nld-latn-10mb-100mb_seed3407/checkpoint-6000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5442670537010159,
  "eval_steps": 3000,
  "global_step": 6000,
  "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
    }
  ],
  "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": 2706712035655680.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}