nld-latn-10mb-100mb_seed3407/checkpoint-21000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9049346879535558,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.742568969726562,
      "epoch": 0.00045355587808418,
      "grad_norm": 5.4375,
      "learning_rate": 2e-06,
      "loss": 10.7846,
      "mean_token_accuracy": 0.0,
      "num_tokens": 10827.0,
      "step": 5
    },
    {
      "entropy": 10.742588996887207,
      "epoch": 0.00090711175616836,
      "grad_norm": 5.25,
      "learning_rate": 4.5e-06,
      "loss": 10.7602,
      "mean_token_accuracy": 6.963788182474673e-05,
      "num_tokens": 21025.0,
      "step": 10
    },
    {
      "entropy": 10.742573070526124,
      "epoch": 0.00136066763425254,
      "grad_norm": 5.25,
      "learning_rate": 7e-06,
      "loss": 10.7437,
      "mean_token_accuracy": 9.000900317914783e-05,
      "num_tokens": 31678.0,
      "step": 15
    },
    {
      "entropy": 10.742661285400391,
      "epoch": 0.00181422351233672,
      "grad_norm": 5.125,
      "learning_rate": 9.5e-06,
      "loss": 10.6591,
      "mean_token_accuracy": 0.0005137635511346161,
      "num_tokens": 41025.0,
      "step": 20
    },
    {
      "entropy": 10.742550945281982,
      "epoch": 0.0022677793904208998,
      "grad_norm": 4.875,
      "learning_rate": 1.2e-05,
      "loss": 10.5792,
      "mean_token_accuracy": 0.006170097610447556,
      "num_tokens": 51270.0,
      "step": 25
    },
    {
      "entropy": 10.742267036437989,
      "epoch": 0.00272133526850508,
      "grad_norm": 3.953125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.4637,
      "mean_token_accuracy": 0.027165642101317646,
      "num_tokens": 62142.0,
      "step": 30
    },
    {
      "entropy": 10.741279792785644,
      "epoch": 0.00317489114658926,
      "grad_norm": 3.390625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3199,
      "mean_token_accuracy": 0.04252696111798286,
      "num_tokens": 72709.0,
      "step": 35
    },
    {
      "entropy": 10.738855934143066,
      "epoch": 0.00362844702467344,
      "grad_norm": 2.625,
      "learning_rate": 1.95e-05,
      "loss": 10.1477,
      "mean_token_accuracy": 0.041494901105761525,
      "num_tokens": 82313.0,
      "step": 40
    },
    {
      "entropy": 10.734966659545899,
      "epoch": 0.00408200290275762,
      "grad_norm": 2.484375,
      "learning_rate": 2.2e-05,
      "loss": 10.0172,
      "mean_token_accuracy": 0.048007883876562116,
      "num_tokens": 92450.0,
      "step": 45
    },
    {
      "entropy": 10.731274795532226,
      "epoch": 0.0045355587808417995,
      "grad_norm": 2.125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.9997,
      "mean_token_accuracy": 0.0419012438505888,
      "num_tokens": 103978.0,
      "step": 50
    },
    {
      "entropy": 10.729759407043456,
      "epoch": 0.00498911465892598,
      "grad_norm": 1.96875,
      "learning_rate": 2.7e-05,
      "loss": 9.8946,
      "mean_token_accuracy": 0.04565211050212383,
      "num_tokens": 114134.0,
      "step": 55
    },
    {
      "entropy": 10.727651214599609,
      "epoch": 0.00544267053701016,
      "grad_norm": 2.015625,
      "learning_rate": 2.95e-05,
      "loss": 9.8734,
      "mean_token_accuracy": 0.04470008127391338,
      "num_tokens": 125037.0,
      "step": 60
    },
    {
      "entropy": 10.725145721435547,
      "epoch": 0.00589622641509434,
      "grad_norm": 1.9375,
      "learning_rate": 3.2e-05,
      "loss": 9.7559,
      "mean_token_accuracy": 0.04514850843697786,
      "num_tokens": 134721.0,
      "step": 65
    },
    {
      "entropy": 10.723241519927978,
      "epoch": 0.00634978229317852,
      "grad_norm": 2.03125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.6935,
      "mean_token_accuracy": 0.04249380584806204,
      "num_tokens": 144448.0,
      "step": 70
    },
    {
      "entropy": 10.721195030212403,
      "epoch": 0.006803338171262699,
      "grad_norm": 2.015625,
      "learning_rate": 3.7e-05,
      "loss": 9.635,
      "mean_token_accuracy": 0.04518970362842083,
      "num_tokens": 154343.0,
      "step": 75
    },
    {
      "entropy": 10.7159348487854,
      "epoch": 0.00725689404934688,
      "grad_norm": 2.0,
      "learning_rate": 3.95e-05,
      "loss": 9.6121,
      "mean_token_accuracy": 0.041934683173894885,
      "num_tokens": 165650.0,
      "step": 80
    },
    {
      "entropy": 10.709981918334961,
      "epoch": 0.007710449927431059,
      "grad_norm": 1.828125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6128,
      "mean_token_accuracy": 0.03882604259997606,
      "num_tokens": 177098.0,
      "step": 85
    },
    {
      "entropy": 10.703641796112061,
      "epoch": 0.00816400580551524,
      "grad_norm": 2.015625,
      "learning_rate": 4.45e-05,
      "loss": 9.4133,
      "mean_token_accuracy": 0.04838715270161629,
      "num_tokens": 187903.0,
      "step": 90
    },
    {
      "entropy": 10.691870975494385,
      "epoch": 0.00861756168359942,
      "grad_norm": 1.9296875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.3442,
      "mean_token_accuracy": 0.04516914151608944,
      "num_tokens": 198077.0,
      "step": 95
    },
    {
      "entropy": 10.677101039886475,
      "epoch": 0.009071117561683599,
      "grad_norm": 1.859375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.3156,
      "mean_token_accuracy": 0.04309112168848515,
      "num_tokens": 208641.0,
      "step": 100
    },
    {
      "entropy": 10.661001300811767,
      "epoch": 0.009524673439767779,
      "grad_norm": 1.9296875,
      "learning_rate": 5.2e-05,
      "loss": 9.1598,
      "mean_token_accuracy": 0.05524700954556465,
      "num_tokens": 218398.0,
      "step": 105
    },
    {
      "entropy": 10.64128761291504,
      "epoch": 0.00997822931785196,
      "grad_norm": 1.7578125,
      "learning_rate": 5.45e-05,
      "loss": 9.035,
      "mean_token_accuracy": 0.06018046177923679,
      "num_tokens": 229059.0,
      "step": 110
    },
    {
      "entropy": 10.614240169525146,
      "epoch": 0.01043178519593614,
      "grad_norm": 1.796875,
      "learning_rate": 5.7e-05,
      "loss": 8.9594,
      "mean_token_accuracy": 0.05966283418238163,
      "num_tokens": 239775.0,
      "step": 115
    },
    {
      "entropy": 10.567374324798584,
      "epoch": 0.01088534107402032,
      "grad_norm": 1.9140625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.8042,
      "mean_token_accuracy": 0.06268535517156124,
      "num_tokens": 249954.0,
      "step": 120
    },
    {
      "entropy": 10.49941110610962,
      "epoch": 0.011338896952104499,
      "grad_norm": 1.71875,
      "learning_rate": 6.2e-05,
      "loss": 8.7019,
      "mean_token_accuracy": 0.06550404466688634,
      "num_tokens": 260056.0,
      "step": 125
    },
    {
      "entropy": 10.433360958099366,
      "epoch": 0.01179245283018868,
      "grad_norm": 1.6015625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.5999,
      "mean_token_accuracy": 0.06795549392700195,
      "num_tokens": 269904.0,
      "step": 130
    },
    {
      "entropy": 10.364013862609863,
      "epoch": 0.01224600870827286,
      "grad_norm": 1.703125,
      "learning_rate": 6.7e-05,
      "loss": 8.4687,
      "mean_token_accuracy": 0.06760800294578076,
      "num_tokens": 280900.0,
      "step": 135
    },
    {
      "entropy": 10.2804931640625,
      "epoch": 0.01269956458635704,
      "grad_norm": 1.59375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.3637,
      "mean_token_accuracy": 0.067552350461483,
      "num_tokens": 290844.0,
      "step": 140
    },
    {
      "entropy": 10.172641277313232,
      "epoch": 0.013153120464441219,
      "grad_norm": 1.3046875,
      "learning_rate": 7.2e-05,
      "loss": 8.2257,
      "mean_token_accuracy": 0.07797299772500992,
      "num_tokens": 301345.0,
      "step": 145
    },
    {
      "entropy": 10.0338849067688,
      "epoch": 0.013606676342525399,
      "grad_norm": 1.484375,
      "learning_rate": 7.45e-05,
      "loss": 8.0893,
      "mean_token_accuracy": 0.07597564421594143,
      "num_tokens": 312681.0,
      "step": 150
    },
    {
      "entropy": 9.874010562896729,
      "epoch": 0.01406023222060958,
      "grad_norm": 1.203125,
      "learning_rate": 7.7e-05,
      "loss": 7.9627,
      "mean_token_accuracy": 0.08046026974916458,
      "num_tokens": 323492.0,
      "step": 155
    },
    {
      "entropy": 9.684037780761718,
      "epoch": 0.01451378809869376,
      "grad_norm": 1.234375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.9031,
      "mean_token_accuracy": 0.07810574918985366,
      "num_tokens": 334521.0,
      "step": 160
    },
    {
      "entropy": 9.470422744750977,
      "epoch": 0.014967343976777939,
      "grad_norm": 1.015625,
      "learning_rate": 8.2e-05,
      "loss": 7.8175,
      "mean_token_accuracy": 0.08020211830735206,
      "num_tokens": 345468.0,
      "step": 165
    },
    {
      "entropy": 9.22849702835083,
      "epoch": 0.015420899854862119,
      "grad_norm": 1.1015625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.6913,
      "mean_token_accuracy": 0.07718781009316444,
      "num_tokens": 355911.0,
      "step": 170
    },
    {
      "entropy": 9.006134128570556,
      "epoch": 0.0158744557329463,
      "grad_norm": 1.03125,
      "learning_rate": 8.7e-05,
      "loss": 7.6633,
      "mean_token_accuracy": 0.07722515687346458,
      "num_tokens": 367537.0,
      "step": 175
    },
    {
      "entropy": 8.818270492553712,
      "epoch": 0.01632801161103048,
      "grad_norm": 0.94921875,
      "learning_rate": 8.95e-05,
      "loss": 7.7245,
      "mean_token_accuracy": 0.07779238075017929,
      "num_tokens": 377918.0,
      "step": 180
    },
    {
      "entropy": 8.593470573425293,
      "epoch": 0.01678156748911466,
      "grad_norm": 0.85546875,
      "learning_rate": 9.2e-05,
      "loss": 7.5619,
      "mean_token_accuracy": 0.07740958072245122,
      "num_tokens": 388461.0,
      "step": 185
    },
    {
      "entropy": 8.449785518646241,
      "epoch": 0.01723512336719884,
      "grad_norm": 0.984375,
      "learning_rate": 9.45e-05,
      "loss": 7.5162,
      "mean_token_accuracy": 0.07913253530859947,
      "num_tokens": 399771.0,
      "step": 190
    },
    {
      "entropy": 8.267493534088135,
      "epoch": 0.01768867924528302,
      "grad_norm": 0.8125,
      "learning_rate": 9.7e-05,
      "loss": 7.3798,
      "mean_token_accuracy": 0.08124997988343238,
      "num_tokens": 410254.0,
      "step": 195
    },
    {
      "entropy": 8.230424785614014,
      "epoch": 0.018142235123367198,
      "grad_norm": 0.84765625,
      "learning_rate": 9.95e-05,
      "loss": 7.3917,
      "mean_token_accuracy": 0.07860286422073841,
      "num_tokens": 419792.0,
      "step": 200
    },
    {
      "entropy": 8.091243362426757,
      "epoch": 0.018595791001451378,
      "grad_norm": 0.96875,
      "learning_rate": 0.000102,
      "loss": 7.3759,
      "mean_token_accuracy": 0.08409927114844322,
      "num_tokens": 429264.0,
      "step": 205
    },
    {
      "entropy": 8.078143405914307,
      "epoch": 0.019049346879535557,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.3197,
      "mean_token_accuracy": 0.08545555844902993,
      "num_tokens": 439624.0,
      "step": 210
    },
    {
      "entropy": 8.048319149017335,
      "epoch": 0.01950290275761974,
      "grad_norm": 0.984375,
      "learning_rate": 0.000107,
      "loss": 7.3071,
      "mean_token_accuracy": 0.08367115408182144,
      "num_tokens": 449556.0,
      "step": 215
    },
    {
      "entropy": 7.937560939788819,
      "epoch": 0.01995645863570392,
      "grad_norm": 1.125,
      "learning_rate": 0.0001095,
      "loss": 7.3073,
      "mean_token_accuracy": 0.08580821454524994,
      "num_tokens": 460243.0,
      "step": 220
    },
    {
      "entropy": 7.98973650932312,
      "epoch": 0.0204100145137881,
      "grad_norm": 1.015625,
      "learning_rate": 0.000112,
      "loss": 7.3523,
      "mean_token_accuracy": 0.08153139427304268,
      "num_tokens": 469767.0,
      "step": 225
    },
    {
      "entropy": 7.94960503578186,
      "epoch": 0.02086357039187228,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0001145,
      "loss": 7.2779,
      "mean_token_accuracy": 0.08782213702797889,
      "num_tokens": 480329.0,
      "step": 230
    },
    {
      "entropy": 7.937411499023438,
      "epoch": 0.02131712626995646,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.2332,
      "mean_token_accuracy": 0.08278771415352822,
      "num_tokens": 490096.0,
      "step": 235
    },
    {
      "entropy": 7.9301097869873045,
      "epoch": 0.02177068214804064,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.2751,
      "mean_token_accuracy": 0.0801933728158474,
      "num_tokens": 501231.0,
      "step": 240
    },
    {
      "entropy": 7.883290576934814,
      "epoch": 0.022224238026124818,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000122,
      "loss": 7.1802,
      "mean_token_accuracy": 0.09071151316165924,
      "num_tokens": 510722.0,
      "step": 245
    },
    {
      "entropy": 7.763730192184449,
      "epoch": 0.022677793904208998,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0001245,
      "loss": 7.225,
      "mean_token_accuracy": 0.08499881774187087,
      "num_tokens": 521452.0,
      "step": 250
    },
    {
      "entropy": 7.808817529678345,
      "epoch": 0.023131349782293177,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000127,
      "loss": 7.2137,
      "mean_token_accuracy": 0.08842154070734978,
      "num_tokens": 532397.0,
      "step": 255
    },
    {
      "entropy": 7.740368890762329,
      "epoch": 0.02358490566037736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001295,
      "loss": 7.1927,
      "mean_token_accuracy": 0.08906038105487823,
      "num_tokens": 542271.0,
      "step": 260
    },
    {
      "entropy": 7.7880730628967285,
      "epoch": 0.02403846153846154,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000132,
      "loss": 7.1677,
      "mean_token_accuracy": 0.09267213717103004,
      "num_tokens": 552411.0,
      "step": 265
    },
    {
      "entropy": 7.73106369972229,
      "epoch": 0.02449201741654572,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.1675,
      "mean_token_accuracy": 0.09035667106509208,
      "num_tokens": 561518.0,
      "step": 270
    },
    {
      "entropy": 7.755526113510132,
      "epoch": 0.0249455732946299,
      "grad_norm": 1.046875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1846,
      "mean_token_accuracy": 0.09218391999602318,
      "num_tokens": 571361.0,
      "step": 275
    },
    {
      "entropy": 7.656434154510498,
      "epoch": 0.02539912917271408,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0001395,
      "loss": 7.0893,
      "mean_token_accuracy": 0.09728347510099411,
      "num_tokens": 580926.0,
      "step": 280
    },
    {
      "entropy": 7.710359334945679,
      "epoch": 0.025852685050798258,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1891,
      "mean_token_accuracy": 0.08993546590209008,
      "num_tokens": 592282.0,
      "step": 285
    },
    {
      "entropy": 7.69242033958435,
      "epoch": 0.026306240928882438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001445,
      "loss": 7.1904,
      "mean_token_accuracy": 0.09070137366652489,
      "num_tokens": 602767.0,
      "step": 290
    },
    {
      "entropy": 7.715097570419312,
      "epoch": 0.026759796806966617,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000147,
      "loss": 7.1064,
      "mean_token_accuracy": 0.0925739623606205,
      "num_tokens": 612763.0,
      "step": 295
    },
    {
      "entropy": 7.563499784469604,
      "epoch": 0.027213352685050797,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0001495,
      "loss": 7.0617,
      "mean_token_accuracy": 0.09330770447850227,
      "num_tokens": 622796.0,
      "step": 300
    },
    {
      "entropy": 7.630935192108154,
      "epoch": 0.027666908563134977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000152,
      "loss": 7.0426,
      "mean_token_accuracy": 0.09754194170236588,
      "num_tokens": 633136.0,
      "step": 305
    },
    {
      "entropy": 7.510959100723267,
      "epoch": 0.02812046444121916,
      "grad_norm": 1.109375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0335,
      "mean_token_accuracy": 0.09690061807632447,
      "num_tokens": 643347.0,
      "step": 310
    },
    {
      "entropy": 7.634088945388794,
      "epoch": 0.02857402031930334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000157,
      "loss": 7.0192,
      "mean_token_accuracy": 0.10197105258703232,
      "num_tokens": 653096.0,
      "step": 315
    },
    {
      "entropy": 7.5339111328125,
      "epoch": 0.02902757619738752,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001595,
      "loss": 7.0519,
      "mean_token_accuracy": 0.09302655458450318,
      "num_tokens": 663557.0,
      "step": 320
    },
    {
      "entropy": 7.588706541061401,
      "epoch": 0.0294811320754717,
      "grad_norm": 1.515625,
      "learning_rate": 0.000162,
      "loss": 7.1186,
      "mean_token_accuracy": 0.0957950733602047,
      "num_tokens": 674395.0,
      "step": 325
    },
    {
      "entropy": 7.544366073608399,
      "epoch": 0.029934687953555878,
      "grad_norm": 1.140625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0165,
      "mean_token_accuracy": 0.0989912673830986,
      "num_tokens": 684776.0,
      "step": 330
    },
    {
      "entropy": 7.589376211166382,
      "epoch": 0.030388243831640058,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9936,
      "mean_token_accuracy": 0.09702750965952874,
      "num_tokens": 694886.0,
      "step": 335
    },
    {
      "entropy": 7.451715087890625,
      "epoch": 0.030841799709724237,
      "grad_norm": 1.234375,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0626,
      "mean_token_accuracy": 0.09623553976416588,
      "num_tokens": 705426.0,
      "step": 340
    },
    {
      "entropy": 7.509190368652344,
      "epoch": 0.03129535558780842,
      "grad_norm": 0.984375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.0567,
      "mean_token_accuracy": 0.09196493923664092,
      "num_tokens": 716271.0,
      "step": 345
    },
    {
      "entropy": 7.5583083152771,
      "epoch": 0.0317489114658926,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0486,
      "mean_token_accuracy": 0.09727451354265212,
      "num_tokens": 726875.0,
      "step": 350
    },
    {
      "entropy": 7.49442138671875,
      "epoch": 0.032202467343976776,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000177,
      "loss": 6.9698,
      "mean_token_accuracy": 0.09514860436320305,
      "num_tokens": 737825.0,
      "step": 355
    },
    {
      "entropy": 7.501305818557739,
      "epoch": 0.03265602322206096,
      "grad_norm": 0.953125,
      "learning_rate": 0.0001795,
      "loss": 6.9607,
      "mean_token_accuracy": 0.0989458367228508,
      "num_tokens": 747424.0,
      "step": 360
    },
    {
      "entropy": 7.407667875289917,
      "epoch": 0.033109579100145135,
      "grad_norm": 1.046875,
      "learning_rate": 0.000182,
      "loss": 6.9748,
      "mean_token_accuracy": 0.09646827578544617,
      "num_tokens": 757649.0,
      "step": 365
    },
    {
      "entropy": 7.480729293823242,
      "epoch": 0.03356313497822932,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001845,
      "loss": 6.946,
      "mean_token_accuracy": 0.09825383350253106,
      "num_tokens": 768556.0,
      "step": 370
    },
    {
      "entropy": 7.410275650024414,
      "epoch": 0.034016690856313495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000187,
      "loss": 6.9456,
      "mean_token_accuracy": 0.0978932335972786,
      "num_tokens": 778331.0,
      "step": 375
    },
    {
      "entropy": 7.457346200942993,
      "epoch": 0.03447024673439768,
      "grad_norm": 1.140625,
      "learning_rate": 0.0001895,
      "loss": 6.9326,
      "mean_token_accuracy": 0.10083384290337563,
      "num_tokens": 788065.0,
      "step": 380
    },
    {
      "entropy": 7.39845495223999,
      "epoch": 0.03492380261248186,
      "grad_norm": 1.078125,
      "learning_rate": 0.000192,
      "loss": 6.9717,
      "mean_token_accuracy": 0.10045002475380897,
      "num_tokens": 798912.0,
      "step": 385
    },
    {
      "entropy": 7.406559228897095,
      "epoch": 0.03537735849056604,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0001945,
      "loss": 6.9143,
      "mean_token_accuracy": 0.10309505760669709,
      "num_tokens": 809580.0,
      "step": 390
    },
    {
      "entropy": 7.444182062149048,
      "epoch": 0.03583091436865022,
      "grad_norm": 1.015625,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.0008,
      "mean_token_accuracy": 0.09738668501377105,
      "num_tokens": 820207.0,
      "step": 395
    },
    {
      "entropy": 7.290267753601074,
      "epoch": 0.036284470246734396,
      "grad_norm": 0.984375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8228,
      "mean_token_accuracy": 0.0949432410299778,
      "num_tokens": 831038.0,
      "step": 400
    },
    {
      "entropy": 7.429319906234741,
      "epoch": 0.03673802612481858,
      "grad_norm": 0.90625,
      "learning_rate": 0.000202,
      "loss": 6.9129,
      "mean_token_accuracy": 0.10030955821275711,
      "num_tokens": 842144.0,
      "step": 405
    },
    {
      "entropy": 7.347342252731323,
      "epoch": 0.037191582002902755,
      "grad_norm": 1.015625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.905,
      "mean_token_accuracy": 0.10373404398560523,
      "num_tokens": 851315.0,
      "step": 410
    },
    {
      "entropy": 7.323047733306884,
      "epoch": 0.03764513788098694,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000207,
      "loss": 6.912,
      "mean_token_accuracy": 0.10021007284522057,
      "num_tokens": 861233.0,
      "step": 415
    },
    {
      "entropy": 7.350712013244629,
      "epoch": 0.038098693759071114,
      "grad_norm": 1.0,
      "learning_rate": 0.0002095,
      "loss": 6.8978,
      "mean_token_accuracy": 0.09711670130491257,
      "num_tokens": 871131.0,
      "step": 420
    },
    {
      "entropy": 7.294861841201782,
      "epoch": 0.0385522496371553,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000212,
      "loss": 6.8749,
      "mean_token_accuracy": 0.09577203691005706,
      "num_tokens": 882195.0,
      "step": 425
    },
    {
      "entropy": 7.353318929672241,
      "epoch": 0.03900580551523948,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002145,
      "loss": 6.8527,
      "mean_token_accuracy": 0.10149310603737831,
      "num_tokens": 891806.0,
      "step": 430
    },
    {
      "entropy": 7.362502574920654,
      "epoch": 0.03945936139332366,
      "grad_norm": 0.953125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.9129,
      "mean_token_accuracy": 0.09761691987514495,
      "num_tokens": 901350.0,
      "step": 435
    },
    {
      "entropy": 7.301689004898071,
      "epoch": 0.03991291727140784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002195,
      "loss": 6.8293,
      "mean_token_accuracy": 0.09842056706547737,
      "num_tokens": 912061.0,
      "step": 440
    },
    {
      "entropy": 7.250009202957154,
      "epoch": 0.040366473149492016,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000222,
      "loss": 6.7954,
      "mean_token_accuracy": 0.09848268181085587,
      "num_tokens": 922220.0,
      "step": 445
    },
    {
      "entropy": 7.2643632888793945,
      "epoch": 0.0408200290275762,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002245,
      "loss": 6.8122,
      "mean_token_accuracy": 0.1012580119073391,
      "num_tokens": 932319.0,
      "step": 450
    },
    {
      "entropy": 7.2017487525939945,
      "epoch": 0.041273584905660375,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8351,
      "mean_token_accuracy": 0.10266126319766045,
      "num_tokens": 943254.0,
      "step": 455
    },
    {
      "entropy": 7.330859661102295,
      "epoch": 0.04172714078374456,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.8445,
      "mean_token_accuracy": 0.09818541109561921,
      "num_tokens": 953921.0,
      "step": 460
    },
    {
      "entropy": 7.193669176101684,
      "epoch": 0.042180696661828734,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7701,
      "mean_token_accuracy": 0.10723233297467231,
      "num_tokens": 963832.0,
      "step": 465
    },
    {
      "entropy": 7.234699535369873,
      "epoch": 0.04263425253991292,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.8614,
      "mean_token_accuracy": 0.09933211356401443,
      "num_tokens": 974066.0,
      "step": 470
    },
    {
      "entropy": 7.233931684494019,
      "epoch": 0.0430878084179971,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000237,
      "loss": 6.8441,
      "mean_token_accuracy": 0.10467069000005721,
      "num_tokens": 984402.0,
      "step": 475
    },
    {
      "entropy": 7.256547212600708,
      "epoch": 0.04354136429608128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002395,
      "loss": 6.7629,
      "mean_token_accuracy": 0.1010866865515709,
      "num_tokens": 993555.0,
      "step": 480
    },
    {
      "entropy": 7.208513498306274,
      "epoch": 0.04399492017416546,
      "grad_norm": 1.015625,
      "learning_rate": 0.000242,
      "loss": 6.7524,
      "mean_token_accuracy": 0.10970519632101058,
      "num_tokens": 1002920.0,
      "step": 485
    },
    {
      "entropy": 7.203861713409424,
      "epoch": 0.044448476052249636,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002445,
      "loss": 6.8285,
      "mean_token_accuracy": 0.09806114360690117,
      "num_tokens": 1013611.0,
      "step": 490
    },
    {
      "entropy": 7.1150740623474125,
      "epoch": 0.04490203193033382,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000247,
      "loss": 6.7083,
      "mean_token_accuracy": 0.10193079337477684,
      "num_tokens": 1024121.0,
      "step": 495
    },
    {
      "entropy": 7.323940992355347,
      "epoch": 0.045355587808417995,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002495,
      "loss": 6.9503,
      "mean_token_accuracy": 0.10487301424145698,
      "num_tokens": 1033845.0,
      "step": 500
    },
    {
      "entropy": 7.164136552810669,
      "epoch": 0.04580914368650218,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000252,
      "loss": 6.7388,
      "mean_token_accuracy": 0.10410365164279937,
      "num_tokens": 1044642.0,
      "step": 505
    },
    {
      "entropy": 7.172935676574707,
      "epoch": 0.046262699564586354,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002545,
      "loss": 6.8158,
      "mean_token_accuracy": 0.10116696879267692,
      "num_tokens": 1054728.0,
      "step": 510
    },
    {
      "entropy": 7.138863849639892,
      "epoch": 0.04671625544267054,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000257,
      "loss": 6.7643,
      "mean_token_accuracy": 0.10025905445218086,
      "num_tokens": 1065743.0,
      "step": 515
    },
    {
      "entropy": 7.10414080619812,
      "epoch": 0.04716981132075472,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002595,
      "loss": 6.6969,
      "mean_token_accuracy": 0.10323216244578362,
      "num_tokens": 1076798.0,
      "step": 520
    },
    {
      "entropy": 7.174085855484009,
      "epoch": 0.047623367198838897,
      "grad_norm": 1.03125,
      "learning_rate": 0.000262,
      "loss": 6.8163,
      "mean_token_accuracy": 0.0997597634792328,
      "num_tokens": 1087445.0,
      "step": 525
    },
    {
      "entropy": 7.094452857971191,
      "epoch": 0.04807692307692308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7197,
      "mean_token_accuracy": 0.10675485357642174,
      "num_tokens": 1098379.0,
      "step": 530
    },
    {
      "entropy": 7.202598142623901,
      "epoch": 0.048530478955007256,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.8289,
      "mean_token_accuracy": 0.0993705414235592,
      "num_tokens": 1108016.0,
      "step": 535
    },
    {
      "entropy": 7.089155960083008,
      "epoch": 0.04898403483309144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.701,
      "mean_token_accuracy": 0.10358337983489037,
      "num_tokens": 1117887.0,
      "step": 540
    },
    {
      "entropy": 7.104513120651245,
      "epoch": 0.049437590711175615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6948,
      "mean_token_accuracy": 0.09933766350150108,
      "num_tokens": 1128497.0,
      "step": 545
    },
    {
      "entropy": 7.055957365036011,
      "epoch": 0.0498911465892598,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002745,
      "loss": 6.7188,
      "mean_token_accuracy": 0.10837908834218979,
      "num_tokens": 1138168.0,
      "step": 550
    },
    {
      "entropy": 7.0923104763031,
      "epoch": 0.050344702467343974,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000277,
      "loss": 6.691,
      "mean_token_accuracy": 0.10448997020721436,
      "num_tokens": 1148357.0,
      "step": 555
    },
    {
      "entropy": 7.086644268035888,
      "epoch": 0.05079825834542816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002795,
      "loss": 6.7185,
      "mean_token_accuracy": 0.10321378186345101,
      "num_tokens": 1159964.0,
      "step": 560
    },
    {
      "entropy": 7.139619493484497,
      "epoch": 0.05125181422351233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.7781,
      "mean_token_accuracy": 0.10231135338544846,
      "num_tokens": 1170683.0,
      "step": 565
    },
    {
      "entropy": 7.115566444396973,
      "epoch": 0.051705370101596516,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002845,
      "loss": 6.7343,
      "mean_token_accuracy": 0.10421703606843949,
      "num_tokens": 1181680.0,
      "step": 570
    },
    {
      "entropy": 7.0173659324646,
      "epoch": 0.0521589259796807,
      "grad_norm": 1.078125,
      "learning_rate": 0.000287,
      "loss": 6.6227,
      "mean_token_accuracy": 0.10645965486764908,
      "num_tokens": 1191619.0,
      "step": 575
    },
    {
      "entropy": 6.977677965164185,
      "epoch": 0.052612481857764876,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002895,
      "loss": 6.6222,
      "mean_token_accuracy": 0.10420064777135848,
      "num_tokens": 1202264.0,
      "step": 580
    },
    {
      "entropy": 7.002698135375977,
      "epoch": 0.05306603773584906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000292,
      "loss": 6.6254,
      "mean_token_accuracy": 0.10796518102288247,
      "num_tokens": 1212119.0,
      "step": 585
    },
    {
      "entropy": 7.092545223236084,
      "epoch": 0.053519593613933235,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002945,
      "loss": 6.7588,
      "mean_token_accuracy": 0.10538023859262466,
      "num_tokens": 1222317.0,
      "step": 590
    },
    {
      "entropy": 7.060644435882568,
      "epoch": 0.05397314949201742,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000297,
      "loss": 6.703,
      "mean_token_accuracy": 0.1080887496471405,
      "num_tokens": 1232914.0,
      "step": 595
    },
    {
      "entropy": 6.919224500656128,
      "epoch": 0.054426705370101594,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002995,
      "loss": 6.5519,
      "mean_token_accuracy": 0.11217400431632996,
      "num_tokens": 1244214.0,
      "step": 600
    },
    {
      "entropy": 7.022872066497802,
      "epoch": 0.05488026124818578,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000302,
      "loss": 6.6491,
      "mean_token_accuracy": 0.10687957927584649,
      "num_tokens": 1254626.0,
      "step": 605
    },
    {
      "entropy": 6.948216915130615,
      "epoch": 0.05533381712626995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003045,
      "loss": 6.7394,
      "mean_token_accuracy": 0.10631354376673699,
      "num_tokens": 1265461.0,
      "step": 610
    },
    {
      "entropy": 7.030997228622437,
      "epoch": 0.055787373004354136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000307,
      "loss": 6.6204,
      "mean_token_accuracy": 0.10827788710594177,
      "num_tokens": 1275810.0,
      "step": 615
    },
    {
      "entropy": 6.895572900772095,
      "epoch": 0.05624092888243832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003095,
      "loss": 6.6276,
      "mean_token_accuracy": 0.11254601404070855,
      "num_tokens": 1285170.0,
      "step": 620
    },
    {
      "entropy": 7.06603364944458,
      "epoch": 0.056694484760522496,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000312,
      "loss": 6.5641,
      "mean_token_accuracy": 0.11045938208699227,
      "num_tokens": 1295460.0,
      "step": 625
    },
    {
      "entropy": 6.950644588470459,
      "epoch": 0.05714804063860668,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003145,
      "loss": 6.687,
      "mean_token_accuracy": 0.10515375509858131,
      "num_tokens": 1306656.0,
      "step": 630
    },
    {
      "entropy": 7.004617834091187,
      "epoch": 0.057601596516690855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000317,
      "loss": 6.6603,
      "mean_token_accuracy": 0.10941670313477517,
      "num_tokens": 1318019.0,
      "step": 635
    },
    {
      "entropy": 6.8719371318817135,
      "epoch": 0.05805515239477504,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003195,
      "loss": 6.5802,
      "mean_token_accuracy": 0.11324906200170518,
      "num_tokens": 1328120.0,
      "step": 640
    },
    {
      "entropy": 7.002581357955933,
      "epoch": 0.058508708272859214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000322,
      "loss": 6.6842,
      "mean_token_accuracy": 0.10679537355899811,
      "num_tokens": 1338354.0,
      "step": 645
    },
    {
      "entropy": 7.0219018936157225,
      "epoch": 0.0589622641509434,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6503,
      "mean_token_accuracy": 0.10833998396992683,
      "num_tokens": 1348666.0,
      "step": 650
    },
    {
      "entropy": 6.952841186523438,
      "epoch": 0.05941582002902757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.6921,
      "mean_token_accuracy": 0.11318251714110375,
      "num_tokens": 1359713.0,
      "step": 655
    },
    {
      "entropy": 6.950813102722168,
      "epoch": 0.059869375907111756,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5549,
      "mean_token_accuracy": 0.11367055401206017,
      "num_tokens": 1369507.0,
      "step": 660
    },
    {
      "entropy": 6.984795665740966,
      "epoch": 0.06032293178519594,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.7111,
      "mean_token_accuracy": 0.10242682620882988,
      "num_tokens": 1380726.0,
      "step": 665
    },
    {
      "entropy": 7.01419677734375,
      "epoch": 0.060776487663280115,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.6325,
      "mean_token_accuracy": 0.11175172179937362,
      "num_tokens": 1391423.0,
      "step": 670
    },
    {
      "entropy": 6.930062580108642,
      "epoch": 0.0612300435413643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000337,
      "loss": 6.6054,
      "mean_token_accuracy": 0.11288354322314262,
      "num_tokens": 1402492.0,
      "step": 675
    },
    {
      "entropy": 6.939368772506714,
      "epoch": 0.061683599419448475,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003395,
      "loss": 6.6504,
      "mean_token_accuracy": 0.1099183864891529,
      "num_tokens": 1412497.0,
      "step": 680
    },
    {
      "entropy": 6.999884557723999,
      "epoch": 0.06213715529753266,
      "grad_norm": 1.15625,
      "learning_rate": 0.000342,
      "loss": 6.6204,
      "mean_token_accuracy": 0.11455832123756408,
      "num_tokens": 1423214.0,
      "step": 685
    },
    {
      "entropy": 6.856761169433594,
      "epoch": 0.06259071117561683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.6202,
      "mean_token_accuracy": 0.11103508621454239,
      "num_tokens": 1434044.0,
      "step": 690
    },
    {
      "entropy": 6.9463762760162355,
      "epoch": 0.06304426705370102,
      "grad_norm": 1.015625,
      "learning_rate": 0.000347,
      "loss": 6.5826,
      "mean_token_accuracy": 0.11131571531295777,
      "num_tokens": 1444387.0,
      "step": 695
    },
    {
      "entropy": 6.848030471801758,
      "epoch": 0.0634978229317852,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003495,
      "loss": 6.5282,
      "mean_token_accuracy": 0.10915688425302505,
      "num_tokens": 1455157.0,
      "step": 700
    },
    {
      "entropy": 6.915933418273926,
      "epoch": 0.06395137880986937,
      "grad_norm": 1.0,
      "learning_rate": 0.000352,
      "loss": 6.6147,
      "mean_token_accuracy": 0.11529198288917542,
      "num_tokens": 1466145.0,
      "step": 705
    },
    {
      "entropy": 6.937824773788452,
      "epoch": 0.06440493468795355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003545,
      "loss": 6.6053,
      "mean_token_accuracy": 0.1114728607237339,
      "num_tokens": 1476713.0,
      "step": 710
    },
    {
      "entropy": 6.905196857452393,
      "epoch": 0.06485849056603774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000357,
      "loss": 6.6113,
      "mean_token_accuracy": 0.11422201618552208,
      "num_tokens": 1486908.0,
      "step": 715
    },
    {
      "entropy": 6.866687297821045,
      "epoch": 0.06531204644412192,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003595,
      "loss": 6.5389,
      "mean_token_accuracy": 0.11563119068741798,
      "num_tokens": 1498185.0,
      "step": 720
    },
    {
      "entropy": 6.873838233947754,
      "epoch": 0.0657656023222061,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000362,
      "loss": 6.5814,
      "mean_token_accuracy": 0.10893168300390244,
      "num_tokens": 1508376.0,
      "step": 725
    },
    {
      "entropy": 6.905275201797485,
      "epoch": 0.06621915820029027,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003645,
      "loss": 6.6695,
      "mean_token_accuracy": 0.11235518902540206,
      "num_tokens": 1519953.0,
      "step": 730
    },
    {
      "entropy": 6.795944786071777,
      "epoch": 0.06667271407837445,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000367,
      "loss": 6.5408,
      "mean_token_accuracy": 0.11527841538190842,
      "num_tokens": 1529793.0,
      "step": 735
    },
    {
      "entropy": 6.8445906162261965,
      "epoch": 0.06712626995645864,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003695,
      "loss": 6.5363,
      "mean_token_accuracy": 0.11552815213799476,
      "num_tokens": 1540859.0,
      "step": 740
    },
    {
      "entropy": 6.8253929138183596,
      "epoch": 0.06757982583454282,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000372,
      "loss": 6.4732,
      "mean_token_accuracy": 0.11656396687030793,
      "num_tokens": 1549783.0,
      "step": 745
    },
    {
      "entropy": 6.756074047088623,
      "epoch": 0.06803338171262699,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003745,
      "loss": 6.5266,
      "mean_token_accuracy": 0.11289834827184678,
      "num_tokens": 1560522.0,
      "step": 750
    },
    {
      "entropy": 6.810197973251343,
      "epoch": 0.06848693759071117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000377,
      "loss": 6.5008,
      "mean_token_accuracy": 0.11671004444360733,
      "num_tokens": 1570668.0,
      "step": 755
    },
    {
      "entropy": 6.83150315284729,
      "epoch": 0.06894049346879536,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003795,
      "loss": 6.6122,
      "mean_token_accuracy": 0.1193370096385479,
      "num_tokens": 1581523.0,
      "step": 760
    },
    {
      "entropy": 6.850863075256347,
      "epoch": 0.06939404934687954,
      "grad_norm": 1.125,
      "learning_rate": 0.000382,
      "loss": 6.5253,
      "mean_token_accuracy": 0.11707687377929688,
      "num_tokens": 1591356.0,
      "step": 765
    },
    {
      "entropy": 6.784787607192993,
      "epoch": 0.06984760522496372,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003845,
      "loss": 6.5113,
      "mean_token_accuracy": 0.11673457846045494,
      "num_tokens": 1601779.0,
      "step": 770
    },
    {
      "entropy": 6.848218965530395,
      "epoch": 0.07030116110304789,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.5091,
      "mean_token_accuracy": 0.11470222994685172,
      "num_tokens": 1611995.0,
      "step": 775
    },
    {
      "entropy": 6.800413322448731,
      "epoch": 0.07075471698113207,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.5369,
      "mean_token_accuracy": 0.11278748363256455,
      "num_tokens": 1621834.0,
      "step": 780
    },
    {
      "entropy": 6.780804920196533,
      "epoch": 0.07120827285921626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.5788,
      "mean_token_accuracy": 0.1113444909453392,
      "num_tokens": 1632881.0,
      "step": 785
    },
    {
      "entropy": 6.707678127288818,
      "epoch": 0.07166182873730044,
      "grad_norm": 1.0,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.5541,
      "mean_token_accuracy": 0.11073025166988373,
      "num_tokens": 1642870.0,
      "step": 790
    },
    {
      "entropy": 6.871548318862915,
      "epoch": 0.07211538461538461,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4402,
      "mean_token_accuracy": 0.12203826904296874,
      "num_tokens": 1653390.0,
      "step": 795
    },
    {
      "entropy": 6.7886035442352295,
      "epoch": 0.07256894049346879,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003995,
      "loss": 6.5612,
      "mean_token_accuracy": 0.11627909317612647,
      "num_tokens": 1663382.0,
      "step": 800
    },
    {
      "entropy": 6.804583787918091,
      "epoch": 0.07302249637155298,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000402,
      "loss": 6.5652,
      "mean_token_accuracy": 0.1136283002793789,
      "num_tokens": 1673601.0,
      "step": 805
    },
    {
      "entropy": 6.703377294540405,
      "epoch": 0.07347605224963716,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004045,
      "loss": 6.4891,
      "mean_token_accuracy": 0.12096939235925674,
      "num_tokens": 1683565.0,
      "step": 810
    },
    {
      "entropy": 6.758061265945434,
      "epoch": 0.07392960812772134,
      "grad_norm": 1.78125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.45,
      "mean_token_accuracy": 0.11798930242657661,
      "num_tokens": 1693636.0,
      "step": 815
    },
    {
      "entropy": 6.882167196273803,
      "epoch": 0.07438316400580551,
      "grad_norm": 1.125,
      "learning_rate": 0.0004095,
      "loss": 6.5623,
      "mean_token_accuracy": 0.11476152688264847,
      "num_tokens": 1703415.0,
      "step": 820
    },
    {
      "entropy": 6.755157995223999,
      "epoch": 0.0748367198838897,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000412,
      "loss": 6.5396,
      "mean_token_accuracy": 0.11611356288194656,
      "num_tokens": 1713579.0,
      "step": 825
    },
    {
      "entropy": 6.7464526176452635,
      "epoch": 0.07529027576197388,
      "grad_norm": 1.0,
      "learning_rate": 0.0004145,
      "loss": 6.5394,
      "mean_token_accuracy": 0.11386027708649635,
      "num_tokens": 1724583.0,
      "step": 830
    },
    {
      "entropy": 6.758458185195923,
      "epoch": 0.07574383164005806,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000417,
      "loss": 6.4527,
      "mean_token_accuracy": 0.12356605678796768,
      "num_tokens": 1735293.0,
      "step": 835
    },
    {
      "entropy": 6.7750757217407225,
      "epoch": 0.07619738751814223,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004195,
      "loss": 6.4786,
      "mean_token_accuracy": 0.12075236141681671,
      "num_tokens": 1745623.0,
      "step": 840
    },
    {
      "entropy": 6.725871658325195,
      "epoch": 0.07665094339622641,
      "grad_norm": 0.96875,
      "learning_rate": 0.000422,
      "loss": 6.5145,
      "mean_token_accuracy": 0.11627071723341942,
      "num_tokens": 1755713.0,
      "step": 845
    },
    {
      "entropy": 6.776280164718628,
      "epoch": 0.0771044992743106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004245,
      "loss": 6.506,
      "mean_token_accuracy": 0.11393775343894959,
      "num_tokens": 1765264.0,
      "step": 850
    },
    {
      "entropy": 6.824870538711548,
      "epoch": 0.07755805515239478,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000427,
      "loss": 6.511,
      "mean_token_accuracy": 0.11867988780140877,
      "num_tokens": 1775769.0,
      "step": 855
    },
    {
      "entropy": 6.691818761825561,
      "epoch": 0.07801161103047896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004295,
      "loss": 6.4441,
      "mean_token_accuracy": 0.11789206489920616,
      "num_tokens": 1786245.0,
      "step": 860
    },
    {
      "entropy": 6.627203035354614,
      "epoch": 0.07846516690856313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000432,
      "loss": 6.4773,
      "mean_token_accuracy": 0.1227649137377739,
      "num_tokens": 1796040.0,
      "step": 865
    },
    {
      "entropy": 6.691270112991333,
      "epoch": 0.07891872278664731,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004345,
      "loss": 6.4935,
      "mean_token_accuracy": 0.11697295233607292,
      "num_tokens": 1806007.0,
      "step": 870
    },
    {
      "entropy": 6.88080358505249,
      "epoch": 0.0793722786647315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000437,
      "loss": 6.5665,
      "mean_token_accuracy": 0.10995617434382439,
      "num_tokens": 1816935.0,
      "step": 875
    },
    {
      "entropy": 6.631897354125977,
      "epoch": 0.07982583454281568,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004395,
      "loss": 6.4244,
      "mean_token_accuracy": 0.11615645289421081,
      "num_tokens": 1827419.0,
      "step": 880
    },
    {
      "entropy": 6.766531658172608,
      "epoch": 0.08027939042089985,
      "grad_norm": 1.15625,
      "learning_rate": 0.000442,
      "loss": 6.533,
      "mean_token_accuracy": 0.11622850596904755,
      "num_tokens": 1837933.0,
      "step": 885
    },
    {
      "entropy": 6.714184427261353,
      "epoch": 0.08073294629898403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004445,
      "loss": 6.5854,
      "mean_token_accuracy": 0.11893809214234352,
      "num_tokens": 1848551.0,
      "step": 890
    },
    {
      "entropy": 6.7861744403839115,
      "epoch": 0.08118650217706821,
      "grad_norm": 1.125,
      "learning_rate": 0.000447,
      "loss": 6.4577,
      "mean_token_accuracy": 0.11503675356507301,
      "num_tokens": 1858380.0,
      "step": 895
    },
    {
      "entropy": 6.660554695129394,
      "epoch": 0.0816400580551524,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3853,
      "mean_token_accuracy": 0.11978982239961625,
      "num_tokens": 1868291.0,
      "step": 900
    },
    {
      "entropy": 6.629644966125488,
      "epoch": 0.08209361393323658,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.4839,
      "mean_token_accuracy": 0.11984607204794884,
      "num_tokens": 1879219.0,
      "step": 905
    },
    {
      "entropy": 6.754321718215943,
      "epoch": 0.08254716981132075,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4652,
      "mean_token_accuracy": 0.1155819833278656,
      "num_tokens": 1889809.0,
      "step": 910
    },
    {
      "entropy": 6.659244441986084,
      "epoch": 0.08300072568940493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4349,
      "mean_token_accuracy": 0.11736063361167907,
      "num_tokens": 1899986.0,
      "step": 915
    },
    {
      "entropy": 6.727784872055054,
      "epoch": 0.08345428156748912,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4192,
      "mean_token_accuracy": 0.11866439804434777,
      "num_tokens": 1909314.0,
      "step": 920
    },
    {
      "entropy": 6.709451198577881,
      "epoch": 0.0839078374455733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000462,
      "loss": 6.4667,
      "mean_token_accuracy": 0.1235547974705696,
      "num_tokens": 1918443.0,
      "step": 925
    },
    {
      "entropy": 6.680801582336426,
      "epoch": 0.08436139332365747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004645,
      "loss": 6.4235,
      "mean_token_accuracy": 0.12087266743183137,
      "num_tokens": 1928683.0,
      "step": 930
    },
    {
      "entropy": 6.541482257843017,
      "epoch": 0.08481494920174165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000467,
      "loss": 6.3312,
      "mean_token_accuracy": 0.11764199584722519,
      "num_tokens": 1938293.0,
      "step": 935
    },
    {
      "entropy": 6.629144906997681,
      "epoch": 0.08526850507982583,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004695,
      "loss": 6.4449,
      "mean_token_accuracy": 0.11702702939510345,
      "num_tokens": 1948233.0,
      "step": 940
    },
    {
      "entropy": 6.660422420501709,
      "epoch": 0.08572206095791002,
      "grad_norm": 1.046875,
      "learning_rate": 0.000472,
      "loss": 6.3444,
      "mean_token_accuracy": 0.12124838382005691,
      "num_tokens": 1958611.0,
      "step": 945
    },
    {
      "entropy": 6.661191082000732,
      "epoch": 0.0861756168359942,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004745,
      "loss": 6.4983,
      "mean_token_accuracy": 0.1220276914536953,
      "num_tokens": 1968957.0,
      "step": 950
    },
    {
      "entropy": 6.65334038734436,
      "epoch": 0.08662917271407837,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477,
      "loss": 6.3551,
      "mean_token_accuracy": 0.12157930657267571,
      "num_tokens": 1978503.0,
      "step": 955
    },
    {
      "entropy": 6.552236032485962,
      "epoch": 0.08708272859216255,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004795,
      "loss": 6.3482,
      "mean_token_accuracy": 0.12029609382152558,
      "num_tokens": 1988458.0,
      "step": 960
    },
    {
      "entropy": 6.655647468566895,
      "epoch": 0.08753628447024674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000482,
      "loss": 6.491,
      "mean_token_accuracy": 0.11715313494205475,
      "num_tokens": 1999984.0,
      "step": 965
    },
    {
      "entropy": 6.510547399520874,
      "epoch": 0.08798984034833092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004845,
      "loss": 6.3364,
      "mean_token_accuracy": 0.12651536986231804,
      "num_tokens": 2010645.0,
      "step": 970
    },
    {
      "entropy": 6.751053190231323,
      "epoch": 0.08844339622641509,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000487,
      "loss": 6.4506,
      "mean_token_accuracy": 0.11614105999469757,
      "num_tokens": 2020196.0,
      "step": 975
    },
    {
      "entropy": 6.587527418136597,
      "epoch": 0.08889695210449927,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004895,
      "loss": 6.3423,
      "mean_token_accuracy": 0.1243398666381836,
      "num_tokens": 2030101.0,
      "step": 980
    },
    {
      "entropy": 6.5954345703125,
      "epoch": 0.08935050798258345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492,
      "loss": 6.4263,
      "mean_token_accuracy": 0.1206196092069149,
      "num_tokens": 2040454.0,
      "step": 985
    },
    {
      "entropy": 6.529821062088013,
      "epoch": 0.08980406386066764,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945,
      "loss": 6.3602,
      "mean_token_accuracy": 0.11969397887587548,
      "num_tokens": 2050378.0,
      "step": 990
    },
    {
      "entropy": 6.656999588012695,
      "epoch": 0.09025761973875182,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497,
      "loss": 6.4186,
      "mean_token_accuracy": 0.11806954741477967,
      "num_tokens": 2060282.0,
      "step": 995
    },
    {
      "entropy": 6.559673261642456,
      "epoch": 0.09071117561683599,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995,
      "loss": 6.4115,
      "mean_token_accuracy": 0.12180240452289581,
      "num_tokens": 2070563.0,
      "step": 1000
    },
    {
      "entropy": 6.661390542984009,
      "epoch": 0.09116473149492017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999985110217,
      "loss": 6.4743,
      "mean_token_accuracy": 0.12212796434760094,
      "num_tokens": 2080153.0,
      "step": 1005
    },
    {
      "entropy": 6.525332069396972,
      "epoch": 0.09161828737300436,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999999924620471,
      "loss": 6.3263,
      "mean_token_accuracy": 0.12526472061872482,
      "num_tokens": 2090590.0,
      "step": 1010
    },
    {
      "entropy": 6.613137102127075,
      "epoch": 0.09207184325108854,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999999817600155,
      "loss": 6.4047,
      "mean_token_accuracy": 0.11499982699751854,
      "num_tokens": 2101945.0,
      "step": 1015
    },
    {
      "entropy": 6.613913679122925,
      "epoch": 0.09252539912917271,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999999664049268,
      "loss": 6.4031,
      "mean_token_accuracy": 0.12405626773834229,
      "num_tokens": 2112514.0,
      "step": 1020
    },
    {
      "entropy": 6.498817586898804,
      "epoch": 0.09297895500725689,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999999463967816,
      "loss": 6.3365,
      "mean_token_accuracy": 0.1212070882320404,
      "num_tokens": 2122296.0,
      "step": 1025
    },
    {
      "entropy": 6.740924310684204,
      "epoch": 0.09343251088534107,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049999992173558,
      "loss": 6.4899,
      "mean_token_accuracy": 0.11208800077438355,
      "num_tokens": 2134512.0,
      "step": 1030
    },
    {
      "entropy": 6.578291606903076,
      "epoch": 0.09388606676342526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999998924213228,
      "loss": 6.4201,
      "mean_token_accuracy": 0.12014863640069962,
      "num_tokens": 2143934.0,
      "step": 1035
    },
    {
      "entropy": 6.489139556884766,
      "epoch": 0.09433962264150944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999998584540105,
      "loss": 6.3024,
      "mean_token_accuracy": 0.12084058821201324,
      "num_tokens": 2155730.0,
      "step": 1040
    },
    {
      "entropy": 6.6251708984375,
      "epoch": 0.09479317851959361,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999998198336437,
      "loss": 6.3417,
      "mean_token_accuracy": 0.1186050683259964,
      "num_tokens": 2166422.0,
      "step": 1045
    },
    {
      "entropy": 6.549373722076416,
      "epoch": 0.09524673439767779,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999997765602233,
      "loss": 6.4281,
      "mean_token_accuracy": 0.11772651746869087,
      "num_tokens": 2177842.0,
      "step": 1050
    },
    {
      "entropy": 6.614485788345337,
      "epoch": 0.09570029027576198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999997286337502,
      "loss": 6.4994,
      "mean_token_accuracy": 0.12015130147337913,
      "num_tokens": 2188614.0,
      "step": 1055
    },
    {
      "entropy": 6.645506238937378,
      "epoch": 0.09615384615384616,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999996760542254,
      "loss": 6.3042,
      "mean_token_accuracy": 0.12434962838888168,
      "num_tokens": 2197886.0,
      "step": 1060
    },
    {
      "entropy": 6.450875282287598,
      "epoch": 0.09660740203193033,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049999961882165,
      "loss": 6.3515,
      "mean_token_accuracy": 0.12345030084252358,
      "num_tokens": 2209930.0,
      "step": 1065
    },
    {
      "entropy": 6.616359043121338,
      "epoch": 0.09706095791001451,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999995569360251,
      "loss": 6.3704,
      "mean_token_accuracy": 0.11785119250416756,
      "num_tokens": 2219767.0,
      "step": 1070
    },
    {
      "entropy": 6.475496864318847,
      "epoch": 0.0975145137880987,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499999490397352,
      "loss": 6.2989,
      "mean_token_accuracy": 0.1218901552259922,
      "num_tokens": 2230082.0,
      "step": 1075
    },
    {
      "entropy": 6.539343738555909,
      "epoch": 0.09796806966618288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999994192056321,
      "loss": 6.3308,
      "mean_token_accuracy": 0.1269037939608097,
      "num_tokens": 2240066.0,
      "step": 1080
    },
    {
      "entropy": 6.536786890029907,
      "epoch": 0.09842162554426705,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999993433608669,
      "loss": 6.3427,
      "mean_token_accuracy": 0.12274787575006485,
      "num_tokens": 2250732.0,
      "step": 1085
    },
    {
      "entropy": 6.519191551208496,
      "epoch": 0.09887518142235123,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999992628630579,
      "loss": 6.2367,
      "mean_token_accuracy": 0.12317501232028008,
      "num_tokens": 2261452.0,
      "step": 1090
    },
    {
      "entropy": 6.529998350143432,
      "epoch": 0.09932873730043541,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999991777122069,
      "loss": 6.3514,
      "mean_token_accuracy": 0.12049490809440613,
      "num_tokens": 2271929.0,
      "step": 1095
    },
    {
      "entropy": 6.48491473197937,
      "epoch": 0.0997822931785196,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999990879083156,
      "loss": 6.3314,
      "mean_token_accuracy": 0.12258056998252868,
      "num_tokens": 2282987.0,
      "step": 1100
    },
    {
      "entropy": 6.453391408920288,
      "epoch": 0.10023584905660378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999989934513857,
      "loss": 6.3044,
      "mean_token_accuracy": 0.12670305147767066,
      "num_tokens": 2292796.0,
      "step": 1105
    },
    {
      "entropy": 6.591043186187744,
      "epoch": 0.10068940493468795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999988943414193,
      "loss": 6.3516,
      "mean_token_accuracy": 0.12217535674571992,
      "num_tokens": 2303057.0,
      "step": 1110
    },
    {
      "entropy": 6.426148223876953,
      "epoch": 0.10114296081277213,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999987905784184,
      "loss": 6.334,
      "mean_token_accuracy": 0.12347318157553673,
      "num_tokens": 2314168.0,
      "step": 1115
    },
    {
      "entropy": 6.5130480289459225,
      "epoch": 0.10159651669085631,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999986821623853,
      "loss": 6.2937,
      "mean_token_accuracy": 0.12410474196076393,
      "num_tokens": 2325241.0,
      "step": 1120
    },
    {
      "entropy": 6.5921142578125,
      "epoch": 0.1020500725689405,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999985690933219,
      "loss": 6.3255,
      "mean_token_accuracy": 0.12369295954704285,
      "num_tokens": 2334866.0,
      "step": 1125
    },
    {
      "entropy": 6.381993150711059,
      "epoch": 0.10250362844702467,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999984513712311,
      "loss": 6.3279,
      "mean_token_accuracy": 0.12018040791153908,
      "num_tokens": 2345422.0,
      "step": 1130
    },
    {
      "entropy": 6.5917377948760985,
      "epoch": 0.10295718432510885,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999983289961146,
      "loss": 6.3473,
      "mean_token_accuracy": 0.12131484672427177,
      "num_tokens": 2356285.0,
      "step": 1135
    },
    {
      "entropy": 6.436669492721558,
      "epoch": 0.10341074020319303,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999982019679755,
      "loss": 6.2615,
      "mean_token_accuracy": 0.12863264828920365,
      "num_tokens": 2366861.0,
      "step": 1140
    },
    {
      "entropy": 6.494740104675293,
      "epoch": 0.10386429608127722,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999980702868164,
      "loss": 6.2841,
      "mean_token_accuracy": 0.12933464497327804,
      "num_tokens": 2376851.0,
      "step": 1145
    },
    {
      "entropy": 6.544817733764648,
      "epoch": 0.1043178519593614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999979339526396,
      "loss": 6.3588,
      "mean_token_accuracy": 0.13027909770607948,
      "num_tokens": 2386857.0,
      "step": 1150
    },
    {
      "entropy": 6.5093776226043705,
      "epoch": 0.10477140783744557,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999977929654484,
      "loss": 6.2838,
      "mean_token_accuracy": 0.12735775336623192,
      "num_tokens": 2397035.0,
      "step": 1155
    },
    {
      "entropy": 6.470782279968262,
      "epoch": 0.10522496371552975,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999976473252453,
      "loss": 6.3559,
      "mean_token_accuracy": 0.12582893818616867,
      "num_tokens": 2408177.0,
      "step": 1160
    },
    {
      "entropy": 6.572890186309815,
      "epoch": 0.10567851959361393,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999974970320338,
      "loss": 6.3173,
      "mean_token_accuracy": 0.12205304279923439,
      "num_tokens": 2418445.0,
      "step": 1165
    },
    {
      "entropy": 6.472396755218506,
      "epoch": 0.10613207547169812,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999973420858165,
      "loss": 6.3095,
      "mean_token_accuracy": 0.1242362380027771,
      "num_tokens": 2428990.0,
      "step": 1170
    },
    {
      "entropy": 6.47665491104126,
      "epoch": 0.10658563134978229,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999971824865968,
      "loss": 6.3435,
      "mean_token_accuracy": 0.11929563954472541,
      "num_tokens": 2439643.0,
      "step": 1175
    },
    {
      "entropy": 6.585198831558228,
      "epoch": 0.10703918722786647,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999970182343782,
      "loss": 6.3812,
      "mean_token_accuracy": 0.12422708719968796,
      "num_tokens": 2450691.0,
      "step": 1180
    },
    {
      "entropy": 6.567696046829224,
      "epoch": 0.10749274310595065,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999968493291638,
      "loss": 6.3221,
      "mean_token_accuracy": 0.12464427575469017,
      "num_tokens": 2460706.0,
      "step": 1185
    },
    {
      "entropy": 6.306331729888916,
      "epoch": 0.10794629898403484,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999966757709573,
      "loss": 6.2099,
      "mean_token_accuracy": 0.1348014645278454,
      "num_tokens": 2470947.0,
      "step": 1190
    },
    {
      "entropy": 6.511581468582153,
      "epoch": 0.10839985486211902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999964975597622,
      "loss": 6.1938,
      "mean_token_accuracy": 0.1275109112262726,
      "num_tokens": 2481132.0,
      "step": 1195
    },
    {
      "entropy": 6.452719688415527,
      "epoch": 0.10885341074020319,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999963146955821,
      "loss": 6.2243,
      "mean_token_accuracy": 0.12854906544089317,
      "num_tokens": 2491701.0,
      "step": 1200
    },
    {
      "entropy": 6.320997333526611,
      "epoch": 0.10930696661828737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499996127178421,
      "loss": 6.2922,
      "mean_token_accuracy": 0.12324495241045952,
      "num_tokens": 2502360.0,
      "step": 1205
    },
    {
      "entropy": 6.524869346618653,
      "epoch": 0.10976052249637155,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999959350082825,
      "loss": 6.1951,
      "mean_token_accuracy": 0.13170049339532852,
      "num_tokens": 2512140.0,
      "step": 1210
    },
    {
      "entropy": 6.4425348281860355,
      "epoch": 0.11021407837445574,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999957381851709,
      "loss": 6.3032,
      "mean_token_accuracy": 0.12556114420294762,
      "num_tokens": 2522721.0,
      "step": 1215
    },
    {
      "entropy": 6.40459713935852,
      "epoch": 0.1106676342525399,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999955367090901,
      "loss": 6.1739,
      "mean_token_accuracy": 0.12824239879846572,
      "num_tokens": 2531786.0,
      "step": 1220
    },
    {
      "entropy": 6.501996803283691,
      "epoch": 0.11112119013062409,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999953305800441,
      "loss": 6.2788,
      "mean_token_accuracy": 0.12112325578927993,
      "num_tokens": 2542895.0,
      "step": 1225
    },
    {
      "entropy": 6.3701423645019535,
      "epoch": 0.11157474600870827,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999951197980374,
      "loss": 6.2802,
      "mean_token_accuracy": 0.12341128513216973,
      "num_tokens": 2554120.0,
      "step": 1230
    },
    {
      "entropy": 6.539844799041748,
      "epoch": 0.11202830188679246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999949043630744,
      "loss": 6.313,
      "mean_token_accuracy": 0.12485537752509117,
      "num_tokens": 2563935.0,
      "step": 1235
    },
    {
      "entropy": 6.355460834503174,
      "epoch": 0.11248185776487664,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999946842751593,
      "loss": 6.2219,
      "mean_token_accuracy": 0.12797416001558304,
      "num_tokens": 2574465.0,
      "step": 1240
    },
    {
      "entropy": 6.420122289657593,
      "epoch": 0.11293541364296081,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999944595342968,
      "loss": 6.1786,
      "mean_token_accuracy": 0.13132491633296012,
      "num_tokens": 2584694.0,
      "step": 1245
    },
    {
      "entropy": 6.382951498031616,
      "epoch": 0.11338896952104499,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999942301404916,
      "loss": 6.1569,
      "mean_token_accuracy": 0.1340431585907936,
      "num_tokens": 2594540.0,
      "step": 1250
    },
    {
      "entropy": 6.32035346031189,
      "epoch": 0.11384252539912917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999939960937484,
      "loss": 6.0947,
      "mean_token_accuracy": 0.13378870859742165,
      "num_tokens": 2604896.0,
      "step": 1255
    },
    {
      "entropy": 6.4604442596435545,
      "epoch": 0.11429608127721336,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999937573940721,
      "loss": 6.2717,
      "mean_token_accuracy": 0.13821497261524202,
      "num_tokens": 2614143.0,
      "step": 1260
    },
    {
      "entropy": 6.316441440582276,
      "epoch": 0.11474963715529753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999935140414674,
      "loss": 6.1643,
      "mean_token_accuracy": 0.12574966847896576,
      "num_tokens": 2623999.0,
      "step": 1265
    },
    {
      "entropy": 6.501933288574219,
      "epoch": 0.11520319303338171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999932660359395,
      "loss": 6.3783,
      "mean_token_accuracy": 0.12551918104290963,
      "num_tokens": 2636208.0,
      "step": 1270
    },
    {
      "entropy": 6.424142551422119,
      "epoch": 0.11565674891146589,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999930133774936,
      "loss": 6.2436,
      "mean_token_accuracy": 0.13297036662697792,
      "num_tokens": 2645854.0,
      "step": 1275
    },
    {
      "entropy": 6.461622619628907,
      "epoch": 0.11611030478955008,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999927560661348,
      "loss": 6.2337,
      "mean_token_accuracy": 0.13188307881355285,
      "num_tokens": 2655877.0,
      "step": 1280
    },
    {
      "entropy": 6.463164520263672,
      "epoch": 0.11656386066763426,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999924941018685,
      "loss": 6.2551,
      "mean_token_accuracy": 0.13034509047865867,
      "num_tokens": 2666157.0,
      "step": 1285
    },
    {
      "entropy": 6.322438955307007,
      "epoch": 0.11701741654571843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999922274847,
      "loss": 6.1759,
      "mean_token_accuracy": 0.12801435366272926,
      "num_tokens": 2676647.0,
      "step": 1290
    },
    {
      "entropy": 6.35759539604187,
      "epoch": 0.11747097242380261,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499991956214635,
      "loss": 6.2134,
      "mean_token_accuracy": 0.13719891905784606,
      "num_tokens": 2686417.0,
      "step": 1295
    },
    {
      "entropy": 6.4111240863800045,
      "epoch": 0.1179245283018868,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499991680291679,
      "loss": 6.2127,
      "mean_token_accuracy": 0.12441546320915223,
      "num_tokens": 2697540.0,
      "step": 1300
    },
    {
      "entropy": 6.4009637355804445,
      "epoch": 0.11837808417997098,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999913997158378,
      "loss": 6.2549,
      "mean_token_accuracy": 0.11752010956406593,
      "num_tokens": 2709397.0,
      "step": 1305
    },
    {
      "entropy": 6.34979658126831,
      "epoch": 0.11883164005805515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999911144871169,
      "loss": 6.2137,
      "mean_token_accuracy": 0.1235676109790802,
      "num_tokens": 2720430.0,
      "step": 1310
    },
    {
      "entropy": 6.358980894088745,
      "epoch": 0.11928519593613933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999908246055226,
      "loss": 6.1748,
      "mean_token_accuracy": 0.12498210072517395,
      "num_tokens": 2730648.0,
      "step": 1315
    },
    {
      "entropy": 6.358032989501953,
      "epoch": 0.11973875181422351,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999905300710607,
      "loss": 6.1528,
      "mean_token_accuracy": 0.12824709638953208,
      "num_tokens": 2739979.0,
      "step": 1320
    },
    {
      "entropy": 6.410165405273437,
      "epoch": 0.1201923076923077,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999902308837373,
      "loss": 6.2777,
      "mean_token_accuracy": 0.1253784976899624,
      "num_tokens": 2750329.0,
      "step": 1325
    },
    {
      "entropy": 6.337530565261841,
      "epoch": 0.12064586357039188,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999899270435584,
      "loss": 6.1525,
      "mean_token_accuracy": 0.13555875271558762,
      "num_tokens": 2760535.0,
      "step": 1330
    },
    {
      "entropy": 6.363765001296997,
      "epoch": 0.12109941944847605,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999896185505307,
      "loss": 6.1902,
      "mean_token_accuracy": 0.12614675164222716,
      "num_tokens": 2770470.0,
      "step": 1335
    },
    {
      "entropy": 6.35250883102417,
      "epoch": 0.12155297532656023,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999893054046603,
      "loss": 6.2192,
      "mean_token_accuracy": 0.12356962338089943,
      "num_tokens": 2781506.0,
      "step": 1340
    },
    {
      "entropy": 6.300785732269287,
      "epoch": 0.12200653120464441,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999889876059537,
      "loss": 6.0809,
      "mean_token_accuracy": 0.13543135821819305,
      "num_tokens": 2791245.0,
      "step": 1345
    },
    {
      "entropy": 6.336817073822021,
      "epoch": 0.1224600870827286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999886651544176,
      "loss": 6.0946,
      "mean_token_accuracy": 0.13693180531263352,
      "num_tokens": 2801680.0,
      "step": 1350
    },
    {
      "entropy": 6.373439359664917,
      "epoch": 0.12291364296081277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999883380500584,
      "loss": 6.1846,
      "mean_token_accuracy": 0.1356772594153881,
      "num_tokens": 2812805.0,
      "step": 1355
    },
    {
      "entropy": 6.366350793838501,
      "epoch": 0.12336719883889695,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999880062928831,
      "loss": 6.2141,
      "mean_token_accuracy": 0.13203392773866654,
      "num_tokens": 2822834.0,
      "step": 1360
    },
    {
      "entropy": 6.361548852920532,
      "epoch": 0.12382075471698113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999876698828985,
      "loss": 6.1554,
      "mean_token_accuracy": 0.1361112043261528,
      "num_tokens": 2832312.0,
      "step": 1365
    },
    {
      "entropy": 6.335978317260742,
      "epoch": 0.12427431059506532,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999873288201116,
      "loss": 6.1665,
      "mean_token_accuracy": 0.13540629148483277,
      "num_tokens": 2842998.0,
      "step": 1370
    },
    {
      "entropy": 6.394313335418701,
      "epoch": 0.1247278664731495,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999869831045294,
      "loss": 6.1961,
      "mean_token_accuracy": 0.13257036805152894,
      "num_tokens": 2854218.0,
      "step": 1375
    },
    {
      "entropy": 6.33300895690918,
      "epoch": 0.12518142235123367,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999866327361589,
      "loss": 6.0892,
      "mean_token_accuracy": 0.13703012093901634,
      "num_tokens": 2864484.0,
      "step": 1380
    },
    {
      "entropy": 6.370788526535034,
      "epoch": 0.12563497822931785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999862777150077,
      "loss": 6.2028,
      "mean_token_accuracy": 0.13148538768291473,
      "num_tokens": 2873800.0,
      "step": 1385
    },
    {
      "entropy": 6.390060663223267,
      "epoch": 0.12608853410740203,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999859180410829,
      "loss": 6.1965,
      "mean_token_accuracy": 0.13013821244239807,
      "num_tokens": 2884247.0,
      "step": 1390
    },
    {
      "entropy": 6.341064691543579,
      "epoch": 0.12654208998548622,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999855537143919,
      "loss": 6.1096,
      "mean_token_accuracy": 0.13113176226615905,
      "num_tokens": 2894739.0,
      "step": 1395
    },
    {
      "entropy": 6.234110736846924,
      "epoch": 0.1269956458635704,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999851847349425,
      "loss": 6.1616,
      "mean_token_accuracy": 0.13198944479227065,
      "num_tokens": 2904376.0,
      "step": 1400
    },
    {
      "entropy": 6.4315849304199215,
      "epoch": 0.12744920174165458,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999848111027419,
      "loss": 6.2265,
      "mean_token_accuracy": 0.12843010425567628,
      "num_tokens": 2914304.0,
      "step": 1405
    },
    {
      "entropy": 6.319644594192505,
      "epoch": 0.12790275761973874,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999844328177983,
      "loss": 6.0959,
      "mean_token_accuracy": 0.13285723477602004,
      "num_tokens": 2924400.0,
      "step": 1410
    },
    {
      "entropy": 6.456520843505859,
      "epoch": 0.12835631349782292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999840498801191,
      "loss": 6.3401,
      "mean_token_accuracy": 0.1338380053639412,
      "num_tokens": 2935040.0,
      "step": 1415
    },
    {
      "entropy": 6.279810905456543,
      "epoch": 0.1288098693759071,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999836622897126,
      "loss": 6.1154,
      "mean_token_accuracy": 0.1344787932932377,
      "num_tokens": 2945536.0,
      "step": 1420
    },
    {
      "entropy": 6.234977674484253,
      "epoch": 0.1292634252539913,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999832700465865,
      "loss": 6.0487,
      "mean_token_accuracy": 0.13450392782688142,
      "num_tokens": 2954969.0,
      "step": 1425
    },
    {
      "entropy": 6.315559005737304,
      "epoch": 0.12971698113207547,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999828731507491,
      "loss": 6.1625,
      "mean_token_accuracy": 0.13059850856661798,
      "num_tokens": 2966218.0,
      "step": 1430
    },
    {
      "entropy": 6.379030179977417,
      "epoch": 0.13017053701015965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999824716022085,
      "loss": 6.1894,
      "mean_token_accuracy": 0.127619219571352,
      "num_tokens": 2976640.0,
      "step": 1435
    },
    {
      "entropy": 6.313711643218994,
      "epoch": 0.13062409288824384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999820654009731,
      "loss": 6.1533,
      "mean_token_accuracy": 0.1332421600818634,
      "num_tokens": 2985883.0,
      "step": 1440
    },
    {
      "entropy": 6.321325731277466,
      "epoch": 0.13107764876632802,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999816545470513,
      "loss": 6.1541,
      "mean_token_accuracy": 0.12692199870944024,
      "num_tokens": 2996695.0,
      "step": 1445
    },
    {
      "entropy": 6.334502124786377,
      "epoch": 0.1315312046444122,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999812390404514,
      "loss": 6.16,
      "mean_token_accuracy": 0.1343579739332199,
      "num_tokens": 3007500.0,
      "step": 1450
    },
    {
      "entropy": 6.184367370605469,
      "epoch": 0.13198476052249636,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999808188811823,
      "loss": 6.0132,
      "mean_token_accuracy": 0.13415692150592803,
      "num_tokens": 3017299.0,
      "step": 1455
    },
    {
      "entropy": 6.403480911254883,
      "epoch": 0.13243831640058054,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999803940692524,
      "loss": 6.3011,
      "mean_token_accuracy": 0.12249834388494492,
      "num_tokens": 3028389.0,
      "step": 1460
    },
    {
      "entropy": 6.317180013656616,
      "epoch": 0.13289187227866472,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999799646046707,
      "loss": 6.1135,
      "mean_token_accuracy": 0.1368947669863701,
      "num_tokens": 3038883.0,
      "step": 1465
    },
    {
      "entropy": 6.403723573684692,
      "epoch": 0.1333454281567489,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999795304874461,
      "loss": 6.2419,
      "mean_token_accuracy": 0.13426837772130967,
      "num_tokens": 3049904.0,
      "step": 1470
    },
    {
      "entropy": 6.2098509788513185,
      "epoch": 0.1337989840348331,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999790917175873,
      "loss": 6.0755,
      "mean_token_accuracy": 0.13349713161587715,
      "num_tokens": 3059788.0,
      "step": 1475
    },
    {
      "entropy": 6.351387977600098,
      "epoch": 0.13425253991291727,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999786482951036,
      "loss": 6.0716,
      "mean_token_accuracy": 0.1320554032921791,
      "num_tokens": 3070413.0,
      "step": 1480
    },
    {
      "entropy": 6.270417833328247,
      "epoch": 0.13470609579100146,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999782002200041,
      "loss": 6.1959,
      "mean_token_accuracy": 0.12884705662727355,
      "num_tokens": 3081093.0,
      "step": 1485
    },
    {
      "entropy": 6.342434740066528,
      "epoch": 0.13515965166908564,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999777474922982,
      "loss": 6.129,
      "mean_token_accuracy": 0.13011036217212676,
      "num_tokens": 3090594.0,
      "step": 1490
    },
    {
      "entropy": 6.322659873962403,
      "epoch": 0.13561320754716982,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999772901119951,
      "loss": 6.2158,
      "mean_token_accuracy": 0.13068215921521187,
      "num_tokens": 3101061.0,
      "step": 1495
    },
    {
      "entropy": 6.294349813461304,
      "epoch": 0.13606676342525398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999768280791043,
      "loss": 6.1383,
      "mean_token_accuracy": 0.13425096422433852,
      "num_tokens": 3111715.0,
      "step": 1500
    },
    {
      "entropy": 6.270824289321899,
      "epoch": 0.13652031930333816,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999763613936352,
      "loss": 6.0633,
      "mean_token_accuracy": 0.13947287648916246,
      "num_tokens": 3121378.0,
      "step": 1505
    },
    {
      "entropy": 6.246769809722901,
      "epoch": 0.13697387518142234,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999758900555979,
      "loss": 5.9928,
      "mean_token_accuracy": 0.1401647448539734,
      "num_tokens": 3131229.0,
      "step": 1510
    },
    {
      "entropy": 6.208933019638062,
      "epoch": 0.13742743105950653,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999754140650017,
      "loss": 6.0878,
      "mean_token_accuracy": 0.13210750818252565,
      "num_tokens": 3140859.0,
      "step": 1515
    },
    {
      "entropy": 6.351792240142823,
      "epoch": 0.1378809869375907,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999749334218567,
      "loss": 6.0861,
      "mean_token_accuracy": 0.13336024880409242,
      "num_tokens": 3151692.0,
      "step": 1520
    },
    {
      "entropy": 6.221499061584472,
      "epoch": 0.1383345428156749,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999744481261726,
      "loss": 6.1524,
      "mean_token_accuracy": 0.13277293592691422,
      "num_tokens": 3161879.0,
      "step": 1525
    },
    {
      "entropy": 6.278037405014038,
      "epoch": 0.13878809869375908,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999739581779597,
      "loss": 6.0267,
      "mean_token_accuracy": 0.13772518262267114,
      "num_tokens": 3171282.0,
      "step": 1530
    },
    {
      "entropy": 6.244936561584472,
      "epoch": 0.13924165457184326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499973463577228,
      "loss": 6.0779,
      "mean_token_accuracy": 0.13918796479701995,
      "num_tokens": 3181398.0,
      "step": 1535
    },
    {
      "entropy": 6.294873476028442,
      "epoch": 0.13969521044992744,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999729643239877,
      "loss": 6.1991,
      "mean_token_accuracy": 0.13689608871936798,
      "num_tokens": 3191431.0,
      "step": 1540
    },
    {
      "entropy": 6.297893095016479,
      "epoch": 0.1401487663280116,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999724604182492,
      "loss": 6.1937,
      "mean_token_accuracy": 0.1275126539170742,
      "num_tokens": 3202147.0,
      "step": 1545
    },
    {
      "entropy": 6.362968301773071,
      "epoch": 0.14060232220609578,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999719518600229,
      "loss": 6.073,
      "mean_token_accuracy": 0.13779982998967172,
      "num_tokens": 3212428.0,
      "step": 1550
    },
    {
      "entropy": 6.222197484970093,
      "epoch": 0.14105587808417996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999714386493192,
      "loss": 6.1203,
      "mean_token_accuracy": 0.13455342650413513,
      "num_tokens": 3222099.0,
      "step": 1555
    },
    {
      "entropy": 6.321711540222168,
      "epoch": 0.14150943396226415,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999709207861489,
      "loss": 6.0726,
      "mean_token_accuracy": 0.13806343004107474,
      "num_tokens": 3232079.0,
      "step": 1560
    },
    {
      "entropy": 6.153911828994751,
      "epoch": 0.14196298984034833,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999703982705227,
      "loss": 5.9721,
      "mean_token_accuracy": 0.14627765491604805,
      "num_tokens": 3241762.0,
      "step": 1565
    },
    {
      "entropy": 6.244314336776734,
      "epoch": 0.1424165457184325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999698711024513,
      "loss": 6.0669,
      "mean_token_accuracy": 0.1394437588751316,
      "num_tokens": 3252571.0,
      "step": 1570
    },
    {
      "entropy": 6.215821599960327,
      "epoch": 0.1428701015965167,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999693392819456,
      "loss": 6.1098,
      "mean_token_accuracy": 0.13461477681994438,
      "num_tokens": 3263542.0,
      "step": 1575
    },
    {
      "entropy": 6.340892887115478,
      "epoch": 0.14332365747460088,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999688028090166,
      "loss": 6.1811,
      "mean_token_accuracy": 0.13378583863377572,
      "num_tokens": 3273295.0,
      "step": 1580
    },
    {
      "entropy": 6.3619109153747555,
      "epoch": 0.14377721335268506,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999682616836755,
      "loss": 6.1952,
      "mean_token_accuracy": 0.136061330139637,
      "num_tokens": 3283605.0,
      "step": 1585
    },
    {
      "entropy": 6.243741178512574,
      "epoch": 0.14423076923076922,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999677159059334,
      "loss": 6.0415,
      "mean_token_accuracy": 0.14292106181383132,
      "num_tokens": 3293837.0,
      "step": 1590
    },
    {
      "entropy": 6.130685424804687,
      "epoch": 0.1446843251088534,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999671654758016,
      "loss": 6.0417,
      "mean_token_accuracy": 0.14042141437530517,
      "num_tokens": 3304131.0,
      "step": 1595
    },
    {
      "entropy": 6.302355670928955,
      "epoch": 0.14513788098693758,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999666103932915,
      "loss": 6.0746,
      "mean_token_accuracy": 0.13077147156000138,
      "num_tokens": 3313543.0,
      "step": 1600
    },
    {
      "entropy": 6.222482824325562,
      "epoch": 0.14559143686502177,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999660506584145,
      "loss": 6.0608,
      "mean_token_accuracy": 0.1355305939912796,
      "num_tokens": 3324487.0,
      "step": 1605
    },
    {
      "entropy": 6.281767272949219,
      "epoch": 0.14604499274310595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999654862711823,
      "loss": 6.0873,
      "mean_token_accuracy": 0.1388099044561386,
      "num_tokens": 3335094.0,
      "step": 1610
    },
    {
      "entropy": 6.182447385787964,
      "epoch": 0.14649854862119013,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999649172316065,
      "loss": 6.0205,
      "mean_token_accuracy": 0.13585399314761162,
      "num_tokens": 3344252.0,
      "step": 1615
    },
    {
      "entropy": 6.219208812713623,
      "epoch": 0.14695210449927432,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499964343539699,
      "loss": 6.0905,
      "mean_token_accuracy": 0.13364078626036643,
      "num_tokens": 3354265.0,
      "step": 1620
    },
    {
      "entropy": 6.3048234462738035,
      "epoch": 0.1474056603773585,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999637651954714,
      "loss": 6.182,
      "mean_token_accuracy": 0.13599163964390754,
      "num_tokens": 3364616.0,
      "step": 1625
    },
    {
      "entropy": 6.33318829536438,
      "epoch": 0.14785921625544268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999631821989358,
      "loss": 6.1253,
      "mean_token_accuracy": 0.14277874156832696,
      "num_tokens": 3373650.0,
      "step": 1630
    },
    {
      "entropy": 6.114496469497681,
      "epoch": 0.14831277213352684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999625945501043,
      "loss": 6.0331,
      "mean_token_accuracy": 0.13961085602641105,
      "num_tokens": 3384590.0,
      "step": 1635
    },
    {
      "entropy": 6.34743185043335,
      "epoch": 0.14876632801161102,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499962002248989,
      "loss": 6.0701,
      "mean_token_accuracy": 0.13647577166557312,
      "num_tokens": 3395616.0,
      "step": 1640
    },
    {
      "entropy": 6.165446662902832,
      "epoch": 0.1492198838896952,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999614052956024,
      "loss": 6.0688,
      "mean_token_accuracy": 0.1349581092596054,
      "num_tokens": 3405810.0,
      "step": 1645
    },
    {
      "entropy": 6.321514320373535,
      "epoch": 0.1496734397677794,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999608036899563,
      "loss": 6.1055,
      "mean_token_accuracy": 0.1329289957880974,
      "num_tokens": 3415932.0,
      "step": 1650
    },
    {
      "entropy": 6.197605752944947,
      "epoch": 0.15012699564586357,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999601974320636,
      "loss": 6.0631,
      "mean_token_accuracy": 0.13647494986653327,
      "num_tokens": 3425626.0,
      "step": 1655
    },
    {
      "entropy": 6.155983877182007,
      "epoch": 0.15058055152394775,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999595865219367,
      "loss": 6.0609,
      "mean_token_accuracy": 0.1348869502544403,
      "num_tokens": 3434837.0,
      "step": 1660
    },
    {
      "entropy": 6.258736085891724,
      "epoch": 0.15103410740203194,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999589709595881,
      "loss": 6.0701,
      "mean_token_accuracy": 0.14022651612758635,
      "num_tokens": 3445238.0,
      "step": 1665
    },
    {
      "entropy": 6.191891717910766,
      "epoch": 0.15148766328011612,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999583507450308,
      "loss": 6.0521,
      "mean_token_accuracy": 0.13351622000336646,
      "num_tokens": 3456738.0,
      "step": 1670
    },
    {
      "entropy": 6.110352945327759,
      "epoch": 0.1519412191582003,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999577258782774,
      "loss": 5.9378,
      "mean_token_accuracy": 0.14009969010949136,
      "num_tokens": 3466296.0,
      "step": 1675
    },
    {
      "entropy": 6.319712066650391,
      "epoch": 0.15239477503628446,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499957096359341,
      "loss": 6.1345,
      "mean_token_accuracy": 0.13245827108621597,
      "num_tokens": 3476989.0,
      "step": 1680
    },
    {
      "entropy": 6.156819915771484,
      "epoch": 0.15284833091436864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999564621882343,
      "loss": 5.97,
      "mean_token_accuracy": 0.1411023445427418,
      "num_tokens": 3487050.0,
      "step": 1685
    },
    {
      "entropy": 6.2730926990509035,
      "epoch": 0.15330188679245282,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999558233649708,
      "loss": 6.188,
      "mean_token_accuracy": 0.12993267178535461,
      "num_tokens": 3497777.0,
      "step": 1690
    },
    {
      "entropy": 6.238672304153442,
      "epoch": 0.153755442670537,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999551798895636,
      "loss": 6.0905,
      "mean_token_accuracy": 0.13315979093313218,
      "num_tokens": 3509066.0,
      "step": 1695
    },
    {
      "entropy": 6.142438983917236,
      "epoch": 0.1542089985486212,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999545317620258,
      "loss": 6.026,
      "mean_token_accuracy": 0.13728305473923683,
      "num_tokens": 3519567.0,
      "step": 1700
    },
    {
      "entropy": 6.209616088867188,
      "epoch": 0.15466255442670537,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999538789823711,
      "loss": 6.0481,
      "mean_token_accuracy": 0.13899025991559028,
      "num_tokens": 3529402.0,
      "step": 1705
    },
    {
      "entropy": 6.3146570205688475,
      "epoch": 0.15511611030478956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999532215506127,
      "loss": 6.0703,
      "mean_token_accuracy": 0.1440271407365799,
      "num_tokens": 3539805.0,
      "step": 1710
    },
    {
      "entropy": 6.143733882904053,
      "epoch": 0.15556966618287374,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999525594667645,
      "loss": 5.9303,
      "mean_token_accuracy": 0.14397806972265242,
      "num_tokens": 3550477.0,
      "step": 1715
    },
    {
      "entropy": 6.110523653030396,
      "epoch": 0.15602322206095792,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049995189273084,
      "loss": 6.0405,
      "mean_token_accuracy": 0.139139074832201,
      "num_tokens": 3560729.0,
      "step": 1720
    },
    {
      "entropy": 6.220071601867676,
      "epoch": 0.15647677793904208,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499951221342853,
      "loss": 6.0973,
      "mean_token_accuracy": 0.13195881322026254,
      "num_tokens": 3571538.0,
      "step": 1725
    },
    {
      "entropy": 6.214411115646362,
      "epoch": 0.15693033381712626,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999505453028174,
      "loss": 6.1078,
      "mean_token_accuracy": 0.13914671614766122,
      "num_tokens": 3583128.0,
      "step": 1730
    },
    {
      "entropy": 6.246854066848755,
      "epoch": 0.15738388969521044,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999498646107472,
      "loss": 6.0354,
      "mean_token_accuracy": 0.14036523252725602,
      "num_tokens": 3594358.0,
      "step": 1735
    },
    {
      "entropy": 6.198376893997192,
      "epoch": 0.15783744557329463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999491792666566,
      "loss": 5.959,
      "mean_token_accuracy": 0.14713839516043664,
      "num_tokens": 3604345.0,
      "step": 1740
    },
    {
      "entropy": 6.148045206069947,
      "epoch": 0.1582910014513788,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999484892705597,
      "loss": 6.0217,
      "mean_token_accuracy": 0.13767840191721917,
      "num_tokens": 3616129.0,
      "step": 1745
    },
    {
      "entropy": 6.18104510307312,
      "epoch": 0.158744557329463,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999477946224705,
      "loss": 6.0602,
      "mean_token_accuracy": 0.13979990035295486,
      "num_tokens": 3627458.0,
      "step": 1750
    },
    {
      "entropy": 6.226989650726319,
      "epoch": 0.15919811320754718,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999470953224037,
      "loss": 5.9442,
      "mean_token_accuracy": 0.14392665922641754,
      "num_tokens": 3636047.0,
      "step": 1755
    },
    {
      "entropy": 6.051787710189819,
      "epoch": 0.15965166908563136,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999463913703735,
      "loss": 5.9397,
      "mean_token_accuracy": 0.1413258582353592,
      "num_tokens": 3646930.0,
      "step": 1760
    },
    {
      "entropy": 6.218035459518433,
      "epoch": 0.16010522496371554,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999456827663948,
      "loss": 6.0338,
      "mean_token_accuracy": 0.14086730852723123,
      "num_tokens": 3657279.0,
      "step": 1765
    },
    {
      "entropy": 6.123547697067261,
      "epoch": 0.1605587808417997,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499944969510482,
      "loss": 5.9464,
      "mean_token_accuracy": 0.1423609510064125,
      "num_tokens": 3667070.0,
      "step": 1770
    },
    {
      "entropy": 6.166107845306397,
      "epoch": 0.16101233671988388,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999442516026498,
      "loss": 5.9613,
      "mean_token_accuracy": 0.1381991073489189,
      "num_tokens": 3677560.0,
      "step": 1775
    },
    {
      "entropy": 6.090453720092773,
      "epoch": 0.16146589259796806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999435290429133,
      "loss": 5.9268,
      "mean_token_accuracy": 0.1474891871213913,
      "num_tokens": 3686701.0,
      "step": 1780
    },
    {
      "entropy": 6.152267122268677,
      "epoch": 0.16191944847605225,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999428018312872,
      "loss": 6.0228,
      "mean_token_accuracy": 0.14551690965890884,
      "num_tokens": 3696338.0,
      "step": 1785
    },
    {
      "entropy": 6.150699281692505,
      "epoch": 0.16237300435413643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999420699677867,
      "loss": 5.9993,
      "mean_token_accuracy": 0.14381766989827155,
      "num_tokens": 3705752.0,
      "step": 1790
    },
    {
      "entropy": 6.229558277130127,
      "epoch": 0.1628265602322206,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999413334524269,
      "loss": 6.084,
      "mean_token_accuracy": 0.13718133196234703,
      "num_tokens": 3716225.0,
      "step": 1795
    },
    {
      "entropy": 6.1720422267913815,
      "epoch": 0.1632801161103048,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999405922852229,
      "loss": 5.9581,
      "mean_token_accuracy": 0.14193314164876938,
      "num_tokens": 3726593.0,
      "step": 1800
    },
    {
      "entropy": 6.101333570480347,
      "epoch": 0.16373367198838898,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999398464661901,
      "loss": 6.0091,
      "mean_token_accuracy": 0.13523678332567216,
      "num_tokens": 3737563.0,
      "step": 1805
    },
    {
      "entropy": 6.307489013671875,
      "epoch": 0.16418722786647316,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999390959953441,
      "loss": 6.0708,
      "mean_token_accuracy": 0.1370485432446003,
      "num_tokens": 3747752.0,
      "step": 1810
    },
    {
      "entropy": 6.057960081100464,
      "epoch": 0.16464078374455732,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999383408727002,
      "loss": 6.0019,
      "mean_token_accuracy": 0.139044688642025,
      "num_tokens": 3758795.0,
      "step": 1815
    },
    {
      "entropy": 6.2216832637786865,
      "epoch": 0.1650943396226415,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999375810982741,
      "loss": 5.9714,
      "mean_token_accuracy": 0.14233563393354415,
      "num_tokens": 3768690.0,
      "step": 1820
    },
    {
      "entropy": 6.090987873077393,
      "epoch": 0.16554789550072568,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999368166720815,
      "loss": 6.0494,
      "mean_token_accuracy": 0.13934776037931443,
      "num_tokens": 3779018.0,
      "step": 1825
    },
    {
      "entropy": 6.219239854812622,
      "epoch": 0.16600145137880987,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999360475941382,
      "loss": 5.9469,
      "mean_token_accuracy": 0.14526350796222687,
      "num_tokens": 3788985.0,
      "step": 1830
    },
    {
      "entropy": 6.173433303833008,
      "epoch": 0.16645500725689405,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049993527386446,
      "loss": 6.0821,
      "mean_token_accuracy": 0.13157572448253632,
      "num_tokens": 3800787.0,
      "step": 1835
    },
    {
      "entropy": 6.16374773979187,
      "epoch": 0.16690856313497823,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999344954830633,
      "loss": 5.9647,
      "mean_token_accuracy": 0.13687085136771202,
      "num_tokens": 3811381.0,
      "step": 1840
    },
    {
      "entropy": 6.157071971893311,
      "epoch": 0.16736211901306242,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999337124499638,
      "loss": 6.0685,
      "mean_token_accuracy": 0.14529063701629638,
      "num_tokens": 3822055.0,
      "step": 1845
    },
    {
      "entropy": 6.190087890625,
      "epoch": 0.1678156748911466,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999329247651777,
      "loss": 6.0321,
      "mean_token_accuracy": 0.1372029162943363,
      "num_tokens": 3831900.0,
      "step": 1850
    },
    {
      "entropy": 6.127886390686035,
      "epoch": 0.16826923076923078,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999321324287213,
      "loss": 5.9863,
      "mean_token_accuracy": 0.14112438708543779,
      "num_tokens": 3842702.0,
      "step": 1855
    },
    {
      "entropy": 6.237508106231689,
      "epoch": 0.16872278664731494,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999313354406114,
      "loss": 5.9599,
      "mean_token_accuracy": 0.14055541008710862,
      "num_tokens": 3852690.0,
      "step": 1860
    },
    {
      "entropy": 6.039614915847778,
      "epoch": 0.16917634252539912,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499930533800864,
      "loss": 5.9466,
      "mean_token_accuracy": 0.1505817547440529,
      "num_tokens": 3863040.0,
      "step": 1865
    },
    {
      "entropy": 6.180584526062011,
      "epoch": 0.1696298984034833,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999297275094958,
      "loss": 6.0401,
      "mean_token_accuracy": 0.13932662308216096,
      "num_tokens": 3873613.0,
      "step": 1870
    },
    {
      "entropy": 6.042843770980835,
      "epoch": 0.1700834542815675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999289165665235,
      "loss": 5.8485,
      "mean_token_accuracy": 0.1480574920773506,
      "num_tokens": 3883936.0,
      "step": 1875
    },
    {
      "entropy": 6.060999536514283,
      "epoch": 0.17053701015965167,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999281009719639,
      "loss": 5.8154,
      "mean_token_accuracy": 0.14890880212187768,
      "num_tokens": 3893788.0,
      "step": 1880
    },
    {
      "entropy": 6.056200647354126,
      "epoch": 0.17099056603773585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999272807258339,
      "loss": 5.8885,
      "mean_token_accuracy": 0.14825719892978667,
      "num_tokens": 3903434.0,
      "step": 1885
    },
    {
      "entropy": 6.1081467628479,
      "epoch": 0.17144412191582004,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999264558281504,
      "loss": 5.895,
      "mean_token_accuracy": 0.1487080954015255,
      "num_tokens": 3913476.0,
      "step": 1890
    },
    {
      "entropy": 5.991112613677979,
      "epoch": 0.17189767779390422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999256262789305,
      "loss": 5.8483,
      "mean_token_accuracy": 0.1485322192311287,
      "num_tokens": 3923301.0,
      "step": 1895
    },
    {
      "entropy": 6.099672555923462,
      "epoch": 0.1723512336719884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999247920781913,
      "loss": 6.0042,
      "mean_token_accuracy": 0.13689797297120093,
      "num_tokens": 3933335.0,
      "step": 1900
    },
    {
      "entropy": 6.197177267074585,
      "epoch": 0.17280478955007256,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999239532259502,
      "loss": 6.0293,
      "mean_token_accuracy": 0.13568304777145385,
      "num_tokens": 3944171.0,
      "step": 1905
    },
    {
      "entropy": 6.177489757537842,
      "epoch": 0.17325834542815674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999231097222244,
      "loss": 6.0829,
      "mean_token_accuracy": 0.14412558674812317,
      "num_tokens": 3954160.0,
      "step": 1910
    },
    {
      "entropy": 6.171565389633178,
      "epoch": 0.17371190130624092,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999222615670312,
      "loss": 5.8844,
      "mean_token_accuracy": 0.14805472120642663,
      "num_tokens": 3964110.0,
      "step": 1915
    },
    {
      "entropy": 6.134349346160889,
      "epoch": 0.1741654571843251,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999214087603885,
      "loss": 5.9675,
      "mean_token_accuracy": 0.1452759712934494,
      "num_tokens": 3974293.0,
      "step": 1920
    },
    {
      "entropy": 6.078616380691528,
      "epoch": 0.1746190130624093,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999205513023135,
      "loss": 5.9293,
      "mean_token_accuracy": 0.15142350345849992,
      "num_tokens": 3984346.0,
      "step": 1925
    },
    {
      "entropy": 6.053181409835815,
      "epoch": 0.17507256894049347,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999196891928245,
      "loss": 5.866,
      "mean_token_accuracy": 0.15365748405456542,
      "num_tokens": 3992897.0,
      "step": 1930
    },
    {
      "entropy": 6.05509238243103,
      "epoch": 0.17552612481857766,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999188224319388,
      "loss": 5.897,
      "mean_token_accuracy": 0.15057138204574586,
      "num_tokens": 4002575.0,
      "step": 1935
    },
    {
      "entropy": 6.070607757568359,
      "epoch": 0.17597968069666184,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999179510196746,
      "loss": 5.9925,
      "mean_token_accuracy": 0.14099667817354203,
      "num_tokens": 4013752.0,
      "step": 1940
    },
    {
      "entropy": 6.107510566711426,
      "epoch": 0.17643323657474602,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999170749560498,
      "loss": 5.8739,
      "mean_token_accuracy": 0.14898938685655594,
      "num_tokens": 4023718.0,
      "step": 1945
    },
    {
      "entropy": 6.064271402359009,
      "epoch": 0.17688679245283018,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999161942410827,
      "loss": 5.9743,
      "mean_token_accuracy": 0.1369916170835495,
      "num_tokens": 4034210.0,
      "step": 1950
    },
    {
      "entropy": 6.2005760192871096,
      "epoch": 0.17734034833091436,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999153088747913,
      "loss": 5.9622,
      "mean_token_accuracy": 0.1424011543393135,
      "num_tokens": 4044868.0,
      "step": 1955
    },
    {
      "entropy": 5.97287278175354,
      "epoch": 0.17779390420899854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499914418857194,
      "loss": 5.9603,
      "mean_token_accuracy": 0.14433022439479828,
      "num_tokens": 4055049.0,
      "step": 1960
    },
    {
      "entropy": 6.208456897735596,
      "epoch": 0.17824746008708273,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999135241883092,
      "loss": 6.0087,
      "mean_token_accuracy": 0.14180623665452002,
      "num_tokens": 4065889.0,
      "step": 1965
    },
    {
      "entropy": 6.090062475204467,
      "epoch": 0.1787010159651669,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999126248681555,
      "loss": 6.0282,
      "mean_token_accuracy": 0.1386457771062851,
      "num_tokens": 4077288.0,
      "step": 1970
    },
    {
      "entropy": 6.084208917617798,
      "epoch": 0.1791545718432511,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999117208967513,
      "loss": 5.9802,
      "mean_token_accuracy": 0.1426486611366272,
      "num_tokens": 4087402.0,
      "step": 1975
    },
    {
      "entropy": 6.152999305725098,
      "epoch": 0.17960812772133528,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999108122741154,
      "loss": 6.027,
      "mean_token_accuracy": 0.13648915737867356,
      "num_tokens": 4098150.0,
      "step": 1980
    },
    {
      "entropy": 6.015424156188965,
      "epoch": 0.18006168359941946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999098990002666,
      "loss": 5.8866,
      "mean_token_accuracy": 0.1459066838026047,
      "num_tokens": 4108134.0,
      "step": 1985
    },
    {
      "entropy": 6.190234994888305,
      "epoch": 0.18051523947750364,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999089810752237,
      "loss": 6.0277,
      "mean_token_accuracy": 0.13577395901083947,
      "num_tokens": 4118251.0,
      "step": 1990
    },
    {
      "entropy": 6.0451069355010985,
      "epoch": 0.1809687953555878,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499908058499006,
      "loss": 5.8609,
      "mean_token_accuracy": 0.1469724267721176,
      "num_tokens": 4129228.0,
      "step": 1995
    },
    {
      "entropy": 6.0451616764068605,
      "epoch": 0.18142235123367198,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999071312716321,
      "loss": 5.9319,
      "mean_token_accuracy": 0.15002663657069207,
      "num_tokens": 4139391.0,
      "step": 2000
    },
    {
      "entropy": 6.0800574779510494,
      "epoch": 0.18187590711175616,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999061993931215,
      "loss": 5.8526,
      "mean_token_accuracy": 0.1524651437997818,
      "num_tokens": 4148891.0,
      "step": 2005
    },
    {
      "entropy": 6.2835156440734865,
      "epoch": 0.18232946298984035,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999052628634934,
      "loss": 6.2064,
      "mean_token_accuracy": 0.138189584761858,
      "num_tokens": 4159551.0,
      "step": 2010
    },
    {
      "entropy": 6.138812065124512,
      "epoch": 0.18278301886792453,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999043216827671,
      "loss": 6.0055,
      "mean_token_accuracy": 0.14381420388817787,
      "num_tokens": 4170055.0,
      "step": 2015
    },
    {
      "entropy": 6.033186531066894,
      "epoch": 0.1832365747460087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999033758509621,
      "loss": 5.8591,
      "mean_token_accuracy": 0.14449694454669954,
      "num_tokens": 4180767.0,
      "step": 2020
    },
    {
      "entropy": 6.07762713432312,
      "epoch": 0.1836901306240929,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499902425368098,
      "loss": 5.9878,
      "mean_token_accuracy": 0.14416737630963325,
      "num_tokens": 4190654.0,
      "step": 2025
    },
    {
      "entropy": 6.103671360015869,
      "epoch": 0.18414368650217708,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999014702341944,
      "loss": 5.9002,
      "mean_token_accuracy": 0.1454018011689186,
      "num_tokens": 4200753.0,
      "step": 2030
    },
    {
      "entropy": 6.093059635162353,
      "epoch": 0.18459724238026126,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999005104492712,
      "loss": 5.9798,
      "mean_token_accuracy": 0.14127395525574685,
      "num_tokens": 4210994.0,
      "step": 2035
    },
    {
      "entropy": 6.107717561721802,
      "epoch": 0.18505079825834542,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998995460133481,
      "loss": 5.9256,
      "mean_token_accuracy": 0.1447222650051117,
      "num_tokens": 4220883.0,
      "step": 2040
    },
    {
      "entropy": 6.123315095901489,
      "epoch": 0.1855043541364296,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998985769264451,
      "loss": 5.9518,
      "mean_token_accuracy": 0.13728086575865744,
      "num_tokens": 4231269.0,
      "step": 2045
    },
    {
      "entropy": 6.056053113937378,
      "epoch": 0.18595791001451378,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998976031885821,
      "loss": 5.9715,
      "mean_token_accuracy": 0.14049455970525743,
      "num_tokens": 4241559.0,
      "step": 2050
    },
    {
      "entropy": 6.045488786697388,
      "epoch": 0.18641146589259797,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998966247997794,
      "loss": 5.8466,
      "mean_token_accuracy": 0.14596481025218963,
      "num_tokens": 4250870.0,
      "step": 2055
    },
    {
      "entropy": 6.027891731262207,
      "epoch": 0.18686502177068215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998956417600572,
      "loss": 5.8125,
      "mean_token_accuracy": 0.1532222166657448,
      "num_tokens": 4260848.0,
      "step": 2060
    },
    {
      "entropy": 6.0480344772338865,
      "epoch": 0.18731857764876633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998946540694358,
      "loss": 5.9379,
      "mean_token_accuracy": 0.14905711114406586,
      "num_tokens": 4270973.0,
      "step": 2065
    },
    {
      "entropy": 6.159943771362305,
      "epoch": 0.18777213352685052,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998936617279357,
      "loss": 6.0104,
      "mean_token_accuracy": 0.14108132272958757,
      "num_tokens": 4281211.0,
      "step": 2070
    },
    {
      "entropy": 6.0061546802520756,
      "epoch": 0.1882256894049347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998926647355773,
      "loss": 5.8728,
      "mean_token_accuracy": 0.14745167195796965,
      "num_tokens": 4291425.0,
      "step": 2075
    },
    {
      "entropy": 6.123989725112915,
      "epoch": 0.18867924528301888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998916630923813,
      "loss": 5.9712,
      "mean_token_accuracy": 0.1410026654601097,
      "num_tokens": 4301079.0,
      "step": 2080
    },
    {
      "entropy": 6.104276657104492,
      "epoch": 0.18913280116110304,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998906567983685,
      "loss": 5.9629,
      "mean_token_accuracy": 0.14320143908262253,
      "num_tokens": 4312374.0,
      "step": 2085
    },
    {
      "entropy": 6.043295049667359,
      "epoch": 0.18958635703918722,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998896458535595,
      "loss": 5.8714,
      "mean_token_accuracy": 0.15539209097623824,
      "num_tokens": 4322810.0,
      "step": 2090
    },
    {
      "entropy": 6.002667427062988,
      "epoch": 0.1900399129172714,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998886302579752,
      "loss": 5.8651,
      "mean_token_accuracy": 0.14834146052598954,
      "num_tokens": 4332082.0,
      "step": 2095
    },
    {
      "entropy": 6.0972264289855955,
      "epoch": 0.19049346879535559,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499887610011637,
      "loss": 5.9636,
      "mean_token_accuracy": 0.14356174021959306,
      "num_tokens": 4342632.0,
      "step": 2100
    },
    {
      "entropy": 6.016445589065552,
      "epoch": 0.19094702467343977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998865851145656,
      "loss": 5.8886,
      "mean_token_accuracy": 0.1432491272687912,
      "num_tokens": 4353544.0,
      "step": 2105
    },
    {
      "entropy": 6.031062507629395,
      "epoch": 0.19140058055152395,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998855555667822,
      "loss": 6.02,
      "mean_token_accuracy": 0.14226981848478318,
      "num_tokens": 4363636.0,
      "step": 2110
    },
    {
      "entropy": 6.133910322189331,
      "epoch": 0.19185413642960814,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998845213683084,
      "loss": 5.8763,
      "mean_token_accuracy": 0.1426288031041622,
      "num_tokens": 4373672.0,
      "step": 2115
    },
    {
      "entropy": 5.938132858276367,
      "epoch": 0.19230769230769232,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998834825191652,
      "loss": 5.8381,
      "mean_token_accuracy": 0.14870256185531616,
      "num_tokens": 4383199.0,
      "step": 2120
    },
    {
      "entropy": 6.071506214141846,
      "epoch": 0.1927612481857765,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998824390193743,
      "loss": 5.9109,
      "mean_token_accuracy": 0.14334884583950042,
      "num_tokens": 4394446.0,
      "step": 2125
    },
    {
      "entropy": 6.126976585388183,
      "epoch": 0.19321480406386066,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998813908689573,
      "loss": 5.9646,
      "mean_token_accuracy": 0.1414870209991932,
      "num_tokens": 4405882.0,
      "step": 2130
    },
    {
      "entropy": 5.956241512298584,
      "epoch": 0.19366835994194484,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998803380679359,
      "loss": 5.8182,
      "mean_token_accuracy": 0.15368922352790831,
      "num_tokens": 4415280.0,
      "step": 2135
    },
    {
      "entropy": 6.030566740036011,
      "epoch": 0.19412191582002902,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998792806163317,
      "loss": 6.0265,
      "mean_token_accuracy": 0.13928496092557907,
      "num_tokens": 4425755.0,
      "step": 2140
    },
    {
      "entropy": 6.020366859436035,
      "epoch": 0.1945754716981132,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998782185141667,
      "loss": 5.8808,
      "mean_token_accuracy": 0.14801339209079742,
      "num_tokens": 4436259.0,
      "step": 2145
    },
    {
      "entropy": 6.056359338760376,
      "epoch": 0.1950290275761974,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998771517614629,
      "loss": 5.917,
      "mean_token_accuracy": 0.14185069352388383,
      "num_tokens": 4446423.0,
      "step": 2150
    },
    {
      "entropy": 6.050050067901611,
      "epoch": 0.19548258345428157,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998760803582422,
      "loss": 5.871,
      "mean_token_accuracy": 0.15231530964374543,
      "num_tokens": 4456678.0,
      "step": 2155
    },
    {
      "entropy": 6.006132411956787,
      "epoch": 0.19593613933236576,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998750043045269,
      "loss": 5.8574,
      "mean_token_accuracy": 0.15113169997930526,
      "num_tokens": 4467031.0,
      "step": 2160
    },
    {
      "entropy": 6.0697108745574955,
      "epoch": 0.19638969521044994,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998739236003392,
      "loss": 5.938,
      "mean_token_accuracy": 0.13495565131306647,
      "num_tokens": 4477771.0,
      "step": 2165
    },
    {
      "entropy": 6.000240993499756,
      "epoch": 0.1968432510885341,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998728382457014,
      "loss": 5.9664,
      "mean_token_accuracy": 0.14246328622102739,
      "num_tokens": 4488426.0,
      "step": 2170
    },
    {
      "entropy": 6.17234845161438,
      "epoch": 0.19729680696661828,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998717482406362,
      "loss": 5.9372,
      "mean_token_accuracy": 0.13979932889342309,
      "num_tokens": 4499735.0,
      "step": 2175
    },
    {
      "entropy": 6.042126989364624,
      "epoch": 0.19775036284470246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998706535851657,
      "loss": 5.8795,
      "mean_token_accuracy": 0.14715081751346587,
      "num_tokens": 4509500.0,
      "step": 2180
    },
    {
      "entropy": 6.014921045303344,
      "epoch": 0.19820391872278664,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998695542793129,
      "loss": 5.8802,
      "mean_token_accuracy": 0.14634579196572303,
      "num_tokens": 4520431.0,
      "step": 2185
    },
    {
      "entropy": 6.034388875961303,
      "epoch": 0.19865747460087083,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998684503231003,
      "loss": 5.9784,
      "mean_token_accuracy": 0.14680968075990677,
      "num_tokens": 4532294.0,
      "step": 2190
    },
    {
      "entropy": 6.020555639266968,
      "epoch": 0.199111030478955,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499867341716551,
      "loss": 5.8983,
      "mean_token_accuracy": 0.14305620416998863,
      "num_tokens": 4542488.0,
      "step": 2195
    },
    {
      "entropy": 6.039434766769409,
      "epoch": 0.1995645863570392,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998662284596878,
      "loss": 5.8402,
      "mean_token_accuracy": 0.13846864104270934,
      "num_tokens": 4552933.0,
      "step": 2200
    },
    {
      "entropy": 6.0450187683105465,
      "epoch": 0.20001814223512338,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998651105525335,
      "loss": 5.8223,
      "mean_token_accuracy": 0.15138240456581115,
      "num_tokens": 4562373.0,
      "step": 2205
    },
    {
      "entropy": 5.943200492858887,
      "epoch": 0.20047169811320756,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998639879951115,
      "loss": 5.9439,
      "mean_token_accuracy": 0.14522400572896005,
      "num_tokens": 4572715.0,
      "step": 2210
    },
    {
      "entropy": 6.120235109329224,
      "epoch": 0.2009252539912917,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499862860787445,
      "loss": 5.8307,
      "mean_token_accuracy": 0.14360565692186356,
      "num_tokens": 4582782.0,
      "step": 2215
    },
    {
      "entropy": 6.041302442550659,
      "epoch": 0.2013788098693759,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998617289295571,
      "loss": 5.9352,
      "mean_token_accuracy": 0.13825485333800316,
      "num_tokens": 4594108.0,
      "step": 2220
    },
    {
      "entropy": 6.074282169342041,
      "epoch": 0.20183236574746008,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998605924214714,
      "loss": 5.8911,
      "mean_token_accuracy": 0.14501106441020967,
      "num_tokens": 4604732.0,
      "step": 2225
    },
    {
      "entropy": 6.039686012268066,
      "epoch": 0.20228592162554426,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998594512632114,
      "loss": 6.0324,
      "mean_token_accuracy": 0.13512664213776587,
      "num_tokens": 4614998.0,
      "step": 2230
    },
    {
      "entropy": 6.029949378967285,
      "epoch": 0.20273947750362845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998583054548006,
      "loss": 5.8536,
      "mean_token_accuracy": 0.14373639225959778,
      "num_tokens": 4624665.0,
      "step": 2235
    },
    {
      "entropy": 6.060518264770508,
      "epoch": 0.20319303338171263,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998571549962628,
      "loss": 5.8374,
      "mean_token_accuracy": 0.15260313004255294,
      "num_tokens": 4633731.0,
      "step": 2240
    },
    {
      "entropy": 6.0299561500549315,
      "epoch": 0.2036465892597968,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998559998876216,
      "loss": 5.8706,
      "mean_token_accuracy": 0.1454593777656555,
      "num_tokens": 4643421.0,
      "step": 2245
    },
    {
      "entropy": 6.054778814315796,
      "epoch": 0.204100145137881,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998548401289012,
      "loss": 5.9553,
      "mean_token_accuracy": 0.1441769301891327,
      "num_tokens": 4653795.0,
      "step": 2250
    },
    {
      "entropy": 6.058734989166259,
      "epoch": 0.20455370101596518,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998536757201253,
      "loss": 5.9276,
      "mean_token_accuracy": 0.1450178414583206,
      "num_tokens": 4664840.0,
      "step": 2255
    },
    {
      "entropy": 6.073224878311157,
      "epoch": 0.20500725689404933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998525066613181,
      "loss": 5.931,
      "mean_token_accuracy": 0.14578341245651244,
      "num_tokens": 4675508.0,
      "step": 2260
    },
    {
      "entropy": 5.989795446395874,
      "epoch": 0.20546081277213352,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998513329525038,
      "loss": 5.8523,
      "mean_token_accuracy": 0.14992504566907883,
      "num_tokens": 4685995.0,
      "step": 2265
    },
    {
      "entropy": 6.086735725402832,
      "epoch": 0.2059143686502177,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998501545937066,
      "loss": 5.8789,
      "mean_token_accuracy": 0.14490637332201003,
      "num_tokens": 4696659.0,
      "step": 2270
    },
    {
      "entropy": 5.985673189163208,
      "epoch": 0.20636792452830188,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499848971584951,
      "loss": 5.8403,
      "mean_token_accuracy": 0.15447261780500413,
      "num_tokens": 4706438.0,
      "step": 2275
    },
    {
      "entropy": 5.944012832641602,
      "epoch": 0.20682148040638607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998477839262611,
      "loss": 5.8513,
      "mean_token_accuracy": 0.1460217833518982,
      "num_tokens": 4716402.0,
      "step": 2280
    },
    {
      "entropy": 6.024405479431152,
      "epoch": 0.20727503628447025,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499846591617662,
      "loss": 5.9409,
      "mean_token_accuracy": 0.14774347990751266,
      "num_tokens": 4728488.0,
      "step": 2285
    },
    {
      "entropy": 6.027887296676636,
      "epoch": 0.20772859216255443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998453946591779,
      "loss": 5.9638,
      "mean_token_accuracy": 0.14974903911352158,
      "num_tokens": 4738693.0,
      "step": 2290
    },
    {
      "entropy": 6.020720100402832,
      "epoch": 0.20818214804063861,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998441930508339,
      "loss": 5.8436,
      "mean_token_accuracy": 0.1517803430557251,
      "num_tokens": 4748895.0,
      "step": 2295
    },
    {
      "entropy": 6.089075136184692,
      "epoch": 0.2086357039187228,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998429867926547,
      "loss": 5.9548,
      "mean_token_accuracy": 0.14135059341788292,
      "num_tokens": 4760135.0,
      "step": 2300
    },
    {
      "entropy": 6.095640230178833,
      "epoch": 0.20908925979680695,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998417758846651,
      "loss": 5.9916,
      "mean_token_accuracy": 0.15003142952919007,
      "num_tokens": 4769935.0,
      "step": 2305
    },
    {
      "entropy": 6.164234828948975,
      "epoch": 0.20954281567489114,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998405603268903,
      "loss": 5.9117,
      "mean_token_accuracy": 0.14601739197969438,
      "num_tokens": 4779520.0,
      "step": 2310
    },
    {
      "entropy": 5.891173362731934,
      "epoch": 0.20999637155297532,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998393401193554,
      "loss": 5.8264,
      "mean_token_accuracy": 0.1525731474161148,
      "num_tokens": 4791256.0,
      "step": 2315
    },
    {
      "entropy": 5.922054576873779,
      "epoch": 0.2104499274310595,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998381152620857,
      "loss": 5.7778,
      "mean_token_accuracy": 0.15643645375967025,
      "num_tokens": 4801334.0,
      "step": 2320
    },
    {
      "entropy": 5.940741872787475,
      "epoch": 0.21090348330914369,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998368857551067,
      "loss": 5.7694,
      "mean_token_accuracy": 0.15321901589632034,
      "num_tokens": 4811361.0,
      "step": 2325
    },
    {
      "entropy": 5.995629167556762,
      "epoch": 0.21135703918722787,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998356515984433,
      "loss": 5.8362,
      "mean_token_accuracy": 0.14930329322814942,
      "num_tokens": 4821190.0,
      "step": 2330
    },
    {
      "entropy": 5.996551895141602,
      "epoch": 0.21181059506531205,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998344127921213,
      "loss": 5.8616,
      "mean_token_accuracy": 0.15249975472688676,
      "num_tokens": 4830694.0,
      "step": 2335
    },
    {
      "entropy": 5.9884954452514645,
      "epoch": 0.21226415094339623,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998331693361666,
      "loss": 5.8478,
      "mean_token_accuracy": 0.14835911467671395,
      "num_tokens": 4841344.0,
      "step": 2340
    },
    {
      "entropy": 5.958621692657471,
      "epoch": 0.21271770682148042,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998319212306046,
      "loss": 5.8214,
      "mean_token_accuracy": 0.1466705918312073,
      "num_tokens": 4851161.0,
      "step": 2345
    },
    {
      "entropy": 5.9765862941741945,
      "epoch": 0.21317126269956457,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499830668475461,
      "loss": 5.8198,
      "mean_token_accuracy": 0.1497928962111473,
      "num_tokens": 4860740.0,
      "step": 2350
    },
    {
      "entropy": 5.9823229789733885,
      "epoch": 0.21362481857764876,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499829411070762,
      "loss": 5.8482,
      "mean_token_accuracy": 0.14942808002233504,
      "num_tokens": 4871733.0,
      "step": 2355
    },
    {
      "entropy": 6.0018017292022705,
      "epoch": 0.21407837445573294,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998281490165335,
      "loss": 5.9064,
      "mean_token_accuracy": 0.14177072122693063,
      "num_tokens": 4883617.0,
      "step": 2360
    },
    {
      "entropy": 6.005364894866943,
      "epoch": 0.21453193033381712,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998268823128016,
      "loss": 5.8483,
      "mean_token_accuracy": 0.14455199390649795,
      "num_tokens": 4894519.0,
      "step": 2365
    },
    {
      "entropy": 5.98149266242981,
      "epoch": 0.2149854862119013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998256109595923,
      "loss": 5.8014,
      "mean_token_accuracy": 0.14887023940682412,
      "num_tokens": 4904965.0,
      "step": 2370
    },
    {
      "entropy": 5.936092567443848,
      "epoch": 0.2154390420899855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998243349569323,
      "loss": 5.7256,
      "mean_token_accuracy": 0.1545273095369339,
      "num_tokens": 4914618.0,
      "step": 2375
    },
    {
      "entropy": 5.868344449996949,
      "epoch": 0.21589259796806967,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998230543048477,
      "loss": 5.7769,
      "mean_token_accuracy": 0.15782049298286438,
      "num_tokens": 4924341.0,
      "step": 2380
    },
    {
      "entropy": 5.880142879486084,
      "epoch": 0.21634615384615385,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499821769003365,
      "loss": 5.7439,
      "mean_token_accuracy": 0.15210559889674186,
      "num_tokens": 4934231.0,
      "step": 2385
    },
    {
      "entropy": 5.974342632293701,
      "epoch": 0.21679970972423804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998204790525109,
      "loss": 5.7777,
      "mean_token_accuracy": 0.1491807922720909,
      "num_tokens": 4944303.0,
      "step": 2390
    },
    {
      "entropy": 5.959984636306762,
      "epoch": 0.2172532656023222,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998191844523119,
      "loss": 5.8578,
      "mean_token_accuracy": 0.15104133933782576,
      "num_tokens": 4954571.0,
      "step": 2395
    },
    {
      "entropy": 5.957349252700806,
      "epoch": 0.21770682148040638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499817885202795,
      "loss": 5.8266,
      "mean_token_accuracy": 0.14804138243198395,
      "num_tokens": 4964418.0,
      "step": 2400
    },
    {
      "entropy": 5.975922727584839,
      "epoch": 0.21816037735849056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998165813039869,
      "loss": 5.8163,
      "mean_token_accuracy": 0.1464727595448494,
      "num_tokens": 4975036.0,
      "step": 2405
    },
    {
      "entropy": 5.98774995803833,
      "epoch": 0.21861393323657474,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998152727559145,
      "loss": 5.9215,
      "mean_token_accuracy": 0.14540747702121734,
      "num_tokens": 4985272.0,
      "step": 2410
    },
    {
      "entropy": 6.043728685379028,
      "epoch": 0.21906748911465893,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499813959558605,
      "loss": 5.8498,
      "mean_token_accuracy": 0.15086397528648376,
      "num_tokens": 4995187.0,
      "step": 2415
    },
    {
      "entropy": 6.0442393779754635,
      "epoch": 0.2195210449927431,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998126417120856,
      "loss": 5.8919,
      "mean_token_accuracy": 0.15320269912481307,
      "num_tokens": 5005013.0,
      "step": 2420
    },
    {
      "entropy": 6.054208087921142,
      "epoch": 0.2199746008708273,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998113192163835,
      "loss": 5.9139,
      "mean_token_accuracy": 0.1428740605711937,
      "num_tokens": 5016789.0,
      "step": 2425
    },
    {
      "entropy": 5.934090423583984,
      "epoch": 0.22042815674891147,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499809992071526,
      "loss": 5.8588,
      "mean_token_accuracy": 0.147943913936615,
      "num_tokens": 5026913.0,
      "step": 2430
    },
    {
      "entropy": 5.962672376632691,
      "epoch": 0.22088171262699566,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998086602775406,
      "loss": 5.9471,
      "mean_token_accuracy": 0.1470245487987995,
      "num_tokens": 5037501.0,
      "step": 2435
    },
    {
      "entropy": 6.027901649475098,
      "epoch": 0.2213352685050798,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998073238344547,
      "loss": 5.7987,
      "mean_token_accuracy": 0.14490654170513154,
      "num_tokens": 5048478.0,
      "step": 2440
    },
    {
      "entropy": 5.919871520996094,
      "epoch": 0.221788824383164,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998059827422962,
      "loss": 5.859,
      "mean_token_accuracy": 0.1465262070298195,
      "num_tokens": 5059118.0,
      "step": 2445
    },
    {
      "entropy": 6.006969404220581,
      "epoch": 0.22224238026124818,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998046370010926,
      "loss": 5.8236,
      "mean_token_accuracy": 0.15497831404209136,
      "num_tokens": 5070180.0,
      "step": 2450
    },
    {
      "entropy": 5.935058832168579,
      "epoch": 0.22269593613933236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998032866108718,
      "loss": 5.8128,
      "mean_token_accuracy": 0.14882795214653016,
      "num_tokens": 5079912.0,
      "step": 2455
    },
    {
      "entropy": 6.052799558639526,
      "epoch": 0.22314949201741655,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998019315716618,
      "loss": 5.9009,
      "mean_token_accuracy": 0.1513648509979248,
      "num_tokens": 5090549.0,
      "step": 2460
    },
    {
      "entropy": 5.9224653244018555,
      "epoch": 0.22360304789550073,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998005718834905,
      "loss": 5.7505,
      "mean_token_accuracy": 0.15711326152086258,
      "num_tokens": 5100880.0,
      "step": 2465
    },
    {
      "entropy": 5.927716445922852,
      "epoch": 0.2240566037735849,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997992075463862,
      "loss": 5.8622,
      "mean_token_accuracy": 0.14909446835517884,
      "num_tokens": 5111796.0,
      "step": 2470
    },
    {
      "entropy": 6.045079755783081,
      "epoch": 0.2245101596516691,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997978385603769,
      "loss": 5.862,
      "mean_token_accuracy": 0.1490783780813217,
      "num_tokens": 5123468.0,
      "step": 2475
    },
    {
      "entropy": 5.968772125244141,
      "epoch": 0.22496371552975328,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997964649254911,
      "loss": 5.7749,
      "mean_token_accuracy": 0.15539813786745071,
      "num_tokens": 5132925.0,
      "step": 2480
    },
    {
      "entropy": 5.98187403678894,
      "epoch": 0.22541727140783743,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997950866417572,
      "loss": 5.8555,
      "mean_token_accuracy": 0.14465042054653168,
      "num_tokens": 5143445.0,
      "step": 2485
    },
    {
      "entropy": 6.025763702392578,
      "epoch": 0.22587082728592162,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997937037092034,
      "loss": 5.7817,
      "mean_token_accuracy": 0.14337050616741182,
      "num_tokens": 5153176.0,
      "step": 2490
    },
    {
      "entropy": 6.005899715423584,
      "epoch": 0.2263243831640058,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997923161278587,
      "loss": 5.8267,
      "mean_token_accuracy": 0.1517674669623375,
      "num_tokens": 5162765.0,
      "step": 2495
    },
    {
      "entropy": 5.946051549911499,
      "epoch": 0.22677793904208998,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997909238977514,
      "loss": 5.8804,
      "mean_token_accuracy": 0.15172123089432715,
      "num_tokens": 5173390.0,
      "step": 2500
    },
    {
      "entropy": 5.946356534957886,
      "epoch": 0.22723149492017417,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997895270189108,
      "loss": 5.7811,
      "mean_token_accuracy": 0.14956301152706147,
      "num_tokens": 5184212.0,
      "step": 2505
    },
    {
      "entropy": 5.886882019042969,
      "epoch": 0.22768505079825835,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997881254913653,
      "loss": 5.7728,
      "mean_token_accuracy": 0.15003687813878058,
      "num_tokens": 5194311.0,
      "step": 2510
    },
    {
      "entropy": 5.916078996658325,
      "epoch": 0.22813860667634253,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997867193151441,
      "loss": 5.7712,
      "mean_token_accuracy": 0.15247755497694016,
      "num_tokens": 5204708.0,
      "step": 2515
    },
    {
      "entropy": 5.87827501296997,
      "epoch": 0.22859216255442671,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997853084902763,
      "loss": 5.7469,
      "mean_token_accuracy": 0.15128237307071685,
      "num_tokens": 5214566.0,
      "step": 2520
    },
    {
      "entropy": 5.943019533157349,
      "epoch": 0.2290457184325109,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997838930167911,
      "loss": 5.7878,
      "mean_token_accuracy": 0.15266707986593248,
      "num_tokens": 5224762.0,
      "step": 2525
    },
    {
      "entropy": 5.9829565525054935,
      "epoch": 0.22949927431059505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997824728947176,
      "loss": 5.8435,
      "mean_token_accuracy": 0.15028240010142327,
      "num_tokens": 5235146.0,
      "step": 2530
    },
    {
      "entropy": 6.023767328262329,
      "epoch": 0.22995283018867924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997810481240853,
      "loss": 5.8108,
      "mean_token_accuracy": 0.14803099185228347,
      "num_tokens": 5243867.0,
      "step": 2535
    },
    {
      "entropy": 5.893141078948974,
      "epoch": 0.23040638606676342,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997796187049238,
      "loss": 5.7234,
      "mean_token_accuracy": 0.1500532016158104,
      "num_tokens": 5254279.0,
      "step": 2540
    },
    {
      "entropy": 6.080553483963013,
      "epoch": 0.2308599419448476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997781846372623,
      "loss": 5.8382,
      "mean_token_accuracy": 0.14982936531305313,
      "num_tokens": 5264769.0,
      "step": 2545
    },
    {
      "entropy": 5.93544192314148,
      "epoch": 0.23131349782293179,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997767459211307,
      "loss": 5.7903,
      "mean_token_accuracy": 0.14275677353143693,
      "num_tokens": 5274291.0,
      "step": 2550
    },
    {
      "entropy": 5.883300113677978,
      "epoch": 0.23176705370101597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997753025565587,
      "loss": 5.7326,
      "mean_token_accuracy": 0.15632595419883727,
      "num_tokens": 5284366.0,
      "step": 2555
    },
    {
      "entropy": 5.95240535736084,
      "epoch": 0.23222060957910015,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997738545435763,
      "loss": 5.8456,
      "mean_token_accuracy": 0.15440944731235504,
      "num_tokens": 5294494.0,
      "step": 2560
    },
    {
      "entropy": 5.928846740722657,
      "epoch": 0.23267416545718433,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997724018822133,
      "loss": 5.8394,
      "mean_token_accuracy": 0.15485012233257295,
      "num_tokens": 5305606.0,
      "step": 2565
    },
    {
      "entropy": 5.960556077957153,
      "epoch": 0.23312772133526852,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997709445724996,
      "loss": 5.8465,
      "mean_token_accuracy": 0.150348149985075,
      "num_tokens": 5315677.0,
      "step": 2570
    },
    {
      "entropy": 5.986758232116699,
      "epoch": 0.23358127721335267,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997694826144656,
      "loss": 5.9049,
      "mean_token_accuracy": 0.1463898465037346,
      "num_tokens": 5326453.0,
      "step": 2575
    },
    {
      "entropy": 5.92100830078125,
      "epoch": 0.23403483309143686,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997680160081414,
      "loss": 5.7906,
      "mean_token_accuracy": 0.15171819776296616,
      "num_tokens": 5337591.0,
      "step": 2580
    },
    {
      "entropy": 5.9538902759552,
      "epoch": 0.23448838896952104,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997665447535573,
      "loss": 5.7791,
      "mean_token_accuracy": 0.15522343516349793,
      "num_tokens": 5349400.0,
      "step": 2585
    },
    {
      "entropy": 6.00141224861145,
      "epoch": 0.23494194484760522,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997650688507438,
      "loss": 5.7634,
      "mean_token_accuracy": 0.1506360337138176,
      "num_tokens": 5359984.0,
      "step": 2590
    },
    {
      "entropy": 5.916184520721435,
      "epoch": 0.2353955007256894,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997635882997314,
      "loss": 5.845,
      "mean_token_accuracy": 0.15854744762182235,
      "num_tokens": 5369254.0,
      "step": 2595
    },
    {
      "entropy": 5.9528069496154785,
      "epoch": 0.2358490566037736,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997621031005507,
      "loss": 5.7798,
      "mean_token_accuracy": 0.15245212465524674,
      "num_tokens": 5379648.0,
      "step": 2600
    },
    {
      "entropy": 5.849475145339966,
      "epoch": 0.23630261248185777,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997606132532325,
      "loss": 5.7372,
      "mean_token_accuracy": 0.154351457208395,
      "num_tokens": 5390321.0,
      "step": 2605
    },
    {
      "entropy": 5.919327926635742,
      "epoch": 0.23675616835994195,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997591187578075,
      "loss": 5.7926,
      "mean_token_accuracy": 0.15511986911296843,
      "num_tokens": 5401078.0,
      "step": 2610
    },
    {
      "entropy": 5.961600828170776,
      "epoch": 0.23720972423802614,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997576196143067,
      "loss": 5.7453,
      "mean_token_accuracy": 0.1566541612148285,
      "num_tokens": 5411977.0,
      "step": 2615
    },
    {
      "entropy": 5.893222713470459,
      "epoch": 0.2376632801161103,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499756115822761,
      "loss": 5.7368,
      "mean_token_accuracy": 0.15138195902109147,
      "num_tokens": 5422967.0,
      "step": 2620
    },
    {
      "entropy": 5.9912902355194095,
      "epoch": 0.23811683599419448,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997546073832015,
      "loss": 5.7835,
      "mean_token_accuracy": 0.16118090003728866,
      "num_tokens": 5432260.0,
      "step": 2625
    },
    {
      "entropy": 5.888670158386231,
      "epoch": 0.23857039187227866,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997530942956595,
      "loss": 5.7564,
      "mean_token_accuracy": 0.15361348688602447,
      "num_tokens": 5441706.0,
      "step": 2630
    },
    {
      "entropy": 5.94991717338562,
      "epoch": 0.23902394775036284,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997515765601662,
      "loss": 5.7622,
      "mean_token_accuracy": 0.1588932454586029,
      "num_tokens": 5452397.0,
      "step": 2635
    },
    {
      "entropy": 5.880772686004638,
      "epoch": 0.23947750362844702,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997500541767531,
      "loss": 5.7218,
      "mean_token_accuracy": 0.15416701436042785,
      "num_tokens": 5463762.0,
      "step": 2640
    },
    {
      "entropy": 5.901000833511352,
      "epoch": 0.2399310595065312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997485271454516,
      "loss": 5.7915,
      "mean_token_accuracy": 0.14654496908187867,
      "num_tokens": 5474622.0,
      "step": 2645
    },
    {
      "entropy": 5.944158935546875,
      "epoch": 0.2403846153846154,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997469954662933,
      "loss": 5.7691,
      "mean_token_accuracy": 0.15286272019147873,
      "num_tokens": 5484697.0,
      "step": 2650
    },
    {
      "entropy": 5.847146940231323,
      "epoch": 0.24083817126269957,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997454591393097,
      "loss": 5.7742,
      "mean_token_accuracy": 0.15464416444301604,
      "num_tokens": 5496433.0,
      "step": 2655
    },
    {
      "entropy": 5.972603416442871,
      "epoch": 0.24129172714078376,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997439181645329,
      "loss": 5.8895,
      "mean_token_accuracy": 0.1483437992632389,
      "num_tokens": 5507917.0,
      "step": 2660
    },
    {
      "entropy": 6.07913875579834,
      "epoch": 0.2417452830188679,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997423725419945,
      "loss": 5.8717,
      "mean_token_accuracy": 0.14761385768651963,
      "num_tokens": 5519159.0,
      "step": 2665
    },
    {
      "entropy": 5.928588962554931,
      "epoch": 0.2421988388969521,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997408222717267,
      "loss": 5.8249,
      "mean_token_accuracy": 0.15743801593780518,
      "num_tokens": 5529180.0,
      "step": 2670
    },
    {
      "entropy": 5.884641027450561,
      "epoch": 0.24265239477503628,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997392673537612,
      "loss": 5.7382,
      "mean_token_accuracy": 0.16192516684532166,
      "num_tokens": 5539363.0,
      "step": 2675
    },
    {
      "entropy": 5.909084510803223,
      "epoch": 0.24310595065312046,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997377077881306,
      "loss": 5.843,
      "mean_token_accuracy": 0.15110581144690513,
      "num_tokens": 5549841.0,
      "step": 2680
    },
    {
      "entropy": 5.940350866317749,
      "epoch": 0.24355950653120464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997361435748668,
      "loss": 5.7156,
      "mean_token_accuracy": 0.161916483938694,
      "num_tokens": 5560299.0,
      "step": 2685
    },
    {
      "entropy": 5.962108039855957,
      "epoch": 0.24401306240928883,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997345747140024,
      "loss": 5.8001,
      "mean_token_accuracy": 0.15469845086336137,
      "num_tokens": 5571081.0,
      "step": 2690
    },
    {
      "entropy": 5.871194362640381,
      "epoch": 0.244466618287373,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997330012055695,
      "loss": 5.7317,
      "mean_token_accuracy": 0.15739612132310868,
      "num_tokens": 5581389.0,
      "step": 2695
    },
    {
      "entropy": 5.881601572036743,
      "epoch": 0.2449201741654572,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997314230496011,
      "loss": 5.779,
      "mean_token_accuracy": 0.14633584022521973,
      "num_tokens": 5592347.0,
      "step": 2700
    },
    {
      "entropy": 5.889206314086914,
      "epoch": 0.24537373004354138,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997298402461296,
      "loss": 5.7548,
      "mean_token_accuracy": 0.15528458803892137,
      "num_tokens": 5603411.0,
      "step": 2705
    },
    {
      "entropy": 5.930398607254029,
      "epoch": 0.24582728592162553,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997282527951876,
      "loss": 5.7452,
      "mean_token_accuracy": 0.15865262895822524,
      "num_tokens": 5614615.0,
      "step": 2710
    },
    {
      "entropy": 5.811875200271606,
      "epoch": 0.24628084179970972,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997266606968081,
      "loss": 5.7026,
      "mean_token_accuracy": 0.15606526732444764,
      "num_tokens": 5624582.0,
      "step": 2715
    },
    {
      "entropy": 5.876440620422363,
      "epoch": 0.2467343976777939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997250639510241,
      "loss": 5.7258,
      "mean_token_accuracy": 0.15305925607681276,
      "num_tokens": 5634310.0,
      "step": 2720
    },
    {
      "entropy": 5.932286167144776,
      "epoch": 0.24718795355587808,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997234625578684,
      "loss": 5.8186,
      "mean_token_accuracy": 0.15080375373363494,
      "num_tokens": 5644018.0,
      "step": 2725
    },
    {
      "entropy": 5.862547779083252,
      "epoch": 0.24764150943396226,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997218565173742,
      "loss": 5.7485,
      "mean_token_accuracy": 0.15690971165895462,
      "num_tokens": 5653592.0,
      "step": 2730
    },
    {
      "entropy": 5.973826742172241,
      "epoch": 0.24809506531204645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997202458295748,
      "loss": 5.7929,
      "mean_token_accuracy": 0.155655175447464,
      "num_tokens": 5662812.0,
      "step": 2735
    },
    {
      "entropy": 5.936570167541504,
      "epoch": 0.24854862119013063,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997186304945034,
      "loss": 5.7758,
      "mean_token_accuracy": 0.1513756588101387,
      "num_tokens": 5672939.0,
      "step": 2740
    },
    {
      "entropy": 5.83735408782959,
      "epoch": 0.24900217706821481,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997170105121935,
      "loss": 5.7543,
      "mean_token_accuracy": 0.15465778708457947,
      "num_tokens": 5683350.0,
      "step": 2745
    },
    {
      "entropy": 5.977609300613404,
      "epoch": 0.249455732946299,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997153858826786,
      "loss": 5.8559,
      "mean_token_accuracy": 0.1457457385957241,
      "num_tokens": 5694516.0,
      "step": 2750
    },
    {
      "entropy": 5.897079944610596,
      "epoch": 0.24990928882438315,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997137566059922,
      "loss": 5.6687,
      "mean_token_accuracy": 0.15890194922685624,
      "num_tokens": 5705443.0,
      "step": 2755
    },
    {
      "entropy": 5.876714992523193,
      "epoch": 0.25036284470246734,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997121226821681,
      "loss": 5.772,
      "mean_token_accuracy": 0.1513215646147728,
      "num_tokens": 5715933.0,
      "step": 2760
    },
    {
      "entropy": 5.92222113609314,
      "epoch": 0.25081640058055155,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049971048411124,
      "loss": 5.835,
      "mean_token_accuracy": 0.14457671642303466,
      "num_tokens": 5726630.0,
      "step": 2765
    },
    {
      "entropy": 6.002235460281372,
      "epoch": 0.2512699564586357,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997088408932418,
      "loss": 5.7697,
      "mean_token_accuracy": 0.1539665475487709,
      "num_tokens": 5736890.0,
      "step": 2770
    },
    {
      "entropy": 5.778037357330322,
      "epoch": 0.25172351233671986,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997071930282076,
      "loss": 5.6726,
      "mean_token_accuracy": 0.1656890943646431,
      "num_tokens": 5747042.0,
      "step": 2775
    },
    {
      "entropy": 5.874831628799439,
      "epoch": 0.25217706821480407,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997055405161712,
      "loss": 5.7902,
      "mean_token_accuracy": 0.1548820838332176,
      "num_tokens": 5757367.0,
      "step": 2780
    },
    {
      "entropy": 5.933564472198486,
      "epoch": 0.2526306240928882,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997038833571671,
      "loss": 5.8466,
      "mean_token_accuracy": 0.14902109056711196,
      "num_tokens": 5769557.0,
      "step": 2785
    },
    {
      "entropy": 5.887054443359375,
      "epoch": 0.25308417997097243,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997022215512295,
      "loss": 5.7514,
      "mean_token_accuracy": 0.15811785459518432,
      "num_tokens": 5780615.0,
      "step": 2790
    },
    {
      "entropy": 5.939122200012207,
      "epoch": 0.2535377358490566,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997005550983925,
      "loss": 5.7361,
      "mean_token_accuracy": 0.1549833208322525,
      "num_tokens": 5791271.0,
      "step": 2795
    },
    {
      "entropy": 5.814103412628174,
      "epoch": 0.2539912917271408,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499698883998691,
      "loss": 5.6636,
      "mean_token_accuracy": 0.15836629867553711,
      "num_tokens": 5801117.0,
      "step": 2800
    },
    {
      "entropy": 5.859375476837158,
      "epoch": 0.25444484760522496,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996972082521591,
      "loss": 5.6955,
      "mean_token_accuracy": 0.15782571583986282,
      "num_tokens": 5811696.0,
      "step": 2805
    },
    {
      "entropy": 5.879450893402099,
      "epoch": 0.25489840348330917,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996955278588319,
      "loss": 5.7319,
      "mean_token_accuracy": 0.16295181214809418,
      "num_tokens": 5822207.0,
      "step": 2810
    },
    {
      "entropy": 5.838560342788696,
      "epoch": 0.2553519593613933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996938428187437,
      "loss": 5.6753,
      "mean_token_accuracy": 0.16222673356533052,
      "num_tokens": 5831927.0,
      "step": 2815
    },
    {
      "entropy": 5.844130134582519,
      "epoch": 0.2558055152394775,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996921531319297,
      "loss": 5.7406,
      "mean_token_accuracy": 0.15631944462656974,
      "num_tokens": 5842093.0,
      "step": 2820
    },
    {
      "entropy": 5.862317848205566,
      "epoch": 0.2562590711175617,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996904587984246,
      "loss": 5.6799,
      "mean_token_accuracy": 0.1637733742594719,
      "num_tokens": 5850551.0,
      "step": 2825
    },
    {
      "entropy": 5.831214189529419,
      "epoch": 0.25671262699564584,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996887598182636,
      "loss": 5.6693,
      "mean_token_accuracy": 0.15734255760908128,
      "num_tokens": 5860651.0,
      "step": 2830
    },
    {
      "entropy": 5.736221647262573,
      "epoch": 0.25716618287373005,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996870561914819,
      "loss": 5.6508,
      "mean_token_accuracy": 0.15711580514907836,
      "num_tokens": 5871500.0,
      "step": 2835
    },
    {
      "entropy": 5.837002754211426,
      "epoch": 0.2576197387518142,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996853479181145,
      "loss": 5.6483,
      "mean_token_accuracy": 0.15363348871469498,
      "num_tokens": 5881523.0,
      "step": 2840
    },
    {
      "entropy": 5.969505023956299,
      "epoch": 0.2580732946298984,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996836349981969,
      "loss": 5.8168,
      "mean_token_accuracy": 0.15440904647111892,
      "num_tokens": 5891162.0,
      "step": 2845
    },
    {
      "entropy": 5.825331020355224,
      "epoch": 0.2585268505079826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996819174317645,
      "loss": 5.6989,
      "mean_token_accuracy": 0.15304600223898887,
      "num_tokens": 5901818.0,
      "step": 2850
    },
    {
      "entropy": 5.837537527084351,
      "epoch": 0.2589804063860668,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996801952188528,
      "loss": 5.7802,
      "mean_token_accuracy": 0.15388137996196746,
      "num_tokens": 5913889.0,
      "step": 2855
    },
    {
      "entropy": 5.898041200637818,
      "epoch": 0.25943396226415094,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996784683594973,
      "loss": 5.7331,
      "mean_token_accuracy": 0.15773886442184448,
      "num_tokens": 5923429.0,
      "step": 2860
    },
    {
      "entropy": 5.9350752353668215,
      "epoch": 0.2598875181422351,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996767368537338,
      "loss": 5.7756,
      "mean_token_accuracy": 0.15665830671787262,
      "num_tokens": 5933847.0,
      "step": 2865
    },
    {
      "entropy": 5.846282577514648,
      "epoch": 0.2603410740203193,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996750007015983,
      "loss": 5.6519,
      "mean_token_accuracy": 0.15802753269672393,
      "num_tokens": 5943874.0,
      "step": 2870
    },
    {
      "entropy": 5.893306922912598,
      "epoch": 0.26079462989840346,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996732599031264,
      "loss": 5.7207,
      "mean_token_accuracy": 0.1564636617898941,
      "num_tokens": 5953653.0,
      "step": 2875
    },
    {
      "entropy": 5.912632322311401,
      "epoch": 0.2612481857764877,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996715144583543,
      "loss": 5.6787,
      "mean_token_accuracy": 0.1598701536655426,
      "num_tokens": 5964039.0,
      "step": 2880
    },
    {
      "entropy": 5.934096145629883,
      "epoch": 0.26170174165457183,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499669764367318,
      "loss": 5.7888,
      "mean_token_accuracy": 0.15126357972621918,
      "num_tokens": 5974083.0,
      "step": 2885
    },
    {
      "entropy": 5.802870368957519,
      "epoch": 0.26215529753265604,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996680096300537,
      "loss": 5.7352,
      "mean_token_accuracy": 0.1535191372036934,
      "num_tokens": 5984629.0,
      "step": 2890
    },
    {
      "entropy": 5.837651777267456,
      "epoch": 0.2626088534107402,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996662502465978,
      "loss": 5.6433,
      "mean_token_accuracy": 0.16809626668691635,
      "num_tokens": 5994295.0,
      "step": 2895
    },
    {
      "entropy": 5.847111940383911,
      "epoch": 0.2630624092888244,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996644862169864,
      "loss": 5.6911,
      "mean_token_accuracy": 0.15560346096754074,
      "num_tokens": 6004156.0,
      "step": 2900
    },
    {
      "entropy": 5.842682647705078,
      "epoch": 0.26351596516690856,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996627175412563,
      "loss": 5.7724,
      "mean_token_accuracy": 0.15826719254255295,
      "num_tokens": 6014188.0,
      "step": 2905
    },
    {
      "entropy": 5.951080179214477,
      "epoch": 0.2639695210449927,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499660944219444,
      "loss": 5.8318,
      "mean_token_accuracy": 0.15130359828472137,
      "num_tokens": 6024879.0,
      "step": 2910
    },
    {
      "entropy": 5.973435401916504,
      "epoch": 0.2644230769230769,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996591662515861,
      "loss": 5.8978,
      "mean_token_accuracy": 0.1513467513024807,
      "num_tokens": 6036374.0,
      "step": 2915
    },
    {
      "entropy": 5.852985286712647,
      "epoch": 0.2648766328011611,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996573836377193,
      "loss": 5.7384,
      "mean_token_accuracy": 0.15605844408273697,
      "num_tokens": 6046432.0,
      "step": 2920
    },
    {
      "entropy": 5.9294966697692875,
      "epoch": 0.2653301886792453,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996555963778806,
      "loss": 5.7428,
      "mean_token_accuracy": 0.14988288134336472,
      "num_tokens": 6056864.0,
      "step": 2925
    },
    {
      "entropy": 5.9699132442474365,
      "epoch": 0.26578374455732945,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499653804472107,
      "loss": 5.7612,
      "mean_token_accuracy": 0.1611533209681511,
      "num_tokens": 6067294.0,
      "step": 2930
    },
    {
      "entropy": 5.882857704162598,
      "epoch": 0.26623730043541366,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996520079204353,
      "loss": 5.7721,
      "mean_token_accuracy": 0.15410346388816834,
      "num_tokens": 6077587.0,
      "step": 2935
    },
    {
      "entropy": 5.921375894546509,
      "epoch": 0.2666908563134978,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499650206722903,
      "loss": 5.7142,
      "mean_token_accuracy": 0.15302591323852538,
      "num_tokens": 6089306.0,
      "step": 2940
    },
    {
      "entropy": 5.801258993148804,
      "epoch": 0.267144412191582,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996484008795471,
      "loss": 5.6718,
      "mean_token_accuracy": 0.15564401000738143,
      "num_tokens": 6099651.0,
      "step": 2945
    },
    {
      "entropy": 5.831615734100342,
      "epoch": 0.2675979680696662,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499646590390405,
      "loss": 5.6786,
      "mean_token_accuracy": 0.15340599417686462,
      "num_tokens": 6110063.0,
      "step": 2950
    },
    {
      "entropy": 5.866582584381104,
      "epoch": 0.26805152394775034,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996447752555142,
      "loss": 5.7069,
      "mean_token_accuracy": 0.1569851815700531,
      "num_tokens": 6120031.0,
      "step": 2955
    },
    {
      "entropy": 5.845797681808472,
      "epoch": 0.26850507982583455,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996429554749122,
      "loss": 5.7447,
      "mean_token_accuracy": 0.1586455672979355,
      "num_tokens": 6131274.0,
      "step": 2960
    },
    {
      "entropy": 5.888718032836914,
      "epoch": 0.2689586357039187,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996411310486367,
      "loss": 5.7833,
      "mean_token_accuracy": 0.15850588232278823,
      "num_tokens": 6140878.0,
      "step": 2965
    },
    {
      "entropy": 5.895853567123413,
      "epoch": 0.2694121915820029,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996393019767253,
      "loss": 5.7917,
      "mean_token_accuracy": 0.15040715038776398,
      "num_tokens": 6152229.0,
      "step": 2970
    },
    {
      "entropy": 5.802438020706177,
      "epoch": 0.26986574746008707,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996374682592158,
      "loss": 5.6873,
      "mean_token_accuracy": 0.15607988238334655,
      "num_tokens": 6162311.0,
      "step": 2975
    },
    {
      "entropy": 5.942750644683838,
      "epoch": 0.2703193033381713,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996356298961463,
      "loss": 5.7936,
      "mean_token_accuracy": 0.15298964083194733,
      "num_tokens": 6172806.0,
      "step": 2980
    },
    {
      "entropy": 5.87895131111145,
      "epoch": 0.27077285921625543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996337868875549,
      "loss": 5.7587,
      "mean_token_accuracy": 0.15776319652795792,
      "num_tokens": 6183479.0,
      "step": 2985
    },
    {
      "entropy": 5.886080932617188,
      "epoch": 0.27122641509433965,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996319392334792,
      "loss": 5.7299,
      "mean_token_accuracy": 0.15342071503400803,
      "num_tokens": 6194201.0,
      "step": 2990
    },
    {
      "entropy": 5.82788724899292,
      "epoch": 0.2716799709724238,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499630086933958,
      "loss": 5.6724,
      "mean_token_accuracy": 0.1564711168408394,
      "num_tokens": 6204774.0,
      "step": 2995
    },
    {
      "entropy": 5.886050224304199,
      "epoch": 0.27213352685050796,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996282299890292,
      "loss": 5.6831,
      "mean_token_accuracy": 0.1526743471622467,
      "num_tokens": 6215056.0,
      "step": 3000
    },
    {
      "epoch": 0.27213352685050796,
      "eval_entropy": 5.7014565256386325,
      "eval_loss": 5.748072624206543,
      "eval_mean_token_accuracy": 0.1622600189245799,
      "eval_num_tokens": 6215056.0,
      "eval_runtime": 20.702,
      "eval_samples_per_second": 1708.048,
      "eval_steps_per_second": 213.506,
      "step": 3000
    },
    {
      "entropy": 5.934528923034668,
      "epoch": 0.27258708272859217,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996263683987316,
      "loss": 5.786,
      "mean_token_accuracy": 0.1579520136117935,
      "num_tokens": 6226264.0,
      "step": 3005
    },
    {
      "entropy": 5.856761884689331,
      "epoch": 0.2730406386066763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996245021631033,
      "loss": 5.7038,
      "mean_token_accuracy": 0.161579892039299,
      "num_tokens": 6235137.0,
      "step": 3010
    },
    {
      "entropy": 5.824236536026001,
      "epoch": 0.27349419448476053,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499622631282183,
      "loss": 5.7119,
      "mean_token_accuracy": 0.15795171409845352,
      "num_tokens": 6245411.0,
      "step": 3015
    },
    {
      "entropy": 5.915331649780273,
      "epoch": 0.2739477503628447,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996207557560095,
      "loss": 5.7743,
      "mean_token_accuracy": 0.1543523669242859,
      "num_tokens": 6255633.0,
      "step": 3020
    },
    {
      "entropy": 5.906075763702392,
      "epoch": 0.2744013062409289,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996188755846216,
      "loss": 5.7388,
      "mean_token_accuracy": 0.1594541311264038,
      "num_tokens": 6265109.0,
      "step": 3025
    },
    {
      "entropy": 5.813056755065918,
      "epoch": 0.27485486211901305,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499616990768058,
      "loss": 5.7737,
      "mean_token_accuracy": 0.1503493830561638,
      "num_tokens": 6275241.0,
      "step": 3030
    },
    {
      "entropy": 5.902055835723877,
      "epoch": 0.27530841799709727,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996151013063579,
      "loss": 5.8127,
      "mean_token_accuracy": 0.15131743252277374,
      "num_tokens": 6286603.0,
      "step": 3035
    },
    {
      "entropy": 5.917137479782104,
      "epoch": 0.2757619738751814,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996132071995601,
      "loss": 5.6609,
      "mean_token_accuracy": 0.16706326305866243,
      "num_tokens": 6297219.0,
      "step": 3040
    },
    {
      "entropy": 5.804540967941284,
      "epoch": 0.2762155297532656,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996113084477042,
      "loss": 5.6991,
      "mean_token_accuracy": 0.16197659075260162,
      "num_tokens": 6307694.0,
      "step": 3045
    },
    {
      "entropy": 5.795589399337769,
      "epoch": 0.2766690856313498,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996094050508291,
      "loss": 5.6305,
      "mean_token_accuracy": 0.16191300749778748,
      "num_tokens": 6317384.0,
      "step": 3050
    },
    {
      "entropy": 5.88090558052063,
      "epoch": 0.27712264150943394,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996074970089741,
      "loss": 5.7749,
      "mean_token_accuracy": 0.15503887385129927,
      "num_tokens": 6327558.0,
      "step": 3055
    },
    {
      "entropy": 5.878848552703857,
      "epoch": 0.27757619738751815,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996055843221789,
      "loss": 5.6264,
      "mean_token_accuracy": 0.16327428370714187,
      "num_tokens": 6337828.0,
      "step": 3060
    },
    {
      "entropy": 5.789236450195313,
      "epoch": 0.2780297532656023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996036669904831,
      "loss": 5.6907,
      "mean_token_accuracy": 0.16211704909801483,
      "num_tokens": 6348678.0,
      "step": 3065
    },
    {
      "entropy": 5.864919281005859,
      "epoch": 0.2784833091436865,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996017450139261,
      "loss": 5.783,
      "mean_token_accuracy": 0.14863884598016738,
      "num_tokens": 6359670.0,
      "step": 3070
    },
    {
      "entropy": 5.968852424621582,
      "epoch": 0.2789368650217707,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995998183925478,
      "loss": 5.7405,
      "mean_token_accuracy": 0.15753814876079558,
      "num_tokens": 6371034.0,
      "step": 3075
    },
    {
      "entropy": 5.818412113189697,
      "epoch": 0.2793904208998549,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995978871263881,
      "loss": 5.6874,
      "mean_token_accuracy": 0.15982511192560195,
      "num_tokens": 6380672.0,
      "step": 3080
    },
    {
      "entropy": 5.8274352073669435,
      "epoch": 0.27984397677793904,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995959512154868,
      "loss": 5.7991,
      "mean_token_accuracy": 0.14778632521629334,
      "num_tokens": 6390799.0,
      "step": 3085
    },
    {
      "entropy": 5.859681892395019,
      "epoch": 0.2802975326560232,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499594010659884,
      "loss": 5.6602,
      "mean_token_accuracy": 0.1630452647805214,
      "num_tokens": 6400965.0,
      "step": 3090
    },
    {
      "entropy": 5.886337471008301,
      "epoch": 0.2807510885341074,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995920654596199,
      "loss": 5.6873,
      "mean_token_accuracy": 0.1525515004992485,
      "num_tokens": 6411567.0,
      "step": 3095
    },
    {
      "entropy": 5.864609479904175,
      "epoch": 0.28120464441219156,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995901156147345,
      "loss": 5.7302,
      "mean_token_accuracy": 0.16160865426063536,
      "num_tokens": 6421193.0,
      "step": 3100
    },
    {
      "entropy": 5.852170133590699,
      "epoch": 0.2816582002902758,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995881611252684,
      "loss": 5.7431,
      "mean_token_accuracy": 0.16411930173635483,
      "num_tokens": 6430281.0,
      "step": 3105
    },
    {
      "entropy": 5.838845443725586,
      "epoch": 0.28211175616835993,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995862019912619,
      "loss": 5.7401,
      "mean_token_accuracy": 0.1545553356409073,
      "num_tokens": 6441055.0,
      "step": 3110
    },
    {
      "entropy": 5.85880446434021,
      "epoch": 0.28256531204644414,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995842382127555,
      "loss": 5.665,
      "mean_token_accuracy": 0.15411336421966554,
      "num_tokens": 6451082.0,
      "step": 3115
    },
    {
      "entropy": 5.826427602767945,
      "epoch": 0.2830188679245283,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995822697897897,
      "loss": 5.7501,
      "mean_token_accuracy": 0.15211194902658462,
      "num_tokens": 6461617.0,
      "step": 3120
    },
    {
      "entropy": 5.894758319854736,
      "epoch": 0.2834724238026125,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995802967224056,
      "loss": 5.7542,
      "mean_token_accuracy": 0.15588101148605346,
      "num_tokens": 6472780.0,
      "step": 3125
    },
    {
      "entropy": 5.758359670639038,
      "epoch": 0.28392597968069666,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995783190106434,
      "loss": 5.6652,
      "mean_token_accuracy": 0.15255783200263978,
      "num_tokens": 6484078.0,
      "step": 3130
    },
    {
      "entropy": 5.893905305862427,
      "epoch": 0.2843795355587808,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995763366545445,
      "loss": 5.7515,
      "mean_token_accuracy": 0.15216615051031113,
      "num_tokens": 6494366.0,
      "step": 3135
    },
    {
      "entropy": 5.754382371902466,
      "epoch": 0.284833091436865,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995743496541497,
      "loss": 5.6839,
      "mean_token_accuracy": 0.16558870375156404,
      "num_tokens": 6505461.0,
      "step": 3140
    },
    {
      "entropy": 5.835577821731567,
      "epoch": 0.2852866473149492,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995723580095001,
      "loss": 5.6808,
      "mean_token_accuracy": 0.15959053486585617,
      "num_tokens": 6516141.0,
      "step": 3145
    },
    {
      "entropy": 5.781411981582641,
      "epoch": 0.2857402031930334,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499570361720637,
      "loss": 5.6851,
      "mean_token_accuracy": 0.154303365200758,
      "num_tokens": 6525536.0,
      "step": 3150
    },
    {
      "entropy": 5.864716672897339,
      "epoch": 0.28619375907111755,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995683607876015,
      "loss": 5.7363,
      "mean_token_accuracy": 0.15480921417474747,
      "num_tokens": 6535852.0,
      "step": 3155
    },
    {
      "entropy": 5.880465698242188,
      "epoch": 0.28664731494920176,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995663552104352,
      "loss": 5.7094,
      "mean_token_accuracy": 0.16345915347337722,
      "num_tokens": 6546353.0,
      "step": 3160
    },
    {
      "entropy": 5.76881217956543,
      "epoch": 0.2871008708272859,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995643449891793,
      "loss": 5.6586,
      "mean_token_accuracy": 0.15820098519325257,
      "num_tokens": 6557782.0,
      "step": 3165
    },
    {
      "entropy": 5.869258546829224,
      "epoch": 0.2875544267053701,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995623301238757,
      "loss": 5.6772,
      "mean_token_accuracy": 0.16326069831848145,
      "num_tokens": 6568100.0,
      "step": 3170
    },
    {
      "entropy": 5.809826850891113,
      "epoch": 0.2880079825834543,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995603106145657,
      "loss": 5.7354,
      "mean_token_accuracy": 0.15384810715913771,
      "num_tokens": 6578698.0,
      "step": 3175
    },
    {
      "entropy": 5.877403926849365,
      "epoch": 0.28846153846153844,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995582864612913,
      "loss": 5.7211,
      "mean_token_accuracy": 0.15826957821846008,
      "num_tokens": 6588908.0,
      "step": 3180
    },
    {
      "entropy": 5.767029571533203,
      "epoch": 0.28891509433962265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995562576640943,
      "loss": 5.577,
      "mean_token_accuracy": 0.16004678606987,
      "num_tokens": 6599445.0,
      "step": 3185
    },
    {
      "entropy": 5.799728536605835,
      "epoch": 0.2893686502177068,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995542242230167,
      "loss": 5.6578,
      "mean_token_accuracy": 0.16209590137004853,
      "num_tokens": 6610246.0,
      "step": 3190
    },
    {
      "entropy": 5.798572778701782,
      "epoch": 0.289822206095791,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995521861381006,
      "loss": 5.6769,
      "mean_token_accuracy": 0.15961335003376007,
      "num_tokens": 6620501.0,
      "step": 3195
    },
    {
      "entropy": 5.898326778411866,
      "epoch": 0.29027576197387517,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499550143409388,
      "loss": 5.7474,
      "mean_token_accuracy": 0.15129139870405198,
      "num_tokens": 6631206.0,
      "step": 3200
    },
    {
      "entropy": 5.829407978057861,
      "epoch": 0.2907293178519594,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995480960369211,
      "loss": 5.6945,
      "mean_token_accuracy": 0.1549079790711403,
      "num_tokens": 6641600.0,
      "step": 3205
    },
    {
      "entropy": 5.8490814685821535,
      "epoch": 0.29118287373004353,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995460440207425,
      "loss": 5.7631,
      "mean_token_accuracy": 0.15415688455104828,
      "num_tokens": 6652319.0,
      "step": 3210
    },
    {
      "entropy": 5.883116722106934,
      "epoch": 0.29163642960812775,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995439873608945,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1500185400247574,
      "num_tokens": 6663702.0,
      "step": 3215
    },
    {
      "entropy": 5.808498239517212,
      "epoch": 0.2920899854862119,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995419260574196,
      "loss": 5.684,
      "mean_token_accuracy": 0.1507360652089119,
      "num_tokens": 6673908.0,
      "step": 3220
    },
    {
      "entropy": 5.82448444366455,
      "epoch": 0.29254354136429606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995398601103603,
      "loss": 5.7033,
      "mean_token_accuracy": 0.16093184649944306,
      "num_tokens": 6683237.0,
      "step": 3225
    },
    {
      "entropy": 5.836985397338867,
      "epoch": 0.29299709724238027,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995377895197597,
      "loss": 5.6351,
      "mean_token_accuracy": 0.1617561772465706,
      "num_tokens": 6693451.0,
      "step": 3230
    },
    {
      "entropy": 5.736857318878174,
      "epoch": 0.2934506531204644,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995357142856602,
      "loss": 5.5948,
      "mean_token_accuracy": 0.16023087352514268,
      "num_tokens": 6703479.0,
      "step": 3235
    },
    {
      "entropy": 5.751489543914795,
      "epoch": 0.29390420899854863,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499533634408105,
      "loss": 5.5968,
      "mean_token_accuracy": 0.16008537858724595,
      "num_tokens": 6715321.0,
      "step": 3240
    },
    {
      "entropy": 5.791557550430298,
      "epoch": 0.2943577648766328,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499531549887137,
      "loss": 5.6655,
      "mean_token_accuracy": 0.1540275439620018,
      "num_tokens": 6725541.0,
      "step": 3245
    },
    {
      "entropy": 5.8105682849884035,
      "epoch": 0.294811320754717,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995294607227993,
      "loss": 5.7043,
      "mean_token_accuracy": 0.16713897138834,
      "num_tokens": 6735165.0,
      "step": 3250
    },
    {
      "entropy": 5.884285402297974,
      "epoch": 0.29526487663280115,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995273669151351,
      "loss": 5.701,
      "mean_token_accuracy": 0.16163005083799362,
      "num_tokens": 6744245.0,
      "step": 3255
    },
    {
      "entropy": 5.83087010383606,
      "epoch": 0.29571843251088537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499525268464188,
      "loss": 5.6645,
      "mean_token_accuracy": 0.16589844822883607,
      "num_tokens": 6754894.0,
      "step": 3260
    },
    {
      "entropy": 5.85270414352417,
      "epoch": 0.2961719883889695,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995231653700009,
      "loss": 5.7307,
      "mean_token_accuracy": 0.15026739612221718,
      "num_tokens": 6764442.0,
      "step": 3265
    },
    {
      "entropy": 5.814293384552002,
      "epoch": 0.2966255442670537,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995210576326176,
      "loss": 5.6807,
      "mean_token_accuracy": 0.1561569720506668,
      "num_tokens": 6775054.0,
      "step": 3270
    },
    {
      "entropy": 5.85151162147522,
      "epoch": 0.2970791001451379,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995189452520816,
      "loss": 5.6647,
      "mean_token_accuracy": 0.1682630956172943,
      "num_tokens": 6784825.0,
      "step": 3275
    },
    {
      "entropy": 5.8156262874603275,
      "epoch": 0.29753265602322204,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995168282284366,
      "loss": 5.7081,
      "mean_token_accuracy": 0.16142811328172685,
      "num_tokens": 6795928.0,
      "step": 3280
    },
    {
      "entropy": 5.77383861541748,
      "epoch": 0.29798621190130625,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995147065617263,
      "loss": 5.6295,
      "mean_token_accuracy": 0.16005017310380937,
      "num_tokens": 6805456.0,
      "step": 3285
    },
    {
      "entropy": 5.866624546051026,
      "epoch": 0.2984397677793904,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995125802519946,
      "loss": 5.7057,
      "mean_token_accuracy": 0.1569285973906517,
      "num_tokens": 6816215.0,
      "step": 3290
    },
    {
      "entropy": 5.855666875839233,
      "epoch": 0.2988933236574746,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995104492992857,
      "loss": 5.7427,
      "mean_token_accuracy": 0.15623879730701445,
      "num_tokens": 6826545.0,
      "step": 3295
    },
    {
      "entropy": 5.702891540527344,
      "epoch": 0.2993468795355588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995083137036434,
      "loss": 5.6479,
      "mean_token_accuracy": 0.16189349293708802,
      "num_tokens": 6836184.0,
      "step": 3300
    },
    {
      "entropy": 5.938313961029053,
      "epoch": 0.299800435413643,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995061734651121,
      "loss": 5.7388,
      "mean_token_accuracy": 0.15550311654806137,
      "num_tokens": 6846646.0,
      "step": 3305
    },
    {
      "entropy": 5.832134151458741,
      "epoch": 0.30025399129172714,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995040285837358,
      "loss": 5.6397,
      "mean_token_accuracy": 0.16627662777900695,
      "num_tokens": 6857597.0,
      "step": 3310
    },
    {
      "entropy": 5.757911109924317,
      "epoch": 0.3007075471698113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995018790595589,
      "loss": 5.7228,
      "mean_token_accuracy": 0.15299862772226333,
      "num_tokens": 6867523.0,
      "step": 3315
    },
    {
      "entropy": 5.863043165206909,
      "epoch": 0.3011611030478955,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994997248926261,
      "loss": 5.8306,
      "mean_token_accuracy": 0.15482011437416077,
      "num_tokens": 6878190.0,
      "step": 3320
    },
    {
      "entropy": 5.831428050994873,
      "epoch": 0.30161465892597966,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994975660829817,
      "loss": 5.6124,
      "mean_token_accuracy": 0.17094137519598007,
      "num_tokens": 6888483.0,
      "step": 3325
    },
    {
      "entropy": 5.726930236816406,
      "epoch": 0.3020682148040639,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994954026306705,
      "loss": 5.611,
      "mean_token_accuracy": 0.16160572692751884,
      "num_tokens": 6898337.0,
      "step": 3330
    },
    {
      "entropy": 5.853848600387574,
      "epoch": 0.30252177068214803,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994932345357371,
      "loss": 5.6829,
      "mean_token_accuracy": 0.1558000400662422,
      "num_tokens": 6909131.0,
      "step": 3335
    },
    {
      "entropy": 5.797689771652221,
      "epoch": 0.30297532656023224,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994910617982264,
      "loss": 5.6578,
      "mean_token_accuracy": 0.15806320160627366,
      "num_tokens": 6919790.0,
      "step": 3340
    },
    {
      "entropy": 5.781959342956543,
      "epoch": 0.3034288824383164,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994888844181833,
      "loss": 5.673,
      "mean_token_accuracy": 0.16309484243392944,
      "num_tokens": 6930072.0,
      "step": 3345
    },
    {
      "entropy": 5.762999105453491,
      "epoch": 0.3038824383164006,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994867023956529,
      "loss": 5.5666,
      "mean_token_accuracy": 0.1615067183971405,
      "num_tokens": 6940646.0,
      "step": 3350
    },
    {
      "entropy": 5.712483549118042,
      "epoch": 0.30433599419448476,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994845157306803,
      "loss": 5.5689,
      "mean_token_accuracy": 0.16492611765861512,
      "num_tokens": 6951645.0,
      "step": 3355
    },
    {
      "entropy": 5.875040483474732,
      "epoch": 0.3047895500725689,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994823244233107,
      "loss": 5.7363,
      "mean_token_accuracy": 0.157509508728981,
      "num_tokens": 6963020.0,
      "step": 3360
    },
    {
      "entropy": 5.842069149017334,
      "epoch": 0.3052431059506531,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994801284735895,
      "loss": 5.6612,
      "mean_token_accuracy": 0.16126682311296464,
      "num_tokens": 6972895.0,
      "step": 3365
    },
    {
      "entropy": 5.881510400772095,
      "epoch": 0.3056966618287373,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994779278815619,
      "loss": 5.8137,
      "mean_token_accuracy": 0.14684042036533357,
      "num_tokens": 6985363.0,
      "step": 3370
    },
    {
      "entropy": 5.853707885742187,
      "epoch": 0.3061502177068215,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994757226472736,
      "loss": 5.7436,
      "mean_token_accuracy": 0.15787117779254914,
      "num_tokens": 6997040.0,
      "step": 3375
    },
    {
      "entropy": 5.848391389846801,
      "epoch": 0.30660377358490565,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994735127707702,
      "loss": 5.6733,
      "mean_token_accuracy": 0.16261067390441894,
      "num_tokens": 7007591.0,
      "step": 3380
    },
    {
      "entropy": 5.860173940658569,
      "epoch": 0.30705732946298986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994712982520972,
      "loss": 5.6754,
      "mean_token_accuracy": 0.16760361343622207,
      "num_tokens": 7017858.0,
      "step": 3385
    },
    {
      "entropy": 5.727178287506104,
      "epoch": 0.307510885341074,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994690790913008,
      "loss": 5.6346,
      "mean_token_accuracy": 0.1600530058145523,
      "num_tokens": 7027365.0,
      "step": 3390
    },
    {
      "entropy": 5.76168851852417,
      "epoch": 0.3079644412191582,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994668552884265,
      "loss": 5.6351,
      "mean_token_accuracy": 0.1617230698466301,
      "num_tokens": 7038728.0,
      "step": 3395
    },
    {
      "entropy": 5.881175231933594,
      "epoch": 0.3084179970972424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994646268435204,
      "loss": 5.5729,
      "mean_token_accuracy": 0.16643185764551163,
      "num_tokens": 7048736.0,
      "step": 3400
    },
    {
      "entropy": 5.75899338722229,
      "epoch": 0.30887155297532654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994623937566286,
      "loss": 5.7002,
      "mean_token_accuracy": 0.1575309082865715,
      "num_tokens": 7059205.0,
      "step": 3405
    },
    {
      "entropy": 5.799173402786255,
      "epoch": 0.30932510885341075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994601560277974,
      "loss": 5.6896,
      "mean_token_accuracy": 0.15942423939704894,
      "num_tokens": 7069513.0,
      "step": 3410
    },
    {
      "entropy": 5.790976524353027,
      "epoch": 0.3097786647314949,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499457913657073,
      "loss": 5.6225,
      "mean_token_accuracy": 0.15869100689888,
      "num_tokens": 7079940.0,
      "step": 3415
    },
    {
      "entropy": 5.831319952011109,
      "epoch": 0.3102322206095791,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994556666445016,
      "loss": 5.6193,
      "mean_token_accuracy": 0.16156373918056488,
      "num_tokens": 7090792.0,
      "step": 3420
    },
    {
      "entropy": 5.755025625228882,
      "epoch": 0.31068577648766327,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994534149901299,
      "loss": 5.577,
      "mean_token_accuracy": 0.16105773597955703,
      "num_tokens": 7101964.0,
      "step": 3425
    },
    {
      "entropy": 5.7506568908691404,
      "epoch": 0.3111393323657475,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994511586940044,
      "loss": 5.6442,
      "mean_token_accuracy": 0.15435996800661086,
      "num_tokens": 7113332.0,
      "step": 3430
    },
    {
      "entropy": 5.91636700630188,
      "epoch": 0.31159288824383163,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994488977561717,
      "loss": 5.7921,
      "mean_token_accuracy": 0.15468702465295792,
      "num_tokens": 7123481.0,
      "step": 3435
    },
    {
      "entropy": 5.73750057220459,
      "epoch": 0.31204644412191584,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994466321766786,
      "loss": 5.5585,
      "mean_token_accuracy": 0.1646214559674263,
      "num_tokens": 7133694.0,
      "step": 3440
    },
    {
      "entropy": 5.804030513763427,
      "epoch": 0.3125,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499444361955572,
      "loss": 5.6664,
      "mean_token_accuracy": 0.1619983956217766,
      "num_tokens": 7144491.0,
      "step": 3445
    },
    {
      "entropy": 5.945888328552246,
      "epoch": 0.31295355587808416,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994420870928988,
      "loss": 5.76,
      "mean_token_accuracy": 0.15487642139196395,
      "num_tokens": 7154716.0,
      "step": 3450
    },
    {
      "entropy": 5.8428285121917725,
      "epoch": 0.31340711175616837,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994398075887059,
      "loss": 5.7553,
      "mean_token_accuracy": 0.15003324151039124,
      "num_tokens": 7164258.0,
      "step": 3455
    },
    {
      "entropy": 5.82437686920166,
      "epoch": 0.3138606676342525,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994375234430407,
      "loss": 5.6108,
      "mean_token_accuracy": 0.15920650959014893,
      "num_tokens": 7175098.0,
      "step": 3460
    },
    {
      "entropy": 5.79690523147583,
      "epoch": 0.31431422351233673,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994352346559504,
      "loss": 5.6851,
      "mean_token_accuracy": 0.15969553738832473,
      "num_tokens": 7186480.0,
      "step": 3465
    },
    {
      "entropy": 5.710554122924805,
      "epoch": 0.3147677793904209,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994329412274821,
      "loss": 5.6593,
      "mean_token_accuracy": 0.16162224411964415,
      "num_tokens": 7197056.0,
      "step": 3470
    },
    {
      "entropy": 5.846120882034302,
      "epoch": 0.3152213352685051,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994306431576835,
      "loss": 5.7725,
      "mean_token_accuracy": 0.15470817685127258,
      "num_tokens": 7206040.0,
      "step": 3475
    },
    {
      "entropy": 5.931183242797852,
      "epoch": 0.31567489114658925,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499428340446602,
      "loss": 5.6142,
      "mean_token_accuracy": 0.16317909955978394,
      "num_tokens": 7216344.0,
      "step": 3480
    },
    {
      "entropy": 5.784151315689087,
      "epoch": 0.31612844702467346,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994260330942851,
      "loss": 5.696,
      "mean_token_accuracy": 0.15416476279497146,
      "num_tokens": 7226729.0,
      "step": 3485
    },
    {
      "entropy": 5.753327083587647,
      "epoch": 0.3165820029027576,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994237211007808,
      "loss": 5.6509,
      "mean_token_accuracy": 0.1596106082201004,
      "num_tokens": 7237616.0,
      "step": 3490
    },
    {
      "entropy": 5.826973915100098,
      "epoch": 0.3170355587808418,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994214044661367,
      "loss": 5.5765,
      "mean_token_accuracy": 0.16474129110574723,
      "num_tokens": 7247857.0,
      "step": 3495
    },
    {
      "entropy": 5.698789978027344,
      "epoch": 0.317489114658926,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994190831904007,
      "loss": 5.5503,
      "mean_token_accuracy": 0.16544652432203294,
      "num_tokens": 7257858.0,
      "step": 3500
    },
    {
      "entropy": 5.665239334106445,
      "epoch": 0.31794267053701014,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499416757273621,
      "loss": 5.5582,
      "mean_token_accuracy": 0.16463755071163177,
      "num_tokens": 7268440.0,
      "step": 3505
    },
    {
      "entropy": 5.83346905708313,
      "epoch": 0.31839622641509435,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994144267158454,
      "loss": 5.6371,
      "mean_token_accuracy": 0.16650597602128983,
      "num_tokens": 7278560.0,
      "step": 3510
    },
    {
      "entropy": 5.785221576690674,
      "epoch": 0.3188497822931785,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994120915171224,
      "loss": 5.6364,
      "mean_token_accuracy": 0.16446116715669631,
      "num_tokens": 7288422.0,
      "step": 3515
    },
    {
      "entropy": 5.840118741989135,
      "epoch": 0.3193033381712627,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994097516775002,
      "loss": 5.7106,
      "mean_token_accuracy": 0.1576638028025627,
      "num_tokens": 7298613.0,
      "step": 3520
    },
    {
      "entropy": 5.828368854522705,
      "epoch": 0.3197568940493469,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994074071970271,
      "loss": 5.7422,
      "mean_token_accuracy": 0.15291873514652252,
      "num_tokens": 7309509.0,
      "step": 3525
    },
    {
      "entropy": 5.860109329223633,
      "epoch": 0.3202104499274311,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994050580757515,
      "loss": 5.6445,
      "mean_token_accuracy": 0.15940439999103545,
      "num_tokens": 7319015.0,
      "step": 3530
    },
    {
      "entropy": 5.721933031082154,
      "epoch": 0.32066400580551524,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994027043137223,
      "loss": 5.61,
      "mean_token_accuracy": 0.1672150120139122,
      "num_tokens": 7329026.0,
      "step": 3535
    },
    {
      "entropy": 5.855372905731201,
      "epoch": 0.3211175616835994,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994003459109879,
      "loss": 5.7137,
      "mean_token_accuracy": 0.16032420545816423,
      "num_tokens": 7339638.0,
      "step": 3540
    },
    {
      "entropy": 5.837479209899902,
      "epoch": 0.3215711175616836,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993979828675972,
      "loss": 5.624,
      "mean_token_accuracy": 0.16279072165489197,
      "num_tokens": 7349979.0,
      "step": 3545
    },
    {
      "entropy": 5.702495336532593,
      "epoch": 0.32202467343976776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499395615183599,
      "loss": 5.5517,
      "mean_token_accuracy": 0.17009643316268921,
      "num_tokens": 7359398.0,
      "step": 3550
    },
    {
      "entropy": 5.7116045475006105,
      "epoch": 0.32247822931785197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993932428590422,
      "loss": 5.5945,
      "mean_token_accuracy": 0.15939887911081313,
      "num_tokens": 7370418.0,
      "step": 3555
    },
    {
      "entropy": 5.781202697753907,
      "epoch": 0.3229317851959361,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993908658939762,
      "loss": 5.6239,
      "mean_token_accuracy": 0.15674008578062057,
      "num_tokens": 7380461.0,
      "step": 3560
    },
    {
      "entropy": 5.814039325714111,
      "epoch": 0.32338534107402034,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993884842884496,
      "loss": 5.647,
      "mean_token_accuracy": 0.16048161089420318,
      "num_tokens": 7390646.0,
      "step": 3565
    },
    {
      "entropy": 5.786873054504395,
      "epoch": 0.3238388969521045,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993860980425121,
      "loss": 5.7203,
      "mean_token_accuracy": 0.1618356540799141,
      "num_tokens": 7402106.0,
      "step": 3570
    },
    {
      "entropy": 5.809912443161011,
      "epoch": 0.3242924528301887,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993837071562131,
      "loss": 5.7043,
      "mean_token_accuracy": 0.16131191849708557,
      "num_tokens": 7412936.0,
      "step": 3575
    },
    {
      "entropy": 5.8115393161773685,
      "epoch": 0.32474600870827286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993813116296017,
      "loss": 5.5708,
      "mean_token_accuracy": 0.16675995141267777,
      "num_tokens": 7422303.0,
      "step": 3580
    },
    {
      "entropy": 5.715788507461548,
      "epoch": 0.325199564586357,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993789114627275,
      "loss": 5.6511,
      "mean_token_accuracy": 0.16085287630558015,
      "num_tokens": 7432829.0,
      "step": 3585
    },
    {
      "entropy": 5.937649488449097,
      "epoch": 0.3256531204644412,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993765066556403,
      "loss": 5.6713,
      "mean_token_accuracy": 0.15823337584733962,
      "num_tokens": 7443218.0,
      "step": 3590
    },
    {
      "entropy": 5.773650455474853,
      "epoch": 0.3261066763425254,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993740972083899,
      "loss": 5.62,
      "mean_token_accuracy": 0.16021817028522492,
      "num_tokens": 7454531.0,
      "step": 3595
    },
    {
      "entropy": 5.790258073806763,
      "epoch": 0.3265602322206096,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993716831210259,
      "loss": 5.6799,
      "mean_token_accuracy": 0.16317007392644883,
      "num_tokens": 7464993.0,
      "step": 3600
    },
    {
      "entropy": 5.729147958755493,
      "epoch": 0.32701378809869375,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993692643935983,
      "loss": 5.7096,
      "mean_token_accuracy": 0.1584396131336689,
      "num_tokens": 7475755.0,
      "step": 3605
    },
    {
      "entropy": 5.771151685714722,
      "epoch": 0.32746734397677796,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993668410261571,
      "loss": 5.6945,
      "mean_token_accuracy": 0.16145843863487244,
      "num_tokens": 7486778.0,
      "step": 3610
    },
    {
      "entropy": 5.907090902328491,
      "epoch": 0.3279208998548621,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993644130187525,
      "loss": 5.6771,
      "mean_token_accuracy": 0.16191897690296173,
      "num_tokens": 7497439.0,
      "step": 3615
    },
    {
      "entropy": 5.734846353530884,
      "epoch": 0.3283744557329463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993619803714347,
      "loss": 5.7075,
      "mean_token_accuracy": 0.15616535544395446,
      "num_tokens": 7508274.0,
      "step": 3620
    },
    {
      "entropy": 5.782626390457153,
      "epoch": 0.3288280116110305,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993595430842539,
      "loss": 5.6874,
      "mean_token_accuracy": 0.16056278795003892,
      "num_tokens": 7519315.0,
      "step": 3625
    },
    {
      "entropy": 5.901585865020752,
      "epoch": 0.32928156748911463,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993571011572606,
      "loss": 5.6808,
      "mean_token_accuracy": 0.1630665123462677,
      "num_tokens": 7529910.0,
      "step": 3630
    },
    {
      "entropy": 5.800387811660767,
      "epoch": 0.32973512336719885,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993546545905053,
      "loss": 5.602,
      "mean_token_accuracy": 0.16489686965942382,
      "num_tokens": 7538966.0,
      "step": 3635
    },
    {
      "entropy": 5.775709533691407,
      "epoch": 0.330188679245283,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993522033840386,
      "loss": 5.6374,
      "mean_token_accuracy": 0.15789183378219604,
      "num_tokens": 7549683.0,
      "step": 3640
    },
    {
      "entropy": 5.700624227523804,
      "epoch": 0.3306422351233672,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499349747537911,
      "loss": 5.6191,
      "mean_token_accuracy": 0.1624668598175049,
      "num_tokens": 7560629.0,
      "step": 3645
    },
    {
      "entropy": 5.806013774871826,
      "epoch": 0.33109579100145137,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993472870521736,
      "loss": 5.5646,
      "mean_token_accuracy": 0.17270171344280244,
      "num_tokens": 7570613.0,
      "step": 3650
    },
    {
      "entropy": 5.79381422996521,
      "epoch": 0.3315493468795356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993448219268771,
      "loss": 5.5499,
      "mean_token_accuracy": 0.16173658668994903,
      "num_tokens": 7580640.0,
      "step": 3655
    },
    {
      "entropy": 5.763698101043701,
      "epoch": 0.33200290275761973,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993423521620726,
      "loss": 5.6381,
      "mean_token_accuracy": 0.16688979268074036,
      "num_tokens": 7590164.0,
      "step": 3660
    },
    {
      "entropy": 5.823714399337769,
      "epoch": 0.33245645863570394,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993398777578109,
      "loss": 5.6959,
      "mean_token_accuracy": 0.16008617281913756,
      "num_tokens": 7600173.0,
      "step": 3665
    },
    {
      "entropy": 5.71161642074585,
      "epoch": 0.3329100145137881,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993373987141435,
      "loss": 5.4841,
      "mean_token_accuracy": 0.17095217257738113,
      "num_tokens": 7609958.0,
      "step": 3670
    },
    {
      "entropy": 5.813071250915527,
      "epoch": 0.33336357039187225,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993349150311214,
      "loss": 5.6411,
      "mean_token_accuracy": 0.157687945663929,
      "num_tokens": 7620237.0,
      "step": 3675
    },
    {
      "entropy": 5.699023199081421,
      "epoch": 0.33381712626995647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993324267087962,
      "loss": 5.5406,
      "mean_token_accuracy": 0.1682781383395195,
      "num_tokens": 7629886.0,
      "step": 3680
    },
    {
      "entropy": 5.728935241699219,
      "epoch": 0.3342706821480406,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993299337472192,
      "loss": 5.6095,
      "mean_token_accuracy": 0.16654768735170364,
      "num_tokens": 7640057.0,
      "step": 3685
    },
    {
      "entropy": 5.7182149410247805,
      "epoch": 0.33472423802612483,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499327436146442,
      "loss": 5.5532,
      "mean_token_accuracy": 0.16529571264982224,
      "num_tokens": 7649766.0,
      "step": 3690
    },
    {
      "entropy": 5.769032526016235,
      "epoch": 0.335177793904209,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993249339065162,
      "loss": 5.6399,
      "mean_token_accuracy": 0.16574999690055847,
      "num_tokens": 7660890.0,
      "step": 3695
    },
    {
      "entropy": 5.773209142684936,
      "epoch": 0.3356313497822932,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993224270274937,
      "loss": 5.6582,
      "mean_token_accuracy": 0.1607873558998108,
      "num_tokens": 7670737.0,
      "step": 3700
    },
    {
      "entropy": 5.7178552627563475,
      "epoch": 0.33608490566037735,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993199155094263,
      "loss": 5.6067,
      "mean_token_accuracy": 0.16417943835258483,
      "num_tokens": 7681407.0,
      "step": 3705
    },
    {
      "entropy": 5.768005990982056,
      "epoch": 0.33653846153846156,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993173993523658,
      "loss": 5.5356,
      "mean_token_accuracy": 0.17343078404664994,
      "num_tokens": 7691758.0,
      "step": 3710
    },
    {
      "entropy": 5.699445009231567,
      "epoch": 0.3369920174165457,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993148785563643,
      "loss": 5.5679,
      "mean_token_accuracy": 0.16385959088802338,
      "num_tokens": 7702934.0,
      "step": 3715
    },
    {
      "entropy": 5.71638708114624,
      "epoch": 0.3374455732946299,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499312353121474,
      "loss": 5.5705,
      "mean_token_accuracy": 0.16445922404527663,
      "num_tokens": 7713001.0,
      "step": 3720
    },
    {
      "entropy": 5.802838611602783,
      "epoch": 0.3378991291727141,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993098230477471,
      "loss": 5.6211,
      "mean_token_accuracy": 0.1572564959526062,
      "num_tokens": 7724005.0,
      "step": 3725
    },
    {
      "entropy": 5.785116243362427,
      "epoch": 0.33835268505079824,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993072883352358,
      "loss": 5.5802,
      "mean_token_accuracy": 0.16274530738592147,
      "num_tokens": 7733743.0,
      "step": 3730
    },
    {
      "entropy": 5.802304887771607,
      "epoch": 0.33880624092888245,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993047489839927,
      "loss": 5.7695,
      "mean_token_accuracy": 0.1579284265637398,
      "num_tokens": 7743762.0,
      "step": 3735
    },
    {
      "entropy": 5.850232982635498,
      "epoch": 0.3392597968069666,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993022049940702,
      "loss": 5.655,
      "mean_token_accuracy": 0.16184027045965194,
      "num_tokens": 7753758.0,
      "step": 3740
    },
    {
      "entropy": 5.739925861358643,
      "epoch": 0.3397133526850508,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992996563655209,
      "loss": 5.6052,
      "mean_token_accuracy": 0.163077212870121,
      "num_tokens": 7764692.0,
      "step": 3745
    },
    {
      "entropy": 5.742031097412109,
      "epoch": 0.340166908563135,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992971030983975,
      "loss": 5.644,
      "mean_token_accuracy": 0.15631235092878343,
      "num_tokens": 7776892.0,
      "step": 3750
    },
    {
      "entropy": 5.875276947021485,
      "epoch": 0.3406204644412192,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992945451927531,
      "loss": 5.6769,
      "mean_token_accuracy": 0.16058121621608734,
      "num_tokens": 7787143.0,
      "step": 3755
    },
    {
      "entropy": 5.723575305938721,
      "epoch": 0.34107402031930334,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992919826486402,
      "loss": 5.5967,
      "mean_token_accuracy": 0.16272515058517456,
      "num_tokens": 7798537.0,
      "step": 3760
    },
    {
      "entropy": 5.713321161270142,
      "epoch": 0.3415275761973875,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992894154661118,
      "loss": 5.5709,
      "mean_token_accuracy": 0.16793359965085983,
      "num_tokens": 7808434.0,
      "step": 3765
    },
    {
      "entropy": 5.79578595161438,
      "epoch": 0.3419811320754717,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992868436452213,
      "loss": 5.6818,
      "mean_token_accuracy": 0.16078411191701888,
      "num_tokens": 7819592.0,
      "step": 3770
    },
    {
      "entropy": 5.809304237365723,
      "epoch": 0.34243468795355586,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992842671860217,
      "loss": 5.5643,
      "mean_token_accuracy": 0.16715734750032424,
      "num_tokens": 7830602.0,
      "step": 3775
    },
    {
      "entropy": 5.786276483535767,
      "epoch": 0.34288824383164007,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992816860885662,
      "loss": 5.7212,
      "mean_token_accuracy": 0.15924838036298752,
      "num_tokens": 7840850.0,
      "step": 3780
    },
    {
      "entropy": 5.884521865844727,
      "epoch": 0.3433417997097242,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992791003529084,
      "loss": 5.6558,
      "mean_token_accuracy": 0.15685309320688248,
      "num_tokens": 7852371.0,
      "step": 3785
    },
    {
      "entropy": 5.715097904205322,
      "epoch": 0.34379535558780844,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992765099791017,
      "loss": 5.6818,
      "mean_token_accuracy": 0.15666263401508332,
      "num_tokens": 7863290.0,
      "step": 3790
    },
    {
      "entropy": 5.71993088722229,
      "epoch": 0.3442489114658926,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992739149671995,
      "loss": 5.509,
      "mean_token_accuracy": 0.16685899645090102,
      "num_tokens": 7873927.0,
      "step": 3795
    },
    {
      "entropy": 5.733199405670166,
      "epoch": 0.3447024673439768,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992713153172556,
      "loss": 5.5502,
      "mean_token_accuracy": 0.16382715702056885,
      "num_tokens": 7884977.0,
      "step": 3800
    },
    {
      "entropy": 5.699240398406983,
      "epoch": 0.34515602322206096,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992687110293238,
      "loss": 5.6304,
      "mean_token_accuracy": 0.16769205555319786,
      "num_tokens": 7894620.0,
      "step": 3805
    },
    {
      "entropy": 5.719654321670532,
      "epoch": 0.3456095791001451,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992661021034579,
      "loss": 5.5507,
      "mean_token_accuracy": 0.16906091570854187,
      "num_tokens": 7904541.0,
      "step": 3810
    },
    {
      "entropy": 5.726424789428711,
      "epoch": 0.3460631349782293,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992634885397118,
      "loss": 5.5794,
      "mean_token_accuracy": 0.1659620225429535,
      "num_tokens": 7915435.0,
      "step": 3815
    },
    {
      "entropy": 5.722615718841553,
      "epoch": 0.3465166908563135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992608703381396,
      "loss": 5.5616,
      "mean_token_accuracy": 0.16741435825824738,
      "num_tokens": 7925870.0,
      "step": 3820
    },
    {
      "entropy": 5.722995281219482,
      "epoch": 0.3469702467343977,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992582474987955,
      "loss": 5.5639,
      "mean_token_accuracy": 0.16274259686470033,
      "num_tokens": 7937046.0,
      "step": 3825
    },
    {
      "entropy": 5.75324854850769,
      "epoch": 0.34742380261248185,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992556200217337,
      "loss": 5.5261,
      "mean_token_accuracy": 0.1641123980283737,
      "num_tokens": 7947150.0,
      "step": 3830
    },
    {
      "entropy": 5.795271539688111,
      "epoch": 0.34787735849056606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992529879070085,
      "loss": 5.618,
      "mean_token_accuracy": 0.16771930009126662,
      "num_tokens": 7956815.0,
      "step": 3835
    },
    {
      "entropy": 5.789908409118652,
      "epoch": 0.3483309143686502,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992503511546745,
      "loss": 5.6657,
      "mean_token_accuracy": 0.15945106595754624,
      "num_tokens": 7968114.0,
      "step": 3840
    },
    {
      "entropy": 5.748790836334228,
      "epoch": 0.3487844702467344,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992477097647859,
      "loss": 5.5656,
      "mean_token_accuracy": 0.1627722293138504,
      "num_tokens": 7977840.0,
      "step": 3845
    },
    {
      "entropy": 5.814347887039185,
      "epoch": 0.3492380261248186,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992450637373976,
      "loss": 5.6907,
      "mean_token_accuracy": 0.16484753638505936,
      "num_tokens": 7988663.0,
      "step": 3850
    },
    {
      "entropy": 5.681796550750732,
      "epoch": 0.34969158200290273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992424130725642,
      "loss": 5.4999,
      "mean_token_accuracy": 0.16489350348711013,
      "num_tokens": 7999693.0,
      "step": 3855
    },
    {
      "entropy": 5.772833013534546,
      "epoch": 0.35014513788098695,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992397577703406,
      "loss": 5.6747,
      "mean_token_accuracy": 0.1532568469643593,
      "num_tokens": 8010154.0,
      "step": 3860
    },
    {
      "entropy": 5.825438737869263,
      "epoch": 0.3505986937590711,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992370978307816,
      "loss": 5.6536,
      "mean_token_accuracy": 0.1622571349143982,
      "num_tokens": 8020921.0,
      "step": 3865
    },
    {
      "entropy": 5.764519309997558,
      "epoch": 0.3510522496371553,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992344332539424,
      "loss": 5.5811,
      "mean_token_accuracy": 0.16872124671936034,
      "num_tokens": 8031404.0,
      "step": 3870
    },
    {
      "entropy": 5.684741640090943,
      "epoch": 0.35150580551523947,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992317640398779,
      "loss": 5.585,
      "mean_token_accuracy": 0.1686838299036026,
      "num_tokens": 8041310.0,
      "step": 3875
    },
    {
      "entropy": 5.710890054702759,
      "epoch": 0.3519593613933237,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992290901886434,
      "loss": 5.57,
      "mean_token_accuracy": 0.1684218779206276,
      "num_tokens": 8051054.0,
      "step": 3880
    },
    {
      "entropy": 5.744958877563477,
      "epoch": 0.35241291727140783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499226411700294,
      "loss": 5.5578,
      "mean_token_accuracy": 0.16676690876483918,
      "num_tokens": 8061063.0,
      "step": 3885
    },
    {
      "entropy": 5.69763445854187,
      "epoch": 0.35286647314949204,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992237285748855,
      "loss": 5.5841,
      "mean_token_accuracy": 0.16478407829999925,
      "num_tokens": 8071715.0,
      "step": 3890
    },
    {
      "entropy": 5.80994873046875,
      "epoch": 0.3533200290275762,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992210408124731,
      "loss": 5.6496,
      "mean_token_accuracy": 0.15725049376487732,
      "num_tokens": 8082679.0,
      "step": 3895
    },
    {
      "entropy": 5.752315807342529,
      "epoch": 0.35377358490566035,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992183484131123,
      "loss": 5.6808,
      "mean_token_accuracy": 0.16136541217565536,
      "num_tokens": 8093785.0,
      "step": 3900
    },
    {
      "entropy": 5.71539158821106,
      "epoch": 0.35422714078374457,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992156513768591,
      "loss": 5.5685,
      "mean_token_accuracy": 0.16779497861862183,
      "num_tokens": 8105822.0,
      "step": 3905
    },
    {
      "entropy": 5.709498691558838,
      "epoch": 0.3546806966618287,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499212949703769,
      "loss": 5.486,
      "mean_token_accuracy": 0.1679369956254959,
      "num_tokens": 8115521.0,
      "step": 3910
    },
    {
      "entropy": 5.726305246353149,
      "epoch": 0.35513425253991293,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992102433938981,
      "loss": 5.5783,
      "mean_token_accuracy": 0.16780869513750077,
      "num_tokens": 8125524.0,
      "step": 3915
    },
    {
      "entropy": 5.712926054000855,
      "epoch": 0.3555878084179971,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992075324473021,
      "loss": 5.482,
      "mean_token_accuracy": 0.17744722813367844,
      "num_tokens": 8135552.0,
      "step": 3920
    },
    {
      "entropy": 5.780481147766113,
      "epoch": 0.3560413642960813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992048168640371,
      "loss": 5.6135,
      "mean_token_accuracy": 0.16508189737796783,
      "num_tokens": 8145617.0,
      "step": 3925
    },
    {
      "entropy": 5.725088024139405,
      "epoch": 0.35649492017416545,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992020966441595,
      "loss": 5.6251,
      "mean_token_accuracy": 0.16034330129623414,
      "num_tokens": 8156729.0,
      "step": 3930
    },
    {
      "entropy": 5.764310026168824,
      "epoch": 0.35694847605224966,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991993717877254,
      "loss": 5.6076,
      "mean_token_accuracy": 0.16174564361572266,
      "num_tokens": 8167637.0,
      "step": 3935
    },
    {
      "entropy": 5.720431089401245,
      "epoch": 0.3574020319303338,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991966422947911,
      "loss": 5.5852,
      "mean_token_accuracy": 0.1702892929315567,
      "num_tokens": 8177354.0,
      "step": 3940
    },
    {
      "entropy": 5.718207550048828,
      "epoch": 0.357855587808418,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991939081654132,
      "loss": 5.5502,
      "mean_token_accuracy": 0.1746809646487236,
      "num_tokens": 8187580.0,
      "step": 3945
    },
    {
      "entropy": 5.708486032485962,
      "epoch": 0.3583091436865022,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499191169399648,
      "loss": 5.6497,
      "mean_token_accuracy": 0.1604743108153343,
      "num_tokens": 8199382.0,
      "step": 3950
    },
    {
      "entropy": 5.712984323501587,
      "epoch": 0.35876269956458634,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991884259975525,
      "loss": 5.5404,
      "mean_token_accuracy": 0.16475948542356492,
      "num_tokens": 8208484.0,
      "step": 3955
    },
    {
      "entropy": 5.727371263504028,
      "epoch": 0.35921625544267055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499185677959183,
      "loss": 5.5838,
      "mean_token_accuracy": 0.16461657136678695,
      "num_tokens": 8219056.0,
      "step": 3960
    },
    {
      "entropy": 5.731614685058593,
      "epoch": 0.3596698113207547,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991829252845966,
      "loss": 5.6047,
      "mean_token_accuracy": 0.16140197962522507,
      "num_tokens": 8230149.0,
      "step": 3965
    },
    {
      "entropy": 5.7445838928222654,
      "epoch": 0.3601233671988389,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991801679738502,
      "loss": 5.6652,
      "mean_token_accuracy": 0.15284715443849564,
      "num_tokens": 8240803.0,
      "step": 3970
    },
    {
      "entropy": 5.766950941085815,
      "epoch": 0.3605769230769231,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991774060270008,
      "loss": 5.6201,
      "mean_token_accuracy": 0.15810701847076417,
      "num_tokens": 8251563.0,
      "step": 3975
    },
    {
      "entropy": 5.701673126220703,
      "epoch": 0.3610304789550073,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991746394441055,
      "loss": 5.5522,
      "mean_token_accuracy": 0.16369420439004898,
      "num_tokens": 8261342.0,
      "step": 3980
    },
    {
      "entropy": 5.753665637969971,
      "epoch": 0.36148403483309144,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991718682252216,
      "loss": 5.6681,
      "mean_token_accuracy": 0.15975360572338104,
      "num_tokens": 8271540.0,
      "step": 3985
    },
    {
      "entropy": 5.782139301300049,
      "epoch": 0.3619375907111756,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991690923704063,
      "loss": 5.58,
      "mean_token_accuracy": 0.16486785262823106,
      "num_tokens": 8280593.0,
      "step": 3990
    },
    {
      "entropy": 5.742940139770508,
      "epoch": 0.3623911465892598,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991663118797171,
      "loss": 5.6104,
      "mean_token_accuracy": 0.16217703521251678,
      "num_tokens": 8290126.0,
      "step": 3995
    },
    {
      "entropy": 5.747898292541504,
      "epoch": 0.36284470246734396,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991635267532113,
      "loss": 5.5394,
      "mean_token_accuracy": 0.16359646618366241,
      "num_tokens": 8301416.0,
      "step": 4000
    },
    {
      "entropy": 5.775230884552002,
      "epoch": 0.36329825834542817,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991607369909468,
      "loss": 5.6135,
      "mean_token_accuracy": 0.1605329006910324,
      "num_tokens": 8311034.0,
      "step": 4005
    },
    {
      "entropy": 5.757994508743286,
      "epoch": 0.3637518142235123,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991579425929811,
      "loss": 5.6266,
      "mean_token_accuracy": 0.16281366795301438,
      "num_tokens": 8322203.0,
      "step": 4010
    },
    {
      "entropy": 5.7276918411254885,
      "epoch": 0.36420537010159654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499155143559372,
      "loss": 5.5117,
      "mean_token_accuracy": 0.16838518381118775,
      "num_tokens": 8330981.0,
      "step": 4015
    },
    {
      "entropy": 5.655632972717285,
      "epoch": 0.3646589259796807,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991523398901773,
      "loss": 5.5437,
      "mean_token_accuracy": 0.16577242016792298,
      "num_tokens": 8341294.0,
      "step": 4020
    },
    {
      "entropy": 5.8728711128234865,
      "epoch": 0.3651124818577649,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991495315854553,
      "loss": 5.6921,
      "mean_token_accuracy": 0.15758731961250305,
      "num_tokens": 8351824.0,
      "step": 4025
    },
    {
      "entropy": 5.734429454803466,
      "epoch": 0.36556603773584906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991467186452637,
      "loss": 5.5595,
      "mean_token_accuracy": 0.16328836232423782,
      "num_tokens": 8361975.0,
      "step": 4030
    },
    {
      "entropy": 5.646515369415283,
      "epoch": 0.3660195936139332,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991439010696609,
      "loss": 5.5053,
      "mean_token_accuracy": 0.16924880295991898,
      "num_tokens": 8372145.0,
      "step": 4035
    },
    {
      "entropy": 5.743934106826782,
      "epoch": 0.3664731494920174,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991410788587051,
      "loss": 5.6034,
      "mean_token_accuracy": 0.1616530880331993,
      "num_tokens": 8382542.0,
      "step": 4040
    },
    {
      "entropy": 5.771177625656128,
      "epoch": 0.3669267053701016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991382520124547,
      "loss": 5.5942,
      "mean_token_accuracy": 0.16173005402088164,
      "num_tokens": 8393450.0,
      "step": 4045
    },
    {
      "entropy": 5.7502647876739506,
      "epoch": 0.3673802612481858,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991354205309681,
      "loss": 5.5407,
      "mean_token_accuracy": 0.1637318655848503,
      "num_tokens": 8404929.0,
      "step": 4050
    },
    {
      "entropy": 5.762520360946655,
      "epoch": 0.36783381712626995,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499132584414304,
      "loss": 5.5531,
      "mean_token_accuracy": 0.1654103085398674,
      "num_tokens": 8415703.0,
      "step": 4055
    },
    {
      "entropy": 5.728594827651977,
      "epoch": 0.36828737300435416,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991297436625207,
      "loss": 5.6347,
      "mean_token_accuracy": 0.16382609456777572,
      "num_tokens": 8426220.0,
      "step": 4060
    },
    {
      "entropy": 5.665254354476929,
      "epoch": 0.3687409288824383,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991268982756774,
      "loss": 5.5135,
      "mean_token_accuracy": 0.16983363032341003,
      "num_tokens": 8436220.0,
      "step": 4065
    },
    {
      "entropy": 5.824220180511475,
      "epoch": 0.3691944847605225,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991240482538327,
      "loss": 5.6576,
      "mean_token_accuracy": 0.1528759405016899,
      "num_tokens": 8446724.0,
      "step": 4070
    },
    {
      "entropy": 5.739574432373047,
      "epoch": 0.3696480406386067,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991211935970455,
      "loss": 5.5215,
      "mean_token_accuracy": 0.16571784317493438,
      "num_tokens": 8457270.0,
      "step": 4075
    },
    {
      "entropy": 5.698843717575073,
      "epoch": 0.37010159651669083,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991183343053748,
      "loss": 5.5808,
      "mean_token_accuracy": 0.16122987121343613,
      "num_tokens": 8468137.0,
      "step": 4080
    },
    {
      "entropy": 5.737043762207032,
      "epoch": 0.37055515239477504,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049911547037888,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16846096962690355,
      "num_tokens": 8477574.0,
      "step": 4085
    },
    {
      "entropy": 5.718803071975708,
      "epoch": 0.3710087082728592,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991126018176202,
      "loss": 5.6211,
      "mean_token_accuracy": 0.1616816744208336,
      "num_tokens": 8487268.0,
      "step": 4090
    },
    {
      "entropy": 5.8145722389221195,
      "epoch": 0.3714622641509434,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991097286216545,
      "loss": 5.6662,
      "mean_token_accuracy": 0.168794247508049,
      "num_tokens": 8497785.0,
      "step": 4095
    },
    {
      "entropy": 5.753116083145142,
      "epoch": 0.37191582002902757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991068507910427,
      "loss": 5.5533,
      "mean_token_accuracy": 0.16418525129556655,
      "num_tokens": 8508169.0,
      "step": 4100
    },
    {
      "entropy": 5.690785503387451,
      "epoch": 0.3723693759071118,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499103968325844,
      "loss": 5.522,
      "mean_token_accuracy": 0.16396052837371827,
      "num_tokens": 8519201.0,
      "step": 4105
    },
    {
      "entropy": 5.7371025562286375,
      "epoch": 0.37282293178519593,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991010812261181,
      "loss": 5.6384,
      "mean_token_accuracy": 0.1556764304637909,
      "num_tokens": 8531093.0,
      "step": 4110
    },
    {
      "entropy": 5.665953826904297,
      "epoch": 0.37327648766328014,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990981894919248,
      "loss": 5.4913,
      "mean_token_accuracy": 0.16958004534244536,
      "num_tokens": 8541730.0,
      "step": 4115
    },
    {
      "entropy": 5.693306016921997,
      "epoch": 0.3737300435413643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990952931233238,
      "loss": 5.5646,
      "mean_token_accuracy": 0.16480782628059387,
      "num_tokens": 8552570.0,
      "step": 4120
    },
    {
      "entropy": 5.754336500167847,
      "epoch": 0.37418359941944845,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499092392120375,
      "loss": 5.4628,
      "mean_token_accuracy": 0.17125755399465561,
      "num_tokens": 8562340.0,
      "step": 4125
    },
    {
      "entropy": 5.735305070877075,
      "epoch": 0.37463715529753266,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990894864831385,
      "loss": 5.6707,
      "mean_token_accuracy": 0.15796474665403365,
      "num_tokens": 8573240.0,
      "step": 4130
    },
    {
      "entropy": 5.769050550460816,
      "epoch": 0.3750907111756168,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990865762116743,
      "loss": 5.5299,
      "mean_token_accuracy": 0.16909209936857222,
      "num_tokens": 8583194.0,
      "step": 4135
    },
    {
      "entropy": 5.719068336486816,
      "epoch": 0.37554426705370103,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990836613060426,
      "loss": 5.5962,
      "mean_token_accuracy": 0.1610783204436302,
      "num_tokens": 8593603.0,
      "step": 4140
    },
    {
      "entropy": 5.664818382263183,
      "epoch": 0.3759978229317852,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990807417663036,
      "loss": 5.5998,
      "mean_token_accuracy": 0.16227242201566697,
      "num_tokens": 8603176.0,
      "step": 4145
    },
    {
      "entropy": 5.7971210956573485,
      "epoch": 0.3764513788098694,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004990778175925179,
      "loss": 5.6154,
      "mean_token_accuracy": 0.16660853177309037,
      "num_tokens": 8614801.0,
      "step": 4150
    },
    {
      "entropy": 5.768546104431152,
      "epoch": 0.37690493468795355,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990748887847457,
      "loss": 5.6418,
      "mean_token_accuracy": 0.16194847822189332,
      "num_tokens": 8624999.0,
      "step": 4155
    },
    {
      "entropy": 5.667234706878662,
      "epoch": 0.37735849056603776,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990719553430477,
      "loss": 5.4948,
      "mean_token_accuracy": 0.17510903775691986,
      "num_tokens": 8635358.0,
      "step": 4160
    },
    {
      "entropy": 5.749764490127563,
      "epoch": 0.3778120464441219,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990690172674847,
      "loss": 5.5679,
      "mean_token_accuracy": 0.1612323611974716,
      "num_tokens": 8645430.0,
      "step": 4165
    },
    {
      "entropy": 5.726269006729126,
      "epoch": 0.3782656023222061,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990660745581172,
      "loss": 5.5234,
      "mean_token_accuracy": 0.17513987571001052,
      "num_tokens": 8655115.0,
      "step": 4170
    },
    {
      "entropy": 5.6731971263885494,
      "epoch": 0.3787191582002903,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990631272150062,
      "loss": 5.5279,
      "mean_token_accuracy": 0.17234957814216614,
      "num_tokens": 8665391.0,
      "step": 4175
    },
    {
      "entropy": 5.764857435226441,
      "epoch": 0.37917271407837444,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004990601752382127,
      "loss": 5.6196,
      "mean_token_accuracy": 0.1626276418566704,
      "num_tokens": 8676819.0,
      "step": 4180
    },
    {
      "entropy": 5.78800048828125,
      "epoch": 0.37962626995645865,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990572186277975,
      "loss": 5.5433,
      "mean_token_accuracy": 0.16920216530561447,
      "num_tokens": 8687052.0,
      "step": 4185
    },
    {
      "entropy": 5.716868305206299,
      "epoch": 0.3800798258345428,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990542573838221,
      "loss": 5.6207,
      "mean_token_accuracy": 0.16125022917985915,
      "num_tokens": 8697756.0,
      "step": 4190
    },
    {
      "entropy": 5.616197013854981,
      "epoch": 0.380533381712627,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990512915063475,
      "loss": 5.5242,
      "mean_token_accuracy": 0.17394591718912125,
      "num_tokens": 8708388.0,
      "step": 4195
    },
    {
      "entropy": 5.806944036483765,
      "epoch": 0.38098693759071117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990483209954352,
      "loss": 5.6288,
      "mean_token_accuracy": 0.15857603251934052,
      "num_tokens": 8719475.0,
      "step": 4200
    },
    {
      "entropy": 5.71453013420105,
      "epoch": 0.3814404934687954,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990453458511464,
      "loss": 5.477,
      "mean_token_accuracy": 0.17523704022169112,
      "num_tokens": 8729395.0,
      "step": 4205
    },
    {
      "entropy": 5.69320387840271,
      "epoch": 0.38189404934687954,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990423660735428,
      "loss": 5.5367,
      "mean_token_accuracy": 0.17120707631111146,
      "num_tokens": 8740351.0,
      "step": 4210
    },
    {
      "entropy": 5.747463369369507,
      "epoch": 0.3823476052249637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499039381662686,
      "loss": 5.5519,
      "mean_token_accuracy": 0.16874152719974517,
      "num_tokens": 8749260.0,
      "step": 4215
    },
    {
      "entropy": 5.784005641937256,
      "epoch": 0.3828011611030479,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990363926186377,
      "loss": 5.6472,
      "mean_token_accuracy": 0.1575682356953621,
      "num_tokens": 8759604.0,
      "step": 4220
    },
    {
      "entropy": 5.778999853134155,
      "epoch": 0.38325471698113206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990333989414597,
      "loss": 5.6582,
      "mean_token_accuracy": 0.1573771618306637,
      "num_tokens": 8769815.0,
      "step": 4225
    },
    {
      "entropy": 5.677911949157715,
      "epoch": 0.38370827285921627,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990304006312138,
      "loss": 5.5816,
      "mean_token_accuracy": 0.16669487804174424,
      "num_tokens": 8780557.0,
      "step": 4230
    },
    {
      "entropy": 5.731473064422607,
      "epoch": 0.3841618287373004,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990273976879622,
      "loss": 5.5429,
      "mean_token_accuracy": 0.17356031835079194,
      "num_tokens": 8790993.0,
      "step": 4235
    },
    {
      "entropy": 5.731623315811158,
      "epoch": 0.38461538461538464,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990243901117669,
      "loss": 5.6024,
      "mean_token_accuracy": 0.15951776504516602,
      "num_tokens": 8800657.0,
      "step": 4240
    },
    {
      "entropy": 5.657181215286255,
      "epoch": 0.3850689404934688,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990213779026903,
      "loss": 5.5996,
      "mean_token_accuracy": 0.1628565177321434,
      "num_tokens": 8812983.0,
      "step": 4245
    },
    {
      "entropy": 5.771100807189941,
      "epoch": 0.385522496371553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990183610607944,
      "loss": 5.571,
      "mean_token_accuracy": 0.1698968768119812,
      "num_tokens": 8822980.0,
      "step": 4250
    },
    {
      "entropy": 5.677217197418213,
      "epoch": 0.38597605224963716,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990153395861415,
      "loss": 5.5484,
      "mean_token_accuracy": 0.16861167252063752,
      "num_tokens": 8833821.0,
      "step": 4255
    },
    {
      "entropy": 5.625637340545654,
      "epoch": 0.3864296081277213,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990123134787946,
      "loss": 5.4898,
      "mean_token_accuracy": 0.17129137367010117,
      "num_tokens": 8843929.0,
      "step": 4260
    },
    {
      "entropy": 5.754845666885376,
      "epoch": 0.3868831640058055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990092827388159,
      "loss": 5.4759,
      "mean_token_accuracy": 0.16929103881120683,
      "num_tokens": 8854180.0,
      "step": 4265
    },
    {
      "entropy": 5.630288219451904,
      "epoch": 0.3873367198838897,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990062473662681,
      "loss": 5.4903,
      "mean_token_accuracy": 0.17342475801706314,
      "num_tokens": 8865018.0,
      "step": 4270
    },
    {
      "entropy": 5.752366018295288,
      "epoch": 0.3877902757619739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499003207361214,
      "loss": 5.6753,
      "mean_token_accuracy": 0.16161389499902726,
      "num_tokens": 8875423.0,
      "step": 4275
    },
    {
      "entropy": 5.819116306304932,
      "epoch": 0.38824383164005805,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990001627237165,
      "loss": 5.6046,
      "mean_token_accuracy": 0.1682436063885689,
      "num_tokens": 8887850.0,
      "step": 4280
    },
    {
      "entropy": 5.644161605834961,
      "epoch": 0.38869738751814226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989971134538386,
      "loss": 5.4972,
      "mean_token_accuracy": 0.17050439417362212,
      "num_tokens": 8897429.0,
      "step": 4285
    },
    {
      "entropy": 5.663009977340698,
      "epoch": 0.3891509433962264,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989940595516432,
      "loss": 5.5253,
      "mean_token_accuracy": 0.16778848469257354,
      "num_tokens": 8907048.0,
      "step": 4290
    },
    {
      "entropy": 5.702516841888428,
      "epoch": 0.3896044992743106,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989910010171938,
      "loss": 5.5746,
      "mean_token_accuracy": 0.16057664453983306,
      "num_tokens": 8917987.0,
      "step": 4295
    },
    {
      "entropy": 5.754789972305298,
      "epoch": 0.3900580551523948,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989879378505532,
      "loss": 5.4806,
      "mean_token_accuracy": 0.16423581093549727,
      "num_tokens": 8929339.0,
      "step": 4300
    },
    {
      "entropy": 5.599805641174316,
      "epoch": 0.39051161103047893,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989848700517849,
      "loss": 5.4908,
      "mean_token_accuracy": 0.1674097627401352,
      "num_tokens": 8939473.0,
      "step": 4305
    },
    {
      "entropy": 5.650110673904419,
      "epoch": 0.39096516690856314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989817976209526,
      "loss": 5.5306,
      "mean_token_accuracy": 0.16720281541347504,
      "num_tokens": 8949173.0,
      "step": 4310
    },
    {
      "entropy": 5.739528751373291,
      "epoch": 0.3914187227866473,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989787205581196,
      "loss": 5.4996,
      "mean_token_accuracy": 0.1708226829767227,
      "num_tokens": 8959879.0,
      "step": 4315
    },
    {
      "entropy": 5.707206296920776,
      "epoch": 0.3918722786647315,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989756388633496,
      "loss": 5.5819,
      "mean_token_accuracy": 0.16486465483903884,
      "num_tokens": 8970643.0,
      "step": 4320
    },
    {
      "entropy": 5.666028881072998,
      "epoch": 0.39232583454281567,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989725525367063,
      "loss": 5.4561,
      "mean_token_accuracy": 0.16958049088716506,
      "num_tokens": 8979882.0,
      "step": 4325
    },
    {
      "entropy": 5.7163489818572994,
      "epoch": 0.3927793904208999,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989694615782534,
      "loss": 5.5912,
      "mean_token_accuracy": 0.1618393614888191,
      "num_tokens": 8990925.0,
      "step": 4330
    },
    {
      "entropy": 5.722356939315796,
      "epoch": 0.39323294629898403,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989663659880552,
      "loss": 5.6084,
      "mean_token_accuracy": 0.1621517851948738,
      "num_tokens": 9001365.0,
      "step": 4335
    },
    {
      "entropy": 5.79314546585083,
      "epoch": 0.3936865021770682,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989632657661754,
      "loss": 5.6,
      "mean_token_accuracy": 0.1683468908071518,
      "num_tokens": 9011942.0,
      "step": 4340
    },
    {
      "entropy": 5.737189674377442,
      "epoch": 0.3941400580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498960160912678,
      "loss": 5.6458,
      "mean_token_accuracy": 0.16355903297662736,
      "num_tokens": 9022610.0,
      "step": 4345
    },
    {
      "entropy": 5.703625679016113,
      "epoch": 0.39459361393323655,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989570514276276,
      "loss": 5.5296,
      "mean_token_accuracy": 0.16396254748106004,
      "num_tokens": 9033034.0,
      "step": 4350
    },
    {
      "entropy": 5.721978998184204,
      "epoch": 0.39504716981132076,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989539373110883,
      "loss": 5.5728,
      "mean_token_accuracy": 0.16446860283613204,
      "num_tokens": 9043418.0,
      "step": 4355
    },
    {
      "entropy": 5.782990789413452,
      "epoch": 0.3955007256894049,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989508185631245,
      "loss": 5.6716,
      "mean_token_accuracy": 0.16027326881885529,
      "num_tokens": 9052708.0,
      "step": 4360
    },
    {
      "entropy": 5.617509841918945,
      "epoch": 0.39595428156748913,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989476951838006,
      "loss": 5.5176,
      "mean_token_accuracy": 0.16624224931001663,
      "num_tokens": 9062883.0,
      "step": 4365
    },
    {
      "entropy": 5.744745779037475,
      "epoch": 0.3964078374455733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989445671731814,
      "loss": 5.4776,
      "mean_token_accuracy": 0.16907311379909515,
      "num_tokens": 9072830.0,
      "step": 4370
    },
    {
      "entropy": 5.634798336029053,
      "epoch": 0.3968613933236575,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989414345313315,
      "loss": 5.4453,
      "mean_token_accuracy": 0.1632262200117111,
      "num_tokens": 9082749.0,
      "step": 4375
    },
    {
      "entropy": 5.70997200012207,
      "epoch": 0.39731494920174165,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004989382972583156,
      "loss": 5.5417,
      "mean_token_accuracy": 0.1624021738767624,
      "num_tokens": 9094188.0,
      "step": 4380
    },
    {
      "entropy": 5.6712054252624515,
      "epoch": 0.3977685050798258,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989351553541986,
      "loss": 5.5543,
      "mean_token_accuracy": 0.1630195915699005,
      "num_tokens": 9105023.0,
      "step": 4385
    },
    {
      "entropy": 5.616395616531372,
      "epoch": 0.39822206095791,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989320088190456,
      "loss": 5.504,
      "mean_token_accuracy": 0.1688954085111618,
      "num_tokens": 9115563.0,
      "step": 4390
    },
    {
      "entropy": 5.713869190216064,
      "epoch": 0.3986756168359942,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989288576529216,
      "loss": 5.5355,
      "mean_token_accuracy": 0.16001897305250168,
      "num_tokens": 9125394.0,
      "step": 4395
    },
    {
      "entropy": 5.724222660064697,
      "epoch": 0.3991291727140784,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989257018558917,
      "loss": 5.531,
      "mean_token_accuracy": 0.16804168224334717,
      "num_tokens": 9137061.0,
      "step": 4400
    },
    {
      "entropy": 5.685125684738159,
      "epoch": 0.39958272859216254,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989225414280213,
      "loss": 5.4906,
      "mean_token_accuracy": 0.168438558280468,
      "num_tokens": 9148196.0,
      "step": 4405
    },
    {
      "entropy": 5.665059661865234,
      "epoch": 0.40003628447024675,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989193763693756,
      "loss": 5.5142,
      "mean_token_accuracy": 0.1706180453300476,
      "num_tokens": 9158013.0,
      "step": 4410
    },
    {
      "entropy": 5.651648664474488,
      "epoch": 0.4004898403483309,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989162066800202,
      "loss": 5.4225,
      "mean_token_accuracy": 0.17173381000757218,
      "num_tokens": 9167813.0,
      "step": 4415
    },
    {
      "entropy": 5.637704706192016,
      "epoch": 0.4009433962264151,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989130323600205,
      "loss": 5.4884,
      "mean_token_accuracy": 0.16999162435531617,
      "num_tokens": 9177783.0,
      "step": 4420
    },
    {
      "entropy": 5.688581085205078,
      "epoch": 0.40139695210449927,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989098534094424,
      "loss": 5.5333,
      "mean_token_accuracy": 0.16801782250404357,
      "num_tokens": 9187556.0,
      "step": 4425
    },
    {
      "entropy": 5.703752279281616,
      "epoch": 0.4018505079825834,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989066698283512,
      "loss": 5.5988,
      "mean_token_accuracy": 0.16409932374954223,
      "num_tokens": 9198659.0,
      "step": 4430
    },
    {
      "entropy": 5.741377305984497,
      "epoch": 0.40230406386066764,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989034816168132,
      "loss": 5.5719,
      "mean_token_accuracy": 0.1686299681663513,
      "num_tokens": 9209546.0,
      "step": 4435
    },
    {
      "entropy": 5.628240251541138,
      "epoch": 0.4027576197387518,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989002887748941,
      "loss": 5.4811,
      "mean_token_accuracy": 0.17123182117938995,
      "num_tokens": 9219339.0,
      "step": 4440
    },
    {
      "entropy": 5.702338361740113,
      "epoch": 0.403211175616836,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988970913026601,
      "loss": 5.6741,
      "mean_token_accuracy": 0.1605461746454239,
      "num_tokens": 9230364.0,
      "step": 4445
    },
    {
      "entropy": 5.706683349609375,
      "epoch": 0.40366473149492016,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988938892001771,
      "loss": 5.5239,
      "mean_token_accuracy": 0.16557129025459288,
      "num_tokens": 9240732.0,
      "step": 4450
    },
    {
      "entropy": 5.59526686668396,
      "epoch": 0.40411828737300437,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988906824675113,
      "loss": 5.4616,
      "mean_token_accuracy": 0.17066627889871597,
      "num_tokens": 9251049.0,
      "step": 4455
    },
    {
      "entropy": 5.65702075958252,
      "epoch": 0.4045718432510885,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988874711047293,
      "loss": 5.5041,
      "mean_token_accuracy": 0.17010140866041185,
      "num_tokens": 9261660.0,
      "step": 4460
    },
    {
      "entropy": 5.707148694992066,
      "epoch": 0.40502539912917274,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988842551118972,
      "loss": 5.5619,
      "mean_token_accuracy": 0.17123832255601884,
      "num_tokens": 9272114.0,
      "step": 4465
    },
    {
      "entropy": 5.657516384124756,
      "epoch": 0.4054789550072569,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988810344890818,
      "loss": 5.5274,
      "mean_token_accuracy": 0.17154074907302858,
      "num_tokens": 9282125.0,
      "step": 4470
    },
    {
      "entropy": 5.755674219131469,
      "epoch": 0.40593251088534105,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988778092363494,
      "loss": 5.5719,
      "mean_token_accuracy": 0.164210744202137,
      "num_tokens": 9294078.0,
      "step": 4475
    },
    {
      "entropy": 5.703670406341553,
      "epoch": 0.40638606676342526,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988745793537671,
      "loss": 5.5273,
      "mean_token_accuracy": 0.16619765460491182,
      "num_tokens": 9304026.0,
      "step": 4480
    },
    {
      "entropy": 5.6847145557403564,
      "epoch": 0.4068396226415094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988713448414011,
      "loss": 5.5717,
      "mean_token_accuracy": 0.17452892959117888,
      "num_tokens": 9313933.0,
      "step": 4485
    },
    {
      "entropy": 5.682414197921753,
      "epoch": 0.4072931785195936,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498868105699319,
      "loss": 5.6038,
      "mean_token_accuracy": 0.1653165489435196,
      "num_tokens": 9323932.0,
      "step": 4490
    },
    {
      "entropy": 5.62096529006958,
      "epoch": 0.4077467343976778,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988648619275872,
      "loss": 5.4957,
      "mean_token_accuracy": 0.17853589355945587,
      "num_tokens": 9333467.0,
      "step": 4495
    },
    {
      "entropy": 5.730397415161133,
      "epoch": 0.408200290275762,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988616135262729,
      "loss": 5.5523,
      "mean_token_accuracy": 0.16236073523759842,
      "num_tokens": 9344909.0,
      "step": 4500
    },
    {
      "entropy": 5.778379106521607,
      "epoch": 0.40865384615384615,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004988583604954436,
      "loss": 5.6145,
      "mean_token_accuracy": 0.1632135108113289,
      "num_tokens": 9355630.0,
      "step": 4505
    },
    {
      "entropy": 5.630618858337402,
      "epoch": 0.40910740203193036,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988551028351663,
      "loss": 5.539,
      "mean_token_accuracy": 0.16523584127426147,
      "num_tokens": 9365700.0,
      "step": 4510
    },
    {
      "entropy": 5.654221391677856,
      "epoch": 0.4095609579100145,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988518405455084,
      "loss": 5.6261,
      "mean_token_accuracy": 0.15864572525024415,
      "num_tokens": 9377088.0,
      "step": 4515
    },
    {
      "entropy": 5.703935956954956,
      "epoch": 0.41001451378809867,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988485736265374,
      "loss": 5.4107,
      "mean_token_accuracy": 0.17571215331554413,
      "num_tokens": 9386855.0,
      "step": 4520
    },
    {
      "entropy": 5.696550416946411,
      "epoch": 0.4104680696661829,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498845302078321,
      "loss": 5.5307,
      "mean_token_accuracy": 0.16638017445802689,
      "num_tokens": 9398661.0,
      "step": 4525
    },
    {
      "entropy": 5.72752366065979,
      "epoch": 0.41092162554426703,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988420259009266,
      "loss": 5.5315,
      "mean_token_accuracy": 0.17306275367736818,
      "num_tokens": 9408670.0,
      "step": 4530
    },
    {
      "entropy": 5.688307809829712,
      "epoch": 0.41137518142235124,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498838745094422,
      "loss": 5.5582,
      "mean_token_accuracy": 0.16831061094999314,
      "num_tokens": 9418740.0,
      "step": 4535
    },
    {
      "entropy": 5.6272059917449955,
      "epoch": 0.4118287373004354,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988354596588751,
      "loss": 5.5311,
      "mean_token_accuracy": 0.16975374519824982,
      "num_tokens": 9429339.0,
      "step": 4540
    },
    {
      "entropy": 5.678937482833862,
      "epoch": 0.4122822931785196,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988321695943539,
      "loss": 5.4738,
      "mean_token_accuracy": 0.1705018848180771,
      "num_tokens": 9438996.0,
      "step": 4545
    },
    {
      "entropy": 5.698333978652954,
      "epoch": 0.41273584905660377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988288749009264,
      "loss": 5.4912,
      "mean_token_accuracy": 0.1687777042388916,
      "num_tokens": 9448669.0,
      "step": 4550
    },
    {
      "entropy": 5.645137596130371,
      "epoch": 0.413189404934688,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988255755786608,
      "loss": 5.4872,
      "mean_token_accuracy": 0.1714872419834137,
      "num_tokens": 9458857.0,
      "step": 4555
    },
    {
      "entropy": 5.650253391265869,
      "epoch": 0.41364296081277213,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988222716276251,
      "loss": 5.4782,
      "mean_token_accuracy": 0.17236211895942688,
      "num_tokens": 9470241.0,
      "step": 4560
    },
    {
      "entropy": 5.731791353225708,
      "epoch": 0.4140965166908563,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498818963047888,
      "loss": 5.6346,
      "mean_token_accuracy": 0.16064894944429398,
      "num_tokens": 9480528.0,
      "step": 4565
    },
    {
      "entropy": 5.75153079032898,
      "epoch": 0.4145500725689405,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988156498395176,
      "loss": 5.5396,
      "mean_token_accuracy": 0.17293335497379303,
      "num_tokens": 9491000.0,
      "step": 4570
    },
    {
      "entropy": 5.723613452911377,
      "epoch": 0.41500362844702465,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004988123320025825,
      "loss": 5.5322,
      "mean_token_accuracy": 0.17224910855293274,
      "num_tokens": 9501612.0,
      "step": 4575
    },
    {
      "entropy": 5.692364740371704,
      "epoch": 0.41545718432510886,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988090095371514,
      "loss": 5.639,
      "mean_token_accuracy": 0.16442762464284896,
      "num_tokens": 9513999.0,
      "step": 4580
    },
    {
      "entropy": 5.704233598709107,
      "epoch": 0.415910740203193,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988056824432928,
      "loss": 5.5106,
      "mean_token_accuracy": 0.1647787868976593,
      "num_tokens": 9524392.0,
      "step": 4585
    },
    {
      "entropy": 5.709184408187866,
      "epoch": 0.41636429608127723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988023507210758,
      "loss": 5.5108,
      "mean_token_accuracy": 0.1714200809597969,
      "num_tokens": 9534227.0,
      "step": 4590
    },
    {
      "entropy": 5.69327826499939,
      "epoch": 0.4168178519593614,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987990143705692,
      "loss": 5.5107,
      "mean_token_accuracy": 0.1758728012442589,
      "num_tokens": 9544242.0,
      "step": 4595
    },
    {
      "entropy": 5.6579484939575195,
      "epoch": 0.4172714078374456,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987956733918418,
      "loss": 5.434,
      "mean_token_accuracy": 0.17048389017581939,
      "num_tokens": 9553921.0,
      "step": 4600
    },
    {
      "entropy": 5.562104272842407,
      "epoch": 0.41772496371552975,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987923277849629,
      "loss": 5.4928,
      "mean_token_accuracy": 0.16425550878047943,
      "num_tokens": 9564768.0,
      "step": 4605
    },
    {
      "entropy": 5.7012450218200685,
      "epoch": 0.4181785195936139,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987889775500017,
      "loss": 5.5184,
      "mean_token_accuracy": 0.16336782574653624,
      "num_tokens": 9576360.0,
      "step": 4610
    },
    {
      "entropy": 5.6748738288879395,
      "epoch": 0.4186320754716981,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987856226870272,
      "loss": 5.54,
      "mean_token_accuracy": 0.16737818121910095,
      "num_tokens": 9587623.0,
      "step": 4615
    },
    {
      "entropy": 5.636663770675659,
      "epoch": 0.4190856313497823,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987822631961092,
      "loss": 5.5691,
      "mean_token_accuracy": 0.1642339438199997,
      "num_tokens": 9597837.0,
      "step": 4620
    },
    {
      "entropy": 5.706202650070191,
      "epoch": 0.4195391872278665,
      "grad_norm": 0.875,
      "learning_rate": 0.000498778899077317,
      "loss": 5.5288,
      "mean_token_accuracy": 0.1656550332903862,
      "num_tokens": 9609509.0,
      "step": 4625
    },
    {
      "entropy": 5.667462539672852,
      "epoch": 0.41999274310595064,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049877553033072,
      "loss": 5.4957,
      "mean_token_accuracy": 0.17181909531354905,
      "num_tokens": 9619954.0,
      "step": 4630
    },
    {
      "entropy": 5.644245195388794,
      "epoch": 0.42044629898403485,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987721569563882,
      "loss": 5.4973,
      "mean_token_accuracy": 0.16737744808197022,
      "num_tokens": 9629742.0,
      "step": 4635
    },
    {
      "entropy": 5.666360330581665,
      "epoch": 0.420899854862119,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498768778954391,
      "loss": 5.515,
      "mean_token_accuracy": 0.1711300268769264,
      "num_tokens": 9641378.0,
      "step": 4640
    },
    {
      "entropy": 5.756295871734619,
      "epoch": 0.4213534107402032,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987653963247985,
      "loss": 5.5852,
      "mean_token_accuracy": 0.16515168696641921,
      "num_tokens": 9652279.0,
      "step": 4645
    },
    {
      "entropy": 5.715899276733398,
      "epoch": 0.42180696661828737,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004987620090676805,
      "loss": 5.5811,
      "mean_token_accuracy": 0.16966685950756072,
      "num_tokens": 9663416.0,
      "step": 4650
    },
    {
      "entropy": 5.65866527557373,
      "epoch": 0.4222605224963715,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987586171831072,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16795586943626403,
      "num_tokens": 9674647.0,
      "step": 4655
    },
    {
      "entropy": 5.646727991104126,
      "epoch": 0.42271407837445574,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987552206711487,
      "loss": 5.3947,
      "mean_token_accuracy": 0.17098441645503043,
      "num_tokens": 9685084.0,
      "step": 4660
    },
    {
      "entropy": 5.602826547622681,
      "epoch": 0.4231676342525399,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987518195318752,
      "loss": 5.451,
      "mean_token_accuracy": 0.17014521807432176,
      "num_tokens": 9695661.0,
      "step": 4665
    },
    {
      "entropy": 5.699631309509277,
      "epoch": 0.4236211901306241,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498748413765357,
      "loss": 5.5111,
      "mean_token_accuracy": 0.1655079945921898,
      "num_tokens": 9705773.0,
      "step": 4670
    },
    {
      "entropy": 5.672311735153198,
      "epoch": 0.42407474600870826,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987450033716646,
      "loss": 5.5543,
      "mean_token_accuracy": 0.1671050012111664,
      "num_tokens": 9715799.0,
      "step": 4675
    },
    {
      "entropy": 5.70002236366272,
      "epoch": 0.42452830188679247,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987415883508686,
      "loss": 5.5117,
      "mean_token_accuracy": 0.17100395262241364,
      "num_tokens": 9727012.0,
      "step": 4680
    },
    {
      "entropy": 5.699238252639771,
      "epoch": 0.4249818577648766,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987381687030394,
      "loss": 5.5134,
      "mean_token_accuracy": 0.16880285739898682,
      "num_tokens": 9737044.0,
      "step": 4685
    },
    {
      "entropy": 5.728292417526245,
      "epoch": 0.42543541364296084,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987347444282479,
      "loss": 5.5911,
      "mean_token_accuracy": 0.16349772363901138,
      "num_tokens": 9747648.0,
      "step": 4690
    },
    {
      "entropy": 5.686186790466309,
      "epoch": 0.425888969521045,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987313155265649,
      "loss": 5.5245,
      "mean_token_accuracy": 0.163296639919281,
      "num_tokens": 9757812.0,
      "step": 4695
    },
    {
      "entropy": 5.753760099411011,
      "epoch": 0.42634252539912915,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987278819980613,
      "loss": 5.5178,
      "mean_token_accuracy": 0.174093596637249,
      "num_tokens": 9768337.0,
      "step": 4700
    },
    {
      "entropy": 5.703652429580688,
      "epoch": 0.42679608127721336,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987244438428081,
      "loss": 5.5634,
      "mean_token_accuracy": 0.16444121152162552,
      "num_tokens": 9779433.0,
      "step": 4705
    },
    {
      "entropy": 5.611905670166015,
      "epoch": 0.4272496371552975,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987210010608764,
      "loss": 5.4602,
      "mean_token_accuracy": 0.1711221307516098,
      "num_tokens": 9790069.0,
      "step": 4710
    },
    {
      "entropy": 5.659437561035157,
      "epoch": 0.4277031930333817,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987175536523373,
      "loss": 5.4316,
      "mean_token_accuracy": 0.17792649418115616,
      "num_tokens": 9799669.0,
      "step": 4715
    },
    {
      "entropy": 5.626423311233521,
      "epoch": 0.4281567489114659,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987141016172622,
      "loss": 5.4731,
      "mean_token_accuracy": 0.1750251442193985,
      "num_tokens": 9809615.0,
      "step": 4720
    },
    {
      "entropy": 5.633890104293823,
      "epoch": 0.4286103047895501,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987106449557225,
      "loss": 5.4379,
      "mean_token_accuracy": 0.17350732386112214,
      "num_tokens": 9819666.0,
      "step": 4725
    },
    {
      "entropy": 5.73306941986084,
      "epoch": 0.42906386066763424,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987071836677896,
      "loss": 5.6122,
      "mean_token_accuracy": 0.16756637394428253,
      "num_tokens": 9831015.0,
      "step": 4730
    },
    {
      "entropy": 5.635014009475708,
      "epoch": 0.42951741654571846,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987037177535353,
      "loss": 5.4679,
      "mean_token_accuracy": 0.1639992356300354,
      "num_tokens": 9841354.0,
      "step": 4735
    },
    {
      "entropy": 5.633705282211304,
      "epoch": 0.4299709724238026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987002472130309,
      "loss": 5.4563,
      "mean_token_accuracy": 0.16753149628639222,
      "num_tokens": 9850467.0,
      "step": 4740
    },
    {
      "entropy": 5.770284414291382,
      "epoch": 0.43042452830188677,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986967720463485,
      "loss": 5.5496,
      "mean_token_accuracy": 0.1623040944337845,
      "num_tokens": 9861115.0,
      "step": 4745
    },
    {
      "entropy": 5.7330851554870605,
      "epoch": 0.430878084179971,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986932922535598,
      "loss": 5.5384,
      "mean_token_accuracy": 0.16642699092626573,
      "num_tokens": 9871172.0,
      "step": 4750
    },
    {
      "entropy": 5.652520322799683,
      "epoch": 0.43133164005805513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986898078347368,
      "loss": 5.4515,
      "mean_token_accuracy": 0.17356225550174714,
      "num_tokens": 9881997.0,
      "step": 4755
    },
    {
      "entropy": 5.61650390625,
      "epoch": 0.43178519593613934,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986863187899516,
      "loss": 5.4211,
      "mean_token_accuracy": 0.1723109319806099,
      "num_tokens": 9892310.0,
      "step": 4760
    },
    {
      "entropy": 5.6054548740386965,
      "epoch": 0.4322387518142235,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986828251192764,
      "loss": 5.4786,
      "mean_token_accuracy": 0.17029564827680588,
      "num_tokens": 9902664.0,
      "step": 4765
    },
    {
      "entropy": 5.6452922344207765,
      "epoch": 0.4326923076923077,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986793268227833,
      "loss": 5.4662,
      "mean_token_accuracy": 0.17340250313282013,
      "num_tokens": 9912491.0,
      "step": 4770
    },
    {
      "entropy": 5.65220012664795,
      "epoch": 0.43314586357039186,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986758239005448,
      "loss": 5.439,
      "mean_token_accuracy": 0.17318663150072097,
      "num_tokens": 9922057.0,
      "step": 4775
    },
    {
      "entropy": 5.640111303329467,
      "epoch": 0.4335994194484761,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986723163526331,
      "loss": 5.4976,
      "mean_token_accuracy": 0.1674135446548462,
      "num_tokens": 9933661.0,
      "step": 4780
    },
    {
      "entropy": 5.577649021148682,
      "epoch": 0.43405297532656023,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986688041791209,
      "loss": 5.4532,
      "mean_token_accuracy": 0.16382990777492523,
      "num_tokens": 9945177.0,
      "step": 4785
    },
    {
      "entropy": 5.671372318267823,
      "epoch": 0.4345065312046444,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498665287380081,
      "loss": 5.479,
      "mean_token_accuracy": 0.1665464699268341,
      "num_tokens": 9956226.0,
      "step": 4790
    },
    {
      "entropy": 5.709822797775269,
      "epoch": 0.4349600870827286,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986617659555859,
      "loss": 5.5528,
      "mean_token_accuracy": 0.16944931149482728,
      "num_tokens": 9965877.0,
      "step": 4795
    },
    {
      "entropy": 5.627245140075684,
      "epoch": 0.43541364296081275,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986582399057085,
      "loss": 5.4958,
      "mean_token_accuracy": 0.17361783385276794,
      "num_tokens": 9975592.0,
      "step": 4800
    },
    {
      "entropy": 5.716971731185913,
      "epoch": 0.43586719883889696,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986547092305216,
      "loss": 5.5188,
      "mean_token_accuracy": 0.16906160414218901,
      "num_tokens": 9986094.0,
      "step": 4805
    },
    {
      "entropy": 5.747332000732422,
      "epoch": 0.4363207547169811,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986511739300984,
      "loss": 5.5563,
      "mean_token_accuracy": 0.16685436069965362,
      "num_tokens": 9997241.0,
      "step": 4810
    },
    {
      "entropy": 5.624169874191284,
      "epoch": 0.43677431059506533,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986476340045119,
      "loss": 5.4805,
      "mean_token_accuracy": 0.1696499690413475,
      "num_tokens": 10008091.0,
      "step": 4815
    },
    {
      "entropy": 5.610916471481323,
      "epoch": 0.4372278664731495,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986440894538354,
      "loss": 5.5358,
      "mean_token_accuracy": 0.16445003151893617,
      "num_tokens": 10018794.0,
      "step": 4820
    },
    {
      "entropy": 5.658578824996948,
      "epoch": 0.4376814223512337,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986405402781421,
      "loss": 5.4595,
      "mean_token_accuracy": 0.17329651415348052,
      "num_tokens": 10028414.0,
      "step": 4825
    },
    {
      "entropy": 5.683524417877197,
      "epoch": 0.43813497822931785,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986369864775054,
      "loss": 5.5255,
      "mean_token_accuracy": 0.1668830245733261,
      "num_tokens": 10038749.0,
      "step": 4830
    },
    {
      "entropy": 5.680928516387939,
      "epoch": 0.438588534107402,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986334280519989,
      "loss": 5.4706,
      "mean_token_accuracy": 0.16898296475410463,
      "num_tokens": 10049866.0,
      "step": 4835
    },
    {
      "entropy": 5.620717573165893,
      "epoch": 0.4390420899854862,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498629865001696,
      "loss": 5.4971,
      "mean_token_accuracy": 0.176370669901371,
      "num_tokens": 10060680.0,
      "step": 4840
    },
    {
      "entropy": 5.6851057529449465,
      "epoch": 0.43949564586357037,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986262973266706,
      "loss": 5.6094,
      "mean_token_accuracy": 0.16637811809778214,
      "num_tokens": 10070931.0,
      "step": 4845
    },
    {
      "entropy": 5.648931550979614,
      "epoch": 0.4399492017416546,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986227250269964,
      "loss": 5.5285,
      "mean_token_accuracy": 0.17198051065206527,
      "num_tokens": 10080189.0,
      "step": 4850
    },
    {
      "entropy": 5.683497285842895,
      "epoch": 0.44040275761973874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004986191481027472,
      "loss": 5.4349,
      "mean_token_accuracy": 0.17278559356927872,
      "num_tokens": 10088813.0,
      "step": 4855
    },
    {
      "entropy": 5.61211371421814,
      "epoch": 0.44085631349782295,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986155665539971,
      "loss": 5.4644,
      "mean_token_accuracy": 0.17080731093883514,
      "num_tokens": 10098125.0,
      "step": 4860
    },
    {
      "entropy": 5.69177656173706,
      "epoch": 0.4413098693759071,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049861198038082,
      "loss": 5.506,
      "mean_token_accuracy": 0.17002110332250595,
      "num_tokens": 10109125.0,
      "step": 4865
    },
    {
      "entropy": 5.673552989959717,
      "epoch": 0.4417634252539913,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986083895832903,
      "loss": 5.4864,
      "mean_token_accuracy": 0.16711554676294327,
      "num_tokens": 10120182.0,
      "step": 4870
    },
    {
      "entropy": 5.619903326034546,
      "epoch": 0.44221698113207547,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986047941614821,
      "loss": 5.5446,
      "mean_token_accuracy": 0.1634003072977066,
      "num_tokens": 10130941.0,
      "step": 4875
    },
    {
      "entropy": 5.683604526519775,
      "epoch": 0.4426705370101596,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986011941154696,
      "loss": 5.5044,
      "mean_token_accuracy": 0.16631961166858672,
      "num_tokens": 10141708.0,
      "step": 4880
    },
    {
      "entropy": 5.637266683578491,
      "epoch": 0.44312409288824384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985975894453275,
      "loss": 5.4431,
      "mean_token_accuracy": 0.16837728172540664,
      "num_tokens": 10151816.0,
      "step": 4885
    },
    {
      "entropy": 5.694721031188965,
      "epoch": 0.443577648766328,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985939801511304,
      "loss": 5.5662,
      "mean_token_accuracy": 0.16548764184117318,
      "num_tokens": 10162470.0,
      "step": 4890
    },
    {
      "entropy": 5.672745275497436,
      "epoch": 0.4440312046444122,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985903662329527,
      "loss": 5.4515,
      "mean_token_accuracy": 0.17449744194746017,
      "num_tokens": 10173171.0,
      "step": 4895
    },
    {
      "entropy": 5.667294073104858,
      "epoch": 0.44448476052249636,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985867476908693,
      "loss": 5.4873,
      "mean_token_accuracy": 0.17015801668167113,
      "num_tokens": 10184185.0,
      "step": 4900
    },
    {
      "entropy": 5.702019119262696,
      "epoch": 0.44493831640058057,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985831245249549,
      "loss": 5.5464,
      "mean_token_accuracy": 0.169259525090456,
      "num_tokens": 10195516.0,
      "step": 4905
    },
    {
      "entropy": 5.6287211894989015,
      "epoch": 0.4453918722786647,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985794967352846,
      "loss": 5.4707,
      "mean_token_accuracy": 0.1748988300561905,
      "num_tokens": 10204936.0,
      "step": 4910
    },
    {
      "entropy": 5.658074283599854,
      "epoch": 0.44584542815674894,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004985758643219334,
      "loss": 5.4998,
      "mean_token_accuracy": 0.17041691839694978,
      "num_tokens": 10216218.0,
      "step": 4915
    },
    {
      "entropy": 5.744098043441772,
      "epoch": 0.4462989840348331,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985722272849762,
      "loss": 5.5099,
      "mean_token_accuracy": 0.16554411202669145,
      "num_tokens": 10226370.0,
      "step": 4920
    },
    {
      "entropy": 5.618660640716553,
      "epoch": 0.44675253991291725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985685856244884,
      "loss": 5.4398,
      "mean_token_accuracy": 0.18081948906183243,
      "num_tokens": 10236170.0,
      "step": 4925
    },
    {
      "entropy": 5.666572141647339,
      "epoch": 0.44720609579100146,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985649393405453,
      "loss": 5.5025,
      "mean_token_accuracy": 0.16895841807127,
      "num_tokens": 10246119.0,
      "step": 4930
    },
    {
      "entropy": 5.671585464477539,
      "epoch": 0.4476596516690856,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004985612884332223,
      "loss": 5.5196,
      "mean_token_accuracy": 0.16341286897659302,
      "num_tokens": 10256762.0,
      "step": 4935
    },
    {
      "entropy": 5.617032051086426,
      "epoch": 0.4481132075471698,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985576329025949,
      "loss": 5.4652,
      "mean_token_accuracy": 0.17075106054544448,
      "num_tokens": 10267317.0,
      "step": 4940
    },
    {
      "entropy": 5.681619024276733,
      "epoch": 0.448566763425254,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985539727487385,
      "loss": 5.603,
      "mean_token_accuracy": 0.15708888173103333,
      "num_tokens": 10279111.0,
      "step": 4945
    },
    {
      "entropy": 5.6958630084991455,
      "epoch": 0.4490203193033382,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985503079717292,
      "loss": 5.4674,
      "mean_token_accuracy": 0.1703536629676819,
      "num_tokens": 10289383.0,
      "step": 4950
    },
    {
      "entropy": 5.6642831325531,
      "epoch": 0.44947387518142234,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985466385716425,
      "loss": 5.4767,
      "mean_token_accuracy": 0.17260171324014664,
      "num_tokens": 10299806.0,
      "step": 4955
    },
    {
      "entropy": 5.578657579421997,
      "epoch": 0.44992743105950656,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985429645485543,
      "loss": 5.5005,
      "mean_token_accuracy": 0.1660196453332901,
      "num_tokens": 10310348.0,
      "step": 4960
    },
    {
      "entropy": 5.674696922302246,
      "epoch": 0.4503809869375907,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985392859025407,
      "loss": 5.4875,
      "mean_token_accuracy": 0.17419226914644242,
      "num_tokens": 10320354.0,
      "step": 4965
    },
    {
      "entropy": 5.604598712921143,
      "epoch": 0.45083454281567487,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985356026336776,
      "loss": 5.4148,
      "mean_token_accuracy": 0.16999066323041917,
      "num_tokens": 10330292.0,
      "step": 4970
    },
    {
      "entropy": 5.621414422988892,
      "epoch": 0.4512880986937591,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985319147420414,
      "loss": 5.3829,
      "mean_token_accuracy": 0.18479472696781157,
      "num_tokens": 10340449.0,
      "step": 4975
    },
    {
      "entropy": 5.649209356307983,
      "epoch": 0.45174165457184323,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985282222277082,
      "loss": 5.4569,
      "mean_token_accuracy": 0.16669374704360962,
      "num_tokens": 10350637.0,
      "step": 4980
    },
    {
      "entropy": 5.581578731536865,
      "epoch": 0.45219521044992744,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985245250907544,
      "loss": 5.3903,
      "mean_token_accuracy": 0.17272900640964509,
      "num_tokens": 10361548.0,
      "step": 4985
    },
    {
      "entropy": 5.573017024993897,
      "epoch": 0.4526487663280116,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985208233312564,
      "loss": 5.4129,
      "mean_token_accuracy": 0.16686842143535613,
      "num_tokens": 10371765.0,
      "step": 4990
    },
    {
      "entropy": 5.696421766281128,
      "epoch": 0.4531023222060958,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985171169492908,
      "loss": 5.5166,
      "mean_token_accuracy": 0.17461331635713578,
      "num_tokens": 10382108.0,
      "step": 4995
    },
    {
      "entropy": 5.630676031112671,
      "epoch": 0.45355587808417996,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985134059449344,
      "loss": 5.4666,
      "mean_token_accuracy": 0.18015743643045426,
      "num_tokens": 10392390.0,
      "step": 5000
    },
    {
      "entropy": 5.6141965866088865,
      "epoch": 0.4540094339622642,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985096903182637,
      "loss": 5.5127,
      "mean_token_accuracy": 0.17537071108818053,
      "num_tokens": 10402094.0,
      "step": 5005
    },
    {
      "entropy": 5.666981887817383,
      "epoch": 0.45446298984034833,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985059700693557,
      "loss": 5.466,
      "mean_token_accuracy": 0.173527130484581,
      "num_tokens": 10413350.0,
      "step": 5010
    },
    {
      "entropy": 5.702916288375855,
      "epoch": 0.4549165457184325,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985022451982872,
      "loss": 5.5916,
      "mean_token_accuracy": 0.16218466758728028,
      "num_tokens": 10424292.0,
      "step": 5015
    },
    {
      "entropy": 5.756633567810058,
      "epoch": 0.4553701015965167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984985157051354,
      "loss": 5.4481,
      "mean_token_accuracy": 0.1688245952129364,
      "num_tokens": 10433724.0,
      "step": 5020
    },
    {
      "entropy": 5.648002099990845,
      "epoch": 0.45582365747460085,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984947815899774,
      "loss": 5.5425,
      "mean_token_accuracy": 0.1708854228258133,
      "num_tokens": 10444724.0,
      "step": 5025
    },
    {
      "entropy": 5.65775728225708,
      "epoch": 0.45627721335268506,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984910428528902,
      "loss": 5.4623,
      "mean_token_accuracy": 0.17359500378370285,
      "num_tokens": 10455086.0,
      "step": 5030
    },
    {
      "entropy": 5.598028898239136,
      "epoch": 0.4567307692307692,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984872994939514,
      "loss": 5.4155,
      "mean_token_accuracy": 0.1689280539751053,
      "num_tokens": 10465076.0,
      "step": 5035
    },
    {
      "entropy": 5.6523706912994385,
      "epoch": 0.45718432510885343,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984835515132382,
      "loss": 5.4321,
      "mean_token_accuracy": 0.17813173532485962,
      "num_tokens": 10474817.0,
      "step": 5040
    },
    {
      "entropy": 5.67584810256958,
      "epoch": 0.4576378809869376,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004984797989108282,
      "loss": 5.4584,
      "mean_token_accuracy": 0.1707393169403076,
      "num_tokens": 10485735.0,
      "step": 5045
    },
    {
      "entropy": 5.613812065124511,
      "epoch": 0.4580914368650218,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984760416867991,
      "loss": 5.5472,
      "mean_token_accuracy": 0.1682056277990341,
      "num_tokens": 10496731.0,
      "step": 5050
    },
    {
      "entropy": 5.568698740005493,
      "epoch": 0.45854499274310595,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984722798412286,
      "loss": 5.4136,
      "mean_token_accuracy": 0.17073957175016402,
      "num_tokens": 10506787.0,
      "step": 5055
    },
    {
      "entropy": 5.659662580490112,
      "epoch": 0.4589985486211901,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984685133741941,
      "loss": 5.4739,
      "mean_token_accuracy": 0.16943865567445754,
      "num_tokens": 10517170.0,
      "step": 5060
    },
    {
      "entropy": 5.622515487670898,
      "epoch": 0.4594521044992743,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498464742285774,
      "loss": 5.4733,
      "mean_token_accuracy": 0.1717009276151657,
      "num_tokens": 10527342.0,
      "step": 5065
    },
    {
      "entropy": 5.642328071594238,
      "epoch": 0.45990566037735847,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498460966576046,
      "loss": 5.4503,
      "mean_token_accuracy": 0.17293796241283416,
      "num_tokens": 10537253.0,
      "step": 5070
    },
    {
      "entropy": 5.698889923095703,
      "epoch": 0.4603592162554427,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984571862450884,
      "loss": 5.465,
      "mean_token_accuracy": 0.1680241197347641,
      "num_tokens": 10547249.0,
      "step": 5075
    },
    {
      "entropy": 5.7043322086334225,
      "epoch": 0.46081277213352684,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004984534012929791,
      "loss": 5.503,
      "mean_token_accuracy": 0.172734771668911,
      "num_tokens": 10558174.0,
      "step": 5080
    },
    {
      "entropy": 5.700099897384644,
      "epoch": 0.46126632801161105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984496117197967,
      "loss": 5.4444,
      "mean_token_accuracy": 0.17787872403860092,
      "num_tokens": 10567610.0,
      "step": 5085
    },
    {
      "entropy": 5.602445316314697,
      "epoch": 0.4617198838896952,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984458175256193,
      "loss": 5.462,
      "mean_token_accuracy": 0.16895331740379332,
      "num_tokens": 10578494.0,
      "step": 5090
    },
    {
      "entropy": 5.690131187438965,
      "epoch": 0.4621734397677794,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984420187105253,
      "loss": 5.5668,
      "mean_token_accuracy": 0.16630938053131103,
      "num_tokens": 10589482.0,
      "step": 5095
    },
    {
      "entropy": 5.727915954589844,
      "epoch": 0.46262699564586357,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984382152745936,
      "loss": 5.5203,
      "mean_token_accuracy": 0.1676805019378662,
      "num_tokens": 10599965.0,
      "step": 5100
    },
    {
      "entropy": 5.638349914550782,
      "epoch": 0.4630805515239477,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984344072179026,
      "loss": 5.5459,
      "mean_token_accuracy": 0.1719616398215294,
      "num_tokens": 10610500.0,
      "step": 5105
    },
    {
      "entropy": 5.737101459503174,
      "epoch": 0.46353410740203194,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984305945405312,
      "loss": 5.5,
      "mean_token_accuracy": 0.17199551463127136,
      "num_tokens": 10619843.0,
      "step": 5110
    },
    {
      "entropy": 5.645010232925415,
      "epoch": 0.4639876632801161,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498426777242558,
      "loss": 5.4291,
      "mean_token_accuracy": 0.1767846554517746,
      "num_tokens": 10631026.0,
      "step": 5115
    },
    {
      "entropy": 5.490650749206543,
      "epoch": 0.4644412191582003,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984229553240623,
      "loss": 5.4348,
      "mean_token_accuracy": 0.18036371171474458,
      "num_tokens": 10641130.0,
      "step": 5120
    },
    {
      "entropy": 5.699585437774658,
      "epoch": 0.46489477503628446,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984191287851228,
      "loss": 5.4958,
      "mean_token_accuracy": 0.17234579771757125,
      "num_tokens": 10652305.0,
      "step": 5125
    },
    {
      "entropy": 5.680755996704102,
      "epoch": 0.46534833091436867,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984152976258188,
      "loss": 5.5328,
      "mean_token_accuracy": 0.17020701766014099,
      "num_tokens": 10663083.0,
      "step": 5130
    },
    {
      "entropy": 5.659653997421264,
      "epoch": 0.4658018867924528,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984114618462296,
      "loss": 5.5545,
      "mean_token_accuracy": 0.16575139313936232,
      "num_tokens": 10673602.0,
      "step": 5135
    },
    {
      "entropy": 5.619678211212158,
      "epoch": 0.46625544267053703,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984076214464343,
      "loss": 5.3841,
      "mean_token_accuracy": 0.1807913675904274,
      "num_tokens": 10684622.0,
      "step": 5140
    },
    {
      "entropy": 5.624106788635254,
      "epoch": 0.4667089985486212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984037764265126,
      "loss": 5.4969,
      "mean_token_accuracy": 0.17183390706777574,
      "num_tokens": 10695129.0,
      "step": 5145
    },
    {
      "entropy": 5.588241958618164,
      "epoch": 0.46716255442670535,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983999267865438,
      "loss": 5.3285,
      "mean_token_accuracy": 0.18249204456806184,
      "num_tokens": 10705236.0,
      "step": 5150
    },
    {
      "entropy": 5.593046188354492,
      "epoch": 0.46761611030478956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983960725266078,
      "loss": 5.4582,
      "mean_token_accuracy": 0.17504026740789413,
      "num_tokens": 10715748.0,
      "step": 5155
    },
    {
      "entropy": 5.624542713165283,
      "epoch": 0.4680696661828737,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983922136467838,
      "loss": 5.3846,
      "mean_token_accuracy": 0.1776614785194397,
      "num_tokens": 10725799.0,
      "step": 5160
    },
    {
      "entropy": 5.6600791931152346,
      "epoch": 0.4685232220609579,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498388350147152,
      "loss": 5.5062,
      "mean_token_accuracy": 0.16867911964654922,
      "num_tokens": 10735668.0,
      "step": 5165
    },
    {
      "entropy": 5.662786149978638,
      "epoch": 0.4689767779390421,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983844820277922,
      "loss": 5.6079,
      "mean_token_accuracy": 0.167015041410923,
      "num_tokens": 10746466.0,
      "step": 5170
    },
    {
      "entropy": 5.752711057662964,
      "epoch": 0.4694303338171263,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983806092887843,
      "loss": 5.5063,
      "mean_token_accuracy": 0.1687478616833687,
      "num_tokens": 10756724.0,
      "step": 5175
    },
    {
      "entropy": 5.664383506774902,
      "epoch": 0.46988388969521044,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983767319302086,
      "loss": 5.5499,
      "mean_token_accuracy": 0.16907021701335906,
      "num_tokens": 10766919.0,
      "step": 5180
    },
    {
      "entropy": 5.576836204528808,
      "epoch": 0.47033744557329465,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983728499521451,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18249264657497405,
      "num_tokens": 10777550.0,
      "step": 5185
    },
    {
      "entropy": 5.655050230026245,
      "epoch": 0.4707910014513788,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983689633546741,
      "loss": 5.4027,
      "mean_token_accuracy": 0.17136975079774858,
      "num_tokens": 10788456.0,
      "step": 5190
    },
    {
      "entropy": 5.663191461563111,
      "epoch": 0.47124455732946297,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983650721378761,
      "loss": 5.4345,
      "mean_token_accuracy": 0.17229533791542054,
      "num_tokens": 10800049.0,
      "step": 5195
    },
    {
      "entropy": 5.655028867721557,
      "epoch": 0.4716981132075472,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983611763018314,
      "loss": 5.4815,
      "mean_token_accuracy": 0.16865273416042328,
      "num_tokens": 10810743.0,
      "step": 5200
    },
    {
      "entropy": 5.753147888183594,
      "epoch": 0.47215166908563133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983572758466208,
      "loss": 5.677,
      "mean_token_accuracy": 0.16416452825069427,
      "num_tokens": 10821897.0,
      "step": 5205
    },
    {
      "entropy": 5.7169407367706295,
      "epoch": 0.47260522496371554,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983533707723247,
      "loss": 5.527,
      "mean_token_accuracy": 0.16855133473873138,
      "num_tokens": 10831589.0,
      "step": 5210
    },
    {
      "entropy": 5.657151603698731,
      "epoch": 0.4730587808417997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983494610790241,
      "loss": 5.429,
      "mean_token_accuracy": 0.16840204745531082,
      "num_tokens": 10841404.0,
      "step": 5215
    },
    {
      "entropy": 5.6251708984375,
      "epoch": 0.4735123367198839,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983455467667996,
      "loss": 5.471,
      "mean_token_accuracy": 0.1659417137503624,
      "num_tokens": 10851577.0,
      "step": 5220
    },
    {
      "entropy": 5.615735912322998,
      "epoch": 0.47396589259796806,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983416278357322,
      "loss": 5.3906,
      "mean_token_accuracy": 0.1765459805727005,
      "num_tokens": 10861403.0,
      "step": 5225
    },
    {
      "entropy": 5.629169273376465,
      "epoch": 0.4744194484760523,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983377042859032,
      "loss": 5.4383,
      "mean_token_accuracy": 0.17094605714082717,
      "num_tokens": 10871933.0,
      "step": 5230
    },
    {
      "entropy": 5.5857017040252686,
      "epoch": 0.47487300435413643,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983337761173936,
      "loss": 5.5079,
      "mean_token_accuracy": 0.16988566964864732,
      "num_tokens": 10881800.0,
      "step": 5235
    },
    {
      "entropy": 5.733130645751953,
      "epoch": 0.4753265602322206,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983298433302843,
      "loss": 5.4496,
      "mean_token_accuracy": 0.17222348898649215,
      "num_tokens": 10891830.0,
      "step": 5240
    },
    {
      "entropy": 5.616529273986816,
      "epoch": 0.4757801161103048,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983259059246572,
      "loss": 5.3783,
      "mean_token_accuracy": 0.18143776208162307,
      "num_tokens": 10901557.0,
      "step": 5245
    },
    {
      "entropy": 5.551515579223633,
      "epoch": 0.47623367198838895,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983219639005934,
      "loss": 5.5078,
      "mean_token_accuracy": 0.1646910712122917,
      "num_tokens": 10912155.0,
      "step": 5250
    },
    {
      "entropy": 5.639435863494873,
      "epoch": 0.47668722786647316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983180172581745,
      "loss": 5.4085,
      "mean_token_accuracy": 0.17542287111282348,
      "num_tokens": 10922291.0,
      "step": 5255
    },
    {
      "entropy": 5.703779172897339,
      "epoch": 0.4771407837445573,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983140659974819,
      "loss": 5.478,
      "mean_token_accuracy": 0.1703026846051216,
      "num_tokens": 10933176.0,
      "step": 5260
    },
    {
      "entropy": 5.653920316696167,
      "epoch": 0.47759433962264153,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983101101185977,
      "loss": 5.5772,
      "mean_token_accuracy": 0.16286463141441346,
      "num_tokens": 10944049.0,
      "step": 5265
    },
    {
      "entropy": 5.619420862197876,
      "epoch": 0.4780478955007257,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983061496216035,
      "loss": 5.4072,
      "mean_token_accuracy": 0.17815487831830978,
      "num_tokens": 10954347.0,
      "step": 5270
    },
    {
      "entropy": 5.614488506317139,
      "epoch": 0.4785014513788099,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983021845065812,
      "loss": 5.4529,
      "mean_token_accuracy": 0.17562718093395233,
      "num_tokens": 10965123.0,
      "step": 5275
    },
    {
      "entropy": 5.580786514282226,
      "epoch": 0.47895500725689405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982982147736128,
      "loss": 5.4492,
      "mean_token_accuracy": 0.1764552414417267,
      "num_tokens": 10976395.0,
      "step": 5280
    },
    {
      "entropy": 5.654614782333374,
      "epoch": 0.4794085631349782,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982942404227805,
      "loss": 5.437,
      "mean_token_accuracy": 0.16845346987247467,
      "num_tokens": 10986797.0,
      "step": 5285
    },
    {
      "entropy": 5.695258712768554,
      "epoch": 0.4798621190130624,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982902614541662,
      "loss": 5.4915,
      "mean_token_accuracy": 0.16865674555301666,
      "num_tokens": 10997229.0,
      "step": 5290
    },
    {
      "entropy": 5.583453416824341,
      "epoch": 0.48031567489114657,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982862778678526,
      "loss": 5.4231,
      "mean_token_accuracy": 0.1796005994081497,
      "num_tokens": 11008180.0,
      "step": 5295
    },
    {
      "entropy": 5.642146348953247,
      "epoch": 0.4807692307692308,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982822896639218,
      "loss": 5.4431,
      "mean_token_accuracy": 0.17733724117279054,
      "num_tokens": 11019248.0,
      "step": 5300
    },
    {
      "entropy": 5.5944068908691404,
      "epoch": 0.48122278664731494,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982782968424563,
      "loss": 5.3671,
      "mean_token_accuracy": 0.1745454654097557,
      "num_tokens": 11029760.0,
      "step": 5305
    },
    {
      "entropy": 5.60427680015564,
      "epoch": 0.48167634252539915,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982742994035388,
      "loss": 5.4575,
      "mean_token_accuracy": 0.17510539293289185,
      "num_tokens": 11039211.0,
      "step": 5310
    },
    {
      "entropy": 5.634851455688477,
      "epoch": 0.4821298984034833,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982702973472518,
      "loss": 5.4293,
      "mean_token_accuracy": 0.17235955595970154,
      "num_tokens": 11049892.0,
      "step": 5315
    },
    {
      "entropy": 5.697804927825928,
      "epoch": 0.4825834542815675,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982662906736783,
      "loss": 5.4824,
      "mean_token_accuracy": 0.17303521037101746,
      "num_tokens": 11060823.0,
      "step": 5320
    },
    {
      "entropy": 5.578884744644165,
      "epoch": 0.48303701015965167,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004982622793829009,
      "loss": 5.4597,
      "mean_token_accuracy": 0.1753349706530571,
      "num_tokens": 11072353.0,
      "step": 5325
    },
    {
      "entropy": 5.626900243759155,
      "epoch": 0.4834905660377358,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982582634750027,
      "loss": 5.4687,
      "mean_token_accuracy": 0.17250408083200455,
      "num_tokens": 11082655.0,
      "step": 5330
    },
    {
      "entropy": 5.67281928062439,
      "epoch": 0.48394412191582004,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982542429500667,
      "loss": 5.4946,
      "mean_token_accuracy": 0.1740492969751358,
      "num_tokens": 11093315.0,
      "step": 5335
    },
    {
      "entropy": 5.53027229309082,
      "epoch": 0.4843976777939042,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982502178081761,
      "loss": 5.3898,
      "mean_token_accuracy": 0.17748946249485015,
      "num_tokens": 11103074.0,
      "step": 5340
    },
    {
      "entropy": 5.641853141784668,
      "epoch": 0.4848512336719884,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982461880494141,
      "loss": 5.4088,
      "mean_token_accuracy": 0.17445126175880432,
      "num_tokens": 11113636.0,
      "step": 5345
    },
    {
      "entropy": 5.629180288314819,
      "epoch": 0.48530478955007256,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498242153673864,
      "loss": 5.3781,
      "mean_token_accuracy": 0.1742332860827446,
      "num_tokens": 11123509.0,
      "step": 5350
    },
    {
      "entropy": 5.594754266738891,
      "epoch": 0.48575834542815677,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982381146816094,
      "loss": 5.4325,
      "mean_token_accuracy": 0.170786514878273,
      "num_tokens": 11133405.0,
      "step": 5355
    },
    {
      "entropy": 5.694587469100952,
      "epoch": 0.4862119013062409,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982340710727336,
      "loss": 5.5156,
      "mean_token_accuracy": 0.16179958879947662,
      "num_tokens": 11144725.0,
      "step": 5360
    },
    {
      "entropy": 5.644291639328003,
      "epoch": 0.48666545718432513,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982300228473203,
      "loss": 5.4028,
      "mean_token_accuracy": 0.17339923679828645,
      "num_tokens": 11155446.0,
      "step": 5365
    },
    {
      "entropy": 5.61758942604065,
      "epoch": 0.4871190130624093,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982259700054533,
      "loss": 5.4547,
      "mean_token_accuracy": 0.18006922155618668,
      "num_tokens": 11165380.0,
      "step": 5370
    },
    {
      "entropy": 5.630903339385986,
      "epoch": 0.48757256894049344,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982219125472163,
      "loss": 5.5089,
      "mean_token_accuracy": 0.1695322260260582,
      "num_tokens": 11175641.0,
      "step": 5375
    },
    {
      "entropy": 5.7712828636169435,
      "epoch": 0.48802612481857766,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982178504726933,
      "loss": 5.6741,
      "mean_token_accuracy": 0.16982539594173432,
      "num_tokens": 11186257.0,
      "step": 5380
    },
    {
      "entropy": 5.703107261657715,
      "epoch": 0.4884796806966618,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982137837819682,
      "loss": 5.4871,
      "mean_token_accuracy": 0.16374782472848892,
      "num_tokens": 11197284.0,
      "step": 5385
    },
    {
      "entropy": 5.680084371566773,
      "epoch": 0.488933236574746,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982097124751252,
      "loss": 5.4638,
      "mean_token_accuracy": 0.17422798573970794,
      "num_tokens": 11208413.0,
      "step": 5390
    },
    {
      "entropy": 5.596026182174683,
      "epoch": 0.4893867924528302,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982056365522486,
      "loss": 5.465,
      "mean_token_accuracy": 0.17459869980812073,
      "num_tokens": 11219016.0,
      "step": 5395
    },
    {
      "entropy": 5.622074842453003,
      "epoch": 0.4898403483309144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982015560134225,
      "loss": 5.425,
      "mean_token_accuracy": 0.17910900115966796,
      "num_tokens": 11228934.0,
      "step": 5400
    },
    {
      "entropy": 5.652887344360352,
      "epoch": 0.49029390420899854,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004981974708587312,
      "loss": 5.346,
      "mean_token_accuracy": 0.1806335538625717,
      "num_tokens": 11239596.0,
      "step": 5405
    },
    {
      "entropy": 5.567725992202758,
      "epoch": 0.49074746008708275,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981933810882594,
      "loss": 5.342,
      "mean_token_accuracy": 0.17635241597890855,
      "num_tokens": 11249560.0,
      "step": 5410
    },
    {
      "entropy": 5.721574974060059,
      "epoch": 0.4912010159651669,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981892867020917,
      "loss": 5.5447,
      "mean_token_accuracy": 0.1738561399281025,
      "num_tokens": 11259787.0,
      "step": 5415
    },
    {
      "entropy": 5.6990868091583256,
      "epoch": 0.49165457184325106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981851877003125,
      "loss": 5.4694,
      "mean_token_accuracy": 0.17295820415019988,
      "num_tokens": 11269717.0,
      "step": 5420
    },
    {
      "entropy": 5.580885410308838,
      "epoch": 0.4921081277213353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981810840830068,
      "loss": 5.4482,
      "mean_token_accuracy": 0.18043429851531984,
      "num_tokens": 11279863.0,
      "step": 5425
    },
    {
      "entropy": 5.6650306224823,
      "epoch": 0.49256168359941943,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981769758502596,
      "loss": 5.5763,
      "mean_token_accuracy": 0.16078406423330308,
      "num_tokens": 11291161.0,
      "step": 5430
    },
    {
      "entropy": 5.6055395126342775,
      "epoch": 0.49301523947750364,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981728630021555,
      "loss": 5.3956,
      "mean_token_accuracy": 0.17572860419750214,
      "num_tokens": 11301151.0,
      "step": 5435
    },
    {
      "entropy": 5.575085067749024,
      "epoch": 0.4934687953555878,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981687455387797,
      "loss": 5.3946,
      "mean_token_accuracy": 0.1716492608189583,
      "num_tokens": 11311466.0,
      "step": 5440
    },
    {
      "entropy": 5.647791528701783,
      "epoch": 0.493922351233672,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981646234602173,
      "loss": 5.3629,
      "mean_token_accuracy": 0.1757534921169281,
      "num_tokens": 11321522.0,
      "step": 5445
    },
    {
      "entropy": 5.705031681060791,
      "epoch": 0.49437590711175616,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981604967665539,
      "loss": 5.4836,
      "mean_token_accuracy": 0.17398695945739745,
      "num_tokens": 11332317.0,
      "step": 5450
    },
    {
      "entropy": 5.611266326904297,
      "epoch": 0.4948294629898404,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981563654578743,
      "loss": 5.4223,
      "mean_token_accuracy": 0.16848283857107163,
      "num_tokens": 11343024.0,
      "step": 5455
    },
    {
      "entropy": 5.585212087631225,
      "epoch": 0.49528301886792453,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981522295342643,
      "loss": 5.381,
      "mean_token_accuracy": 0.17383768558502197,
      "num_tokens": 11353114.0,
      "step": 5460
    },
    {
      "entropy": 5.634729814529419,
      "epoch": 0.4957365747460087,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981480889958092,
      "loss": 5.4311,
      "mean_token_accuracy": 0.17489740252494812,
      "num_tokens": 11363446.0,
      "step": 5465
    },
    {
      "entropy": 5.55680570602417,
      "epoch": 0.4961901306240929,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981439438425949,
      "loss": 5.4316,
      "mean_token_accuracy": 0.17906904518604277,
      "num_tokens": 11373413.0,
      "step": 5470
    },
    {
      "entropy": 5.541783714294434,
      "epoch": 0.49664368650217705,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498139794074707,
      "loss": 5.3189,
      "mean_token_accuracy": 0.180606409907341,
      "num_tokens": 11383142.0,
      "step": 5475
    },
    {
      "entropy": 5.681115627288818,
      "epoch": 0.49709724238026126,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981356396922313,
      "loss": 5.4639,
      "mean_token_accuracy": 0.17151203900575637,
      "num_tokens": 11394635.0,
      "step": 5480
    },
    {
      "entropy": 5.54714994430542,
      "epoch": 0.4975507982583454,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981314806952535,
      "loss": 5.3762,
      "mean_token_accuracy": 0.17595487385988234,
      "num_tokens": 11404442.0,
      "step": 5485
    },
    {
      "entropy": 5.550731897354126,
      "epoch": 0.49800435413642963,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049812731708386,
      "loss": 5.3204,
      "mean_token_accuracy": 0.18081480413675308,
      "num_tokens": 11414217.0,
      "step": 5490
    },
    {
      "entropy": 5.680741786956787,
      "epoch": 0.4984579100145138,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981231488581367,
      "loss": 5.456,
      "mean_token_accuracy": 0.17206876575946808,
      "num_tokens": 11423834.0,
      "step": 5495
    },
    {
      "entropy": 5.532857179641724,
      "epoch": 0.498911465892598,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981189760181699,
      "loss": 5.3341,
      "mean_token_accuracy": 0.17510698288679122,
      "num_tokens": 11433613.0,
      "step": 5500
    },
    {
      "entropy": 5.559586668014527,
      "epoch": 0.49936502177068215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981147985640456,
      "loss": 5.412,
      "mean_token_accuracy": 0.17702824473381043,
      "num_tokens": 11443915.0,
      "step": 5505
    },
    {
      "entropy": 5.699520635604858,
      "epoch": 0.4998185776487663,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004981106164958506,
      "loss": 5.4326,
      "mean_token_accuracy": 0.17155378311872482,
      "num_tokens": 11454588.0,
      "step": 5510
    },
    {
      "entropy": 5.580562543869019,
      "epoch": 0.5002721335268505,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981064298136712,
      "loss": 5.4901,
      "mean_token_accuracy": 0.17163256406784058,
      "num_tokens": 11465533.0,
      "step": 5515
    },
    {
      "entropy": 5.621740531921387,
      "epoch": 0.5007256894049347,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981022385175939,
      "loss": 5.4445,
      "mean_token_accuracy": 0.1629154846072197,
      "num_tokens": 11475776.0,
      "step": 5520
    },
    {
      "entropy": 5.647769451141357,
      "epoch": 0.5011792452830188,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980980426077054,
      "loss": 5.417,
      "mean_token_accuracy": 0.18146403580904008,
      "num_tokens": 11486200.0,
      "step": 5525
    },
    {
      "entropy": 5.586569929122925,
      "epoch": 0.5016328011611031,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980938420840926,
      "loss": 5.4449,
      "mean_token_accuracy": 0.16791134476661682,
      "num_tokens": 11496768.0,
      "step": 5530
    },
    {
      "entropy": 5.628451347351074,
      "epoch": 0.5020863570391872,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980896369468423,
      "loss": 5.4208,
      "mean_token_accuracy": 0.17588136047124864,
      "num_tokens": 11507096.0,
      "step": 5535
    },
    {
      "entropy": 5.637207078933716,
      "epoch": 0.5025399129172714,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980854271960415,
      "loss": 5.4631,
      "mean_token_accuracy": 0.17342848032712938,
      "num_tokens": 11517301.0,
      "step": 5540
    },
    {
      "entropy": 5.666760635375977,
      "epoch": 0.5029934687953556,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980812128317772,
      "loss": 5.5259,
      "mean_token_accuracy": 0.1743517428636551,
      "num_tokens": 11528616.0,
      "step": 5545
    },
    {
      "entropy": 5.676811027526855,
      "epoch": 0.5034470246734397,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004980769938541364,
      "loss": 5.4487,
      "mean_token_accuracy": 0.17927418798208236,
      "num_tokens": 11539344.0,
      "step": 5550
    },
    {
      "entropy": 5.635175704956055,
      "epoch": 0.503900580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980727702632066,
      "loss": 5.4897,
      "mean_token_accuracy": 0.17068908959627152,
      "num_tokens": 11549511.0,
      "step": 5555
    },
    {
      "entropy": 5.6430299282073975,
      "epoch": 0.5043541364296081,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980685420590752,
      "loss": 5.4369,
      "mean_token_accuracy": 0.17500991076231004,
      "num_tokens": 11559425.0,
      "step": 5560
    },
    {
      "entropy": 5.707560014724732,
      "epoch": 0.5048076923076923,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980643092418294,
      "loss": 5.5405,
      "mean_token_accuracy": 0.16627538055181504,
      "num_tokens": 11570406.0,
      "step": 5565
    },
    {
      "entropy": 5.634152126312256,
      "epoch": 0.5052612481857764,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980600718115568,
      "loss": 5.4295,
      "mean_token_accuracy": 0.175694976747036,
      "num_tokens": 11580509.0,
      "step": 5570
    },
    {
      "entropy": 5.635722541809082,
      "epoch": 0.5057148040638607,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980558297683452,
      "loss": 5.4389,
      "mean_token_accuracy": 0.1730503410100937,
      "num_tokens": 11590860.0,
      "step": 5575
    },
    {
      "entropy": 5.6468902111053465,
      "epoch": 0.5061683599419449,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004980515831122822,
      "loss": 5.4468,
      "mean_token_accuracy": 0.17353652268648148,
      "num_tokens": 11601954.0,
      "step": 5580
    },
    {
      "entropy": 5.652958011627197,
      "epoch": 0.506621915820029,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004980473318434555,
      "loss": 5.4462,
      "mean_token_accuracy": 0.17210652977228164,
      "num_tokens": 11612581.0,
      "step": 5585
    },
    {
      "entropy": 5.573605346679687,
      "epoch": 0.5070754716981132,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004980430759619532,
      "loss": 5.3652,
      "mean_token_accuracy": 0.17937352061271666,
      "num_tokens": 11622599.0,
      "step": 5590
    },
    {
      "entropy": 5.543113279342651,
      "epoch": 0.5075290275761973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980388154678633,
      "loss": 5.3908,
      "mean_token_accuracy": 0.17439862638711928,
      "num_tokens": 11633611.0,
      "step": 5595
    },
    {
      "entropy": 5.585327100753784,
      "epoch": 0.5079825834542816,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980345503612738,
      "loss": 5.365,
      "mean_token_accuracy": 0.17676355540752411,
      "num_tokens": 11643159.0,
      "step": 5600
    },
    {
      "entropy": 5.5928394317626955,
      "epoch": 0.5084361393323658,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980302806422729,
      "loss": 5.3449,
      "mean_token_accuracy": 0.177506685256958,
      "num_tokens": 11653174.0,
      "step": 5605
    },
    {
      "entropy": 5.6118120670318605,
      "epoch": 0.5088896952104499,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980260063109491,
      "loss": 5.4038,
      "mean_token_accuracy": 0.17569431215524672,
      "num_tokens": 11662621.0,
      "step": 5610
    },
    {
      "entropy": 5.528201961517334,
      "epoch": 0.5093432510885341,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004980217273673906,
      "loss": 5.3599,
      "mean_token_accuracy": 0.17619221061468124,
      "num_tokens": 11673538.0,
      "step": 5615
    },
    {
      "entropy": 5.644331693649292,
      "epoch": 0.5097968069666183,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980174438116859,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1794475644826889,
      "num_tokens": 11683203.0,
      "step": 5620
    },
    {
      "entropy": 5.552001285552978,
      "epoch": 0.5102503628447025,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980131556439236,
      "loss": 5.3636,
      "mean_token_accuracy": 0.18283542692661287,
      "num_tokens": 11693313.0,
      "step": 5625
    },
    {
      "entropy": 5.612477016448975,
      "epoch": 0.5107039187227866,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980088628641925,
      "loss": 5.4534,
      "mean_token_accuracy": 0.17441582083702087,
      "num_tokens": 11704194.0,
      "step": 5630
    },
    {
      "entropy": 5.56378436088562,
      "epoch": 0.5111574746008708,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980045654725812,
      "loss": 5.3146,
      "mean_token_accuracy": 0.17968470156192778,
      "num_tokens": 11715109.0,
      "step": 5635
    },
    {
      "entropy": 5.534660005569458,
      "epoch": 0.511611030478955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980002634691787,
      "loss": 5.3489,
      "mean_token_accuracy": 0.18387647569179535,
      "num_tokens": 11724961.0,
      "step": 5640
    },
    {
      "entropy": 5.55992374420166,
      "epoch": 0.5120645863570392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497995956854074,
      "loss": 5.3785,
      "mean_token_accuracy": 0.1765351787209511,
      "num_tokens": 11735334.0,
      "step": 5645
    },
    {
      "entropy": 5.569534349441528,
      "epoch": 0.5125181422351234,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979916456273561,
      "loss": 5.3789,
      "mean_token_accuracy": 0.1730036035180092,
      "num_tokens": 11746858.0,
      "step": 5650
    },
    {
      "entropy": 5.661505699157715,
      "epoch": 0.5129716981132075,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979873297891141,
      "loss": 5.4903,
      "mean_token_accuracy": 0.17127009034156798,
      "num_tokens": 11757819.0,
      "step": 5655
    },
    {
      "entropy": 5.618195486068726,
      "epoch": 0.5134252539912917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979830093394372,
      "loss": 5.4053,
      "mean_token_accuracy": 0.17758016139268876,
      "num_tokens": 11767470.0,
      "step": 5660
    },
    {
      "entropy": 5.549928283691406,
      "epoch": 0.513878809869376,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979786842784149,
      "loss": 5.3273,
      "mean_token_accuracy": 0.1853337362408638,
      "num_tokens": 11777060.0,
      "step": 5665
    },
    {
      "entropy": 5.577957963943481,
      "epoch": 0.5143323657474601,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979743546061366,
      "loss": 5.3962,
      "mean_token_accuracy": 0.17992057055234909,
      "num_tokens": 11787176.0,
      "step": 5670
    },
    {
      "entropy": 5.539264678955078,
      "epoch": 0.5147859216255443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979700203226918,
      "loss": 5.2699,
      "mean_token_accuracy": 0.18505061715841292,
      "num_tokens": 11796419.0,
      "step": 5675
    },
    {
      "entropy": 5.661307001113892,
      "epoch": 0.5152394775036284,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979656814281702,
      "loss": 5.4999,
      "mean_token_accuracy": 0.17273307591676712,
      "num_tokens": 11807279.0,
      "step": 5680
    },
    {
      "entropy": 5.561037874221801,
      "epoch": 0.5156930333817126,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979613379226615,
      "loss": 5.3961,
      "mean_token_accuracy": 0.17279176861047746,
      "num_tokens": 11817669.0,
      "step": 5685
    },
    {
      "entropy": 5.53184027671814,
      "epoch": 0.5161465892597968,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979569898062554,
      "loss": 5.3938,
      "mean_token_accuracy": 0.18203373849391938,
      "num_tokens": 11827247.0,
      "step": 5690
    },
    {
      "entropy": 5.618873500823975,
      "epoch": 0.516600145137881,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497952637079042,
      "loss": 5.3803,
      "mean_token_accuracy": 0.17191717773675919,
      "num_tokens": 11838201.0,
      "step": 5695
    },
    {
      "entropy": 5.568843269348145,
      "epoch": 0.5170537010159652,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979482797411114,
      "loss": 5.322,
      "mean_token_accuracy": 0.18213216811418534,
      "num_tokens": 11847561.0,
      "step": 5700
    },
    {
      "entropy": 5.6255693435668945,
      "epoch": 0.5175072568940493,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979439177925534,
      "loss": 5.4481,
      "mean_token_accuracy": 0.180612213909626,
      "num_tokens": 11858272.0,
      "step": 5705
    },
    {
      "entropy": 5.612488603591919,
      "epoch": 0.5179608127721336,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979395512334584,
      "loss": 5.4823,
      "mean_token_accuracy": 0.17200103104114534,
      "num_tokens": 11869660.0,
      "step": 5710
    },
    {
      "entropy": 5.6795634746551515,
      "epoch": 0.5184143686502177,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979351800639166,
      "loss": 5.5096,
      "mean_token_accuracy": 0.1676321730017662,
      "num_tokens": 11880239.0,
      "step": 5715
    },
    {
      "entropy": 5.712689352035523,
      "epoch": 0.5188679245283019,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979308042840186,
      "loss": 5.4816,
      "mean_token_accuracy": 0.16780400574207305,
      "num_tokens": 11890630.0,
      "step": 5720
    },
    {
      "entropy": 5.669008064270019,
      "epoch": 0.519321480406386,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979264238938547,
      "loss": 5.5112,
      "mean_token_accuracy": 0.1703075349330902,
      "num_tokens": 11900948.0,
      "step": 5725
    },
    {
      "entropy": 5.541024541854858,
      "epoch": 0.5197750362844702,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979220388935155,
      "loss": 5.3925,
      "mean_token_accuracy": 0.17539578080177307,
      "num_tokens": 11912528.0,
      "step": 5730
    },
    {
      "entropy": 5.5685014724731445,
      "epoch": 0.5202285921625545,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979176492830918,
      "loss": 5.353,
      "mean_token_accuracy": 0.17433730214834214,
      "num_tokens": 11922211.0,
      "step": 5735
    },
    {
      "entropy": 5.598578691482544,
      "epoch": 0.5206821480406386,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979132550626742,
      "loss": 5.4084,
      "mean_token_accuracy": 0.17142235189676286,
      "num_tokens": 11932520.0,
      "step": 5740
    },
    {
      "entropy": 5.59015007019043,
      "epoch": 0.5211357039187228,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979088562323538,
      "loss": 5.3221,
      "mean_token_accuracy": 0.18398308157920837,
      "num_tokens": 11943312.0,
      "step": 5745
    },
    {
      "entropy": 5.559282016754151,
      "epoch": 0.5215892597968069,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979044527922214,
      "loss": 5.4661,
      "mean_token_accuracy": 0.17238451689481735,
      "num_tokens": 11954455.0,
      "step": 5750
    },
    {
      "entropy": 5.626161718368531,
      "epoch": 0.5220428156748912,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979000447423681,
      "loss": 5.3677,
      "mean_token_accuracy": 0.18028335571289061,
      "num_tokens": 11964650.0,
      "step": 5755
    },
    {
      "entropy": 5.579872035980225,
      "epoch": 0.5224963715529753,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978956320828851,
      "loss": 5.4013,
      "mean_token_accuracy": 0.17466039061546326,
      "num_tokens": 11975134.0,
      "step": 5760
    },
    {
      "entropy": 5.607253074645996,
      "epoch": 0.5229499274310595,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978912148138637,
      "loss": 5.45,
      "mean_token_accuracy": 0.17419570237398146,
      "num_tokens": 11986486.0,
      "step": 5765
    },
    {
      "entropy": 5.655194902420044,
      "epoch": 0.5234034833091437,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497886792935395,
      "loss": 5.4488,
      "mean_token_accuracy": 0.1717308521270752,
      "num_tokens": 11998045.0,
      "step": 5770
    },
    {
      "entropy": 5.663045930862427,
      "epoch": 0.5238570391872278,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978823664475707,
      "loss": 5.4513,
      "mean_token_accuracy": 0.17971017062664033,
      "num_tokens": 12008130.0,
      "step": 5775
    },
    {
      "entropy": 5.596103620529175,
      "epoch": 0.5243105950653121,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978779353504824,
      "loss": 5.5171,
      "mean_token_accuracy": 0.16514042466878892,
      "num_tokens": 12018753.0,
      "step": 5780
    },
    {
      "entropy": 5.621715688705445,
      "epoch": 0.5247641509433962,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978734996442214,
      "loss": 5.3943,
      "mean_token_accuracy": 0.18264525085687638,
      "num_tokens": 12028202.0,
      "step": 5785
    },
    {
      "entropy": 5.640986490249634,
      "epoch": 0.5252177068214804,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978690593288798,
      "loss": 5.4546,
      "mean_token_accuracy": 0.17733746469020845,
      "num_tokens": 12039652.0,
      "step": 5790
    },
    {
      "entropy": 5.637750244140625,
      "epoch": 0.5256712626995645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978646144045491,
      "loss": 5.3854,
      "mean_token_accuracy": 0.17197972387075425,
      "num_tokens": 12051269.0,
      "step": 5795
    },
    {
      "entropy": 5.482382678985596,
      "epoch": 0.5261248185776488,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978601648713215,
      "loss": 5.2795,
      "mean_token_accuracy": 0.17774615287780762,
      "num_tokens": 12060647.0,
      "step": 5800
    },
    {
      "entropy": 5.569826936721801,
      "epoch": 0.526578374455733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978557107292889,
      "loss": 5.408,
      "mean_token_accuracy": 0.17919009327888488,
      "num_tokens": 12071669.0,
      "step": 5805
    },
    {
      "entropy": 5.547264099121094,
      "epoch": 0.5270319303338171,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978512519785433,
      "loss": 5.282,
      "mean_token_accuracy": 0.18562869876623153,
      "num_tokens": 12081971.0,
      "step": 5810
    },
    {
      "entropy": 5.5526330947875975,
      "epoch": 0.5274854862119013,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497846788619177,
      "loss": 5.4195,
      "mean_token_accuracy": 0.17687285095453262,
      "num_tokens": 12092465.0,
      "step": 5815
    },
    {
      "entropy": 5.617220735549926,
      "epoch": 0.5279390420899854,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978423206512825,
      "loss": 5.3735,
      "mean_token_accuracy": 0.17665890902280806,
      "num_tokens": 12101752.0,
      "step": 5820
    },
    {
      "entropy": 5.661663913726807,
      "epoch": 0.5283925979680697,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497837848074952,
      "loss": 5.456,
      "mean_token_accuracy": 0.17533332258462905,
      "num_tokens": 12111805.0,
      "step": 5825
    },
    {
      "entropy": 5.665722894668579,
      "epoch": 0.5288461538461539,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497833370890278,
      "loss": 5.4739,
      "mean_token_accuracy": 0.17663151323795317,
      "num_tokens": 12122511.0,
      "step": 5830
    },
    {
      "entropy": 5.596258163452148,
      "epoch": 0.529299709724238,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978288890973529,
      "loss": 5.3516,
      "mean_token_accuracy": 0.17615063041448592,
      "num_tokens": 12132510.0,
      "step": 5835
    },
    {
      "entropy": 5.578837966918945,
      "epoch": 0.5297532656023222,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978244026962698,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1739246740937233,
      "num_tokens": 12142419.0,
      "step": 5840
    },
    {
      "entropy": 5.594649934768677,
      "epoch": 0.5302068214804064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978199116871212,
      "loss": 5.4149,
      "mean_token_accuracy": 0.17248719334602355,
      "num_tokens": 12152759.0,
      "step": 5845
    },
    {
      "entropy": 5.554829692840576,
      "epoch": 0.5306603773584906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049781541607,
      "loss": 5.3653,
      "mean_token_accuracy": 0.17687786668539046,
      "num_tokens": 12162758.0,
      "step": 5850
    },
    {
      "entropy": 5.583818531036377,
      "epoch": 0.5311139332365747,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978109158449993,
      "loss": 5.4229,
      "mean_token_accuracy": 0.17397976219654082,
      "num_tokens": 12172664.0,
      "step": 5855
    },
    {
      "entropy": 5.605308771133423,
      "epoch": 0.5315674891146589,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978064110122121,
      "loss": 5.3847,
      "mean_token_accuracy": 0.17937863022089004,
      "num_tokens": 12182028.0,
      "step": 5860
    },
    {
      "entropy": 5.624365615844726,
      "epoch": 0.532021044992743,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978019015717314,
      "loss": 5.4322,
      "mean_token_accuracy": 0.17571227252483368,
      "num_tokens": 12192168.0,
      "step": 5865
    },
    {
      "entropy": 5.526874589920044,
      "epoch": 0.5324746008708273,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977973875236508,
      "loss": 5.3325,
      "mean_token_accuracy": 0.17963843196630477,
      "num_tokens": 12202889.0,
      "step": 5870
    },
    {
      "entropy": 5.54812593460083,
      "epoch": 0.5329281567489115,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977928688680634,
      "loss": 5.3732,
      "mean_token_accuracy": 0.18322223126888276,
      "num_tokens": 12212514.0,
      "step": 5875
    },
    {
      "entropy": 5.584365177154541,
      "epoch": 0.5333817126269956,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977883456050626,
      "loss": 5.247,
      "mean_token_accuracy": 0.1861517012119293,
      "num_tokens": 12222353.0,
      "step": 5880
    },
    {
      "entropy": 5.6489053726196286,
      "epoch": 0.5338352685050798,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497783817734742,
      "loss": 5.5219,
      "mean_token_accuracy": 0.16939357966184615,
      "num_tokens": 12232845.0,
      "step": 5885
    },
    {
      "entropy": 5.61934814453125,
      "epoch": 0.534288824383164,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977792852571954,
      "loss": 5.45,
      "mean_token_accuracy": 0.16973866671323776,
      "num_tokens": 12242950.0,
      "step": 5890
    },
    {
      "entropy": 5.633284950256348,
      "epoch": 0.5347423802612482,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977747481725164,
      "loss": 5.4321,
      "mean_token_accuracy": 0.1735300526022911,
      "num_tokens": 12253132.0,
      "step": 5895
    },
    {
      "entropy": 5.599405288696289,
      "epoch": 0.5351959361393324,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977702064807987,
      "loss": 5.3785,
      "mean_token_accuracy": 0.17767370939254762,
      "num_tokens": 12263214.0,
      "step": 5900
    },
    {
      "entropy": 5.526389789581299,
      "epoch": 0.5356494920174165,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977656601821366,
      "loss": 5.3095,
      "mean_token_accuracy": 0.18471042364835738,
      "num_tokens": 12273250.0,
      "step": 5905
    },
    {
      "entropy": 5.716156530380249,
      "epoch": 0.5361030478955007,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977611092766238,
      "loss": 5.4942,
      "mean_token_accuracy": 0.1808107778429985,
      "num_tokens": 12283767.0,
      "step": 5910
    },
    {
      "entropy": 5.523726463317871,
      "epoch": 0.5365566037735849,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977565537643544,
      "loss": 5.3421,
      "mean_token_accuracy": 0.17266201227903366,
      "num_tokens": 12294612.0,
      "step": 5915
    },
    {
      "entropy": 5.554253149032593,
      "epoch": 0.5370101596516691,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977519936454227,
      "loss": 5.4225,
      "mean_token_accuracy": 0.18037798404693603,
      "num_tokens": 12305614.0,
      "step": 5920
    },
    {
      "entropy": 5.550849533081054,
      "epoch": 0.5374637155297532,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977474289199231,
      "loss": 5.2966,
      "mean_token_accuracy": 0.17565731406211854,
      "num_tokens": 12316152.0,
      "step": 5925
    },
    {
      "entropy": 5.522589588165284,
      "epoch": 0.5379172714078374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977428595879499,
      "loss": 5.3751,
      "mean_token_accuracy": 0.17792995423078536,
      "num_tokens": 12325940.0,
      "step": 5930
    },
    {
      "entropy": 5.570098352432251,
      "epoch": 0.5383708272859217,
      "grad_norm": 1.125,
      "learning_rate": 0.0004977382856495975,
      "loss": 5.3606,
      "mean_token_accuracy": 0.17704337388277053,
      "num_tokens": 12335482.0,
      "step": 5935
    },
    {
      "entropy": 5.6550994396209715,
      "epoch": 0.5388243831640058,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977337071049606,
      "loss": 5.4564,
      "mean_token_accuracy": 0.17251523584127426,
      "num_tokens": 12346386.0,
      "step": 5940
    },
    {
      "entropy": 5.668819808959961,
      "epoch": 0.53927793904209,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497729123954134,
      "loss": 5.4197,
      "mean_token_accuracy": 0.1771680533885956,
      "num_tokens": 12356282.0,
      "step": 5945
    },
    {
      "entropy": 5.580939483642578,
      "epoch": 0.5397314949201741,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977245361972123,
      "loss": 5.3956,
      "mean_token_accuracy": 0.17961778491735458,
      "num_tokens": 12366994.0,
      "step": 5950
    },
    {
      "entropy": 5.550912141799927,
      "epoch": 0.5401850507982583,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977199438342904,
      "loss": 5.2834,
      "mean_token_accuracy": 0.1829832002520561,
      "num_tokens": 12377524.0,
      "step": 5955
    },
    {
      "entropy": 5.6025049686431885,
      "epoch": 0.5406386066763426,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977153468654634,
      "loss": 5.4502,
      "mean_token_accuracy": 0.18007041662931442,
      "num_tokens": 12387593.0,
      "step": 5960
    },
    {
      "entropy": 5.677999258041382,
      "epoch": 0.5410921625544267,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004977107452908262,
      "loss": 5.4079,
      "mean_token_accuracy": 0.18105369359254836,
      "num_tokens": 12398732.0,
      "step": 5965
    },
    {
      "entropy": 5.5877197265625,
      "epoch": 0.5415457184325109,
      "grad_norm": 1.0,
      "learning_rate": 0.000497706139110474,
      "loss": 5.405,
      "mean_token_accuracy": 0.18087859451770782,
      "num_tokens": 12409821.0,
      "step": 5970
    },
    {
      "entropy": 5.5568990230560305,
      "epoch": 0.541999274310595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977015283245021,
      "loss": 5.4217,
      "mean_token_accuracy": 0.17138040512800218,
      "num_tokens": 12420044.0,
      "step": 5975
    },
    {
      "entropy": 5.519027042388916,
      "epoch": 0.5424528301886793,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976969129330059,
      "loss": 5.3454,
      "mean_token_accuracy": 0.1738667070865631,
      "num_tokens": 12430980.0,
      "step": 5980
    },
    {
      "entropy": 5.6030182361602785,
      "epoch": 0.5429063860667634,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004976922929360808,
      "loss": 5.3948,
      "mean_token_accuracy": 0.18232907652854918,
      "num_tokens": 12441999.0,
      "step": 5985
    },
    {
      "entropy": 5.621866083145141,
      "epoch": 0.5433599419448476,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976876683338222,
      "loss": 5.4362,
      "mean_token_accuracy": 0.17455211877822877,
      "num_tokens": 12452837.0,
      "step": 5990
    },
    {
      "entropy": 5.572230195999145,
      "epoch": 0.5438134978229318,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497683039126326,
      "loss": 5.3547,
      "mean_token_accuracy": 0.18271825313568116,
      "num_tokens": 12462142.0,
      "step": 5995
    },
    {
      "entropy": 5.600415372848511,
      "epoch": 0.5442670537010159,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976784053136878,
      "loss": 5.4043,
      "mean_token_accuracy": 0.17925617694854737,
      "num_tokens": 12472756.0,
      "step": 6000
    },
    {
      "epoch": 0.5442670537010159,
      "eval_entropy": 5.379152549122254,
      "eval_loss": 5.41491174697876,
      "eval_mean_token_accuracy": 0.18314881672433733,
      "eval_num_tokens": 12472756.0,
      "eval_runtime": 20.6957,
      "eval_samples_per_second": 1708.571,
      "eval_steps_per_second": 213.571,
      "step": 6000
    },
    {
      "entropy": 5.644152069091797,
      "epoch": 0.5447206095791002,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976737668960034,
      "loss": 5.5029,
      "mean_token_accuracy": 0.17155803442001344,
      "num_tokens": 12483115.0,
      "step": 6005
    },
    {
      "entropy": 5.652310562133789,
      "epoch": 0.5451741654571843,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976691238733688,
      "loss": 5.3646,
      "mean_token_accuracy": 0.17381614148616792,
      "num_tokens": 12493280.0,
      "step": 6010
    },
    {
      "entropy": 5.596336126327515,
      "epoch": 0.5456277213352685,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049766447624588,
      "loss": 5.5069,
      "mean_token_accuracy": 0.1748632788658142,
      "num_tokens": 12503665.0,
      "step": 6015
    },
    {
      "entropy": 5.5472839832305905,
      "epoch": 0.5460812772133526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497659824013633,
      "loss": 5.4058,
      "mean_token_accuracy": 0.17200566828250885,
      "num_tokens": 12513381.0,
      "step": 6020
    },
    {
      "entropy": 5.680693578720093,
      "epoch": 0.5465348330914369,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976551671767241,
      "loss": 5.4393,
      "mean_token_accuracy": 0.17311064749956132,
      "num_tokens": 12524121.0,
      "step": 6025
    },
    {
      "entropy": 5.601163005828857,
      "epoch": 0.5469883889695211,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976505057352496,
      "loss": 5.4147,
      "mean_token_accuracy": 0.18124686181545258,
      "num_tokens": 12533972.0,
      "step": 6030
    },
    {
      "entropy": 5.644373226165771,
      "epoch": 0.5474419448476052,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976458396893059,
      "loss": 5.5213,
      "mean_token_accuracy": 0.16715265363454818,
      "num_tokens": 12544110.0,
      "step": 6035
    },
    {
      "entropy": 5.5955469608306885,
      "epoch": 0.5478955007256894,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976411690389896,
      "loss": 5.3403,
      "mean_token_accuracy": 0.17307349890470505,
      "num_tokens": 12554273.0,
      "step": 6040
    },
    {
      "entropy": 5.661379146575928,
      "epoch": 0.5483490566037735,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497636493784397,
      "loss": 5.4215,
      "mean_token_accuracy": 0.17500090450048447,
      "num_tokens": 12564633.0,
      "step": 6045
    },
    {
      "entropy": 5.586629152297974,
      "epoch": 0.5488026124818578,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976318139256251,
      "loss": 5.3537,
      "mean_token_accuracy": 0.178970268368721,
      "num_tokens": 12573790.0,
      "step": 6050
    },
    {
      "entropy": 5.6261004447937015,
      "epoch": 0.549256168359942,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976271294627706,
      "loss": 5.4317,
      "mean_token_accuracy": 0.17064266204833983,
      "num_tokens": 12584265.0,
      "step": 6055
    },
    {
      "entropy": 5.570170545578003,
      "epoch": 0.5497097242380261,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976224403959303,
      "loss": 5.2914,
      "mean_token_accuracy": 0.1787312373518944,
      "num_tokens": 12594413.0,
      "step": 6060
    },
    {
      "entropy": 5.510761976242065,
      "epoch": 0.5501632801161103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976177467252012,
      "loss": 5.3683,
      "mean_token_accuracy": 0.18546044677495957,
      "num_tokens": 12604637.0,
      "step": 6065
    },
    {
      "entropy": 5.589938545227051,
      "epoch": 0.5506168359941945,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976130484506804,
      "loss": 5.4391,
      "mean_token_accuracy": 0.1787260204553604,
      "num_tokens": 12614873.0,
      "step": 6070
    },
    {
      "entropy": 5.601373863220215,
      "epoch": 0.5510703918722787,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976083455724649,
      "loss": 5.3712,
      "mean_token_accuracy": 0.1769331619143486,
      "num_tokens": 12626182.0,
      "step": 6075
    },
    {
      "entropy": 5.540160465240478,
      "epoch": 0.5515239477503628,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976036380906523,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1818635180592537,
      "num_tokens": 12635418.0,
      "step": 6080
    },
    {
      "entropy": 5.5531182289123535,
      "epoch": 0.551977503628447,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975989260053395,
      "loss": 5.4516,
      "mean_token_accuracy": 0.16748463958501816,
      "num_tokens": 12645942.0,
      "step": 6085
    },
    {
      "entropy": 5.6305876731872555,
      "epoch": 0.5524310595065312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975942093166242,
      "loss": 5.3756,
      "mean_token_accuracy": 0.17721232026815414,
      "num_tokens": 12656460.0,
      "step": 6090
    },
    {
      "entropy": 5.610075330734253,
      "epoch": 0.5528846153846154,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975894880246041,
      "loss": 5.3241,
      "mean_token_accuracy": 0.17855470925569533,
      "num_tokens": 12666916.0,
      "step": 6095
    },
    {
      "entropy": 5.539130449295044,
      "epoch": 0.5533381712626996,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975847621293766,
      "loss": 5.3682,
      "mean_token_accuracy": 0.18464813530445098,
      "num_tokens": 12677562.0,
      "step": 6100
    },
    {
      "entropy": 5.5453942775726315,
      "epoch": 0.5537917271407837,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975800316310394,
      "loss": 5.3443,
      "mean_token_accuracy": 0.1776389956474304,
      "num_tokens": 12687349.0,
      "step": 6105
    },
    {
      "entropy": 5.601655769348144,
      "epoch": 0.5542452830188679,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975752965296906,
      "loss": 5.4005,
      "mean_token_accuracy": 0.17570678889751434,
      "num_tokens": 12697850.0,
      "step": 6110
    },
    {
      "entropy": 5.50960373878479,
      "epoch": 0.5546988388969522,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975705568254278,
      "loss": 5.2554,
      "mean_token_accuracy": 0.1816887453198433,
      "num_tokens": 12708283.0,
      "step": 6115
    },
    {
      "entropy": 5.553288125991822,
      "epoch": 0.5551523947750363,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975658125183493,
      "loss": 5.3604,
      "mean_token_accuracy": 0.1766258791089058,
      "num_tokens": 12718201.0,
      "step": 6120
    },
    {
      "entropy": 5.5428098201751705,
      "epoch": 0.5556059506531205,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497561063608553,
      "loss": 5.3093,
      "mean_token_accuracy": 0.18314361572265625,
      "num_tokens": 12728860.0,
      "step": 6125
    },
    {
      "entropy": 5.565022611618042,
      "epoch": 0.5560595065312046,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004975563100961372,
      "loss": 5.3704,
      "mean_token_accuracy": 0.17591018676757814,
      "num_tokens": 12739007.0,
      "step": 6130
    },
    {
      "entropy": 5.546350717544556,
      "epoch": 0.5565130624092888,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004975515519812002,
      "loss": 5.2309,
      "mean_token_accuracy": 0.1811644896864891,
      "num_tokens": 12749004.0,
      "step": 6135
    },
    {
      "entropy": 5.555451583862305,
      "epoch": 0.556966618287373,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975467892638402,
      "loss": 5.3761,
      "mean_token_accuracy": 0.1802150845527649,
      "num_tokens": 12760085.0,
      "step": 6140
    },
    {
      "entropy": 5.568343782424927,
      "epoch": 0.5574201741654572,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975420219441561,
      "loss": 5.3294,
      "mean_token_accuracy": 0.18599051386117935,
      "num_tokens": 12770303.0,
      "step": 6145
    },
    {
      "entropy": 5.574814367294311,
      "epoch": 0.5578737300435413,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975372500222462,
      "loss": 5.3613,
      "mean_token_accuracy": 0.18204200863838196,
      "num_tokens": 12780334.0,
      "step": 6150
    },
    {
      "entropy": 5.568981695175171,
      "epoch": 0.5583272859216255,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975324734982094,
      "loss": 5.4068,
      "mean_token_accuracy": 0.1729448139667511,
      "num_tokens": 12790980.0,
      "step": 6155
    },
    {
      "entropy": 5.61474723815918,
      "epoch": 0.5587808417997098,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975276923721442,
      "loss": 5.4477,
      "mean_token_accuracy": 0.17725956439971924,
      "num_tokens": 12801713.0,
      "step": 6160
    },
    {
      "entropy": 5.613516807556152,
      "epoch": 0.5592343976777939,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975229066441496,
      "loss": 5.4632,
      "mean_token_accuracy": 0.17113297879695893,
      "num_tokens": 12813040.0,
      "step": 6165
    },
    {
      "entropy": 5.557818269729614,
      "epoch": 0.5596879535558781,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975181163143247,
      "loss": 5.3897,
      "mean_token_accuracy": 0.17339551895856858,
      "num_tokens": 12824048.0,
      "step": 6170
    },
    {
      "entropy": 5.607717227935791,
      "epoch": 0.5601415094339622,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975133213827684,
      "loss": 5.3425,
      "mean_token_accuracy": 0.17274201065301895,
      "num_tokens": 12834297.0,
      "step": 6175
    },
    {
      "entropy": 5.6809532165527346,
      "epoch": 0.5605950653120464,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975085218495798,
      "loss": 5.4172,
      "mean_token_accuracy": 0.17106309533119202,
      "num_tokens": 12844574.0,
      "step": 6180
    },
    {
      "entropy": 5.524085903167725,
      "epoch": 0.5610486211901307,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975037177148585,
      "loss": 5.3589,
      "mean_token_accuracy": 0.1753137394785881,
      "num_tokens": 12854670.0,
      "step": 6185
    },
    {
      "entropy": 5.51950306892395,
      "epoch": 0.5615021770682148,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974989089787035,
      "loss": 5.3713,
      "mean_token_accuracy": 0.18194560408592225,
      "num_tokens": 12865281.0,
      "step": 6190
    },
    {
      "entropy": 5.6557981967926025,
      "epoch": 0.561955732946299,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974940956412143,
      "loss": 5.3725,
      "mean_token_accuracy": 0.18128142803907393,
      "num_tokens": 12875222.0,
      "step": 6195
    },
    {
      "entropy": 5.714055490493775,
      "epoch": 0.5624092888243831,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974892777024905,
      "loss": 5.4478,
      "mean_token_accuracy": 0.1658184453845024,
      "num_tokens": 12885079.0,
      "step": 6200
    },
    {
      "entropy": 5.553147268295288,
      "epoch": 0.5628628447024674,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004974844551626318,
      "loss": 5.3775,
      "mean_token_accuracy": 0.18091828525066375,
      "num_tokens": 12896402.0,
      "step": 6205
    },
    {
      "entropy": 5.556194925308228,
      "epoch": 0.5633164005805515,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974796280217379,
      "loss": 5.3496,
      "mean_token_accuracy": 0.18283513188362122,
      "num_tokens": 12906403.0,
      "step": 6210
    },
    {
      "entropy": 5.598826265335083,
      "epoch": 0.5637699564586357,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974747962799086,
      "loss": 5.3687,
      "mean_token_accuracy": 0.1756142258644104,
      "num_tokens": 12917061.0,
      "step": 6215
    },
    {
      "entropy": 5.568496322631836,
      "epoch": 0.5642235123367199,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974699599372437,
      "loss": 5.3911,
      "mean_token_accuracy": 0.1795428067445755,
      "num_tokens": 12927182.0,
      "step": 6220
    },
    {
      "entropy": 5.529246139526367,
      "epoch": 0.564677068214804,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974651189938434,
      "loss": 5.361,
      "mean_token_accuracy": 0.17926921993494033,
      "num_tokens": 12937371.0,
      "step": 6225
    },
    {
      "entropy": 5.573128747940063,
      "epoch": 0.5651306240928883,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974602734498079,
      "loss": 5.4119,
      "mean_token_accuracy": 0.1759198173880577,
      "num_tokens": 12948569.0,
      "step": 6230
    },
    {
      "entropy": 5.610108995437622,
      "epoch": 0.5655841799709724,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000497455423305237,
      "loss": 5.3696,
      "mean_token_accuracy": 0.1703425258398056,
      "num_tokens": 12958503.0,
      "step": 6235
    },
    {
      "entropy": 5.612152004241944,
      "epoch": 0.5660377358490566,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974505685602314,
      "loss": 5.3857,
      "mean_token_accuracy": 0.1823065236210823,
      "num_tokens": 12968714.0,
      "step": 6240
    },
    {
      "entropy": 5.555819892883301,
      "epoch": 0.5664912917271407,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974457092148912,
      "loss": 5.4102,
      "mean_token_accuracy": 0.17348227500915528,
      "num_tokens": 12978857.0,
      "step": 6245
    },
    {
      "entropy": 5.670070648193359,
      "epoch": 0.566944847605225,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974408452693172,
      "loss": 5.5099,
      "mean_token_accuracy": 0.17125375419855118,
      "num_tokens": 12989863.0,
      "step": 6250
    },
    {
      "entropy": 5.5948718070983885,
      "epoch": 0.5673984034833092,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974359767236099,
      "loss": 5.3118,
      "mean_token_accuracy": 0.17757333517074586,
      "num_tokens": 13000307.0,
      "step": 6255
    },
    {
      "entropy": 5.592511749267578,
      "epoch": 0.5678519593613933,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004974311035778698,
      "loss": 5.3883,
      "mean_token_accuracy": 0.18470816314220428,
      "num_tokens": 13009053.0,
      "step": 6260
    },
    {
      "entropy": 5.578414249420166,
      "epoch": 0.5683055152394775,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974262258321979,
      "loss": 5.4205,
      "mean_token_accuracy": 0.172482830286026,
      "num_tokens": 13019136.0,
      "step": 6265
    },
    {
      "entropy": 5.6188208103179935,
      "epoch": 0.5687590711175616,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004974213434866949,
      "loss": 5.4601,
      "mean_token_accuracy": 0.17041391432285308,
      "num_tokens": 13030356.0,
      "step": 6270
    },
    {
      "entropy": 5.613703393936158,
      "epoch": 0.5692126269956459,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974164565414617,
      "loss": 5.4759,
      "mean_token_accuracy": 0.17600406408309938,
      "num_tokens": 13040477.0,
      "step": 6275
    },
    {
      "entropy": 5.604294204711914,
      "epoch": 0.56966618287373,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974115649965998,
      "loss": 5.3422,
      "mean_token_accuracy": 0.18165747225284576,
      "num_tokens": 13051755.0,
      "step": 6280
    },
    {
      "entropy": 5.5737049102783205,
      "epoch": 0.5701197387518142,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974066688522098,
      "loss": 5.3627,
      "mean_token_accuracy": 0.1862332984805107,
      "num_tokens": 13061670.0,
      "step": 6285
    },
    {
      "entropy": 5.556718826293945,
      "epoch": 0.5705732946298984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974017681083934,
      "loss": 5.3469,
      "mean_token_accuracy": 0.18105981051921843,
      "num_tokens": 13070917.0,
      "step": 6290
    },
    {
      "entropy": 5.554030179977417,
      "epoch": 0.5710268505079826,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004973968627652516,
      "loss": 5.3933,
      "mean_token_accuracy": 0.17703206092119217,
      "num_tokens": 13080577.0,
      "step": 6295
    },
    {
      "entropy": 5.614200782775879,
      "epoch": 0.5714804063860668,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497391952822886,
      "loss": 5.4314,
      "mean_token_accuracy": 0.17582101076841355,
      "num_tokens": 13091446.0,
      "step": 6300
    },
    {
      "entropy": 5.623776531219482,
      "epoch": 0.5719339622641509,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973870382813983,
      "loss": 5.4603,
      "mean_token_accuracy": 0.17539089173078537,
      "num_tokens": 13102752.0,
      "step": 6305
    },
    {
      "entropy": 5.665589141845703,
      "epoch": 0.5723875181422351,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973821191408899,
      "loss": 5.465,
      "mean_token_accuracy": 0.17193345874547958,
      "num_tokens": 13113944.0,
      "step": 6310
    },
    {
      "entropy": 5.614248418807984,
      "epoch": 0.5728410740203193,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973771954014626,
      "loss": 5.399,
      "mean_token_accuracy": 0.17831117212772368,
      "num_tokens": 13124316.0,
      "step": 6315
    },
    {
      "entropy": 5.65424370765686,
      "epoch": 0.5732946298984035,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004973722670632182,
      "loss": 5.4797,
      "mean_token_accuracy": 0.17098383456468583,
      "num_tokens": 13133567.0,
      "step": 6320
    },
    {
      "entropy": 5.5662015914917,
      "epoch": 0.5737481857764877,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973673341262587,
      "loss": 5.3917,
      "mean_token_accuracy": 0.1759439915418625,
      "num_tokens": 13144279.0,
      "step": 6325
    },
    {
      "entropy": 5.604485321044922,
      "epoch": 0.5742017416545718,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497362396590686,
      "loss": 5.4699,
      "mean_token_accuracy": 0.17275095134973525,
      "num_tokens": 13154463.0,
      "step": 6330
    },
    {
      "entropy": 5.647492170333862,
      "epoch": 0.574655297532656,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973574544566024,
      "loss": 5.4281,
      "mean_token_accuracy": 0.16383579671382903,
      "num_tokens": 13165565.0,
      "step": 6335
    },
    {
      "entropy": 5.578436040878296,
      "epoch": 0.5751088534107403,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973525077241098,
      "loss": 5.3417,
      "mean_token_accuracy": 0.18116174638271332,
      "num_tokens": 13175839.0,
      "step": 6340
    },
    {
      "entropy": 5.509803342819214,
      "epoch": 0.5755624092888244,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973475563933109,
      "loss": 5.3268,
      "mean_token_accuracy": 0.17926134169101715,
      "num_tokens": 13186372.0,
      "step": 6345
    },
    {
      "entropy": 5.569689559936523,
      "epoch": 0.5760159651669086,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973426004643077,
      "loss": 5.3894,
      "mean_token_accuracy": 0.1798923596739769,
      "num_tokens": 13196259.0,
      "step": 6350
    },
    {
      "entropy": 5.578006315231323,
      "epoch": 0.5764695210449927,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973376399372029,
      "loss": 5.3612,
      "mean_token_accuracy": 0.1823232352733612,
      "num_tokens": 13206497.0,
      "step": 6355
    },
    {
      "entropy": 5.582531070709228,
      "epoch": 0.5769230769230769,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973326748120991,
      "loss": 5.4179,
      "mean_token_accuracy": 0.17883315533399582,
      "num_tokens": 13217188.0,
      "step": 6360
    },
    {
      "entropy": 5.578738641738892,
      "epoch": 0.5773766328011611,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973277050890988,
      "loss": 5.4042,
      "mean_token_accuracy": 0.17636267542839051,
      "num_tokens": 13226544.0,
      "step": 6365
    },
    {
      "entropy": 5.636392402648926,
      "epoch": 0.5778301886792453,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973227307683049,
      "loss": 5.4464,
      "mean_token_accuracy": 0.17221278697252274,
      "num_tokens": 13236572.0,
      "step": 6370
    },
    {
      "entropy": 5.705385446548462,
      "epoch": 0.5782837445573294,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973177518498204,
      "loss": 5.4333,
      "mean_token_accuracy": 0.1773147389292717,
      "num_tokens": 13247958.0,
      "step": 6375
    },
    {
      "entropy": 5.586960983276367,
      "epoch": 0.5787373004354136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973127683337482,
      "loss": 5.342,
      "mean_token_accuracy": 0.17918977737426758,
      "num_tokens": 13258255.0,
      "step": 6380
    },
    {
      "entropy": 5.52049298286438,
      "epoch": 0.5791908563134979,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973077802201911,
      "loss": 5.3718,
      "mean_token_accuracy": 0.17743289172649385,
      "num_tokens": 13268416.0,
      "step": 6385
    },
    {
      "entropy": 5.5609814643859865,
      "epoch": 0.579644412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973027875092526,
      "loss": 5.346,
      "mean_token_accuracy": 0.17577312141656876,
      "num_tokens": 13278087.0,
      "step": 6390
    },
    {
      "entropy": 5.551668643951416,
      "epoch": 0.5800979680696662,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972977902010357,
      "loss": 5.2501,
      "mean_token_accuracy": 0.18471986949443817,
      "num_tokens": 13288121.0,
      "step": 6395
    },
    {
      "entropy": 5.666141223907471,
      "epoch": 0.5805515239477503,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972927882956439,
      "loss": 5.4785,
      "mean_token_accuracy": 0.16452504992485045,
      "num_tokens": 13298620.0,
      "step": 6400
    },
    {
      "entropy": 5.551081323623658,
      "epoch": 0.5810050798258345,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972877817931807,
      "loss": 5.2865,
      "mean_token_accuracy": 0.17660645842552186,
      "num_tokens": 13308413.0,
      "step": 6405
    },
    {
      "entropy": 5.522010469436646,
      "epoch": 0.5814586357039188,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972827706937493,
      "loss": 5.3758,
      "mean_token_accuracy": 0.17947073578834533,
      "num_tokens": 13319387.0,
      "step": 6410
    },
    {
      "entropy": 5.620809888839721,
      "epoch": 0.5819121915820029,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972777549974537,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17694490402936935,
      "num_tokens": 13330494.0,
      "step": 6415
    },
    {
      "entropy": 5.556739234924317,
      "epoch": 0.5823657474600871,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972727347043975,
      "loss": 5.3374,
      "mean_token_accuracy": 0.17858588695526123,
      "num_tokens": 13341009.0,
      "step": 6420
    },
    {
      "entropy": 5.526559066772461,
      "epoch": 0.5828193033381712,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972677098146845,
      "loss": 5.2915,
      "mean_token_accuracy": 0.18030916303396224,
      "num_tokens": 13351269.0,
      "step": 6425
    },
    {
      "entropy": 5.536026811599731,
      "epoch": 0.5832728592162555,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972626803284185,
      "loss": 5.4053,
      "mean_token_accuracy": 0.17801292687654496,
      "num_tokens": 13362160.0,
      "step": 6430
    },
    {
      "entropy": 5.641278648376465,
      "epoch": 0.5837264150943396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972576462457038,
      "loss": 5.4702,
      "mean_token_accuracy": 0.17617461383342742,
      "num_tokens": 13372497.0,
      "step": 6435
    },
    {
      "entropy": 5.673011207580567,
      "epoch": 0.5841799709724238,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972526075666442,
      "loss": 5.3986,
      "mean_token_accuracy": 0.1806288555264473,
      "num_tokens": 13381665.0,
      "step": 6440
    },
    {
      "entropy": 5.523349046707153,
      "epoch": 0.584633526850508,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972475642913442,
      "loss": 5.3026,
      "mean_token_accuracy": 0.1852041155099869,
      "num_tokens": 13392507.0,
      "step": 6445
    },
    {
      "entropy": 5.5107386112213135,
      "epoch": 0.5850870827285921,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972425164199078,
      "loss": 5.2899,
      "mean_token_accuracy": 0.18328872472047805,
      "num_tokens": 13402171.0,
      "step": 6450
    },
    {
      "entropy": 5.651026678085327,
      "epoch": 0.5855406386066764,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972374639524396,
      "loss": 5.5147,
      "mean_token_accuracy": 0.1725245714187622,
      "num_tokens": 13412933.0,
      "step": 6455
    },
    {
      "entropy": 5.612742567062378,
      "epoch": 0.5859941944847605,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000497232406889044,
      "loss": 5.3512,
      "mean_token_accuracy": 0.1728028178215027,
      "num_tokens": 13423563.0,
      "step": 6460
    },
    {
      "entropy": 5.638546085357666,
      "epoch": 0.5864477503628447,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972273452298255,
      "loss": 5.36,
      "mean_token_accuracy": 0.18153707385063172,
      "num_tokens": 13433998.0,
      "step": 6465
    },
    {
      "entropy": 5.532230949401855,
      "epoch": 0.5869013062409288,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972222789748888,
      "loss": 5.3105,
      "mean_token_accuracy": 0.18037805259227752,
      "num_tokens": 13443878.0,
      "step": 6470
    },
    {
      "entropy": 5.536468267440796,
      "epoch": 0.5873548621190131,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004972172081243389,
      "loss": 5.3923,
      "mean_token_accuracy": 0.17271317690610885,
      "num_tokens": 13456051.0,
      "step": 6475
    },
    {
      "entropy": 5.576761245727539,
      "epoch": 0.5878084179970973,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972121326782806,
      "loss": 5.3829,
      "mean_token_accuracy": 0.18032484352588654,
      "num_tokens": 13467276.0,
      "step": 6480
    },
    {
      "entropy": 5.621661853790283,
      "epoch": 0.5882619738751814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972070526368186,
      "loss": 5.3517,
      "mean_token_accuracy": 0.17507115453481675,
      "num_tokens": 13478331.0,
      "step": 6485
    },
    {
      "entropy": 5.605046653747559,
      "epoch": 0.5887155297532656,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972019680000582,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18173345625400544,
      "num_tokens": 13489879.0,
      "step": 6490
    },
    {
      "entropy": 5.548620986938476,
      "epoch": 0.5891690856313497,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971968787681044,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1835034653544426,
      "num_tokens": 13500879.0,
      "step": 6495
    },
    {
      "entropy": 5.66513090133667,
      "epoch": 0.589622641509434,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971917849410627,
      "loss": 5.3983,
      "mean_token_accuracy": 0.1771375998854637,
      "num_tokens": 13511719.0,
      "step": 6500
    },
    {
      "entropy": 5.545218276977539,
      "epoch": 0.5900761973875182,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971866865190381,
      "loss": 5.3364,
      "mean_token_accuracy": 0.18135370761156083,
      "num_tokens": 13521850.0,
      "step": 6505
    },
    {
      "entropy": 5.597626638412476,
      "epoch": 0.5905297532656023,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004971815835021362,
      "loss": 5.3715,
      "mean_token_accuracy": 0.1825823336839676,
      "num_tokens": 13532095.0,
      "step": 6510
    },
    {
      "entropy": 5.5470359325408936,
      "epoch": 0.5909833091436865,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971764758904626,
      "loss": 5.3513,
      "mean_token_accuracy": 0.17569283097982408,
      "num_tokens": 13543133.0,
      "step": 6515
    },
    {
      "entropy": 5.5723918914794925,
      "epoch": 0.5914368650217707,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971713636841229,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1692360073328018,
      "num_tokens": 13553771.0,
      "step": 6520
    },
    {
      "entropy": 5.536196327209472,
      "epoch": 0.5918904208998549,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971662468832228,
      "loss": 5.3591,
      "mean_token_accuracy": 0.18100160360336304,
      "num_tokens": 13564548.0,
      "step": 6525
    },
    {
      "entropy": 5.500099849700928,
      "epoch": 0.592343976777939,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497161125487868,
      "loss": 5.343,
      "mean_token_accuracy": 0.17980441451072693,
      "num_tokens": 13574567.0,
      "step": 6530
    },
    {
      "entropy": 5.531034755706787,
      "epoch": 0.5927975326560232,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971559994981645,
      "loss": 5.3804,
      "mean_token_accuracy": 0.1790984570980072,
      "num_tokens": 13585302.0,
      "step": 6535
    },
    {
      "entropy": 5.616375589370728,
      "epoch": 0.5932510885341074,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971508689142184,
      "loss": 5.4028,
      "mean_token_accuracy": 0.1724355548620224,
      "num_tokens": 13594754.0,
      "step": 6540
    },
    {
      "entropy": 5.548051548004151,
      "epoch": 0.5937046444121916,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971457337361356,
      "loss": 5.3406,
      "mean_token_accuracy": 0.17818880081176758,
      "num_tokens": 13604691.0,
      "step": 6545
    },
    {
      "entropy": 5.455195903778076,
      "epoch": 0.5941582002902758,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971405939640225,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18700708597898483,
      "num_tokens": 13613911.0,
      "step": 6550
    },
    {
      "entropy": 5.672217941284179,
      "epoch": 0.5946117561683599,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971354495979854,
      "loss": 5.456,
      "mean_token_accuracy": 0.17892757952213287,
      "num_tokens": 13624285.0,
      "step": 6555
    },
    {
      "entropy": 5.586663722991943,
      "epoch": 0.5950653120464441,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971303006381305,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1812276229262352,
      "num_tokens": 13635434.0,
      "step": 6560
    },
    {
      "entropy": 5.5121111392974855,
      "epoch": 0.5955188679245284,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971251470845643,
      "loss": 5.361,
      "mean_token_accuracy": 0.17556140720844268,
      "num_tokens": 13645582.0,
      "step": 6565
    },
    {
      "entropy": 5.546522235870361,
      "epoch": 0.5959724238026125,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004971199889373935,
      "loss": 5.3174,
      "mean_token_accuracy": 0.17810338735580444,
      "num_tokens": 13655029.0,
      "step": 6570
    },
    {
      "entropy": 5.550983810424805,
      "epoch": 0.5964259796806967,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971148261967249,
      "loss": 5.309,
      "mean_token_accuracy": 0.1805609330534935,
      "num_tokens": 13665204.0,
      "step": 6575
    },
    {
      "entropy": 5.571804332733154,
      "epoch": 0.5968795355587808,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971096588626649,
      "loss": 5.4215,
      "mean_token_accuracy": 0.17430134862661362,
      "num_tokens": 13676281.0,
      "step": 6580
    },
    {
      "entropy": 5.537905597686768,
      "epoch": 0.597333091436865,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004971044869353206,
      "loss": 5.3613,
      "mean_token_accuracy": 0.18184176832437515,
      "num_tokens": 13687144.0,
      "step": 6585
    },
    {
      "entropy": 5.562548398971558,
      "epoch": 0.5977866473149492,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497099310414799,
      "loss": 5.3553,
      "mean_token_accuracy": 0.18140233010053636,
      "num_tokens": 13696371.0,
      "step": 6590
    },
    {
      "entropy": 5.626462841033936,
      "epoch": 0.5982402031930334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970941293012069,
      "loss": 5.3001,
      "mean_token_accuracy": 0.1769891545176506,
      "num_tokens": 13706900.0,
      "step": 6595
    },
    {
      "entropy": 5.530804586410523,
      "epoch": 0.5986937590711175,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970889435946518,
      "loss": 5.3671,
      "mean_token_accuracy": 0.1814924567937851,
      "num_tokens": 13717874.0,
      "step": 6600
    },
    {
      "entropy": 5.539168167114258,
      "epoch": 0.5991473149492017,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970837532952406,
      "loss": 5.3369,
      "mean_token_accuracy": 0.18162619471549987,
      "num_tokens": 13728020.0,
      "step": 6605
    },
    {
      "entropy": 5.58751859664917,
      "epoch": 0.599600870827286,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970785584030808,
      "loss": 5.3918,
      "mean_token_accuracy": 0.17638065218925475,
      "num_tokens": 13738007.0,
      "step": 6610
    },
    {
      "entropy": 5.566587495803833,
      "epoch": 0.6000544267053701,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970733589182799,
      "loss": 5.3877,
      "mean_token_accuracy": 0.17494157403707505,
      "num_tokens": 13749429.0,
      "step": 6615
    },
    {
      "entropy": 5.558034038543701,
      "epoch": 0.6005079825834543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970681548409452,
      "loss": 5.262,
      "mean_token_accuracy": 0.18570734858512877,
      "num_tokens": 13760279.0,
      "step": 6620
    },
    {
      "entropy": 5.600715589523316,
      "epoch": 0.6009615384615384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970629461711846,
      "loss": 5.4823,
      "mean_token_accuracy": 0.17011196166276932,
      "num_tokens": 13770575.0,
      "step": 6625
    },
    {
      "entropy": 5.498355531692505,
      "epoch": 0.6014150943396226,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970577329091057,
      "loss": 5.3063,
      "mean_token_accuracy": 0.1857804089784622,
      "num_tokens": 13781068.0,
      "step": 6630
    },
    {
      "entropy": 5.5711182117462155,
      "epoch": 0.6018686502177069,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970525150548162,
      "loss": 5.3957,
      "mean_token_accuracy": 0.18031402826309204,
      "num_tokens": 13790745.0,
      "step": 6635
    },
    {
      "entropy": 5.585378217697143,
      "epoch": 0.602322206095791,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970472926084242,
      "loss": 5.3303,
      "mean_token_accuracy": 0.17876851856708526,
      "num_tokens": 13800196.0,
      "step": 6640
    },
    {
      "entropy": 5.521693229675293,
      "epoch": 0.6027757619738752,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970420655700375,
      "loss": 5.2459,
      "mean_token_accuracy": 0.1828948974609375,
      "num_tokens": 13809954.0,
      "step": 6645
    },
    {
      "entropy": 5.5161940574646,
      "epoch": 0.6032293178519593,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970368339397643,
      "loss": 5.4225,
      "mean_token_accuracy": 0.1718314230442047,
      "num_tokens": 13820281.0,
      "step": 6650
    },
    {
      "entropy": 5.634420919418335,
      "epoch": 0.6036828737300436,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004970315977177129,
      "loss": 5.3772,
      "mean_token_accuracy": 0.17908216267824173,
      "num_tokens": 13830614.0,
      "step": 6655
    },
    {
      "entropy": 5.5539734840393065,
      "epoch": 0.6041364296081277,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004970263569039914,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18047489523887633,
      "num_tokens": 13840551.0,
      "step": 6660
    },
    {
      "entropy": 5.62788462638855,
      "epoch": 0.6045899854862119,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970211114987083,
      "loss": 5.4994,
      "mean_token_accuracy": 0.1773001953959465,
      "num_tokens": 13851830.0,
      "step": 6665
    },
    {
      "entropy": 5.5775432109832765,
      "epoch": 0.6050435413642961,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970158615019719,
      "loss": 5.2978,
      "mean_token_accuracy": 0.18300541788339614,
      "num_tokens": 13862299.0,
      "step": 6670
    },
    {
      "entropy": 5.607674980163575,
      "epoch": 0.6054970972423802,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970106069138912,
      "loss": 5.4325,
      "mean_token_accuracy": 0.1705495685338974,
      "num_tokens": 13872852.0,
      "step": 6675
    },
    {
      "entropy": 5.530221462249756,
      "epoch": 0.6059506531204645,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004970053477345743,
      "loss": 5.3784,
      "mean_token_accuracy": 0.18458971232175828,
      "num_tokens": 13883601.0,
      "step": 6680
    },
    {
      "entropy": 5.578758192062378,
      "epoch": 0.6064042089985486,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004970000839641304,
      "loss": 5.3524,
      "mean_token_accuracy": 0.18701042830944062,
      "num_tokens": 13894246.0,
      "step": 6685
    },
    {
      "entropy": 5.521432018280029,
      "epoch": 0.6068577648766328,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969948156026681,
      "loss": 5.3868,
      "mean_token_accuracy": 0.1802896812558174,
      "num_tokens": 13904823.0,
      "step": 6690
    },
    {
      "entropy": 5.647825717926025,
      "epoch": 0.6073113207547169,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969895426502965,
      "loss": 5.403,
      "mean_token_accuracy": 0.1710766464471817,
      "num_tokens": 13915338.0,
      "step": 6695
    },
    {
      "entropy": 5.546239614486694,
      "epoch": 0.6077648766328012,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969842651071246,
      "loss": 5.3611,
      "mean_token_accuracy": 0.17356563061475755,
      "num_tokens": 13925838.0,
      "step": 6700
    },
    {
      "entropy": 5.606841611862182,
      "epoch": 0.6082184325108854,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969789829732615,
      "loss": 5.3653,
      "mean_token_accuracy": 0.17559958398342132,
      "num_tokens": 13936341.0,
      "step": 6705
    },
    {
      "entropy": 5.611944770812988,
      "epoch": 0.6086719883889695,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969736962488164,
      "loss": 5.3672,
      "mean_token_accuracy": 0.1796737715601921,
      "num_tokens": 13945438.0,
      "step": 6710
    },
    {
      "entropy": 5.558967113494873,
      "epoch": 0.6091255442670537,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004969684049338987,
      "loss": 5.2584,
      "mean_token_accuracy": 0.18215785771608353,
      "num_tokens": 13956029.0,
      "step": 6715
    },
    {
      "entropy": 5.596481418609619,
      "epoch": 0.6095791001451378,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496963109028618,
      "loss": 5.3783,
      "mean_token_accuracy": 0.1746341571211815,
      "num_tokens": 13966498.0,
      "step": 6720
    },
    {
      "entropy": 5.53505163192749,
      "epoch": 0.6100326560232221,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969578085330834,
      "loss": 5.3298,
      "mean_token_accuracy": 0.18196677714586257,
      "num_tokens": 13976997.0,
      "step": 6725
    },
    {
      "entropy": 5.615462064743042,
      "epoch": 0.6104862119013063,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969525034474049,
      "loss": 5.3307,
      "mean_token_accuracy": 0.17902466505765915,
      "num_tokens": 13986664.0,
      "step": 6730
    },
    {
      "entropy": 5.498200178146362,
      "epoch": 0.6109397677793904,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496947193771692,
      "loss": 5.3209,
      "mean_token_accuracy": 0.18532713502645493,
      "num_tokens": 13997828.0,
      "step": 6735
    },
    {
      "entropy": 5.508232593536377,
      "epoch": 0.6113933236574746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969418795060548,
      "loss": 5.2761,
      "mean_token_accuracy": 0.17970932871103287,
      "num_tokens": 14008293.0,
      "step": 6740
    },
    {
      "entropy": 5.507388973236084,
      "epoch": 0.6118468795355588,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969365606506027,
      "loss": 5.2646,
      "mean_token_accuracy": 0.18298047930002212,
      "num_tokens": 14018031.0,
      "step": 6745
    },
    {
      "entropy": 5.491404008865357,
      "epoch": 0.612300435413643,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969312372054461,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1883564054965973,
      "num_tokens": 14028013.0,
      "step": 6750
    },
    {
      "entropy": 5.524714374542237,
      "epoch": 0.6127539912917271,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004969259091706948,
      "loss": 5.3723,
      "mean_token_accuracy": 0.17930554449558259,
      "num_tokens": 14039787.0,
      "step": 6755
    },
    {
      "entropy": 5.5531907081604,
      "epoch": 0.6132075471698113,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004969205765464593,
      "loss": 5.2734,
      "mean_token_accuracy": 0.18047384172677994,
      "num_tokens": 14049223.0,
      "step": 6760
    },
    {
      "entropy": 5.597138452529907,
      "epoch": 0.6136611030478955,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969152393328497,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1850910186767578,
      "num_tokens": 14059004.0,
      "step": 6765
    },
    {
      "entropy": 5.561625194549561,
      "epoch": 0.6141146589259797,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969098975299764,
      "loss": 5.4108,
      "mean_token_accuracy": 0.1744755983352661,
      "num_tokens": 14069615.0,
      "step": 6770
    },
    {
      "entropy": 5.550989723205566,
      "epoch": 0.6145682148040639,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969045511379499,
      "loss": 5.3455,
      "mean_token_accuracy": 0.17937582731246948,
      "num_tokens": 14078886.0,
      "step": 6775
    },
    {
      "entropy": 5.646713638305664,
      "epoch": 0.615021770682148,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004968992001568806,
      "loss": 5.4293,
      "mean_token_accuracy": 0.1722983568906784,
      "num_tokens": 14089021.0,
      "step": 6780
    },
    {
      "entropy": 5.55102949142456,
      "epoch": 0.6154753265602322,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968938445868795,
      "loss": 5.235,
      "mean_token_accuracy": 0.19036040008068084,
      "num_tokens": 14098316.0,
      "step": 6785
    },
    {
      "entropy": 5.502194881439209,
      "epoch": 0.6159288824383164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968884844280569,
      "loss": 5.3202,
      "mean_token_accuracy": 0.1710915148258209,
      "num_tokens": 14108874.0,
      "step": 6790
    },
    {
      "entropy": 5.5451150894165036,
      "epoch": 0.6163824383164006,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004968831196805242,
      "loss": 5.3731,
      "mean_token_accuracy": 0.1814247488975525,
      "num_tokens": 14119613.0,
      "step": 6795
    },
    {
      "entropy": 5.577469968795777,
      "epoch": 0.6168359941944848,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968777503443918,
      "loss": 5.3335,
      "mean_token_accuracy": 0.1749727502465248,
      "num_tokens": 14129085.0,
      "step": 6800
    },
    {
      "entropy": 5.4897855758667,
      "epoch": 0.6172895500725689,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968723764197711,
      "loss": 5.2292,
      "mean_token_accuracy": 0.19681656509637832,
      "num_tokens": 14138461.0,
      "step": 6805
    },
    {
      "entropy": 5.5140174388885494,
      "epoch": 0.6177431059506531,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496866997906773,
      "loss": 5.2926,
      "mean_token_accuracy": 0.18786997944116593,
      "num_tokens": 14149126.0,
      "step": 6810
    },
    {
      "entropy": 5.499987840652466,
      "epoch": 0.6181966618287373,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004968616148055089,
      "loss": 5.306,
      "mean_token_accuracy": 0.1850553944706917,
      "num_tokens": 14160068.0,
      "step": 6815
    },
    {
      "entropy": 5.559629440307617,
      "epoch": 0.6186502177068215,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004968562271160901,
      "loss": 5.3981,
      "mean_token_accuracy": 0.18361826986074448,
      "num_tokens": 14170288.0,
      "step": 6820
    },
    {
      "entropy": 5.63637866973877,
      "epoch": 0.6191037735849056,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496850834838628,
      "loss": 5.3922,
      "mean_token_accuracy": 0.17993116676807402,
      "num_tokens": 14181231.0,
      "step": 6825
    },
    {
      "entropy": 5.592602682113648,
      "epoch": 0.6195573294629898,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004968454379732342,
      "loss": 5.294,
      "mean_token_accuracy": 0.18232411295175552,
      "num_tokens": 14191715.0,
      "step": 6830
    },
    {
      "entropy": 5.52168231010437,
      "epoch": 0.6200108853410741,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049684003652002,
      "loss": 5.2998,
      "mean_token_accuracy": 0.17813656479120255,
      "num_tokens": 14203416.0,
      "step": 6835
    },
    {
      "entropy": 5.446109342575073,
      "epoch": 0.6204644412191582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968346304790973,
      "loss": 5.2524,
      "mean_token_accuracy": 0.19115934818983077,
      "num_tokens": 14213658.0,
      "step": 6840
    },
    {
      "entropy": 5.593048906326294,
      "epoch": 0.6209179970972424,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968292198505781,
      "loss": 5.4235,
      "mean_token_accuracy": 0.17608387172222137,
      "num_tokens": 14223956.0,
      "step": 6845
    },
    {
      "entropy": 5.618917846679688,
      "epoch": 0.6213715529753265,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000496823804634574,
      "loss": 5.4129,
      "mean_token_accuracy": 0.17932555377483367,
      "num_tokens": 14236801.0,
      "step": 6850
    },
    {
      "entropy": 5.529294729232788,
      "epoch": 0.6218251088534107,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968183848311972,
      "loss": 5.2253,
      "mean_token_accuracy": 0.1876459464430809,
      "num_tokens": 14247382.0,
      "step": 6855
    },
    {
      "entropy": 5.560374402999878,
      "epoch": 0.622278664731495,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968129604405596,
      "loss": 5.3169,
      "mean_token_accuracy": 0.18302041888237,
      "num_tokens": 14258627.0,
      "step": 6860
    },
    {
      "entropy": 5.464026880264282,
      "epoch": 0.6227322206095791,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968075314627734,
      "loss": 5.3135,
      "mean_token_accuracy": 0.1861542820930481,
      "num_tokens": 14269013.0,
      "step": 6865
    },
    {
      "entropy": 5.536890077590942,
      "epoch": 0.6231857764876633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496802097897951,
      "loss": 5.3071,
      "mean_token_accuracy": 0.18000025004148484,
      "num_tokens": 14279620.0,
      "step": 6870
    },
    {
      "entropy": 5.5682155132293705,
      "epoch": 0.6236393323657474,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004967966597462046,
      "loss": 5.2997,
      "mean_token_accuracy": 0.18479396849870683,
      "num_tokens": 14289448.0,
      "step": 6875
    },
    {
      "entropy": 5.561263513565064,
      "epoch": 0.6240928882438317,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967912170076469,
      "loss": 5.2893,
      "mean_token_accuracy": 0.17907041907310486,
      "num_tokens": 14299999.0,
      "step": 6880
    },
    {
      "entropy": 5.501129531860352,
      "epoch": 0.6245464441219158,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004967857696823901,
      "loss": 5.2876,
      "mean_token_accuracy": 0.18401148319244384,
      "num_tokens": 14310384.0,
      "step": 6885
    },
    {
      "entropy": 5.550896310806275,
      "epoch": 0.625,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967803177705472,
      "loss": 5.4017,
      "mean_token_accuracy": 0.17491451203823088,
      "num_tokens": 14320644.0,
      "step": 6890
    },
    {
      "entropy": 5.607715702056884,
      "epoch": 0.6254535558780842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004967748612722308,
      "loss": 5.3716,
      "mean_token_accuracy": 0.18048068434000014,
      "num_tokens": 14330077.0,
      "step": 6895
    },
    {
      "entropy": 5.613873863220215,
      "epoch": 0.6259071117561683,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967694001875538,
      "loss": 5.4525,
      "mean_token_accuracy": 0.17149625718593597,
      "num_tokens": 14341549.0,
      "step": 6900
    },
    {
      "entropy": 5.585393762588501,
      "epoch": 0.6263606676342526,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496763934516629,
      "loss": 5.3118,
      "mean_token_accuracy": 0.18692418187856674,
      "num_tokens": 14352146.0,
      "step": 6905
    },
    {
      "entropy": 5.495177698135376,
      "epoch": 0.6268142235123367,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004967584642595696,
      "loss": 5.3372,
      "mean_token_accuracy": 0.1760642945766449,
      "num_tokens": 14362521.0,
      "step": 6910
    },
    {
      "entropy": 5.564804315567017,
      "epoch": 0.6272677793904209,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004967529894164887,
      "loss": 5.3685,
      "mean_token_accuracy": 0.1765128716826439,
      "num_tokens": 14373216.0,
      "step": 6915
    },
    {
      "entropy": 5.5402006149292,
      "epoch": 0.627721335268505,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967475099874992,
      "loss": 5.2748,
      "mean_token_accuracy": 0.18698276132345198,
      "num_tokens": 14383104.0,
      "step": 6920
    },
    {
      "entropy": 5.488488435745239,
      "epoch": 0.6281748911465893,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496742025972715,
      "loss": 5.31,
      "mean_token_accuracy": 0.1918911427259445,
      "num_tokens": 14393854.0,
      "step": 6925
    },
    {
      "entropy": 5.564506006240845,
      "epoch": 0.6286284470246735,
      "grad_norm": 1.046875,
      "learning_rate": 0.000496736537372249,
      "loss": 5.2489,
      "mean_token_accuracy": 0.18805742710828782,
      "num_tokens": 14403213.0,
      "step": 6930
    },
    {
      "entropy": 5.641831588745117,
      "epoch": 0.6290820029027576,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004967310441862151,
      "loss": 5.478,
      "mean_token_accuracy": 0.17123783379793167,
      "num_tokens": 14412792.0,
      "step": 6935
    },
    {
      "entropy": 5.584077024459839,
      "epoch": 0.6295355587808418,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967255464147265,
      "loss": 5.3367,
      "mean_token_accuracy": 0.18502951860427858,
      "num_tokens": 14423061.0,
      "step": 6940
    },
    {
      "entropy": 5.563900518417358,
      "epoch": 0.6299891146589259,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967200440578972,
      "loss": 5.2726,
      "mean_token_accuracy": 0.1876469150185585,
      "num_tokens": 14433338.0,
      "step": 6945
    },
    {
      "entropy": 5.481202793121338,
      "epoch": 0.6304426705370102,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967145371158408,
      "loss": 5.2585,
      "mean_token_accuracy": 0.19061126410961152,
      "num_tokens": 14444124.0,
      "step": 6950
    },
    {
      "entropy": 5.514369153976441,
      "epoch": 0.6308962264150944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967090255886714,
      "loss": 5.3887,
      "mean_token_accuracy": 0.18010223805904388,
      "num_tokens": 14453868.0,
      "step": 6955
    },
    {
      "entropy": 5.5535437107086185,
      "epoch": 0.6313497822931785,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004967035094765028,
      "loss": 5.2781,
      "mean_token_accuracy": 0.17700921893119811,
      "num_tokens": 14465015.0,
      "step": 6960
    },
    {
      "entropy": 5.504481458663941,
      "epoch": 0.6318033381712627,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966979887794491,
      "loss": 5.2372,
      "mean_token_accuracy": 0.19093358367681504,
      "num_tokens": 14475523.0,
      "step": 6965
    },
    {
      "entropy": 5.507241201400757,
      "epoch": 0.6322568940493469,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966924634976245,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18237321078777313,
      "num_tokens": 14485523.0,
      "step": 6970
    },
    {
      "entropy": 5.608642530441284,
      "epoch": 0.6327104499274311,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966869336311433,
      "loss": 5.371,
      "mean_token_accuracy": 0.1829627588391304,
      "num_tokens": 14495237.0,
      "step": 6975
    },
    {
      "entropy": 5.584758949279785,
      "epoch": 0.6331640058055152,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966813991801197,
      "loss": 5.5108,
      "mean_token_accuracy": 0.17335500121116637,
      "num_tokens": 14506431.0,
      "step": 6980
    },
    {
      "entropy": 5.615596199035645,
      "epoch": 0.6336175616835994,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966758601446684,
      "loss": 5.4274,
      "mean_token_accuracy": 0.17262813448905945,
      "num_tokens": 14517445.0,
      "step": 6985
    },
    {
      "entropy": 5.597373199462891,
      "epoch": 0.6340711175616836,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966703165249038,
      "loss": 5.3624,
      "mean_token_accuracy": 0.17349915504455565,
      "num_tokens": 14527818.0,
      "step": 6990
    },
    {
      "entropy": 5.553152656555175,
      "epoch": 0.6345246734397678,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966647683209406,
      "loss": 5.4443,
      "mean_token_accuracy": 0.16868529468774796,
      "num_tokens": 14538235.0,
      "step": 6995
    },
    {
      "entropy": 5.693959903717041,
      "epoch": 0.634978229317852,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966592155328935,
      "loss": 5.4539,
      "mean_token_accuracy": 0.1710901975631714,
      "num_tokens": 14549784.0,
      "step": 7000
    },
    {
      "entropy": 5.5767427444458,
      "epoch": 0.6354317851959361,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004966536581608771,
      "loss": 5.4413,
      "mean_token_accuracy": 0.16910024285316466,
      "num_tokens": 14561057.0,
      "step": 7005
    },
    {
      "entropy": 5.555035877227783,
      "epoch": 0.6358853410740203,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966480962050069,
      "loss": 5.3529,
      "mean_token_accuracy": 0.1803895801305771,
      "num_tokens": 14570859.0,
      "step": 7010
    },
    {
      "entropy": 5.552107143402099,
      "epoch": 0.6363388969521045,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966425296653974,
      "loss": 5.353,
      "mean_token_accuracy": 0.1830982819199562,
      "num_tokens": 14581066.0,
      "step": 7015
    },
    {
      "entropy": 5.559674692153931,
      "epoch": 0.6367924528301887,
      "grad_norm": 0.90625,
      "learning_rate": 0.000496636958542164,
      "loss": 5.3239,
      "mean_token_accuracy": 0.17896756231784822,
      "num_tokens": 14590709.0,
      "step": 7020
    },
    {
      "entropy": 5.556455469131469,
      "epoch": 0.6372460087082729,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004966313828354218,
      "loss": 5.3089,
      "mean_token_accuracy": 0.18295130729675294,
      "num_tokens": 14600849.0,
      "step": 7025
    },
    {
      "entropy": 5.481048727035523,
      "epoch": 0.637699564586357,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496625802545286,
      "loss": 5.1954,
      "mean_token_accuracy": 0.18557194024324417,
      "num_tokens": 14611427.0,
      "step": 7030
    },
    {
      "entropy": 5.520011472702026,
      "epoch": 0.6381531204644412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966202176718723,
      "loss": 5.3651,
      "mean_token_accuracy": 0.18566489666700364,
      "num_tokens": 14623860.0,
      "step": 7035
    },
    {
      "entropy": 5.556991815567017,
      "epoch": 0.6386066763425254,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004966146282152959,
      "loss": 5.2855,
      "mean_token_accuracy": 0.1784844383597374,
      "num_tokens": 14633481.0,
      "step": 7040
    },
    {
      "entropy": 5.5632165431976315,
      "epoch": 0.6390602322206096,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966090341756726,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19155942499637604,
      "num_tokens": 14643519.0,
      "step": 7045
    },
    {
      "entropy": 5.432686424255371,
      "epoch": 0.6395137880986937,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496603435553118,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1885633021593094,
      "num_tokens": 14654570.0,
      "step": 7050
    },
    {
      "entropy": 5.48898401260376,
      "epoch": 0.6399673439767779,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965978323477478,
      "loss": 5.3028,
      "mean_token_accuracy": 0.1765861302614212,
      "num_tokens": 14665596.0,
      "step": 7055
    },
    {
      "entropy": 5.582192134857178,
      "epoch": 0.6404208998548622,
      "grad_norm": 1.03125,
      "learning_rate": 0.000496592224559678,
      "loss": 5.2517,
      "mean_token_accuracy": 0.18795465528964997,
      "num_tokens": 14675660.0,
      "step": 7060
    },
    {
      "entropy": 5.563787317276001,
      "epoch": 0.6408744557329463,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965866121890245,
      "loss": 5.4095,
      "mean_token_accuracy": 0.18029894530773163,
      "num_tokens": 14685229.0,
      "step": 7065
    },
    {
      "entropy": 5.557628488540649,
      "epoch": 0.6413280116110305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965809952359034,
      "loss": 5.3122,
      "mean_token_accuracy": 0.17769250571727752,
      "num_tokens": 14695672.0,
      "step": 7070
    },
    {
      "entropy": 5.514695119857788,
      "epoch": 0.6417815674891146,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965753737004309,
      "loss": 5.2615,
      "mean_token_accuracy": 0.18418668806552888,
      "num_tokens": 14705160.0,
      "step": 7075
    },
    {
      "entropy": 5.52557144165039,
      "epoch": 0.6422351233671988,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965697475827231,
      "loss": 5.3097,
      "mean_token_accuracy": 0.18284096121788024,
      "num_tokens": 14715487.0,
      "step": 7080
    },
    {
      "entropy": 5.577736806869507,
      "epoch": 0.6426886792452831,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965641168828965,
      "loss": 5.3544,
      "mean_token_accuracy": 0.18357906490564346,
      "num_tokens": 14726987.0,
      "step": 7085
    },
    {
      "entropy": 5.4941871643066404,
      "epoch": 0.6431422351233672,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965584816010676,
      "loss": 5.3026,
      "mean_token_accuracy": 0.18404889553785325,
      "num_tokens": 14737810.0,
      "step": 7090
    },
    {
      "entropy": 5.579785442352295,
      "epoch": 0.6435957910014514,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965528417373528,
      "loss": 5.3193,
      "mean_token_accuracy": 0.18094680607318878,
      "num_tokens": 14748997.0,
      "step": 7095
    },
    {
      "entropy": 5.553784418106079,
      "epoch": 0.6440493468795355,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965471972918688,
      "loss": 5.3318,
      "mean_token_accuracy": 0.17817866504192353,
      "num_tokens": 14760439.0,
      "step": 7100
    },
    {
      "entropy": 5.445217418670654,
      "epoch": 0.6445029027576198,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004965415482647322,
      "loss": 5.2362,
      "mean_token_accuracy": 0.18102772533893585,
      "num_tokens": 14770828.0,
      "step": 7105
    },
    {
      "entropy": 5.490187788009644,
      "epoch": 0.6449564586357039,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049653589465606,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18126174360513686,
      "num_tokens": 14781753.0,
      "step": 7110
    },
    {
      "entropy": 5.619928598403931,
      "epoch": 0.6454100145137881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000496530236465969,
      "loss": 5.3987,
      "mean_token_accuracy": 0.1711141899228096,
      "num_tokens": 14792241.0,
      "step": 7115
    },
    {
      "entropy": 5.504394292831421,
      "epoch": 0.6458635703918723,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004965245736945763,
      "loss": 5.2172,
      "mean_token_accuracy": 0.1904732972383499,
      "num_tokens": 14802580.0,
      "step": 7120
    },
    {
      "entropy": 5.5241124629974365,
      "epoch": 0.6463171262699564,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965189063419989,
      "loss": 5.4283,
      "mean_token_accuracy": 0.17418646961450576,
      "num_tokens": 14813103.0,
      "step": 7125
    },
    {
      "entropy": 5.571353006362915,
      "epoch": 0.6467706821480407,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965132344083541,
      "loss": 5.3258,
      "mean_token_accuracy": 0.17736517637968063,
      "num_tokens": 14823334.0,
      "step": 7130
    },
    {
      "entropy": 5.56951904296875,
      "epoch": 0.6472242380261248,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965075578937593,
      "loss": 5.336,
      "mean_token_accuracy": 0.17852421551942826,
      "num_tokens": 14834850.0,
      "step": 7135
    },
    {
      "entropy": 5.485752820968628,
      "epoch": 0.647677793904209,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965018767983316,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18051422238349915,
      "num_tokens": 14844276.0,
      "step": 7140
    },
    {
      "entropy": 5.440044736862182,
      "epoch": 0.6481313497822931,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964961911221886,
      "loss": 5.1867,
      "mean_token_accuracy": 0.19364167600870133,
      "num_tokens": 14854352.0,
      "step": 7145
    },
    {
      "entropy": 5.579373216629028,
      "epoch": 0.6485849056603774,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496490500865448,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18060466051101684,
      "num_tokens": 14864395.0,
      "step": 7150
    },
    {
      "entropy": 5.513666248321533,
      "epoch": 0.6490384615384616,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004964848060282275,
      "loss": 5.2513,
      "mean_token_accuracy": 0.18520902544260026,
      "num_tokens": 14874373.0,
      "step": 7155
    },
    {
      "entropy": 5.515260982513428,
      "epoch": 0.6494920174165457,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964791066106447,
      "loss": 5.2496,
      "mean_token_accuracy": 0.17809188216924668,
      "num_tokens": 14884932.0,
      "step": 7160
    },
    {
      "entropy": 5.572841691970825,
      "epoch": 0.6499455732946299,
      "grad_norm": 0.875,
      "learning_rate": 0.0004964734026128175,
      "loss": 5.3991,
      "mean_token_accuracy": 0.17553634494543074,
      "num_tokens": 14896254.0,
      "step": 7165
    },
    {
      "entropy": 5.601190614700317,
      "epoch": 0.650399129172714,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964676940348638,
      "loss": 5.3319,
      "mean_token_accuracy": 0.17750634700059892,
      "num_tokens": 14907392.0,
      "step": 7170
    },
    {
      "entropy": 5.573651218414307,
      "epoch": 0.6508526850507983,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000496461980876902,
      "loss": 5.4395,
      "mean_token_accuracy": 0.17827751934528352,
      "num_tokens": 14917676.0,
      "step": 7175
    },
    {
      "entropy": 5.549517297744751,
      "epoch": 0.6513062409288825,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964562631390498,
      "loss": 5.2987,
      "mean_token_accuracy": 0.18335486352443695,
      "num_tokens": 14928500.0,
      "step": 7180
    },
    {
      "entropy": 5.538741636276245,
      "epoch": 0.6517597968069666,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964505408214257,
      "loss": 5.2492,
      "mean_token_accuracy": 0.18961041122674943,
      "num_tokens": 14938970.0,
      "step": 7185
    },
    {
      "entropy": 5.511117887496948,
      "epoch": 0.6522133526850508,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000496444813924148,
      "loss": 5.311,
      "mean_token_accuracy": 0.17935390323400496,
      "num_tokens": 14949017.0,
      "step": 7190
    },
    {
      "entropy": 5.51752781867981,
      "epoch": 0.652666908563135,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004964390824473351,
      "loss": 5.2404,
      "mean_token_accuracy": 0.18981554210186005,
      "num_tokens": 14958529.0,
      "step": 7195
    },
    {
      "entropy": 5.509324216842652,
      "epoch": 0.6531204644412192,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964333463911056,
      "loss": 5.3481,
      "mean_token_accuracy": 0.1777355194091797,
      "num_tokens": 14968044.0,
      "step": 7200
    },
    {
      "entropy": 5.569104433059692,
      "epoch": 0.6535740203193033,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964276057555779,
      "loss": 5.365,
      "mean_token_accuracy": 0.17823416888713836,
      "num_tokens": 14977867.0,
      "step": 7205
    },
    {
      "entropy": 5.6319859504699705,
      "epoch": 0.6540275761973875,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496421860540871,
      "loss": 5.3719,
      "mean_token_accuracy": 0.1820246011018753,
      "num_tokens": 14986936.0,
      "step": 7210
    },
    {
      "entropy": 5.530765676498413,
      "epoch": 0.6544811320754716,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964161107471036,
      "loss": 5.331,
      "mean_token_accuracy": 0.18487488329410554,
      "num_tokens": 14996681.0,
      "step": 7215
    },
    {
      "entropy": 5.552242612838745,
      "epoch": 0.6549346879535559,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964103563743946,
      "loss": 5.3224,
      "mean_token_accuracy": 0.1786444827914238,
      "num_tokens": 15007100.0,
      "step": 7220
    },
    {
      "entropy": 5.599308681488037,
      "epoch": 0.6553882438316401,
      "grad_norm": 1.0,
      "learning_rate": 0.000496404597422863,
      "loss": 5.3806,
      "mean_token_accuracy": 0.18596629649400712,
      "num_tokens": 15017452.0,
      "step": 7225
    },
    {
      "entropy": 5.516591310501099,
      "epoch": 0.6558417997097242,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496398833892628,
      "loss": 5.2531,
      "mean_token_accuracy": 0.18542855978012085,
      "num_tokens": 15028362.0,
      "step": 7230
    },
    {
      "entropy": 5.581466817855835,
      "epoch": 0.6562953555878084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963930657838086,
      "loss": 5.3469,
      "mean_token_accuracy": 0.17714816331863403,
      "num_tokens": 15038691.0,
      "step": 7235
    },
    {
      "entropy": 5.552061080932617,
      "epoch": 0.6567489114658926,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004963872930965242,
      "loss": 5.3117,
      "mean_token_accuracy": 0.1878270536661148,
      "num_tokens": 15050037.0,
      "step": 7240
    },
    {
      "entropy": 5.552956056594849,
      "epoch": 0.6572024673439768,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963815158308941,
      "loss": 5.2747,
      "mean_token_accuracy": 0.18535771667957307,
      "num_tokens": 15059793.0,
      "step": 7245
    },
    {
      "entropy": 5.5456095218658445,
      "epoch": 0.657656023222061,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004963757339870379,
      "loss": 5.3749,
      "mean_token_accuracy": 0.17560529112815856,
      "num_tokens": 15070500.0,
      "step": 7250
    },
    {
      "entropy": 5.579104375839234,
      "epoch": 0.6581095791001451,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963699475650751,
      "loss": 5.4274,
      "mean_token_accuracy": 0.1789287731051445,
      "num_tokens": 15081109.0,
      "step": 7255
    },
    {
      "entropy": 5.514655733108521,
      "epoch": 0.6585631349782293,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963641565651255,
      "loss": 5.3029,
      "mean_token_accuracy": 0.18567312955856324,
      "num_tokens": 15092315.0,
      "step": 7260
    },
    {
      "entropy": 5.513403844833374,
      "epoch": 0.6590166908563135,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963583609873084,
      "loss": 5.3694,
      "mean_token_accuracy": 0.18358769416809081,
      "num_tokens": 15103957.0,
      "step": 7265
    },
    {
      "entropy": 5.562368488311767,
      "epoch": 0.6594702467343977,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963525608317443,
      "loss": 5.362,
      "mean_token_accuracy": 0.1775441661477089,
      "num_tokens": 15114996.0,
      "step": 7270
    },
    {
      "entropy": 5.581370878219604,
      "epoch": 0.6599238026124818,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004963467560985527,
      "loss": 5.2938,
      "mean_token_accuracy": 0.17815622687339783,
      "num_tokens": 15125098.0,
      "step": 7275
    },
    {
      "entropy": 5.480411434173584,
      "epoch": 0.660377358490566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963409467878538,
      "loss": 5.2984,
      "mean_token_accuracy": 0.1813468173146248,
      "num_tokens": 15134746.0,
      "step": 7280
    },
    {
      "entropy": 5.59424524307251,
      "epoch": 0.6608309143686503,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004963351328997678,
      "loss": 5.3253,
      "mean_token_accuracy": 0.17637452483177185,
      "num_tokens": 15145698.0,
      "step": 7285
    },
    {
      "entropy": 5.577039194107056,
      "epoch": 0.6612844702467344,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963293144344148,
      "loss": 5.3688,
      "mean_token_accuracy": 0.17920732200145723,
      "num_tokens": 15156026.0,
      "step": 7290
    },
    {
      "entropy": 5.529924106597901,
      "epoch": 0.6617380261248186,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963234913919151,
      "loss": 5.3084,
      "mean_token_accuracy": 0.17793872058391572,
      "num_tokens": 15165900.0,
      "step": 7295
    },
    {
      "entropy": 5.5461030960083,
      "epoch": 0.6621915820029027,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963176637723893,
      "loss": 5.3744,
      "mean_token_accuracy": 0.182407346367836,
      "num_tokens": 15176693.0,
      "step": 7300
    },
    {
      "entropy": 5.591224098205567,
      "epoch": 0.6626451378809869,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963118315759578,
      "loss": 5.3901,
      "mean_token_accuracy": 0.16922723799943923,
      "num_tokens": 15186535.0,
      "step": 7305
    },
    {
      "entropy": 5.577974796295166,
      "epoch": 0.6630986937590712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963059948027411,
      "loss": 5.3013,
      "mean_token_accuracy": 0.182105216383934,
      "num_tokens": 15196019.0,
      "step": 7310
    },
    {
      "entropy": 5.474636697769165,
      "epoch": 0.6635522496371553,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963001534528601,
      "loss": 5.1884,
      "mean_token_accuracy": 0.192403119802475,
      "num_tokens": 15205273.0,
      "step": 7315
    },
    {
      "entropy": 5.537912750244141,
      "epoch": 0.6640058055152395,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962943075264355,
      "loss": 5.3325,
      "mean_token_accuracy": 0.1802896723151207,
      "num_tokens": 15216716.0,
      "step": 7320
    },
    {
      "entropy": 5.575633764266968,
      "epoch": 0.6644593613933236,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004962884570235884,
      "loss": 5.3761,
      "mean_token_accuracy": 0.17417128533124923,
      "num_tokens": 15227306.0,
      "step": 7325
    },
    {
      "entropy": 5.511094474792481,
      "epoch": 0.6649129172714079,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962826019444394,
      "loss": 5.2652,
      "mean_token_accuracy": 0.1794412538409233,
      "num_tokens": 15237302.0,
      "step": 7330
    },
    {
      "entropy": 5.517401790618896,
      "epoch": 0.665366473149492,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962767422891099,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18371451795101165,
      "num_tokens": 15247064.0,
      "step": 7335
    },
    {
      "entropy": 5.537144470214844,
      "epoch": 0.6658200290275762,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496270878057721,
      "loss": 5.3427,
      "mean_token_accuracy": 0.18440829068422318,
      "num_tokens": 15257006.0,
      "step": 7340
    },
    {
      "entropy": 5.58497314453125,
      "epoch": 0.6662735849056604,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962650092503939,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1943359598517418,
      "num_tokens": 15266720.0,
      "step": 7345
    },
    {
      "entropy": 5.543534851074218,
      "epoch": 0.6667271407837445,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962591358672502,
      "loss": 5.3427,
      "mean_token_accuracy": 0.18019769638776778,
      "num_tokens": 15277330.0,
      "step": 7350
    },
    {
      "entropy": 5.5373796939849855,
      "epoch": 0.6671806966618288,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496253257908411,
      "loss": 5.271,
      "mean_token_accuracy": 0.1839846909046173,
      "num_tokens": 15287401.0,
      "step": 7355
    },
    {
      "entropy": 5.489133691787719,
      "epoch": 0.6676342525399129,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962473753739981,
      "loss": 5.2292,
      "mean_token_accuracy": 0.1890961855649948,
      "num_tokens": 15297426.0,
      "step": 7360
    },
    {
      "entropy": 5.556583213806152,
      "epoch": 0.6680878084179971,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496241488264133,
      "loss": 5.2272,
      "mean_token_accuracy": 0.1845378175377846,
      "num_tokens": 15308412.0,
      "step": 7365
    },
    {
      "entropy": 5.527732086181641,
      "epoch": 0.6685413642960812,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004962355965789377,
      "loss": 5.2293,
      "mean_token_accuracy": 0.18310726881027223,
      "num_tokens": 15318821.0,
      "step": 7370
    },
    {
      "entropy": 5.476294469833374,
      "epoch": 0.6689949201741655,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962297003185338,
      "loss": 5.2916,
      "mean_token_accuracy": 0.1801626518368721,
      "num_tokens": 15328737.0,
      "step": 7375
    },
    {
      "entropy": 5.622493934631348,
      "epoch": 0.6694484760522497,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962237994830433,
      "loss": 5.393,
      "mean_token_accuracy": 0.17233246713876724,
      "num_tokens": 15339695.0,
      "step": 7380
    },
    {
      "entropy": 5.615177869796753,
      "epoch": 0.6699020319303338,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962178940725883,
      "loss": 5.364,
      "mean_token_accuracy": 0.18082518577575685,
      "num_tokens": 15350270.0,
      "step": 7385
    },
    {
      "entropy": 5.520081806182861,
      "epoch": 0.670355587808418,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962119840872908,
      "loss": 5.3439,
      "mean_token_accuracy": 0.18389475494623184,
      "num_tokens": 15360986.0,
      "step": 7390
    },
    {
      "entropy": 5.516814851760865,
      "epoch": 0.6708091436865021,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004962060695272733,
      "loss": 5.3391,
      "mean_token_accuracy": 0.18298070579767228,
      "num_tokens": 15371786.0,
      "step": 7395
    },
    {
      "entropy": 5.55599799156189,
      "epoch": 0.6712626995645864,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004962001503926577,
      "loss": 5.2993,
      "mean_token_accuracy": 0.1882503256201744,
      "num_tokens": 15382960.0,
      "step": 7400
    },
    {
      "entropy": 5.526637887954712,
      "epoch": 0.6717162554426706,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961942266835667,
      "loss": 5.287,
      "mean_token_accuracy": 0.18534947335720062,
      "num_tokens": 15392807.0,
      "step": 7405
    },
    {
      "entropy": 5.559031534194946,
      "epoch": 0.6721698113207547,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004961882984001227,
      "loss": 5.3307,
      "mean_token_accuracy": 0.1767762467265129,
      "num_tokens": 15403712.0,
      "step": 7410
    },
    {
      "entropy": 5.584656190872193,
      "epoch": 0.6726233671988389,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961823655424484,
      "loss": 5.3916,
      "mean_token_accuracy": 0.17293067425489425,
      "num_tokens": 15414839.0,
      "step": 7415
    },
    {
      "entropy": 5.518201398849487,
      "epoch": 0.6730769230769231,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961764281106664,
      "loss": 5.2838,
      "mean_token_accuracy": 0.1870151072740555,
      "num_tokens": 15425320.0,
      "step": 7420
    },
    {
      "entropy": 5.517269277572632,
      "epoch": 0.6735304789550073,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961704861048995,
      "loss": 5.2887,
      "mean_token_accuracy": 0.17715439349412918,
      "num_tokens": 15435476.0,
      "step": 7425
    },
    {
      "entropy": 5.53749532699585,
      "epoch": 0.6739840348330914,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961645395252705,
      "loss": 5.3947,
      "mean_token_accuracy": 0.17859402149915696,
      "num_tokens": 15445786.0,
      "step": 7430
    },
    {
      "entropy": 5.562483930587769,
      "epoch": 0.6744375907111756,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961585883719025,
      "loss": 5.3762,
      "mean_token_accuracy": 0.18556168675422668,
      "num_tokens": 15455915.0,
      "step": 7435
    },
    {
      "entropy": 5.612591075897217,
      "epoch": 0.6748911465892597,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961526326449186,
      "loss": 5.3614,
      "mean_token_accuracy": 0.1754249468445778,
      "num_tokens": 15466523.0,
      "step": 7440
    },
    {
      "entropy": 5.561124181747436,
      "epoch": 0.675344702467344,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004961466723444419,
      "loss": 5.3425,
      "mean_token_accuracy": 0.18115431517362596,
      "num_tokens": 15476821.0,
      "step": 7445
    },
    {
      "entropy": 5.506720113754272,
      "epoch": 0.6757982583454282,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961407074705955,
      "loss": 5.2759,
      "mean_token_accuracy": 0.1748933583498001,
      "num_tokens": 15487182.0,
      "step": 7450
    },
    {
      "entropy": 5.5438395023345945,
      "epoch": 0.6762518142235123,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961347380235031,
      "loss": 5.267,
      "mean_token_accuracy": 0.18452272266149522,
      "num_tokens": 15497600.0,
      "step": 7455
    },
    {
      "entropy": 5.560304641723633,
      "epoch": 0.6767053701015965,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961287640032879,
      "loss": 5.41,
      "mean_token_accuracy": 0.17247417122125625,
      "num_tokens": 15508278.0,
      "step": 7460
    },
    {
      "entropy": 5.6194220066070555,
      "epoch": 0.6771589259796807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961227854100734,
      "loss": 5.4261,
      "mean_token_accuracy": 0.17710963487625123,
      "num_tokens": 15518458.0,
      "step": 7465
    },
    {
      "entropy": 5.551078796386719,
      "epoch": 0.6776124818577649,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961168022439835,
      "loss": 5.2497,
      "mean_token_accuracy": 0.1808410331606865,
      "num_tokens": 15528738.0,
      "step": 7470
    },
    {
      "entropy": 5.536972951889038,
      "epoch": 0.6780660377358491,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961108145051417,
      "loss": 5.2783,
      "mean_token_accuracy": 0.1874865859746933,
      "num_tokens": 15538656.0,
      "step": 7475
    },
    {
      "entropy": 5.481096410751343,
      "epoch": 0.6785195936139332,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496104822193672,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18348010629415512,
      "num_tokens": 15549285.0,
      "step": 7480
    },
    {
      "entropy": 5.595572376251221,
      "epoch": 0.6789731494920174,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960988253096982,
      "loss": 5.3666,
      "mean_token_accuracy": 0.18583495169878006,
      "num_tokens": 15559766.0,
      "step": 7485
    },
    {
      "entropy": 5.505124187469482,
      "epoch": 0.6794267053701016,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004960928238533443,
      "loss": 5.2299,
      "mean_token_accuracy": 0.19369809031486512,
      "num_tokens": 15570873.0,
      "step": 7490
    },
    {
      "entropy": 5.5578302383422855,
      "epoch": 0.6798802612481858,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004960868178247344,
      "loss": 5.4746,
      "mean_token_accuracy": 0.17800090312957764,
      "num_tokens": 15581229.0,
      "step": 7495
    },
    {
      "entropy": 5.530143356323242,
      "epoch": 0.68033381712627,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960808072239929,
      "loss": 5.3422,
      "mean_token_accuracy": 0.17350580245256425,
      "num_tokens": 15591686.0,
      "step": 7500
    },
    {
      "entropy": 5.646744251251221,
      "epoch": 0.6807873730043541,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960747920512439,
      "loss": 5.3575,
      "mean_token_accuracy": 0.17898468375205995,
      "num_tokens": 15601869.0,
      "step": 7505
    },
    {
      "entropy": 5.519121694564819,
      "epoch": 0.6812409288824384,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000496068772306612,
      "loss": 5.3459,
      "mean_token_accuracy": 0.17619211375713348,
      "num_tokens": 15612946.0,
      "step": 7510
    },
    {
      "entropy": 5.554516172409057,
      "epoch": 0.6816944847605225,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960627479902214,
      "loss": 5.3769,
      "mean_token_accuracy": 0.17832256257534027,
      "num_tokens": 15623761.0,
      "step": 7515
    },
    {
      "entropy": 5.549900197982788,
      "epoch": 0.6821480406386067,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496056719102197,
      "loss": 5.369,
      "mean_token_accuracy": 0.17480377554893495,
      "num_tokens": 15634500.0,
      "step": 7520
    },
    {
      "entropy": 5.474475002288818,
      "epoch": 0.6826015965166908,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960506856426631,
      "loss": 5.2228,
      "mean_token_accuracy": 0.18139137476682662,
      "num_tokens": 15644896.0,
      "step": 7525
    },
    {
      "entropy": 5.5556963920593265,
      "epoch": 0.683055152394775,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960446476117448,
      "loss": 5.3439,
      "mean_token_accuracy": 0.18209579437971116,
      "num_tokens": 15655008.0,
      "step": 7530
    },
    {
      "entropy": 5.526656913757324,
      "epoch": 0.6835087082728593,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960386050095668,
      "loss": 5.265,
      "mean_token_accuracy": 0.17771433889865876,
      "num_tokens": 15665944.0,
      "step": 7535
    },
    {
      "entropy": 5.55998010635376,
      "epoch": 0.6839622641509434,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004960325578362543,
      "loss": 5.3639,
      "mean_token_accuracy": 0.18044693470001222,
      "num_tokens": 15677099.0,
      "step": 7540
    },
    {
      "entropy": 5.593668031692505,
      "epoch": 0.6844158200290276,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496026506091932,
      "loss": 5.3198,
      "mean_token_accuracy": 0.1848937004804611,
      "num_tokens": 15687531.0,
      "step": 7545
    },
    {
      "entropy": 5.586200761795044,
      "epoch": 0.6848693759071117,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960204497767253,
      "loss": 5.3321,
      "mean_token_accuracy": 0.17955249547958374,
      "num_tokens": 15697002.0,
      "step": 7550
    },
    {
      "entropy": 5.43728461265564,
      "epoch": 0.685322931785196,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960143888907594,
      "loss": 5.2166,
      "mean_token_accuracy": 0.18282410353422165,
      "num_tokens": 15707249.0,
      "step": 7555
    },
    {
      "entropy": 5.5209822177886965,
      "epoch": 0.6857764876632801,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960083234341595,
      "loss": 5.2848,
      "mean_token_accuracy": 0.18320000618696214,
      "num_tokens": 15718764.0,
      "step": 7560
    },
    {
      "entropy": 5.546800947189331,
      "epoch": 0.6862300435413643,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960022534070513,
      "loss": 5.2361,
      "mean_token_accuracy": 0.1883462518453598,
      "num_tokens": 15728854.0,
      "step": 7565
    },
    {
      "entropy": 5.43729796409607,
      "epoch": 0.6866835994194485,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049599617880956,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18559020906686782,
      "num_tokens": 15739682.0,
      "step": 7570
    },
    {
      "entropy": 5.473732614517212,
      "epoch": 0.6871371552975326,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959900996418115,
      "loss": 5.2534,
      "mean_token_accuracy": 0.18659544587135315,
      "num_tokens": 15749628.0,
      "step": 7575
    },
    {
      "entropy": 5.554725790023804,
      "epoch": 0.6875907111756169,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959840159039315,
      "loss": 5.3221,
      "mean_token_accuracy": 0.1806378498673439,
      "num_tokens": 15760137.0,
      "step": 7580
    },
    {
      "entropy": 5.579864168167115,
      "epoch": 0.688044267053701,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959779275960456,
      "loss": 5.2711,
      "mean_token_accuracy": 0.18420590311288834,
      "num_tokens": 15769889.0,
      "step": 7585
    },
    {
      "entropy": 5.514402961730957,
      "epoch": 0.6884978229317852,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049597183471828,
      "loss": 5.3859,
      "mean_token_accuracy": 0.18321892023086547,
      "num_tokens": 15780912.0,
      "step": 7590
    },
    {
      "entropy": 5.531727409362793,
      "epoch": 0.6889513788098693,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004959657372707604,
      "loss": 5.2578,
      "mean_token_accuracy": 0.18849852085113525,
      "num_tokens": 15791719.0,
      "step": 7595
    },
    {
      "entropy": 5.572351932525635,
      "epoch": 0.6894049346879536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959596352536132,
      "loss": 5.2987,
      "mean_token_accuracy": 0.18423290997743608,
      "num_tokens": 15802551.0,
      "step": 7600
    },
    {
      "entropy": 5.585467529296875,
      "epoch": 0.6898584905660378,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959535286669643,
      "loss": 5.3539,
      "mean_token_accuracy": 0.1799274355173111,
      "num_tokens": 15813551.0,
      "step": 7605
    },
    {
      "entropy": 5.518627882003784,
      "epoch": 0.6903120464441219,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959474175109402,
      "loss": 5.3222,
      "mean_token_accuracy": 0.1847722724080086,
      "num_tokens": 15824262.0,
      "step": 7610
    },
    {
      "entropy": 5.595053720474243,
      "epoch": 0.6907656023222061,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959413017856672,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18128857016563416,
      "num_tokens": 15835493.0,
      "step": 7615
    },
    {
      "entropy": 5.448233413696289,
      "epoch": 0.6912191582002902,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004959351814912716,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18287449926137925,
      "num_tokens": 15846224.0,
      "step": 7620
    },
    {
      "entropy": 5.5843593120574955,
      "epoch": 0.6916727140783745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959290566278803,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1838727205991745,
      "num_tokens": 15855622.0,
      "step": 7625
    },
    {
      "entropy": 5.5864767074584964,
      "epoch": 0.6921262699564587,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959229271956199,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18438090682029723,
      "num_tokens": 15865835.0,
      "step": 7630
    },
    {
      "entropy": 5.511325550079346,
      "epoch": 0.6925798258345428,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959167931946169,
      "loss": 5.2916,
      "mean_token_accuracy": 0.18278788179159164,
      "num_tokens": 15875915.0,
      "step": 7635
    },
    {
      "entropy": 5.570894479751587,
      "epoch": 0.693033381712627,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959106546249984,
      "loss": 5.301,
      "mean_token_accuracy": 0.18494617342948913,
      "num_tokens": 15887111.0,
      "step": 7640
    },
    {
      "entropy": 5.589294576644898,
      "epoch": 0.6934869375907112,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004959045114868913,
      "loss": 5.4105,
      "mean_token_accuracy": 0.17258675992488862,
      "num_tokens": 15899528.0,
      "step": 7645
    },
    {
      "entropy": 5.547265100479126,
      "epoch": 0.6939404934687954,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958983637804225,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18939780741930007,
      "num_tokens": 15910881.0,
      "step": 7650
    },
    {
      "entropy": 5.6097711563110355,
      "epoch": 0.6943940493468795,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958922115057194,
      "loss": 5.3506,
      "mean_token_accuracy": 0.17834327965974808,
      "num_tokens": 15921189.0,
      "step": 7655
    },
    {
      "entropy": 5.611061525344849,
      "epoch": 0.6948476052249637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958860546629089,
      "loss": 5.3497,
      "mean_token_accuracy": 0.18507591933012008,
      "num_tokens": 15931276.0,
      "step": 7660
    },
    {
      "entropy": 5.58590931892395,
      "epoch": 0.6953011611030478,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004958798932521186,
      "loss": 5.429,
      "mean_token_accuracy": 0.1739620342850685,
      "num_tokens": 15941506.0,
      "step": 7665
    },
    {
      "entropy": 5.493170309066772,
      "epoch": 0.6957547169811321,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958737272734759,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18339498937129975,
      "num_tokens": 15951158.0,
      "step": 7670
    },
    {
      "entropy": 5.588201665878296,
      "epoch": 0.6962082728592163,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004958675567271081,
      "loss": 5.2997,
      "mean_token_accuracy": 0.18696109801530839,
      "num_tokens": 15960852.0,
      "step": 7675
    },
    {
      "entropy": 5.544498586654663,
      "epoch": 0.6966618287373004,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000495861381613143,
      "loss": 5.2936,
      "mean_token_accuracy": 0.17795197516679764,
      "num_tokens": 15971291.0,
      "step": 7680
    },
    {
      "entropy": 5.560691261291504,
      "epoch": 0.6971153846153846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958552019317082,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18023389130830764,
      "num_tokens": 15982180.0,
      "step": 7685
    },
    {
      "entropy": 5.486464738845825,
      "epoch": 0.6975689404934688,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004958490176829316,
      "loss": 5.2506,
      "mean_token_accuracy": 0.18482301533222198,
      "num_tokens": 15992363.0,
      "step": 7690
    },
    {
      "entropy": 5.53675217628479,
      "epoch": 0.698022496371553,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495842828866941,
      "loss": 5.3368,
      "mean_token_accuracy": 0.1778530851006508,
      "num_tokens": 16003477.0,
      "step": 7695
    },
    {
      "entropy": 5.585207319259643,
      "epoch": 0.6984760522496372,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958366354838645,
      "loss": 5.3135,
      "mean_token_accuracy": 0.1820820912718773,
      "num_tokens": 16013747.0,
      "step": 7700
    },
    {
      "entropy": 5.542598628997803,
      "epoch": 0.6989296081277213,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049583043753383,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1875500798225403,
      "num_tokens": 16025662.0,
      "step": 7705
    },
    {
      "entropy": 5.572959136962891,
      "epoch": 0.6993831640058055,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004958242350169658,
      "loss": 5.3992,
      "mean_token_accuracy": 0.1829013928771019,
      "num_tokens": 16037315.0,
      "step": 7710
    },
    {
      "entropy": 5.502584838867188,
      "epoch": 0.6998367198838897,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004958180279334001,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1828617572784424,
      "num_tokens": 16047297.0,
      "step": 7715
    },
    {
      "entropy": 5.482805871963501,
      "epoch": 0.7002902757619739,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958118162832614,
      "loss": 5.2115,
      "mean_token_accuracy": 0.18582056909799577,
      "num_tokens": 16057573.0,
      "step": 7720
    },
    {
      "entropy": 5.51785831451416,
      "epoch": 0.700743831640058,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958056000666781,
      "loss": 5.3708,
      "mean_token_accuracy": 0.17752822488546371,
      "num_tokens": 16067926.0,
      "step": 7725
    },
    {
      "entropy": 5.560018396377563,
      "epoch": 0.7011973875181422,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957993792837786,
      "loss": 5.3089,
      "mean_token_accuracy": 0.18352362066507338,
      "num_tokens": 16077938.0,
      "step": 7730
    },
    {
      "entropy": 5.509559869766235,
      "epoch": 0.7016509433962265,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004957931539346917,
      "loss": 5.1887,
      "mean_token_accuracy": 0.18910111486911774,
      "num_tokens": 16088163.0,
      "step": 7735
    },
    {
      "entropy": 5.6044848442077635,
      "epoch": 0.7021044992743106,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957869240195462,
      "loss": 5.3964,
      "mean_token_accuracy": 0.17759514451026917,
      "num_tokens": 16100640.0,
      "step": 7740
    },
    {
      "entropy": 5.525788402557373,
      "epoch": 0.7025580551523948,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004957806895384707,
      "loss": 5.22,
      "mean_token_accuracy": 0.1873020723462105,
      "num_tokens": 16111018.0,
      "step": 7745
    },
    {
      "entropy": 5.509130525588989,
      "epoch": 0.7030116110304789,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957744504915944,
      "loss": 5.2634,
      "mean_token_accuracy": 0.17722718566656112,
      "num_tokens": 16120873.0,
      "step": 7750
    },
    {
      "entropy": 5.563656759262085,
      "epoch": 0.7034651669085631,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957682068790463,
      "loss": 5.2845,
      "mean_token_accuracy": 0.18420600146055222,
      "num_tokens": 16131178.0,
      "step": 7755
    },
    {
      "entropy": 5.503865575790405,
      "epoch": 0.7039187227866474,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004957619587009553,
      "loss": 5.2961,
      "mean_token_accuracy": 0.18365350663661956,
      "num_tokens": 16141156.0,
      "step": 7760
    },
    {
      "entropy": 5.544031953811645,
      "epoch": 0.7043722786647315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004957557059574507,
      "loss": 5.397,
      "mean_token_accuracy": 0.18116726726293564,
      "num_tokens": 16150077.0,
      "step": 7765
    },
    {
      "entropy": 5.521406173706055,
      "epoch": 0.7048258345428157,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957494486486618,
      "loss": 5.307,
      "mean_token_accuracy": 0.1804828479886055,
      "num_tokens": 16160514.0,
      "step": 7770
    },
    {
      "entropy": 5.567109632492065,
      "epoch": 0.7052793904208998,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957431867747182,
      "loss": 5.3432,
      "mean_token_accuracy": 0.187213496863842,
      "num_tokens": 16170811.0,
      "step": 7775
    },
    {
      "entropy": 5.480603742599487,
      "epoch": 0.7057329462989841,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957369203357492,
      "loss": 5.2402,
      "mean_token_accuracy": 0.18310723453760147,
      "num_tokens": 16181553.0,
      "step": 7780
    },
    {
      "entropy": 5.523513269424439,
      "epoch": 0.7061865021770682,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004957306493318844,
      "loss": 5.2696,
      "mean_token_accuracy": 0.18861202895641327,
      "num_tokens": 16191308.0,
      "step": 7785
    },
    {
      "entropy": 5.526185083389282,
      "epoch": 0.7066400580551524,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957243737632535,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18830116391181945,
      "num_tokens": 16201311.0,
      "step": 7790
    },
    {
      "entropy": 5.584646558761596,
      "epoch": 0.7070936139332366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957180936299863,
      "loss": 5.2822,
      "mean_token_accuracy": 0.19124689400196077,
      "num_tokens": 16211105.0,
      "step": 7795
    },
    {
      "entropy": 5.554888391494751,
      "epoch": 0.7075471698113207,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957118089322127,
      "loss": 5.3012,
      "mean_token_accuracy": 0.1761303499341011,
      "num_tokens": 16220836.0,
      "step": 7800
    },
    {
      "entropy": 5.484191131591797,
      "epoch": 0.708000725689405,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957055196700628,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18652159571647645,
      "num_tokens": 16231630.0,
      "step": 7805
    },
    {
      "entropy": 5.544210433959961,
      "epoch": 0.7084542815674891,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004956992258436663,
      "loss": 5.304,
      "mean_token_accuracy": 0.18482640385627747,
      "num_tokens": 16242272.0,
      "step": 7810
    },
    {
      "entropy": 5.429587078094483,
      "epoch": 0.7089078374455733,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956929274531536,
      "loss": 5.2099,
      "mean_token_accuracy": 0.19073296040296556,
      "num_tokens": 16252555.0,
      "step": 7815
    },
    {
      "entropy": 5.518241500854492,
      "epoch": 0.7093613933236574,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000495686624498655,
      "loss": 5.331,
      "mean_token_accuracy": 0.187027183175087,
      "num_tokens": 16261964.0,
      "step": 7820
    },
    {
      "entropy": 5.617752361297607,
      "epoch": 0.7098149492017417,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004956803169803007,
      "loss": 5.3859,
      "mean_token_accuracy": 0.17545270025730134,
      "num_tokens": 16274543.0,
      "step": 7825
    },
    {
      "entropy": 5.507696151733398,
      "epoch": 0.7102685050798259,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004956740048982212,
      "loss": 5.2941,
      "mean_token_accuracy": 0.1758976697921753,
      "num_tokens": 16285124.0,
      "step": 7830
    },
    {
      "entropy": 5.528030920028686,
      "epoch": 0.71072206095791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956676882525472,
      "loss": 5.3634,
      "mean_token_accuracy": 0.18425541818141938,
      "num_tokens": 16295727.0,
      "step": 7835
    },
    {
      "entropy": 5.628829622268677,
      "epoch": 0.7111756168359942,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495661367043409,
      "loss": 5.2686,
      "mean_token_accuracy": 0.18808830976486207,
      "num_tokens": 16305496.0,
      "step": 7840
    },
    {
      "entropy": 5.535170650482177,
      "epoch": 0.7116291727140783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956550412709376,
      "loss": 5.2455,
      "mean_token_accuracy": 0.19257238060235976,
      "num_tokens": 16315112.0,
      "step": 7845
    },
    {
      "entropy": 5.514770650863648,
      "epoch": 0.7120827285921626,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956487109352637,
      "loss": 5.2948,
      "mean_token_accuracy": 0.18559942841529847,
      "num_tokens": 16325613.0,
      "step": 7850
    },
    {
      "entropy": 5.510515832901001,
      "epoch": 0.7125362844702468,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956423760365181,
      "loss": 5.2236,
      "mean_token_accuracy": 0.18841205388307572,
      "num_tokens": 16335317.0,
      "step": 7855
    },
    {
      "entropy": 5.594292879104614,
      "epoch": 0.7129898403483309,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004956360365748321,
      "loss": 5.3684,
      "mean_token_accuracy": 0.17692358642816544,
      "num_tokens": 16345003.0,
      "step": 7860
    },
    {
      "entropy": 5.5583514213562015,
      "epoch": 0.7134433962264151,
      "grad_norm": 0.875,
      "learning_rate": 0.0004956296925503367,
      "loss": 5.3597,
      "mean_token_accuracy": 0.1829149454832077,
      "num_tokens": 16356108.0,
      "step": 7865
    },
    {
      "entropy": 5.483655261993408,
      "epoch": 0.7138969521044993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004956233439631628,
      "loss": 5.1666,
      "mean_token_accuracy": 0.19129244983196259,
      "num_tokens": 16364733.0,
      "step": 7870
    },
    {
      "entropy": 5.558326196670532,
      "epoch": 0.7143505079825835,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956169908134421,
      "loss": 5.273,
      "mean_token_accuracy": 0.18959595412015914,
      "num_tokens": 16375229.0,
      "step": 7875
    },
    {
      "entropy": 5.519115591049195,
      "epoch": 0.7148040638606676,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004956106331013057,
      "loss": 5.3594,
      "mean_token_accuracy": 0.18587884157896042,
      "num_tokens": 16386366.0,
      "step": 7880
    },
    {
      "entropy": 5.490304279327392,
      "epoch": 0.7152576197387518,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956042708268853,
      "loss": 5.2427,
      "mean_token_accuracy": 0.1905361145734787,
      "num_tokens": 16397067.0,
      "step": 7885
    },
    {
      "entropy": 5.452323913574219,
      "epoch": 0.715711175616836,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955979039903123,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18583076745271682,
      "num_tokens": 16407256.0,
      "step": 7890
    },
    {
      "entropy": 5.501726055145264,
      "epoch": 0.7161647314949202,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004955915325917183,
      "loss": 5.2573,
      "mean_token_accuracy": 0.18402386009693145,
      "num_tokens": 16418701.0,
      "step": 7895
    },
    {
      "entropy": 5.472748422622681,
      "epoch": 0.7166182873730044,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004955851566312354,
      "loss": 5.1796,
      "mean_token_accuracy": 0.19416148364543914,
      "num_tokens": 16428019.0,
      "step": 7900
    },
    {
      "entropy": 5.482213354110717,
      "epoch": 0.7170718432510885,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955787761089952,
      "loss": 5.2658,
      "mean_token_accuracy": 0.179888778924942,
      "num_tokens": 16440122.0,
      "step": 7905
    },
    {
      "entropy": 5.606745100021362,
      "epoch": 0.7175253991291727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004955723910251296,
      "loss": 5.3425,
      "mean_token_accuracy": 0.1781800538301468,
      "num_tokens": 16449801.0,
      "step": 7910
    },
    {
      "entropy": 5.498953914642334,
      "epoch": 0.717978955007257,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955660013797709,
      "loss": 5.2049,
      "mean_token_accuracy": 0.18978928178548812,
      "num_tokens": 16460155.0,
      "step": 7915
    },
    {
      "entropy": 5.5611310482025145,
      "epoch": 0.7184325108853411,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495559607173051,
      "loss": 5.4094,
      "mean_token_accuracy": 0.17007175385951995,
      "num_tokens": 16470656.0,
      "step": 7920
    },
    {
      "entropy": 5.565587759017944,
      "epoch": 0.7188860667634253,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955532084051022,
      "loss": 5.3028,
      "mean_token_accuracy": 0.18778245747089387,
      "num_tokens": 16481449.0,
      "step": 7925
    },
    {
      "entropy": 5.544949960708618,
      "epoch": 0.7193396226415094,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955468050760568,
      "loss": 5.3213,
      "mean_token_accuracy": 0.18307759761810302,
      "num_tokens": 16492032.0,
      "step": 7930
    },
    {
      "entropy": 5.559630060195923,
      "epoch": 0.7197931785195936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955403971860474,
      "loss": 5.2449,
      "mean_token_accuracy": 0.18228569626808167,
      "num_tokens": 16502221.0,
      "step": 7935
    },
    {
      "entropy": 5.516655302047729,
      "epoch": 0.7202467343976778,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955339847352063,
      "loss": 5.2255,
      "mean_token_accuracy": 0.18727567791938782,
      "num_tokens": 16512021.0,
      "step": 7940
    },
    {
      "entropy": 5.526650905609131,
      "epoch": 0.720700290275762,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955275677236662,
      "loss": 5.2781,
      "mean_token_accuracy": 0.18725835978984834,
      "num_tokens": 16522060.0,
      "step": 7945
    },
    {
      "entropy": 5.522590780258179,
      "epoch": 0.7211538461538461,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955211461515599,
      "loss": 5.2851,
      "mean_token_accuracy": 0.18848953992128373,
      "num_tokens": 16532827.0,
      "step": 7950
    },
    {
      "entropy": 5.5282035827636715,
      "epoch": 0.7216074020319303,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049551472001902,
      "loss": 5.3291,
      "mean_token_accuracy": 0.18570098727941514,
      "num_tokens": 16542838.0,
      "step": 7955
    },
    {
      "entropy": 5.49766263961792,
      "epoch": 0.7220609579100146,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955082893261796,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18782561421394348,
      "num_tokens": 16552615.0,
      "step": 7960
    },
    {
      "entropy": 5.5685553550720215,
      "epoch": 0.7225145137880987,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955018540731714,
      "loss": 5.3358,
      "mean_token_accuracy": 0.18197497874498367,
      "num_tokens": 16561777.0,
      "step": 7965
    },
    {
      "entropy": 5.561930131912232,
      "epoch": 0.7229680696661829,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954954142601288,
      "loss": 5.3046,
      "mean_token_accuracy": 0.1859089732170105,
      "num_tokens": 16572111.0,
      "step": 7970
    },
    {
      "entropy": 5.553031778335571,
      "epoch": 0.723421625544267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004954889698871849,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1823531538248062,
      "num_tokens": 16581674.0,
      "step": 7975
    },
    {
      "entropy": 5.5707708358764645,
      "epoch": 0.7238751814223512,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004954825209544728,
      "loss": 5.3265,
      "mean_token_accuracy": 0.18853056877851487,
      "num_tokens": 16591802.0,
      "step": 7980
    },
    {
      "entropy": 5.4827385425567625,
      "epoch": 0.7243287373004355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495476067462126,
      "loss": 5.2194,
      "mean_token_accuracy": 0.1914340078830719,
      "num_tokens": 16602772.0,
      "step": 7985
    },
    {
      "entropy": 5.56507248878479,
      "epoch": 0.7247822931785196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495469609410278,
      "loss": 5.2697,
      "mean_token_accuracy": 0.18687217831611633,
      "num_tokens": 16613526.0,
      "step": 7990
    },
    {
      "entropy": 5.475165891647339,
      "epoch": 0.7252358490566038,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954631467990622,
      "loss": 5.2738,
      "mean_token_accuracy": 0.1871074303984642,
      "num_tokens": 16623583.0,
      "step": 7995
    },
    {
      "entropy": 5.543488788604736,
      "epoch": 0.7256894049346879,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954566796286124,
      "loss": 5.3322,
      "mean_token_accuracy": 0.1807237446308136,
      "num_tokens": 16635068.0,
      "step": 8000
    },
    {
      "entropy": 5.588771057128906,
      "epoch": 0.7261429608127722,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954502078990623,
      "loss": 5.2332,
      "mean_token_accuracy": 0.18706244379281997,
      "num_tokens": 16645021.0,
      "step": 8005
    },
    {
      "entropy": 5.526928997039795,
      "epoch": 0.7265965166908563,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004954437316105457,
      "loss": 5.3204,
      "mean_token_accuracy": 0.17738464921712876,
      "num_tokens": 16656021.0,
      "step": 8010
    },
    {
      "entropy": 5.476295518875122,
      "epoch": 0.7270500725689405,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954372507631965,
      "loss": 5.2173,
      "mean_token_accuracy": 0.19314096868038177,
      "num_tokens": 16665236.0,
      "step": 8015
    },
    {
      "entropy": 5.545669603347778,
      "epoch": 0.7275036284470247,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954307653571489,
      "loss": 5.4112,
      "mean_token_accuracy": 0.17193434983491898,
      "num_tokens": 16676690.0,
      "step": 8020
    },
    {
      "entropy": 5.553662109375,
      "epoch": 0.7279571843251088,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954242753925367,
      "loss": 5.2901,
      "mean_token_accuracy": 0.17670878767967224,
      "num_tokens": 16687288.0,
      "step": 8025
    },
    {
      "entropy": 5.5189173221588135,
      "epoch": 0.7284107402031931,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954177808694946,
      "loss": 5.3052,
      "mean_token_accuracy": 0.1775049388408661,
      "num_tokens": 16697873.0,
      "step": 8030
    },
    {
      "entropy": 5.551232862472534,
      "epoch": 0.7288642960812772,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004954112817881565,
      "loss": 5.3207,
      "mean_token_accuracy": 0.18077222257852554,
      "num_tokens": 16709002.0,
      "step": 8035
    },
    {
      "entropy": 5.493309879302979,
      "epoch": 0.7293178519593614,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954047781486568,
      "loss": 5.278,
      "mean_token_accuracy": 0.18425150364637374,
      "num_tokens": 16718989.0,
      "step": 8040
    },
    {
      "entropy": 5.620672035217285,
      "epoch": 0.7297714078374455,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953982699511303,
      "loss": 5.3692,
      "mean_token_accuracy": 0.18326299786567687,
      "num_tokens": 16730425.0,
      "step": 8045
    },
    {
      "entropy": 5.562078189849854,
      "epoch": 0.7302249637155298,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953917571957114,
      "loss": 5.2638,
      "mean_token_accuracy": 0.18367500156164168,
      "num_tokens": 16740216.0,
      "step": 8050
    },
    {
      "entropy": 5.507948780059815,
      "epoch": 0.730678519593614,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953852398825346,
      "loss": 5.2517,
      "mean_token_accuracy": 0.18657175153493882,
      "num_tokens": 16752436.0,
      "step": 8055
    },
    {
      "entropy": 5.471053600311279,
      "epoch": 0.7311320754716981,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495378718011735,
      "loss": 5.2782,
      "mean_token_accuracy": 0.17842504680156707,
      "num_tokens": 16762631.0,
      "step": 8060
    },
    {
      "entropy": 5.580714368820191,
      "epoch": 0.7315856313497823,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004953721915834473,
      "loss": 5.2878,
      "mean_token_accuracy": 0.18535203933715821,
      "num_tokens": 16772613.0,
      "step": 8065
    },
    {
      "entropy": 5.585743761062622,
      "epoch": 0.7320391872278664,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953656605978067,
      "loss": 5.2995,
      "mean_token_accuracy": 0.17919224500656128,
      "num_tokens": 16782963.0,
      "step": 8070
    },
    {
      "entropy": 5.497876787185669,
      "epoch": 0.7324927431059507,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000495359125054948,
      "loss": 5.2248,
      "mean_token_accuracy": 0.18719250857830047,
      "num_tokens": 16793892.0,
      "step": 8075
    },
    {
      "entropy": 5.433376407623291,
      "epoch": 0.7329462989840348,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004953525849550063,
      "loss": 5.2243,
      "mean_token_accuracy": 0.1793593570590019,
      "num_tokens": 16803388.0,
      "step": 8080
    },
    {
      "entropy": 5.505691719055176,
      "epoch": 0.733399854862119,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953460402981172,
      "loss": 5.1801,
      "mean_token_accuracy": 0.19310829937458038,
      "num_tokens": 16813201.0,
      "step": 8085
    },
    {
      "entropy": 5.602481555938721,
      "epoch": 0.7338534107402032,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953394910844157,
      "loss": 5.4435,
      "mean_token_accuracy": 0.18122223764657974,
      "num_tokens": 16823265.0,
      "step": 8090
    },
    {
      "entropy": 5.53153166770935,
      "epoch": 0.7343069666182874,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953329373140374,
      "loss": 5.2311,
      "mean_token_accuracy": 0.18478136509656906,
      "num_tokens": 16833057.0,
      "step": 8095
    },
    {
      "entropy": 5.486098194122315,
      "epoch": 0.7347605224963716,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004953263789871179,
      "loss": 5.1915,
      "mean_token_accuracy": 0.18385787308216095,
      "num_tokens": 16843488.0,
      "step": 8100
    },
    {
      "entropy": 5.463052749633789,
      "epoch": 0.7352140783744557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953198161037925,
      "loss": 5.2102,
      "mean_token_accuracy": 0.19679825603961945,
      "num_tokens": 16852448.0,
      "step": 8105
    },
    {
      "entropy": 5.431389713287354,
      "epoch": 0.7356676342525399,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004953132486641975,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1843101128935814,
      "num_tokens": 16863944.0,
      "step": 8110
    },
    {
      "entropy": 5.52769889831543,
      "epoch": 0.736121190130624,
      "grad_norm": 1.046875,
      "learning_rate": 0.000495306676668468,
      "loss": 5.1792,
      "mean_token_accuracy": 0.18727706223726273,
      "num_tokens": 16873923.0,
      "step": 8115
    },
    {
      "entropy": 5.547335052490235,
      "epoch": 0.7365747460087083,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953001001167406,
      "loss": 5.2912,
      "mean_token_accuracy": 0.18075185418128967,
      "num_tokens": 16884373.0,
      "step": 8120
    },
    {
      "entropy": 5.570522165298462,
      "epoch": 0.7370283018867925,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004952935190091508,
      "loss": 5.4223,
      "mean_token_accuracy": 0.17607153505086898,
      "num_tokens": 16895869.0,
      "step": 8125
    },
    {
      "entropy": 5.455164957046509,
      "epoch": 0.7374818577648766,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952869333458349,
      "loss": 5.266,
      "mean_token_accuracy": 0.18525791615247728,
      "num_tokens": 16906737.0,
      "step": 8130
    },
    {
      "entropy": 5.48662314414978,
      "epoch": 0.7379354136429608,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952803431269292,
      "loss": 5.1975,
      "mean_token_accuracy": 0.19422420412302016,
      "num_tokens": 16916083.0,
      "step": 8135
    },
    {
      "entropy": 5.597974872589111,
      "epoch": 0.738388969521045,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004952737483525699,
      "loss": 5.3612,
      "mean_token_accuracy": 0.18378081768751145,
      "num_tokens": 16926419.0,
      "step": 8140
    },
    {
      "entropy": 5.5718834400177,
      "epoch": 0.7388425253991292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004952671490228932,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1853596895933151,
      "num_tokens": 16936067.0,
      "step": 8145
    },
    {
      "entropy": 5.585398387908936,
      "epoch": 0.7392960812772134,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004952605451380357,
      "loss": 5.2489,
      "mean_token_accuracy": 0.18981460332870484,
      "num_tokens": 16946571.0,
      "step": 8150
    },
    {
      "entropy": 5.453392887115479,
      "epoch": 0.7397496371552975,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004952539366981342,
      "loss": 5.2966,
      "mean_token_accuracy": 0.18083768635988234,
      "num_tokens": 16956764.0,
      "step": 8155
    },
    {
      "entropy": 5.519579267501831,
      "epoch": 0.7402031930333817,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000495247323703325,
      "loss": 5.2249,
      "mean_token_accuracy": 0.1882965698838234,
      "num_tokens": 16966002.0,
      "step": 8160
    },
    {
      "entropy": 5.572782278060913,
      "epoch": 0.7406567489114659,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495240706153745,
      "loss": 5.3257,
      "mean_token_accuracy": 0.18548445403575897,
      "num_tokens": 16975705.0,
      "step": 8165
    },
    {
      "entropy": 5.522975587844849,
      "epoch": 0.7411103047895501,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952340840495312,
      "loss": 5.2265,
      "mean_token_accuracy": 0.1855018049478531,
      "num_tokens": 16985945.0,
      "step": 8170
    },
    {
      "entropy": 5.4940248966217045,
      "epoch": 0.7415638606676342,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952274573908203,
      "loss": 5.2005,
      "mean_token_accuracy": 0.18602324575185775,
      "num_tokens": 16995475.0,
      "step": 8175
    },
    {
      "entropy": 5.510499334335327,
      "epoch": 0.7420174165457184,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952208261777496,
      "loss": 5.2237,
      "mean_token_accuracy": 0.18946883529424668,
      "num_tokens": 17005028.0,
      "step": 8180
    },
    {
      "entropy": 5.531285905838013,
      "epoch": 0.7424709724238027,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495214190410456,
      "loss": 5.2916,
      "mean_token_accuracy": 0.18801498115062715,
      "num_tokens": 17014786.0,
      "step": 8185
    },
    {
      "entropy": 5.563917732238769,
      "epoch": 0.7429245283018868,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952075500890768,
      "loss": 5.2945,
      "mean_token_accuracy": 0.18754068166017532,
      "num_tokens": 17024431.0,
      "step": 8190
    },
    {
      "entropy": 5.541070652008057,
      "epoch": 0.743378084179971,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952009052137493,
      "loss": 5.2722,
      "mean_token_accuracy": 0.18842438012361526,
      "num_tokens": 17034368.0,
      "step": 8195
    },
    {
      "entropy": 5.53280725479126,
      "epoch": 0.7438316400580551,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951942557846112,
      "loss": 5.3039,
      "mean_token_accuracy": 0.182603819668293,
      "num_tokens": 17043818.0,
      "step": 8200
    },
    {
      "entropy": 5.568868350982666,
      "epoch": 0.7442851959361393,
      "grad_norm": 1.0,
      "learning_rate": 0.0004951876018017997,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18415388315916062,
      "num_tokens": 17054843.0,
      "step": 8205
    },
    {
      "entropy": 5.5368658065795895,
      "epoch": 0.7447387518142236,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004951809432654524,
      "loss": 5.3268,
      "mean_token_accuracy": 0.1834358587861061,
      "num_tokens": 17065539.0,
      "step": 8210
    },
    {
      "entropy": 5.491271352767944,
      "epoch": 0.7451923076923077,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495174280175707,
      "loss": 5.2211,
      "mean_token_accuracy": 0.18677894920110702,
      "num_tokens": 17075443.0,
      "step": 8215
    },
    {
      "entropy": 5.592065238952637,
      "epoch": 0.7456458635703919,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004951676125327015,
      "loss": 5.3549,
      "mean_token_accuracy": 0.18309056311845778,
      "num_tokens": 17085490.0,
      "step": 8220
    },
    {
      "entropy": 5.606427907943726,
      "epoch": 0.746099419448476,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951609403365737,
      "loss": 5.3441,
      "mean_token_accuracy": 0.1845713585615158,
      "num_tokens": 17094803.0,
      "step": 8225
    },
    {
      "entropy": 5.5416947364807125,
      "epoch": 0.7465529753265603,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004951542635874615,
      "loss": 5.2032,
      "mean_token_accuracy": 0.1835956171154976,
      "num_tokens": 17104373.0,
      "step": 8230
    },
    {
      "entropy": 5.5435961246490475,
      "epoch": 0.7470065312046444,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951475822855029,
      "loss": 5.2741,
      "mean_token_accuracy": 0.18503042459487914,
      "num_tokens": 17115469.0,
      "step": 8235
    },
    {
      "entropy": 5.504471731185913,
      "epoch": 0.7474600870827286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004951408964308364,
      "loss": 5.2491,
      "mean_token_accuracy": 0.1830831825733185,
      "num_tokens": 17125443.0,
      "step": 8240
    },
    {
      "entropy": 5.5065467834472654,
      "epoch": 0.7479136429608128,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951342060236,
      "loss": 5.2229,
      "mean_token_accuracy": 0.19070644974708556,
      "num_tokens": 17136253.0,
      "step": 8245
    },
    {
      "entropy": 5.5766904830932615,
      "epoch": 0.7483671988388969,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951275110639321,
      "loss": 5.2796,
      "mean_token_accuracy": 0.18655992895364762,
      "num_tokens": 17145544.0,
      "step": 8250
    },
    {
      "entropy": 5.52774977684021,
      "epoch": 0.7488207547169812,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951208115519712,
      "loss": 5.2905,
      "mean_token_accuracy": 0.18376002162694932,
      "num_tokens": 17155043.0,
      "step": 8255
    },
    {
      "entropy": 5.510969734191894,
      "epoch": 0.7492743105950653,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951141074878558,
      "loss": 5.3221,
      "mean_token_accuracy": 0.183181294798851,
      "num_tokens": 17164885.0,
      "step": 8260
    },
    {
      "entropy": 5.609808874130249,
      "epoch": 0.7497278664731495,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951073988717246,
      "loss": 5.3253,
      "mean_token_accuracy": 0.18268342316150665,
      "num_tokens": 17175681.0,
      "step": 8265
    },
    {
      "entropy": 5.468179035186767,
      "epoch": 0.7501814223512336,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951006857037163,
      "loss": 5.2009,
      "mean_token_accuracy": 0.18825145214796066,
      "num_tokens": 17186002.0,
      "step": 8270
    },
    {
      "entropy": 5.501715898513794,
      "epoch": 0.7506349782293179,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004950939679839696,
      "loss": 5.2434,
      "mean_token_accuracy": 0.18338658064603805,
      "num_tokens": 17195499.0,
      "step": 8275
    },
    {
      "entropy": 5.538295269012451,
      "epoch": 0.7510885341074021,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004950872457126237,
      "loss": 5.2498,
      "mean_token_accuracy": 0.18770517855882646,
      "num_tokens": 17204684.0,
      "step": 8280
    },
    {
      "entropy": 5.519716119766235,
      "epoch": 0.7515420899854862,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950805188898175,
      "loss": 5.3005,
      "mean_token_accuracy": 0.18001003712415695,
      "num_tokens": 17213987.0,
      "step": 8285
    },
    {
      "entropy": 5.542646598815918,
      "epoch": 0.7519956458635704,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049507378751569,
      "loss": 5.3083,
      "mean_token_accuracy": 0.18153341710567475,
      "num_tokens": 17224512.0,
      "step": 8290
    },
    {
      "entropy": 5.592179584503174,
      "epoch": 0.7524492017416545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004950670515903805,
      "loss": 5.3213,
      "mean_token_accuracy": 0.18630932569503783,
      "num_tokens": 17234362.0,
      "step": 8295
    },
    {
      "entropy": 5.5631856441497805,
      "epoch": 0.7529027576197388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950603111140283,
      "loss": 5.3047,
      "mean_token_accuracy": 0.18705005645751954,
      "num_tokens": 17244241.0,
      "step": 8300
    },
    {
      "entropy": 5.497237014770508,
      "epoch": 0.753356313497823,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950535660867728,
      "loss": 5.2429,
      "mean_token_accuracy": 0.1937210038304329,
      "num_tokens": 17254689.0,
      "step": 8305
    },
    {
      "entropy": 5.5569257736206055,
      "epoch": 0.7538098693759071,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950468165087535,
      "loss": 5.3237,
      "mean_token_accuracy": 0.1863400772213936,
      "num_tokens": 17265506.0,
      "step": 8310
    },
    {
      "entropy": 5.533684968948364,
      "epoch": 0.7542634252539913,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004950400623801098,
      "loss": 5.3973,
      "mean_token_accuracy": 0.18226426541805268,
      "num_tokens": 17276376.0,
      "step": 8315
    },
    {
      "entropy": 5.59690899848938,
      "epoch": 0.7547169811320755,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950333037009818,
      "loss": 5.2811,
      "mean_token_accuracy": 0.17492688000202178,
      "num_tokens": 17287005.0,
      "step": 8320
    },
    {
      "entropy": 5.592009592056274,
      "epoch": 0.7551705370101597,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950265404715089,
      "loss": 5.2924,
      "mean_token_accuracy": 0.1817374274134636,
      "num_tokens": 17297595.0,
      "step": 8325
    },
    {
      "entropy": 5.455151844024658,
      "epoch": 0.7556240928882438,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495019772691831,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1883062943816185,
      "num_tokens": 17307319.0,
      "step": 8330
    },
    {
      "entropy": 5.569844818115234,
      "epoch": 0.756077648766328,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950130003620882,
      "loss": 5.3959,
      "mean_token_accuracy": 0.18379779458045958,
      "num_tokens": 17317978.0,
      "step": 8335
    },
    {
      "entropy": 5.532192945480347,
      "epoch": 0.7565312046444121,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950062234824204,
      "loss": 5.2826,
      "mean_token_accuracy": 0.1818419873714447,
      "num_tokens": 17328339.0,
      "step": 8340
    },
    {
      "entropy": 5.53981204032898,
      "epoch": 0.7569847605224964,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949994420529679,
      "loss": 5.3248,
      "mean_token_accuracy": 0.18118658065795898,
      "num_tokens": 17338668.0,
      "step": 8345
    },
    {
      "entropy": 5.4830098152160645,
      "epoch": 0.7574383164005806,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004949926560738709,
      "loss": 5.2692,
      "mean_token_accuracy": 0.17884783148765565,
      "num_tokens": 17348609.0,
      "step": 8350
    },
    {
      "entropy": 5.543556213378906,
      "epoch": 0.7578918722786647,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004949858655452697,
      "loss": 5.2878,
      "mean_token_accuracy": 0.18408870846033096,
      "num_tokens": 17358014.0,
      "step": 8355
    },
    {
      "entropy": 5.600489568710327,
      "epoch": 0.7583454281567489,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004949790704673048,
      "loss": 5.2905,
      "mean_token_accuracy": 0.18661998957395554,
      "num_tokens": 17367975.0,
      "step": 8360
    },
    {
      "entropy": 5.554596185684204,
      "epoch": 0.7587989840348331,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004949722708401166,
      "loss": 5.2372,
      "mean_token_accuracy": 0.18600034564733506,
      "num_tokens": 17377960.0,
      "step": 8365
    },
    {
      "entropy": 5.555735635757446,
      "epoch": 0.7592525399129173,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949654666638457,
      "loss": 5.3269,
      "mean_token_accuracy": 0.18860286325216294,
      "num_tokens": 17387334.0,
      "step": 8370
    },
    {
      "entropy": 5.5340049266815186,
      "epoch": 0.7597060957910015,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494958657938633,
      "loss": 5.2708,
      "mean_token_accuracy": 0.18800106942653655,
      "num_tokens": 17396521.0,
      "step": 8375
    },
    {
      "entropy": 5.478679418563843,
      "epoch": 0.7601596516690856,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949518446646192,
      "loss": 5.2504,
      "mean_token_accuracy": 0.18262986093759537,
      "num_tokens": 17407217.0,
      "step": 8380
    },
    {
      "entropy": 5.634729862213135,
      "epoch": 0.7606132075471698,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949450268419451,
      "loss": 5.3708,
      "mean_token_accuracy": 0.17887072116136551,
      "num_tokens": 17416982.0,
      "step": 8385
    },
    {
      "entropy": 5.479625225067139,
      "epoch": 0.761066763425254,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494938204470752,
      "loss": 5.2013,
      "mean_token_accuracy": 0.18697880059480668,
      "num_tokens": 17427636.0,
      "step": 8390
    },
    {
      "entropy": 5.509683084487915,
      "epoch": 0.7615203193033382,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004949313775511806,
      "loss": 5.3342,
      "mean_token_accuracy": 0.179373337328434,
      "num_tokens": 17437488.0,
      "step": 8395
    },
    {
      "entropy": 5.50303807258606,
      "epoch": 0.7619738751814223,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004949245460833723,
      "loss": 5.1156,
      "mean_token_accuracy": 0.1915082797408104,
      "num_tokens": 17447838.0,
      "step": 8400
    },
    {
      "entropy": 5.572708368301392,
      "epoch": 0.7624274310595065,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949177100674684,
      "loss": 5.34,
      "mean_token_accuracy": 0.18229371309280396,
      "num_tokens": 17457257.0,
      "step": 8405
    },
    {
      "entropy": 5.531822872161865,
      "epoch": 0.7628809869375908,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949108695036102,
      "loss": 5.2602,
      "mean_token_accuracy": 0.18731117099523545,
      "num_tokens": 17467729.0,
      "step": 8410
    },
    {
      "entropy": 5.531580114364624,
      "epoch": 0.7633345428156749,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949040243919392,
      "loss": 5.2793,
      "mean_token_accuracy": 0.18056346476078033,
      "num_tokens": 17477692.0,
      "step": 8415
    },
    {
      "entropy": 5.491760206222534,
      "epoch": 0.7637880986937591,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948971747325968,
      "loss": 5.2807,
      "mean_token_accuracy": 0.18631262183189393,
      "num_tokens": 17488730.0,
      "step": 8420
    },
    {
      "entropy": 5.567406415939331,
      "epoch": 0.7642416545718432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000494890320525725,
      "loss": 5.3518,
      "mean_token_accuracy": 0.18223876208066941,
      "num_tokens": 17498374.0,
      "step": 8425
    },
    {
      "entropy": 5.554366874694824,
      "epoch": 0.7646952104499274,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948834617714652,
      "loss": 5.2025,
      "mean_token_accuracy": 0.1882804214954376,
      "num_tokens": 17508601.0,
      "step": 8430
    },
    {
      "entropy": 5.512381458282471,
      "epoch": 0.7651487663280117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948765984699595,
      "loss": 5.2487,
      "mean_token_accuracy": 0.18550052344799042,
      "num_tokens": 17518164.0,
      "step": 8435
    },
    {
      "entropy": 5.567288446426391,
      "epoch": 0.7656023222060958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004948697306213494,
      "loss": 5.2225,
      "mean_token_accuracy": 0.19012609124183655,
      "num_tokens": 17527244.0,
      "step": 8440
    },
    {
      "entropy": 5.549527502059936,
      "epoch": 0.76605587808418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004948628582257775,
      "loss": 5.2897,
      "mean_token_accuracy": 0.1844577297568321,
      "num_tokens": 17538045.0,
      "step": 8445
    },
    {
      "entropy": 5.501225328445434,
      "epoch": 0.7665094339622641,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948559812833857,
      "loss": 5.2492,
      "mean_token_accuracy": 0.18562017679214476,
      "num_tokens": 17547584.0,
      "step": 8450
    },
    {
      "entropy": 5.550030612945557,
      "epoch": 0.7669629898403484,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004948490997943162,
      "loss": 5.3011,
      "mean_token_accuracy": 0.18173202872276306,
      "num_tokens": 17558925.0,
      "step": 8455
    },
    {
      "entropy": 5.558056306838989,
      "epoch": 0.7674165457184325,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004948422137587111,
      "loss": 5.3712,
      "mean_token_accuracy": 0.17508765757083894,
      "num_tokens": 17569355.0,
      "step": 8460
    },
    {
      "entropy": 5.5319421768188475,
      "epoch": 0.7678701015965167,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004948353231767131,
      "loss": 5.2348,
      "mean_token_accuracy": 0.18913586884737016,
      "num_tokens": 17579782.0,
      "step": 8465
    },
    {
      "entropy": 5.539266347885132,
      "epoch": 0.7683236574746009,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004948284280484645,
      "loss": 5.2996,
      "mean_token_accuracy": 0.19015204161405563,
      "num_tokens": 17589772.0,
      "step": 8470
    },
    {
      "entropy": 5.481799364089966,
      "epoch": 0.768777213352685,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004948215283741081,
      "loss": 5.2051,
      "mean_token_accuracy": 0.18939271271228791,
      "num_tokens": 17599613.0,
      "step": 8475
    },
    {
      "entropy": 5.554580450057983,
      "epoch": 0.7692307692307693,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948146241537866,
      "loss": 5.2984,
      "mean_token_accuracy": 0.180326946079731,
      "num_tokens": 17609056.0,
      "step": 8480
    },
    {
      "entropy": 5.499348592758179,
      "epoch": 0.7696843251088534,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004948077153876425,
      "loss": 5.2124,
      "mean_token_accuracy": 0.1857810825109482,
      "num_tokens": 17618620.0,
      "step": 8485
    },
    {
      "entropy": 5.500620794296265,
      "epoch": 0.7701378809869376,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948008020758188,
      "loss": 5.1845,
      "mean_token_accuracy": 0.19046938568353652,
      "num_tokens": 17627630.0,
      "step": 8490
    },
    {
      "entropy": 5.409573554992676,
      "epoch": 0.7705914368650217,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004947938842184585,
      "loss": 5.1788,
      "mean_token_accuracy": 0.1935472920536995,
      "num_tokens": 17638903.0,
      "step": 8495
    },
    {
      "entropy": 5.469343042373657,
      "epoch": 0.771044992743106,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947869618157048,
      "loss": 5.2378,
      "mean_token_accuracy": 0.18261487334966658,
      "num_tokens": 17649257.0,
      "step": 8500
    },
    {
      "entropy": 5.481932640075684,
      "epoch": 0.7714985486211902,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947800348677006,
      "loss": 5.2129,
      "mean_token_accuracy": 0.19850903004407883,
      "num_tokens": 17659495.0,
      "step": 8505
    },
    {
      "entropy": 5.464404773712158,
      "epoch": 0.7719521044992743,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947731033745893,
      "loss": 5.1901,
      "mean_token_accuracy": 0.18522430658340455,
      "num_tokens": 17669671.0,
      "step": 8510
    },
    {
      "entropy": 5.5648661136627195,
      "epoch": 0.7724056603773585,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947661673365142,
      "loss": 5.3616,
      "mean_token_accuracy": 0.18332356959581375,
      "num_tokens": 17680927.0,
      "step": 8515
    },
    {
      "entropy": 5.506349372863769,
      "epoch": 0.7728592162554426,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004947592267536188,
      "loss": 5.2245,
      "mean_token_accuracy": 0.19083321392536162,
      "num_tokens": 17691378.0,
      "step": 8520
    },
    {
      "entropy": 5.522361421585083,
      "epoch": 0.7733127721335269,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004947522816260466,
      "loss": 5.3236,
      "mean_token_accuracy": 0.18290676027536393,
      "num_tokens": 17700942.0,
      "step": 8525
    },
    {
      "entropy": 5.499392032623291,
      "epoch": 0.773766328011611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947453319539411,
      "loss": 5.2144,
      "mean_token_accuracy": 0.18726398050785065,
      "num_tokens": 17710076.0,
      "step": 8530
    },
    {
      "entropy": 5.495902681350708,
      "epoch": 0.7742198838896952,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947383777374462,
      "loss": 5.2539,
      "mean_token_accuracy": 0.1863068625330925,
      "num_tokens": 17721009.0,
      "step": 8535
    },
    {
      "entropy": 5.479369497299194,
      "epoch": 0.7746734397677794,
      "grad_norm": 0.875,
      "learning_rate": 0.0004947314189767058,
      "loss": 5.2782,
      "mean_token_accuracy": 0.18749626576900483,
      "num_tokens": 17731821.0,
      "step": 8540
    },
    {
      "entropy": 5.466885328292847,
      "epoch": 0.7751269956458636,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947244556718635,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19538922607898712,
      "num_tokens": 17741425.0,
      "step": 8545
    },
    {
      "entropy": 5.487680578231812,
      "epoch": 0.7755805515239478,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947174878230635,
      "loss": 5.2167,
      "mean_token_accuracy": 0.19239004105329513,
      "num_tokens": 17752381.0,
      "step": 8550
    },
    {
      "entropy": 5.448737573623657,
      "epoch": 0.7760341074020319,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947105154304498,
      "loss": 5.2126,
      "mean_token_accuracy": 0.18385005593299866,
      "num_tokens": 17762768.0,
      "step": 8555
    },
    {
      "entropy": 5.496819543838501,
      "epoch": 0.7764876632801161,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004947035384941667,
      "loss": 5.2306,
      "mean_token_accuracy": 0.1835770294070244,
      "num_tokens": 17774032.0,
      "step": 8560
    },
    {
      "entropy": 5.567213773727417,
      "epoch": 0.7769412191582002,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004946965570143584,
      "loss": 5.3774,
      "mean_token_accuracy": 0.17739971727132797,
      "num_tokens": 17785098.0,
      "step": 8565
    },
    {
      "entropy": 5.56190505027771,
      "epoch": 0.7773947750362845,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946895709911695,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1863827884197235,
      "num_tokens": 17795487.0,
      "step": 8570
    },
    {
      "entropy": 5.525628662109375,
      "epoch": 0.7778483309143687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946825804247442,
      "loss": 5.3538,
      "mean_token_accuracy": 0.17812175750732423,
      "num_tokens": 17806045.0,
      "step": 8575
    },
    {
      "entropy": 5.567047452926635,
      "epoch": 0.7783018867924528,
      "grad_norm": 0.890625,
      "learning_rate": 0.000494675585315227,
      "loss": 5.314,
      "mean_token_accuracy": 0.18278959393501282,
      "num_tokens": 17816948.0,
      "step": 8580
    },
    {
      "entropy": 5.596051406860352,
      "epoch": 0.778755442670537,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004946685856627628,
      "loss": 5.2593,
      "mean_token_accuracy": 0.18712812513113022,
      "num_tokens": 17826929.0,
      "step": 8585
    },
    {
      "entropy": 5.555044555664063,
      "epoch": 0.7792089985486212,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946615814674963,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1852679133415222,
      "num_tokens": 17837780.0,
      "step": 8590
    },
    {
      "entropy": 5.394699335098267,
      "epoch": 0.7796625544267054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004946545727295724,
      "loss": 5.1144,
      "mean_token_accuracy": 0.19915736168622972,
      "num_tokens": 17847835.0,
      "step": 8595
    },
    {
      "entropy": 5.485675191879272,
      "epoch": 0.7801161103047896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946475594491359,
      "loss": 5.2043,
      "mean_token_accuracy": 0.19405315816402435,
      "num_tokens": 17858324.0,
      "step": 8600
    },
    {
      "entropy": 5.496988296508789,
      "epoch": 0.7805696661828737,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946405416263319,
      "loss": 5.2818,
      "mean_token_accuracy": 0.17678985744714737,
      "num_tokens": 17868455.0,
      "step": 8605
    },
    {
      "entropy": 5.545461845397949,
      "epoch": 0.7810232220609579,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946335192613054,
      "loss": 5.2465,
      "mean_token_accuracy": 0.18767861723899842,
      "num_tokens": 17879038.0,
      "step": 8610
    },
    {
      "entropy": 5.460289812088012,
      "epoch": 0.7814767779390421,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946264923542019,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19217866361141206,
      "num_tokens": 17888153.0,
      "step": 8615
    },
    {
      "entropy": 5.4844536781311035,
      "epoch": 0.7819303338171263,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004946194609051664,
      "loss": 5.2519,
      "mean_token_accuracy": 0.19185152649879456,
      "num_tokens": 17898289.0,
      "step": 8620
    },
    {
      "entropy": 5.501482677459717,
      "epoch": 0.7823838896952104,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004946124249143447,
      "loss": 5.2547,
      "mean_token_accuracy": 0.18857731074094772,
      "num_tokens": 17909192.0,
      "step": 8625
    },
    {
      "entropy": 5.4674238681793215,
      "epoch": 0.7828374455732946,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494605384381882,
      "loss": 5.2012,
      "mean_token_accuracy": 0.18121700286865233,
      "num_tokens": 17918937.0,
      "step": 8630
    },
    {
      "entropy": 5.441970586776733,
      "epoch": 0.7832910014513788,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945983393079239,
      "loss": 5.2272,
      "mean_token_accuracy": 0.1953953266143799,
      "num_tokens": 17929346.0,
      "step": 8635
    },
    {
      "entropy": 5.503742122650147,
      "epoch": 0.783744557329463,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004945912896926163,
      "loss": 5.2471,
      "mean_token_accuracy": 0.1864580824971199,
      "num_tokens": 17939479.0,
      "step": 8640
    },
    {
      "entropy": 5.592639398574829,
      "epoch": 0.7841981132075472,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945842355361049,
      "loss": 5.3252,
      "mean_token_accuracy": 0.18374345749616622,
      "num_tokens": 17949841.0,
      "step": 8645
    },
    {
      "entropy": 5.471678686141968,
      "epoch": 0.7846516690856313,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945771768385355,
      "loss": 5.2526,
      "mean_token_accuracy": 0.19229422211647035,
      "num_tokens": 17960334.0,
      "step": 8650
    },
    {
      "entropy": 5.478883457183838,
      "epoch": 0.7851052249637155,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494570113600054,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19275954067707063,
      "num_tokens": 17970595.0,
      "step": 8655
    },
    {
      "entropy": 5.50038251876831,
      "epoch": 0.7855587808417998,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004945630458208068,
      "loss": 5.3421,
      "mean_token_accuracy": 0.17421732395887374,
      "num_tokens": 17981301.0,
      "step": 8660
    },
    {
      "entropy": 5.457895803451538,
      "epoch": 0.7860123367198839,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945559735009397,
      "loss": 5.1658,
      "mean_token_accuracy": 0.191603122651577,
      "num_tokens": 17991262.0,
      "step": 8665
    },
    {
      "entropy": 5.568162202835083,
      "epoch": 0.7864658925979681,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945488966405993,
      "loss": 5.2024,
      "mean_token_accuracy": 0.18861326873302459,
      "num_tokens": 18001212.0,
      "step": 8670
    },
    {
      "entropy": 5.501408624649048,
      "epoch": 0.7869194484760522,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004945418152399317,
      "loss": 5.2229,
      "mean_token_accuracy": 0.19007271379232407,
      "num_tokens": 18010432.0,
      "step": 8675
    },
    {
      "entropy": 5.5106878757476805,
      "epoch": 0.7873730043541364,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945347292990833,
      "loss": 5.2655,
      "mean_token_accuracy": 0.18418308347463608,
      "num_tokens": 18020902.0,
      "step": 8680
    },
    {
      "entropy": 5.512370157241821,
      "epoch": 0.7878265602322206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004945276388182008,
      "loss": 5.132,
      "mean_token_accuracy": 0.1858739361166954,
      "num_tokens": 18031133.0,
      "step": 8685
    },
    {
      "entropy": 5.567989873886108,
      "epoch": 0.7882801161103048,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945205437974309,
      "loss": 5.2472,
      "mean_token_accuracy": 0.18936766386032106,
      "num_tokens": 18042050.0,
      "step": 8690
    },
    {
      "entropy": 5.428337478637696,
      "epoch": 0.788733671988389,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004945134442369201,
      "loss": 5.214,
      "mean_token_accuracy": 0.19088352620601653,
      "num_tokens": 18051642.0,
      "step": 8695
    },
    {
      "entropy": 5.552801275253296,
      "epoch": 0.7891872278664731,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945063401368154,
      "loss": 5.2664,
      "mean_token_accuracy": 0.1811126798391342,
      "num_tokens": 18062180.0,
      "step": 8700
    },
    {
      "entropy": 5.50035719871521,
      "epoch": 0.7896407837445574,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944992314972637,
      "loss": 5.2154,
      "mean_token_accuracy": 0.18759729713201523,
      "num_tokens": 18072302.0,
      "step": 8705
    },
    {
      "entropy": 5.486184597015381,
      "epoch": 0.7900943396226415,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004944921183184118,
      "loss": 5.2179,
      "mean_token_accuracy": 0.18937012553215027,
      "num_tokens": 18081824.0,
      "step": 8710
    },
    {
      "entropy": 5.53223443031311,
      "epoch": 0.7905478955007257,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494485000600407,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1890762209892273,
      "num_tokens": 18092696.0,
      "step": 8715
    },
    {
      "entropy": 5.482760906219482,
      "epoch": 0.7910014513788098,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944778783433965,
      "loss": 5.1892,
      "mean_token_accuracy": 0.1936694249510765,
      "num_tokens": 18101790.0,
      "step": 8720
    },
    {
      "entropy": 5.550623178482056,
      "epoch": 0.791455007256894,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944707515475276,
      "loss": 5.3132,
      "mean_token_accuracy": 0.1843470126390457,
      "num_tokens": 18111572.0,
      "step": 8725
    },
    {
      "entropy": 5.5540313720703125,
      "epoch": 0.7919085631349783,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944636202129475,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1926032617688179,
      "num_tokens": 18120650.0,
      "step": 8730
    },
    {
      "entropy": 5.47811222076416,
      "epoch": 0.7923621190130624,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004944564843398039,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1985896944999695,
      "num_tokens": 18131055.0,
      "step": 8735
    },
    {
      "entropy": 5.573776483535767,
      "epoch": 0.7928156748911466,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004944493439282443,
      "loss": 5.2739,
      "mean_token_accuracy": 0.18772944211959838,
      "num_tokens": 18141423.0,
      "step": 8740
    },
    {
      "entropy": 5.4522857666015625,
      "epoch": 0.7932692307692307,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004944421989784162,
      "loss": 5.1932,
      "mean_token_accuracy": 0.18533381223678588,
      "num_tokens": 18152015.0,
      "step": 8745
    },
    {
      "entropy": 5.539948177337647,
      "epoch": 0.793722786647315,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004944350494904676,
      "loss": 5.2892,
      "mean_token_accuracy": 0.18949988037347792,
      "num_tokens": 18161934.0,
      "step": 8750
    },
    {
      "entropy": 5.496873092651367,
      "epoch": 0.7941763425253991,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944278954645463,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18725937306880952,
      "num_tokens": 18172287.0,
      "step": 8755
    },
    {
      "entropy": 5.482796335220337,
      "epoch": 0.7946298984034833,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004944207369008002,
      "loss": 5.2505,
      "mean_token_accuracy": 0.19790160655975342,
      "num_tokens": 18181015.0,
      "step": 8760
    },
    {
      "entropy": 5.577034568786621,
      "epoch": 0.7950834542815675,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004944135737993773,
      "loss": 5.3461,
      "mean_token_accuracy": 0.18178513646125793,
      "num_tokens": 18192382.0,
      "step": 8765
    },
    {
      "entropy": 5.488105583190918,
      "epoch": 0.7955370101596516,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004944064061604258,
      "loss": 5.2109,
      "mean_token_accuracy": 0.19367620944976807,
      "num_tokens": 18202597.0,
      "step": 8770
    },
    {
      "entropy": 5.523237991333008,
      "epoch": 0.7959905660377359,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943992339840939,
      "loss": 5.2774,
      "mean_token_accuracy": 0.18589625358581544,
      "num_tokens": 18213756.0,
      "step": 8775
    },
    {
      "entropy": 5.434082651138306,
      "epoch": 0.79644412191582,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049439205727053,
      "loss": 5.1884,
      "mean_token_accuracy": 0.18429725021123886,
      "num_tokens": 18223669.0,
      "step": 8780
    },
    {
      "entropy": 5.497088479995727,
      "epoch": 0.7968976777939042,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004943848760198823,
      "loss": 5.2104,
      "mean_token_accuracy": 0.19213344007730485,
      "num_tokens": 18234884.0,
      "step": 8785
    },
    {
      "entropy": 5.466268301010132,
      "epoch": 0.7973512336719883,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943776902322995,
      "loss": 5.1506,
      "mean_token_accuracy": 0.1955432653427124,
      "num_tokens": 18245155.0,
      "step": 8790
    },
    {
      "entropy": 5.37814736366272,
      "epoch": 0.7978047895500726,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943704999079302,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19864733517169952,
      "num_tokens": 18254686.0,
      "step": 8795
    },
    {
      "entropy": 5.471942377090454,
      "epoch": 0.7982583454281568,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943633050469231,
      "loss": 5.121,
      "mean_token_accuracy": 0.19447069764137268,
      "num_tokens": 18265473.0,
      "step": 8800
    },
    {
      "entropy": 5.53497724533081,
      "epoch": 0.7987119013062409,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943561056494269,
      "loss": 5.3336,
      "mean_token_accuracy": 0.18408775627613067,
      "num_tokens": 18276024.0,
      "step": 8805
    },
    {
      "entropy": 5.5106282234191895,
      "epoch": 0.7991654571843251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004943489017155906,
      "loss": 5.2888,
      "mean_token_accuracy": 0.1785351738333702,
      "num_tokens": 18287261.0,
      "step": 8810
    },
    {
      "entropy": 5.560506153106689,
      "epoch": 0.7996190130624092,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004943416932455631,
      "loss": 5.305,
      "mean_token_accuracy": 0.18382198810577394,
      "num_tokens": 18299374.0,
      "step": 8815
    },
    {
      "entropy": 5.472782135009766,
      "epoch": 0.8000725689404935,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004943344802394935,
      "loss": 5.2679,
      "mean_token_accuracy": 0.1808692380785942,
      "num_tokens": 18309344.0,
      "step": 8820
    },
    {
      "entropy": 5.548999357223511,
      "epoch": 0.8005261248185777,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943272626975309,
      "loss": 5.2525,
      "mean_token_accuracy": 0.18682713359594344,
      "num_tokens": 18319778.0,
      "step": 8825
    },
    {
      "entropy": 5.542529487609864,
      "epoch": 0.8009796806966618,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004943200406198246,
      "loss": 5.2544,
      "mean_token_accuracy": 0.18820199817419053,
      "num_tokens": 18329787.0,
      "step": 8830
    },
    {
      "entropy": 5.463485479354858,
      "epoch": 0.801433236574746,
      "grad_norm": 0.921875,
      "learning_rate": 0.000494312814006524,
      "loss": 5.2068,
      "mean_token_accuracy": 0.19309862405061723,
      "num_tokens": 18340670.0,
      "step": 8835
    },
    {
      "entropy": 5.48480076789856,
      "epoch": 0.8018867924528302,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004943055828577784,
      "loss": 5.2014,
      "mean_token_accuracy": 0.18602313101291656,
      "num_tokens": 18352484.0,
      "step": 8840
    },
    {
      "entropy": 5.469893169403076,
      "epoch": 0.8023403483309144,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004942983471737376,
      "loss": 5.1777,
      "mean_token_accuracy": 0.19207196831703185,
      "num_tokens": 18363103.0,
      "step": 8845
    },
    {
      "entropy": 5.4201044082641605,
      "epoch": 0.8027939042089985,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004942911069545511,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19178254008293152,
      "num_tokens": 18373925.0,
      "step": 8850
    },
    {
      "entropy": 5.5190746784210205,
      "epoch": 0.8032474600870827,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942838622003686,
      "loss": 5.288,
      "mean_token_accuracy": 0.18306922763586045,
      "num_tokens": 18384081.0,
      "step": 8855
    },
    {
      "entropy": 5.582727241516113,
      "epoch": 0.8037010159651669,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942766129113399,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19331018328666688,
      "num_tokens": 18392510.0,
      "step": 8860
    },
    {
      "entropy": 5.567781639099121,
      "epoch": 0.8041545718432511,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942693590876151,
      "loss": 5.3186,
      "mean_token_accuracy": 0.182000532746315,
      "num_tokens": 18402809.0,
      "step": 8865
    },
    {
      "entropy": 5.5828429698944095,
      "epoch": 0.8046081277213353,
      "grad_norm": 1.078125,
      "learning_rate": 0.000494262100729344,
      "loss": 5.3322,
      "mean_token_accuracy": 0.18444409668445588,
      "num_tokens": 18413342.0,
      "step": 8870
    },
    {
      "entropy": 5.521259212493897,
      "epoch": 0.8050616835994194,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004942548378366768,
      "loss": 5.3185,
      "mean_token_accuracy": 0.1856078937649727,
      "num_tokens": 18424735.0,
      "step": 8875
    },
    {
      "entropy": 5.5181769847869875,
      "epoch": 0.8055152394775036,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942475704097635,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18817635774612426,
      "num_tokens": 18434797.0,
      "step": 8880
    },
    {
      "entropy": 5.514129638671875,
      "epoch": 0.8059687953555879,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942402984487548,
      "loss": 5.2634,
      "mean_token_accuracy": 0.18100712597370147,
      "num_tokens": 18445456.0,
      "step": 8885
    },
    {
      "entropy": 5.485908126831054,
      "epoch": 0.806422351233672,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004942330219538008,
      "loss": 5.2398,
      "mean_token_accuracy": 0.18712347149848937,
      "num_tokens": 18456728.0,
      "step": 8890
    },
    {
      "entropy": 5.682804965972901,
      "epoch": 0.8068759071117562,
      "grad_norm": 1.125,
      "learning_rate": 0.000494225740925052,
      "loss": 5.3729,
      "mean_token_accuracy": 0.17928425073623658,
      "num_tokens": 18468813.0,
      "step": 8895
    },
    {
      "entropy": 5.553705883026123,
      "epoch": 0.8073294629898403,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004942184553626591,
      "loss": 5.2608,
      "mean_token_accuracy": 0.185865418612957,
      "num_tokens": 18477912.0,
      "step": 8900
    },
    {
      "entropy": 5.464331388473511,
      "epoch": 0.8077830188679245,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942111652667726,
      "loss": 5.2155,
      "mean_token_accuracy": 0.19424942433834075,
      "num_tokens": 18488115.0,
      "step": 8905
    },
    {
      "entropy": 5.492030239105224,
      "epoch": 0.8082365747460087,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942038706375433,
      "loss": 5.2499,
      "mean_token_accuracy": 0.18399515151977539,
      "num_tokens": 18498853.0,
      "step": 8910
    },
    {
      "entropy": 5.5277331352233885,
      "epoch": 0.8086901306240929,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941965714751222,
      "loss": 5.2947,
      "mean_token_accuracy": 0.18574539422988892,
      "num_tokens": 18510047.0,
      "step": 8915
    },
    {
      "entropy": 5.511319303512574,
      "epoch": 0.809143686502177,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941892677796601,
      "loss": 5.2691,
      "mean_token_accuracy": 0.18705060333013535,
      "num_tokens": 18519673.0,
      "step": 8920
    },
    {
      "entropy": 5.497614240646362,
      "epoch": 0.8095972423802612,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004941819595513081,
      "loss": 5.2878,
      "mean_token_accuracy": 0.18286352902650832,
      "num_tokens": 18530090.0,
      "step": 8925
    },
    {
      "entropy": 5.597482538223266,
      "epoch": 0.8100507982583455,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004941746467902173,
      "loss": 5.3243,
      "mean_token_accuracy": 0.1817130759358406,
      "num_tokens": 18541862.0,
      "step": 8930
    },
    {
      "entropy": 5.573345136642456,
      "epoch": 0.8105043541364296,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000494167329496539,
      "loss": 5.3369,
      "mean_token_accuracy": 0.17775724679231644,
      "num_tokens": 18551624.0,
      "step": 8935
    },
    {
      "entropy": 5.514851140975952,
      "epoch": 0.8109579100145138,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941600076704244,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18574765473604202,
      "num_tokens": 18561784.0,
      "step": 8940
    },
    {
      "entropy": 5.55006742477417,
      "epoch": 0.8114114658925979,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941526813120251,
      "loss": 5.318,
      "mean_token_accuracy": 0.1817800968885422,
      "num_tokens": 18572110.0,
      "step": 8945
    },
    {
      "entropy": 5.550695466995239,
      "epoch": 0.8118650217706821,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941453504214924,
      "loss": 5.2702,
      "mean_token_accuracy": 0.1783347800374031,
      "num_tokens": 18582012.0,
      "step": 8950
    },
    {
      "entropy": 5.4474639892578125,
      "epoch": 0.8123185776487664,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004941380149989781,
      "loss": 5.1115,
      "mean_token_accuracy": 0.1958514556288719,
      "num_tokens": 18592850.0,
      "step": 8955
    },
    {
      "entropy": 5.4674022674560545,
      "epoch": 0.8127721335268505,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004941306750446336,
      "loss": 5.2945,
      "mean_token_accuracy": 0.1863408148288727,
      "num_tokens": 18602590.0,
      "step": 8960
    },
    {
      "entropy": 5.463942813873291,
      "epoch": 0.8132256894049347,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494123330558611,
      "loss": 5.1962,
      "mean_token_accuracy": 0.19270357340574265,
      "num_tokens": 18612607.0,
      "step": 8965
    },
    {
      "entropy": 5.4990180969238285,
      "epoch": 0.8136792452830188,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941159815410622,
      "loss": 5.1769,
      "mean_token_accuracy": 0.19223012775182724,
      "num_tokens": 18622936.0,
      "step": 8970
    },
    {
      "entropy": 5.516388368606568,
      "epoch": 0.8141328011611031,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000494108627992139,
      "loss": 5.3316,
      "mean_token_accuracy": 0.18003688007593155,
      "num_tokens": 18633997.0,
      "step": 8975
    },
    {
      "entropy": 5.481217002868652,
      "epoch": 0.8145863570391872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004941012699119936,
      "loss": 5.1633,
      "mean_token_accuracy": 0.19657521694898605,
      "num_tokens": 18644347.0,
      "step": 8980
    },
    {
      "entropy": 5.528944206237793,
      "epoch": 0.8150399129172714,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004940939073007779,
      "loss": 5.2399,
      "mean_token_accuracy": 0.18880648016929627,
      "num_tokens": 18654989.0,
      "step": 8985
    },
    {
      "entropy": 5.4825615882873535,
      "epoch": 0.8154934687953556,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940865401586446,
      "loss": 5.2116,
      "mean_token_accuracy": 0.19149848967790603,
      "num_tokens": 18665526.0,
      "step": 8990
    },
    {
      "entropy": 5.465846824645996,
      "epoch": 0.8159470246734397,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940791684857458,
      "loss": 5.2131,
      "mean_token_accuracy": 0.19184787720441818,
      "num_tokens": 18676226.0,
      "step": 8995
    },
    {
      "entropy": 5.5053871154785154,
      "epoch": 0.816400580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940717922822339,
      "loss": 5.1511,
      "mean_token_accuracy": 0.1933094546198845,
      "num_tokens": 18686499.0,
      "step": 9000
    },
    {
      "epoch": 0.816400580551524,
      "eval_entropy": 5.358133128541627,
      "eval_loss": 5.252007961273193,
      "eval_mean_token_accuracy": 0.19481117964819142,
      "eval_num_tokens": 18686499.0,
      "eval_runtime": 20.7388,
      "eval_samples_per_second": 1705.014,
      "eval_steps_per_second": 213.127,
      "step": 9000
    },
    {
      "entropy": 5.569316291809082,
      "epoch": 0.8168541364296081,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004940644115482615,
      "loss": 5.2286,
      "mean_token_accuracy": 0.18744052350521087,
      "num_tokens": 18696543.0,
      "step": 9005
    },
    {
      "entropy": 5.459391164779663,
      "epoch": 0.8173076923076923,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940570262839812,
      "loss": 5.2701,
      "mean_token_accuracy": 0.18510979413986206,
      "num_tokens": 18707029.0,
      "step": 9010
    },
    {
      "entropy": 5.494412612915039,
      "epoch": 0.8177612481857764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940496364895457,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18970807492733002,
      "num_tokens": 18716937.0,
      "step": 9015
    },
    {
      "entropy": 5.612600326538086,
      "epoch": 0.8182148040638607,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940422421651082,
      "loss": 5.2543,
      "mean_token_accuracy": 0.1858663633465767,
      "num_tokens": 18727478.0,
      "step": 9020
    },
    {
      "entropy": 5.491196346282959,
      "epoch": 0.8186683599419449,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004940348433108211,
      "loss": 5.1684,
      "mean_token_accuracy": 0.18780348002910613,
      "num_tokens": 18737339.0,
      "step": 9025
    },
    {
      "entropy": 5.48309326171875,
      "epoch": 0.819121915820029,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004940274399268377,
      "loss": 5.276,
      "mean_token_accuracy": 0.18714603930711746,
      "num_tokens": 18747810.0,
      "step": 9030
    },
    {
      "entropy": 5.493037223815918,
      "epoch": 0.8195754716981132,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940200320133109,
      "loss": 5.2579,
      "mean_token_accuracy": 0.18732280135154725,
      "num_tokens": 18757564.0,
      "step": 9035
    },
    {
      "entropy": 5.595683336257935,
      "epoch": 0.8200290275761973,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940126195703941,
      "loss": 5.2798,
      "mean_token_accuracy": 0.18251087218523027,
      "num_tokens": 18768196.0,
      "step": 9040
    },
    {
      "entropy": 5.569827604293823,
      "epoch": 0.8204825834542816,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004940052025982405,
      "loss": 5.2491,
      "mean_token_accuracy": 0.19636135995388032,
      "num_tokens": 18778350.0,
      "step": 9045
    },
    {
      "entropy": 5.475778579711914,
      "epoch": 0.8209361393323658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004939977810970035,
      "loss": 5.2048,
      "mean_token_accuracy": 0.18468899726867677,
      "num_tokens": 18789825.0,
      "step": 9050
    },
    {
      "entropy": 5.49319748878479,
      "epoch": 0.8213896952104499,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939903550668366,
      "loss": 5.2059,
      "mean_token_accuracy": 0.18663357943296432,
      "num_tokens": 18799793.0,
      "step": 9055
    },
    {
      "entropy": 5.511374473571777,
      "epoch": 0.8218432510885341,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004939829245078934,
      "loss": 5.2397,
      "mean_token_accuracy": 0.19193035662174224,
      "num_tokens": 18809771.0,
      "step": 9060
    },
    {
      "entropy": 5.4916544437408445,
      "epoch": 0.8222968069666183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004939754894203275,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18900091648101808,
      "num_tokens": 18820529.0,
      "step": 9065
    },
    {
      "entropy": 5.425688123703003,
      "epoch": 0.8227503628447025,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939680498042926,
      "loss": 5.1706,
      "mean_token_accuracy": 0.19591638892889024,
      "num_tokens": 18830797.0,
      "step": 9070
    },
    {
      "entropy": 5.465984773635864,
      "epoch": 0.8232039187227866,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004939606056599428,
      "loss": 5.1989,
      "mean_token_accuracy": 0.1867137596011162,
      "num_tokens": 18840526.0,
      "step": 9075
    },
    {
      "entropy": 5.483081436157226,
      "epoch": 0.8236574746008708,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004939531569874316,
      "loss": 5.2148,
      "mean_token_accuracy": 0.1837407886981964,
      "num_tokens": 18851201.0,
      "step": 9080
    },
    {
      "entropy": 5.4127593517303465,
      "epoch": 0.824111030478955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939457037869135,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20760098844766617,
      "num_tokens": 18861676.0,
      "step": 9085
    },
    {
      "entropy": 5.430079507827759,
      "epoch": 0.8245645863570392,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004939382460585424,
      "loss": 5.178,
      "mean_token_accuracy": 0.184494811296463,
      "num_tokens": 18872669.0,
      "step": 9090
    },
    {
      "entropy": 5.434750080108643,
      "epoch": 0.8250181422351234,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939307838024726,
      "loss": 5.2201,
      "mean_token_accuracy": 0.1835554853081703,
      "num_tokens": 18882130.0,
      "step": 9095
    },
    {
      "entropy": 5.456328392028809,
      "epoch": 0.8254716981132075,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004939233170188585,
      "loss": 5.2109,
      "mean_token_accuracy": 0.1889977216720581,
      "num_tokens": 18892457.0,
      "step": 9100
    },
    {
      "entropy": 5.537774133682251,
      "epoch": 0.8259252539912917,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004939158457078543,
      "loss": 5.2316,
      "mean_token_accuracy": 0.18953631073236465,
      "num_tokens": 18903152.0,
      "step": 9105
    },
    {
      "entropy": 5.513712215423584,
      "epoch": 0.826378809869376,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004939083698696146,
      "loss": 5.1924,
      "mean_token_accuracy": 0.18583756536245347,
      "num_tokens": 18913740.0,
      "step": 9110
    },
    {
      "entropy": 5.487645053863526,
      "epoch": 0.8268323657474601,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493900889504294,
      "loss": 5.2403,
      "mean_token_accuracy": 0.18907872438430787,
      "num_tokens": 18923501.0,
      "step": 9115
    },
    {
      "entropy": 5.49286789894104,
      "epoch": 0.8272859216255443,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004938934046120473,
      "loss": 5.2371,
      "mean_token_accuracy": 0.18701096922159194,
      "num_tokens": 18933255.0,
      "step": 9120
    },
    {
      "entropy": 5.486018419265747,
      "epoch": 0.8277394775036284,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004938859151930293,
      "loss": 5.1943,
      "mean_token_accuracy": 0.19105144888162612,
      "num_tokens": 18943649.0,
      "step": 9125
    },
    {
      "entropy": 5.471429109573364,
      "epoch": 0.8281930333817126,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938784212473947,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19930363148450853,
      "num_tokens": 18953187.0,
      "step": 9130
    },
    {
      "entropy": 5.544489240646362,
      "epoch": 0.8286465892597968,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938709227752985,
      "loss": 5.2041,
      "mean_token_accuracy": 0.1965079352259636,
      "num_tokens": 18963015.0,
      "step": 9135
    },
    {
      "entropy": 5.482017135620117,
      "epoch": 0.829100145137881,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493863419776896,
      "loss": 5.2664,
      "mean_token_accuracy": 0.1819402039051056,
      "num_tokens": 18974184.0,
      "step": 9140
    },
    {
      "entropy": 5.475565242767334,
      "epoch": 0.8295537010159652,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004938559122523421,
      "loss": 5.2073,
      "mean_token_accuracy": 0.1881326138973236,
      "num_tokens": 18985036.0,
      "step": 9145
    },
    {
      "entropy": 5.559447145462036,
      "epoch": 0.8300072568940493,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938484002017921,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1855930745601654,
      "num_tokens": 18995237.0,
      "step": 9150
    },
    {
      "entropy": 5.519651746749878,
      "epoch": 0.8304608127721336,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938408836254015,
      "loss": 5.3199,
      "mean_token_accuracy": 0.17952737510204314,
      "num_tokens": 19006272.0,
      "step": 9155
    },
    {
      "entropy": 5.461530971527099,
      "epoch": 0.8309143686502177,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938333625233258,
      "loss": 5.2331,
      "mean_token_accuracy": 0.18802201896905898,
      "num_tokens": 19016728.0,
      "step": 9160
    },
    {
      "entropy": 5.531713771820068,
      "epoch": 0.8313679245283019,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004938258368957202,
      "loss": 5.2856,
      "mean_token_accuracy": 0.17781373262405395,
      "num_tokens": 19028024.0,
      "step": 9165
    },
    {
      "entropy": 5.528930377960205,
      "epoch": 0.831821480406386,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938183067427406,
      "loss": 5.2252,
      "mean_token_accuracy": 0.1947782501578331,
      "num_tokens": 19038218.0,
      "step": 9170
    },
    {
      "entropy": 5.461911010742187,
      "epoch": 0.8322750362844702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004938107720645426,
      "loss": 5.254,
      "mean_token_accuracy": 0.18719436675310136,
      "num_tokens": 19048574.0,
      "step": 9175
    },
    {
      "entropy": 5.498357486724854,
      "epoch": 0.8327285921625545,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938032328612821,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18767004311084748,
      "num_tokens": 19057721.0,
      "step": 9180
    },
    {
      "entropy": 5.501523447036743,
      "epoch": 0.8331821480406386,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493795689133115,
      "loss": 5.2002,
      "mean_token_accuracy": 0.1808902621269226,
      "num_tokens": 19067582.0,
      "step": 9185
    },
    {
      "entropy": 5.48468132019043,
      "epoch": 0.8336357039187228,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937881408801974,
      "loss": 5.1627,
      "mean_token_accuracy": 0.19163392037153243,
      "num_tokens": 19077896.0,
      "step": 9190
    },
    {
      "entropy": 5.507964372634888,
      "epoch": 0.8340892597968069,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493780588102685,
      "loss": 5.2563,
      "mean_token_accuracy": 0.192536298930645,
      "num_tokens": 19088082.0,
      "step": 9195
    },
    {
      "entropy": 5.55622878074646,
      "epoch": 0.8345428156748912,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937730308007346,
      "loss": 5.2814,
      "mean_token_accuracy": 0.18960726112127305,
      "num_tokens": 19097176.0,
      "step": 9200
    },
    {
      "entropy": 5.497657394409179,
      "epoch": 0.8349963715529753,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004937654689745021,
      "loss": 5.1633,
      "mean_token_accuracy": 0.19420606195926665,
      "num_tokens": 19107835.0,
      "step": 9205
    },
    {
      "entropy": 5.424288654327393,
      "epoch": 0.8354499274310595,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004937579026241439,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19160815477371215,
      "num_tokens": 19118028.0,
      "step": 9210
    },
    {
      "entropy": 5.48676643371582,
      "epoch": 0.8359034833091437,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004937503317498165,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19056888222694396,
      "num_tokens": 19129223.0,
      "step": 9215
    },
    {
      "entropy": 5.492285490036011,
      "epoch": 0.8363570391872278,
      "grad_norm": 0.875,
      "learning_rate": 0.0004937427563516765,
      "loss": 5.3239,
      "mean_token_accuracy": 0.17992374151945115,
      "num_tokens": 19140048.0,
      "step": 9220
    },
    {
      "entropy": 5.564370107650757,
      "epoch": 0.8368105950653121,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937351764298805,
      "loss": 5.287,
      "mean_token_accuracy": 0.178375805914402,
      "num_tokens": 19150853.0,
      "step": 9225
    },
    {
      "entropy": 5.543662738800049,
      "epoch": 0.8372641509433962,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937275919845855,
      "loss": 5.2718,
      "mean_token_accuracy": 0.18463384956121445,
      "num_tokens": 19161645.0,
      "step": 9230
    },
    {
      "entropy": 5.514537143707275,
      "epoch": 0.8377177068214804,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004937200030159481,
      "loss": 5.3572,
      "mean_token_accuracy": 0.18064510822296143,
      "num_tokens": 19173276.0,
      "step": 9235
    },
    {
      "entropy": 5.551925563812256,
      "epoch": 0.8381712626995645,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937124095241252,
      "loss": 5.2393,
      "mean_token_accuracy": 0.19420024901628494,
      "num_tokens": 19182949.0,
      "step": 9240
    },
    {
      "entropy": 5.564868927001953,
      "epoch": 0.8386248185776488,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004937048115092739,
      "loss": 5.2393,
      "mean_token_accuracy": 0.1857432618737221,
      "num_tokens": 19193504.0,
      "step": 9245
    },
    {
      "entropy": 5.488748836517334,
      "epoch": 0.839078374455733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004936972089715515,
      "loss": 5.3068,
      "mean_token_accuracy": 0.1808287188410759,
      "num_tokens": 19203406.0,
      "step": 9250
    },
    {
      "entropy": 5.481642866134644,
      "epoch": 0.8395319303338171,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493689601911115,
      "loss": 5.2549,
      "mean_token_accuracy": 0.18709076792001725,
      "num_tokens": 19214325.0,
      "step": 9255
    },
    {
      "entropy": 5.523412084579467,
      "epoch": 0.8399854862119013,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936819903281219,
      "loss": 5.3033,
      "mean_token_accuracy": 0.17142430394887925,
      "num_tokens": 19226466.0,
      "step": 9260
    },
    {
      "entropy": 5.605291700363159,
      "epoch": 0.8404390420899854,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004936743742227295,
      "loss": 5.3239,
      "mean_token_accuracy": 0.18541250377893448,
      "num_tokens": 19237032.0,
      "step": 9265
    },
    {
      "entropy": 5.553051137924195,
      "epoch": 0.8408925979680697,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004936667535950952,
      "loss": 5.3008,
      "mean_token_accuracy": 0.18086010366678237,
      "num_tokens": 19248009.0,
      "step": 9270
    },
    {
      "entropy": 5.570973443984985,
      "epoch": 0.8413461538461539,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936591284453768,
      "loss": 5.3585,
      "mean_token_accuracy": 0.18207215815782546,
      "num_tokens": 19259363.0,
      "step": 9275
    },
    {
      "entropy": 5.544232416152954,
      "epoch": 0.841799709724238,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936514987737319,
      "loss": 5.2609,
      "mean_token_accuracy": 0.18669627606868744,
      "num_tokens": 19269191.0,
      "step": 9280
    },
    {
      "entropy": 5.55557804107666,
      "epoch": 0.8422532656023222,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004936438645803182,
      "loss": 5.2303,
      "mean_token_accuracy": 0.18535475730895995,
      "num_tokens": 19280278.0,
      "step": 9285
    },
    {
      "entropy": 5.522267198562622,
      "epoch": 0.8427068214804064,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004936362258652938,
      "loss": 5.2129,
      "mean_token_accuracy": 0.1926440954208374,
      "num_tokens": 19290130.0,
      "step": 9290
    },
    {
      "entropy": 5.5636334896087645,
      "epoch": 0.8431603773584906,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004936285826288165,
      "loss": 5.3232,
      "mean_token_accuracy": 0.19174562841653825,
      "num_tokens": 19301669.0,
      "step": 9295
    },
    {
      "entropy": 5.453507995605468,
      "epoch": 0.8436139332365747,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004936209348710444,
      "loss": 5.1815,
      "mean_token_accuracy": 0.195254348218441,
      "num_tokens": 19312490.0,
      "step": 9300
    },
    {
      "entropy": 5.478431653976441,
      "epoch": 0.8440674891146589,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004936132825921355,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1872256428003311,
      "num_tokens": 19321743.0,
      "step": 9305
    },
    {
      "entropy": 5.496931743621826,
      "epoch": 0.844521044992743,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004936056257922484,
      "loss": 5.1403,
      "mean_token_accuracy": 0.193279492855072,
      "num_tokens": 19332149.0,
      "step": 9310
    },
    {
      "entropy": 5.508009672164917,
      "epoch": 0.8449746008708273,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935979644715412,
      "loss": 5.2505,
      "mean_token_accuracy": 0.18201816231012344,
      "num_tokens": 19341655.0,
      "step": 9315
    },
    {
      "entropy": 5.538322639465332,
      "epoch": 0.8454281567489115,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935902986301724,
      "loss": 5.282,
      "mean_token_accuracy": 0.1822471797466278,
      "num_tokens": 19351404.0,
      "step": 9320
    },
    {
      "entropy": 5.544085264205933,
      "epoch": 0.8458817126269956,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935826282683005,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18208804428577424,
      "num_tokens": 19361788.0,
      "step": 9325
    },
    {
      "entropy": 5.524391937255859,
      "epoch": 0.8463352685050798,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004935749533860842,
      "loss": 5.2728,
      "mean_token_accuracy": 0.18818075209856033,
      "num_tokens": 19372902.0,
      "step": 9330
    },
    {
      "entropy": 5.45745267868042,
      "epoch": 0.846788824383164,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493567273983682,
      "loss": 5.1709,
      "mean_token_accuracy": 0.1850125476717949,
      "num_tokens": 19382181.0,
      "step": 9335
    },
    {
      "entropy": 5.514938163757324,
      "epoch": 0.8472423802612482,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000493559590061253,
      "loss": 5.229,
      "mean_token_accuracy": 0.18603635281324388,
      "num_tokens": 19393692.0,
      "step": 9340
    },
    {
      "entropy": 5.539198207855224,
      "epoch": 0.8476959361393324,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004935519016189559,
      "loss": 5.2581,
      "mean_token_accuracy": 0.19449927061796188,
      "num_tokens": 19404587.0,
      "step": 9345
    },
    {
      "entropy": 5.51031002998352,
      "epoch": 0.8481494920174165,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935442086569498,
      "loss": 5.2329,
      "mean_token_accuracy": 0.1873893231153488,
      "num_tokens": 19415296.0,
      "step": 9350
    },
    {
      "entropy": 5.568622016906739,
      "epoch": 0.8486030478955007,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935365111753939,
      "loss": 5.2907,
      "mean_token_accuracy": 0.18625998497009277,
      "num_tokens": 19426119.0,
      "step": 9355
    },
    {
      "entropy": 5.55921778678894,
      "epoch": 0.8490566037735849,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004935288091744472,
      "loss": 5.2751,
      "mean_token_accuracy": 0.183457387983799,
      "num_tokens": 19436828.0,
      "step": 9360
    },
    {
      "entropy": 5.553300952911377,
      "epoch": 0.8495101596516691,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493521102654269,
      "loss": 5.2551,
      "mean_token_accuracy": 0.1870663121342659,
      "num_tokens": 19446583.0,
      "step": 9365
    },
    {
      "entropy": 5.452865266799927,
      "epoch": 0.8499637155297532,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935133916150187,
      "loss": 5.1336,
      "mean_token_accuracy": 0.19529592543840407,
      "num_tokens": 19456757.0,
      "step": 9370
    },
    {
      "entropy": 5.455078649520874,
      "epoch": 0.8504172714078374,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004935056760568558,
      "loss": 5.2172,
      "mean_token_accuracy": 0.18394441306591033,
      "num_tokens": 19467624.0,
      "step": 9375
    },
    {
      "entropy": 5.49309778213501,
      "epoch": 0.8508708272859217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934979559799399,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19133689105510712,
      "num_tokens": 19477759.0,
      "step": 9380
    },
    {
      "entropy": 5.515308380126953,
      "epoch": 0.8513243831640058,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004934902313844306,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19061967581510544,
      "num_tokens": 19488877.0,
      "step": 9385
    },
    {
      "entropy": 5.406462621688843,
      "epoch": 0.85177793904209,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004934825022704875,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19388926029205322,
      "num_tokens": 19499170.0,
      "step": 9390
    },
    {
      "entropy": 5.4668279647827145,
      "epoch": 0.8522314949201741,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004934747686382707,
      "loss": 5.2111,
      "mean_token_accuracy": 0.18104320466518403,
      "num_tokens": 19509944.0,
      "step": 9395
    },
    {
      "entropy": 5.513626003265381,
      "epoch": 0.8526850507982583,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934670304879398,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19153966307640075,
      "num_tokens": 19519014.0,
      "step": 9400
    },
    {
      "entropy": 5.448259162902832,
      "epoch": 0.8531386066763426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934592878196553,
      "loss": 5.187,
      "mean_token_accuracy": 0.19212067872285843,
      "num_tokens": 19528586.0,
      "step": 9405
    },
    {
      "entropy": 5.49515585899353,
      "epoch": 0.8535921625544267,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493451540633577,
      "loss": 5.3225,
      "mean_token_accuracy": 0.18525902032852173,
      "num_tokens": 19539912.0,
      "step": 9410
    },
    {
      "entropy": 5.518271636962891,
      "epoch": 0.8540457184325109,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004934437889298651,
      "loss": 5.2805,
      "mean_token_accuracy": 0.1793683409690857,
      "num_tokens": 19549860.0,
      "step": 9415
    },
    {
      "entropy": 5.540545892715454,
      "epoch": 0.854499274310595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004934360327086799,
      "loss": 5.2327,
      "mean_token_accuracy": 0.18420504480600358,
      "num_tokens": 19559446.0,
      "step": 9420
    },
    {
      "entropy": 5.539361572265625,
      "epoch": 0.8549528301886793,
      "grad_norm": 0.921875,
      "learning_rate": 0.000493428271970182,
      "loss": 5.2417,
      "mean_token_accuracy": 0.1881583198904991,
      "num_tokens": 19570282.0,
      "step": 9425
    },
    {
      "entropy": 5.510746669769287,
      "epoch": 0.8554063860667634,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004934205067145318,
      "loss": 5.1678,
      "mean_token_accuracy": 0.18670809119939805,
      "num_tokens": 19581087.0,
      "step": 9430
    },
    {
      "entropy": 5.492790555953979,
      "epoch": 0.8558599419448476,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004934127369418898,
      "loss": 5.2616,
      "mean_token_accuracy": 0.18760718405246735,
      "num_tokens": 19592100.0,
      "step": 9435
    },
    {
      "entropy": 5.487104320526123,
      "epoch": 0.8563134978229318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934049626524168,
      "loss": 5.2409,
      "mean_token_accuracy": 0.18586532920598983,
      "num_tokens": 19602557.0,
      "step": 9440
    },
    {
      "entropy": 5.5392913818359375,
      "epoch": 0.8567670537010159,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004933971838462734,
      "loss": 5.2899,
      "mean_token_accuracy": 0.18158393800258638,
      "num_tokens": 19613643.0,
      "step": 9445
    },
    {
      "entropy": 5.4967609405517575,
      "epoch": 0.8572206095791002,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933894005236207,
      "loss": 5.2001,
      "mean_token_accuracy": 0.1887534514069557,
      "num_tokens": 19624180.0,
      "step": 9450
    },
    {
      "entropy": 5.516611337661743,
      "epoch": 0.8576741654571843,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933816126846195,
      "loss": 5.2246,
      "mean_token_accuracy": 0.18624820560216904,
      "num_tokens": 19634799.0,
      "step": 9455
    },
    {
      "entropy": 5.485091066360473,
      "epoch": 0.8581277213352685,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933738203294309,
      "loss": 5.1903,
      "mean_token_accuracy": 0.19355441778898239,
      "num_tokens": 19644618.0,
      "step": 9460
    },
    {
      "entropy": 5.47304892539978,
      "epoch": 0.8585812772133526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493366023458216,
      "loss": 5.1999,
      "mean_token_accuracy": 0.18136429488658906,
      "num_tokens": 19655550.0,
      "step": 9465
    },
    {
      "entropy": 5.517861652374267,
      "epoch": 0.8590348330914369,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004933582220711362,
      "loss": 5.2673,
      "mean_token_accuracy": 0.1912809804081917,
      "num_tokens": 19665734.0,
      "step": 9470
    },
    {
      "entropy": 5.552363920211792,
      "epoch": 0.8594883889695211,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004933504161683526,
      "loss": 5.2249,
      "mean_token_accuracy": 0.18164515048265456,
      "num_tokens": 19677130.0,
      "step": 9475
    },
    {
      "entropy": 5.463605499267578,
      "epoch": 0.8599419448476052,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004933426057500269,
      "loss": 5.1415,
      "mean_token_accuracy": 0.19647871553897858,
      "num_tokens": 19687374.0,
      "step": 9480
    },
    {
      "entropy": 5.548071193695068,
      "epoch": 0.8603955007256894,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933347908163203,
      "loss": 5.2359,
      "mean_token_accuracy": 0.1858404666185379,
      "num_tokens": 19698182.0,
      "step": 9485
    },
    {
      "entropy": 5.489673233032226,
      "epoch": 0.8608490566037735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933269713673947,
      "loss": 5.209,
      "mean_token_accuracy": 0.18983528316020964,
      "num_tokens": 19707664.0,
      "step": 9490
    },
    {
      "entropy": 5.418010711669922,
      "epoch": 0.8613026124818578,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933191474034116,
      "loss": 5.2464,
      "mean_token_accuracy": 0.18493668735027313,
      "num_tokens": 19717647.0,
      "step": 9495
    },
    {
      "entropy": 5.481851577758789,
      "epoch": 0.861756168359942,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493311318924533,
      "loss": 5.2824,
      "mean_token_accuracy": 0.18406570553779603,
      "num_tokens": 19727444.0,
      "step": 9500
    },
    {
      "entropy": 5.547173070907593,
      "epoch": 0.8622097242380261,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933034859309208,
      "loss": 5.3011,
      "mean_token_accuracy": 0.1775568827986717,
      "num_tokens": 19738719.0,
      "step": 9505
    },
    {
      "entropy": 5.583591938018799,
      "epoch": 0.8626632801161103,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932956484227367,
      "loss": 5.2774,
      "mean_token_accuracy": 0.19096940159797668,
      "num_tokens": 19749786.0,
      "step": 9510
    },
    {
      "entropy": 5.515065097808838,
      "epoch": 0.8631168359941945,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000493287806400143,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19534773230552674,
      "num_tokens": 19759623.0,
      "step": 9515
    },
    {
      "entropy": 5.536817026138306,
      "epoch": 0.8635703918722787,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004932799598633018,
      "loss": 5.2199,
      "mean_token_accuracy": 0.18721261024475097,
      "num_tokens": 19770117.0,
      "step": 9520
    },
    {
      "entropy": 5.51656551361084,
      "epoch": 0.8640239477503628,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004932721088123753,
      "loss": 5.3074,
      "mean_token_accuracy": 0.18952126950025558,
      "num_tokens": 19780748.0,
      "step": 9525
    },
    {
      "entropy": 5.547321510314942,
      "epoch": 0.864477503628447,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004932642532475262,
      "loss": 5.2737,
      "mean_token_accuracy": 0.18902790993452073,
      "num_tokens": 19791816.0,
      "step": 9530
    },
    {
      "entropy": 5.57507176399231,
      "epoch": 0.8649310595065312,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932563931689168,
      "loss": 5.3212,
      "mean_token_accuracy": 0.18597906678915024,
      "num_tokens": 19801096.0,
      "step": 9535
    },
    {
      "entropy": 5.51464467048645,
      "epoch": 0.8653846153846154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932485285767093,
      "loss": 5.1896,
      "mean_token_accuracy": 0.19145326763391496,
      "num_tokens": 19811258.0,
      "step": 9540
    },
    {
      "entropy": 5.531577491760254,
      "epoch": 0.8658381712626996,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004932406594710667,
      "loss": 5.2594,
      "mean_token_accuracy": 0.19362047761678697,
      "num_tokens": 19821765.0,
      "step": 9545
    },
    {
      "entropy": 5.4913736343383786,
      "epoch": 0.8662917271407837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932327858521517,
      "loss": 5.1323,
      "mean_token_accuracy": 0.19672255367040634,
      "num_tokens": 19831537.0,
      "step": 9550
    },
    {
      "entropy": 5.447400045394898,
      "epoch": 0.8667452830188679,
      "grad_norm": 0.984375,
      "learning_rate": 0.000493224907720127,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19029980450868605,
      "num_tokens": 19841830.0,
      "step": 9555
    },
    {
      "entropy": 5.512387990951538,
      "epoch": 0.8671988388969522,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004932170250751556,
      "loss": 5.256,
      "mean_token_accuracy": 0.18012520521879197,
      "num_tokens": 19851896.0,
      "step": 9560
    },
    {
      "entropy": 5.537576293945312,
      "epoch": 0.8676523947750363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932091379174006,
      "loss": 5.289,
      "mean_token_accuracy": 0.1798422083258629,
      "num_tokens": 19861735.0,
      "step": 9565
    },
    {
      "entropy": 5.497203254699707,
      "epoch": 0.8681059506531205,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493201246247025,
      "loss": 5.1747,
      "mean_token_accuracy": 0.19459885060787202,
      "num_tokens": 19872333.0,
      "step": 9570
    },
    {
      "entropy": 5.521367979049683,
      "epoch": 0.8685595065312046,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000493193350064192,
      "loss": 5.1918,
      "mean_token_accuracy": 0.18699753284454346,
      "num_tokens": 19882896.0,
      "step": 9575
    },
    {
      "entropy": 5.4799529075622555,
      "epoch": 0.8690130624092888,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004931854493690649,
      "loss": 5.1517,
      "mean_token_accuracy": 0.189018677175045,
      "num_tokens": 19893767.0,
      "step": 9580
    },
    {
      "entropy": 5.41310601234436,
      "epoch": 0.869466618287373,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004931775441618071,
      "loss": 5.1264,
      "mean_token_accuracy": 0.19416602104902267,
      "num_tokens": 19904242.0,
      "step": 9585
    },
    {
      "entropy": 5.478859567642212,
      "epoch": 0.8699201741654572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004931696344425821,
      "loss": 5.2633,
      "mean_token_accuracy": 0.188524429500103,
      "num_tokens": 19913898.0,
      "step": 9590
    },
    {
      "entropy": 5.614615726470947,
      "epoch": 0.8703737300435413,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931617202115535,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19075826853513717,
      "num_tokens": 19923756.0,
      "step": 9595
    },
    {
      "entropy": 5.524111700057984,
      "epoch": 0.8708272859216255,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004931538014688849,
      "loss": 5.1939,
      "mean_token_accuracy": 0.1916771948337555,
      "num_tokens": 19932720.0,
      "step": 9600
    },
    {
      "entropy": 5.558842515945434,
      "epoch": 0.8712808417997098,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004931458782147401,
      "loss": 5.4063,
      "mean_token_accuracy": 0.17672424167394638,
      "num_tokens": 19942962.0,
      "step": 9605
    },
    {
      "entropy": 5.560279273986817,
      "epoch": 0.8717343976777939,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000493137950449283,
      "loss": 5.276,
      "mean_token_accuracy": 0.19128557592630385,
      "num_tokens": 19953735.0,
      "step": 9610
    },
    {
      "entropy": 5.505196857452392,
      "epoch": 0.8721879535558781,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931300181726775,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18588921427726746,
      "num_tokens": 19964416.0,
      "step": 9615
    },
    {
      "entropy": 5.517534017562866,
      "epoch": 0.8726415094339622,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004931220813850875,
      "loss": 5.2602,
      "mean_token_accuracy": 0.18303596675395967,
      "num_tokens": 19974531.0,
      "step": 9620
    },
    {
      "entropy": 5.540431547164917,
      "epoch": 0.8730950653120464,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004931141400866773,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19180816859006883,
      "num_tokens": 19984964.0,
      "step": 9625
    },
    {
      "entropy": 5.482637929916382,
      "epoch": 0.8735486211901307,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004931061942776112,
      "loss": 5.1951,
      "mean_token_accuracy": 0.19486953020095826,
      "num_tokens": 19995568.0,
      "step": 9630
    },
    {
      "entropy": 5.608856678009033,
      "epoch": 0.8740021770682148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930982439580534,
      "loss": 5.3692,
      "mean_token_accuracy": 0.17587941586971284,
      "num_tokens": 20005856.0,
      "step": 9635
    },
    {
      "entropy": 5.535829830169678,
      "epoch": 0.874455732946299,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004930902891281683,
      "loss": 5.2749,
      "mean_token_accuracy": 0.18487472534179689,
      "num_tokens": 20015683.0,
      "step": 9640
    },
    {
      "entropy": 5.421100807189942,
      "epoch": 0.8749092888243831,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004930823297881205,
      "loss": 5.0891,
      "mean_token_accuracy": 0.1941447973251343,
      "num_tokens": 20025998.0,
      "step": 9645
    },
    {
      "entropy": 5.542952680587769,
      "epoch": 0.8753628447024674,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004930743659380745,
      "loss": 5.2547,
      "mean_token_accuracy": 0.18763321340084077,
      "num_tokens": 20036486.0,
      "step": 9650
    },
    {
      "entropy": 5.459162616729737,
      "epoch": 0.8758164005805515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493066397578195,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19946982264518737,
      "num_tokens": 20046784.0,
      "step": 9655
    },
    {
      "entropy": 5.504218053817749,
      "epoch": 0.8762699564586357,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004930584247086468,
      "loss": 5.2238,
      "mean_token_accuracy": 0.18850288093090056,
      "num_tokens": 20056999.0,
      "step": 9660
    },
    {
      "entropy": 5.464111852645874,
      "epoch": 0.8767235123367199,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493050447329595,
      "loss": 5.1184,
      "mean_token_accuracy": 0.19502724707126617,
      "num_tokens": 20067612.0,
      "step": 9665
    },
    {
      "entropy": 5.520261335372925,
      "epoch": 0.877177068214804,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004930424654412043,
      "loss": 5.2666,
      "mean_token_accuracy": 0.1904868870973587,
      "num_tokens": 20077330.0,
      "step": 9670
    },
    {
      "entropy": 5.511299180984497,
      "epoch": 0.8776306240928883,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930344790436398,
      "loss": 5.26,
      "mean_token_accuracy": 0.18457795083522796,
      "num_tokens": 20086801.0,
      "step": 9675
    },
    {
      "entropy": 5.532819318771362,
      "epoch": 0.8780841799709724,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004930264881370666,
      "loss": 5.2115,
      "mean_token_accuracy": 0.18478081822395326,
      "num_tokens": 20097047.0,
      "step": 9680
    },
    {
      "entropy": 5.5098706722259525,
      "epoch": 0.8785377358490566,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930184927216504,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19049520045518875,
      "num_tokens": 20107185.0,
      "step": 9685
    },
    {
      "entropy": 5.523552942276001,
      "epoch": 0.8789912917271407,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930104927975559,
      "loss": 5.186,
      "mean_token_accuracy": 0.1903187081217766,
      "num_tokens": 20118493.0,
      "step": 9690
    },
    {
      "entropy": 5.510832071304321,
      "epoch": 0.879444847605225,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004930024883649489,
      "loss": 5.2423,
      "mean_token_accuracy": 0.1870877966284752,
      "num_tokens": 20129294.0,
      "step": 9695
    },
    {
      "entropy": 5.439377927780152,
      "epoch": 0.8798984034833092,
      "grad_norm": 1.0,
      "learning_rate": 0.000492994479423995,
      "loss": 5.2038,
      "mean_token_accuracy": 0.19338049739599228,
      "num_tokens": 20139194.0,
      "step": 9700
    },
    {
      "entropy": 5.4923159122467045,
      "epoch": 0.8803519593613933,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004929864659748597,
      "loss": 5.2241,
      "mean_token_accuracy": 0.18959469497203826,
      "num_tokens": 20149681.0,
      "step": 9705
    },
    {
      "entropy": 5.547377967834473,
      "epoch": 0.8808055152394775,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004929784480177087,
      "loss": 5.3143,
      "mean_token_accuracy": 0.18003763556480407,
      "num_tokens": 20160557.0,
      "step": 9710
    },
    {
      "entropy": 5.4973267078399655,
      "epoch": 0.8812590711175616,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004929704255527078,
      "loss": 5.1858,
      "mean_token_accuracy": 0.18966007977724075,
      "num_tokens": 20171047.0,
      "step": 9715
    },
    {
      "entropy": 5.519015216827393,
      "epoch": 0.8817126269956459,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492962398580023,
      "loss": 5.3549,
      "mean_token_accuracy": 0.18190275877714157,
      "num_tokens": 20182148.0,
      "step": 9720
    },
    {
      "entropy": 5.434145259857178,
      "epoch": 0.88216618287373,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929543670998203,
      "loss": 5.107,
      "mean_token_accuracy": 0.19520194828510284,
      "num_tokens": 20192131.0,
      "step": 9725
    },
    {
      "entropy": 5.524809503555298,
      "epoch": 0.8826197387518142,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004929463311122657,
      "loss": 5.228,
      "mean_token_accuracy": 0.1856295123696327,
      "num_tokens": 20202805.0,
      "step": 9730
    },
    {
      "entropy": 5.528243207931519,
      "epoch": 0.8830732946298984,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004929382906175253,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19349892288446427,
      "num_tokens": 20212702.0,
      "step": 9735
    },
    {
      "entropy": 5.468289422988891,
      "epoch": 0.8835268505079826,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929302456157657,
      "loss": 5.1914,
      "mean_token_accuracy": 0.1951130896806717,
      "num_tokens": 20222661.0,
      "step": 9740
    },
    {
      "entropy": 5.48810658454895,
      "epoch": 0.8839804063860668,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929221961071531,
      "loss": 5.239,
      "mean_token_accuracy": 0.1886950433254242,
      "num_tokens": 20233012.0,
      "step": 9745
    },
    {
      "entropy": 5.531930875778198,
      "epoch": 0.8844339622641509,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929141420918538,
      "loss": 5.2469,
      "mean_token_accuracy": 0.1919715628027916,
      "num_tokens": 20242918.0,
      "step": 9750
    },
    {
      "entropy": 5.4830118179321286,
      "epoch": 0.8848875181422351,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929060835700347,
      "loss": 5.1766,
      "mean_token_accuracy": 0.19638995528221131,
      "num_tokens": 20253254.0,
      "step": 9755
    },
    {
      "entropy": 5.4783000469207765,
      "epoch": 0.8853410740203193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492898020541862,
      "loss": 5.2378,
      "mean_token_accuracy": 0.18468834161758424,
      "num_tokens": 20263355.0,
      "step": 9760
    },
    {
      "entropy": 5.460058784484863,
      "epoch": 0.8857946298984035,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004928899530075029,
      "loss": 5.1601,
      "mean_token_accuracy": 0.19061171710491182,
      "num_tokens": 20274716.0,
      "step": 9765
    },
    {
      "entropy": 5.433362579345703,
      "epoch": 0.8862481857764877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004928818809671241,
      "loss": 5.1507,
      "mean_token_accuracy": 0.19014403820037842,
      "num_tokens": 20285631.0,
      "step": 9770
    },
    {
      "entropy": 5.545337724685669,
      "epoch": 0.8867017416545718,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928738044208923,
      "loss": 5.2384,
      "mean_token_accuracy": 0.19255442023277283,
      "num_tokens": 20295717.0,
      "step": 9775
    },
    {
      "entropy": 5.456331539154053,
      "epoch": 0.887155297532656,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928657233689748,
      "loss": 5.2041,
      "mean_token_accuracy": 0.1925482451915741,
      "num_tokens": 20306637.0,
      "step": 9780
    },
    {
      "entropy": 5.484044456481934,
      "epoch": 0.8876088534107403,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004928576378115387,
      "loss": 5.1831,
      "mean_token_accuracy": 0.1889451801776886,
      "num_tokens": 20316629.0,
      "step": 9785
    },
    {
      "entropy": 5.5872455596923825,
      "epoch": 0.8880624092888244,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928495477487509,
      "loss": 5.355,
      "mean_token_accuracy": 0.17307652533054352,
      "num_tokens": 20327935.0,
      "step": 9790
    },
    {
      "entropy": 5.530547428131103,
      "epoch": 0.8885159651669086,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928414531807792,
      "loss": 5.1728,
      "mean_token_accuracy": 0.1830437198281288,
      "num_tokens": 20339142.0,
      "step": 9795
    },
    {
      "entropy": 5.4817322254180905,
      "epoch": 0.8889695210449927,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004928333541077906,
      "loss": 5.172,
      "mean_token_accuracy": 0.2016322910785675,
      "num_tokens": 20348474.0,
      "step": 9800
    },
    {
      "entropy": 5.479598665237427,
      "epoch": 0.8894230769230769,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928252505299527,
      "loss": 5.2418,
      "mean_token_accuracy": 0.19005548357963561,
      "num_tokens": 20359440.0,
      "step": 9805
    },
    {
      "entropy": 5.569894742965698,
      "epoch": 0.8898766328011611,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004928171424474332,
      "loss": 5.2767,
      "mean_token_accuracy": 0.1847678691148758,
      "num_tokens": 20370355.0,
      "step": 9810
    },
    {
      "entropy": 5.4881508350372314,
      "epoch": 0.8903301886792453,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004928090298603996,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1922898545861244,
      "num_tokens": 20380214.0,
      "step": 9815
    },
    {
      "entropy": 5.526920175552368,
      "epoch": 0.8907837445573294,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004928009127690199,
      "loss": 5.1952,
      "mean_token_accuracy": 0.18694787621498107,
      "num_tokens": 20389190.0,
      "step": 9820
    },
    {
      "entropy": 5.458229827880859,
      "epoch": 0.8912373004354136,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927927911734617,
      "loss": 5.232,
      "mean_token_accuracy": 0.18914854824542998,
      "num_tokens": 20399531.0,
      "step": 9825
    },
    {
      "entropy": 5.465860223770141,
      "epoch": 0.8916908563134979,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004927846650738931,
      "loss": 5.2215,
      "mean_token_accuracy": 0.18792601376771928,
      "num_tokens": 20410201.0,
      "step": 9830
    },
    {
      "entropy": 5.48523736000061,
      "epoch": 0.892144412191582,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004927765344704821,
      "loss": 5.1693,
      "mean_token_accuracy": 0.19596148431301116,
      "num_tokens": 20421579.0,
      "step": 9835
    },
    {
      "entropy": 5.474275684356689,
      "epoch": 0.8925979680696662,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004927683993633969,
      "loss": 5.2427,
      "mean_token_accuracy": 0.19045363068580629,
      "num_tokens": 20432048.0,
      "step": 9840
    },
    {
      "entropy": 5.553377866744995,
      "epoch": 0.8930515239477503,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004927602597528057,
      "loss": 5.2663,
      "mean_token_accuracy": 0.18120875656604768,
      "num_tokens": 20441752.0,
      "step": 9845
    },
    {
      "entropy": 5.456312799453736,
      "epoch": 0.8935050798258345,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004927521156388769,
      "loss": 5.13,
      "mean_token_accuracy": 0.1914264976978302,
      "num_tokens": 20451693.0,
      "step": 9850
    },
    {
      "entropy": 5.502097320556641,
      "epoch": 0.8939586357039188,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004927439670217788,
      "loss": 5.2738,
      "mean_token_accuracy": 0.18670168668031692,
      "num_tokens": 20462312.0,
      "step": 9855
    },
    {
      "entropy": 5.483501243591308,
      "epoch": 0.8944121915820029,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049273581390168,
      "loss": 5.1804,
      "mean_token_accuracy": 0.18460638374090194,
      "num_tokens": 20472546.0,
      "step": 9860
    },
    {
      "entropy": 5.506111812591553,
      "epoch": 0.8948657474600871,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000492727656278749,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1930461436510086,
      "num_tokens": 20483266.0,
      "step": 9865
    },
    {
      "entropy": 5.464397954940796,
      "epoch": 0.8953193033381712,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004927194941531547,
      "loss": 5.1512,
      "mean_token_accuracy": 0.18788621425628663,
      "num_tokens": 20492490.0,
      "step": 9870
    },
    {
      "entropy": 5.485383033752441,
      "epoch": 0.8957728592162555,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004927113275250657,
      "loss": 5.1922,
      "mean_token_accuracy": 0.19176370054483413,
      "num_tokens": 20503077.0,
      "step": 9875
    },
    {
      "entropy": 5.464278745651245,
      "epoch": 0.8962264150943396,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004927031563946511,
      "loss": 5.2078,
      "mean_token_accuracy": 0.18573808223009108,
      "num_tokens": 20513771.0,
      "step": 9880
    },
    {
      "entropy": 5.532713842391968,
      "epoch": 0.8966799709724238,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004926949807620796,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1865783840417862,
      "num_tokens": 20523616.0,
      "step": 9885
    },
    {
      "entropy": 5.4971997261047365,
      "epoch": 0.897133526850508,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926868006275204,
      "loss": 5.0925,
      "mean_token_accuracy": 0.19763742238283158,
      "num_tokens": 20533613.0,
      "step": 9890
    },
    {
      "entropy": 5.4906604290008545,
      "epoch": 0.8975870827285921,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004926786159911427,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1880497470498085,
      "num_tokens": 20543812.0,
      "step": 9895
    },
    {
      "entropy": 5.460688543319702,
      "epoch": 0.8980406386066764,
      "grad_norm": 1.0,
      "learning_rate": 0.000492670426853116,
      "loss": 5.1363,
      "mean_token_accuracy": 0.192536161839962,
      "num_tokens": 20553785.0,
      "step": 9900
    },
    {
      "entropy": 5.562257194519043,
      "epoch": 0.8984941944847605,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926622332136092,
      "loss": 5.2559,
      "mean_token_accuracy": 0.19221415668725966,
      "num_tokens": 20563326.0,
      "step": 9905
    },
    {
      "entropy": 5.436643981933594,
      "epoch": 0.8989477503628447,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004926540350727918,
      "loss": 5.1005,
      "mean_token_accuracy": 0.19912851303815843,
      "num_tokens": 20573643.0,
      "step": 9910
    },
    {
      "entropy": 5.426002407073975,
      "epoch": 0.8994013062409288,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926458324308337,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19306659698486328,
      "num_tokens": 20583916.0,
      "step": 9915
    },
    {
      "entropy": 5.516725540161133,
      "epoch": 0.8998548621190131,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004926376252879043,
      "loss": 5.2894,
      "mean_token_accuracy": 0.19082354307174682,
      "num_tokens": 20593984.0,
      "step": 9920
    },
    {
      "entropy": 5.518711185455322,
      "epoch": 0.9003084179970973,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004926294136441733,
      "loss": 5.1265,
      "mean_token_accuracy": 0.1930823028087616,
      "num_tokens": 20603927.0,
      "step": 9925
    },
    {
      "entropy": 5.496563529968261,
      "epoch": 0.9007619738751814,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004926211974998106,
      "loss": 5.2612,
      "mean_token_accuracy": 0.1923177346587181,
      "num_tokens": 20614147.0,
      "step": 9930
    },
    {
      "entropy": 5.523381233215332,
      "epoch": 0.9012155297532656,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000492612976854986,
      "loss": 5.2238,
      "mean_token_accuracy": 0.1915097326040268,
      "num_tokens": 20623631.0,
      "step": 9935
    },
    {
      "entropy": 5.465784215927124,
      "epoch": 0.9016690856313497,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926047517098697,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19526478499174119,
      "num_tokens": 20632566.0,
      "step": 9940
    },
    {
      "entropy": 5.512724876403809,
      "epoch": 0.902122641509434,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004925965220646315,
      "loss": 5.2151,
      "mean_token_accuracy": 0.18857671916484833,
      "num_tokens": 20643246.0,
      "step": 9945
    },
    {
      "entropy": 5.507947826385498,
      "epoch": 0.9025761973875182,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925882879194419,
      "loss": 5.296,
      "mean_token_accuracy": 0.18535405844449998,
      "num_tokens": 20653372.0,
      "step": 9950
    },
    {
      "entropy": 5.5078880310058596,
      "epoch": 0.9030297532656023,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004925800492744712,
      "loss": 5.2362,
      "mean_token_accuracy": 0.18510301411151886,
      "num_tokens": 20662687.0,
      "step": 9955
    },
    {
      "entropy": 5.463442754745484,
      "epoch": 0.9034833091436865,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004925718061298895,
      "loss": 5.1509,
      "mean_token_accuracy": 0.196966715157032,
      "num_tokens": 20673168.0,
      "step": 9960
    },
    {
      "entropy": 5.560179567337036,
      "epoch": 0.9039368650217707,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925635584858675,
      "loss": 5.267,
      "mean_token_accuracy": 0.1841796949505806,
      "num_tokens": 20683100.0,
      "step": 9965
    },
    {
      "entropy": 5.587424993515015,
      "epoch": 0.9043904208998549,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925553063425755,
      "loss": 5.2772,
      "mean_token_accuracy": 0.186713108420372,
      "num_tokens": 20694097.0,
      "step": 9970
    },
    {
      "entropy": 5.5363075733184814,
      "epoch": 0.904843976777939,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925470497001844,
      "loss": 5.2393,
      "mean_token_accuracy": 0.19200791716575621,
      "num_tokens": 20704640.0,
      "step": 9975
    },
    {
      "entropy": 5.471827363967895,
      "epoch": 0.9052975326560232,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925387885588648,
      "loss": 5.1802,
      "mean_token_accuracy": 0.18870692253112792,
      "num_tokens": 20714419.0,
      "step": 9980
    },
    {
      "entropy": 5.544799852371216,
      "epoch": 0.9057510885341074,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004925305229187877,
      "loss": 5.2693,
      "mean_token_accuracy": 0.18884519636631011,
      "num_tokens": 20724705.0,
      "step": 9985
    },
    {
      "entropy": 5.597160959243775,
      "epoch": 0.9062046444121916,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492522252780124,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18726316094398499,
      "num_tokens": 20734753.0,
      "step": 9990
    },
    {
      "entropy": 5.465805244445801,
      "epoch": 0.9066582002902758,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004925139781430446,
      "loss": 5.1156,
      "mean_token_accuracy": 0.1963346555829048,
      "num_tokens": 20745389.0,
      "step": 9995
    },
    {
      "entropy": 5.480003547668457,
      "epoch": 0.9071117561683599,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004925056990077207,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1931140884757042,
      "num_tokens": 20755702.0,
      "step": 10000
    },
    {
      "entropy": 5.498392629623413,
      "epoch": 0.9075653120464441,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004924974153743235,
      "loss": 5.0984,
      "mean_token_accuracy": 0.19833627194166184,
      "num_tokens": 20766046.0,
      "step": 10005
    },
    {
      "entropy": 5.50446720123291,
      "epoch": 0.9080188679245284,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924891272430244,
      "loss": 5.2141,
      "mean_token_accuracy": 0.19061006605625153,
      "num_tokens": 20776654.0,
      "step": 10010
    },
    {
      "entropy": 5.5594000816345215,
      "epoch": 0.9084724238026125,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004924808346139948,
      "loss": 5.2713,
      "mean_token_accuracy": 0.19167041033506393,
      "num_tokens": 20787066.0,
      "step": 10015
    },
    {
      "entropy": 5.47709379196167,
      "epoch": 0.9089259796806967,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924725374874059,
      "loss": 5.1736,
      "mean_token_accuracy": 0.1896710604429245,
      "num_tokens": 20796822.0,
      "step": 10020
    },
    {
      "entropy": 5.495741128921509,
      "epoch": 0.9093795355587808,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924642358634297,
      "loss": 5.2297,
      "mean_token_accuracy": 0.18855800926685334,
      "num_tokens": 20807602.0,
      "step": 10025
    },
    {
      "entropy": 5.5386439800262455,
      "epoch": 0.909833091436865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924559297422376,
      "loss": 5.1864,
      "mean_token_accuracy": 0.19395318776369094,
      "num_tokens": 20819046.0,
      "step": 10030
    },
    {
      "entropy": 5.51975998878479,
      "epoch": 0.9102866473149492,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004924476191240014,
      "loss": 5.2153,
      "mean_token_accuracy": 0.19031708985567092,
      "num_tokens": 20828989.0,
      "step": 10035
    },
    {
      "entropy": 5.446819305419922,
      "epoch": 0.9107402031930334,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492439304008893,
      "loss": 5.2111,
      "mean_token_accuracy": 0.18754609823226928,
      "num_tokens": 20839366.0,
      "step": 10040
    },
    {
      "entropy": 5.519721603393554,
      "epoch": 0.9111937590711175,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004924309843970847,
      "loss": 5.2112,
      "mean_token_accuracy": 0.19083768427371978,
      "num_tokens": 20849513.0,
      "step": 10045
    },
    {
      "entropy": 5.559159517288208,
      "epoch": 0.9116473149492017,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004924226602887479,
      "loss": 5.2674,
      "mean_token_accuracy": 0.18467978686094283,
      "num_tokens": 20861406.0,
      "step": 10050
    },
    {
      "entropy": 5.47953314781189,
      "epoch": 0.912100870827286,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004924143316840552,
      "loss": 5.1645,
      "mean_token_accuracy": 0.19939935952425003,
      "num_tokens": 20871190.0,
      "step": 10055
    },
    {
      "entropy": 5.531355714797973,
      "epoch": 0.9125544267053701,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924059985831787,
      "loss": 5.3149,
      "mean_token_accuracy": 0.1820805475115776,
      "num_tokens": 20881589.0,
      "step": 10060
    },
    {
      "entropy": 5.5043120861053465,
      "epoch": 0.9130079825834543,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923976609862908,
      "loss": 5.249,
      "mean_token_accuracy": 0.17866823971271514,
      "num_tokens": 20892072.0,
      "step": 10065
    },
    {
      "entropy": 5.541967296600342,
      "epoch": 0.9134615384615384,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923893188935638,
      "loss": 5.2361,
      "mean_token_accuracy": 0.18548781424760818,
      "num_tokens": 20902616.0,
      "step": 10070
    },
    {
      "entropy": 5.578791618347168,
      "epoch": 0.9139150943396226,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004923809723051702,
      "loss": 5.231,
      "mean_token_accuracy": 0.1868339642882347,
      "num_tokens": 20913832.0,
      "step": 10075
    },
    {
      "entropy": 5.451777458190918,
      "epoch": 0.9143686502177069,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004923726212212829,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1962735816836357,
      "num_tokens": 20923322.0,
      "step": 10080
    },
    {
      "entropy": 5.44694151878357,
      "epoch": 0.914822206095791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923642656420742,
      "loss": 5.2078,
      "mean_token_accuracy": 0.19073516130447388,
      "num_tokens": 20933334.0,
      "step": 10085
    },
    {
      "entropy": 5.462447452545166,
      "epoch": 0.9152757619738752,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923559055677173,
      "loss": 5.1788,
      "mean_token_accuracy": 0.1864377275109291,
      "num_tokens": 20944171.0,
      "step": 10090
    },
    {
      "entropy": 5.524150848388672,
      "epoch": 0.9157293178519593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923475409983848,
      "loss": 5.215,
      "mean_token_accuracy": 0.18702605217695237,
      "num_tokens": 20954200.0,
      "step": 10095
    },
    {
      "entropy": 5.490021753311157,
      "epoch": 0.9161828737300436,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004923391719342497,
      "loss": 5.2339,
      "mean_token_accuracy": 0.1857544004917145,
      "num_tokens": 20964935.0,
      "step": 10100
    },
    {
      "entropy": 5.470995235443115,
      "epoch": 0.9166364296081277,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923307983754852,
      "loss": 5.2232,
      "mean_token_accuracy": 0.18978999555110931,
      "num_tokens": 20975499.0,
      "step": 10105
    },
    {
      "entropy": 5.406945657730103,
      "epoch": 0.9170899854862119,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923224203222645,
      "loss": 5.1461,
      "mean_token_accuracy": 0.19023018777370454,
      "num_tokens": 20986199.0,
      "step": 10110
    },
    {
      "entropy": 5.485232019424439,
      "epoch": 0.9175435413642961,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923140377747606,
      "loss": 5.2263,
      "mean_token_accuracy": 0.18715826719999312,
      "num_tokens": 20996313.0,
      "step": 10115
    },
    {
      "entropy": 5.515059518814087,
      "epoch": 0.9179970972423802,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923056507331471,
      "loss": 5.2129,
      "mean_token_accuracy": 0.18567863702774048,
      "num_tokens": 21006133.0,
      "step": 10120
    },
    {
      "entropy": 5.515254068374634,
      "epoch": 0.9184506531204645,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004922972591975974,
      "loss": 5.1735,
      "mean_token_accuracy": 0.18915437906980515,
      "num_tokens": 21016448.0,
      "step": 10125
    },
    {
      "entropy": 5.455716228485107,
      "epoch": 0.9189042089985486,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922888631682849,
      "loss": 5.1148,
      "mean_token_accuracy": 0.18959181159734725,
      "num_tokens": 21026458.0,
      "step": 10130
    },
    {
      "entropy": 5.464995288848877,
      "epoch": 0.9193577648766328,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004922804626453834,
      "loss": 5.1567,
      "mean_token_accuracy": 0.19512831717729567,
      "num_tokens": 21037111.0,
      "step": 10135
    },
    {
      "entropy": 5.4613899230957035,
      "epoch": 0.9198113207547169,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004922720576290665,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19339829832315444,
      "num_tokens": 21048041.0,
      "step": 10140
    },
    {
      "entropy": 5.438198232650757,
      "epoch": 0.9202648766328012,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004922636481195082,
      "loss": 5.0888,
      "mean_token_accuracy": 0.20385636240243912,
      "num_tokens": 21058099.0,
      "step": 10145
    },
    {
      "entropy": 5.448825025558472,
      "epoch": 0.9207184325108854,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004922552341168823,
      "loss": 5.2224,
      "mean_token_accuracy": 0.18477773368358613,
      "num_tokens": 21068137.0,
      "step": 10150
    },
    {
      "entropy": 5.417740678787231,
      "epoch": 0.9211719883889695,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922468156213627,
      "loss": 5.115,
      "mean_token_accuracy": 0.19422985017299652,
      "num_tokens": 21077556.0,
      "step": 10155
    },
    {
      "entropy": 5.478008985519409,
      "epoch": 0.9216255442670537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004922383926331236,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19280444234609603,
      "num_tokens": 21087524.0,
      "step": 10160
    },
    {
      "entropy": 5.425194454193115,
      "epoch": 0.9220791001451378,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922299651523392,
      "loss": 5.2124,
      "mean_token_accuracy": 0.18724745512008667,
      "num_tokens": 21098446.0,
      "step": 10165
    },
    {
      "entropy": 5.459459257125855,
      "epoch": 0.9225326560232221,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922215331791837,
      "loss": 5.2302,
      "mean_token_accuracy": 0.18599504977464676,
      "num_tokens": 21109291.0,
      "step": 10170
    },
    {
      "entropy": 5.501635313034058,
      "epoch": 0.9229862119013063,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004922130967138316,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19838399291038514,
      "num_tokens": 21120043.0,
      "step": 10175
    },
    {
      "entropy": 5.518260431289673,
      "epoch": 0.9234397677793904,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004922046557564574,
      "loss": 5.2318,
      "mean_token_accuracy": 0.18464371412992478,
      "num_tokens": 21130339.0,
      "step": 10180
    },
    {
      "entropy": 5.4879247665405275,
      "epoch": 0.9238933236574746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921962103072355,
      "loss": 5.2982,
      "mean_token_accuracy": 0.18253571540117264,
      "num_tokens": 21142307.0,
      "step": 10185
    },
    {
      "entropy": 5.516984462738037,
      "epoch": 0.9243468795355588,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921877603663406,
      "loss": 5.1916,
      "mean_token_accuracy": 0.19168877601623535,
      "num_tokens": 21151843.0,
      "step": 10190
    },
    {
      "entropy": 5.52185640335083,
      "epoch": 0.924800435413643,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921793059339474,
      "loss": 5.1917,
      "mean_token_accuracy": 0.1947432205080986,
      "num_tokens": 21162574.0,
      "step": 10195
    },
    {
      "entropy": 5.514074325561523,
      "epoch": 0.9252539912917271,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921708470102309,
      "loss": 5.1502,
      "mean_token_accuracy": 0.19119355827569962,
      "num_tokens": 21172637.0,
      "step": 10200
    },
    {
      "entropy": 5.465252828598023,
      "epoch": 0.9257075471698113,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492162383595366,
      "loss": 5.1695,
      "mean_token_accuracy": 0.19286154508590697,
      "num_tokens": 21182272.0,
      "step": 10205
    },
    {
      "entropy": 5.5177301406860355,
      "epoch": 0.9261611030478955,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921539156895275,
      "loss": 5.287,
      "mean_token_accuracy": 0.18583095222711563,
      "num_tokens": 21193712.0,
      "step": 10210
    },
    {
      "entropy": 5.565571260452271,
      "epoch": 0.9266146589259797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004921454432928908,
      "loss": 5.3146,
      "mean_token_accuracy": 0.18066612184047698,
      "num_tokens": 21205104.0,
      "step": 10215
    },
    {
      "entropy": 5.5436365604400635,
      "epoch": 0.9270682148040639,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000492136966405631,
      "loss": 5.1809,
      "mean_token_accuracy": 0.1893284574151039,
      "num_tokens": 21215495.0,
      "step": 10220
    },
    {
      "entropy": 5.496471214294433,
      "epoch": 0.927521770682148,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004921284850279233,
      "loss": 5.2052,
      "mean_token_accuracy": 0.18886639475822448,
      "num_tokens": 21226047.0,
      "step": 10225
    },
    {
      "entropy": 5.502618789672852,
      "epoch": 0.9279753265602322,
      "grad_norm": 1.0,
      "learning_rate": 0.0004921199991599433,
      "loss": 5.2528,
      "mean_token_accuracy": 0.18845335245132447,
      "num_tokens": 21235784.0,
      "step": 10230
    },
    {
      "entropy": 5.557482576370239,
      "epoch": 0.9284288824383164,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921115088018665,
      "loss": 5.2097,
      "mean_token_accuracy": 0.19081354588270188,
      "num_tokens": 21245493.0,
      "step": 10235
    },
    {
      "entropy": 5.612878608703613,
      "epoch": 0.9288824383164006,
      "grad_norm": 0.96875,
      "learning_rate": 0.000492103013953868,
      "loss": 5.2906,
      "mean_token_accuracy": 0.1868404284119606,
      "num_tokens": 21255541.0,
      "step": 10240
    },
    {
      "entropy": 5.442387437820434,
      "epoch": 0.9293359941944848,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004920945146161242,
      "loss": 5.1463,
      "mean_token_accuracy": 0.19286169111728668,
      "num_tokens": 21265427.0,
      "step": 10245
    },
    {
      "entropy": 5.464549398422241,
      "epoch": 0.9297895500725689,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920860107888105,
      "loss": 5.1543,
      "mean_token_accuracy": 0.1940089598298073,
      "num_tokens": 21276597.0,
      "step": 10250
    },
    {
      "entropy": 5.460808277130127,
      "epoch": 0.9302431059506531,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004920775024721026,
      "loss": 5.1745,
      "mean_token_accuracy": 0.18666350245475768,
      "num_tokens": 21286416.0,
      "step": 10255
    },
    {
      "entropy": 5.482029771804809,
      "epoch": 0.9306966618287373,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920689896661767,
      "loss": 5.1954,
      "mean_token_accuracy": 0.19428568035364152,
      "num_tokens": 21296064.0,
      "step": 10260
    },
    {
      "entropy": 5.459726476669312,
      "epoch": 0.9311502177068215,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920604723712088,
      "loss": 5.2188,
      "mean_token_accuracy": 0.19196224957704544,
      "num_tokens": 21306836.0,
      "step": 10265
    },
    {
      "entropy": 5.48441743850708,
      "epoch": 0.9316037735849056,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920519505873751,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19369848519563676,
      "num_tokens": 21316936.0,
      "step": 10270
    },
    {
      "entropy": 5.464742708206177,
      "epoch": 0.9320573294629898,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004920434243148515,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20192406624555587,
      "num_tokens": 21326892.0,
      "step": 10275
    },
    {
      "entropy": 5.472868013381958,
      "epoch": 0.9325108853410741,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920348935538147,
      "loss": 5.1444,
      "mean_token_accuracy": 0.1884632021188736,
      "num_tokens": 21337270.0,
      "step": 10280
    },
    {
      "entropy": 5.532982063293457,
      "epoch": 0.9329644412191582,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000492026358304441,
      "loss": 5.2328,
      "mean_token_accuracy": 0.19010204523801805,
      "num_tokens": 21347968.0,
      "step": 10285
    },
    {
      "entropy": 5.47656831741333,
      "epoch": 0.9334179970972424,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000492017818566907,
      "loss": 5.1746,
      "mean_token_accuracy": 0.19259967803955078,
      "num_tokens": 21358312.0,
      "step": 10290
    },
    {
      "entropy": 5.470859527587891,
      "epoch": 0.9338715529753265,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920092743413892,
      "loss": 5.142,
      "mean_token_accuracy": 0.19385140240192414,
      "num_tokens": 21368794.0,
      "step": 10295
    },
    {
      "entropy": 5.508439350128174,
      "epoch": 0.9343251088534107,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920007256280642,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1947756752371788,
      "num_tokens": 21380080.0,
      "step": 10300
    },
    {
      "entropy": 5.497937631607056,
      "epoch": 0.934778664731495,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491992172427109,
      "loss": 5.1906,
      "mean_token_accuracy": 0.19269357919692992,
      "num_tokens": 21389961.0,
      "step": 10305
    },
    {
      "entropy": 5.450817489624024,
      "epoch": 0.9352322206095791,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004919836147387002,
      "loss": 5.2161,
      "mean_token_accuracy": 0.1923079416155815,
      "num_tokens": 21399869.0,
      "step": 10310
    },
    {
      "entropy": 5.519410276412964,
      "epoch": 0.9356857764876633,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004919750525630151,
      "loss": 5.2044,
      "mean_token_accuracy": 0.1905330315232277,
      "num_tokens": 21410343.0,
      "step": 10315
    },
    {
      "entropy": 5.512972068786621,
      "epoch": 0.9361393323657474,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919664859002306,
      "loss": 5.1847,
      "mean_token_accuracy": 0.18935477286577224,
      "num_tokens": 21420915.0,
      "step": 10320
    },
    {
      "entropy": 5.480722951889038,
      "epoch": 0.9365928882438317,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004919579147505239,
      "loss": 5.138,
      "mean_token_accuracy": 0.19782340973615647,
      "num_tokens": 21430879.0,
      "step": 10325
    },
    {
      "entropy": 5.5278362274169925,
      "epoch": 0.9370464441219158,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004919493391140721,
      "loss": 5.2364,
      "mean_token_accuracy": 0.186504028737545,
      "num_tokens": 21440843.0,
      "step": 10330
    },
    {
      "entropy": 5.567490196228027,
      "epoch": 0.9375,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919407589910529,
      "loss": 5.4008,
      "mean_token_accuracy": 0.18015343099832534,
      "num_tokens": 21452132.0,
      "step": 10335
    },
    {
      "entropy": 5.494717454910278,
      "epoch": 0.9379535558780842,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004919321743816434,
      "loss": 5.185,
      "mean_token_accuracy": 0.1896796151995659,
      "num_tokens": 21461861.0,
      "step": 10340
    },
    {
      "entropy": 5.5109964370727536,
      "epoch": 0.9384071117561683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004919235852860213,
      "loss": 5.1841,
      "mean_token_accuracy": 0.1865747943520546,
      "num_tokens": 21472244.0,
      "step": 10345
    },
    {
      "entropy": 5.528714323043824,
      "epoch": 0.9388606676342526,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004919149917043641,
      "loss": 5.2286,
      "mean_token_accuracy": 0.18693966567516326,
      "num_tokens": 21482394.0,
      "step": 10350
    },
    {
      "entropy": 5.504456090927124,
      "epoch": 0.9393142235123367,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004919063936368496,
      "loss": 5.2679,
      "mean_token_accuracy": 0.18453455120325088,
      "num_tokens": 21493334.0,
      "step": 10355
    },
    {
      "entropy": 5.5536915302276615,
      "epoch": 0.9397677793904209,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918977910836557,
      "loss": 5.2333,
      "mean_token_accuracy": 0.18519501984119416,
      "num_tokens": 21504443.0,
      "step": 10360
    },
    {
      "entropy": 5.539934301376343,
      "epoch": 0.940221335268505,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004918891840449601,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18833032846450806,
      "num_tokens": 21513908.0,
      "step": 10365
    },
    {
      "entropy": 5.542459392547608,
      "epoch": 0.9406748911465893,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000491880572520941,
      "loss": 5.2724,
      "mean_token_accuracy": 0.18035230338573455,
      "num_tokens": 21524349.0,
      "step": 10370
    },
    {
      "entropy": 5.6526354312896725,
      "epoch": 0.9411284470246735,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004918719565117763,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1903665080666542,
      "num_tokens": 21535003.0,
      "step": 10375
    },
    {
      "entropy": 5.616309881210327,
      "epoch": 0.9415820029027576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004918633360176444,
      "loss": 5.3643,
      "mean_token_accuracy": 0.1793342188000679,
      "num_tokens": 21545045.0,
      "step": 10380
    },
    {
      "entropy": 5.4643491268157955,
      "epoch": 0.9420355587808418,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918547110387234,
      "loss": 5.182,
      "mean_token_accuracy": 0.19717971086502076,
      "num_tokens": 21554830.0,
      "step": 10385
    },
    {
      "entropy": 5.5400168895721436,
      "epoch": 0.9424891146589259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004918460815751916,
      "loss": 5.1437,
      "mean_token_accuracy": 0.19479870349168776,
      "num_tokens": 21564693.0,
      "step": 10390
    },
    {
      "entropy": 5.491382455825805,
      "epoch": 0.9429426705370102,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004918374476272277,
      "loss": 5.167,
      "mean_token_accuracy": 0.187149715423584,
      "num_tokens": 21575475.0,
      "step": 10395
    },
    {
      "entropy": 5.4525188446044925,
      "epoch": 0.9433962264150944,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004918288091950101,
      "loss": 5.1887,
      "mean_token_accuracy": 0.18847269862890242,
      "num_tokens": 21586944.0,
      "step": 10400
    },
    {
      "entropy": 5.544979858398437,
      "epoch": 0.9438497822931785,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004918201662787174,
      "loss": 5.2002,
      "mean_token_accuracy": 0.1872720867395401,
      "num_tokens": 21598170.0,
      "step": 10405
    },
    {
      "entropy": 5.511740016937256,
      "epoch": 0.9443033381712627,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004918115188785285,
      "loss": 5.2188,
      "mean_token_accuracy": 0.1880117490887642,
      "num_tokens": 21609170.0,
      "step": 10410
    },
    {
      "entropy": 5.515832614898682,
      "epoch": 0.9447568940493469,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004918028669946221,
      "loss": 5.2956,
      "mean_token_accuracy": 0.18484372794628143,
      "num_tokens": 21620171.0,
      "step": 10415
    },
    {
      "entropy": 5.486926412582397,
      "epoch": 0.9452104499274311,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917942106271771,
      "loss": 5.2104,
      "mean_token_accuracy": 0.19340168982744216,
      "num_tokens": 21631072.0,
      "step": 10420
    },
    {
      "entropy": 5.4726005554199215,
      "epoch": 0.9456640058055152,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004917855497763727,
      "loss": 5.0846,
      "mean_token_accuracy": 0.1976454183459282,
      "num_tokens": 21641543.0,
      "step": 10425
    },
    {
      "entropy": 5.4695212841033936,
      "epoch": 0.9461175616835994,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004917768844423878,
      "loss": 5.2123,
      "mean_token_accuracy": 0.19243092685937882,
      "num_tokens": 21651342.0,
      "step": 10430
    },
    {
      "entropy": 5.524942016601562,
      "epoch": 0.9465711175616836,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004917682146254017,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19580309987068176,
      "num_tokens": 21661043.0,
      "step": 10435
    },
    {
      "entropy": 5.4407624244689945,
      "epoch": 0.9470246734397678,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004917595403255938,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1993831992149353,
      "num_tokens": 21671197.0,
      "step": 10440
    },
    {
      "entropy": 5.421433401107788,
      "epoch": 0.947478229317852,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917508615431431,
      "loss": 5.1253,
      "mean_token_accuracy": 0.19686731249094008,
      "num_tokens": 21682066.0,
      "step": 10445
    },
    {
      "entropy": 5.554642391204834,
      "epoch": 0.9479317851959361,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004917421782782296,
      "loss": 5.1912,
      "mean_token_accuracy": 0.19019131511449813,
      "num_tokens": 21692117.0,
      "step": 10450
    },
    {
      "entropy": 5.5294716358184814,
      "epoch": 0.9483853410740203,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917334905310325,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18246906250715256,
      "num_tokens": 21703544.0,
      "step": 10455
    },
    {
      "entropy": 5.445235729217529,
      "epoch": 0.9488388969521045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004917247983017317,
      "loss": 5.1233,
      "mean_token_accuracy": 0.19221487492322922,
      "num_tokens": 21712850.0,
      "step": 10460
    },
    {
      "entropy": 5.474435758590698,
      "epoch": 0.9492924528301887,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004917161015905067,
      "loss": 5.2636,
      "mean_token_accuracy": 0.18826578557491302,
      "num_tokens": 21723015.0,
      "step": 10465
    },
    {
      "entropy": 5.513061094284057,
      "epoch": 0.9497460087082729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917074003975376,
      "loss": 5.2182,
      "mean_token_accuracy": 0.1881333664059639,
      "num_tokens": 21732453.0,
      "step": 10470
    },
    {
      "entropy": 5.551966810226441,
      "epoch": 0.950199564586357,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004916986947230042,
      "loss": 5.2908,
      "mean_token_accuracy": 0.18900395184755325,
      "num_tokens": 21743251.0,
      "step": 10475
    },
    {
      "entropy": 5.515570878982544,
      "epoch": 0.9506531204644412,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916899845670866,
      "loss": 5.2076,
      "mean_token_accuracy": 0.18819339275360109,
      "num_tokens": 21752891.0,
      "step": 10480
    },
    {
      "entropy": 5.543520164489746,
      "epoch": 0.9511066763425254,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004916812699299649,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19229596108198166,
      "num_tokens": 21763230.0,
      "step": 10485
    },
    {
      "entropy": 5.49464168548584,
      "epoch": 0.9515602322206096,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916725508118193,
      "loss": 5.1474,
      "mean_token_accuracy": 0.1914653703570366,
      "num_tokens": 21773663.0,
      "step": 10490
    },
    {
      "entropy": 5.490184593200683,
      "epoch": 0.9520137880986937,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004916638272128301,
      "loss": 5.216,
      "mean_token_accuracy": 0.19228923469781875,
      "num_tokens": 21784265.0,
      "step": 10495
    },
    {
      "entropy": 5.456570482254028,
      "epoch": 0.9524673439767779,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916550991331777,
      "loss": 5.1794,
      "mean_token_accuracy": 0.19147132486104965,
      "num_tokens": 21795056.0,
      "step": 10500
    },
    {
      "entropy": 5.484953880310059,
      "epoch": 0.9529208998548622,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004916463665730427,
      "loss": 5.1838,
      "mean_token_accuracy": 0.1929536059498787,
      "num_tokens": 21806204.0,
      "step": 10505
    },
    {
      "entropy": 5.4476096630096436,
      "epoch": 0.9533744557329463,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916376295326057,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19114798158407212,
      "num_tokens": 21816993.0,
      "step": 10510
    },
    {
      "entropy": 5.462368154525757,
      "epoch": 0.9538280116110305,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004916288880120472,
      "loss": 5.0671,
      "mean_token_accuracy": 0.19932386875152588,
      "num_tokens": 21827457.0,
      "step": 10515
    },
    {
      "entropy": 5.480139207839966,
      "epoch": 0.9542815674891146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916201420115482,
      "loss": 5.1847,
      "mean_token_accuracy": 0.18778882622718812,
      "num_tokens": 21837555.0,
      "step": 10520
    },
    {
      "entropy": 5.487935304641724,
      "epoch": 0.9547351233671988,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916113915312895,
      "loss": 5.2163,
      "mean_token_accuracy": 0.19396420568227768,
      "num_tokens": 21847074.0,
      "step": 10525
    },
    {
      "entropy": 5.557778024673462,
      "epoch": 0.9551886792452831,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916026365714519,
      "loss": 5.2066,
      "mean_token_accuracy": 0.18701234608888626,
      "num_tokens": 21858333.0,
      "step": 10530
    },
    {
      "entropy": 5.46713604927063,
      "epoch": 0.9556422351233672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915938771322165,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19187063127756118,
      "num_tokens": 21869132.0,
      "step": 10535
    },
    {
      "entropy": 5.460751390457153,
      "epoch": 0.9560957910014514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915851132137646,
      "loss": 5.2371,
      "mean_token_accuracy": 0.189180788397789,
      "num_tokens": 21879153.0,
      "step": 10540
    },
    {
      "entropy": 5.506651306152344,
      "epoch": 0.9565493468795355,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004915763448162774,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19192200899124146,
      "num_tokens": 21889063.0,
      "step": 10545
    },
    {
      "entropy": 5.5077903270721436,
      "epoch": 0.9570029027576198,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004915675719399362,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1865275591611862,
      "num_tokens": 21899005.0,
      "step": 10550
    },
    {
      "entropy": 5.494836759567261,
      "epoch": 0.9574564586357039,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915587945849225,
      "loss": 5.1099,
      "mean_token_accuracy": 0.2017269641160965,
      "num_tokens": 21908391.0,
      "step": 10555
    },
    {
      "entropy": 5.501866388320923,
      "epoch": 0.9579100145137881,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004915500127514176,
      "loss": 5.1823,
      "mean_token_accuracy": 0.1938887879252434,
      "num_tokens": 21917655.0,
      "step": 10560
    },
    {
      "entropy": 5.5588977336883545,
      "epoch": 0.9583635703918723,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004915412264396034,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19208064824342727,
      "num_tokens": 21928038.0,
      "step": 10565
    },
    {
      "entropy": 5.502170753479004,
      "epoch": 0.9588171262699564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915324356496613,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19868719577789307,
      "num_tokens": 21938106.0,
      "step": 10570
    },
    {
      "entropy": 5.459872579574585,
      "epoch": 0.9592706821480407,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915236403817733,
      "loss": 5.1307,
      "mean_token_accuracy": 0.19554639905691146,
      "num_tokens": 21947484.0,
      "step": 10575
    },
    {
      "entropy": 5.483737945556641,
      "epoch": 0.9597242380261248,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004915148406361213,
      "loss": 5.2229,
      "mean_token_accuracy": 0.18886553943157197,
      "num_tokens": 21957139.0,
      "step": 10580
    },
    {
      "entropy": 5.475294065475464,
      "epoch": 0.960177793904209,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004915060364128872,
      "loss": 5.219,
      "mean_token_accuracy": 0.18840531557798385,
      "num_tokens": 21967030.0,
      "step": 10585
    },
    {
      "entropy": 5.461275815963745,
      "epoch": 0.9606313497822931,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914972277122532,
      "loss": 5.1727,
      "mean_token_accuracy": 0.19003562033176422,
      "num_tokens": 21976418.0,
      "step": 10590
    },
    {
      "entropy": 5.482760334014893,
      "epoch": 0.9610849056603774,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914884145344012,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19301039576530457,
      "num_tokens": 21987284.0,
      "step": 10595
    },
    {
      "entropy": 5.4282470703125,
      "epoch": 0.9615384615384616,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914795968795137,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19979993253946304,
      "num_tokens": 21997106.0,
      "step": 10600
    },
    {
      "entropy": 5.441180753707886,
      "epoch": 0.9619920174165457,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914707747477728,
      "loss": 5.0567,
      "mean_token_accuracy": 0.19796952307224275,
      "num_tokens": 22007589.0,
      "step": 10605
    },
    {
      "entropy": 5.470518398284912,
      "epoch": 0.9624455732946299,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004914619481393613,
      "loss": 5.2181,
      "mean_token_accuracy": 0.1860770180821419,
      "num_tokens": 22017669.0,
      "step": 10610
    },
    {
      "entropy": 5.617669343948364,
      "epoch": 0.962899129172714,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914531170544615,
      "loss": 5.3642,
      "mean_token_accuracy": 0.18315273523330688,
      "num_tokens": 22029330.0,
      "step": 10615
    },
    {
      "entropy": 5.45698356628418,
      "epoch": 0.9633526850507983,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004914442814932561,
      "loss": 5.1359,
      "mean_token_accuracy": 0.19763439446687697,
      "num_tokens": 22039877.0,
      "step": 10620
    },
    {
      "entropy": 5.481981801986694,
      "epoch": 0.9638062409288825,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004914354414559278,
      "loss": 5.1859,
      "mean_token_accuracy": 0.19522749185562133,
      "num_tokens": 22049579.0,
      "step": 10625
    },
    {
      "entropy": 5.393762636184692,
      "epoch": 0.9642597968069666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914265969426594,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20582508593797683,
      "num_tokens": 22059608.0,
      "step": 10630
    },
    {
      "entropy": 5.472809171676635,
      "epoch": 0.9647133526850508,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914177479536338,
      "loss": 5.2292,
      "mean_token_accuracy": 0.19040220081806183,
      "num_tokens": 22070036.0,
      "step": 10635
    },
    {
      "entropy": 5.493038988113403,
      "epoch": 0.965166908563135,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000491408894489034,
      "loss": 5.2222,
      "mean_token_accuracy": 0.18583654314279557,
      "num_tokens": 22080679.0,
      "step": 10640
    },
    {
      "entropy": 5.500444984436035,
      "epoch": 0.9656204644412192,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004914000365490432,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20014223903417588,
      "num_tokens": 22092192.0,
      "step": 10645
    },
    {
      "entropy": 5.490348196029663,
      "epoch": 0.9660740203193033,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913911741338445,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1852397084236145,
      "num_tokens": 22102722.0,
      "step": 10650
    },
    {
      "entropy": 5.406291580200195,
      "epoch": 0.9665275761973875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004913823072436211,
      "loss": 5.1155,
      "mean_token_accuracy": 0.19892334789037705,
      "num_tokens": 22113521.0,
      "step": 10655
    },
    {
      "entropy": 5.501695013046264,
      "epoch": 0.9669811320754716,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004913734358785565,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19285878390073777,
      "num_tokens": 22124891.0,
      "step": 10660
    },
    {
      "entropy": 5.486810874938965,
      "epoch": 0.9674346879535559,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004913645600388341,
      "loss": 5.2107,
      "mean_token_accuracy": 0.18844253867864608,
      "num_tokens": 22135972.0,
      "step": 10665
    },
    {
      "entropy": 5.540691900253296,
      "epoch": 0.9678882438316401,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913556797246374,
      "loss": 5.2649,
      "mean_token_accuracy": 0.18179991245269775,
      "num_tokens": 22146623.0,
      "step": 10670
    },
    {
      "entropy": 5.5169305324554445,
      "epoch": 0.9683417997097242,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913467949361503,
      "loss": 5.1767,
      "mean_token_accuracy": 0.18900265991687776,
      "num_tokens": 22157619.0,
      "step": 10675
    },
    {
      "entropy": 5.546804666519165,
      "epoch": 0.9687953555878084,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913379056735561,
      "loss": 5.1826,
      "mean_token_accuracy": 0.18539999574422836,
      "num_tokens": 22168655.0,
      "step": 10680
    },
    {
      "entropy": 5.47217059135437,
      "epoch": 0.9692489114658926,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000491329011937039,
      "loss": 5.2184,
      "mean_token_accuracy": 0.18761395961046218,
      "num_tokens": 22178500.0,
      "step": 10685
    },
    {
      "entropy": 5.515644979476929,
      "epoch": 0.9697024673439768,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004913201137267828,
      "loss": 5.224,
      "mean_token_accuracy": 0.18653703182935716,
      "num_tokens": 22188873.0,
      "step": 10690
    },
    {
      "entropy": 5.454254674911499,
      "epoch": 0.970156023222061,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913112110429715,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19080322682857515,
      "num_tokens": 22199743.0,
      "step": 10695
    },
    {
      "entropy": 5.465599822998047,
      "epoch": 0.9706095791001451,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004913023038857892,
      "loss": 5.0907,
      "mean_token_accuracy": 0.2004644140601158,
      "num_tokens": 22210301.0,
      "step": 10700
    },
    {
      "entropy": 5.4484978199005125,
      "epoch": 0.9710631349782293,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004912933922554203,
      "loss": 5.1006,
      "mean_token_accuracy": 0.1963333547115326,
      "num_tokens": 22220771.0,
      "step": 10705
    },
    {
      "entropy": 5.459470510482788,
      "epoch": 0.9715166908563135,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004912844761520487,
      "loss": 5.2071,
      "mean_token_accuracy": 0.1855422705411911,
      "num_tokens": 22231100.0,
      "step": 10710
    },
    {
      "entropy": 5.4956214427948,
      "epoch": 0.9719702467343977,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912755555758592,
      "loss": 5.1826,
      "mean_token_accuracy": 0.19422222226858138,
      "num_tokens": 22240997.0,
      "step": 10715
    },
    {
      "entropy": 5.514835929870605,
      "epoch": 0.9724238026124818,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491266630527036,
      "loss": 5.2124,
      "mean_token_accuracy": 0.189374640583992,
      "num_tokens": 22251795.0,
      "step": 10720
    },
    {
      "entropy": 5.549368095397949,
      "epoch": 0.972877358490566,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004912577010057638,
      "loss": 5.257,
      "mean_token_accuracy": 0.18586982041597366,
      "num_tokens": 22262465.0,
      "step": 10725
    },
    {
      "entropy": 5.599781894683838,
      "epoch": 0.9733309143686503,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004912487670122273,
      "loss": 5.3566,
      "mean_token_accuracy": 0.18138834536075593,
      "num_tokens": 22272399.0,
      "step": 10730
    },
    {
      "entropy": 5.5480303287506105,
      "epoch": 0.9737844702467344,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912398285466111,
      "loss": 5.2657,
      "mean_token_accuracy": 0.18869337886571885,
      "num_tokens": 22282545.0,
      "step": 10735
    },
    {
      "entropy": 5.521822595596314,
      "epoch": 0.9742380261248186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912308856091002,
      "loss": 5.2233,
      "mean_token_accuracy": 0.19182628244161606,
      "num_tokens": 22293154.0,
      "step": 10740
    },
    {
      "entropy": 5.5208924293518065,
      "epoch": 0.9746915820029027,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004912219381998795,
      "loss": 5.143,
      "mean_token_accuracy": 0.19636851102113723,
      "num_tokens": 22302309.0,
      "step": 10745
    },
    {
      "entropy": 5.558265972137451,
      "epoch": 0.9751451378809869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004912129863191339,
      "loss": 5.2559,
      "mean_token_accuracy": 0.178457310795784,
      "num_tokens": 22312316.0,
      "step": 10750
    },
    {
      "entropy": 5.517702722549439,
      "epoch": 0.9755986937590712,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912040299670487,
      "loss": 5.2337,
      "mean_token_accuracy": 0.1895495042204857,
      "num_tokens": 22322387.0,
      "step": 10755
    },
    {
      "entropy": 5.497195672988892,
      "epoch": 0.9760522496371553,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911950691438091,
      "loss": 5.22,
      "mean_token_accuracy": 0.19919940531253816,
      "num_tokens": 22332759.0,
      "step": 10760
    },
    {
      "entropy": 5.516394901275635,
      "epoch": 0.9765058055152395,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004911861038496004,
      "loss": 5.2637,
      "mean_token_accuracy": 0.18964977264404298,
      "num_tokens": 22344056.0,
      "step": 10765
    },
    {
      "entropy": 5.451268243789673,
      "epoch": 0.9769593613933236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004911771340846079,
      "loss": 5.2089,
      "mean_token_accuracy": 0.1823955789208412,
      "num_tokens": 22353877.0,
      "step": 10770
    },
    {
      "entropy": 5.557831621170044,
      "epoch": 0.9774129172714079,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911681598490172,
      "loss": 5.2351,
      "mean_token_accuracy": 0.18687503039836884,
      "num_tokens": 22363549.0,
      "step": 10775
    },
    {
      "entropy": 5.593202066421509,
      "epoch": 0.977866473149492,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911591811430139,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18995759636163712,
      "num_tokens": 22374040.0,
      "step": 10780
    },
    {
      "entropy": 5.54970760345459,
      "epoch": 0.9783200290275762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911501979667836,
      "loss": 5.3196,
      "mean_token_accuracy": 0.1861735463142395,
      "num_tokens": 22384819.0,
      "step": 10785
    },
    {
      "entropy": 5.487377309799195,
      "epoch": 0.9787735849056604,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004911412103205122,
      "loss": 5.1873,
      "mean_token_accuracy": 0.1898432657122612,
      "num_tokens": 22395130.0,
      "step": 10790
    },
    {
      "entropy": 5.503686380386353,
      "epoch": 0.9792271407837445,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911322182043855,
      "loss": 5.2081,
      "mean_token_accuracy": 0.18677622675895691,
      "num_tokens": 22405093.0,
      "step": 10795
    },
    {
      "entropy": 5.51582441329956,
      "epoch": 0.9796806966618288,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004911232216185893,
      "loss": 5.2269,
      "mean_token_accuracy": 0.18393291234970094,
      "num_tokens": 22415008.0,
      "step": 10800
    },
    {
      "entropy": 5.5575262069702145,
      "epoch": 0.9801342525399129,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004911142205633099,
      "loss": 5.2651,
      "mean_token_accuracy": 0.18396617025136947,
      "num_tokens": 22426486.0,
      "step": 10805
    },
    {
      "entropy": 5.4911479473114015,
      "epoch": 0.9805878084179971,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004911052150387334,
      "loss": 5.1708,
      "mean_token_accuracy": 0.19237590581178665,
      "num_tokens": 22436844.0,
      "step": 10810
    },
    {
      "entropy": 5.568431949615478,
      "epoch": 0.9810413642960812,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004910962050450458,
      "loss": 5.3003,
      "mean_token_accuracy": 0.1872747138142586,
      "num_tokens": 22446630.0,
      "step": 10815
    },
    {
      "entropy": 5.56194052696228,
      "epoch": 0.9814949201741655,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910871905824338,
      "loss": 5.1379,
      "mean_token_accuracy": 0.1935226395726204,
      "num_tokens": 22456406.0,
      "step": 10820
    },
    {
      "entropy": 5.460543298721314,
      "epoch": 0.9819484760522497,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004910781716510837,
      "loss": 5.1251,
      "mean_token_accuracy": 0.19543748646974562,
      "num_tokens": 22466210.0,
      "step": 10825
    },
    {
      "entropy": 5.49596643447876,
      "epoch": 0.9824020319303338,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910691482511817,
      "loss": 5.2273,
      "mean_token_accuracy": 0.1921169638633728,
      "num_tokens": 22476978.0,
      "step": 10830
    },
    {
      "entropy": 5.588016319274902,
      "epoch": 0.982855587808418,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004910601203829148,
      "loss": 5.1929,
      "mean_token_accuracy": 0.18680335581302643,
      "num_tokens": 22488613.0,
      "step": 10835
    },
    {
      "entropy": 5.482051849365234,
      "epoch": 0.9833091436865021,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004910510880464696,
      "loss": 5.0861,
      "mean_token_accuracy": 0.2025364950299263,
      "num_tokens": 22498312.0,
      "step": 10840
    },
    {
      "entropy": 5.485942792892456,
      "epoch": 0.9837626995645864,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004910420512420327,
      "loss": 5.2127,
      "mean_token_accuracy": 0.1877727434039116,
      "num_tokens": 22508131.0,
      "step": 10845
    },
    {
      "entropy": 5.513615369796753,
      "epoch": 0.9842162554426706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910330099697913,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1947124034166336,
      "num_tokens": 22517842.0,
      "step": 10850
    },
    {
      "entropy": 5.540058898925781,
      "epoch": 0.9846698113207547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910239642299321,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19113553911447526,
      "num_tokens": 22528239.0,
      "step": 10855
    },
    {
      "entropy": 5.472961711883545,
      "epoch": 0.9851233671988389,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004910149140226422,
      "loss": 5.1559,
      "mean_token_accuracy": 0.19106702506542206,
      "num_tokens": 22539734.0,
      "step": 10860
    },
    {
      "entropy": 5.481329441070557,
      "epoch": 0.9855769230769231,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004910058593481089,
      "loss": 5.1549,
      "mean_token_accuracy": 0.19573762267827988,
      "num_tokens": 22549796.0,
      "step": 10865
    },
    {
      "entropy": 5.457813835144043,
      "epoch": 0.9860304789550073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909968002065195,
      "loss": 5.137,
      "mean_token_accuracy": 0.19690064191818238,
      "num_tokens": 22559934.0,
      "step": 10870
    },
    {
      "entropy": 5.597090387344361,
      "epoch": 0.9864840348330914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000490987736598061,
      "loss": 5.26,
      "mean_token_accuracy": 0.1757528841495514,
      "num_tokens": 22569864.0,
      "step": 10875
    },
    {
      "entropy": 5.525681018829346,
      "epoch": 0.9869375907111756,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909786685229211,
      "loss": 5.1874,
      "mean_token_accuracy": 0.19665819704532622,
      "num_tokens": 22580687.0,
      "step": 10880
    },
    {
      "entropy": 5.555548810958863,
      "epoch": 0.9873911465892597,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909695959812875,
      "loss": 5.2377,
      "mean_token_accuracy": 0.18621382713317872,
      "num_tokens": 22590507.0,
      "step": 10885
    },
    {
      "entropy": 5.540699434280396,
      "epoch": 0.987844702467344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004909605189733474,
      "loss": 5.2675,
      "mean_token_accuracy": 0.19158151596784592,
      "num_tokens": 22601901.0,
      "step": 10890
    },
    {
      "entropy": 5.478123521804809,
      "epoch": 0.9882982583454282,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909514374992889,
      "loss": 5.2022,
      "mean_token_accuracy": 0.18934640884399415,
      "num_tokens": 22613003.0,
      "step": 10895
    },
    {
      "entropy": 5.444561386108399,
      "epoch": 0.9887518142235123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909423515592996,
      "loss": 5.0531,
      "mean_token_accuracy": 0.1984160289168358,
      "num_tokens": 22623401.0,
      "step": 10900
    },
    {
      "entropy": 5.533397102355957,
      "epoch": 0.9892053701015965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004909332611535675,
      "loss": 5.2067,
      "mean_token_accuracy": 0.1915249705314636,
      "num_tokens": 22633147.0,
      "step": 10905
    },
    {
      "entropy": 5.486864948272705,
      "epoch": 0.9896589259796807,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909241662822804,
      "loss": 5.0547,
      "mean_token_accuracy": 0.20758973360061644,
      "num_tokens": 22642990.0,
      "step": 10910
    },
    {
      "entropy": 5.489868307113648,
      "epoch": 0.9901124818577649,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909150669456265,
      "loss": 5.2756,
      "mean_token_accuracy": 0.18449527472257615,
      "num_tokens": 22652734.0,
      "step": 10915
    },
    {
      "entropy": 5.500755023956299,
      "epoch": 0.9905660377358491,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000490905963143794,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18280779719352722,
      "num_tokens": 22662968.0,
      "step": 10920
    },
    {
      "entropy": 5.448152780532837,
      "epoch": 0.9910195936139332,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908968548769713,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20330245196819305,
      "num_tokens": 22673157.0,
      "step": 10925
    },
    {
      "entropy": 5.517993688583374,
      "epoch": 0.9914731494920174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908877421453466,
      "loss": 5.2567,
      "mean_token_accuracy": 0.19120330810546876,
      "num_tokens": 22683105.0,
      "step": 10930
    },
    {
      "entropy": 5.48880033493042,
      "epoch": 0.9919267053701016,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908786249491082,
      "loss": 5.1906,
      "mean_token_accuracy": 0.19872619658708573,
      "num_tokens": 22693793.0,
      "step": 10935
    },
    {
      "entropy": 5.597190093994141,
      "epoch": 0.9923802612481858,
      "grad_norm": 1.015625,
      "learning_rate": 0.000490869503288445,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19000887721776963,
      "num_tokens": 22703884.0,
      "step": 10940
    },
    {
      "entropy": 5.532585954666137,
      "epoch": 0.99283381712627,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908603771635454,
      "loss": 5.2745,
      "mean_token_accuracy": 0.18670693933963775,
      "num_tokens": 22713699.0,
      "step": 10945
    },
    {
      "entropy": 5.5036516189575195,
      "epoch": 0.9932873730043541,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908512465745982,
      "loss": 5.1147,
      "mean_token_accuracy": 0.19385309815406798,
      "num_tokens": 22722871.0,
      "step": 10950
    },
    {
      "entropy": 5.504653596878052,
      "epoch": 0.9937409288824384,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004908421115217922,
      "loss": 5.1247,
      "mean_token_accuracy": 0.18782687783241273,
      "num_tokens": 22733314.0,
      "step": 10955
    },
    {
      "entropy": 5.431400537490845,
      "epoch": 0.9941944847605225,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004908329720053164,
      "loss": 5.1209,
      "mean_token_accuracy": 0.1992374211549759,
      "num_tokens": 22743255.0,
      "step": 10960
    },
    {
      "entropy": 5.420324754714966,
      "epoch": 0.9946480406386067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004908238280253597,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19695128947496415,
      "num_tokens": 22753857.0,
      "step": 10965
    },
    {
      "entropy": 5.496116971969604,
      "epoch": 0.9951015965166908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908146795821112,
      "loss": 5.1057,
      "mean_token_accuracy": 0.19125611186027527,
      "num_tokens": 22764326.0,
      "step": 10970
    },
    {
      "entropy": 5.502989768981934,
      "epoch": 0.995555152394775,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908055266757602,
      "loss": 5.1898,
      "mean_token_accuracy": 0.19550111889839172,
      "num_tokens": 22774958.0,
      "step": 10975
    },
    {
      "entropy": 5.408831548690796,
      "epoch": 0.9960087082728593,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004907963693064958,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1952947497367859,
      "num_tokens": 22785616.0,
      "step": 10980
    },
    {
      "entropy": 5.565986251831054,
      "epoch": 0.9964622641509434,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004907872074745075,
      "loss": 5.2869,
      "mean_token_accuracy": 0.18623248487710953,
      "num_tokens": 22795967.0,
      "step": 10985
    },
    {
      "entropy": 5.514637279510498,
      "epoch": 0.9969158200290276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907780411799849,
      "loss": 5.2722,
      "mean_token_accuracy": 0.18776020854711534,
      "num_tokens": 22806335.0,
      "step": 10990
    },
    {
      "entropy": 5.554602861404419,
      "epoch": 0.9973693759071117,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907688704231173,
      "loss": 5.2089,
      "mean_token_accuracy": 0.19053253084421157,
      "num_tokens": 22816390.0,
      "step": 10995
    },
    {
      "entropy": 5.597799396514892,
      "epoch": 0.997822931785196,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004907596952040945,
      "loss": 5.2777,
      "mean_token_accuracy": 0.18841657638549805,
      "num_tokens": 22826395.0,
      "step": 11000
    },
    {
      "entropy": 5.564309597015381,
      "epoch": 0.9982764876632801,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004907505155231063,
      "loss": 5.2449,
      "mean_token_accuracy": 0.18635112643241883,
      "num_tokens": 22836790.0,
      "step": 11005
    },
    {
      "entropy": 5.494991683959961,
      "epoch": 0.9987300435413643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907413313803424,
      "loss": 5.1386,
      "mean_token_accuracy": 0.19103351086378098,
      "num_tokens": 22846687.0,
      "step": 11010
    },
    {
      "entropy": 5.538417148590088,
      "epoch": 0.9991835994194485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907321427759928,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19118976593017578,
      "num_tokens": 22857208.0,
      "step": 11015
    },
    {
      "entropy": 5.539492750167847,
      "epoch": 0.9996371552975326,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004907229497102475,
      "loss": 5.2696,
      "mean_token_accuracy": 0.1830685555934906,
      "num_tokens": 22868556.0,
      "step": 11020
    },
    {
      "entropy": 5.554223203659058,
      "epoch": 1.0000907111756168,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004907137521832966,
      "loss": 5.2441,
      "mean_token_accuracy": 0.18953093886375427,
      "num_tokens": 22880081.0,
      "step": 11025
    },
    {
      "entropy": 5.496646070480347,
      "epoch": 1.000544267053701,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004907045501953304,
      "loss": 5.0765,
      "mean_token_accuracy": 0.19675455540418624,
      "num_tokens": 22892141.0,
      "step": 11030
    },
    {
      "entropy": 5.515450572967529,
      "epoch": 1.0009978229317853,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000490695343746539,
      "loss": 5.0916,
      "mean_token_accuracy": 0.19881475418806077,
      "num_tokens": 22902688.0,
      "step": 11035
    },
    {
      "entropy": 5.431853437423706,
      "epoch": 1.0014513788098693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004906861328371131,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19932159036397934,
      "num_tokens": 22912453.0,
      "step": 11040
    },
    {
      "entropy": 5.550728511810303,
      "epoch": 1.0019049346879536,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004906769174672429,
      "loss": 5.1846,
      "mean_token_accuracy": 0.18711721748113633,
      "num_tokens": 22923167.0,
      "step": 11045
    },
    {
      "entropy": 5.52794919013977,
      "epoch": 1.0023584905660377,
      "grad_norm": 0.96875,
      "learning_rate": 0.000490667697637119,
      "loss": 5.1534,
      "mean_token_accuracy": 0.18911569565534592,
      "num_tokens": 22934753.0,
      "step": 11050
    },
    {
      "entropy": 5.456609916687012,
      "epoch": 1.002812046444122,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004906584733469324,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20015866309404373,
      "num_tokens": 22945277.0,
      "step": 11055
    },
    {
      "entropy": 5.559716701507568,
      "epoch": 1.0032656023222062,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004906492445968735,
      "loss": 5.1634,
      "mean_token_accuracy": 0.18062861263751984,
      "num_tokens": 22955797.0,
      "step": 11060
    },
    {
      "entropy": 5.544274425506591,
      "epoch": 1.0037191582002902,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906400113871332,
      "loss": 5.132,
      "mean_token_accuracy": 0.19073731899261476,
      "num_tokens": 22965969.0,
      "step": 11065
    },
    {
      "entropy": 5.493846797943116,
      "epoch": 1.0041727140783745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004906307737179026,
      "loss": 5.1033,
      "mean_token_accuracy": 0.20635355561971663,
      "num_tokens": 22975323.0,
      "step": 11070
    },
    {
      "entropy": 5.447731399536133,
      "epoch": 1.0046262699564585,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004906215315893726,
      "loss": 5.108,
      "mean_token_accuracy": 0.19876318871974946,
      "num_tokens": 22985063.0,
      "step": 11075
    },
    {
      "entropy": 5.5594007015228275,
      "epoch": 1.0050798258345428,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004906122850017344,
      "loss": 5.2382,
      "mean_token_accuracy": 0.18069565892219544,
      "num_tokens": 22996001.0,
      "step": 11080
    },
    {
      "entropy": 5.500598669052124,
      "epoch": 1.005533381712627,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906030339551792,
      "loss": 5.1468,
      "mean_token_accuracy": 0.19543736428022385,
      "num_tokens": 23006896.0,
      "step": 11085
    },
    {
      "entropy": 5.514100217819214,
      "epoch": 1.0059869375907111,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905937784498983,
      "loss": 5.1205,
      "mean_token_accuracy": 0.1983324557542801,
      "num_tokens": 23017080.0,
      "step": 11090
    },
    {
      "entropy": 5.526914501190186,
      "epoch": 1.0064404934687954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004905845184860831,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19216082692146302,
      "num_tokens": 23027168.0,
      "step": 11095
    },
    {
      "entropy": 5.4901073455810545,
      "epoch": 1.0068940493468794,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905752540639252,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19402697086334228,
      "num_tokens": 23037364.0,
      "step": 11100
    },
    {
      "entropy": 5.444597339630127,
      "epoch": 1.0073476052249637,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004905659851836161,
      "loss": 5.0147,
      "mean_token_accuracy": 0.2009423241019249,
      "num_tokens": 23047609.0,
      "step": 11105
    },
    {
      "entropy": 5.493267250061035,
      "epoch": 1.007801161103048,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004905567118453475,
      "loss": 5.1704,
      "mean_token_accuracy": 0.18859416395425796,
      "num_tokens": 23057942.0,
      "step": 11110
    },
    {
      "entropy": 5.469370317459107,
      "epoch": 1.008254716981132,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004905474340493111,
      "loss": 5.0764,
      "mean_token_accuracy": 0.1908912554383278,
      "num_tokens": 23069119.0,
      "step": 11115
    },
    {
      "entropy": 5.45906629562378,
      "epoch": 1.0087082728592163,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004905381517956988,
      "loss": 5.0678,
      "mean_token_accuracy": 0.20038813650608062,
      "num_tokens": 23080060.0,
      "step": 11120
    },
    {
      "entropy": 5.434975862503052,
      "epoch": 1.0091618287373005,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004905288650847027,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19017168432474135,
      "num_tokens": 23090645.0,
      "step": 11125
    },
    {
      "entropy": 5.517891359329224,
      "epoch": 1.0096153846153846,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004905195739165146,
      "loss": 5.108,
      "mean_token_accuracy": 0.19634836316108703,
      "num_tokens": 23100449.0,
      "step": 11130
    },
    {
      "entropy": 5.488073348999023,
      "epoch": 1.0100689404934688,
      "grad_norm": 1.0,
      "learning_rate": 0.000490510278291327,
      "loss": 5.0931,
      "mean_token_accuracy": 0.1908187672495842,
      "num_tokens": 23110415.0,
      "step": 11135
    },
    {
      "entropy": 5.465580987930298,
      "epoch": 1.010522496371553,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004905009782093318,
      "loss": 5.08,
      "mean_token_accuracy": 0.19305008202791213,
      "num_tokens": 23120336.0,
      "step": 11140
    },
    {
      "entropy": 5.442973518371582,
      "epoch": 1.0109760522496372,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004904916736707214,
      "loss": 5.157,
      "mean_token_accuracy": 0.18449659645557404,
      "num_tokens": 23130568.0,
      "step": 11145
    },
    {
      "entropy": 5.543105840682983,
      "epoch": 1.0114296081277214,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004904823646756885,
      "loss": 5.112,
      "mean_token_accuracy": 0.1943865343928337,
      "num_tokens": 23140080.0,
      "step": 11150
    },
    {
      "entropy": 5.483130693435669,
      "epoch": 1.0118831640058055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004904730512244252,
      "loss": 5.0452,
      "mean_token_accuracy": 0.1947724238038063,
      "num_tokens": 23150360.0,
      "step": 11155
    },
    {
      "entropy": 5.424828720092774,
      "epoch": 1.0123367198838897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904637333171243,
      "loss": 5.0556,
      "mean_token_accuracy": 0.19245949983596802,
      "num_tokens": 23160620.0,
      "step": 11160
    },
    {
      "entropy": 5.550334787368774,
      "epoch": 1.0127902757619738,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004904544109539785,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19757008105516433,
      "num_tokens": 23171920.0,
      "step": 11165
    },
    {
      "entropy": 5.491248846054077,
      "epoch": 1.013243831640058,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004904450841351805,
      "loss": 5.1023,
      "mean_token_accuracy": 0.1991717129945755,
      "num_tokens": 23182929.0,
      "step": 11170
    },
    {
      "entropy": 5.514200067520141,
      "epoch": 1.0136973875181423,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004904357528609236,
      "loss": 5.0722,
      "mean_token_accuracy": 0.19128975570201873,
      "num_tokens": 23193020.0,
      "step": 11175
    },
    {
      "entropy": 5.448251628875733,
      "epoch": 1.0141509433962264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004904264171314,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20164622366428375,
      "num_tokens": 23204073.0,
      "step": 11180
    },
    {
      "entropy": 5.421363639831543,
      "epoch": 1.0146044992743106,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004904170769468035,
      "loss": 4.9913,
      "mean_token_accuracy": 0.19620340317487717,
      "num_tokens": 23214376.0,
      "step": 11185
    },
    {
      "entropy": 5.5082276344299315,
      "epoch": 1.0150580551523947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004904077323073268,
      "loss": 5.1497,
      "mean_token_accuracy": 0.19836536943912506,
      "num_tokens": 23224304.0,
      "step": 11190
    },
    {
      "entropy": 5.492418527603149,
      "epoch": 1.015511611030479,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004903983832131633,
      "loss": 5.1656,
      "mean_token_accuracy": 0.1937434956431389,
      "num_tokens": 23235324.0,
      "step": 11195
    },
    {
      "entropy": 5.493016624450684,
      "epoch": 1.0159651669085632,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004903890296645063,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19382944256067275,
      "num_tokens": 23245774.0,
      "step": 11200
    },
    {
      "entropy": 5.507924365997314,
      "epoch": 1.0164187227866472,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004903796716615495,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19998153001070024,
      "num_tokens": 23256274.0,
      "step": 11205
    },
    {
      "entropy": 5.483238792419433,
      "epoch": 1.0168722786647315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903703092044861,
      "loss": 5.0927,
      "mean_token_accuracy": 0.19459670037031174,
      "num_tokens": 23266304.0,
      "step": 11210
    },
    {
      "entropy": 5.416814374923706,
      "epoch": 1.0173258345428158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903609422935097,
      "loss": 5.0196,
      "mean_token_accuracy": 0.19843120574951173,
      "num_tokens": 23276552.0,
      "step": 11215
    },
    {
      "entropy": 5.462734937667847,
      "epoch": 1.0177793904208998,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004903515709288143,
      "loss": 5.1272,
      "mean_token_accuracy": 0.1896622136235237,
      "num_tokens": 23287799.0,
      "step": 11220
    },
    {
      "entropy": 5.425547027587891,
      "epoch": 1.018232946298984,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004903421951105933,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2000567466020584,
      "num_tokens": 23297958.0,
      "step": 11225
    },
    {
      "entropy": 5.490804672241211,
      "epoch": 1.0186865021770681,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004903328148390411,
      "loss": 5.1606,
      "mean_token_accuracy": 0.18824684619903564,
      "num_tokens": 23307523.0,
      "step": 11230
    },
    {
      "entropy": 5.533596992492676,
      "epoch": 1.0191400580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004903234301143513,
      "loss": 5.1305,
      "mean_token_accuracy": 0.1967354640364647,
      "num_tokens": 23317970.0,
      "step": 11235
    },
    {
      "entropy": 5.465874528884887,
      "epoch": 1.0195936139332367,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004903140409367181,
      "loss": 5.1225,
      "mean_token_accuracy": 0.19505420625209807,
      "num_tokens": 23328612.0,
      "step": 11240
    },
    {
      "entropy": 5.415059518814087,
      "epoch": 1.0200471698113207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004903046473063357,
      "loss": 5.0724,
      "mean_token_accuracy": 0.18936332017183305,
      "num_tokens": 23338104.0,
      "step": 11245
    },
    {
      "entropy": 5.445874977111816,
      "epoch": 1.020500725689405,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004902952492233983,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21072589308023454,
      "num_tokens": 23348105.0,
      "step": 11250
    },
    {
      "entropy": 5.442014884948731,
      "epoch": 1.020954281567489,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004902858466881002,
      "loss": 4.9613,
      "mean_token_accuracy": 0.19927150905132293,
      "num_tokens": 23357150.0,
      "step": 11255
    },
    {
      "entropy": 5.465661239624024,
      "epoch": 1.0214078374455733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490276439700636,
      "loss": 5.1693,
      "mean_token_accuracy": 0.1910180225968361,
      "num_tokens": 23367223.0,
      "step": 11260
    },
    {
      "entropy": 5.492317008972168,
      "epoch": 1.0218613933236576,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004902670282612002,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1948421135544777,
      "num_tokens": 23377664.0,
      "step": 11265
    },
    {
      "entropy": 5.424583578109742,
      "epoch": 1.0223149492017416,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004902576123699873,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20003152638673782,
      "num_tokens": 23388057.0,
      "step": 11270
    },
    {
      "entropy": 5.3437347412109375,
      "epoch": 1.0227685050798259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902481920271922,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20076482594013215,
      "num_tokens": 23397808.0,
      "step": 11275
    },
    {
      "entropy": 5.3924784660339355,
      "epoch": 1.02322206095791,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004902387672330097,
      "loss": 5.0345,
      "mean_token_accuracy": 0.19002968221902847,
      "num_tokens": 23406979.0,
      "step": 11280
    },
    {
      "entropy": 5.443336629867554,
      "epoch": 1.0236756168359942,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902293379876345,
      "loss": 5.0838,
      "mean_token_accuracy": 0.19750918447971344,
      "num_tokens": 23416545.0,
      "step": 11285
    },
    {
      "entropy": 5.479193782806396,
      "epoch": 1.0241291727140784,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004902199042912618,
      "loss": 5.1001,
      "mean_token_accuracy": 0.18947370052337648,
      "num_tokens": 23427464.0,
      "step": 11290
    },
    {
      "entropy": 5.497503757476807,
      "epoch": 1.0245827285921625,
      "grad_norm": 1.0,
      "learning_rate": 0.0004902104661440868,
      "loss": 5.1211,
      "mean_token_accuracy": 0.19749965518712997,
      "num_tokens": 23436690.0,
      "step": 11295
    },
    {
      "entropy": 5.464810943603515,
      "epoch": 1.0250362844702468,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902010235463042,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19768522083759307,
      "num_tokens": 23446701.0,
      "step": 11300
    },
    {
      "entropy": 5.534040021896362,
      "epoch": 1.025489840348331,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004901915764981098,
      "loss": 5.1473,
      "mean_token_accuracy": 0.19012537896633147,
      "num_tokens": 23456966.0,
      "step": 11305
    },
    {
      "entropy": 5.495480489730835,
      "epoch": 1.025943396226415,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004901821249996988,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19729931354522706,
      "num_tokens": 23467300.0,
      "step": 11310
    },
    {
      "entropy": 5.5246494770050045,
      "epoch": 1.0263969521044993,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004901726690512665,
      "loss": 5.1462,
      "mean_token_accuracy": 0.1901402324438095,
      "num_tokens": 23478457.0,
      "step": 11315
    },
    {
      "entropy": 5.463982629776001,
      "epoch": 1.0268505079825834,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004901632086530087,
      "loss": 5.0403,
      "mean_token_accuracy": 0.19925176948308945,
      "num_tokens": 23488066.0,
      "step": 11320
    },
    {
      "entropy": 5.4931072235107425,
      "epoch": 1.0273040638606676,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004901537438051207,
      "loss": 5.1304,
      "mean_token_accuracy": 0.1888103410601616,
      "num_tokens": 23498856.0,
      "step": 11325
    },
    {
      "entropy": 5.479607105255127,
      "epoch": 1.027757619738752,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004901442745077986,
      "loss": 5.0765,
      "mean_token_accuracy": 0.196649968624115,
      "num_tokens": 23509845.0,
      "step": 11330
    },
    {
      "entropy": 5.557180452346802,
      "epoch": 1.028211175616836,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004901348007612381,
      "loss": 5.2062,
      "mean_token_accuracy": 0.19007708877325058,
      "num_tokens": 23520322.0,
      "step": 11335
    },
    {
      "entropy": 5.499701595306396,
      "epoch": 1.0286647314949202,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901253225656351,
      "loss": 5.1263,
      "mean_token_accuracy": 0.19733863621950148,
      "num_tokens": 23530030.0,
      "step": 11340
    },
    {
      "entropy": 5.559247732162476,
      "epoch": 1.0291182873730043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004901158399211855,
      "loss": 5.1838,
      "mean_token_accuracy": 0.19454871863126755,
      "num_tokens": 23541063.0,
      "step": 11345
    },
    {
      "entropy": 5.469040298461914,
      "epoch": 1.0295718432510885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004901063528280856,
      "loss": 5.0867,
      "mean_token_accuracy": 0.1939779043197632,
      "num_tokens": 23551693.0,
      "step": 11350
    },
    {
      "entropy": 5.4881370067596436,
      "epoch": 1.0300253991291728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900968612865314,
      "loss": 5.1351,
      "mean_token_accuracy": 0.19694630354642867,
      "num_tokens": 23561758.0,
      "step": 11355
    },
    {
      "entropy": 5.429140520095825,
      "epoch": 1.0304789550072568,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900873652967195,
      "loss": 5.0771,
      "mean_token_accuracy": 0.18856350779533387,
      "num_tokens": 23571912.0,
      "step": 11360
    },
    {
      "entropy": 5.442412996292115,
      "epoch": 1.030932510885341,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004900778648588459,
      "loss": 5.0333,
      "mean_token_accuracy": 0.20146304965019227,
      "num_tokens": 23582600.0,
      "step": 11365
    },
    {
      "entropy": 5.472327375411988,
      "epoch": 1.0313860667634251,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004900683599731073,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20228659957647324,
      "num_tokens": 23592761.0,
      "step": 11370
    },
    {
      "entropy": 5.564744472503662,
      "epoch": 1.0318396226415094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900588506397002,
      "loss": 5.159,
      "mean_token_accuracy": 0.19236104786396027,
      "num_tokens": 23603119.0,
      "step": 11375
    },
    {
      "entropy": 5.404077386856079,
      "epoch": 1.0322931785195937,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004900493368588213,
      "loss": 5.068,
      "mean_token_accuracy": 0.20131881684064865,
      "num_tokens": 23613916.0,
      "step": 11380
    },
    {
      "entropy": 5.535603141784668,
      "epoch": 1.0327467343976777,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004900398186306672,
      "loss": 5.2023,
      "mean_token_accuracy": 0.19502190947532655,
      "num_tokens": 23623956.0,
      "step": 11385
    },
    {
      "entropy": 5.452536010742188,
      "epoch": 1.033200290275762,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000490030295955435,
      "loss": 5.0969,
      "mean_token_accuracy": 0.19548225551843643,
      "num_tokens": 23634666.0,
      "step": 11390
    },
    {
      "entropy": 5.414992809295654,
      "epoch": 1.0336538461538463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900207688333215,
      "loss": 5.0505,
      "mean_token_accuracy": 0.19136851727962495,
      "num_tokens": 23645601.0,
      "step": 11395
    },
    {
      "entropy": 5.4911417961120605,
      "epoch": 1.0341074020319303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900112372645236,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19436682015657425,
      "num_tokens": 23656339.0,
      "step": 11400
    },
    {
      "entropy": 5.497488927841187,
      "epoch": 1.0345609579100146,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004900017012492386,
      "loss": 5.0899,
      "mean_token_accuracy": 0.1950056940317154,
      "num_tokens": 23667433.0,
      "step": 11405
    },
    {
      "entropy": 5.528972911834717,
      "epoch": 1.0350145137880986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004899921607876634,
      "loss": 5.1392,
      "mean_token_accuracy": 0.1908893272280693,
      "num_tokens": 23678186.0,
      "step": 11410
    },
    {
      "entropy": 5.481750106811523,
      "epoch": 1.0354680696661829,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004899826158799958,
      "loss": 5.0724,
      "mean_token_accuracy": 0.19224554300308228,
      "num_tokens": 23688317.0,
      "step": 11415
    },
    {
      "entropy": 5.4784728527069095,
      "epoch": 1.0359216255442671,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899730665264328,
      "loss": 5.0224,
      "mean_token_accuracy": 0.19699247032403946,
      "num_tokens": 23697752.0,
      "step": 11420
    },
    {
      "entropy": 5.447753953933716,
      "epoch": 1.0363751814223512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899635127271721,
      "loss": 4.9843,
      "mean_token_accuracy": 0.20179656744003296,
      "num_tokens": 23707803.0,
      "step": 11425
    },
    {
      "entropy": 5.401472043991089,
      "epoch": 1.0368287373004355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899539544824111,
      "loss": 5.0116,
      "mean_token_accuracy": 0.2051843896508217,
      "num_tokens": 23717630.0,
      "step": 11430
    },
    {
      "entropy": 5.473198318481446,
      "epoch": 1.0372822931785195,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899443917923476,
      "loss": 5.1287,
      "mean_token_accuracy": 0.19700049012899398,
      "num_tokens": 23728836.0,
      "step": 11435
    },
    {
      "entropy": 5.452097368240357,
      "epoch": 1.0377358490566038,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004899348246571792,
      "loss": 5.1174,
      "mean_token_accuracy": 0.1917497396469116,
      "num_tokens": 23740145.0,
      "step": 11440
    },
    {
      "entropy": 5.4935462951660154,
      "epoch": 1.038189404934688,
      "grad_norm": 0.96875,
      "learning_rate": 0.000489925253077104,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19191719591617584,
      "num_tokens": 23751162.0,
      "step": 11445
    },
    {
      "entropy": 5.504214334487915,
      "epoch": 1.038642960812772,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004899156770523196,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19370600432157517,
      "num_tokens": 23761634.0,
      "step": 11450
    },
    {
      "entropy": 5.47665491104126,
      "epoch": 1.0390965166908563,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004899060965830244,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20089899897575378,
      "num_tokens": 23772483.0,
      "step": 11455
    },
    {
      "entropy": 5.452212142944336,
      "epoch": 1.0395500725689404,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004898965116694162,
      "loss": 5.0453,
      "mean_token_accuracy": 0.1946641504764557,
      "num_tokens": 23782298.0,
      "step": 11460
    },
    {
      "entropy": 5.494813919067383,
      "epoch": 1.0400036284470247,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898869223116935,
      "loss": 5.1309,
      "mean_token_accuracy": 0.18946383744478226,
      "num_tokens": 23792492.0,
      "step": 11465
    },
    {
      "entropy": 5.54174313545227,
      "epoch": 1.040457184325109,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898773285100546,
      "loss": 5.256,
      "mean_token_accuracy": 0.18755862563848497,
      "num_tokens": 23802999.0,
      "step": 11470
    },
    {
      "entropy": 5.487423419952393,
      "epoch": 1.040910740203193,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004898677302646975,
      "loss": 5.0665,
      "mean_token_accuracy": 0.1984709843993187,
      "num_tokens": 23812541.0,
      "step": 11475
    },
    {
      "entropy": 5.466430902481079,
      "epoch": 1.0413642960812772,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004898581275758211,
      "loss": 5.1706,
      "mean_token_accuracy": 0.18461524546146393,
      "num_tokens": 23823244.0,
      "step": 11480
    },
    {
      "entropy": 5.467811203002929,
      "epoch": 1.0418178519593615,
      "grad_norm": 1.0,
      "learning_rate": 0.0004898485204436239,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20237330198287964,
      "num_tokens": 23832346.0,
      "step": 11485
    },
    {
      "entropy": 5.515140390396118,
      "epoch": 1.0422714078374455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004898389088683047,
      "loss": 5.1301,
      "mean_token_accuracy": 0.20025138109922408,
      "num_tokens": 23842554.0,
      "step": 11490
    },
    {
      "entropy": 5.392171144485474,
      "epoch": 1.0427249637155298,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004898292928500619,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20212062299251557,
      "num_tokens": 23852938.0,
      "step": 11495
    },
    {
      "entropy": 5.421096229553223,
      "epoch": 1.0431785195936139,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004898196723890947,
      "loss": 4.9989,
      "mean_token_accuracy": 0.20295965373516084,
      "num_tokens": 23862367.0,
      "step": 11500
    },
    {
      "entropy": 5.464499378204346,
      "epoch": 1.0436320754716981,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898100474856019,
      "loss": 5.1109,
      "mean_token_accuracy": 0.1982620269060135,
      "num_tokens": 23873249.0,
      "step": 11505
    },
    {
      "entropy": 5.475397109985352,
      "epoch": 1.0440856313497824,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004898004181397824,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1877864271402359,
      "num_tokens": 23883830.0,
      "step": 11510
    },
    {
      "entropy": 5.5174485206604,
      "epoch": 1.0445391872278664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004897907843518358,
      "loss": 5.0633,
      "mean_token_accuracy": 0.19433184564113617,
      "num_tokens": 23894359.0,
      "step": 11515
    },
    {
      "entropy": 5.46463451385498,
      "epoch": 1.0449927431059507,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004897811461219609,
      "loss": 5.0531,
      "mean_token_accuracy": 0.19569095522165297,
      "num_tokens": 23903779.0,
      "step": 11520
    },
    {
      "entropy": 5.448574495315552,
      "epoch": 1.0454462989840347,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004897715034503571,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19483239948749542,
      "num_tokens": 23914523.0,
      "step": 11525
    },
    {
      "entropy": 5.456052017211914,
      "epoch": 1.045899854862119,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000489761856337224,
      "loss": 5.0868,
      "mean_token_accuracy": 0.19752579927444458,
      "num_tokens": 23924996.0,
      "step": 11530
    },
    {
      "entropy": 5.469106912612915,
      "epoch": 1.0463534107402033,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004897522047827609,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20350758582353592,
      "num_tokens": 23935843.0,
      "step": 11535
    },
    {
      "entropy": 5.433482885360718,
      "epoch": 1.0468069666182873,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004897425487871676,
      "loss": 5.0343,
      "mean_token_accuracy": 0.1945498988032341,
      "num_tokens": 23945044.0,
      "step": 11540
    },
    {
      "entropy": 5.449752473831177,
      "epoch": 1.0472605224963716,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897328883506436,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19367363899946213,
      "num_tokens": 23955933.0,
      "step": 11545
    },
    {
      "entropy": 5.481488513946533,
      "epoch": 1.0477140783744556,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004897232234733887,
      "loss": 5.1493,
      "mean_token_accuracy": 0.18738794922828675,
      "num_tokens": 23966320.0,
      "step": 11550
    },
    {
      "entropy": 5.456253623962402,
      "epoch": 1.04816763425254,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897135541556028,
      "loss": 5.0468,
      "mean_token_accuracy": 0.20325277298688887,
      "num_tokens": 23976233.0,
      "step": 11555
    },
    {
      "entropy": 5.49235692024231,
      "epoch": 1.0486211901306242,
      "grad_norm": 0.84375,
      "learning_rate": 0.000489703880397486,
      "loss": 5.11,
      "mean_token_accuracy": 0.19220408350229262,
      "num_tokens": 23987854.0,
      "step": 11560
    },
    {
      "entropy": 5.457041692733765,
      "epoch": 1.0490747460087082,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004896942021992382,
      "loss": 5.1154,
      "mean_token_accuracy": 0.19077137857675552,
      "num_tokens": 23997973.0,
      "step": 11565
    },
    {
      "entropy": 5.482615661621094,
      "epoch": 1.0495283018867925,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004896845195610595,
      "loss": 5.1169,
      "mean_token_accuracy": 0.1928719624876976,
      "num_tokens": 24007313.0,
      "step": 11570
    },
    {
      "entropy": 5.476922655105591,
      "epoch": 1.0499818577648767,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004896748324831504,
      "loss": 5.1244,
      "mean_token_accuracy": 0.19298154413700103,
      "num_tokens": 24018254.0,
      "step": 11575
    },
    {
      "entropy": 5.492932891845703,
      "epoch": 1.0504354136429608,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004896651409657109,
      "loss": 5.1368,
      "mean_token_accuracy": 0.19103600829839706,
      "num_tokens": 24029799.0,
      "step": 11580
    },
    {
      "entropy": 5.414491605758667,
      "epoch": 1.050888969521045,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004896554450089417,
      "loss": 5.0258,
      "mean_token_accuracy": 0.19909809976816178,
      "num_tokens": 24040972.0,
      "step": 11585
    },
    {
      "entropy": 5.435051107406617,
      "epoch": 1.051342525399129,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896457446130431,
      "loss": 4.958,
      "mean_token_accuracy": 0.2022219255566597,
      "num_tokens": 24050781.0,
      "step": 11590
    },
    {
      "entropy": 5.418264770507813,
      "epoch": 1.0517960812772134,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896360397782159,
      "loss": 5.1011,
      "mean_token_accuracy": 0.1896982803940773,
      "num_tokens": 24060184.0,
      "step": 11595
    },
    {
      "entropy": 5.513707160949707,
      "epoch": 1.0522496371552976,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004896263305046607,
      "loss": 5.1531,
      "mean_token_accuracy": 0.1932757616043091,
      "num_tokens": 24070777.0,
      "step": 11600
    },
    {
      "entropy": 5.481814289093018,
      "epoch": 1.0527031930333817,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896166167925782,
      "loss": 5.1822,
      "mean_token_accuracy": 0.19279300570487976,
      "num_tokens": 24080878.0,
      "step": 11605
    },
    {
      "entropy": 5.519545316696167,
      "epoch": 1.053156748911466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004896068986421693,
      "loss": 5.1407,
      "mean_token_accuracy": 0.19935741722583772,
      "num_tokens": 24090038.0,
      "step": 11610
    },
    {
      "entropy": 5.561168384552002,
      "epoch": 1.05361030478955,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004895971760536351,
      "loss": 5.1541,
      "mean_token_accuracy": 0.19477766305208205,
      "num_tokens": 24100425.0,
      "step": 11615
    },
    {
      "entropy": 5.538584423065186,
      "epoch": 1.0540638606676342,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004895874490271768,
      "loss": 5.2186,
      "mean_token_accuracy": 0.18994971215724946,
      "num_tokens": 24112182.0,
      "step": 11620
    },
    {
      "entropy": 5.481795358657837,
      "epoch": 1.0545174165457185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895777175629952,
      "loss": 5.0765,
      "mean_token_accuracy": 0.1948103278875351,
      "num_tokens": 24122182.0,
      "step": 11625
    },
    {
      "entropy": 5.5433516025543215,
      "epoch": 1.0549709724238026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004895679816612917,
      "loss": 5.0685,
      "mean_token_accuracy": 0.2016040116548538,
      "num_tokens": 24131930.0,
      "step": 11630
    },
    {
      "entropy": 5.5081311702728275,
      "epoch": 1.0554245283018868,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895582413222678,
      "loss": 5.0693,
      "mean_token_accuracy": 0.20441591888666152,
      "num_tokens": 24141575.0,
      "step": 11635
    },
    {
      "entropy": 5.475733232498169,
      "epoch": 1.0558780841799709,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895484965461247,
      "loss": 5.2086,
      "mean_token_accuracy": 0.1897387534379959,
      "num_tokens": 24151720.0,
      "step": 11640
    },
    {
      "entropy": 5.4533342838287355,
      "epoch": 1.0563316400580551,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000489538747333064,
      "loss": 5.089,
      "mean_token_accuracy": 0.19781393110752105,
      "num_tokens": 24161881.0,
      "step": 11645
    },
    {
      "entropy": 5.43628077507019,
      "epoch": 1.0567851959361394,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895289936832874,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20072407871484757,
      "num_tokens": 24171474.0,
      "step": 11650
    },
    {
      "entropy": 5.535496854782105,
      "epoch": 1.0572387518142234,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004895192355969966,
      "loss": 5.1662,
      "mean_token_accuracy": 0.19123655706644058,
      "num_tokens": 24182186.0,
      "step": 11655
    },
    {
      "entropy": 5.506235218048095,
      "epoch": 1.0576923076923077,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004895094730743932,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19827568531036377,
      "num_tokens": 24192318.0,
      "step": 11660
    },
    {
      "entropy": 5.52546124458313,
      "epoch": 1.058145863570392,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004894997061156793,
      "loss": 5.1616,
      "mean_token_accuracy": 0.19303964376449584,
      "num_tokens": 24202918.0,
      "step": 11665
    },
    {
      "entropy": 5.439881277084351,
      "epoch": 1.058599419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894899347210567,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20049711018800737,
      "num_tokens": 24213218.0,
      "step": 11670
    },
    {
      "entropy": 5.467959308624268,
      "epoch": 1.0590529753265603,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004894801588907278,
      "loss": 5.0387,
      "mean_token_accuracy": 0.198946575820446,
      "num_tokens": 24223229.0,
      "step": 11675
    },
    {
      "entropy": 5.52664999961853,
      "epoch": 1.0595065312046443,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004894703786248944,
      "loss": 5.1111,
      "mean_token_accuracy": 0.19159167855978013,
      "num_tokens": 24233637.0,
      "step": 11680
    },
    {
      "entropy": 5.478523540496826,
      "epoch": 1.0599600870827286,
      "grad_norm": 1.0,
      "learning_rate": 0.000489460593923759,
      "loss": 5.1528,
      "mean_token_accuracy": 0.1973455563187599,
      "num_tokens": 24244251.0,
      "step": 11685
    },
    {
      "entropy": 5.44070839881897,
      "epoch": 1.0604136429608129,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004894508047875237,
      "loss": 5.0526,
      "mean_token_accuracy": 0.19362615346908568,
      "num_tokens": 24254380.0,
      "step": 11690
    },
    {
      "entropy": 5.355499649047852,
      "epoch": 1.060867198838897,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004894410112163913,
      "loss": 5.0026,
      "mean_token_accuracy": 0.2004845768213272,
      "num_tokens": 24264841.0,
      "step": 11695
    },
    {
      "entropy": 5.414450836181641,
      "epoch": 1.0613207547169812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004894312132105641,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20084351897239686,
      "num_tokens": 24274383.0,
      "step": 11700
    },
    {
      "entropy": 5.5343756675720215,
      "epoch": 1.0617743105950652,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894214107702447,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19480592012405396,
      "num_tokens": 24284140.0,
      "step": 11705
    },
    {
      "entropy": 5.590111207962036,
      "epoch": 1.0622278664731495,
      "grad_norm": 1.109375,
      "learning_rate": 0.000489411603895636,
      "loss": 5.1771,
      "mean_token_accuracy": 0.19371833950281142,
      "num_tokens": 24295044.0,
      "step": 11710
    },
    {
      "entropy": 5.4418053150177,
      "epoch": 1.0626814223512338,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894017925869406,
      "loss": 5.08,
      "mean_token_accuracy": 0.19503361880779266,
      "num_tokens": 24305080.0,
      "step": 11715
    },
    {
      "entropy": 5.4726255416870115,
      "epoch": 1.0631349782293178,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004893919768443615,
      "loss": 5.1257,
      "mean_token_accuracy": 0.19226745814085006,
      "num_tokens": 24316306.0,
      "step": 11720
    },
    {
      "entropy": 5.4423802375793455,
      "epoch": 1.063588534107402,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004893821566681017,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19782619476318358,
      "num_tokens": 24326740.0,
      "step": 11725
    },
    {
      "entropy": 5.503826522827149,
      "epoch": 1.0640420899854863,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004893723320583644,
      "loss": 5.1457,
      "mean_token_accuracy": 0.18667171746492386,
      "num_tokens": 24337621.0,
      "step": 11730
    },
    {
      "entropy": 5.399421215057373,
      "epoch": 1.0644956458635704,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004893625030153525,
      "loss": 4.9887,
      "mean_token_accuracy": 0.20595765560865403,
      "num_tokens": 24347732.0,
      "step": 11735
    },
    {
      "entropy": 5.4055805683135985,
      "epoch": 1.0649492017416546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893526695392694,
      "loss": 5.0222,
      "mean_token_accuracy": 0.20757206678390502,
      "num_tokens": 24359173.0,
      "step": 11740
    },
    {
      "entropy": 5.422066068649292,
      "epoch": 1.0654027576197387,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893428316303185,
      "loss": 5.0202,
      "mean_token_accuracy": 0.19757631570100784,
      "num_tokens": 24369465.0,
      "step": 11745
    },
    {
      "entropy": 5.412892198562622,
      "epoch": 1.065856313497823,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004893329892887031,
      "loss": 5.0619,
      "mean_token_accuracy": 0.1981126308441162,
      "num_tokens": 24378899.0,
      "step": 11750
    },
    {
      "entropy": 5.484836912155151,
      "epoch": 1.0663098693759072,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004893231425146271,
      "loss": 5.1799,
      "mean_token_accuracy": 0.19192711710929872,
      "num_tokens": 24388919.0,
      "step": 11755
    },
    {
      "entropy": 5.4312244892120365,
      "epoch": 1.0667634252539913,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004893132913082937,
      "loss": 5.0478,
      "mean_token_accuracy": 0.1983172506093979,
      "num_tokens": 24399632.0,
      "step": 11760
    },
    {
      "entropy": 5.488347053527832,
      "epoch": 1.0672169811320755,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004893034356699069,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19779288917779922,
      "num_tokens": 24410817.0,
      "step": 11765
    },
    {
      "entropy": 5.417764663696289,
      "epoch": 1.0676705370101596,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004892935755996705,
      "loss": 5.0904,
      "mean_token_accuracy": 0.1991182193160057,
      "num_tokens": 24421058.0,
      "step": 11770
    },
    {
      "entropy": 5.408098793029785,
      "epoch": 1.0681240928882438,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892837110977883,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2003073051571846,
      "num_tokens": 24431522.0,
      "step": 11775
    },
    {
      "entropy": 5.462185478210449,
      "epoch": 1.068577648766328,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004892738421644644,
      "loss": 5.0522,
      "mean_token_accuracy": 0.19488768875598908,
      "num_tokens": 24441746.0,
      "step": 11780
    },
    {
      "entropy": 5.397897434234619,
      "epoch": 1.0690312046444121,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892639687999027,
      "loss": 4.9586,
      "mean_token_accuracy": 0.20120252817869186,
      "num_tokens": 24451916.0,
      "step": 11785
    },
    {
      "entropy": 5.439264392852783,
      "epoch": 1.0694847605224964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004892540910043076,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20249505788087846,
      "num_tokens": 24461935.0,
      "step": 11790
    },
    {
      "entropy": 5.457518911361694,
      "epoch": 1.0699383164005805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892442087778833,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20152089595794678,
      "num_tokens": 24472437.0,
      "step": 11795
    },
    {
      "entropy": 5.457108974456787,
      "epoch": 1.0703918722786647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004892343221208342,
      "loss": 5.0965,
      "mean_token_accuracy": 0.20119676738977432,
      "num_tokens": 24481726.0,
      "step": 11800
    },
    {
      "entropy": 5.491053771972656,
      "epoch": 1.070845428156749,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004892244310333646,
      "loss": 5.1704,
      "mean_token_accuracy": 0.19343004524707794,
      "num_tokens": 24492340.0,
      "step": 11805
    },
    {
      "entropy": 5.452194881439209,
      "epoch": 1.071298984034833,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004892145355156793,
      "loss": 5.0366,
      "mean_token_accuracy": 0.19774752855300903,
      "num_tokens": 24502138.0,
      "step": 11810
    },
    {
      "entropy": 5.4785271167755125,
      "epoch": 1.0717525399129173,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004892046355679829,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19350825250148773,
      "num_tokens": 24512723.0,
      "step": 11815
    },
    {
      "entropy": 5.504654312133789,
      "epoch": 1.0722060957910013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004891947311904799,
      "loss": 5.1181,
      "mean_token_accuracy": 0.19506742358207702,
      "num_tokens": 24523766.0,
      "step": 11820
    },
    {
      "entropy": 5.500041055679321,
      "epoch": 1.0726596516690856,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004891848223833753,
      "loss": 5.1052,
      "mean_token_accuracy": 0.19698472768068315,
      "num_tokens": 24534571.0,
      "step": 11825
    },
    {
      "entropy": 5.482295513153076,
      "epoch": 1.0731132075471699,
      "grad_norm": 1.03125,
      "learning_rate": 0.000489174909146874,
      "loss": 5.1054,
      "mean_token_accuracy": 0.19994303733110427,
      "num_tokens": 24545265.0,
      "step": 11830
    },
    {
      "entropy": 5.5266765594482425,
      "epoch": 1.073566763425254,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004891649914811811,
      "loss": 5.1877,
      "mean_token_accuracy": 0.18802001029253007,
      "num_tokens": 24555524.0,
      "step": 11835
    },
    {
      "entropy": 5.397583484649658,
      "epoch": 1.0740203193033382,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891550693865014,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21245964467525483,
      "num_tokens": 24565420.0,
      "step": 11840
    },
    {
      "entropy": 5.469696044921875,
      "epoch": 1.0744738751814222,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004891451428630405,
      "loss": 5.0954,
      "mean_token_accuracy": 0.1935744747519493,
      "num_tokens": 24576794.0,
      "step": 11845
    },
    {
      "entropy": 5.500405263900757,
      "epoch": 1.0749274310595065,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004891352119110033,
      "loss": 5.0969,
      "mean_token_accuracy": 0.19519586563110353,
      "num_tokens": 24586320.0,
      "step": 11850
    },
    {
      "entropy": 5.446904039382934,
      "epoch": 1.0753809869375908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004891252765305954,
      "loss": 5.0769,
      "mean_token_accuracy": 0.2048339292407036,
      "num_tokens": 24596524.0,
      "step": 11855
    },
    {
      "entropy": 5.4598876953125,
      "epoch": 1.0758345428156748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891153367220222,
      "loss": 5.016,
      "mean_token_accuracy": 0.20519405454397202,
      "num_tokens": 24605877.0,
      "step": 11860
    },
    {
      "entropy": 5.514345026016235,
      "epoch": 1.076288098693759,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004891053924854894,
      "loss": 5.2141,
      "mean_token_accuracy": 0.1887012705206871,
      "num_tokens": 24616719.0,
      "step": 11865
    },
    {
      "entropy": 5.420893049240112,
      "epoch": 1.0767416545718433,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890954438212024,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20212965607643127,
      "num_tokens": 24626882.0,
      "step": 11870
    },
    {
      "entropy": 5.469028377532959,
      "epoch": 1.0771952104499274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890854907293671,
      "loss": 5.0921,
      "mean_token_accuracy": 0.19691340178251265,
      "num_tokens": 24637884.0,
      "step": 11875
    },
    {
      "entropy": 5.448418807983399,
      "epoch": 1.0776487663280117,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004890755332101893,
      "loss": 5.0798,
      "mean_token_accuracy": 0.1930866315960884,
      "num_tokens": 24647389.0,
      "step": 11880
    },
    {
      "entropy": 5.449471950531006,
      "epoch": 1.0781023222060957,
      "grad_norm": 0.875,
      "learning_rate": 0.0004890655712638749,
      "loss": 5.0972,
      "mean_token_accuracy": 0.19339723885059357,
      "num_tokens": 24657595.0,
      "step": 11885
    },
    {
      "entropy": 5.456168222427368,
      "epoch": 1.07855587808418,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048905560489063,
      "loss": 5.0011,
      "mean_token_accuracy": 0.19946667104959487,
      "num_tokens": 24667683.0,
      "step": 11890
    },
    {
      "entropy": 5.491331672668457,
      "epoch": 1.0790094339622642,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004890456340906605,
      "loss": 5.157,
      "mean_token_accuracy": 0.1912636011838913,
      "num_tokens": 24678041.0,
      "step": 11895
    },
    {
      "entropy": 5.445166206359863,
      "epoch": 1.0794629898403483,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004890356588641729,
      "loss": 5.1521,
      "mean_token_accuracy": 0.20095144957304,
      "num_tokens": 24688569.0,
      "step": 11900
    },
    {
      "entropy": 5.475362777709961,
      "epoch": 1.0799165457184325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004890256792113733,
      "loss": 5.0967,
      "mean_token_accuracy": 0.19915975630283356,
      "num_tokens": 24699913.0,
      "step": 11905
    },
    {
      "entropy": 5.547744035720825,
      "epoch": 1.0803701015965168,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004890156951324681,
      "loss": 5.1223,
      "mean_token_accuracy": 0.19804803878068925,
      "num_tokens": 24710682.0,
      "step": 11910
    },
    {
      "entropy": 5.558585834503174,
      "epoch": 1.0808236574746009,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004890057066276636,
      "loss": 5.1974,
      "mean_token_accuracy": 0.18895184248685837,
      "num_tokens": 24721087.0,
      "step": 11915
    },
    {
      "entropy": 5.424392366409302,
      "epoch": 1.0812772133526851,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004889957136971668,
      "loss": 5.015,
      "mean_token_accuracy": 0.20212464183568954,
      "num_tokens": 24731323.0,
      "step": 11920
    },
    {
      "entropy": 5.424842023849488,
      "epoch": 1.0817307692307692,
      "grad_norm": 1.0,
      "learning_rate": 0.000488985716341184,
      "loss": 5.1052,
      "mean_token_accuracy": 0.1901997834444046,
      "num_tokens": 24742559.0,
      "step": 11925
    },
    {
      "entropy": 5.492843246459961,
      "epoch": 1.0821843251088534,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004889757145599222,
      "loss": 5.0918,
      "mean_token_accuracy": 0.1986592561006546,
      "num_tokens": 24752307.0,
      "step": 11930
    },
    {
      "entropy": 5.383856630325317,
      "epoch": 1.0826378809869377,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000488965708353588,
      "loss": 5.0516,
      "mean_token_accuracy": 0.19449937641620635,
      "num_tokens": 24762792.0,
      "step": 11935
    },
    {
      "entropy": 5.493750333786011,
      "epoch": 1.0830914368650217,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004889556977223883,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19200471192598342,
      "num_tokens": 24774377.0,
      "step": 11940
    },
    {
      "entropy": 5.503545761108398,
      "epoch": 1.083544992743106,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004889456826665304,
      "loss": 5.078,
      "mean_token_accuracy": 0.19847277849912642,
      "num_tokens": 24785919.0,
      "step": 11945
    },
    {
      "entropy": 5.51596851348877,
      "epoch": 1.08399854862119,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004889356631862211,
      "loss": 5.115,
      "mean_token_accuracy": 0.18979937434196473,
      "num_tokens": 24796594.0,
      "step": 11950
    },
    {
      "entropy": 5.460939598083496,
      "epoch": 1.0844521044992743,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000488925639281668,
      "loss": 5.1161,
      "mean_token_accuracy": 0.1949406310915947,
      "num_tokens": 24807322.0,
      "step": 11955
    },
    {
      "entropy": 5.417977905273437,
      "epoch": 1.0849056603773586,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004889156109530781,
      "loss": 4.9998,
      "mean_token_accuracy": 0.19890448153018953,
      "num_tokens": 24817593.0,
      "step": 11960
    },
    {
      "entropy": 5.4494860649108885,
      "epoch": 1.0853592162554426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889055782006588,
      "loss": 5.0115,
      "mean_token_accuracy": 0.20230089426040648,
      "num_tokens": 24827577.0,
      "step": 11965
    },
    {
      "entropy": 5.441661405563354,
      "epoch": 1.085812772133527,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004888955410246176,
      "loss": 5.0568,
      "mean_token_accuracy": 0.19938754588365554,
      "num_tokens": 24838148.0,
      "step": 11970
    },
    {
      "entropy": 5.443475580215454,
      "epoch": 1.086266328011611,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004888854994251621,
      "loss": 5.0108,
      "mean_token_accuracy": 0.19843291789293288,
      "num_tokens": 24848451.0,
      "step": 11975
    },
    {
      "entropy": 5.492741966247559,
      "epoch": 1.0867198838896952,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004888754534025,
      "loss": 5.1404,
      "mean_token_accuracy": 0.18598994314670564,
      "num_tokens": 24859619.0,
      "step": 11980
    },
    {
      "entropy": 5.499451589584351,
      "epoch": 1.0871734397677795,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888654029568391,
      "loss": 5.1711,
      "mean_token_accuracy": 0.18404650390148164,
      "num_tokens": 24869863.0,
      "step": 11985
    },
    {
      "entropy": 5.457519960403443,
      "epoch": 1.0876269956458635,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004888553480883871,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1899152174592018,
      "num_tokens": 24880524.0,
      "step": 11990
    },
    {
      "entropy": 5.415886926651001,
      "epoch": 1.0880805515239478,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000488845288797352,
      "loss": 5.0696,
      "mean_token_accuracy": 0.19850811511278152,
      "num_tokens": 24891514.0,
      "step": 11995
    },
    {
      "entropy": 5.4387938499450685,
      "epoch": 1.0885341074020318,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888352250839418,
      "loss": 5.0251,
      "mean_token_accuracy": 0.19936315417289735,
      "num_tokens": 24900842.0,
      "step": 12000
    },
    {
      "epoch": 1.0885341074020318,
      "eval_entropy": 5.235634247102349,
      "eval_loss": 5.16903829574585,
      "eval_mean_token_accuracy": 0.20154844025714636,
      "eval_num_tokens": 24900842.0,
      "eval_runtime": 20.6538,
      "eval_samples_per_second": 1712.03,
      "eval_steps_per_second": 214.004,
      "step": 12000
    },
    {
      "entropy": 5.4245994091033936,
      "epoch": 1.088987663280116,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004888251569483647,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20064777582883836,
      "num_tokens": 24911845.0,
      "step": 12005
    },
    {
      "entropy": 5.382421827316284,
      "epoch": 1.0894412191582004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004888150843908288,
      "loss": 5.0402,
      "mean_token_accuracy": 0.20876970440149306,
      "num_tokens": 24922211.0,
      "step": 12010
    },
    {
      "entropy": 5.518812799453736,
      "epoch": 1.0898947750362844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888050074115426,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1951376900076866,
      "num_tokens": 24932129.0,
      "step": 12015
    },
    {
      "entropy": 5.550019693374634,
      "epoch": 1.0903483309143687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004887949260107143,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19230424612760544,
      "num_tokens": 24942869.0,
      "step": 12020
    },
    {
      "entropy": 5.4164728164672855,
      "epoch": 1.0908018867924527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887848401885524,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19477627873420716,
      "num_tokens": 24952621.0,
      "step": 12025
    },
    {
      "entropy": 5.556595325469971,
      "epoch": 1.091255442670537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004887747499452656,
      "loss": 5.195,
      "mean_token_accuracy": 0.1833399146795273,
      "num_tokens": 24963587.0,
      "step": 12030
    },
    {
      "entropy": 5.515872573852539,
      "epoch": 1.0917089985486212,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004887646552810624,
      "loss": 5.0398,
      "mean_token_accuracy": 0.2013555198907852,
      "num_tokens": 24974122.0,
      "step": 12035
    },
    {
      "entropy": 5.483128356933594,
      "epoch": 1.0921625544267053,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004887545561961517,
      "loss": 5.1285,
      "mean_token_accuracy": 0.1937446415424347,
      "num_tokens": 24985145.0,
      "step": 12040
    },
    {
      "entropy": 5.429468202590942,
      "epoch": 1.0926161103047896,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004887444526907424,
      "loss": 5.0495,
      "mean_token_accuracy": 0.1973423257470131,
      "num_tokens": 24995511.0,
      "step": 12045
    },
    {
      "entropy": 5.477809238433838,
      "epoch": 1.0930696661828738,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004887343447650432,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19301147162914276,
      "num_tokens": 25006193.0,
      "step": 12050
    },
    {
      "entropy": 5.505136442184448,
      "epoch": 1.0935232220609579,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004887242324192634,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19657760858535767,
      "num_tokens": 25016444.0,
      "step": 12055
    },
    {
      "entropy": 5.528617286682129,
      "epoch": 1.0939767779390421,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004887141156536119,
      "loss": 5.0888,
      "mean_token_accuracy": 0.19392160177230836,
      "num_tokens": 25027898.0,
      "step": 12060
    },
    {
      "entropy": 5.447785425186157,
      "epoch": 1.0944303338171262,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004887039944682981,
      "loss": 4.9868,
      "mean_token_accuracy": 0.20039299726486207,
      "num_tokens": 25038201.0,
      "step": 12065
    },
    {
      "entropy": 5.549479961395264,
      "epoch": 1.0948838896952104,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886938688635313,
      "loss": 5.1931,
      "mean_token_accuracy": 0.1823815271258354,
      "num_tokens": 25048466.0,
      "step": 12070
    },
    {
      "entropy": 5.569545698165894,
      "epoch": 1.0953374455732947,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004886837388395207,
      "loss": 5.1347,
      "mean_token_accuracy": 0.19485244303941726,
      "num_tokens": 25059064.0,
      "step": 12075
    },
    {
      "entropy": 5.543216323852539,
      "epoch": 1.0957910014513788,
      "grad_norm": 1.078125,
      "learning_rate": 0.000488673604396476,
      "loss": 5.2136,
      "mean_token_accuracy": 0.18050719946622848,
      "num_tokens": 25068834.0,
      "step": 12080
    },
    {
      "entropy": 5.506651210784912,
      "epoch": 1.096244557329463,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886634655346067,
      "loss": 5.155,
      "mean_token_accuracy": 0.18910235464572905,
      "num_tokens": 25080058.0,
      "step": 12085
    },
    {
      "entropy": 5.440000104904175,
      "epoch": 1.0966981132075473,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886533222541225,
      "loss": 5.0182,
      "mean_token_accuracy": 0.20164837390184404,
      "num_tokens": 25089751.0,
      "step": 12090
    },
    {
      "entropy": 5.478103876113892,
      "epoch": 1.0971516690856313,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886431745552331,
      "loss": 5.1127,
      "mean_token_accuracy": 0.19630647152662278,
      "num_tokens": 25099055.0,
      "step": 12095
    },
    {
      "entropy": 5.54117488861084,
      "epoch": 1.0976052249637156,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004886330224381485,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18562830686569215,
      "num_tokens": 25109793.0,
      "step": 12100
    },
    {
      "entropy": 5.521268510818482,
      "epoch": 1.0980587808417996,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004886228659030785,
      "loss": 5.1163,
      "mean_token_accuracy": 0.19521568566560746,
      "num_tokens": 25119814.0,
      "step": 12105
    },
    {
      "entropy": 5.5408790588378904,
      "epoch": 1.098512336719884,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004886127049502331,
      "loss": 5.1591,
      "mean_token_accuracy": 0.19464686661958694,
      "num_tokens": 25130747.0,
      "step": 12110
    },
    {
      "entropy": 5.454093742370605,
      "epoch": 1.0989658925979682,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886025395798228,
      "loss": 5.0522,
      "mean_token_accuracy": 0.1918802246451378,
      "num_tokens": 25142235.0,
      "step": 12115
    },
    {
      "entropy": 5.366468334197998,
      "epoch": 1.0994194484760522,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004885923697920573,
      "loss": 4.9928,
      "mean_token_accuracy": 0.205183207988739,
      "num_tokens": 25152047.0,
      "step": 12120
    },
    {
      "entropy": 5.431280422210693,
      "epoch": 1.0998730043541365,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004885821955871472,
      "loss": 5.0471,
      "mean_token_accuracy": 0.19808025658130646,
      "num_tokens": 25162678.0,
      "step": 12125
    },
    {
      "entropy": 5.49901065826416,
      "epoch": 1.1003265602322205,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885720169653028,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19658662378787994,
      "num_tokens": 25172973.0,
      "step": 12130
    },
    {
      "entropy": 5.507623529434204,
      "epoch": 1.1007801161103048,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004885618339267348,
      "loss": 5.1716,
      "mean_token_accuracy": 0.1899319440126419,
      "num_tokens": 25183458.0,
      "step": 12135
    },
    {
      "entropy": 5.445809078216553,
      "epoch": 1.101233671988389,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885516464716537,
      "loss": 5.0571,
      "mean_token_accuracy": 0.19627972692251205,
      "num_tokens": 25193827.0,
      "step": 12140
    },
    {
      "entropy": 5.5591919898986815,
      "epoch": 1.101687227866473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048854145460027,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19065470844507218,
      "num_tokens": 25204448.0,
      "step": 12145
    },
    {
      "entropy": 5.519454479217529,
      "epoch": 1.1021407837445574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885312583127946,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20112146586179733,
      "num_tokens": 25215041.0,
      "step": 12150
    },
    {
      "entropy": 5.444091129302978,
      "epoch": 1.1025943396226414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004885210576094385,
      "loss": 5.0821,
      "mean_token_accuracy": 0.1961773082613945,
      "num_tokens": 25225937.0,
      "step": 12155
    },
    {
      "entropy": 5.418184804916382,
      "epoch": 1.1030478955007257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885108524904124,
      "loss": 4.995,
      "mean_token_accuracy": 0.1989628866314888,
      "num_tokens": 25235270.0,
      "step": 12160
    },
    {
      "entropy": 5.467058181762695,
      "epoch": 1.10350145137881,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004885006429559275,
      "loss": 5.101,
      "mean_token_accuracy": 0.18867916166782378,
      "num_tokens": 25247080.0,
      "step": 12165
    },
    {
      "entropy": 5.534938716888428,
      "epoch": 1.103955007256894,
      "grad_norm": 1.125,
      "learning_rate": 0.0004884904290061949,
      "loss": 5.0948,
      "mean_token_accuracy": 0.18896181732416154,
      "num_tokens": 25256718.0,
      "step": 12170
    },
    {
      "entropy": 5.4963476181030275,
      "epoch": 1.1044085631349783,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004884802106414258,
      "loss": 5.2132,
      "mean_token_accuracy": 0.1859460324048996,
      "num_tokens": 25266632.0,
      "step": 12175
    },
    {
      "entropy": 5.516941118240356,
      "epoch": 1.1048621190130623,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004884699878618316,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19886712729930878,
      "num_tokens": 25275598.0,
      "step": 12180
    },
    {
      "entropy": 5.540178489685059,
      "epoch": 1.1053156748911466,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004884597606676236,
      "loss": 5.1522,
      "mean_token_accuracy": 0.1940113976597786,
      "num_tokens": 25286598.0,
      "step": 12185
    },
    {
      "entropy": 5.459210348129273,
      "epoch": 1.1057692307692308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884495290590134,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19530683755874634,
      "num_tokens": 25296730.0,
      "step": 12190
    },
    {
      "entropy": 5.447724723815918,
      "epoch": 1.1062227866473149,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004884392930362124,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20545725524425507,
      "num_tokens": 25306326.0,
      "step": 12195
    },
    {
      "entropy": 5.494918966293335,
      "epoch": 1.1066763425253991,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004884290525994326,
      "loss": 5.0998,
      "mean_token_accuracy": 0.19826816022396088,
      "num_tokens": 25316342.0,
      "step": 12200
    },
    {
      "entropy": 5.579006290435791,
      "epoch": 1.1071298984034832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004884188077488855,
      "loss": 5.1396,
      "mean_token_accuracy": 0.1880671426653862,
      "num_tokens": 25326129.0,
      "step": 12205
    },
    {
      "entropy": 5.491881370544434,
      "epoch": 1.1075834542815675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884085584847831,
      "loss": 5.1088,
      "mean_token_accuracy": 0.18871529400348663,
      "num_tokens": 25337156.0,
      "step": 12210
    },
    {
      "entropy": 5.553006219863891,
      "epoch": 1.1080370101596517,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004883983048073375,
      "loss": 5.2199,
      "mean_token_accuracy": 0.18386245816946029,
      "num_tokens": 25347632.0,
      "step": 12215
    },
    {
      "entropy": 5.458075666427613,
      "epoch": 1.1084905660377358,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004883880467167603,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20122691094875336,
      "num_tokens": 25359186.0,
      "step": 12220
    },
    {
      "entropy": 5.496602344512939,
      "epoch": 1.10894412191582,
      "grad_norm": 1.15625,
      "learning_rate": 0.000488377784213264,
      "loss": 5.0975,
      "mean_token_accuracy": 0.1946451857686043,
      "num_tokens": 25368993.0,
      "step": 12225
    },
    {
      "entropy": 5.43999285697937,
      "epoch": 1.1093976777939043,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004883675172970608,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19383164942264558,
      "num_tokens": 25379880.0,
      "step": 12230
    },
    {
      "entropy": 5.533678007125855,
      "epoch": 1.1098512336719883,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488357245968363,
      "loss": 5.2312,
      "mean_token_accuracy": 0.18909236490726472,
      "num_tokens": 25389942.0,
      "step": 12235
    },
    {
      "entropy": 5.496716022491455,
      "epoch": 1.1103047895500726,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004883469702273829,
      "loss": 5.045,
      "mean_token_accuracy": 0.19642258435487747,
      "num_tokens": 25400836.0,
      "step": 12240
    },
    {
      "entropy": 5.479660177230835,
      "epoch": 1.1107583454281567,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004883366900743331,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20178468227386476,
      "num_tokens": 25410502.0,
      "step": 12245
    },
    {
      "entropy": 5.4430962085723875,
      "epoch": 1.111211901306241,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004883264055094262,
      "loss": 5.0206,
      "mean_token_accuracy": 0.20242179334163665,
      "num_tokens": 25420301.0,
      "step": 12250
    },
    {
      "entropy": 5.464849042892456,
      "epoch": 1.1116654571843252,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004883161165328749,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19221176207065582,
      "num_tokens": 25431107.0,
      "step": 12255
    },
    {
      "entropy": 5.491602897644043,
      "epoch": 1.1121190130624092,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000488305823144892,
      "loss": 5.0869,
      "mean_token_accuracy": 0.1981050953269005,
      "num_tokens": 25441256.0,
      "step": 12260
    },
    {
      "entropy": 5.526266479492188,
      "epoch": 1.1125725689404935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004882955253456902,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20367671996355058,
      "num_tokens": 25451627.0,
      "step": 12265
    },
    {
      "entropy": 5.511344575881958,
      "epoch": 1.1130261248185778,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004882852231354826,
      "loss": 5.1622,
      "mean_token_accuracy": 0.18984195441007615,
      "num_tokens": 25461273.0,
      "step": 12270
    },
    {
      "entropy": 5.3871073722839355,
      "epoch": 1.1134796806966618,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004882749165144822,
      "loss": 5.0437,
      "mean_token_accuracy": 0.2021900862455368,
      "num_tokens": 25471949.0,
      "step": 12275
    },
    {
      "entropy": 5.513221025466919,
      "epoch": 1.113933236574746,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004882646054829023,
      "loss": 5.0712,
      "mean_token_accuracy": 0.19517802745103835,
      "num_tokens": 25482175.0,
      "step": 12280
    },
    {
      "entropy": 5.539937973022461,
      "epoch": 1.1143867924528301,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048825429004095597,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1953285738825798,
      "num_tokens": 25491786.0,
      "step": 12285
    },
    {
      "entropy": 5.4934309959411625,
      "epoch": 1.1148403483309144,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004882439701888566,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2028183802962303,
      "num_tokens": 25502103.0,
      "step": 12290
    },
    {
      "entropy": 5.489936685562133,
      "epoch": 1.1152939042089987,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048823364592681754,
      "loss": 5.0345,
      "mean_token_accuracy": 0.20078462064266206,
      "num_tokens": 25512530.0,
      "step": 12295
    },
    {
      "entropy": 5.453662967681884,
      "epoch": 1.1157474600870827,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048822331725505234,
      "loss": 5.0326,
      "mean_token_accuracy": 0.1963891550898552,
      "num_tokens": 25522905.0,
      "step": 12300
    },
    {
      "entropy": 5.455841970443726,
      "epoch": 1.116201015965167,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048821298417377456,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2051209181547165,
      "num_tokens": 25532963.0,
      "step": 12305
    },
    {
      "entropy": 5.400241136550903,
      "epoch": 1.116654571843251,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000488202646683198,
      "loss": 5.0859,
      "mean_token_accuracy": 0.19909514039754866,
      "num_tokens": 25542583.0,
      "step": 12310
    },
    {
      "entropy": 5.4752288341522215,
      "epoch": 1.1171081277213353,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004881923047835364,
      "loss": 5.0598,
      "mean_token_accuracy": 0.19813199937343598,
      "num_tokens": 25553367.0,
      "step": 12315
    },
    {
      "entropy": 5.614972066879273,
      "epoch": 1.1175616835994195,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004881819584750036,
      "loss": 5.158,
      "mean_token_accuracy": 0.18798512667417527,
      "num_tokens": 25564823.0,
      "step": 12320
    },
    {
      "entropy": 5.5026294708251955,
      "epoch": 1.1180152394775036,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048817160775781356,
      "loss": 5.1412,
      "mean_token_accuracy": 0.19281987398862838,
      "num_tokens": 25574946.0,
      "step": 12325
    },
    {
      "entropy": 5.5025474548339846,
      "epoch": 1.1184687953555879,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048816125263218023,
      "loss": 5.1247,
      "mean_token_accuracy": 0.19272203445434571,
      "num_tokens": 25584779.0,
      "step": 12330
    },
    {
      "entropy": 5.55981478691101,
      "epoch": 1.118922351233672,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048815089309831803,
      "loss": 5.1748,
      "mean_token_accuracy": 0.1895157814025879,
      "num_tokens": 25594725.0,
      "step": 12335
    },
    {
      "entropy": 5.490996026992798,
      "epoch": 1.1193759071117562,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488140529156441,
      "loss": 5.1544,
      "mean_token_accuracy": 0.19034090340137483,
      "num_tokens": 25605499.0,
      "step": 12340
    },
    {
      "entropy": 5.4678388118743895,
      "epoch": 1.1198294629898404,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004881301608067635,
      "loss": 5.099,
      "mean_token_accuracy": 0.19867289513349534,
      "num_tokens": 25616443.0,
      "step": 12345
    },
    {
      "entropy": 5.508192157745361,
      "epoch": 1.1202830188679245,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048811978804949997,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19435406923294068,
      "num_tokens": 25627762.0,
      "step": 12350
    },
    {
      "entropy": 5.572100734710693,
      "epoch": 1.1207365747460087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000488109410884865,
      "loss": 5.1585,
      "mean_token_accuracy": 0.19211570024490357,
      "num_tokens": 25637904.0,
      "step": 12355
    },
    {
      "entropy": 5.481449365615845,
      "epoch": 1.1211901306240928,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048809902931307294,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20028771758079528,
      "num_tokens": 25649058.0,
      "step": 12360
    },
    {
      "entropy": 5.452664804458618,
      "epoch": 1.121643686502177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004880886433343388,
      "loss": 5.0991,
      "mean_token_accuracy": 0.19796349555253984,
      "num_tokens": 25659161.0,
      "step": 12365
    },
    {
      "entropy": 5.456683731079101,
      "epoch": 1.1220972423802613,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004880782529488771,
      "loss": 4.9979,
      "mean_token_accuracy": 0.20418920665979384,
      "num_tokens": 25669336.0,
      "step": 12370
    },
    {
      "entropy": 5.531540966033935,
      "epoch": 1.1225507982583454,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004880678581569029,
      "loss": 5.1348,
      "mean_token_accuracy": 0.19526318609714508,
      "num_tokens": 25680443.0,
      "step": 12375
    },
    {
      "entropy": 5.457948493957519,
      "epoch": 1.1230043541364296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048805745895863105,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1991736814379692,
      "num_tokens": 25690202.0,
      "step": 12380
    },
    {
      "entropy": 5.457512521743775,
      "epoch": 1.1234579100145137,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004880470553542767,
      "loss": 5.1085,
      "mean_token_accuracy": 0.1963775172829628,
      "num_tokens": 25699255.0,
      "step": 12385
    },
    {
      "entropy": 5.506438302993774,
      "epoch": 1.123911465892598,
      "grad_norm": 0.875,
      "learning_rate": 0.00048803664734405495,
      "loss": 5.0875,
      "mean_token_accuracy": 0.18788405805826186,
      "num_tokens": 25710714.0,
      "step": 12390
    },
    {
      "entropy": 5.501832008361816,
      "epoch": 1.1243650217706822,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488026234928181,
      "loss": 5.0692,
      "mean_token_accuracy": 0.1955667182803154,
      "num_tokens": 25721722.0,
      "step": 12395
    },
    {
      "entropy": 5.461233329772949,
      "epoch": 1.1248185776487662,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004880158181068703,
      "loss": 5.0539,
      "mean_token_accuracy": 0.20558907687664033,
      "num_tokens": 25732096.0,
      "step": 12400
    },
    {
      "entropy": 5.483524560928345,
      "epoch": 1.1252721335268505,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048800539688033807,
      "loss": 5.1225,
      "mean_token_accuracy": 0.19797305017709732,
      "num_tokens": 25741499.0,
      "step": 12405
    },
    {
      "entropy": 5.534729719161987,
      "epoch": 1.1257256894049348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004879949712488001,
      "loss": 5.1901,
      "mean_token_accuracy": 0.1946818009018898,
      "num_tokens": 25751913.0,
      "step": 12410
    },
    {
      "entropy": 5.53972749710083,
      "epoch": 1.1261792452830188,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048798454121247177,
      "loss": 5.131,
      "mean_token_accuracy": 0.19206802695989608,
      "num_tokens": 25762602.0,
      "step": 12415
    },
    {
      "entropy": 5.440932893753052,
      "epoch": 1.126632801161103,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004879741067715688,
      "loss": 5.0934,
      "mean_token_accuracy": 0.19573008865118027,
      "num_tokens": 25773872.0,
      "step": 12420
    },
    {
      "entropy": 5.512517929077148,
      "epoch": 1.1270863570391871,
      "grad_norm": 1.0,
      "learning_rate": 0.00048796366792630715,
      "loss": 5.0935,
      "mean_token_accuracy": 0.19780833125114441,
      "num_tokens": 25784723.0,
      "step": 12425
    },
    {
      "entropy": 5.485009574890137,
      "epoch": 1.1275399129172714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004879532246769025,
      "loss": 5.1044,
      "mean_token_accuracy": 0.19494618028402327,
      "num_tokens": 25794897.0,
      "step": 12430
    },
    {
      "entropy": 5.4504047393798825,
      "epoch": 1.1279934687953557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004879427770235709,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20593176931142806,
      "num_tokens": 25804445.0,
      "step": 12435
    },
    {
      "entropy": 5.5492393493652346,
      "epoch": 1.1284470246734397,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048793232496652846,
      "loss": 5.1673,
      "mean_token_accuracy": 0.1864362522959709,
      "num_tokens": 25815495.0,
      "step": 12440
    },
    {
      "entropy": 5.485481834411621,
      "epoch": 1.128900580551524,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048792186850599124,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20277364999055864,
      "num_tokens": 25826120.0,
      "step": 12445
    },
    {
      "entropy": 5.441395616531372,
      "epoch": 1.1293541364296082,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004879114076421755,
      "loss": 5.0743,
      "mean_token_accuracy": 0.2006630629301071,
      "num_tokens": 25835926.0,
      "step": 12450
    },
    {
      "entropy": 5.51393871307373,
      "epoch": 1.1298076923076923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048790094237529763,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19224480241537095,
      "num_tokens": 25846095.0,
      "step": 12455
    },
    {
      "entropy": 5.568841123580933,
      "epoch": 1.1302612481857766,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487890472705574,
      "loss": 5.1286,
      "mean_token_accuracy": 0.19136714339256286,
      "num_tokens": 25858541.0,
      "step": 12460
    },
    {
      "entropy": 5.4886997699737545,
      "epoch": 1.1307148040638606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048787999863322113,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2051388591527939,
      "num_tokens": 25867664.0,
      "step": 12465
    },
    {
      "entropy": 5.474790382385254,
      "epoch": 1.1311683599419449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004878695201584557,
      "loss": 5.1188,
      "mean_token_accuracy": 0.19833892285823823,
      "num_tokens": 25877300.0,
      "step": 12470
    },
    {
      "entropy": 5.444518423080444,
      "epoch": 1.1316219158200291,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004878590372814943,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20307615250349045,
      "num_tokens": 25888009.0,
      "step": 12475
    },
    {
      "entropy": 5.532946348190308,
      "epoch": 1.1320754716981132,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048784855000255374,
      "loss": 5.1264,
      "mean_token_accuracy": 0.18937529623508453,
      "num_tokens": 25897285.0,
      "step": 12480
    },
    {
      "entropy": 5.478674602508545,
      "epoch": 1.1325290275761974,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048783805832185097,
      "loss": 5.0157,
      "mean_token_accuracy": 0.201152403652668,
      "num_tokens": 25907355.0,
      "step": 12485
    },
    {
      "entropy": 5.494596529006958,
      "epoch": 1.1329825834542815,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004878275622396029,
      "loss": 5.0919,
      "mean_token_accuracy": 0.20091527253389357,
      "num_tokens": 25918566.0,
      "step": 12490
    },
    {
      "entropy": 5.486082220077515,
      "epoch": 1.1334361393323658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048781706175602665,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19996363818645477,
      "num_tokens": 25929145.0,
      "step": 12495
    },
    {
      "entropy": 5.460976219177246,
      "epoch": 1.13388969521045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048780655687133927,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20151668787002563,
      "num_tokens": 25938163.0,
      "step": 12500
    },
    {
      "entropy": 5.457347440719604,
      "epoch": 1.134343251088534,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048779604758575806,
      "loss": 5.046,
      "mean_token_accuracy": 0.19633741229772567,
      "num_tokens": 25948962.0,
      "step": 12505
    },
    {
      "entropy": 5.419883298873901,
      "epoch": 1.1347968069666183,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004877855338995004,
      "loss": 5.033,
      "mean_token_accuracy": 0.19193218201398848,
      "num_tokens": 25959653.0,
      "step": 12510
    },
    {
      "entropy": 5.5417643070220945,
      "epoch": 1.1352503628447024,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048777501581278367,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19666285663843155,
      "num_tokens": 25970268.0,
      "step": 12515
    },
    {
      "entropy": 5.472268581390381,
      "epoch": 1.1357039187227866,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004877644933258254,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1935422420501709,
      "num_tokens": 25980321.0,
      "step": 12520
    },
    {
      "entropy": 5.421860313415527,
      "epoch": 1.136157474600871,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004877539664388432,
      "loss": 5.0697,
      "mean_token_accuracy": 0.19183527082204818,
      "num_tokens": 25992466.0,
      "step": 12525
    },
    {
      "entropy": 5.475657939910889,
      "epoch": 1.136611030478955,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048774343515205464,
      "loss": 5.0452,
      "mean_token_accuracy": 0.2010459080338478,
      "num_tokens": 26003553.0,
      "step": 12530
    },
    {
      "entropy": 5.488823318481446,
      "epoch": 1.1370645863570392,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004877328994656778,
      "loss": 5.0703,
      "mean_token_accuracy": 0.19986444860696792,
      "num_tokens": 26013299.0,
      "step": 12535
    },
    {
      "entropy": 5.454457712173462,
      "epoch": 1.1375181422351233,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004877223593799303,
      "loss": 5.0957,
      "mean_token_accuracy": 0.19835387021303177,
      "num_tokens": 26022913.0,
      "step": 12540
    },
    {
      "entropy": 5.421596193313599,
      "epoch": 1.1379716981132075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004877118148950302,
      "loss": 5.0688,
      "mean_token_accuracy": 0.19932219684123992,
      "num_tokens": 26033191.0,
      "step": 12545
    },
    {
      "entropy": 5.6301860332489015,
      "epoch": 1.1384252539912918,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048770126601119555,
      "loss": 5.2148,
      "mean_token_accuracy": 0.18730915635824202,
      "num_tokens": 26043517.0,
      "step": 12550
    },
    {
      "entropy": 5.577491188049317,
      "epoch": 1.1388788098693758,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048769071272864454,
      "loss": 5.1624,
      "mean_token_accuracy": 0.19309705942869188,
      "num_tokens": 26054173.0,
      "step": 12555
    },
    {
      "entropy": 5.546094846725464,
      "epoch": 1.13933236574746,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048768015504759536,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1877343937754631,
      "num_tokens": 26064929.0,
      "step": 12560
    },
    {
      "entropy": 5.445685005187988,
      "epoch": 1.1397859216255442,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004876695929682665,
      "loss": 5.0438,
      "mean_token_accuracy": 0.200204998254776,
      "num_tokens": 26074872.0,
      "step": 12565
    },
    {
      "entropy": 5.451004409790039,
      "epoch": 1.1402394775036284,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048765902649087605,
      "loss": 4.9624,
      "mean_token_accuracy": 0.20139967799186706,
      "num_tokens": 26086328.0,
      "step": 12570
    },
    {
      "entropy": 5.524568271636963,
      "epoch": 1.1406930333817127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048764845561564285,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19644412994384766,
      "num_tokens": 26096477.0,
      "step": 12575
    },
    {
      "entropy": 5.468381452560425,
      "epoch": 1.1411465892597967,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048763788034278545,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20553031712770461,
      "num_tokens": 26106183.0,
      "step": 12580
    },
    {
      "entropy": 5.454225015640259,
      "epoch": 1.141600145137881,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004876273006725225,
      "loss": 5.084,
      "mean_token_accuracy": 0.1978095680475235,
      "num_tokens": 26117932.0,
      "step": 12585
    },
    {
      "entropy": 5.446175527572632,
      "epoch": 1.1420537010159653,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048761671660507266,
      "loss": 5.0421,
      "mean_token_accuracy": 0.20245159715414046,
      "num_tokens": 26128616.0,
      "step": 12590
    },
    {
      "entropy": 5.502018308639526,
      "epoch": 1.1425072568940493,
      "grad_norm": 0.953125,
      "learning_rate": 0.000487606128140655,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19867226481437683,
      "num_tokens": 26138758.0,
      "step": 12595
    },
    {
      "entropy": 5.4364453792572025,
      "epoch": 1.1429608127721336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004875955352794885,
      "loss": 5.065,
      "mean_token_accuracy": 0.19398841261863708,
      "num_tokens": 26149199.0,
      "step": 12600
    },
    {
      "entropy": 5.430867147445679,
      "epoch": 1.1434143686502176,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004875849380217921,
      "loss": 5.1569,
      "mean_token_accuracy": 0.19733918458223343,
      "num_tokens": 26160276.0,
      "step": 12605
    },
    {
      "entropy": 5.500235271453858,
      "epoch": 1.1438679245283019,
      "grad_norm": 1.046875,
      "learning_rate": 0.000487574336367785,
      "loss": 5.0283,
      "mean_token_accuracy": 0.19995264261960982,
      "num_tokens": 26169564.0,
      "step": 12610
    },
    {
      "entropy": 5.42881498336792,
      "epoch": 1.1443214804063861,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004875637303176865,
      "loss": 5.0007,
      "mean_token_accuracy": 0.20410393923521042,
      "num_tokens": 26179135.0,
      "step": 12615
    },
    {
      "entropy": 5.4993397235870365,
      "epoch": 1.1447750362844702,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048755311987171585,
      "loss": 5.1338,
      "mean_token_accuracy": 0.19475421607494353,
      "num_tokens": 26190116.0,
      "step": 12620
    },
    {
      "entropy": 5.511917066574097,
      "epoch": 1.1452285921625545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048754250503009244,
      "loss": 5.1992,
      "mean_token_accuracy": 0.18957407921552658,
      "num_tokens": 26200847.0,
      "step": 12625
    },
    {
      "entropy": 5.599766254425049,
      "epoch": 1.1456821480406387,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048753188579303596,
      "loss": 5.155,
      "mean_token_accuracy": 0.19580694288015366,
      "num_tokens": 26211151.0,
      "step": 12630
    },
    {
      "entropy": 5.524540090560913,
      "epoch": 1.1461357039187228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004875212621607659,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2049766629934311,
      "num_tokens": 26221273.0,
      "step": 12635
    },
    {
      "entropy": 5.4686830043792725,
      "epoch": 1.146589259796807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048751063413350203,
      "loss": 5.1737,
      "mean_token_accuracy": 0.19148514568805694,
      "num_tokens": 26231770.0,
      "step": 12640
    },
    {
      "entropy": 5.545756721496582,
      "epoch": 1.147042815674891,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000487500001711464,
      "loss": 5.1424,
      "mean_token_accuracy": 0.18901480883359909,
      "num_tokens": 26241386.0,
      "step": 12645
    },
    {
      "entropy": 5.642010736465454,
      "epoch": 1.1474963715529753,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004874893648948719,
      "loss": 5.2114,
      "mean_token_accuracy": 0.19006966054439545,
      "num_tokens": 26251991.0,
      "step": 12650
    },
    {
      "entropy": 5.627796792984009,
      "epoch": 1.1479499274310596,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048747872368394546,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19527844488620758,
      "num_tokens": 26262331.0,
      "step": 12655
    },
    {
      "entropy": 5.561524677276611,
      "epoch": 1.1484034833091437,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048746807807890496,
      "loss": 5.1325,
      "mean_token_accuracy": 0.19464536160230636,
      "num_tokens": 26272939.0,
      "step": 12660
    },
    {
      "entropy": 5.510550403594971,
      "epoch": 1.148857039187228,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048745742807997046,
      "loss": 5.0807,
      "mean_token_accuracy": 0.2004735991358757,
      "num_tokens": 26282849.0,
      "step": 12665
    },
    {
      "entropy": 5.467938804626465,
      "epoch": 1.149310595065312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048744677368736216,
      "loss": 5.0885,
      "mean_token_accuracy": 0.2036068171262741,
      "num_tokens": 26292449.0,
      "step": 12670
    },
    {
      "entropy": 5.4095618724823,
      "epoch": 1.1497641509433962,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048743611490130044,
      "loss": 5.034,
      "mean_token_accuracy": 0.19792603999376296,
      "num_tokens": 26302506.0,
      "step": 12675
    },
    {
      "entropy": 5.50772762298584,
      "epoch": 1.1502177068214805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048742545172200576,
      "loss": 5.1061,
      "mean_token_accuracy": 0.2004696920514107,
      "num_tokens": 26312346.0,
      "step": 12680
    },
    {
      "entropy": 5.542489147186279,
      "epoch": 1.1506712626995645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004874147841496986,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19783084094524384,
      "num_tokens": 26323173.0,
      "step": 12685
    },
    {
      "entropy": 5.547864484786987,
      "epoch": 1.1511248185776488,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004874041121845996,
      "loss": 5.1629,
      "mean_token_accuracy": 0.187813962996006,
      "num_tokens": 26334371.0,
      "step": 12690
    },
    {
      "entropy": 5.475184345245362,
      "epoch": 1.1515783744557329,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004873934358269294,
      "loss": 5.0875,
      "mean_token_accuracy": 0.1992534339427948,
      "num_tokens": 26344506.0,
      "step": 12695
    },
    {
      "entropy": 5.564321422576905,
      "epoch": 1.1520319303338171,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004873827550769088,
      "loss": 5.1226,
      "mean_token_accuracy": 0.1877172037959099,
      "num_tokens": 26354534.0,
      "step": 12700
    },
    {
      "entropy": 5.496835136413575,
      "epoch": 1.1524854862119014,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004873720699347587,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19809603989124297,
      "num_tokens": 26365682.0,
      "step": 12705
    },
    {
      "entropy": 5.570609045028687,
      "epoch": 1.1529390420899854,
      "grad_norm": 1.0,
      "learning_rate": 0.0004873613804007001,
      "loss": 5.1936,
      "mean_token_accuracy": 0.19731561243534088,
      "num_tokens": 26376702.0,
      "step": 12710
    },
    {
      "entropy": 5.4602783679962155,
      "epoch": 1.1533925979680697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004873506864749541,
      "loss": 5.0067,
      "mean_token_accuracy": 0.19946054369211197,
      "num_tokens": 26387154.0,
      "step": 12715
    },
    {
      "entropy": 5.552713298797608,
      "epoch": 1.1538461538461537,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004873399881577417,
      "loss": 5.1788,
      "mean_token_accuracy": 0.19176933765411378,
      "num_tokens": 26397883.0,
      "step": 12720
    },
    {
      "entropy": 5.427668762207031,
      "epoch": 1.154299709724238,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004873292854492842,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20469096153974534,
      "num_tokens": 26407613.0,
      "step": 12725
    },
    {
      "entropy": 5.396756219863891,
      "epoch": 1.1547532656023223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004873185783498031,
      "loss": 5.0023,
      "mean_token_accuracy": 0.20199891477823256,
      "num_tokens": 26417221.0,
      "step": 12730
    },
    {
      "entropy": 5.4925392150878904,
      "epoch": 1.1552068214804063,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004873078668595197,
      "loss": 5.0976,
      "mean_token_accuracy": 0.19460710734128953,
      "num_tokens": 26427976.0,
      "step": 12735
    },
    {
      "entropy": 5.484690761566162,
      "epoch": 1.1556603773584906,
      "grad_norm": 1.0,
      "learning_rate": 0.00048729715097865534,
      "loss": 5.1287,
      "mean_token_accuracy": 0.19827549159526825,
      "num_tokens": 26439213.0,
      "step": 12740
    },
    {
      "entropy": 5.478594827651977,
      "epoch": 1.1561139332365746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004872864307074319,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19928483068943023,
      "num_tokens": 26449763.0,
      "step": 12745
    },
    {
      "entropy": 5.546472692489624,
      "epoch": 1.156567489114659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048727570604607096,
      "loss": 5.1252,
      "mean_token_accuracy": 0.19816723614931106,
      "num_tokens": 26460133.0,
      "step": 12750
    },
    {
      "entropy": 5.5383056640625,
      "epoch": 1.1570210449927432,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048726497699479437,
      "loss": 5.0922,
      "mean_token_accuracy": 0.19691014140844346,
      "num_tokens": 26470164.0,
      "step": 12755
    },
    {
      "entropy": 5.475792503356933,
      "epoch": 1.1574746008708272,
      "grad_norm": 0.921875,
      "learning_rate": 0.000487254243553824,
      "loss": 5.1044,
      "mean_token_accuracy": 0.19779773503541948,
      "num_tokens": 26480314.0,
      "step": 12760
    },
    {
      "entropy": 5.404000902175904,
      "epoch": 1.1579281567489115,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004872435057233817,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20014721006155015,
      "num_tokens": 26491184.0,
      "step": 12765
    },
    {
      "entropy": 5.449006414413452,
      "epoch": 1.1583817126269957,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048723276350368957,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20397381633520126,
      "num_tokens": 26501432.0,
      "step": 12770
    },
    {
      "entropy": 5.49757866859436,
      "epoch": 1.1588352685050798,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048722201689496984,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19721294641494752,
      "num_tokens": 26510379.0,
      "step": 12775
    },
    {
      "entropy": 5.46393346786499,
      "epoch": 1.159288824383164,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004872112658974447,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20006239414215088,
      "num_tokens": 26519454.0,
      "step": 12780
    },
    {
      "entropy": 5.439549779891967,
      "epoch": 1.159742380261248,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048720051051133657,
      "loss": 4.9887,
      "mean_token_accuracy": 0.19962143003940583,
      "num_tokens": 26531325.0,
      "step": 12785
    },
    {
      "entropy": 5.460226392745971,
      "epoch": 1.1601959361393324,
      "grad_norm": 1.0,
      "learning_rate": 0.0004871897507368678,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19462902396917342,
      "num_tokens": 26541510.0,
      "step": 12790
    },
    {
      "entropy": 5.522227096557617,
      "epoch": 1.1606494920174166,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048717898657426093,
      "loss": 5.2614,
      "mean_token_accuracy": 0.18738264143466948,
      "num_tokens": 26553400.0,
      "step": 12795
    },
    {
      "entropy": 5.558783531188965,
      "epoch": 1.1611030478955007,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048716821802373856,
      "loss": 5.0776,
      "mean_token_accuracy": 0.1960080683231354,
      "num_tokens": 26564340.0,
      "step": 12800
    },
    {
      "entropy": 5.512545728683472,
      "epoch": 1.161556603773585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048715744508552334,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19665155112743377,
      "num_tokens": 26573615.0,
      "step": 12805
    },
    {
      "entropy": 5.526737642288208,
      "epoch": 1.1620101596516692,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004871466677598381,
      "loss": 5.198,
      "mean_token_accuracy": 0.19085295498371124,
      "num_tokens": 26584149.0,
      "step": 12810
    },
    {
      "entropy": 5.475385999679565,
      "epoch": 1.1624637155297532,
      "grad_norm": 1.0,
      "learning_rate": 0.0004871358860469057,
      "loss": 5.1147,
      "mean_token_accuracy": 0.18644985556602478,
      "num_tokens": 26595064.0,
      "step": 12815
    },
    {
      "entropy": 5.487485313415528,
      "epoch": 1.1629172714078375,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048712509994694903,
      "loss": 5.1004,
      "mean_token_accuracy": 0.1979370340704918,
      "num_tokens": 26605636.0,
      "step": 12820
    },
    {
      "entropy": 5.573808574676514,
      "epoch": 1.1633708272859216,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004871143094601914,
      "loss": 5.196,
      "mean_token_accuracy": 0.19262153655290604,
      "num_tokens": 26616065.0,
      "step": 12825
    },
    {
      "entropy": 5.570623970031738,
      "epoch": 1.1638243831640058,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004871035145868557,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19735285639762878,
      "num_tokens": 26627246.0,
      "step": 12830
    },
    {
      "entropy": 5.500070810317993,
      "epoch": 1.16427793904209,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048709271532716524,
      "loss": 5.1282,
      "mean_token_accuracy": 0.19226828664541246,
      "num_tokens": 26637630.0,
      "step": 12835
    },
    {
      "entropy": 5.54427752494812,
      "epoch": 1.1647314949201741,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004870819116813434,
      "loss": 5.159,
      "mean_token_accuracy": 0.19256802797317504,
      "num_tokens": 26648683.0,
      "step": 12840
    },
    {
      "entropy": 5.546771287918091,
      "epoch": 1.1651850507982584,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048707110364961363,
      "loss": 5.1174,
      "mean_token_accuracy": 0.195330211520195,
      "num_tokens": 26659573.0,
      "step": 12845
    },
    {
      "entropy": 5.445801401138306,
      "epoch": 1.1656386066763424,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048706029123219937,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2026770994067192,
      "num_tokens": 26669384.0,
      "step": 12850
    },
    {
      "entropy": 5.5239941596984865,
      "epoch": 1.1660921625544267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048704947442932426,
      "loss": 5.195,
      "mean_token_accuracy": 0.1876979649066925,
      "num_tokens": 26679721.0,
      "step": 12855
    },
    {
      "entropy": 5.582709121704101,
      "epoch": 1.166545718432511,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048703865324121196,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20531719774007798,
      "num_tokens": 26689923.0,
      "step": 12860
    },
    {
      "entropy": 5.421995735168457,
      "epoch": 1.166999274310595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048702782766808637,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19754623025655746,
      "num_tokens": 26699953.0,
      "step": 12865
    },
    {
      "entropy": 5.512957429885864,
      "epoch": 1.1674528301886793,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004870169977101711,
      "loss": 5.2723,
      "mean_token_accuracy": 0.18300295621156693,
      "num_tokens": 26712099.0,
      "step": 12870
    },
    {
      "entropy": 5.5708049774169925,
      "epoch": 1.1679063860667633,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048700616336769037,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19754792302846907,
      "num_tokens": 26722334.0,
      "step": 12875
    },
    {
      "entropy": 5.538000631332397,
      "epoch": 1.1683599419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.00048699532464086817,
      "loss": 5.0668,
      "mean_token_accuracy": 0.19858947843313218,
      "num_tokens": 26732470.0,
      "step": 12880
    },
    {
      "entropy": 5.49172830581665,
      "epoch": 1.1688134978229319,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048698448152992865,
      "loss": 5.0952,
      "mean_token_accuracy": 0.19598378092050553,
      "num_tokens": 26742015.0,
      "step": 12885
    },
    {
      "entropy": 5.415465831756592,
      "epoch": 1.169267053701016,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048697363403509605,
      "loss": 5.1135,
      "mean_token_accuracy": 0.19333448261022568,
      "num_tokens": 26753231.0,
      "step": 12890
    },
    {
      "entropy": 5.515175199508667,
      "epoch": 1.1697206095791002,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004869627821565946,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1952577516436577,
      "num_tokens": 26763895.0,
      "step": 12895
    },
    {
      "entropy": 5.568731832504272,
      "epoch": 1.1701741654571842,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004869519258946489,
      "loss": 5.1751,
      "mean_token_accuracy": 0.19621602445840836,
      "num_tokens": 26774208.0,
      "step": 12900
    },
    {
      "entropy": 5.484986734390259,
      "epoch": 1.1706277213352685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048694106524948324,
      "loss": 5.1065,
      "mean_token_accuracy": 0.19364592730998992,
      "num_tokens": 26785554.0,
      "step": 12905
    },
    {
      "entropy": 5.521382761001587,
      "epoch": 1.1710812772133528,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048693020022132243,
      "loss": 5.0204,
      "mean_token_accuracy": 0.1997543454170227,
      "num_tokens": 26795488.0,
      "step": 12910
    },
    {
      "entropy": 5.552805805206299,
      "epoch": 1.1715348330914368,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000486919330810391,
      "loss": 5.1837,
      "mean_token_accuracy": 0.18431039452552794,
      "num_tokens": 26806801.0,
      "step": 12915
    },
    {
      "entropy": 5.510914707183838,
      "epoch": 1.171988388969521,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004869084570169139,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19960006475448608,
      "num_tokens": 26816869.0,
      "step": 12920
    },
    {
      "entropy": 5.481137228012085,
      "epoch": 1.172441944847605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048689757884111585,
      "loss": 5.0411,
      "mean_token_accuracy": 0.1998990297317505,
      "num_tokens": 26827003.0,
      "step": 12925
    },
    {
      "entropy": 5.433103561401367,
      "epoch": 1.1728955007256894,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048688669628322185,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19865425080060958,
      "num_tokens": 26837065.0,
      "step": 12930
    },
    {
      "entropy": 5.410325002670288,
      "epoch": 1.1733490566037736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000486875809343457,
      "loss": 4.9967,
      "mean_token_accuracy": 0.19845315665006638,
      "num_tokens": 26847487.0,
      "step": 12935
    },
    {
      "entropy": 5.475924348831176,
      "epoch": 1.1738026124818577,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048686491802204645,
      "loss": 5.0499,
      "mean_token_accuracy": 0.2035646066069603,
      "num_tokens": 26858407.0,
      "step": 12940
    },
    {
      "entropy": 5.54250168800354,
      "epoch": 1.174256168359942,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004868540223192154,
      "loss": 5.1214,
      "mean_token_accuracy": 0.19381057024002074,
      "num_tokens": 26868469.0,
      "step": 12945
    },
    {
      "entropy": 5.530332469940186,
      "epoch": 1.1747097242380262,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004868431222351892,
      "loss": 5.1229,
      "mean_token_accuracy": 0.19985790103673934,
      "num_tokens": 26879580.0,
      "step": 12950
    },
    {
      "entropy": 5.5076759338378904,
      "epoch": 1.1751632801161103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048683221777019313,
      "loss": 5.0943,
      "mean_token_accuracy": 0.2020096644759178,
      "num_tokens": 26889513.0,
      "step": 12955
    },
    {
      "entropy": 5.486361122131347,
      "epoch": 1.1756168359941945,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048682130892445293,
      "loss": 5.1776,
      "mean_token_accuracy": 0.19359630793333055,
      "num_tokens": 26900792.0,
      "step": 12960
    },
    {
      "entropy": 5.453381633758545,
      "epoch": 1.1760703918722786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000486810395698194,
      "loss": 5.0583,
      "mean_token_accuracy": 0.2014825612306595,
      "num_tokens": 26911298.0,
      "step": 12965
    },
    {
      "entropy": 5.533105802536011,
      "epoch": 1.1765239477503628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048679947809164223,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19168972223997116,
      "num_tokens": 26921545.0,
      "step": 12970
    },
    {
      "entropy": 5.528624534606934,
      "epoch": 1.176977503628447,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004867885561050232,
      "loss": 5.08,
      "mean_token_accuracy": 0.20872305780649186,
      "num_tokens": 26932140.0,
      "step": 12975
    },
    {
      "entropy": 5.496430492401123,
      "epoch": 1.1774310595065312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867776297385629,
      "loss": 5.1293,
      "mean_token_accuracy": 0.18912578672170638,
      "num_tokens": 26942634.0,
      "step": 12980
    },
    {
      "entropy": 5.506092357635498,
      "epoch": 1.1778846153846154,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048676669899248723,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20136663019657136,
      "num_tokens": 26952758.0,
      "step": 12985
    },
    {
      "entropy": 5.442593050003052,
      "epoch": 1.1783381712626997,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004867557638670222,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20214931666851044,
      "num_tokens": 26963439.0,
      "step": 12990
    },
    {
      "entropy": 5.461094951629638,
      "epoch": 1.1787917271407837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867448243623941,
      "loss": 5.099,
      "mean_token_accuracy": 0.19264838695526124,
      "num_tokens": 26974331.0,
      "step": 12995
    },
    {
      "entropy": 5.447110414505005,
      "epoch": 1.179245283018868,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000486733880478829,
      "loss": 4.967,
      "mean_token_accuracy": 0.2060053676366806,
      "num_tokens": 26983416.0,
      "step": 13000
    },
    {
      "entropy": 5.51430835723877,
      "epoch": 1.179698838896952,
      "grad_norm": 1.0,
      "learning_rate": 0.00048672293221655333,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19942551702260972,
      "num_tokens": 26994176.0,
      "step": 13005
    },
    {
      "entropy": 5.465344715118408,
      "epoch": 1.1801523947750363,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004867119795757935,
      "loss": 5.085,
      "mean_token_accuracy": 0.1964393898844719,
      "num_tokens": 27003295.0,
      "step": 13010
    },
    {
      "entropy": 5.476764297485351,
      "epoch": 1.1806059506531206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000486701022556776,
      "loss": 5.0424,
      "mean_token_accuracy": 0.19700005501508713,
      "num_tokens": 27014018.0,
      "step": 13015
    },
    {
      "entropy": 5.484636688232422,
      "epoch": 1.1810595065312046,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004866900611597273,
      "loss": 5.071,
      "mean_token_accuracy": 0.19935835003852845,
      "num_tokens": 27024706.0,
      "step": 13020
    },
    {
      "entropy": 5.532466125488281,
      "epoch": 1.1815130624092889,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048667909538487427,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20182140171527863,
      "num_tokens": 27035801.0,
      "step": 13025
    },
    {
      "entropy": 5.4090821743011475,
      "epoch": 1.181966618287373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004866681252324436,
      "loss": 4.9705,
      "mean_token_accuracy": 0.20396371632814408,
      "num_tokens": 27046596.0,
      "step": 13030
    },
    {
      "entropy": 5.454617643356324,
      "epoch": 1.1824201741654572,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048665715070266214,
      "loss": 5.0772,
      "mean_token_accuracy": 0.1967050701379776,
      "num_tokens": 27057057.0,
      "step": 13035
    },
    {
      "entropy": 5.402617454528809,
      "epoch": 1.1828737300435415,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048664617179575685,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20518202632665633,
      "num_tokens": 27067947.0,
      "step": 13040
    },
    {
      "entropy": 5.530779886245727,
      "epoch": 1.1833272859216255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004866351885119548,
      "loss": 5.0709,
      "mean_token_accuracy": 0.19731810688972473,
      "num_tokens": 27077503.0,
      "step": 13045
    },
    {
      "entropy": 5.535081672668457,
      "epoch": 1.1837808417997098,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004866242008514832,
      "loss": 5.1897,
      "mean_token_accuracy": 0.1931100070476532,
      "num_tokens": 27088298.0,
      "step": 13050
    },
    {
      "entropy": 5.475606679916382,
      "epoch": 1.1842343976777938,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048661320881456904,
      "loss": 5.0551,
      "mean_token_accuracy": 0.19839247316122055,
      "num_tokens": 27097841.0,
      "step": 13055
    },
    {
      "entropy": 5.406886243820191,
      "epoch": 1.184687953555878,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004866022124014399,
      "loss": 4.994,
      "mean_token_accuracy": 0.20290929526090623,
      "num_tokens": 27108263.0,
      "step": 13060
    },
    {
      "entropy": 5.4522699356079105,
      "epoch": 1.1851415094339623,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048659121161232313,
      "loss": 5.0589,
      "mean_token_accuracy": 0.20132856369018554,
      "num_tokens": 27119339.0,
      "step": 13065
    },
    {
      "entropy": 5.508952474594116,
      "epoch": 1.1855950653120464,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004865802064474461,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19557663500308992,
      "num_tokens": 27130393.0,
      "step": 13070
    },
    {
      "entropy": 5.534052133560181,
      "epoch": 1.1860486211901307,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048656919690703653,
      "loss": 5.1063,
      "mean_token_accuracy": 0.19433179795742034,
      "num_tokens": 27141156.0,
      "step": 13075
    },
    {
      "entropy": 5.542227840423584,
      "epoch": 1.1865021770682147,
      "grad_norm": 1.015625,
      "learning_rate": 0.000486558182991322,
      "loss": 5.1431,
      "mean_token_accuracy": 0.19479037076234818,
      "num_tokens": 27151630.0,
      "step": 13080
    },
    {
      "entropy": 5.474994468688965,
      "epoch": 1.186955732946299,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004865471647005304,
      "loss": 5.0169,
      "mean_token_accuracy": 0.20152200609445572,
      "num_tokens": 27160734.0,
      "step": 13085
    },
    {
      "entropy": 5.467565727233887,
      "epoch": 1.1874092888243832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048653614203488947,
      "loss": 5.0981,
      "mean_token_accuracy": 0.19816022962331772,
      "num_tokens": 27171584.0,
      "step": 13090
    },
    {
      "entropy": 5.464590835571289,
      "epoch": 1.1878628447024673,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004865251149946273,
      "loss": 5.005,
      "mean_token_accuracy": 0.20658984631299973,
      "num_tokens": 27180817.0,
      "step": 13095
    },
    {
      "entropy": 5.542662477493286,
      "epoch": 1.1883164005805515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048651408357997173,
      "loss": 5.0887,
      "mean_token_accuracy": 0.1919817328453064,
      "num_tokens": 27190363.0,
      "step": 13100
    },
    {
      "entropy": 5.467042398452759,
      "epoch": 1.1887699564586356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048650304779115114,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20326141864061356,
      "num_tokens": 27202181.0,
      "step": 13105
    },
    {
      "entropy": 5.43560643196106,
      "epoch": 1.1892235123367199,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004864920076283935,
      "loss": 5.0895,
      "mean_token_accuracy": 0.19626662284135818,
      "num_tokens": 27212514.0,
      "step": 13110
    },
    {
      "entropy": 5.550775957107544,
      "epoch": 1.1896770682148041,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004864809630919273,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20287249684333802,
      "num_tokens": 27223352.0,
      "step": 13115
    },
    {
      "entropy": 5.518157911300659,
      "epoch": 1.1901306240928882,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004864699141819809,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2069821238517761,
      "num_tokens": 27232643.0,
      "step": 13120
    },
    {
      "entropy": 5.41427264213562,
      "epoch": 1.1905841799709724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048645886089878276,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20254046469926834,
      "num_tokens": 27243793.0,
      "step": 13125
    },
    {
      "entropy": 5.4802384853363035,
      "epoch": 1.1910377358490567,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048644780324256153,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1999229371547699,
      "num_tokens": 27254554.0,
      "step": 13130
    },
    {
      "entropy": 5.487999629974365,
      "epoch": 1.1914912917271407,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004864367412135458,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19469751119613649,
      "num_tokens": 27265842.0,
      "step": 13135
    },
    {
      "entropy": 5.496365976333618,
      "epoch": 1.191944847605225,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048642567481196446,
      "loss": 5.1129,
      "mean_token_accuracy": 0.19370853155851364,
      "num_tokens": 27277617.0,
      "step": 13140
    },
    {
      "entropy": 5.44011173248291,
      "epoch": 1.192398403483309,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004864146040380462,
      "loss": 5.045,
      "mean_token_accuracy": 0.19978711754083633,
      "num_tokens": 27287450.0,
      "step": 13145
    },
    {
      "entropy": 5.47112717628479,
      "epoch": 1.1928519593613933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004864035288920202,
      "loss": 5.0671,
      "mean_token_accuracy": 0.20327106565237046,
      "num_tokens": 27297010.0,
      "step": 13150
    },
    {
      "entropy": 5.463698434829712,
      "epoch": 1.1933055152394776,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048639244937411516,
      "loss": 5.1043,
      "mean_token_accuracy": 0.19503520727157592,
      "num_tokens": 27306874.0,
      "step": 13155
    },
    {
      "entropy": 5.5497582912445065,
      "epoch": 1.1937590711175616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048638136548456045,
      "loss": 5.1019,
      "mean_token_accuracy": 0.19800865054130554,
      "num_tokens": 27318762.0,
      "step": 13160
    },
    {
      "entropy": 5.442694759368896,
      "epoch": 1.194212626995646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048637027722358536,
      "loss": 4.981,
      "mean_token_accuracy": 0.21250585168600084,
      "num_tokens": 27327231.0,
      "step": 13165
    },
    {
      "entropy": 5.385290479660034,
      "epoch": 1.1946661828737302,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048635918459141893,
      "loss": 5.0192,
      "mean_token_accuracy": 0.19913043081760406,
      "num_tokens": 27338517.0,
      "step": 13170
    },
    {
      "entropy": 5.528426170349121,
      "epoch": 1.1951197387518142,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048634808758829073,
      "loss": 5.0883,
      "mean_token_accuracy": 0.19603673368692398,
      "num_tokens": 27348861.0,
      "step": 13175
    },
    {
      "entropy": 5.511196565628052,
      "epoch": 1.1955732946298985,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048633698621443024,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20928379148244858,
      "num_tokens": 27358610.0,
      "step": 13180
    },
    {
      "entropy": 5.475197076797485,
      "epoch": 1.1960268505079825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000486325880470067,
      "loss": 5.1288,
      "mean_token_accuracy": 0.1985629081726074,
      "num_tokens": 27369707.0,
      "step": 13185
    },
    {
      "entropy": 5.4696228981018065,
      "epoch": 1.1964804063860668,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004863147703554307,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19705890119075775,
      "num_tokens": 27381030.0,
      "step": 13190
    },
    {
      "entropy": 5.519057321548462,
      "epoch": 1.196933962264151,
      "grad_norm": 1.0,
      "learning_rate": 0.00048630365587075115,
      "loss": 5.0961,
      "mean_token_accuracy": 0.19812251776456832,
      "num_tokens": 27390866.0,
      "step": 13195
    },
    {
      "entropy": 5.493912649154663,
      "epoch": 1.197387518142235,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048629253701625816,
      "loss": 5.0357,
      "mean_token_accuracy": 0.20332785397768022,
      "num_tokens": 27400724.0,
      "step": 13200
    },
    {
      "entropy": 5.473924732208252,
      "epoch": 1.1978410740203194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004862814137921816,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19959327727556228,
      "num_tokens": 27411212.0,
      "step": 13205
    },
    {
      "entropy": 5.467393922805786,
      "epoch": 1.1982946298984034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048627028619875157,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1994468241930008,
      "num_tokens": 27421279.0,
      "step": 13210
    },
    {
      "entropy": 5.483545017242432,
      "epoch": 1.1987481857764877,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004862591542361982,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19779787510633468,
      "num_tokens": 27432520.0,
      "step": 13215
    },
    {
      "entropy": 5.4180910110473635,
      "epoch": 1.199201741654572,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004862480179047516,
      "loss": 4.989,
      "mean_token_accuracy": 0.20875904560089112,
      "num_tokens": 27442246.0,
      "step": 13220
    },
    {
      "entropy": 5.482895421981811,
      "epoch": 1.199655297532656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004862368772046423,
      "loss": 5.0607,
      "mean_token_accuracy": 0.19732182919979097,
      "num_tokens": 27452134.0,
      "step": 13225
    },
    {
      "entropy": 5.496697950363159,
      "epoch": 1.2001088534107403,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048622573213610057,
      "loss": 5.1461,
      "mean_token_accuracy": 0.188792884349823,
      "num_tokens": 27462253.0,
      "step": 13230
    },
    {
      "entropy": 5.511988878250122,
      "epoch": 1.2005624092888243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048621458269935676,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20037435442209245,
      "num_tokens": 27473131.0,
      "step": 13235
    },
    {
      "entropy": 5.489395809173584,
      "epoch": 1.2010159651669086,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048620342889464153,
      "loss": 5.0756,
      "mean_token_accuracy": 0.2012531042098999,
      "num_tokens": 27483683.0,
      "step": 13240
    },
    {
      "entropy": 5.384080266952514,
      "epoch": 1.2014695210449928,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048619227072218565,
      "loss": 5.052,
      "mean_token_accuracy": 0.1951747328042984,
      "num_tokens": 27493679.0,
      "step": 13245
    },
    {
      "entropy": 5.4413477897644045,
      "epoch": 1.2019230769230769,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004861811081822198,
      "loss": 5.0351,
      "mean_token_accuracy": 0.20437403321266173,
      "num_tokens": 27503897.0,
      "step": 13250
    },
    {
      "entropy": 5.532030296325684,
      "epoch": 1.2023766328011611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004861699412749748,
      "loss": 5.1421,
      "mean_token_accuracy": 0.192557655274868,
      "num_tokens": 27514560.0,
      "step": 13255
    },
    {
      "entropy": 5.4998753547668455,
      "epoch": 1.2028301886792452,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004861587700006816,
      "loss": 5.0427,
      "mean_token_accuracy": 0.1978370800614357,
      "num_tokens": 27525594.0,
      "step": 13260
    },
    {
      "entropy": 5.426636695861816,
      "epoch": 1.2032837445573294,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048614759435957114,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2014489009976387,
      "num_tokens": 27535676.0,
      "step": 13265
    },
    {
      "entropy": 5.413556241989136,
      "epoch": 1.2037373004354137,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004861364143518747,
      "loss": 5.0281,
      "mean_token_accuracy": 0.19157828241586686,
      "num_tokens": 27545520.0,
      "step": 13270
    },
    {
      "entropy": 5.573331117630005,
      "epoch": 1.2041908563134978,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004861252299778234,
      "loss": 5.0973,
      "mean_token_accuracy": 0.18991344869136811,
      "num_tokens": 27556866.0,
      "step": 13275
    },
    {
      "entropy": 5.47489185333252,
      "epoch": 1.204644412191582,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048611404123764855,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19900879561901091,
      "num_tokens": 27566951.0,
      "step": 13280
    },
    {
      "entropy": 5.470123195648194,
      "epoch": 1.205097968069666,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048610284813158154,
      "loss": 5.1365,
      "mean_token_accuracy": 0.1957337498664856,
      "num_tokens": 27576796.0,
      "step": 13285
    },
    {
      "entropy": 5.402993488311767,
      "epoch": 1.2055515239477503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004860916506598538,
      "loss": 5.0034,
      "mean_token_accuracy": 0.1991758719086647,
      "num_tokens": 27586902.0,
      "step": 13290
    },
    {
      "entropy": 5.585123109817505,
      "epoch": 1.2060050798258346,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048608044882269695,
      "loss": 5.1957,
      "mean_token_accuracy": 0.18857692927122116,
      "num_tokens": 27599232.0,
      "step": 13295
    },
    {
      "entropy": 5.482767248153687,
      "epoch": 1.2064586357039186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004860692426203427,
      "loss": 4.9931,
      "mean_token_accuracy": 0.1983819544315338,
      "num_tokens": 27609022.0,
      "step": 13300
    },
    {
      "entropy": 5.515181207656861,
      "epoch": 1.206912191582003,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004860580320530226,
      "loss": 5.1371,
      "mean_token_accuracy": 0.19917869865894317,
      "num_tokens": 27619829.0,
      "step": 13305
    },
    {
      "entropy": 5.515312433242798,
      "epoch": 1.2073657474600872,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004860468171209687,
      "loss": 5.1298,
      "mean_token_accuracy": 0.20136113315820695,
      "num_tokens": 27629337.0,
      "step": 13310
    },
    {
      "entropy": 5.546958637237549,
      "epoch": 1.2078193033381712,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004860355978244128,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19766863137483598,
      "num_tokens": 27639377.0,
      "step": 13315
    },
    {
      "entropy": 5.500257062911987,
      "epoch": 1.2082728592162555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000486024374163587,
      "loss": 5.0404,
      "mean_token_accuracy": 0.19715555906295776,
      "num_tokens": 27649989.0,
      "step": 13320
    },
    {
      "entropy": 5.49214277267456,
      "epoch": 1.2087264150943395,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048601314613872334,
      "loss": 5.1191,
      "mean_token_accuracy": 0.19113418459892273,
      "num_tokens": 27659967.0,
      "step": 13325
    },
    {
      "entropy": 5.509144020080567,
      "epoch": 1.2091799709724238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000486001913750054,
      "loss": 5.0724,
      "mean_token_accuracy": 0.19962707608938218,
      "num_tokens": 27670099.0,
      "step": 13330
    },
    {
      "entropy": 5.463675832748413,
      "epoch": 1.209633526850508,
      "grad_norm": 1.0,
      "learning_rate": 0.0004859906769978113,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1901517853140831,
      "num_tokens": 27681497.0,
      "step": 13335
    },
    {
      "entropy": 5.572259140014649,
      "epoch": 1.2100870827285921,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048597943588222776,
      "loss": 5.2052,
      "mean_token_accuracy": 0.19154019355773927,
      "num_tokens": 27691828.0,
      "step": 13340
    },
    {
      "entropy": 5.538029193878174,
      "epoch": 1.2105406386066764,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048596819040353564,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20628093034029008,
      "num_tokens": 27701426.0,
      "step": 13345
    },
    {
      "entropy": 5.493690490722656,
      "epoch": 1.2109941944847606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859569405619676,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20149533450603485,
      "num_tokens": 27711112.0,
      "step": 13350
    },
    {
      "entropy": 5.52337875366211,
      "epoch": 1.2114477503628447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004859456863577562,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19508373439311982,
      "num_tokens": 27723160.0,
      "step": 13355
    },
    {
      "entropy": 5.522914218902588,
      "epoch": 1.211901306240929,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048593442779113437,
      "loss": 5.1034,
      "mean_token_accuracy": 0.19623059332370757,
      "num_tokens": 27733143.0,
      "step": 13360
    },
    {
      "entropy": 5.5026891231536865,
      "epoch": 1.212354862119013,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004859231648623348,
      "loss": 5.0884,
      "mean_token_accuracy": 0.19636708945035936,
      "num_tokens": 27743956.0,
      "step": 13365
    },
    {
      "entropy": 5.522414779663086,
      "epoch": 1.2128084179970973,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048591189757159035,
      "loss": 5.0851,
      "mean_token_accuracy": 0.2020998165011406,
      "num_tokens": 27754192.0,
      "step": 13370
    },
    {
      "entropy": 5.487301158905029,
      "epoch": 1.2132619738751815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004859006259191342,
      "loss": 5.1205,
      "mean_token_accuracy": 0.18899906873703004,
      "num_tokens": 27763939.0,
      "step": 13375
    },
    {
      "entropy": 5.485107564926148,
      "epoch": 1.2137155297532656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004858893499051993,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20167715549468995,
      "num_tokens": 27774743.0,
      "step": 13380
    },
    {
      "entropy": 5.508217763900757,
      "epoch": 1.2141690856313498,
      "grad_norm": 1.125,
      "learning_rate": 0.0004858780695300189,
      "loss": 5.0722,
      "mean_token_accuracy": 0.19787210524082183,
      "num_tokens": 27785344.0,
      "step": 13385
    },
    {
      "entropy": 5.520311784744263,
      "epoch": 1.2146226415094339,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004858667847938263,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19476209133863448,
      "num_tokens": 27797101.0,
      "step": 13390
    },
    {
      "entropy": 5.452220058441162,
      "epoch": 1.2150761973875182,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048585549569685486,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2015383094549179,
      "num_tokens": 27807581.0,
      "step": 13395
    },
    {
      "entropy": 5.462075805664062,
      "epoch": 1.2155297532656024,
      "grad_norm": 1.0,
      "learning_rate": 0.00048584420223933805,
      "loss": 5.0023,
      "mean_token_accuracy": 0.19964679926633835,
      "num_tokens": 27818170.0,
      "step": 13400
    },
    {
      "entropy": 5.432729625701905,
      "epoch": 1.2159833091436865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004858329044215094,
      "loss": 5.029,
      "mean_token_accuracy": 0.20274798572063446,
      "num_tokens": 27828153.0,
      "step": 13405
    },
    {
      "entropy": 5.487482118606567,
      "epoch": 1.2164368650217707,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048582160224360253,
      "loss": 4.9698,
      "mean_token_accuracy": 0.206896835565567,
      "num_tokens": 27837681.0,
      "step": 13410
    },
    {
      "entropy": 5.494462537765503,
      "epoch": 1.2168904208998548,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004858102957058513,
      "loss": 5.108,
      "mean_token_accuracy": 0.19647631645202637,
      "num_tokens": 27847447.0,
      "step": 13415
    },
    {
      "entropy": 5.444395112991333,
      "epoch": 1.217343976777939,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048579898480848933,
      "loss": 4.9716,
      "mean_token_accuracy": 0.2026344954967499,
      "num_tokens": 27858586.0,
      "step": 13420
    },
    {
      "entropy": 5.495455551147461,
      "epoch": 1.2177975326560233,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004857876695517507,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19661417752504348,
      "num_tokens": 27868832.0,
      "step": 13425
    },
    {
      "entropy": 5.492994785308838,
      "epoch": 1.2182510885341074,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048577634993586923,
      "loss": 5.0952,
      "mean_token_accuracy": 0.19540033489465714,
      "num_tokens": 27879339.0,
      "step": 13430
    },
    {
      "entropy": 5.515705108642578,
      "epoch": 1.2187046444121916,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048576502596107924,
      "loss": 5.0881,
      "mean_token_accuracy": 0.19840986430644988,
      "num_tokens": 27889964.0,
      "step": 13435
    },
    {
      "entropy": 5.5269997119903564,
      "epoch": 1.2191582002902757,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857536976276148,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19223980158567427,
      "num_tokens": 27901004.0,
      "step": 13440
    },
    {
      "entropy": 5.484214353561401,
      "epoch": 1.21961175616836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004857423649357102,
      "loss": 5.0613,
      "mean_token_accuracy": 0.1969600275158882,
      "num_tokens": 27910788.0,
      "step": 13445
    },
    {
      "entropy": 5.544168424606323,
      "epoch": 1.2200653120464442,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048573102788559974,
      "loss": 5.1035,
      "mean_token_accuracy": 0.2005355179309845,
      "num_tokens": 27921360.0,
      "step": 13450
    },
    {
      "entropy": 5.482368850708008,
      "epoch": 1.2205188679245282,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004857196864775179,
      "loss": 5.0467,
      "mean_token_accuracy": 0.20181792825460435,
      "num_tokens": 27931889.0,
      "step": 13455
    },
    {
      "entropy": 5.456076622009277,
      "epoch": 1.2209724238026125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048570834071169926,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20518947392702103,
      "num_tokens": 27942049.0,
      "step": 13460
    },
    {
      "entropy": 5.538109827041626,
      "epoch": 1.2214259796806965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048569699058837856,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1968417227268219,
      "num_tokens": 27952416.0,
      "step": 13465
    },
    {
      "entropy": 5.467518997192383,
      "epoch": 1.2218795355587808,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004856856361077903,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20182372778654098,
      "num_tokens": 27962009.0,
      "step": 13470
    },
    {
      "entropy": 5.550679159164429,
      "epoch": 1.222333091436865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004856742772701694,
      "loss": 5.0959,
      "mean_token_accuracy": 0.19606231898069382,
      "num_tokens": 27972175.0,
      "step": 13475
    },
    {
      "entropy": 5.5194611072540285,
      "epoch": 1.2227866473149491,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048566291407575073,
      "loss": 5.1141,
      "mean_token_accuracy": 0.2015317127108574,
      "num_tokens": 27982899.0,
      "step": 13480
    },
    {
      "entropy": 5.513904333114624,
      "epoch": 1.2232402031930334,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048565154652476937,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20250820070505143,
      "num_tokens": 27993231.0,
      "step": 13485
    },
    {
      "entropy": 5.432183265686035,
      "epoch": 1.2236937590711174,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004856401746174603,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20159731060266495,
      "num_tokens": 28003623.0,
      "step": 13490
    },
    {
      "entropy": 5.488606214523315,
      "epoch": 1.2241473149492017,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048562879835405874,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20178948640823363,
      "num_tokens": 28013878.0,
      "step": 13495
    },
    {
      "entropy": 5.447732400894165,
      "epoch": 1.224600870827286,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004856174177348,
      "loss": 5.0941,
      "mean_token_accuracy": 0.19857164174318315,
      "num_tokens": 28024224.0,
      "step": 13500
    },
    {
      "entropy": 5.470084142684937,
      "epoch": 1.22505442670537,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004856060327599194,
      "loss": 5.0161,
      "mean_token_accuracy": 0.1966568797826767,
      "num_tokens": 28034819.0,
      "step": 13505
    },
    {
      "entropy": 5.463940000534057,
      "epoch": 1.2255079825834543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048559464342965233,
      "loss": 5.0775,
      "mean_token_accuracy": 0.19706240743398667,
      "num_tokens": 28044240.0,
      "step": 13510
    },
    {
      "entropy": 5.565493249893189,
      "epoch": 1.2259615384615385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004855832497442344,
      "loss": 5.2113,
      "mean_token_accuracy": 0.18766228705644608,
      "num_tokens": 28055046.0,
      "step": 13515
    },
    {
      "entropy": 5.462039518356323,
      "epoch": 1.2264150943396226,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004855718517039012,
      "loss": 5.0052,
      "mean_token_accuracy": 0.205356565117836,
      "num_tokens": 28065272.0,
      "step": 13520
    },
    {
      "entropy": 5.517233610153198,
      "epoch": 1.2268686502177069,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048556044930888844,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2079099714756012,
      "num_tokens": 28076255.0,
      "step": 13525
    },
    {
      "entropy": 5.515420389175415,
      "epoch": 1.2273222060957911,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004855490425594319,
      "loss": 5.1358,
      "mean_token_accuracy": 0.19787602573633195,
      "num_tokens": 28086258.0,
      "step": 13530
    },
    {
      "entropy": 5.494659328460694,
      "epoch": 1.2277757619738752,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004855376314557675,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20258419066667557,
      "num_tokens": 28096497.0,
      "step": 13535
    },
    {
      "entropy": 5.510882329940796,
      "epoch": 1.2282293178519594,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004855262159981313,
      "loss": 5.0602,
      "mean_token_accuracy": 0.2041010618209839,
      "num_tokens": 28106282.0,
      "step": 13540
    },
    {
      "entropy": 5.450311946868896,
      "epoch": 1.2286828737300435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048551479618675925,
      "loss": 5.1113,
      "mean_token_accuracy": 0.19611413329839705,
      "num_tokens": 28115771.0,
      "step": 13545
    },
    {
      "entropy": 5.526232290267944,
      "epoch": 1.2291364296081277,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048550337202188756,
      "loss": 5.0407,
      "mean_token_accuracy": 0.2037350058555603,
      "num_tokens": 28125477.0,
      "step": 13550
    },
    {
      "entropy": 5.55634880065918,
      "epoch": 1.229589985486212,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004854919435037525,
      "loss": 5.0788,
      "mean_token_accuracy": 0.19526431262493132,
      "num_tokens": 28135800.0,
      "step": 13555
    },
    {
      "entropy": 5.429639768600464,
      "epoch": 1.230043541364296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004854805106325904,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2002708464860916,
      "num_tokens": 28146779.0,
      "step": 13560
    },
    {
      "entropy": 5.4886777877807615,
      "epoch": 1.2304970972423803,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004854690734086378,
      "loss": 5.1104,
      "mean_token_accuracy": 0.19459391683340072,
      "num_tokens": 28156819.0,
      "step": 13565
    },
    {
      "entropy": 5.470524597167969,
      "epoch": 1.2309506531204644,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004854576318321311,
      "loss": 5.0871,
      "mean_token_accuracy": 0.1952335223555565,
      "num_tokens": 28168553.0,
      "step": 13570
    },
    {
      "entropy": 5.530834293365478,
      "epoch": 1.2314042089985486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048544618590330685,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19884106814861296,
      "num_tokens": 28178064.0,
      "step": 13575
    },
    {
      "entropy": 5.565642642974853,
      "epoch": 1.231857764876633,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048543473562240194,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19776492416858674,
      "num_tokens": 28188580.0,
      "step": 13580
    },
    {
      "entropy": 5.524669981002807,
      "epoch": 1.232311320754717,
      "grad_norm": 0.984375,
      "learning_rate": 0.000485423280989653,
      "loss": 5.1629,
      "mean_token_accuracy": 0.19327968508005142,
      "num_tokens": 28199955.0,
      "step": 13585
    },
    {
      "entropy": 5.5645153522491455,
      "epoch": 1.2327648766328012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000485411822005297,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1941158130764961,
      "num_tokens": 28210506.0,
      "step": 13590
    },
    {
      "entropy": 5.493167734146118,
      "epoch": 1.2332184325108853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048540035866957097,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2134607493877411,
      "num_tokens": 28220486.0,
      "step": 13595
    },
    {
      "entropy": 5.521177005767822,
      "epoch": 1.2336719883889695,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004853888909827118,
      "loss": 5.1623,
      "mean_token_accuracy": 0.18835390061140062,
      "num_tokens": 28232668.0,
      "step": 13600
    },
    {
      "entropy": 5.507457447052002,
      "epoch": 1.2341255442670538,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853774189449568,
      "loss": 5.1072,
      "mean_token_accuracy": 0.1947883576154709,
      "num_tokens": 28243733.0,
      "step": 13605
    },
    {
      "entropy": 5.572214269638062,
      "epoch": 1.2345791001451378,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048536594255654315,
      "loss": 5.1413,
      "mean_token_accuracy": 0.19594713300466537,
      "num_tokens": 28254953.0,
      "step": 13610
    },
    {
      "entropy": 5.530012893676758,
      "epoch": 1.235032656023222,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004853544618177083,
      "loss": 5.0832,
      "mean_token_accuracy": 0.20127384960651398,
      "num_tokens": 28264959.0,
      "step": 13615
    },
    {
      "entropy": 5.54983115196228,
      "epoch": 1.2354862119013061,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048534297672868945,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19841146767139434,
      "num_tokens": 28275509.0,
      "step": 13620
    },
    {
      "entropy": 5.500857210159301,
      "epoch": 1.2359397677793904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048533148728972434,
      "loss": 5.0277,
      "mean_token_accuracy": 0.2020431414246559,
      "num_tokens": 28285983.0,
      "step": 13625
    },
    {
      "entropy": 5.510316038131714,
      "epoch": 1.2363933236574747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048531999350105037,
      "loss": 5.1071,
      "mean_token_accuracy": 0.1961878165602684,
      "num_tokens": 28296178.0,
      "step": 13630
    },
    {
      "entropy": 5.491313409805298,
      "epoch": 1.2368468795355587,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048530849536290537,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19448001086711883,
      "num_tokens": 28305741.0,
      "step": 13635
    },
    {
      "entropy": 5.42606258392334,
      "epoch": 1.237300435413643,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004852969928755271,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2011515825986862,
      "num_tokens": 28315607.0,
      "step": 13640
    },
    {
      "entropy": 5.420935916900635,
      "epoch": 1.237753991291727,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048528548603915344,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20115619599819184,
      "num_tokens": 28327349.0,
      "step": 13645
    },
    {
      "entropy": 5.52898759841919,
      "epoch": 1.2382075471698113,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048527397485402236,
      "loss": 5.0485,
      "mean_token_accuracy": 0.19680120646953583,
      "num_tokens": 28339949.0,
      "step": 13650
    },
    {
      "entropy": 5.542382383346558,
      "epoch": 1.2386611030478956,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048526245932037186,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20272745341062545,
      "num_tokens": 28351340.0,
      "step": 13655
    },
    {
      "entropy": 5.385624742507934,
      "epoch": 1.2391146589259796,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048525093943844013,
      "loss": 4.9409,
      "mean_token_accuracy": 0.20911214351654053,
      "num_tokens": 28360929.0,
      "step": 13660
    },
    {
      "entropy": 5.426797580718994,
      "epoch": 1.2395682148040639,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004852394152084654,
      "loss": 5.0391,
      "mean_token_accuracy": 0.2021891802549362,
      "num_tokens": 28370477.0,
      "step": 13665
    },
    {
      "entropy": 5.446661186218262,
      "epoch": 1.240021770682148,
      "grad_norm": 0.984375,
      "learning_rate": 0.000485227886630686,
      "loss": 4.966,
      "mean_token_accuracy": 0.21486925035715104,
      "num_tokens": 28380876.0,
      "step": 13670
    },
    {
      "entropy": 5.394935607910156,
      "epoch": 1.2404753265602322,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004852163537053403,
      "loss": 4.9434,
      "mean_token_accuracy": 0.20210244953632356,
      "num_tokens": 28390685.0,
      "step": 13675
    },
    {
      "entropy": 5.438671875,
      "epoch": 1.2409288824383164,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004852048164326669,
      "loss": 5.0539,
      "mean_token_accuracy": 0.19393928349018097,
      "num_tokens": 28400236.0,
      "step": 13680
    },
    {
      "entropy": 5.503456926345825,
      "epoch": 1.2413824383164005,
      "grad_norm": 1.125,
      "learning_rate": 0.0004851932748129043,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20332229733467103,
      "num_tokens": 28410353.0,
      "step": 13685
    },
    {
      "entropy": 5.507759523391724,
      "epoch": 1.2418359941944848,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004851817288462912,
      "loss": 5.0384,
      "mean_token_accuracy": 0.1971975803375244,
      "num_tokens": 28420241.0,
      "step": 13690
    },
    {
      "entropy": 5.471080160140991,
      "epoch": 1.242289550072569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004851701785330663,
      "loss": 5.0943,
      "mean_token_accuracy": 0.19966191947460174,
      "num_tokens": 28430610.0,
      "step": 13695
    },
    {
      "entropy": 5.473397445678711,
      "epoch": 1.242743105950653,
      "grad_norm": 1.0,
      "learning_rate": 0.0004851586238734687,
      "loss": 5.0184,
      "mean_token_accuracy": 0.19985776245594025,
      "num_tokens": 28440026.0,
      "step": 13700
    },
    {
      "entropy": 5.48423581123352,
      "epoch": 1.2431966618287373,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004851470648677371,
      "loss": 5.0913,
      "mean_token_accuracy": 0.1909073993563652,
      "num_tokens": 28450707.0,
      "step": 13705
    },
    {
      "entropy": 5.484880018234253,
      "epoch": 1.2436502177068216,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004851355015161107,
      "loss": 5.04,
      "mean_token_accuracy": 0.19988873600959778,
      "num_tokens": 28462378.0,
      "step": 13710
    },
    {
      "entropy": 5.562987518310547,
      "epoch": 1.2441037735849056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004851239338188286,
      "loss": 5.1173,
      "mean_token_accuracy": 0.19390659928321838,
      "num_tokens": 28473084.0,
      "step": 13715
    },
    {
      "entropy": 5.484537076950073,
      "epoch": 1.24455732946299,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048511236177612994,
      "loss": 5.041,
      "mean_token_accuracy": 0.1961607202887535,
      "num_tokens": 28482767.0,
      "step": 13720
    },
    {
      "entropy": 5.44678144454956,
      "epoch": 1.245010885341074,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048510078538825413,
      "loss": 5.0312,
      "mean_token_accuracy": 0.20292534232139586,
      "num_tokens": 28493274.0,
      "step": 13725
    },
    {
      "entropy": 5.520133829116821,
      "epoch": 1.2454644412191582,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048508920465544056,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20110096782445908,
      "num_tokens": 28503823.0,
      "step": 13730
    },
    {
      "entropy": 5.439535331726074,
      "epoch": 1.2459179970972425,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004850776195779286,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20129793733358384,
      "num_tokens": 28514209.0,
      "step": 13735
    },
    {
      "entropy": 5.532947778701782,
      "epoch": 1.2463715529753265,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000485066030155958,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19237526655197143,
      "num_tokens": 28524665.0,
      "step": 13740
    },
    {
      "entropy": 5.485327291488647,
      "epoch": 1.2468251088534108,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004850544363897684,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20430061221122742,
      "num_tokens": 28533639.0,
      "step": 13745
    },
    {
      "entropy": 5.405498456954956,
      "epoch": 1.2472786647314948,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048504283827959947,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20359645783901215,
      "num_tokens": 28544469.0,
      "step": 13750
    },
    {
      "entropy": 5.532805395126343,
      "epoch": 1.2477322206095791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850312358256912,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20535819679498674,
      "num_tokens": 28554637.0,
      "step": 13755
    },
    {
      "entropy": 5.463185548782349,
      "epoch": 1.2481857764876634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004850196290282834,
      "loss": 5.0384,
      "mean_token_accuracy": 0.20495663583278656,
      "num_tokens": 28564752.0,
      "step": 13760
    },
    {
      "entropy": 5.503740978240967,
      "epoch": 1.2486393323657474,
      "grad_norm": 1.0,
      "learning_rate": 0.00048500801788761613,
      "loss": 5.1586,
      "mean_token_accuracy": 0.19918418526649476,
      "num_tokens": 28575007.0,
      "step": 13765
    },
    {
      "entropy": 5.5523582935333256,
      "epoch": 1.2490928882438317,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004849964024039296,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19649922996759414,
      "num_tokens": 28584907.0,
      "step": 13770
    },
    {
      "entropy": 5.572278308868408,
      "epoch": 1.2495464441219157,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048498478257746387,
      "loss": 5.1286,
      "mean_token_accuracy": 0.20043385326862334,
      "num_tokens": 28594244.0,
      "step": 13775
    },
    {
      "entropy": 5.506260251998901,
      "epoch": 1.25,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048497315840845934,
      "loss": 5.0921,
      "mean_token_accuracy": 0.1991775080561638,
      "num_tokens": 28603816.0,
      "step": 13780
    },
    {
      "entropy": 5.560829496383667,
      "epoch": 1.2504535558780843,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004849615298971565,
      "loss": 5.1309,
      "mean_token_accuracy": 0.19353427439928056,
      "num_tokens": 28613067.0,
      "step": 13785
    },
    {
      "entropy": 5.500925207138062,
      "epoch": 1.2509071117561683,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048494989704379556,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20042569041252137,
      "num_tokens": 28623752.0,
      "step": 13790
    },
    {
      "entropy": 5.450966548919678,
      "epoch": 1.2513606676342526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004849382598486173,
      "loss": 5.0284,
      "mean_token_accuracy": 0.2054324969649315,
      "num_tokens": 28633190.0,
      "step": 13795
    },
    {
      "entropy": 5.589546823501587,
      "epoch": 1.2518142235123366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048492661831186243,
      "loss": 5.1193,
      "mean_token_accuracy": 0.19111987203359604,
      "num_tokens": 28644567.0,
      "step": 13800
    },
    {
      "entropy": 5.544061851501465,
      "epoch": 1.2522677793904209,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004849149724337715,
      "loss": 5.1056,
      "mean_token_accuracy": 0.19287900030612945,
      "num_tokens": 28654785.0,
      "step": 13805
    },
    {
      "entropy": 5.419747304916382,
      "epoch": 1.2527213352685052,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048490332221458543,
      "loss": 4.9955,
      "mean_token_accuracy": 0.19568415731191635,
      "num_tokens": 28665286.0,
      "step": 13810
    },
    {
      "entropy": 5.51972770690918,
      "epoch": 1.2531748911465892,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004848916676545452,
      "loss": 5.1157,
      "mean_token_accuracy": 0.19823004603385924,
      "num_tokens": 28675830.0,
      "step": 13815
    },
    {
      "entropy": 5.568089532852173,
      "epoch": 1.2536284470246735,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048488000875389174,
      "loss": 5.1713,
      "mean_token_accuracy": 0.18988842964172364,
      "num_tokens": 28686640.0,
      "step": 13820
    },
    {
      "entropy": 5.505718946456909,
      "epoch": 1.2540820029027575,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004848683455128663,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20179916322231292,
      "num_tokens": 28697677.0,
      "step": 13825
    },
    {
      "entropy": 5.494057321548462,
      "epoch": 1.2545355587808418,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048485667793170994,
      "loss": 5.0571,
      "mean_token_accuracy": 0.19772837162017823,
      "num_tokens": 28708831.0,
      "step": 13830
    },
    {
      "entropy": 5.5372607707977295,
      "epoch": 1.254989114658926,
      "grad_norm": 1.0,
      "learning_rate": 0.000484845006010664,
      "loss": 5.1174,
      "mean_token_accuracy": 0.19325800836086274,
      "num_tokens": 28718959.0,
      "step": 13835
    },
    {
      "entropy": 5.494212627410889,
      "epoch": 1.25544267053701,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004848333297499699,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20306683331727982,
      "num_tokens": 28728719.0,
      "step": 13840
    },
    {
      "entropy": 5.565085124969483,
      "epoch": 1.2558962264150944,
      "grad_norm": 1.0,
      "learning_rate": 0.00048482164914986903,
      "loss": 5.2087,
      "mean_token_accuracy": 0.18758925199508666,
      "num_tokens": 28740032.0,
      "step": 13845
    },
    {
      "entropy": 5.449249124526977,
      "epoch": 1.2563497822931784,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048480996421060307,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2051217496395111,
      "num_tokens": 28749374.0,
      "step": 13850
    },
    {
      "entropy": 5.466462850570679,
      "epoch": 1.2568033381712627,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048479827493241347,
      "loss": 5.0566,
      "mean_token_accuracy": 0.19801553338766098,
      "num_tokens": 28760038.0,
      "step": 13855
    },
    {
      "entropy": 5.49345588684082,
      "epoch": 1.257256894049347,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048478658131554216,
      "loss": 5.0803,
      "mean_token_accuracy": 0.19837768524885177,
      "num_tokens": 28770247.0,
      "step": 13860
    },
    {
      "entropy": 5.488525819778443,
      "epoch": 1.2577104499274312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004847748833602308,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20438348203897477,
      "num_tokens": 28780251.0,
      "step": 13865
    },
    {
      "entropy": 5.443554973602295,
      "epoch": 1.2581640058055152,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004847631810667215,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2038698151707649,
      "num_tokens": 28790829.0,
      "step": 13870
    },
    {
      "entropy": 5.4590709686279295,
      "epoch": 1.2586175616835993,
      "grad_norm": 1.0,
      "learning_rate": 0.0004847514744352561,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2023104190826416,
      "num_tokens": 28801036.0,
      "step": 13875
    },
    {
      "entropy": 5.403080654144287,
      "epoch": 1.2590711175616836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004847397634660768,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20570943951606752,
      "num_tokens": 28811013.0,
      "step": 13880
    },
    {
      "entropy": 5.440387964248657,
      "epoch": 1.2595246734397678,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048472804815942575,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20063267648220062,
      "num_tokens": 28821367.0,
      "step": 13885
    },
    {
      "entropy": 5.50364317893982,
      "epoch": 1.259978229317852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048471632851554517,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20409667789936065,
      "num_tokens": 28831077.0,
      "step": 13890
    },
    {
      "entropy": 5.5035868167877195,
      "epoch": 1.2604317851959361,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004847046045346775,
      "loss": 5.0865,
      "mean_token_accuracy": 0.1920575052499771,
      "num_tokens": 28841490.0,
      "step": 13895
    },
    {
      "entropy": 5.508995771408081,
      "epoch": 1.2608853410740204,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048469287621706516,
      "loss": 5.0859,
      "mean_token_accuracy": 0.19784659445285796,
      "num_tokens": 28852510.0,
      "step": 13900
    },
    {
      "entropy": 5.496377563476562,
      "epoch": 1.2613388969521044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004846811435629507,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20141980946063995,
      "num_tokens": 28863388.0,
      "step": 13905
    },
    {
      "entropy": 5.401807069778442,
      "epoch": 1.2617924528301887,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004846694065725768,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20184410065412522,
      "num_tokens": 28873798.0,
      "step": 13910
    },
    {
      "entropy": 5.483473587036133,
      "epoch": 1.262246008708273,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004846576652461861,
      "loss": 5.0442,
      "mean_token_accuracy": 0.19962013363838196,
      "num_tokens": 28883482.0,
      "step": 13915
    },
    {
      "entropy": 5.4887090682983395,
      "epoch": 1.262699564586357,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048464591958402147,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20471181720495224,
      "num_tokens": 28893573.0,
      "step": 13920
    },
    {
      "entropy": 5.560998773574829,
      "epoch": 1.2631531204644413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048463416958632573,
      "loss": 5.1454,
      "mean_token_accuracy": 0.1967838451266289,
      "num_tokens": 28904064.0,
      "step": 13925
    },
    {
      "entropy": 5.505468559265137,
      "epoch": 1.2636066763425253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048462241525334206,
      "loss": 5.1,
      "mean_token_accuracy": 0.19794451743364333,
      "num_tokens": 28915329.0,
      "step": 13930
    },
    {
      "entropy": 5.49501748085022,
      "epoch": 1.2640602322206096,
      "grad_norm": 1.0,
      "learning_rate": 0.0004846106565853133,
      "loss": 5.1859,
      "mean_token_accuracy": 0.19487343728542328,
      "num_tokens": 28925228.0,
      "step": 13935
    },
    {
      "entropy": 5.477839565277099,
      "epoch": 1.2645137880986939,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048459889358248284,
      "loss": 5.023,
      "mean_token_accuracy": 0.19917779564857482,
      "num_tokens": 28935891.0,
      "step": 13940
    },
    {
      "entropy": 5.4873875141143795,
      "epoch": 1.264967343976778,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048458712624509383,
      "loss": 4.985,
      "mean_token_accuracy": 0.20131113827228547,
      "num_tokens": 28946372.0,
      "step": 13945
    },
    {
      "entropy": 5.5058043003082275,
      "epoch": 1.2654208998548622,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048457535457338964,
      "loss": 5.0739,
      "mean_token_accuracy": 0.1947830557823181,
      "num_tokens": 28957874.0,
      "step": 13950
    },
    {
      "entropy": 5.5295580387115475,
      "epoch": 1.2658744557329462,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004845635785676137,
      "loss": 5.0528,
      "mean_token_accuracy": 0.19305006712675093,
      "num_tokens": 28968023.0,
      "step": 13955
    },
    {
      "entropy": 5.480109596252442,
      "epoch": 1.2663280116110305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048455179822800957,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2045432895421982,
      "num_tokens": 28977744.0,
      "step": 13960
    },
    {
      "entropy": 5.56711974143982,
      "epoch": 1.2667815674891147,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004845400135548208,
      "loss": 5.2127,
      "mean_token_accuracy": 0.191379314661026,
      "num_tokens": 28988875.0,
      "step": 13965
    },
    {
      "entropy": 5.415848731994629,
      "epoch": 1.2672351233671988,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004845282245482912,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20306566059589387,
      "num_tokens": 28998562.0,
      "step": 13970
    },
    {
      "entropy": 5.4908302307128904,
      "epoch": 1.267688679245283,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845164312086646,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20640750229358673,
      "num_tokens": 29008783.0,
      "step": 13975
    },
    {
      "entropy": 5.436105823516845,
      "epoch": 1.268142235123367,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004845046335361847,
      "loss": 4.9642,
      "mean_token_accuracy": 0.19867230355739593,
      "num_tokens": 29019303.0,
      "step": 13980
    },
    {
      "entropy": 5.5175525665283205,
      "epoch": 1.2685957910014514,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048449283153109566,
      "loss": 5.1912,
      "mean_token_accuracy": 0.18820945173501968,
      "num_tokens": 29029843.0,
      "step": 13985
    },
    {
      "entropy": 5.526941013336182,
      "epoch": 1.2690493468795356,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004844810251936413,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19493067264556885,
      "num_tokens": 29040603.0,
      "step": 13990
    },
    {
      "entropy": 5.484497308731079,
      "epoch": 1.2695029027576197,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004844692145240661,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20133136510848998,
      "num_tokens": 29050599.0,
      "step": 13995
    },
    {
      "entropy": 5.470039415359497,
      "epoch": 1.269956458635704,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004844573995226142,
      "loss": 5.0875,
      "mean_token_accuracy": 0.195917047560215,
      "num_tokens": 29061292.0,
      "step": 14000
    },
    {
      "entropy": 5.417318868637085,
      "epoch": 1.270410014513788,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048444558018952984,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2037899911403656,
      "num_tokens": 29071779.0,
      "step": 14005
    },
    {
      "entropy": 5.4463835716247555,
      "epoch": 1.2708635703918723,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048443375652505754,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20027604550123215,
      "num_tokens": 29082073.0,
      "step": 14010
    },
    {
      "entropy": 5.47108268737793,
      "epoch": 1.2713171262699565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004844219285294418,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20421914309263228,
      "num_tokens": 29092032.0,
      "step": 14015
    },
    {
      "entropy": 5.4538318634033205,
      "epoch": 1.2717706821480406,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048441009620292714,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2062854215502739,
      "num_tokens": 29102936.0,
      "step": 14020
    },
    {
      "entropy": 5.460636281967163,
      "epoch": 1.2722242380261248,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004843982595457584,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20491472631692886,
      "num_tokens": 29113385.0,
      "step": 14025
    },
    {
      "entropy": 5.457720708847046,
      "epoch": 1.2726777939042089,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048438641855818026,
      "loss": 5.0877,
      "mean_token_accuracy": 0.19535764306783676,
      "num_tokens": 29125075.0,
      "step": 14030
    },
    {
      "entropy": 5.503909730911255,
      "epoch": 1.2731313497822931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004843745732404376,
      "loss": 5.0182,
      "mean_token_accuracy": 0.1980168715119362,
      "num_tokens": 29135098.0,
      "step": 14035
    },
    {
      "entropy": 5.545841121673584,
      "epoch": 1.2735849056603774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004843627235927755,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19769282042980194,
      "num_tokens": 29145577.0,
      "step": 14040
    },
    {
      "entropy": 5.463726425170899,
      "epoch": 1.2740384615384617,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004843508696154388,
      "loss": 5.0457,
      "mean_token_accuracy": 0.20328806787729264,
      "num_tokens": 29155905.0,
      "step": 14045
    },
    {
      "entropy": 5.406875324249268,
      "epoch": 1.2744920174165457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048433901130867283,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20879150778055192,
      "num_tokens": 29164926.0,
      "step": 14050
    },
    {
      "entropy": 5.431298923492432,
      "epoch": 1.2749455732946298,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004843271486727228,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2010043293237686,
      "num_tokens": 29175050.0,
      "step": 14055
    },
    {
      "entropy": 5.49884524345398,
      "epoch": 1.275399129172714,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048431528170783395,
      "loss": 5.0853,
      "mean_token_accuracy": 0.19557402282953262,
      "num_tokens": 29186255.0,
      "step": 14060
    },
    {
      "entropy": 5.4883829116821286,
      "epoch": 1.2758526850507983,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004843034104142517,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2044265389442444,
      "num_tokens": 29195639.0,
      "step": 14065
    },
    {
      "entropy": 5.477971744537354,
      "epoch": 1.2763062409288826,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004842915347922216,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1940547063946724,
      "num_tokens": 29206482.0,
      "step": 14070
    },
    {
      "entropy": 5.554836750030518,
      "epoch": 1.2767597968069666,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004842796548419893,
      "loss": 5.0606,
      "mean_token_accuracy": 0.19817731380462647,
      "num_tokens": 29217433.0,
      "step": 14075
    },
    {
      "entropy": 5.548083972930908,
      "epoch": 1.2772133526850509,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048426777056380035,
      "loss": 5.1262,
      "mean_token_accuracy": 0.18795846849679948,
      "num_tokens": 29226193.0,
      "step": 14080
    },
    {
      "entropy": 5.44550051689148,
      "epoch": 1.277666908563135,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004842558819579006,
      "loss": 5.0557,
      "mean_token_accuracy": 0.20124146789312364,
      "num_tokens": 29237061.0,
      "step": 14085
    },
    {
      "entropy": 5.469652271270752,
      "epoch": 1.2781204644412192,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004842439890245358,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20593013614416122,
      "num_tokens": 29247689.0,
      "step": 14090
    },
    {
      "entropy": 5.469670963287354,
      "epoch": 1.2785740203193035,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048423209176395204,
      "loss": 5.018,
      "mean_token_accuracy": 0.2025768503546715,
      "num_tokens": 29258235.0,
      "step": 14095
    },
    {
      "entropy": 5.454458236694336,
      "epoch": 1.2790275761973875,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004842201901763954,
      "loss": 4.9898,
      "mean_token_accuracy": 0.20647548735141755,
      "num_tokens": 29268884.0,
      "step": 14100
    },
    {
      "entropy": 5.475712776184082,
      "epoch": 1.2794811320754718,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004842082842621118,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19233165234327315,
      "num_tokens": 29280484.0,
      "step": 14105
    },
    {
      "entropy": 5.504487562179565,
      "epoch": 1.2799346879535558,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004841963740213476,
      "loss": 5.0951,
      "mean_token_accuracy": 0.199015374481678,
      "num_tokens": 29291199.0,
      "step": 14110
    },
    {
      "entropy": 5.559184598922729,
      "epoch": 1.28038824383164,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004841844594543491,
      "loss": 5.0801,
      "mean_token_accuracy": 0.19844071418046952,
      "num_tokens": 29302007.0,
      "step": 14115
    },
    {
      "entropy": 5.513727045059204,
      "epoch": 1.2808417997097243,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841725405613626,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20648347586393356,
      "num_tokens": 29312127.0,
      "step": 14120
    },
    {
      "entropy": 5.5265700340271,
      "epoch": 1.2812953555878084,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004841606173426347,
      "loss": 5.1555,
      "mean_token_accuracy": 0.19536625295877458,
      "num_tokens": 29323691.0,
      "step": 14125
    },
    {
      "entropy": 5.46632194519043,
      "epoch": 1.2817489114658926,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048414868979841197,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1954666033387184,
      "num_tokens": 29335203.0,
      "step": 14130
    },
    {
      "entropy": 5.524719381332398,
      "epoch": 1.2822024673439767,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048413675792894105,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19794932007789612,
      "num_tokens": 29346422.0,
      "step": 14135
    },
    {
      "entropy": 5.46589846611023,
      "epoch": 1.282656023222061,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004841248217344686,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20230863988399506,
      "num_tokens": 29356084.0,
      "step": 14140
    },
    {
      "entropy": 5.420564365386963,
      "epoch": 1.2831095791001452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841128812152415,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20358027517795563,
      "num_tokens": 29366093.0,
      "step": 14145
    },
    {
      "entropy": 5.505201244354248,
      "epoch": 1.2835631349782293,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841009363715068,
      "loss": 5.0912,
      "mean_token_accuracy": 0.19935815930366516,
      "num_tokens": 29376621.0,
      "step": 14150
    },
    {
      "entropy": 5.556482458114624,
      "epoch": 1.2840166908563135,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004840889872035115,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19459037482738495,
      "num_tokens": 29387119.0,
      "step": 14155
    },
    {
      "entropy": 5.484884357452392,
      "epoch": 1.2844702467343976,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004840770337115026,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20135198682546615,
      "num_tokens": 29396981.0,
      "step": 14160
    },
    {
      "entropy": 5.525591802597046,
      "epoch": 1.2849238026124818,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048406507589572747,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20318254977464675,
      "num_tokens": 29406737.0,
      "step": 14165
    },
    {
      "entropy": 5.500586366653442,
      "epoch": 1.2853773584905661,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004840531137564331,
      "loss": 5.0564,
      "mean_token_accuracy": 0.19997634142637252,
      "num_tokens": 29417300.0,
      "step": 14170
    },
    {
      "entropy": 5.476732492446899,
      "epoch": 1.2858309143686502,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004840411472938672,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2080481082201004,
      "num_tokens": 29426749.0,
      "step": 14175
    },
    {
      "entropy": 5.405505704879761,
      "epoch": 1.2862844702467344,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004840291765082771,
      "loss": 5.0034,
      "mean_token_accuracy": 0.19875230491161347,
      "num_tokens": 29437264.0,
      "step": 14180
    },
    {
      "entropy": 5.474243116378784,
      "epoch": 1.2867380261248185,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048401720139991036,
      "loss": 4.9954,
      "mean_token_accuracy": 0.19903332740068436,
      "num_tokens": 29448192.0,
      "step": 14185
    },
    {
      "entropy": 5.5092840671539305,
      "epoch": 1.2871915820029027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004840052219690146,
      "loss": 4.9105,
      "mean_token_accuracy": 0.20989975184202195,
      "num_tokens": 29457923.0,
      "step": 14190
    },
    {
      "entropy": 5.3856312274932865,
      "epoch": 1.287645137880987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004839932382158376,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20904698222875595,
      "num_tokens": 29466914.0,
      "step": 14195
    },
    {
      "entropy": 5.449931573867798,
      "epoch": 1.288098693759071,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004839812501406272,
      "loss": 5.0846,
      "mean_token_accuracy": 0.1940669134259224,
      "num_tokens": 29477648.0,
      "step": 14200
    },
    {
      "entropy": 5.52697377204895,
      "epoch": 1.2885522496371553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004839692577436313,
      "loss": 5.0756,
      "mean_token_accuracy": 0.2011621966958046,
      "num_tokens": 29487947.0,
      "step": 14205
    },
    {
      "entropy": 5.589422082901001,
      "epoch": 1.2890058055152394,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004839572610250978,
      "loss": 5.1548,
      "mean_token_accuracy": 0.1920746996998787,
      "num_tokens": 29497667.0,
      "step": 14210
    },
    {
      "entropy": 5.475475311279297,
      "epoch": 1.2894593613933236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048394525998527496,
      "loss": 5.0363,
      "mean_token_accuracy": 0.1968516707420349,
      "num_tokens": 29507065.0,
      "step": 14215
    },
    {
      "entropy": 5.54481086730957,
      "epoch": 1.2899129172714079,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004839332546244109,
      "loss": 5.1283,
      "mean_token_accuracy": 0.1999591186642647,
      "num_tokens": 29517752.0,
      "step": 14220
    },
    {
      "entropy": 5.490834951400757,
      "epoch": 1.2903664731494922,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048392124494275385,
      "loss": 5.0457,
      "mean_token_accuracy": 0.19647222459316255,
      "num_tokens": 29527894.0,
      "step": 14225
    },
    {
      "entropy": 5.454130935668945,
      "epoch": 1.2908200290275762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004839092309405523,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19558317065238953,
      "num_tokens": 29538136.0,
      "step": 14230
    },
    {
      "entropy": 5.496419763565063,
      "epoch": 1.2912735849056602,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048389721261805457,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20487109571695328,
      "num_tokens": 29549193.0,
      "step": 14235
    },
    {
      "entropy": 5.461962795257568,
      "epoch": 1.2917271407837445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048388518997550926,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2094683215022087,
      "num_tokens": 29558889.0,
      "step": 14240
    },
    {
      "entropy": 5.4891866683959964,
      "epoch": 1.2921806966618288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048387316301316493,
      "loss": 5.0708,
      "mean_token_accuracy": 0.1955946922302246,
      "num_tokens": 29569691.0,
      "step": 14245
    },
    {
      "entropy": 5.508250522613525,
      "epoch": 1.292634252539913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838611317312705,
      "loss": 5.1483,
      "mean_token_accuracy": 0.19559630900621414,
      "num_tokens": 29580459.0,
      "step": 14250
    },
    {
      "entropy": 5.51726884841919,
      "epoch": 1.293087808417997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004838490961300745,
      "loss": 5.0637,
      "mean_token_accuracy": 0.19881267249584197,
      "num_tokens": 29590652.0,
      "step": 14255
    },
    {
      "entropy": 5.512678384780884,
      "epoch": 1.2935413642960814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048383705620982606,
      "loss": 5.089,
      "mean_token_accuracy": 0.2042189583182335,
      "num_tokens": 29600353.0,
      "step": 14260
    },
    {
      "entropy": 5.496693420410156,
      "epoch": 1.2939949201741654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048382501197077405,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20726697742938996,
      "num_tokens": 29610561.0,
      "step": 14265
    },
    {
      "entropy": 5.438076114654541,
      "epoch": 1.2944484760522497,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004838129634131675,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2033169150352478,
      "num_tokens": 29621741.0,
      "step": 14270
    },
    {
      "entropy": 5.44814715385437,
      "epoch": 1.294902031930334,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004838009105372558,
      "loss": 4.967,
      "mean_token_accuracy": 0.201059290766716,
      "num_tokens": 29633097.0,
      "step": 14275
    },
    {
      "entropy": 5.48220272064209,
      "epoch": 1.295355587808418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048378885334328793,
      "loss": 5.088,
      "mean_token_accuracy": 0.19085302352905273,
      "num_tokens": 29644290.0,
      "step": 14280
    },
    {
      "entropy": 5.585373497009277,
      "epoch": 1.2958091436865022,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004837767918315135,
      "loss": 5.1258,
      "mean_token_accuracy": 0.19204142391681672,
      "num_tokens": 29654977.0,
      "step": 14285
    },
    {
      "entropy": 5.5744030475616455,
      "epoch": 1.2962626995645863,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048376472600218174,
      "loss": 5.1235,
      "mean_token_accuracy": 0.19505043774843217,
      "num_tokens": 29665020.0,
      "step": 14290
    },
    {
      "entropy": 5.475365161895752,
      "epoch": 1.2967162554426706,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004837526558555422,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20274775326251984,
      "num_tokens": 29675098.0,
      "step": 14295
    },
    {
      "entropy": 5.428688049316406,
      "epoch": 1.2971698113207548,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004837405813918446,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2038100093603134,
      "num_tokens": 29685450.0,
      "step": 14300
    },
    {
      "entropy": 5.453429365158081,
      "epoch": 1.2976233671988389,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048372850261133856,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2004016637802124,
      "num_tokens": 29694818.0,
      "step": 14305
    },
    {
      "entropy": 5.5270758152008055,
      "epoch": 1.2980769230769231,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048371641951427393,
      "loss": 5.1108,
      "mean_token_accuracy": 0.19570721834897994,
      "num_tokens": 29705573.0,
      "step": 14310
    },
    {
      "entropy": 5.487916421890259,
      "epoch": 1.2985304789550072,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048370433210090055,
      "loss": 5.0108,
      "mean_token_accuracy": 0.19875363856554032,
      "num_tokens": 29716366.0,
      "step": 14315
    },
    {
      "entropy": 5.469477128982544,
      "epoch": 1.2989840348330914,
      "grad_norm": 1.0,
      "learning_rate": 0.0004836922403714683,
      "loss": 5.042,
      "mean_token_accuracy": 0.19777244478464126,
      "num_tokens": 29727181.0,
      "step": 14320
    },
    {
      "entropy": 5.452441024780273,
      "epoch": 1.2994375907111757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048368014432622746,
      "loss": 5.0523,
      "mean_token_accuracy": 0.20698187947273256,
      "num_tokens": 29736960.0,
      "step": 14325
    },
    {
      "entropy": 5.445215511322021,
      "epoch": 1.2998911465892597,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000483668043965428,
      "loss": 4.9363,
      "mean_token_accuracy": 0.21604397892951965,
      "num_tokens": 29746354.0,
      "step": 14330
    },
    {
      "entropy": 5.471987867355347,
      "epoch": 1.300344702467344,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004836559392893203,
      "loss": 5.0542,
      "mean_token_accuracy": 0.1957583099603653,
      "num_tokens": 29756627.0,
      "step": 14335
    },
    {
      "entropy": 5.464911985397339,
      "epoch": 1.300798258345428,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004836438302981545,
      "loss": 5.0398,
      "mean_token_accuracy": 0.20956512540578842,
      "num_tokens": 29766684.0,
      "step": 14340
    },
    {
      "entropy": 5.470847320556641,
      "epoch": 1.3012518142235123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004836317169921811,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19975050240755082,
      "num_tokens": 29776556.0,
      "step": 14345
    },
    {
      "entropy": 5.553652000427246,
      "epoch": 1.3017053701015966,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048361959937165075,
      "loss": 5.1958,
      "mean_token_accuracy": 0.19425188899040222,
      "num_tokens": 29786896.0,
      "step": 14350
    },
    {
      "entropy": 5.496381521224976,
      "epoch": 1.3021589259796806,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004836074774368138,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2041577160358429,
      "num_tokens": 29797332.0,
      "step": 14355
    },
    {
      "entropy": 5.503427219390869,
      "epoch": 1.302612481857765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048359535118792116,
      "loss": 5.0903,
      "mean_token_accuracy": 0.20044789463281631,
      "num_tokens": 29806428.0,
      "step": 14360
    },
    {
      "entropy": 5.4628736019134525,
      "epoch": 1.303066037735849,
      "grad_norm": 1.0,
      "learning_rate": 0.0004835832206252235,
      "loss": 5.0628,
      "mean_token_accuracy": 0.19847192913293837,
      "num_tokens": 29816794.0,
      "step": 14365
    },
    {
      "entropy": 5.562831878662109,
      "epoch": 1.3035195936139332,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004835710857489716,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20265541970729828,
      "num_tokens": 29827077.0,
      "step": 14370
    },
    {
      "entropy": 5.462686586380005,
      "epoch": 1.3039731494920175,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004835589465594165,
      "loss": 5.0597,
      "mean_token_accuracy": 0.1975301370024681,
      "num_tokens": 29837868.0,
      "step": 14375
    },
    {
      "entropy": 5.480144453048706,
      "epoch": 1.3044267053701015,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048354680305680926,
      "loss": 5.0848,
      "mean_token_accuracy": 0.2043821007013321,
      "num_tokens": 29846929.0,
      "step": 14380
    },
    {
      "entropy": 5.444469165802002,
      "epoch": 1.3048802612481858,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048353465524140105,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20302888005971909,
      "num_tokens": 29857905.0,
      "step": 14385
    },
    {
      "entropy": 5.474423599243164,
      "epoch": 1.3053338171262698,
      "grad_norm": 1.078125,
      "learning_rate": 0.000483522503113443,
      "loss": 5.0751,
      "mean_token_accuracy": 0.1963631957769394,
      "num_tokens": 29868230.0,
      "step": 14390
    },
    {
      "entropy": 5.482108020782471,
      "epoch": 1.305787373004354,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048351034667318634,
      "loss": 5.0315,
      "mean_token_accuracy": 0.20345700830221175,
      "num_tokens": 29879224.0,
      "step": 14395
    },
    {
      "entropy": 5.5577699661254885,
      "epoch": 1.3062409288824384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004834981859208827,
      "loss": 5.1748,
      "mean_token_accuracy": 0.19315745383501054,
      "num_tokens": 29890416.0,
      "step": 14400
    },
    {
      "entropy": 5.52885799407959,
      "epoch": 1.3066944847605226,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048348602085678346,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19536293894052506,
      "num_tokens": 29901064.0,
      "step": 14405
    },
    {
      "entropy": 5.577747058868408,
      "epoch": 1.3071480406386067,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048347385148114004,
      "loss": 5.1489,
      "mean_token_accuracy": 0.19691503047943115,
      "num_tokens": 29911084.0,
      "step": 14410
    },
    {
      "entropy": 5.467166519165039,
      "epoch": 1.3076015965166907,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048346167779420436,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21268735975027084,
      "num_tokens": 29921104.0,
      "step": 14415
    },
    {
      "entropy": 5.474800539016724,
      "epoch": 1.308055152394775,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048344949979622806,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20343236178159713,
      "num_tokens": 29932014.0,
      "step": 14420
    },
    {
      "entropy": 5.46609206199646,
      "epoch": 1.3085087082728593,
      "grad_norm": 1.125,
      "learning_rate": 0.00048343731748746294,
      "loss": 5.0518,
      "mean_token_accuracy": 0.2032724514603615,
      "num_tokens": 29941707.0,
      "step": 14425
    },
    {
      "entropy": 5.475191116333008,
      "epoch": 1.3089622641509435,
      "grad_norm": 1.09375,
      "learning_rate": 0.000483425130868161,
      "loss": 5.0461,
      "mean_token_accuracy": 0.19366600662469863,
      "num_tokens": 29952082.0,
      "step": 14430
    },
    {
      "entropy": 5.422572040557862,
      "epoch": 1.3094158200290276,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004834129399385742,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22088147699832916,
      "num_tokens": 29961744.0,
      "step": 14435
    },
    {
      "entropy": 5.479946565628052,
      "epoch": 1.3098693759071118,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004834007446989548,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2101757764816284,
      "num_tokens": 29972611.0,
      "step": 14440
    },
    {
      "entropy": 5.459532022476196,
      "epoch": 1.3103229317851959,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048338854514955474,
      "loss": 4.9624,
      "mean_token_accuracy": 0.208077135682106,
      "num_tokens": 29982660.0,
      "step": 14445
    },
    {
      "entropy": 5.445099782943726,
      "epoch": 1.3107764876632801,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048337634129062655,
      "loss": 5.0747,
      "mean_token_accuracy": 0.1888832926750183,
      "num_tokens": 29993349.0,
      "step": 14450
    },
    {
      "entropy": 5.562231159210205,
      "epoch": 1.3112300435413644,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048336413312242255,
      "loss": 5.1659,
      "mean_token_accuracy": 0.19416691362857819,
      "num_tokens": 30004474.0,
      "step": 14455
    },
    {
      "entropy": 5.558346843719482,
      "epoch": 1.3116835994194485,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048335192064519514,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19719318449497222,
      "num_tokens": 30015189.0,
      "step": 14460
    },
    {
      "entropy": 5.511405658721924,
      "epoch": 1.3121371552975327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048333970385919695,
      "loss": 4.9939,
      "mean_token_accuracy": 0.19969361424446105,
      "num_tokens": 30025491.0,
      "step": 14465
    },
    {
      "entropy": 5.478708028793335,
      "epoch": 1.3125907111756168,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048332748276468053,
      "loss": 5.0612,
      "mean_token_accuracy": 0.20458794683218,
      "num_tokens": 30034683.0,
      "step": 14470
    },
    {
      "entropy": 5.471246767044067,
      "epoch": 1.313044267053701,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004833152573618987,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20007718205451966,
      "num_tokens": 30044135.0,
      "step": 14475
    },
    {
      "entropy": 5.51534218788147,
      "epoch": 1.3134978229317853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048330302765110436,
      "loss": 5.1094,
      "mean_token_accuracy": 0.2011711835861206,
      "num_tokens": 30054520.0,
      "step": 14480
    },
    {
      "entropy": 5.505444955825806,
      "epoch": 1.3139513788098693,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048329079363255024,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19864712804555892,
      "num_tokens": 30065439.0,
      "step": 14485
    },
    {
      "entropy": 5.502534627914429,
      "epoch": 1.3144049346879536,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048327855530648934,
      "loss": 5.0746,
      "mean_token_accuracy": 0.1971224755048752,
      "num_tokens": 30075362.0,
      "step": 14490
    },
    {
      "entropy": 5.471915912628174,
      "epoch": 1.3148584905660377,
      "grad_norm": 1.015625,
      "learning_rate": 0.000483266312673175,
      "loss": 5.0361,
      "mean_token_accuracy": 0.19733360409736633,
      "num_tokens": 30084913.0,
      "step": 14495
    },
    {
      "entropy": 5.471127033233643,
      "epoch": 1.315312046444122,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004832540657328602,
      "loss": 5.0596,
      "mean_token_accuracy": 0.19881338328123094,
      "num_tokens": 30095704.0,
      "step": 14500
    },
    {
      "entropy": 5.488482666015625,
      "epoch": 1.3157656023222062,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004832418144857982,
      "loss": 5.0117,
      "mean_token_accuracy": 0.20118715465068818,
      "num_tokens": 30105981.0,
      "step": 14505
    },
    {
      "entropy": 5.4902448654174805,
      "epoch": 1.3162191582002902,
      "grad_norm": 1.0,
      "learning_rate": 0.00048322955893224245,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19059209823608397,
      "num_tokens": 30115772.0,
      "step": 14510
    },
    {
      "entropy": 5.492611694335937,
      "epoch": 1.3166727140783745,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048321729907244626,
      "loss": 5.016,
      "mean_token_accuracy": 0.2023007556796074,
      "num_tokens": 30127050.0,
      "step": 14515
    },
    {
      "entropy": 5.49431209564209,
      "epoch": 1.3171262699564585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004832050349066634,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20058691054582595,
      "num_tokens": 30138221.0,
      "step": 14520
    },
    {
      "entropy": 5.533912324905396,
      "epoch": 1.3175798258345428,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004831927664351473,
      "loss": 5.1336,
      "mean_token_accuracy": 0.195409394800663,
      "num_tokens": 30150181.0,
      "step": 14525
    },
    {
      "entropy": 5.490771627426147,
      "epoch": 1.318033381712627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004831804936581518,
      "loss": 5.0948,
      "mean_token_accuracy": 0.20345331132411956,
      "num_tokens": 30160521.0,
      "step": 14530
    },
    {
      "entropy": 5.524169111251831,
      "epoch": 1.3184869375907111,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004831682165759305,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19717661589384078,
      "num_tokens": 30170611.0,
      "step": 14535
    },
    {
      "entropy": 5.526472377777099,
      "epoch": 1.3189404934687954,
      "grad_norm": 1.0,
      "learning_rate": 0.00048315593518873757,
      "loss": 5.0375,
      "mean_token_accuracy": 0.209712053835392,
      "num_tokens": 30179914.0,
      "step": 14540
    },
    {
      "entropy": 5.466052293777466,
      "epoch": 1.3193940493468794,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004831436494968268,
      "loss": 5.0337,
      "mean_token_accuracy": 0.20637008249759675,
      "num_tokens": 30191426.0,
      "step": 14545
    },
    {
      "entropy": 5.4549181938171385,
      "epoch": 1.3198476052249637,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048313135950045234,
      "loss": 5.0122,
      "mean_token_accuracy": 0.19953661859035493,
      "num_tokens": 30202813.0,
      "step": 14550
    },
    {
      "entropy": 5.475877141952514,
      "epoch": 1.320301161103048,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048311906519986824,
      "loss": 4.9444,
      "mean_token_accuracy": 0.20676357299089432,
      "num_tokens": 30213044.0,
      "step": 14555
    },
    {
      "entropy": 5.5254028797149655,
      "epoch": 1.320754716981132,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000483106766595329,
      "loss": 5.0314,
      "mean_token_accuracy": 0.1970205068588257,
      "num_tokens": 30223419.0,
      "step": 14560
    },
    {
      "entropy": 5.553525733947754,
      "epoch": 1.3212082728592163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004830944636870887,
      "loss": 5.0777,
      "mean_token_accuracy": 0.19376981556415557,
      "num_tokens": 30233203.0,
      "step": 14565
    },
    {
      "entropy": 5.487842988967896,
      "epoch": 1.3216618287373003,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048308215647540183,
      "loss": 5.0705,
      "mean_token_accuracy": 0.19618242979049683,
      "num_tokens": 30244604.0,
      "step": 14570
    },
    {
      "entropy": 5.5587037086486815,
      "epoch": 1.3221153846153846,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048306984496052297,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20165631622076036,
      "num_tokens": 30255355.0,
      "step": 14575
    },
    {
      "entropy": 5.577435302734375,
      "epoch": 1.3225689404934688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048305752914270665,
      "loss": 5.1656,
      "mean_token_accuracy": 0.19230901449918747,
      "num_tokens": 30265665.0,
      "step": 14580
    },
    {
      "entropy": 5.564355754852295,
      "epoch": 1.3230224963715531,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048304520902220767,
      "loss": 5.129,
      "mean_token_accuracy": 0.20115920305252075,
      "num_tokens": 30274347.0,
      "step": 14585
    },
    {
      "entropy": 5.51794400215149,
      "epoch": 1.3234760522496372,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004830328845992808,
      "loss": 4.9679,
      "mean_token_accuracy": 0.20817530900239944,
      "num_tokens": 30284243.0,
      "step": 14590
    },
    {
      "entropy": 5.437684631347656,
      "epoch": 1.3239296081277212,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048302055587418074,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20018340796232223,
      "num_tokens": 30294327.0,
      "step": 14595
    },
    {
      "entropy": 5.445873594284057,
      "epoch": 1.3243831640058055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004830082228471626,
      "loss": 5.0842,
      "mean_token_accuracy": 0.19650617390871047,
      "num_tokens": 30304188.0,
      "step": 14600
    },
    {
      "entropy": 5.5108343124389645,
      "epoch": 1.3248367198838897,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048299588551848146,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2086288645863533,
      "num_tokens": 30314917.0,
      "step": 14605
    },
    {
      "entropy": 5.482002687454224,
      "epoch": 1.325290275761974,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048298354388839237,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2028503105044365,
      "num_tokens": 30325221.0,
      "step": 14610
    },
    {
      "entropy": 5.512210655212402,
      "epoch": 1.325743831640058,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004829711979571506,
      "loss": 5.1134,
      "mean_token_accuracy": 0.19776519536972045,
      "num_tokens": 30336196.0,
      "step": 14615
    },
    {
      "entropy": 5.444050025939942,
      "epoch": 1.3261973875181423,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004829588477250115,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21473861038684844,
      "num_tokens": 30345546.0,
      "step": 14620
    },
    {
      "entropy": 5.52281084060669,
      "epoch": 1.3266509433962264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004829464931922304,
      "loss": 5.0554,
      "mean_token_accuracy": 0.1991855889558792,
      "num_tokens": 30355672.0,
      "step": 14625
    },
    {
      "entropy": 5.505038070678711,
      "epoch": 1.3271044992743106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048293413435906285,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20044143497943878,
      "num_tokens": 30366233.0,
      "step": 14630
    },
    {
      "entropy": 5.467237281799316,
      "epoch": 1.327558055152395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004829217712257644,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20227838754653932,
      "num_tokens": 30377366.0,
      "step": 14635
    },
    {
      "entropy": 5.473698282241822,
      "epoch": 1.328011611030479,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048290940379259073,
      "loss": 5.101,
      "mean_token_accuracy": 0.19721851050853728,
      "num_tokens": 30388022.0,
      "step": 14640
    },
    {
      "entropy": 5.483129501342773,
      "epoch": 1.3284651669085632,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048289703205979756,
      "loss": 4.9687,
      "mean_token_accuracy": 0.20514749586582184,
      "num_tokens": 30399235.0,
      "step": 14645
    },
    {
      "entropy": 5.588254690170288,
      "epoch": 1.3289187227866472,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048288465602764084,
      "loss": 5.122,
      "mean_token_accuracy": 0.20071659833192826,
      "num_tokens": 30409112.0,
      "step": 14650
    },
    {
      "entropy": 5.452871370315552,
      "epoch": 1.3293722786647315,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048287227569637645,
      "loss": 5.0219,
      "mean_token_accuracy": 0.2019849956035614,
      "num_tokens": 30420202.0,
      "step": 14655
    },
    {
      "entropy": 5.532520389556884,
      "epoch": 1.3298258345428158,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048285989106626036,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20122217237949372,
      "num_tokens": 30431210.0,
      "step": 14660
    },
    {
      "entropy": 5.497643756866455,
      "epoch": 1.3302793904208998,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004828475021375489,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20784106552600862,
      "num_tokens": 30441291.0,
      "step": 14665
    },
    {
      "entropy": 5.457648134231567,
      "epoch": 1.330732946298984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004828351089104981,
      "loss": 4.964,
      "mean_token_accuracy": 0.20614739060401915,
      "num_tokens": 30450285.0,
      "step": 14670
    },
    {
      "entropy": 5.443894147872925,
      "epoch": 1.3311865021770681,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048282271138536424,
      "loss": 5.0606,
      "mean_token_accuracy": 0.1993221625685692,
      "num_tokens": 30460311.0,
      "step": 14675
    },
    {
      "entropy": 5.560179233551025,
      "epoch": 1.3316400580551524,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004828103095624037,
      "loss": 5.072,
      "mean_token_accuracy": 0.20389872342348098,
      "num_tokens": 30470188.0,
      "step": 14680
    },
    {
      "entropy": 5.470388412475586,
      "epoch": 1.3320936139332367,
      "grad_norm": 0.921875,
      "learning_rate": 0.000482797903441873,
      "loss": 5.024,
      "mean_token_accuracy": 0.1966773748397827,
      "num_tokens": 30481406.0,
      "step": 14685
    },
    {
      "entropy": 5.407244253158569,
      "epoch": 1.3325471698113207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004827854930240288,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20737527012825013,
      "num_tokens": 30492314.0,
      "step": 14690
    },
    {
      "entropy": 5.552218914031982,
      "epoch": 1.333000725689405,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004827730783091276,
      "loss": 5.1477,
      "mean_token_accuracy": 0.19321692138910293,
      "num_tokens": 30503273.0,
      "step": 14695
    },
    {
      "entropy": 5.546647071838379,
      "epoch": 1.333454281567489,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048276065929742616,
      "loss": 5.0477,
      "mean_token_accuracy": 0.1968608483672142,
      "num_tokens": 30513582.0,
      "step": 14700
    },
    {
      "entropy": 5.440297269821167,
      "epoch": 1.3339078374455733,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004827482359891814,
      "loss": 4.97,
      "mean_token_accuracy": 0.21266762018203736,
      "num_tokens": 30523261.0,
      "step": 14705
    },
    {
      "entropy": 5.500866556167603,
      "epoch": 1.3343613933236576,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048273580838465016,
      "loss": 5.0851,
      "mean_token_accuracy": 0.19987517595291138,
      "num_tokens": 30533774.0,
      "step": 14710
    },
    {
      "entropy": 5.502251482009887,
      "epoch": 1.3348149492017416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004827233764840894,
      "loss": 5.1351,
      "mean_token_accuracy": 0.19586722254753114,
      "num_tokens": 30544580.0,
      "step": 14715
    },
    {
      "entropy": 5.525274324417114,
      "epoch": 1.3352685050798259,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004827109402877564,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20630441159009932,
      "num_tokens": 30555105.0,
      "step": 14720
    },
    {
      "entropy": 5.514051389694214,
      "epoch": 1.33572206095791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004826984997959082,
      "loss": 5.1256,
      "mean_token_accuracy": 0.18905919790267944,
      "num_tokens": 30564912.0,
      "step": 14725
    },
    {
      "entropy": 5.463503980636597,
      "epoch": 1.3361756168359942,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000482686055008802,
      "loss": 5.0573,
      "mean_token_accuracy": 0.1945282906293869,
      "num_tokens": 30575862.0,
      "step": 14730
    },
    {
      "entropy": 5.500078582763672,
      "epoch": 1.3366291727140784,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048267360592669535,
      "loss": 5.1625,
      "mean_token_accuracy": 0.19513950496912003,
      "num_tokens": 30586951.0,
      "step": 14735
    },
    {
      "entropy": 5.529357385635376,
      "epoch": 1.3370827285921625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004826611525498456,
      "loss": 5.0766,
      "mean_token_accuracy": 0.1965434119105339,
      "num_tokens": 30597604.0,
      "step": 14740
    },
    {
      "entropy": 5.536353826522827,
      "epoch": 1.3375362844702468,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004826486948785103,
      "loss": 5.1232,
      "mean_token_accuracy": 0.19308859407901763,
      "num_tokens": 30608157.0,
      "step": 14745
    },
    {
      "entropy": 5.521725034713745,
      "epoch": 1.3379898403483308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048263623291294706,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20122882276773452,
      "num_tokens": 30618527.0,
      "step": 14750
    },
    {
      "entropy": 5.541813707351684,
      "epoch": 1.338443396226415,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048262376665341355,
      "loss": 5.1482,
      "mean_token_accuracy": 0.1948135241866112,
      "num_tokens": 30629253.0,
      "step": 14755
    },
    {
      "entropy": 5.508347606658935,
      "epoch": 1.3388969521044993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004826112961001677,
      "loss": 5.1093,
      "mean_token_accuracy": 0.18560464829206466,
      "num_tokens": 30639015.0,
      "step": 14760
    },
    {
      "entropy": 5.557197237014771,
      "epoch": 1.3393505079825834,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048259882125346735,
      "loss": 5.1851,
      "mean_token_accuracy": 0.1878888040781021,
      "num_tokens": 30650035.0,
      "step": 14765
    },
    {
      "entropy": 5.557344198226929,
      "epoch": 1.3398040638606676,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048258634211357045,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1993729680776596,
      "num_tokens": 30659813.0,
      "step": 14770
    },
    {
      "entropy": 5.504110193252563,
      "epoch": 1.3402576197387517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004825738586807351,
      "loss": 5.0459,
      "mean_token_accuracy": 0.1985906630754471,
      "num_tokens": 30670243.0,
      "step": 14775
    },
    {
      "entropy": 5.452295589447021,
      "epoch": 1.340711175616836,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048256137095521945,
      "loss": 5.0215,
      "mean_token_accuracy": 0.1977318599820137,
      "num_tokens": 30680056.0,
      "step": 14780
    },
    {
      "entropy": 5.476433897018433,
      "epoch": 1.3411647314949202,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048254887893728175,
      "loss": 4.9717,
      "mean_token_accuracy": 0.2092755541205406,
      "num_tokens": 30690536.0,
      "step": 14785
    },
    {
      "entropy": 5.443423271179199,
      "epoch": 1.3416182873730045,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004825363826271804,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1987844005227089,
      "num_tokens": 30700200.0,
      "step": 14790
    },
    {
      "entropy": 5.44862470626831,
      "epoch": 1.3420718432510885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004825238820251737,
      "loss": 4.9579,
      "mean_token_accuracy": 0.20584656149148942,
      "num_tokens": 30710576.0,
      "step": 14795
    },
    {
      "entropy": 5.482344055175782,
      "epoch": 1.3425253991291728,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048251137713152027,
      "loss": 5.0585,
      "mean_token_accuracy": 0.2019256144762039,
      "num_tokens": 30721495.0,
      "step": 14800
    },
    {
      "entropy": 5.519159507751465,
      "epoch": 1.3429789550072568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004824988679464787,
      "loss": 5.0938,
      "mean_token_accuracy": 0.1964031144976616,
      "num_tokens": 30731700.0,
      "step": 14805
    },
    {
      "entropy": 5.527659940719604,
      "epoch": 1.343432510885341,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048248635447030767,
      "loss": 5.0728,
      "mean_token_accuracy": 0.19750734865665437,
      "num_tokens": 30743421.0,
      "step": 14810
    },
    {
      "entropy": 5.488374805450439,
      "epoch": 1.3438860667634254,
      "grad_norm": 1.0,
      "learning_rate": 0.0004824738367032659,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2078663155436516,
      "num_tokens": 30754238.0,
      "step": 14815
    },
    {
      "entropy": 5.480378723144531,
      "epoch": 1.3443396226415094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004824613146456123,
      "loss": 5.0715,
      "mean_token_accuracy": 0.19379124492406846,
      "num_tokens": 30763938.0,
      "step": 14820
    },
    {
      "entropy": 5.573088788986206,
      "epoch": 1.3447931785195937,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004824487882976059,
      "loss": 5.1257,
      "mean_token_accuracy": 0.19805540144443512,
      "num_tokens": 30773654.0,
      "step": 14825
    },
    {
      "entropy": 5.551896667480468,
      "epoch": 1.3452467343976777,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048243625765950564,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19453760385513305,
      "num_tokens": 30784659.0,
      "step": 14830
    },
    {
      "entropy": 5.469422769546509,
      "epoch": 1.345700290275762,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004824237227315708,
      "loss": 4.992,
      "mean_token_accuracy": 0.19851766228675843,
      "num_tokens": 30795194.0,
      "step": 14835
    },
    {
      "entropy": 5.410362339019775,
      "epoch": 1.3461538461538463,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004824111835140604,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21054262965917586,
      "num_tokens": 30805623.0,
      "step": 14840
    },
    {
      "entropy": 5.441026401519776,
      "epoch": 1.3466074020319303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000482398640007234,
      "loss": 4.9817,
      "mean_token_accuracy": 0.2019845500588417,
      "num_tokens": 30814899.0,
      "step": 14845
    },
    {
      "entropy": 5.4771339893341064,
      "epoch": 1.3470609579100146,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048238609221135076,
      "loss": 5.1093,
      "mean_token_accuracy": 0.19643137007951736,
      "num_tokens": 30825546.0,
      "step": 14850
    },
    {
      "entropy": 5.455335998535157,
      "epoch": 1.3475145137880986,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004823735401266704,
      "loss": 4.9939,
      "mean_token_accuracy": 0.19967146068811417,
      "num_tokens": 30836406.0,
      "step": 14855
    },
    {
      "entropy": 5.473223638534546,
      "epoch": 1.3479680696661829,
      "grad_norm": 1.0,
      "learning_rate": 0.00048236098375345227,
      "loss": 5.0861,
      "mean_token_accuracy": 0.19803501665592194,
      "num_tokens": 30847966.0,
      "step": 14860
    },
    {
      "entropy": 5.569869422912598,
      "epoch": 1.3484216255442671,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004823484230919563,
      "loss": 5.1078,
      "mean_token_accuracy": 0.19745918065309526,
      "num_tokens": 30856831.0,
      "step": 14865
    },
    {
      "entropy": 5.444964838027954,
      "epoch": 1.3488751814223512,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048233585814244195,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21360531598329544,
      "num_tokens": 30866076.0,
      "step": 14870
    },
    {
      "entropy": 5.4761899471282955,
      "epoch": 1.3493287373004355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004823232889051693,
      "loss": 5.0081,
      "mean_token_accuracy": 0.2073754221200943,
      "num_tokens": 30876845.0,
      "step": 14875
    },
    {
      "entropy": 5.450684928894043,
      "epoch": 1.3497822931785195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048231071538039816,
      "loss": 5.0523,
      "mean_token_accuracy": 0.19669134616851808,
      "num_tokens": 30887448.0,
      "step": 14880
    },
    {
      "entropy": 5.546138429641724,
      "epoch": 1.3502358490566038,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048229813756838863,
      "loss": 5.0474,
      "mean_token_accuracy": 0.2002974957227707,
      "num_tokens": 30898414.0,
      "step": 14885
    },
    {
      "entropy": 5.486542892456055,
      "epoch": 1.350689404934688,
      "grad_norm": 1.0,
      "learning_rate": 0.0004822855554694007,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20294709950685502,
      "num_tokens": 30908559.0,
      "step": 14890
    },
    {
      "entropy": 5.514283227920532,
      "epoch": 1.351142960812772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048227296908369475,
      "loss": 5.0293,
      "mean_token_accuracy": 0.20701266676187516,
      "num_tokens": 30918655.0,
      "step": 14895
    },
    {
      "entropy": 5.453053760528564,
      "epoch": 1.3515965166908563,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048226037841153104,
      "loss": 5.0473,
      "mean_token_accuracy": 0.20324480831623076,
      "num_tokens": 30929188.0,
      "step": 14900
    },
    {
      "entropy": 5.520646667480468,
      "epoch": 1.3520500725689404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004822477834531698,
      "loss": 5.1263,
      "mean_token_accuracy": 0.2014889895915985,
      "num_tokens": 30939902.0,
      "step": 14905
    },
    {
      "entropy": 5.496817684173584,
      "epoch": 1.3525036284470247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004822351842088716,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20315656661987305,
      "num_tokens": 30951289.0,
      "step": 14910
    },
    {
      "entropy": 5.581944751739502,
      "epoch": 1.352957184325109,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004822225806788969,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1878611609339714,
      "num_tokens": 30961233.0,
      "step": 14915
    },
    {
      "entropy": 5.486229324340821,
      "epoch": 1.353410740203193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048220997286350654,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20341320782899858,
      "num_tokens": 30971922.0,
      "step": 14920
    },
    {
      "entropy": 5.460137271881104,
      "epoch": 1.3538642960812772,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048219736076296114,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20320184230804444,
      "num_tokens": 30981358.0,
      "step": 14925
    },
    {
      "entropy": 5.512710285186768,
      "epoch": 1.3543178519593613,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004821847443775214,
      "loss": 5.1231,
      "mean_token_accuracy": 0.19762219935655595,
      "num_tokens": 30991984.0,
      "step": 14930
    },
    {
      "entropy": 5.4981756687164305,
      "epoch": 1.3547714078374455,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048217212370744853,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20261044204235076,
      "num_tokens": 31002790.0,
      "step": 14935
    },
    {
      "entropy": 5.4207916259765625,
      "epoch": 1.3552249637155298,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048215949875300327,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2109668642282486,
      "num_tokens": 31012827.0,
      "step": 14940
    },
    {
      "entropy": 5.430930042266846,
      "epoch": 1.3556785195936139,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004821468695144467,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2006464809179306,
      "num_tokens": 31023075.0,
      "step": 14945
    },
    {
      "entropy": 5.520316028594971,
      "epoch": 1.3561320754716981,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048213423599204015,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20287319719791413,
      "num_tokens": 31032454.0,
      "step": 14950
    },
    {
      "entropy": 5.57982611656189,
      "epoch": 1.3565856313497822,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048212159818604477,
      "loss": 5.1899,
      "mean_token_accuracy": 0.18757683485746385,
      "num_tokens": 31042498.0,
      "step": 14955
    },
    {
      "entropy": 5.469901990890503,
      "epoch": 1.3570391872278664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000482108956096722,
      "loss": 5.0336,
      "mean_token_accuracy": 0.2002289518713951,
      "num_tokens": 31053188.0,
      "step": 14960
    },
    {
      "entropy": 5.497510814666748,
      "epoch": 1.3574927431059507,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048209630972433326,
      "loss": 5.012,
      "mean_token_accuracy": 0.2035995379090309,
      "num_tokens": 31062840.0,
      "step": 14965
    },
    {
      "entropy": 5.561136054992676,
      "epoch": 1.357946298984035,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048208365906914,
      "loss": 5.1582,
      "mean_token_accuracy": 0.1987423777580261,
      "num_tokens": 31072646.0,
      "step": 14970
    },
    {
      "entropy": 5.48033299446106,
      "epoch": 1.358399854862119,
      "grad_norm": 1.125,
      "learning_rate": 0.00048207100413140397,
      "loss": 4.9932,
      "mean_token_accuracy": 0.1994940683245659,
      "num_tokens": 31082700.0,
      "step": 14975
    },
    {
      "entropy": 5.541996240615845,
      "epoch": 1.3588534107402033,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004820583449113868,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19881935864686967,
      "num_tokens": 31093743.0,
      "step": 14980
    },
    {
      "entropy": 5.503304290771484,
      "epoch": 1.3593069666182873,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048204568140935023,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20480051189661025,
      "num_tokens": 31104293.0,
      "step": 14985
    },
    {
      "entropy": 5.443846797943115,
      "epoch": 1.3597605224963716,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004820330136255563,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2066284880042076,
      "num_tokens": 31114244.0,
      "step": 14990
    },
    {
      "entropy": 5.46366810798645,
      "epoch": 1.3602140783744558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048202034156026683,
      "loss": 5.0191,
      "mean_token_accuracy": 0.19951775372028352,
      "num_tokens": 31123596.0,
      "step": 14995
    },
    {
      "entropy": 5.5296409130096436,
      "epoch": 1.36066763425254,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048200766521374396,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1923889234662056,
      "num_tokens": 31134258.0,
      "step": 15000
    },
    {
      "epoch": 1.36066763425254,
      "eval_entropy": 5.295666738009561,
      "eval_loss": 5.112661838531494,
      "eval_mean_token_accuracy": 0.2063279388017312,
      "eval_num_tokens": 31134258.0,
      "eval_runtime": 20.7071,
      "eval_samples_per_second": 1707.628,
      "eval_steps_per_second": 213.453,
      "step": 15000
    },
    {
      "entropy": 5.491491031646729,
      "epoch": 1.3611211901306242,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004819949845862498,
      "loss": 5.0584,
      "mean_token_accuracy": 0.2033733531832695,
      "num_tokens": 31144806.0,
      "step": 15005
    },
    {
      "entropy": 5.542679929733277,
      "epoch": 1.3615747460087082,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004819822996780466,
      "loss": 5.1565,
      "mean_token_accuracy": 0.1965965673327446,
      "num_tokens": 31155642.0,
      "step": 15010
    },
    {
      "entropy": 5.541298770904541,
      "epoch": 1.3620283018867925,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004819696104893967,
      "loss": 4.9988,
      "mean_token_accuracy": 0.20615871697664262,
      "num_tokens": 31165034.0,
      "step": 15015
    },
    {
      "entropy": 5.529401683807373,
      "epoch": 1.3624818577648767,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004819569170205625,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1988907501101494,
      "num_tokens": 31176491.0,
      "step": 15020
    },
    {
      "entropy": 5.515815877914429,
      "epoch": 1.3629354136429608,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048194421927180654,
      "loss": 5.0767,
      "mean_token_accuracy": 0.1967637598514557,
      "num_tokens": 31186198.0,
      "step": 15025
    },
    {
      "entropy": 5.515884733200073,
      "epoch": 1.363388969521045,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004819315172433914,
      "loss": 5.0597,
      "mean_token_accuracy": 0.19971127957105636,
      "num_tokens": 31197094.0,
      "step": 15030
    },
    {
      "entropy": 5.552103185653687,
      "epoch": 1.363842525399129,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048191881093557977,
      "loss": 5.0781,
      "mean_token_accuracy": 0.19090546071529388,
      "num_tokens": 31207505.0,
      "step": 15035
    },
    {
      "entropy": 5.576592445373535,
      "epoch": 1.3642960812772134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048190610034863434,
      "loss": 5.1897,
      "mean_token_accuracy": 0.18819540590047837,
      "num_tokens": 31218809.0,
      "step": 15040
    },
    {
      "entropy": 5.542851114273072,
      "epoch": 1.3647496371552976,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004818933854828181,
      "loss": 5.1072,
      "mean_token_accuracy": 0.19087555557489394,
      "num_tokens": 31230281.0,
      "step": 15045
    },
    {
      "entropy": 5.482969665527344,
      "epoch": 1.3652031930333817,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004818806663383939,
      "loss": 4.992,
      "mean_token_accuracy": 0.20569752156734467,
      "num_tokens": 31239573.0,
      "step": 15050
    },
    {
      "entropy": 5.471133422851563,
      "epoch": 1.365656748911466,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004818679429156248,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2037968084216118,
      "num_tokens": 31249671.0,
      "step": 15055
    },
    {
      "entropy": 5.448171234130859,
      "epoch": 1.36611030478955,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048185521521477385,
      "loss": 5.0496,
      "mean_token_accuracy": 0.19569021314382554,
      "num_tokens": 31259601.0,
      "step": 15060
    },
    {
      "entropy": 5.488245248794556,
      "epoch": 1.3665638606676342,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004818424832361044,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2052937850356102,
      "num_tokens": 31269085.0,
      "step": 15065
    },
    {
      "entropy": 5.454237413406372,
      "epoch": 1.3670174165457185,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004818297469798797,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19666333496570587,
      "num_tokens": 31280177.0,
      "step": 15070
    },
    {
      "entropy": 5.516110610961914,
      "epoch": 1.3674709724238026,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048181700644636314,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20633066445589066,
      "num_tokens": 31290545.0,
      "step": 15075
    },
    {
      "entropy": 5.434341859817505,
      "epoch": 1.3679245283018868,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048180426163581816,
      "loss": 4.9031,
      "mean_token_accuracy": 0.20979585647583007,
      "num_tokens": 31301143.0,
      "step": 15080
    },
    {
      "entropy": 5.4555473804473875,
      "epoch": 1.3683780841799709,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004817915125485083,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20275210589170456,
      "num_tokens": 31311071.0,
      "step": 15085
    },
    {
      "entropy": 5.551762104034424,
      "epoch": 1.3688316400580551,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048177875918469735,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19941299557685851,
      "num_tokens": 31322263.0,
      "step": 15090
    },
    {
      "entropy": 5.529943132400513,
      "epoch": 1.3692851959361394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048176600154464897,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1998817577958107,
      "num_tokens": 31332975.0,
      "step": 15095
    },
    {
      "entropy": 5.581736898422241,
      "epoch": 1.3697387518142234,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004817532396286269,
      "loss": 5.1303,
      "mean_token_accuracy": 0.1915099710226059,
      "num_tokens": 31345252.0,
      "step": 15100
    },
    {
      "entropy": 5.494141530990601,
      "epoch": 1.3701923076923077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004817404734368952,
      "loss": 4.9906,
      "mean_token_accuracy": 0.20686235874891282,
      "num_tokens": 31355479.0,
      "step": 15105
    },
    {
      "entropy": 5.5393835544586185,
      "epoch": 1.3706458635703918,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048172770296971785,
      "loss": 5.136,
      "mean_token_accuracy": 0.19656438529491424,
      "num_tokens": 31366020.0,
      "step": 15110
    },
    {
      "entropy": 5.513412237167358,
      "epoch": 1.371099419448476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004817149282273589,
      "loss": 5.046,
      "mean_token_accuracy": 0.1999151214957237,
      "num_tokens": 31376055.0,
      "step": 15115
    },
    {
      "entropy": 5.548714685440063,
      "epoch": 1.3715529753265603,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004817021492100825,
      "loss": 5.178,
      "mean_token_accuracy": 0.18876050859689714,
      "num_tokens": 31386179.0,
      "step": 15120
    },
    {
      "entropy": 5.466830825805664,
      "epoch": 1.3720065312046443,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000481689365918153,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20302161574363708,
      "num_tokens": 31395850.0,
      "step": 15125
    },
    {
      "entropy": 5.502423667907715,
      "epoch": 1.3724600870827286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004816765783518348,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21120330691337585,
      "num_tokens": 31405053.0,
      "step": 15130
    },
    {
      "entropy": 5.439601993560791,
      "epoch": 1.3729136429608126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004816637865113923,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20858091861009598,
      "num_tokens": 31415153.0,
      "step": 15135
    },
    {
      "entropy": 5.446499586105347,
      "epoch": 1.373367198838897,
      "grad_norm": 1.0,
      "learning_rate": 0.00048165099039708994,
      "loss": 5.0592,
      "mean_token_accuracy": 0.19852253049612045,
      "num_tokens": 31425451.0,
      "step": 15140
    },
    {
      "entropy": 5.445763492584229,
      "epoch": 1.3738207547169812,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004816381900091925,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20059703588485717,
      "num_tokens": 31436298.0,
      "step": 15145
    },
    {
      "entropy": 5.520010280609131,
      "epoch": 1.3742743105950654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048162538534796465,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20523983836174012,
      "num_tokens": 31446932.0,
      "step": 15150
    },
    {
      "entropy": 5.551976919174194,
      "epoch": 1.3747278664731495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004816125764136711,
      "loss": 5.0848,
      "mean_token_accuracy": 0.19165421277284622,
      "num_tokens": 31457582.0,
      "step": 15155
    },
    {
      "entropy": 5.560914659500122,
      "epoch": 1.3751814223512338,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004815997632065769,
      "loss": 5.0433,
      "mean_token_accuracy": 0.201973794400692,
      "num_tokens": 31468168.0,
      "step": 15160
    },
    {
      "entropy": 5.4446817398071286,
      "epoch": 1.3756349782293178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048158694572694694,
      "loss": 5.0329,
      "mean_token_accuracy": 0.207179257273674,
      "num_tokens": 31478524.0,
      "step": 15165
    },
    {
      "entropy": 5.456293964385987,
      "epoch": 1.376088534107402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048157412397504623,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20530054569244385,
      "num_tokens": 31488608.0,
      "step": 15170
    },
    {
      "entropy": 5.493011617660523,
      "epoch": 1.3765420899854863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004815612979511401,
      "loss": 5.0507,
      "mean_token_accuracy": 0.20859668999910355,
      "num_tokens": 31498984.0,
      "step": 15175
    },
    {
      "entropy": 5.5580322265625,
      "epoch": 1.3769956458635704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048154846765549353,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20434010922908782,
      "num_tokens": 31509110.0,
      "step": 15180
    },
    {
      "entropy": 5.479401540756226,
      "epoch": 1.3774492017416546,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048153563308837215,
      "loss": 5.0164,
      "mean_token_accuracy": 0.20211873799562455,
      "num_tokens": 31519266.0,
      "step": 15185
    },
    {
      "entropy": 5.491009759902954,
      "epoch": 1.3779027576197387,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048152279425004124,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20235521346330643,
      "num_tokens": 31529043.0,
      "step": 15190
    },
    {
      "entropy": 5.520250082015991,
      "epoch": 1.378356313497823,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004815099511407663,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20950424671173096,
      "num_tokens": 31538650.0,
      "step": 15195
    },
    {
      "entropy": 5.458363056182861,
      "epoch": 1.3788098693759072,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000481497103760813,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20308187007904052,
      "num_tokens": 31549392.0,
      "step": 15200
    },
    {
      "entropy": 5.550771903991699,
      "epoch": 1.3792634252539913,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048148425211044684,
      "loss": 5.1663,
      "mean_token_accuracy": 0.18810785859823226,
      "num_tokens": 31559999.0,
      "step": 15205
    },
    {
      "entropy": 5.559471702575683,
      "epoch": 1.3797169811320755,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004814713961899338,
      "loss": 5.0556,
      "mean_token_accuracy": 0.1991157650947571,
      "num_tokens": 31570108.0,
      "step": 15210
    },
    {
      "entropy": 5.515550661087036,
      "epoch": 1.3801705370101596,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048145853599953966,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20343615710735322,
      "num_tokens": 31579698.0,
      "step": 15215
    },
    {
      "entropy": 5.457998752593994,
      "epoch": 1.3806240928882438,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048144567153953036,
      "loss": 5.0727,
      "mean_token_accuracy": 0.19539187401533126,
      "num_tokens": 31589292.0,
      "step": 15220
    },
    {
      "entropy": 5.559013748168946,
      "epoch": 1.381077648766328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048143280281017197,
      "loss": 5.1465,
      "mean_token_accuracy": 0.19716491550207138,
      "num_tokens": 31599848.0,
      "step": 15225
    },
    {
      "entropy": 5.541344213485718,
      "epoch": 1.3815312046444121,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004814199298117307,
      "loss": 5.0811,
      "mean_token_accuracy": 0.19916218519210815,
      "num_tokens": 31611109.0,
      "step": 15230
    },
    {
      "entropy": 5.4703916072845455,
      "epoch": 1.3819847605224964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048140705254447256,
      "loss": 4.9954,
      "mean_token_accuracy": 0.203843155503273,
      "num_tokens": 31621140.0,
      "step": 15235
    },
    {
      "entropy": 5.583371543884278,
      "epoch": 1.3824383164005805,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048139417100866405,
      "loss": 5.0945,
      "mean_token_accuracy": 0.19653286635875702,
      "num_tokens": 31631268.0,
      "step": 15240
    },
    {
      "entropy": 5.551007652282715,
      "epoch": 1.3828918722786647,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048138128520457146,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20167967230081557,
      "num_tokens": 31640758.0,
      "step": 15245
    },
    {
      "entropy": 5.47036771774292,
      "epoch": 1.383345428156749,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004813683951324613,
      "loss": 5.0601,
      "mean_token_accuracy": 0.19761260747909545,
      "num_tokens": 31651060.0,
      "step": 15250
    },
    {
      "entropy": 5.516869258880615,
      "epoch": 1.383798984034833,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048135550079260014,
      "loss": 5.134,
      "mean_token_accuracy": 0.20071062743663787,
      "num_tokens": 31662282.0,
      "step": 15255
    },
    {
      "entropy": 5.504281425476075,
      "epoch": 1.3842525399129173,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004813426021852547,
      "loss": 5.0203,
      "mean_token_accuracy": 0.19429908096790313,
      "num_tokens": 31672364.0,
      "step": 15260
    },
    {
      "entropy": 5.517607593536377,
      "epoch": 1.3847060957910013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048132969931069156,
      "loss": 5.0738,
      "mean_token_accuracy": 0.1958681508898735,
      "num_tokens": 31683094.0,
      "step": 15265
    },
    {
      "entropy": 5.505269622802734,
      "epoch": 1.3851596516690856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004813167921691777,
      "loss": 5.0391,
      "mean_token_accuracy": 0.2022292897105217,
      "num_tokens": 31693087.0,
      "step": 15270
    },
    {
      "entropy": 5.5196010112762455,
      "epoch": 1.3856132075471699,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048130388076098003,
      "loss": 5.0301,
      "mean_token_accuracy": 0.19693872481584548,
      "num_tokens": 31703708.0,
      "step": 15275
    },
    {
      "entropy": 5.46891303062439,
      "epoch": 1.386066763425254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048129096508636547,
      "loss": 4.9565,
      "mean_token_accuracy": 0.211345899105072,
      "num_tokens": 31713640.0,
      "step": 15280
    },
    {
      "entropy": 5.512295389175415,
      "epoch": 1.3865203193033382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048127804514560126,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19729174673557281,
      "num_tokens": 31723526.0,
      "step": 15285
    },
    {
      "entropy": 5.444248962402344,
      "epoch": 1.3869738751814222,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048126512093895447,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20083393156528473,
      "num_tokens": 31733265.0,
      "step": 15290
    },
    {
      "entropy": 5.519571924209595,
      "epoch": 1.3874274310595065,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004812521924666925,
      "loss": 4.9972,
      "mean_token_accuracy": 0.20516951829195024,
      "num_tokens": 31743832.0,
      "step": 15295
    },
    {
      "entropy": 5.533500385284424,
      "epoch": 1.3878809869375908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004812392597290825,
      "loss": 5.0691,
      "mean_token_accuracy": 0.2090713858604431,
      "num_tokens": 31754443.0,
      "step": 15300
    },
    {
      "entropy": 5.4089892387390135,
      "epoch": 1.3883345428156748,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004812263227263921,
      "loss": 5.019,
      "mean_token_accuracy": 0.20233907401561738,
      "num_tokens": 31765304.0,
      "step": 15305
    },
    {
      "entropy": 5.474977779388428,
      "epoch": 1.388788098693759,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004812133814588888,
      "loss": 5.0712,
      "mean_token_accuracy": 0.19324804544448854,
      "num_tokens": 31775809.0,
      "step": 15310
    },
    {
      "entropy": 5.53752875328064,
      "epoch": 1.3892416545718431,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004812004359268402,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20598687529563903,
      "num_tokens": 31786271.0,
      "step": 15315
    },
    {
      "entropy": 5.527387714385986,
      "epoch": 1.3896952104499274,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004811874861305141,
      "loss": 5.0821,
      "mean_token_accuracy": 0.20123516917228698,
      "num_tokens": 31797358.0,
      "step": 15320
    },
    {
      "entropy": 5.475525903701782,
      "epoch": 1.3901487663280117,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004811745320701782,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20617760568857194,
      "num_tokens": 31808303.0,
      "step": 15325
    },
    {
      "entropy": 5.433736848831177,
      "epoch": 1.390602322206096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004811615737461004,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20380660891532898,
      "num_tokens": 31818837.0,
      "step": 15330
    },
    {
      "entropy": 5.55443115234375,
      "epoch": 1.39105587808418,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048114861115854877,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1971920147538185,
      "num_tokens": 31829729.0,
      "step": 15335
    },
    {
      "entropy": 5.499588584899902,
      "epoch": 1.3915094339622642,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048113564430779136,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20167743265628815,
      "num_tokens": 31839996.0,
      "step": 15340
    },
    {
      "entropy": 5.401311779022217,
      "epoch": 1.3919629898403483,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004811226731940962,
      "loss": 4.9661,
      "mean_token_accuracy": 0.20950472056865693,
      "num_tokens": 31850298.0,
      "step": 15345
    },
    {
      "entropy": 5.48508653640747,
      "epoch": 1.3924165457184325,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048110969781773173,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20816129744052886,
      "num_tokens": 31862170.0,
      "step": 15350
    },
    {
      "entropy": 5.5605326175689695,
      "epoch": 1.3928701015965168,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048109671817896614,
      "loss": 5.1177,
      "mean_token_accuracy": 0.20055409371852875,
      "num_tokens": 31872081.0,
      "step": 15355
    },
    {
      "entropy": 5.497061347961425,
      "epoch": 1.3933236574746009,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048108373427806797,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2065151587128639,
      "num_tokens": 31882715.0,
      "step": 15360
    },
    {
      "entropy": 5.526398038864135,
      "epoch": 1.3937772133526851,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048107074611530555,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20854706615209578,
      "num_tokens": 31891925.0,
      "step": 15365
    },
    {
      "entropy": 5.541769695281983,
      "epoch": 1.3942307692307692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048105775369094756,
      "loss": 5.0347,
      "mean_token_accuracy": 0.20411615669727326,
      "num_tokens": 31902320.0,
      "step": 15370
    },
    {
      "entropy": 5.45694899559021,
      "epoch": 1.3946843251088534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004810447570052628,
      "loss": 4.9559,
      "mean_token_accuracy": 0.20837390571832656,
      "num_tokens": 31912726.0,
      "step": 15375
    },
    {
      "entropy": 5.457587814331054,
      "epoch": 1.3951378809869377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048103175605851997,
      "loss": 5.0727,
      "mean_token_accuracy": 0.19942245334386827,
      "num_tokens": 31924039.0,
      "step": 15380
    },
    {
      "entropy": 5.43327054977417,
      "epoch": 1.3955914368650217,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004810187508509879,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2109137788414955,
      "num_tokens": 31933280.0,
      "step": 15385
    },
    {
      "entropy": 5.482158946990967,
      "epoch": 1.396044992743106,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004810057413829355,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19854603260755538,
      "num_tokens": 31943182.0,
      "step": 15390
    },
    {
      "entropy": 5.606261444091797,
      "epoch": 1.39649854862119,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048099272765463193,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20472536087036133,
      "num_tokens": 31953256.0,
      "step": 15395
    },
    {
      "entropy": 5.608113241195679,
      "epoch": 1.3969521044992743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004809797096663463,
      "loss": 5.1651,
      "mean_token_accuracy": 0.19737764447927475,
      "num_tokens": 31964231.0,
      "step": 15400
    },
    {
      "entropy": 5.514317989349365,
      "epoch": 1.3974056603773586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004809666874183477,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20222418010234833,
      "num_tokens": 31974322.0,
      "step": 15405
    },
    {
      "entropy": 5.4731401920318605,
      "epoch": 1.3978592162554426,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004809536609109056,
      "loss": 5.0691,
      "mean_token_accuracy": 0.2095978870987892,
      "num_tokens": 31984239.0,
      "step": 15410
    },
    {
      "entropy": 5.476027631759644,
      "epoch": 1.398312772133527,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048094063014428925,
      "loss": 5.0785,
      "mean_token_accuracy": 0.19442455768585204,
      "num_tokens": 31994366.0,
      "step": 15415
    },
    {
      "entropy": 5.566170167922974,
      "epoch": 1.398766328011611,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004809275951187682,
      "loss": 5.16,
      "mean_token_accuracy": 0.19522467702627183,
      "num_tokens": 32004921.0,
      "step": 15420
    },
    {
      "entropy": 5.466505813598633,
      "epoch": 1.3992198838896952,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048091455583461204,
      "loss": 5.0773,
      "mean_token_accuracy": 0.19962003231048583,
      "num_tokens": 32016385.0,
      "step": 15425
    },
    {
      "entropy": 5.520288276672363,
      "epoch": 1.3996734397677795,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048090151229209036,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20249414145946504,
      "num_tokens": 32027398.0,
      "step": 15430
    },
    {
      "entropy": 5.5420714855194095,
      "epoch": 1.4001269956458635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004808884644914729,
      "loss": 5.0636,
      "mean_token_accuracy": 0.20085171461105347,
      "num_tokens": 32038326.0,
      "step": 15435
    },
    {
      "entropy": 5.516499710083008,
      "epoch": 1.4005805515239478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004808754124330296,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19790878891944885,
      "num_tokens": 32048744.0,
      "step": 15440
    },
    {
      "entropy": 5.500717973709106,
      "epoch": 1.4010341074020318,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004808623561170302,
      "loss": 4.9933,
      "mean_token_accuracy": 0.20921683311462402,
      "num_tokens": 32059133.0,
      "step": 15445
    },
    {
      "entropy": 5.502972602844238,
      "epoch": 1.401487663280116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004808492955437449,
      "loss": 5.0357,
      "mean_token_accuracy": 0.20092233568429946,
      "num_tokens": 32069277.0,
      "step": 15450
    },
    {
      "entropy": 5.505456733703613,
      "epoch": 1.4019412191582004,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048083623071344374,
      "loss": 4.9932,
      "mean_token_accuracy": 0.20528948158025742,
      "num_tokens": 32079802.0,
      "step": 15455
    },
    {
      "entropy": 5.508525562286377,
      "epoch": 1.4023947750362844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048082316162639685,
      "loss": 5.1042,
      "mean_token_accuracy": 0.1987621620297432,
      "num_tokens": 32090962.0,
      "step": 15460
    },
    {
      "entropy": 5.478303909301758,
      "epoch": 1.4028483309143687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048081008828287447,
      "loss": 5.077,
      "mean_token_accuracy": 0.20106507390737532,
      "num_tokens": 32101058.0,
      "step": 15465
    },
    {
      "entropy": 5.55407304763794,
      "epoch": 1.4033018867924527,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048079701068314706,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20511925965547562,
      "num_tokens": 32111763.0,
      "step": 15470
    },
    {
      "entropy": 5.53937177658081,
      "epoch": 1.403755442670537,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048078392882748507,
      "loss": 5.1272,
      "mean_token_accuracy": 0.2001139998435974,
      "num_tokens": 32123556.0,
      "step": 15475
    },
    {
      "entropy": 5.495906591415405,
      "epoch": 1.4042089985486212,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048077084271615895,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20298214703798295,
      "num_tokens": 32134012.0,
      "step": 15480
    },
    {
      "entropy": 5.424602937698364,
      "epoch": 1.4046625544267053,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004807577523494394,
      "loss": 5.0046,
      "mean_token_accuracy": 0.2091123193502426,
      "num_tokens": 32144540.0,
      "step": 15485
    },
    {
      "entropy": 5.516141748428344,
      "epoch": 1.4051161103047896,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048074465772759707,
      "loss": 5.0977,
      "mean_token_accuracy": 0.20051607191562654,
      "num_tokens": 32154655.0,
      "step": 15490
    },
    {
      "entropy": 5.478973913192749,
      "epoch": 1.4055696661828736,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048073155885090276,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20102933645248414,
      "num_tokens": 32165191.0,
      "step": 15495
    },
    {
      "entropy": 5.515355730056763,
      "epoch": 1.4060232220609579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048071845571962733,
      "loss": 5.0933,
      "mean_token_accuracy": 0.1930096685886383,
      "num_tokens": 32175961.0,
      "step": 15500
    },
    {
      "entropy": 5.549034500122071,
      "epoch": 1.4064767779390421,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004807053483340419,
      "loss": 5.0587,
      "mean_token_accuracy": 0.2007862851023674,
      "num_tokens": 32187073.0,
      "step": 15505
    },
    {
      "entropy": 5.495439291000366,
      "epoch": 1.4069303338171264,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004806922366944175,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19969787448644638,
      "num_tokens": 32197426.0,
      "step": 15510
    },
    {
      "entropy": 5.536189460754395,
      "epoch": 1.4073838896952104,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048067912080102513,
      "loss": 5.032,
      "mean_token_accuracy": 0.2082691878080368,
      "num_tokens": 32206847.0,
      "step": 15515
    },
    {
      "entropy": 5.478679609298706,
      "epoch": 1.4078374455732947,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004806660006541362,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1986638382077217,
      "num_tokens": 32217647.0,
      "step": 15520
    },
    {
      "entropy": 5.412083864212036,
      "epoch": 1.4082910014513788,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004806528762540219,
      "loss": 4.9054,
      "mean_token_accuracy": 0.21444422006607056,
      "num_tokens": 32227133.0,
      "step": 15525
    },
    {
      "entropy": 5.521481990814209,
      "epoch": 1.408744557329463,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048063974760095374,
      "loss": 5.1197,
      "mean_token_accuracy": 0.19907931387424468,
      "num_tokens": 32237378.0,
      "step": 15530
    },
    {
      "entropy": 5.450902986526489,
      "epoch": 1.4091981132075473,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048062661469520327,
      "loss": 4.9959,
      "mean_token_accuracy": 0.2066905677318573,
      "num_tokens": 32247213.0,
      "step": 15535
    },
    {
      "entropy": 5.555349111557007,
      "epoch": 1.4096516690856313,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048061347753704193,
      "loss": 5.0782,
      "mean_token_accuracy": 0.2022297203540802,
      "num_tokens": 32257100.0,
      "step": 15540
    },
    {
      "entropy": 5.5293927669525145,
      "epoch": 1.4101052249637156,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004806003361267415,
      "loss": 5.0423,
      "mean_token_accuracy": 0.1960933342576027,
      "num_tokens": 32267456.0,
      "step": 15545
    },
    {
      "entropy": 5.456585788726807,
      "epoch": 1.4105587808417996,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004805871904645737,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2102513313293457,
      "num_tokens": 32278220.0,
      "step": 15550
    },
    {
      "entropy": 5.444500207901001,
      "epoch": 1.411012336719884,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004805740405508103,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2062092110514641,
      "num_tokens": 32289036.0,
      "step": 15555
    },
    {
      "entropy": 5.468543577194214,
      "epoch": 1.4114658925979682,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004805608863857236,
      "loss": 4.9818,
      "mean_token_accuracy": 0.19995850175619126,
      "num_tokens": 32298909.0,
      "step": 15560
    },
    {
      "entropy": 5.513676548004151,
      "epoch": 1.4119194484760522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048054772796958517,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20487444698810578,
      "num_tokens": 32309851.0,
      "step": 15565
    },
    {
      "entropy": 5.464004564285278,
      "epoch": 1.4123730043541365,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004805345653026675,
      "loss": 4.99,
      "mean_token_accuracy": 0.21017162203788758,
      "num_tokens": 32319336.0,
      "step": 15570
    },
    {
      "entropy": 5.510065793991089,
      "epoch": 1.4128265602322205,
      "grad_norm": 1.0,
      "learning_rate": 0.00048052139838524263,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20483591854572297,
      "num_tokens": 32329898.0,
      "step": 15575
    },
    {
      "entropy": 5.492889308929444,
      "epoch": 1.4132801161103048,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048050822721758274,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20117055475711823,
      "num_tokens": 32340100.0,
      "step": 15580
    },
    {
      "entropy": 5.4790565967559814,
      "epoch": 1.413733671988389,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004804950517999604,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19834206104278565,
      "num_tokens": 32349663.0,
      "step": 15585
    },
    {
      "entropy": 5.524758529663086,
      "epoch": 1.414187227866473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004804818721326481,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20059896111488343,
      "num_tokens": 32360461.0,
      "step": 15590
    },
    {
      "entropy": 5.545775079727173,
      "epoch": 1.4146407837445574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004804686882159183,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20205838233232498,
      "num_tokens": 32370910.0,
      "step": 15595
    },
    {
      "entropy": 5.578698396682739,
      "epoch": 1.4150943396226414,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048045550005004355,
      "loss": 5.0914,
      "mean_token_accuracy": 0.20864028632640838,
      "num_tokens": 32381669.0,
      "step": 15600
    },
    {
      "entropy": 5.462166213989258,
      "epoch": 1.4155478955007257,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004804423076352968,
      "loss": 5.0099,
      "mean_token_accuracy": 0.19899781793355942,
      "num_tokens": 32392024.0,
      "step": 15605
    },
    {
      "entropy": 5.416386508941651,
      "epoch": 1.41600145137881,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048042911097195075,
      "loss": 5.0229,
      "mean_token_accuracy": 0.20628166496753692,
      "num_tokens": 32401872.0,
      "step": 15610
    },
    {
      "entropy": 5.426913595199585,
      "epoch": 1.416455007256894,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004804159100602784,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2084158256649971,
      "num_tokens": 32411749.0,
      "step": 15615
    },
    {
      "entropy": 5.584346342086792,
      "epoch": 1.4169085631349783,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048040270490055264,
      "loss": 5.1609,
      "mean_token_accuracy": 0.19357605278491974,
      "num_tokens": 32421358.0,
      "step": 15620
    },
    {
      "entropy": 5.543413543701172,
      "epoch": 1.4173621190130623,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004803894954930465,
      "loss": 5.055,
      "mean_token_accuracy": 0.19952864944934845,
      "num_tokens": 32432322.0,
      "step": 15625
    },
    {
      "entropy": 5.500397825241089,
      "epoch": 1.4178156748911466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004803762818380334,
      "loss": 5.0745,
      "mean_token_accuracy": 0.19714537113904954,
      "num_tokens": 32440930.0,
      "step": 15630
    },
    {
      "entropy": 5.511843538284301,
      "epoch": 1.4182692307692308,
      "grad_norm": 1.0,
      "learning_rate": 0.00048036306393578644,
      "loss": 5.0697,
      "mean_token_accuracy": 0.19900647252798082,
      "num_tokens": 32451124.0,
      "step": 15635
    },
    {
      "entropy": 5.549473667144776,
      "epoch": 1.4187227866473149,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004803498417865789,
      "loss": 5.0478,
      "mean_token_accuracy": 0.19772396683692933,
      "num_tokens": 32460636.0,
      "step": 15640
    },
    {
      "entropy": 5.521689081192017,
      "epoch": 1.4191763425253991,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004803366153906844,
      "loss": 5.0847,
      "mean_token_accuracy": 0.19507592916488647,
      "num_tokens": 32470501.0,
      "step": 15645
    },
    {
      "entropy": 5.53672423362732,
      "epoch": 1.4196298984034832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004803233847483763,
      "loss": 5.1524,
      "mean_token_accuracy": 0.19808802306652068,
      "num_tokens": 32481559.0,
      "step": 15650
    },
    {
      "entropy": 5.6585461616516115,
      "epoch": 1.4200834542815675,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048031014985992825,
      "loss": 5.2276,
      "mean_token_accuracy": 0.19318990856409074,
      "num_tokens": 32493000.0,
      "step": 15655
    },
    {
      "entropy": 5.520326471328735,
      "epoch": 1.4205370101596517,
      "grad_norm": 1.03125,
      "learning_rate": 0.000480296910725614,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21055828481912614,
      "num_tokens": 32502773.0,
      "step": 15660
    },
    {
      "entropy": 5.480014324188232,
      "epoch": 1.4209905660377358,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004802836673457074,
      "loss": 5.1155,
      "mean_token_accuracy": 0.19522669911384583,
      "num_tokens": 32513430.0,
      "step": 15665
    },
    {
      "entropy": 5.4377953052520756,
      "epoch": 1.42144412191582,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048027041972048225,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20099482089281082,
      "num_tokens": 32524409.0,
      "step": 15670
    },
    {
      "entropy": 5.553106594085693,
      "epoch": 1.421897677793904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004802571678502124,
      "loss": 5.0725,
      "mean_token_accuracy": 0.19653349220752717,
      "num_tokens": 32534696.0,
      "step": 15675
    },
    {
      "entropy": 5.576407051086425,
      "epoch": 1.4223512336719883,
      "grad_norm": 0.9375,
      "learning_rate": 0.000480243911735172,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20046114325523376,
      "num_tokens": 32544782.0,
      "step": 15680
    },
    {
      "entropy": 5.475587797164917,
      "epoch": 1.4228047895500726,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048023065137563534,
      "loss": 5.0548,
      "mean_token_accuracy": 0.19665246605873107,
      "num_tokens": 32554818.0,
      "step": 15685
    },
    {
      "entropy": 5.473144388198852,
      "epoch": 1.4232583454281569,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004802173867718764,
      "loss": 5.0834,
      "mean_token_accuracy": 0.20439634025096892,
      "num_tokens": 32565007.0,
      "step": 15690
    },
    {
      "entropy": 5.523764944076538,
      "epoch": 1.423711901306241,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004802041179241697,
      "loss": 5.0382,
      "mean_token_accuracy": 0.19968950599431992,
      "num_tokens": 32575657.0,
      "step": 15695
    },
    {
      "entropy": 5.4944602966308596,
      "epoch": 1.4241654571843252,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004801908448327895,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19597385972738265,
      "num_tokens": 32586676.0,
      "step": 15700
    },
    {
      "entropy": 5.4961262226104735,
      "epoch": 1.4246190130624092,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004801775674980103,
      "loss": 5.021,
      "mean_token_accuracy": 0.20194448232650758,
      "num_tokens": 32596191.0,
      "step": 15705
    },
    {
      "entropy": 5.567626953125,
      "epoch": 1.4250725689404935,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048016428592010676,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1960490435361862,
      "num_tokens": 32607178.0,
      "step": 15710
    },
    {
      "entropy": 5.550439023971558,
      "epoch": 1.4255261248185778,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048015100009935347,
      "loss": 5.0184,
      "mean_token_accuracy": 0.2020075500011444,
      "num_tokens": 32617301.0,
      "step": 15715
    },
    {
      "entropy": 5.430491018295288,
      "epoch": 1.4259796806966618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004801377100360253,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2087884947657585,
      "num_tokens": 32627395.0,
      "step": 15720
    },
    {
      "entropy": 5.501881122589111,
      "epoch": 1.426433236574746,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000480124415730397,
      "loss": 5.0287,
      "mean_token_accuracy": 0.206740340590477,
      "num_tokens": 32638460.0,
      "step": 15725
    },
    {
      "entropy": 5.589461088180542,
      "epoch": 1.4268867924528301,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004801111171827434,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20110534280538558,
      "num_tokens": 32649319.0,
      "step": 15730
    },
    {
      "entropy": 5.498560190200806,
      "epoch": 1.4273403483309144,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004800978143933398,
      "loss": 5.0293,
      "mean_token_accuracy": 0.20413393080234526,
      "num_tokens": 32659976.0,
      "step": 15735
    },
    {
      "entropy": 5.433808660507202,
      "epoch": 1.4277939042089987,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048008450736246094,
      "loss": 4.9683,
      "mean_token_accuracy": 0.20098088234663009,
      "num_tokens": 32670772.0,
      "step": 15740
    },
    {
      "entropy": 5.5375607967376705,
      "epoch": 1.4282474600870827,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048007119609038237,
      "loss": 5.0968,
      "mean_token_accuracy": 0.19511828124523162,
      "num_tokens": 32682368.0,
      "step": 15745
    },
    {
      "entropy": 5.439204216003418,
      "epoch": 1.428701015965167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048005788057737907,
      "loss": 4.9274,
      "mean_token_accuracy": 0.20723052322864532,
      "num_tokens": 32692649.0,
      "step": 15750
    },
    {
      "entropy": 5.533323192596436,
      "epoch": 1.429154571843251,
      "grad_norm": 1.0,
      "learning_rate": 0.0004800445608237266,
      "loss": 5.1454,
      "mean_token_accuracy": 0.20151389092206956,
      "num_tokens": 32704075.0,
      "step": 15755
    },
    {
      "entropy": 5.5563231945037845,
      "epoch": 1.4296081277213353,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048003123682970045,
      "loss": 5.1597,
      "mean_token_accuracy": 0.19443312734365464,
      "num_tokens": 32714670.0,
      "step": 15760
    },
    {
      "entropy": 5.512262058258057,
      "epoch": 1.4300616835994195,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048001790859557596,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2086060270667076,
      "num_tokens": 32724601.0,
      "step": 15765
    },
    {
      "entropy": 5.458773851394653,
      "epoch": 1.4305152394775036,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004800045761216289,
      "loss": 4.9602,
      "mean_token_accuracy": 0.20292164236307145,
      "num_tokens": 32735015.0,
      "step": 15770
    },
    {
      "entropy": 5.500267553329468,
      "epoch": 1.4309687953555879,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000479991239408135,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20426508635282517,
      "num_tokens": 32745899.0,
      "step": 15775
    },
    {
      "entropy": 5.536904525756836,
      "epoch": 1.431422351233672,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047997789845537003,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20640725493431092,
      "num_tokens": 32756052.0,
      "step": 15780
    },
    {
      "entropy": 5.4844502925872805,
      "epoch": 1.4318759071117562,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004799645532636098,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2041822388768196,
      "num_tokens": 32767118.0,
      "step": 15785
    },
    {
      "entropy": 5.410877799987793,
      "epoch": 1.4323294629898404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047995120383313044,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21355968564748765,
      "num_tokens": 32777069.0,
      "step": 15790
    },
    {
      "entropy": 5.477860069274902,
      "epoch": 1.4327830188679245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047993785016420796,
      "loss": 5.1263,
      "mean_token_accuracy": 0.1909132644534111,
      "num_tokens": 32786967.0,
      "step": 15795
    },
    {
      "entropy": 5.523899364471435,
      "epoch": 1.4332365747460087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047992449225711854,
      "loss": 5.002,
      "mean_token_accuracy": 0.20324475318193436,
      "num_tokens": 32797563.0,
      "step": 15800
    },
    {
      "entropy": 5.5151763439178465,
      "epoch": 1.4336901306240928,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047991113011213837,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20701471120119094,
      "num_tokens": 32808730.0,
      "step": 15805
    },
    {
      "entropy": 5.444163513183594,
      "epoch": 1.434143686502177,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004798977637295438,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20792941600084305,
      "num_tokens": 32818050.0,
      "step": 15810
    },
    {
      "entropy": 5.459983968734742,
      "epoch": 1.4345972423802613,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798843931096113,
      "loss": 5.0233,
      "mean_token_accuracy": 0.20102835297584534,
      "num_tokens": 32828993.0,
      "step": 15815
    },
    {
      "entropy": 5.4861907958984375,
      "epoch": 1.4350507982583454,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004798710182526173,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20848754793405533,
      "num_tokens": 32839331.0,
      "step": 15820
    },
    {
      "entropy": 5.5293779373168945,
      "epoch": 1.4355043541364296,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047985763915883847,
      "loss": 5.0278,
      "mean_token_accuracy": 0.19959765970706939,
      "num_tokens": 32849946.0,
      "step": 15825
    },
    {
      "entropy": 5.438718461990357,
      "epoch": 1.4359579100145137,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798442558285514,
      "loss": 4.965,
      "mean_token_accuracy": 0.2041826367378235,
      "num_tokens": 32859960.0,
      "step": 15830
    },
    {
      "entropy": 5.485565805435181,
      "epoch": 1.436411465892598,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004798308682620329,
      "loss": 4.9937,
      "mean_token_accuracy": 0.19982496798038482,
      "num_tokens": 32870317.0,
      "step": 15835
    },
    {
      "entropy": 5.587348508834839,
      "epoch": 1.4368650217706822,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047981747645955996,
      "loss": 5.1683,
      "mean_token_accuracy": 0.19669103622436523,
      "num_tokens": 32880906.0,
      "step": 15840
    },
    {
      "entropy": 5.593022584915161,
      "epoch": 1.4373185776487662,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004798040804214094,
      "loss": 5.1978,
      "mean_token_accuracy": 0.1926955610513687,
      "num_tokens": 32891038.0,
      "step": 15845
    },
    {
      "entropy": 5.585289764404297,
      "epoch": 1.4377721335268505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797906801478582,
      "loss": 5.136,
      "mean_token_accuracy": 0.20103828907012938,
      "num_tokens": 32901013.0,
      "step": 15850
    },
    {
      "entropy": 5.58568844795227,
      "epoch": 1.4382256894049346,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004797772756391836,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1981079399585724,
      "num_tokens": 32911592.0,
      "step": 15855
    },
    {
      "entropy": 5.4849381923675535,
      "epoch": 1.4386792452830188,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004797638668956627,
      "loss": 4.9591,
      "mean_token_accuracy": 0.20855915248394014,
      "num_tokens": 32922087.0,
      "step": 15860
    },
    {
      "entropy": 5.482039165496826,
      "epoch": 1.439132801161103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004797504539175729,
      "loss": 5.0489,
      "mean_token_accuracy": 0.20400669425725937,
      "num_tokens": 32932039.0,
      "step": 15865
    },
    {
      "entropy": 5.499445915222168,
      "epoch": 1.4395863570391874,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004797370367051915,
      "loss": 5.0049,
      "mean_token_accuracy": 0.2056662067770958,
      "num_tokens": 32941896.0,
      "step": 15870
    },
    {
      "entropy": 5.50581727027893,
      "epoch": 1.4400399129172714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000479723615258796,
      "loss": 5.0322,
      "mean_token_accuracy": 0.19427276104688646,
      "num_tokens": 32951675.0,
      "step": 15875
    },
    {
      "entropy": 5.531674242019653,
      "epoch": 1.4404934687953557,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047971018957866396,
      "loss": 5.1049,
      "mean_token_accuracy": 0.20822160840034484,
      "num_tokens": 32961378.0,
      "step": 15880
    },
    {
      "entropy": 5.497912979125976,
      "epoch": 1.4409470246734397,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047969675966507314,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20277913063764572,
      "num_tokens": 32971770.0,
      "step": 15885
    },
    {
      "entropy": 5.465924024581909,
      "epoch": 1.441400580551524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000479683325518301,
      "loss": 5.0215,
      "mean_token_accuracy": 0.2001161202788353,
      "num_tokens": 32982934.0,
      "step": 15890
    },
    {
      "entropy": 5.492756032943726,
      "epoch": 1.4418541364296082,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004796698871386257,
      "loss": 5.0294,
      "mean_token_accuracy": 0.1993359848856926,
      "num_tokens": 32993252.0,
      "step": 15895
    },
    {
      "entropy": 5.536276340484619,
      "epoch": 1.4423076923076923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004796564445263249,
      "loss": 5.0486,
      "mean_token_accuracy": 0.2013053610920906,
      "num_tokens": 33003610.0,
      "step": 15900
    },
    {
      "entropy": 5.499846506118774,
      "epoch": 1.4427612481857766,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047964299768167665,
      "loss": 5.0061,
      "mean_token_accuracy": 0.2022084638476372,
      "num_tokens": 33014037.0,
      "step": 15905
    },
    {
      "entropy": 5.495324516296387,
      "epoch": 1.4432148040638606,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004796295466049591,
      "loss": 5.031,
      "mean_token_accuracy": 0.1982549697160721,
      "num_tokens": 33023299.0,
      "step": 15910
    },
    {
      "entropy": 5.528369045257568,
      "epoch": 1.4436683599419449,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047961609129645036,
      "loss": 5.085,
      "mean_token_accuracy": 0.19658140540122987,
      "num_tokens": 33034194.0,
      "step": 15915
    },
    {
      "entropy": 5.492931365966797,
      "epoch": 1.4441219158200291,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047960263175642867,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21053073406219483,
      "num_tokens": 33044715.0,
      "step": 15920
    },
    {
      "entropy": 5.429548358917236,
      "epoch": 1.4445754716981132,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004795891679851724,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21614315807819368,
      "num_tokens": 33055214.0,
      "step": 15925
    },
    {
      "entropy": 5.474163293838501,
      "epoch": 1.4450290275761974,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004795756999829601,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20264820456504823,
      "num_tokens": 33065026.0,
      "step": 15930
    },
    {
      "entropy": 5.55479736328125,
      "epoch": 1.4454825834542815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047956222775007016,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1960279420018196,
      "num_tokens": 33075138.0,
      "step": 15935
    },
    {
      "entropy": 5.525855541229248,
      "epoch": 1.4459361393323658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004795487512867812,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20468559861183167,
      "num_tokens": 33086625.0,
      "step": 15940
    },
    {
      "entropy": 5.53939790725708,
      "epoch": 1.44638969521045,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047953527059337194,
      "loss": 5.0861,
      "mean_token_accuracy": 0.19433015435934067,
      "num_tokens": 33097076.0,
      "step": 15945
    },
    {
      "entropy": 5.514823246002197,
      "epoch": 1.446843251088534,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004795217856701212,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20547079890966416,
      "num_tokens": 33107168.0,
      "step": 15950
    },
    {
      "entropy": 5.537642621994019,
      "epoch": 1.4472968069666183,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047950829651730774,
      "loss": 5.0654,
      "mean_token_accuracy": 0.20469075739383696,
      "num_tokens": 33117862.0,
      "step": 15955
    },
    {
      "entropy": 5.523198652267456,
      "epoch": 1.4477503628447024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004794948031352107,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20375883877277373,
      "num_tokens": 33127144.0,
      "step": 15960
    },
    {
      "entropy": 5.587871456146241,
      "epoch": 1.4482039187227866,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047948130552410897,
      "loss": 5.1368,
      "mean_token_accuracy": 0.19235823452472686,
      "num_tokens": 33139180.0,
      "step": 15965
    },
    {
      "entropy": 5.546544790267944,
      "epoch": 1.448657474600871,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047946780368428175,
      "loss": 5.0157,
      "mean_token_accuracy": 0.1992461785674095,
      "num_tokens": 33149467.0,
      "step": 15970
    },
    {
      "entropy": 5.458764553070068,
      "epoch": 1.449111030478955,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004794542976160081,
      "loss": 4.8907,
      "mean_token_accuracy": 0.21425037384033202,
      "num_tokens": 33158752.0,
      "step": 15975
    },
    {
      "entropy": 5.4563695907592775,
      "epoch": 1.4495645863570392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047944078731956765,
      "loss": 4.978,
      "mean_token_accuracy": 0.2071639657020569,
      "num_tokens": 33168242.0,
      "step": 15980
    },
    {
      "entropy": 5.450711584091186,
      "epoch": 1.4500181422351233,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047942727279523957,
      "loss": 4.995,
      "mean_token_accuracy": 0.20726195424795152,
      "num_tokens": 33179134.0,
      "step": 15985
    },
    {
      "entropy": 5.507195711135864,
      "epoch": 1.4504716981132075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004794137540433034,
      "loss": 5.1177,
      "mean_token_accuracy": 0.19952171742916108,
      "num_tokens": 33188388.0,
      "step": 15990
    },
    {
      "entropy": 5.557099103927612,
      "epoch": 1.4509252539912918,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004794002310640386,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2005619153380394,
      "num_tokens": 33198750.0,
      "step": 15995
    },
    {
      "entropy": 5.621443796157837,
      "epoch": 1.4513788098693758,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047938670385772503,
      "loss": 5.1347,
      "mean_token_accuracy": 0.19036090821027757,
      "num_tokens": 33209532.0,
      "step": 16000
    },
    {
      "entropy": 5.533468580245971,
      "epoch": 1.45183236574746,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047937317242464244,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20237762480974197,
      "num_tokens": 33219675.0,
      "step": 16005
    },
    {
      "entropy": 5.518578052520752,
      "epoch": 1.4522859216255442,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004793596367650704,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19858584105968474,
      "num_tokens": 33230919.0,
      "step": 16010
    },
    {
      "entropy": 5.495726633071899,
      "epoch": 1.4527394775036284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004793460968792891,
      "loss": 4.9979,
      "mean_token_accuracy": 0.20685409456491471,
      "num_tokens": 33241146.0,
      "step": 16015
    },
    {
      "entropy": 5.535656595230103,
      "epoch": 1.4531930333817127,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047933255276757836,
      "loss": 5.0513,
      "mean_token_accuracy": 0.19511209279298783,
      "num_tokens": 33251434.0,
      "step": 16020
    },
    {
      "entropy": 5.569311809539795,
      "epoch": 1.4536465892597967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004793190044302184,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19259205162525178,
      "num_tokens": 33261868.0,
      "step": 16025
    },
    {
      "entropy": 5.501216983795166,
      "epoch": 1.454100145137881,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047930545186748936,
      "loss": 4.9708,
      "mean_token_accuracy": 0.2036798745393753,
      "num_tokens": 33271798.0,
      "step": 16030
    },
    {
      "entropy": 5.495710563659668,
      "epoch": 1.454553701015965,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047929189507967153,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20195723026990892,
      "num_tokens": 33282614.0,
      "step": 16035
    },
    {
      "entropy": 5.459695625305176,
      "epoch": 1.4550072568940493,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047927833406704516,
      "loss": 4.9892,
      "mean_token_accuracy": 0.19896757304668428,
      "num_tokens": 33292906.0,
      "step": 16040
    },
    {
      "entropy": 5.509294652938843,
      "epoch": 1.4554608127721336,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004792647688298909,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20415859073400497,
      "num_tokens": 33302720.0,
      "step": 16045
    },
    {
      "entropy": 5.459511947631836,
      "epoch": 1.4559143686502178,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004792511993684891,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2102614790201187,
      "num_tokens": 33312959.0,
      "step": 16050
    },
    {
      "entropy": 5.46536283493042,
      "epoch": 1.4563679245283019,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047923762568312053,
      "loss": 4.9342,
      "mean_token_accuracy": 0.20821652561426163,
      "num_tokens": 33322737.0,
      "step": 16055
    },
    {
      "entropy": 5.4693663120269775,
      "epoch": 1.4568214804063861,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004792240477740657,
      "loss": 4.9625,
      "mean_token_accuracy": 0.21078547537326814,
      "num_tokens": 33333185.0,
      "step": 16060
    },
    {
      "entropy": 5.519652271270752,
      "epoch": 1.4572750362844702,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004792104656416056,
      "loss": 5.0883,
      "mean_token_accuracy": 0.1959143102169037,
      "num_tokens": 33343414.0,
      "step": 16065
    },
    {
      "entropy": 5.496297025680542,
      "epoch": 1.4577285921625545,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047919687928602104,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2069736033678055,
      "num_tokens": 33352527.0,
      "step": 16070
    },
    {
      "entropy": 5.553343343734741,
      "epoch": 1.4581821480406387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047918328870759293,
      "loss": 5.0628,
      "mean_token_accuracy": 0.19901107400655746,
      "num_tokens": 33363526.0,
      "step": 16075
    },
    {
      "entropy": 5.490126514434815,
      "epoch": 1.4586357039187228,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004791696939066024,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20202584862709044,
      "num_tokens": 33374302.0,
      "step": 16080
    },
    {
      "entropy": 5.5225855827331545,
      "epoch": 1.459089259796807,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791560948833306,
      "loss": 5.0604,
      "mean_token_accuracy": 0.19944061189889908,
      "num_tokens": 33384786.0,
      "step": 16085
    },
    {
      "entropy": 5.548208236694336,
      "epoch": 1.459542815674891,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004791424916380586,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2090217351913452,
      "num_tokens": 33394464.0,
      "step": 16090
    },
    {
      "entropy": 5.509014225006103,
      "epoch": 1.4599963715529753,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000479128884171068,
      "loss": 5.0608,
      "mean_token_accuracy": 0.20015767216682434,
      "num_tokens": 33405476.0,
      "step": 16095
    },
    {
      "entropy": 5.522106266021728,
      "epoch": 1.4604499274310596,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047911527248264003,
      "loss": 5.0873,
      "mean_token_accuracy": 0.20099602192640303,
      "num_tokens": 33415758.0,
      "step": 16100
    },
    {
      "entropy": 5.497673892974854,
      "epoch": 1.4609034833091437,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047910165657305613,
      "loss": 5.06,
      "mean_token_accuracy": 0.19588928818702697,
      "num_tokens": 33424845.0,
      "step": 16105
    },
    {
      "entropy": 5.547488403320313,
      "epoch": 1.461357039187228,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047908803644259804,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20185463279485702,
      "num_tokens": 33435928.0,
      "step": 16110
    },
    {
      "entropy": 5.537172937393189,
      "epoch": 1.461810595065312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047907441209154726,
      "loss": 4.9964,
      "mean_token_accuracy": 0.19959668070077896,
      "num_tokens": 33447307.0,
      "step": 16115
    },
    {
      "entropy": 5.52874789237976,
      "epoch": 1.4622641509433962,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047906078352018575,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19590460509061813,
      "num_tokens": 33458116.0,
      "step": 16120
    },
    {
      "entropy": 5.383693361282349,
      "epoch": 1.4627177068214805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004790471507287952,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2084435611963272,
      "num_tokens": 33469065.0,
      "step": 16125
    },
    {
      "entropy": 5.535000514984131,
      "epoch": 1.4631712626995645,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047903351371765757,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20379114598035813,
      "num_tokens": 33479118.0,
      "step": 16130
    },
    {
      "entropy": 5.554023694992066,
      "epoch": 1.4636248185776488,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047901987248705484,
      "loss": 5.064,
      "mean_token_accuracy": 0.19928945004940032,
      "num_tokens": 33489642.0,
      "step": 16135
    },
    {
      "entropy": 5.477730655670166,
      "epoch": 1.4640783744557329,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004790062270372693,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19506554156541825,
      "num_tokens": 33499771.0,
      "step": 16140
    },
    {
      "entropy": 5.470936298370361,
      "epoch": 1.4645319303338171,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004789925773685828,
      "loss": 5.027,
      "mean_token_accuracy": 0.19897197037935258,
      "num_tokens": 33509975.0,
      "step": 16145
    },
    {
      "entropy": 5.5395927906036375,
      "epoch": 1.4649854862119014,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004789789234812779,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1954951375722885,
      "num_tokens": 33520017.0,
      "step": 16150
    },
    {
      "entropy": 5.571431016921997,
      "epoch": 1.4654390420899854,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047896526537563686,
      "loss": 5.1316,
      "mean_token_accuracy": 0.1930692121386528,
      "num_tokens": 33531977.0,
      "step": 16155
    },
    {
      "entropy": 5.512016248703003,
      "epoch": 1.4658925979680697,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047895160305194223,
      "loss": 5.0951,
      "mean_token_accuracy": 0.19491859078407286,
      "num_tokens": 33543660.0,
      "step": 16160
    },
    {
      "entropy": 5.518211936950683,
      "epoch": 1.4663461538461537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789379365104764,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2014843299984932,
      "num_tokens": 33553814.0,
      "step": 16165
    },
    {
      "entropy": 5.52564640045166,
      "epoch": 1.466799709724238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004789242657515221,
      "loss": 5.0457,
      "mean_token_accuracy": 0.2034329056739807,
      "num_tokens": 33564250.0,
      "step": 16170
    },
    {
      "entropy": 5.5023092269897464,
      "epoch": 1.4672532656023223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004789105907753621,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20267768502235411,
      "num_tokens": 33574755.0,
      "step": 16175
    },
    {
      "entropy": 5.480043220520019,
      "epoch": 1.4677068214804063,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000478896911582279,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2060309335589409,
      "num_tokens": 33585268.0,
      "step": 16180
    },
    {
      "entropy": 5.487579488754273,
      "epoch": 1.4681603773584906,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004788832281725558,
      "loss": 5.0891,
      "mean_token_accuracy": 0.2038735941052437,
      "num_tokens": 33595753.0,
      "step": 16185
    },
    {
      "entropy": 5.461152362823486,
      "epoch": 1.4686139332365746,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004788695405464755,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20444866567850112,
      "num_tokens": 33605270.0,
      "step": 16190
    },
    {
      "entropy": 5.515757703781128,
      "epoch": 1.469067489114659,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047885584870432127,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20079802423715593,
      "num_tokens": 33615096.0,
      "step": 16195
    },
    {
      "entropy": 5.5126025676727295,
      "epoch": 1.4695210449927432,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047884215264637606,
      "loss": 4.968,
      "mean_token_accuracy": 0.20840421617031096,
      "num_tokens": 33624770.0,
      "step": 16200
    },
    {
      "entropy": 5.545363903045654,
      "epoch": 1.4699746008708272,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004788284523729232,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19470954835414886,
      "num_tokens": 33634545.0,
      "step": 16205
    },
    {
      "entropy": 5.569434070587159,
      "epoch": 1.4704281567489115,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047881474788424595,
      "loss": 5.0661,
      "mean_token_accuracy": 0.19960723966360092,
      "num_tokens": 33644950.0,
      "step": 16210
    },
    {
      "entropy": 5.549499034881592,
      "epoch": 1.4708817126269955,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047880103918062776,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20500340163707734,
      "num_tokens": 33654369.0,
      "step": 16215
    },
    {
      "entropy": 5.4006843090057375,
      "epoch": 1.4713352685050798,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047878732626235225,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19897183328866958,
      "num_tokens": 33664687.0,
      "step": 16220
    },
    {
      "entropy": 5.475517129898071,
      "epoch": 1.471788824383164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004787736091297028,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2159730851650238,
      "num_tokens": 33675832.0,
      "step": 16225
    },
    {
      "entropy": 5.559790515899659,
      "epoch": 1.4722423802612483,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004787598877829633,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20152770280838012,
      "num_tokens": 33686788.0,
      "step": 16230
    },
    {
      "entropy": 5.472647285461425,
      "epoch": 1.4726959361393324,
      "grad_norm": 1.125,
      "learning_rate": 0.0004787461622224173,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20888569504022597,
      "num_tokens": 33696384.0,
      "step": 16235
    },
    {
      "entropy": 5.463289499282837,
      "epoch": 1.4731494920174166,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004787324324483488,
      "loss": 5.0216,
      "mean_token_accuracy": 0.2022559866309166,
      "num_tokens": 33707758.0,
      "step": 16240
    },
    {
      "entropy": 5.502991962432861,
      "epoch": 1.4736030478955007,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004787186984610416,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2065095692873001,
      "num_tokens": 33717747.0,
      "step": 16245
    },
    {
      "entropy": 5.4916908740997314,
      "epoch": 1.474056603773585,
      "grad_norm": 1.0,
      "learning_rate": 0.00047870496026077984,
      "loss": 4.9906,
      "mean_token_accuracy": 0.20281320214271545,
      "num_tokens": 33727960.0,
      "step": 16250
    },
    {
      "entropy": 5.448631668090821,
      "epoch": 1.4745101596516692,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004786912178478476,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20749327540397644,
      "num_tokens": 33738874.0,
      "step": 16255
    },
    {
      "entropy": 5.473556089401245,
      "epoch": 1.4749637155297532,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004786774712225291,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20806999057531356,
      "num_tokens": 33749071.0,
      "step": 16260
    },
    {
      "entropy": 5.497419309616089,
      "epoch": 1.4754172714078375,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004786637203851086,
      "loss": 5.0007,
      "mean_token_accuracy": 0.20123066157102584,
      "num_tokens": 33759509.0,
      "step": 16265
    },
    {
      "entropy": 5.53020339012146,
      "epoch": 1.4758708272859216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004786499653358705,
      "loss": 5.0372,
      "mean_token_accuracy": 0.20973011404275893,
      "num_tokens": 33769974.0,
      "step": 16270
    },
    {
      "entropy": 5.570294809341431,
      "epoch": 1.4763243831640058,
      "grad_norm": 1.0,
      "learning_rate": 0.0004786362060750991,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20195408910512924,
      "num_tokens": 33780403.0,
      "step": 16275
    },
    {
      "entropy": 5.508374261856079,
      "epoch": 1.47677793904209,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047862244260307913,
      "loss": 5.0793,
      "mean_token_accuracy": 0.20363739430904387,
      "num_tokens": 33791364.0,
      "step": 16280
    },
    {
      "entropy": 5.478105258941651,
      "epoch": 1.4772314949201741,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004786086749200951,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20485219210386277,
      "num_tokens": 33801964.0,
      "step": 16285
    },
    {
      "entropy": 5.54928183555603,
      "epoch": 1.4776850507982584,
      "grad_norm": 1.0,
      "learning_rate": 0.0004785949030264319,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20498123168945312,
      "num_tokens": 33813132.0,
      "step": 16290
    },
    {
      "entropy": 5.491539859771729,
      "epoch": 1.4781386066763424,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004785811269223741,
      "loss": 5.0086,
      "mean_token_accuracy": 0.2109767660498619,
      "num_tokens": 33823069.0,
      "step": 16295
    },
    {
      "entropy": 5.547567510604859,
      "epoch": 1.4785921625544267,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004785673466082068,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2017671912908554,
      "num_tokens": 33832626.0,
      "step": 16300
    },
    {
      "entropy": 5.47188663482666,
      "epoch": 1.479045718432511,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047855356208421484,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20665343701839448,
      "num_tokens": 33843384.0,
      "step": 16305
    },
    {
      "entropy": 5.470773935317993,
      "epoch": 1.479499274310595,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004785397733506834,
      "loss": 4.9992,
      "mean_token_accuracy": 0.20403469055891038,
      "num_tokens": 33854286.0,
      "step": 16310
    },
    {
      "entropy": 5.445315361022949,
      "epoch": 1.4799528301886793,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004785259804078976,
      "loss": 4.965,
      "mean_token_accuracy": 0.2035436064004898,
      "num_tokens": 33863379.0,
      "step": 16315
    },
    {
      "entropy": 5.542968368530273,
      "epoch": 1.4804063860667633,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004785121832561426,
      "loss": 5.2038,
      "mean_token_accuracy": 0.1929345428943634,
      "num_tokens": 33875641.0,
      "step": 16320
    },
    {
      "entropy": 5.525194787979126,
      "epoch": 1.4808599419448476,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047849838189570375,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20928843170404435,
      "num_tokens": 33885023.0,
      "step": 16325
    },
    {
      "entropy": 5.618498134613037,
      "epoch": 1.4813134978229319,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004784845763268666,
      "loss": 5.1197,
      "mean_token_accuracy": 0.19594130516052247,
      "num_tokens": 33895562.0,
      "step": 16330
    },
    {
      "entropy": 5.488453531265259,
      "epoch": 1.481767053701016,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004784707665499165,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20698830634355544,
      "num_tokens": 33907095.0,
      "step": 16335
    },
    {
      "entropy": 5.407530736923218,
      "epoch": 1.4822206095791002,
      "grad_norm": 1.0,
      "learning_rate": 0.0004784569525651391,
      "loss": 4.8663,
      "mean_token_accuracy": 0.21668605953454972,
      "num_tokens": 33917475.0,
      "step": 16340
    },
    {
      "entropy": 5.445207118988037,
      "epoch": 1.4826741654571842,
      "grad_norm": 1.0,
      "learning_rate": 0.00047844313437282005,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20546100437641143,
      "num_tokens": 33927740.0,
      "step": 16345
    },
    {
      "entropy": 5.458725690841675,
      "epoch": 1.4831277213352685,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047842931197324525,
      "loss": 4.9693,
      "mean_token_accuracy": 0.20588107407093048,
      "num_tokens": 33937416.0,
      "step": 16350
    },
    {
      "entropy": 5.481399345397949,
      "epoch": 1.4835812772133528,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047841548536670033,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21526847928762435,
      "num_tokens": 33946968.0,
      "step": 16355
    },
    {
      "entropy": 5.577416038513183,
      "epoch": 1.4840348330914368,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047840165455347136,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19545600265264512,
      "num_tokens": 33957877.0,
      "step": 16360
    },
    {
      "entropy": 5.556639623641968,
      "epoch": 1.484488388969521,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047838781953384436,
      "loss": 5.0587,
      "mean_token_accuracy": 0.2009688213467598,
      "num_tokens": 33967890.0,
      "step": 16365
    },
    {
      "entropy": 5.463122367858887,
      "epoch": 1.484941944847605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004783739803081054,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20946183651685715,
      "num_tokens": 33977117.0,
      "step": 16370
    },
    {
      "entropy": 5.498557996749878,
      "epoch": 1.4853955007256894,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047836013687654075,
      "loss": 5.1039,
      "mean_token_accuracy": 0.19548790007829667,
      "num_tokens": 33986943.0,
      "step": 16375
    },
    {
      "entropy": 5.518389129638672,
      "epoch": 1.4858490566037736,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047834628923943654,
      "loss": 5.0073,
      "mean_token_accuracy": 0.19906968772411346,
      "num_tokens": 33996499.0,
      "step": 16380
    },
    {
      "entropy": 5.553841161727905,
      "epoch": 1.4863026124818577,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004783324373970793,
      "loss": 5.1192,
      "mean_token_accuracy": 0.20226586610078812,
      "num_tokens": 34006749.0,
      "step": 16385
    },
    {
      "entropy": 5.565635728836059,
      "epoch": 1.486756168359942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004783185813497555,
      "loss": 5.1152,
      "mean_token_accuracy": 0.19189804196357726,
      "num_tokens": 34017194.0,
      "step": 16390
    },
    {
      "entropy": 5.573387622833252,
      "epoch": 1.487209724238026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004783047210977516,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20614250153303146,
      "num_tokens": 34027386.0,
      "step": 16395
    },
    {
      "entropy": 5.6038854122161865,
      "epoch": 1.4876632801161103,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047829085664135424,
      "loss": 5.091,
      "mean_token_accuracy": 0.19450503885746,
      "num_tokens": 34037540.0,
      "step": 16400
    },
    {
      "entropy": 5.462180089950562,
      "epoch": 1.4881168359941945,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047827698798085015,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20876026153564453,
      "num_tokens": 34048008.0,
      "step": 16405
    },
    {
      "entropy": 5.512149477005005,
      "epoch": 1.4885703918722788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004782631151165262,
      "loss": 5.0296,
      "mean_token_accuracy": 0.2054557114839554,
      "num_tokens": 34058009.0,
      "step": 16410
    },
    {
      "entropy": 5.547857332229614,
      "epoch": 1.4890239477503628,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047824923804866914,
      "loss": 4.9857,
      "mean_token_accuracy": 0.20190584361553193,
      "num_tokens": 34067625.0,
      "step": 16415
    },
    {
      "entropy": 5.48768720626831,
      "epoch": 1.4894775036284469,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047823535677756607,
      "loss": 5.0204,
      "mean_token_accuracy": 0.19690577685832977,
      "num_tokens": 34077349.0,
      "step": 16420
    },
    {
      "entropy": 5.555475282669067,
      "epoch": 1.4899310595065312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004782214713035041,
      "loss": 5.139,
      "mean_token_accuracy": 0.19575213491916657,
      "num_tokens": 34088636.0,
      "step": 16425
    },
    {
      "entropy": 5.538937330245972,
      "epoch": 1.4903846153846154,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047820758162677026,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20493769496679307,
      "num_tokens": 34098806.0,
      "step": 16430
    },
    {
      "entropy": 5.502752494812012,
      "epoch": 1.4908381712626997,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047819368774765187,
      "loss": 5.0268,
      "mean_token_accuracy": 0.2061465784907341,
      "num_tokens": 34109713.0,
      "step": 16435
    },
    {
      "entropy": 5.475600719451904,
      "epoch": 1.4912917271407837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047817978966643626,
      "loss": 4.9947,
      "mean_token_accuracy": 0.205531345307827,
      "num_tokens": 34119585.0,
      "step": 16440
    },
    {
      "entropy": 5.568090486526489,
      "epoch": 1.491745283018868,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004781658873834108,
      "loss": 5.0499,
      "mean_token_accuracy": 0.1953028440475464,
      "num_tokens": 34131245.0,
      "step": 16445
    },
    {
      "entropy": 5.523230075836182,
      "epoch": 1.492198838896952,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047815198089886296,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2036787435412407,
      "num_tokens": 34141229.0,
      "step": 16450
    },
    {
      "entropy": 5.5111089706420895,
      "epoch": 1.4926523947750363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004781380702130805,
      "loss": 5.0672,
      "mean_token_accuracy": 0.19387389421463014,
      "num_tokens": 34151781.0,
      "step": 16455
    },
    {
      "entropy": 5.496467018127442,
      "epoch": 1.4931059506531206,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047812415532635087,
      "loss": 5.0182,
      "mean_token_accuracy": 0.20649980157613754,
      "num_tokens": 34162305.0,
      "step": 16460
    },
    {
      "entropy": 5.5562403202056885,
      "epoch": 1.4935595065312046,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000478110236238962,
      "loss": 5.0943,
      "mean_token_accuracy": 0.19907167553901672,
      "num_tokens": 34173613.0,
      "step": 16465
    },
    {
      "entropy": 5.488120412826538,
      "epoch": 1.4940130624092889,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047809631295120164,
      "loss": 5.021,
      "mean_token_accuracy": 0.20047862976789474,
      "num_tokens": 34183540.0,
      "step": 16470
    },
    {
      "entropy": 5.479513454437256,
      "epoch": 1.494466618287373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047808238546335786,
      "loss": 5.0361,
      "mean_token_accuracy": 0.20338803231716157,
      "num_tokens": 34193292.0,
      "step": 16475
    },
    {
      "entropy": 5.552792882919311,
      "epoch": 1.4949201741654572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047806845377571856,
      "loss": 5.0713,
      "mean_token_accuracy": 0.21045733243227005,
      "num_tokens": 34202697.0,
      "step": 16480
    },
    {
      "entropy": 5.5124249935150145,
      "epoch": 1.4953737300435415,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047805451788857187,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20582942068576812,
      "num_tokens": 34213396.0,
      "step": 16485
    },
    {
      "entropy": 5.520803928375244,
      "epoch": 1.4958272859216255,
      "grad_norm": 1.15625,
      "learning_rate": 0.000478040577802206,
      "loss": 4.9824,
      "mean_token_accuracy": 0.20087953060865402,
      "num_tokens": 34223020.0,
      "step": 16490
    },
    {
      "entropy": 5.445318698883057,
      "epoch": 1.4962808417997098,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004780266335169093,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20431505143642426,
      "num_tokens": 34233249.0,
      "step": 16495
    },
    {
      "entropy": 5.581749963760376,
      "epoch": 1.4967343976777938,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047801268503297,
      "loss": 5.0777,
      "mean_token_accuracy": 0.19654079228639604,
      "num_tokens": 34244091.0,
      "step": 16500
    },
    {
      "entropy": 5.560769701004029,
      "epoch": 1.497187953555878,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047799873235067667,
      "loss": 5.072,
      "mean_token_accuracy": 0.204006327688694,
      "num_tokens": 34255060.0,
      "step": 16505
    },
    {
      "entropy": 5.464819622039795,
      "epoch": 1.4976415094339623,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004779847754703178,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2080100730061531,
      "num_tokens": 34264361.0,
      "step": 16510
    },
    {
      "entropy": 5.492960977554321,
      "epoch": 1.4980950653120464,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004779708143921821,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20396582782268524,
      "num_tokens": 34274480.0,
      "step": 16515
    },
    {
      "entropy": 5.490897560119629,
      "epoch": 1.4985486211901307,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047795684911655824,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20501819401979446,
      "num_tokens": 34285951.0,
      "step": 16520
    },
    {
      "entropy": 5.553942012786865,
      "epoch": 1.4990021770682147,
      "grad_norm": 1.015625,
      "learning_rate": 0.000477942879643735,
      "loss": 5.0925,
      "mean_token_accuracy": 0.19646671563386917,
      "num_tokens": 34296384.0,
      "step": 16525
    },
    {
      "entropy": 5.501747560501099,
      "epoch": 1.499455732946299,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047792890597400125,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20746974349021913,
      "num_tokens": 34306776.0,
      "step": 16530
    },
    {
      "entropy": 5.557483053207397,
      "epoch": 1.4999092888243832,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047791492810764604,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20399336218833924,
      "num_tokens": 34317047.0,
      "step": 16535
    },
    {
      "entropy": 5.4808516025543215,
      "epoch": 1.5003628447024675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004779009460449584,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21083079278469086,
      "num_tokens": 34326392.0,
      "step": 16540
    },
    {
      "entropy": 5.536378717422485,
      "epoch": 1.5008164005805515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047788695978622756,
      "loss": 5.0283,
      "mean_token_accuracy": 0.206308513879776,
      "num_tokens": 34336580.0,
      "step": 16545
    },
    {
      "entropy": 5.4815671920776365,
      "epoch": 1.5012699564586356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047787296933174265,
      "loss": 5.0454,
      "mean_token_accuracy": 0.2009593978524208,
      "num_tokens": 34346792.0,
      "step": 16550
    },
    {
      "entropy": 5.531548976898193,
      "epoch": 1.5017235123367199,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000477858974681793,
      "loss": 5.073,
      "mean_token_accuracy": 0.19875523447990417,
      "num_tokens": 34356675.0,
      "step": 16555
    },
    {
      "entropy": 5.547983121871948,
      "epoch": 1.5021770682148041,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004778449758366681,
      "loss": 5.0216,
      "mean_token_accuracy": 0.2152870386838913,
      "num_tokens": 34366357.0,
      "step": 16560
    },
    {
      "entropy": 5.482605981826782,
      "epoch": 1.5026306240928884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004778309727966574,
      "loss": 5.0302,
      "mean_token_accuracy": 0.2065482184290886,
      "num_tokens": 34377229.0,
      "step": 16565
    },
    {
      "entropy": 5.479886150360107,
      "epoch": 1.5030841799709724,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047781696556205045,
      "loss": 4.9632,
      "mean_token_accuracy": 0.2029298186302185,
      "num_tokens": 34387873.0,
      "step": 16570
    },
    {
      "entropy": 5.515002155303955,
      "epoch": 1.5035377358490565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047780295413313696,
      "loss": 4.9855,
      "mean_token_accuracy": 0.20847856849431992,
      "num_tokens": 34398889.0,
      "step": 16575
    },
    {
      "entropy": 5.46763858795166,
      "epoch": 1.5039912917271407,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047778893851020676,
      "loss": 4.9349,
      "mean_token_accuracy": 0.207319737970829,
      "num_tokens": 34408508.0,
      "step": 16580
    },
    {
      "entropy": 5.492764616012574,
      "epoch": 1.504444847605225,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004777749186935496,
      "loss": 5.0973,
      "mean_token_accuracy": 0.19956476986408234,
      "num_tokens": 34418928.0,
      "step": 16585
    },
    {
      "entropy": 5.5559296131134035,
      "epoch": 1.5048984034833093,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004777608946834555,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20337532162666322,
      "num_tokens": 34429789.0,
      "step": 16590
    },
    {
      "entropy": 5.575822639465332,
      "epoch": 1.5053519593613933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004777468664802144,
      "loss": 5.1585,
      "mean_token_accuracy": 0.19455186128616334,
      "num_tokens": 34440795.0,
      "step": 16595
    },
    {
      "entropy": 5.568664598464966,
      "epoch": 1.5058055152394774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047773283408411636,
      "loss": 5.0648,
      "mean_token_accuracy": 0.19236605167388915,
      "num_tokens": 34450772.0,
      "step": 16600
    },
    {
      "entropy": 5.51489109992981,
      "epoch": 1.5062590711175616,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047771879749545175,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20160395056009292,
      "num_tokens": 34461273.0,
      "step": 16605
    },
    {
      "entropy": 5.520502233505249,
      "epoch": 1.506712626995646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047770475671451073,
      "loss": 5.0666,
      "mean_token_accuracy": 0.20057346075773239,
      "num_tokens": 34470942.0,
      "step": 16610
    },
    {
      "entropy": 5.560528707504273,
      "epoch": 1.5071661828737302,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004776907117415837,
      "loss": 4.994,
      "mean_token_accuracy": 0.2034139022231102,
      "num_tokens": 34481117.0,
      "step": 16615
    },
    {
      "entropy": 5.469490098953247,
      "epoch": 1.5076197387518142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047767666257696103,
      "loss": 4.9706,
      "mean_token_accuracy": 0.20528270900249482,
      "num_tokens": 34490953.0,
      "step": 16620
    },
    {
      "entropy": 5.440951585769653,
      "epoch": 1.5080732946298983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047766260922093337,
      "loss": 4.925,
      "mean_token_accuracy": 0.21110389679670333,
      "num_tokens": 34502765.0,
      "step": 16625
    },
    {
      "entropy": 5.6015337944030765,
      "epoch": 1.5085268505079825,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004776485516737913,
      "loss": 5.1148,
      "mean_token_accuracy": 0.19609806537628174,
      "num_tokens": 34514473.0,
      "step": 16630
    },
    {
      "entropy": 5.567944288253784,
      "epoch": 1.5089804063860668,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004776344899358256,
      "loss": 5.1329,
      "mean_token_accuracy": 0.1925711914896965,
      "num_tokens": 34524218.0,
      "step": 16635
    },
    {
      "entropy": 5.536440992355347,
      "epoch": 1.509433962264151,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004776204240073269,
      "loss": 5.0156,
      "mean_token_accuracy": 0.20336877703666686,
      "num_tokens": 34535697.0,
      "step": 16640
    },
    {
      "entropy": 5.533255910873413,
      "epoch": 1.509887518142235,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047760635388858624,
      "loss": 5.0585,
      "mean_token_accuracy": 0.2072887033224106,
      "num_tokens": 34545130.0,
      "step": 16645
    },
    {
      "entropy": 5.475963020324707,
      "epoch": 1.5103410740203191,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047759227957989465,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20716975182294844,
      "num_tokens": 34554974.0,
      "step": 16650
    },
    {
      "entropy": 5.521869230270386,
      "epoch": 1.5107946298984034,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477578201081543,
      "loss": 5.024,
      "mean_token_accuracy": 0.20081094801425933,
      "num_tokens": 34565660.0,
      "step": 16655
    },
    {
      "entropy": 5.5908403396606445,
      "epoch": 1.5112481857764877,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004775641183938225,
      "loss": 5.0622,
      "mean_token_accuracy": 0.1991403043270111,
      "num_tokens": 34576447.0,
      "step": 16660
    },
    {
      "entropy": 5.5940087795257565,
      "epoch": 1.511701741654572,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047755003151702443,
      "loss": 5.1877,
      "mean_token_accuracy": 0.19777849465608596,
      "num_tokens": 34588333.0,
      "step": 16665
    },
    {
      "entropy": 5.453550243377686,
      "epoch": 1.512155297532656,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047753594045144015,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20184141397476196,
      "num_tokens": 34598166.0,
      "step": 16670
    },
    {
      "entropy": 5.485975837707519,
      "epoch": 1.5126088534107403,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004775218451973609,
      "loss": 4.968,
      "mean_token_accuracy": 0.20365387797355652,
      "num_tokens": 34608615.0,
      "step": 16675
    },
    {
      "entropy": 5.497578477859497,
      "epoch": 1.5130624092888243,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004775077457550784,
      "loss": 5.0384,
      "mean_token_accuracy": 0.20195758193731309,
      "num_tokens": 34618618.0,
      "step": 16680
    },
    {
      "entropy": 5.5292384147644045,
      "epoch": 1.5135159651669086,
      "grad_norm": 1.0,
      "learning_rate": 0.00047749364212488403,
      "loss": 5.0519,
      "mean_token_accuracy": 0.19552540332078933,
      "num_tokens": 34629412.0,
      "step": 16685
    },
    {
      "entropy": 5.576356315612793,
      "epoch": 1.5139695210449928,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004774795343070696,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20159600377082826,
      "num_tokens": 34640487.0,
      "step": 16690
    },
    {
      "entropy": 5.537057590484619,
      "epoch": 1.5144230769230769,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004774654223019268,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20246966928243637,
      "num_tokens": 34650126.0,
      "step": 16695
    },
    {
      "entropy": 5.455569219589234,
      "epoch": 1.5148766328011611,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004774513061097474,
      "loss": 4.9078,
      "mean_token_accuracy": 0.20852980017662048,
      "num_tokens": 34659721.0,
      "step": 16700
    },
    {
      "entropy": 5.494598388671875,
      "epoch": 1.5153301886792452,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004774371857308235,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20723765790462495,
      "num_tokens": 34670611.0,
      "step": 16705
    },
    {
      "entropy": 5.525382232666016,
      "epoch": 1.5157837445573294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047742306116544693,
      "loss": 5.007,
      "mean_token_accuracy": 0.2060346469283104,
      "num_tokens": 34680173.0,
      "step": 16710
    },
    {
      "entropy": 5.528834962844849,
      "epoch": 1.5162373004354137,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047740893241390994,
      "loss": 5.0667,
      "mean_token_accuracy": 0.2004269316792488,
      "num_tokens": 34690176.0,
      "step": 16715
    },
    {
      "entropy": 5.544941663742065,
      "epoch": 1.516690856313498,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004773947994765046,
      "loss": 4.9883,
      "mean_token_accuracy": 0.205637526512146,
      "num_tokens": 34700481.0,
      "step": 16720
    },
    {
      "entropy": 5.481034755706787,
      "epoch": 1.517144412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004773806623535232,
      "loss": 5.1565,
      "mean_token_accuracy": 0.19801995754241944,
      "num_tokens": 34710643.0,
      "step": 16725
    },
    {
      "entropy": 5.548236465454101,
      "epoch": 1.517597968069666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047736652104525816,
      "loss": 5.0252,
      "mean_token_accuracy": 0.20002883970737456,
      "num_tokens": 34720998.0,
      "step": 16730
    },
    {
      "entropy": 5.598523807525635,
      "epoch": 1.5180515239477503,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004773523755520019,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20067711025476456,
      "num_tokens": 34731071.0,
      "step": 16735
    },
    {
      "entropy": 5.519377326965332,
      "epoch": 1.5185050798258346,
      "grad_norm": 1.046875,
      "learning_rate": 0.000477338225874047,
      "loss": 4.9939,
      "mean_token_accuracy": 0.20087372213602067,
      "num_tokens": 34741346.0,
      "step": 16740
    },
    {
      "entropy": 5.559559297561646,
      "epoch": 1.5189586357039189,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047732407201168595,
      "loss": 4.9857,
      "mean_token_accuracy": 0.20293824523687362,
      "num_tokens": 34751092.0,
      "step": 16745
    },
    {
      "entropy": 5.604568767547607,
      "epoch": 1.519412191582003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004773099139652116,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20174917578697205,
      "num_tokens": 34761477.0,
      "step": 16750
    },
    {
      "entropy": 5.471523380279541,
      "epoch": 1.519865747460087,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004772957517349167,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2046545147895813,
      "num_tokens": 34772491.0,
      "step": 16755
    },
    {
      "entropy": 5.5182709217071535,
      "epoch": 1.5203193033381712,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004772815853210939,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19876878559589387,
      "num_tokens": 34783192.0,
      "step": 16760
    },
    {
      "entropy": 5.534299516677857,
      "epoch": 1.5207728592162555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047726741472403655,
      "loss": 5.0353,
      "mean_token_accuracy": 0.2047033876180649,
      "num_tokens": 34794164.0,
      "step": 16765
    },
    {
      "entropy": 5.557278108596802,
      "epoch": 1.5212264150943398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047725323994403753,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19490733742713928,
      "num_tokens": 34804390.0,
      "step": 16770
    },
    {
      "entropy": 5.639868545532226,
      "epoch": 1.5216799709724238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047723906098138993,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19346656948328017,
      "num_tokens": 34814751.0,
      "step": 16775
    },
    {
      "entropy": 5.527458047866821,
      "epoch": 1.5221335268505078,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047722487783638697,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20670805424451827,
      "num_tokens": 34824977.0,
      "step": 16780
    },
    {
      "entropy": 5.530150556564331,
      "epoch": 1.5225870827285921,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000477210690509322,
      "loss": 5.0629,
      "mean_token_accuracy": 0.20371615886688232,
      "num_tokens": 34835540.0,
      "step": 16785
    },
    {
      "entropy": 5.575498342514038,
      "epoch": 1.5230406386066764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047719649900048843,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20107826590538025,
      "num_tokens": 34845233.0,
      "step": 16790
    },
    {
      "entropy": 5.594610977172851,
      "epoch": 1.5234941944847606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047718230331017975,
      "loss": 5.1333,
      "mean_token_accuracy": 0.19620306491851808,
      "num_tokens": 34855756.0,
      "step": 16795
    },
    {
      "entropy": 5.552404689788818,
      "epoch": 1.5239477503628447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047716810343868956,
      "loss": 5.0549,
      "mean_token_accuracy": 0.19141069501638414,
      "num_tokens": 34867869.0,
      "step": 16800
    },
    {
      "entropy": 5.617313003540039,
      "epoch": 1.5244013062409287,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047715389938631143,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19923090040683747,
      "num_tokens": 34878625.0,
      "step": 16805
    },
    {
      "entropy": 5.596384859085083,
      "epoch": 1.524854862119013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047713969115333916,
      "loss": 5.101,
      "mean_token_accuracy": 0.1899292528629303,
      "num_tokens": 34889408.0,
      "step": 16810
    },
    {
      "entropy": 5.51461124420166,
      "epoch": 1.5253084179970973,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004771254787400666,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19528281986713408,
      "num_tokens": 34899602.0,
      "step": 16815
    },
    {
      "entropy": 5.465745878219605,
      "epoch": 1.5257619738751815,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004771112621467876,
      "loss": 4.9482,
      "mean_token_accuracy": 0.20518770664930344,
      "num_tokens": 34910215.0,
      "step": 16820
    },
    {
      "entropy": 5.495348262786865,
      "epoch": 1.5262155297532656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004770970413737961,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2078547567129135,
      "num_tokens": 34919641.0,
      "step": 16825
    },
    {
      "entropy": 5.527891826629639,
      "epoch": 1.5266690856313496,
      "grad_norm": 1.0,
      "learning_rate": 0.0004770828164213865,
      "loss": 4.9251,
      "mean_token_accuracy": 0.2131982073187828,
      "num_tokens": 34930243.0,
      "step": 16830
    },
    {
      "entropy": 5.534064245223999,
      "epoch": 1.5271226415094339,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004770685872898525,
      "loss": 5.1001,
      "mean_token_accuracy": 0.20086862593889238,
      "num_tokens": 34941057.0,
      "step": 16835
    },
    {
      "entropy": 5.490671062469483,
      "epoch": 1.5275761973875182,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047705435397948876,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21125524044036864,
      "num_tokens": 34951150.0,
      "step": 16840
    },
    {
      "entropy": 5.574343013763428,
      "epoch": 1.5280297532656024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004770401164905895,
      "loss": 5.0603,
      "mean_token_accuracy": 0.203806009888649,
      "num_tokens": 34961648.0,
      "step": 16845
    },
    {
      "entropy": 5.55496039390564,
      "epoch": 1.5284833091436865,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047702587482344915,
      "loss": 5.0569,
      "mean_token_accuracy": 0.1965765506029129,
      "num_tokens": 34971573.0,
      "step": 16850
    },
    {
      "entropy": 5.556545734405518,
      "epoch": 1.5289368650217707,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004770116289783622,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1982572853565216,
      "num_tokens": 34981984.0,
      "step": 16855
    },
    {
      "entropy": 5.547101306915283,
      "epoch": 1.5293904208998548,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004769973789556233,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1957728385925293,
      "num_tokens": 34990861.0,
      "step": 16860
    },
    {
      "entropy": 5.530113744735718,
      "epoch": 1.529843976777939,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047698312475552714,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20197166949510575,
      "num_tokens": 35001504.0,
      "step": 16865
    },
    {
      "entropy": 5.549115228652954,
      "epoch": 1.5302975326560233,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004769688663783684,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2118733510375023,
      "num_tokens": 35012425.0,
      "step": 16870
    },
    {
      "entropy": 5.511075735092163,
      "epoch": 1.5307510885341074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004769546038244421,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21017370223999024,
      "num_tokens": 35022854.0,
      "step": 16875
    },
    {
      "entropy": 5.544015026092529,
      "epoch": 1.5312046444121916,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047694033709404313,
      "loss": 5.0909,
      "mean_token_accuracy": 0.196853344142437,
      "num_tokens": 35032694.0,
      "step": 16880
    },
    {
      "entropy": 5.522608804702759,
      "epoch": 1.5316582002902757,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047692606618746655,
      "loss": 4.9996,
      "mean_token_accuracy": 0.20808610320091248,
      "num_tokens": 35042217.0,
      "step": 16885
    },
    {
      "entropy": 5.5298590660095215,
      "epoch": 1.53211175616836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047691179110500735,
      "loss": 5.0889,
      "mean_token_accuracy": 0.1910834416747093,
      "num_tokens": 35052831.0,
      "step": 16890
    },
    {
      "entropy": 5.586901235580444,
      "epoch": 1.5325653120464442,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047689751184696097,
      "loss": 5.1123,
      "mean_token_accuracy": 0.1989368662238121,
      "num_tokens": 35061955.0,
      "step": 16895
    },
    {
      "entropy": 5.543337059020996,
      "epoch": 1.5330188679245285,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004768832284136226,
      "loss": 5.0147,
      "mean_token_accuracy": 0.2031451642513275,
      "num_tokens": 35072526.0,
      "step": 16900
    },
    {
      "entropy": 5.501619005203247,
      "epoch": 1.5334724238026125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004768689408052876,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20547995269298552,
      "num_tokens": 35082415.0,
      "step": 16905
    },
    {
      "entropy": 5.475245141983033,
      "epoch": 1.5339259796806965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047685464902225144,
      "loss": 4.982,
      "mean_token_accuracy": 0.20835471153259277,
      "num_tokens": 35092805.0,
      "step": 16910
    },
    {
      "entropy": 5.542777490615845,
      "epoch": 1.5343795355587808,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004768403530648097,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20715077519416808,
      "num_tokens": 35102963.0,
      "step": 16915
    },
    {
      "entropy": 5.54917802810669,
      "epoch": 1.534833091436865,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047682605293325806,
      "loss": 5.0578,
      "mean_token_accuracy": 0.2008623957633972,
      "num_tokens": 35114140.0,
      "step": 16920
    },
    {
      "entropy": 5.588230037689209,
      "epoch": 1.5352866473149493,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047681174862789215,
      "loss": 5.1263,
      "mean_token_accuracy": 0.20138452351093292,
      "num_tokens": 35125314.0,
      "step": 16925
    },
    {
      "entropy": 5.523447322845459,
      "epoch": 1.5357402031930334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047679744014900795,
      "loss": 5.0299,
      "mean_token_accuracy": 0.2043992131948471,
      "num_tokens": 35134934.0,
      "step": 16930
    },
    {
      "entropy": 5.487507247924805,
      "epoch": 1.5361937590711174,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004767831274969011,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20349218994379042,
      "num_tokens": 35145341.0,
      "step": 16935
    },
    {
      "entropy": 5.455434989929199,
      "epoch": 1.5366473149492017,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047676881067186793,
      "loss": 4.9612,
      "mean_token_accuracy": 0.2022953748703003,
      "num_tokens": 35156046.0,
      "step": 16940
    },
    {
      "entropy": 5.558455896377564,
      "epoch": 1.537100870827286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047675448967420426,
      "loss": 5.1179,
      "mean_token_accuracy": 0.19479937106370926,
      "num_tokens": 35168232.0,
      "step": 16945
    },
    {
      "entropy": 5.607294321060181,
      "epoch": 1.5375544267053702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004767401645042064,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20226863920688629,
      "num_tokens": 35178585.0,
      "step": 16950
    },
    {
      "entropy": 5.532871150970459,
      "epoch": 1.5380079825834543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047672583516217043,
      "loss": 4.9313,
      "mean_token_accuracy": 0.20752260088920593,
      "num_tokens": 35189417.0,
      "step": 16955
    },
    {
      "entropy": 5.495153856277466,
      "epoch": 1.5384615384615383,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004767115016483928,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2012806624174118,
      "num_tokens": 35200167.0,
      "step": 16960
    },
    {
      "entropy": 5.490826416015625,
      "epoch": 1.5389150943396226,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004766971639631699,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20241640508174896,
      "num_tokens": 35210711.0,
      "step": 16965
    },
    {
      "entropy": 5.481350374221802,
      "epoch": 1.5393686502177069,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047668282210679825,
      "loss": 5.0353,
      "mean_token_accuracy": 0.1973751053214073,
      "num_tokens": 35220866.0,
      "step": 16970
    },
    {
      "entropy": 5.4429432392120365,
      "epoch": 1.5398222060957911,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004766684760795745,
      "loss": 4.992,
      "mean_token_accuracy": 0.20854220241308213,
      "num_tokens": 35231478.0,
      "step": 16975
    },
    {
      "entropy": 5.476144599914551,
      "epoch": 1.5402757619738752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047665412588179524,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21443094462156295,
      "num_tokens": 35242233.0,
      "step": 16980
    },
    {
      "entropy": 5.492010641098022,
      "epoch": 1.5407293178519592,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004766397715137573,
      "loss": 4.9949,
      "mean_token_accuracy": 0.20129960477352143,
      "num_tokens": 35253097.0,
      "step": 16985
    },
    {
      "entropy": 5.592983293533325,
      "epoch": 1.5411828737300435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047662541297575743,
      "loss": 5.1592,
      "mean_token_accuracy": 0.19676952362060546,
      "num_tokens": 35264258.0,
      "step": 16990
    },
    {
      "entropy": 5.667860507965088,
      "epoch": 1.5416364296081277,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047661105026809274,
      "loss": 5.1293,
      "mean_token_accuracy": 0.19510892480611802,
      "num_tokens": 35273501.0,
      "step": 16995
    },
    {
      "entropy": 5.5120124340057375,
      "epoch": 1.542089985486212,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004765966833910601,
      "loss": 5.0757,
      "mean_token_accuracy": 0.19924862831830978,
      "num_tokens": 35284125.0,
      "step": 17000
    },
    {
      "entropy": 5.618932294845581,
      "epoch": 1.542543541364296,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004765823123449567,
      "loss": 5.073,
      "mean_token_accuracy": 0.1988307297229767,
      "num_tokens": 35296039.0,
      "step": 17005
    },
    {
      "entropy": 5.568522262573242,
      "epoch": 1.54299709724238,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047656793713007965,
      "loss": 5.0089,
      "mean_token_accuracy": 0.19973995089530944,
      "num_tokens": 35306178.0,
      "step": 17010
    },
    {
      "entropy": 5.4798565864562985,
      "epoch": 1.5434506531204644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047655355774672636,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20508571118116378,
      "num_tokens": 35316445.0,
      "step": 17015
    },
    {
      "entropy": 5.537961816787719,
      "epoch": 1.5439042089985486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047653917419519413,
      "loss": 5.0381,
      "mean_token_accuracy": 0.1992551326751709,
      "num_tokens": 35327274.0,
      "step": 17020
    },
    {
      "entropy": 5.526545619964599,
      "epoch": 1.544357764876633,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004765247864757804,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20724283307790756,
      "num_tokens": 35339007.0,
      "step": 17025
    },
    {
      "entropy": 5.55948657989502,
      "epoch": 1.544811320754717,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004765103945887827,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19256114810705185,
      "num_tokens": 35349303.0,
      "step": 17030
    },
    {
      "entropy": 5.5475461959838865,
      "epoch": 1.5452648766328012,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004764959985344987,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2082339644432068,
      "num_tokens": 35359556.0,
      "step": 17035
    },
    {
      "entropy": 5.494559288024902,
      "epoch": 1.5457184325108853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004764815983132261,
      "loss": 4.9836,
      "mean_token_accuracy": 0.20900323241949081,
      "num_tokens": 35370043.0,
      "step": 17040
    },
    {
      "entropy": 5.523154592514038,
      "epoch": 1.5461719883889695,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004764671939252628,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21071674376726152,
      "num_tokens": 35380031.0,
      "step": 17045
    },
    {
      "entropy": 5.532931280136109,
      "epoch": 1.5466255442670538,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004764527853709064,
      "loss": 5.0761,
      "mean_token_accuracy": 0.1993743047118187,
      "num_tokens": 35390481.0,
      "step": 17050
    },
    {
      "entropy": 5.51744589805603,
      "epoch": 1.5470791001451378,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004764383726504552,
      "loss": 5.06,
      "mean_token_accuracy": 0.20070597678422927,
      "num_tokens": 35401756.0,
      "step": 17055
    },
    {
      "entropy": 5.545562696456909,
      "epoch": 1.547532656023222,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004764239557642071,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21284027099609376,
      "num_tokens": 35411045.0,
      "step": 17060
    },
    {
      "entropy": 5.501655769348145,
      "epoch": 1.5479862119013061,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004764095347124603,
      "loss": 5.0474,
      "mean_token_accuracy": 0.19385991096496583,
      "num_tokens": 35421256.0,
      "step": 17065
    },
    {
      "entropy": 5.5252477645874025,
      "epoch": 1.5484397677793904,
      "grad_norm": 1.0,
      "learning_rate": 0.0004763951094955129,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20608636289834975,
      "num_tokens": 35430798.0,
      "step": 17070
    },
    {
      "entropy": 5.514478445053101,
      "epoch": 1.5488933236574747,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004763806801136634,
      "loss": 4.9919,
      "mean_token_accuracy": 0.2099059194326401,
      "num_tokens": 35440459.0,
      "step": 17075
    },
    {
      "entropy": 5.513672113418579,
      "epoch": 1.549346879535559,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004763662465672101,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20351307094097137,
      "num_tokens": 35451586.0,
      "step": 17080
    },
    {
      "entropy": 5.521123647689819,
      "epoch": 1.549800435413643,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047635180885645146,
      "loss": 5.0499,
      "mean_token_accuracy": 0.2032095104455948,
      "num_tokens": 35462197.0,
      "step": 17085
    },
    {
      "entropy": 5.549882459640503,
      "epoch": 1.550253991291727,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004763373669816862,
      "loss": 4.9743,
      "mean_token_accuracy": 0.20916669368743895,
      "num_tokens": 35471737.0,
      "step": 17090
    },
    {
      "entropy": 5.562024736404419,
      "epoch": 1.5507075471698113,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004763229209432128,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1963101416826248,
      "num_tokens": 35482016.0,
      "step": 17095
    },
    {
      "entropy": 5.572429990768432,
      "epoch": 1.5511611030478956,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004763084707413301,
      "loss": 5.1605,
      "mean_token_accuracy": 0.18899236917495726,
      "num_tokens": 35492622.0,
      "step": 17100
    },
    {
      "entropy": 5.624724960327148,
      "epoch": 1.5516146589259798,
      "grad_norm": 1.125,
      "learning_rate": 0.00047629401637633696,
      "loss": 5.1854,
      "mean_token_accuracy": 0.19349545240402222,
      "num_tokens": 35504468.0,
      "step": 17105
    },
    {
      "entropy": 5.528221845626831,
      "epoch": 1.5520682148040639,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047627955784853226,
      "loss": 5.0141,
      "mean_token_accuracy": 0.199432609975338,
      "num_tokens": 35514272.0,
      "step": 17110
    },
    {
      "entropy": 5.519846725463867,
      "epoch": 1.552521770682148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000476265095158215,
      "loss": 5.0043,
      "mean_token_accuracy": 0.1996614083647728,
      "num_tokens": 35524110.0,
      "step": 17115
    },
    {
      "entropy": 5.548833322525025,
      "epoch": 1.5529753265602322,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004762506283056843,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21215675473213197,
      "num_tokens": 35535504.0,
      "step": 17120
    },
    {
      "entropy": 5.624251079559326,
      "epoch": 1.5534288824383164,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004762361572912393,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19390741139650344,
      "num_tokens": 35544885.0,
      "step": 17125
    },
    {
      "entropy": 5.543760204315186,
      "epoch": 1.5538824383164007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047622168211517926,
      "loss": 5.1441,
      "mean_token_accuracy": 0.20457107722759246,
      "num_tokens": 35555485.0,
      "step": 17130
    },
    {
      "entropy": 5.468977546691894,
      "epoch": 1.5543359941944848,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047620720277780353,
      "loss": 4.9531,
      "mean_token_accuracy": 0.20908924788236619,
      "num_tokens": 35567197.0,
      "step": 17135
    },
    {
      "entropy": 5.547029113769531,
      "epoch": 1.5547895500725688,
      "grad_norm": 1.0,
      "learning_rate": 0.00047619271927941165,
      "loss": 4.9914,
      "mean_token_accuracy": 0.20580670535564421,
      "num_tokens": 35577576.0,
      "step": 17140
    },
    {
      "entropy": 5.532487440109253,
      "epoch": 1.555243105950653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000476178231620303,
      "loss": 5.0558,
      "mean_token_accuracy": 0.2027304396033287,
      "num_tokens": 35588012.0,
      "step": 17145
    },
    {
      "entropy": 5.498031139373779,
      "epoch": 1.5556966618287373,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004761637398007774,
      "loss": 5.0329,
      "mean_token_accuracy": 0.19690879434347153,
      "num_tokens": 35598076.0,
      "step": 17150
    },
    {
      "entropy": 5.444629192352295,
      "epoch": 1.5561502177068216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004761492438211342,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2136305958032608,
      "num_tokens": 35607170.0,
      "step": 17155
    },
    {
      "entropy": 5.514594650268554,
      "epoch": 1.5566037735849056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047613474368167355,
      "loss": 4.9972,
      "mean_token_accuracy": 0.1959199920296669,
      "num_tokens": 35617865.0,
      "step": 17160
    },
    {
      "entropy": 5.528052568435669,
      "epoch": 1.5570573294629897,
      "grad_norm": 1.0625,
      "learning_rate": 0.000476120239382695,
      "loss": 5.0517,
      "mean_token_accuracy": 0.20201874971389772,
      "num_tokens": 35629329.0,
      "step": 17165
    },
    {
      "entropy": 5.517793083190918,
      "epoch": 1.557510885341074,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047610573092449877,
      "loss": 5.0536,
      "mean_token_accuracy": 0.19934562891721724,
      "num_tokens": 35640044.0,
      "step": 17170
    },
    {
      "entropy": 5.552383804321289,
      "epoch": 1.5579644412191582,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004760912183073847,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20094917714595795,
      "num_tokens": 35650435.0,
      "step": 17175
    },
    {
      "entropy": 5.532912158966065,
      "epoch": 1.5584179970972425,
      "grad_norm": 0.953125,
      "learning_rate": 0.000476076701531653,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20271799117326736,
      "num_tokens": 35659613.0,
      "step": 17180
    },
    {
      "entropy": 5.494220638275147,
      "epoch": 1.5588715529753265,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004760621805976039,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20519281178712845,
      "num_tokens": 35670987.0,
      "step": 17185
    },
    {
      "entropy": 5.498604202270508,
      "epoch": 1.5593251088534106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047604765550553766,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20504542142152787,
      "num_tokens": 35682505.0,
      "step": 17190
    },
    {
      "entropy": 5.552568626403809,
      "epoch": 1.5597786647314948,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047603312625575466,
      "loss": 4.9551,
      "mean_token_accuracy": 0.20232332348823548,
      "num_tokens": 35693056.0,
      "step": 17195
    },
    {
      "entropy": 5.518586921691894,
      "epoch": 1.5602322206095791,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047601859284855544,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21319104433059693,
      "num_tokens": 35703659.0,
      "step": 17200
    },
    {
      "entropy": 5.4766816139221195,
      "epoch": 1.5606857764876634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004760040552842403,
      "loss": 5.025,
      "mean_token_accuracy": 0.2015480637550354,
      "num_tokens": 35714367.0,
      "step": 17205
    },
    {
      "entropy": 5.4776341915130615,
      "epoch": 1.5611393323657474,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047598951356311035,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2010535180568695,
      "num_tokens": 35724773.0,
      "step": 17210
    },
    {
      "entropy": 5.569267797470093,
      "epoch": 1.5615928882438317,
      "grad_norm": 1.0,
      "learning_rate": 0.0004759749676854658,
      "loss": 5.1312,
      "mean_token_accuracy": 0.19799581170082092,
      "num_tokens": 35735622.0,
      "step": 17215
    },
    {
      "entropy": 5.543905401229859,
      "epoch": 1.5620464441219157,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004759604176516079,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2084849089384079,
      "num_tokens": 35744967.0,
      "step": 17220
    },
    {
      "entropy": 5.555732345581054,
      "epoch": 1.5625,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004759458634618372,
      "loss": 5.0294,
      "mean_token_accuracy": 0.20423816591501237,
      "num_tokens": 35755647.0,
      "step": 17225
    },
    {
      "entropy": 5.529198217391968,
      "epoch": 1.5629535558780843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004759313051164549,
      "loss": 5.063,
      "mean_token_accuracy": 0.1968506619334221,
      "num_tokens": 35765595.0,
      "step": 17230
    },
    {
      "entropy": 5.545705795288086,
      "epoch": 1.5634071117561683,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000475916742615762,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2100907638669014,
      "num_tokens": 35775804.0,
      "step": 17235
    },
    {
      "entropy": 5.504986000061035,
      "epoch": 1.5638606676342526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004759021759600597,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20709893405437468,
      "num_tokens": 35786227.0,
      "step": 17240
    },
    {
      "entropy": 5.567078065872193,
      "epoch": 1.5643142235123366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004758876051496492,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20364011228084564,
      "num_tokens": 35797511.0,
      "step": 17245
    },
    {
      "entropy": 5.560240745544434,
      "epoch": 1.5647677793904209,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047587303018483177,
      "loss": 5.051,
      "mean_token_accuracy": 0.19859770983457564,
      "num_tokens": 35807248.0,
      "step": 17250
    },
    {
      "entropy": 5.52660493850708,
      "epoch": 1.5652213352685052,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047585845106590897,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19810502827167512,
      "num_tokens": 35818164.0,
      "step": 17255
    },
    {
      "entropy": 5.515972757339478,
      "epoch": 1.5656748911465894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047584386779318214,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2109167367219925,
      "num_tokens": 35828746.0,
      "step": 17260
    },
    {
      "entropy": 5.546813488006592,
      "epoch": 1.5661284470246735,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000475829280366953,
      "loss": 5.0758,
      "mean_token_accuracy": 0.1961701601743698,
      "num_tokens": 35839325.0,
      "step": 17265
    },
    {
      "entropy": 5.536845350265503,
      "epoch": 1.5665820029027575,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047581468878752313,
      "loss": 5.001,
      "mean_token_accuracy": 0.20544881075620652,
      "num_tokens": 35849406.0,
      "step": 17270
    },
    {
      "entropy": 5.454688835144043,
      "epoch": 1.5670355587808418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047580009305519427,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2047882094979286,
      "num_tokens": 35858715.0,
      "step": 17275
    },
    {
      "entropy": 5.4782956600189205,
      "epoch": 1.567489114658926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004757854931702684,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2048191696405411,
      "num_tokens": 35869063.0,
      "step": 17280
    },
    {
      "entropy": 5.48824028968811,
      "epoch": 1.5679426705370103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047577088913304733,
      "loss": 4.9003,
      "mean_token_accuracy": 0.20872622430324556,
      "num_tokens": 35878428.0,
      "step": 17285
    },
    {
      "entropy": 5.558001232147217,
      "epoch": 1.5683962264150944,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047575628094383317,
      "loss": 5.035,
      "mean_token_accuracy": 0.2058235377073288,
      "num_tokens": 35888801.0,
      "step": 17290
    },
    {
      "entropy": 5.473654413223267,
      "epoch": 1.5688497822931784,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004757416686029279,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20353106558322906,
      "num_tokens": 35899381.0,
      "step": 17295
    },
    {
      "entropy": 5.556720638275147,
      "epoch": 1.5693033381712627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004757270521106338,
      "loss": 5.0738,
      "mean_token_accuracy": 0.20061299651861192,
      "num_tokens": 35909546.0,
      "step": 17300
    },
    {
      "entropy": 5.520714902877808,
      "epoch": 1.569756894049347,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004757124314672531,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2015165463089943,
      "num_tokens": 35919301.0,
      "step": 17305
    },
    {
      "entropy": 5.58653039932251,
      "epoch": 1.5702104499274312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047569780667308816,
      "loss": 5.0878,
      "mean_token_accuracy": 0.20005113929510115,
      "num_tokens": 35928603.0,
      "step": 17310
    },
    {
      "entropy": 5.58700156211853,
      "epoch": 1.5706640058055152,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047568317772844147,
      "loss": 5.067,
      "mean_token_accuracy": 0.20702016204595566,
      "num_tokens": 35939005.0,
      "step": 17315
    },
    {
      "entropy": 5.575606870651245,
      "epoch": 1.5711175616835993,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047566854463361544,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20989421457052232,
      "num_tokens": 35948471.0,
      "step": 17320
    },
    {
      "entropy": 5.5107934951782225,
      "epoch": 1.5715711175616836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004756539073889128,
      "loss": 5.0578,
      "mean_token_accuracy": 0.20375427305698396,
      "num_tokens": 35958488.0,
      "step": 17325
    },
    {
      "entropy": 5.51311354637146,
      "epoch": 1.5720246734397678,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047563926599463623,
      "loss": 5.1086,
      "mean_token_accuracy": 0.19656799137592315,
      "num_tokens": 35968291.0,
      "step": 17330
    },
    {
      "entropy": 5.564052724838257,
      "epoch": 1.572478229317852,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004756246204510886,
      "loss": 5.0249,
      "mean_token_accuracy": 0.2046962261199951,
      "num_tokens": 35978565.0,
      "step": 17335
    },
    {
      "entropy": 5.559260129928589,
      "epoch": 1.5729317851959361,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047560997075857253,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20554231852293015,
      "num_tokens": 35987483.0,
      "step": 17340
    },
    {
      "entropy": 5.522519826889038,
      "epoch": 1.5733853410740202,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004755953169173913,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19977228343486786,
      "num_tokens": 35998355.0,
      "step": 17345
    },
    {
      "entropy": 5.510653686523438,
      "epoch": 1.5738388969521044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004755806589278477,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2097836673259735,
      "num_tokens": 36008365.0,
      "step": 17350
    },
    {
      "entropy": 5.576305627822876,
      "epoch": 1.5742924528301887,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047556599679024497,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19741983562707902,
      "num_tokens": 36018880.0,
      "step": 17355
    },
    {
      "entropy": 5.55736198425293,
      "epoch": 1.574746008708273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047555133050488637,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19512535780668258,
      "num_tokens": 36029335.0,
      "step": 17360
    },
    {
      "entropy": 5.5182085037231445,
      "epoch": 1.575199564586357,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047553666007207516,
      "loss": 4.9967,
      "mean_token_accuracy": 0.20793434828519822,
      "num_tokens": 36039774.0,
      "step": 17365
    },
    {
      "entropy": 5.588819742202759,
      "epoch": 1.575653120464441,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047552198549211463,
      "loss": 5.1406,
      "mean_token_accuracy": 0.201657672226429,
      "num_tokens": 36049864.0,
      "step": 17370
    },
    {
      "entropy": 5.533351182937622,
      "epoch": 1.5761066763425253,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004755073067653084,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19713248908519745,
      "num_tokens": 36059224.0,
      "step": 17375
    },
    {
      "entropy": 5.431200313568115,
      "epoch": 1.5765602322206096,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004754926238919599,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21095066368579865,
      "num_tokens": 36069243.0,
      "step": 17380
    },
    {
      "entropy": 5.51810622215271,
      "epoch": 1.5770137880986939,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004754779368723729,
      "loss": 5.0266,
      "mean_token_accuracy": 0.1996670052409172,
      "num_tokens": 36079833.0,
      "step": 17385
    },
    {
      "entropy": 5.519850492477417,
      "epoch": 1.577467343976778,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004754632457068511,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20477883368730546,
      "num_tokens": 36090673.0,
      "step": 17390
    },
    {
      "entropy": 5.55263671875,
      "epoch": 1.5779208998548622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004754485503956982,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20432986617088317,
      "num_tokens": 36101719.0,
      "step": 17395
    },
    {
      "entropy": 5.518518877029419,
      "epoch": 1.5783744557329462,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754338509392183,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20799205005168914,
      "num_tokens": 36112352.0,
      "step": 17400
    },
    {
      "entropy": 5.403399133682251,
      "epoch": 1.5788280116110305,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004754191473377153,
      "loss": 4.9002,
      "mean_token_accuracy": 0.2080320432782173,
      "num_tokens": 36121797.0,
      "step": 17405
    },
    {
      "entropy": 5.507850456237793,
      "epoch": 1.5792815674891147,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047540443959149316,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20719701945781707,
      "num_tokens": 36132094.0,
      "step": 17410
    },
    {
      "entropy": 5.516197729110718,
      "epoch": 1.5797351233671988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004753897277008562,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2037370592355728,
      "num_tokens": 36142860.0,
      "step": 17415
    },
    {
      "entropy": 5.527059650421142,
      "epoch": 1.580188679245283,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004753750116661086,
      "loss": 5.0152,
      "mean_token_accuracy": 0.19879119247198104,
      "num_tokens": 36154555.0,
      "step": 17420
    },
    {
      "entropy": 5.602051639556885,
      "epoch": 1.580642235123367,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047536029148755477,
      "loss": 5.145,
      "mean_token_accuracy": 0.1905045822262764,
      "num_tokens": 36165479.0,
      "step": 17425
    },
    {
      "entropy": 5.570267295837402,
      "epoch": 1.5810957910014514,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047534556716549893,
      "loss": 5.0555,
      "mean_token_accuracy": 0.1985606536269188,
      "num_tokens": 36175873.0,
      "step": 17430
    },
    {
      "entropy": 5.489364194869995,
      "epoch": 1.5815493468795356,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047533083870024575,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20992728173732758,
      "num_tokens": 36186792.0,
      "step": 17435
    },
    {
      "entropy": 5.63928256034851,
      "epoch": 1.58200290275762,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004753161060920998,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1986440509557724,
      "num_tokens": 36196691.0,
      "step": 17440
    },
    {
      "entropy": 5.606499433517456,
      "epoch": 1.582456458635704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004753013693413657,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20787551999092102,
      "num_tokens": 36207173.0,
      "step": 17445
    },
    {
      "entropy": 5.48338131904602,
      "epoch": 1.582910014513788,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047528662844834827,
      "loss": 5.029,
      "mean_token_accuracy": 0.21019935011863708,
      "num_tokens": 36217912.0,
      "step": 17450
    },
    {
      "entropy": 5.420902919769287,
      "epoch": 1.5833635703918723,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047527188341335233,
      "loss": 4.9371,
      "mean_token_accuracy": 0.21276613771915437,
      "num_tokens": 36227714.0,
      "step": 17455
    },
    {
      "entropy": 5.499198532104492,
      "epoch": 1.5838171262699565,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004752571342366828,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2063855767250061,
      "num_tokens": 36239114.0,
      "step": 17460
    },
    {
      "entropy": 5.530270004272461,
      "epoch": 1.5842706821480408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004752423809186446,
      "loss": 5.0342,
      "mean_token_accuracy": 0.19996798485517503,
      "num_tokens": 36249436.0,
      "step": 17465
    },
    {
      "entropy": 5.529721927642822,
      "epoch": 1.5847242380261248,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047522762345954305,
      "loss": 4.9974,
      "mean_token_accuracy": 0.20384032130241395,
      "num_tokens": 36259541.0,
      "step": 17470
    },
    {
      "entropy": 5.571310377120971,
      "epoch": 1.5851777939042089,
      "grad_norm": 1.0,
      "learning_rate": 0.0004752128618596832,
      "loss": 5.046,
      "mean_token_accuracy": 0.19789719432592393,
      "num_tokens": 36270689.0,
      "step": 17475
    },
    {
      "entropy": 5.60941514968872,
      "epoch": 1.5856313497822931,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047519809611937036,
      "loss": 5.0863,
      "mean_token_accuracy": 0.2013432890176773,
      "num_tokens": 36281191.0,
      "step": 17480
    },
    {
      "entropy": 5.47082896232605,
      "epoch": 1.5860849056603774,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751833262389098,
      "loss": 4.9994,
      "mean_token_accuracy": 0.20849235653877257,
      "num_tokens": 36291022.0,
      "step": 17485
    },
    {
      "entropy": 5.536634874343872,
      "epoch": 1.5865384615384617,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047516855221860715,
      "loss": 5.0558,
      "mean_token_accuracy": 0.2028082311153412,
      "num_tokens": 36301237.0,
      "step": 17490
    },
    {
      "entropy": 5.563267469406128,
      "epoch": 1.5869920174165457,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047515377405876774,
      "loss": 5.021,
      "mean_token_accuracy": 0.1983124077320099,
      "num_tokens": 36311792.0,
      "step": 17495
    },
    {
      "entropy": 5.497449779510498,
      "epoch": 1.5874455732946298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004751389917596973,
      "loss": 5.003,
      "mean_token_accuracy": 0.2117833748459816,
      "num_tokens": 36321847.0,
      "step": 17500
    },
    {
      "entropy": 5.534449243545533,
      "epoch": 1.587899129172714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047512420532170154,
      "loss": 5.1105,
      "mean_token_accuracy": 0.19348909854888915,
      "num_tokens": 36332735.0,
      "step": 17505
    },
    {
      "entropy": 5.604989242553711,
      "epoch": 1.5883526850507983,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751094147450862,
      "loss": 5.1031,
      "mean_token_accuracy": 0.19945639073848725,
      "num_tokens": 36343149.0,
      "step": 17510
    },
    {
      "entropy": 5.57188925743103,
      "epoch": 1.5888062409288826,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004750946200301571,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20454846769571305,
      "num_tokens": 36353081.0,
      "step": 17515
    },
    {
      "entropy": 5.513240242004395,
      "epoch": 1.5892597968069666,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047507982117722035,
      "loss": 5.0678,
      "mean_token_accuracy": 0.1978520154953003,
      "num_tokens": 36363624.0,
      "step": 17520
    },
    {
      "entropy": 5.562574768066407,
      "epoch": 1.5897133526850507,
      "grad_norm": 1.03125,
      "learning_rate": 0.000475065018186582,
      "loss": 5.0985,
      "mean_token_accuracy": 0.19767258316278458,
      "num_tokens": 36374081.0,
      "step": 17525
    },
    {
      "entropy": 5.577431631088257,
      "epoch": 1.590166908563135,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047505021105854794,
      "loss": 5.1493,
      "mean_token_accuracy": 0.19680466055870055,
      "num_tokens": 36384360.0,
      "step": 17530
    },
    {
      "entropy": 5.541051292419434,
      "epoch": 1.5906204644412192,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047503539979342456,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20463062524795533,
      "num_tokens": 36395924.0,
      "step": 17535
    },
    {
      "entropy": 5.468084049224854,
      "epoch": 1.5910740203193035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004750205843915182,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21147719472646714,
      "num_tokens": 36405246.0,
      "step": 17540
    },
    {
      "entropy": 5.489252662658691,
      "epoch": 1.5915275761973875,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004750057648531351,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20914095193147658,
      "num_tokens": 36414964.0,
      "step": 17545
    },
    {
      "entropy": 5.516128301620483,
      "epoch": 1.5919811320754715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004749909411785819,
      "loss": 5.0587,
      "mean_token_accuracy": 0.19703762233257294,
      "num_tokens": 36425011.0,
      "step": 17550
    },
    {
      "entropy": 5.495929622650147,
      "epoch": 1.5924346879535558,
      "grad_norm": 1.078125,
      "learning_rate": 0.000474976113368165,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2121456131339073,
      "num_tokens": 36434734.0,
      "step": 17555
    },
    {
      "entropy": 5.524063920974731,
      "epoch": 1.59288824383164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004749612814221912,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2045255184173584,
      "num_tokens": 36444748.0,
      "step": 17560
    },
    {
      "entropy": 5.616097640991211,
      "epoch": 1.5933417997097243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047494644534096703,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19496432393789292,
      "num_tokens": 36455076.0,
      "step": 17565
    },
    {
      "entropy": 5.543533039093018,
      "epoch": 1.5937953555878084,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004749316051247995,
      "loss": 4.9717,
      "mean_token_accuracy": 0.2077886164188385,
      "num_tokens": 36465302.0,
      "step": 17570
    },
    {
      "entropy": 5.510424327850342,
      "epoch": 1.5942489114658926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047491676077399543,
      "loss": 5.0872,
      "mean_token_accuracy": 0.19867370873689652,
      "num_tokens": 36476666.0,
      "step": 17575
    },
    {
      "entropy": 5.487771701812744,
      "epoch": 1.5947024673439767,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004749019122888618,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2082185000181198,
      "num_tokens": 36487379.0,
      "step": 17580
    },
    {
      "entropy": 5.527111959457398,
      "epoch": 1.595156023222061,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047488705966970565,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2124447599053383,
      "num_tokens": 36497897.0,
      "step": 17585
    },
    {
      "entropy": 5.54830584526062,
      "epoch": 1.5956095791001452,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004748722029168342,
      "loss": 5.0678,
      "mean_token_accuracy": 0.2106671467423439,
      "num_tokens": 36508528.0,
      "step": 17590
    },
    {
      "entropy": 5.5429380416870115,
      "epoch": 1.5960631349782293,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004748573420305546,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19697727560997008,
      "num_tokens": 36519196.0,
      "step": 17595
    },
    {
      "entropy": 5.5159355163574215,
      "epoch": 1.5965166908563135,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047484247701117427,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20312255024909973,
      "num_tokens": 36529871.0,
      "step": 17600
    },
    {
      "entropy": 5.626338863372803,
      "epoch": 1.5969702467343976,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004748276078590006,
      "loss": 5.1236,
      "mean_token_accuracy": 0.1941250815987587,
      "num_tokens": 36541188.0,
      "step": 17605
    },
    {
      "entropy": 5.58647198677063,
      "epoch": 1.5974238026124818,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004748127345743411,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19909974187612534,
      "num_tokens": 36551968.0,
      "step": 17610
    },
    {
      "entropy": 5.530744457244873,
      "epoch": 1.5978773584905661,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004747978571575033,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20223728865385054,
      "num_tokens": 36562870.0,
      "step": 17615
    },
    {
      "entropy": 5.479086828231812,
      "epoch": 1.5983309143686504,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004747829756087949,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20745939463377,
      "num_tokens": 36573295.0,
      "step": 17620
    },
    {
      "entropy": 5.479890394210815,
      "epoch": 1.5987844702467344,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047476808992852364,
      "loss": 4.9653,
      "mean_token_accuracy": 0.20873234122991563,
      "num_tokens": 36583830.0,
      "step": 17625
    },
    {
      "entropy": 5.517432165145874,
      "epoch": 1.5992380261248185,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004747532001169974,
      "loss": 5.0462,
      "mean_token_accuracy": 0.2130327671766281,
      "num_tokens": 36594052.0,
      "step": 17630
    },
    {
      "entropy": 5.520537948608398,
      "epoch": 1.5996915820029027,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047473830617452415,
      "loss": 5.0696,
      "mean_token_accuracy": 0.20406648516654968,
      "num_tokens": 36604378.0,
      "step": 17635
    },
    {
      "entropy": 5.596294307708741,
      "epoch": 1.600145137880987,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047472340810141163,
      "loss": 5.1623,
      "mean_token_accuracy": 0.19078289419412614,
      "num_tokens": 36615588.0,
      "step": 17640
    },
    {
      "entropy": 5.626055479049683,
      "epoch": 1.6005986937590713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047470850589796833,
      "loss": 5.0402,
      "mean_token_accuracy": 0.1986627995967865,
      "num_tokens": 36626369.0,
      "step": 17645
    },
    {
      "entropy": 5.5786247730255125,
      "epoch": 1.6010522496371553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047469359956450215,
      "loss": 5.0791,
      "mean_token_accuracy": 0.1997091665863991,
      "num_tokens": 36636565.0,
      "step": 17650
    },
    {
      "entropy": 5.611868000030517,
      "epoch": 1.6015058055152394,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047467868910132144,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1966838225722313,
      "num_tokens": 36647721.0,
      "step": 17655
    },
    {
      "entropy": 5.495266342163086,
      "epoch": 1.6019593613933236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004746637745087346,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21024642437696456,
      "num_tokens": 36657887.0,
      "step": 17660
    },
    {
      "entropy": 5.604389238357544,
      "epoch": 1.6024129172714079,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047464885578705,
      "loss": 5.0494,
      "mean_token_accuracy": 0.19974400252103805,
      "num_tokens": 36667918.0,
      "step": 17665
    },
    {
      "entropy": 5.472668647766113,
      "epoch": 1.6028664731494922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047463393293657616,
      "loss": 4.9976,
      "mean_token_accuracy": 0.2072445273399353,
      "num_tokens": 36678491.0,
      "step": 17670
    },
    {
      "entropy": 5.5294928550720215,
      "epoch": 1.6033200290275762,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004746190059576217,
      "loss": 5.0594,
      "mean_token_accuracy": 0.2025817811489105,
      "num_tokens": 36688638.0,
      "step": 17675
    },
    {
      "entropy": 5.555827474594116,
      "epoch": 1.6037735849056602,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004746040748504954,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20663116574287416,
      "num_tokens": 36699344.0,
      "step": 17680
    },
    {
      "entropy": 5.555319690704346,
      "epoch": 1.6042271407837445,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047458913961550593,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20222714096307753,
      "num_tokens": 36709606.0,
      "step": 17685
    },
    {
      "entropy": 5.463090229034424,
      "epoch": 1.6046806966618288,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004745742002529622,
      "loss": 4.9625,
      "mean_token_accuracy": 0.20935799330472946,
      "num_tokens": 36718740.0,
      "step": 17690
    },
    {
      "entropy": 5.522600412368774,
      "epoch": 1.605134252539913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004745592567631731,
      "loss": 5.0515,
      "mean_token_accuracy": 0.19677862673997878,
      "num_tokens": 36729960.0,
      "step": 17695
    },
    {
      "entropy": 5.4693926811218265,
      "epoch": 1.605587808417997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047454430914644773,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21306394040584564,
      "num_tokens": 36740221.0,
      "step": 17700
    },
    {
      "entropy": 5.427061891555786,
      "epoch": 1.6060413642960811,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004745293574030952,
      "loss": 4.9615,
      "mean_token_accuracy": 0.213801009953022,
      "num_tokens": 36750841.0,
      "step": 17705
    },
    {
      "entropy": 5.497584199905395,
      "epoch": 1.6064949201741654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047451440153342477,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19926677495241166,
      "num_tokens": 36760944.0,
      "step": 17710
    },
    {
      "entropy": 5.571436548233033,
      "epoch": 1.6069484760522497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004744994415377457,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20537643432617186,
      "num_tokens": 36770922.0,
      "step": 17715
    },
    {
      "entropy": 5.579338598251343,
      "epoch": 1.607402031930334,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047448447741636724,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20064235776662825,
      "num_tokens": 36780702.0,
      "step": 17720
    },
    {
      "entropy": 5.577346754074097,
      "epoch": 1.607855587808418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047446950916959905,
      "loss": 5.082,
      "mean_token_accuracy": 0.19528306573629378,
      "num_tokens": 36791868.0,
      "step": 17725
    },
    {
      "entropy": 5.585596323013306,
      "epoch": 1.608309143686502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004744545367977505,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2042024776339531,
      "num_tokens": 36802330.0,
      "step": 17730
    },
    {
      "entropy": 5.558025169372558,
      "epoch": 1.6087626995645863,
      "grad_norm": 1.125,
      "learning_rate": 0.00047443956030113134,
      "loss": 5.0891,
      "mean_token_accuracy": 0.1996786117553711,
      "num_tokens": 36812936.0,
      "step": 17735
    },
    {
      "entropy": 5.577855634689331,
      "epoch": 1.6092162554426706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047442457968005123,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20214971750974656,
      "num_tokens": 36823123.0,
      "step": 17740
    },
    {
      "entropy": 5.4617472171783445,
      "epoch": 1.6096698113207548,
      "grad_norm": 1.125,
      "learning_rate": 0.00047440959493481996,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2160551816225052,
      "num_tokens": 36833050.0,
      "step": 17745
    },
    {
      "entropy": 5.5303796291351315,
      "epoch": 1.6101233671988389,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047439460606574745,
      "loss": 5.0548,
      "mean_token_accuracy": 0.20608482956886293,
      "num_tokens": 36842520.0,
      "step": 17750
    },
    {
      "entropy": 5.546290588378906,
      "epoch": 1.6105769230769231,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004743796130731437,
      "loss": 5.0162,
      "mean_token_accuracy": 0.2015488028526306,
      "num_tokens": 36853024.0,
      "step": 17755
    },
    {
      "entropy": 5.547762823104859,
      "epoch": 1.6110304789550072,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004743646159573188,
      "loss": 4.9728,
      "mean_token_accuracy": 0.21105591356754302,
      "num_tokens": 36862534.0,
      "step": 17760
    },
    {
      "entropy": 5.516180324554443,
      "epoch": 1.6114840348330914,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004743496147185828,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21105688959360122,
      "num_tokens": 36872356.0,
      "step": 17765
    },
    {
      "entropy": 5.461830472946167,
      "epoch": 1.6119375907111757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047433460935724595,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20244371742010117,
      "num_tokens": 36882643.0,
      "step": 17770
    },
    {
      "entropy": 5.5579465389251705,
      "epoch": 1.6123911465892597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004743195998736186,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20161498188972474,
      "num_tokens": 36893349.0,
      "step": 17775
    },
    {
      "entropy": 5.6073261260986325,
      "epoch": 1.612844702467344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047430458626801107,
      "loss": 5.0289,
      "mean_token_accuracy": 0.20888873934745789,
      "num_tokens": 36903713.0,
      "step": 17780
    },
    {
      "entropy": 5.512869548797608,
      "epoch": 1.613298258345428,
      "grad_norm": 1.109375,
      "learning_rate": 0.000474289568540734,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21190437525510789,
      "num_tokens": 36912999.0,
      "step": 17785
    },
    {
      "entropy": 5.560475158691406,
      "epoch": 1.6137518142235123,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047427454669209785,
      "loss": 5.0708,
      "mean_token_accuracy": 0.2030900776386261,
      "num_tokens": 36923598.0,
      "step": 17790
    },
    {
      "entropy": 5.521582317352295,
      "epoch": 1.6142053701015966,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047425952072241326,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20506981611251832,
      "num_tokens": 36933036.0,
      "step": 17795
    },
    {
      "entropy": 5.536417770385742,
      "epoch": 1.6146589259796809,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000474244490631991,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19817693680524825,
      "num_tokens": 36943137.0,
      "step": 17800
    },
    {
      "entropy": 5.649659061431885,
      "epoch": 1.615112481857765,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004742294564211419,
      "loss": 5.2031,
      "mean_token_accuracy": 0.19474415481090546,
      "num_tokens": 36955619.0,
      "step": 17805
    },
    {
      "entropy": 5.543700122833252,
      "epoch": 1.615566037735849,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004742144180901769,
      "loss": 4.973,
      "mean_token_accuracy": 0.21407785415649414,
      "num_tokens": 36966377.0,
      "step": 17810
    },
    {
      "entropy": 5.564962434768677,
      "epoch": 1.6160195936139332,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047419937563940697,
      "loss": 5.0333,
      "mean_token_accuracy": 0.2056509166955948,
      "num_tokens": 36977100.0,
      "step": 17815
    },
    {
      "entropy": 5.557198762893677,
      "epoch": 1.6164731494920175,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047418432906914314,
      "loss": 5.1014,
      "mean_token_accuracy": 0.20024712532758712,
      "num_tokens": 36987120.0,
      "step": 17820
    },
    {
      "entropy": 5.593332052230835,
      "epoch": 1.6169267053701017,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004741692783796967,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20203897207975388,
      "num_tokens": 36996885.0,
      "step": 17825
    },
    {
      "entropy": 5.544119739532471,
      "epoch": 1.6173802612481858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047415422357137876,
      "loss": 4.9721,
      "mean_token_accuracy": 0.21137814223766327,
      "num_tokens": 37006570.0,
      "step": 17830
    },
    {
      "entropy": 5.594113206863403,
      "epoch": 1.6178338171262698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004741391646445008,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20022266656160354,
      "num_tokens": 37017185.0,
      "step": 17835
    },
    {
      "entropy": 5.547800064086914,
      "epoch": 1.618287373004354,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004741241015993741,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20494076907634734,
      "num_tokens": 37028386.0,
      "step": 17840
    },
    {
      "entropy": 5.493295669555664,
      "epoch": 1.6187409288824384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047410903443631026,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20483069121837616,
      "num_tokens": 37038326.0,
      "step": 17845
    },
    {
      "entropy": 5.548564386367798,
      "epoch": 1.6191944847605226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004740939631556209,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20101993530988693,
      "num_tokens": 37048570.0,
      "step": 17850
    },
    {
      "entropy": 5.505143356323242,
      "epoch": 1.6196480406386067,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004740788877576176,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2146276816725731,
      "num_tokens": 37058909.0,
      "step": 17855
    },
    {
      "entropy": 5.533849716186523,
      "epoch": 1.6201015965166907,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004740638082426121,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21495533287525176,
      "num_tokens": 37068957.0,
      "step": 17860
    },
    {
      "entropy": 5.539355087280273,
      "epoch": 1.620555152394775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047404872461091643,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20215528160333635,
      "num_tokens": 37080187.0,
      "step": 17865
    },
    {
      "entropy": 5.5432701587677,
      "epoch": 1.6210087082728593,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047403363686284236,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20310061573982238,
      "num_tokens": 37090314.0,
      "step": 17870
    },
    {
      "entropy": 5.5888636112213135,
      "epoch": 1.6214622641509435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047401854499870197,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20264121890068054,
      "num_tokens": 37099907.0,
      "step": 17875
    },
    {
      "entropy": 5.561094284057617,
      "epoch": 1.6219158200290276,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047400344901880735,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1975569471716881,
      "num_tokens": 37110495.0,
      "step": 17880
    },
    {
      "entropy": 5.553846216201782,
      "epoch": 1.6223693759071116,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004739883489234707,
      "loss": 5.065,
      "mean_token_accuracy": 0.19802882373332978,
      "num_tokens": 37121957.0,
      "step": 17885
    },
    {
      "entropy": 5.539548778533936,
      "epoch": 1.6228229317851959,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047397324471300424,
      "loss": 4.9772,
      "mean_token_accuracy": 0.210438971221447,
      "num_tokens": 37132743.0,
      "step": 17890
    },
    {
      "entropy": 5.562828397750854,
      "epoch": 1.6232764876632801,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047395813638772047,
      "loss": 5.0776,
      "mean_token_accuracy": 0.1967625856399536,
      "num_tokens": 37143947.0,
      "step": 17895
    },
    {
      "entropy": 5.6736279964447025,
      "epoch": 1.6237300435413644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047394302394793164,
      "loss": 5.172,
      "mean_token_accuracy": 0.1933992624282837,
      "num_tokens": 37154316.0,
      "step": 17900
    },
    {
      "entropy": 5.604281616210938,
      "epoch": 1.6241835994194485,
      "grad_norm": 1.0,
      "learning_rate": 0.0004739279073939504,
      "loss": 5.073,
      "mean_token_accuracy": 0.20024010837078093,
      "num_tokens": 37164999.0,
      "step": 17905
    },
    {
      "entropy": 5.56506929397583,
      "epoch": 1.6246371552975325,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004739127867260894,
      "loss": 5.075,
      "mean_token_accuracy": 0.20138372778892516,
      "num_tokens": 37174392.0,
      "step": 17910
    },
    {
      "entropy": 5.48519606590271,
      "epoch": 1.6250907111756168,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004738976619446613,
      "loss": 5.056,
      "mean_token_accuracy": 0.20201866924762726,
      "num_tokens": 37185341.0,
      "step": 17915
    },
    {
      "entropy": 5.580339765548706,
      "epoch": 1.625544267053701,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047388253304997874,
      "loss": 5.0779,
      "mean_token_accuracy": 0.1985635995864868,
      "num_tokens": 37196725.0,
      "step": 17920
    },
    {
      "entropy": 5.555712938308716,
      "epoch": 1.6259978229317853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004738674000423548,
      "loss": 4.9466,
      "mean_token_accuracy": 0.20997316688299178,
      "num_tokens": 37206877.0,
      "step": 17925
    },
    {
      "entropy": 5.458779954910279,
      "epoch": 1.6264513788098693,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004738522629221023,
      "loss": 4.9571,
      "mean_token_accuracy": 0.20690305680036544,
      "num_tokens": 37217277.0,
      "step": 17930
    },
    {
      "entropy": 5.475347471237183,
      "epoch": 1.6269049346879536,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004738371216895344,
      "loss": 4.934,
      "mean_token_accuracy": 0.2105925515294075,
      "num_tokens": 37227716.0,
      "step": 17935
    },
    {
      "entropy": 5.506228351593018,
      "epoch": 1.6273584905660377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004738219763449641,
      "loss": 5.047,
      "mean_token_accuracy": 0.20263891220092772,
      "num_tokens": 37238178.0,
      "step": 17940
    },
    {
      "entropy": 5.525771284103394,
      "epoch": 1.627812046444122,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004738068268887047,
      "loss": 5.0732,
      "mean_token_accuracy": 0.2064121648669243,
      "num_tokens": 37248408.0,
      "step": 17945
    },
    {
      "entropy": 5.564480209350586,
      "epoch": 1.6282656023222062,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004737916733210694,
      "loss": 5.1805,
      "mean_token_accuracy": 0.1948343485593796,
      "num_tokens": 37259176.0,
      "step": 17950
    },
    {
      "entropy": 5.55413498878479,
      "epoch": 1.6287191582002902,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047377651564237165,
      "loss": 4.9403,
      "mean_token_accuracy": 0.21000248938798904,
      "num_tokens": 37269205.0,
      "step": 17955
    },
    {
      "entropy": 5.556760597229004,
      "epoch": 1.6291727140783745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047376135385292494,
      "loss": 5.0615,
      "mean_token_accuracy": 0.20742124766111375,
      "num_tokens": 37278841.0,
      "step": 17960
    },
    {
      "entropy": 5.571680688858033,
      "epoch": 1.6296262699564585,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004737461879530428,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20085595697164535,
      "num_tokens": 37288487.0,
      "step": 17965
    },
    {
      "entropy": 5.593708944320679,
      "epoch": 1.6300798258345428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047373101794303884,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2009998604655266,
      "num_tokens": 37300086.0,
      "step": 17970
    },
    {
      "entropy": 5.503368282318116,
      "epoch": 1.630533381712627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004737158438232267,
      "loss": 5.0912,
      "mean_token_accuracy": 0.19563519060611725,
      "num_tokens": 37310072.0,
      "step": 17975
    },
    {
      "entropy": 5.5453534603118895,
      "epoch": 1.6309869375907113,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047370066559392033,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21010988503694533,
      "num_tokens": 37320278.0,
      "step": 17980
    },
    {
      "entropy": 5.560372066497803,
      "epoch": 1.6314404934687954,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047368548325543366,
      "loss": 4.9913,
      "mean_token_accuracy": 0.19833416491746902,
      "num_tokens": 37330741.0,
      "step": 17985
    },
    {
      "entropy": 5.562993144989013,
      "epoch": 1.6318940493468794,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004736702968080804,
      "loss": 5.076,
      "mean_token_accuracy": 0.19929848611354828,
      "num_tokens": 37341213.0,
      "step": 17990
    },
    {
      "entropy": 5.496298837661743,
      "epoch": 1.6323476052249637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047365510625217483,
      "loss": 4.9298,
      "mean_token_accuracy": 0.20807789713144303,
      "num_tokens": 37350528.0,
      "step": 17995
    },
    {
      "entropy": 5.528117418289185,
      "epoch": 1.632801161103048,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047363991158803117,
      "loss": 5.0309,
      "mean_token_accuracy": 0.20054963678121568,
      "num_tokens": 37359938.0,
      "step": 18000
    },
    {
      "epoch": 1.632801161103048,
      "eval_entropy": 5.307341233098129,
      "eval_loss": 5.06912899017334,
      "eval_mean_token_accuracy": 0.21036261755180843,
      "eval_num_tokens": 37359938.0,
      "eval_runtime": 20.6219,
      "eval_samples_per_second": 1714.681,
      "eval_steps_per_second": 214.335,
      "step": 18000
    },
    {
      "entropy": 5.560543966293335,
      "epoch": 1.6332547169811322,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047362471281596334,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20620202720165254,
      "num_tokens": 37369744.0,
      "step": 18005
    },
    {
      "entropy": 5.596627044677734,
      "epoch": 1.6337082728592163,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047360950993628585,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20341364592313765,
      "num_tokens": 37380578.0,
      "step": 18010
    },
    {
      "entropy": 5.550653743743896,
      "epoch": 1.6341618287373003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004735943029493132,
      "loss": 5.0317,
      "mean_token_accuracy": 0.2036425366997719,
      "num_tokens": 37390237.0,
      "step": 18015
    },
    {
      "entropy": 5.574190616607666,
      "epoch": 1.6346153846153846,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047357909185535965,
      "loss": 5.0872,
      "mean_token_accuracy": 0.1972410872578621,
      "num_tokens": 37400261.0,
      "step": 18020
    },
    {
      "entropy": 5.560031843185425,
      "epoch": 1.6350689404934688,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004735638766547399,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2051742523908615,
      "num_tokens": 37410012.0,
      "step": 18025
    },
    {
      "entropy": 5.576835584640503,
      "epoch": 1.6355224963715531,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047354865734776857,
      "loss": 5.1316,
      "mean_token_accuracy": 0.192862069606781,
      "num_tokens": 37420530.0,
      "step": 18030
    },
    {
      "entropy": 5.5550755023956295,
      "epoch": 1.6359760522496372,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004735334339347605,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21534911990165712,
      "num_tokens": 37431146.0,
      "step": 18035
    },
    {
      "entropy": 5.548226690292358,
      "epoch": 1.6364296081277212,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047351820641603035,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20894050747156143,
      "num_tokens": 37441253.0,
      "step": 18040
    },
    {
      "entropy": 5.463728237152099,
      "epoch": 1.6368831640058055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004735029747918931,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20038239955902098,
      "num_tokens": 37451326.0,
      "step": 18045
    },
    {
      "entropy": 5.466217184066773,
      "epoch": 1.6373367198838897,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047348773906266374,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2092629000544548,
      "num_tokens": 37462462.0,
      "step": 18050
    },
    {
      "entropy": 5.486484622955322,
      "epoch": 1.637790275761974,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004734724992286574,
      "loss": 4.9832,
      "mean_token_accuracy": 0.20355282127857208,
      "num_tokens": 37473016.0,
      "step": 18055
    },
    {
      "entropy": 5.569339752197266,
      "epoch": 1.638243831640058,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004734572552901892,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20118896216154097,
      "num_tokens": 37483737.0,
      "step": 18060
    },
    {
      "entropy": 5.526478099822998,
      "epoch": 1.638697387518142,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004734420072475743,
      "loss": 4.9928,
      "mean_token_accuracy": 0.20455218851566315,
      "num_tokens": 37493070.0,
      "step": 18065
    },
    {
      "entropy": 5.51882038116455,
      "epoch": 1.6391509433962264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004734267551011282,
      "loss": 4.995,
      "mean_token_accuracy": 0.20797305405139924,
      "num_tokens": 37503696.0,
      "step": 18070
    },
    {
      "entropy": 5.495640802383423,
      "epoch": 1.6396044992743106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004734114988511662,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20355913639068604,
      "num_tokens": 37513468.0,
      "step": 18075
    },
    {
      "entropy": 5.449446725845337,
      "epoch": 1.640058055152395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004733962384980039,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2130780801177025,
      "num_tokens": 37523631.0,
      "step": 18080
    },
    {
      "entropy": 5.569212102890015,
      "epoch": 1.640511611030479,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004733809740419568,
      "loss": 5.0759,
      "mean_token_accuracy": 0.20668964982032775,
      "num_tokens": 37534602.0,
      "step": 18085
    },
    {
      "entropy": 5.457507658004761,
      "epoch": 1.640965166908563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047336570548334053,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2021588295698166,
      "num_tokens": 37544547.0,
      "step": 18090
    },
    {
      "entropy": 5.548616027832031,
      "epoch": 1.6414187227866472,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047335043282247105,
      "loss": 5.0301,
      "mean_token_accuracy": 0.20326164364814758,
      "num_tokens": 37554903.0,
      "step": 18095
    },
    {
      "entropy": 5.55193772315979,
      "epoch": 1.6418722786647315,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047333515605966396,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2056132212281227,
      "num_tokens": 37565586.0,
      "step": 18100
    },
    {
      "entropy": 5.536584043502808,
      "epoch": 1.6423258345428158,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047331987519523533,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2078459829092026,
      "num_tokens": 37575996.0,
      "step": 18105
    },
    {
      "entropy": 5.482731580734253,
      "epoch": 1.6427793904208998,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004733045902295012,
      "loss": 4.941,
      "mean_token_accuracy": 0.2008867606520653,
      "num_tokens": 37586441.0,
      "step": 18110
    },
    {
      "entropy": 5.551346015930176,
      "epoch": 1.643232946298984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047328930116277755,
      "loss": 5.055,
      "mean_token_accuracy": 0.2052132710814476,
      "num_tokens": 37598226.0,
      "step": 18115
    },
    {
      "entropy": 5.542083644866944,
      "epoch": 1.6436865021770681,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004732740079953806,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20656532049179077,
      "num_tokens": 37609102.0,
      "step": 18120
    },
    {
      "entropy": 5.545897531509399,
      "epoch": 1.6441400580551524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004732587107276267,
      "loss": 5.0499,
      "mean_token_accuracy": 0.2018830582499504,
      "num_tokens": 37619349.0,
      "step": 18125
    },
    {
      "entropy": 5.512788105010986,
      "epoch": 1.6445936139332367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047324340935983225,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20966263711452485,
      "num_tokens": 37629844.0,
      "step": 18130
    },
    {
      "entropy": 5.482539653778076,
      "epoch": 1.6450471698113207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047322810389231333,
      "loss": 4.9394,
      "mean_token_accuracy": 0.20833370983600616,
      "num_tokens": 37639339.0,
      "step": 18135
    },
    {
      "entropy": 5.502611827850342,
      "epoch": 1.645500725689405,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004732127943253868,
      "loss": 4.978,
      "mean_token_accuracy": 0.21004469245672225,
      "num_tokens": 37648836.0,
      "step": 18140
    },
    {
      "entropy": 5.584582233428955,
      "epoch": 1.645954281567489,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004731974806593693,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2051238164305687,
      "num_tokens": 37658424.0,
      "step": 18145
    },
    {
      "entropy": 5.549859142303466,
      "epoch": 1.6464078374455733,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047318216289457727,
      "loss": 5.0074,
      "mean_token_accuracy": 0.21039813309907912,
      "num_tokens": 37669849.0,
      "step": 18150
    },
    {
      "entropy": 5.505876684188843,
      "epoch": 1.6468613933236576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047316684103132766,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20256547927856444,
      "num_tokens": 37678987.0,
      "step": 18155
    },
    {
      "entropy": 5.555306529998779,
      "epoch": 1.6473149492017418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004731515150699372,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20816829204559326,
      "num_tokens": 37689692.0,
      "step": 18160
    },
    {
      "entropy": 5.585298776626587,
      "epoch": 1.6477685050798259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047313618501072303,
      "loss": 5.1711,
      "mean_token_accuracy": 0.19286765158176422,
      "num_tokens": 37700092.0,
      "step": 18165
    },
    {
      "entropy": 5.5218486309051515,
      "epoch": 1.64822206095791,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047312085085400194,
      "loss": 5.065,
      "mean_token_accuracy": 0.20425543636083604,
      "num_tokens": 37710305.0,
      "step": 18170
    },
    {
      "entropy": 5.658654689788818,
      "epoch": 1.6486756168359942,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047310551260009126,
      "loss": 5.1307,
      "mean_token_accuracy": 0.1929813876748085,
      "num_tokens": 37720451.0,
      "step": 18175
    },
    {
      "entropy": 5.577204895019531,
      "epoch": 1.6491291727140784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000473090170249308,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20589849203824998,
      "num_tokens": 37731952.0,
      "step": 18180
    },
    {
      "entropy": 5.489793586730957,
      "epoch": 1.6495827285921627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004730748238019696,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20154588669538498,
      "num_tokens": 37742201.0,
      "step": 18185
    },
    {
      "entropy": 5.474636507034302,
      "epoch": 1.6500362844702468,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047305947325839337,
      "loss": 4.9928,
      "mean_token_accuracy": 0.20665855407714845,
      "num_tokens": 37753317.0,
      "step": 18190
    },
    {
      "entropy": 5.58493447303772,
      "epoch": 1.6504898403483308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004730441186188968,
      "loss": 5.0549,
      "mean_token_accuracy": 0.2018160879611969,
      "num_tokens": 37763452.0,
      "step": 18195
    },
    {
      "entropy": 5.551719331741333,
      "epoch": 1.650943396226415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047302875988379734,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2069983720779419,
      "num_tokens": 37772804.0,
      "step": 18200
    },
    {
      "entropy": 5.548997497558593,
      "epoch": 1.6513969521044993,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047301339705341266,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19575041085481643,
      "num_tokens": 37784441.0,
      "step": 18205
    },
    {
      "entropy": 5.529232168197632,
      "epoch": 1.6518505079825836,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047299803012806056,
      "loss": 5.0436,
      "mean_token_accuracy": 0.2065920427441597,
      "num_tokens": 37795103.0,
      "step": 18210
    },
    {
      "entropy": 5.487066984176636,
      "epoch": 1.6523040638606676,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004729826591080586,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20657159388065338,
      "num_tokens": 37806654.0,
      "step": 18215
    },
    {
      "entropy": 5.526598072052002,
      "epoch": 1.6527576197387517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047296728399372483,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20016657412052155,
      "num_tokens": 37816673.0,
      "step": 18220
    },
    {
      "entropy": 5.555908298492431,
      "epoch": 1.653211175616836,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047295190478537725,
      "loss": 5.0503,
      "mean_token_accuracy": 0.2029751569032669,
      "num_tokens": 37827389.0,
      "step": 18225
    },
    {
      "entropy": 5.658416748046875,
      "epoch": 1.6536647314949202,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047293652148333376,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20344164222478867,
      "num_tokens": 37838087.0,
      "step": 18230
    },
    {
      "entropy": 5.4327747344970705,
      "epoch": 1.6541182873730045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047292113408791266,
      "loss": 4.9477,
      "mean_token_accuracy": 0.20138545483350753,
      "num_tokens": 37848630.0,
      "step": 18235
    },
    {
      "entropy": 5.453988885879516,
      "epoch": 1.6545718432510885,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047290574259943195,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22004650682210922,
      "num_tokens": 37858689.0,
      "step": 18240
    },
    {
      "entropy": 5.6077797412872314,
      "epoch": 1.6550253991291726,
      "grad_norm": 1.0,
      "learning_rate": 0.00047289034701821013,
      "loss": 5.0897,
      "mean_token_accuracy": 0.20234557688236238,
      "num_tokens": 37869049.0,
      "step": 18245
    },
    {
      "entropy": 5.5167817115783695,
      "epoch": 1.6554789550072568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047287494734456545,
      "loss": 4.9994,
      "mean_token_accuracy": 0.2003646969795227,
      "num_tokens": 37879495.0,
      "step": 18250
    },
    {
      "entropy": 5.544087553024292,
      "epoch": 1.655932510885341,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047285954357881646,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2015921652317047,
      "num_tokens": 37889442.0,
      "step": 18255
    },
    {
      "entropy": 5.598650503158569,
      "epoch": 1.6563860667634254,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004728441357212817,
      "loss": 5.1712,
      "mean_token_accuracy": 0.1892197161912918,
      "num_tokens": 37900147.0,
      "step": 18260
    },
    {
      "entropy": 5.6057878017425535,
      "epoch": 1.6568396226415094,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004728287237722798,
      "loss": 4.9953,
      "mean_token_accuracy": 0.20190392732620238,
      "num_tokens": 37910307.0,
      "step": 18265
    },
    {
      "entropy": 5.492194032669067,
      "epoch": 1.6572931785195935,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004728133077321295,
      "loss": 4.9683,
      "mean_token_accuracy": 0.20334598273038865,
      "num_tokens": 37920838.0,
      "step": 18270
    },
    {
      "entropy": 5.546195602416992,
      "epoch": 1.6577467343976777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047279788760114954,
      "loss": 5.0489,
      "mean_token_accuracy": 0.19957167655229568,
      "num_tokens": 37930717.0,
      "step": 18275
    },
    {
      "entropy": 5.565794944763184,
      "epoch": 1.658200290275762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004727824633796589,
      "loss": 5.0443,
      "mean_token_accuracy": 0.2027788817882538,
      "num_tokens": 37941254.0,
      "step": 18280
    },
    {
      "entropy": 5.510905408859253,
      "epoch": 1.6586538461538463,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004727670350679765,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20850671380758284,
      "num_tokens": 37952132.0,
      "step": 18285
    },
    {
      "entropy": 5.513995122909546,
      "epoch": 1.6591074020319303,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004727516026664214,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20742790400981903,
      "num_tokens": 37962026.0,
      "step": 18290
    },
    {
      "entropy": 5.5256506443023685,
      "epoch": 1.6595609579100146,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004727361661753128,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20238326042890548,
      "num_tokens": 37972911.0,
      "step": 18295
    },
    {
      "entropy": 5.5586512088775635,
      "epoch": 1.6600145137880986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004727207255949699,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20900478065013886,
      "num_tokens": 37983577.0,
      "step": 18300
    },
    {
      "entropy": 5.5725830078125,
      "epoch": 1.6604680696661829,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047270528092571197,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20097507536411285,
      "num_tokens": 37994294.0,
      "step": 18305
    },
    {
      "entropy": 5.538248825073242,
      "epoch": 1.6609216255442671,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004726898321678584,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20685731917619704,
      "num_tokens": 38005144.0,
      "step": 18310
    },
    {
      "entropy": 5.4818648338317875,
      "epoch": 1.6613751814223512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726743793217288,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20080667734146118,
      "num_tokens": 38015277.0,
      "step": 18315
    },
    {
      "entropy": 5.5463134288787845,
      "epoch": 1.6618287373004355,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047265892238764265,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20633944720029831,
      "num_tokens": 38027530.0,
      "step": 18320
    },
    {
      "entropy": 5.617530441284179,
      "epoch": 1.6622822931785195,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004726434613659196,
      "loss": 5.0891,
      "mean_token_accuracy": 0.20411063283681868,
      "num_tokens": 38036895.0,
      "step": 18325
    },
    {
      "entropy": 5.583403873443603,
      "epoch": 1.6627358490566038,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004726279962568794,
      "loss": 5.1012,
      "mean_token_accuracy": 0.2044348746538162,
      "num_tokens": 38047001.0,
      "step": 18330
    },
    {
      "entropy": 5.554082775115967,
      "epoch": 1.663189404934688,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004726125270608419,
      "loss": 4.975,
      "mean_token_accuracy": 0.2108441025018692,
      "num_tokens": 38056815.0,
      "step": 18335
    },
    {
      "entropy": 5.568605566024781,
      "epoch": 1.6636429608127723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047259705377812703,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20433802753686905,
      "num_tokens": 38067139.0,
      "step": 18340
    },
    {
      "entropy": 5.548576307296753,
      "epoch": 1.6640965166908563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004725815764090546,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20833301097154616,
      "num_tokens": 38076824.0,
      "step": 18345
    },
    {
      "entropy": 5.536114835739136,
      "epoch": 1.6645500725689404,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047256609495394494,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20151522010564804,
      "num_tokens": 38087517.0,
      "step": 18350
    },
    {
      "entropy": 5.568885850906372,
      "epoch": 1.6650036284470247,
      "grad_norm": 1.03125,
      "learning_rate": 0.000472550609413118,
      "loss": 5.0186,
      "mean_token_accuracy": 0.20044526159763337,
      "num_tokens": 38098499.0,
      "step": 18355
    },
    {
      "entropy": 5.48836989402771,
      "epoch": 1.665457184325109,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004725351197868942,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2095501482486725,
      "num_tokens": 38108327.0,
      "step": 18360
    },
    {
      "entropy": 5.598924016952514,
      "epoch": 1.6659107402031932,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047251962607559373,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20512547791004182,
      "num_tokens": 38118667.0,
      "step": 18365
    },
    {
      "entropy": 5.5005449771881105,
      "epoch": 1.6663642960812772,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047250412827953717,
      "loss": 4.9998,
      "mean_token_accuracy": 0.20354425311088561,
      "num_tokens": 38128571.0,
      "step": 18370
    },
    {
      "entropy": 5.588938045501709,
      "epoch": 1.6668178519593613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004724886263990448,
      "loss": 5.1049,
      "mean_token_accuracy": 0.195250703394413,
      "num_tokens": 38138921.0,
      "step": 18375
    },
    {
      "entropy": 5.591016340255737,
      "epoch": 1.6672714078374455,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004724731204344372,
      "loss": 5.0368,
      "mean_token_accuracy": 0.20534040927886962,
      "num_tokens": 38148858.0,
      "step": 18380
    },
    {
      "entropy": 5.462696266174317,
      "epoch": 1.6677249637155298,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047245761038603536,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20529389679431914,
      "num_tokens": 38159107.0,
      "step": 18385
    },
    {
      "entropy": 5.618238306045532,
      "epoch": 1.668178519593614,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047244209625415975,
      "loss": 5.0921,
      "mean_token_accuracy": 0.20097823292016984,
      "num_tokens": 38169228.0,
      "step": 18390
    },
    {
      "entropy": 5.603679180145264,
      "epoch": 1.6686320754716981,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004724265780391312,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2068103551864624,
      "num_tokens": 38179175.0,
      "step": 18395
    },
    {
      "entropy": 5.556495428085327,
      "epoch": 1.6690856313497822,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004724110557412708,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20790047347545623,
      "num_tokens": 38188537.0,
      "step": 18400
    },
    {
      "entropy": 5.514718437194825,
      "epoch": 1.6695391872278664,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047239552936089943,
      "loss": 5.0496,
      "mean_token_accuracy": 0.2000635713338852,
      "num_tokens": 38198972.0,
      "step": 18405
    },
    {
      "entropy": 5.53039960861206,
      "epoch": 1.6699927431059507,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047237999889833816,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2063676580786705,
      "num_tokens": 38209472.0,
      "step": 18410
    },
    {
      "entropy": 5.461790990829468,
      "epoch": 1.670446298984035,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047236446435390826,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2051655277609825,
      "num_tokens": 38219977.0,
      "step": 18415
    },
    {
      "entropy": 5.519774389266968,
      "epoch": 1.670899854862119,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047234892572793093,
      "loss": 5.0577,
      "mean_token_accuracy": 0.2033174842596054,
      "num_tokens": 38230492.0,
      "step": 18420
    },
    {
      "entropy": 5.529501247406006,
      "epoch": 1.671353410740203,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004723333830207275,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1979853630065918,
      "num_tokens": 38240799.0,
      "step": 18425
    },
    {
      "entropy": 5.518443870544433,
      "epoch": 1.6718069666182873,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047231783623261954,
      "loss": 4.9815,
      "mean_token_accuracy": 0.1997287839651108,
      "num_tokens": 38251439.0,
      "step": 18430
    },
    {
      "entropy": 5.596939945220948,
      "epoch": 1.6722605224963716,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004723022853639283,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19885122925043106,
      "num_tokens": 38261795.0,
      "step": 18435
    },
    {
      "entropy": 5.516361951828003,
      "epoch": 1.6727140783744558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004722867304149756,
      "loss": 4.9921,
      "mean_token_accuracy": 0.20650445967912673,
      "num_tokens": 38272172.0,
      "step": 18440
    },
    {
      "entropy": 5.552573108673096,
      "epoch": 1.67316763425254,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004722711713860831,
      "loss": 5.083,
      "mean_token_accuracy": 0.19582296311855316,
      "num_tokens": 38282829.0,
      "step": 18445
    },
    {
      "entropy": 5.580194520950317,
      "epoch": 1.673621190130624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004722556082775724,
      "loss": 5.0806,
      "mean_token_accuracy": 0.20674073845148086,
      "num_tokens": 38293793.0,
      "step": 18450
    },
    {
      "entropy": 5.5810974597930905,
      "epoch": 1.6740747460087082,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004722400410897655,
      "loss": 5.083,
      "mean_token_accuracy": 0.20138343721628188,
      "num_tokens": 38302577.0,
      "step": 18455
    },
    {
      "entropy": 5.53628568649292,
      "epoch": 1.6745283018867925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004722244698229842,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20656457394361497,
      "num_tokens": 38312571.0,
      "step": 18460
    },
    {
      "entropy": 5.527368402481079,
      "epoch": 1.6749818577648767,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047220889447755073,
      "loss": 4.966,
      "mean_token_accuracy": 0.2101530909538269,
      "num_tokens": 38323544.0,
      "step": 18465
    },
    {
      "entropy": 5.582181930541992,
      "epoch": 1.6754354136429608,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004721933150537869,
      "loss": 5.0528,
      "mean_token_accuracy": 0.20263455361127852,
      "num_tokens": 38334509.0,
      "step": 18470
    },
    {
      "entropy": 5.493896722793579,
      "epoch": 1.675888969521045,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047217773155201526,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20418312698602675,
      "num_tokens": 38344747.0,
      "step": 18475
    },
    {
      "entropy": 5.586251354217529,
      "epoch": 1.676342525399129,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047216214397255776,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1967146709561348,
      "num_tokens": 38355671.0,
      "step": 18480
    },
    {
      "entropy": 5.565404796600342,
      "epoch": 1.6767960812772134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047214655231573693,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20337298065423964,
      "num_tokens": 38366368.0,
      "step": 18485
    },
    {
      "entropy": 5.563398122787476,
      "epoch": 1.6772496371552976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004721309565818752,
      "loss": 5.1038,
      "mean_token_accuracy": 0.20038592368364333,
      "num_tokens": 38377322.0,
      "step": 18490
    },
    {
      "entropy": 5.608231687545777,
      "epoch": 1.6777031930333817,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047211535677129496,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20255047976970672,
      "num_tokens": 38389204.0,
      "step": 18495
    },
    {
      "entropy": 5.548760461807251,
      "epoch": 1.678156748911466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047209975288431893,
      "loss": 5.053,
      "mean_token_accuracy": 0.2019473984837532,
      "num_tokens": 38399117.0,
      "step": 18500
    },
    {
      "entropy": 5.537803888320923,
      "epoch": 1.67861030478955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047208414492126986,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2017965391278267,
      "num_tokens": 38410341.0,
      "step": 18505
    },
    {
      "entropy": 5.595143365859985,
      "epoch": 1.6790638606676342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047206853288247043,
      "loss": 5.0836,
      "mean_token_accuracy": 0.19759543389081954,
      "num_tokens": 38422742.0,
      "step": 18510
    },
    {
      "entropy": 5.56275897026062,
      "epoch": 1.6795174165457185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047205291676824345,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19663460105657576,
      "num_tokens": 38433022.0,
      "step": 18515
    },
    {
      "entropy": 5.593740558624267,
      "epoch": 1.6799709724238028,
      "grad_norm": 1.0,
      "learning_rate": 0.000472037296578912,
      "loss": 5.0561,
      "mean_token_accuracy": 0.20002095848321916,
      "num_tokens": 38444325.0,
      "step": 18520
    },
    {
      "entropy": 5.537397623062134,
      "epoch": 1.6804245283018868,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000472021672314799,
      "loss": 4.9606,
      "mean_token_accuracy": 0.20704559981822968,
      "num_tokens": 38454576.0,
      "step": 18525
    },
    {
      "entropy": 5.525679063796997,
      "epoch": 1.6808780841799709,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047200604397622763,
      "loss": 4.9752,
      "mean_token_accuracy": 0.20736278742551803,
      "num_tokens": 38465090.0,
      "step": 18530
    },
    {
      "entropy": 5.576778554916382,
      "epoch": 1.6813316400580551,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004719904115635211,
      "loss": 5.2162,
      "mean_token_accuracy": 0.1895391821861267,
      "num_tokens": 38477253.0,
      "step": 18535
    },
    {
      "entropy": 5.618865823745727,
      "epoch": 1.6817851959361394,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004719747750770026,
      "loss": 5.1115,
      "mean_token_accuracy": 0.19903451353311538,
      "num_tokens": 38488040.0,
      "step": 18540
    },
    {
      "entropy": 5.549099969863891,
      "epoch": 1.6822387518142237,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004719591345169956,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20792275220155715,
      "num_tokens": 38498016.0,
      "step": 18545
    },
    {
      "entropy": 5.586939668655395,
      "epoch": 1.6826923076923077,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047194348988382345,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20007387101650237,
      "num_tokens": 38508996.0,
      "step": 18550
    },
    {
      "entropy": 5.544355249404907,
      "epoch": 1.6831458635703918,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004719278411778097,
      "loss": 5.0735,
      "mean_token_accuracy": 0.20666104257106782,
      "num_tokens": 38519608.0,
      "step": 18555
    },
    {
      "entropy": 5.548975801467895,
      "epoch": 1.683599419448476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047191218839927817,
      "loss": 5.0226,
      "mean_token_accuracy": 0.19572516828775405,
      "num_tokens": 38530737.0,
      "step": 18560
    },
    {
      "entropy": 5.543293809890747,
      "epoch": 1.6840529753265603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004718965315485523,
      "loss": 4.9857,
      "mean_token_accuracy": 0.204212287068367,
      "num_tokens": 38540925.0,
      "step": 18565
    },
    {
      "entropy": 5.55231466293335,
      "epoch": 1.6845065312046446,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004718808706259561,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19747941792011262,
      "num_tokens": 38553176.0,
      "step": 18570
    },
    {
      "entropy": 5.599540758132934,
      "epoch": 1.6849600870827286,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047186520563181325,
      "loss": 5.1021,
      "mean_token_accuracy": 0.19440075755119324,
      "num_tokens": 38563394.0,
      "step": 18575
    },
    {
      "entropy": 5.613511562347412,
      "epoch": 1.6854136429608126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004718495365664477,
      "loss": 5.0467,
      "mean_token_accuracy": 0.20269798040390014,
      "num_tokens": 38573243.0,
      "step": 18580
    },
    {
      "entropy": 5.544748973846436,
      "epoch": 1.685867198838897,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047183386343018364,
      "loss": 5.0687,
      "mean_token_accuracy": 0.19938255101442337,
      "num_tokens": 38584014.0,
      "step": 18585
    },
    {
      "entropy": 5.572828197479248,
      "epoch": 1.6863207547169812,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004718181862233451,
      "loss": 4.996,
      "mean_token_accuracy": 0.20601304918527602,
      "num_tokens": 38594467.0,
      "step": 18590
    },
    {
      "entropy": 5.595120668411255,
      "epoch": 1.6867743105950654,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004718025049462563,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20386261492967606,
      "num_tokens": 38604407.0,
      "step": 18595
    },
    {
      "entropy": 5.581459188461304,
      "epoch": 1.6872278664731495,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004717868195992416,
      "loss": 5.0629,
      "mean_token_accuracy": 0.20675850361585618,
      "num_tokens": 38614508.0,
      "step": 18600
    },
    {
      "entropy": 5.6191541194915775,
      "epoch": 1.6876814223512335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004717711301826253,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20611653476953506,
      "num_tokens": 38626244.0,
      "step": 18605
    },
    {
      "entropy": 5.520521974563598,
      "epoch": 1.6881349782293178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047175543669673186,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20494984090328217,
      "num_tokens": 38636694.0,
      "step": 18610
    },
    {
      "entropy": 5.6410387516021725,
      "epoch": 1.688588534107402,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047173973914188587,
      "loss": 5.1134,
      "mean_token_accuracy": 0.20223284512758255,
      "num_tokens": 38647414.0,
      "step": 18615
    },
    {
      "entropy": 5.5597906589508055,
      "epoch": 1.6890420899854863,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047172403751841196,
      "loss": 5.0246,
      "mean_token_accuracy": 0.20177806913852692,
      "num_tokens": 38657425.0,
      "step": 18620
    },
    {
      "entropy": 5.58294153213501,
      "epoch": 1.6894956458635704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047170833182663485,
      "loss": 5.0429,
      "mean_token_accuracy": 0.2058634117245674,
      "num_tokens": 38667751.0,
      "step": 18625
    },
    {
      "entropy": 5.461794662475586,
      "epoch": 1.6899492017416544,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004716926220668793,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20780032575130464,
      "num_tokens": 38676918.0,
      "step": 18630
    },
    {
      "entropy": 5.486989974975586,
      "epoch": 1.6904027576197387,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047167690823947015,
      "loss": 5.025,
      "mean_token_accuracy": 0.21017662882804872,
      "num_tokens": 38685887.0,
      "step": 18635
    },
    {
      "entropy": 5.512678337097168,
      "epoch": 1.690856313497823,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004716611903447325,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20788081139326095,
      "num_tokens": 38696179.0,
      "step": 18640
    },
    {
      "entropy": 5.5653375625610355,
      "epoch": 1.6913098693759072,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004716454683829912,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20205404609441757,
      "num_tokens": 38705777.0,
      "step": 18645
    },
    {
      "entropy": 5.53246579170227,
      "epoch": 1.6917634252539913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047162974235457163,
      "loss": 5.023,
      "mean_token_accuracy": 0.20526780039072037,
      "num_tokens": 38717709.0,
      "step": 18650
    },
    {
      "entropy": 5.516905975341797,
      "epoch": 1.6922169811320755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004716140122597988,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20970158874988556,
      "num_tokens": 38729517.0,
      "step": 18655
    },
    {
      "entropy": 5.529263734817505,
      "epoch": 1.6926705370101596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004715982780989981,
      "loss": 5.032,
      "mean_token_accuracy": 0.2079072117805481,
      "num_tokens": 38740878.0,
      "step": 18660
    },
    {
      "entropy": 5.517541885375977,
      "epoch": 1.6931240928882438,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047158253987249487,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2089331105351448,
      "num_tokens": 38751291.0,
      "step": 18665
    },
    {
      "entropy": 5.509375333786011,
      "epoch": 1.693577648766328,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047156679758061465,
      "loss": 4.959,
      "mean_token_accuracy": 0.2138681709766388,
      "num_tokens": 38760940.0,
      "step": 18670
    },
    {
      "entropy": 5.549563932418823,
      "epoch": 1.6940312046444121,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047155105122368297,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21167542934417724,
      "num_tokens": 38770240.0,
      "step": 18675
    },
    {
      "entropy": 5.466352033615112,
      "epoch": 1.6944847605224964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004715353008020255,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20637205243110657,
      "num_tokens": 38780452.0,
      "step": 18680
    },
    {
      "entropy": 5.537557363510132,
      "epoch": 1.6949383164005805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047151954631596794,
      "loss": 5.0745,
      "mean_token_accuracy": 0.199966262280941,
      "num_tokens": 38790763.0,
      "step": 18685
    },
    {
      "entropy": 5.533513927459717,
      "epoch": 1.6953918722786647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047150378776583603,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21286385506391525,
      "num_tokens": 38802164.0,
      "step": 18690
    },
    {
      "entropy": 5.537321853637695,
      "epoch": 1.695845428156749,
      "grad_norm": 1.0,
      "learning_rate": 0.0004714880251519557,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20215222984552383,
      "num_tokens": 38812901.0,
      "step": 18695
    },
    {
      "entropy": 5.508861923217774,
      "epoch": 1.6962989840348333,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000471472258474653,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2096114531159401,
      "num_tokens": 38823305.0,
      "step": 18700
    },
    {
      "entropy": 5.539385175704956,
      "epoch": 1.6967525399129173,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047145648773425384,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20409904271364213,
      "num_tokens": 38833602.0,
      "step": 18705
    },
    {
      "entropy": 5.576952886581421,
      "epoch": 1.6972060957910013,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047144071293108456,
      "loss": 5.071,
      "mean_token_accuracy": 0.2042400449514389,
      "num_tokens": 38843950.0,
      "step": 18710
    },
    {
      "entropy": 5.52810320854187,
      "epoch": 1.6976596516690856,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004714249340654712,
      "loss": 4.9925,
      "mean_token_accuracy": 0.2058522433042526,
      "num_tokens": 38853753.0,
      "step": 18715
    },
    {
      "entropy": 5.467414999008179,
      "epoch": 1.6981132075471699,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004714091511377402,
      "loss": 4.9304,
      "mean_token_accuracy": 0.20456690341234207,
      "num_tokens": 38863394.0,
      "step": 18720
    },
    {
      "entropy": 5.521736574172974,
      "epoch": 1.6985667634252541,
      "grad_norm": 1.125,
      "learning_rate": 0.00047139336414821784,
      "loss": 5.024,
      "mean_token_accuracy": 0.20001382976770402,
      "num_tokens": 38872972.0,
      "step": 18725
    },
    {
      "entropy": 5.5153477668762205,
      "epoch": 1.6990203193033382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047137757309723073,
      "loss": 5.0092,
      "mean_token_accuracy": 0.212299245595932,
      "num_tokens": 38883251.0,
      "step": 18730
    },
    {
      "entropy": 5.516787052154541,
      "epoch": 1.6994738751814222,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004713617779851053,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20373117178678513,
      "num_tokens": 38893274.0,
      "step": 18735
    },
    {
      "entropy": 5.61175503730774,
      "epoch": 1.6999274310595065,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004713459788121683,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19592057019472123,
      "num_tokens": 38903870.0,
      "step": 18740
    },
    {
      "entropy": 5.572488260269165,
      "epoch": 1.7003809869375908,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004713301755787464,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1994834527373314,
      "num_tokens": 38914921.0,
      "step": 18745
    },
    {
      "entropy": 5.570324230194092,
      "epoch": 1.700834542815675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004713143682851665,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20570874065160752,
      "num_tokens": 38924630.0,
      "step": 18750
    },
    {
      "entropy": 5.477497339248657,
      "epoch": 1.701288098693759,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712985569317553,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20439738780260086,
      "num_tokens": 38934926.0,
      "step": 18755
    },
    {
      "entropy": 5.603098392486572,
      "epoch": 1.7017416545718431,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004712827415188401,
      "loss": 5.1481,
      "mean_token_accuracy": 0.19430297911167144,
      "num_tokens": 38944513.0,
      "step": 18760
    },
    {
      "entropy": 5.575858306884766,
      "epoch": 1.7021952104499274,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004712669220467476,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20788175910711287,
      "num_tokens": 38955918.0,
      "step": 18765
    },
    {
      "entropy": 5.568398094177246,
      "epoch": 1.7026487663280117,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047125109851580533,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20533353090286255,
      "num_tokens": 38966221.0,
      "step": 18770
    },
    {
      "entropy": 5.535524511337281,
      "epoch": 1.703102322206096,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047123527092634025,
      "loss": 4.9866,
      "mean_token_accuracy": 0.20195710361003877,
      "num_tokens": 38976463.0,
      "step": 18775
    },
    {
      "entropy": 5.479678440093994,
      "epoch": 1.70355587808418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004712194392786797,
      "loss": 4.9273,
      "mean_token_accuracy": 0.20722113698720931,
      "num_tokens": 38986375.0,
      "step": 18780
    },
    {
      "entropy": 5.625212955474853,
      "epoch": 1.704009433962264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047120360357315127,
      "loss": 5.064,
      "mean_token_accuracy": 0.20052340477705002,
      "num_tokens": 38996116.0,
      "step": 18785
    },
    {
      "entropy": 5.509733486175537,
      "epoch": 1.7044629898403483,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047118776381008216,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20391646027565002,
      "num_tokens": 39005846.0,
      "step": 18790
    },
    {
      "entropy": 5.573023700714112,
      "epoch": 1.7049165457184325,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047117191998980026,
      "loss": 5.0213,
      "mean_token_accuracy": 0.19692962765693664,
      "num_tokens": 39016744.0,
      "step": 18795
    },
    {
      "entropy": 5.5535383224487305,
      "epoch": 1.7053701015965168,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047115607211263295,
      "loss": 5.0736,
      "mean_token_accuracy": 0.19864329397678376,
      "num_tokens": 39026725.0,
      "step": 18800
    },
    {
      "entropy": 5.5790722370147705,
      "epoch": 1.7058236574746009,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047114022017890814,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20246905237436294,
      "num_tokens": 39036851.0,
      "step": 18805
    },
    {
      "entropy": 5.526825666427612,
      "epoch": 1.706277213352685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004711243641889536,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2043837919831276,
      "num_tokens": 39047179.0,
      "step": 18810
    },
    {
      "entropy": 5.49183235168457,
      "epoch": 1.7067307692307692,
      "grad_norm": 1.0,
      "learning_rate": 0.00047110850414309714,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2154756158590317,
      "num_tokens": 39057758.0,
      "step": 18815
    },
    {
      "entropy": 5.5515491485595705,
      "epoch": 1.7071843251088534,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047109264004166696,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20021470189094542,
      "num_tokens": 39068142.0,
      "step": 18820
    },
    {
      "entropy": 5.5086527347564695,
      "epoch": 1.7076378809869377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004710767718849909,
      "loss": 4.9844,
      "mean_token_accuracy": 0.19743564873933792,
      "num_tokens": 39078157.0,
      "step": 18825
    },
    {
      "entropy": 5.514392566680908,
      "epoch": 1.7080914368650217,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004710608996733973,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21087042093276978,
      "num_tokens": 39089090.0,
      "step": 18830
    },
    {
      "entropy": 5.553383779525757,
      "epoch": 1.708544992743106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047104502340721433,
      "loss": 5.0734,
      "mean_token_accuracy": 0.19710729867219925,
      "num_tokens": 39099079.0,
      "step": 18835
    },
    {
      "entropy": 5.547780752182007,
      "epoch": 1.70899854862119,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047102914308677033,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20297769010066985,
      "num_tokens": 39109927.0,
      "step": 18840
    },
    {
      "entropy": 5.581135034561157,
      "epoch": 1.7094521044992743,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047101325871239363,
      "loss": 5.0103,
      "mean_token_accuracy": 0.2026897057890892,
      "num_tokens": 39119806.0,
      "step": 18845
    },
    {
      "entropy": 5.597395563125611,
      "epoch": 1.7099056603773586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004709973702844128,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20226970463991165,
      "num_tokens": 39130147.0,
      "step": 18850
    },
    {
      "entropy": 5.476788091659546,
      "epoch": 1.7103592162554426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047098147780315645,
      "loss": 4.902,
      "mean_token_accuracy": 0.21620516031980513,
      "num_tokens": 39139937.0,
      "step": 18855
    },
    {
      "entropy": 5.548338937759399,
      "epoch": 1.710812772133527,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004709655812689532,
      "loss": 5.023,
      "mean_token_accuracy": 0.20639331191778182,
      "num_tokens": 39149977.0,
      "step": 18860
    },
    {
      "entropy": 5.53417010307312,
      "epoch": 1.711266328011611,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004709496806821318,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20468510836362838,
      "num_tokens": 39161076.0,
      "step": 18865
    },
    {
      "entropy": 5.57164454460144,
      "epoch": 1.7117198838896952,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047093377604302106,
      "loss": 5.0879,
      "mean_token_accuracy": 0.19266926050186156,
      "num_tokens": 39170776.0,
      "step": 18870
    },
    {
      "entropy": 5.578165769577026,
      "epoch": 1.7121734397677795,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004709178673519499,
      "loss": 5.0447,
      "mean_token_accuracy": 0.19984612464904786,
      "num_tokens": 39181134.0,
      "step": 18875
    },
    {
      "entropy": 5.5719358921051025,
      "epoch": 1.7126269956458637,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004709019546092473,
      "loss": 5.1256,
      "mean_token_accuracy": 0.19469009190797806,
      "num_tokens": 39193331.0,
      "step": 18880
    },
    {
      "entropy": 5.624257659912109,
      "epoch": 1.7130805515239478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047088603781524237,
      "loss": 5.0845,
      "mean_token_accuracy": 0.19632348269224167,
      "num_tokens": 39204372.0,
      "step": 18885
    },
    {
      "entropy": 5.576696538925171,
      "epoch": 1.7135341074020318,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004708701169702642,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2091236263513565,
      "num_tokens": 39214725.0,
      "step": 18890
    },
    {
      "entropy": 5.4728844165802,
      "epoch": 1.713987663280116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004708541920746422,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20199199467897416,
      "num_tokens": 39225119.0,
      "step": 18895
    },
    {
      "entropy": 5.503029870986938,
      "epoch": 1.7144412191582004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004708382631287055,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20356596112251282,
      "num_tokens": 39235021.0,
      "step": 18900
    },
    {
      "entropy": 5.5845558643341064,
      "epoch": 1.7148947750362846,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004708223301327837,
      "loss": 5.1454,
      "mean_token_accuracy": 0.19812427461147308,
      "num_tokens": 39245830.0,
      "step": 18905
    },
    {
      "entropy": 5.646688604354859,
      "epoch": 1.7153483309143687,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047080639308720617,
      "loss": 5.1165,
      "mean_token_accuracy": 0.2034643605351448,
      "num_tokens": 39255264.0,
      "step": 18910
    },
    {
      "entropy": 5.553671073913574,
      "epoch": 1.7158018867924527,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047079045199230254,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20753138065338134,
      "num_tokens": 39264966.0,
      "step": 18915
    },
    {
      "entropy": 5.5494787216186525,
      "epoch": 1.716255442670537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047077450684840247,
      "loss": 5.1004,
      "mean_token_accuracy": 0.1983710289001465,
      "num_tokens": 39275082.0,
      "step": 18920
    },
    {
      "entropy": 5.622508811950683,
      "epoch": 1.7167089985486212,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004707585576558357,
      "loss": 5.1766,
      "mean_token_accuracy": 0.19712687581777572,
      "num_tokens": 39285426.0,
      "step": 18925
    },
    {
      "entropy": 5.640629720687866,
      "epoch": 1.7171625544267055,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004707426044149321,
      "loss": 5.1012,
      "mean_token_accuracy": 0.20073680877685546,
      "num_tokens": 39296633.0,
      "step": 18930
    },
    {
      "entropy": 5.516549062728882,
      "epoch": 1.7176161103047896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047072664712602153,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21242664009332657,
      "num_tokens": 39306775.0,
      "step": 18935
    },
    {
      "entropy": 5.538043451309204,
      "epoch": 1.7180696661828736,
      "grad_norm": 1.0625,
      "learning_rate": 0.000470710685789434,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1989818185567856,
      "num_tokens": 39317917.0,
      "step": 18940
    },
    {
      "entropy": 5.572309398651123,
      "epoch": 1.7185232220609579,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047069472040549965,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20398209542036055,
      "num_tokens": 39328387.0,
      "step": 18945
    },
    {
      "entropy": 5.482699537277222,
      "epoch": 1.7189767779390421,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004706787509745486,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21320468336343765,
      "num_tokens": 39337639.0,
      "step": 18950
    },
    {
      "entropy": 5.552534103393555,
      "epoch": 1.7194303338171264,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047066277749691115,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20591485649347305,
      "num_tokens": 39347971.0,
      "step": 18955
    },
    {
      "entropy": 5.576763486862182,
      "epoch": 1.7198838896952104,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004706467999729175,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20291970521211625,
      "num_tokens": 39358038.0,
      "step": 18960
    },
    {
      "entropy": 5.5647767066955565,
      "epoch": 1.7203374455732945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004706308184028983,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20378267467021943,
      "num_tokens": 39367834.0,
      "step": 18965
    },
    {
      "entropy": 5.572733592987061,
      "epoch": 1.7207910014513788,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004706148327871839,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2063360780477524,
      "num_tokens": 39378924.0,
      "step": 18970
    },
    {
      "entropy": 5.635785675048828,
      "epoch": 1.721244557329463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047059884312610487,
      "loss": 5.1586,
      "mean_token_accuracy": 0.19298899471759795,
      "num_tokens": 39390082.0,
      "step": 18975
    },
    {
      "entropy": 5.568527746200561,
      "epoch": 1.7216981132075473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047058284941999185,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20253492444753646,
      "num_tokens": 39400485.0,
      "step": 18980
    },
    {
      "entropy": 5.5739670753479,
      "epoch": 1.7221516690856313,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004705668516691758,
      "loss": 5.0413,
      "mean_token_accuracy": 0.20526162534952164,
      "num_tokens": 39411601.0,
      "step": 18985
    },
    {
      "entropy": 5.5383148193359375,
      "epoch": 1.7226052249637154,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047055084987398735,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20863622277975083,
      "num_tokens": 39421625.0,
      "step": 18990
    },
    {
      "entropy": 5.5457957744598385,
      "epoch": 1.7230587808417996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004705348440347575,
      "loss": 5.0186,
      "mean_token_accuracy": 0.19834876507520677,
      "num_tokens": 39432033.0,
      "step": 18995
    },
    {
      "entropy": 5.519546318054199,
      "epoch": 1.723512336719884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047051883415181724,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20518908947706221,
      "num_tokens": 39442514.0,
      "step": 19000
    },
    {
      "entropy": 5.529253387451172,
      "epoch": 1.7239658925979682,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004705028202254977,
      "loss": 4.897,
      "mean_token_accuracy": 0.21023046970367432,
      "num_tokens": 39452497.0,
      "step": 19005
    },
    {
      "entropy": 5.531752395629883,
      "epoch": 1.7244194484760522,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047048680225612996,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20395348966121674,
      "num_tokens": 39462173.0,
      "step": 19010
    },
    {
      "entropy": 5.590306568145752,
      "epoch": 1.7248730043541363,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004704707802440454,
      "loss": 5.0729,
      "mean_token_accuracy": 0.20942844450473785,
      "num_tokens": 39473123.0,
      "step": 19015
    },
    {
      "entropy": 5.588532257080078,
      "epoch": 1.7253265602322205,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047045475418957524,
      "loss": 5.0522,
      "mean_token_accuracy": 0.19402004182338714,
      "num_tokens": 39484142.0,
      "step": 19020
    },
    {
      "entropy": 5.568538188934326,
      "epoch": 1.7257801161103048,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004704387240930509,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19874946773052216,
      "num_tokens": 39494926.0,
      "step": 19025
    },
    {
      "entropy": 5.536000108718872,
      "epoch": 1.726233671988389,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004704226899548041,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20445523411035538,
      "num_tokens": 39505153.0,
      "step": 19030
    },
    {
      "entropy": 5.509835290908813,
      "epoch": 1.726687227866473,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047040665177516616,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20506981313228606,
      "num_tokens": 39515905.0,
      "step": 19035
    },
    {
      "entropy": 5.510123252868652,
      "epoch": 1.7271407837445574,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047039060955446886,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20858711749315262,
      "num_tokens": 39526762.0,
      "step": 19040
    },
    {
      "entropy": 5.503658151626587,
      "epoch": 1.7275943396226414,
      "grad_norm": 1.078125,
      "learning_rate": 0.000470374563293044,
      "loss": 4.988,
      "mean_token_accuracy": 0.20179152339696885,
      "num_tokens": 39537340.0,
      "step": 19045
    },
    {
      "entropy": 5.5714703559875485,
      "epoch": 1.7280478955007257,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004703585129912234,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20270833969116211,
      "num_tokens": 39547144.0,
      "step": 19050
    },
    {
      "entropy": 5.653240966796875,
      "epoch": 1.72850145137881,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047034245864933894,
      "loss": 5.1215,
      "mean_token_accuracy": 0.19415885508060454,
      "num_tokens": 39557707.0,
      "step": 19055
    },
    {
      "entropy": 5.619681930541992,
      "epoch": 1.728955007256894,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047032640026772274,
      "loss": 5.1011,
      "mean_token_accuracy": 0.2001511499285698,
      "num_tokens": 39567629.0,
      "step": 19060
    },
    {
      "entropy": 5.5302807807922365,
      "epoch": 1.7294085631349783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004703103378467067,
      "loss": 5.0007,
      "mean_token_accuracy": 0.20744604766368865,
      "num_tokens": 39577918.0,
      "step": 19065
    },
    {
      "entropy": 5.571910429000854,
      "epoch": 1.7298621190130623,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047029427138662314,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21134694665670395,
      "num_tokens": 39587838.0,
      "step": 19070
    },
    {
      "entropy": 5.525185012817383,
      "epoch": 1.7303156748911466,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004702782008878042,
      "loss": 4.9286,
      "mean_token_accuracy": 0.20921002179384232,
      "num_tokens": 39597353.0,
      "step": 19075
    },
    {
      "entropy": 5.522932624816894,
      "epoch": 1.7307692307692308,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004702621263505824,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2087055891752243,
      "num_tokens": 39607349.0,
      "step": 19080
    },
    {
      "entropy": 5.562011480331421,
      "epoch": 1.731222786647315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004702460477752901,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20134317576885224,
      "num_tokens": 39617810.0,
      "step": 19085
    },
    {
      "entropy": 5.5258690357208256,
      "epoch": 1.7316763425253991,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004702299651622597,
      "loss": 4.942,
      "mean_token_accuracy": 0.21373670399188996,
      "num_tokens": 39626805.0,
      "step": 19090
    },
    {
      "entropy": 5.464047527313232,
      "epoch": 1.7321298984034832,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004702138785118239,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21195108890533448,
      "num_tokens": 39637839.0,
      "step": 19095
    },
    {
      "entropy": 5.493241214752198,
      "epoch": 1.7325834542815675,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004701977878243153,
      "loss": 4.9499,
      "mean_token_accuracy": 0.20881755650043488,
      "num_tokens": 39648697.0,
      "step": 19100
    },
    {
      "entropy": 5.490741825103759,
      "epoch": 1.7330370101596517,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004701816931000668,
      "loss": 4.991,
      "mean_token_accuracy": 0.20672685503959656,
      "num_tokens": 39658154.0,
      "step": 19105
    },
    {
      "entropy": 5.542169046401978,
      "epoch": 1.733490566037736,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047016559433941106,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2038709044456482,
      "num_tokens": 39667939.0,
      "step": 19110
    },
    {
      "entropy": 5.519157743453979,
      "epoch": 1.73394412191582,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004701494915426812,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20618734061717986,
      "num_tokens": 39678289.0,
      "step": 19115
    },
    {
      "entropy": 5.559554719924927,
      "epoch": 1.734397677793904,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047013338471021,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2046155884861946,
      "num_tokens": 39689901.0,
      "step": 19120
    },
    {
      "entropy": 5.5800446510314945,
      "epoch": 1.7348512336719883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047011727384233074,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20256204903125763,
      "num_tokens": 39700831.0,
      "step": 19125
    },
    {
      "entropy": 5.562351942062378,
      "epoch": 1.7353047895500726,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004701011589393765,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20052081495523452,
      "num_tokens": 39711794.0,
      "step": 19130
    },
    {
      "entropy": 5.521185445785522,
      "epoch": 1.7357583454281569,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047008504000168054,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2035038337111473,
      "num_tokens": 39721889.0,
      "step": 19135
    },
    {
      "entropy": 5.497358751296997,
      "epoch": 1.736211901306241,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004700689170295763,
      "loss": 5.074,
      "mean_token_accuracy": 0.20340992957353593,
      "num_tokens": 39732562.0,
      "step": 19140
    },
    {
      "entropy": 5.610089254379273,
      "epoch": 1.736665457184325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047005279002339717,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20637455582618713,
      "num_tokens": 39742286.0,
      "step": 19145
    },
    {
      "entropy": 5.5739748001098635,
      "epoch": 1.7371190130624092,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047003665898347663,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20306781083345413,
      "num_tokens": 39754336.0,
      "step": 19150
    },
    {
      "entropy": 5.496878910064697,
      "epoch": 1.7375725689404935,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004700205239101483,
      "loss": 4.936,
      "mean_token_accuracy": 0.21169695258140564,
      "num_tokens": 39763545.0,
      "step": 19155
    },
    {
      "entropy": 5.546338224411011,
      "epoch": 1.7380261248185778,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004700043848037458,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19563563019037247,
      "num_tokens": 39774536.0,
      "step": 19160
    },
    {
      "entropy": 5.5876946449279785,
      "epoch": 1.7384796806966618,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004699882416646029,
      "loss": 4.9862,
      "mean_token_accuracy": 0.20746959298849105,
      "num_tokens": 39785377.0,
      "step": 19165
    },
    {
      "entropy": 5.5691807746887205,
      "epoch": 1.7389332365747459,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004699720944930535,
      "loss": 5.106,
      "mean_token_accuracy": 0.19034550935029984,
      "num_tokens": 39796692.0,
      "step": 19170
    },
    {
      "entropy": 5.473299121856689,
      "epoch": 1.7393867924528301,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046995594328943147,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2059233844280243,
      "num_tokens": 39807927.0,
      "step": 19175
    },
    {
      "entropy": 5.669158411026001,
      "epoch": 1.7398403483309144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004699397880540709,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20216887295246125,
      "num_tokens": 39817617.0,
      "step": 19180
    },
    {
      "entropy": 5.580893421173096,
      "epoch": 1.7402939042089987,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004699236287873058,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21136983335018159,
      "num_tokens": 39827092.0,
      "step": 19185
    },
    {
      "entropy": 5.529440689086914,
      "epoch": 1.7407474600870827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004699074654894705,
      "loss": 4.9738,
      "mean_token_accuracy": 0.20527196526527405,
      "num_tokens": 39837378.0,
      "step": 19190
    },
    {
      "entropy": 5.53071403503418,
      "epoch": 1.7412010159651667,
      "grad_norm": 1.03125,
      "learning_rate": 0.000469891298160899,
      "loss": 5.0009,
      "mean_token_accuracy": 0.19492065608501435,
      "num_tokens": 39848191.0,
      "step": 19195
    },
    {
      "entropy": 5.512627172470093,
      "epoch": 1.741654571843251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046987512680192586,
      "loss": 4.9916,
      "mean_token_accuracy": 0.204746875166893,
      "num_tokens": 39857886.0,
      "step": 19200
    },
    {
      "entropy": 5.600387763977051,
      "epoch": 1.7421081277213353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046985895141288546,
      "loss": 5.0785,
      "mean_token_accuracy": 0.1978824257850647,
      "num_tokens": 39868348.0,
      "step": 19205
    },
    {
      "entropy": 5.620454978942871,
      "epoch": 1.7425616835994195,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046984277199411224,
      "loss": 5.0972,
      "mean_token_accuracy": 0.20216984152793885,
      "num_tokens": 39879129.0,
      "step": 19210
    },
    {
      "entropy": 5.51452603340149,
      "epoch": 1.7430152394775036,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046982658854594083,
      "loss": 4.984,
      "mean_token_accuracy": 0.20440545082092285,
      "num_tokens": 39890239.0,
      "step": 19215
    },
    {
      "entropy": 5.4759642601013185,
      "epoch": 1.7434687953555879,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046981040106870595,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2083138108253479,
      "num_tokens": 39900907.0,
      "step": 19220
    },
    {
      "entropy": 5.566823959350586,
      "epoch": 1.743922351233672,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004697942095627423,
      "loss": 5.0202,
      "mean_token_accuracy": 0.20522894561290742,
      "num_tokens": 39911564.0,
      "step": 19225
    },
    {
      "entropy": 5.534403371810913,
      "epoch": 1.7443759071117562,
      "grad_norm": 1.125,
      "learning_rate": 0.0004697780140283848,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2044839709997177,
      "num_tokens": 39922168.0,
      "step": 19230
    },
    {
      "entropy": 5.536777639389038,
      "epoch": 1.7448294629898404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004697618144659683,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20845214426517486,
      "num_tokens": 39932690.0,
      "step": 19235
    },
    {
      "entropy": 5.645267963409424,
      "epoch": 1.7452830188679245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004697456108758278,
      "loss": 5.149,
      "mean_token_accuracy": 0.20028664767742158,
      "num_tokens": 39942970.0,
      "step": 19240
    },
    {
      "entropy": 5.525376605987549,
      "epoch": 1.7457365747460087,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004697294032582986,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20895958989858626,
      "num_tokens": 39953255.0,
      "step": 19245
    },
    {
      "entropy": 5.507604694366455,
      "epoch": 1.7461901306240928,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046971319161371564,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21043440699577332,
      "num_tokens": 39963521.0,
      "step": 19250
    },
    {
      "entropy": 5.5434962749481205,
      "epoch": 1.746643686502177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046969697594241426,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20170967131853104,
      "num_tokens": 39973494.0,
      "step": 19255
    },
    {
      "entropy": 5.534239292144775,
      "epoch": 1.7470972423802613,
      "grad_norm": 1.0,
      "learning_rate": 0.0004696807562447298,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19806908071041107,
      "num_tokens": 39984652.0,
      "step": 19260
    },
    {
      "entropy": 5.531661176681519,
      "epoch": 1.7475507982583456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046966453252099763,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20599662363529206,
      "num_tokens": 39995490.0,
      "step": 19265
    },
    {
      "entropy": 5.502593040466309,
      "epoch": 1.7480043541364296,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046964830477155335,
      "loss": 4.8317,
      "mean_token_accuracy": 0.22880762964487075,
      "num_tokens": 40006447.0,
      "step": 19270
    },
    {
      "entropy": 5.607112073898316,
      "epoch": 1.7484579100145137,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046963207299673257,
      "loss": 5.0919,
      "mean_token_accuracy": 0.20307234823703765,
      "num_tokens": 40015733.0,
      "step": 19275
    },
    {
      "entropy": 5.598391199111939,
      "epoch": 1.748911465892598,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046961583719687085,
      "loss": 4.9486,
      "mean_token_accuracy": 0.2105601742863655,
      "num_tokens": 40025844.0,
      "step": 19280
    },
    {
      "entropy": 5.516723155975342,
      "epoch": 1.7493650217706822,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046959959737230414,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21180037707090377,
      "num_tokens": 40035942.0,
      "step": 19285
    },
    {
      "entropy": 5.472798824310303,
      "epoch": 1.7498185776487665,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046958335352336806,
      "loss": 4.8877,
      "mean_token_accuracy": 0.21567040979862212,
      "num_tokens": 40045551.0,
      "step": 19290
    },
    {
      "entropy": 5.558046770095825,
      "epoch": 1.7502721335268505,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004695671056503987,
      "loss": 5.0308,
      "mean_token_accuracy": 0.2100541278719902,
      "num_tokens": 40055803.0,
      "step": 19295
    },
    {
      "entropy": 5.535646677017212,
      "epoch": 1.7507256894049346,
      "grad_norm": 1.03125,
      "learning_rate": 0.000469550853753732,
      "loss": 5.0046,
      "mean_token_accuracy": 0.2040293887257576,
      "num_tokens": 40066004.0,
      "step": 19300
    },
    {
      "entropy": 5.547567701339721,
      "epoch": 1.7511792452830188,
      "grad_norm": 1.09375,
      "learning_rate": 0.000469534597833704,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2093586027622223,
      "num_tokens": 40075522.0,
      "step": 19305
    },
    {
      "entropy": 5.5241851806640625,
      "epoch": 1.751632801161103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000469518337890651,
      "loss": 5.0075,
      "mean_token_accuracy": 0.2129354938864708,
      "num_tokens": 40084839.0,
      "step": 19310
    },
    {
      "entropy": 5.571526432037354,
      "epoch": 1.7520863570391874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046950207392490917,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20345304161310196,
      "num_tokens": 40094392.0,
      "step": 19315
    },
    {
      "entropy": 5.500971698760987,
      "epoch": 1.7525399129172714,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046948580593681496,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20845315605401993,
      "num_tokens": 40105990.0,
      "step": 19320
    },
    {
      "entropy": 5.486760139465332,
      "epoch": 1.7529934687953554,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004694695339267047,
      "loss": 4.9008,
      "mean_token_accuracy": 0.21510082483291626,
      "num_tokens": 40116143.0,
      "step": 19325
    },
    {
      "entropy": 5.655039691925049,
      "epoch": 1.7534470246734397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046945325789491483,
      "loss": 5.089,
      "mean_token_accuracy": 0.20208169966936113,
      "num_tokens": 40126849.0,
      "step": 19330
    },
    {
      "entropy": 5.551968955993653,
      "epoch": 1.753900580551524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004694369778417822,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20378676950931549,
      "num_tokens": 40137259.0,
      "step": 19335
    },
    {
      "entropy": 5.43944149017334,
      "epoch": 1.7543541364296082,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004694206937676431,
      "loss": 4.9077,
      "mean_token_accuracy": 0.20813882648944854,
      "num_tokens": 40148578.0,
      "step": 19340
    },
    {
      "entropy": 5.605654096603393,
      "epoch": 1.7548076923076923,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004694044056728347,
      "loss": 5.0896,
      "mean_token_accuracy": 0.1949380248785019,
      "num_tokens": 40159754.0,
      "step": 19345
    },
    {
      "entropy": 5.568848085403443,
      "epoch": 1.7552612481857763,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046938811355769363,
      "loss": 5.0174,
      "mean_token_accuracy": 0.20421231389045716,
      "num_tokens": 40169388.0,
      "step": 19350
    },
    {
      "entropy": 5.539310693740845,
      "epoch": 1.7557148040638606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046937181742255673,
      "loss": 5.0168,
      "mean_token_accuracy": 0.2036700502038002,
      "num_tokens": 40179271.0,
      "step": 19355
    },
    {
      "entropy": 5.590096855163575,
      "epoch": 1.7561683599419449,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004693555172677612,
      "loss": 5.0602,
      "mean_token_accuracy": 0.20551477372646332,
      "num_tokens": 40189292.0,
      "step": 19360
    },
    {
      "entropy": 5.516025447845459,
      "epoch": 1.7566219158200291,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000469339213093644,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20286233127117156,
      "num_tokens": 40199271.0,
      "step": 19365
    },
    {
      "entropy": 5.531123161315918,
      "epoch": 1.7570754716981132,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046932290490054235,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20643097013235093,
      "num_tokens": 40210386.0,
      "step": 19370
    },
    {
      "entropy": 5.578841209411621,
      "epoch": 1.7575290275761972,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004693065926887936,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20037015676498413,
      "num_tokens": 40220892.0,
      "step": 19375
    },
    {
      "entropy": 5.576885557174682,
      "epoch": 1.7579825834542815,
      "grad_norm": 1.0,
      "learning_rate": 0.00046929027645873484,
      "loss": 4.9987,
      "mean_token_accuracy": 0.20828428864479065,
      "num_tokens": 40231737.0,
      "step": 19380
    },
    {
      "entropy": 5.5091084957122805,
      "epoch": 1.7584361393323658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004692739562107038,
      "loss": 5.042,
      "mean_token_accuracy": 0.1980927422642708,
      "num_tokens": 40242218.0,
      "step": 19385
    },
    {
      "entropy": 5.525993061065674,
      "epoch": 1.75888969521045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046925763194503775,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20683387964963912,
      "num_tokens": 40253659.0,
      "step": 19390
    },
    {
      "entropy": 5.507544612884521,
      "epoch": 1.759343251088534,
      "grad_norm": 1.0,
      "learning_rate": 0.0004692413036620744,
      "loss": 4.8609,
      "mean_token_accuracy": 0.21369356215000151,
      "num_tokens": 40264328.0,
      "step": 19395
    },
    {
      "entropy": 5.502504682540893,
      "epoch": 1.7597968069666183,
      "grad_norm": 1.0,
      "learning_rate": 0.0004692249713621514,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2095779672265053,
      "num_tokens": 40275188.0,
      "step": 19400
    },
    {
      "entropy": 5.608981037139893,
      "epoch": 1.7602503628447024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046920863504560647,
      "loss": 5.1346,
      "mean_token_accuracy": 0.1956270754337311,
      "num_tokens": 40287094.0,
      "step": 19405
    },
    {
      "entropy": 5.5839334487915036,
      "epoch": 1.7607039187227866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004691922947127775,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20392605662345886,
      "num_tokens": 40296551.0,
      "step": 19410
    },
    {
      "entropy": 5.523501968383789,
      "epoch": 1.761157474600871,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004691759503640024,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20223620533943176,
      "num_tokens": 40306920.0,
      "step": 19415
    },
    {
      "entropy": 5.511826515197754,
      "epoch": 1.761611030478955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004691596019996191,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21371311396360398,
      "num_tokens": 40316797.0,
      "step": 19420
    },
    {
      "entropy": 5.458982658386231,
      "epoch": 1.7620645863570392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004691432496199658,
      "loss": 4.951,
      "mean_token_accuracy": 0.20571546405553817,
      "num_tokens": 40327002.0,
      "step": 19425
    },
    {
      "entropy": 5.651781749725342,
      "epoch": 1.7625181422351233,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046912689322538064,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20137424170970916,
      "num_tokens": 40337247.0,
      "step": 19430
    },
    {
      "entropy": 5.558581399917602,
      "epoch": 1.7629716981132075,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046911053281620187,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20686520040035247,
      "num_tokens": 40346960.0,
      "step": 19435
    },
    {
      "entropy": 5.588541746139526,
      "epoch": 1.7634252539912918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046909416839276775,
      "loss": 5.0532,
      "mean_token_accuracy": 0.2004593625664711,
      "num_tokens": 40357169.0,
      "step": 19440
    },
    {
      "entropy": 5.501184940338135,
      "epoch": 1.763878809869376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004690777999554168,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2171902596950531,
      "num_tokens": 40367660.0,
      "step": 19445
    },
    {
      "entropy": 5.5190881252288815,
      "epoch": 1.76433236574746,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004690614275044875,
      "loss": 4.914,
      "mean_token_accuracy": 0.20759038627147675,
      "num_tokens": 40378217.0,
      "step": 19450
    },
    {
      "entropy": 5.514788818359375,
      "epoch": 1.7647859216255442,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004690450510403184,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20546943843364715,
      "num_tokens": 40388249.0,
      "step": 19455
    },
    {
      "entropy": 5.5890326499938965,
      "epoch": 1.7652394775036284,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004690286705632483,
      "loss": 5.064,
      "mean_token_accuracy": 0.1979004979133606,
      "num_tokens": 40398239.0,
      "step": 19460
    },
    {
      "entropy": 5.568636322021485,
      "epoch": 1.7656930333817127,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004690122860736157,
      "loss": 5.0626,
      "mean_token_accuracy": 0.200815749168396,
      "num_tokens": 40408473.0,
      "step": 19465
    },
    {
      "entropy": 5.640263843536377,
      "epoch": 1.766146589259797,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046899589757175964,
      "loss": 5.1298,
      "mean_token_accuracy": 0.19818018674850463,
      "num_tokens": 40418674.0,
      "step": 19470
    },
    {
      "entropy": 5.501596450805664,
      "epoch": 1.766600145137881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000468979505058019,
      "loss": 4.97,
      "mean_token_accuracy": 0.20711494535207747,
      "num_tokens": 40429815.0,
      "step": 19475
    },
    {
      "entropy": 5.587862634658814,
      "epoch": 1.767053701015965,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046896310853273274,
      "loss": 5.0866,
      "mean_token_accuracy": 0.20570432543754577,
      "num_tokens": 40439809.0,
      "step": 19480
    },
    {
      "entropy": 5.6429685115814205,
      "epoch": 1.7675072568940493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046894670799624,
      "loss": 5.0691,
      "mean_token_accuracy": 0.20091277956962586,
      "num_tokens": 40449503.0,
      "step": 19485
    },
    {
      "entropy": 5.549521589279175,
      "epoch": 1.7679608127721336,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046893030344887995,
      "loss": 5.0881,
      "mean_token_accuracy": 0.1946729063987732,
      "num_tokens": 40459794.0,
      "step": 19490
    },
    {
      "entropy": 5.5187304496765135,
      "epoch": 1.7684143686502178,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046891389489099166,
      "loss": 4.948,
      "mean_token_accuracy": 0.2143867790699005,
      "num_tokens": 40470044.0,
      "step": 19495
    },
    {
      "entropy": 5.54579644203186,
      "epoch": 1.7688679245283019,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046889748232291476,
      "loss": 5.002,
      "mean_token_accuracy": 0.2024047553539276,
      "num_tokens": 40480709.0,
      "step": 19500
    },
    {
      "entropy": 5.475869750976562,
      "epoch": 1.769321480406386,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046888106574498843,
      "loss": 4.843,
      "mean_token_accuracy": 0.21536738574504852,
      "num_tokens": 40490906.0,
      "step": 19505
    },
    {
      "entropy": 5.504793357849121,
      "epoch": 1.7697750362844702,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004688646451575522,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2155078321695328,
      "num_tokens": 40501971.0,
      "step": 19510
    },
    {
      "entropy": 5.525714445114136,
      "epoch": 1.7702285921625545,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046884822056094587,
      "loss": 5.0446,
      "mean_token_accuracy": 0.1978493884205818,
      "num_tokens": 40512685.0,
      "step": 19515
    },
    {
      "entropy": 5.506925630569458,
      "epoch": 1.7706821480406387,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046883179195550883,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20833256095647812,
      "num_tokens": 40523460.0,
      "step": 19520
    },
    {
      "entropy": 5.576283740997314,
      "epoch": 1.7711357039187228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046881535934158094,
      "loss": 4.9669,
      "mean_token_accuracy": 0.20880083590745926,
      "num_tokens": 40533594.0,
      "step": 19525
    },
    {
      "entropy": 5.504837560653686,
      "epoch": 1.7715892597968068,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004687989227195021,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20841706395149232,
      "num_tokens": 40543504.0,
      "step": 19530
    },
    {
      "entropy": 5.526325607299805,
      "epoch": 1.772042815674891,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046878248208961204,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2034996911883354,
      "num_tokens": 40553711.0,
      "step": 19535
    },
    {
      "entropy": 5.5185425758361815,
      "epoch": 1.7724963715529753,
      "grad_norm": 1.09375,
      "learning_rate": 0.000468766037452251,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20368110984563828,
      "num_tokens": 40563731.0,
      "step": 19540
    },
    {
      "entropy": 5.586743545532227,
      "epoch": 1.7729499274310596,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046874958880775886,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20407156348228456,
      "num_tokens": 40574235.0,
      "step": 19545
    },
    {
      "entropy": 5.57427830696106,
      "epoch": 1.7734034833091437,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004687331361564758,
      "loss": 4.969,
      "mean_token_accuracy": 0.20999808609485626,
      "num_tokens": 40584459.0,
      "step": 19550
    },
    {
      "entropy": 5.531437730789184,
      "epoch": 1.7738570391872277,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004687166794987422,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20473010540008546,
      "num_tokens": 40594268.0,
      "step": 19555
    },
    {
      "entropy": 5.495528221130371,
      "epoch": 1.774310595065312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004687002188348983,
      "loss": 4.9667,
      "mean_token_accuracy": 0.2113693654537201,
      "num_tokens": 40604673.0,
      "step": 19560
    },
    {
      "entropy": 5.454802703857422,
      "epoch": 1.7747641509433962,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046868375416528443,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2076304167509079,
      "num_tokens": 40614998.0,
      "step": 19565
    },
    {
      "entropy": 5.547639560699463,
      "epoch": 1.7752177068214805,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046866728549024123,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20194907188415528,
      "num_tokens": 40624729.0,
      "step": 19570
    },
    {
      "entropy": 5.563374614715576,
      "epoch": 1.7756712626995645,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004686508128101092,
      "loss": 5.0957,
      "mean_token_accuracy": 0.20164481103420256,
      "num_tokens": 40635673.0,
      "step": 19575
    },
    {
      "entropy": 5.579333925247193,
      "epoch": 1.7761248185776488,
      "grad_norm": 0.984375,
      "learning_rate": 0.000468634336125229,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20136806219816208,
      "num_tokens": 40645990.0,
      "step": 19580
    },
    {
      "entropy": 5.5818102836608885,
      "epoch": 1.7765783744557329,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046861785543594143,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21454309225082396,
      "num_tokens": 40656339.0,
      "step": 19585
    },
    {
      "entropy": 5.520548629760742,
      "epoch": 1.7770319303338171,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004686013707425873,
      "loss": 4.9276,
      "mean_token_accuracy": 0.20953194946050643,
      "num_tokens": 40666371.0,
      "step": 19590
    },
    {
      "entropy": 5.559475946426391,
      "epoch": 1.7774854862119014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046858488204550735,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20916937738657,
      "num_tokens": 40676423.0,
      "step": 19595
    },
    {
      "entropy": 5.511946535110473,
      "epoch": 1.7779390420899854,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004685683893450428,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21064889430999756,
      "num_tokens": 40686331.0,
      "step": 19600
    },
    {
      "entropy": 5.496103239059448,
      "epoch": 1.7783925979680697,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004685518926415346,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21334135830402373,
      "num_tokens": 40697408.0,
      "step": 19605
    },
    {
      "entropy": 5.439297533035278,
      "epoch": 1.7788461538461537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000468535391935324,
      "loss": 4.8678,
      "mean_token_accuracy": 0.21968182027339936,
      "num_tokens": 40707078.0,
      "step": 19610
    },
    {
      "entropy": 5.552730417251587,
      "epoch": 1.779299709724238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685188872267521,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2048732966184616,
      "num_tokens": 40718116.0,
      "step": 19615
    },
    {
      "entropy": 5.568325138092041,
      "epoch": 1.7797532656023223,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685023785161603,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20626704692840575,
      "num_tokens": 40727737.0,
      "step": 19620
    },
    {
      "entropy": 5.5643861293792725,
      "epoch": 1.7802068214804065,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004684858658038901,
      "loss": 4.9754,
      "mean_token_accuracy": 0.20463858544826508,
      "num_tokens": 40737455.0,
      "step": 19625
    },
    {
      "entropy": 5.59386043548584,
      "epoch": 1.7806603773584906,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004684693490902828,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2064111277461052,
      "num_tokens": 40746715.0,
      "step": 19630
    },
    {
      "entropy": 5.49885196685791,
      "epoch": 1.7811139332365746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046845282837568,
      "loss": 4.8931,
      "mean_token_accuracy": 0.21309821605682372,
      "num_tokens": 40756432.0,
      "step": 19635
    },
    {
      "entropy": 5.446016025543213,
      "epoch": 1.781567489114659,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046843630366042355,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20433081090450286,
      "num_tokens": 40766629.0,
      "step": 19640
    },
    {
      "entropy": 5.593451738357544,
      "epoch": 1.7820210449927432,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046841977494485506,
      "loss": 5.0983,
      "mean_token_accuracy": 0.19296467453241348,
      "num_tokens": 40776650.0,
      "step": 19645
    },
    {
      "entropy": 5.60643949508667,
      "epoch": 1.7824746008708274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004684032422293162,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21133604347705842,
      "num_tokens": 40786722.0,
      "step": 19650
    },
    {
      "entropy": 5.441478490829468,
      "epoch": 1.7829281567489115,
      "grad_norm": 1.015625,
      "learning_rate": 0.000468386705514149,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22297567427158355,
      "num_tokens": 40797058.0,
      "step": 19655
    },
    {
      "entropy": 5.481105852127075,
      "epoch": 1.7833817126269955,
      "grad_norm": 1.125,
      "learning_rate": 0.0004683701647996956,
      "loss": 5.041,
      "mean_token_accuracy": 0.20970151722431182,
      "num_tokens": 40808207.0,
      "step": 19660
    },
    {
      "entropy": 5.586927700042724,
      "epoch": 1.7838352685050798,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004683536200862978,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20260453671216966,
      "num_tokens": 40819655.0,
      "step": 19665
    },
    {
      "entropy": 5.59092755317688,
      "epoch": 1.784288824383164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004683370713742979,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20382409691810607,
      "num_tokens": 40831015.0,
      "step": 19670
    },
    {
      "entropy": 5.512878561019898,
      "epoch": 1.7847423802612483,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046832051866403814,
      "loss": 4.955,
      "mean_token_accuracy": 0.20847117751836777,
      "num_tokens": 40841763.0,
      "step": 19675
    },
    {
      "entropy": 5.506123781204224,
      "epoch": 1.7851959361393324,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004683039619558607,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2064852684736252,
      "num_tokens": 40852479.0,
      "step": 19680
    },
    {
      "entropy": 5.579826307296753,
      "epoch": 1.7856494920174164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046828740125010813,
      "loss": 5.016,
      "mean_token_accuracy": 0.20882578641176225,
      "num_tokens": 40862486.0,
      "step": 19685
    },
    {
      "entropy": 5.611700773239136,
      "epoch": 1.7861030478955007,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004682708365471228,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20229277312755584,
      "num_tokens": 40872182.0,
      "step": 19690
    },
    {
      "entropy": 5.558951568603516,
      "epoch": 1.786556603773585,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046825426784724734,
      "loss": 5.0676,
      "mean_token_accuracy": 0.20649306178092958,
      "num_tokens": 40883036.0,
      "step": 19695
    },
    {
      "entropy": 5.594229221343994,
      "epoch": 1.7870101596516692,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004682376951508244,
      "loss": 5.1016,
      "mean_token_accuracy": 0.1905641257762909,
      "num_tokens": 40893684.0,
      "step": 19700
    },
    {
      "entropy": 5.629052925109863,
      "epoch": 1.7874637155297532,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046822111845819665,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20329589396715164,
      "num_tokens": 40904540.0,
      "step": 19705
    },
    {
      "entropy": 5.60902042388916,
      "epoch": 1.7879172714078373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046820453776970697,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20841806381940842,
      "num_tokens": 40915269.0,
      "step": 19710
    },
    {
      "entropy": 5.5748247623443605,
      "epoch": 1.7883708272859216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046818795308569815,
      "loss": 5.0823,
      "mean_token_accuracy": 0.19952248632907868,
      "num_tokens": 40925987.0,
      "step": 19715
    },
    {
      "entropy": 5.497185230255127,
      "epoch": 1.7888243831640058,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004681713644065133,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21137189120054245,
      "num_tokens": 40936130.0,
      "step": 19720
    },
    {
      "entropy": 5.468851041793823,
      "epoch": 1.78927793904209,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004681547717324954,
      "loss": 4.8934,
      "mean_token_accuracy": 0.21224325299263,
      "num_tokens": 40945476.0,
      "step": 19725
    },
    {
      "entropy": 5.450008964538574,
      "epoch": 1.7897314949201741,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046813817506398755,
      "loss": 4.8696,
      "mean_token_accuracy": 0.217603899538517,
      "num_tokens": 40956492.0,
      "step": 19730
    },
    {
      "entropy": 5.5896382331848145,
      "epoch": 1.7901850507982582,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046812157440133303,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20275839120149614,
      "num_tokens": 40966990.0,
      "step": 19735
    },
    {
      "entropy": 5.575714921951294,
      "epoch": 1.7906386066763424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004681049697448752,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21206290721893312,
      "num_tokens": 40977065.0,
      "step": 19740
    },
    {
      "entropy": 5.519251680374145,
      "epoch": 1.7910921625544267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046808836109495734,
      "loss": 5.0133,
      "mean_token_accuracy": 0.20591092705726624,
      "num_tokens": 40987207.0,
      "step": 19745
    },
    {
      "entropy": 5.5362590789794925,
      "epoch": 1.791545718432511,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004680717484519229,
      "loss": 4.9674,
      "mean_token_accuracy": 0.20424300730228423,
      "num_tokens": 40997541.0,
      "step": 19750
    },
    {
      "entropy": 5.5382311820983885,
      "epoch": 1.791999274310595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004680551318161156,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20712293982505797,
      "num_tokens": 41009044.0,
      "step": 19755
    },
    {
      "entropy": 5.559739017486573,
      "epoch": 1.7924528301886793,
      "grad_norm": 1.0,
      "learning_rate": 0.00046803851118787897,
      "loss": 5.008,
      "mean_token_accuracy": 0.20548251569271087,
      "num_tokens": 41018165.0,
      "step": 19760
    },
    {
      "entropy": 5.497653388977051,
      "epoch": 1.7929063860667633,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004680218865675567,
      "loss": 4.9652,
      "mean_token_accuracy": 0.20819184333086013,
      "num_tokens": 41028972.0,
      "step": 19765
    },
    {
      "entropy": 5.507817316055298,
      "epoch": 1.7933599419448476,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004680052579554926,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20903472155332564,
      "num_tokens": 41039552.0,
      "step": 19770
    },
    {
      "entropy": 5.620876598358154,
      "epoch": 1.7938134978229319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004679886253520307,
      "loss": 5.0528,
      "mean_token_accuracy": 0.1981100857257843,
      "num_tokens": 41049808.0,
      "step": 19775
    },
    {
      "entropy": 5.539993143081665,
      "epoch": 1.794267053701016,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046797198875751474,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2042449876666069,
      "num_tokens": 41060574.0,
      "step": 19780
    },
    {
      "entropy": 5.614523315429688,
      "epoch": 1.7947206095791002,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004679553481722889,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20428696572780608,
      "num_tokens": 41071239.0,
      "step": 19785
    },
    {
      "entropy": 5.615858888626098,
      "epoch": 1.7951741654571842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046793870359669733,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20591361224651336,
      "num_tokens": 41080819.0,
      "step": 19790
    },
    {
      "entropy": 5.587080287933349,
      "epoch": 1.7956277213352685,
      "grad_norm": 1.25,
      "learning_rate": 0.0004679220550310842,
      "loss": 4.933,
      "mean_token_accuracy": 0.2169566422700882,
      "num_tokens": 41090457.0,
      "step": 19795
    },
    {
      "entropy": 5.530445909500122,
      "epoch": 1.7960812772133528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004679054024757937,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21040300279855728,
      "num_tokens": 41101210.0,
      "step": 19800
    },
    {
      "entropy": 5.537619400024414,
      "epoch": 1.796534833091437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004678887459311704,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20775469839572908,
      "num_tokens": 41110566.0,
      "step": 19805
    },
    {
      "entropy": 5.47896146774292,
      "epoch": 1.796988388969521,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004678720853975587,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2113064005970955,
      "num_tokens": 41120000.0,
      "step": 19810
    },
    {
      "entropy": 5.469798278808594,
      "epoch": 1.797441944847605,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046785542087530307,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2107544869184494,
      "num_tokens": 41129944.0,
      "step": 19815
    },
    {
      "entropy": 5.485931491851806,
      "epoch": 1.7978955007256894,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004678387523647483,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2056867852807045,
      "num_tokens": 41138803.0,
      "step": 19820
    },
    {
      "entropy": 5.560794973373413,
      "epoch": 1.7983490566037736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046782207986623885,
      "loss": 5.0807,
      "mean_token_accuracy": 0.19582369774580002,
      "num_tokens": 41150365.0,
      "step": 19825
    },
    {
      "entropy": 5.664840650558472,
      "epoch": 1.798802612481858,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046780540338011974,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20080285668373107,
      "num_tokens": 41160310.0,
      "step": 19830
    },
    {
      "entropy": 5.512536096572876,
      "epoch": 1.799256168359942,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004677887229067358,
      "loss": 4.9104,
      "mean_token_accuracy": 0.20896457135677338,
      "num_tokens": 41170506.0,
      "step": 19835
    },
    {
      "entropy": 5.555658769607544,
      "epoch": 1.799709724238026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046777203844643184,
      "loss": 5.0348,
      "mean_token_accuracy": 0.1990685984492302,
      "num_tokens": 41181293.0,
      "step": 19840
    },
    {
      "entropy": 5.489247608184814,
      "epoch": 1.8001632801161103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046775534999955303,
      "loss": 4.9688,
      "mean_token_accuracy": 0.20351264625787735,
      "num_tokens": 41191637.0,
      "step": 19845
    },
    {
      "entropy": 5.479731512069702,
      "epoch": 1.8006168359941945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004677386575664445,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2177043989300728,
      "num_tokens": 41201833.0,
      "step": 19850
    },
    {
      "entropy": 5.57675142288208,
      "epoch": 1.8010703918722788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046772196114745133,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2109220579266548,
      "num_tokens": 41211243.0,
      "step": 19855
    },
    {
      "entropy": 5.58609938621521,
      "epoch": 1.8015239477503628,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004677052607429189,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21167860627174379,
      "num_tokens": 41221111.0,
      "step": 19860
    },
    {
      "entropy": 5.501392126083374,
      "epoch": 1.8019775036284469,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004676885563531927,
      "loss": 5.0146,
      "mean_token_accuracy": 0.2025648310780525,
      "num_tokens": 41231578.0,
      "step": 19865
    },
    {
      "entropy": 5.509230232238769,
      "epoch": 1.8024310595065312,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046767184797861795,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20938555002212525,
      "num_tokens": 41242130.0,
      "step": 19870
    },
    {
      "entropy": 5.59357099533081,
      "epoch": 1.8028846153846154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046765513561954026,
      "loss": 5.0226,
      "mean_token_accuracy": 0.19962058812379838,
      "num_tokens": 41252885.0,
      "step": 19875
    },
    {
      "entropy": 5.595680999755859,
      "epoch": 1.8033381712626997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046763841927630534,
      "loss": 5.0556,
      "mean_token_accuracy": 0.20472382009029388,
      "num_tokens": 41263532.0,
      "step": 19880
    },
    {
      "entropy": 5.59276933670044,
      "epoch": 1.8037917271407837,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046762169894925876,
      "loss": 5.0079,
      "mean_token_accuracy": 0.20627274066209794,
      "num_tokens": 41273469.0,
      "step": 19885
    },
    {
      "entropy": 5.55303692817688,
      "epoch": 1.8042452830188678,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004676049746387464,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20153881907463073,
      "num_tokens": 41283663.0,
      "step": 19890
    },
    {
      "entropy": 5.510688352584839,
      "epoch": 1.804698838896952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004675882463451141,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21141158491373063,
      "num_tokens": 41294657.0,
      "step": 19895
    },
    {
      "entropy": 5.568413734436035,
      "epoch": 1.8051523947750363,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004675715140687078,
      "loss": 5.0009,
      "mean_token_accuracy": 0.20919404029846192,
      "num_tokens": 41305124.0,
      "step": 19900
    },
    {
      "entropy": 5.4937211036682125,
      "epoch": 1.8056059506531206,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046755477780987344,
      "loss": 4.9496,
      "mean_token_accuracy": 0.20851007848978043,
      "num_tokens": 41315839.0,
      "step": 19905
    },
    {
      "entropy": 5.574371242523194,
      "epoch": 1.8060595065312046,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046753803756895726,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20170085430145263,
      "num_tokens": 41326388.0,
      "step": 19910
    },
    {
      "entropy": 5.527889442443848,
      "epoch": 1.8065130624092887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004675212933463054,
      "loss": 5.0737,
      "mean_token_accuracy": 0.1926412522792816,
      "num_tokens": 41337333.0,
      "step": 19915
    },
    {
      "entropy": 5.618911981582642,
      "epoch": 1.806966618287373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004675045451422641,
      "loss": 5.0657,
      "mean_token_accuracy": 0.19672111868858339,
      "num_tokens": 41348353.0,
      "step": 19920
    },
    {
      "entropy": 5.586848640441895,
      "epoch": 1.8074201741654572,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046748779295717973,
      "loss": 5.0304,
      "mean_token_accuracy": 0.213688226044178,
      "num_tokens": 41358797.0,
      "step": 19925
    },
    {
      "entropy": 5.5675770282745365,
      "epoch": 1.8078737300435415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004674710367913989,
      "loss": 5.0237,
      "mean_token_accuracy": 0.2066128134727478,
      "num_tokens": 41368818.0,
      "step": 19930
    },
    {
      "entropy": 5.5882336616516115,
      "epoch": 1.8083272859216255,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004674542766452678,
      "loss": 5.0116,
      "mean_token_accuracy": 0.2071378633379936,
      "num_tokens": 41378949.0,
      "step": 19935
    },
    {
      "entropy": 5.566381454467773,
      "epoch": 1.8087808417997098,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046743751251913327,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20560411214828492,
      "num_tokens": 41388892.0,
      "step": 19940
    },
    {
      "entropy": 5.5644384860992435,
      "epoch": 1.8092343976777938,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046742074441334197,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20125612020492553,
      "num_tokens": 41399197.0,
      "step": 19945
    },
    {
      "entropy": 5.52111587524414,
      "epoch": 1.809687953555878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004674039723282406,
      "loss": 5.0377,
      "mean_token_accuracy": 0.1966809868812561,
      "num_tokens": 41409551.0,
      "step": 19950
    },
    {
      "entropy": 5.518880939483642,
      "epoch": 1.8101415094339623,
      "grad_norm": 1.125,
      "learning_rate": 0.00046738719626417607,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20883867889642715,
      "num_tokens": 41418900.0,
      "step": 19955
    },
    {
      "entropy": 5.52384581565857,
      "epoch": 1.8105950653120464,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046737041622149536,
      "loss": 4.949,
      "mean_token_accuracy": 0.20624009966850282,
      "num_tokens": 41428204.0,
      "step": 19960
    },
    {
      "entropy": 5.538127660751343,
      "epoch": 1.8110486211901307,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004673536322005454,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20278254002332688,
      "num_tokens": 41437922.0,
      "step": 19965
    },
    {
      "entropy": 5.561640787124634,
      "epoch": 1.8115021770682147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004673368442016732,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20119996517896652,
      "num_tokens": 41447411.0,
      "step": 19970
    },
    {
      "entropy": 5.607109880447387,
      "epoch": 1.811955732946299,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004673200522252261,
      "loss": 5.093,
      "mean_token_accuracy": 0.20437152981758117,
      "num_tokens": 41458756.0,
      "step": 19975
    },
    {
      "entropy": 5.579017496109008,
      "epoch": 1.8124092888243832,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004673032562715514,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2080783352255821,
      "num_tokens": 41468610.0,
      "step": 19980
    },
    {
      "entropy": 5.555026006698609,
      "epoch": 1.8128628447024675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046728645634099627,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19671774953603743,
      "num_tokens": 41479849.0,
      "step": 19985
    },
    {
      "entropy": 5.526775121688843,
      "epoch": 1.8133164005805515,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046726965243390825,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2027814969420433,
      "num_tokens": 41490945.0,
      "step": 19990
    },
    {
      "entropy": 5.486761856079101,
      "epoch": 1.8137699564586356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004672528445506348,
      "loss": 4.9515,
      "mean_token_accuracy": 0.20101569592952728,
      "num_tokens": 41501479.0,
      "step": 19995
    },
    {
      "entropy": 5.525984382629394,
      "epoch": 1.8142235123367199,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004672360326915236,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21505804359912872,
      "num_tokens": 41512447.0,
      "step": 20000
    },
    {
      "entropy": 5.509391450881958,
      "epoch": 1.8146770682148041,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004672192168569222,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21713926196098327,
      "num_tokens": 41521905.0,
      "step": 20005
    },
    {
      "entropy": 5.505262660980224,
      "epoch": 1.8151306240928884,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046720239704717845,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20470809191465378,
      "num_tokens": 41533143.0,
      "step": 20010
    },
    {
      "entropy": 5.530465078353882,
      "epoch": 1.8155841799709724,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046718557326264,
      "loss": 4.9591,
      "mean_token_accuracy": 0.20491113215684892,
      "num_tokens": 41542234.0,
      "step": 20015
    },
    {
      "entropy": 5.579422378540039,
      "epoch": 1.8160377358490565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004671687455036551,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2017393171787262,
      "num_tokens": 41552632.0,
      "step": 20020
    },
    {
      "entropy": 5.54284815788269,
      "epoch": 1.8164912917271407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004671519137705715,
      "loss": 5.016,
      "mean_token_accuracy": 0.21005022525787354,
      "num_tokens": 41563723.0,
      "step": 20025
    },
    {
      "entropy": 5.551046705245971,
      "epoch": 1.816944847605225,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046713507806373736,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21051034778356553,
      "num_tokens": 41573680.0,
      "step": 20030
    },
    {
      "entropy": 5.553695774078369,
      "epoch": 1.8173984034833093,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046711823838350086,
      "loss": 5.0444,
      "mean_token_accuracy": 0.20358869880437852,
      "num_tokens": 41584722.0,
      "step": 20035
    },
    {
      "entropy": 5.532588720321655,
      "epoch": 1.8178519593613933,
      "grad_norm": 1.0,
      "learning_rate": 0.00046710139473021025,
      "loss": 5.0871,
      "mean_token_accuracy": 0.19790272265672684,
      "num_tokens": 41595785.0,
      "step": 20040
    },
    {
      "entropy": 5.569847583770752,
      "epoch": 1.8183055152394774,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046708454710421376,
      "loss": 5.107,
      "mean_token_accuracy": 0.1956186145544052,
      "num_tokens": 41606023.0,
      "step": 20045
    },
    {
      "entropy": 5.566348218917847,
      "epoch": 1.8187590711175616,
      "grad_norm": 1.125,
      "learning_rate": 0.00046706769550586,
      "loss": 5.0154,
      "mean_token_accuracy": 0.20516480803489684,
      "num_tokens": 41616010.0,
      "step": 20050
    },
    {
      "entropy": 5.557003307342529,
      "epoch": 1.819212626995646,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046705083993549723,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20764904767274855,
      "num_tokens": 41625105.0,
      "step": 20055
    },
    {
      "entropy": 5.5250067710876465,
      "epoch": 1.8196661828737302,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004670339803934743,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2096752032637596,
      "num_tokens": 41634989.0,
      "step": 20060
    },
    {
      "entropy": 5.5185634136199955,
      "epoch": 1.8201197387518142,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004670171168801396,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2066364035010338,
      "num_tokens": 41645736.0,
      "step": 20065
    },
    {
      "entropy": 5.605870532989502,
      "epoch": 1.8205732946298983,
      "grad_norm": 1.0625,
      "learning_rate": 0.000467000249395842,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20920195430517197,
      "num_tokens": 41655505.0,
      "step": 20070
    },
    {
      "entropy": 5.528751707077026,
      "epoch": 1.8210268505079825,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004669833779409303,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20744871944189072,
      "num_tokens": 41665732.0,
      "step": 20075
    },
    {
      "entropy": 5.504740619659424,
      "epoch": 1.8214804063860668,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004669665025157535,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2010861560702324,
      "num_tokens": 41676050.0,
      "step": 20080
    },
    {
      "entropy": 5.614622354507446,
      "epoch": 1.821933962264151,
      "grad_norm": 1.125,
      "learning_rate": 0.00046694962312066046,
      "loss": 5.1475,
      "mean_token_accuracy": 0.19905100166797637,
      "num_tokens": 41686787.0,
      "step": 20085
    },
    {
      "entropy": 5.545039892196655,
      "epoch": 1.822387518142235,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046693273975600036,
      "loss": 4.944,
      "mean_token_accuracy": 0.20888380259275435,
      "num_tokens": 41697168.0,
      "step": 20090
    },
    {
      "entropy": 5.503949499130249,
      "epoch": 1.8228410740203191,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004669158524221222,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20895278602838516,
      "num_tokens": 41707524.0,
      "step": 20095
    },
    {
      "entropy": 5.587649965286255,
      "epoch": 1.8232946298984034,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046689896111937537,
      "loss": 5.0626,
      "mean_token_accuracy": 0.2100888043642044,
      "num_tokens": 41718109.0,
      "step": 20100
    },
    {
      "entropy": 5.64175763130188,
      "epoch": 1.8237481857764877,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004668820658481091,
      "loss": 5.0354,
      "mean_token_accuracy": 0.20424672663211824,
      "num_tokens": 41728862.0,
      "step": 20105
    },
    {
      "entropy": 5.652068376541138,
      "epoch": 1.824201741654572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046686516660867283,
      "loss": 5.1023,
      "mean_token_accuracy": 0.1998430848121643,
      "num_tokens": 41739024.0,
      "step": 20110
    },
    {
      "entropy": 5.565501117706299,
      "epoch": 1.824655297532656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004668482634014159,
      "loss": 5.0277,
      "mean_token_accuracy": 0.20229787528514862,
      "num_tokens": 41748979.0,
      "step": 20115
    },
    {
      "entropy": 5.510527801513672,
      "epoch": 1.8251088534107403,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004668313562266882,
      "loss": 5.036,
      "mean_token_accuracy": 0.208828841149807,
      "num_tokens": 41759972.0,
      "step": 20120
    },
    {
      "entropy": 5.5156959056854244,
      "epoch": 1.8255624092888243,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000466814445084839,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21409412175416948,
      "num_tokens": 41770150.0,
      "step": 20125
    },
    {
      "entropy": 5.549961090087891,
      "epoch": 1.8260159651669086,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004667975299762183,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21211618036031724,
      "num_tokens": 41780405.0,
      "step": 20130
    },
    {
      "entropy": 5.5162074089050295,
      "epoch": 1.8264695210449928,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046678061090117576,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21304504275321962,
      "num_tokens": 41791269.0,
      "step": 20135
    },
    {
      "entropy": 5.442075777053833,
      "epoch": 1.8269230769230769,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004667636878600613,
      "loss": 4.917,
      "mean_token_accuracy": 0.20943055897951127,
      "num_tokens": 41802513.0,
      "step": 20140
    },
    {
      "entropy": 5.578953790664673,
      "epoch": 1.8273766328011611,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004667467608532249,
      "loss": 5.0589,
      "mean_token_accuracy": 0.20094809979200362,
      "num_tokens": 41814149.0,
      "step": 20145
    },
    {
      "entropy": 5.572923946380615,
      "epoch": 1.8278301886792452,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004667298298810167,
      "loss": 5.035,
      "mean_token_accuracy": 0.20357026308774948,
      "num_tokens": 41825763.0,
      "step": 20150
    },
    {
      "entropy": 5.522440624237061,
      "epoch": 1.8282837445573294,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004667128949437867,
      "loss": 5.0115,
      "mean_token_accuracy": 0.20984289944171905,
      "num_tokens": 41834543.0,
      "step": 20155
    },
    {
      "entropy": 5.519170665740967,
      "epoch": 1.8287373004354137,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004666959560418852,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21681906580924987,
      "num_tokens": 41844940.0,
      "step": 20160
    },
    {
      "entropy": 5.498330211639404,
      "epoch": 1.829190856313498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046667901317566245,
      "loss": 4.9249,
      "mean_token_accuracy": 0.20911678224802016,
      "num_tokens": 41855532.0,
      "step": 20165
    },
    {
      "entropy": 5.592134714126587,
      "epoch": 1.829644412191582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004666620663454689,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20335020124912262,
      "num_tokens": 41865661.0,
      "step": 20170
    },
    {
      "entropy": 5.5471094131469725,
      "epoch": 1.830097968069666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046664511555165496,
      "loss": 5.0347,
      "mean_token_accuracy": 0.2055257648229599,
      "num_tokens": 41875662.0,
      "step": 20175
    },
    {
      "entropy": 5.522885942459107,
      "epoch": 1.8305515239477503,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046662816079457114,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21466411501169205,
      "num_tokens": 41886211.0,
      "step": 20180
    },
    {
      "entropy": 5.623113489151001,
      "epoch": 1.8310050798258346,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004666112020745682,
      "loss": 5.1163,
      "mean_token_accuracy": 0.19894250482320786,
      "num_tokens": 41896737.0,
      "step": 20185
    },
    {
      "entropy": 5.552207946777344,
      "epoch": 1.8314586357039189,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004665942393919967,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20558253079652786,
      "num_tokens": 41907768.0,
      "step": 20190
    },
    {
      "entropy": 5.5546571731567385,
      "epoch": 1.831912191582003,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046657727274720755,
      "loss": 4.9649,
      "mean_token_accuracy": 0.20418288111686705,
      "num_tokens": 41917587.0,
      "step": 20195
    },
    {
      "entropy": 5.475027942657471,
      "epoch": 1.832365747460087,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004665603021405516,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21469545364379883,
      "num_tokens": 41927842.0,
      "step": 20200
    },
    {
      "entropy": 5.485632419586182,
      "epoch": 1.8328193033381712,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004665433275723798,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21441649943590163,
      "num_tokens": 41939899.0,
      "step": 20205
    },
    {
      "entropy": 5.527506351470947,
      "epoch": 1.8332728592162555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004665263490430431,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20336755961179734,
      "num_tokens": 41950932.0,
      "step": 20210
    },
    {
      "entropy": 5.5579382419586185,
      "epoch": 1.8337264150943398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004665093665528928,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20819393843412398,
      "num_tokens": 41963512.0,
      "step": 20215
    },
    {
      "entropy": 5.636223602294922,
      "epoch": 1.8341799709724238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046649238010227983,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21151686012744902,
      "num_tokens": 41974677.0,
      "step": 20220
    },
    {
      "entropy": 5.485407495498658,
      "epoch": 1.8346335268505078,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004664753896915558,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21072100549936296,
      "num_tokens": 41984582.0,
      "step": 20225
    },
    {
      "entropy": 5.543138790130615,
      "epoch": 1.8350870827285921,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004664583953210719,
      "loss": 4.9368,
      "mean_token_accuracy": 0.20556853264570235,
      "num_tokens": 41994624.0,
      "step": 20230
    },
    {
      "entropy": 5.525238418579102,
      "epoch": 1.8355406386066764,
      "grad_norm": 1.125,
      "learning_rate": 0.0004664413969911795,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21315260231494904,
      "num_tokens": 42004835.0,
      "step": 20235
    },
    {
      "entropy": 5.474159622192383,
      "epoch": 1.8359941944847606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004664243947022303,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20505469739437104,
      "num_tokens": 42014727.0,
      "step": 20240
    },
    {
      "entropy": 5.56663556098938,
      "epoch": 1.8364477503628447,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046640738845457586,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20703188925981522,
      "num_tokens": 42023628.0,
      "step": 20245
    },
    {
      "entropy": 5.59652886390686,
      "epoch": 1.8369013062409287,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004663903782485678,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20446922183036803,
      "num_tokens": 42033587.0,
      "step": 20250
    },
    {
      "entropy": 5.5812348365783695,
      "epoch": 1.837354862119013,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046637336408455793,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20150297582149507,
      "num_tokens": 42043765.0,
      "step": 20255
    },
    {
      "entropy": 5.608831453323364,
      "epoch": 1.8378084179970973,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046635634596289816,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20444309413433076,
      "num_tokens": 42054205.0,
      "step": 20260
    },
    {
      "entropy": 5.487960290908814,
      "epoch": 1.8382619738751815,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004663393238839404,
      "loss": 4.9104,
      "mean_token_accuracy": 0.20541877448558807,
      "num_tokens": 42064512.0,
      "step": 20265
    },
    {
      "entropy": 5.507561206817627,
      "epoch": 1.8387155297532656,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004663222978480366,
      "loss": 4.9937,
      "mean_token_accuracy": 0.205137300491333,
      "num_tokens": 42074954.0,
      "step": 20270
    },
    {
      "entropy": 5.568202161788941,
      "epoch": 1.8391690856313496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000466305267855539,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20168593227863313,
      "num_tokens": 42084914.0,
      "step": 20275
    },
    {
      "entropy": 5.576024961471558,
      "epoch": 1.8396226415094339,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004662882339067996,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20538800060749055,
      "num_tokens": 42095454.0,
      "step": 20280
    },
    {
      "entropy": 5.587305212020874,
      "epoch": 1.8400761973875182,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004662711960021709,
      "loss": 5.0882,
      "mean_token_accuracy": 0.1997831478714943,
      "num_tokens": 42106960.0,
      "step": 20285
    },
    {
      "entropy": 5.525085353851319,
      "epoch": 1.8405297532656024,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046625415414200513,
      "loss": 4.9626,
      "mean_token_accuracy": 0.20938224345445633,
      "num_tokens": 42117009.0,
      "step": 20290
    },
    {
      "entropy": 5.577379465103149,
      "epoch": 1.8409833091436865,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004662371083266546,
      "loss": 5.002,
      "mean_token_accuracy": 0.2110588639974594,
      "num_tokens": 42126371.0,
      "step": 20295
    },
    {
      "entropy": 5.58545708656311,
      "epoch": 1.8414368650217707,
      "grad_norm": 1.125,
      "learning_rate": 0.000466220058556472,
      "loss": 5.031,
      "mean_token_accuracy": 0.20728013813495635,
      "num_tokens": 42135890.0,
      "step": 20300
    },
    {
      "entropy": 5.6177905082702635,
      "epoch": 1.8418904208998548,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046620300483180986,
      "loss": 5.1298,
      "mean_token_accuracy": 0.19759823381900787,
      "num_tokens": 42145696.0,
      "step": 20305
    },
    {
      "entropy": 5.548390865325928,
      "epoch": 1.842343976777939,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004661859471530209,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21023260056972504,
      "num_tokens": 42156236.0,
      "step": 20310
    },
    {
      "entropy": 5.597132444381714,
      "epoch": 1.8427975326560233,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004661688855204578,
      "loss": 5.0606,
      "mean_token_accuracy": 0.20116368681192398,
      "num_tokens": 42166179.0,
      "step": 20315
    },
    {
      "entropy": 5.510843658447266,
      "epoch": 1.8432510885341074,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046615181993447344,
      "loss": 4.9412,
      "mean_token_accuracy": 0.20522765815258026,
      "num_tokens": 42176643.0,
      "step": 20320
    },
    {
      "entropy": 5.596554136276245,
      "epoch": 1.8437046444121916,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004661347503954207,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19032192081212998,
      "num_tokens": 42188328.0,
      "step": 20325
    },
    {
      "entropy": 5.601656866073609,
      "epoch": 1.8441582002902757,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046611767690365265,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21157871037721634,
      "num_tokens": 42198110.0,
      "step": 20330
    },
    {
      "entropy": 5.527664279937744,
      "epoch": 1.84461175616836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004661005994595223,
      "loss": 4.9823,
      "mean_token_accuracy": 0.19972015023231507,
      "num_tokens": 42208438.0,
      "step": 20335
    },
    {
      "entropy": 5.560026597976685,
      "epoch": 1.8450653120464442,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004660835180633829,
      "loss": 4.9784,
      "mean_token_accuracy": 0.2038367673754692,
      "num_tokens": 42220052.0,
      "step": 20340
    },
    {
      "entropy": 5.538200902938843,
      "epoch": 1.8455188679245285,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004660664327155876,
      "loss": 4.9749,
      "mean_token_accuracy": 0.20943403989076614,
      "num_tokens": 42229497.0,
      "step": 20345
    },
    {
      "entropy": 5.496162986755371,
      "epoch": 1.8459724238026125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004660493434164898,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20182014554738997,
      "num_tokens": 42240725.0,
      "step": 20350
    },
    {
      "entropy": 5.565572261810303,
      "epoch": 1.8464259796806965,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004660322501664429,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20267361104488374,
      "num_tokens": 42251114.0,
      "step": 20355
    },
    {
      "entropy": 5.619518995285034,
      "epoch": 1.8468795355587808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004660151529658004,
      "loss": 5.1897,
      "mean_token_accuracy": 0.19703874588012696,
      "num_tokens": 42262825.0,
      "step": 20360
    },
    {
      "entropy": 5.617018222808838,
      "epoch": 1.847333091436865,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004659980518149158,
      "loss": 5.0151,
      "mean_token_accuracy": 0.21149738729000092,
      "num_tokens": 42272176.0,
      "step": 20365
    },
    {
      "entropy": 5.602189350128174,
      "epoch": 1.8477866473149493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046598094671414285,
      "loss": 5.066,
      "mean_token_accuracy": 0.2062876984477043,
      "num_tokens": 42281934.0,
      "step": 20370
    },
    {
      "entropy": 5.540198755264282,
      "epoch": 1.8482402031930334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046596383766383533,
      "loss": 4.9589,
      "mean_token_accuracy": 0.20870660990476608,
      "num_tokens": 42291927.0,
      "step": 20375
    },
    {
      "entropy": 5.5219872951507565,
      "epoch": 1.8486937590711174,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004659467246643469,
      "loss": 4.9532,
      "mean_token_accuracy": 0.20699819028377534,
      "num_tokens": 42301407.0,
      "step": 20380
    },
    {
      "entropy": 5.455481624603271,
      "epoch": 1.8491473149492017,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046592960771603154,
      "loss": 4.9019,
      "mean_token_accuracy": 0.21629468202590943,
      "num_tokens": 42311199.0,
      "step": 20385
    },
    {
      "entropy": 5.533198833465576,
      "epoch": 1.849600870827286,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004659124868192433,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20851015746593476,
      "num_tokens": 42320725.0,
      "step": 20390
    },
    {
      "entropy": 5.621488285064697,
      "epoch": 1.8500544267053702,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004658953619743361,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20408866852521895,
      "num_tokens": 42330857.0,
      "step": 20395
    },
    {
      "entropy": 5.526382637023926,
      "epoch": 1.8505079825834543,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046587823318166427,
      "loss": 4.9196,
      "mean_token_accuracy": 0.20995397865772247,
      "num_tokens": 42340241.0,
      "step": 20400
    },
    {
      "entropy": 5.619599676132202,
      "epoch": 1.8509615384615383,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046586110044158184,
      "loss": 5.1337,
      "mean_token_accuracy": 0.19478118121623994,
      "num_tokens": 42351756.0,
      "step": 20405
    },
    {
      "entropy": 5.573098421096802,
      "epoch": 1.8514150943396226,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046584396375444335,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21466975510120392,
      "num_tokens": 42361649.0,
      "step": 20410
    },
    {
      "entropy": 5.54986982345581,
      "epoch": 1.8518686502177069,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046582682312060295,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20996458381414412,
      "num_tokens": 42371273.0,
      "step": 20415
    },
    {
      "entropy": 5.4714539527893065,
      "epoch": 1.8523222060957911,
      "grad_norm": 1.0,
      "learning_rate": 0.0004658096785404152,
      "loss": 4.9663,
      "mean_token_accuracy": 0.20906607210636138,
      "num_tokens": 42382275.0,
      "step": 20420
    },
    {
      "entropy": 5.61240701675415,
      "epoch": 1.8527757619738752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046579253001423473,
      "loss": 5.0182,
      "mean_token_accuracy": 0.201275971531868,
      "num_tokens": 42392360.0,
      "step": 20425
    },
    {
      "entropy": 5.576451110839844,
      "epoch": 1.8532293178519592,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046577537754241604,
      "loss": 5.0072,
      "mean_token_accuracy": 0.19999832063913345,
      "num_tokens": 42403695.0,
      "step": 20430
    },
    {
      "entropy": 5.557355308532715,
      "epoch": 1.8536828737300435,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000465758221125314,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20686184316873552,
      "num_tokens": 42414199.0,
      "step": 20435
    },
    {
      "entropy": 5.566457271575928,
      "epoch": 1.8541364296081277,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046574106076328333,
      "loss": 5.0809,
      "mean_token_accuracy": 0.2030840277671814,
      "num_tokens": 42423655.0,
      "step": 20440
    },
    {
      "entropy": 5.526126050949097,
      "epoch": 1.854589985486212,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004657238964566789,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21088251173496247,
      "num_tokens": 42434235.0,
      "step": 20445
    },
    {
      "entropy": 5.480642986297608,
      "epoch": 1.855043541364296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004657067282058557,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20678504556417465,
      "num_tokens": 42444701.0,
      "step": 20450
    },
    {
      "entropy": 5.4687940120697025,
      "epoch": 1.85549709724238,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004656895560111687,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20691045373678207,
      "num_tokens": 42454499.0,
      "step": 20455
    },
    {
      "entropy": 5.5942786693572994,
      "epoch": 1.8559506531204644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046567237987297323,
      "loss": 4.9735,
      "mean_token_accuracy": 0.1980825424194336,
      "num_tokens": 42465391.0,
      "step": 20460
    },
    {
      "entropy": 5.624168586730957,
      "epoch": 1.8564042089985486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004656551997916243,
      "loss": 4.9827,
      "mean_token_accuracy": 0.1953688681125641,
      "num_tokens": 42476126.0,
      "step": 20465
    },
    {
      "entropy": 5.570189476013184,
      "epoch": 1.856857764876633,
      "grad_norm": 1.0,
      "learning_rate": 0.00046563801576747716,
      "loss": 5.0589,
      "mean_token_accuracy": 0.2046019032597542,
      "num_tokens": 42486833.0,
      "step": 20470
    },
    {
      "entropy": 5.648507928848266,
      "epoch": 1.857311320754717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046562082780088733,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20140296071767808,
      "num_tokens": 42496986.0,
      "step": 20475
    },
    {
      "entropy": 5.571720695495605,
      "epoch": 1.8577648766328012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004656036358922103,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20405839681625365,
      "num_tokens": 42507352.0,
      "step": 20480
    },
    {
      "entropy": 5.565200138092041,
      "epoch": 1.8582184325108853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004655864400418014,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20897690653800965,
      "num_tokens": 42517655.0,
      "step": 20485
    },
    {
      "entropy": 5.513050222396851,
      "epoch": 1.8586719883889695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004655692402500163,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21601160764694213,
      "num_tokens": 42527502.0,
      "step": 20490
    },
    {
      "entropy": 5.529283952713013,
      "epoch": 1.8591255442670538,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046555203651721085,
      "loss": 4.9749,
      "mean_token_accuracy": 0.2061847746372223,
      "num_tokens": 42538893.0,
      "step": 20495
    },
    {
      "entropy": 5.429849100112915,
      "epoch": 1.8595791001451378,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004655348288437407,
      "loss": 4.861,
      "mean_token_accuracy": 0.21934590935707093,
      "num_tokens": 42550372.0,
      "step": 20500
    },
    {
      "entropy": 5.584519672393799,
      "epoch": 1.860032656023222,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004655176172299618,
      "loss": 5.0942,
      "mean_token_accuracy": 0.19986526370048524,
      "num_tokens": 42560814.0,
      "step": 20505
    },
    {
      "entropy": 5.634581851959228,
      "epoch": 1.8604862119013061,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046550040167622986,
      "loss": 5.0769,
      "mean_token_accuracy": 0.19995513409376145,
      "num_tokens": 42570993.0,
      "step": 20510
    },
    {
      "entropy": 5.559375667572022,
      "epoch": 1.8609397677793904,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046548318218290124,
      "loss": 4.9351,
      "mean_token_accuracy": 0.20816790759563447,
      "num_tokens": 42581357.0,
      "step": 20515
    },
    {
      "entropy": 5.523706865310669,
      "epoch": 1.8613933236574747,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004654659587503318,
      "loss": 5.0758,
      "mean_token_accuracy": 0.19584809392690658,
      "num_tokens": 42591653.0,
      "step": 20520
    },
    {
      "entropy": 5.612924957275391,
      "epoch": 1.861846879535559,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004654487313788778,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19749281257390977,
      "num_tokens": 42602232.0,
      "step": 20525
    },
    {
      "entropy": 5.571268939971924,
      "epoch": 1.862300435413643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046543150006889546,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20954685807228088,
      "num_tokens": 42613283.0,
      "step": 20530
    },
    {
      "entropy": 5.507944822311401,
      "epoch": 1.862753991291727,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004654142648207412,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21491098999977112,
      "num_tokens": 42623594.0,
      "step": 20535
    },
    {
      "entropy": 5.565047264099121,
      "epoch": 1.8632075471698113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004653970256347715,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20384258925914764,
      "num_tokens": 42633587.0,
      "step": 20540
    },
    {
      "entropy": 5.593127250671387,
      "epoch": 1.8636611030478956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046537978251134265,
      "loss": 4.9958,
      "mean_token_accuracy": 0.20397452265024185,
      "num_tokens": 42643426.0,
      "step": 20545
    },
    {
      "entropy": 5.48998236656189,
      "epoch": 1.8641146589259798,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004653625354508115,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20894977897405625,
      "num_tokens": 42653521.0,
      "step": 20550
    },
    {
      "entropy": 5.5964569568634035,
      "epoch": 1.8645682148040639,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004653452844535346,
      "loss": 5.0268,
      "mean_token_accuracy": 0.2043809026479721,
      "num_tokens": 42664818.0,
      "step": 20555
    },
    {
      "entropy": 5.667501926422119,
      "epoch": 1.865021770682148,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046532802951986863,
      "loss": 5.0684,
      "mean_token_accuracy": 0.20821174383163452,
      "num_tokens": 42674738.0,
      "step": 20560
    },
    {
      "entropy": 5.529953861236573,
      "epoch": 1.8654753265602322,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004653107706501705,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21852142363786697,
      "num_tokens": 42685003.0,
      "step": 20565
    },
    {
      "entropy": 5.484201097488404,
      "epoch": 1.8659288824383164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004652935078447973,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20671058297157288,
      "num_tokens": 42695428.0,
      "step": 20570
    },
    {
      "entropy": 5.577590703964233,
      "epoch": 1.8663824383164007,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046527624110410574,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20736818611621857,
      "num_tokens": 42704901.0,
      "step": 20575
    },
    {
      "entropy": 5.609286212921143,
      "epoch": 1.8668359941944848,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000465258970428453,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20195237398147584,
      "num_tokens": 42714379.0,
      "step": 20580
    },
    {
      "entropy": 5.624383544921875,
      "epoch": 1.8672895500725688,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046524169581819633,
      "loss": 5.1681,
      "mean_token_accuracy": 0.19813774675130844,
      "num_tokens": 42724813.0,
      "step": 20585
    },
    {
      "entropy": 5.512622976303101,
      "epoch": 1.867743105950653,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004652244172736929,
      "loss": 5.0003,
      "mean_token_accuracy": 0.20564983189105987,
      "num_tokens": 42736590.0,
      "step": 20590
    },
    {
      "entropy": 5.616318225860596,
      "epoch": 1.8681966618287373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004652071347953,
      "loss": 5.0857,
      "mean_token_accuracy": 0.1994020715355873,
      "num_tokens": 42746718.0,
      "step": 20595
    },
    {
      "entropy": 5.560640478134156,
      "epoch": 1.8686502177068216,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046518984838337513,
      "loss": 4.9689,
      "mean_token_accuracy": 0.20223457664251326,
      "num_tokens": 42756484.0,
      "step": 20600
    },
    {
      "entropy": 5.579528045654297,
      "epoch": 1.8691037735849056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004651725580382757,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20161826014518738,
      "num_tokens": 42768512.0,
      "step": 20605
    },
    {
      "entropy": 5.533065414428711,
      "epoch": 1.8695573294629897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046515526376035934,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21082388907670974,
      "num_tokens": 42779812.0,
      "step": 20610
    },
    {
      "entropy": 5.4643162250518795,
      "epoch": 1.870010885341074,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004651379655499837,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21543058305978774,
      "num_tokens": 42791163.0,
      "step": 20615
    },
    {
      "entropy": 5.540845632553101,
      "epoch": 1.8704644412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046512066340750636,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20470986813306807,
      "num_tokens": 42801810.0,
      "step": 20620
    },
    {
      "entropy": 5.603699636459351,
      "epoch": 1.8709179970972425,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004651033573332853,
      "loss": 5.0062,
      "mean_token_accuracy": 0.20591593086719512,
      "num_tokens": 42813358.0,
      "step": 20625
    },
    {
      "entropy": 5.460213851928711,
      "epoch": 1.8713715529753265,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004650860473276784,
      "loss": 4.8519,
      "mean_token_accuracy": 0.21921395212411882,
      "num_tokens": 42823519.0,
      "step": 20630
    },
    {
      "entropy": 5.571874475479126,
      "epoch": 1.8718251088534106,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004650687333910436,
      "loss": 5.0334,
      "mean_token_accuracy": 0.20522348582744598,
      "num_tokens": 42833257.0,
      "step": 20635
    },
    {
      "entropy": 5.595556211471558,
      "epoch": 1.8722786647314948,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046505141552373897,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21454567611217498,
      "num_tokens": 42843536.0,
      "step": 20640
    },
    {
      "entropy": 5.54931116104126,
      "epoch": 1.8727322206095791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004650340937261226,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2033155769109726,
      "num_tokens": 42853911.0,
      "step": 20645
    },
    {
      "entropy": 5.641363668441772,
      "epoch": 1.8731857764876634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004650167679985528,
      "loss": 5.1005,
      "mean_token_accuracy": 0.19384931772947311,
      "num_tokens": 42864173.0,
      "step": 20650
    },
    {
      "entropy": 5.5625217914581295,
      "epoch": 1.8736393323657474,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004649994383413877,
      "loss": 4.9542,
      "mean_token_accuracy": 0.20902465730905534,
      "num_tokens": 42873470.0,
      "step": 20655
    },
    {
      "entropy": 5.509566593170166,
      "epoch": 1.8740928882438317,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046498210475498595,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21467747837305068,
      "num_tokens": 42882496.0,
      "step": 20660
    },
    {
      "entropy": 5.505311059951782,
      "epoch": 1.8745464441219157,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046496476723970574,
      "loss": 4.9549,
      "mean_token_accuracy": 0.20503194034099578,
      "num_tokens": 42893163.0,
      "step": 20665
    },
    {
      "entropy": 5.522883796691895,
      "epoch": 1.875,
      "grad_norm": 1.125,
      "learning_rate": 0.0004649474257959058,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20853859633207322,
      "num_tokens": 42903406.0,
      "step": 20670
    },
    {
      "entropy": 5.593277215957642,
      "epoch": 1.8754535558780843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004649300804239447,
      "loss": 4.9718,
      "mean_token_accuracy": 0.20735423564910888,
      "num_tokens": 42913702.0,
      "step": 20675
    },
    {
      "entropy": 5.506184911727905,
      "epoch": 1.8759071117561683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046491273112418105,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2134859025478363,
      "num_tokens": 42923877.0,
      "step": 20680
    },
    {
      "entropy": 5.5144805908203125,
      "epoch": 1.8763606676342526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046489537789697386,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20901622474193574,
      "num_tokens": 42934393.0,
      "step": 20685
    },
    {
      "entropy": 5.567466974258423,
      "epoch": 1.8768142235123366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046487802074268177,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21161091774702073,
      "num_tokens": 42945803.0,
      "step": 20690
    },
    {
      "entropy": 5.5185154438018795,
      "epoch": 1.8772677793904209,
      "grad_norm": 1.125,
      "learning_rate": 0.00046486065966166377,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21167248189449311,
      "num_tokens": 42955829.0,
      "step": 20695
    },
    {
      "entropy": 5.543769025802613,
      "epoch": 1.8777213352685052,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004648432946542791,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20266976952552795,
      "num_tokens": 42967080.0,
      "step": 20700
    },
    {
      "entropy": 5.512664699554444,
      "epoch": 1.8781748911465894,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046482592572088654,
      "loss": 4.9818,
      "mean_token_accuracy": 0.2005089059472084,
      "num_tokens": 42978082.0,
      "step": 20705
    },
    {
      "entropy": 5.572058773040771,
      "epoch": 1.8786284470246735,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046480855286184565,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20274498760700227,
      "num_tokens": 42988982.0,
      "step": 20710
    },
    {
      "entropy": 5.5833902835845945,
      "epoch": 1.8790820029027575,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046479117607751535,
      "loss": 4.9986,
      "mean_token_accuracy": 0.1983781024813652,
      "num_tokens": 42998982.0,
      "step": 20715
    },
    {
      "entropy": 5.55176157951355,
      "epoch": 1.8795355587808418,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046477379536825534,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20915919542312622,
      "num_tokens": 43009574.0,
      "step": 20720
    },
    {
      "entropy": 5.560165309906006,
      "epoch": 1.879989114658926,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004647564107344247,
      "loss": 4.9931,
      "mean_token_accuracy": 0.20352196842432022,
      "num_tokens": 43019335.0,
      "step": 20725
    },
    {
      "entropy": 5.481369352340698,
      "epoch": 1.8804426705370103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046473902217638327,
      "loss": 5.0163,
      "mean_token_accuracy": 0.20774939954280852,
      "num_tokens": 43029264.0,
      "step": 20730
    },
    {
      "entropy": 5.516702032089233,
      "epoch": 1.8808962264150944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046472162969449037,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2099528893828392,
      "num_tokens": 43040519.0,
      "step": 20735
    },
    {
      "entropy": 5.604666137695313,
      "epoch": 1.8813497822931784,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046470423328910594,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20703690648078918,
      "num_tokens": 43051146.0,
      "step": 20740
    },
    {
      "entropy": 5.6006255626678465,
      "epoch": 1.8818033381712627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046468683296058957,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20637607127428054,
      "num_tokens": 43061339.0,
      "step": 20745
    },
    {
      "entropy": 5.504037857055664,
      "epoch": 1.882256894049347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046466942870930126,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20894727855920792,
      "num_tokens": 43072624.0,
      "step": 20750
    },
    {
      "entropy": 5.590467262268066,
      "epoch": 1.8827104499274312,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004646520205356007,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2081509128212929,
      "num_tokens": 43082120.0,
      "step": 20755
    },
    {
      "entropy": 5.649250793457031,
      "epoch": 1.8831640058055152,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046463460843984805,
      "loss": 5.0907,
      "mean_token_accuracy": 0.20072863548994063,
      "num_tokens": 43093202.0,
      "step": 20760
    },
    {
      "entropy": 5.665876722335815,
      "epoch": 1.8836175616835993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004646171924224034,
      "loss": 5.1642,
      "mean_token_accuracy": 0.2048816204071045,
      "num_tokens": 43104801.0,
      "step": 20765
    },
    {
      "entropy": 5.5828657150268555,
      "epoch": 1.8840711175616836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046459977248362687,
      "loss": 4.9738,
      "mean_token_accuracy": 0.2094803050160408,
      "num_tokens": 43114855.0,
      "step": 20770
    },
    {
      "entropy": 5.570040464401245,
      "epoch": 1.8845246734397678,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004645823486238788,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20368480086326599,
      "num_tokens": 43124759.0,
      "step": 20775
    },
    {
      "entropy": 5.521081876754761,
      "epoch": 1.884978229317852,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004645649208435193,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21389394551515578,
      "num_tokens": 43135526.0,
      "step": 20780
    },
    {
      "entropy": 5.5732262134552,
      "epoch": 1.8854317851959361,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046454748914290906,
      "loss": 5.0718,
      "mean_token_accuracy": 0.20253242701292037,
      "num_tokens": 43145605.0,
      "step": 20785
    },
    {
      "entropy": 5.642473983764648,
      "epoch": 1.8858853410740202,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046453005352240835,
      "loss": 5.1235,
      "mean_token_accuracy": 0.1973668545484543,
      "num_tokens": 43156578.0,
      "step": 20790
    },
    {
      "entropy": 5.60530800819397,
      "epoch": 1.8863388969521044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000464512613982378,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20926489681005478,
      "num_tokens": 43166870.0,
      "step": 20795
    },
    {
      "entropy": 5.562167406082153,
      "epoch": 1.8867924528301887,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004644951705231783,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2154717341065407,
      "num_tokens": 43176406.0,
      "step": 20800
    },
    {
      "entropy": 5.656374311447143,
      "epoch": 1.887246008708273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046447772314517024,
      "loss": 5.1511,
      "mean_token_accuracy": 0.2020910695195198,
      "num_tokens": 43187502.0,
      "step": 20805
    },
    {
      "entropy": 5.501628065109253,
      "epoch": 1.887699564586357,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046446027184871464,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21230237782001496,
      "num_tokens": 43197928.0,
      "step": 20810
    },
    {
      "entropy": 5.533857202529907,
      "epoch": 1.888153120464441,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046444281663417236,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21144766956567765,
      "num_tokens": 43209083.0,
      "step": 20815
    },
    {
      "entropy": 5.538638877868652,
      "epoch": 1.8886066763425253,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004644253575019043,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2015242889523506,
      "num_tokens": 43219089.0,
      "step": 20820
    },
    {
      "entropy": 5.602917671203613,
      "epoch": 1.8890602322206096,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004644078944522716,
      "loss": 5.051,
      "mean_token_accuracy": 0.20531389713287354,
      "num_tokens": 43229440.0,
      "step": 20825
    },
    {
      "entropy": 5.597049617767334,
      "epoch": 1.8895137880986939,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004643904274856353,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20621335208415986,
      "num_tokens": 43239995.0,
      "step": 20830
    },
    {
      "entropy": 5.554989004135132,
      "epoch": 1.889967343976778,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046437295660235674,
      "loss": 4.977,
      "mean_token_accuracy": 0.21172338724136353,
      "num_tokens": 43249722.0,
      "step": 20835
    },
    {
      "entropy": 5.4992584705352785,
      "epoch": 1.8904208998548622,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004643554818027972,
      "loss": 5.0197,
      "mean_token_accuracy": 0.2075279548764229,
      "num_tokens": 43260029.0,
      "step": 20840
    },
    {
      "entropy": 5.554016780853272,
      "epoch": 1.8908744557329462,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004643380030873181,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20793743878602983,
      "num_tokens": 43270571.0,
      "step": 20845
    },
    {
      "entropy": 5.570363712310791,
      "epoch": 1.8913280116110305,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004643205204562808,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21176953464746476,
      "num_tokens": 43281310.0,
      "step": 20850
    },
    {
      "entropy": 5.55697021484375,
      "epoch": 1.8917815674891147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046430303391004687,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20941325575113295,
      "num_tokens": 43292904.0,
      "step": 20855
    },
    {
      "entropy": 5.554282379150391,
      "epoch": 1.8922351233671988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004642855434489779,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2037817120552063,
      "num_tokens": 43303745.0,
      "step": 20860
    },
    {
      "entropy": 5.586295175552368,
      "epoch": 1.892688679245283,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046426804907343574,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20697399526834487,
      "num_tokens": 43314652.0,
      "step": 20865
    },
    {
      "entropy": 5.527583217620849,
      "epoch": 1.893142235123367,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046425055078378206,
      "loss": 4.8744,
      "mean_token_accuracy": 0.21875638663768768,
      "num_tokens": 43324242.0,
      "step": 20870
    },
    {
      "entropy": 5.523820161819458,
      "epoch": 1.8935957910014514,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046423304858037885,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21595541387796402,
      "num_tokens": 43334054.0,
      "step": 20875
    },
    {
      "entropy": 5.469158792495728,
      "epoch": 1.8940493468795356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004642155424635878,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2215532347559929,
      "num_tokens": 43343389.0,
      "step": 20880
    },
    {
      "entropy": 5.477652502059937,
      "epoch": 1.89450290275762,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046419803243377125,
      "loss": 4.9423,
      "mean_token_accuracy": 0.2153505265712738,
      "num_tokens": 43353892.0,
      "step": 20885
    },
    {
      "entropy": 5.501353168487549,
      "epoch": 1.894956458635704,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046418051849129115,
      "loss": 4.972,
      "mean_token_accuracy": 0.214233835041523,
      "num_tokens": 43363912.0,
      "step": 20890
    },
    {
      "entropy": 5.545388412475586,
      "epoch": 1.895410014513788,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004641630006365097,
      "loss": 5.0428,
      "mean_token_accuracy": 0.20356074273586272,
      "num_tokens": 43374719.0,
      "step": 20895
    },
    {
      "entropy": 5.580736112594605,
      "epoch": 1.8958635703918723,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046414547886978926,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2027682676911354,
      "num_tokens": 43385549.0,
      "step": 20900
    },
    {
      "entropy": 5.600714111328125,
      "epoch": 1.8963171262699565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046412795319149204,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21062333434820174,
      "num_tokens": 43395503.0,
      "step": 20905
    },
    {
      "entropy": 5.538871383666992,
      "epoch": 1.8967706821480408,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046411042360198053,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2010958045721054,
      "num_tokens": 43406913.0,
      "step": 20910
    },
    {
      "entropy": 5.562172079086304,
      "epoch": 1.8972242380261248,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004640928901016173,
      "loss": 4.947,
      "mean_token_accuracy": 0.20708273500204086,
      "num_tokens": 43416459.0,
      "step": 20915
    },
    {
      "entropy": 5.53361029624939,
      "epoch": 1.8976777939042089,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046407535269076493,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20667988210916519,
      "num_tokens": 43426280.0,
      "step": 20920
    },
    {
      "entropy": 5.610696172714233,
      "epoch": 1.8981313497822931,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046405781136978606,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20656325072050094,
      "num_tokens": 43436676.0,
      "step": 20925
    },
    {
      "entropy": 5.660415744781494,
      "epoch": 1.8985849056603774,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046404026613904357,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2016023129224777,
      "num_tokens": 43447019.0,
      "step": 20930
    },
    {
      "entropy": 5.605124664306641,
      "epoch": 1.8990384615384617,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004640227169989001,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2014377772808075,
      "num_tokens": 43456609.0,
      "step": 20935
    },
    {
      "entropy": 5.576879262924194,
      "epoch": 1.8994920174165457,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004640051639497187,
      "loss": 5.0459,
      "mean_token_accuracy": 0.2026502937078476,
      "num_tokens": 43467616.0,
      "step": 20940
    },
    {
      "entropy": 5.485981082916259,
      "epoch": 1.8999455732946298,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004639876069918624,
      "loss": 5.0186,
      "mean_token_accuracy": 0.20398564338684083,
      "num_tokens": 43478209.0,
      "step": 20945
    },
    {
      "entropy": 5.527855396270752,
      "epoch": 1.900399129172714,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046397004612569416,
      "loss": 4.9072,
      "mean_token_accuracy": 0.20824503153562546,
      "num_tokens": 43487278.0,
      "step": 20950
    },
    {
      "entropy": 5.507301187515258,
      "epoch": 1.9008526850507983,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046395248135157715,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2197753131389618,
      "num_tokens": 43497457.0,
      "step": 20955
    },
    {
      "entropy": 5.562612533569336,
      "epoch": 1.9013062409288826,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046393491266987484,
      "loss": 5.0487,
      "mean_token_accuracy": 0.20178602188825606,
      "num_tokens": 43507761.0,
      "step": 20960
    },
    {
      "entropy": 5.497231721878052,
      "epoch": 1.9017597968069666,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004639173400809502,
      "loss": 4.9672,
      "mean_token_accuracy": 0.20882880091667175,
      "num_tokens": 43517198.0,
      "step": 20965
    },
    {
      "entropy": 5.540993881225586,
      "epoch": 1.9022133526850507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004638997635851669,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20050388872623442,
      "num_tokens": 43527429.0,
      "step": 20970
    },
    {
      "entropy": 5.612675666809082,
      "epoch": 1.902666908563135,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046388218318288844,
      "loss": 5.0583,
      "mean_token_accuracy": 0.20635149627923965,
      "num_tokens": 43537867.0,
      "step": 20975
    },
    {
      "entropy": 5.600853681564331,
      "epoch": 1.9031204644412192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004638645988744782,
      "loss": 4.995,
      "mean_token_accuracy": 0.20810700356960296,
      "num_tokens": 43547284.0,
      "step": 20980
    },
    {
      "entropy": 5.544989919662475,
      "epoch": 1.9035740203193035,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004638470106603,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20433826893568038,
      "num_tokens": 43557693.0,
      "step": 20985
    },
    {
      "entropy": 5.580530118942261,
      "epoch": 1.9040275761973875,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004638294185407174,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21300598680973054,
      "num_tokens": 43566909.0,
      "step": 20990
    },
    {
      "entropy": 5.587985372543335,
      "epoch": 1.9044811320754715,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638118225160944,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21214412152767181,
      "num_tokens": 43576963.0,
      "step": 20995
    },
    {
      "entropy": 5.497220659255982,
      "epoch": 1.9049346879535558,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046379422258679477,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20666191428899766,
      "num_tokens": 43586649.0,
      "step": 21000
    },
    {
      "epoch": 1.9049346879535558,
      "eval_entropy": 5.2827021436993356,
      "eval_loss": 5.041497230529785,
      "eval_mean_token_accuracy": 0.21340517185426136,
      "eval_num_tokens": 43586649.0,
      "eval_runtime": 20.6472,
      "eval_samples_per_second": 1712.58,
      "eval_steps_per_second": 214.072,
      "step": 21000
    }
  ],
  "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": 9472777408696320.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}