nld-latn-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-24000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.1770682148040637,
  "eval_steps": 3000,
  "global_step": 24000,
  "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
    }
  ],
  "logging_steps": 5,
  "max_steps": 110230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.081703127564288e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}