy": 0.21534696221351624,
      "num_tokens": 84189496.0,
      "step": 40570
    },
    {
      "entropy": 5.581364393234253,
      "epoch": 3.6806059506531206,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003693269784392816,
      "loss": 4.8492,
      "mean_token_accuracy": 0.23357266634702684,
      "num_tokens": 84200396.0,
      "step": 40575
    },
    {
      "entropy": 5.544690084457398,
      "epoch": 3.681059506531205,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00036929760173077863,
      "loss": 4.8182,
      "mean_token_accuracy": 0.22782830595970155,
      "num_tokens": 84210114.0,
      "step": 40580
    },
    {
      "entropy": 5.5797666072845455,
      "epoch": 3.6815130624092887,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003692682230721773,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2247671291232109,
      "num_tokens": 84220472.0,
      "step": 40585
    },
    {
      "entropy": 5.604016971588135,
      "epoch": 3.681966618287373,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003692388424640851,
      "loss": 4.9013,
      "mean_token_accuracy": 0.21643133610486984,
      "num_tokens": 84230718.0,
      "step": 40590
    },
    {
      "entropy": 5.6349588394165036,
      "epoch": 3.682420174165457,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036920945990710964,
      "loss": 4.8863,
      "mean_token_accuracy": 0.22070770114660263,
      "num_tokens": 84241736.0,
      "step": 40595
    },
    {
      "entropy": 5.527383279800415,
      "epoch": 3.6828737300435415,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036918007540185846,
      "loss": 4.8109,
      "mean_token_accuracy": 0.22870029509067535,
      "num_tokens": 84251770.0,
      "step": 40600
    },
    {
      "entropy": 5.584906721115113,
      "epoch": 3.6833272859216253,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003691506889489394,
      "loss": 4.9194,
      "mean_token_accuracy": 0.21823448687791824,
      "num_tokens": 84263355.0,
      "step": 40605
    },
    {
      "entropy": 5.59528341293335,
      "epoch": 3.6837808417997095,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036912130054896005,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21586682349443437,
      "num_tokens": 84274835.0,
      "step": 40610
    },
    {
      "entropy": 5.686047029495239,
      "epoch": 3.684234397677794,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036909191020252825,
      "loss": 4.99,
      "mean_token_accuracy": 0.20589088499546052,
      "num_tokens": 84286099.0,
      "step": 40615
    },
    {
      "entropy": 5.654107522964478,
      "epoch": 3.684687953555878,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003690625179102517,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22320015281438826,
      "num_tokens": 84296168.0,
      "step": 40620
    },
    {
      "entropy": 5.69366192817688,
      "epoch": 3.6851415094339623,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003690331236727384,
      "loss": 4.9978,
      "mean_token_accuracy": 0.21234170496463775,
      "num_tokens": 84307352.0,
      "step": 40625
    },
    {
      "entropy": 5.598083782196045,
      "epoch": 3.6855950653120466,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003690037274905961,
      "loss": 4.9229,
      "mean_token_accuracy": 0.21975927352905272,
      "num_tokens": 84318496.0,
      "step": 40630
    },
    {
      "entropy": 5.544028568267822,
      "epoch": 3.686048621190131,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003689743293644327,
      "loss": 4.8453,
      "mean_token_accuracy": 0.22161079794168473,
      "num_tokens": 84329101.0,
      "step": 40635
    },
    {
      "entropy": 5.554684019088745,
      "epoch": 3.6865021770682147,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003689449292948563,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22812528014183045,
      "num_tokens": 84339143.0,
      "step": 40640
    },
    {
      "entropy": 5.654756116867065,
      "epoch": 3.686955732946299,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036891552728247483,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22400348335504533,
      "num_tokens": 84349693.0,
      "step": 40645
    },
    {
      "entropy": 5.63770751953125,
      "epoch": 3.6874092888243832,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036888612332789636,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21775434613227845,
      "num_tokens": 84359799.0,
      "step": 40650
    },
    {
      "entropy": 5.5892339706420895,
      "epoch": 3.6878628447024675,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036885671743172875,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22512342631816865,
      "num_tokens": 84370695.0,
      "step": 40655
    },
    {
      "entropy": 5.6030360698699955,
      "epoch": 3.6883164005805513,
      "grad_norm": 1.25,
      "learning_rate": 0.0003688273095945805,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21315661668777466,
      "num_tokens": 84380205.0,
      "step": 40660
    },
    {
      "entropy": 5.648444652557373,
      "epoch": 3.6887699564586356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036879789981705954,
      "loss": 4.884,
      "mean_token_accuracy": 0.2327017068862915,
      "num_tokens": 84391034.0,
      "step": 40665
    },
    {
      "entropy": 5.663483715057373,
      "epoch": 3.68922351233672,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036876848809977404,
      "loss": 4.9597,
      "mean_token_accuracy": 0.2124026194214821,
      "num_tokens": 84401452.0,
      "step": 40670
    },
    {
      "entropy": 5.553037786483765,
      "epoch": 3.689677068214804,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036873907444333234,
      "loss": 4.8233,
      "mean_token_accuracy": 0.2222222939133644,
      "num_tokens": 84411126.0,
      "step": 40675
    },
    {
      "entropy": 5.577631330490112,
      "epoch": 3.6901306240928884,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036870965884834274,
      "loss": 4.876,
      "mean_token_accuracy": 0.22277166098356246,
      "num_tokens": 84422740.0,
      "step": 40680
    },
    {
      "entropy": 5.591572284698486,
      "epoch": 3.6905841799709727,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00036868024131541337,
      "loss": 4.908,
      "mean_token_accuracy": 0.2163054034113884,
      "num_tokens": 84432831.0,
      "step": 40685
    },
    {
      "entropy": 5.65256814956665,
      "epoch": 3.6910377358490565,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036865082184515296,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22705880254507066,
      "num_tokens": 84443176.0,
      "step": 40690
    },
    {
      "entropy": 5.660401153564453,
      "epoch": 3.6914912917271407,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036862140043816957,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21292325109243393,
      "num_tokens": 84453185.0,
      "step": 40695
    },
    {
      "entropy": 5.6245016098022464,
      "epoch": 3.691944847605225,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003685919770950718,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22527073919773102,
      "num_tokens": 84464303.0,
      "step": 40700
    },
    {
      "entropy": 5.615199756622315,
      "epoch": 3.6923984034833093,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036856255181646797,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21953359246253967,
      "num_tokens": 84474463.0,
      "step": 40705
    },
    {
      "entropy": 5.5809777736663815,
      "epoch": 3.692851959361393,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003685331246029668,
      "loss": 4.8759,
      "mean_token_accuracy": 0.21850893944501876,
      "num_tokens": 84484682.0,
      "step": 40710
    },
    {
      "entropy": 5.6068253993988035,
      "epoch": 3.6933055152394774,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036850369545517685,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22223106324672698,
      "num_tokens": 84494830.0,
      "step": 40715
    },
    {
      "entropy": 5.557840204238891,
      "epoch": 3.6937590711175616,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003684742643737065,
      "loss": 4.8395,
      "mean_token_accuracy": 0.22403172701597213,
      "num_tokens": 84505449.0,
      "step": 40720
    },
    {
      "entropy": 5.653700113296509,
      "epoch": 3.694212626995646,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036844483135916467,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21957908123731612,
      "num_tokens": 84515833.0,
      "step": 40725
    },
    {
      "entropy": 5.6320253849029545,
      "epoch": 3.69466618287373,
      "grad_norm": 1.25,
      "learning_rate": 0.0003684153964121599,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21804351806640626,
      "num_tokens": 84526677.0,
      "step": 40730
    },
    {
      "entropy": 5.591643905639648,
      "epoch": 3.6951197387518144,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036838595953330095,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22896821796894073,
      "num_tokens": 84536656.0,
      "step": 40735
    },
    {
      "entropy": 5.6335373878479,
      "epoch": 3.6955732946298983,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036835652072319646,
      "loss": 4.8216,
      "mean_token_accuracy": 0.22985567897558212,
      "num_tokens": 84546506.0,
      "step": 40740
    },
    {
      "entropy": 5.633129644393921,
      "epoch": 3.6960268505079825,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003683270799824554,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2214898645877838,
      "num_tokens": 84556921.0,
      "step": 40745
    },
    {
      "entropy": 5.63296070098877,
      "epoch": 3.696480406386067,
      "grad_norm": 1.25,
      "learning_rate": 0.0003682976373116865,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2270765036344528,
      "num_tokens": 84566325.0,
      "step": 40750
    },
    {
      "entropy": 5.5508081912994385,
      "epoch": 3.696933962264151,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036826819271149877,
      "loss": 4.8423,
      "mean_token_accuracy": 0.21793505251407624,
      "num_tokens": 84577167.0,
      "step": 40755
    },
    {
      "entropy": 5.5803662776947025,
      "epoch": 3.697387518142235,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00036823874618250105,
      "loss": 4.8549,
      "mean_token_accuracy": 0.22092701345682145,
      "num_tokens": 84586764.0,
      "step": 40760
    },
    {
      "entropy": 5.549526786804199,
      "epoch": 3.697841074020319,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036820929772530224,
      "loss": 4.8193,
      "mean_token_accuracy": 0.22408065795898438,
      "num_tokens": 84598903.0,
      "step": 40765
    },
    {
      "entropy": 5.548995923995972,
      "epoch": 3.6982946298984034,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003681798473405114,
      "loss": 4.8251,
      "mean_token_accuracy": 0.2342318758368492,
      "num_tokens": 84608695.0,
      "step": 40770
    },
    {
      "entropy": 5.534334802627564,
      "epoch": 3.6987481857764877,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036815039502873756,
      "loss": 4.8386,
      "mean_token_accuracy": 0.22162525057792665,
      "num_tokens": 84619377.0,
      "step": 40775
    },
    {
      "entropy": 5.542198753356933,
      "epoch": 3.699201741654572,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036812094079058985,
      "loss": 4.8288,
      "mean_token_accuracy": 0.2200347140431404,
      "num_tokens": 84629716.0,
      "step": 40780
    },
    {
      "entropy": 5.508095455169678,
      "epoch": 3.699655297532656,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003680914846266773,
      "loss": 4.7496,
      "mean_token_accuracy": 0.2336232766509056,
      "num_tokens": 84640214.0,
      "step": 40785
    },
    {
      "entropy": 5.599783134460449,
      "epoch": 3.7001088534107405,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003680620265376092,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22074273973703384,
      "num_tokens": 84650403.0,
      "step": 40790
    },
    {
      "entropy": 5.627480411529541,
      "epoch": 3.7005624092888243,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036803256652399456,
      "loss": 4.8247,
      "mean_token_accuracy": 0.22606240063905716,
      "num_tokens": 84660912.0,
      "step": 40795
    },
    {
      "entropy": 5.625306177139282,
      "epoch": 3.7010159651669086,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036800310458644285,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21461918205022812,
      "num_tokens": 84672194.0,
      "step": 40800
    },
    {
      "entropy": 5.614631938934326,
      "epoch": 3.701469521044993,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036797364072556316,
      "loss": 4.9908,
      "mean_token_accuracy": 0.2116982087492943,
      "num_tokens": 84681850.0,
      "step": 40805
    },
    {
      "entropy": 5.65272536277771,
      "epoch": 3.7019230769230766,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000367944174941965,
      "loss": 4.8874,
      "mean_token_accuracy": 0.2261722967028618,
      "num_tokens": 84691271.0,
      "step": 40810
    },
    {
      "entropy": 5.553043413162231,
      "epoch": 3.702376632801161,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036791470723625746,
      "loss": 4.8746,
      "mean_token_accuracy": 0.2184649884700775,
      "num_tokens": 84702189.0,
      "step": 40815
    },
    {
      "entropy": 5.589780235290528,
      "epoch": 3.702830188679245,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003678852376090502,
      "loss": 4.7905,
      "mean_token_accuracy": 0.23754313141107558,
      "num_tokens": 84712399.0,
      "step": 40820
    },
    {
      "entropy": 5.629763269424439,
      "epoch": 3.7032837445573294,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003678557660609525,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21353862136602403,
      "num_tokens": 84723174.0,
      "step": 40825
    },
    {
      "entropy": 5.609356689453125,
      "epoch": 3.7037373004354137,
      "grad_norm": 1.265625,
      "learning_rate": 0.000367826292592574,
      "loss": 4.8693,
      "mean_token_accuracy": 0.23255186676979064,
      "num_tokens": 84732498.0,
      "step": 40830
    },
    {
      "entropy": 5.632840919494629,
      "epoch": 3.704190856313498,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003677968172045241,
      "loss": 4.905,
      "mean_token_accuracy": 0.21941236555576324,
      "num_tokens": 84742362.0,
      "step": 40835
    },
    {
      "entropy": 5.547799777984619,
      "epoch": 3.7046444121915822,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003677673398974122,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22006913870573044,
      "num_tokens": 84753379.0,
      "step": 40840
    },
    {
      "entropy": 5.632326984405518,
      "epoch": 3.705097968069666,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003677378606718482,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21458609998226166,
      "num_tokens": 84763245.0,
      "step": 40845
    },
    {
      "entropy": 5.62682991027832,
      "epoch": 3.7055515239477503,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003677083795284416,
      "loss": 4.9174,
      "mean_token_accuracy": 0.2148534044623375,
      "num_tokens": 84773929.0,
      "step": 40850
    },
    {
      "entropy": 5.661520957946777,
      "epoch": 3.7060050798258346,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036767889646780204,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21308955252170564,
      "num_tokens": 84784733.0,
      "step": 40855
    },
    {
      "entropy": 5.595984411239624,
      "epoch": 3.706458635703919,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003676494114905394,
      "loss": 4.8916,
      "mean_token_accuracy": 0.21934501230716705,
      "num_tokens": 84794254.0,
      "step": 40860
    },
    {
      "entropy": 5.61206283569336,
      "epoch": 3.7069121915820027,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036761992459726325,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21068903356790541,
      "num_tokens": 84804361.0,
      "step": 40865
    },
    {
      "entropy": 5.622202396392822,
      "epoch": 3.707365747460087,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036759043578858344,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22559751868247985,
      "num_tokens": 84814497.0,
      "step": 40870
    },
    {
      "entropy": 5.615522575378418,
      "epoch": 3.7078193033381712,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036756094506510984,
      "loss": 4.8614,
      "mean_token_accuracy": 0.22450143992900848,
      "num_tokens": 84826871.0,
      "step": 40875
    },
    {
      "entropy": 5.606543302536011,
      "epoch": 3.7082728592162555,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036753145242745226,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22307068556547166,
      "num_tokens": 84837053.0,
      "step": 40880
    },
    {
      "entropy": 5.5344452381134035,
      "epoch": 3.7087264150943398,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003675019578762208,
      "loss": 4.8277,
      "mean_token_accuracy": 0.229698546230793,
      "num_tokens": 84846681.0,
      "step": 40885
    },
    {
      "entropy": 5.547759103775024,
      "epoch": 3.709179970972424,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003674724614120251,
      "loss": 4.8033,
      "mean_token_accuracy": 0.22371743470430375,
      "num_tokens": 84856651.0,
      "step": 40890
    },
    {
      "entropy": 5.582463359832763,
      "epoch": 3.709633526850508,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036744296303547547,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21309611052274705,
      "num_tokens": 84867531.0,
      "step": 40895
    },
    {
      "entropy": 5.544451951980591,
      "epoch": 3.710087082728592,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003674134627471818,
      "loss": 4.8513,
      "mean_token_accuracy": 0.228191015124321,
      "num_tokens": 84879132.0,
      "step": 40900
    },
    {
      "entropy": 5.617182207107544,
      "epoch": 3.7105406386066764,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000367383960547754,
      "loss": 4.8898,
      "mean_token_accuracy": 0.2310369625687599,
      "num_tokens": 84888873.0,
      "step": 40905
    },
    {
      "entropy": 5.610183572769165,
      "epoch": 3.7109941944847606,
      "grad_norm": 1.375,
      "learning_rate": 0.00036735445643780256,
      "loss": 4.912,
      "mean_token_accuracy": 0.21800245046615602,
      "num_tokens": 84898926.0,
      "step": 40910
    },
    {
      "entropy": 5.5954345703125,
      "epoch": 3.7114477503628445,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003673249504179374,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22153220772743226,
      "num_tokens": 84908502.0,
      "step": 40915
    },
    {
      "entropy": 5.597489547729492,
      "epoch": 3.7119013062409287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003672954424887687,
      "loss": 4.8927,
      "mean_token_accuracy": 0.21910693496465683,
      "num_tokens": 84919681.0,
      "step": 40920
    },
    {
      "entropy": 5.509079360961914,
      "epoch": 3.712354862119013,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003672659326509066,
      "loss": 4.7974,
      "mean_token_accuracy": 0.22543679177761078,
      "num_tokens": 84930149.0,
      "step": 40925
    },
    {
      "entropy": 5.541243362426758,
      "epoch": 3.7128084179970973,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036723642090496166,
      "loss": 4.851,
      "mean_token_accuracy": 0.221881964802742,
      "num_tokens": 84940086.0,
      "step": 40930
    },
    {
      "entropy": 5.631255722045898,
      "epoch": 3.7132619738751815,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036720690725154404,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22272559404373168,
      "num_tokens": 84949248.0,
      "step": 40935
    },
    {
      "entropy": 5.568168640136719,
      "epoch": 3.713715529753266,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036717739169126395,
      "loss": 4.8365,
      "mean_token_accuracy": 0.2264125943183899,
      "num_tokens": 84960892.0,
      "step": 40940
    },
    {
      "entropy": 5.5828814029693605,
      "epoch": 3.7141690856313496,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036714787422473206,
      "loss": 4.871,
      "mean_token_accuracy": 0.21667643636465073,
      "num_tokens": 84971024.0,
      "step": 40945
    },
    {
      "entropy": 5.526698207855224,
      "epoch": 3.714622641509434,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003671183548525586,
      "loss": 4.7954,
      "mean_token_accuracy": 0.2248763158917427,
      "num_tokens": 84980945.0,
      "step": 40950
    },
    {
      "entropy": 5.60327000617981,
      "epoch": 3.715076197387518,
      "grad_norm": 1.234375,
      "learning_rate": 0.000367088833575354,
      "loss": 4.9402,
      "mean_token_accuracy": 0.21621231585741044,
      "num_tokens": 84992055.0,
      "step": 40955
    },
    {
      "entropy": 5.602426671981812,
      "epoch": 3.7155297532656024,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036705931039372904,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22144952267408372,
      "num_tokens": 85002619.0,
      "step": 40960
    },
    {
      "entropy": 5.6058001041412355,
      "epoch": 3.7159833091436862,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003670297853082939,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22271185368299484,
      "num_tokens": 85013317.0,
      "step": 40965
    },
    {
      "entropy": 5.60260238647461,
      "epoch": 3.7164368650217705,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036700025831965937,
      "loss": 4.8557,
      "mean_token_accuracy": 0.22290650755167007,
      "num_tokens": 85023457.0,
      "step": 40970
    },
    {
      "entropy": 5.591012859344483,
      "epoch": 3.7168904208998548,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036697072942843616,
      "loss": 4.8519,
      "mean_token_accuracy": 0.21889072507619858,
      "num_tokens": 85033498.0,
      "step": 40975
    },
    {
      "entropy": 5.718241167068482,
      "epoch": 3.717343976777939,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003669411986352348,
      "loss": 5.0748,
      "mean_token_accuracy": 0.21237341016530992,
      "num_tokens": 85043687.0,
      "step": 40980
    },
    {
      "entropy": 5.645121812820435,
      "epoch": 3.7177975326560233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000366911665940666,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22174159288406373,
      "num_tokens": 85055045.0,
      "step": 40985
    },
    {
      "entropy": 5.598646831512451,
      "epoch": 3.7182510885341076,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036688213134534044,
      "loss": 4.812,
      "mean_token_accuracy": 0.23610011339187623,
      "num_tokens": 85064625.0,
      "step": 40990
    },
    {
      "entropy": 5.700275182723999,
      "epoch": 3.718704644412192,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003668525948498691,
      "loss": 5.0392,
      "mean_token_accuracy": 0.2058858245611191,
      "num_tokens": 85075249.0,
      "step": 40995
    },
    {
      "entropy": 5.59577374458313,
      "epoch": 3.7191582002902757,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036682305645486266,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22311525642871857,
      "num_tokens": 85085900.0,
      "step": 41000
    },
    {
      "entropy": 5.603970909118653,
      "epoch": 3.71961175616836,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036679351616093194,
      "loss": 4.9202,
      "mean_token_accuracy": 0.22257013618946075,
      "num_tokens": 85095772.0,
      "step": 41005
    },
    {
      "entropy": 5.557195234298706,
      "epoch": 3.720065312046444,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036676397396868797,
      "loss": 4.8347,
      "mean_token_accuracy": 0.22497595250606536,
      "num_tokens": 85105653.0,
      "step": 41010
    },
    {
      "entropy": 5.6220947265625,
      "epoch": 3.7205188679245285,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003667344298787417,
      "loss": 4.9634,
      "mean_token_accuracy": 0.21566734611988067,
      "num_tokens": 85115432.0,
      "step": 41015
    },
    {
      "entropy": 5.69744758605957,
      "epoch": 3.7209724238026123,
      "grad_norm": 1.25,
      "learning_rate": 0.00036670488389170395,
      "loss": 5.0565,
      "mean_token_accuracy": 0.20855958312749862,
      "num_tokens": 85126572.0,
      "step": 41020
    },
    {
      "entropy": 5.544163751602173,
      "epoch": 3.7214259796806965,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036667533600818586,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22635802775621414,
      "num_tokens": 85136809.0,
      "step": 41025
    },
    {
      "entropy": 5.763647508621216,
      "epoch": 3.721879535558781,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003666457862287985,
      "loss": 5.0487,
      "mean_token_accuracy": 0.20916630029678346,
      "num_tokens": 85148346.0,
      "step": 41030
    },
    {
      "entropy": 5.5963341236114506,
      "epoch": 3.722333091436865,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036661623455415296,
      "loss": 4.8949,
      "mean_token_accuracy": 0.21946420073509215,
      "num_tokens": 85158807.0,
      "step": 41035
    },
    {
      "entropy": 5.659514570236206,
      "epoch": 3.7227866473149493,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003665866809848602,
      "loss": 4.9871,
      "mean_token_accuracy": 0.20989061743021012,
      "num_tokens": 85170752.0,
      "step": 41040
    },
    {
      "entropy": 5.614282131195068,
      "epoch": 3.7232402031930336,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036655712552153164,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22241588532924653,
      "num_tokens": 85180779.0,
      "step": 41045
    },
    {
      "entropy": 5.623584985733032,
      "epoch": 3.7236937590711174,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003665275681647784,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22459035813808442,
      "num_tokens": 85191391.0,
      "step": 41050
    },
    {
      "entropy": 5.610738611221313,
      "epoch": 3.7241473149492017,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003664980089152116,
      "loss": 4.863,
      "mean_token_accuracy": 0.22108899652957917,
      "num_tokens": 85201357.0,
      "step": 41055
    },
    {
      "entropy": 5.587913513183594,
      "epoch": 3.724600870827286,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036646844777344284,
      "loss": 4.8902,
      "mean_token_accuracy": 0.22266859263181688,
      "num_tokens": 85212193.0,
      "step": 41060
    },
    {
      "entropy": 5.615016460418701,
      "epoch": 3.7250544267053702,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036643888474008313,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21867918223142624,
      "num_tokens": 85223809.0,
      "step": 41065
    },
    {
      "entropy": 5.601362895965576,
      "epoch": 3.725507982583454,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003664093198157441,
      "loss": 4.88,
      "mean_token_accuracy": 0.22628283500671387,
      "num_tokens": 85234992.0,
      "step": 41070
    },
    {
      "entropy": 5.660713815689087,
      "epoch": 3.7259615384615383,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003663797530010369,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21391869336366653,
      "num_tokens": 85244157.0,
      "step": 41075
    },
    {
      "entropy": 5.583652782440185,
      "epoch": 3.7264150943396226,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003663501842965732,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22759334444999696,
      "num_tokens": 85255946.0,
      "step": 41080
    },
    {
      "entropy": 5.556160831451416,
      "epoch": 3.726868650217707,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003663206137029645,
      "loss": 4.8473,
      "mean_token_accuracy": 0.22499336004257203,
      "num_tokens": 85265273.0,
      "step": 41085
    },
    {
      "entropy": 5.48696665763855,
      "epoch": 3.727322206095791,
      "grad_norm": 1.25,
      "learning_rate": 0.00036629104122082215,
      "loss": 4.7173,
      "mean_token_accuracy": 0.2295747622847557,
      "num_tokens": 85275794.0,
      "step": 41090
    },
    {
      "entropy": 5.577328300476074,
      "epoch": 3.7277757619738754,
      "grad_norm": 1.25,
      "learning_rate": 0.00036626146685075775,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2084342882037163,
      "num_tokens": 85287334.0,
      "step": 41095
    },
    {
      "entropy": 5.612257575988769,
      "epoch": 3.728229317851959,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000366231890593383,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22396960407495498,
      "num_tokens": 85297320.0,
      "step": 41100
    },
    {
      "entropy": 5.553638029098511,
      "epoch": 3.7286828737300435,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003662023124493095,
      "loss": 4.7923,
      "mean_token_accuracy": 0.22827279567718506,
      "num_tokens": 85308071.0,
      "step": 41105
    },
    {
      "entropy": 5.593637084960937,
      "epoch": 3.7291364296081277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000366172732419149,
      "loss": 4.908,
      "mean_token_accuracy": 0.21741290986537934,
      "num_tokens": 85317790.0,
      "step": 41110
    },
    {
      "entropy": 5.621043968200683,
      "epoch": 3.729589985486212,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036614315050351317,
      "loss": 4.9721,
      "mean_token_accuracy": 0.21697428822517395,
      "num_tokens": 85329008.0,
      "step": 41115
    },
    {
      "entropy": 5.719133090972901,
      "epoch": 3.730043541364296,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003661135667030137,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21347654163837432,
      "num_tokens": 85339095.0,
      "step": 41120
    },
    {
      "entropy": 5.656979417800903,
      "epoch": 3.73049709724238,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003660839810182624,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21583237648010253,
      "num_tokens": 85349833.0,
      "step": 41125
    },
    {
      "entropy": 5.642375040054321,
      "epoch": 3.7309506531204644,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036605439344987125,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2127848371863365,
      "num_tokens": 85359512.0,
      "step": 41130
    },
    {
      "entropy": 5.5954430103302,
      "epoch": 3.7314042089985486,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000366024803998452,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22528755962848662,
      "num_tokens": 85370053.0,
      "step": 41135
    },
    {
      "entropy": 5.6010528087615965,
      "epoch": 3.731857764876633,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036599521266461653,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2228283017873764,
      "num_tokens": 85380020.0,
      "step": 41140
    },
    {
      "entropy": 5.624449300765991,
      "epoch": 3.732311320754717,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036596561944897703,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22425800412893296,
      "num_tokens": 85390232.0,
      "step": 41145
    },
    {
      "entropy": 5.536047077178955,
      "epoch": 3.7327648766328014,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003659360243521453,
      "loss": 4.8097,
      "mean_token_accuracy": 0.2284854993224144,
      "num_tokens": 85400401.0,
      "step": 41150
    },
    {
      "entropy": 5.606580114364624,
      "epoch": 3.7332184325108853,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003659064273747334,
      "loss": 4.917,
      "mean_token_accuracy": 0.21334304958581923,
      "num_tokens": 85411749.0,
      "step": 41155
    },
    {
      "entropy": 5.663390398025513,
      "epoch": 3.7336719883889695,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036587682851735336,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21453898549079894,
      "num_tokens": 85424065.0,
      "step": 41160
    },
    {
      "entropy": 5.632834243774414,
      "epoch": 3.734125544267054,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036584722778061743,
      "loss": 4.9147,
      "mean_token_accuracy": 0.21882192492485047,
      "num_tokens": 85434412.0,
      "step": 41165
    },
    {
      "entropy": 5.617663812637329,
      "epoch": 3.7345791001451376,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003658176251651376,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22405829578638076,
      "num_tokens": 85444489.0,
      "step": 41170
    },
    {
      "entropy": 5.638788795471191,
      "epoch": 3.735032656023222,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036578802067152615,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22335279434919358,
      "num_tokens": 85454556.0,
      "step": 41175
    },
    {
      "entropy": 5.5778968334198,
      "epoch": 3.735486211901306,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003657584143003954,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22520850449800492,
      "num_tokens": 85464701.0,
      "step": 41180
    },
    {
      "entropy": 5.616639232635498,
      "epoch": 3.7359397677793904,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003657288060523574,
      "loss": 4.9437,
      "mean_token_accuracy": 0.22143598794937133,
      "num_tokens": 85475657.0,
      "step": 41185
    },
    {
      "entropy": 5.609533166885376,
      "epoch": 3.7363933236574747,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036569919592802466,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22944324612617492,
      "num_tokens": 85486358.0,
      "step": 41190
    },
    {
      "entropy": 5.630937242507935,
      "epoch": 3.736846879535559,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003656695839280093,
      "loss": 4.8928,
      "mean_token_accuracy": 0.2160073608160019,
      "num_tokens": 85497913.0,
      "step": 41195
    },
    {
      "entropy": 5.606608200073242,
      "epoch": 3.737300435413643,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00036563997005292396,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22670585811138153,
      "num_tokens": 85508314.0,
      "step": 41200
    },
    {
      "entropy": 5.583343982696533,
      "epoch": 3.737753991291727,
      "grad_norm": 1.25,
      "learning_rate": 0.00036561035430338084,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2173798844218254,
      "num_tokens": 85518142.0,
      "step": 41205
    },
    {
      "entropy": 5.590759515762329,
      "epoch": 3.7382075471698113,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036558073667999257,
      "loss": 4.7717,
      "mean_token_accuracy": 0.22751502841711044,
      "num_tokens": 85526854.0,
      "step": 41210
    },
    {
      "entropy": 5.52944507598877,
      "epoch": 3.7386611030478956,
      "grad_norm": 1.25,
      "learning_rate": 0.00036555111718337166,
      "loss": 4.7815,
      "mean_token_accuracy": 0.231533545255661,
      "num_tokens": 85537153.0,
      "step": 41215
    },
    {
      "entropy": 5.643707227706909,
      "epoch": 3.73911465892598,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003655214958141305,
      "loss": 4.9817,
      "mean_token_accuracy": 0.20915819555521012,
      "num_tokens": 85548583.0,
      "step": 41220
    },
    {
      "entropy": 5.635319185256958,
      "epoch": 3.7395682148040637,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036549187257288164,
      "loss": 4.9336,
      "mean_token_accuracy": 0.23079233914613723,
      "num_tokens": 85559228.0,
      "step": 41225
    },
    {
      "entropy": 5.664715909957886,
      "epoch": 3.740021770682148,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036546224746023795,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22583277374505997,
      "num_tokens": 85569757.0,
      "step": 41230
    },
    {
      "entropy": 5.582386016845703,
      "epoch": 3.740475326560232,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036543262047681186,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22723791897296905,
      "num_tokens": 85582013.0,
      "step": 41235
    },
    {
      "entropy": 5.600093841552734,
      "epoch": 3.7409288824383164,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003654029916232161,
      "loss": 4.8691,
      "mean_token_accuracy": 0.2359455034136772,
      "num_tokens": 85594049.0,
      "step": 41240
    },
    {
      "entropy": 5.612578725814819,
      "epoch": 3.7413824383164007,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00036537336090006347,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2179932862520218,
      "num_tokens": 85605253.0,
      "step": 41245
    },
    {
      "entropy": 5.597386074066162,
      "epoch": 3.741835994194485,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036534372830796674,
      "loss": 4.8405,
      "mean_token_accuracy": 0.22601506114006042,
      "num_tokens": 85615032.0,
      "step": 41250
    },
    {
      "entropy": 5.59451150894165,
      "epoch": 3.742289550072569,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003653140938475386,
      "loss": 4.868,
      "mean_token_accuracy": 0.2282341718673706,
      "num_tokens": 85625497.0,
      "step": 41255
    },
    {
      "entropy": 5.621959257125854,
      "epoch": 3.742743105950653,
      "grad_norm": 1.125,
      "learning_rate": 0.00036528445751939206,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22247463017702102,
      "num_tokens": 85636058.0,
      "step": 41260
    },
    {
      "entropy": 5.5920909404754635,
      "epoch": 3.7431966618287373,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036525481932413997,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22008536010980606,
      "num_tokens": 85646505.0,
      "step": 41265
    },
    {
      "entropy": 5.6329492092132565,
      "epoch": 3.7436502177068216,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036522517926239513,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2221252754330635,
      "num_tokens": 85656864.0,
      "step": 41270
    },
    {
      "entropy": 5.547658443450928,
      "epoch": 3.7441037735849054,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036519553733477055,
      "loss": 4.8174,
      "mean_token_accuracy": 0.2276010200381279,
      "num_tokens": 85666783.0,
      "step": 41275
    },
    {
      "entropy": 5.49616322517395,
      "epoch": 3.7445573294629897,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003651658935418794,
      "loss": 4.7677,
      "mean_token_accuracy": 0.2311669573187828,
      "num_tokens": 85677102.0,
      "step": 41280
    },
    {
      "entropy": 5.536429309844971,
      "epoch": 3.745010885341074,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036513624788433454,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23184246569871902,
      "num_tokens": 85687048.0,
      "step": 41285
    },
    {
      "entropy": 5.565611505508423,
      "epoch": 3.7454644412191582,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003651066003627491,
      "loss": 4.8782,
      "mean_token_accuracy": 0.21524805575609207,
      "num_tokens": 85698314.0,
      "step": 41290
    },
    {
      "entropy": 5.6669567108154295,
      "epoch": 3.7459179970972425,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003650769509777362,
      "loss": 4.945,
      "mean_token_accuracy": 0.2164806380867958,
      "num_tokens": 85708039.0,
      "step": 41295
    },
    {
      "entropy": 5.601735973358155,
      "epoch": 3.7463715529753268,
      "grad_norm": 1.5,
      "learning_rate": 0.00036504729972990904,
      "loss": 4.8028,
      "mean_token_accuracy": 0.23303242027759552,
      "num_tokens": 85717741.0,
      "step": 41300
    },
    {
      "entropy": 5.613136911392212,
      "epoch": 3.7468251088534106,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003650176466198807,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2254712015390396,
      "num_tokens": 85728286.0,
      "step": 41305
    },
    {
      "entropy": 5.569131183624267,
      "epoch": 3.747278664731495,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036498799164826453,
      "loss": 4.863,
      "mean_token_accuracy": 0.22300679981708527,
      "num_tokens": 85738970.0,
      "step": 41310
    },
    {
      "entropy": 5.543033075332642,
      "epoch": 3.747732220609579,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036495833481567375,
      "loss": 4.8252,
      "mean_token_accuracy": 0.22323786318302155,
      "num_tokens": 85749284.0,
      "step": 41315
    },
    {
      "entropy": 5.646782684326172,
      "epoch": 3.7481857764876634,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036492867612272176,
      "loss": 4.929,
      "mean_token_accuracy": 0.21498074531555175,
      "num_tokens": 85759595.0,
      "step": 41320
    },
    {
      "entropy": 5.688265657424926,
      "epoch": 3.748639332365747,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003648990155700217,
      "loss": 4.976,
      "mean_token_accuracy": 0.21302862763404845,
      "num_tokens": 85771049.0,
      "step": 41325
    },
    {
      "entropy": 5.629099130630493,
      "epoch": 3.7490928882438315,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036486935315818713,
      "loss": 4.8803,
      "mean_token_accuracy": 0.21610768288373947,
      "num_tokens": 85781656.0,
      "step": 41330
    },
    {
      "entropy": 5.5362084865570065,
      "epoch": 3.7495464441219157,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036483968888783154,
      "loss": 4.7499,
      "mean_token_accuracy": 0.22503260970115663,
      "num_tokens": 85790922.0,
      "step": 41335
    },
    {
      "entropy": 5.564342403411866,
      "epoch": 3.75,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036481002275956813,
      "loss": 4.8265,
      "mean_token_accuracy": 0.22990502566099166,
      "num_tokens": 85799928.0,
      "step": 41340
    },
    {
      "entropy": 5.589816522598267,
      "epoch": 3.7504535558780843,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003647803547740106,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2163300946354866,
      "num_tokens": 85810594.0,
      "step": 41345
    },
    {
      "entropy": 5.574264621734619,
      "epoch": 3.7509071117561685,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036475068493177254,
      "loss": 4.8984,
      "mean_token_accuracy": 0.21630749255418777,
      "num_tokens": 85820690.0,
      "step": 41350
    },
    {
      "entropy": 5.577137517929077,
      "epoch": 3.751360667634253,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003647210132334674,
      "loss": 4.8409,
      "mean_token_accuracy": 0.22375607639551162,
      "num_tokens": 85830314.0,
      "step": 41355
    },
    {
      "entropy": 5.5901806354522705,
      "epoch": 3.7518142235123366,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003646913396797088,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2082374632358551,
      "num_tokens": 85841099.0,
      "step": 41360
    },
    {
      "entropy": 5.651024913787841,
      "epoch": 3.752267779390421,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036466166427111046,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22038891017436982,
      "num_tokens": 85852813.0,
      "step": 41365
    },
    {
      "entropy": 5.607216024398804,
      "epoch": 3.752721335268505,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036463198700828616,
      "loss": 4.8257,
      "mean_token_accuracy": 0.22262684255838394,
      "num_tokens": 85862778.0,
      "step": 41370
    },
    {
      "entropy": 5.571303033828736,
      "epoch": 3.7531748911465894,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003646023078918494,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21781998723745347,
      "num_tokens": 85872729.0,
      "step": 41375
    },
    {
      "entropy": 5.60386028289795,
      "epoch": 3.7536284470246732,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003645726269224141,
      "loss": 4.8656,
      "mean_token_accuracy": 0.22781819850206375,
      "num_tokens": 85883289.0,
      "step": 41380
    },
    {
      "entropy": 5.535169172286987,
      "epoch": 3.7540820029027575,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036454294410059407,
      "loss": 4.8025,
      "mean_token_accuracy": 0.23517142683267594,
      "num_tokens": 85892318.0,
      "step": 41385
    },
    {
      "entropy": 5.547143268585205,
      "epoch": 3.7545355587808418,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003645132594270032,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2212044984102249,
      "num_tokens": 85902719.0,
      "step": 41390
    },
    {
      "entropy": 5.563756227493286,
      "epoch": 3.754989114658926,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036448357290225523,
      "loss": 4.8392,
      "mean_token_accuracy": 0.2256179392337799,
      "num_tokens": 85913271.0,
      "step": 41395
    },
    {
      "entropy": 5.550149583816529,
      "epoch": 3.7554426705370103,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003644538845269642,
      "loss": 4.7868,
      "mean_token_accuracy": 0.2275686427950859,
      "num_tokens": 85922397.0,
      "step": 41400
    },
    {
      "entropy": 5.639441061019897,
      "epoch": 3.7558962264150946,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000364424194301744,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22138051986694335,
      "num_tokens": 85933261.0,
      "step": 41405
    },
    {
      "entropy": 5.669304132461548,
      "epoch": 3.7563497822931784,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003643945022272087,
      "loss": 4.954,
      "mean_token_accuracy": 0.22452425211668015,
      "num_tokens": 85943860.0,
      "step": 41410
    },
    {
      "entropy": 5.656178712844849,
      "epoch": 3.7568033381712627,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003643648083039724,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21566831767559053,
      "num_tokens": 85953984.0,
      "step": 41415
    },
    {
      "entropy": 5.608934211730957,
      "epoch": 3.757256894049347,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000364335112532649,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22674254924058915,
      "num_tokens": 85964771.0,
      "step": 41420
    },
    {
      "entropy": 5.598434019088745,
      "epoch": 3.757710449927431,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003643054149138527,
      "loss": 4.925,
      "mean_token_accuracy": 0.2164538323879242,
      "num_tokens": 85975859.0,
      "step": 41425
    },
    {
      "entropy": 5.680511379241944,
      "epoch": 3.758164005805515,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036427571544819765,
      "loss": 4.9526,
      "mean_token_accuracy": 0.2212819054722786,
      "num_tokens": 85987022.0,
      "step": 41430
    },
    {
      "entropy": 5.573312187194825,
      "epoch": 3.7586175616835993,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003642460141362981,
      "loss": 4.8312,
      "mean_token_accuracy": 0.2280511274933815,
      "num_tokens": 85997056.0,
      "step": 41435
    },
    {
      "entropy": 5.6508503437042235,
      "epoch": 3.7590711175616836,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036421631097876823,
      "loss": 4.947,
      "mean_token_accuracy": 0.21813933849334716,
      "num_tokens": 86008853.0,
      "step": 41440
    },
    {
      "entropy": 5.5742310047149655,
      "epoch": 3.759524673439768,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003641866059762223,
      "loss": 4.9,
      "mean_token_accuracy": 0.21943695843219757,
      "num_tokens": 86020358.0,
      "step": 41445
    },
    {
      "entropy": 5.610048627853393,
      "epoch": 3.759978229317852,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003641568991292747,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21778797805309297,
      "num_tokens": 86030706.0,
      "step": 41450
    },
    {
      "entropy": 5.681662559509277,
      "epoch": 3.7604317851959363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036412719043853956,
      "loss": 4.9622,
      "mean_token_accuracy": 0.216257506608963,
      "num_tokens": 86041964.0,
      "step": 41455
    },
    {
      "entropy": 5.657385921478271,
      "epoch": 3.76088534107402,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036409747990463146,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22602810114622116,
      "num_tokens": 86052213.0,
      "step": 41460
    },
    {
      "entropy": 5.554709625244141,
      "epoch": 3.7613388969521044,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003640677675281649,
      "loss": 4.8898,
      "mean_token_accuracy": 0.2167268455028534,
      "num_tokens": 86061419.0,
      "step": 41465
    },
    {
      "entropy": 5.6276383876800535,
      "epoch": 3.7617924528301887,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003640380533097541,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22321978062391282,
      "num_tokens": 86071585.0,
      "step": 41470
    },
    {
      "entropy": 5.6155191421508786,
      "epoch": 3.762246008708273,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036400833725001364,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22024742513895035,
      "num_tokens": 86081175.0,
      "step": 41475
    },
    {
      "entropy": 5.626105546951294,
      "epoch": 3.762699564586357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003639786193495581,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2163862630724907,
      "num_tokens": 86092407.0,
      "step": 41480
    },
    {
      "entropy": 5.689416694641113,
      "epoch": 3.763153120464441,
      "grad_norm": 1.25,
      "learning_rate": 0.00036394889960900215,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22176722139120103,
      "num_tokens": 86102218.0,
      "step": 41485
    },
    {
      "entropy": 5.670872545242309,
      "epoch": 3.7636066763425253,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003639191780289602,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22324185371398925,
      "num_tokens": 86112631.0,
      "step": 41490
    },
    {
      "entropy": 5.625333642959594,
      "epoch": 3.7640602322206096,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003638894546100469,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21200647652149202,
      "num_tokens": 86123675.0,
      "step": 41495
    },
    {
      "entropy": 5.553248262405395,
      "epoch": 3.764513788098694,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036385972935287715,
      "loss": 4.8318,
      "mean_token_accuracy": 0.22540275752544403,
      "num_tokens": 86133909.0,
      "step": 41500
    },
    {
      "entropy": 5.62788987159729,
      "epoch": 3.764967343976778,
      "grad_norm": 1.25,
      "learning_rate": 0.0003638300022580656,
      "loss": 4.907,
      "mean_token_accuracy": 0.22037242203950883,
      "num_tokens": 86144560.0,
      "step": 41505
    },
    {
      "entropy": 5.704502773284912,
      "epoch": 3.7654208998548624,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036380027332622684,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21510438472032548,
      "num_tokens": 86155140.0,
      "step": 41510
    },
    {
      "entropy": 5.684950113296509,
      "epoch": 3.765874455732946,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036377054255797584,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2167373463511467,
      "num_tokens": 86165555.0,
      "step": 41515
    },
    {
      "entropy": 5.618059492111206,
      "epoch": 3.7663280116110305,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036374080995392737,
      "loss": 4.881,
      "mean_token_accuracy": 0.22100454270839692,
      "num_tokens": 86176204.0,
      "step": 41520
    },
    {
      "entropy": 5.633462381362915,
      "epoch": 3.7667815674891147,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003637110755146964,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21545106470584868,
      "num_tokens": 86186888.0,
      "step": 41525
    },
    {
      "entropy": 5.6192621231079105,
      "epoch": 3.7672351233671986,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036368133924089766,
      "loss": 4.8518,
      "mean_token_accuracy": 0.22867828607559204,
      "num_tokens": 86196654.0,
      "step": 41530
    },
    {
      "entropy": 5.594656562805175,
      "epoch": 3.767688679245283,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003636516011331464,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22529308795928954,
      "num_tokens": 86207045.0,
      "step": 41535
    },
    {
      "entropy": 5.666325998306275,
      "epoch": 3.768142235123367,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036362186119205734,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21006513386964798,
      "num_tokens": 86216906.0,
      "step": 41540
    },
    {
      "entropy": 5.570976591110229,
      "epoch": 3.7685957910014514,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003635921194182457,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2234368920326233,
      "num_tokens": 86227380.0,
      "step": 41545
    },
    {
      "entropy": 5.647460842132569,
      "epoch": 3.7690493468795356,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003635623758123263,
      "loss": 5.0061,
      "mean_token_accuracy": 0.22267422527074815,
      "num_tokens": 86236948.0,
      "step": 41550
    },
    {
      "entropy": 5.616719341278076,
      "epoch": 3.76950290275762,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036353263037491445,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21824501901865007,
      "num_tokens": 86246784.0,
      "step": 41555
    },
    {
      "entropy": 5.5459810256958,
      "epoch": 3.769956458635704,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003635028831066253,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22633751332759858,
      "num_tokens": 86257381.0,
      "step": 41560
    },
    {
      "entropy": 5.549964570999146,
      "epoch": 3.770410014513788,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003634731340080739,
      "loss": 4.8325,
      "mean_token_accuracy": 0.2245854392647743,
      "num_tokens": 86268117.0,
      "step": 41565
    },
    {
      "entropy": 5.739038562774658,
      "epoch": 3.7708635703918723,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003634433830798756,
      "loss": 5.0312,
      "mean_token_accuracy": 0.21311648935079575,
      "num_tokens": 86279195.0,
      "step": 41570
    },
    {
      "entropy": 5.553050518035889,
      "epoch": 3.7713171262699565,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003634136303226457,
      "loss": 4.7345,
      "mean_token_accuracy": 0.23312177509069443,
      "num_tokens": 86288154.0,
      "step": 41575
    },
    {
      "entropy": 5.605372524261474,
      "epoch": 3.771770682148041,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003633838757369992,
      "loss": 4.8931,
      "mean_token_accuracy": 0.21574170440435408,
      "num_tokens": 86297524.0,
      "step": 41580
    },
    {
      "entropy": 5.591262769699097,
      "epoch": 3.7722242380261246,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003633541193235517,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21798088252544404,
      "num_tokens": 86307031.0,
      "step": 41585
    },
    {
      "entropy": 5.597372198104859,
      "epoch": 3.772677793904209,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003633243610829185,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22384639978408813,
      "num_tokens": 86317629.0,
      "step": 41590
    },
    {
      "entropy": 5.647684097290039,
      "epoch": 3.773131349782293,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036329460101571496,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21849690973758698,
      "num_tokens": 86328243.0,
      "step": 41595
    },
    {
      "entropy": 5.707601404190063,
      "epoch": 3.7735849056603774,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003632648391225568,
      "loss": 4.9868,
      "mean_token_accuracy": 0.2123067244887352,
      "num_tokens": 86338905.0,
      "step": 41600
    },
    {
      "entropy": 5.664737367630005,
      "epoch": 3.7740384615384617,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036323507540405903,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2215026304125786,
      "num_tokens": 86350433.0,
      "step": 41605
    },
    {
      "entropy": 5.654308700561524,
      "epoch": 3.774492017416546,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003632053098608374,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21244916915893555,
      "num_tokens": 86360664.0,
      "step": 41610
    },
    {
      "entropy": 5.6497961521148685,
      "epoch": 3.7749455732946298,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036317554249350767,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21997612714767456,
      "num_tokens": 86372101.0,
      "step": 41615
    },
    {
      "entropy": 5.619952440261841,
      "epoch": 3.775399129172714,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003631457733026851,
      "loss": 4.8059,
      "mean_token_accuracy": 0.22608379274606705,
      "num_tokens": 86382494.0,
      "step": 41620
    },
    {
      "entropy": 5.709770059585571,
      "epoch": 3.7758526850507983,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036311600228898554,
      "loss": 5.047,
      "mean_token_accuracy": 0.20622465312480925,
      "num_tokens": 86393292.0,
      "step": 41625
    },
    {
      "entropy": 5.5456939220428465,
      "epoch": 3.7763062409288826,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036308622945302455,
      "loss": 4.8329,
      "mean_token_accuracy": 0.22456151843070984,
      "num_tokens": 86402940.0,
      "step": 41630
    },
    {
      "entropy": 5.56774411201477,
      "epoch": 3.7767597968069664,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000363056454795418,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2190259128808975,
      "num_tokens": 86413382.0,
      "step": 41635
    },
    {
      "entropy": 5.671607112884521,
      "epoch": 3.7772133526850507,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003630266783167814,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21687548905611037,
      "num_tokens": 86423866.0,
      "step": 41640
    },
    {
      "entropy": 5.694989013671875,
      "epoch": 3.777666908563135,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003629969000177307,
      "loss": 4.9868,
      "mean_token_accuracy": 0.2119221195578575,
      "num_tokens": 86434908.0,
      "step": 41645
    },
    {
      "entropy": 5.531735038757324,
      "epoch": 3.778120464441219,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003629671198988817,
      "loss": 4.8215,
      "mean_token_accuracy": 0.2278204768896103,
      "num_tokens": 86445604.0,
      "step": 41650
    },
    {
      "entropy": 5.51928448677063,
      "epoch": 3.7785740203193035,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003629373379608502,
      "loss": 4.7648,
      "mean_token_accuracy": 0.2316206708550453,
      "num_tokens": 86455813.0,
      "step": 41655
    },
    {
      "entropy": 5.491696166992187,
      "epoch": 3.7790275761973877,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036290755420425214,
      "loss": 4.7358,
      "mean_token_accuracy": 0.23947824984788896,
      "num_tokens": 86465256.0,
      "step": 41660
    },
    {
      "entropy": 5.573163175582886,
      "epoch": 3.7794811320754715,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036287776862970356,
      "loss": 4.8656,
      "mean_token_accuracy": 0.22518586218357087,
      "num_tokens": 86474605.0,
      "step": 41665
    },
    {
      "entropy": 5.625118589401245,
      "epoch": 3.779934687953556,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036284798123782024,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2187691405415535,
      "num_tokens": 86484886.0,
      "step": 41670
    },
    {
      "entropy": 5.593328475952148,
      "epoch": 3.78038824383164,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003628181920292182,
      "loss": 4.9561,
      "mean_token_accuracy": 0.213165220618248,
      "num_tokens": 86495816.0,
      "step": 41675
    },
    {
      "entropy": 5.617562866210937,
      "epoch": 3.7808417997097243,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003627884010045138,
      "loss": 4.8571,
      "mean_token_accuracy": 0.221609865128994,
      "num_tokens": 86506319.0,
      "step": 41680
    },
    {
      "entropy": 5.680721759796143,
      "epoch": 3.781295355587808,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003627586081643227,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2204956293106079,
      "num_tokens": 86516952.0,
      "step": 41685
    },
    {
      "entropy": 5.664397478103638,
      "epoch": 3.7817489114658924,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036272881350926135,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21619092673063278,
      "num_tokens": 86527327.0,
      "step": 41690
    },
    {
      "entropy": 5.610509920120239,
      "epoch": 3.7822024673439767,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003626990170399457,
      "loss": 4.8389,
      "mean_token_accuracy": 0.22601523697376252,
      "num_tokens": 86538281.0,
      "step": 41695
    },
    {
      "entropy": 5.613770961761475,
      "epoch": 3.782656023222061,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036266921875699205,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21222818195819854,
      "num_tokens": 86549034.0,
      "step": 41700
    },
    {
      "entropy": 5.616689920425415,
      "epoch": 3.7831095791001452,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003626394186610166,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2178095191717148,
      "num_tokens": 86559306.0,
      "step": 41705
    },
    {
      "entropy": 5.672933721542359,
      "epoch": 3.7835631349782295,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003626096167526356,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2207697793841362,
      "num_tokens": 86569414.0,
      "step": 41710
    },
    {
      "entropy": 5.525878763198852,
      "epoch": 3.7840166908563138,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003625798130324655,
      "loss": 4.7753,
      "mean_token_accuracy": 0.2287873148918152,
      "num_tokens": 86578554.0,
      "step": 41715
    },
    {
      "entropy": 5.64386396408081,
      "epoch": 3.7844702467343976,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036255000750112256,
      "loss": 4.8927,
      "mean_token_accuracy": 0.22212022095918654,
      "num_tokens": 86588624.0,
      "step": 41720
    },
    {
      "entropy": 5.629253625869751,
      "epoch": 3.784923802612482,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036252020015922304,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2229263499379158,
      "num_tokens": 86599266.0,
      "step": 41725
    },
    {
      "entropy": 5.642392110824585,
      "epoch": 3.785377358490566,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036249039100738357,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21143231093883513,
      "num_tokens": 86609853.0,
      "step": 41730
    },
    {
      "entropy": 5.550648355484009,
      "epoch": 3.7858309143686504,
      "grad_norm": 1.25,
      "learning_rate": 0.0003624605800462205,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22352551221847533,
      "num_tokens": 86619927.0,
      "step": 41735
    },
    {
      "entropy": 5.6466514587402346,
      "epoch": 3.786284470246734,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036243076727635027,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2233829215168953,
      "num_tokens": 86630079.0,
      "step": 41740
    },
    {
      "entropy": 5.610857915878296,
      "epoch": 3.7867380261248185,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036240095269838955,
      "loss": 4.8717,
      "mean_token_accuracy": 0.22407594323158264,
      "num_tokens": 86639744.0,
      "step": 41745
    },
    {
      "entropy": 5.5412956237792965,
      "epoch": 3.7871915820029027,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036237113631295485,
      "loss": 4.9416,
      "mean_token_accuracy": 0.2198496863245964,
      "num_tokens": 86650866.0,
      "step": 41750
    },
    {
      "entropy": 5.583549928665161,
      "epoch": 3.787645137880987,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036234131812066286,
      "loss": 4.7977,
      "mean_token_accuracy": 0.22846141308546067,
      "num_tokens": 86659864.0,
      "step": 41755
    },
    {
      "entropy": 5.566881942749023,
      "epoch": 3.7880986937590713,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036231149812213006,
      "loss": 4.8138,
      "mean_token_accuracy": 0.22403739988803864,
      "num_tokens": 86670039.0,
      "step": 41760
    },
    {
      "entropy": 5.586530494689941,
      "epoch": 3.7885522496371555,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003622816763179732,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22727217376232148,
      "num_tokens": 86680434.0,
      "step": 41765
    },
    {
      "entropy": 5.625749588012695,
      "epoch": 3.7890058055152394,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003622518527088091,
      "loss": 4.9194,
      "mean_token_accuracy": 0.21443816274404526,
      "num_tokens": 86691026.0,
      "step": 41770
    },
    {
      "entropy": 5.656641674041748,
      "epoch": 3.7894593613933236,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036222202729525453,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21857694387435914,
      "num_tokens": 86701323.0,
      "step": 41775
    },
    {
      "entropy": 5.592256879806518,
      "epoch": 3.789912917271408,
      "grad_norm": 1.234375,
      "learning_rate": 0.000362192200077926,
      "loss": 4.8391,
      "mean_token_accuracy": 0.23345235884189605,
      "num_tokens": 86711152.0,
      "step": 41780
    },
    {
      "entropy": 5.656635856628418,
      "epoch": 3.790366473149492,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036216237105744076,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22142889201641083,
      "num_tokens": 86720814.0,
      "step": 41785
    },
    {
      "entropy": 5.622056865692139,
      "epoch": 3.790820029027576,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003621325402344154,
      "loss": 4.8665,
      "mean_token_accuracy": 0.22930116802453995,
      "num_tokens": 86731229.0,
      "step": 41790
    },
    {
      "entropy": 5.598729991912842,
      "epoch": 3.7912735849056602,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036210270760946684,
      "loss": 4.869,
      "mean_token_accuracy": 0.21661020666360856,
      "num_tokens": 86741440.0,
      "step": 41795
    },
    {
      "entropy": 5.6839375495910645,
      "epoch": 3.7917271407837445,
      "grad_norm": 1.25,
      "learning_rate": 0.0003620728731832122,
      "loss": 4.968,
      "mean_token_accuracy": 0.21746186017990113,
      "num_tokens": 86752022.0,
      "step": 41800
    },
    {
      "entropy": 5.636864900588989,
      "epoch": 3.7921806966618288,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003620430369562683,
      "loss": 4.8604,
      "mean_token_accuracy": 0.22020047754049302,
      "num_tokens": 86763090.0,
      "step": 41805
    },
    {
      "entropy": 5.5830785751342775,
      "epoch": 3.792634252539913,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036201319892925214,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22351621389389037,
      "num_tokens": 86773357.0,
      "step": 41810
    },
    {
      "entropy": 5.552571392059326,
      "epoch": 3.7930878084179973,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000361983359102781,
      "loss": 4.8261,
      "mean_token_accuracy": 0.2198467567563057,
      "num_tokens": 86784402.0,
      "step": 41815
    },
    {
      "entropy": 5.612296438217163,
      "epoch": 3.793541364296081,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036195351747747177,
      "loss": 4.9221,
      "mean_token_accuracy": 0.22276830524206162,
      "num_tokens": 86794945.0,
      "step": 41820
    },
    {
      "entropy": 5.604290246963501,
      "epoch": 3.7939949201741654,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036192367405394157,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22098950743675233,
      "num_tokens": 86806936.0,
      "step": 41825
    },
    {
      "entropy": 5.688607788085937,
      "epoch": 3.7944484760522497,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003618938288328077,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21362111270427703,
      "num_tokens": 86817776.0,
      "step": 41830
    },
    {
      "entropy": 5.538198375701905,
      "epoch": 3.794902031930334,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003618639818146873,
      "loss": 4.7681,
      "mean_token_accuracy": 0.22893187254667283,
      "num_tokens": 86828054.0,
      "step": 41835
    },
    {
      "entropy": 5.578504753112793,
      "epoch": 3.7953555878084178,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036183413300019767,
      "loss": 4.8372,
      "mean_token_accuracy": 0.22775990068912505,
      "num_tokens": 86838132.0,
      "step": 41840
    },
    {
      "entropy": 5.565424203872681,
      "epoch": 3.795809143686502,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036180428238995595,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21613416969776153,
      "num_tokens": 86848219.0,
      "step": 41845
    },
    {
      "entropy": 5.53728985786438,
      "epoch": 3.7962626995645863,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003617744299845796,
      "loss": 4.8115,
      "mean_token_accuracy": 0.22700901478528976,
      "num_tokens": 86858559.0,
      "step": 41850
    },
    {
      "entropy": 5.556297063827515,
      "epoch": 3.7967162554426706,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003617445757846859,
      "loss": 4.9213,
      "mean_token_accuracy": 0.22252996563911437,
      "num_tokens": 86868573.0,
      "step": 41855
    },
    {
      "entropy": 5.682421875,
      "epoch": 3.797169811320755,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036171471979089227,
      "loss": 4.8899,
      "mean_token_accuracy": 0.2246537908911705,
      "num_tokens": 86878641.0,
      "step": 41860
    },
    {
      "entropy": 5.632895517349243,
      "epoch": 3.797623367198839,
      "grad_norm": 1.375,
      "learning_rate": 0.0003616848620038162,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21448186039924622,
      "num_tokens": 86888644.0,
      "step": 41865
    },
    {
      "entropy": 5.66579532623291,
      "epoch": 3.7980769230769234,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000361655002424075,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20469139218330384,
      "num_tokens": 86899428.0,
      "step": 41870
    },
    {
      "entropy": 5.666243648529052,
      "epoch": 3.798530478955007,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036162514105228635,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22285107523202896,
      "num_tokens": 86910155.0,
      "step": 41875
    },
    {
      "entropy": 5.588804864883423,
      "epoch": 3.7989840348330914,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003615952778890676,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22400906383991243,
      "num_tokens": 86919783.0,
      "step": 41880
    },
    {
      "entropy": 5.592104291915893,
      "epoch": 3.7994375907111757,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036156541293503653,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2187621384859085,
      "num_tokens": 86929647.0,
      "step": 41885
    },
    {
      "entropy": 5.6078362464904785,
      "epoch": 3.7998911465892595,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036153554619081064,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2204468935728073,
      "num_tokens": 86940113.0,
      "step": 41890
    },
    {
      "entropy": 5.641484308242798,
      "epoch": 3.800344702467344,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003615056776570076,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2210904464125633,
      "num_tokens": 86949950.0,
      "step": 41895
    },
    {
      "entropy": 5.573281574249267,
      "epoch": 3.800798258345428,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036147580733424513,
      "loss": 4.8035,
      "mean_token_accuracy": 0.23234857767820358,
      "num_tokens": 86959286.0,
      "step": 41900
    },
    {
      "entropy": 5.572811555862427,
      "epoch": 3.8012518142235123,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003614459352231409,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22602326720952987,
      "num_tokens": 86970165.0,
      "step": 41905
    },
    {
      "entropy": 5.557807970046997,
      "epoch": 3.8017053701015966,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003614160613243127,
      "loss": 4.8658,
      "mean_token_accuracy": 0.21917305439710616,
      "num_tokens": 86980331.0,
      "step": 41910
    },
    {
      "entropy": 5.663860750198364,
      "epoch": 3.802158925979681,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003613861856383783,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2141605168581009,
      "num_tokens": 86991790.0,
      "step": 41915
    },
    {
      "entropy": 5.647479629516601,
      "epoch": 3.802612481857765,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003613563081659556,
      "loss": 5.0078,
      "mean_token_accuracy": 0.22258283346891403,
      "num_tokens": 87001220.0,
      "step": 41920
    },
    {
      "entropy": 5.649091958999634,
      "epoch": 3.803066037735849,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003613264289076625,
      "loss": 4.8792,
      "mean_token_accuracy": 0.22127438336610794,
      "num_tokens": 87012720.0,
      "step": 41925
    },
    {
      "entropy": 5.628114414215088,
      "epoch": 3.803519593613933,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003612965478641167,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2181042566895485,
      "num_tokens": 87021891.0,
      "step": 41930
    },
    {
      "entropy": 5.670102119445801,
      "epoch": 3.8039731494920175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036126666503593643,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21555988043546676,
      "num_tokens": 87032546.0,
      "step": 41935
    },
    {
      "entropy": 5.596625947952271,
      "epoch": 3.8044267053701017,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003612367804237395,
      "loss": 4.8524,
      "mean_token_accuracy": 0.21960129141807555,
      "num_tokens": 87042173.0,
      "step": 41940
    },
    {
      "entropy": 5.633037757873535,
      "epoch": 3.8048802612481856,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003612068940281439,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22041576355695724,
      "num_tokens": 87052475.0,
      "step": 41945
    },
    {
      "entropy": 5.606350660324097,
      "epoch": 3.80533381712627,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003611770058497679,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2164686515927315,
      "num_tokens": 87062698.0,
      "step": 41950
    },
    {
      "entropy": 5.584737634658813,
      "epoch": 3.805787373004354,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036114711588922945,
      "loss": 4.8606,
      "mean_token_accuracy": 0.21688618808984755,
      "num_tokens": 87073168.0,
      "step": 41955
    },
    {
      "entropy": 5.666655540466309,
      "epoch": 3.8062409288824384,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003611172241471466,
      "loss": 4.9757,
      "mean_token_accuracy": 0.2102243885397911,
      "num_tokens": 87084270.0,
      "step": 41960
    },
    {
      "entropy": 5.58031587600708,
      "epoch": 3.8066944847605226,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003610873306241376,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2205825537443161,
      "num_tokens": 87093869.0,
      "step": 41965
    },
    {
      "entropy": 5.577521228790284,
      "epoch": 3.807148040638607,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036105743532082075,
      "loss": 4.8544,
      "mean_token_accuracy": 0.22242291420698165,
      "num_tokens": 87103371.0,
      "step": 41970
    },
    {
      "entropy": 5.623032855987549,
      "epoch": 3.8076015965166907,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003610275382378142,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22170430272817612,
      "num_tokens": 87113384.0,
      "step": 41975
    },
    {
      "entropy": 5.661807250976563,
      "epoch": 3.808055152394775,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003609976393757362,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22047605812549592,
      "num_tokens": 87123925.0,
      "step": 41980
    },
    {
      "entropy": 5.576672172546386,
      "epoch": 3.8085087082728593,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036096773873520513,
      "loss": 4.8598,
      "mean_token_accuracy": 0.21996969878673553,
      "num_tokens": 87134608.0,
      "step": 41985
    },
    {
      "entropy": 5.598281478881836,
      "epoch": 3.8089622641509435,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036093783631683927,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22203450947999953,
      "num_tokens": 87145184.0,
      "step": 41990
    },
    {
      "entropy": 5.55139741897583,
      "epoch": 3.8094158200290273,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003609079321212571,
      "loss": 4.787,
      "mean_token_accuracy": 0.22273022383451463,
      "num_tokens": 87155483.0,
      "step": 41995
    },
    {
      "entropy": 5.607876491546631,
      "epoch": 3.8098693759071116,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036087802614907696,
      "loss": 4.979,
      "mean_token_accuracy": 0.2144949719309807,
      "num_tokens": 87166615.0,
      "step": 42000
    },
    {
      "epoch": 3.8098693759071116,
      "eval_entropy": 5.346203438081353,
      "eval_loss": 4.952507972717285,
      "eval_mean_token_accuracy": 0.22764768383351927,
      "eval_num_tokens": 87166615.0,
      "eval_runtime": 20.6758,
      "eval_samples_per_second": 1710.213,
      "eval_steps_per_second": 213.777,
      "step": 42000
    },
    {
      "entropy": 5.634857940673828,
      "epoch": 3.810322931785196,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036084811840091736,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2212119996547699,
      "num_tokens": 87177151.0,
      "step": 42005
    },
    {
      "entropy": 5.584418439865113,
      "epoch": 3.81077648766328,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003608182088773969,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22364019155502318,
      "num_tokens": 87188085.0,
      "step": 42010
    },
    {
      "entropy": 5.588363838195801,
      "epoch": 3.8112300435413644,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003607882975791338,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21974351853132248,
      "num_tokens": 87197240.0,
      "step": 42015
    },
    {
      "entropy": 5.598347997665405,
      "epoch": 3.8116835994194487,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000360758384506747,
      "loss": 4.8729,
      "mean_token_accuracy": 0.2303866043686867,
      "num_tokens": 87208064.0,
      "step": 42020
    },
    {
      "entropy": 5.596074771881104,
      "epoch": 3.8121371552975325,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036072846966085496,
      "loss": 4.8006,
      "mean_token_accuracy": 0.22313464432954788,
      "num_tokens": 87218386.0,
      "step": 42025
    },
    {
      "entropy": 5.579715347290039,
      "epoch": 3.8125907111756168,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036069855304207634,
      "loss": 4.7776,
      "mean_token_accuracy": 0.2290860891342163,
      "num_tokens": 87228459.0,
      "step": 42030
    },
    {
      "entropy": 5.61886830329895,
      "epoch": 3.813044267053701,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003606686346510297,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21134898513555528,
      "num_tokens": 87240317.0,
      "step": 42035
    },
    {
      "entropy": 5.523982286453247,
      "epoch": 3.8134978229317853,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000360638714488334,
      "loss": 4.789,
      "mean_token_accuracy": 0.22310695350170134,
      "num_tokens": 87250197.0,
      "step": 42040
    },
    {
      "entropy": 5.551533365249634,
      "epoch": 3.813951378809869,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036060879255460766,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22053295224905015,
      "num_tokens": 87261266.0,
      "step": 42045
    },
    {
      "entropy": 5.530144309997558,
      "epoch": 3.8144049346879534,
      "grad_norm": 1.375,
      "learning_rate": 0.0003605788688504699,
      "loss": 4.7593,
      "mean_token_accuracy": 0.23171026557683944,
      "num_tokens": 87270490.0,
      "step": 42050
    },
    {
      "entropy": 5.625289344787598,
      "epoch": 3.8148584905660377,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003605489433765392,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22309321761131287,
      "num_tokens": 87281260.0,
      "step": 42055
    },
    {
      "entropy": 5.584912395477295,
      "epoch": 3.815312046444122,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003605190161334345,
      "loss": 4.8134,
      "mean_token_accuracy": 0.23128933757543563,
      "num_tokens": 87290577.0,
      "step": 42060
    },
    {
      "entropy": 5.535187673568726,
      "epoch": 3.815765602322206,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003604890871217748,
      "loss": 4.7415,
      "mean_token_accuracy": 0.23366574198007584,
      "num_tokens": 87300185.0,
      "step": 42065
    },
    {
      "entropy": 5.625460577011109,
      "epoch": 3.8162191582002905,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003604591563421791,
      "loss": 4.9756,
      "mean_token_accuracy": 0.22205374389886856,
      "num_tokens": 87310088.0,
      "step": 42070
    },
    {
      "entropy": 5.612613296508789,
      "epoch": 3.8166727140783747,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003604292237952662,
      "loss": 4.896,
      "mean_token_accuracy": 0.21813660860061646,
      "num_tokens": 87319942.0,
      "step": 42075
    },
    {
      "entropy": 5.635054016113282,
      "epoch": 3.8171262699564585,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003603992894816551,
      "loss": 4.8672,
      "mean_token_accuracy": 0.21989766508340836,
      "num_tokens": 87329175.0,
      "step": 42080
    },
    {
      "entropy": 5.5353446960449215,
      "epoch": 3.817579825834543,
      "grad_norm": 1.40625,
      "learning_rate": 0.000360369353401965,
      "loss": 4.7841,
      "mean_token_accuracy": 0.22539449185132981,
      "num_tokens": 87338555.0,
      "step": 42085
    },
    {
      "entropy": 5.573533773422241,
      "epoch": 3.818033381712627,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036033941555681486,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22953293919563295,
      "num_tokens": 87348424.0,
      "step": 42090
    },
    {
      "entropy": 5.584469223022461,
      "epoch": 3.8184869375907113,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036030947594682395,
      "loss": 4.7945,
      "mean_token_accuracy": 0.22611795663833617,
      "num_tokens": 87359724.0,
      "step": 42095
    },
    {
      "entropy": 5.604579305648803,
      "epoch": 3.818940493468795,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003602795345726113,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22013984769582748,
      "num_tokens": 87371324.0,
      "step": 42100
    },
    {
      "entropy": 5.614258718490601,
      "epoch": 3.8193940493468794,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003602495914347962,
      "loss": 4.7892,
      "mean_token_accuracy": 0.2261779174208641,
      "num_tokens": 87380770.0,
      "step": 42105
    },
    {
      "entropy": 5.556986856460571,
      "epoch": 3.8198476052249637,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036021964653399773,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22143764197826385,
      "num_tokens": 87391123.0,
      "step": 42110
    },
    {
      "entropy": 5.616132926940918,
      "epoch": 3.820301161103048,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036018969987083524,
      "loss": 4.923,
      "mean_token_accuracy": 0.22146390974521638,
      "num_tokens": 87401357.0,
      "step": 42115
    },
    {
      "entropy": 5.616885948181152,
      "epoch": 3.8207547169811322,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003601597514459281,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22252754718065262,
      "num_tokens": 87411373.0,
      "step": 42120
    },
    {
      "entropy": 5.560558414459228,
      "epoch": 3.8212082728592165,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036012980125989567,
      "loss": 4.8456,
      "mean_token_accuracy": 0.230593766272068,
      "num_tokens": 87421915.0,
      "step": 42125
    },
    {
      "entropy": 5.585498666763305,
      "epoch": 3.8216618287373003,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003600998493133571,
      "loss": 4.8833,
      "mean_token_accuracy": 0.2263690635561943,
      "num_tokens": 87432148.0,
      "step": 42130
    },
    {
      "entropy": 5.585041570663452,
      "epoch": 3.8221153846153846,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003600698956069321,
      "loss": 4.9194,
      "mean_token_accuracy": 0.21915029883384704,
      "num_tokens": 87442605.0,
      "step": 42135
    },
    {
      "entropy": 5.627900886535644,
      "epoch": 3.822568940493469,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003600399401412399,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2261598750948906,
      "num_tokens": 87452482.0,
      "step": 42140
    },
    {
      "entropy": 5.51025619506836,
      "epoch": 3.823022496371553,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036000998291690005,
      "loss": 4.7473,
      "mean_token_accuracy": 0.22577544003725053,
      "num_tokens": 87462074.0,
      "step": 42145
    },
    {
      "entropy": 5.617627239227295,
      "epoch": 3.823476052249637,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003599800239345321,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22161577194929122,
      "num_tokens": 87472377.0,
      "step": 42150
    },
    {
      "entropy": 5.569552326202393,
      "epoch": 3.823929608127721,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003599500631947557,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22076617181301117,
      "num_tokens": 87483436.0,
      "step": 42155
    },
    {
      "entropy": 5.6514184951782225,
      "epoch": 3.8243831640058055,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035992010069819025,
      "loss": 4.8555,
      "mean_token_accuracy": 0.2278494119644165,
      "num_tokens": 87492231.0,
      "step": 42160
    },
    {
      "entropy": 5.73201904296875,
      "epoch": 3.8248367198838897,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035989013644545555,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22374685406684874,
      "num_tokens": 87503416.0,
      "step": 42165
    },
    {
      "entropy": 5.58912296295166,
      "epoch": 3.825290275761974,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035986017043717114,
      "loss": 4.8504,
      "mean_token_accuracy": 0.226153764128685,
      "num_tokens": 87513737.0,
      "step": 42170
    },
    {
      "entropy": 5.605336904525757,
      "epoch": 3.8257438316400583,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003598302026739568,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21784147918224334,
      "num_tokens": 87523532.0,
      "step": 42175
    },
    {
      "entropy": 5.557153987884521,
      "epoch": 3.826197387518142,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035980023315643217,
      "loss": 4.7681,
      "mean_token_accuracy": 0.2319904774427414,
      "num_tokens": 87534066.0,
      "step": 42180
    },
    {
      "entropy": 5.6845635890960695,
      "epoch": 3.8266509433962264,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003597702618852172,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2124808356165886,
      "num_tokens": 87544520.0,
      "step": 42185
    },
    {
      "entropy": 5.640701150894165,
      "epoch": 3.8271044992743106,
      "grad_norm": 1.25,
      "learning_rate": 0.0003597402888609316,
      "loss": 4.903,
      "mean_token_accuracy": 0.22138697952032088,
      "num_tokens": 87553913.0,
      "step": 42190
    },
    {
      "entropy": 5.626333475112915,
      "epoch": 3.827558055152395,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035971031408419524,
      "loss": 4.8907,
      "mean_token_accuracy": 0.21907430440187453,
      "num_tokens": 87563434.0,
      "step": 42195
    },
    {
      "entropy": 5.5932670593261715,
      "epoch": 3.8280116110304787,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035968033755562795,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22602132558822632,
      "num_tokens": 87573137.0,
      "step": 42200
    },
    {
      "entropy": 5.663025760650635,
      "epoch": 3.828465166908563,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003596503592758497,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22563652992248534,
      "num_tokens": 87582610.0,
      "step": 42205
    },
    {
      "entropy": 5.5894327640533445,
      "epoch": 3.8289187227866472,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003596203792454805,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2237766295671463,
      "num_tokens": 87593062.0,
      "step": 42210
    },
    {
      "entropy": 5.595774793624878,
      "epoch": 3.8293722786647315,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003595903974651403,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21620270162820815,
      "num_tokens": 87604250.0,
      "step": 42215
    },
    {
      "entropy": 5.656678199768066,
      "epoch": 3.8298258345428158,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003595604139354491,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2179296463727951,
      "num_tokens": 87615434.0,
      "step": 42220
    },
    {
      "entropy": 5.6424988269805905,
      "epoch": 3.8302793904209,
      "grad_norm": 1.328125,
      "learning_rate": 0.000359530428657027,
      "loss": 4.8676,
      "mean_token_accuracy": 0.22537012696266173,
      "num_tokens": 87625005.0,
      "step": 42225
    },
    {
      "entropy": 5.560103893280029,
      "epoch": 3.8307329462989843,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003595004416304941,
      "loss": 4.8385,
      "mean_token_accuracy": 0.22304467260837554,
      "num_tokens": 87635486.0,
      "step": 42230
    },
    {
      "entropy": 5.58238525390625,
      "epoch": 3.831186502177068,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035947045285647045,
      "loss": 4.858,
      "mean_token_accuracy": 0.230358025431633,
      "num_tokens": 87645771.0,
      "step": 42235
    },
    {
      "entropy": 5.687296152114868,
      "epoch": 3.8316400580551524,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003594404623355765,
      "loss": 4.9382,
      "mean_token_accuracy": 0.22328181862831115,
      "num_tokens": 87656116.0,
      "step": 42240
    },
    {
      "entropy": 5.795941877365112,
      "epoch": 3.8320936139332367,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003594104700684321,
      "loss": 5.0791,
      "mean_token_accuracy": 0.2093728557229042,
      "num_tokens": 87666476.0,
      "step": 42245
    },
    {
      "entropy": 5.611774349212647,
      "epoch": 3.8325471698113205,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035938047605565764,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23027875125408173,
      "num_tokens": 87677406.0,
      "step": 42250
    },
    {
      "entropy": 5.543580770492554,
      "epoch": 3.8330007256894048,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035935048029787356,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2273533046245575,
      "num_tokens": 87687461.0,
      "step": 42255
    },
    {
      "entropy": 5.577033376693725,
      "epoch": 3.833454281567489,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003593204827957,
      "loss": 4.839,
      "mean_token_accuracy": 0.21658724248409272,
      "num_tokens": 87696470.0,
      "step": 42260
    },
    {
      "entropy": 5.603527355194092,
      "epoch": 3.8339078374455733,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003592904835497574,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2153046727180481,
      "num_tokens": 87706598.0,
      "step": 42265
    },
    {
      "entropy": 5.726874208450317,
      "epoch": 3.8343613933236576,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003592604825606661,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21628989577293395,
      "num_tokens": 87717749.0,
      "step": 42270
    },
    {
      "entropy": 5.596792888641358,
      "epoch": 3.834814949201742,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003592304798290465,
      "loss": 4.8275,
      "mean_token_accuracy": 0.23229033052921294,
      "num_tokens": 87727893.0,
      "step": 42275
    },
    {
      "entropy": 5.527669095993042,
      "epoch": 3.835268505079826,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003592004753555191,
      "loss": 4.923,
      "mean_token_accuracy": 0.2199908211827278,
      "num_tokens": 87737980.0,
      "step": 42280
    },
    {
      "entropy": 5.604449844360351,
      "epoch": 3.83572206095791,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003591704691407045,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2235417366027832,
      "num_tokens": 87749271.0,
      "step": 42285
    },
    {
      "entropy": 5.605932092666626,
      "epoch": 3.836175616835994,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035914046118522313,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22407016307115554,
      "num_tokens": 87760523.0,
      "step": 42290
    },
    {
      "entropy": 5.550204610824585,
      "epoch": 3.8366291727140784,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035911045148969556,
      "loss": 4.8497,
      "mean_token_accuracy": 0.2258607879281044,
      "num_tokens": 87770084.0,
      "step": 42295
    },
    {
      "entropy": 5.676536130905151,
      "epoch": 3.8370827285921627,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035908044005474244,
      "loss": 4.9586,
      "mean_token_accuracy": 0.2140437588095665,
      "num_tokens": 87780163.0,
      "step": 42300
    },
    {
      "entropy": 5.551030397415161,
      "epoch": 3.8375362844702465,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003590504268809843,
      "loss": 4.8266,
      "mean_token_accuracy": 0.22786989510059358,
      "num_tokens": 87791382.0,
      "step": 42305
    },
    {
      "entropy": 5.489073753356934,
      "epoch": 3.837989840348331,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035902041196904204,
      "loss": 4.8434,
      "mean_token_accuracy": 0.21914518624544144,
      "num_tokens": 87802143.0,
      "step": 42310
    },
    {
      "entropy": 5.599003648757934,
      "epoch": 3.838443396226415,
      "grad_norm": 1.375,
      "learning_rate": 0.0003589903953195362,
      "loss": 4.8953,
      "mean_token_accuracy": 0.21948555558919908,
      "num_tokens": 87812255.0,
      "step": 42315
    },
    {
      "entropy": 5.648344659805298,
      "epoch": 3.8388969521044993,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035896037693308757,
      "loss": 4.8991,
      "mean_token_accuracy": 0.2194198936223984,
      "num_tokens": 87822744.0,
      "step": 42320
    },
    {
      "entropy": 5.715447425842285,
      "epoch": 3.8393505079825836,
      "grad_norm": 1.34375,
      "learning_rate": 0.000358930356810317,
      "loss": 4.9623,
      "mean_token_accuracy": 0.21412782967090607,
      "num_tokens": 87833701.0,
      "step": 42325
    },
    {
      "entropy": 5.644897508621216,
      "epoch": 3.839804063860668,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003589003349518452,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22096876353025435,
      "num_tokens": 87844215.0,
      "step": 42330
    },
    {
      "entropy": 5.67422251701355,
      "epoch": 3.8402576197387517,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035887031135829305,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21518833935260773,
      "num_tokens": 87854281.0,
      "step": 42335
    },
    {
      "entropy": 5.652591276168823,
      "epoch": 3.840711175616836,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003588402860302816,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21988693177700042,
      "num_tokens": 87866003.0,
      "step": 42340
    },
    {
      "entropy": 5.616757011413574,
      "epoch": 3.84116473149492,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035881025896843165,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2238581120967865,
      "num_tokens": 87875854.0,
      "step": 42345
    },
    {
      "entropy": 5.679847669601441,
      "epoch": 3.8416182873730045,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035878023017336415,
      "loss": 5.0027,
      "mean_token_accuracy": 0.21685487478971482,
      "num_tokens": 87887242.0,
      "step": 42350
    },
    {
      "entropy": 5.658097791671753,
      "epoch": 3.8420718432510883,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003587501996457001,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22040608674287795,
      "num_tokens": 87897859.0,
      "step": 42355
    },
    {
      "entropy": 5.607671070098877,
      "epoch": 3.8425253991291726,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035872016738606067,
      "loss": 4.8571,
      "mean_token_accuracy": 0.22060235142707824,
      "num_tokens": 87908945.0,
      "step": 42360
    },
    {
      "entropy": 5.603793907165527,
      "epoch": 3.842978955007257,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035869013339506674,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22771041691303254,
      "num_tokens": 87919572.0,
      "step": 42365
    },
    {
      "entropy": 5.633954954147339,
      "epoch": 3.843432510885341,
      "grad_norm": 1.484375,
      "learning_rate": 0.00035866009767333964,
      "loss": 4.9335,
      "mean_token_accuracy": 0.2216442972421646,
      "num_tokens": 87929337.0,
      "step": 42370
    },
    {
      "entropy": 5.671731328964233,
      "epoch": 3.8438860667634254,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035863006022150034,
      "loss": 4.9754,
      "mean_token_accuracy": 0.214978164434433,
      "num_tokens": 87940513.0,
      "step": 42375
    },
    {
      "entropy": 5.585833549499512,
      "epoch": 3.8443396226415096,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035860002104017014,
      "loss": 4.821,
      "mean_token_accuracy": 0.2299963966012001,
      "num_tokens": 87951181.0,
      "step": 42380
    },
    {
      "entropy": 5.624216175079345,
      "epoch": 3.8447931785195935,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003585699801299701,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22295817881822586,
      "num_tokens": 87961925.0,
      "step": 42385
    },
    {
      "entropy": 5.618380832672119,
      "epoch": 3.8452467343976777,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035853993749152165,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22466915994882583,
      "num_tokens": 87972233.0,
      "step": 42390
    },
    {
      "entropy": 5.619912910461426,
      "epoch": 3.845700290275762,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035850989312544596,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2176641121506691,
      "num_tokens": 87983131.0,
      "step": 42395
    },
    {
      "entropy": 5.586140537261963,
      "epoch": 3.8461538461538463,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003584798470323644,
      "loss": 4.8411,
      "mean_token_accuracy": 0.22479360699653625,
      "num_tokens": 87993926.0,
      "step": 42400
    },
    {
      "entropy": 5.684434318542481,
      "epoch": 3.84660740203193,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003584497992128984,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21059598326683043,
      "num_tokens": 88004403.0,
      "step": 42405
    },
    {
      "entropy": 5.605659770965576,
      "epoch": 3.8470609579100143,
      "grad_norm": 1.25,
      "learning_rate": 0.0003584197496676692,
      "loss": 4.8373,
      "mean_token_accuracy": 0.22597959637641907,
      "num_tokens": 88015289.0,
      "step": 42410
    },
    {
      "entropy": 5.56340274810791,
      "epoch": 3.8475145137880986,
      "grad_norm": 1.25,
      "learning_rate": 0.00035838969839729846,
      "loss": 4.8294,
      "mean_token_accuracy": 0.2276433750987053,
      "num_tokens": 88025724.0,
      "step": 42415
    },
    {
      "entropy": 5.577134847640991,
      "epoch": 3.847968069666183,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003583596454024073,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21406975239515305,
      "num_tokens": 88036444.0,
      "step": 42420
    },
    {
      "entropy": 5.650828456878662,
      "epoch": 3.848421625544267,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003583295906836176,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2251036211848259,
      "num_tokens": 88047334.0,
      "step": 42425
    },
    {
      "entropy": 5.623916578292847,
      "epoch": 3.8488751814223514,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035829953424155065,
      "loss": 4.8689,
      "mean_token_accuracy": 0.21903757452964784,
      "num_tokens": 88057268.0,
      "step": 42430
    },
    {
      "entropy": 5.573695802688599,
      "epoch": 3.8493287373004357,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003582694760768281,
      "loss": 4.846,
      "mean_token_accuracy": 0.22270508706569672,
      "num_tokens": 88067615.0,
      "step": 42435
    },
    {
      "entropy": 5.645818090438842,
      "epoch": 3.8497822931785195,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035823941619007155,
      "loss": 4.9704,
      "mean_token_accuracy": 0.217561174929142,
      "num_tokens": 88077754.0,
      "step": 42440
    },
    {
      "entropy": 5.588379621505737,
      "epoch": 3.8502358490566038,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035820935458190275,
      "loss": 4.8614,
      "mean_token_accuracy": 0.21765046566724777,
      "num_tokens": 88088346.0,
      "step": 42445
    },
    {
      "entropy": 5.544213390350341,
      "epoch": 3.850689404934688,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035817929125294323,
      "loss": 4.8296,
      "mean_token_accuracy": 0.2274036228656769,
      "num_tokens": 88098534.0,
      "step": 42450
    },
    {
      "entropy": 5.563419008255005,
      "epoch": 3.8511429608127723,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003581492262038148,
      "loss": 4.9101,
      "mean_token_accuracy": 0.2232576996088028,
      "num_tokens": 88109264.0,
      "step": 42455
    },
    {
      "entropy": 5.633353281021118,
      "epoch": 3.851596516690856,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003581191594351392,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2245269924402237,
      "num_tokens": 88120407.0,
      "step": 42460
    },
    {
      "entropy": 5.640510177612304,
      "epoch": 3.8520500725689404,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003580890909475381,
      "loss": 4.8684,
      "mean_token_accuracy": 0.217959201335907,
      "num_tokens": 88131908.0,
      "step": 42465
    },
    {
      "entropy": 5.549888801574707,
      "epoch": 3.8525036284470247,
      "grad_norm": 1.25,
      "learning_rate": 0.00035805902074163354,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2214904472231865,
      "num_tokens": 88142559.0,
      "step": 42470
    },
    {
      "entropy": 5.587997722625732,
      "epoch": 3.852957184325109,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003580289488180472,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22519796192646027,
      "num_tokens": 88153441.0,
      "step": 42475
    },
    {
      "entropy": 5.607268524169922,
      "epoch": 3.853410740203193,
      "grad_norm": 1.828125,
      "learning_rate": 0.00035799887517740103,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22467581629753114,
      "num_tokens": 88164153.0,
      "step": 42480
    },
    {
      "entropy": 5.675697660446167,
      "epoch": 3.8538642960812775,
      "grad_norm": 1.375,
      "learning_rate": 0.000357968799820317,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21547329127788545,
      "num_tokens": 88174138.0,
      "step": 42485
    },
    {
      "entropy": 5.620659351348877,
      "epoch": 3.8543178519593613,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003579387227474171,
      "loss": 4.8975,
      "mean_token_accuracy": 0.21933926939964293,
      "num_tokens": 88184704.0,
      "step": 42490
    },
    {
      "entropy": 5.7066975116729735,
      "epoch": 3.8547714078374455,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003579086439593232,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21374958902597427,
      "num_tokens": 88194225.0,
      "step": 42495
    },
    {
      "entropy": 5.623697280883789,
      "epoch": 3.85522496371553,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003578785634566574,
      "loss": 4.906,
      "mean_token_accuracy": 0.2221931904554367,
      "num_tokens": 88206125.0,
      "step": 42500
    },
    {
      "entropy": 5.568353939056396,
      "epoch": 3.855678519593614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035784848124004187,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23359186500310897,
      "num_tokens": 88217085.0,
      "step": 42505
    },
    {
      "entropy": 5.607264709472656,
      "epoch": 3.856132075471698,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003578183973100986,
      "loss": 4.8882,
      "mean_token_accuracy": 0.21627986580133438,
      "num_tokens": 88228126.0,
      "step": 42510
    },
    {
      "entropy": 5.624561405181884,
      "epoch": 3.856585631349782,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035778831166744966,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21362484842538834,
      "num_tokens": 88238829.0,
      "step": 42515
    },
    {
      "entropy": 5.630802345275879,
      "epoch": 3.8570391872278664,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003577582243127175,
      "loss": 4.8998,
      "mean_token_accuracy": 0.21610333323478698,
      "num_tokens": 88249821.0,
      "step": 42520
    },
    {
      "entropy": 5.595686197280884,
      "epoch": 3.8574927431059507,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003577281352465241,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22959115207195283,
      "num_tokens": 88260126.0,
      "step": 42525
    },
    {
      "entropy": 5.572345542907715,
      "epoch": 3.857946298984035,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035769804446949177,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22181571125984192,
      "num_tokens": 88271492.0,
      "step": 42530
    },
    {
      "entropy": 5.691086673736573,
      "epoch": 3.8583998548621192,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003576679519822428,
      "loss": 4.969,
      "mean_token_accuracy": 0.21242373585700988,
      "num_tokens": 88281966.0,
      "step": 42535
    },
    {
      "entropy": 5.581513977050781,
      "epoch": 3.858853410740203,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035763785778539963,
      "loss": 4.7902,
      "mean_token_accuracy": 0.23529008626937867,
      "num_tokens": 88292805.0,
      "step": 42540
    },
    {
      "entropy": 5.567488479614258,
      "epoch": 3.8593069666182873,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003576077618795844,
      "loss": 4.8324,
      "mean_token_accuracy": 0.2294877603650093,
      "num_tokens": 88302727.0,
      "step": 42545
    },
    {
      "entropy": 5.4796217918396,
      "epoch": 3.8597605224963716,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003575776642654196,
      "loss": 4.8388,
      "mean_token_accuracy": 0.22537384182214737,
      "num_tokens": 88313150.0,
      "step": 42550
    },
    {
      "entropy": 5.578943204879761,
      "epoch": 3.860214078374456,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035754756494352767,
      "loss": 4.8207,
      "mean_token_accuracy": 0.2296580821275711,
      "num_tokens": 88323060.0,
      "step": 42555
    },
    {
      "entropy": 5.683638620376587,
      "epoch": 3.8606676342525397,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003575174639145311,
      "loss": 4.882,
      "mean_token_accuracy": 0.22596658617258072,
      "num_tokens": 88332641.0,
      "step": 42560
    },
    {
      "entropy": 5.577805376052856,
      "epoch": 3.861121190130624,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003574873611790523,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22531621903181076,
      "num_tokens": 88342807.0,
      "step": 42565
    },
    {
      "entropy": 5.518455886840821,
      "epoch": 3.861574746008708,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035745725673771387,
      "loss": 4.7744,
      "mean_token_accuracy": 0.22867559641599655,
      "num_tokens": 88352958.0,
      "step": 42570
    },
    {
      "entropy": 5.647450113296509,
      "epoch": 3.8620283018867925,
      "grad_norm": 1.25,
      "learning_rate": 0.00035742715059113836,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2163737565279007,
      "num_tokens": 88363184.0,
      "step": 42575
    },
    {
      "entropy": 5.539021968841553,
      "epoch": 3.8624818577648767,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003573970427399484,
      "loss": 4.8186,
      "mean_token_accuracy": 0.23205912560224534,
      "num_tokens": 88373991.0,
      "step": 42580
    },
    {
      "entropy": 5.643276977539062,
      "epoch": 3.862935413642961,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035736693318476646,
      "loss": 4.8984,
      "mean_token_accuracy": 0.21713860630989074,
      "num_tokens": 88384518.0,
      "step": 42585
    },
    {
      "entropy": 5.520178174972534,
      "epoch": 3.8633889695210453,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035733682192621555,
      "loss": 4.7745,
      "mean_token_accuracy": 0.22728700935840607,
      "num_tokens": 88394506.0,
      "step": 42590
    },
    {
      "entropy": 5.602962684631348,
      "epoch": 3.863842525399129,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003573067089649181,
      "loss": 4.8815,
      "mean_token_accuracy": 0.22077593356370925,
      "num_tokens": 88404625.0,
      "step": 42595
    },
    {
      "entropy": 5.618331527709961,
      "epoch": 3.8642960812772134,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035727659430149686,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2227043554186821,
      "num_tokens": 88413754.0,
      "step": 42600
    },
    {
      "entropy": 5.689132881164551,
      "epoch": 3.8647496371552976,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003572464779365748,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21008697301149368,
      "num_tokens": 88424613.0,
      "step": 42605
    },
    {
      "entropy": 5.630249786376953,
      "epoch": 3.8652031930333814,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035721635987077455,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22755245864391327,
      "num_tokens": 88435248.0,
      "step": 42610
    },
    {
      "entropy": 5.574333429336548,
      "epoch": 3.8656567489114657,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035718624010471904,
      "loss": 4.7617,
      "mean_token_accuracy": 0.23457664400339126,
      "num_tokens": 88444802.0,
      "step": 42615
    },
    {
      "entropy": 5.616395473480225,
      "epoch": 3.86611030478955,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003571561186390311,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21520706862211228,
      "num_tokens": 88455335.0,
      "step": 42620
    },
    {
      "entropy": 5.671471786499024,
      "epoch": 3.8665638606676342,
      "grad_norm": 1.25,
      "learning_rate": 0.0003571259954743338,
      "loss": 4.9188,
      "mean_token_accuracy": 0.2242258682847023,
      "num_tokens": 88465857.0,
      "step": 42625
    },
    {
      "entropy": 5.633995532989502,
      "epoch": 3.8670174165457185,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035709587061124985,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21358892172574998,
      "num_tokens": 88475930.0,
      "step": 42630
    },
    {
      "entropy": 5.6251749992370605,
      "epoch": 3.8674709724238028,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035706574405040245,
      "loss": 4.8879,
      "mean_token_accuracy": 0.225510673224926,
      "num_tokens": 88486732.0,
      "step": 42635
    },
    {
      "entropy": 5.623756456375122,
      "epoch": 3.867924528301887,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003570356157924145,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22028617560863495,
      "num_tokens": 88497017.0,
      "step": 42640
    },
    {
      "entropy": 5.591638708114624,
      "epoch": 3.868378084179971,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003570054858379092,
      "loss": 4.7742,
      "mean_token_accuracy": 0.2406635031104088,
      "num_tokens": 88507423.0,
      "step": 42645
    },
    {
      "entropy": 5.601606988906861,
      "epoch": 3.868831640058055,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035697535418750955,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2259859561920166,
      "num_tokens": 88517795.0,
      "step": 42650
    },
    {
      "entropy": 5.49961256980896,
      "epoch": 3.8692851959361394,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035694522084183864,
      "loss": 4.7821,
      "mean_token_accuracy": 0.22595662176609038,
      "num_tokens": 88527641.0,
      "step": 42655
    },
    {
      "entropy": 5.601541566848755,
      "epoch": 3.8697387518142237,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00035691508580151966,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21696866750717164,
      "num_tokens": 88537874.0,
      "step": 42660
    },
    {
      "entropy": 5.632071924209595,
      "epoch": 3.8701923076923075,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003568849490671759,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21993334740400314,
      "num_tokens": 88548108.0,
      "step": 42665
    },
    {
      "entropy": 5.547215557098388,
      "epoch": 3.8706458635703918,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003568548106394305,
      "loss": 4.7737,
      "mean_token_accuracy": 0.22869363576173782,
      "num_tokens": 88558284.0,
      "step": 42670
    },
    {
      "entropy": 5.596485996246338,
      "epoch": 3.871099419448476,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003568246705189067,
      "loss": 4.902,
      "mean_token_accuracy": 0.21695830523967743,
      "num_tokens": 88568635.0,
      "step": 42675
    },
    {
      "entropy": 5.612693834304809,
      "epoch": 3.8715529753265603,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003567945287062279,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22160447984933854,
      "num_tokens": 88577968.0,
      "step": 42680
    },
    {
      "entropy": 5.68268346786499,
      "epoch": 3.8720065312046446,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035676438520201745,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21818131506443023,
      "num_tokens": 88590562.0,
      "step": 42685
    },
    {
      "entropy": 5.59676947593689,
      "epoch": 3.872460087082729,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035673424000689857,
      "loss": 4.907,
      "mean_token_accuracy": 0.21371458768844603,
      "num_tokens": 88601187.0,
      "step": 42690
    },
    {
      "entropy": 5.54228458404541,
      "epoch": 3.8729136429608126,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035670409312149495,
      "loss": 4.7973,
      "mean_token_accuracy": 0.22271920442581178,
      "num_tokens": 88611459.0,
      "step": 42695
    },
    {
      "entropy": 5.61933388710022,
      "epoch": 3.873367198838897,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035667394454642963,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2229720264673233,
      "num_tokens": 88621652.0,
      "step": 42700
    },
    {
      "entropy": 5.608958053588867,
      "epoch": 3.873820754716981,
      "grad_norm": 1.25,
      "learning_rate": 0.0003566437942823265,
      "loss": 4.7675,
      "mean_token_accuracy": 0.22596181631088258,
      "num_tokens": 88632425.0,
      "step": 42705
    },
    {
      "entropy": 5.549476957321167,
      "epoch": 3.8742743105950654,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035661364232980884,
      "loss": 4.8,
      "mean_token_accuracy": 0.2315898969769478,
      "num_tokens": 88641263.0,
      "step": 42710
    },
    {
      "entropy": 5.544644546508789,
      "epoch": 3.8747278664731493,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035658348868950027,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2201856940984726,
      "num_tokens": 88651665.0,
      "step": 42715
    },
    {
      "entropy": 5.582266569137573,
      "epoch": 3.8751814223512335,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035655333336202436,
      "loss": 4.819,
      "mean_token_accuracy": 0.22780657410621644,
      "num_tokens": 88660791.0,
      "step": 42720
    },
    {
      "entropy": 5.523458242416382,
      "epoch": 3.875634978229318,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003565231763480047,
      "loss": 4.8035,
      "mean_token_accuracy": 0.23371667712926864,
      "num_tokens": 88671341.0,
      "step": 42725
    },
    {
      "entropy": 5.58234486579895,
      "epoch": 3.876088534107402,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003564930176480651,
      "loss": 4.8736,
      "mean_token_accuracy": 0.21932352483272552,
      "num_tokens": 88682500.0,
      "step": 42730
    },
    {
      "entropy": 5.628639030456543,
      "epoch": 3.8765420899854863,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035646285726282896,
      "loss": 4.885,
      "mean_token_accuracy": 0.21412044316530227,
      "num_tokens": 88693513.0,
      "step": 42735
    },
    {
      "entropy": 5.654679775238037,
      "epoch": 3.8769956458635706,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035643269519292027,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21985863596200944,
      "num_tokens": 88703498.0,
      "step": 42740
    },
    {
      "entropy": 5.5956048488616945,
      "epoch": 3.8774492017416544,
      "grad_norm": 1.25,
      "learning_rate": 0.0003564025314389627,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2211844354867935,
      "num_tokens": 88714217.0,
      "step": 42745
    },
    {
      "entropy": 5.601555681228637,
      "epoch": 3.8779027576197387,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035637236600157994,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2154860958456993,
      "num_tokens": 88725006.0,
      "step": 42750
    },
    {
      "entropy": 5.568734645843506,
      "epoch": 3.878356313497823,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000356342198881396,
      "loss": 4.8159,
      "mean_token_accuracy": 0.2271187946200371,
      "num_tokens": 88735253.0,
      "step": 42755
    },
    {
      "entropy": 5.623343515396118,
      "epoch": 3.878809869375907,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035631203007903465,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22137145102024078,
      "num_tokens": 88746313.0,
      "step": 42760
    },
    {
      "entropy": 5.581971788406372,
      "epoch": 3.879263425253991,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003562818595951198,
      "loss": 4.8289,
      "mean_token_accuracy": 0.23292960375547409,
      "num_tokens": 88757448.0,
      "step": 42765
    },
    {
      "entropy": 5.611841106414795,
      "epoch": 3.8797169811320753,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003562516874302754,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22864297330379485,
      "num_tokens": 88767426.0,
      "step": 42770
    },
    {
      "entropy": 5.600691318511963,
      "epoch": 3.8801705370101596,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003562215135851253,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2213900551199913,
      "num_tokens": 88778188.0,
      "step": 42775
    },
    {
      "entropy": 5.686941814422608,
      "epoch": 3.880624092888244,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035619133806029374,
      "loss": 4.9315,
      "mean_token_accuracy": 0.22943966537714006,
      "num_tokens": 88788927.0,
      "step": 42780
    },
    {
      "entropy": 5.673157167434693,
      "epoch": 3.881077648766328,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035616116085640455,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21913135945796966,
      "num_tokens": 88798809.0,
      "step": 42785
    },
    {
      "entropy": 5.476976585388184,
      "epoch": 3.8815312046444124,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003561309819740819,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23172022849321366,
      "num_tokens": 88808688.0,
      "step": 42790
    },
    {
      "entropy": 5.565728044509887,
      "epoch": 3.8819847605224966,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035610080141394983,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22325649112462997,
      "num_tokens": 88818400.0,
      "step": 42795
    },
    {
      "entropy": 5.569725275039673,
      "epoch": 3.8824383164005805,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035607061917663254,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22431000024080278,
      "num_tokens": 88828032.0,
      "step": 42800
    },
    {
      "entropy": 5.621572113037109,
      "epoch": 3.8828918722786647,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035604043526275413,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22619543671607972,
      "num_tokens": 88839856.0,
      "step": 42805
    },
    {
      "entropy": 5.619917345046997,
      "epoch": 3.883345428156749,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000356010249672939,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2292220786213875,
      "num_tokens": 88851016.0,
      "step": 42810
    },
    {
      "entropy": 5.585749530792237,
      "epoch": 3.8837989840348333,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003559800624078111,
      "loss": 4.9095,
      "mean_token_accuracy": 0.22026924043893814,
      "num_tokens": 88859795.0,
      "step": 42815
    },
    {
      "entropy": 5.582521104812622,
      "epoch": 3.884252539912917,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003559498734679949,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22029165923595428,
      "num_tokens": 88869595.0,
      "step": 42820
    },
    {
      "entropy": 5.6647968769073485,
      "epoch": 3.8847060957910013,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003559196828541148,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21099997609853743,
      "num_tokens": 88880857.0,
      "step": 42825
    },
    {
      "entropy": 5.575960636138916,
      "epoch": 3.8851596516690856,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000355889490566795,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22916074842214584,
      "num_tokens": 88890687.0,
      "step": 42830
    },
    {
      "entropy": 5.610993528366089,
      "epoch": 3.88561320754717,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035585929660665984,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22015308290719987,
      "num_tokens": 88899600.0,
      "step": 42835
    },
    {
      "entropy": 5.622123050689697,
      "epoch": 3.886066763425254,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003558291009743339,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22689017057418823,
      "num_tokens": 88909521.0,
      "step": 42840
    },
    {
      "entropy": 5.6108520984649655,
      "epoch": 3.8865203193033384,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00035579890367044153,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2173363208770752,
      "num_tokens": 88920061.0,
      "step": 42845
    },
    {
      "entropy": 5.547650098800659,
      "epoch": 3.8869738751814222,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003557687046956073,
      "loss": 4.7846,
      "mean_token_accuracy": 0.22666936963796616,
      "num_tokens": 88929504.0,
      "step": 42850
    },
    {
      "entropy": 5.693078422546387,
      "epoch": 3.8874274310595065,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035573850405045554,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20609140247106553,
      "num_tokens": 88940653.0,
      "step": 42855
    },
    {
      "entropy": 5.657821607589722,
      "epoch": 3.8878809869375908,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003557083017356111,
      "loss": 4.951,
      "mean_token_accuracy": 0.22172697633504868,
      "num_tokens": 88952133.0,
      "step": 42860
    },
    {
      "entropy": 5.539954137802124,
      "epoch": 3.888334542815675,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003556780977516983,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2256540283560753,
      "num_tokens": 88963190.0,
      "step": 42865
    },
    {
      "entropy": 5.483289194107056,
      "epoch": 3.888788098693759,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035564789209934183,
      "loss": 4.7792,
      "mean_token_accuracy": 0.2274247869849205,
      "num_tokens": 88973034.0,
      "step": 42870
    },
    {
      "entropy": 5.570822381973267,
      "epoch": 3.889241654571843,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035561768477916655,
      "loss": 4.7845,
      "mean_token_accuracy": 0.23778816014528276,
      "num_tokens": 88982809.0,
      "step": 42875
    },
    {
      "entropy": 5.594953203201294,
      "epoch": 3.8896952104499274,
      "grad_norm": 1.1875,
      "learning_rate": 0.000355587475791797,
      "loss": 4.8642,
      "mean_token_accuracy": 0.21591373682022094,
      "num_tokens": 88993563.0,
      "step": 42880
    },
    {
      "entropy": 5.526847839355469,
      "epoch": 3.8901487663280117,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035555726513785785,
      "loss": 4.7875,
      "mean_token_accuracy": 0.22998472154140473,
      "num_tokens": 89002408.0,
      "step": 42885
    },
    {
      "entropy": 5.555732107162475,
      "epoch": 3.890602322206096,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003555270528179739,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22390047013759612,
      "num_tokens": 89013060.0,
      "step": 42890
    },
    {
      "entropy": 5.658052062988281,
      "epoch": 3.89105587808418,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035549683883277,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22229683250188828,
      "num_tokens": 89023406.0,
      "step": 42895
    },
    {
      "entropy": 5.638285827636719,
      "epoch": 3.891509433962264,
      "grad_norm": 1.296875,
      "learning_rate": 0.000355466623182871,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2270367130637169,
      "num_tokens": 89033800.0,
      "step": 42900
    },
    {
      "entropy": 5.686974906921387,
      "epoch": 3.8919629898403483,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003554364058689017,
      "loss": 4.8898,
      "mean_token_accuracy": 0.21828233152627946,
      "num_tokens": 89044601.0,
      "step": 42905
    },
    {
      "entropy": 5.666297817230225,
      "epoch": 3.8924165457184325,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000355406186891487,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21877190917730333,
      "num_tokens": 89056138.0,
      "step": 42910
    },
    {
      "entropy": 5.594916248321534,
      "epoch": 3.892870101596517,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003553759662512519,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22407578825950622,
      "num_tokens": 89065776.0,
      "step": 42915
    },
    {
      "entropy": 5.749632835388184,
      "epoch": 3.8933236574746006,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035534574394882135,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20963463932275772,
      "num_tokens": 89076926.0,
      "step": 42920
    },
    {
      "entropy": 5.625717687606811,
      "epoch": 3.893777213352685,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003553155199848203,
      "loss": 4.8919,
      "mean_token_accuracy": 0.21228738278150558,
      "num_tokens": 89088036.0,
      "step": 42925
    },
    {
      "entropy": 5.573708152770996,
      "epoch": 3.894230769230769,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003552852943598739,
      "loss": 4.8444,
      "mean_token_accuracy": 0.22454096823930741,
      "num_tokens": 89098606.0,
      "step": 42930
    },
    {
      "entropy": 5.560185337066651,
      "epoch": 3.8946843251088534,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000355255067074607,
      "loss": 4.812,
      "mean_token_accuracy": 0.2280314013361931,
      "num_tokens": 89107733.0,
      "step": 42935
    },
    {
      "entropy": 5.562902212142944,
      "epoch": 3.8951378809869377,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035522483812964496,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22810633778572081,
      "num_tokens": 89119094.0,
      "step": 42940
    },
    {
      "entropy": 5.572535228729248,
      "epoch": 3.895591436865022,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035519460752561285,
      "loss": 4.7463,
      "mean_token_accuracy": 0.22831853181123735,
      "num_tokens": 89129254.0,
      "step": 42945
    },
    {
      "entropy": 5.6140917301177975,
      "epoch": 3.8960449927431062,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003551643752631358,
      "loss": 4.8151,
      "mean_token_accuracy": 0.2297599047422409,
      "num_tokens": 89139609.0,
      "step": 42950
    },
    {
      "entropy": 5.611825609207154,
      "epoch": 3.89649854862119,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003551341413428391,
      "loss": 4.8528,
      "mean_token_accuracy": 0.2127300500869751,
      "num_tokens": 89149504.0,
      "step": 42955
    },
    {
      "entropy": 5.58812403678894,
      "epoch": 3.8969521044992743,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003551039057653478,
      "loss": 4.8525,
      "mean_token_accuracy": 0.22652725428342818,
      "num_tokens": 89160036.0,
      "step": 42960
    },
    {
      "entropy": 5.605102396011352,
      "epoch": 3.8974056603773586,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003550736685312874,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21726494580507277,
      "num_tokens": 89170960.0,
      "step": 42965
    },
    {
      "entropy": 5.591526651382447,
      "epoch": 3.8978592162554424,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003550434296412831,
      "loss": 4.8343,
      "mean_token_accuracy": 0.2309944897890091,
      "num_tokens": 89179630.0,
      "step": 42970
    },
    {
      "entropy": 5.619865083694458,
      "epoch": 3.8983127721335267,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003550131890959603,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22770145535469055,
      "num_tokens": 89188880.0,
      "step": 42975
    },
    {
      "entropy": 5.657916927337647,
      "epoch": 3.898766328011611,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035498294689594435,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22078754156827926,
      "num_tokens": 89199585.0,
      "step": 42980
    },
    {
      "entropy": 5.610892724990845,
      "epoch": 3.899219883889695,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035495270304186074,
      "loss": 4.8557,
      "mean_token_accuracy": 0.21994916200637818,
      "num_tokens": 89209812.0,
      "step": 42985
    },
    {
      "entropy": 5.582829093933105,
      "epoch": 3.8996734397677795,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003549224575343347,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22428322434425355,
      "num_tokens": 89220244.0,
      "step": 42990
    },
    {
      "entropy": 5.575133275985718,
      "epoch": 3.9001269956458637,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035489221037399203,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2173730969429016,
      "num_tokens": 89230618.0,
      "step": 42995
    },
    {
      "entropy": 5.6275060176849365,
      "epoch": 3.900580551523948,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035486196156145805,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22284162044525146,
      "num_tokens": 89242041.0,
      "step": 43000
    },
    {
      "entropy": 5.6372627258300785,
      "epoch": 3.901034107402032,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003548317110973583,
      "loss": 4.8987,
      "mean_token_accuracy": 0.2224171355366707,
      "num_tokens": 89252817.0,
      "step": 43005
    },
    {
      "entropy": 5.6234947681427006,
      "epoch": 3.901487663280116,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003548014589823185,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22318633198738097,
      "num_tokens": 89263702.0,
      "step": 43010
    },
    {
      "entropy": 5.589557647705078,
      "epoch": 3.9019412191582004,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035477120521696417,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2276522174477577,
      "num_tokens": 89273822.0,
      "step": 43015
    },
    {
      "entropy": 5.666953277587891,
      "epoch": 3.9023947750362846,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003547409498019209,
      "loss": 4.9919,
      "mean_token_accuracy": 0.2130616530776024,
      "num_tokens": 89283973.0,
      "step": 43020
    },
    {
      "entropy": 5.644901943206787,
      "epoch": 3.9028483309143684,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035471069273781454,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22813804000616072,
      "num_tokens": 89294059.0,
      "step": 43025
    },
    {
      "entropy": 5.612426280975342,
      "epoch": 3.9033018867924527,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003546804340252707,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22520266473293304,
      "num_tokens": 89304049.0,
      "step": 43030
    },
    {
      "entropy": 5.539799785614013,
      "epoch": 3.903755442670537,
      "grad_norm": 1.375,
      "learning_rate": 0.00035465017366491527,
      "loss": 4.7644,
      "mean_token_accuracy": 0.23296934813261033,
      "num_tokens": 89314468.0,
      "step": 43035
    },
    {
      "entropy": 5.536766910552979,
      "epoch": 3.9042089985486212,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035461991165737386,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22753619253635407,
      "num_tokens": 89324580.0,
      "step": 43040
    },
    {
      "entropy": 5.684184741973877,
      "epoch": 3.9046625544267055,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035458964800327245,
      "loss": 4.9416,
      "mean_token_accuracy": 0.2114985004067421,
      "num_tokens": 89334976.0,
      "step": 43045
    },
    {
      "entropy": 5.650997495651245,
      "epoch": 3.90511611030479,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035455938270323684,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22097509503364562,
      "num_tokens": 89345210.0,
      "step": 43050
    },
    {
      "entropy": 5.56312198638916,
      "epoch": 3.9055696661828736,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003545291157578928,
      "loss": 4.7676,
      "mean_token_accuracy": 0.2340549498796463,
      "num_tokens": 89354623.0,
      "step": 43055
    },
    {
      "entropy": 5.564368391036988,
      "epoch": 3.906023222060958,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035449884716786653,
      "loss": 4.8374,
      "mean_token_accuracy": 0.22510685473680497,
      "num_tokens": 89364684.0,
      "step": 43060
    },
    {
      "entropy": 5.591606330871582,
      "epoch": 3.906476777939042,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035446857693378377,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21740567088127136,
      "num_tokens": 89375415.0,
      "step": 43065
    },
    {
      "entropy": 5.642599248886109,
      "epoch": 3.9069303338171264,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003544383050562706,
      "loss": 4.9515,
      "mean_token_accuracy": 0.22354326099157334,
      "num_tokens": 89385791.0,
      "step": 43070
    },
    {
      "entropy": 5.660962057113648,
      "epoch": 3.90738388969521,
      "grad_norm": 1.296875,
      "learning_rate": 0.000354408031535953,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21786124259233475,
      "num_tokens": 89395345.0,
      "step": 43075
    },
    {
      "entropy": 5.597086715698242,
      "epoch": 3.9078374455732945,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035437775637345715,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22820939719676972,
      "num_tokens": 89405413.0,
      "step": 43080
    },
    {
      "entropy": 5.61079158782959,
      "epoch": 3.9082910014513788,
      "grad_norm": 1.421875,
      "learning_rate": 0.00035434747956940906,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22560741007328033,
      "num_tokens": 89415168.0,
      "step": 43085
    },
    {
      "entropy": 5.5550049304962155,
      "epoch": 3.908744557329463,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035431720112443475,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22644688338041305,
      "num_tokens": 89426438.0,
      "step": 43090
    },
    {
      "entropy": 5.56920862197876,
      "epoch": 3.9091981132075473,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035428692103916067,
      "loss": 4.7966,
      "mean_token_accuracy": 0.22562383860349655,
      "num_tokens": 89437454.0,
      "step": 43095
    },
    {
      "entropy": 5.634692764282226,
      "epoch": 3.9096516690856316,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003542566393142128,
      "loss": 4.8437,
      "mean_token_accuracy": 0.22463493496179582,
      "num_tokens": 89448082.0,
      "step": 43100
    },
    {
      "entropy": 5.540117311477661,
      "epoch": 3.9101052249637154,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035422635595021745,
      "loss": 4.8047,
      "mean_token_accuracy": 0.22532156109809875,
      "num_tokens": 89458227.0,
      "step": 43105
    },
    {
      "entropy": 5.570941543579101,
      "epoch": 3.9105587808417996,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035419607094780084,
      "loss": 4.8766,
      "mean_token_accuracy": 0.21982712745666505,
      "num_tokens": 89468374.0,
      "step": 43110
    },
    {
      "entropy": 5.603132629394532,
      "epoch": 3.911012336719884,
      "grad_norm": 1.421875,
      "learning_rate": 0.00035416578430758934,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22127149403095245,
      "num_tokens": 89477728.0,
      "step": 43115
    },
    {
      "entropy": 5.648571634292603,
      "epoch": 3.911465892597968,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035413549603020915,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22545821815729142,
      "num_tokens": 89487107.0,
      "step": 43120
    },
    {
      "entropy": 5.643214702606201,
      "epoch": 3.911919448476052,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035410520611628676,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2316722571849823,
      "num_tokens": 89497075.0,
      "step": 43125
    },
    {
      "entropy": 5.498446226119995,
      "epoch": 3.9123730043541363,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003540749145664486,
      "loss": 4.8233,
      "mean_token_accuracy": 0.22406744360923767,
      "num_tokens": 89508068.0,
      "step": 43130
    },
    {
      "entropy": 5.584011602401733,
      "epoch": 3.9128265602322205,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003540446213813211,
      "loss": 4.8085,
      "mean_token_accuracy": 0.23416467607021332,
      "num_tokens": 89518371.0,
      "step": 43135
    },
    {
      "entropy": 5.592210197448731,
      "epoch": 3.913280116110305,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035401432656153057,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22430134266614915,
      "num_tokens": 89529024.0,
      "step": 43140
    },
    {
      "entropy": 5.5411238193511965,
      "epoch": 3.913733671988389,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035398403010770377,
      "loss": 4.761,
      "mean_token_accuracy": 0.2330788031220436,
      "num_tokens": 89540331.0,
      "step": 43145
    },
    {
      "entropy": 5.501801633834839,
      "epoch": 3.9141872278664733,
      "grad_norm": 1.34375,
      "learning_rate": 0.000353953732020467,
      "loss": 4.7783,
      "mean_token_accuracy": 0.23395952433347703,
      "num_tokens": 89550385.0,
      "step": 43150
    },
    {
      "entropy": 5.620708036422729,
      "epoch": 3.9146407837445576,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035392343230044687,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2191256195306778,
      "num_tokens": 89561480.0,
      "step": 43155
    },
    {
      "entropy": 5.612206840515137,
      "epoch": 3.9150943396226414,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035389313094827017,
      "loss": 4.8586,
      "mean_token_accuracy": 0.21825949102640152,
      "num_tokens": 89571897.0,
      "step": 43160
    },
    {
      "entropy": 5.62544960975647,
      "epoch": 3.9155478955007257,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035386282796456343,
      "loss": 4.8826,
      "mean_token_accuracy": 0.22333759516477586,
      "num_tokens": 89582019.0,
      "step": 43165
    },
    {
      "entropy": 5.599987411499024,
      "epoch": 3.91600145137881,
      "grad_norm": 1.375,
      "learning_rate": 0.0003538325233499532,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22556322813034058,
      "num_tokens": 89591806.0,
      "step": 43170
    },
    {
      "entropy": 5.566720676422119,
      "epoch": 3.9164550072568938,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035380221710506636,
      "loss": 4.8526,
      "mean_token_accuracy": 0.22628452032804489,
      "num_tokens": 89601845.0,
      "step": 43175
    },
    {
      "entropy": 5.627076959609985,
      "epoch": 3.916908563134978,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035377190923052964,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22301062643527986,
      "num_tokens": 89612331.0,
      "step": 43180
    },
    {
      "entropy": 5.6248725891113285,
      "epoch": 3.9173621190130623,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035374159972696966,
      "loss": 4.9231,
      "mean_token_accuracy": 0.22055679112672805,
      "num_tokens": 89622631.0,
      "step": 43185
    },
    {
      "entropy": 5.614597463607788,
      "epoch": 3.9178156748911466,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003537112885950133,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21597190350294113,
      "num_tokens": 89633065.0,
      "step": 43190
    },
    {
      "entropy": 5.6464110851287845,
      "epoch": 3.918269230769231,
      "grad_norm": 1.375,
      "learning_rate": 0.00035368097583528756,
      "loss": 4.901,
      "mean_token_accuracy": 0.22663470059633256,
      "num_tokens": 89643451.0,
      "step": 43195
    },
    {
      "entropy": 5.7089378356933596,
      "epoch": 3.918722786647315,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003536506614484192,
      "loss": 4.9347,
      "mean_token_accuracy": 0.2236470028758049,
      "num_tokens": 89654786.0,
      "step": 43200
    },
    {
      "entropy": 5.659652137756348,
      "epoch": 3.9191763425253994,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003536203454350351,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2184249758720398,
      "num_tokens": 89665459.0,
      "step": 43205
    },
    {
      "entropy": 5.573454332351685,
      "epoch": 3.919629898403483,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003535900277957621,
      "loss": 4.8664,
      "mean_token_accuracy": 0.22049006521701814,
      "num_tokens": 89675149.0,
      "step": 43210
    },
    {
      "entropy": 5.593321323394775,
      "epoch": 3.9200834542815675,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035355970853122743,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22758912444114685,
      "num_tokens": 89684994.0,
      "step": 43215
    },
    {
      "entropy": 5.693453598022461,
      "epoch": 3.9205370101596517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035352938764205793,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20932717621326447,
      "num_tokens": 89694751.0,
      "step": 43220
    },
    {
      "entropy": 5.626496887207031,
      "epoch": 3.920990566037736,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003534990651288807,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22037066370248795,
      "num_tokens": 89704976.0,
      "step": 43225
    },
    {
      "entropy": 5.643860149383545,
      "epoch": 3.92144412191582,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035346874099232274,
      "loss": 4.9573,
      "mean_token_accuracy": 0.2149613082408905,
      "num_tokens": 89715863.0,
      "step": 43230
    },
    {
      "entropy": 5.603780126571655,
      "epoch": 3.921897677793904,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035343841523301125,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2233002558350563,
      "num_tokens": 89725856.0,
      "step": 43235
    },
    {
      "entropy": 5.630465793609619,
      "epoch": 3.9223512336719883,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003534080878515733,
      "loss": 4.9236,
      "mean_token_accuracy": 0.211833655834198,
      "num_tokens": 89736345.0,
      "step": 43240
    },
    {
      "entropy": 5.666217374801636,
      "epoch": 3.9228047895500726,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003533777588486362,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22685311287641524,
      "num_tokens": 89746960.0,
      "step": 43245
    },
    {
      "entropy": 5.605423307418823,
      "epoch": 3.923258345428157,
      "grad_norm": 1.265625,
      "learning_rate": 0.000353347428224827,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2177856221795082,
      "num_tokens": 89757373.0,
      "step": 43250
    },
    {
      "entropy": 5.629535055160522,
      "epoch": 3.923711901306241,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035331709598077304,
      "loss": 4.891,
      "mean_token_accuracy": 0.22138093113899232,
      "num_tokens": 89767895.0,
      "step": 43255
    },
    {
      "entropy": 5.629093265533447,
      "epoch": 3.924165457184325,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003532867621171016,
      "loss": 4.8588,
      "mean_token_accuracy": 0.2249281495809555,
      "num_tokens": 89777512.0,
      "step": 43260
    },
    {
      "entropy": 5.617240476608276,
      "epoch": 3.9246190130624092,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035325642663443994,
      "loss": 4.8926,
      "mean_token_accuracy": 0.2241021066904068,
      "num_tokens": 89787308.0,
      "step": 43265
    },
    {
      "entropy": 5.626300096511841,
      "epoch": 3.9250725689404935,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003532260895334155,
      "loss": 4.9148,
      "mean_token_accuracy": 0.2241753175854683,
      "num_tokens": 89798384.0,
      "step": 43270
    },
    {
      "entropy": 5.663939094543457,
      "epoch": 3.9255261248185778,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035319575081465556,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22019025534391404,
      "num_tokens": 89808757.0,
      "step": 43275
    },
    {
      "entropy": 5.652631711959839,
      "epoch": 3.9259796806966616,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035316541047878753,
      "loss": 4.9014,
      "mean_token_accuracy": 0.21873293519020082,
      "num_tokens": 89818925.0,
      "step": 43280
    },
    {
      "entropy": 5.674756193161011,
      "epoch": 3.926433236574746,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003531350685264389,
      "loss": 4.9429,
      "mean_token_accuracy": 0.223711359500885,
      "num_tokens": 89828647.0,
      "step": 43285
    },
    {
      "entropy": 5.71081509590149,
      "epoch": 3.92688679245283,
      "grad_norm": 1.25,
      "learning_rate": 0.00035310472495823713,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21380311399698257,
      "num_tokens": 89838784.0,
      "step": 43290
    },
    {
      "entropy": 5.616066265106201,
      "epoch": 3.9273403483309144,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003530743797748098,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22013625949621202,
      "num_tokens": 89848464.0,
      "step": 43295
    },
    {
      "entropy": 5.553391647338867,
      "epoch": 3.9277939042089987,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003530440329767844,
      "loss": 4.852,
      "mean_token_accuracy": 0.21535703986883165,
      "num_tokens": 89859594.0,
      "step": 43300
    },
    {
      "entropy": 5.585629415512085,
      "epoch": 3.928247460087083,
      "grad_norm": 1.25,
      "learning_rate": 0.0003530136845647885,
      "loss": 4.8314,
      "mean_token_accuracy": 0.22574584782123566,
      "num_tokens": 89869869.0,
      "step": 43305
    },
    {
      "entropy": 5.662320613861084,
      "epoch": 3.9287010159651667,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035298333453944976,
      "loss": 4.935,
      "mean_token_accuracy": 0.22433721423149108,
      "num_tokens": 89879604.0,
      "step": 43310
    },
    {
      "entropy": 5.520637559890747,
      "epoch": 3.929154571843251,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003529529829013957,
      "loss": 4.752,
      "mean_token_accuracy": 0.22736752331256865,
      "num_tokens": 89889200.0,
      "step": 43315
    },
    {
      "entropy": 5.616128206253052,
      "epoch": 3.9296081277213353,
      "grad_norm": 1.25,
      "learning_rate": 0.0003529226296512541,
      "loss": 4.8919,
      "mean_token_accuracy": 0.2294008955359459,
      "num_tokens": 89898525.0,
      "step": 43320
    },
    {
      "entropy": 5.56722674369812,
      "epoch": 3.9300616835994195,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035289227478965267,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21578328162431717,
      "num_tokens": 89909053.0,
      "step": 43325
    },
    {
      "entropy": 5.534801054000854,
      "epoch": 3.9305152394775034,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003528619183172192,
      "loss": 4.8041,
      "mean_token_accuracy": 0.23143748342990875,
      "num_tokens": 89918375.0,
      "step": 43330
    },
    {
      "entropy": 5.692203330993652,
      "epoch": 3.9309687953555876,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003528315602345814,
      "loss": 4.9413,
      "mean_token_accuracy": 0.22652362287044525,
      "num_tokens": 89928252.0,
      "step": 43335
    },
    {
      "entropy": 5.590328788757324,
      "epoch": 3.931422351233672,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003528012005423671,
      "loss": 4.8659,
      "mean_token_accuracy": 0.21630339622497557,
      "num_tokens": 89938613.0,
      "step": 43340
    },
    {
      "entropy": 5.671624946594238,
      "epoch": 3.931875907111756,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035277083924120406,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22579684406518935,
      "num_tokens": 89948781.0,
      "step": 43345
    },
    {
      "entropy": 5.597576761245728,
      "epoch": 3.9323294629898404,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035274047633172037,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21653368771076204,
      "num_tokens": 89959928.0,
      "step": 43350
    },
    {
      "entropy": 5.660564517974853,
      "epoch": 3.9327830188679247,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003527101118145437,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21858150213956834,
      "num_tokens": 89970756.0,
      "step": 43355
    },
    {
      "entropy": 5.648901176452637,
      "epoch": 3.933236574746009,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003526797456903022,
      "loss": 4.9171,
      "mean_token_accuracy": 0.220272496342659,
      "num_tokens": 89981332.0,
      "step": 43360
    },
    {
      "entropy": 5.574952888488769,
      "epoch": 3.933690130624093,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035264937795962374,
      "loss": 4.8494,
      "mean_token_accuracy": 0.23076258301734925,
      "num_tokens": 89991427.0,
      "step": 43365
    },
    {
      "entropy": 5.6070410251617435,
      "epoch": 3.934143686502177,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003526190086231364,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22866371273994446,
      "num_tokens": 90001599.0,
      "step": 43370
    },
    {
      "entropy": 5.64363956451416,
      "epoch": 3.9345972423802613,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035258863768146816,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22396067082881926,
      "num_tokens": 90011932.0,
      "step": 43375
    },
    {
      "entropy": 5.650518226623535,
      "epoch": 3.9350507982583456,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003525582651352471,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22305236607789994,
      "num_tokens": 90022123.0,
      "step": 43380
    },
    {
      "entropy": 5.555824899673462,
      "epoch": 3.9355043541364294,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003525278909851013,
      "loss": 4.8177,
      "mean_token_accuracy": 0.22943018078804017,
      "num_tokens": 90033024.0,
      "step": 43385
    },
    {
      "entropy": 5.595191574096679,
      "epoch": 3.9359579100145137,
      "grad_norm": 1.25,
      "learning_rate": 0.000352497515231659,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22776588946580886,
      "num_tokens": 90043742.0,
      "step": 43390
    },
    {
      "entropy": 5.565085983276367,
      "epoch": 3.936411465892598,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003524671378755484,
      "loss": 4.8433,
      "mean_token_accuracy": 0.22127674371004105,
      "num_tokens": 90054508.0,
      "step": 43395
    },
    {
      "entropy": 5.605257177352906,
      "epoch": 3.936865021770682,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003524367589173976,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2281677857041359,
      "num_tokens": 90065058.0,
      "step": 43400
    },
    {
      "entropy": 5.578793859481811,
      "epoch": 3.9373185776487665,
      "grad_norm": 1.453125,
      "learning_rate": 0.00035240637835783485,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22612740099430084,
      "num_tokens": 90075813.0,
      "step": 43405
    },
    {
      "entropy": 5.642021417617798,
      "epoch": 3.9377721335268507,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035237599619748853,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2187727153301239,
      "num_tokens": 90087064.0,
      "step": 43410
    },
    {
      "entropy": 5.599623918533325,
      "epoch": 3.9382256894049346,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003523456124369868,
      "loss": 4.8188,
      "mean_token_accuracy": 0.2304860919713974,
      "num_tokens": 90096668.0,
      "step": 43415
    },
    {
      "entropy": 5.5289144039154055,
      "epoch": 3.938679245283019,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003523152270769582,
      "loss": 4.7487,
      "mean_token_accuracy": 0.23337374329566957,
      "num_tokens": 90105569.0,
      "step": 43420
    },
    {
      "entropy": 5.593250465393067,
      "epoch": 3.939132801161103,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000352284840118031,
      "loss": 4.9402,
      "mean_token_accuracy": 0.21771066933870314,
      "num_tokens": 90116040.0,
      "step": 43425
    },
    {
      "entropy": 5.646475744247437,
      "epoch": 3.9395863570391874,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003522544515608335,
      "loss": 4.9025,
      "mean_token_accuracy": 0.22183049470186234,
      "num_tokens": 90125929.0,
      "step": 43430
    },
    {
      "entropy": 5.740047788619995,
      "epoch": 3.940039912917271,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003522240614059944,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21673423200845718,
      "num_tokens": 90136251.0,
      "step": 43435
    },
    {
      "entropy": 5.603198289871216,
      "epoch": 3.9404934687953554,
      "grad_norm": 1.28125,
      "learning_rate": 0.000352193669654142,
      "loss": 4.8475,
      "mean_token_accuracy": 0.22729016989469528,
      "num_tokens": 90146480.0,
      "step": 43440
    },
    {
      "entropy": 5.598655843734742,
      "epoch": 3.9409470246734397,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035216327630590476,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22182249575853347,
      "num_tokens": 90157192.0,
      "step": 43445
    },
    {
      "entropy": 5.636499357223511,
      "epoch": 3.941400580551524,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035213288136191136,
      "loss": 4.8841,
      "mean_token_accuracy": 0.221979621052742,
      "num_tokens": 90167120.0,
      "step": 43450
    },
    {
      "entropy": 5.607434034347534,
      "epoch": 3.9418541364296082,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035210248482279034,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2237686723470688,
      "num_tokens": 90176276.0,
      "step": 43455
    },
    {
      "entropy": 5.585240411758423,
      "epoch": 3.9423076923076925,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035207208668917026,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22544508725404738,
      "num_tokens": 90187150.0,
      "step": 43460
    },
    {
      "entropy": 5.543673086166382,
      "epoch": 3.9427612481857763,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035204168696167975,
      "loss": 4.8106,
      "mean_token_accuracy": 0.21664329320192338,
      "num_tokens": 90198112.0,
      "step": 43465
    },
    {
      "entropy": 5.625420093536377,
      "epoch": 3.9432148040638606,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035201128564094753,
      "loss": 4.937,
      "mean_token_accuracy": 0.2220859482884407,
      "num_tokens": 90208940.0,
      "step": 43470
    },
    {
      "entropy": 5.62300295829773,
      "epoch": 3.943668359941945,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003519808827276023,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22518924474716187,
      "num_tokens": 90220073.0,
      "step": 43475
    },
    {
      "entropy": 5.6139250755310055,
      "epoch": 3.944121915820029,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035195047822227283,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21644351184368132,
      "num_tokens": 90230059.0,
      "step": 43480
    },
    {
      "entropy": 5.562581777572632,
      "epoch": 3.944575471698113,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035192007212558787,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22481199204921723,
      "num_tokens": 90240486.0,
      "step": 43485
    },
    {
      "entropy": 5.6876708507537845,
      "epoch": 3.9450290275761972,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003518896644381762,
      "loss": 4.9541,
      "mean_token_accuracy": 0.21668176651000975,
      "num_tokens": 90250014.0,
      "step": 43490
    },
    {
      "entropy": 5.610002088546753,
      "epoch": 3.9454825834542815,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003518592551606666,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2138165980577469,
      "num_tokens": 90261040.0,
      "step": 43495
    },
    {
      "entropy": 5.649934434890747,
      "epoch": 3.9459361393323658,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003518288442936881,
      "loss": 4.926,
      "mean_token_accuracy": 0.22308849394321442,
      "num_tokens": 90271243.0,
      "step": 43500
    },
    {
      "entropy": 5.58831000328064,
      "epoch": 3.94638969521045,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003517984318378695,
      "loss": 4.8798,
      "mean_token_accuracy": 0.2233734294772148,
      "num_tokens": 90281176.0,
      "step": 43505
    },
    {
      "entropy": 5.69565896987915,
      "epoch": 3.9468432510885343,
      "grad_norm": 1.375,
      "learning_rate": 0.0003517680177938398,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22264595776796342,
      "num_tokens": 90292465.0,
      "step": 43510
    },
    {
      "entropy": 5.627174997329712,
      "epoch": 3.9472968069666186,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003517376021622279,
      "loss": 4.8007,
      "mean_token_accuracy": 0.2277838781476021,
      "num_tokens": 90303482.0,
      "step": 43515
    },
    {
      "entropy": 5.613353633880616,
      "epoch": 3.9477503628447024,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003517071849436629,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22391087412834168,
      "num_tokens": 90313696.0,
      "step": 43520
    },
    {
      "entropy": 5.635232210159302,
      "epoch": 3.9482039187227866,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035167676613877366,
      "loss": 4.917,
      "mean_token_accuracy": 0.21487147212028504,
      "num_tokens": 90324564.0,
      "step": 43525
    },
    {
      "entropy": 5.59060640335083,
      "epoch": 3.948657474600871,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003516463457481894,
      "loss": 4.7768,
      "mean_token_accuracy": 0.23128025680780412,
      "num_tokens": 90334440.0,
      "step": 43530
    },
    {
      "entropy": 5.630839014053345,
      "epoch": 3.9491110304789547,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003516159237725392,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22471659034490585,
      "num_tokens": 90344029.0,
      "step": 43535
    },
    {
      "entropy": 5.655629587173462,
      "epoch": 3.949564586357039,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003515855002124521,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22030624598264695,
      "num_tokens": 90354628.0,
      "step": 43540
    },
    {
      "entropy": 5.632492780685425,
      "epoch": 3.9500181422351233,
      "grad_norm": 1.25,
      "learning_rate": 0.0003515550750685573,
      "loss": 4.793,
      "mean_token_accuracy": 0.23002799898386,
      "num_tokens": 90365091.0,
      "step": 43545
    },
    {
      "entropy": 5.620274877548217,
      "epoch": 3.9504716981132075,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003515246483414842,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22337178885936737,
      "num_tokens": 90374197.0,
      "step": 43550
    },
    {
      "entropy": 5.639148283004761,
      "epoch": 3.950925253991292,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035149422003186176,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22177066057920455,
      "num_tokens": 90384686.0,
      "step": 43555
    },
    {
      "entropy": 5.666125106811523,
      "epoch": 3.951378809869376,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035146379014031943,
      "loss": 4.957,
      "mean_token_accuracy": 0.22066957950592042,
      "num_tokens": 90396329.0,
      "step": 43560
    },
    {
      "entropy": 5.674965333938599,
      "epoch": 3.9518323657474603,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003514333586674864,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22092576026916505,
      "num_tokens": 90408957.0,
      "step": 43565
    },
    {
      "entropy": 5.564086246490478,
      "epoch": 3.952285921625544,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000351402925613992,
      "loss": 4.8415,
      "mean_token_accuracy": 0.22564305514097213,
      "num_tokens": 90420194.0,
      "step": 43570
    },
    {
      "entropy": 5.55805721282959,
      "epoch": 3.9527394775036284,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003513724909804656,
      "loss": 4.809,
      "mean_token_accuracy": 0.22613648176193238,
      "num_tokens": 90430881.0,
      "step": 43575
    },
    {
      "entropy": 5.617742824554443,
      "epoch": 3.9531930333817127,
      "grad_norm": 1.421875,
      "learning_rate": 0.00035134205476753674,
      "loss": 4.8712,
      "mean_token_accuracy": 0.21894871592521667,
      "num_tokens": 90440713.0,
      "step": 43580
    },
    {
      "entropy": 5.631799030303955,
      "epoch": 3.953646589259797,
      "grad_norm": 1.25,
      "learning_rate": 0.0003513116169758347,
      "loss": 4.952,
      "mean_token_accuracy": 0.21696564108133315,
      "num_tokens": 90451274.0,
      "step": 43585
    },
    {
      "entropy": 5.663276195526123,
      "epoch": 3.9541001451378808,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003512811776059889,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21524891257286072,
      "num_tokens": 90461503.0,
      "step": 43590
    },
    {
      "entropy": 5.613519525527954,
      "epoch": 3.954553701015965,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000351250736658629,
      "loss": 4.8126,
      "mean_token_accuracy": 0.23150897920131683,
      "num_tokens": 90471178.0,
      "step": 43595
    },
    {
      "entropy": 5.599734497070313,
      "epoch": 3.9550072568940493,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003512202941343844,
      "loss": 4.8378,
      "mean_token_accuracy": 0.22066653966903688,
      "num_tokens": 90481573.0,
      "step": 43600
    },
    {
      "entropy": 5.627104997634888,
      "epoch": 3.9554608127721336,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003511898500338847,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21594603806734086,
      "num_tokens": 90491853.0,
      "step": 43605
    },
    {
      "entropy": 5.621861267089844,
      "epoch": 3.955914368650218,
      "grad_norm": 1.25,
      "learning_rate": 0.0003511594043577594,
      "loss": 4.9176,
      "mean_token_accuracy": 0.216946080327034,
      "num_tokens": 90502958.0,
      "step": 43610
    },
    {
      "entropy": 5.5644608497619625,
      "epoch": 3.956367924528302,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035112895710663833,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22364986836910247,
      "num_tokens": 90512555.0,
      "step": 43615
    },
    {
      "entropy": 5.618684196472168,
      "epoch": 3.956821480406386,
      "grad_norm": 1.25,
      "learning_rate": 0.00035109850828115105,
      "loss": 4.917,
      "mean_token_accuracy": 0.22495315223932266,
      "num_tokens": 90522722.0,
      "step": 43620
    },
    {
      "entropy": 5.617133665084839,
      "epoch": 3.95727503628447,
      "grad_norm": 1.25,
      "learning_rate": 0.00035106805788192707,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22280225306749343,
      "num_tokens": 90533182.0,
      "step": 43625
    },
    {
      "entropy": 5.64388222694397,
      "epoch": 3.9577285921625545,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003510376059095963,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21308707296848298,
      "num_tokens": 90543495.0,
      "step": 43630
    },
    {
      "entropy": 5.537411499023437,
      "epoch": 3.9581821480406387,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003510071523647886,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22778836339712144,
      "num_tokens": 90553680.0,
      "step": 43635
    },
    {
      "entropy": 5.564325761795044,
      "epoch": 3.9586357039187225,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003509766972481336,
      "loss": 4.7937,
      "mean_token_accuracy": 0.2248803362250328,
      "num_tokens": 90563513.0,
      "step": 43640
    },
    {
      "entropy": 5.716092538833618,
      "epoch": 3.959089259796807,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000350946240560261,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22279592007398605,
      "num_tokens": 90573337.0,
      "step": 43645
    },
    {
      "entropy": 5.585870885848999,
      "epoch": 3.959542815674891,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003509157823018009,
      "loss": 4.8612,
      "mean_token_accuracy": 0.2253480836749077,
      "num_tokens": 90582979.0,
      "step": 43650
    },
    {
      "entropy": 5.651793909072876,
      "epoch": 3.9599963715529753,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035088532247338306,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2131386250257492,
      "num_tokens": 90593552.0,
      "step": 43655
    },
    {
      "entropy": 5.642519187927246,
      "epoch": 3.9604499274310596,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035085486107563733,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2189141944050789,
      "num_tokens": 90603813.0,
      "step": 43660
    },
    {
      "entropy": 5.630860424041748,
      "epoch": 3.960903483309144,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035082439810919385,
      "loss": 4.8384,
      "mean_token_accuracy": 0.23279846757650374,
      "num_tokens": 90613885.0,
      "step": 43665
    },
    {
      "entropy": 5.609619855880737,
      "epoch": 3.9613570391872277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003507939335746824,
      "loss": 4.7896,
      "mean_token_accuracy": 0.22308997809886932,
      "num_tokens": 90624665.0,
      "step": 43670
    },
    {
      "entropy": 5.5984782695770265,
      "epoch": 3.961810595065312,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003507634674727332,
      "loss": 4.8625,
      "mean_token_accuracy": 0.228514201939106,
      "num_tokens": 90635496.0,
      "step": 43675
    },
    {
      "entropy": 5.627188634872437,
      "epoch": 3.9622641509433962,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003507329998039762,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22334334254264832,
      "num_tokens": 90645502.0,
      "step": 43680
    },
    {
      "entropy": 5.528412914276123,
      "epoch": 3.9627177068214805,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035070253056904133,
      "loss": 4.7945,
      "mean_token_accuracy": 0.2294122874736786,
      "num_tokens": 90654724.0,
      "step": 43685
    },
    {
      "entropy": 5.556365966796875,
      "epoch": 3.9631712626995643,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003506720597685589,
      "loss": 4.8657,
      "mean_token_accuracy": 0.2272373169660568,
      "num_tokens": 90665280.0,
      "step": 43690
    },
    {
      "entropy": 5.593201398849487,
      "epoch": 3.9636248185776486,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035064158740315894,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22146636843681336,
      "num_tokens": 90675804.0,
      "step": 43695
    },
    {
      "entropy": 5.5918159008026125,
      "epoch": 3.964078374455733,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035061111347347174,
      "loss": 4.8398,
      "mean_token_accuracy": 0.22862047404050828,
      "num_tokens": 90686636.0,
      "step": 43700
    },
    {
      "entropy": 5.549758052825927,
      "epoch": 3.964531930333817,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035058063798012746,
      "loss": 4.869,
      "mean_token_accuracy": 0.22825548350811004,
      "num_tokens": 90697666.0,
      "step": 43705
    },
    {
      "entropy": 5.662643098831177,
      "epoch": 3.9649854862119014,
      "grad_norm": 1.625,
      "learning_rate": 0.0003505501609237563,
      "loss": 4.9682,
      "mean_token_accuracy": 0.22011823654174806,
      "num_tokens": 90708111.0,
      "step": 43710
    },
    {
      "entropy": 5.609377479553222,
      "epoch": 3.9654390420899857,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035051968230498853,
      "loss": 4.8207,
      "mean_token_accuracy": 0.2346563994884491,
      "num_tokens": 90718711.0,
      "step": 43715
    },
    {
      "entropy": 5.534402322769165,
      "epoch": 3.96589259796807,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035048920212445454,
      "loss": 4.8126,
      "mean_token_accuracy": 0.2302231967449188,
      "num_tokens": 90729585.0,
      "step": 43720
    },
    {
      "entropy": 5.51228404045105,
      "epoch": 3.9663461538461537,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035045872038278467,
      "loss": 4.7503,
      "mean_token_accuracy": 0.233260902762413,
      "num_tokens": 90739654.0,
      "step": 43725
    },
    {
      "entropy": 5.585073328018188,
      "epoch": 3.966799709724238,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003504282370806091,
      "loss": 4.8141,
      "mean_token_accuracy": 0.22703779637813568,
      "num_tokens": 90749456.0,
      "step": 43730
    },
    {
      "entropy": 5.567838859558106,
      "epoch": 3.9672532656023223,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035039775221855845,
      "loss": 4.8333,
      "mean_token_accuracy": 0.22745248377323152,
      "num_tokens": 90760051.0,
      "step": 43735
    },
    {
      "entropy": 5.556958723068237,
      "epoch": 3.9677068214804065,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000350367265797263,
      "loss": 4.8055,
      "mean_token_accuracy": 0.22849767208099364,
      "num_tokens": 90770452.0,
      "step": 43740
    },
    {
      "entropy": 5.621649742126465,
      "epoch": 3.9681603773584904,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035033677781735336,
      "loss": 4.89,
      "mean_token_accuracy": 0.22210618555545808,
      "num_tokens": 90780150.0,
      "step": 43745
    },
    {
      "entropy": 5.615019655227661,
      "epoch": 3.9686139332365746,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003503062882794599,
      "loss": 4.8035,
      "mean_token_accuracy": 0.23081523776054383,
      "num_tokens": 90790338.0,
      "step": 43750
    },
    {
      "entropy": 5.560942983627319,
      "epoch": 3.969067489114659,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035027579718421327,
      "loss": 4.8521,
      "mean_token_accuracy": 0.22717443555593492,
      "num_tokens": 90800195.0,
      "step": 43755
    },
    {
      "entropy": 5.606147575378418,
      "epoch": 3.969521044992743,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003502453045322439,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21842382699251175,
      "num_tokens": 90810872.0,
      "step": 43760
    },
    {
      "entropy": 5.641963720321655,
      "epoch": 3.9699746008708274,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035021481032418254,
      "loss": 4.8509,
      "mean_token_accuracy": 0.22185555398464202,
      "num_tokens": 90821018.0,
      "step": 43765
    },
    {
      "entropy": 5.622490787506104,
      "epoch": 3.9704281567489117,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003501843145606597,
      "loss": 4.8855,
      "mean_token_accuracy": 0.21883311867713928,
      "num_tokens": 90830274.0,
      "step": 43770
    },
    {
      "entropy": 5.5336771488189695,
      "epoch": 3.9708817126269955,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000350153817242306,
      "loss": 4.7872,
      "mean_token_accuracy": 0.23493994623422623,
      "num_tokens": 90839926.0,
      "step": 43775
    },
    {
      "entropy": 5.505676937103272,
      "epoch": 3.97133526850508,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003501233183697523,
      "loss": 4.7734,
      "mean_token_accuracy": 0.2265172004699707,
      "num_tokens": 90850263.0,
      "step": 43780
    },
    {
      "entropy": 5.618466997146607,
      "epoch": 3.971788824383164,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035009281794362926,
      "loss": 4.8948,
      "mean_token_accuracy": 0.21449514329433442,
      "num_tokens": 90861747.0,
      "step": 43785
    },
    {
      "entropy": 5.631987762451172,
      "epoch": 3.9722423802612483,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035006231596456755,
      "loss": 4.9566,
      "mean_token_accuracy": 0.214010851085186,
      "num_tokens": 90872186.0,
      "step": 43790
    },
    {
      "entropy": 5.639399576187134,
      "epoch": 3.972695936139332,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035003181243319797,
      "loss": 4.8572,
      "mean_token_accuracy": 0.2287888765335083,
      "num_tokens": 90881088.0,
      "step": 43795
    },
    {
      "entropy": 5.534959077835083,
      "epoch": 3.9731494920174164,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003500013073501514,
      "loss": 4.8357,
      "mean_token_accuracy": 0.21795334666967392,
      "num_tokens": 90890589.0,
      "step": 43800
    },
    {
      "entropy": 5.587713146209717,
      "epoch": 3.9736030478955007,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003499708007160588,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2233035922050476,
      "num_tokens": 90901692.0,
      "step": 43805
    },
    {
      "entropy": 5.6533256530761715,
      "epoch": 3.974056603773585,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003499402925315509,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22145114243030548,
      "num_tokens": 90911811.0,
      "step": 43810
    },
    {
      "entropy": 5.628856372833252,
      "epoch": 3.974510159651669,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034990978279725846,
      "loss": 4.7872,
      "mean_token_accuracy": 0.23011803925037383,
      "num_tokens": 90922323.0,
      "step": 43815
    },
    {
      "entropy": 5.713240766525269,
      "epoch": 3.9749637155297535,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034987927151381283,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21182865351438523,
      "num_tokens": 90932834.0,
      "step": 43820
    },
    {
      "entropy": 5.61557903289795,
      "epoch": 3.9754172714078373,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034984875868184466,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22673559933900833,
      "num_tokens": 90943616.0,
      "step": 43825
    },
    {
      "entropy": 5.638593292236328,
      "epoch": 3.9758708272859216,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034981824430198517,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2186393156647682,
      "num_tokens": 90954468.0,
      "step": 43830
    },
    {
      "entropy": 5.6248821258544925,
      "epoch": 3.976324383164006,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034978772837486536,
      "loss": 4.8069,
      "mean_token_accuracy": 0.22970810830593108,
      "num_tokens": 90965109.0,
      "step": 43835
    },
    {
      "entropy": 5.629035425186157,
      "epoch": 3.97677793904209,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00034975721090111625,
      "loss": 4.8839,
      "mean_token_accuracy": 0.21781204789876937,
      "num_tokens": 90974789.0,
      "step": 43840
    },
    {
      "entropy": 5.625867033004761,
      "epoch": 3.977231494920174,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003497266918813689,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2187602326273918,
      "num_tokens": 90985154.0,
      "step": 43845
    },
    {
      "entropy": 5.6458518505096436,
      "epoch": 3.977685050798258,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003496961713162545,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22161955535411834,
      "num_tokens": 90996007.0,
      "step": 43850
    },
    {
      "entropy": 5.682658338546753,
      "epoch": 3.9781386066763424,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003496656492064043,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21685384064912797,
      "num_tokens": 91006253.0,
      "step": 43855
    },
    {
      "entropy": 5.67601203918457,
      "epoch": 3.9785921625544267,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003496351255524494,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21520382910966873,
      "num_tokens": 91017169.0,
      "step": 43860
    },
    {
      "entropy": 5.604124069213867,
      "epoch": 3.979045718432511,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034960460035502117,
      "loss": 4.8655,
      "mean_token_accuracy": 0.21704234778881074,
      "num_tokens": 91026684.0,
      "step": 43865
    },
    {
      "entropy": 5.592619466781616,
      "epoch": 3.9794992743105952,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003495740736147508,
      "loss": 4.8255,
      "mean_token_accuracy": 0.22522745430469512,
      "num_tokens": 91037462.0,
      "step": 43870
    },
    {
      "entropy": 5.65322093963623,
      "epoch": 3.9799528301886795,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034954354533226953,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21626212000846862,
      "num_tokens": 91049418.0,
      "step": 43875
    },
    {
      "entropy": 5.601533889770508,
      "epoch": 3.9804063860667633,
      "grad_norm": 1.25,
      "learning_rate": 0.0003495130155082087,
      "loss": 4.8336,
      "mean_token_accuracy": 0.22192179411649704,
      "num_tokens": 91061013.0,
      "step": 43880
    },
    {
      "entropy": 5.604547071456909,
      "epoch": 3.9808599419448476,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003494824841431998,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22662772834300995,
      "num_tokens": 91070974.0,
      "step": 43885
    },
    {
      "entropy": 5.643239450454712,
      "epoch": 3.981313497822932,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003494519512378741,
      "loss": 4.9304,
      "mean_token_accuracy": 0.2264939308166504,
      "num_tokens": 91082681.0,
      "step": 43890
    },
    {
      "entropy": 5.616656112670898,
      "epoch": 3.9817670537010157,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034942141679286307,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23242186903953552,
      "num_tokens": 91093552.0,
      "step": 43895
    },
    {
      "entropy": 5.6163019180297855,
      "epoch": 3.9822206095791,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003493908808087982,
      "loss": 4.9083,
      "mean_token_accuracy": 0.21618666648864746,
      "num_tokens": 91105470.0,
      "step": 43900
    },
    {
      "entropy": 5.5929603099823,
      "epoch": 3.9826741654571842,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000349360343286311,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2280385360121727,
      "num_tokens": 91115822.0,
      "step": 43905
    },
    {
      "entropy": 5.586233758926392,
      "epoch": 3.9831277213352685,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034932980422603283,
      "loss": 4.8282,
      "mean_token_accuracy": 0.23132265210151673,
      "num_tokens": 91125722.0,
      "step": 43910
    },
    {
      "entropy": 5.731109285354615,
      "epoch": 3.9835812772133528,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003492992636285955,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22006487846374512,
      "num_tokens": 91135783.0,
      "step": 43915
    },
    {
      "entropy": 5.635255813598633,
      "epoch": 3.984034833091437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003492687214946304,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22833815068006516,
      "num_tokens": 91146151.0,
      "step": 43920
    },
    {
      "entropy": 5.610350942611694,
      "epoch": 3.9844883889695213,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034923817782476927,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21957208812236786,
      "num_tokens": 91156711.0,
      "step": 43925
    },
    {
      "entropy": 5.601725435256958,
      "epoch": 3.984941944847605,
      "grad_norm": 1.25,
      "learning_rate": 0.0003492076326196436,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21859576553106308,
      "num_tokens": 91166601.0,
      "step": 43930
    },
    {
      "entropy": 5.695220947265625,
      "epoch": 3.9853955007256894,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034917708587988523,
      "loss": 4.9556,
      "mean_token_accuracy": 0.20916615724563598,
      "num_tokens": 91176556.0,
      "step": 43935
    },
    {
      "entropy": 5.638649225234985,
      "epoch": 3.9858490566037736,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034914653760612584,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22388535290956496,
      "num_tokens": 91186499.0,
      "step": 43940
    },
    {
      "entropy": 5.600866508483887,
      "epoch": 3.986302612481858,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003491159877989971,
      "loss": 4.8493,
      "mean_token_accuracy": 0.21494744569063187,
      "num_tokens": 91197088.0,
      "step": 43945
    },
    {
      "entropy": 5.588364124298096,
      "epoch": 3.9867561683599417,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000349085436459131,
      "loss": 4.891,
      "mean_token_accuracy": 0.2266418904066086,
      "num_tokens": 91207188.0,
      "step": 43950
    },
    {
      "entropy": 5.591065263748169,
      "epoch": 3.987209724238026,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034905488358715907,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22398226857185363,
      "num_tokens": 91217349.0,
      "step": 43955
    },
    {
      "entropy": 5.618248128890992,
      "epoch": 3.9876632801161103,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003490243291837133,
      "loss": 4.9292,
      "mean_token_accuracy": 0.21863684207201003,
      "num_tokens": 91227775.0,
      "step": 43960
    },
    {
      "entropy": 5.582199478149414,
      "epoch": 3.9881168359941945,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034899377324942556,
      "loss": 4.7826,
      "mean_token_accuracy": 0.23247582018375396,
      "num_tokens": 91238315.0,
      "step": 43965
    },
    {
      "entropy": 5.63126769065857,
      "epoch": 3.988570391872279,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034896321578492767,
      "loss": 4.9903,
      "mean_token_accuracy": 0.2124732404947281,
      "num_tokens": 91249178.0,
      "step": 43970
    },
    {
      "entropy": 5.5965455055236815,
      "epoch": 3.989023947750363,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003489326567908516,
      "loss": 4.8568,
      "mean_token_accuracy": 0.22392703890800475,
      "num_tokens": 91257911.0,
      "step": 43975
    },
    {
      "entropy": 5.600323295593261,
      "epoch": 3.989477503628447,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034890209626782943,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22421206682920455,
      "num_tokens": 91267491.0,
      "step": 43980
    },
    {
      "entropy": 5.609611511230469,
      "epoch": 3.989931059506531,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003488715342164931,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22713396400213243,
      "num_tokens": 91278292.0,
      "step": 43985
    },
    {
      "entropy": 5.659009599685669,
      "epoch": 3.9903846153846154,
      "grad_norm": 1.515625,
      "learning_rate": 0.00034884097063747453,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2096550226211548,
      "num_tokens": 91287871.0,
      "step": 43990
    },
    {
      "entropy": 5.5706806659698485,
      "epoch": 3.9908381712626997,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003488104055314059,
      "loss": 4.8183,
      "mean_token_accuracy": 0.22714080810546874,
      "num_tokens": 91297201.0,
      "step": 43995
    },
    {
      "entropy": 5.556038570404053,
      "epoch": 3.9912917271407835,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003487798388989193,
      "loss": 4.768,
      "mean_token_accuracy": 0.23064112663269043,
      "num_tokens": 91306449.0,
      "step": 44000
    },
    {
      "entropy": 5.528042125701904,
      "epoch": 3.9917452830188678,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034874927074064684,
      "loss": 4.7893,
      "mean_token_accuracy": 0.22544007301330565,
      "num_tokens": 91317516.0,
      "step": 44005
    },
    {
      "entropy": 5.604190683364868,
      "epoch": 3.992198838896952,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003487187010572206,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22188632786273957,
      "num_tokens": 91326850.0,
      "step": 44010
    },
    {
      "entropy": 5.619228744506836,
      "epoch": 3.9926523947750363,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003486881298492729,
      "loss": 4.8915,
      "mean_token_accuracy": 0.2217486783862114,
      "num_tokens": 91338068.0,
      "step": 44015
    },
    {
      "entropy": 5.70633544921875,
      "epoch": 3.9931059506531206,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003486575571174359,
      "loss": 4.866,
      "mean_token_accuracy": 0.22041468173265458,
      "num_tokens": 91349627.0,
      "step": 44020
    },
    {
      "entropy": 5.641649675369263,
      "epoch": 3.993559506531205,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034862698286234185,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2111356645822525,
      "num_tokens": 91361229.0,
      "step": 44025
    },
    {
      "entropy": 5.634420776367188,
      "epoch": 3.9940130624092887,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003485964070846229,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2139379173517227,
      "num_tokens": 91372348.0,
      "step": 44030
    },
    {
      "entropy": 5.579347276687622,
      "epoch": 3.994466618287373,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003485658297849117,
      "loss": 4.805,
      "mean_token_accuracy": 0.2221580997109413,
      "num_tokens": 91383028.0,
      "step": 44035
    },
    {
      "entropy": 5.642390203475952,
      "epoch": 3.994920174165457,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034853525096384034,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22041716277599335,
      "num_tokens": 91393263.0,
      "step": 44040
    },
    {
      "entropy": 5.646244668960572,
      "epoch": 3.9953737300435415,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003485046706220411,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22217684537172316,
      "num_tokens": 91403881.0,
      "step": 44045
    },
    {
      "entropy": 5.627064895629883,
      "epoch": 3.9958272859216253,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003484740887601467,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2178072839975357,
      "num_tokens": 91413723.0,
      "step": 44050
    },
    {
      "entropy": 5.6279621601104735,
      "epoch": 3.9962808417997095,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034844350537878937,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2247815638780594,
      "num_tokens": 91424168.0,
      "step": 44055
    },
    {
      "entropy": 5.70906138420105,
      "epoch": 3.996734397677794,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034841292047860164,
      "loss": 4.9668,
      "mean_token_accuracy": 0.211597840487957,
      "num_tokens": 91433133.0,
      "step": 44060
    },
    {
      "entropy": 5.647665500640869,
      "epoch": 3.997187953555878,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000348382334060216,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2101356267929077,
      "num_tokens": 91444521.0,
      "step": 44065
    },
    {
      "entropy": 5.554708433151245,
      "epoch": 3.9976415094339623,
      "grad_norm": 1.453125,
      "learning_rate": 0.000348351746124265,
      "loss": 4.8556,
      "mean_token_accuracy": 0.22607757449150084,
      "num_tokens": 91454870.0,
      "step": 44070
    },
    {
      "entropy": 5.604092788696289,
      "epoch": 3.9980950653120466,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003483211566713812,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22667310535907745,
      "num_tokens": 91465472.0,
      "step": 44075
    },
    {
      "entropy": 5.649698495864868,
      "epoch": 3.998548621190131,
      "grad_norm": 1.25,
      "learning_rate": 0.0003482905657021971,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22566566467285157,
      "num_tokens": 91475467.0,
      "step": 44080
    },
    {
      "entropy": 5.64481897354126,
      "epoch": 3.9990021770682147,
      "grad_norm": 1.25,
      "learning_rate": 0.0003482599732173456,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22375485002994538,
      "num_tokens": 91485615.0,
      "step": 44085
    },
    {
      "entropy": 5.614963483810425,
      "epoch": 3.999455732946299,
      "grad_norm": 1.25,
      "learning_rate": 0.0003482293792174591,
      "loss": 4.8549,
      "mean_token_accuracy": 0.2235678941011429,
      "num_tokens": 91495790.0,
      "step": 44090
    },
    {
      "entropy": 5.609225845336914,
      "epoch": 3.9999092888243832,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003481987837031703,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2185494065284729,
      "num_tokens": 91507164.0,
      "step": 44095
    },
    {
      "entropy": 5.582514142990112,
      "epoch": 4.000362844702467,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003481681866751121,
      "loss": 4.8098,
      "mean_token_accuracy": 0.2285863161087036,
      "num_tokens": 91516732.0,
      "step": 44100
    },
    {
      "entropy": 5.543740081787109,
      "epoch": 4.000816400580551,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034813758813391716,
      "loss": 4.7628,
      "mean_token_accuracy": 0.23362062573432923,
      "num_tokens": 91527073.0,
      "step": 44105
    },
    {
      "entropy": 5.572063875198364,
      "epoch": 4.001269956458636,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003481069880802183,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2260878473520279,
      "num_tokens": 91538314.0,
      "step": 44110
    },
    {
      "entropy": 5.617223691940308,
      "epoch": 4.00172351233672,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034807638651464815,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22718000262975693,
      "num_tokens": 91549159.0,
      "step": 44115
    },
    {
      "entropy": 5.580330181121826,
      "epoch": 4.002177068214804,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003480457834378398,
      "loss": 4.854,
      "mean_token_accuracy": 0.2230929210782051,
      "num_tokens": 91560901.0,
      "step": 44120
    },
    {
      "entropy": 5.57982988357544,
      "epoch": 4.002630624092888,
      "grad_norm": 1.25,
      "learning_rate": 0.00034801517885042597,
      "loss": 4.8124,
      "mean_token_accuracy": 0.2222102627158165,
      "num_tokens": 91570797.0,
      "step": 44125
    },
    {
      "entropy": 5.699436950683594,
      "epoch": 4.003084179970973,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003479845727530397,
      "loss": 4.8978,
      "mean_token_accuracy": 0.216018408536911,
      "num_tokens": 91580971.0,
      "step": 44130
    },
    {
      "entropy": 5.624577331542969,
      "epoch": 4.003537735849057,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034795396514631386,
      "loss": 4.8633,
      "mean_token_accuracy": 0.227254319190979,
      "num_tokens": 91590834.0,
      "step": 44135
    },
    {
      "entropy": 5.582599925994873,
      "epoch": 4.003991291727141,
      "grad_norm": 1.125,
      "learning_rate": 0.00034792335603088135,
      "loss": 4.8733,
      "mean_token_accuracy": 0.2203483372926712,
      "num_tokens": 91602099.0,
      "step": 44140
    },
    {
      "entropy": 5.551930618286133,
      "epoch": 4.004444847605225,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034789274540737535,
      "loss": 4.7796,
      "mean_token_accuracy": 0.23713664412498475,
      "num_tokens": 91612269.0,
      "step": 44145
    },
    {
      "entropy": 5.619314098358155,
      "epoch": 4.004898403483309,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034786213327642874,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22387367635965347,
      "num_tokens": 91623644.0,
      "step": 44150
    },
    {
      "entropy": 5.621955919265747,
      "epoch": 4.005351959361393,
      "grad_norm": 1.375,
      "learning_rate": 0.0003478315196386747,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23409025818109513,
      "num_tokens": 91633563.0,
      "step": 44155
    },
    {
      "entropy": 5.571321296691894,
      "epoch": 4.005805515239477,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003478009044947463,
      "loss": 4.7949,
      "mean_token_accuracy": 0.23007688969373702,
      "num_tokens": 91644648.0,
      "step": 44160
    },
    {
      "entropy": 5.563946771621704,
      "epoch": 4.006259071117562,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003477702878452766,
      "loss": 4.6941,
      "mean_token_accuracy": 0.24915118217468263,
      "num_tokens": 91655366.0,
      "step": 44165
    },
    {
      "entropy": 5.575225687026977,
      "epoch": 4.006712626995646,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003477396696908988,
      "loss": 4.815,
      "mean_token_accuracy": 0.22737202942371368,
      "num_tokens": 91665474.0,
      "step": 44170
    },
    {
      "entropy": 5.612418365478516,
      "epoch": 4.00716618287373,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034770905003224606,
      "loss": 4.8344,
      "mean_token_accuracy": 0.22141342163085936,
      "num_tokens": 91675291.0,
      "step": 44175
    },
    {
      "entropy": 5.51589879989624,
      "epoch": 4.007619738751814,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003476784288699517,
      "loss": 4.7581,
      "mean_token_accuracy": 0.2309844359755516,
      "num_tokens": 91686472.0,
      "step": 44180
    },
    {
      "entropy": 5.618040609359741,
      "epoch": 4.008073294629899,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034764780620464886,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22249191403388976,
      "num_tokens": 91696347.0,
      "step": 44185
    },
    {
      "entropy": 5.587293434143066,
      "epoch": 4.008526850507983,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003476171820369709,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22103931456804277,
      "num_tokens": 91706504.0,
      "step": 44190
    },
    {
      "entropy": 5.57837142944336,
      "epoch": 4.008980406386066,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034758655636755103,
      "loss": 4.7817,
      "mean_token_accuracy": 0.2345270812511444,
      "num_tokens": 91716559.0,
      "step": 44195
    },
    {
      "entropy": 5.6016185760498045,
      "epoch": 4.009433962264151,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034755592919702275,
      "loss": 4.8205,
      "mean_token_accuracy": 0.22436016350984572,
      "num_tokens": 91726312.0,
      "step": 44200
    },
    {
      "entropy": 5.559885931015015,
      "epoch": 4.009887518142235,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034752530052601944,
      "loss": 4.7849,
      "mean_token_accuracy": 0.22821685373783113,
      "num_tokens": 91736882.0,
      "step": 44205
    },
    {
      "entropy": 5.601106786727906,
      "epoch": 4.010341074020319,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034749467035517437,
      "loss": 4.8686,
      "mean_token_accuracy": 0.220924474298954,
      "num_tokens": 91746905.0,
      "step": 44210
    },
    {
      "entropy": 5.5938637256622314,
      "epoch": 4.010794629898403,
      "grad_norm": 1.25,
      "learning_rate": 0.00034746403868512094,
      "loss": 4.877,
      "mean_token_accuracy": 0.2210596099495888,
      "num_tokens": 91757609.0,
      "step": 44215
    },
    {
      "entropy": 5.613859844207764,
      "epoch": 4.011248185776488,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034743340551649285,
      "loss": 4.7877,
      "mean_token_accuracy": 0.23170709758996963,
      "num_tokens": 91768017.0,
      "step": 44220
    },
    {
      "entropy": 5.549668645858764,
      "epoch": 4.011701741654572,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034740277084992346,
      "loss": 4.7206,
      "mean_token_accuracy": 0.23830063939094542,
      "num_tokens": 91777820.0,
      "step": 44225
    },
    {
      "entropy": 5.545172929763794,
      "epoch": 4.012155297532656,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034737213468604626,
      "loss": 4.836,
      "mean_token_accuracy": 0.22365311682224273,
      "num_tokens": 91789011.0,
      "step": 44230
    },
    {
      "entropy": 5.6277398586273195,
      "epoch": 4.0126088534107405,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003473414970254949,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2198638513684273,
      "num_tokens": 91799165.0,
      "step": 44235
    },
    {
      "entropy": 5.5939546585083,
      "epoch": 4.013062409288825,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034731085786890306,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22711962312459946,
      "num_tokens": 91809602.0,
      "step": 44240
    },
    {
      "entropy": 5.636578607559204,
      "epoch": 4.013515965166909,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034728021721690407,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22376101613044738,
      "num_tokens": 91820092.0,
      "step": 44245
    },
    {
      "entropy": 5.566355657577515,
      "epoch": 4.013969521044992,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003472495750701318,
      "loss": 4.8429,
      "mean_token_accuracy": 0.2324467420578003,
      "num_tokens": 91829730.0,
      "step": 44250
    },
    {
      "entropy": 5.582024049758911,
      "epoch": 4.014423076923077,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034721893142922,
      "loss": 4.8655,
      "mean_token_accuracy": 0.21847952157258987,
      "num_tokens": 91840697.0,
      "step": 44255
    },
    {
      "entropy": 5.621986532211304,
      "epoch": 4.014876632801161,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003471882862948023,
      "loss": 4.8544,
      "mean_token_accuracy": 0.22468733191490173,
      "num_tokens": 91852145.0,
      "step": 44260
    },
    {
      "entropy": 5.625450325012207,
      "epoch": 4.015330188679245,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034715763966751233,
      "loss": 4.8055,
      "mean_token_accuracy": 0.2212170660495758,
      "num_tokens": 91861940.0,
      "step": 44265
    },
    {
      "entropy": 5.652124118804932,
      "epoch": 4.0157837445573294,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034712699154798403,
      "loss": 4.8696,
      "mean_token_accuracy": 0.2269722878932953,
      "num_tokens": 91872179.0,
      "step": 44270
    },
    {
      "entropy": 5.623732948303223,
      "epoch": 4.016237300435414,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034709634193685126,
      "loss": 4.7919,
      "mean_token_accuracy": 0.2293704628944397,
      "num_tokens": 91882051.0,
      "step": 44275
    },
    {
      "entropy": 5.582018136978149,
      "epoch": 4.016690856313498,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003470656908347476,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22427183240652085,
      "num_tokens": 91892672.0,
      "step": 44280
    },
    {
      "entropy": 5.5499889850616455,
      "epoch": 4.017144412191582,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003470350382423072,
      "loss": 4.7432,
      "mean_token_accuracy": 0.23241445869207383,
      "num_tokens": 91903418.0,
      "step": 44285
    },
    {
      "entropy": 5.5989936828613285,
      "epoch": 4.0175979680696665,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003470043841601639,
      "loss": 4.835,
      "mean_token_accuracy": 0.22673787474632262,
      "num_tokens": 91913076.0,
      "step": 44290
    },
    {
      "entropy": 5.575576210021973,
      "epoch": 4.018051523947751,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034697372858895147,
      "loss": 4.8282,
      "mean_token_accuracy": 0.23406293541193007,
      "num_tokens": 91923464.0,
      "step": 44295
    },
    {
      "entropy": 5.636105346679687,
      "epoch": 4.018505079825834,
      "grad_norm": 1.28125,
      "learning_rate": 0.000346943071529304,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22333446890115738,
      "num_tokens": 91934399.0,
      "step": 44300
    },
    {
      "entropy": 5.643493175506592,
      "epoch": 4.018958635703918,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034691241298185553,
      "loss": 4.8653,
      "mean_token_accuracy": 0.23206893652677535,
      "num_tokens": 91945330.0,
      "step": 44305
    },
    {
      "entropy": 5.616344308853149,
      "epoch": 4.019412191582003,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034688175294724004,
      "loss": 4.8694,
      "mean_token_accuracy": 0.21859676986932755,
      "num_tokens": 91956379.0,
      "step": 44310
    },
    {
      "entropy": 5.492795801162719,
      "epoch": 4.019865747460087,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034685109142609154,
      "loss": 4.7131,
      "mean_token_accuracy": 0.2326076954603195,
      "num_tokens": 91965969.0,
      "step": 44315
    },
    {
      "entropy": 5.6541228771209715,
      "epoch": 4.020319303338171,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003468204284190442,
      "loss": 4.9645,
      "mean_token_accuracy": 0.2119796097278595,
      "num_tokens": 91977005.0,
      "step": 44320
    },
    {
      "entropy": 5.628622007369995,
      "epoch": 4.0207728592162555,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034678976392673206,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22776288986206056,
      "num_tokens": 91988735.0,
      "step": 44325
    },
    {
      "entropy": 5.637476634979248,
      "epoch": 4.02122641509434,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034675909794978926,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2270573452115059,
      "num_tokens": 91998743.0,
      "step": 44330
    },
    {
      "entropy": 5.626567125320435,
      "epoch": 4.021679970972424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034672843048885017,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22716895639896392,
      "num_tokens": 92009465.0,
      "step": 44335
    },
    {
      "entropy": 5.552109956741333,
      "epoch": 4.022133526850508,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034669776154454875,
      "loss": 4.7965,
      "mean_token_accuracy": 0.2357358902692795,
      "num_tokens": 92020818.0,
      "step": 44340
    },
    {
      "entropy": 5.636726760864258,
      "epoch": 4.0225870827285926,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034666709111751945,
      "loss": 4.9121,
      "mean_token_accuracy": 0.21912719905376435,
      "num_tokens": 92031528.0,
      "step": 44345
    },
    {
      "entropy": 5.6711774349212645,
      "epoch": 4.023040638606676,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003466364192083963,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22255193442106247,
      "num_tokens": 92041856.0,
      "step": 44350
    },
    {
      "entropy": 5.607060337066651,
      "epoch": 4.02349419448476,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034660574581781386,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23123977333307266,
      "num_tokens": 92052117.0,
      "step": 44355
    },
    {
      "entropy": 5.630417537689209,
      "epoch": 4.0239477503628445,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003465750709464063,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22280800193548203,
      "num_tokens": 92062405.0,
      "step": 44360
    },
    {
      "entropy": 5.642234516143799,
      "epoch": 4.024401306240929,
      "grad_norm": 1.25,
      "learning_rate": 0.000346544394594808,
      "loss": 4.8742,
      "mean_token_accuracy": 0.233217491209507,
      "num_tokens": 92073269.0,
      "step": 44365
    },
    {
      "entropy": 5.5918628692626955,
      "epoch": 4.024854862119013,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034651371676365343,
      "loss": 4.7818,
      "mean_token_accuracy": 0.23179157078266144,
      "num_tokens": 92083645.0,
      "step": 44370
    },
    {
      "entropy": 5.697069072723389,
      "epoch": 4.025308417997097,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034648303745357705,
      "loss": 4.843,
      "mean_token_accuracy": 0.22801963686943055,
      "num_tokens": 92093696.0,
      "step": 44375
    },
    {
      "entropy": 5.589551019668579,
      "epoch": 4.0257619738751815,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003464523566652131,
      "loss": 4.8853,
      "mean_token_accuracy": 0.21721942275762557,
      "num_tokens": 92103217.0,
      "step": 44380
    },
    {
      "entropy": 5.560887670516967,
      "epoch": 4.026215529753266,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034642167439919634,
      "loss": 4.7791,
      "mean_token_accuracy": 0.2266951397061348,
      "num_tokens": 92112934.0,
      "step": 44385
    },
    {
      "entropy": 5.580127954483032,
      "epoch": 4.02666908563135,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003463909906561611,
      "loss": 4.8336,
      "mean_token_accuracy": 0.23183880001306534,
      "num_tokens": 92123066.0,
      "step": 44390
    },
    {
      "entropy": 5.598183727264404,
      "epoch": 4.027122641509434,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003463603054367419,
      "loss": 4.8225,
      "mean_token_accuracy": 0.2302472174167633,
      "num_tokens": 92133911.0,
      "step": 44395
    },
    {
      "entropy": 5.528662443161011,
      "epoch": 4.027576197387518,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003463296187415735,
      "loss": 4.7644,
      "mean_token_accuracy": 0.23149650394916535,
      "num_tokens": 92143941.0,
      "step": 44400
    },
    {
      "entropy": 5.542422437667847,
      "epoch": 4.028029753265602,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003462989305712904,
      "loss": 4.7833,
      "mean_token_accuracy": 0.22673832774162292,
      "num_tokens": 92153934.0,
      "step": 44405
    },
    {
      "entropy": 5.578616762161255,
      "epoch": 4.028483309143686,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003462682409265272,
      "loss": 4.7807,
      "mean_token_accuracy": 0.23435907661914826,
      "num_tokens": 92163642.0,
      "step": 44410
    },
    {
      "entropy": 5.5748560428619385,
      "epoch": 4.0289368650217705,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003462375498079186,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22381222248077393,
      "num_tokens": 92174034.0,
      "step": 44415
    },
    {
      "entropy": 5.613165950775146,
      "epoch": 4.029390420899855,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034620685721609937,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2274747982621193,
      "num_tokens": 92183714.0,
      "step": 44420
    },
    {
      "entropy": 5.590362501144409,
      "epoch": 4.029843976777939,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034617616315170413,
      "loss": 4.8878,
      "mean_token_accuracy": 0.21772707849740983,
      "num_tokens": 92195291.0,
      "step": 44425
    },
    {
      "entropy": 5.592410326004028,
      "epoch": 4.030297532656023,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003461454676153678,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22856836169958114,
      "num_tokens": 92206159.0,
      "step": 44430
    },
    {
      "entropy": 5.641799545288086,
      "epoch": 4.030751088534108,
      "grad_norm": 1.296875,
      "learning_rate": 0.000346114770607725,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2268071249127388,
      "num_tokens": 92216928.0,
      "step": 44435
    },
    {
      "entropy": 5.547517538070679,
      "epoch": 4.031204644412192,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003460840721294106,
      "loss": 4.8071,
      "mean_token_accuracy": 0.2300657734274864,
      "num_tokens": 92227029.0,
      "step": 44440
    },
    {
      "entropy": 5.594676828384399,
      "epoch": 4.031658200290276,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034605337218105955,
      "loss": 4.866,
      "mean_token_accuracy": 0.22630172669887544,
      "num_tokens": 92237310.0,
      "step": 44445
    },
    {
      "entropy": 5.556257152557373,
      "epoch": 4.03211175616836,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003460226707633066,
      "loss": 4.8093,
      "mean_token_accuracy": 0.23246752470731735,
      "num_tokens": 92247841.0,
      "step": 44450
    },
    {
      "entropy": 5.634732961654663,
      "epoch": 4.032565312046444,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034599196787678667,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22523575723171235,
      "num_tokens": 92258513.0,
      "step": 44455
    },
    {
      "entropy": 5.5847461223602295,
      "epoch": 4.033018867924528,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003459612635221349,
      "loss": 4.848,
      "mean_token_accuracy": 0.224467670917511,
      "num_tokens": 92268659.0,
      "step": 44460
    },
    {
      "entropy": 5.564281558990478,
      "epoch": 4.033472423802612,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034593055769998595,
      "loss": 4.8349,
      "mean_token_accuracy": 0.2213011547923088,
      "num_tokens": 92279810.0,
      "step": 44465
    },
    {
      "entropy": 5.588127374649048,
      "epoch": 4.0339259796806965,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003458998504109751,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22737037390470505,
      "num_tokens": 92289625.0,
      "step": 44470
    },
    {
      "entropy": 5.552697372436524,
      "epoch": 4.034379535558781,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003458691416557372,
      "loss": 4.8091,
      "mean_token_accuracy": 0.22761324793100357,
      "num_tokens": 92300669.0,
      "step": 44475
    },
    {
      "entropy": 5.56818642616272,
      "epoch": 4.034833091436865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034583843143490744,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22722165137529374,
      "num_tokens": 92312160.0,
      "step": 44480
    },
    {
      "entropy": 5.636919116973877,
      "epoch": 4.035286647314949,
      "grad_norm": 1.25,
      "learning_rate": 0.0003458077197491208,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2149924010038376,
      "num_tokens": 92322677.0,
      "step": 44485
    },
    {
      "entropy": 5.592791986465454,
      "epoch": 4.035740203193034,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003457770065990126,
      "loss": 4.7668,
      "mean_token_accuracy": 0.23507199883461,
      "num_tokens": 92333148.0,
      "step": 44490
    },
    {
      "entropy": 5.636591196060181,
      "epoch": 4.036193759071118,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003457462919852178,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22432505190372468,
      "num_tokens": 92344189.0,
      "step": 44495
    },
    {
      "entropy": 5.619293260574341,
      "epoch": 4.036647314949202,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003457155759083717,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23194621950387956,
      "num_tokens": 92356136.0,
      "step": 44500
    },
    {
      "entropy": 5.564880847930908,
      "epoch": 4.0371008708272855,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003456848583691094,
      "loss": 4.7302,
      "mean_token_accuracy": 0.23218549489974977,
      "num_tokens": 92366393.0,
      "step": 44505
    },
    {
      "entropy": 5.578526401519776,
      "epoch": 4.03755442670537,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003456541393680662,
      "loss": 4.7763,
      "mean_token_accuracy": 0.2345937356352806,
      "num_tokens": 92376819.0,
      "step": 44510
    },
    {
      "entropy": 5.5824384689331055,
      "epoch": 4.038007982583454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034562341890587747,
      "loss": 4.8716,
      "mean_token_accuracy": 0.22493281066417695,
      "num_tokens": 92386704.0,
      "step": 44515
    },
    {
      "entropy": 5.567879343032837,
      "epoch": 4.038461538461538,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003455926969831784,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22270402163267136,
      "num_tokens": 92397412.0,
      "step": 44520
    },
    {
      "entropy": 5.690203094482422,
      "epoch": 4.038915094339623,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034556197360060436,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21730535477399826,
      "num_tokens": 92408641.0,
      "step": 44525
    },
    {
      "entropy": 5.5965500831604,
      "epoch": 4.039368650217707,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003455312487587908,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22844212800264357,
      "num_tokens": 92419341.0,
      "step": 44530
    },
    {
      "entropy": 5.59076623916626,
      "epoch": 4.039822206095791,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003455005224583729,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2282400280237198,
      "num_tokens": 92429297.0,
      "step": 44535
    },
    {
      "entropy": 5.656727695465088,
      "epoch": 4.040275761973875,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034546979469998637,
      "loss": 4.8971,
      "mean_token_accuracy": 0.22270358800888063,
      "num_tokens": 92439851.0,
      "step": 44540
    },
    {
      "entropy": 5.524363327026367,
      "epoch": 4.04072931785196,
      "grad_norm": 1.25,
      "learning_rate": 0.00034543906548426646,
      "loss": 4.7875,
      "mean_token_accuracy": 0.23762448728084565,
      "num_tokens": 92451096.0,
      "step": 44545
    },
    {
      "entropy": 5.58683295249939,
      "epoch": 4.041182873730044,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034540833481184867,
      "loss": 4.81,
      "mean_token_accuracy": 0.2267359808087349,
      "num_tokens": 92460516.0,
      "step": 44550
    },
    {
      "entropy": 5.683700799942017,
      "epoch": 4.041636429608127,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034537760268336866,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21936065107584,
      "num_tokens": 92471429.0,
      "step": 44555
    },
    {
      "entropy": 5.634250497817993,
      "epoch": 4.042089985486212,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003453468690994619,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22756273597478865,
      "num_tokens": 92482047.0,
      "step": 44560
    },
    {
      "entropy": 5.549635553359986,
      "epoch": 4.042543541364296,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034531613406076387,
      "loss": 4.7829,
      "mean_token_accuracy": 0.23348269164562224,
      "num_tokens": 92492206.0,
      "step": 44565
    },
    {
      "entropy": 5.520943689346313,
      "epoch": 4.04299709724238,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003452853975679103,
      "loss": 4.7487,
      "mean_token_accuracy": 0.2299257054924965,
      "num_tokens": 92501979.0,
      "step": 44570
    },
    {
      "entropy": 5.703576183319091,
      "epoch": 4.043450653120464,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034525465962153683,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22335177659988403,
      "num_tokens": 92512388.0,
      "step": 44575
    },
    {
      "entropy": 5.6362231254577635,
      "epoch": 4.043904208998549,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034522392022227915,
      "loss": 4.8882,
      "mean_token_accuracy": 0.21817947626113893,
      "num_tokens": 92522805.0,
      "step": 44580
    },
    {
      "entropy": 5.559390687942505,
      "epoch": 4.044357764876633,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034519317937077283,
      "loss": 4.8378,
      "mean_token_accuracy": 0.22282641977071763,
      "num_tokens": 92532957.0,
      "step": 44585
    },
    {
      "entropy": 5.5748978614807125,
      "epoch": 4.044811320754717,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003451624370676537,
      "loss": 4.8579,
      "mean_token_accuracy": 0.22243332117795944,
      "num_tokens": 92543451.0,
      "step": 44590
    },
    {
      "entropy": 5.67425012588501,
      "epoch": 4.045264876632801,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034513169331355754,
      "loss": 4.8421,
      "mean_token_accuracy": 0.22604949921369552,
      "num_tokens": 92553856.0,
      "step": 44595
    },
    {
      "entropy": 5.565695333480835,
      "epoch": 4.045718432510886,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034510094810911995,
      "loss": 4.7656,
      "mean_token_accuracy": 0.2331593319773674,
      "num_tokens": 92563816.0,
      "step": 44600
    },
    {
      "entropy": 5.567518901824951,
      "epoch": 4.04617198838897,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034507020145497705,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23186744302511214,
      "num_tokens": 92574062.0,
      "step": 44605
    },
    {
      "entropy": 5.563264179229736,
      "epoch": 4.046625544267053,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034503945335176444,
      "loss": 4.7909,
      "mean_token_accuracy": 0.23605702221393585,
      "num_tokens": 92583935.0,
      "step": 44610
    },
    {
      "entropy": 5.6631800651550295,
      "epoch": 4.047079100145138,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034500870380011807,
      "loss": 4.849,
      "mean_token_accuracy": 0.23103433698415757,
      "num_tokens": 92594037.0,
      "step": 44615
    },
    {
      "entropy": 5.601016330718994,
      "epoch": 4.047532656023222,
      "grad_norm": 1.25,
      "learning_rate": 0.00034497795280067386,
      "loss": 4.809,
      "mean_token_accuracy": 0.22476974576711656,
      "num_tokens": 92603336.0,
      "step": 44620
    },
    {
      "entropy": 5.511468505859375,
      "epoch": 4.047986211901306,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034494720035406777,
      "loss": 4.7471,
      "mean_token_accuracy": 0.23324310928583145,
      "num_tokens": 92614210.0,
      "step": 44625
    },
    {
      "entropy": 5.60916223526001,
      "epoch": 4.04843976777939,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003449164464609358,
      "loss": 4.859,
      "mean_token_accuracy": 0.22311047911643983,
      "num_tokens": 92624036.0,
      "step": 44630
    },
    {
      "entropy": 5.618900632858276,
      "epoch": 4.048893323657475,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003448856911219139,
      "loss": 4.917,
      "mean_token_accuracy": 0.217339688539505,
      "num_tokens": 92634588.0,
      "step": 44635
    },
    {
      "entropy": 5.587625026702881,
      "epoch": 4.049346879535559,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034485493433763804,
      "loss": 4.797,
      "mean_token_accuracy": 0.22843301445245742,
      "num_tokens": 92645027.0,
      "step": 44640
    },
    {
      "entropy": 5.68034029006958,
      "epoch": 4.049800435413643,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034482417610874437,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22312044501304626,
      "num_tokens": 92656669.0,
      "step": 44645
    },
    {
      "entropy": 5.618550252914429,
      "epoch": 4.0502539912917275,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003447934164358689,
      "loss": 4.9292,
      "mean_token_accuracy": 0.21590907126665115,
      "num_tokens": 92666558.0,
      "step": 44650
    },
    {
      "entropy": 5.585630559921265,
      "epoch": 4.050707547169812,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034476265531964786,
      "loss": 4.86,
      "mean_token_accuracy": 0.22730345726013185,
      "num_tokens": 92677384.0,
      "step": 44655
    },
    {
      "entropy": 5.540721607208252,
      "epoch": 4.051161103047895,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003447318927607173,
      "loss": 4.782,
      "mean_token_accuracy": 0.23331023901700973,
      "num_tokens": 92687619.0,
      "step": 44660
    },
    {
      "entropy": 5.562377786636352,
      "epoch": 4.051614658925979,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003447011287597134,
      "loss": 4.7177,
      "mean_token_accuracy": 0.2337102934718132,
      "num_tokens": 92697784.0,
      "step": 44665
    },
    {
      "entropy": 5.625196075439453,
      "epoch": 4.052068214804064,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003446703633172724,
      "loss": 4.9048,
      "mean_token_accuracy": 0.21917217522859572,
      "num_tokens": 92709255.0,
      "step": 44670
    },
    {
      "entropy": 5.54682035446167,
      "epoch": 4.052521770682148,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003446395964340305,
      "loss": 4.7761,
      "mean_token_accuracy": 0.2355020135641098,
      "num_tokens": 92720714.0,
      "step": 44675
    },
    {
      "entropy": 5.58401517868042,
      "epoch": 4.052975326560232,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034460882811062417,
      "loss": 4.818,
      "mean_token_accuracy": 0.22225105464458467,
      "num_tokens": 92730783.0,
      "step": 44680
    },
    {
      "entropy": 5.645687866210937,
      "epoch": 4.0534288824383164,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034457805834768937,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21911788135766982,
      "num_tokens": 92741943.0,
      "step": 44685
    },
    {
      "entropy": 5.591673183441162,
      "epoch": 4.053882438316401,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003445472871458627,
      "loss": 4.8384,
      "mean_token_accuracy": 0.22499633580446243,
      "num_tokens": 92753510.0,
      "step": 44690
    },
    {
      "entropy": 5.610969400405883,
      "epoch": 4.054335994194485,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003445165145057803,
      "loss": 4.7921,
      "mean_token_accuracy": 0.22890833914279937,
      "num_tokens": 92763655.0,
      "step": 44695
    },
    {
      "entropy": 5.608323335647583,
      "epoch": 4.054789550072569,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034448574042807866,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22084259241819382,
      "num_tokens": 92773732.0,
      "step": 44700
    },
    {
      "entropy": 5.595209646224975,
      "epoch": 4.0552431059506535,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003444549649133943,
      "loss": 4.7939,
      "mean_token_accuracy": 0.22537048757076264,
      "num_tokens": 92784822.0,
      "step": 44705
    },
    {
      "entropy": 5.660883045196533,
      "epoch": 4.055696661828737,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003444241879623636,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21364960074424744,
      "num_tokens": 92795270.0,
      "step": 44710
    },
    {
      "entropy": 5.616207551956177,
      "epoch": 4.056150217706821,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003443934095756229,
      "loss": 4.815,
      "mean_token_accuracy": 0.2324299544095993,
      "num_tokens": 92806578.0,
      "step": 44715
    },
    {
      "entropy": 5.643567323684692,
      "epoch": 4.056603773584905,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003443626297538088,
      "loss": 4.8371,
      "mean_token_accuracy": 0.22598257958889006,
      "num_tokens": 92815932.0,
      "step": 44720
    },
    {
      "entropy": 5.570905971527099,
      "epoch": 4.05705732946299,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003443318484975579,
      "loss": 4.8185,
      "mean_token_accuracy": 0.2312711015343666,
      "num_tokens": 92825536.0,
      "step": 44725
    },
    {
      "entropy": 5.587768459320069,
      "epoch": 4.057510885341074,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003443010658075066,
      "loss": 4.8029,
      "mean_token_accuracy": 0.2293898046016693,
      "num_tokens": 92836451.0,
      "step": 44730
    },
    {
      "entropy": 5.645005178451538,
      "epoch": 4.057964441219158,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003442702816842917,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22377101629972457,
      "num_tokens": 92846923.0,
      "step": 44735
    },
    {
      "entropy": 5.606642961502075,
      "epoch": 4.0584179970972425,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034423949612854974,
      "loss": 4.7842,
      "mean_token_accuracy": 0.22843304574489592,
      "num_tokens": 92857549.0,
      "step": 44740
    },
    {
      "entropy": 5.64547815322876,
      "epoch": 4.058871552975327,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003442087091409173,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22554138749837876,
      "num_tokens": 92867059.0,
      "step": 44745
    },
    {
      "entropy": 5.592003726959229,
      "epoch": 4.059325108853411,
      "grad_norm": 1.25,
      "learning_rate": 0.0003441779207220311,
      "loss": 4.8499,
      "mean_token_accuracy": 0.21685340851545334,
      "num_tokens": 92878138.0,
      "step": 44750
    },
    {
      "entropy": 5.614434385299683,
      "epoch": 4.059778664731495,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003441471308725279,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22340454906225204,
      "num_tokens": 92888931.0,
      "step": 44755
    },
    {
      "entropy": 5.648858308792114,
      "epoch": 4.060232220609579,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034411633959304444,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21605476438999177,
      "num_tokens": 92899718.0,
      "step": 44760
    },
    {
      "entropy": 5.642678594589233,
      "epoch": 4.060685776487663,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034408554688421743,
      "loss": 4.8025,
      "mean_token_accuracy": 0.2328519642353058,
      "num_tokens": 92910381.0,
      "step": 44765
    },
    {
      "entropy": 5.556215381622314,
      "epoch": 4.061139332365747,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003440547527466837,
      "loss": 4.8009,
      "mean_token_accuracy": 0.23354393392801284,
      "num_tokens": 92919905.0,
      "step": 44770
    },
    {
      "entropy": 5.516737604141236,
      "epoch": 4.0615928882438315,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00034402395718108016,
      "loss": 4.7908,
      "mean_token_accuracy": 0.23669462502002717,
      "num_tokens": 92930210.0,
      "step": 44775
    },
    {
      "entropy": 5.574579095840454,
      "epoch": 4.062046444121916,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034399316018804356,
      "loss": 4.8216,
      "mean_token_accuracy": 0.22903160154819488,
      "num_tokens": 92940832.0,
      "step": 44780
    },
    {
      "entropy": 5.694894027709961,
      "epoch": 4.0625,
      "grad_norm": 1.375,
      "learning_rate": 0.0003439623617682107,
      "loss": 4.94,
      "mean_token_accuracy": 0.21604613065719605,
      "num_tokens": 92951349.0,
      "step": 44785
    },
    {
      "entropy": 5.618463516235352,
      "epoch": 4.062953555878084,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003439315619222188,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22501133382320404,
      "num_tokens": 92961642.0,
      "step": 44790
    },
    {
      "entropy": 5.584527444839478,
      "epoch": 4.0634071117561685,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003439007606507046,
      "loss": 4.8515,
      "mean_token_accuracy": 0.23371994644403457,
      "num_tokens": 92971061.0,
      "step": 44795
    },
    {
      "entropy": 5.682553768157959,
      "epoch": 4.063860667634253,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003438699579543051,
      "loss": 4.948,
      "mean_token_accuracy": 0.21575381308794023,
      "num_tokens": 92981732.0,
      "step": 44800
    },
    {
      "entropy": 5.651588821411133,
      "epoch": 4.064314223512337,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00034383915383365726,
      "loss": 4.9186,
      "mean_token_accuracy": 0.22133039832115173,
      "num_tokens": 92991570.0,
      "step": 44805
    },
    {
      "entropy": 5.574545621871948,
      "epoch": 4.064767779390421,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003438083482893982,
      "loss": 4.8497,
      "mean_token_accuracy": 0.22949228435754776,
      "num_tokens": 93001623.0,
      "step": 44810
    },
    {
      "entropy": 5.62262716293335,
      "epoch": 4.065221335268505,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034377754132216496,
      "loss": 4.8669,
      "mean_token_accuracy": 0.2263652801513672,
      "num_tokens": 93012167.0,
      "step": 44815
    },
    {
      "entropy": 5.565275096893311,
      "epoch": 4.065674891146589,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003437467329325947,
      "loss": 4.7806,
      "mean_token_accuracy": 0.23684606254100798,
      "num_tokens": 93022090.0,
      "step": 44820
    },
    {
      "entropy": 5.594990682601929,
      "epoch": 4.066128447024673,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003437159231213244,
      "loss": 4.7163,
      "mean_token_accuracy": 0.23756403774023055,
      "num_tokens": 93030917.0,
      "step": 44825
    },
    {
      "entropy": 5.604778718948364,
      "epoch": 4.0665820029027575,
      "grad_norm": 1.375,
      "learning_rate": 0.0003436851118889914,
      "loss": 4.818,
      "mean_token_accuracy": 0.22406309992074966,
      "num_tokens": 93040113.0,
      "step": 44830
    },
    {
      "entropy": 5.606387186050415,
      "epoch": 4.067035558780842,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003436542992362326,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22789676636457443,
      "num_tokens": 93050053.0,
      "step": 44835
    },
    {
      "entropy": 5.662514591217041,
      "epoch": 4.067489114658926,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003436234851636856,
      "loss": 4.9171,
      "mean_token_accuracy": 0.21070526242256166,
      "num_tokens": 93060581.0,
      "step": 44840
    },
    {
      "entropy": 5.676267814636231,
      "epoch": 4.06794267053701,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034359266967198734,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2218409851193428,
      "num_tokens": 93070158.0,
      "step": 44845
    },
    {
      "entropy": 5.544855451583862,
      "epoch": 4.068396226415095,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003435618527617752,
      "loss": 4.7987,
      "mean_token_accuracy": 0.22716328352689744,
      "num_tokens": 93080804.0,
      "step": 44850
    },
    {
      "entropy": 5.623077774047852,
      "epoch": 4.068849782293179,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003435310344336865,
      "loss": 4.8208,
      "mean_token_accuracy": 0.2248335748910904,
      "num_tokens": 93091346.0,
      "step": 44855
    },
    {
      "entropy": 5.574751853942871,
      "epoch": 4.069303338171263,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003435002146883585,
      "loss": 4.7939,
      "mean_token_accuracy": 0.2331780195236206,
      "num_tokens": 93100931.0,
      "step": 44860
    },
    {
      "entropy": 5.57326979637146,
      "epoch": 4.0697568940493465,
      "grad_norm": 1.375,
      "learning_rate": 0.0003434693935264286,
      "loss": 4.8226,
      "mean_token_accuracy": 0.22805947810411453,
      "num_tokens": 93110181.0,
      "step": 44865
    },
    {
      "entropy": 5.639777803421021,
      "epoch": 4.070210449927431,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003434385709485343,
      "loss": 4.8003,
      "mean_token_accuracy": 0.22709694057703017,
      "num_tokens": 93120140.0,
      "step": 44870
    },
    {
      "entropy": 5.657888698577881,
      "epoch": 4.070664005805515,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003434077469553128,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2222645714879036,
      "num_tokens": 93130853.0,
      "step": 44875
    },
    {
      "entropy": 5.595185089111328,
      "epoch": 4.071117561683599,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003433769215474017,
      "loss": 4.82,
      "mean_token_accuracy": 0.2288125991821289,
      "num_tokens": 93141430.0,
      "step": 44880
    },
    {
      "entropy": 5.681582641601563,
      "epoch": 4.0715711175616836,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034334609472543845,
      "loss": 4.937,
      "mean_token_accuracy": 0.21889784783124924,
      "num_tokens": 93152270.0,
      "step": 44885
    },
    {
      "entropy": 5.67920503616333,
      "epoch": 4.072024673439768,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003433152664900605,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2194622799754143,
      "num_tokens": 93162850.0,
      "step": 44890
    },
    {
      "entropy": 5.597731351852417,
      "epoch": 4.072478229317852,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003432844368419055,
      "loss": 4.7315,
      "mean_token_accuracy": 0.23338056951761246,
      "num_tokens": 93172265.0,
      "step": 44895
    },
    {
      "entropy": 5.566187191009521,
      "epoch": 4.072931785195936,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003432536057816108,
      "loss": 4.7823,
      "mean_token_accuracy": 0.23623788952827454,
      "num_tokens": 93182756.0,
      "step": 44900
    },
    {
      "entropy": 5.600265026092529,
      "epoch": 4.073385341074021,
      "grad_norm": 1.375,
      "learning_rate": 0.00034322277330981426,
      "loss": 4.8544,
      "mean_token_accuracy": 0.2216146245598793,
      "num_tokens": 93192710.0,
      "step": 44905
    },
    {
      "entropy": 5.594810914993286,
      "epoch": 4.073838896952105,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034319193942715343,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23140716254711152,
      "num_tokens": 93202531.0,
      "step": 44910
    },
    {
      "entropy": 5.670984601974487,
      "epoch": 4.074292452830188,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034316110413426586,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22163503617048264,
      "num_tokens": 93212208.0,
      "step": 44915
    },
    {
      "entropy": 5.701069784164429,
      "epoch": 4.0747460087082725,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003431302674317892,
      "loss": 4.906,
      "mean_token_accuracy": 0.22078319042921066,
      "num_tokens": 93222932.0,
      "step": 44920
    },
    {
      "entropy": 5.602761936187744,
      "epoch": 4.075199564586357,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034309942932036136,
      "loss": 4.7429,
      "mean_token_accuracy": 0.23693148493766786,
      "num_tokens": 93232049.0,
      "step": 44925
    },
    {
      "entropy": 5.5965917110443115,
      "epoch": 4.075653120464441,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034306858980061994,
      "loss": 4.8676,
      "mean_token_accuracy": 0.22742598056793212,
      "num_tokens": 93244230.0,
      "step": 44930
    },
    {
      "entropy": 5.574835777282715,
      "epoch": 4.076106676342525,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034303774887320274,
      "loss": 4.8634,
      "mean_token_accuracy": 0.22267157435417176,
      "num_tokens": 93254159.0,
      "step": 44935
    },
    {
      "entropy": 5.602105331420899,
      "epoch": 4.07656023222061,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034300690653874753,
      "loss": 4.7669,
      "mean_token_accuracy": 0.2311967983841896,
      "num_tokens": 93263742.0,
      "step": 44940
    },
    {
      "entropy": 5.607001781463623,
      "epoch": 4.077013788098694,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034297606279789215,
      "loss": 4.9495,
      "mean_token_accuracy": 0.2125396728515625,
      "num_tokens": 93274446.0,
      "step": 44945
    },
    {
      "entropy": 5.619918727874756,
      "epoch": 4.077467343976778,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003429452176512745,
      "loss": 4.922,
      "mean_token_accuracy": 0.22096005082130432,
      "num_tokens": 93284102.0,
      "step": 44950
    },
    {
      "entropy": 5.5941229343414305,
      "epoch": 4.077920899854862,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034291437109953244,
      "loss": 4.7977,
      "mean_token_accuracy": 0.23862214982509614,
      "num_tokens": 93295096.0,
      "step": 44955
    },
    {
      "entropy": 5.588586807250977,
      "epoch": 4.078374455732947,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003428835231433039,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22536150068044664,
      "num_tokens": 93305377.0,
      "step": 44960
    },
    {
      "entropy": 5.707796287536621,
      "epoch": 4.07882801161103,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034285267378322674,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2246744900941849,
      "num_tokens": 93315859.0,
      "step": 44965
    },
    {
      "entropy": 5.621111679077148,
      "epoch": 4.079281567489114,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034282182301993903,
      "loss": 4.8052,
      "mean_token_accuracy": 0.23054684996604918,
      "num_tokens": 93326519.0,
      "step": 44970
    },
    {
      "entropy": 5.615607500076294,
      "epoch": 4.079735123367199,
      "grad_norm": 1.25,
      "learning_rate": 0.0003427909708540787,
      "loss": 4.8266,
      "mean_token_accuracy": 0.22684440165758132,
      "num_tokens": 93337055.0,
      "step": 44975
    },
    {
      "entropy": 5.615185117721557,
      "epoch": 4.080188679245283,
      "grad_norm": 1.25,
      "learning_rate": 0.00034276011728628393,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22140184044837952,
      "num_tokens": 93346565.0,
      "step": 44980
    },
    {
      "entropy": 5.597073793411255,
      "epoch": 4.080642235123367,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034272926231719255,
      "loss": 4.8439,
      "mean_token_accuracy": 0.22581083923578263,
      "num_tokens": 93356366.0,
      "step": 44985
    },
    {
      "entropy": 5.63536901473999,
      "epoch": 4.081095791001451,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003426984059474428,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2221727505326271,
      "num_tokens": 93367317.0,
      "step": 44990
    },
    {
      "entropy": 5.579619979858398,
      "epoch": 4.081549346879536,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003426675481776728,
      "loss": 4.8531,
      "mean_token_accuracy": 0.22672135084867479,
      "num_tokens": 93379601.0,
      "step": 44995
    },
    {
      "entropy": 5.598932266235352,
      "epoch": 4.08200290275762,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034263668900852065,
      "loss": 4.8177,
      "mean_token_accuracy": 0.23660581856966018,
      "num_tokens": 93389740.0,
      "step": 45000
    },
    {
      "epoch": 4.08200290275762,
      "eval_entropy": 5.36609978341409,
      "eval_loss": 4.948644161224365,
      "eval_mean_token_accuracy": 0.22913637396403297,
      "eval_num_tokens": 93389740.0,
      "eval_runtime": 21.2153,
      "eval_samples_per_second": 1666.726,
      "eval_steps_per_second": 208.341,
      "step": 45000
    },
    {
      "entropy": 5.710444498062134,
      "epoch": 4.082456458635704,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003426058284406244,
      "loss": 4.9773,
      "mean_token_accuracy": 0.2175597682595253,
      "num_tokens": 93400232.0,
      "step": 45005
    },
    {
      "entropy": 5.644765424728393,
      "epoch": 4.082910014513788,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003425749664746226,
      "loss": 4.8779,
      "mean_token_accuracy": 0.2313398540019989,
      "num_tokens": 93410662.0,
      "step": 45010
    },
    {
      "entropy": 5.593817806243896,
      "epoch": 4.083363570391873,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003425441031111532,
      "loss": 4.7336,
      "mean_token_accuracy": 0.2376194566488266,
      "num_tokens": 93420328.0,
      "step": 45015
    },
    {
      "entropy": 5.637911415100097,
      "epoch": 4.083817126269956,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034251323835085453,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21632947772741318,
      "num_tokens": 93430878.0,
      "step": 45020
    },
    {
      "entropy": 5.7028522968292235,
      "epoch": 4.08427068214804,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003424823721943649,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22205618917942047,
      "num_tokens": 93441784.0,
      "step": 45025
    },
    {
      "entropy": 5.586560821533203,
      "epoch": 4.084724238026125,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003424515046423226,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2274570271372795,
      "num_tokens": 93453058.0,
      "step": 45030
    },
    {
      "entropy": 5.647066354751587,
      "epoch": 4.085177793904209,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003424206356953659,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2251850873231888,
      "num_tokens": 93463089.0,
      "step": 45035
    },
    {
      "entropy": 5.644313907623291,
      "epoch": 4.085631349782293,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034238976535413333,
      "loss": 4.854,
      "mean_token_accuracy": 0.22697433680295945,
      "num_tokens": 93473315.0,
      "step": 45040
    },
    {
      "entropy": 5.695970964431763,
      "epoch": 4.086084905660377,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003423588936192633,
      "loss": 4.9513,
      "mean_token_accuracy": 0.21366531848907472,
      "num_tokens": 93485113.0,
      "step": 45045
    },
    {
      "entropy": 5.586354827880859,
      "epoch": 4.086538461538462,
      "grad_norm": 1.375,
      "learning_rate": 0.0003423280204913941,
      "loss": 4.7959,
      "mean_token_accuracy": 0.22895257622003556,
      "num_tokens": 93494269.0,
      "step": 45050
    },
    {
      "entropy": 5.5170365333557125,
      "epoch": 4.086992017416546,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003422971459711644,
      "loss": 4.7712,
      "mean_token_accuracy": 0.23256022930145265,
      "num_tokens": 93505052.0,
      "step": 45055
    },
    {
      "entropy": 5.584046220779419,
      "epoch": 4.08744557329463,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034226627005921244,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22312091588973998,
      "num_tokens": 93514986.0,
      "step": 45060
    },
    {
      "entropy": 5.612998580932617,
      "epoch": 4.0878991291727145,
      "grad_norm": 1.359375,
      "learning_rate": 0.000342235392756177,
      "loss": 4.8336,
      "mean_token_accuracy": 0.22653710842132568,
      "num_tokens": 93525286.0,
      "step": 45065
    },
    {
      "entropy": 5.6412458419799805,
      "epoch": 4.088352685050798,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034220451406269636,
      "loss": 4.8933,
      "mean_token_accuracy": 0.22484995275735856,
      "num_tokens": 93536034.0,
      "step": 45070
    },
    {
      "entropy": 5.556498432159424,
      "epoch": 4.088806240928882,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034217363397940937,
      "loss": 4.8231,
      "mean_token_accuracy": 0.2260088637471199,
      "num_tokens": 93546283.0,
      "step": 45075
    },
    {
      "entropy": 5.550257349014283,
      "epoch": 4.089259796806966,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034214275250695446,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22520696073770524,
      "num_tokens": 93555811.0,
      "step": 45080
    },
    {
      "entropy": 5.652612257003784,
      "epoch": 4.089713352685051,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003421118696459703,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22132974714040757,
      "num_tokens": 93567154.0,
      "step": 45085
    },
    {
      "entropy": 5.576481866836548,
      "epoch": 4.090166908563135,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034208098539709566,
      "loss": 4.8081,
      "mean_token_accuracy": 0.2351209968328476,
      "num_tokens": 93577654.0,
      "step": 45090
    },
    {
      "entropy": 5.597405958175659,
      "epoch": 4.090620464441219,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034205009976096906,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22180677056312562,
      "num_tokens": 93588712.0,
      "step": 45095
    },
    {
      "entropy": 5.607003450393677,
      "epoch": 4.0910740203193035,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003420192127382293,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22677280157804489,
      "num_tokens": 93599186.0,
      "step": 45100
    },
    {
      "entropy": 5.566450786590576,
      "epoch": 4.091527576197388,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034198832432951517,
      "loss": 4.7618,
      "mean_token_accuracy": 0.23677219897508622,
      "num_tokens": 93609439.0,
      "step": 45105
    },
    {
      "entropy": 5.61141357421875,
      "epoch": 4.091981132075472,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003419574345354654,
      "loss": 4.8394,
      "mean_token_accuracy": 0.22290150970220565,
      "num_tokens": 93619903.0,
      "step": 45110
    },
    {
      "entropy": 5.593566179275513,
      "epoch": 4.092434687953556,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034192654335671893,
      "loss": 4.7891,
      "mean_token_accuracy": 0.23174709379673003,
      "num_tokens": 93629895.0,
      "step": 45115
    },
    {
      "entropy": 5.62216010093689,
      "epoch": 4.09288824383164,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034189565079391436,
      "loss": 4.9324,
      "mean_token_accuracy": 0.21850364357233049,
      "num_tokens": 93640677.0,
      "step": 45120
    },
    {
      "entropy": 5.639234447479248,
      "epoch": 4.093341799709724,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003418647568476907,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22498167008161546,
      "num_tokens": 93651541.0,
      "step": 45125
    },
    {
      "entropy": 5.6587926864624025,
      "epoch": 4.093795355587808,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003418338615186869,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22495889961719512,
      "num_tokens": 93661901.0,
      "step": 45130
    },
    {
      "entropy": 5.607352924346924,
      "epoch": 4.094248911465892,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034180296480754174,
      "loss": 4.8266,
      "mean_token_accuracy": 0.22662619203329087,
      "num_tokens": 93672307.0,
      "step": 45135
    },
    {
      "entropy": 5.593094539642334,
      "epoch": 4.094702467343977,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003417720667148943,
      "loss": 4.808,
      "mean_token_accuracy": 0.2198405683040619,
      "num_tokens": 93683802.0,
      "step": 45140
    },
    {
      "entropy": 5.651087665557862,
      "epoch": 4.095156023222061,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003417411672413835,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21566251814365386,
      "num_tokens": 93693467.0,
      "step": 45145
    },
    {
      "entropy": 5.596789216995239,
      "epoch": 4.095609579100145,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034171026638764833,
      "loss": 4.8371,
      "mean_token_accuracy": 0.233934685587883,
      "num_tokens": 93703858.0,
      "step": 45150
    },
    {
      "entropy": 5.607862901687622,
      "epoch": 4.0960631349782295,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034167936415432785,
      "loss": 4.8331,
      "mean_token_accuracy": 0.23498803824186326,
      "num_tokens": 93714457.0,
      "step": 45155
    },
    {
      "entropy": 5.629857301712036,
      "epoch": 4.096516690856314,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003416484605420611,
      "loss": 4.8718,
      "mean_token_accuracy": 0.22573629766702652,
      "num_tokens": 93725009.0,
      "step": 45160
    },
    {
      "entropy": 5.656674814224243,
      "epoch": 4.096970246734398,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034161755555148725,
      "loss": 4.8889,
      "mean_token_accuracy": 0.21739200502634048,
      "num_tokens": 93736357.0,
      "step": 45165
    },
    {
      "entropy": 5.569134664535523,
      "epoch": 4.097423802612482,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034158664918324536,
      "loss": 4.8135,
      "mean_token_accuracy": 0.22434503734111785,
      "num_tokens": 93746711.0,
      "step": 45170
    },
    {
      "entropy": 5.63999662399292,
      "epoch": 4.097877358490566,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034155574143797455,
      "loss": 4.9177,
      "mean_token_accuracy": 0.22342435866594315,
      "num_tokens": 93757732.0,
      "step": 45175
    },
    {
      "entropy": 5.607414245605469,
      "epoch": 4.09833091436865,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034152483231631413,
      "loss": 4.8217,
      "mean_token_accuracy": 0.22524635195732118,
      "num_tokens": 93767048.0,
      "step": 45180
    },
    {
      "entropy": 5.734642267227173,
      "epoch": 4.098784470246734,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003414939218189032,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21261272132396697,
      "num_tokens": 93778283.0,
      "step": 45185
    },
    {
      "entropy": 5.678839778900146,
      "epoch": 4.0992380261248185,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000341463009946381,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22047129273414612,
      "num_tokens": 93788776.0,
      "step": 45190
    },
    {
      "entropy": 5.647962331771851,
      "epoch": 4.099691582002903,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034143209669938685,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22222113460302353,
      "num_tokens": 93798695.0,
      "step": 45195
    },
    {
      "entropy": 5.706395959854126,
      "epoch": 4.100145137880987,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034140118207856,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2230614736676216,
      "num_tokens": 93808993.0,
      "step": 45200
    },
    {
      "entropy": 5.695864343643189,
      "epoch": 4.100598693759071,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034137026608453973,
      "loss": 4.9702,
      "mean_token_accuracy": 0.21833188235759735,
      "num_tokens": 93818361.0,
      "step": 45205
    },
    {
      "entropy": 5.686879348754883,
      "epoch": 4.1010522496371555,
      "grad_norm": 1.25,
      "learning_rate": 0.0003413393487179655,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23343277275562285,
      "num_tokens": 93828988.0,
      "step": 45210
    },
    {
      "entropy": 5.71076135635376,
      "epoch": 4.10150580551524,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003413084299794767,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21829482316970825,
      "num_tokens": 93839350.0,
      "step": 45215
    },
    {
      "entropy": 5.649133825302124,
      "epoch": 4.101959361393324,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034127750986971255,
      "loss": 4.8254,
      "mean_token_accuracy": 0.22525520175695418,
      "num_tokens": 93849939.0,
      "step": 45220
    },
    {
      "entropy": 5.634019613265991,
      "epoch": 4.1024129172714074,
      "grad_norm": 1.25,
      "learning_rate": 0.0003412465883893126,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22501737475395203,
      "num_tokens": 93861278.0,
      "step": 45225
    },
    {
      "entropy": 5.656541252136231,
      "epoch": 4.102866473149492,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003412156655389164,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22136138528585433,
      "num_tokens": 93871693.0,
      "step": 45230
    },
    {
      "entropy": 5.619652318954468,
      "epoch": 4.103320029027576,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003411847413191633,
      "loss": 4.8714,
      "mean_token_accuracy": 0.22605828195810318,
      "num_tokens": 93883098.0,
      "step": 45235
    },
    {
      "entropy": 5.5858306884765625,
      "epoch": 4.10377358490566,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034115381573069297,
      "loss": 4.7786,
      "mean_token_accuracy": 0.22957106977701186,
      "num_tokens": 93893347.0,
      "step": 45240
    },
    {
      "entropy": 5.666557168960571,
      "epoch": 4.1042271407837445,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034112288877414485,
      "loss": 4.9143,
      "mean_token_accuracy": 0.2185812160372734,
      "num_tokens": 93903964.0,
      "step": 45245
    },
    {
      "entropy": 5.638802194595337,
      "epoch": 4.104680696661829,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003410919604501585,
      "loss": 4.8894,
      "mean_token_accuracy": 0.2264538064599037,
      "num_tokens": 93914526.0,
      "step": 45250
    },
    {
      "entropy": 5.680412769317627,
      "epoch": 4.105134252539913,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003410610307593735,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22640000581741332,
      "num_tokens": 93924430.0,
      "step": 45255
    },
    {
      "entropy": 5.664857053756714,
      "epoch": 4.105587808417997,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003410300997024297,
      "loss": 4.8457,
      "mean_token_accuracy": 0.2309491828083992,
      "num_tokens": 93935150.0,
      "step": 45260
    },
    {
      "entropy": 5.577393102645874,
      "epoch": 4.106041364296082,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003409991672799665,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23433299362659454,
      "num_tokens": 93944496.0,
      "step": 45265
    },
    {
      "entropy": 5.626067447662353,
      "epoch": 4.106494920174166,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034096823349262376,
      "loss": 4.8876,
      "mean_token_accuracy": 0.21984762996435164,
      "num_tokens": 93955745.0,
      "step": 45270
    },
    {
      "entropy": 5.588831377029419,
      "epoch": 4.106948476052249,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003409372983410411,
      "loss": 4.7869,
      "mean_token_accuracy": 0.22538850903511048,
      "num_tokens": 93965930.0,
      "step": 45275
    },
    {
      "entropy": 5.58941764831543,
      "epoch": 4.1074020319303335,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003409063618258583,
      "loss": 4.7725,
      "mean_token_accuracy": 0.23182172179222107,
      "num_tokens": 93976435.0,
      "step": 45280
    },
    {
      "entropy": 5.490685892105103,
      "epoch": 4.107855587808418,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034087542394771524,
      "loss": 4.7205,
      "mean_token_accuracy": 0.23362548649311066,
      "num_tokens": 93987294.0,
      "step": 45285
    },
    {
      "entropy": 5.602276277542114,
      "epoch": 4.108309143686502,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034084448470725147,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22531571686267854,
      "num_tokens": 93997177.0,
      "step": 45290
    },
    {
      "entropy": 5.717010641098023,
      "epoch": 4.108762699564586,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003408135441051071,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21590640246868134,
      "num_tokens": 94007663.0,
      "step": 45295
    },
    {
      "entropy": 5.642841053009033,
      "epoch": 4.1092162554426706,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003407826021419218,
      "loss": 4.7962,
      "mean_token_accuracy": 0.22395060509443282,
      "num_tokens": 94016899.0,
      "step": 45300
    },
    {
      "entropy": 5.6121316909790036,
      "epoch": 4.109669811320755,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003407516588183356,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2247838243842125,
      "num_tokens": 94027696.0,
      "step": 45305
    },
    {
      "entropy": 5.559666490554809,
      "epoch": 4.110123367198839,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003407207141349882,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22182020545005798,
      "num_tokens": 94038075.0,
      "step": 45310
    },
    {
      "entropy": 5.650737190246582,
      "epoch": 4.110576923076923,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003406897680925199,
      "loss": 4.8893,
      "mean_token_accuracy": 0.21868476420640945,
      "num_tokens": 94049082.0,
      "step": 45315
    },
    {
      "entropy": 5.63407564163208,
      "epoch": 4.111030478955008,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034065882069157023,
      "loss": 4.8827,
      "mean_token_accuracy": 0.21910038143396376,
      "num_tokens": 94060367.0,
      "step": 45320
    },
    {
      "entropy": 5.593263769149781,
      "epoch": 4.111484034833092,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034062787193277954,
      "loss": 4.8198,
      "mean_token_accuracy": 0.22931315004825592,
      "num_tokens": 94069415.0,
      "step": 45325
    },
    {
      "entropy": 5.646347761154175,
      "epoch": 4.111937590711175,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034059692181678775,
      "loss": 4.8752,
      "mean_token_accuracy": 0.21984780877828597,
      "num_tokens": 94079047.0,
      "step": 45330
    },
    {
      "entropy": 5.586153554916382,
      "epoch": 4.1123911465892595,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034056597034423484,
      "loss": 4.8025,
      "mean_token_accuracy": 0.23010508567094803,
      "num_tokens": 94089631.0,
      "step": 45335
    },
    {
      "entropy": 5.592344141006469,
      "epoch": 4.112844702467344,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034053501751576097,
      "loss": 4.8093,
      "mean_token_accuracy": 0.23061895370483398,
      "num_tokens": 94099265.0,
      "step": 45340
    },
    {
      "entropy": 5.5652707576751705,
      "epoch": 4.113298258345428,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003405040633320063,
      "loss": 4.8397,
      "mean_token_accuracy": 0.21754233986139299,
      "num_tokens": 94109673.0,
      "step": 45345
    },
    {
      "entropy": 5.582605218887329,
      "epoch": 4.113751814223512,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003404731077936108,
      "loss": 4.7602,
      "mean_token_accuracy": 0.23436962068080902,
      "num_tokens": 94120004.0,
      "step": 45350
    },
    {
      "entropy": 5.579771518707275,
      "epoch": 4.114205370101597,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003404421509012148,
      "loss": 4.8614,
      "mean_token_accuracy": 0.22182814478874208,
      "num_tokens": 94129935.0,
      "step": 45355
    },
    {
      "entropy": 5.639489364624024,
      "epoch": 4.114658925979681,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034041119265545847,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22056747674942018,
      "num_tokens": 94139035.0,
      "step": 45360
    },
    {
      "entropy": 5.640705394744873,
      "epoch": 4.115112481857765,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034038023305698196,
      "loss": 4.8299,
      "mean_token_accuracy": 0.222282837331295,
      "num_tokens": 94149281.0,
      "step": 45365
    },
    {
      "entropy": 5.585346937179565,
      "epoch": 4.115566037735849,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003403492721064256,
      "loss": 4.841,
      "mean_token_accuracy": 0.2291246846318245,
      "num_tokens": 94160455.0,
      "step": 45370
    },
    {
      "entropy": 5.635524654388428,
      "epoch": 4.116019593613934,
      "grad_norm": 1.375,
      "learning_rate": 0.00034031830980442955,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22047477215528488,
      "num_tokens": 94170334.0,
      "step": 45375
    },
    {
      "entropy": 5.588205337524414,
      "epoch": 4.116473149492017,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034028734615163436,
      "loss": 4.8022,
      "mean_token_accuracy": 0.22666919082403184,
      "num_tokens": 94181055.0,
      "step": 45380
    },
    {
      "entropy": 5.61060848236084,
      "epoch": 4.116926705370101,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003402563811486802,
      "loss": 4.8168,
      "mean_token_accuracy": 0.23083567917346953,
      "num_tokens": 94191182.0,
      "step": 45385
    },
    {
      "entropy": 5.630852937698364,
      "epoch": 4.117380261248186,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003402254147962073,
      "loss": 4.892,
      "mean_token_accuracy": 0.2232121467590332,
      "num_tokens": 94201229.0,
      "step": 45390
    },
    {
      "entropy": 5.5671977519989015,
      "epoch": 4.11783381712627,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003401944470948563,
      "loss": 4.8266,
      "mean_token_accuracy": 0.22683941274881364,
      "num_tokens": 94210956.0,
      "step": 45395
    },
    {
      "entropy": 5.615626382827759,
      "epoch": 4.118287373004354,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034016347804526753,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21808965504169464,
      "num_tokens": 94220740.0,
      "step": 45400
    },
    {
      "entropy": 5.545680952072144,
      "epoch": 4.118740928882438,
      "grad_norm": 1.125,
      "learning_rate": 0.00034013250764808133,
      "loss": 4.7386,
      "mean_token_accuracy": 0.24268535226583482,
      "num_tokens": 94231497.0,
      "step": 45405
    },
    {
      "entropy": 5.6001890182495115,
      "epoch": 4.119194484760523,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003401015359039384,
      "loss": 4.8733,
      "mean_token_accuracy": 0.22684278935194016,
      "num_tokens": 94240492.0,
      "step": 45410
    },
    {
      "entropy": 5.573484802246094,
      "epoch": 4.119648040638607,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003400705628134791,
      "loss": 4.8168,
      "mean_token_accuracy": 0.23307552188634872,
      "num_tokens": 94251586.0,
      "step": 45415
    },
    {
      "entropy": 5.596332550048828,
      "epoch": 4.120101596516691,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000340039588377344,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22828202247619628,
      "num_tokens": 94261635.0,
      "step": 45420
    },
    {
      "entropy": 5.654912996292114,
      "epoch": 4.120555152394775,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003400086125961735,
      "loss": 4.8793,
      "mean_token_accuracy": 0.2236225724220276,
      "num_tokens": 94269881.0,
      "step": 45425
    },
    {
      "entropy": 5.603367567062378,
      "epoch": 4.121008708272859,
      "grad_norm": 1.484375,
      "learning_rate": 0.00033997763547060847,
      "loss": 4.7995,
      "mean_token_accuracy": 0.22961757332086563,
      "num_tokens": 94279412.0,
      "step": 45430
    },
    {
      "entropy": 5.591487407684326,
      "epoch": 4.121462264150943,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003399466570012893,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22344161719083785,
      "num_tokens": 94290157.0,
      "step": 45435
    },
    {
      "entropy": 5.644675588607788,
      "epoch": 4.121915820029027,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033991567718885674,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22159571647644044,
      "num_tokens": 94299693.0,
      "step": 45440
    },
    {
      "entropy": 5.718258047103882,
      "epoch": 4.122369375907112,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003398846960339515,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21719103902578354,
      "num_tokens": 94310245.0,
      "step": 45445
    },
    {
      "entropy": 5.635201644897461,
      "epoch": 4.122822931785196,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003398537135372141,
      "loss": 4.8571,
      "mean_token_accuracy": 0.22374897450208664,
      "num_tokens": 94321190.0,
      "step": 45450
    },
    {
      "entropy": 5.662248420715332,
      "epoch": 4.12327648766328,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003398227296992854,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22854887694120407,
      "num_tokens": 94330273.0,
      "step": 45455
    },
    {
      "entropy": 5.691571760177612,
      "epoch": 4.123730043541364,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033979174452080616,
      "loss": 4.9643,
      "mean_token_accuracy": 0.214274799823761,
      "num_tokens": 94341097.0,
      "step": 45460
    },
    {
      "entropy": 5.617294502258301,
      "epoch": 4.124183599419449,
      "grad_norm": 1.25,
      "learning_rate": 0.0003397607580024171,
      "loss": 4.8362,
      "mean_token_accuracy": 0.22768158465623856,
      "num_tokens": 94352329.0,
      "step": 45465
    },
    {
      "entropy": 5.60767970085144,
      "epoch": 4.124637155297533,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003397297701447591,
      "loss": 4.8171,
      "mean_token_accuracy": 0.2288493424654007,
      "num_tokens": 94362551.0,
      "step": 45470
    },
    {
      "entropy": 5.612308597564697,
      "epoch": 4.125090711175617,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000339698780948473,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2220309406518936,
      "num_tokens": 94372405.0,
      "step": 45475
    },
    {
      "entropy": 5.639502143859863,
      "epoch": 4.125544267053701,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003396677904141995,
      "loss": 4.8262,
      "mean_token_accuracy": 0.23347653299570084,
      "num_tokens": 94382347.0,
      "step": 45480
    },
    {
      "entropy": 5.585342025756836,
      "epoch": 4.125997822931785,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033963679854257965,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22623097896575928,
      "num_tokens": 94393428.0,
      "step": 45485
    },
    {
      "entropy": 5.619420003890991,
      "epoch": 4.126451378809869,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003396058053342544,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23079249262809753,
      "num_tokens": 94403941.0,
      "step": 45490
    },
    {
      "entropy": 5.626084327697754,
      "epoch": 4.126904934687953,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033957481078986465,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2249835506081581,
      "num_tokens": 94414206.0,
      "step": 45495
    },
    {
      "entropy": 5.637312936782837,
      "epoch": 4.127358490566038,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033954381491005126,
      "loss": 4.842,
      "mean_token_accuracy": 0.23241206705570222,
      "num_tokens": 94424481.0,
      "step": 45500
    },
    {
      "entropy": 5.599806642532348,
      "epoch": 4.127812046444122,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003395128176954554,
      "loss": 4.8179,
      "mean_token_accuracy": 0.23077050000429153,
      "num_tokens": 94434714.0,
      "step": 45505
    },
    {
      "entropy": 5.646073293685913,
      "epoch": 4.128265602322206,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000339481819146718,
      "loss": 4.9024,
      "mean_token_accuracy": 0.22023053616285324,
      "num_tokens": 94445267.0,
      "step": 45510
    },
    {
      "entropy": 5.662341642379761,
      "epoch": 4.1287191582002905,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033945081926448013,
      "loss": 4.8633,
      "mean_token_accuracy": 0.2238749921321869,
      "num_tokens": 94455724.0,
      "step": 45515
    },
    {
      "entropy": 5.65594630241394,
      "epoch": 4.129172714078375,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033941981804938294,
      "loss": 4.922,
      "mean_token_accuracy": 0.21989777833223342,
      "num_tokens": 94466655.0,
      "step": 45520
    },
    {
      "entropy": 5.571654033660889,
      "epoch": 4.129626269956459,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033938881550206744,
      "loss": 4.8064,
      "mean_token_accuracy": 0.23060028553009032,
      "num_tokens": 94477699.0,
      "step": 45525
    },
    {
      "entropy": 5.639128065109253,
      "epoch": 4.130079825834542,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033935781162317485,
      "loss": 4.8531,
      "mean_token_accuracy": 0.22494232058525085,
      "num_tokens": 94487639.0,
      "step": 45530
    },
    {
      "entropy": 5.742417192459106,
      "epoch": 4.130533381712627,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033932680641334635,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21176039427518845,
      "num_tokens": 94498687.0,
      "step": 45535
    },
    {
      "entropy": 5.652783441543579,
      "epoch": 4.130986937590711,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003392957998732231,
      "loss": 4.8257,
      "mean_token_accuracy": 0.22805264890193938,
      "num_tokens": 94508706.0,
      "step": 45540
    },
    {
      "entropy": 5.5647481918334964,
      "epoch": 4.131440493468795,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033926479200344626,
      "loss": 4.7688,
      "mean_token_accuracy": 0.23213642388582229,
      "num_tokens": 94517246.0,
      "step": 45545
    },
    {
      "entropy": 5.564794063568115,
      "epoch": 4.131894049346879,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003392337828046572,
      "loss": 4.7715,
      "mean_token_accuracy": 0.23055853545665742,
      "num_tokens": 94527956.0,
      "step": 45550
    },
    {
      "entropy": 5.643964004516602,
      "epoch": 4.132347605224964,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003392027722774971,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21730686128139495,
      "num_tokens": 94539101.0,
      "step": 45555
    },
    {
      "entropy": 5.575935745239258,
      "epoch": 4.132801161103048,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033917176042260735,
      "loss": 4.8055,
      "mean_token_accuracy": 0.2231568768620491,
      "num_tokens": 94549597.0,
      "step": 45560
    },
    {
      "entropy": 5.658811426162719,
      "epoch": 4.133254716981132,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003391407472406292,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22433864921331406,
      "num_tokens": 94559911.0,
      "step": 45565
    },
    {
      "entropy": 5.580729246139526,
      "epoch": 4.1337082728592165,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033910973273220416,
      "loss": 4.794,
      "mean_token_accuracy": 0.22668695896863938,
      "num_tokens": 94569396.0,
      "step": 45570
    },
    {
      "entropy": 5.595262908935547,
      "epoch": 4.134161828737301,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033907871689797337,
      "loss": 4.8972,
      "mean_token_accuracy": 0.23455162942409516,
      "num_tokens": 94579817.0,
      "step": 45575
    },
    {
      "entropy": 5.528309059143067,
      "epoch": 4.134615384615385,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003390476997385785,
      "loss": 4.8145,
      "mean_token_accuracy": 0.22697588801383972,
      "num_tokens": 94590398.0,
      "step": 45580
    },
    {
      "entropy": 5.582088613510132,
      "epoch": 4.135068940493468,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003390166812546608,
      "loss": 4.8406,
      "mean_token_accuracy": 0.2227940618991852,
      "num_tokens": 94600528.0,
      "step": 45585
    },
    {
      "entropy": 5.622423362731934,
      "epoch": 4.135522496371553,
      "grad_norm": 1.25,
      "learning_rate": 0.00033898566144686184,
      "loss": 4.8677,
      "mean_token_accuracy": 0.22018461376428605,
      "num_tokens": 94611107.0,
      "step": 45590
    },
    {
      "entropy": 5.671023368835449,
      "epoch": 4.135976052249637,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003389546403158231,
      "loss": 4.9198,
      "mean_token_accuracy": 0.21760229468345643,
      "num_tokens": 94620932.0,
      "step": 45595
    },
    {
      "entropy": 5.670893430709839,
      "epoch": 4.136429608127721,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033892361786218604,
      "loss": 4.937,
      "mean_token_accuracy": 0.21279033422470092,
      "num_tokens": 94630476.0,
      "step": 45600
    },
    {
      "entropy": 5.578131294250488,
      "epoch": 4.1368831640058055,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003388925940865924,
      "loss": 4.8534,
      "mean_token_accuracy": 0.2264867141842842,
      "num_tokens": 94641487.0,
      "step": 45605
    },
    {
      "entropy": 5.642728519439697,
      "epoch": 4.13733671988389,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033886156898968353,
      "loss": 4.7656,
      "mean_token_accuracy": 0.2355020150542259,
      "num_tokens": 94651668.0,
      "step": 45610
    },
    {
      "entropy": 5.608333396911621,
      "epoch": 4.137790275761974,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033883054257210116,
      "loss": 4.8836,
      "mean_token_accuracy": 0.21826896965503692,
      "num_tokens": 94661218.0,
      "step": 45615
    },
    {
      "entropy": 5.54527096748352,
      "epoch": 4.138243831640058,
      "grad_norm": 1.265625,
      "learning_rate": 0.000338799514834487,
      "loss": 4.8159,
      "mean_token_accuracy": 0.2202413335442543,
      "num_tokens": 94672412.0,
      "step": 45620
    },
    {
      "entropy": 5.632928800582886,
      "epoch": 4.1386973875181425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003387684857774824,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21495541036128998,
      "num_tokens": 94683102.0,
      "step": 45625
    },
    {
      "entropy": 5.6283186912536625,
      "epoch": 4.139150943396227,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003387374554017294,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22799694091081618,
      "num_tokens": 94692869.0,
      "step": 45630
    },
    {
      "entropy": 5.622475051879883,
      "epoch": 4.13960449927431,
      "grad_norm": 1.375,
      "learning_rate": 0.00033870642370786966,
      "loss": 4.8067,
      "mean_token_accuracy": 0.22699427753686904,
      "num_tokens": 94702437.0,
      "step": 45635
    },
    {
      "entropy": 5.577946043014526,
      "epoch": 4.1400580551523944,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033867539069654476,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22075816243886948,
      "num_tokens": 94713130.0,
      "step": 45640
    },
    {
      "entropy": 5.660130596160888,
      "epoch": 4.140511611030479,
      "grad_norm": 1.375,
      "learning_rate": 0.00033864435636839647,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2173692613840103,
      "num_tokens": 94723251.0,
      "step": 45645
    },
    {
      "entropy": 5.639151096343994,
      "epoch": 4.140965166908563,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003386133207240668,
      "loss": 4.8597,
      "mean_token_accuracy": 0.22838103622198105,
      "num_tokens": 94733805.0,
      "step": 45650
    },
    {
      "entropy": 5.593966388702393,
      "epoch": 4.141418722786647,
      "grad_norm": 1.375,
      "learning_rate": 0.0003385822837641974,
      "loss": 4.7803,
      "mean_token_accuracy": 0.22447696328163147,
      "num_tokens": 94743654.0,
      "step": 45655
    },
    {
      "entropy": 5.562912321090698,
      "epoch": 4.1418722786647315,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033855124548943025,
      "loss": 4.8225,
      "mean_token_accuracy": 0.22763589322566985,
      "num_tokens": 94754635.0,
      "step": 45660
    },
    {
      "entropy": 5.610265064239502,
      "epoch": 4.142325834542816,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00033852020590040705,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22455063164234162,
      "num_tokens": 94765371.0,
      "step": 45665
    },
    {
      "entropy": 5.573143053054809,
      "epoch": 4.1427793904209,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003384891649977698,
      "loss": 4.7858,
      "mean_token_accuracy": 0.23540292233228682,
      "num_tokens": 94775780.0,
      "step": 45670
    },
    {
      "entropy": 5.632145833969116,
      "epoch": 4.143232946298984,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033845812278216045,
      "loss": 4.9031,
      "mean_token_accuracy": 0.21835450679063798,
      "num_tokens": 94785345.0,
      "step": 45675
    },
    {
      "entropy": 5.563596296310425,
      "epoch": 4.143686502177069,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033842707925422095,
      "loss": 4.8151,
      "mean_token_accuracy": 0.22559059262275696,
      "num_tokens": 94796059.0,
      "step": 45680
    },
    {
      "entropy": 5.601997995376587,
      "epoch": 4.144140058055152,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033839603441459333,
      "loss": 4.8307,
      "mean_token_accuracy": 0.22807716727256774,
      "num_tokens": 94805867.0,
      "step": 45685
    },
    {
      "entropy": 5.57486138343811,
      "epoch": 4.144593613933236,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003383649882639195,
      "loss": 4.8358,
      "mean_token_accuracy": 0.22382062673568726,
      "num_tokens": 94817172.0,
      "step": 45690
    },
    {
      "entropy": 5.598248434066773,
      "epoch": 4.1450471698113205,
      "grad_norm": 1.25,
      "learning_rate": 0.0003383339408028416,
      "loss": 4.8493,
      "mean_token_accuracy": 0.22410875409841538,
      "num_tokens": 94828748.0,
      "step": 45695
    },
    {
      "entropy": 5.608453464508057,
      "epoch": 4.145500725689405,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003383028920320017,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2250341072678566,
      "num_tokens": 94838704.0,
      "step": 45700
    },
    {
      "entropy": 5.6524840831756595,
      "epoch": 4.145954281567489,
      "grad_norm": 1.25,
      "learning_rate": 0.00033827184195204174,
      "loss": 4.92,
      "mean_token_accuracy": 0.2193364068865776,
      "num_tokens": 94849555.0,
      "step": 45705
    },
    {
      "entropy": 5.66926383972168,
      "epoch": 4.146407837445573,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003382407905636041,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2217804491519928,
      "num_tokens": 94862238.0,
      "step": 45710
    },
    {
      "entropy": 5.519079160690308,
      "epoch": 4.1468613933236576,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003382097378673308,
      "loss": 4.7381,
      "mean_token_accuracy": 0.23946598172187805,
      "num_tokens": 94872960.0,
      "step": 45715
    },
    {
      "entropy": 5.5377778053283695,
      "epoch": 4.147314949201742,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000338178683863864,
      "loss": 4.8245,
      "mean_token_accuracy": 0.22847038209438325,
      "num_tokens": 94883612.0,
      "step": 45720
    },
    {
      "entropy": 5.558489465713501,
      "epoch": 4.147768505079826,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003381476285538458,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23069373965263368,
      "num_tokens": 94893176.0,
      "step": 45725
    },
    {
      "entropy": 5.599456262588501,
      "epoch": 4.14822206095791,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003381165719379187,
      "loss": 4.7911,
      "mean_token_accuracy": 0.2260712653398514,
      "num_tokens": 94903860.0,
      "step": 45730
    },
    {
      "entropy": 5.585466527938843,
      "epoch": 4.148675616835995,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003380855140167247,
      "loss": 4.7905,
      "mean_token_accuracy": 0.23095814883708954,
      "num_tokens": 94915681.0,
      "step": 45735
    },
    {
      "entropy": 5.581074523925781,
      "epoch": 4.149129172714078,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003380544547909061,
      "loss": 4.8418,
      "mean_token_accuracy": 0.22475317120552063,
      "num_tokens": 94927062.0,
      "step": 45740
    },
    {
      "entropy": 5.578788137435913,
      "epoch": 4.149582728592162,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003380233942611055,
      "loss": 4.842,
      "mean_token_accuracy": 0.23134032934904097,
      "num_tokens": 94938838.0,
      "step": 45745
    },
    {
      "entropy": 5.543794631958008,
      "epoch": 4.1500362844702465,
      "grad_norm": 1.265625,
      "learning_rate": 0.000337992332427965,
      "loss": 4.7998,
      "mean_token_accuracy": 0.2270699843764305,
      "num_tokens": 94949678.0,
      "step": 45750
    },
    {
      "entropy": 5.620090818405151,
      "epoch": 4.150489840348331,
      "grad_norm": 1.359375,
      "learning_rate": 0.000337961269292127,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22675600796937942,
      "num_tokens": 94960314.0,
      "step": 45755
    },
    {
      "entropy": 5.637551307678223,
      "epoch": 4.150943396226415,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033793020485423385,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2249660849571228,
      "num_tokens": 94969831.0,
      "step": 45760
    },
    {
      "entropy": 5.611756801605225,
      "epoch": 4.151396952104499,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033789913911492814,
      "loss": 4.8422,
      "mean_token_accuracy": 0.22016721814870835,
      "num_tokens": 94980046.0,
      "step": 45765
    },
    {
      "entropy": 5.5561370849609375,
      "epoch": 4.151850507982584,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003378680720748522,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2225993365049362,
      "num_tokens": 94990850.0,
      "step": 45770
    },
    {
      "entropy": 5.640976905822754,
      "epoch": 4.152304063860668,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003378370037346484,
      "loss": 4.9026,
      "mean_token_accuracy": 0.218935127556324,
      "num_tokens": 95001658.0,
      "step": 45775
    },
    {
      "entropy": 5.609769248962403,
      "epoch": 4.152757619738752,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00033780593409495936,
      "loss": 4.8185,
      "mean_token_accuracy": 0.22747478783130645,
      "num_tokens": 95011912.0,
      "step": 45780
    },
    {
      "entropy": 5.602326822280884,
      "epoch": 4.153211175616836,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003377748631564276,
      "loss": 4.8285,
      "mean_token_accuracy": 0.22545932829380036,
      "num_tokens": 95021299.0,
      "step": 45785
    },
    {
      "entropy": 5.595681524276733,
      "epoch": 4.15366473149492,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033774379091969583,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23171577453613282,
      "num_tokens": 95031520.0,
      "step": 45790
    },
    {
      "entropy": 5.524148607254029,
      "epoch": 4.154118287373004,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033771271738540634,
      "loss": 4.7754,
      "mean_token_accuracy": 0.23335529267787933,
      "num_tokens": 95042127.0,
      "step": 45795
    },
    {
      "entropy": 5.603129100799561,
      "epoch": 4.154571843251088,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033768164255420184,
      "loss": 4.8357,
      "mean_token_accuracy": 0.22408168911933898,
      "num_tokens": 95052181.0,
      "step": 45800
    },
    {
      "entropy": 5.644991445541382,
      "epoch": 4.155025399129173,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00033765056642672507,
      "loss": 4.8793,
      "mean_token_accuracy": 0.2261883482336998,
      "num_tokens": 95062583.0,
      "step": 45805
    },
    {
      "entropy": 5.634037876129151,
      "epoch": 4.155478955007257,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033761948900361857,
      "loss": 4.8718,
      "mean_token_accuracy": 0.2233712390065193,
      "num_tokens": 95073250.0,
      "step": 45810
    },
    {
      "entropy": 5.648438405990601,
      "epoch": 4.155932510885341,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003375884102855251,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22480896413326262,
      "num_tokens": 95083544.0,
      "step": 45815
    },
    {
      "entropy": 5.569507360458374,
      "epoch": 4.156386066763425,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033755733027308746,
      "loss": 4.8458,
      "mean_token_accuracy": 0.23124369531869887,
      "num_tokens": 95093933.0,
      "step": 45820
    },
    {
      "entropy": 5.582110977172851,
      "epoch": 4.15683962264151,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003375262489669481,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22808850556612015,
      "num_tokens": 95104468.0,
      "step": 45825
    },
    {
      "entropy": 5.671491861343384,
      "epoch": 4.157293178519594,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033749516636775014,
      "loss": 4.8697,
      "mean_token_accuracy": 0.2247892677783966,
      "num_tokens": 95114285.0,
      "step": 45830
    },
    {
      "entropy": 5.672903919219971,
      "epoch": 4.157746734397678,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033746408247613616,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22037472873926162,
      "num_tokens": 95123984.0,
      "step": 45835
    },
    {
      "entropy": 5.657049703598022,
      "epoch": 4.1582002902757615,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000337432997292749,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2229042574763298,
      "num_tokens": 95135466.0,
      "step": 45840
    },
    {
      "entropy": 5.560149383544922,
      "epoch": 4.158653846153846,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003374019108182315,
      "loss": 4.7636,
      "mean_token_accuracy": 0.22950555980205536,
      "num_tokens": 95145223.0,
      "step": 45845
    },
    {
      "entropy": 5.675199937820435,
      "epoch": 4.15910740203193,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003373708230532267,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2283041924238205,
      "num_tokens": 95154916.0,
      "step": 45850
    },
    {
      "entropy": 5.594363594055176,
      "epoch": 4.159560957910014,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033733973399837735,
      "loss": 4.7886,
      "mean_token_accuracy": 0.22393823862075807,
      "num_tokens": 95164717.0,
      "step": 45855
    },
    {
      "entropy": 5.685700798034668,
      "epoch": 4.160014513788099,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003373086436543263,
      "loss": 4.8918,
      "mean_token_accuracy": 0.21790481805801393,
      "num_tokens": 95175544.0,
      "step": 45860
    },
    {
      "entropy": 5.59934253692627,
      "epoch": 4.160468069666183,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003372775520217168,
      "loss": 4.916,
      "mean_token_accuracy": 0.2247701659798622,
      "num_tokens": 95185997.0,
      "step": 45865
    },
    {
      "entropy": 5.584728765487671,
      "epoch": 4.160921625544267,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003372464591011916,
      "loss": 4.858,
      "mean_token_accuracy": 0.21790525317192078,
      "num_tokens": 95196967.0,
      "step": 45870
    },
    {
      "entropy": 5.612652730941773,
      "epoch": 4.161375181422351,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033721536489339374,
      "loss": 4.7876,
      "mean_token_accuracy": 0.2336038902401924,
      "num_tokens": 95206788.0,
      "step": 45875
    },
    {
      "entropy": 5.697801208496093,
      "epoch": 4.161828737300436,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033718426939896627,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2253885641694069,
      "num_tokens": 95217806.0,
      "step": 45880
    },
    {
      "entropy": 5.598578548431396,
      "epoch": 4.16228229317852,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033715317261855235,
      "loss": 4.8075,
      "mean_token_accuracy": 0.23400939255952835,
      "num_tokens": 95228311.0,
      "step": 45885
    },
    {
      "entropy": 5.5558905601501465,
      "epoch": 4.162735849056604,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033712207455279485,
      "loss": 4.828,
      "mean_token_accuracy": 0.22068298906087874,
      "num_tokens": 95238872.0,
      "step": 45890
    },
    {
      "entropy": 5.541109371185303,
      "epoch": 4.163189404934688,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003370909752023371,
      "loss": 4.7887,
      "mean_token_accuracy": 0.24139557629823685,
      "num_tokens": 95248575.0,
      "step": 45895
    },
    {
      "entropy": 5.6295863628387455,
      "epoch": 4.163642960812772,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003370598745678221,
      "loss": 4.875,
      "mean_token_accuracy": 0.22322934120893478,
      "num_tokens": 95259323.0,
      "step": 45900
    },
    {
      "entropy": 5.64022126197815,
      "epoch": 4.164096516690856,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003370287726498932,
      "loss": 4.8171,
      "mean_token_accuracy": 0.22574844658374787,
      "num_tokens": 95270018.0,
      "step": 45905
    },
    {
      "entropy": 5.616472387313843,
      "epoch": 4.16455007256894,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033699766944919325,
      "loss": 4.8156,
      "mean_token_accuracy": 0.22607270181179046,
      "num_tokens": 95279783.0,
      "step": 45910
    },
    {
      "entropy": 5.637238597869873,
      "epoch": 4.165003628447025,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033696656496636586,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22588370740413666,
      "num_tokens": 95291200.0,
      "step": 45915
    },
    {
      "entropy": 5.637813568115234,
      "epoch": 4.165457184325109,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033693545920205416,
      "loss": 4.9013,
      "mean_token_accuracy": 0.21645888388156892,
      "num_tokens": 95301641.0,
      "step": 45920
    },
    {
      "entropy": 5.558434963226318,
      "epoch": 4.165910740203193,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033690435215690125,
      "loss": 4.7665,
      "mean_token_accuracy": 0.22870970368385315,
      "num_tokens": 95312636.0,
      "step": 45925
    },
    {
      "entropy": 5.554112434387207,
      "epoch": 4.1663642960812775,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003368732438315506,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23078914880752563,
      "num_tokens": 95322168.0,
      "step": 45930
    },
    {
      "entropy": 5.543451738357544,
      "epoch": 4.166817851959362,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033684213422664554,
      "loss": 4.79,
      "mean_token_accuracy": 0.22979618608951569,
      "num_tokens": 95333162.0,
      "step": 45935
    },
    {
      "entropy": 5.5841187000274655,
      "epoch": 4.167271407837446,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003368110233428294,
      "loss": 4.8167,
      "mean_token_accuracy": 0.22716125398874282,
      "num_tokens": 95343806.0,
      "step": 45940
    },
    {
      "entropy": 5.594476699829102,
      "epoch": 4.167724963715529,
      "grad_norm": 1.25,
      "learning_rate": 0.00033677991118074554,
      "loss": 4.7724,
      "mean_token_accuracy": 0.23300998508930207,
      "num_tokens": 95353906.0,
      "step": 45945
    },
    {
      "entropy": 5.688653516769409,
      "epoch": 4.168178519593614,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003367487977410373,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2176999792456627,
      "num_tokens": 95363801.0,
      "step": 45950
    },
    {
      "entropy": 5.643514347076416,
      "epoch": 4.168632075471698,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033671768302434823,
      "loss": 4.8635,
      "mean_token_accuracy": 0.2222139984369278,
      "num_tokens": 95375012.0,
      "step": 45955
    },
    {
      "entropy": 5.641274452209473,
      "epoch": 4.169085631349782,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033668656703132174,
      "loss": 4.8736,
      "mean_token_accuracy": 0.22985648214817048,
      "num_tokens": 95386429.0,
      "step": 45960
    },
    {
      "entropy": 5.576644325256348,
      "epoch": 4.169539187227866,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003366554497626014,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2208377465605736,
      "num_tokens": 95396534.0,
      "step": 45965
    },
    {
      "entropy": 5.609455728530884,
      "epoch": 4.169992743105951,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033662433121883057,
      "loss": 4.8672,
      "mean_token_accuracy": 0.2271244779229164,
      "num_tokens": 95406621.0,
      "step": 45970
    },
    {
      "entropy": 5.612331533432007,
      "epoch": 4.170446298984035,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003365932114006528,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22392307072877884,
      "num_tokens": 95417587.0,
      "step": 45975
    },
    {
      "entropy": 5.531369972229004,
      "epoch": 4.170899854862119,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033656209030871184,
      "loss": 4.7398,
      "mean_token_accuracy": 0.23544736802577973,
      "num_tokens": 95428288.0,
      "step": 45980
    },
    {
      "entropy": 5.607578229904175,
      "epoch": 4.1713534107402035,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033653096794365115,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22697918564081193,
      "num_tokens": 95438093.0,
      "step": 45985
    },
    {
      "entropy": 5.532621574401856,
      "epoch": 4.171806966618288,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033649984430611437,
      "loss": 4.765,
      "mean_token_accuracy": 0.23091351389884948,
      "num_tokens": 95448397.0,
      "step": 45990
    },
    {
      "entropy": 5.677552223205566,
      "epoch": 4.172260522496371,
      "grad_norm": 1.25,
      "learning_rate": 0.00033646871939674506,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21065406799316405,
      "num_tokens": 95459533.0,
      "step": 45995
    },
    {
      "entropy": 5.616369962692261,
      "epoch": 4.172714078374455,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000336437593216187,
      "loss": 4.8815,
      "mean_token_accuracy": 0.2241342380642891,
      "num_tokens": 95469507.0,
      "step": 46000
    },
    {
      "entropy": 5.700064706802368,
      "epoch": 4.17316763425254,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003364064657650839,
      "loss": 4.9062,
      "mean_token_accuracy": 0.22164378464221954,
      "num_tokens": 95479742.0,
      "step": 46005
    },
    {
      "entropy": 5.546701383590698,
      "epoch": 4.173621190130624,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033637533704407936,
      "loss": 4.7883,
      "mean_token_accuracy": 0.23581935614347457,
      "num_tokens": 95489751.0,
      "step": 46010
    },
    {
      "entropy": 5.65342082977295,
      "epoch": 4.174074746008708,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003363442070538173,
      "loss": 4.8951,
      "mean_token_accuracy": 0.2164279818534851,
      "num_tokens": 95500580.0,
      "step": 46015
    },
    {
      "entropy": 5.684735584259033,
      "epoch": 4.1745283018867925,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033631307579494137,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22643557339906692,
      "num_tokens": 95511018.0,
      "step": 46020
    },
    {
      "entropy": 5.660071516036988,
      "epoch": 4.174981857764877,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033628194326809543,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22902738004922868,
      "num_tokens": 95521220.0,
      "step": 46025
    },
    {
      "entropy": 5.595278072357178,
      "epoch": 4.175435413642961,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033625080947392324,
      "loss": 4.8485,
      "mean_token_accuracy": 0.221150678396225,
      "num_tokens": 95531426.0,
      "step": 46030
    },
    {
      "entropy": 5.6453203678131105,
      "epoch": 4.175888969521045,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003362196744130687,
      "loss": 4.8696,
      "mean_token_accuracy": 0.23197310864925386,
      "num_tokens": 95542698.0,
      "step": 46035
    },
    {
      "entropy": 5.65197491645813,
      "epoch": 4.1763425253991295,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033618853808617573,
      "loss": 4.7847,
      "mean_token_accuracy": 0.23697241991758347,
      "num_tokens": 95553545.0,
      "step": 46040
    },
    {
      "entropy": 5.683568334579467,
      "epoch": 4.176796081277214,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003361574004938882,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22699520736932755,
      "num_tokens": 95564452.0,
      "step": 46045
    },
    {
      "entropy": 5.603371286392212,
      "epoch": 4.177249637155297,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033612626163685,
      "loss": 4.8784,
      "mean_token_accuracy": 0.2335369810461998,
      "num_tokens": 95574756.0,
      "step": 46050
    },
    {
      "entropy": 5.707676362991333,
      "epoch": 4.1777031930333814,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033609512151570517,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2178369402885437,
      "num_tokens": 95585147.0,
      "step": 46055
    },
    {
      "entropy": 5.664637660980224,
      "epoch": 4.178156748911466,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033606398013109765,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22499686926603318,
      "num_tokens": 95596025.0,
      "step": 46060
    },
    {
      "entropy": 5.605254125595093,
      "epoch": 4.17861030478955,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003360328374836715,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2207189157605171,
      "num_tokens": 95607191.0,
      "step": 46065
    },
    {
      "entropy": 5.6414680004119875,
      "epoch": 4.179063860667634,
      "grad_norm": 1.375,
      "learning_rate": 0.0003360016935740706,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22180553376674653,
      "num_tokens": 95617256.0,
      "step": 46070
    },
    {
      "entropy": 5.650389909744263,
      "epoch": 4.1795174165457185,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033597054840293927,
      "loss": 4.895,
      "mean_token_accuracy": 0.2214730605483055,
      "num_tokens": 95627638.0,
      "step": 46075
    },
    {
      "entropy": 5.617570304870606,
      "epoch": 4.179970972423803,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003359394019709214,
      "loss": 4.8009,
      "mean_token_accuracy": 0.23019987791776658,
      "num_tokens": 95637662.0,
      "step": 46080
    },
    {
      "entropy": 5.606051015853882,
      "epoch": 4.180424528301887,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033590825427866117,
      "loss": 4.8229,
      "mean_token_accuracy": 0.22204454690217973,
      "num_tokens": 95647505.0,
      "step": 46085
    },
    {
      "entropy": 5.584885215759277,
      "epoch": 4.180878084179971,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003358771053268027,
      "loss": 4.7797,
      "mean_token_accuracy": 0.21810621470212938,
      "num_tokens": 95657810.0,
      "step": 46090
    },
    {
      "entropy": 5.5727630138397215,
      "epoch": 4.181331640058056,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033584595511599024,
      "loss": 4.8052,
      "mean_token_accuracy": 0.2289726361632347,
      "num_tokens": 95667838.0,
      "step": 46095
    },
    {
      "entropy": 5.6028602600097654,
      "epoch": 4.181785195936139,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003358148036468679,
      "loss": 4.8372,
      "mean_token_accuracy": 0.22938428670167924,
      "num_tokens": 95678497.0,
      "step": 46100
    },
    {
      "entropy": 5.565297842025757,
      "epoch": 4.182238751814223,
      "grad_norm": 1.375,
      "learning_rate": 0.00033578365092007986,
      "loss": 4.7651,
      "mean_token_accuracy": 0.22888371497392654,
      "num_tokens": 95688656.0,
      "step": 46105
    },
    {
      "entropy": 5.6484519958496096,
      "epoch": 4.1826923076923075,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003357524969362705,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21997195184230806,
      "num_tokens": 95699156.0,
      "step": 46110
    },
    {
      "entropy": 5.636931991577148,
      "epoch": 4.183145863570392,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033572134169608395,
      "loss": 4.8628,
      "mean_token_accuracy": 0.23053984493017196,
      "num_tokens": 95709680.0,
      "step": 46115
    },
    {
      "entropy": 5.726464891433716,
      "epoch": 4.183599419448476,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033569018520016465,
      "loss": 4.878,
      "mean_token_accuracy": 0.22551177442073822,
      "num_tokens": 95719142.0,
      "step": 46120
    },
    {
      "entropy": 5.669403219223023,
      "epoch": 4.18405297532656,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003356590274491568,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21749901920557022,
      "num_tokens": 95729415.0,
      "step": 46125
    },
    {
      "entropy": 5.560628461837768,
      "epoch": 4.1845065312046446,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003356278684437049,
      "loss": 4.8496,
      "mean_token_accuracy": 0.22188816219568253,
      "num_tokens": 95739775.0,
      "step": 46130
    },
    {
      "entropy": 5.631206464767456,
      "epoch": 4.184960087082729,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033559670818445315,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22157105952501296,
      "num_tokens": 95751540.0,
      "step": 46135
    },
    {
      "entropy": 5.723659801483154,
      "epoch": 4.185413642960813,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000335565546672046,
      "loss": 5.0343,
      "mean_token_accuracy": 0.22000080496072769,
      "num_tokens": 95761885.0,
      "step": 46140
    },
    {
      "entropy": 5.688881492614746,
      "epoch": 4.185867198838897,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000335534383907128,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22621696889400483,
      "num_tokens": 95772284.0,
      "step": 46145
    },
    {
      "entropy": 5.615299892425537,
      "epoch": 4.186320754716981,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003355032198903434,
      "loss": 4.7933,
      "mean_token_accuracy": 0.22674741894006728,
      "num_tokens": 95783373.0,
      "step": 46150
    },
    {
      "entropy": 5.53031325340271,
      "epoch": 4.186774310595065,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033547205462233694,
      "loss": 4.7451,
      "mean_token_accuracy": 0.23516739308834075,
      "num_tokens": 95793226.0,
      "step": 46155
    },
    {
      "entropy": 5.478276348114013,
      "epoch": 4.187227866473149,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033544088810375287,
      "loss": 4.7657,
      "mean_token_accuracy": 0.22290200740098953,
      "num_tokens": 95803559.0,
      "step": 46160
    },
    {
      "entropy": 5.567489528656006,
      "epoch": 4.1876814223512335,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003354097203352358,
      "loss": 4.8381,
      "mean_token_accuracy": 0.2199566587805748,
      "num_tokens": 95814620.0,
      "step": 46165
    },
    {
      "entropy": 5.565248250961304,
      "epoch": 4.188134978229318,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003353785513174305,
      "loss": 4.8271,
      "mean_token_accuracy": 0.23366139829158783,
      "num_tokens": 95825024.0,
      "step": 46170
    },
    {
      "entropy": 5.650697135925293,
      "epoch": 4.188588534107402,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033534738105098124,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23154536187648772,
      "num_tokens": 95837373.0,
      "step": 46175
    },
    {
      "entropy": 5.651026916503906,
      "epoch": 4.189042089985486,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033531620953653276,
      "loss": 4.8613,
      "mean_token_accuracy": 0.2234511435031891,
      "num_tokens": 95846962.0,
      "step": 46180
    },
    {
      "entropy": 5.64197883605957,
      "epoch": 4.189495645863571,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003352850367747297,
      "loss": 4.85,
      "mean_token_accuracy": 0.22786208838224412,
      "num_tokens": 95857492.0,
      "step": 46185
    },
    {
      "entropy": 5.584845352172851,
      "epoch": 4.189949201741655,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003352538627662168,
      "loss": 4.7973,
      "mean_token_accuracy": 0.22742410004138947,
      "num_tokens": 95867903.0,
      "step": 46190
    },
    {
      "entropy": 5.680199432373047,
      "epoch": 4.190402757619739,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003352226875116386,
      "loss": 4.9611,
      "mean_token_accuracy": 0.22232498824596406,
      "num_tokens": 95878720.0,
      "step": 46195
    },
    {
      "entropy": 5.628942489624023,
      "epoch": 4.1908563134978225,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003351915110116399,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2179182857275009,
      "num_tokens": 95888406.0,
      "step": 46200
    },
    {
      "entropy": 5.619317293167114,
      "epoch": 4.191309869375907,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033516033326686537,
      "loss": 4.875,
      "mean_token_accuracy": 0.2120461568236351,
      "num_tokens": 95898431.0,
      "step": 46205
    },
    {
      "entropy": 5.645869255065918,
      "epoch": 4.191763425253991,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033512915427795975,
      "loss": 4.8291,
      "mean_token_accuracy": 0.23380981236696244,
      "num_tokens": 95909089.0,
      "step": 46210
    },
    {
      "entropy": 5.5506126403808596,
      "epoch": 4.192216981132075,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000335097974045568,
      "loss": 4.8197,
      "mean_token_accuracy": 0.23206972181797028,
      "num_tokens": 95919787.0,
      "step": 46215
    },
    {
      "entropy": 5.644887828826905,
      "epoch": 4.19267053701016,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00033506679257033485,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22265405356884002,
      "num_tokens": 95930620.0,
      "step": 46220
    },
    {
      "entropy": 5.647461748123169,
      "epoch": 4.193124092888244,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003350356098529051,
      "loss": 4.8166,
      "mean_token_accuracy": 0.2285475715994835,
      "num_tokens": 95940258.0,
      "step": 46225
    },
    {
      "entropy": 5.618821620941162,
      "epoch": 4.193577648766328,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003350044258939236,
      "loss": 4.917,
      "mean_token_accuracy": 0.22301074415445327,
      "num_tokens": 95950062.0,
      "step": 46230
    },
    {
      "entropy": 5.578625726699829,
      "epoch": 4.194031204644412,
      "grad_norm": 1.375,
      "learning_rate": 0.00033497324069403536,
      "loss": 4.7613,
      "mean_token_accuracy": 0.23030263781547547,
      "num_tokens": 95960658.0,
      "step": 46235
    },
    {
      "entropy": 5.6136847019195555,
      "epoch": 4.194484760522497,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003349420542538851,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2284703478217125,
      "num_tokens": 95970497.0,
      "step": 46240
    },
    {
      "entropy": 5.627645826339721,
      "epoch": 4.194938316400581,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033491086657411797,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2201028972864151,
      "num_tokens": 95980405.0,
      "step": 46245
    },
    {
      "entropy": 5.681105709075927,
      "epoch": 4.195391872278664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003348796776553788,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21626714020967483,
      "num_tokens": 95991111.0,
      "step": 46250
    },
    {
      "entropy": 5.69240436553955,
      "epoch": 4.1958454281567485,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003348484874983127,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2281465694308281,
      "num_tokens": 96001962.0,
      "step": 46255
    },
    {
      "entropy": 5.628242635726929,
      "epoch": 4.196298984034833,
      "grad_norm": 1.25,
      "learning_rate": 0.0003348172961035646,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22331026941537857,
      "num_tokens": 96011774.0,
      "step": 46260
    },
    {
      "entropy": 5.594936656951904,
      "epoch": 4.196752539912917,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003347861034717795,
      "loss": 4.8738,
      "mean_token_accuracy": 0.21719496697187424,
      "num_tokens": 96021915.0,
      "step": 46265
    },
    {
      "entropy": 5.572141695022583,
      "epoch": 4.197206095791001,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033475490960360266,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22516145408153534,
      "num_tokens": 96032851.0,
      "step": 46270
    },
    {
      "entropy": 5.572245931625366,
      "epoch": 4.197659651669086,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000334723714499679,
      "loss": 4.8446,
      "mean_token_accuracy": 0.2316009894013405,
      "num_tokens": 96042630.0,
      "step": 46275
    },
    {
      "entropy": 5.586907863616943,
      "epoch": 4.19811320754717,
      "grad_norm": 1.5,
      "learning_rate": 0.00033469251816065375,
      "loss": 4.7752,
      "mean_token_accuracy": 0.23266879022121428,
      "num_tokens": 96052154.0,
      "step": 46280
    },
    {
      "entropy": 5.62572979927063,
      "epoch": 4.198566763425254,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000334661320587172,
      "loss": 4.8426,
      "mean_token_accuracy": 0.22974378019571304,
      "num_tokens": 96062301.0,
      "step": 46285
    },
    {
      "entropy": 5.636858224868774,
      "epoch": 4.199020319303338,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000334630121779879,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22020387947559356,
      "num_tokens": 96072792.0,
      "step": 46290
    },
    {
      "entropy": 5.672989988327027,
      "epoch": 4.199473875181423,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033459892173941973,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22287983149290086,
      "num_tokens": 96082610.0,
      "step": 46295
    },
    {
      "entropy": 5.711738729476929,
      "epoch": 4.199927431059507,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003345677204664397,
      "loss": 4.9351,
      "mean_token_accuracy": 0.22334103435277938,
      "num_tokens": 96093095.0,
      "step": 46300
    },
    {
      "entropy": 5.599697780609131,
      "epoch": 4.20038098693759,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033453651796158397,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22855852246284486,
      "num_tokens": 96103218.0,
      "step": 46305
    },
    {
      "entropy": 5.59517469406128,
      "epoch": 4.200834542815675,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033450531422549786,
      "loss": 4.8197,
      "mean_token_accuracy": 0.23066069334745407,
      "num_tokens": 96113937.0,
      "step": 46310
    },
    {
      "entropy": 5.704042100906372,
      "epoch": 4.201288098693759,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00033447410925882673,
      "loss": 4.9983,
      "mean_token_accuracy": 0.21831278502941132,
      "num_tokens": 96124378.0,
      "step": 46315
    },
    {
      "entropy": 5.546307516098023,
      "epoch": 4.201741654571843,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003344429030622159,
      "loss": 4.7423,
      "mean_token_accuracy": 0.23220301270484925,
      "num_tokens": 96133657.0,
      "step": 46320
    },
    {
      "entropy": 5.572737169265747,
      "epoch": 4.202195210449927,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033441169563631055,
      "loss": 4.8569,
      "mean_token_accuracy": 0.22928023487329482,
      "num_tokens": 96143669.0,
      "step": 46325
    },
    {
      "entropy": 5.593768835067749,
      "epoch": 4.202648766328012,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033438048698175625,
      "loss": 4.8454,
      "mean_token_accuracy": 0.2256556198000908,
      "num_tokens": 96154782.0,
      "step": 46330
    },
    {
      "entropy": 5.628937530517578,
      "epoch": 4.203102322206096,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003343492770991984,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22844437658786773,
      "num_tokens": 96165128.0,
      "step": 46335
    },
    {
      "entropy": 5.605251693725586,
      "epoch": 4.20355587808418,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003343180659892824,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22459150850772858,
      "num_tokens": 96176909.0,
      "step": 46340
    },
    {
      "entropy": 5.595322895050049,
      "epoch": 4.2040094339622645,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003342868536526536,
      "loss": 4.8292,
      "mean_token_accuracy": 0.2325044170022011,
      "num_tokens": 96188287.0,
      "step": 46345
    },
    {
      "entropy": 5.5453588485717775,
      "epoch": 4.204462989840349,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033425564008995764,
      "loss": 4.8218,
      "mean_token_accuracy": 0.22864848524332046,
      "num_tokens": 96198852.0,
      "step": 46350
    },
    {
      "entropy": 5.5845318794250485,
      "epoch": 4.204916545718432,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003342244253018399,
      "loss": 4.8497,
      "mean_token_accuracy": 0.22177359759807586,
      "num_tokens": 96208111.0,
      "step": 46355
    },
    {
      "entropy": 5.603541517257691,
      "epoch": 4.205370101596516,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003341932092889459,
      "loss": 4.832,
      "mean_token_accuracy": 0.23271377831697465,
      "num_tokens": 96218925.0,
      "step": 46360
    },
    {
      "entropy": 5.65958423614502,
      "epoch": 4.205823657474601,
      "grad_norm": 1.375,
      "learning_rate": 0.0003341619920519213,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2287338137626648,
      "num_tokens": 96228573.0,
      "step": 46365
    },
    {
      "entropy": 5.6383861064910885,
      "epoch": 4.206277213352685,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033413077359141166,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22606369554996492,
      "num_tokens": 96239109.0,
      "step": 46370
    },
    {
      "entropy": 5.636392259597779,
      "epoch": 4.206730769230769,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003340995539080625,
      "loss": 4.9087,
      "mean_token_accuracy": 0.21735179871320726,
      "num_tokens": 96249286.0,
      "step": 46375
    },
    {
      "entropy": 5.573814249038696,
      "epoch": 4.207184325108853,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033406833300251947,
      "loss": 4.8427,
      "mean_token_accuracy": 0.2256662905216217,
      "num_tokens": 96259515.0,
      "step": 46380
    },
    {
      "entropy": 5.587550258636474,
      "epoch": 4.207637880986938,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003340371108754284,
      "loss": 4.787,
      "mean_token_accuracy": 0.23248307406902313,
      "num_tokens": 96269136.0,
      "step": 46385
    },
    {
      "entropy": 5.633391857147217,
      "epoch": 4.208091436865022,
      "grad_norm": 1.25,
      "learning_rate": 0.0003340058875274347,
      "loss": 4.906,
      "mean_token_accuracy": 0.22721685022115706,
      "num_tokens": 96279792.0,
      "step": 46390
    },
    {
      "entropy": 5.626743793487549,
      "epoch": 4.208544992743106,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003339746629591842,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2232711672782898,
      "num_tokens": 96289606.0,
      "step": 46395
    },
    {
      "entropy": 5.575702571868897,
      "epoch": 4.2089985486211905,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033394343717132275,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2215135470032692,
      "num_tokens": 96299903.0,
      "step": 46400
    },
    {
      "entropy": 5.67324743270874,
      "epoch": 4.209452104499274,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033391221016449595,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21800197660923004,
      "num_tokens": 96310179.0,
      "step": 46405
    },
    {
      "entropy": 5.6481225967407225,
      "epoch": 4.209905660377358,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033388098193934965,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2240140601992607,
      "num_tokens": 96320948.0,
      "step": 46410
    },
    {
      "entropy": 5.599363946914673,
      "epoch": 4.210359216255442,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003338497524965296,
      "loss": 4.8162,
      "mean_token_accuracy": 0.2256555289030075,
      "num_tokens": 96332046.0,
      "step": 46415
    },
    {
      "entropy": 5.595290660858154,
      "epoch": 4.210812772133527,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033381852183668166,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23457978665828705,
      "num_tokens": 96342327.0,
      "step": 46420
    },
    {
      "entropy": 5.577626752853393,
      "epoch": 4.211266328011611,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003337872899604517,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22184203565120697,
      "num_tokens": 96352048.0,
      "step": 46425
    },
    {
      "entropy": 5.581136846542359,
      "epoch": 4.211719883889695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033375605686848566,
      "loss": 4.7774,
      "mean_token_accuracy": 0.2347755402326584,
      "num_tokens": 96362917.0,
      "step": 46430
    },
    {
      "entropy": 5.596902561187744,
      "epoch": 4.2121734397677795,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003337248225614293,
      "loss": 4.8669,
      "mean_token_accuracy": 0.22739448696374892,
      "num_tokens": 96373261.0,
      "step": 46435
    },
    {
      "entropy": 5.649684619903565,
      "epoch": 4.212626995645864,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003336935870399288,
      "loss": 4.8937,
      "mean_token_accuracy": 0.21745210289955139,
      "num_tokens": 96384279.0,
      "step": 46440
    },
    {
      "entropy": 5.686035633087158,
      "epoch": 4.213080551523948,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003336623503046299,
      "loss": 4.8988,
      "mean_token_accuracy": 0.21928188502788543,
      "num_tokens": 96394071.0,
      "step": 46445
    },
    {
      "entropy": 5.6166078567504885,
      "epoch": 4.213534107402032,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00033363111235617854,
      "loss": 4.8575,
      "mean_token_accuracy": 0.23086592108011245,
      "num_tokens": 96404233.0,
      "step": 46450
    },
    {
      "entropy": 5.641084957122803,
      "epoch": 4.2139876632801165,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033359987319522095,
      "loss": 4.8597,
      "mean_token_accuracy": 0.23304994851350785,
      "num_tokens": 96415388.0,
      "step": 46455
    },
    {
      "entropy": 5.667003488540649,
      "epoch": 4.2144412191582,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033356863282240295,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22276459038257598,
      "num_tokens": 96425744.0,
      "step": 46460
    },
    {
      "entropy": 5.649685525894165,
      "epoch": 4.214894775036284,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003335373912383707,
      "loss": 4.8338,
      "mean_token_accuracy": 0.23602993339300155,
      "num_tokens": 96436340.0,
      "step": 46465
    },
    {
      "entropy": 5.6969311237335205,
      "epoch": 4.2153483309143684,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003335061484437704,
      "loss": 4.9761,
      "mean_token_accuracy": 0.21371111273765564,
      "num_tokens": 96447384.0,
      "step": 46470
    },
    {
      "entropy": 5.588261890411377,
      "epoch": 4.215801886792453,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000333474904439248,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23110016584396362,
      "num_tokens": 96456938.0,
      "step": 46475
    },
    {
      "entropy": 5.579414510726929,
      "epoch": 4.216255442670537,
      "grad_norm": 1.25,
      "learning_rate": 0.0003334436592254496,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22365673929452895,
      "num_tokens": 96467343.0,
      "step": 46480
    },
    {
      "entropy": 5.614346218109131,
      "epoch": 4.216708998548621,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033341241280302153,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22690063267946242,
      "num_tokens": 96477443.0,
      "step": 46485
    },
    {
      "entropy": 5.614342451095581,
      "epoch": 4.2171625544267055,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033338116517260985,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21869966834783555,
      "num_tokens": 96487711.0,
      "step": 46490
    },
    {
      "entropy": 5.715102195739746,
      "epoch": 4.21761611030479,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003333499163348608,
      "loss": 4.989,
      "mean_token_accuracy": 0.21690969914197922,
      "num_tokens": 96497839.0,
      "step": 46495
    },
    {
      "entropy": 5.674069929122925,
      "epoch": 4.218069666182874,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033331866629042055,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22335280627012252,
      "num_tokens": 96508237.0,
      "step": 46500
    },
    {
      "entropy": 5.599286413192749,
      "epoch": 4.218523222060958,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003332874150399355,
      "loss": 4.7658,
      "mean_token_accuracy": 0.22440260350704194,
      "num_tokens": 96518877.0,
      "step": 46505
    },
    {
      "entropy": 5.582112836837768,
      "epoch": 4.218976777939042,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003332561625840519,
      "loss": 4.8301,
      "mean_token_accuracy": 0.22762729376554489,
      "num_tokens": 96530734.0,
      "step": 46510
    },
    {
      "entropy": 5.650498723983764,
      "epoch": 4.219430333817126,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003332249089234159,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22359474897384643,
      "num_tokens": 96540983.0,
      "step": 46515
    },
    {
      "entropy": 5.65999174118042,
      "epoch": 4.21988388969521,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033319365405867394,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2213198259472847,
      "num_tokens": 96551866.0,
      "step": 46520
    },
    {
      "entropy": 5.63258695602417,
      "epoch": 4.2203374455732945,
      "grad_norm": 1.375,
      "learning_rate": 0.00033316239799047247,
      "loss": 4.8438,
      "mean_token_accuracy": 0.224632066488266,
      "num_tokens": 96562515.0,
      "step": 46525
    },
    {
      "entropy": 5.585324811935425,
      "epoch": 4.220791001451379,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033313114071945764,
      "loss": 4.8294,
      "mean_token_accuracy": 0.22754328548908234,
      "num_tokens": 96573266.0,
      "step": 46530
    },
    {
      "entropy": 5.581152009963989,
      "epoch": 4.221244557329463,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003330998822462762,
      "loss": 4.8078,
      "mean_token_accuracy": 0.2302495464682579,
      "num_tokens": 96583070.0,
      "step": 46535
    },
    {
      "entropy": 5.592003917694091,
      "epoch": 4.221698113207547,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033306862257157427,
      "loss": 4.8005,
      "mean_token_accuracy": 0.23104698359966278,
      "num_tokens": 96593582.0,
      "step": 46540
    },
    {
      "entropy": 5.632849550247192,
      "epoch": 4.2221516690856316,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003330373616959984,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23188619017601014,
      "num_tokens": 96603233.0,
      "step": 46545
    },
    {
      "entropy": 5.595374917984008,
      "epoch": 4.222605224963716,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033300609962019523,
      "loss": 4.8079,
      "mean_token_accuracy": 0.22700596153736113,
      "num_tokens": 96614361.0,
      "step": 46550
    },
    {
      "entropy": 5.631172037124633,
      "epoch": 4.2230587808418,
      "grad_norm": 1.359375,
      "learning_rate": 0.000332974836344811,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22467966377735138,
      "num_tokens": 96624161.0,
      "step": 46555
    },
    {
      "entropy": 5.667362403869629,
      "epoch": 4.2235123367198835,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033294357187049246,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2176077112555504,
      "num_tokens": 96634329.0,
      "step": 46560
    },
    {
      "entropy": 5.684721851348877,
      "epoch": 4.223965892597968,
      "grad_norm": 1.25,
      "learning_rate": 0.00033291230619788604,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22551367580890655,
      "num_tokens": 96645575.0,
      "step": 46565
    },
    {
      "entropy": 5.622018527984619,
      "epoch": 4.224419448476052,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033288103932763846,
      "loss": 4.8777,
      "mean_token_accuracy": 0.21770126670598983,
      "num_tokens": 96656224.0,
      "step": 46570
    },
    {
      "entropy": 5.649169206619263,
      "epoch": 4.224873004354136,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003328497712603962,
      "loss": 4.8357,
      "mean_token_accuracy": 0.23603827953338624,
      "num_tokens": 96665950.0,
      "step": 46575
    },
    {
      "entropy": 5.548826503753662,
      "epoch": 4.2253265602322205,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033281850199680586,
      "loss": 4.7317,
      "mean_token_accuracy": 0.23584811836481095,
      "num_tokens": 96676916.0,
      "step": 46580
    },
    {
      "entropy": 5.633476829528808,
      "epoch": 4.225780116110305,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003327872315375142,
      "loss": 4.9482,
      "mean_token_accuracy": 0.216978257894516,
      "num_tokens": 96687289.0,
      "step": 46585
    },
    {
      "entropy": 5.635323095321655,
      "epoch": 4.226233671988389,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003327559598831679,
      "loss": 4.884,
      "mean_token_accuracy": 0.232246695458889,
      "num_tokens": 96698149.0,
      "step": 46590
    },
    {
      "entropy": 5.617386722564698,
      "epoch": 4.226687227866473,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033272468703441363,
      "loss": 4.8393,
      "mean_token_accuracy": 0.2260541468858719,
      "num_tokens": 96709466.0,
      "step": 46595
    },
    {
      "entropy": 5.636447095870972,
      "epoch": 4.227140783744558,
      "grad_norm": 1.328125,
      "learning_rate": 0.000332693412991898,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22354019433259964,
      "num_tokens": 96720179.0,
      "step": 46600
    },
    {
      "entropy": 5.667103099822998,
      "epoch": 4.227594339622642,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003326621377562681,
      "loss": 4.915,
      "mean_token_accuracy": 0.21815070062875747,
      "num_tokens": 96730783.0,
      "step": 46605
    },
    {
      "entropy": 5.667810440063477,
      "epoch": 4.228047895500726,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033263086132817035,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21745525449514388,
      "num_tokens": 96740903.0,
      "step": 46610
    },
    {
      "entropy": 5.6163427352905275,
      "epoch": 4.2285014513788095,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033259958370825165,
      "loss": 4.8408,
      "mean_token_accuracy": 0.2320155084133148,
      "num_tokens": 96750600.0,
      "step": 46615
    },
    {
      "entropy": 5.607594966888428,
      "epoch": 4.228955007256894,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033256830489715905,
      "loss": 4.7968,
      "mean_token_accuracy": 0.23158927261829376,
      "num_tokens": 96760550.0,
      "step": 46620
    },
    {
      "entropy": 5.64610743522644,
      "epoch": 4.229408563134978,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003325370248955392,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22881616801023483,
      "num_tokens": 96770784.0,
      "step": 46625
    },
    {
      "entropy": 5.592218637466431,
      "epoch": 4.229862119013062,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003325057437040389,
      "loss": 4.779,
      "mean_token_accuracy": 0.23100534975528716,
      "num_tokens": 96780572.0,
      "step": 46630
    },
    {
      "entropy": 5.558398485183716,
      "epoch": 4.230315674891147,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003324744613233052,
      "loss": 4.7839,
      "mean_token_accuracy": 0.22809428721666336,
      "num_tokens": 96790504.0,
      "step": 46635
    },
    {
      "entropy": 5.607027769088745,
      "epoch": 4.230769230769231,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033244317775398506,
      "loss": 4.8517,
      "mean_token_accuracy": 0.2235510990023613,
      "num_tokens": 96800856.0,
      "step": 46640
    },
    {
      "entropy": 5.607737350463867,
      "epoch": 4.231222786647315,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003324118929967253,
      "loss": 4.8413,
      "mean_token_accuracy": 0.22567494809627534,
      "num_tokens": 96811293.0,
      "step": 46645
    },
    {
      "entropy": 5.545803785324097,
      "epoch": 4.231676342525399,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000332380607052173,
      "loss": 4.7748,
      "mean_token_accuracy": 0.23581669628620147,
      "num_tokens": 96821844.0,
      "step": 46650
    },
    {
      "entropy": 5.617425441741943,
      "epoch": 4.232129898403484,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003323493199209751,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22280262261629105,
      "num_tokens": 96833438.0,
      "step": 46655
    },
    {
      "entropy": 5.566705417633057,
      "epoch": 4.232583454281568,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003323180316037787,
      "loss": 4.7965,
      "mean_token_accuracy": 0.23031732738018035,
      "num_tokens": 96843086.0,
      "step": 46660
    },
    {
      "entropy": 5.622716808319092,
      "epoch": 4.233037010159651,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003322867421012308,
      "loss": 4.8382,
      "mean_token_accuracy": 0.225467449426651,
      "num_tokens": 96853064.0,
      "step": 46665
    },
    {
      "entropy": 5.722280883789063,
      "epoch": 4.2334905660377355,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003322554514139784,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22267532348632812,
      "num_tokens": 96864213.0,
      "step": 46670
    },
    {
      "entropy": 5.716294384002685,
      "epoch": 4.23394412191582,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033222415954266873,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21526502072811127,
      "num_tokens": 96875344.0,
      "step": 46675
    },
    {
      "entropy": 5.651243209838867,
      "epoch": 4.234397677793904,
      "grad_norm": 1.375,
      "learning_rate": 0.00033219286648794886,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22914940118789673,
      "num_tokens": 96885985.0,
      "step": 46680
    },
    {
      "entropy": 5.575485610961914,
      "epoch": 4.234851233671988,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003321615722504659,
      "loss": 4.7791,
      "mean_token_accuracy": 0.23782141208648683,
      "num_tokens": 96897512.0,
      "step": 46685
    },
    {
      "entropy": 5.632569694519043,
      "epoch": 4.235304789550073,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003321302768308671,
      "loss": 4.9268,
      "mean_token_accuracy": 0.21563323736190795,
      "num_tokens": 96907630.0,
      "step": 46690
    },
    {
      "entropy": 5.567236518859863,
      "epoch": 4.235758345428157,
      "grad_norm": 1.515625,
      "learning_rate": 0.00033209898022979964,
      "loss": 4.7823,
      "mean_token_accuracy": 0.23321677446365358,
      "num_tokens": 96917400.0,
      "step": 46695
    },
    {
      "entropy": 5.7093218803405765,
      "epoch": 4.236211901306241,
      "grad_norm": 1.375,
      "learning_rate": 0.0003320676824479106,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21596077382564544,
      "num_tokens": 96927700.0,
      "step": 46700
    },
    {
      "entropy": 5.681747579574585,
      "epoch": 4.236665457184325,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003320363834858474,
      "loss": 4.8215,
      "mean_token_accuracy": 0.2279474437236786,
      "num_tokens": 96938925.0,
      "step": 46705
    },
    {
      "entropy": 5.596373081207275,
      "epoch": 4.23711901306241,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003320050833442573,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23098774403333663,
      "num_tokens": 96948956.0,
      "step": 46710
    },
    {
      "entropy": 5.572036790847778,
      "epoch": 4.237572568940493,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003319737820237875,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22173832058906556,
      "num_tokens": 96958565.0,
      "step": 46715
    },
    {
      "entropy": 5.584075260162353,
      "epoch": 4.238026124818577,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033194247952508544,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22097350358963014,
      "num_tokens": 96968261.0,
      "step": 46720
    },
    {
      "entropy": 5.65211706161499,
      "epoch": 4.238479680696662,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033191117584879836,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22369531393051148,
      "num_tokens": 96978621.0,
      "step": 46725
    },
    {
      "entropy": 5.609901428222656,
      "epoch": 4.238933236574746,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003318798709955736,
      "loss": 4.8318,
      "mean_token_accuracy": 0.22153929322957994,
      "num_tokens": 96989081.0,
      "step": 46730
    },
    {
      "entropy": 5.7107829570770265,
      "epoch": 4.23938679245283,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003318485649660586,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2107331097126007,
      "num_tokens": 97000518.0,
      "step": 46735
    },
    {
      "entropy": 5.610215139389038,
      "epoch": 4.239840348330914,
      "grad_norm": 1.375,
      "learning_rate": 0.00033181725776090093,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2301383525133133,
      "num_tokens": 97012038.0,
      "step": 46740
    },
    {
      "entropy": 5.646110248565674,
      "epoch": 4.240293904208999,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033178594938074787,
      "loss": 4.8378,
      "mean_token_accuracy": 0.2226180449128151,
      "num_tokens": 97022256.0,
      "step": 46745
    },
    {
      "entropy": 5.6544441223144535,
      "epoch": 4.240747460087083,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003317546398262468,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21915076971054076,
      "num_tokens": 97033063.0,
      "step": 46750
    },
    {
      "entropy": 5.590554904937744,
      "epoch": 4.241201015965167,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003317233290980454,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22357482016086577,
      "num_tokens": 97043711.0,
      "step": 46755
    },
    {
      "entropy": 5.62206335067749,
      "epoch": 4.2416545718432515,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00033169201719679105,
      "loss": 4.8251,
      "mean_token_accuracy": 0.2312798485159874,
      "num_tokens": 97053582.0,
      "step": 46760
    },
    {
      "entropy": 5.6606837749481205,
      "epoch": 4.242108127721336,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033166070412313133,
      "loss": 4.881,
      "mean_token_accuracy": 0.22771119624376296,
      "num_tokens": 97064535.0,
      "step": 46765
    },
    {
      "entropy": 5.622524118423462,
      "epoch": 4.242561683599419,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003316293898777138,
      "loss": 4.886,
      "mean_token_accuracy": 0.2198437586426735,
      "num_tokens": 97074295.0,
      "step": 46770
    },
    {
      "entropy": 5.583196592330933,
      "epoch": 4.243015239477503,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033159807446118614,
      "loss": 4.8723,
      "mean_token_accuracy": 0.2204604908823967,
      "num_tokens": 97084781.0,
      "step": 46775
    },
    {
      "entropy": 5.6988152980804445,
      "epoch": 4.243468795355588,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003315667578741958,
      "loss": 4.9175,
      "mean_token_accuracy": 0.21672960370779037,
      "num_tokens": 97094848.0,
      "step": 46780
    },
    {
      "entropy": 5.639906501770019,
      "epoch": 4.243922351233672,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003315354401173905,
      "loss": 4.8578,
      "mean_token_accuracy": 0.22393720597028732,
      "num_tokens": 97104747.0,
      "step": 46785
    },
    {
      "entropy": 5.5903740406036375,
      "epoch": 4.244375907111756,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003315041211914179,
      "loss": 4.8171,
      "mean_token_accuracy": 0.22050128281116485,
      "num_tokens": 97115599.0,
      "step": 46790
    },
    {
      "entropy": 5.628864336013794,
      "epoch": 4.24482946298984,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003314728010969257,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22254079431295395,
      "num_tokens": 97127492.0,
      "step": 46795
    },
    {
      "entropy": 5.621570634841919,
      "epoch": 4.245283018867925,
      "grad_norm": 1.375,
      "learning_rate": 0.00033144147983456154,
      "loss": 4.8922,
      "mean_token_accuracy": 0.21863047182559966,
      "num_tokens": 97138336.0,
      "step": 46800
    },
    {
      "entropy": 5.606234979629517,
      "epoch": 4.245736574746009,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003314101574049732,
      "loss": 4.819,
      "mean_token_accuracy": 0.22951818108558655,
      "num_tokens": 97149196.0,
      "step": 46805
    },
    {
      "entropy": 5.669437599182129,
      "epoch": 4.246190130624093,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00033137883380880847,
      "loss": 4.911,
      "mean_token_accuracy": 0.22391146868467332,
      "num_tokens": 97159622.0,
      "step": 46810
    },
    {
      "entropy": 5.655715608596802,
      "epoch": 4.2466436865021775,
      "grad_norm": 1.28125,
      "learning_rate": 0.000331347509046715,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21366937160491944,
      "num_tokens": 97170149.0,
      "step": 46815
    },
    {
      "entropy": 5.67915711402893,
      "epoch": 4.247097242380261,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003313161831193408,
      "loss": 4.8783,
      "mean_token_accuracy": 0.22023363113403321,
      "num_tokens": 97180409.0,
      "step": 46820
    },
    {
      "entropy": 5.647928762435913,
      "epoch": 4.247550798258345,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033128485602733354,
      "loss": 4.9133,
      "mean_token_accuracy": 0.2233542025089264,
      "num_tokens": 97190899.0,
      "step": 46825
    },
    {
      "entropy": 5.652402639389038,
      "epoch": 4.248004354136429,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003312535277713411,
      "loss": 4.8936,
      "mean_token_accuracy": 0.23104675859212875,
      "num_tokens": 97200759.0,
      "step": 46830
    },
    {
      "entropy": 5.636211824417114,
      "epoch": 4.248457910014514,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033122219835201135,
      "loss": 4.8167,
      "mean_token_accuracy": 0.23237361907958984,
      "num_tokens": 97211476.0,
      "step": 46835
    },
    {
      "entropy": 5.684138679504395,
      "epoch": 4.248911465892598,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003311908677699923,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2201432392001152,
      "num_tokens": 97222374.0,
      "step": 46840
    },
    {
      "entropy": 5.630644226074219,
      "epoch": 4.249365021770682,
      "grad_norm": 1.375,
      "learning_rate": 0.0003311595360259317,
      "loss": 4.884,
      "mean_token_accuracy": 0.2238625779747963,
      "num_tokens": 97233079.0,
      "step": 46845
    },
    {
      "entropy": 5.526723623275757,
      "epoch": 4.2498185776487665,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033112820312047756,
      "loss": 4.7935,
      "mean_token_accuracy": 0.23256900012493134,
      "num_tokens": 97243796.0,
      "step": 46850
    },
    {
      "entropy": 5.591722869873047,
      "epoch": 4.250272133526851,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033109686905427794,
      "loss": 4.7516,
      "mean_token_accuracy": 0.23437889218330382,
      "num_tokens": 97253341.0,
      "step": 46855
    },
    {
      "entropy": 5.685336875915527,
      "epoch": 4.250725689404935,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033106553382798075,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21711051762104033,
      "num_tokens": 97264244.0,
      "step": 46860
    },
    {
      "entropy": 5.584280347824096,
      "epoch": 4.251179245283019,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033103419744223413,
      "loss": 4.8377,
      "mean_token_accuracy": 0.22694845050573348,
      "num_tokens": 97274239.0,
      "step": 46865
    },
    {
      "entropy": 5.557263422012329,
      "epoch": 4.251632801161103,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003310028598976859,
      "loss": 4.7531,
      "mean_token_accuracy": 0.2406819999217987,
      "num_tokens": 97284709.0,
      "step": 46870
    },
    {
      "entropy": 5.583110523223877,
      "epoch": 4.252086357039187,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033097152119498425,
      "loss": 4.794,
      "mean_token_accuracy": 0.22957405894994737,
      "num_tokens": 97295262.0,
      "step": 46875
    },
    {
      "entropy": 5.620219612121582,
      "epoch": 4.252539912917271,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003309401813347774,
      "loss": 4.8837,
      "mean_token_accuracy": 0.22242826074361802,
      "num_tokens": 97306143.0,
      "step": 46880
    },
    {
      "entropy": 5.647772598266601,
      "epoch": 4.2529934687953554,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003309088403177132,
      "loss": 4.7979,
      "mean_token_accuracy": 0.23372879922389983,
      "num_tokens": 97316248.0,
      "step": 46885
    },
    {
      "entropy": 5.699823570251465,
      "epoch": 4.25344702467344,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033087749814444,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22064039260149002,
      "num_tokens": 97327704.0,
      "step": 46890
    },
    {
      "entropy": 5.683367729187012,
      "epoch": 4.253900580551524,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033084615481560585,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2149379163980484,
      "num_tokens": 97337817.0,
      "step": 46895
    },
    {
      "entropy": 5.701856851577759,
      "epoch": 4.254354136429608,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033081481033185903,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22350728064775466,
      "num_tokens": 97348315.0,
      "step": 46900
    },
    {
      "entropy": 5.631969785690307,
      "epoch": 4.2548076923076925,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003307834646938477,
      "loss": 4.8709,
      "mean_token_accuracy": 0.22382384538650513,
      "num_tokens": 97358842.0,
      "step": 46905
    },
    {
      "entropy": 5.721827650070191,
      "epoch": 4.255261248185777,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003307521179022201,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2211735412478447,
      "num_tokens": 97369247.0,
      "step": 46910
    },
    {
      "entropy": 5.618614673614502,
      "epoch": 4.255714804063861,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003307207699576246,
      "loss": 4.8405,
      "mean_token_accuracy": 0.22973832935094834,
      "num_tokens": 97379668.0,
      "step": 46915
    },
    {
      "entropy": 5.610636806488037,
      "epoch": 4.256168359941945,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033068942086070924,
      "loss": 4.8271,
      "mean_token_accuracy": 0.22910839766263963,
      "num_tokens": 97389131.0,
      "step": 46920
    },
    {
      "entropy": 5.678656578063965,
      "epoch": 4.256621915820029,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033065807061212254,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21673500537872314,
      "num_tokens": 97400583.0,
      "step": 46925
    },
    {
      "entropy": 5.6806488037109375,
      "epoch": 4.257075471698113,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003306267192125128,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22242211252450944,
      "num_tokens": 97410828.0,
      "step": 46930
    },
    {
      "entropy": 5.672375011444092,
      "epoch": 4.257529027576197,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003305953666625283,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2207650989294052,
      "num_tokens": 97421190.0,
      "step": 46935
    },
    {
      "entropy": 5.632593011856079,
      "epoch": 4.2579825834542815,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033056401296281733,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21688589751720427,
      "num_tokens": 97432257.0,
      "step": 46940
    },
    {
      "entropy": 5.632818651199341,
      "epoch": 4.258436139332366,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003305326581140286,
      "loss": 4.9149,
      "mean_token_accuracy": 0.22278669029474257,
      "num_tokens": 97442272.0,
      "step": 46945
    },
    {
      "entropy": 5.630366802215576,
      "epoch": 4.25888969521045,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033050130211681017,
      "loss": 4.8423,
      "mean_token_accuracy": 0.2283510223031044,
      "num_tokens": 97451462.0,
      "step": 46950
    },
    {
      "entropy": 5.596278810501099,
      "epoch": 4.259343251088534,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033046994497181077,
      "loss": 4.7661,
      "mean_token_accuracy": 0.2302933692932129,
      "num_tokens": 97461579.0,
      "step": 46955
    },
    {
      "entropy": 5.6997950077056885,
      "epoch": 4.2597968069666186,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033043858667967875,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22260741591453553,
      "num_tokens": 97472892.0,
      "step": 46960
    },
    {
      "entropy": 5.577630805969238,
      "epoch": 4.260250362844703,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033040722724106257,
      "loss": 4.8248,
      "mean_token_accuracy": 0.2254362776875496,
      "num_tokens": 97484278.0,
      "step": 46965
    },
    {
      "entropy": 5.607456016540527,
      "epoch": 4.260703918722786,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033037586665661094,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2226048082113266,
      "num_tokens": 97494127.0,
      "step": 46970
    },
    {
      "entropy": 5.574779033660889,
      "epoch": 4.2611574746008705,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003303445049269722,
      "loss": 4.7991,
      "mean_token_accuracy": 0.22869897037744522,
      "num_tokens": 97504213.0,
      "step": 46975
    },
    {
      "entropy": 5.681661367416382,
      "epoch": 4.261611030478955,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033031314205279503,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22021500915288925,
      "num_tokens": 97515242.0,
      "step": 46980
    },
    {
      "entropy": 5.712862491607666,
      "epoch": 4.262064586357039,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033028177803472793,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22420736104249955,
      "num_tokens": 97525956.0,
      "step": 46985
    },
    {
      "entropy": 5.608055257797242,
      "epoch": 4.262518142235123,
      "grad_norm": 1.375,
      "learning_rate": 0.0003302504128734196,
      "loss": 4.7928,
      "mean_token_accuracy": 0.23298343420028686,
      "num_tokens": 97535973.0,
      "step": 46990
    },
    {
      "entropy": 5.560281419754029,
      "epoch": 4.2629716981132075,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033021904656951867,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22866786122322083,
      "num_tokens": 97547580.0,
      "step": 46995
    },
    {
      "entropy": 5.601482152938843,
      "epoch": 4.263425253991292,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033018767912367376,
      "loss": 4.817,
      "mean_token_accuracy": 0.2252979412674904,
      "num_tokens": 97558050.0,
      "step": 47000
    },
    {
      "entropy": 5.6489685535430905,
      "epoch": 4.263878809869376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033015631053653357,
      "loss": 4.8191,
      "mean_token_accuracy": 0.2322539657354355,
      "num_tokens": 97567997.0,
      "step": 47005
    },
    {
      "entropy": 5.5774085521698,
      "epoch": 4.26433236574746,
      "grad_norm": 1.5,
      "learning_rate": 0.0003301249408087469,
      "loss": 4.8228,
      "mean_token_accuracy": 0.23039344251155852,
      "num_tokens": 97577220.0,
      "step": 47010
    },
    {
      "entropy": 5.6266707420349125,
      "epoch": 4.264785921625545,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033009356994096237,
      "loss": 4.8535,
      "mean_token_accuracy": 0.2310619294643402,
      "num_tokens": 97588421.0,
      "step": 47015
    },
    {
      "entropy": 5.571185636520386,
      "epoch": 4.265239477503629,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033006219793382867,
      "loss": 4.7781,
      "mean_token_accuracy": 0.23053307384252547,
      "num_tokens": 97598624.0,
      "step": 47020
    },
    {
      "entropy": 5.632955598831177,
      "epoch": 4.265693033381712,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003300308247879948,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22222653329372405,
      "num_tokens": 97609128.0,
      "step": 47025
    },
    {
      "entropy": 5.614797782897949,
      "epoch": 4.2661465892597965,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003299994505041094,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22944799065589905,
      "num_tokens": 97619549.0,
      "step": 47030
    },
    {
      "entropy": 5.608154439926148,
      "epoch": 4.266600145137881,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032996807508282134,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22114461660385132,
      "num_tokens": 97630184.0,
      "step": 47035
    },
    {
      "entropy": 5.61085877418518,
      "epoch": 4.267053701015965,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032993669852477954,
      "loss": 4.8414,
      "mean_token_accuracy": 0.2295462816953659,
      "num_tokens": 97640569.0,
      "step": 47040
    },
    {
      "entropy": 5.601362705230713,
      "epoch": 4.267507256894049,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003299053208306328,
      "loss": 4.8117,
      "mean_token_accuracy": 0.22550351023674012,
      "num_tokens": 97651006.0,
      "step": 47045
    },
    {
      "entropy": 5.5818705558776855,
      "epoch": 4.267960812772134,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003298739420010299,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22643264681100844,
      "num_tokens": 97661567.0,
      "step": 47050
    },
    {
      "entropy": 5.611021089553833,
      "epoch": 4.268414368650218,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003298425620366201,
      "loss": 4.8544,
      "mean_token_accuracy": 0.23071752786636351,
      "num_tokens": 97673058.0,
      "step": 47055
    },
    {
      "entropy": 5.763935279846192,
      "epoch": 4.268867924528302,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003298111809380521,
      "loss": 4.9774,
      "mean_token_accuracy": 0.21452312022447587,
      "num_tokens": 97684509.0,
      "step": 47060
    },
    {
      "entropy": 5.715897846221924,
      "epoch": 4.269321480406386,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032977979870597477,
      "loss": 4.878,
      "mean_token_accuracy": 0.22203823029994965,
      "num_tokens": 97693954.0,
      "step": 47065
    },
    {
      "entropy": 5.618963384628296,
      "epoch": 4.269775036284471,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003297484153410374,
      "loss": 4.8155,
      "mean_token_accuracy": 0.23069721907377244,
      "num_tokens": 97704345.0,
      "step": 47070
    },
    {
      "entropy": 5.523329162597657,
      "epoch": 4.270228592162554,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032971703084388885,
      "loss": 4.7321,
      "mean_token_accuracy": 0.24109630584716796,
      "num_tokens": 97714857.0,
      "step": 47075
    },
    {
      "entropy": 5.629623126983643,
      "epoch": 4.270682148040638,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000329685645215178,
      "loss": 4.8823,
      "mean_token_accuracy": 0.221777980029583,
      "num_tokens": 97725921.0,
      "step": 47080
    },
    {
      "entropy": 5.6375984191894535,
      "epoch": 4.2711357039187225,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032965425845555413,
      "loss": 4.8406,
      "mean_token_accuracy": 0.23915466219186782,
      "num_tokens": 97735361.0,
      "step": 47085
    },
    {
      "entropy": 5.6213250160217285,
      "epoch": 4.271589259796807,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032962287056566633,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22300980389118194,
      "num_tokens": 97745915.0,
      "step": 47090
    },
    {
      "entropy": 5.651946210861206,
      "epoch": 4.272042815674891,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003295914815461637,
      "loss": 4.8481,
      "mean_token_accuracy": 0.2349666476249695,
      "num_tokens": 97755663.0,
      "step": 47095
    },
    {
      "entropy": 5.667207431793213,
      "epoch": 4.272496371552975,
      "grad_norm": 1.25,
      "learning_rate": 0.00032956009139769514,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21652076840400697,
      "num_tokens": 97766300.0,
      "step": 47100
    },
    {
      "entropy": 5.641920900344848,
      "epoch": 4.27294992743106,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032952870012091013,
      "loss": 4.8668,
      "mean_token_accuracy": 0.22219767421483994,
      "num_tokens": 97776584.0,
      "step": 47105
    },
    {
      "entropy": 5.577944898605347,
      "epoch": 4.273403483309144,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032949730771645767,
      "loss": 4.8095,
      "mean_token_accuracy": 0.2248141884803772,
      "num_tokens": 97786230.0,
      "step": 47110
    },
    {
      "entropy": 5.6517528057098385,
      "epoch": 4.273857039187228,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000329465914184987,
      "loss": 4.9155,
      "mean_token_accuracy": 0.2197134718298912,
      "num_tokens": 97796671.0,
      "step": 47115
    },
    {
      "entropy": 5.561695194244384,
      "epoch": 4.274310595065312,
      "grad_norm": 1.25,
      "learning_rate": 0.0003294345195271473,
      "loss": 4.8376,
      "mean_token_accuracy": 0.22531707137823104,
      "num_tokens": 97807569.0,
      "step": 47120
    },
    {
      "entropy": 5.699811267852783,
      "epoch": 4.274764150943396,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032940312374358794,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21824986934661866,
      "num_tokens": 97818680.0,
      "step": 47125
    },
    {
      "entropy": 5.643488836288452,
      "epoch": 4.27521770682148,
      "grad_norm": 1.453125,
      "learning_rate": 0.000329371726834958,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2279416427016258,
      "num_tokens": 97828955.0,
      "step": 47130
    },
    {
      "entropy": 5.619002914428711,
      "epoch": 4.275671262699564,
      "grad_norm": 1.28125,
      "learning_rate": 0.000329340328801907,
      "loss": 4.8586,
      "mean_token_accuracy": 0.230656997859478,
      "num_tokens": 97839000.0,
      "step": 47135
    },
    {
      "entropy": 5.612377119064331,
      "epoch": 4.276124818577649,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003293089296450841,
      "loss": 4.905,
      "mean_token_accuracy": 0.21601884812116623,
      "num_tokens": 97850140.0,
      "step": 47140
    },
    {
      "entropy": 5.584545993804932,
      "epoch": 4.276578374455733,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003292775293651388,
      "loss": 4.7768,
      "mean_token_accuracy": 0.22782012969255447,
      "num_tokens": 97860887.0,
      "step": 47145
    },
    {
      "entropy": 5.638222980499267,
      "epoch": 4.277031930333817,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003292461279627203,
      "loss": 4.8321,
      "mean_token_accuracy": 0.22459143847227098,
      "num_tokens": 97871190.0,
      "step": 47150
    },
    {
      "entropy": 5.612672281265259,
      "epoch": 4.277485486211901,
      "grad_norm": 1.375,
      "learning_rate": 0.00032921472543847804,
      "loss": 4.7907,
      "mean_token_accuracy": 0.2291889727115631,
      "num_tokens": 97881479.0,
      "step": 47155
    },
    {
      "entropy": 5.583629131317139,
      "epoch": 4.277939042089986,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003291833217930614,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22547540366649627,
      "num_tokens": 97892340.0,
      "step": 47160
    },
    {
      "entropy": 5.6450413227081295,
      "epoch": 4.27839259796807,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032915191702711995,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22380622178316117,
      "num_tokens": 97902882.0,
      "step": 47165
    },
    {
      "entropy": 5.578026628494262,
      "epoch": 4.278846153846154,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032912051114130297,
      "loss": 4.7808,
      "mean_token_accuracy": 0.2286839008331299,
      "num_tokens": 97913187.0,
      "step": 47170
    },
    {
      "entropy": 5.593288612365723,
      "epoch": 4.2792997097242385,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00032908910413626,
      "loss": 4.8155,
      "mean_token_accuracy": 0.23624345958232879,
      "num_tokens": 97922434.0,
      "step": 47175
    },
    {
      "entropy": 5.520088052749633,
      "epoch": 4.279753265602322,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003290576960126407,
      "loss": 4.7372,
      "mean_token_accuracy": 0.23658651113510132,
      "num_tokens": 97933020.0,
      "step": 47180
    },
    {
      "entropy": 5.563190746307373,
      "epoch": 4.280206821480406,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032902628677109434,
      "loss": 4.831,
      "mean_token_accuracy": 0.2287142440676689,
      "num_tokens": 97944103.0,
      "step": 47185
    },
    {
      "entropy": 5.643956089019776,
      "epoch": 4.28066037735849,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003289948764122707,
      "loss": 4.8405,
      "mean_token_accuracy": 0.233802430331707,
      "num_tokens": 97953692.0,
      "step": 47190
    },
    {
      "entropy": 5.667386245727539,
      "epoch": 4.281113933236575,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003289634649368193,
      "loss": 4.8634,
      "mean_token_accuracy": 0.22631232738494872,
      "num_tokens": 97963592.0,
      "step": 47195
    },
    {
      "entropy": 5.701091527938843,
      "epoch": 4.281567489114659,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003289320523453895,
      "loss": 5.0211,
      "mean_token_accuracy": 0.22137184143066407,
      "num_tokens": 97974092.0,
      "step": 47200
    },
    {
      "entropy": 5.547875785827637,
      "epoch": 4.282021044992743,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032890063863863126,
      "loss": 4.7527,
      "mean_token_accuracy": 0.23776146918535232,
      "num_tokens": 97984175.0,
      "step": 47205
    },
    {
      "entropy": 5.574573183059693,
      "epoch": 4.282474600870827,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003288692238171941,
      "loss": 4.8731,
      "mean_token_accuracy": 0.23747948408126832,
      "num_tokens": 97994993.0,
      "step": 47210
    },
    {
      "entropy": 5.609606742858887,
      "epoch": 4.282928156748912,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032883780788172756,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22149042934179305,
      "num_tokens": 98005892.0,
      "step": 47215
    },
    {
      "entropy": 5.590387296676636,
      "epoch": 4.283381712626996,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003288063908328814,
      "loss": 4.7835,
      "mean_token_accuracy": 0.2324290931224823,
      "num_tokens": 98016726.0,
      "step": 47220
    },
    {
      "entropy": 5.6688902378082275,
      "epoch": 4.28383526850508,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003287749726713055,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2265734165906906,
      "num_tokens": 98028500.0,
      "step": 47225
    },
    {
      "entropy": 5.69581937789917,
      "epoch": 4.284288824383164,
      "grad_norm": 1.375,
      "learning_rate": 0.0003287435533976494,
      "loss": 4.9695,
      "mean_token_accuracy": 0.2148450180888176,
      "num_tokens": 98039304.0,
      "step": 47230
    },
    {
      "entropy": 5.616467523574829,
      "epoch": 4.284742380261248,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003287121330125629,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22336160987615586,
      "num_tokens": 98049448.0,
      "step": 47235
    },
    {
      "entropy": 5.620266008377075,
      "epoch": 4.285195936139332,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032868071151669585,
      "loss": 4.861,
      "mean_token_accuracy": 0.23323998153209685,
      "num_tokens": 98059532.0,
      "step": 47240
    },
    {
      "entropy": 5.555338048934937,
      "epoch": 4.285649492017416,
      "grad_norm": 1.59375,
      "learning_rate": 0.000328649288910698,
      "loss": 4.7716,
      "mean_token_accuracy": 0.22919470071792603,
      "num_tokens": 98069385.0,
      "step": 47245
    },
    {
      "entropy": 5.6446723461151125,
      "epoch": 4.286103047895501,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003286178651952192,
      "loss": 4.8401,
      "mean_token_accuracy": 0.23199428468942643,
      "num_tokens": 98079354.0,
      "step": 47250
    },
    {
      "entropy": 5.665226650238037,
      "epoch": 4.286556603773585,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032858644037090924,
      "loss": 4.9654,
      "mean_token_accuracy": 0.2168981611728668,
      "num_tokens": 98090864.0,
      "step": 47255
    },
    {
      "entropy": 5.592876720428467,
      "epoch": 4.287010159651669,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003285550144384181,
      "loss": 4.8211,
      "mean_token_accuracy": 0.22947462797164916,
      "num_tokens": 98101515.0,
      "step": 47260
    },
    {
      "entropy": 5.642464303970337,
      "epoch": 4.2874637155297535,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00032852358739839555,
      "loss": 4.8456,
      "mean_token_accuracy": 0.22512095868587495,
      "num_tokens": 98111786.0,
      "step": 47265
    },
    {
      "entropy": 5.624148607254028,
      "epoch": 4.287917271407838,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032849215925149155,
      "loss": 4.8045,
      "mean_token_accuracy": 0.23568045645952224,
      "num_tokens": 98121732.0,
      "step": 47270
    },
    {
      "entropy": 5.66390061378479,
      "epoch": 4.288370827285922,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003284607299983561,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21933106780052186,
      "num_tokens": 98131417.0,
      "step": 47275
    },
    {
      "entropy": 5.6697639465332035,
      "epoch": 4.288824383164005,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032842929963963915,
      "loss": 4.8369,
      "mean_token_accuracy": 0.22621381729841233,
      "num_tokens": 98141664.0,
      "step": 47280
    },
    {
      "entropy": 5.639224195480347,
      "epoch": 4.28927793904209,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032839786817599065,
      "loss": 4.8144,
      "mean_token_accuracy": 0.22321742326021193,
      "num_tokens": 98150959.0,
      "step": 47285
    },
    {
      "entropy": 5.633822250366211,
      "epoch": 4.289731494920174,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003283664356080606,
      "loss": 4.9463,
      "mean_token_accuracy": 0.2225288286805153,
      "num_tokens": 98161957.0,
      "step": 47290
    },
    {
      "entropy": 5.563166761398316,
      "epoch": 4.290185050798258,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032833500193649916,
      "loss": 4.7925,
      "mean_token_accuracy": 0.2278713896870613,
      "num_tokens": 98172053.0,
      "step": 47295
    },
    {
      "entropy": 5.6144476413726805,
      "epoch": 4.2906386066763424,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032830356716195624,
      "loss": 4.8124,
      "mean_token_accuracy": 0.2267252117395401,
      "num_tokens": 98182074.0,
      "step": 47300
    },
    {
      "entropy": 5.666690111160278,
      "epoch": 4.291092162554427,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003282721312850819,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22365194857120513,
      "num_tokens": 98192591.0,
      "step": 47305
    },
    {
      "entropy": 5.579317474365235,
      "epoch": 4.291545718432511,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003282406943065264,
      "loss": 4.7609,
      "mean_token_accuracy": 0.23524457961320877,
      "num_tokens": 98203426.0,
      "step": 47310
    },
    {
      "entropy": 5.577765941619873,
      "epoch": 4.291999274310595,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032820925622693977,
      "loss": 4.8241,
      "mean_token_accuracy": 0.2242076277732849,
      "num_tokens": 98214342.0,
      "step": 47315
    },
    {
      "entropy": 5.621197652816773,
      "epoch": 4.2924528301886795,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032817781704697204,
      "loss": 4.8302,
      "mean_token_accuracy": 0.22377626299858094,
      "num_tokens": 98224155.0,
      "step": 47320
    },
    {
      "entropy": 5.613556289672852,
      "epoch": 4.292906386066764,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003281463767672736,
      "loss": 4.8224,
      "mean_token_accuracy": 0.23124950379133224,
      "num_tokens": 98233613.0,
      "step": 47325
    },
    {
      "entropy": 5.63717794418335,
      "epoch": 4.293359941944848,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003281149353884946,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23010460138320923,
      "num_tokens": 98243117.0,
      "step": 47330
    },
    {
      "entropy": 5.607906103134155,
      "epoch": 4.293813497822931,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003280834929112851,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23018937557935715,
      "num_tokens": 98253619.0,
      "step": 47335
    },
    {
      "entropy": 5.614600086212159,
      "epoch": 4.294267053701016,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003280520493362955,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22821273803710937,
      "num_tokens": 98264221.0,
      "step": 47340
    },
    {
      "entropy": 5.583437299728393,
      "epoch": 4.2947206095791,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003280206046641759,
      "loss": 4.7972,
      "mean_token_accuracy": 0.22635616660118102,
      "num_tokens": 98275530.0,
      "step": 47345
    },
    {
      "entropy": 5.701711416244507,
      "epoch": 4.295174165457184,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032798915889557687,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22117282599210739,
      "num_tokens": 98285937.0,
      "step": 47350
    },
    {
      "entropy": 5.661172199249267,
      "epoch": 4.2956277213352685,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003279577120311483,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22456271946430206,
      "num_tokens": 98295690.0,
      "step": 47355
    },
    {
      "entropy": 5.596270227432251,
      "epoch": 4.296081277213353,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000327926264071541,
      "loss": 4.8471,
      "mean_token_accuracy": 0.22586225718259811,
      "num_tokens": 98305073.0,
      "step": 47360
    },
    {
      "entropy": 5.622945833206177,
      "epoch": 4.296534833091437,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003278948150174049,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22098020613193511,
      "num_tokens": 98316008.0,
      "step": 47365
    },
    {
      "entropy": 5.596741580963135,
      "epoch": 4.296988388969521,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003278633648693906,
      "loss": 4.8177,
      "mean_token_accuracy": 0.22668539583683014,
      "num_tokens": 98327032.0,
      "step": 47370
    },
    {
      "entropy": 5.658023023605347,
      "epoch": 4.2974419448476056,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003278319136281484,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22983046025037765,
      "num_tokens": 98336085.0,
      "step": 47375
    },
    {
      "entropy": 5.631104278564453,
      "epoch": 4.297895500725689,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032780046129432885,
      "loss": 4.8717,
      "mean_token_accuracy": 0.23373648524284363,
      "num_tokens": 98346180.0,
      "step": 47380
    },
    {
      "entropy": 5.616884231567383,
      "epoch": 4.298349056603773,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00032776900786858214,
      "loss": 4.7993,
      "mean_token_accuracy": 0.22891128063201904,
      "num_tokens": 98356001.0,
      "step": 47385
    },
    {
      "entropy": 5.690781450271606,
      "epoch": 4.2988026124818575,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000327737553351559,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22270919382572174,
      "num_tokens": 98366744.0,
      "step": 47390
    },
    {
      "entropy": 5.577809047698975,
      "epoch": 4.299256168359942,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032770609774390985,
      "loss": 4.7953,
      "mean_token_accuracy": 0.22995573729276658,
      "num_tokens": 98377350.0,
      "step": 47395
    },
    {
      "entropy": 5.563135385513306,
      "epoch": 4.299709724238026,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003276746410462852,
      "loss": 4.8349,
      "mean_token_accuracy": 0.2301477536559105,
      "num_tokens": 98387926.0,
      "step": 47400
    },
    {
      "entropy": 5.642178678512574,
      "epoch": 4.30016328011611,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032764318325933546,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2242929980158806,
      "num_tokens": 98397117.0,
      "step": 47405
    },
    {
      "entropy": 5.663260269165039,
      "epoch": 4.3006168359941945,
      "grad_norm": 1.25,
      "learning_rate": 0.0003276117243837113,
      "loss": 4.887,
      "mean_token_accuracy": 0.22340668588876725,
      "num_tokens": 98407983.0,
      "step": 47410
    },
    {
      "entropy": 5.636649036407471,
      "epoch": 4.301070391872279,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003275802644200633,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22215288430452346,
      "num_tokens": 98419940.0,
      "step": 47415
    },
    {
      "entropy": 5.577565765380859,
      "epoch": 4.301523947750363,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000327548803369042,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23608071208000184,
      "num_tokens": 98430487.0,
      "step": 47420
    },
    {
      "entropy": 5.61991605758667,
      "epoch": 4.301977503628447,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032751734123129814,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2345798820257187,
      "num_tokens": 98440231.0,
      "step": 47425
    },
    {
      "entropy": 5.654092407226562,
      "epoch": 4.302431059506532,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003274858780074822,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22632853090763091,
      "num_tokens": 98449741.0,
      "step": 47430
    },
    {
      "entropy": 5.631859397888183,
      "epoch": 4.302884615384615,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032745441369824485,
      "loss": 4.7825,
      "mean_token_accuracy": 0.24184008538722992,
      "num_tokens": 98461068.0,
      "step": 47435
    },
    {
      "entropy": 5.628634214401245,
      "epoch": 4.303338171262699,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032742294830423696,
      "loss": 4.8759,
      "mean_token_accuracy": 0.23079689741134643,
      "num_tokens": 98472585.0,
      "step": 47440
    },
    {
      "entropy": 5.635098791122436,
      "epoch": 4.3037917271407835,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003273914818261092,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22806455940008163,
      "num_tokens": 98482874.0,
      "step": 47445
    },
    {
      "entropy": 5.622268915176392,
      "epoch": 4.304245283018868,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032736001426451216,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2275979846715927,
      "num_tokens": 98492995.0,
      "step": 47450
    },
    {
      "entropy": 5.689875555038452,
      "epoch": 4.304698838896952,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003273285456200966,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2230018436908722,
      "num_tokens": 98503965.0,
      "step": 47455
    },
    {
      "entropy": 5.615868616104126,
      "epoch": 4.305152394775036,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003272970758935135,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22644599825143813,
      "num_tokens": 98515584.0,
      "step": 47460
    },
    {
      "entropy": 5.662915802001953,
      "epoch": 4.305605950653121,
      "grad_norm": 1.375,
      "learning_rate": 0.00032726560508541356,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22782246321439742,
      "num_tokens": 98527134.0,
      "step": 47465
    },
    {
      "entropy": 5.660779619216919,
      "epoch": 4.306059506531205,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032723413319644756,
      "loss": 4.8944,
      "mean_token_accuracy": 0.2248922899365425,
      "num_tokens": 98537189.0,
      "step": 47470
    },
    {
      "entropy": 5.623597192764282,
      "epoch": 4.306513062409289,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003272026602272663,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22595262974500657,
      "num_tokens": 98547803.0,
      "step": 47475
    },
    {
      "entropy": 5.588962078094482,
      "epoch": 4.306966618287373,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032717118617852086,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22469249963760377,
      "num_tokens": 98558332.0,
      "step": 47480
    },
    {
      "entropy": 5.693544149398804,
      "epoch": 4.307420174165458,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003271397110508619,
      "loss": 4.9721,
      "mean_token_accuracy": 0.22422416061162947,
      "num_tokens": 98569046.0,
      "step": 47485
    },
    {
      "entropy": 5.611083698272705,
      "epoch": 4.307873730043541,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003271082348449405,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22270580530166625,
      "num_tokens": 98579358.0,
      "step": 47490
    },
    {
      "entropy": 5.606311464309693,
      "epoch": 4.308327285921625,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003270767575614074,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22420376241207124,
      "num_tokens": 98589224.0,
      "step": 47495
    },
    {
      "entropy": 5.683764123916626,
      "epoch": 4.3087808417997095,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003270452792009138,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2254382088780403,
      "num_tokens": 98598954.0,
      "step": 47500
    },
    {
      "entropy": 5.607494497299195,
      "epoch": 4.309234397677794,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032701379976411043,
      "loss": 4.7893,
      "mean_token_accuracy": 0.22426560819149016,
      "num_tokens": 98608813.0,
      "step": 47505
    },
    {
      "entropy": 5.590865993499756,
      "epoch": 4.309687953555878,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003269823192516486,
      "loss": 4.8367,
      "mean_token_accuracy": 0.2300844520330429,
      "num_tokens": 98618746.0,
      "step": 47510
    },
    {
      "entropy": 5.645672464370728,
      "epoch": 4.310141509433962,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032695083766417913,
      "loss": 4.851,
      "mean_token_accuracy": 0.23096059411764144,
      "num_tokens": 98628804.0,
      "step": 47515
    },
    {
      "entropy": 5.6246910572052,
      "epoch": 4.310595065312047,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000326919355002353,
      "loss": 4.8475,
      "mean_token_accuracy": 0.22641227543354034,
      "num_tokens": 98640056.0,
      "step": 47520
    },
    {
      "entropy": 5.5767802715301515,
      "epoch": 4.311048621190131,
      "grad_norm": 1.375,
      "learning_rate": 0.00032688787126682143,
      "loss": 4.8077,
      "mean_token_accuracy": 0.2302025318145752,
      "num_tokens": 98651042.0,
      "step": 47525
    },
    {
      "entropy": 5.624698495864868,
      "epoch": 4.311502177068215,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003268563864582354,
      "loss": 4.8416,
      "mean_token_accuracy": 0.22820381671190262,
      "num_tokens": 98661405.0,
      "step": 47530
    },
    {
      "entropy": 5.684638738632202,
      "epoch": 4.3119557329462985,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032682490057724616,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21758223623037337,
      "num_tokens": 98672323.0,
      "step": 47535
    },
    {
      "entropy": 5.685717296600342,
      "epoch": 4.312409288824383,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003267934136245047,
      "loss": 4.8873,
      "mean_token_accuracy": 0.22413784265518188,
      "num_tokens": 98681431.0,
      "step": 47540
    },
    {
      "entropy": 5.542545032501221,
      "epoch": 4.312862844702467,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003267619256006623,
      "loss": 4.7825,
      "mean_token_accuracy": 0.24008105844259262,
      "num_tokens": 98692345.0,
      "step": 47545
    },
    {
      "entropy": 5.616249036788941,
      "epoch": 4.313316400580551,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003267304365063701,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22397732138633727,
      "num_tokens": 98703425.0,
      "step": 47550
    },
    {
      "entropy": 5.645273351669312,
      "epoch": 4.313769956458636,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003266989463422793,
      "loss": 4.8688,
      "mean_token_accuracy": 0.2304901361465454,
      "num_tokens": 98714193.0,
      "step": 47555
    },
    {
      "entropy": 5.6491199970245365,
      "epoch": 4.31422351233672,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00032666745510904105,
      "loss": 4.839,
      "mean_token_accuracy": 0.23138196915388107,
      "num_tokens": 98723550.0,
      "step": 47560
    },
    {
      "entropy": 5.604430389404297,
      "epoch": 4.314677068214804,
      "grad_norm": 1.375,
      "learning_rate": 0.00032663596280730673,
      "loss": 4.8606,
      "mean_token_accuracy": 0.2220545455813408,
      "num_tokens": 98734075.0,
      "step": 47565
    },
    {
      "entropy": 5.6208476543426515,
      "epoch": 4.315130624092888,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032660446943772755,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22627621740102768,
      "num_tokens": 98745447.0,
      "step": 47570
    },
    {
      "entropy": 5.654650783538818,
      "epoch": 4.315584179970973,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003265729750009547,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22317731082439424,
      "num_tokens": 98756169.0,
      "step": 47575
    },
    {
      "entropy": 5.642724514007568,
      "epoch": 4.316037735849057,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003265414794976397,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21905523538589478,
      "num_tokens": 98767112.0,
      "step": 47580
    },
    {
      "entropy": 5.667980194091797,
      "epoch": 4.316491291727141,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003265099829284337,
      "loss": 4.917,
      "mean_token_accuracy": 0.22542796283960342,
      "num_tokens": 98777533.0,
      "step": 47585
    },
    {
      "entropy": 5.6909300804138185,
      "epoch": 4.316944847605225,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003264784852939881,
      "loss": 4.926,
      "mean_token_accuracy": 0.21785698384046553,
      "num_tokens": 98788101.0,
      "step": 47590
    },
    {
      "entropy": 5.72018575668335,
      "epoch": 4.317398403483309,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032644698659495437,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22534625381231307,
      "num_tokens": 98799438.0,
      "step": 47595
    },
    {
      "entropy": 5.60253529548645,
      "epoch": 4.317851959361393,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032641548683198384,
      "loss": 4.8303,
      "mean_token_accuracy": 0.223910416662693,
      "num_tokens": 98810325.0,
      "step": 47600
    },
    {
      "entropy": 5.565122652053833,
      "epoch": 4.318305515239477,
      "grad_norm": 1.25,
      "learning_rate": 0.0003263839860057279,
      "loss": 4.8446,
      "mean_token_accuracy": 0.2276192158460617,
      "num_tokens": 98821897.0,
      "step": 47605
    },
    {
      "entropy": 5.631326532363891,
      "epoch": 4.318759071117562,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000326352484116838,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22253056466579438,
      "num_tokens": 98833338.0,
      "step": 47610
    },
    {
      "entropy": 5.671670246124267,
      "epoch": 4.319212626995646,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032632098116596576,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22501097321510316,
      "num_tokens": 98844308.0,
      "step": 47615
    },
    {
      "entropy": 5.645916938781738,
      "epoch": 4.31966618287373,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032628947715376247,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22481424510478973,
      "num_tokens": 98854491.0,
      "step": 47620
    },
    {
      "entropy": 5.641106796264649,
      "epoch": 4.320119738751814,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003262579720808797,
      "loss": 4.9093,
      "mean_token_accuracy": 0.21963253766298294,
      "num_tokens": 98864702.0,
      "step": 47625
    },
    {
      "entropy": 5.614544343948364,
      "epoch": 4.320573294629899,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032622646594796905,
      "loss": 4.8557,
      "mean_token_accuracy": 0.23078173100948335,
      "num_tokens": 98874682.0,
      "step": 47630
    },
    {
      "entropy": 5.679851627349853,
      "epoch": 4.321026850507983,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003261949587556821,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21658227145671843,
      "num_tokens": 98885130.0,
      "step": 47635
    },
    {
      "entropy": 5.649454355239868,
      "epoch": 4.321480406386067,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032616345050467016,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22533013224601744,
      "num_tokens": 98896398.0,
      "step": 47640
    },
    {
      "entropy": 5.610025882720947,
      "epoch": 4.321933962264151,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003261319411955852,
      "loss": 4.8259,
      "mean_token_accuracy": 0.231671604514122,
      "num_tokens": 98906571.0,
      "step": 47645
    },
    {
      "entropy": 5.600841045379639,
      "epoch": 4.322387518142235,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003261004308290786,
      "loss": 4.8145,
      "mean_token_accuracy": 0.2319784477353096,
      "num_tokens": 98916666.0,
      "step": 47650
    },
    {
      "entropy": 5.6467290878295895,
      "epoch": 4.322841074020319,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003260689194058021,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22290383726358415,
      "num_tokens": 98926599.0,
      "step": 47655
    },
    {
      "entropy": 5.583265256881714,
      "epoch": 4.323294629898403,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032603740692640733,
      "loss": 4.7869,
      "mean_token_accuracy": 0.2283250778913498,
      "num_tokens": 98936512.0,
      "step": 47660
    },
    {
      "entropy": 5.662972354888916,
      "epoch": 4.323748185776488,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003260058933915459,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20911809653043748,
      "num_tokens": 98946604.0,
      "step": 47665
    },
    {
      "entropy": 5.690374565124512,
      "epoch": 4.324201741654572,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003259743788018697,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22004904448986054,
      "num_tokens": 98956257.0,
      "step": 47670
    },
    {
      "entropy": 5.6734363555908205,
      "epoch": 4.324655297532656,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032594286315803026,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21805552989244462,
      "num_tokens": 98966346.0,
      "step": 47675
    },
    {
      "entropy": 5.603524446487427,
      "epoch": 4.3251088534107405,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003259113464606795,
      "loss": 4.8226,
      "mean_token_accuracy": 0.22663767635822296,
      "num_tokens": 98977410.0,
      "step": 47680
    },
    {
      "entropy": 5.682003498077393,
      "epoch": 4.325562409288825,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003258798287104691,
      "loss": 4.914,
      "mean_token_accuracy": 0.22437324076890947,
      "num_tokens": 98987488.0,
      "step": 47685
    },
    {
      "entropy": 5.648275899887085,
      "epoch": 4.326015965166908,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032584830990805094,
      "loss": 4.8011,
      "mean_token_accuracy": 0.2328224390745163,
      "num_tokens": 98996792.0,
      "step": 47690
    },
    {
      "entropy": 5.62788577079773,
      "epoch": 4.326469521044992,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032581679005407667,
      "loss": 4.7842,
      "mean_token_accuracy": 0.23430884182453154,
      "num_tokens": 99007425.0,
      "step": 47695
    },
    {
      "entropy": 5.588452243804932,
      "epoch": 4.326923076923077,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003257852691491983,
      "loss": 4.8227,
      "mean_token_accuracy": 0.22192427515983582,
      "num_tokens": 99018275.0,
      "step": 47700
    },
    {
      "entropy": 5.630432653427124,
      "epoch": 4.327376632801161,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032575374719406755,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22239874303340912,
      "num_tokens": 99028853.0,
      "step": 47705
    },
    {
      "entropy": 5.645051670074463,
      "epoch": 4.327830188679245,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032572222418933644,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2207654371857643,
      "num_tokens": 99039423.0,
      "step": 47710
    },
    {
      "entropy": 5.6546203136444095,
      "epoch": 4.3282837445573294,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003256907001356568,
      "loss": 4.9077,
      "mean_token_accuracy": 0.21334689408540725,
      "num_tokens": 99050041.0,
      "step": 47715
    },
    {
      "entropy": 5.638952016830444,
      "epoch": 4.328737300435414,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003256591750336806,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21801101118326188,
      "num_tokens": 99062050.0,
      "step": 47720
    },
    {
      "entropy": 5.5602985382080075,
      "epoch": 4.329190856313498,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003256276488840596,
      "loss": 4.7685,
      "mean_token_accuracy": 0.22365766912698745,
      "num_tokens": 99072337.0,
      "step": 47725
    },
    {
      "entropy": 5.53347692489624,
      "epoch": 4.329644412191582,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032559612168744613,
      "loss": 4.7715,
      "mean_token_accuracy": 0.22890606969594957,
      "num_tokens": 99082468.0,
      "step": 47730
    },
    {
      "entropy": 5.61519455909729,
      "epoch": 4.3300979680696665,
      "grad_norm": 1.5,
      "learning_rate": 0.0003255645934444918,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22253265976905823,
      "num_tokens": 99091757.0,
      "step": 47735
    },
    {
      "entropy": 5.673792457580566,
      "epoch": 4.330551523947751,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003255330641558489,
      "loss": 4.9282,
      "mean_token_accuracy": 0.2171151965856552,
      "num_tokens": 99103192.0,
      "step": 47740
    },
    {
      "entropy": 5.58910026550293,
      "epoch": 4.331005079825834,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003255015338221693,
      "loss": 4.7745,
      "mean_token_accuracy": 0.23132523596286775,
      "num_tokens": 99112660.0,
      "step": 47745
    },
    {
      "entropy": 5.679754590988159,
      "epoch": 4.331458635703918,
      "grad_norm": 1.25,
      "learning_rate": 0.00032547000244410515,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22515864670276642,
      "num_tokens": 99123867.0,
      "step": 47750
    },
    {
      "entropy": 5.605358648300171,
      "epoch": 4.331912191582003,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032543847002230845,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22274049520492553,
      "num_tokens": 99134445.0,
      "step": 47755
    },
    {
      "entropy": 5.586302995681763,
      "epoch": 4.332365747460087,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003254069365574313,
      "loss": 4.8214,
      "mean_token_accuracy": 0.22941377460956575,
      "num_tokens": 99143483.0,
      "step": 47760
    },
    {
      "entropy": 5.5938245296478275,
      "epoch": 4.332819303338171,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003253754020501259,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23127358108758928,
      "num_tokens": 99154514.0,
      "step": 47765
    },
    {
      "entropy": 5.6041340827941895,
      "epoch": 4.3332728592162555,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032534386650104444,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21638872474431992,
      "num_tokens": 99165969.0,
      "step": 47770
    },
    {
      "entropy": 5.569817352294922,
      "epoch": 4.33372641509434,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003253123299108388,
      "loss": 4.7691,
      "mean_token_accuracy": 0.23499057590961456,
      "num_tokens": 99176595.0,
      "step": 47775
    },
    {
      "entropy": 5.667035531997681,
      "epoch": 4.334179970972424,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003252807922801615,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22287643253803252,
      "num_tokens": 99186916.0,
      "step": 47780
    },
    {
      "entropy": 5.676402997970581,
      "epoch": 4.334633526850508,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003252492536096646,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21792744249105453,
      "num_tokens": 99197892.0,
      "step": 47785
    },
    {
      "entropy": 5.662909650802613,
      "epoch": 4.3350870827285926,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003252177139000003,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22544106543064119,
      "num_tokens": 99208373.0,
      "step": 47790
    },
    {
      "entropy": 5.678479528427124,
      "epoch": 4.335540638606676,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003251861731518209,
      "loss": 4.8863,
      "mean_token_accuracy": 0.22625533640384674,
      "num_tokens": 99218788.0,
      "step": 47795
    },
    {
      "entropy": 5.726149368286133,
      "epoch": 4.33599419448476,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00032515463136577864,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22453456073999406,
      "num_tokens": 99228653.0,
      "step": 47800
    },
    {
      "entropy": 5.603572177886963,
      "epoch": 4.3364477503628445,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032512308854252577,
      "loss": 4.7935,
      "mean_token_accuracy": 0.23357176780700684,
      "num_tokens": 99239925.0,
      "step": 47805
    },
    {
      "entropy": 5.5493011474609375,
      "epoch": 4.336901306240929,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003250915446827147,
      "loss": 4.75,
      "mean_token_accuracy": 0.2331490099430084,
      "num_tokens": 99250091.0,
      "step": 47810
    },
    {
      "entropy": 5.590400695800781,
      "epoch": 4.337354862119013,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003250599997869977,
      "loss": 4.8831,
      "mean_token_accuracy": 0.21178607344627381,
      "num_tokens": 99260799.0,
      "step": 47815
    },
    {
      "entropy": 5.582299137115479,
      "epoch": 4.337808417997097,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003250284538560272,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23411945402622222,
      "num_tokens": 99270338.0,
      "step": 47820
    },
    {
      "entropy": 5.596254873275757,
      "epoch": 4.3382619738751815,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003249969068904554,
      "loss": 4.84,
      "mean_token_accuracy": 0.2245519995689392,
      "num_tokens": 99280419.0,
      "step": 47825
    },
    {
      "entropy": 5.681427955627441,
      "epoch": 4.338715529753266,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032496535889093496,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22201812416315078,
      "num_tokens": 99291590.0,
      "step": 47830
    },
    {
      "entropy": 5.658459854125977,
      "epoch": 4.33916908563135,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032493380985811814,
      "loss": 4.923,
      "mean_token_accuracy": 0.2200256645679474,
      "num_tokens": 99301987.0,
      "step": 47835
    },
    {
      "entropy": 5.669340324401856,
      "epoch": 4.339622641509434,
      "grad_norm": 1.484375,
      "learning_rate": 0.00032490225979265734,
      "loss": 4.8797,
      "mean_token_accuracy": 0.2253356993198395,
      "num_tokens": 99312231.0,
      "step": 47840
    },
    {
      "entropy": 5.664461231231689,
      "epoch": 4.340076197387518,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032487070869520517,
      "loss": 4.8157,
      "mean_token_accuracy": 0.22471612840890884,
      "num_tokens": 99322784.0,
      "step": 47845
    },
    {
      "entropy": 5.6572594165802,
      "epoch": 4.340529753265602,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000324839156566414,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22448582351207733,
      "num_tokens": 99333989.0,
      "step": 47850
    },
    {
      "entropy": 5.561170625686645,
      "epoch": 4.340983309143686,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003248076034069364,
      "loss": 4.8384,
      "mean_token_accuracy": 0.2304004117846489,
      "num_tokens": 99345034.0,
      "step": 47855
    },
    {
      "entropy": 5.585924291610718,
      "epoch": 4.3414368650217705,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003247760492174248,
      "loss": 4.7924,
      "mean_token_accuracy": 0.22956569194793702,
      "num_tokens": 99356062.0,
      "step": 47860
    },
    {
      "entropy": 5.632416915893555,
      "epoch": 4.341890420899855,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003247444939985319,
      "loss": 4.823,
      "mean_token_accuracy": 0.2353198766708374,
      "num_tokens": 99365315.0,
      "step": 47865
    },
    {
      "entropy": 5.703348827362061,
      "epoch": 4.342343976777939,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032471293775091013,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22791737020015718,
      "num_tokens": 99374897.0,
      "step": 47870
    },
    {
      "entropy": 5.608999347686767,
      "epoch": 4.342797532656023,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032468138047521216,
      "loss": 4.8179,
      "mean_token_accuracy": 0.23082666099071503,
      "num_tokens": 99383624.0,
      "step": 47875
    },
    {
      "entropy": 5.569965076446533,
      "epoch": 4.343251088534108,
      "grad_norm": 1.25,
      "learning_rate": 0.0003246498221720906,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22204968631267546,
      "num_tokens": 99393791.0,
      "step": 47880
    },
    {
      "entropy": 5.612374258041382,
      "epoch": 4.343704644412192,
      "grad_norm": 1.328125,
      "learning_rate": 0.000324618262842198,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22458772957324982,
      "num_tokens": 99403687.0,
      "step": 47885
    },
    {
      "entropy": 5.640883684158325,
      "epoch": 4.344158200290276,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032458670248618714,
      "loss": 4.8325,
      "mean_token_accuracy": 0.22591176480054856,
      "num_tokens": 99413411.0,
      "step": 47890
    },
    {
      "entropy": 5.655157232284546,
      "epoch": 4.34461175616836,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003245551411047106,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22235150933265685,
      "num_tokens": 99425563.0,
      "step": 47895
    },
    {
      "entropy": 5.627981042861938,
      "epoch": 4.345065312046444,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032452357869842117,
      "loss": 4.8445,
      "mean_token_accuracy": 0.23448598086833955,
      "num_tokens": 99436297.0,
      "step": 47900
    },
    {
      "entropy": 5.6689050674438475,
      "epoch": 4.345518867924528,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003244920152679715,
      "loss": 4.8553,
      "mean_token_accuracy": 0.22996119260787964,
      "num_tokens": 99445745.0,
      "step": 47905
    },
    {
      "entropy": 5.600391578674317,
      "epoch": 4.345972423802612,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032446045081401426,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22004247456789017,
      "num_tokens": 99456915.0,
      "step": 47910
    },
    {
      "entropy": 5.675147581100464,
      "epoch": 4.3464259796806965,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003244288853372024,
      "loss": 4.8823,
      "mean_token_accuracy": 0.2235233649611473,
      "num_tokens": 99466778.0,
      "step": 47915
    },
    {
      "entropy": 5.653891849517822,
      "epoch": 4.346879535558781,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003243973188381885,
      "loss": 4.8367,
      "mean_token_accuracy": 0.22305303663015366,
      "num_tokens": 99475984.0,
      "step": 47920
    },
    {
      "entropy": 5.539736604690551,
      "epoch": 4.347333091436865,
      "grad_norm": 1.25,
      "learning_rate": 0.00032436575131762553,
      "loss": 4.8357,
      "mean_token_accuracy": 0.220150326192379,
      "num_tokens": 99486238.0,
      "step": 47925
    },
    {
      "entropy": 5.7392138004302975,
      "epoch": 4.347786647314949,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00032433418277616623,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2254258319735527,
      "num_tokens": 99497806.0,
      "step": 47930
    },
    {
      "entropy": 5.635987758636475,
      "epoch": 4.348240203193034,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003243026132144635,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22160115987062454,
      "num_tokens": 99507214.0,
      "step": 47935
    },
    {
      "entropy": 5.640987300872803,
      "epoch": 4.348693759071118,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003242710426331701,
      "loss": 4.9046,
      "mean_token_accuracy": 0.2243996098637581,
      "num_tokens": 99517025.0,
      "step": 47940
    },
    {
      "entropy": 5.667987871170044,
      "epoch": 4.349147314949202,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032423947103293893,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22466300427913666,
      "num_tokens": 99526972.0,
      "step": 47945
    },
    {
      "entropy": 5.664741563796997,
      "epoch": 4.3496008708272855,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032420789841442306,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21778185069561004,
      "num_tokens": 99538583.0,
      "step": 47950
    },
    {
      "entropy": 5.582291889190674,
      "epoch": 4.35005442670537,
      "grad_norm": 1.375,
      "learning_rate": 0.00032417632477827526,
      "loss": 4.7995,
      "mean_token_accuracy": 0.23142951130867004,
      "num_tokens": 99549178.0,
      "step": 47955
    },
    {
      "entropy": 5.615753269195556,
      "epoch": 4.350507982583454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032414475012514847,
      "loss": 4.8412,
      "mean_token_accuracy": 0.2285893440246582,
      "num_tokens": 99558491.0,
      "step": 47960
    },
    {
      "entropy": 5.6170494556427,
      "epoch": 4.350961538461538,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003241131744556958,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22427558600902558,
      "num_tokens": 99568711.0,
      "step": 47965
    },
    {
      "entropy": 5.59133734703064,
      "epoch": 4.351415094339623,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003240815977705702,
      "loss": 4.868,
      "mean_token_accuracy": 0.22320684492588044,
      "num_tokens": 99578329.0,
      "step": 47970
    },
    {
      "entropy": 5.5686053276062015,
      "epoch": 4.351868650217707,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032405002007042453,
      "loss": 4.7353,
      "mean_token_accuracy": 0.2318221852183342,
      "num_tokens": 99588646.0,
      "step": 47975
    },
    {
      "entropy": 5.669190549850464,
      "epoch": 4.352322206095791,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032401844135591213,
      "loss": 4.9034,
      "mean_token_accuracy": 0.21720054745674133,
      "num_tokens": 99598780.0,
      "step": 47980
    },
    {
      "entropy": 5.647005319595337,
      "epoch": 4.352775761973875,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003239868616276857,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22623246908187866,
      "num_tokens": 99609528.0,
      "step": 47985
    },
    {
      "entropy": 5.583690595626831,
      "epoch": 4.35322931785196,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003239552808863986,
      "loss": 4.7842,
      "mean_token_accuracy": 0.23177415877580643,
      "num_tokens": 99619928.0,
      "step": 47990
    },
    {
      "entropy": 5.569075393676758,
      "epoch": 4.353682873730044,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032392369913270376,
      "loss": 4.7651,
      "mean_token_accuracy": 0.22978690415620803,
      "num_tokens": 99630135.0,
      "step": 47995
    },
    {
      "entropy": 5.651469612121582,
      "epoch": 4.354136429608127,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003238921163672544,
      "loss": 4.9806,
      "mean_token_accuracy": 0.2099696949124336,
      "num_tokens": 99641557.0,
      "step": 48000
    },
    {
      "epoch": 4.354136429608127,
      "eval_entropy": 5.325137837224417,
      "eval_loss": 4.948969841003418,
      "eval_mean_token_accuracy": 0.22987959253605944,
      "eval_num_tokens": 99641557.0,
      "eval_runtime": 20.6212,
      "eval_samples_per_second": 1714.743,
      "eval_steps_per_second": 214.343,
      "step": 48000
    },
    {
      "entropy": 5.640565633773804,
      "epoch": 4.354589985486212,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003238605325907035,
      "loss": 4.8686,
      "mean_token_accuracy": 0.22544457018375397,
      "num_tokens": 99651244.0,
      "step": 48005
    },
    {
      "entropy": 5.641449737548828,
      "epoch": 4.355043541364296,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032382894780370446,
      "loss": 4.8361,
      "mean_token_accuracy": 0.221346116065979,
      "num_tokens": 99662317.0,
      "step": 48010
    },
    {
      "entropy": 5.622298240661621,
      "epoch": 4.35549709724238,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032379736200691036,
      "loss": 4.8179,
      "mean_token_accuracy": 0.22855726927518843,
      "num_tokens": 99671624.0,
      "step": 48015
    },
    {
      "entropy": 5.6669697761535645,
      "epoch": 4.355950653120464,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032376577520097427,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21547969430685043,
      "num_tokens": 99682076.0,
      "step": 48020
    },
    {
      "entropy": 5.590432596206665,
      "epoch": 4.356404208998549,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032373418738654953,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22628352642059327,
      "num_tokens": 99692076.0,
      "step": 48025
    },
    {
      "entropy": 5.6119671821594235,
      "epoch": 4.356857764876633,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003237025985642895,
      "loss": 4.8407,
      "mean_token_accuracy": 0.22803518921136856,
      "num_tokens": 99701455.0,
      "step": 48030
    },
    {
      "entropy": 5.666763830184936,
      "epoch": 4.357311320754717,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032367100873484725,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22491520047187805,
      "num_tokens": 99712034.0,
      "step": 48035
    },
    {
      "entropy": 5.665150785446167,
      "epoch": 4.357764876632801,
      "grad_norm": 1.25,
      "learning_rate": 0.000323639417898876,
      "loss": 4.8873,
      "mean_token_accuracy": 0.22376833111047745,
      "num_tokens": 99722965.0,
      "step": 48040
    },
    {
      "entropy": 5.55571665763855,
      "epoch": 4.358218432510886,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003236078260570293,
      "loss": 4.7836,
      "mean_token_accuracy": 0.22473234236240386,
      "num_tokens": 99732847.0,
      "step": 48045
    },
    {
      "entropy": 5.576132440567017,
      "epoch": 4.35867198838897,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003235762332099604,
      "loss": 4.7986,
      "mean_token_accuracy": 0.23672358095645904,
      "num_tokens": 99743639.0,
      "step": 48050
    },
    {
      "entropy": 5.617038106918335,
      "epoch": 4.359125544267053,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003235446393583227,
      "loss": 4.775,
      "mean_token_accuracy": 0.23903438150882722,
      "num_tokens": 99753374.0,
      "step": 48055
    },
    {
      "entropy": 5.62092695236206,
      "epoch": 4.359579100145138,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003235130445027693,
      "loss": 4.8089,
      "mean_token_accuracy": 0.23170243352651596,
      "num_tokens": 99763361.0,
      "step": 48060
    },
    {
      "entropy": 5.679444265365601,
      "epoch": 4.360032656023222,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032348144864395383,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22190809696912767,
      "num_tokens": 99773813.0,
      "step": 48065
    },
    {
      "entropy": 5.625652742385864,
      "epoch": 4.360486211901306,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003234498517825297,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22292246371507646,
      "num_tokens": 99784812.0,
      "step": 48070
    },
    {
      "entropy": 5.630132246017456,
      "epoch": 4.36093976777939,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032341825391915026,
      "loss": 4.8308,
      "mean_token_accuracy": 0.23031337261199952,
      "num_tokens": 99795197.0,
      "step": 48075
    },
    {
      "entropy": 5.553213262557984,
      "epoch": 4.361393323657475,
      "grad_norm": 1.3125,
      "learning_rate": 0.000323386655054469,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22989968955516815,
      "num_tokens": 99805883.0,
      "step": 48080
    },
    {
      "entropy": 5.625237941741943,
      "epoch": 4.361846879535559,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003233550551891394,
      "loss": 4.8582,
      "mean_token_accuracy": 0.2221593976020813,
      "num_tokens": 99816940.0,
      "step": 48085
    },
    {
      "entropy": 5.608311748504638,
      "epoch": 4.362300435413643,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032332345432381493,
      "loss": 4.8283,
      "mean_token_accuracy": 0.22529232650995254,
      "num_tokens": 99827926.0,
      "step": 48090
    },
    {
      "entropy": 5.570586490631103,
      "epoch": 4.3627539912917275,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003232918524591491,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23092973232269287,
      "num_tokens": 99837900.0,
      "step": 48095
    },
    {
      "entropy": 5.670150232315064,
      "epoch": 4.363207547169811,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003232602495957955,
      "loss": 4.9058,
      "mean_token_accuracy": 0.21870438009500504,
      "num_tokens": 99848655.0,
      "step": 48100
    },
    {
      "entropy": 5.597477293014526,
      "epoch": 4.363661103047895,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032322864573440757,
      "loss": 4.8162,
      "mean_token_accuracy": 0.22718170136213303,
      "num_tokens": 99859715.0,
      "step": 48105
    },
    {
      "entropy": 5.598419189453125,
      "epoch": 4.364114658925979,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003231970408756389,
      "loss": 4.806,
      "mean_token_accuracy": 0.23490022271871566,
      "num_tokens": 99870175.0,
      "step": 48110
    },
    {
      "entropy": 5.666688537597656,
      "epoch": 4.364568214804064,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003231654350201432,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22053509801626206,
      "num_tokens": 99879794.0,
      "step": 48115
    },
    {
      "entropy": 5.642016839981079,
      "epoch": 4.365021770682148,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003231338281685741,
      "loss": 4.8487,
      "mean_token_accuracy": 0.23228411227464676,
      "num_tokens": 99889418.0,
      "step": 48120
    },
    {
      "entropy": 5.670563364028931,
      "epoch": 4.365475326560232,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003231022203215851,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22479935437440873,
      "num_tokens": 99898741.0,
      "step": 48125
    },
    {
      "entropy": 5.580864524841308,
      "epoch": 4.3659288824383164,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032307061147982994,
      "loss": 4.809,
      "mean_token_accuracy": 0.2266858011484146,
      "num_tokens": 99909167.0,
      "step": 48130
    },
    {
      "entropy": 5.601598405838013,
      "epoch": 4.366382438316401,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003230390016439623,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2315944418311119,
      "num_tokens": 99918786.0,
      "step": 48135
    },
    {
      "entropy": 5.6030748844146725,
      "epoch": 4.366835994194485,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003230073908146359,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22128552496433257,
      "num_tokens": 99928758.0,
      "step": 48140
    },
    {
      "entropy": 5.604924726486206,
      "epoch": 4.367289550072569,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003229757789925044,
      "loss": 4.8243,
      "mean_token_accuracy": 0.22608945071697234,
      "num_tokens": 99939721.0,
      "step": 48145
    },
    {
      "entropy": 5.573545455932617,
      "epoch": 4.3677431059506535,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003229441661782215,
      "loss": 4.7936,
      "mean_token_accuracy": 0.228642600774765,
      "num_tokens": 99949308.0,
      "step": 48150
    },
    {
      "entropy": 5.606756067276001,
      "epoch": 4.368196661828737,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032291255237244104,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22624740302562713,
      "num_tokens": 99960284.0,
      "step": 48155
    },
    {
      "entropy": 5.563920307159424,
      "epoch": 4.368650217706821,
      "grad_norm": 1.515625,
      "learning_rate": 0.00032288093757581684,
      "loss": 4.7561,
      "mean_token_accuracy": 0.23164529204368592,
      "num_tokens": 99970701.0,
      "step": 48160
    },
    {
      "entropy": 5.634937143325805,
      "epoch": 4.369103773584905,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003228493217890026,
      "loss": 4.826,
      "mean_token_accuracy": 0.22586302906274797,
      "num_tokens": 99981362.0,
      "step": 48165
    },
    {
      "entropy": 5.651753520965576,
      "epoch": 4.36955732946299,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003228177050126523,
      "loss": 4.8626,
      "mean_token_accuracy": 0.2321810856461525,
      "num_tokens": 99991383.0,
      "step": 48170
    },
    {
      "entropy": 5.6074323654174805,
      "epoch": 4.370010885341074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032278608724741956,
      "loss": 4.8668,
      "mean_token_accuracy": 0.22491192668676377,
      "num_tokens": 100001176.0,
      "step": 48175
    },
    {
      "entropy": 5.602209663391113,
      "epoch": 4.370464441219158,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003227544684939584,
      "loss": 4.8113,
      "mean_token_accuracy": 0.23598344027996063,
      "num_tokens": 100011467.0,
      "step": 48180
    },
    {
      "entropy": 5.6369462490081785,
      "epoch": 4.3709179970972425,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003227228487529226,
      "loss": 4.8329,
      "mean_token_accuracy": 0.2289295732975006,
      "num_tokens": 100020976.0,
      "step": 48185
    },
    {
      "entropy": 5.586886930465698,
      "epoch": 4.371371552975327,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003226912280249662,
      "loss": 4.8476,
      "mean_token_accuracy": 0.2286357969045639,
      "num_tokens": 100032250.0,
      "step": 48190
    },
    {
      "entropy": 5.636065053939819,
      "epoch": 4.371825108853411,
      "grad_norm": 1.328125,
      "learning_rate": 0.000322659606310743,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22684965431690216,
      "num_tokens": 100043383.0,
      "step": 48195
    },
    {
      "entropy": 5.58125524520874,
      "epoch": 4.372278664731495,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032262798361090705,
      "loss": 4.7675,
      "mean_token_accuracy": 0.23555166125297547,
      "num_tokens": 100053396.0,
      "step": 48200
    },
    {
      "entropy": 5.605345296859741,
      "epoch": 4.37273222060958,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003225963599261123,
      "loss": 4.8466,
      "mean_token_accuracy": 0.2257437750697136,
      "num_tokens": 100064159.0,
      "step": 48205
    },
    {
      "entropy": 5.5784539699554445,
      "epoch": 4.373185776487663,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032256473525701257,
      "loss": 4.7913,
      "mean_token_accuracy": 0.2432887449860573,
      "num_tokens": 100074722.0,
      "step": 48210
    },
    {
      "entropy": 5.565586423873901,
      "epoch": 4.373639332365747,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032253310960426215,
      "loss": 4.8129,
      "mean_token_accuracy": 0.23467101156711578,
      "num_tokens": 100085067.0,
      "step": 48215
    },
    {
      "entropy": 5.620814037322998,
      "epoch": 4.3740928882438315,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032250148296851484,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22941384464502335,
      "num_tokens": 100095806.0,
      "step": 48220
    },
    {
      "entropy": 5.7151954650878904,
      "epoch": 4.374546444121916,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003224698553504247,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21616954654455184,
      "num_tokens": 100106069.0,
      "step": 48225
    },
    {
      "entropy": 5.68791880607605,
      "epoch": 4.375,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032243822675064585,
      "loss": 4.9077,
      "mean_token_accuracy": 0.21458782106637955,
      "num_tokens": 100117136.0,
      "step": 48230
    },
    {
      "entropy": 5.684746265411377,
      "epoch": 4.375453555878084,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003224065971698324,
      "loss": 4.8903,
      "mean_token_accuracy": 0.211772383749485,
      "num_tokens": 100126757.0,
      "step": 48235
    },
    {
      "entropy": 5.774072074890137,
      "epoch": 4.3759071117561685,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003223749666086385,
      "loss": 4.9884,
      "mean_token_accuracy": 0.21950967162847518,
      "num_tokens": 100138431.0,
      "step": 48240
    },
    {
      "entropy": 5.654997777938843,
      "epoch": 4.376360667634253,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003223433350677182,
      "loss": 4.8716,
      "mean_token_accuracy": 0.22089108526706697,
      "num_tokens": 100148919.0,
      "step": 48245
    },
    {
      "entropy": 5.688544416427613,
      "epoch": 4.376814223512337,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003223117025477256,
      "loss": 4.9795,
      "mean_token_accuracy": 0.2220478057861328,
      "num_tokens": 100159175.0,
      "step": 48250
    },
    {
      "entropy": 5.616968393325806,
      "epoch": 4.37726777939042,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003222800690493151,
      "loss": 4.8261,
      "mean_token_accuracy": 0.22433843165636064,
      "num_tokens": 100169451.0,
      "step": 48255
    },
    {
      "entropy": 5.649549961090088,
      "epoch": 4.377721335268505,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003222484345731405,
      "loss": 4.8607,
      "mean_token_accuracy": 0.21963613480329514,
      "num_tokens": 100180077.0,
      "step": 48260
    },
    {
      "entropy": 5.629762649536133,
      "epoch": 4.378174891146589,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032221679911985634,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22469949424266816,
      "num_tokens": 100190175.0,
      "step": 48265
    },
    {
      "entropy": 5.5940595149993895,
      "epoch": 4.378628447024673,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032218516269011676,
      "loss": 4.7848,
      "mean_token_accuracy": 0.2327679067850113,
      "num_tokens": 100200285.0,
      "step": 48270
    },
    {
      "entropy": 5.471023225784302,
      "epoch": 4.3790820029027575,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032215352528457597,
      "loss": 4.6874,
      "mean_token_accuracy": 0.23192292749881743,
      "num_tokens": 100210685.0,
      "step": 48275
    },
    {
      "entropy": 5.593541526794434,
      "epoch": 4.379535558780842,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032212188690388825,
      "loss": 4.8065,
      "mean_token_accuracy": 0.2345779076218605,
      "num_tokens": 100220486.0,
      "step": 48280
    },
    {
      "entropy": 5.596224594116211,
      "epoch": 4.379989114658926,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032209024754870796,
      "loss": 4.7713,
      "mean_token_accuracy": 0.22448755353689193,
      "num_tokens": 100229007.0,
      "step": 48285
    },
    {
      "entropy": 5.603769111633301,
      "epoch": 4.38044267053701,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003220586072196893,
      "loss": 4.7928,
      "mean_token_accuracy": 0.22924472391605377,
      "num_tokens": 100240363.0,
      "step": 48290
    },
    {
      "entropy": 5.555106067657471,
      "epoch": 4.380896226415095,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032202696591748667,
      "loss": 4.8133,
      "mean_token_accuracy": 0.22545038312673568,
      "num_tokens": 100251193.0,
      "step": 48295
    },
    {
      "entropy": 5.593208265304566,
      "epoch": 4.381349782293179,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032199532364275436,
      "loss": 4.8372,
      "mean_token_accuracy": 0.2314404234290123,
      "num_tokens": 100261248.0,
      "step": 48300
    },
    {
      "entropy": 5.630082082748413,
      "epoch": 4.381803338171263,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003219636803961468,
      "loss": 4.8222,
      "mean_token_accuracy": 0.2231413394212723,
      "num_tokens": 100271174.0,
      "step": 48305
    },
    {
      "entropy": 5.6366759777069095,
      "epoch": 4.3822568940493465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032193203617831835,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22499779909849166,
      "num_tokens": 100282428.0,
      "step": 48310
    },
    {
      "entropy": 5.639781951904297,
      "epoch": 4.382710449927431,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032190039098992345,
      "loss": 4.8884,
      "mean_token_accuracy": 0.2247499167919159,
      "num_tokens": 100292559.0,
      "step": 48315
    },
    {
      "entropy": 5.567109680175781,
      "epoch": 4.383164005805515,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003218687448316166,
      "loss": 4.7679,
      "mean_token_accuracy": 0.23415702134370803,
      "num_tokens": 100303956.0,
      "step": 48320
    },
    {
      "entropy": 5.557625818252563,
      "epoch": 4.383617561683599,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032183709770405204,
      "loss": 4.8472,
      "mean_token_accuracy": 0.22660403400659562,
      "num_tokens": 100315281.0,
      "step": 48325
    },
    {
      "entropy": 5.647139406204223,
      "epoch": 4.3840711175616836,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003218054496078844,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2235768496990204,
      "num_tokens": 100326761.0,
      "step": 48330
    },
    {
      "entropy": 5.664307975769043,
      "epoch": 4.384524673439768,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032177380054376813,
      "loss": 4.914,
      "mean_token_accuracy": 0.22473444789648056,
      "num_tokens": 100337219.0,
      "step": 48335
    },
    {
      "entropy": 5.655197381973267,
      "epoch": 4.384978229317852,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032174215051235776,
      "loss": 4.8446,
      "mean_token_accuracy": 0.2313897728919983,
      "num_tokens": 100347135.0,
      "step": 48340
    },
    {
      "entropy": 5.535141563415527,
      "epoch": 4.385431785195936,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032171049951430775,
      "loss": 4.7254,
      "mean_token_accuracy": 0.24118565171957015,
      "num_tokens": 100357476.0,
      "step": 48345
    },
    {
      "entropy": 5.618880414962769,
      "epoch": 4.385885341074021,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032167884755027285,
      "loss": 4.8117,
      "mean_token_accuracy": 0.22324383705854417,
      "num_tokens": 100367860.0,
      "step": 48350
    },
    {
      "entropy": 5.626428651809692,
      "epoch": 4.386338896952105,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003216471946209074,
      "loss": 4.8189,
      "mean_token_accuracy": 0.2282093808054924,
      "num_tokens": 100378613.0,
      "step": 48355
    },
    {
      "entropy": 5.607592153549194,
      "epoch": 4.386792452830189,
      "grad_norm": 1.296875,
      "learning_rate": 0.000321615540726866,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2256610095500946,
      "num_tokens": 100389499.0,
      "step": 48360
    },
    {
      "entropy": 5.673888397216797,
      "epoch": 4.3872460087082725,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032158388586880346,
      "loss": 4.889,
      "mean_token_accuracy": 0.2260989099740982,
      "num_tokens": 100400538.0,
      "step": 48365
    },
    {
      "entropy": 5.6334020614624025,
      "epoch": 4.387699564586357,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032155223004737423,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23078583031892777,
      "num_tokens": 100411528.0,
      "step": 48370
    },
    {
      "entropy": 5.612975835800171,
      "epoch": 4.388153120464441,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032152057326323305,
      "loss": 4.861,
      "mean_token_accuracy": 0.21818068921566008,
      "num_tokens": 100421349.0,
      "step": 48375
    },
    {
      "entropy": 5.582864046096802,
      "epoch": 4.388606676342525,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032148891551703445,
      "loss": 4.7939,
      "mean_token_accuracy": 0.22781539112329482,
      "num_tokens": 100431132.0,
      "step": 48380
    },
    {
      "entropy": 5.573835468292236,
      "epoch": 4.38906023222061,
      "grad_norm": 1.375,
      "learning_rate": 0.0003214572568094334,
      "loss": 4.8524,
      "mean_token_accuracy": 0.231012025475502,
      "num_tokens": 100441436.0,
      "step": 48385
    },
    {
      "entropy": 5.561656141281128,
      "epoch": 4.389513788098694,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003214255971410843,
      "loss": 4.716,
      "mean_token_accuracy": 0.23499952703714372,
      "num_tokens": 100451711.0,
      "step": 48390
    },
    {
      "entropy": 5.75250096321106,
      "epoch": 4.389967343976778,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000321393936512642,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21917416900396347,
      "num_tokens": 100461177.0,
      "step": 48395
    },
    {
      "entropy": 5.627883100509644,
      "epoch": 4.390420899854862,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032136227492476145,
      "loss": 4.8678,
      "mean_token_accuracy": 0.22619658708572388,
      "num_tokens": 100471391.0,
      "step": 48400
    },
    {
      "entropy": 5.669579935073853,
      "epoch": 4.390874455732947,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003213306123780971,
      "loss": 4.8398,
      "mean_token_accuracy": 0.2283684179186821,
      "num_tokens": 100480747.0,
      "step": 48405
    },
    {
      "entropy": 5.640640926361084,
      "epoch": 4.39132801161103,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032129894887330396,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23353632390499116,
      "num_tokens": 100491451.0,
      "step": 48410
    },
    {
      "entropy": 5.686919069290161,
      "epoch": 4.391781567489114,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003212672844110367,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2256743222475052,
      "num_tokens": 100501801.0,
      "step": 48415
    },
    {
      "entropy": 5.588483047485352,
      "epoch": 4.392235123367199,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032123561899195026,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2326078712940216,
      "num_tokens": 100511857.0,
      "step": 48420
    },
    {
      "entropy": 5.662438011169433,
      "epoch": 4.392688679245283,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003212039526166994,
      "loss": 4.8449,
      "mean_token_accuracy": 0.22411953359842302,
      "num_tokens": 100523654.0,
      "step": 48425
    },
    {
      "entropy": 5.734838819503784,
      "epoch": 4.393142235123367,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032117228528593907,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2273124173283577,
      "num_tokens": 100533030.0,
      "step": 48430
    },
    {
      "entropy": 5.675047111511231,
      "epoch": 4.393595791001451,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000321140617000324,
      "loss": 4.8835,
      "mean_token_accuracy": 0.21827754825353624,
      "num_tokens": 100543414.0,
      "step": 48435
    },
    {
      "entropy": 5.617260217666626,
      "epoch": 4.394049346879536,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032110894776050936,
      "loss": 4.8376,
      "mean_token_accuracy": 0.22989493012428283,
      "num_tokens": 100553303.0,
      "step": 48440
    },
    {
      "entropy": 5.556719779968262,
      "epoch": 4.39450290275762,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032107727756714984,
      "loss": 4.8243,
      "mean_token_accuracy": 0.22862074226140977,
      "num_tokens": 100563030.0,
      "step": 48445
    },
    {
      "entropy": 5.6923400402069095,
      "epoch": 4.394956458635704,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003210456064209005,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22064014822244643,
      "num_tokens": 100573216.0,
      "step": 48450
    },
    {
      "entropy": 5.606204080581665,
      "epoch": 4.395410014513788,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032101393432241635,
      "loss": 4.7836,
      "mean_token_accuracy": 0.2295920431613922,
      "num_tokens": 100582646.0,
      "step": 48455
    },
    {
      "entropy": 5.625340700149536,
      "epoch": 4.395863570391873,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032098226127235215,
      "loss": 4.8925,
      "mean_token_accuracy": 0.21771706640720367,
      "num_tokens": 100593689.0,
      "step": 48460
    },
    {
      "entropy": 5.629131698608399,
      "epoch": 4.396317126269956,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003209505872713632,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2198092445731163,
      "num_tokens": 100604721.0,
      "step": 48465
    },
    {
      "entropy": 5.6555229187011715,
      "epoch": 4.39677068214804,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032091891232010434,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2258932814002037,
      "num_tokens": 100614684.0,
      "step": 48470
    },
    {
      "entropy": 5.651744222640991,
      "epoch": 4.397224238026125,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032088723641923066,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23021347522735597,
      "num_tokens": 100624434.0,
      "step": 48475
    },
    {
      "entropy": 5.723307943344116,
      "epoch": 4.397677793904209,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032085555956939726,
      "loss": 4.9884,
      "mean_token_accuracy": 0.21019589602947236,
      "num_tokens": 100635656.0,
      "step": 48480
    },
    {
      "entropy": 5.53769326210022,
      "epoch": 4.398131349782293,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003208238817712593,
      "loss": 4.7237,
      "mean_token_accuracy": 0.23821058869361877,
      "num_tokens": 100645749.0,
      "step": 48485
    },
    {
      "entropy": 5.54173264503479,
      "epoch": 4.398584905660377,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003207922030254717,
      "loss": 4.8195,
      "mean_token_accuracy": 0.22556515634059907,
      "num_tokens": 100655814.0,
      "step": 48490
    },
    {
      "entropy": 5.586210060119629,
      "epoch": 4.399038461538462,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032076052333268964,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2310555875301361,
      "num_tokens": 100665579.0,
      "step": 48495
    },
    {
      "entropy": 5.72037353515625,
      "epoch": 4.399492017416546,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003207288426935683,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21578879058361053,
      "num_tokens": 100676421.0,
      "step": 48500
    },
    {
      "entropy": 5.623065233230591,
      "epoch": 4.39994557329463,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003206971611087629,
      "loss": 4.8135,
      "mean_token_accuracy": 0.23469114899635315,
      "num_tokens": 100686689.0,
      "step": 48505
    },
    {
      "entropy": 5.720742797851562,
      "epoch": 4.4003991291727145,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003206654785789285,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22444673776626586,
      "num_tokens": 100697324.0,
      "step": 48510
    },
    {
      "entropy": 5.561285924911499,
      "epoch": 4.400852685050798,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032063379510472036,
      "loss": 4.7743,
      "mean_token_accuracy": 0.23481216579675673,
      "num_tokens": 100706003.0,
      "step": 48515
    },
    {
      "entropy": 5.662320470809936,
      "epoch": 4.401306240928882,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003206021106867938,
      "loss": 4.898,
      "mean_token_accuracy": 0.23185317516326903,
      "num_tokens": 100716755.0,
      "step": 48520
    },
    {
      "entropy": 5.63327169418335,
      "epoch": 4.401759796806966,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003205704253258039,
      "loss": 4.896,
      "mean_token_accuracy": 0.22048550546169282,
      "num_tokens": 100727846.0,
      "step": 48525
    },
    {
      "entropy": 5.641690111160278,
      "epoch": 4.402213352685051,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000320538739022406,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22127152979373932,
      "num_tokens": 100738293.0,
      "step": 48530
    },
    {
      "entropy": 5.623243618011474,
      "epoch": 4.402666908563135,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003205070517772554,
      "loss": 4.7829,
      "mean_token_accuracy": 0.23062654286623002,
      "num_tokens": 100748031.0,
      "step": 48535
    },
    {
      "entropy": 5.581604242324829,
      "epoch": 4.403120464441219,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003204753635910073,
      "loss": 4.8023,
      "mean_token_accuracy": 0.2342190071940422,
      "num_tokens": 100758825.0,
      "step": 48540
    },
    {
      "entropy": 5.630904817581177,
      "epoch": 4.4035740203193035,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032044367446431714,
      "loss": 4.827,
      "mean_token_accuracy": 0.23125179708003998,
      "num_tokens": 100768226.0,
      "step": 48545
    },
    {
      "entropy": 5.584793138504028,
      "epoch": 4.404027576197388,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032041198439784025,
      "loss": 4.7871,
      "mean_token_accuracy": 0.2283258855342865,
      "num_tokens": 100778125.0,
      "step": 48550
    },
    {
      "entropy": 5.6403961181640625,
      "epoch": 4.404481132075472,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003203802933922319,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23356873244047166,
      "num_tokens": 100788670.0,
      "step": 48555
    },
    {
      "entropy": 5.636342239379883,
      "epoch": 4.404934687953556,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032034860144814756,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22310814261436462,
      "num_tokens": 100800077.0,
      "step": 48560
    },
    {
      "entropy": 5.631391000747681,
      "epoch": 4.40538824383164,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003203169085662425,
      "loss": 4.849,
      "mean_token_accuracy": 0.2280895307660103,
      "num_tokens": 100809432.0,
      "step": 48565
    },
    {
      "entropy": 5.612020587921142,
      "epoch": 4.405841799709724,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003202852147471723,
      "loss": 4.8318,
      "mean_token_accuracy": 0.23678984493017197,
      "num_tokens": 100819675.0,
      "step": 48570
    },
    {
      "entropy": 5.653158473968506,
      "epoch": 4.406295355587808,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032025351999159237,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22723868042230605,
      "num_tokens": 100829779.0,
      "step": 48575
    },
    {
      "entropy": 5.624128246307373,
      "epoch": 4.406748911465892,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032022182430015795,
      "loss": 4.9132,
      "mean_token_accuracy": 0.21592796295881272,
      "num_tokens": 100839840.0,
      "step": 48580
    },
    {
      "entropy": 5.60003342628479,
      "epoch": 4.407202467343977,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003201901276735248,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2246597155928612,
      "num_tokens": 100849495.0,
      "step": 48585
    },
    {
      "entropy": 5.600904083251953,
      "epoch": 4.407656023222061,
      "grad_norm": 1.375,
      "learning_rate": 0.00032015843011234826,
      "loss": 4.843,
      "mean_token_accuracy": 0.22320834249258042,
      "num_tokens": 100860214.0,
      "step": 48590
    },
    {
      "entropy": 5.736521768569946,
      "epoch": 4.408109579100145,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032012673161728396,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20974743068218232,
      "num_tokens": 100870773.0,
      "step": 48595
    },
    {
      "entropy": 5.676349830627442,
      "epoch": 4.4085631349782295,
      "grad_norm": 1.375,
      "learning_rate": 0.00032009503218898716,
      "loss": 4.8599,
      "mean_token_accuracy": 0.23422354012727736,
      "num_tokens": 100880654.0,
      "step": 48600
    },
    {
      "entropy": 5.573979330062866,
      "epoch": 4.409016690856314,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003200633318281138,
      "loss": 4.827,
      "mean_token_accuracy": 0.22613150775432586,
      "num_tokens": 100891217.0,
      "step": 48605
    },
    {
      "entropy": 5.600195121765137,
      "epoch": 4.409470246734398,
      "grad_norm": 1.5,
      "learning_rate": 0.00032003163053531913,
      "loss": 4.7571,
      "mean_token_accuracy": 0.23134440630674363,
      "num_tokens": 100901424.0,
      "step": 48610
    },
    {
      "entropy": 5.663672685623169,
      "epoch": 4.409923802612482,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031999992831125893,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22636909037828445,
      "num_tokens": 100912069.0,
      "step": 48615
    },
    {
      "entropy": 5.597124862670898,
      "epoch": 4.410377358490566,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003199682251565887,
      "loss": 4.8514,
      "mean_token_accuracy": 0.21679269075393676,
      "num_tokens": 100922750.0,
      "step": 48620
    },
    {
      "entropy": 5.66830530166626,
      "epoch": 4.41083091436865,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031993652107196416,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22561971694231034,
      "num_tokens": 100933052.0,
      "step": 48625
    },
    {
      "entropy": 5.667844963073731,
      "epoch": 4.411284470246734,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031990481605804084,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22934581339359283,
      "num_tokens": 100944519.0,
      "step": 48630
    },
    {
      "entropy": 5.597740507125854,
      "epoch": 4.4117380261248185,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031987311011547455,
      "loss": 4.8908,
      "mean_token_accuracy": 0.21607525944709777,
      "num_tokens": 100954901.0,
      "step": 48635
    },
    {
      "entropy": 5.611607837677002,
      "epoch": 4.412191582002903,
      "grad_norm": 1.25,
      "learning_rate": 0.0003198414032449209,
      "loss": 4.829,
      "mean_token_accuracy": 0.22785748392343522,
      "num_tokens": 100964931.0,
      "step": 48640
    },
    {
      "entropy": 5.671269702911377,
      "epoch": 4.412645137880987,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003198096954470356,
      "loss": 4.8379,
      "mean_token_accuracy": 0.23043961226940154,
      "num_tokens": 100974877.0,
      "step": 48645
    },
    {
      "entropy": 5.6868133544921875,
      "epoch": 4.413098693759071,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031977798672247444,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2288626477122307,
      "num_tokens": 100986122.0,
      "step": 48650
    },
    {
      "entropy": 5.669177055358887,
      "epoch": 4.4135522496371555,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031974627707189303,
      "loss": 4.8532,
      "mean_token_accuracy": 0.22193720042705536,
      "num_tokens": 100996489.0,
      "step": 48655
    },
    {
      "entropy": 5.616918563842773,
      "epoch": 4.41400580551524,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031971456649594723,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22649308443069457,
      "num_tokens": 101006453.0,
      "step": 48660
    },
    {
      "entropy": 5.669553613662719,
      "epoch": 4.414459361393324,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003196828549952927,
      "loss": 4.8972,
      "mean_token_accuracy": 0.2235180765390396,
      "num_tokens": 101015609.0,
      "step": 48665
    },
    {
      "entropy": 5.659882640838623,
      "epoch": 4.4149129172714074,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003196511425705855,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2221067875623703,
      "num_tokens": 101027406.0,
      "step": 48670
    },
    {
      "entropy": 5.593004560470581,
      "epoch": 4.415366473149492,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003196194292224812,
      "loss": 4.8528,
      "mean_token_accuracy": 0.2239632248878479,
      "num_tokens": 101038255.0,
      "step": 48675
    },
    {
      "entropy": 5.602800512313843,
      "epoch": 4.415820029027576,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031958771495163574,
      "loss": 4.8444,
      "mean_token_accuracy": 0.22615189254283904,
      "num_tokens": 101048857.0,
      "step": 48680
    },
    {
      "entropy": 5.675091505050659,
      "epoch": 4.41627358490566,
      "grad_norm": 1.359375,
      "learning_rate": 0.000319555999758705,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22014941275119781,
      "num_tokens": 101058640.0,
      "step": 48685
    },
    {
      "entropy": 5.681172180175781,
      "epoch": 4.4167271407837445,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031952428364434484,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21964388638734816,
      "num_tokens": 101069537.0,
      "step": 48690
    },
    {
      "entropy": 5.660367965698242,
      "epoch": 4.417180696661829,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031949256660921104,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2220981553196907,
      "num_tokens": 101080202.0,
      "step": 48695
    },
    {
      "entropy": 5.637272119522095,
      "epoch": 4.417634252539913,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031946084865395975,
      "loss": 4.8585,
      "mean_token_accuracy": 0.22784958630800248,
      "num_tokens": 101089500.0,
      "step": 48700
    },
    {
      "entropy": 5.727157306671143,
      "epoch": 4.418087808417997,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031942912977924674,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2177243188023567,
      "num_tokens": 101099858.0,
      "step": 48705
    },
    {
      "entropy": 5.644932270050049,
      "epoch": 4.418541364296082,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031939740998572806,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2193149819970131,
      "num_tokens": 101110834.0,
      "step": 48710
    },
    {
      "entropy": 5.660656452178955,
      "epoch": 4.418994920174166,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003193656892740595,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2229199677705765,
      "num_tokens": 101121003.0,
      "step": 48715
    },
    {
      "entropy": 5.686791944503784,
      "epoch": 4.419448476052249,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003193339676448973,
      "loss": 4.897,
      "mean_token_accuracy": 0.21884216070175172,
      "num_tokens": 101131925.0,
      "step": 48720
    },
    {
      "entropy": 5.611762952804566,
      "epoch": 4.4199020319303335,
      "grad_norm": 1.375,
      "learning_rate": 0.0003193022450988973,
      "loss": 4.796,
      "mean_token_accuracy": 0.23181460350751876,
      "num_tokens": 101143130.0,
      "step": 48725
    },
    {
      "entropy": 5.5750497341156,
      "epoch": 4.420355587808418,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003192705216367156,
      "loss": 4.8359,
      "mean_token_accuracy": 0.2320162758231163,
      "num_tokens": 101155524.0,
      "step": 48730
    },
    {
      "entropy": 5.571474218368531,
      "epoch": 4.420809143686502,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031923879725900817,
      "loss": 4.8557,
      "mean_token_accuracy": 0.22663215100765227,
      "num_tokens": 101164815.0,
      "step": 48735
    },
    {
      "entropy": 5.652619743347168,
      "epoch": 4.421262699564586,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031920707196643125,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21900852769613266,
      "num_tokens": 101174910.0,
      "step": 48740
    },
    {
      "entropy": 5.55000147819519,
      "epoch": 4.4217162554426706,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003191753457596407,
      "loss": 4.7272,
      "mean_token_accuracy": 0.22915629744529725,
      "num_tokens": 101185698.0,
      "step": 48745
    },
    {
      "entropy": 5.594257402420044,
      "epoch": 4.422169811320755,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031914361863929276,
      "loss": 4.8415,
      "mean_token_accuracy": 0.22395770251750946,
      "num_tokens": 101197001.0,
      "step": 48750
    },
    {
      "entropy": 5.574169874191284,
      "epoch": 4.422623367198839,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003191118906060436,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22597149461507798,
      "num_tokens": 101206705.0,
      "step": 48755
    },
    {
      "entropy": 5.633812665939331,
      "epoch": 4.423076923076923,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031908016166054933,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22382844239473343,
      "num_tokens": 101216484.0,
      "step": 48760
    },
    {
      "entropy": 5.622669839859009,
      "epoch": 4.423530478955008,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003190484318034659,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22478090971708298,
      "num_tokens": 101227169.0,
      "step": 48765
    },
    {
      "entropy": 5.647678089141846,
      "epoch": 4.423984034833092,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003190167010354499,
      "loss": 4.8307,
      "mean_token_accuracy": 0.2298034504055977,
      "num_tokens": 101236956.0,
      "step": 48770
    },
    {
      "entropy": 5.619843292236328,
      "epoch": 4.424437590711175,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003189849693571572,
      "loss": 4.8792,
      "mean_token_accuracy": 0.2212349757552147,
      "num_tokens": 101248681.0,
      "step": 48775
    },
    {
      "entropy": 5.657643365859985,
      "epoch": 4.4248911465892595,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003189532367692441,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2212299734354019,
      "num_tokens": 101258375.0,
      "step": 48780
    },
    {
      "entropy": 5.5852197170257565,
      "epoch": 4.425344702467344,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031892150327236693,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22789230197668076,
      "num_tokens": 101268513.0,
      "step": 48785
    },
    {
      "entropy": 5.657761764526367,
      "epoch": 4.425798258345428,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003188897688671818,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21678464412689208,
      "num_tokens": 101278998.0,
      "step": 48790
    },
    {
      "entropy": 5.725422000885009,
      "epoch": 4.426251814223512,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031885803355434516,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2221237987279892,
      "num_tokens": 101289090.0,
      "step": 48795
    },
    {
      "entropy": 5.651026725769043,
      "epoch": 4.426705370101597,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003188262973345131,
      "loss": 4.8678,
      "mean_token_accuracy": 0.22505754232406616,
      "num_tokens": 101299697.0,
      "step": 48800
    },
    {
      "entropy": 5.648601436614991,
      "epoch": 4.427158925979681,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031879456020834213,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21541815996170044,
      "num_tokens": 101310470.0,
      "step": 48805
    },
    {
      "entropy": 5.655812358856201,
      "epoch": 4.427612481857765,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031876282217648845,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22268507033586502,
      "num_tokens": 101321864.0,
      "step": 48810
    },
    {
      "entropy": 5.646479701995849,
      "epoch": 4.428066037735849,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031873108323960845,
      "loss": 4.8525,
      "mean_token_accuracy": 0.225327168405056,
      "num_tokens": 101331302.0,
      "step": 48815
    },
    {
      "entropy": 5.667787790298462,
      "epoch": 4.428519593613933,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031869934339835856,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2220663234591484,
      "num_tokens": 101341387.0,
      "step": 48820
    },
    {
      "entropy": 5.652537775039673,
      "epoch": 4.428973149492017,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031866760265339514,
      "loss": 4.8276,
      "mean_token_accuracy": 0.22939931303262712,
      "num_tokens": 101351235.0,
      "step": 48825
    },
    {
      "entropy": 5.658036518096924,
      "epoch": 4.429426705370101,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031863586100537453,
      "loss": 4.891,
      "mean_token_accuracy": 0.22458130419254302,
      "num_tokens": 101361289.0,
      "step": 48830
    },
    {
      "entropy": 5.603162002563477,
      "epoch": 4.429880261248186,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003186041184549531,
      "loss": 4.8626,
      "mean_token_accuracy": 0.229951174557209,
      "num_tokens": 101371666.0,
      "step": 48835
    },
    {
      "entropy": 5.606611347198486,
      "epoch": 4.43033381712627,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003185723750027875,
      "loss": 4.8467,
      "mean_token_accuracy": 0.23008151054382325,
      "num_tokens": 101381195.0,
      "step": 48840
    },
    {
      "entropy": 5.698768281936646,
      "epoch": 4.430787373004354,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031854063064953403,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22436307817697526,
      "num_tokens": 101392478.0,
      "step": 48845
    },
    {
      "entropy": 5.67167329788208,
      "epoch": 4.431240928882438,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003185088853958492,
      "loss": 4.9418,
      "mean_token_accuracy": 0.2198147550225258,
      "num_tokens": 101405005.0,
      "step": 48850
    },
    {
      "entropy": 5.592788219451904,
      "epoch": 4.431694484760523,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031847713924238965,
      "loss": 4.862,
      "mean_token_accuracy": 0.22108488380908967,
      "num_tokens": 101416588.0,
      "step": 48855
    },
    {
      "entropy": 5.6366715908050535,
      "epoch": 4.432148040638607,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031844539218981165,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22000187039375305,
      "num_tokens": 101426784.0,
      "step": 48860
    },
    {
      "entropy": 5.615900897979737,
      "epoch": 4.432601596516691,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003184136442387719,
      "loss": 4.8164,
      "mean_token_accuracy": 0.22730119675397872,
      "num_tokens": 101437919.0,
      "step": 48865
    },
    {
      "entropy": 5.627296447753906,
      "epoch": 4.433055152394775,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031838189538992693,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2275494933128357,
      "num_tokens": 101448654.0,
      "step": 48870
    },
    {
      "entropy": 5.633132028579712,
      "epoch": 4.433508708272859,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003183501456439334,
      "loss": 4.8008,
      "mean_token_accuracy": 0.2259764090180397,
      "num_tokens": 101458648.0,
      "step": 48875
    },
    {
      "entropy": 5.589241456985474,
      "epoch": 4.433962264150943,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031831839500144767,
      "loss": 4.8193,
      "mean_token_accuracy": 0.22275903970003127,
      "num_tokens": 101469339.0,
      "step": 48880
    },
    {
      "entropy": 5.649479675292969,
      "epoch": 4.434415820029027,
      "grad_norm": 1.4375,
      "learning_rate": 0.00031828664346312657,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23030509799718857,
      "num_tokens": 101479452.0,
      "step": 48885
    },
    {
      "entropy": 5.638439702987671,
      "epoch": 4.434869375907112,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003182548910296267,
      "loss": 4.8299,
      "mean_token_accuracy": 0.2271818533539772,
      "num_tokens": 101489147.0,
      "step": 48890
    },
    {
      "entropy": 5.565455913543701,
      "epoch": 4.435322931785196,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003182231377016046,
      "loss": 4.7551,
      "mean_token_accuracy": 0.23633746653795243,
      "num_tokens": 101498958.0,
      "step": 48895
    },
    {
      "entropy": 5.670753431320191,
      "epoch": 4.43577648766328,
      "grad_norm": 1.34375,
      "learning_rate": 0.000318191383479717,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21781085878610612,
      "num_tokens": 101509197.0,
      "step": 48900
    },
    {
      "entropy": 5.655673694610596,
      "epoch": 4.436230043541364,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003181596283646205,
      "loss": 4.8764,
      "mean_token_accuracy": 0.2277288943529129,
      "num_tokens": 101520018.0,
      "step": 48905
    },
    {
      "entropy": 5.654379034042359,
      "epoch": 4.436683599419449,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031812787235697206,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22256461381912232,
      "num_tokens": 101530755.0,
      "step": 48910
    },
    {
      "entropy": 5.6234485626220705,
      "epoch": 4.437137155297533,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031809611545742804,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2171500876545906,
      "num_tokens": 101541028.0,
      "step": 48915
    },
    {
      "entropy": 5.651374053955078,
      "epoch": 4.437590711175617,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003180643576666455,
      "loss": 4.8321,
      "mean_token_accuracy": 0.226201656460762,
      "num_tokens": 101550826.0,
      "step": 48920
    },
    {
      "entropy": 5.647074222564697,
      "epoch": 4.4380442670537015,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003180325989852811,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22674667984247207,
      "num_tokens": 101560657.0,
      "step": 48925
    },
    {
      "entropy": 5.578207206726074,
      "epoch": 4.438497822931785,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003180008394139914,
      "loss": 4.837,
      "mean_token_accuracy": 0.23128877580165863,
      "num_tokens": 101570504.0,
      "step": 48930
    },
    {
      "entropy": 5.678635597229004,
      "epoch": 4.438951378809869,
      "grad_norm": 1.453125,
      "learning_rate": 0.00031796907895343356,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21462297290563584,
      "num_tokens": 101581729.0,
      "step": 48935
    },
    {
      "entropy": 5.721919441223145,
      "epoch": 4.439404934687953,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031793731760426414,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2192406252026558,
      "num_tokens": 101592649.0,
      "step": 48940
    },
    {
      "entropy": 5.7387183666229244,
      "epoch": 4.439858490566038,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031790555536714013,
      "loss": 4.9929,
      "mean_token_accuracy": 0.22132169008255004,
      "num_tokens": 101603824.0,
      "step": 48945
    },
    {
      "entropy": 5.660867166519165,
      "epoch": 4.440312046444122,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003178737922427182,
      "loss": 4.9039,
      "mean_token_accuracy": 0.23030058443546295,
      "num_tokens": 101615659.0,
      "step": 48950
    },
    {
      "entropy": 5.729944133758545,
      "epoch": 4.440765602322206,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003178420282316554,
      "loss": 4.9629,
      "mean_token_accuracy": 0.22028013616800307,
      "num_tokens": 101625937.0,
      "step": 48955
    },
    {
      "entropy": 5.648879146575927,
      "epoch": 4.4412191582002905,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003178102633346085,
      "loss": 4.8356,
      "mean_token_accuracy": 0.2290242061018944,
      "num_tokens": 101637352.0,
      "step": 48960
    },
    {
      "entropy": 5.6489893913269045,
      "epoch": 4.441672714078375,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003177784975522344,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23597368597984314,
      "num_tokens": 101647939.0,
      "step": 48965
    },
    {
      "entropy": 5.688599014282227,
      "epoch": 4.442126269956459,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00031774673088519014,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2230267271399498,
      "num_tokens": 101658200.0,
      "step": 48970
    },
    {
      "entropy": 5.644593620300293,
      "epoch": 4.442579825834542,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003177149633341326,
      "loss": 4.8842,
      "mean_token_accuracy": 0.2232554480433464,
      "num_tokens": 101668624.0,
      "step": 48975
    },
    {
      "entropy": 5.705611896514893,
      "epoch": 4.443033381712627,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003176831948997187,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2138810083270073,
      "num_tokens": 101678939.0,
      "step": 48980
    },
    {
      "entropy": 5.621998834609985,
      "epoch": 4.443486937590711,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003176514255826054,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22602296024560928,
      "num_tokens": 101689365.0,
      "step": 48985
    },
    {
      "entropy": 5.61886625289917,
      "epoch": 4.443940493468795,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003176196553834498,
      "loss": 4.8516,
      "mean_token_accuracy": 0.21929872632026673,
      "num_tokens": 101699429.0,
      "step": 48990
    },
    {
      "entropy": 5.633785915374756,
      "epoch": 4.444394049346879,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031758788430290893,
      "loss": 4.906,
      "mean_token_accuracy": 0.2211710035800934,
      "num_tokens": 101709452.0,
      "step": 48995
    },
    {
      "entropy": 5.617898654937744,
      "epoch": 4.444847605224964,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031755611234163966,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23007694482803345,
      "num_tokens": 101719783.0,
      "step": 49000
    },
    {
      "entropy": 5.615143823623657,
      "epoch": 4.445301161103048,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003175243395002992,
      "loss": 4.8238,
      "mean_token_accuracy": 0.23166000097990036,
      "num_tokens": 101729100.0,
      "step": 49005
    },
    {
      "entropy": 5.610653877258301,
      "epoch": 4.445754716981132,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031749256577954453,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2270632952451706,
      "num_tokens": 101738711.0,
      "step": 49010
    },
    {
      "entropy": 5.61120023727417,
      "epoch": 4.4462082728592165,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003174607911800328,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2248609408736229,
      "num_tokens": 101749624.0,
      "step": 49015
    },
    {
      "entropy": 5.615170574188232,
      "epoch": 4.446661828737301,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031742901570242107,
      "loss": 4.8207,
      "mean_token_accuracy": 0.22903938889503478,
      "num_tokens": 101760951.0,
      "step": 49020
    },
    {
      "entropy": 5.630980491638184,
      "epoch": 4.447115384615385,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031739723934736655,
      "loss": 4.8086,
      "mean_token_accuracy": 0.2319727584719658,
      "num_tokens": 101770773.0,
      "step": 49025
    },
    {
      "entropy": 5.645061874389649,
      "epoch": 4.447568940493468,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003173654621155262,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22596594095230102,
      "num_tokens": 101781566.0,
      "step": 49030
    },
    {
      "entropy": 5.59080319404602,
      "epoch": 4.448022496371553,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031733368400755735,
      "loss": 4.8697,
      "mean_token_accuracy": 0.23380719423294066,
      "num_tokens": 101792268.0,
      "step": 49035
    },
    {
      "entropy": 5.683953714370728,
      "epoch": 4.448476052249637,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031730190502411715,
      "loss": 4.927,
      "mean_token_accuracy": 0.22095780223608016,
      "num_tokens": 101802500.0,
      "step": 49040
    },
    {
      "entropy": 5.675297451019287,
      "epoch": 4.448929608127721,
      "grad_norm": 1.25,
      "learning_rate": 0.0003172701251658628,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22274777442216873,
      "num_tokens": 101813851.0,
      "step": 49045
    },
    {
      "entropy": 5.708823013305664,
      "epoch": 4.4493831640058055,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031723834443345134,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2196418285369873,
      "num_tokens": 101824774.0,
      "step": 49050
    },
    {
      "entropy": 5.590535831451416,
      "epoch": 4.44983671988389,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00031720656282754027,
      "loss": 4.7472,
      "mean_token_accuracy": 0.2464963510632515,
      "num_tokens": 101835975.0,
      "step": 49055
    },
    {
      "entropy": 5.597509336471558,
      "epoch": 4.450290275761974,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031717478034878676,
      "loss": 4.8215,
      "mean_token_accuracy": 0.2300755113363266,
      "num_tokens": 101846764.0,
      "step": 49060
    },
    {
      "entropy": 5.602775716781617,
      "epoch": 4.450743831640058,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031714299699784796,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22558782398700714,
      "num_tokens": 101857968.0,
      "step": 49065
    },
    {
      "entropy": 5.631313753128052,
      "epoch": 4.4511973875181425,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003171112127753813,
      "loss": 4.848,
      "mean_token_accuracy": 0.2340001717209816,
      "num_tokens": 101868484.0,
      "step": 49070
    },
    {
      "entropy": 5.667361736297607,
      "epoch": 4.451650943396227,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003170794276820441,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22456134557724,
      "num_tokens": 101877969.0,
      "step": 49075
    },
    {
      "entropy": 5.5901871681213375,
      "epoch": 4.452104499274311,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031704764171849344,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2274169236421585,
      "num_tokens": 101888635.0,
      "step": 49080
    },
    {
      "entropy": 5.620852613449097,
      "epoch": 4.4525580551523944,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031701585488538695,
      "loss": 4.8135,
      "mean_token_accuracy": 0.22826258093118668,
      "num_tokens": 101898340.0,
      "step": 49085
    },
    {
      "entropy": 5.688434362411499,
      "epoch": 4.453011611030479,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003169840671833819,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21651571244001389,
      "num_tokens": 101908993.0,
      "step": 49090
    },
    {
      "entropy": 5.582010126113891,
      "epoch": 4.453465166908563,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003169522786131355,
      "loss": 4.7992,
      "mean_token_accuracy": 0.22893144190311432,
      "num_tokens": 101919675.0,
      "step": 49095
    },
    {
      "entropy": 5.680526208877564,
      "epoch": 4.453918722786647,
      "grad_norm": 1.46875,
      "learning_rate": 0.00031692048917530535,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2221715658903122,
      "num_tokens": 101929949.0,
      "step": 49100
    },
    {
      "entropy": 5.628272008895874,
      "epoch": 4.4543722786647315,
      "grad_norm": 1.4375,
      "learning_rate": 0.00031688869887054887,
      "loss": 4.7412,
      "mean_token_accuracy": 0.23930002003908157,
      "num_tokens": 101940025.0,
      "step": 49105
    },
    {
      "entropy": 5.6292273044586185,
      "epoch": 4.454825834542816,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031685690769952336,
      "loss": 4.815,
      "mean_token_accuracy": 0.23405244201421738,
      "num_tokens": 101950344.0,
      "step": 49110
    },
    {
      "entropy": 5.677900505065918,
      "epoch": 4.4552793904209,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003168251156628863,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22621249109506608,
      "num_tokens": 101961405.0,
      "step": 49115
    },
    {
      "entropy": 5.570927143096924,
      "epoch": 4.455732946298984,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031679332276129523,
      "loss": 4.7409,
      "mean_token_accuracy": 0.23360927402973175,
      "num_tokens": 101972164.0,
      "step": 49120
    },
    {
      "entropy": 5.599050092697143,
      "epoch": 4.456186502177069,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00031676152899540763,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2223188355565071,
      "num_tokens": 101981573.0,
      "step": 49125
    },
    {
      "entropy": 5.5814799785614015,
      "epoch": 4.456640058055152,
      "grad_norm": 1.40625,
      "learning_rate": 0.000316729734365881,
      "loss": 4.8287,
      "mean_token_accuracy": 0.2263931378722191,
      "num_tokens": 101991051.0,
      "step": 49130
    },
    {
      "entropy": 5.601623439788819,
      "epoch": 4.457093613933236,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003166979388733727,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23287868201732637,
      "num_tokens": 102000003.0,
      "step": 49135
    },
    {
      "entropy": 5.647953748703003,
      "epoch": 4.4575471698113205,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003166661425185405,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2145568162202835,
      "num_tokens": 102010489.0,
      "step": 49140
    },
    {
      "entropy": 5.630386590957642,
      "epoch": 4.458000725689405,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031663434530204183,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21789473444223403,
      "num_tokens": 102020486.0,
      "step": 49145
    },
    {
      "entropy": 5.665774726867676,
      "epoch": 4.458454281567489,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003166025472245342,
      "loss": 4.8541,
      "mean_token_accuracy": 0.23006440103054046,
      "num_tokens": 102030098.0,
      "step": 49150
    },
    {
      "entropy": 5.693413496017456,
      "epoch": 4.458907837445573,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003165707482866754,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2243760734796524,
      "num_tokens": 102040850.0,
      "step": 49155
    },
    {
      "entropy": 5.605059003829956,
      "epoch": 4.4593613933236576,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031653894848912296,
      "loss": 4.7881,
      "mean_token_accuracy": 0.22748537957668305,
      "num_tokens": 102051056.0,
      "step": 49160
    },
    {
      "entropy": 5.659085988998413,
      "epoch": 4.459814949201742,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003165071478325344,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22048922926187514,
      "num_tokens": 102061727.0,
      "step": 49165
    },
    {
      "entropy": 5.617341613769531,
      "epoch": 4.460268505079826,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003164753463175675,
      "loss": 4.911,
      "mean_token_accuracy": 0.22240567058324814,
      "num_tokens": 102071501.0,
      "step": 49170
    },
    {
      "entropy": 5.624067211151123,
      "epoch": 4.46072206095791,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031644354394487977,
      "loss": 4.789,
      "mean_token_accuracy": 0.2374361515045166,
      "num_tokens": 102082511.0,
      "step": 49175
    },
    {
      "entropy": 5.680125379562378,
      "epoch": 4.461175616835995,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031641174071512905,
      "loss": 4.8582,
      "mean_token_accuracy": 0.23295336812734604,
      "num_tokens": 102093010.0,
      "step": 49180
    },
    {
      "entropy": 5.596185827255249,
      "epoch": 4.461629172714078,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000316379936628973,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22490974217653276,
      "num_tokens": 102102691.0,
      "step": 49185
    },
    {
      "entropy": 5.623129749298096,
      "epoch": 4.462082728592162,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031634813168706933,
      "loss": 4.828,
      "mean_token_accuracy": 0.23011025190353393,
      "num_tokens": 102112726.0,
      "step": 49190
    },
    {
      "entropy": 5.613856220245362,
      "epoch": 4.4625362844702465,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031631632589007575,
      "loss": 4.8665,
      "mean_token_accuracy": 0.2294480696320534,
      "num_tokens": 102122879.0,
      "step": 49195
    },
    {
      "entropy": 5.670864009857178,
      "epoch": 4.462989840348331,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031628451923864997,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22382826507091522,
      "num_tokens": 102133485.0,
      "step": 49200
    },
    {
      "entropy": 5.685948038101197,
      "epoch": 4.463443396226415,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031625271173344984,
      "loss": 4.9558,
      "mean_token_accuracy": 0.22063474357128143,
      "num_tokens": 102143439.0,
      "step": 49205
    },
    {
      "entropy": 5.599951553344726,
      "epoch": 4.463896952104499,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031622090337513314,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23295689672231673,
      "num_tokens": 102153808.0,
      "step": 49210
    },
    {
      "entropy": 5.707747554779052,
      "epoch": 4.464350507982584,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031618909416435766,
      "loss": 4.9486,
      "mean_token_accuracy": 0.21678301095962524,
      "num_tokens": 102163375.0,
      "step": 49215
    },
    {
      "entropy": 5.644668865203857,
      "epoch": 4.464804063860668,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003161572841017811,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2234131932258606,
      "num_tokens": 102174402.0,
      "step": 49220
    },
    {
      "entropy": 5.554730939865112,
      "epoch": 4.465257619738752,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031612547318806154,
      "loss": 4.7507,
      "mean_token_accuracy": 0.23497088253498077,
      "num_tokens": 102184446.0,
      "step": 49225
    },
    {
      "entropy": 5.63445553779602,
      "epoch": 4.465711175616836,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031609366142385675,
      "loss": 4.855,
      "mean_token_accuracy": 0.22719788253307344,
      "num_tokens": 102194398.0,
      "step": 49230
    },
    {
      "entropy": 5.575746154785156,
      "epoch": 4.46616473149492,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003160618488098245,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23080240488052367,
      "num_tokens": 102205953.0,
      "step": 49235
    },
    {
      "entropy": 5.59887204170227,
      "epoch": 4.466618287373004,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031603003534662273,
      "loss": 4.7981,
      "mean_token_accuracy": 0.23858145326375962,
      "num_tokens": 102215815.0,
      "step": 49240
    },
    {
      "entropy": 5.645167064666748,
      "epoch": 4.467071843251088,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031599822103490946,
      "loss": 4.8025,
      "mean_token_accuracy": 0.22809929847717286,
      "num_tokens": 102225703.0,
      "step": 49245
    },
    {
      "entropy": 5.690086841583252,
      "epoch": 4.467525399129173,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003159664058753425,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2100346252322197,
      "num_tokens": 102236627.0,
      "step": 49250
    },
    {
      "entropy": 5.60487847328186,
      "epoch": 4.467978955007257,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003159345898685797,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22969655096530914,
      "num_tokens": 102246539.0,
      "step": 49255
    },
    {
      "entropy": 5.640522813796997,
      "epoch": 4.468432510885341,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003159027730152793,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22354926466941832,
      "num_tokens": 102256734.0,
      "step": 49260
    },
    {
      "entropy": 5.699540758132935,
      "epoch": 4.468886066763425,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003158709553160991,
      "loss": 4.9858,
      "mean_token_accuracy": 0.21694791465997695,
      "num_tokens": 102267469.0,
      "step": 49265
    },
    {
      "entropy": 5.66622142791748,
      "epoch": 4.46933962264151,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031583913677169705,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22464902997016906,
      "num_tokens": 102278435.0,
      "step": 49270
    },
    {
      "entropy": 5.68406834602356,
      "epoch": 4.469793178519594,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031580731738273135,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22519890069961548,
      "num_tokens": 102289171.0,
      "step": 49275
    },
    {
      "entropy": 5.635126352310181,
      "epoch": 4.470246734397678,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003157754971498599,
      "loss": 4.8132,
      "mean_token_accuracy": 0.2342111811041832,
      "num_tokens": 102299254.0,
      "step": 49280
    },
    {
      "entropy": 5.675919866561889,
      "epoch": 4.4707002902757615,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003157436760737408,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22464556097984315,
      "num_tokens": 102308471.0,
      "step": 49285
    },
    {
      "entropy": 5.6680444240570065,
      "epoch": 4.471153846153846,
      "grad_norm": 1.375,
      "learning_rate": 0.000315711854155032,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22468231320381166,
      "num_tokens": 102319005.0,
      "step": 49290
    },
    {
      "entropy": 5.656484651565552,
      "epoch": 4.47160740203193,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031568003139439185,
      "loss": 4.8522,
      "mean_token_accuracy": 0.23002826273441315,
      "num_tokens": 102329568.0,
      "step": 49295
    },
    {
      "entropy": 5.712450790405273,
      "epoch": 4.472060957910014,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003156482077924781,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2175341084599495,
      "num_tokens": 102340810.0,
      "step": 49300
    },
    {
      "entropy": 5.591939115524292,
      "epoch": 4.472514513788099,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003156163833499491,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2177632838487625,
      "num_tokens": 102351305.0,
      "step": 49305
    },
    {
      "entropy": 5.6574924945831295,
      "epoch": 4.472968069666183,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031558455806746306,
      "loss": 4.8702,
      "mean_token_accuracy": 0.22235261052846908,
      "num_tokens": 102362057.0,
      "step": 49310
    },
    {
      "entropy": 5.718958139419556,
      "epoch": 4.473421625544267,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003155527319456779,
      "loss": 4.9747,
      "mean_token_accuracy": 0.21234172731637954,
      "num_tokens": 102372612.0,
      "step": 49315
    },
    {
      "entropy": 5.604626178741455,
      "epoch": 4.473875181422351,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031552090498525195,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22858582735061644,
      "num_tokens": 102383337.0,
      "step": 49320
    },
    {
      "entropy": 5.644073534011841,
      "epoch": 4.474328737300436,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003154890771868433,
      "loss": 4.8705,
      "mean_token_accuracy": 0.221617491543293,
      "num_tokens": 102394196.0,
      "step": 49325
    },
    {
      "entropy": 5.694830465316772,
      "epoch": 4.47478229317852,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031545724855111025,
      "loss": 4.8972,
      "mean_token_accuracy": 0.23030000925064087,
      "num_tokens": 102404677.0,
      "step": 49330
    },
    {
      "entropy": 5.659955978393555,
      "epoch": 4.475235849056604,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031542541907871095,
      "loss": 4.8113,
      "mean_token_accuracy": 0.23593172878026963,
      "num_tokens": 102413445.0,
      "step": 49335
    },
    {
      "entropy": 5.605733919143677,
      "epoch": 4.475689404934688,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003153935887703038,
      "loss": 4.8546,
      "mean_token_accuracy": 0.22654233127832413,
      "num_tokens": 102424527.0,
      "step": 49340
    },
    {
      "entropy": 5.644651460647583,
      "epoch": 4.476142960812772,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003153617576265469,
      "loss": 4.8434,
      "mean_token_accuracy": 0.2225853055715561,
      "num_tokens": 102434940.0,
      "step": 49345
    },
    {
      "entropy": 5.713822889328003,
      "epoch": 4.476596516690856,
      "grad_norm": 1.546875,
      "learning_rate": 0.00031532992564809847,
      "loss": 4.946,
      "mean_token_accuracy": 0.22491227984428405,
      "num_tokens": 102446678.0,
      "step": 49350
    },
    {
      "entropy": 5.625935125350952,
      "epoch": 4.47705007256894,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000315298092835617,
      "loss": 4.837,
      "mean_token_accuracy": 0.23035448789596558,
      "num_tokens": 102457255.0,
      "step": 49355
    },
    {
      "entropy": 5.704388761520386,
      "epoch": 4.477503628447025,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031526625918976066,
      "loss": 4.9994,
      "mean_token_accuracy": 0.22258975505828857,
      "num_tokens": 102468385.0,
      "step": 49360
    },
    {
      "entropy": 5.642435932159424,
      "epoch": 4.477957184325109,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031523442471118783,
      "loss": 4.8271,
      "mean_token_accuracy": 0.22323777824640273,
      "num_tokens": 102478204.0,
      "step": 49365
    },
    {
      "entropy": 5.683664608001709,
      "epoch": 4.478410740203193,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031520258940055684,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22842807918787003,
      "num_tokens": 102488492.0,
      "step": 49370
    },
    {
      "entropy": 5.612175607681275,
      "epoch": 4.4788642960812775,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031517075325852606,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22520501911640167,
      "num_tokens": 102500309.0,
      "step": 49375
    },
    {
      "entropy": 5.581572723388672,
      "epoch": 4.479317851959362,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003151389162857539,
      "loss": 4.8606,
      "mean_token_accuracy": 0.2259416624903679,
      "num_tokens": 102511196.0,
      "step": 49380
    },
    {
      "entropy": 5.625988721847534,
      "epoch": 4.479771407837446,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031510707848289864,
      "loss": 4.8306,
      "mean_token_accuracy": 0.2278372675180435,
      "num_tokens": 102522100.0,
      "step": 49385
    },
    {
      "entropy": 5.601570415496826,
      "epoch": 4.480224963715529,
      "grad_norm": 1.375,
      "learning_rate": 0.00031507523985061886,
      "loss": 4.8243,
      "mean_token_accuracy": 0.23635229766368865,
      "num_tokens": 102531677.0,
      "step": 49390
    },
    {
      "entropy": 5.576582527160644,
      "epoch": 4.480678519593614,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031504340038957294,
      "loss": 4.7561,
      "mean_token_accuracy": 0.2405527040362358,
      "num_tokens": 102541854.0,
      "step": 49395
    },
    {
      "entropy": 5.635606670379639,
      "epoch": 4.481132075471698,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003150115601004192,
      "loss": 4.8901,
      "mean_token_accuracy": 0.22413284480571746,
      "num_tokens": 102551187.0,
      "step": 49400
    },
    {
      "entropy": 5.677726697921753,
      "epoch": 4.481585631349782,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003149797189838162,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2218853160738945,
      "num_tokens": 102562568.0,
      "step": 49405
    },
    {
      "entropy": 5.649278974533081,
      "epoch": 4.482039187227866,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031494787704042243,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22124607861042023,
      "num_tokens": 102573170.0,
      "step": 49410
    },
    {
      "entropy": 5.554230117797852,
      "epoch": 4.482492743105951,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031491603427089644,
      "loss": 4.7148,
      "mean_token_accuracy": 0.23635787069797515,
      "num_tokens": 102583950.0,
      "step": 49415
    },
    {
      "entropy": 5.628840160369873,
      "epoch": 4.482946298984035,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031488419067589665,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22833748757839203,
      "num_tokens": 102594370.0,
      "step": 49420
    },
    {
      "entropy": 5.638507461547851,
      "epoch": 4.483399854862119,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003148523462560816,
      "loss": 4.8328,
      "mean_token_accuracy": 0.2318490266799927,
      "num_tokens": 102604837.0,
      "step": 49425
    },
    {
      "entropy": 5.680521154403687,
      "epoch": 4.4838534107402035,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003148205010121099,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22288707345724107,
      "num_tokens": 102616170.0,
      "step": 49430
    },
    {
      "entropy": 5.62971076965332,
      "epoch": 4.484306966618288,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003147886549446401,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22801705449819565,
      "num_tokens": 102627283.0,
      "step": 49435
    },
    {
      "entropy": 5.692225599288941,
      "epoch": 4.484760522496371,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003147568080543308,
      "loss": 4.896,
      "mean_token_accuracy": 0.223686183989048,
      "num_tokens": 102637087.0,
      "step": 49440
    },
    {
      "entropy": 5.645506477355957,
      "epoch": 4.485214078374455,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003147249603418406,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22555406391620636,
      "num_tokens": 102646818.0,
      "step": 49445
    },
    {
      "entropy": 5.580931901931763,
      "epoch": 4.48566763425254,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003146931118078281,
      "loss": 4.7905,
      "mean_token_accuracy": 0.2350690558552742,
      "num_tokens": 102656940.0,
      "step": 49450
    },
    {
      "entropy": 5.596077346801758,
      "epoch": 4.486121190130624,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003146612624529518,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22616479992866517,
      "num_tokens": 102666841.0,
      "step": 49455
    },
    {
      "entropy": 5.658067512512207,
      "epoch": 4.486574746008708,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031462941227787056,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22293926775455475,
      "num_tokens": 102676821.0,
      "step": 49460
    },
    {
      "entropy": 5.750235271453858,
      "epoch": 4.4870283018867925,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003145975612832431,
      "loss": 4.9707,
      "mean_token_accuracy": 0.2208462417125702,
      "num_tokens": 102687340.0,
      "step": 49465
    },
    {
      "entropy": 5.6162749290466305,
      "epoch": 4.487481857764877,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003145657094697278,
      "loss": 4.8779,
      "mean_token_accuracy": 0.2263274222612381,
      "num_tokens": 102698320.0,
      "step": 49470
    },
    {
      "entropy": 5.654202461242676,
      "epoch": 4.487935413642961,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003145338568379836,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2264561250805855,
      "num_tokens": 102708685.0,
      "step": 49475
    },
    {
      "entropy": 5.609914350509643,
      "epoch": 4.488388969521045,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003145020033886692,
      "loss": 4.8445,
      "mean_token_accuracy": 0.21920885145664215,
      "num_tokens": 102719680.0,
      "step": 49480
    },
    {
      "entropy": 5.587669897079468,
      "epoch": 4.4888425253991295,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031447014912244325,
      "loss": 4.8158,
      "mean_token_accuracy": 0.2316541463136673,
      "num_tokens": 102731156.0,
      "step": 49485
    },
    {
      "entropy": 5.629397773742676,
      "epoch": 4.489296081277214,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003144382940399646,
      "loss": 4.7971,
      "mean_token_accuracy": 0.23393794596195222,
      "num_tokens": 102740790.0,
      "step": 49490
    },
    {
      "entropy": 5.637163686752319,
      "epoch": 4.489749637155297,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000314406438141892,
      "loss": 4.8562,
      "mean_token_accuracy": 0.23081200271844865,
      "num_tokens": 102750672.0,
      "step": 49495
    },
    {
      "entropy": 5.592739486694336,
      "epoch": 4.4902031930333814,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003143745814288842,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21943145841360093,
      "num_tokens": 102760364.0,
      "step": 49500
    },
    {
      "entropy": 5.728304004669189,
      "epoch": 4.490656748911466,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003143427239016001,
      "loss": 4.949,
      "mean_token_accuracy": 0.2228900372982025,
      "num_tokens": 102770415.0,
      "step": 49505
    },
    {
      "entropy": 5.655384111404419,
      "epoch": 4.49111030478955,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003143108655606983,
      "loss": 4.8355,
      "mean_token_accuracy": 0.2335874170064926,
      "num_tokens": 102781203.0,
      "step": 49510
    },
    {
      "entropy": 5.63679141998291,
      "epoch": 4.491563860667634,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031427900640683783,
      "loss": 4.8081,
      "mean_token_accuracy": 0.22937019914388657,
      "num_tokens": 102791071.0,
      "step": 49515
    },
    {
      "entropy": 5.684739351272583,
      "epoch": 4.4920174165457185,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031424714644067754,
      "loss": 4.9223,
      "mean_token_accuracy": 0.22481405287981032,
      "num_tokens": 102801755.0,
      "step": 49520
    },
    {
      "entropy": 5.635535001754761,
      "epoch": 4.492470972423803,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031421528566287623,
      "loss": 4.8982,
      "mean_token_accuracy": 0.22298934906721116,
      "num_tokens": 102812947.0,
      "step": 49525
    },
    {
      "entropy": 5.610058689117432,
      "epoch": 4.492924528301887,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003141834240740928,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22876159697771073,
      "num_tokens": 102822917.0,
      "step": 49530
    },
    {
      "entropy": 5.652454137802124,
      "epoch": 4.493378084179971,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003141515616749862,
      "loss": 4.8553,
      "mean_token_accuracy": 0.2267722964286804,
      "num_tokens": 102833102.0,
      "step": 49535
    },
    {
      "entropy": 5.694706106185913,
      "epoch": 4.493831640058055,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003141196984662153,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23061597794294358,
      "num_tokens": 102843888.0,
      "step": 49540
    },
    {
      "entropy": 5.646428012847901,
      "epoch": 4.494285195936139,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003140878344484391,
      "loss": 4.9251,
      "mean_token_accuracy": 0.22708246111869812,
      "num_tokens": 102854417.0,
      "step": 49545
    },
    {
      "entropy": 5.618627071380615,
      "epoch": 4.494738751814223,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003140559696223165,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22899708449840545,
      "num_tokens": 102865206.0,
      "step": 49550
    },
    {
      "entropy": 5.674997758865357,
      "epoch": 4.4951923076923075,
      "grad_norm": 1.375,
      "learning_rate": 0.0003140241039885065,
      "loss": 4.8268,
      "mean_token_accuracy": 0.23099810779094695,
      "num_tokens": 102874967.0,
      "step": 49555
    },
    {
      "entropy": 5.6058753490447994,
      "epoch": 4.495645863570392,
      "grad_norm": 1.25,
      "learning_rate": 0.0003139922375476681,
      "loss": 4.8259,
      "mean_token_accuracy": 0.2290083423256874,
      "num_tokens": 102884808.0,
      "step": 49560
    },
    {
      "entropy": 5.555175876617431,
      "epoch": 4.496099419448476,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003139603703004603,
      "loss": 4.7562,
      "mean_token_accuracy": 0.2335077315568924,
      "num_tokens": 102895206.0,
      "step": 49565
    },
    {
      "entropy": 5.690444755554199,
      "epoch": 4.49655297532656,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003139285022475421,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22447886019945146,
      "num_tokens": 102906658.0,
      "step": 49570
    },
    {
      "entropy": 5.63068265914917,
      "epoch": 4.4970065312046446,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031389663338957254,
      "loss": 4.856,
      "mean_token_accuracy": 0.22486236840486526,
      "num_tokens": 102917071.0,
      "step": 49575
    },
    {
      "entropy": 5.625520706176758,
      "epoch": 4.497460087082729,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031386476372721067,
      "loss": 4.813,
      "mean_token_accuracy": 0.22809524834156036,
      "num_tokens": 102927280.0,
      "step": 49580
    },
    {
      "entropy": 5.639458465576172,
      "epoch": 4.497913642960813,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003138328932611156,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2187660366296768,
      "num_tokens": 102937670.0,
      "step": 49585
    },
    {
      "entropy": 5.6291083812713625,
      "epoch": 4.498367198838897,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031380102199194646,
      "loss": 4.8932,
      "mean_token_accuracy": 0.21949890106916428,
      "num_tokens": 102947874.0,
      "step": 49590
    },
    {
      "entropy": 5.686755800247193,
      "epoch": 4.498820754716981,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003137691499203622,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2216712489724159,
      "num_tokens": 102959311.0,
      "step": 49595
    },
    {
      "entropy": 5.60427303314209,
      "epoch": 4.499274310595065,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031373727704702217,
      "loss": 4.8266,
      "mean_token_accuracy": 0.22990832924842836,
      "num_tokens": 102969866.0,
      "step": 49600
    },
    {
      "entropy": 5.611121797561646,
      "epoch": 4.499727866473149,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031370540337258523,
      "loss": 4.787,
      "mean_token_accuracy": 0.23213734924793245,
      "num_tokens": 102979672.0,
      "step": 49605
    },
    {
      "entropy": 5.598658466339112,
      "epoch": 4.5001814223512335,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003136735288977108,
      "loss": 4.8597,
      "mean_token_accuracy": 0.22234815508127212,
      "num_tokens": 102991050.0,
      "step": 49610
    },
    {
      "entropy": 5.628060626983642,
      "epoch": 4.500634978229318,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031364165362305783,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22139324247837067,
      "num_tokens": 103001700.0,
      "step": 49615
    },
    {
      "entropy": 5.663005304336548,
      "epoch": 4.501088534107402,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003136097775492857,
      "loss": 4.8897,
      "mean_token_accuracy": 0.220831236243248,
      "num_tokens": 103013013.0,
      "step": 49620
    },
    {
      "entropy": 5.626133155822754,
      "epoch": 4.501542089985486,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003135779006770535,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22515156716108323,
      "num_tokens": 103024110.0,
      "step": 49625
    },
    {
      "entropy": 5.700735235214234,
      "epoch": 4.501995645863571,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003135460230070205,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22093861550092697,
      "num_tokens": 103034991.0,
      "step": 49630
    },
    {
      "entropy": 5.684466886520386,
      "epoch": 4.502449201741655,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031351414453984595,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22591807693243027,
      "num_tokens": 103045634.0,
      "step": 49635
    },
    {
      "entropy": 5.600563383102417,
      "epoch": 4.502902757619739,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000313482265276189,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2290247216820717,
      "num_tokens": 103054999.0,
      "step": 49640
    },
    {
      "entropy": 5.622740888595581,
      "epoch": 4.503356313497823,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031345038521670907,
      "loss": 4.8967,
      "mean_token_accuracy": 0.21916372925043107,
      "num_tokens": 103065163.0,
      "step": 49645
    },
    {
      "entropy": 5.695866346359253,
      "epoch": 4.503809869375907,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003134185043620654,
      "loss": 4.8515,
      "mean_token_accuracy": 0.23310244232416152,
      "num_tokens": 103075979.0,
      "step": 49650
    },
    {
      "entropy": 5.591418838500976,
      "epoch": 4.504263425253991,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003133866227129171,
      "loss": 4.7729,
      "mean_token_accuracy": 0.23156407624483108,
      "num_tokens": 103086124.0,
      "step": 49655
    },
    {
      "entropy": 5.583737373352051,
      "epoch": 4.504716981132075,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031335474026992377,
      "loss": 4.8357,
      "mean_token_accuracy": 0.2247992992401123,
      "num_tokens": 103095330.0,
      "step": 49660
    },
    {
      "entropy": 5.6357416152954105,
      "epoch": 4.50517053701016,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003133228570337447,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22448360025882722,
      "num_tokens": 103106863.0,
      "step": 49665
    },
    {
      "entropy": 5.7022209644317625,
      "epoch": 4.505624092888244,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003132909730050391,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22539257854223252,
      "num_tokens": 103117256.0,
      "step": 49670
    },
    {
      "entropy": 5.669220733642578,
      "epoch": 4.506077648766328,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031325908818446643,
      "loss": 4.866,
      "mean_token_accuracy": 0.22613943964242936,
      "num_tokens": 103127601.0,
      "step": 49675
    },
    {
      "entropy": 5.62586145401001,
      "epoch": 4.506531204644412,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00031322720257268607,
      "loss": 4.804,
      "mean_token_accuracy": 0.22655246257781983,
      "num_tokens": 103136868.0,
      "step": 49680
    },
    {
      "entropy": 5.610585689544678,
      "epoch": 4.506984760522497,
      "grad_norm": 1.375,
      "learning_rate": 0.00031319531617035747,
      "loss": 4.8382,
      "mean_token_accuracy": 0.22649689316749572,
      "num_tokens": 103146908.0,
      "step": 49685
    },
    {
      "entropy": 5.585486793518067,
      "epoch": 4.507438316400581,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031316342897813994,
      "loss": 4.8268,
      "mean_token_accuracy": 0.22273413985967636,
      "num_tokens": 103157244.0,
      "step": 49690
    },
    {
      "entropy": 5.591665554046631,
      "epoch": 4.507891872278664,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031313154099669294,
      "loss": 4.8233,
      "mean_token_accuracy": 0.22315495014190673,
      "num_tokens": 103168502.0,
      "step": 49695
    },
    {
      "entropy": 5.618940162658691,
      "epoch": 4.5083454281567485,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003130996522266761,
      "loss": 4.8866,
      "mean_token_accuracy": 0.21917984038591384,
      "num_tokens": 103179015.0,
      "step": 49700
    },
    {
      "entropy": 5.6941381931304935,
      "epoch": 4.508798984034833,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003130677626687485,
      "loss": 4.8598,
      "mean_token_accuracy": 0.23338279128074646,
      "num_tokens": 103189200.0,
      "step": 49705
    },
    {
      "entropy": 5.717541599273682,
      "epoch": 4.509252539912917,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003130358723235702,
      "loss": 5.0001,
      "mean_token_accuracy": 0.21582192182540894,
      "num_tokens": 103200683.0,
      "step": 49710
    },
    {
      "entropy": 5.694086122512817,
      "epoch": 4.509706095791001,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031300398119180013,
      "loss": 4.8874,
      "mean_token_accuracy": 0.23135345727205275,
      "num_tokens": 103210992.0,
      "step": 49715
    },
    {
      "entropy": 5.619565343856811,
      "epoch": 4.510159651669086,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003129720892740981,
      "loss": 4.811,
      "mean_token_accuracy": 0.2283020481467247,
      "num_tokens": 103220783.0,
      "step": 49720
    },
    {
      "entropy": 5.580638074874878,
      "epoch": 4.51061320754717,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003129401965711236,
      "loss": 4.8205,
      "mean_token_accuracy": 0.2291475862264633,
      "num_tokens": 103231593.0,
      "step": 49725
    },
    {
      "entropy": 5.640674829483032,
      "epoch": 4.511066763425254,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003129083030835361,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22020932137966157,
      "num_tokens": 103242416.0,
      "step": 49730
    },
    {
      "entropy": 5.580939674377442,
      "epoch": 4.511520319303338,
      "grad_norm": 1.5,
      "learning_rate": 0.00031287640881199534,
      "loss": 4.7938,
      "mean_token_accuracy": 0.23655528724193572,
      "num_tokens": 103252535.0,
      "step": 49735
    },
    {
      "entropy": 5.688248538970948,
      "epoch": 4.511973875181423,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003128445137571607,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21797917038202286,
      "num_tokens": 103263039.0,
      "step": 49740
    },
    {
      "entropy": 5.6719725131988525,
      "epoch": 4.512427431059507,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031281261791969196,
      "loss": 4.8322,
      "mean_token_accuracy": 0.2270725890994072,
      "num_tokens": 103273304.0,
      "step": 49745
    },
    {
      "entropy": 5.695727968215943,
      "epoch": 4.51288098693759,
      "grad_norm": 1.25,
      "learning_rate": 0.00031278072130024856,
      "loss": 4.967,
      "mean_token_accuracy": 0.22278642058372497,
      "num_tokens": 103283688.0,
      "step": 49750
    },
    {
      "entropy": 5.70925235748291,
      "epoch": 4.513334542815675,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003127488238994902,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2273602768778801,
      "num_tokens": 103293572.0,
      "step": 49755
    },
    {
      "entropy": 5.63466739654541,
      "epoch": 4.513788098693759,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003127169257180766,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22488928735256195,
      "num_tokens": 103303474.0,
      "step": 49760
    },
    {
      "entropy": 5.742396974563599,
      "epoch": 4.514241654571843,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003126850267566674,
      "loss": 4.988,
      "mean_token_accuracy": 0.21695507764816285,
      "num_tokens": 103314343.0,
      "step": 49765
    },
    {
      "entropy": 5.656146144866943,
      "epoch": 4.514695210449927,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003126531270159221,
      "loss": 4.8449,
      "mean_token_accuracy": 0.23239742666482927,
      "num_tokens": 103323942.0,
      "step": 49770
    },
    {
      "entropy": 5.68074107170105,
      "epoch": 4.515148766328012,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003126212264965007,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2215219870209694,
      "num_tokens": 103333785.0,
      "step": 49775
    },
    {
      "entropy": 5.585204124450684,
      "epoch": 4.515602322206096,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031258932519906266,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23268220573663712,
      "num_tokens": 103344184.0,
      "step": 49780
    },
    {
      "entropy": 5.66937026977539,
      "epoch": 4.51605587808418,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031255742312426786,
      "loss": 4.937,
      "mean_token_accuracy": 0.21990395188331605,
      "num_tokens": 103355085.0,
      "step": 49785
    },
    {
      "entropy": 5.599531793594361,
      "epoch": 4.5165094339622645,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031252552027277595,
      "loss": 4.8183,
      "mean_token_accuracy": 0.22387731373310088,
      "num_tokens": 103365942.0,
      "step": 49790
    },
    {
      "entropy": 5.601382255554199,
      "epoch": 4.516962989840349,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003124936166452467,
      "loss": 4.7904,
      "mean_token_accuracy": 0.2368833526968956,
      "num_tokens": 103375844.0,
      "step": 49795
    },
    {
      "entropy": 5.6356090068817135,
      "epoch": 4.517416545718433,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003124617122423399,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23091549575328826,
      "num_tokens": 103385871.0,
      "step": 49800
    },
    {
      "entropy": 5.629478406906128,
      "epoch": 4.517870101596516,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003124298070647154,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22487613409757615,
      "num_tokens": 103396888.0,
      "step": 49805
    },
    {
      "entropy": 5.637585067749024,
      "epoch": 4.518323657474601,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003123979011130329,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22664490342140198,
      "num_tokens": 103406913.0,
      "step": 49810
    },
    {
      "entropy": 5.637938594818115,
      "epoch": 4.518777213352685,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003123659943879523,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23366693258285523,
      "num_tokens": 103416485.0,
      "step": 49815
    },
    {
      "entropy": 5.569645833969116,
      "epoch": 4.519230769230769,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003123340868901334,
      "loss": 4.7754,
      "mean_token_accuracy": 0.23003117740154266,
      "num_tokens": 103427476.0,
      "step": 49820
    },
    {
      "entropy": 5.574479389190674,
      "epoch": 4.519684325108853,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031230217862023604,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22834824174642562,
      "num_tokens": 103437500.0,
      "step": 49825
    },
    {
      "entropy": 5.60947322845459,
      "epoch": 4.520137880986938,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031227026957892013,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23218510150909424,
      "num_tokens": 103447646.0,
      "step": 49830
    },
    {
      "entropy": 5.601420116424561,
      "epoch": 4.520591436865022,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031223835976684555,
      "loss": 4.8183,
      "mean_token_accuracy": 0.22533762454986572,
      "num_tokens": 103457698.0,
      "step": 49835
    },
    {
      "entropy": 5.6715521812438965,
      "epoch": 4.521044992743106,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031220644918467224,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22608720511198044,
      "num_tokens": 103467566.0,
      "step": 49840
    },
    {
      "entropy": 5.644864559173584,
      "epoch": 4.5214985486211905,
      "grad_norm": 1.5,
      "learning_rate": 0.00031217453783306007,
      "loss": 4.8748,
      "mean_token_accuracy": 0.22261758148670197,
      "num_tokens": 103477724.0,
      "step": 49845
    },
    {
      "entropy": 5.638568878173828,
      "epoch": 4.521952104499274,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003121426257126689,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22927780747413634,
      "num_tokens": 103487864.0,
      "step": 49850
    },
    {
      "entropy": 5.585676527023315,
      "epoch": 4.522405660377358,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031211071282415885,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23295720666646957,
      "num_tokens": 103497229.0,
      "step": 49855
    },
    {
      "entropy": 5.5994843482971195,
      "epoch": 4.522859216255442,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003120787991681898,
      "loss": 4.7835,
      "mean_token_accuracy": 0.23967131227254868,
      "num_tokens": 103506984.0,
      "step": 49860
    },
    {
      "entropy": 5.579676389694214,
      "epoch": 4.523312772133527,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003120468847454217,
      "loss": 4.8178,
      "mean_token_accuracy": 0.22419394552707672,
      "num_tokens": 103517287.0,
      "step": 49865
    },
    {
      "entropy": 5.592616367340088,
      "epoch": 4.523766328011611,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031201496955651466,
      "loss": 4.8432,
      "mean_token_accuracy": 0.22904193699359893,
      "num_tokens": 103528575.0,
      "step": 49870
    },
    {
      "entropy": 5.696830129623413,
      "epoch": 4.524219883889695,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031198305360212854,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21811314672231674,
      "num_tokens": 103538406.0,
      "step": 49875
    },
    {
      "entropy": 5.683716821670532,
      "epoch": 4.5246734397677795,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003119511368829235,
      "loss": 4.921,
      "mean_token_accuracy": 0.22130908071994781,
      "num_tokens": 103548738.0,
      "step": 49880
    },
    {
      "entropy": 5.566265296936035,
      "epoch": 4.525126995645864,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031191921939955947,
      "loss": 4.8034,
      "mean_token_accuracy": 0.23026317954063416,
      "num_tokens": 103558755.0,
      "step": 49885
    },
    {
      "entropy": 5.665856456756591,
      "epoch": 4.525580551523948,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003118873011526966,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22525896281003951,
      "num_tokens": 103569115.0,
      "step": 49890
    },
    {
      "entropy": 5.73694109916687,
      "epoch": 4.526034107402032,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003118553821429949,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22743822485208512,
      "num_tokens": 103580453.0,
      "step": 49895
    },
    {
      "entropy": 5.6315490245819095,
      "epoch": 4.5264876632801165,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003118234623711146,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2268270879983902,
      "num_tokens": 103592995.0,
      "step": 49900
    },
    {
      "entropy": 5.558250665664673,
      "epoch": 4.5269412191582,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031179154183771576,
      "loss": 4.7673,
      "mean_token_accuracy": 0.23652720153331758,
      "num_tokens": 103603496.0,
      "step": 49905
    },
    {
      "entropy": 5.6663806438446045,
      "epoch": 4.527394775036284,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003117596205434583,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22872476130723954,
      "num_tokens": 103615028.0,
      "step": 49910
    },
    {
      "entropy": 5.623172378540039,
      "epoch": 4.5278483309143684,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031172769848900256,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22549275904893876,
      "num_tokens": 103625430.0,
      "step": 49915
    },
    {
      "entropy": 5.70122332572937,
      "epoch": 4.528301886792453,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003116957756750087,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22823155224323272,
      "num_tokens": 103636409.0,
      "step": 49920
    },
    {
      "entropy": 5.663666248321533,
      "epoch": 4.528755442670537,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031166385210213684,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2307484656572342,
      "num_tokens": 103646869.0,
      "step": 49925
    },
    {
      "entropy": 5.594880676269531,
      "epoch": 4.529208998548621,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031163192777104715,
      "loss": 4.7476,
      "mean_token_accuracy": 0.23180926740169525,
      "num_tokens": 103657098.0,
      "step": 49930
    },
    {
      "entropy": 5.684582281112671,
      "epoch": 4.5296625544267055,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003116000026823999,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2240087404847145,
      "num_tokens": 103667331.0,
      "step": 49935
    },
    {
      "entropy": 5.601498985290528,
      "epoch": 4.53011611030479,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003115680768368552,
      "loss": 4.8078,
      "mean_token_accuracy": 0.22878775000572205,
      "num_tokens": 103676755.0,
      "step": 49940
    },
    {
      "entropy": 5.587102651596069,
      "epoch": 4.530569666182874,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003115361502350734,
      "loss": 4.8025,
      "mean_token_accuracy": 0.22458022087812424,
      "num_tokens": 103686334.0,
      "step": 49945
    },
    {
      "entropy": 5.626953935623169,
      "epoch": 4.531023222060957,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003115042228777148,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2226430743932724,
      "num_tokens": 103697127.0,
      "step": 49950
    },
    {
      "entropy": 5.576815462112426,
      "epoch": 4.531476777939043,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031147229476543935,
      "loss": 4.8299,
      "mean_token_accuracy": 0.21763428002595903,
      "num_tokens": 103707860.0,
      "step": 49955
    },
    {
      "entropy": 5.68440580368042,
      "epoch": 4.531930333817126,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031144036589890765,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22012140303850175,
      "num_tokens": 103718816.0,
      "step": 49960
    },
    {
      "entropy": 5.571541881561279,
      "epoch": 4.53238388969521,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031140843627878003,
      "loss": 4.7915,
      "mean_token_accuracy": 0.23026351630687714,
      "num_tokens": 103729203.0,
      "step": 49965
    },
    {
      "entropy": 5.695663595199585,
      "epoch": 4.5328374455732945,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031137650590571655,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22381720095872878,
      "num_tokens": 103740830.0,
      "step": 49970
    },
    {
      "entropy": 5.626502752304077,
      "epoch": 4.533291001451379,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003113445747803776,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2242555558681488,
      "num_tokens": 103752881.0,
      "step": 49975
    },
    {
      "entropy": 5.634502363204956,
      "epoch": 4.533744557329463,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003113126429034237,
      "loss": 4.8858,
      "mean_token_accuracy": 0.2244726300239563,
      "num_tokens": 103762875.0,
      "step": 49980
    },
    {
      "entropy": 5.653916740417481,
      "epoch": 4.534198113207547,
      "grad_norm": 1.375,
      "learning_rate": 0.00031128071027551504,
      "loss": 4.8666,
      "mean_token_accuracy": 0.2260408416390419,
      "num_tokens": 103772147.0,
      "step": 49985
    },
    {
      "entropy": 5.588986253738403,
      "epoch": 4.5346516690856316,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000311248776897312,
      "loss": 4.7996,
      "mean_token_accuracy": 0.22437098175287246,
      "num_tokens": 103782904.0,
      "step": 49990
    },
    {
      "entropy": 5.55466890335083,
      "epoch": 4.535105224963716,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031121684276947504,
      "loss": 4.8401,
      "mean_token_accuracy": 0.22920847684144974,
      "num_tokens": 103793202.0,
      "step": 49995
    },
    {
      "entropy": 5.725629711151123,
      "epoch": 4.5355587808418,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003111849078926646,
      "loss": 5.0622,
      "mean_token_accuracy": 0.21530501395463944,
      "num_tokens": 103805796.0,
      "step": 50000
    },
    {
      "entropy": 5.695513439178467,
      "epoch": 4.5360123367198835,
      "grad_norm": 1.296875,
      "learning_rate": 0.000311152972267541,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2160744145512581,
      "num_tokens": 103816622.0,
      "step": 50005
    },
    {
      "entropy": 5.622814130783081,
      "epoch": 4.536465892597968,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003111210358947647,
      "loss": 4.7889,
      "mean_token_accuracy": 0.23170362114906312,
      "num_tokens": 103826174.0,
      "step": 50010
    },
    {
      "entropy": 5.689163017272949,
      "epoch": 4.536919448476052,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003110890987749962,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22693859934806823,
      "num_tokens": 103836217.0,
      "step": 50015
    },
    {
      "entropy": 5.599464750289917,
      "epoch": 4.537373004354136,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003110571609088959,
      "loss": 4.7556,
      "mean_token_accuracy": 0.23380465507507325,
      "num_tokens": 103846661.0,
      "step": 50020
    },
    {
      "entropy": 5.624052667617798,
      "epoch": 4.5378265602322205,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003110252222971244,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22045110613107682,
      "num_tokens": 103857323.0,
      "step": 50025
    },
    {
      "entropy": 5.679291677474976,
      "epoch": 4.538280116110305,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003109932829403421,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21745751202106475,
      "num_tokens": 103867409.0,
      "step": 50030
    },
    {
      "entropy": 5.694183301925659,
      "epoch": 4.538733671988389,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003109613428392095,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22579509019851685,
      "num_tokens": 103877603.0,
      "step": 50035
    },
    {
      "entropy": 5.635189580917358,
      "epoch": 4.539187227866473,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031092940199438715,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22360528707504274,
      "num_tokens": 103888236.0,
      "step": 50040
    },
    {
      "entropy": 5.614944314956665,
      "epoch": 4.539640783744558,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003108974604065357,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23066894710063934,
      "num_tokens": 103898474.0,
      "step": 50045
    },
    {
      "entropy": 5.59013352394104,
      "epoch": 4.540094339622642,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031086551807631553,
      "loss": 4.8483,
      "mean_token_accuracy": 0.2291017308831215,
      "num_tokens": 103909854.0,
      "step": 50050
    },
    {
      "entropy": 5.614906072616577,
      "epoch": 4.540547895500726,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031083357500438736,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22067913562059402,
      "num_tokens": 103919227.0,
      "step": 50055
    },
    {
      "entropy": 5.596865224838257,
      "epoch": 4.5410014513788095,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031080163119141167,
      "loss": 4.8235,
      "mean_token_accuracy": 0.23617361783981322,
      "num_tokens": 103929022.0,
      "step": 50060
    },
    {
      "entropy": 5.678083848953247,
      "epoch": 4.541455007256894,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031076968663804917,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22406742870807647,
      "num_tokens": 103939441.0,
      "step": 50065
    },
    {
      "entropy": 5.613299608230591,
      "epoch": 4.541908563134978,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031073774134496047,
      "loss": 4.8371,
      "mean_token_accuracy": 0.22294163703918457,
      "num_tokens": 103950312.0,
      "step": 50070
    },
    {
      "entropy": 5.654742097854614,
      "epoch": 4.542362119013062,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003107057953128061,
      "loss": 4.891,
      "mean_token_accuracy": 0.2182607188820839,
      "num_tokens": 103960415.0,
      "step": 50075
    },
    {
      "entropy": 5.590717744827271,
      "epoch": 4.542815674891147,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003106738485422468,
      "loss": 4.8273,
      "mean_token_accuracy": 0.23070331811904907,
      "num_tokens": 103971183.0,
      "step": 50080
    },
    {
      "entropy": 5.689915752410888,
      "epoch": 4.543269230769231,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003106419010339433,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21720528304576875,
      "num_tokens": 103982120.0,
      "step": 50085
    },
    {
      "entropy": 5.618250799179077,
      "epoch": 4.543722786647315,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031060995278855616,
      "loss": 4.7858,
      "mean_token_accuracy": 0.23471929579973222,
      "num_tokens": 103992116.0,
      "step": 50090
    },
    {
      "entropy": 5.595202493667602,
      "epoch": 4.544176342525399,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031057800380674604,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22609275728464126,
      "num_tokens": 104002646.0,
      "step": 50095
    },
    {
      "entropy": 5.652548551559448,
      "epoch": 4.544629898403484,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003105460540891738,
      "loss": 4.9338,
      "mean_token_accuracy": 0.2199845403432846,
      "num_tokens": 104013158.0,
      "step": 50100
    },
    {
      "entropy": 5.617697525024414,
      "epoch": 4.545083454281567,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003105141036365001,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22477666288614273,
      "num_tokens": 104023485.0,
      "step": 50105
    },
    {
      "entropy": 5.638106393814087,
      "epoch": 4.545537010159651,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003104821524493857,
      "loss": 4.8713,
      "mean_token_accuracy": 0.2265024185180664,
      "num_tokens": 104034588.0,
      "step": 50110
    },
    {
      "entropy": 5.650418949127197,
      "epoch": 4.5459905660377355,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003104502005284913,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2245439887046814,
      "num_tokens": 104045081.0,
      "step": 50115
    },
    {
      "entropy": 5.663956785202027,
      "epoch": 4.54644412191582,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003104182478744778,
      "loss": 4.8442,
      "mean_token_accuracy": 0.2238982453942299,
      "num_tokens": 104056166.0,
      "step": 50120
    },
    {
      "entropy": 5.651707506179809,
      "epoch": 4.546897677793904,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003103862944880058,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23217601031064988,
      "num_tokens": 104066699.0,
      "step": 50125
    },
    {
      "entropy": 5.642619228363037,
      "epoch": 4.547351233671988,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003103543403697363,
      "loss": 4.8646,
      "mean_token_accuracy": 0.2291841760277748,
      "num_tokens": 104077061.0,
      "step": 50130
    },
    {
      "entropy": 5.642163801193237,
      "epoch": 4.547804789550073,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003103223855203301,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22244041115045549,
      "num_tokens": 104087877.0,
      "step": 50135
    },
    {
      "entropy": 5.730192852020264,
      "epoch": 4.548258345428157,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00031029042994044784,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21927947402000428,
      "num_tokens": 104099483.0,
      "step": 50140
    },
    {
      "entropy": 5.643180799484253,
      "epoch": 4.548711901306241,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031025847363075056,
      "loss": 4.845,
      "mean_token_accuracy": 0.219778211414814,
      "num_tokens": 104109530.0,
      "step": 50145
    },
    {
      "entropy": 5.63972659111023,
      "epoch": 4.549165457184325,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031022651659189907,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22663553506135942,
      "num_tokens": 104119610.0,
      "step": 50150
    },
    {
      "entropy": 5.666290330886841,
      "epoch": 4.54961901306241,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003101945588245541,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22288836985826493,
      "num_tokens": 104130337.0,
      "step": 50155
    },
    {
      "entropy": 5.710297584533691,
      "epoch": 4.550072568940493,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003101626003293768,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22102049738168716,
      "num_tokens": 104139462.0,
      "step": 50160
    },
    {
      "entropy": 5.6157739639282225,
      "epoch": 4.550526124818577,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003101306411070279,
      "loss": 4.7781,
      "mean_token_accuracy": 0.23931246846914292,
      "num_tokens": 104148640.0,
      "step": 50165
    },
    {
      "entropy": 5.6210558891296385,
      "epoch": 4.550979680696662,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031009868115816847,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22993152290582658,
      "num_tokens": 104159400.0,
      "step": 50170
    },
    {
      "entropy": 5.587764930725098,
      "epoch": 4.551433236574746,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031006672048345936,
      "loss": 4.835,
      "mean_token_accuracy": 0.22996535450220107,
      "num_tokens": 104170149.0,
      "step": 50175
    },
    {
      "entropy": 5.5866437435150145,
      "epoch": 4.55188679245283,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003100347590835615,
      "loss": 4.8109,
      "mean_token_accuracy": 0.23505755215883256,
      "num_tokens": 104180787.0,
      "step": 50180
    },
    {
      "entropy": 5.697358322143555,
      "epoch": 4.552340348330914,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003100027969591359,
      "loss": 4.893,
      "mean_token_accuracy": 0.22432547211647033,
      "num_tokens": 104190961.0,
      "step": 50185
    },
    {
      "entropy": 5.6299303531646725,
      "epoch": 4.552793904208999,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030997083411084356,
      "loss": 4.83,
      "mean_token_accuracy": 0.2322741612792015,
      "num_tokens": 104201886.0,
      "step": 50190
    },
    {
      "entropy": 5.67264347076416,
      "epoch": 4.553247460087083,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003099388705393454,
      "loss": 4.8649,
      "mean_token_accuracy": 0.23176738023757934,
      "num_tokens": 104212834.0,
      "step": 50195
    },
    {
      "entropy": 5.656153774261474,
      "epoch": 4.553701015965167,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003099069062453025,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22573157548904418,
      "num_tokens": 104222384.0,
      "step": 50200
    },
    {
      "entropy": 5.604507780075073,
      "epoch": 4.5541545718432515,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030987494122937604,
      "loss": 4.8342,
      "mean_token_accuracy": 0.23019215017557143,
      "num_tokens": 104232971.0,
      "step": 50205
    },
    {
      "entropy": 5.640227365493774,
      "epoch": 4.554608127721336,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003098429754922267,
      "loss": 4.7692,
      "mean_token_accuracy": 0.22722506225109101,
      "num_tokens": 104245239.0,
      "step": 50210
    },
    {
      "entropy": 5.655685997009277,
      "epoch": 4.555061683599419,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030981100903451587,
      "loss": 4.893,
      "mean_token_accuracy": 0.22190751731395722,
      "num_tokens": 104256323.0,
      "step": 50215
    },
    {
      "entropy": 5.65981388092041,
      "epoch": 4.555515239477503,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003097790418569045,
      "loss": 4.88,
      "mean_token_accuracy": 0.22497054189443588,
      "num_tokens": 104267106.0,
      "step": 50220
    },
    {
      "entropy": 5.637773275375366,
      "epoch": 4.555968795355588,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003097470739600536,
      "loss": 4.8742,
      "mean_token_accuracy": 0.2281886577606201,
      "num_tokens": 104277336.0,
      "step": 50225
    },
    {
      "entropy": 5.629691505432129,
      "epoch": 4.556422351233672,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003097151053446244,
      "loss": 4.8452,
      "mean_token_accuracy": 0.232522414624691,
      "num_tokens": 104288809.0,
      "step": 50230
    },
    {
      "entropy": 5.621202754974365,
      "epoch": 4.556875907111756,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003096831360112781,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22104736417531967,
      "num_tokens": 104299434.0,
      "step": 50235
    },
    {
      "entropy": 5.586041593551636,
      "epoch": 4.55732946298984,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003096511659606756,
      "loss": 4.8075,
      "mean_token_accuracy": 0.22792518883943558,
      "num_tokens": 104309426.0,
      "step": 50240
    },
    {
      "entropy": 5.567373275756836,
      "epoch": 4.557783018867925,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003096191951934781,
      "loss": 4.8113,
      "mean_token_accuracy": 0.230711629986763,
      "num_tokens": 104319530.0,
      "step": 50245
    },
    {
      "entropy": 5.598849821090698,
      "epoch": 4.558236574746009,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003095872237103468,
      "loss": 4.7843,
      "mean_token_accuracy": 0.23187736570835113,
      "num_tokens": 104329198.0,
      "step": 50250
    },
    {
      "entropy": 5.706061220169067,
      "epoch": 4.558690130624093,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030955525151194304,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22357378751039506,
      "num_tokens": 104340481.0,
      "step": 50255
    },
    {
      "entropy": 5.689530372619629,
      "epoch": 4.559143686502177,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003095232785989277,
      "loss": 4.8532,
      "mean_token_accuracy": 0.2265137791633606,
      "num_tokens": 104350382.0,
      "step": 50260
    },
    {
      "entropy": 5.632962656021118,
      "epoch": 4.559597242380261,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003094913049719623,
      "loss": 4.7713,
      "mean_token_accuracy": 0.23010687679052352,
      "num_tokens": 104360967.0,
      "step": 50265
    },
    {
      "entropy": 5.639765930175781,
      "epoch": 4.560050798258345,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030945933063170793,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22832264602184296,
      "num_tokens": 104371283.0,
      "step": 50270
    },
    {
      "entropy": 5.644532346725464,
      "epoch": 4.560504354136429,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030942735557882575,
      "loss": 4.8416,
      "mean_token_accuracy": 0.23075293451547624,
      "num_tokens": 104381545.0,
      "step": 50275
    },
    {
      "entropy": 5.6883077144622805,
      "epoch": 4.560957910014514,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030939537981397716,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21853375434875488,
      "num_tokens": 104392533.0,
      "step": 50280
    },
    {
      "entropy": 5.658603715896606,
      "epoch": 4.561411465892598,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00030936340333782326,
      "loss": 4.9104,
      "mean_token_accuracy": 0.2233912840485573,
      "num_tokens": 104402230.0,
      "step": 50285
    },
    {
      "entropy": 5.678667974472046,
      "epoch": 4.561865021770682,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003093314261510254,
      "loss": 4.8569,
      "mean_token_accuracy": 0.22987056970596315,
      "num_tokens": 104412029.0,
      "step": 50290
    },
    {
      "entropy": 5.613131093978882,
      "epoch": 4.5623185776487665,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000309299448254245,
      "loss": 4.8344,
      "mean_token_accuracy": 0.2229466825723648,
      "num_tokens": 104422107.0,
      "step": 50295
    },
    {
      "entropy": 5.6594932079315186,
      "epoch": 4.562772133526851,
      "grad_norm": 1.46875,
      "learning_rate": 0.00030926746964814324,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22183825224637985,
      "num_tokens": 104432073.0,
      "step": 50300
    },
    {
      "entropy": 5.62965202331543,
      "epoch": 4.563225689404935,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030923549033338147,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22795804291963578,
      "num_tokens": 104441998.0,
      "step": 50305
    },
    {
      "entropy": 5.541490364074707,
      "epoch": 4.563679245283019,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000309203510310621,
      "loss": 4.7359,
      "mean_token_accuracy": 0.23800933063030244,
      "num_tokens": 104451474.0,
      "step": 50310
    },
    {
      "entropy": 5.596472311019897,
      "epoch": 4.564132801161103,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003091715295805233,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23936792016029357,
      "num_tokens": 104461351.0,
      "step": 50315
    },
    {
      "entropy": 5.6548563003540036,
      "epoch": 4.564586357039187,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003091395481437496,
      "loss": 4.8555,
      "mean_token_accuracy": 0.2317029356956482,
      "num_tokens": 104471013.0,
      "step": 50320
    },
    {
      "entropy": 5.650713777542114,
      "epoch": 4.565039912917271,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003091075660009614,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22257944345474243,
      "num_tokens": 104481863.0,
      "step": 50325
    },
    {
      "entropy": 5.6875556945800785,
      "epoch": 4.5654934687953554,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030907558315282,
      "loss": 4.8692,
      "mean_token_accuracy": 0.2235629677772522,
      "num_tokens": 104492718.0,
      "step": 50330
    },
    {
      "entropy": 5.656477260589599,
      "epoch": 4.56594702467344,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030904359959998685,
      "loss": 4.8381,
      "mean_token_accuracy": 0.2265316739678383,
      "num_tokens": 104502837.0,
      "step": 50335
    },
    {
      "entropy": 5.655040979385376,
      "epoch": 4.566400580551524,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003090116153431234,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2287306621670723,
      "num_tokens": 104514202.0,
      "step": 50340
    },
    {
      "entropy": 5.6685301780700685,
      "epoch": 4.566854136429608,
      "grad_norm": 1.46875,
      "learning_rate": 0.000308979630382891,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22518843412399292,
      "num_tokens": 104525037.0,
      "step": 50345
    },
    {
      "entropy": 5.594941473007202,
      "epoch": 4.5673076923076925,
      "grad_norm": 1.375,
      "learning_rate": 0.00030894764471995137,
      "loss": 4.8136,
      "mean_token_accuracy": 0.2302778795361519,
      "num_tokens": 104534883.0,
      "step": 50350
    },
    {
      "entropy": 5.61194429397583,
      "epoch": 4.567761248185777,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003089156583549657,
      "loss": 4.8525,
      "mean_token_accuracy": 0.22601610124111177,
      "num_tokens": 104544845.0,
      "step": 50355
    },
    {
      "entropy": 5.673996114730835,
      "epoch": 4.568214804063861,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003088836712885955,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22464786171913148,
      "num_tokens": 104555654.0,
      "step": 50360
    },
    {
      "entropy": 5.622860527038574,
      "epoch": 4.568668359941945,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030885168352150243,
      "loss": 4.894,
      "mean_token_accuracy": 0.21621053516864777,
      "num_tokens": 104565015.0,
      "step": 50365
    },
    {
      "entropy": 5.542737817764282,
      "epoch": 4.569121915820029,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030881969505434795,
      "loss": 4.8087,
      "mean_token_accuracy": 0.23138071596622467,
      "num_tokens": 104575934.0,
      "step": 50370
    },
    {
      "entropy": 5.5371763706207275,
      "epoch": 4.569575471698113,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030878770588779335,
      "loss": 4.7698,
      "mean_token_accuracy": 0.2305763065814972,
      "num_tokens": 104585915.0,
      "step": 50375
    },
    {
      "entropy": 5.687968730926514,
      "epoch": 4.570029027576197,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030875571602250064,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2215704843401909,
      "num_tokens": 104595605.0,
      "step": 50380
    },
    {
      "entropy": 5.669393491744995,
      "epoch": 4.5704825834542815,
      "grad_norm": 1.203125,
      "learning_rate": 0.000308723725459131,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21701785624027253,
      "num_tokens": 104608251.0,
      "step": 50385
    },
    {
      "entropy": 5.697407293319702,
      "epoch": 4.570936139332366,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003086917341983461,
      "loss": 4.8366,
      "mean_token_accuracy": 0.23312151581048965,
      "num_tokens": 104618586.0,
      "step": 50390
    },
    {
      "entropy": 5.632842588424682,
      "epoch": 4.57138969521045,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003086597422408076,
      "loss": 4.858,
      "mean_token_accuracy": 0.23026376962661743,
      "num_tokens": 104629718.0,
      "step": 50395
    },
    {
      "entropy": 5.607403087615967,
      "epoch": 4.571843251088534,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000308627749587177,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22479713708162308,
      "num_tokens": 104639923.0,
      "step": 50400
    },
    {
      "entropy": 5.643717241287232,
      "epoch": 4.5722968069666186,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000308595756238116,
      "loss": 4.876,
      "mean_token_accuracy": 0.22416977286338807,
      "num_tokens": 104650890.0,
      "step": 50405
    },
    {
      "entropy": 5.618989181518555,
      "epoch": 4.572750362844703,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003085637621942862,
      "loss": 4.8349,
      "mean_token_accuracy": 0.2296306625008583,
      "num_tokens": 104661231.0,
      "step": 50410
    },
    {
      "entropy": 5.599440956115723,
      "epoch": 4.573203918722786,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030853176745634925,
      "loss": 4.798,
      "mean_token_accuracy": 0.22579531222581864,
      "num_tokens": 104671641.0,
      "step": 50415
    },
    {
      "entropy": 5.6189836978912355,
      "epoch": 4.5736574746008705,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003084997720249668,
      "loss": 4.8426,
      "mean_token_accuracy": 0.22637235671281813,
      "num_tokens": 104682310.0,
      "step": 50420
    },
    {
      "entropy": 5.694850969314575,
      "epoch": 4.574111030478955,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030846777590080057,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2208185538649559,
      "num_tokens": 104692414.0,
      "step": 50425
    },
    {
      "entropy": 5.628934669494629,
      "epoch": 4.574564586357039,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030843577908451215,
      "loss": 4.9161,
      "mean_token_accuracy": 0.215856671333313,
      "num_tokens": 104703617.0,
      "step": 50430
    },
    {
      "entropy": 5.578206396102905,
      "epoch": 4.575018142235123,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003084037815767633,
      "loss": 4.8448,
      "mean_token_accuracy": 0.22974322140216827,
      "num_tokens": 104713494.0,
      "step": 50435
    },
    {
      "entropy": 5.628090286254883,
      "epoch": 4.5754716981132075,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003083717833782158,
      "loss": 4.8264,
      "mean_token_accuracy": 0.2248722121119499,
      "num_tokens": 104723349.0,
      "step": 50440
    },
    {
      "entropy": 5.599398183822632,
      "epoch": 4.575925253991292,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003083397844895312,
      "loss": 4.7865,
      "mean_token_accuracy": 0.24185170233249664,
      "num_tokens": 104732903.0,
      "step": 50445
    },
    {
      "entropy": 5.559778690338135,
      "epoch": 4.576378809869376,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003083077849113715,
      "loss": 4.8229,
      "mean_token_accuracy": 0.22757029086351394,
      "num_tokens": 104742533.0,
      "step": 50450
    },
    {
      "entropy": 5.561368942260742,
      "epoch": 4.57683236574746,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030827578464439823,
      "loss": 4.7459,
      "mean_token_accuracy": 0.2268357753753662,
      "num_tokens": 104752161.0,
      "step": 50455
    },
    {
      "entropy": 5.586967658996582,
      "epoch": 4.577285921625545,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030824378368927324,
      "loss": 4.8154,
      "mean_token_accuracy": 0.23313124030828475,
      "num_tokens": 104762630.0,
      "step": 50460
    },
    {
      "entropy": 5.586753797531128,
      "epoch": 4.577739477503629,
      "grad_norm": 1.375,
      "learning_rate": 0.00030821178204665836,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23075274378061295,
      "num_tokens": 104773529.0,
      "step": 50465
    },
    {
      "entropy": 5.589892578125,
      "epoch": 4.578193033381712,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003081797797172154,
      "loss": 4.7911,
      "mean_token_accuracy": 0.22131619453430176,
      "num_tokens": 104783386.0,
      "step": 50470
    },
    {
      "entropy": 5.650656270980835,
      "epoch": 4.5786465892597965,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003081477767016061,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2253374382853508,
      "num_tokens": 104793254.0,
      "step": 50475
    },
    {
      "entropy": 5.618346500396728,
      "epoch": 4.579100145137881,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030811577300049226,
      "loss": 4.881,
      "mean_token_accuracy": 0.2295272544026375,
      "num_tokens": 104804345.0,
      "step": 50480
    },
    {
      "entropy": 5.623326349258423,
      "epoch": 4.579553701015965,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003080837686145359,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2219862625002861,
      "num_tokens": 104814891.0,
      "step": 50485
    },
    {
      "entropy": 5.59203577041626,
      "epoch": 4.580007256894049,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003080517635443987,
      "loss": 4.8135,
      "mean_token_accuracy": 0.22337402701377868,
      "num_tokens": 104825015.0,
      "step": 50490
    },
    {
      "entropy": 5.640541505813599,
      "epoch": 4.580460812772134,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003080197577907427,
      "loss": 4.8804,
      "mean_token_accuracy": 0.21845125257968903,
      "num_tokens": 104835735.0,
      "step": 50495
    },
    {
      "entropy": 5.6017937660217285,
      "epoch": 4.580914368650218,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030798775135422966,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22304065078496932,
      "num_tokens": 104847484.0,
      "step": 50500
    },
    {
      "entropy": 5.621297693252563,
      "epoch": 4.581367924528302,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003079557442355215,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23170382380485535,
      "num_tokens": 104857783.0,
      "step": 50505
    },
    {
      "entropy": 5.580574655532837,
      "epoch": 4.581821480406386,
      "grad_norm": 1.375,
      "learning_rate": 0.0003079237364352801,
      "loss": 4.7989,
      "mean_token_accuracy": 0.24066220968961716,
      "num_tokens": 104867498.0,
      "step": 50510
    },
    {
      "entropy": 5.662361574172974,
      "epoch": 4.58227503628447,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003078917279541675,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22711084485054017,
      "num_tokens": 104876505.0,
      "step": 50515
    },
    {
      "entropy": 5.649198818206787,
      "epoch": 4.582728592162555,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003078597187928456,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21648503839969635,
      "num_tokens": 104886932.0,
      "step": 50520
    },
    {
      "entropy": 5.677640056610107,
      "epoch": 4.583182148040638,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003078277089519764,
      "loss": 4.8453,
      "mean_token_accuracy": 0.23290291279554368,
      "num_tokens": 104896733.0,
      "step": 50525
    },
    {
      "entropy": 5.649524068832397,
      "epoch": 4.5836357039187225,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030779569843222165,
      "loss": 4.9014,
      "mean_token_accuracy": 0.21404143124818803,
      "num_tokens": 104906846.0,
      "step": 50530
    },
    {
      "entropy": 5.53101544380188,
      "epoch": 4.584089259796807,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030776368723424365,
      "loss": 4.6734,
      "mean_token_accuracy": 0.24461723864078522,
      "num_tokens": 104916136.0,
      "step": 50535
    },
    {
      "entropy": 5.6068034172058105,
      "epoch": 4.584542815674891,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003077316753587042,
      "loss": 4.819,
      "mean_token_accuracy": 0.2325960412621498,
      "num_tokens": 104926145.0,
      "step": 50540
    },
    {
      "entropy": 5.660551738739014,
      "epoch": 4.584996371552975,
      "grad_norm": 1.46875,
      "learning_rate": 0.00030769966280626526,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21992480754852295,
      "num_tokens": 104935576.0,
      "step": 50545
    },
    {
      "entropy": 5.595847082138062,
      "epoch": 4.58544992743106,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030766764957758906,
      "loss": 4.7507,
      "mean_token_accuracy": 0.23778180181980133,
      "num_tokens": 104944846.0,
      "step": 50550
    },
    {
      "entropy": 5.5945477962493895,
      "epoch": 4.585903483309144,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030763563567333754,
      "loss": 4.7951,
      "mean_token_accuracy": 0.23055403381586076,
      "num_tokens": 104954604.0,
      "step": 50555
    },
    {
      "entropy": 5.601663875579834,
      "epoch": 4.586357039187228,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030760362109417277,
      "loss": 4.8355,
      "mean_token_accuracy": 0.22555627822875976,
      "num_tokens": 104964858.0,
      "step": 50560
    },
    {
      "entropy": 5.664407968521118,
      "epoch": 4.586810595065312,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003075716058407568,
      "loss": 4.8979,
      "mean_token_accuracy": 0.219611294567585,
      "num_tokens": 104974997.0,
      "step": 50565
    },
    {
      "entropy": 5.619162082672119,
      "epoch": 4.587264150943396,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030753958991375176,
      "loss": 4.8066,
      "mean_token_accuracy": 0.2315882548689842,
      "num_tokens": 104984585.0,
      "step": 50570
    },
    {
      "entropy": 5.599903774261475,
      "epoch": 4.58771770682148,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030750757331381964,
      "loss": 4.8143,
      "mean_token_accuracy": 0.2357783392071724,
      "num_tokens": 104994843.0,
      "step": 50575
    },
    {
      "entropy": 5.598966503143311,
      "epoch": 4.588171262699564,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003074755560416226,
      "loss": 4.884,
      "mean_token_accuracy": 0.22850777506828307,
      "num_tokens": 105006082.0,
      "step": 50580
    },
    {
      "entropy": 5.606411933898926,
      "epoch": 4.588624818577649,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030744353809782283,
      "loss": 4.8404,
      "mean_token_accuracy": 0.22787705063819885,
      "num_tokens": 105016704.0,
      "step": 50585
    },
    {
      "entropy": 5.6820660591125485,
      "epoch": 4.589078374455733,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030741151948308237,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22554327845573424,
      "num_tokens": 105027810.0,
      "step": 50590
    },
    {
      "entropy": 5.599266195297242,
      "epoch": 4.589531930333817,
      "grad_norm": 1.375,
      "learning_rate": 0.0003073795001980635,
      "loss": 4.8174,
      "mean_token_accuracy": 0.23282620906829835,
      "num_tokens": 105037488.0,
      "step": 50595
    },
    {
      "entropy": 5.639221477508545,
      "epoch": 4.589985486211901,
      "grad_norm": 1.5,
      "learning_rate": 0.0003073474802434282,
      "loss": 4.8748,
      "mean_token_accuracy": 0.2260706529021263,
      "num_tokens": 105047461.0,
      "step": 50600
    },
    {
      "entropy": 5.656062459945678,
      "epoch": 4.590439042089986,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000307315459619839,
      "loss": 4.8886,
      "mean_token_accuracy": 0.2277732253074646,
      "num_tokens": 105058020.0,
      "step": 50605
    },
    {
      "entropy": 5.729408550262451,
      "epoch": 4.59089259796807,
      "grad_norm": 1.375,
      "learning_rate": 0.0003072834383279576,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2208922639489174,
      "num_tokens": 105069638.0,
      "step": 50610
    },
    {
      "entropy": 5.664794111251831,
      "epoch": 4.591346153846154,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003072514163684466,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22586833834648132,
      "num_tokens": 105078686.0,
      "step": 50615
    },
    {
      "entropy": 5.637793970108032,
      "epoch": 4.5917997097242385,
      "grad_norm": 1.25,
      "learning_rate": 0.0003072193937419681,
      "loss": 4.851,
      "mean_token_accuracy": 0.22906333208084106,
      "num_tokens": 105089739.0,
      "step": 50620
    },
    {
      "entropy": 5.62109146118164,
      "epoch": 4.592253265602322,
      "grad_norm": 1.25,
      "learning_rate": 0.00030718737044918435,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22694653570652007,
      "num_tokens": 105100673.0,
      "step": 50625
    },
    {
      "entropy": 5.621618318557739,
      "epoch": 4.592706821480406,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030715534649075745,
      "loss": 4.8432,
      "mean_token_accuracy": 0.22862128168344498,
      "num_tokens": 105110750.0,
      "step": 50630
    },
    {
      "entropy": 5.562719249725342,
      "epoch": 4.59316037735849,
      "grad_norm": 1.375,
      "learning_rate": 0.00030712332186734985,
      "loss": 4.7615,
      "mean_token_accuracy": 0.2326630249619484,
      "num_tokens": 105120809.0,
      "step": 50635
    },
    {
      "entropy": 5.637806558609009,
      "epoch": 4.593613933236575,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003070912965796238,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22420564144849778,
      "num_tokens": 105131106.0,
      "step": 50640
    },
    {
      "entropy": 5.66543025970459,
      "epoch": 4.594067489114659,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030705927062824154,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22100044339895247,
      "num_tokens": 105140970.0,
      "step": 50645
    },
    {
      "entropy": 5.667979335784912,
      "epoch": 4.594521044992743,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003070272440138654,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22011010050773622,
      "num_tokens": 105152429.0,
      "step": 50650
    },
    {
      "entropy": 5.692554950714111,
      "epoch": 4.594974600870827,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003069952167371577,
      "loss": 4.931,
      "mean_token_accuracy": 0.22218783050775529,
      "num_tokens": 105163534.0,
      "step": 50655
    },
    {
      "entropy": 5.590262937545776,
      "epoch": 4.595428156748912,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030696318879878064,
      "loss": 4.782,
      "mean_token_accuracy": 0.2283254384994507,
      "num_tokens": 105172798.0,
      "step": 50660
    },
    {
      "entropy": 5.657283449172974,
      "epoch": 4.595881712626996,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030693116019939686,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22765248119831086,
      "num_tokens": 105182576.0,
      "step": 50665
    },
    {
      "entropy": 5.653711795806885,
      "epoch": 4.596335268505079,
      "grad_norm": 1.59375,
      "learning_rate": 0.00030689913093966847,
      "loss": 4.8666,
      "mean_token_accuracy": 0.23224549293518065,
      "num_tokens": 105192070.0,
      "step": 50670
    },
    {
      "entropy": 5.656721878051758,
      "epoch": 4.5967888243831645,
      "grad_norm": 1.296875,
      "learning_rate": 0.000306867101020258,
      "loss": 4.8532,
      "mean_token_accuracy": 0.23058617860078812,
      "num_tokens": 105203151.0,
      "step": 50675
    },
    {
      "entropy": 5.658195447921753,
      "epoch": 4.597242380261248,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003068350704418277,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2186349704861641,
      "num_tokens": 105213217.0,
      "step": 50680
    },
    {
      "entropy": 5.5655436515808105,
      "epoch": 4.597695936139332,
      "grad_norm": 1.375,
      "learning_rate": 0.00030680303920504006,
      "loss": 4.7665,
      "mean_token_accuracy": 0.23543991148471832,
      "num_tokens": 105223450.0,
      "step": 50685
    },
    {
      "entropy": 5.671842050552368,
      "epoch": 4.598149492017416,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003067710073105574,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2234810084104538,
      "num_tokens": 105234705.0,
      "step": 50690
    },
    {
      "entropy": 5.7554931640625,
      "epoch": 4.598603047895501,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00030673897475904223,
      "loss": 5.0656,
      "mean_token_accuracy": 0.20581818521022796,
      "num_tokens": 105245042.0,
      "step": 50695
    },
    {
      "entropy": 5.695953893661499,
      "epoch": 4.599056603773585,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030670694155115704,
      "loss": 4.8653,
      "mean_token_accuracy": 0.21814588457345963,
      "num_tokens": 105255830.0,
      "step": 50700
    },
    {
      "entropy": 5.627609300613403,
      "epoch": 4.599510159651669,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030667490768756416,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22714640945196152,
      "num_tokens": 105265844.0,
      "step": 50705
    },
    {
      "entropy": 5.583739233016968,
      "epoch": 4.5999637155297535,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003066428731689262,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2303543657064438,
      "num_tokens": 105276846.0,
      "step": 50710
    },
    {
      "entropy": 5.686843967437744,
      "epoch": 4.600417271407838,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030661083799590544,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22183436304330825,
      "num_tokens": 105286740.0,
      "step": 50715
    },
    {
      "entropy": 5.616544103622436,
      "epoch": 4.600870827285922,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030657880216916454,
      "loss": 4.8342,
      "mean_token_accuracy": 0.2319315105676651,
      "num_tokens": 105296710.0,
      "step": 50720
    },
    {
      "entropy": 5.65447359085083,
      "epoch": 4.601324383164005,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000306546765689366,
      "loss": 4.8954,
      "mean_token_accuracy": 0.2313593491911888,
      "num_tokens": 105307076.0,
      "step": 50725
    },
    {
      "entropy": 5.641586399078369,
      "epoch": 4.60177793904209,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003065147285571722,
      "loss": 4.8199,
      "mean_token_accuracy": 0.2286187544465065,
      "num_tokens": 105317486.0,
      "step": 50730
    },
    {
      "entropy": 5.608083820343017,
      "epoch": 4.602231494920174,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003064826907732459,
      "loss": 4.8444,
      "mean_token_accuracy": 0.22638851255178452,
      "num_tokens": 105327771.0,
      "step": 50735
    },
    {
      "entropy": 5.630862045288086,
      "epoch": 4.602685050798258,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030645065233824955,
      "loss": 4.8947,
      "mean_token_accuracy": 0.21975306868553163,
      "num_tokens": 105339191.0,
      "step": 50740
    },
    {
      "entropy": 5.657925939559936,
      "epoch": 4.6031386066763424,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030641861325284557,
      "loss": 4.8748,
      "mean_token_accuracy": 0.2318240746855736,
      "num_tokens": 105349949.0,
      "step": 50745
    },
    {
      "entropy": 5.584345293045044,
      "epoch": 4.603592162554427,
      "grad_norm": 1.25,
      "learning_rate": 0.00030638657351769673,
      "loss": 4.8463,
      "mean_token_accuracy": 0.21807619482278823,
      "num_tokens": 105360357.0,
      "step": 50750
    },
    {
      "entropy": 5.709790229797363,
      "epoch": 4.604045718432511,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030635453313346554,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22389622330665587,
      "num_tokens": 105370765.0,
      "step": 50755
    },
    {
      "entropy": 5.690171098709106,
      "epoch": 4.604499274310595,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003063224921008146,
      "loss": 4.8821,
      "mean_token_accuracy": 0.2295605272054672,
      "num_tokens": 105380797.0,
      "step": 50760
    },
    {
      "entropy": 5.661124229431152,
      "epoch": 4.6049528301886795,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003062904504204066,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22322365939617156,
      "num_tokens": 105390653.0,
      "step": 50765
    },
    {
      "entropy": 5.747948074340821,
      "epoch": 4.605406386066764,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030625840809290404,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22046414017677307,
      "num_tokens": 105401308.0,
      "step": 50770
    },
    {
      "entropy": 5.618389749526978,
      "epoch": 4.605859941944848,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030622636511896977,
      "loss": 4.8647,
      "mean_token_accuracy": 0.2199568271636963,
      "num_tokens": 105412412.0,
      "step": 50775
    },
    {
      "entropy": 5.686307191848755,
      "epoch": 4.606313497822931,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003061943214992662,
      "loss": 4.9366,
      "mean_token_accuracy": 0.216050323843956,
      "num_tokens": 105424789.0,
      "step": 50780
    },
    {
      "entropy": 5.640134382247925,
      "epoch": 4.606767053701016,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003061622772344561,
      "loss": 4.9423,
      "mean_token_accuracy": 0.22565828561782836,
      "num_tokens": 105434380.0,
      "step": 50785
    },
    {
      "entropy": 5.647423553466797,
      "epoch": 4.6072206095791,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003061302323252022,
      "loss": 4.8343,
      "mean_token_accuracy": 0.2336523249745369,
      "num_tokens": 105443889.0,
      "step": 50790
    },
    {
      "entropy": 5.681927680969238,
      "epoch": 4.607674165457184,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003060981867721671,
      "loss": 4.8466,
      "mean_token_accuracy": 0.22723437249660491,
      "num_tokens": 105454336.0,
      "step": 50795
    },
    {
      "entropy": 5.714066362380981,
      "epoch": 4.6081277213352685,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030606614057601374,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21961478143930435,
      "num_tokens": 105464601.0,
      "step": 50800
    },
    {
      "entropy": 5.698313045501709,
      "epoch": 4.608581277213353,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030603409373740456,
      "loss": 4.8885,
      "mean_token_accuracy": 0.21760042011737823,
      "num_tokens": 105475265.0,
      "step": 50805
    },
    {
      "entropy": 5.569851303100586,
      "epoch": 4.609034833091437,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030600204625700234,
      "loss": 4.7688,
      "mean_token_accuracy": 0.23256535083055496,
      "num_tokens": 105485491.0,
      "step": 50810
    },
    {
      "entropy": 5.617128133773804,
      "epoch": 4.609488388969521,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030596999813547007,
      "loss": 4.8314,
      "mean_token_accuracy": 0.2247454509139061,
      "num_tokens": 105495859.0,
      "step": 50815
    },
    {
      "entropy": 5.57914924621582,
      "epoch": 4.6099419448476056,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003059379493734702,
      "loss": 4.7879,
      "mean_token_accuracy": 0.23186047971248627,
      "num_tokens": 105506385.0,
      "step": 50820
    },
    {
      "entropy": 5.629433870315552,
      "epoch": 4.610395500725689,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030590589997166575,
      "loss": 4.7819,
      "mean_token_accuracy": 0.23234861940145493,
      "num_tokens": 105515829.0,
      "step": 50825
    },
    {
      "entropy": 5.612813234329224,
      "epoch": 4.610849056603773,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003058738499307193,
      "loss": 4.7791,
      "mean_token_accuracy": 0.24024191349744797,
      "num_tokens": 105525950.0,
      "step": 50830
    },
    {
      "entropy": 5.621317148208618,
      "epoch": 4.6113026124818575,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030584179925129384,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22495245486497878,
      "num_tokens": 105536536.0,
      "step": 50835
    },
    {
      "entropy": 5.555648565292358,
      "epoch": 4.611756168359942,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030580974793405213,
      "loss": 4.7801,
      "mean_token_accuracy": 0.23644742518663406,
      "num_tokens": 105546069.0,
      "step": 50840
    },
    {
      "entropy": 5.582544040679932,
      "epoch": 4.612209724238026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030577769597965696,
      "loss": 4.8426,
      "mean_token_accuracy": 0.226922869682312,
      "num_tokens": 105556522.0,
      "step": 50845
    },
    {
      "entropy": 5.595562171936035,
      "epoch": 4.61266328011611,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003057456433887712,
      "loss": 4.7835,
      "mean_token_accuracy": 0.2273197665810585,
      "num_tokens": 105566431.0,
      "step": 50850
    },
    {
      "entropy": 5.600449514389038,
      "epoch": 4.6131168359941945,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030571359016205767,
      "loss": 4.812,
      "mean_token_accuracy": 0.2331511363387108,
      "num_tokens": 105576270.0,
      "step": 50855
    },
    {
      "entropy": 5.608708238601684,
      "epoch": 4.613570391872279,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003056815363001793,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22699575275182723,
      "num_tokens": 105587144.0,
      "step": 50860
    },
    {
      "entropy": 5.59222993850708,
      "epoch": 4.614023947750363,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030564948180379897,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22785727977752684,
      "num_tokens": 105596636.0,
      "step": 50865
    },
    {
      "entropy": 5.638585472106934,
      "epoch": 4.614477503628447,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003056174266735795,
      "loss": 4.8591,
      "mean_token_accuracy": 0.2252470225095749,
      "num_tokens": 105605825.0,
      "step": 50870
    },
    {
      "entropy": 5.735712242126465,
      "epoch": 4.614931059506532,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003055853709101839,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22258786410093306,
      "num_tokens": 105615990.0,
      "step": 50875
    },
    {
      "entropy": 5.643759775161743,
      "epoch": 4.615384615384615,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030555331451427494,
      "loss": 4.8783,
      "mean_token_accuracy": 0.22072271555662154,
      "num_tokens": 105626673.0,
      "step": 50880
    },
    {
      "entropy": 5.653816652297974,
      "epoch": 4.615838171262699,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030552125748651573,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21463381201028825,
      "num_tokens": 105636752.0,
      "step": 50885
    },
    {
      "entropy": 5.709087991714478,
      "epoch": 4.6162917271407835,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030548919982756914,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21918269097805024,
      "num_tokens": 105647476.0,
      "step": 50890
    },
    {
      "entropy": 5.666573143005371,
      "epoch": 4.616745283018868,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030545714153809805,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22858808785676957,
      "num_tokens": 105658135.0,
      "step": 50895
    },
    {
      "entropy": 5.595974922180176,
      "epoch": 4.617198838896952,
      "grad_norm": 1.375,
      "learning_rate": 0.0003054250826187655,
      "loss": 4.7757,
      "mean_token_accuracy": 0.23189259320497513,
      "num_tokens": 105667802.0,
      "step": 50900
    },
    {
      "entropy": 5.671021270751953,
      "epoch": 4.617652394775036,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003053930230702346,
      "loss": 4.9068,
      "mean_token_accuracy": 0.228813798725605,
      "num_tokens": 105679161.0,
      "step": 50905
    },
    {
      "entropy": 5.656111526489258,
      "epoch": 4.618105950653121,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003053609628931682,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21804093569517136,
      "num_tokens": 105688093.0,
      "step": 50910
    },
    {
      "entropy": 5.678687047958374,
      "epoch": 4.618559506531205,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003053289020882293,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22115679532289506,
      "num_tokens": 105697982.0,
      "step": 50915
    },
    {
      "entropy": 5.667180919647217,
      "epoch": 4.619013062409289,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030529684065608103,
      "loss": 4.7907,
      "mean_token_accuracy": 0.230059252679348,
      "num_tokens": 105708729.0,
      "step": 50920
    },
    {
      "entropy": 5.581840705871582,
      "epoch": 4.619466618287373,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030526477859738635,
      "loss": 4.7864,
      "mean_token_accuracy": 0.22572413980960845,
      "num_tokens": 105718919.0,
      "step": 50925
    },
    {
      "entropy": 5.60128345489502,
      "epoch": 4.619920174165458,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003052327159128083,
      "loss": 4.7875,
      "mean_token_accuracy": 0.23378977328538894,
      "num_tokens": 105727787.0,
      "step": 50930
    },
    {
      "entropy": 5.555875396728515,
      "epoch": 4.620373730043541,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030520065260301003,
      "loss": 4.8092,
      "mean_token_accuracy": 0.22311647236347198,
      "num_tokens": 105738282.0,
      "step": 50935
    },
    {
      "entropy": 5.624907159805298,
      "epoch": 4.620827285921625,
      "grad_norm": 1.375,
      "learning_rate": 0.0003051685886686546,
      "loss": 4.8059,
      "mean_token_accuracy": 0.22512520998716354,
      "num_tokens": 105748786.0,
      "step": 50940
    },
    {
      "entropy": 5.6374183177948,
      "epoch": 4.6212808417997095,
      "grad_norm": 1.296875,
      "learning_rate": 0.000305136524110405,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22338816821575164,
      "num_tokens": 105759156.0,
      "step": 50945
    },
    {
      "entropy": 5.617321157455445,
      "epoch": 4.621734397677794,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003051044589289245,
      "loss": 4.9121,
      "mean_token_accuracy": 0.219149149954319,
      "num_tokens": 105769434.0,
      "step": 50950
    },
    {
      "entropy": 5.680973052978516,
      "epoch": 4.622187953555878,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00030507239312487595,
      "loss": 4.995,
      "mean_token_accuracy": 0.21889229416847228,
      "num_tokens": 105779408.0,
      "step": 50955
    },
    {
      "entropy": 5.6847470760345455,
      "epoch": 4.622641509433962,
      "grad_norm": 1.375,
      "learning_rate": 0.00030504032669892276,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21904740184545518,
      "num_tokens": 105789517.0,
      "step": 50960
    },
    {
      "entropy": 5.668029403686523,
      "epoch": 4.623095065312047,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030500825965172797,
      "loss": 4.8435,
      "mean_token_accuracy": 0.2260068491101265,
      "num_tokens": 105799823.0,
      "step": 50965
    },
    {
      "entropy": 5.6476218700408936,
      "epoch": 4.623548621190131,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030497619198395466,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22505230605602264,
      "num_tokens": 105810038.0,
      "step": 50970
    },
    {
      "entropy": 5.645989513397216,
      "epoch": 4.624002177068215,
      "grad_norm": 1.375,
      "learning_rate": 0.00030494412369626615,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2184138759970665,
      "num_tokens": 105819940.0,
      "step": 50975
    },
    {
      "entropy": 5.619664335250855,
      "epoch": 4.6244557329462985,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003049120547893255,
      "loss": 4.8377,
      "mean_token_accuracy": 0.2263239488005638,
      "num_tokens": 105829829.0,
      "step": 50980
    },
    {
      "entropy": 5.628520584106445,
      "epoch": 4.624909288824383,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003048799852637958,
      "loss": 4.8201,
      "mean_token_accuracy": 0.23618641793727874,
      "num_tokens": 105839463.0,
      "step": 50985
    },
    {
      "entropy": 5.722886323928833,
      "epoch": 4.625362844702467,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030484791512034063,
      "loss": 5.0337,
      "mean_token_accuracy": 0.21005859822034836,
      "num_tokens": 105849916.0,
      "step": 50990
    },
    {
      "entropy": 5.611998462677002,
      "epoch": 4.625816400580551,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003048158443596228,
      "loss": 4.8298,
      "mean_token_accuracy": 0.2257625624537468,
      "num_tokens": 105860699.0,
      "step": 50995
    },
    {
      "entropy": 5.658289623260498,
      "epoch": 4.626269956458636,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003047837729823058,
      "loss": 4.8218,
      "mean_token_accuracy": 0.22070784419775008,
      "num_tokens": 105870332.0,
      "step": 51000
    },
    {
      "epoch": 4.626269956458636,
      "eval_entropy": 5.314673195813037,
      "eval_loss": 4.943879127502441,
      "eval_mean_token_accuracy": 0.23091070976327446,
      "eval_num_tokens": 105870332.0,
      "eval_runtime": 20.5109,
      "eval_samples_per_second": 1723.961,
      "eval_steps_per_second": 215.495,
      "step": 51000
    },
    {
      "entropy": 5.5747912406921385,
      "epoch": 4.62672351233672,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003047517009890527,
      "loss": 4.8402,
      "mean_token_accuracy": 0.23035148084163665,
      "num_tokens": 105879364.0,
      "step": 51005
    },
    {
      "entropy": 5.654276990890503,
      "epoch": 4.627177068214804,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030471962838052693,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21786375790834428,
      "num_tokens": 105889554.0,
      "step": 51010
    },
    {
      "entropy": 5.637483835220337,
      "epoch": 4.627630624092888,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003046875551573916,
      "loss": 4.7727,
      "mean_token_accuracy": 0.2379957765340805,
      "num_tokens": 105900426.0,
      "step": 51015
    },
    {
      "entropy": 5.530906009674072,
      "epoch": 4.628084179970973,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003046554813203101,
      "loss": 4.6728,
      "mean_token_accuracy": 0.23877642303705215,
      "num_tokens": 105909541.0,
      "step": 51020
    },
    {
      "entropy": 5.646114540100098,
      "epoch": 4.628537735849057,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003046234068699457,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22120093554258347,
      "num_tokens": 105919951.0,
      "step": 51025
    },
    {
      "entropy": 5.574805164337159,
      "epoch": 4.628991291727141,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003045913318069616,
      "loss": 4.7788,
      "mean_token_accuracy": 0.2319977819919586,
      "num_tokens": 105929048.0,
      "step": 51030
    },
    {
      "entropy": 5.6524749279022215,
      "epoch": 4.629444847605225,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003045592561320213,
      "loss": 4.7947,
      "mean_token_accuracy": 0.23470182120800018,
      "num_tokens": 105938105.0,
      "step": 51035
    },
    {
      "entropy": 5.650112152099609,
      "epoch": 4.629898403483309,
      "grad_norm": 1.4375,
      "learning_rate": 0.000304527179845788,
      "loss": 4.8671,
      "mean_token_accuracy": 0.22331915199756622,
      "num_tokens": 105948167.0,
      "step": 51040
    },
    {
      "entropy": 5.68101396560669,
      "epoch": 4.630351959361393,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003044951029489252,
      "loss": 4.889,
      "mean_token_accuracy": 0.2294978141784668,
      "num_tokens": 105959387.0,
      "step": 51045
    },
    {
      "entropy": 5.567787599563599,
      "epoch": 4.630805515239477,
      "grad_norm": 1.375,
      "learning_rate": 0.000304463025442096,
      "loss": 4.7713,
      "mean_token_accuracy": 0.2325567200779915,
      "num_tokens": 105969436.0,
      "step": 51050
    },
    {
      "entropy": 5.6713114261627195,
      "epoch": 4.631259071117562,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003044309473259641,
      "loss": 4.8774,
      "mean_token_accuracy": 0.2301829233765602,
      "num_tokens": 105979102.0,
      "step": 51055
    },
    {
      "entropy": 5.568918037414551,
      "epoch": 4.631712626995646,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030439886860119265,
      "loss": 4.779,
      "mean_token_accuracy": 0.23363327533006667,
      "num_tokens": 105989796.0,
      "step": 51060
    },
    {
      "entropy": 5.647738695144653,
      "epoch": 4.63216618287373,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000304366789268445,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22614934146404267,
      "num_tokens": 106000194.0,
      "step": 51065
    },
    {
      "entropy": 5.656801319122314,
      "epoch": 4.632619738751814,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003043347093283848,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22746677100658416,
      "num_tokens": 106009108.0,
      "step": 51070
    },
    {
      "entropy": 5.601779508590698,
      "epoch": 4.633073294629899,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030430262878167533,
      "loss": 4.8111,
      "mean_token_accuracy": 0.2262859523296356,
      "num_tokens": 106019342.0,
      "step": 51075
    },
    {
      "entropy": 5.689200305938721,
      "epoch": 4.633526850507983,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003042705476289799,
      "loss": 4.9175,
      "mean_token_accuracy": 0.21309079527854918,
      "num_tokens": 106029079.0,
      "step": 51080
    },
    {
      "entropy": 5.5901631832122805,
      "epoch": 4.633980406386067,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003042384658709622,
      "loss": 4.8004,
      "mean_token_accuracy": 0.23674729317426682,
      "num_tokens": 106039406.0,
      "step": 51085
    },
    {
      "entropy": 5.640896844863891,
      "epoch": 4.634433962264151,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003042063835082856,
      "loss": 4.873,
      "mean_token_accuracy": 0.2213134765625,
      "num_tokens": 106049603.0,
      "step": 51090
    },
    {
      "entropy": 5.630808973312378,
      "epoch": 4.634887518142235,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003041743005416135,
      "loss": 4.8253,
      "mean_token_accuracy": 0.2368577390909195,
      "num_tokens": 106058692.0,
      "step": 51095
    },
    {
      "entropy": 5.585863924026489,
      "epoch": 4.635341074020319,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003041422169716094,
      "loss": 4.8192,
      "mean_token_accuracy": 0.2365136831998825,
      "num_tokens": 106069307.0,
      "step": 51100
    },
    {
      "entropy": 5.667902517318725,
      "epoch": 4.635794629898403,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003041101327989368,
      "loss": 4.8613,
      "mean_token_accuracy": 0.2321127638220787,
      "num_tokens": 106080831.0,
      "step": 51105
    },
    {
      "entropy": 5.644593000411987,
      "epoch": 4.636248185776488,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003040780480242593,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22098470181226731,
      "num_tokens": 106090007.0,
      "step": 51110
    },
    {
      "entropy": 5.662961959838867,
      "epoch": 4.636701741654572,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030404596264824033,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22854455113410949,
      "num_tokens": 106099790.0,
      "step": 51115
    },
    {
      "entropy": 5.63086256980896,
      "epoch": 4.637155297532656,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030401387667154347,
      "loss": 4.797,
      "mean_token_accuracy": 0.23479324728250503,
      "num_tokens": 106110364.0,
      "step": 51120
    },
    {
      "entropy": 5.543121385574341,
      "epoch": 4.6376088534107405,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003039817900948322,
      "loss": 4.7613,
      "mean_token_accuracy": 0.23455163985490798,
      "num_tokens": 106121353.0,
      "step": 51125
    },
    {
      "entropy": 5.60977725982666,
      "epoch": 4.638062409288825,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003039497029187702,
      "loss": 4.8482,
      "mean_token_accuracy": 0.22174757421016694,
      "num_tokens": 106130750.0,
      "step": 51130
    },
    {
      "entropy": 5.628994560241699,
      "epoch": 4.638515965166908,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030391761514402096,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23218538612127304,
      "num_tokens": 106142014.0,
      "step": 51135
    },
    {
      "entropy": 5.636723756790161,
      "epoch": 4.638969521044992,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030388552677124804,
      "loss": 4.8404,
      "mean_token_accuracy": 0.2345106691122055,
      "num_tokens": 106152419.0,
      "step": 51140
    },
    {
      "entropy": 5.611972284317017,
      "epoch": 4.639423076923077,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000303853437801115,
      "loss": 4.7852,
      "mean_token_accuracy": 0.22788599133491516,
      "num_tokens": 106163365.0,
      "step": 51145
    },
    {
      "entropy": 5.590104341506958,
      "epoch": 4.639876632801161,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003038213482342856,
      "loss": 4.7932,
      "mean_token_accuracy": 0.23682606667280198,
      "num_tokens": 106173324.0,
      "step": 51150
    },
    {
      "entropy": 5.71228404045105,
      "epoch": 4.640330188679245,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030378925807142334,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22683367133140564,
      "num_tokens": 106183427.0,
      "step": 51155
    },
    {
      "entropy": 5.668336677551269,
      "epoch": 4.6407837445573294,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030375716731319193,
      "loss": 4.9095,
      "mean_token_accuracy": 0.21664217710494996,
      "num_tokens": 106192894.0,
      "step": 51160
    },
    {
      "entropy": 5.622656917572021,
      "epoch": 4.641237300435414,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030372507596025493,
      "loss": 4.8237,
      "mean_token_accuracy": 0.2273292899131775,
      "num_tokens": 106203185.0,
      "step": 51165
    },
    {
      "entropy": 5.736173963546753,
      "epoch": 4.641690856313498,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000303692984013276,
      "loss": 4.965,
      "mean_token_accuracy": 0.21472627073526382,
      "num_tokens": 106213817.0,
      "step": 51170
    },
    {
      "entropy": 5.59650616645813,
      "epoch": 4.642144412191582,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003036608914729189,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23001222610473632,
      "num_tokens": 106224367.0,
      "step": 51175
    },
    {
      "entropy": 5.516629028320312,
      "epoch": 4.6425979680696665,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003036287983398473,
      "loss": 4.7681,
      "mean_token_accuracy": 0.2355932265520096,
      "num_tokens": 106234146.0,
      "step": 51180
    },
    {
      "entropy": 5.654841756820678,
      "epoch": 4.643051523947751,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030359670461472475,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22010868638753892,
      "num_tokens": 106244910.0,
      "step": 51185
    },
    {
      "entropy": 5.71726245880127,
      "epoch": 4.643505079825834,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030356461029821516,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21086065471172333,
      "num_tokens": 106254267.0,
      "step": 51190
    },
    {
      "entropy": 5.688195371627808,
      "epoch": 4.643958635703918,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003035325153909821,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22264937609434127,
      "num_tokens": 106264755.0,
      "step": 51195
    },
    {
      "entropy": 5.587097215652466,
      "epoch": 4.644412191582003,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003035004198936894,
      "loss": 4.8377,
      "mean_token_accuracy": 0.2265124186873436,
      "num_tokens": 106273964.0,
      "step": 51200
    },
    {
      "entropy": 5.652390336990356,
      "epoch": 4.644865747460087,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003034683238070008,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22290389537811278,
      "num_tokens": 106283933.0,
      "step": 51205
    },
    {
      "entropy": 5.675181293487549,
      "epoch": 4.645319303338171,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030343622713157985,
      "loss": 4.8377,
      "mean_token_accuracy": 0.23244220912456512,
      "num_tokens": 106294075.0,
      "step": 51210
    },
    {
      "entropy": 5.637624263763428,
      "epoch": 4.6457728592162555,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003034041298680906,
      "loss": 4.814,
      "mean_token_accuracy": 0.230797703564167,
      "num_tokens": 106304970.0,
      "step": 51215
    },
    {
      "entropy": 5.629592323303223,
      "epoch": 4.64622641509434,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003033720320171966,
      "loss": 4.8345,
      "mean_token_accuracy": 0.22397781312465667,
      "num_tokens": 106315610.0,
      "step": 51220
    },
    {
      "entropy": 5.670551395416259,
      "epoch": 4.646679970972424,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003033399335795619,
      "loss": 4.8843,
      "mean_token_accuracy": 0.2246227502822876,
      "num_tokens": 106326302.0,
      "step": 51225
    },
    {
      "entropy": 5.612484550476074,
      "epoch": 4.647133526850508,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030330783455585,
      "loss": 4.8228,
      "mean_token_accuracy": 0.22766892313957215,
      "num_tokens": 106336350.0,
      "step": 51230
    },
    {
      "entropy": 5.5970076560974125,
      "epoch": 4.647587082728592,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003032757349467249,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2233673706650734,
      "num_tokens": 106348233.0,
      "step": 51235
    },
    {
      "entropy": 5.582151556015015,
      "epoch": 4.648040638606677,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030324363475285046,
      "loss": 4.779,
      "mean_token_accuracy": 0.23166831135749816,
      "num_tokens": 106357992.0,
      "step": 51240
    },
    {
      "entropy": 5.645804071426392,
      "epoch": 4.64849419448476,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003032115339748904,
      "loss": 4.7545,
      "mean_token_accuracy": 0.2370196595788002,
      "num_tokens": 106367683.0,
      "step": 51245
    },
    {
      "entropy": 5.655009746551514,
      "epoch": 4.6489477503628445,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003031794326135086,
      "loss": 4.8324,
      "mean_token_accuracy": 0.23029607087373732,
      "num_tokens": 106378275.0,
      "step": 51250
    },
    {
      "entropy": 5.642401790618896,
      "epoch": 4.649401306240929,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030314733066936903,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23249487429857255,
      "num_tokens": 106387811.0,
      "step": 51255
    },
    {
      "entropy": 5.60443959236145,
      "epoch": 4.649854862119013,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003031152281431355,
      "loss": 4.879,
      "mean_token_accuracy": 0.224543035030365,
      "num_tokens": 106398327.0,
      "step": 51260
    },
    {
      "entropy": 5.569450092315674,
      "epoch": 4.650308417997097,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030308312503547183,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2261110723018646,
      "num_tokens": 106408602.0,
      "step": 51265
    },
    {
      "entropy": 5.5927026748657225,
      "epoch": 4.6507619738751815,
      "grad_norm": 1.28125,
      "learning_rate": 0.000303051021347042,
      "loss": 4.7968,
      "mean_token_accuracy": 0.2282913878560066,
      "num_tokens": 106418717.0,
      "step": 51270
    },
    {
      "entropy": 5.585525941848755,
      "epoch": 4.651215529753266,
      "grad_norm": 1.515625,
      "learning_rate": 0.00030301891707850987,
      "loss": 4.7392,
      "mean_token_accuracy": 0.23537057787179946,
      "num_tokens": 106428311.0,
      "step": 51275
    },
    {
      "entropy": 5.630226087570191,
      "epoch": 4.65166908563135,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003029868122305395,
      "loss": 4.8502,
      "mean_token_accuracy": 0.22298333346843718,
      "num_tokens": 106438772.0,
      "step": 51280
    },
    {
      "entropy": 5.631834506988525,
      "epoch": 4.652122641509434,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030295470680379456,
      "loss": 4.8358,
      "mean_token_accuracy": 0.23331484645605088,
      "num_tokens": 106448895.0,
      "step": 51285
    },
    {
      "entropy": 5.60748438835144,
      "epoch": 4.652576197387518,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003029226007989393,
      "loss": 4.913,
      "mean_token_accuracy": 0.22834351062774658,
      "num_tokens": 106459711.0,
      "step": 51290
    },
    {
      "entropy": 5.642921686172485,
      "epoch": 4.653029753265602,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030289049421663744,
      "loss": 4.8264,
      "mean_token_accuracy": 0.23212611228227614,
      "num_tokens": 106469691.0,
      "step": 51295
    },
    {
      "entropy": 5.647533226013183,
      "epoch": 4.653483309143686,
      "grad_norm": 1.375,
      "learning_rate": 0.0003028583870575531,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22049656510353088,
      "num_tokens": 106479894.0,
      "step": 51300
    },
    {
      "entropy": 5.634368848800659,
      "epoch": 4.6539368650217705,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003028262793223502,
      "loss": 4.7924,
      "mean_token_accuracy": 0.23526830077171326,
      "num_tokens": 106491444.0,
      "step": 51305
    },
    {
      "entropy": 5.638995265960693,
      "epoch": 4.654390420899855,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030279417101169276,
      "loss": 4.8552,
      "mean_token_accuracy": 0.23144645392894744,
      "num_tokens": 106501633.0,
      "step": 51310
    },
    {
      "entropy": 5.5927633285522464,
      "epoch": 4.654843976777939,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003027620621262448,
      "loss": 4.8099,
      "mean_token_accuracy": 0.23860193490982057,
      "num_tokens": 106511272.0,
      "step": 51315
    },
    {
      "entropy": 5.614375925064087,
      "epoch": 4.655297532656023,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003027299526666703,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2255490705370903,
      "num_tokens": 106521977.0,
      "step": 51320
    },
    {
      "entropy": 5.65538911819458,
      "epoch": 4.655751088534108,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00030269784263363327,
      "loss": 4.8744,
      "mean_token_accuracy": 0.2264071598649025,
      "num_tokens": 106532802.0,
      "step": 51325
    },
    {
      "entropy": 5.617374420166016,
      "epoch": 4.656204644412192,
      "grad_norm": 1.453125,
      "learning_rate": 0.00030266573202779784,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22736508399248123,
      "num_tokens": 106542249.0,
      "step": 51330
    },
    {
      "entropy": 5.615729951858521,
      "epoch": 4.656658200290276,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030263362084982805,
      "loss": 4.8583,
      "mean_token_accuracy": 0.2309017613530159,
      "num_tokens": 106552501.0,
      "step": 51335
    },
    {
      "entropy": 5.60629186630249,
      "epoch": 4.65711175616836,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030260150910038793,
      "loss": 4.8084,
      "mean_token_accuracy": 0.23878553062677382,
      "num_tokens": 106561901.0,
      "step": 51340
    },
    {
      "entropy": 5.602769184112549,
      "epoch": 4.657565312046444,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030256939678014155,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22884787172079085,
      "num_tokens": 106572104.0,
      "step": 51345
    },
    {
      "entropy": 5.616610479354859,
      "epoch": 4.658018867924528,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003025372838897529,
      "loss": 4.8568,
      "mean_token_accuracy": 0.22390409111976622,
      "num_tokens": 106583436.0,
      "step": 51350
    },
    {
      "entropy": 5.6590341091156,
      "epoch": 4.658472423802612,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003025051704298864,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22177911251783372,
      "num_tokens": 106593879.0,
      "step": 51355
    },
    {
      "entropy": 5.649652194976807,
      "epoch": 4.6589259796806965,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030247305640120585,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22046148627996445,
      "num_tokens": 106604774.0,
      "step": 51360
    },
    {
      "entropy": 5.700344800949097,
      "epoch": 4.659379535558781,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003024409418043756,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21837965101003648,
      "num_tokens": 106616125.0,
      "step": 51365
    },
    {
      "entropy": 5.657768249511719,
      "epoch": 4.659833091436865,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003024088266400595,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22354645431041717,
      "num_tokens": 106625171.0,
      "step": 51370
    },
    {
      "entropy": 5.609553813934326,
      "epoch": 4.660286647314949,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030237671090892204,
      "loss": 4.8394,
      "mean_token_accuracy": 0.2277699053287506,
      "num_tokens": 106636455.0,
      "step": 51375
    },
    {
      "entropy": 5.618097639083862,
      "epoch": 4.660740203193034,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003023445946116272,
      "loss": 4.825,
      "mean_token_accuracy": 0.22416653037071227,
      "num_tokens": 106645956.0,
      "step": 51380
    },
    {
      "entropy": 5.649156665802002,
      "epoch": 4.661193759071118,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030231247774883913,
      "loss": 4.902,
      "mean_token_accuracy": 0.2244000717997551,
      "num_tokens": 106655562.0,
      "step": 51385
    },
    {
      "entropy": 5.661860227584839,
      "epoch": 4.661647314949201,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003022803603212221,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2173143208026886,
      "num_tokens": 106666620.0,
      "step": 51390
    },
    {
      "entropy": 5.654275894165039,
      "epoch": 4.662100870827286,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003022482423294403,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2250022143125534,
      "num_tokens": 106676881.0,
      "step": 51395
    },
    {
      "entropy": 5.619033575057983,
      "epoch": 4.66255442670537,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003022161237741578,
      "loss": 4.7935,
      "mean_token_accuracy": 0.22898587584495544,
      "num_tokens": 106687765.0,
      "step": 51400
    },
    {
      "entropy": 5.599718284606934,
      "epoch": 4.663007982583454,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030218400465603895,
      "loss": 4.8292,
      "mean_token_accuracy": 0.22485257238149642,
      "num_tokens": 106698796.0,
      "step": 51405
    },
    {
      "entropy": 5.624629354476928,
      "epoch": 4.663461538461538,
      "grad_norm": 1.46875,
      "learning_rate": 0.000302151884975748,
      "loss": 4.827,
      "mean_token_accuracy": 0.23007217049598694,
      "num_tokens": 106709439.0,
      "step": 51410
    },
    {
      "entropy": 5.678779888153076,
      "epoch": 4.663915094339623,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003021197647339492,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21878890693187714,
      "num_tokens": 106720325.0,
      "step": 51415
    },
    {
      "entropy": 5.640367412567139,
      "epoch": 4.664368650217707,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003020876439313066,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22896699160337447,
      "num_tokens": 106730526.0,
      "step": 51420
    },
    {
      "entropy": 5.620300006866455,
      "epoch": 4.664822206095791,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030205552256848476,
      "loss": 4.834,
      "mean_token_accuracy": 0.2285284236073494,
      "num_tokens": 106741142.0,
      "step": 51425
    },
    {
      "entropy": 5.717283916473389,
      "epoch": 4.665275761973875,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030202340064614783,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21465857923030854,
      "num_tokens": 106751843.0,
      "step": 51430
    },
    {
      "entropy": 5.580708694458008,
      "epoch": 4.66572931785196,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003019912781649599,
      "loss": 4.7107,
      "mean_token_accuracy": 0.23825723975896834,
      "num_tokens": 106761726.0,
      "step": 51435
    },
    {
      "entropy": 5.613479375839233,
      "epoch": 4.666182873730044,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030195915512558564,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22970918864011763,
      "num_tokens": 106771829.0,
      "step": 51440
    },
    {
      "entropy": 5.567964363098144,
      "epoch": 4.666636429608127,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003019270315286892,
      "loss": 4.7633,
      "mean_token_accuracy": 0.23079639673233032,
      "num_tokens": 106783457.0,
      "step": 51445
    },
    {
      "entropy": 5.623996210098267,
      "epoch": 4.667089985486212,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030189490737493477,
      "loss": 4.834,
      "mean_token_accuracy": 0.2352101758122444,
      "num_tokens": 106793290.0,
      "step": 51450
    },
    {
      "entropy": 5.639394664764405,
      "epoch": 4.667543541364296,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030186278266498677,
      "loss": 4.8147,
      "mean_token_accuracy": 0.22420746982097625,
      "num_tokens": 106804412.0,
      "step": 51455
    },
    {
      "entropy": 5.701258373260498,
      "epoch": 4.66799709724238,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003018306573995097,
      "loss": 4.9477,
      "mean_token_accuracy": 0.22613338828086854,
      "num_tokens": 106815353.0,
      "step": 51460
    },
    {
      "entropy": 5.642526912689209,
      "epoch": 4.668450653120464,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030179853157916777,
      "loss": 4.8515,
      "mean_token_accuracy": 0.2293083295226097,
      "num_tokens": 106826350.0,
      "step": 51465
    },
    {
      "entropy": 5.664902782440185,
      "epoch": 4.668904208998549,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003017664052046253,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22597613632678987,
      "num_tokens": 106837434.0,
      "step": 51470
    },
    {
      "entropy": 5.6065673828125,
      "epoch": 4.669357764876633,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030173427827654683,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22734495848417283,
      "num_tokens": 106847829.0,
      "step": 51475
    },
    {
      "entropy": 5.718232297897339,
      "epoch": 4.669811320754717,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003017021507955967,
      "loss": 4.969,
      "mean_token_accuracy": 0.21826019883155823,
      "num_tokens": 106858519.0,
      "step": 51480
    },
    {
      "entropy": 5.668228054046631,
      "epoch": 4.670264876632801,
      "grad_norm": 1.375,
      "learning_rate": 0.00030167002276243915,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2310011550784111,
      "num_tokens": 106868614.0,
      "step": 51485
    },
    {
      "entropy": 5.670083093643188,
      "epoch": 4.670718432510886,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030163789417773885,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21401710510253907,
      "num_tokens": 106879592.0,
      "step": 51490
    },
    {
      "entropy": 5.62853102684021,
      "epoch": 4.67117198838897,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003016057650421601,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22642157822847367,
      "num_tokens": 106890083.0,
      "step": 51495
    },
    {
      "entropy": 5.682906818389893,
      "epoch": 4.671625544267053,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030157363535636726,
      "loss": 4.9272,
      "mean_token_accuracy": 0.22532157599925995,
      "num_tokens": 106900672.0,
      "step": 51500
    },
    {
      "entropy": 5.646746587753296,
      "epoch": 4.672079100145138,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000301541505121025,
      "loss": 4.8084,
      "mean_token_accuracy": 0.23173405081033707,
      "num_tokens": 106910396.0,
      "step": 51505
    },
    {
      "entropy": 5.683377122879028,
      "epoch": 4.672532656023222,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003015093743367976,
      "loss": 4.8956,
      "mean_token_accuracy": 0.22286300212144852,
      "num_tokens": 106919742.0,
      "step": 51510
    },
    {
      "entropy": 5.58050537109375,
      "epoch": 4.672986211901306,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003014772430043495,
      "loss": 4.7813,
      "mean_token_accuracy": 0.2332137331366539,
      "num_tokens": 106930837.0,
      "step": 51515
    },
    {
      "entropy": 5.594314289093018,
      "epoch": 4.67343976777939,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030144511112434536,
      "loss": 4.8212,
      "mean_token_accuracy": 0.22822803556919097,
      "num_tokens": 106942023.0,
      "step": 51520
    },
    {
      "entropy": 5.657700920104981,
      "epoch": 4.673893323657475,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030141297869744956,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22226927429437637,
      "num_tokens": 106951374.0,
      "step": 51525
    },
    {
      "entropy": 5.726312208175659,
      "epoch": 4.674346879535559,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030138084572432663,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21815831661224366,
      "num_tokens": 106961714.0,
      "step": 51530
    },
    {
      "entropy": 5.664337968826294,
      "epoch": 4.674800435413643,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003013487122056411,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2296933963894844,
      "num_tokens": 106971783.0,
      "step": 51535
    },
    {
      "entropy": 5.606761455535889,
      "epoch": 4.6752539912917275,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003013165781420575,
      "loss": 4.8038,
      "mean_token_accuracy": 0.2361462250351906,
      "num_tokens": 106982047.0,
      "step": 51540
    },
    {
      "entropy": 5.62720046043396,
      "epoch": 4.675707547169811,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030128444353424034,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2292511984705925,
      "num_tokens": 106993063.0,
      "step": 51545
    },
    {
      "entropy": 5.661016845703125,
      "epoch": 4.676161103047895,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030125230838285415,
      "loss": 4.8635,
      "mean_token_accuracy": 0.2229092761874199,
      "num_tokens": 107002865.0,
      "step": 51550
    },
    {
      "entropy": 5.679088735580445,
      "epoch": 4.676614658925979,
      "grad_norm": 1.25,
      "learning_rate": 0.00030122017268856353,
      "loss": 4.8237,
      "mean_token_accuracy": 0.23492152839899064,
      "num_tokens": 107013007.0,
      "step": 51555
    },
    {
      "entropy": 5.681649780273437,
      "epoch": 4.677068214804064,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00030118803645203315,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22368464469909669,
      "num_tokens": 107023109.0,
      "step": 51560
    },
    {
      "entropy": 5.611205339431763,
      "epoch": 4.677521770682148,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003011558996739274,
      "loss": 4.8341,
      "mean_token_accuracy": 0.2304351270198822,
      "num_tokens": 107033824.0,
      "step": 51565
    },
    {
      "entropy": 5.634835815429687,
      "epoch": 4.677975326560232,
      "grad_norm": 1.359375,
      "learning_rate": 0.000301123762354911,
      "loss": 4.879,
      "mean_token_accuracy": 0.224780935049057,
      "num_tokens": 107043529.0,
      "step": 51570
    },
    {
      "entropy": 5.651682662963867,
      "epoch": 4.6784288824383164,
      "grad_norm": 1.25,
      "learning_rate": 0.0003010916244956486,
      "loss": 4.8935,
      "mean_token_accuracy": 0.2331716611981392,
      "num_tokens": 107054378.0,
      "step": 51575
    },
    {
      "entropy": 5.619350814819336,
      "epoch": 4.678882438316401,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003010594860968048,
      "loss": 4.776,
      "mean_token_accuracy": 0.22617520838975907,
      "num_tokens": 107064300.0,
      "step": 51580
    },
    {
      "entropy": 5.706398916244507,
      "epoch": 4.679335994194485,
      "grad_norm": 1.375,
      "learning_rate": 0.000301027347159044,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2200508400797844,
      "num_tokens": 107074560.0,
      "step": 51585
    },
    {
      "entropy": 5.645678901672364,
      "epoch": 4.679789550072569,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003009952076830312,
      "loss": 4.8285,
      "mean_token_accuracy": 0.22575079798698425,
      "num_tokens": 107084574.0,
      "step": 51590
    },
    {
      "entropy": 5.611685180664063,
      "epoch": 4.6802431059506535,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003009630676694308,
      "loss": 4.8555,
      "mean_token_accuracy": 0.21997918635606767,
      "num_tokens": 107095207.0,
      "step": 51595
    },
    {
      "entropy": 5.603780698776245,
      "epoch": 4.680696661828737,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030093092711890756,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2265464410185814,
      "num_tokens": 107106091.0,
      "step": 51600
    },
    {
      "entropy": 5.641469144821167,
      "epoch": 4.681150217706821,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003008987860321261,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22050147503614426,
      "num_tokens": 107117753.0,
      "step": 51605
    },
    {
      "entropy": 5.575567531585693,
      "epoch": 4.681603773584905,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003008666444097512,
      "loss": 4.7769,
      "mean_token_accuracy": 0.2383032351732254,
      "num_tokens": 107128131.0,
      "step": 51610
    },
    {
      "entropy": 5.665344142913819,
      "epoch": 4.68205732946299,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003008345022524475,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21450660675764083,
      "num_tokens": 107139447.0,
      "step": 51615
    },
    {
      "entropy": 5.652925729751587,
      "epoch": 4.682510885341074,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030080235956087963,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22520839273929597,
      "num_tokens": 107149721.0,
      "step": 51620
    },
    {
      "entropy": 5.65245680809021,
      "epoch": 4.682964441219158,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003007702163357125,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22065943628549575,
      "num_tokens": 107159814.0,
      "step": 51625
    },
    {
      "entropy": 5.6026547908782955,
      "epoch": 4.6834179970972425,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030073807257761074,
      "loss": 4.8702,
      "mean_token_accuracy": 0.2243686869740486,
      "num_tokens": 107169012.0,
      "step": 51630
    },
    {
      "entropy": 5.6826526641845705,
      "epoch": 4.683871552975327,
      "grad_norm": 1.375,
      "learning_rate": 0.00030070592828723903,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22034205198287965,
      "num_tokens": 107179571.0,
      "step": 51635
    },
    {
      "entropy": 5.6335468769073485,
      "epoch": 4.684325108853411,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003006737834652621,
      "loss": 4.8554,
      "mean_token_accuracy": 0.22734544724225997,
      "num_tokens": 107190535.0,
      "step": 51640
    },
    {
      "entropy": 5.680628442764283,
      "epoch": 4.684778664731495,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003006416381123449,
      "loss": 4.8684,
      "mean_token_accuracy": 0.23113794773817062,
      "num_tokens": 107200361.0,
      "step": 51645
    },
    {
      "entropy": 5.650760126113892,
      "epoch": 4.68523222060958,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030060949222915205,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2257312372326851,
      "num_tokens": 107211080.0,
      "step": 51650
    },
    {
      "entropy": 5.662133693695068,
      "epoch": 4.685685776487663,
      "grad_norm": 1.375,
      "learning_rate": 0.0003005773458163484,
      "loss": 4.9025,
      "mean_token_accuracy": 0.22987064868211746,
      "num_tokens": 107220556.0,
      "step": 51655
    },
    {
      "entropy": 5.620728015899658,
      "epoch": 4.686139332365747,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030054519887459864,
      "loss": 4.8388,
      "mean_token_accuracy": 0.22177089601755143,
      "num_tokens": 107230693.0,
      "step": 51660
    },
    {
      "entropy": 5.674018096923828,
      "epoch": 4.6865928882438315,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030051305140456775,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21869321167469025,
      "num_tokens": 107240926.0,
      "step": 51665
    },
    {
      "entropy": 5.658247566223144,
      "epoch": 4.687046444121916,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030048090340692034,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22213871777057648,
      "num_tokens": 107251425.0,
      "step": 51670
    },
    {
      "entropy": 5.667091035842896,
      "epoch": 4.6875,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030044875488232145,
      "loss": 4.863,
      "mean_token_accuracy": 0.23359666019678116,
      "num_tokens": 107262046.0,
      "step": 51675
    },
    {
      "entropy": 5.566891098022461,
      "epoch": 4.687953555878084,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003004166058314358,
      "loss": 4.8042,
      "mean_token_accuracy": 0.232314233481884,
      "num_tokens": 107273141.0,
      "step": 51680
    },
    {
      "entropy": 5.659048557281494,
      "epoch": 4.6884071117561685,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003003844562549282,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22723358273506164,
      "num_tokens": 107282497.0,
      "step": 51685
    },
    {
      "entropy": 5.608068990707397,
      "epoch": 4.688860667634253,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030035230615346373,
      "loss": 4.7954,
      "mean_token_accuracy": 0.2298545867204666,
      "num_tokens": 107292710.0,
      "step": 51690
    },
    {
      "entropy": 5.599319124221802,
      "epoch": 4.689314223512337,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030032015552770695,
      "loss": 4.7207,
      "mean_token_accuracy": 0.2362816259264946,
      "num_tokens": 107303157.0,
      "step": 51695
    },
    {
      "entropy": 5.613744306564331,
      "epoch": 4.68976777939042,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030028800437832296,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22394119650125505,
      "num_tokens": 107313503.0,
      "step": 51700
    },
    {
      "entropy": 5.585458850860595,
      "epoch": 4.690221335268505,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003002558527059766,
      "loss": 4.8314,
      "mean_token_accuracy": 0.23098168671131133,
      "num_tokens": 107322276.0,
      "step": 51705
    },
    {
      "entropy": 5.629023647308349,
      "epoch": 4.690674891146589,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003002237005113328,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22017191350460052,
      "num_tokens": 107332876.0,
      "step": 51710
    },
    {
      "entropy": 5.620489931106567,
      "epoch": 4.691128447024673,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030019154779505644,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22940749675035477,
      "num_tokens": 107343262.0,
      "step": 51715
    },
    {
      "entropy": 5.563034105300903,
      "epoch": 4.6915820029027575,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003001593945578124,
      "loss": 4.8301,
      "mean_token_accuracy": 0.22212015390396117,
      "num_tokens": 107354281.0,
      "step": 51720
    },
    {
      "entropy": 5.5351745128631595,
      "epoch": 4.692035558780842,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003001272408002657,
      "loss": 4.8087,
      "mean_token_accuracy": 0.23216308504343033,
      "num_tokens": 107364011.0,
      "step": 51725
    },
    {
      "entropy": 5.610294485092163,
      "epoch": 4.692489114658926,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030009508652308124,
      "loss": 4.8117,
      "mean_token_accuracy": 0.2267228439450264,
      "num_tokens": 107372943.0,
      "step": 51730
    },
    {
      "entropy": 5.631402111053466,
      "epoch": 4.69294267053701,
      "grad_norm": 1.375,
      "learning_rate": 0.00030006293172692405,
      "loss": 4.888,
      "mean_token_accuracy": 0.22653815746307374,
      "num_tokens": 107382838.0,
      "step": 51735
    },
    {
      "entropy": 5.685107564926147,
      "epoch": 4.693396226415095,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030003077641245897,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22574774026870728,
      "num_tokens": 107393250.0,
      "step": 51740
    },
    {
      "entropy": 5.656757497787476,
      "epoch": 4.693849782293179,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029999862058035116,
      "loss": 4.758,
      "mean_token_accuracy": 0.2325229451060295,
      "num_tokens": 107403153.0,
      "step": 51745
    },
    {
      "entropy": 5.644321346282959,
      "epoch": 4.694303338171263,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029996646423126544,
      "loss": 4.849,
      "mean_token_accuracy": 0.22409629076719284,
      "num_tokens": 107413092.0,
      "step": 51750
    },
    {
      "entropy": 5.641963911056519,
      "epoch": 4.6947568940493465,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002999343073658669,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2267318844795227,
      "num_tokens": 107423561.0,
      "step": 51755
    },
    {
      "entropy": 5.6783990383148195,
      "epoch": 4.695210449927431,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029990214998482046,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22900304794311524,
      "num_tokens": 107433954.0,
      "step": 51760
    },
    {
      "entropy": 5.666752862930298,
      "epoch": 4.695664005805515,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002998699920887913,
      "loss": 4.8088,
      "mean_token_accuracy": 0.23007324635982512,
      "num_tokens": 107443150.0,
      "step": 51765
    },
    {
      "entropy": 5.603961086273193,
      "epoch": 4.696117561683599,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002998378336784443,
      "loss": 4.7993,
      "mean_token_accuracy": 0.2279962792992592,
      "num_tokens": 107454250.0,
      "step": 51770
    },
    {
      "entropy": 5.529533720016479,
      "epoch": 4.6965711175616836,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002998056747544446,
      "loss": 4.7849,
      "mean_token_accuracy": 0.23800675421953202,
      "num_tokens": 107465058.0,
      "step": 51775
    },
    {
      "entropy": 5.555084848403931,
      "epoch": 4.697024673439768,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002997735153174572,
      "loss": 4.793,
      "mean_token_accuracy": 0.2332381308078766,
      "num_tokens": 107475449.0,
      "step": 51780
    },
    {
      "entropy": 5.741664552688599,
      "epoch": 4.697478229317852,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029974135536814725,
      "loss": 5.0299,
      "mean_token_accuracy": 0.21843865215778352,
      "num_tokens": 107486645.0,
      "step": 51785
    },
    {
      "entropy": 5.641718864440918,
      "epoch": 4.697931785195936,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002997091949071797,
      "loss": 4.8598,
      "mean_token_accuracy": 0.2276191607117653,
      "num_tokens": 107497242.0,
      "step": 51790
    },
    {
      "entropy": 5.704970455169677,
      "epoch": 4.698385341074021,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002996770339352197,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22413975298404692,
      "num_tokens": 107507596.0,
      "step": 51795
    },
    {
      "entropy": 5.702626132965088,
      "epoch": 4.698838896952105,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002996448724529324,
      "loss": 4.9421,
      "mean_token_accuracy": 0.2197820097208023,
      "num_tokens": 107517795.0,
      "step": 51800
    },
    {
      "entropy": 5.651201486587524,
      "epoch": 4.699292452830189,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029961271046098284,
      "loss": 4.835,
      "mean_token_accuracy": 0.22721916139125825,
      "num_tokens": 107527805.0,
      "step": 51805
    },
    {
      "entropy": 5.650051498413086,
      "epoch": 4.6997460087082725,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029958054796003615,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22574893683195113,
      "num_tokens": 107537396.0,
      "step": 51810
    },
    {
      "entropy": 5.622155714035034,
      "epoch": 4.700199564586357,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002995483849507575,
      "loss": 4.8446,
      "mean_token_accuracy": 0.22888796478509904,
      "num_tokens": 107546876.0,
      "step": 51815
    },
    {
      "entropy": 5.637112808227539,
      "epoch": 4.700653120464441,
      "grad_norm": 1.5,
      "learning_rate": 0.000299516221433812,
      "loss": 4.8056,
      "mean_token_accuracy": 0.23163412660360336,
      "num_tokens": 107556865.0,
      "step": 51820
    },
    {
      "entropy": 5.624751329421997,
      "epoch": 4.701106676342525,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029948405740986466,
      "loss": 4.8444,
      "mean_token_accuracy": 0.22313083708286285,
      "num_tokens": 107566305.0,
      "step": 51825
    },
    {
      "entropy": 5.691544008255005,
      "epoch": 4.70156023222061,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002994518928795809,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22502174973487854,
      "num_tokens": 107575640.0,
      "step": 51830
    },
    {
      "entropy": 5.668277931213379,
      "epoch": 4.702013788098694,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029941972784362574,
      "loss": 4.936,
      "mean_token_accuracy": 0.2319621965289116,
      "num_tokens": 107585955.0,
      "step": 51835
    },
    {
      "entropy": 5.672263479232788,
      "epoch": 4.702467343976778,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029938756230266436,
      "loss": 4.8753,
      "mean_token_accuracy": 0.22293990701436997,
      "num_tokens": 107596911.0,
      "step": 51840
    },
    {
      "entropy": 5.607419157028199,
      "epoch": 4.702920899854862,
      "grad_norm": 1.4375,
      "learning_rate": 0.000299355396257362,
      "loss": 4.8033,
      "mean_token_accuracy": 0.23086132407188414,
      "num_tokens": 107606117.0,
      "step": 51845
    },
    {
      "entropy": 5.583935165405274,
      "epoch": 4.703374455732947,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002993232297083838,
      "loss": 4.7101,
      "mean_token_accuracy": 0.245095694065094,
      "num_tokens": 107616363.0,
      "step": 51850
    },
    {
      "entropy": 5.581477355957031,
      "epoch": 4.70382801161103,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002992910626563951,
      "loss": 4.7819,
      "mean_token_accuracy": 0.2300983339548111,
      "num_tokens": 107627402.0,
      "step": 51855
    },
    {
      "entropy": 5.5818596363067625,
      "epoch": 4.704281567489114,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029925889510206094,
      "loss": 4.8472,
      "mean_token_accuracy": 0.2269759252667427,
      "num_tokens": 107637446.0,
      "step": 51860
    },
    {
      "entropy": 5.59421820640564,
      "epoch": 4.704735123367199,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002992267270460467,
      "loss": 4.8227,
      "mean_token_accuracy": 0.22895709723234176,
      "num_tokens": 107647386.0,
      "step": 51865
    },
    {
      "entropy": 5.653500080108643,
      "epoch": 4.705188679245283,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002991945584890175,
      "loss": 4.9016,
      "mean_token_accuracy": 0.2291799858212471,
      "num_tokens": 107657881.0,
      "step": 51870
    },
    {
      "entropy": 5.662353372573852,
      "epoch": 4.705642235123367,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029916238943163874,
      "loss": 4.8513,
      "mean_token_accuracy": 0.2335037425160408,
      "num_tokens": 107668478.0,
      "step": 51875
    },
    {
      "entropy": 5.7356596946716305,
      "epoch": 4.706095791001451,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002991302198745756,
      "loss": 4.9793,
      "mean_token_accuracy": 0.22458555400371552,
      "num_tokens": 107679149.0,
      "step": 51880
    },
    {
      "entropy": 5.643193197250366,
      "epoch": 4.706549346879536,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002990980498184934,
      "loss": 4.8728,
      "mean_token_accuracy": 0.21858871430158616,
      "num_tokens": 107689471.0,
      "step": 51885
    },
    {
      "entropy": 5.626543474197388,
      "epoch": 4.70700290275762,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029906587926405734,
      "loss": 4.8394,
      "mean_token_accuracy": 0.2364283725619316,
      "num_tokens": 107700604.0,
      "step": 51890
    },
    {
      "entropy": 5.548242378234863,
      "epoch": 4.707456458635704,
      "grad_norm": 1.375,
      "learning_rate": 0.0002990337082119327,
      "loss": 4.7879,
      "mean_token_accuracy": 0.23130279928445815,
      "num_tokens": 107710268.0,
      "step": 51895
    },
    {
      "entropy": 5.67280330657959,
      "epoch": 4.707910014513788,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029900153666278503,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22368043959140776,
      "num_tokens": 107720328.0,
      "step": 51900
    },
    {
      "entropy": 5.634679508209229,
      "epoch": 4.708363570391873,
      "grad_norm": 1.375,
      "learning_rate": 0.00029896936461727924,
      "loss": 4.8433,
      "mean_token_accuracy": 0.22590590864419938,
      "num_tokens": 107730295.0,
      "step": 51905
    },
    {
      "entropy": 5.620717620849609,
      "epoch": 4.708817126269956,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029893719207608105,
      "loss": 4.8385,
      "mean_token_accuracy": 0.22394608110189437,
      "num_tokens": 107741334.0,
      "step": 51910
    },
    {
      "entropy": 5.647150039672852,
      "epoch": 4.70927068214804,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029890501903985555,
      "loss": 4.8459,
      "mean_token_accuracy": 0.22673273235559463,
      "num_tokens": 107752060.0,
      "step": 51915
    },
    {
      "entropy": 5.636652040481567,
      "epoch": 4.709724238026125,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029887284550926817,
      "loss": 4.8357,
      "mean_token_accuracy": 0.22352114021778108,
      "num_tokens": 107762188.0,
      "step": 51920
    },
    {
      "entropy": 5.6312517642974855,
      "epoch": 4.710177793904209,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002988406714849843,
      "loss": 4.8843,
      "mean_token_accuracy": 0.22537585645914077,
      "num_tokens": 107771670.0,
      "step": 51925
    },
    {
      "entropy": 5.603356742858887,
      "epoch": 4.710631349782293,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002988084969676692,
      "loss": 4.8196,
      "mean_token_accuracy": 0.22919293344020844,
      "num_tokens": 107781734.0,
      "step": 51930
    },
    {
      "entropy": 5.68241286277771,
      "epoch": 4.711084905660377,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029877632195798843,
      "loss": 4.867,
      "mean_token_accuracy": 0.21960362941026687,
      "num_tokens": 107792159.0,
      "step": 51935
    },
    {
      "entropy": 5.645976877212524,
      "epoch": 4.711538461538462,
      "grad_norm": 1.375,
      "learning_rate": 0.00029874414645660713,
      "loss": 4.85,
      "mean_token_accuracy": 0.22973107248544694,
      "num_tokens": 107803153.0,
      "step": 51940
    },
    {
      "entropy": 5.609997415542603,
      "epoch": 4.711992017416546,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002987119704641909,
      "loss": 4.8271,
      "mean_token_accuracy": 0.2304958611726761,
      "num_tokens": 107813214.0,
      "step": 51945
    },
    {
      "entropy": 5.618137407302856,
      "epoch": 4.71244557329463,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002986797939814051,
      "loss": 4.8632,
      "mean_token_accuracy": 0.2296712353825569,
      "num_tokens": 107823976.0,
      "step": 51950
    },
    {
      "entropy": 5.630889177322388,
      "epoch": 4.712899129172714,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029864761700891507,
      "loss": 4.8704,
      "mean_token_accuracy": 0.23087739795446396,
      "num_tokens": 107834127.0,
      "step": 51955
    },
    {
      "entropy": 5.688243627548218,
      "epoch": 4.713352685050799,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029861543954738635,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21547903567552568,
      "num_tokens": 107843753.0,
      "step": 51960
    },
    {
      "entropy": 5.696873903274536,
      "epoch": 4.713806240928882,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029858326159748434,
      "loss": 4.857,
      "mean_token_accuracy": 0.2304224044084549,
      "num_tokens": 107854690.0,
      "step": 51965
    },
    {
      "entropy": 5.6429298400878904,
      "epoch": 4.714259796806966,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029855108315987437,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23052936941385269,
      "num_tokens": 107865235.0,
      "step": 51970
    },
    {
      "entropy": 5.636487579345703,
      "epoch": 4.714713352685051,
      "grad_norm": 1.1875,
      "learning_rate": 0.000298518904235222,
      "loss": 4.8261,
      "mean_token_accuracy": 0.23074483424425124,
      "num_tokens": 107876052.0,
      "step": 51975
    },
    {
      "entropy": 5.5969034194946286,
      "epoch": 4.715166908563135,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002984867248241928,
      "loss": 4.863,
      "mean_token_accuracy": 0.23013104647397994,
      "num_tokens": 107886187.0,
      "step": 51980
    },
    {
      "entropy": 5.635718250274659,
      "epoch": 4.715620464441219,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002984545449274521,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22768629491329193,
      "num_tokens": 107896995.0,
      "step": 51985
    },
    {
      "entropy": 5.681853103637695,
      "epoch": 4.7160740203193035,
      "grad_norm": 1.375,
      "learning_rate": 0.0002984223645456655,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2227205827832222,
      "num_tokens": 107907868.0,
      "step": 51990
    },
    {
      "entropy": 5.669187068939209,
      "epoch": 4.716527576197388,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029839018367949825,
      "loss": 4.877,
      "mean_token_accuracy": 0.22584448754787445,
      "num_tokens": 107917564.0,
      "step": 51995
    },
    {
      "entropy": 5.639579010009766,
      "epoch": 4.716981132075472,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002983580023296162,
      "loss": 4.812,
      "mean_token_accuracy": 0.2401526689529419,
      "num_tokens": 107926853.0,
      "step": 52000
    },
    {
      "entropy": 5.58984408378601,
      "epoch": 4.717434687953556,
      "grad_norm": 1.25,
      "learning_rate": 0.00029832582049668465,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2184126004576683,
      "num_tokens": 107937440.0,
      "step": 52005
    },
    {
      "entropy": 5.632284355163574,
      "epoch": 4.71788824383164,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029829363818136917,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2248371109366417,
      "num_tokens": 107947750.0,
      "step": 52010
    },
    {
      "entropy": 5.681309700012207,
      "epoch": 4.718341799709724,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002982614553843354,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22780147194862366,
      "num_tokens": 107959646.0,
      "step": 52015
    },
    {
      "entropy": 5.660957336425781,
      "epoch": 4.718795355587808,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029822927210624873,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2240862786769867,
      "num_tokens": 107969678.0,
      "step": 52020
    },
    {
      "entropy": 5.67547492980957,
      "epoch": 4.719248911465892,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029819708834777483,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22985138297080993,
      "num_tokens": 107979589.0,
      "step": 52025
    },
    {
      "entropy": 5.5802422046661375,
      "epoch": 4.719702467343977,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002981649041095792,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23040609657764435,
      "num_tokens": 107990682.0,
      "step": 52030
    },
    {
      "entropy": 5.649995183944702,
      "epoch": 4.720156023222061,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002981327193923275,
      "loss": 4.8105,
      "mean_token_accuracy": 0.22992743104696273,
      "num_tokens": 108000999.0,
      "step": 52035
    },
    {
      "entropy": 5.59195613861084,
      "epoch": 4.720609579100145,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002981005341966852,
      "loss": 4.771,
      "mean_token_accuracy": 0.2258443132042885,
      "num_tokens": 108010709.0,
      "step": 52040
    },
    {
      "entropy": 5.573256778717041,
      "epoch": 4.7210631349782295,
      "grad_norm": 1.3125,
      "learning_rate": 0.000298068348523318,
      "loss": 4.806,
      "mean_token_accuracy": 0.2329244941473007,
      "num_tokens": 108020522.0,
      "step": 52045
    },
    {
      "entropy": 5.561178493499756,
      "epoch": 4.721516690856314,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002980361623728915,
      "loss": 4.8041,
      "mean_token_accuracy": 0.23204840272665023,
      "num_tokens": 108031387.0,
      "step": 52050
    },
    {
      "entropy": 5.634180974960327,
      "epoch": 4.721970246734398,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002980039757460713,
      "loss": 4.848,
      "mean_token_accuracy": 0.22833117842674255,
      "num_tokens": 108041653.0,
      "step": 52055
    },
    {
      "entropy": 5.7578123092651365,
      "epoch": 4.722423802612482,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029797178864352306,
      "loss": 4.9881,
      "mean_token_accuracy": 0.22197122275829315,
      "num_tokens": 108051969.0,
      "step": 52060
    },
    {
      "entropy": 5.66195330619812,
      "epoch": 4.722877358490566,
      "grad_norm": 1.375,
      "learning_rate": 0.00029793960106591226,
      "loss": 4.8605,
      "mean_token_accuracy": 0.22459030598402024,
      "num_tokens": 108061364.0,
      "step": 52065
    },
    {
      "entropy": 5.621276426315307,
      "epoch": 4.72333091436865,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002979074130139047,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22405014783143998,
      "num_tokens": 108071140.0,
      "step": 52070
    },
    {
      "entropy": 5.633353614807129,
      "epoch": 4.723784470246734,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029787522448816603,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22373841404914857,
      "num_tokens": 108082599.0,
      "step": 52075
    },
    {
      "entropy": 5.694049119949341,
      "epoch": 4.7242380261248185,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029784303548936183,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2292081505060196,
      "num_tokens": 108092017.0,
      "step": 52080
    },
    {
      "entropy": 5.713012647628784,
      "epoch": 4.724691582002903,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002978108460181579,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22288725078105925,
      "num_tokens": 108103853.0,
      "step": 52085
    },
    {
      "entropy": 5.634487247467041,
      "epoch": 4.725145137880987,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002977786560752198,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2305181309580803,
      "num_tokens": 108114727.0,
      "step": 52090
    },
    {
      "entropy": 5.647755765914917,
      "epoch": 4.725598693759071,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002977464656612133,
      "loss": 4.8238,
      "mean_token_accuracy": 0.23937030583620073,
      "num_tokens": 108125653.0,
      "step": 52095
    },
    {
      "entropy": 5.664167785644532,
      "epoch": 4.7260522496371555,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029771427477680415,
      "loss": 4.8556,
      "mean_token_accuracy": 0.22824369817972184,
      "num_tokens": 108135281.0,
      "step": 52100
    },
    {
      "entropy": 5.680458498001099,
      "epoch": 4.72650580551524,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029768208342265796,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2255040317773819,
      "num_tokens": 108145005.0,
      "step": 52105
    },
    {
      "entropy": 5.7186840057373045,
      "epoch": 4.726959361393323,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002976498915994406,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21808644831180574,
      "num_tokens": 108156477.0,
      "step": 52110
    },
    {
      "entropy": 5.657406044006348,
      "epoch": 4.727412917271408,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00029761769930781757,
      "loss": 4.8906,
      "mean_token_accuracy": 0.22826910018920898,
      "num_tokens": 108166947.0,
      "step": 52115
    },
    {
      "entropy": 5.601456117630005,
      "epoch": 4.727866473149492,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002975855065484548,
      "loss": 4.7483,
      "mean_token_accuracy": 0.2431057244539261,
      "num_tokens": 108175889.0,
      "step": 52120
    },
    {
      "entropy": 5.720250129699707,
      "epoch": 4.728320029027576,
      "grad_norm": 1.34375,
      "learning_rate": 0.000297553313322018,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21897458732128144,
      "num_tokens": 108186276.0,
      "step": 52125
    },
    {
      "entropy": 5.61057858467102,
      "epoch": 4.72877358490566,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029752111962917297,
      "loss": 4.8715,
      "mean_token_accuracy": 0.2258746713399887,
      "num_tokens": 108196960.0,
      "step": 52130
    },
    {
      "entropy": 5.69613299369812,
      "epoch": 4.7292271407837445,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002974889254705854,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22844289988279343,
      "num_tokens": 108207442.0,
      "step": 52135
    },
    {
      "entropy": 5.700296211242676,
      "epoch": 4.729680696661829,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029745673084692116,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23810974061489104,
      "num_tokens": 108217919.0,
      "step": 52140
    },
    {
      "entropy": 5.61967396736145,
      "epoch": 4.730134252539913,
      "grad_norm": 1.40625,
      "learning_rate": 0.000297424535758846,
      "loss": 4.8101,
      "mean_token_accuracy": 0.2320191130042076,
      "num_tokens": 108228106.0,
      "step": 52145
    },
    {
      "entropy": 5.594845867156982,
      "epoch": 4.730587808417997,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029739234020702564,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22701286077499389,
      "num_tokens": 108238398.0,
      "step": 52150
    },
    {
      "entropy": 5.6471456527709964,
      "epoch": 4.731041364296082,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029736014419212605,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22432231456041335,
      "num_tokens": 108248825.0,
      "step": 52155
    },
    {
      "entropy": 5.5855762481689455,
      "epoch": 4.731494920174166,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029732794771481305,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22723645567893982,
      "num_tokens": 108258989.0,
      "step": 52160
    },
    {
      "entropy": 5.558561038970947,
      "epoch": 4.731948476052249,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002972957507757523,
      "loss": 4.8145,
      "mean_token_accuracy": 0.2274627685546875,
      "num_tokens": 108268251.0,
      "step": 52165
    },
    {
      "entropy": 5.636238050460816,
      "epoch": 4.7324020319303335,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029726355337560985,
      "loss": 4.8642,
      "mean_token_accuracy": 0.2265596345067024,
      "num_tokens": 108278290.0,
      "step": 52170
    },
    {
      "entropy": 5.6630393981933596,
      "epoch": 4.732855587808418,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029723135551505133,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2266159325838089,
      "num_tokens": 108288727.0,
      "step": 52175
    },
    {
      "entropy": 5.624604034423828,
      "epoch": 4.733309143686502,
      "grad_norm": 1.25,
      "learning_rate": 0.00029719915719474274,
      "loss": 4.8458,
      "mean_token_accuracy": 0.22840378284454346,
      "num_tokens": 108299781.0,
      "step": 52180
    },
    {
      "entropy": 5.68928017616272,
      "epoch": 4.733762699564586,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002971669584153499,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2257700800895691,
      "num_tokens": 108310703.0,
      "step": 52185
    },
    {
      "entropy": 5.603392601013184,
      "epoch": 4.7342162554426706,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002971347591775388,
      "loss": 4.8128,
      "mean_token_accuracy": 0.22706200778484345,
      "num_tokens": 108321096.0,
      "step": 52190
    },
    {
      "entropy": 5.629251527786255,
      "epoch": 4.734669811320755,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002971025594819751,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22437026351690292,
      "num_tokens": 108331899.0,
      "step": 52195
    },
    {
      "entropy": 5.556859636306763,
      "epoch": 4.735123367198839,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029707035932932495,
      "loss": 4.7325,
      "mean_token_accuracy": 0.2379524141550064,
      "num_tokens": 108342248.0,
      "step": 52200
    },
    {
      "entropy": 5.663289356231689,
      "epoch": 4.735576923076923,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029703815872025413,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22665054947137833,
      "num_tokens": 108355264.0,
      "step": 52205
    },
    {
      "entropy": 5.613776206970215,
      "epoch": 4.736030478955008,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002970059576554285,
      "loss": 4.8454,
      "mean_token_accuracy": 0.222385972738266,
      "num_tokens": 108365929.0,
      "step": 52210
    },
    {
      "entropy": 5.606583547592163,
      "epoch": 4.736484034833092,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002969737561355142,
      "loss": 4.8316,
      "mean_token_accuracy": 0.22769063711166382,
      "num_tokens": 108376874.0,
      "step": 52215
    },
    {
      "entropy": 5.604950714111328,
      "epoch": 4.736937590711175,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002969415541611769,
      "loss": 4.8122,
      "mean_token_accuracy": 0.22499992102384567,
      "num_tokens": 108387277.0,
      "step": 52220
    },
    {
      "entropy": 5.689796257019043,
      "epoch": 4.7373911465892595,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002969093517330827,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22008468508720397,
      "num_tokens": 108399088.0,
      "step": 52225
    },
    {
      "entropy": 5.709333419799805,
      "epoch": 4.737844702467344,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002968771488518975,
      "loss": 4.9516,
      "mean_token_accuracy": 0.21810991764068605,
      "num_tokens": 108409875.0,
      "step": 52230
    },
    {
      "entropy": 5.67483811378479,
      "epoch": 4.738298258345428,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002968449455182874,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22787051945924758,
      "num_tokens": 108421133.0,
      "step": 52235
    },
    {
      "entropy": 5.771256923675537,
      "epoch": 4.738751814223512,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029681274173291817,
      "loss": 5.0528,
      "mean_token_accuracy": 0.21301034241914749,
      "num_tokens": 108433408.0,
      "step": 52240
    },
    {
      "entropy": 5.606645488739014,
      "epoch": 4.739205370101597,
      "grad_norm": 1.375,
      "learning_rate": 0.00029678053749645583,
      "loss": 4.7893,
      "mean_token_accuracy": 0.2333521917462349,
      "num_tokens": 108443783.0,
      "step": 52245
    },
    {
      "entropy": 5.521975469589234,
      "epoch": 4.739658925979681,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002967483328095665,
      "loss": 4.7319,
      "mean_token_accuracy": 0.23259187489748,
      "num_tokens": 108454770.0,
      "step": 52250
    },
    {
      "entropy": 5.612837791442871,
      "epoch": 4.740112481857765,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002967161276729161,
      "loss": 4.828,
      "mean_token_accuracy": 0.23543968796730042,
      "num_tokens": 108464812.0,
      "step": 52255
    },
    {
      "entropy": 5.6790975570678714,
      "epoch": 4.740566037735849,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002966839220871707,
      "loss": 4.8768,
      "mean_token_accuracy": 0.22128035873174667,
      "num_tokens": 108474642.0,
      "step": 52260
    },
    {
      "entropy": 5.605567121505738,
      "epoch": 4.741019593613933,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002966517160529962,
      "loss": 4.8317,
      "mean_token_accuracy": 0.23071384578943252,
      "num_tokens": 108484143.0,
      "step": 52265
    },
    {
      "entropy": 5.715630578994751,
      "epoch": 4.741473149492017,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002966195095710587,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2218109369277954,
      "num_tokens": 108494850.0,
      "step": 52270
    },
    {
      "entropy": 5.671351766586303,
      "epoch": 4.741926705370101,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002965873026420242,
      "loss": 4.9,
      "mean_token_accuracy": 0.2212483450770378,
      "num_tokens": 108505441.0,
      "step": 52275
    },
    {
      "entropy": 5.691192626953125,
      "epoch": 4.742380261248186,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029655509526655885,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21869966238737107,
      "num_tokens": 108517350.0,
      "step": 52280
    },
    {
      "entropy": 5.620677804946899,
      "epoch": 4.74283381712627,
      "grad_norm": 1.25,
      "learning_rate": 0.00029652288744532856,
      "loss": 4.8345,
      "mean_token_accuracy": 0.23160264492034913,
      "num_tokens": 108528472.0,
      "step": 52285
    },
    {
      "entropy": 5.66132698059082,
      "epoch": 4.743287373004354,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002964906791789995,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22255614250898362,
      "num_tokens": 108539834.0,
      "step": 52290
    },
    {
      "entropy": 5.682744455337525,
      "epoch": 4.743740928882438,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002964584704682377,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2275290846824646,
      "num_tokens": 108551478.0,
      "step": 52295
    },
    {
      "entropy": 5.575284194946289,
      "epoch": 4.744194484760523,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029642626131370936,
      "loss": 4.718,
      "mean_token_accuracy": 0.24244055896997452,
      "num_tokens": 108560957.0,
      "step": 52300
    },
    {
      "entropy": 5.658299398422241,
      "epoch": 4.744648040638607,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002963940517160804,
      "loss": 4.8928,
      "mean_token_accuracy": 0.2236693888902664,
      "num_tokens": 108571146.0,
      "step": 52305
    },
    {
      "entropy": 5.631214618682861,
      "epoch": 4.745101596516691,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029636184167601695,
      "loss": 4.8298,
      "mean_token_accuracy": 0.231575645506382,
      "num_tokens": 108581085.0,
      "step": 52310
    },
    {
      "entropy": 5.5773498058319095,
      "epoch": 4.745555152394775,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00029632963119418526,
      "loss": 4.8301,
      "mean_token_accuracy": 0.22852738946676254,
      "num_tokens": 108591103.0,
      "step": 52315
    },
    {
      "entropy": 5.639823913574219,
      "epoch": 4.746008708272859,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029629742027125134,
      "loss": 4.8417,
      "mean_token_accuracy": 0.2250826671719551,
      "num_tokens": 108601433.0,
      "step": 52320
    },
    {
      "entropy": 5.618091058731079,
      "epoch": 4.746462264150943,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029626520890788127,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22619408816099168,
      "num_tokens": 108611728.0,
      "step": 52325
    },
    {
      "entropy": 5.633877992630005,
      "epoch": 4.746915820029027,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029623299710474134,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22156716287136077,
      "num_tokens": 108621521.0,
      "step": 52330
    },
    {
      "entropy": 5.60551905632019,
      "epoch": 4.747369375907112,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029620078486249763,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21905920654535294,
      "num_tokens": 108632921.0,
      "step": 52335
    },
    {
      "entropy": 5.638763523101806,
      "epoch": 4.747822931785196,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002961685721818163,
      "loss": 4.8775,
      "mean_token_accuracy": 0.23296921104192733,
      "num_tokens": 108643244.0,
      "step": 52340
    },
    {
      "entropy": 5.6242602348327635,
      "epoch": 4.74827648766328,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002961363590633634,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22921345382928848,
      "num_tokens": 108653036.0,
      "step": 52345
    },
    {
      "entropy": 5.717230272293091,
      "epoch": 4.748730043541364,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002961041455078053,
      "loss": 4.9,
      "mean_token_accuracy": 0.22478835582733153,
      "num_tokens": 108663317.0,
      "step": 52350
    },
    {
      "entropy": 5.615634965896606,
      "epoch": 4.749183599419449,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002960719315158081,
      "loss": 4.7987,
      "mean_token_accuracy": 0.23187349438667298,
      "num_tokens": 108673657.0,
      "step": 52355
    },
    {
      "entropy": 5.597364902496338,
      "epoch": 4.749637155297533,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002960397170880378,
      "loss": 4.8257,
      "mean_token_accuracy": 0.2216600552201271,
      "num_tokens": 108683831.0,
      "step": 52360
    },
    {
      "entropy": 5.591486215591431,
      "epoch": 4.750090711175617,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029600750222516097,
      "loss": 4.8151,
      "mean_token_accuracy": 0.2277290180325508,
      "num_tokens": 108694062.0,
      "step": 52365
    },
    {
      "entropy": 5.6233659267425535,
      "epoch": 4.7505442670537015,
      "grad_norm": 1.25,
      "learning_rate": 0.00029597528692784365,
      "loss": 4.8318,
      "mean_token_accuracy": 0.2283254459500313,
      "num_tokens": 108704712.0,
      "step": 52370
    },
    {
      "entropy": 5.621436262130738,
      "epoch": 4.750997822931785,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029594307119675195,
      "loss": 4.833,
      "mean_token_accuracy": 0.22577468901872635,
      "num_tokens": 108714741.0,
      "step": 52375
    },
    {
      "entropy": 5.6719067096710205,
      "epoch": 4.751451378809869,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002959108550325523,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22586423009634018,
      "num_tokens": 108724860.0,
      "step": 52380
    },
    {
      "entropy": 5.614187955856323,
      "epoch": 4.751904934687953,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029587863843591076,
      "loss": 4.7853,
      "mean_token_accuracy": 0.2436222866177559,
      "num_tokens": 108735527.0,
      "step": 52385
    },
    {
      "entropy": 5.655112266540527,
      "epoch": 4.752358490566038,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029584642140749375,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21367600709199905,
      "num_tokens": 108745748.0,
      "step": 52390
    },
    {
      "entropy": 5.658301782608032,
      "epoch": 4.752812046444122,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029581420394796733,
      "loss": 4.8843,
      "mean_token_accuracy": 0.2323250263929367,
      "num_tokens": 108756635.0,
      "step": 52395
    },
    {
      "entropy": 5.618615913391113,
      "epoch": 4.753265602322206,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029578198605799796,
      "loss": 4.8197,
      "mean_token_accuracy": 0.22975172847509384,
      "num_tokens": 108765912.0,
      "step": 52400
    },
    {
      "entropy": 5.688010931015015,
      "epoch": 4.7537191582002905,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029574976773825176,
      "loss": 4.8405,
      "mean_token_accuracy": 0.22769033014774323,
      "num_tokens": 108775831.0,
      "step": 52405
    },
    {
      "entropy": 5.652090215682984,
      "epoch": 4.754172714078375,
      "grad_norm": 1.4375,
      "learning_rate": 0.000295717548989395,
      "loss": 4.825,
      "mean_token_accuracy": 0.2263962894678116,
      "num_tokens": 108785937.0,
      "step": 52410
    },
    {
      "entropy": 5.584827613830567,
      "epoch": 4.754626269956459,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002956853298120942,
      "loss": 4.8236,
      "mean_token_accuracy": 0.2291240692138672,
      "num_tokens": 108796289.0,
      "step": 52415
    },
    {
      "entropy": 5.6686793804168705,
      "epoch": 4.755079825834542,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029565311020701544,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2231221467256546,
      "num_tokens": 108806040.0,
      "step": 52420
    },
    {
      "entropy": 5.6595433235168455,
      "epoch": 4.755533381712627,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029562089017482515,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22468622922897338,
      "num_tokens": 108815440.0,
      "step": 52425
    },
    {
      "entropy": 5.715126371383667,
      "epoch": 4.755986937590711,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002955886697161895,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22231387495994567,
      "num_tokens": 108826059.0,
      "step": 52430
    },
    {
      "entropy": 5.678690433502197,
      "epoch": 4.756440493468795,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029555644883177496,
      "loss": 4.834,
      "mean_token_accuracy": 0.23017715513706208,
      "num_tokens": 108836373.0,
      "step": 52435
    },
    {
      "entropy": 5.571122503280639,
      "epoch": 4.756894049346879,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029552422752224787,
      "loss": 4.7736,
      "mean_token_accuracy": 0.23998536467552184,
      "num_tokens": 108847364.0,
      "step": 52440
    },
    {
      "entropy": 5.599294233322143,
      "epoch": 4.757347605224964,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029549200578827444,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22038956582546235,
      "num_tokens": 108856893.0,
      "step": 52445
    },
    {
      "entropy": 5.665643215179443,
      "epoch": 4.757801161103048,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002954597836305212,
      "loss": 4.7963,
      "mean_token_accuracy": 0.23858331143856049,
      "num_tokens": 108867218.0,
      "step": 52450
    },
    {
      "entropy": 5.636919736862183,
      "epoch": 4.758254716981132,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029542756104965434,
      "loss": 4.8448,
      "mean_token_accuracy": 0.22345684319734574,
      "num_tokens": 108876938.0,
      "step": 52455
    },
    {
      "entropy": 5.6562642574310305,
      "epoch": 4.7587082728592165,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029539533804634027,
      "loss": 4.8352,
      "mean_token_accuracy": 0.2263913244009018,
      "num_tokens": 108887831.0,
      "step": 52460
    },
    {
      "entropy": 5.617428112030029,
      "epoch": 4.759161828737301,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029536311462124553,
      "loss": 4.8209,
      "mean_token_accuracy": 0.23740432411432266,
      "num_tokens": 108897634.0,
      "step": 52465
    },
    {
      "entropy": 5.603098392486572,
      "epoch": 4.759615384615385,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002953308907750363,
      "loss": 4.8138,
      "mean_token_accuracy": 0.22858071625232695,
      "num_tokens": 108907976.0,
      "step": 52470
    },
    {
      "entropy": 5.609667205810547,
      "epoch": 4.760068940493468,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002952986665083792,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22344138771295546,
      "num_tokens": 108918755.0,
      "step": 52475
    },
    {
      "entropy": 5.630168914794922,
      "epoch": 4.760522496371553,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002952664418219404,
      "loss": 4.86,
      "mean_token_accuracy": 0.23377442061901094,
      "num_tokens": 108928349.0,
      "step": 52480
    },
    {
      "entropy": 5.64237642288208,
      "epoch": 4.760976052249637,
      "grad_norm": 1.25,
      "learning_rate": 0.00029523421671638644,
      "loss": 4.9178,
      "mean_token_accuracy": 0.21967661678791045,
      "num_tokens": 108938429.0,
      "step": 52485
    },
    {
      "entropy": 5.680794095993042,
      "epoch": 4.761429608127721,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002952019911923837,
      "loss": 4.861,
      "mean_token_accuracy": 0.22872537970542908,
      "num_tokens": 108948378.0,
      "step": 52490
    },
    {
      "entropy": 5.604490852355957,
      "epoch": 4.7618831640058055,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029516976525059864,
      "loss": 4.8009,
      "mean_token_accuracy": 0.23161621242761612,
      "num_tokens": 108958667.0,
      "step": 52495
    },
    {
      "entropy": 5.708456802368164,
      "epoch": 4.76233671988389,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002951375388916977,
      "loss": 4.9266,
      "mean_token_accuracy": 0.21775925159454346,
      "num_tokens": 108968955.0,
      "step": 52500
    },
    {
      "entropy": 5.672987747192383,
      "epoch": 4.762790275761974,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002951053121163474,
      "loss": 4.808,
      "mean_token_accuracy": 0.2305819123983383,
      "num_tokens": 108979565.0,
      "step": 52505
    },
    {
      "entropy": 5.645192050933838,
      "epoch": 4.763243831640058,
      "grad_norm": 1.359375,
      "learning_rate": 0.000295073084925214,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2242574080824852,
      "num_tokens": 108989547.0,
      "step": 52510
    },
    {
      "entropy": 5.612680530548095,
      "epoch": 4.7636973875181425,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029504085731896413,
      "loss": 4.8449,
      "mean_token_accuracy": 0.22904817759990692,
      "num_tokens": 109000286.0,
      "step": 52515
    },
    {
      "entropy": 5.664691066741943,
      "epoch": 4.764150943396227,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002950086292982642,
      "loss": 4.8014,
      "mean_token_accuracy": 0.23575116395950318,
      "num_tokens": 109010676.0,
      "step": 52520
    },
    {
      "entropy": 5.710342931747436,
      "epoch": 4.764604499274311,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029497640086378077,
      "loss": 4.939,
      "mean_token_accuracy": 0.21703298836946489,
      "num_tokens": 109020800.0,
      "step": 52525
    },
    {
      "entropy": 5.676192808151245,
      "epoch": 4.7650580551523944,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029494417201618027,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2198370188474655,
      "num_tokens": 109031158.0,
      "step": 52530
    },
    {
      "entropy": 5.596923875808716,
      "epoch": 4.765511611030479,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002949119427561292,
      "loss": 4.8142,
      "mean_token_accuracy": 0.23129542469978331,
      "num_tokens": 109041664.0,
      "step": 52535
    },
    {
      "entropy": 5.633503389358521,
      "epoch": 4.765965166908563,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029487971308429415,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23392236083745957,
      "num_tokens": 109052633.0,
      "step": 52540
    },
    {
      "entropy": 5.6323254108428955,
      "epoch": 4.766418722786647,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002948474830013414,
      "loss": 4.8541,
      "mean_token_accuracy": 0.22733922451734542,
      "num_tokens": 109062687.0,
      "step": 52545
    },
    {
      "entropy": 5.6883176326751705,
      "epoch": 4.7668722786647315,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029481525250793777,
      "loss": 4.8864,
      "mean_token_accuracy": 0.23309983760118486,
      "num_tokens": 109073348.0,
      "step": 52550
    },
    {
      "entropy": 5.5975922584533695,
      "epoch": 4.767325834542816,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002947830216047497,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22087957561016083,
      "num_tokens": 109083276.0,
      "step": 52555
    },
    {
      "entropy": 5.594065952301025,
      "epoch": 4.7677793904209,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029475079029244356,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2287531539797783,
      "num_tokens": 109094342.0,
      "step": 52560
    },
    {
      "entropy": 5.63740291595459,
      "epoch": 4.768232946298984,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029471855857168616,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22886939942836762,
      "num_tokens": 109105178.0,
      "step": 52565
    },
    {
      "entropy": 5.708528757095337,
      "epoch": 4.768686502177069,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002946863264431439,
      "loss": 4.96,
      "mean_token_accuracy": 0.2153998762369156,
      "num_tokens": 109116205.0,
      "step": 52570
    },
    {
      "entropy": 5.624877452850342,
      "epoch": 4.769140058055152,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002946540939074834,
      "loss": 4.8425,
      "mean_token_accuracy": 0.2220785692334175,
      "num_tokens": 109126299.0,
      "step": 52575
    },
    {
      "entropy": 5.661425447463989,
      "epoch": 4.769593613933236,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029462186096537116,
      "loss": 4.946,
      "mean_token_accuracy": 0.22503869235515594,
      "num_tokens": 109137675.0,
      "step": 52580
    },
    {
      "entropy": 5.661192083358765,
      "epoch": 4.7700471698113205,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029458962761747394,
      "loss": 4.8313,
      "mean_token_accuracy": 0.22939475774765014,
      "num_tokens": 109147716.0,
      "step": 52585
    },
    {
      "entropy": 5.644068670272827,
      "epoch": 4.770500725689405,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002945573938644582,
      "loss": 4.8441,
      "mean_token_accuracy": 0.22903202921152116,
      "num_tokens": 109159211.0,
      "step": 52590
    },
    {
      "entropy": 5.64030876159668,
      "epoch": 4.770954281567489,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029452515970699054,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21943735033273698,
      "num_tokens": 109169490.0,
      "step": 52595
    },
    {
      "entropy": 5.687409257888794,
      "epoch": 4.771407837445573,
      "grad_norm": 1.375,
      "learning_rate": 0.00029449292514573767,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21967633217573165,
      "num_tokens": 109179687.0,
      "step": 52600
    },
    {
      "entropy": 5.633784818649292,
      "epoch": 4.7718613933236576,
      "grad_norm": 1.5,
      "learning_rate": 0.00029446069018136613,
      "loss": 4.8385,
      "mean_token_accuracy": 0.22660400867462158,
      "num_tokens": 109189415.0,
      "step": 52605
    },
    {
      "entropy": 5.588267850875854,
      "epoch": 4.772314949201742,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002944284548145424,
      "loss": 4.7364,
      "mean_token_accuracy": 0.23210244178771972,
      "num_tokens": 109199281.0,
      "step": 52610
    },
    {
      "entropy": 5.615257644653321,
      "epoch": 4.772768505079826,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002943962190459335,
      "loss": 4.8257,
      "mean_token_accuracy": 0.22793096899986268,
      "num_tokens": 109208347.0,
      "step": 52615
    },
    {
      "entropy": 5.587856960296631,
      "epoch": 4.77322206095791,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029436398287620573,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22335030734539033,
      "num_tokens": 109218026.0,
      "step": 52620
    },
    {
      "entropy": 5.588596200942993,
      "epoch": 4.773675616835995,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002943317463060259,
      "loss": 4.8293,
      "mean_token_accuracy": 0.22319174259901048,
      "num_tokens": 109229457.0,
      "step": 52625
    },
    {
      "entropy": 5.63396520614624,
      "epoch": 4.774129172714078,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002942995093360606,
      "loss": 4.8164,
      "mean_token_accuracy": 0.22723015993833542,
      "num_tokens": 109239504.0,
      "step": 52630
    },
    {
      "entropy": 5.576156854629517,
      "epoch": 4.774582728592162,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002942672719669765,
      "loss": 4.7457,
      "mean_token_accuracy": 0.23487858921289445,
      "num_tokens": 109249745.0,
      "step": 52635
    },
    {
      "entropy": 5.589773654937744,
      "epoch": 4.7750362844702465,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002942350341994404,
      "loss": 4.8135,
      "mean_token_accuracy": 0.2322341486811638,
      "num_tokens": 109259767.0,
      "step": 52640
    },
    {
      "entropy": 5.754535484313965,
      "epoch": 4.775489840348331,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029420279603411875,
      "loss": 5.0165,
      "mean_token_accuracy": 0.22071036398410798,
      "num_tokens": 109269611.0,
      "step": 52645
    },
    {
      "entropy": 5.758963298797608,
      "epoch": 4.775943396226415,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002941705574716785,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2198594555258751,
      "num_tokens": 109281396.0,
      "step": 52650
    },
    {
      "entropy": 5.69974193572998,
      "epoch": 4.776396952104499,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002941383185127862,
      "loss": 4.8143,
      "mean_token_accuracy": 0.22679316252470016,
      "num_tokens": 109292346.0,
      "step": 52655
    },
    {
      "entropy": 5.730378246307373,
      "epoch": 4.776850507982584,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029410607915810854,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21880850195884705,
      "num_tokens": 109302966.0,
      "step": 52660
    },
    {
      "entropy": 5.557135391235351,
      "epoch": 4.777304063860668,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002940738394083124,
      "loss": 4.8042,
      "mean_token_accuracy": 0.22933248430490494,
      "num_tokens": 109312558.0,
      "step": 52665
    },
    {
      "entropy": 5.661647653579712,
      "epoch": 4.777757619738752,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00029404159926406436,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22697875052690505,
      "num_tokens": 109322364.0,
      "step": 52670
    },
    {
      "entropy": 5.722963619232178,
      "epoch": 4.7782111756168355,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002940093587260312,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2255360946059227,
      "num_tokens": 109332449.0,
      "step": 52675
    },
    {
      "entropy": 5.692411041259765,
      "epoch": 4.778664731494921,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029397711779487954,
      "loss": 4.822,
      "mean_token_accuracy": 0.22846415191888808,
      "num_tokens": 109343262.0,
      "step": 52680
    },
    {
      "entropy": 5.625509357452392,
      "epoch": 4.779118287373004,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029394487647127645,
      "loss": 4.8185,
      "mean_token_accuracy": 0.22888773679733276,
      "num_tokens": 109352951.0,
      "step": 52685
    },
    {
      "entropy": 5.586233711242675,
      "epoch": 4.779571843251088,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029391263475588844,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2228041797876358,
      "num_tokens": 109363913.0,
      "step": 52690
    },
    {
      "entropy": 5.616233587265015,
      "epoch": 4.780025399129173,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029388039264938224,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2242326870560646,
      "num_tokens": 109374405.0,
      "step": 52695
    },
    {
      "entropy": 5.726517915725708,
      "epoch": 4.780478955007257,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002938481501524247,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2200401321053505,
      "num_tokens": 109385696.0,
      "step": 52700
    },
    {
      "entropy": 5.6907796382904055,
      "epoch": 4.780932510885341,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00029381590726568267,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22477795630693437,
      "num_tokens": 109395061.0,
      "step": 52705
    },
    {
      "entropy": 5.647987651824951,
      "epoch": 4.781386066763425,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029378366398982286,
      "loss": 4.8668,
      "mean_token_accuracy": 0.2233182057738304,
      "num_tokens": 109406152.0,
      "step": 52710
    },
    {
      "entropy": 5.7147167205810545,
      "epoch": 4.78183962264151,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029375142032551216,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22091188728809358,
      "num_tokens": 109416596.0,
      "step": 52715
    },
    {
      "entropy": 5.740494871139527,
      "epoch": 4.782293178519594,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029371917627341726,
      "loss": 4.927,
      "mean_token_accuracy": 0.2245527222752571,
      "num_tokens": 109426169.0,
      "step": 52720
    },
    {
      "entropy": 5.68250412940979,
      "epoch": 4.782746734397678,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000293686931834205,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2283185362815857,
      "num_tokens": 109436359.0,
      "step": 52725
    },
    {
      "entropy": 5.646133947372436,
      "epoch": 4.7832002902757615,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002936546870085422,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22126978784799575,
      "num_tokens": 109446050.0,
      "step": 52730
    },
    {
      "entropy": 5.678741693496704,
      "epoch": 4.783653846153846,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029362244179709576,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23047959059476852,
      "num_tokens": 109457124.0,
      "step": 52735
    },
    {
      "entropy": 5.586958599090576,
      "epoch": 4.78410740203193,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029359019620053253,
      "loss": 4.7762,
      "mean_token_accuracy": 0.23238855004310607,
      "num_tokens": 109467642.0,
      "step": 52740
    },
    {
      "entropy": 5.686911344528198,
      "epoch": 4.784560957910014,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002935579502195192,
      "loss": 4.8981,
      "mean_token_accuracy": 0.21863505989313126,
      "num_tokens": 109477877.0,
      "step": 52745
    },
    {
      "entropy": 5.657322072982788,
      "epoch": 4.785014513788099,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029352570385472276,
      "loss": 4.8389,
      "mean_token_accuracy": 0.22490146309137343,
      "num_tokens": 109489302.0,
      "step": 52750
    },
    {
      "entropy": 5.644135427474976,
      "epoch": 4.785468069666183,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002934934571068101,
      "loss": 4.8396,
      "mean_token_accuracy": 0.22730441838502885,
      "num_tokens": 109500023.0,
      "step": 52755
    },
    {
      "entropy": 5.60171365737915,
      "epoch": 4.785921625544267,
      "grad_norm": 1.28125,
      "learning_rate": 0.000293461209976448,
      "loss": 4.8723,
      "mean_token_accuracy": 0.22833799421787263,
      "num_tokens": 109510588.0,
      "step": 52760
    },
    {
      "entropy": 5.701880645751953,
      "epoch": 4.786375181422351,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002934289624643034,
      "loss": 4.8959,
      "mean_token_accuracy": 0.22276590019464493,
      "num_tokens": 109520162.0,
      "step": 52765
    },
    {
      "entropy": 5.662490940093994,
      "epoch": 4.786828737300436,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002933967145710431,
      "loss": 4.851,
      "mean_token_accuracy": 0.22318645268678666,
      "num_tokens": 109530155.0,
      "step": 52770
    },
    {
      "entropy": 5.621389865875244,
      "epoch": 4.78728229317852,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002933644662973341,
      "loss": 4.8987,
      "mean_token_accuracy": 0.2277026280760765,
      "num_tokens": 109539897.0,
      "step": 52775
    },
    {
      "entropy": 5.595298528671265,
      "epoch": 4.787735849056604,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029333221764384326,
      "loss": 4.849,
      "mean_token_accuracy": 0.2212102323770523,
      "num_tokens": 109549995.0,
      "step": 52780
    },
    {
      "entropy": 5.654567289352417,
      "epoch": 4.788189404934688,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029329996861123745,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22533482313156128,
      "num_tokens": 109560446.0,
      "step": 52785
    },
    {
      "entropy": 5.607658958435058,
      "epoch": 4.788642960812772,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029326771920018367,
      "loss": 4.8325,
      "mean_token_accuracy": 0.22632743269205094,
      "num_tokens": 109570864.0,
      "step": 52790
    },
    {
      "entropy": 5.735541725158692,
      "epoch": 4.789096516690856,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029323546941134875,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22914797216653823,
      "num_tokens": 109581015.0,
      "step": 52795
    },
    {
      "entropy": 5.6438463687896725,
      "epoch": 4.78955007256894,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029320321924539974,
      "loss": 4.865,
      "mean_token_accuracy": 0.22209956496953964,
      "num_tokens": 109591695.0,
      "step": 52800
    },
    {
      "entropy": 5.578674936294556,
      "epoch": 4.790003628447025,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002931709687030035,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23232684284448624,
      "num_tokens": 109602274.0,
      "step": 52805
    },
    {
      "entropy": 5.688881921768188,
      "epoch": 4.790457184325109,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029313871778482693,
      "loss": 4.871,
      "mean_token_accuracy": 0.23010208308696747,
      "num_tokens": 109612995.0,
      "step": 52810
    },
    {
      "entropy": 5.6651569366455075,
      "epoch": 4.790910740203193,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002931064664915371,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22488640993833542,
      "num_tokens": 109623133.0,
      "step": 52815
    },
    {
      "entropy": 5.584497308731079,
      "epoch": 4.7913642960812775,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000293074214823801,
      "loss": 4.8159,
      "mean_token_accuracy": 0.22636962532997132,
      "num_tokens": 109634312.0,
      "step": 52820
    },
    {
      "entropy": 5.6789247512817385,
      "epoch": 4.791817851959362,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029304196278228544,
      "loss": 4.8794,
      "mean_token_accuracy": 0.21459644585847854,
      "num_tokens": 109644333.0,
      "step": 52825
    },
    {
      "entropy": 5.616401338577271,
      "epoch": 4.792271407837445,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002930097103676575,
      "loss": 4.7708,
      "mean_token_accuracy": 0.23262296319007875,
      "num_tokens": 109654730.0,
      "step": 52830
    },
    {
      "entropy": 5.589736461639404,
      "epoch": 4.79272496371553,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029297745758058414,
      "loss": 4.8119,
      "mean_token_accuracy": 0.22892464250326156,
      "num_tokens": 109664867.0,
      "step": 52835
    },
    {
      "entropy": 5.6731462478637695,
      "epoch": 4.793178519593614,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029294520442173244,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22929418236017227,
      "num_tokens": 109675090.0,
      "step": 52840
    },
    {
      "entropy": 5.610694503784179,
      "epoch": 4.793632075471698,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002929129508917693,
      "loss": 4.8027,
      "mean_token_accuracy": 0.23342251628637314,
      "num_tokens": 109685046.0,
      "step": 52845
    },
    {
      "entropy": 5.749180030822754,
      "epoch": 4.794085631349782,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002928806969913618,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22772050648927689,
      "num_tokens": 109696240.0,
      "step": 52850
    },
    {
      "entropy": 5.643613719940186,
      "epoch": 4.794539187227866,
      "grad_norm": 1.375,
      "learning_rate": 0.0002928484427211769,
      "loss": 4.8658,
      "mean_token_accuracy": 0.2210042417049408,
      "num_tokens": 109707676.0,
      "step": 52855
    },
    {
      "entropy": 5.719730567932129,
      "epoch": 4.794992743105951,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002928161880818817,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22191641926765443,
      "num_tokens": 109718168.0,
      "step": 52860
    },
    {
      "entropy": 5.606037950515747,
      "epoch": 4.795446298984035,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002927839330741431,
      "loss": 4.8023,
      "mean_token_accuracy": 0.228772135078907,
      "num_tokens": 109728624.0,
      "step": 52865
    },
    {
      "entropy": 5.739957571029663,
      "epoch": 4.795899854862119,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029275167769862834,
      "loss": 4.9481,
      "mean_token_accuracy": 0.21613523811101915,
      "num_tokens": 109738301.0,
      "step": 52870
    },
    {
      "entropy": 5.684234857559204,
      "epoch": 4.7963534107402035,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002927194219560043,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21721005588769912,
      "num_tokens": 109749269.0,
      "step": 52875
    },
    {
      "entropy": 5.6338677406311035,
      "epoch": 4.796806966618288,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029268716584693804,
      "loss": 4.8271,
      "mean_token_accuracy": 0.22504261285066604,
      "num_tokens": 109758335.0,
      "step": 52880
    },
    {
      "entropy": 5.642784070968628,
      "epoch": 4.797260522496371,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002926549093720968,
      "loss": 4.803,
      "mean_token_accuracy": 0.23389253318309783,
      "num_tokens": 109768848.0,
      "step": 52885
    },
    {
      "entropy": 5.611158227920532,
      "epoch": 4.797714078374455,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029262265253214754,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22540179640054703,
      "num_tokens": 109779331.0,
      "step": 52890
    },
    {
      "entropy": 5.672042322158814,
      "epoch": 4.79816763425254,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029259039532775725,
      "loss": 4.8444,
      "mean_token_accuracy": 0.22843475341796876,
      "num_tokens": 109789331.0,
      "step": 52895
    },
    {
      "entropy": 5.6949248790740965,
      "epoch": 4.798621190130624,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002925581377595931,
      "loss": 4.8273,
      "mean_token_accuracy": 0.2307179570198059,
      "num_tokens": 109798630.0,
      "step": 52900
    },
    {
      "entropy": 5.636219453811646,
      "epoch": 4.799074746008708,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029252587982832225,
      "loss": 4.8567,
      "mean_token_accuracy": 0.23180412948131562,
      "num_tokens": 109808501.0,
      "step": 52905
    },
    {
      "entropy": 5.624536037445068,
      "epoch": 4.7995283018867925,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002924936215346117,
      "loss": 4.8406,
      "mean_token_accuracy": 0.23255689591169357,
      "num_tokens": 109818847.0,
      "step": 52910
    },
    {
      "entropy": 5.6615558624267575,
      "epoch": 4.799981857764877,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029246136287912865,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22287191152572633,
      "num_tokens": 109830017.0,
      "step": 52915
    },
    {
      "entropy": 5.6327965259552,
      "epoch": 4.800435413642961,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002924291038625402,
      "loss": 4.8718,
      "mean_token_accuracy": 0.22852483689785003,
      "num_tokens": 109839795.0,
      "step": 52920
    },
    {
      "entropy": 5.659586000442505,
      "epoch": 4.800888969521045,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029239684448551336,
      "loss": 4.82,
      "mean_token_accuracy": 0.23215188831090927,
      "num_tokens": 109850202.0,
      "step": 52925
    },
    {
      "entropy": 5.627269077301025,
      "epoch": 4.8013425253991295,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002923645847487153,
      "loss": 4.8792,
      "mean_token_accuracy": 0.22786654978990556,
      "num_tokens": 109860625.0,
      "step": 52930
    },
    {
      "entropy": 5.652566432952881,
      "epoch": 4.801796081277214,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002923323246528133,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23703712821006775,
      "num_tokens": 109871241.0,
      "step": 52935
    },
    {
      "entropy": 5.67678804397583,
      "epoch": 4.802249637155297,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029230006419847434,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22326949387788772,
      "num_tokens": 109881627.0,
      "step": 52940
    },
    {
      "entropy": 5.680752754211426,
      "epoch": 4.8027031930333814,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002922678033863656,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22708907276391982,
      "num_tokens": 109893247.0,
      "step": 52945
    },
    {
      "entropy": 5.6523042678833,
      "epoch": 4.803156748911466,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002922355422171544,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22537523806095122,
      "num_tokens": 109904112.0,
      "step": 52950
    },
    {
      "entropy": 5.628271102905273,
      "epoch": 4.80361030478955,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029220328069150774,
      "loss": 4.7525,
      "mean_token_accuracy": 0.23634327650070192,
      "num_tokens": 109913854.0,
      "step": 52955
    },
    {
      "entropy": 5.724130582809448,
      "epoch": 4.804063860667634,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029217101881009284,
      "loss": 4.9162,
      "mean_token_accuracy": 0.2245335251092911,
      "num_tokens": 109923947.0,
      "step": 52960
    },
    {
      "entropy": 5.647067928314209,
      "epoch": 4.8045174165457185,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002921387565735769,
      "loss": 4.8456,
      "mean_token_accuracy": 0.22952244579792022,
      "num_tokens": 109934590.0,
      "step": 52965
    },
    {
      "entropy": 5.656402206420898,
      "epoch": 4.804970972423803,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002921064939826272,
      "loss": 4.8853,
      "mean_token_accuracy": 0.2241079479455948,
      "num_tokens": 109944224.0,
      "step": 52970
    },
    {
      "entropy": 5.651070690155029,
      "epoch": 4.805424528301887,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002920742310379108,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22470937222242354,
      "num_tokens": 109954414.0,
      "step": 52975
    },
    {
      "entropy": 5.716910934448242,
      "epoch": 4.805878084179971,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029204196774009476,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22939649522304534,
      "num_tokens": 109964304.0,
      "step": 52980
    },
    {
      "entropy": 5.694309234619141,
      "epoch": 4.806331640058055,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029200970408984673,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21731321960687638,
      "num_tokens": 109975401.0,
      "step": 52985
    },
    {
      "entropy": 5.710887670516968,
      "epoch": 4.806785195936139,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002919774400878336,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22552456557750702,
      "num_tokens": 109985964.0,
      "step": 52990
    },
    {
      "entropy": 5.718891954421997,
      "epoch": 4.807238751814223,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002919451757347226,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22393586188554765,
      "num_tokens": 109995476.0,
      "step": 52995
    },
    {
      "entropy": 5.668258762359619,
      "epoch": 4.8076923076923075,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002919129110311812,
      "loss": 4.7986,
      "mean_token_accuracy": 0.23034249395132064,
      "num_tokens": 110005238.0,
      "step": 53000
    },
    {
      "entropy": 5.672264099121094,
      "epoch": 4.808145863570392,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002918806459778764,
      "loss": 4.954,
      "mean_token_accuracy": 0.2096040815114975,
      "num_tokens": 110016120.0,
      "step": 53005
    },
    {
      "entropy": 5.653968715667725,
      "epoch": 4.808599419448476,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002918483805754756,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22260472625494004,
      "num_tokens": 110026739.0,
      "step": 53010
    },
    {
      "entropy": 5.735321283340454,
      "epoch": 4.80905297532656,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002918161148246459,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22493019551038743,
      "num_tokens": 110037491.0,
      "step": 53015
    },
    {
      "entropy": 5.677459287643432,
      "epoch": 4.8095065312046446,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029178384872605465,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22524028420448303,
      "num_tokens": 110048550.0,
      "step": 53020
    },
    {
      "entropy": 5.659679794311524,
      "epoch": 4.809960087082729,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029175158228036924,
      "loss": 4.875,
      "mean_token_accuracy": 0.2251047223806381,
      "num_tokens": 110058472.0,
      "step": 53025
    },
    {
      "entropy": 5.635154581069946,
      "epoch": 4.810413642960813,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002917193154882568,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23143436163663864,
      "num_tokens": 110070766.0,
      "step": 53030
    },
    {
      "entropy": 5.690961742401123,
      "epoch": 4.810867198838897,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002916870483503847,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22758369892835617,
      "num_tokens": 110081739.0,
      "step": 53035
    },
    {
      "entropy": 5.661534929275513,
      "epoch": 4.811320754716981,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029165478086742006,
      "loss": 4.8568,
      "mean_token_accuracy": 0.22312567979097367,
      "num_tokens": 110091075.0,
      "step": 53040
    },
    {
      "entropy": 5.64955472946167,
      "epoch": 4.811774310595065,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029162251304003044,
      "loss": 4.8298,
      "mean_token_accuracy": 0.23109465837478638,
      "num_tokens": 110100530.0,
      "step": 53045
    },
    {
      "entropy": 5.575726413726807,
      "epoch": 4.812227866473149,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002915902448688829,
      "loss": 4.8317,
      "mean_token_accuracy": 0.22661267220973969,
      "num_tokens": 110110866.0,
      "step": 53050
    },
    {
      "entropy": 5.620332527160644,
      "epoch": 4.8126814223512335,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029155797635464493,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2304461568593979,
      "num_tokens": 110121344.0,
      "step": 53055
    },
    {
      "entropy": 5.579096889495849,
      "epoch": 4.813134978229318,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002915257074979838,
      "loss": 4.7919,
      "mean_token_accuracy": 0.22958470582962037,
      "num_tokens": 110130529.0,
      "step": 53060
    },
    {
      "entropy": 5.629188013076782,
      "epoch": 4.813588534107402,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002914934382995668,
      "loss": 4.8154,
      "mean_token_accuracy": 0.2263242557644844,
      "num_tokens": 110141536.0,
      "step": 53065
    },
    {
      "entropy": 5.69946141242981,
      "epoch": 4.814042089985486,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002914611687600613,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22829544842243193,
      "num_tokens": 110152364.0,
      "step": 53070
    },
    {
      "entropy": 5.641432285308838,
      "epoch": 4.814495645863571,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002914288988801347,
      "loss": 4.8794,
      "mean_token_accuracy": 0.23028891533613205,
      "num_tokens": 110163800.0,
      "step": 53075
    },
    {
      "entropy": 5.5953834533691404,
      "epoch": 4.814949201741655,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002913966286604541,
      "loss": 4.7856,
      "mean_token_accuracy": 0.22586278468370438,
      "num_tokens": 110174590.0,
      "step": 53080
    },
    {
      "entropy": 5.676026296615601,
      "epoch": 4.815402757619739,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029136435810168726,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2233936756849289,
      "num_tokens": 110185142.0,
      "step": 53085
    },
    {
      "entropy": 5.5707542419433596,
      "epoch": 4.815856313497823,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002913320872045012,
      "loss": 4.7358,
      "mean_token_accuracy": 0.24365141838788987,
      "num_tokens": 110194013.0,
      "step": 53090
    },
    {
      "entropy": 5.630069255828857,
      "epoch": 4.816309869375907,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029129981596956353,
      "loss": 4.8225,
      "mean_token_accuracy": 0.23806393891572952,
      "num_tokens": 110204087.0,
      "step": 53095
    },
    {
      "entropy": 5.6932535648345945,
      "epoch": 4.816763425253991,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029126754439754137,
      "loss": 4.871,
      "mean_token_accuracy": 0.22689005732536316,
      "num_tokens": 110214066.0,
      "step": 53100
    },
    {
      "entropy": 5.64039249420166,
      "epoch": 4.817216981132075,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002912352724891023,
      "loss": 4.8547,
      "mean_token_accuracy": 0.23029600977897643,
      "num_tokens": 110224378.0,
      "step": 53105
    },
    {
      "entropy": 5.601255083084107,
      "epoch": 4.81767053701016,
      "grad_norm": 1.375,
      "learning_rate": 0.0002912030002449137,
      "loss": 4.7646,
      "mean_token_accuracy": 0.23258891254663466,
      "num_tokens": 110235493.0,
      "step": 53110
    },
    {
      "entropy": 5.632140111923218,
      "epoch": 4.818124092888244,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029117072766564285,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21587228924036025,
      "num_tokens": 110246953.0,
      "step": 53115
    },
    {
      "entropy": 5.606414318084717,
      "epoch": 4.818577648766328,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002911384547519573,
      "loss": 4.7601,
      "mean_token_accuracy": 0.23433198034763336,
      "num_tokens": 110257387.0,
      "step": 53120
    },
    {
      "entropy": 5.647055053710938,
      "epoch": 4.819031204644412,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029110618150452437,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22693479657173157,
      "num_tokens": 110268152.0,
      "step": 53125
    },
    {
      "entropy": 5.675388526916504,
      "epoch": 4.819484760522497,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00029107390792401145,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22676258981227876,
      "num_tokens": 110277933.0,
      "step": 53130
    },
    {
      "entropy": 5.59682936668396,
      "epoch": 4.819938316400581,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002910416340110861,
      "loss": 4.7749,
      "mean_token_accuracy": 0.23103955239057541,
      "num_tokens": 110288219.0,
      "step": 53135
    },
    {
      "entropy": 5.64211254119873,
      "epoch": 4.820391872278664,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002910093597664156,
      "loss": 4.8665,
      "mean_token_accuracy": 0.2302780494093895,
      "num_tokens": 110299209.0,
      "step": 53140
    },
    {
      "entropy": 5.615989542007446,
      "epoch": 4.8208454281567485,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002909770851906675,
      "loss": 4.8562,
      "mean_token_accuracy": 0.2281659036874771,
      "num_tokens": 110309482.0,
      "step": 53145
    },
    {
      "entropy": 5.6454430103302,
      "epoch": 4.821298984034833,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029094481028450926,
      "loss": 4.8266,
      "mean_token_accuracy": 0.22884132713079453,
      "num_tokens": 110319499.0,
      "step": 53150
    },
    {
      "entropy": 5.629137086868286,
      "epoch": 4.821752539912917,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002909125350486082,
      "loss": 4.7801,
      "mean_token_accuracy": 0.2349223241209984,
      "num_tokens": 110329453.0,
      "step": 53155
    },
    {
      "entropy": 5.605965089797974,
      "epoch": 4.822206095791001,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002908802594836319,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22566077560186387,
      "num_tokens": 110339343.0,
      "step": 53160
    },
    {
      "entropy": 5.658988904953003,
      "epoch": 4.822659651669086,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002908479835902477,
      "loss": 4.9364,
      "mean_token_accuracy": 0.22918891459703444,
      "num_tokens": 110349101.0,
      "step": 53165
    },
    {
      "entropy": 5.6077546119689945,
      "epoch": 4.82311320754717,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002908157073691233,
      "loss": 4.8382,
      "mean_token_accuracy": 0.22752186357975007,
      "num_tokens": 110359377.0,
      "step": 53170
    },
    {
      "entropy": 5.632033443450927,
      "epoch": 4.823566763425254,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029078343082092594,
      "loss": 4.8228,
      "mean_token_accuracy": 0.23501441478729249,
      "num_tokens": 110370132.0,
      "step": 53175
    },
    {
      "entropy": 5.629738712310791,
      "epoch": 4.824020319303338,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029075115394632325,
      "loss": 4.8531,
      "mean_token_accuracy": 0.22202265858650208,
      "num_tokens": 110381046.0,
      "step": 53180
    },
    {
      "entropy": 5.707213115692139,
      "epoch": 4.824473875181423,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029071887674598265,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22030968070030213,
      "num_tokens": 110392309.0,
      "step": 53185
    },
    {
      "entropy": 5.739593172073365,
      "epoch": 4.824927431059507,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002906865992205717,
      "loss": 4.9803,
      "mean_token_accuracy": 0.2210324689745903,
      "num_tokens": 110402464.0,
      "step": 53190
    },
    {
      "entropy": 5.5837983131408695,
      "epoch": 4.82538098693759,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002906543213707579,
      "loss": 4.7567,
      "mean_token_accuracy": 0.23068068027496338,
      "num_tokens": 110412272.0,
      "step": 53195
    },
    {
      "entropy": 5.62268443107605,
      "epoch": 4.825834542815675,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029062204319720865,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2244965136051178,
      "num_tokens": 110423425.0,
      "step": 53200
    },
    {
      "entropy": 5.624871110916137,
      "epoch": 4.826288098693759,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002905897647005917,
      "loss": 4.8744,
      "mean_token_accuracy": 0.2204594284296036,
      "num_tokens": 110434469.0,
      "step": 53205
    },
    {
      "entropy": 5.633252859115601,
      "epoch": 4.826741654571843,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002905574858815744,
      "loss": 4.8281,
      "mean_token_accuracy": 0.23326743841171266,
      "num_tokens": 110443777.0,
      "step": 53210
    },
    {
      "entropy": 5.653081798553467,
      "epoch": 4.827195210449927,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002905252067408242,
      "loss": 4.8876,
      "mean_token_accuracy": 0.214824116230011,
      "num_tokens": 110454481.0,
      "step": 53215
    },
    {
      "entropy": 5.662503099441528,
      "epoch": 4.827648766328012,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002904929272790089,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22359896898269654,
      "num_tokens": 110464789.0,
      "step": 53220
    },
    {
      "entropy": 5.578028202056885,
      "epoch": 4.828102322206096,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002904606474967959,
      "loss": 4.7404,
      "mean_token_accuracy": 0.2374787762761116,
      "num_tokens": 110474276.0,
      "step": 53225
    },
    {
      "entropy": 5.672610330581665,
      "epoch": 4.82855587808418,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029042836739485273,
      "loss": 4.9467,
      "mean_token_accuracy": 0.225629061460495,
      "num_tokens": 110486193.0,
      "step": 53230
    },
    {
      "entropy": 5.68757586479187,
      "epoch": 4.8290094339622645,
      "grad_norm": 1.234375,
      "learning_rate": 0.000290396086973847,
      "loss": 4.8749,
      "mean_token_accuracy": 0.23025979697704316,
      "num_tokens": 110496619.0,
      "step": 53235
    },
    {
      "entropy": 5.650324487686158,
      "epoch": 4.829462989840349,
      "grad_norm": 1.203125,
      "learning_rate": 0.00029036380623444636,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22178814560174942,
      "num_tokens": 110507103.0,
      "step": 53240
    },
    {
      "entropy": 5.668415975570679,
      "epoch": 4.829916545718433,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002903315251773182,
      "loss": 4.8656,
      "mean_token_accuracy": 0.22661312222480773,
      "num_tokens": 110516855.0,
      "step": 53245
    },
    {
      "entropy": 5.644951009750367,
      "epoch": 4.830370101596516,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029029924380313024,
      "loss": 4.8484,
      "mean_token_accuracy": 0.23292219787836074,
      "num_tokens": 110526220.0,
      "step": 53250
    },
    {
      "entropy": 5.662230443954468,
      "epoch": 4.830823657474601,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029026696211255,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2226686730980873,
      "num_tokens": 110537029.0,
      "step": 53255
    },
    {
      "entropy": 5.680279779434204,
      "epoch": 4.831277213352685,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029023468010624503,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22495268583297728,
      "num_tokens": 110548334.0,
      "step": 53260
    },
    {
      "entropy": 5.670314931869507,
      "epoch": 4.831730769230769,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029020239778488303,
      "loss": 4.8948,
      "mean_token_accuracy": 0.23200928568840026,
      "num_tokens": 110557999.0,
      "step": 53265
    },
    {
      "entropy": 5.634663105010986,
      "epoch": 4.832184325108853,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002901701151491316,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21856084913015367,
      "num_tokens": 110568743.0,
      "step": 53270
    },
    {
      "entropy": 5.664666318893433,
      "epoch": 4.832637880986938,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029013783219965835,
      "loss": 4.8804,
      "mean_token_accuracy": 0.2312145173549652,
      "num_tokens": 110578961.0,
      "step": 53275
    },
    {
      "entropy": 5.705579662322998,
      "epoch": 4.833091436865022,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029010554893713086,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2212559312582016,
      "num_tokens": 110590496.0,
      "step": 53280
    },
    {
      "entropy": 5.736619758605957,
      "epoch": 4.833544992743106,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002900732653622167,
      "loss": 4.8561,
      "mean_token_accuracy": 0.23212044835090637,
      "num_tokens": 110601018.0,
      "step": 53285
    },
    {
      "entropy": 5.631805896759033,
      "epoch": 4.8339985486211905,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002900409814755836,
      "loss": 4.854,
      "mean_token_accuracy": 0.22247184067964554,
      "num_tokens": 110611369.0,
      "step": 53290
    },
    {
      "entropy": 5.643290996551514,
      "epoch": 4.834452104499274,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029000869727789923,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22937071174383164,
      "num_tokens": 110621223.0,
      "step": 53295
    },
    {
      "entropy": 5.578810214996338,
      "epoch": 4.834905660377358,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028997641276983107,
      "loss": 4.7975,
      "mean_token_accuracy": 0.23001529574394225,
      "num_tokens": 110631033.0,
      "step": 53300
    },
    {
      "entropy": 5.6046302318573,
      "epoch": 4.835359216255442,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028994412795204695,
      "loss": 4.8408,
      "mean_token_accuracy": 0.22722357511520386,
      "num_tokens": 110642368.0,
      "step": 53305
    },
    {
      "entropy": 5.6320836544036865,
      "epoch": 4.835812772133527,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002899118428252144,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22379519492387773,
      "num_tokens": 110653317.0,
      "step": 53310
    },
    {
      "entropy": 5.6341780662536625,
      "epoch": 4.836266328011611,
      "grad_norm": 1.515625,
      "learning_rate": 0.00028987955739000123,
      "loss": 4.8637,
      "mean_token_accuracy": 0.22359674721956252,
      "num_tokens": 110663143.0,
      "step": 53315
    },
    {
      "entropy": 5.6660521030426025,
      "epoch": 4.836719883889695,
      "grad_norm": 1.296875,
      "learning_rate": 0.000289847271647075,
      "loss": 4.8869,
      "mean_token_accuracy": 0.23104428499937057,
      "num_tokens": 110673383.0,
      "step": 53320
    },
    {
      "entropy": 5.654747581481933,
      "epoch": 4.8371734397677795,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002898149855971034,
      "loss": 4.8092,
      "mean_token_accuracy": 0.22666301280260087,
      "num_tokens": 110683576.0,
      "step": 53325
    },
    {
      "entropy": 5.622866678237915,
      "epoch": 4.837626995645864,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028978269924075404,
      "loss": 4.8106,
      "mean_token_accuracy": 0.23146373629570008,
      "num_tokens": 110693986.0,
      "step": 53330
    },
    {
      "entropy": 5.608734178543091,
      "epoch": 4.838080551523948,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002897504125786948,
      "loss": 4.8161,
      "mean_token_accuracy": 0.2283170700073242,
      "num_tokens": 110703724.0,
      "step": 53335
    },
    {
      "entropy": 5.605955123901367,
      "epoch": 4.838534107402032,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028971812561159324,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22396186292171477,
      "num_tokens": 110714251.0,
      "step": 53340
    },
    {
      "entropy": 5.654089164733887,
      "epoch": 4.8389876632801165,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00028968583834011714,
      "loss": 4.9075,
      "mean_token_accuracy": 0.21755824238061905,
      "num_tokens": 110724011.0,
      "step": 53345
    },
    {
      "entropy": 5.68786301612854,
      "epoch": 4.8394412191582,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028965355076493405,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22417781949043275,
      "num_tokens": 110733627.0,
      "step": 53350
    },
    {
      "entropy": 5.690388917922974,
      "epoch": 4.839894775036284,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002896212628867118,
      "loss": 4.8746,
      "mean_token_accuracy": 0.221796652674675,
      "num_tokens": 110744514.0,
      "step": 53355
    },
    {
      "entropy": 5.744256544113159,
      "epoch": 4.8403483309143684,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028958897470611825,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2280839651823044,
      "num_tokens": 110754257.0,
      "step": 53360
    },
    {
      "entropy": 5.6345189094543455,
      "epoch": 4.840801886792453,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002895566862238208,
      "loss": 4.831,
      "mean_token_accuracy": 0.22824896425008773,
      "num_tokens": 110764383.0,
      "step": 53365
    },
    {
      "entropy": 5.709937667846679,
      "epoch": 4.841255442670537,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002895243974404875,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2222899913787842,
      "num_tokens": 110775006.0,
      "step": 53370
    },
    {
      "entropy": 5.695196342468262,
      "epoch": 4.841708998548621,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002894921083567859,
      "loss": 4.8837,
      "mean_token_accuracy": 0.2254146322607994,
      "num_tokens": 110784764.0,
      "step": 53375
    },
    {
      "entropy": 5.669567394256592,
      "epoch": 4.8421625544267055,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002894598189733838,
      "loss": 4.8555,
      "mean_token_accuracy": 0.23375797569751738,
      "num_tokens": 110794942.0,
      "step": 53380
    },
    {
      "entropy": 5.611394691467285,
      "epoch": 4.84261611030479,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000289427529290949,
      "loss": 4.7969,
      "mean_token_accuracy": 0.23847442865371704,
      "num_tokens": 110804696.0,
      "step": 53385
    },
    {
      "entropy": 5.56563138961792,
      "epoch": 4.843069666182874,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028939523931014924,
      "loss": 4.766,
      "mean_token_accuracy": 0.23961413353681565,
      "num_tokens": 110814867.0,
      "step": 53390
    },
    {
      "entropy": 5.6385478496551515,
      "epoch": 4.843523222060957,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028936294903165224,
      "loss": 4.8406,
      "mean_token_accuracy": 0.22437868118286133,
      "num_tokens": 110824738.0,
      "step": 53395
    },
    {
      "entropy": 5.668552875518799,
      "epoch": 4.843976777939043,
      "grad_norm": 1.375,
      "learning_rate": 0.0002893306584561257,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2252178356051445,
      "num_tokens": 110834646.0,
      "step": 53400
    },
    {
      "entropy": 5.683118200302124,
      "epoch": 4.844430333817126,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028929836758423763,
      "loss": 4.9294,
      "mean_token_accuracy": 0.22451551258563995,
      "num_tokens": 110843907.0,
      "step": 53405
    },
    {
      "entropy": 5.645402336120606,
      "epoch": 4.84488388969521,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002892660764166556,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22451913207769394,
      "num_tokens": 110854930.0,
      "step": 53410
    },
    {
      "entropy": 5.597357892990113,
      "epoch": 4.8453374455732945,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028923378495404747,
      "loss": 4.819,
      "mean_token_accuracy": 0.22951989322900773,
      "num_tokens": 110865256.0,
      "step": 53415
    },
    {
      "entropy": 5.70387716293335,
      "epoch": 4.845791001451379,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028920149319708105,
      "loss": 4.9338,
      "mean_token_accuracy": 0.22250669747591018,
      "num_tokens": 110875600.0,
      "step": 53420
    },
    {
      "entropy": 5.739201879501342,
      "epoch": 4.846244557329463,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028916920114642415,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22582444697618484,
      "num_tokens": 110886856.0,
      "step": 53425
    },
    {
      "entropy": 5.69910306930542,
      "epoch": 4.846698113207547,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028913690880274454,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23255734145641327,
      "num_tokens": 110896908.0,
      "step": 53430
    },
    {
      "entropy": 5.662793827056885,
      "epoch": 4.8471516690856316,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028910461616671004,
      "loss": 4.8434,
      "mean_token_accuracy": 0.23335074186325072,
      "num_tokens": 110908987.0,
      "step": 53435
    },
    {
      "entropy": 5.6475951194763185,
      "epoch": 4.847605224963716,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028907232323898855,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23138083666563034,
      "num_tokens": 110920036.0,
      "step": 53440
    },
    {
      "entropy": 5.674916696548462,
      "epoch": 4.8480587808418,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028904003002024775,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22258280813694,
      "num_tokens": 110931170.0,
      "step": 53445
    },
    {
      "entropy": 5.698870468139648,
      "epoch": 4.8485123367198835,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002890077365111556,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2184234470129013,
      "num_tokens": 110942698.0,
      "step": 53450
    },
    {
      "entropy": 5.693228721618652,
      "epoch": 4.848965892597968,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028897544271237985,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22374559938907623,
      "num_tokens": 110953400.0,
      "step": 53455
    },
    {
      "entropy": 5.699455738067627,
      "epoch": 4.849419448476052,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028894314862458844,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2258833661675453,
      "num_tokens": 110963792.0,
      "step": 53460
    },
    {
      "entropy": 5.681771612167358,
      "epoch": 4.849873004354136,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028891085424844913,
      "loss": 4.8587,
      "mean_token_accuracy": 0.22460289746522905,
      "num_tokens": 110973636.0,
      "step": 53465
    },
    {
      "entropy": 5.630780982971191,
      "epoch": 4.8503265602322205,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002888785595846298,
      "loss": 4.8782,
      "mean_token_accuracy": 0.23026085495948792,
      "num_tokens": 110983163.0,
      "step": 53470
    },
    {
      "entropy": 5.642491054534912,
      "epoch": 4.850780116110305,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002888462646337983,
      "loss": 4.8386,
      "mean_token_accuracy": 0.22442800998687745,
      "num_tokens": 110993344.0,
      "step": 53475
    },
    {
      "entropy": 5.685483407974243,
      "epoch": 4.851233671988389,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028881396939662256,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2227097272872925,
      "num_tokens": 111004380.0,
      "step": 53480
    },
    {
      "entropy": 5.679120683670044,
      "epoch": 4.851687227866473,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002887816738737704,
      "loss": 4.8254,
      "mean_token_accuracy": 0.23204307109117508,
      "num_tokens": 111015661.0,
      "step": 53485
    },
    {
      "entropy": 5.6261943817138675,
      "epoch": 4.852140783744558,
      "grad_norm": 1.25,
      "learning_rate": 0.0002887493780659097,
      "loss": 4.8347,
      "mean_token_accuracy": 0.2298179194331169,
      "num_tokens": 111026496.0,
      "step": 53490
    },
    {
      "entropy": 5.590049123764038,
      "epoch": 4.852594339622642,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028871708197370844,
      "loss": 4.7986,
      "mean_token_accuracy": 0.22922968715429307,
      "num_tokens": 111036356.0,
      "step": 53495
    },
    {
      "entropy": 5.650780582427979,
      "epoch": 4.853047895500726,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028868478559783433,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2245974138379097,
      "num_tokens": 111045967.0,
      "step": 53500
    },
    {
      "entropy": 5.665289688110351,
      "epoch": 4.8535014513788095,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028865248893895543,
      "loss": 4.862,
      "mean_token_accuracy": 0.22514557093381882,
      "num_tokens": 111056459.0,
      "step": 53505
    },
    {
      "entropy": 5.640075063705444,
      "epoch": 4.853955007256894,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028862019199773955,
      "loss": 4.8011,
      "mean_token_accuracy": 0.22611933648586274,
      "num_tokens": 111065901.0,
      "step": 53510
    },
    {
      "entropy": 5.612060165405273,
      "epoch": 4.854408563134978,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002885878947748546,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22793640494346618,
      "num_tokens": 111075721.0,
      "step": 53515
    },
    {
      "entropy": 5.608594131469727,
      "epoch": 4.854862119013062,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028855559727096854,
      "loss": 4.8365,
      "mean_token_accuracy": 0.23345322608947755,
      "num_tokens": 111086453.0,
      "step": 53520
    },
    {
      "entropy": 5.657582759857178,
      "epoch": 4.855315674891147,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028852329948674926,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21954546570777894,
      "num_tokens": 111099130.0,
      "step": 53525
    },
    {
      "entropy": 5.619465255737305,
      "epoch": 4.855769230769231,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002884910014228648,
      "loss": 4.8237,
      "mean_token_accuracy": 0.23293367177248,
      "num_tokens": 111108304.0,
      "step": 53530
    },
    {
      "entropy": 5.685942316055298,
      "epoch": 4.856222786647315,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002884587030799829,
      "loss": 4.817,
      "mean_token_accuracy": 0.22885674387216567,
      "num_tokens": 111118369.0,
      "step": 53535
    },
    {
      "entropy": 5.609336519241333,
      "epoch": 4.856676342525399,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00028842640445877153,
      "loss": 4.8673,
      "mean_token_accuracy": 0.2258546084165573,
      "num_tokens": 111128550.0,
      "step": 53540
    },
    {
      "entropy": 5.712727737426758,
      "epoch": 4.857129898403484,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028839410555989885,
      "loss": 4.961,
      "mean_token_accuracy": 0.215118145942688,
      "num_tokens": 111140824.0,
      "step": 53545
    },
    {
      "entropy": 5.641538810729981,
      "epoch": 4.857583454281567,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028836180638403257,
      "loss": 4.8538,
      "mean_token_accuracy": 0.21999497413635255,
      "num_tokens": 111152241.0,
      "step": 53550
    },
    {
      "entropy": 5.675179672241211,
      "epoch": 4.858037010159651,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002883295069318407,
      "loss": 4.92,
      "mean_token_accuracy": 0.22193428874015808,
      "num_tokens": 111161897.0,
      "step": 53555
    },
    {
      "entropy": 5.676577758789063,
      "epoch": 4.8584905660377355,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002882972072039914,
      "loss": 4.856,
      "mean_token_accuracy": 0.22808670848608018,
      "num_tokens": 111172586.0,
      "step": 53560
    },
    {
      "entropy": 5.69434962272644,
      "epoch": 4.85894412191582,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028826490720115224,
      "loss": 4.9773,
      "mean_token_accuracy": 0.21963163912296296,
      "num_tokens": 111182734.0,
      "step": 53565
    },
    {
      "entropy": 5.586197423934936,
      "epoch": 4.859397677793904,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028823260692399156,
      "loss": 4.7656,
      "mean_token_accuracy": 0.2346462905406952,
      "num_tokens": 111193551.0,
      "step": 53570
    },
    {
      "entropy": 5.629780721664429,
      "epoch": 4.859851233671988,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002882003063731771,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23312482982873917,
      "num_tokens": 111202604.0,
      "step": 53575
    },
    {
      "entropy": 5.60758867263794,
      "epoch": 4.860304789550073,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028816800554937703,
      "loss": 4.8317,
      "mean_token_accuracy": 0.23280305713415145,
      "num_tokens": 111213675.0,
      "step": 53580
    },
    {
      "entropy": 5.7010571479797365,
      "epoch": 4.860758345428157,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028813570445325915,
      "loss": 4.8162,
      "mean_token_accuracy": 0.23637080788612366,
      "num_tokens": 111223354.0,
      "step": 53585
    },
    {
      "entropy": 5.664448404312134,
      "epoch": 4.861211901306241,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028810340308549164,
      "loss": 4.7789,
      "mean_token_accuracy": 0.23842962086200714,
      "num_tokens": 111232865.0,
      "step": 53590
    },
    {
      "entropy": 5.580556964874267,
      "epoch": 4.861665457184325,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028807110144674243,
      "loss": 4.8342,
      "mean_token_accuracy": 0.2297973409295082,
      "num_tokens": 111242924.0,
      "step": 53595
    },
    {
      "entropy": 5.619889211654663,
      "epoch": 4.86211901306241,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002880387995376794,
      "loss": 4.79,
      "mean_token_accuracy": 0.23052696734666825,
      "num_tokens": 111252921.0,
      "step": 53600
    },
    {
      "entropy": 5.664149570465088,
      "epoch": 4.862572568940493,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002880064973589708,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22271490693092347,
      "num_tokens": 111263176.0,
      "step": 53605
    },
    {
      "entropy": 5.661705493927002,
      "epoch": 4.863026124818577,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028797419491128446,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22154775261878967,
      "num_tokens": 111273043.0,
      "step": 53610
    },
    {
      "entropy": 5.66186752319336,
      "epoch": 4.863479680696662,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028794189219528846,
      "loss": 4.8243,
      "mean_token_accuracy": 0.22848019003868103,
      "num_tokens": 111282720.0,
      "step": 53615
    },
    {
      "entropy": 5.560529899597168,
      "epoch": 4.863933236574746,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028790958921165086,
      "loss": 4.8601,
      "mean_token_accuracy": 0.21905989348888397,
      "num_tokens": 111293946.0,
      "step": 53620
    },
    {
      "entropy": 5.6806902408599855,
      "epoch": 4.86438679245283,
      "grad_norm": 1.25,
      "learning_rate": 0.0002878772859610396,
      "loss": 4.9235,
      "mean_token_accuracy": 0.21922473907470702,
      "num_tokens": 111305052.0,
      "step": 53625
    },
    {
      "entropy": 5.7242518901824955,
      "epoch": 4.864840348330914,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028784498244412284,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22676833868026733,
      "num_tokens": 111314216.0,
      "step": 53630
    },
    {
      "entropy": 5.644245386123657,
      "epoch": 4.865293904208999,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028781267866156854,
      "loss": 4.8221,
      "mean_token_accuracy": 0.2293958067893982,
      "num_tokens": 111324939.0,
      "step": 53635
    },
    {
      "entropy": 5.676898956298828,
      "epoch": 4.865747460087083,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028778037461404477,
      "loss": 4.8591,
      "mean_token_accuracy": 0.2203165650367737,
      "num_tokens": 111334908.0,
      "step": 53640
    },
    {
      "entropy": 5.595578622817993,
      "epoch": 4.866201015965167,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028774807030221966,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22695845514535903,
      "num_tokens": 111345560.0,
      "step": 53645
    },
    {
      "entropy": 5.609410619735717,
      "epoch": 4.8666545718432515,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002877157657267611,
      "loss": 4.8004,
      "mean_token_accuracy": 0.23474580496549607,
      "num_tokens": 111355955.0,
      "step": 53650
    },
    {
      "entropy": 5.7086262702941895,
      "epoch": 4.867108127721336,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028768346088833737,
      "loss": 4.943,
      "mean_token_accuracy": 0.22919799238443375,
      "num_tokens": 111365567.0,
      "step": 53655
    },
    {
      "entropy": 5.5942497730255125,
      "epoch": 4.867561683599419,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002876511557876164,
      "loss": 4.7982,
      "mean_token_accuracy": 0.23076252192258834,
      "num_tokens": 111375740.0,
      "step": 53660
    },
    {
      "entropy": 5.624006175994873,
      "epoch": 4.868015239477503,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028761885042526635,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22788144201040267,
      "num_tokens": 111385593.0,
      "step": 53665
    },
    {
      "entropy": 5.691855001449585,
      "epoch": 4.868468795355588,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028758654480195517,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21506067365407944,
      "num_tokens": 111395731.0,
      "step": 53670
    },
    {
      "entropy": 5.70223708152771,
      "epoch": 4.868922351233672,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002875542389183511,
      "loss": 4.8385,
      "mean_token_accuracy": 0.23649916648864747,
      "num_tokens": 111406140.0,
      "step": 53675
    },
    {
      "entropy": 5.67141375541687,
      "epoch": 4.869375907111756,
      "grad_norm": 1.375,
      "learning_rate": 0.00028752193277512217,
      "loss": 4.8837,
      "mean_token_accuracy": 0.22787901312112807,
      "num_tokens": 111416853.0,
      "step": 53680
    },
    {
      "entropy": 5.623068809509277,
      "epoch": 4.86982946298984,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002874896263729364,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22982754409313202,
      "num_tokens": 111427043.0,
      "step": 53685
    },
    {
      "entropy": 5.734610271453858,
      "epoch": 4.870283018867925,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002874573197124621,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22058325558900832,
      "num_tokens": 111436791.0,
      "step": 53690
    },
    {
      "entropy": 5.615300846099854,
      "epoch": 4.870736574746009,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002874250127943672,
      "loss": 4.7994,
      "mean_token_accuracy": 0.23612288236618043,
      "num_tokens": 111446603.0,
      "step": 53695
    },
    {
      "entropy": 5.6245026111602785,
      "epoch": 4.871190130624093,
      "grad_norm": 1.40625,
      "learning_rate": 0.00028739270561931984,
      "loss": 4.7723,
      "mean_token_accuracy": 0.24160352796316148,
      "num_tokens": 111456942.0,
      "step": 53700
    },
    {
      "entropy": 5.633362865447998,
      "epoch": 4.871643686502177,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028736039818798813,
      "loss": 4.863,
      "mean_token_accuracy": 0.22618399560451508,
      "num_tokens": 111467671.0,
      "step": 53705
    },
    {
      "entropy": 5.6662682056427,
      "epoch": 4.872097242380261,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002873280905010403,
      "loss": 4.9423,
      "mean_token_accuracy": 0.22038195580244063,
      "num_tokens": 111476495.0,
      "step": 53710
    },
    {
      "entropy": 5.606502056121826,
      "epoch": 4.872550798258345,
      "grad_norm": 1.375,
      "learning_rate": 0.00028729578255914446,
      "loss": 4.7263,
      "mean_token_accuracy": 0.24102134853601456,
      "num_tokens": 111487491.0,
      "step": 53715
    },
    {
      "entropy": 5.629835605621338,
      "epoch": 4.873004354136429,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002872634743629686,
      "loss": 4.7942,
      "mean_token_accuracy": 0.22751702219247819,
      "num_tokens": 111497970.0,
      "step": 53720
    },
    {
      "entropy": 5.653442001342773,
      "epoch": 4.873457910014514,
      "grad_norm": 1.25,
      "learning_rate": 0.000287231165913181,
      "loss": 4.878,
      "mean_token_accuracy": 0.22399977147579192,
      "num_tokens": 111507635.0,
      "step": 53725
    },
    {
      "entropy": 5.607708263397217,
      "epoch": 4.873911465892598,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002871988572104498,
      "loss": 4.8338,
      "mean_token_accuracy": 0.2302190124988556,
      "num_tokens": 111516751.0,
      "step": 53730
    },
    {
      "entropy": 5.629733562469482,
      "epoch": 4.874365021770682,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028716654825544316,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22682311087846757,
      "num_tokens": 111527748.0,
      "step": 53735
    },
    {
      "entropy": 5.671847105026245,
      "epoch": 4.8748185776487665,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002871342390488291,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22485494762659072,
      "num_tokens": 111537718.0,
      "step": 53740
    },
    {
      "entropy": 5.574645090103149,
      "epoch": 4.875272133526851,
      "grad_norm": 1.375,
      "learning_rate": 0.00028710192959127597,
      "loss": 4.738,
      "mean_token_accuracy": 0.2382654994726181,
      "num_tokens": 111546631.0,
      "step": 53745
    },
    {
      "entropy": 5.654329633712768,
      "epoch": 4.875725689404935,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002870696198834518,
      "loss": 4.836,
      "mean_token_accuracy": 0.2285819411277771,
      "num_tokens": 111557276.0,
      "step": 53750
    },
    {
      "entropy": 5.643083333969116,
      "epoch": 4.876179245283019,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028703730992602486,
      "loss": 4.8454,
      "mean_token_accuracy": 0.23144342750310898,
      "num_tokens": 111567146.0,
      "step": 53755
    },
    {
      "entropy": 5.668443632125855,
      "epoch": 4.876632801161103,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028700499971966327,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2191551461815834,
      "num_tokens": 111576887.0,
      "step": 53760
    },
    {
      "entropy": 5.678356742858886,
      "epoch": 4.877086357039187,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028697268926503527,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22250308394432067,
      "num_tokens": 111588212.0,
      "step": 53765
    },
    {
      "entropy": 5.66697325706482,
      "epoch": 4.877539912917271,
      "grad_norm": 1.40625,
      "learning_rate": 0.00028694037856280894,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22134220898151397,
      "num_tokens": 111598308.0,
      "step": 53770
    },
    {
      "entropy": 5.704410457611084,
      "epoch": 4.8779934687953554,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002869080676136525,
      "loss": 4.9024,
      "mean_token_accuracy": 0.22673436105251313,
      "num_tokens": 111608252.0,
      "step": 53775
    },
    {
      "entropy": 5.655329608917237,
      "epoch": 4.87844702467344,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002868757564182343,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22151461094617844,
      "num_tokens": 111619555.0,
      "step": 53780
    },
    {
      "entropy": 5.5927958488464355,
      "epoch": 4.878900580551524,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002868434449772223,
      "loss": 4.814,
      "mean_token_accuracy": 0.2296259507536888,
      "num_tokens": 111629578.0,
      "step": 53785
    },
    {
      "entropy": 5.664986181259155,
      "epoch": 4.879354136429608,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002868111332912849,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22563903629779816,
      "num_tokens": 111639389.0,
      "step": 53790
    },
    {
      "entropy": 5.737447500228882,
      "epoch": 4.8798076923076925,
      "grad_norm": 1.375,
      "learning_rate": 0.0002867788213610903,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22411651760339737,
      "num_tokens": 111649855.0,
      "step": 53795
    },
    {
      "entropy": 5.733330297470093,
      "epoch": 4.880261248185777,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002867465091873066,
      "loss": 4.9426,
      "mean_token_accuracy": 0.2224299877882004,
      "num_tokens": 111660315.0,
      "step": 53800
    },
    {
      "entropy": 5.644560289382935,
      "epoch": 4.880714804063861,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002867141967706021,
      "loss": 4.901,
      "mean_token_accuracy": 0.22353803962469102,
      "num_tokens": 111669883.0,
      "step": 53805
    },
    {
      "entropy": 5.598710584640503,
      "epoch": 4.881168359941945,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028668188411164506,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21956893503665925,
      "num_tokens": 111680598.0,
      "step": 53810
    },
    {
      "entropy": 5.684199619293213,
      "epoch": 4.881621915820029,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028664957121110374,
      "loss": 4.8619,
      "mean_token_accuracy": 0.2267229065299034,
      "num_tokens": 111689805.0,
      "step": 53815
    },
    {
      "entropy": 5.688586664199829,
      "epoch": 4.882075471698113,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028661725806964624,
      "loss": 4.8165,
      "mean_token_accuracy": 0.23283295184373856,
      "num_tokens": 111700313.0,
      "step": 53820
    },
    {
      "entropy": 5.702735757827758,
      "epoch": 4.882529027576197,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002865849446879409,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2273788720369339,
      "num_tokens": 111711301.0,
      "step": 53825
    },
    {
      "entropy": 5.629639577865601,
      "epoch": 4.8829825834542815,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002865526310666559,
      "loss": 4.8226,
      "mean_token_accuracy": 0.2293864443898201,
      "num_tokens": 111720800.0,
      "step": 53830
    },
    {
      "entropy": 5.610587787628174,
      "epoch": 4.883436139332366,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028652031720645963,
      "loss": 4.8015,
      "mean_token_accuracy": 0.2298102542757988,
      "num_tokens": 111731665.0,
      "step": 53835
    },
    {
      "entropy": 5.6298058986663815,
      "epoch": 4.88388969521045,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002864880031080202,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22697973102331162,
      "num_tokens": 111742149.0,
      "step": 53840
    },
    {
      "entropy": 5.664055109024048,
      "epoch": 4.884343251088534,
      "grad_norm": 1.25,
      "learning_rate": 0.00028645568877200603,
      "loss": 4.8289,
      "mean_token_accuracy": 0.23299773782491684,
      "num_tokens": 111753089.0,
      "step": 53845
    },
    {
      "entropy": 5.602269268035888,
      "epoch": 4.8847968069666186,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002864233741990852,
      "loss": 4.8088,
      "mean_token_accuracy": 0.23182955682277678,
      "num_tokens": 111762937.0,
      "step": 53850
    },
    {
      "entropy": 5.593773221969604,
      "epoch": 4.885250362844703,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002863910593899261,
      "loss": 4.7835,
      "mean_token_accuracy": 0.23938246071338654,
      "num_tokens": 111772920.0,
      "step": 53855
    },
    {
      "entropy": 5.619352960586548,
      "epoch": 4.885703918722786,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000286358744345197,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23044109493494033,
      "num_tokens": 111783117.0,
      "step": 53860
    },
    {
      "entropy": 5.6672797203063965,
      "epoch": 4.8861574746008705,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002863264290655662,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2186119884252548,
      "num_tokens": 111795033.0,
      "step": 53865
    },
    {
      "entropy": 5.643439626693725,
      "epoch": 4.886611030478955,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028629411355170193,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22066473364830017,
      "num_tokens": 111805156.0,
      "step": 53870
    },
    {
      "entropy": 5.675177097320557,
      "epoch": 4.887064586357039,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028626179780427246,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2216375082731247,
      "num_tokens": 111815363.0,
      "step": 53875
    },
    {
      "entropy": 5.678713941574097,
      "epoch": 4.887518142235123,
      "grad_norm": 1.71875,
      "learning_rate": 0.00028622948182394623,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2297086462378502,
      "num_tokens": 111824983.0,
      "step": 53880
    },
    {
      "entropy": 5.6028724193573,
      "epoch": 4.8879716981132075,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028619716561139143,
      "loss": 4.8609,
      "mean_token_accuracy": 0.22374506294727325,
      "num_tokens": 111836468.0,
      "step": 53885
    },
    {
      "entropy": 5.670812129974365,
      "epoch": 4.888425253991292,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002861648491672764,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22587817162275314,
      "num_tokens": 111847659.0,
      "step": 53890
    },
    {
      "entropy": 5.694909477233887,
      "epoch": 4.888878809869376,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028613253249226944,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2182258978486061,
      "num_tokens": 111858742.0,
      "step": 53895
    },
    {
      "entropy": 5.675878095626831,
      "epoch": 4.88933236574746,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028610021558703885,
      "loss": 4.8993,
      "mean_token_accuracy": 0.225470133125782,
      "num_tokens": 111869692.0,
      "step": 53900
    },
    {
      "entropy": 5.649547863006592,
      "epoch": 4.889785921625545,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028606789845225295,
      "loss": 4.8567,
      "mean_token_accuracy": 0.2254531130194664,
      "num_tokens": 111880532.0,
      "step": 53905
    },
    {
      "entropy": 5.7409216403961185,
      "epoch": 4.890239477503629,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028603558108858016,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22071272879838943,
      "num_tokens": 111891178.0,
      "step": 53910
    },
    {
      "entropy": 5.698775577545166,
      "epoch": 4.890693033381712,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002860032634966887,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22841224670410157,
      "num_tokens": 111902043.0,
      "step": 53915
    },
    {
      "entropy": 5.665069675445556,
      "epoch": 4.8911465892597965,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000285970945677247,
      "loss": 4.8285,
      "mean_token_accuracy": 0.2294325515627861,
      "num_tokens": 111912170.0,
      "step": 53920
    },
    {
      "entropy": 5.584176445007325,
      "epoch": 4.891600145137881,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028593862763092327,
      "loss": 4.8474,
      "mean_token_accuracy": 0.22930973768234253,
      "num_tokens": 111922415.0,
      "step": 53925
    },
    {
      "entropy": 5.684861183166504,
      "epoch": 4.892053701015965,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028590630935838596,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2220746174454689,
      "num_tokens": 111933338.0,
      "step": 53930
    },
    {
      "entropy": 5.628093194961548,
      "epoch": 4.892507256894049,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002858739908603034,
      "loss": 4.8354,
      "mean_token_accuracy": 0.23648880124092103,
      "num_tokens": 111942225.0,
      "step": 53935
    },
    {
      "entropy": 5.622928619384766,
      "epoch": 4.892960812772134,
      "grad_norm": 1.375,
      "learning_rate": 0.0002858416721373439,
      "loss": 4.8303,
      "mean_token_accuracy": 0.23444282114505768,
      "num_tokens": 111953356.0,
      "step": 53940
    },
    {
      "entropy": 5.607836151123047,
      "epoch": 4.893414368650218,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028580935319017593,
      "loss": 4.7744,
      "mean_token_accuracy": 0.22941247671842574,
      "num_tokens": 111964275.0,
      "step": 53945
    },
    {
      "entropy": 5.642651796340942,
      "epoch": 4.893867924528302,
      "grad_norm": 1.40625,
      "learning_rate": 0.00028577703401946777,
      "loss": 4.854,
      "mean_token_accuracy": 0.2289108619093895,
      "num_tokens": 111974006.0,
      "step": 53950
    },
    {
      "entropy": 5.700316572189331,
      "epoch": 4.894321480406386,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028574471462588776,
      "loss": 4.9643,
      "mean_token_accuracy": 0.21803802847862244,
      "num_tokens": 111985085.0,
      "step": 53955
    },
    {
      "entropy": 5.69154839515686,
      "epoch": 4.89477503628447,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002857123950101043,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21635894030332564,
      "num_tokens": 111995162.0,
      "step": 53960
    },
    {
      "entropy": 5.702213287353516,
      "epoch": 4.895228592162555,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002856800751727858,
      "loss": 4.8814,
      "mean_token_accuracy": 0.2226106643676758,
      "num_tokens": 112005625.0,
      "step": 53965
    },
    {
      "entropy": 5.644302225112915,
      "epoch": 4.895682148040638,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028564775511460066,
      "loss": 4.851,
      "mean_token_accuracy": 0.22957917749881745,
      "num_tokens": 112015977.0,
      "step": 53970
    },
    {
      "entropy": 5.625647401809692,
      "epoch": 4.8961357039187225,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002856154348362173,
      "loss": 4.8423,
      "mean_token_accuracy": 0.229736790060997,
      "num_tokens": 112025655.0,
      "step": 53975
    },
    {
      "entropy": 5.596014165878296,
      "epoch": 4.896589259796807,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028558311433830386,
      "loss": 4.8221,
      "mean_token_accuracy": 0.23546603620052337,
      "num_tokens": 112036986.0,
      "step": 53980
    },
    {
      "entropy": 5.649260568618774,
      "epoch": 4.897042815674891,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000285550793621529,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22766707241535186,
      "num_tokens": 112047541.0,
      "step": 53985
    },
    {
      "entropy": 5.653165197372436,
      "epoch": 4.897496371552975,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002855184726865611,
      "loss": 4.8301,
      "mean_token_accuracy": 0.2266516476869583,
      "num_tokens": 112058125.0,
      "step": 53990
    },
    {
      "entropy": 5.724337196350097,
      "epoch": 4.89794992743106,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028548615153406844,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2241588830947876,
      "num_tokens": 112068712.0,
      "step": 53995
    },
    {
      "entropy": 5.612188911437988,
      "epoch": 4.898403483309144,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028545383016471954,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22237393856048585,
      "num_tokens": 112078529.0,
      "step": 54000
    },
    {
      "epoch": 4.898403483309144,
      "eval_entropy": 5.365193633472218,
      "eval_loss": 4.938462734222412,
      "eval_mean_token_accuracy": 0.23199300347612453,
      "eval_num_tokens": 112078529.0,
      "eval_runtime": 20.659,
      "eval_samples_per_second": 1711.6,
      "eval_steps_per_second": 213.95,
      "step": 54000
    }
  ],
  "logging_steps": 5,
  "max_steps": 110230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.431833283436544e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}