{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.354136429608127,
  "eval_steps": 3000,
  "global_step": 48000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.742568969726562,
      "epoch": 0.00045355587808418,
      "grad_norm": 5.4375,
      "learning_rate": 2e-06,
      "loss": 10.7846,
      "mean_token_accuracy": 0.0,
      "num_tokens": 10827.0,
      "step": 5
    },
    {
      "entropy": 10.742588996887207,
      "epoch": 0.00090711175616836,
      "grad_norm": 5.25,
      "learning_rate": 4.5e-06,
      "loss": 10.7602,
      "mean_token_accuracy": 6.963788182474673e-05,
      "num_tokens": 21025.0,
      "step": 10
    },
    {
      "entropy": 10.742573070526124,
      "epoch": 0.00136066763425254,
      "grad_norm": 5.25,
      "learning_rate": 7e-06,
      "loss": 10.7437,
      "mean_token_accuracy": 9.000900317914783e-05,
      "num_tokens": 31678.0,
      "step": 15
    },
    {
      "entropy": 10.742661285400391,
      "epoch": 0.00181422351233672,
      "grad_norm": 5.125,
      "learning_rate": 9.5e-06,
      "loss": 10.6591,
      "mean_token_accuracy": 0.0005137635511346161,
      "num_tokens": 41025.0,
      "step": 20
    },
    {
      "entropy": 10.742550945281982,
      "epoch": 0.0022677793904208998,
      "grad_norm": 4.875,
      "learning_rate": 1.2e-05,
      "loss": 10.5792,
      "mean_token_accuracy": 0.006170097610447556,
      "num_tokens": 51270.0,
      "step": 25
    },
    {
      "entropy": 10.742267036437989,
      "epoch": 0.00272133526850508,
      "grad_norm": 3.953125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.4637,
      "mean_token_accuracy": 0.027165642101317646,
      "num_tokens": 62142.0,
      "step": 30
    },
    {
      "entropy": 10.741279792785644,
      "epoch": 0.00317489114658926,
      "grad_norm": 3.390625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3199,
      "mean_token_accuracy": 0.04252696111798286,
      "num_tokens": 72709.0,
      "step": 35
    },
    {
      "entropy": 10.738855934143066,
      "epoch": 0.00362844702467344,
      "grad_norm": 2.625,
      "learning_rate": 1.95e-05,
      "loss": 10.1477,
      "mean_token_accuracy": 0.041494901105761525,
      "num_tokens": 82313.0,
      "step": 40
    },
    {
      "entropy": 10.734966659545899,
      "epoch": 0.00408200290275762,
      "grad_norm": 2.484375,
      "learning_rate": 2.2e-05,
      "loss": 10.0172,
      "mean_token_accuracy": 0.048007883876562116,
      "num_tokens": 92450.0,
      "step": 45
    },
    {
      "entropy": 10.731274795532226,
      "epoch": 0.0045355587808417995,
      "grad_norm": 2.125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.9997,
      "mean_token_accuracy": 0.0419012438505888,
      "num_tokens": 103978.0,
      "step": 50
    },
    {
      "entropy": 10.729759407043456,
      "epoch": 0.00498911465892598,
      "grad_norm": 1.96875,
      "learning_rate": 2.7e-05,
      "loss": 9.8946,
      "mean_token_accuracy": 0.04565211050212383,
      "num_tokens": 114134.0,
      "step": 55
    },
    {
      "entropy": 10.727651214599609,
      "epoch": 0.00544267053701016,
      "grad_norm": 2.015625,
      "learning_rate": 2.95e-05,
      "loss": 9.8734,
      "mean_token_accuracy": 0.04470008127391338,
      "num_tokens": 125037.0,
      "step": 60
    },
    {
      "entropy": 10.725145721435547,
      "epoch": 0.00589622641509434,
      "grad_norm": 1.9375,
      "learning_rate": 3.2e-05,
      "loss": 9.7559,
      "mean_token_accuracy": 0.04514850843697786,
      "num_tokens": 134721.0,
      "step": 65
    },
    {
      "entropy": 10.723241519927978,
      "epoch": 0.00634978229317852,
      "grad_norm": 2.03125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.6935,
      "mean_token_accuracy": 0.04249380584806204,
      "num_tokens": 144448.0,
      "step": 70
    },
    {
      "entropy": 10.721195030212403,
      "epoch": 0.006803338171262699,
      "grad_norm": 2.015625,
      "learning_rate": 3.7e-05,
      "loss": 9.635,
      "mean_token_accuracy": 0.04518970362842083,
      "num_tokens": 154343.0,
      "step": 75
    },
    {
      "entropy": 10.7159348487854,
      "epoch": 0.00725689404934688,
      "grad_norm": 2.0,
      "learning_rate": 3.95e-05,
      "loss": 9.6121,
      "mean_token_accuracy": 0.041934683173894885,
      "num_tokens": 165650.0,
      "step": 80
    },
    {
      "entropy": 10.709981918334961,
      "epoch": 0.007710449927431059,
      "grad_norm": 1.828125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6128,
      "mean_token_accuracy": 0.03882604259997606,
      "num_tokens": 177098.0,
      "step": 85
    },
    {
      "entropy": 10.703641796112061,
      "epoch": 0.00816400580551524,
      "grad_norm": 2.015625,
      "learning_rate": 4.45e-05,
      "loss": 9.4133,
      "mean_token_accuracy": 0.04838715270161629,
      "num_tokens": 187903.0,
      "step": 90
    },
    {
      "entropy": 10.691870975494385,
      "epoch": 0.00861756168359942,
      "grad_norm": 1.9296875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.3442,
      "mean_token_accuracy": 0.04516914151608944,
      "num_tokens": 198077.0,
      "step": 95
    },
    {
      "entropy": 10.677101039886475,
      "epoch": 0.009071117561683599,
      "grad_norm": 1.859375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.3156,
      "mean_token_accuracy": 0.04309112168848515,
      "num_tokens": 208641.0,
      "step": 100
    },
    {
      "entropy": 10.661001300811767,
      "epoch": 0.009524673439767779,
      "grad_norm": 1.9296875,
      "learning_rate": 5.2e-05,
      "loss": 9.1598,
      "mean_token_accuracy": 0.05524700954556465,
      "num_tokens": 218398.0,
      "step": 105
    },
    {
      "entropy": 10.64128761291504,
      "epoch": 0.00997822931785196,
      "grad_norm": 1.7578125,
      "learning_rate": 5.45e-05,
      "loss": 9.035,
      "mean_token_accuracy": 0.06018046177923679,
      "num_tokens": 229059.0,
      "step": 110
    },
    {
      "entropy": 10.614240169525146,
      "epoch": 0.01043178519593614,
      "grad_norm": 1.796875,
      "learning_rate": 5.7e-05,
      "loss": 8.9594,
      "mean_token_accuracy": 0.05966283418238163,
      "num_tokens": 239775.0,
      "step": 115
    },
    {
      "entropy": 10.567374324798584,
      "epoch": 0.01088534107402032,
      "grad_norm": 1.9140625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.8042,
      "mean_token_accuracy": 0.06268535517156124,
      "num_tokens": 249954.0,
      "step": 120
    },
    {
      "entropy": 10.49941110610962,
      "epoch": 0.011338896952104499,
      "grad_norm": 1.71875,
      "learning_rate": 6.2e-05,
      "loss": 8.7019,
      "mean_token_accuracy": 0.06550404466688634,
      "num_tokens": 260056.0,
      "step": 125
    },
    {
      "entropy": 10.433360958099366,
      "epoch": 0.01179245283018868,
      "grad_norm": 1.6015625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.5999,
      "mean_token_accuracy": 0.06795549392700195,
      "num_tokens": 269904.0,
      "step": 130
    },
    {
      "entropy": 10.364013862609863,
      "epoch": 0.01224600870827286,
      "grad_norm": 1.703125,
      "learning_rate": 6.7e-05,
      "loss": 8.4687,
      "mean_token_accuracy": 0.06760800294578076,
      "num_tokens": 280900.0,
      "step": 135
    },
    {
      "entropy": 10.2804931640625,
      "epoch": 0.01269956458635704,
      "grad_norm": 1.59375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.3637,
      "mean_token_accuracy": 0.067552350461483,
      "num_tokens": 290844.0,
      "step": 140
    },
    {
      "entropy": 10.172641277313232,
      "epoch": 0.013153120464441219,
      "grad_norm": 1.3046875,
      "learning_rate": 7.2e-05,
      "loss": 8.2257,
      "mean_token_accuracy": 0.07797299772500992,
      "num_tokens": 301345.0,
      "step": 145
    },
    {
      "entropy": 10.0338849067688,
      "epoch": 0.013606676342525399,
      "grad_norm": 1.484375,
      "learning_rate": 7.45e-05,
      "loss": 8.0893,
      "mean_token_accuracy": 0.07597564421594143,
      "num_tokens": 312681.0,
      "step": 150
    },
    {
      "entropy": 9.874010562896729,
      "epoch": 0.01406023222060958,
      "grad_norm": 1.203125,
      "learning_rate": 7.7e-05,
      "loss": 7.9627,
      "mean_token_accuracy": 0.08046026974916458,
      "num_tokens": 323492.0,
      "step": 155
    },
    {
      "entropy": 9.684037780761718,
      "epoch": 0.01451378809869376,
      "grad_norm": 1.234375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.9031,
      "mean_token_accuracy": 0.07810574918985366,
      "num_tokens": 334521.0,
      "step": 160
    },
    {
      "entropy": 9.470422744750977,
      "epoch": 0.014967343976777939,
      "grad_norm": 1.015625,
      "learning_rate": 8.2e-05,
      "loss": 7.8175,
      "mean_token_accuracy": 0.08020211830735206,
      "num_tokens": 345468.0,
      "step": 165
    },
    {
      "entropy": 9.22849702835083,
      "epoch": 0.015420899854862119,
      "grad_norm": 1.1015625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.6913,
      "mean_token_accuracy": 0.07718781009316444,
      "num_tokens": 355911.0,
      "step": 170
    },
    {
      "entropy": 9.006134128570556,
      "epoch": 0.0158744557329463,
      "grad_norm": 1.03125,
      "learning_rate": 8.7e-05,
      "loss": 7.6633,
      "mean_token_accuracy": 0.07722515687346458,
      "num_tokens": 367537.0,
      "step": 175
    },
    {
      "entropy": 8.818270492553712,
      "epoch": 0.01632801161103048,
      "grad_norm": 0.94921875,
      "learning_rate": 8.95e-05,
      "loss": 7.7245,
      "mean_token_accuracy": 0.07779238075017929,
      "num_tokens": 377918.0,
      "step": 180
    },
    {
      "entropy": 8.593470573425293,
      "epoch": 0.01678156748911466,
      "grad_norm": 0.85546875,
      "learning_rate": 9.2e-05,
      "loss": 7.5619,
      "mean_token_accuracy": 0.07740958072245122,
      "num_tokens": 388461.0,
      "step": 185
    },
    {
      "entropy": 8.449785518646241,
      "epoch": 0.01723512336719884,
      "grad_norm": 0.984375,
      "learning_rate": 9.45e-05,
      "loss": 7.5162,
      "mean_token_accuracy": 0.07913253530859947,
      "num_tokens": 399771.0,
      "step": 190
    },
    {
      "entropy": 8.267493534088135,
      "epoch": 0.01768867924528302,
      "grad_norm": 0.8125,
      "learning_rate": 9.7e-05,
      "loss": 7.3798,
      "mean_token_accuracy": 0.08124997988343238,
      "num_tokens": 410254.0,
      "step": 195
    },
    {
      "entropy": 8.230424785614014,
      "epoch": 0.018142235123367198,
      "grad_norm": 0.84765625,
      "learning_rate": 9.95e-05,
      "loss": 7.3917,
      "mean_token_accuracy": 0.07860286422073841,
      "num_tokens": 419792.0,
      "step": 200
    },
    {
      "entropy": 8.091243362426757,
      "epoch": 0.018595791001451378,
      "grad_norm": 0.96875,
      "learning_rate": 0.000102,
      "loss": 7.3759,
      "mean_token_accuracy": 0.08409927114844322,
      "num_tokens": 429264.0,
      "step": 205
    },
    {
      "entropy": 8.078143405914307,
      "epoch": 0.019049346879535557,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.3197,
      "mean_token_accuracy": 0.08545555844902993,
      "num_tokens": 439624.0,
      "step": 210
    },
    {
      "entropy": 8.048319149017335,
      "epoch": 0.01950290275761974,
      "grad_norm": 0.984375,
      "learning_rate": 0.000107,
      "loss": 7.3071,
      "mean_token_accuracy": 0.08367115408182144,
      "num_tokens": 449556.0,
      "step": 215
    },
    {
      "entropy": 7.937560939788819,
      "epoch": 0.01995645863570392,
      "grad_norm": 1.125,
      "learning_rate": 0.0001095,
      "loss": 7.3073,
      "mean_token_accuracy": 0.08580821454524994,
      "num_tokens": 460243.0,
      "step": 220
    },
    {
      "entropy": 7.98973650932312,
      "epoch": 0.0204100145137881,
      "grad_norm": 1.015625,
      "learning_rate": 0.000112,
      "loss": 7.3523,
      "mean_token_accuracy": 0.08153139427304268,
      "num_tokens": 469767.0,
      "step": 225
    },
    {
      "entropy": 7.94960503578186,
      "epoch": 0.02086357039187228,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0001145,
      "loss": 7.2779,
      "mean_token_accuracy": 0.08782213702797889,
      "num_tokens": 480329.0,
      "step": 230
    },
    {
      "entropy": 7.937411499023438,
      "epoch": 0.02131712626995646,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.2332,
      "mean_token_accuracy": 0.08278771415352822,
      "num_tokens": 490096.0,
      "step": 235
    },
    {
      "entropy": 7.9301097869873045,
      "epoch": 0.02177068214804064,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.2751,
      "mean_token_accuracy": 0.0801933728158474,
      "num_tokens": 501231.0,
      "step": 240
    },
    {
      "entropy": 7.883290576934814,
      "epoch": 0.022224238026124818,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000122,
      "loss": 7.1802,
      "mean_token_accuracy": 0.09071151316165924,
      "num_tokens": 510722.0,
      "step": 245
    },
    {
      "entropy": 7.763730192184449,
      "epoch": 0.022677793904208998,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0001245,
      "loss": 7.225,
      "mean_token_accuracy": 0.08499881774187087,
      "num_tokens": 521452.0,
      "step": 250
    },
    {
      "entropy": 7.808817529678345,
      "epoch": 0.023131349782293177,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000127,
      "loss": 7.2137,
      "mean_token_accuracy": 0.08842154070734978,
      "num_tokens": 532397.0,
      "step": 255
    },
    {
      "entropy": 7.740368890762329,
      "epoch": 0.02358490566037736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001295,
      "loss": 7.1927,
      "mean_token_accuracy": 0.08906038105487823,
      "num_tokens": 542271.0,
      "step": 260
    },
    {
      "entropy": 7.7880730628967285,
      "epoch": 0.02403846153846154,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000132,
      "loss": 7.1677,
      "mean_token_accuracy": 0.09267213717103004,
      "num_tokens": 552411.0,
      "step": 265
    },
    {
      "entropy": 7.73106369972229,
      "epoch": 0.02449201741654572,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.1675,
      "mean_token_accuracy": 0.09035667106509208,
      "num_tokens": 561518.0,
      "step": 270
    },
    {
      "entropy": 7.755526113510132,
      "epoch": 0.0249455732946299,
      "grad_norm": 1.046875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1846,
      "mean_token_accuracy": 0.09218391999602318,
      "num_tokens": 571361.0,
      "step": 275
    },
    {
      "entropy": 7.656434154510498,
      "epoch": 0.02539912917271408,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0001395,
      "loss": 7.0893,
      "mean_token_accuracy": 0.09728347510099411,
      "num_tokens": 580926.0,
      "step": 280
    },
    {
      "entropy": 7.710359334945679,
      "epoch": 0.025852685050798258,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1891,
      "mean_token_accuracy": 0.08993546590209008,
      "num_tokens": 592282.0,
      "step": 285
    },
    {
      "entropy": 7.69242033958435,
      "epoch": 0.026306240928882438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001445,
      "loss": 7.1904,
      "mean_token_accuracy": 0.09070137366652489,
      "num_tokens": 602767.0,
      "step": 290
    },
    {
      "entropy": 7.715097570419312,
      "epoch": 0.026759796806966617,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000147,
      "loss": 7.1064,
      "mean_token_accuracy": 0.0925739623606205,
      "num_tokens": 612763.0,
      "step": 295
    },
    {
      "entropy": 7.563499784469604,
      "epoch": 0.027213352685050797,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0001495,
      "loss": 7.0617,
      "mean_token_accuracy": 0.09330770447850227,
      "num_tokens": 622796.0,
      "step": 300
    },
    {
      "entropy": 7.630935192108154,
      "epoch": 0.027666908563134977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000152,
      "loss": 7.0426,
      "mean_token_accuracy": 0.09754194170236588,
      "num_tokens": 633136.0,
      "step": 305
    },
    {
      "entropy": 7.510959100723267,
      "epoch": 0.02812046444121916,
      "grad_norm": 1.109375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0335,
      "mean_token_accuracy": 0.09690061807632447,
      "num_tokens": 643347.0,
      "step": 310
    },
    {
      "entropy": 7.634088945388794,
      "epoch": 0.02857402031930334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000157,
      "loss": 7.0192,
      "mean_token_accuracy": 0.10197105258703232,
      "num_tokens": 653096.0,
      "step": 315
    },
    {
      "entropy": 7.5339111328125,
      "epoch": 0.02902757619738752,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001595,
      "loss": 7.0519,
      "mean_token_accuracy": 0.09302655458450318,
      "num_tokens": 663557.0,
      "step": 320
    },
    {
      "entropy": 7.588706541061401,
      "epoch": 0.0294811320754717,
      "grad_norm": 1.515625,
      "learning_rate": 0.000162,
      "loss": 7.1186,
      "mean_token_accuracy": 0.0957950733602047,
      "num_tokens": 674395.0,
      "step": 325
    },
    {
      "entropy": 7.544366073608399,
      "epoch": 0.029934687953555878,
      "grad_norm": 1.140625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0165,
      "mean_token_accuracy": 0.0989912673830986,
      "num_tokens": 684776.0,
      "step": 330
    },
    {
      "entropy": 7.589376211166382,
      "epoch": 0.030388243831640058,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9936,
      "mean_token_accuracy": 0.09702750965952874,
      "num_tokens": 694886.0,
      "step": 335
    },
    {
      "entropy": 7.451715087890625,
      "epoch": 0.030841799709724237,
      "grad_norm": 1.234375,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0626,
      "mean_token_accuracy": 0.09623553976416588,
      "num_tokens": 705426.0,
      "step": 340
    },
    {
      "entropy": 7.509190368652344,
      "epoch": 0.03129535558780842,
      "grad_norm": 0.984375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.0567,
      "mean_token_accuracy": 0.09196493923664092,
      "num_tokens": 716271.0,
      "step": 345
    },
    {
      "entropy": 7.5583083152771,
      "epoch": 0.0317489114658926,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0486,
      "mean_token_accuracy": 0.09727451354265212,
      "num_tokens": 726875.0,
      "step": 350
    },
    {
      "entropy": 7.49442138671875,
      "epoch": 0.032202467343976776,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000177,
      "loss": 6.9698,
      "mean_token_accuracy": 0.09514860436320305,
      "num_tokens": 737825.0,
      "step": 355
    },
    {
      "entropy": 7.501305818557739,
      "epoch": 0.03265602322206096,
      "grad_norm": 0.953125,
      "learning_rate": 0.0001795,
      "loss": 6.9607,
      "mean_token_accuracy": 0.0989458367228508,
      "num_tokens": 747424.0,
      "step": 360
    },
    {
      "entropy": 7.407667875289917,
      "epoch": 0.033109579100145135,
      "grad_norm": 1.046875,
      "learning_rate": 0.000182,
      "loss": 6.9748,
      "mean_token_accuracy": 0.09646827578544617,
      "num_tokens": 757649.0,
      "step": 365
    },
    {
      "entropy": 7.480729293823242,
      "epoch": 0.03356313497822932,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001845,
      "loss": 6.946,
      "mean_token_accuracy": 0.09825383350253106,
      "num_tokens": 768556.0,
      "step": 370
    },
    {
      "entropy": 7.410275650024414,
      "epoch": 0.034016690856313495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000187,
      "loss": 6.9456,
      "mean_token_accuracy": 0.0978932335972786,
      "num_tokens": 778331.0,
      "step": 375
    },
    {
      "entropy": 7.457346200942993,
      "epoch": 0.03447024673439768,
      "grad_norm": 1.140625,
      "learning_rate": 0.0001895,
      "loss": 6.9326,
      "mean_token_accuracy": 0.10083384290337563,
      "num_tokens": 788065.0,
      "step": 380
    },
    {
      "entropy": 7.39845495223999,
      "epoch": 0.03492380261248186,
      "grad_norm": 1.078125,
      "learning_rate": 0.000192,
      "loss": 6.9717,
      "mean_token_accuracy": 0.10045002475380897,
      "num_tokens": 798912.0,
      "step": 385
    },
    {
      "entropy": 7.406559228897095,
      "epoch": 0.03537735849056604,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0001945,
      "loss": 6.9143,
      "mean_token_accuracy": 0.10309505760669709,
      "num_tokens": 809580.0,
      "step": 390
    },
    {
      "entropy": 7.444182062149048,
      "epoch": 0.03583091436865022,
      "grad_norm": 1.015625,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.0008,
      "mean_token_accuracy": 0.09738668501377105,
      "num_tokens": 820207.0,
      "step": 395
    },
    {
      "entropy": 7.290267753601074,
      "epoch": 0.036284470246734396,
      "grad_norm": 0.984375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8228,
      "mean_token_accuracy": 0.0949432410299778,
      "num_tokens": 831038.0,
      "step": 400
    },
    {
      "entropy": 7.429319906234741,
      "epoch": 0.03673802612481858,
      "grad_norm": 0.90625,
      "learning_rate": 0.000202,
      "loss": 6.9129,
      "mean_token_accuracy": 0.10030955821275711,
      "num_tokens": 842144.0,
      "step": 405
    },
    {
      "entropy": 7.347342252731323,
      "epoch": 0.037191582002902755,
      "grad_norm": 1.015625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.905,
      "mean_token_accuracy": 0.10373404398560523,
      "num_tokens": 851315.0,
      "step": 410
    },
    {
      "entropy": 7.323047733306884,
      "epoch": 0.03764513788098694,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000207,
      "loss": 6.912,
      "mean_token_accuracy": 0.10021007284522057,
      "num_tokens": 861233.0,
      "step": 415
    },
    {
      "entropy": 7.350712013244629,
      "epoch": 0.038098693759071114,
      "grad_norm": 1.0,
      "learning_rate": 0.0002095,
      "loss": 6.8978,
      "mean_token_accuracy": 0.09711670130491257,
      "num_tokens": 871131.0,
      "step": 420
    },
    {
      "entropy": 7.294861841201782,
      "epoch": 0.0385522496371553,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000212,
      "loss": 6.8749,
      "mean_token_accuracy": 0.09577203691005706,
      "num_tokens": 882195.0,
      "step": 425
    },
    {
      "entropy": 7.353318929672241,
      "epoch": 0.03900580551523948,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002145,
      "loss": 6.8527,
      "mean_token_accuracy": 0.10149310603737831,
      "num_tokens": 891806.0,
      "step": 430
    },
    {
      "entropy": 7.362502574920654,
      "epoch": 0.03945936139332366,
      "grad_norm": 0.953125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.9129,
      "mean_token_accuracy": 0.09761691987514495,
      "num_tokens": 901350.0,
      "step": 435
    },
    {
      "entropy": 7.301689004898071,
      "epoch": 0.03991291727140784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002195,
      "loss": 6.8293,
      "mean_token_accuracy": 0.09842056706547737,
      "num_tokens": 912061.0,
      "step": 440
    },
    {
      "entropy": 7.250009202957154,
      "epoch": 0.040366473149492016,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000222,
      "loss": 6.7954,
      "mean_token_accuracy": 0.09848268181085587,
      "num_tokens": 922220.0,
      "step": 445
    },
    {
      "entropy": 7.2643632888793945,
      "epoch": 0.0408200290275762,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002245,
      "loss": 6.8122,
      "mean_token_accuracy": 0.1012580119073391,
      "num_tokens": 932319.0,
      "step": 450
    },
    {
      "entropy": 7.2017487525939945,
      "epoch": 0.041273584905660375,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8351,
      "mean_token_accuracy": 0.10266126319766045,
      "num_tokens": 943254.0,
      "step": 455
    },
    {
      "entropy": 7.330859661102295,
      "epoch": 0.04172714078374456,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.8445,
      "mean_token_accuracy": 0.09818541109561921,
      "num_tokens": 953921.0,
      "step": 460
    },
    {
      "entropy": 7.193669176101684,
      "epoch": 0.042180696661828734,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7701,
      "mean_token_accuracy": 0.10723233297467231,
      "num_tokens": 963832.0,
      "step": 465
    },
    {
      "entropy": 7.234699535369873,
      "epoch": 0.04263425253991292,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.8614,
      "mean_token_accuracy": 0.09933211356401443,
      "num_tokens": 974066.0,
      "step": 470
    },
    {
      "entropy": 7.233931684494019,
      "epoch": 0.0430878084179971,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000237,
      "loss": 6.8441,
      "mean_token_accuracy": 0.10467069000005721,
      "num_tokens": 984402.0,
      "step": 475
    },
    {
      "entropy": 7.256547212600708,
      "epoch": 0.04354136429608128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002395,
      "loss": 6.7629,
      "mean_token_accuracy": 0.1010866865515709,
      "num_tokens": 993555.0,
      "step": 480
    },
    {
      "entropy": 7.208513498306274,
      "epoch": 0.04399492017416546,
      "grad_norm": 1.015625,
      "learning_rate": 0.000242,
      "loss": 6.7524,
      "mean_token_accuracy": 0.10970519632101058,
      "num_tokens": 1002920.0,
      "step": 485
    },
    {
      "entropy": 7.203861713409424,
      "epoch": 0.044448476052249636,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002445,
      "loss": 6.8285,
      "mean_token_accuracy": 0.09806114360690117,
      "num_tokens": 1013611.0,
      "step": 490
    },
    {
      "entropy": 7.1150740623474125,
      "epoch": 0.04490203193033382,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000247,
      "loss": 6.7083,
      "mean_token_accuracy": 0.10193079337477684,
      "num_tokens": 1024121.0,
      "step": 495
    },
    {
      "entropy": 7.323940992355347,
      "epoch": 0.045355587808417995,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002495,
      "loss": 6.9503,
      "mean_token_accuracy": 0.10487301424145698,
      "num_tokens": 1033845.0,
      "step": 500
    },
    {
      "entropy": 7.164136552810669,
      "epoch": 0.04580914368650218,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000252,
      "loss": 6.7388,
      "mean_token_accuracy": 0.10410365164279937,
      "num_tokens": 1044642.0,
      "step": 505
    },
    {
      "entropy": 7.172935676574707,
      "epoch": 0.046262699564586354,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002545,
      "loss": 6.8158,
      "mean_token_accuracy": 0.10116696879267692,
      "num_tokens": 1054728.0,
      "step": 510
    },
    {
      "entropy": 7.138863849639892,
      "epoch": 0.04671625544267054,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000257,
      "loss": 6.7643,
      "mean_token_accuracy": 0.10025905445218086,
      "num_tokens": 1065743.0,
      "step": 515
    },
    {
      "entropy": 7.10414080619812,
      "epoch": 0.04716981132075472,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002595,
      "loss": 6.6969,
      "mean_token_accuracy": 0.10323216244578362,
      "num_tokens": 1076798.0,
      "step": 520
    },
    {
      "entropy": 7.174085855484009,
      "epoch": 0.047623367198838897,
      "grad_norm": 1.03125,
      "learning_rate": 0.000262,
      "loss": 6.8163,
      "mean_token_accuracy": 0.0997597634792328,
      "num_tokens": 1087445.0,
      "step": 525
    },
    {
      "entropy": 7.094452857971191,
      "epoch": 0.04807692307692308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7197,
      "mean_token_accuracy": 0.10675485357642174,
      "num_tokens": 1098379.0,
      "step": 530
    },
    {
      "entropy": 7.202598142623901,
      "epoch": 0.048530478955007256,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.8289,
      "mean_token_accuracy": 0.0993705414235592,
      "num_tokens": 1108016.0,
      "step": 535
    },
    {
      "entropy": 7.089155960083008,
      "epoch": 0.04898403483309144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.701,
      "mean_token_accuracy": 0.10358337983489037,
      "num_tokens": 1117887.0,
      "step": 540
    },
    {
      "entropy": 7.104513120651245,
      "epoch": 0.049437590711175615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6948,
      "mean_token_accuracy": 0.09933766350150108,
      "num_tokens": 1128497.0,
      "step": 545
    },
    {
      "entropy": 7.055957365036011,
      "epoch": 0.0498911465892598,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002745,
      "loss": 6.7188,
      "mean_token_accuracy": 0.10837908834218979,
      "num_tokens": 1138168.0,
      "step": 550
    },
    {
      "entropy": 7.0923104763031,
      "epoch": 0.050344702467343974,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000277,
      "loss": 6.691,
      "mean_token_accuracy": 0.10448997020721436,
      "num_tokens": 1148357.0,
      "step": 555
    },
    {
      "entropy": 7.086644268035888,
      "epoch": 0.05079825834542816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002795,
      "loss": 6.7185,
      "mean_token_accuracy": 0.10321378186345101,
      "num_tokens": 1159964.0,
      "step": 560
    },
    {
      "entropy": 7.139619493484497,
      "epoch": 0.05125181422351233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.7781,
      "mean_token_accuracy": 0.10231135338544846,
      "num_tokens": 1170683.0,
      "step": 565
    },
    {
      "entropy": 7.115566444396973,
      "epoch": 0.051705370101596516,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002845,
      "loss": 6.7343,
      "mean_token_accuracy": 0.10421703606843949,
      "num_tokens": 1181680.0,
      "step": 570
    },
    {
      "entropy": 7.0173659324646,
      "epoch": 0.0521589259796807,
      "grad_norm": 1.078125,
      "learning_rate": 0.000287,
      "loss": 6.6227,
      "mean_token_accuracy": 0.10645965486764908,
      "num_tokens": 1191619.0,
      "step": 575
    },
    {
      "entropy": 6.977677965164185,
      "epoch": 0.052612481857764876,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002895,
      "loss": 6.6222,
      "mean_token_accuracy": 0.10420064777135848,
      "num_tokens": 1202264.0,
      "step": 580
    },
    {
      "entropy": 7.002698135375977,
      "epoch": 0.05306603773584906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000292,
      "loss": 6.6254,
      "mean_token_accuracy": 0.10796518102288247,
      "num_tokens": 1212119.0,
      "step": 585
    },
    {
      "entropy": 7.092545223236084,
      "epoch": 0.053519593613933235,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002945,
      "loss": 6.7588,
      "mean_token_accuracy": 0.10538023859262466,
      "num_tokens": 1222317.0,
      "step": 590
    },
    {
      "entropy": 7.060644435882568,
      "epoch": 0.05397314949201742,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000297,
      "loss": 6.703,
      "mean_token_accuracy": 0.1080887496471405,
      "num_tokens": 1232914.0,
      "step": 595
    },
    {
      "entropy": 6.919224500656128,
      "epoch": 0.054426705370101594,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002995,
      "loss": 6.5519,
      "mean_token_accuracy": 0.11217400431632996,
      "num_tokens": 1244214.0,
      "step": 600
    },
    {
      "entropy": 7.022872066497802,
      "epoch": 0.05488026124818578,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000302,
      "loss": 6.6491,
      "mean_token_accuracy": 0.10687957927584649,
      "num_tokens": 1254626.0,
      "step": 605
    },
    {
      "entropy": 6.948216915130615,
      "epoch": 0.05533381712626995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003045,
      "loss": 6.7394,
      "mean_token_accuracy": 0.10631354376673699,
      "num_tokens": 1265461.0,
      "step": 610
    },
    {
      "entropy": 7.030997228622437,
      "epoch": 0.055787373004354136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000307,
      "loss": 6.6204,
      "mean_token_accuracy": 0.10827788710594177,
      "num_tokens": 1275810.0,
      "step": 615
    },
    {
      "entropy": 6.895572900772095,
      "epoch": 0.05624092888243832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003095,
      "loss": 6.6276,
      "mean_token_accuracy": 0.11254601404070855,
      "num_tokens": 1285170.0,
      "step": 620
    },
    {
      "entropy": 7.06603364944458,
      "epoch": 0.056694484760522496,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000312,
      "loss": 6.5641,
      "mean_token_accuracy": 0.11045938208699227,
      "num_tokens": 1295460.0,
      "step": 625
    },
    {
      "entropy": 6.950644588470459,
      "epoch": 0.05714804063860668,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003145,
      "loss": 6.687,
      "mean_token_accuracy": 0.10515375509858131,
      "num_tokens": 1306656.0,
      "step": 630
    },
    {
      "entropy": 7.004617834091187,
      "epoch": 0.057601596516690855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000317,
      "loss": 6.6603,
      "mean_token_accuracy": 0.10941670313477517,
      "num_tokens": 1318019.0,
      "step": 635
    },
    {
      "entropy": 6.8719371318817135,
      "epoch": 0.05805515239477504,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003195,
      "loss": 6.5802,
      "mean_token_accuracy": 0.11324906200170518,
      "num_tokens": 1328120.0,
      "step": 640
    },
    {
      "entropy": 7.002581357955933,
      "epoch": 0.058508708272859214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000322,
      "loss": 6.6842,
      "mean_token_accuracy": 0.10679537355899811,
      "num_tokens": 1338354.0,
      "step": 645
    },
    {
      "entropy": 7.0219018936157225,
      "epoch": 0.0589622641509434,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6503,
      "mean_token_accuracy": 0.10833998396992683,
      "num_tokens": 1348666.0,
      "step": 650
    },
    {
      "entropy": 6.952841186523438,
      "epoch": 0.05941582002902757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.6921,
      "mean_token_accuracy": 0.11318251714110375,
      "num_tokens": 1359713.0,
      "step": 655
    },
    {
      "entropy": 6.950813102722168,
      "epoch": 0.059869375907111756,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5549,
      "mean_token_accuracy": 0.11367055401206017,
      "num_tokens": 1369507.0,
      "step": 660
    },
    {
      "entropy": 6.984795665740966,
      "epoch": 0.06032293178519594,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.7111,
      "mean_token_accuracy": 0.10242682620882988,
      "num_tokens": 1380726.0,
      "step": 665
    },
    {
      "entropy": 7.01419677734375,
      "epoch": 0.060776487663280115,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.6325,
      "mean_token_accuracy": 0.11175172179937362,
      "num_tokens": 1391423.0,
      "step": 670
    },
    {
      "entropy": 6.930062580108642,
      "epoch": 0.0612300435413643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000337,
      "loss": 6.6054,
      "mean_token_accuracy": 0.11288354322314262,
      "num_tokens": 1402492.0,
      "step": 675
    },
    {
      "entropy": 6.939368772506714,
      "epoch": 0.061683599419448475,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003395,
      "loss": 6.6504,
      "mean_token_accuracy": 0.1099183864891529,
      "num_tokens": 1412497.0,
      "step": 680
    },
    {
      "entropy": 6.999884557723999,
      "epoch": 0.06213715529753266,
      "grad_norm": 1.15625,
      "learning_rate": 0.000342,
      "loss": 6.6204,
      "mean_token_accuracy": 0.11455832123756408,
      "num_tokens": 1423214.0,
      "step": 685
    },
    {
      "entropy": 6.856761169433594,
      "epoch": 0.06259071117561683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.6202,
      "mean_token_accuracy": 0.11103508621454239,
      "num_tokens": 1434044.0,
      "step": 690
    },
    {
      "entropy": 6.9463762760162355,
      "epoch": 0.06304426705370102,
      "grad_norm": 1.015625,
      "learning_rate": 0.000347,
      "loss": 6.5826,
      "mean_token_accuracy": 0.11131571531295777,
      "num_tokens": 1444387.0,
      "step": 695
    },
    {
      "entropy": 6.848030471801758,
      "epoch": 0.0634978229317852,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003495,
      "loss": 6.5282,
      "mean_token_accuracy": 0.10915688425302505,
      "num_tokens": 1455157.0,
      "step": 700
    },
    {
      "entropy": 6.915933418273926,
      "epoch": 0.06395137880986937,
      "grad_norm": 1.0,
      "learning_rate": 0.000352,
      "loss": 6.6147,
      "mean_token_accuracy": 0.11529198288917542,
      "num_tokens": 1466145.0,
      "step": 705
    },
    {
      "entropy": 6.937824773788452,
      "epoch": 0.06440493468795355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003545,
      "loss": 6.6053,
      "mean_token_accuracy": 0.1114728607237339,
      "num_tokens": 1476713.0,
      "step": 710
    },
    {
      "entropy": 6.905196857452393,
      "epoch": 0.06485849056603774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000357,
      "loss": 6.6113,
      "mean_token_accuracy": 0.11422201618552208,
      "num_tokens": 1486908.0,
      "step": 715
    },
    {
      "entropy": 6.866687297821045,
      "epoch": 0.06531204644412192,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003595,
      "loss": 6.5389,
      "mean_token_accuracy": 0.11563119068741798,
      "num_tokens": 1498185.0,
      "step": 720
    },
    {
      "entropy": 6.873838233947754,
      "epoch": 0.0657656023222061,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000362,
      "loss": 6.5814,
      "mean_token_accuracy": 0.10893168300390244,
      "num_tokens": 1508376.0,
      "step": 725
    },
    {
      "entropy": 6.905275201797485,
      "epoch": 0.06621915820029027,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003645,
      "loss": 6.6695,
      "mean_token_accuracy": 0.11235518902540206,
      "num_tokens": 1519953.0,
      "step": 730
    },
    {
      "entropy": 6.795944786071777,
      "epoch": 0.06667271407837445,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000367,
      "loss": 6.5408,
      "mean_token_accuracy": 0.11527841538190842,
      "num_tokens": 1529793.0,
      "step": 735
    },
    {
      "entropy": 6.8445906162261965,
      "epoch": 0.06712626995645864,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003695,
      "loss": 6.5363,
      "mean_token_accuracy": 0.11552815213799476,
      "num_tokens": 1540859.0,
      "step": 740
    },
    {
      "entropy": 6.8253929138183596,
      "epoch": 0.06757982583454282,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000372,
      "loss": 6.4732,
      "mean_token_accuracy": 0.11656396687030793,
      "num_tokens": 1549783.0,
      "step": 745
    },
    {
      "entropy": 6.756074047088623,
      "epoch": 0.06803338171262699,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003745,
      "loss": 6.5266,
      "mean_token_accuracy": 0.11289834827184678,
      "num_tokens": 1560522.0,
      "step": 750
    },
    {
      "entropy": 6.810197973251343,
      "epoch": 0.06848693759071117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000377,
      "loss": 6.5008,
      "mean_token_accuracy": 0.11671004444360733,
      "num_tokens": 1570668.0,
      "step": 755
    },
    {
      "entropy": 6.83150315284729,
      "epoch": 0.06894049346879536,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003795,
      "loss": 6.6122,
      "mean_token_accuracy": 0.1193370096385479,
      "num_tokens": 1581523.0,
      "step": 760
    },
    {
      "entropy": 6.850863075256347,
      "epoch": 0.06939404934687954,
      "grad_norm": 1.125,
      "learning_rate": 0.000382,
      "loss": 6.5253,
      "mean_token_accuracy": 0.11707687377929688,
      "num_tokens": 1591356.0,
      "step": 765
    },
    {
      "entropy": 6.784787607192993,
      "epoch": 0.06984760522496372,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003845,
      "loss": 6.5113,
      "mean_token_accuracy": 0.11673457846045494,
      "num_tokens": 1601779.0,
      "step": 770
    },
    {
      "entropy": 6.848218965530395,
      "epoch": 0.07030116110304789,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.5091,
      "mean_token_accuracy": 0.11470222994685172,
      "num_tokens": 1611995.0,
      "step": 775
    },
    {
      "entropy": 6.800413322448731,
      "epoch": 0.07075471698113207,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.5369,
      "mean_token_accuracy": 0.11278748363256455,
      "num_tokens": 1621834.0,
      "step": 780
    },
    {
      "entropy": 6.780804920196533,
      "epoch": 0.07120827285921626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.5788,
      "mean_token_accuracy": 0.1113444909453392,
      "num_tokens": 1632881.0,
      "step": 785
    },
    {
      "entropy": 6.707678127288818,
      "epoch": 0.07166182873730044,
      "grad_norm": 1.0,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.5541,
      "mean_token_accuracy": 0.11073025166988373,
      "num_tokens": 1642870.0,
      "step": 790
    },
    {
      "entropy": 6.871548318862915,
      "epoch": 0.07211538461538461,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4402,
      "mean_token_accuracy": 0.12203826904296874,
      "num_tokens": 1653390.0,
      "step": 795
    },
    {
      "entropy": 6.7886035442352295,
      "epoch": 0.07256894049346879,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003995,
      "loss": 6.5612,
      "mean_token_accuracy": 0.11627909317612647,
      "num_tokens": 1663382.0,
      "step": 800
    },
    {
      "entropy": 6.804583787918091,
      "epoch": 0.07302249637155298,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000402,
      "loss": 6.5652,
      "mean_token_accuracy": 0.1136283002793789,
      "num_tokens": 1673601.0,
      "step": 805
    },
    {
      "entropy": 6.703377294540405,
      "epoch": 0.07347605224963716,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004045,
      "loss": 6.4891,
      "mean_token_accuracy": 0.12096939235925674,
      "num_tokens": 1683565.0,
      "step": 810
    },
    {
      "entropy": 6.758061265945434,
      "epoch": 0.07392960812772134,
      "grad_norm": 1.78125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.45,
      "mean_token_accuracy": 0.11798930242657661,
      "num_tokens": 1693636.0,
      "step": 815
    },
    {
      "entropy": 6.882167196273803,
      "epoch": 0.07438316400580551,
      "grad_norm": 1.125,
      "learning_rate": 0.0004095,
      "loss": 6.5623,
      "mean_token_accuracy": 0.11476152688264847,
      "num_tokens": 1703415.0,
      "step": 820
    },
    {
      "entropy": 6.755157995223999,
      "epoch": 0.0748367198838897,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000412,
      "loss": 6.5396,
      "mean_token_accuracy": 0.11611356288194656,
      "num_tokens": 1713579.0,
      "step": 825
    },
    {
      "entropy": 6.7464526176452635,
      "epoch": 0.07529027576197388,
      "grad_norm": 1.0,
      "learning_rate": 0.0004145,
      "loss": 6.5394,
      "mean_token_accuracy": 0.11386027708649635,
      "num_tokens": 1724583.0,
      "step": 830
    },
    {
      "entropy": 6.758458185195923,
      "epoch": 0.07574383164005806,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000417,
      "loss": 6.4527,
      "mean_token_accuracy": 0.12356605678796768,
      "num_tokens": 1735293.0,
      "step": 835
    },
    {
      "entropy": 6.7750757217407225,
      "epoch": 0.07619738751814223,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004195,
      "loss": 6.4786,
      "mean_token_accuracy": 0.12075236141681671,
      "num_tokens": 1745623.0,
      "step": 840
    },
    {
      "entropy": 6.725871658325195,
      "epoch": 0.07665094339622641,
      "grad_norm": 0.96875,
      "learning_rate": 0.000422,
      "loss": 6.5145,
      "mean_token_accuracy": 0.11627071723341942,
      "num_tokens": 1755713.0,
      "step": 845
    },
    {
      "entropy": 6.776280164718628,
      "epoch": 0.0771044992743106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004245,
      "loss": 6.506,
      "mean_token_accuracy": 0.11393775343894959,
      "num_tokens": 1765264.0,
      "step": 850
    },
    {
      "entropy": 6.824870538711548,
      "epoch": 0.07755805515239478,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000427,
      "loss": 6.511,
      "mean_token_accuracy": 0.11867988780140877,
      "num_tokens": 1775769.0,
      "step": 855
    },
    {
      "entropy": 6.691818761825561,
      "epoch": 0.07801161103047896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004295,
      "loss": 6.4441,
      "mean_token_accuracy": 0.11789206489920616,
      "num_tokens": 1786245.0,
      "step": 860
    },
    {
      "entropy": 6.627203035354614,
      "epoch": 0.07846516690856313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000432,
      "loss": 6.4773,
      "mean_token_accuracy": 0.1227649137377739,
      "num_tokens": 1796040.0,
      "step": 865
    },
    {
      "entropy": 6.691270112991333,
      "epoch": 0.07891872278664731,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004345,
      "loss": 6.4935,
      "mean_token_accuracy": 0.11697295233607292,
      "num_tokens": 1806007.0,
      "step": 870
    },
    {
      "entropy": 6.88080358505249,
      "epoch": 0.0793722786647315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000437,
      "loss": 6.5665,
      "mean_token_accuracy": 0.10995617434382439,
      "num_tokens": 1816935.0,
      "step": 875
    },
    {
      "entropy": 6.631897354125977,
      "epoch": 0.07982583454281568,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004395,
      "loss": 6.4244,
      "mean_token_accuracy": 0.11615645289421081,
      "num_tokens": 1827419.0,
      "step": 880
    },
    {
      "entropy": 6.766531658172608,
      "epoch": 0.08027939042089985,
      "grad_norm": 1.15625,
      "learning_rate": 0.000442,
      "loss": 6.533,
      "mean_token_accuracy": 0.11622850596904755,
      "num_tokens": 1837933.0,
      "step": 885
    },
    {
      "entropy": 6.714184427261353,
      "epoch": 0.08073294629898403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004445,
      "loss": 6.5854,
      "mean_token_accuracy": 0.11893809214234352,
      "num_tokens": 1848551.0,
      "step": 890
    },
    {
      "entropy": 6.7861744403839115,
      "epoch": 0.08118650217706821,
      "grad_norm": 1.125,
      "learning_rate": 0.000447,
      "loss": 6.4577,
      "mean_token_accuracy": 0.11503675356507301,
      "num_tokens": 1858380.0,
      "step": 895
    },
    {
      "entropy": 6.660554695129394,
      "epoch": 0.0816400580551524,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3853,
      "mean_token_accuracy": 0.11978982239961625,
      "num_tokens": 1868291.0,
      "step": 900
    },
    {
      "entropy": 6.629644966125488,
      "epoch": 0.08209361393323658,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.4839,
      "mean_token_accuracy": 0.11984607204794884,
      "num_tokens": 1879219.0,
      "step": 905
    },
    {
      "entropy": 6.754321718215943,
      "epoch": 0.08254716981132075,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4652,
      "mean_token_accuracy": 0.1155819833278656,
      "num_tokens": 1889809.0,
      "step": 910
    },
    {
      "entropy": 6.659244441986084,
      "epoch": 0.08300072568940493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4349,
      "mean_token_accuracy": 0.11736063361167907,
      "num_tokens": 1899986.0,
      "step": 915
    },
    {
      "entropy": 6.727784872055054,
      "epoch": 0.08345428156748912,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4192,
      "mean_token_accuracy": 0.11866439804434777,
      "num_tokens": 1909314.0,
      "step": 920
    },
    {
      "entropy": 6.709451198577881,
      "epoch": 0.0839078374455733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000462,
      "loss": 6.4667,
      "mean_token_accuracy": 0.1235547974705696,
      "num_tokens": 1918443.0,
      "step": 925
    },
    {
      "entropy": 6.680801582336426,
      "epoch": 0.08436139332365747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004645,
      "loss": 6.4235,
      "mean_token_accuracy": 0.12087266743183137,
      "num_tokens": 1928683.0,
      "step": 930
    },
    {
      "entropy": 6.541482257843017,
      "epoch": 0.08481494920174165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000467,
      "loss": 6.3312,
      "mean_token_accuracy": 0.11764199584722519,
      "num_tokens": 1938293.0,
      "step": 935
    },
    {
      "entropy": 6.629144906997681,
      "epoch": 0.08526850507982583,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004695,
      "loss": 6.4449,
      "mean_token_accuracy": 0.11702702939510345,
      "num_tokens": 1948233.0,
      "step": 940
    },
    {
      "entropy": 6.660422420501709,
      "epoch": 0.08572206095791002,
      "grad_norm": 1.046875,
      "learning_rate": 0.000472,
      "loss": 6.3444,
      "mean_token_accuracy": 0.12124838382005691,
      "num_tokens": 1958611.0,
      "step": 945
    },
    {
      "entropy": 6.661191082000732,
      "epoch": 0.0861756168359942,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004745,
      "loss": 6.4983,
      "mean_token_accuracy": 0.1220276914536953,
      "num_tokens": 1968957.0,
      "step": 950
    },
    {
      "entropy": 6.65334038734436,
      "epoch": 0.08662917271407837,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477,
      "loss": 6.3551,
      "mean_token_accuracy": 0.12157930657267571,
      "num_tokens": 1978503.0,
      "step": 955
    },
    {
      "entropy": 6.552236032485962,
      "epoch": 0.08708272859216255,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004795,
      "loss": 6.3482,
      "mean_token_accuracy": 0.12029609382152558,
      "num_tokens": 1988458.0,
      "step": 960
    },
    {
      "entropy": 6.655647468566895,
      "epoch": 0.08753628447024674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000482,
      "loss": 6.491,
      "mean_token_accuracy": 0.11715313494205475,
      "num_tokens": 1999984.0,
      "step": 965
    },
    {
      "entropy": 6.510547399520874,
      "epoch": 0.08798984034833092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004845,
      "loss": 6.3364,
      "mean_token_accuracy": 0.12651536986231804,
      "num_tokens": 2010645.0,
      "step": 970
    },
    {
      "entropy": 6.751053190231323,
      "epoch": 0.08844339622641509,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000487,
      "loss": 6.4506,
      "mean_token_accuracy": 0.11614105999469757,
      "num_tokens": 2020196.0,
      "step": 975
    },
    {
      "entropy": 6.587527418136597,
      "epoch": 0.08889695210449927,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004895,
      "loss": 6.3423,
      "mean_token_accuracy": 0.1243398666381836,
      "num_tokens": 2030101.0,
      "step": 980
    },
    {
      "entropy": 6.5954345703125,
      "epoch": 0.08935050798258345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492,
      "loss": 6.4263,
      "mean_token_accuracy": 0.1206196092069149,
      "num_tokens": 2040454.0,
      "step": 985
    },
    {
      "entropy": 6.529821062088013,
      "epoch": 0.08980406386066764,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945,
      "loss": 6.3602,
      "mean_token_accuracy": 0.11969397887587548,
      "num_tokens": 2050378.0,
      "step": 990
    },
    {
      "entropy": 6.656999588012695,
      "epoch": 0.09025761973875182,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497,
      "loss": 6.4186,
      "mean_token_accuracy": 0.11806954741477967,
      "num_tokens": 2060282.0,
      "step": 995
    },
    {
      "entropy": 6.559673261642456,
      "epoch": 0.09071117561683599,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995,
      "loss": 6.4115,
      "mean_token_accuracy": 0.12180240452289581,
      "num_tokens": 2070563.0,
      "step": 1000
    },
    {
      "entropy": 6.661390542984009,
      "epoch": 0.09116473149492017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999985110217,
      "loss": 6.4743,
      "mean_token_accuracy": 0.12212796434760094,
      "num_tokens": 2080153.0,
      "step": 1005
    },
    {
      "entropy": 6.525332069396972,
      "epoch": 0.09161828737300436,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999999924620471,
      "loss": 6.3263,
      "mean_token_accuracy": 0.12526472061872482,
      "num_tokens": 2090590.0,
      "step": 1010
    },
    {
      "entropy": 6.613137102127075,
      "epoch": 0.09207184325108854,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999999817600155,
      "loss": 6.4047,
      "mean_token_accuracy": 0.11499982699751854,
      "num_tokens": 2101945.0,
      "step": 1015
    },
    {
      "entropy": 6.613913679122925,
      "epoch": 0.09252539912917271,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999999664049268,
      "loss": 6.4031,
      "mean_token_accuracy": 0.12405626773834229,
      "num_tokens": 2112514.0,
      "step": 1020
    },
    {
      "entropy": 6.498817586898804,
      "epoch": 0.09297895500725689,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999999463967816,
      "loss": 6.3365,
      "mean_token_accuracy": 0.1212070882320404,
      "num_tokens": 2122296.0,
      "step": 1025
    },
    {
      "entropy": 6.740924310684204,
      "epoch": 0.09343251088534107,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049999992173558,
      "loss": 6.4899,
      "mean_token_accuracy": 0.11208800077438355,
      "num_tokens": 2134512.0,
      "step": 1030
    },
    {
      "entropy": 6.578291606903076,
      "epoch": 0.09388606676342526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999998924213228,
      "loss": 6.4201,
      "mean_token_accuracy": 0.12014863640069962,
      "num_tokens": 2143934.0,
      "step": 1035
    },
    {
      "entropy": 6.489139556884766,
      "epoch": 0.09433962264150944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999998584540105,
      "loss": 6.3024,
      "mean_token_accuracy": 0.12084058821201324,
      "num_tokens": 2155730.0,
      "step": 1040
    },
    {
      "entropy": 6.6251708984375,
      "epoch": 0.09479317851959361,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999998198336437,
      "loss": 6.3417,
      "mean_token_accuracy": 0.1186050683259964,
      "num_tokens": 2166422.0,
      "step": 1045
    },
    {
      "entropy": 6.549373722076416,
      "epoch": 0.09524673439767779,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999997765602233,
      "loss": 6.4281,
      "mean_token_accuracy": 0.11772651746869087,
      "num_tokens": 2177842.0,
      "step": 1050
    },
    {
      "entropy": 6.614485788345337,
      "epoch": 0.09570029027576198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999997286337502,
      "loss": 6.4994,
      "mean_token_accuracy": 0.12015130147337913,
      "num_tokens": 2188614.0,
      "step": 1055
    },
    {
      "entropy": 6.645506238937378,
      "epoch": 0.09615384615384616,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999996760542254,
      "loss": 6.3042,
      "mean_token_accuracy": 0.12434962838888168,
      "num_tokens": 2197886.0,
      "step": 1060
    },
    {
      "entropy": 6.450875282287598,
      "epoch": 0.09660740203193033,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049999961882165,
      "loss": 6.3515,
      "mean_token_accuracy": 0.12345030084252358,
      "num_tokens": 2209930.0,
      "step": 1065
    },
    {
      "entropy": 6.616359043121338,
      "epoch": 0.09706095791001451,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999995569360251,
      "loss": 6.3704,
      "mean_token_accuracy": 0.11785119250416756,
      "num_tokens": 2219767.0,
      "step": 1070
    },
    {
      "entropy": 6.475496864318847,
      "epoch": 0.0975145137880987,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499999490397352,
      "loss": 6.2989,
      "mean_token_accuracy": 0.1218901552259922,
      "num_tokens": 2230082.0,
      "step": 1075
    },
    {
      "entropy": 6.539343738555909,
      "epoch": 0.09796806966618288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999994192056321,
      "loss": 6.3308,
      "mean_token_accuracy": 0.1269037939608097,
      "num_tokens": 2240066.0,
      "step": 1080
    },
    {
      "entropy": 6.536786890029907,
      "epoch": 0.09842162554426705,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999993433608669,
      "loss": 6.3427,
      "mean_token_accuracy": 0.12274787575006485,
      "num_tokens": 2250732.0,
      "step": 1085
    },
    {
      "entropy": 6.519191551208496,
      "epoch": 0.09887518142235123,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999992628630579,
      "loss": 6.2367,
      "mean_token_accuracy": 0.12317501232028008,
      "num_tokens": 2261452.0,
      "step": 1090
    },
    {
      "entropy": 6.529998350143432,
      "epoch": 0.09932873730043541,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999991777122069,
      "loss": 6.3514,
      "mean_token_accuracy": 0.12049490809440613,
      "num_tokens": 2271929.0,
      "step": 1095
    },
    {
      "entropy": 6.48491473197937,
      "epoch": 0.0997822931785196,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999990879083156,
      "loss": 6.3314,
      "mean_token_accuracy": 0.12258056998252868,
      "num_tokens": 2282987.0,
      "step": 1100
    },
    {
      "entropy": 6.453391408920288,
      "epoch": 0.10023584905660378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999989934513857,
      "loss": 6.3044,
      "mean_token_accuracy": 0.12670305147767066,
      "num_tokens": 2292796.0,
      "step": 1105
    },
    {
      "entropy": 6.591043186187744,
      "epoch": 0.10068940493468795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999988943414193,
      "loss": 6.3516,
      "mean_token_accuracy": 0.12217535674571992,
      "num_tokens": 2303057.0,
      "step": 1110
    },
    {
      "entropy": 6.426148223876953,
      "epoch": 0.10114296081277213,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999987905784184,
      "loss": 6.334,
      "mean_token_accuracy": 0.12347318157553673,
      "num_tokens": 2314168.0,
      "step": 1115
    },
    {
      "entropy": 6.5130480289459225,
      "epoch": 0.10159651669085631,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999986821623853,
      "loss": 6.2937,
      "mean_token_accuracy": 0.12410474196076393,
      "num_tokens": 2325241.0,
      "step": 1120
    },
    {
      "entropy": 6.5921142578125,
      "epoch": 0.1020500725689405,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999985690933219,
      "loss": 6.3255,
      "mean_token_accuracy": 0.12369295954704285,
      "num_tokens": 2334866.0,
      "step": 1125
    },
    {
      "entropy": 6.381993150711059,
      "epoch": 0.10250362844702467,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999984513712311,
      "loss": 6.3279,
      "mean_token_accuracy": 0.12018040791153908,
      "num_tokens": 2345422.0,
      "step": 1130
    },
    {
      "entropy": 6.5917377948760985,
      "epoch": 0.10295718432510885,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999983289961146,
      "loss": 6.3473,
      "mean_token_accuracy": 0.12131484672427177,
      "num_tokens": 2356285.0,
      "step": 1135
    },
    {
      "entropy": 6.436669492721558,
      "epoch": 0.10341074020319303,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999982019679755,
      "loss": 6.2615,
      "mean_token_accuracy": 0.12863264828920365,
      "num_tokens": 2366861.0,
      "step": 1140
    },
    {
      "entropy": 6.494740104675293,
      "epoch": 0.10386429608127722,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999980702868164,
      "loss": 6.2841,
      "mean_token_accuracy": 0.12933464497327804,
      "num_tokens": 2376851.0,
      "step": 1145
    },
    {
      "entropy": 6.544817733764648,
      "epoch": 0.1043178519593614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999979339526396,
      "loss": 6.3588,
      "mean_token_accuracy": 0.13027909770607948,
      "num_tokens": 2386857.0,
      "step": 1150
    },
    {
      "entropy": 6.5093776226043705,
      "epoch": 0.10477140783744557,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999977929654484,
      "loss": 6.2838,
      "mean_token_accuracy": 0.12735775336623192,
      "num_tokens": 2397035.0,
      "step": 1155
    },
    {
      "entropy": 6.470782279968262,
      "epoch": 0.10522496371552975,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999976473252453,
      "loss": 6.3559,
      "mean_token_accuracy": 0.12582893818616867,
      "num_tokens": 2408177.0,
      "step": 1160
    },
    {
      "entropy": 6.572890186309815,
      "epoch": 0.10567851959361393,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999974970320338,
      "loss": 6.3173,
      "mean_token_accuracy": 0.12205304279923439,
      "num_tokens": 2418445.0,
      "step": 1165
    },
    {
      "entropy": 6.472396755218506,
      "epoch": 0.10613207547169812,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999973420858165,
      "loss": 6.3095,
      "mean_token_accuracy": 0.1242362380027771,
      "num_tokens": 2428990.0,
      "step": 1170
    },
    {
      "entropy": 6.47665491104126,
      "epoch": 0.10658563134978229,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999971824865968,
      "loss": 6.3435,
      "mean_token_accuracy": 0.11929563954472541,
      "num_tokens": 2439643.0,
      "step": 1175
    },
    {
      "entropy": 6.585198831558228,
      "epoch": 0.10703918722786647,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999970182343782,
      "loss": 6.3812,
      "mean_token_accuracy": 0.12422708719968796,
      "num_tokens": 2450691.0,
      "step": 1180
    },
    {
      "entropy": 6.567696046829224,
      "epoch": 0.10749274310595065,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999968493291638,
      "loss": 6.3221,
      "mean_token_accuracy": 0.12464427575469017,
      "num_tokens": 2460706.0,
      "step": 1185
    },
    {
      "entropy": 6.306331729888916,
      "epoch": 0.10794629898403484,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999966757709573,
      "loss": 6.2099,
      "mean_token_accuracy": 0.1348014645278454,
      "num_tokens": 2470947.0,
      "step": 1190
    },
    {
      "entropy": 6.511581468582153,
      "epoch": 0.10839985486211902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999964975597622,
      "loss": 6.1938,
      "mean_token_accuracy": 0.1275109112262726,
      "num_tokens": 2481132.0,
      "step": 1195
    },
    {
      "entropy": 6.452719688415527,
      "epoch": 0.10885341074020319,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999963146955821,
      "loss": 6.2243,
      "mean_token_accuracy": 0.12854906544089317,
      "num_tokens": 2491701.0,
      "step": 1200
    },
    {
      "entropy": 6.320997333526611,
      "epoch": 0.10930696661828737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499996127178421,
      "loss": 6.2922,
      "mean_token_accuracy": 0.12324495241045952,
      "num_tokens": 2502360.0,
      "step": 1205
    },
    {
      "entropy": 6.524869346618653,
      "epoch": 0.10976052249637155,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999959350082825,
      "loss": 6.1951,
      "mean_token_accuracy": 0.13170049339532852,
      "num_tokens": 2512140.0,
      "step": 1210
    },
    {
      "entropy": 6.4425348281860355,
      "epoch": 0.11021407837445574,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999957381851709,
      "loss": 6.3032,
      "mean_token_accuracy": 0.12556114420294762,
      "num_tokens": 2522721.0,
      "step": 1215
    },
    {
      "entropy": 6.40459713935852,
      "epoch": 0.1106676342525399,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999955367090901,
      "loss": 6.1739,
      "mean_token_accuracy": 0.12824239879846572,
      "num_tokens": 2531786.0,
      "step": 1220
    },
    {
      "entropy": 6.501996803283691,
      "epoch": 0.11112119013062409,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999953305800441,
      "loss": 6.2788,
      "mean_token_accuracy": 0.12112325578927993,
      "num_tokens": 2542895.0,
      "step": 1225
    },
    {
      "entropy": 6.3701423645019535,
      "epoch": 0.11157474600870827,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999951197980374,
      "loss": 6.2802,
      "mean_token_accuracy": 0.12341128513216973,
      "num_tokens": 2554120.0,
      "step": 1230
    },
    {
      "entropy": 6.539844799041748,
      "epoch": 0.11202830188679246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999949043630744,
      "loss": 6.313,
      "mean_token_accuracy": 0.12485537752509117,
      "num_tokens": 2563935.0,
      "step": 1235
    },
    {
      "entropy": 6.355460834503174,
      "epoch": 0.11248185776487664,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999946842751593,
      "loss": 6.2219,
      "mean_token_accuracy": 0.12797416001558304,
      "num_tokens": 2574465.0,
      "step": 1240
    },
    {
      "entropy": 6.420122289657593,
      "epoch": 0.11293541364296081,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999944595342968,
      "loss": 6.1786,
      "mean_token_accuracy": 0.13132491633296012,
      "num_tokens": 2584694.0,
      "step": 1245
    },
    {
      "entropy": 6.382951498031616,
      "epoch": 0.11338896952104499,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999942301404916,
      "loss": 6.1569,
      "mean_token_accuracy": 0.1340431585907936,
      "num_tokens": 2594540.0,
      "step": 1250
    },
    {
      "entropy": 6.32035346031189,
      "epoch": 0.11384252539912917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999939960937484,
      "loss": 6.0947,
      "mean_token_accuracy": 0.13378870859742165,
      "num_tokens": 2604896.0,
      "step": 1255
    },
    {
      "entropy": 6.4604442596435545,
      "epoch": 0.11429608127721336,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999937573940721,
      "loss": 6.2717,
      "mean_token_accuracy": 0.13821497261524202,
      "num_tokens": 2614143.0,
      "step": 1260
    },
    {
      "entropy": 6.316441440582276,
      "epoch": 0.11474963715529753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999935140414674,
      "loss": 6.1643,
      "mean_token_accuracy": 0.12574966847896576,
      "num_tokens": 2623999.0,
      "step": 1265
    },
    {
      "entropy": 6.501933288574219,
      "epoch": 0.11520319303338171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999932660359395,
      "loss": 6.3783,
      "mean_token_accuracy": 0.12551918104290963,
      "num_tokens": 2636208.0,
      "step": 1270
    },
    {
      "entropy": 6.424142551422119,
      "epoch": 0.11565674891146589,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999930133774936,
      "loss": 6.2436,
      "mean_token_accuracy": 0.13297036662697792,
      "num_tokens": 2645854.0,
      "step": 1275
    },
    {
      "entropy": 6.461622619628907,
      "epoch": 0.11611030478955008,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999927560661348,
      "loss": 6.2337,
      "mean_token_accuracy": 0.13188307881355285,
      "num_tokens": 2655877.0,
      "step": 1280
    },
    {
      "entropy": 6.463164520263672,
      "epoch": 0.11656386066763426,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999924941018685,
      "loss": 6.2551,
      "mean_token_accuracy": 0.13034509047865867,
      "num_tokens": 2666157.0,
      "step": 1285
    },
    {
      "entropy": 6.322438955307007,
      "epoch": 0.11701741654571843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999922274847,
      "loss": 6.1759,
      "mean_token_accuracy": 0.12801435366272926,
      "num_tokens": 2676647.0,
      "step": 1290
    },
    {
      "entropy": 6.35759539604187,
      "epoch": 0.11747097242380261,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499991956214635,
      "loss": 6.2134,
      "mean_token_accuracy": 0.13719891905784606,
      "num_tokens": 2686417.0,
      "step": 1295
    },
    {
      "entropy": 6.4111240863800045,
      "epoch": 0.1179245283018868,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499991680291679,
      "loss": 6.2127,
      "mean_token_accuracy": 0.12441546320915223,
      "num_tokens": 2697540.0,
      "step": 1300
    },
    {
      "entropy": 6.4009637355804445,
      "epoch": 0.11837808417997098,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999913997158378,
      "loss": 6.2549,
      "mean_token_accuracy": 0.11752010956406593,
      "num_tokens": 2709397.0,
      "step": 1305
    },
    {
      "entropy": 6.34979658126831,
      "epoch": 0.11883164005805515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999911144871169,
      "loss": 6.2137,
      "mean_token_accuracy": 0.1235676109790802,
      "num_tokens": 2720430.0,
      "step": 1310
    },
    {
      "entropy": 6.358980894088745,
      "epoch": 0.11928519593613933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999908246055226,
      "loss": 6.1748,
      "mean_token_accuracy": 0.12498210072517395,
      "num_tokens": 2730648.0,
      "step": 1315
    },
    {
      "entropy": 6.358032989501953,
      "epoch": 0.11973875181422351,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999905300710607,
      "loss": 6.1528,
      "mean_token_accuracy": 0.12824709638953208,
      "num_tokens": 2739979.0,
      "step": 1320
    },
    {
      "entropy": 6.410165405273437,
      "epoch": 0.1201923076923077,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999902308837373,
      "loss": 6.2777,
      "mean_token_accuracy": 0.1253784976899624,
      "num_tokens": 2750329.0,
      "step": 1325
    },
    {
      "entropy": 6.337530565261841,
      "epoch": 0.12064586357039188,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999899270435584,
      "loss": 6.1525,
      "mean_token_accuracy": 0.13555875271558762,
      "num_tokens": 2760535.0,
      "step": 1330
    },
    {
      "entropy": 6.363765001296997,
      "epoch": 0.12109941944847605,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999896185505307,
      "loss": 6.1902,
      "mean_token_accuracy": 0.12614675164222716,
      "num_tokens": 2770470.0,
      "step": 1335
    },
    {
      "entropy": 6.35250883102417,
      "epoch": 0.12155297532656023,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999893054046603,
      "loss": 6.2192,
      "mean_token_accuracy": 0.12356962338089943,
      "num_tokens": 2781506.0,
      "step": 1340
    },
    {
      "entropy": 6.300785732269287,
      "epoch": 0.12200653120464441,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999889876059537,
      "loss": 6.0809,
      "mean_token_accuracy": 0.13543135821819305,
      "num_tokens": 2791245.0,
      "step": 1345
    },
    {
      "entropy": 6.336817073822021,
      "epoch": 0.1224600870827286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999886651544176,
      "loss": 6.0946,
      "mean_token_accuracy": 0.13693180531263352,
      "num_tokens": 2801680.0,
      "step": 1350
    },
    {
      "entropy": 6.373439359664917,
      "epoch": 0.12291364296081277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999883380500584,
      "loss": 6.1846,
      "mean_token_accuracy": 0.1356772594153881,
      "num_tokens": 2812805.0,
      "step": 1355
    },
    {
      "entropy": 6.366350793838501,
      "epoch": 0.12336719883889695,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999880062928831,
      "loss": 6.2141,
      "mean_token_accuracy": 0.13203392773866654,
      "num_tokens": 2822834.0,
      "step": 1360
    },
    {
      "entropy": 6.361548852920532,
      "epoch": 0.12382075471698113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999876698828985,
      "loss": 6.1554,
      "mean_token_accuracy": 0.1361112043261528,
      "num_tokens": 2832312.0,
      "step": 1365
    },
    {
      "entropy": 6.335978317260742,
      "epoch": 0.12427431059506532,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999873288201116,
      "loss": 6.1665,
      "mean_token_accuracy": 0.13540629148483277,
      "num_tokens": 2842998.0,
      "step": 1370
    },
    {
      "entropy": 6.394313335418701,
      "epoch": 0.1247278664731495,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999869831045294,
      "loss": 6.1961,
      "mean_token_accuracy": 0.13257036805152894,
      "num_tokens": 2854218.0,
      "step": 1375
    },
    {
      "entropy": 6.33300895690918,
      "epoch": 0.12518142235123367,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999866327361589,
      "loss": 6.0892,
      "mean_token_accuracy": 0.13703012093901634,
      "num_tokens": 2864484.0,
      "step": 1380
    },
    {
      "entropy": 6.370788526535034,
      "epoch": 0.12563497822931785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999862777150077,
      "loss": 6.2028,
      "mean_token_accuracy": 0.13148538768291473,
      "num_tokens": 2873800.0,
      "step": 1385
    },
    {
      "entropy": 6.390060663223267,
      "epoch": 0.12608853410740203,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999859180410829,
      "loss": 6.1965,
      "mean_token_accuracy": 0.13013821244239807,
      "num_tokens": 2884247.0,
      "step": 1390
    },
    {
      "entropy": 6.341064691543579,
      "epoch": 0.12654208998548622,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999855537143919,
      "loss": 6.1096,
      "mean_token_accuracy": 0.13113176226615905,
      "num_tokens": 2894739.0,
      "step": 1395
    },
    {
      "entropy": 6.234110736846924,
      "epoch": 0.1269956458635704,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999851847349425,
      "loss": 6.1616,
      "mean_token_accuracy": 0.13198944479227065,
      "num_tokens": 2904376.0,
      "step": 1400
    },
    {
      "entropy": 6.4315849304199215,
      "epoch": 0.12744920174165458,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999848111027419,
      "loss": 6.2265,
      "mean_token_accuracy": 0.12843010425567628,
      "num_tokens": 2914304.0,
      "step": 1405
    },
    {
      "entropy": 6.319644594192505,
      "epoch": 0.12790275761973874,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999844328177983,
      "loss": 6.0959,
      "mean_token_accuracy": 0.13285723477602004,
      "num_tokens": 2924400.0,
      "step": 1410
    },
    {
      "entropy": 6.456520843505859,
      "epoch": 0.12835631349782292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999840498801191,
      "loss": 6.3401,
      "mean_token_accuracy": 0.1338380053639412,
      "num_tokens": 2935040.0,
      "step": 1415
    },
    {
      "entropy": 6.279810905456543,
      "epoch": 0.1288098693759071,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999836622897126,
      "loss": 6.1154,
      "mean_token_accuracy": 0.1344787932932377,
      "num_tokens": 2945536.0,
      "step": 1420
    },
    {
      "entropy": 6.234977674484253,
      "epoch": 0.1292634252539913,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999832700465865,
      "loss": 6.0487,
      "mean_token_accuracy": 0.13450392782688142,
      "num_tokens": 2954969.0,
      "step": 1425
    },
    {
      "entropy": 6.315559005737304,
      "epoch": 0.12971698113207547,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999828731507491,
      "loss": 6.1625,
      "mean_token_accuracy": 0.13059850856661798,
      "num_tokens": 2966218.0,
      "step": 1430
    },
    {
      "entropy": 6.379030179977417,
      "epoch": 0.13017053701015965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999824716022085,
      "loss": 6.1894,
      "mean_token_accuracy": 0.127619219571352,
      "num_tokens": 2976640.0,
      "step": 1435
    },
    {
      "entropy": 6.313711643218994,
      "epoch": 0.13062409288824384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999820654009731,
      "loss": 6.1533,
      "mean_token_accuracy": 0.1332421600818634,
      "num_tokens": 2985883.0,
      "step": 1440
    },
    {
      "entropy": 6.321325731277466,
      "epoch": 0.13107764876632802,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999816545470513,
      "loss": 6.1541,
      "mean_token_accuracy": 0.12692199870944024,
      "num_tokens": 2996695.0,
      "step": 1445
    },
    {
      "entropy": 6.334502124786377,
      "epoch": 0.1315312046444122,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999812390404514,
      "loss": 6.16,
      "mean_token_accuracy": 0.1343579739332199,
      "num_tokens": 3007500.0,
      "step": 1450
    },
    {
      "entropy": 6.184367370605469,
      "epoch": 0.13198476052249636,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999808188811823,
      "loss": 6.0132,
      "mean_token_accuracy": 0.13415692150592803,
      "num_tokens": 3017299.0,
      "step": 1455
    },
    {
      "entropy": 6.403480911254883,
      "epoch": 0.13243831640058054,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999803940692524,
      "loss": 6.3011,
      "mean_token_accuracy": 0.12249834388494492,
      "num_tokens": 3028389.0,
      "step": 1460
    },
    {
      "entropy": 6.317180013656616,
      "epoch": 0.13289187227866472,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999799646046707,
      "loss": 6.1135,
      "mean_token_accuracy": 0.1368947669863701,
      "num_tokens": 3038883.0,
      "step": 1465
    },
    {
      "entropy": 6.403723573684692,
      "epoch": 0.1333454281567489,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999795304874461,
      "loss": 6.2419,
      "mean_token_accuracy": 0.13426837772130967,
      "num_tokens": 3049904.0,
      "step": 1470
    },
    {
      "entropy": 6.2098509788513185,
      "epoch": 0.1337989840348331,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999790917175873,
      "loss": 6.0755,
      "mean_token_accuracy": 0.13349713161587715,
      "num_tokens": 3059788.0,
      "step": 1475
    },
    {
      "entropy": 6.351387977600098,
      "epoch": 0.13425253991291727,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999786482951036,
      "loss": 6.0716,
      "mean_token_accuracy": 0.1320554032921791,
      "num_tokens": 3070413.0,
      "step": 1480
    },
    {
      "entropy": 6.270417833328247,
      "epoch": 0.13470609579100146,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999782002200041,
      "loss": 6.1959,
      "mean_token_accuracy": 0.12884705662727355,
      "num_tokens": 3081093.0,
      "step": 1485
    },
    {
      "entropy": 6.342434740066528,
      "epoch": 0.13515965166908564,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999777474922982,
      "loss": 6.129,
      "mean_token_accuracy": 0.13011036217212676,
      "num_tokens": 3090594.0,
      "step": 1490
    },
    {
      "entropy": 6.322659873962403,
      "epoch": 0.13561320754716982,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999772901119951,
      "loss": 6.2158,
      "mean_token_accuracy": 0.13068215921521187,
      "num_tokens": 3101061.0,
      "step": 1495
    },
    {
      "entropy": 6.294349813461304,
      "epoch": 0.13606676342525398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999768280791043,
      "loss": 6.1383,
      "mean_token_accuracy": 0.13425096422433852,
      "num_tokens": 3111715.0,
      "step": 1500
    },
    {
      "entropy": 6.270824289321899,
      "epoch": 0.13652031930333816,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999763613936352,
      "loss": 6.0633,
      "mean_token_accuracy": 0.13947287648916246,
      "num_tokens": 3121378.0,
      "step": 1505
    },
    {
      "entropy": 6.246769809722901,
      "epoch": 0.13697387518142234,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999758900555979,
      "loss": 5.9928,
      "mean_token_accuracy": 0.1401647448539734,
      "num_tokens": 3131229.0,
      "step": 1510
    },
    {
      "entropy": 6.208933019638062,
      "epoch": 0.13742743105950653,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999754140650017,
      "loss": 6.0878,
      "mean_token_accuracy": 0.13210750818252565,
      "num_tokens": 3140859.0,
      "step": 1515
    },
    {
      "entropy": 6.351792240142823,
      "epoch": 0.1378809869375907,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999749334218567,
      "loss": 6.0861,
      "mean_token_accuracy": 0.13336024880409242,
      "num_tokens": 3151692.0,
      "step": 1520
    },
    {
      "entropy": 6.221499061584472,
      "epoch": 0.1383345428156749,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999744481261726,
      "loss": 6.1524,
      "mean_token_accuracy": 0.13277293592691422,
      "num_tokens": 3161879.0,
      "step": 1525
    },
    {
      "entropy": 6.278037405014038,
      "epoch": 0.13878809869375908,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999739581779597,
      "loss": 6.0267,
      "mean_token_accuracy": 0.13772518262267114,
      "num_tokens": 3171282.0,
      "step": 1530
    },
    {
      "entropy": 6.244936561584472,
      "epoch": 0.13924165457184326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499973463577228,
      "loss": 6.0779,
      "mean_token_accuracy": 0.13918796479701995,
      "num_tokens": 3181398.0,
      "step": 1535
    },
    {
      "entropy": 6.294873476028442,
      "epoch": 0.13969521044992744,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999729643239877,
      "loss": 6.1991,
      "mean_token_accuracy": 0.13689608871936798,
      "num_tokens": 3191431.0,
      "step": 1540
    },
    {
      "entropy": 6.297893095016479,
      "epoch": 0.1401487663280116,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999724604182492,
      "loss": 6.1937,
      "mean_token_accuracy": 0.1275126539170742,
      "num_tokens": 3202147.0,
      "step": 1545
    },
    {
      "entropy": 6.362968301773071,
      "epoch": 0.14060232220609578,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999719518600229,
      "loss": 6.073,
      "mean_token_accuracy": 0.13779982998967172,
      "num_tokens": 3212428.0,
      "step": 1550
    },
    {
      "entropy": 6.222197484970093,
      "epoch": 0.14105587808417996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999714386493192,
      "loss": 6.1203,
      "mean_token_accuracy": 0.13455342650413513,
      "num_tokens": 3222099.0,
      "step": 1555
    },
    {
      "entropy": 6.321711540222168,
      "epoch": 0.14150943396226415,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999709207861489,
      "loss": 6.0726,
      "mean_token_accuracy": 0.13806343004107474,
      "num_tokens": 3232079.0,
      "step": 1560
    },
    {
      "entropy": 6.153911828994751,
      "epoch": 0.14196298984034833,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999703982705227,
      "loss": 5.9721,
      "mean_token_accuracy": 0.14627765491604805,
      "num_tokens": 3241762.0,
      "step": 1565
    },
    {
      "entropy": 6.244314336776734,
      "epoch": 0.1424165457184325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999698711024513,
      "loss": 6.0669,
      "mean_token_accuracy": 0.1394437588751316,
      "num_tokens": 3252571.0,
      "step": 1570
    },
    {
      "entropy": 6.215821599960327,
      "epoch": 0.1428701015965167,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999693392819456,
      "loss": 6.1098,
      "mean_token_accuracy": 0.13461477681994438,
      "num_tokens": 3263542.0,
      "step": 1575
    },
    {
      "entropy": 6.340892887115478,
      "epoch": 0.14332365747460088,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999688028090166,
      "loss": 6.1811,
      "mean_token_accuracy": 0.13378583863377572,
      "num_tokens": 3273295.0,
      "step": 1580
    },
    {
      "entropy": 6.3619109153747555,
      "epoch": 0.14377721335268506,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999682616836755,
      "loss": 6.1952,
      "mean_token_accuracy": 0.136061330139637,
      "num_tokens": 3283605.0,
      "step": 1585
    },
    {
      "entropy": 6.243741178512574,
      "epoch": 0.14423076923076922,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999677159059334,
      "loss": 6.0415,
      "mean_token_accuracy": 0.14292106181383132,
      "num_tokens": 3293837.0,
      "step": 1590
    },
    {
      "entropy": 6.130685424804687,
      "epoch": 0.1446843251088534,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999671654758016,
      "loss": 6.0417,
      "mean_token_accuracy": 0.14042141437530517,
      "num_tokens": 3304131.0,
      "step": 1595
    },
    {
      "entropy": 6.302355670928955,
      "epoch": 0.14513788098693758,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999666103932915,
      "loss": 6.0746,
      "mean_token_accuracy": 0.13077147156000138,
      "num_tokens": 3313543.0,
      "step": 1600
    },
    {
      "entropy": 6.222482824325562,
      "epoch": 0.14559143686502177,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999660506584145,
      "loss": 6.0608,
      "mean_token_accuracy": 0.1355305939912796,
      "num_tokens": 3324487.0,
      "step": 1605
    },
    {
      "entropy": 6.281767272949219,
      "epoch": 0.14604499274310595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999654862711823,
      "loss": 6.0873,
      "mean_token_accuracy": 0.1388099044561386,
      "num_tokens": 3335094.0,
      "step": 1610
    },
    {
      "entropy": 6.182447385787964,
      "epoch": 0.14649854862119013,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999649172316065,
      "loss": 6.0205,
      "mean_token_accuracy": 0.13585399314761162,
      "num_tokens": 3344252.0,
      "step": 1615
    },
    {
      "entropy": 6.219208812713623,
      "epoch": 0.14695210449927432,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499964343539699,
      "loss": 6.0905,
      "mean_token_accuracy": 0.13364078626036643,
      "num_tokens": 3354265.0,
      "step": 1620
    },
    {
      "entropy": 6.3048234462738035,
      "epoch": 0.1474056603773585,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999637651954714,
      "loss": 6.182,
      "mean_token_accuracy": 0.13599163964390754,
      "num_tokens": 3364616.0,
      "step": 1625
    },
    {
      "entropy": 6.33318829536438,
      "epoch": 0.14785921625544268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999631821989358,
      "loss": 6.1253,
      "mean_token_accuracy": 0.14277874156832696,
      "num_tokens": 3373650.0,
      "step": 1630
    },
    {
      "entropy": 6.114496469497681,
      "epoch": 0.14831277213352684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999625945501043,
      "loss": 6.0331,
      "mean_token_accuracy": 0.13961085602641105,
      "num_tokens": 3384590.0,
      "step": 1635
    },
    {
      "entropy": 6.34743185043335,
      "epoch": 0.14876632801161102,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499962002248989,
      "loss": 6.0701,
      "mean_token_accuracy": 0.13647577166557312,
      "num_tokens": 3395616.0,
      "step": 1640
    },
    {
      "entropy": 6.165446662902832,
      "epoch": 0.1492198838896952,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999614052956024,
      "loss": 6.0688,
      "mean_token_accuracy": 0.1349581092596054,
      "num_tokens": 3405810.0,
      "step": 1645
    },
    {
      "entropy": 6.321514320373535,
      "epoch": 0.1496734397677794,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999608036899563,
      "loss": 6.1055,
      "mean_token_accuracy": 0.1329289957880974,
      "num_tokens": 3415932.0,
      "step": 1650
    },
    {
      "entropy": 6.197605752944947,
      "epoch": 0.15012699564586357,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999601974320636,
      "loss": 6.0631,
      "mean_token_accuracy": 0.13647494986653327,
      "num_tokens": 3425626.0,
      "step": 1655
    },
    {
      "entropy": 6.155983877182007,
      "epoch": 0.15058055152394775,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999595865219367,
      "loss": 6.0609,
      "mean_token_accuracy": 0.1348869502544403,
      "num_tokens": 3434837.0,
      "step": 1660
    },
    {
      "entropy": 6.258736085891724,
      "epoch": 0.15103410740203194,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999589709595881,
      "loss": 6.0701,
      "mean_token_accuracy": 0.14022651612758635,
      "num_tokens": 3445238.0,
      "step": 1665
    },
    {
      "entropy": 6.191891717910766,
      "epoch": 0.15148766328011612,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999583507450308,
      "loss": 6.0521,
      "mean_token_accuracy": 0.13351622000336646,
      "num_tokens": 3456738.0,
      "step": 1670
    },
    {
      "entropy": 6.110352945327759,
      "epoch": 0.1519412191582003,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999577258782774,
      "loss": 5.9378,
      "mean_token_accuracy": 0.14009969010949136,
      "num_tokens": 3466296.0,
      "step": 1675
    },
    {
      "entropy": 6.319712066650391,
      "epoch": 0.15239477503628446,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499957096359341,
      "loss": 6.1345,
      "mean_token_accuracy": 0.13245827108621597,
      "num_tokens": 3476989.0,
      "step": 1680
    },
    {
      "entropy": 6.156819915771484,
      "epoch": 0.15284833091436864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999564621882343,
      "loss": 5.97,
      "mean_token_accuracy": 0.1411023445427418,
      "num_tokens": 3487050.0,
      "step": 1685
    },
    {
      "entropy": 6.2730926990509035,
      "epoch": 0.15330188679245282,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999558233649708,
      "loss": 6.188,
      "mean_token_accuracy": 0.12993267178535461,
      "num_tokens": 3497777.0,
      "step": 1690
    },
    {
      "entropy": 6.238672304153442,
      "epoch": 0.153755442670537,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999551798895636,
      "loss": 6.0905,
      "mean_token_accuracy": 0.13315979093313218,
      "num_tokens": 3509066.0,
      "step": 1695
    },
    {
      "entropy": 6.142438983917236,
      "epoch": 0.1542089985486212,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999545317620258,
      "loss": 6.026,
      "mean_token_accuracy": 0.13728305473923683,
      "num_tokens": 3519567.0,
      "step": 1700
    },
    {
      "entropy": 6.209616088867188,
      "epoch": 0.15466255442670537,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999538789823711,
      "loss": 6.0481,
      "mean_token_accuracy": 0.13899025991559028,
      "num_tokens": 3529402.0,
      "step": 1705
    },
    {
      "entropy": 6.3146570205688475,
      "epoch": 0.15511611030478956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999532215506127,
      "loss": 6.0703,
      "mean_token_accuracy": 0.1440271407365799,
      "num_tokens": 3539805.0,
      "step": 1710
    },
    {
      "entropy": 6.143733882904053,
      "epoch": 0.15556966618287374,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999525594667645,
      "loss": 5.9303,
      "mean_token_accuracy": 0.14397806972265242,
      "num_tokens": 3550477.0,
      "step": 1715
    },
    {
      "entropy": 6.110523653030396,
      "epoch": 0.15602322206095792,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049995189273084,
      "loss": 6.0405,
      "mean_token_accuracy": 0.139139074832201,
      "num_tokens": 3560729.0,
      "step": 1720
    },
    {
      "entropy": 6.220071601867676,
      "epoch": 0.15647677793904208,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499951221342853,
      "loss": 6.0973,
      "mean_token_accuracy": 0.13195881322026254,
      "num_tokens": 3571538.0,
      "step": 1725
    },
    {
      "entropy": 6.214411115646362,
      "epoch": 0.15693033381712626,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999505453028174,
      "loss": 6.1078,
      "mean_token_accuracy": 0.13914671614766122,
      "num_tokens": 3583128.0,
      "step": 1730
    },
    {
      "entropy": 6.246854066848755,
      "epoch": 0.15738388969521044,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999498646107472,
      "loss": 6.0354,
      "mean_token_accuracy": 0.14036523252725602,
      "num_tokens": 3594358.0,
      "step": 1735
    },
    {
      "entropy": 6.198376893997192,
      "epoch": 0.15783744557329463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999491792666566,
      "loss": 5.959,
      "mean_token_accuracy": 0.14713839516043664,
      "num_tokens": 3604345.0,
      "step": 1740
    },
    {
      "entropy": 6.148045206069947,
      "epoch": 0.1582910014513788,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999484892705597,
      "loss": 6.0217,
      "mean_token_accuracy": 0.13767840191721917,
      "num_tokens": 3616129.0,
      "step": 1745
    },
    {
      "entropy": 6.18104510307312,
      "epoch": 0.158744557329463,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999477946224705,
      "loss": 6.0602,
      "mean_token_accuracy": 0.13979990035295486,
      "num_tokens": 3627458.0,
      "step": 1750
    },
    {
      "entropy": 6.226989650726319,
      "epoch": 0.15919811320754718,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999470953224037,
      "loss": 5.9442,
      "mean_token_accuracy": 0.14392665922641754,
      "num_tokens": 3636047.0,
      "step": 1755
    },
    {
      "entropy": 6.051787710189819,
      "epoch": 0.15965166908563136,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999463913703735,
      "loss": 5.9397,
      "mean_token_accuracy": 0.1413258582353592,
      "num_tokens": 3646930.0,
      "step": 1760
    },
    {
      "entropy": 6.218035459518433,
      "epoch": 0.16010522496371554,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999456827663948,
      "loss": 6.0338,
      "mean_token_accuracy": 0.14086730852723123,
      "num_tokens": 3657279.0,
      "step": 1765
    },
    {
      "entropy": 6.123547697067261,
      "epoch": 0.1605587808417997,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499944969510482,
      "loss": 5.9464,
      "mean_token_accuracy": 0.1423609510064125,
      "num_tokens": 3667070.0,
      "step": 1770
    },
    {
      "entropy": 6.166107845306397,
      "epoch": 0.16101233671988388,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999442516026498,
      "loss": 5.9613,
      "mean_token_accuracy": 0.1381991073489189,
      "num_tokens": 3677560.0,
      "step": 1775
    },
    {
      "entropy": 6.090453720092773,
      "epoch": 0.16146589259796806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999435290429133,
      "loss": 5.9268,
      "mean_token_accuracy": 0.1474891871213913,
      "num_tokens": 3686701.0,
      "step": 1780
    },
    {
      "entropy": 6.152267122268677,
      "epoch": 0.16191944847605225,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999428018312872,
      "loss": 6.0228,
      "mean_token_accuracy": 0.14551690965890884,
      "num_tokens": 3696338.0,
      "step": 1785
    },
    {
      "entropy": 6.150699281692505,
      "epoch": 0.16237300435413643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999420699677867,
      "loss": 5.9993,
      "mean_token_accuracy": 0.14381766989827155,
      "num_tokens": 3705752.0,
      "step": 1790
    },
    {
      "entropy": 6.229558277130127,
      "epoch": 0.1628265602322206,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999413334524269,
      "loss": 6.084,
      "mean_token_accuracy": 0.13718133196234703,
      "num_tokens": 3716225.0,
      "step": 1795
    },
    {
      "entropy": 6.1720422267913815,
      "epoch": 0.1632801161103048,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999405922852229,
      "loss": 5.9581,
      "mean_token_accuracy": 0.14193314164876938,
      "num_tokens": 3726593.0,
      "step": 1800
    },
    {
      "entropy": 6.101333570480347,
      "epoch": 0.16373367198838898,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999398464661901,
      "loss": 6.0091,
      "mean_token_accuracy": 0.13523678332567216,
      "num_tokens": 3737563.0,
      "step": 1805
    },
    {
      "entropy": 6.307489013671875,
      "epoch": 0.16418722786647316,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999390959953441,
      "loss": 6.0708,
      "mean_token_accuracy": 0.1370485432446003,
      "num_tokens": 3747752.0,
      "step": 1810
    },
    {
      "entropy": 6.057960081100464,
      "epoch": 0.16464078374455732,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999383408727002,
      "loss": 6.0019,
      "mean_token_accuracy": 0.139044688642025,
      "num_tokens": 3758795.0,
      "step": 1815
    },
    {
      "entropy": 6.2216832637786865,
      "epoch": 0.1650943396226415,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999375810982741,
      "loss": 5.9714,
      "mean_token_accuracy": 0.14233563393354415,
      "num_tokens": 3768690.0,
      "step": 1820
    },
    {
      "entropy": 6.090987873077393,
      "epoch": 0.16554789550072568,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999368166720815,
      "loss": 6.0494,
      "mean_token_accuracy": 0.13934776037931443,
      "num_tokens": 3779018.0,
      "step": 1825
    },
    {
      "entropy": 6.219239854812622,
      "epoch": 0.16600145137880987,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999360475941382,
      "loss": 5.9469,
      "mean_token_accuracy": 0.14526350796222687,
      "num_tokens": 3788985.0,
      "step": 1830
    },
    {
      "entropy": 6.173433303833008,
      "epoch": 0.16645500725689405,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049993527386446,
      "loss": 6.0821,
      "mean_token_accuracy": 0.13157572448253632,
      "num_tokens": 3800787.0,
      "step": 1835
    },
    {
      "entropy": 6.16374773979187,
      "epoch": 0.16690856313497823,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999344954830633,
      "loss": 5.9647,
      "mean_token_accuracy": 0.13687085136771202,
      "num_tokens": 3811381.0,
      "step": 1840
    },
    {
      "entropy": 6.157071971893311,
      "epoch": 0.16736211901306242,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999337124499638,
      "loss": 6.0685,
      "mean_token_accuracy": 0.14529063701629638,
      "num_tokens": 3822055.0,
      "step": 1845
    },
    {
      "entropy": 6.190087890625,
      "epoch": 0.1678156748911466,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999329247651777,
      "loss": 6.0321,
      "mean_token_accuracy": 0.1372029162943363,
      "num_tokens": 3831900.0,
      "step": 1850
    },
    {
      "entropy": 6.127886390686035,
      "epoch": 0.16826923076923078,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999321324287213,
      "loss": 5.9863,
      "mean_token_accuracy": 0.14112438708543779,
      "num_tokens": 3842702.0,
      "step": 1855
    },
    {
      "entropy": 6.237508106231689,
      "epoch": 0.16872278664731494,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999313354406114,
      "loss": 5.9599,
      "mean_token_accuracy": 0.14055541008710862,
      "num_tokens": 3852690.0,
      "step": 1860
    },
    {
      "entropy": 6.039614915847778,
      "epoch": 0.16917634252539912,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499930533800864,
      "loss": 5.9466,
      "mean_token_accuracy": 0.1505817547440529,
      "num_tokens": 3863040.0,
      "step": 1865
    },
    {
      "entropy": 6.180584526062011,
      "epoch": 0.1696298984034833,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999297275094958,
      "loss": 6.0401,
      "mean_token_accuracy": 0.13932662308216096,
      "num_tokens": 3873613.0,
      "step": 1870
    },
    {
      "entropy": 6.042843770980835,
      "epoch": 0.1700834542815675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999289165665235,
      "loss": 5.8485,
      "mean_token_accuracy": 0.1480574920773506,
      "num_tokens": 3883936.0,
      "step": 1875
    },
    {
      "entropy": 6.060999536514283,
      "epoch": 0.17053701015965167,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999281009719639,
      "loss": 5.8154,
      "mean_token_accuracy": 0.14890880212187768,
      "num_tokens": 3893788.0,
      "step": 1880
    },
    {
      "entropy": 6.056200647354126,
      "epoch": 0.17099056603773585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999272807258339,
      "loss": 5.8885,
      "mean_token_accuracy": 0.14825719892978667,
      "num_tokens": 3903434.0,
      "step": 1885
    },
    {
      "entropy": 6.1081467628479,
      "epoch": 0.17144412191582004,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999264558281504,
      "loss": 5.895,
      "mean_token_accuracy": 0.1487080954015255,
      "num_tokens": 3913476.0,
      "step": 1890
    },
    {
      "entropy": 5.991112613677979,
      "epoch": 0.17189767779390422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999256262789305,
      "loss": 5.8483,
      "mean_token_accuracy": 0.1485322192311287,
      "num_tokens": 3923301.0,
      "step": 1895
    },
    {
      "entropy": 6.099672555923462,
      "epoch": 0.1723512336719884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999247920781913,
      "loss": 6.0042,
      "mean_token_accuracy": 0.13689797297120093,
      "num_tokens": 3933335.0,
      "step": 1900
    },
    {
      "entropy": 6.197177267074585,
      "epoch": 0.17280478955007256,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999239532259502,
      "loss": 6.0293,
      "mean_token_accuracy": 0.13568304777145385,
      "num_tokens": 3944171.0,
      "step": 1905
    },
    {
      "entropy": 6.177489757537842,
      "epoch": 0.17325834542815674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999231097222244,
      "loss": 6.0829,
      "mean_token_accuracy": 0.14412558674812317,
      "num_tokens": 3954160.0,
      "step": 1910
    },
    {
      "entropy": 6.171565389633178,
      "epoch": 0.17371190130624092,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999222615670312,
      "loss": 5.8844,
      "mean_token_accuracy": 0.14805472120642663,
      "num_tokens": 3964110.0,
      "step": 1915
    },
    {
      "entropy": 6.134349346160889,
      "epoch": 0.1741654571843251,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999214087603885,
      "loss": 5.9675,
      "mean_token_accuracy": 0.1452759712934494,
      "num_tokens": 3974293.0,
      "step": 1920
    },
    {
      "entropy": 6.078616380691528,
      "epoch": 0.1746190130624093,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999205513023135,
      "loss": 5.9293,
      "mean_token_accuracy": 0.15142350345849992,
      "num_tokens": 3984346.0,
      "step": 1925
    },
    {
      "entropy": 6.053181409835815,
      "epoch": 0.17507256894049347,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999196891928245,
      "loss": 5.866,
      "mean_token_accuracy": 0.15365748405456542,
      "num_tokens": 3992897.0,
      "step": 1930
    },
    {
      "entropy": 6.05509238243103,
      "epoch": 0.17552612481857766,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999188224319388,
      "loss": 5.897,
      "mean_token_accuracy": 0.15057138204574586,
      "num_tokens": 4002575.0,
      "step": 1935
    },
    {
      "entropy": 6.070607757568359,
      "epoch": 0.17597968069666184,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999179510196746,
      "loss": 5.9925,
      "mean_token_accuracy": 0.14099667817354203,
      "num_tokens": 4013752.0,
      "step": 1940
    },
    {
      "entropy": 6.107510566711426,
      "epoch": 0.17643323657474602,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999170749560498,
      "loss": 5.8739,
      "mean_token_accuracy": 0.14898938685655594,
      "num_tokens": 4023718.0,
      "step": 1945
    },
    {
      "entropy": 6.064271402359009,
      "epoch": 0.17688679245283018,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999161942410827,
      "loss": 5.9743,
      "mean_token_accuracy": 0.1369916170835495,
      "num_tokens": 4034210.0,
      "step": 1950
    },
    {
      "entropy": 6.2005760192871096,
      "epoch": 0.17734034833091436,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999153088747913,
      "loss": 5.9622,
      "mean_token_accuracy": 0.1424011543393135,
      "num_tokens": 4044868.0,
      "step": 1955
    },
    {
      "entropy": 5.97287278175354,
      "epoch": 0.17779390420899854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499914418857194,
      "loss": 5.9603,
      "mean_token_accuracy": 0.14433022439479828,
      "num_tokens": 4055049.0,
      "step": 1960
    },
    {
      "entropy": 6.208456897735596,
      "epoch": 0.17824746008708273,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999135241883092,
      "loss": 6.0087,
      "mean_token_accuracy": 0.14180623665452002,
      "num_tokens": 4065889.0,
      "step": 1965
    },
    {
      "entropy": 6.090062475204467,
      "epoch": 0.1787010159651669,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999126248681555,
      "loss": 6.0282,
      "mean_token_accuracy": 0.1386457771062851,
      "num_tokens": 4077288.0,
      "step": 1970
    },
    {
      "entropy": 6.084208917617798,
      "epoch": 0.1791545718432511,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999117208967513,
      "loss": 5.9802,
      "mean_token_accuracy": 0.1426486611366272,
      "num_tokens": 4087402.0,
      "step": 1975
    },
    {
      "entropy": 6.152999305725098,
      "epoch": 0.17960812772133528,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999108122741154,
      "loss": 6.027,
      "mean_token_accuracy": 0.13648915737867356,
      "num_tokens": 4098150.0,
      "step": 1980
    },
    {
      "entropy": 6.015424156188965,
      "epoch": 0.18006168359941946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999098990002666,
      "loss": 5.8866,
      "mean_token_accuracy": 0.1459066838026047,
      "num_tokens": 4108134.0,
      "step": 1985
    },
    {
      "entropy": 6.190234994888305,
      "epoch": 0.18051523947750364,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999089810752237,
      "loss": 6.0277,
      "mean_token_accuracy": 0.13577395901083947,
      "num_tokens": 4118251.0,
      "step": 1990
    },
    {
      "entropy": 6.0451069355010985,
      "epoch": 0.1809687953555878,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499908058499006,
      "loss": 5.8609,
      "mean_token_accuracy": 0.1469724267721176,
      "num_tokens": 4129228.0,
      "step": 1995
    },
    {
      "entropy": 6.0451616764068605,
      "epoch": 0.18142235123367198,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999071312716321,
      "loss": 5.9319,
      "mean_token_accuracy": 0.15002663657069207,
      "num_tokens": 4139391.0,
      "step": 2000
    },
    {
      "entropy": 6.0800574779510494,
      "epoch": 0.18187590711175616,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999061993931215,
      "loss": 5.8526,
      "mean_token_accuracy": 0.1524651437997818,
      "num_tokens": 4148891.0,
      "step": 2005
    },
    {
      "entropy": 6.2835156440734865,
      "epoch": 0.18232946298984035,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999052628634934,
      "loss": 6.2064,
      "mean_token_accuracy": 0.138189584761858,
      "num_tokens": 4159551.0,
      "step": 2010
    },
    {
      "entropy": 6.138812065124512,
      "epoch": 0.18278301886792453,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999043216827671,
      "loss": 6.0055,
      "mean_token_accuracy": 0.14381420388817787,
      "num_tokens": 4170055.0,
      "step": 2015
    },
    {
      "entropy": 6.033186531066894,
      "epoch": 0.1832365747460087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999033758509621,
      "loss": 5.8591,
      "mean_token_accuracy": 0.14449694454669954,
      "num_tokens": 4180767.0,
      "step": 2020
    },
    {
      "entropy": 6.07762713432312,
      "epoch": 0.1836901306240929,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499902425368098,
      "loss": 5.9878,
      "mean_token_accuracy": 0.14416737630963325,
      "num_tokens": 4190654.0,
      "step": 2025
    },
    {
      "entropy": 6.103671360015869,
      "epoch": 0.18414368650217708,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999014702341944,
      "loss": 5.9002,
      "mean_token_accuracy": 0.1454018011689186,
      "num_tokens": 4200753.0,
      "step": 2030
    },
    {
      "entropy": 6.093059635162353,
      "epoch": 0.18459724238026126,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999005104492712,
      "loss": 5.9798,
      "mean_token_accuracy": 0.14127395525574685,
      "num_tokens": 4210994.0,
      "step": 2035
    },
    {
      "entropy": 6.107717561721802,
      "epoch": 0.18505079825834542,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998995460133481,
      "loss": 5.9256,
      "mean_token_accuracy": 0.1447222650051117,
      "num_tokens": 4220883.0,
      "step": 2040
    },
    {
      "entropy": 6.123315095901489,
      "epoch": 0.1855043541364296,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998985769264451,
      "loss": 5.9518,
      "mean_token_accuracy": 0.13728086575865744,
      "num_tokens": 4231269.0,
      "step": 2045
    },
    {
      "entropy": 6.056053113937378,
      "epoch": 0.18595791001451378,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998976031885821,
      "loss": 5.9715,
      "mean_token_accuracy": 0.14049455970525743,
      "num_tokens": 4241559.0,
      "step": 2050
    },
    {
      "entropy": 6.045488786697388,
      "epoch": 0.18641146589259797,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998966247997794,
      "loss": 5.8466,
      "mean_token_accuracy": 0.14596481025218963,
      "num_tokens": 4250870.0,
      "step": 2055
    },
    {
      "entropy": 6.027891731262207,
      "epoch": 0.18686502177068215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998956417600572,
      "loss": 5.8125,
      "mean_token_accuracy": 0.1532222166657448,
      "num_tokens": 4260848.0,
      "step": 2060
    },
    {
      "entropy": 6.0480344772338865,
      "epoch": 0.18731857764876633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998946540694358,
      "loss": 5.9379,
      "mean_token_accuracy": 0.14905711114406586,
      "num_tokens": 4270973.0,
      "step": 2065
    },
    {
      "entropy": 6.159943771362305,
      "epoch": 0.18777213352685052,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998936617279357,
      "loss": 6.0104,
      "mean_token_accuracy": 0.14108132272958757,
      "num_tokens": 4281211.0,
      "step": 2070
    },
    {
      "entropy": 6.0061546802520756,
      "epoch": 0.1882256894049347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998926647355773,
      "loss": 5.8728,
      "mean_token_accuracy": 0.14745167195796965,
      "num_tokens": 4291425.0,
      "step": 2075
    },
    {
      "entropy": 6.123989725112915,
      "epoch": 0.18867924528301888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998916630923813,
      "loss": 5.9712,
      "mean_token_accuracy": 0.1410026654601097,
      "num_tokens": 4301079.0,
      "step": 2080
    },
    {
      "entropy": 6.104276657104492,
      "epoch": 0.18913280116110304,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998906567983685,
      "loss": 5.9629,
      "mean_token_accuracy": 0.14320143908262253,
      "num_tokens": 4312374.0,
      "step": 2085
    },
    {
      "entropy": 6.043295049667359,
      "epoch": 0.18958635703918722,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998896458535595,
      "loss": 5.8714,
      "mean_token_accuracy": 0.15539209097623824,
      "num_tokens": 4322810.0,
      "step": 2090
    },
    {
      "entropy": 6.002667427062988,
      "epoch": 0.1900399129172714,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998886302579752,
      "loss": 5.8651,
      "mean_token_accuracy": 0.14834146052598954,
      "num_tokens": 4332082.0,
      "step": 2095
    },
    {
      "entropy": 6.0972264289855955,
      "epoch": 0.19049346879535559,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499887610011637,
      "loss": 5.9636,
      "mean_token_accuracy": 0.14356174021959306,
      "num_tokens": 4342632.0,
      "step": 2100
    },
    {
      "entropy": 6.016445589065552,
      "epoch": 0.19094702467343977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998865851145656,
      "loss": 5.8886,
      "mean_token_accuracy": 0.1432491272687912,
      "num_tokens": 4353544.0,
      "step": 2105
    },
    {
      "entropy": 6.031062507629395,
      "epoch": 0.19140058055152395,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998855555667822,
      "loss": 6.02,
      "mean_token_accuracy": 0.14226981848478318,
      "num_tokens": 4363636.0,
      "step": 2110
    },
    {
      "entropy": 6.133910322189331,
      "epoch": 0.19185413642960814,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998845213683084,
      "loss": 5.8763,
      "mean_token_accuracy": 0.1426288031041622,
      "num_tokens": 4373672.0,
      "step": 2115
    },
    {
      "entropy": 5.938132858276367,
      "epoch": 0.19230769230769232,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998834825191652,
      "loss": 5.8381,
      "mean_token_accuracy": 0.14870256185531616,
      "num_tokens": 4383199.0,
      "step": 2120
    },
    {
      "entropy": 6.071506214141846,
      "epoch": 0.1927612481857765,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998824390193743,
      "loss": 5.9109,
      "mean_token_accuracy": 0.14334884583950042,
      "num_tokens": 4394446.0,
      "step": 2125
    },
    {
      "entropy": 6.126976585388183,
      "epoch": 0.19321480406386066,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998813908689573,
      "loss": 5.9646,
      "mean_token_accuracy": 0.1414870209991932,
      "num_tokens": 4405882.0,
      "step": 2130
    },
    {
      "entropy": 5.956241512298584,
      "epoch": 0.19366835994194484,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998803380679359,
      "loss": 5.8182,
      "mean_token_accuracy": 0.15368922352790831,
      "num_tokens": 4415280.0,
      "step": 2135
    },
    {
      "entropy": 6.030566740036011,
      "epoch": 0.19412191582002902,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998792806163317,
      "loss": 6.0265,
      "mean_token_accuracy": 0.13928496092557907,
      "num_tokens": 4425755.0,
      "step": 2140
    },
    {
      "entropy": 6.020366859436035,
      "epoch": 0.1945754716981132,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998782185141667,
      "loss": 5.8808,
      "mean_token_accuracy": 0.14801339209079742,
      "num_tokens": 4436259.0,
      "step": 2145
    },
    {
      "entropy": 6.056359338760376,
      "epoch": 0.1950290275761974,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998771517614629,
      "loss": 5.917,
      "mean_token_accuracy": 0.14185069352388383,
      "num_tokens": 4446423.0,
      "step": 2150
    },
    {
      "entropy": 6.050050067901611,
      "epoch": 0.19548258345428157,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998760803582422,
      "loss": 5.871,
      "mean_token_accuracy": 0.15231530964374543,
      "num_tokens": 4456678.0,
      "step": 2155
    },
    {
      "entropy": 6.006132411956787,
      "epoch": 0.19593613933236576,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998750043045269,
      "loss": 5.8574,
      "mean_token_accuracy": 0.15113169997930526,
      "num_tokens": 4467031.0,
      "step": 2160
    },
    {
      "entropy": 6.0697108745574955,
      "epoch": 0.19638969521044994,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998739236003392,
      "loss": 5.938,
      "mean_token_accuracy": 0.13495565131306647,
      "num_tokens": 4477771.0,
      "step": 2165
    },
    {
      "entropy": 6.000240993499756,
      "epoch": 0.1968432510885341,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998728382457014,
      "loss": 5.9664,
      "mean_token_accuracy": 0.14246328622102739,
      "num_tokens": 4488426.0,
      "step": 2170
    },
    {
      "entropy": 6.17234845161438,
      "epoch": 0.19729680696661828,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998717482406362,
      "loss": 5.9372,
      "mean_token_accuracy": 0.13979932889342309,
      "num_tokens": 4499735.0,
      "step": 2175
    },
    {
      "entropy": 6.042126989364624,
      "epoch": 0.19775036284470246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998706535851657,
      "loss": 5.8795,
      "mean_token_accuracy": 0.14715081751346587,
      "num_tokens": 4509500.0,
      "step": 2180
    },
    {
      "entropy": 6.014921045303344,
      "epoch": 0.19820391872278664,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998695542793129,
      "loss": 5.8802,
      "mean_token_accuracy": 0.14634579196572303,
      "num_tokens": 4520431.0,
      "step": 2185
    },
    {
      "entropy": 6.034388875961303,
      "epoch": 0.19865747460087083,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998684503231003,
      "loss": 5.9784,
      "mean_token_accuracy": 0.14680968075990677,
      "num_tokens": 4532294.0,
      "step": 2190
    },
    {
      "entropy": 6.020555639266968,
      "epoch": 0.199111030478955,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499867341716551,
      "loss": 5.8983,
      "mean_token_accuracy": 0.14305620416998863,
      "num_tokens": 4542488.0,
      "step": 2195
    },
    {
      "entropy": 6.039434766769409,
      "epoch": 0.1995645863570392,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998662284596878,
      "loss": 5.8402,
      "mean_token_accuracy": 0.13846864104270934,
      "num_tokens": 4552933.0,
      "step": 2200
    },
    {
      "entropy": 6.0450187683105465,
      "epoch": 0.20001814223512338,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998651105525335,
      "loss": 5.8223,
      "mean_token_accuracy": 0.15138240456581115,
      "num_tokens": 4562373.0,
      "step": 2205
    },
    {
      "entropy": 5.943200492858887,
      "epoch": 0.20047169811320756,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998639879951115,
      "loss": 5.9439,
      "mean_token_accuracy": 0.14522400572896005,
      "num_tokens": 4572715.0,
      "step": 2210
    },
    {
      "entropy": 6.120235109329224,
      "epoch": 0.2009252539912917,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499862860787445,
      "loss": 5.8307,
      "mean_token_accuracy": 0.14360565692186356,
      "num_tokens": 4582782.0,
      "step": 2215
    },
    {
      "entropy": 6.041302442550659,
      "epoch": 0.2013788098693759,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998617289295571,
      "loss": 5.9352,
      "mean_token_accuracy": 0.13825485333800316,
      "num_tokens": 4594108.0,
      "step": 2220
    },
    {
      "entropy": 6.074282169342041,
      "epoch": 0.20183236574746008,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998605924214714,
      "loss": 5.8911,
      "mean_token_accuracy": 0.14501106441020967,
      "num_tokens": 4604732.0,
      "step": 2225
    },
    {
      "entropy": 6.039686012268066,
      "epoch": 0.20228592162554426,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998594512632114,
      "loss": 6.0324,
      "mean_token_accuracy": 0.13512664213776587,
      "num_tokens": 4614998.0,
      "step": 2230
    },
    {
      "entropy": 6.029949378967285,
      "epoch": 0.20273947750362845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998583054548006,
      "loss": 5.8536,
      "mean_token_accuracy": 0.14373639225959778,
      "num_tokens": 4624665.0,
      "step": 2235
    },
    {
      "entropy": 6.060518264770508,
      "epoch": 0.20319303338171263,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998571549962628,
      "loss": 5.8374,
      "mean_token_accuracy": 0.15260313004255294,
      "num_tokens": 4633731.0,
      "step": 2240
    },
    {
      "entropy": 6.0299561500549315,
      "epoch": 0.2036465892597968,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998559998876216,
      "loss": 5.8706,
      "mean_token_accuracy": 0.1454593777656555,
      "num_tokens": 4643421.0,
      "step": 2245
    },
    {
      "entropy": 6.054778814315796,
      "epoch": 0.204100145137881,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998548401289012,
      "loss": 5.9553,
      "mean_token_accuracy": 0.1441769301891327,
      "num_tokens": 4653795.0,
      "step": 2250
    },
    {
      "entropy": 6.058734989166259,
      "epoch": 0.20455370101596518,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998536757201253,
      "loss": 5.9276,
      "mean_token_accuracy": 0.1450178414583206,
      "num_tokens": 4664840.0,
      "step": 2255
    },
    {
      "entropy": 6.073224878311157,
      "epoch": 0.20500725689404933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998525066613181,
      "loss": 5.931,
      "mean_token_accuracy": 0.14578341245651244,
      "num_tokens": 4675508.0,
      "step": 2260
    },
    {
      "entropy": 5.989795446395874,
      "epoch": 0.20546081277213352,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998513329525038,
      "loss": 5.8523,
      "mean_token_accuracy": 0.14992504566907883,
      "num_tokens": 4685995.0,
      "step": 2265
    },
    {
      "entropy": 6.086735725402832,
      "epoch": 0.2059143686502177,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998501545937066,
      "loss": 5.8789,
      "mean_token_accuracy": 0.14490637332201003,
      "num_tokens": 4696659.0,
      "step": 2270
    },
    {
      "entropy": 5.985673189163208,
      "epoch": 0.20636792452830188,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499848971584951,
      "loss": 5.8403,
      "mean_token_accuracy": 0.15447261780500413,
      "num_tokens": 4706438.0,
      "step": 2275
    },
    {
      "entropy": 5.944012832641602,
      "epoch": 0.20682148040638607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998477839262611,
      "loss": 5.8513,
      "mean_token_accuracy": 0.1460217833518982,
      "num_tokens": 4716402.0,
      "step": 2280
    },
    {
      "entropy": 6.024405479431152,
      "epoch": 0.20727503628447025,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499846591617662,
      "loss": 5.9409,
      "mean_token_accuracy": 0.14774347990751266,
      "num_tokens": 4728488.0,
      "step": 2285
    },
    {
      "entropy": 6.027887296676636,
      "epoch": 0.20772859216255443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998453946591779,
      "loss": 5.9638,
      "mean_token_accuracy": 0.14974903911352158,
      "num_tokens": 4738693.0,
      "step": 2290
    },
    {
      "entropy": 6.020720100402832,
      "epoch": 0.20818214804063861,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998441930508339,
      "loss": 5.8436,
      "mean_token_accuracy": 0.1517803430557251,
      "num_tokens": 4748895.0,
      "step": 2295
    },
    {
      "entropy": 6.089075136184692,
      "epoch": 0.2086357039187228,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998429867926547,
      "loss": 5.9548,
      "mean_token_accuracy": 0.14135059341788292,
      "num_tokens": 4760135.0,
      "step": 2300
    },
    {
      "entropy": 6.095640230178833,
      "epoch": 0.20908925979680695,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998417758846651,
      "loss": 5.9916,
      "mean_token_accuracy": 0.15003142952919007,
      "num_tokens": 4769935.0,
      "step": 2305
    },
    {
      "entropy": 6.164234828948975,
      "epoch": 0.20954281567489114,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998405603268903,
      "loss": 5.9117,
      "mean_token_accuracy": 0.14601739197969438,
      "num_tokens": 4779520.0,
      "step": 2310
    },
    {
      "entropy": 5.891173362731934,
      "epoch": 0.20999637155297532,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998393401193554,
      "loss": 5.8264,
      "mean_token_accuracy": 0.1525731474161148,
      "num_tokens": 4791256.0,
      "step": 2315
    },
    {
      "entropy": 5.922054576873779,
      "epoch": 0.2104499274310595,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998381152620857,
      "loss": 5.7778,
      "mean_token_accuracy": 0.15643645375967025,
      "num_tokens": 4801334.0,
      "step": 2320
    },
    {
      "entropy": 5.940741872787475,
      "epoch": 0.21090348330914369,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998368857551067,
      "loss": 5.7694,
      "mean_token_accuracy": 0.15321901589632034,
      "num_tokens": 4811361.0,
      "step": 2325
    },
    {
      "entropy": 5.995629167556762,
      "epoch": 0.21135703918722787,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998356515984433,
      "loss": 5.8362,
      "mean_token_accuracy": 0.14930329322814942,
      "num_tokens": 4821190.0,
      "step": 2330
    },
    {
      "entropy": 5.996551895141602,
      "epoch": 0.21181059506531205,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998344127921213,
      "loss": 5.8616,
      "mean_token_accuracy": 0.15249975472688676,
      "num_tokens": 4830694.0,
      "step": 2335
    },
    {
      "entropy": 5.9884954452514645,
      "epoch": 0.21226415094339623,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998331693361666,
      "loss": 5.8478,
      "mean_token_accuracy": 0.14835911467671395,
      "num_tokens": 4841344.0,
      "step": 2340
    },
    {
      "entropy": 5.958621692657471,
      "epoch": 0.21271770682148042,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998319212306046,
      "loss": 5.8214,
      "mean_token_accuracy": 0.1466705918312073,
      "num_tokens": 4851161.0,
      "step": 2345
    },
    {
      "entropy": 5.9765862941741945,
      "epoch": 0.21317126269956457,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499830668475461,
      "loss": 5.8198,
      "mean_token_accuracy": 0.1497928962111473,
      "num_tokens": 4860740.0,
      "step": 2350
    },
    {
      "entropy": 5.9823229789733885,
      "epoch": 0.21362481857764876,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499829411070762,
      "loss": 5.8482,
      "mean_token_accuracy": 0.14942808002233504,
      "num_tokens": 4871733.0,
      "step": 2355
    },
    {
      "entropy": 6.0018017292022705,
      "epoch": 0.21407837445573294,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998281490165335,
      "loss": 5.9064,
      "mean_token_accuracy": 0.14177072122693063,
      "num_tokens": 4883617.0,
      "step": 2360
    },
    {
      "entropy": 6.005364894866943,
      "epoch": 0.21453193033381712,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998268823128016,
      "loss": 5.8483,
      "mean_token_accuracy": 0.14455199390649795,
      "num_tokens": 4894519.0,
      "step": 2365
    },
    {
      "entropy": 5.98149266242981,
      "epoch": 0.2149854862119013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998256109595923,
      "loss": 5.8014,
      "mean_token_accuracy": 0.14887023940682412,
      "num_tokens": 4904965.0,
      "step": 2370
    },
    {
      "entropy": 5.936092567443848,
      "epoch": 0.2154390420899855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998243349569323,
      "loss": 5.7256,
      "mean_token_accuracy": 0.1545273095369339,
      "num_tokens": 4914618.0,
      "step": 2375
    },
    {
      "entropy": 5.868344449996949,
      "epoch": 0.21589259796806967,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998230543048477,
      "loss": 5.7769,
      "mean_token_accuracy": 0.15782049298286438,
      "num_tokens": 4924341.0,
      "step": 2380
    },
    {
      "entropy": 5.880142879486084,
      "epoch": 0.21634615384615385,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499821769003365,
      "loss": 5.7439,
      "mean_token_accuracy": 0.15210559889674186,
      "num_tokens": 4934231.0,
      "step": 2385
    },
    {
      "entropy": 5.974342632293701,
      "epoch": 0.21679970972423804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998204790525109,
      "loss": 5.7777,
      "mean_token_accuracy": 0.1491807922720909,
      "num_tokens": 4944303.0,
      "step": 2390
    },
    {
      "entropy": 5.959984636306762,
      "epoch": 0.2172532656023222,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998191844523119,
      "loss": 5.8578,
      "mean_token_accuracy": 0.15104133933782576,
      "num_tokens": 4954571.0,
      "step": 2395
    },
    {
      "entropy": 5.957349252700806,
      "epoch": 0.21770682148040638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499817885202795,
      "loss": 5.8266,
      "mean_token_accuracy": 0.14804138243198395,
      "num_tokens": 4964418.0,
      "step": 2400
    },
    {
      "entropy": 5.975922727584839,
      "epoch": 0.21816037735849056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998165813039869,
      "loss": 5.8163,
      "mean_token_accuracy": 0.1464727595448494,
      "num_tokens": 4975036.0,
      "step": 2405
    },
    {
      "entropy": 5.98774995803833,
      "epoch": 0.21861393323657474,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998152727559145,
      "loss": 5.9215,
      "mean_token_accuracy": 0.14540747702121734,
      "num_tokens": 4985272.0,
      "step": 2410
    },
    {
      "entropy": 6.043728685379028,
      "epoch": 0.21906748911465893,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499813959558605,
      "loss": 5.8498,
      "mean_token_accuracy": 0.15086397528648376,
      "num_tokens": 4995187.0,
      "step": 2415
    },
    {
      "entropy": 6.0442393779754635,
      "epoch": 0.2195210449927431,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998126417120856,
      "loss": 5.8919,
      "mean_token_accuracy": 0.15320269912481307,
      "num_tokens": 5005013.0,
      "step": 2420
    },
    {
      "entropy": 6.054208087921142,
      "epoch": 0.2199746008708273,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998113192163835,
      "loss": 5.9139,
      "mean_token_accuracy": 0.1428740605711937,
      "num_tokens": 5016789.0,
      "step": 2425
    },
    {
      "entropy": 5.934090423583984,
      "epoch": 0.22042815674891147,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499809992071526,
      "loss": 5.8588,
      "mean_token_accuracy": 0.147943913936615,
      "num_tokens": 5026913.0,
      "step": 2430
    },
    {
      "entropy": 5.962672376632691,
      "epoch": 0.22088171262699566,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998086602775406,
      "loss": 5.9471,
      "mean_token_accuracy": 0.1470245487987995,
      "num_tokens": 5037501.0,
      "step": 2435
    },
    {
      "entropy": 6.027901649475098,
      "epoch": 0.2213352685050798,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998073238344547,
      "loss": 5.7987,
      "mean_token_accuracy": 0.14490654170513154,
      "num_tokens": 5048478.0,
      "step": 2440
    },
    {
      "entropy": 5.919871520996094,
      "epoch": 0.221788824383164,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998059827422962,
      "loss": 5.859,
      "mean_token_accuracy": 0.1465262070298195,
      "num_tokens": 5059118.0,
      "step": 2445
    },
    {
      "entropy": 6.006969404220581,
      "epoch": 0.22224238026124818,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998046370010926,
      "loss": 5.8236,
      "mean_token_accuracy": 0.15497831404209136,
      "num_tokens": 5070180.0,
      "step": 2450
    },
    {
      "entropy": 5.935058832168579,
      "epoch": 0.22269593613933236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998032866108718,
      "loss": 5.8128,
      "mean_token_accuracy": 0.14882795214653016,
      "num_tokens": 5079912.0,
      "step": 2455
    },
    {
      "entropy": 6.052799558639526,
      "epoch": 0.22314949201741655,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998019315716618,
      "loss": 5.9009,
      "mean_token_accuracy": 0.1513648509979248,
      "num_tokens": 5090549.0,
      "step": 2460
    },
    {
      "entropy": 5.9224653244018555,
      "epoch": 0.22360304789550073,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998005718834905,
      "loss": 5.7505,
      "mean_token_accuracy": 0.15711326152086258,
      "num_tokens": 5100880.0,
      "step": 2465
    },
    {
      "entropy": 5.927716445922852,
      "epoch": 0.2240566037735849,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997992075463862,
      "loss": 5.8622,
      "mean_token_accuracy": 0.14909446835517884,
      "num_tokens": 5111796.0,
      "step": 2470
    },
    {
      "entropy": 6.045079755783081,
      "epoch": 0.2245101596516691,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997978385603769,
      "loss": 5.862,
      "mean_token_accuracy": 0.1490783780813217,
      "num_tokens": 5123468.0,
      "step": 2475
    },
    {
      "entropy": 5.968772125244141,
      "epoch": 0.22496371552975328,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997964649254911,
      "loss": 5.7749,
      "mean_token_accuracy": 0.15539813786745071,
      "num_tokens": 5132925.0,
      "step": 2480
    },
    {
      "entropy": 5.98187403678894,
      "epoch": 0.22541727140783743,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997950866417572,
      "loss": 5.8555,
      "mean_token_accuracy": 0.14465042054653168,
      "num_tokens": 5143445.0,
      "step": 2485
    },
    {
      "entropy": 6.025763702392578,
      "epoch": 0.22587082728592162,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997937037092034,
      "loss": 5.7817,
      "mean_token_accuracy": 0.14337050616741182,
      "num_tokens": 5153176.0,
      "step": 2490
    },
    {
      "entropy": 6.005899715423584,
      "epoch": 0.2263243831640058,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997923161278587,
      "loss": 5.8267,
      "mean_token_accuracy": 0.1517674669623375,
      "num_tokens": 5162765.0,
      "step": 2495
    },
    {
      "entropy": 5.946051549911499,
      "epoch": 0.22677793904208998,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997909238977514,
      "loss": 5.8804,
      "mean_token_accuracy": 0.15172123089432715,
      "num_tokens": 5173390.0,
      "step": 2500
    },
    {
      "entropy": 5.946356534957886,
      "epoch": 0.22723149492017417,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997895270189108,
      "loss": 5.7811,
      "mean_token_accuracy": 0.14956301152706147,
      "num_tokens": 5184212.0,
      "step": 2505
    },
    {
      "entropy": 5.886882019042969,
      "epoch": 0.22768505079825835,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997881254913653,
      "loss": 5.7728,
      "mean_token_accuracy": 0.15003687813878058,
      "num_tokens": 5194311.0,
      "step": 2510
    },
    {
      "entropy": 5.916078996658325,
      "epoch": 0.22813860667634253,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997867193151441,
      "loss": 5.7712,
      "mean_token_accuracy": 0.15247755497694016,
      "num_tokens": 5204708.0,
      "step": 2515
    },
    {
      "entropy": 5.87827501296997,
      "epoch": 0.22859216255442671,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997853084902763,
      "loss": 5.7469,
      "mean_token_accuracy": 0.15128237307071685,
      "num_tokens": 5214566.0,
      "step": 2520
    },
    {
      "entropy": 5.943019533157349,
      "epoch": 0.2290457184325109,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997838930167911,
      "loss": 5.7878,
      "mean_token_accuracy": 0.15266707986593248,
      "num_tokens": 5224762.0,
      "step": 2525
    },
    {
      "entropy": 5.9829565525054935,
      "epoch": 0.22949927431059505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997824728947176,
      "loss": 5.8435,
      "mean_token_accuracy": 0.15028240010142327,
      "num_tokens": 5235146.0,
      "step": 2530
    },
    {
      "entropy": 6.023767328262329,
      "epoch": 0.22995283018867924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997810481240853,
      "loss": 5.8108,
      "mean_token_accuracy": 0.14803099185228347,
      "num_tokens": 5243867.0,
      "step": 2535
    },
    {
      "entropy": 5.893141078948974,
      "epoch": 0.23040638606676342,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997796187049238,
      "loss": 5.7234,
      "mean_token_accuracy": 0.1500532016158104,
      "num_tokens": 5254279.0,
      "step": 2540
    },
    {
      "entropy": 6.080553483963013,
      "epoch": 0.2308599419448476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997781846372623,
      "loss": 5.8382,
      "mean_token_accuracy": 0.14982936531305313,
      "num_tokens": 5264769.0,
      "step": 2545
    },
    {
      "entropy": 5.93544192314148,
      "epoch": 0.23131349782293179,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997767459211307,
      "loss": 5.7903,
      "mean_token_accuracy": 0.14275677353143693,
      "num_tokens": 5274291.0,
      "step": 2550
    },
    {
      "entropy": 5.883300113677978,
      "epoch": 0.23176705370101597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997753025565587,
      "loss": 5.7326,
      "mean_token_accuracy": 0.15632595419883727,
      "num_tokens": 5284366.0,
      "step": 2555
    },
    {
      "entropy": 5.95240535736084,
      "epoch": 0.23222060957910015,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997738545435763,
      "loss": 5.8456,
      "mean_token_accuracy": 0.15440944731235504,
      "num_tokens": 5294494.0,
      "step": 2560
    },
    {
      "entropy": 5.928846740722657,
      "epoch": 0.23267416545718433,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997724018822133,
      "loss": 5.8394,
      "mean_token_accuracy": 0.15485012233257295,
      "num_tokens": 5305606.0,
      "step": 2565
    },
    {
      "entropy": 5.960556077957153,
      "epoch": 0.23312772133526852,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997709445724996,
      "loss": 5.8465,
      "mean_token_accuracy": 0.150348149985075,
      "num_tokens": 5315677.0,
      "step": 2570
    },
    {
      "entropy": 5.986758232116699,
      "epoch": 0.23358127721335267,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997694826144656,
      "loss": 5.9049,
      "mean_token_accuracy": 0.1463898465037346,
      "num_tokens": 5326453.0,
      "step": 2575
    },
    {
      "entropy": 5.92100830078125,
      "epoch": 0.23403483309143686,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997680160081414,
      "loss": 5.7906,
      "mean_token_accuracy": 0.15171819776296616,
      "num_tokens": 5337591.0,
      "step": 2580
    },
    {
      "entropy": 5.9538902759552,
      "epoch": 0.23448838896952104,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997665447535573,
      "loss": 5.7791,
      "mean_token_accuracy": 0.15522343516349793,
      "num_tokens": 5349400.0,
      "step": 2585
    },
    {
      "entropy": 6.00141224861145,
      "epoch": 0.23494194484760522,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997650688507438,
      "loss": 5.7634,
      "mean_token_accuracy": 0.1506360337138176,
      "num_tokens": 5359984.0,
      "step": 2590
    },
    {
      "entropy": 5.916184520721435,
      "epoch": 0.2353955007256894,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997635882997314,
      "loss": 5.845,
      "mean_token_accuracy": 0.15854744762182235,
      "num_tokens": 5369254.0,
      "step": 2595
    },
    {
      "entropy": 5.9528069496154785,
      "epoch": 0.2358490566037736,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997621031005507,
      "loss": 5.7798,
      "mean_token_accuracy": 0.15245212465524674,
      "num_tokens": 5379648.0,
      "step": 2600
    },
    {
      "entropy": 5.849475145339966,
      "epoch": 0.23630261248185777,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997606132532325,
      "loss": 5.7372,
      "mean_token_accuracy": 0.154351457208395,
      "num_tokens": 5390321.0,
      "step": 2605
    },
    {
      "entropy": 5.919327926635742,
      "epoch": 0.23675616835994195,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997591187578075,
      "loss": 5.7926,
      "mean_token_accuracy": 0.15511986911296843,
      "num_tokens": 5401078.0,
      "step": 2610
    },
    {
      "entropy": 5.961600828170776,
      "epoch": 0.23720972423802614,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997576196143067,
      "loss": 5.7453,
      "mean_token_accuracy": 0.1566541612148285,
      "num_tokens": 5411977.0,
      "step": 2615
    },
    {
      "entropy": 5.893222713470459,
      "epoch": 0.2376632801161103,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499756115822761,
      "loss": 5.7368,
      "mean_token_accuracy": 0.15138195902109147,
      "num_tokens": 5422967.0,
      "step": 2620
    },
    {
      "entropy": 5.9912902355194095,
      "epoch": 0.23811683599419448,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997546073832015,
      "loss": 5.7835,
      "mean_token_accuracy": 0.16118090003728866,
      "num_tokens": 5432260.0,
      "step": 2625
    },
    {
      "entropy": 5.888670158386231,
      "epoch": 0.23857039187227866,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997530942956595,
      "loss": 5.7564,
      "mean_token_accuracy": 0.15361348688602447,
      "num_tokens": 5441706.0,
      "step": 2630
    },
    {
      "entropy": 5.94991717338562,
      "epoch": 0.23902394775036284,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997515765601662,
      "loss": 5.7622,
      "mean_token_accuracy": 0.1588932454586029,
      "num_tokens": 5452397.0,
      "step": 2635
    },
    {
      "entropy": 5.880772686004638,
      "epoch": 0.23947750362844702,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997500541767531,
      "loss": 5.7218,
      "mean_token_accuracy": 0.15416701436042785,
      "num_tokens": 5463762.0,
      "step": 2640
    },
    {
      "entropy": 5.901000833511352,
      "epoch": 0.2399310595065312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997485271454516,
      "loss": 5.7915,
      "mean_token_accuracy": 0.14654496908187867,
      "num_tokens": 5474622.0,
      "step": 2645
    },
    {
      "entropy": 5.944158935546875,
      "epoch": 0.2403846153846154,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997469954662933,
      "loss": 5.7691,
      "mean_token_accuracy": 0.15286272019147873,
      "num_tokens": 5484697.0,
      "step": 2650
    },
    {
      "entropy": 5.847146940231323,
      "epoch": 0.24083817126269957,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997454591393097,
      "loss": 5.7742,
      "mean_token_accuracy": 0.15464416444301604,
      "num_tokens": 5496433.0,
      "step": 2655
    },
    {
      "entropy": 5.972603416442871,
      "epoch": 0.24129172714078376,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997439181645329,
      "loss": 5.8895,
      "mean_token_accuracy": 0.1483437992632389,
      "num_tokens": 5507917.0,
      "step": 2660
    },
    {
      "entropy": 6.07913875579834,
      "epoch": 0.2417452830188679,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997423725419945,
      "loss": 5.8717,
      "mean_token_accuracy": 0.14761385768651963,
      "num_tokens": 5519159.0,
      "step": 2665
    },
    {
      "entropy": 5.928588962554931,
      "epoch": 0.2421988388969521,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997408222717267,
      "loss": 5.8249,
      "mean_token_accuracy": 0.15743801593780518,
      "num_tokens": 5529180.0,
      "step": 2670
    },
    {
      "entropy": 5.884641027450561,
      "epoch": 0.24265239477503628,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997392673537612,
      "loss": 5.7382,
      "mean_token_accuracy": 0.16192516684532166,
      "num_tokens": 5539363.0,
      "step": 2675
    },
    {
      "entropy": 5.909084510803223,
      "epoch": 0.24310595065312046,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997377077881306,
      "loss": 5.843,
      "mean_token_accuracy": 0.15110581144690513,
      "num_tokens": 5549841.0,
      "step": 2680
    },
    {
      "entropy": 5.940350866317749,
      "epoch": 0.24355950653120464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997361435748668,
      "loss": 5.7156,
      "mean_token_accuracy": 0.161916483938694,
      "num_tokens": 5560299.0,
      "step": 2685
    },
    {
      "entropy": 5.962108039855957,
      "epoch": 0.24401306240928883,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997345747140024,
      "loss": 5.8001,
      "mean_token_accuracy": 0.15469845086336137,
      "num_tokens": 5571081.0,
      "step": 2690
    },
    {
      "entropy": 5.871194362640381,
      "epoch": 0.244466618287373,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997330012055695,
      "loss": 5.7317,
      "mean_token_accuracy": 0.15739612132310868,
      "num_tokens": 5581389.0,
      "step": 2695
    },
    {
      "entropy": 5.881601572036743,
      "epoch": 0.2449201741654572,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997314230496011,
      "loss": 5.779,
      "mean_token_accuracy": 0.14633584022521973,
      "num_tokens": 5592347.0,
      "step": 2700
    },
    {
      "entropy": 5.889206314086914,
      "epoch": 0.24537373004354138,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997298402461296,
      "loss": 5.7548,
      "mean_token_accuracy": 0.15528458803892137,
      "num_tokens": 5603411.0,
      "step": 2705
    },
    {
      "entropy": 5.930398607254029,
      "epoch": 0.24582728592162553,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997282527951876,
      "loss": 5.7452,
      "mean_token_accuracy": 0.15865262895822524,
      "num_tokens": 5614615.0,
      "step": 2710
    },
    {
      "entropy": 5.811875200271606,
      "epoch": 0.24628084179970972,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997266606968081,
      "loss": 5.7026,
      "mean_token_accuracy": 0.15606526732444764,
      "num_tokens": 5624582.0,
      "step": 2715
    },
    {
      "entropy": 5.876440620422363,
      "epoch": 0.2467343976777939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997250639510241,
      "loss": 5.7258,
      "mean_token_accuracy": 0.15305925607681276,
      "num_tokens": 5634310.0,
      "step": 2720
    },
    {
      "entropy": 5.932286167144776,
      "epoch": 0.24718795355587808,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997234625578684,
      "loss": 5.8186,
      "mean_token_accuracy": 0.15080375373363494,
      "num_tokens": 5644018.0,
      "step": 2725
    },
    {
      "entropy": 5.862547779083252,
      "epoch": 0.24764150943396226,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997218565173742,
      "loss": 5.7485,
      "mean_token_accuracy": 0.15690971165895462,
      "num_tokens": 5653592.0,
      "step": 2730
    },
    {
      "entropy": 5.973826742172241,
      "epoch": 0.24809506531204645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997202458295748,
      "loss": 5.7929,
      "mean_token_accuracy": 0.155655175447464,
      "num_tokens": 5662812.0,
      "step": 2735
    },
    {
      "entropy": 5.936570167541504,
      "epoch": 0.24854862119013063,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997186304945034,
      "loss": 5.7758,
      "mean_token_accuracy": 0.1513756588101387,
      "num_tokens": 5672939.0,
      "step": 2740
    },
    {
      "entropy": 5.83735408782959,
      "epoch": 0.24900217706821481,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997170105121935,
      "loss": 5.7543,
      "mean_token_accuracy": 0.15465778708457947,
      "num_tokens": 5683350.0,
      "step": 2745
    },
    {
      "entropy": 5.977609300613404,
      "epoch": 0.249455732946299,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997153858826786,
      "loss": 5.8559,
      "mean_token_accuracy": 0.1457457385957241,
      "num_tokens": 5694516.0,
      "step": 2750
    },
    {
      "entropy": 5.897079944610596,
      "epoch": 0.24990928882438315,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997137566059922,
      "loss": 5.6687,
      "mean_token_accuracy": 0.15890194922685624,
      "num_tokens": 5705443.0,
      "step": 2755
    },
    {
      "entropy": 5.876714992523193,
      "epoch": 0.25036284470246734,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997121226821681,
      "loss": 5.772,
      "mean_token_accuracy": 0.1513215646147728,
      "num_tokens": 5715933.0,
      "step": 2760
    },
    {
      "entropy": 5.92222113609314,
      "epoch": 0.25081640058055155,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049971048411124,
      "loss": 5.835,
      "mean_token_accuracy": 0.14457671642303466,
      "num_tokens": 5726630.0,
      "step": 2765
    },
    {
      "entropy": 6.002235460281372,
      "epoch": 0.2512699564586357,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997088408932418,
      "loss": 5.7697,
      "mean_token_accuracy": 0.1539665475487709,
      "num_tokens": 5736890.0,
      "step": 2770
    },
    {
      "entropy": 5.778037357330322,
      "epoch": 0.25172351233671986,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997071930282076,
      "loss": 5.6726,
      "mean_token_accuracy": 0.1656890943646431,
      "num_tokens": 5747042.0,
      "step": 2775
    },
    {
      "entropy": 5.874831628799439,
      "epoch": 0.25217706821480407,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997055405161712,
      "loss": 5.7902,
      "mean_token_accuracy": 0.1548820838332176,
      "num_tokens": 5757367.0,
      "step": 2780
    },
    {
      "entropy": 5.933564472198486,
      "epoch": 0.2526306240928882,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997038833571671,
      "loss": 5.8466,
      "mean_token_accuracy": 0.14902109056711196,
      "num_tokens": 5769557.0,
      "step": 2785
    },
    {
      "entropy": 5.887054443359375,
      "epoch": 0.25308417997097243,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997022215512295,
      "loss": 5.7514,
      "mean_token_accuracy": 0.15811785459518432,
      "num_tokens": 5780615.0,
      "step": 2790
    },
    {
      "entropy": 5.939122200012207,
      "epoch": 0.2535377358490566,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997005550983925,
      "loss": 5.7361,
      "mean_token_accuracy": 0.1549833208322525,
      "num_tokens": 5791271.0,
      "step": 2795
    },
    {
      "entropy": 5.814103412628174,
      "epoch": 0.2539912917271408,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499698883998691,
      "loss": 5.6636,
      "mean_token_accuracy": 0.15836629867553711,
      "num_tokens": 5801117.0,
      "step": 2800
    },
    {
      "entropy": 5.859375476837158,
      "epoch": 0.25444484760522496,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996972082521591,
      "loss": 5.6955,
      "mean_token_accuracy": 0.15782571583986282,
      "num_tokens": 5811696.0,
      "step": 2805
    },
    {
      "entropy": 5.879450893402099,
      "epoch": 0.25489840348330917,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996955278588319,
      "loss": 5.7319,
      "mean_token_accuracy": 0.16295181214809418,
      "num_tokens": 5822207.0,
      "step": 2810
    },
    {
      "entropy": 5.838560342788696,
      "epoch": 0.2553519593613933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996938428187437,
      "loss": 5.6753,
      "mean_token_accuracy": 0.16222673356533052,
      "num_tokens": 5831927.0,
      "step": 2815
    },
    {
      "entropy": 5.844130134582519,
      "epoch": 0.2558055152394775,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996921531319297,
      "loss": 5.7406,
      "mean_token_accuracy": 0.15631944462656974,
      "num_tokens": 5842093.0,
      "step": 2820
    },
    {
      "entropy": 5.862317848205566,
      "epoch": 0.2562590711175617,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996904587984246,
      "loss": 5.6799,
      "mean_token_accuracy": 0.1637733742594719,
      "num_tokens": 5850551.0,
      "step": 2825
    },
    {
      "entropy": 5.831214189529419,
      "epoch": 0.25671262699564584,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996887598182636,
      "loss": 5.6693,
      "mean_token_accuracy": 0.15734255760908128,
      "num_tokens": 5860651.0,
      "step": 2830
    },
    {
      "entropy": 5.736221647262573,
      "epoch": 0.25716618287373005,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996870561914819,
      "loss": 5.6508,
      "mean_token_accuracy": 0.15711580514907836,
      "num_tokens": 5871500.0,
      "step": 2835
    },
    {
      "entropy": 5.837002754211426,
      "epoch": 0.2576197387518142,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996853479181145,
      "loss": 5.6483,
      "mean_token_accuracy": 0.15363348871469498,
      "num_tokens": 5881523.0,
      "step": 2840
    },
    {
      "entropy": 5.969505023956299,
      "epoch": 0.2580732946298984,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996836349981969,
      "loss": 5.8168,
      "mean_token_accuracy": 0.15440904647111892,
      "num_tokens": 5891162.0,
      "step": 2845
    },
    {
      "entropy": 5.825331020355224,
      "epoch": 0.2585268505079826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996819174317645,
      "loss": 5.6989,
      "mean_token_accuracy": 0.15304600223898887,
      "num_tokens": 5901818.0,
      "step": 2850
    },
    {
      "entropy": 5.837537527084351,
      "epoch": 0.2589804063860668,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996801952188528,
      "loss": 5.7802,
      "mean_token_accuracy": 0.15388137996196746,
      "num_tokens": 5913889.0,
      "step": 2855
    },
    {
      "entropy": 5.898041200637818,
      "epoch": 0.25943396226415094,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996784683594973,
      "loss": 5.7331,
      "mean_token_accuracy": 0.15773886442184448,
      "num_tokens": 5923429.0,
      "step": 2860
    },
    {
      "entropy": 5.9350752353668215,
      "epoch": 0.2598875181422351,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996767368537338,
      "loss": 5.7756,
      "mean_token_accuracy": 0.15665830671787262,
      "num_tokens": 5933847.0,
      "step": 2865
    },
    {
      "entropy": 5.846282577514648,
      "epoch": 0.2603410740203193,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996750007015983,
      "loss": 5.6519,
      "mean_token_accuracy": 0.15802753269672393,
      "num_tokens": 5943874.0,
      "step": 2870
    },
    {
      "entropy": 5.893306922912598,
      "epoch": 0.26079462989840346,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996732599031264,
      "loss": 5.7207,
      "mean_token_accuracy": 0.1564636617898941,
      "num_tokens": 5953653.0,
      "step": 2875
    },
    {
      "entropy": 5.912632322311401,
      "epoch": 0.2612481857764877,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996715144583543,
      "loss": 5.6787,
      "mean_token_accuracy": 0.1598701536655426,
      "num_tokens": 5964039.0,
      "step": 2880
    },
    {
      "entropy": 5.934096145629883,
      "epoch": 0.26170174165457183,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499669764367318,
      "loss": 5.7888,
      "mean_token_accuracy": 0.15126357972621918,
      "num_tokens": 5974083.0,
      "step": 2885
    },
    {
      "entropy": 5.802870368957519,
      "epoch": 0.26215529753265604,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996680096300537,
      "loss": 5.7352,
      "mean_token_accuracy": 0.1535191372036934,
      "num_tokens": 5984629.0,
      "step": 2890
    },
    {
      "entropy": 5.837651777267456,
      "epoch": 0.2626088534107402,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996662502465978,
      "loss": 5.6433,
      "mean_token_accuracy": 0.16809626668691635,
      "num_tokens": 5994295.0,
      "step": 2895
    },
    {
      "entropy": 5.847111940383911,
      "epoch": 0.2630624092888244,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996644862169864,
      "loss": 5.6911,
      "mean_token_accuracy": 0.15560346096754074,
      "num_tokens": 6004156.0,
      "step": 2900
    },
    {
      "entropy": 5.842682647705078,
      "epoch": 0.26351596516690856,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996627175412563,
      "loss": 5.7724,
      "mean_token_accuracy": 0.15826719254255295,
      "num_tokens": 6014188.0,
      "step": 2905
    },
    {
      "entropy": 5.951080179214477,
      "epoch": 0.2639695210449927,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499660944219444,
      "loss": 5.8318,
      "mean_token_accuracy": 0.15130359828472137,
      "num_tokens": 6024879.0,
      "step": 2910
    },
    {
      "entropy": 5.973435401916504,
      "epoch": 0.2644230769230769,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996591662515861,
      "loss": 5.8978,
      "mean_token_accuracy": 0.1513467513024807,
      "num_tokens": 6036374.0,
      "step": 2915
    },
    {
      "entropy": 5.852985286712647,
      "epoch": 0.2648766328011611,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996573836377193,
      "loss": 5.7384,
      "mean_token_accuracy": 0.15605844408273697,
      "num_tokens": 6046432.0,
      "step": 2920
    },
    {
      "entropy": 5.9294966697692875,
      "epoch": 0.2653301886792453,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996555963778806,
      "loss": 5.7428,
      "mean_token_accuracy": 0.14988288134336472,
      "num_tokens": 6056864.0,
      "step": 2925
    },
    {
      "entropy": 5.9699132442474365,
      "epoch": 0.26578374455732945,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499653804472107,
      "loss": 5.7612,
      "mean_token_accuracy": 0.1611533209681511,
      "num_tokens": 6067294.0,
      "step": 2930
    },
    {
      "entropy": 5.882857704162598,
      "epoch": 0.26623730043541366,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996520079204353,
      "loss": 5.7721,
      "mean_token_accuracy": 0.15410346388816834,
      "num_tokens": 6077587.0,
      "step": 2935
    },
    {
      "entropy": 5.921375894546509,
      "epoch": 0.2666908563134978,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499650206722903,
      "loss": 5.7142,
      "mean_token_accuracy": 0.15302591323852538,
      "num_tokens": 6089306.0,
      "step": 2940
    },
    {
      "entropy": 5.801258993148804,
      "epoch": 0.267144412191582,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996484008795471,
      "loss": 5.6718,
      "mean_token_accuracy": 0.15564401000738143,
      "num_tokens": 6099651.0,
      "step": 2945
    },
    {
      "entropy": 5.831615734100342,
      "epoch": 0.2675979680696662,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499646590390405,
      "loss": 5.6786,
      "mean_token_accuracy": 0.15340599417686462,
      "num_tokens": 6110063.0,
      "step": 2950
    },
    {
      "entropy": 5.866582584381104,
      "epoch": 0.26805152394775034,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996447752555142,
      "loss": 5.7069,
      "mean_token_accuracy": 0.1569851815700531,
      "num_tokens": 6120031.0,
      "step": 2955
    },
    {
      "entropy": 5.845797681808472,
      "epoch": 0.26850507982583455,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996429554749122,
      "loss": 5.7447,
      "mean_token_accuracy": 0.1586455672979355,
      "num_tokens": 6131274.0,
      "step": 2960
    },
    {
      "entropy": 5.888718032836914,
      "epoch": 0.2689586357039187,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996411310486367,
      "loss": 5.7833,
      "mean_token_accuracy": 0.15850588232278823,
      "num_tokens": 6140878.0,
      "step": 2965
    },
    {
      "entropy": 5.895853567123413,
      "epoch": 0.2694121915820029,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996393019767253,
      "loss": 5.7917,
      "mean_token_accuracy": 0.15040715038776398,
      "num_tokens": 6152229.0,
      "step": 2970
    },
    {
      "entropy": 5.802438020706177,
      "epoch": 0.26986574746008707,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996374682592158,
      "loss": 5.6873,
      "mean_token_accuracy": 0.15607988238334655,
      "num_tokens": 6162311.0,
      "step": 2975
    },
    {
      "entropy": 5.942750644683838,
      "epoch": 0.2703193033381713,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996356298961463,
      "loss": 5.7936,
      "mean_token_accuracy": 0.15298964083194733,
      "num_tokens": 6172806.0,
      "step": 2980
    },
    {
      "entropy": 5.87895131111145,
      "epoch": 0.27077285921625543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996337868875549,
      "loss": 5.7587,
      "mean_token_accuracy": 0.15776319652795792,
      "num_tokens": 6183479.0,
      "step": 2985
    },
    {
      "entropy": 5.886080932617188,
      "epoch": 0.27122641509433965,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996319392334792,
      "loss": 5.7299,
      "mean_token_accuracy": 0.15342071503400803,
      "num_tokens": 6194201.0,
      "step": 2990
    },
    {
      "entropy": 5.82788724899292,
      "epoch": 0.2716799709724238,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499630086933958,
      "loss": 5.6724,
      "mean_token_accuracy": 0.1564711168408394,
      "num_tokens": 6204774.0,
      "step": 2995
    },
    {
      "entropy": 5.886050224304199,
      "epoch": 0.27213352685050796,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996282299890292,
      "loss": 5.6831,
      "mean_token_accuracy": 0.1526743471622467,
      "num_tokens": 6215056.0,
      "step": 3000
    },
    {
      "epoch": 0.27213352685050796,
      "eval_entropy": 5.7014565256386325,
      "eval_loss": 5.748072624206543,
      "eval_mean_token_accuracy": 0.1622600189245799,
      "eval_num_tokens": 6215056.0,
      "eval_runtime": 20.702,
      "eval_samples_per_second": 1708.048,
      "eval_steps_per_second": 213.506,
      "step": 3000
    },
    {
      "entropy": 5.934528923034668,
      "epoch": 0.27258708272859217,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996263683987316,
      "loss": 5.786,
      "mean_token_accuracy": 0.1579520136117935,
      "num_tokens": 6226264.0,
      "step": 3005
    },
    {
      "entropy": 5.856761884689331,
      "epoch": 0.2730406386066763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996245021631033,
      "loss": 5.7038,
      "mean_token_accuracy": 0.161579892039299,
      "num_tokens": 6235137.0,
      "step": 3010
    },
    {
      "entropy": 5.824236536026001,
      "epoch": 0.27349419448476053,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499622631282183,
      "loss": 5.7119,
      "mean_token_accuracy": 0.15795171409845352,
      "num_tokens": 6245411.0,
      "step": 3015
    },
    {
      "entropy": 5.915331649780273,
      "epoch": 0.2739477503628447,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996207557560095,
      "loss": 5.7743,
      "mean_token_accuracy": 0.1543523669242859,
      "num_tokens": 6255633.0,
      "step": 3020
    },
    {
      "entropy": 5.906075763702392,
      "epoch": 0.2744013062409289,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996188755846216,
      "loss": 5.7388,
      "mean_token_accuracy": 0.1594541311264038,
      "num_tokens": 6265109.0,
      "step": 3025
    },
    {
      "entropy": 5.813056755065918,
      "epoch": 0.27485486211901305,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499616990768058,
      "loss": 5.7737,
      "mean_token_accuracy": 0.1503493830561638,
      "num_tokens": 6275241.0,
      "step": 3030
    },
    {
      "entropy": 5.902055835723877,
      "epoch": 0.27530841799709727,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996151013063579,
      "loss": 5.8127,
      "mean_token_accuracy": 0.15131743252277374,
      "num_tokens": 6286603.0,
      "step": 3035
    },
    {
      "entropy": 5.917137479782104,
      "epoch": 0.2757619738751814,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996132071995601,
      "loss": 5.6609,
      "mean_token_accuracy": 0.16706326305866243,
      "num_tokens": 6297219.0,
      "step": 3040
    },
    {
      "entropy": 5.804540967941284,
      "epoch": 0.2762155297532656,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996113084477042,
      "loss": 5.6991,
      "mean_token_accuracy": 0.16197659075260162,
      "num_tokens": 6307694.0,
      "step": 3045
    },
    {
      "entropy": 5.795589399337769,
      "epoch": 0.2766690856313498,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996094050508291,
      "loss": 5.6305,
      "mean_token_accuracy": 0.16191300749778748,
      "num_tokens": 6317384.0,
      "step": 3050
    },
    {
      "entropy": 5.88090558052063,
      "epoch": 0.27712264150943394,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996074970089741,
      "loss": 5.7749,
      "mean_token_accuracy": 0.15503887385129927,
      "num_tokens": 6327558.0,
      "step": 3055
    },
    {
      "entropy": 5.878848552703857,
      "epoch": 0.27757619738751815,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996055843221789,
      "loss": 5.6264,
      "mean_token_accuracy": 0.16327428370714187,
      "num_tokens": 6337828.0,
      "step": 3060
    },
    {
      "entropy": 5.789236450195313,
      "epoch": 0.2780297532656023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996036669904831,
      "loss": 5.6907,
      "mean_token_accuracy": 0.16211704909801483,
      "num_tokens": 6348678.0,
      "step": 3065
    },
    {
      "entropy": 5.864919281005859,
      "epoch": 0.2784833091436865,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996017450139261,
      "loss": 5.783,
      "mean_token_accuracy": 0.14863884598016738,
      "num_tokens": 6359670.0,
      "step": 3070
    },
    {
      "entropy": 5.968852424621582,
      "epoch": 0.2789368650217707,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995998183925478,
      "loss": 5.7405,
      "mean_token_accuracy": 0.15753814876079558,
      "num_tokens": 6371034.0,
      "step": 3075
    },
    {
      "entropy": 5.818412113189697,
      "epoch": 0.2793904208998549,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995978871263881,
      "loss": 5.6874,
      "mean_token_accuracy": 0.15982511192560195,
      "num_tokens": 6380672.0,
      "step": 3080
    },
    {
      "entropy": 5.8274352073669435,
      "epoch": 0.27984397677793904,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995959512154868,
      "loss": 5.7991,
      "mean_token_accuracy": 0.14778632521629334,
      "num_tokens": 6390799.0,
      "step": 3085
    },
    {
      "entropy": 5.859681892395019,
      "epoch": 0.2802975326560232,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499594010659884,
      "loss": 5.6602,
      "mean_token_accuracy": 0.1630452647805214,
      "num_tokens": 6400965.0,
      "step": 3090
    },
    {
      "entropy": 5.886337471008301,
      "epoch": 0.2807510885341074,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995920654596199,
      "loss": 5.6873,
      "mean_token_accuracy": 0.1525515004992485,
      "num_tokens": 6411567.0,
      "step": 3095
    },
    {
      "entropy": 5.864609479904175,
      "epoch": 0.28120464441219156,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995901156147345,
      "loss": 5.7302,
      "mean_token_accuracy": 0.16160865426063536,
      "num_tokens": 6421193.0,
      "step": 3100
    },
    {
      "entropy": 5.852170133590699,
      "epoch": 0.2816582002902758,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995881611252684,
      "loss": 5.7431,
      "mean_token_accuracy": 0.16411930173635483,
      "num_tokens": 6430281.0,
      "step": 3105
    },
    {
      "entropy": 5.838845443725586,
      "epoch": 0.28211175616835993,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995862019912619,
      "loss": 5.7401,
      "mean_token_accuracy": 0.1545553356409073,
      "num_tokens": 6441055.0,
      "step": 3110
    },
    {
      "entropy": 5.85880446434021,
      "epoch": 0.28256531204644414,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995842382127555,
      "loss": 5.665,
      "mean_token_accuracy": 0.15411336421966554,
      "num_tokens": 6451082.0,
      "step": 3115
    },
    {
      "entropy": 5.826427602767945,
      "epoch": 0.2830188679245283,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995822697897897,
      "loss": 5.7501,
      "mean_token_accuracy": 0.15211194902658462,
      "num_tokens": 6461617.0,
      "step": 3120
    },
    {
      "entropy": 5.894758319854736,
      "epoch": 0.2834724238026125,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995802967224056,
      "loss": 5.7542,
      "mean_token_accuracy": 0.15588101148605346,
      "num_tokens": 6472780.0,
      "step": 3125
    },
    {
      "entropy": 5.758359670639038,
      "epoch": 0.28392597968069666,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995783190106434,
      "loss": 5.6652,
      "mean_token_accuracy": 0.15255783200263978,
      "num_tokens": 6484078.0,
      "step": 3130
    },
    {
      "entropy": 5.893905305862427,
      "epoch": 0.2843795355587808,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995763366545445,
      "loss": 5.7515,
      "mean_token_accuracy": 0.15216615051031113,
      "num_tokens": 6494366.0,
      "step": 3135
    },
    {
      "entropy": 5.754382371902466,
      "epoch": 0.284833091436865,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995743496541497,
      "loss": 5.6839,
      "mean_token_accuracy": 0.16558870375156404,
      "num_tokens": 6505461.0,
      "step": 3140
    },
    {
      "entropy": 5.835577821731567,
      "epoch": 0.2852866473149492,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995723580095001,
      "loss": 5.6808,
      "mean_token_accuracy": 0.15959053486585617,
      "num_tokens": 6516141.0,
      "step": 3145
    },
    {
      "entropy": 5.781411981582641,
      "epoch": 0.2857402031930334,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499570361720637,
      "loss": 5.6851,
      "mean_token_accuracy": 0.154303365200758,
      "num_tokens": 6525536.0,
      "step": 3150
    },
    {
      "entropy": 5.864716672897339,
      "epoch": 0.28619375907111755,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995683607876015,
      "loss": 5.7363,
      "mean_token_accuracy": 0.15480921417474747,
      "num_tokens": 6535852.0,
      "step": 3155
    },
    {
      "entropy": 5.880465698242188,
      "epoch": 0.28664731494920176,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995663552104352,
      "loss": 5.7094,
      "mean_token_accuracy": 0.16345915347337722,
      "num_tokens": 6546353.0,
      "step": 3160
    },
    {
      "entropy": 5.76881217956543,
      "epoch": 0.2871008708272859,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995643449891793,
      "loss": 5.6586,
      "mean_token_accuracy": 0.15820098519325257,
      "num_tokens": 6557782.0,
      "step": 3165
    },
    {
      "entropy": 5.869258546829224,
      "epoch": 0.2875544267053701,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995623301238757,
      "loss": 5.6772,
      "mean_token_accuracy": 0.16326069831848145,
      "num_tokens": 6568100.0,
      "step": 3170
    },
    {
      "entropy": 5.809826850891113,
      "epoch": 0.2880079825834543,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995603106145657,
      "loss": 5.7354,
      "mean_token_accuracy": 0.15384810715913771,
      "num_tokens": 6578698.0,
      "step": 3175
    },
    {
      "entropy": 5.877403926849365,
      "epoch": 0.28846153846153844,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995582864612913,
      "loss": 5.7211,
      "mean_token_accuracy": 0.15826957821846008,
      "num_tokens": 6588908.0,
      "step": 3180
    },
    {
      "entropy": 5.767029571533203,
      "epoch": 0.28891509433962265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995562576640943,
      "loss": 5.577,
      "mean_token_accuracy": 0.16004678606987,
      "num_tokens": 6599445.0,
      "step": 3185
    },
    {
      "entropy": 5.799728536605835,
      "epoch": 0.2893686502177068,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995542242230167,
      "loss": 5.6578,
      "mean_token_accuracy": 0.16209590137004853,
      "num_tokens": 6610246.0,
      "step": 3190
    },
    {
      "entropy": 5.798572778701782,
      "epoch": 0.289822206095791,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995521861381006,
      "loss": 5.6769,
      "mean_token_accuracy": 0.15961335003376007,
      "num_tokens": 6620501.0,
      "step": 3195
    },
    {
      "entropy": 5.898326778411866,
      "epoch": 0.29027576197387517,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499550143409388,
      "loss": 5.7474,
      "mean_token_accuracy": 0.15129139870405198,
      "num_tokens": 6631206.0,
      "step": 3200
    },
    {
      "entropy": 5.829407978057861,
      "epoch": 0.2907293178519594,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995480960369211,
      "loss": 5.6945,
      "mean_token_accuracy": 0.1549079790711403,
      "num_tokens": 6641600.0,
      "step": 3205
    },
    {
      "entropy": 5.8490814685821535,
      "epoch": 0.29118287373004353,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995460440207425,
      "loss": 5.7631,
      "mean_token_accuracy": 0.15415688455104828,
      "num_tokens": 6652319.0,
      "step": 3210
    },
    {
      "entropy": 5.883116722106934,
      "epoch": 0.29163642960812775,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995439873608945,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1500185400247574,
      "num_tokens": 6663702.0,
      "step": 3215
    },
    {
      "entropy": 5.808498239517212,
      "epoch": 0.2920899854862119,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995419260574196,
      "loss": 5.684,
      "mean_token_accuracy": 0.1507360652089119,
      "num_tokens": 6673908.0,
      "step": 3220
    },
    {
      "entropy": 5.82448444366455,
      "epoch": 0.29254354136429606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995398601103603,
      "loss": 5.7033,
      "mean_token_accuracy": 0.16093184649944306,
      "num_tokens": 6683237.0,
      "step": 3225
    },
    {
      "entropy": 5.836985397338867,
      "epoch": 0.29299709724238027,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995377895197597,
      "loss": 5.6351,
      "mean_token_accuracy": 0.1617561772465706,
      "num_tokens": 6693451.0,
      "step": 3230
    },
    {
      "entropy": 5.736857318878174,
      "epoch": 0.2934506531204644,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995357142856602,
      "loss": 5.5948,
      "mean_token_accuracy": 0.16023087352514268,
      "num_tokens": 6703479.0,
      "step": 3235
    },
    {
      "entropy": 5.751489543914795,
      "epoch": 0.29390420899854863,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499533634408105,
      "loss": 5.5968,
      "mean_token_accuracy": 0.16008537858724595,
      "num_tokens": 6715321.0,
      "step": 3240
    },
    {
      "entropy": 5.791557550430298,
      "epoch": 0.2943577648766328,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499531549887137,
      "loss": 5.6655,
      "mean_token_accuracy": 0.1540275439620018,
      "num_tokens": 6725541.0,
      "step": 3245
    },
    {
      "entropy": 5.8105682849884035,
      "epoch": 0.294811320754717,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995294607227993,
      "loss": 5.7043,
      "mean_token_accuracy": 0.16713897138834,
      "num_tokens": 6735165.0,
      "step": 3250
    },
    {
      "entropy": 5.884285402297974,
      "epoch": 0.29526487663280115,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995273669151351,
      "loss": 5.701,
      "mean_token_accuracy": 0.16163005083799362,
      "num_tokens": 6744245.0,
      "step": 3255
    },
    {
      "entropy": 5.83087010383606,
      "epoch": 0.29571843251088537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499525268464188,
      "loss": 5.6645,
      "mean_token_accuracy": 0.16589844822883607,
      "num_tokens": 6754894.0,
      "step": 3260
    },
    {
      "entropy": 5.85270414352417,
      "epoch": 0.2961719883889695,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995231653700009,
      "loss": 5.7307,
      "mean_token_accuracy": 0.15026739612221718,
      "num_tokens": 6764442.0,
      "step": 3265
    },
    {
      "entropy": 5.814293384552002,
      "epoch": 0.2966255442670537,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995210576326176,
      "loss": 5.6807,
      "mean_token_accuracy": 0.1561569720506668,
      "num_tokens": 6775054.0,
      "step": 3270
    },
    {
      "entropy": 5.85151162147522,
      "epoch": 0.2970791001451379,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995189452520816,
      "loss": 5.6647,
      "mean_token_accuracy": 0.1682630956172943,
      "num_tokens": 6784825.0,
      "step": 3275
    },
    {
      "entropy": 5.8156262874603275,
      "epoch": 0.29753265602322204,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995168282284366,
      "loss": 5.7081,
      "mean_token_accuracy": 0.16142811328172685,
      "num_tokens": 6795928.0,
      "step": 3280
    },
    {
      "entropy": 5.77383861541748,
      "epoch": 0.29798621190130625,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995147065617263,
      "loss": 5.6295,
      "mean_token_accuracy": 0.16005017310380937,
      "num_tokens": 6805456.0,
      "step": 3285
    },
    {
      "entropy": 5.866624546051026,
      "epoch": 0.2984397677793904,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995125802519946,
      "loss": 5.7057,
      "mean_token_accuracy": 0.1569285973906517,
      "num_tokens": 6816215.0,
      "step": 3290
    },
    {
      "entropy": 5.855666875839233,
      "epoch": 0.2988933236574746,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995104492992857,
      "loss": 5.7427,
      "mean_token_accuracy": 0.15623879730701445,
      "num_tokens": 6826545.0,
      "step": 3295
    },
    {
      "entropy": 5.702891540527344,
      "epoch": 0.2993468795355588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995083137036434,
      "loss": 5.6479,
      "mean_token_accuracy": 0.16189349293708802,
      "num_tokens": 6836184.0,
      "step": 3300
    },
    {
      "entropy": 5.938313961029053,
      "epoch": 0.299800435413643,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995061734651121,
      "loss": 5.7388,
      "mean_token_accuracy": 0.15550311654806137,
      "num_tokens": 6846646.0,
      "step": 3305
    },
    {
      "entropy": 5.832134151458741,
      "epoch": 0.30025399129172714,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995040285837358,
      "loss": 5.6397,
      "mean_token_accuracy": 0.16627662777900695,
      "num_tokens": 6857597.0,
      "step": 3310
    },
    {
      "entropy": 5.757911109924317,
      "epoch": 0.3007075471698113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995018790595589,
      "loss": 5.7228,
      "mean_token_accuracy": 0.15299862772226333,
      "num_tokens": 6867523.0,
      "step": 3315
    },
    {
      "entropy": 5.863043165206909,
      "epoch": 0.3011611030478955,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994997248926261,
      "loss": 5.8306,
      "mean_token_accuracy": 0.15482011437416077,
      "num_tokens": 6878190.0,
      "step": 3320
    },
    {
      "entropy": 5.831428050994873,
      "epoch": 0.30161465892597966,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994975660829817,
      "loss": 5.6124,
      "mean_token_accuracy": 0.17094137519598007,
      "num_tokens": 6888483.0,
      "step": 3325
    },
    {
      "entropy": 5.726930236816406,
      "epoch": 0.3020682148040639,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994954026306705,
      "loss": 5.611,
      "mean_token_accuracy": 0.16160572692751884,
      "num_tokens": 6898337.0,
      "step": 3330
    },
    {
      "entropy": 5.853848600387574,
      "epoch": 0.30252177068214803,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994932345357371,
      "loss": 5.6829,
      "mean_token_accuracy": 0.1558000400662422,
      "num_tokens": 6909131.0,
      "step": 3335
    },
    {
      "entropy": 5.797689771652221,
      "epoch": 0.30297532656023224,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994910617982264,
      "loss": 5.6578,
      "mean_token_accuracy": 0.15806320160627366,
      "num_tokens": 6919790.0,
      "step": 3340
    },
    {
      "entropy": 5.781959342956543,
      "epoch": 0.3034288824383164,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994888844181833,
      "loss": 5.673,
      "mean_token_accuracy": 0.16309484243392944,
      "num_tokens": 6930072.0,
      "step": 3345
    },
    {
      "entropy": 5.762999105453491,
      "epoch": 0.3038824383164006,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994867023956529,
      "loss": 5.5666,
      "mean_token_accuracy": 0.1615067183971405,
      "num_tokens": 6940646.0,
      "step": 3350
    },
    {
      "entropy": 5.712483549118042,
      "epoch": 0.30433599419448476,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994845157306803,
      "loss": 5.5689,
      "mean_token_accuracy": 0.16492611765861512,
      "num_tokens": 6951645.0,
      "step": 3355
    },
    {
      "entropy": 5.875040483474732,
      "epoch": 0.3047895500725689,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994823244233107,
      "loss": 5.7363,
      "mean_token_accuracy": 0.157509508728981,
      "num_tokens": 6963020.0,
      "step": 3360
    },
    {
      "entropy": 5.842069149017334,
      "epoch": 0.3052431059506531,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994801284735895,
      "loss": 5.6612,
      "mean_token_accuracy": 0.16126682311296464,
      "num_tokens": 6972895.0,
      "step": 3365
    },
    {
      "entropy": 5.881510400772095,
      "epoch": 0.3056966618287373,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994779278815619,
      "loss": 5.8137,
      "mean_token_accuracy": 0.14684042036533357,
      "num_tokens": 6985363.0,
      "step": 3370
    },
    {
      "entropy": 5.853707885742187,
      "epoch": 0.3061502177068215,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994757226472736,
      "loss": 5.7436,
      "mean_token_accuracy": 0.15787117779254914,
      "num_tokens": 6997040.0,
      "step": 3375
    },
    {
      "entropy": 5.848391389846801,
      "epoch": 0.30660377358490565,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994735127707702,
      "loss": 5.6733,
      "mean_token_accuracy": 0.16261067390441894,
      "num_tokens": 7007591.0,
      "step": 3380
    },
    {
      "entropy": 5.860173940658569,
      "epoch": 0.30705732946298986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994712982520972,
      "loss": 5.6754,
      "mean_token_accuracy": 0.16760361343622207,
      "num_tokens": 7017858.0,
      "step": 3385
    },
    {
      "entropy": 5.727178287506104,
      "epoch": 0.307510885341074,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994690790913008,
      "loss": 5.6346,
      "mean_token_accuracy": 0.1600530058145523,
      "num_tokens": 7027365.0,
      "step": 3390
    },
    {
      "entropy": 5.76168851852417,
      "epoch": 0.3079644412191582,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994668552884265,
      "loss": 5.6351,
      "mean_token_accuracy": 0.1617230698466301,
      "num_tokens": 7038728.0,
      "step": 3395
    },
    {
      "entropy": 5.881175231933594,
      "epoch": 0.3084179970972424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994646268435204,
      "loss": 5.5729,
      "mean_token_accuracy": 0.16643185764551163,
      "num_tokens": 7048736.0,
      "step": 3400
    },
    {
      "entropy": 5.75899338722229,
      "epoch": 0.30887155297532654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994623937566286,
      "loss": 5.7002,
      "mean_token_accuracy": 0.1575309082865715,
      "num_tokens": 7059205.0,
      "step": 3405
    },
    {
      "entropy": 5.799173402786255,
      "epoch": 0.30932510885341075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994601560277974,
      "loss": 5.6896,
      "mean_token_accuracy": 0.15942423939704894,
      "num_tokens": 7069513.0,
      "step": 3410
    },
    {
      "entropy": 5.790976524353027,
      "epoch": 0.3097786647314949,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499457913657073,
      "loss": 5.6225,
      "mean_token_accuracy": 0.15869100689888,
      "num_tokens": 7079940.0,
      "step": 3415
    },
    {
      "entropy": 5.831319952011109,
      "epoch": 0.3102322206095791,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994556666445016,
      "loss": 5.6193,
      "mean_token_accuracy": 0.16156373918056488,
      "num_tokens": 7090792.0,
      "step": 3420
    },
    {
      "entropy": 5.755025625228882,
      "epoch": 0.31068577648766327,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994534149901299,
      "loss": 5.577,
      "mean_token_accuracy": 0.16105773597955703,
      "num_tokens": 7101964.0,
      "step": 3425
    },
    {
      "entropy": 5.7506568908691404,
      "epoch": 0.3111393323657475,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994511586940044,
      "loss": 5.6442,
      "mean_token_accuracy": 0.15435996800661086,
      "num_tokens": 7113332.0,
      "step": 3430
    },
    {
      "entropy": 5.91636700630188,
      "epoch": 0.31159288824383163,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994488977561717,
      "loss": 5.7921,
      "mean_token_accuracy": 0.15468702465295792,
      "num_tokens": 7123481.0,
      "step": 3435
    },
    {
      "entropy": 5.73750057220459,
      "epoch": 0.31204644412191584,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994466321766786,
      "loss": 5.5585,
      "mean_token_accuracy": 0.1646214559674263,
      "num_tokens": 7133694.0,
      "step": 3440
    },
    {
      "entropy": 5.804030513763427,
      "epoch": 0.3125,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499444361955572,
      "loss": 5.6664,
      "mean_token_accuracy": 0.1619983956217766,
      "num_tokens": 7144491.0,
      "step": 3445
    },
    {
      "entropy": 5.945888328552246,
      "epoch": 0.31295355587808416,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994420870928988,
      "loss": 5.76,
      "mean_token_accuracy": 0.15487642139196395,
      "num_tokens": 7154716.0,
      "step": 3450
    },
    {
      "entropy": 5.8428285121917725,
      "epoch": 0.31340711175616837,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994398075887059,
      "loss": 5.7553,
      "mean_token_accuracy": 0.15003324151039124,
      "num_tokens": 7164258.0,
      "step": 3455
    },
    {
      "entropy": 5.82437686920166,
      "epoch": 0.3138606676342525,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994375234430407,
      "loss": 5.6108,
      "mean_token_accuracy": 0.15920650959014893,
      "num_tokens": 7175098.0,
      "step": 3460
    },
    {
      "entropy": 5.79690523147583,
      "epoch": 0.31431422351233673,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994352346559504,
      "loss": 5.6851,
      "mean_token_accuracy": 0.15969553738832473,
      "num_tokens": 7186480.0,
      "step": 3465
    },
    {
      "entropy": 5.710554122924805,
      "epoch": 0.3147677793904209,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994329412274821,
      "loss": 5.6593,
      "mean_token_accuracy": 0.16162224411964415,
      "num_tokens": 7197056.0,
      "step": 3470
    },
    {
      "entropy": 5.846120882034302,
      "epoch": 0.3152213352685051,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994306431576835,
      "loss": 5.7725,
      "mean_token_accuracy": 0.15470817685127258,
      "num_tokens": 7206040.0,
      "step": 3475
    },
    {
      "entropy": 5.931183242797852,
      "epoch": 0.31567489114658925,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499428340446602,
      "loss": 5.6142,
      "mean_token_accuracy": 0.16317909955978394,
      "num_tokens": 7216344.0,
      "step": 3480
    },
    {
      "entropy": 5.784151315689087,
      "epoch": 0.31612844702467346,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994260330942851,
      "loss": 5.696,
      "mean_token_accuracy": 0.15416476279497146,
      "num_tokens": 7226729.0,
      "step": 3485
    },
    {
      "entropy": 5.753327083587647,
      "epoch": 0.3165820029027576,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994237211007808,
      "loss": 5.6509,
      "mean_token_accuracy": 0.1596106082201004,
      "num_tokens": 7237616.0,
      "step": 3490
    },
    {
      "entropy": 5.826973915100098,
      "epoch": 0.3170355587808418,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994214044661367,
      "loss": 5.5765,
      "mean_token_accuracy": 0.16474129110574723,
      "num_tokens": 7247857.0,
      "step": 3495
    },
    {
      "entropy": 5.698789978027344,
      "epoch": 0.317489114658926,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994190831904007,
      "loss": 5.5503,
      "mean_token_accuracy": 0.16544652432203294,
      "num_tokens": 7257858.0,
      "step": 3500
    },
    {
      "entropy": 5.665239334106445,
      "epoch": 0.31794267053701014,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499416757273621,
      "loss": 5.5582,
      "mean_token_accuracy": 0.16463755071163177,
      "num_tokens": 7268440.0,
      "step": 3505
    },
    {
      "entropy": 5.83346905708313,
      "epoch": 0.31839622641509435,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994144267158454,
      "loss": 5.6371,
      "mean_token_accuracy": 0.16650597602128983,
      "num_tokens": 7278560.0,
      "step": 3510
    },
    {
      "entropy": 5.785221576690674,
      "epoch": 0.3188497822931785,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994120915171224,
      "loss": 5.6364,
      "mean_token_accuracy": 0.16446116715669631,
      "num_tokens": 7288422.0,
      "step": 3515
    },
    {
      "entropy": 5.840118741989135,
      "epoch": 0.3193033381712627,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994097516775002,
      "loss": 5.7106,
      "mean_token_accuracy": 0.1576638028025627,
      "num_tokens": 7298613.0,
      "step": 3520
    },
    {
      "entropy": 5.828368854522705,
      "epoch": 0.3197568940493469,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994074071970271,
      "loss": 5.7422,
      "mean_token_accuracy": 0.15291873514652252,
      "num_tokens": 7309509.0,
      "step": 3525
    },
    {
      "entropy": 5.860109329223633,
      "epoch": 0.3202104499274311,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994050580757515,
      "loss": 5.6445,
      "mean_token_accuracy": 0.15940439999103545,
      "num_tokens": 7319015.0,
      "step": 3530
    },
    {
      "entropy": 5.721933031082154,
      "epoch": 0.32066400580551524,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994027043137223,
      "loss": 5.61,
      "mean_token_accuracy": 0.1672150120139122,
      "num_tokens": 7329026.0,
      "step": 3535
    },
    {
      "entropy": 5.855372905731201,
      "epoch": 0.3211175616835994,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994003459109879,
      "loss": 5.7137,
      "mean_token_accuracy": 0.16032420545816423,
      "num_tokens": 7339638.0,
      "step": 3540
    },
    {
      "entropy": 5.837479209899902,
      "epoch": 0.3215711175616836,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993979828675972,
      "loss": 5.624,
      "mean_token_accuracy": 0.16279072165489197,
      "num_tokens": 7349979.0,
      "step": 3545
    },
    {
      "entropy": 5.702495336532593,
      "epoch": 0.32202467343976776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499395615183599,
      "loss": 5.5517,
      "mean_token_accuracy": 0.17009643316268921,
      "num_tokens": 7359398.0,
      "step": 3550
    },
    {
      "entropy": 5.7116045475006105,
      "epoch": 0.32247822931785197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993932428590422,
      "loss": 5.5945,
      "mean_token_accuracy": 0.15939887911081313,
      "num_tokens": 7370418.0,
      "step": 3555
    },
    {
      "entropy": 5.781202697753907,
      "epoch": 0.3229317851959361,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993908658939762,
      "loss": 5.6239,
      "mean_token_accuracy": 0.15674008578062057,
      "num_tokens": 7380461.0,
      "step": 3560
    },
    {
      "entropy": 5.814039325714111,
      "epoch": 0.32338534107402034,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993884842884496,
      "loss": 5.647,
      "mean_token_accuracy": 0.16048161089420318,
      "num_tokens": 7390646.0,
      "step": 3565
    },
    {
      "entropy": 5.786873054504395,
      "epoch": 0.3238388969521045,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993860980425121,
      "loss": 5.7203,
      "mean_token_accuracy": 0.1618356540799141,
      "num_tokens": 7402106.0,
      "step": 3570
    },
    {
      "entropy": 5.809912443161011,
      "epoch": 0.3242924528301887,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993837071562131,
      "loss": 5.7043,
      "mean_token_accuracy": 0.16131191849708557,
      "num_tokens": 7412936.0,
      "step": 3575
    },
    {
      "entropy": 5.8115393161773685,
      "epoch": 0.32474600870827286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993813116296017,
      "loss": 5.5708,
      "mean_token_accuracy": 0.16675995141267777,
      "num_tokens": 7422303.0,
      "step": 3580
    },
    {
      "entropy": 5.715788507461548,
      "epoch": 0.325199564586357,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993789114627275,
      "loss": 5.6511,
      "mean_token_accuracy": 0.16085287630558015,
      "num_tokens": 7432829.0,
      "step": 3585
    },
    {
      "entropy": 5.937649488449097,
      "epoch": 0.3256531204644412,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993765066556403,
      "loss": 5.6713,
      "mean_token_accuracy": 0.15823337584733962,
      "num_tokens": 7443218.0,
      "step": 3590
    },
    {
      "entropy": 5.773650455474853,
      "epoch": 0.3261066763425254,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993740972083899,
      "loss": 5.62,
      "mean_token_accuracy": 0.16021817028522492,
      "num_tokens": 7454531.0,
      "step": 3595
    },
    {
      "entropy": 5.790258073806763,
      "epoch": 0.3265602322206096,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993716831210259,
      "loss": 5.6799,
      "mean_token_accuracy": 0.16317007392644883,
      "num_tokens": 7464993.0,
      "step": 3600
    },
    {
      "entropy": 5.729147958755493,
      "epoch": 0.32701378809869375,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993692643935983,
      "loss": 5.7096,
      "mean_token_accuracy": 0.1584396131336689,
      "num_tokens": 7475755.0,
      "step": 3605
    },
    {
      "entropy": 5.771151685714722,
      "epoch": 0.32746734397677796,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993668410261571,
      "loss": 5.6945,
      "mean_token_accuracy": 0.16145843863487244,
      "num_tokens": 7486778.0,
      "step": 3610
    },
    {
      "entropy": 5.907090902328491,
      "epoch": 0.3279208998548621,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993644130187525,
      "loss": 5.6771,
      "mean_token_accuracy": 0.16191897690296173,
      "num_tokens": 7497439.0,
      "step": 3615
    },
    {
      "entropy": 5.734846353530884,
      "epoch": 0.3283744557329463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993619803714347,
      "loss": 5.7075,
      "mean_token_accuracy": 0.15616535544395446,
      "num_tokens": 7508274.0,
      "step": 3620
    },
    {
      "entropy": 5.782626390457153,
      "epoch": 0.3288280116110305,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993595430842539,
      "loss": 5.6874,
      "mean_token_accuracy": 0.16056278795003892,
      "num_tokens": 7519315.0,
      "step": 3625
    },
    {
      "entropy": 5.901585865020752,
      "epoch": 0.32928156748911463,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993571011572606,
      "loss": 5.6808,
      "mean_token_accuracy": 0.1630665123462677,
      "num_tokens": 7529910.0,
      "step": 3630
    },
    {
      "entropy": 5.800387811660767,
      "epoch": 0.32973512336719885,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993546545905053,
      "loss": 5.602,
      "mean_token_accuracy": 0.16489686965942382,
      "num_tokens": 7538966.0,
      "step": 3635
    },
    {
      "entropy": 5.775709533691407,
      "epoch": 0.330188679245283,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993522033840386,
      "loss": 5.6374,
      "mean_token_accuracy": 0.15789183378219604,
      "num_tokens": 7549683.0,
      "step": 3640
    },
    {
      "entropy": 5.700624227523804,
      "epoch": 0.3306422351233672,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499349747537911,
      "loss": 5.6191,
      "mean_token_accuracy": 0.1624668598175049,
      "num_tokens": 7560629.0,
      "step": 3645
    },
    {
      "entropy": 5.806013774871826,
      "epoch": 0.33109579100145137,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993472870521736,
      "loss": 5.5646,
      "mean_token_accuracy": 0.17270171344280244,
      "num_tokens": 7570613.0,
      "step": 3650
    },
    {
      "entropy": 5.79381422996521,
      "epoch": 0.3315493468795356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993448219268771,
      "loss": 5.5499,
      "mean_token_accuracy": 0.16173658668994903,
      "num_tokens": 7580640.0,
      "step": 3655
    },
    {
      "entropy": 5.763698101043701,
      "epoch": 0.33200290275761973,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993423521620726,
      "loss": 5.6381,
      "mean_token_accuracy": 0.16688979268074036,
      "num_tokens": 7590164.0,
      "step": 3660
    },
    {
      "entropy": 5.823714399337769,
      "epoch": 0.33245645863570394,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993398777578109,
      "loss": 5.6959,
      "mean_token_accuracy": 0.16008617281913756,
      "num_tokens": 7600173.0,
      "step": 3665
    },
    {
      "entropy": 5.71161642074585,
      "epoch": 0.3329100145137881,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993373987141435,
      "loss": 5.4841,
      "mean_token_accuracy": 0.17095217257738113,
      "num_tokens": 7609958.0,
      "step": 3670
    },
    {
      "entropy": 5.813071250915527,
      "epoch": 0.33336357039187225,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993349150311214,
      "loss": 5.6411,
      "mean_token_accuracy": 0.157687945663929,
      "num_tokens": 7620237.0,
      "step": 3675
    },
    {
      "entropy": 5.699023199081421,
      "epoch": 0.33381712626995647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993324267087962,
      "loss": 5.5406,
      "mean_token_accuracy": 0.1682781383395195,
      "num_tokens": 7629886.0,
      "step": 3680
    },
    {
      "entropy": 5.728935241699219,
      "epoch": 0.3342706821480406,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993299337472192,
      "loss": 5.6095,
      "mean_token_accuracy": 0.16654768735170364,
      "num_tokens": 7640057.0,
      "step": 3685
    },
    {
      "entropy": 5.7182149410247805,
      "epoch": 0.33472423802612483,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499327436146442,
      "loss": 5.5532,
      "mean_token_accuracy": 0.16529571264982224,
      "num_tokens": 7649766.0,
      "step": 3690
    },
    {
      "entropy": 5.769032526016235,
      "epoch": 0.335177793904209,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993249339065162,
      "loss": 5.6399,
      "mean_token_accuracy": 0.16574999690055847,
      "num_tokens": 7660890.0,
      "step": 3695
    },
    {
      "entropy": 5.773209142684936,
      "epoch": 0.3356313497822932,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993224270274937,
      "loss": 5.6582,
      "mean_token_accuracy": 0.1607873558998108,
      "num_tokens": 7670737.0,
      "step": 3700
    },
    {
      "entropy": 5.7178552627563475,
      "epoch": 0.33608490566037735,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993199155094263,
      "loss": 5.6067,
      "mean_token_accuracy": 0.16417943835258483,
      "num_tokens": 7681407.0,
      "step": 3705
    },
    {
      "entropy": 5.768005990982056,
      "epoch": 0.33653846153846156,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993173993523658,
      "loss": 5.5356,
      "mean_token_accuracy": 0.17343078404664994,
      "num_tokens": 7691758.0,
      "step": 3710
    },
    {
      "entropy": 5.699445009231567,
      "epoch": 0.3369920174165457,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993148785563643,
      "loss": 5.5679,
      "mean_token_accuracy": 0.16385959088802338,
      "num_tokens": 7702934.0,
      "step": 3715
    },
    {
      "entropy": 5.71638708114624,
      "epoch": 0.3374455732946299,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499312353121474,
      "loss": 5.5705,
      "mean_token_accuracy": 0.16445922404527663,
      "num_tokens": 7713001.0,
      "step": 3720
    },
    {
      "entropy": 5.802838611602783,
      "epoch": 0.3378991291727141,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993098230477471,
      "loss": 5.6211,
      "mean_token_accuracy": 0.1572564959526062,
      "num_tokens": 7724005.0,
      "step": 3725
    },
    {
      "entropy": 5.785116243362427,
      "epoch": 0.33835268505079824,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993072883352358,
      "loss": 5.5802,
      "mean_token_accuracy": 0.16274530738592147,
      "num_tokens": 7733743.0,
      "step": 3730
    },
    {
      "entropy": 5.802304887771607,
      "epoch": 0.33880624092888245,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993047489839927,
      "loss": 5.7695,
      "mean_token_accuracy": 0.1579284265637398,
      "num_tokens": 7743762.0,
      "step": 3735
    },
    {
      "entropy": 5.850232982635498,
      "epoch": 0.3392597968069666,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993022049940702,
      "loss": 5.655,
      "mean_token_accuracy": 0.16184027045965194,
      "num_tokens": 7753758.0,
      "step": 3740
    },
    {
      "entropy": 5.739925861358643,
      "epoch": 0.3397133526850508,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992996563655209,
      "loss": 5.6052,
      "mean_token_accuracy": 0.163077212870121,
      "num_tokens": 7764692.0,
      "step": 3745
    },
    {
      "entropy": 5.742031097412109,
      "epoch": 0.340166908563135,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992971030983975,
      "loss": 5.644,
      "mean_token_accuracy": 0.15631235092878343,
      "num_tokens": 7776892.0,
      "step": 3750
    },
    {
      "entropy": 5.875276947021485,
      "epoch": 0.3406204644412192,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992945451927531,
      "loss": 5.6769,
      "mean_token_accuracy": 0.16058121621608734,
      "num_tokens": 7787143.0,
      "step": 3755
    },
    {
      "entropy": 5.723575305938721,
      "epoch": 0.34107402031930334,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992919826486402,
      "loss": 5.5967,
      "mean_token_accuracy": 0.16272515058517456,
      "num_tokens": 7798537.0,
      "step": 3760
    },
    {
      "entropy": 5.713321161270142,
      "epoch": 0.3415275761973875,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992894154661118,
      "loss": 5.5709,
      "mean_token_accuracy": 0.16793359965085983,
      "num_tokens": 7808434.0,
      "step": 3765
    },
    {
      "entropy": 5.79578595161438,
      "epoch": 0.3419811320754717,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992868436452213,
      "loss": 5.6818,
      "mean_token_accuracy": 0.16078411191701888,
      "num_tokens": 7819592.0,
      "step": 3770
    },
    {
      "entropy": 5.809304237365723,
      "epoch": 0.34243468795355586,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992842671860217,
      "loss": 5.5643,
      "mean_token_accuracy": 0.16715734750032424,
      "num_tokens": 7830602.0,
      "step": 3775
    },
    {
      "entropy": 5.786276483535767,
      "epoch": 0.34288824383164007,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992816860885662,
      "loss": 5.7212,
      "mean_token_accuracy": 0.15924838036298752,
      "num_tokens": 7840850.0,
      "step": 3780
    },
    {
      "entropy": 5.884521865844727,
      "epoch": 0.3433417997097242,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992791003529084,
      "loss": 5.6558,
      "mean_token_accuracy": 0.15685309320688248,
      "num_tokens": 7852371.0,
      "step": 3785
    },
    {
      "entropy": 5.715097904205322,
      "epoch": 0.34379535558780844,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992765099791017,
      "loss": 5.6818,
      "mean_token_accuracy": 0.15666263401508332,
      "num_tokens": 7863290.0,
      "step": 3790
    },
    {
      "entropy": 5.71993088722229,
      "epoch": 0.3442489114658926,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992739149671995,
      "loss": 5.509,
      "mean_token_accuracy": 0.16685899645090102,
      "num_tokens": 7873927.0,
      "step": 3795
    },
    {
      "entropy": 5.733199405670166,
      "epoch": 0.3447024673439768,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992713153172556,
      "loss": 5.5502,
      "mean_token_accuracy": 0.16382715702056885,
      "num_tokens": 7884977.0,
      "step": 3800
    },
    {
      "entropy": 5.699240398406983,
      "epoch": 0.34515602322206096,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992687110293238,
      "loss": 5.6304,
      "mean_token_accuracy": 0.16769205555319786,
      "num_tokens": 7894620.0,
      "step": 3805
    },
    {
      "entropy": 5.719654321670532,
      "epoch": 0.3456095791001451,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992661021034579,
      "loss": 5.5507,
      "mean_token_accuracy": 0.16906091570854187,
      "num_tokens": 7904541.0,
      "step": 3810
    },
    {
      "entropy": 5.726424789428711,
      "epoch": 0.3460631349782293,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992634885397118,
      "loss": 5.5794,
      "mean_token_accuracy": 0.1659620225429535,
      "num_tokens": 7915435.0,
      "step": 3815
    },
    {
      "entropy": 5.722615718841553,
      "epoch": 0.3465166908563135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992608703381396,
      "loss": 5.5616,
      "mean_token_accuracy": 0.16741435825824738,
      "num_tokens": 7925870.0,
      "step": 3820
    },
    {
      "entropy": 5.722995281219482,
      "epoch": 0.3469702467343977,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992582474987955,
      "loss": 5.5639,
      "mean_token_accuracy": 0.16274259686470033,
      "num_tokens": 7937046.0,
      "step": 3825
    },
    {
      "entropy": 5.75324854850769,
      "epoch": 0.34742380261248185,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992556200217337,
      "loss": 5.5261,
      "mean_token_accuracy": 0.1641123980283737,
      "num_tokens": 7947150.0,
      "step": 3830
    },
    {
      "entropy": 5.795271539688111,
      "epoch": 0.34787735849056606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992529879070085,
      "loss": 5.618,
      "mean_token_accuracy": 0.16771930009126662,
      "num_tokens": 7956815.0,
      "step": 3835
    },
    {
      "entropy": 5.789908409118652,
      "epoch": 0.3483309143686502,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992503511546745,
      "loss": 5.6657,
      "mean_token_accuracy": 0.15945106595754624,
      "num_tokens": 7968114.0,
      "step": 3840
    },
    {
      "entropy": 5.748790836334228,
      "epoch": 0.3487844702467344,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992477097647859,
      "loss": 5.5656,
      "mean_token_accuracy": 0.1627722293138504,
      "num_tokens": 7977840.0,
      "step": 3845
    },
    {
      "entropy": 5.814347887039185,
      "epoch": 0.3492380261248186,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992450637373976,
      "loss": 5.6907,
      "mean_token_accuracy": 0.16484753638505936,
      "num_tokens": 7988663.0,
      "step": 3850
    },
    {
      "entropy": 5.681796550750732,
      "epoch": 0.34969158200290273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992424130725642,
      "loss": 5.4999,
      "mean_token_accuracy": 0.16489350348711013,
      "num_tokens": 7999693.0,
      "step": 3855
    },
    {
      "entropy": 5.772833013534546,
      "epoch": 0.35014513788098695,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992397577703406,
      "loss": 5.6747,
      "mean_token_accuracy": 0.1532568469643593,
      "num_tokens": 8010154.0,
      "step": 3860
    },
    {
      "entropy": 5.825438737869263,
      "epoch": 0.3505986937590711,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992370978307816,
      "loss": 5.6536,
      "mean_token_accuracy": 0.1622571349143982,
      "num_tokens": 8020921.0,
      "step": 3865
    },
    {
      "entropy": 5.764519309997558,
      "epoch": 0.3510522496371553,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992344332539424,
      "loss": 5.5811,
      "mean_token_accuracy": 0.16872124671936034,
      "num_tokens": 8031404.0,
      "step": 3870
    },
    {
      "entropy": 5.684741640090943,
      "epoch": 0.35150580551523947,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992317640398779,
      "loss": 5.585,
      "mean_token_accuracy": 0.1686838299036026,
      "num_tokens": 8041310.0,
      "step": 3875
    },
    {
      "entropy": 5.710890054702759,
      "epoch": 0.3519593613933237,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992290901886434,
      "loss": 5.57,
      "mean_token_accuracy": 0.1684218779206276,
      "num_tokens": 8051054.0,
      "step": 3880
    },
    {
      "entropy": 5.744958877563477,
      "epoch": 0.35241291727140783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499226411700294,
      "loss": 5.5578,
      "mean_token_accuracy": 0.16676690876483918,
      "num_tokens": 8061063.0,
      "step": 3885
    },
    {
      "entropy": 5.69763445854187,
      "epoch": 0.35286647314949204,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992237285748855,
      "loss": 5.5841,
      "mean_token_accuracy": 0.16478407829999925,
      "num_tokens": 8071715.0,
      "step": 3890
    },
    {
      "entropy": 5.80994873046875,
      "epoch": 0.3533200290275762,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992210408124731,
      "loss": 5.6496,
      "mean_token_accuracy": 0.15725049376487732,
      "num_tokens": 8082679.0,
      "step": 3895
    },
    {
      "entropy": 5.752315807342529,
      "epoch": 0.35377358490566035,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992183484131123,
      "loss": 5.6808,
      "mean_token_accuracy": 0.16136541217565536,
      "num_tokens": 8093785.0,
      "step": 3900
    },
    {
      "entropy": 5.71539158821106,
      "epoch": 0.35422714078374457,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992156513768591,
      "loss": 5.5685,
      "mean_token_accuracy": 0.16779497861862183,
      "num_tokens": 8105822.0,
      "step": 3905
    },
    {
      "entropy": 5.709498691558838,
      "epoch": 0.3546806966618287,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499212949703769,
      "loss": 5.486,
      "mean_token_accuracy": 0.1679369956254959,
      "num_tokens": 8115521.0,
      "step": 3910
    },
    {
      "entropy": 5.726305246353149,
      "epoch": 0.35513425253991293,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992102433938981,
      "loss": 5.5783,
      "mean_token_accuracy": 0.16780869513750077,
      "num_tokens": 8125524.0,
      "step": 3915
    },
    {
      "entropy": 5.712926054000855,
      "epoch": 0.3555878084179971,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992075324473021,
      "loss": 5.482,
      "mean_token_accuracy": 0.17744722813367844,
      "num_tokens": 8135552.0,
      "step": 3920
    },
    {
      "entropy": 5.780481147766113,
      "epoch": 0.3560413642960813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992048168640371,
      "loss": 5.6135,
      "mean_token_accuracy": 0.16508189737796783,
      "num_tokens": 8145617.0,
      "step": 3925
    },
    {
      "entropy": 5.725088024139405,
      "epoch": 0.35649492017416545,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992020966441595,
      "loss": 5.6251,
      "mean_token_accuracy": 0.16034330129623414,
      "num_tokens": 8156729.0,
      "step": 3930
    },
    {
      "entropy": 5.764310026168824,
      "epoch": 0.35694847605224966,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991993717877254,
      "loss": 5.6076,
      "mean_token_accuracy": 0.16174564361572266,
      "num_tokens": 8167637.0,
      "step": 3935
    },
    {
      "entropy": 5.720431089401245,
      "epoch": 0.3574020319303338,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991966422947911,
      "loss": 5.5852,
      "mean_token_accuracy": 0.1702892929315567,
      "num_tokens": 8177354.0,
      "step": 3940
    },
    {
      "entropy": 5.718207550048828,
      "epoch": 0.357855587808418,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991939081654132,
      "loss": 5.5502,
      "mean_token_accuracy": 0.1746809646487236,
      "num_tokens": 8187580.0,
      "step": 3945
    },
    {
      "entropy": 5.708486032485962,
      "epoch": 0.3583091436865022,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499191169399648,
      "loss": 5.6497,
      "mean_token_accuracy": 0.1604743108153343,
      "num_tokens": 8199382.0,
      "step": 3950
    },
    {
      "entropy": 5.712984323501587,
      "epoch": 0.35876269956458634,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991884259975525,
      "loss": 5.5404,
      "mean_token_accuracy": 0.16475948542356492,
      "num_tokens": 8208484.0,
      "step": 3955
    },
    {
      "entropy": 5.727371263504028,
      "epoch": 0.35921625544267055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499185677959183,
      "loss": 5.5838,
      "mean_token_accuracy": 0.16461657136678695,
      "num_tokens": 8219056.0,
      "step": 3960
    },
    {
      "entropy": 5.731614685058593,
      "epoch": 0.3596698113207547,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991829252845966,
      "loss": 5.6047,
      "mean_token_accuracy": 0.16140197962522507,
      "num_tokens": 8230149.0,
      "step": 3965
    },
    {
      "entropy": 5.7445838928222654,
      "epoch": 0.3601233671988389,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991801679738502,
      "loss": 5.6652,
      "mean_token_accuracy": 0.15284715443849564,
      "num_tokens": 8240803.0,
      "step": 3970
    },
    {
      "entropy": 5.766950941085815,
      "epoch": 0.3605769230769231,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991774060270008,
      "loss": 5.6201,
      "mean_token_accuracy": 0.15810701847076417,
      "num_tokens": 8251563.0,
      "step": 3975
    },
    {
      "entropy": 5.701673126220703,
      "epoch": 0.3610304789550073,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991746394441055,
      "loss": 5.5522,
      "mean_token_accuracy": 0.16369420439004898,
      "num_tokens": 8261342.0,
      "step": 3980
    },
    {
      "entropy": 5.753665637969971,
      "epoch": 0.36148403483309144,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991718682252216,
      "loss": 5.6681,
      "mean_token_accuracy": 0.15975360572338104,
      "num_tokens": 8271540.0,
      "step": 3985
    },
    {
      "entropy": 5.782139301300049,
      "epoch": 0.3619375907111756,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991690923704063,
      "loss": 5.58,
      "mean_token_accuracy": 0.16486785262823106,
      "num_tokens": 8280593.0,
      "step": 3990
    },
    {
      "entropy": 5.742940139770508,
      "epoch": 0.3623911465892598,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991663118797171,
      "loss": 5.6104,
      "mean_token_accuracy": 0.16217703521251678,
      "num_tokens": 8290126.0,
      "step": 3995
    },
    {
      "entropy": 5.747898292541504,
      "epoch": 0.36284470246734396,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991635267532113,
      "loss": 5.5394,
      "mean_token_accuracy": 0.16359646618366241,
      "num_tokens": 8301416.0,
      "step": 4000
    },
    {
      "entropy": 5.775230884552002,
      "epoch": 0.36329825834542817,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991607369909468,
      "loss": 5.6135,
      "mean_token_accuracy": 0.1605329006910324,
      "num_tokens": 8311034.0,
      "step": 4005
    },
    {
      "entropy": 5.757994508743286,
      "epoch": 0.3637518142235123,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991579425929811,
      "loss": 5.6266,
      "mean_token_accuracy": 0.16281366795301438,
      "num_tokens": 8322203.0,
      "step": 4010
    },
    {
      "entropy": 5.7276918411254885,
      "epoch": 0.36420537010159654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499155143559372,
      "loss": 5.5117,
      "mean_token_accuracy": 0.16838518381118775,
      "num_tokens": 8330981.0,
      "step": 4015
    },
    {
      "entropy": 5.655632972717285,
      "epoch": 0.3646589259796807,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991523398901773,
      "loss": 5.5437,
      "mean_token_accuracy": 0.16577242016792298,
      "num_tokens": 8341294.0,
      "step": 4020
    },
    {
      "entropy": 5.8728711128234865,
      "epoch": 0.3651124818577649,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991495315854553,
      "loss": 5.6921,
      "mean_token_accuracy": 0.15758731961250305,
      "num_tokens": 8351824.0,
      "step": 4025
    },
    {
      "entropy": 5.734429454803466,
      "epoch": 0.36556603773584906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991467186452637,
      "loss": 5.5595,
      "mean_token_accuracy": 0.16328836232423782,
      "num_tokens": 8361975.0,
      "step": 4030
    },
    {
      "entropy": 5.646515369415283,
      "epoch": 0.3660195936139332,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991439010696609,
      "loss": 5.5053,
      "mean_token_accuracy": 0.16924880295991898,
      "num_tokens": 8372145.0,
      "step": 4035
    },
    {
      "entropy": 5.743934106826782,
      "epoch": 0.3664731494920174,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991410788587051,
      "loss": 5.6034,
      "mean_token_accuracy": 0.1616530880331993,
      "num_tokens": 8382542.0,
      "step": 4040
    },
    {
      "entropy": 5.771177625656128,
      "epoch": 0.3669267053701016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991382520124547,
      "loss": 5.5942,
      "mean_token_accuracy": 0.16173005402088164,
      "num_tokens": 8393450.0,
      "step": 4045
    },
    {
      "entropy": 5.7502647876739506,
      "epoch": 0.3673802612481858,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991354205309681,
      "loss": 5.5407,
      "mean_token_accuracy": 0.1637318655848503,
      "num_tokens": 8404929.0,
      "step": 4050
    },
    {
      "entropy": 5.762520360946655,
      "epoch": 0.36783381712626995,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499132584414304,
      "loss": 5.5531,
      "mean_token_accuracy": 0.1654103085398674,
      "num_tokens": 8415703.0,
      "step": 4055
    },
    {
      "entropy": 5.728594827651977,
      "epoch": 0.36828737300435416,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991297436625207,
      "loss": 5.6347,
      "mean_token_accuracy": 0.16382609456777572,
      "num_tokens": 8426220.0,
      "step": 4060
    },
    {
      "entropy": 5.665254354476929,
      "epoch": 0.3687409288824383,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991268982756774,
      "loss": 5.5135,
      "mean_token_accuracy": 0.16983363032341003,
      "num_tokens": 8436220.0,
      "step": 4065
    },
    {
      "entropy": 5.824220180511475,
      "epoch": 0.3691944847605225,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991240482538327,
      "loss": 5.6576,
      "mean_token_accuracy": 0.1528759405016899,
      "num_tokens": 8446724.0,
      "step": 4070
    },
    {
      "entropy": 5.739574432373047,
      "epoch": 0.3696480406386067,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991211935970455,
      "loss": 5.5215,
      "mean_token_accuracy": 0.16571784317493438,
      "num_tokens": 8457270.0,
      "step": 4075
    },
    {
      "entropy": 5.698843717575073,
      "epoch": 0.37010159651669083,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991183343053748,
      "loss": 5.5808,
      "mean_token_accuracy": 0.16122987121343613,
      "num_tokens": 8468137.0,
      "step": 4080
    },
    {
      "entropy": 5.737043762207032,
      "epoch": 0.37055515239477504,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049911547037888,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16846096962690355,
      "num_tokens": 8477574.0,
      "step": 4085
    },
    {
      "entropy": 5.718803071975708,
      "epoch": 0.3710087082728592,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991126018176202,
      "loss": 5.6211,
      "mean_token_accuracy": 0.1616816744208336,
      "num_tokens": 8487268.0,
      "step": 4090
    },
    {
      "entropy": 5.8145722389221195,
      "epoch": 0.3714622641509434,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991097286216545,
      "loss": 5.6662,
      "mean_token_accuracy": 0.168794247508049,
      "num_tokens": 8497785.0,
      "step": 4095
    },
    {
      "entropy": 5.753116083145142,
      "epoch": 0.37191582002902757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991068507910427,
      "loss": 5.5533,
      "mean_token_accuracy": 0.16418525129556655,
      "num_tokens": 8508169.0,
      "step": 4100
    },
    {
      "entropy": 5.690785503387451,
      "epoch": 0.3723693759071118,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499103968325844,
      "loss": 5.522,
      "mean_token_accuracy": 0.16396052837371827,
      "num_tokens": 8519201.0,
      "step": 4105
    },
    {
      "entropy": 5.7371025562286375,
      "epoch": 0.37282293178519593,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991010812261181,
      "loss": 5.6384,
      "mean_token_accuracy": 0.1556764304637909,
      "num_tokens": 8531093.0,
      "step": 4110
    },
    {
      "entropy": 5.665953826904297,
      "epoch": 0.37327648766328014,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990981894919248,
      "loss": 5.4913,
      "mean_token_accuracy": 0.16958004534244536,
      "num_tokens": 8541730.0,
      "step": 4115
    },
    {
      "entropy": 5.693306016921997,
      "epoch": 0.3737300435413643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990952931233238,
      "loss": 5.5646,
      "mean_token_accuracy": 0.16480782628059387,
      "num_tokens": 8552570.0,
      "step": 4120
    },
    {
      "entropy": 5.754336500167847,
      "epoch": 0.37418359941944845,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499092392120375,
      "loss": 5.4628,
      "mean_token_accuracy": 0.17125755399465561,
      "num_tokens": 8562340.0,
      "step": 4125
    },
    {
      "entropy": 5.735305070877075,
      "epoch": 0.37463715529753266,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990894864831385,
      "loss": 5.6707,
      "mean_token_accuracy": 0.15796474665403365,
      "num_tokens": 8573240.0,
      "step": 4130
    },
    {
      "entropy": 5.769050550460816,
      "epoch": 0.3750907111756168,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990865762116743,
      "loss": 5.5299,
      "mean_token_accuracy": 0.16909209936857222,
      "num_tokens": 8583194.0,
      "step": 4135
    },
    {
      "entropy": 5.719068336486816,
      "epoch": 0.37554426705370103,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990836613060426,
      "loss": 5.5962,
      "mean_token_accuracy": 0.1610783204436302,
      "num_tokens": 8593603.0,
      "step": 4140
    },
    {
      "entropy": 5.664818382263183,
      "epoch": 0.3759978229317852,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990807417663036,
      "loss": 5.5998,
      "mean_token_accuracy": 0.16227242201566697,
      "num_tokens": 8603176.0,
      "step": 4145
    },
    {
      "entropy": 5.7971210956573485,
      "epoch": 0.3764513788098694,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004990778175925179,
      "loss": 5.6154,
      "mean_token_accuracy": 0.16660853177309037,
      "num_tokens": 8614801.0,
      "step": 4150
    },
    {
      "entropy": 5.768546104431152,
      "epoch": 0.37690493468795355,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990748887847457,
      "loss": 5.6418,
      "mean_token_accuracy": 0.16194847822189332,
      "num_tokens": 8624999.0,
      "step": 4155
    },
    {
      "entropy": 5.667234706878662,
      "epoch": 0.37735849056603776,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990719553430477,
      "loss": 5.4948,
      "mean_token_accuracy": 0.17510903775691986,
      "num_tokens": 8635358.0,
      "step": 4160
    },
    {
      "entropy": 5.749764490127563,
      "epoch": 0.3778120464441219,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990690172674847,
      "loss": 5.5679,
      "mean_token_accuracy": 0.1612323611974716,
      "num_tokens": 8645430.0,
      "step": 4165
    },
    {
      "entropy": 5.726269006729126,
      "epoch": 0.3782656023222061,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990660745581172,
      "loss": 5.5234,
      "mean_token_accuracy": 0.17513987571001052,
      "num_tokens": 8655115.0,
      "step": 4170
    },
    {
      "entropy": 5.6731971263885494,
      "epoch": 0.3787191582002903,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990631272150062,
      "loss": 5.5279,
      "mean_token_accuracy": 0.17234957814216614,
      "num_tokens": 8665391.0,
      "step": 4175
    },
    {
      "entropy": 5.764857435226441,
      "epoch": 0.37917271407837444,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004990601752382127,
      "loss": 5.6196,
      "mean_token_accuracy": 0.1626276418566704,
      "num_tokens": 8676819.0,
      "step": 4180
    },
    {
      "entropy": 5.78800048828125,
      "epoch": 0.37962626995645865,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990572186277975,
      "loss": 5.5433,
      "mean_token_accuracy": 0.16920216530561447,
      "num_tokens": 8687052.0,
      "step": 4185
    },
    {
      "entropy": 5.716868305206299,
      "epoch": 0.3800798258345428,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990542573838221,
      "loss": 5.6207,
      "mean_token_accuracy": 0.16125022917985915,
      "num_tokens": 8697756.0,
      "step": 4190
    },
    {
      "entropy": 5.616197013854981,
      "epoch": 0.380533381712627,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990512915063475,
      "loss": 5.5242,
      "mean_token_accuracy": 0.17394591718912125,
      "num_tokens": 8708388.0,
      "step": 4195
    },
    {
      "entropy": 5.806944036483765,
      "epoch": 0.38098693759071117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990483209954352,
      "loss": 5.6288,
      "mean_token_accuracy": 0.15857603251934052,
      "num_tokens": 8719475.0,
      "step": 4200
    },
    {
      "entropy": 5.71453013420105,
      "epoch": 0.3814404934687954,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990453458511464,
      "loss": 5.477,
      "mean_token_accuracy": 0.17523704022169112,
      "num_tokens": 8729395.0,
      "step": 4205
    },
    {
      "entropy": 5.69320387840271,
      "epoch": 0.38189404934687954,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990423660735428,
      "loss": 5.5367,
      "mean_token_accuracy": 0.17120707631111146,
      "num_tokens": 8740351.0,
      "step": 4210
    },
    {
      "entropy": 5.747463369369507,
      "epoch": 0.3823476052249637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499039381662686,
      "loss": 5.5519,
      "mean_token_accuracy": 0.16874152719974517,
      "num_tokens": 8749260.0,
      "step": 4215
    },
    {
      "entropy": 5.784005641937256,
      "epoch": 0.3828011611030479,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990363926186377,
      "loss": 5.6472,
      "mean_token_accuracy": 0.1575682356953621,
      "num_tokens": 8759604.0,
      "step": 4220
    },
    {
      "entropy": 5.778999853134155,
      "epoch": 0.38325471698113206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990333989414597,
      "loss": 5.6582,
      "mean_token_accuracy": 0.1573771618306637,
      "num_tokens": 8769815.0,
      "step": 4225
    },
    {
      "entropy": 5.677911949157715,
      "epoch": 0.38370827285921627,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990304006312138,
      "loss": 5.5816,
      "mean_token_accuracy": 0.16669487804174424,
      "num_tokens": 8780557.0,
      "step": 4230
    },
    {
      "entropy": 5.731473064422607,
      "epoch": 0.3841618287373004,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990273976879622,
      "loss": 5.5429,
      "mean_token_accuracy": 0.17356031835079194,
      "num_tokens": 8790993.0,
      "step": 4235
    },
    {
      "entropy": 5.731623315811158,
      "epoch": 0.38461538461538464,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990243901117669,
      "loss": 5.6024,
      "mean_token_accuracy": 0.15951776504516602,
      "num_tokens": 8800657.0,
      "step": 4240
    },
    {
      "entropy": 5.657181215286255,
      "epoch": 0.3850689404934688,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990213779026903,
      "loss": 5.5996,
      "mean_token_accuracy": 0.1628565177321434,
      "num_tokens": 8812983.0,
      "step": 4245
    },
    {
      "entropy": 5.771100807189941,
      "epoch": 0.385522496371553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990183610607944,
      "loss": 5.571,
      "mean_token_accuracy": 0.1698968768119812,
      "num_tokens": 8822980.0,
      "step": 4250
    },
    {
      "entropy": 5.677217197418213,
      "epoch": 0.38597605224963716,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990153395861415,
      "loss": 5.5484,
      "mean_token_accuracy": 0.16861167252063752,
      "num_tokens": 8833821.0,
      "step": 4255
    },
    {
      "entropy": 5.625637340545654,
      "epoch": 0.3864296081277213,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990123134787946,
      "loss": 5.4898,
      "mean_token_accuracy": 0.17129137367010117,
      "num_tokens": 8843929.0,
      "step": 4260
    },
    {
      "entropy": 5.754845666885376,
      "epoch": 0.3868831640058055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990092827388159,
      "loss": 5.4759,
      "mean_token_accuracy": 0.16929103881120683,
      "num_tokens": 8854180.0,
      "step": 4265
    },
    {
      "entropy": 5.630288219451904,
      "epoch": 0.3873367198838897,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990062473662681,
      "loss": 5.4903,
      "mean_token_accuracy": 0.17342475801706314,
      "num_tokens": 8865018.0,
      "step": 4270
    },
    {
      "entropy": 5.752366018295288,
      "epoch": 0.3877902757619739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499003207361214,
      "loss": 5.6753,
      "mean_token_accuracy": 0.16161389499902726,
      "num_tokens": 8875423.0,
      "step": 4275
    },
    {
      "entropy": 5.819116306304932,
      "epoch": 0.38824383164005805,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990001627237165,
      "loss": 5.6046,
      "mean_token_accuracy": 0.1682436063885689,
      "num_tokens": 8887850.0,
      "step": 4280
    },
    {
      "entropy": 5.644161605834961,
      "epoch": 0.38869738751814226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989971134538386,
      "loss": 5.4972,
      "mean_token_accuracy": 0.17050439417362212,
      "num_tokens": 8897429.0,
      "step": 4285
    },
    {
      "entropy": 5.663009977340698,
      "epoch": 0.3891509433962264,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989940595516432,
      "loss": 5.5253,
      "mean_token_accuracy": 0.16778848469257354,
      "num_tokens": 8907048.0,
      "step": 4290
    },
    {
      "entropy": 5.702516841888428,
      "epoch": 0.3896044992743106,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989910010171938,
      "loss": 5.5746,
      "mean_token_accuracy": 0.16057664453983306,
      "num_tokens": 8917987.0,
      "step": 4295
    },
    {
      "entropy": 5.754789972305298,
      "epoch": 0.3900580551523948,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989879378505532,
      "loss": 5.4806,
      "mean_token_accuracy": 0.16423581093549727,
      "num_tokens": 8929339.0,
      "step": 4300
    },
    {
      "entropy": 5.599805641174316,
      "epoch": 0.39051161103047893,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989848700517849,
      "loss": 5.4908,
      "mean_token_accuracy": 0.1674097627401352,
      "num_tokens": 8939473.0,
      "step": 4305
    },
    {
      "entropy": 5.650110673904419,
      "epoch": 0.39096516690856314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989817976209526,
      "loss": 5.5306,
      "mean_token_accuracy": 0.16720281541347504,
      "num_tokens": 8949173.0,
      "step": 4310
    },
    {
      "entropy": 5.739528751373291,
      "epoch": 0.3914187227866473,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989787205581196,
      "loss": 5.4996,
      "mean_token_accuracy": 0.1708226829767227,
      "num_tokens": 8959879.0,
      "step": 4315
    },
    {
      "entropy": 5.707206296920776,
      "epoch": 0.3918722786647315,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989756388633496,
      "loss": 5.5819,
      "mean_token_accuracy": 0.16486465483903884,
      "num_tokens": 8970643.0,
      "step": 4320
    },
    {
      "entropy": 5.666028881072998,
      "epoch": 0.39232583454281567,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989725525367063,
      "loss": 5.4561,
      "mean_token_accuracy": 0.16958049088716506,
      "num_tokens": 8979882.0,
      "step": 4325
    },
    {
      "entropy": 5.7163489818572994,
      "epoch": 0.3927793904208999,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989694615782534,
      "loss": 5.5912,
      "mean_token_accuracy": 0.1618393614888191,
      "num_tokens": 8990925.0,
      "step": 4330
    },
    {
      "entropy": 5.722356939315796,
      "epoch": 0.39323294629898403,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989663659880552,
      "loss": 5.6084,
      "mean_token_accuracy": 0.1621517851948738,
      "num_tokens": 9001365.0,
      "step": 4335
    },
    {
      "entropy": 5.79314546585083,
      "epoch": 0.3936865021770682,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989632657661754,
      "loss": 5.6,
      "mean_token_accuracy": 0.1683468908071518,
      "num_tokens": 9011942.0,
      "step": 4340
    },
    {
      "entropy": 5.737189674377442,
      "epoch": 0.3941400580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498960160912678,
      "loss": 5.6458,
      "mean_token_accuracy": 0.16355903297662736,
      "num_tokens": 9022610.0,
      "step": 4345
    },
    {
      "entropy": 5.703625679016113,
      "epoch": 0.39459361393323655,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989570514276276,
      "loss": 5.5296,
      "mean_token_accuracy": 0.16396254748106004,
      "num_tokens": 9033034.0,
      "step": 4350
    },
    {
      "entropy": 5.721978998184204,
      "epoch": 0.39504716981132076,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989539373110883,
      "loss": 5.5728,
      "mean_token_accuracy": 0.16446860283613204,
      "num_tokens": 9043418.0,
      "step": 4355
    },
    {
      "entropy": 5.782990789413452,
      "epoch": 0.3955007256894049,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989508185631245,
      "loss": 5.6716,
      "mean_token_accuracy": 0.16027326881885529,
      "num_tokens": 9052708.0,
      "step": 4360
    },
    {
      "entropy": 5.617509841918945,
      "epoch": 0.39595428156748913,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989476951838006,
      "loss": 5.5176,
      "mean_token_accuracy": 0.16624224931001663,
      "num_tokens": 9062883.0,
      "step": 4365
    },
    {
      "entropy": 5.744745779037475,
      "epoch": 0.3964078374455733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989445671731814,
      "loss": 5.4776,
      "mean_token_accuracy": 0.16907311379909515,
      "num_tokens": 9072830.0,
      "step": 4370
    },
    {
      "entropy": 5.634798336029053,
      "epoch": 0.3968613933236575,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989414345313315,
      "loss": 5.4453,
      "mean_token_accuracy": 0.1632262200117111,
      "num_tokens": 9082749.0,
      "step": 4375
    },
    {
      "entropy": 5.70997200012207,
      "epoch": 0.39731494920174165,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004989382972583156,
      "loss": 5.5417,
      "mean_token_accuracy": 0.1624021738767624,
      "num_tokens": 9094188.0,
      "step": 4380
    },
    {
      "entropy": 5.6712054252624515,
      "epoch": 0.3977685050798258,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989351553541986,
      "loss": 5.5543,
      "mean_token_accuracy": 0.1630195915699005,
      "num_tokens": 9105023.0,
      "step": 4385
    },
    {
      "entropy": 5.616395616531372,
      "epoch": 0.39822206095791,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989320088190456,
      "loss": 5.504,
      "mean_token_accuracy": 0.1688954085111618,
      "num_tokens": 9115563.0,
      "step": 4390
    },
    {
      "entropy": 5.713869190216064,
      "epoch": 0.3986756168359942,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989288576529216,
      "loss": 5.5355,
      "mean_token_accuracy": 0.16001897305250168,
      "num_tokens": 9125394.0,
      "step": 4395
    },
    {
      "entropy": 5.724222660064697,
      "epoch": 0.3991291727140784,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989257018558917,
      "loss": 5.531,
      "mean_token_accuracy": 0.16804168224334717,
      "num_tokens": 9137061.0,
      "step": 4400
    },
    {
      "entropy": 5.685125684738159,
      "epoch": 0.39958272859216254,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989225414280213,
      "loss": 5.4906,
      "mean_token_accuracy": 0.168438558280468,
      "num_tokens": 9148196.0,
      "step": 4405
    },
    {
      "entropy": 5.665059661865234,
      "epoch": 0.40003628447024675,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989193763693756,
      "loss": 5.5142,
      "mean_token_accuracy": 0.1706180453300476,
      "num_tokens": 9158013.0,
      "step": 4410
    },
    {
      "entropy": 5.651648664474488,
      "epoch": 0.4004898403483309,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989162066800202,
      "loss": 5.4225,
      "mean_token_accuracy": 0.17173381000757218,
      "num_tokens": 9167813.0,
      "step": 4415
    },
    {
      "entropy": 5.637704706192016,
      "epoch": 0.4009433962264151,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989130323600205,
      "loss": 5.4884,
      "mean_token_accuracy": 0.16999162435531617,
      "num_tokens": 9177783.0,
      "step": 4420
    },
    {
      "entropy": 5.688581085205078,
      "epoch": 0.40139695210449927,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989098534094424,
      "loss": 5.5333,
      "mean_token_accuracy": 0.16801782250404357,
      "num_tokens": 9187556.0,
      "step": 4425
    },
    {
      "entropy": 5.703752279281616,
      "epoch": 0.4018505079825834,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989066698283512,
      "loss": 5.5988,
      "mean_token_accuracy": 0.16409932374954223,
      "num_tokens": 9198659.0,
      "step": 4430
    },
    {
      "entropy": 5.741377305984497,
      "epoch": 0.40230406386066764,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989034816168132,
      "loss": 5.5719,
      "mean_token_accuracy": 0.1686299681663513,
      "num_tokens": 9209546.0,
      "step": 4435
    },
    {
      "entropy": 5.628240251541138,
      "epoch": 0.4027576197387518,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989002887748941,
      "loss": 5.4811,
      "mean_token_accuracy": 0.17123182117938995,
      "num_tokens": 9219339.0,
      "step": 4440
    },
    {
      "entropy": 5.702338361740113,
      "epoch": 0.403211175616836,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988970913026601,
      "loss": 5.6741,
      "mean_token_accuracy": 0.1605461746454239,
      "num_tokens": 9230364.0,
      "step": 4445
    },
    {
      "entropy": 5.706683349609375,
      "epoch": 0.40366473149492016,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988938892001771,
      "loss": 5.5239,
      "mean_token_accuracy": 0.16557129025459288,
      "num_tokens": 9240732.0,
      "step": 4450
    },
    {
      "entropy": 5.59526686668396,
      "epoch": 0.40411828737300437,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988906824675113,
      "loss": 5.4616,
      "mean_token_accuracy": 0.17066627889871597,
      "num_tokens": 9251049.0,
      "step": 4455
    },
    {
      "entropy": 5.65702075958252,
      "epoch": 0.4045718432510885,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988874711047293,
      "loss": 5.5041,
      "mean_token_accuracy": 0.17010140866041185,
      "num_tokens": 9261660.0,
      "step": 4460
    },
    {
      "entropy": 5.707148694992066,
      "epoch": 0.40502539912917274,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988842551118972,
      "loss": 5.5619,
      "mean_token_accuracy": 0.17123832255601884,
      "num_tokens": 9272114.0,
      "step": 4465
    },
    {
      "entropy": 5.657516384124756,
      "epoch": 0.4054789550072569,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988810344890818,
      "loss": 5.5274,
      "mean_token_accuracy": 0.17154074907302858,
      "num_tokens": 9282125.0,
      "step": 4470
    },
    {
      "entropy": 5.755674219131469,
      "epoch": 0.40593251088534105,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988778092363494,
      "loss": 5.5719,
      "mean_token_accuracy": 0.164210744202137,
      "num_tokens": 9294078.0,
      "step": 4475
    },
    {
      "entropy": 5.703670406341553,
      "epoch": 0.40638606676342526,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988745793537671,
      "loss": 5.5273,
      "mean_token_accuracy": 0.16619765460491182,
      "num_tokens": 9304026.0,
      "step": 4480
    },
    {
      "entropy": 5.6847145557403564,
      "epoch": 0.4068396226415094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988713448414011,
      "loss": 5.5717,
      "mean_token_accuracy": 0.17452892959117888,
      "num_tokens": 9313933.0,
      "step": 4485
    },
    {
      "entropy": 5.682414197921753,
      "epoch": 0.4072931785195936,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498868105699319,
      "loss": 5.6038,
      "mean_token_accuracy": 0.1653165489435196,
      "num_tokens": 9323932.0,
      "step": 4490
    },
    {
      "entropy": 5.62096529006958,
      "epoch": 0.4077467343976778,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988648619275872,
      "loss": 5.4957,
      "mean_token_accuracy": 0.17853589355945587,
      "num_tokens": 9333467.0,
      "step": 4495
    },
    {
      "entropy": 5.730397415161133,
      "epoch": 0.408200290275762,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988616135262729,
      "loss": 5.5523,
      "mean_token_accuracy": 0.16236073523759842,
      "num_tokens": 9344909.0,
      "step": 4500
    },
    {
      "entropy": 5.778379106521607,
      "epoch": 0.40865384615384615,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004988583604954436,
      "loss": 5.6145,
      "mean_token_accuracy": 0.1632135108113289,
      "num_tokens": 9355630.0,
      "step": 4505
    },
    {
      "entropy": 5.630618858337402,
      "epoch": 0.40910740203193036,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988551028351663,
      "loss": 5.539,
      "mean_token_accuracy": 0.16523584127426147,
      "num_tokens": 9365700.0,
      "step": 4510
    },
    {
      "entropy": 5.654221391677856,
      "epoch": 0.4095609579100145,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988518405455084,
      "loss": 5.6261,
      "mean_token_accuracy": 0.15864572525024415,
      "num_tokens": 9377088.0,
      "step": 4515
    },
    {
      "entropy": 5.703935956954956,
      "epoch": 0.41001451378809867,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988485736265374,
      "loss": 5.4107,
      "mean_token_accuracy": 0.17571215331554413,
      "num_tokens": 9386855.0,
      "step": 4520
    },
    {
      "entropy": 5.696550416946411,
      "epoch": 0.4104680696661829,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498845302078321,
      "loss": 5.5307,
      "mean_token_accuracy": 0.16638017445802689,
      "num_tokens": 9398661.0,
      "step": 4525
    },
    {
      "entropy": 5.72752366065979,
      "epoch": 0.41092162554426703,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988420259009266,
      "loss": 5.5315,
      "mean_token_accuracy": 0.17306275367736818,
      "num_tokens": 9408670.0,
      "step": 4530
    },
    {
      "entropy": 5.688307809829712,
      "epoch": 0.41137518142235124,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498838745094422,
      "loss": 5.5582,
      "mean_token_accuracy": 0.16831061094999314,
      "num_tokens": 9418740.0,
      "step": 4535
    },
    {
      "entropy": 5.6272059917449955,
      "epoch": 0.4118287373004354,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988354596588751,
      "loss": 5.5311,
      "mean_token_accuracy": 0.16975374519824982,
      "num_tokens": 9429339.0,
      "step": 4540
    },
    {
      "entropy": 5.678937482833862,
      "epoch": 0.4122822931785196,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988321695943539,
      "loss": 5.4738,
      "mean_token_accuracy": 0.1705018848180771,
      "num_tokens": 9438996.0,
      "step": 4545
    },
    {
      "entropy": 5.698333978652954,
      "epoch": 0.41273584905660377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988288749009264,
      "loss": 5.4912,
      "mean_token_accuracy": 0.1687777042388916,
      "num_tokens": 9448669.0,
      "step": 4550
    },
    {
      "entropy": 5.645137596130371,
      "epoch": 0.413189404934688,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988255755786608,
      "loss": 5.4872,
      "mean_token_accuracy": 0.1714872419834137,
      "num_tokens": 9458857.0,
      "step": 4555
    },
    {
      "entropy": 5.650253391265869,
      "epoch": 0.41364296081277213,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988222716276251,
      "loss": 5.4782,
      "mean_token_accuracy": 0.17236211895942688,
      "num_tokens": 9470241.0,
      "step": 4560
    },
    {
      "entropy": 5.731791353225708,
      "epoch": 0.4140965166908563,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498818963047888,
      "loss": 5.6346,
      "mean_token_accuracy": 0.16064894944429398,
      "num_tokens": 9480528.0,
      "step": 4565
    },
    {
      "entropy": 5.75153079032898,
      "epoch": 0.4145500725689405,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988156498395176,
      "loss": 5.5396,
      "mean_token_accuracy": 0.17293335497379303,
      "num_tokens": 9491000.0,
      "step": 4570
    },
    {
      "entropy": 5.723613452911377,
      "epoch": 0.41500362844702465,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004988123320025825,
      "loss": 5.5322,
      "mean_token_accuracy": 0.17224910855293274,
      "num_tokens": 9501612.0,
      "step": 4575
    },
    {
      "entropy": 5.692364740371704,
      "epoch": 0.41545718432510886,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988090095371514,
      "loss": 5.639,
      "mean_token_accuracy": 0.16442762464284896,
      "num_tokens": 9513999.0,
      "step": 4580
    },
    {
      "entropy": 5.704233598709107,
      "epoch": 0.415910740203193,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988056824432928,
      "loss": 5.5106,
      "mean_token_accuracy": 0.1647787868976593,
      "num_tokens": 9524392.0,
      "step": 4585
    },
    {
      "entropy": 5.709184408187866,
      "epoch": 0.41636429608127723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988023507210758,
      "loss": 5.5108,
      "mean_token_accuracy": 0.1714200809597969,
      "num_tokens": 9534227.0,
      "step": 4590
    },
    {
      "entropy": 5.69327826499939,
      "epoch": 0.4168178519593614,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987990143705692,
      "loss": 5.5107,
      "mean_token_accuracy": 0.1758728012442589,
      "num_tokens": 9544242.0,
      "step": 4595
    },
    {
      "entropy": 5.6579484939575195,
      "epoch": 0.4172714078374456,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987956733918418,
      "loss": 5.434,
      "mean_token_accuracy": 0.17048389017581939,
      "num_tokens": 9553921.0,
      "step": 4600
    },
    {
      "entropy": 5.562104272842407,
      "epoch": 0.41772496371552975,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987923277849629,
      "loss": 5.4928,
      "mean_token_accuracy": 0.16425550878047943,
      "num_tokens": 9564768.0,
      "step": 4605
    },
    {
      "entropy": 5.7012450218200685,
      "epoch": 0.4181785195936139,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987889775500017,
      "loss": 5.5184,
      "mean_token_accuracy": 0.16336782574653624,
      "num_tokens": 9576360.0,
      "step": 4610
    },
    {
      "entropy": 5.6748738288879395,
      "epoch": 0.4186320754716981,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987856226870272,
      "loss": 5.54,
      "mean_token_accuracy": 0.16737818121910095,
      "num_tokens": 9587623.0,
      "step": 4615
    },
    {
      "entropy": 5.636663770675659,
      "epoch": 0.4190856313497823,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987822631961092,
      "loss": 5.5691,
      "mean_token_accuracy": 0.1642339438199997,
      "num_tokens": 9597837.0,
      "step": 4620
    },
    {
      "entropy": 5.706202650070191,
      "epoch": 0.4195391872278665,
      "grad_norm": 0.875,
      "learning_rate": 0.000498778899077317,
      "loss": 5.5288,
      "mean_token_accuracy": 0.1656550332903862,
      "num_tokens": 9609509.0,
      "step": 4625
    },
    {
      "entropy": 5.667462539672852,
      "epoch": 0.41999274310595064,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049877553033072,
      "loss": 5.4957,
      "mean_token_accuracy": 0.17181909531354905,
      "num_tokens": 9619954.0,
      "step": 4630
    },
    {
      "entropy": 5.644245195388794,
      "epoch": 0.42044629898403485,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987721569563882,
      "loss": 5.4973,
      "mean_token_accuracy": 0.16737744808197022,
      "num_tokens": 9629742.0,
      "step": 4635
    },
    {
      "entropy": 5.666360330581665,
      "epoch": 0.420899854862119,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498768778954391,
      "loss": 5.515,
      "mean_token_accuracy": 0.1711300268769264,
      "num_tokens": 9641378.0,
      "step": 4640
    },
    {
      "entropy": 5.756295871734619,
      "epoch": 0.4213534107402032,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987653963247985,
      "loss": 5.5852,
      "mean_token_accuracy": 0.16515168696641921,
      "num_tokens": 9652279.0,
      "step": 4645
    },
    {
      "entropy": 5.715899276733398,
      "epoch": 0.42180696661828737,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004987620090676805,
      "loss": 5.5811,
      "mean_token_accuracy": 0.16966685950756072,
      "num_tokens": 9663416.0,
      "step": 4650
    },
    {
      "entropy": 5.65866527557373,
      "epoch": 0.4222605224963715,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987586171831072,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16795586943626403,
      "num_tokens": 9674647.0,
      "step": 4655
    },
    {
      "entropy": 5.646727991104126,
      "epoch": 0.42271407837445574,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987552206711487,
      "loss": 5.3947,
      "mean_token_accuracy": 0.17098441645503043,
      "num_tokens": 9685084.0,
      "step": 4660
    },
    {
      "entropy": 5.602826547622681,
      "epoch": 0.4231676342525399,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987518195318752,
      "loss": 5.451,
      "mean_token_accuracy": 0.17014521807432176,
      "num_tokens": 9695661.0,
      "step": 4665
    },
    {
      "entropy": 5.699631309509277,
      "epoch": 0.4236211901306241,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498748413765357,
      "loss": 5.5111,
      "mean_token_accuracy": 0.1655079945921898,
      "num_tokens": 9705773.0,
      "step": 4670
    },
    {
      "entropy": 5.672311735153198,
      "epoch": 0.42407474600870826,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987450033716646,
      "loss": 5.5543,
      "mean_token_accuracy": 0.1671050012111664,
      "num_tokens": 9715799.0,
      "step": 4675
    },
    {
      "entropy": 5.70002236366272,
      "epoch": 0.42452830188679247,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987415883508686,
      "loss": 5.5117,
      "mean_token_accuracy": 0.17100395262241364,
      "num_tokens": 9727012.0,
      "step": 4680
    },
    {
      "entropy": 5.699238252639771,
      "epoch": 0.4249818577648766,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987381687030394,
      "loss": 5.5134,
      "mean_token_accuracy": 0.16880285739898682,
      "num_tokens": 9737044.0,
      "step": 4685
    },
    {
      "entropy": 5.728292417526245,
      "epoch": 0.42543541364296084,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987347444282479,
      "loss": 5.5911,
      "mean_token_accuracy": 0.16349772363901138,
      "num_tokens": 9747648.0,
      "step": 4690
    },
    {
      "entropy": 5.686186790466309,
      "epoch": 0.425888969521045,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987313155265649,
      "loss": 5.5245,
      "mean_token_accuracy": 0.163296639919281,
      "num_tokens": 9757812.0,
      "step": 4695
    },
    {
      "entropy": 5.753760099411011,
      "epoch": 0.42634252539912915,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987278819980613,
      "loss": 5.5178,
      "mean_token_accuracy": 0.174093596637249,
      "num_tokens": 9768337.0,
      "step": 4700
    },
    {
      "entropy": 5.703652429580688,
      "epoch": 0.42679608127721336,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987244438428081,
      "loss": 5.5634,
      "mean_token_accuracy": 0.16444121152162552,
      "num_tokens": 9779433.0,
      "step": 4705
    },
    {
      "entropy": 5.611905670166015,
      "epoch": 0.4272496371552975,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987210010608764,
      "loss": 5.4602,
      "mean_token_accuracy": 0.1711221307516098,
      "num_tokens": 9790069.0,
      "step": 4710
    },
    {
      "entropy": 5.659437561035157,
      "epoch": 0.4277031930333817,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987175536523373,
      "loss": 5.4316,
      "mean_token_accuracy": 0.17792649418115616,
      "num_tokens": 9799669.0,
      "step": 4715
    },
    {
      "entropy": 5.626423311233521,
      "epoch": 0.4281567489114659,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987141016172622,
      "loss": 5.4731,
      "mean_token_accuracy": 0.1750251442193985,
      "num_tokens": 9809615.0,
      "step": 4720
    },
    {
      "entropy": 5.633890104293823,
      "epoch": 0.4286103047895501,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987106449557225,
      "loss": 5.4379,
      "mean_token_accuracy": 0.17350732386112214,
      "num_tokens": 9819666.0,
      "step": 4725
    },
    {
      "entropy": 5.73306941986084,
      "epoch": 0.42906386066763424,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987071836677896,
      "loss": 5.6122,
      "mean_token_accuracy": 0.16756637394428253,
      "num_tokens": 9831015.0,
      "step": 4730
    },
    {
      "entropy": 5.635014009475708,
      "epoch": 0.42951741654571846,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987037177535353,
      "loss": 5.4679,
      "mean_token_accuracy": 0.1639992356300354,
      "num_tokens": 9841354.0,
      "step": 4735
    },
    {
      "entropy": 5.633705282211304,
      "epoch": 0.4299709724238026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987002472130309,
      "loss": 5.4563,
      "mean_token_accuracy": 0.16753149628639222,
      "num_tokens": 9850467.0,
      "step": 4740
    },
    {
      "entropy": 5.770284414291382,
      "epoch": 0.43042452830188677,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986967720463485,
      "loss": 5.5496,
      "mean_token_accuracy": 0.1623040944337845,
      "num_tokens": 9861115.0,
      "step": 4745
    },
    {
      "entropy": 5.7330851554870605,
      "epoch": 0.430878084179971,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986932922535598,
      "loss": 5.5384,
      "mean_token_accuracy": 0.16642699092626573,
      "num_tokens": 9871172.0,
      "step": 4750
    },
    {
      "entropy": 5.652520322799683,
      "epoch": 0.43133164005805513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986898078347368,
      "loss": 5.4515,
      "mean_token_accuracy": 0.17356225550174714,
      "num_tokens": 9881997.0,
      "step": 4755
    },
    {
      "entropy": 5.61650390625,
      "epoch": 0.43178519593613934,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986863187899516,
      "loss": 5.4211,
      "mean_token_accuracy": 0.1723109319806099,
      "num_tokens": 9892310.0,
      "step": 4760
    },
    {
      "entropy": 5.6054548740386965,
      "epoch": 0.4322387518142235,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986828251192764,
      "loss": 5.4786,
      "mean_token_accuracy": 0.17029564827680588,
      "num_tokens": 9902664.0,
      "step": 4765
    },
    {
      "entropy": 5.6452922344207765,
      "epoch": 0.4326923076923077,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986793268227833,
      "loss": 5.4662,
      "mean_token_accuracy": 0.17340250313282013,
      "num_tokens": 9912491.0,
      "step": 4770
    },
    {
      "entropy": 5.65220012664795,
      "epoch": 0.43314586357039186,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986758239005448,
      "loss": 5.439,
      "mean_token_accuracy": 0.17318663150072097,
      "num_tokens": 9922057.0,
      "step": 4775
    },
    {
      "entropy": 5.640111303329467,
      "epoch": 0.4335994194484761,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986723163526331,
      "loss": 5.4976,
      "mean_token_accuracy": 0.1674135446548462,
      "num_tokens": 9933661.0,
      "step": 4780
    },
    {
      "entropy": 5.577649021148682,
      "epoch": 0.43405297532656023,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986688041791209,
      "loss": 5.4532,
      "mean_token_accuracy": 0.16382990777492523,
      "num_tokens": 9945177.0,
      "step": 4785
    },
    {
      "entropy": 5.671372318267823,
      "epoch": 0.4345065312046444,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498665287380081,
      "loss": 5.479,
      "mean_token_accuracy": 0.1665464699268341,
      "num_tokens": 9956226.0,
      "step": 4790
    },
    {
      "entropy": 5.709822797775269,
      "epoch": 0.4349600870827286,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986617659555859,
      "loss": 5.5528,
      "mean_token_accuracy": 0.16944931149482728,
      "num_tokens": 9965877.0,
      "step": 4795
    },
    {
      "entropy": 5.627245140075684,
      "epoch": 0.43541364296081275,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986582399057085,
      "loss": 5.4958,
      "mean_token_accuracy": 0.17361783385276794,
      "num_tokens": 9975592.0,
      "step": 4800
    },
    {
      "entropy": 5.716971731185913,
      "epoch": 0.43586719883889696,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986547092305216,
      "loss": 5.5188,
      "mean_token_accuracy": 0.16906160414218901,
      "num_tokens": 9986094.0,
      "step": 4805
    },
    {
      "entropy": 5.747332000732422,
      "epoch": 0.4363207547169811,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986511739300984,
      "loss": 5.5563,
      "mean_token_accuracy": 0.16685436069965362,
      "num_tokens": 9997241.0,
      "step": 4810
    },
    {
      "entropy": 5.624169874191284,
      "epoch": 0.43677431059506533,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986476340045119,
      "loss": 5.4805,
      "mean_token_accuracy": 0.1696499690413475,
      "num_tokens": 10008091.0,
      "step": 4815
    },
    {
      "entropy": 5.610916471481323,
      "epoch": 0.4372278664731495,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986440894538354,
      "loss": 5.5358,
      "mean_token_accuracy": 0.16445003151893617,
      "num_tokens": 10018794.0,
      "step": 4820
    },
    {
      "entropy": 5.658578824996948,
      "epoch": 0.4376814223512337,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986405402781421,
      "loss": 5.4595,
      "mean_token_accuracy": 0.17329651415348052,
      "num_tokens": 10028414.0,
      "step": 4825
    },
    {
      "entropy": 5.683524417877197,
      "epoch": 0.43813497822931785,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986369864775054,
      "loss": 5.5255,
      "mean_token_accuracy": 0.1668830245733261,
      "num_tokens": 10038749.0,
      "step": 4830
    },
    {
      "entropy": 5.680928516387939,
      "epoch": 0.438588534107402,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986334280519989,
      "loss": 5.4706,
      "mean_token_accuracy": 0.16898296475410463,
      "num_tokens": 10049866.0,
      "step": 4835
    },
    {
      "entropy": 5.620717573165893,
      "epoch": 0.4390420899854862,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498629865001696,
      "loss": 5.4971,
      "mean_token_accuracy": 0.176370669901371,
      "num_tokens": 10060680.0,
      "step": 4840
    },
    {
      "entropy": 5.6851057529449465,
      "epoch": 0.43949564586357037,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986262973266706,
      "loss": 5.6094,
      "mean_token_accuracy": 0.16637811809778214,
      "num_tokens": 10070931.0,
      "step": 4845
    },
    {
      "entropy": 5.648931550979614,
      "epoch": 0.4399492017416546,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986227250269964,
      "loss": 5.5285,
      "mean_token_accuracy": 0.17198051065206527,
      "num_tokens": 10080189.0,
      "step": 4850
    },
    {
      "entropy": 5.683497285842895,
      "epoch": 0.44040275761973874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004986191481027472,
      "loss": 5.4349,
      "mean_token_accuracy": 0.17278559356927872,
      "num_tokens": 10088813.0,
      "step": 4855
    },
    {
      "entropy": 5.61211371421814,
      "epoch": 0.44085631349782295,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986155665539971,
      "loss": 5.4644,
      "mean_token_accuracy": 0.17080731093883514,
      "num_tokens": 10098125.0,
      "step": 4860
    },
    {
      "entropy": 5.69177656173706,
      "epoch": 0.4413098693759071,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049861198038082,
      "loss": 5.506,
      "mean_token_accuracy": 0.17002110332250595,
      "num_tokens": 10109125.0,
      "step": 4865
    },
    {
      "entropy": 5.673552989959717,
      "epoch": 0.4417634252539913,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986083895832903,
      "loss": 5.4864,
      "mean_token_accuracy": 0.16711554676294327,
      "num_tokens": 10120182.0,
      "step": 4870
    },
    {
      "entropy": 5.619903326034546,
      "epoch": 0.44221698113207547,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986047941614821,
      "loss": 5.5446,
      "mean_token_accuracy": 0.1634003072977066,
      "num_tokens": 10130941.0,
      "step": 4875
    },
    {
      "entropy": 5.683604526519775,
      "epoch": 0.4426705370101596,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986011941154696,
      "loss": 5.5044,
      "mean_token_accuracy": 0.16631961166858672,
      "num_tokens": 10141708.0,
      "step": 4880
    },
    {
      "entropy": 5.637266683578491,
      "epoch": 0.44312409288824384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985975894453275,
      "loss": 5.4431,
      "mean_token_accuracy": 0.16837728172540664,
      "num_tokens": 10151816.0,
      "step": 4885
    },
    {
      "entropy": 5.694721031188965,
      "epoch": 0.443577648766328,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985939801511304,
      "loss": 5.5662,
      "mean_token_accuracy": 0.16548764184117318,
      "num_tokens": 10162470.0,
      "step": 4890
    },
    {
      "entropy": 5.672745275497436,
      "epoch": 0.4440312046444122,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985903662329527,
      "loss": 5.4515,
      "mean_token_accuracy": 0.17449744194746017,
      "num_tokens": 10173171.0,
      "step": 4895
    },
    {
      "entropy": 5.667294073104858,
      "epoch": 0.44448476052249636,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985867476908693,
      "loss": 5.4873,
      "mean_token_accuracy": 0.17015801668167113,
      "num_tokens": 10184185.0,
      "step": 4900
    },
    {
      "entropy": 5.702019119262696,
      "epoch": 0.44493831640058057,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985831245249549,
      "loss": 5.5464,
      "mean_token_accuracy": 0.169259525090456,
      "num_tokens": 10195516.0,
      "step": 4905
    },
    {
      "entropy": 5.6287211894989015,
      "epoch": 0.4453918722786647,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985794967352846,
      "loss": 5.4707,
      "mean_token_accuracy": 0.1748988300561905,
      "num_tokens": 10204936.0,
      "step": 4910
    },
    {
      "entropy": 5.658074283599854,
      "epoch": 0.44584542815674894,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004985758643219334,
      "loss": 5.4998,
      "mean_token_accuracy": 0.17041691839694978,
      "num_tokens": 10216218.0,
      "step": 4915
    },
    {
      "entropy": 5.744098043441772,
      "epoch": 0.4462989840348331,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985722272849762,
      "loss": 5.5099,
      "mean_token_accuracy": 0.16554411202669145,
      "num_tokens": 10226370.0,
      "step": 4920
    },
    {
      "entropy": 5.618660640716553,
      "epoch": 0.44675253991291725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985685856244884,
      "loss": 5.4398,
      "mean_token_accuracy": 0.18081948906183243,
      "num_tokens": 10236170.0,
      "step": 4925
    },
    {
      "entropy": 5.666572141647339,
      "epoch": 0.44720609579100146,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985649393405453,
      "loss": 5.5025,
      "mean_token_accuracy": 0.16895841807127,
      "num_tokens": 10246119.0,
      "step": 4930
    },
    {
      "entropy": 5.671585464477539,
      "epoch": 0.4476596516690856,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004985612884332223,
      "loss": 5.5196,
      "mean_token_accuracy": 0.16341286897659302,
      "num_tokens": 10256762.0,
      "step": 4935
    },
    {
      "entropy": 5.617032051086426,
      "epoch": 0.4481132075471698,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985576329025949,
      "loss": 5.4652,
      "mean_token_accuracy": 0.17075106054544448,
      "num_tokens": 10267317.0,
      "step": 4940
    },
    {
      "entropy": 5.681619024276733,
      "epoch": 0.448566763425254,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985539727487385,
      "loss": 5.603,
      "mean_token_accuracy": 0.15708888173103333,
      "num_tokens": 10279111.0,
      "step": 4945
    },
    {
      "entropy": 5.6958630084991455,
      "epoch": 0.4490203193033382,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985503079717292,
      "loss": 5.4674,
      "mean_token_accuracy": 0.1703536629676819,
      "num_tokens": 10289383.0,
      "step": 4950
    },
    {
      "entropy": 5.6642831325531,
      "epoch": 0.44947387518142234,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985466385716425,
      "loss": 5.4767,
      "mean_token_accuracy": 0.17260171324014664,
      "num_tokens": 10299806.0,
      "step": 4955
    },
    {
      "entropy": 5.578657579421997,
      "epoch": 0.44992743105950656,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985429645485543,
      "loss": 5.5005,
      "mean_token_accuracy": 0.1660196453332901,
      "num_tokens": 10310348.0,
      "step": 4960
    },
    {
      "entropy": 5.674696922302246,
      "epoch": 0.4503809869375907,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985392859025407,
      "loss": 5.4875,
      "mean_token_accuracy": 0.17419226914644242,
      "num_tokens": 10320354.0,
      "step": 4965
    },
    {
      "entropy": 5.604598712921143,
      "epoch": 0.45083454281567487,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985356026336776,
      "loss": 5.4148,
      "mean_token_accuracy": 0.16999066323041917,
      "num_tokens": 10330292.0,
      "step": 4970
    },
    {
      "entropy": 5.621414422988892,
      "epoch": 0.4512880986937591,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985319147420414,
      "loss": 5.3829,
      "mean_token_accuracy": 0.18479472696781157,
      "num_tokens": 10340449.0,
      "step": 4975
    },
    {
      "entropy": 5.649209356307983,
      "epoch": 0.45174165457184323,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985282222277082,
      "loss": 5.4569,
      "mean_token_accuracy": 0.16669374704360962,
      "num_tokens": 10350637.0,
      "step": 4980
    },
    {
      "entropy": 5.581578731536865,
      "epoch": 0.45219521044992744,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985245250907544,
      "loss": 5.3903,
      "mean_token_accuracy": 0.17272900640964509,
      "num_tokens": 10361548.0,
      "step": 4985
    },
    {
      "entropy": 5.573017024993897,
      "epoch": 0.4526487663280116,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985208233312564,
      "loss": 5.4129,
      "mean_token_accuracy": 0.16686842143535613,
      "num_tokens": 10371765.0,
      "step": 4990
    },
    {
      "entropy": 5.696421766281128,
      "epoch": 0.4531023222060958,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985171169492908,
      "loss": 5.5166,
      "mean_token_accuracy": 0.17461331635713578,
      "num_tokens": 10382108.0,
      "step": 4995
    },
    {
      "entropy": 5.630676031112671,
      "epoch": 0.45355587808417996,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985134059449344,
      "loss": 5.4666,
      "mean_token_accuracy": 0.18015743643045426,
      "num_tokens": 10392390.0,
      "step": 5000
    },
    {
      "entropy": 5.6141965866088865,
      "epoch": 0.4540094339622642,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985096903182637,
      "loss": 5.5127,
      "mean_token_accuracy": 0.17537071108818053,
      "num_tokens": 10402094.0,
      "step": 5005
    },
    {
      "entropy": 5.666981887817383,
      "epoch": 0.45446298984034833,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985059700693557,
      "loss": 5.466,
      "mean_token_accuracy": 0.173527130484581,
      "num_tokens": 10413350.0,
      "step": 5010
    },
    {
      "entropy": 5.702916288375855,
      "epoch": 0.4549165457184325,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985022451982872,
      "loss": 5.5916,
      "mean_token_accuracy": 0.16218466758728028,
      "num_tokens": 10424292.0,
      "step": 5015
    },
    {
      "entropy": 5.756633567810058,
      "epoch": 0.4553701015965167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984985157051354,
      "loss": 5.4481,
      "mean_token_accuracy": 0.1688245952129364,
      "num_tokens": 10433724.0,
      "step": 5020
    },
    {
      "entropy": 5.648002099990845,
      "epoch": 0.45582365747460085,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984947815899774,
      "loss": 5.5425,
      "mean_token_accuracy": 0.1708854228258133,
      "num_tokens": 10444724.0,
      "step": 5025
    },
    {
      "entropy": 5.65775728225708,
      "epoch": 0.45627721335268506,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984910428528902,
      "loss": 5.4623,
      "mean_token_accuracy": 0.17359500378370285,
      "num_tokens": 10455086.0,
      "step": 5030
    },
    {
      "entropy": 5.598028898239136,
      "epoch": 0.4567307692307692,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984872994939514,
      "loss": 5.4155,
      "mean_token_accuracy": 0.1689280539751053,
      "num_tokens": 10465076.0,
      "step": 5035
    },
    {
      "entropy": 5.6523706912994385,
      "epoch": 0.45718432510885343,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984835515132382,
      "loss": 5.4321,
      "mean_token_accuracy": 0.17813173532485962,
      "num_tokens": 10474817.0,
      "step": 5040
    },
    {
      "entropy": 5.67584810256958,
      "epoch": 0.4576378809869376,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004984797989108282,
      "loss": 5.4584,
      "mean_token_accuracy": 0.1707393169403076,
      "num_tokens": 10485735.0,
      "step": 5045
    },
    {
      "entropy": 5.613812065124511,
      "epoch": 0.4580914368650218,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984760416867991,
      "loss": 5.5472,
      "mean_token_accuracy": 0.1682056277990341,
      "num_tokens": 10496731.0,
      "step": 5050
    },
    {
      "entropy": 5.568698740005493,
      "epoch": 0.45854499274310595,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984722798412286,
      "loss": 5.4136,
      "mean_token_accuracy": 0.17073957175016402,
      "num_tokens": 10506787.0,
      "step": 5055
    },
    {
      "entropy": 5.659662580490112,
      "epoch": 0.4589985486211901,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984685133741941,
      "loss": 5.4739,
      "mean_token_accuracy": 0.16943865567445754,
      "num_tokens": 10517170.0,
      "step": 5060
    },
    {
      "entropy": 5.622515487670898,
      "epoch": 0.4594521044992743,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498464742285774,
      "loss": 5.4733,
      "mean_token_accuracy": 0.1717009276151657,
      "num_tokens": 10527342.0,
      "step": 5065
    },
    {
      "entropy": 5.642328071594238,
      "epoch": 0.45990566037735847,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498460966576046,
      "loss": 5.4503,
      "mean_token_accuracy": 0.17293796241283416,
      "num_tokens": 10537253.0,
      "step": 5070
    },
    {
      "entropy": 5.698889923095703,
      "epoch": 0.4603592162554427,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984571862450884,
      "loss": 5.465,
      "mean_token_accuracy": 0.1680241197347641,
      "num_tokens": 10547249.0,
      "step": 5075
    },
    {
      "entropy": 5.7043322086334225,
      "epoch": 0.46081277213352684,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004984534012929791,
      "loss": 5.503,
      "mean_token_accuracy": 0.172734771668911,
      "num_tokens": 10558174.0,
      "step": 5080
    },
    {
      "entropy": 5.700099897384644,
      "epoch": 0.46126632801161105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984496117197967,
      "loss": 5.4444,
      "mean_token_accuracy": 0.17787872403860092,
      "num_tokens": 10567610.0,
      "step": 5085
    },
    {
      "entropy": 5.602445316314697,
      "epoch": 0.4617198838896952,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984458175256193,
      "loss": 5.462,
      "mean_token_accuracy": 0.16895331740379332,
      "num_tokens": 10578494.0,
      "step": 5090
    },
    {
      "entropy": 5.690131187438965,
      "epoch": 0.4621734397677794,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984420187105253,
      "loss": 5.5668,
      "mean_token_accuracy": 0.16630938053131103,
      "num_tokens": 10589482.0,
      "step": 5095
    },
    {
      "entropy": 5.727915954589844,
      "epoch": 0.46262699564586357,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984382152745936,
      "loss": 5.5203,
      "mean_token_accuracy": 0.1676805019378662,
      "num_tokens": 10599965.0,
      "step": 5100
    },
    {
      "entropy": 5.638349914550782,
      "epoch": 0.4630805515239477,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984344072179026,
      "loss": 5.5459,
      "mean_token_accuracy": 0.1719616398215294,
      "num_tokens": 10610500.0,
      "step": 5105
    },
    {
      "entropy": 5.737101459503174,
      "epoch": 0.46353410740203194,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984305945405312,
      "loss": 5.5,
      "mean_token_accuracy": 0.17199551463127136,
      "num_tokens": 10619843.0,
      "step": 5110
    },
    {
      "entropy": 5.645010232925415,
      "epoch": 0.4639876632801161,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498426777242558,
      "loss": 5.4291,
      "mean_token_accuracy": 0.1767846554517746,
      "num_tokens": 10631026.0,
      "step": 5115
    },
    {
      "entropy": 5.490650749206543,
      "epoch": 0.4644412191582003,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984229553240623,
      "loss": 5.4348,
      "mean_token_accuracy": 0.18036371171474458,
      "num_tokens": 10641130.0,
      "step": 5120
    },
    {
      "entropy": 5.699585437774658,
      "epoch": 0.46489477503628446,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984191287851228,
      "loss": 5.4958,
      "mean_token_accuracy": 0.17234579771757125,
      "num_tokens": 10652305.0,
      "step": 5125
    },
    {
      "entropy": 5.680755996704102,
      "epoch": 0.46534833091436867,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984152976258188,
      "loss": 5.5328,
      "mean_token_accuracy": 0.17020701766014099,
      "num_tokens": 10663083.0,
      "step": 5130
    },
    {
      "entropy": 5.659653997421264,
      "epoch": 0.4658018867924528,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984114618462296,
      "loss": 5.5545,
      "mean_token_accuracy": 0.16575139313936232,
      "num_tokens": 10673602.0,
      "step": 5135
    },
    {
      "entropy": 5.619678211212158,
      "epoch": 0.46625544267053703,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984076214464343,
      "loss": 5.3841,
      "mean_token_accuracy": 0.1807913675904274,
      "num_tokens": 10684622.0,
      "step": 5140
    },
    {
      "entropy": 5.624106788635254,
      "epoch": 0.4667089985486212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984037764265126,
      "loss": 5.4969,
      "mean_token_accuracy": 0.17183390706777574,
      "num_tokens": 10695129.0,
      "step": 5145
    },
    {
      "entropy": 5.588241958618164,
      "epoch": 0.46716255442670535,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983999267865438,
      "loss": 5.3285,
      "mean_token_accuracy": 0.18249204456806184,
      "num_tokens": 10705236.0,
      "step": 5150
    },
    {
      "entropy": 5.593046188354492,
      "epoch": 0.46761611030478956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983960725266078,
      "loss": 5.4582,
      "mean_token_accuracy": 0.17504026740789413,
      "num_tokens": 10715748.0,
      "step": 5155
    },
    {
      "entropy": 5.624542713165283,
      "epoch": 0.4680696661828737,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983922136467838,
      "loss": 5.3846,
      "mean_token_accuracy": 0.1776614785194397,
      "num_tokens": 10725799.0,
      "step": 5160
    },
    {
      "entropy": 5.6600791931152346,
      "epoch": 0.4685232220609579,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498388350147152,
      "loss": 5.5062,
      "mean_token_accuracy": 0.16867911964654922,
      "num_tokens": 10735668.0,
      "step": 5165
    },
    {
      "entropy": 5.662786149978638,
      "epoch": 0.4689767779390421,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983844820277922,
      "loss": 5.6079,
      "mean_token_accuracy": 0.167015041410923,
      "num_tokens": 10746466.0,
      "step": 5170
    },
    {
      "entropy": 5.752711057662964,
      "epoch": 0.4694303338171263,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983806092887843,
      "loss": 5.5063,
      "mean_token_accuracy": 0.1687478616833687,
      "num_tokens": 10756724.0,
      "step": 5175
    },
    {
      "entropy": 5.664383506774902,
      "epoch": 0.46988388969521044,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983767319302086,
      "loss": 5.5499,
      "mean_token_accuracy": 0.16907021701335906,
      "num_tokens": 10766919.0,
      "step": 5180
    },
    {
      "entropy": 5.576836204528808,
      "epoch": 0.47033744557329465,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983728499521451,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18249264657497405,
      "num_tokens": 10777550.0,
      "step": 5185
    },
    {
      "entropy": 5.655050230026245,
      "epoch": 0.4707910014513788,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983689633546741,
      "loss": 5.4027,
      "mean_token_accuracy": 0.17136975079774858,
      "num_tokens": 10788456.0,
      "step": 5190
    },
    {
      "entropy": 5.663191461563111,
      "epoch": 0.47124455732946297,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983650721378761,
      "loss": 5.4345,
      "mean_token_accuracy": 0.17229533791542054,
      "num_tokens": 10800049.0,
      "step": 5195
    },
    {
      "entropy": 5.655028867721557,
      "epoch": 0.4716981132075472,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983611763018314,
      "loss": 5.4815,
      "mean_token_accuracy": 0.16865273416042328,
      "num_tokens": 10810743.0,
      "step": 5200
    },
    {
      "entropy": 5.753147888183594,
      "epoch": 0.47215166908563133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983572758466208,
      "loss": 5.677,
      "mean_token_accuracy": 0.16416452825069427,
      "num_tokens": 10821897.0,
      "step": 5205
    },
    {
      "entropy": 5.7169407367706295,
      "epoch": 0.47260522496371554,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983533707723247,
      "loss": 5.527,
      "mean_token_accuracy": 0.16855133473873138,
      "num_tokens": 10831589.0,
      "step": 5210
    },
    {
      "entropy": 5.657151603698731,
      "epoch": 0.4730587808417997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983494610790241,
      "loss": 5.429,
      "mean_token_accuracy": 0.16840204745531082,
      "num_tokens": 10841404.0,
      "step": 5215
    },
    {
      "entropy": 5.6251708984375,
      "epoch": 0.4735123367198839,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983455467667996,
      "loss": 5.471,
      "mean_token_accuracy": 0.1659417137503624,
      "num_tokens": 10851577.0,
      "step": 5220
    },
    {
      "entropy": 5.615735912322998,
      "epoch": 0.47396589259796806,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983416278357322,
      "loss": 5.3906,
      "mean_token_accuracy": 0.1765459805727005,
      "num_tokens": 10861403.0,
      "step": 5225
    },
    {
      "entropy": 5.629169273376465,
      "epoch": 0.4744194484760523,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983377042859032,
      "loss": 5.4383,
      "mean_token_accuracy": 0.17094605714082717,
      "num_tokens": 10871933.0,
      "step": 5230
    },
    {
      "entropy": 5.5857017040252686,
      "epoch": 0.47487300435413643,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983337761173936,
      "loss": 5.5079,
      "mean_token_accuracy": 0.16988566964864732,
      "num_tokens": 10881800.0,
      "step": 5235
    },
    {
      "entropy": 5.733130645751953,
      "epoch": 0.4753265602322206,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983298433302843,
      "loss": 5.4496,
      "mean_token_accuracy": 0.17222348898649215,
      "num_tokens": 10891830.0,
      "step": 5240
    },
    {
      "entropy": 5.616529273986816,
      "epoch": 0.4757801161103048,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983259059246572,
      "loss": 5.3783,
      "mean_token_accuracy": 0.18143776208162307,
      "num_tokens": 10901557.0,
      "step": 5245
    },
    {
      "entropy": 5.551515579223633,
      "epoch": 0.47623367198838895,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983219639005934,
      "loss": 5.5078,
      "mean_token_accuracy": 0.1646910712122917,
      "num_tokens": 10912155.0,
      "step": 5250
    },
    {
      "entropy": 5.639435863494873,
      "epoch": 0.47668722786647316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983180172581745,
      "loss": 5.4085,
      "mean_token_accuracy": 0.17542287111282348,
      "num_tokens": 10922291.0,
      "step": 5255
    },
    {
      "entropy": 5.703779172897339,
      "epoch": 0.4771407837445573,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983140659974819,
      "loss": 5.478,
      "mean_token_accuracy": 0.1703026846051216,
      "num_tokens": 10933176.0,
      "step": 5260
    },
    {
      "entropy": 5.653920316696167,
      "epoch": 0.47759433962264153,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983101101185977,
      "loss": 5.5772,
      "mean_token_accuracy": 0.16286463141441346,
      "num_tokens": 10944049.0,
      "step": 5265
    },
    {
      "entropy": 5.619420862197876,
      "epoch": 0.4780478955007257,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983061496216035,
      "loss": 5.4072,
      "mean_token_accuracy": 0.17815487831830978,
      "num_tokens": 10954347.0,
      "step": 5270
    },
    {
      "entropy": 5.614488506317139,
      "epoch": 0.4785014513788099,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983021845065812,
      "loss": 5.4529,
      "mean_token_accuracy": 0.17562718093395233,
      "num_tokens": 10965123.0,
      "step": 5275
    },
    {
      "entropy": 5.580786514282226,
      "epoch": 0.47895500725689405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982982147736128,
      "loss": 5.4492,
      "mean_token_accuracy": 0.1764552414417267,
      "num_tokens": 10976395.0,
      "step": 5280
    },
    {
      "entropy": 5.654614782333374,
      "epoch": 0.4794085631349782,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982942404227805,
      "loss": 5.437,
      "mean_token_accuracy": 0.16845346987247467,
      "num_tokens": 10986797.0,
      "step": 5285
    },
    {
      "entropy": 5.695258712768554,
      "epoch": 0.4798621190130624,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982902614541662,
      "loss": 5.4915,
      "mean_token_accuracy": 0.16865674555301666,
      "num_tokens": 10997229.0,
      "step": 5290
    },
    {
      "entropy": 5.583453416824341,
      "epoch": 0.48031567489114657,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982862778678526,
      "loss": 5.4231,
      "mean_token_accuracy": 0.1796005994081497,
      "num_tokens": 11008180.0,
      "step": 5295
    },
    {
      "entropy": 5.642146348953247,
      "epoch": 0.4807692307692308,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982822896639218,
      "loss": 5.4431,
      "mean_token_accuracy": 0.17733724117279054,
      "num_tokens": 11019248.0,
      "step": 5300
    },
    {
      "entropy": 5.5944068908691404,
      "epoch": 0.48122278664731494,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982782968424563,
      "loss": 5.3671,
      "mean_token_accuracy": 0.1745454654097557,
      "num_tokens": 11029760.0,
      "step": 5305
    },
    {
      "entropy": 5.60427680015564,
      "epoch": 0.48167634252539915,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982742994035388,
      "loss": 5.4575,
      "mean_token_accuracy": 0.17510539293289185,
      "num_tokens": 11039211.0,
      "step": 5310
    },
    {
      "entropy": 5.634851455688477,
      "epoch": 0.4821298984034833,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982702973472518,
      "loss": 5.4293,
      "mean_token_accuracy": 0.17235955595970154,
      "num_tokens": 11049892.0,
      "step": 5315
    },
    {
      "entropy": 5.697804927825928,
      "epoch": 0.4825834542815675,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982662906736783,
      "loss": 5.4824,
      "mean_token_accuracy": 0.17303521037101746,
      "num_tokens": 11060823.0,
      "step": 5320
    },
    {
      "entropy": 5.578884744644165,
      "epoch": 0.48303701015965167,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004982622793829009,
      "loss": 5.4597,
      "mean_token_accuracy": 0.1753349706530571,
      "num_tokens": 11072353.0,
      "step": 5325
    },
    {
      "entropy": 5.626900243759155,
      "epoch": 0.4834905660377358,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982582634750027,
      "loss": 5.4687,
      "mean_token_accuracy": 0.17250408083200455,
      "num_tokens": 11082655.0,
      "step": 5330
    },
    {
      "entropy": 5.67281928062439,
      "epoch": 0.48394412191582004,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982542429500667,
      "loss": 5.4946,
      "mean_token_accuracy": 0.1740492969751358,
      "num_tokens": 11093315.0,
      "step": 5335
    },
    {
      "entropy": 5.53027229309082,
      "epoch": 0.4843976777939042,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982502178081761,
      "loss": 5.3898,
      "mean_token_accuracy": 0.17748946249485015,
      "num_tokens": 11103074.0,
      "step": 5340
    },
    {
      "entropy": 5.641853141784668,
      "epoch": 0.4848512336719884,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982461880494141,
      "loss": 5.4088,
      "mean_token_accuracy": 0.17445126175880432,
      "num_tokens": 11113636.0,
      "step": 5345
    },
    {
      "entropy": 5.629180288314819,
      "epoch": 0.48530478955007256,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498242153673864,
      "loss": 5.3781,
      "mean_token_accuracy": 0.1742332860827446,
      "num_tokens": 11123509.0,
      "step": 5350
    },
    {
      "entropy": 5.594754266738891,
      "epoch": 0.48575834542815677,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982381146816094,
      "loss": 5.4325,
      "mean_token_accuracy": 0.170786514878273,
      "num_tokens": 11133405.0,
      "step": 5355
    },
    {
      "entropy": 5.694587469100952,
      "epoch": 0.4862119013062409,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982340710727336,
      "loss": 5.5156,
      "mean_token_accuracy": 0.16179958879947662,
      "num_tokens": 11144725.0,
      "step": 5360
    },
    {
      "entropy": 5.644291639328003,
      "epoch": 0.48666545718432513,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982300228473203,
      "loss": 5.4028,
      "mean_token_accuracy": 0.17339923679828645,
      "num_tokens": 11155446.0,
      "step": 5365
    },
    {
      "entropy": 5.61758942604065,
      "epoch": 0.4871190130624093,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982259700054533,
      "loss": 5.4547,
      "mean_token_accuracy": 0.18006922155618668,
      "num_tokens": 11165380.0,
      "step": 5370
    },
    {
      "entropy": 5.630903339385986,
      "epoch": 0.48757256894049344,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982219125472163,
      "loss": 5.5089,
      "mean_token_accuracy": 0.1695322260260582,
      "num_tokens": 11175641.0,
      "step": 5375
    },
    {
      "entropy": 5.7712828636169435,
      "epoch": 0.48802612481857766,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982178504726933,
      "loss": 5.6741,
      "mean_token_accuracy": 0.16982539594173432,
      "num_tokens": 11186257.0,
      "step": 5380
    },
    {
      "entropy": 5.703107261657715,
      "epoch": 0.4884796806966618,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982137837819682,
      "loss": 5.4871,
      "mean_token_accuracy": 0.16374782472848892,
      "num_tokens": 11197284.0,
      "step": 5385
    },
    {
      "entropy": 5.680084371566773,
      "epoch": 0.488933236574746,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982097124751252,
      "loss": 5.4638,
      "mean_token_accuracy": 0.17422798573970794,
      "num_tokens": 11208413.0,
      "step": 5390
    },
    {
      "entropy": 5.596026182174683,
      "epoch": 0.4893867924528302,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982056365522486,
      "loss": 5.465,
      "mean_token_accuracy": 0.17459869980812073,
      "num_tokens": 11219016.0,
      "step": 5395
    },
    {
      "entropy": 5.622074842453003,
      "epoch": 0.4898403483309144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982015560134225,
      "loss": 5.425,
      "mean_token_accuracy": 0.17910900115966796,
      "num_tokens": 11228934.0,
      "step": 5400
    },
    {
      "entropy": 5.652887344360352,
      "epoch": 0.49029390420899854,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004981974708587312,
      "loss": 5.346,
      "mean_token_accuracy": 0.1806335538625717,
      "num_tokens": 11239596.0,
      "step": 5405
    },
    {
      "entropy": 5.567725992202758,
      "epoch": 0.49074746008708275,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981933810882594,
      "loss": 5.342,
      "mean_token_accuracy": 0.17635241597890855,
      "num_tokens": 11249560.0,
      "step": 5410
    },
    {
      "entropy": 5.721574974060059,
      "epoch": 0.4912010159651669,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981892867020917,
      "loss": 5.5447,
      "mean_token_accuracy": 0.1738561399281025,
      "num_tokens": 11259787.0,
      "step": 5415
    },
    {
      "entropy": 5.6990868091583256,
      "epoch": 0.49165457184325106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981851877003125,
      "loss": 5.4694,
      "mean_token_accuracy": 0.17295820415019988,
      "num_tokens": 11269717.0,
      "step": 5420
    },
    {
      "entropy": 5.580885410308838,
      "epoch": 0.4921081277213353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981810840830068,
      "loss": 5.4482,
      "mean_token_accuracy": 0.18043429851531984,
      "num_tokens": 11279863.0,
      "step": 5425
    },
    {
      "entropy": 5.6650306224823,
      "epoch": 0.49256168359941943,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981769758502596,
      "loss": 5.5763,
      "mean_token_accuracy": 0.16078406423330308,
      "num_tokens": 11291161.0,
      "step": 5430
    },
    {
      "entropy": 5.6055395126342775,
      "epoch": 0.49301523947750364,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981728630021555,
      "loss": 5.3956,
      "mean_token_accuracy": 0.17572860419750214,
      "num_tokens": 11301151.0,
      "step": 5435
    },
    {
      "entropy": 5.575085067749024,
      "epoch": 0.4934687953555878,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981687455387797,
      "loss": 5.3946,
      "mean_token_accuracy": 0.1716492608189583,
      "num_tokens": 11311466.0,
      "step": 5440
    },
    {
      "entropy": 5.647791528701783,
      "epoch": 0.493922351233672,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981646234602173,
      "loss": 5.3629,
      "mean_token_accuracy": 0.1757534921169281,
      "num_tokens": 11321522.0,
      "step": 5445
    },
    {
      "entropy": 5.705031681060791,
      "epoch": 0.49437590711175616,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981604967665539,
      "loss": 5.4836,
      "mean_token_accuracy": 0.17398695945739745,
      "num_tokens": 11332317.0,
      "step": 5450
    },
    {
      "entropy": 5.611266326904297,
      "epoch": 0.4948294629898404,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981563654578743,
      "loss": 5.4223,
      "mean_token_accuracy": 0.16848283857107163,
      "num_tokens": 11343024.0,
      "step": 5455
    },
    {
      "entropy": 5.585212087631225,
      "epoch": 0.49528301886792453,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981522295342643,
      "loss": 5.381,
      "mean_token_accuracy": 0.17383768558502197,
      "num_tokens": 11353114.0,
      "step": 5460
    },
    {
      "entropy": 5.634729814529419,
      "epoch": 0.4957365747460087,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981480889958092,
      "loss": 5.4311,
      "mean_token_accuracy": 0.17489740252494812,
      "num_tokens": 11363446.0,
      "step": 5465
    },
    {
      "entropy": 5.55680570602417,
      "epoch": 0.4961901306240929,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981439438425949,
      "loss": 5.4316,
      "mean_token_accuracy": 0.17906904518604277,
      "num_tokens": 11373413.0,
      "step": 5470
    },
    {
      "entropy": 5.541783714294434,
      "epoch": 0.49664368650217705,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498139794074707,
      "loss": 5.3189,
      "mean_token_accuracy": 0.180606409907341,
      "num_tokens": 11383142.0,
      "step": 5475
    },
    {
      "entropy": 5.681115627288818,
      "epoch": 0.49709724238026126,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981356396922313,
      "loss": 5.4639,
      "mean_token_accuracy": 0.17151203900575637,
      "num_tokens": 11394635.0,
      "step": 5480
    },
    {
      "entropy": 5.54714994430542,
      "epoch": 0.4975507982583454,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981314806952535,
      "loss": 5.3762,
      "mean_token_accuracy": 0.17595487385988234,
      "num_tokens": 11404442.0,
      "step": 5485
    },
    {
      "entropy": 5.550731897354126,
      "epoch": 0.49800435413642963,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049812731708386,
      "loss": 5.3204,
      "mean_token_accuracy": 0.18081480413675308,
      "num_tokens": 11414217.0,
      "step": 5490
    },
    {
      "entropy": 5.680741786956787,
      "epoch": 0.4984579100145138,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981231488581367,
      "loss": 5.456,
      "mean_token_accuracy": 0.17206876575946808,
      "num_tokens": 11423834.0,
      "step": 5495
    },
    {
      "entropy": 5.532857179641724,
      "epoch": 0.498911465892598,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981189760181699,
      "loss": 5.3341,
      "mean_token_accuracy": 0.17510698288679122,
      "num_tokens": 11433613.0,
      "step": 5500
    },
    {
      "entropy": 5.559586668014527,
      "epoch": 0.49936502177068215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981147985640456,
      "loss": 5.412,
      "mean_token_accuracy": 0.17702824473381043,
      "num_tokens": 11443915.0,
      "step": 5505
    },
    {
      "entropy": 5.699520635604858,
      "epoch": 0.4998185776487663,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004981106164958506,
      "loss": 5.4326,
      "mean_token_accuracy": 0.17155378311872482,
      "num_tokens": 11454588.0,
      "step": 5510
    },
    {
      "entropy": 5.580562543869019,
      "epoch": 0.5002721335268505,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981064298136712,
      "loss": 5.4901,
      "mean_token_accuracy": 0.17163256406784058,
      "num_tokens": 11465533.0,
      "step": 5515
    },
    {
      "entropy": 5.621740531921387,
      "epoch": 0.5007256894049347,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981022385175939,
      "loss": 5.4445,
      "mean_token_accuracy": 0.1629154846072197,
      "num_tokens": 11475776.0,
      "step": 5520
    },
    {
      "entropy": 5.647769451141357,
      "epoch": 0.5011792452830188,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980980426077054,
      "loss": 5.417,
      "mean_token_accuracy": 0.18146403580904008,
      "num_tokens": 11486200.0,
      "step": 5525
    },
    {
      "entropy": 5.586569929122925,
      "epoch": 0.5016328011611031,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980938420840926,
      "loss": 5.4449,
      "mean_token_accuracy": 0.16791134476661682,
      "num_tokens": 11496768.0,
      "step": 5530
    },
    {
      "entropy": 5.628451347351074,
      "epoch": 0.5020863570391872,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980896369468423,
      "loss": 5.4208,
      "mean_token_accuracy": 0.17588136047124864,
      "num_tokens": 11507096.0,
      "step": 5535
    },
    {
      "entropy": 5.637207078933716,
      "epoch": 0.5025399129172714,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980854271960415,
      "loss": 5.4631,
      "mean_token_accuracy": 0.17342848032712938,
      "num_tokens": 11517301.0,
      "step": 5540
    },
    {
      "entropy": 5.666760635375977,
      "epoch": 0.5029934687953556,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980812128317772,
      "loss": 5.5259,
      "mean_token_accuracy": 0.1743517428636551,
      "num_tokens": 11528616.0,
      "step": 5545
    },
    {
      "entropy": 5.676811027526855,
      "epoch": 0.5034470246734397,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004980769938541364,
      "loss": 5.4487,
      "mean_token_accuracy": 0.17927418798208236,
      "num_tokens": 11539344.0,
      "step": 5550
    },
    {
      "entropy": 5.635175704956055,
      "epoch": 0.503900580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980727702632066,
      "loss": 5.4897,
      "mean_token_accuracy": 0.17068908959627152,
      "num_tokens": 11549511.0,
      "step": 5555
    },
    {
      "entropy": 5.6430299282073975,
      "epoch": 0.5043541364296081,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980685420590752,
      "loss": 5.4369,
      "mean_token_accuracy": 0.17500991076231004,
      "num_tokens": 11559425.0,
      "step": 5560
    },
    {
      "entropy": 5.707560014724732,
      "epoch": 0.5048076923076923,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980643092418294,
      "loss": 5.5405,
      "mean_token_accuracy": 0.16627538055181504,
      "num_tokens": 11570406.0,
      "step": 5565
    },
    {
      "entropy": 5.634152126312256,
      "epoch": 0.5052612481857764,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980600718115568,
      "loss": 5.4295,
      "mean_token_accuracy": 0.175694976747036,
      "num_tokens": 11580509.0,
      "step": 5570
    },
    {
      "entropy": 5.635722541809082,
      "epoch": 0.5057148040638607,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980558297683452,
      "loss": 5.4389,
      "mean_token_accuracy": 0.1730503410100937,
      "num_tokens": 11590860.0,
      "step": 5575
    },
    {
      "entropy": 5.6468902111053465,
      "epoch": 0.5061683599419449,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004980515831122822,
      "loss": 5.4468,
      "mean_token_accuracy": 0.17353652268648148,
      "num_tokens": 11601954.0,
      "step": 5580
    },
    {
      "entropy": 5.652958011627197,
      "epoch": 0.506621915820029,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004980473318434555,
      "loss": 5.4462,
      "mean_token_accuracy": 0.17210652977228164,
      "num_tokens": 11612581.0,
      "step": 5585
    },
    {
      "entropy": 5.573605346679687,
      "epoch": 0.5070754716981132,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004980430759619532,
      "loss": 5.3652,
      "mean_token_accuracy": 0.17937352061271666,
      "num_tokens": 11622599.0,
      "step": 5590
    },
    {
      "entropy": 5.543113279342651,
      "epoch": 0.5075290275761973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980388154678633,
      "loss": 5.3908,
      "mean_token_accuracy": 0.17439862638711928,
      "num_tokens": 11633611.0,
      "step": 5595
    },
    {
      "entropy": 5.585327100753784,
      "epoch": 0.5079825834542816,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980345503612738,
      "loss": 5.365,
      "mean_token_accuracy": 0.17676355540752411,
      "num_tokens": 11643159.0,
      "step": 5600
    },
    {
      "entropy": 5.5928394317626955,
      "epoch": 0.5084361393323658,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980302806422729,
      "loss": 5.3449,
      "mean_token_accuracy": 0.177506685256958,
      "num_tokens": 11653174.0,
      "step": 5605
    },
    {
      "entropy": 5.6118120670318605,
      "epoch": 0.5088896952104499,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980260063109491,
      "loss": 5.4038,
      "mean_token_accuracy": 0.17569431215524672,
      "num_tokens": 11662621.0,
      "step": 5610
    },
    {
      "entropy": 5.528201961517334,
      "epoch": 0.5093432510885341,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004980217273673906,
      "loss": 5.3599,
      "mean_token_accuracy": 0.17619221061468124,
      "num_tokens": 11673538.0,
      "step": 5615
    },
    {
      "entropy": 5.644331693649292,
      "epoch": 0.5097968069666183,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980174438116859,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1794475644826889,
      "num_tokens": 11683203.0,
      "step": 5620
    },
    {
      "entropy": 5.552001285552978,
      "epoch": 0.5102503628447025,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980131556439236,
      "loss": 5.3636,
      "mean_token_accuracy": 0.18283542692661287,
      "num_tokens": 11693313.0,
      "step": 5625
    },
    {
      "entropy": 5.612477016448975,
      "epoch": 0.5107039187227866,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980088628641925,
      "loss": 5.4534,
      "mean_token_accuracy": 0.17441582083702087,
      "num_tokens": 11704194.0,
      "step": 5630
    },
    {
      "entropy": 5.56378436088562,
      "epoch": 0.5111574746008708,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980045654725812,
      "loss": 5.3146,
      "mean_token_accuracy": 0.17968470156192778,
      "num_tokens": 11715109.0,
      "step": 5635
    },
    {
      "entropy": 5.534660005569458,
      "epoch": 0.511611030478955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980002634691787,
      "loss": 5.3489,
      "mean_token_accuracy": 0.18387647569179535,
      "num_tokens": 11724961.0,
      "step": 5640
    },
    {
      "entropy": 5.55992374420166,
      "epoch": 0.5120645863570392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497995956854074,
      "loss": 5.3785,
      "mean_token_accuracy": 0.1765351787209511,
      "num_tokens": 11735334.0,
      "step": 5645
    },
    {
      "entropy": 5.569534349441528,
      "epoch": 0.5125181422351234,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979916456273561,
      "loss": 5.3789,
      "mean_token_accuracy": 0.1730036035180092,
      "num_tokens": 11746858.0,
      "step": 5650
    },
    {
      "entropy": 5.661505699157715,
      "epoch": 0.5129716981132075,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979873297891141,
      "loss": 5.4903,
      "mean_token_accuracy": 0.17127009034156798,
      "num_tokens": 11757819.0,
      "step": 5655
    },
    {
      "entropy": 5.618195486068726,
      "epoch": 0.5134252539912917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979830093394372,
      "loss": 5.4053,
      "mean_token_accuracy": 0.17758016139268876,
      "num_tokens": 11767470.0,
      "step": 5660
    },
    {
      "entropy": 5.549928283691406,
      "epoch": 0.513878809869376,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979786842784149,
      "loss": 5.3273,
      "mean_token_accuracy": 0.1853337362408638,
      "num_tokens": 11777060.0,
      "step": 5665
    },
    {
      "entropy": 5.577957963943481,
      "epoch": 0.5143323657474601,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979743546061366,
      "loss": 5.3962,
      "mean_token_accuracy": 0.17992057055234909,
      "num_tokens": 11787176.0,
      "step": 5670
    },
    {
      "entropy": 5.539264678955078,
      "epoch": 0.5147859216255443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979700203226918,
      "loss": 5.2699,
      "mean_token_accuracy": 0.18505061715841292,
      "num_tokens": 11796419.0,
      "step": 5675
    },
    {
      "entropy": 5.661307001113892,
      "epoch": 0.5152394775036284,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979656814281702,
      "loss": 5.4999,
      "mean_token_accuracy": 0.17273307591676712,
      "num_tokens": 11807279.0,
      "step": 5680
    },
    {
      "entropy": 5.561037874221801,
      "epoch": 0.5156930333817126,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979613379226615,
      "loss": 5.3961,
      "mean_token_accuracy": 0.17279176861047746,
      "num_tokens": 11817669.0,
      "step": 5685
    },
    {
      "entropy": 5.53184027671814,
      "epoch": 0.5161465892597968,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979569898062554,
      "loss": 5.3938,
      "mean_token_accuracy": 0.18203373849391938,
      "num_tokens": 11827247.0,
      "step": 5690
    },
    {
      "entropy": 5.618873500823975,
      "epoch": 0.516600145137881,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497952637079042,
      "loss": 5.3803,
      "mean_token_accuracy": 0.17191717773675919,
      "num_tokens": 11838201.0,
      "step": 5695
    },
    {
      "entropy": 5.568843269348145,
      "epoch": 0.5170537010159652,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979482797411114,
      "loss": 5.322,
      "mean_token_accuracy": 0.18213216811418534,
      "num_tokens": 11847561.0,
      "step": 5700
    },
    {
      "entropy": 5.6255693435668945,
      "epoch": 0.5175072568940493,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979439177925534,
      "loss": 5.4481,
      "mean_token_accuracy": 0.180612213909626,
      "num_tokens": 11858272.0,
      "step": 5705
    },
    {
      "entropy": 5.612488603591919,
      "epoch": 0.5179608127721336,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979395512334584,
      "loss": 5.4823,
      "mean_token_accuracy": 0.17200103104114534,
      "num_tokens": 11869660.0,
      "step": 5710
    },
    {
      "entropy": 5.6795634746551515,
      "epoch": 0.5184143686502177,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979351800639166,
      "loss": 5.5096,
      "mean_token_accuracy": 0.1676321730017662,
      "num_tokens": 11880239.0,
      "step": 5715
    },
    {
      "entropy": 5.712689352035523,
      "epoch": 0.5188679245283019,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979308042840186,
      "loss": 5.4816,
      "mean_token_accuracy": 0.16780400574207305,
      "num_tokens": 11890630.0,
      "step": 5720
    },
    {
      "entropy": 5.669008064270019,
      "epoch": 0.519321480406386,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979264238938547,
      "loss": 5.5112,
      "mean_token_accuracy": 0.1703075349330902,
      "num_tokens": 11900948.0,
      "step": 5725
    },
    {
      "entropy": 5.541024541854858,
      "epoch": 0.5197750362844702,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979220388935155,
      "loss": 5.3925,
      "mean_token_accuracy": 0.17539578080177307,
      "num_tokens": 11912528.0,
      "step": 5730
    },
    {
      "entropy": 5.5685014724731445,
      "epoch": 0.5202285921625545,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979176492830918,
      "loss": 5.353,
      "mean_token_accuracy": 0.17433730214834214,
      "num_tokens": 11922211.0,
      "step": 5735
    },
    {
      "entropy": 5.598578691482544,
      "epoch": 0.5206821480406386,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979132550626742,
      "loss": 5.4084,
      "mean_token_accuracy": 0.17142235189676286,
      "num_tokens": 11932520.0,
      "step": 5740
    },
    {
      "entropy": 5.59015007019043,
      "epoch": 0.5211357039187228,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979088562323538,
      "loss": 5.3221,
      "mean_token_accuracy": 0.18398308157920837,
      "num_tokens": 11943312.0,
      "step": 5745
    },
    {
      "entropy": 5.559282016754151,
      "epoch": 0.5215892597968069,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979044527922214,
      "loss": 5.4661,
      "mean_token_accuracy": 0.17238451689481735,
      "num_tokens": 11954455.0,
      "step": 5750
    },
    {
      "entropy": 5.626161718368531,
      "epoch": 0.5220428156748912,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979000447423681,
      "loss": 5.3677,
      "mean_token_accuracy": 0.18028335571289061,
      "num_tokens": 11964650.0,
      "step": 5755
    },
    {
      "entropy": 5.579872035980225,
      "epoch": 0.5224963715529753,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978956320828851,
      "loss": 5.4013,
      "mean_token_accuracy": 0.17466039061546326,
      "num_tokens": 11975134.0,
      "step": 5760
    },
    {
      "entropy": 5.607253074645996,
      "epoch": 0.5229499274310595,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978912148138637,
      "loss": 5.45,
      "mean_token_accuracy": 0.17419570237398146,
      "num_tokens": 11986486.0,
      "step": 5765
    },
    {
      "entropy": 5.655194902420044,
      "epoch": 0.5234034833091437,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497886792935395,
      "loss": 5.4488,
      "mean_token_accuracy": 0.1717308521270752,
      "num_tokens": 11998045.0,
      "step": 5770
    },
    {
      "entropy": 5.663045930862427,
      "epoch": 0.5238570391872278,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978823664475707,
      "loss": 5.4513,
      "mean_token_accuracy": 0.17971017062664033,
      "num_tokens": 12008130.0,
      "step": 5775
    },
    {
      "entropy": 5.596103620529175,
      "epoch": 0.5243105950653121,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978779353504824,
      "loss": 5.5171,
      "mean_token_accuracy": 0.16514042466878892,
      "num_tokens": 12018753.0,
      "step": 5780
    },
    {
      "entropy": 5.621715688705445,
      "epoch": 0.5247641509433962,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978734996442214,
      "loss": 5.3943,
      "mean_token_accuracy": 0.18264525085687638,
      "num_tokens": 12028202.0,
      "step": 5785
    },
    {
      "entropy": 5.640986490249634,
      "epoch": 0.5252177068214804,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978690593288798,
      "loss": 5.4546,
      "mean_token_accuracy": 0.17733746469020845,
      "num_tokens": 12039652.0,
      "step": 5790
    },
    {
      "entropy": 5.637750244140625,
      "epoch": 0.5256712626995645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978646144045491,
      "loss": 5.3854,
      "mean_token_accuracy": 0.17197972387075425,
      "num_tokens": 12051269.0,
      "step": 5795
    },
    {
      "entropy": 5.482382678985596,
      "epoch": 0.5261248185776488,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978601648713215,
      "loss": 5.2795,
      "mean_token_accuracy": 0.17774615287780762,
      "num_tokens": 12060647.0,
      "step": 5800
    },
    {
      "entropy": 5.569826936721801,
      "epoch": 0.526578374455733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978557107292889,
      "loss": 5.408,
      "mean_token_accuracy": 0.17919009327888488,
      "num_tokens": 12071669.0,
      "step": 5805
    },
    {
      "entropy": 5.547264099121094,
      "epoch": 0.5270319303338171,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978512519785433,
      "loss": 5.282,
      "mean_token_accuracy": 0.18562869876623153,
      "num_tokens": 12081971.0,
      "step": 5810
    },
    {
      "entropy": 5.5526330947875975,
      "epoch": 0.5274854862119013,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497846788619177,
      "loss": 5.4195,
      "mean_token_accuracy": 0.17687285095453262,
      "num_tokens": 12092465.0,
      "step": 5815
    },
    {
      "entropy": 5.617220735549926,
      "epoch": 0.5279390420899854,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978423206512825,
      "loss": 5.3735,
      "mean_token_accuracy": 0.17665890902280806,
      "num_tokens": 12101752.0,
      "step": 5820
    },
    {
      "entropy": 5.661663913726807,
      "epoch": 0.5283925979680697,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497837848074952,
      "loss": 5.456,
      "mean_token_accuracy": 0.17533332258462905,
      "num_tokens": 12111805.0,
      "step": 5825
    },
    {
      "entropy": 5.665722894668579,
      "epoch": 0.5288461538461539,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497833370890278,
      "loss": 5.4739,
      "mean_token_accuracy": 0.17663151323795317,
      "num_tokens": 12122511.0,
      "step": 5830
    },
    {
      "entropy": 5.596258163452148,
      "epoch": 0.529299709724238,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978288890973529,
      "loss": 5.3516,
      "mean_token_accuracy": 0.17615063041448592,
      "num_tokens": 12132510.0,
      "step": 5835
    },
    {
      "entropy": 5.578837966918945,
      "epoch": 0.5297532656023222,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978244026962698,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1739246740937233,
      "num_tokens": 12142419.0,
      "step": 5840
    },
    {
      "entropy": 5.594649934768677,
      "epoch": 0.5302068214804064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978199116871212,
      "loss": 5.4149,
      "mean_token_accuracy": 0.17248719334602355,
      "num_tokens": 12152759.0,
      "step": 5845
    },
    {
      "entropy": 5.554829692840576,
      "epoch": 0.5306603773584906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049781541607,
      "loss": 5.3653,
      "mean_token_accuracy": 0.17687786668539046,
      "num_tokens": 12162758.0,
      "step": 5850
    },
    {
      "entropy": 5.583818531036377,
      "epoch": 0.5311139332365747,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978109158449993,
      "loss": 5.4229,
      "mean_token_accuracy": 0.17397976219654082,
      "num_tokens": 12172664.0,
      "step": 5855
    },
    {
      "entropy": 5.605308771133423,
      "epoch": 0.5315674891146589,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978064110122121,
      "loss": 5.3847,
      "mean_token_accuracy": 0.17937863022089004,
      "num_tokens": 12182028.0,
      "step": 5860
    },
    {
      "entropy": 5.624365615844726,
      "epoch": 0.532021044992743,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978019015717314,
      "loss": 5.4322,
      "mean_token_accuracy": 0.17571227252483368,
      "num_tokens": 12192168.0,
      "step": 5865
    },
    {
      "entropy": 5.526874589920044,
      "epoch": 0.5324746008708273,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977973875236508,
      "loss": 5.3325,
      "mean_token_accuracy": 0.17963843196630477,
      "num_tokens": 12202889.0,
      "step": 5870
    },
    {
      "entropy": 5.54812593460083,
      "epoch": 0.5329281567489115,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977928688680634,
      "loss": 5.3732,
      "mean_token_accuracy": 0.18322223126888276,
      "num_tokens": 12212514.0,
      "step": 5875
    },
    {
      "entropy": 5.584365177154541,
      "epoch": 0.5333817126269956,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977883456050626,
      "loss": 5.247,
      "mean_token_accuracy": 0.1861517012119293,
      "num_tokens": 12222353.0,
      "step": 5880
    },
    {
      "entropy": 5.6489053726196286,
      "epoch": 0.5338352685050798,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497783817734742,
      "loss": 5.5219,
      "mean_token_accuracy": 0.16939357966184615,
      "num_tokens": 12232845.0,
      "step": 5885
    },
    {
      "entropy": 5.61934814453125,
      "epoch": 0.534288824383164,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977792852571954,
      "loss": 5.45,
      "mean_token_accuracy": 0.16973866671323776,
      "num_tokens": 12242950.0,
      "step": 5890
    },
    {
      "entropy": 5.633284950256348,
      "epoch": 0.5347423802612482,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977747481725164,
      "loss": 5.4321,
      "mean_token_accuracy": 0.1735300526022911,
      "num_tokens": 12253132.0,
      "step": 5895
    },
    {
      "entropy": 5.599405288696289,
      "epoch": 0.5351959361393324,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977702064807987,
      "loss": 5.3785,
      "mean_token_accuracy": 0.17767370939254762,
      "num_tokens": 12263214.0,
      "step": 5900
    },
    {
      "entropy": 5.526389789581299,
      "epoch": 0.5356494920174165,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977656601821366,
      "loss": 5.3095,
      "mean_token_accuracy": 0.18471042364835738,
      "num_tokens": 12273250.0,
      "step": 5905
    },
    {
      "entropy": 5.716156530380249,
      "epoch": 0.5361030478955007,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977611092766238,
      "loss": 5.4942,
      "mean_token_accuracy": 0.1808107778429985,
      "num_tokens": 12283767.0,
      "step": 5910
    },
    {
      "entropy": 5.523726463317871,
      "epoch": 0.5365566037735849,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977565537643544,
      "loss": 5.3421,
      "mean_token_accuracy": 0.17266201227903366,
      "num_tokens": 12294612.0,
      "step": 5915
    },
    {
      "entropy": 5.554253149032593,
      "epoch": 0.5370101596516691,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977519936454227,
      "loss": 5.4225,
      "mean_token_accuracy": 0.18037798404693603,
      "num_tokens": 12305614.0,
      "step": 5920
    },
    {
      "entropy": 5.550849533081054,
      "epoch": 0.5374637155297532,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977474289199231,
      "loss": 5.2966,
      "mean_token_accuracy": 0.17565731406211854,
      "num_tokens": 12316152.0,
      "step": 5925
    },
    {
      "entropy": 5.522589588165284,
      "epoch": 0.5379172714078374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977428595879499,
      "loss": 5.3751,
      "mean_token_accuracy": 0.17792995423078536,
      "num_tokens": 12325940.0,
      "step": 5930
    },
    {
      "entropy": 5.570098352432251,
      "epoch": 0.5383708272859217,
      "grad_norm": 1.125,
      "learning_rate": 0.0004977382856495975,
      "loss": 5.3606,
      "mean_token_accuracy": 0.17704337388277053,
      "num_tokens": 12335482.0,
      "step": 5935
    },
    {
      "entropy": 5.6550994396209715,
      "epoch": 0.5388243831640058,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977337071049606,
      "loss": 5.4564,
      "mean_token_accuracy": 0.17251523584127426,
      "num_tokens": 12346386.0,
      "step": 5940
    },
    {
      "entropy": 5.668819808959961,
      "epoch": 0.53927793904209,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497729123954134,
      "loss": 5.4197,
      "mean_token_accuracy": 0.1771680533885956,
      "num_tokens": 12356282.0,
      "step": 5945
    },
    {
      "entropy": 5.580939483642578,
      "epoch": 0.5397314949201741,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977245361972123,
      "loss": 5.3956,
      "mean_token_accuracy": 0.17961778491735458,
      "num_tokens": 12366994.0,
      "step": 5950
    },
    {
      "entropy": 5.550912141799927,
      "epoch": 0.5401850507982583,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977199438342904,
      "loss": 5.2834,
      "mean_token_accuracy": 0.1829832002520561,
      "num_tokens": 12377524.0,
      "step": 5955
    },
    {
      "entropy": 5.6025049686431885,
      "epoch": 0.5406386066763426,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977153468654634,
      "loss": 5.4502,
      "mean_token_accuracy": 0.18007041662931442,
      "num_tokens": 12387593.0,
      "step": 5960
    },
    {
      "entropy": 5.677999258041382,
      "epoch": 0.5410921625544267,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004977107452908262,
      "loss": 5.4079,
      "mean_token_accuracy": 0.18105369359254836,
      "num_tokens": 12398732.0,
      "step": 5965
    },
    {
      "entropy": 5.5877197265625,
      "epoch": 0.5415457184325109,
      "grad_norm": 1.0,
      "learning_rate": 0.000497706139110474,
      "loss": 5.405,
      "mean_token_accuracy": 0.18087859451770782,
      "num_tokens": 12409821.0,
      "step": 5970
    },
    {
      "entropy": 5.5568990230560305,
      "epoch": 0.541999274310595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977015283245021,
      "loss": 5.4217,
      "mean_token_accuracy": 0.17138040512800218,
      "num_tokens": 12420044.0,
      "step": 5975
    },
    {
      "entropy": 5.519027042388916,
      "epoch": 0.5424528301886793,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976969129330059,
      "loss": 5.3454,
      "mean_token_accuracy": 0.1738667070865631,
      "num_tokens": 12430980.0,
      "step": 5980
    },
    {
      "entropy": 5.6030182361602785,
      "epoch": 0.5429063860667634,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004976922929360808,
      "loss": 5.3948,
      "mean_token_accuracy": 0.18232907652854918,
      "num_tokens": 12441999.0,
      "step": 5985
    },
    {
      "entropy": 5.621866083145141,
      "epoch": 0.5433599419448476,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976876683338222,
      "loss": 5.4362,
      "mean_token_accuracy": 0.17455211877822877,
      "num_tokens": 12452837.0,
      "step": 5990
    },
    {
      "entropy": 5.572230195999145,
      "epoch": 0.5438134978229318,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497683039126326,
      "loss": 5.3547,
      "mean_token_accuracy": 0.18271825313568116,
      "num_tokens": 12462142.0,
      "step": 5995
    },
    {
      "entropy": 5.600415372848511,
      "epoch": 0.5442670537010159,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976784053136878,
      "loss": 5.4043,
      "mean_token_accuracy": 0.17925617694854737,
      "num_tokens": 12472756.0,
      "step": 6000
    },
    {
      "epoch": 0.5442670537010159,
      "eval_entropy": 5.379152549122254,
      "eval_loss": 5.41491174697876,
      "eval_mean_token_accuracy": 0.18314881672433733,
      "eval_num_tokens": 12472756.0,
      "eval_runtime": 20.6957,
      "eval_samples_per_second": 1708.571,
      "eval_steps_per_second": 213.571,
      "step": 6000
    },
    {
      "entropy": 5.644152069091797,
      "epoch": 0.5447206095791002,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976737668960034,
      "loss": 5.5029,
      "mean_token_accuracy": 0.17155803442001344,
      "num_tokens": 12483115.0,
      "step": 6005
    },
    {
      "entropy": 5.652310562133789,
      "epoch": 0.5451741654571843,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976691238733688,
      "loss": 5.3646,
      "mean_token_accuracy": 0.17381614148616792,
      "num_tokens": 12493280.0,
      "step": 6010
    },
    {
      "entropy": 5.596336126327515,
      "epoch": 0.5456277213352685,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049766447624588,
      "loss": 5.5069,
      "mean_token_accuracy": 0.1748632788658142,
      "num_tokens": 12503665.0,
      "step": 6015
    },
    {
      "entropy": 5.5472839832305905,
      "epoch": 0.5460812772133526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497659824013633,
      "loss": 5.4058,
      "mean_token_accuracy": 0.17200566828250885,
      "num_tokens": 12513381.0,
      "step": 6020
    },
    {
      "entropy": 5.680693578720093,
      "epoch": 0.5465348330914369,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976551671767241,
      "loss": 5.4393,
      "mean_token_accuracy": 0.17311064749956132,
      "num_tokens": 12524121.0,
      "step": 6025
    },
    {
      "entropy": 5.601163005828857,
      "epoch": 0.5469883889695211,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976505057352496,
      "loss": 5.4147,
      "mean_token_accuracy": 0.18124686181545258,
      "num_tokens": 12533972.0,
      "step": 6030
    },
    {
      "entropy": 5.644373226165771,
      "epoch": 0.5474419448476052,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976458396893059,
      "loss": 5.5213,
      "mean_token_accuracy": 0.16715265363454818,
      "num_tokens": 12544110.0,
      "step": 6035
    },
    {
      "entropy": 5.5955469608306885,
      "epoch": 0.5478955007256894,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976411690389896,
      "loss": 5.3403,
      "mean_token_accuracy": 0.17307349890470505,
      "num_tokens": 12554273.0,
      "step": 6040
    },
    {
      "entropy": 5.661379146575928,
      "epoch": 0.5483490566037735,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497636493784397,
      "loss": 5.4215,
      "mean_token_accuracy": 0.17500090450048447,
      "num_tokens": 12564633.0,
      "step": 6045
    },
    {
      "entropy": 5.586629152297974,
      "epoch": 0.5488026124818578,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976318139256251,
      "loss": 5.3537,
      "mean_token_accuracy": 0.178970268368721,
      "num_tokens": 12573790.0,
      "step": 6050
    },
    {
      "entropy": 5.6261004447937015,
      "epoch": 0.549256168359942,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976271294627706,
      "loss": 5.4317,
      "mean_token_accuracy": 0.17064266204833983,
      "num_tokens": 12584265.0,
      "step": 6055
    },
    {
      "entropy": 5.570170545578003,
      "epoch": 0.5497097242380261,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976224403959303,
      "loss": 5.2914,
      "mean_token_accuracy": 0.1787312373518944,
      "num_tokens": 12594413.0,
      "step": 6060
    },
    {
      "entropy": 5.510761976242065,
      "epoch": 0.5501632801161103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976177467252012,
      "loss": 5.3683,
      "mean_token_accuracy": 0.18546044677495957,
      "num_tokens": 12604637.0,
      "step": 6065
    },
    {
      "entropy": 5.589938545227051,
      "epoch": 0.5506168359941945,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976130484506804,
      "loss": 5.4391,
      "mean_token_accuracy": 0.1787260204553604,
      "num_tokens": 12614873.0,
      "step": 6070
    },
    {
      "entropy": 5.601373863220215,
      "epoch": 0.5510703918722787,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976083455724649,
      "loss": 5.3712,
      "mean_token_accuracy": 0.1769331619143486,
      "num_tokens": 12626182.0,
      "step": 6075
    },
    {
      "entropy": 5.540160465240478,
      "epoch": 0.5515239477503628,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976036380906523,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1818635180592537,
      "num_tokens": 12635418.0,
      "step": 6080
    },
    {
      "entropy": 5.5531182289123535,
      "epoch": 0.551977503628447,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975989260053395,
      "loss": 5.4516,
      "mean_token_accuracy": 0.16748463958501816,
      "num_tokens": 12645942.0,
      "step": 6085
    },
    {
      "entropy": 5.6305876731872555,
      "epoch": 0.5524310595065312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975942093166242,
      "loss": 5.3756,
      "mean_token_accuracy": 0.17721232026815414,
      "num_tokens": 12656460.0,
      "step": 6090
    },
    {
      "entropy": 5.610075330734253,
      "epoch": 0.5528846153846154,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975894880246041,
      "loss": 5.3241,
      "mean_token_accuracy": 0.17855470925569533,
      "num_tokens": 12666916.0,
      "step": 6095
    },
    {
      "entropy": 5.539130449295044,
      "epoch": 0.5533381712626996,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975847621293766,
      "loss": 5.3682,
      "mean_token_accuracy": 0.18464813530445098,
      "num_tokens": 12677562.0,
      "step": 6100
    },
    {
      "entropy": 5.5453942775726315,
      "epoch": 0.5537917271407837,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975800316310394,
      "loss": 5.3443,
      "mean_token_accuracy": 0.1776389956474304,
      "num_tokens": 12687349.0,
      "step": 6105
    },
    {
      "entropy": 5.601655769348144,
      "epoch": 0.5542452830188679,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975752965296906,
      "loss": 5.4005,
      "mean_token_accuracy": 0.17570678889751434,
      "num_tokens": 12697850.0,
      "step": 6110
    },
    {
      "entropy": 5.50960373878479,
      "epoch": 0.5546988388969522,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975705568254278,
      "loss": 5.2554,
      "mean_token_accuracy": 0.1816887453198433,
      "num_tokens": 12708283.0,
      "step": 6115
    },
    {
      "entropy": 5.553288125991822,
      "epoch": 0.5551523947750363,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975658125183493,
      "loss": 5.3604,
      "mean_token_accuracy": 0.1766258791089058,
      "num_tokens": 12718201.0,
      "step": 6120
    },
    {
      "entropy": 5.5428098201751705,
      "epoch": 0.5556059506531205,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497561063608553,
      "loss": 5.3093,
      "mean_token_accuracy": 0.18314361572265625,
      "num_tokens": 12728860.0,
      "step": 6125
    },
    {
      "entropy": 5.565022611618042,
      "epoch": 0.5560595065312046,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004975563100961372,
      "loss": 5.3704,
      "mean_token_accuracy": 0.17591018676757814,
      "num_tokens": 12739007.0,
      "step": 6130
    },
    {
      "entropy": 5.546350717544556,
      "epoch": 0.5565130624092888,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004975515519812002,
      "loss": 5.2309,
      "mean_token_accuracy": 0.1811644896864891,
      "num_tokens": 12749004.0,
      "step": 6135
    },
    {
      "entropy": 5.555451583862305,
      "epoch": 0.556966618287373,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975467892638402,
      "loss": 5.3761,
      "mean_token_accuracy": 0.1802150845527649,
      "num_tokens": 12760085.0,
      "step": 6140
    },
    {
      "entropy": 5.568343782424927,
      "epoch": 0.5574201741654572,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975420219441561,
      "loss": 5.3294,
      "mean_token_accuracy": 0.18599051386117935,
      "num_tokens": 12770303.0,
      "step": 6145
    },
    {
      "entropy": 5.574814367294311,
      "epoch": 0.5578737300435413,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975372500222462,
      "loss": 5.3613,
      "mean_token_accuracy": 0.18204200863838196,
      "num_tokens": 12780334.0,
      "step": 6150
    },
    {
      "entropy": 5.568981695175171,
      "epoch": 0.5583272859216255,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975324734982094,
      "loss": 5.4068,
      "mean_token_accuracy": 0.1729448139667511,
      "num_tokens": 12790980.0,
      "step": 6155
    },
    {
      "entropy": 5.61474723815918,
      "epoch": 0.5587808417997098,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975276923721442,
      "loss": 5.4477,
      "mean_token_accuracy": 0.17725956439971924,
      "num_tokens": 12801713.0,
      "step": 6160
    },
    {
      "entropy": 5.613516807556152,
      "epoch": 0.5592343976777939,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975229066441496,
      "loss": 5.4632,
      "mean_token_accuracy": 0.17113297879695893,
      "num_tokens": 12813040.0,
      "step": 6165
    },
    {
      "entropy": 5.557818269729614,
      "epoch": 0.5596879535558781,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975181163143247,
      "loss": 5.3897,
      "mean_token_accuracy": 0.17339551895856858,
      "num_tokens": 12824048.0,
      "step": 6170
    },
    {
      "entropy": 5.607717227935791,
      "epoch": 0.5601415094339622,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975133213827684,
      "loss": 5.3425,
      "mean_token_accuracy": 0.17274201065301895,
      "num_tokens": 12834297.0,
      "step": 6175
    },
    {
      "entropy": 5.6809532165527346,
      "epoch": 0.5605950653120464,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975085218495798,
      "loss": 5.4172,
      "mean_token_accuracy": 0.17106309533119202,
      "num_tokens": 12844574.0,
      "step": 6180
    },
    {
      "entropy": 5.524085903167725,
      "epoch": 0.5610486211901307,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975037177148585,
      "loss": 5.3589,
      "mean_token_accuracy": 0.1753137394785881,
      "num_tokens": 12854670.0,
      "step": 6185
    },
    {
      "entropy": 5.51950306892395,
      "epoch": 0.5615021770682148,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974989089787035,
      "loss": 5.3713,
      "mean_token_accuracy": 0.18194560408592225,
      "num_tokens": 12865281.0,
      "step": 6190
    },
    {
      "entropy": 5.6557981967926025,
      "epoch": 0.561955732946299,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974940956412143,
      "loss": 5.3725,
      "mean_token_accuracy": 0.18128142803907393,
      "num_tokens": 12875222.0,
      "step": 6195
    },
    {
      "entropy": 5.714055490493775,
      "epoch": 0.5624092888243831,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974892777024905,
      "loss": 5.4478,
      "mean_token_accuracy": 0.1658184453845024,
      "num_tokens": 12885079.0,
      "step": 6200
    },
    {
      "entropy": 5.553147268295288,
      "epoch": 0.5628628447024674,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004974844551626318,
      "loss": 5.3775,
      "mean_token_accuracy": 0.18091828525066375,
      "num_tokens": 12896402.0,
      "step": 6205
    },
    {
      "entropy": 5.556194925308228,
      "epoch": 0.5633164005805515,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974796280217379,
      "loss": 5.3496,
      "mean_token_accuracy": 0.18283513188362122,
      "num_tokens": 12906403.0,
      "step": 6210
    },
    {
      "entropy": 5.598826265335083,
      "epoch": 0.5637699564586357,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974747962799086,
      "loss": 5.3687,
      "mean_token_accuracy": 0.1756142258644104,
      "num_tokens": 12917061.0,
      "step": 6215
    },
    {
      "entropy": 5.568496322631836,
      "epoch": 0.5642235123367199,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974699599372437,
      "loss": 5.3911,
      "mean_token_accuracy": 0.1795428067445755,
      "num_tokens": 12927182.0,
      "step": 6220
    },
    {
      "entropy": 5.529246139526367,
      "epoch": 0.564677068214804,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974651189938434,
      "loss": 5.361,
      "mean_token_accuracy": 0.17926921993494033,
      "num_tokens": 12937371.0,
      "step": 6225
    },
    {
      "entropy": 5.573128747940063,
      "epoch": 0.5651306240928883,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974602734498079,
      "loss": 5.4119,
      "mean_token_accuracy": 0.1759198173880577,
      "num_tokens": 12948569.0,
      "step": 6230
    },
    {
      "entropy": 5.610108995437622,
      "epoch": 0.5655841799709724,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000497455423305237,
      "loss": 5.3696,
      "mean_token_accuracy": 0.1703425258398056,
      "num_tokens": 12958503.0,
      "step": 6235
    },
    {
      "entropy": 5.612152004241944,
      "epoch": 0.5660377358490566,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974505685602314,
      "loss": 5.3857,
      "mean_token_accuracy": 0.1823065236210823,
      "num_tokens": 12968714.0,
      "step": 6240
    },
    {
      "entropy": 5.555819892883301,
      "epoch": 0.5664912917271407,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974457092148912,
      "loss": 5.4102,
      "mean_token_accuracy": 0.17348227500915528,
      "num_tokens": 12978857.0,
      "step": 6245
    },
    {
      "entropy": 5.670070648193359,
      "epoch": 0.566944847605225,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974408452693172,
      "loss": 5.5099,
      "mean_token_accuracy": 0.17125375419855118,
      "num_tokens": 12989863.0,
      "step": 6250
    },
    {
      "entropy": 5.5948718070983885,
      "epoch": 0.5673984034833092,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974359767236099,
      "loss": 5.3118,
      "mean_token_accuracy": 0.17757333517074586,
      "num_tokens": 13000307.0,
      "step": 6255
    },
    {
      "entropy": 5.592511749267578,
      "epoch": 0.5678519593613933,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004974311035778698,
      "loss": 5.3883,
      "mean_token_accuracy": 0.18470816314220428,
      "num_tokens": 13009053.0,
      "step": 6260
    },
    {
      "entropy": 5.578414249420166,
      "epoch": 0.5683055152394775,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974262258321979,
      "loss": 5.4205,
      "mean_token_accuracy": 0.172482830286026,
      "num_tokens": 13019136.0,
      "step": 6265
    },
    {
      "entropy": 5.6188208103179935,
      "epoch": 0.5687590711175616,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004974213434866949,
      "loss": 5.4601,
      "mean_token_accuracy": 0.17041391432285308,
      "num_tokens": 13030356.0,
      "step": 6270
    },
    {
      "entropy": 5.613703393936158,
      "epoch": 0.5692126269956459,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974164565414617,
      "loss": 5.4759,
      "mean_token_accuracy": 0.17600406408309938,
      "num_tokens": 13040477.0,
      "step": 6275
    },
    {
      "entropy": 5.604294204711914,
      "epoch": 0.56966618287373,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974115649965998,
      "loss": 5.3422,
      "mean_token_accuracy": 0.18165747225284576,
      "num_tokens": 13051755.0,
      "step": 6280
    },
    {
      "entropy": 5.5737049102783205,
      "epoch": 0.5701197387518142,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974066688522098,
      "loss": 5.3627,
      "mean_token_accuracy": 0.1862332984805107,
      "num_tokens": 13061670.0,
      "step": 6285
    },
    {
      "entropy": 5.556718826293945,
      "epoch": 0.5705732946298984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974017681083934,
      "loss": 5.3469,
      "mean_token_accuracy": 0.18105981051921843,
      "num_tokens": 13070917.0,
      "step": 6290
    },
    {
      "entropy": 5.554030179977417,
      "epoch": 0.5710268505079826,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004973968627652516,
      "loss": 5.3933,
      "mean_token_accuracy": 0.17703206092119217,
      "num_tokens": 13080577.0,
      "step": 6295
    },
    {
      "entropy": 5.614200782775879,
      "epoch": 0.5714804063860668,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497391952822886,
      "loss": 5.4314,
      "mean_token_accuracy": 0.17582101076841355,
      "num_tokens": 13091446.0,
      "step": 6300
    },
    {
      "entropy": 5.623776531219482,
      "epoch": 0.5719339622641509,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973870382813983,
      "loss": 5.4603,
      "mean_token_accuracy": 0.17539089173078537,
      "num_tokens": 13102752.0,
      "step": 6305
    },
    {
      "entropy": 5.665589141845703,
      "epoch": 0.5723875181422351,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973821191408899,
      "loss": 5.465,
      "mean_token_accuracy": 0.17193345874547958,
      "num_tokens": 13113944.0,
      "step": 6310
    },
    {
      "entropy": 5.614248418807984,
      "epoch": 0.5728410740203193,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973771954014626,
      "loss": 5.399,
      "mean_token_accuracy": 0.17831117212772368,
      "num_tokens": 13124316.0,
      "step": 6315
    },
    {
      "entropy": 5.65424370765686,
      "epoch": 0.5732946298984035,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004973722670632182,
      "loss": 5.4797,
      "mean_token_accuracy": 0.17098383456468583,
      "num_tokens": 13133567.0,
      "step": 6320
    },
    {
      "entropy": 5.5662015914917,
      "epoch": 0.5737481857764877,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973673341262587,
      "loss": 5.3917,
      "mean_token_accuracy": 0.1759439915418625,
      "num_tokens": 13144279.0,
      "step": 6325
    },
    {
      "entropy": 5.604485321044922,
      "epoch": 0.5742017416545718,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497362396590686,
      "loss": 5.4699,
      "mean_token_accuracy": 0.17275095134973525,
      "num_tokens": 13154463.0,
      "step": 6330
    },
    {
      "entropy": 5.647492170333862,
      "epoch": 0.574655297532656,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973574544566024,
      "loss": 5.4281,
      "mean_token_accuracy": 0.16383579671382903,
      "num_tokens": 13165565.0,
      "step": 6335
    },
    {
      "entropy": 5.578436040878296,
      "epoch": 0.5751088534107403,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973525077241098,
      "loss": 5.3417,
      "mean_token_accuracy": 0.18116174638271332,
      "num_tokens": 13175839.0,
      "step": 6340
    },
    {
      "entropy": 5.509803342819214,
      "epoch": 0.5755624092888244,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973475563933109,
      "loss": 5.3268,
      "mean_token_accuracy": 0.17926134169101715,
      "num_tokens": 13186372.0,
      "step": 6345
    },
    {
      "entropy": 5.569689559936523,
      "epoch": 0.5760159651669086,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973426004643077,
      "loss": 5.3894,
      "mean_token_accuracy": 0.1798923596739769,
      "num_tokens": 13196259.0,
      "step": 6350
    },
    {
      "entropy": 5.578006315231323,
      "epoch": 0.5764695210449927,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973376399372029,
      "loss": 5.3612,
      "mean_token_accuracy": 0.1823232352733612,
      "num_tokens": 13206497.0,
      "step": 6355
    },
    {
      "entropy": 5.582531070709228,
      "epoch": 0.5769230769230769,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973326748120991,
      "loss": 5.4179,
      "mean_token_accuracy": 0.17883315533399582,
      "num_tokens": 13217188.0,
      "step": 6360
    },
    {
      "entropy": 5.578738641738892,
      "epoch": 0.5773766328011611,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973277050890988,
      "loss": 5.4042,
      "mean_token_accuracy": 0.17636267542839051,
      "num_tokens": 13226544.0,
      "step": 6365
    },
    {
      "entropy": 5.636392402648926,
      "epoch": 0.5778301886792453,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973227307683049,
      "loss": 5.4464,
      "mean_token_accuracy": 0.17221278697252274,
      "num_tokens": 13236572.0,
      "step": 6370
    },
    {
      "entropy": 5.705385446548462,
      "epoch": 0.5782837445573294,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973177518498204,
      "loss": 5.4333,
      "mean_token_accuracy": 0.1773147389292717,
      "num_tokens": 13247958.0,
      "step": 6375
    },
    {
      "entropy": 5.586960983276367,
      "epoch": 0.5787373004354136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973127683337482,
      "loss": 5.342,
      "mean_token_accuracy": 0.17918977737426758,
      "num_tokens": 13258255.0,
      "step": 6380
    },
    {
      "entropy": 5.52049298286438,
      "epoch": 0.5791908563134979,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973077802201911,
      "loss": 5.3718,
      "mean_token_accuracy": 0.17743289172649385,
      "num_tokens": 13268416.0,
      "step": 6385
    },
    {
      "entropy": 5.5609814643859865,
      "epoch": 0.579644412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973027875092526,
      "loss": 5.346,
      "mean_token_accuracy": 0.17577312141656876,
      "num_tokens": 13278087.0,
      "step": 6390
    },
    {
      "entropy": 5.551668643951416,
      "epoch": 0.5800979680696662,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972977902010357,
      "loss": 5.2501,
      "mean_token_accuracy": 0.18471986949443817,
      "num_tokens": 13288121.0,
      "step": 6395
    },
    {
      "entropy": 5.666141223907471,
      "epoch": 0.5805515239477503,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972927882956439,
      "loss": 5.4785,
      "mean_token_accuracy": 0.16452504992485045,
      "num_tokens": 13298620.0,
      "step": 6400
    },
    {
      "entropy": 5.551081323623658,
      "epoch": 0.5810050798258345,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972877817931807,
      "loss": 5.2865,
      "mean_token_accuracy": 0.17660645842552186,
      "num_tokens": 13308413.0,
      "step": 6405
    },
    {
      "entropy": 5.522010469436646,
      "epoch": 0.5814586357039188,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972827706937493,
      "loss": 5.3758,
      "mean_token_accuracy": 0.17947073578834533,
      "num_tokens": 13319387.0,
      "step": 6410
    },
    {
      "entropy": 5.620809888839721,
      "epoch": 0.5819121915820029,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972777549974537,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17694490402936935,
      "num_tokens": 13330494.0,
      "step": 6415
    },
    {
      "entropy": 5.556739234924317,
      "epoch": 0.5823657474600871,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972727347043975,
      "loss": 5.3374,
      "mean_token_accuracy": 0.17858588695526123,
      "num_tokens": 13341009.0,
      "step": 6420
    },
    {
      "entropy": 5.526559066772461,
      "epoch": 0.5828193033381712,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972677098146845,
      "loss": 5.2915,
      "mean_token_accuracy": 0.18030916303396224,
      "num_tokens": 13351269.0,
      "step": 6425
    },
    {
      "entropy": 5.536026811599731,
      "epoch": 0.5832728592162555,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972626803284185,
      "loss": 5.4053,
      "mean_token_accuracy": 0.17801292687654496,
      "num_tokens": 13362160.0,
      "step": 6430
    },
    {
      "entropy": 5.641278648376465,
      "epoch": 0.5837264150943396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972576462457038,
      "loss": 5.4702,
      "mean_token_accuracy": 0.17617461383342742,
      "num_tokens": 13372497.0,
      "step": 6435
    },
    {
      "entropy": 5.673011207580567,
      "epoch": 0.5841799709724238,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972526075666442,
      "loss": 5.3986,
      "mean_token_accuracy": 0.1806288555264473,
      "num_tokens": 13381665.0,
      "step": 6440
    },
    {
      "entropy": 5.523349046707153,
      "epoch": 0.584633526850508,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972475642913442,
      "loss": 5.3026,
      "mean_token_accuracy": 0.1852041155099869,
      "num_tokens": 13392507.0,
      "step": 6445
    },
    {
      "entropy": 5.5107386112213135,
      "epoch": 0.5850870827285921,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972425164199078,
      "loss": 5.2899,
      "mean_token_accuracy": 0.18328872472047805,
      "num_tokens": 13402171.0,
      "step": 6450
    },
    {
      "entropy": 5.651026678085327,
      "epoch": 0.5855406386066764,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972374639524396,
      "loss": 5.5147,
      "mean_token_accuracy": 0.1725245714187622,
      "num_tokens": 13412933.0,
      "step": 6455
    },
    {
      "entropy": 5.612742567062378,
      "epoch": 0.5859941944847605,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000497232406889044,
      "loss": 5.3512,
      "mean_token_accuracy": 0.1728028178215027,
      "num_tokens": 13423563.0,
      "step": 6460
    },
    {
      "entropy": 5.638546085357666,
      "epoch": 0.5864477503628447,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972273452298255,
      "loss": 5.36,
      "mean_token_accuracy": 0.18153707385063172,
      "num_tokens": 13433998.0,
      "step": 6465
    },
    {
      "entropy": 5.532230949401855,
      "epoch": 0.5869013062409288,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972222789748888,
      "loss": 5.3105,
      "mean_token_accuracy": 0.18037805259227752,
      "num_tokens": 13443878.0,
      "step": 6470
    },
    {
      "entropy": 5.536468267440796,
      "epoch": 0.5873548621190131,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004972172081243389,
      "loss": 5.3923,
      "mean_token_accuracy": 0.17271317690610885,
      "num_tokens": 13456051.0,
      "step": 6475
    },
    {
      "entropy": 5.576761245727539,
      "epoch": 0.5878084179970973,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972121326782806,
      "loss": 5.3829,
      "mean_token_accuracy": 0.18032484352588654,
      "num_tokens": 13467276.0,
      "step": 6480
    },
    {
      "entropy": 5.621661853790283,
      "epoch": 0.5882619738751814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972070526368186,
      "loss": 5.3517,
      "mean_token_accuracy": 0.17507115453481675,
      "num_tokens": 13478331.0,
      "step": 6485
    },
    {
      "entropy": 5.605046653747559,
      "epoch": 0.5887155297532656,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972019680000582,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18173345625400544,
      "num_tokens": 13489879.0,
      "step": 6490
    },
    {
      "entropy": 5.548620986938476,
      "epoch": 0.5891690856313497,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971968787681044,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1835034653544426,
      "num_tokens": 13500879.0,
      "step": 6495
    },
    {
      "entropy": 5.66513090133667,
      "epoch": 0.589622641509434,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971917849410627,
      "loss": 5.3983,
      "mean_token_accuracy": 0.1771375998854637,
      "num_tokens": 13511719.0,
      "step": 6500
    },
    {
      "entropy": 5.545218276977539,
      "epoch": 0.5900761973875182,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971866865190381,
      "loss": 5.3364,
      "mean_token_accuracy": 0.18135370761156083,
      "num_tokens": 13521850.0,
      "step": 6505
    },
    {
      "entropy": 5.597626638412476,
      "epoch": 0.5905297532656023,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004971815835021362,
      "loss": 5.3715,
      "mean_token_accuracy": 0.1825823336839676,
      "num_tokens": 13532095.0,
      "step": 6510
    },
    {
      "entropy": 5.5470359325408936,
      "epoch": 0.5909833091436865,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971764758904626,
      "loss": 5.3513,
      "mean_token_accuracy": 0.17569283097982408,
      "num_tokens": 13543133.0,
      "step": 6515
    },
    {
      "entropy": 5.5723918914794925,
      "epoch": 0.5914368650217707,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971713636841229,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1692360073328018,
      "num_tokens": 13553771.0,
      "step": 6520
    },
    {
      "entropy": 5.536196327209472,
      "epoch": 0.5918904208998549,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971662468832228,
      "loss": 5.3591,
      "mean_token_accuracy": 0.18100160360336304,
      "num_tokens": 13564548.0,
      "step": 6525
    },
    {
      "entropy": 5.500099849700928,
      "epoch": 0.592343976777939,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497161125487868,
      "loss": 5.343,
      "mean_token_accuracy": 0.17980441451072693,
      "num_tokens": 13574567.0,
      "step": 6530
    },
    {
      "entropy": 5.531034755706787,
      "epoch": 0.5927975326560232,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971559994981645,
      "loss": 5.3804,
      "mean_token_accuracy": 0.1790984570980072,
      "num_tokens": 13585302.0,
      "step": 6535
    },
    {
      "entropy": 5.616375589370728,
      "epoch": 0.5932510885341074,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971508689142184,
      "loss": 5.4028,
      "mean_token_accuracy": 0.1724355548620224,
      "num_tokens": 13594754.0,
      "step": 6540
    },
    {
      "entropy": 5.548051548004151,
      "epoch": 0.5937046444121916,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971457337361356,
      "loss": 5.3406,
      "mean_token_accuracy": 0.17818880081176758,
      "num_tokens": 13604691.0,
      "step": 6545
    },
    {
      "entropy": 5.455195903778076,
      "epoch": 0.5941582002902758,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971405939640225,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18700708597898483,
      "num_tokens": 13613911.0,
      "step": 6550
    },
    {
      "entropy": 5.672217941284179,
      "epoch": 0.5946117561683599,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971354495979854,
      "loss": 5.456,
      "mean_token_accuracy": 0.17892757952213287,
      "num_tokens": 13624285.0,
      "step": 6555
    },
    {
      "entropy": 5.586663722991943,
      "epoch": 0.5950653120464441,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971303006381305,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1812276229262352,
      "num_tokens": 13635434.0,
      "step": 6560
    },
    {
      "entropy": 5.5121111392974855,
      "epoch": 0.5955188679245284,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971251470845643,
      "loss": 5.361,
      "mean_token_accuracy": 0.17556140720844268,
      "num_tokens": 13645582.0,
      "step": 6565
    },
    {
      "entropy": 5.546522235870361,
      "epoch": 0.5959724238026125,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004971199889373935,
      "loss": 5.3174,
      "mean_token_accuracy": 0.17810338735580444,
      "num_tokens": 13655029.0,
      "step": 6570
    },
    {
      "entropy": 5.550983810424805,
      "epoch": 0.5964259796806967,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971148261967249,
      "loss": 5.309,
      "mean_token_accuracy": 0.1805609330534935,
      "num_tokens": 13665204.0,
      "step": 6575
    },
    {
      "entropy": 5.571804332733154,
      "epoch": 0.5968795355587808,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971096588626649,
      "loss": 5.4215,
      "mean_token_accuracy": 0.17430134862661362,
      "num_tokens": 13676281.0,
      "step": 6580
    },
    {
      "entropy": 5.537905597686768,
      "epoch": 0.597333091436865,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004971044869353206,
      "loss": 5.3613,
      "mean_token_accuracy": 0.18184176832437515,
      "num_tokens": 13687144.0,
      "step": 6585
    },
    {
      "entropy": 5.562548398971558,
      "epoch": 0.5977866473149492,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497099310414799,
      "loss": 5.3553,
      "mean_token_accuracy": 0.18140233010053636,
      "num_tokens": 13696371.0,
      "step": 6590
    },
    {
      "entropy": 5.626462841033936,
      "epoch": 0.5982402031930334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970941293012069,
      "loss": 5.3001,
      "mean_token_accuracy": 0.1769891545176506,
      "num_tokens": 13706900.0,
      "step": 6595
    },
    {
      "entropy": 5.530804586410523,
      "epoch": 0.5986937590711175,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970889435946518,
      "loss": 5.3671,
      "mean_token_accuracy": 0.1814924567937851,
      "num_tokens": 13717874.0,
      "step": 6600
    },
    {
      "entropy": 5.539168167114258,
      "epoch": 0.5991473149492017,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970837532952406,
      "loss": 5.3369,
      "mean_token_accuracy": 0.18162619471549987,
      "num_tokens": 13728020.0,
      "step": 6605
    },
    {
      "entropy": 5.58751859664917,
      "epoch": 0.599600870827286,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970785584030808,
      "loss": 5.3918,
      "mean_token_accuracy": 0.17638065218925475,
      "num_tokens": 13738007.0,
      "step": 6610
    },
    {
      "entropy": 5.566587495803833,
      "epoch": 0.6000544267053701,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970733589182799,
      "loss": 5.3877,
      "mean_token_accuracy": 0.17494157403707505,
      "num_tokens": 13749429.0,
      "step": 6615
    },
    {
      "entropy": 5.558034038543701,
      "epoch": 0.6005079825834543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970681548409452,
      "loss": 5.262,
      "mean_token_accuracy": 0.18570734858512877,
      "num_tokens": 13760279.0,
      "step": 6620
    },
    {
      "entropy": 5.600715589523316,
      "epoch": 0.6009615384615384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970629461711846,
      "loss": 5.4823,
      "mean_token_accuracy": 0.17011196166276932,
      "num_tokens": 13770575.0,
      "step": 6625
    },
    {
      "entropy": 5.498355531692505,
      "epoch": 0.6014150943396226,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970577329091057,
      "loss": 5.3063,
      "mean_token_accuracy": 0.1857804089784622,
      "num_tokens": 13781068.0,
      "step": 6630
    },
    {
      "entropy": 5.5711182117462155,
      "epoch": 0.6018686502177069,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970525150548162,
      "loss": 5.3957,
      "mean_token_accuracy": 0.18031402826309204,
      "num_tokens": 13790745.0,
      "step": 6635
    },
    {
      "entropy": 5.585378217697143,
      "epoch": 0.602322206095791,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970472926084242,
      "loss": 5.3303,
      "mean_token_accuracy": 0.17876851856708526,
      "num_tokens": 13800196.0,
      "step": 6640
    },
    {
      "entropy": 5.521693229675293,
      "epoch": 0.6027757619738752,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970420655700375,
      "loss": 5.2459,
      "mean_token_accuracy": 0.1828948974609375,
      "num_tokens": 13809954.0,
      "step": 6645
    },
    {
      "entropy": 5.5161940574646,
      "epoch": 0.6032293178519593,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970368339397643,
      "loss": 5.4225,
      "mean_token_accuracy": 0.1718314230442047,
      "num_tokens": 13820281.0,
      "step": 6650
    },
    {
      "entropy": 5.634420919418335,
      "epoch": 0.6036828737300436,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004970315977177129,
      "loss": 5.3772,
      "mean_token_accuracy": 0.17908216267824173,
      "num_tokens": 13830614.0,
      "step": 6655
    },
    {
      "entropy": 5.5539734840393065,
      "epoch": 0.6041364296081277,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004970263569039914,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18047489523887633,
      "num_tokens": 13840551.0,
      "step": 6660
    },
    {
      "entropy": 5.62788462638855,
      "epoch": 0.6045899854862119,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970211114987083,
      "loss": 5.4994,
      "mean_token_accuracy": 0.1773001953959465,
      "num_tokens": 13851830.0,
      "step": 6665
    },
    {
      "entropy": 5.5775432109832765,
      "epoch": 0.6050435413642961,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970158615019719,
      "loss": 5.2978,
      "mean_token_accuracy": 0.18300541788339614,
      "num_tokens": 13862299.0,
      "step": 6670
    },
    {
      "entropy": 5.607674980163575,
      "epoch": 0.6054970972423802,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970106069138912,
      "loss": 5.4325,
      "mean_token_accuracy": 0.1705495685338974,
      "num_tokens": 13872852.0,
      "step": 6675
    },
    {
      "entropy": 5.530221462249756,
      "epoch": 0.6059506531204645,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004970053477345743,
      "loss": 5.3784,
      "mean_token_accuracy": 0.18458971232175828,
      "num_tokens": 13883601.0,
      "step": 6680
    },
    {
      "entropy": 5.578758192062378,
      "epoch": 0.6064042089985486,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004970000839641304,
      "loss": 5.3524,
      "mean_token_accuracy": 0.18701042830944062,
      "num_tokens": 13894246.0,
      "step": 6685
    },
    {
      "entropy": 5.521432018280029,
      "epoch": 0.6068577648766328,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969948156026681,
      "loss": 5.3868,
      "mean_token_accuracy": 0.1802896812558174,
      "num_tokens": 13904823.0,
      "step": 6690
    },
    {
      "entropy": 5.647825717926025,
      "epoch": 0.6073113207547169,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969895426502965,
      "loss": 5.403,
      "mean_token_accuracy": 0.1710766464471817,
      "num_tokens": 13915338.0,
      "step": 6695
    },
    {
      "entropy": 5.546239614486694,
      "epoch": 0.6077648766328012,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969842651071246,
      "loss": 5.3611,
      "mean_token_accuracy": 0.17356563061475755,
      "num_tokens": 13925838.0,
      "step": 6700
    },
    {
      "entropy": 5.606841611862182,
      "epoch": 0.6082184325108854,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969789829732615,
      "loss": 5.3653,
      "mean_token_accuracy": 0.17559958398342132,
      "num_tokens": 13936341.0,
      "step": 6705
    },
    {
      "entropy": 5.611944770812988,
      "epoch": 0.6086719883889695,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969736962488164,
      "loss": 5.3672,
      "mean_token_accuracy": 0.1796737715601921,
      "num_tokens": 13945438.0,
      "step": 6710
    },
    {
      "entropy": 5.558967113494873,
      "epoch": 0.6091255442670537,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004969684049338987,
      "loss": 5.2584,
      "mean_token_accuracy": 0.18215785771608353,
      "num_tokens": 13956029.0,
      "step": 6715
    },
    {
      "entropy": 5.596481418609619,
      "epoch": 0.6095791001451378,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496963109028618,
      "loss": 5.3783,
      "mean_token_accuracy": 0.1746341571211815,
      "num_tokens": 13966498.0,
      "step": 6720
    },
    {
      "entropy": 5.53505163192749,
      "epoch": 0.6100326560232221,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969578085330834,
      "loss": 5.3298,
      "mean_token_accuracy": 0.18196677714586257,
      "num_tokens": 13976997.0,
      "step": 6725
    },
    {
      "entropy": 5.615462064743042,
      "epoch": 0.6104862119013063,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969525034474049,
      "loss": 5.3307,
      "mean_token_accuracy": 0.17902466505765915,
      "num_tokens": 13986664.0,
      "step": 6730
    },
    {
      "entropy": 5.498200178146362,
      "epoch": 0.6109397677793904,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496947193771692,
      "loss": 5.3209,
      "mean_token_accuracy": 0.18532713502645493,
      "num_tokens": 13997828.0,
      "step": 6735
    },
    {
      "entropy": 5.508232593536377,
      "epoch": 0.6113933236574746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969418795060548,
      "loss": 5.2761,
      "mean_token_accuracy": 0.17970932871103287,
      "num_tokens": 14008293.0,
      "step": 6740
    },
    {
      "entropy": 5.507388973236084,
      "epoch": 0.6118468795355588,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969365606506027,
      "loss": 5.2646,
      "mean_token_accuracy": 0.18298047930002212,
      "num_tokens": 14018031.0,
      "step": 6745
    },
    {
      "entropy": 5.491404008865357,
      "epoch": 0.612300435413643,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969312372054461,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1883564054965973,
      "num_tokens": 14028013.0,
      "step": 6750
    },
    {
      "entropy": 5.524714374542237,
      "epoch": 0.6127539912917271,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004969259091706948,
      "loss": 5.3723,
      "mean_token_accuracy": 0.17930554449558259,
      "num_tokens": 14039787.0,
      "step": 6755
    },
    {
      "entropy": 5.5531907081604,
      "epoch": 0.6132075471698113,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004969205765464593,
      "loss": 5.2734,
      "mean_token_accuracy": 0.18047384172677994,
      "num_tokens": 14049223.0,
      "step": 6760
    },
    {
      "entropy": 5.597138452529907,
      "epoch": 0.6136611030478955,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969152393328497,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1850910186767578,
      "num_tokens": 14059004.0,
      "step": 6765
    },
    {
      "entropy": 5.561625194549561,
      "epoch": 0.6141146589259797,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969098975299764,
      "loss": 5.4108,
      "mean_token_accuracy": 0.1744755983352661,
      "num_tokens": 14069615.0,
      "step": 6770
    },
    {
      "entropy": 5.550989723205566,
      "epoch": 0.6145682148040639,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969045511379499,
      "loss": 5.3455,
      "mean_token_accuracy": 0.17937582731246948,
      "num_tokens": 14078886.0,
      "step": 6775
    },
    {
      "entropy": 5.646713638305664,
      "epoch": 0.615021770682148,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004968992001568806,
      "loss": 5.4293,
      "mean_token_accuracy": 0.1722983568906784,
      "num_tokens": 14089021.0,
      "step": 6780
    },
    {
      "entropy": 5.55102949142456,
      "epoch": 0.6154753265602322,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968938445868795,
      "loss": 5.235,
      "mean_token_accuracy": 0.19036040008068084,
      "num_tokens": 14098316.0,
      "step": 6785
    },
    {
      "entropy": 5.502194881439209,
      "epoch": 0.6159288824383164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968884844280569,
      "loss": 5.3202,
      "mean_token_accuracy": 0.1710915148258209,
      "num_tokens": 14108874.0,
      "step": 6790
    },
    {
      "entropy": 5.5451150894165036,
      "epoch": 0.6163824383164006,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004968831196805242,
      "loss": 5.3731,
      "mean_token_accuracy": 0.1814247488975525,
      "num_tokens": 14119613.0,
      "step": 6795
    },
    {
      "entropy": 5.577469968795777,
      "epoch": 0.6168359941944848,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968777503443918,
      "loss": 5.3335,
      "mean_token_accuracy": 0.1749727502465248,
      "num_tokens": 14129085.0,
      "step": 6800
    },
    {
      "entropy": 5.4897855758667,
      "epoch": 0.6172895500725689,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968723764197711,
      "loss": 5.2292,
      "mean_token_accuracy": 0.19681656509637832,
      "num_tokens": 14138461.0,
      "step": 6805
    },
    {
      "entropy": 5.5140174388885494,
      "epoch": 0.6177431059506531,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496866997906773,
      "loss": 5.2926,
      "mean_token_accuracy": 0.18786997944116593,
      "num_tokens": 14149126.0,
      "step": 6810
    },
    {
      "entropy": 5.499987840652466,
      "epoch": 0.6181966618287373,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004968616148055089,
      "loss": 5.306,
      "mean_token_accuracy": 0.1850553944706917,
      "num_tokens": 14160068.0,
      "step": 6815
    },
    {
      "entropy": 5.559629440307617,
      "epoch": 0.6186502177068215,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004968562271160901,
      "loss": 5.3981,
      "mean_token_accuracy": 0.18361826986074448,
      "num_tokens": 14170288.0,
      "step": 6820
    },
    {
      "entropy": 5.63637866973877,
      "epoch": 0.6191037735849056,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496850834838628,
      "loss": 5.3922,
      "mean_token_accuracy": 0.17993116676807402,
      "num_tokens": 14181231.0,
      "step": 6825
    },
    {
      "entropy": 5.592602682113648,
      "epoch": 0.6195573294629898,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004968454379732342,
      "loss": 5.294,
      "mean_token_accuracy": 0.18232411295175552,
      "num_tokens": 14191715.0,
      "step": 6830
    },
    {
      "entropy": 5.52168231010437,
      "epoch": 0.6200108853410741,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049684003652002,
      "loss": 5.2998,
      "mean_token_accuracy": 0.17813656479120255,
      "num_tokens": 14203416.0,
      "step": 6835
    },
    {
      "entropy": 5.446109342575073,
      "epoch": 0.6204644412191582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968346304790973,
      "loss": 5.2524,
      "mean_token_accuracy": 0.19115934818983077,
      "num_tokens": 14213658.0,
      "step": 6840
    },
    {
      "entropy": 5.593048906326294,
      "epoch": 0.6209179970972424,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968292198505781,
      "loss": 5.4235,
      "mean_token_accuracy": 0.17608387172222137,
      "num_tokens": 14223956.0,
      "step": 6845
    },
    {
      "entropy": 5.618917846679688,
      "epoch": 0.6213715529753265,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000496823804634574,
      "loss": 5.4129,
      "mean_token_accuracy": 0.17932555377483367,
      "num_tokens": 14236801.0,
      "step": 6850
    },
    {
      "entropy": 5.529294729232788,
      "epoch": 0.6218251088534107,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968183848311972,
      "loss": 5.2253,
      "mean_token_accuracy": 0.1876459464430809,
      "num_tokens": 14247382.0,
      "step": 6855
    },
    {
      "entropy": 5.560374402999878,
      "epoch": 0.622278664731495,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968129604405596,
      "loss": 5.3169,
      "mean_token_accuracy": 0.18302041888237,
      "num_tokens": 14258627.0,
      "step": 6860
    },
    {
      "entropy": 5.464026880264282,
      "epoch": 0.6227322206095791,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968075314627734,
      "loss": 5.3135,
      "mean_token_accuracy": 0.1861542820930481,
      "num_tokens": 14269013.0,
      "step": 6865
    },
    {
      "entropy": 5.536890077590942,
      "epoch": 0.6231857764876633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496802097897951,
      "loss": 5.3071,
      "mean_token_accuracy": 0.18000025004148484,
      "num_tokens": 14279620.0,
      "step": 6870
    },
    {
      "entropy": 5.5682155132293705,
      "epoch": 0.6236393323657474,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004967966597462046,
      "loss": 5.2997,
      "mean_token_accuracy": 0.18479396849870683,
      "num_tokens": 14289448.0,
      "step": 6875
    },
    {
      "entropy": 5.561263513565064,
      "epoch": 0.6240928882438317,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967912170076469,
      "loss": 5.2893,
      "mean_token_accuracy": 0.17907041907310486,
      "num_tokens": 14299999.0,
      "step": 6880
    },
    {
      "entropy": 5.501129531860352,
      "epoch": 0.6245464441219158,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004967857696823901,
      "loss": 5.2876,
      "mean_token_accuracy": 0.18401148319244384,
      "num_tokens": 14310384.0,
      "step": 6885
    },
    {
      "entropy": 5.550896310806275,
      "epoch": 0.625,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967803177705472,
      "loss": 5.4017,
      "mean_token_accuracy": 0.17491451203823088,
      "num_tokens": 14320644.0,
      "step": 6890
    },
    {
      "entropy": 5.607715702056884,
      "epoch": 0.6254535558780842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004967748612722308,
      "loss": 5.3716,
      "mean_token_accuracy": 0.18048068434000014,
      "num_tokens": 14330077.0,
      "step": 6895
    },
    {
      "entropy": 5.613873863220215,
      "epoch": 0.6259071117561683,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967694001875538,
      "loss": 5.4525,
      "mean_token_accuracy": 0.17149625718593597,
      "num_tokens": 14341549.0,
      "step": 6900
    },
    {
      "entropy": 5.585393762588501,
      "epoch": 0.6263606676342526,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496763934516629,
      "loss": 5.3118,
      "mean_token_accuracy": 0.18692418187856674,
      "num_tokens": 14352146.0,
      "step": 6905
    },
    {
      "entropy": 5.495177698135376,
      "epoch": 0.6268142235123367,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004967584642595696,
      "loss": 5.3372,
      "mean_token_accuracy": 0.1760642945766449,
      "num_tokens": 14362521.0,
      "step": 6910
    },
    {
      "entropy": 5.564804315567017,
      "epoch": 0.6272677793904209,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004967529894164887,
      "loss": 5.3685,
      "mean_token_accuracy": 0.1765128716826439,
      "num_tokens": 14373216.0,
      "step": 6915
    },
    {
      "entropy": 5.5402006149292,
      "epoch": 0.627721335268505,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967475099874992,
      "loss": 5.2748,
      "mean_token_accuracy": 0.18698276132345198,
      "num_tokens": 14383104.0,
      "step": 6920
    },
    {
      "entropy": 5.488488435745239,
      "epoch": 0.6281748911465893,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496742025972715,
      "loss": 5.31,
      "mean_token_accuracy": 0.1918911427259445,
      "num_tokens": 14393854.0,
      "step": 6925
    },
    {
      "entropy": 5.564506006240845,
      "epoch": 0.6286284470246735,
      "grad_norm": 1.046875,
      "learning_rate": 0.000496736537372249,
      "loss": 5.2489,
      "mean_token_accuracy": 0.18805742710828782,
      "num_tokens": 14403213.0,
      "step": 6930
    },
    {
      "entropy": 5.641831588745117,
      "epoch": 0.6290820029027576,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004967310441862151,
      "loss": 5.478,
      "mean_token_accuracy": 0.17123783379793167,
      "num_tokens": 14412792.0,
      "step": 6935
    },
    {
      "entropy": 5.584077024459839,
      "epoch": 0.6295355587808418,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967255464147265,
      "loss": 5.3367,
      "mean_token_accuracy": 0.18502951860427858,
      "num_tokens": 14423061.0,
      "step": 6940
    },
    {
      "entropy": 5.563900518417358,
      "epoch": 0.6299891146589259,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967200440578972,
      "loss": 5.2726,
      "mean_token_accuracy": 0.1876469150185585,
      "num_tokens": 14433338.0,
      "step": 6945
    },
    {
      "entropy": 5.481202793121338,
      "epoch": 0.6304426705370102,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967145371158408,
      "loss": 5.2585,
      "mean_token_accuracy": 0.19061126410961152,
      "num_tokens": 14444124.0,
      "step": 6950
    },
    {
      "entropy": 5.514369153976441,
      "epoch": 0.6308962264150944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967090255886714,
      "loss": 5.3887,
      "mean_token_accuracy": 0.18010223805904388,
      "num_tokens": 14453868.0,
      "step": 6955
    },
    {
      "entropy": 5.5535437107086185,
      "epoch": 0.6313497822931785,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004967035094765028,
      "loss": 5.2781,
      "mean_token_accuracy": 0.17700921893119811,
      "num_tokens": 14465015.0,
      "step": 6960
    },
    {
      "entropy": 5.504481458663941,
      "epoch": 0.6318033381712627,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966979887794491,
      "loss": 5.2372,
      "mean_token_accuracy": 0.19093358367681504,
      "num_tokens": 14475523.0,
      "step": 6965
    },
    {
      "entropy": 5.507241201400757,
      "epoch": 0.6322568940493469,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966924634976245,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18237321078777313,
      "num_tokens": 14485523.0,
      "step": 6970
    },
    {
      "entropy": 5.608642530441284,
      "epoch": 0.6327104499274311,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966869336311433,
      "loss": 5.371,
      "mean_token_accuracy": 0.1829627588391304,
      "num_tokens": 14495237.0,
      "step": 6975
    },
    {
      "entropy": 5.584758949279785,
      "epoch": 0.6331640058055152,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966813991801197,
      "loss": 5.5108,
      "mean_token_accuracy": 0.17335500121116637,
      "num_tokens": 14506431.0,
      "step": 6980
    },
    {
      "entropy": 5.615596199035645,
      "epoch": 0.6336175616835994,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966758601446684,
      "loss": 5.4274,
      "mean_token_accuracy": 0.17262813448905945,
      "num_tokens": 14517445.0,
      "step": 6985
    },
    {
      "entropy": 5.597373199462891,
      "epoch": 0.6340711175616836,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966703165249038,
      "loss": 5.3624,
      "mean_token_accuracy": 0.17349915504455565,
      "num_tokens": 14527818.0,
      "step": 6990
    },
    {
      "entropy": 5.553152656555175,
      "epoch": 0.6345246734397678,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966647683209406,
      "loss": 5.4443,
      "mean_token_accuracy": 0.16868529468774796,
      "num_tokens": 14538235.0,
      "step": 6995
    },
    {
      "entropy": 5.693959903717041,
      "epoch": 0.634978229317852,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966592155328935,
      "loss": 5.4539,
      "mean_token_accuracy": 0.1710901975631714,
      "num_tokens": 14549784.0,
      "step": 7000
    },
    {
      "entropy": 5.5767427444458,
      "epoch": 0.6354317851959361,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004966536581608771,
      "loss": 5.4413,
      "mean_token_accuracy": 0.16910024285316466,
      "num_tokens": 14561057.0,
      "step": 7005
    },
    {
      "entropy": 5.555035877227783,
      "epoch": 0.6358853410740203,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966480962050069,
      "loss": 5.3529,
      "mean_token_accuracy": 0.1803895801305771,
      "num_tokens": 14570859.0,
      "step": 7010
    },
    {
      "entropy": 5.552107143402099,
      "epoch": 0.6363388969521045,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966425296653974,
      "loss": 5.353,
      "mean_token_accuracy": 0.1830982819199562,
      "num_tokens": 14581066.0,
      "step": 7015
    },
    {
      "entropy": 5.559674692153931,
      "epoch": 0.6367924528301887,
      "grad_norm": 0.90625,
      "learning_rate": 0.000496636958542164,
      "loss": 5.3239,
      "mean_token_accuracy": 0.17896756231784822,
      "num_tokens": 14590709.0,
      "step": 7020
    },
    {
      "entropy": 5.556455469131469,
      "epoch": 0.6372460087082729,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004966313828354218,
      "loss": 5.3089,
      "mean_token_accuracy": 0.18295130729675294,
      "num_tokens": 14600849.0,
      "step": 7025
    },
    {
      "entropy": 5.481048727035523,
      "epoch": 0.637699564586357,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496625802545286,
      "loss": 5.1954,
      "mean_token_accuracy": 0.18557194024324417,
      "num_tokens": 14611427.0,
      "step": 7030
    },
    {
      "entropy": 5.520011472702026,
      "epoch": 0.6381531204644412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966202176718723,
      "loss": 5.3651,
      "mean_token_accuracy": 0.18566489666700364,
      "num_tokens": 14623860.0,
      "step": 7035
    },
    {
      "entropy": 5.556991815567017,
      "epoch": 0.6386066763425254,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004966146282152959,
      "loss": 5.2855,
      "mean_token_accuracy": 0.1784844383597374,
      "num_tokens": 14633481.0,
      "step": 7040
    },
    {
      "entropy": 5.5632165431976315,
      "epoch": 0.6390602322206096,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966090341756726,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19155942499637604,
      "num_tokens": 14643519.0,
      "step": 7045
    },
    {
      "entropy": 5.432686424255371,
      "epoch": 0.6395137880986937,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496603435553118,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1885633021593094,
      "num_tokens": 14654570.0,
      "step": 7050
    },
    {
      "entropy": 5.48898401260376,
      "epoch": 0.6399673439767779,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965978323477478,
      "loss": 5.3028,
      "mean_token_accuracy": 0.1765861302614212,
      "num_tokens": 14665596.0,
      "step": 7055
    },
    {
      "entropy": 5.582192134857178,
      "epoch": 0.6404208998548622,
      "grad_norm": 1.03125,
      "learning_rate": 0.000496592224559678,
      "loss": 5.2517,
      "mean_token_accuracy": 0.18795465528964997,
      "num_tokens": 14675660.0,
      "step": 7060
    },
    {
      "entropy": 5.563787317276001,
      "epoch": 0.6408744557329463,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965866121890245,
      "loss": 5.4095,
      "mean_token_accuracy": 0.18029894530773163,
      "num_tokens": 14685229.0,
      "step": 7065
    },
    {
      "entropy": 5.557628488540649,
      "epoch": 0.6413280116110305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965809952359034,
      "loss": 5.3122,
      "mean_token_accuracy": 0.17769250571727752,
      "num_tokens": 14695672.0,
      "step": 7070
    },
    {
      "entropy": 5.514695119857788,
      "epoch": 0.6417815674891146,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965753737004309,
      "loss": 5.2615,
      "mean_token_accuracy": 0.18418668806552888,
      "num_tokens": 14705160.0,
      "step": 7075
    },
    {
      "entropy": 5.52557144165039,
      "epoch": 0.6422351233671988,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965697475827231,
      "loss": 5.3097,
      "mean_token_accuracy": 0.18284096121788024,
      "num_tokens": 14715487.0,
      "step": 7080
    },
    {
      "entropy": 5.577736806869507,
      "epoch": 0.6426886792452831,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965641168828965,
      "loss": 5.3544,
      "mean_token_accuracy": 0.18357906490564346,
      "num_tokens": 14726987.0,
      "step": 7085
    },
    {
      "entropy": 5.4941871643066404,
      "epoch": 0.6431422351233672,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965584816010676,
      "loss": 5.3026,
      "mean_token_accuracy": 0.18404889553785325,
      "num_tokens": 14737810.0,
      "step": 7090
    },
    {
      "entropy": 5.579785442352295,
      "epoch": 0.6435957910014514,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965528417373528,
      "loss": 5.3193,
      "mean_token_accuracy": 0.18094680607318878,
      "num_tokens": 14748997.0,
      "step": 7095
    },
    {
      "entropy": 5.553784418106079,
      "epoch": 0.6440493468795355,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965471972918688,
      "loss": 5.3318,
      "mean_token_accuracy": 0.17817866504192353,
      "num_tokens": 14760439.0,
      "step": 7100
    },
    {
      "entropy": 5.445217418670654,
      "epoch": 0.6445029027576198,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004965415482647322,
      "loss": 5.2362,
      "mean_token_accuracy": 0.18102772533893585,
      "num_tokens": 14770828.0,
      "step": 7105
    },
    {
      "entropy": 5.490187788009644,
      "epoch": 0.6449564586357039,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049653589465606,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18126174360513686,
      "num_tokens": 14781753.0,
      "step": 7110
    },
    {
      "entropy": 5.619928598403931,
      "epoch": 0.6454100145137881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000496530236465969,
      "loss": 5.3987,
      "mean_token_accuracy": 0.1711141899228096,
      "num_tokens": 14792241.0,
      "step": 7115
    },
    {
      "entropy": 5.504394292831421,
      "epoch": 0.6458635703918723,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004965245736945763,
      "loss": 5.2172,
      "mean_token_accuracy": 0.1904732972383499,
      "num_tokens": 14802580.0,
      "step": 7120
    },
    {
      "entropy": 5.5241124629974365,
      "epoch": 0.6463171262699564,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965189063419989,
      "loss": 5.4283,
      "mean_token_accuracy": 0.17418646961450576,
      "num_tokens": 14813103.0,
      "step": 7125
    },
    {
      "entropy": 5.571353006362915,
      "epoch": 0.6467706821480407,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965132344083541,
      "loss": 5.3258,
      "mean_token_accuracy": 0.17736517637968063,
      "num_tokens": 14823334.0,
      "step": 7130
    },
    {
      "entropy": 5.56951904296875,
      "epoch": 0.6472242380261248,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965075578937593,
      "loss": 5.336,
      "mean_token_accuracy": 0.17852421551942826,
      "num_tokens": 14834850.0,
      "step": 7135
    },
    {
      "entropy": 5.485752820968628,
      "epoch": 0.647677793904209,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965018767983316,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18051422238349915,
      "num_tokens": 14844276.0,
      "step": 7140
    },
    {
      "entropy": 5.440044736862182,
      "epoch": 0.6481313497822931,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964961911221886,
      "loss": 5.1867,
      "mean_token_accuracy": 0.19364167600870133,
      "num_tokens": 14854352.0,
      "step": 7145
    },
    {
      "entropy": 5.579373216629028,
      "epoch": 0.6485849056603774,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496490500865448,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18060466051101684,
      "num_tokens": 14864395.0,
      "step": 7150
    },
    {
      "entropy": 5.513666248321533,
      "epoch": 0.6490384615384616,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004964848060282275,
      "loss": 5.2513,
      "mean_token_accuracy": 0.18520902544260026,
      "num_tokens": 14874373.0,
      "step": 7155
    },
    {
      "entropy": 5.515260982513428,
      "epoch": 0.6494920174165457,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964791066106447,
      "loss": 5.2496,
      "mean_token_accuracy": 0.17809188216924668,
      "num_tokens": 14884932.0,
      "step": 7160
    },
    {
      "entropy": 5.572841691970825,
      "epoch": 0.6499455732946299,
      "grad_norm": 0.875,
      "learning_rate": 0.0004964734026128175,
      "loss": 5.3991,
      "mean_token_accuracy": 0.17553634494543074,
      "num_tokens": 14896254.0,
      "step": 7165
    },
    {
      "entropy": 5.601190614700317,
      "epoch": 0.650399129172714,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964676940348638,
      "loss": 5.3319,
      "mean_token_accuracy": 0.17750634700059892,
      "num_tokens": 14907392.0,
      "step": 7170
    },
    {
      "entropy": 5.573651218414307,
      "epoch": 0.6508526850507983,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000496461980876902,
      "loss": 5.4395,
      "mean_token_accuracy": 0.17827751934528352,
      "num_tokens": 14917676.0,
      "step": 7175
    },
    {
      "entropy": 5.549517297744751,
      "epoch": 0.6513062409288825,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964562631390498,
      "loss": 5.2987,
      "mean_token_accuracy": 0.18335486352443695,
      "num_tokens": 14928500.0,
      "step": 7180
    },
    {
      "entropy": 5.538741636276245,
      "epoch": 0.6517597968069666,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964505408214257,
      "loss": 5.2492,
      "mean_token_accuracy": 0.18961041122674943,
      "num_tokens": 14938970.0,
      "step": 7185
    },
    {
      "entropy": 5.511117887496948,
      "epoch": 0.6522133526850508,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000496444813924148,
      "loss": 5.311,
      "mean_token_accuracy": 0.17935390323400496,
      "num_tokens": 14949017.0,
      "step": 7190
    },
    {
      "entropy": 5.51752781867981,
      "epoch": 0.652666908563135,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004964390824473351,
      "loss": 5.2404,
      "mean_token_accuracy": 0.18981554210186005,
      "num_tokens": 14958529.0,
      "step": 7195
    },
    {
      "entropy": 5.509324216842652,
      "epoch": 0.6531204644412192,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964333463911056,
      "loss": 5.3481,
      "mean_token_accuracy": 0.1777355194091797,
      "num_tokens": 14968044.0,
      "step": 7200
    },
    {
      "entropy": 5.569104433059692,
      "epoch": 0.6535740203193033,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964276057555779,
      "loss": 5.365,
      "mean_token_accuracy": 0.17823416888713836,
      "num_tokens": 14977867.0,
      "step": 7205
    },
    {
      "entropy": 5.6319859504699705,
      "epoch": 0.6540275761973875,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496421860540871,
      "loss": 5.3719,
      "mean_token_accuracy": 0.1820246011018753,
      "num_tokens": 14986936.0,
      "step": 7210
    },
    {
      "entropy": 5.530765676498413,
      "epoch": 0.6544811320754716,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964161107471036,
      "loss": 5.331,
      "mean_token_accuracy": 0.18487488329410554,
      "num_tokens": 14996681.0,
      "step": 7215
    },
    {
      "entropy": 5.552242612838745,
      "epoch": 0.6549346879535559,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964103563743946,
      "loss": 5.3224,
      "mean_token_accuracy": 0.1786444827914238,
      "num_tokens": 15007100.0,
      "step": 7220
    },
    {
      "entropy": 5.599308681488037,
      "epoch": 0.6553882438316401,
      "grad_norm": 1.0,
      "learning_rate": 0.000496404597422863,
      "loss": 5.3806,
      "mean_token_accuracy": 0.18596629649400712,
      "num_tokens": 15017452.0,
      "step": 7225
    },
    {
      "entropy": 5.516591310501099,
      "epoch": 0.6558417997097242,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496398833892628,
      "loss": 5.2531,
      "mean_token_accuracy": 0.18542855978012085,
      "num_tokens": 15028362.0,
      "step": 7230
    },
    {
      "entropy": 5.581466817855835,
      "epoch": 0.6562953555878084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963930657838086,
      "loss": 5.3469,
      "mean_token_accuracy": 0.17714816331863403,
      "num_tokens": 15038691.0,
      "step": 7235
    },
    {
      "entropy": 5.552061080932617,
      "epoch": 0.6567489114658926,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004963872930965242,
      "loss": 5.3117,
      "mean_token_accuracy": 0.1878270536661148,
      "num_tokens": 15050037.0,
      "step": 7240
    },
    {
      "entropy": 5.552956056594849,
      "epoch": 0.6572024673439768,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963815158308941,
      "loss": 5.2747,
      "mean_token_accuracy": 0.18535771667957307,
      "num_tokens": 15059793.0,
      "step": 7245
    },
    {
      "entropy": 5.5456095218658445,
      "epoch": 0.657656023222061,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004963757339870379,
      "loss": 5.3749,
      "mean_token_accuracy": 0.17560529112815856,
      "num_tokens": 15070500.0,
      "step": 7250
    },
    {
      "entropy": 5.579104375839234,
      "epoch": 0.6581095791001451,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963699475650751,
      "loss": 5.4274,
      "mean_token_accuracy": 0.1789287731051445,
      "num_tokens": 15081109.0,
      "step": 7255
    },
    {
      "entropy": 5.514655733108521,
      "epoch": 0.6585631349782293,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963641565651255,
      "loss": 5.3029,
      "mean_token_accuracy": 0.18567312955856324,
      "num_tokens": 15092315.0,
      "step": 7260
    },
    {
      "entropy": 5.513403844833374,
      "epoch": 0.6590166908563135,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963583609873084,
      "loss": 5.3694,
      "mean_token_accuracy": 0.18358769416809081,
      "num_tokens": 15103957.0,
      "step": 7265
    },
    {
      "entropy": 5.562368488311767,
      "epoch": 0.6594702467343977,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963525608317443,
      "loss": 5.362,
      "mean_token_accuracy": 0.1775441661477089,
      "num_tokens": 15114996.0,
      "step": 7270
    },
    {
      "entropy": 5.581370878219604,
      "epoch": 0.6599238026124818,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004963467560985527,
      "loss": 5.2938,
      "mean_token_accuracy": 0.17815622687339783,
      "num_tokens": 15125098.0,
      "step": 7275
    },
    {
      "entropy": 5.480411434173584,
      "epoch": 0.660377358490566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963409467878538,
      "loss": 5.2984,
      "mean_token_accuracy": 0.1813468173146248,
      "num_tokens": 15134746.0,
      "step": 7280
    },
    {
      "entropy": 5.59424524307251,
      "epoch": 0.6608309143686503,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004963351328997678,
      "loss": 5.3253,
      "mean_token_accuracy": 0.17637452483177185,
      "num_tokens": 15145698.0,
      "step": 7285
    },
    {
      "entropy": 5.577039194107056,
      "epoch": 0.6612844702467344,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963293144344148,
      "loss": 5.3688,
      "mean_token_accuracy": 0.17920732200145723,
      "num_tokens": 15156026.0,
      "step": 7290
    },
    {
      "entropy": 5.529924106597901,
      "epoch": 0.6617380261248186,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963234913919151,
      "loss": 5.3084,
      "mean_token_accuracy": 0.17793872058391572,
      "num_tokens": 15165900.0,
      "step": 7295
    },
    {
      "entropy": 5.5461030960083,
      "epoch": 0.6621915820029027,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963176637723893,
      "loss": 5.3744,
      "mean_token_accuracy": 0.182407346367836,
      "num_tokens": 15176693.0,
      "step": 7300
    },
    {
      "entropy": 5.591224098205567,
      "epoch": 0.6626451378809869,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963118315759578,
      "loss": 5.3901,
      "mean_token_accuracy": 0.16922723799943923,
      "num_tokens": 15186535.0,
      "step": 7305
    },
    {
      "entropy": 5.577974796295166,
      "epoch": 0.6630986937590712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963059948027411,
      "loss": 5.3013,
      "mean_token_accuracy": 0.182105216383934,
      "num_tokens": 15196019.0,
      "step": 7310
    },
    {
      "entropy": 5.474636697769165,
      "epoch": 0.6635522496371553,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963001534528601,
      "loss": 5.1884,
      "mean_token_accuracy": 0.192403119802475,
      "num_tokens": 15205273.0,
      "step": 7315
    },
    {
      "entropy": 5.537912750244141,
      "epoch": 0.6640058055152395,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962943075264355,
      "loss": 5.3325,
      "mean_token_accuracy": 0.1802896723151207,
      "num_tokens": 15216716.0,
      "step": 7320
    },
    {
      "entropy": 5.575633764266968,
      "epoch": 0.6644593613933236,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004962884570235884,
      "loss": 5.3761,
      "mean_token_accuracy": 0.17417128533124923,
      "num_tokens": 15227306.0,
      "step": 7325
    },
    {
      "entropy": 5.511094474792481,
      "epoch": 0.6649129172714079,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962826019444394,
      "loss": 5.2652,
      "mean_token_accuracy": 0.1794412538409233,
      "num_tokens": 15237302.0,
      "step": 7330
    },
    {
      "entropy": 5.517401790618896,
      "epoch": 0.665366473149492,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962767422891099,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18371451795101165,
      "num_tokens": 15247064.0,
      "step": 7335
    },
    {
      "entropy": 5.537144470214844,
      "epoch": 0.6658200290275762,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496270878057721,
      "loss": 5.3427,
      "mean_token_accuracy": 0.18440829068422318,
      "num_tokens": 15257006.0,
      "step": 7340
    },
    {
      "entropy": 5.58497314453125,
      "epoch": 0.6662735849056604,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962650092503939,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1943359598517418,
      "num_tokens": 15266720.0,
      "step": 7345
    },
    {
      "entropy": 5.543534851074218,
      "epoch": 0.6667271407837445,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962591358672502,
      "loss": 5.3427,
      "mean_token_accuracy": 0.18019769638776778,
      "num_tokens": 15277330.0,
      "step": 7350
    },
    {
      "entropy": 5.5373796939849855,
      "epoch": 0.6671806966618288,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496253257908411,
      "loss": 5.271,
      "mean_token_accuracy": 0.1839846909046173,
      "num_tokens": 15287401.0,
      "step": 7355
    },
    {
      "entropy": 5.489133691787719,
      "epoch": 0.6676342525399129,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962473753739981,
      "loss": 5.2292,
      "mean_token_accuracy": 0.1890961855649948,
      "num_tokens": 15297426.0,
      "step": 7360
    },
    {
      "entropy": 5.556583213806152,
      "epoch": 0.6680878084179971,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496241488264133,
      "loss": 5.2272,
      "mean_token_accuracy": 0.1845378175377846,
      "num_tokens": 15308412.0,
      "step": 7365
    },
    {
      "entropy": 5.527732086181641,
      "epoch": 0.6685413642960812,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004962355965789377,
      "loss": 5.2293,
      "mean_token_accuracy": 0.18310726881027223,
      "num_tokens": 15318821.0,
      "step": 7370
    },
    {
      "entropy": 5.476294469833374,
      "epoch": 0.6689949201741655,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962297003185338,
      "loss": 5.2916,
      "mean_token_accuracy": 0.1801626518368721,
      "num_tokens": 15328737.0,
      "step": 7375
    },
    {
      "entropy": 5.622493934631348,
      "epoch": 0.6694484760522497,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962237994830433,
      "loss": 5.393,
      "mean_token_accuracy": 0.17233246713876724,
      "num_tokens": 15339695.0,
      "step": 7380
    },
    {
      "entropy": 5.615177869796753,
      "epoch": 0.6699020319303338,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962178940725883,
      "loss": 5.364,
      "mean_token_accuracy": 0.18082518577575685,
      "num_tokens": 15350270.0,
      "step": 7385
    },
    {
      "entropy": 5.520081806182861,
      "epoch": 0.670355587808418,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962119840872908,
      "loss": 5.3439,
      "mean_token_accuracy": 0.18389475494623184,
      "num_tokens": 15360986.0,
      "step": 7390
    },
    {
      "entropy": 5.516814851760865,
      "epoch": 0.6708091436865021,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004962060695272733,
      "loss": 5.3391,
      "mean_token_accuracy": 0.18298070579767228,
      "num_tokens": 15371786.0,
      "step": 7395
    },
    {
      "entropy": 5.55599799156189,
      "epoch": 0.6712626995645864,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004962001503926577,
      "loss": 5.2993,
      "mean_token_accuracy": 0.1882503256201744,
      "num_tokens": 15382960.0,
      "step": 7400
    },
    {
      "entropy": 5.526637887954712,
      "epoch": 0.6717162554426706,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961942266835667,
      "loss": 5.287,
      "mean_token_accuracy": 0.18534947335720062,
      "num_tokens": 15392807.0,
      "step": 7405
    },
    {
      "entropy": 5.559031534194946,
      "epoch": 0.6721698113207547,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004961882984001227,
      "loss": 5.3307,
      "mean_token_accuracy": 0.1767762467265129,
      "num_tokens": 15403712.0,
      "step": 7410
    },
    {
      "entropy": 5.584656190872193,
      "epoch": 0.6726233671988389,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961823655424484,
      "loss": 5.3916,
      "mean_token_accuracy": 0.17293067425489425,
      "num_tokens": 15414839.0,
      "step": 7415
    },
    {
      "entropy": 5.518201398849487,
      "epoch": 0.6730769230769231,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961764281106664,
      "loss": 5.2838,
      "mean_token_accuracy": 0.1870151072740555,
      "num_tokens": 15425320.0,
      "step": 7420
    },
    {
      "entropy": 5.517269277572632,
      "epoch": 0.6735304789550073,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961704861048995,
      "loss": 5.2887,
      "mean_token_accuracy": 0.17715439349412918,
      "num_tokens": 15435476.0,
      "step": 7425
    },
    {
      "entropy": 5.53749532699585,
      "epoch": 0.6739840348330914,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961645395252705,
      "loss": 5.3947,
      "mean_token_accuracy": 0.17859402149915696,
      "num_tokens": 15445786.0,
      "step": 7430
    },
    {
      "entropy": 5.562483930587769,
      "epoch": 0.6744375907111756,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961585883719025,
      "loss": 5.3762,
      "mean_token_accuracy": 0.18556168675422668,
      "num_tokens": 15455915.0,
      "step": 7435
    },
    {
      "entropy": 5.612591075897217,
      "epoch": 0.6748911465892597,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961526326449186,
      "loss": 5.3614,
      "mean_token_accuracy": 0.1754249468445778,
      "num_tokens": 15466523.0,
      "step": 7440
    },
    {
      "entropy": 5.561124181747436,
      "epoch": 0.675344702467344,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004961466723444419,
      "loss": 5.3425,
      "mean_token_accuracy": 0.18115431517362596,
      "num_tokens": 15476821.0,
      "step": 7445
    },
    {
      "entropy": 5.506720113754272,
      "epoch": 0.6757982583454282,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961407074705955,
      "loss": 5.2759,
      "mean_token_accuracy": 0.1748933583498001,
      "num_tokens": 15487182.0,
      "step": 7450
    },
    {
      "entropy": 5.5438395023345945,
      "epoch": 0.6762518142235123,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961347380235031,
      "loss": 5.267,
      "mean_token_accuracy": 0.18452272266149522,
      "num_tokens": 15497600.0,
      "step": 7455
    },
    {
      "entropy": 5.560304641723633,
      "epoch": 0.6767053701015965,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961287640032879,
      "loss": 5.41,
      "mean_token_accuracy": 0.17247417122125625,
      "num_tokens": 15508278.0,
      "step": 7460
    },
    {
      "entropy": 5.6194220066070555,
      "epoch": 0.6771589259796807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961227854100734,
      "loss": 5.4261,
      "mean_token_accuracy": 0.17710963487625123,
      "num_tokens": 15518458.0,
      "step": 7465
    },
    {
      "entropy": 5.551078796386719,
      "epoch": 0.6776124818577649,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961168022439835,
      "loss": 5.2497,
      "mean_token_accuracy": 0.1808410331606865,
      "num_tokens": 15528738.0,
      "step": 7470
    },
    {
      "entropy": 5.536972951889038,
      "epoch": 0.6780660377358491,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961108145051417,
      "loss": 5.2783,
      "mean_token_accuracy": 0.1874865859746933,
      "num_tokens": 15538656.0,
      "step": 7475
    },
    {
      "entropy": 5.481096410751343,
      "epoch": 0.6785195936139332,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496104822193672,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18348010629415512,
      "num_tokens": 15549285.0,
      "step": 7480
    },
    {
      "entropy": 5.595572376251221,
      "epoch": 0.6789731494920174,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960988253096982,
      "loss": 5.3666,
      "mean_token_accuracy": 0.18583495169878006,
      "num_tokens": 15559766.0,
      "step": 7485
    },
    {
      "entropy": 5.505124187469482,
      "epoch": 0.6794267053701016,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004960928238533443,
      "loss": 5.2299,
      "mean_token_accuracy": 0.19369809031486512,
      "num_tokens": 15570873.0,
      "step": 7490
    },
    {
      "entropy": 5.5578302383422855,
      "epoch": 0.6798802612481858,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004960868178247344,
      "loss": 5.4746,
      "mean_token_accuracy": 0.17800090312957764,
      "num_tokens": 15581229.0,
      "step": 7495
    },
    {
      "entropy": 5.530143356323242,
      "epoch": 0.68033381712627,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960808072239929,
      "loss": 5.3422,
      "mean_token_accuracy": 0.17350580245256425,
      "num_tokens": 15591686.0,
      "step": 7500
    },
    {
      "entropy": 5.646744251251221,
      "epoch": 0.6807873730043541,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960747920512439,
      "loss": 5.3575,
      "mean_token_accuracy": 0.17898468375205995,
      "num_tokens": 15601869.0,
      "step": 7505
    },
    {
      "entropy": 5.519121694564819,
      "epoch": 0.6812409288824384,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000496068772306612,
      "loss": 5.3459,
      "mean_token_accuracy": 0.17619211375713348,
      "num_tokens": 15612946.0,
      "step": 7510
    },
    {
      "entropy": 5.554516172409057,
      "epoch": 0.6816944847605225,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960627479902214,
      "loss": 5.3769,
      "mean_token_accuracy": 0.17832256257534027,
      "num_tokens": 15623761.0,
      "step": 7515
    },
    {
      "entropy": 5.549900197982788,
      "epoch": 0.6821480406386067,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496056719102197,
      "loss": 5.369,
      "mean_token_accuracy": 0.17480377554893495,
      "num_tokens": 15634500.0,
      "step": 7520
    },
    {
      "entropy": 5.474475002288818,
      "epoch": 0.6826015965166908,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960506856426631,
      "loss": 5.2228,
      "mean_token_accuracy": 0.18139137476682662,
      "num_tokens": 15644896.0,
      "step": 7525
    },
    {
      "entropy": 5.5556963920593265,
      "epoch": 0.683055152394775,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960446476117448,
      "loss": 5.3439,
      "mean_token_accuracy": 0.18209579437971116,
      "num_tokens": 15655008.0,
      "step": 7530
    },
    {
      "entropy": 5.526656913757324,
      "epoch": 0.6835087082728593,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960386050095668,
      "loss": 5.265,
      "mean_token_accuracy": 0.17771433889865876,
      "num_tokens": 15665944.0,
      "step": 7535
    },
    {
      "entropy": 5.55998010635376,
      "epoch": 0.6839622641509434,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004960325578362543,
      "loss": 5.3639,
      "mean_token_accuracy": 0.18044693470001222,
      "num_tokens": 15677099.0,
      "step": 7540
    },
    {
      "entropy": 5.593668031692505,
      "epoch": 0.6844158200290276,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496026506091932,
      "loss": 5.3198,
      "mean_token_accuracy": 0.1848937004804611,
      "num_tokens": 15687531.0,
      "step": 7545
    },
    {
      "entropy": 5.586200761795044,
      "epoch": 0.6848693759071117,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960204497767253,
      "loss": 5.3321,
      "mean_token_accuracy": 0.17955249547958374,
      "num_tokens": 15697002.0,
      "step": 7550
    },
    {
      "entropy": 5.43728461265564,
      "epoch": 0.685322931785196,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960143888907594,
      "loss": 5.2166,
      "mean_token_accuracy": 0.18282410353422165,
      "num_tokens": 15707249.0,
      "step": 7555
    },
    {
      "entropy": 5.5209822177886965,
      "epoch": 0.6857764876632801,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960083234341595,
      "loss": 5.2848,
      "mean_token_accuracy": 0.18320000618696214,
      "num_tokens": 15718764.0,
      "step": 7560
    },
    {
      "entropy": 5.546800947189331,
      "epoch": 0.6862300435413643,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960022534070513,
      "loss": 5.2361,
      "mean_token_accuracy": 0.1883462518453598,
      "num_tokens": 15728854.0,
      "step": 7565
    },
    {
      "entropy": 5.43729796409607,
      "epoch": 0.6866835994194485,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049599617880956,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18559020906686782,
      "num_tokens": 15739682.0,
      "step": 7570
    },
    {
      "entropy": 5.473732614517212,
      "epoch": 0.6871371552975326,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959900996418115,
      "loss": 5.2534,
      "mean_token_accuracy": 0.18659544587135315,
      "num_tokens": 15749628.0,
      "step": 7575
    },
    {
      "entropy": 5.554725790023804,
      "epoch": 0.6875907111756169,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959840159039315,
      "loss": 5.3221,
      "mean_token_accuracy": 0.1806378498673439,
      "num_tokens": 15760137.0,
      "step": 7580
    },
    {
      "entropy": 5.579864168167115,
      "epoch": 0.688044267053701,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959779275960456,
      "loss": 5.2711,
      "mean_token_accuracy": 0.18420590311288834,
      "num_tokens": 15769889.0,
      "step": 7585
    },
    {
      "entropy": 5.514402961730957,
      "epoch": 0.6884978229317852,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049597183471828,
      "loss": 5.3859,
      "mean_token_accuracy": 0.18321892023086547,
      "num_tokens": 15780912.0,
      "step": 7590
    },
    {
      "entropy": 5.531727409362793,
      "epoch": 0.6889513788098693,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004959657372707604,
      "loss": 5.2578,
      "mean_token_accuracy": 0.18849852085113525,
      "num_tokens": 15791719.0,
      "step": 7595
    },
    {
      "entropy": 5.572351932525635,
      "epoch": 0.6894049346879536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959596352536132,
      "loss": 5.2987,
      "mean_token_accuracy": 0.18423290997743608,
      "num_tokens": 15802551.0,
      "step": 7600
    },
    {
      "entropy": 5.585467529296875,
      "epoch": 0.6898584905660378,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959535286669643,
      "loss": 5.3539,
      "mean_token_accuracy": 0.1799274355173111,
      "num_tokens": 15813551.0,
      "step": 7605
    },
    {
      "entropy": 5.518627882003784,
      "epoch": 0.6903120464441219,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959474175109402,
      "loss": 5.3222,
      "mean_token_accuracy": 0.1847722724080086,
      "num_tokens": 15824262.0,
      "step": 7610
    },
    {
      "entropy": 5.595053720474243,
      "epoch": 0.6907656023222061,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959413017856672,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18128857016563416,
      "num_tokens": 15835493.0,
      "step": 7615
    },
    {
      "entropy": 5.448233413696289,
      "epoch": 0.6912191582002902,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004959351814912716,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18287449926137925,
      "num_tokens": 15846224.0,
      "step": 7620
    },
    {
      "entropy": 5.5843593120574955,
      "epoch": 0.6916727140783745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959290566278803,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1838727205991745,
      "num_tokens": 15855622.0,
      "step": 7625
    },
    {
      "entropy": 5.5864767074584964,
      "epoch": 0.6921262699564587,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959229271956199,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18438090682029723,
      "num_tokens": 15865835.0,
      "step": 7630
    },
    {
      "entropy": 5.511325550079346,
      "epoch": 0.6925798258345428,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959167931946169,
      "loss": 5.2916,
      "mean_token_accuracy": 0.18278788179159164,
      "num_tokens": 15875915.0,
      "step": 7635
    },
    {
      "entropy": 5.570894479751587,
      "epoch": 0.693033381712627,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959106546249984,
      "loss": 5.301,
      "mean_token_accuracy": 0.18494617342948913,
      "num_tokens": 15887111.0,
      "step": 7640
    },
    {
      "entropy": 5.589294576644898,
      "epoch": 0.6934869375907112,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004959045114868913,
      "loss": 5.4105,
      "mean_token_accuracy": 0.17258675992488862,
      "num_tokens": 15899528.0,
      "step": 7645
    },
    {
      "entropy": 5.547265100479126,
      "epoch": 0.6939404934687954,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958983637804225,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18939780741930007,
      "num_tokens": 15910881.0,
      "step": 7650
    },
    {
      "entropy": 5.6097711563110355,
      "epoch": 0.6943940493468795,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958922115057194,
      "loss": 5.3506,
      "mean_token_accuracy": 0.17834327965974808,
      "num_tokens": 15921189.0,
      "step": 7655
    },
    {
      "entropy": 5.611061525344849,
      "epoch": 0.6948476052249637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958860546629089,
      "loss": 5.3497,
      "mean_token_accuracy": 0.18507591933012008,
      "num_tokens": 15931276.0,
      "step": 7660
    },
    {
      "entropy": 5.58590931892395,
      "epoch": 0.6953011611030478,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004958798932521186,
      "loss": 5.429,
      "mean_token_accuracy": 0.1739620342850685,
      "num_tokens": 15941506.0,
      "step": 7665
    },
    {
      "entropy": 5.493170309066772,
      "epoch": 0.6957547169811321,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958737272734759,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18339498937129975,
      "num_tokens": 15951158.0,
      "step": 7670
    },
    {
      "entropy": 5.588201665878296,
      "epoch": 0.6962082728592163,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004958675567271081,
      "loss": 5.2997,
      "mean_token_accuracy": 0.18696109801530839,
      "num_tokens": 15960852.0,
      "step": 7675
    },
    {
      "entropy": 5.544498586654663,
      "epoch": 0.6966618287373004,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000495861381613143,
      "loss": 5.2936,
      "mean_token_accuracy": 0.17795197516679764,
      "num_tokens": 15971291.0,
      "step": 7680
    },
    {
      "entropy": 5.560691261291504,
      "epoch": 0.6971153846153846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958552019317082,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18023389130830764,
      "num_tokens": 15982180.0,
      "step": 7685
    },
    {
      "entropy": 5.486464738845825,
      "epoch": 0.6975689404934688,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004958490176829316,
      "loss": 5.2506,
      "mean_token_accuracy": 0.18482301533222198,
      "num_tokens": 15992363.0,
      "step": 7690
    },
    {
      "entropy": 5.53675217628479,
      "epoch": 0.698022496371553,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495842828866941,
      "loss": 5.3368,
      "mean_token_accuracy": 0.1778530851006508,
      "num_tokens": 16003477.0,
      "step": 7695
    },
    {
      "entropy": 5.585207319259643,
      "epoch": 0.6984760522496372,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958366354838645,
      "loss": 5.3135,
      "mean_token_accuracy": 0.1820820912718773,
      "num_tokens": 16013747.0,
      "step": 7700
    },
    {
      "entropy": 5.542598628997803,
      "epoch": 0.6989296081277213,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049583043753383,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1875500798225403,
      "num_tokens": 16025662.0,
      "step": 7705
    },
    {
      "entropy": 5.572959136962891,
      "epoch": 0.6993831640058055,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004958242350169658,
      "loss": 5.3992,
      "mean_token_accuracy": 0.1829013928771019,
      "num_tokens": 16037315.0,
      "step": 7710
    },
    {
      "entropy": 5.502584838867188,
      "epoch": 0.6998367198838897,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004958180279334001,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1828617572784424,
      "num_tokens": 16047297.0,
      "step": 7715
    },
    {
      "entropy": 5.482805871963501,
      "epoch": 0.7002902757619739,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958118162832614,
      "loss": 5.2115,
      "mean_token_accuracy": 0.18582056909799577,
      "num_tokens": 16057573.0,
      "step": 7720
    },
    {
      "entropy": 5.51785831451416,
      "epoch": 0.700743831640058,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958056000666781,
      "loss": 5.3708,
      "mean_token_accuracy": 0.17752822488546371,
      "num_tokens": 16067926.0,
      "step": 7725
    },
    {
      "entropy": 5.560018396377563,
      "epoch": 0.7011973875181422,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957993792837786,
      "loss": 5.3089,
      "mean_token_accuracy": 0.18352362066507338,
      "num_tokens": 16077938.0,
      "step": 7730
    },
    {
      "entropy": 5.509559869766235,
      "epoch": 0.7016509433962265,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004957931539346917,
      "loss": 5.1887,
      "mean_token_accuracy": 0.18910111486911774,
      "num_tokens": 16088163.0,
      "step": 7735
    },
    {
      "entropy": 5.6044848442077635,
      "epoch": 0.7021044992743106,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957869240195462,
      "loss": 5.3964,
      "mean_token_accuracy": 0.17759514451026917,
      "num_tokens": 16100640.0,
      "step": 7740
    },
    {
      "entropy": 5.525788402557373,
      "epoch": 0.7025580551523948,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004957806895384707,
      "loss": 5.22,
      "mean_token_accuracy": 0.1873020723462105,
      "num_tokens": 16111018.0,
      "step": 7745
    },
    {
      "entropy": 5.509130525588989,
      "epoch": 0.7030116110304789,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957744504915944,
      "loss": 5.2634,
      "mean_token_accuracy": 0.17722718566656112,
      "num_tokens": 16120873.0,
      "step": 7750
    },
    {
      "entropy": 5.563656759262085,
      "epoch": 0.7034651669085631,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957682068790463,
      "loss": 5.2845,
      "mean_token_accuracy": 0.18420600146055222,
      "num_tokens": 16131178.0,
      "step": 7755
    },
    {
      "entropy": 5.503865575790405,
      "epoch": 0.7039187227866474,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004957619587009553,
      "loss": 5.2961,
      "mean_token_accuracy": 0.18365350663661956,
      "num_tokens": 16141156.0,
      "step": 7760
    },
    {
      "entropy": 5.544031953811645,
      "epoch": 0.7043722786647315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004957557059574507,
      "loss": 5.397,
      "mean_token_accuracy": 0.18116726726293564,
      "num_tokens": 16150077.0,
      "step": 7765
    },
    {
      "entropy": 5.521406173706055,
      "epoch": 0.7048258345428157,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957494486486618,
      "loss": 5.307,
      "mean_token_accuracy": 0.1804828479886055,
      "num_tokens": 16160514.0,
      "step": 7770
    },
    {
      "entropy": 5.567109632492065,
      "epoch": 0.7052793904208998,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957431867747182,
      "loss": 5.3432,
      "mean_token_accuracy": 0.187213496863842,
      "num_tokens": 16170811.0,
      "step": 7775
    },
    {
      "entropy": 5.480603742599487,
      "epoch": 0.7057329462989841,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957369203357492,
      "loss": 5.2402,
      "mean_token_accuracy": 0.18310723453760147,
      "num_tokens": 16181553.0,
      "step": 7780
    },
    {
      "entropy": 5.523513269424439,
      "epoch": 0.7061865021770682,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004957306493318844,
      "loss": 5.2696,
      "mean_token_accuracy": 0.18861202895641327,
      "num_tokens": 16191308.0,
      "step": 7785
    },
    {
      "entropy": 5.526185083389282,
      "epoch": 0.7066400580551524,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957243737632535,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18830116391181945,
      "num_tokens": 16201311.0,
      "step": 7790
    },
    {
      "entropy": 5.584646558761596,
      "epoch": 0.7070936139332366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957180936299863,
      "loss": 5.2822,
      "mean_token_accuracy": 0.19124689400196077,
      "num_tokens": 16211105.0,
      "step": 7795
    },
    {
      "entropy": 5.554888391494751,
      "epoch": 0.7075471698113207,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957118089322127,
      "loss": 5.3012,
      "mean_token_accuracy": 0.1761303499341011,
      "num_tokens": 16220836.0,
      "step": 7800
    },
    {
      "entropy": 5.484191131591797,
      "epoch": 0.708000725689405,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957055196700628,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18652159571647645,
      "num_tokens": 16231630.0,
      "step": 7805
    },
    {
      "entropy": 5.544210433959961,
      "epoch": 0.7084542815674891,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004956992258436663,
      "loss": 5.304,
      "mean_token_accuracy": 0.18482640385627747,
      "num_tokens": 16242272.0,
      "step": 7810
    },
    {
      "entropy": 5.429587078094483,
      "epoch": 0.7089078374455733,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956929274531536,
      "loss": 5.2099,
      "mean_token_accuracy": 0.19073296040296556,
      "num_tokens": 16252555.0,
      "step": 7815
    },
    {
      "entropy": 5.518241500854492,
      "epoch": 0.7093613933236574,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000495686624498655,
      "loss": 5.331,
      "mean_token_accuracy": 0.187027183175087,
      "num_tokens": 16261964.0,
      "step": 7820
    },
    {
      "entropy": 5.617752361297607,
      "epoch": 0.7098149492017417,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004956803169803007,
      "loss": 5.3859,
      "mean_token_accuracy": 0.17545270025730134,
      "num_tokens": 16274543.0,
      "step": 7825
    },
    {
      "entropy": 5.507696151733398,
      "epoch": 0.7102685050798259,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004956740048982212,
      "loss": 5.2941,
      "mean_token_accuracy": 0.1758976697921753,
      "num_tokens": 16285124.0,
      "step": 7830
    },
    {
      "entropy": 5.528030920028686,
      "epoch": 0.71072206095791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956676882525472,
      "loss": 5.3634,
      "mean_token_accuracy": 0.18425541818141938,
      "num_tokens": 16295727.0,
      "step": 7835
    },
    {
      "entropy": 5.628829622268677,
      "epoch": 0.7111756168359942,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495661367043409,
      "loss": 5.2686,
      "mean_token_accuracy": 0.18808830976486207,
      "num_tokens": 16305496.0,
      "step": 7840
    },
    {
      "entropy": 5.535170650482177,
      "epoch": 0.7116291727140783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956550412709376,
      "loss": 5.2455,
      "mean_token_accuracy": 0.19257238060235976,
      "num_tokens": 16315112.0,
      "step": 7845
    },
    {
      "entropy": 5.514770650863648,
      "epoch": 0.7120827285921626,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956487109352637,
      "loss": 5.2948,
      "mean_token_accuracy": 0.18559942841529847,
      "num_tokens": 16325613.0,
      "step": 7850
    },
    {
      "entropy": 5.510515832901001,
      "epoch": 0.7125362844702468,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956423760365181,
      "loss": 5.2236,
      "mean_token_accuracy": 0.18841205388307572,
      "num_tokens": 16335317.0,
      "step": 7855
    },
    {
      "entropy": 5.594292879104614,
      "epoch": 0.7129898403483309,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004956360365748321,
      "loss": 5.3684,
      "mean_token_accuracy": 0.17692358642816544,
      "num_tokens": 16345003.0,
      "step": 7860
    },
    {
      "entropy": 5.5583514213562015,
      "epoch": 0.7134433962264151,
      "grad_norm": 0.875,
      "learning_rate": 0.0004956296925503367,
      "loss": 5.3597,
      "mean_token_accuracy": 0.1829149454832077,
      "num_tokens": 16356108.0,
      "step": 7865
    },
    {
      "entropy": 5.483655261993408,
      "epoch": 0.7138969521044993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004956233439631628,
      "loss": 5.1666,
      "mean_token_accuracy": 0.19129244983196259,
      "num_tokens": 16364733.0,
      "step": 7870
    },
    {
      "entropy": 5.558326196670532,
      "epoch": 0.7143505079825835,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956169908134421,
      "loss": 5.273,
      "mean_token_accuracy": 0.18959595412015914,
      "num_tokens": 16375229.0,
      "step": 7875
    },
    {
      "entropy": 5.519115591049195,
      "epoch": 0.7148040638606676,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004956106331013057,
      "loss": 5.3594,
      "mean_token_accuracy": 0.18587884157896042,
      "num_tokens": 16386366.0,
      "step": 7880
    },
    {
      "entropy": 5.490304279327392,
      "epoch": 0.7152576197387518,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956042708268853,
      "loss": 5.2427,
      "mean_token_accuracy": 0.1905361145734787,
      "num_tokens": 16397067.0,
      "step": 7885
    },
    {
      "entropy": 5.452323913574219,
      "epoch": 0.715711175616836,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955979039903123,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18583076745271682,
      "num_tokens": 16407256.0,
      "step": 7890
    },
    {
      "entropy": 5.501726055145264,
      "epoch": 0.7161647314949202,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004955915325917183,
      "loss": 5.2573,
      "mean_token_accuracy": 0.18402386009693145,
      "num_tokens": 16418701.0,
      "step": 7895
    },
    {
      "entropy": 5.472748422622681,
      "epoch": 0.7166182873730044,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004955851566312354,
      "loss": 5.1796,
      "mean_token_accuracy": 0.19416148364543914,
      "num_tokens": 16428019.0,
      "step": 7900
    },
    {
      "entropy": 5.482213354110717,
      "epoch": 0.7170718432510885,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955787761089952,
      "loss": 5.2658,
      "mean_token_accuracy": 0.179888778924942,
      "num_tokens": 16440122.0,
      "step": 7905
    },
    {
      "entropy": 5.606745100021362,
      "epoch": 0.7175253991291727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004955723910251296,
      "loss": 5.3425,
      "mean_token_accuracy": 0.1781800538301468,
      "num_tokens": 16449801.0,
      "step": 7910
    },
    {
      "entropy": 5.498953914642334,
      "epoch": 0.717978955007257,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955660013797709,
      "loss": 5.2049,
      "mean_token_accuracy": 0.18978928178548812,
      "num_tokens": 16460155.0,
      "step": 7915
    },
    {
      "entropy": 5.5611310482025145,
      "epoch": 0.7184325108853411,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495559607173051,
      "loss": 5.4094,
      "mean_token_accuracy": 0.17007175385951995,
      "num_tokens": 16470656.0,
      "step": 7920
    },
    {
      "entropy": 5.565587759017944,
      "epoch": 0.7188860667634253,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955532084051022,
      "loss": 5.3028,
      "mean_token_accuracy": 0.18778245747089387,
      "num_tokens": 16481449.0,
      "step": 7925
    },
    {
      "entropy": 5.544949960708618,
      "epoch": 0.7193396226415094,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955468050760568,
      "loss": 5.3213,
      "mean_token_accuracy": 0.18307759761810302,
      "num_tokens": 16492032.0,
      "step": 7930
    },
    {
      "entropy": 5.559630060195923,
      "epoch": 0.7197931785195936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955403971860474,
      "loss": 5.2449,
      "mean_token_accuracy": 0.18228569626808167,
      "num_tokens": 16502221.0,
      "step": 7935
    },
    {
      "entropy": 5.516655302047729,
      "epoch": 0.7202467343976778,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955339847352063,
      "loss": 5.2255,
      "mean_token_accuracy": 0.18727567791938782,
      "num_tokens": 16512021.0,
      "step": 7940
    },
    {
      "entropy": 5.526650905609131,
      "epoch": 0.720700290275762,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955275677236662,
      "loss": 5.2781,
      "mean_token_accuracy": 0.18725835978984834,
      "num_tokens": 16522060.0,
      "step": 7945
    },
    {
      "entropy": 5.522590780258179,
      "epoch": 0.7211538461538461,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955211461515599,
      "loss": 5.2851,
      "mean_token_accuracy": 0.18848953992128373,
      "num_tokens": 16532827.0,
      "step": 7950
    },
    {
      "entropy": 5.5282035827636715,
      "epoch": 0.7216074020319303,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049551472001902,
      "loss": 5.3291,
      "mean_token_accuracy": 0.18570098727941514,
      "num_tokens": 16542838.0,
      "step": 7955
    },
    {
      "entropy": 5.49766263961792,
      "epoch": 0.7220609579100146,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955082893261796,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18782561421394348,
      "num_tokens": 16552615.0,
      "step": 7960
    },
    {
      "entropy": 5.5685553550720215,
      "epoch": 0.7225145137880987,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955018540731714,
      "loss": 5.3358,
      "mean_token_accuracy": 0.18197497874498367,
      "num_tokens": 16561777.0,
      "step": 7965
    },
    {
      "entropy": 5.561930131912232,
      "epoch": 0.7229680696661829,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954954142601288,
      "loss": 5.3046,
      "mean_token_accuracy": 0.1859089732170105,
      "num_tokens": 16572111.0,
      "step": 7970
    },
    {
      "entropy": 5.553031778335571,
      "epoch": 0.723421625544267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004954889698871849,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1823531538248062,
      "num_tokens": 16581674.0,
      "step": 7975
    },
    {
      "entropy": 5.5707708358764645,
      "epoch": 0.7238751814223512,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004954825209544728,
      "loss": 5.3265,
      "mean_token_accuracy": 0.18853056877851487,
      "num_tokens": 16591802.0,
      "step": 7980
    },
    {
      "entropy": 5.4827385425567625,
      "epoch": 0.7243287373004355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495476067462126,
      "loss": 5.2194,
      "mean_token_accuracy": 0.1914340078830719,
      "num_tokens": 16602772.0,
      "step": 7985
    },
    {
      "entropy": 5.56507248878479,
      "epoch": 0.7247822931785196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495469609410278,
      "loss": 5.2697,
      "mean_token_accuracy": 0.18687217831611633,
      "num_tokens": 16613526.0,
      "step": 7990
    },
    {
      "entropy": 5.475165891647339,
      "epoch": 0.7252358490566038,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954631467990622,
      "loss": 5.2738,
      "mean_token_accuracy": 0.1871074303984642,
      "num_tokens": 16623583.0,
      "step": 7995
    },
    {
      "entropy": 5.543488788604736,
      "epoch": 0.7256894049346879,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954566796286124,
      "loss": 5.3322,
      "mean_token_accuracy": 0.1807237446308136,
      "num_tokens": 16635068.0,
      "step": 8000
    },
    {
      "entropy": 5.588771057128906,
      "epoch": 0.7261429608127722,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954502078990623,
      "loss": 5.2332,
      "mean_token_accuracy": 0.18706244379281997,
      "num_tokens": 16645021.0,
      "step": 8005
    },
    {
      "entropy": 5.526928997039795,
      "epoch": 0.7265965166908563,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004954437316105457,
      "loss": 5.3204,
      "mean_token_accuracy": 0.17738464921712876,
      "num_tokens": 16656021.0,
      "step": 8010
    },
    {
      "entropy": 5.476295518875122,
      "epoch": 0.7270500725689405,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954372507631965,
      "loss": 5.2173,
      "mean_token_accuracy": 0.19314096868038177,
      "num_tokens": 16665236.0,
      "step": 8015
    },
    {
      "entropy": 5.545669603347778,
      "epoch": 0.7275036284470247,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954307653571489,
      "loss": 5.4112,
      "mean_token_accuracy": 0.17193434983491898,
      "num_tokens": 16676690.0,
      "step": 8020
    },
    {
      "entropy": 5.553662109375,
      "epoch": 0.7279571843251088,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954242753925367,
      "loss": 5.2901,
      "mean_token_accuracy": 0.17670878767967224,
      "num_tokens": 16687288.0,
      "step": 8025
    },
    {
      "entropy": 5.5189173221588135,
      "epoch": 0.7284107402031931,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954177808694946,
      "loss": 5.3052,
      "mean_token_accuracy": 0.1775049388408661,
      "num_tokens": 16697873.0,
      "step": 8030
    },
    {
      "entropy": 5.551232862472534,
      "epoch": 0.7288642960812772,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004954112817881565,
      "loss": 5.3207,
      "mean_token_accuracy": 0.18077222257852554,
      "num_tokens": 16709002.0,
      "step": 8035
    },
    {
      "entropy": 5.493309879302979,
      "epoch": 0.7293178519593614,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954047781486568,
      "loss": 5.278,
      "mean_token_accuracy": 0.18425150364637374,
      "num_tokens": 16718989.0,
      "step": 8040
    },
    {
      "entropy": 5.620672035217285,
      "epoch": 0.7297714078374455,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953982699511303,
      "loss": 5.3692,
      "mean_token_accuracy": 0.18326299786567687,
      "num_tokens": 16730425.0,
      "step": 8045
    },
    {
      "entropy": 5.562078189849854,
      "epoch": 0.7302249637155298,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953917571957114,
      "loss": 5.2638,
      "mean_token_accuracy": 0.18367500156164168,
      "num_tokens": 16740216.0,
      "step": 8050
    },
    {
      "entropy": 5.507948780059815,
      "epoch": 0.730678519593614,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953852398825346,
      "loss": 5.2517,
      "mean_token_accuracy": 0.18657175153493882,
      "num_tokens": 16752436.0,
      "step": 8055
    },
    {
      "entropy": 5.471053600311279,
      "epoch": 0.7311320754716981,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495378718011735,
      "loss": 5.2782,
      "mean_token_accuracy": 0.17842504680156707,
      "num_tokens": 16762631.0,
      "step": 8060
    },
    {
      "entropy": 5.580714368820191,
      "epoch": 0.7315856313497823,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004953721915834473,
      "loss": 5.2878,
      "mean_token_accuracy": 0.18535203933715821,
      "num_tokens": 16772613.0,
      "step": 8065
    },
    {
      "entropy": 5.585743761062622,
      "epoch": 0.7320391872278664,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953656605978067,
      "loss": 5.2995,
      "mean_token_accuracy": 0.17919224500656128,
      "num_tokens": 16782963.0,
      "step": 8070
    },
    {
      "entropy": 5.497876787185669,
      "epoch": 0.7324927431059507,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000495359125054948,
      "loss": 5.2248,
      "mean_token_accuracy": 0.18719250857830047,
      "num_tokens": 16793892.0,
      "step": 8075
    },
    {
      "entropy": 5.433376407623291,
      "epoch": 0.7329462989840348,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004953525849550063,
      "loss": 5.2243,
      "mean_token_accuracy": 0.1793593570590019,
      "num_tokens": 16803388.0,
      "step": 8080
    },
    {
      "entropy": 5.505691719055176,
      "epoch": 0.733399854862119,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953460402981172,
      "loss": 5.1801,
      "mean_token_accuracy": 0.19310829937458038,
      "num_tokens": 16813201.0,
      "step": 8085
    },
    {
      "entropy": 5.602481555938721,
      "epoch": 0.7338534107402032,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953394910844157,
      "loss": 5.4435,
      "mean_token_accuracy": 0.18122223764657974,
      "num_tokens": 16823265.0,
      "step": 8090
    },
    {
      "entropy": 5.53153166770935,
      "epoch": 0.7343069666182874,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953329373140374,
      "loss": 5.2311,
      "mean_token_accuracy": 0.18478136509656906,
      "num_tokens": 16833057.0,
      "step": 8095
    },
    {
      "entropy": 5.486098194122315,
      "epoch": 0.7347605224963716,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004953263789871179,
      "loss": 5.1915,
      "mean_token_accuracy": 0.18385787308216095,
      "num_tokens": 16843488.0,
      "step": 8100
    },
    {
      "entropy": 5.463052749633789,
      "epoch": 0.7352140783744557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953198161037925,
      "loss": 5.2102,
      "mean_token_accuracy": 0.19679825603961945,
      "num_tokens": 16852448.0,
      "step": 8105
    },
    {
      "entropy": 5.431389713287354,
      "epoch": 0.7356676342525399,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004953132486641975,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1843101128935814,
      "num_tokens": 16863944.0,
      "step": 8110
    },
    {
      "entropy": 5.52769889831543,
      "epoch": 0.736121190130624,
      "grad_norm": 1.046875,
      "learning_rate": 0.000495306676668468,
      "loss": 5.1792,
      "mean_token_accuracy": 0.18727706223726273,
      "num_tokens": 16873923.0,
      "step": 8115
    },
    {
      "entropy": 5.547335052490235,
      "epoch": 0.7365747460087083,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953001001167406,
      "loss": 5.2912,
      "mean_token_accuracy": 0.18075185418128967,
      "num_tokens": 16884373.0,
      "step": 8120
    },
    {
      "entropy": 5.570522165298462,
      "epoch": 0.7370283018867925,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004952935190091508,
      "loss": 5.4223,
      "mean_token_accuracy": 0.17607153505086898,
      "num_tokens": 16895869.0,
      "step": 8125
    },
    {
      "entropy": 5.455164957046509,
      "epoch": 0.7374818577648766,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952869333458349,
      "loss": 5.266,
      "mean_token_accuracy": 0.18525791615247728,
      "num_tokens": 16906737.0,
      "step": 8130
    },
    {
      "entropy": 5.48662314414978,
      "epoch": 0.7379354136429608,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952803431269292,
      "loss": 5.1975,
      "mean_token_accuracy": 0.19422420412302016,
      "num_tokens": 16916083.0,
      "step": 8135
    },
    {
      "entropy": 5.597974872589111,
      "epoch": 0.738388969521045,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004952737483525699,
      "loss": 5.3612,
      "mean_token_accuracy": 0.18378081768751145,
      "num_tokens": 16926419.0,
      "step": 8140
    },
    {
      "entropy": 5.5718834400177,
      "epoch": 0.7388425253991292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004952671490228932,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1853596895933151,
      "num_tokens": 16936067.0,
      "step": 8145
    },
    {
      "entropy": 5.585398387908936,
      "epoch": 0.7392960812772134,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004952605451380357,
      "loss": 5.2489,
      "mean_token_accuracy": 0.18981460332870484,
      "num_tokens": 16946571.0,
      "step": 8150
    },
    {
      "entropy": 5.453392887115479,
      "epoch": 0.7397496371552975,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004952539366981342,
      "loss": 5.2966,
      "mean_token_accuracy": 0.18083768635988234,
      "num_tokens": 16956764.0,
      "step": 8155
    },
    {
      "entropy": 5.519579267501831,
      "epoch": 0.7402031930333817,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000495247323703325,
      "loss": 5.2249,
      "mean_token_accuracy": 0.1882965698838234,
      "num_tokens": 16966002.0,
      "step": 8160
    },
    {
      "entropy": 5.572782278060913,
      "epoch": 0.7406567489114659,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495240706153745,
      "loss": 5.3257,
      "mean_token_accuracy": 0.18548445403575897,
      "num_tokens": 16975705.0,
      "step": 8165
    },
    {
      "entropy": 5.522975587844849,
      "epoch": 0.7411103047895501,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952340840495312,
      "loss": 5.2265,
      "mean_token_accuracy": 0.1855018049478531,
      "num_tokens": 16985945.0,
      "step": 8170
    },
    {
      "entropy": 5.4940248966217045,
      "epoch": 0.7415638606676342,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952274573908203,
      "loss": 5.2005,
      "mean_token_accuracy": 0.18602324575185775,
      "num_tokens": 16995475.0,
      "step": 8175
    },
    {
      "entropy": 5.510499334335327,
      "epoch": 0.7420174165457184,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952208261777496,
      "loss": 5.2237,
      "mean_token_accuracy": 0.18946883529424668,
      "num_tokens": 17005028.0,
      "step": 8180
    },
    {
      "entropy": 5.531285905838013,
      "epoch": 0.7424709724238027,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495214190410456,
      "loss": 5.2916,
      "mean_token_accuracy": 0.18801498115062715,
      "num_tokens": 17014786.0,
      "step": 8185
    },
    {
      "entropy": 5.563917732238769,
      "epoch": 0.7429245283018868,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952075500890768,
      "loss": 5.2945,
      "mean_token_accuracy": 0.18754068166017532,
      "num_tokens": 17024431.0,
      "step": 8190
    },
    {
      "entropy": 5.541070652008057,
      "epoch": 0.743378084179971,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952009052137493,
      "loss": 5.2722,
      "mean_token_accuracy": 0.18842438012361526,
      "num_tokens": 17034368.0,
      "step": 8195
    },
    {
      "entropy": 5.53280725479126,
      "epoch": 0.7438316400580551,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951942557846112,
      "loss": 5.3039,
      "mean_token_accuracy": 0.182603819668293,
      "num_tokens": 17043818.0,
      "step": 8200
    },
    {
      "entropy": 5.568868350982666,
      "epoch": 0.7442851959361393,
      "grad_norm": 1.0,
      "learning_rate": 0.0004951876018017997,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18415388315916062,
      "num_tokens": 17054843.0,
      "step": 8205
    },
    {
      "entropy": 5.5368658065795895,
      "epoch": 0.7447387518142236,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004951809432654524,
      "loss": 5.3268,
      "mean_token_accuracy": 0.1834358587861061,
      "num_tokens": 17065539.0,
      "step": 8210
    },
    {
      "entropy": 5.491271352767944,
      "epoch": 0.7451923076923077,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495174280175707,
      "loss": 5.2211,
      "mean_token_accuracy": 0.18677894920110702,
      "num_tokens": 17075443.0,
      "step": 8215
    },
    {
      "entropy": 5.592065238952637,
      "epoch": 0.7456458635703919,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004951676125327015,
      "loss": 5.3549,
      "mean_token_accuracy": 0.18309056311845778,
      "num_tokens": 17085490.0,
      "step": 8220
    },
    {
      "entropy": 5.606427907943726,
      "epoch": 0.746099419448476,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951609403365737,
      "loss": 5.3441,
      "mean_token_accuracy": 0.1845713585615158,
      "num_tokens": 17094803.0,
      "step": 8225
    },
    {
      "entropy": 5.5416947364807125,
      "epoch": 0.7465529753265603,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004951542635874615,
      "loss": 5.2032,
      "mean_token_accuracy": 0.1835956171154976,
      "num_tokens": 17104373.0,
      "step": 8230
    },
    {
      "entropy": 5.5435961246490475,
      "epoch": 0.7470065312046444,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951475822855029,
      "loss": 5.2741,
      "mean_token_accuracy": 0.18503042459487914,
      "num_tokens": 17115469.0,
      "step": 8235
    },
    {
      "entropy": 5.504471731185913,
      "epoch": 0.7474600870827286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004951408964308364,
      "loss": 5.2491,
      "mean_token_accuracy": 0.1830831825733185,
      "num_tokens": 17125443.0,
      "step": 8240
    },
    {
      "entropy": 5.5065467834472654,
      "epoch": 0.7479136429608128,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951342060236,
      "loss": 5.2229,
      "mean_token_accuracy": 0.19070644974708556,
      "num_tokens": 17136253.0,
      "step": 8245
    },
    {
      "entropy": 5.5766904830932615,
      "epoch": 0.7483671988388969,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951275110639321,
      "loss": 5.2796,
      "mean_token_accuracy": 0.18655992895364762,
      "num_tokens": 17145544.0,
      "step": 8250
    },
    {
      "entropy": 5.52774977684021,
      "epoch": 0.7488207547169812,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951208115519712,
      "loss": 5.2905,
      "mean_token_accuracy": 0.18376002162694932,
      "num_tokens": 17155043.0,
      "step": 8255
    },
    {
      "entropy": 5.510969734191894,
      "epoch": 0.7492743105950653,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951141074878558,
      "loss": 5.3221,
      "mean_token_accuracy": 0.183181294798851,
      "num_tokens": 17164885.0,
      "step": 8260
    },
    {
      "entropy": 5.609808874130249,
      "epoch": 0.7497278664731495,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951073988717246,
      "loss": 5.3253,
      "mean_token_accuracy": 0.18268342316150665,
      "num_tokens": 17175681.0,
      "step": 8265
    },
    {
      "entropy": 5.468179035186767,
      "epoch": 0.7501814223512336,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951006857037163,
      "loss": 5.2009,
      "mean_token_accuracy": 0.18825145214796066,
      "num_tokens": 17186002.0,
      "step": 8270
    },
    {
      "entropy": 5.501715898513794,
      "epoch": 0.7506349782293179,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004950939679839696,
      "loss": 5.2434,
      "mean_token_accuracy": 0.18338658064603805,
      "num_tokens": 17195499.0,
      "step": 8275
    },
    {
      "entropy": 5.538295269012451,
      "epoch": 0.7510885341074021,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004950872457126237,
      "loss": 5.2498,
      "mean_token_accuracy": 0.18770517855882646,
      "num_tokens": 17204684.0,
      "step": 8280
    },
    {
      "entropy": 5.519716119766235,
      "epoch": 0.7515420899854862,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950805188898175,
      "loss": 5.3005,
      "mean_token_accuracy": 0.18001003712415695,
      "num_tokens": 17213987.0,
      "step": 8285
    },
    {
      "entropy": 5.542646598815918,
      "epoch": 0.7519956458635704,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049507378751569,
      "loss": 5.3083,
      "mean_token_accuracy": 0.18153341710567475,
      "num_tokens": 17224512.0,
      "step": 8290
    },
    {
      "entropy": 5.592179584503174,
      "epoch": 0.7524492017416545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004950670515903805,
      "loss": 5.3213,
      "mean_token_accuracy": 0.18630932569503783,
      "num_tokens": 17234362.0,
      "step": 8295
    },
    {
      "entropy": 5.5631856441497805,
      "epoch": 0.7529027576197388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950603111140283,
      "loss": 5.3047,
      "mean_token_accuracy": 0.18705005645751954,
      "num_tokens": 17244241.0,
      "step": 8300
    },
    {
      "entropy": 5.497237014770508,
      "epoch": 0.753356313497823,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950535660867728,
      "loss": 5.2429,
      "mean_token_accuracy": 0.1937210038304329,
      "num_tokens": 17254689.0,
      "step": 8305
    },
    {
      "entropy": 5.5569257736206055,
      "epoch": 0.7538098693759071,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950468165087535,
      "loss": 5.3237,
      "mean_token_accuracy": 0.1863400772213936,
      "num_tokens": 17265506.0,
      "step": 8310
    },
    {
      "entropy": 5.533684968948364,
      "epoch": 0.7542634252539913,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004950400623801098,
      "loss": 5.3973,
      "mean_token_accuracy": 0.18226426541805268,
      "num_tokens": 17276376.0,
      "step": 8315
    },
    {
      "entropy": 5.59690899848938,
      "epoch": 0.7547169811320755,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950333037009818,
      "loss": 5.2811,
      "mean_token_accuracy": 0.17492688000202178,
      "num_tokens": 17287005.0,
      "step": 8320
    },
    {
      "entropy": 5.592009592056274,
      "epoch": 0.7551705370101597,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950265404715089,
      "loss": 5.2924,
      "mean_token_accuracy": 0.1817374274134636,
      "num_tokens": 17297595.0,
      "step": 8325
    },
    {
      "entropy": 5.455151844024658,
      "epoch": 0.7556240928882438,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495019772691831,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1883062943816185,
      "num_tokens": 17307319.0,
      "step": 8330
    },
    {
      "entropy": 5.569844818115234,
      "epoch": 0.756077648766328,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950130003620882,
      "loss": 5.3959,
      "mean_token_accuracy": 0.18379779458045958,
      "num_tokens": 17317978.0,
      "step": 8335
    },
    {
      "entropy": 5.532192945480347,
      "epoch": 0.7565312046444121,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950062234824204,
      "loss": 5.2826,
      "mean_token_accuracy": 0.1818419873714447,
      "num_tokens": 17328339.0,
      "step": 8340
    },
    {
      "entropy": 5.53981204032898,
      "epoch": 0.7569847605224964,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949994420529679,
      "loss": 5.3248,
      "mean_token_accuracy": 0.18118658065795898,
      "num_tokens": 17338668.0,
      "step": 8345
    },
    {
      "entropy": 5.4830098152160645,
      "epoch": 0.7574383164005806,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004949926560738709,
      "loss": 5.2692,
      "mean_token_accuracy": 0.17884783148765565,
      "num_tokens": 17348609.0,
      "step": 8350
    },
    {
      "entropy": 5.543556213378906,
      "epoch": 0.7578918722786647,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004949858655452697,
      "loss": 5.2878,
      "mean_token_accuracy": 0.18408870846033096,
      "num_tokens": 17358014.0,
      "step": 8355
    },
    {
      "entropy": 5.600489568710327,
      "epoch": 0.7583454281567489,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004949790704673048,
      "loss": 5.2905,
      "mean_token_accuracy": 0.18661998957395554,
      "num_tokens": 17367975.0,
      "step": 8360
    },
    {
      "entropy": 5.554596185684204,
      "epoch": 0.7587989840348331,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004949722708401166,
      "loss": 5.2372,
      "mean_token_accuracy": 0.18600034564733506,
      "num_tokens": 17377960.0,
      "step": 8365
    },
    {
      "entropy": 5.555735635757446,
      "epoch": 0.7592525399129173,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949654666638457,
      "loss": 5.3269,
      "mean_token_accuracy": 0.18860286325216294,
      "num_tokens": 17387334.0,
      "step": 8370
    },
    {
      "entropy": 5.5340049266815186,
      "epoch": 0.7597060957910015,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494958657938633,
      "loss": 5.2708,
      "mean_token_accuracy": 0.18800106942653655,
      "num_tokens": 17396521.0,
      "step": 8375
    },
    {
      "entropy": 5.478679418563843,
      "epoch": 0.7601596516690856,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949518446646192,
      "loss": 5.2504,
      "mean_token_accuracy": 0.18262986093759537,
      "num_tokens": 17407217.0,
      "step": 8380
    },
    {
      "entropy": 5.634729862213135,
      "epoch": 0.7606132075471698,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949450268419451,
      "loss": 5.3708,
      "mean_token_accuracy": 0.17887072116136551,
      "num_tokens": 17416982.0,
      "step": 8385
    },
    {
      "entropy": 5.479625225067139,
      "epoch": 0.761066763425254,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494938204470752,
      "loss": 5.2013,
      "mean_token_accuracy": 0.18697880059480668,
      "num_tokens": 17427636.0,
      "step": 8390
    },
    {
      "entropy": 5.509683084487915,
      "epoch": 0.7615203193033382,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004949313775511806,
      "loss": 5.3342,
      "mean_token_accuracy": 0.179373337328434,
      "num_tokens": 17437488.0,
      "step": 8395
    },
    {
      "entropy": 5.50303807258606,
      "epoch": 0.7619738751814223,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004949245460833723,
      "loss": 5.1156,
      "mean_token_accuracy": 0.1915082797408104,
      "num_tokens": 17447838.0,
      "step": 8400
    },
    {
      "entropy": 5.572708368301392,
      "epoch": 0.7624274310595065,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949177100674684,
      "loss": 5.34,
      "mean_token_accuracy": 0.18229371309280396,
      "num_tokens": 17457257.0,
      "step": 8405
    },
    {
      "entropy": 5.531822872161865,
      "epoch": 0.7628809869375908,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949108695036102,
      "loss": 5.2602,
      "mean_token_accuracy": 0.18731117099523545,
      "num_tokens": 17467729.0,
      "step": 8410
    },
    {
      "entropy": 5.531580114364624,
      "epoch": 0.7633345428156749,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949040243919392,
      "loss": 5.2793,
      "mean_token_accuracy": 0.18056346476078033,
      "num_tokens": 17477692.0,
      "step": 8415
    },
    {
      "entropy": 5.491760206222534,
      "epoch": 0.7637880986937591,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948971747325968,
      "loss": 5.2807,
      "mean_token_accuracy": 0.18631262183189393,
      "num_tokens": 17488730.0,
      "step": 8420
    },
    {
      "entropy": 5.567406415939331,
      "epoch": 0.7642416545718432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000494890320525725,
      "loss": 5.3518,
      "mean_token_accuracy": 0.18223876208066941,
      "num_tokens": 17498374.0,
      "step": 8425
    },
    {
      "entropy": 5.554366874694824,
      "epoch": 0.7646952104499274,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948834617714652,
      "loss": 5.2025,
      "mean_token_accuracy": 0.1882804214954376,
      "num_tokens": 17508601.0,
      "step": 8430
    },
    {
      "entropy": 5.512381458282471,
      "epoch": 0.7651487663280117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948765984699595,
      "loss": 5.2487,
      "mean_token_accuracy": 0.18550052344799042,
      "num_tokens": 17518164.0,
      "step": 8435
    },
    {
      "entropy": 5.567288446426391,
      "epoch": 0.7656023222060958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004948697306213494,
      "loss": 5.2225,
      "mean_token_accuracy": 0.19012609124183655,
      "num_tokens": 17527244.0,
      "step": 8440
    },
    {
      "entropy": 5.549527502059936,
      "epoch": 0.76605587808418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004948628582257775,
      "loss": 5.2897,
      "mean_token_accuracy": 0.1844577297568321,
      "num_tokens": 17538045.0,
      "step": 8445
    },
    {
      "entropy": 5.501225328445434,
      "epoch": 0.7665094339622641,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948559812833857,
      "loss": 5.2492,
      "mean_token_accuracy": 0.18562017679214476,
      "num_tokens": 17547584.0,
      "step": 8450
    },
    {
      "entropy": 5.550030612945557,
      "epoch": 0.7669629898403484,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004948490997943162,
      "loss": 5.3011,
      "mean_token_accuracy": 0.18173202872276306,
      "num_tokens": 17558925.0,
      "step": 8455
    },
    {
      "entropy": 5.558056306838989,
      "epoch": 0.7674165457184325,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004948422137587111,
      "loss": 5.3712,
      "mean_token_accuracy": 0.17508765757083894,
      "num_tokens": 17569355.0,
      "step": 8460
    },
    {
      "entropy": 5.5319421768188475,
      "epoch": 0.7678701015965167,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004948353231767131,
      "loss": 5.2348,
      "mean_token_accuracy": 0.18913586884737016,
      "num_tokens": 17579782.0,
      "step": 8465
    },
    {
      "entropy": 5.539266347885132,
      "epoch": 0.7683236574746009,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004948284280484645,
      "loss": 5.2996,
      "mean_token_accuracy": 0.19015204161405563,
      "num_tokens": 17589772.0,
      "step": 8470
    },
    {
      "entropy": 5.481799364089966,
      "epoch": 0.768777213352685,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004948215283741081,
      "loss": 5.2051,
      "mean_token_accuracy": 0.18939271271228791,
      "num_tokens": 17599613.0,
      "step": 8475
    },
    {
      "entropy": 5.554580450057983,
      "epoch": 0.7692307692307693,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948146241537866,
      "loss": 5.2984,
      "mean_token_accuracy": 0.180326946079731,
      "num_tokens": 17609056.0,
      "step": 8480
    },
    {
      "entropy": 5.499348592758179,
      "epoch": 0.7696843251088534,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004948077153876425,
      "loss": 5.2124,
      "mean_token_accuracy": 0.1857810825109482,
      "num_tokens": 17618620.0,
      "step": 8485
    },
    {
      "entropy": 5.500620794296265,
      "epoch": 0.7701378809869376,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948008020758188,
      "loss": 5.1845,
      "mean_token_accuracy": 0.19046938568353652,
      "num_tokens": 17627630.0,
      "step": 8490
    },
    {
      "entropy": 5.409573554992676,
      "epoch": 0.7705914368650217,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004947938842184585,
      "loss": 5.1788,
      "mean_token_accuracy": 0.1935472920536995,
      "num_tokens": 17638903.0,
      "step": 8495
    },
    {
      "entropy": 5.469343042373657,
      "epoch": 0.771044992743106,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947869618157048,
      "loss": 5.2378,
      "mean_token_accuracy": 0.18261487334966658,
      "num_tokens": 17649257.0,
      "step": 8500
    },
    {
      "entropy": 5.481932640075684,
      "epoch": 0.7714985486211902,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947800348677006,
      "loss": 5.2129,
      "mean_token_accuracy": 0.19850903004407883,
      "num_tokens": 17659495.0,
      "step": 8505
    },
    {
      "entropy": 5.464404773712158,
      "epoch": 0.7719521044992743,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947731033745893,
      "loss": 5.1901,
      "mean_token_accuracy": 0.18522430658340455,
      "num_tokens": 17669671.0,
      "step": 8510
    },
    {
      "entropy": 5.5648661136627195,
      "epoch": 0.7724056603773585,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947661673365142,
      "loss": 5.3616,
      "mean_token_accuracy": 0.18332356959581375,
      "num_tokens": 17680927.0,
      "step": 8515
    },
    {
      "entropy": 5.506349372863769,
      "epoch": 0.7728592162554426,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004947592267536188,
      "loss": 5.2245,
      "mean_token_accuracy": 0.19083321392536162,
      "num_tokens": 17691378.0,
      "step": 8520
    },
    {
      "entropy": 5.522361421585083,
      "epoch": 0.7733127721335269,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004947522816260466,
      "loss": 5.3236,
      "mean_token_accuracy": 0.18290676027536393,
      "num_tokens": 17700942.0,
      "step": 8525
    },
    {
      "entropy": 5.499392032623291,
      "epoch": 0.773766328011611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947453319539411,
      "loss": 5.2144,
      "mean_token_accuracy": 0.18726398050785065,
      "num_tokens": 17710076.0,
      "step": 8530
    },
    {
      "entropy": 5.495902681350708,
      "epoch": 0.7742198838896952,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947383777374462,
      "loss": 5.2539,
      "mean_token_accuracy": 0.1863068625330925,
      "num_tokens": 17721009.0,
      "step": 8535
    },
    {
      "entropy": 5.479369497299194,
      "epoch": 0.7746734397677794,
      "grad_norm": 0.875,
      "learning_rate": 0.0004947314189767058,
      "loss": 5.2782,
      "mean_token_accuracy": 0.18749626576900483,
      "num_tokens": 17731821.0,
      "step": 8540
    },
    {
      "entropy": 5.466885328292847,
      "epoch": 0.7751269956458636,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947244556718635,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19538922607898712,
      "num_tokens": 17741425.0,
      "step": 8545
    },
    {
      "entropy": 5.487680578231812,
      "epoch": 0.7755805515239478,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947174878230635,
      "loss": 5.2167,
      "mean_token_accuracy": 0.19239004105329513,
      "num_tokens": 17752381.0,
      "step": 8550
    },
    {
      "entropy": 5.448737573623657,
      "epoch": 0.7760341074020319,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947105154304498,
      "loss": 5.2126,
      "mean_token_accuracy": 0.18385005593299866,
      "num_tokens": 17762768.0,
      "step": 8555
    },
    {
      "entropy": 5.496819543838501,
      "epoch": 0.7764876632801161,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004947035384941667,
      "loss": 5.2306,
      "mean_token_accuracy": 0.1835770294070244,
      "num_tokens": 17774032.0,
      "step": 8560
    },
    {
      "entropy": 5.567213773727417,
      "epoch": 0.7769412191582002,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004946965570143584,
      "loss": 5.3774,
      "mean_token_accuracy": 0.17739971727132797,
      "num_tokens": 17785098.0,
      "step": 8565
    },
    {
      "entropy": 5.56190505027771,
      "epoch": 0.7773947750362845,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946895709911695,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1863827884197235,
      "num_tokens": 17795487.0,
      "step": 8570
    },
    {
      "entropy": 5.525628662109375,
      "epoch": 0.7778483309143687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946825804247442,
      "loss": 5.3538,
      "mean_token_accuracy": 0.17812175750732423,
      "num_tokens": 17806045.0,
      "step": 8575
    },
    {
      "entropy": 5.567047452926635,
      "epoch": 0.7783018867924528,
      "grad_norm": 0.890625,
      "learning_rate": 0.000494675585315227,
      "loss": 5.314,
      "mean_token_accuracy": 0.18278959393501282,
      "num_tokens": 17816948.0,
      "step": 8580
    },
    {
      "entropy": 5.596051406860352,
      "epoch": 0.778755442670537,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004946685856627628,
      "loss": 5.2593,
      "mean_token_accuracy": 0.18712812513113022,
      "num_tokens": 17826929.0,
      "step": 8585
    },
    {
      "entropy": 5.555044555664063,
      "epoch": 0.7792089985486212,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946615814674963,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1852679133415222,
      "num_tokens": 17837780.0,
      "step": 8590
    },
    {
      "entropy": 5.394699335098267,
      "epoch": 0.7796625544267054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004946545727295724,
      "loss": 5.1144,
      "mean_token_accuracy": 0.19915736168622972,
      "num_tokens": 17847835.0,
      "step": 8595
    },
    {
      "entropy": 5.485675191879272,
      "epoch": 0.7801161103047896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946475594491359,
      "loss": 5.2043,
      "mean_token_accuracy": 0.19405315816402435,
      "num_tokens": 17858324.0,
      "step": 8600
    },
    {
      "entropy": 5.496988296508789,
      "epoch": 0.7805696661828737,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946405416263319,
      "loss": 5.2818,
      "mean_token_accuracy": 0.17678985744714737,
      "num_tokens": 17868455.0,
      "step": 8605
    },
    {
      "entropy": 5.545461845397949,
      "epoch": 0.7810232220609579,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946335192613054,
      "loss": 5.2465,
      "mean_token_accuracy": 0.18767861723899842,
      "num_tokens": 17879038.0,
      "step": 8610
    },
    {
      "entropy": 5.460289812088012,
      "epoch": 0.7814767779390421,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946264923542019,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19217866361141206,
      "num_tokens": 17888153.0,
      "step": 8615
    },
    {
      "entropy": 5.4844536781311035,
      "epoch": 0.7819303338171263,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004946194609051664,
      "loss": 5.2519,
      "mean_token_accuracy": 0.19185152649879456,
      "num_tokens": 17898289.0,
      "step": 8620
    },
    {
      "entropy": 5.501482677459717,
      "epoch": 0.7823838896952104,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004946124249143447,
      "loss": 5.2547,
      "mean_token_accuracy": 0.18857731074094772,
      "num_tokens": 17909192.0,
      "step": 8625
    },
    {
      "entropy": 5.4674238681793215,
      "epoch": 0.7828374455732946,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494605384381882,
      "loss": 5.2012,
      "mean_token_accuracy": 0.18121700286865233,
      "num_tokens": 17918937.0,
      "step": 8630
    },
    {
      "entropy": 5.441970586776733,
      "epoch": 0.7832910014513788,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945983393079239,
      "loss": 5.2272,
      "mean_token_accuracy": 0.1953953266143799,
      "num_tokens": 17929346.0,
      "step": 8635
    },
    {
      "entropy": 5.503742122650147,
      "epoch": 0.783744557329463,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004945912896926163,
      "loss": 5.2471,
      "mean_token_accuracy": 0.1864580824971199,
      "num_tokens": 17939479.0,
      "step": 8640
    },
    {
      "entropy": 5.592639398574829,
      "epoch": 0.7841981132075472,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945842355361049,
      "loss": 5.3252,
      "mean_token_accuracy": 0.18374345749616622,
      "num_tokens": 17949841.0,
      "step": 8645
    },
    {
      "entropy": 5.471678686141968,
      "epoch": 0.7846516690856313,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945771768385355,
      "loss": 5.2526,
      "mean_token_accuracy": 0.19229422211647035,
      "num_tokens": 17960334.0,
      "step": 8650
    },
    {
      "entropy": 5.478883457183838,
      "epoch": 0.7851052249637155,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494570113600054,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19275954067707063,
      "num_tokens": 17970595.0,
      "step": 8655
    },
    {
      "entropy": 5.50038251876831,
      "epoch": 0.7855587808417998,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004945630458208068,
      "loss": 5.3421,
      "mean_token_accuracy": 0.17421732395887374,
      "num_tokens": 17981301.0,
      "step": 8660
    },
    {
      "entropy": 5.457895803451538,
      "epoch": 0.7860123367198839,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945559735009397,
      "loss": 5.1658,
      "mean_token_accuracy": 0.191603122651577,
      "num_tokens": 17991262.0,
      "step": 8665
    },
    {
      "entropy": 5.568162202835083,
      "epoch": 0.7864658925979681,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945488966405993,
      "loss": 5.2024,
      "mean_token_accuracy": 0.18861326873302459,
      "num_tokens": 18001212.0,
      "step": 8670
    },
    {
      "entropy": 5.501408624649048,
      "epoch": 0.7869194484760522,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004945418152399317,
      "loss": 5.2229,
      "mean_token_accuracy": 0.19007271379232407,
      "num_tokens": 18010432.0,
      "step": 8675
    },
    {
      "entropy": 5.5106878757476805,
      "epoch": 0.7873730043541364,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945347292990833,
      "loss": 5.2655,
      "mean_token_accuracy": 0.18418308347463608,
      "num_tokens": 18020902.0,
      "step": 8680
    },
    {
      "entropy": 5.512370157241821,
      "epoch": 0.7878265602322206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004945276388182008,
      "loss": 5.132,
      "mean_token_accuracy": 0.1858739361166954,
      "num_tokens": 18031133.0,
      "step": 8685
    },
    {
      "entropy": 5.567989873886108,
      "epoch": 0.7882801161103048,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945205437974309,
      "loss": 5.2472,
      "mean_token_accuracy": 0.18936766386032106,
      "num_tokens": 18042050.0,
      "step": 8690
    },
    {
      "entropy": 5.428337478637696,
      "epoch": 0.788733671988389,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004945134442369201,
      "loss": 5.214,
      "mean_token_accuracy": 0.19088352620601653,
      "num_tokens": 18051642.0,
      "step": 8695
    },
    {
      "entropy": 5.552801275253296,
      "epoch": 0.7891872278664731,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945063401368154,
      "loss": 5.2664,
      "mean_token_accuracy": 0.1811126798391342,
      "num_tokens": 18062180.0,
      "step": 8700
    },
    {
      "entropy": 5.50035719871521,
      "epoch": 0.7896407837445574,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944992314972637,
      "loss": 5.2154,
      "mean_token_accuracy": 0.18759729713201523,
      "num_tokens": 18072302.0,
      "step": 8705
    },
    {
      "entropy": 5.486184597015381,
      "epoch": 0.7900943396226415,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004944921183184118,
      "loss": 5.2179,
      "mean_token_accuracy": 0.18937012553215027,
      "num_tokens": 18081824.0,
      "step": 8710
    },
    {
      "entropy": 5.53223443031311,
      "epoch": 0.7905478955007257,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494485000600407,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1890762209892273,
      "num_tokens": 18092696.0,
      "step": 8715
    },
    {
      "entropy": 5.482760906219482,
      "epoch": 0.7910014513788098,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944778783433965,
      "loss": 5.1892,
      "mean_token_accuracy": 0.1936694249510765,
      "num_tokens": 18101790.0,
      "step": 8720
    },
    {
      "entropy": 5.550623178482056,
      "epoch": 0.791455007256894,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944707515475276,
      "loss": 5.3132,
      "mean_token_accuracy": 0.1843470126390457,
      "num_tokens": 18111572.0,
      "step": 8725
    },
    {
      "entropy": 5.5540313720703125,
      "epoch": 0.7919085631349783,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944636202129475,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1926032617688179,
      "num_tokens": 18120650.0,
      "step": 8730
    },
    {
      "entropy": 5.47811222076416,
      "epoch": 0.7923621190130624,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004944564843398039,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1985896944999695,
      "num_tokens": 18131055.0,
      "step": 8735
    },
    {
      "entropy": 5.573776483535767,
      "epoch": 0.7928156748911466,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004944493439282443,
      "loss": 5.2739,
      "mean_token_accuracy": 0.18772944211959838,
      "num_tokens": 18141423.0,
      "step": 8740
    },
    {
      "entropy": 5.4522857666015625,
      "epoch": 0.7932692307692307,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004944421989784162,
      "loss": 5.1932,
      "mean_token_accuracy": 0.18533381223678588,
      "num_tokens": 18152015.0,
      "step": 8745
    },
    {
      "entropy": 5.539948177337647,
      "epoch": 0.793722786647315,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004944350494904676,
      "loss": 5.2892,
      "mean_token_accuracy": 0.18949988037347792,
      "num_tokens": 18161934.0,
      "step": 8750
    },
    {
      "entropy": 5.496873092651367,
      "epoch": 0.7941763425253991,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944278954645463,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18725937306880952,
      "num_tokens": 18172287.0,
      "step": 8755
    },
    {
      "entropy": 5.482796335220337,
      "epoch": 0.7946298984034833,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004944207369008002,
      "loss": 5.2505,
      "mean_token_accuracy": 0.19790160655975342,
      "num_tokens": 18181015.0,
      "step": 8760
    },
    {
      "entropy": 5.577034568786621,
      "epoch": 0.7950834542815675,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004944135737993773,
      "loss": 5.3461,
      "mean_token_accuracy": 0.18178513646125793,
      "num_tokens": 18192382.0,
      "step": 8765
    },
    {
      "entropy": 5.488105583190918,
      "epoch": 0.7955370101596516,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004944064061604258,
      "loss": 5.2109,
      "mean_token_accuracy": 0.19367620944976807,
      "num_tokens": 18202597.0,
      "step": 8770
    },
    {
      "entropy": 5.523237991333008,
      "epoch": 0.7959905660377359,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943992339840939,
      "loss": 5.2774,
      "mean_token_accuracy": 0.18589625358581544,
      "num_tokens": 18213756.0,
      "step": 8775
    },
    {
      "entropy": 5.434082651138306,
      "epoch": 0.79644412191582,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049439205727053,
      "loss": 5.1884,
      "mean_token_accuracy": 0.18429725021123886,
      "num_tokens": 18223669.0,
      "step": 8780
    },
    {
      "entropy": 5.497088479995727,
      "epoch": 0.7968976777939042,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004943848760198823,
      "loss": 5.2104,
      "mean_token_accuracy": 0.19213344007730485,
      "num_tokens": 18234884.0,
      "step": 8785
    },
    {
      "entropy": 5.466268301010132,
      "epoch": 0.7973512336719883,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943776902322995,
      "loss": 5.1506,
      "mean_token_accuracy": 0.1955432653427124,
      "num_tokens": 18245155.0,
      "step": 8790
    },
    {
      "entropy": 5.37814736366272,
      "epoch": 0.7978047895500726,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943704999079302,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19864733517169952,
      "num_tokens": 18254686.0,
      "step": 8795
    },
    {
      "entropy": 5.471942377090454,
      "epoch": 0.7982583454281568,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943633050469231,
      "loss": 5.121,
      "mean_token_accuracy": 0.19447069764137268,
      "num_tokens": 18265473.0,
      "step": 8800
    },
    {
      "entropy": 5.53497724533081,
      "epoch": 0.7987119013062409,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943561056494269,
      "loss": 5.3336,
      "mean_token_accuracy": 0.18408775627613067,
      "num_tokens": 18276024.0,
      "step": 8805
    },
    {
      "entropy": 5.5106282234191895,
      "epoch": 0.7991654571843251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004943489017155906,
      "loss": 5.2888,
      "mean_token_accuracy": 0.1785351738333702,
      "num_tokens": 18287261.0,
      "step": 8810
    },
    {
      "entropy": 5.560506153106689,
      "epoch": 0.7996190130624092,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004943416932455631,
      "loss": 5.305,
      "mean_token_accuracy": 0.18382198810577394,
      "num_tokens": 18299374.0,
      "step": 8815
    },
    {
      "entropy": 5.472782135009766,
      "epoch": 0.8000725689404935,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004943344802394935,
      "loss": 5.2679,
      "mean_token_accuracy": 0.1808692380785942,
      "num_tokens": 18309344.0,
      "step": 8820
    },
    {
      "entropy": 5.548999357223511,
      "epoch": 0.8005261248185777,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943272626975309,
      "loss": 5.2525,
      "mean_token_accuracy": 0.18682713359594344,
      "num_tokens": 18319778.0,
      "step": 8825
    },
    {
      "entropy": 5.542529487609864,
      "epoch": 0.8009796806966618,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004943200406198246,
      "loss": 5.2544,
      "mean_token_accuracy": 0.18820199817419053,
      "num_tokens": 18329787.0,
      "step": 8830
    },
    {
      "entropy": 5.463485479354858,
      "epoch": 0.801433236574746,
      "grad_norm": 0.921875,
      "learning_rate": 0.000494312814006524,
      "loss": 5.2068,
      "mean_token_accuracy": 0.19309862405061723,
      "num_tokens": 18340670.0,
      "step": 8835
    },
    {
      "entropy": 5.48480076789856,
      "epoch": 0.8018867924528302,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004943055828577784,
      "loss": 5.2014,
      "mean_token_accuracy": 0.18602313101291656,
      "num_tokens": 18352484.0,
      "step": 8840
    },
    {
      "entropy": 5.469893169403076,
      "epoch": 0.8023403483309144,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004942983471737376,
      "loss": 5.1777,
      "mean_token_accuracy": 0.19207196831703185,
      "num_tokens": 18363103.0,
      "step": 8845
    },
    {
      "entropy": 5.4201044082641605,
      "epoch": 0.8027939042089985,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004942911069545511,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19178254008293152,
      "num_tokens": 18373925.0,
      "step": 8850
    },
    {
      "entropy": 5.5190746784210205,
      "epoch": 0.8032474600870827,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942838622003686,
      "loss": 5.288,
      "mean_token_accuracy": 0.18306922763586045,
      "num_tokens": 18384081.0,
      "step": 8855
    },
    {
      "entropy": 5.582727241516113,
      "epoch": 0.8037010159651669,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942766129113399,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19331018328666688,
      "num_tokens": 18392510.0,
      "step": 8860
    },
    {
      "entropy": 5.567781639099121,
      "epoch": 0.8041545718432511,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942693590876151,
      "loss": 5.3186,
      "mean_token_accuracy": 0.182000532746315,
      "num_tokens": 18402809.0,
      "step": 8865
    },
    {
      "entropy": 5.5828429698944095,
      "epoch": 0.8046081277213353,
      "grad_norm": 1.078125,
      "learning_rate": 0.000494262100729344,
      "loss": 5.3322,
      "mean_token_accuracy": 0.18444409668445588,
      "num_tokens": 18413342.0,
      "step": 8870
    },
    {
      "entropy": 5.521259212493897,
      "epoch": 0.8050616835994194,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004942548378366768,
      "loss": 5.3185,
      "mean_token_accuracy": 0.1856078937649727,
      "num_tokens": 18424735.0,
      "step": 8875
    },
    {
      "entropy": 5.5181769847869875,
      "epoch": 0.8055152394775036,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942475704097635,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18817635774612426,
      "num_tokens": 18434797.0,
      "step": 8880
    },
    {
      "entropy": 5.514129638671875,
      "epoch": 0.8059687953555879,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942402984487548,
      "loss": 5.2634,
      "mean_token_accuracy": 0.18100712597370147,
      "num_tokens": 18445456.0,
      "step": 8885
    },
    {
      "entropy": 5.485908126831054,
      "epoch": 0.806422351233672,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004942330219538008,
      "loss": 5.2398,
      "mean_token_accuracy": 0.18712347149848937,
      "num_tokens": 18456728.0,
      "step": 8890
    },
    {
      "entropy": 5.682804965972901,
      "epoch": 0.8068759071117562,
      "grad_norm": 1.125,
      "learning_rate": 0.000494225740925052,
      "loss": 5.3729,
      "mean_token_accuracy": 0.17928425073623658,
      "num_tokens": 18468813.0,
      "step": 8895
    },
    {
      "entropy": 5.553705883026123,
      "epoch": 0.8073294629898403,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004942184553626591,
      "loss": 5.2608,
      "mean_token_accuracy": 0.185865418612957,
      "num_tokens": 18477912.0,
      "step": 8900
    },
    {
      "entropy": 5.464331388473511,
      "epoch": 0.8077830188679245,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942111652667726,
      "loss": 5.2155,
      "mean_token_accuracy": 0.19424942433834075,
      "num_tokens": 18488115.0,
      "step": 8905
    },
    {
      "entropy": 5.492030239105224,
      "epoch": 0.8082365747460087,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942038706375433,
      "loss": 5.2499,
      "mean_token_accuracy": 0.18399515151977539,
      "num_tokens": 18498853.0,
      "step": 8910
    },
    {
      "entropy": 5.5277331352233885,
      "epoch": 0.8086901306240929,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941965714751222,
      "loss": 5.2947,
      "mean_token_accuracy": 0.18574539422988892,
      "num_tokens": 18510047.0,
      "step": 8915
    },
    {
      "entropy": 5.511319303512574,
      "epoch": 0.809143686502177,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941892677796601,
      "loss": 5.2691,
      "mean_token_accuracy": 0.18705060333013535,
      "num_tokens": 18519673.0,
      "step": 8920
    },
    {
      "entropy": 5.497614240646362,
      "epoch": 0.8095972423802612,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004941819595513081,
      "loss": 5.2878,
      "mean_token_accuracy": 0.18286352902650832,
      "num_tokens": 18530090.0,
      "step": 8925
    },
    {
      "entropy": 5.597482538223266,
      "epoch": 0.8100507982583455,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004941746467902173,
      "loss": 5.3243,
      "mean_token_accuracy": 0.1817130759358406,
      "num_tokens": 18541862.0,
      "step": 8930
    },
    {
      "entropy": 5.573345136642456,
      "epoch": 0.8105043541364296,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000494167329496539,
      "loss": 5.3369,
      "mean_token_accuracy": 0.17775724679231644,
      "num_tokens": 18551624.0,
      "step": 8935
    },
    {
      "entropy": 5.514851140975952,
      "epoch": 0.8109579100145138,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941600076704244,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18574765473604202,
      "num_tokens": 18561784.0,
      "step": 8940
    },
    {
      "entropy": 5.55006742477417,
      "epoch": 0.8114114658925979,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941526813120251,
      "loss": 5.318,
      "mean_token_accuracy": 0.1817800968885422,
      "num_tokens": 18572110.0,
      "step": 8945
    },
    {
      "entropy": 5.550695466995239,
      "epoch": 0.8118650217706821,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941453504214924,
      "loss": 5.2702,
      "mean_token_accuracy": 0.1783347800374031,
      "num_tokens": 18582012.0,
      "step": 8950
    },
    {
      "entropy": 5.4474639892578125,
      "epoch": 0.8123185776487664,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004941380149989781,
      "loss": 5.1115,
      "mean_token_accuracy": 0.1958514556288719,
      "num_tokens": 18592850.0,
      "step": 8955
    },
    {
      "entropy": 5.4674022674560545,
      "epoch": 0.8127721335268505,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004941306750446336,
      "loss": 5.2945,
      "mean_token_accuracy": 0.1863408148288727,
      "num_tokens": 18602590.0,
      "step": 8960
    },
    {
      "entropy": 5.463942813873291,
      "epoch": 0.8132256894049347,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494123330558611,
      "loss": 5.1962,
      "mean_token_accuracy": 0.19270357340574265,
      "num_tokens": 18612607.0,
      "step": 8965
    },
    {
      "entropy": 5.4990180969238285,
      "epoch": 0.8136792452830188,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941159815410622,
      "loss": 5.1769,
      "mean_token_accuracy": 0.19223012775182724,
      "num_tokens": 18622936.0,
      "step": 8970
    },
    {
      "entropy": 5.516388368606568,
      "epoch": 0.8141328011611031,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000494108627992139,
      "loss": 5.3316,
      "mean_token_accuracy": 0.18003688007593155,
      "num_tokens": 18633997.0,
      "step": 8975
    },
    {
      "entropy": 5.481217002868652,
      "epoch": 0.8145863570391872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004941012699119936,
      "loss": 5.1633,
      "mean_token_accuracy": 0.19657521694898605,
      "num_tokens": 18644347.0,
      "step": 8980
    },
    {
      "entropy": 5.528944206237793,
      "epoch": 0.8150399129172714,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004940939073007779,
      "loss": 5.2399,
      "mean_token_accuracy": 0.18880648016929627,
      "num_tokens": 18654989.0,
      "step": 8985
    },
    {
      "entropy": 5.4825615882873535,
      "epoch": 0.8154934687953556,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940865401586446,
      "loss": 5.2116,
      "mean_token_accuracy": 0.19149848967790603,
      "num_tokens": 18665526.0,
      "step": 8990
    },
    {
      "entropy": 5.465846824645996,
      "epoch": 0.8159470246734397,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940791684857458,
      "loss": 5.2131,
      "mean_token_accuracy": 0.19184787720441818,
      "num_tokens": 18676226.0,
      "step": 8995
    },
    {
      "entropy": 5.5053871154785154,
      "epoch": 0.816400580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940717922822339,
      "loss": 5.1511,
      "mean_token_accuracy": 0.1933094546198845,
      "num_tokens": 18686499.0,
      "step": 9000
    },
    {
      "epoch": 0.816400580551524,
      "eval_entropy": 5.358133128541627,
      "eval_loss": 5.252007961273193,
      "eval_mean_token_accuracy": 0.19481117964819142,
      "eval_num_tokens": 18686499.0,
      "eval_runtime": 20.7388,
      "eval_samples_per_second": 1705.014,
      "eval_steps_per_second": 213.127,
      "step": 9000
    },
    {
      "entropy": 5.569316291809082,
      "epoch": 0.8168541364296081,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004940644115482615,
      "loss": 5.2286,
      "mean_token_accuracy": 0.18744052350521087,
      "num_tokens": 18696543.0,
      "step": 9005
    },
    {
      "entropy": 5.459391164779663,
      "epoch": 0.8173076923076923,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940570262839812,
      "loss": 5.2701,
      "mean_token_accuracy": 0.18510979413986206,
      "num_tokens": 18707029.0,
      "step": 9010
    },
    {
      "entropy": 5.494412612915039,
      "epoch": 0.8177612481857764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940496364895457,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18970807492733002,
      "num_tokens": 18716937.0,
      "step": 9015
    },
    {
      "entropy": 5.612600326538086,
      "epoch": 0.8182148040638607,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940422421651082,
      "loss": 5.2543,
      "mean_token_accuracy": 0.1858663633465767,
      "num_tokens": 18727478.0,
      "step": 9020
    },
    {
      "entropy": 5.491196346282959,
      "epoch": 0.8186683599419449,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004940348433108211,
      "loss": 5.1684,
      "mean_token_accuracy": 0.18780348002910613,
      "num_tokens": 18737339.0,
      "step": 9025
    },
    {
      "entropy": 5.48309326171875,
      "epoch": 0.819121915820029,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004940274399268377,
      "loss": 5.276,
      "mean_token_accuracy": 0.18714603930711746,
      "num_tokens": 18747810.0,
      "step": 9030
    },
    {
      "entropy": 5.493037223815918,
      "epoch": 0.8195754716981132,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940200320133109,
      "loss": 5.2579,
      "mean_token_accuracy": 0.18732280135154725,
      "num_tokens": 18757564.0,
      "step": 9035
    },
    {
      "entropy": 5.595683336257935,
      "epoch": 0.8200290275761973,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940126195703941,
      "loss": 5.2798,
      "mean_token_accuracy": 0.18251087218523027,
      "num_tokens": 18768196.0,
      "step": 9040
    },
    {
      "entropy": 5.569827604293823,
      "epoch": 0.8204825834542816,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004940052025982405,
      "loss": 5.2491,
      "mean_token_accuracy": 0.19636135995388032,
      "num_tokens": 18778350.0,
      "step": 9045
    },
    {
      "entropy": 5.475778579711914,
      "epoch": 0.8209361393323658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004939977810970035,
      "loss": 5.2048,
      "mean_token_accuracy": 0.18468899726867677,
      "num_tokens": 18789825.0,
      "step": 9050
    },
    {
      "entropy": 5.49319748878479,
      "epoch": 0.8213896952104499,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939903550668366,
      "loss": 5.2059,
      "mean_token_accuracy": 0.18663357943296432,
      "num_tokens": 18799793.0,
      "step": 9055
    },
    {
      "entropy": 5.511374473571777,
      "epoch": 0.8218432510885341,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004939829245078934,
      "loss": 5.2397,
      "mean_token_accuracy": 0.19193035662174224,
      "num_tokens": 18809771.0,
      "step": 9060
    },
    {
      "entropy": 5.4916544437408445,
      "epoch": 0.8222968069666183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004939754894203275,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18900091648101808,
      "num_tokens": 18820529.0,
      "step": 9065
    },
    {
      "entropy": 5.425688123703003,
      "epoch": 0.8227503628447025,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939680498042926,
      "loss": 5.1706,
      "mean_token_accuracy": 0.19591638892889024,
      "num_tokens": 18830797.0,
      "step": 9070
    },
    {
      "entropy": 5.465984773635864,
      "epoch": 0.8232039187227866,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004939606056599428,
      "loss": 5.1989,
      "mean_token_accuracy": 0.1867137596011162,
      "num_tokens": 18840526.0,
      "step": 9075
    },
    {
      "entropy": 5.483081436157226,
      "epoch": 0.8236574746008708,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004939531569874316,
      "loss": 5.2148,
      "mean_token_accuracy": 0.1837407886981964,
      "num_tokens": 18851201.0,
      "step": 9080
    },
    {
      "entropy": 5.4127593517303465,
      "epoch": 0.824111030478955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939457037869135,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20760098844766617,
      "num_tokens": 18861676.0,
      "step": 9085
    },
    {
      "entropy": 5.430079507827759,
      "epoch": 0.8245645863570392,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004939382460585424,
      "loss": 5.178,
      "mean_token_accuracy": 0.184494811296463,
      "num_tokens": 18872669.0,
      "step": 9090
    },
    {
      "entropy": 5.434750080108643,
      "epoch": 0.8250181422351234,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939307838024726,
      "loss": 5.2201,
      "mean_token_accuracy": 0.1835554853081703,
      "num_tokens": 18882130.0,
      "step": 9095
    },
    {
      "entropy": 5.456328392028809,
      "epoch": 0.8254716981132075,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004939233170188585,
      "loss": 5.2109,
      "mean_token_accuracy": 0.1889977216720581,
      "num_tokens": 18892457.0,
      "step": 9100
    },
    {
      "entropy": 5.537774133682251,
      "epoch": 0.8259252539912917,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004939158457078543,
      "loss": 5.2316,
      "mean_token_accuracy": 0.18953631073236465,
      "num_tokens": 18903152.0,
      "step": 9105
    },
    {
      "entropy": 5.513712215423584,
      "epoch": 0.826378809869376,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004939083698696146,
      "loss": 5.1924,
      "mean_token_accuracy": 0.18583756536245347,
      "num_tokens": 18913740.0,
      "step": 9110
    },
    {
      "entropy": 5.487645053863526,
      "epoch": 0.8268323657474601,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493900889504294,
      "loss": 5.2403,
      "mean_token_accuracy": 0.18907872438430787,
      "num_tokens": 18923501.0,
      "step": 9115
    },
    {
      "entropy": 5.49286789894104,
      "epoch": 0.8272859216255443,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004938934046120473,
      "loss": 5.2371,
      "mean_token_accuracy": 0.18701096922159194,
      "num_tokens": 18933255.0,
      "step": 9120
    },
    {
      "entropy": 5.486018419265747,
      "epoch": 0.8277394775036284,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004938859151930293,
      "loss": 5.1943,
      "mean_token_accuracy": 0.19105144888162612,
      "num_tokens": 18943649.0,
      "step": 9125
    },
    {
      "entropy": 5.471429109573364,
      "epoch": 0.8281930333817126,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938784212473947,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19930363148450853,
      "num_tokens": 18953187.0,
      "step": 9130
    },
    {
      "entropy": 5.544489240646362,
      "epoch": 0.8286465892597968,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938709227752985,
      "loss": 5.2041,
      "mean_token_accuracy": 0.1965079352259636,
      "num_tokens": 18963015.0,
      "step": 9135
    },
    {
      "entropy": 5.482017135620117,
      "epoch": 0.829100145137881,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493863419776896,
      "loss": 5.2664,
      "mean_token_accuracy": 0.1819402039051056,
      "num_tokens": 18974184.0,
      "step": 9140
    },
    {
      "entropy": 5.475565242767334,
      "epoch": 0.8295537010159652,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004938559122523421,
      "loss": 5.2073,
      "mean_token_accuracy": 0.1881326138973236,
      "num_tokens": 18985036.0,
      "step": 9145
    },
    {
      "entropy": 5.559447145462036,
      "epoch": 0.8300072568940493,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938484002017921,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1855930745601654,
      "num_tokens": 18995237.0,
      "step": 9150
    },
    {
      "entropy": 5.519651746749878,
      "epoch": 0.8304608127721336,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938408836254015,
      "loss": 5.3199,
      "mean_token_accuracy": 0.17952737510204314,
      "num_tokens": 19006272.0,
      "step": 9155
    },
    {
      "entropy": 5.461530971527099,
      "epoch": 0.8309143686502177,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938333625233258,
      "loss": 5.2331,
      "mean_token_accuracy": 0.18802201896905898,
      "num_tokens": 19016728.0,
      "step": 9160
    },
    {
      "entropy": 5.531713771820068,
      "epoch": 0.8313679245283019,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004938258368957202,
      "loss": 5.2856,
      "mean_token_accuracy": 0.17781373262405395,
      "num_tokens": 19028024.0,
      "step": 9165
    },
    {
      "entropy": 5.528930377960205,
      "epoch": 0.831821480406386,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938183067427406,
      "loss": 5.2252,
      "mean_token_accuracy": 0.1947782501578331,
      "num_tokens": 19038218.0,
      "step": 9170
    },
    {
      "entropy": 5.461911010742187,
      "epoch": 0.8322750362844702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004938107720645426,
      "loss": 5.254,
      "mean_token_accuracy": 0.18719436675310136,
      "num_tokens": 19048574.0,
      "step": 9175
    },
    {
      "entropy": 5.498357486724854,
      "epoch": 0.8327285921625545,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938032328612821,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18767004311084748,
      "num_tokens": 19057721.0,
      "step": 9180
    },
    {
      "entropy": 5.501523447036743,
      "epoch": 0.8331821480406386,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493795689133115,
      "loss": 5.2002,
      "mean_token_accuracy": 0.1808902621269226,
      "num_tokens": 19067582.0,
      "step": 9185
    },
    {
      "entropy": 5.48468132019043,
      "epoch": 0.8336357039187228,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937881408801974,
      "loss": 5.1627,
      "mean_token_accuracy": 0.19163392037153243,
      "num_tokens": 19077896.0,
      "step": 9190
    },
    {
      "entropy": 5.507964372634888,
      "epoch": 0.8340892597968069,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493780588102685,
      "loss": 5.2563,
      "mean_token_accuracy": 0.192536298930645,
      "num_tokens": 19088082.0,
      "step": 9195
    },
    {
      "entropy": 5.55622878074646,
      "epoch": 0.8345428156748912,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937730308007346,
      "loss": 5.2814,
      "mean_token_accuracy": 0.18960726112127305,
      "num_tokens": 19097176.0,
      "step": 9200
    },
    {
      "entropy": 5.497657394409179,
      "epoch": 0.8349963715529753,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004937654689745021,
      "loss": 5.1633,
      "mean_token_accuracy": 0.19420606195926665,
      "num_tokens": 19107835.0,
      "step": 9205
    },
    {
      "entropy": 5.424288654327393,
      "epoch": 0.8354499274310595,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004937579026241439,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19160815477371215,
      "num_tokens": 19118028.0,
      "step": 9210
    },
    {
      "entropy": 5.48676643371582,
      "epoch": 0.8359034833091437,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004937503317498165,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19056888222694396,
      "num_tokens": 19129223.0,
      "step": 9215
    },
    {
      "entropy": 5.492285490036011,
      "epoch": 0.8363570391872278,
      "grad_norm": 0.875,
      "learning_rate": 0.0004937427563516765,
      "loss": 5.3239,
      "mean_token_accuracy": 0.17992374151945115,
      "num_tokens": 19140048.0,
      "step": 9220
    },
    {
      "entropy": 5.564370107650757,
      "epoch": 0.8368105950653121,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937351764298805,
      "loss": 5.287,
      "mean_token_accuracy": 0.178375805914402,
      "num_tokens": 19150853.0,
      "step": 9225
    },
    {
      "entropy": 5.543662738800049,
      "epoch": 0.8372641509433962,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937275919845855,
      "loss": 5.2718,
      "mean_token_accuracy": 0.18463384956121445,
      "num_tokens": 19161645.0,
      "step": 9230
    },
    {
      "entropy": 5.514537143707275,
      "epoch": 0.8377177068214804,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004937200030159481,
      "loss": 5.3572,
      "mean_token_accuracy": 0.18064510822296143,
      "num_tokens": 19173276.0,
      "step": 9235
    },
    {
      "entropy": 5.551925563812256,
      "epoch": 0.8381712626995645,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937124095241252,
      "loss": 5.2393,
      "mean_token_accuracy": 0.19420024901628494,
      "num_tokens": 19182949.0,
      "step": 9240
    },
    {
      "entropy": 5.564868927001953,
      "epoch": 0.8386248185776488,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004937048115092739,
      "loss": 5.2393,
      "mean_token_accuracy": 0.1857432618737221,
      "num_tokens": 19193504.0,
      "step": 9245
    },
    {
      "entropy": 5.488748836517334,
      "epoch": 0.839078374455733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004936972089715515,
      "loss": 5.3068,
      "mean_token_accuracy": 0.1808287188410759,
      "num_tokens": 19203406.0,
      "step": 9250
    },
    {
      "entropy": 5.481642866134644,
      "epoch": 0.8395319303338171,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493689601911115,
      "loss": 5.2549,
      "mean_token_accuracy": 0.18709076792001725,
      "num_tokens": 19214325.0,
      "step": 9255
    },
    {
      "entropy": 5.523412084579467,
      "epoch": 0.8399854862119013,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936819903281219,
      "loss": 5.3033,
      "mean_token_accuracy": 0.17142430394887925,
      "num_tokens": 19226466.0,
      "step": 9260
    },
    {
      "entropy": 5.605291700363159,
      "epoch": 0.8404390420899854,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004936743742227295,
      "loss": 5.3239,
      "mean_token_accuracy": 0.18541250377893448,
      "num_tokens": 19237032.0,
      "step": 9265
    },
    {
      "entropy": 5.553051137924195,
      "epoch": 0.8408925979680697,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004936667535950952,
      "loss": 5.3008,
      "mean_token_accuracy": 0.18086010366678237,
      "num_tokens": 19248009.0,
      "step": 9270
    },
    {
      "entropy": 5.570973443984985,
      "epoch": 0.8413461538461539,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936591284453768,
      "loss": 5.3585,
      "mean_token_accuracy": 0.18207215815782546,
      "num_tokens": 19259363.0,
      "step": 9275
    },
    {
      "entropy": 5.544232416152954,
      "epoch": 0.841799709724238,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936514987737319,
      "loss": 5.2609,
      "mean_token_accuracy": 0.18669627606868744,
      "num_tokens": 19269191.0,
      "step": 9280
    },
    {
      "entropy": 5.55557804107666,
      "epoch": 0.8422532656023222,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004936438645803182,
      "loss": 5.2303,
      "mean_token_accuracy": 0.18535475730895995,
      "num_tokens": 19280278.0,
      "step": 9285
    },
    {
      "entropy": 5.522267198562622,
      "epoch": 0.8427068214804064,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004936362258652938,
      "loss": 5.2129,
      "mean_token_accuracy": 0.1926440954208374,
      "num_tokens": 19290130.0,
      "step": 9290
    },
    {
      "entropy": 5.5636334896087645,
      "epoch": 0.8431603773584906,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004936285826288165,
      "loss": 5.3232,
      "mean_token_accuracy": 0.19174562841653825,
      "num_tokens": 19301669.0,
      "step": 9295
    },
    {
      "entropy": 5.453507995605468,
      "epoch": 0.8436139332365747,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004936209348710444,
      "loss": 5.1815,
      "mean_token_accuracy": 0.195254348218441,
      "num_tokens": 19312490.0,
      "step": 9300
    },
    {
      "entropy": 5.478431653976441,
      "epoch": 0.8440674891146589,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004936132825921355,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1872256428003311,
      "num_tokens": 19321743.0,
      "step": 9305
    },
    {
      "entropy": 5.496931743621826,
      "epoch": 0.844521044992743,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004936056257922484,
      "loss": 5.1403,
      "mean_token_accuracy": 0.193279492855072,
      "num_tokens": 19332149.0,
      "step": 9310
    },
    {
      "entropy": 5.508009672164917,
      "epoch": 0.8449746008708273,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935979644715412,
      "loss": 5.2505,
      "mean_token_accuracy": 0.18201816231012344,
      "num_tokens": 19341655.0,
      "step": 9315
    },
    {
      "entropy": 5.538322639465332,
      "epoch": 0.8454281567489115,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935902986301724,
      "loss": 5.282,
      "mean_token_accuracy": 0.1822471797466278,
      "num_tokens": 19351404.0,
      "step": 9320
    },
    {
      "entropy": 5.544085264205933,
      "epoch": 0.8458817126269956,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935826282683005,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18208804428577424,
      "num_tokens": 19361788.0,
      "step": 9325
    },
    {
      "entropy": 5.524391937255859,
      "epoch": 0.8463352685050798,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004935749533860842,
      "loss": 5.2728,
      "mean_token_accuracy": 0.18818075209856033,
      "num_tokens": 19372902.0,
      "step": 9330
    },
    {
      "entropy": 5.45745267868042,
      "epoch": 0.846788824383164,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493567273983682,
      "loss": 5.1709,
      "mean_token_accuracy": 0.1850125476717949,
      "num_tokens": 19382181.0,
      "step": 9335
    },
    {
      "entropy": 5.514938163757324,
      "epoch": 0.8472423802612482,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000493559590061253,
      "loss": 5.229,
      "mean_token_accuracy": 0.18603635281324388,
      "num_tokens": 19393692.0,
      "step": 9340
    },
    {
      "entropy": 5.539198207855224,
      "epoch": 0.8476959361393324,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004935519016189559,
      "loss": 5.2581,
      "mean_token_accuracy": 0.19449927061796188,
      "num_tokens": 19404587.0,
      "step": 9345
    },
    {
      "entropy": 5.51031002998352,
      "epoch": 0.8481494920174165,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935442086569498,
      "loss": 5.2329,
      "mean_token_accuracy": 0.1873893231153488,
      "num_tokens": 19415296.0,
      "step": 9350
    },
    {
      "entropy": 5.568622016906739,
      "epoch": 0.8486030478955007,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935365111753939,
      "loss": 5.2907,
      "mean_token_accuracy": 0.18625998497009277,
      "num_tokens": 19426119.0,
      "step": 9355
    },
    {
      "entropy": 5.55921778678894,
      "epoch": 0.8490566037735849,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004935288091744472,
      "loss": 5.2751,
      "mean_token_accuracy": 0.183457387983799,
      "num_tokens": 19436828.0,
      "step": 9360
    },
    {
      "entropy": 5.553300952911377,
      "epoch": 0.8495101596516691,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493521102654269,
      "loss": 5.2551,
      "mean_token_accuracy": 0.1870663121342659,
      "num_tokens": 19446583.0,
      "step": 9365
    },
    {
      "entropy": 5.452865266799927,
      "epoch": 0.8499637155297532,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935133916150187,
      "loss": 5.1336,
      "mean_token_accuracy": 0.19529592543840407,
      "num_tokens": 19456757.0,
      "step": 9370
    },
    {
      "entropy": 5.455078649520874,
      "epoch": 0.8504172714078374,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004935056760568558,
      "loss": 5.2172,
      "mean_token_accuracy": 0.18394441306591033,
      "num_tokens": 19467624.0,
      "step": 9375
    },
    {
      "entropy": 5.49309778213501,
      "epoch": 0.8508708272859217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934979559799399,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19133689105510712,
      "num_tokens": 19477759.0,
      "step": 9380
    },
    {
      "entropy": 5.515308380126953,
      "epoch": 0.8513243831640058,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004934902313844306,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19061967581510544,
      "num_tokens": 19488877.0,
      "step": 9385
    },
    {
      "entropy": 5.406462621688843,
      "epoch": 0.85177793904209,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004934825022704875,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19388926029205322,
      "num_tokens": 19499170.0,
      "step": 9390
    },
    {
      "entropy": 5.4668279647827145,
      "epoch": 0.8522314949201741,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004934747686382707,
      "loss": 5.2111,
      "mean_token_accuracy": 0.18104320466518403,
      "num_tokens": 19509944.0,
      "step": 9395
    },
    {
      "entropy": 5.513626003265381,
      "epoch": 0.8526850507982583,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934670304879398,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19153966307640075,
      "num_tokens": 19519014.0,
      "step": 9400
    },
    {
      "entropy": 5.448259162902832,
      "epoch": 0.8531386066763426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934592878196553,
      "loss": 5.187,
      "mean_token_accuracy": 0.19212067872285843,
      "num_tokens": 19528586.0,
      "step": 9405
    },
    {
      "entropy": 5.49515585899353,
      "epoch": 0.8535921625544267,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493451540633577,
      "loss": 5.3225,
      "mean_token_accuracy": 0.18525902032852173,
      "num_tokens": 19539912.0,
      "step": 9410
    },
    {
      "entropy": 5.518271636962891,
      "epoch": 0.8540457184325109,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004934437889298651,
      "loss": 5.2805,
      "mean_token_accuracy": 0.1793683409690857,
      "num_tokens": 19549860.0,
      "step": 9415
    },
    {
      "entropy": 5.540545892715454,
      "epoch": 0.854499274310595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004934360327086799,
      "loss": 5.2327,
      "mean_token_accuracy": 0.18420504480600358,
      "num_tokens": 19559446.0,
      "step": 9420
    },
    {
      "entropy": 5.539361572265625,
      "epoch": 0.8549528301886793,
      "grad_norm": 0.921875,
      "learning_rate": 0.000493428271970182,
      "loss": 5.2417,
      "mean_token_accuracy": 0.1881583198904991,
      "num_tokens": 19570282.0,
      "step": 9425
    },
    {
      "entropy": 5.510746669769287,
      "epoch": 0.8554063860667634,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004934205067145318,
      "loss": 5.1678,
      "mean_token_accuracy": 0.18670809119939805,
      "num_tokens": 19581087.0,
      "step": 9430
    },
    {
      "entropy": 5.492790555953979,
      "epoch": 0.8558599419448476,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004934127369418898,
      "loss": 5.2616,
      "mean_token_accuracy": 0.18760718405246735,
      "num_tokens": 19592100.0,
      "step": 9435
    },
    {
      "entropy": 5.487104320526123,
      "epoch": 0.8563134978229318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934049626524168,
      "loss": 5.2409,
      "mean_token_accuracy": 0.18586532920598983,
      "num_tokens": 19602557.0,
      "step": 9440
    },
    {
      "entropy": 5.5392913818359375,
      "epoch": 0.8567670537010159,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004933971838462734,
      "loss": 5.2899,
      "mean_token_accuracy": 0.18158393800258638,
      "num_tokens": 19613643.0,
      "step": 9445
    },
    {
      "entropy": 5.4967609405517575,
      "epoch": 0.8572206095791002,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933894005236207,
      "loss": 5.2001,
      "mean_token_accuracy": 0.1887534514069557,
      "num_tokens": 19624180.0,
      "step": 9450
    },
    {
      "entropy": 5.516611337661743,
      "epoch": 0.8576741654571843,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933816126846195,
      "loss": 5.2246,
      "mean_token_accuracy": 0.18624820560216904,
      "num_tokens": 19634799.0,
      "step": 9455
    },
    {
      "entropy": 5.485091066360473,
      "epoch": 0.8581277213352685,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933738203294309,
      "loss": 5.1903,
      "mean_token_accuracy": 0.19355441778898239,
      "num_tokens": 19644618.0,
      "step": 9460
    },
    {
      "entropy": 5.47304892539978,
      "epoch": 0.8585812772133526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493366023458216,
      "loss": 5.1999,
      "mean_token_accuracy": 0.18136429488658906,
      "num_tokens": 19655550.0,
      "step": 9465
    },
    {
      "entropy": 5.517861652374267,
      "epoch": 0.8590348330914369,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004933582220711362,
      "loss": 5.2673,
      "mean_token_accuracy": 0.1912809804081917,
      "num_tokens": 19665734.0,
      "step": 9470
    },
    {
      "entropy": 5.552363920211792,
      "epoch": 0.8594883889695211,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004933504161683526,
      "loss": 5.2249,
      "mean_token_accuracy": 0.18164515048265456,
      "num_tokens": 19677130.0,
      "step": 9475
    },
    {
      "entropy": 5.463605499267578,
      "epoch": 0.8599419448476052,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004933426057500269,
      "loss": 5.1415,
      "mean_token_accuracy": 0.19647871553897858,
      "num_tokens": 19687374.0,
      "step": 9480
    },
    {
      "entropy": 5.548071193695068,
      "epoch": 0.8603955007256894,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933347908163203,
      "loss": 5.2359,
      "mean_token_accuracy": 0.1858404666185379,
      "num_tokens": 19698182.0,
      "step": 9485
    },
    {
      "entropy": 5.489673233032226,
      "epoch": 0.8608490566037735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933269713673947,
      "loss": 5.209,
      "mean_token_accuracy": 0.18983528316020964,
      "num_tokens": 19707664.0,
      "step": 9490
    },
    {
      "entropy": 5.418010711669922,
      "epoch": 0.8613026124818578,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933191474034116,
      "loss": 5.2464,
      "mean_token_accuracy": 0.18493668735027313,
      "num_tokens": 19717647.0,
      "step": 9495
    },
    {
      "entropy": 5.481851577758789,
      "epoch": 0.861756168359942,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493311318924533,
      "loss": 5.2824,
      "mean_token_accuracy": 0.18406570553779603,
      "num_tokens": 19727444.0,
      "step": 9500
    },
    {
      "entropy": 5.547173070907593,
      "epoch": 0.8622097242380261,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933034859309208,
      "loss": 5.3011,
      "mean_token_accuracy": 0.1775568827986717,
      "num_tokens": 19738719.0,
      "step": 9505
    },
    {
      "entropy": 5.583591938018799,
      "epoch": 0.8626632801161103,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932956484227367,
      "loss": 5.2774,
      "mean_token_accuracy": 0.19096940159797668,
      "num_tokens": 19749786.0,
      "step": 9510
    },
    {
      "entropy": 5.515065097808838,
      "epoch": 0.8631168359941945,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000493287806400143,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19534773230552674,
      "num_tokens": 19759623.0,
      "step": 9515
    },
    {
      "entropy": 5.536817026138306,
      "epoch": 0.8635703918722787,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004932799598633018,
      "loss": 5.2199,
      "mean_token_accuracy": 0.18721261024475097,
      "num_tokens": 19770117.0,
      "step": 9520
    },
    {
      "entropy": 5.51656551361084,
      "epoch": 0.8640239477503628,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004932721088123753,
      "loss": 5.3074,
      "mean_token_accuracy": 0.18952126950025558,
      "num_tokens": 19780748.0,
      "step": 9525
    },
    {
      "entropy": 5.547321510314942,
      "epoch": 0.864477503628447,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004932642532475262,
      "loss": 5.2737,
      "mean_token_accuracy": 0.18902790993452073,
      "num_tokens": 19791816.0,
      "step": 9530
    },
    {
      "entropy": 5.57507176399231,
      "epoch": 0.8649310595065312,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932563931689168,
      "loss": 5.3212,
      "mean_token_accuracy": 0.18597906678915024,
      "num_tokens": 19801096.0,
      "step": 9535
    },
    {
      "entropy": 5.51464467048645,
      "epoch": 0.8653846153846154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932485285767093,
      "loss": 5.1896,
      "mean_token_accuracy": 0.19145326763391496,
      "num_tokens": 19811258.0,
      "step": 9540
    },
    {
      "entropy": 5.531577491760254,
      "epoch": 0.8658381712626996,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004932406594710667,
      "loss": 5.2594,
      "mean_token_accuracy": 0.19362047761678697,
      "num_tokens": 19821765.0,
      "step": 9545
    },
    {
      "entropy": 5.4913736343383786,
      "epoch": 0.8662917271407837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932327858521517,
      "loss": 5.1323,
      "mean_token_accuracy": 0.19672255367040634,
      "num_tokens": 19831537.0,
      "step": 9550
    },
    {
      "entropy": 5.447400045394898,
      "epoch": 0.8667452830188679,
      "grad_norm": 0.984375,
      "learning_rate": 0.000493224907720127,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19029980450868605,
      "num_tokens": 19841830.0,
      "step": 9555
    },
    {
      "entropy": 5.512387990951538,
      "epoch": 0.8671988388969522,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004932170250751556,
      "loss": 5.256,
      "mean_token_accuracy": 0.18012520521879197,
      "num_tokens": 19851896.0,
      "step": 9560
    },
    {
      "entropy": 5.537576293945312,
      "epoch": 0.8676523947750363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932091379174006,
      "loss": 5.289,
      "mean_token_accuracy": 0.1798422083258629,
      "num_tokens": 19861735.0,
      "step": 9565
    },
    {
      "entropy": 5.497203254699707,
      "epoch": 0.8681059506531205,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493201246247025,
      "loss": 5.1747,
      "mean_token_accuracy": 0.19459885060787202,
      "num_tokens": 19872333.0,
      "step": 9570
    },
    {
      "entropy": 5.521367979049683,
      "epoch": 0.8685595065312046,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000493193350064192,
      "loss": 5.1918,
      "mean_token_accuracy": 0.18699753284454346,
      "num_tokens": 19882896.0,
      "step": 9575
    },
    {
      "entropy": 5.4799529075622555,
      "epoch": 0.8690130624092888,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004931854493690649,
      "loss": 5.1517,
      "mean_token_accuracy": 0.189018677175045,
      "num_tokens": 19893767.0,
      "step": 9580
    },
    {
      "entropy": 5.41310601234436,
      "epoch": 0.869466618287373,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004931775441618071,
      "loss": 5.1264,
      "mean_token_accuracy": 0.19416602104902267,
      "num_tokens": 19904242.0,
      "step": 9585
    },
    {
      "entropy": 5.478859567642212,
      "epoch": 0.8699201741654572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004931696344425821,
      "loss": 5.2633,
      "mean_token_accuracy": 0.188524429500103,
      "num_tokens": 19913898.0,
      "step": 9590
    },
    {
      "entropy": 5.614615726470947,
      "epoch": 0.8703737300435413,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931617202115535,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19075826853513717,
      "num_tokens": 19923756.0,
      "step": 9595
    },
    {
      "entropy": 5.524111700057984,
      "epoch": 0.8708272859216255,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004931538014688849,
      "loss": 5.1939,
      "mean_token_accuracy": 0.1916771948337555,
      "num_tokens": 19932720.0,
      "step": 9600
    },
    {
      "entropy": 5.558842515945434,
      "epoch": 0.8712808417997098,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004931458782147401,
      "loss": 5.4063,
      "mean_token_accuracy": 0.17672424167394638,
      "num_tokens": 19942962.0,
      "step": 9605
    },
    {
      "entropy": 5.560279273986817,
      "epoch": 0.8717343976777939,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000493137950449283,
      "loss": 5.276,
      "mean_token_accuracy": 0.19128557592630385,
      "num_tokens": 19953735.0,
      "step": 9610
    },
    {
      "entropy": 5.505196857452392,
      "epoch": 0.8721879535558781,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931300181726775,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18588921427726746,
      "num_tokens": 19964416.0,
      "step": 9615
    },
    {
      "entropy": 5.517534017562866,
      "epoch": 0.8726415094339622,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004931220813850875,
      "loss": 5.2602,
      "mean_token_accuracy": 0.18303596675395967,
      "num_tokens": 19974531.0,
      "step": 9620
    },
    {
      "entropy": 5.540431547164917,
      "epoch": 0.8730950653120464,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004931141400866773,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19180816859006883,
      "num_tokens": 19984964.0,
      "step": 9625
    },
    {
      "entropy": 5.482637929916382,
      "epoch": 0.8735486211901307,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004931061942776112,
      "loss": 5.1951,
      "mean_token_accuracy": 0.19486953020095826,
      "num_tokens": 19995568.0,
      "step": 9630
    },
    {
      "entropy": 5.608856678009033,
      "epoch": 0.8740021770682148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930982439580534,
      "loss": 5.3692,
      "mean_token_accuracy": 0.17587941586971284,
      "num_tokens": 20005856.0,
      "step": 9635
    },
    {
      "entropy": 5.535829830169678,
      "epoch": 0.874455732946299,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004930902891281683,
      "loss": 5.2749,
      "mean_token_accuracy": 0.18487472534179689,
      "num_tokens": 20015683.0,
      "step": 9640
    },
    {
      "entropy": 5.421100807189942,
      "epoch": 0.8749092888243831,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004930823297881205,
      "loss": 5.0891,
      "mean_token_accuracy": 0.1941447973251343,
      "num_tokens": 20025998.0,
      "step": 9645
    },
    {
      "entropy": 5.542952680587769,
      "epoch": 0.8753628447024674,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004930743659380745,
      "loss": 5.2547,
      "mean_token_accuracy": 0.18763321340084077,
      "num_tokens": 20036486.0,
      "step": 9650
    },
    {
      "entropy": 5.459162616729737,
      "epoch": 0.8758164005805515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493066397578195,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19946982264518737,
      "num_tokens": 20046784.0,
      "step": 9655
    },
    {
      "entropy": 5.504218053817749,
      "epoch": 0.8762699564586357,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004930584247086468,
      "loss": 5.2238,
      "mean_token_accuracy": 0.18850288093090056,
      "num_tokens": 20056999.0,
      "step": 9660
    },
    {
      "entropy": 5.464111852645874,
      "epoch": 0.8767235123367199,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493050447329595,
      "loss": 5.1184,
      "mean_token_accuracy": 0.19502724707126617,
      "num_tokens": 20067612.0,
      "step": 9665
    },
    {
      "entropy": 5.520261335372925,
      "epoch": 0.877177068214804,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004930424654412043,
      "loss": 5.2666,
      "mean_token_accuracy": 0.1904868870973587,
      "num_tokens": 20077330.0,
      "step": 9670
    },
    {
      "entropy": 5.511299180984497,
      "epoch": 0.8776306240928883,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930344790436398,
      "loss": 5.26,
      "mean_token_accuracy": 0.18457795083522796,
      "num_tokens": 20086801.0,
      "step": 9675
    },
    {
      "entropy": 5.532819318771362,
      "epoch": 0.8780841799709724,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004930264881370666,
      "loss": 5.2115,
      "mean_token_accuracy": 0.18478081822395326,
      "num_tokens": 20097047.0,
      "step": 9680
    },
    {
      "entropy": 5.5098706722259525,
      "epoch": 0.8785377358490566,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930184927216504,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19049520045518875,
      "num_tokens": 20107185.0,
      "step": 9685
    },
    {
      "entropy": 5.523552942276001,
      "epoch": 0.8789912917271407,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930104927975559,
      "loss": 5.186,
      "mean_token_accuracy": 0.1903187081217766,
      "num_tokens": 20118493.0,
      "step": 9690
    },
    {
      "entropy": 5.510832071304321,
      "epoch": 0.879444847605225,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004930024883649489,
      "loss": 5.2423,
      "mean_token_accuracy": 0.1870877966284752,
      "num_tokens": 20129294.0,
      "step": 9695
    },
    {
      "entropy": 5.439377927780152,
      "epoch": 0.8798984034833092,
      "grad_norm": 1.0,
      "learning_rate": 0.000492994479423995,
      "loss": 5.2038,
      "mean_token_accuracy": 0.19338049739599228,
      "num_tokens": 20139194.0,
      "step": 9700
    },
    {
      "entropy": 5.4923159122467045,
      "epoch": 0.8803519593613933,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004929864659748597,
      "loss": 5.2241,
      "mean_token_accuracy": 0.18959469497203826,
      "num_tokens": 20149681.0,
      "step": 9705
    },
    {
      "entropy": 5.547377967834473,
      "epoch": 0.8808055152394775,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004929784480177087,
      "loss": 5.3143,
      "mean_token_accuracy": 0.18003763556480407,
      "num_tokens": 20160557.0,
      "step": 9710
    },
    {
      "entropy": 5.4973267078399655,
      "epoch": 0.8812590711175616,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004929704255527078,
      "loss": 5.1858,
      "mean_token_accuracy": 0.18966007977724075,
      "num_tokens": 20171047.0,
      "step": 9715
    },
    {
      "entropy": 5.519015216827393,
      "epoch": 0.8817126269956459,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492962398580023,
      "loss": 5.3549,
      "mean_token_accuracy": 0.18190275877714157,
      "num_tokens": 20182148.0,
      "step": 9720
    },
    {
      "entropy": 5.434145259857178,
      "epoch": 0.88216618287373,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929543670998203,
      "loss": 5.107,
      "mean_token_accuracy": 0.19520194828510284,
      "num_tokens": 20192131.0,
      "step": 9725
    },
    {
      "entropy": 5.524809503555298,
      "epoch": 0.8826197387518142,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004929463311122657,
      "loss": 5.228,
      "mean_token_accuracy": 0.1856295123696327,
      "num_tokens": 20202805.0,
      "step": 9730
    },
    {
      "entropy": 5.528243207931519,
      "epoch": 0.8830732946298984,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004929382906175253,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19349892288446427,
      "num_tokens": 20212702.0,
      "step": 9735
    },
    {
      "entropy": 5.468289422988891,
      "epoch": 0.8835268505079826,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929302456157657,
      "loss": 5.1914,
      "mean_token_accuracy": 0.1951130896806717,
      "num_tokens": 20222661.0,
      "step": 9740
    },
    {
      "entropy": 5.48810658454895,
      "epoch": 0.8839804063860668,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929221961071531,
      "loss": 5.239,
      "mean_token_accuracy": 0.1886950433254242,
      "num_tokens": 20233012.0,
      "step": 9745
    },
    {
      "entropy": 5.531930875778198,
      "epoch": 0.8844339622641509,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929141420918538,
      "loss": 5.2469,
      "mean_token_accuracy": 0.1919715628027916,
      "num_tokens": 20242918.0,
      "step": 9750
    },
    {
      "entropy": 5.4830118179321286,
      "epoch": 0.8848875181422351,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929060835700347,
      "loss": 5.1766,
      "mean_token_accuracy": 0.19638995528221131,
      "num_tokens": 20253254.0,
      "step": 9755
    },
    {
      "entropy": 5.4783000469207765,
      "epoch": 0.8853410740203193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492898020541862,
      "loss": 5.2378,
      "mean_token_accuracy": 0.18468834161758424,
      "num_tokens": 20263355.0,
      "step": 9760
    },
    {
      "entropy": 5.460058784484863,
      "epoch": 0.8857946298984035,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004928899530075029,
      "loss": 5.1601,
      "mean_token_accuracy": 0.19061171710491182,
      "num_tokens": 20274716.0,
      "step": 9765
    },
    {
      "entropy": 5.433362579345703,
      "epoch": 0.8862481857764877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004928818809671241,
      "loss": 5.1507,
      "mean_token_accuracy": 0.19014403820037842,
      "num_tokens": 20285631.0,
      "step": 9770
    },
    {
      "entropy": 5.545337724685669,
      "epoch": 0.8867017416545718,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928738044208923,
      "loss": 5.2384,
      "mean_token_accuracy": 0.19255442023277283,
      "num_tokens": 20295717.0,
      "step": 9775
    },
    {
      "entropy": 5.456331539154053,
      "epoch": 0.887155297532656,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928657233689748,
      "loss": 5.2041,
      "mean_token_accuracy": 0.1925482451915741,
      "num_tokens": 20306637.0,
      "step": 9780
    },
    {
      "entropy": 5.484044456481934,
      "epoch": 0.8876088534107403,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004928576378115387,
      "loss": 5.1831,
      "mean_token_accuracy": 0.1889451801776886,
      "num_tokens": 20316629.0,
      "step": 9785
    },
    {
      "entropy": 5.5872455596923825,
      "epoch": 0.8880624092888244,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928495477487509,
      "loss": 5.355,
      "mean_token_accuracy": 0.17307652533054352,
      "num_tokens": 20327935.0,
      "step": 9790
    },
    {
      "entropy": 5.530547428131103,
      "epoch": 0.8885159651669086,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928414531807792,
      "loss": 5.1728,
      "mean_token_accuracy": 0.1830437198281288,
      "num_tokens": 20339142.0,
      "step": 9795
    },
    {
      "entropy": 5.4817322254180905,
      "epoch": 0.8889695210449927,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004928333541077906,
      "loss": 5.172,
      "mean_token_accuracy": 0.2016322910785675,
      "num_tokens": 20348474.0,
      "step": 9800
    },
    {
      "entropy": 5.479598665237427,
      "epoch": 0.8894230769230769,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928252505299527,
      "loss": 5.2418,
      "mean_token_accuracy": 0.19005548357963561,
      "num_tokens": 20359440.0,
      "step": 9805
    },
    {
      "entropy": 5.569894742965698,
      "epoch": 0.8898766328011611,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004928171424474332,
      "loss": 5.2767,
      "mean_token_accuracy": 0.1847678691148758,
      "num_tokens": 20370355.0,
      "step": 9810
    },
    {
      "entropy": 5.4881508350372314,
      "epoch": 0.8903301886792453,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004928090298603996,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1922898545861244,
      "num_tokens": 20380214.0,
      "step": 9815
    },
    {
      "entropy": 5.526920175552368,
      "epoch": 0.8907837445573294,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004928009127690199,
      "loss": 5.1952,
      "mean_token_accuracy": 0.18694787621498107,
      "num_tokens": 20389190.0,
      "step": 9820
    },
    {
      "entropy": 5.458229827880859,
      "epoch": 0.8912373004354136,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927927911734617,
      "loss": 5.232,
      "mean_token_accuracy": 0.18914854824542998,
      "num_tokens": 20399531.0,
      "step": 9825
    },
    {
      "entropy": 5.465860223770141,
      "epoch": 0.8916908563134979,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004927846650738931,
      "loss": 5.2215,
      "mean_token_accuracy": 0.18792601376771928,
      "num_tokens": 20410201.0,
      "step": 9830
    },
    {
      "entropy": 5.48523736000061,
      "epoch": 0.892144412191582,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004927765344704821,
      "loss": 5.1693,
      "mean_token_accuracy": 0.19596148431301116,
      "num_tokens": 20421579.0,
      "step": 9835
    },
    {
      "entropy": 5.474275684356689,
      "epoch": 0.8925979680696662,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004927683993633969,
      "loss": 5.2427,
      "mean_token_accuracy": 0.19045363068580629,
      "num_tokens": 20432048.0,
      "step": 9840
    },
    {
      "entropy": 5.553377866744995,
      "epoch": 0.8930515239477503,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004927602597528057,
      "loss": 5.2663,
      "mean_token_accuracy": 0.18120875656604768,
      "num_tokens": 20441752.0,
      "step": 9845
    },
    {
      "entropy": 5.456312799453736,
      "epoch": 0.8935050798258345,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004927521156388769,
      "loss": 5.13,
      "mean_token_accuracy": 0.1914264976978302,
      "num_tokens": 20451693.0,
      "step": 9850
    },
    {
      "entropy": 5.502097320556641,
      "epoch": 0.8939586357039188,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004927439670217788,
      "loss": 5.2738,
      "mean_token_accuracy": 0.18670168668031692,
      "num_tokens": 20462312.0,
      "step": 9855
    },
    {
      "entropy": 5.483501243591308,
      "epoch": 0.8944121915820029,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049273581390168,
      "loss": 5.1804,
      "mean_token_accuracy": 0.18460638374090194,
      "num_tokens": 20472546.0,
      "step": 9860
    },
    {
      "entropy": 5.506111812591553,
      "epoch": 0.8948657474600871,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000492727656278749,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1930461436510086,
      "num_tokens": 20483266.0,
      "step": 9865
    },
    {
      "entropy": 5.464397954940796,
      "epoch": 0.8953193033381712,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004927194941531547,
      "loss": 5.1512,
      "mean_token_accuracy": 0.18788621425628663,
      "num_tokens": 20492490.0,
      "step": 9870
    },
    {
      "entropy": 5.485383033752441,
      "epoch": 0.8957728592162555,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004927113275250657,
      "loss": 5.1922,
      "mean_token_accuracy": 0.19176370054483413,
      "num_tokens": 20503077.0,
      "step": 9875
    },
    {
      "entropy": 5.464278745651245,
      "epoch": 0.8962264150943396,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004927031563946511,
      "loss": 5.2078,
      "mean_token_accuracy": 0.18573808223009108,
      "num_tokens": 20513771.0,
      "step": 9880
    },
    {
      "entropy": 5.532713842391968,
      "epoch": 0.8966799709724238,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004926949807620796,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1865783840417862,
      "num_tokens": 20523616.0,
      "step": 9885
    },
    {
      "entropy": 5.4971997261047365,
      "epoch": 0.897133526850508,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926868006275204,
      "loss": 5.0925,
      "mean_token_accuracy": 0.19763742238283158,
      "num_tokens": 20533613.0,
      "step": 9890
    },
    {
      "entropy": 5.4906604290008545,
      "epoch": 0.8975870827285921,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004926786159911427,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1880497470498085,
      "num_tokens": 20543812.0,
      "step": 9895
    },
    {
      "entropy": 5.460688543319702,
      "epoch": 0.8980406386066764,
      "grad_norm": 1.0,
      "learning_rate": 0.000492670426853116,
      "loss": 5.1363,
      "mean_token_accuracy": 0.192536161839962,
      "num_tokens": 20553785.0,
      "step": 9900
    },
    {
      "entropy": 5.562257194519043,
      "epoch": 0.8984941944847605,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926622332136092,
      "loss": 5.2559,
      "mean_token_accuracy": 0.19221415668725966,
      "num_tokens": 20563326.0,
      "step": 9905
    },
    {
      "entropy": 5.436643981933594,
      "epoch": 0.8989477503628447,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004926540350727918,
      "loss": 5.1005,
      "mean_token_accuracy": 0.19912851303815843,
      "num_tokens": 20573643.0,
      "step": 9910
    },
    {
      "entropy": 5.426002407073975,
      "epoch": 0.8994013062409288,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926458324308337,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19306659698486328,
      "num_tokens": 20583916.0,
      "step": 9915
    },
    {
      "entropy": 5.516725540161133,
      "epoch": 0.8998548621190131,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004926376252879043,
      "loss": 5.2894,
      "mean_token_accuracy": 0.19082354307174682,
      "num_tokens": 20593984.0,
      "step": 9920
    },
    {
      "entropy": 5.518711185455322,
      "epoch": 0.9003084179970973,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004926294136441733,
      "loss": 5.1265,
      "mean_token_accuracy": 0.1930823028087616,
      "num_tokens": 20603927.0,
      "step": 9925
    },
    {
      "entropy": 5.496563529968261,
      "epoch": 0.9007619738751814,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004926211974998106,
      "loss": 5.2612,
      "mean_token_accuracy": 0.1923177346587181,
      "num_tokens": 20614147.0,
      "step": 9930
    },
    {
      "entropy": 5.523381233215332,
      "epoch": 0.9012155297532656,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000492612976854986,
      "loss": 5.2238,
      "mean_token_accuracy": 0.1915097326040268,
      "num_tokens": 20623631.0,
      "step": 9935
    },
    {
      "entropy": 5.465784215927124,
      "epoch": 0.9016690856313497,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926047517098697,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19526478499174119,
      "num_tokens": 20632566.0,
      "step": 9940
    },
    {
      "entropy": 5.512724876403809,
      "epoch": 0.902122641509434,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004925965220646315,
      "loss": 5.2151,
      "mean_token_accuracy": 0.18857671916484833,
      "num_tokens": 20643246.0,
      "step": 9945
    },
    {
      "entropy": 5.507947826385498,
      "epoch": 0.9025761973875182,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925882879194419,
      "loss": 5.296,
      "mean_token_accuracy": 0.18535405844449998,
      "num_tokens": 20653372.0,
      "step": 9950
    },
    {
      "entropy": 5.5078880310058596,
      "epoch": 0.9030297532656023,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004925800492744712,
      "loss": 5.2362,
      "mean_token_accuracy": 0.18510301411151886,
      "num_tokens": 20662687.0,
      "step": 9955
    },
    {
      "entropy": 5.463442754745484,
      "epoch": 0.9034833091436865,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004925718061298895,
      "loss": 5.1509,
      "mean_token_accuracy": 0.196966715157032,
      "num_tokens": 20673168.0,
      "step": 9960
    },
    {
      "entropy": 5.560179567337036,
      "epoch": 0.9039368650217707,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925635584858675,
      "loss": 5.267,
      "mean_token_accuracy": 0.1841796949505806,
      "num_tokens": 20683100.0,
      "step": 9965
    },
    {
      "entropy": 5.587424993515015,
      "epoch": 0.9043904208998549,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925553063425755,
      "loss": 5.2772,
      "mean_token_accuracy": 0.186713108420372,
      "num_tokens": 20694097.0,
      "step": 9970
    },
    {
      "entropy": 5.5363075733184814,
      "epoch": 0.904843976777939,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925470497001844,
      "loss": 5.2393,
      "mean_token_accuracy": 0.19200791716575621,
      "num_tokens": 20704640.0,
      "step": 9975
    },
    {
      "entropy": 5.471827363967895,
      "epoch": 0.9052975326560232,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925387885588648,
      "loss": 5.1802,
      "mean_token_accuracy": 0.18870692253112792,
      "num_tokens": 20714419.0,
      "step": 9980
    },
    {
      "entropy": 5.544799852371216,
      "epoch": 0.9057510885341074,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004925305229187877,
      "loss": 5.2693,
      "mean_token_accuracy": 0.18884519636631011,
      "num_tokens": 20724705.0,
      "step": 9985
    },
    {
      "entropy": 5.597160959243775,
      "epoch": 0.9062046444121916,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492522252780124,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18726316094398499,
      "num_tokens": 20734753.0,
      "step": 9990
    },
    {
      "entropy": 5.465805244445801,
      "epoch": 0.9066582002902758,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004925139781430446,
      "loss": 5.1156,
      "mean_token_accuracy": 0.1963346555829048,
      "num_tokens": 20745389.0,
      "step": 9995
    },
    {
      "entropy": 5.480003547668457,
      "epoch": 0.9071117561683599,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004925056990077207,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1931140884757042,
      "num_tokens": 20755702.0,
      "step": 10000
    },
    {
      "entropy": 5.498392629623413,
      "epoch": 0.9075653120464441,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004924974153743235,
      "loss": 5.0984,
      "mean_token_accuracy": 0.19833627194166184,
      "num_tokens": 20766046.0,
      "step": 10005
    },
    {
      "entropy": 5.50446720123291,
      "epoch": 0.9080188679245284,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924891272430244,
      "loss": 5.2141,
      "mean_token_accuracy": 0.19061006605625153,
      "num_tokens": 20776654.0,
      "step": 10010
    },
    {
      "entropy": 5.5594000816345215,
      "epoch": 0.9084724238026125,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004924808346139948,
      "loss": 5.2713,
      "mean_token_accuracy": 0.19167041033506393,
      "num_tokens": 20787066.0,
      "step": 10015
    },
    {
      "entropy": 5.47709379196167,
      "epoch": 0.9089259796806967,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924725374874059,
      "loss": 5.1736,
      "mean_token_accuracy": 0.1896710604429245,
      "num_tokens": 20796822.0,
      "step": 10020
    },
    {
      "entropy": 5.495741128921509,
      "epoch": 0.9093795355587808,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924642358634297,
      "loss": 5.2297,
      "mean_token_accuracy": 0.18855800926685334,
      "num_tokens": 20807602.0,
      "step": 10025
    },
    {
      "entropy": 5.5386439800262455,
      "epoch": 0.909833091436865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924559297422376,
      "loss": 5.1864,
      "mean_token_accuracy": 0.19395318776369094,
      "num_tokens": 20819046.0,
      "step": 10030
    },
    {
      "entropy": 5.51975998878479,
      "epoch": 0.9102866473149492,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004924476191240014,
      "loss": 5.2153,
      "mean_token_accuracy": 0.19031708985567092,
      "num_tokens": 20828989.0,
      "step": 10035
    },
    {
      "entropy": 5.446819305419922,
      "epoch": 0.9107402031930334,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492439304008893,
      "loss": 5.2111,
      "mean_token_accuracy": 0.18754609823226928,
      "num_tokens": 20839366.0,
      "step": 10040
    },
    {
      "entropy": 5.519721603393554,
      "epoch": 0.9111937590711175,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004924309843970847,
      "loss": 5.2112,
      "mean_token_accuracy": 0.19083768427371978,
      "num_tokens": 20849513.0,
      "step": 10045
    },
    {
      "entropy": 5.559159517288208,
      "epoch": 0.9116473149492017,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004924226602887479,
      "loss": 5.2674,
      "mean_token_accuracy": 0.18467978686094283,
      "num_tokens": 20861406.0,
      "step": 10050
    },
    {
      "entropy": 5.47953314781189,
      "epoch": 0.912100870827286,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004924143316840552,
      "loss": 5.1645,
      "mean_token_accuracy": 0.19939935952425003,
      "num_tokens": 20871190.0,
      "step": 10055
    },
    {
      "entropy": 5.531355714797973,
      "epoch": 0.9125544267053701,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924059985831787,
      "loss": 5.3149,
      "mean_token_accuracy": 0.1820805475115776,
      "num_tokens": 20881589.0,
      "step": 10060
    },
    {
      "entropy": 5.5043120861053465,
      "epoch": 0.9130079825834543,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923976609862908,
      "loss": 5.249,
      "mean_token_accuracy": 0.17866823971271514,
      "num_tokens": 20892072.0,
      "step": 10065
    },
    {
      "entropy": 5.541967296600342,
      "epoch": 0.9134615384615384,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923893188935638,
      "loss": 5.2361,
      "mean_token_accuracy": 0.18548781424760818,
      "num_tokens": 20902616.0,
      "step": 10070
    },
    {
      "entropy": 5.578791618347168,
      "epoch": 0.9139150943396226,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004923809723051702,
      "loss": 5.231,
      "mean_token_accuracy": 0.1868339642882347,
      "num_tokens": 20913832.0,
      "step": 10075
    },
    {
      "entropy": 5.451777458190918,
      "epoch": 0.9143686502177069,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004923726212212829,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1962735816836357,
      "num_tokens": 20923322.0,
      "step": 10080
    },
    {
      "entropy": 5.44694151878357,
      "epoch": 0.914822206095791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923642656420742,
      "loss": 5.2078,
      "mean_token_accuracy": 0.19073516130447388,
      "num_tokens": 20933334.0,
      "step": 10085
    },
    {
      "entropy": 5.462447452545166,
      "epoch": 0.9152757619738752,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923559055677173,
      "loss": 5.1788,
      "mean_token_accuracy": 0.1864377275109291,
      "num_tokens": 20944171.0,
      "step": 10090
    },
    {
      "entropy": 5.524150848388672,
      "epoch": 0.9157293178519593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923475409983848,
      "loss": 5.215,
      "mean_token_accuracy": 0.18702605217695237,
      "num_tokens": 20954200.0,
      "step": 10095
    },
    {
      "entropy": 5.490021753311157,
      "epoch": 0.9161828737300436,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004923391719342497,
      "loss": 5.2339,
      "mean_token_accuracy": 0.1857544004917145,
      "num_tokens": 20964935.0,
      "step": 10100
    },
    {
      "entropy": 5.470995235443115,
      "epoch": 0.9166364296081277,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923307983754852,
      "loss": 5.2232,
      "mean_token_accuracy": 0.18978999555110931,
      "num_tokens": 20975499.0,
      "step": 10105
    },
    {
      "entropy": 5.406945657730103,
      "epoch": 0.9170899854862119,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923224203222645,
      "loss": 5.1461,
      "mean_token_accuracy": 0.19023018777370454,
      "num_tokens": 20986199.0,
      "step": 10110
    },
    {
      "entropy": 5.485232019424439,
      "epoch": 0.9175435413642961,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923140377747606,
      "loss": 5.2263,
      "mean_token_accuracy": 0.18715826719999312,
      "num_tokens": 20996313.0,
      "step": 10115
    },
    {
      "entropy": 5.515059518814087,
      "epoch": 0.9179970972423802,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923056507331471,
      "loss": 5.2129,
      "mean_token_accuracy": 0.18567863702774048,
      "num_tokens": 21006133.0,
      "step": 10120
    },
    {
      "entropy": 5.515254068374634,
      "epoch": 0.9184506531204645,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004922972591975974,
      "loss": 5.1735,
      "mean_token_accuracy": 0.18915437906980515,
      "num_tokens": 21016448.0,
      "step": 10125
    },
    {
      "entropy": 5.455716228485107,
      "epoch": 0.9189042089985486,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922888631682849,
      "loss": 5.1148,
      "mean_token_accuracy": 0.18959181159734725,
      "num_tokens": 21026458.0,
      "step": 10130
    },
    {
      "entropy": 5.464995288848877,
      "epoch": 0.9193577648766328,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004922804626453834,
      "loss": 5.1567,
      "mean_token_accuracy": 0.19512831717729567,
      "num_tokens": 21037111.0,
      "step": 10135
    },
    {
      "entropy": 5.4613899230957035,
      "epoch": 0.9198113207547169,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004922720576290665,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19339829832315444,
      "num_tokens": 21048041.0,
      "step": 10140
    },
    {
      "entropy": 5.438198232650757,
      "epoch": 0.9202648766328012,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004922636481195082,
      "loss": 5.0888,
      "mean_token_accuracy": 0.20385636240243912,
      "num_tokens": 21058099.0,
      "step": 10145
    },
    {
      "entropy": 5.448825025558472,
      "epoch": 0.9207184325108854,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004922552341168823,
      "loss": 5.2224,
      "mean_token_accuracy": 0.18477773368358613,
      "num_tokens": 21068137.0,
      "step": 10150
    },
    {
      "entropy": 5.417740678787231,
      "epoch": 0.9211719883889695,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922468156213627,
      "loss": 5.115,
      "mean_token_accuracy": 0.19422985017299652,
      "num_tokens": 21077556.0,
      "step": 10155
    },
    {
      "entropy": 5.478008985519409,
      "epoch": 0.9216255442670537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004922383926331236,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19280444234609603,
      "num_tokens": 21087524.0,
      "step": 10160
    },
    {
      "entropy": 5.425194454193115,
      "epoch": 0.9220791001451378,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922299651523392,
      "loss": 5.2124,
      "mean_token_accuracy": 0.18724745512008667,
      "num_tokens": 21098446.0,
      "step": 10165
    },
    {
      "entropy": 5.459459257125855,
      "epoch": 0.9225326560232221,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922215331791837,
      "loss": 5.2302,
      "mean_token_accuracy": 0.18599504977464676,
      "num_tokens": 21109291.0,
      "step": 10170
    },
    {
      "entropy": 5.501635313034058,
      "epoch": 0.9229862119013063,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004922130967138316,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19838399291038514,
      "num_tokens": 21120043.0,
      "step": 10175
    },
    {
      "entropy": 5.518260431289673,
      "epoch": 0.9234397677793904,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004922046557564574,
      "loss": 5.2318,
      "mean_token_accuracy": 0.18464371412992478,
      "num_tokens": 21130339.0,
      "step": 10180
    },
    {
      "entropy": 5.4879247665405275,
      "epoch": 0.9238933236574746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921962103072355,
      "loss": 5.2982,
      "mean_token_accuracy": 0.18253571540117264,
      "num_tokens": 21142307.0,
      "step": 10185
    },
    {
      "entropy": 5.516984462738037,
      "epoch": 0.9243468795355588,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921877603663406,
      "loss": 5.1916,
      "mean_token_accuracy": 0.19168877601623535,
      "num_tokens": 21151843.0,
      "step": 10190
    },
    {
      "entropy": 5.52185640335083,
      "epoch": 0.924800435413643,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921793059339474,
      "loss": 5.1917,
      "mean_token_accuracy": 0.1947432205080986,
      "num_tokens": 21162574.0,
      "step": 10195
    },
    {
      "entropy": 5.514074325561523,
      "epoch": 0.9252539912917271,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921708470102309,
      "loss": 5.1502,
      "mean_token_accuracy": 0.19119355827569962,
      "num_tokens": 21172637.0,
      "step": 10200
    },
    {
      "entropy": 5.465252828598023,
      "epoch": 0.9257075471698113,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492162383595366,
      "loss": 5.1695,
      "mean_token_accuracy": 0.19286154508590697,
      "num_tokens": 21182272.0,
      "step": 10205
    },
    {
      "entropy": 5.5177301406860355,
      "epoch": 0.9261611030478955,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921539156895275,
      "loss": 5.287,
      "mean_token_accuracy": 0.18583095222711563,
      "num_tokens": 21193712.0,
      "step": 10210
    },
    {
      "entropy": 5.565571260452271,
      "epoch": 0.9266146589259797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004921454432928908,
      "loss": 5.3146,
      "mean_token_accuracy": 0.18066612184047698,
      "num_tokens": 21205104.0,
      "step": 10215
    },
    {
      "entropy": 5.5436365604400635,
      "epoch": 0.9270682148040639,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000492136966405631,
      "loss": 5.1809,
      "mean_token_accuracy": 0.1893284574151039,
      "num_tokens": 21215495.0,
      "step": 10220
    },
    {
      "entropy": 5.496471214294433,
      "epoch": 0.927521770682148,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004921284850279233,
      "loss": 5.2052,
      "mean_token_accuracy": 0.18886639475822448,
      "num_tokens": 21226047.0,
      "step": 10225
    },
    {
      "entropy": 5.502618789672852,
      "epoch": 0.9279753265602322,
      "grad_norm": 1.0,
      "learning_rate": 0.0004921199991599433,
      "loss": 5.2528,
      "mean_token_accuracy": 0.18845335245132447,
      "num_tokens": 21235784.0,
      "step": 10230
    },
    {
      "entropy": 5.557482576370239,
      "epoch": 0.9284288824383164,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921115088018665,
      "loss": 5.2097,
      "mean_token_accuracy": 0.19081354588270188,
      "num_tokens": 21245493.0,
      "step": 10235
    },
    {
      "entropy": 5.612878608703613,
      "epoch": 0.9288824383164006,
      "grad_norm": 0.96875,
      "learning_rate": 0.000492103013953868,
      "loss": 5.2906,
      "mean_token_accuracy": 0.1868404284119606,
      "num_tokens": 21255541.0,
      "step": 10240
    },
    {
      "entropy": 5.442387437820434,
      "epoch": 0.9293359941944848,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004920945146161242,
      "loss": 5.1463,
      "mean_token_accuracy": 0.19286169111728668,
      "num_tokens": 21265427.0,
      "step": 10245
    },
    {
      "entropy": 5.464549398422241,
      "epoch": 0.9297895500725689,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920860107888105,
      "loss": 5.1543,
      "mean_token_accuracy": 0.1940089598298073,
      "num_tokens": 21276597.0,
      "step": 10250
    },
    {
      "entropy": 5.460808277130127,
      "epoch": 0.9302431059506531,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004920775024721026,
      "loss": 5.1745,
      "mean_token_accuracy": 0.18666350245475768,
      "num_tokens": 21286416.0,
      "step": 10255
    },
    {
      "entropy": 5.482029771804809,
      "epoch": 0.9306966618287373,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920689896661767,
      "loss": 5.1954,
      "mean_token_accuracy": 0.19428568035364152,
      "num_tokens": 21296064.0,
      "step": 10260
    },
    {
      "entropy": 5.459726476669312,
      "epoch": 0.9311502177068215,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920604723712088,
      "loss": 5.2188,
      "mean_token_accuracy": 0.19196224957704544,
      "num_tokens": 21306836.0,
      "step": 10265
    },
    {
      "entropy": 5.48441743850708,
      "epoch": 0.9316037735849056,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920519505873751,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19369848519563676,
      "num_tokens": 21316936.0,
      "step": 10270
    },
    {
      "entropy": 5.464742708206177,
      "epoch": 0.9320573294629898,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004920434243148515,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20192406624555587,
      "num_tokens": 21326892.0,
      "step": 10275
    },
    {
      "entropy": 5.472868013381958,
      "epoch": 0.9325108853410741,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920348935538147,
      "loss": 5.1444,
      "mean_token_accuracy": 0.1884632021188736,
      "num_tokens": 21337270.0,
      "step": 10280
    },
    {
      "entropy": 5.532982063293457,
      "epoch": 0.9329644412191582,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000492026358304441,
      "loss": 5.2328,
      "mean_token_accuracy": 0.19010204523801805,
      "num_tokens": 21347968.0,
      "step": 10285
    },
    {
      "entropy": 5.47656831741333,
      "epoch": 0.9334179970972424,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000492017818566907,
      "loss": 5.1746,
      "mean_token_accuracy": 0.19259967803955078,
      "num_tokens": 21358312.0,
      "step": 10290
    },
    {
      "entropy": 5.470859527587891,
      "epoch": 0.9338715529753265,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920092743413892,
      "loss": 5.142,
      "mean_token_accuracy": 0.19385140240192414,
      "num_tokens": 21368794.0,
      "step": 10295
    },
    {
      "entropy": 5.508439350128174,
      "epoch": 0.9343251088534107,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920007256280642,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1947756752371788,
      "num_tokens": 21380080.0,
      "step": 10300
    },
    {
      "entropy": 5.497937631607056,
      "epoch": 0.934778664731495,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491992172427109,
      "loss": 5.1906,
      "mean_token_accuracy": 0.19269357919692992,
      "num_tokens": 21389961.0,
      "step": 10305
    },
    {
      "entropy": 5.450817489624024,
      "epoch": 0.9352322206095791,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004919836147387002,
      "loss": 5.2161,
      "mean_token_accuracy": 0.1923079416155815,
      "num_tokens": 21399869.0,
      "step": 10310
    },
    {
      "entropy": 5.519410276412964,
      "epoch": 0.9356857764876633,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004919750525630151,
      "loss": 5.2044,
      "mean_token_accuracy": 0.1905330315232277,
      "num_tokens": 21410343.0,
      "step": 10315
    },
    {
      "entropy": 5.512972068786621,
      "epoch": 0.9361393323657474,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919664859002306,
      "loss": 5.1847,
      "mean_token_accuracy": 0.18935477286577224,
      "num_tokens": 21420915.0,
      "step": 10320
    },
    {
      "entropy": 5.480722951889038,
      "epoch": 0.9365928882438317,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004919579147505239,
      "loss": 5.138,
      "mean_token_accuracy": 0.19782340973615647,
      "num_tokens": 21430879.0,
      "step": 10325
    },
    {
      "entropy": 5.5278362274169925,
      "epoch": 0.9370464441219158,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004919493391140721,
      "loss": 5.2364,
      "mean_token_accuracy": 0.186504028737545,
      "num_tokens": 21440843.0,
      "step": 10330
    },
    {
      "entropy": 5.567490196228027,
      "epoch": 0.9375,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919407589910529,
      "loss": 5.4008,
      "mean_token_accuracy": 0.18015343099832534,
      "num_tokens": 21452132.0,
      "step": 10335
    },
    {
      "entropy": 5.494717454910278,
      "epoch": 0.9379535558780842,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004919321743816434,
      "loss": 5.185,
      "mean_token_accuracy": 0.1896796151995659,
      "num_tokens": 21461861.0,
      "step": 10340
    },
    {
      "entropy": 5.5109964370727536,
      "epoch": 0.9384071117561683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004919235852860213,
      "loss": 5.1841,
      "mean_token_accuracy": 0.1865747943520546,
      "num_tokens": 21472244.0,
      "step": 10345
    },
    {
      "entropy": 5.528714323043824,
      "epoch": 0.9388606676342526,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004919149917043641,
      "loss": 5.2286,
      "mean_token_accuracy": 0.18693966567516326,
      "num_tokens": 21482394.0,
      "step": 10350
    },
    {
      "entropy": 5.504456090927124,
      "epoch": 0.9393142235123367,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004919063936368496,
      "loss": 5.2679,
      "mean_token_accuracy": 0.18453455120325088,
      "num_tokens": 21493334.0,
      "step": 10355
    },
    {
      "entropy": 5.5536915302276615,
      "epoch": 0.9397677793904209,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918977910836557,
      "loss": 5.2333,
      "mean_token_accuracy": 0.18519501984119416,
      "num_tokens": 21504443.0,
      "step": 10360
    },
    {
      "entropy": 5.539934301376343,
      "epoch": 0.940221335268505,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004918891840449601,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18833032846450806,
      "num_tokens": 21513908.0,
      "step": 10365
    },
    {
      "entropy": 5.542459392547608,
      "epoch": 0.9406748911465893,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000491880572520941,
      "loss": 5.2724,
      "mean_token_accuracy": 0.18035230338573455,
      "num_tokens": 21524349.0,
      "step": 10370
    },
    {
      "entropy": 5.6526354312896725,
      "epoch": 0.9411284470246735,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004918719565117763,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1903665080666542,
      "num_tokens": 21535003.0,
      "step": 10375
    },
    {
      "entropy": 5.616309881210327,
      "epoch": 0.9415820029027576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004918633360176444,
      "loss": 5.3643,
      "mean_token_accuracy": 0.1793342188000679,
      "num_tokens": 21545045.0,
      "step": 10380
    },
    {
      "entropy": 5.4643491268157955,
      "epoch": 0.9420355587808418,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918547110387234,
      "loss": 5.182,
      "mean_token_accuracy": 0.19717971086502076,
      "num_tokens": 21554830.0,
      "step": 10385
    },
    {
      "entropy": 5.5400168895721436,
      "epoch": 0.9424891146589259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004918460815751916,
      "loss": 5.1437,
      "mean_token_accuracy": 0.19479870349168776,
      "num_tokens": 21564693.0,
      "step": 10390
    },
    {
      "entropy": 5.491382455825805,
      "epoch": 0.9429426705370102,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004918374476272277,
      "loss": 5.167,
      "mean_token_accuracy": 0.187149715423584,
      "num_tokens": 21575475.0,
      "step": 10395
    },
    {
      "entropy": 5.4525188446044925,
      "epoch": 0.9433962264150944,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004918288091950101,
      "loss": 5.1887,
      "mean_token_accuracy": 0.18847269862890242,
      "num_tokens": 21586944.0,
      "step": 10400
    },
    {
      "entropy": 5.544979858398437,
      "epoch": 0.9438497822931785,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004918201662787174,
      "loss": 5.2002,
      "mean_token_accuracy": 0.1872720867395401,
      "num_tokens": 21598170.0,
      "step": 10405
    },
    {
      "entropy": 5.511740016937256,
      "epoch": 0.9443033381712627,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004918115188785285,
      "loss": 5.2188,
      "mean_token_accuracy": 0.1880117490887642,
      "num_tokens": 21609170.0,
      "step": 10410
    },
    {
      "entropy": 5.515832614898682,
      "epoch": 0.9447568940493469,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004918028669946221,
      "loss": 5.2956,
      "mean_token_accuracy": 0.18484372794628143,
      "num_tokens": 21620171.0,
      "step": 10415
    },
    {
      "entropy": 5.486926412582397,
      "epoch": 0.9452104499274311,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917942106271771,
      "loss": 5.2104,
      "mean_token_accuracy": 0.19340168982744216,
      "num_tokens": 21631072.0,
      "step": 10420
    },
    {
      "entropy": 5.4726005554199215,
      "epoch": 0.9456640058055152,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004917855497763727,
      "loss": 5.0846,
      "mean_token_accuracy": 0.1976454183459282,
      "num_tokens": 21641543.0,
      "step": 10425
    },
    {
      "entropy": 5.4695212841033936,
      "epoch": 0.9461175616835994,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004917768844423878,
      "loss": 5.2123,
      "mean_token_accuracy": 0.19243092685937882,
      "num_tokens": 21651342.0,
      "step": 10430
    },
    {
      "entropy": 5.524942016601562,
      "epoch": 0.9465711175616836,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004917682146254017,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19580309987068176,
      "num_tokens": 21661043.0,
      "step": 10435
    },
    {
      "entropy": 5.4407624244689945,
      "epoch": 0.9470246734397678,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004917595403255938,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1993831992149353,
      "num_tokens": 21671197.0,
      "step": 10440
    },
    {
      "entropy": 5.421433401107788,
      "epoch": 0.947478229317852,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917508615431431,
      "loss": 5.1253,
      "mean_token_accuracy": 0.19686731249094008,
      "num_tokens": 21682066.0,
      "step": 10445
    },
    {
      "entropy": 5.554642391204834,
      "epoch": 0.9479317851959361,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004917421782782296,
      "loss": 5.1912,
      "mean_token_accuracy": 0.19019131511449813,
      "num_tokens": 21692117.0,
      "step": 10450
    },
    {
      "entropy": 5.5294716358184814,
      "epoch": 0.9483853410740203,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917334905310325,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18246906250715256,
      "num_tokens": 21703544.0,
      "step": 10455
    },
    {
      "entropy": 5.445235729217529,
      "epoch": 0.9488388969521045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004917247983017317,
      "loss": 5.1233,
      "mean_token_accuracy": 0.19221487492322922,
      "num_tokens": 21712850.0,
      "step": 10460
    },
    {
      "entropy": 5.474435758590698,
      "epoch": 0.9492924528301887,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004917161015905067,
      "loss": 5.2636,
      "mean_token_accuracy": 0.18826578557491302,
      "num_tokens": 21723015.0,
      "step": 10465
    },
    {
      "entropy": 5.513061094284057,
      "epoch": 0.9497460087082729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917074003975376,
      "loss": 5.2182,
      "mean_token_accuracy": 0.1881333664059639,
      "num_tokens": 21732453.0,
      "step": 10470
    },
    {
      "entropy": 5.551966810226441,
      "epoch": 0.950199564586357,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004916986947230042,
      "loss": 5.2908,
      "mean_token_accuracy": 0.18900395184755325,
      "num_tokens": 21743251.0,
      "step": 10475
    },
    {
      "entropy": 5.515570878982544,
      "epoch": 0.9506531204644412,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916899845670866,
      "loss": 5.2076,
      "mean_token_accuracy": 0.18819339275360109,
      "num_tokens": 21752891.0,
      "step": 10480
    },
    {
      "entropy": 5.543520164489746,
      "epoch": 0.9511066763425254,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004916812699299649,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19229596108198166,
      "num_tokens": 21763230.0,
      "step": 10485
    },
    {
      "entropy": 5.49464168548584,
      "epoch": 0.9515602322206096,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916725508118193,
      "loss": 5.1474,
      "mean_token_accuracy": 0.1914653703570366,
      "num_tokens": 21773663.0,
      "step": 10490
    },
    {
      "entropy": 5.490184593200683,
      "epoch": 0.9520137880986937,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004916638272128301,
      "loss": 5.216,
      "mean_token_accuracy": 0.19228923469781875,
      "num_tokens": 21784265.0,
      "step": 10495
    },
    {
      "entropy": 5.456570482254028,
      "epoch": 0.9524673439767779,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916550991331777,
      "loss": 5.1794,
      "mean_token_accuracy": 0.19147132486104965,
      "num_tokens": 21795056.0,
      "step": 10500
    },
    {
      "entropy": 5.484953880310059,
      "epoch": 0.9529208998548622,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004916463665730427,
      "loss": 5.1838,
      "mean_token_accuracy": 0.1929536059498787,
      "num_tokens": 21806204.0,
      "step": 10505
    },
    {
      "entropy": 5.4476096630096436,
      "epoch": 0.9533744557329463,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916376295326057,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19114798158407212,
      "num_tokens": 21816993.0,
      "step": 10510
    },
    {
      "entropy": 5.462368154525757,
      "epoch": 0.9538280116110305,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004916288880120472,
      "loss": 5.0671,
      "mean_token_accuracy": 0.19932386875152588,
      "num_tokens": 21827457.0,
      "step": 10515
    },
    {
      "entropy": 5.480139207839966,
      "epoch": 0.9542815674891146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916201420115482,
      "loss": 5.1847,
      "mean_token_accuracy": 0.18778882622718812,
      "num_tokens": 21837555.0,
      "step": 10520
    },
    {
      "entropy": 5.487935304641724,
      "epoch": 0.9547351233671988,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916113915312895,
      "loss": 5.2163,
      "mean_token_accuracy": 0.19396420568227768,
      "num_tokens": 21847074.0,
      "step": 10525
    },
    {
      "entropy": 5.557778024673462,
      "epoch": 0.9551886792452831,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916026365714519,
      "loss": 5.2066,
      "mean_token_accuracy": 0.18701234608888626,
      "num_tokens": 21858333.0,
      "step": 10530
    },
    {
      "entropy": 5.46713604927063,
      "epoch": 0.9556422351233672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915938771322165,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19187063127756118,
      "num_tokens": 21869132.0,
      "step": 10535
    },
    {
      "entropy": 5.460751390457153,
      "epoch": 0.9560957910014514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915851132137646,
      "loss": 5.2371,
      "mean_token_accuracy": 0.189180788397789,
      "num_tokens": 21879153.0,
      "step": 10540
    },
    {
      "entropy": 5.506651306152344,
      "epoch": 0.9565493468795355,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004915763448162774,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19192200899124146,
      "num_tokens": 21889063.0,
      "step": 10545
    },
    {
      "entropy": 5.5077903270721436,
      "epoch": 0.9570029027576198,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004915675719399362,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1865275591611862,
      "num_tokens": 21899005.0,
      "step": 10550
    },
    {
      "entropy": 5.494836759567261,
      "epoch": 0.9574564586357039,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915587945849225,
      "loss": 5.1099,
      "mean_token_accuracy": 0.2017269641160965,
      "num_tokens": 21908391.0,
      "step": 10555
    },
    {
      "entropy": 5.501866388320923,
      "epoch": 0.9579100145137881,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004915500127514176,
      "loss": 5.1823,
      "mean_token_accuracy": 0.1938887879252434,
      "num_tokens": 21917655.0,
      "step": 10560
    },
    {
      "entropy": 5.5588977336883545,
      "epoch": 0.9583635703918723,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004915412264396034,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19208064824342727,
      "num_tokens": 21928038.0,
      "step": 10565
    },
    {
      "entropy": 5.502170753479004,
      "epoch": 0.9588171262699564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915324356496613,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19868719577789307,
      "num_tokens": 21938106.0,
      "step": 10570
    },
    {
      "entropy": 5.459872579574585,
      "epoch": 0.9592706821480407,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915236403817733,
      "loss": 5.1307,
      "mean_token_accuracy": 0.19554639905691146,
      "num_tokens": 21947484.0,
      "step": 10575
    },
    {
      "entropy": 5.483737945556641,
      "epoch": 0.9597242380261248,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004915148406361213,
      "loss": 5.2229,
      "mean_token_accuracy": 0.18886553943157197,
      "num_tokens": 21957139.0,
      "step": 10580
    },
    {
      "entropy": 5.475294065475464,
      "epoch": 0.960177793904209,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004915060364128872,
      "loss": 5.219,
      "mean_token_accuracy": 0.18840531557798385,
      "num_tokens": 21967030.0,
      "step": 10585
    },
    {
      "entropy": 5.461275815963745,
      "epoch": 0.9606313497822931,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914972277122532,
      "loss": 5.1727,
      "mean_token_accuracy": 0.19003562033176422,
      "num_tokens": 21976418.0,
      "step": 10590
    },
    {
      "entropy": 5.482760334014893,
      "epoch": 0.9610849056603774,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914884145344012,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19301039576530457,
      "num_tokens": 21987284.0,
      "step": 10595
    },
    {
      "entropy": 5.4282470703125,
      "epoch": 0.9615384615384616,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914795968795137,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19979993253946304,
      "num_tokens": 21997106.0,
      "step": 10600
    },
    {
      "entropy": 5.441180753707886,
      "epoch": 0.9619920174165457,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914707747477728,
      "loss": 5.0567,
      "mean_token_accuracy": 0.19796952307224275,
      "num_tokens": 22007589.0,
      "step": 10605
    },
    {
      "entropy": 5.470518398284912,
      "epoch": 0.9624455732946299,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004914619481393613,
      "loss": 5.2181,
      "mean_token_accuracy": 0.1860770180821419,
      "num_tokens": 22017669.0,
      "step": 10610
    },
    {
      "entropy": 5.617669343948364,
      "epoch": 0.962899129172714,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914531170544615,
      "loss": 5.3642,
      "mean_token_accuracy": 0.18315273523330688,
      "num_tokens": 22029330.0,
      "step": 10615
    },
    {
      "entropy": 5.45698356628418,
      "epoch": 0.9633526850507983,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004914442814932561,
      "loss": 5.1359,
      "mean_token_accuracy": 0.19763439446687697,
      "num_tokens": 22039877.0,
      "step": 10620
    },
    {
      "entropy": 5.481981801986694,
      "epoch": 0.9638062409288825,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004914354414559278,
      "loss": 5.1859,
      "mean_token_accuracy": 0.19522749185562133,
      "num_tokens": 22049579.0,
      "step": 10625
    },
    {
      "entropy": 5.393762636184692,
      "epoch": 0.9642597968069666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914265969426594,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20582508593797683,
      "num_tokens": 22059608.0,
      "step": 10630
    },
    {
      "entropy": 5.472809171676635,
      "epoch": 0.9647133526850508,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914177479536338,
      "loss": 5.2292,
      "mean_token_accuracy": 0.19040220081806183,
      "num_tokens": 22070036.0,
      "step": 10635
    },
    {
      "entropy": 5.493038988113403,
      "epoch": 0.965166908563135,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000491408894489034,
      "loss": 5.2222,
      "mean_token_accuracy": 0.18583654314279557,
      "num_tokens": 22080679.0,
      "step": 10640
    },
    {
      "entropy": 5.500444984436035,
      "epoch": 0.9656204644412192,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004914000365490432,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20014223903417588,
      "num_tokens": 22092192.0,
      "step": 10645
    },
    {
      "entropy": 5.490348196029663,
      "epoch": 0.9660740203193033,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913911741338445,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1852397084236145,
      "num_tokens": 22102722.0,
      "step": 10650
    },
    {
      "entropy": 5.406291580200195,
      "epoch": 0.9665275761973875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004913823072436211,
      "loss": 5.1155,
      "mean_token_accuracy": 0.19892334789037705,
      "num_tokens": 22113521.0,
      "step": 10655
    },
    {
      "entropy": 5.501695013046264,
      "epoch": 0.9669811320754716,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004913734358785565,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19285878390073777,
      "num_tokens": 22124891.0,
      "step": 10660
    },
    {
      "entropy": 5.486810874938965,
      "epoch": 0.9674346879535559,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004913645600388341,
      "loss": 5.2107,
      "mean_token_accuracy": 0.18844253867864608,
      "num_tokens": 22135972.0,
      "step": 10665
    },
    {
      "entropy": 5.540691900253296,
      "epoch": 0.9678882438316401,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913556797246374,
      "loss": 5.2649,
      "mean_token_accuracy": 0.18179991245269775,
      "num_tokens": 22146623.0,
      "step": 10670
    },
    {
      "entropy": 5.5169305324554445,
      "epoch": 0.9683417997097242,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913467949361503,
      "loss": 5.1767,
      "mean_token_accuracy": 0.18900265991687776,
      "num_tokens": 22157619.0,
      "step": 10675
    },
    {
      "entropy": 5.546804666519165,
      "epoch": 0.9687953555878084,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913379056735561,
      "loss": 5.1826,
      "mean_token_accuracy": 0.18539999574422836,
      "num_tokens": 22168655.0,
      "step": 10680
    },
    {
      "entropy": 5.47217059135437,
      "epoch": 0.9692489114658926,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000491329011937039,
      "loss": 5.2184,
      "mean_token_accuracy": 0.18761395961046218,
      "num_tokens": 22178500.0,
      "step": 10685
    },
    {
      "entropy": 5.515644979476929,
      "epoch": 0.9697024673439768,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004913201137267828,
      "loss": 5.224,
      "mean_token_accuracy": 0.18653703182935716,
      "num_tokens": 22188873.0,
      "step": 10690
    },
    {
      "entropy": 5.454254674911499,
      "epoch": 0.970156023222061,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913112110429715,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19080322682857515,
      "num_tokens": 22199743.0,
      "step": 10695
    },
    {
      "entropy": 5.465599822998047,
      "epoch": 0.9706095791001451,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004913023038857892,
      "loss": 5.0907,
      "mean_token_accuracy": 0.2004644140601158,
      "num_tokens": 22210301.0,
      "step": 10700
    },
    {
      "entropy": 5.4484978199005125,
      "epoch": 0.9710631349782293,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004912933922554203,
      "loss": 5.1006,
      "mean_token_accuracy": 0.1963333547115326,
      "num_tokens": 22220771.0,
      "step": 10705
    },
    {
      "entropy": 5.459470510482788,
      "epoch": 0.9715166908563135,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004912844761520487,
      "loss": 5.2071,
      "mean_token_accuracy": 0.1855422705411911,
      "num_tokens": 22231100.0,
      "step": 10710
    },
    {
      "entropy": 5.4956214427948,
      "epoch": 0.9719702467343977,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912755555758592,
      "loss": 5.1826,
      "mean_token_accuracy": 0.19422222226858138,
      "num_tokens": 22240997.0,
      "step": 10715
    },
    {
      "entropy": 5.514835929870605,
      "epoch": 0.9724238026124818,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491266630527036,
      "loss": 5.2124,
      "mean_token_accuracy": 0.189374640583992,
      "num_tokens": 22251795.0,
      "step": 10720
    },
    {
      "entropy": 5.549368095397949,
      "epoch": 0.972877358490566,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004912577010057638,
      "loss": 5.257,
      "mean_token_accuracy": 0.18586982041597366,
      "num_tokens": 22262465.0,
      "step": 10725
    },
    {
      "entropy": 5.599781894683838,
      "epoch": 0.9733309143686503,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004912487670122273,
      "loss": 5.3566,
      "mean_token_accuracy": 0.18138834536075593,
      "num_tokens": 22272399.0,
      "step": 10730
    },
    {
      "entropy": 5.5480303287506105,
      "epoch": 0.9737844702467344,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912398285466111,
      "loss": 5.2657,
      "mean_token_accuracy": 0.18869337886571885,
      "num_tokens": 22282545.0,
      "step": 10735
    },
    {
      "entropy": 5.521822595596314,
      "epoch": 0.9742380261248186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912308856091002,
      "loss": 5.2233,
      "mean_token_accuracy": 0.19182628244161606,
      "num_tokens": 22293154.0,
      "step": 10740
    },
    {
      "entropy": 5.5208924293518065,
      "epoch": 0.9746915820029027,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004912219381998795,
      "loss": 5.143,
      "mean_token_accuracy": 0.19636851102113723,
      "num_tokens": 22302309.0,
      "step": 10745
    },
    {
      "entropy": 5.558265972137451,
      "epoch": 0.9751451378809869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004912129863191339,
      "loss": 5.2559,
      "mean_token_accuracy": 0.178457310795784,
      "num_tokens": 22312316.0,
      "step": 10750
    },
    {
      "entropy": 5.517702722549439,
      "epoch": 0.9755986937590712,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912040299670487,
      "loss": 5.2337,
      "mean_token_accuracy": 0.1895495042204857,
      "num_tokens": 22322387.0,
      "step": 10755
    },
    {
      "entropy": 5.497195672988892,
      "epoch": 0.9760522496371553,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911950691438091,
      "loss": 5.22,
      "mean_token_accuracy": 0.19919940531253816,
      "num_tokens": 22332759.0,
      "step": 10760
    },
    {
      "entropy": 5.516394901275635,
      "epoch": 0.9765058055152395,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004911861038496004,
      "loss": 5.2637,
      "mean_token_accuracy": 0.18964977264404298,
      "num_tokens": 22344056.0,
      "step": 10765
    },
    {
      "entropy": 5.451268243789673,
      "epoch": 0.9769593613933236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004911771340846079,
      "loss": 5.2089,
      "mean_token_accuracy": 0.1823955789208412,
      "num_tokens": 22353877.0,
      "step": 10770
    },
    {
      "entropy": 5.557831621170044,
      "epoch": 0.9774129172714079,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911681598490172,
      "loss": 5.2351,
      "mean_token_accuracy": 0.18687503039836884,
      "num_tokens": 22363549.0,
      "step": 10775
    },
    {
      "entropy": 5.593202066421509,
      "epoch": 0.977866473149492,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911591811430139,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18995759636163712,
      "num_tokens": 22374040.0,
      "step": 10780
    },
    {
      "entropy": 5.54970760345459,
      "epoch": 0.9783200290275762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911501979667836,
      "loss": 5.3196,
      "mean_token_accuracy": 0.1861735463142395,
      "num_tokens": 22384819.0,
      "step": 10785
    },
    {
      "entropy": 5.487377309799195,
      "epoch": 0.9787735849056604,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004911412103205122,
      "loss": 5.1873,
      "mean_token_accuracy": 0.1898432657122612,
      "num_tokens": 22395130.0,
      "step": 10790
    },
    {
      "entropy": 5.503686380386353,
      "epoch": 0.9792271407837445,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911322182043855,
      "loss": 5.2081,
      "mean_token_accuracy": 0.18677622675895691,
      "num_tokens": 22405093.0,
      "step": 10795
    },
    {
      "entropy": 5.51582441329956,
      "epoch": 0.9796806966618288,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004911232216185893,
      "loss": 5.2269,
      "mean_token_accuracy": 0.18393291234970094,
      "num_tokens": 22415008.0,
      "step": 10800
    },
    {
      "entropy": 5.5575262069702145,
      "epoch": 0.9801342525399129,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004911142205633099,
      "loss": 5.2651,
      "mean_token_accuracy": 0.18396617025136947,
      "num_tokens": 22426486.0,
      "step": 10805
    },
    {
      "entropy": 5.4911479473114015,
      "epoch": 0.9805878084179971,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004911052150387334,
      "loss": 5.1708,
      "mean_token_accuracy": 0.19237590581178665,
      "num_tokens": 22436844.0,
      "step": 10810
    },
    {
      "entropy": 5.568431949615478,
      "epoch": 0.9810413642960812,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004910962050450458,
      "loss": 5.3003,
      "mean_token_accuracy": 0.1872747138142586,
      "num_tokens": 22446630.0,
      "step": 10815
    },
    {
      "entropy": 5.56194052696228,
      "epoch": 0.9814949201741655,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910871905824338,
      "loss": 5.1379,
      "mean_token_accuracy": 0.1935226395726204,
      "num_tokens": 22456406.0,
      "step": 10820
    },
    {
      "entropy": 5.460543298721314,
      "epoch": 0.9819484760522497,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004910781716510837,
      "loss": 5.1251,
      "mean_token_accuracy": 0.19543748646974562,
      "num_tokens": 22466210.0,
      "step": 10825
    },
    {
      "entropy": 5.49596643447876,
      "epoch": 0.9824020319303338,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910691482511817,
      "loss": 5.2273,
      "mean_token_accuracy": 0.1921169638633728,
      "num_tokens": 22476978.0,
      "step": 10830
    },
    {
      "entropy": 5.588016319274902,
      "epoch": 0.982855587808418,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004910601203829148,
      "loss": 5.1929,
      "mean_token_accuracy": 0.18680335581302643,
      "num_tokens": 22488613.0,
      "step": 10835
    },
    {
      "entropy": 5.482051849365234,
      "epoch": 0.9833091436865021,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004910510880464696,
      "loss": 5.0861,
      "mean_token_accuracy": 0.2025364950299263,
      "num_tokens": 22498312.0,
      "step": 10840
    },
    {
      "entropy": 5.485942792892456,
      "epoch": 0.9837626995645864,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004910420512420327,
      "loss": 5.2127,
      "mean_token_accuracy": 0.1877727434039116,
      "num_tokens": 22508131.0,
      "step": 10845
    },
    {
      "entropy": 5.513615369796753,
      "epoch": 0.9842162554426706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910330099697913,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1947124034166336,
      "num_tokens": 22517842.0,
      "step": 10850
    },
    {
      "entropy": 5.540058898925781,
      "epoch": 0.9846698113207547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910239642299321,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19113553911447526,
      "num_tokens": 22528239.0,
      "step": 10855
    },
    {
      "entropy": 5.472961711883545,
      "epoch": 0.9851233671988389,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004910149140226422,
      "loss": 5.1559,
      "mean_token_accuracy": 0.19106702506542206,
      "num_tokens": 22539734.0,
      "step": 10860
    },
    {
      "entropy": 5.481329441070557,
      "epoch": 0.9855769230769231,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004910058593481089,
      "loss": 5.1549,
      "mean_token_accuracy": 0.19573762267827988,
      "num_tokens": 22549796.0,
      "step": 10865
    },
    {
      "entropy": 5.457813835144043,
      "epoch": 0.9860304789550073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909968002065195,
      "loss": 5.137,
      "mean_token_accuracy": 0.19690064191818238,
      "num_tokens": 22559934.0,
      "step": 10870
    },
    {
      "entropy": 5.597090387344361,
      "epoch": 0.9864840348330914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000490987736598061,
      "loss": 5.26,
      "mean_token_accuracy": 0.1757528841495514,
      "num_tokens": 22569864.0,
      "step": 10875
    },
    {
      "entropy": 5.525681018829346,
      "epoch": 0.9869375907111756,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909786685229211,
      "loss": 5.1874,
      "mean_token_accuracy": 0.19665819704532622,
      "num_tokens": 22580687.0,
      "step": 10880
    },
    {
      "entropy": 5.555548810958863,
      "epoch": 0.9873911465892597,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909695959812875,
      "loss": 5.2377,
      "mean_token_accuracy": 0.18621382713317872,
      "num_tokens": 22590507.0,
      "step": 10885
    },
    {
      "entropy": 5.540699434280396,
      "epoch": 0.987844702467344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004909605189733474,
      "loss": 5.2675,
      "mean_token_accuracy": 0.19158151596784592,
      "num_tokens": 22601901.0,
      "step": 10890
    },
    {
      "entropy": 5.478123521804809,
      "epoch": 0.9882982583454282,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909514374992889,
      "loss": 5.2022,
      "mean_token_accuracy": 0.18934640884399415,
      "num_tokens": 22613003.0,
      "step": 10895
    },
    {
      "entropy": 5.444561386108399,
      "epoch": 0.9887518142235123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909423515592996,
      "loss": 5.0531,
      "mean_token_accuracy": 0.1984160289168358,
      "num_tokens": 22623401.0,
      "step": 10900
    },
    {
      "entropy": 5.533397102355957,
      "epoch": 0.9892053701015965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004909332611535675,
      "loss": 5.2067,
      "mean_token_accuracy": 0.1915249705314636,
      "num_tokens": 22633147.0,
      "step": 10905
    },
    {
      "entropy": 5.486864948272705,
      "epoch": 0.9896589259796807,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909241662822804,
      "loss": 5.0547,
      "mean_token_accuracy": 0.20758973360061644,
      "num_tokens": 22642990.0,
      "step": 10910
    },
    {
      "entropy": 5.489868307113648,
      "epoch": 0.9901124818577649,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909150669456265,
      "loss": 5.2756,
      "mean_token_accuracy": 0.18449527472257615,
      "num_tokens": 22652734.0,
      "step": 10915
    },
    {
      "entropy": 5.500755023956299,
      "epoch": 0.9905660377358491,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000490905963143794,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18280779719352722,
      "num_tokens": 22662968.0,
      "step": 10920
    },
    {
      "entropy": 5.448152780532837,
      "epoch": 0.9910195936139332,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908968548769713,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20330245196819305,
      "num_tokens": 22673157.0,
      "step": 10925
    },
    {
      "entropy": 5.517993688583374,
      "epoch": 0.9914731494920174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908877421453466,
      "loss": 5.2567,
      "mean_token_accuracy": 0.19120330810546876,
      "num_tokens": 22683105.0,
      "step": 10930
    },
    {
      "entropy": 5.48880033493042,
      "epoch": 0.9919267053701016,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908786249491082,
      "loss": 5.1906,
      "mean_token_accuracy": 0.19872619658708573,
      "num_tokens": 22693793.0,
      "step": 10935
    },
    {
      "entropy": 5.597190093994141,
      "epoch": 0.9923802612481858,
      "grad_norm": 1.015625,
      "learning_rate": 0.000490869503288445,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19000887721776963,
      "num_tokens": 22703884.0,
      "step": 10940
    },
    {
      "entropy": 5.532585954666137,
      "epoch": 0.99283381712627,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908603771635454,
      "loss": 5.2745,
      "mean_token_accuracy": 0.18670693933963775,
      "num_tokens": 22713699.0,
      "step": 10945
    },
    {
      "entropy": 5.5036516189575195,
      "epoch": 0.9932873730043541,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908512465745982,
      "loss": 5.1147,
      "mean_token_accuracy": 0.19385309815406798,
      "num_tokens": 22722871.0,
      "step": 10950
    },
    {
      "entropy": 5.504653596878052,
      "epoch": 0.9937409288824384,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004908421115217922,
      "loss": 5.1247,
      "mean_token_accuracy": 0.18782687783241273,
      "num_tokens": 22733314.0,
      "step": 10955
    },
    {
      "entropy": 5.431400537490845,
      "epoch": 0.9941944847605225,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004908329720053164,
      "loss": 5.1209,
      "mean_token_accuracy": 0.1992374211549759,
      "num_tokens": 22743255.0,
      "step": 10960
    },
    {
      "entropy": 5.420324754714966,
      "epoch": 0.9946480406386067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004908238280253597,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19695128947496415,
      "num_tokens": 22753857.0,
      "step": 10965
    },
    {
      "entropy": 5.496116971969604,
      "epoch": 0.9951015965166908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908146795821112,
      "loss": 5.1057,
      "mean_token_accuracy": 0.19125611186027527,
      "num_tokens": 22764326.0,
      "step": 10970
    },
    {
      "entropy": 5.502989768981934,
      "epoch": 0.995555152394775,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908055266757602,
      "loss": 5.1898,
      "mean_token_accuracy": 0.19550111889839172,
      "num_tokens": 22774958.0,
      "step": 10975
    },
    {
      "entropy": 5.408831548690796,
      "epoch": 0.9960087082728593,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004907963693064958,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1952947497367859,
      "num_tokens": 22785616.0,
      "step": 10980
    },
    {
      "entropy": 5.565986251831054,
      "epoch": 0.9964622641509434,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004907872074745075,
      "loss": 5.2869,
      "mean_token_accuracy": 0.18623248487710953,
      "num_tokens": 22795967.0,
      "step": 10985
    },
    {
      "entropy": 5.514637279510498,
      "epoch": 0.9969158200290276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907780411799849,
      "loss": 5.2722,
      "mean_token_accuracy": 0.18776020854711534,
      "num_tokens": 22806335.0,
      "step": 10990
    },
    {
      "entropy": 5.554602861404419,
      "epoch": 0.9973693759071117,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907688704231173,
      "loss": 5.2089,
      "mean_token_accuracy": 0.19053253084421157,
      "num_tokens": 22816390.0,
      "step": 10995
    },
    {
      "entropy": 5.597799396514892,
      "epoch": 0.997822931785196,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004907596952040945,
      "loss": 5.2777,
      "mean_token_accuracy": 0.18841657638549805,
      "num_tokens": 22826395.0,
      "step": 11000
    },
    {
      "entropy": 5.564309597015381,
      "epoch": 0.9982764876632801,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004907505155231063,
      "loss": 5.2449,
      "mean_token_accuracy": 0.18635112643241883,
      "num_tokens": 22836790.0,
      "step": 11005
    },
    {
      "entropy": 5.494991683959961,
      "epoch": 0.9987300435413643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907413313803424,
      "loss": 5.1386,
      "mean_token_accuracy": 0.19103351086378098,
      "num_tokens": 22846687.0,
      "step": 11010
    },
    {
      "entropy": 5.538417148590088,
      "epoch": 0.9991835994194485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907321427759928,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19118976593017578,
      "num_tokens": 22857208.0,
      "step": 11015
    },
    {
      "entropy": 5.539492750167847,
      "epoch": 0.9996371552975326,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004907229497102475,
      "loss": 5.2696,
      "mean_token_accuracy": 0.1830685555934906,
      "num_tokens": 22868556.0,
      "step": 11020
    },
    {
      "entropy": 5.554223203659058,
      "epoch": 1.0000907111756168,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004907137521832966,
      "loss": 5.2441,
      "mean_token_accuracy": 0.18953093886375427,
      "num_tokens": 22880081.0,
      "step": 11025
    },
    {
      "entropy": 5.496646070480347,
      "epoch": 1.000544267053701,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004907045501953304,
      "loss": 5.0765,
      "mean_token_accuracy": 0.19675455540418624,
      "num_tokens": 22892141.0,
      "step": 11030
    },
    {
      "entropy": 5.515450572967529,
      "epoch": 1.0009978229317853,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000490695343746539,
      "loss": 5.0916,
      "mean_token_accuracy": 0.19881475418806077,
      "num_tokens": 22902688.0,
      "step": 11035
    },
    {
      "entropy": 5.431853437423706,
      "epoch": 1.0014513788098693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004906861328371131,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19932159036397934,
      "num_tokens": 22912453.0,
      "step": 11040
    },
    {
      "entropy": 5.550728511810303,
      "epoch": 1.0019049346879536,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004906769174672429,
      "loss": 5.1846,
      "mean_token_accuracy": 0.18711721748113633,
      "num_tokens": 22923167.0,
      "step": 11045
    },
    {
      "entropy": 5.52794919013977,
      "epoch": 1.0023584905660377,
      "grad_norm": 0.96875,
      "learning_rate": 0.000490667697637119,
      "loss": 5.1534,
      "mean_token_accuracy": 0.18911569565534592,
      "num_tokens": 22934753.0,
      "step": 11050
    },
    {
      "entropy": 5.456609916687012,
      "epoch": 1.002812046444122,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004906584733469324,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20015866309404373,
      "num_tokens": 22945277.0,
      "step": 11055
    },
    {
      "entropy": 5.559716701507568,
      "epoch": 1.0032656023222062,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004906492445968735,
      "loss": 5.1634,
      "mean_token_accuracy": 0.18062861263751984,
      "num_tokens": 22955797.0,
      "step": 11060
    },
    {
      "entropy": 5.544274425506591,
      "epoch": 1.0037191582002902,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906400113871332,
      "loss": 5.132,
      "mean_token_accuracy": 0.19073731899261476,
      "num_tokens": 22965969.0,
      "step": 11065
    },
    {
      "entropy": 5.493846797943116,
      "epoch": 1.0041727140783745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004906307737179026,
      "loss": 5.1033,
      "mean_token_accuracy": 0.20635355561971663,
      "num_tokens": 22975323.0,
      "step": 11070
    },
    {
      "entropy": 5.447731399536133,
      "epoch": 1.0046262699564585,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004906215315893726,
      "loss": 5.108,
      "mean_token_accuracy": 0.19876318871974946,
      "num_tokens": 22985063.0,
      "step": 11075
    },
    {
      "entropy": 5.5594007015228275,
      "epoch": 1.0050798258345428,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004906122850017344,
      "loss": 5.2382,
      "mean_token_accuracy": 0.18069565892219544,
      "num_tokens": 22996001.0,
      "step": 11080
    },
    {
      "entropy": 5.500598669052124,
      "epoch": 1.005533381712627,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906030339551792,
      "loss": 5.1468,
      "mean_token_accuracy": 0.19543736428022385,
      "num_tokens": 23006896.0,
      "step": 11085
    },
    {
      "entropy": 5.514100217819214,
      "epoch": 1.0059869375907111,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905937784498983,
      "loss": 5.1205,
      "mean_token_accuracy": 0.1983324557542801,
      "num_tokens": 23017080.0,
      "step": 11090
    },
    {
      "entropy": 5.526914501190186,
      "epoch": 1.0064404934687954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004905845184860831,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19216082692146302,
      "num_tokens": 23027168.0,
      "step": 11095
    },
    {
      "entropy": 5.4901073455810545,
      "epoch": 1.0068940493468794,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905752540639252,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19402697086334228,
      "num_tokens": 23037364.0,
      "step": 11100
    },
    {
      "entropy": 5.444597339630127,
      "epoch": 1.0073476052249637,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004905659851836161,
      "loss": 5.0147,
      "mean_token_accuracy": 0.2009423241019249,
      "num_tokens": 23047609.0,
      "step": 11105
    },
    {
      "entropy": 5.493267250061035,
      "epoch": 1.007801161103048,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004905567118453475,
      "loss": 5.1704,
      "mean_token_accuracy": 0.18859416395425796,
      "num_tokens": 23057942.0,
      "step": 11110
    },
    {
      "entropy": 5.469370317459107,
      "epoch": 1.008254716981132,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004905474340493111,
      "loss": 5.0764,
      "mean_token_accuracy": 0.1908912554383278,
      "num_tokens": 23069119.0,
      "step": 11115
    },
    {
      "entropy": 5.45906629562378,
      "epoch": 1.0087082728592163,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004905381517956988,
      "loss": 5.0678,
      "mean_token_accuracy": 0.20038813650608062,
      "num_tokens": 23080060.0,
      "step": 11120
    },
    {
      "entropy": 5.434975862503052,
      "epoch": 1.0091618287373005,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004905288650847027,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19017168432474135,
      "num_tokens": 23090645.0,
      "step": 11125
    },
    {
      "entropy": 5.517891359329224,
      "epoch": 1.0096153846153846,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004905195739165146,
      "loss": 5.108,
      "mean_token_accuracy": 0.19634836316108703,
      "num_tokens": 23100449.0,
      "step": 11130
    },
    {
      "entropy": 5.488073348999023,
      "epoch": 1.0100689404934688,
      "grad_norm": 1.0,
      "learning_rate": 0.000490510278291327,
      "loss": 5.0931,
      "mean_token_accuracy": 0.1908187672495842,
      "num_tokens": 23110415.0,
      "step": 11135
    },
    {
      "entropy": 5.465580987930298,
      "epoch": 1.010522496371553,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004905009782093318,
      "loss": 5.08,
      "mean_token_accuracy": 0.19305008202791213,
      "num_tokens": 23120336.0,
      "step": 11140
    },
    {
      "entropy": 5.442973518371582,
      "epoch": 1.0109760522496372,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004904916736707214,
      "loss": 5.157,
      "mean_token_accuracy": 0.18449659645557404,
      "num_tokens": 23130568.0,
      "step": 11145
    },
    {
      "entropy": 5.543105840682983,
      "epoch": 1.0114296081277214,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004904823646756885,
      "loss": 5.112,
      "mean_token_accuracy": 0.1943865343928337,
      "num_tokens": 23140080.0,
      "step": 11150
    },
    {
      "entropy": 5.483130693435669,
      "epoch": 1.0118831640058055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004904730512244252,
      "loss": 5.0452,
      "mean_token_accuracy": 0.1947724238038063,
      "num_tokens": 23150360.0,
      "step": 11155
    },
    {
      "entropy": 5.424828720092774,
      "epoch": 1.0123367198838897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904637333171243,
      "loss": 5.0556,
      "mean_token_accuracy": 0.19245949983596802,
      "num_tokens": 23160620.0,
      "step": 11160
    },
    {
      "entropy": 5.550334787368774,
      "epoch": 1.0127902757619738,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004904544109539785,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19757008105516433,
      "num_tokens": 23171920.0,
      "step": 11165
    },
    {
      "entropy": 5.491248846054077,
      "epoch": 1.013243831640058,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004904450841351805,
      "loss": 5.1023,
      "mean_token_accuracy": 0.1991717129945755,
      "num_tokens": 23182929.0,
      "step": 11170
    },
    {
      "entropy": 5.514200067520141,
      "epoch": 1.0136973875181423,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004904357528609236,
      "loss": 5.0722,
      "mean_token_accuracy": 0.19128975570201873,
      "num_tokens": 23193020.0,
      "step": 11175
    },
    {
      "entropy": 5.448251628875733,
      "epoch": 1.0141509433962264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004904264171314,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20164622366428375,
      "num_tokens": 23204073.0,
      "step": 11180
    },
    {
      "entropy": 5.421363639831543,
      "epoch": 1.0146044992743106,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004904170769468035,
      "loss": 4.9913,
      "mean_token_accuracy": 0.19620340317487717,
      "num_tokens": 23214376.0,
      "step": 11185
    },
    {
      "entropy": 5.5082276344299315,
      "epoch": 1.0150580551523947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004904077323073268,
      "loss": 5.1497,
      "mean_token_accuracy": 0.19836536943912506,
      "num_tokens": 23224304.0,
      "step": 11190
    },
    {
      "entropy": 5.492418527603149,
      "epoch": 1.015511611030479,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004903983832131633,
      "loss": 5.1656,
      "mean_token_accuracy": 0.1937434956431389,
      "num_tokens": 23235324.0,
      "step": 11195
    },
    {
      "entropy": 5.493016624450684,
      "epoch": 1.0159651669085632,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004903890296645063,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19382944256067275,
      "num_tokens": 23245774.0,
      "step": 11200
    },
    {
      "entropy": 5.507924365997314,
      "epoch": 1.0164187227866472,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004903796716615495,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19998153001070024,
      "num_tokens": 23256274.0,
      "step": 11205
    },
    {
      "entropy": 5.483238792419433,
      "epoch": 1.0168722786647315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903703092044861,
      "loss": 5.0927,
      "mean_token_accuracy": 0.19459670037031174,
      "num_tokens": 23266304.0,
      "step": 11210
    },
    {
      "entropy": 5.416814374923706,
      "epoch": 1.0173258345428158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903609422935097,
      "loss": 5.0196,
      "mean_token_accuracy": 0.19843120574951173,
      "num_tokens": 23276552.0,
      "step": 11215
    },
    {
      "entropy": 5.462734937667847,
      "epoch": 1.0177793904208998,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004903515709288143,
      "loss": 5.1272,
      "mean_token_accuracy": 0.1896622136235237,
      "num_tokens": 23287799.0,
      "step": 11220
    },
    {
      "entropy": 5.425547027587891,
      "epoch": 1.018232946298984,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004903421951105933,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2000567466020584,
      "num_tokens": 23297958.0,
      "step": 11225
    },
    {
      "entropy": 5.490804672241211,
      "epoch": 1.0186865021770681,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004903328148390411,
      "loss": 5.1606,
      "mean_token_accuracy": 0.18824684619903564,
      "num_tokens": 23307523.0,
      "step": 11230
    },
    {
      "entropy": 5.533596992492676,
      "epoch": 1.0191400580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004903234301143513,
      "loss": 5.1305,
      "mean_token_accuracy": 0.1967354640364647,
      "num_tokens": 23317970.0,
      "step": 11235
    },
    {
      "entropy": 5.465874528884887,
      "epoch": 1.0195936139332367,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004903140409367181,
      "loss": 5.1225,
      "mean_token_accuracy": 0.19505420625209807,
      "num_tokens": 23328612.0,
      "step": 11240
    },
    {
      "entropy": 5.415059518814087,
      "epoch": 1.0200471698113207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004903046473063357,
      "loss": 5.0724,
      "mean_token_accuracy": 0.18936332017183305,
      "num_tokens": 23338104.0,
      "step": 11245
    },
    {
      "entropy": 5.445874977111816,
      "epoch": 1.020500725689405,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004902952492233983,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21072589308023454,
      "num_tokens": 23348105.0,
      "step": 11250
    },
    {
      "entropy": 5.442014884948731,
      "epoch": 1.020954281567489,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004902858466881002,
      "loss": 4.9613,
      "mean_token_accuracy": 0.19927150905132293,
      "num_tokens": 23357150.0,
      "step": 11255
    },
    {
      "entropy": 5.465661239624024,
      "epoch": 1.0214078374455733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490276439700636,
      "loss": 5.1693,
      "mean_token_accuracy": 0.1910180225968361,
      "num_tokens": 23367223.0,
      "step": 11260
    },
    {
      "entropy": 5.492317008972168,
      "epoch": 1.0218613933236576,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004902670282612002,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1948421135544777,
      "num_tokens": 23377664.0,
      "step": 11265
    },
    {
      "entropy": 5.424583578109742,
      "epoch": 1.0223149492017416,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004902576123699873,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20003152638673782,
      "num_tokens": 23388057.0,
      "step": 11270
    },
    {
      "entropy": 5.3437347412109375,
      "epoch": 1.0227685050798259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902481920271922,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20076482594013215,
      "num_tokens": 23397808.0,
      "step": 11275
    },
    {
      "entropy": 5.3924784660339355,
      "epoch": 1.02322206095791,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004902387672330097,
      "loss": 5.0345,
      "mean_token_accuracy": 0.19002968221902847,
      "num_tokens": 23406979.0,
      "step": 11280
    },
    {
      "entropy": 5.443336629867554,
      "epoch": 1.0236756168359942,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902293379876345,
      "loss": 5.0838,
      "mean_token_accuracy": 0.19750918447971344,
      "num_tokens": 23416545.0,
      "step": 11285
    },
    {
      "entropy": 5.479193782806396,
      "epoch": 1.0241291727140784,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004902199042912618,
      "loss": 5.1001,
      "mean_token_accuracy": 0.18947370052337648,
      "num_tokens": 23427464.0,
      "step": 11290
    },
    {
      "entropy": 5.497503757476807,
      "epoch": 1.0245827285921625,
      "grad_norm": 1.0,
      "learning_rate": 0.0004902104661440868,
      "loss": 5.1211,
      "mean_token_accuracy": 0.19749965518712997,
      "num_tokens": 23436690.0,
      "step": 11295
    },
    {
      "entropy": 5.464810943603515,
      "epoch": 1.0250362844702468,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902010235463042,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19768522083759307,
      "num_tokens": 23446701.0,
      "step": 11300
    },
    {
      "entropy": 5.534040021896362,
      "epoch": 1.025489840348331,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004901915764981098,
      "loss": 5.1473,
      "mean_token_accuracy": 0.19012537896633147,
      "num_tokens": 23456966.0,
      "step": 11305
    },
    {
      "entropy": 5.495480489730835,
      "epoch": 1.025943396226415,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004901821249996988,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19729931354522706,
      "num_tokens": 23467300.0,
      "step": 11310
    },
    {
      "entropy": 5.5246494770050045,
      "epoch": 1.0263969521044993,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004901726690512665,
      "loss": 5.1462,
      "mean_token_accuracy": 0.1901402324438095,
      "num_tokens": 23478457.0,
      "step": 11315
    },
    {
      "entropy": 5.463982629776001,
      "epoch": 1.0268505079825834,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004901632086530087,
      "loss": 5.0403,
      "mean_token_accuracy": 0.19925176948308945,
      "num_tokens": 23488066.0,
      "step": 11320
    },
    {
      "entropy": 5.4931072235107425,
      "epoch": 1.0273040638606676,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004901537438051207,
      "loss": 5.1304,
      "mean_token_accuracy": 0.1888103410601616,
      "num_tokens": 23498856.0,
      "step": 11325
    },
    {
      "entropy": 5.479607105255127,
      "epoch": 1.027757619738752,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004901442745077986,
      "loss": 5.0765,
      "mean_token_accuracy": 0.196649968624115,
      "num_tokens": 23509845.0,
      "step": 11330
    },
    {
      "entropy": 5.557180452346802,
      "epoch": 1.028211175616836,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004901348007612381,
      "loss": 5.2062,
      "mean_token_accuracy": 0.19007708877325058,
      "num_tokens": 23520322.0,
      "step": 11335
    },
    {
      "entropy": 5.499701595306396,
      "epoch": 1.0286647314949202,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901253225656351,
      "loss": 5.1263,
      "mean_token_accuracy": 0.19733863621950148,
      "num_tokens": 23530030.0,
      "step": 11340
    },
    {
      "entropy": 5.559247732162476,
      "epoch": 1.0291182873730043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004901158399211855,
      "loss": 5.1838,
      "mean_token_accuracy": 0.19454871863126755,
      "num_tokens": 23541063.0,
      "step": 11345
    },
    {
      "entropy": 5.469040298461914,
      "epoch": 1.0295718432510885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004901063528280856,
      "loss": 5.0867,
      "mean_token_accuracy": 0.1939779043197632,
      "num_tokens": 23551693.0,
      "step": 11350
    },
    {
      "entropy": 5.4881370067596436,
      "epoch": 1.0300253991291728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900968612865314,
      "loss": 5.1351,
      "mean_token_accuracy": 0.19694630354642867,
      "num_tokens": 23561758.0,
      "step": 11355
    },
    {
      "entropy": 5.429140520095825,
      "epoch": 1.0304789550072568,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900873652967195,
      "loss": 5.0771,
      "mean_token_accuracy": 0.18856350779533387,
      "num_tokens": 23571912.0,
      "step": 11360
    },
    {
      "entropy": 5.442412996292115,
      "epoch": 1.030932510885341,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004900778648588459,
      "loss": 5.0333,
      "mean_token_accuracy": 0.20146304965019227,
      "num_tokens": 23582600.0,
      "step": 11365
    },
    {
      "entropy": 5.472327375411988,
      "epoch": 1.0313860667634251,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004900683599731073,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20228659957647324,
      "num_tokens": 23592761.0,
      "step": 11370
    },
    {
      "entropy": 5.564744472503662,
      "epoch": 1.0318396226415094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900588506397002,
      "loss": 5.159,
      "mean_token_accuracy": 0.19236104786396027,
      "num_tokens": 23603119.0,
      "step": 11375
    },
    {
      "entropy": 5.404077386856079,
      "epoch": 1.0322931785195937,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004900493368588213,
      "loss": 5.068,
      "mean_token_accuracy": 0.20131881684064865,
      "num_tokens": 23613916.0,
      "step": 11380
    },
    {
      "entropy": 5.535603141784668,
      "epoch": 1.0327467343976777,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004900398186306672,
      "loss": 5.2023,
      "mean_token_accuracy": 0.19502190947532655,
      "num_tokens": 23623956.0,
      "step": 11385
    },
    {
      "entropy": 5.452536010742188,
      "epoch": 1.033200290275762,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000490030295955435,
      "loss": 5.0969,
      "mean_token_accuracy": 0.19548225551843643,
      "num_tokens": 23634666.0,
      "step": 11390
    },
    {
      "entropy": 5.414992809295654,
      "epoch": 1.0336538461538463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900207688333215,
      "loss": 5.0505,
      "mean_token_accuracy": 0.19136851727962495,
      "num_tokens": 23645601.0,
      "step": 11395
    },
    {
      "entropy": 5.4911417961120605,
      "epoch": 1.0341074020319303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900112372645236,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19436682015657425,
      "num_tokens": 23656339.0,
      "step": 11400
    },
    {
      "entropy": 5.497488927841187,
      "epoch": 1.0345609579100146,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004900017012492386,
      "loss": 5.0899,
      "mean_token_accuracy": 0.1950056940317154,
      "num_tokens": 23667433.0,
      "step": 11405
    },
    {
      "entropy": 5.528972911834717,
      "epoch": 1.0350145137880986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004899921607876634,
      "loss": 5.1392,
      "mean_token_accuracy": 0.1908893272280693,
      "num_tokens": 23678186.0,
      "step": 11410
    },
    {
      "entropy": 5.481750106811523,
      "epoch": 1.0354680696661829,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004899826158799958,
      "loss": 5.0724,
      "mean_token_accuracy": 0.19224554300308228,
      "num_tokens": 23688317.0,
      "step": 11415
    },
    {
      "entropy": 5.4784728527069095,
      "epoch": 1.0359216255442671,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899730665264328,
      "loss": 5.0224,
      "mean_token_accuracy": 0.19699247032403946,
      "num_tokens": 23697752.0,
      "step": 11420
    },
    {
      "entropy": 5.447753953933716,
      "epoch": 1.0363751814223512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899635127271721,
      "loss": 4.9843,
      "mean_token_accuracy": 0.20179656744003296,
      "num_tokens": 23707803.0,
      "step": 11425
    },
    {
      "entropy": 5.401472043991089,
      "epoch": 1.0368287373004355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899539544824111,
      "loss": 5.0116,
      "mean_token_accuracy": 0.2051843896508217,
      "num_tokens": 23717630.0,
      "step": 11430
    },
    {
      "entropy": 5.473198318481446,
      "epoch": 1.0372822931785195,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899443917923476,
      "loss": 5.1287,
      "mean_token_accuracy": 0.19700049012899398,
      "num_tokens": 23728836.0,
      "step": 11435
    },
    {
      "entropy": 5.452097368240357,
      "epoch": 1.0377358490566038,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004899348246571792,
      "loss": 5.1174,
      "mean_token_accuracy": 0.1917497396469116,
      "num_tokens": 23740145.0,
      "step": 11440
    },
    {
      "entropy": 5.4935462951660154,
      "epoch": 1.038189404934688,
      "grad_norm": 0.96875,
      "learning_rate": 0.000489925253077104,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19191719591617584,
      "num_tokens": 23751162.0,
      "step": 11445
    },
    {
      "entropy": 5.504214334487915,
      "epoch": 1.038642960812772,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004899156770523196,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19370600432157517,
      "num_tokens": 23761634.0,
      "step": 11450
    },
    {
      "entropy": 5.47665491104126,
      "epoch": 1.0390965166908563,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004899060965830244,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20089899897575378,
      "num_tokens": 23772483.0,
      "step": 11455
    },
    {
      "entropy": 5.452212142944336,
      "epoch": 1.0395500725689404,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004898965116694162,
      "loss": 5.0453,
      "mean_token_accuracy": 0.1946641504764557,
      "num_tokens": 23782298.0,
      "step": 11460
    },
    {
      "entropy": 5.494813919067383,
      "epoch": 1.0400036284470247,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898869223116935,
      "loss": 5.1309,
      "mean_token_accuracy": 0.18946383744478226,
      "num_tokens": 23792492.0,
      "step": 11465
    },
    {
      "entropy": 5.54174313545227,
      "epoch": 1.040457184325109,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898773285100546,
      "loss": 5.256,
      "mean_token_accuracy": 0.18755862563848497,
      "num_tokens": 23802999.0,
      "step": 11470
    },
    {
      "entropy": 5.487423419952393,
      "epoch": 1.040910740203193,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004898677302646975,
      "loss": 5.0665,
      "mean_token_accuracy": 0.1984709843993187,
      "num_tokens": 23812541.0,
      "step": 11475
    },
    {
      "entropy": 5.466430902481079,
      "epoch": 1.0413642960812772,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004898581275758211,
      "loss": 5.1706,
      "mean_token_accuracy": 0.18461524546146393,
      "num_tokens": 23823244.0,
      "step": 11480
    },
    {
      "entropy": 5.467811203002929,
      "epoch": 1.0418178519593615,
      "grad_norm": 1.0,
      "learning_rate": 0.0004898485204436239,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20237330198287964,
      "num_tokens": 23832346.0,
      "step": 11485
    },
    {
      "entropy": 5.515140390396118,
      "epoch": 1.0422714078374455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004898389088683047,
      "loss": 5.1301,
      "mean_token_accuracy": 0.20025138109922408,
      "num_tokens": 23842554.0,
      "step": 11490
    },
    {
      "entropy": 5.392171144485474,
      "epoch": 1.0427249637155298,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004898292928500619,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20212062299251557,
      "num_tokens": 23852938.0,
      "step": 11495
    },
    {
      "entropy": 5.421096229553223,
      "epoch": 1.0431785195936139,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004898196723890947,
      "loss": 4.9989,
      "mean_token_accuracy": 0.20295965373516084,
      "num_tokens": 23862367.0,
      "step": 11500
    },
    {
      "entropy": 5.464499378204346,
      "epoch": 1.0436320754716981,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898100474856019,
      "loss": 5.1109,
      "mean_token_accuracy": 0.1982620269060135,
      "num_tokens": 23873249.0,
      "step": 11505
    },
    {
      "entropy": 5.475397109985352,
      "epoch": 1.0440856313497824,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004898004181397824,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1877864271402359,
      "num_tokens": 23883830.0,
      "step": 11510
    },
    {
      "entropy": 5.5174485206604,
      "epoch": 1.0445391872278664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004897907843518358,
      "loss": 5.0633,
      "mean_token_accuracy": 0.19433184564113617,
      "num_tokens": 23894359.0,
      "step": 11515
    },
    {
      "entropy": 5.46463451385498,
      "epoch": 1.0449927431059507,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004897811461219609,
      "loss": 5.0531,
      "mean_token_accuracy": 0.19569095522165297,
      "num_tokens": 23903779.0,
      "step": 11520
    },
    {
      "entropy": 5.448574495315552,
      "epoch": 1.0454462989840347,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004897715034503571,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19483239948749542,
      "num_tokens": 23914523.0,
      "step": 11525
    },
    {
      "entropy": 5.456052017211914,
      "epoch": 1.045899854862119,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000489761856337224,
      "loss": 5.0868,
      "mean_token_accuracy": 0.19752579927444458,
      "num_tokens": 23924996.0,
      "step": 11530
    },
    {
      "entropy": 5.469106912612915,
      "epoch": 1.0463534107402033,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004897522047827609,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20350758582353592,
      "num_tokens": 23935843.0,
      "step": 11535
    },
    {
      "entropy": 5.433482885360718,
      "epoch": 1.0468069666182873,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004897425487871676,
      "loss": 5.0343,
      "mean_token_accuracy": 0.1945498988032341,
      "num_tokens": 23945044.0,
      "step": 11540
    },
    {
      "entropy": 5.449752473831177,
      "epoch": 1.0472605224963716,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897328883506436,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19367363899946213,
      "num_tokens": 23955933.0,
      "step": 11545
    },
    {
      "entropy": 5.481488513946533,
      "epoch": 1.0477140783744556,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004897232234733887,
      "loss": 5.1493,
      "mean_token_accuracy": 0.18738794922828675,
      "num_tokens": 23966320.0,
      "step": 11550
    },
    {
      "entropy": 5.456253623962402,
      "epoch": 1.04816763425254,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897135541556028,
      "loss": 5.0468,
      "mean_token_accuracy": 0.20325277298688887,
      "num_tokens": 23976233.0,
      "step": 11555
    },
    {
      "entropy": 5.49235692024231,
      "epoch": 1.0486211901306242,
      "grad_norm": 0.84375,
      "learning_rate": 0.000489703880397486,
      "loss": 5.11,
      "mean_token_accuracy": 0.19220408350229262,
      "num_tokens": 23987854.0,
      "step": 11560
    },
    {
      "entropy": 5.457041692733765,
      "epoch": 1.0490747460087082,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004896942021992382,
      "loss": 5.1154,
      "mean_token_accuracy": 0.19077137857675552,
      "num_tokens": 23997973.0,
      "step": 11565
    },
    {
      "entropy": 5.482615661621094,
      "epoch": 1.0495283018867925,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004896845195610595,
      "loss": 5.1169,
      "mean_token_accuracy": 0.1928719624876976,
      "num_tokens": 24007313.0,
      "step": 11570
    },
    {
      "entropy": 5.476922655105591,
      "epoch": 1.0499818577648767,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004896748324831504,
      "loss": 5.1244,
      "mean_token_accuracy": 0.19298154413700103,
      "num_tokens": 24018254.0,
      "step": 11575
    },
    {
      "entropy": 5.492932891845703,
      "epoch": 1.0504354136429608,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004896651409657109,
      "loss": 5.1368,
      "mean_token_accuracy": 0.19103600829839706,
      "num_tokens": 24029799.0,
      "step": 11580
    },
    {
      "entropy": 5.414491605758667,
      "epoch": 1.050888969521045,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004896554450089417,
      "loss": 5.0258,
      "mean_token_accuracy": 0.19909809976816178,
      "num_tokens": 24040972.0,
      "step": 11585
    },
    {
      "entropy": 5.435051107406617,
      "epoch": 1.051342525399129,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896457446130431,
      "loss": 4.958,
      "mean_token_accuracy": 0.2022219255566597,
      "num_tokens": 24050781.0,
      "step": 11590
    },
    {
      "entropy": 5.418264770507813,
      "epoch": 1.0517960812772134,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896360397782159,
      "loss": 5.1011,
      "mean_token_accuracy": 0.1896982803940773,
      "num_tokens": 24060184.0,
      "step": 11595
    },
    {
      "entropy": 5.513707160949707,
      "epoch": 1.0522496371552976,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004896263305046607,
      "loss": 5.1531,
      "mean_token_accuracy": 0.1932757616043091,
      "num_tokens": 24070777.0,
      "step": 11600
    },
    {
      "entropy": 5.481814289093018,
      "epoch": 1.0527031930333817,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896166167925782,
      "loss": 5.1822,
      "mean_token_accuracy": 0.19279300570487976,
      "num_tokens": 24080878.0,
      "step": 11605
    },
    {
      "entropy": 5.519545316696167,
      "epoch": 1.053156748911466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004896068986421693,
      "loss": 5.1407,
      "mean_token_accuracy": 0.19935741722583772,
      "num_tokens": 24090038.0,
      "step": 11610
    },
    {
      "entropy": 5.561168384552002,
      "epoch": 1.05361030478955,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004895971760536351,
      "loss": 5.1541,
      "mean_token_accuracy": 0.19477766305208205,
      "num_tokens": 24100425.0,
      "step": 11615
    },
    {
      "entropy": 5.538584423065186,
      "epoch": 1.0540638606676342,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004895874490271768,
      "loss": 5.2186,
      "mean_token_accuracy": 0.18994971215724946,
      "num_tokens": 24112182.0,
      "step": 11620
    },
    {
      "entropy": 5.481795358657837,
      "epoch": 1.0545174165457185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895777175629952,
      "loss": 5.0765,
      "mean_token_accuracy": 0.1948103278875351,
      "num_tokens": 24122182.0,
      "step": 11625
    },
    {
      "entropy": 5.5433516025543215,
      "epoch": 1.0549709724238026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004895679816612917,
      "loss": 5.0685,
      "mean_token_accuracy": 0.2016040116548538,
      "num_tokens": 24131930.0,
      "step": 11630
    },
    {
      "entropy": 5.5081311702728275,
      "epoch": 1.0554245283018868,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895582413222678,
      "loss": 5.0693,
      "mean_token_accuracy": 0.20441591888666152,
      "num_tokens": 24141575.0,
      "step": 11635
    },
    {
      "entropy": 5.475733232498169,
      "epoch": 1.0558780841799709,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895484965461247,
      "loss": 5.2086,
      "mean_token_accuracy": 0.1897387534379959,
      "num_tokens": 24151720.0,
      "step": 11640
    },
    {
      "entropy": 5.4533342838287355,
      "epoch": 1.0563316400580551,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000489538747333064,
      "loss": 5.089,
      "mean_token_accuracy": 0.19781393110752105,
      "num_tokens": 24161881.0,
      "step": 11645
    },
    {
      "entropy": 5.43628077507019,
      "epoch": 1.0567851959361394,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895289936832874,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20072407871484757,
      "num_tokens": 24171474.0,
      "step": 11650
    },
    {
      "entropy": 5.535496854782105,
      "epoch": 1.0572387518142234,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004895192355969966,
      "loss": 5.1662,
      "mean_token_accuracy": 0.19123655706644058,
      "num_tokens": 24182186.0,
      "step": 11655
    },
    {
      "entropy": 5.506235218048095,
      "epoch": 1.0576923076923077,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004895094730743932,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19827568531036377,
      "num_tokens": 24192318.0,
      "step": 11660
    },
    {
      "entropy": 5.52546124458313,
      "epoch": 1.058145863570392,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004894997061156793,
      "loss": 5.1616,
      "mean_token_accuracy": 0.19303964376449584,
      "num_tokens": 24202918.0,
      "step": 11665
    },
    {
      "entropy": 5.439881277084351,
      "epoch": 1.058599419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894899347210567,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20049711018800737,
      "num_tokens": 24213218.0,
      "step": 11670
    },
    {
      "entropy": 5.467959308624268,
      "epoch": 1.0590529753265603,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004894801588907278,
      "loss": 5.0387,
      "mean_token_accuracy": 0.198946575820446,
      "num_tokens": 24223229.0,
      "step": 11675
    },
    {
      "entropy": 5.52664999961853,
      "epoch": 1.0595065312046443,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004894703786248944,
      "loss": 5.1111,
      "mean_token_accuracy": 0.19159167855978013,
      "num_tokens": 24233637.0,
      "step": 11680
    },
    {
      "entropy": 5.478523540496826,
      "epoch": 1.0599600870827286,
      "grad_norm": 1.0,
      "learning_rate": 0.000489460593923759,
      "loss": 5.1528,
      "mean_token_accuracy": 0.1973455563187599,
      "num_tokens": 24244251.0,
      "step": 11685
    },
    {
      "entropy": 5.44070839881897,
      "epoch": 1.0604136429608129,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004894508047875237,
      "loss": 5.0526,
      "mean_token_accuracy": 0.19362615346908568,
      "num_tokens": 24254380.0,
      "step": 11690
    },
    {
      "entropy": 5.355499649047852,
      "epoch": 1.060867198838897,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004894410112163913,
      "loss": 5.0026,
      "mean_token_accuracy": 0.2004845768213272,
      "num_tokens": 24264841.0,
      "step": 11695
    },
    {
      "entropy": 5.414450836181641,
      "epoch": 1.0613207547169812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004894312132105641,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20084351897239686,
      "num_tokens": 24274383.0,
      "step": 11700
    },
    {
      "entropy": 5.5343756675720215,
      "epoch": 1.0617743105950652,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894214107702447,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19480592012405396,
      "num_tokens": 24284140.0,
      "step": 11705
    },
    {
      "entropy": 5.590111207962036,
      "epoch": 1.0622278664731495,
      "grad_norm": 1.109375,
      "learning_rate": 0.000489411603895636,
      "loss": 5.1771,
      "mean_token_accuracy": 0.19371833950281142,
      "num_tokens": 24295044.0,
      "step": 11710
    },
    {
      "entropy": 5.4418053150177,
      "epoch": 1.0626814223512338,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894017925869406,
      "loss": 5.08,
      "mean_token_accuracy": 0.19503361880779266,
      "num_tokens": 24305080.0,
      "step": 11715
    },
    {
      "entropy": 5.4726255416870115,
      "epoch": 1.0631349782293178,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004893919768443615,
      "loss": 5.1257,
      "mean_token_accuracy": 0.19226745814085006,
      "num_tokens": 24316306.0,
      "step": 11720
    },
    {
      "entropy": 5.4423802375793455,
      "epoch": 1.063588534107402,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004893821566681017,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19782619476318358,
      "num_tokens": 24326740.0,
      "step": 11725
    },
    {
      "entropy": 5.503826522827149,
      "epoch": 1.0640420899854863,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004893723320583644,
      "loss": 5.1457,
      "mean_token_accuracy": 0.18667171746492386,
      "num_tokens": 24337621.0,
      "step": 11730
    },
    {
      "entropy": 5.399421215057373,
      "epoch": 1.0644956458635704,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004893625030153525,
      "loss": 4.9887,
      "mean_token_accuracy": 0.20595765560865403,
      "num_tokens": 24347732.0,
      "step": 11735
    },
    {
      "entropy": 5.4055805683135985,
      "epoch": 1.0649492017416546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893526695392694,
      "loss": 5.0222,
      "mean_token_accuracy": 0.20757206678390502,
      "num_tokens": 24359173.0,
      "step": 11740
    },
    {
      "entropy": 5.422066068649292,
      "epoch": 1.0654027576197387,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893428316303185,
      "loss": 5.0202,
      "mean_token_accuracy": 0.19757631570100784,
      "num_tokens": 24369465.0,
      "step": 11745
    },
    {
      "entropy": 5.412892198562622,
      "epoch": 1.065856313497823,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004893329892887031,
      "loss": 5.0619,
      "mean_token_accuracy": 0.1981126308441162,
      "num_tokens": 24378899.0,
      "step": 11750
    },
    {
      "entropy": 5.484836912155151,
      "epoch": 1.0663098693759072,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004893231425146271,
      "loss": 5.1799,
      "mean_token_accuracy": 0.19192711710929872,
      "num_tokens": 24388919.0,
      "step": 11755
    },
    {
      "entropy": 5.4312244892120365,
      "epoch": 1.0667634252539913,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004893132913082937,
      "loss": 5.0478,
      "mean_token_accuracy": 0.1983172506093979,
      "num_tokens": 24399632.0,
      "step": 11760
    },
    {
      "entropy": 5.488347053527832,
      "epoch": 1.0672169811320755,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004893034356699069,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19779288917779922,
      "num_tokens": 24410817.0,
      "step": 11765
    },
    {
      "entropy": 5.417764663696289,
      "epoch": 1.0676705370101596,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004892935755996705,
      "loss": 5.0904,
      "mean_token_accuracy": 0.1991182193160057,
      "num_tokens": 24421058.0,
      "step": 11770
    },
    {
      "entropy": 5.408098793029785,
      "epoch": 1.0681240928882438,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892837110977883,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2003073051571846,
      "num_tokens": 24431522.0,
      "step": 11775
    },
    {
      "entropy": 5.462185478210449,
      "epoch": 1.068577648766328,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004892738421644644,
      "loss": 5.0522,
      "mean_token_accuracy": 0.19488768875598908,
      "num_tokens": 24441746.0,
      "step": 11780
    },
    {
      "entropy": 5.397897434234619,
      "epoch": 1.0690312046444121,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892639687999027,
      "loss": 4.9586,
      "mean_token_accuracy": 0.20120252817869186,
      "num_tokens": 24451916.0,
      "step": 11785
    },
    {
      "entropy": 5.439264392852783,
      "epoch": 1.0694847605224964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004892540910043076,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20249505788087846,
      "num_tokens": 24461935.0,
      "step": 11790
    },
    {
      "entropy": 5.457518911361694,
      "epoch": 1.0699383164005805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892442087778833,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20152089595794678,
      "num_tokens": 24472437.0,
      "step": 11795
    },
    {
      "entropy": 5.457108974456787,
      "epoch": 1.0703918722786647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004892343221208342,
      "loss": 5.0965,
      "mean_token_accuracy": 0.20119676738977432,
      "num_tokens": 24481726.0,
      "step": 11800
    },
    {
      "entropy": 5.491053771972656,
      "epoch": 1.070845428156749,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004892244310333646,
      "loss": 5.1704,
      "mean_token_accuracy": 0.19343004524707794,
      "num_tokens": 24492340.0,
      "step": 11805
    },
    {
      "entropy": 5.452194881439209,
      "epoch": 1.071298984034833,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004892145355156793,
      "loss": 5.0366,
      "mean_token_accuracy": 0.19774752855300903,
      "num_tokens": 24502138.0,
      "step": 11810
    },
    {
      "entropy": 5.4785271167755125,
      "epoch": 1.0717525399129173,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004892046355679829,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19350825250148773,
      "num_tokens": 24512723.0,
      "step": 11815
    },
    {
      "entropy": 5.504654312133789,
      "epoch": 1.0722060957910013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004891947311904799,
      "loss": 5.1181,
      "mean_token_accuracy": 0.19506742358207702,
      "num_tokens": 24523766.0,
      "step": 11820
    },
    {
      "entropy": 5.500041055679321,
      "epoch": 1.0726596516690856,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004891848223833753,
      "loss": 5.1052,
      "mean_token_accuracy": 0.19698472768068315,
      "num_tokens": 24534571.0,
      "step": 11825
    },
    {
      "entropy": 5.482295513153076,
      "epoch": 1.0731132075471699,
      "grad_norm": 1.03125,
      "learning_rate": 0.000489174909146874,
      "loss": 5.1054,
      "mean_token_accuracy": 0.19994303733110427,
      "num_tokens": 24545265.0,
      "step": 11830
    },
    {
      "entropy": 5.5266765594482425,
      "epoch": 1.073566763425254,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004891649914811811,
      "loss": 5.1877,
      "mean_token_accuracy": 0.18802001029253007,
      "num_tokens": 24555524.0,
      "step": 11835
    },
    {
      "entropy": 5.397583484649658,
      "epoch": 1.0740203193033382,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891550693865014,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21245964467525483,
      "num_tokens": 24565420.0,
      "step": 11840
    },
    {
      "entropy": 5.469696044921875,
      "epoch": 1.0744738751814222,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004891451428630405,
      "loss": 5.0954,
      "mean_token_accuracy": 0.1935744747519493,
      "num_tokens": 24576794.0,
      "step": 11845
    },
    {
      "entropy": 5.500405263900757,
      "epoch": 1.0749274310595065,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004891352119110033,
      "loss": 5.0969,
      "mean_token_accuracy": 0.19519586563110353,
      "num_tokens": 24586320.0,
      "step": 11850
    },
    {
      "entropy": 5.446904039382934,
      "epoch": 1.0753809869375908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004891252765305954,
      "loss": 5.0769,
      "mean_token_accuracy": 0.2048339292407036,
      "num_tokens": 24596524.0,
      "step": 11855
    },
    {
      "entropy": 5.4598876953125,
      "epoch": 1.0758345428156748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891153367220222,
      "loss": 5.016,
      "mean_token_accuracy": 0.20519405454397202,
      "num_tokens": 24605877.0,
      "step": 11860
    },
    {
      "entropy": 5.514345026016235,
      "epoch": 1.076288098693759,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004891053924854894,
      "loss": 5.2141,
      "mean_token_accuracy": 0.1887012705206871,
      "num_tokens": 24616719.0,
      "step": 11865
    },
    {
      "entropy": 5.420893049240112,
      "epoch": 1.0767416545718433,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890954438212024,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20212965607643127,
      "num_tokens": 24626882.0,
      "step": 11870
    },
    {
      "entropy": 5.469028377532959,
      "epoch": 1.0771952104499274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890854907293671,
      "loss": 5.0921,
      "mean_token_accuracy": 0.19691340178251265,
      "num_tokens": 24637884.0,
      "step": 11875
    },
    {
      "entropy": 5.448418807983399,
      "epoch": 1.0776487663280117,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004890755332101893,
      "loss": 5.0798,
      "mean_token_accuracy": 0.1930866315960884,
      "num_tokens": 24647389.0,
      "step": 11880
    },
    {
      "entropy": 5.449471950531006,
      "epoch": 1.0781023222060957,
      "grad_norm": 0.875,
      "learning_rate": 0.0004890655712638749,
      "loss": 5.0972,
      "mean_token_accuracy": 0.19339723885059357,
      "num_tokens": 24657595.0,
      "step": 11885
    },
    {
      "entropy": 5.456168222427368,
      "epoch": 1.07855587808418,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048905560489063,
      "loss": 5.0011,
      "mean_token_accuracy": 0.19946667104959487,
      "num_tokens": 24667683.0,
      "step": 11890
    },
    {
      "entropy": 5.491331672668457,
      "epoch": 1.0790094339622642,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004890456340906605,
      "loss": 5.157,
      "mean_token_accuracy": 0.1912636011838913,
      "num_tokens": 24678041.0,
      "step": 11895
    },
    {
      "entropy": 5.445166206359863,
      "epoch": 1.0794629898403483,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004890356588641729,
      "loss": 5.1521,
      "mean_token_accuracy": 0.20095144957304,
      "num_tokens": 24688569.0,
      "step": 11900
    },
    {
      "entropy": 5.475362777709961,
      "epoch": 1.0799165457184325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004890256792113733,
      "loss": 5.0967,
      "mean_token_accuracy": 0.19915975630283356,
      "num_tokens": 24699913.0,
      "step": 11905
    },
    {
      "entropy": 5.547744035720825,
      "epoch": 1.0803701015965168,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004890156951324681,
      "loss": 5.1223,
      "mean_token_accuracy": 0.19804803878068925,
      "num_tokens": 24710682.0,
      "step": 11910
    },
    {
      "entropy": 5.558585834503174,
      "epoch": 1.0808236574746009,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004890057066276636,
      "loss": 5.1974,
      "mean_token_accuracy": 0.18895184248685837,
      "num_tokens": 24721087.0,
      "step": 11915
    },
    {
      "entropy": 5.424392366409302,
      "epoch": 1.0812772133526851,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004889957136971668,
      "loss": 5.015,
      "mean_token_accuracy": 0.20212464183568954,
      "num_tokens": 24731323.0,
      "step": 11920
    },
    {
      "entropy": 5.424842023849488,
      "epoch": 1.0817307692307692,
      "grad_norm": 1.0,
      "learning_rate": 0.000488985716341184,
      "loss": 5.1052,
      "mean_token_accuracy": 0.1901997834444046,
      "num_tokens": 24742559.0,
      "step": 11925
    },
    {
      "entropy": 5.492843246459961,
      "epoch": 1.0821843251088534,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004889757145599222,
      "loss": 5.0918,
      "mean_token_accuracy": 0.1986592561006546,
      "num_tokens": 24752307.0,
      "step": 11930
    },
    {
      "entropy": 5.383856630325317,
      "epoch": 1.0826378809869377,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000488965708353588,
      "loss": 5.0516,
      "mean_token_accuracy": 0.19449937641620635,
      "num_tokens": 24762792.0,
      "step": 11935
    },
    {
      "entropy": 5.493750333786011,
      "epoch": 1.0830914368650217,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004889556977223883,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19200471192598342,
      "num_tokens": 24774377.0,
      "step": 11940
    },
    {
      "entropy": 5.503545761108398,
      "epoch": 1.083544992743106,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004889456826665304,
      "loss": 5.078,
      "mean_token_accuracy": 0.19847277849912642,
      "num_tokens": 24785919.0,
      "step": 11945
    },
    {
      "entropy": 5.51596851348877,
      "epoch": 1.08399854862119,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004889356631862211,
      "loss": 5.115,
      "mean_token_accuracy": 0.18979937434196473,
      "num_tokens": 24796594.0,
      "step": 11950
    },
    {
      "entropy": 5.460939598083496,
      "epoch": 1.0844521044992743,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000488925639281668,
      "loss": 5.1161,
      "mean_token_accuracy": 0.1949406310915947,
      "num_tokens": 24807322.0,
      "step": 11955
    },
    {
      "entropy": 5.417977905273437,
      "epoch": 1.0849056603773586,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004889156109530781,
      "loss": 4.9998,
      "mean_token_accuracy": 0.19890448153018953,
      "num_tokens": 24817593.0,
      "step": 11960
    },
    {
      "entropy": 5.4494860649108885,
      "epoch": 1.0853592162554426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889055782006588,
      "loss": 5.0115,
      "mean_token_accuracy": 0.20230089426040648,
      "num_tokens": 24827577.0,
      "step": 11965
    },
    {
      "entropy": 5.441661405563354,
      "epoch": 1.085812772133527,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004888955410246176,
      "loss": 5.0568,
      "mean_token_accuracy": 0.19938754588365554,
      "num_tokens": 24838148.0,
      "step": 11970
    },
    {
      "entropy": 5.443475580215454,
      "epoch": 1.086266328011611,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004888854994251621,
      "loss": 5.0108,
      "mean_token_accuracy": 0.19843291789293288,
      "num_tokens": 24848451.0,
      "step": 11975
    },
    {
      "entropy": 5.492741966247559,
      "epoch": 1.0867198838896952,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004888754534025,
      "loss": 5.1404,
      "mean_token_accuracy": 0.18598994314670564,
      "num_tokens": 24859619.0,
      "step": 11980
    },
    {
      "entropy": 5.499451589584351,
      "epoch": 1.0871734397677795,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888654029568391,
      "loss": 5.1711,
      "mean_token_accuracy": 0.18404650390148164,
      "num_tokens": 24869863.0,
      "step": 11985
    },
    {
      "entropy": 5.457519960403443,
      "epoch": 1.0876269956458635,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004888553480883871,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1899152174592018,
      "num_tokens": 24880524.0,
      "step": 11990
    },
    {
      "entropy": 5.415886926651001,
      "epoch": 1.0880805515239478,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000488845288797352,
      "loss": 5.0696,
      "mean_token_accuracy": 0.19850811511278152,
      "num_tokens": 24891514.0,
      "step": 11995
    },
    {
      "entropy": 5.4387938499450685,
      "epoch": 1.0885341074020318,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888352250839418,
      "loss": 5.0251,
      "mean_token_accuracy": 0.19936315417289735,
      "num_tokens": 24900842.0,
      "step": 12000
    },
    {
      "epoch": 1.0885341074020318,
      "eval_entropy": 5.235634247102349,
      "eval_loss": 5.16903829574585,
      "eval_mean_token_accuracy": 0.20154844025714636,
      "eval_num_tokens": 24900842.0,
      "eval_runtime": 20.6538,
      "eval_samples_per_second": 1712.03,
      "eval_steps_per_second": 214.004,
      "step": 12000
    },
    {
      "entropy": 5.4245994091033936,
      "epoch": 1.088987663280116,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004888251569483647,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20064777582883836,
      "num_tokens": 24911845.0,
      "step": 12005
    },
    {
      "entropy": 5.382421827316284,
      "epoch": 1.0894412191582004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004888150843908288,
      "loss": 5.0402,
      "mean_token_accuracy": 0.20876970440149306,
      "num_tokens": 24922211.0,
      "step": 12010
    },
    {
      "entropy": 5.518812799453736,
      "epoch": 1.0898947750362844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888050074115426,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1951376900076866,
      "num_tokens": 24932129.0,
      "step": 12015
    },
    {
      "entropy": 5.550019693374634,
      "epoch": 1.0903483309143687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004887949260107143,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19230424612760544,
      "num_tokens": 24942869.0,
      "step": 12020
    },
    {
      "entropy": 5.4164728164672855,
      "epoch": 1.0908018867924527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887848401885524,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19477627873420716,
      "num_tokens": 24952621.0,
      "step": 12025
    },
    {
      "entropy": 5.556595325469971,
      "epoch": 1.091255442670537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004887747499452656,
      "loss": 5.195,
      "mean_token_accuracy": 0.1833399146795273,
      "num_tokens": 24963587.0,
      "step": 12030
    },
    {
      "entropy": 5.515872573852539,
      "epoch": 1.0917089985486212,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004887646552810624,
      "loss": 5.0398,
      "mean_token_accuracy": 0.2013555198907852,
      "num_tokens": 24974122.0,
      "step": 12035
    },
    {
      "entropy": 5.483128356933594,
      "epoch": 1.0921625544267053,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004887545561961517,
      "loss": 5.1285,
      "mean_token_accuracy": 0.1937446415424347,
      "num_tokens": 24985145.0,
      "step": 12040
    },
    {
      "entropy": 5.429468202590942,
      "epoch": 1.0926161103047896,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004887444526907424,
      "loss": 5.0495,
      "mean_token_accuracy": 0.1973423257470131,
      "num_tokens": 24995511.0,
      "step": 12045
    },
    {
      "entropy": 5.477809238433838,
      "epoch": 1.0930696661828738,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004887343447650432,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19301147162914276,
      "num_tokens": 25006193.0,
      "step": 12050
    },
    {
      "entropy": 5.505136442184448,
      "epoch": 1.0935232220609579,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004887242324192634,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19657760858535767,
      "num_tokens": 25016444.0,
      "step": 12055
    },
    {
      "entropy": 5.528617286682129,
      "epoch": 1.0939767779390421,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004887141156536119,
      "loss": 5.0888,
      "mean_token_accuracy": 0.19392160177230836,
      "num_tokens": 25027898.0,
      "step": 12060
    },
    {
      "entropy": 5.447785425186157,
      "epoch": 1.0944303338171262,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004887039944682981,
      "loss": 4.9868,
      "mean_token_accuracy": 0.20039299726486207,
      "num_tokens": 25038201.0,
      "step": 12065
    },
    {
      "entropy": 5.549479961395264,
      "epoch": 1.0948838896952104,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886938688635313,
      "loss": 5.1931,
      "mean_token_accuracy": 0.1823815271258354,
      "num_tokens": 25048466.0,
      "step": 12070
    },
    {
      "entropy": 5.569545698165894,
      "epoch": 1.0953374455732947,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004886837388395207,
      "loss": 5.1347,
      "mean_token_accuracy": 0.19485244303941726,
      "num_tokens": 25059064.0,
      "step": 12075
    },
    {
      "entropy": 5.543216323852539,
      "epoch": 1.0957910014513788,
      "grad_norm": 1.078125,
      "learning_rate": 0.000488673604396476,
      "loss": 5.2136,
      "mean_token_accuracy": 0.18050719946622848,
      "num_tokens": 25068834.0,
      "step": 12080
    },
    {
      "entropy": 5.506651210784912,
      "epoch": 1.096244557329463,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886634655346067,
      "loss": 5.155,
      "mean_token_accuracy": 0.18910235464572905,
      "num_tokens": 25080058.0,
      "step": 12085
    },
    {
      "entropy": 5.440000104904175,
      "epoch": 1.0966981132075473,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886533222541225,
      "loss": 5.0182,
      "mean_token_accuracy": 0.20164837390184404,
      "num_tokens": 25089751.0,
      "step": 12090
    },
    {
      "entropy": 5.478103876113892,
      "epoch": 1.0971516690856313,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886431745552331,
      "loss": 5.1127,
      "mean_token_accuracy": 0.19630647152662278,
      "num_tokens": 25099055.0,
      "step": 12095
    },
    {
      "entropy": 5.54117488861084,
      "epoch": 1.0976052249637156,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004886330224381485,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18562830686569215,
      "num_tokens": 25109793.0,
      "step": 12100
    },
    {
      "entropy": 5.521268510818482,
      "epoch": 1.0980587808417996,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004886228659030785,
      "loss": 5.1163,
      "mean_token_accuracy": 0.19521568566560746,
      "num_tokens": 25119814.0,
      "step": 12105
    },
    {
      "entropy": 5.5408790588378904,
      "epoch": 1.098512336719884,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004886127049502331,
      "loss": 5.1591,
      "mean_token_accuracy": 0.19464686661958694,
      "num_tokens": 25130747.0,
      "step": 12110
    },
    {
      "entropy": 5.454093742370605,
      "epoch": 1.0989658925979682,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886025395798228,
      "loss": 5.0522,
      "mean_token_accuracy": 0.1918802246451378,
      "num_tokens": 25142235.0,
      "step": 12115
    },
    {
      "entropy": 5.366468334197998,
      "epoch": 1.0994194484760522,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004885923697920573,
      "loss": 4.9928,
      "mean_token_accuracy": 0.205183207988739,
      "num_tokens": 25152047.0,
      "step": 12120
    },
    {
      "entropy": 5.431280422210693,
      "epoch": 1.0998730043541365,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004885821955871472,
      "loss": 5.0471,
      "mean_token_accuracy": 0.19808025658130646,
      "num_tokens": 25162678.0,
      "step": 12125
    },
    {
      "entropy": 5.49901065826416,
      "epoch": 1.1003265602322205,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885720169653028,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19658662378787994,
      "num_tokens": 25172973.0,
      "step": 12130
    },
    {
      "entropy": 5.507623529434204,
      "epoch": 1.1007801161103048,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004885618339267348,
      "loss": 5.1716,
      "mean_token_accuracy": 0.1899319440126419,
      "num_tokens": 25183458.0,
      "step": 12135
    },
    {
      "entropy": 5.445809078216553,
      "epoch": 1.101233671988389,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885516464716537,
      "loss": 5.0571,
      "mean_token_accuracy": 0.19627972692251205,
      "num_tokens": 25193827.0,
      "step": 12140
    },
    {
      "entropy": 5.5591919898986815,
      "epoch": 1.101687227866473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048854145460027,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19065470844507218,
      "num_tokens": 25204448.0,
      "step": 12145
    },
    {
      "entropy": 5.519454479217529,
      "epoch": 1.1021407837445574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885312583127946,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20112146586179733,
      "num_tokens": 25215041.0,
      "step": 12150
    },
    {
      "entropy": 5.444091129302978,
      "epoch": 1.1025943396226414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004885210576094385,
      "loss": 5.0821,
      "mean_token_accuracy": 0.1961773082613945,
      "num_tokens": 25225937.0,
      "step": 12155
    },
    {
      "entropy": 5.418184804916382,
      "epoch": 1.1030478955007257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885108524904124,
      "loss": 4.995,
      "mean_token_accuracy": 0.1989628866314888,
      "num_tokens": 25235270.0,
      "step": 12160
    },
    {
      "entropy": 5.467058181762695,
      "epoch": 1.10350145137881,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004885006429559275,
      "loss": 5.101,
      "mean_token_accuracy": 0.18867916166782378,
      "num_tokens": 25247080.0,
      "step": 12165
    },
    {
      "entropy": 5.534938716888428,
      "epoch": 1.103955007256894,
      "grad_norm": 1.125,
      "learning_rate": 0.0004884904290061949,
      "loss": 5.0948,
      "mean_token_accuracy": 0.18896181732416154,
      "num_tokens": 25256718.0,
      "step": 12170
    },
    {
      "entropy": 5.4963476181030275,
      "epoch": 1.1044085631349783,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004884802106414258,
      "loss": 5.2132,
      "mean_token_accuracy": 0.1859460324048996,
      "num_tokens": 25266632.0,
      "step": 12175
    },
    {
      "entropy": 5.516941118240356,
      "epoch": 1.1048621190130623,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004884699878618316,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19886712729930878,
      "num_tokens": 25275598.0,
      "step": 12180
    },
    {
      "entropy": 5.540178489685059,
      "epoch": 1.1053156748911466,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004884597606676236,
      "loss": 5.1522,
      "mean_token_accuracy": 0.1940113976597786,
      "num_tokens": 25286598.0,
      "step": 12185
    },
    {
      "entropy": 5.459210348129273,
      "epoch": 1.1057692307692308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884495290590134,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19530683755874634,
      "num_tokens": 25296730.0,
      "step": 12190
    },
    {
      "entropy": 5.447724723815918,
      "epoch": 1.1062227866473149,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004884392930362124,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20545725524425507,
      "num_tokens": 25306326.0,
      "step": 12195
    },
    {
      "entropy": 5.494918966293335,
      "epoch": 1.1066763425253991,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004884290525994326,
      "loss": 5.0998,
      "mean_token_accuracy": 0.19826816022396088,
      "num_tokens": 25316342.0,
      "step": 12200
    },
    {
      "entropy": 5.579006290435791,
      "epoch": 1.1071298984034832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004884188077488855,
      "loss": 5.1396,
      "mean_token_accuracy": 0.1880671426653862,
      "num_tokens": 25326129.0,
      "step": 12205
    },
    {
      "entropy": 5.491881370544434,
      "epoch": 1.1075834542815675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884085584847831,
      "loss": 5.1088,
      "mean_token_accuracy": 0.18871529400348663,
      "num_tokens": 25337156.0,
      "step": 12210
    },
    {
      "entropy": 5.553006219863891,
      "epoch": 1.1080370101596517,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004883983048073375,
      "loss": 5.2199,
      "mean_token_accuracy": 0.18386245816946029,
      "num_tokens": 25347632.0,
      "step": 12215
    },
    {
      "entropy": 5.458075666427613,
      "epoch": 1.1084905660377358,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004883880467167603,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20122691094875336,
      "num_tokens": 25359186.0,
      "step": 12220
    },
    {
      "entropy": 5.496602344512939,
      "epoch": 1.10894412191582,
      "grad_norm": 1.15625,
      "learning_rate": 0.000488377784213264,
      "loss": 5.0975,
      "mean_token_accuracy": 0.1946451857686043,
      "num_tokens": 25368993.0,
      "step": 12225
    },
    {
      "entropy": 5.43999285697937,
      "epoch": 1.1093976777939043,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004883675172970608,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19383164942264558,
      "num_tokens": 25379880.0,
      "step": 12230
    },
    {
      "entropy": 5.533678007125855,
      "epoch": 1.1098512336719883,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488357245968363,
      "loss": 5.2312,
      "mean_token_accuracy": 0.18909236490726472,
      "num_tokens": 25389942.0,
      "step": 12235
    },
    {
      "entropy": 5.496716022491455,
      "epoch": 1.1103047895500726,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004883469702273829,
      "loss": 5.045,
      "mean_token_accuracy": 0.19642258435487747,
      "num_tokens": 25400836.0,
      "step": 12240
    },
    {
      "entropy": 5.479660177230835,
      "epoch": 1.1107583454281567,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004883366900743331,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20178468227386476,
      "num_tokens": 25410502.0,
      "step": 12245
    },
    {
      "entropy": 5.4430962085723875,
      "epoch": 1.111211901306241,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004883264055094262,
      "loss": 5.0206,
      "mean_token_accuracy": 0.20242179334163665,
      "num_tokens": 25420301.0,
      "step": 12250
    },
    {
      "entropy": 5.464849042892456,
      "epoch": 1.1116654571843252,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004883161165328749,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19221176207065582,
      "num_tokens": 25431107.0,
      "step": 12255
    },
    {
      "entropy": 5.491602897644043,
      "epoch": 1.1121190130624092,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000488305823144892,
      "loss": 5.0869,
      "mean_token_accuracy": 0.1981050953269005,
      "num_tokens": 25441256.0,
      "step": 12260
    },
    {
      "entropy": 5.526266479492188,
      "epoch": 1.1125725689404935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004882955253456902,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20367671996355058,
      "num_tokens": 25451627.0,
      "step": 12265
    },
    {
      "entropy": 5.511344575881958,
      "epoch": 1.1130261248185778,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004882852231354826,
      "loss": 5.1622,
      "mean_token_accuracy": 0.18984195441007615,
      "num_tokens": 25461273.0,
      "step": 12270
    },
    {
      "entropy": 5.3871073722839355,
      "epoch": 1.1134796806966618,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004882749165144822,
      "loss": 5.0437,
      "mean_token_accuracy": 0.2021900862455368,
      "num_tokens": 25471949.0,
      "step": 12275
    },
    {
      "entropy": 5.513221025466919,
      "epoch": 1.113933236574746,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004882646054829023,
      "loss": 5.0712,
      "mean_token_accuracy": 0.19517802745103835,
      "num_tokens": 25482175.0,
      "step": 12280
    },
    {
      "entropy": 5.539937973022461,
      "epoch": 1.1143867924528301,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048825429004095597,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1953285738825798,
      "num_tokens": 25491786.0,
      "step": 12285
    },
    {
      "entropy": 5.4934309959411625,
      "epoch": 1.1148403483309144,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004882439701888566,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2028183802962303,
      "num_tokens": 25502103.0,
      "step": 12290
    },
    {
      "entropy": 5.489936685562133,
      "epoch": 1.1152939042089987,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048823364592681754,
      "loss": 5.0345,
      "mean_token_accuracy": 0.20078462064266206,
      "num_tokens": 25512530.0,
      "step": 12295
    },
    {
      "entropy": 5.453662967681884,
      "epoch": 1.1157474600870827,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048822331725505234,
      "loss": 5.0326,
      "mean_token_accuracy": 0.1963891550898552,
      "num_tokens": 25522905.0,
      "step": 12300
    },
    {
      "entropy": 5.455841970443726,
      "epoch": 1.116201015965167,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048821298417377456,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2051209181547165,
      "num_tokens": 25532963.0,
      "step": 12305
    },
    {
      "entropy": 5.400241136550903,
      "epoch": 1.116654571843251,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000488202646683198,
      "loss": 5.0859,
      "mean_token_accuracy": 0.19909514039754866,
      "num_tokens": 25542583.0,
      "step": 12310
    },
    {
      "entropy": 5.4752288341522215,
      "epoch": 1.1171081277213353,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004881923047835364,
      "loss": 5.0598,
      "mean_token_accuracy": 0.19813199937343598,
      "num_tokens": 25553367.0,
      "step": 12315
    },
    {
      "entropy": 5.614972066879273,
      "epoch": 1.1175616835994195,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004881819584750036,
      "loss": 5.158,
      "mean_token_accuracy": 0.18798512667417527,
      "num_tokens": 25564823.0,
      "step": 12320
    },
    {
      "entropy": 5.5026294708251955,
      "epoch": 1.1180152394775036,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048817160775781356,
      "loss": 5.1412,
      "mean_token_accuracy": 0.19281987398862838,
      "num_tokens": 25574946.0,
      "step": 12325
    },
    {
      "entropy": 5.5025474548339846,
      "epoch": 1.1184687953555879,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048816125263218023,
      "loss": 5.1247,
      "mean_token_accuracy": 0.19272203445434571,
      "num_tokens": 25584779.0,
      "step": 12330
    },
    {
      "entropy": 5.55981478691101,
      "epoch": 1.118922351233672,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048815089309831803,
      "loss": 5.1748,
      "mean_token_accuracy": 0.1895157814025879,
      "num_tokens": 25594725.0,
      "step": 12335
    },
    {
      "entropy": 5.490996026992798,
      "epoch": 1.1193759071117562,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488140529156441,
      "loss": 5.1544,
      "mean_token_accuracy": 0.19034090340137483,
      "num_tokens": 25605499.0,
      "step": 12340
    },
    {
      "entropy": 5.4678388118743895,
      "epoch": 1.1198294629898404,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004881301608067635,
      "loss": 5.099,
      "mean_token_accuracy": 0.19867289513349534,
      "num_tokens": 25616443.0,
      "step": 12345
    },
    {
      "entropy": 5.508192157745361,
      "epoch": 1.1202830188679245,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048811978804949997,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19435406923294068,
      "num_tokens": 25627762.0,
      "step": 12350
    },
    {
      "entropy": 5.572100734710693,
      "epoch": 1.1207365747460087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000488109410884865,
      "loss": 5.1585,
      "mean_token_accuracy": 0.19211570024490357,
      "num_tokens": 25637904.0,
      "step": 12355
    },
    {
      "entropy": 5.481449365615845,
      "epoch": 1.1211901306240928,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048809902931307294,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20028771758079528,
      "num_tokens": 25649058.0,
      "step": 12360
    },
    {
      "entropy": 5.452664804458618,
      "epoch": 1.121643686502177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004880886433343388,
      "loss": 5.0991,
      "mean_token_accuracy": 0.19796349555253984,
      "num_tokens": 25659161.0,
      "step": 12365
    },
    {
      "entropy": 5.456683731079101,
      "epoch": 1.1220972423802613,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004880782529488771,
      "loss": 4.9979,
      "mean_token_accuracy": 0.20418920665979384,
      "num_tokens": 25669336.0,
      "step": 12370
    },
    {
      "entropy": 5.531540966033935,
      "epoch": 1.1225507982583454,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004880678581569029,
      "loss": 5.1348,
      "mean_token_accuracy": 0.19526318609714508,
      "num_tokens": 25680443.0,
      "step": 12375
    },
    {
      "entropy": 5.457948493957519,
      "epoch": 1.1230043541364296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048805745895863105,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1991736814379692,
      "num_tokens": 25690202.0,
      "step": 12380
    },
    {
      "entropy": 5.457512521743775,
      "epoch": 1.1234579100145137,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004880470553542767,
      "loss": 5.1085,
      "mean_token_accuracy": 0.1963775172829628,
      "num_tokens": 25699255.0,
      "step": 12385
    },
    {
      "entropy": 5.506438302993774,
      "epoch": 1.123911465892598,
      "grad_norm": 0.875,
      "learning_rate": 0.00048803664734405495,
      "loss": 5.0875,
      "mean_token_accuracy": 0.18788405805826186,
      "num_tokens": 25710714.0,
      "step": 12390
    },
    {
      "entropy": 5.501832008361816,
      "epoch": 1.1243650217706822,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488026234928181,
      "loss": 5.0692,
      "mean_token_accuracy": 0.1955667182803154,
      "num_tokens": 25721722.0,
      "step": 12395
    },
    {
      "entropy": 5.461233329772949,
      "epoch": 1.1248185776487662,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004880158181068703,
      "loss": 5.0539,
      "mean_token_accuracy": 0.20558907687664033,
      "num_tokens": 25732096.0,
      "step": 12400
    },
    {
      "entropy": 5.483524560928345,
      "epoch": 1.1252721335268505,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048800539688033807,
      "loss": 5.1225,
      "mean_token_accuracy": 0.19797305017709732,
      "num_tokens": 25741499.0,
      "step": 12405
    },
    {
      "entropy": 5.534729719161987,
      "epoch": 1.1257256894049348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004879949712488001,
      "loss": 5.1901,
      "mean_token_accuracy": 0.1946818009018898,
      "num_tokens": 25751913.0,
      "step": 12410
    },
    {
      "entropy": 5.53972749710083,
      "epoch": 1.1261792452830188,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048798454121247177,
      "loss": 5.131,
      "mean_token_accuracy": 0.19206802695989608,
      "num_tokens": 25762602.0,
      "step": 12415
    },
    {
      "entropy": 5.440932893753052,
      "epoch": 1.126632801161103,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004879741067715688,
      "loss": 5.0934,
      "mean_token_accuracy": 0.19573008865118027,
      "num_tokens": 25773872.0,
      "step": 12420
    },
    {
      "entropy": 5.512517929077148,
      "epoch": 1.1270863570391871,
      "grad_norm": 1.0,
      "learning_rate": 0.00048796366792630715,
      "loss": 5.0935,
      "mean_token_accuracy": 0.19780833125114441,
      "num_tokens": 25784723.0,
      "step": 12425
    },
    {
      "entropy": 5.485009574890137,
      "epoch": 1.1275399129172714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004879532246769025,
      "loss": 5.1044,
      "mean_token_accuracy": 0.19494618028402327,
      "num_tokens": 25794897.0,
      "step": 12430
    },
    {
      "entropy": 5.4504047393798825,
      "epoch": 1.1279934687953557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004879427770235709,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20593176931142806,
      "num_tokens": 25804445.0,
      "step": 12435
    },
    {
      "entropy": 5.5492393493652346,
      "epoch": 1.1284470246734397,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048793232496652846,
      "loss": 5.1673,
      "mean_token_accuracy": 0.1864362522959709,
      "num_tokens": 25815495.0,
      "step": 12440
    },
    {
      "entropy": 5.485481834411621,
      "epoch": 1.128900580551524,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048792186850599124,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20277364999055864,
      "num_tokens": 25826120.0,
      "step": 12445
    },
    {
      "entropy": 5.441395616531372,
      "epoch": 1.1293541364296082,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004879114076421755,
      "loss": 5.0743,
      "mean_token_accuracy": 0.2006630629301071,
      "num_tokens": 25835926.0,
      "step": 12450
    },
    {
      "entropy": 5.51393871307373,
      "epoch": 1.1298076923076923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048790094237529763,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19224480241537095,
      "num_tokens": 25846095.0,
      "step": 12455
    },
    {
      "entropy": 5.568841123580933,
      "epoch": 1.1302612481857766,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487890472705574,
      "loss": 5.1286,
      "mean_token_accuracy": 0.19136714339256286,
      "num_tokens": 25858541.0,
      "step": 12460
    },
    {
      "entropy": 5.4886997699737545,
      "epoch": 1.1307148040638606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048787999863322113,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2051388591527939,
      "num_tokens": 25867664.0,
      "step": 12465
    },
    {
      "entropy": 5.474790382385254,
      "epoch": 1.1311683599419449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004878695201584557,
      "loss": 5.1188,
      "mean_token_accuracy": 0.19833892285823823,
      "num_tokens": 25877300.0,
      "step": 12470
    },
    {
      "entropy": 5.444518423080444,
      "epoch": 1.1316219158200291,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004878590372814943,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20307615250349045,
      "num_tokens": 25888009.0,
      "step": 12475
    },
    {
      "entropy": 5.532946348190308,
      "epoch": 1.1320754716981132,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048784855000255374,
      "loss": 5.1264,
      "mean_token_accuracy": 0.18937529623508453,
      "num_tokens": 25897285.0,
      "step": 12480
    },
    {
      "entropy": 5.478674602508545,
      "epoch": 1.1325290275761974,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048783805832185097,
      "loss": 5.0157,
      "mean_token_accuracy": 0.201152403652668,
      "num_tokens": 25907355.0,
      "step": 12485
    },
    {
      "entropy": 5.494596529006958,
      "epoch": 1.1329825834542815,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004878275622396029,
      "loss": 5.0919,
      "mean_token_accuracy": 0.20091527253389357,
      "num_tokens": 25918566.0,
      "step": 12490
    },
    {
      "entropy": 5.486082220077515,
      "epoch": 1.1334361393323658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048781706175602665,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19996363818645477,
      "num_tokens": 25929145.0,
      "step": 12495
    },
    {
      "entropy": 5.460976219177246,
      "epoch": 1.13388969521045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048780655687133927,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20151668787002563,
      "num_tokens": 25938163.0,
      "step": 12500
    },
    {
      "entropy": 5.457347440719604,
      "epoch": 1.134343251088534,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048779604758575806,
      "loss": 5.046,
      "mean_token_accuracy": 0.19633741229772567,
      "num_tokens": 25948962.0,
      "step": 12505
    },
    {
      "entropy": 5.419883298873901,
      "epoch": 1.1347968069666183,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004877855338995004,
      "loss": 5.033,
      "mean_token_accuracy": 0.19193218201398848,
      "num_tokens": 25959653.0,
      "step": 12510
    },
    {
      "entropy": 5.5417643070220945,
      "epoch": 1.1352503628447024,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048777501581278367,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19666285663843155,
      "num_tokens": 25970268.0,
      "step": 12515
    },
    {
      "entropy": 5.472268581390381,
      "epoch": 1.1357039187227866,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004877644933258254,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1935422420501709,
      "num_tokens": 25980321.0,
      "step": 12520
    },
    {
      "entropy": 5.421860313415527,
      "epoch": 1.136157474600871,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004877539664388432,
      "loss": 5.0697,
      "mean_token_accuracy": 0.19183527082204818,
      "num_tokens": 25992466.0,
      "step": 12525
    },
    {
      "entropy": 5.475657939910889,
      "epoch": 1.136611030478955,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048774343515205464,
      "loss": 5.0452,
      "mean_token_accuracy": 0.2010459080338478,
      "num_tokens": 26003553.0,
      "step": 12530
    },
    {
      "entropy": 5.488823318481446,
      "epoch": 1.1370645863570392,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004877328994656778,
      "loss": 5.0703,
      "mean_token_accuracy": 0.19986444860696792,
      "num_tokens": 26013299.0,
      "step": 12535
    },
    {
      "entropy": 5.454457712173462,
      "epoch": 1.1375181422351233,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004877223593799303,
      "loss": 5.0957,
      "mean_token_accuracy": 0.19835387021303177,
      "num_tokens": 26022913.0,
      "step": 12540
    },
    {
      "entropy": 5.421596193313599,
      "epoch": 1.1379716981132075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004877118148950302,
      "loss": 5.0688,
      "mean_token_accuracy": 0.19932219684123992,
      "num_tokens": 26033191.0,
      "step": 12545
    },
    {
      "entropy": 5.6301860332489015,
      "epoch": 1.1384252539912918,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048770126601119555,
      "loss": 5.2148,
      "mean_token_accuracy": 0.18730915635824202,
      "num_tokens": 26043517.0,
      "step": 12550
    },
    {
      "entropy": 5.577491188049317,
      "epoch": 1.1388788098693758,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048769071272864454,
      "loss": 5.1624,
      "mean_token_accuracy": 0.19309705942869188,
      "num_tokens": 26054173.0,
      "step": 12555
    },
    {
      "entropy": 5.546094846725464,
      "epoch": 1.13933236574746,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048768015504759536,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1877343937754631,
      "num_tokens": 26064929.0,
      "step": 12560
    },
    {
      "entropy": 5.445685005187988,
      "epoch": 1.1397859216255442,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004876695929682665,
      "loss": 5.0438,
      "mean_token_accuracy": 0.200204998254776,
      "num_tokens": 26074872.0,
      "step": 12565
    },
    {
      "entropy": 5.451004409790039,
      "epoch": 1.1402394775036284,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048765902649087605,
      "loss": 4.9624,
      "mean_token_accuracy": 0.20139967799186706,
      "num_tokens": 26086328.0,
      "step": 12570
    },
    {
      "entropy": 5.524568271636963,
      "epoch": 1.1406930333817127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048764845561564285,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19644412994384766,
      "num_tokens": 26096477.0,
      "step": 12575
    },
    {
      "entropy": 5.468381452560425,
      "epoch": 1.1411465892597967,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048763788034278545,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20553031712770461,
      "num_tokens": 26106183.0,
      "step": 12580
    },
    {
      "entropy": 5.454225015640259,
      "epoch": 1.141600145137881,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004876273006725225,
      "loss": 5.084,
      "mean_token_accuracy": 0.1978095680475235,
      "num_tokens": 26117932.0,
      "step": 12585
    },
    {
      "entropy": 5.446175527572632,
      "epoch": 1.1420537010159653,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048761671660507266,
      "loss": 5.0421,
      "mean_token_accuracy": 0.20245159715414046,
      "num_tokens": 26128616.0,
      "step": 12590
    },
    {
      "entropy": 5.502018308639526,
      "epoch": 1.1425072568940493,
      "grad_norm": 0.953125,
      "learning_rate": 0.000487606128140655,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19867226481437683,
      "num_tokens": 26138758.0,
      "step": 12595
    },
    {
      "entropy": 5.4364453792572025,
      "epoch": 1.1429608127721336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004875955352794885,
      "loss": 5.065,
      "mean_token_accuracy": 0.19398841261863708,
      "num_tokens": 26149199.0,
      "step": 12600
    },
    {
      "entropy": 5.430867147445679,
      "epoch": 1.1434143686502176,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004875849380217921,
      "loss": 5.1569,
      "mean_token_accuracy": 0.19733918458223343,
      "num_tokens": 26160276.0,
      "step": 12605
    },
    {
      "entropy": 5.500235271453858,
      "epoch": 1.1438679245283019,
      "grad_norm": 1.046875,
      "learning_rate": 0.000487574336367785,
      "loss": 5.0283,
      "mean_token_accuracy": 0.19995264261960982,
      "num_tokens": 26169564.0,
      "step": 12610
    },
    {
      "entropy": 5.42881498336792,
      "epoch": 1.1443214804063861,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004875637303176865,
      "loss": 5.0007,
      "mean_token_accuracy": 0.20410393923521042,
      "num_tokens": 26179135.0,
      "step": 12615
    },
    {
      "entropy": 5.4993397235870365,
      "epoch": 1.1447750362844702,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048755311987171585,
      "loss": 5.1338,
      "mean_token_accuracy": 0.19475421607494353,
      "num_tokens": 26190116.0,
      "step": 12620
    },
    {
      "entropy": 5.511917066574097,
      "epoch": 1.1452285921625545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048754250503009244,
      "loss": 5.1992,
      "mean_token_accuracy": 0.18957407921552658,
      "num_tokens": 26200847.0,
      "step": 12625
    },
    {
      "entropy": 5.599766254425049,
      "epoch": 1.1456821480406387,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048753188579303596,
      "loss": 5.155,
      "mean_token_accuracy": 0.19580694288015366,
      "num_tokens": 26211151.0,
      "step": 12630
    },
    {
      "entropy": 5.524540090560913,
      "epoch": 1.1461357039187228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004875212621607659,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2049766629934311,
      "num_tokens": 26221273.0,
      "step": 12635
    },
    {
      "entropy": 5.4686830043792725,
      "epoch": 1.146589259796807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048751063413350203,
      "loss": 5.1737,
      "mean_token_accuracy": 0.19148514568805694,
      "num_tokens": 26231770.0,
      "step": 12640
    },
    {
      "entropy": 5.545756721496582,
      "epoch": 1.147042815674891,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000487500001711464,
      "loss": 5.1424,
      "mean_token_accuracy": 0.18901480883359909,
      "num_tokens": 26241386.0,
      "step": 12645
    },
    {
      "entropy": 5.642010736465454,
      "epoch": 1.1474963715529753,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004874893648948719,
      "loss": 5.2114,
      "mean_token_accuracy": 0.19006966054439545,
      "num_tokens": 26251991.0,
      "step": 12650
    },
    {
      "entropy": 5.627796792984009,
      "epoch": 1.1479499274310596,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048747872368394546,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19527844488620758,
      "num_tokens": 26262331.0,
      "step": 12655
    },
    {
      "entropy": 5.561524677276611,
      "epoch": 1.1484034833091437,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048746807807890496,
      "loss": 5.1325,
      "mean_token_accuracy": 0.19464536160230636,
      "num_tokens": 26272939.0,
      "step": 12660
    },
    {
      "entropy": 5.510550403594971,
      "epoch": 1.148857039187228,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048745742807997046,
      "loss": 5.0807,
      "mean_token_accuracy": 0.2004735991358757,
      "num_tokens": 26282849.0,
      "step": 12665
    },
    {
      "entropy": 5.467938804626465,
      "epoch": 1.149310595065312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048744677368736216,
      "loss": 5.0885,
      "mean_token_accuracy": 0.2036068171262741,
      "num_tokens": 26292449.0,
      "step": 12670
    },
    {
      "entropy": 5.4095618724823,
      "epoch": 1.1497641509433962,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048743611490130044,
      "loss": 5.034,
      "mean_token_accuracy": 0.19792603999376296,
      "num_tokens": 26302506.0,
      "step": 12675
    },
    {
      "entropy": 5.50772762298584,
      "epoch": 1.1502177068214805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048742545172200576,
      "loss": 5.1061,
      "mean_token_accuracy": 0.2004696920514107,
      "num_tokens": 26312346.0,
      "step": 12680
    },
    {
      "entropy": 5.542489147186279,
      "epoch": 1.1506712626995645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004874147841496986,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19783084094524384,
      "num_tokens": 26323173.0,
      "step": 12685
    },
    {
      "entropy": 5.547864484786987,
      "epoch": 1.1511248185776488,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004874041121845996,
      "loss": 5.1629,
      "mean_token_accuracy": 0.187813962996006,
      "num_tokens": 26334371.0,
      "step": 12690
    },
    {
      "entropy": 5.475184345245362,
      "epoch": 1.1515783744557329,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004873934358269294,
      "loss": 5.0875,
      "mean_token_accuracy": 0.1992534339427948,
      "num_tokens": 26344506.0,
      "step": 12695
    },
    {
      "entropy": 5.564321422576905,
      "epoch": 1.1520319303338171,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004873827550769088,
      "loss": 5.1226,
      "mean_token_accuracy": 0.1877172037959099,
      "num_tokens": 26354534.0,
      "step": 12700
    },
    {
      "entropy": 5.496835136413575,
      "epoch": 1.1524854862119014,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004873720699347587,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19809603989124297,
      "num_tokens": 26365682.0,
      "step": 12705
    },
    {
      "entropy": 5.570609045028687,
      "epoch": 1.1529390420899854,
      "grad_norm": 1.0,
      "learning_rate": 0.0004873613804007001,
      "loss": 5.1936,
      "mean_token_accuracy": 0.19731561243534088,
      "num_tokens": 26376702.0,
      "step": 12710
    },
    {
      "entropy": 5.4602783679962155,
      "epoch": 1.1533925979680697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004873506864749541,
      "loss": 5.0067,
      "mean_token_accuracy": 0.19946054369211197,
      "num_tokens": 26387154.0,
      "step": 12715
    },
    {
      "entropy": 5.552713298797608,
      "epoch": 1.1538461538461537,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004873399881577417,
      "loss": 5.1788,
      "mean_token_accuracy": 0.19176933765411378,
      "num_tokens": 26397883.0,
      "step": 12720
    },
    {
      "entropy": 5.427668762207031,
      "epoch": 1.154299709724238,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004873292854492842,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20469096153974534,
      "num_tokens": 26407613.0,
      "step": 12725
    },
    {
      "entropy": 5.396756219863891,
      "epoch": 1.1547532656023223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004873185783498031,
      "loss": 5.0023,
      "mean_token_accuracy": 0.20199891477823256,
      "num_tokens": 26417221.0,
      "step": 12730
    },
    {
      "entropy": 5.4925392150878904,
      "epoch": 1.1552068214804063,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004873078668595197,
      "loss": 5.0976,
      "mean_token_accuracy": 0.19460710734128953,
      "num_tokens": 26427976.0,
      "step": 12735
    },
    {
      "entropy": 5.484690761566162,
      "epoch": 1.1556603773584906,
      "grad_norm": 1.0,
      "learning_rate": 0.00048729715097865534,
      "loss": 5.1287,
      "mean_token_accuracy": 0.19827549159526825,
      "num_tokens": 26439213.0,
      "step": 12740
    },
    {
      "entropy": 5.478594827651977,
      "epoch": 1.1561139332365746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004872864307074319,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19928483068943023,
      "num_tokens": 26449763.0,
      "step": 12745
    },
    {
      "entropy": 5.546472692489624,
      "epoch": 1.156567489114659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048727570604607096,
      "loss": 5.1252,
      "mean_token_accuracy": 0.19816723614931106,
      "num_tokens": 26460133.0,
      "step": 12750
    },
    {
      "entropy": 5.5383056640625,
      "epoch": 1.1570210449927432,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048726497699479437,
      "loss": 5.0922,
      "mean_token_accuracy": 0.19691014140844346,
      "num_tokens": 26470164.0,
      "step": 12755
    },
    {
      "entropy": 5.475792503356933,
      "epoch": 1.1574746008708272,
      "grad_norm": 0.921875,
      "learning_rate": 0.000487254243553824,
      "loss": 5.1044,
      "mean_token_accuracy": 0.19779773503541948,
      "num_tokens": 26480314.0,
      "step": 12760
    },
    {
      "entropy": 5.404000902175904,
      "epoch": 1.1579281567489115,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004872435057233817,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20014721006155015,
      "num_tokens": 26491184.0,
      "step": 12765
    },
    {
      "entropy": 5.449006414413452,
      "epoch": 1.1583817126269957,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048723276350368957,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20397381633520126,
      "num_tokens": 26501432.0,
      "step": 12770
    },
    {
      "entropy": 5.49757866859436,
      "epoch": 1.1588352685050798,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048722201689496984,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19721294641494752,
      "num_tokens": 26510379.0,
      "step": 12775
    },
    {
      "entropy": 5.46393346786499,
      "epoch": 1.159288824383164,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004872112658974447,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20006239414215088,
      "num_tokens": 26519454.0,
      "step": 12780
    },
    {
      "entropy": 5.439549779891967,
      "epoch": 1.159742380261248,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048720051051133657,
      "loss": 4.9887,
      "mean_token_accuracy": 0.19962143003940583,
      "num_tokens": 26531325.0,
      "step": 12785
    },
    {
      "entropy": 5.460226392745971,
      "epoch": 1.1601959361393324,
      "grad_norm": 1.0,
      "learning_rate": 0.0004871897507368678,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19462902396917342,
      "num_tokens": 26541510.0,
      "step": 12790
    },
    {
      "entropy": 5.522227096557617,
      "epoch": 1.1606494920174166,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048717898657426093,
      "loss": 5.2614,
      "mean_token_accuracy": 0.18738264143466948,
      "num_tokens": 26553400.0,
      "step": 12795
    },
    {
      "entropy": 5.558783531188965,
      "epoch": 1.1611030478955007,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048716821802373856,
      "loss": 5.0776,
      "mean_token_accuracy": 0.1960080683231354,
      "num_tokens": 26564340.0,
      "step": 12800
    },
    {
      "entropy": 5.512545728683472,
      "epoch": 1.161556603773585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048715744508552334,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19665155112743377,
      "num_tokens": 26573615.0,
      "step": 12805
    },
    {
      "entropy": 5.526737642288208,
      "epoch": 1.1620101596516692,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004871466677598381,
      "loss": 5.198,
      "mean_token_accuracy": 0.19085295498371124,
      "num_tokens": 26584149.0,
      "step": 12810
    },
    {
      "entropy": 5.475385999679565,
      "epoch": 1.1624637155297532,
      "grad_norm": 1.0,
      "learning_rate": 0.0004871358860469057,
      "loss": 5.1147,
      "mean_token_accuracy": 0.18644985556602478,
      "num_tokens": 26595064.0,
      "step": 12815
    },
    {
      "entropy": 5.487485313415528,
      "epoch": 1.1629172714078375,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048712509994694903,
      "loss": 5.1004,
      "mean_token_accuracy": 0.1979370340704918,
      "num_tokens": 26605636.0,
      "step": 12820
    },
    {
      "entropy": 5.573808574676514,
      "epoch": 1.1633708272859216,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004871143094601914,
      "loss": 5.196,
      "mean_token_accuracy": 0.19262153655290604,
      "num_tokens": 26616065.0,
      "step": 12825
    },
    {
      "entropy": 5.570623970031738,
      "epoch": 1.1638243831640058,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004871035145868557,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19735285639762878,
      "num_tokens": 26627246.0,
      "step": 12830
    },
    {
      "entropy": 5.500070810317993,
      "epoch": 1.16427793904209,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048709271532716524,
      "loss": 5.1282,
      "mean_token_accuracy": 0.19226828664541246,
      "num_tokens": 26637630.0,
      "step": 12835
    },
    {
      "entropy": 5.54427752494812,
      "epoch": 1.1647314949201741,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004870819116813434,
      "loss": 5.159,
      "mean_token_accuracy": 0.19256802797317504,
      "num_tokens": 26648683.0,
      "step": 12840
    },
    {
      "entropy": 5.546771287918091,
      "epoch": 1.1651850507982584,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048707110364961363,
      "loss": 5.1174,
      "mean_token_accuracy": 0.195330211520195,
      "num_tokens": 26659573.0,
      "step": 12845
    },
    {
      "entropy": 5.445801401138306,
      "epoch": 1.1656386066763424,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048706029123219937,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2026770994067192,
      "num_tokens": 26669384.0,
      "step": 12850
    },
    {
      "entropy": 5.5239941596984865,
      "epoch": 1.1660921625544267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048704947442932426,
      "loss": 5.195,
      "mean_token_accuracy": 0.1876979649066925,
      "num_tokens": 26679721.0,
      "step": 12855
    },
    {
      "entropy": 5.582709121704101,
      "epoch": 1.166545718432511,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048703865324121196,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20531719774007798,
      "num_tokens": 26689923.0,
      "step": 12860
    },
    {
      "entropy": 5.421995735168457,
      "epoch": 1.166999274310595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048702782766808637,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19754623025655746,
      "num_tokens": 26699953.0,
      "step": 12865
    },
    {
      "entropy": 5.512957429885864,
      "epoch": 1.1674528301886793,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004870169977101711,
      "loss": 5.2723,
      "mean_token_accuracy": 0.18300295621156693,
      "num_tokens": 26712099.0,
      "step": 12870
    },
    {
      "entropy": 5.5708049774169925,
      "epoch": 1.1679063860667633,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048700616336769037,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19754792302846907,
      "num_tokens": 26722334.0,
      "step": 12875
    },
    {
      "entropy": 5.538000631332397,
      "epoch": 1.1683599419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.00048699532464086817,
      "loss": 5.0668,
      "mean_token_accuracy": 0.19858947843313218,
      "num_tokens": 26732470.0,
      "step": 12880
    },
    {
      "entropy": 5.49172830581665,
      "epoch": 1.1688134978229319,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048698448152992865,
      "loss": 5.0952,
      "mean_token_accuracy": 0.19598378092050553,
      "num_tokens": 26742015.0,
      "step": 12885
    },
    {
      "entropy": 5.415465831756592,
      "epoch": 1.169267053701016,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048697363403509605,
      "loss": 5.1135,
      "mean_token_accuracy": 0.19333448261022568,
      "num_tokens": 26753231.0,
      "step": 12890
    },
    {
      "entropy": 5.515175199508667,
      "epoch": 1.1697206095791002,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004869627821565946,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1952577516436577,
      "num_tokens": 26763895.0,
      "step": 12895
    },
    {
      "entropy": 5.568731832504272,
      "epoch": 1.1701741654571842,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004869519258946489,
      "loss": 5.1751,
      "mean_token_accuracy": 0.19621602445840836,
      "num_tokens": 26774208.0,
      "step": 12900
    },
    {
      "entropy": 5.484986734390259,
      "epoch": 1.1706277213352685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048694106524948324,
      "loss": 5.1065,
      "mean_token_accuracy": 0.19364592730998992,
      "num_tokens": 26785554.0,
      "step": 12905
    },
    {
      "entropy": 5.521382761001587,
      "epoch": 1.1710812772133528,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048693020022132243,
      "loss": 5.0204,
      "mean_token_accuracy": 0.1997543454170227,
      "num_tokens": 26795488.0,
      "step": 12910
    },
    {
      "entropy": 5.552805805206299,
      "epoch": 1.1715348330914368,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000486919330810391,
      "loss": 5.1837,
      "mean_token_accuracy": 0.18431039452552794,
      "num_tokens": 26806801.0,
      "step": 12915
    },
    {
      "entropy": 5.510914707183838,
      "epoch": 1.171988388969521,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004869084570169139,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19960006475448608,
      "num_tokens": 26816869.0,
      "step": 12920
    },
    {
      "entropy": 5.481137228012085,
      "epoch": 1.172441944847605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048689757884111585,
      "loss": 5.0411,
      "mean_token_accuracy": 0.1998990297317505,
      "num_tokens": 26827003.0,
      "step": 12925
    },
    {
      "entropy": 5.433103561401367,
      "epoch": 1.1728955007256894,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048688669628322185,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19865425080060958,
      "num_tokens": 26837065.0,
      "step": 12930
    },
    {
      "entropy": 5.410325002670288,
      "epoch": 1.1733490566037736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000486875809343457,
      "loss": 4.9967,
      "mean_token_accuracy": 0.19845315665006638,
      "num_tokens": 26847487.0,
      "step": 12935
    },
    {
      "entropy": 5.475924348831176,
      "epoch": 1.1738026124818577,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048686491802204645,
      "loss": 5.0499,
      "mean_token_accuracy": 0.2035646066069603,
      "num_tokens": 26858407.0,
      "step": 12940
    },
    {
      "entropy": 5.54250168800354,
      "epoch": 1.174256168359942,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004868540223192154,
      "loss": 5.1214,
      "mean_token_accuracy": 0.19381057024002074,
      "num_tokens": 26868469.0,
      "step": 12945
    },
    {
      "entropy": 5.530332469940186,
      "epoch": 1.1747097242380262,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004868431222351892,
      "loss": 5.1229,
      "mean_token_accuracy": 0.19985790103673934,
      "num_tokens": 26879580.0,
      "step": 12950
    },
    {
      "entropy": 5.5076759338378904,
      "epoch": 1.1751632801161103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048683221777019313,
      "loss": 5.0943,
      "mean_token_accuracy": 0.2020096644759178,
      "num_tokens": 26889513.0,
      "step": 12955
    },
    {
      "entropy": 5.486361122131347,
      "epoch": 1.1756168359941945,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048682130892445293,
      "loss": 5.1776,
      "mean_token_accuracy": 0.19359630793333055,
      "num_tokens": 26900792.0,
      "step": 12960
    },
    {
      "entropy": 5.453381633758545,
      "epoch": 1.1760703918722786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000486810395698194,
      "loss": 5.0583,
      "mean_token_accuracy": 0.2014825612306595,
      "num_tokens": 26911298.0,
      "step": 12965
    },
    {
      "entropy": 5.533105802536011,
      "epoch": 1.1765239477503628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048679947809164223,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19168972223997116,
      "num_tokens": 26921545.0,
      "step": 12970
    },
    {
      "entropy": 5.528624534606934,
      "epoch": 1.176977503628447,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004867885561050232,
      "loss": 5.08,
      "mean_token_accuracy": 0.20872305780649186,
      "num_tokens": 26932140.0,
      "step": 12975
    },
    {
      "entropy": 5.496430492401123,
      "epoch": 1.1774310595065312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867776297385629,
      "loss": 5.1293,
      "mean_token_accuracy": 0.18912578672170638,
      "num_tokens": 26942634.0,
      "step": 12980
    },
    {
      "entropy": 5.506092357635498,
      "epoch": 1.1778846153846154,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048676669899248723,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20136663019657136,
      "num_tokens": 26952758.0,
      "step": 12985
    },
    {
      "entropy": 5.442593050003052,
      "epoch": 1.1783381712626997,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004867557638670222,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20214931666851044,
      "num_tokens": 26963439.0,
      "step": 12990
    },
    {
      "entropy": 5.461094951629638,
      "epoch": 1.1787917271407837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867448243623941,
      "loss": 5.099,
      "mean_token_accuracy": 0.19264838695526124,
      "num_tokens": 26974331.0,
      "step": 12995
    },
    {
      "entropy": 5.447110414505005,
      "epoch": 1.179245283018868,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000486733880478829,
      "loss": 4.967,
      "mean_token_accuracy": 0.2060053676366806,
      "num_tokens": 26983416.0,
      "step": 13000
    },
    {
      "entropy": 5.51430835723877,
      "epoch": 1.179698838896952,
      "grad_norm": 1.0,
      "learning_rate": 0.00048672293221655333,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19942551702260972,
      "num_tokens": 26994176.0,
      "step": 13005
    },
    {
      "entropy": 5.465344715118408,
      "epoch": 1.1801523947750363,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004867119795757935,
      "loss": 5.085,
      "mean_token_accuracy": 0.1964393898844719,
      "num_tokens": 27003295.0,
      "step": 13010
    },
    {
      "entropy": 5.476764297485351,
      "epoch": 1.1806059506531206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000486701022556776,
      "loss": 5.0424,
      "mean_token_accuracy": 0.19700005501508713,
      "num_tokens": 27014018.0,
      "step": 13015
    },
    {
      "entropy": 5.484636688232422,
      "epoch": 1.1810595065312046,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004866900611597273,
      "loss": 5.071,
      "mean_token_accuracy": 0.19935835003852845,
      "num_tokens": 27024706.0,
      "step": 13020
    },
    {
      "entropy": 5.532466125488281,
      "epoch": 1.1815130624092889,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048667909538487427,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20182140171527863,
      "num_tokens": 27035801.0,
      "step": 13025
    },
    {
      "entropy": 5.4090821743011475,
      "epoch": 1.181966618287373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004866681252324436,
      "loss": 4.9705,
      "mean_token_accuracy": 0.20396371632814408,
      "num_tokens": 27046596.0,
      "step": 13030
    },
    {
      "entropy": 5.454617643356324,
      "epoch": 1.1824201741654572,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048665715070266214,
      "loss": 5.0772,
      "mean_token_accuracy": 0.1967050701379776,
      "num_tokens": 27057057.0,
      "step": 13035
    },
    {
      "entropy": 5.402617454528809,
      "epoch": 1.1828737300435415,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048664617179575685,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20518202632665633,
      "num_tokens": 27067947.0,
      "step": 13040
    },
    {
      "entropy": 5.530779886245727,
      "epoch": 1.1833272859216255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004866351885119548,
      "loss": 5.0709,
      "mean_token_accuracy": 0.19731810688972473,
      "num_tokens": 27077503.0,
      "step": 13045
    },
    {
      "entropy": 5.535081672668457,
      "epoch": 1.1837808417997098,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004866242008514832,
      "loss": 5.1897,
      "mean_token_accuracy": 0.1931100070476532,
      "num_tokens": 27088298.0,
      "step": 13050
    },
    {
      "entropy": 5.475606679916382,
      "epoch": 1.1842343976777938,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048661320881456904,
      "loss": 5.0551,
      "mean_token_accuracy": 0.19839247316122055,
      "num_tokens": 27097841.0,
      "step": 13055
    },
    {
      "entropy": 5.406886243820191,
      "epoch": 1.184687953555878,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004866022124014399,
      "loss": 4.994,
      "mean_token_accuracy": 0.20290929526090623,
      "num_tokens": 27108263.0,
      "step": 13060
    },
    {
      "entropy": 5.4522699356079105,
      "epoch": 1.1851415094339623,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048659121161232313,
      "loss": 5.0589,
      "mean_token_accuracy": 0.20132856369018554,
      "num_tokens": 27119339.0,
      "step": 13065
    },
    {
      "entropy": 5.508952474594116,
      "epoch": 1.1855950653120464,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004865802064474461,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19557663500308992,
      "num_tokens": 27130393.0,
      "step": 13070
    },
    {
      "entropy": 5.534052133560181,
      "epoch": 1.1860486211901307,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048656919690703653,
      "loss": 5.1063,
      "mean_token_accuracy": 0.19433179795742034,
      "num_tokens": 27141156.0,
      "step": 13075
    },
    {
      "entropy": 5.542227840423584,
      "epoch": 1.1865021770682147,
      "grad_norm": 1.015625,
      "learning_rate": 0.000486558182991322,
      "loss": 5.1431,
      "mean_token_accuracy": 0.19479037076234818,
      "num_tokens": 27151630.0,
      "step": 13080
    },
    {
      "entropy": 5.474994468688965,
      "epoch": 1.186955732946299,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004865471647005304,
      "loss": 5.0169,
      "mean_token_accuracy": 0.20152200609445572,
      "num_tokens": 27160734.0,
      "step": 13085
    },
    {
      "entropy": 5.467565727233887,
      "epoch": 1.1874092888243832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048653614203488947,
      "loss": 5.0981,
      "mean_token_accuracy": 0.19816022962331772,
      "num_tokens": 27171584.0,
      "step": 13090
    },
    {
      "entropy": 5.464590835571289,
      "epoch": 1.1878628447024673,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004865251149946273,
      "loss": 5.005,
      "mean_token_accuracy": 0.20658984631299973,
      "num_tokens": 27180817.0,
      "step": 13095
    },
    {
      "entropy": 5.542662477493286,
      "epoch": 1.1883164005805515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048651408357997173,
      "loss": 5.0887,
      "mean_token_accuracy": 0.1919817328453064,
      "num_tokens": 27190363.0,
      "step": 13100
    },
    {
      "entropy": 5.467042398452759,
      "epoch": 1.1887699564586356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048650304779115114,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20326141864061356,
      "num_tokens": 27202181.0,
      "step": 13105
    },
    {
      "entropy": 5.43560643196106,
      "epoch": 1.1892235123367199,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004864920076283935,
      "loss": 5.0895,
      "mean_token_accuracy": 0.19626662284135818,
      "num_tokens": 27212514.0,
      "step": 13110
    },
    {
      "entropy": 5.550775957107544,
      "epoch": 1.1896770682148041,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004864809630919273,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20287249684333802,
      "num_tokens": 27223352.0,
      "step": 13115
    },
    {
      "entropy": 5.518157911300659,
      "epoch": 1.1901306240928882,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004864699141819809,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2069821238517761,
      "num_tokens": 27232643.0,
      "step": 13120
    },
    {
      "entropy": 5.41427264213562,
      "epoch": 1.1905841799709724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048645886089878276,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20254046469926834,
      "num_tokens": 27243793.0,
      "step": 13125
    },
    {
      "entropy": 5.4802384853363035,
      "epoch": 1.1910377358490567,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048644780324256153,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1999229371547699,
      "num_tokens": 27254554.0,
      "step": 13130
    },
    {
      "entropy": 5.487999629974365,
      "epoch": 1.1914912917271407,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004864367412135458,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19469751119613649,
      "num_tokens": 27265842.0,
      "step": 13135
    },
    {
      "entropy": 5.496365976333618,
      "epoch": 1.191944847605225,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048642567481196446,
      "loss": 5.1129,
      "mean_token_accuracy": 0.19370853155851364,
      "num_tokens": 27277617.0,
      "step": 13140
    },
    {
      "entropy": 5.44011173248291,
      "epoch": 1.192398403483309,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004864146040380462,
      "loss": 5.045,
      "mean_token_accuracy": 0.19978711754083633,
      "num_tokens": 27287450.0,
      "step": 13145
    },
    {
      "entropy": 5.47112717628479,
      "epoch": 1.1928519593613933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004864035288920202,
      "loss": 5.0671,
      "mean_token_accuracy": 0.20327106565237046,
      "num_tokens": 27297010.0,
      "step": 13150
    },
    {
      "entropy": 5.463698434829712,
      "epoch": 1.1933055152394776,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048639244937411516,
      "loss": 5.1043,
      "mean_token_accuracy": 0.19503520727157592,
      "num_tokens": 27306874.0,
      "step": 13155
    },
    {
      "entropy": 5.5497582912445065,
      "epoch": 1.1937590711175616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048638136548456045,
      "loss": 5.1019,
      "mean_token_accuracy": 0.19800865054130554,
      "num_tokens": 27318762.0,
      "step": 13160
    },
    {
      "entropy": 5.442694759368896,
      "epoch": 1.194212626995646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048637027722358536,
      "loss": 4.981,
      "mean_token_accuracy": 0.21250585168600084,
      "num_tokens": 27327231.0,
      "step": 13165
    },
    {
      "entropy": 5.385290479660034,
      "epoch": 1.1946661828737302,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048635918459141893,
      "loss": 5.0192,
      "mean_token_accuracy": 0.19913043081760406,
      "num_tokens": 27338517.0,
      "step": 13170
    },
    {
      "entropy": 5.528426170349121,
      "epoch": 1.1951197387518142,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048634808758829073,
      "loss": 5.0883,
      "mean_token_accuracy": 0.19603673368692398,
      "num_tokens": 27348861.0,
      "step": 13175
    },
    {
      "entropy": 5.511196565628052,
      "epoch": 1.1955732946298985,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048633698621443024,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20928379148244858,
      "num_tokens": 27358610.0,
      "step": 13180
    },
    {
      "entropy": 5.475197076797485,
      "epoch": 1.1960268505079825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000486325880470067,
      "loss": 5.1288,
      "mean_token_accuracy": 0.1985629081726074,
      "num_tokens": 27369707.0,
      "step": 13185
    },
    {
      "entropy": 5.4696228981018065,
      "epoch": 1.1964804063860668,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004863147703554307,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19705890119075775,
      "num_tokens": 27381030.0,
      "step": 13190
    },
    {
      "entropy": 5.519057321548462,
      "epoch": 1.196933962264151,
      "grad_norm": 1.0,
      "learning_rate": 0.00048630365587075115,
      "loss": 5.0961,
      "mean_token_accuracy": 0.19812251776456832,
      "num_tokens": 27390866.0,
      "step": 13195
    },
    {
      "entropy": 5.493912649154663,
      "epoch": 1.197387518142235,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048629253701625816,
      "loss": 5.0357,
      "mean_token_accuracy": 0.20332785397768022,
      "num_tokens": 27400724.0,
      "step": 13200
    },
    {
      "entropy": 5.473924732208252,
      "epoch": 1.1978410740203194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004862814137921816,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19959327727556228,
      "num_tokens": 27411212.0,
      "step": 13205
    },
    {
      "entropy": 5.467393922805786,
      "epoch": 1.1982946298984034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048627028619875157,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1994468241930008,
      "num_tokens": 27421279.0,
      "step": 13210
    },
    {
      "entropy": 5.483545017242432,
      "epoch": 1.1987481857764877,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004862591542361982,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19779787510633468,
      "num_tokens": 27432520.0,
      "step": 13215
    },
    {
      "entropy": 5.4180910110473635,
      "epoch": 1.199201741654572,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004862480179047516,
      "loss": 4.989,
      "mean_token_accuracy": 0.20875904560089112,
      "num_tokens": 27442246.0,
      "step": 13220
    },
    {
      "entropy": 5.482895421981811,
      "epoch": 1.199655297532656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004862368772046423,
      "loss": 5.0607,
      "mean_token_accuracy": 0.19732182919979097,
      "num_tokens": 27452134.0,
      "step": 13225
    },
    {
      "entropy": 5.496697950363159,
      "epoch": 1.2001088534107403,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048622573213610057,
      "loss": 5.1461,
      "mean_token_accuracy": 0.188792884349823,
      "num_tokens": 27462253.0,
      "step": 13230
    },
    {
      "entropy": 5.511988878250122,
      "epoch": 1.2005624092888243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048621458269935676,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20037435442209245,
      "num_tokens": 27473131.0,
      "step": 13235
    },
    {
      "entropy": 5.489395809173584,
      "epoch": 1.2010159651669086,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048620342889464153,
      "loss": 5.0756,
      "mean_token_accuracy": 0.2012531042098999,
      "num_tokens": 27483683.0,
      "step": 13240
    },
    {
      "entropy": 5.384080266952514,
      "epoch": 1.2014695210449928,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048619227072218565,
      "loss": 5.052,
      "mean_token_accuracy": 0.1951747328042984,
      "num_tokens": 27493679.0,
      "step": 13245
    },
    {
      "entropy": 5.4413477897644045,
      "epoch": 1.2019230769230769,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004861811081822198,
      "loss": 5.0351,
      "mean_token_accuracy": 0.20437403321266173,
      "num_tokens": 27503897.0,
      "step": 13250
    },
    {
      "entropy": 5.532030296325684,
      "epoch": 1.2023766328011611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004861699412749748,
      "loss": 5.1421,
      "mean_token_accuracy": 0.192557655274868,
      "num_tokens": 27514560.0,
      "step": 13255
    },
    {
      "entropy": 5.4998753547668455,
      "epoch": 1.2028301886792452,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004861587700006816,
      "loss": 5.0427,
      "mean_token_accuracy": 0.1978370800614357,
      "num_tokens": 27525594.0,
      "step": 13260
    },
    {
      "entropy": 5.426636695861816,
      "epoch": 1.2032837445573294,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048614759435957114,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2014489009976387,
      "num_tokens": 27535676.0,
      "step": 13265
    },
    {
      "entropy": 5.413556241989136,
      "epoch": 1.2037373004354137,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004861364143518747,
      "loss": 5.0281,
      "mean_token_accuracy": 0.19157828241586686,
      "num_tokens": 27545520.0,
      "step": 13270
    },
    {
      "entropy": 5.573331117630005,
      "epoch": 1.2041908563134978,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004861252299778234,
      "loss": 5.0973,
      "mean_token_accuracy": 0.18991344869136811,
      "num_tokens": 27556866.0,
      "step": 13275
    },
    {
      "entropy": 5.47489185333252,
      "epoch": 1.204644412191582,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048611404123764855,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19900879561901091,
      "num_tokens": 27566951.0,
      "step": 13280
    },
    {
      "entropy": 5.470123195648194,
      "epoch": 1.205097968069666,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048610284813158154,
      "loss": 5.1365,
      "mean_token_accuracy": 0.1957337498664856,
      "num_tokens": 27576796.0,
      "step": 13285
    },
    {
      "entropy": 5.402993488311767,
      "epoch": 1.2055515239477503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004860916506598538,
      "loss": 5.0034,
      "mean_token_accuracy": 0.1991758719086647,
      "num_tokens": 27586902.0,
      "step": 13290
    },
    {
      "entropy": 5.585123109817505,
      "epoch": 1.2060050798258346,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048608044882269695,
      "loss": 5.1957,
      "mean_token_accuracy": 0.18857692927122116,
      "num_tokens": 27599232.0,
      "step": 13295
    },
    {
      "entropy": 5.482767248153687,
      "epoch": 1.2064586357039186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004860692426203427,
      "loss": 4.9931,
      "mean_token_accuracy": 0.1983819544315338,
      "num_tokens": 27609022.0,
      "step": 13300
    },
    {
      "entropy": 5.515181207656861,
      "epoch": 1.206912191582003,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004860580320530226,
      "loss": 5.1371,
      "mean_token_accuracy": 0.19917869865894317,
      "num_tokens": 27619829.0,
      "step": 13305
    },
    {
      "entropy": 5.515312433242798,
      "epoch": 1.2073657474600872,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004860468171209687,
      "loss": 5.1298,
      "mean_token_accuracy": 0.20136113315820695,
      "num_tokens": 27629337.0,
      "step": 13310
    },
    {
      "entropy": 5.546958637237549,
      "epoch": 1.2078193033381712,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004860355978244128,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19766863137483598,
      "num_tokens": 27639377.0,
      "step": 13315
    },
    {
      "entropy": 5.500257062911987,
      "epoch": 1.2082728592162555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000486024374163587,
      "loss": 5.0404,
      "mean_token_accuracy": 0.19715555906295776,
      "num_tokens": 27649989.0,
      "step": 13320
    },
    {
      "entropy": 5.49214277267456,
      "epoch": 1.2087264150943395,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048601314613872334,
      "loss": 5.1191,
      "mean_token_accuracy": 0.19113418459892273,
      "num_tokens": 27659967.0,
      "step": 13325
    },
    {
      "entropy": 5.509144020080567,
      "epoch": 1.2091799709724238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000486001913750054,
      "loss": 5.0724,
      "mean_token_accuracy": 0.19962707608938218,
      "num_tokens": 27670099.0,
      "step": 13330
    },
    {
      "entropy": 5.463675832748413,
      "epoch": 1.209633526850508,
      "grad_norm": 1.0,
      "learning_rate": 0.0004859906769978113,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1901517853140831,
      "num_tokens": 27681497.0,
      "step": 13335
    },
    {
      "entropy": 5.572259140014649,
      "epoch": 1.2100870827285921,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048597943588222776,
      "loss": 5.2052,
      "mean_token_accuracy": 0.19154019355773927,
      "num_tokens": 27691828.0,
      "step": 13340
    },
    {
      "entropy": 5.538029193878174,
      "epoch": 1.2105406386066764,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048596819040353564,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20628093034029008,
      "num_tokens": 27701426.0,
      "step": 13345
    },
    {
      "entropy": 5.493690490722656,
      "epoch": 1.2109941944847606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859569405619676,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20149533450603485,
      "num_tokens": 27711112.0,
      "step": 13350
    },
    {
      "entropy": 5.52337875366211,
      "epoch": 1.2114477503628447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004859456863577562,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19508373439311982,
      "num_tokens": 27723160.0,
      "step": 13355
    },
    {
      "entropy": 5.522914218902588,
      "epoch": 1.211901306240929,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048593442779113437,
      "loss": 5.1034,
      "mean_token_accuracy": 0.19623059332370757,
      "num_tokens": 27733143.0,
      "step": 13360
    },
    {
      "entropy": 5.5026891231536865,
      "epoch": 1.212354862119013,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004859231648623348,
      "loss": 5.0884,
      "mean_token_accuracy": 0.19636708945035936,
      "num_tokens": 27743956.0,
      "step": 13365
    },
    {
      "entropy": 5.522414779663086,
      "epoch": 1.2128084179970973,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048591189757159035,
      "loss": 5.0851,
      "mean_token_accuracy": 0.2020998165011406,
      "num_tokens": 27754192.0,
      "step": 13370
    },
    {
      "entropy": 5.487301158905029,
      "epoch": 1.2132619738751815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004859006259191342,
      "loss": 5.1205,
      "mean_token_accuracy": 0.18899906873703004,
      "num_tokens": 27763939.0,
      "step": 13375
    },
    {
      "entropy": 5.485107564926148,
      "epoch": 1.2137155297532656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004858893499051993,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20167715549468995,
      "num_tokens": 27774743.0,
      "step": 13380
    },
    {
      "entropy": 5.508217763900757,
      "epoch": 1.2141690856313498,
      "grad_norm": 1.125,
      "learning_rate": 0.0004858780695300189,
      "loss": 5.0722,
      "mean_token_accuracy": 0.19787210524082183,
      "num_tokens": 27785344.0,
      "step": 13385
    },
    {
      "entropy": 5.520311784744263,
      "epoch": 1.2146226415094339,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004858667847938263,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19476209133863448,
      "num_tokens": 27797101.0,
      "step": 13390
    },
    {
      "entropy": 5.452220058441162,
      "epoch": 1.2150761973875182,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048585549569685486,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2015383094549179,
      "num_tokens": 27807581.0,
      "step": 13395
    },
    {
      "entropy": 5.462075805664062,
      "epoch": 1.2155297532656024,
      "grad_norm": 1.0,
      "learning_rate": 0.00048584420223933805,
      "loss": 5.0023,
      "mean_token_accuracy": 0.19964679926633835,
      "num_tokens": 27818170.0,
      "step": 13400
    },
    {
      "entropy": 5.432729625701905,
      "epoch": 1.2159833091436865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004858329044215094,
      "loss": 5.029,
      "mean_token_accuracy": 0.20274798572063446,
      "num_tokens": 27828153.0,
      "step": 13405
    },
    {
      "entropy": 5.487482118606567,
      "epoch": 1.2164368650217707,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048582160224360253,
      "loss": 4.9698,
      "mean_token_accuracy": 0.206896835565567,
      "num_tokens": 27837681.0,
      "step": 13410
    },
    {
      "entropy": 5.494462537765503,
      "epoch": 1.2168904208998548,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004858102957058513,
      "loss": 5.108,
      "mean_token_accuracy": 0.19647631645202637,
      "num_tokens": 27847447.0,
      "step": 13415
    },
    {
      "entropy": 5.444395112991333,
      "epoch": 1.217343976777939,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048579898480848933,
      "loss": 4.9716,
      "mean_token_accuracy": 0.2026344954967499,
      "num_tokens": 27858586.0,
      "step": 13420
    },
    {
      "entropy": 5.495455551147461,
      "epoch": 1.2177975326560233,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004857876695517507,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19661417752504348,
      "num_tokens": 27868832.0,
      "step": 13425
    },
    {
      "entropy": 5.492994785308838,
      "epoch": 1.2182510885341074,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048577634993586923,
      "loss": 5.0952,
      "mean_token_accuracy": 0.19540033489465714,
      "num_tokens": 27879339.0,
      "step": 13430
    },
    {
      "entropy": 5.515705108642578,
      "epoch": 1.2187046444121916,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048576502596107924,
      "loss": 5.0881,
      "mean_token_accuracy": 0.19840986430644988,
      "num_tokens": 27889964.0,
      "step": 13435
    },
    {
      "entropy": 5.5269997119903564,
      "epoch": 1.2191582002902757,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857536976276148,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19223980158567427,
      "num_tokens": 27901004.0,
      "step": 13440
    },
    {
      "entropy": 5.484214353561401,
      "epoch": 1.21961175616836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004857423649357102,
      "loss": 5.0613,
      "mean_token_accuracy": 0.1969600275158882,
      "num_tokens": 27910788.0,
      "step": 13445
    },
    {
      "entropy": 5.544168424606323,
      "epoch": 1.2200653120464442,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048573102788559974,
      "loss": 5.1035,
      "mean_token_accuracy": 0.2005355179309845,
      "num_tokens": 27921360.0,
      "step": 13450
    },
    {
      "entropy": 5.482368850708008,
      "epoch": 1.2205188679245282,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004857196864775179,
      "loss": 5.0467,
      "mean_token_accuracy": 0.20181792825460435,
      "num_tokens": 27931889.0,
      "step": 13455
    },
    {
      "entropy": 5.456076622009277,
      "epoch": 1.2209724238026125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048570834071169926,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20518947392702103,
      "num_tokens": 27942049.0,
      "step": 13460
    },
    {
      "entropy": 5.538109827041626,
      "epoch": 1.2214259796806965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048569699058837856,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1968417227268219,
      "num_tokens": 27952416.0,
      "step": 13465
    },
    {
      "entropy": 5.467518997192383,
      "epoch": 1.2218795355587808,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004856856361077903,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20182372778654098,
      "num_tokens": 27962009.0,
      "step": 13470
    },
    {
      "entropy": 5.550679159164429,
      "epoch": 1.222333091436865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004856742772701694,
      "loss": 5.0959,
      "mean_token_accuracy": 0.19606231898069382,
      "num_tokens": 27972175.0,
      "step": 13475
    },
    {
      "entropy": 5.5194611072540285,
      "epoch": 1.2227866473149491,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048566291407575073,
      "loss": 5.1141,
      "mean_token_accuracy": 0.2015317127108574,
      "num_tokens": 27982899.0,
      "step": 13480
    },
    {
      "entropy": 5.513904333114624,
      "epoch": 1.2232402031930334,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048565154652476937,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20250820070505143,
      "num_tokens": 27993231.0,
      "step": 13485
    },
    {
      "entropy": 5.432183265686035,
      "epoch": 1.2236937590711174,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004856401746174603,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20159731060266495,
      "num_tokens": 28003623.0,
      "step": 13490
    },
    {
      "entropy": 5.488606214523315,
      "epoch": 1.2241473149492017,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048562879835405874,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20178948640823363,
      "num_tokens": 28013878.0,
      "step": 13495
    },
    {
      "entropy": 5.447732400894165,
      "epoch": 1.224600870827286,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004856174177348,
      "loss": 5.0941,
      "mean_token_accuracy": 0.19857164174318315,
      "num_tokens": 28024224.0,
      "step": 13500
    },
    {
      "entropy": 5.470084142684937,
      "epoch": 1.22505442670537,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004856060327599194,
      "loss": 5.0161,
      "mean_token_accuracy": 0.1966568797826767,
      "num_tokens": 28034819.0,
      "step": 13505
    },
    {
      "entropy": 5.463940000534057,
      "epoch": 1.2255079825834543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048559464342965233,
      "loss": 5.0775,
      "mean_token_accuracy": 0.19706240743398667,
      "num_tokens": 28044240.0,
      "step": 13510
    },
    {
      "entropy": 5.565493249893189,
      "epoch": 1.2259615384615385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004855832497442344,
      "loss": 5.2113,
      "mean_token_accuracy": 0.18766228705644608,
      "num_tokens": 28055046.0,
      "step": 13515
    },
    {
      "entropy": 5.462039518356323,
      "epoch": 1.2264150943396226,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004855718517039012,
      "loss": 5.0052,
      "mean_token_accuracy": 0.205356565117836,
      "num_tokens": 28065272.0,
      "step": 13520
    },
    {
      "entropy": 5.517233610153198,
      "epoch": 1.2268686502177069,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048556044930888844,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2079099714756012,
      "num_tokens": 28076255.0,
      "step": 13525
    },
    {
      "entropy": 5.515420389175415,
      "epoch": 1.2273222060957911,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004855490425594319,
      "loss": 5.1358,
      "mean_token_accuracy": 0.19787602573633195,
      "num_tokens": 28086258.0,
      "step": 13530
    },
    {
      "entropy": 5.494659328460694,
      "epoch": 1.2277757619738752,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004855376314557675,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20258419066667557,
      "num_tokens": 28096497.0,
      "step": 13535
    },
    {
      "entropy": 5.510882329940796,
      "epoch": 1.2282293178519594,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004855262159981313,
      "loss": 5.0602,
      "mean_token_accuracy": 0.2041010618209839,
      "num_tokens": 28106282.0,
      "step": 13540
    },
    {
      "entropy": 5.450311946868896,
      "epoch": 1.2286828737300435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048551479618675925,
      "loss": 5.1113,
      "mean_token_accuracy": 0.19611413329839705,
      "num_tokens": 28115771.0,
      "step": 13545
    },
    {
      "entropy": 5.526232290267944,
      "epoch": 1.2291364296081277,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048550337202188756,
      "loss": 5.0407,
      "mean_token_accuracy": 0.2037350058555603,
      "num_tokens": 28125477.0,
      "step": 13550
    },
    {
      "entropy": 5.55634880065918,
      "epoch": 1.229589985486212,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004854919435037525,
      "loss": 5.0788,
      "mean_token_accuracy": 0.19526431262493132,
      "num_tokens": 28135800.0,
      "step": 13555
    },
    {
      "entropy": 5.429639768600464,
      "epoch": 1.230043541364296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004854805106325904,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2002708464860916,
      "num_tokens": 28146779.0,
      "step": 13560
    },
    {
      "entropy": 5.4886777877807615,
      "epoch": 1.2304970972423803,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004854690734086378,
      "loss": 5.1104,
      "mean_token_accuracy": 0.19459391683340072,
      "num_tokens": 28156819.0,
      "step": 13565
    },
    {
      "entropy": 5.470524597167969,
      "epoch": 1.2309506531204644,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004854576318321311,
      "loss": 5.0871,
      "mean_token_accuracy": 0.1952335223555565,
      "num_tokens": 28168553.0,
      "step": 13570
    },
    {
      "entropy": 5.530834293365478,
      "epoch": 1.2314042089985486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048544618590330685,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19884106814861296,
      "num_tokens": 28178064.0,
      "step": 13575
    },
    {
      "entropy": 5.565642642974853,
      "epoch": 1.231857764876633,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048543473562240194,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19776492416858674,
      "num_tokens": 28188580.0,
      "step": 13580
    },
    {
      "entropy": 5.524669981002807,
      "epoch": 1.232311320754717,
      "grad_norm": 0.984375,
      "learning_rate": 0.000485423280989653,
      "loss": 5.1629,
      "mean_token_accuracy": 0.19327968508005142,
      "num_tokens": 28199955.0,
      "step": 13585
    },
    {
      "entropy": 5.5645153522491455,
      "epoch": 1.2327648766328012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000485411822005297,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1941158130764961,
      "num_tokens": 28210506.0,
      "step": 13590
    },
    {
      "entropy": 5.493167734146118,
      "epoch": 1.2332184325108853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048540035866957097,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2134607493877411,
      "num_tokens": 28220486.0,
      "step": 13595
    },
    {
      "entropy": 5.521177005767822,
      "epoch": 1.2336719883889695,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004853888909827118,
      "loss": 5.1623,
      "mean_token_accuracy": 0.18835390061140062,
      "num_tokens": 28232668.0,
      "step": 13600
    },
    {
      "entropy": 5.507457447052002,
      "epoch": 1.2341255442670538,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853774189449568,
      "loss": 5.1072,
      "mean_token_accuracy": 0.1947883576154709,
      "num_tokens": 28243733.0,
      "step": 13605
    },
    {
      "entropy": 5.572214269638062,
      "epoch": 1.2345791001451378,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048536594255654315,
      "loss": 5.1413,
      "mean_token_accuracy": 0.19594713300466537,
      "num_tokens": 28254953.0,
      "step": 13610
    },
    {
      "entropy": 5.530012893676758,
      "epoch": 1.235032656023222,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004853544618177083,
      "loss": 5.0832,
      "mean_token_accuracy": 0.20127384960651398,
      "num_tokens": 28264959.0,
      "step": 13615
    },
    {
      "entropy": 5.54983115196228,
      "epoch": 1.2354862119013061,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048534297672868945,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19841146767139434,
      "num_tokens": 28275509.0,
      "step": 13620
    },
    {
      "entropy": 5.500857210159301,
      "epoch": 1.2359397677793904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048533148728972434,
      "loss": 5.0277,
      "mean_token_accuracy": 0.2020431414246559,
      "num_tokens": 28285983.0,
      "step": 13625
    },
    {
      "entropy": 5.510316038131714,
      "epoch": 1.2363933236574747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048531999350105037,
      "loss": 5.1071,
      "mean_token_accuracy": 0.1961878165602684,
      "num_tokens": 28296178.0,
      "step": 13630
    },
    {
      "entropy": 5.491313409805298,
      "epoch": 1.2368468795355587,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048530849536290537,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19448001086711883,
      "num_tokens": 28305741.0,
      "step": 13635
    },
    {
      "entropy": 5.42606258392334,
      "epoch": 1.237300435413643,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004852969928755271,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2011515825986862,
      "num_tokens": 28315607.0,
      "step": 13640
    },
    {
      "entropy": 5.420935916900635,
      "epoch": 1.237753991291727,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048528548603915344,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20115619599819184,
      "num_tokens": 28327349.0,
      "step": 13645
    },
    {
      "entropy": 5.52898759841919,
      "epoch": 1.2382075471698113,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048527397485402236,
      "loss": 5.0485,
      "mean_token_accuracy": 0.19680120646953583,
      "num_tokens": 28339949.0,
      "step": 13650
    },
    {
      "entropy": 5.542382383346558,
      "epoch": 1.2386611030478956,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048526245932037186,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20272745341062545,
      "num_tokens": 28351340.0,
      "step": 13655
    },
    {
      "entropy": 5.385624742507934,
      "epoch": 1.2391146589259796,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048525093943844013,
      "loss": 4.9409,
      "mean_token_accuracy": 0.20911214351654053,
      "num_tokens": 28360929.0,
      "step": 13660
    },
    {
      "entropy": 5.426797580718994,
      "epoch": 1.2395682148040639,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004852394152084654,
      "loss": 5.0391,
      "mean_token_accuracy": 0.2021891802549362,
      "num_tokens": 28370477.0,
      "step": 13665
    },
    {
      "entropy": 5.446661186218262,
      "epoch": 1.240021770682148,
      "grad_norm": 0.984375,
      "learning_rate": 0.000485227886630686,
      "loss": 4.966,
      "mean_token_accuracy": 0.21486925035715104,
      "num_tokens": 28380876.0,
      "step": 13670
    },
    {
      "entropy": 5.394935607910156,
      "epoch": 1.2404753265602322,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004852163537053403,
      "loss": 4.9434,
      "mean_token_accuracy": 0.20210244953632356,
      "num_tokens": 28390685.0,
      "step": 13675
    },
    {
      "entropy": 5.438671875,
      "epoch": 1.2409288824383164,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004852048164326669,
      "loss": 5.0539,
      "mean_token_accuracy": 0.19393928349018097,
      "num_tokens": 28400236.0,
      "step": 13680
    },
    {
      "entropy": 5.503456926345825,
      "epoch": 1.2413824383164005,
      "grad_norm": 1.125,
      "learning_rate": 0.0004851932748129043,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20332229733467103,
      "num_tokens": 28410353.0,
      "step": 13685
    },
    {
      "entropy": 5.507759523391724,
      "epoch": 1.2418359941944848,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004851817288462912,
      "loss": 5.0384,
      "mean_token_accuracy": 0.1971975803375244,
      "num_tokens": 28420241.0,
      "step": 13690
    },
    {
      "entropy": 5.471080160140991,
      "epoch": 1.242289550072569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004851701785330663,
      "loss": 5.0943,
      "mean_token_accuracy": 0.19966191947460174,
      "num_tokens": 28430610.0,
      "step": 13695
    },
    {
      "entropy": 5.473397445678711,
      "epoch": 1.242743105950653,
      "grad_norm": 1.0,
      "learning_rate": 0.0004851586238734687,
      "loss": 5.0184,
      "mean_token_accuracy": 0.19985776245594025,
      "num_tokens": 28440026.0,
      "step": 13700
    },
    {
      "entropy": 5.48423581123352,
      "epoch": 1.2431966618287373,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004851470648677371,
      "loss": 5.0913,
      "mean_token_accuracy": 0.1909073993563652,
      "num_tokens": 28450707.0,
      "step": 13705
    },
    {
      "entropy": 5.484880018234253,
      "epoch": 1.2436502177068216,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004851355015161107,
      "loss": 5.04,
      "mean_token_accuracy": 0.19988873600959778,
      "num_tokens": 28462378.0,
      "step": 13710
    },
    {
      "entropy": 5.562987518310547,
      "epoch": 1.2441037735849056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004851239338188286,
      "loss": 5.1173,
      "mean_token_accuracy": 0.19390659928321838,
      "num_tokens": 28473084.0,
      "step": 13715
    },
    {
      "entropy": 5.484537076950073,
      "epoch": 1.24455732946299,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048511236177612994,
      "loss": 5.041,
      "mean_token_accuracy": 0.1961607202887535,
      "num_tokens": 28482767.0,
      "step": 13720
    },
    {
      "entropy": 5.44678144454956,
      "epoch": 1.245010885341074,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048510078538825413,
      "loss": 5.0312,
      "mean_token_accuracy": 0.20292534232139586,
      "num_tokens": 28493274.0,
      "step": 13725
    },
    {
      "entropy": 5.520133829116821,
      "epoch": 1.2454644412191582,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048508920465544056,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20110096782445908,
      "num_tokens": 28503823.0,
      "step": 13730
    },
    {
      "entropy": 5.439535331726074,
      "epoch": 1.2459179970972425,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004850776195779286,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20129793733358384,
      "num_tokens": 28514209.0,
      "step": 13735
    },
    {
      "entropy": 5.532947778701782,
      "epoch": 1.2463715529753265,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000485066030155958,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19237526655197143,
      "num_tokens": 28524665.0,
      "step": 13740
    },
    {
      "entropy": 5.485327291488647,
      "epoch": 1.2468251088534108,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004850544363897684,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20430061221122742,
      "num_tokens": 28533639.0,
      "step": 13745
    },
    {
      "entropy": 5.405498456954956,
      "epoch": 1.2472786647314948,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048504283827959947,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20359645783901215,
      "num_tokens": 28544469.0,
      "step": 13750
    },
    {
      "entropy": 5.532805395126343,
      "epoch": 1.2477322206095791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850312358256912,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20535819679498674,
      "num_tokens": 28554637.0,
      "step": 13755
    },
    {
      "entropy": 5.463185548782349,
      "epoch": 1.2481857764876634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004850196290282834,
      "loss": 5.0384,
      "mean_token_accuracy": 0.20495663583278656,
      "num_tokens": 28564752.0,
      "step": 13760
    },
    {
      "entropy": 5.503740978240967,
      "epoch": 1.2486393323657474,
      "grad_norm": 1.0,
      "learning_rate": 0.00048500801788761613,
      "loss": 5.1586,
      "mean_token_accuracy": 0.19918418526649476,
      "num_tokens": 28575007.0,
      "step": 13765
    },
    {
      "entropy": 5.5523582935333256,
      "epoch": 1.2490928882438317,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004849964024039296,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19649922996759414,
      "num_tokens": 28584907.0,
      "step": 13770
    },
    {
      "entropy": 5.572278308868408,
      "epoch": 1.2495464441219157,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048498478257746387,
      "loss": 5.1286,
      "mean_token_accuracy": 0.20043385326862334,
      "num_tokens": 28594244.0,
      "step": 13775
    },
    {
      "entropy": 5.506260251998901,
      "epoch": 1.25,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048497315840845934,
      "loss": 5.0921,
      "mean_token_accuracy": 0.1991775080561638,
      "num_tokens": 28603816.0,
      "step": 13780
    },
    {
      "entropy": 5.560829496383667,
      "epoch": 1.2504535558780843,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004849615298971565,
      "loss": 5.1309,
      "mean_token_accuracy": 0.19353427439928056,
      "num_tokens": 28613067.0,
      "step": 13785
    },
    {
      "entropy": 5.500925207138062,
      "epoch": 1.2509071117561683,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048494989704379556,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20042569041252137,
      "num_tokens": 28623752.0,
      "step": 13790
    },
    {
      "entropy": 5.450966548919678,
      "epoch": 1.2513606676342526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004849382598486173,
      "loss": 5.0284,
      "mean_token_accuracy": 0.2054324969649315,
      "num_tokens": 28633190.0,
      "step": 13795
    },
    {
      "entropy": 5.589546823501587,
      "epoch": 1.2518142235123366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048492661831186243,
      "loss": 5.1193,
      "mean_token_accuracy": 0.19111987203359604,
      "num_tokens": 28644567.0,
      "step": 13800
    },
    {
      "entropy": 5.544061851501465,
      "epoch": 1.2522677793904209,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004849149724337715,
      "loss": 5.1056,
      "mean_token_accuracy": 0.19287900030612945,
      "num_tokens": 28654785.0,
      "step": 13805
    },
    {
      "entropy": 5.419747304916382,
      "epoch": 1.2527213352685052,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048490332221458543,
      "loss": 4.9955,
      "mean_token_accuracy": 0.19568415731191635,
      "num_tokens": 28665286.0,
      "step": 13810
    },
    {
      "entropy": 5.51972770690918,
      "epoch": 1.2531748911465892,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004848916676545452,
      "loss": 5.1157,
      "mean_token_accuracy": 0.19823004603385924,
      "num_tokens": 28675830.0,
      "step": 13815
    },
    {
      "entropy": 5.568089532852173,
      "epoch": 1.2536284470246735,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048488000875389174,
      "loss": 5.1713,
      "mean_token_accuracy": 0.18988842964172364,
      "num_tokens": 28686640.0,
      "step": 13820
    },
    {
      "entropy": 5.505718946456909,
      "epoch": 1.2540820029027575,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004848683455128663,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20179916322231292,
      "num_tokens": 28697677.0,
      "step": 13825
    },
    {
      "entropy": 5.494057321548462,
      "epoch": 1.2545355587808418,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048485667793170994,
      "loss": 5.0571,
      "mean_token_accuracy": 0.19772837162017823,
      "num_tokens": 28708831.0,
      "step": 13830
    },
    {
      "entropy": 5.5372607707977295,
      "epoch": 1.254989114658926,
      "grad_norm": 1.0,
      "learning_rate": 0.000484845006010664,
      "loss": 5.1174,
      "mean_token_accuracy": 0.19325800836086274,
      "num_tokens": 28718959.0,
      "step": 13835
    },
    {
      "entropy": 5.494212627410889,
      "epoch": 1.25544267053701,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004848333297499699,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20306683331727982,
      "num_tokens": 28728719.0,
      "step": 13840
    },
    {
      "entropy": 5.565085124969483,
      "epoch": 1.2558962264150944,
      "grad_norm": 1.0,
      "learning_rate": 0.00048482164914986903,
      "loss": 5.2087,
      "mean_token_accuracy": 0.18758925199508666,
      "num_tokens": 28740032.0,
      "step": 13845
    },
    {
      "entropy": 5.449249124526977,
      "epoch": 1.2563497822931784,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048480996421060307,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2051217496395111,
      "num_tokens": 28749374.0,
      "step": 13850
    },
    {
      "entropy": 5.466462850570679,
      "epoch": 1.2568033381712627,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048479827493241347,
      "loss": 5.0566,
      "mean_token_accuracy": 0.19801553338766098,
      "num_tokens": 28760038.0,
      "step": 13855
    },
    {
      "entropy": 5.49345588684082,
      "epoch": 1.257256894049347,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048478658131554216,
      "loss": 5.0803,
      "mean_token_accuracy": 0.19837768524885177,
      "num_tokens": 28770247.0,
      "step": 13860
    },
    {
      "entropy": 5.488525819778443,
      "epoch": 1.2577104499274312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004847748833602308,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20438348203897477,
      "num_tokens": 28780251.0,
      "step": 13865
    },
    {
      "entropy": 5.443554973602295,
      "epoch": 1.2581640058055152,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004847631810667215,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2038698151707649,
      "num_tokens": 28790829.0,
      "step": 13870
    },
    {
      "entropy": 5.4590709686279295,
      "epoch": 1.2586175616835993,
      "grad_norm": 1.0,
      "learning_rate": 0.0004847514744352561,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2023104190826416,
      "num_tokens": 28801036.0,
      "step": 13875
    },
    {
      "entropy": 5.403080654144287,
      "epoch": 1.2590711175616836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004847397634660768,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20570943951606752,
      "num_tokens": 28811013.0,
      "step": 13880
    },
    {
      "entropy": 5.440387964248657,
      "epoch": 1.2595246734397678,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048472804815942575,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20063267648220062,
      "num_tokens": 28821367.0,
      "step": 13885
    },
    {
      "entropy": 5.50364317893982,
      "epoch": 1.259978229317852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048471632851554517,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20409667789936065,
      "num_tokens": 28831077.0,
      "step": 13890
    },
    {
      "entropy": 5.5035868167877195,
      "epoch": 1.2604317851959361,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004847046045346775,
      "loss": 5.0865,
      "mean_token_accuracy": 0.1920575052499771,
      "num_tokens": 28841490.0,
      "step": 13895
    },
    {
      "entropy": 5.508995771408081,
      "epoch": 1.2608853410740204,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048469287621706516,
      "loss": 5.0859,
      "mean_token_accuracy": 0.19784659445285796,
      "num_tokens": 28852510.0,
      "step": 13900
    },
    {
      "entropy": 5.496377563476562,
      "epoch": 1.2613388969521044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004846811435629507,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20141980946063995,
      "num_tokens": 28863388.0,
      "step": 13905
    },
    {
      "entropy": 5.401807069778442,
      "epoch": 1.2617924528301887,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004846694065725768,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20184410065412522,
      "num_tokens": 28873798.0,
      "step": 13910
    },
    {
      "entropy": 5.483473587036133,
      "epoch": 1.262246008708273,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004846576652461861,
      "loss": 5.0442,
      "mean_token_accuracy": 0.19962013363838196,
      "num_tokens": 28883482.0,
      "step": 13915
    },
    {
      "entropy": 5.4887090682983395,
      "epoch": 1.262699564586357,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048464591958402147,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20471181720495224,
      "num_tokens": 28893573.0,
      "step": 13920
    },
    {
      "entropy": 5.560998773574829,
      "epoch": 1.2631531204644413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048463416958632573,
      "loss": 5.1454,
      "mean_token_accuracy": 0.1967838451266289,
      "num_tokens": 28904064.0,
      "step": 13925
    },
    {
      "entropy": 5.505468559265137,
      "epoch": 1.2636066763425253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048462241525334206,
      "loss": 5.1,
      "mean_token_accuracy": 0.19794451743364333,
      "num_tokens": 28915329.0,
      "step": 13930
    },
    {
      "entropy": 5.49501748085022,
      "epoch": 1.2640602322206096,
      "grad_norm": 1.0,
      "learning_rate": 0.0004846106565853133,
      "loss": 5.1859,
      "mean_token_accuracy": 0.19487343728542328,
      "num_tokens": 28925228.0,
      "step": 13935
    },
    {
      "entropy": 5.477839565277099,
      "epoch": 1.2645137880986939,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048459889358248284,
      "loss": 5.023,
      "mean_token_accuracy": 0.19917779564857482,
      "num_tokens": 28935891.0,
      "step": 13940
    },
    {
      "entropy": 5.4873875141143795,
      "epoch": 1.264967343976778,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048458712624509383,
      "loss": 4.985,
      "mean_token_accuracy": 0.20131113827228547,
      "num_tokens": 28946372.0,
      "step": 13945
    },
    {
      "entropy": 5.5058043003082275,
      "epoch": 1.2654208998548622,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048457535457338964,
      "loss": 5.0739,
      "mean_token_accuracy": 0.1947830557823181,
      "num_tokens": 28957874.0,
      "step": 13950
    },
    {
      "entropy": 5.5295580387115475,
      "epoch": 1.2658744557329462,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004845635785676137,
      "loss": 5.0528,
      "mean_token_accuracy": 0.19305006712675093,
      "num_tokens": 28968023.0,
      "step": 13955
    },
    {
      "entropy": 5.480109596252442,
      "epoch": 1.2663280116110305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048455179822800957,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2045432895421982,
      "num_tokens": 28977744.0,
      "step": 13960
    },
    {
      "entropy": 5.56711974143982,
      "epoch": 1.2667815674891147,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004845400135548208,
      "loss": 5.2127,
      "mean_token_accuracy": 0.191379314661026,
      "num_tokens": 28988875.0,
      "step": 13965
    },
    {
      "entropy": 5.415848731994629,
      "epoch": 1.2672351233671988,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004845282245482912,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20306566059589387,
      "num_tokens": 28998562.0,
      "step": 13970
    },
    {
      "entropy": 5.4908302307128904,
      "epoch": 1.267688679245283,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845164312086646,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20640750229358673,
      "num_tokens": 29008783.0,
      "step": 13975
    },
    {
      "entropy": 5.436105823516845,
      "epoch": 1.268142235123367,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004845046335361847,
      "loss": 4.9642,
      "mean_token_accuracy": 0.19867230355739593,
      "num_tokens": 29019303.0,
      "step": 13980
    },
    {
      "entropy": 5.5175525665283205,
      "epoch": 1.2685957910014514,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048449283153109566,
      "loss": 5.1912,
      "mean_token_accuracy": 0.18820945173501968,
      "num_tokens": 29029843.0,
      "step": 13985
    },
    {
      "entropy": 5.526941013336182,
      "epoch": 1.2690493468795356,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004844810251936413,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19493067264556885,
      "num_tokens": 29040603.0,
      "step": 13990
    },
    {
      "entropy": 5.484497308731079,
      "epoch": 1.2695029027576197,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004844692145240661,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20133136510848998,
      "num_tokens": 29050599.0,
      "step": 13995
    },
    {
      "entropy": 5.470039415359497,
      "epoch": 1.269956458635704,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004844573995226142,
      "loss": 5.0875,
      "mean_token_accuracy": 0.195917047560215,
      "num_tokens": 29061292.0,
      "step": 14000
    },
    {
      "entropy": 5.417318868637085,
      "epoch": 1.270410014513788,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048444558018952984,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2037899911403656,
      "num_tokens": 29071779.0,
      "step": 14005
    },
    {
      "entropy": 5.4463835716247555,
      "epoch": 1.2708635703918723,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048443375652505754,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20027604550123215,
      "num_tokens": 29082073.0,
      "step": 14010
    },
    {
      "entropy": 5.47108268737793,
      "epoch": 1.2713171262699565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004844219285294418,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20421914309263228,
      "num_tokens": 29092032.0,
      "step": 14015
    },
    {
      "entropy": 5.4538318634033205,
      "epoch": 1.2717706821480406,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048441009620292714,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2062854215502739,
      "num_tokens": 29102936.0,
      "step": 14020
    },
    {
      "entropy": 5.460636281967163,
      "epoch": 1.2722242380261248,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004843982595457584,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20491472631692886,
      "num_tokens": 29113385.0,
      "step": 14025
    },
    {
      "entropy": 5.457720708847046,
      "epoch": 1.2726777939042089,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048438641855818026,
      "loss": 5.0877,
      "mean_token_accuracy": 0.19535764306783676,
      "num_tokens": 29125075.0,
      "step": 14030
    },
    {
      "entropy": 5.503909730911255,
      "epoch": 1.2731313497822931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004843745732404376,
      "loss": 5.0182,
      "mean_token_accuracy": 0.1980168715119362,
      "num_tokens": 29135098.0,
      "step": 14035
    },
    {
      "entropy": 5.545841121673584,
      "epoch": 1.2735849056603774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004843627235927755,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19769282042980194,
      "num_tokens": 29145577.0,
      "step": 14040
    },
    {
      "entropy": 5.463726425170899,
      "epoch": 1.2740384615384617,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004843508696154388,
      "loss": 5.0457,
      "mean_token_accuracy": 0.20328806787729264,
      "num_tokens": 29155905.0,
      "step": 14045
    },
    {
      "entropy": 5.406875324249268,
      "epoch": 1.2744920174165457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048433901130867283,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20879150778055192,
      "num_tokens": 29164926.0,
      "step": 14050
    },
    {
      "entropy": 5.431298923492432,
      "epoch": 1.2749455732946298,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004843271486727228,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2010043293237686,
      "num_tokens": 29175050.0,
      "step": 14055
    },
    {
      "entropy": 5.49884524345398,
      "epoch": 1.275399129172714,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048431528170783395,
      "loss": 5.0853,
      "mean_token_accuracy": 0.19557402282953262,
      "num_tokens": 29186255.0,
      "step": 14060
    },
    {
      "entropy": 5.4883829116821286,
      "epoch": 1.2758526850507983,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004843034104142517,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2044265389442444,
      "num_tokens": 29195639.0,
      "step": 14065
    },
    {
      "entropy": 5.477971744537354,
      "epoch": 1.2763062409288826,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004842915347922216,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1940547063946724,
      "num_tokens": 29206482.0,
      "step": 14070
    },
    {
      "entropy": 5.554836750030518,
      "epoch": 1.2767597968069666,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004842796548419893,
      "loss": 5.0606,
      "mean_token_accuracy": 0.19817731380462647,
      "num_tokens": 29217433.0,
      "step": 14075
    },
    {
      "entropy": 5.548083972930908,
      "epoch": 1.2772133526850509,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048426777056380035,
      "loss": 5.1262,
      "mean_token_accuracy": 0.18795846849679948,
      "num_tokens": 29226193.0,
      "step": 14080
    },
    {
      "entropy": 5.44550051689148,
      "epoch": 1.277666908563135,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004842558819579006,
      "loss": 5.0557,
      "mean_token_accuracy": 0.20124146789312364,
      "num_tokens": 29237061.0,
      "step": 14085
    },
    {
      "entropy": 5.469652271270752,
      "epoch": 1.2781204644412192,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004842439890245358,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20593013614416122,
      "num_tokens": 29247689.0,
      "step": 14090
    },
    {
      "entropy": 5.469670963287354,
      "epoch": 1.2785740203193035,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048423209176395204,
      "loss": 5.018,
      "mean_token_accuracy": 0.2025768503546715,
      "num_tokens": 29258235.0,
      "step": 14095
    },
    {
      "entropy": 5.454458236694336,
      "epoch": 1.2790275761973875,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004842201901763954,
      "loss": 4.9898,
      "mean_token_accuracy": 0.20647548735141755,
      "num_tokens": 29268884.0,
      "step": 14100
    },
    {
      "entropy": 5.475712776184082,
      "epoch": 1.2794811320754718,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004842082842621118,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19233165234327315,
      "num_tokens": 29280484.0,
      "step": 14105
    },
    {
      "entropy": 5.504487562179565,
      "epoch": 1.2799346879535558,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004841963740213476,
      "loss": 5.0951,
      "mean_token_accuracy": 0.199015374481678,
      "num_tokens": 29291199.0,
      "step": 14110
    },
    {
      "entropy": 5.559184598922729,
      "epoch": 1.28038824383164,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004841844594543491,
      "loss": 5.0801,
      "mean_token_accuracy": 0.19844071418046952,
      "num_tokens": 29302007.0,
      "step": 14115
    },
    {
      "entropy": 5.513727045059204,
      "epoch": 1.2808417997097243,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841725405613626,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20648347586393356,
      "num_tokens": 29312127.0,
      "step": 14120
    },
    {
      "entropy": 5.5265700340271,
      "epoch": 1.2812953555878084,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004841606173426347,
      "loss": 5.1555,
      "mean_token_accuracy": 0.19536625295877458,
      "num_tokens": 29323691.0,
      "step": 14125
    },
    {
      "entropy": 5.46632194519043,
      "epoch": 1.2817489114658926,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048414868979841197,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1954666033387184,
      "num_tokens": 29335203.0,
      "step": 14130
    },
    {
      "entropy": 5.524719381332398,
      "epoch": 1.2822024673439767,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048413675792894105,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19794932007789612,
      "num_tokens": 29346422.0,
      "step": 14135
    },
    {
      "entropy": 5.46589846611023,
      "epoch": 1.282656023222061,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004841248217344686,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20230863988399506,
      "num_tokens": 29356084.0,
      "step": 14140
    },
    {
      "entropy": 5.420564365386963,
      "epoch": 1.2831095791001452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841128812152415,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20358027517795563,
      "num_tokens": 29366093.0,
      "step": 14145
    },
    {
      "entropy": 5.505201244354248,
      "epoch": 1.2835631349782293,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841009363715068,
      "loss": 5.0912,
      "mean_token_accuracy": 0.19935815930366516,
      "num_tokens": 29376621.0,
      "step": 14150
    },
    {
      "entropy": 5.556482458114624,
      "epoch": 1.2840166908563135,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004840889872035115,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19459037482738495,
      "num_tokens": 29387119.0,
      "step": 14155
    },
    {
      "entropy": 5.484884357452392,
      "epoch": 1.2844702467343976,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004840770337115026,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20135198682546615,
      "num_tokens": 29396981.0,
      "step": 14160
    },
    {
      "entropy": 5.525591802597046,
      "epoch": 1.2849238026124818,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048406507589572747,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20318254977464675,
      "num_tokens": 29406737.0,
      "step": 14165
    },
    {
      "entropy": 5.500586366653442,
      "epoch": 1.2853773584905661,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004840531137564331,
      "loss": 5.0564,
      "mean_token_accuracy": 0.19997634142637252,
      "num_tokens": 29417300.0,
      "step": 14170
    },
    {
      "entropy": 5.476732492446899,
      "epoch": 1.2858309143686502,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004840411472938672,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2080481082201004,
      "num_tokens": 29426749.0,
      "step": 14175
    },
    {
      "entropy": 5.405505704879761,
      "epoch": 1.2862844702467344,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004840291765082771,
      "loss": 5.0034,
      "mean_token_accuracy": 0.19875230491161347,
      "num_tokens": 29437264.0,
      "step": 14180
    },
    {
      "entropy": 5.474243116378784,
      "epoch": 1.2867380261248185,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048401720139991036,
      "loss": 4.9954,
      "mean_token_accuracy": 0.19903332740068436,
      "num_tokens": 29448192.0,
      "step": 14185
    },
    {
      "entropy": 5.5092840671539305,
      "epoch": 1.2871915820029027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004840052219690146,
      "loss": 4.9105,
      "mean_token_accuracy": 0.20989975184202195,
      "num_tokens": 29457923.0,
      "step": 14190
    },
    {
      "entropy": 5.3856312274932865,
      "epoch": 1.287645137880987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004839932382158376,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20904698222875595,
      "num_tokens": 29466914.0,
      "step": 14195
    },
    {
      "entropy": 5.449931573867798,
      "epoch": 1.288098693759071,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004839812501406272,
      "loss": 5.0846,
      "mean_token_accuracy": 0.1940669134259224,
      "num_tokens": 29477648.0,
      "step": 14200
    },
    {
      "entropy": 5.52697377204895,
      "epoch": 1.2885522496371553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004839692577436313,
      "loss": 5.0756,
      "mean_token_accuracy": 0.2011621966958046,
      "num_tokens": 29487947.0,
      "step": 14205
    },
    {
      "entropy": 5.589422082901001,
      "epoch": 1.2890058055152394,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004839572610250978,
      "loss": 5.1548,
      "mean_token_accuracy": 0.1920746996998787,
      "num_tokens": 29497667.0,
      "step": 14210
    },
    {
      "entropy": 5.475475311279297,
      "epoch": 1.2894593613933236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048394525998527496,
      "loss": 5.0363,
      "mean_token_accuracy": 0.1968516707420349,
      "num_tokens": 29507065.0,
      "step": 14215
    },
    {
      "entropy": 5.54481086730957,
      "epoch": 1.2899129172714079,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004839332546244109,
      "loss": 5.1283,
      "mean_token_accuracy": 0.1999591186642647,
      "num_tokens": 29517752.0,
      "step": 14220
    },
    {
      "entropy": 5.490834951400757,
      "epoch": 1.2903664731494922,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048392124494275385,
      "loss": 5.0457,
      "mean_token_accuracy": 0.19647222459316255,
      "num_tokens": 29527894.0,
      "step": 14225
    },
    {
      "entropy": 5.454130935668945,
      "epoch": 1.2908200290275762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004839092309405523,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19558317065238953,
      "num_tokens": 29538136.0,
      "step": 14230
    },
    {
      "entropy": 5.496419763565063,
      "epoch": 1.2912735849056602,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048389721261805457,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20487109571695328,
      "num_tokens": 29549193.0,
      "step": 14235
    },
    {
      "entropy": 5.461962795257568,
      "epoch": 1.2917271407837445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048388518997550926,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2094683215022087,
      "num_tokens": 29558889.0,
      "step": 14240
    },
    {
      "entropy": 5.4891866683959964,
      "epoch": 1.2921806966618288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048387316301316493,
      "loss": 5.0708,
      "mean_token_accuracy": 0.1955946922302246,
      "num_tokens": 29569691.0,
      "step": 14245
    },
    {
      "entropy": 5.508250522613525,
      "epoch": 1.292634252539913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838611317312705,
      "loss": 5.1483,
      "mean_token_accuracy": 0.19559630900621414,
      "num_tokens": 29580459.0,
      "step": 14250
    },
    {
      "entropy": 5.51726884841919,
      "epoch": 1.293087808417997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004838490961300745,
      "loss": 5.0637,
      "mean_token_accuracy": 0.19881267249584197,
      "num_tokens": 29590652.0,
      "step": 14255
    },
    {
      "entropy": 5.512678384780884,
      "epoch": 1.2935413642960814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048383705620982606,
      "loss": 5.089,
      "mean_token_accuracy": 0.2042189583182335,
      "num_tokens": 29600353.0,
      "step": 14260
    },
    {
      "entropy": 5.496693420410156,
      "epoch": 1.2939949201741654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048382501197077405,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20726697742938996,
      "num_tokens": 29610561.0,
      "step": 14265
    },
    {
      "entropy": 5.438076114654541,
      "epoch": 1.2944484760522497,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004838129634131675,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2033169150352478,
      "num_tokens": 29621741.0,
      "step": 14270
    },
    {
      "entropy": 5.44814715385437,
      "epoch": 1.294902031930334,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004838009105372558,
      "loss": 4.967,
      "mean_token_accuracy": 0.201059290766716,
      "num_tokens": 29633097.0,
      "step": 14275
    },
    {
      "entropy": 5.48220272064209,
      "epoch": 1.295355587808418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048378885334328793,
      "loss": 5.088,
      "mean_token_accuracy": 0.19085302352905273,
      "num_tokens": 29644290.0,
      "step": 14280
    },
    {
      "entropy": 5.585373497009277,
      "epoch": 1.2958091436865022,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004837767918315135,
      "loss": 5.1258,
      "mean_token_accuracy": 0.19204142391681672,
      "num_tokens": 29654977.0,
      "step": 14285
    },
    {
      "entropy": 5.5744030475616455,
      "epoch": 1.2962626995645863,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048376472600218174,
      "loss": 5.1235,
      "mean_token_accuracy": 0.19505043774843217,
      "num_tokens": 29665020.0,
      "step": 14290
    },
    {
      "entropy": 5.475365161895752,
      "epoch": 1.2967162554426706,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004837526558555422,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20274775326251984,
      "num_tokens": 29675098.0,
      "step": 14295
    },
    {
      "entropy": 5.428688049316406,
      "epoch": 1.2971698113207548,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004837405813918446,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2038100093603134,
      "num_tokens": 29685450.0,
      "step": 14300
    },
    {
      "entropy": 5.453429365158081,
      "epoch": 1.2976233671988389,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048372850261133856,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2004016637802124,
      "num_tokens": 29694818.0,
      "step": 14305
    },
    {
      "entropy": 5.5270758152008055,
      "epoch": 1.2980769230769231,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048371641951427393,
      "loss": 5.1108,
      "mean_token_accuracy": 0.19570721834897994,
      "num_tokens": 29705573.0,
      "step": 14310
    },
    {
      "entropy": 5.487916421890259,
      "epoch": 1.2985304789550072,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048370433210090055,
      "loss": 5.0108,
      "mean_token_accuracy": 0.19875363856554032,
      "num_tokens": 29716366.0,
      "step": 14315
    },
    {
      "entropy": 5.469477128982544,
      "epoch": 1.2989840348330914,
      "grad_norm": 1.0,
      "learning_rate": 0.0004836922403714683,
      "loss": 5.042,
      "mean_token_accuracy": 0.19777244478464126,
      "num_tokens": 29727181.0,
      "step": 14320
    },
    {
      "entropy": 5.452441024780273,
      "epoch": 1.2994375907111757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048368014432622746,
      "loss": 5.0523,
      "mean_token_accuracy": 0.20698187947273256,
      "num_tokens": 29736960.0,
      "step": 14325
    },
    {
      "entropy": 5.445215511322021,
      "epoch": 1.2998911465892597,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000483668043965428,
      "loss": 4.9363,
      "mean_token_accuracy": 0.21604397892951965,
      "num_tokens": 29746354.0,
      "step": 14330
    },
    {
      "entropy": 5.471987867355347,
      "epoch": 1.300344702467344,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004836559392893203,
      "loss": 5.0542,
      "mean_token_accuracy": 0.1957583099603653,
      "num_tokens": 29756627.0,
      "step": 14335
    },
    {
      "entropy": 5.464911985397339,
      "epoch": 1.300798258345428,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004836438302981545,
      "loss": 5.0398,
      "mean_token_accuracy": 0.20956512540578842,
      "num_tokens": 29766684.0,
      "step": 14340
    },
    {
      "entropy": 5.470847320556641,
      "epoch": 1.3012518142235123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004836317169921811,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19975050240755082,
      "num_tokens": 29776556.0,
      "step": 14345
    },
    {
      "entropy": 5.553652000427246,
      "epoch": 1.3017053701015966,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048361959937165075,
      "loss": 5.1958,
      "mean_token_accuracy": 0.19425188899040222,
      "num_tokens": 29786896.0,
      "step": 14350
    },
    {
      "entropy": 5.496381521224976,
      "epoch": 1.3021589259796806,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004836074774368138,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2041577160358429,
      "num_tokens": 29797332.0,
      "step": 14355
    },
    {
      "entropy": 5.503427219390869,
      "epoch": 1.302612481857765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048359535118792116,
      "loss": 5.0903,
      "mean_token_accuracy": 0.20044789463281631,
      "num_tokens": 29806428.0,
      "step": 14360
    },
    {
      "entropy": 5.4628736019134525,
      "epoch": 1.303066037735849,
      "grad_norm": 1.0,
      "learning_rate": 0.0004835832206252235,
      "loss": 5.0628,
      "mean_token_accuracy": 0.19847192913293837,
      "num_tokens": 29816794.0,
      "step": 14365
    },
    {
      "entropy": 5.562831878662109,
      "epoch": 1.3035195936139332,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004835710857489716,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20265541970729828,
      "num_tokens": 29827077.0,
      "step": 14370
    },
    {
      "entropy": 5.462686586380005,
      "epoch": 1.3039731494920175,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004835589465594165,
      "loss": 5.0597,
      "mean_token_accuracy": 0.1975301370024681,
      "num_tokens": 29837868.0,
      "step": 14375
    },
    {
      "entropy": 5.480144453048706,
      "epoch": 1.3044267053701015,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048354680305680926,
      "loss": 5.0848,
      "mean_token_accuracy": 0.2043821007013321,
      "num_tokens": 29846929.0,
      "step": 14380
    },
    {
      "entropy": 5.444469165802002,
      "epoch": 1.3048802612481858,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048353465524140105,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20302888005971909,
      "num_tokens": 29857905.0,
      "step": 14385
    },
    {
      "entropy": 5.474423599243164,
      "epoch": 1.3053338171262698,
      "grad_norm": 1.078125,
      "learning_rate": 0.000483522503113443,
      "loss": 5.0751,
      "mean_token_accuracy": 0.1963631957769394,
      "num_tokens": 29868230.0,
      "step": 14390
    },
    {
      "entropy": 5.482108020782471,
      "epoch": 1.305787373004354,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048351034667318634,
      "loss": 5.0315,
      "mean_token_accuracy": 0.20345700830221175,
      "num_tokens": 29879224.0,
      "step": 14395
    },
    {
      "entropy": 5.5577699661254885,
      "epoch": 1.3062409288824384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004834981859208827,
      "loss": 5.1748,
      "mean_token_accuracy": 0.19315745383501054,
      "num_tokens": 29890416.0,
      "step": 14400
    },
    {
      "entropy": 5.52885799407959,
      "epoch": 1.3066944847605226,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048348602085678346,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19536293894052506,
      "num_tokens": 29901064.0,
      "step": 14405
    },
    {
      "entropy": 5.577747058868408,
      "epoch": 1.3071480406386067,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048347385148114004,
      "loss": 5.1489,
      "mean_token_accuracy": 0.19691503047943115,
      "num_tokens": 29911084.0,
      "step": 14410
    },
    {
      "entropy": 5.467166519165039,
      "epoch": 1.3076015965166907,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048346167779420436,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21268735975027084,
      "num_tokens": 29921104.0,
      "step": 14415
    },
    {
      "entropy": 5.474800539016724,
      "epoch": 1.308055152394775,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048344949979622806,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20343236178159713,
      "num_tokens": 29932014.0,
      "step": 14420
    },
    {
      "entropy": 5.46609206199646,
      "epoch": 1.3085087082728593,
      "grad_norm": 1.125,
      "learning_rate": 0.00048343731748746294,
      "loss": 5.0518,
      "mean_token_accuracy": 0.2032724514603615,
      "num_tokens": 29941707.0,
      "step": 14425
    },
    {
      "entropy": 5.475191116333008,
      "epoch": 1.3089622641509435,
      "grad_norm": 1.09375,
      "learning_rate": 0.000483425130868161,
      "loss": 5.0461,
      "mean_token_accuracy": 0.19366600662469863,
      "num_tokens": 29952082.0,
      "step": 14430
    },
    {
      "entropy": 5.422572040557862,
      "epoch": 1.3094158200290276,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004834129399385742,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22088147699832916,
      "num_tokens": 29961744.0,
      "step": 14435
    },
    {
      "entropy": 5.479946565628052,
      "epoch": 1.3098693759071118,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004834007446989548,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2101757764816284,
      "num_tokens": 29972611.0,
      "step": 14440
    },
    {
      "entropy": 5.459532022476196,
      "epoch": 1.3103229317851959,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048338854514955474,
      "loss": 4.9624,
      "mean_token_accuracy": 0.208077135682106,
      "num_tokens": 29982660.0,
      "step": 14445
    },
    {
      "entropy": 5.445099782943726,
      "epoch": 1.3107764876632801,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048337634129062655,
      "loss": 5.0747,
      "mean_token_accuracy": 0.1888832926750183,
      "num_tokens": 29993349.0,
      "step": 14450
    },
    {
      "entropy": 5.562231159210205,
      "epoch": 1.3112300435413644,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048336413312242255,
      "loss": 5.1659,
      "mean_token_accuracy": 0.19416691362857819,
      "num_tokens": 30004474.0,
      "step": 14455
    },
    {
      "entropy": 5.558346843719482,
      "epoch": 1.3116835994194485,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048335192064519514,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19719318449497222,
      "num_tokens": 30015189.0,
      "step": 14460
    },
    {
      "entropy": 5.511405658721924,
      "epoch": 1.3121371552975327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048333970385919695,
      "loss": 4.9939,
      "mean_token_accuracy": 0.19969361424446105,
      "num_tokens": 30025491.0,
      "step": 14465
    },
    {
      "entropy": 5.478708028793335,
      "epoch": 1.3125907111756168,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048332748276468053,
      "loss": 5.0612,
      "mean_token_accuracy": 0.20458794683218,
      "num_tokens": 30034683.0,
      "step": 14470
    },
    {
      "entropy": 5.471246767044067,
      "epoch": 1.313044267053701,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004833152573618987,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20007718205451966,
      "num_tokens": 30044135.0,
      "step": 14475
    },
    {
      "entropy": 5.51534218788147,
      "epoch": 1.3134978229317853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048330302765110436,
      "loss": 5.1094,
      "mean_token_accuracy": 0.2011711835861206,
      "num_tokens": 30054520.0,
      "step": 14480
    },
    {
      "entropy": 5.505444955825806,
      "epoch": 1.3139513788098693,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048329079363255024,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19864712804555892,
      "num_tokens": 30065439.0,
      "step": 14485
    },
    {
      "entropy": 5.502534627914429,
      "epoch": 1.3144049346879536,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048327855530648934,
      "loss": 5.0746,
      "mean_token_accuracy": 0.1971224755048752,
      "num_tokens": 30075362.0,
      "step": 14490
    },
    {
      "entropy": 5.471915912628174,
      "epoch": 1.3148584905660377,
      "grad_norm": 1.015625,
      "learning_rate": 0.000483266312673175,
      "loss": 5.0361,
      "mean_token_accuracy": 0.19733360409736633,
      "num_tokens": 30084913.0,
      "step": 14495
    },
    {
      "entropy": 5.471127033233643,
      "epoch": 1.315312046444122,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004832540657328602,
      "loss": 5.0596,
      "mean_token_accuracy": 0.19881338328123094,
      "num_tokens": 30095704.0,
      "step": 14500
    },
    {
      "entropy": 5.488482666015625,
      "epoch": 1.3157656023222062,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004832418144857982,
      "loss": 5.0117,
      "mean_token_accuracy": 0.20118715465068818,
      "num_tokens": 30105981.0,
      "step": 14505
    },
    {
      "entropy": 5.4902448654174805,
      "epoch": 1.3162191582002902,
      "grad_norm": 1.0,
      "learning_rate": 0.00048322955893224245,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19059209823608397,
      "num_tokens": 30115772.0,
      "step": 14510
    },
    {
      "entropy": 5.492611694335937,
      "epoch": 1.3166727140783745,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048321729907244626,
      "loss": 5.016,
      "mean_token_accuracy": 0.2023007556796074,
      "num_tokens": 30127050.0,
      "step": 14515
    },
    {
      "entropy": 5.49431209564209,
      "epoch": 1.3171262699564585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004832050349066634,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20058691054582595,
      "num_tokens": 30138221.0,
      "step": 14520
    },
    {
      "entropy": 5.533912324905396,
      "epoch": 1.3175798258345428,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004831927664351473,
      "loss": 5.1336,
      "mean_token_accuracy": 0.195409394800663,
      "num_tokens": 30150181.0,
      "step": 14525
    },
    {
      "entropy": 5.490771627426147,
      "epoch": 1.318033381712627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004831804936581518,
      "loss": 5.0948,
      "mean_token_accuracy": 0.20345331132411956,
      "num_tokens": 30160521.0,
      "step": 14530
    },
    {
      "entropy": 5.524169111251831,
      "epoch": 1.3184869375907111,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004831682165759305,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19717661589384078,
      "num_tokens": 30170611.0,
      "step": 14535
    },
    {
      "entropy": 5.526472377777099,
      "epoch": 1.3189404934687954,
      "grad_norm": 1.0,
      "learning_rate": 0.00048315593518873757,
      "loss": 5.0375,
      "mean_token_accuracy": 0.209712053835392,
      "num_tokens": 30179914.0,
      "step": 14540
    },
    {
      "entropy": 5.466052293777466,
      "epoch": 1.3193940493468794,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004831436494968268,
      "loss": 5.0337,
      "mean_token_accuracy": 0.20637008249759675,
      "num_tokens": 30191426.0,
      "step": 14545
    },
    {
      "entropy": 5.4549181938171385,
      "epoch": 1.3198476052249637,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048313135950045234,
      "loss": 5.0122,
      "mean_token_accuracy": 0.19953661859035493,
      "num_tokens": 30202813.0,
      "step": 14550
    },
    {
      "entropy": 5.475877141952514,
      "epoch": 1.320301161103048,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048311906519986824,
      "loss": 4.9444,
      "mean_token_accuracy": 0.20676357299089432,
      "num_tokens": 30213044.0,
      "step": 14555
    },
    {
      "entropy": 5.5254028797149655,
      "epoch": 1.320754716981132,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000483106766595329,
      "loss": 5.0314,
      "mean_token_accuracy": 0.1970205068588257,
      "num_tokens": 30223419.0,
      "step": 14560
    },
    {
      "entropy": 5.553525733947754,
      "epoch": 1.3212082728592163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004830944636870887,
      "loss": 5.0777,
      "mean_token_accuracy": 0.19376981556415557,
      "num_tokens": 30233203.0,
      "step": 14565
    },
    {
      "entropy": 5.487842988967896,
      "epoch": 1.3216618287373003,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048308215647540183,
      "loss": 5.0705,
      "mean_token_accuracy": 0.19618242979049683,
      "num_tokens": 30244604.0,
      "step": 14570
    },
    {
      "entropy": 5.5587037086486815,
      "epoch": 1.3221153846153846,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048306984496052297,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20165631622076036,
      "num_tokens": 30255355.0,
      "step": 14575
    },
    {
      "entropy": 5.577435302734375,
      "epoch": 1.3225689404934688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048305752914270665,
      "loss": 5.1656,
      "mean_token_accuracy": 0.19230901449918747,
      "num_tokens": 30265665.0,
      "step": 14580
    },
    {
      "entropy": 5.564355754852295,
      "epoch": 1.3230224963715531,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048304520902220767,
      "loss": 5.129,
      "mean_token_accuracy": 0.20115920305252075,
      "num_tokens": 30274347.0,
      "step": 14585
    },
    {
      "entropy": 5.51794400215149,
      "epoch": 1.3234760522496372,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004830328845992808,
      "loss": 4.9679,
      "mean_token_accuracy": 0.20817530900239944,
      "num_tokens": 30284243.0,
      "step": 14590
    },
    {
      "entropy": 5.437684631347656,
      "epoch": 1.3239296081277212,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048302055587418074,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20018340796232223,
      "num_tokens": 30294327.0,
      "step": 14595
    },
    {
      "entropy": 5.445873594284057,
      "epoch": 1.3243831640058055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004830082228471626,
      "loss": 5.0842,
      "mean_token_accuracy": 0.19650617390871047,
      "num_tokens": 30304188.0,
      "step": 14600
    },
    {
      "entropy": 5.5108343124389645,
      "epoch": 1.3248367198838897,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048299588551848146,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2086288645863533,
      "num_tokens": 30314917.0,
      "step": 14605
    },
    {
      "entropy": 5.482002687454224,
      "epoch": 1.325290275761974,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048298354388839237,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2028503105044365,
      "num_tokens": 30325221.0,
      "step": 14610
    },
    {
      "entropy": 5.512210655212402,
      "epoch": 1.325743831640058,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004829711979571506,
      "loss": 5.1134,
      "mean_token_accuracy": 0.19776519536972045,
      "num_tokens": 30336196.0,
      "step": 14615
    },
    {
      "entropy": 5.444050025939942,
      "epoch": 1.3261973875181423,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004829588477250115,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21473861038684844,
      "num_tokens": 30345546.0,
      "step": 14620
    },
    {
      "entropy": 5.52281084060669,
      "epoch": 1.3266509433962264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004829464931922304,
      "loss": 5.0554,
      "mean_token_accuracy": 0.1991855889558792,
      "num_tokens": 30355672.0,
      "step": 14625
    },
    {
      "entropy": 5.505038070678711,
      "epoch": 1.3271044992743106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048293413435906285,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20044143497943878,
      "num_tokens": 30366233.0,
      "step": 14630
    },
    {
      "entropy": 5.467237281799316,
      "epoch": 1.327558055152395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004829217712257644,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20227838754653932,
      "num_tokens": 30377366.0,
      "step": 14635
    },
    {
      "entropy": 5.473698282241822,
      "epoch": 1.328011611030479,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048290940379259073,
      "loss": 5.101,
      "mean_token_accuracy": 0.19721851050853728,
      "num_tokens": 30388022.0,
      "step": 14640
    },
    {
      "entropy": 5.483129501342773,
      "epoch": 1.3284651669085632,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048289703205979756,
      "loss": 4.9687,
      "mean_token_accuracy": 0.20514749586582184,
      "num_tokens": 30399235.0,
      "step": 14645
    },
    {
      "entropy": 5.588254690170288,
      "epoch": 1.3289187227866472,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048288465602764084,
      "loss": 5.122,
      "mean_token_accuracy": 0.20071659833192826,
      "num_tokens": 30409112.0,
      "step": 14650
    },
    {
      "entropy": 5.452871370315552,
      "epoch": 1.3293722786647315,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048287227569637645,
      "loss": 5.0219,
      "mean_token_accuracy": 0.2019849956035614,
      "num_tokens": 30420202.0,
      "step": 14655
    },
    {
      "entropy": 5.532520389556884,
      "epoch": 1.3298258345428158,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048285989106626036,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20122217237949372,
      "num_tokens": 30431210.0,
      "step": 14660
    },
    {
      "entropy": 5.497643756866455,
      "epoch": 1.3302793904208998,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004828475021375489,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20784106552600862,
      "num_tokens": 30441291.0,
      "step": 14665
    },
    {
      "entropy": 5.457648134231567,
      "epoch": 1.330732946298984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004828351089104981,
      "loss": 4.964,
      "mean_token_accuracy": 0.20614739060401915,
      "num_tokens": 30450285.0,
      "step": 14670
    },
    {
      "entropy": 5.443894147872925,
      "epoch": 1.3311865021770681,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048282271138536424,
      "loss": 5.0606,
      "mean_token_accuracy": 0.1993221625685692,
      "num_tokens": 30460311.0,
      "step": 14675
    },
    {
      "entropy": 5.560179233551025,
      "epoch": 1.3316400580551524,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004828103095624037,
      "loss": 5.072,
      "mean_token_accuracy": 0.20389872342348098,
      "num_tokens": 30470188.0,
      "step": 14680
    },
    {
      "entropy": 5.470388412475586,
      "epoch": 1.3320936139332367,
      "grad_norm": 0.921875,
      "learning_rate": 0.000482797903441873,
      "loss": 5.024,
      "mean_token_accuracy": 0.1966773748397827,
      "num_tokens": 30481406.0,
      "step": 14685
    },
    {
      "entropy": 5.407244253158569,
      "epoch": 1.3325471698113207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004827854930240288,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20737527012825013,
      "num_tokens": 30492314.0,
      "step": 14690
    },
    {
      "entropy": 5.552218914031982,
      "epoch": 1.333000725689405,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004827730783091276,
      "loss": 5.1477,
      "mean_token_accuracy": 0.19321692138910293,
      "num_tokens": 30503273.0,
      "step": 14695
    },
    {
      "entropy": 5.546647071838379,
      "epoch": 1.333454281567489,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048276065929742616,
      "loss": 5.0477,
      "mean_token_accuracy": 0.1968608483672142,
      "num_tokens": 30513582.0,
      "step": 14700
    },
    {
      "entropy": 5.440297269821167,
      "epoch": 1.3339078374455733,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004827482359891814,
      "loss": 4.97,
      "mean_token_accuracy": 0.21266762018203736,
      "num_tokens": 30523261.0,
      "step": 14705
    },
    {
      "entropy": 5.500866556167603,
      "epoch": 1.3343613933236576,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048273580838465016,
      "loss": 5.0851,
      "mean_token_accuracy": 0.19987517595291138,
      "num_tokens": 30533774.0,
      "step": 14710
    },
    {
      "entropy": 5.502251482009887,
      "epoch": 1.3348149492017416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004827233764840894,
      "loss": 5.1351,
      "mean_token_accuracy": 0.19586722254753114,
      "num_tokens": 30544580.0,
      "step": 14715
    },
    {
      "entropy": 5.525274324417114,
      "epoch": 1.3352685050798259,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004827109402877564,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20630441159009932,
      "num_tokens": 30555105.0,
      "step": 14720
    },
    {
      "entropy": 5.514051389694214,
      "epoch": 1.33572206095791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004826984997959082,
      "loss": 5.1256,
      "mean_token_accuracy": 0.18905919790267944,
      "num_tokens": 30564912.0,
      "step": 14725
    },
    {
      "entropy": 5.463503980636597,
      "epoch": 1.3361756168359942,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000482686055008802,
      "loss": 5.0573,
      "mean_token_accuracy": 0.1945282906293869,
      "num_tokens": 30575862.0,
      "step": 14730
    },
    {
      "entropy": 5.500078582763672,
      "epoch": 1.3366291727140784,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048267360592669535,
      "loss": 5.1625,
      "mean_token_accuracy": 0.19513950496912003,
      "num_tokens": 30586951.0,
      "step": 14735
    },
    {
      "entropy": 5.529357385635376,
      "epoch": 1.3370827285921625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004826611525498456,
      "loss": 5.0766,
      "mean_token_accuracy": 0.1965434119105339,
      "num_tokens": 30597604.0,
      "step": 14740
    },
    {
      "entropy": 5.536353826522827,
      "epoch": 1.3375362844702468,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004826486948785103,
      "loss": 5.1232,
      "mean_token_accuracy": 0.19308859407901763,
      "num_tokens": 30608157.0,
      "step": 14745
    },
    {
      "entropy": 5.521725034713745,
      "epoch": 1.3379898403483308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048263623291294706,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20122882276773452,
      "num_tokens": 30618527.0,
      "step": 14750
    },
    {
      "entropy": 5.541813707351684,
      "epoch": 1.338443396226415,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048262376665341355,
      "loss": 5.1482,
      "mean_token_accuracy": 0.1948135241866112,
      "num_tokens": 30629253.0,
      "step": 14755
    },
    {
      "entropy": 5.508347606658935,
      "epoch": 1.3388969521044993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004826112961001677,
      "loss": 5.1093,
      "mean_token_accuracy": 0.18560464829206466,
      "num_tokens": 30639015.0,
      "step": 14760
    },
    {
      "entropy": 5.557197237014771,
      "epoch": 1.3393505079825834,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048259882125346735,
      "loss": 5.1851,
      "mean_token_accuracy": 0.1878888040781021,
      "num_tokens": 30650035.0,
      "step": 14765
    },
    {
      "entropy": 5.557344198226929,
      "epoch": 1.3398040638606676,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048258634211357045,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1993729680776596,
      "num_tokens": 30659813.0,
      "step": 14770
    },
    {
      "entropy": 5.504110193252563,
      "epoch": 1.3402576197387517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004825738586807351,
      "loss": 5.0459,
      "mean_token_accuracy": 0.1985906630754471,
      "num_tokens": 30670243.0,
      "step": 14775
    },
    {
      "entropy": 5.452295589447021,
      "epoch": 1.340711175616836,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048256137095521945,
      "loss": 5.0215,
      "mean_token_accuracy": 0.1977318599820137,
      "num_tokens": 30680056.0,
      "step": 14780
    },
    {
      "entropy": 5.476433897018433,
      "epoch": 1.3411647314949202,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048254887893728175,
      "loss": 4.9717,
      "mean_token_accuracy": 0.2092755541205406,
      "num_tokens": 30690536.0,
      "step": 14785
    },
    {
      "entropy": 5.443423271179199,
      "epoch": 1.3416182873730045,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004825363826271804,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1987844005227089,
      "num_tokens": 30700200.0,
      "step": 14790
    },
    {
      "entropy": 5.44862470626831,
      "epoch": 1.3420718432510885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004825238820251737,
      "loss": 4.9579,
      "mean_token_accuracy": 0.20584656149148942,
      "num_tokens": 30710576.0,
      "step": 14795
    },
    {
      "entropy": 5.482344055175782,
      "epoch": 1.3425253991291728,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048251137713152027,
      "loss": 5.0585,
      "mean_token_accuracy": 0.2019256144762039,
      "num_tokens": 30721495.0,
      "step": 14800
    },
    {
      "entropy": 5.519159507751465,
      "epoch": 1.3429789550072568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004824988679464787,
      "loss": 5.0938,
      "mean_token_accuracy": 0.1964031144976616,
      "num_tokens": 30731700.0,
      "step": 14805
    },
    {
      "entropy": 5.527659940719604,
      "epoch": 1.343432510885341,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048248635447030767,
      "loss": 5.0728,
      "mean_token_accuracy": 0.19750734865665437,
      "num_tokens": 30743421.0,
      "step": 14810
    },
    {
      "entropy": 5.488374805450439,
      "epoch": 1.3438860667634254,
      "grad_norm": 1.0,
      "learning_rate": 0.0004824738367032659,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2078663155436516,
      "num_tokens": 30754238.0,
      "step": 14815
    },
    {
      "entropy": 5.480378723144531,
      "epoch": 1.3443396226415094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004824613146456123,
      "loss": 5.0715,
      "mean_token_accuracy": 0.19379124492406846,
      "num_tokens": 30763938.0,
      "step": 14820
    },
    {
      "entropy": 5.573088788986206,
      "epoch": 1.3447931785195937,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004824487882976059,
      "loss": 5.1257,
      "mean_token_accuracy": 0.19805540144443512,
      "num_tokens": 30773654.0,
      "step": 14825
    },
    {
      "entropy": 5.551896667480468,
      "epoch": 1.3452467343976777,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048243625765950564,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19453760385513305,
      "num_tokens": 30784659.0,
      "step": 14830
    },
    {
      "entropy": 5.469422769546509,
      "epoch": 1.345700290275762,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004824237227315708,
      "loss": 4.992,
      "mean_token_accuracy": 0.19851766228675843,
      "num_tokens": 30795194.0,
      "step": 14835
    },
    {
      "entropy": 5.410362339019775,
      "epoch": 1.3461538461538463,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004824111835140604,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21054262965917586,
      "num_tokens": 30805623.0,
      "step": 14840
    },
    {
      "entropy": 5.441026401519776,
      "epoch": 1.3466074020319303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000482398640007234,
      "loss": 4.9817,
      "mean_token_accuracy": 0.2019845500588417,
      "num_tokens": 30814899.0,
      "step": 14845
    },
    {
      "entropy": 5.4771339893341064,
      "epoch": 1.3470609579100146,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048238609221135076,
      "loss": 5.1093,
      "mean_token_accuracy": 0.19643137007951736,
      "num_tokens": 30825546.0,
      "step": 14850
    },
    {
      "entropy": 5.455335998535157,
      "epoch": 1.3475145137880986,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004823735401266704,
      "loss": 4.9939,
      "mean_token_accuracy": 0.19967146068811417,
      "num_tokens": 30836406.0,
      "step": 14855
    },
    {
      "entropy": 5.473223638534546,
      "epoch": 1.3479680696661829,
      "grad_norm": 1.0,
      "learning_rate": 0.00048236098375345227,
      "loss": 5.0861,
      "mean_token_accuracy": 0.19803501665592194,
      "num_tokens": 30847966.0,
      "step": 14860
    },
    {
      "entropy": 5.569869422912598,
      "epoch": 1.3484216255442671,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004823484230919563,
      "loss": 5.1078,
      "mean_token_accuracy": 0.19745918065309526,
      "num_tokens": 30856831.0,
      "step": 14865
    },
    {
      "entropy": 5.444964838027954,
      "epoch": 1.3488751814223512,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048233585814244195,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21360531598329544,
      "num_tokens": 30866076.0,
      "step": 14870
    },
    {
      "entropy": 5.4761899471282955,
      "epoch": 1.3493287373004355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004823232889051693,
      "loss": 5.0081,
      "mean_token_accuracy": 0.2073754221200943,
      "num_tokens": 30876845.0,
      "step": 14875
    },
    {
      "entropy": 5.450684928894043,
      "epoch": 1.3497822931785195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048231071538039816,
      "loss": 5.0523,
      "mean_token_accuracy": 0.19669134616851808,
      "num_tokens": 30887448.0,
      "step": 14880
    },
    {
      "entropy": 5.546138429641724,
      "epoch": 1.3502358490566038,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048229813756838863,
      "loss": 5.0474,
      "mean_token_accuracy": 0.2002974957227707,
      "num_tokens": 30898414.0,
      "step": 14885
    },
    {
      "entropy": 5.486542892456055,
      "epoch": 1.350689404934688,
      "grad_norm": 1.0,
      "learning_rate": 0.0004822855554694007,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20294709950685502,
      "num_tokens": 30908559.0,
      "step": 14890
    },
    {
      "entropy": 5.514283227920532,
      "epoch": 1.351142960812772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048227296908369475,
      "loss": 5.0293,
      "mean_token_accuracy": 0.20701266676187516,
      "num_tokens": 30918655.0,
      "step": 14895
    },
    {
      "entropy": 5.453053760528564,
      "epoch": 1.3515965166908563,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048226037841153104,
      "loss": 5.0473,
      "mean_token_accuracy": 0.20324480831623076,
      "num_tokens": 30929188.0,
      "step": 14900
    },
    {
      "entropy": 5.520646667480468,
      "epoch": 1.3520500725689404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004822477834531698,
      "loss": 5.1263,
      "mean_token_accuracy": 0.2014889895915985,
      "num_tokens": 30939902.0,
      "step": 14905
    },
    {
      "entropy": 5.496817684173584,
      "epoch": 1.3525036284470247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004822351842088716,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20315656661987305,
      "num_tokens": 30951289.0,
      "step": 14910
    },
    {
      "entropy": 5.581944751739502,
      "epoch": 1.352957184325109,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004822225806788969,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1878611609339714,
      "num_tokens": 30961233.0,
      "step": 14915
    },
    {
      "entropy": 5.486229324340821,
      "epoch": 1.353410740203193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048220997286350654,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20341320782899858,
      "num_tokens": 30971922.0,
      "step": 14920
    },
    {
      "entropy": 5.460137271881104,
      "epoch": 1.3538642960812772,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048219736076296114,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20320184230804444,
      "num_tokens": 30981358.0,
      "step": 14925
    },
    {
      "entropy": 5.512710285186768,
      "epoch": 1.3543178519593613,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004821847443775214,
      "loss": 5.1231,
      "mean_token_accuracy": 0.19762219935655595,
      "num_tokens": 30991984.0,
      "step": 14930
    },
    {
      "entropy": 5.4981756687164305,
      "epoch": 1.3547714078374455,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048217212370744853,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20261044204235076,
      "num_tokens": 31002790.0,
      "step": 14935
    },
    {
      "entropy": 5.4207916259765625,
      "epoch": 1.3552249637155298,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048215949875300327,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2109668642282486,
      "num_tokens": 31012827.0,
      "step": 14940
    },
    {
      "entropy": 5.430930042266846,
      "epoch": 1.3556785195936139,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004821468695144467,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2006464809179306,
      "num_tokens": 31023075.0,
      "step": 14945
    },
    {
      "entropy": 5.520316028594971,
      "epoch": 1.3561320754716981,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048213423599204015,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20287319719791413,
      "num_tokens": 31032454.0,
      "step": 14950
    },
    {
      "entropy": 5.57982611656189,
      "epoch": 1.3565856313497822,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048212159818604477,
      "loss": 5.1899,
      "mean_token_accuracy": 0.18757683485746385,
      "num_tokens": 31042498.0,
      "step": 14955
    },
    {
      "entropy": 5.469901990890503,
      "epoch": 1.3570391872278664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000482108956096722,
      "loss": 5.0336,
      "mean_token_accuracy": 0.2002289518713951,
      "num_tokens": 31053188.0,
      "step": 14960
    },
    {
      "entropy": 5.497510814666748,
      "epoch": 1.3574927431059507,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048209630972433326,
      "loss": 5.012,
      "mean_token_accuracy": 0.2035995379090309,
      "num_tokens": 31062840.0,
      "step": 14965
    },
    {
      "entropy": 5.561136054992676,
      "epoch": 1.357946298984035,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048208365906914,
      "loss": 5.1582,
      "mean_token_accuracy": 0.1987423777580261,
      "num_tokens": 31072646.0,
      "step": 14970
    },
    {
      "entropy": 5.48033299446106,
      "epoch": 1.358399854862119,
      "grad_norm": 1.125,
      "learning_rate": 0.00048207100413140397,
      "loss": 4.9932,
      "mean_token_accuracy": 0.1994940683245659,
      "num_tokens": 31082700.0,
      "step": 14975
    },
    {
      "entropy": 5.541996240615845,
      "epoch": 1.3588534107402033,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004820583449113868,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19881935864686967,
      "num_tokens": 31093743.0,
      "step": 14980
    },
    {
      "entropy": 5.503304290771484,
      "epoch": 1.3593069666182873,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048204568140935023,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20480051189661025,
      "num_tokens": 31104293.0,
      "step": 14985
    },
    {
      "entropy": 5.443846797943115,
      "epoch": 1.3597605224963716,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004820330136255563,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2066284880042076,
      "num_tokens": 31114244.0,
      "step": 14990
    },
    {
      "entropy": 5.46366810798645,
      "epoch": 1.3602140783744558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048202034156026683,
      "loss": 5.0191,
      "mean_token_accuracy": 0.19951775372028352,
      "num_tokens": 31123596.0,
      "step": 14995
    },
    {
      "entropy": 5.5296409130096436,
      "epoch": 1.36066763425254,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048200766521374396,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1923889234662056,
      "num_tokens": 31134258.0,
      "step": 15000
    },
    {
      "epoch": 1.36066763425254,
      "eval_entropy": 5.295666738009561,
      "eval_loss": 5.112661838531494,
      "eval_mean_token_accuracy": 0.2063279388017312,
      "eval_num_tokens": 31134258.0,
      "eval_runtime": 20.7071,
      "eval_samples_per_second": 1707.628,
      "eval_steps_per_second": 213.453,
      "step": 15000
    },
    {
      "entropy": 5.491491031646729,
      "epoch": 1.3611211901306242,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004819949845862498,
      "loss": 5.0584,
      "mean_token_accuracy": 0.2033733531832695,
      "num_tokens": 31144806.0,
      "step": 15005
    },
    {
      "entropy": 5.542679929733277,
      "epoch": 1.3615747460087082,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004819822996780466,
      "loss": 5.1565,
      "mean_token_accuracy": 0.1965965673327446,
      "num_tokens": 31155642.0,
      "step": 15010
    },
    {
      "entropy": 5.541298770904541,
      "epoch": 1.3620283018867925,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004819696104893967,
      "loss": 4.9988,
      "mean_token_accuracy": 0.20615871697664262,
      "num_tokens": 31165034.0,
      "step": 15015
    },
    {
      "entropy": 5.529401683807373,
      "epoch": 1.3624818577648767,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004819569170205625,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1988907501101494,
      "num_tokens": 31176491.0,
      "step": 15020
    },
    {
      "entropy": 5.515815877914429,
      "epoch": 1.3629354136429608,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048194421927180654,
      "loss": 5.0767,
      "mean_token_accuracy": 0.1967637598514557,
      "num_tokens": 31186198.0,
      "step": 15025
    },
    {
      "entropy": 5.515884733200073,
      "epoch": 1.363388969521045,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004819315172433914,
      "loss": 5.0597,
      "mean_token_accuracy": 0.19971127957105636,
      "num_tokens": 31197094.0,
      "step": 15030
    },
    {
      "entropy": 5.552103185653687,
      "epoch": 1.363842525399129,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048191881093557977,
      "loss": 5.0781,
      "mean_token_accuracy": 0.19090546071529388,
      "num_tokens": 31207505.0,
      "step": 15035
    },
    {
      "entropy": 5.576592445373535,
      "epoch": 1.3642960812772134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048190610034863434,
      "loss": 5.1897,
      "mean_token_accuracy": 0.18819540590047837,
      "num_tokens": 31218809.0,
      "step": 15040
    },
    {
      "entropy": 5.542851114273072,
      "epoch": 1.3647496371552976,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004818933854828181,
      "loss": 5.1072,
      "mean_token_accuracy": 0.19087555557489394,
      "num_tokens": 31230281.0,
      "step": 15045
    },
    {
      "entropy": 5.482969665527344,
      "epoch": 1.3652031930333817,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004818806663383939,
      "loss": 4.992,
      "mean_token_accuracy": 0.20569752156734467,
      "num_tokens": 31239573.0,
      "step": 15050
    },
    {
      "entropy": 5.471133422851563,
      "epoch": 1.365656748911466,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004818679429156248,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2037968084216118,
      "num_tokens": 31249671.0,
      "step": 15055
    },
    {
      "entropy": 5.448171234130859,
      "epoch": 1.36611030478955,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048185521521477385,
      "loss": 5.0496,
      "mean_token_accuracy": 0.19569021314382554,
      "num_tokens": 31259601.0,
      "step": 15060
    },
    {
      "entropy": 5.488245248794556,
      "epoch": 1.3665638606676342,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004818424832361044,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2052937850356102,
      "num_tokens": 31269085.0,
      "step": 15065
    },
    {
      "entropy": 5.454237413406372,
      "epoch": 1.3670174165457185,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004818297469798797,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19666333496570587,
      "num_tokens": 31280177.0,
      "step": 15070
    },
    {
      "entropy": 5.516110610961914,
      "epoch": 1.3674709724238026,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048181700644636314,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20633066445589066,
      "num_tokens": 31290545.0,
      "step": 15075
    },
    {
      "entropy": 5.434341859817505,
      "epoch": 1.3679245283018868,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048180426163581816,
      "loss": 4.9031,
      "mean_token_accuracy": 0.20979585647583007,
      "num_tokens": 31301143.0,
      "step": 15080
    },
    {
      "entropy": 5.4555473804473875,
      "epoch": 1.3683780841799709,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004817915125485083,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20275210589170456,
      "num_tokens": 31311071.0,
      "step": 15085
    },
    {
      "entropy": 5.551762104034424,
      "epoch": 1.3688316400580551,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048177875918469735,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19941299557685851,
      "num_tokens": 31322263.0,
      "step": 15090
    },
    {
      "entropy": 5.529943132400513,
      "epoch": 1.3692851959361394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048176600154464897,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1998817577958107,
      "num_tokens": 31332975.0,
      "step": 15095
    },
    {
      "entropy": 5.581736898422241,
      "epoch": 1.3697387518142234,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004817532396286269,
      "loss": 5.1303,
      "mean_token_accuracy": 0.1915099710226059,
      "num_tokens": 31345252.0,
      "step": 15100
    },
    {
      "entropy": 5.494141530990601,
      "epoch": 1.3701923076923077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004817404734368952,
      "loss": 4.9906,
      "mean_token_accuracy": 0.20686235874891282,
      "num_tokens": 31355479.0,
      "step": 15105
    },
    {
      "entropy": 5.5393835544586185,
      "epoch": 1.3706458635703918,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048172770296971785,
      "loss": 5.136,
      "mean_token_accuracy": 0.19656438529491424,
      "num_tokens": 31366020.0,
      "step": 15110
    },
    {
      "entropy": 5.513412237167358,
      "epoch": 1.371099419448476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004817149282273589,
      "loss": 5.046,
      "mean_token_accuracy": 0.1999151214957237,
      "num_tokens": 31376055.0,
      "step": 15115
    },
    {
      "entropy": 5.548714685440063,
      "epoch": 1.3715529753265603,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004817021492100825,
      "loss": 5.178,
      "mean_token_accuracy": 0.18876050859689714,
      "num_tokens": 31386179.0,
      "step": 15120
    },
    {
      "entropy": 5.466830825805664,
      "epoch": 1.3720065312046443,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000481689365918153,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20302161574363708,
      "num_tokens": 31395850.0,
      "step": 15125
    },
    {
      "entropy": 5.502423667907715,
      "epoch": 1.3724600870827286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004816765783518348,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21120330691337585,
      "num_tokens": 31405053.0,
      "step": 15130
    },
    {
      "entropy": 5.439601993560791,
      "epoch": 1.3729136429608126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004816637865113923,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20858091861009598,
      "num_tokens": 31415153.0,
      "step": 15135
    },
    {
      "entropy": 5.446499586105347,
      "epoch": 1.373367198838897,
      "grad_norm": 1.0,
      "learning_rate": 0.00048165099039708994,
      "loss": 5.0592,
      "mean_token_accuracy": 0.19852253049612045,
      "num_tokens": 31425451.0,
      "step": 15140
    },
    {
      "entropy": 5.445763492584229,
      "epoch": 1.3738207547169812,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004816381900091925,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20059703588485717,
      "num_tokens": 31436298.0,
      "step": 15145
    },
    {
      "entropy": 5.520010280609131,
      "epoch": 1.3742743105950654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048162538534796465,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20523983836174012,
      "num_tokens": 31446932.0,
      "step": 15150
    },
    {
      "entropy": 5.551976919174194,
      "epoch": 1.3747278664731495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004816125764136711,
      "loss": 5.0848,
      "mean_token_accuracy": 0.19165421277284622,
      "num_tokens": 31457582.0,
      "step": 15155
    },
    {
      "entropy": 5.560914659500122,
      "epoch": 1.3751814223512338,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004815997632065769,
      "loss": 5.0433,
      "mean_token_accuracy": 0.201973794400692,
      "num_tokens": 31468168.0,
      "step": 15160
    },
    {
      "entropy": 5.4446817398071286,
      "epoch": 1.3756349782293178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048158694572694694,
      "loss": 5.0329,
      "mean_token_accuracy": 0.207179257273674,
      "num_tokens": 31478524.0,
      "step": 15165
    },
    {
      "entropy": 5.456293964385987,
      "epoch": 1.376088534107402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048157412397504623,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20530054569244385,
      "num_tokens": 31488608.0,
      "step": 15170
    },
    {
      "entropy": 5.493011617660523,
      "epoch": 1.3765420899854863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004815612979511401,
      "loss": 5.0507,
      "mean_token_accuracy": 0.20859668999910355,
      "num_tokens": 31498984.0,
      "step": 15175
    },
    {
      "entropy": 5.5580322265625,
      "epoch": 1.3769956458635704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048154846765549353,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20434010922908782,
      "num_tokens": 31509110.0,
      "step": 15180
    },
    {
      "entropy": 5.479401540756226,
      "epoch": 1.3774492017416546,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048153563308837215,
      "loss": 5.0164,
      "mean_token_accuracy": 0.20211873799562455,
      "num_tokens": 31519266.0,
      "step": 15185
    },
    {
      "entropy": 5.491009759902954,
      "epoch": 1.3779027576197387,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048152279425004124,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20235521346330643,
      "num_tokens": 31529043.0,
      "step": 15190
    },
    {
      "entropy": 5.520250082015991,
      "epoch": 1.378356313497823,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004815099511407663,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20950424671173096,
      "num_tokens": 31538650.0,
      "step": 15195
    },
    {
      "entropy": 5.458363056182861,
      "epoch": 1.3788098693759072,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000481497103760813,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20308187007904052,
      "num_tokens": 31549392.0,
      "step": 15200
    },
    {
      "entropy": 5.550771903991699,
      "epoch": 1.3792634252539913,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048148425211044684,
      "loss": 5.1663,
      "mean_token_accuracy": 0.18810785859823226,
      "num_tokens": 31559999.0,
      "step": 15205
    },
    {
      "entropy": 5.559471702575683,
      "epoch": 1.3797169811320755,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004814713961899338,
      "loss": 5.0556,
      "mean_token_accuracy": 0.1991157650947571,
      "num_tokens": 31570108.0,
      "step": 15210
    },
    {
      "entropy": 5.515550661087036,
      "epoch": 1.3801705370101596,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048145853599953966,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20343615710735322,
      "num_tokens": 31579698.0,
      "step": 15215
    },
    {
      "entropy": 5.457998752593994,
      "epoch": 1.3806240928882438,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048144567153953036,
      "loss": 5.0727,
      "mean_token_accuracy": 0.19539187401533126,
      "num_tokens": 31589292.0,
      "step": 15220
    },
    {
      "entropy": 5.559013748168946,
      "epoch": 1.381077648766328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048143280281017197,
      "loss": 5.1465,
      "mean_token_accuracy": 0.19716491550207138,
      "num_tokens": 31599848.0,
      "step": 15225
    },
    {
      "entropy": 5.541344213485718,
      "epoch": 1.3815312046444121,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004814199298117307,
      "loss": 5.0811,
      "mean_token_accuracy": 0.19916218519210815,
      "num_tokens": 31611109.0,
      "step": 15230
    },
    {
      "entropy": 5.4703916072845455,
      "epoch": 1.3819847605224964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048140705254447256,
      "loss": 4.9954,
      "mean_token_accuracy": 0.203843155503273,
      "num_tokens": 31621140.0,
      "step": 15235
    },
    {
      "entropy": 5.583371543884278,
      "epoch": 1.3824383164005805,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048139417100866405,
      "loss": 5.0945,
      "mean_token_accuracy": 0.19653286635875702,
      "num_tokens": 31631268.0,
      "step": 15240
    },
    {
      "entropy": 5.551007652282715,
      "epoch": 1.3828918722786647,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048138128520457146,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20167967230081557,
      "num_tokens": 31640758.0,
      "step": 15245
    },
    {
      "entropy": 5.47036771774292,
      "epoch": 1.383345428156749,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004813683951324613,
      "loss": 5.0601,
      "mean_token_accuracy": 0.19761260747909545,
      "num_tokens": 31651060.0,
      "step": 15250
    },
    {
      "entropy": 5.516869258880615,
      "epoch": 1.383798984034833,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048135550079260014,
      "loss": 5.134,
      "mean_token_accuracy": 0.20071062743663787,
      "num_tokens": 31662282.0,
      "step": 15255
    },
    {
      "entropy": 5.504281425476075,
      "epoch": 1.3842525399129173,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004813426021852547,
      "loss": 5.0203,
      "mean_token_accuracy": 0.19429908096790313,
      "num_tokens": 31672364.0,
      "step": 15260
    },
    {
      "entropy": 5.517607593536377,
      "epoch": 1.3847060957910013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048132969931069156,
      "loss": 5.0738,
      "mean_token_accuracy": 0.1958681508898735,
      "num_tokens": 31683094.0,
      "step": 15265
    },
    {
      "entropy": 5.505269622802734,
      "epoch": 1.3851596516690856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004813167921691777,
      "loss": 5.0391,
      "mean_token_accuracy": 0.2022292897105217,
      "num_tokens": 31693087.0,
      "step": 15270
    },
    {
      "entropy": 5.5196010112762455,
      "epoch": 1.3856132075471699,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048130388076098003,
      "loss": 5.0301,
      "mean_token_accuracy": 0.19693872481584548,
      "num_tokens": 31703708.0,
      "step": 15275
    },
    {
      "entropy": 5.46891303062439,
      "epoch": 1.386066763425254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048129096508636547,
      "loss": 4.9565,
      "mean_token_accuracy": 0.211345899105072,
      "num_tokens": 31713640.0,
      "step": 15280
    },
    {
      "entropy": 5.512295389175415,
      "epoch": 1.3865203193033382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048127804514560126,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19729174673557281,
      "num_tokens": 31723526.0,
      "step": 15285
    },
    {
      "entropy": 5.444248962402344,
      "epoch": 1.3869738751814222,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048126512093895447,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20083393156528473,
      "num_tokens": 31733265.0,
      "step": 15290
    },
    {
      "entropy": 5.519571924209595,
      "epoch": 1.3874274310595065,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004812521924666925,
      "loss": 4.9972,
      "mean_token_accuracy": 0.20516951829195024,
      "num_tokens": 31743832.0,
      "step": 15295
    },
    {
      "entropy": 5.533500385284424,
      "epoch": 1.3878809869375908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004812392597290825,
      "loss": 5.0691,
      "mean_token_accuracy": 0.2090713858604431,
      "num_tokens": 31754443.0,
      "step": 15300
    },
    {
      "entropy": 5.4089892387390135,
      "epoch": 1.3883345428156748,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004812263227263921,
      "loss": 5.019,
      "mean_token_accuracy": 0.20233907401561738,
      "num_tokens": 31765304.0,
      "step": 15305
    },
    {
      "entropy": 5.474977779388428,
      "epoch": 1.388788098693759,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004812133814588888,
      "loss": 5.0712,
      "mean_token_accuracy": 0.19324804544448854,
      "num_tokens": 31775809.0,
      "step": 15310
    },
    {
      "entropy": 5.53752875328064,
      "epoch": 1.3892416545718431,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004812004359268402,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20598687529563903,
      "num_tokens": 31786271.0,
      "step": 15315
    },
    {
      "entropy": 5.527387714385986,
      "epoch": 1.3896952104499274,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004811874861305141,
      "loss": 5.0821,
      "mean_token_accuracy": 0.20123516917228698,
      "num_tokens": 31797358.0,
      "step": 15320
    },
    {
      "entropy": 5.475525903701782,
      "epoch": 1.3901487663280117,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004811745320701782,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20617760568857194,
      "num_tokens": 31808303.0,
      "step": 15325
    },
    {
      "entropy": 5.433736848831177,
      "epoch": 1.390602322206096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004811615737461004,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20380660891532898,
      "num_tokens": 31818837.0,
      "step": 15330
    },
    {
      "entropy": 5.55443115234375,
      "epoch": 1.39105587808418,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048114861115854877,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1971920147538185,
      "num_tokens": 31829729.0,
      "step": 15335
    },
    {
      "entropy": 5.499588584899902,
      "epoch": 1.3915094339622642,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048113564430779136,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20167743265628815,
      "num_tokens": 31839996.0,
      "step": 15340
    },
    {
      "entropy": 5.401311779022217,
      "epoch": 1.3919629898403483,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004811226731940962,
      "loss": 4.9661,
      "mean_token_accuracy": 0.20950472056865693,
      "num_tokens": 31850298.0,
      "step": 15345
    },
    {
      "entropy": 5.48508653640747,
      "epoch": 1.3924165457184325,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048110969781773173,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20816129744052886,
      "num_tokens": 31862170.0,
      "step": 15350
    },
    {
      "entropy": 5.5605326175689695,
      "epoch": 1.3928701015965168,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048109671817896614,
      "loss": 5.1177,
      "mean_token_accuracy": 0.20055409371852875,
      "num_tokens": 31872081.0,
      "step": 15355
    },
    {
      "entropy": 5.497061347961425,
      "epoch": 1.3933236574746009,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048108373427806797,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2065151587128639,
      "num_tokens": 31882715.0,
      "step": 15360
    },
    {
      "entropy": 5.526398038864135,
      "epoch": 1.3937772133526851,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048107074611530555,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20854706615209578,
      "num_tokens": 31891925.0,
      "step": 15365
    },
    {
      "entropy": 5.541769695281983,
      "epoch": 1.3942307692307692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048105775369094756,
      "loss": 5.0347,
      "mean_token_accuracy": 0.20411615669727326,
      "num_tokens": 31902320.0,
      "step": 15370
    },
    {
      "entropy": 5.45694899559021,
      "epoch": 1.3946843251088534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004810447570052628,
      "loss": 4.9559,
      "mean_token_accuracy": 0.20837390571832656,
      "num_tokens": 31912726.0,
      "step": 15375
    },
    {
      "entropy": 5.457587814331054,
      "epoch": 1.3951378809869377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048103175605851997,
      "loss": 5.0727,
      "mean_token_accuracy": 0.19942245334386827,
      "num_tokens": 31924039.0,
      "step": 15380
    },
    {
      "entropy": 5.43327054977417,
      "epoch": 1.3955914368650217,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004810187508509879,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2109137788414955,
      "num_tokens": 31933280.0,
      "step": 15385
    },
    {
      "entropy": 5.482158946990967,
      "epoch": 1.396044992743106,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004810057413829355,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19854603260755538,
      "num_tokens": 31943182.0,
      "step": 15390
    },
    {
      "entropy": 5.606261444091797,
      "epoch": 1.39649854862119,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048099272765463193,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20472536087036133,
      "num_tokens": 31953256.0,
      "step": 15395
    },
    {
      "entropy": 5.608113241195679,
      "epoch": 1.3969521044992743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004809797096663463,
      "loss": 5.1651,
      "mean_token_accuracy": 0.19737764447927475,
      "num_tokens": 31964231.0,
      "step": 15400
    },
    {
      "entropy": 5.514317989349365,
      "epoch": 1.3974056603773586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004809666874183477,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20222418010234833,
      "num_tokens": 31974322.0,
      "step": 15405
    },
    {
      "entropy": 5.4731401920318605,
      "epoch": 1.3978592162554426,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004809536609109056,
      "loss": 5.0691,
      "mean_token_accuracy": 0.2095978870987892,
      "num_tokens": 31984239.0,
      "step": 15410
    },
    {
      "entropy": 5.476027631759644,
      "epoch": 1.398312772133527,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048094063014428925,
      "loss": 5.0785,
      "mean_token_accuracy": 0.19442455768585204,
      "num_tokens": 31994366.0,
      "step": 15415
    },
    {
      "entropy": 5.566170167922974,
      "epoch": 1.398766328011611,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004809275951187682,
      "loss": 5.16,
      "mean_token_accuracy": 0.19522467702627183,
      "num_tokens": 32004921.0,
      "step": 15420
    },
    {
      "entropy": 5.466505813598633,
      "epoch": 1.3992198838896952,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048091455583461204,
      "loss": 5.0773,
      "mean_token_accuracy": 0.19962003231048583,
      "num_tokens": 32016385.0,
      "step": 15425
    },
    {
      "entropy": 5.520288276672363,
      "epoch": 1.3996734397677795,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048090151229209036,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20249414145946504,
      "num_tokens": 32027398.0,
      "step": 15430
    },
    {
      "entropy": 5.5420714855194095,
      "epoch": 1.4001269956458635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004808884644914729,
      "loss": 5.0636,
      "mean_token_accuracy": 0.20085171461105347,
      "num_tokens": 32038326.0,
      "step": 15435
    },
    {
      "entropy": 5.516499710083008,
      "epoch": 1.4005805515239478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004808754124330296,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19790878891944885,
      "num_tokens": 32048744.0,
      "step": 15440
    },
    {
      "entropy": 5.500717973709106,
      "epoch": 1.4010341074020318,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004808623561170302,
      "loss": 4.9933,
      "mean_token_accuracy": 0.20921683311462402,
      "num_tokens": 32059133.0,
      "step": 15445
    },
    {
      "entropy": 5.502972602844238,
      "epoch": 1.401487663280116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004808492955437449,
      "loss": 5.0357,
      "mean_token_accuracy": 0.20092233568429946,
      "num_tokens": 32069277.0,
      "step": 15450
    },
    {
      "entropy": 5.505456733703613,
      "epoch": 1.4019412191582004,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048083623071344374,
      "loss": 4.9932,
      "mean_token_accuracy": 0.20528948158025742,
      "num_tokens": 32079802.0,
      "step": 15455
    },
    {
      "entropy": 5.508525562286377,
      "epoch": 1.4023947750362844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048082316162639685,
      "loss": 5.1042,
      "mean_token_accuracy": 0.1987621620297432,
      "num_tokens": 32090962.0,
      "step": 15460
    },
    {
      "entropy": 5.478303909301758,
      "epoch": 1.4028483309143687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048081008828287447,
      "loss": 5.077,
      "mean_token_accuracy": 0.20106507390737532,
      "num_tokens": 32101058.0,
      "step": 15465
    },
    {
      "entropy": 5.55407304763794,
      "epoch": 1.4033018867924527,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048079701068314706,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20511925965547562,
      "num_tokens": 32111763.0,
      "step": 15470
    },
    {
      "entropy": 5.53937177658081,
      "epoch": 1.403755442670537,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048078392882748507,
      "loss": 5.1272,
      "mean_token_accuracy": 0.2001139998435974,
      "num_tokens": 32123556.0,
      "step": 15475
    },
    {
      "entropy": 5.495906591415405,
      "epoch": 1.4042089985486212,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048077084271615895,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20298214703798295,
      "num_tokens": 32134012.0,
      "step": 15480
    },
    {
      "entropy": 5.424602937698364,
      "epoch": 1.4046625544267053,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004807577523494394,
      "loss": 5.0046,
      "mean_token_accuracy": 0.2091123193502426,
      "num_tokens": 32144540.0,
      "step": 15485
    },
    {
      "entropy": 5.516141748428344,
      "epoch": 1.4051161103047896,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048074465772759707,
      "loss": 5.0977,
      "mean_token_accuracy": 0.20051607191562654,
      "num_tokens": 32154655.0,
      "step": 15490
    },
    {
      "entropy": 5.478973913192749,
      "epoch": 1.4055696661828736,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048073155885090276,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20102933645248414,
      "num_tokens": 32165191.0,
      "step": 15495
    },
    {
      "entropy": 5.515355730056763,
      "epoch": 1.4060232220609579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048071845571962733,
      "loss": 5.0933,
      "mean_token_accuracy": 0.1930096685886383,
      "num_tokens": 32175961.0,
      "step": 15500
    },
    {
      "entropy": 5.549034500122071,
      "epoch": 1.4064767779390421,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004807053483340419,
      "loss": 5.0587,
      "mean_token_accuracy": 0.2007862851023674,
      "num_tokens": 32187073.0,
      "step": 15505
    },
    {
      "entropy": 5.495439291000366,
      "epoch": 1.4069303338171264,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004806922366944175,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19969787448644638,
      "num_tokens": 32197426.0,
      "step": 15510
    },
    {
      "entropy": 5.536189460754395,
      "epoch": 1.4073838896952104,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048067912080102513,
      "loss": 5.032,
      "mean_token_accuracy": 0.2082691878080368,
      "num_tokens": 32206847.0,
      "step": 15515
    },
    {
      "entropy": 5.478679609298706,
      "epoch": 1.4078374455732947,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004806660006541362,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1986638382077217,
      "num_tokens": 32217647.0,
      "step": 15520
    },
    {
      "entropy": 5.412083864212036,
      "epoch": 1.4082910014513788,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004806528762540219,
      "loss": 4.9054,
      "mean_token_accuracy": 0.21444422006607056,
      "num_tokens": 32227133.0,
      "step": 15525
    },
    {
      "entropy": 5.521481990814209,
      "epoch": 1.408744557329463,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048063974760095374,
      "loss": 5.1197,
      "mean_token_accuracy": 0.19907931387424468,
      "num_tokens": 32237378.0,
      "step": 15530
    },
    {
      "entropy": 5.450902986526489,
      "epoch": 1.4091981132075473,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048062661469520327,
      "loss": 4.9959,
      "mean_token_accuracy": 0.2066905677318573,
      "num_tokens": 32247213.0,
      "step": 15535
    },
    {
      "entropy": 5.555349111557007,
      "epoch": 1.4096516690856313,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048061347753704193,
      "loss": 5.0782,
      "mean_token_accuracy": 0.2022297203540802,
      "num_tokens": 32257100.0,
      "step": 15540
    },
    {
      "entropy": 5.5293927669525145,
      "epoch": 1.4101052249637156,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004806003361267415,
      "loss": 5.0423,
      "mean_token_accuracy": 0.1960933342576027,
      "num_tokens": 32267456.0,
      "step": 15545
    },
    {
      "entropy": 5.456585788726807,
      "epoch": 1.4105587808417996,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004805871904645737,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2102513313293457,
      "num_tokens": 32278220.0,
      "step": 15550
    },
    {
      "entropy": 5.444500207901001,
      "epoch": 1.411012336719884,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004805740405508103,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2062092110514641,
      "num_tokens": 32289036.0,
      "step": 15555
    },
    {
      "entropy": 5.468543577194214,
      "epoch": 1.4114658925979682,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004805608863857236,
      "loss": 4.9818,
      "mean_token_accuracy": 0.19995850175619126,
      "num_tokens": 32298909.0,
      "step": 15560
    },
    {
      "entropy": 5.513676548004151,
      "epoch": 1.4119194484760522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048054772796958517,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20487444698810578,
      "num_tokens": 32309851.0,
      "step": 15565
    },
    {
      "entropy": 5.464004564285278,
      "epoch": 1.4123730043541365,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004805345653026675,
      "loss": 4.99,
      "mean_token_accuracy": 0.21017162203788758,
      "num_tokens": 32319336.0,
      "step": 15570
    },
    {
      "entropy": 5.510065793991089,
      "epoch": 1.4128265602322205,
      "grad_norm": 1.0,
      "learning_rate": 0.00048052139838524263,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20483591854572297,
      "num_tokens": 32329898.0,
      "step": 15575
    },
    {
      "entropy": 5.492889308929444,
      "epoch": 1.4132801161103048,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048050822721758274,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20117055475711823,
      "num_tokens": 32340100.0,
      "step": 15580
    },
    {
      "entropy": 5.4790565967559814,
      "epoch": 1.413733671988389,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004804950517999604,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19834206104278565,
      "num_tokens": 32349663.0,
      "step": 15585
    },
    {
      "entropy": 5.524758529663086,
      "epoch": 1.414187227866473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004804818721326481,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20059896111488343,
      "num_tokens": 32360461.0,
      "step": 15590
    },
    {
      "entropy": 5.545775079727173,
      "epoch": 1.4146407837445574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004804686882159183,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20205838233232498,
      "num_tokens": 32370910.0,
      "step": 15595
    },
    {
      "entropy": 5.578698396682739,
      "epoch": 1.4150943396226414,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048045550005004355,
      "loss": 5.0914,
      "mean_token_accuracy": 0.20864028632640838,
      "num_tokens": 32381669.0,
      "step": 15600
    },
    {
      "entropy": 5.462166213989258,
      "epoch": 1.4155478955007257,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004804423076352968,
      "loss": 5.0099,
      "mean_token_accuracy": 0.19899781793355942,
      "num_tokens": 32392024.0,
      "step": 15605
    },
    {
      "entropy": 5.416386508941651,
      "epoch": 1.41600145137881,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048042911097195075,
      "loss": 5.0229,
      "mean_token_accuracy": 0.20628166496753692,
      "num_tokens": 32401872.0,
      "step": 15610
    },
    {
      "entropy": 5.426913595199585,
      "epoch": 1.416455007256894,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004804159100602784,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2084158256649971,
      "num_tokens": 32411749.0,
      "step": 15615
    },
    {
      "entropy": 5.584346342086792,
      "epoch": 1.4169085631349783,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048040270490055264,
      "loss": 5.1609,
      "mean_token_accuracy": 0.19357605278491974,
      "num_tokens": 32421358.0,
      "step": 15620
    },
    {
      "entropy": 5.543413543701172,
      "epoch": 1.4173621190130623,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004803894954930465,
      "loss": 5.055,
      "mean_token_accuracy": 0.19952864944934845,
      "num_tokens": 32432322.0,
      "step": 15625
    },
    {
      "entropy": 5.500397825241089,
      "epoch": 1.4178156748911466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004803762818380334,
      "loss": 5.0745,
      "mean_token_accuracy": 0.19714537113904954,
      "num_tokens": 32440930.0,
      "step": 15630
    },
    {
      "entropy": 5.511843538284301,
      "epoch": 1.4182692307692308,
      "grad_norm": 1.0,
      "learning_rate": 0.00048036306393578644,
      "loss": 5.0697,
      "mean_token_accuracy": 0.19900647252798082,
      "num_tokens": 32451124.0,
      "step": 15635
    },
    {
      "entropy": 5.549473667144776,
      "epoch": 1.4187227866473149,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004803498417865789,
      "loss": 5.0478,
      "mean_token_accuracy": 0.19772396683692933,
      "num_tokens": 32460636.0,
      "step": 15640
    },
    {
      "entropy": 5.521689081192017,
      "epoch": 1.4191763425253991,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004803366153906844,
      "loss": 5.0847,
      "mean_token_accuracy": 0.19507592916488647,
      "num_tokens": 32470501.0,
      "step": 15645
    },
    {
      "entropy": 5.53672423362732,
      "epoch": 1.4196298984034832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004803233847483763,
      "loss": 5.1524,
      "mean_token_accuracy": 0.19808802306652068,
      "num_tokens": 32481559.0,
      "step": 15650
    },
    {
      "entropy": 5.6585461616516115,
      "epoch": 1.4200834542815675,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048031014985992825,
      "loss": 5.2276,
      "mean_token_accuracy": 0.19318990856409074,
      "num_tokens": 32493000.0,
      "step": 15655
    },
    {
      "entropy": 5.520326471328735,
      "epoch": 1.4205370101596517,
      "grad_norm": 1.03125,
      "learning_rate": 0.000480296910725614,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21055828481912614,
      "num_tokens": 32502773.0,
      "step": 15660
    },
    {
      "entropy": 5.480014324188232,
      "epoch": 1.4209905660377358,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004802836673457074,
      "loss": 5.1155,
      "mean_token_accuracy": 0.19522669911384583,
      "num_tokens": 32513430.0,
      "step": 15665
    },
    {
      "entropy": 5.4377953052520756,
      "epoch": 1.42144412191582,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048027041972048225,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20099482089281082,
      "num_tokens": 32524409.0,
      "step": 15670
    },
    {
      "entropy": 5.553106594085693,
      "epoch": 1.421897677793904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004802571678502124,
      "loss": 5.0725,
      "mean_token_accuracy": 0.19653349220752717,
      "num_tokens": 32534696.0,
      "step": 15675
    },
    {
      "entropy": 5.576407051086425,
      "epoch": 1.4223512336719883,
      "grad_norm": 0.9375,
      "learning_rate": 0.000480243911735172,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20046114325523376,
      "num_tokens": 32544782.0,
      "step": 15680
    },
    {
      "entropy": 5.475587797164917,
      "epoch": 1.4228047895500726,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048023065137563534,
      "loss": 5.0548,
      "mean_token_accuracy": 0.19665246605873107,
      "num_tokens": 32554818.0,
      "step": 15685
    },
    {
      "entropy": 5.473144388198852,
      "epoch": 1.4232583454281569,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004802173867718764,
      "loss": 5.0834,
      "mean_token_accuracy": 0.20439634025096892,
      "num_tokens": 32565007.0,
      "step": 15690
    },
    {
      "entropy": 5.523764944076538,
      "epoch": 1.423711901306241,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004802041179241697,
      "loss": 5.0382,
      "mean_token_accuracy": 0.19968950599431992,
      "num_tokens": 32575657.0,
      "step": 15695
    },
    {
      "entropy": 5.4944602966308596,
      "epoch": 1.4241654571843252,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004801908448327895,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19597385972738265,
      "num_tokens": 32586676.0,
      "step": 15700
    },
    {
      "entropy": 5.4961262226104735,
      "epoch": 1.4246190130624092,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004801775674980103,
      "loss": 5.021,
      "mean_token_accuracy": 0.20194448232650758,
      "num_tokens": 32596191.0,
      "step": 15705
    },
    {
      "entropy": 5.567626953125,
      "epoch": 1.4250725689404935,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048016428592010676,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1960490435361862,
      "num_tokens": 32607178.0,
      "step": 15710
    },
    {
      "entropy": 5.550439023971558,
      "epoch": 1.4255261248185778,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048015100009935347,
      "loss": 5.0184,
      "mean_token_accuracy": 0.2020075500011444,
      "num_tokens": 32617301.0,
      "step": 15715
    },
    {
      "entropy": 5.430491018295288,
      "epoch": 1.4259796806966618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004801377100360253,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2087884947657585,
      "num_tokens": 32627395.0,
      "step": 15720
    },
    {
      "entropy": 5.501881122589111,
      "epoch": 1.426433236574746,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000480124415730397,
      "loss": 5.0287,
      "mean_token_accuracy": 0.206740340590477,
      "num_tokens": 32638460.0,
      "step": 15725
    },
    {
      "entropy": 5.589461088180542,
      "epoch": 1.4268867924528301,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004801111171827434,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20110534280538558,
      "num_tokens": 32649319.0,
      "step": 15730
    },
    {
      "entropy": 5.498560190200806,
      "epoch": 1.4273403483309144,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004800978143933398,
      "loss": 5.0293,
      "mean_token_accuracy": 0.20413393080234526,
      "num_tokens": 32659976.0,
      "step": 15735
    },
    {
      "entropy": 5.433808660507202,
      "epoch": 1.4277939042089987,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048008450736246094,
      "loss": 4.9683,
      "mean_token_accuracy": 0.20098088234663009,
      "num_tokens": 32670772.0,
      "step": 15740
    },
    {
      "entropy": 5.5375607967376705,
      "epoch": 1.4282474600870827,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048007119609038237,
      "loss": 5.0968,
      "mean_token_accuracy": 0.19511828124523162,
      "num_tokens": 32682368.0,
      "step": 15745
    },
    {
      "entropy": 5.439204216003418,
      "epoch": 1.428701015965167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048005788057737907,
      "loss": 4.9274,
      "mean_token_accuracy": 0.20723052322864532,
      "num_tokens": 32692649.0,
      "step": 15750
    },
    {
      "entropy": 5.533323192596436,
      "epoch": 1.429154571843251,
      "grad_norm": 1.0,
      "learning_rate": 0.0004800445608237266,
      "loss": 5.1454,
      "mean_token_accuracy": 0.20151389092206956,
      "num_tokens": 32704075.0,
      "step": 15755
    },
    {
      "entropy": 5.5563231945037845,
      "epoch": 1.4296081277213353,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048003123682970045,
      "loss": 5.1597,
      "mean_token_accuracy": 0.19443312734365464,
      "num_tokens": 32714670.0,
      "step": 15760
    },
    {
      "entropy": 5.512262058258057,
      "epoch": 1.4300616835994195,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048001790859557596,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2086060270667076,
      "num_tokens": 32724601.0,
      "step": 15765
    },
    {
      "entropy": 5.458773851394653,
      "epoch": 1.4305152394775036,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004800045761216289,
      "loss": 4.9602,
      "mean_token_accuracy": 0.20292164236307145,
      "num_tokens": 32735015.0,
      "step": 15770
    },
    {
      "entropy": 5.500267553329468,
      "epoch": 1.4309687953555879,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000479991239408135,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20426508635282517,
      "num_tokens": 32745899.0,
      "step": 15775
    },
    {
      "entropy": 5.536904525756836,
      "epoch": 1.431422351233672,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047997789845537003,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20640725493431092,
      "num_tokens": 32756052.0,
      "step": 15780
    },
    {
      "entropy": 5.4844502925872805,
      "epoch": 1.4318759071117562,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004799645532636098,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2041822388768196,
      "num_tokens": 32767118.0,
      "step": 15785
    },
    {
      "entropy": 5.410877799987793,
      "epoch": 1.4323294629898404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047995120383313044,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21355968564748765,
      "num_tokens": 32777069.0,
      "step": 15790
    },
    {
      "entropy": 5.477860069274902,
      "epoch": 1.4327830188679245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047993785016420796,
      "loss": 5.1263,
      "mean_token_accuracy": 0.1909132644534111,
      "num_tokens": 32786967.0,
      "step": 15795
    },
    {
      "entropy": 5.523899364471435,
      "epoch": 1.4332365747460087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047992449225711854,
      "loss": 5.002,
      "mean_token_accuracy": 0.20324475318193436,
      "num_tokens": 32797563.0,
      "step": 15800
    },
    {
      "entropy": 5.5151763439178465,
      "epoch": 1.4336901306240928,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047991113011213837,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20701471120119094,
      "num_tokens": 32808730.0,
      "step": 15805
    },
    {
      "entropy": 5.444163513183594,
      "epoch": 1.434143686502177,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004798977637295438,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20792941600084305,
      "num_tokens": 32818050.0,
      "step": 15810
    },
    {
      "entropy": 5.459983968734742,
      "epoch": 1.4345972423802613,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798843931096113,
      "loss": 5.0233,
      "mean_token_accuracy": 0.20102835297584534,
      "num_tokens": 32828993.0,
      "step": 15815
    },
    {
      "entropy": 5.4861907958984375,
      "epoch": 1.4350507982583454,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004798710182526173,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20848754793405533,
      "num_tokens": 32839331.0,
      "step": 15820
    },
    {
      "entropy": 5.5293779373168945,
      "epoch": 1.4355043541364296,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047985763915883847,
      "loss": 5.0278,
      "mean_token_accuracy": 0.19959765970706939,
      "num_tokens": 32849946.0,
      "step": 15825
    },
    {
      "entropy": 5.438718461990357,
      "epoch": 1.4359579100145137,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798442558285514,
      "loss": 4.965,
      "mean_token_accuracy": 0.2041826367378235,
      "num_tokens": 32859960.0,
      "step": 15830
    },
    {
      "entropy": 5.485565805435181,
      "epoch": 1.436411465892598,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004798308682620329,
      "loss": 4.9937,
      "mean_token_accuracy": 0.19982496798038482,
      "num_tokens": 32870317.0,
      "step": 15835
    },
    {
      "entropy": 5.587348508834839,
      "epoch": 1.4368650217706822,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047981747645955996,
      "loss": 5.1683,
      "mean_token_accuracy": 0.19669103622436523,
      "num_tokens": 32880906.0,
      "step": 15840
    },
    {
      "entropy": 5.593022584915161,
      "epoch": 1.4373185776487662,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004798040804214094,
      "loss": 5.1978,
      "mean_token_accuracy": 0.1926955610513687,
      "num_tokens": 32891038.0,
      "step": 15845
    },
    {
      "entropy": 5.585289764404297,
      "epoch": 1.4377721335268505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797906801478582,
      "loss": 5.136,
      "mean_token_accuracy": 0.20103828907012938,
      "num_tokens": 32901013.0,
      "step": 15850
    },
    {
      "entropy": 5.58568844795227,
      "epoch": 1.4382256894049346,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004797772756391836,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1981079399585724,
      "num_tokens": 32911592.0,
      "step": 15855
    },
    {
      "entropy": 5.4849381923675535,
      "epoch": 1.4386792452830188,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004797638668956627,
      "loss": 4.9591,
      "mean_token_accuracy": 0.20855915248394014,
      "num_tokens": 32922087.0,
      "step": 15860
    },
    {
      "entropy": 5.482039165496826,
      "epoch": 1.439132801161103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004797504539175729,
      "loss": 5.0489,
      "mean_token_accuracy": 0.20400669425725937,
      "num_tokens": 32932039.0,
      "step": 15865
    },
    {
      "entropy": 5.499445915222168,
      "epoch": 1.4395863570391874,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004797370367051915,
      "loss": 5.0049,
      "mean_token_accuracy": 0.2056662067770958,
      "num_tokens": 32941896.0,
      "step": 15870
    },
    {
      "entropy": 5.50581727027893,
      "epoch": 1.4400399129172714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000479723615258796,
      "loss": 5.0322,
      "mean_token_accuracy": 0.19427276104688646,
      "num_tokens": 32951675.0,
      "step": 15875
    },
    {
      "entropy": 5.531674242019653,
      "epoch": 1.4404934687953557,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047971018957866396,
      "loss": 5.1049,
      "mean_token_accuracy": 0.20822160840034484,
      "num_tokens": 32961378.0,
      "step": 15880
    },
    {
      "entropy": 5.497912979125976,
      "epoch": 1.4409470246734397,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047969675966507314,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20277913063764572,
      "num_tokens": 32971770.0,
      "step": 15885
    },
    {
      "entropy": 5.465924024581909,
      "epoch": 1.441400580551524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000479683325518301,
      "loss": 5.0215,
      "mean_token_accuracy": 0.2001161202788353,
      "num_tokens": 32982934.0,
      "step": 15890
    },
    {
      "entropy": 5.492756032943726,
      "epoch": 1.4418541364296082,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004796698871386257,
      "loss": 5.0294,
      "mean_token_accuracy": 0.1993359848856926,
      "num_tokens": 32993252.0,
      "step": 15895
    },
    {
      "entropy": 5.536276340484619,
      "epoch": 1.4423076923076923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004796564445263249,
      "loss": 5.0486,
      "mean_token_accuracy": 0.2013053610920906,
      "num_tokens": 33003610.0,
      "step": 15900
    },
    {
      "entropy": 5.499846506118774,
      "epoch": 1.4427612481857766,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047964299768167665,
      "loss": 5.0061,
      "mean_token_accuracy": 0.2022084638476372,
      "num_tokens": 33014037.0,
      "step": 15905
    },
    {
      "entropy": 5.495324516296387,
      "epoch": 1.4432148040638606,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004796295466049591,
      "loss": 5.031,
      "mean_token_accuracy": 0.1982549697160721,
      "num_tokens": 33023299.0,
      "step": 15910
    },
    {
      "entropy": 5.528369045257568,
      "epoch": 1.4436683599419449,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047961609129645036,
      "loss": 5.085,
      "mean_token_accuracy": 0.19658140540122987,
      "num_tokens": 33034194.0,
      "step": 15915
    },
    {
      "entropy": 5.492931365966797,
      "epoch": 1.4441219158200291,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047960263175642867,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21053073406219483,
      "num_tokens": 33044715.0,
      "step": 15920
    },
    {
      "entropy": 5.429548358917236,
      "epoch": 1.4445754716981132,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004795891679851724,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21614315807819368,
      "num_tokens": 33055214.0,
      "step": 15925
    },
    {
      "entropy": 5.474163293838501,
      "epoch": 1.4450290275761974,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004795756999829601,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20264820456504823,
      "num_tokens": 33065026.0,
      "step": 15930
    },
    {
      "entropy": 5.55479736328125,
      "epoch": 1.4454825834542815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047956222775007016,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1960279420018196,
      "num_tokens": 33075138.0,
      "step": 15935
    },
    {
      "entropy": 5.525855541229248,
      "epoch": 1.4459361393323658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004795487512867812,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20468559861183167,
      "num_tokens": 33086625.0,
      "step": 15940
    },
    {
      "entropy": 5.53939790725708,
      "epoch": 1.44638969521045,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047953527059337194,
      "loss": 5.0861,
      "mean_token_accuracy": 0.19433015435934067,
      "num_tokens": 33097076.0,
      "step": 15945
    },
    {
      "entropy": 5.514823246002197,
      "epoch": 1.446843251088534,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004795217856701212,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20547079890966416,
      "num_tokens": 33107168.0,
      "step": 15950
    },
    {
      "entropy": 5.537642621994019,
      "epoch": 1.4472968069666183,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047950829651730774,
      "loss": 5.0654,
      "mean_token_accuracy": 0.20469075739383696,
      "num_tokens": 33117862.0,
      "step": 15955
    },
    {
      "entropy": 5.523198652267456,
      "epoch": 1.4477503628447024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004794948031352107,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20375883877277373,
      "num_tokens": 33127144.0,
      "step": 15960
    },
    {
      "entropy": 5.587871456146241,
      "epoch": 1.4482039187227866,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047948130552410897,
      "loss": 5.1368,
      "mean_token_accuracy": 0.19235823452472686,
      "num_tokens": 33139180.0,
      "step": 15965
    },
    {
      "entropy": 5.546544790267944,
      "epoch": 1.448657474600871,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047946780368428175,
      "loss": 5.0157,
      "mean_token_accuracy": 0.1992461785674095,
      "num_tokens": 33149467.0,
      "step": 15970
    },
    {
      "entropy": 5.458764553070068,
      "epoch": 1.449111030478955,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004794542976160081,
      "loss": 4.8907,
      "mean_token_accuracy": 0.21425037384033202,
      "num_tokens": 33158752.0,
      "step": 15975
    },
    {
      "entropy": 5.4563695907592775,
      "epoch": 1.4495645863570392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047944078731956765,
      "loss": 4.978,
      "mean_token_accuracy": 0.2071639657020569,
      "num_tokens": 33168242.0,
      "step": 15980
    },
    {
      "entropy": 5.450711584091186,
      "epoch": 1.4500181422351233,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047942727279523957,
      "loss": 4.995,
      "mean_token_accuracy": 0.20726195424795152,
      "num_tokens": 33179134.0,
      "step": 15985
    },
    {
      "entropy": 5.507195711135864,
      "epoch": 1.4504716981132075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004794137540433034,
      "loss": 5.1177,
      "mean_token_accuracy": 0.19952171742916108,
      "num_tokens": 33188388.0,
      "step": 15990
    },
    {
      "entropy": 5.557099103927612,
      "epoch": 1.4509252539912918,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004794002310640386,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2005619153380394,
      "num_tokens": 33198750.0,
      "step": 15995
    },
    {
      "entropy": 5.621443796157837,
      "epoch": 1.4513788098693758,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047938670385772503,
      "loss": 5.1347,
      "mean_token_accuracy": 0.19036090821027757,
      "num_tokens": 33209532.0,
      "step": 16000
    },
    {
      "entropy": 5.533468580245971,
      "epoch": 1.45183236574746,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047937317242464244,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20237762480974197,
      "num_tokens": 33219675.0,
      "step": 16005
    },
    {
      "entropy": 5.518578052520752,
      "epoch": 1.4522859216255442,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004793596367650704,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19858584105968474,
      "num_tokens": 33230919.0,
      "step": 16010
    },
    {
      "entropy": 5.495726633071899,
      "epoch": 1.4527394775036284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004793460968792891,
      "loss": 4.9979,
      "mean_token_accuracy": 0.20685409456491471,
      "num_tokens": 33241146.0,
      "step": 16015
    },
    {
      "entropy": 5.535656595230103,
      "epoch": 1.4531930333817127,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047933255276757836,
      "loss": 5.0513,
      "mean_token_accuracy": 0.19511209279298783,
      "num_tokens": 33251434.0,
      "step": 16020
    },
    {
      "entropy": 5.569311809539795,
      "epoch": 1.4536465892597967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004793190044302184,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19259205162525178,
      "num_tokens": 33261868.0,
      "step": 16025
    },
    {
      "entropy": 5.501216983795166,
      "epoch": 1.454100145137881,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047930545186748936,
      "loss": 4.9708,
      "mean_token_accuracy": 0.2036798745393753,
      "num_tokens": 33271798.0,
      "step": 16030
    },
    {
      "entropy": 5.495710563659668,
      "epoch": 1.454553701015965,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047929189507967153,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20195723026990892,
      "num_tokens": 33282614.0,
      "step": 16035
    },
    {
      "entropy": 5.459695625305176,
      "epoch": 1.4550072568940493,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047927833406704516,
      "loss": 4.9892,
      "mean_token_accuracy": 0.19896757304668428,
      "num_tokens": 33292906.0,
      "step": 16040
    },
    {
      "entropy": 5.509294652938843,
      "epoch": 1.4554608127721336,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004792647688298909,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20415859073400497,
      "num_tokens": 33302720.0,
      "step": 16045
    },
    {
      "entropy": 5.459511947631836,
      "epoch": 1.4559143686502178,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004792511993684891,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2102614790201187,
      "num_tokens": 33312959.0,
      "step": 16050
    },
    {
      "entropy": 5.46536283493042,
      "epoch": 1.4563679245283019,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047923762568312053,
      "loss": 4.9342,
      "mean_token_accuracy": 0.20821652561426163,
      "num_tokens": 33322737.0,
      "step": 16055
    },
    {
      "entropy": 5.4693663120269775,
      "epoch": 1.4568214804063861,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004792240477740657,
      "loss": 4.9625,
      "mean_token_accuracy": 0.21078547537326814,
      "num_tokens": 33333185.0,
      "step": 16060
    },
    {
      "entropy": 5.519652271270752,
      "epoch": 1.4572750362844702,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004792104656416056,
      "loss": 5.0883,
      "mean_token_accuracy": 0.1959143102169037,
      "num_tokens": 33343414.0,
      "step": 16065
    },
    {
      "entropy": 5.496297025680542,
      "epoch": 1.4577285921625545,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047919687928602104,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2069736033678055,
      "num_tokens": 33352527.0,
      "step": 16070
    },
    {
      "entropy": 5.553343343734741,
      "epoch": 1.4581821480406387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047918328870759293,
      "loss": 5.0628,
      "mean_token_accuracy": 0.19901107400655746,
      "num_tokens": 33363526.0,
      "step": 16075
    },
    {
      "entropy": 5.490126514434815,
      "epoch": 1.4586357039187228,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004791696939066024,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20202584862709044,
      "num_tokens": 33374302.0,
      "step": 16080
    },
    {
      "entropy": 5.5225855827331545,
      "epoch": 1.459089259796807,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791560948833306,
      "loss": 5.0604,
      "mean_token_accuracy": 0.19944061189889908,
      "num_tokens": 33384786.0,
      "step": 16085
    },
    {
      "entropy": 5.548208236694336,
      "epoch": 1.459542815674891,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004791424916380586,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2090217351913452,
      "num_tokens": 33394464.0,
      "step": 16090
    },
    {
      "entropy": 5.509014225006103,
      "epoch": 1.4599963715529753,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000479128884171068,
      "loss": 5.0608,
      "mean_token_accuracy": 0.20015767216682434,
      "num_tokens": 33405476.0,
      "step": 16095
    },
    {
      "entropy": 5.522106266021728,
      "epoch": 1.4604499274310596,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047911527248264003,
      "loss": 5.0873,
      "mean_token_accuracy": 0.20099602192640303,
      "num_tokens": 33415758.0,
      "step": 16100
    },
    {
      "entropy": 5.497673892974854,
      "epoch": 1.4609034833091437,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047910165657305613,
      "loss": 5.06,
      "mean_token_accuracy": 0.19588928818702697,
      "num_tokens": 33424845.0,
      "step": 16105
    },
    {
      "entropy": 5.547488403320313,
      "epoch": 1.461357039187228,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047908803644259804,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20185463279485702,
      "num_tokens": 33435928.0,
      "step": 16110
    },
    {
      "entropy": 5.537172937393189,
      "epoch": 1.461810595065312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047907441209154726,
      "loss": 4.9964,
      "mean_token_accuracy": 0.19959668070077896,
      "num_tokens": 33447307.0,
      "step": 16115
    },
    {
      "entropy": 5.52874789237976,
      "epoch": 1.4622641509433962,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047906078352018575,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19590460509061813,
      "num_tokens": 33458116.0,
      "step": 16120
    },
    {
      "entropy": 5.383693361282349,
      "epoch": 1.4627177068214805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004790471507287952,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2084435611963272,
      "num_tokens": 33469065.0,
      "step": 16125
    },
    {
      "entropy": 5.535000514984131,
      "epoch": 1.4631712626995645,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047903351371765757,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20379114598035813,
      "num_tokens": 33479118.0,
      "step": 16130
    },
    {
      "entropy": 5.554023694992066,
      "epoch": 1.4636248185776488,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047901987248705484,
      "loss": 5.064,
      "mean_token_accuracy": 0.19928945004940032,
      "num_tokens": 33489642.0,
      "step": 16135
    },
    {
      "entropy": 5.477730655670166,
      "epoch": 1.4640783744557329,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004790062270372693,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19506554156541825,
      "num_tokens": 33499771.0,
      "step": 16140
    },
    {
      "entropy": 5.470936298370361,
      "epoch": 1.4645319303338171,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004789925773685828,
      "loss": 5.027,
      "mean_token_accuracy": 0.19897197037935258,
      "num_tokens": 33509975.0,
      "step": 16145
    },
    {
      "entropy": 5.5395927906036375,
      "epoch": 1.4649854862119014,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004789789234812779,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1954951375722885,
      "num_tokens": 33520017.0,
      "step": 16150
    },
    {
      "entropy": 5.571431016921997,
      "epoch": 1.4654390420899854,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047896526537563686,
      "loss": 5.1316,
      "mean_token_accuracy": 0.1930692121386528,
      "num_tokens": 33531977.0,
      "step": 16155
    },
    {
      "entropy": 5.512016248703003,
      "epoch": 1.4658925979680697,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047895160305194223,
      "loss": 5.0951,
      "mean_token_accuracy": 0.19491859078407286,
      "num_tokens": 33543660.0,
      "step": 16160
    },
    {
      "entropy": 5.518211936950683,
      "epoch": 1.4663461538461537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789379365104764,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2014843299984932,
      "num_tokens": 33553814.0,
      "step": 16165
    },
    {
      "entropy": 5.52564640045166,
      "epoch": 1.466799709724238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004789242657515221,
      "loss": 5.0457,
      "mean_token_accuracy": 0.2034329056739807,
      "num_tokens": 33564250.0,
      "step": 16170
    },
    {
      "entropy": 5.5023092269897464,
      "epoch": 1.4672532656023223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004789105907753621,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20267768502235411,
      "num_tokens": 33574755.0,
      "step": 16175
    },
    {
      "entropy": 5.480043220520019,
      "epoch": 1.4677068214804063,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000478896911582279,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2060309335589409,
      "num_tokens": 33585268.0,
      "step": 16180
    },
    {
      "entropy": 5.487579488754273,
      "epoch": 1.4681603773584906,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004788832281725558,
      "loss": 5.0891,
      "mean_token_accuracy": 0.2038735941052437,
      "num_tokens": 33595753.0,
      "step": 16185
    },
    {
      "entropy": 5.461152362823486,
      "epoch": 1.4686139332365746,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004788695405464755,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20444866567850112,
      "num_tokens": 33605270.0,
      "step": 16190
    },
    {
      "entropy": 5.515757703781128,
      "epoch": 1.469067489114659,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047885584870432127,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20079802423715593,
      "num_tokens": 33615096.0,
      "step": 16195
    },
    {
      "entropy": 5.5126025676727295,
      "epoch": 1.4695210449927432,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047884215264637606,
      "loss": 4.968,
      "mean_token_accuracy": 0.20840421617031096,
      "num_tokens": 33624770.0,
      "step": 16200
    },
    {
      "entropy": 5.545363903045654,
      "epoch": 1.4699746008708272,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004788284523729232,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19470954835414886,
      "num_tokens": 33634545.0,
      "step": 16205
    },
    {
      "entropy": 5.569434070587159,
      "epoch": 1.4704281567489115,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047881474788424595,
      "loss": 5.0661,
      "mean_token_accuracy": 0.19960723966360092,
      "num_tokens": 33644950.0,
      "step": 16210
    },
    {
      "entropy": 5.549499034881592,
      "epoch": 1.4708817126269955,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047880103918062776,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20500340163707734,
      "num_tokens": 33654369.0,
      "step": 16215
    },
    {
      "entropy": 5.4006843090057375,
      "epoch": 1.4713352685050798,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047878732626235225,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19897183328866958,
      "num_tokens": 33664687.0,
      "step": 16220
    },
    {
      "entropy": 5.475517129898071,
      "epoch": 1.471788824383164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004787736091297028,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2159730851650238,
      "num_tokens": 33675832.0,
      "step": 16225
    },
    {
      "entropy": 5.559790515899659,
      "epoch": 1.4722423802612483,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004787598877829633,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20152770280838012,
      "num_tokens": 33686788.0,
      "step": 16230
    },
    {
      "entropy": 5.472647285461425,
      "epoch": 1.4726959361393324,
      "grad_norm": 1.125,
      "learning_rate": 0.0004787461622224173,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20888569504022597,
      "num_tokens": 33696384.0,
      "step": 16235
    },
    {
      "entropy": 5.463289499282837,
      "epoch": 1.4731494920174166,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004787324324483488,
      "loss": 5.0216,
      "mean_token_accuracy": 0.2022559866309166,
      "num_tokens": 33707758.0,
      "step": 16240
    },
    {
      "entropy": 5.502991962432861,
      "epoch": 1.4736030478955007,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004787186984610416,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2065095692873001,
      "num_tokens": 33717747.0,
      "step": 16245
    },
    {
      "entropy": 5.4916908740997314,
      "epoch": 1.474056603773585,
      "grad_norm": 1.0,
      "learning_rate": 0.00047870496026077984,
      "loss": 4.9906,
      "mean_token_accuracy": 0.20281320214271545,
      "num_tokens": 33727960.0,
      "step": 16250
    },
    {
      "entropy": 5.448631668090821,
      "epoch": 1.4745101596516692,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004786912178478476,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20749327540397644,
      "num_tokens": 33738874.0,
      "step": 16255
    },
    {
      "entropy": 5.473556089401245,
      "epoch": 1.4749637155297532,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004786774712225291,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20806999057531356,
      "num_tokens": 33749071.0,
      "step": 16260
    },
    {
      "entropy": 5.497419309616089,
      "epoch": 1.4754172714078375,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004786637203851086,
      "loss": 5.0007,
      "mean_token_accuracy": 0.20123066157102584,
      "num_tokens": 33759509.0,
      "step": 16265
    },
    {
      "entropy": 5.53020339012146,
      "epoch": 1.4758708272859216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004786499653358705,
      "loss": 5.0372,
      "mean_token_accuracy": 0.20973011404275893,
      "num_tokens": 33769974.0,
      "step": 16270
    },
    {
      "entropy": 5.570294809341431,
      "epoch": 1.4763243831640058,
      "grad_norm": 1.0,
      "learning_rate": 0.0004786362060750991,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20195408910512924,
      "num_tokens": 33780403.0,
      "step": 16275
    },
    {
      "entropy": 5.508374261856079,
      "epoch": 1.47677793904209,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047862244260307913,
      "loss": 5.0793,
      "mean_token_accuracy": 0.20363739430904387,
      "num_tokens": 33791364.0,
      "step": 16280
    },
    {
      "entropy": 5.478105258941651,
      "epoch": 1.4772314949201741,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004786086749200951,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20485219210386277,
      "num_tokens": 33801964.0,
      "step": 16285
    },
    {
      "entropy": 5.54928183555603,
      "epoch": 1.4776850507982584,
      "grad_norm": 1.0,
      "learning_rate": 0.0004785949030264319,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20498123168945312,
      "num_tokens": 33813132.0,
      "step": 16290
    },
    {
      "entropy": 5.491539859771729,
      "epoch": 1.4781386066763424,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004785811269223741,
      "loss": 5.0086,
      "mean_token_accuracy": 0.2109767660498619,
      "num_tokens": 33823069.0,
      "step": 16295
    },
    {
      "entropy": 5.547567510604859,
      "epoch": 1.4785921625544267,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004785673466082068,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2017671912908554,
      "num_tokens": 33832626.0,
      "step": 16300
    },
    {
      "entropy": 5.47188663482666,
      "epoch": 1.479045718432511,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047855356208421484,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20665343701839448,
      "num_tokens": 33843384.0,
      "step": 16305
    },
    {
      "entropy": 5.470773935317993,
      "epoch": 1.479499274310595,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004785397733506834,
      "loss": 4.9992,
      "mean_token_accuracy": 0.20403469055891038,
      "num_tokens": 33854286.0,
      "step": 16310
    },
    {
      "entropy": 5.445315361022949,
      "epoch": 1.4799528301886793,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004785259804078976,
      "loss": 4.965,
      "mean_token_accuracy": 0.2035436064004898,
      "num_tokens": 33863379.0,
      "step": 16315
    },
    {
      "entropy": 5.542968368530273,
      "epoch": 1.4804063860667633,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004785121832561426,
      "loss": 5.2038,
      "mean_token_accuracy": 0.1929345428943634,
      "num_tokens": 33875641.0,
      "step": 16320
    },
    {
      "entropy": 5.525194787979126,
      "epoch": 1.4808599419448476,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047849838189570375,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20928843170404435,
      "num_tokens": 33885023.0,
      "step": 16325
    },
    {
      "entropy": 5.618498134613037,
      "epoch": 1.4813134978229319,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004784845763268666,
      "loss": 5.1197,
      "mean_token_accuracy": 0.19594130516052247,
      "num_tokens": 33895562.0,
      "step": 16330
    },
    {
      "entropy": 5.488453531265259,
      "epoch": 1.481767053701016,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004784707665499165,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20698830634355544,
      "num_tokens": 33907095.0,
      "step": 16335
    },
    {
      "entropy": 5.407530736923218,
      "epoch": 1.4822206095791002,
      "grad_norm": 1.0,
      "learning_rate": 0.0004784569525651391,
      "loss": 4.8663,
      "mean_token_accuracy": 0.21668605953454972,
      "num_tokens": 33917475.0,
      "step": 16340
    },
    {
      "entropy": 5.445207118988037,
      "epoch": 1.4826741654571842,
      "grad_norm": 1.0,
      "learning_rate": 0.00047844313437282005,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20546100437641143,
      "num_tokens": 33927740.0,
      "step": 16345
    },
    {
      "entropy": 5.458725690841675,
      "epoch": 1.4831277213352685,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047842931197324525,
      "loss": 4.9693,
      "mean_token_accuracy": 0.20588107407093048,
      "num_tokens": 33937416.0,
      "step": 16350
    },
    {
      "entropy": 5.481399345397949,
      "epoch": 1.4835812772133528,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047841548536670033,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21526847928762435,
      "num_tokens": 33946968.0,
      "step": 16355
    },
    {
      "entropy": 5.577416038513183,
      "epoch": 1.4840348330914368,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047840165455347136,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19545600265264512,
      "num_tokens": 33957877.0,
      "step": 16360
    },
    {
      "entropy": 5.556639623641968,
      "epoch": 1.484488388969521,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047838781953384436,
      "loss": 5.0587,
      "mean_token_accuracy": 0.2009688213467598,
      "num_tokens": 33967890.0,
      "step": 16365
    },
    {
      "entropy": 5.463122367858887,
      "epoch": 1.484941944847605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004783739803081054,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20946183651685715,
      "num_tokens": 33977117.0,
      "step": 16370
    },
    {
      "entropy": 5.498557996749878,
      "epoch": 1.4853955007256894,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047836013687654075,
      "loss": 5.1039,
      "mean_token_accuracy": 0.19548790007829667,
      "num_tokens": 33986943.0,
      "step": 16375
    },
    {
      "entropy": 5.518389129638672,
      "epoch": 1.4858490566037736,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047834628923943654,
      "loss": 5.0073,
      "mean_token_accuracy": 0.19906968772411346,
      "num_tokens": 33996499.0,
      "step": 16380
    },
    {
      "entropy": 5.553841161727905,
      "epoch": 1.4863026124818577,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004783324373970793,
      "loss": 5.1192,
      "mean_token_accuracy": 0.20226586610078812,
      "num_tokens": 34006749.0,
      "step": 16385
    },
    {
      "entropy": 5.565635728836059,
      "epoch": 1.486756168359942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004783185813497555,
      "loss": 5.1152,
      "mean_token_accuracy": 0.19189804196357726,
      "num_tokens": 34017194.0,
      "step": 16390
    },
    {
      "entropy": 5.573387622833252,
      "epoch": 1.487209724238026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004783047210977516,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20614250153303146,
      "num_tokens": 34027386.0,
      "step": 16395
    },
    {
      "entropy": 5.6038854122161865,
      "epoch": 1.4876632801161103,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047829085664135424,
      "loss": 5.091,
      "mean_token_accuracy": 0.19450503885746,
      "num_tokens": 34037540.0,
      "step": 16400
    },
    {
      "entropy": 5.462180089950562,
      "epoch": 1.4881168359941945,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047827698798085015,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20876026153564453,
      "num_tokens": 34048008.0,
      "step": 16405
    },
    {
      "entropy": 5.512149477005005,
      "epoch": 1.4885703918722788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004782631151165262,
      "loss": 5.0296,
      "mean_token_accuracy": 0.2054557114839554,
      "num_tokens": 34058009.0,
      "step": 16410
    },
    {
      "entropy": 5.547857332229614,
      "epoch": 1.4890239477503628,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047824923804866914,
      "loss": 4.9857,
      "mean_token_accuracy": 0.20190584361553193,
      "num_tokens": 34067625.0,
      "step": 16415
    },
    {
      "entropy": 5.48768720626831,
      "epoch": 1.4894775036284469,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047823535677756607,
      "loss": 5.0204,
      "mean_token_accuracy": 0.19690577685832977,
      "num_tokens": 34077349.0,
      "step": 16420
    },
    {
      "entropy": 5.555475282669067,
      "epoch": 1.4899310595065312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004782214713035041,
      "loss": 5.139,
      "mean_token_accuracy": 0.19575213491916657,
      "num_tokens": 34088636.0,
      "step": 16425
    },
    {
      "entropy": 5.538937330245972,
      "epoch": 1.4903846153846154,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047820758162677026,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20493769496679307,
      "num_tokens": 34098806.0,
      "step": 16430
    },
    {
      "entropy": 5.502752494812012,
      "epoch": 1.4908381712626997,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047819368774765187,
      "loss": 5.0268,
      "mean_token_accuracy": 0.2061465784907341,
      "num_tokens": 34109713.0,
      "step": 16435
    },
    {
      "entropy": 5.475600719451904,
      "epoch": 1.4912917271407837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047817978966643626,
      "loss": 4.9947,
      "mean_token_accuracy": 0.205531345307827,
      "num_tokens": 34119585.0,
      "step": 16440
    },
    {
      "entropy": 5.568090486526489,
      "epoch": 1.491745283018868,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004781658873834108,
      "loss": 5.0499,
      "mean_token_accuracy": 0.1953028440475464,
      "num_tokens": 34131245.0,
      "step": 16445
    },
    {
      "entropy": 5.523230075836182,
      "epoch": 1.492198838896952,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047815198089886296,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2036787435412407,
      "num_tokens": 34141229.0,
      "step": 16450
    },
    {
      "entropy": 5.5111089706420895,
      "epoch": 1.4926523947750363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004781380702130805,
      "loss": 5.0672,
      "mean_token_accuracy": 0.19387389421463014,
      "num_tokens": 34151781.0,
      "step": 16455
    },
    {
      "entropy": 5.496467018127442,
      "epoch": 1.4931059506531206,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047812415532635087,
      "loss": 5.0182,
      "mean_token_accuracy": 0.20649980157613754,
      "num_tokens": 34162305.0,
      "step": 16460
    },
    {
      "entropy": 5.5562403202056885,
      "epoch": 1.4935595065312046,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000478110236238962,
      "loss": 5.0943,
      "mean_token_accuracy": 0.19907167553901672,
      "num_tokens": 34173613.0,
      "step": 16465
    },
    {
      "entropy": 5.488120412826538,
      "epoch": 1.4940130624092889,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047809631295120164,
      "loss": 5.021,
      "mean_token_accuracy": 0.20047862976789474,
      "num_tokens": 34183540.0,
      "step": 16470
    },
    {
      "entropy": 5.479513454437256,
      "epoch": 1.494466618287373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047808238546335786,
      "loss": 5.0361,
      "mean_token_accuracy": 0.20338803231716157,
      "num_tokens": 34193292.0,
      "step": 16475
    },
    {
      "entropy": 5.552792882919311,
      "epoch": 1.4949201741654572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047806845377571856,
      "loss": 5.0713,
      "mean_token_accuracy": 0.21045733243227005,
      "num_tokens": 34202697.0,
      "step": 16480
    },
    {
      "entropy": 5.5124249935150145,
      "epoch": 1.4953737300435415,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047805451788857187,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20582942068576812,
      "num_tokens": 34213396.0,
      "step": 16485
    },
    {
      "entropy": 5.520803928375244,
      "epoch": 1.4958272859216255,
      "grad_norm": 1.15625,
      "learning_rate": 0.000478040577802206,
      "loss": 4.9824,
      "mean_token_accuracy": 0.20087953060865402,
      "num_tokens": 34223020.0,
      "step": 16490
    },
    {
      "entropy": 5.445318698883057,
      "epoch": 1.4962808417997098,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004780266335169093,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20431505143642426,
      "num_tokens": 34233249.0,
      "step": 16495
    },
    {
      "entropy": 5.581749963760376,
      "epoch": 1.4967343976777938,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047801268503297,
      "loss": 5.0777,
      "mean_token_accuracy": 0.19654079228639604,
      "num_tokens": 34244091.0,
      "step": 16500
    },
    {
      "entropy": 5.560769701004029,
      "epoch": 1.497187953555878,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047799873235067667,
      "loss": 5.072,
      "mean_token_accuracy": 0.204006327688694,
      "num_tokens": 34255060.0,
      "step": 16505
    },
    {
      "entropy": 5.464819622039795,
      "epoch": 1.4976415094339623,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004779847754703178,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2080100730061531,
      "num_tokens": 34264361.0,
      "step": 16510
    },
    {
      "entropy": 5.492960977554321,
      "epoch": 1.4980950653120464,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004779708143921821,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20396582782268524,
      "num_tokens": 34274480.0,
      "step": 16515
    },
    {
      "entropy": 5.490897560119629,
      "epoch": 1.4985486211901307,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047795684911655824,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20501819401979446,
      "num_tokens": 34285951.0,
      "step": 16520
    },
    {
      "entropy": 5.553942012786865,
      "epoch": 1.4990021770682147,
      "grad_norm": 1.015625,
      "learning_rate": 0.000477942879643735,
      "loss": 5.0925,
      "mean_token_accuracy": 0.19646671563386917,
      "num_tokens": 34296384.0,
      "step": 16525
    },
    {
      "entropy": 5.501747560501099,
      "epoch": 1.499455732946299,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047792890597400125,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20746974349021913,
      "num_tokens": 34306776.0,
      "step": 16530
    },
    {
      "entropy": 5.557483053207397,
      "epoch": 1.4999092888243832,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047791492810764604,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20399336218833924,
      "num_tokens": 34317047.0,
      "step": 16535
    },
    {
      "entropy": 5.4808516025543215,
      "epoch": 1.5003628447024675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004779009460449584,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21083079278469086,
      "num_tokens": 34326392.0,
      "step": 16540
    },
    {
      "entropy": 5.536378717422485,
      "epoch": 1.5008164005805515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047788695978622756,
      "loss": 5.0283,
      "mean_token_accuracy": 0.206308513879776,
      "num_tokens": 34336580.0,
      "step": 16545
    },
    {
      "entropy": 5.4815671920776365,
      "epoch": 1.5012699564586356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047787296933174265,
      "loss": 5.0454,
      "mean_token_accuracy": 0.2009593978524208,
      "num_tokens": 34346792.0,
      "step": 16550
    },
    {
      "entropy": 5.531548976898193,
      "epoch": 1.5017235123367199,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000477858974681793,
      "loss": 5.073,
      "mean_token_accuracy": 0.19875523447990417,
      "num_tokens": 34356675.0,
      "step": 16555
    },
    {
      "entropy": 5.547983121871948,
      "epoch": 1.5021770682148041,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004778449758366681,
      "loss": 5.0216,
      "mean_token_accuracy": 0.2152870386838913,
      "num_tokens": 34366357.0,
      "step": 16560
    },
    {
      "entropy": 5.482605981826782,
      "epoch": 1.5026306240928884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004778309727966574,
      "loss": 5.0302,
      "mean_token_accuracy": 0.2065482184290886,
      "num_tokens": 34377229.0,
      "step": 16565
    },
    {
      "entropy": 5.479886150360107,
      "epoch": 1.5030841799709724,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047781696556205045,
      "loss": 4.9632,
      "mean_token_accuracy": 0.2029298186302185,
      "num_tokens": 34387873.0,
      "step": 16570
    },
    {
      "entropy": 5.515002155303955,
      "epoch": 1.5035377358490565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047780295413313696,
      "loss": 4.9855,
      "mean_token_accuracy": 0.20847856849431992,
      "num_tokens": 34398889.0,
      "step": 16575
    },
    {
      "entropy": 5.46763858795166,
      "epoch": 1.5039912917271407,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047778893851020676,
      "loss": 4.9349,
      "mean_token_accuracy": 0.207319737970829,
      "num_tokens": 34408508.0,
      "step": 16580
    },
    {
      "entropy": 5.492764616012574,
      "epoch": 1.504444847605225,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004777749186935496,
      "loss": 5.0973,
      "mean_token_accuracy": 0.19956476986408234,
      "num_tokens": 34418928.0,
      "step": 16585
    },
    {
      "entropy": 5.5559296131134035,
      "epoch": 1.5048984034833093,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004777608946834555,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20337532162666322,
      "num_tokens": 34429789.0,
      "step": 16590
    },
    {
      "entropy": 5.575822639465332,
      "epoch": 1.5053519593613933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004777468664802144,
      "loss": 5.1585,
      "mean_token_accuracy": 0.19455186128616334,
      "num_tokens": 34440795.0,
      "step": 16595
    },
    {
      "entropy": 5.568664598464966,
      "epoch": 1.5058055152394774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047773283408411636,
      "loss": 5.0648,
      "mean_token_accuracy": 0.19236605167388915,
      "num_tokens": 34450772.0,
      "step": 16600
    },
    {
      "entropy": 5.51489109992981,
      "epoch": 1.5062590711175616,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047771879749545175,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20160395056009292,
      "num_tokens": 34461273.0,
      "step": 16605
    },
    {
      "entropy": 5.520502233505249,
      "epoch": 1.506712626995646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047770475671451073,
      "loss": 5.0666,
      "mean_token_accuracy": 0.20057346075773239,
      "num_tokens": 34470942.0,
      "step": 16610
    },
    {
      "entropy": 5.560528707504273,
      "epoch": 1.5071661828737302,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004776907117415837,
      "loss": 4.994,
      "mean_token_accuracy": 0.2034139022231102,
      "num_tokens": 34481117.0,
      "step": 16615
    },
    {
      "entropy": 5.469490098953247,
      "epoch": 1.5076197387518142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047767666257696103,
      "loss": 4.9706,
      "mean_token_accuracy": 0.20528270900249482,
      "num_tokens": 34490953.0,
      "step": 16620
    },
    {
      "entropy": 5.440951585769653,
      "epoch": 1.5080732946298983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047766260922093337,
      "loss": 4.925,
      "mean_token_accuracy": 0.21110389679670333,
      "num_tokens": 34502765.0,
      "step": 16625
    },
    {
      "entropy": 5.6015337944030765,
      "epoch": 1.5085268505079825,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004776485516737913,
      "loss": 5.1148,
      "mean_token_accuracy": 0.19609806537628174,
      "num_tokens": 34514473.0,
      "step": 16630
    },
    {
      "entropy": 5.567944288253784,
      "epoch": 1.5089804063860668,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004776344899358256,
      "loss": 5.1329,
      "mean_token_accuracy": 0.1925711914896965,
      "num_tokens": 34524218.0,
      "step": 16635
    },
    {
      "entropy": 5.536440992355347,
      "epoch": 1.509433962264151,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004776204240073269,
      "loss": 5.0156,
      "mean_token_accuracy": 0.20336877703666686,
      "num_tokens": 34535697.0,
      "step": 16640
    },
    {
      "entropy": 5.533255910873413,
      "epoch": 1.509887518142235,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047760635388858624,
      "loss": 5.0585,
      "mean_token_accuracy": 0.2072887033224106,
      "num_tokens": 34545130.0,
      "step": 16645
    },
    {
      "entropy": 5.475963020324707,
      "epoch": 1.5103410740203191,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047759227957989465,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20716975182294844,
      "num_tokens": 34554974.0,
      "step": 16650
    },
    {
      "entropy": 5.521869230270386,
      "epoch": 1.5107946298984034,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477578201081543,
      "loss": 5.024,
      "mean_token_accuracy": 0.20081094801425933,
      "num_tokens": 34565660.0,
      "step": 16655
    },
    {
      "entropy": 5.5908403396606445,
      "epoch": 1.5112481857764877,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004775641183938225,
      "loss": 5.0622,
      "mean_token_accuracy": 0.1991403043270111,
      "num_tokens": 34576447.0,
      "step": 16660
    },
    {
      "entropy": 5.5940087795257565,
      "epoch": 1.511701741654572,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047755003151702443,
      "loss": 5.1877,
      "mean_token_accuracy": 0.19777849465608596,
      "num_tokens": 34588333.0,
      "step": 16665
    },
    {
      "entropy": 5.453550243377686,
      "epoch": 1.512155297532656,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047753594045144015,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20184141397476196,
      "num_tokens": 34598166.0,
      "step": 16670
    },
    {
      "entropy": 5.485975837707519,
      "epoch": 1.5126088534107403,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004775218451973609,
      "loss": 4.968,
      "mean_token_accuracy": 0.20365387797355652,
      "num_tokens": 34608615.0,
      "step": 16675
    },
    {
      "entropy": 5.497578477859497,
      "epoch": 1.5130624092888243,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004775077457550784,
      "loss": 5.0384,
      "mean_token_accuracy": 0.20195758193731309,
      "num_tokens": 34618618.0,
      "step": 16680
    },
    {
      "entropy": 5.5292384147644045,
      "epoch": 1.5135159651669086,
      "grad_norm": 1.0,
      "learning_rate": 0.00047749364212488403,
      "loss": 5.0519,
      "mean_token_accuracy": 0.19552540332078933,
      "num_tokens": 34629412.0,
      "step": 16685
    },
    {
      "entropy": 5.576356315612793,
      "epoch": 1.5139695210449928,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004774795343070696,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20159600377082826,
      "num_tokens": 34640487.0,
      "step": 16690
    },
    {
      "entropy": 5.537057590484619,
      "epoch": 1.5144230769230769,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004774654223019268,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20246966928243637,
      "num_tokens": 34650126.0,
      "step": 16695
    },
    {
      "entropy": 5.455569219589234,
      "epoch": 1.5148766328011611,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004774513061097474,
      "loss": 4.9078,
      "mean_token_accuracy": 0.20852980017662048,
      "num_tokens": 34659721.0,
      "step": 16700
    },
    {
      "entropy": 5.494598388671875,
      "epoch": 1.5153301886792452,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004774371857308235,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20723765790462495,
      "num_tokens": 34670611.0,
      "step": 16705
    },
    {
      "entropy": 5.525382232666016,
      "epoch": 1.5157837445573294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047742306116544693,
      "loss": 5.007,
      "mean_token_accuracy": 0.2060346469283104,
      "num_tokens": 34680173.0,
      "step": 16710
    },
    {
      "entropy": 5.528834962844849,
      "epoch": 1.5162373004354137,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047740893241390994,
      "loss": 5.0667,
      "mean_token_accuracy": 0.2004269316792488,
      "num_tokens": 34690176.0,
      "step": 16715
    },
    {
      "entropy": 5.544941663742065,
      "epoch": 1.516690856313498,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004773947994765046,
      "loss": 4.9883,
      "mean_token_accuracy": 0.205637526512146,
      "num_tokens": 34700481.0,
      "step": 16720
    },
    {
      "entropy": 5.481034755706787,
      "epoch": 1.517144412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004773806623535232,
      "loss": 5.1565,
      "mean_token_accuracy": 0.19801995754241944,
      "num_tokens": 34710643.0,
      "step": 16725
    },
    {
      "entropy": 5.548236465454101,
      "epoch": 1.517597968069666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047736652104525816,
      "loss": 5.0252,
      "mean_token_accuracy": 0.20002883970737456,
      "num_tokens": 34720998.0,
      "step": 16730
    },
    {
      "entropy": 5.598523807525635,
      "epoch": 1.5180515239477503,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004773523755520019,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20067711025476456,
      "num_tokens": 34731071.0,
      "step": 16735
    },
    {
      "entropy": 5.519377326965332,
      "epoch": 1.5185050798258346,
      "grad_norm": 1.046875,
      "learning_rate": 0.000477338225874047,
      "loss": 4.9939,
      "mean_token_accuracy": 0.20087372213602067,
      "num_tokens": 34741346.0,
      "step": 16740
    },
    {
      "entropy": 5.559559297561646,
      "epoch": 1.5189586357039189,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047732407201168595,
      "loss": 4.9857,
      "mean_token_accuracy": 0.20293824523687362,
      "num_tokens": 34751092.0,
      "step": 16745
    },
    {
      "entropy": 5.604568767547607,
      "epoch": 1.519412191582003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004773099139652116,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20174917578697205,
      "num_tokens": 34761477.0,
      "step": 16750
    },
    {
      "entropy": 5.471523380279541,
      "epoch": 1.519865747460087,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004772957517349167,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2046545147895813,
      "num_tokens": 34772491.0,
      "step": 16755
    },
    {
      "entropy": 5.5182709217071535,
      "epoch": 1.5203193033381712,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004772815853210939,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19876878559589387,
      "num_tokens": 34783192.0,
      "step": 16760
    },
    {
      "entropy": 5.534299516677857,
      "epoch": 1.5207728592162555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047726741472403655,
      "loss": 5.0353,
      "mean_token_accuracy": 0.2047033876180649,
      "num_tokens": 34794164.0,
      "step": 16765
    },
    {
      "entropy": 5.557278108596802,
      "epoch": 1.5212264150943398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047725323994403753,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19490733742713928,
      "num_tokens": 34804390.0,
      "step": 16770
    },
    {
      "entropy": 5.639868545532226,
      "epoch": 1.5216799709724238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047723906098138993,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19346656948328017,
      "num_tokens": 34814751.0,
      "step": 16775
    },
    {
      "entropy": 5.527458047866821,
      "epoch": 1.5221335268505078,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047722487783638697,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20670805424451827,
      "num_tokens": 34824977.0,
      "step": 16780
    },
    {
      "entropy": 5.530150556564331,
      "epoch": 1.5225870827285921,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000477210690509322,
      "loss": 5.0629,
      "mean_token_accuracy": 0.20371615886688232,
      "num_tokens": 34835540.0,
      "step": 16785
    },
    {
      "entropy": 5.575498342514038,
      "epoch": 1.5230406386066764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047719649900048843,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20107826590538025,
      "num_tokens": 34845233.0,
      "step": 16790
    },
    {
      "entropy": 5.594610977172851,
      "epoch": 1.5234941944847606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047718230331017975,
      "loss": 5.1333,
      "mean_token_accuracy": 0.19620306491851808,
      "num_tokens": 34855756.0,
      "step": 16795
    },
    {
      "entropy": 5.552404689788818,
      "epoch": 1.5239477503628447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047716810343868956,
      "loss": 5.0549,
      "mean_token_accuracy": 0.19141069501638414,
      "num_tokens": 34867869.0,
      "step": 16800
    },
    {
      "entropy": 5.617313003540039,
      "epoch": 1.5244013062409287,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047715389938631143,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19923090040683747,
      "num_tokens": 34878625.0,
      "step": 16805
    },
    {
      "entropy": 5.596384859085083,
      "epoch": 1.524854862119013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047713969115333916,
      "loss": 5.101,
      "mean_token_accuracy": 0.1899292528629303,
      "num_tokens": 34889408.0,
      "step": 16810
    },
    {
      "entropy": 5.51461124420166,
      "epoch": 1.5253084179970973,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004771254787400666,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19528281986713408,
      "num_tokens": 34899602.0,
      "step": 16815
    },
    {
      "entropy": 5.465745878219605,
      "epoch": 1.5257619738751815,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004771112621467876,
      "loss": 4.9482,
      "mean_token_accuracy": 0.20518770664930344,
      "num_tokens": 34910215.0,
      "step": 16820
    },
    {
      "entropy": 5.495348262786865,
      "epoch": 1.5262155297532656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004770970413737961,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2078547567129135,
      "num_tokens": 34919641.0,
      "step": 16825
    },
    {
      "entropy": 5.527891826629639,
      "epoch": 1.5266690856313496,
      "grad_norm": 1.0,
      "learning_rate": 0.0004770828164213865,
      "loss": 4.9251,
      "mean_token_accuracy": 0.2131982073187828,
      "num_tokens": 34930243.0,
      "step": 16830
    },
    {
      "entropy": 5.534064245223999,
      "epoch": 1.5271226415094339,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004770685872898525,
      "loss": 5.1001,
      "mean_token_accuracy": 0.20086862593889238,
      "num_tokens": 34941057.0,
      "step": 16835
    },
    {
      "entropy": 5.490671062469483,
      "epoch": 1.5275761973875182,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047705435397948876,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21125524044036864,
      "num_tokens": 34951150.0,
      "step": 16840
    },
    {
      "entropy": 5.574343013763428,
      "epoch": 1.5280297532656024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004770401164905895,
      "loss": 5.0603,
      "mean_token_accuracy": 0.203806009888649,
      "num_tokens": 34961648.0,
      "step": 16845
    },
    {
      "entropy": 5.55496039390564,
      "epoch": 1.5284833091436865,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047702587482344915,
      "loss": 5.0569,
      "mean_token_accuracy": 0.1965765506029129,
      "num_tokens": 34971573.0,
      "step": 16850
    },
    {
      "entropy": 5.556545734405518,
      "epoch": 1.5289368650217707,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004770116289783622,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1982572853565216,
      "num_tokens": 34981984.0,
      "step": 16855
    },
    {
      "entropy": 5.547101306915283,
      "epoch": 1.5293904208998548,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004769973789556233,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1957728385925293,
      "num_tokens": 34990861.0,
      "step": 16860
    },
    {
      "entropy": 5.530113744735718,
      "epoch": 1.529843976777939,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047698312475552714,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20197166949510575,
      "num_tokens": 35001504.0,
      "step": 16865
    },
    {
      "entropy": 5.549115228652954,
      "epoch": 1.5302975326560233,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004769688663783684,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2118733510375023,
      "num_tokens": 35012425.0,
      "step": 16870
    },
    {
      "entropy": 5.511075735092163,
      "epoch": 1.5307510885341074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004769546038244421,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21017370223999024,
      "num_tokens": 35022854.0,
      "step": 16875
    },
    {
      "entropy": 5.544015026092529,
      "epoch": 1.5312046444121916,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047694033709404313,
      "loss": 5.0909,
      "mean_token_accuracy": 0.196853344142437,
      "num_tokens": 35032694.0,
      "step": 16880
    },
    {
      "entropy": 5.522608804702759,
      "epoch": 1.5316582002902757,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047692606618746655,
      "loss": 4.9996,
      "mean_token_accuracy": 0.20808610320091248,
      "num_tokens": 35042217.0,
      "step": 16885
    },
    {
      "entropy": 5.5298590660095215,
      "epoch": 1.53211175616836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047691179110500735,
      "loss": 5.0889,
      "mean_token_accuracy": 0.1910834416747093,
      "num_tokens": 35052831.0,
      "step": 16890
    },
    {
      "entropy": 5.586901235580444,
      "epoch": 1.5325653120464442,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047689751184696097,
      "loss": 5.1123,
      "mean_token_accuracy": 0.1989368662238121,
      "num_tokens": 35061955.0,
      "step": 16895
    },
    {
      "entropy": 5.543337059020996,
      "epoch": 1.5330188679245285,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004768832284136226,
      "loss": 5.0147,
      "mean_token_accuracy": 0.2031451642513275,
      "num_tokens": 35072526.0,
      "step": 16900
    },
    {
      "entropy": 5.501619005203247,
      "epoch": 1.5334724238026125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004768689408052876,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20547995269298552,
      "num_tokens": 35082415.0,
      "step": 16905
    },
    {
      "entropy": 5.475245141983033,
      "epoch": 1.5339259796806965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047685464902225144,
      "loss": 4.982,
      "mean_token_accuracy": 0.20835471153259277,
      "num_tokens": 35092805.0,
      "step": 16910
    },
    {
      "entropy": 5.542777490615845,
      "epoch": 1.5343795355587808,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004768403530648097,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20715077519416808,
      "num_tokens": 35102963.0,
      "step": 16915
    },
    {
      "entropy": 5.54917802810669,
      "epoch": 1.534833091436865,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047682605293325806,
      "loss": 5.0578,
      "mean_token_accuracy": 0.2008623957633972,
      "num_tokens": 35114140.0,
      "step": 16920
    },
    {
      "entropy": 5.588230037689209,
      "epoch": 1.5352866473149493,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047681174862789215,
      "loss": 5.1263,
      "mean_token_accuracy": 0.20138452351093292,
      "num_tokens": 35125314.0,
      "step": 16925
    },
    {
      "entropy": 5.523447322845459,
      "epoch": 1.5357402031930334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047679744014900795,
      "loss": 5.0299,
      "mean_token_accuracy": 0.2043992131948471,
      "num_tokens": 35134934.0,
      "step": 16930
    },
    {
      "entropy": 5.487507247924805,
      "epoch": 1.5361937590711174,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004767831274969011,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20349218994379042,
      "num_tokens": 35145341.0,
      "step": 16935
    },
    {
      "entropy": 5.455434989929199,
      "epoch": 1.5366473149492017,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047676881067186793,
      "loss": 4.9612,
      "mean_token_accuracy": 0.2022953748703003,
      "num_tokens": 35156046.0,
      "step": 16940
    },
    {
      "entropy": 5.558455896377564,
      "epoch": 1.537100870827286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047675448967420426,
      "loss": 5.1179,
      "mean_token_accuracy": 0.19479937106370926,
      "num_tokens": 35168232.0,
      "step": 16945
    },
    {
      "entropy": 5.607294321060181,
      "epoch": 1.5375544267053702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004767401645042064,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20226863920688629,
      "num_tokens": 35178585.0,
      "step": 16950
    },
    {
      "entropy": 5.532871150970459,
      "epoch": 1.5380079825834543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047672583516217043,
      "loss": 4.9313,
      "mean_token_accuracy": 0.20752260088920593,
      "num_tokens": 35189417.0,
      "step": 16955
    },
    {
      "entropy": 5.495153856277466,
      "epoch": 1.5384615384615383,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004767115016483928,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2012806624174118,
      "num_tokens": 35200167.0,
      "step": 16960
    },
    {
      "entropy": 5.490826416015625,
      "epoch": 1.5389150943396226,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004766971639631699,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20241640508174896,
      "num_tokens": 35210711.0,
      "step": 16965
    },
    {
      "entropy": 5.481350374221802,
      "epoch": 1.5393686502177069,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047668282210679825,
      "loss": 5.0353,
      "mean_token_accuracy": 0.1973751053214073,
      "num_tokens": 35220866.0,
      "step": 16970
    },
    {
      "entropy": 5.4429432392120365,
      "epoch": 1.5398222060957911,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004766684760795745,
      "loss": 4.992,
      "mean_token_accuracy": 0.20854220241308213,
      "num_tokens": 35231478.0,
      "step": 16975
    },
    {
      "entropy": 5.476144599914551,
      "epoch": 1.5402757619738752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047665412588179524,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21443094462156295,
      "num_tokens": 35242233.0,
      "step": 16980
    },
    {
      "entropy": 5.492010641098022,
      "epoch": 1.5407293178519592,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004766397715137573,
      "loss": 4.9949,
      "mean_token_accuracy": 0.20129960477352143,
      "num_tokens": 35253097.0,
      "step": 16985
    },
    {
      "entropy": 5.592983293533325,
      "epoch": 1.5411828737300435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047662541297575743,
      "loss": 5.1592,
      "mean_token_accuracy": 0.19676952362060546,
      "num_tokens": 35264258.0,
      "step": 16990
    },
    {
      "entropy": 5.667860507965088,
      "epoch": 1.5416364296081277,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047661105026809274,
      "loss": 5.1293,
      "mean_token_accuracy": 0.19510892480611802,
      "num_tokens": 35273501.0,
      "step": 16995
    },
    {
      "entropy": 5.5120124340057375,
      "epoch": 1.542089985486212,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004765966833910601,
      "loss": 5.0757,
      "mean_token_accuracy": 0.19924862831830978,
      "num_tokens": 35284125.0,
      "step": 17000
    },
    {
      "entropy": 5.618932294845581,
      "epoch": 1.542543541364296,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004765823123449567,
      "loss": 5.073,
      "mean_token_accuracy": 0.1988307297229767,
      "num_tokens": 35296039.0,
      "step": 17005
    },
    {
      "entropy": 5.568522262573242,
      "epoch": 1.54299709724238,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047656793713007965,
      "loss": 5.0089,
      "mean_token_accuracy": 0.19973995089530944,
      "num_tokens": 35306178.0,
      "step": 17010
    },
    {
      "entropy": 5.4798565864562985,
      "epoch": 1.5434506531204644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047655355774672636,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20508571118116378,
      "num_tokens": 35316445.0,
      "step": 17015
    },
    {
      "entropy": 5.537961816787719,
      "epoch": 1.5439042089985486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047653917419519413,
      "loss": 5.0381,
      "mean_token_accuracy": 0.1992551326751709,
      "num_tokens": 35327274.0,
      "step": 17020
    },
    {
      "entropy": 5.526545619964599,
      "epoch": 1.544357764876633,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004765247864757804,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20724283307790756,
      "num_tokens": 35339007.0,
      "step": 17025
    },
    {
      "entropy": 5.55948657989502,
      "epoch": 1.544811320754717,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004765103945887827,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19256114810705185,
      "num_tokens": 35349303.0,
      "step": 17030
    },
    {
      "entropy": 5.5475461959838865,
      "epoch": 1.5452648766328012,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004764959985344987,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2082339644432068,
      "num_tokens": 35359556.0,
      "step": 17035
    },
    {
      "entropy": 5.494559288024902,
      "epoch": 1.5457184325108853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004764815983132261,
      "loss": 4.9836,
      "mean_token_accuracy": 0.20900323241949081,
      "num_tokens": 35370043.0,
      "step": 17040
    },
    {
      "entropy": 5.523154592514038,
      "epoch": 1.5461719883889695,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004764671939252628,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21071674376726152,
      "num_tokens": 35380031.0,
      "step": 17045
    },
    {
      "entropy": 5.532931280136109,
      "epoch": 1.5466255442670538,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004764527853709064,
      "loss": 5.0761,
      "mean_token_accuracy": 0.1993743047118187,
      "num_tokens": 35390481.0,
      "step": 17050
    },
    {
      "entropy": 5.51744589805603,
      "epoch": 1.5470791001451378,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004764383726504552,
      "loss": 5.06,
      "mean_token_accuracy": 0.20070597678422927,
      "num_tokens": 35401756.0,
      "step": 17055
    },
    {
      "entropy": 5.545562696456909,
      "epoch": 1.547532656023222,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004764239557642071,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21284027099609376,
      "num_tokens": 35411045.0,
      "step": 17060
    },
    {
      "entropy": 5.501655769348145,
      "epoch": 1.5479862119013061,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004764095347124603,
      "loss": 5.0474,
      "mean_token_accuracy": 0.19385991096496583,
      "num_tokens": 35421256.0,
      "step": 17065
    },
    {
      "entropy": 5.5252477645874025,
      "epoch": 1.5484397677793904,
      "grad_norm": 1.0,
      "learning_rate": 0.0004763951094955129,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20608636289834975,
      "num_tokens": 35430798.0,
      "step": 17070
    },
    {
      "entropy": 5.514478445053101,
      "epoch": 1.5488933236574747,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004763806801136634,
      "loss": 4.9919,
      "mean_token_accuracy": 0.2099059194326401,
      "num_tokens": 35440459.0,
      "step": 17075
    },
    {
      "entropy": 5.513672113418579,
      "epoch": 1.549346879535559,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004763662465672101,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20351307094097137,
      "num_tokens": 35451586.0,
      "step": 17080
    },
    {
      "entropy": 5.521123647689819,
      "epoch": 1.549800435413643,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047635180885645146,
      "loss": 5.0499,
      "mean_token_accuracy": 0.2032095104455948,
      "num_tokens": 35462197.0,
      "step": 17085
    },
    {
      "entropy": 5.549882459640503,
      "epoch": 1.550253991291727,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004763373669816862,
      "loss": 4.9743,
      "mean_token_accuracy": 0.20916669368743895,
      "num_tokens": 35471737.0,
      "step": 17090
    },
    {
      "entropy": 5.562024736404419,
      "epoch": 1.5507075471698113,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004763229209432128,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1963101416826248,
      "num_tokens": 35482016.0,
      "step": 17095
    },
    {
      "entropy": 5.572429990768432,
      "epoch": 1.5511611030478956,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004763084707413301,
      "loss": 5.1605,
      "mean_token_accuracy": 0.18899236917495726,
      "num_tokens": 35492622.0,
      "step": 17100
    },
    {
      "entropy": 5.624724960327148,
      "epoch": 1.5516146589259798,
      "grad_norm": 1.125,
      "learning_rate": 0.00047629401637633696,
      "loss": 5.1854,
      "mean_token_accuracy": 0.19349545240402222,
      "num_tokens": 35504468.0,
      "step": 17105
    },
    {
      "entropy": 5.528221845626831,
      "epoch": 1.5520682148040639,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047627955784853226,
      "loss": 5.0141,
      "mean_token_accuracy": 0.199432609975338,
      "num_tokens": 35514272.0,
      "step": 17110
    },
    {
      "entropy": 5.519846725463867,
      "epoch": 1.552521770682148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000476265095158215,
      "loss": 5.0043,
      "mean_token_accuracy": 0.1996614083647728,
      "num_tokens": 35524110.0,
      "step": 17115
    },
    {
      "entropy": 5.548833322525025,
      "epoch": 1.5529753265602322,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004762506283056843,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21215675473213197,
      "num_tokens": 35535504.0,
      "step": 17120
    },
    {
      "entropy": 5.624251079559326,
      "epoch": 1.5534288824383164,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004762361572912393,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19390741139650344,
      "num_tokens": 35544885.0,
      "step": 17125
    },
    {
      "entropy": 5.543760204315186,
      "epoch": 1.5538824383164007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047622168211517926,
      "loss": 5.1441,
      "mean_token_accuracy": 0.20457107722759246,
      "num_tokens": 35555485.0,
      "step": 17130
    },
    {
      "entropy": 5.468977546691894,
      "epoch": 1.5543359941944848,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047620720277780353,
      "loss": 4.9531,
      "mean_token_accuracy": 0.20908924788236619,
      "num_tokens": 35567197.0,
      "step": 17135
    },
    {
      "entropy": 5.547029113769531,
      "epoch": 1.5547895500725688,
      "grad_norm": 1.0,
      "learning_rate": 0.00047619271927941165,
      "loss": 4.9914,
      "mean_token_accuracy": 0.20580670535564421,
      "num_tokens": 35577576.0,
      "step": 17140
    },
    {
      "entropy": 5.532487440109253,
      "epoch": 1.555243105950653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000476178231620303,
      "loss": 5.0558,
      "mean_token_accuracy": 0.2027304396033287,
      "num_tokens": 35588012.0,
      "step": 17145
    },
    {
      "entropy": 5.498031139373779,
      "epoch": 1.5556966618287373,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004761637398007774,
      "loss": 5.0329,
      "mean_token_accuracy": 0.19690879434347153,
      "num_tokens": 35598076.0,
      "step": 17150
    },
    {
      "entropy": 5.444629192352295,
      "epoch": 1.5561502177068216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004761492438211342,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2136305958032608,
      "num_tokens": 35607170.0,
      "step": 17155
    },
    {
      "entropy": 5.514594650268554,
      "epoch": 1.5566037735849056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047613474368167355,
      "loss": 4.9972,
      "mean_token_accuracy": 0.1959199920296669,
      "num_tokens": 35617865.0,
      "step": 17160
    },
    {
      "entropy": 5.528052568435669,
      "epoch": 1.5570573294629897,
      "grad_norm": 1.0625,
      "learning_rate": 0.000476120239382695,
      "loss": 5.0517,
      "mean_token_accuracy": 0.20201874971389772,
      "num_tokens": 35629329.0,
      "step": 17165
    },
    {
      "entropy": 5.517793083190918,
      "epoch": 1.557510885341074,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047610573092449877,
      "loss": 5.0536,
      "mean_token_accuracy": 0.19934562891721724,
      "num_tokens": 35640044.0,
      "step": 17170
    },
    {
      "entropy": 5.552383804321289,
      "epoch": 1.5579644412191582,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004760912183073847,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20094917714595795,
      "num_tokens": 35650435.0,
      "step": 17175
    },
    {
      "entropy": 5.532912158966065,
      "epoch": 1.5584179970972425,
      "grad_norm": 0.953125,
      "learning_rate": 0.000476076701531653,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20271799117326736,
      "num_tokens": 35659613.0,
      "step": 17180
    },
    {
      "entropy": 5.494220638275147,
      "epoch": 1.5588715529753265,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004760621805976039,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20519281178712845,
      "num_tokens": 35670987.0,
      "step": 17185
    },
    {
      "entropy": 5.498604202270508,
      "epoch": 1.5593251088534106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047604765550553766,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20504542142152787,
      "num_tokens": 35682505.0,
      "step": 17190
    },
    {
      "entropy": 5.552568626403809,
      "epoch": 1.5597786647314948,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047603312625575466,
      "loss": 4.9551,
      "mean_token_accuracy": 0.20232332348823548,
      "num_tokens": 35693056.0,
      "step": 17195
    },
    {
      "entropy": 5.518586921691894,
      "epoch": 1.5602322206095791,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047601859284855544,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21319104433059693,
      "num_tokens": 35703659.0,
      "step": 17200
    },
    {
      "entropy": 5.4766816139221195,
      "epoch": 1.5606857764876634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004760040552842403,
      "loss": 5.025,
      "mean_token_accuracy": 0.2015480637550354,
      "num_tokens": 35714367.0,
      "step": 17205
    },
    {
      "entropy": 5.4776341915130615,
      "epoch": 1.5611393323657474,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047598951356311035,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2010535180568695,
      "num_tokens": 35724773.0,
      "step": 17210
    },
    {
      "entropy": 5.569267797470093,
      "epoch": 1.5615928882438317,
      "grad_norm": 1.0,
      "learning_rate": 0.0004759749676854658,
      "loss": 5.1312,
      "mean_token_accuracy": 0.19799581170082092,
      "num_tokens": 35735622.0,
      "step": 17215
    },
    {
      "entropy": 5.543905401229859,
      "epoch": 1.5620464441219157,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004759604176516079,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2084849089384079,
      "num_tokens": 35744967.0,
      "step": 17220
    },
    {
      "entropy": 5.555732345581054,
      "epoch": 1.5625,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004759458634618372,
      "loss": 5.0294,
      "mean_token_accuracy": 0.20423816591501237,
      "num_tokens": 35755647.0,
      "step": 17225
    },
    {
      "entropy": 5.529198217391968,
      "epoch": 1.5629535558780843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004759313051164549,
      "loss": 5.063,
      "mean_token_accuracy": 0.1968506619334221,
      "num_tokens": 35765595.0,
      "step": 17230
    },
    {
      "entropy": 5.545705795288086,
      "epoch": 1.5634071117561683,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000475916742615762,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2100907638669014,
      "num_tokens": 35775804.0,
      "step": 17235
    },
    {
      "entropy": 5.504986000061035,
      "epoch": 1.5638606676342526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004759021759600597,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20709893405437468,
      "num_tokens": 35786227.0,
      "step": 17240
    },
    {
      "entropy": 5.567078065872193,
      "epoch": 1.5643142235123366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004758876051496492,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20364011228084564,
      "num_tokens": 35797511.0,
      "step": 17245
    },
    {
      "entropy": 5.560240745544434,
      "epoch": 1.5647677793904209,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047587303018483177,
      "loss": 5.051,
      "mean_token_accuracy": 0.19859770983457564,
      "num_tokens": 35807248.0,
      "step": 17250
    },
    {
      "entropy": 5.52660493850708,
      "epoch": 1.5652213352685052,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047585845106590897,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19810502827167512,
      "num_tokens": 35818164.0,
      "step": 17255
    },
    {
      "entropy": 5.515972757339478,
      "epoch": 1.5656748911465894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047584386779318214,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2109167367219925,
      "num_tokens": 35828746.0,
      "step": 17260
    },
    {
      "entropy": 5.546813488006592,
      "epoch": 1.5661284470246735,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000475829280366953,
      "loss": 5.0758,
      "mean_token_accuracy": 0.1961701601743698,
      "num_tokens": 35839325.0,
      "step": 17265
    },
    {
      "entropy": 5.536845350265503,
      "epoch": 1.5665820029027575,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047581468878752313,
      "loss": 5.001,
      "mean_token_accuracy": 0.20544881075620652,
      "num_tokens": 35849406.0,
      "step": 17270
    },
    {
      "entropy": 5.454688835144043,
      "epoch": 1.5670355587808418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047580009305519427,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2047882094979286,
      "num_tokens": 35858715.0,
      "step": 17275
    },
    {
      "entropy": 5.4782956600189205,
      "epoch": 1.567489114658926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004757854931702684,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2048191696405411,
      "num_tokens": 35869063.0,
      "step": 17280
    },
    {
      "entropy": 5.48824028968811,
      "epoch": 1.5679426705370103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047577088913304733,
      "loss": 4.9003,
      "mean_token_accuracy": 0.20872622430324556,
      "num_tokens": 35878428.0,
      "step": 17285
    },
    {
      "entropy": 5.558001232147217,
      "epoch": 1.5683962264150944,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047575628094383317,
      "loss": 5.035,
      "mean_token_accuracy": 0.2058235377073288,
      "num_tokens": 35888801.0,
      "step": 17290
    },
    {
      "entropy": 5.473654413223267,
      "epoch": 1.5688497822931784,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004757416686029279,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20353106558322906,
      "num_tokens": 35899381.0,
      "step": 17295
    },
    {
      "entropy": 5.556720638275147,
      "epoch": 1.5693033381712627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004757270521106338,
      "loss": 5.0738,
      "mean_token_accuracy": 0.20061299651861192,
      "num_tokens": 35909546.0,
      "step": 17300
    },
    {
      "entropy": 5.520714902877808,
      "epoch": 1.569756894049347,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004757124314672531,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2015165463089943,
      "num_tokens": 35919301.0,
      "step": 17305
    },
    {
      "entropy": 5.58653039932251,
      "epoch": 1.5702104499274312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047569780667308816,
      "loss": 5.0878,
      "mean_token_accuracy": 0.20005113929510115,
      "num_tokens": 35928603.0,
      "step": 17310
    },
    {
      "entropy": 5.58700156211853,
      "epoch": 1.5706640058055152,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047568317772844147,
      "loss": 5.067,
      "mean_token_accuracy": 0.20702016204595566,
      "num_tokens": 35939005.0,
      "step": 17315
    },
    {
      "entropy": 5.575606870651245,
      "epoch": 1.5711175616835993,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047566854463361544,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20989421457052232,
      "num_tokens": 35948471.0,
      "step": 17320
    },
    {
      "entropy": 5.5107934951782225,
      "epoch": 1.5715711175616836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004756539073889128,
      "loss": 5.0578,
      "mean_token_accuracy": 0.20375427305698396,
      "num_tokens": 35958488.0,
      "step": 17325
    },
    {
      "entropy": 5.51311354637146,
      "epoch": 1.5720246734397678,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047563926599463623,
      "loss": 5.1086,
      "mean_token_accuracy": 0.19656799137592315,
      "num_tokens": 35968291.0,
      "step": 17330
    },
    {
      "entropy": 5.564052724838257,
      "epoch": 1.572478229317852,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004756246204510886,
      "loss": 5.0249,
      "mean_token_accuracy": 0.2046962261199951,
      "num_tokens": 35978565.0,
      "step": 17335
    },
    {
      "entropy": 5.559260129928589,
      "epoch": 1.5729317851959361,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047560997075857253,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20554231852293015,
      "num_tokens": 35987483.0,
      "step": 17340
    },
    {
      "entropy": 5.522519826889038,
      "epoch": 1.5733853410740202,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004755953169173913,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19977228343486786,
      "num_tokens": 35998355.0,
      "step": 17345
    },
    {
      "entropy": 5.510653686523438,
      "epoch": 1.5738388969521044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004755806589278477,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2097836673259735,
      "num_tokens": 36008365.0,
      "step": 17350
    },
    {
      "entropy": 5.576305627822876,
      "epoch": 1.5742924528301887,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047556599679024497,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19741983562707902,
      "num_tokens": 36018880.0,
      "step": 17355
    },
    {
      "entropy": 5.55736198425293,
      "epoch": 1.574746008708273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047555133050488637,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19512535780668258,
      "num_tokens": 36029335.0,
      "step": 17360
    },
    {
      "entropy": 5.5182085037231445,
      "epoch": 1.575199564586357,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047553666007207516,
      "loss": 4.9967,
      "mean_token_accuracy": 0.20793434828519822,
      "num_tokens": 36039774.0,
      "step": 17365
    },
    {
      "entropy": 5.588819742202759,
      "epoch": 1.575653120464441,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047552198549211463,
      "loss": 5.1406,
      "mean_token_accuracy": 0.201657672226429,
      "num_tokens": 36049864.0,
      "step": 17370
    },
    {
      "entropy": 5.533351182937622,
      "epoch": 1.5761066763425253,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004755073067653084,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19713248908519745,
      "num_tokens": 36059224.0,
      "step": 17375
    },
    {
      "entropy": 5.431200313568115,
      "epoch": 1.5765602322206096,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004754926238919599,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21095066368579865,
      "num_tokens": 36069243.0,
      "step": 17380
    },
    {
      "entropy": 5.51810622215271,
      "epoch": 1.5770137880986939,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004754779368723729,
      "loss": 5.0266,
      "mean_token_accuracy": 0.1996670052409172,
      "num_tokens": 36079833.0,
      "step": 17385
    },
    {
      "entropy": 5.519850492477417,
      "epoch": 1.577467343976778,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004754632457068511,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20477883368730546,
      "num_tokens": 36090673.0,
      "step": 17390
    },
    {
      "entropy": 5.55263671875,
      "epoch": 1.5779208998548622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004754485503956982,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20432986617088317,
      "num_tokens": 36101719.0,
      "step": 17395
    },
    {
      "entropy": 5.518518877029419,
      "epoch": 1.5783744557329462,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754338509392183,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20799205005168914,
      "num_tokens": 36112352.0,
      "step": 17400
    },
    {
      "entropy": 5.403399133682251,
      "epoch": 1.5788280116110305,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004754191473377153,
      "loss": 4.9002,
      "mean_token_accuracy": 0.2080320432782173,
      "num_tokens": 36121797.0,
      "step": 17405
    },
    {
      "entropy": 5.507850456237793,
      "epoch": 1.5792815674891147,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047540443959149316,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20719701945781707,
      "num_tokens": 36132094.0,
      "step": 17410
    },
    {
      "entropy": 5.516197729110718,
      "epoch": 1.5797351233671988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004753897277008562,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2037370592355728,
      "num_tokens": 36142860.0,
      "step": 17415
    },
    {
      "entropy": 5.527059650421142,
      "epoch": 1.580188679245283,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004753750116661086,
      "loss": 5.0152,
      "mean_token_accuracy": 0.19879119247198104,
      "num_tokens": 36154555.0,
      "step": 17420
    },
    {
      "entropy": 5.602051639556885,
      "epoch": 1.580642235123367,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047536029148755477,
      "loss": 5.145,
      "mean_token_accuracy": 0.1905045822262764,
      "num_tokens": 36165479.0,
      "step": 17425
    },
    {
      "entropy": 5.570267295837402,
      "epoch": 1.5810957910014514,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047534556716549893,
      "loss": 5.0555,
      "mean_token_accuracy": 0.1985606536269188,
      "num_tokens": 36175873.0,
      "step": 17430
    },
    {
      "entropy": 5.489364194869995,
      "epoch": 1.5815493468795356,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047533083870024575,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20992728173732758,
      "num_tokens": 36186792.0,
      "step": 17435
    },
    {
      "entropy": 5.63928256034851,
      "epoch": 1.58200290275762,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004753161060920998,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1986440509557724,
      "num_tokens": 36196691.0,
      "step": 17440
    },
    {
      "entropy": 5.606499433517456,
      "epoch": 1.582456458635704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004753013693413657,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20787551999092102,
      "num_tokens": 36207173.0,
      "step": 17445
    },
    {
      "entropy": 5.48338131904602,
      "epoch": 1.582910014513788,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047528662844834827,
      "loss": 5.029,
      "mean_token_accuracy": 0.21019935011863708,
      "num_tokens": 36217912.0,
      "step": 17450
    },
    {
      "entropy": 5.420902919769287,
      "epoch": 1.5833635703918723,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047527188341335233,
      "loss": 4.9371,
      "mean_token_accuracy": 0.21276613771915437,
      "num_tokens": 36227714.0,
      "step": 17455
    },
    {
      "entropy": 5.499198532104492,
      "epoch": 1.5838171262699565,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004752571342366828,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2063855767250061,
      "num_tokens": 36239114.0,
      "step": 17460
    },
    {
      "entropy": 5.530270004272461,
      "epoch": 1.5842706821480408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004752423809186446,
      "loss": 5.0342,
      "mean_token_accuracy": 0.19996798485517503,
      "num_tokens": 36249436.0,
      "step": 17465
    },
    {
      "entropy": 5.529721927642822,
      "epoch": 1.5847242380261248,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047522762345954305,
      "loss": 4.9974,
      "mean_token_accuracy": 0.20384032130241395,
      "num_tokens": 36259541.0,
      "step": 17470
    },
    {
      "entropy": 5.571310377120971,
      "epoch": 1.5851777939042089,
      "grad_norm": 1.0,
      "learning_rate": 0.0004752128618596832,
      "loss": 5.046,
      "mean_token_accuracy": 0.19789719432592393,
      "num_tokens": 36270689.0,
      "step": 17475
    },
    {
      "entropy": 5.60941514968872,
      "epoch": 1.5856313497822931,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047519809611937036,
      "loss": 5.0863,
      "mean_token_accuracy": 0.2013432890176773,
      "num_tokens": 36281191.0,
      "step": 17480
    },
    {
      "entropy": 5.47082896232605,
      "epoch": 1.5860849056603774,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751833262389098,
      "loss": 4.9994,
      "mean_token_accuracy": 0.20849235653877257,
      "num_tokens": 36291022.0,
      "step": 17485
    },
    {
      "entropy": 5.536634874343872,
      "epoch": 1.5865384615384617,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047516855221860715,
      "loss": 5.0558,
      "mean_token_accuracy": 0.2028082311153412,
      "num_tokens": 36301237.0,
      "step": 17490
    },
    {
      "entropy": 5.563267469406128,
      "epoch": 1.5869920174165457,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047515377405876774,
      "loss": 5.021,
      "mean_token_accuracy": 0.1983124077320099,
      "num_tokens": 36311792.0,
      "step": 17495
    },
    {
      "entropy": 5.497449779510498,
      "epoch": 1.5874455732946298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004751389917596973,
      "loss": 5.003,
      "mean_token_accuracy": 0.2117833748459816,
      "num_tokens": 36321847.0,
      "step": 17500
    },
    {
      "entropy": 5.534449243545533,
      "epoch": 1.587899129172714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047512420532170154,
      "loss": 5.1105,
      "mean_token_accuracy": 0.19348909854888915,
      "num_tokens": 36332735.0,
      "step": 17505
    },
    {
      "entropy": 5.604989242553711,
      "epoch": 1.5883526850507983,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751094147450862,
      "loss": 5.1031,
      "mean_token_accuracy": 0.19945639073848725,
      "num_tokens": 36343149.0,
      "step": 17510
    },
    {
      "entropy": 5.57188925743103,
      "epoch": 1.5888062409288826,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004750946200301571,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20454846769571305,
      "num_tokens": 36353081.0,
      "step": 17515
    },
    {
      "entropy": 5.513240242004395,
      "epoch": 1.5892597968069666,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047507982117722035,
      "loss": 5.0678,
      "mean_token_accuracy": 0.1978520154953003,
      "num_tokens": 36363624.0,
      "step": 17520
    },
    {
      "entropy": 5.562574768066407,
      "epoch": 1.5897133526850507,
      "grad_norm": 1.03125,
      "learning_rate": 0.000475065018186582,
      "loss": 5.0985,
      "mean_token_accuracy": 0.19767258316278458,
      "num_tokens": 36374081.0,
      "step": 17525
    },
    {
      "entropy": 5.577431631088257,
      "epoch": 1.590166908563135,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047505021105854794,
      "loss": 5.1493,
      "mean_token_accuracy": 0.19680466055870055,
      "num_tokens": 36384360.0,
      "step": 17530
    },
    {
      "entropy": 5.541051292419434,
      "epoch": 1.5906204644412192,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047503539979342456,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20463062524795533,
      "num_tokens": 36395924.0,
      "step": 17535
    },
    {
      "entropy": 5.468084049224854,
      "epoch": 1.5910740203193035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004750205843915182,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21147719472646714,
      "num_tokens": 36405246.0,
      "step": 17540
    },
    {
      "entropy": 5.489252662658691,
      "epoch": 1.5915275761973875,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004750057648531351,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20914095193147658,
      "num_tokens": 36414964.0,
      "step": 17545
    },
    {
      "entropy": 5.516128301620483,
      "epoch": 1.5919811320754715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004749909411785819,
      "loss": 5.0587,
      "mean_token_accuracy": 0.19703762233257294,
      "num_tokens": 36425011.0,
      "step": 17550
    },
    {
      "entropy": 5.495929622650147,
      "epoch": 1.5924346879535558,
      "grad_norm": 1.078125,
      "learning_rate": 0.000474976113368165,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2121456131339073,
      "num_tokens": 36434734.0,
      "step": 17555
    },
    {
      "entropy": 5.524063920974731,
      "epoch": 1.59288824383164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004749612814221912,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2045255184173584,
      "num_tokens": 36444748.0,
      "step": 17560
    },
    {
      "entropy": 5.616097640991211,
      "epoch": 1.5933417997097243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047494644534096703,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19496432393789292,
      "num_tokens": 36455076.0,
      "step": 17565
    },
    {
      "entropy": 5.543533039093018,
      "epoch": 1.5937953555878084,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004749316051247995,
      "loss": 4.9717,
      "mean_token_accuracy": 0.2077886164188385,
      "num_tokens": 36465302.0,
      "step": 17570
    },
    {
      "entropy": 5.510424327850342,
      "epoch": 1.5942489114658926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047491676077399543,
      "loss": 5.0872,
      "mean_token_accuracy": 0.19867370873689652,
      "num_tokens": 36476666.0,
      "step": 17575
    },
    {
      "entropy": 5.487771701812744,
      "epoch": 1.5947024673439767,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004749019122888618,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2082185000181198,
      "num_tokens": 36487379.0,
      "step": 17580
    },
    {
      "entropy": 5.527111959457398,
      "epoch": 1.595156023222061,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047488705966970565,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2124447599053383,
      "num_tokens": 36497897.0,
      "step": 17585
    },
    {
      "entropy": 5.54830584526062,
      "epoch": 1.5956095791001452,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004748722029168342,
      "loss": 5.0678,
      "mean_token_accuracy": 0.2106671467423439,
      "num_tokens": 36508528.0,
      "step": 17590
    },
    {
      "entropy": 5.5429380416870115,
      "epoch": 1.5960631349782293,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004748573420305546,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19697727560997008,
      "num_tokens": 36519196.0,
      "step": 17595
    },
    {
      "entropy": 5.5159355163574215,
      "epoch": 1.5965166908563135,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047484247701117427,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20312255024909973,
      "num_tokens": 36529871.0,
      "step": 17600
    },
    {
      "entropy": 5.626338863372803,
      "epoch": 1.5969702467343976,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004748276078590006,
      "loss": 5.1236,
      "mean_token_accuracy": 0.1941250815987587,
      "num_tokens": 36541188.0,
      "step": 17605
    },
    {
      "entropy": 5.58647198677063,
      "epoch": 1.5974238026124818,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004748127345743411,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19909974187612534,
      "num_tokens": 36551968.0,
      "step": 17610
    },
    {
      "entropy": 5.530744457244873,
      "epoch": 1.5978773584905661,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004747978571575033,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20223728865385054,
      "num_tokens": 36562870.0,
      "step": 17615
    },
    {
      "entropy": 5.479086828231812,
      "epoch": 1.5983309143686504,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004747829756087949,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20745939463377,
      "num_tokens": 36573295.0,
      "step": 17620
    },
    {
      "entropy": 5.479890394210815,
      "epoch": 1.5987844702467344,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047476808992852364,
      "loss": 4.9653,
      "mean_token_accuracy": 0.20873234122991563,
      "num_tokens": 36583830.0,
      "step": 17625
    },
    {
      "entropy": 5.517432165145874,
      "epoch": 1.5992380261248185,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004747532001169974,
      "loss": 5.0462,
      "mean_token_accuracy": 0.2130327671766281,
      "num_tokens": 36594052.0,
      "step": 17630
    },
    {
      "entropy": 5.520537948608398,
      "epoch": 1.5996915820029027,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047473830617452415,
      "loss": 5.0696,
      "mean_token_accuracy": 0.20406648516654968,
      "num_tokens": 36604378.0,
      "step": 17635
    },
    {
      "entropy": 5.596294307708741,
      "epoch": 1.600145137880987,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047472340810141163,
      "loss": 5.1623,
      "mean_token_accuracy": 0.19078289419412614,
      "num_tokens": 36615588.0,
      "step": 17640
    },
    {
      "entropy": 5.626055479049683,
      "epoch": 1.6005986937590713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047470850589796833,
      "loss": 5.0402,
      "mean_token_accuracy": 0.1986627995967865,
      "num_tokens": 36626369.0,
      "step": 17645
    },
    {
      "entropy": 5.5786247730255125,
      "epoch": 1.6010522496371553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047469359956450215,
      "loss": 5.0791,
      "mean_token_accuracy": 0.1997091665863991,
      "num_tokens": 36636565.0,
      "step": 17650
    },
    {
      "entropy": 5.611868000030517,
      "epoch": 1.6015058055152394,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047467868910132144,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1966838225722313,
      "num_tokens": 36647721.0,
      "step": 17655
    },
    {
      "entropy": 5.495266342163086,
      "epoch": 1.6019593613933236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004746637745087346,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21024642437696456,
      "num_tokens": 36657887.0,
      "step": 17660
    },
    {
      "entropy": 5.604389238357544,
      "epoch": 1.6024129172714079,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047464885578705,
      "loss": 5.0494,
      "mean_token_accuracy": 0.19974400252103805,
      "num_tokens": 36667918.0,
      "step": 17665
    },
    {
      "entropy": 5.472668647766113,
      "epoch": 1.6028664731494922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047463393293657616,
      "loss": 4.9976,
      "mean_token_accuracy": 0.2072445273399353,
      "num_tokens": 36678491.0,
      "step": 17670
    },
    {
      "entropy": 5.5294928550720215,
      "epoch": 1.6033200290275762,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004746190059576217,
      "loss": 5.0594,
      "mean_token_accuracy": 0.2025817811489105,
      "num_tokens": 36688638.0,
      "step": 17675
    },
    {
      "entropy": 5.555827474594116,
      "epoch": 1.6037735849056602,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004746040748504954,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20663116574287416,
      "num_tokens": 36699344.0,
      "step": 17680
    },
    {
      "entropy": 5.555319690704346,
      "epoch": 1.6042271407837445,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047458913961550593,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20222714096307753,
      "num_tokens": 36709606.0,
      "step": 17685
    },
    {
      "entropy": 5.463090229034424,
      "epoch": 1.6046806966618288,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004745742002529622,
      "loss": 4.9625,
      "mean_token_accuracy": 0.20935799330472946,
      "num_tokens": 36718740.0,
      "step": 17690
    },
    {
      "entropy": 5.522600412368774,
      "epoch": 1.605134252539913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004745592567631731,
      "loss": 5.0515,
      "mean_token_accuracy": 0.19677862673997878,
      "num_tokens": 36729960.0,
      "step": 17695
    },
    {
      "entropy": 5.4693926811218265,
      "epoch": 1.605587808417997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047454430914644773,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21306394040584564,
      "num_tokens": 36740221.0,
      "step": 17700
    },
    {
      "entropy": 5.427061891555786,
      "epoch": 1.6060413642960811,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004745293574030952,
      "loss": 4.9615,
      "mean_token_accuracy": 0.213801009953022,
      "num_tokens": 36750841.0,
      "step": 17705
    },
    {
      "entropy": 5.497584199905395,
      "epoch": 1.6064949201741654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047451440153342477,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19926677495241166,
      "num_tokens": 36760944.0,
      "step": 17710
    },
    {
      "entropy": 5.571436548233033,
      "epoch": 1.6069484760522497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004744994415377457,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20537643432617186,
      "num_tokens": 36770922.0,
      "step": 17715
    },
    {
      "entropy": 5.579338598251343,
      "epoch": 1.607402031930334,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047448447741636724,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20064235776662825,
      "num_tokens": 36780702.0,
      "step": 17720
    },
    {
      "entropy": 5.577346754074097,
      "epoch": 1.607855587808418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047446950916959905,
      "loss": 5.082,
      "mean_token_accuracy": 0.19528306573629378,
      "num_tokens": 36791868.0,
      "step": 17725
    },
    {
      "entropy": 5.585596323013306,
      "epoch": 1.608309143686502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004744545367977505,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2042024776339531,
      "num_tokens": 36802330.0,
      "step": 17730
    },
    {
      "entropy": 5.558025169372558,
      "epoch": 1.6087626995645863,
      "grad_norm": 1.125,
      "learning_rate": 0.00047443956030113134,
      "loss": 5.0891,
      "mean_token_accuracy": 0.1996786117553711,
      "num_tokens": 36812936.0,
      "step": 17735
    },
    {
      "entropy": 5.577855634689331,
      "epoch": 1.6092162554426706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047442457968005123,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20214971750974656,
      "num_tokens": 36823123.0,
      "step": 17740
    },
    {
      "entropy": 5.4617472171783445,
      "epoch": 1.6096698113207548,
      "grad_norm": 1.125,
      "learning_rate": 0.00047440959493481996,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2160551816225052,
      "num_tokens": 36833050.0,
      "step": 17745
    },
    {
      "entropy": 5.5303796291351315,
      "epoch": 1.6101233671988389,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047439460606574745,
      "loss": 5.0548,
      "mean_token_accuracy": 0.20608482956886293,
      "num_tokens": 36842520.0,
      "step": 17750
    },
    {
      "entropy": 5.546290588378906,
      "epoch": 1.6105769230769231,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004743796130731437,
      "loss": 5.0162,
      "mean_token_accuracy": 0.2015488028526306,
      "num_tokens": 36853024.0,
      "step": 17755
    },
    {
      "entropy": 5.547762823104859,
      "epoch": 1.6110304789550072,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004743646159573188,
      "loss": 4.9728,
      "mean_token_accuracy": 0.21105591356754302,
      "num_tokens": 36862534.0,
      "step": 17760
    },
    {
      "entropy": 5.516180324554443,
      "epoch": 1.6114840348330914,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004743496147185828,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21105688959360122,
      "num_tokens": 36872356.0,
      "step": 17765
    },
    {
      "entropy": 5.461830472946167,
      "epoch": 1.6119375907111757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047433460935724595,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20244371742010117,
      "num_tokens": 36882643.0,
      "step": 17770
    },
    {
      "entropy": 5.5579465389251705,
      "epoch": 1.6123911465892597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004743195998736186,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20161498188972474,
      "num_tokens": 36893349.0,
      "step": 17775
    },
    {
      "entropy": 5.6073261260986325,
      "epoch": 1.612844702467344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047430458626801107,
      "loss": 5.0289,
      "mean_token_accuracy": 0.20888873934745789,
      "num_tokens": 36903713.0,
      "step": 17780
    },
    {
      "entropy": 5.512869548797608,
      "epoch": 1.613298258345428,
      "grad_norm": 1.109375,
      "learning_rate": 0.000474289568540734,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21190437525510789,
      "num_tokens": 36912999.0,
      "step": 17785
    },
    {
      "entropy": 5.560475158691406,
      "epoch": 1.6137518142235123,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047427454669209785,
      "loss": 5.0708,
      "mean_token_accuracy": 0.2030900776386261,
      "num_tokens": 36923598.0,
      "step": 17790
    },
    {
      "entropy": 5.521582317352295,
      "epoch": 1.6142053701015966,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047425952072241326,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20506981611251832,
      "num_tokens": 36933036.0,
      "step": 17795
    },
    {
      "entropy": 5.536417770385742,
      "epoch": 1.6146589259796809,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000474244490631991,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19817693680524825,
      "num_tokens": 36943137.0,
      "step": 17800
    },
    {
      "entropy": 5.649659061431885,
      "epoch": 1.615112481857765,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004742294564211419,
      "loss": 5.2031,
      "mean_token_accuracy": 0.19474415481090546,
      "num_tokens": 36955619.0,
      "step": 17805
    },
    {
      "entropy": 5.543700122833252,
      "epoch": 1.615566037735849,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004742144180901769,
      "loss": 4.973,
      "mean_token_accuracy": 0.21407785415649414,
      "num_tokens": 36966377.0,
      "step": 17810
    },
    {
      "entropy": 5.564962434768677,
      "epoch": 1.6160195936139332,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047419937563940697,
      "loss": 5.0333,
      "mean_token_accuracy": 0.2056509166955948,
      "num_tokens": 36977100.0,
      "step": 17815
    },
    {
      "entropy": 5.557198762893677,
      "epoch": 1.6164731494920175,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047418432906914314,
      "loss": 5.1014,
      "mean_token_accuracy": 0.20024712532758712,
      "num_tokens": 36987120.0,
      "step": 17820
    },
    {
      "entropy": 5.593332052230835,
      "epoch": 1.6169267053701017,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004741692783796967,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20203897207975388,
      "num_tokens": 36996885.0,
      "step": 17825
    },
    {
      "entropy": 5.544119739532471,
      "epoch": 1.6173802612481858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047415422357137876,
      "loss": 4.9721,
      "mean_token_accuracy": 0.21137814223766327,
      "num_tokens": 37006570.0,
      "step": 17830
    },
    {
      "entropy": 5.594113206863403,
      "epoch": 1.6178338171262698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004741391646445008,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20022266656160354,
      "num_tokens": 37017185.0,
      "step": 17835
    },
    {
      "entropy": 5.547800064086914,
      "epoch": 1.618287373004354,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004741241015993741,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20494076907634734,
      "num_tokens": 37028386.0,
      "step": 17840
    },
    {
      "entropy": 5.493295669555664,
      "epoch": 1.6187409288824384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047410903443631026,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20483069121837616,
      "num_tokens": 37038326.0,
      "step": 17845
    },
    {
      "entropy": 5.548564386367798,
      "epoch": 1.6191944847605226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004740939631556209,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20101993530988693,
      "num_tokens": 37048570.0,
      "step": 17850
    },
    {
      "entropy": 5.505143356323242,
      "epoch": 1.6196480406386067,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004740788877576176,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2146276816725731,
      "num_tokens": 37058909.0,
      "step": 17855
    },
    {
      "entropy": 5.533849716186523,
      "epoch": 1.6201015965166907,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004740638082426121,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21495533287525176,
      "num_tokens": 37068957.0,
      "step": 17860
    },
    {
      "entropy": 5.539355087280273,
      "epoch": 1.620555152394775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047404872461091643,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20215528160333635,
      "num_tokens": 37080187.0,
      "step": 17865
    },
    {
      "entropy": 5.5432701587677,
      "epoch": 1.6210087082728593,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047403363686284236,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20310061573982238,
      "num_tokens": 37090314.0,
      "step": 17870
    },
    {
      "entropy": 5.5888636112213135,
      "epoch": 1.6214622641509435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047401854499870197,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20264121890068054,
      "num_tokens": 37099907.0,
      "step": 17875
    },
    {
      "entropy": 5.561094284057617,
      "epoch": 1.6219158200290276,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047400344901880735,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1975569471716881,
      "num_tokens": 37110495.0,
      "step": 17880
    },
    {
      "entropy": 5.553846216201782,
      "epoch": 1.6223693759071116,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004739883489234707,
      "loss": 5.065,
      "mean_token_accuracy": 0.19802882373332978,
      "num_tokens": 37121957.0,
      "step": 17885
    },
    {
      "entropy": 5.539548778533936,
      "epoch": 1.6228229317851959,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047397324471300424,
      "loss": 4.9772,
      "mean_token_accuracy": 0.210438971221447,
      "num_tokens": 37132743.0,
      "step": 17890
    },
    {
      "entropy": 5.562828397750854,
      "epoch": 1.6232764876632801,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047395813638772047,
      "loss": 5.0776,
      "mean_token_accuracy": 0.1967625856399536,
      "num_tokens": 37143947.0,
      "step": 17895
    },
    {
      "entropy": 5.6736279964447025,
      "epoch": 1.6237300435413644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047394302394793164,
      "loss": 5.172,
      "mean_token_accuracy": 0.1933992624282837,
      "num_tokens": 37154316.0,
      "step": 17900
    },
    {
      "entropy": 5.604281616210938,
      "epoch": 1.6241835994194485,
      "grad_norm": 1.0,
      "learning_rate": 0.0004739279073939504,
      "loss": 5.073,
      "mean_token_accuracy": 0.20024010837078093,
      "num_tokens": 37164999.0,
      "step": 17905
    },
    {
      "entropy": 5.56506929397583,
      "epoch": 1.6246371552975325,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004739127867260894,
      "loss": 5.075,
      "mean_token_accuracy": 0.20138372778892516,
      "num_tokens": 37174392.0,
      "step": 17910
    },
    {
      "entropy": 5.48519606590271,
      "epoch": 1.6250907111756168,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004738976619446613,
      "loss": 5.056,
      "mean_token_accuracy": 0.20201866924762726,
      "num_tokens": 37185341.0,
      "step": 17915
    },
    {
      "entropy": 5.580339765548706,
      "epoch": 1.625544267053701,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047388253304997874,
      "loss": 5.0779,
      "mean_token_accuracy": 0.1985635995864868,
      "num_tokens": 37196725.0,
      "step": 17920
    },
    {
      "entropy": 5.555712938308716,
      "epoch": 1.6259978229317853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004738674000423548,
      "loss": 4.9466,
      "mean_token_accuracy": 0.20997316688299178,
      "num_tokens": 37206877.0,
      "step": 17925
    },
    {
      "entropy": 5.458779954910279,
      "epoch": 1.6264513788098693,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004738522629221023,
      "loss": 4.9571,
      "mean_token_accuracy": 0.20690305680036544,
      "num_tokens": 37217277.0,
      "step": 17930
    },
    {
      "entropy": 5.475347471237183,
      "epoch": 1.6269049346879536,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004738371216895344,
      "loss": 4.934,
      "mean_token_accuracy": 0.2105925515294075,
      "num_tokens": 37227716.0,
      "step": 17935
    },
    {
      "entropy": 5.506228351593018,
      "epoch": 1.6273584905660377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004738219763449641,
      "loss": 5.047,
      "mean_token_accuracy": 0.20263891220092772,
      "num_tokens": 37238178.0,
      "step": 17940
    },
    {
      "entropy": 5.525771284103394,
      "epoch": 1.627812046444122,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004738068268887047,
      "loss": 5.0732,
      "mean_token_accuracy": 0.2064121648669243,
      "num_tokens": 37248408.0,
      "step": 17945
    },
    {
      "entropy": 5.564480209350586,
      "epoch": 1.6282656023222062,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004737916733210694,
      "loss": 5.1805,
      "mean_token_accuracy": 0.1948343485593796,
      "num_tokens": 37259176.0,
      "step": 17950
    },
    {
      "entropy": 5.55413498878479,
      "epoch": 1.6287191582002902,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047377651564237165,
      "loss": 4.9403,
      "mean_token_accuracy": 0.21000248938798904,
      "num_tokens": 37269205.0,
      "step": 17955
    },
    {
      "entropy": 5.556760597229004,
      "epoch": 1.6291727140783745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047376135385292494,
      "loss": 5.0615,
      "mean_token_accuracy": 0.20742124766111375,
      "num_tokens": 37278841.0,
      "step": 17960
    },
    {
      "entropy": 5.571680688858033,
      "epoch": 1.6296262699564585,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004737461879530428,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20085595697164535,
      "num_tokens": 37288487.0,
      "step": 17965
    },
    {
      "entropy": 5.593708944320679,
      "epoch": 1.6300798258345428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047373101794303884,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2009998604655266,
      "num_tokens": 37300086.0,
      "step": 17970
    },
    {
      "entropy": 5.503368282318116,
      "epoch": 1.630533381712627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004737158438232267,
      "loss": 5.0912,
      "mean_token_accuracy": 0.19563519060611725,
      "num_tokens": 37310072.0,
      "step": 17975
    },
    {
      "entropy": 5.5453534603118895,
      "epoch": 1.6309869375907113,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047370066559392033,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21010988503694533,
      "num_tokens": 37320278.0,
      "step": 17980
    },
    {
      "entropy": 5.560372066497803,
      "epoch": 1.6314404934687954,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047368548325543366,
      "loss": 4.9913,
      "mean_token_accuracy": 0.19833416491746902,
      "num_tokens": 37330741.0,
      "step": 17985
    },
    {
      "entropy": 5.562993144989013,
      "epoch": 1.6318940493468794,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004736702968080804,
      "loss": 5.076,
      "mean_token_accuracy": 0.19929848611354828,
      "num_tokens": 37341213.0,
      "step": 17990
    },
    {
      "entropy": 5.496298837661743,
      "epoch": 1.6323476052249637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047365510625217483,
      "loss": 4.9298,
      "mean_token_accuracy": 0.20807789713144303,
      "num_tokens": 37350528.0,
      "step": 17995
    },
    {
      "entropy": 5.528117418289185,
      "epoch": 1.632801161103048,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047363991158803117,
      "loss": 5.0309,
      "mean_token_accuracy": 0.20054963678121568,
      "num_tokens": 37359938.0,
      "step": 18000
    },
    {
      "epoch": 1.632801161103048,
      "eval_entropy": 5.307341233098129,
      "eval_loss": 5.06912899017334,
      "eval_mean_token_accuracy": 0.21036261755180843,
      "eval_num_tokens": 37359938.0,
      "eval_runtime": 20.6219,
      "eval_samples_per_second": 1714.681,
      "eval_steps_per_second": 214.335,
      "step": 18000
    },
    {
      "entropy": 5.560543966293335,
      "epoch": 1.6332547169811322,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047362471281596334,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20620202720165254,
      "num_tokens": 37369744.0,
      "step": 18005
    },
    {
      "entropy": 5.596627044677734,
      "epoch": 1.6337082728592163,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047360950993628585,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20341364592313765,
      "num_tokens": 37380578.0,
      "step": 18010
    },
    {
      "entropy": 5.550653743743896,
      "epoch": 1.6341618287373003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004735943029493132,
      "loss": 5.0317,
      "mean_token_accuracy": 0.2036425366997719,
      "num_tokens": 37390237.0,
      "step": 18015
    },
    {
      "entropy": 5.574190616607666,
      "epoch": 1.6346153846153846,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047357909185535965,
      "loss": 5.0872,
      "mean_token_accuracy": 0.1972410872578621,
      "num_tokens": 37400261.0,
      "step": 18020
    },
    {
      "entropy": 5.560031843185425,
      "epoch": 1.6350689404934688,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004735638766547399,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2051742523908615,
      "num_tokens": 37410012.0,
      "step": 18025
    },
    {
      "entropy": 5.576835584640503,
      "epoch": 1.6355224963715531,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047354865734776857,
      "loss": 5.1316,
      "mean_token_accuracy": 0.192862069606781,
      "num_tokens": 37420530.0,
      "step": 18030
    },
    {
      "entropy": 5.5550755023956295,
      "epoch": 1.6359760522496372,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004735334339347605,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21534911990165712,
      "num_tokens": 37431146.0,
      "step": 18035
    },
    {
      "entropy": 5.548226690292358,
      "epoch": 1.6364296081277212,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047351820641603035,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20894050747156143,
      "num_tokens": 37441253.0,
      "step": 18040
    },
    {
      "entropy": 5.463728237152099,
      "epoch": 1.6368831640058055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004735029747918931,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20038239955902098,
      "num_tokens": 37451326.0,
      "step": 18045
    },
    {
      "entropy": 5.466217184066773,
      "epoch": 1.6373367198838897,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047348773906266374,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2092629000544548,
      "num_tokens": 37462462.0,
      "step": 18050
    },
    {
      "entropy": 5.486484622955322,
      "epoch": 1.637790275761974,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004734724992286574,
      "loss": 4.9832,
      "mean_token_accuracy": 0.20355282127857208,
      "num_tokens": 37473016.0,
      "step": 18055
    },
    {
      "entropy": 5.569339752197266,
      "epoch": 1.638243831640058,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004734572552901892,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20118896216154097,
      "num_tokens": 37483737.0,
      "step": 18060
    },
    {
      "entropy": 5.526478099822998,
      "epoch": 1.638697387518142,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004734420072475743,
      "loss": 4.9928,
      "mean_token_accuracy": 0.20455218851566315,
      "num_tokens": 37493070.0,
      "step": 18065
    },
    {
      "entropy": 5.51882038116455,
      "epoch": 1.6391509433962264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004734267551011282,
      "loss": 4.995,
      "mean_token_accuracy": 0.20797305405139924,
      "num_tokens": 37503696.0,
      "step": 18070
    },
    {
      "entropy": 5.495640802383423,
      "epoch": 1.6396044992743106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004734114988511662,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20355913639068604,
      "num_tokens": 37513468.0,
      "step": 18075
    },
    {
      "entropy": 5.449446725845337,
      "epoch": 1.640058055152395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004733962384980039,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2130780801177025,
      "num_tokens": 37523631.0,
      "step": 18080
    },
    {
      "entropy": 5.569212102890015,
      "epoch": 1.640511611030479,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004733809740419568,
      "loss": 5.0759,
      "mean_token_accuracy": 0.20668964982032775,
      "num_tokens": 37534602.0,
      "step": 18085
    },
    {
      "entropy": 5.457507658004761,
      "epoch": 1.640965166908563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047336570548334053,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2021588295698166,
      "num_tokens": 37544547.0,
      "step": 18090
    },
    {
      "entropy": 5.548616027832031,
      "epoch": 1.6414187227866472,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047335043282247105,
      "loss": 5.0301,
      "mean_token_accuracy": 0.20326164364814758,
      "num_tokens": 37554903.0,
      "step": 18095
    },
    {
      "entropy": 5.55193772315979,
      "epoch": 1.6418722786647315,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047333515605966396,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2056132212281227,
      "num_tokens": 37565586.0,
      "step": 18100
    },
    {
      "entropy": 5.536584043502808,
      "epoch": 1.6423258345428158,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047331987519523533,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2078459829092026,
      "num_tokens": 37575996.0,
      "step": 18105
    },
    {
      "entropy": 5.482731580734253,
      "epoch": 1.6427793904208998,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004733045902295012,
      "loss": 4.941,
      "mean_token_accuracy": 0.2008867606520653,
      "num_tokens": 37586441.0,
      "step": 18110
    },
    {
      "entropy": 5.551346015930176,
      "epoch": 1.643232946298984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047328930116277755,
      "loss": 5.055,
      "mean_token_accuracy": 0.2052132710814476,
      "num_tokens": 37598226.0,
      "step": 18115
    },
    {
      "entropy": 5.542083644866944,
      "epoch": 1.6436865021770681,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004732740079953806,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20656532049179077,
      "num_tokens": 37609102.0,
      "step": 18120
    },
    {
      "entropy": 5.545897531509399,
      "epoch": 1.6441400580551524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004732587107276267,
      "loss": 5.0499,
      "mean_token_accuracy": 0.2018830582499504,
      "num_tokens": 37619349.0,
      "step": 18125
    },
    {
      "entropy": 5.512788105010986,
      "epoch": 1.6445936139332367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047324340935983225,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20966263711452485,
      "num_tokens": 37629844.0,
      "step": 18130
    },
    {
      "entropy": 5.482539653778076,
      "epoch": 1.6450471698113207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047322810389231333,
      "loss": 4.9394,
      "mean_token_accuracy": 0.20833370983600616,
      "num_tokens": 37639339.0,
      "step": 18135
    },
    {
      "entropy": 5.502611827850342,
      "epoch": 1.645500725689405,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004732127943253868,
      "loss": 4.978,
      "mean_token_accuracy": 0.21004469245672225,
      "num_tokens": 37648836.0,
      "step": 18140
    },
    {
      "entropy": 5.584582233428955,
      "epoch": 1.645954281567489,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004731974806593693,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2051238164305687,
      "num_tokens": 37658424.0,
      "step": 18145
    },
    {
      "entropy": 5.549859142303466,
      "epoch": 1.6464078374455733,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047318216289457727,
      "loss": 5.0074,
      "mean_token_accuracy": 0.21039813309907912,
      "num_tokens": 37669849.0,
      "step": 18150
    },
    {
      "entropy": 5.505876684188843,
      "epoch": 1.6468613933236576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047316684103132766,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20256547927856444,
      "num_tokens": 37678987.0,
      "step": 18155
    },
    {
      "entropy": 5.555306529998779,
      "epoch": 1.6473149492017418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004731515150699372,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20816829204559326,
      "num_tokens": 37689692.0,
      "step": 18160
    },
    {
      "entropy": 5.585298776626587,
      "epoch": 1.6477685050798259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047313618501072303,
      "loss": 5.1711,
      "mean_token_accuracy": 0.19286765158176422,
      "num_tokens": 37700092.0,
      "step": 18165
    },
    {
      "entropy": 5.5218486309051515,
      "epoch": 1.64822206095791,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047312085085400194,
      "loss": 5.065,
      "mean_token_accuracy": 0.20425543636083604,
      "num_tokens": 37710305.0,
      "step": 18170
    },
    {
      "entropy": 5.658654689788818,
      "epoch": 1.6486756168359942,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047310551260009126,
      "loss": 5.1307,
      "mean_token_accuracy": 0.1929813876748085,
      "num_tokens": 37720451.0,
      "step": 18175
    },
    {
      "entropy": 5.577204895019531,
      "epoch": 1.6491291727140784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000473090170249308,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20589849203824998,
      "num_tokens": 37731952.0,
      "step": 18180
    },
    {
      "entropy": 5.489793586730957,
      "epoch": 1.6495827285921627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004730748238019696,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20154588669538498,
      "num_tokens": 37742201.0,
      "step": 18185
    },
    {
      "entropy": 5.474636507034302,
      "epoch": 1.6500362844702468,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047305947325839337,
      "loss": 4.9928,
      "mean_token_accuracy": 0.20665855407714845,
      "num_tokens": 37753317.0,
      "step": 18190
    },
    {
      "entropy": 5.58493447303772,
      "epoch": 1.6504898403483308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004730441186188968,
      "loss": 5.0549,
      "mean_token_accuracy": 0.2018160879611969,
      "num_tokens": 37763452.0,
      "step": 18195
    },
    {
      "entropy": 5.551719331741333,
      "epoch": 1.650943396226415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047302875988379734,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2069983720779419,
      "num_tokens": 37772804.0,
      "step": 18200
    },
    {
      "entropy": 5.548997497558593,
      "epoch": 1.6513969521044993,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047301339705341266,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19575041085481643,
      "num_tokens": 37784441.0,
      "step": 18205
    },
    {
      "entropy": 5.529232168197632,
      "epoch": 1.6518505079825836,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047299803012806056,
      "loss": 5.0436,
      "mean_token_accuracy": 0.2065920427441597,
      "num_tokens": 37795103.0,
      "step": 18210
    },
    {
      "entropy": 5.487066984176636,
      "epoch": 1.6523040638606676,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004729826591080586,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20657159388065338,
      "num_tokens": 37806654.0,
      "step": 18215
    },
    {
      "entropy": 5.526598072052002,
      "epoch": 1.6527576197387517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047296728399372483,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20016657412052155,
      "num_tokens": 37816673.0,
      "step": 18220
    },
    {
      "entropy": 5.555908298492431,
      "epoch": 1.653211175616836,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047295190478537725,
      "loss": 5.0503,
      "mean_token_accuracy": 0.2029751569032669,
      "num_tokens": 37827389.0,
      "step": 18225
    },
    {
      "entropy": 5.658416748046875,
      "epoch": 1.6536647314949202,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047293652148333376,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20344164222478867,
      "num_tokens": 37838087.0,
      "step": 18230
    },
    {
      "entropy": 5.4327747344970705,
      "epoch": 1.6541182873730045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047292113408791266,
      "loss": 4.9477,
      "mean_token_accuracy": 0.20138545483350753,
      "num_tokens": 37848630.0,
      "step": 18235
    },
    {
      "entropy": 5.453988885879516,
      "epoch": 1.6545718432510885,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047290574259943195,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22004650682210922,
      "num_tokens": 37858689.0,
      "step": 18240
    },
    {
      "entropy": 5.6077797412872314,
      "epoch": 1.6550253991291726,
      "grad_norm": 1.0,
      "learning_rate": 0.00047289034701821013,
      "loss": 5.0897,
      "mean_token_accuracy": 0.20234557688236238,
      "num_tokens": 37869049.0,
      "step": 18245
    },
    {
      "entropy": 5.5167817115783695,
      "epoch": 1.6554789550072568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047287494734456545,
      "loss": 4.9994,
      "mean_token_accuracy": 0.2003646969795227,
      "num_tokens": 37879495.0,
      "step": 18250
    },
    {
      "entropy": 5.544087553024292,
      "epoch": 1.655932510885341,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047285954357881646,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2015921652317047,
      "num_tokens": 37889442.0,
      "step": 18255
    },
    {
      "entropy": 5.598650503158569,
      "epoch": 1.6563860667634254,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004728441357212817,
      "loss": 5.1712,
      "mean_token_accuracy": 0.1892197161912918,
      "num_tokens": 37900147.0,
      "step": 18260
    },
    {
      "entropy": 5.6057878017425535,
      "epoch": 1.6568396226415094,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004728287237722798,
      "loss": 4.9953,
      "mean_token_accuracy": 0.20190392732620238,
      "num_tokens": 37910307.0,
      "step": 18265
    },
    {
      "entropy": 5.492194032669067,
      "epoch": 1.6572931785195935,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004728133077321295,
      "loss": 4.9683,
      "mean_token_accuracy": 0.20334598273038865,
      "num_tokens": 37920838.0,
      "step": 18270
    },
    {
      "entropy": 5.546195602416992,
      "epoch": 1.6577467343976777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047279788760114954,
      "loss": 5.0489,
      "mean_token_accuracy": 0.19957167655229568,
      "num_tokens": 37930717.0,
      "step": 18275
    },
    {
      "entropy": 5.565794944763184,
      "epoch": 1.658200290275762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004727824633796589,
      "loss": 5.0443,
      "mean_token_accuracy": 0.2027788817882538,
      "num_tokens": 37941254.0,
      "step": 18280
    },
    {
      "entropy": 5.510905408859253,
      "epoch": 1.6586538461538463,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004727670350679765,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20850671380758284,
      "num_tokens": 37952132.0,
      "step": 18285
    },
    {
      "entropy": 5.513995122909546,
      "epoch": 1.6591074020319303,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004727516026664214,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20742790400981903,
      "num_tokens": 37962026.0,
      "step": 18290
    },
    {
      "entropy": 5.5256506443023685,
      "epoch": 1.6595609579100146,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004727361661753128,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20238326042890548,
      "num_tokens": 37972911.0,
      "step": 18295
    },
    {
      "entropy": 5.5586512088775635,
      "epoch": 1.6600145137880986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004727207255949699,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20900478065013886,
      "num_tokens": 37983577.0,
      "step": 18300
    },
    {
      "entropy": 5.5725830078125,
      "epoch": 1.6604680696661829,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047270528092571197,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20097507536411285,
      "num_tokens": 37994294.0,
      "step": 18305
    },
    {
      "entropy": 5.538248825073242,
      "epoch": 1.6609216255442671,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004726898321678584,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20685731917619704,
      "num_tokens": 38005144.0,
      "step": 18310
    },
    {
      "entropy": 5.4818648338317875,
      "epoch": 1.6613751814223512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726743793217288,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20080667734146118,
      "num_tokens": 38015277.0,
      "step": 18315
    },
    {
      "entropy": 5.5463134288787845,
      "epoch": 1.6618287373004355,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047265892238764265,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20633944720029831,
      "num_tokens": 38027530.0,
      "step": 18320
    },
    {
      "entropy": 5.617530441284179,
      "epoch": 1.6622822931785195,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004726434613659196,
      "loss": 5.0891,
      "mean_token_accuracy": 0.20411063283681868,
      "num_tokens": 38036895.0,
      "step": 18325
    },
    {
      "entropy": 5.583403873443603,
      "epoch": 1.6627358490566038,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004726279962568794,
      "loss": 5.1012,
      "mean_token_accuracy": 0.2044348746538162,
      "num_tokens": 38047001.0,
      "step": 18330
    },
    {
      "entropy": 5.554082775115967,
      "epoch": 1.663189404934688,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004726125270608419,
      "loss": 4.975,
      "mean_token_accuracy": 0.2108441025018692,
      "num_tokens": 38056815.0,
      "step": 18335
    },
    {
      "entropy": 5.568605566024781,
      "epoch": 1.6636429608127723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047259705377812703,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20433802753686905,
      "num_tokens": 38067139.0,
      "step": 18340
    },
    {
      "entropy": 5.548576307296753,
      "epoch": 1.6640965166908563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004725815764090546,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20833301097154616,
      "num_tokens": 38076824.0,
      "step": 18345
    },
    {
      "entropy": 5.536114835739136,
      "epoch": 1.6645500725689404,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047256609495394494,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20151522010564804,
      "num_tokens": 38087517.0,
      "step": 18350
    },
    {
      "entropy": 5.568885850906372,
      "epoch": 1.6650036284470247,
      "grad_norm": 1.03125,
      "learning_rate": 0.000472550609413118,
      "loss": 5.0186,
      "mean_token_accuracy": 0.20044526159763337,
      "num_tokens": 38098499.0,
      "step": 18355
    },
    {
      "entropy": 5.48836989402771,
      "epoch": 1.665457184325109,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004725351197868942,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2095501482486725,
      "num_tokens": 38108327.0,
      "step": 18360
    },
    {
      "entropy": 5.598924016952514,
      "epoch": 1.6659107402031932,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047251962607559373,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20512547791004182,
      "num_tokens": 38118667.0,
      "step": 18365
    },
    {
      "entropy": 5.5005449771881105,
      "epoch": 1.6663642960812772,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047250412827953717,
      "loss": 4.9998,
      "mean_token_accuracy": 0.20354425311088561,
      "num_tokens": 38128571.0,
      "step": 18370
    },
    {
      "entropy": 5.588938045501709,
      "epoch": 1.6668178519593613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004724886263990448,
      "loss": 5.1049,
      "mean_token_accuracy": 0.195250703394413,
      "num_tokens": 38138921.0,
      "step": 18375
    },
    {
      "entropy": 5.591016340255737,
      "epoch": 1.6672714078374455,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004724731204344372,
      "loss": 5.0368,
      "mean_token_accuracy": 0.20534040927886962,
      "num_tokens": 38148858.0,
      "step": 18380
    },
    {
      "entropy": 5.462696266174317,
      "epoch": 1.6677249637155298,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047245761038603536,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20529389679431914,
      "num_tokens": 38159107.0,
      "step": 18385
    },
    {
      "entropy": 5.618238306045532,
      "epoch": 1.668178519593614,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047244209625415975,
      "loss": 5.0921,
      "mean_token_accuracy": 0.20097823292016984,
      "num_tokens": 38169228.0,
      "step": 18390
    },
    {
      "entropy": 5.603679180145264,
      "epoch": 1.6686320754716981,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004724265780391312,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2068103551864624,
      "num_tokens": 38179175.0,
      "step": 18395
    },
    {
      "entropy": 5.556495428085327,
      "epoch": 1.6690856313497822,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004724110557412708,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20790047347545623,
      "num_tokens": 38188537.0,
      "step": 18400
    },
    {
      "entropy": 5.514718437194825,
      "epoch": 1.6695391872278664,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047239552936089943,
      "loss": 5.0496,
      "mean_token_accuracy": 0.2000635713338852,
      "num_tokens": 38198972.0,
      "step": 18405
    },
    {
      "entropy": 5.53039960861206,
      "epoch": 1.6699927431059507,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047237999889833816,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2063676580786705,
      "num_tokens": 38209472.0,
      "step": 18410
    },
    {
      "entropy": 5.461790990829468,
      "epoch": 1.670446298984035,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047236446435390826,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2051655277609825,
      "num_tokens": 38219977.0,
      "step": 18415
    },
    {
      "entropy": 5.519774389266968,
      "epoch": 1.670899854862119,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047234892572793093,
      "loss": 5.0577,
      "mean_token_accuracy": 0.2033174842596054,
      "num_tokens": 38230492.0,
      "step": 18420
    },
    {
      "entropy": 5.529501247406006,
      "epoch": 1.671353410740203,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004723333830207275,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1979853630065918,
      "num_tokens": 38240799.0,
      "step": 18425
    },
    {
      "entropy": 5.518443870544433,
      "epoch": 1.6718069666182873,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047231783623261954,
      "loss": 4.9815,
      "mean_token_accuracy": 0.1997287839651108,
      "num_tokens": 38251439.0,
      "step": 18430
    },
    {
      "entropy": 5.596939945220948,
      "epoch": 1.6722605224963716,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004723022853639283,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19885122925043106,
      "num_tokens": 38261795.0,
      "step": 18435
    },
    {
      "entropy": 5.516361951828003,
      "epoch": 1.6727140783744558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004722867304149756,
      "loss": 4.9921,
      "mean_token_accuracy": 0.20650445967912673,
      "num_tokens": 38272172.0,
      "step": 18440
    },
    {
      "entropy": 5.552573108673096,
      "epoch": 1.67316763425254,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004722711713860831,
      "loss": 5.083,
      "mean_token_accuracy": 0.19582296311855316,
      "num_tokens": 38282829.0,
      "step": 18445
    },
    {
      "entropy": 5.580194520950317,
      "epoch": 1.673621190130624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004722556082775724,
      "loss": 5.0806,
      "mean_token_accuracy": 0.20674073845148086,
      "num_tokens": 38293793.0,
      "step": 18450
    },
    {
      "entropy": 5.5810974597930905,
      "epoch": 1.6740747460087082,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004722400410897655,
      "loss": 5.083,
      "mean_token_accuracy": 0.20138343721628188,
      "num_tokens": 38302577.0,
      "step": 18455
    },
    {
      "entropy": 5.53628568649292,
      "epoch": 1.6745283018867925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004722244698229842,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20656457394361497,
      "num_tokens": 38312571.0,
      "step": 18460
    },
    {
      "entropy": 5.527368402481079,
      "epoch": 1.6749818577648767,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047220889447755073,
      "loss": 4.966,
      "mean_token_accuracy": 0.2101530909538269,
      "num_tokens": 38323544.0,
      "step": 18465
    },
    {
      "entropy": 5.582181930541992,
      "epoch": 1.6754354136429608,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004721933150537869,
      "loss": 5.0528,
      "mean_token_accuracy": 0.20263455361127852,
      "num_tokens": 38334509.0,
      "step": 18470
    },
    {
      "entropy": 5.493896722793579,
      "epoch": 1.675888969521045,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047217773155201526,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20418312698602675,
      "num_tokens": 38344747.0,
      "step": 18475
    },
    {
      "entropy": 5.586251354217529,
      "epoch": 1.676342525399129,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047216214397255776,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1967146709561348,
      "num_tokens": 38355671.0,
      "step": 18480
    },
    {
      "entropy": 5.565404796600342,
      "epoch": 1.6767960812772134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047214655231573693,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20337298065423964,
      "num_tokens": 38366368.0,
      "step": 18485
    },
    {
      "entropy": 5.563398122787476,
      "epoch": 1.6772496371552976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004721309565818752,
      "loss": 5.1038,
      "mean_token_accuracy": 0.20038592368364333,
      "num_tokens": 38377322.0,
      "step": 18490
    },
    {
      "entropy": 5.608231687545777,
      "epoch": 1.6777031930333817,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047211535677129496,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20255047976970672,
      "num_tokens": 38389204.0,
      "step": 18495
    },
    {
      "entropy": 5.548760461807251,
      "epoch": 1.678156748911466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047209975288431893,
      "loss": 5.053,
      "mean_token_accuracy": 0.2019473984837532,
      "num_tokens": 38399117.0,
      "step": 18500
    },
    {
      "entropy": 5.537803888320923,
      "epoch": 1.67861030478955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047208414492126986,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2017965391278267,
      "num_tokens": 38410341.0,
      "step": 18505
    },
    {
      "entropy": 5.595143365859985,
      "epoch": 1.6790638606676342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047206853288247043,
      "loss": 5.0836,
      "mean_token_accuracy": 0.19759543389081954,
      "num_tokens": 38422742.0,
      "step": 18510
    },
    {
      "entropy": 5.56275897026062,
      "epoch": 1.6795174165457185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047205291676824345,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19663460105657576,
      "num_tokens": 38433022.0,
      "step": 18515
    },
    {
      "entropy": 5.593740558624267,
      "epoch": 1.6799709724238028,
      "grad_norm": 1.0,
      "learning_rate": 0.000472037296578912,
      "loss": 5.0561,
      "mean_token_accuracy": 0.20002095848321916,
      "num_tokens": 38444325.0,
      "step": 18520
    },
    {
      "entropy": 5.537397623062134,
      "epoch": 1.6804245283018868,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000472021672314799,
      "loss": 4.9606,
      "mean_token_accuracy": 0.20704559981822968,
      "num_tokens": 38454576.0,
      "step": 18525
    },
    {
      "entropy": 5.525679063796997,
      "epoch": 1.6808780841799709,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047200604397622763,
      "loss": 4.9752,
      "mean_token_accuracy": 0.20736278742551803,
      "num_tokens": 38465090.0,
      "step": 18530
    },
    {
      "entropy": 5.576778554916382,
      "epoch": 1.6813316400580551,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004719904115635211,
      "loss": 5.2162,
      "mean_token_accuracy": 0.1895391821861267,
      "num_tokens": 38477253.0,
      "step": 18535
    },
    {
      "entropy": 5.618865823745727,
      "epoch": 1.6817851959361394,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004719747750770026,
      "loss": 5.1115,
      "mean_token_accuracy": 0.19903451353311538,
      "num_tokens": 38488040.0,
      "step": 18540
    },
    {
      "entropy": 5.549099969863891,
      "epoch": 1.6822387518142237,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004719591345169956,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20792275220155715,
      "num_tokens": 38498016.0,
      "step": 18545
    },
    {
      "entropy": 5.586939668655395,
      "epoch": 1.6826923076923077,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047194348988382345,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20007387101650237,
      "num_tokens": 38508996.0,
      "step": 18550
    },
    {
      "entropy": 5.544355249404907,
      "epoch": 1.6831458635703918,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004719278411778097,
      "loss": 5.0735,
      "mean_token_accuracy": 0.20666104257106782,
      "num_tokens": 38519608.0,
      "step": 18555
    },
    {
      "entropy": 5.548975801467895,
      "epoch": 1.683599419448476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047191218839927817,
      "loss": 5.0226,
      "mean_token_accuracy": 0.19572516828775405,
      "num_tokens": 38530737.0,
      "step": 18560
    },
    {
      "entropy": 5.543293809890747,
      "epoch": 1.6840529753265603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004718965315485523,
      "loss": 4.9857,
      "mean_token_accuracy": 0.204212287068367,
      "num_tokens": 38540925.0,
      "step": 18565
    },
    {
      "entropy": 5.55231466293335,
      "epoch": 1.6845065312046446,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004718808706259561,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19747941792011262,
      "num_tokens": 38553176.0,
      "step": 18570
    },
    {
      "entropy": 5.599540758132934,
      "epoch": 1.6849600870827286,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047186520563181325,
      "loss": 5.1021,
      "mean_token_accuracy": 0.19440075755119324,
      "num_tokens": 38563394.0,
      "step": 18575
    },
    {
      "entropy": 5.613511562347412,
      "epoch": 1.6854136429608126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004718495365664477,
      "loss": 5.0467,
      "mean_token_accuracy": 0.20269798040390014,
      "num_tokens": 38573243.0,
      "step": 18580
    },
    {
      "entropy": 5.544748973846436,
      "epoch": 1.685867198838897,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047183386343018364,
      "loss": 5.0687,
      "mean_token_accuracy": 0.19938255101442337,
      "num_tokens": 38584014.0,
      "step": 18585
    },
    {
      "entropy": 5.572828197479248,
      "epoch": 1.6863207547169812,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004718181862233451,
      "loss": 4.996,
      "mean_token_accuracy": 0.20601304918527602,
      "num_tokens": 38594467.0,
      "step": 18590
    },
    {
      "entropy": 5.595120668411255,
      "epoch": 1.6867743105950654,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004718025049462563,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20386261492967606,
      "num_tokens": 38604407.0,
      "step": 18595
    },
    {
      "entropy": 5.581459188461304,
      "epoch": 1.6872278664731495,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004717868195992416,
      "loss": 5.0629,
      "mean_token_accuracy": 0.20675850361585618,
      "num_tokens": 38614508.0,
      "step": 18600
    },
    {
      "entropy": 5.6191541194915775,
      "epoch": 1.6876814223512335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004717711301826253,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20611653476953506,
      "num_tokens": 38626244.0,
      "step": 18605
    },
    {
      "entropy": 5.520521974563598,
      "epoch": 1.6881349782293178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047175543669673186,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20494984090328217,
      "num_tokens": 38636694.0,
      "step": 18610
    },
    {
      "entropy": 5.6410387516021725,
      "epoch": 1.688588534107402,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047173973914188587,
      "loss": 5.1134,
      "mean_token_accuracy": 0.20223284512758255,
      "num_tokens": 38647414.0,
      "step": 18615
    },
    {
      "entropy": 5.5597906589508055,
      "epoch": 1.6890420899854863,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047172403751841196,
      "loss": 5.0246,
      "mean_token_accuracy": 0.20177806913852692,
      "num_tokens": 38657425.0,
      "step": 18620
    },
    {
      "entropy": 5.58294153213501,
      "epoch": 1.6894956458635704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047170833182663485,
      "loss": 5.0429,
      "mean_token_accuracy": 0.2058634117245674,
      "num_tokens": 38667751.0,
      "step": 18625
    },
    {
      "entropy": 5.461794662475586,
      "epoch": 1.6899492017416544,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004716926220668793,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20780032575130464,
      "num_tokens": 38676918.0,
      "step": 18630
    },
    {
      "entropy": 5.486989974975586,
      "epoch": 1.6904027576197387,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047167690823947015,
      "loss": 5.025,
      "mean_token_accuracy": 0.21017662882804872,
      "num_tokens": 38685887.0,
      "step": 18635
    },
    {
      "entropy": 5.512678337097168,
      "epoch": 1.690856313497823,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004716611903447325,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20788081139326095,
      "num_tokens": 38696179.0,
      "step": 18640
    },
    {
      "entropy": 5.5653375625610355,
      "epoch": 1.6913098693759072,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004716454683829912,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20205404609441757,
      "num_tokens": 38705777.0,
      "step": 18645
    },
    {
      "entropy": 5.53246579170227,
      "epoch": 1.6917634252539913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047162974235457163,
      "loss": 5.023,
      "mean_token_accuracy": 0.20526780039072037,
      "num_tokens": 38717709.0,
      "step": 18650
    },
    {
      "entropy": 5.516905975341797,
      "epoch": 1.6922169811320755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004716140122597988,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20970158874988556,
      "num_tokens": 38729517.0,
      "step": 18655
    },
    {
      "entropy": 5.529263734817505,
      "epoch": 1.6926705370101596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004715982780989981,
      "loss": 5.032,
      "mean_token_accuracy": 0.2079072117805481,
      "num_tokens": 38740878.0,
      "step": 18660
    },
    {
      "entropy": 5.517541885375977,
      "epoch": 1.6931240928882438,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047158253987249487,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2089331105351448,
      "num_tokens": 38751291.0,
      "step": 18665
    },
    {
      "entropy": 5.509375333786011,
      "epoch": 1.693577648766328,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047156679758061465,
      "loss": 4.959,
      "mean_token_accuracy": 0.2138681709766388,
      "num_tokens": 38760940.0,
      "step": 18670
    },
    {
      "entropy": 5.549563932418823,
      "epoch": 1.6940312046444121,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047155105122368297,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21167542934417724,
      "num_tokens": 38770240.0,
      "step": 18675
    },
    {
      "entropy": 5.466352033615112,
      "epoch": 1.6944847605224964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004715353008020255,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20637205243110657,
      "num_tokens": 38780452.0,
      "step": 18680
    },
    {
      "entropy": 5.537557363510132,
      "epoch": 1.6949383164005805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047151954631596794,
      "loss": 5.0745,
      "mean_token_accuracy": 0.199966262280941,
      "num_tokens": 38790763.0,
      "step": 18685
    },
    {
      "entropy": 5.533513927459717,
      "epoch": 1.6953918722786647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047150378776583603,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21286385506391525,
      "num_tokens": 38802164.0,
      "step": 18690
    },
    {
      "entropy": 5.537321853637695,
      "epoch": 1.695845428156749,
      "grad_norm": 1.0,
      "learning_rate": 0.0004714880251519557,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20215222984552383,
      "num_tokens": 38812901.0,
      "step": 18695
    },
    {
      "entropy": 5.508861923217774,
      "epoch": 1.6962989840348333,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000471472258474653,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2096114531159401,
      "num_tokens": 38823305.0,
      "step": 18700
    },
    {
      "entropy": 5.539385175704956,
      "epoch": 1.6967525399129173,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047145648773425384,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20409904271364213,
      "num_tokens": 38833602.0,
      "step": 18705
    },
    {
      "entropy": 5.576952886581421,
      "epoch": 1.6972060957910013,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047144071293108456,
      "loss": 5.071,
      "mean_token_accuracy": 0.2042400449514389,
      "num_tokens": 38843950.0,
      "step": 18710
    },
    {
      "entropy": 5.52810320854187,
      "epoch": 1.6976596516690856,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004714249340654712,
      "loss": 4.9925,
      "mean_token_accuracy": 0.2058522433042526,
      "num_tokens": 38853753.0,
      "step": 18715
    },
    {
      "entropy": 5.467414999008179,
      "epoch": 1.6981132075471699,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004714091511377402,
      "loss": 4.9304,
      "mean_token_accuracy": 0.20456690341234207,
      "num_tokens": 38863394.0,
      "step": 18720
    },
    {
      "entropy": 5.521736574172974,
      "epoch": 1.6985667634252541,
      "grad_norm": 1.125,
      "learning_rate": 0.00047139336414821784,
      "loss": 5.024,
      "mean_token_accuracy": 0.20001382976770402,
      "num_tokens": 38872972.0,
      "step": 18725
    },
    {
      "entropy": 5.5153477668762205,
      "epoch": 1.6990203193033382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047137757309723073,
      "loss": 5.0092,
      "mean_token_accuracy": 0.212299245595932,
      "num_tokens": 38883251.0,
      "step": 18730
    },
    {
      "entropy": 5.516787052154541,
      "epoch": 1.6994738751814222,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004713617779851053,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20373117178678513,
      "num_tokens": 38893274.0,
      "step": 18735
    },
    {
      "entropy": 5.61175503730774,
      "epoch": 1.6999274310595065,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004713459788121683,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19592057019472123,
      "num_tokens": 38903870.0,
      "step": 18740
    },
    {
      "entropy": 5.572488260269165,
      "epoch": 1.7003809869375908,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004713301755787464,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1994834527373314,
      "num_tokens": 38914921.0,
      "step": 18745
    },
    {
      "entropy": 5.570324230194092,
      "epoch": 1.700834542815675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004713143682851665,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20570874065160752,
      "num_tokens": 38924630.0,
      "step": 18750
    },
    {
      "entropy": 5.477497339248657,
      "epoch": 1.701288098693759,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712985569317553,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20439738780260086,
      "num_tokens": 38934926.0,
      "step": 18755
    },
    {
      "entropy": 5.603098392486572,
      "epoch": 1.7017416545718431,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004712827415188401,
      "loss": 5.1481,
      "mean_token_accuracy": 0.19430297911167144,
      "num_tokens": 38944513.0,
      "step": 18760
    },
    {
      "entropy": 5.575858306884766,
      "epoch": 1.7021952104499274,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004712669220467476,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20788175910711287,
      "num_tokens": 38955918.0,
      "step": 18765
    },
    {
      "entropy": 5.568398094177246,
      "epoch": 1.7026487663280117,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047125109851580533,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20533353090286255,
      "num_tokens": 38966221.0,
      "step": 18770
    },
    {
      "entropy": 5.535524511337281,
      "epoch": 1.703102322206096,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047123527092634025,
      "loss": 4.9866,
      "mean_token_accuracy": 0.20195710361003877,
      "num_tokens": 38976463.0,
      "step": 18775
    },
    {
      "entropy": 5.479678440093994,
      "epoch": 1.70355587808418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004712194392786797,
      "loss": 4.9273,
      "mean_token_accuracy": 0.20722113698720931,
      "num_tokens": 38986375.0,
      "step": 18780
    },
    {
      "entropy": 5.625212955474853,
      "epoch": 1.704009433962264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047120360357315127,
      "loss": 5.064,
      "mean_token_accuracy": 0.20052340477705002,
      "num_tokens": 38996116.0,
      "step": 18785
    },
    {
      "entropy": 5.509733486175537,
      "epoch": 1.7044629898403483,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047118776381008216,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20391646027565002,
      "num_tokens": 39005846.0,
      "step": 18790
    },
    {
      "entropy": 5.573023700714112,
      "epoch": 1.7049165457184325,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047117191998980026,
      "loss": 5.0213,
      "mean_token_accuracy": 0.19692962765693664,
      "num_tokens": 39016744.0,
      "step": 18795
    },
    {
      "entropy": 5.5535383224487305,
      "epoch": 1.7053701015965168,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047115607211263295,
      "loss": 5.0736,
      "mean_token_accuracy": 0.19864329397678376,
      "num_tokens": 39026725.0,
      "step": 18800
    },
    {
      "entropy": 5.5790722370147705,
      "epoch": 1.7058236574746009,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047114022017890814,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20246905237436294,
      "num_tokens": 39036851.0,
      "step": 18805
    },
    {
      "entropy": 5.526825666427612,
      "epoch": 1.706277213352685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004711243641889536,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2043837919831276,
      "num_tokens": 39047179.0,
      "step": 18810
    },
    {
      "entropy": 5.49183235168457,
      "epoch": 1.7067307692307692,
      "grad_norm": 1.0,
      "learning_rate": 0.00047110850414309714,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2154756158590317,
      "num_tokens": 39057758.0,
      "step": 18815
    },
    {
      "entropy": 5.5515491485595705,
      "epoch": 1.7071843251088534,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047109264004166696,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20021470189094542,
      "num_tokens": 39068142.0,
      "step": 18820
    },
    {
      "entropy": 5.5086527347564695,
      "epoch": 1.7076378809869377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004710767718849909,
      "loss": 4.9844,
      "mean_token_accuracy": 0.19743564873933792,
      "num_tokens": 39078157.0,
      "step": 18825
    },
    {
      "entropy": 5.514392566680908,
      "epoch": 1.7080914368650217,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004710608996733973,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21087042093276978,
      "num_tokens": 39089090.0,
      "step": 18830
    },
    {
      "entropy": 5.553383779525757,
      "epoch": 1.708544992743106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047104502340721433,
      "loss": 5.0734,
      "mean_token_accuracy": 0.19710729867219925,
      "num_tokens": 39099079.0,
      "step": 18835
    },
    {
      "entropy": 5.547780752182007,
      "epoch": 1.70899854862119,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047102914308677033,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20297769010066985,
      "num_tokens": 39109927.0,
      "step": 18840
    },
    {
      "entropy": 5.581135034561157,
      "epoch": 1.7094521044992743,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047101325871239363,
      "loss": 5.0103,
      "mean_token_accuracy": 0.2026897057890892,
      "num_tokens": 39119806.0,
      "step": 18845
    },
    {
      "entropy": 5.597395563125611,
      "epoch": 1.7099056603773586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004709973702844128,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20226970463991165,
      "num_tokens": 39130147.0,
      "step": 18850
    },
    {
      "entropy": 5.476788091659546,
      "epoch": 1.7103592162554426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047098147780315645,
      "loss": 4.902,
      "mean_token_accuracy": 0.21620516031980513,
      "num_tokens": 39139937.0,
      "step": 18855
    },
    {
      "entropy": 5.548338937759399,
      "epoch": 1.710812772133527,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004709655812689532,
      "loss": 5.023,
      "mean_token_accuracy": 0.20639331191778182,
      "num_tokens": 39149977.0,
      "step": 18860
    },
    {
      "entropy": 5.53417010307312,
      "epoch": 1.711266328011611,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004709496806821318,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20468510836362838,
      "num_tokens": 39161076.0,
      "step": 18865
    },
    {
      "entropy": 5.57164454460144,
      "epoch": 1.7117198838896952,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047093377604302106,
      "loss": 5.0879,
      "mean_token_accuracy": 0.19266926050186156,
      "num_tokens": 39170776.0,
      "step": 18870
    },
    {
      "entropy": 5.578165769577026,
      "epoch": 1.7121734397677795,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004709178673519499,
      "loss": 5.0447,
      "mean_token_accuracy": 0.19984612464904786,
      "num_tokens": 39181134.0,
      "step": 18875
    },
    {
      "entropy": 5.5719358921051025,
      "epoch": 1.7126269956458637,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004709019546092473,
      "loss": 5.1256,
      "mean_token_accuracy": 0.19469009190797806,
      "num_tokens": 39193331.0,
      "step": 18880
    },
    {
      "entropy": 5.624257659912109,
      "epoch": 1.7130805515239478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047088603781524237,
      "loss": 5.0845,
      "mean_token_accuracy": 0.19632348269224167,
      "num_tokens": 39204372.0,
      "step": 18885
    },
    {
      "entropy": 5.576696538925171,
      "epoch": 1.7135341074020318,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004708701169702642,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2091236263513565,
      "num_tokens": 39214725.0,
      "step": 18890
    },
    {
      "entropy": 5.4728844165802,
      "epoch": 1.713987663280116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004708541920746422,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20199199467897416,
      "num_tokens": 39225119.0,
      "step": 18895
    },
    {
      "entropy": 5.503029870986938,
      "epoch": 1.7144412191582004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004708382631287055,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20356596112251282,
      "num_tokens": 39235021.0,
      "step": 18900
    },
    {
      "entropy": 5.5845558643341064,
      "epoch": 1.7148947750362846,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004708223301327837,
      "loss": 5.1454,
      "mean_token_accuracy": 0.19812427461147308,
      "num_tokens": 39245830.0,
      "step": 18905
    },
    {
      "entropy": 5.646688604354859,
      "epoch": 1.7153483309143687,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047080639308720617,
      "loss": 5.1165,
      "mean_token_accuracy": 0.2034643605351448,
      "num_tokens": 39255264.0,
      "step": 18910
    },
    {
      "entropy": 5.553671073913574,
      "epoch": 1.7158018867924527,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047079045199230254,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20753138065338134,
      "num_tokens": 39264966.0,
      "step": 18915
    },
    {
      "entropy": 5.5494787216186525,
      "epoch": 1.716255442670537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047077450684840247,
      "loss": 5.1004,
      "mean_token_accuracy": 0.1983710289001465,
      "num_tokens": 39275082.0,
      "step": 18920
    },
    {
      "entropy": 5.622508811950683,
      "epoch": 1.7167089985486212,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004707585576558357,
      "loss": 5.1766,
      "mean_token_accuracy": 0.19712687581777572,
      "num_tokens": 39285426.0,
      "step": 18925
    },
    {
      "entropy": 5.640629720687866,
      "epoch": 1.7171625544267055,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004707426044149321,
      "loss": 5.1012,
      "mean_token_accuracy": 0.20073680877685546,
      "num_tokens": 39296633.0,
      "step": 18930
    },
    {
      "entropy": 5.516549062728882,
      "epoch": 1.7176161103047896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047072664712602153,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21242664009332657,
      "num_tokens": 39306775.0,
      "step": 18935
    },
    {
      "entropy": 5.538043451309204,
      "epoch": 1.7180696661828736,
      "grad_norm": 1.0625,
      "learning_rate": 0.000470710685789434,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1989818185567856,
      "num_tokens": 39317917.0,
      "step": 18940
    },
    {
      "entropy": 5.572309398651123,
      "epoch": 1.7185232220609579,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047069472040549965,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20398209542036055,
      "num_tokens": 39328387.0,
      "step": 18945
    },
    {
      "entropy": 5.482699537277222,
      "epoch": 1.7189767779390421,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004706787509745486,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21320468336343765,
      "num_tokens": 39337639.0,
      "step": 18950
    },
    {
      "entropy": 5.552534103393555,
      "epoch": 1.7194303338171264,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047066277749691115,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20591485649347305,
      "num_tokens": 39347971.0,
      "step": 18955
    },
    {
      "entropy": 5.576763486862182,
      "epoch": 1.7198838896952104,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004706467999729175,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20291970521211625,
      "num_tokens": 39358038.0,
      "step": 18960
    },
    {
      "entropy": 5.5647767066955565,
      "epoch": 1.7203374455732945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004706308184028983,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20378267467021943,
      "num_tokens": 39367834.0,
      "step": 18965
    },
    {
      "entropy": 5.572733592987061,
      "epoch": 1.7207910014513788,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004706148327871839,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2063360780477524,
      "num_tokens": 39378924.0,
      "step": 18970
    },
    {
      "entropy": 5.635785675048828,
      "epoch": 1.721244557329463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047059884312610487,
      "loss": 5.1586,
      "mean_token_accuracy": 0.19298899471759795,
      "num_tokens": 39390082.0,
      "step": 18975
    },
    {
      "entropy": 5.568527746200561,
      "epoch": 1.7216981132075473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047058284941999185,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20253492444753646,
      "num_tokens": 39400485.0,
      "step": 18980
    },
    {
      "entropy": 5.5739670753479,
      "epoch": 1.7221516690856313,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004705668516691758,
      "loss": 5.0413,
      "mean_token_accuracy": 0.20526162534952164,
      "num_tokens": 39411601.0,
      "step": 18985
    },
    {
      "entropy": 5.5383148193359375,
      "epoch": 1.7226052249637154,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047055084987398735,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20863622277975083,
      "num_tokens": 39421625.0,
      "step": 18990
    },
    {
      "entropy": 5.5457957744598385,
      "epoch": 1.7230587808417996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004705348440347575,
      "loss": 5.0186,
      "mean_token_accuracy": 0.19834876507520677,
      "num_tokens": 39432033.0,
      "step": 18995
    },
    {
      "entropy": 5.519546318054199,
      "epoch": 1.723512336719884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047051883415181724,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20518908947706221,
      "num_tokens": 39442514.0,
      "step": 19000
    },
    {
      "entropy": 5.529253387451172,
      "epoch": 1.7239658925979682,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004705028202254977,
      "loss": 4.897,
      "mean_token_accuracy": 0.21023046970367432,
      "num_tokens": 39452497.0,
      "step": 19005
    },
    {
      "entropy": 5.531752395629883,
      "epoch": 1.7244194484760522,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047048680225612996,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20395348966121674,
      "num_tokens": 39462173.0,
      "step": 19010
    },
    {
      "entropy": 5.590306568145752,
      "epoch": 1.7248730043541363,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004704707802440454,
      "loss": 5.0729,
      "mean_token_accuracy": 0.20942844450473785,
      "num_tokens": 39473123.0,
      "step": 19015
    },
    {
      "entropy": 5.588532257080078,
      "epoch": 1.7253265602322205,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047045475418957524,
      "loss": 5.0522,
      "mean_token_accuracy": 0.19402004182338714,
      "num_tokens": 39484142.0,
      "step": 19020
    },
    {
      "entropy": 5.568538188934326,
      "epoch": 1.7257801161103048,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004704387240930509,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19874946773052216,
      "num_tokens": 39494926.0,
      "step": 19025
    },
    {
      "entropy": 5.536000108718872,
      "epoch": 1.726233671988389,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004704226899548041,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20445523411035538,
      "num_tokens": 39505153.0,
      "step": 19030
    },
    {
      "entropy": 5.509835290908813,
      "epoch": 1.726687227866473,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047040665177516616,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20506981313228606,
      "num_tokens": 39515905.0,
      "step": 19035
    },
    {
      "entropy": 5.510123252868652,
      "epoch": 1.7271407837445574,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047039060955446886,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20858711749315262,
      "num_tokens": 39526762.0,
      "step": 19040
    },
    {
      "entropy": 5.503658151626587,
      "epoch": 1.7275943396226414,
      "grad_norm": 1.078125,
      "learning_rate": 0.000470374563293044,
      "loss": 4.988,
      "mean_token_accuracy": 0.20179152339696885,
      "num_tokens": 39537340.0,
      "step": 19045
    },
    {
      "entropy": 5.5714703559875485,
      "epoch": 1.7280478955007257,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004703585129912234,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20270833969116211,
      "num_tokens": 39547144.0,
      "step": 19050
    },
    {
      "entropy": 5.653240966796875,
      "epoch": 1.72850145137881,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047034245864933894,
      "loss": 5.1215,
      "mean_token_accuracy": 0.19415885508060454,
      "num_tokens": 39557707.0,
      "step": 19055
    },
    {
      "entropy": 5.619681930541992,
      "epoch": 1.728955007256894,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047032640026772274,
      "loss": 5.1011,
      "mean_token_accuracy": 0.2001511499285698,
      "num_tokens": 39567629.0,
      "step": 19060
    },
    {
      "entropy": 5.5302807807922365,
      "epoch": 1.7294085631349783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004703103378467067,
      "loss": 5.0007,
      "mean_token_accuracy": 0.20744604766368865,
      "num_tokens": 39577918.0,
      "step": 19065
    },
    {
      "entropy": 5.571910429000854,
      "epoch": 1.7298621190130623,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047029427138662314,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21134694665670395,
      "num_tokens": 39587838.0,
      "step": 19070
    },
    {
      "entropy": 5.525185012817383,
      "epoch": 1.7303156748911466,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004702782008878042,
      "loss": 4.9286,
      "mean_token_accuracy": 0.20921002179384232,
      "num_tokens": 39597353.0,
      "step": 19075
    },
    {
      "entropy": 5.522932624816894,
      "epoch": 1.7307692307692308,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004702621263505824,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2087055891752243,
      "num_tokens": 39607349.0,
      "step": 19080
    },
    {
      "entropy": 5.562011480331421,
      "epoch": 1.731222786647315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004702460477752901,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20134317576885224,
      "num_tokens": 39617810.0,
      "step": 19085
    },
    {
      "entropy": 5.5258690357208256,
      "epoch": 1.7316763425253991,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004702299651622597,
      "loss": 4.942,
      "mean_token_accuracy": 0.21373670399188996,
      "num_tokens": 39626805.0,
      "step": 19090
    },
    {
      "entropy": 5.464047527313232,
      "epoch": 1.7321298984034832,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004702138785118239,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21195108890533448,
      "num_tokens": 39637839.0,
      "step": 19095
    },
    {
      "entropy": 5.493241214752198,
      "epoch": 1.7325834542815675,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004701977878243153,
      "loss": 4.9499,
      "mean_token_accuracy": 0.20881755650043488,
      "num_tokens": 39648697.0,
      "step": 19100
    },
    {
      "entropy": 5.490741825103759,
      "epoch": 1.7330370101596517,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004701816931000668,
      "loss": 4.991,
      "mean_token_accuracy": 0.20672685503959656,
      "num_tokens": 39658154.0,
      "step": 19105
    },
    {
      "entropy": 5.542169046401978,
      "epoch": 1.733490566037736,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047016559433941106,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2038709044456482,
      "num_tokens": 39667939.0,
      "step": 19110
    },
    {
      "entropy": 5.519157743453979,
      "epoch": 1.73394412191582,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004701494915426812,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20618734061717986,
      "num_tokens": 39678289.0,
      "step": 19115
    },
    {
      "entropy": 5.559554719924927,
      "epoch": 1.734397677793904,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047013338471021,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2046155884861946,
      "num_tokens": 39689901.0,
      "step": 19120
    },
    {
      "entropy": 5.5800446510314945,
      "epoch": 1.7348512336719883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047011727384233074,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20256204903125763,
      "num_tokens": 39700831.0,
      "step": 19125
    },
    {
      "entropy": 5.562351942062378,
      "epoch": 1.7353047895500726,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004701011589393765,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20052081495523452,
      "num_tokens": 39711794.0,
      "step": 19130
    },
    {
      "entropy": 5.521185445785522,
      "epoch": 1.7357583454281569,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047008504000168054,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2035038337111473,
      "num_tokens": 39721889.0,
      "step": 19135
    },
    {
      "entropy": 5.497358751296997,
      "epoch": 1.736211901306241,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004700689170295763,
      "loss": 5.074,
      "mean_token_accuracy": 0.20340992957353593,
      "num_tokens": 39732562.0,
      "step": 19140
    },
    {
      "entropy": 5.610089254379273,
      "epoch": 1.736665457184325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047005279002339717,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20637455582618713,
      "num_tokens": 39742286.0,
      "step": 19145
    },
    {
      "entropy": 5.5739748001098635,
      "epoch": 1.7371190130624092,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047003665898347663,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20306781083345413,
      "num_tokens": 39754336.0,
      "step": 19150
    },
    {
      "entropy": 5.496878910064697,
      "epoch": 1.7375725689404935,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004700205239101483,
      "loss": 4.936,
      "mean_token_accuracy": 0.21169695258140564,
      "num_tokens": 39763545.0,
      "step": 19155
    },
    {
      "entropy": 5.546338224411011,
      "epoch": 1.7380261248185778,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004700043848037458,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19563563019037247,
      "num_tokens": 39774536.0,
      "step": 19160
    },
    {
      "entropy": 5.5876946449279785,
      "epoch": 1.7384796806966618,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004699882416646029,
      "loss": 4.9862,
      "mean_token_accuracy": 0.20746959298849105,
      "num_tokens": 39785377.0,
      "step": 19165
    },
    {
      "entropy": 5.5691807746887205,
      "epoch": 1.7389332365747459,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004699720944930535,
      "loss": 5.106,
      "mean_token_accuracy": 0.19034550935029984,
      "num_tokens": 39796692.0,
      "step": 19170
    },
    {
      "entropy": 5.473299121856689,
      "epoch": 1.7393867924528301,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046995594328943147,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2059233844280243,
      "num_tokens": 39807927.0,
      "step": 19175
    },
    {
      "entropy": 5.669158411026001,
      "epoch": 1.7398403483309144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004699397880540709,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20216887295246125,
      "num_tokens": 39817617.0,
      "step": 19180
    },
    {
      "entropy": 5.580893421173096,
      "epoch": 1.7402939042089987,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004699236287873058,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21136983335018159,
      "num_tokens": 39827092.0,
      "step": 19185
    },
    {
      "entropy": 5.529440689086914,
      "epoch": 1.7407474600870827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004699074654894705,
      "loss": 4.9738,
      "mean_token_accuracy": 0.20527196526527405,
      "num_tokens": 39837378.0,
      "step": 19190
    },
    {
      "entropy": 5.53071403503418,
      "epoch": 1.7412010159651667,
      "grad_norm": 1.03125,
      "learning_rate": 0.000469891298160899,
      "loss": 5.0009,
      "mean_token_accuracy": 0.19492065608501435,
      "num_tokens": 39848191.0,
      "step": 19195
    },
    {
      "entropy": 5.512627172470093,
      "epoch": 1.741654571843251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046987512680192586,
      "loss": 4.9916,
      "mean_token_accuracy": 0.204746875166893,
      "num_tokens": 39857886.0,
      "step": 19200
    },
    {
      "entropy": 5.600387763977051,
      "epoch": 1.7421081277213353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046985895141288546,
      "loss": 5.0785,
      "mean_token_accuracy": 0.1978824257850647,
      "num_tokens": 39868348.0,
      "step": 19205
    },
    {
      "entropy": 5.620454978942871,
      "epoch": 1.7425616835994195,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046984277199411224,
      "loss": 5.0972,
      "mean_token_accuracy": 0.20216984152793885,
      "num_tokens": 39879129.0,
      "step": 19210
    },
    {
      "entropy": 5.51452603340149,
      "epoch": 1.7430152394775036,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046982658854594083,
      "loss": 4.984,
      "mean_token_accuracy": 0.20440545082092285,
      "num_tokens": 39890239.0,
      "step": 19215
    },
    {
      "entropy": 5.4759642601013185,
      "epoch": 1.7434687953555879,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046981040106870595,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2083138108253479,
      "num_tokens": 39900907.0,
      "step": 19220
    },
    {
      "entropy": 5.566823959350586,
      "epoch": 1.743922351233672,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004697942095627423,
      "loss": 5.0202,
      "mean_token_accuracy": 0.20522894561290742,
      "num_tokens": 39911564.0,
      "step": 19225
    },
    {
      "entropy": 5.534403371810913,
      "epoch": 1.7443759071117562,
      "grad_norm": 1.125,
      "learning_rate": 0.0004697780140283848,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2044839709997177,
      "num_tokens": 39922168.0,
      "step": 19230
    },
    {
      "entropy": 5.536777639389038,
      "epoch": 1.7448294629898404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004697618144659683,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20845214426517486,
      "num_tokens": 39932690.0,
      "step": 19235
    },
    {
      "entropy": 5.645267963409424,
      "epoch": 1.7452830188679245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004697456108758278,
      "loss": 5.149,
      "mean_token_accuracy": 0.20028664767742158,
      "num_tokens": 39942970.0,
      "step": 19240
    },
    {
      "entropy": 5.525376605987549,
      "epoch": 1.7457365747460087,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004697294032582986,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20895958989858626,
      "num_tokens": 39953255.0,
      "step": 19245
    },
    {
      "entropy": 5.507604694366455,
      "epoch": 1.7461901306240928,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046971319161371564,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21043440699577332,
      "num_tokens": 39963521.0,
      "step": 19250
    },
    {
      "entropy": 5.5434962749481205,
      "epoch": 1.746643686502177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046969697594241426,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20170967131853104,
      "num_tokens": 39973494.0,
      "step": 19255
    },
    {
      "entropy": 5.534239292144775,
      "epoch": 1.7470972423802613,
      "grad_norm": 1.0,
      "learning_rate": 0.0004696807562447298,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19806908071041107,
      "num_tokens": 39984652.0,
      "step": 19260
    },
    {
      "entropy": 5.531661176681519,
      "epoch": 1.7475507982583456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046966453252099763,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20599662363529206,
      "num_tokens": 39995490.0,
      "step": 19265
    },
    {
      "entropy": 5.502593040466309,
      "epoch": 1.7480043541364296,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046964830477155335,
      "loss": 4.8317,
      "mean_token_accuracy": 0.22880762964487075,
      "num_tokens": 40006447.0,
      "step": 19270
    },
    {
      "entropy": 5.607112073898316,
      "epoch": 1.7484579100145137,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046963207299673257,
      "loss": 5.0919,
      "mean_token_accuracy": 0.20307234823703765,
      "num_tokens": 40015733.0,
      "step": 19275
    },
    {
      "entropy": 5.598391199111939,
      "epoch": 1.748911465892598,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046961583719687085,
      "loss": 4.9486,
      "mean_token_accuracy": 0.2105601742863655,
      "num_tokens": 40025844.0,
      "step": 19280
    },
    {
      "entropy": 5.516723155975342,
      "epoch": 1.7493650217706822,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046959959737230414,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21180037707090377,
      "num_tokens": 40035942.0,
      "step": 19285
    },
    {
      "entropy": 5.472798824310303,
      "epoch": 1.7498185776487665,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046958335352336806,
      "loss": 4.8877,
      "mean_token_accuracy": 0.21567040979862212,
      "num_tokens": 40045551.0,
      "step": 19290
    },
    {
      "entropy": 5.558046770095825,
      "epoch": 1.7502721335268505,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004695671056503987,
      "loss": 5.0308,
      "mean_token_accuracy": 0.2100541278719902,
      "num_tokens": 40055803.0,
      "step": 19295
    },
    {
      "entropy": 5.535646677017212,
      "epoch": 1.7507256894049346,
      "grad_norm": 1.03125,
      "learning_rate": 0.000469550853753732,
      "loss": 5.0046,
      "mean_token_accuracy": 0.2040293887257576,
      "num_tokens": 40066004.0,
      "step": 19300
    },
    {
      "entropy": 5.547567701339721,
      "epoch": 1.7511792452830188,
      "grad_norm": 1.09375,
      "learning_rate": 0.000469534597833704,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2093586027622223,
      "num_tokens": 40075522.0,
      "step": 19305
    },
    {
      "entropy": 5.5241851806640625,
      "epoch": 1.751632801161103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000469518337890651,
      "loss": 5.0075,
      "mean_token_accuracy": 0.2129354938864708,
      "num_tokens": 40084839.0,
      "step": 19310
    },
    {
      "entropy": 5.571526432037354,
      "epoch": 1.7520863570391874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046950207392490917,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20345304161310196,
      "num_tokens": 40094392.0,
      "step": 19315
    },
    {
      "entropy": 5.500971698760987,
      "epoch": 1.7525399129172714,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046948580593681496,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20845315605401993,
      "num_tokens": 40105990.0,
      "step": 19320
    },
    {
      "entropy": 5.486760139465332,
      "epoch": 1.7529934687953554,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004694695339267047,
      "loss": 4.9008,
      "mean_token_accuracy": 0.21510082483291626,
      "num_tokens": 40116143.0,
      "step": 19325
    },
    {
      "entropy": 5.655039691925049,
      "epoch": 1.7534470246734397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046945325789491483,
      "loss": 5.089,
      "mean_token_accuracy": 0.20208169966936113,
      "num_tokens": 40126849.0,
      "step": 19330
    },
    {
      "entropy": 5.551968955993653,
      "epoch": 1.753900580551524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004694369778417822,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20378676950931549,
      "num_tokens": 40137259.0,
      "step": 19335
    },
    {
      "entropy": 5.43944149017334,
      "epoch": 1.7543541364296082,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004694206937676431,
      "loss": 4.9077,
      "mean_token_accuracy": 0.20813882648944854,
      "num_tokens": 40148578.0,
      "step": 19340
    },
    {
      "entropy": 5.605654096603393,
      "epoch": 1.7548076923076923,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004694044056728347,
      "loss": 5.0896,
      "mean_token_accuracy": 0.1949380248785019,
      "num_tokens": 40159754.0,
      "step": 19345
    },
    {
      "entropy": 5.568848085403443,
      "epoch": 1.7552612481857763,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046938811355769363,
      "loss": 5.0174,
      "mean_token_accuracy": 0.20421231389045716,
      "num_tokens": 40169388.0,
      "step": 19350
    },
    {
      "entropy": 5.539310693740845,
      "epoch": 1.7557148040638606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046937181742255673,
      "loss": 5.0168,
      "mean_token_accuracy": 0.2036700502038002,
      "num_tokens": 40179271.0,
      "step": 19355
    },
    {
      "entropy": 5.590096855163575,
      "epoch": 1.7561683599419449,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004693555172677612,
      "loss": 5.0602,
      "mean_token_accuracy": 0.20551477372646332,
      "num_tokens": 40189292.0,
      "step": 19360
    },
    {
      "entropy": 5.516025447845459,
      "epoch": 1.7566219158200291,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000469339213093644,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20286233127117156,
      "num_tokens": 40199271.0,
      "step": 19365
    },
    {
      "entropy": 5.531123161315918,
      "epoch": 1.7570754716981132,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046932290490054235,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20643097013235093,
      "num_tokens": 40210386.0,
      "step": 19370
    },
    {
      "entropy": 5.578841209411621,
      "epoch": 1.7575290275761972,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004693065926887936,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20037015676498413,
      "num_tokens": 40220892.0,
      "step": 19375
    },
    {
      "entropy": 5.576885557174682,
      "epoch": 1.7579825834542815,
      "grad_norm": 1.0,
      "learning_rate": 0.00046929027645873484,
      "loss": 4.9987,
      "mean_token_accuracy": 0.20828428864479065,
      "num_tokens": 40231737.0,
      "step": 19380
    },
    {
      "entropy": 5.5091084957122805,
      "epoch": 1.7584361393323658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004692739562107038,
      "loss": 5.042,
      "mean_token_accuracy": 0.1980927422642708,
      "num_tokens": 40242218.0,
      "step": 19385
    },
    {
      "entropy": 5.525993061065674,
      "epoch": 1.75888969521045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046925763194503775,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20683387964963912,
      "num_tokens": 40253659.0,
      "step": 19390
    },
    {
      "entropy": 5.507544612884521,
      "epoch": 1.759343251088534,
      "grad_norm": 1.0,
      "learning_rate": 0.0004692413036620744,
      "loss": 4.8609,
      "mean_token_accuracy": 0.21369356215000151,
      "num_tokens": 40264328.0,
      "step": 19395
    },
    {
      "entropy": 5.502504682540893,
      "epoch": 1.7597968069666183,
      "grad_norm": 1.0,
      "learning_rate": 0.0004692249713621514,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2095779672265053,
      "num_tokens": 40275188.0,
      "step": 19400
    },
    {
      "entropy": 5.608981037139893,
      "epoch": 1.7602503628447024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046920863504560647,
      "loss": 5.1346,
      "mean_token_accuracy": 0.1956270754337311,
      "num_tokens": 40287094.0,
      "step": 19405
    },
    {
      "entropy": 5.5839334487915036,
      "epoch": 1.7607039187227866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004691922947127775,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20392605662345886,
      "num_tokens": 40296551.0,
      "step": 19410
    },
    {
      "entropy": 5.523501968383789,
      "epoch": 1.761157474600871,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004691759503640024,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20223620533943176,
      "num_tokens": 40306920.0,
      "step": 19415
    },
    {
      "entropy": 5.511826515197754,
      "epoch": 1.761611030478955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004691596019996191,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21371311396360398,
      "num_tokens": 40316797.0,
      "step": 19420
    },
    {
      "entropy": 5.458982658386231,
      "epoch": 1.7620645863570392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004691432496199658,
      "loss": 4.951,
      "mean_token_accuracy": 0.20571546405553817,
      "num_tokens": 40327002.0,
      "step": 19425
    },
    {
      "entropy": 5.651781749725342,
      "epoch": 1.7625181422351233,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046912689322538064,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20137424170970916,
      "num_tokens": 40337247.0,
      "step": 19430
    },
    {
      "entropy": 5.558581399917602,
      "epoch": 1.7629716981132075,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046911053281620187,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20686520040035247,
      "num_tokens": 40346960.0,
      "step": 19435
    },
    {
      "entropy": 5.588541746139526,
      "epoch": 1.7634252539912918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046909416839276775,
      "loss": 5.0532,
      "mean_token_accuracy": 0.2004593625664711,
      "num_tokens": 40357169.0,
      "step": 19440
    },
    {
      "entropy": 5.501184940338135,
      "epoch": 1.763878809869376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004690777999554168,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2171902596950531,
      "num_tokens": 40367660.0,
      "step": 19445
    },
    {
      "entropy": 5.5190881252288815,
      "epoch": 1.76433236574746,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004690614275044875,
      "loss": 4.914,
      "mean_token_accuracy": 0.20759038627147675,
      "num_tokens": 40378217.0,
      "step": 19450
    },
    {
      "entropy": 5.514788818359375,
      "epoch": 1.7647859216255442,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004690450510403184,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20546943843364715,
      "num_tokens": 40388249.0,
      "step": 19455
    },
    {
      "entropy": 5.5890326499938965,
      "epoch": 1.7652394775036284,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004690286705632483,
      "loss": 5.064,
      "mean_token_accuracy": 0.1979004979133606,
      "num_tokens": 40398239.0,
      "step": 19460
    },
    {
      "entropy": 5.568636322021485,
      "epoch": 1.7656930333817127,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004690122860736157,
      "loss": 5.0626,
      "mean_token_accuracy": 0.200815749168396,
      "num_tokens": 40408473.0,
      "step": 19465
    },
    {
      "entropy": 5.640263843536377,
      "epoch": 1.766146589259797,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046899589757175964,
      "loss": 5.1298,
      "mean_token_accuracy": 0.19818018674850463,
      "num_tokens": 40418674.0,
      "step": 19470
    },
    {
      "entropy": 5.501596450805664,
      "epoch": 1.766600145137881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000468979505058019,
      "loss": 4.97,
      "mean_token_accuracy": 0.20711494535207747,
      "num_tokens": 40429815.0,
      "step": 19475
    },
    {
      "entropy": 5.587862634658814,
      "epoch": 1.767053701015965,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046896310853273274,
      "loss": 5.0866,
      "mean_token_accuracy": 0.20570432543754577,
      "num_tokens": 40439809.0,
      "step": 19480
    },
    {
      "entropy": 5.6429685115814205,
      "epoch": 1.7675072568940493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046894670799624,
      "loss": 5.0691,
      "mean_token_accuracy": 0.20091277956962586,
      "num_tokens": 40449503.0,
      "step": 19485
    },
    {
      "entropy": 5.549521589279175,
      "epoch": 1.7679608127721336,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046893030344887995,
      "loss": 5.0881,
      "mean_token_accuracy": 0.1946729063987732,
      "num_tokens": 40459794.0,
      "step": 19490
    },
    {
      "entropy": 5.5187304496765135,
      "epoch": 1.7684143686502178,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046891389489099166,
      "loss": 4.948,
      "mean_token_accuracy": 0.2143867790699005,
      "num_tokens": 40470044.0,
      "step": 19495
    },
    {
      "entropy": 5.54579644203186,
      "epoch": 1.7688679245283019,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046889748232291476,
      "loss": 5.002,
      "mean_token_accuracy": 0.2024047553539276,
      "num_tokens": 40480709.0,
      "step": 19500
    },
    {
      "entropy": 5.475869750976562,
      "epoch": 1.769321480406386,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046888106574498843,
      "loss": 4.843,
      "mean_token_accuracy": 0.21536738574504852,
      "num_tokens": 40490906.0,
      "step": 19505
    },
    {
      "entropy": 5.504793357849121,
      "epoch": 1.7697750362844702,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004688646451575522,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2155078321695328,
      "num_tokens": 40501971.0,
      "step": 19510
    },
    {
      "entropy": 5.525714445114136,
      "epoch": 1.7702285921625545,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046884822056094587,
      "loss": 5.0446,
      "mean_token_accuracy": 0.1978493884205818,
      "num_tokens": 40512685.0,
      "step": 19515
    },
    {
      "entropy": 5.506925630569458,
      "epoch": 1.7706821480406387,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046883179195550883,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20833256095647812,
      "num_tokens": 40523460.0,
      "step": 19520
    },
    {
      "entropy": 5.576283740997314,
      "epoch": 1.7711357039187228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046881535934158094,
      "loss": 4.9669,
      "mean_token_accuracy": 0.20880083590745926,
      "num_tokens": 40533594.0,
      "step": 19525
    },
    {
      "entropy": 5.504837560653686,
      "epoch": 1.7715892597968068,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004687989227195021,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20841706395149232,
      "num_tokens": 40543504.0,
      "step": 19530
    },
    {
      "entropy": 5.526325607299805,
      "epoch": 1.772042815674891,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046878248208961204,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2034996911883354,
      "num_tokens": 40553711.0,
      "step": 19535
    },
    {
      "entropy": 5.5185425758361815,
      "epoch": 1.7724963715529753,
      "grad_norm": 1.09375,
      "learning_rate": 0.000468766037452251,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20368110984563828,
      "num_tokens": 40563731.0,
      "step": 19540
    },
    {
      "entropy": 5.586743545532227,
      "epoch": 1.7729499274310596,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046874958880775886,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20407156348228456,
      "num_tokens": 40574235.0,
      "step": 19545
    },
    {
      "entropy": 5.57427830696106,
      "epoch": 1.7734034833091437,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004687331361564758,
      "loss": 4.969,
      "mean_token_accuracy": 0.20999808609485626,
      "num_tokens": 40584459.0,
      "step": 19550
    },
    {
      "entropy": 5.531437730789184,
      "epoch": 1.7738570391872277,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004687166794987422,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20473010540008546,
      "num_tokens": 40594268.0,
      "step": 19555
    },
    {
      "entropy": 5.495528221130371,
      "epoch": 1.774310595065312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004687002188348983,
      "loss": 4.9667,
      "mean_token_accuracy": 0.2113693654537201,
      "num_tokens": 40604673.0,
      "step": 19560
    },
    {
      "entropy": 5.454802703857422,
      "epoch": 1.7747641509433962,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046868375416528443,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2076304167509079,
      "num_tokens": 40614998.0,
      "step": 19565
    },
    {
      "entropy": 5.547639560699463,
      "epoch": 1.7752177068214805,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046866728549024123,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20194907188415528,
      "num_tokens": 40624729.0,
      "step": 19570
    },
    {
      "entropy": 5.563374614715576,
      "epoch": 1.7756712626995645,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004686508128101092,
      "loss": 5.0957,
      "mean_token_accuracy": 0.20164481103420256,
      "num_tokens": 40635673.0,
      "step": 19575
    },
    {
      "entropy": 5.579333925247193,
      "epoch": 1.7761248185776488,
      "grad_norm": 0.984375,
      "learning_rate": 0.000468634336125229,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20136806219816208,
      "num_tokens": 40645990.0,
      "step": 19580
    },
    {
      "entropy": 5.5818102836608885,
      "epoch": 1.7765783744557329,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046861785543594143,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21454309225082396,
      "num_tokens": 40656339.0,
      "step": 19585
    },
    {
      "entropy": 5.520548629760742,
      "epoch": 1.7770319303338171,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004686013707425873,
      "loss": 4.9276,
      "mean_token_accuracy": 0.20953194946050643,
      "num_tokens": 40666371.0,
      "step": 19590
    },
    {
      "entropy": 5.559475946426391,
      "epoch": 1.7774854862119014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046858488204550735,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20916937738657,
      "num_tokens": 40676423.0,
      "step": 19595
    },
    {
      "entropy": 5.511946535110473,
      "epoch": 1.7779390420899854,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004685683893450428,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21064889430999756,
      "num_tokens": 40686331.0,
      "step": 19600
    },
    {
      "entropy": 5.496103239059448,
      "epoch": 1.7783925979680697,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004685518926415346,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21334135830402373,
      "num_tokens": 40697408.0,
      "step": 19605
    },
    {
      "entropy": 5.439297533035278,
      "epoch": 1.7788461538461537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000468535391935324,
      "loss": 4.8678,
      "mean_token_accuracy": 0.21968182027339936,
      "num_tokens": 40707078.0,
      "step": 19610
    },
    {
      "entropy": 5.552730417251587,
      "epoch": 1.779299709724238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685188872267521,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2048732966184616,
      "num_tokens": 40718116.0,
      "step": 19615
    },
    {
      "entropy": 5.568325138092041,
      "epoch": 1.7797532656023223,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685023785161603,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20626704692840575,
      "num_tokens": 40727737.0,
      "step": 19620
    },
    {
      "entropy": 5.5643861293792725,
      "epoch": 1.7802068214804065,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004684858658038901,
      "loss": 4.9754,
      "mean_token_accuracy": 0.20463858544826508,
      "num_tokens": 40737455.0,
      "step": 19625
    },
    {
      "entropy": 5.59386043548584,
      "epoch": 1.7806603773584906,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004684693490902828,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2064111277461052,
      "num_tokens": 40746715.0,
      "step": 19630
    },
    {
      "entropy": 5.49885196685791,
      "epoch": 1.7811139332365746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046845282837568,
      "loss": 4.8931,
      "mean_token_accuracy": 0.21309821605682372,
      "num_tokens": 40756432.0,
      "step": 19635
    },
    {
      "entropy": 5.446016025543213,
      "epoch": 1.781567489114659,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046843630366042355,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20433081090450286,
      "num_tokens": 40766629.0,
      "step": 19640
    },
    {
      "entropy": 5.593451738357544,
      "epoch": 1.7820210449927432,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046841977494485506,
      "loss": 5.0983,
      "mean_token_accuracy": 0.19296467453241348,
      "num_tokens": 40776650.0,
      "step": 19645
    },
    {
      "entropy": 5.60643949508667,
      "epoch": 1.7824746008708274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004684032422293162,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21133604347705842,
      "num_tokens": 40786722.0,
      "step": 19650
    },
    {
      "entropy": 5.441478490829468,
      "epoch": 1.7829281567489115,
      "grad_norm": 1.015625,
      "learning_rate": 0.000468386705514149,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22297567427158355,
      "num_tokens": 40797058.0,
      "step": 19655
    },
    {
      "entropy": 5.481105852127075,
      "epoch": 1.7833817126269955,
      "grad_norm": 1.125,
      "learning_rate": 0.0004683701647996956,
      "loss": 5.041,
      "mean_token_accuracy": 0.20970151722431182,
      "num_tokens": 40808207.0,
      "step": 19660
    },
    {
      "entropy": 5.586927700042724,
      "epoch": 1.7838352685050798,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004683536200862978,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20260453671216966,
      "num_tokens": 40819655.0,
      "step": 19665
    },
    {
      "entropy": 5.59092755317688,
      "epoch": 1.784288824383164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004683370713742979,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20382409691810607,
      "num_tokens": 40831015.0,
      "step": 19670
    },
    {
      "entropy": 5.512878561019898,
      "epoch": 1.7847423802612483,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046832051866403814,
      "loss": 4.955,
      "mean_token_accuracy": 0.20847117751836777,
      "num_tokens": 40841763.0,
      "step": 19675
    },
    {
      "entropy": 5.506123781204224,
      "epoch": 1.7851959361393324,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004683039619558607,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2064852684736252,
      "num_tokens": 40852479.0,
      "step": 19680
    },
    {
      "entropy": 5.579826307296753,
      "epoch": 1.7856494920174164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046828740125010813,
      "loss": 5.016,
      "mean_token_accuracy": 0.20882578641176225,
      "num_tokens": 40862486.0,
      "step": 19685
    },
    {
      "entropy": 5.611700773239136,
      "epoch": 1.7861030478955007,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004682708365471228,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20229277312755584,
      "num_tokens": 40872182.0,
      "step": 19690
    },
    {
      "entropy": 5.558951568603516,
      "epoch": 1.786556603773585,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046825426784724734,
      "loss": 5.0676,
      "mean_token_accuracy": 0.20649306178092958,
      "num_tokens": 40883036.0,
      "step": 19695
    },
    {
      "entropy": 5.594229221343994,
      "epoch": 1.7870101596516692,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004682376951508244,
      "loss": 5.1016,
      "mean_token_accuracy": 0.1905641257762909,
      "num_tokens": 40893684.0,
      "step": 19700
    },
    {
      "entropy": 5.629052925109863,
      "epoch": 1.7874637155297532,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046822111845819665,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20329589396715164,
      "num_tokens": 40904540.0,
      "step": 19705
    },
    {
      "entropy": 5.60902042388916,
      "epoch": 1.7879172714078373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046820453776970697,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20841806381940842,
      "num_tokens": 40915269.0,
      "step": 19710
    },
    {
      "entropy": 5.5748247623443605,
      "epoch": 1.7883708272859216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046818795308569815,
      "loss": 5.0823,
      "mean_token_accuracy": 0.19952248632907868,
      "num_tokens": 40925987.0,
      "step": 19715
    },
    {
      "entropy": 5.497185230255127,
      "epoch": 1.7888243831640058,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004681713644065133,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21137189120054245,
      "num_tokens": 40936130.0,
      "step": 19720
    },
    {
      "entropy": 5.468851041793823,
      "epoch": 1.78927793904209,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004681547717324954,
      "loss": 4.8934,
      "mean_token_accuracy": 0.21224325299263,
      "num_tokens": 40945476.0,
      "step": 19725
    },
    {
      "entropy": 5.450008964538574,
      "epoch": 1.7897314949201741,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046813817506398755,
      "loss": 4.8696,
      "mean_token_accuracy": 0.217603899538517,
      "num_tokens": 40956492.0,
      "step": 19730
    },
    {
      "entropy": 5.5896382331848145,
      "epoch": 1.7901850507982582,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046812157440133303,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20275839120149614,
      "num_tokens": 40966990.0,
      "step": 19735
    },
    {
      "entropy": 5.575714921951294,
      "epoch": 1.7906386066763424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004681049697448752,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21206290721893312,
      "num_tokens": 40977065.0,
      "step": 19740
    },
    {
      "entropy": 5.519251680374145,
      "epoch": 1.7910921625544267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046808836109495734,
      "loss": 5.0133,
      "mean_token_accuracy": 0.20591092705726624,
      "num_tokens": 40987207.0,
      "step": 19745
    },
    {
      "entropy": 5.5362590789794925,
      "epoch": 1.791545718432511,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004680717484519229,
      "loss": 4.9674,
      "mean_token_accuracy": 0.20424300730228423,
      "num_tokens": 40997541.0,
      "step": 19750
    },
    {
      "entropy": 5.5382311820983885,
      "epoch": 1.791999274310595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004680551318161156,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20712293982505797,
      "num_tokens": 41009044.0,
      "step": 19755
    },
    {
      "entropy": 5.559739017486573,
      "epoch": 1.7924528301886793,
      "grad_norm": 1.0,
      "learning_rate": 0.00046803851118787897,
      "loss": 5.008,
      "mean_token_accuracy": 0.20548251569271087,
      "num_tokens": 41018165.0,
      "step": 19760
    },
    {
      "entropy": 5.497653388977051,
      "epoch": 1.7929063860667633,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004680218865675567,
      "loss": 4.9652,
      "mean_token_accuracy": 0.20819184333086013,
      "num_tokens": 41028972.0,
      "step": 19765
    },
    {
      "entropy": 5.507817316055298,
      "epoch": 1.7933599419448476,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004680052579554926,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20903472155332564,
      "num_tokens": 41039552.0,
      "step": 19770
    },
    {
      "entropy": 5.620876598358154,
      "epoch": 1.7938134978229319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004679886253520307,
      "loss": 5.0528,
      "mean_token_accuracy": 0.1981100857257843,
      "num_tokens": 41049808.0,
      "step": 19775
    },
    {
      "entropy": 5.539993143081665,
      "epoch": 1.794267053701016,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046797198875751474,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2042449876666069,
      "num_tokens": 41060574.0,
      "step": 19780
    },
    {
      "entropy": 5.614523315429688,
      "epoch": 1.7947206095791002,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004679553481722889,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20428696572780608,
      "num_tokens": 41071239.0,
      "step": 19785
    },
    {
      "entropy": 5.615858888626098,
      "epoch": 1.7951741654571842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046793870359669733,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20591361224651336,
      "num_tokens": 41080819.0,
      "step": 19790
    },
    {
      "entropy": 5.587080287933349,
      "epoch": 1.7956277213352685,
      "grad_norm": 1.25,
      "learning_rate": 0.0004679220550310842,
      "loss": 4.933,
      "mean_token_accuracy": 0.2169566422700882,
      "num_tokens": 41090457.0,
      "step": 19795
    },
    {
      "entropy": 5.530445909500122,
      "epoch": 1.7960812772133528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004679054024757937,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21040300279855728,
      "num_tokens": 41101210.0,
      "step": 19800
    },
    {
      "entropy": 5.537619400024414,
      "epoch": 1.796534833091437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004678887459311704,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20775469839572908,
      "num_tokens": 41110566.0,
      "step": 19805
    },
    {
      "entropy": 5.47896146774292,
      "epoch": 1.796988388969521,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004678720853975587,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2113064005970955,
      "num_tokens": 41120000.0,
      "step": 19810
    },
    {
      "entropy": 5.469798278808594,
      "epoch": 1.797441944847605,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046785542087530307,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2107544869184494,
      "num_tokens": 41129944.0,
      "step": 19815
    },
    {
      "entropy": 5.485931491851806,
      "epoch": 1.7978955007256894,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004678387523647483,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2056867852807045,
      "num_tokens": 41138803.0,
      "step": 19820
    },
    {
      "entropy": 5.560794973373413,
      "epoch": 1.7983490566037736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046782207986623885,
      "loss": 5.0807,
      "mean_token_accuracy": 0.19582369774580002,
      "num_tokens": 41150365.0,
      "step": 19825
    },
    {
      "entropy": 5.664840650558472,
      "epoch": 1.798802612481858,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046780540338011974,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20080285668373107,
      "num_tokens": 41160310.0,
      "step": 19830
    },
    {
      "entropy": 5.512536096572876,
      "epoch": 1.799256168359942,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004677887229067358,
      "loss": 4.9104,
      "mean_token_accuracy": 0.20896457135677338,
      "num_tokens": 41170506.0,
      "step": 19835
    },
    {
      "entropy": 5.555658769607544,
      "epoch": 1.799709724238026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046777203844643184,
      "loss": 5.0348,
      "mean_token_accuracy": 0.1990685984492302,
      "num_tokens": 41181293.0,
      "step": 19840
    },
    {
      "entropy": 5.489247608184814,
      "epoch": 1.8001632801161103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046775534999955303,
      "loss": 4.9688,
      "mean_token_accuracy": 0.20351264625787735,
      "num_tokens": 41191637.0,
      "step": 19845
    },
    {
      "entropy": 5.479731512069702,
      "epoch": 1.8006168359941945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004677386575664445,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2177043989300728,
      "num_tokens": 41201833.0,
      "step": 19850
    },
    {
      "entropy": 5.57675142288208,
      "epoch": 1.8010703918722788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046772196114745133,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2109220579266548,
      "num_tokens": 41211243.0,
      "step": 19855
    },
    {
      "entropy": 5.58609938621521,
      "epoch": 1.8015239477503628,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004677052607429189,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21167860627174379,
      "num_tokens": 41221111.0,
      "step": 19860
    },
    {
      "entropy": 5.501392126083374,
      "epoch": 1.8019775036284469,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004676885563531927,
      "loss": 5.0146,
      "mean_token_accuracy": 0.2025648310780525,
      "num_tokens": 41231578.0,
      "step": 19865
    },
    {
      "entropy": 5.509230232238769,
      "epoch": 1.8024310595065312,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046767184797861795,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20938555002212525,
      "num_tokens": 41242130.0,
      "step": 19870
    },
    {
      "entropy": 5.59357099533081,
      "epoch": 1.8028846153846154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046765513561954026,
      "loss": 5.0226,
      "mean_token_accuracy": 0.19962058812379838,
      "num_tokens": 41252885.0,
      "step": 19875
    },
    {
      "entropy": 5.595680999755859,
      "epoch": 1.8033381712626997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046763841927630534,
      "loss": 5.0556,
      "mean_token_accuracy": 0.20472382009029388,
      "num_tokens": 41263532.0,
      "step": 19880
    },
    {
      "entropy": 5.59276933670044,
      "epoch": 1.8037917271407837,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046762169894925876,
      "loss": 5.0079,
      "mean_token_accuracy": 0.20627274066209794,
      "num_tokens": 41273469.0,
      "step": 19885
    },
    {
      "entropy": 5.55303692817688,
      "epoch": 1.8042452830188678,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004676049746387464,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20153881907463073,
      "num_tokens": 41283663.0,
      "step": 19890
    },
    {
      "entropy": 5.510688352584839,
      "epoch": 1.804698838896952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004675882463451141,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21141158491373063,
      "num_tokens": 41294657.0,
      "step": 19895
    },
    {
      "entropy": 5.568413734436035,
      "epoch": 1.8051523947750363,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004675715140687078,
      "loss": 5.0009,
      "mean_token_accuracy": 0.20919404029846192,
      "num_tokens": 41305124.0,
      "step": 19900
    },
    {
      "entropy": 5.4937211036682125,
      "epoch": 1.8056059506531206,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046755477780987344,
      "loss": 4.9496,
      "mean_token_accuracy": 0.20851007848978043,
      "num_tokens": 41315839.0,
      "step": 19905
    },
    {
      "entropy": 5.574371242523194,
      "epoch": 1.8060595065312046,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046753803756895726,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20170085430145263,
      "num_tokens": 41326388.0,
      "step": 19910
    },
    {
      "entropy": 5.527889442443848,
      "epoch": 1.8065130624092887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004675212933463054,
      "loss": 5.0737,
      "mean_token_accuracy": 0.1926412522792816,
      "num_tokens": 41337333.0,
      "step": 19915
    },
    {
      "entropy": 5.618911981582642,
      "epoch": 1.806966618287373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004675045451422641,
      "loss": 5.0657,
      "mean_token_accuracy": 0.19672111868858339,
      "num_tokens": 41348353.0,
      "step": 19920
    },
    {
      "entropy": 5.586848640441895,
      "epoch": 1.8074201741654572,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046748779295717973,
      "loss": 5.0304,
      "mean_token_accuracy": 0.213688226044178,
      "num_tokens": 41358797.0,
      "step": 19925
    },
    {
      "entropy": 5.5675770282745365,
      "epoch": 1.8078737300435415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004674710367913989,
      "loss": 5.0237,
      "mean_token_accuracy": 0.2066128134727478,
      "num_tokens": 41368818.0,
      "step": 19930
    },
    {
      "entropy": 5.5882336616516115,
      "epoch": 1.8083272859216255,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004674542766452678,
      "loss": 5.0116,
      "mean_token_accuracy": 0.2071378633379936,
      "num_tokens": 41378949.0,
      "step": 19935
    },
    {
      "entropy": 5.566381454467773,
      "epoch": 1.8087808417997098,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046743751251913327,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20560411214828492,
      "num_tokens": 41388892.0,
      "step": 19940
    },
    {
      "entropy": 5.5644384860992435,
      "epoch": 1.8092343976777938,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046742074441334197,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20125612020492553,
      "num_tokens": 41399197.0,
      "step": 19945
    },
    {
      "entropy": 5.52111587524414,
      "epoch": 1.809687953555878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004674039723282406,
      "loss": 5.0377,
      "mean_token_accuracy": 0.1966809868812561,
      "num_tokens": 41409551.0,
      "step": 19950
    },
    {
      "entropy": 5.518880939483642,
      "epoch": 1.8101415094339623,
      "grad_norm": 1.125,
      "learning_rate": 0.00046738719626417607,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20883867889642715,
      "num_tokens": 41418900.0,
      "step": 19955
    },
    {
      "entropy": 5.52384581565857,
      "epoch": 1.8105950653120464,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046737041622149536,
      "loss": 4.949,
      "mean_token_accuracy": 0.20624009966850282,
      "num_tokens": 41428204.0,
      "step": 19960
    },
    {
      "entropy": 5.538127660751343,
      "epoch": 1.8110486211901307,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004673536322005454,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20278254002332688,
      "num_tokens": 41437922.0,
      "step": 19965
    },
    {
      "entropy": 5.561640787124634,
      "epoch": 1.8115021770682147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004673368442016732,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20119996517896652,
      "num_tokens": 41447411.0,
      "step": 19970
    },
    {
      "entropy": 5.607109880447387,
      "epoch": 1.811955732946299,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004673200522252261,
      "loss": 5.093,
      "mean_token_accuracy": 0.20437152981758117,
      "num_tokens": 41458756.0,
      "step": 19975
    },
    {
      "entropy": 5.579017496109008,
      "epoch": 1.8124092888243832,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004673032562715514,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2080783352255821,
      "num_tokens": 41468610.0,
      "step": 19980
    },
    {
      "entropy": 5.555026006698609,
      "epoch": 1.8128628447024675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046728645634099627,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19671774953603743,
      "num_tokens": 41479849.0,
      "step": 19985
    },
    {
      "entropy": 5.526775121688843,
      "epoch": 1.8133164005805515,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046726965243390825,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2027814969420433,
      "num_tokens": 41490945.0,
      "step": 19990
    },
    {
      "entropy": 5.486761856079101,
      "epoch": 1.8137699564586356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004672528445506348,
      "loss": 4.9515,
      "mean_token_accuracy": 0.20101569592952728,
      "num_tokens": 41501479.0,
      "step": 19995
    },
    {
      "entropy": 5.525984382629394,
      "epoch": 1.8142235123367199,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004672360326915236,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21505804359912872,
      "num_tokens": 41512447.0,
      "step": 20000
    },
    {
      "entropy": 5.509391450881958,
      "epoch": 1.8146770682148041,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004672192168569222,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21713926196098327,
      "num_tokens": 41521905.0,
      "step": 20005
    },
    {
      "entropy": 5.505262660980224,
      "epoch": 1.8151306240928884,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046720239704717845,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20470809191465378,
      "num_tokens": 41533143.0,
      "step": 20010
    },
    {
      "entropy": 5.530465078353882,
      "epoch": 1.8155841799709724,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046718557326264,
      "loss": 4.9591,
      "mean_token_accuracy": 0.20491113215684892,
      "num_tokens": 41542234.0,
      "step": 20015
    },
    {
      "entropy": 5.579422378540039,
      "epoch": 1.8160377358490565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004671687455036551,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2017393171787262,
      "num_tokens": 41552632.0,
      "step": 20020
    },
    {
      "entropy": 5.54284815788269,
      "epoch": 1.8164912917271407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004671519137705715,
      "loss": 5.016,
      "mean_token_accuracy": 0.21005022525787354,
      "num_tokens": 41563723.0,
      "step": 20025
    },
    {
      "entropy": 5.551046705245971,
      "epoch": 1.816944847605225,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046713507806373736,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21051034778356553,
      "num_tokens": 41573680.0,
      "step": 20030
    },
    {
      "entropy": 5.553695774078369,
      "epoch": 1.8173984034833093,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046711823838350086,
      "loss": 5.0444,
      "mean_token_accuracy": 0.20358869880437852,
      "num_tokens": 41584722.0,
      "step": 20035
    },
    {
      "entropy": 5.532588720321655,
      "epoch": 1.8178519593613933,
      "grad_norm": 1.0,
      "learning_rate": 0.00046710139473021025,
      "loss": 5.0871,
      "mean_token_accuracy": 0.19790272265672684,
      "num_tokens": 41595785.0,
      "step": 20040
    },
    {
      "entropy": 5.569847583770752,
      "epoch": 1.8183055152394774,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046708454710421376,
      "loss": 5.107,
      "mean_token_accuracy": 0.1956186145544052,
      "num_tokens": 41606023.0,
      "step": 20045
    },
    {
      "entropy": 5.566348218917847,
      "epoch": 1.8187590711175616,
      "grad_norm": 1.125,
      "learning_rate": 0.00046706769550586,
      "loss": 5.0154,
      "mean_token_accuracy": 0.20516480803489684,
      "num_tokens": 41616010.0,
      "step": 20050
    },
    {
      "entropy": 5.557003307342529,
      "epoch": 1.819212626995646,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046705083993549723,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20764904767274855,
      "num_tokens": 41625105.0,
      "step": 20055
    },
    {
      "entropy": 5.5250067710876465,
      "epoch": 1.8196661828737302,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004670339803934743,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2096752032637596,
      "num_tokens": 41634989.0,
      "step": 20060
    },
    {
      "entropy": 5.5185634136199955,
      "epoch": 1.8201197387518142,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004670171168801396,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2066364035010338,
      "num_tokens": 41645736.0,
      "step": 20065
    },
    {
      "entropy": 5.605870532989502,
      "epoch": 1.8205732946298983,
      "grad_norm": 1.0625,
      "learning_rate": 0.000467000249395842,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20920195430517197,
      "num_tokens": 41655505.0,
      "step": 20070
    },
    {
      "entropy": 5.528751707077026,
      "epoch": 1.8210268505079825,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004669833779409303,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20744871944189072,
      "num_tokens": 41665732.0,
      "step": 20075
    },
    {
      "entropy": 5.504740619659424,
      "epoch": 1.8214804063860668,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004669665025157535,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2010861560702324,
      "num_tokens": 41676050.0,
      "step": 20080
    },
    {
      "entropy": 5.614622354507446,
      "epoch": 1.821933962264151,
      "grad_norm": 1.125,
      "learning_rate": 0.00046694962312066046,
      "loss": 5.1475,
      "mean_token_accuracy": 0.19905100166797637,
      "num_tokens": 41686787.0,
      "step": 20085
    },
    {
      "entropy": 5.545039892196655,
      "epoch": 1.822387518142235,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046693273975600036,
      "loss": 4.944,
      "mean_token_accuracy": 0.20888380259275435,
      "num_tokens": 41697168.0,
      "step": 20090
    },
    {
      "entropy": 5.503949499130249,
      "epoch": 1.8228410740203191,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004669158524221222,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20895278602838516,
      "num_tokens": 41707524.0,
      "step": 20095
    },
    {
      "entropy": 5.587649965286255,
      "epoch": 1.8232946298984034,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046689896111937537,
      "loss": 5.0626,
      "mean_token_accuracy": 0.2100888043642044,
      "num_tokens": 41718109.0,
      "step": 20100
    },
    {
      "entropy": 5.64175763130188,
      "epoch": 1.8237481857764877,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004668820658481091,
      "loss": 5.0354,
      "mean_token_accuracy": 0.20424672663211824,
      "num_tokens": 41728862.0,
      "step": 20105
    },
    {
      "entropy": 5.652068376541138,
      "epoch": 1.824201741654572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046686516660867283,
      "loss": 5.1023,
      "mean_token_accuracy": 0.1998430848121643,
      "num_tokens": 41739024.0,
      "step": 20110
    },
    {
      "entropy": 5.565501117706299,
      "epoch": 1.824655297532656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004668482634014159,
      "loss": 5.0277,
      "mean_token_accuracy": 0.20229787528514862,
      "num_tokens": 41748979.0,
      "step": 20115
    },
    {
      "entropy": 5.510527801513672,
      "epoch": 1.8251088534107403,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004668313562266882,
      "loss": 5.036,
      "mean_token_accuracy": 0.208828841149807,
      "num_tokens": 41759972.0,
      "step": 20120
    },
    {
      "entropy": 5.5156959056854244,
      "epoch": 1.8255624092888243,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000466814445084839,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21409412175416948,
      "num_tokens": 41770150.0,
      "step": 20125
    },
    {
      "entropy": 5.549961090087891,
      "epoch": 1.8260159651669086,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004667975299762183,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21211618036031724,
      "num_tokens": 41780405.0,
      "step": 20130
    },
    {
      "entropy": 5.5162074089050295,
      "epoch": 1.8264695210449928,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046678061090117576,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21304504275321962,
      "num_tokens": 41791269.0,
      "step": 20135
    },
    {
      "entropy": 5.442075777053833,
      "epoch": 1.8269230769230769,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004667636878600613,
      "loss": 4.917,
      "mean_token_accuracy": 0.20943055897951127,
      "num_tokens": 41802513.0,
      "step": 20140
    },
    {
      "entropy": 5.578953790664673,
      "epoch": 1.8273766328011611,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004667467608532249,
      "loss": 5.0589,
      "mean_token_accuracy": 0.20094809979200362,
      "num_tokens": 41814149.0,
      "step": 20145
    },
    {
      "entropy": 5.572923946380615,
      "epoch": 1.8278301886792452,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004667298298810167,
      "loss": 5.035,
      "mean_token_accuracy": 0.20357026308774948,
      "num_tokens": 41825763.0,
      "step": 20150
    },
    {
      "entropy": 5.522440624237061,
      "epoch": 1.8282837445573294,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004667128949437867,
      "loss": 5.0115,
      "mean_token_accuracy": 0.20984289944171905,
      "num_tokens": 41834543.0,
      "step": 20155
    },
    {
      "entropy": 5.519170665740967,
      "epoch": 1.8287373004354137,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004666959560418852,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21681906580924987,
      "num_tokens": 41844940.0,
      "step": 20160
    },
    {
      "entropy": 5.498330211639404,
      "epoch": 1.829190856313498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046667901317566245,
      "loss": 4.9249,
      "mean_token_accuracy": 0.20911678224802016,
      "num_tokens": 41855532.0,
      "step": 20165
    },
    {
      "entropy": 5.592134714126587,
      "epoch": 1.829644412191582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004666620663454689,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20335020124912262,
      "num_tokens": 41865661.0,
      "step": 20170
    },
    {
      "entropy": 5.5471094131469725,
      "epoch": 1.830097968069666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046664511555165496,
      "loss": 5.0347,
      "mean_token_accuracy": 0.2055257648229599,
      "num_tokens": 41875662.0,
      "step": 20175
    },
    {
      "entropy": 5.522885942459107,
      "epoch": 1.8305515239477503,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046662816079457114,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21466411501169205,
      "num_tokens": 41886211.0,
      "step": 20180
    },
    {
      "entropy": 5.623113489151001,
      "epoch": 1.8310050798258346,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004666112020745682,
      "loss": 5.1163,
      "mean_token_accuracy": 0.19894250482320786,
      "num_tokens": 41896737.0,
      "step": 20185
    },
    {
      "entropy": 5.552207946777344,
      "epoch": 1.8314586357039189,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004665942393919967,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20558253079652786,
      "num_tokens": 41907768.0,
      "step": 20190
    },
    {
      "entropy": 5.5546571731567385,
      "epoch": 1.831912191582003,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046657727274720755,
      "loss": 4.9649,
      "mean_token_accuracy": 0.20418288111686705,
      "num_tokens": 41917587.0,
      "step": 20195
    },
    {
      "entropy": 5.475027942657471,
      "epoch": 1.832365747460087,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004665603021405516,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21469545364379883,
      "num_tokens": 41927842.0,
      "step": 20200
    },
    {
      "entropy": 5.485632419586182,
      "epoch": 1.8328193033381712,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004665433275723798,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21441649943590163,
      "num_tokens": 41939899.0,
      "step": 20205
    },
    {
      "entropy": 5.527506351470947,
      "epoch": 1.8332728592162555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004665263490430431,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20336755961179734,
      "num_tokens": 41950932.0,
      "step": 20210
    },
    {
      "entropy": 5.5579382419586185,
      "epoch": 1.8337264150943398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004665093665528928,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20819393843412398,
      "num_tokens": 41963512.0,
      "step": 20215
    },
    {
      "entropy": 5.636223602294922,
      "epoch": 1.8341799709724238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046649238010227983,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21151686012744902,
      "num_tokens": 41974677.0,
      "step": 20220
    },
    {
      "entropy": 5.485407495498658,
      "epoch": 1.8346335268505078,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004664753896915558,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21072100549936296,
      "num_tokens": 41984582.0,
      "step": 20225
    },
    {
      "entropy": 5.543138790130615,
      "epoch": 1.8350870827285921,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004664583953210719,
      "loss": 4.9368,
      "mean_token_accuracy": 0.20556853264570235,
      "num_tokens": 41994624.0,
      "step": 20230
    },
    {
      "entropy": 5.525238418579102,
      "epoch": 1.8355406386066764,
      "grad_norm": 1.125,
      "learning_rate": 0.0004664413969911795,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21315260231494904,
      "num_tokens": 42004835.0,
      "step": 20235
    },
    {
      "entropy": 5.474159622192383,
      "epoch": 1.8359941944847606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004664243947022303,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20505469739437104,
      "num_tokens": 42014727.0,
      "step": 20240
    },
    {
      "entropy": 5.56663556098938,
      "epoch": 1.8364477503628447,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046640738845457586,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20703188925981522,
      "num_tokens": 42023628.0,
      "step": 20245
    },
    {
      "entropy": 5.59652886390686,
      "epoch": 1.8369013062409287,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004663903782485678,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20446922183036803,
      "num_tokens": 42033587.0,
      "step": 20250
    },
    {
      "entropy": 5.5812348365783695,
      "epoch": 1.837354862119013,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046637336408455793,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20150297582149507,
      "num_tokens": 42043765.0,
      "step": 20255
    },
    {
      "entropy": 5.608831453323364,
      "epoch": 1.8378084179970973,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046635634596289816,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20444309413433076,
      "num_tokens": 42054205.0,
      "step": 20260
    },
    {
      "entropy": 5.487960290908814,
      "epoch": 1.8382619738751815,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004663393238839404,
      "loss": 4.9104,
      "mean_token_accuracy": 0.20541877448558807,
      "num_tokens": 42064512.0,
      "step": 20265
    },
    {
      "entropy": 5.507561206817627,
      "epoch": 1.8387155297532656,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004663222978480366,
      "loss": 4.9937,
      "mean_token_accuracy": 0.205137300491333,
      "num_tokens": 42074954.0,
      "step": 20270
    },
    {
      "entropy": 5.568202161788941,
      "epoch": 1.8391690856313496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000466305267855539,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20168593227863313,
      "num_tokens": 42084914.0,
      "step": 20275
    },
    {
      "entropy": 5.576024961471558,
      "epoch": 1.8396226415094339,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004662882339067996,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20538800060749055,
      "num_tokens": 42095454.0,
      "step": 20280
    },
    {
      "entropy": 5.587305212020874,
      "epoch": 1.8400761973875182,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004662711960021709,
      "loss": 5.0882,
      "mean_token_accuracy": 0.1997831478714943,
      "num_tokens": 42106960.0,
      "step": 20285
    },
    {
      "entropy": 5.525085353851319,
      "epoch": 1.8405297532656024,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046625415414200513,
      "loss": 4.9626,
      "mean_token_accuracy": 0.20938224345445633,
      "num_tokens": 42117009.0,
      "step": 20290
    },
    {
      "entropy": 5.577379465103149,
      "epoch": 1.8409833091436865,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004662371083266546,
      "loss": 5.002,
      "mean_token_accuracy": 0.2110588639974594,
      "num_tokens": 42126371.0,
      "step": 20295
    },
    {
      "entropy": 5.58545708656311,
      "epoch": 1.8414368650217707,
      "grad_norm": 1.125,
      "learning_rate": 0.000466220058556472,
      "loss": 5.031,
      "mean_token_accuracy": 0.20728013813495635,
      "num_tokens": 42135890.0,
      "step": 20300
    },
    {
      "entropy": 5.6177905082702635,
      "epoch": 1.8418904208998548,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046620300483180986,
      "loss": 5.1298,
      "mean_token_accuracy": 0.19759823381900787,
      "num_tokens": 42145696.0,
      "step": 20305
    },
    {
      "entropy": 5.548390865325928,
      "epoch": 1.842343976777939,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004661859471530209,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21023260056972504,
      "num_tokens": 42156236.0,
      "step": 20310
    },
    {
      "entropy": 5.597132444381714,
      "epoch": 1.8427975326560233,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004661688855204578,
      "loss": 5.0606,
      "mean_token_accuracy": 0.20116368681192398,
      "num_tokens": 42166179.0,
      "step": 20315
    },
    {
      "entropy": 5.510843658447266,
      "epoch": 1.8432510885341074,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046615181993447344,
      "loss": 4.9412,
      "mean_token_accuracy": 0.20522765815258026,
      "num_tokens": 42176643.0,
      "step": 20320
    },
    {
      "entropy": 5.596554136276245,
      "epoch": 1.8437046444121916,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004661347503954207,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19032192081212998,
      "num_tokens": 42188328.0,
      "step": 20325
    },
    {
      "entropy": 5.601656866073609,
      "epoch": 1.8441582002902757,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046611767690365265,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21157871037721634,
      "num_tokens": 42198110.0,
      "step": 20330
    },
    {
      "entropy": 5.527664279937744,
      "epoch": 1.84461175616836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004661005994595223,
      "loss": 4.9823,
      "mean_token_accuracy": 0.19972015023231507,
      "num_tokens": 42208438.0,
      "step": 20335
    },
    {
      "entropy": 5.560026597976685,
      "epoch": 1.8450653120464442,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004660835180633829,
      "loss": 4.9784,
      "mean_token_accuracy": 0.2038367673754692,
      "num_tokens": 42220052.0,
      "step": 20340
    },
    {
      "entropy": 5.538200902938843,
      "epoch": 1.8455188679245285,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004660664327155876,
      "loss": 4.9749,
      "mean_token_accuracy": 0.20943403989076614,
      "num_tokens": 42229497.0,
      "step": 20345
    },
    {
      "entropy": 5.496162986755371,
      "epoch": 1.8459724238026125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004660493434164898,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20182014554738997,
      "num_tokens": 42240725.0,
      "step": 20350
    },
    {
      "entropy": 5.565572261810303,
      "epoch": 1.8464259796806965,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004660322501664429,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20267361104488374,
      "num_tokens": 42251114.0,
      "step": 20355
    },
    {
      "entropy": 5.619518995285034,
      "epoch": 1.8468795355587808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004660151529658004,
      "loss": 5.1897,
      "mean_token_accuracy": 0.19703874588012696,
      "num_tokens": 42262825.0,
      "step": 20360
    },
    {
      "entropy": 5.617018222808838,
      "epoch": 1.847333091436865,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004659980518149158,
      "loss": 5.0151,
      "mean_token_accuracy": 0.21149738729000092,
      "num_tokens": 42272176.0,
      "step": 20365
    },
    {
      "entropy": 5.602189350128174,
      "epoch": 1.8477866473149493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046598094671414285,
      "loss": 5.066,
      "mean_token_accuracy": 0.2062876984477043,
      "num_tokens": 42281934.0,
      "step": 20370
    },
    {
      "entropy": 5.540198755264282,
      "epoch": 1.8482402031930334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046596383766383533,
      "loss": 4.9589,
      "mean_token_accuracy": 0.20870660990476608,
      "num_tokens": 42291927.0,
      "step": 20375
    },
    {
      "entropy": 5.5219872951507565,
      "epoch": 1.8486937590711174,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004659467246643469,
      "loss": 4.9532,
      "mean_token_accuracy": 0.20699819028377534,
      "num_tokens": 42301407.0,
      "step": 20380
    },
    {
      "entropy": 5.455481624603271,
      "epoch": 1.8491473149492017,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046592960771603154,
      "loss": 4.9019,
      "mean_token_accuracy": 0.21629468202590943,
      "num_tokens": 42311199.0,
      "step": 20385
    },
    {
      "entropy": 5.533198833465576,
      "epoch": 1.849600870827286,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004659124868192433,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20851015746593476,
      "num_tokens": 42320725.0,
      "step": 20390
    },
    {
      "entropy": 5.621488285064697,
      "epoch": 1.8500544267053702,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004658953619743361,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20408866852521895,
      "num_tokens": 42330857.0,
      "step": 20395
    },
    {
      "entropy": 5.526382637023926,
      "epoch": 1.8505079825834543,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046587823318166427,
      "loss": 4.9196,
      "mean_token_accuracy": 0.20995397865772247,
      "num_tokens": 42340241.0,
      "step": 20400
    },
    {
      "entropy": 5.619599676132202,
      "epoch": 1.8509615384615383,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046586110044158184,
      "loss": 5.1337,
      "mean_token_accuracy": 0.19478118121623994,
      "num_tokens": 42351756.0,
      "step": 20405
    },
    {
      "entropy": 5.573098421096802,
      "epoch": 1.8514150943396226,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046584396375444335,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21466975510120392,
      "num_tokens": 42361649.0,
      "step": 20410
    },
    {
      "entropy": 5.54986982345581,
      "epoch": 1.8518686502177069,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046582682312060295,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20996458381414412,
      "num_tokens": 42371273.0,
      "step": 20415
    },
    {
      "entropy": 5.4714539527893065,
      "epoch": 1.8523222060957911,
      "grad_norm": 1.0,
      "learning_rate": 0.0004658096785404152,
      "loss": 4.9663,
      "mean_token_accuracy": 0.20906607210636138,
      "num_tokens": 42382275.0,
      "step": 20420
    },
    {
      "entropy": 5.61240701675415,
      "epoch": 1.8527757619738752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046579253001423473,
      "loss": 5.0182,
      "mean_token_accuracy": 0.201275971531868,
      "num_tokens": 42392360.0,
      "step": 20425
    },
    {
      "entropy": 5.576451110839844,
      "epoch": 1.8532293178519592,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046577537754241604,
      "loss": 5.0072,
      "mean_token_accuracy": 0.19999832063913345,
      "num_tokens": 42403695.0,
      "step": 20430
    },
    {
      "entropy": 5.557355308532715,
      "epoch": 1.8536828737300435,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000465758221125314,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20686184316873552,
      "num_tokens": 42414199.0,
      "step": 20435
    },
    {
      "entropy": 5.566457271575928,
      "epoch": 1.8541364296081277,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046574106076328333,
      "loss": 5.0809,
      "mean_token_accuracy": 0.2030840277671814,
      "num_tokens": 42423655.0,
      "step": 20440
    },
    {
      "entropy": 5.526126050949097,
      "epoch": 1.854589985486212,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004657238964566789,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21088251173496247,
      "num_tokens": 42434235.0,
      "step": 20445
    },
    {
      "entropy": 5.480642986297608,
      "epoch": 1.855043541364296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004657067282058557,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20678504556417465,
      "num_tokens": 42444701.0,
      "step": 20450
    },
    {
      "entropy": 5.4687940120697025,
      "epoch": 1.85549709724238,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004656895560111687,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20691045373678207,
      "num_tokens": 42454499.0,
      "step": 20455
    },
    {
      "entropy": 5.5942786693572994,
      "epoch": 1.8559506531204644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046567237987297323,
      "loss": 4.9735,
      "mean_token_accuracy": 0.1980825424194336,
      "num_tokens": 42465391.0,
      "step": 20460
    },
    {
      "entropy": 5.624168586730957,
      "epoch": 1.8564042089985486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004656551997916243,
      "loss": 4.9827,
      "mean_token_accuracy": 0.1953688681125641,
      "num_tokens": 42476126.0,
      "step": 20465
    },
    {
      "entropy": 5.570189476013184,
      "epoch": 1.856857764876633,
      "grad_norm": 1.0,
      "learning_rate": 0.00046563801576747716,
      "loss": 5.0589,
      "mean_token_accuracy": 0.2046019032597542,
      "num_tokens": 42486833.0,
      "step": 20470
    },
    {
      "entropy": 5.648507928848266,
      "epoch": 1.857311320754717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046562082780088733,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20140296071767808,
      "num_tokens": 42496986.0,
      "step": 20475
    },
    {
      "entropy": 5.571720695495605,
      "epoch": 1.8577648766328012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004656036358922103,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20405839681625365,
      "num_tokens": 42507352.0,
      "step": 20480
    },
    {
      "entropy": 5.565200138092041,
      "epoch": 1.8582184325108853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004655864400418014,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20897690653800965,
      "num_tokens": 42517655.0,
      "step": 20485
    },
    {
      "entropy": 5.513050222396851,
      "epoch": 1.8586719883889695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004655692402500163,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21601160764694213,
      "num_tokens": 42527502.0,
      "step": 20490
    },
    {
      "entropy": 5.529283952713013,
      "epoch": 1.8591255442670538,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046555203651721085,
      "loss": 4.9749,
      "mean_token_accuracy": 0.2061847746372223,
      "num_tokens": 42538893.0,
      "step": 20495
    },
    {
      "entropy": 5.429849100112915,
      "epoch": 1.8595791001451378,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004655348288437407,
      "loss": 4.861,
      "mean_token_accuracy": 0.21934590935707093,
      "num_tokens": 42550372.0,
      "step": 20500
    },
    {
      "entropy": 5.584519672393799,
      "epoch": 1.860032656023222,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004655176172299618,
      "loss": 5.0942,
      "mean_token_accuracy": 0.19986526370048524,
      "num_tokens": 42560814.0,
      "step": 20505
    },
    {
      "entropy": 5.634581851959228,
      "epoch": 1.8604862119013061,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046550040167622986,
      "loss": 5.0769,
      "mean_token_accuracy": 0.19995513409376145,
      "num_tokens": 42570993.0,
      "step": 20510
    },
    {
      "entropy": 5.559375667572022,
      "epoch": 1.8609397677793904,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046548318218290124,
      "loss": 4.9351,
      "mean_token_accuracy": 0.20816790759563447,
      "num_tokens": 42581357.0,
      "step": 20515
    },
    {
      "entropy": 5.523706865310669,
      "epoch": 1.8613933236574747,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004654659587503318,
      "loss": 5.0758,
      "mean_token_accuracy": 0.19584809392690658,
      "num_tokens": 42591653.0,
      "step": 20520
    },
    {
      "entropy": 5.612924957275391,
      "epoch": 1.861846879535559,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004654487313788778,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19749281257390977,
      "num_tokens": 42602232.0,
      "step": 20525
    },
    {
      "entropy": 5.571268939971924,
      "epoch": 1.862300435413643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046543150006889546,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20954685807228088,
      "num_tokens": 42613283.0,
      "step": 20530
    },
    {
      "entropy": 5.507944822311401,
      "epoch": 1.862753991291727,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004654142648207412,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21491098999977112,
      "num_tokens": 42623594.0,
      "step": 20535
    },
    {
      "entropy": 5.565047264099121,
      "epoch": 1.8632075471698113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004653970256347715,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20384258925914764,
      "num_tokens": 42633587.0,
      "step": 20540
    },
    {
      "entropy": 5.593127250671387,
      "epoch": 1.8636611030478956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046537978251134265,
      "loss": 4.9958,
      "mean_token_accuracy": 0.20397452265024185,
      "num_tokens": 42643426.0,
      "step": 20545
    },
    {
      "entropy": 5.48998236656189,
      "epoch": 1.8641146589259798,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004653625354508115,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20894977897405625,
      "num_tokens": 42653521.0,
      "step": 20550
    },
    {
      "entropy": 5.5964569568634035,
      "epoch": 1.8645682148040639,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004653452844535346,
      "loss": 5.0268,
      "mean_token_accuracy": 0.2043809026479721,
      "num_tokens": 42664818.0,
      "step": 20555
    },
    {
      "entropy": 5.667501926422119,
      "epoch": 1.865021770682148,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046532802951986863,
      "loss": 5.0684,
      "mean_token_accuracy": 0.20821174383163452,
      "num_tokens": 42674738.0,
      "step": 20560
    },
    {
      "entropy": 5.529953861236573,
      "epoch": 1.8654753265602322,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004653107706501705,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21852142363786697,
      "num_tokens": 42685003.0,
      "step": 20565
    },
    {
      "entropy": 5.484201097488404,
      "epoch": 1.8659288824383164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004652935078447973,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20671058297157288,
      "num_tokens": 42695428.0,
      "step": 20570
    },
    {
      "entropy": 5.577590703964233,
      "epoch": 1.8663824383164007,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046527624110410574,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20736818611621857,
      "num_tokens": 42704901.0,
      "step": 20575
    },
    {
      "entropy": 5.609286212921143,
      "epoch": 1.8668359941944848,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000465258970428453,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20195237398147584,
      "num_tokens": 42714379.0,
      "step": 20580
    },
    {
      "entropy": 5.624383544921875,
      "epoch": 1.8672895500725688,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046524169581819633,
      "loss": 5.1681,
      "mean_token_accuracy": 0.19813774675130844,
      "num_tokens": 42724813.0,
      "step": 20585
    },
    {
      "entropy": 5.512622976303101,
      "epoch": 1.867743105950653,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004652244172736929,
      "loss": 5.0003,
      "mean_token_accuracy": 0.20564983189105987,
      "num_tokens": 42736590.0,
      "step": 20590
    },
    {
      "entropy": 5.616318225860596,
      "epoch": 1.8681966618287373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004652071347953,
      "loss": 5.0857,
      "mean_token_accuracy": 0.1994020715355873,
      "num_tokens": 42746718.0,
      "step": 20595
    },
    {
      "entropy": 5.560640478134156,
      "epoch": 1.8686502177068216,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046518984838337513,
      "loss": 4.9689,
      "mean_token_accuracy": 0.20223457664251326,
      "num_tokens": 42756484.0,
      "step": 20600
    },
    {
      "entropy": 5.579528045654297,
      "epoch": 1.8691037735849056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004651725580382757,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20161826014518738,
      "num_tokens": 42768512.0,
      "step": 20605
    },
    {
      "entropy": 5.533065414428711,
      "epoch": 1.8695573294629897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046515526376035934,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21082388907670974,
      "num_tokens": 42779812.0,
      "step": 20610
    },
    {
      "entropy": 5.4643162250518795,
      "epoch": 1.870010885341074,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004651379655499837,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21543058305978774,
      "num_tokens": 42791163.0,
      "step": 20615
    },
    {
      "entropy": 5.540845632553101,
      "epoch": 1.8704644412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046512066340750636,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20470986813306807,
      "num_tokens": 42801810.0,
      "step": 20620
    },
    {
      "entropy": 5.603699636459351,
      "epoch": 1.8709179970972425,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004651033573332853,
      "loss": 5.0062,
      "mean_token_accuracy": 0.20591593086719512,
      "num_tokens": 42813358.0,
      "step": 20625
    },
    {
      "entropy": 5.460213851928711,
      "epoch": 1.8713715529753265,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004650860473276784,
      "loss": 4.8519,
      "mean_token_accuracy": 0.21921395212411882,
      "num_tokens": 42823519.0,
      "step": 20630
    },
    {
      "entropy": 5.571874475479126,
      "epoch": 1.8718251088534106,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004650687333910436,
      "loss": 5.0334,
      "mean_token_accuracy": 0.20522348582744598,
      "num_tokens": 42833257.0,
      "step": 20635
    },
    {
      "entropy": 5.595556211471558,
      "epoch": 1.8722786647314948,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046505141552373897,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21454567611217498,
      "num_tokens": 42843536.0,
      "step": 20640
    },
    {
      "entropy": 5.54931116104126,
      "epoch": 1.8727322206095791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004650340937261226,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2033155769109726,
      "num_tokens": 42853911.0,
      "step": 20645
    },
    {
      "entropy": 5.641363668441772,
      "epoch": 1.8731857764876634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004650167679985528,
      "loss": 5.1005,
      "mean_token_accuracy": 0.19384931772947311,
      "num_tokens": 42864173.0,
      "step": 20650
    },
    {
      "entropy": 5.5625217914581295,
      "epoch": 1.8736393323657474,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004649994383413877,
      "loss": 4.9542,
      "mean_token_accuracy": 0.20902465730905534,
      "num_tokens": 42873470.0,
      "step": 20655
    },
    {
      "entropy": 5.509566593170166,
      "epoch": 1.8740928882438317,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046498210475498595,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21467747837305068,
      "num_tokens": 42882496.0,
      "step": 20660
    },
    {
      "entropy": 5.505311059951782,
      "epoch": 1.8745464441219157,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046496476723970574,
      "loss": 4.9549,
      "mean_token_accuracy": 0.20503194034099578,
      "num_tokens": 42893163.0,
      "step": 20665
    },
    {
      "entropy": 5.522883796691895,
      "epoch": 1.875,
      "grad_norm": 1.125,
      "learning_rate": 0.0004649474257959058,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20853859633207322,
      "num_tokens": 42903406.0,
      "step": 20670
    },
    {
      "entropy": 5.593277215957642,
      "epoch": 1.8754535558780843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004649300804239447,
      "loss": 4.9718,
      "mean_token_accuracy": 0.20735423564910888,
      "num_tokens": 42913702.0,
      "step": 20675
    },
    {
      "entropy": 5.506184911727905,
      "epoch": 1.8759071117561683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046491273112418105,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2134859025478363,
      "num_tokens": 42923877.0,
      "step": 20680
    },
    {
      "entropy": 5.5144805908203125,
      "epoch": 1.8763606676342526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046489537789697386,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20901622474193574,
      "num_tokens": 42934393.0,
      "step": 20685
    },
    {
      "entropy": 5.567466974258423,
      "epoch": 1.8768142235123366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046487802074268177,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21161091774702073,
      "num_tokens": 42945803.0,
      "step": 20690
    },
    {
      "entropy": 5.5185154438018795,
      "epoch": 1.8772677793904209,
      "grad_norm": 1.125,
      "learning_rate": 0.00046486065966166377,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21167248189449311,
      "num_tokens": 42955829.0,
      "step": 20695
    },
    {
      "entropy": 5.543769025802613,
      "epoch": 1.8777213352685052,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004648432946542791,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20266976952552795,
      "num_tokens": 42967080.0,
      "step": 20700
    },
    {
      "entropy": 5.512664699554444,
      "epoch": 1.8781748911465894,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046482592572088654,
      "loss": 4.9818,
      "mean_token_accuracy": 0.2005089059472084,
      "num_tokens": 42978082.0,
      "step": 20705
    },
    {
      "entropy": 5.572058773040771,
      "epoch": 1.8786284470246735,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046480855286184565,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20274498760700227,
      "num_tokens": 42988982.0,
      "step": 20710
    },
    {
      "entropy": 5.5833902835845945,
      "epoch": 1.8790820029027575,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046479117607751535,
      "loss": 4.9986,
      "mean_token_accuracy": 0.1983781024813652,
      "num_tokens": 42998982.0,
      "step": 20715
    },
    {
      "entropy": 5.55176157951355,
      "epoch": 1.8795355587808418,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046477379536825534,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20915919542312622,
      "num_tokens": 43009574.0,
      "step": 20720
    },
    {
      "entropy": 5.560165309906006,
      "epoch": 1.879989114658926,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004647564107344247,
      "loss": 4.9931,
      "mean_token_accuracy": 0.20352196842432022,
      "num_tokens": 43019335.0,
      "step": 20725
    },
    {
      "entropy": 5.481369352340698,
      "epoch": 1.8804426705370103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046473902217638327,
      "loss": 5.0163,
      "mean_token_accuracy": 0.20774939954280852,
      "num_tokens": 43029264.0,
      "step": 20730
    },
    {
      "entropy": 5.516702032089233,
      "epoch": 1.8808962264150944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046472162969449037,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2099528893828392,
      "num_tokens": 43040519.0,
      "step": 20735
    },
    {
      "entropy": 5.604666137695313,
      "epoch": 1.8813497822931784,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046470423328910594,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20703690648078918,
      "num_tokens": 43051146.0,
      "step": 20740
    },
    {
      "entropy": 5.6006255626678465,
      "epoch": 1.8818033381712627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046468683296058957,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20637607127428054,
      "num_tokens": 43061339.0,
      "step": 20745
    },
    {
      "entropy": 5.504037857055664,
      "epoch": 1.882256894049347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046466942870930126,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20894727855920792,
      "num_tokens": 43072624.0,
      "step": 20750
    },
    {
      "entropy": 5.590467262268066,
      "epoch": 1.8827104499274312,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004646520205356007,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2081509128212929,
      "num_tokens": 43082120.0,
      "step": 20755
    },
    {
      "entropy": 5.649250793457031,
      "epoch": 1.8831640058055152,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046463460843984805,
      "loss": 5.0907,
      "mean_token_accuracy": 0.20072863548994063,
      "num_tokens": 43093202.0,
      "step": 20760
    },
    {
      "entropy": 5.665876722335815,
      "epoch": 1.8836175616835993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004646171924224034,
      "loss": 5.1642,
      "mean_token_accuracy": 0.2048816204071045,
      "num_tokens": 43104801.0,
      "step": 20765
    },
    {
      "entropy": 5.5828657150268555,
      "epoch": 1.8840711175616836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046459977248362687,
      "loss": 4.9738,
      "mean_token_accuracy": 0.2094803050160408,
      "num_tokens": 43114855.0,
      "step": 20770
    },
    {
      "entropy": 5.570040464401245,
      "epoch": 1.8845246734397678,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004645823486238788,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20368480086326599,
      "num_tokens": 43124759.0,
      "step": 20775
    },
    {
      "entropy": 5.521081876754761,
      "epoch": 1.884978229317852,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004645649208435193,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21389394551515578,
      "num_tokens": 43135526.0,
      "step": 20780
    },
    {
      "entropy": 5.5732262134552,
      "epoch": 1.8854317851959361,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046454748914290906,
      "loss": 5.0718,
      "mean_token_accuracy": 0.20253242701292037,
      "num_tokens": 43145605.0,
      "step": 20785
    },
    {
      "entropy": 5.642473983764648,
      "epoch": 1.8858853410740202,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046453005352240835,
      "loss": 5.1235,
      "mean_token_accuracy": 0.1973668545484543,
      "num_tokens": 43156578.0,
      "step": 20790
    },
    {
      "entropy": 5.60530800819397,
      "epoch": 1.8863388969521044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000464512613982378,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20926489681005478,
      "num_tokens": 43166870.0,
      "step": 20795
    },
    {
      "entropy": 5.562167406082153,
      "epoch": 1.8867924528301887,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004644951705231783,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2154717341065407,
      "num_tokens": 43176406.0,
      "step": 20800
    },
    {
      "entropy": 5.656374311447143,
      "epoch": 1.887246008708273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046447772314517024,
      "loss": 5.1511,
      "mean_token_accuracy": 0.2020910695195198,
      "num_tokens": 43187502.0,
      "step": 20805
    },
    {
      "entropy": 5.501628065109253,
      "epoch": 1.887699564586357,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046446027184871464,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21230237782001496,
      "num_tokens": 43197928.0,
      "step": 20810
    },
    {
      "entropy": 5.533857202529907,
      "epoch": 1.888153120464441,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046444281663417236,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21144766956567765,
      "num_tokens": 43209083.0,
      "step": 20815
    },
    {
      "entropy": 5.538638877868652,
      "epoch": 1.8886066763425253,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004644253575019043,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2015242889523506,
      "num_tokens": 43219089.0,
      "step": 20820
    },
    {
      "entropy": 5.602917671203613,
      "epoch": 1.8890602322206096,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004644078944522716,
      "loss": 5.051,
      "mean_token_accuracy": 0.20531389713287354,
      "num_tokens": 43229440.0,
      "step": 20825
    },
    {
      "entropy": 5.597049617767334,
      "epoch": 1.8895137880986939,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004643904274856353,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20621335208415986,
      "num_tokens": 43239995.0,
      "step": 20830
    },
    {
      "entropy": 5.554989004135132,
      "epoch": 1.889967343976778,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046437295660235674,
      "loss": 4.977,
      "mean_token_accuracy": 0.21172338724136353,
      "num_tokens": 43249722.0,
      "step": 20835
    },
    {
      "entropy": 5.4992584705352785,
      "epoch": 1.8904208998548622,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004643554818027972,
      "loss": 5.0197,
      "mean_token_accuracy": 0.2075279548764229,
      "num_tokens": 43260029.0,
      "step": 20840
    },
    {
      "entropy": 5.554016780853272,
      "epoch": 1.8908744557329462,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004643380030873181,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20793743878602983,
      "num_tokens": 43270571.0,
      "step": 20845
    },
    {
      "entropy": 5.570363712310791,
      "epoch": 1.8913280116110305,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004643205204562808,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21176953464746476,
      "num_tokens": 43281310.0,
      "step": 20850
    },
    {
      "entropy": 5.55697021484375,
      "epoch": 1.8917815674891147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046430303391004687,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20941325575113295,
      "num_tokens": 43292904.0,
      "step": 20855
    },
    {
      "entropy": 5.554282379150391,
      "epoch": 1.8922351233671988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004642855434489779,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2037817120552063,
      "num_tokens": 43303745.0,
      "step": 20860
    },
    {
      "entropy": 5.586295175552368,
      "epoch": 1.892688679245283,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046426804907343574,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20697399526834487,
      "num_tokens": 43314652.0,
      "step": 20865
    },
    {
      "entropy": 5.527583217620849,
      "epoch": 1.893142235123367,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046425055078378206,
      "loss": 4.8744,
      "mean_token_accuracy": 0.21875638663768768,
      "num_tokens": 43324242.0,
      "step": 20870
    },
    {
      "entropy": 5.523820161819458,
      "epoch": 1.8935957910014514,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046423304858037885,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21595541387796402,
      "num_tokens": 43334054.0,
      "step": 20875
    },
    {
      "entropy": 5.469158792495728,
      "epoch": 1.8940493468795356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004642155424635878,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2215532347559929,
      "num_tokens": 43343389.0,
      "step": 20880
    },
    {
      "entropy": 5.477652502059937,
      "epoch": 1.89450290275762,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046419803243377125,
      "loss": 4.9423,
      "mean_token_accuracy": 0.2153505265712738,
      "num_tokens": 43353892.0,
      "step": 20885
    },
    {
      "entropy": 5.501353168487549,
      "epoch": 1.894956458635704,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046418051849129115,
      "loss": 4.972,
      "mean_token_accuracy": 0.214233835041523,
      "num_tokens": 43363912.0,
      "step": 20890
    },
    {
      "entropy": 5.545388412475586,
      "epoch": 1.895410014513788,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004641630006365097,
      "loss": 5.0428,
      "mean_token_accuracy": 0.20356074273586272,
      "num_tokens": 43374719.0,
      "step": 20895
    },
    {
      "entropy": 5.580736112594605,
      "epoch": 1.8958635703918723,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046414547886978926,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2027682676911354,
      "num_tokens": 43385549.0,
      "step": 20900
    },
    {
      "entropy": 5.600714111328125,
      "epoch": 1.8963171262699565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046412795319149204,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21062333434820174,
      "num_tokens": 43395503.0,
      "step": 20905
    },
    {
      "entropy": 5.538871383666992,
      "epoch": 1.8967706821480408,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046411042360198053,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2010958045721054,
      "num_tokens": 43406913.0,
      "step": 20910
    },
    {
      "entropy": 5.562172079086304,
      "epoch": 1.8972242380261248,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004640928901016173,
      "loss": 4.947,
      "mean_token_accuracy": 0.20708273500204086,
      "num_tokens": 43416459.0,
      "step": 20915
    },
    {
      "entropy": 5.53361029624939,
      "epoch": 1.8976777939042089,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046407535269076493,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20667988210916519,
      "num_tokens": 43426280.0,
      "step": 20920
    },
    {
      "entropy": 5.610696172714233,
      "epoch": 1.8981313497822931,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046405781136978606,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20656325072050094,
      "num_tokens": 43436676.0,
      "step": 20925
    },
    {
      "entropy": 5.660415744781494,
      "epoch": 1.8985849056603774,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046404026613904357,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2016023129224777,
      "num_tokens": 43447019.0,
      "step": 20930
    },
    {
      "entropy": 5.605124664306641,
      "epoch": 1.8990384615384617,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004640227169989001,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2014377772808075,
      "num_tokens": 43456609.0,
      "step": 20935
    },
    {
      "entropy": 5.576879262924194,
      "epoch": 1.8994920174165457,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004640051639497187,
      "loss": 5.0459,
      "mean_token_accuracy": 0.2026502937078476,
      "num_tokens": 43467616.0,
      "step": 20940
    },
    {
      "entropy": 5.485981082916259,
      "epoch": 1.8999455732946298,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004639876069918624,
      "loss": 5.0186,
      "mean_token_accuracy": 0.20398564338684083,
      "num_tokens": 43478209.0,
      "step": 20945
    },
    {
      "entropy": 5.527855396270752,
      "epoch": 1.900399129172714,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046397004612569416,
      "loss": 4.9072,
      "mean_token_accuracy": 0.20824503153562546,
      "num_tokens": 43487278.0,
      "step": 20950
    },
    {
      "entropy": 5.507301187515258,
      "epoch": 1.9008526850507983,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046395248135157715,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2197753131389618,
      "num_tokens": 43497457.0,
      "step": 20955
    },
    {
      "entropy": 5.562612533569336,
      "epoch": 1.9013062409288826,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046393491266987484,
      "loss": 5.0487,
      "mean_token_accuracy": 0.20178602188825606,
      "num_tokens": 43507761.0,
      "step": 20960
    },
    {
      "entropy": 5.497231721878052,
      "epoch": 1.9017597968069666,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004639173400809502,
      "loss": 4.9672,
      "mean_token_accuracy": 0.20882880091667175,
      "num_tokens": 43517198.0,
      "step": 20965
    },
    {
      "entropy": 5.540993881225586,
      "epoch": 1.9022133526850507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004638997635851669,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20050388872623442,
      "num_tokens": 43527429.0,
      "step": 20970
    },
    {
      "entropy": 5.612675666809082,
      "epoch": 1.902666908563135,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046388218318288844,
      "loss": 5.0583,
      "mean_token_accuracy": 0.20635149627923965,
      "num_tokens": 43537867.0,
      "step": 20975
    },
    {
      "entropy": 5.600853681564331,
      "epoch": 1.9031204644412192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004638645988744782,
      "loss": 4.995,
      "mean_token_accuracy": 0.20810700356960296,
      "num_tokens": 43547284.0,
      "step": 20980
    },
    {
      "entropy": 5.544989919662475,
      "epoch": 1.9035740203193035,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004638470106603,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20433826893568038,
      "num_tokens": 43557693.0,
      "step": 20985
    },
    {
      "entropy": 5.580530118942261,
      "epoch": 1.9040275761973875,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004638294185407174,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21300598680973054,
      "num_tokens": 43566909.0,
      "step": 20990
    },
    {
      "entropy": 5.587985372543335,
      "epoch": 1.9044811320754715,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638118225160944,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21214412152767181,
      "num_tokens": 43576963.0,
      "step": 20995
    },
    {
      "entropy": 5.497220659255982,
      "epoch": 1.9049346879535558,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046379422258679477,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20666191428899766,
      "num_tokens": 43586649.0,
      "step": 21000
    },
    {
      "epoch": 1.9049346879535558,
      "eval_entropy": 5.2827021436993356,
      "eval_loss": 5.041497230529785,
      "eval_mean_token_accuracy": 0.21340517185426136,
      "eval_num_tokens": 43586649.0,
      "eval_runtime": 20.6472,
      "eval_samples_per_second": 1712.58,
      "eval_steps_per_second": 214.072,
      "step": 21000
    },
    {
      "entropy": 5.486915922164917,
      "epoch": 1.90538824383164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004637766187531825,
      "loss": 4.9612,
      "mean_token_accuracy": 0.21056651920080185,
      "num_tokens": 43596506.0,
      "step": 21005
    },
    {
      "entropy": 5.500951766967773,
      "epoch": 1.9058417997097243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046375901101562163,
      "loss": 5.0464,
      "mean_token_accuracy": 0.20540109127759934,
      "num_tokens": 43607188.0,
      "step": 21010
    },
    {
      "entropy": 5.562234878540039,
      "epoch": 1.9062953555878084,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004637413993744763,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21434430927038192,
      "num_tokens": 43617465.0,
      "step": 21015
    },
    {
      "entropy": 5.558161783218384,
      "epoch": 1.9067489114658926,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046372378383011077,
      "loss": 4.9468,
      "mean_token_accuracy": 0.20854316800832748,
      "num_tokens": 43627692.0,
      "step": 21020
    },
    {
      "entropy": 5.5884510517120365,
      "epoch": 1.9072024673439767,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046370616438288933,
      "loss": 5.0386,
      "mean_token_accuracy": 0.20229142010211945,
      "num_tokens": 43637350.0,
      "step": 21025
    },
    {
      "entropy": 5.564635181427002,
      "epoch": 1.907656023222061,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046368854103317624,
      "loss": 5.0286,
      "mean_token_accuracy": 0.2014291226863861,
      "num_tokens": 43647092.0,
      "step": 21030
    },
    {
      "entropy": 5.542835474014282,
      "epoch": 1.9081095791001452,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046367091378133615,
      "loss": 4.986,
      "mean_token_accuracy": 0.20898470431566238,
      "num_tokens": 43657237.0,
      "step": 21035
    },
    {
      "entropy": 5.577055358886719,
      "epoch": 1.9085631349782293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004636532826277333,
      "loss": 4.9968,
      "mean_token_accuracy": 0.20737866908311844,
      "num_tokens": 43667966.0,
      "step": 21040
    },
    {
      "entropy": 5.530750703811646,
      "epoch": 1.9090166908563135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046363564757273264,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2092711329460144,
      "num_tokens": 43678059.0,
      "step": 21045
    },
    {
      "entropy": 5.525274276733398,
      "epoch": 1.9094702467343976,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046361800861669867,
      "loss": 4.9845,
      "mean_token_accuracy": 0.2132541298866272,
      "num_tokens": 43687556.0,
      "step": 21050
    },
    {
      "entropy": 5.587990045547485,
      "epoch": 1.9099238026124818,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046360036575999623,
      "loss": 5.0244,
      "mean_token_accuracy": 0.20550597310066224,
      "num_tokens": 43697791.0,
      "step": 21055
    },
    {
      "entropy": 5.578096628189087,
      "epoch": 1.9103773584905661,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004635827190029902,
      "loss": 4.9358,
      "mean_token_accuracy": 0.20452708899974822,
      "num_tokens": 43707390.0,
      "step": 21060
    },
    {
      "entropy": 5.6091104507446286,
      "epoch": 1.9108309143686504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004635650683460454,
      "loss": 5.062,
      "mean_token_accuracy": 0.20398611426353455,
      "num_tokens": 43716842.0,
      "step": 21065
    },
    {
      "entropy": 5.482925319671631,
      "epoch": 1.9112844702467344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000463547413789527,
      "loss": 4.9841,
      "mean_token_accuracy": 0.20588076561689378,
      "num_tokens": 43726967.0,
      "step": 21070
    },
    {
      "entropy": 5.621833944320679,
      "epoch": 1.9117380261248185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046352975533379997,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21100803166627885,
      "num_tokens": 43737425.0,
      "step": 21075
    },
    {
      "entropy": 5.5032435894012455,
      "epoch": 1.9121915820029027,
      "grad_norm": 1.125,
      "learning_rate": 0.00046351209297922964,
      "loss": 4.9277,
      "mean_token_accuracy": 0.215091273188591,
      "num_tokens": 43747395.0,
      "step": 21080
    },
    {
      "entropy": 5.527018260955811,
      "epoch": 1.912645137880987,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046349442672618114,
      "loss": 5.0381,
      "mean_token_accuracy": 0.2003874033689499,
      "num_tokens": 43756348.0,
      "step": 21085
    },
    {
      "entropy": 5.550302171707154,
      "epoch": 1.9130986937590713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046347675657501986,
      "loss": 4.9845,
      "mean_token_accuracy": 0.20703417658805848,
      "num_tokens": 43767629.0,
      "step": 21090
    },
    {
      "entropy": 5.559140729904175,
      "epoch": 1.9135522496371553,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004634590825261113,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21757467538118364,
      "num_tokens": 43777398.0,
      "step": 21095
    },
    {
      "entropy": 5.564572525024414,
      "epoch": 1.9140058055152394,
      "grad_norm": 1.125,
      "learning_rate": 0.00046344140457982075,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20926016867160796,
      "num_tokens": 43787605.0,
      "step": 21100
    },
    {
      "entropy": 5.522140550613403,
      "epoch": 1.9144593613933236,
      "grad_norm": 1.109375,
      "learning_rate": 0.000463423722736514,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20708602368831636,
      "num_tokens": 43796989.0,
      "step": 21105
    },
    {
      "entropy": 5.550064373016357,
      "epoch": 1.9149129172714079,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004634060369965567,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20980975776910782,
      "num_tokens": 43806653.0,
      "step": 21110
    },
    {
      "entropy": 5.5119102001190186,
      "epoch": 1.9153664731494922,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046338834736031446,
      "loss": 4.9469,
      "mean_token_accuracy": 0.21025218218564987,
      "num_tokens": 43817152.0,
      "step": 21115
    },
    {
      "entropy": 5.545574283599853,
      "epoch": 1.9158200290275762,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046337065382815324,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20765405297279357,
      "num_tokens": 43827341.0,
      "step": 21120
    },
    {
      "entropy": 5.510351037979126,
      "epoch": 1.9162735849056602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046335295640043887,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21740228086709976,
      "num_tokens": 43837195.0,
      "step": 21125
    },
    {
      "entropy": 5.531359481811523,
      "epoch": 1.9167271407837445,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046333525507753743,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21550391167402266,
      "num_tokens": 43847473.0,
      "step": 21130
    },
    {
      "entropy": 5.509198904037476,
      "epoch": 1.9171806966618288,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004633175498598148,
      "loss": 4.9701,
      "mean_token_accuracy": 0.20922050327062608,
      "num_tokens": 43857590.0,
      "step": 21135
    },
    {
      "entropy": 5.563808107376099,
      "epoch": 1.917634252539913,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046329984074763733,
      "loss": 5.1241,
      "mean_token_accuracy": 0.1958355963230133,
      "num_tokens": 43867085.0,
      "step": 21140
    },
    {
      "entropy": 5.573216676712036,
      "epoch": 1.918087808417997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004632821277413711,
      "loss": 4.9602,
      "mean_token_accuracy": 0.20524263232946396,
      "num_tokens": 43877153.0,
      "step": 21145
    },
    {
      "entropy": 5.632125520706177,
      "epoch": 1.9185413642960811,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004632644108413825,
      "loss": 5.0833,
      "mean_token_accuracy": 0.19107387959957123,
      "num_tokens": 43887977.0,
      "step": 21150
    },
    {
      "entropy": 5.476994419097901,
      "epoch": 1.9189949201741654,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046324669004803803,
      "loss": 4.9789,
      "mean_token_accuracy": 0.20683950334787368,
      "num_tokens": 43898649.0,
      "step": 21155
    },
    {
      "entropy": 5.6364538192749025,
      "epoch": 1.9194484760522497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000463228965361704,
      "loss": 5.0881,
      "mean_token_accuracy": 0.2023804023861885,
      "num_tokens": 43909369.0,
      "step": 21160
    },
    {
      "entropy": 5.600297927856445,
      "epoch": 1.919902031930334,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004632112367827469,
      "loss": 4.9866,
      "mean_token_accuracy": 0.20831674337387085,
      "num_tokens": 43919941.0,
      "step": 21165
    },
    {
      "entropy": 5.576290559768677,
      "epoch": 1.920355587808418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004631935043115335,
      "loss": 5.0463,
      "mean_token_accuracy": 0.20568810999393464,
      "num_tokens": 43929710.0,
      "step": 21170
    },
    {
      "entropy": 5.52694091796875,
      "epoch": 1.920809143686502,
      "grad_norm": 1.125,
      "learning_rate": 0.00046317576794843056,
      "loss": 4.9718,
      "mean_token_accuracy": 0.20809148699045182,
      "num_tokens": 43939850.0,
      "step": 21175
    },
    {
      "entropy": 5.530804395675659,
      "epoch": 1.9212626995645863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046315802769380474,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2124344989657402,
      "num_tokens": 43950067.0,
      "step": 21180
    },
    {
      "entropy": 5.560138988494873,
      "epoch": 1.9217162554426706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004631402835480229,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21008018553256988,
      "num_tokens": 43960029.0,
      "step": 21185
    },
    {
      "entropy": 5.544347286224365,
      "epoch": 1.9221698113207548,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046312253551145213,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20792224556207656,
      "num_tokens": 43971221.0,
      "step": 21190
    },
    {
      "entropy": 5.570632648468018,
      "epoch": 1.9226233671988389,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046310478358445935,
      "loss": 5.0277,
      "mean_token_accuracy": 0.2045084208250046,
      "num_tokens": 43981579.0,
      "step": 21195
    },
    {
      "entropy": 5.587553834915161,
      "epoch": 1.9230769230769231,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004630870277674117,
      "loss": 5.123,
      "mean_token_accuracy": 0.19563973993062972,
      "num_tokens": 43991132.0,
      "step": 21200
    },
    {
      "entropy": 5.548740816116333,
      "epoch": 1.9235304789550072,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004630692680606764,
      "loss": 4.9196,
      "mean_token_accuracy": 0.20882577300071717,
      "num_tokens": 44001255.0,
      "step": 21205
    },
    {
      "entropy": 5.490432262420654,
      "epoch": 1.9239840348330914,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004630515044646207,
      "loss": 4.836,
      "mean_token_accuracy": 0.21653300970792771,
      "num_tokens": 44011275.0,
      "step": 21210
    },
    {
      "entropy": 5.536353349685669,
      "epoch": 1.9244375907111757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000463033736979612,
      "loss": 5.0134,
      "mean_token_accuracy": 0.20539278090000151,
      "num_tokens": 44022524.0,
      "step": 21215
    },
    {
      "entropy": 5.524604558944702,
      "epoch": 1.9248911465892597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046301596560601767,
      "loss": 4.918,
      "mean_token_accuracy": 0.20774720907211303,
      "num_tokens": 44033955.0,
      "step": 21220
    },
    {
      "entropy": 5.55246868133545,
      "epoch": 1.925344702467344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004629981903442052,
      "loss": 5.0268,
      "mean_token_accuracy": 0.2021826833486557,
      "num_tokens": 44045149.0,
      "step": 21225
    },
    {
      "entropy": 5.6093236923217775,
      "epoch": 1.925798258345428,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004629804111945424,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20155078619718553,
      "num_tokens": 44055280.0,
      "step": 21230
    },
    {
      "entropy": 5.553763055801392,
      "epoch": 1.9262518142235123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004629626281573966,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21317025125026703,
      "num_tokens": 44065578.0,
      "step": 21235
    },
    {
      "entropy": 5.47047119140625,
      "epoch": 1.9267053701015966,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004629448412331359,
      "loss": 4.9284,
      "mean_token_accuracy": 0.21356330811977386,
      "num_tokens": 44075454.0,
      "step": 21240
    },
    {
      "entropy": 5.518563747406006,
      "epoch": 1.9271589259796809,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004629270504221279,
      "loss": 5.0179,
      "mean_token_accuracy": 0.2049138143658638,
      "num_tokens": 44085680.0,
      "step": 21245
    },
    {
      "entropy": 5.4879231452941895,
      "epoch": 1.927612481857765,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004629092557247407,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21290490329265593,
      "num_tokens": 44097177.0,
      "step": 21250
    },
    {
      "entropy": 5.588929128646851,
      "epoch": 1.928066037735849,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004628914571413422,
      "loss": 5.05,
      "mean_token_accuracy": 0.20613875538110732,
      "num_tokens": 44107251.0,
      "step": 21255
    },
    {
      "entropy": 5.552894973754883,
      "epoch": 1.9285195936139332,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046287365467230035,
      "loss": 5.0305,
      "mean_token_accuracy": 0.20274046510457994,
      "num_tokens": 44117943.0,
      "step": 21260
    },
    {
      "entropy": 5.620401859283447,
      "epoch": 1.9289731494920175,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046285584831798354,
      "loss": 5.0251,
      "mean_token_accuracy": 0.21080010682344436,
      "num_tokens": 44128608.0,
      "step": 21265
    },
    {
      "entropy": 5.584255838394165,
      "epoch": 1.9294267053701017,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046283803807875996,
      "loss": 4.9946,
      "mean_token_accuracy": 0.21257734894752503,
      "num_tokens": 44138809.0,
      "step": 21270
    },
    {
      "entropy": 5.531477785110473,
      "epoch": 1.9298802612481858,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046282022395499785,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2101381629705429,
      "num_tokens": 44148532.0,
      "step": 21275
    },
    {
      "entropy": 5.580505990982056,
      "epoch": 1.9303338171262698,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004628024059470656,
      "loss": 5.0275,
      "mean_token_accuracy": 0.20669040381908416,
      "num_tokens": 44158628.0,
      "step": 21280
    },
    {
      "entropy": 5.5577842712402346,
      "epoch": 1.930787373004354,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004627845840553318,
      "loss": 4.8918,
      "mean_token_accuracy": 0.21354376524686813,
      "num_tokens": 44168574.0,
      "step": 21285
    },
    {
      "entropy": 5.599446964263916,
      "epoch": 1.9312409288824384,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046276675828016493,
      "loss": 5.0614,
      "mean_token_accuracy": 0.20820466727018355,
      "num_tokens": 44180207.0,
      "step": 21290
    },
    {
      "entropy": 5.609013271331787,
      "epoch": 1.9316944847605226,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004627489286219336,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1978868678212166,
      "num_tokens": 44190473.0,
      "step": 21295
    },
    {
      "entropy": 5.6146135330200195,
      "epoch": 1.9321480406386067,
      "grad_norm": 1.0,
      "learning_rate": 0.0004627310950810066,
      "loss": 5.0343,
      "mean_token_accuracy": 0.20951875001192094,
      "num_tokens": 44201150.0,
      "step": 21300
    },
    {
      "entropy": 5.520677757263184,
      "epoch": 1.9326015965166907,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004627132576577528,
      "loss": 4.9987,
      "mean_token_accuracy": 0.19714128077030182,
      "num_tokens": 44211272.0,
      "step": 21305
    },
    {
      "entropy": 5.577132272720337,
      "epoch": 1.933055152394775,
      "grad_norm": 0.96875,
      "learning_rate": 0.000462695416352541,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20158825367689132,
      "num_tokens": 44223556.0,
      "step": 21310
    },
    {
      "entropy": 5.546791362762451,
      "epoch": 1.9335087082728593,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046267757116574,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2087857723236084,
      "num_tokens": 44233911.0,
      "step": 21315
    },
    {
      "entropy": 5.603171873092651,
      "epoch": 1.9339622641509435,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004626597220977191,
      "loss": 5.0477,
      "mean_token_accuracy": 0.21053002923727035,
      "num_tokens": 44244108.0,
      "step": 21320
    },
    {
      "entropy": 5.486507225036621,
      "epoch": 1.9344158200290276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046264186914884737,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2079022005200386,
      "num_tokens": 44253604.0,
      "step": 21325
    },
    {
      "entropy": 5.514385032653808,
      "epoch": 1.9348693759071116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046262401231949386,
      "loss": 5.0501,
      "mean_token_accuracy": 0.199178808927536,
      "num_tokens": 44263573.0,
      "step": 21330
    },
    {
      "entropy": 5.542400693893432,
      "epoch": 1.9353229317851959,
      "grad_norm": 1.0,
      "learning_rate": 0.00046260615161002803,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2072915554046631,
      "num_tokens": 44274071.0,
      "step": 21335
    },
    {
      "entropy": 5.650053405761719,
      "epoch": 1.9357764876632801,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004625882870208192,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20556609183549882,
      "num_tokens": 44283997.0,
      "step": 21340
    },
    {
      "entropy": 5.605504035949707,
      "epoch": 1.9362300435413644,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046257041855223674,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20844724476337434,
      "num_tokens": 44295024.0,
      "step": 21345
    },
    {
      "entropy": 5.49817156791687,
      "epoch": 1.9366835994194485,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046255254620465033,
      "loss": 4.9489,
      "mean_token_accuracy": 0.21543079614639282,
      "num_tokens": 44305476.0,
      "step": 21350
    },
    {
      "entropy": 5.509833717346192,
      "epoch": 1.9371371552975325,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046253466997842936,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21113743185997008,
      "num_tokens": 44316101.0,
      "step": 21355
    },
    {
      "entropy": 5.503489017486572,
      "epoch": 1.9375907111756168,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004625167898739436,
      "loss": 4.8728,
      "mean_token_accuracy": 0.2139115959405899,
      "num_tokens": 44325058.0,
      "step": 21360
    },
    {
      "entropy": 5.665496063232422,
      "epoch": 1.938044267053701,
      "grad_norm": 1.015625,
      "learning_rate": 0.000462498905891563,
      "loss": 5.1797,
      "mean_token_accuracy": 0.18986935168504715,
      "num_tokens": 44336585.0,
      "step": 21365
    },
    {
      "entropy": 5.543269491195678,
      "epoch": 1.9384978229317853,
      "grad_norm": 1.125,
      "learning_rate": 0.0004624810180316571,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2091907352209091,
      "num_tokens": 44346911.0,
      "step": 21370
    },
    {
      "entropy": 5.515858936309814,
      "epoch": 1.9389513788098693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046246312629459596,
      "loss": 4.9494,
      "mean_token_accuracy": 0.207963527739048,
      "num_tokens": 44358040.0,
      "step": 21375
    },
    {
      "entropy": 5.564346837997436,
      "epoch": 1.9394049346879536,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004624452306807497,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21364262849092483,
      "num_tokens": 44368781.0,
      "step": 21380
    },
    {
      "entropy": 5.488349914550781,
      "epoch": 1.9398584905660377,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004624273311904882,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21095094680786133,
      "num_tokens": 44378809.0,
      "step": 21385
    },
    {
      "entropy": 5.605105686187744,
      "epoch": 1.940312046444122,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046240942782418184,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20438190251588823,
      "num_tokens": 44388517.0,
      "step": 21390
    },
    {
      "entropy": 5.631004095077515,
      "epoch": 1.9407656023222062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004623915205822007,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20357698798179627,
      "num_tokens": 44399364.0,
      "step": 21395
    },
    {
      "entropy": 5.540302753448486,
      "epoch": 1.9412191582002902,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046237360946491514,
      "loss": 4.9337,
      "mean_token_accuracy": 0.20093414485454558,
      "num_tokens": 44409506.0,
      "step": 21400
    },
    {
      "entropy": 5.5960570812225345,
      "epoch": 1.9416727140783745,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004623556944726956,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20793051570653914,
      "num_tokens": 44419446.0,
      "step": 21405
    },
    {
      "entropy": 5.544157934188843,
      "epoch": 1.9421262699564585,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046233777560591256,
      "loss": 4.9739,
      "mean_token_accuracy": 0.20520949959754944,
      "num_tokens": 44429560.0,
      "step": 21410
    },
    {
      "entropy": 5.546304750442505,
      "epoch": 1.9425798258345428,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046231985286493654,
      "loss": 4.975,
      "mean_token_accuracy": 0.20405519604682923,
      "num_tokens": 44439640.0,
      "step": 21415
    },
    {
      "entropy": 5.578876399993897,
      "epoch": 1.943033381712627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046230192625013836,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20514049381017685,
      "num_tokens": 44450480.0,
      "step": 21420
    },
    {
      "entropy": 5.479057550430298,
      "epoch": 1.9434869375907113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046228399576188844,
      "loss": 4.892,
      "mean_token_accuracy": 0.20939820855855942,
      "num_tokens": 44460633.0,
      "step": 21425
    },
    {
      "entropy": 5.556435871124267,
      "epoch": 1.9439404934687954,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046226606140055784,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2110549733042717,
      "num_tokens": 44470917.0,
      "step": 21430
    },
    {
      "entropy": 5.555643892288208,
      "epoch": 1.9443940493468794,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046224812316651736,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2043233558535576,
      "num_tokens": 44482538.0,
      "step": 21435
    },
    {
      "entropy": 5.5876607418060305,
      "epoch": 1.9448476052249637,
      "grad_norm": 1.0625,
      "learning_rate": 0.000462230181060138,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20266881585121155,
      "num_tokens": 44493518.0,
      "step": 21440
    },
    {
      "entropy": 5.587052583694458,
      "epoch": 1.945301161103048,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046221223508179077,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20541617274284363,
      "num_tokens": 44503685.0,
      "step": 21445
    },
    {
      "entropy": 5.538943290710449,
      "epoch": 1.9457547169811322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046219428523184685,
      "loss": 4.956,
      "mean_token_accuracy": 0.21194851249456406,
      "num_tokens": 44514015.0,
      "step": 21450
    },
    {
      "entropy": 5.5282660007476805,
      "epoch": 1.9462082728592163,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004621763315106774,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2088385656476021,
      "num_tokens": 44525187.0,
      "step": 21455
    },
    {
      "entropy": 5.470666694641113,
      "epoch": 1.9466618287373003,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046215837391865364,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2128278523683548,
      "num_tokens": 44535964.0,
      "step": 21460
    },
    {
      "entropy": 5.503521203994751,
      "epoch": 1.9471153846153846,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046214041245614713,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2047021880745888,
      "num_tokens": 44546985.0,
      "step": 21465
    },
    {
      "entropy": 5.594559001922607,
      "epoch": 1.9475689404934688,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004621224471235291,
      "loss": 5.0258,
      "mean_token_accuracy": 0.19286321550607682,
      "num_tokens": 44556888.0,
      "step": 21470
    },
    {
      "entropy": 5.569971132278442,
      "epoch": 1.9480224963715531,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046210447792117114,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20596435964107512,
      "num_tokens": 44566857.0,
      "step": 21475
    },
    {
      "entropy": 5.581793975830078,
      "epoch": 1.9484760522496372,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000462086504849445,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20321155339479446,
      "num_tokens": 44578152.0,
      "step": 21480
    },
    {
      "entropy": 5.5432069301605225,
      "epoch": 1.9489296081277212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004620685279087222,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21598812192678452,
      "num_tokens": 44588814.0,
      "step": 21485
    },
    {
      "entropy": 5.535352325439453,
      "epoch": 1.9493831640058055,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046205054709937474,
      "loss": 4.9684,
      "mean_token_accuracy": 0.20474015027284623,
      "num_tokens": 44599366.0,
      "step": 21490
    },
    {
      "entropy": 5.661087274551392,
      "epoch": 1.9498367198838897,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004620325624217741,
      "loss": 5.0773,
      "mean_token_accuracy": 0.1957104668021202,
      "num_tokens": 44609962.0,
      "step": 21495
    },
    {
      "entropy": 5.627064371109009,
      "epoch": 1.950290275761974,
      "grad_norm": 1.125,
      "learning_rate": 0.0004620145738762925,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20886536687612534,
      "num_tokens": 44619917.0,
      "step": 21500
    },
    {
      "entropy": 5.508686161041259,
      "epoch": 1.950743831640058,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004619965814633019,
      "loss": 4.982,
      "mean_token_accuracy": 0.21039706617593765,
      "num_tokens": 44629723.0,
      "step": 21505
    },
    {
      "entropy": 5.548584842681885,
      "epoch": 1.951197387518142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004619785851831742,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20363247990608216,
      "num_tokens": 44641011.0,
      "step": 21510
    },
    {
      "entropy": 5.573685359954834,
      "epoch": 1.9516509433962264,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046196058503628176,
      "loss": 4.963,
      "mean_token_accuracy": 0.20982809662818908,
      "num_tokens": 44651154.0,
      "step": 21515
    },
    {
      "entropy": 5.5663313388824465,
      "epoch": 1.9521044992743106,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004619425810229968,
      "loss": 5.0547,
      "mean_token_accuracy": 0.2016923740506172,
      "num_tokens": 44661672.0,
      "step": 21520
    },
    {
      "entropy": 5.506641054153443,
      "epoch": 1.952558055152395,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004619245731436917,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21523777395486832,
      "num_tokens": 44671449.0,
      "step": 21525
    },
    {
      "entropy": 5.567077016830444,
      "epoch": 1.953011611030479,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046190656139873863,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21322334855794906,
      "num_tokens": 44681977.0,
      "step": 21530
    },
    {
      "entropy": 5.548317098617554,
      "epoch": 1.953465166908563,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004618885457885103,
      "loss": 4.9549,
      "mean_token_accuracy": 0.21130643784999847,
      "num_tokens": 44691946.0,
      "step": 21535
    },
    {
      "entropy": 5.526917743682861,
      "epoch": 1.9539187227866472,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004618705263133793,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20774240046739578,
      "num_tokens": 44702012.0,
      "step": 21540
    },
    {
      "entropy": 5.534388828277588,
      "epoch": 1.9543722786647315,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004618525029737182,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21082423031330108,
      "num_tokens": 44711605.0,
      "step": 21545
    },
    {
      "entropy": 5.611813592910766,
      "epoch": 1.9548258345428158,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046183447576989966,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20988747030496596,
      "num_tokens": 44722765.0,
      "step": 21550
    },
    {
      "entropy": 5.5836060523986815,
      "epoch": 1.9552793904208998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046181644470229656,
      "loss": 4.9701,
      "mean_token_accuracy": 0.208165480196476,
      "num_tokens": 44732082.0,
      "step": 21555
    },
    {
      "entropy": 5.461510801315308,
      "epoch": 1.955732946298984,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004617984097712817,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21089304983615875,
      "num_tokens": 44742054.0,
      "step": 21560
    },
    {
      "entropy": 5.486566877365112,
      "epoch": 1.9561865021770681,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004617803709772282,
      "loss": 5.0066,
      "mean_token_accuracy": 0.2106909289956093,
      "num_tokens": 44752954.0,
      "step": 21565
    },
    {
      "entropy": 5.573259687423706,
      "epoch": 1.9566400580551524,
      "grad_norm": 1.0,
      "learning_rate": 0.0004617623283205091,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20604926198720933,
      "num_tokens": 44764277.0,
      "step": 21570
    },
    {
      "entropy": 5.6039666652679445,
      "epoch": 1.9570936139332367,
      "grad_norm": 1.125,
      "learning_rate": 0.00046174428180149733,
      "loss": 5.057,
      "mean_token_accuracy": 0.20239367187023163,
      "num_tokens": 44775429.0,
      "step": 21575
    },
    {
      "entropy": 5.613869333267212,
      "epoch": 1.9575471698113207,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004617262314205663,
      "loss": 5.0372,
      "mean_token_accuracy": 0.20427386611700057,
      "num_tokens": 44786390.0,
      "step": 21580
    },
    {
      "entropy": 5.583804893493652,
      "epoch": 1.958000725689405,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004617081771780893,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20702233165502548,
      "num_tokens": 44796473.0,
      "step": 21585
    },
    {
      "entropy": 5.519788503646851,
      "epoch": 1.958454281567489,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046169011907443944,
      "loss": 4.9429,
      "mean_token_accuracy": 0.20918683409690858,
      "num_tokens": 44806379.0,
      "step": 21590
    },
    {
      "entropy": 5.5597998142242435,
      "epoch": 1.9589078374455733,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004616720571099904,
      "loss": 5.0267,
      "mean_token_accuracy": 0.20871667712926864,
      "num_tokens": 44817659.0,
      "step": 21595
    },
    {
      "entropy": 5.581849050521851,
      "epoch": 1.9593613933236576,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004616539912851157,
      "loss": 4.9463,
      "mean_token_accuracy": 0.20679994970560073,
      "num_tokens": 44827073.0,
      "step": 21600
    },
    {
      "entropy": 5.507561159133911,
      "epoch": 1.9598149492017418,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046163592160018886,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20796812921762467,
      "num_tokens": 44837019.0,
      "step": 21605
    },
    {
      "entropy": 5.568103218078614,
      "epoch": 1.9602685050798259,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004616178480555836,
      "loss": 4.9602,
      "mean_token_accuracy": 0.21116245090961455,
      "num_tokens": 44846670.0,
      "step": 21610
    },
    {
      "entropy": 5.479678535461426,
      "epoch": 1.96072206095791,
      "grad_norm": 1.125,
      "learning_rate": 0.0004615997706516737,
      "loss": 4.9048,
      "mean_token_accuracy": 0.21223325580358504,
      "num_tokens": 44857016.0,
      "step": 21615
    },
    {
      "entropy": 5.618633842468261,
      "epoch": 1.9611756168359942,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000461581689388833,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19679954797029495,
      "num_tokens": 44868180.0,
      "step": 21620
    },
    {
      "entropy": 5.514395332336425,
      "epoch": 1.9616291727140784,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046156360426743536,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2127016082406044,
      "num_tokens": 44877864.0,
      "step": 21625
    },
    {
      "entropy": 5.584148025512695,
      "epoch": 1.9620827285921627,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004615455152878549,
      "loss": 4.9796,
      "mean_token_accuracy": 0.2177381232380867,
      "num_tokens": 44887737.0,
      "step": 21630
    },
    {
      "entropy": 5.482257509231568,
      "epoch": 1.9625362844702468,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004615274224504657,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20395839661359788,
      "num_tokens": 44898669.0,
      "step": 21635
    },
    {
      "entropy": 5.585608720779419,
      "epoch": 1.9629898403483308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046150932575564173,
      "loss": 5.0484,
      "mean_token_accuracy": 0.21015833765268327,
      "num_tokens": 44909074.0,
      "step": 21640
    },
    {
      "entropy": 5.535607528686524,
      "epoch": 1.963443396226415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046149122520375746,
      "loss": 4.9414,
      "mean_token_accuracy": 0.20969487279653548,
      "num_tokens": 44919468.0,
      "step": 21645
    },
    {
      "entropy": 5.601579999923706,
      "epoch": 1.9638969521044993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004614731207951871,
      "loss": 5.0555,
      "mean_token_accuracy": 0.19970683753490448,
      "num_tokens": 44930470.0,
      "step": 21650
    },
    {
      "entropy": 5.551943969726563,
      "epoch": 1.9643505079825836,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004614550125303051,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20635733604431153,
      "num_tokens": 44940611.0,
      "step": 21655
    },
    {
      "entropy": 5.551587009429932,
      "epoch": 1.9648040638606676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004614369004094859,
      "loss": 4.9948,
      "mean_token_accuracy": 0.20839242786169052,
      "num_tokens": 44950985.0,
      "step": 21660
    },
    {
      "entropy": 5.5767988681793215,
      "epoch": 1.9652576197387517,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046141878443310415,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20357128083705903,
      "num_tokens": 44961858.0,
      "step": 21665
    },
    {
      "entropy": 5.585680866241455,
      "epoch": 1.965711175616836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046140066460153444,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2094133287668228,
      "num_tokens": 44972048.0,
      "step": 21670
    },
    {
      "entropy": 5.550045347213745,
      "epoch": 1.9661647314949202,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046138254091515136,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2133963719010353,
      "num_tokens": 44981835.0,
      "step": 21675
    },
    {
      "entropy": 5.527277660369873,
      "epoch": 1.9666182873730045,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046136441337432996,
      "loss": 4.902,
      "mean_token_accuracy": 0.22058788537979127,
      "num_tokens": 44992108.0,
      "step": 21680
    },
    {
      "entropy": 5.657939195632935,
      "epoch": 1.9670718432510885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046134628197944496,
      "loss": 5.068,
      "mean_token_accuracy": 0.20214518755674363,
      "num_tokens": 45002105.0,
      "step": 21685
    },
    {
      "entropy": 5.52752161026001,
      "epoch": 1.9675253991291726,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046132814673087143,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20533880591392517,
      "num_tokens": 45013934.0,
      "step": 21690
    },
    {
      "entropy": 5.514544820785522,
      "epoch": 1.9679789550072568,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004613100076289842,
      "loss": 4.9806,
      "mean_token_accuracy": 0.2086989015340805,
      "num_tokens": 45023851.0,
      "step": 21695
    },
    {
      "entropy": 5.508292245864868,
      "epoch": 1.968432510885341,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004612918646741586,
      "loss": 4.9118,
      "mean_token_accuracy": 0.20739118456840516,
      "num_tokens": 45034140.0,
      "step": 21700
    },
    {
      "entropy": 5.541641807556152,
      "epoch": 1.9688860667634254,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004612737178667698,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20653848052024842,
      "num_tokens": 45044452.0,
      "step": 21705
    },
    {
      "entropy": 5.540272903442383,
      "epoch": 1.9693396226415094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046125556720719304,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2012146830558777,
      "num_tokens": 45053845.0,
      "step": 21710
    },
    {
      "entropy": 5.551776456832886,
      "epoch": 1.9697931785195935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004612374126958037,
      "loss": 4.9341,
      "mean_token_accuracy": 0.20701974779367446,
      "num_tokens": 45063973.0,
      "step": 21715
    },
    {
      "entropy": 5.608538961410522,
      "epoch": 1.9702467343976777,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004612192543329771,
      "loss": 5.0878,
      "mean_token_accuracy": 0.20605275481939317,
      "num_tokens": 45074879.0,
      "step": 21720
    },
    {
      "entropy": 5.523556232452393,
      "epoch": 1.970700290275762,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000461201092119089,
      "loss": 4.9229,
      "mean_token_accuracy": 0.21459337770938874,
      "num_tokens": 45085161.0,
      "step": 21725
    },
    {
      "entropy": 5.583820486068726,
      "epoch": 1.9711538461538463,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004611829260545148,
      "loss": 5.063,
      "mean_token_accuracy": 0.20150579512119293,
      "num_tokens": 45095862.0,
      "step": 21730
    },
    {
      "entropy": 5.590506267547608,
      "epoch": 1.9716074020319303,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046116475613963026,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2055190920829773,
      "num_tokens": 45106396.0,
      "step": 21735
    },
    {
      "entropy": 5.557551670074463,
      "epoch": 1.9720609579100146,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046114658237481114,
      "loss": 4.9051,
      "mean_token_accuracy": 0.21439230144023896,
      "num_tokens": 45117178.0,
      "step": 21740
    },
    {
      "entropy": 5.4734259128570555,
      "epoch": 1.9725145137880986,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004611284047604332,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21700416952371598,
      "num_tokens": 45126571.0,
      "step": 21745
    },
    {
      "entropy": 5.553644704818725,
      "epoch": 1.9729680696661829,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046111022329687246,
      "loss": 5.0282,
      "mean_token_accuracy": 0.21026648730039596,
      "num_tokens": 45137247.0,
      "step": 21750
    },
    {
      "entropy": 5.633781290054321,
      "epoch": 1.9734216255442671,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046109203798450483,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2073136553168297,
      "num_tokens": 45147531.0,
      "step": 21755
    },
    {
      "entropy": 5.5364891529083256,
      "epoch": 1.9738751814223512,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004610738488237065,
      "loss": 4.9597,
      "mean_token_accuracy": 0.2094478502869606,
      "num_tokens": 45158312.0,
      "step": 21760
    },
    {
      "entropy": 5.513592100143432,
      "epoch": 1.9743287373004355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004610556558148535,
      "loss": 5.0161,
      "mean_token_accuracy": 0.2071317106485367,
      "num_tokens": 45169754.0,
      "step": 21765
    },
    {
      "entropy": 5.587693548202514,
      "epoch": 1.9747822931785195,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004610374589583222,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20725469440221786,
      "num_tokens": 45180428.0,
      "step": 21770
    },
    {
      "entropy": 5.55584135055542,
      "epoch": 1.9752358490566038,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004610192582544888,
      "loss": 4.9517,
      "mean_token_accuracy": 0.2135287344455719,
      "num_tokens": 45191558.0,
      "step": 21775
    },
    {
      "entropy": 5.486757230758667,
      "epoch": 1.975689404934688,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004610010537037297,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21148865520954133,
      "num_tokens": 45201567.0,
      "step": 21780
    },
    {
      "entropy": 5.618310594558716,
      "epoch": 1.9761429608127723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004609828453064214,
      "loss": 5.0631,
      "mean_token_accuracy": 0.20185273587703706,
      "num_tokens": 45211359.0,
      "step": 21785
    },
    {
      "entropy": 5.617689752578736,
      "epoch": 1.9765965166908563,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046096463306294057,
      "loss": 4.98,
      "mean_token_accuracy": 0.20351368337869644,
      "num_tokens": 45222106.0,
      "step": 21790
    },
    {
      "entropy": 5.480583572387696,
      "epoch": 1.9770500725689404,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046094641697366366,
      "loss": 4.9967,
      "mean_token_accuracy": 0.20901469141244888,
      "num_tokens": 45233169.0,
      "step": 21795
    },
    {
      "entropy": 5.605236577987671,
      "epoch": 1.9775036284470247,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004609281970389675,
      "loss": 5.064,
      "mean_token_accuracy": 0.19933681339025497,
      "num_tokens": 45243627.0,
      "step": 21800
    },
    {
      "entropy": 5.594206094741821,
      "epoch": 1.977957184325109,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046090997325922887,
      "loss": 4.922,
      "mean_token_accuracy": 0.21322551965713502,
      "num_tokens": 45253730.0,
      "step": 21805
    },
    {
      "entropy": 5.533895969390869,
      "epoch": 1.9784107402031932,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046089174563482456,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20635027885437013,
      "num_tokens": 45263443.0,
      "step": 21810
    },
    {
      "entropy": 5.496335220336914,
      "epoch": 1.9788642960812772,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004608735141661316,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21626696288585662,
      "num_tokens": 45275298.0,
      "step": 21815
    },
    {
      "entropy": 5.553706550598145,
      "epoch": 1.9793178519593613,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046085527885352694,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20689027458429338,
      "num_tokens": 45285628.0,
      "step": 21820
    },
    {
      "entropy": 5.535992097854614,
      "epoch": 1.9797714078374455,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004608370396973878,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21273459494113922,
      "num_tokens": 45295559.0,
      "step": 21825
    },
    {
      "entropy": 5.507489633560181,
      "epoch": 1.9802249637155298,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046081879669809133,
      "loss": 4.9325,
      "mean_token_accuracy": 0.20965660661458968,
      "num_tokens": 45305395.0,
      "step": 21830
    },
    {
      "entropy": 5.523382043838501,
      "epoch": 1.980678519593614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004608005498560148,
      "loss": 4.9574,
      "mean_token_accuracy": 0.20966893136501313,
      "num_tokens": 45316763.0,
      "step": 21835
    },
    {
      "entropy": 5.581750822067261,
      "epoch": 1.9811320754716981,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004607822991715356,
      "loss": 5.095,
      "mean_token_accuracy": 0.20398128926753997,
      "num_tokens": 45326484.0,
      "step": 21840
    },
    {
      "entropy": 5.575086307525635,
      "epoch": 1.9815856313497822,
      "grad_norm": 1.21875,
      "learning_rate": 0.000460764044645031,
      "loss": 5.0174,
      "mean_token_accuracy": 0.2053569659590721,
      "num_tokens": 45335792.0,
      "step": 21845
    },
    {
      "entropy": 5.5670332431793215,
      "epoch": 1.9820391872278664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004607457862768787,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2041582301259041,
      "num_tokens": 45347247.0,
      "step": 21850
    },
    {
      "entropy": 5.608470106124878,
      "epoch": 1.9824927431059507,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004607275240674562,
      "loss": 5.063,
      "mean_token_accuracy": 0.20145944505929947,
      "num_tokens": 45357994.0,
      "step": 21855
    },
    {
      "entropy": 5.502914667129517,
      "epoch": 1.982946298984035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004607092580171411,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20760190933942796,
      "num_tokens": 45368170.0,
      "step": 21860
    },
    {
      "entropy": 5.592855215072632,
      "epoch": 1.983399854862119,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004606909881263113,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20415429025888443,
      "num_tokens": 45378696.0,
      "step": 21865
    },
    {
      "entropy": 5.6079676151275635,
      "epoch": 1.983853410740203,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046067271439534445,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20656584650278093,
      "num_tokens": 45388578.0,
      "step": 21870
    },
    {
      "entropy": 5.5629298210144045,
      "epoch": 1.9843069666182873,
      "grad_norm": 1.125,
      "learning_rate": 0.00046065443682461864,
      "loss": 5.0487,
      "mean_token_accuracy": 0.20561712235212326,
      "num_tokens": 45398835.0,
      "step": 21875
    },
    {
      "entropy": 5.595419931411743,
      "epoch": 1.9847605224963716,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046063615541451177,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20945125967264175,
      "num_tokens": 45409032.0,
      "step": 21880
    },
    {
      "entropy": 5.625519227981568,
      "epoch": 1.9852140783744558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046061787016540184,
      "loss": 5.1106,
      "mean_token_accuracy": 0.1957710400223732,
      "num_tokens": 45420661.0,
      "step": 21885
    },
    {
      "entropy": 5.5140228271484375,
      "epoch": 1.98566763425254,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046059958107766706,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22318844944238664,
      "num_tokens": 45429751.0,
      "step": 21890
    },
    {
      "entropy": 5.571798849105835,
      "epoch": 1.986121190130624,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004605812881516857,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2143436923623085,
      "num_tokens": 45439551.0,
      "step": 21895
    },
    {
      "entropy": 5.482573986053467,
      "epoch": 1.9865747460087082,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046056299138783595,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22369229346513747,
      "num_tokens": 45449543.0,
      "step": 21900
    },
    {
      "entropy": 5.5848533630371096,
      "epoch": 1.9870283018867925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004605446907864963,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20833965241909028,
      "num_tokens": 45459069.0,
      "step": 21905
    },
    {
      "entropy": 5.585950422286987,
      "epoch": 1.9874818577648767,
      "grad_norm": 1.0,
      "learning_rate": 0.0004605263863480452,
      "loss": 5.074,
      "mean_token_accuracy": 0.20062010884284973,
      "num_tokens": 45468987.0,
      "step": 21910
    },
    {
      "entropy": 5.604973936080933,
      "epoch": 1.9879354136429608,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046050807807286105,
      "loss": 4.9927,
      "mean_token_accuracy": 0.20938576012849808,
      "num_tokens": 45479781.0,
      "step": 21915
    },
    {
      "entropy": 5.662037038803101,
      "epoch": 1.988388969521045,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046048976596132265,
      "loss": 5.1257,
      "mean_token_accuracy": 0.19111909568309784,
      "num_tokens": 45490318.0,
      "step": 21920
    },
    {
      "entropy": 5.551020193099975,
      "epoch": 1.988842525399129,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046047145001380856,
      "loss": 4.9609,
      "mean_token_accuracy": 0.20929338186979293,
      "num_tokens": 45501514.0,
      "step": 21925
    },
    {
      "entropy": 5.570407772064209,
      "epoch": 1.9892960812772134,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046045313023069764,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20886309146881105,
      "num_tokens": 45511727.0,
      "step": 21930
    },
    {
      "entropy": 5.60915699005127,
      "epoch": 1.9897496371552976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004604348066123688,
      "loss": 5.0612,
      "mean_token_accuracy": 0.20543305873870848,
      "num_tokens": 45521807.0,
      "step": 21935
    },
    {
      "entropy": 5.625822973251343,
      "epoch": 1.9902031930333817,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004604164791592008,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19708407819271087,
      "num_tokens": 45532033.0,
      "step": 21940
    },
    {
      "entropy": 5.567453861236572,
      "epoch": 1.990656748911466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046039814787157283,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20609915554523467,
      "num_tokens": 45542806.0,
      "step": 21945
    },
    {
      "entropy": 5.53896164894104,
      "epoch": 1.99111030478955,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004603798127498639,
      "loss": 4.89,
      "mean_token_accuracy": 0.22646187990903854,
      "num_tokens": 45551982.0,
      "step": 21950
    },
    {
      "entropy": 5.554815864562988,
      "epoch": 1.9915638606676342,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004603614737944532,
      "loss": 5.002,
      "mean_token_accuracy": 0.20761286318302155,
      "num_tokens": 45561998.0,
      "step": 21955
    },
    {
      "entropy": 5.531924295425415,
      "epoch": 1.9920174165457185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046034313100571994,
      "loss": 5.0012,
      "mean_token_accuracy": 0.20836948603391647,
      "num_tokens": 45572246.0,
      "step": 21960
    },
    {
      "entropy": 5.587581634521484,
      "epoch": 1.9924709724238028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004603247843840436,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20864470452070236,
      "num_tokens": 45582251.0,
      "step": 21965
    },
    {
      "entropy": 5.560734987258911,
      "epoch": 1.9929245283018868,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004603064339298034,
      "loss": 4.9069,
      "mean_token_accuracy": 0.20998208373785018,
      "num_tokens": 45592749.0,
      "step": 21970
    },
    {
      "entropy": 5.526675510406494,
      "epoch": 1.9933780841799709,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004602880796433789,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21031934171915054,
      "num_tokens": 45602412.0,
      "step": 21975
    },
    {
      "entropy": 5.597551918029785,
      "epoch": 1.9938316400580551,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046026972152514976,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20619491785764693,
      "num_tokens": 45612736.0,
      "step": 21980
    },
    {
      "entropy": 5.570060729980469,
      "epoch": 1.9942851959361394,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004602513595754956,
      "loss": 4.9978,
      "mean_token_accuracy": 0.20696091502904893,
      "num_tokens": 45622952.0,
      "step": 21985
    },
    {
      "entropy": 5.615493488311768,
      "epoch": 1.9947387518142237,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046023299379479606,
      "loss": 5.0068,
      "mean_token_accuracy": 0.20682632327079772,
      "num_tokens": 45633569.0,
      "step": 21990
    },
    {
      "entropy": 5.508515644073486,
      "epoch": 1.9951923076923077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046021462418343104,
      "loss": 4.9777,
      "mean_token_accuracy": 0.21096506118774414,
      "num_tokens": 45644347.0,
      "step": 21995
    },
    {
      "entropy": 5.6098042011260985,
      "epoch": 1.9956458635703918,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004601962507417804,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20653981119394302,
      "num_tokens": 45655268.0,
      "step": 22000
    },
    {
      "entropy": 5.545614004135132,
      "epoch": 1.996099419448476,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046017787347022403,
      "loss": 4.9611,
      "mean_token_accuracy": 0.20893973112106323,
      "num_tokens": 45664900.0,
      "step": 22005
    },
    {
      "entropy": 5.592525959014893,
      "epoch": 1.9965529753265603,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046015949236914217,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21572151631116868,
      "num_tokens": 45675252.0,
      "step": 22010
    },
    {
      "entropy": 5.647371912002564,
      "epoch": 1.9970065312046446,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004601411074389147,
      "loss": 4.9595,
      "mean_token_accuracy": 0.2131752163171768,
      "num_tokens": 45686027.0,
      "step": 22015
    },
    {
      "entropy": 5.564004421234131,
      "epoch": 1.9974600870827286,
      "grad_norm": 1.125,
      "learning_rate": 0.000460122718679922,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20969758331775665,
      "num_tokens": 45696436.0,
      "step": 22020
    },
    {
      "entropy": 5.594499158859253,
      "epoch": 1.9979136429608126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004601043260925443,
      "loss": 5.0641,
      "mean_token_accuracy": 0.21029372215270997,
      "num_tokens": 45706509.0,
      "step": 22025
    },
    {
      "entropy": 5.629207944869995,
      "epoch": 1.998367198838897,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046008592967716197,
      "loss": 5.0873,
      "mean_token_accuracy": 0.20248838514089584,
      "num_tokens": 45717933.0,
      "step": 22030
    },
    {
      "entropy": 5.560246419906616,
      "epoch": 1.9988207547169812,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004600675294341554,
      "loss": 5.0413,
      "mean_token_accuracy": 0.19791806489229202,
      "num_tokens": 45727966.0,
      "step": 22035
    },
    {
      "entropy": 5.554030799865723,
      "epoch": 1.9992743105950654,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046004912536390526,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20639860033988952,
      "num_tokens": 45737649.0,
      "step": 22040
    },
    {
      "entropy": 5.598674964904785,
      "epoch": 1.9997278664731495,
      "grad_norm": 1.15625,
      "learning_rate": 0.000460030717466792,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2102483779191971,
      "num_tokens": 45747561.0,
      "step": 22045
    },
    {
      "entropy": 5.536699819564819,
      "epoch": 2.0001814223512335,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004600123057431963,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21040636897087098,
      "num_tokens": 45758183.0,
      "step": 22050
    },
    {
      "entropy": 5.535328960418701,
      "epoch": 2.000634978229318,
      "grad_norm": 1.171875,
      "learning_rate": 0.000459993890193499,
      "loss": 4.8764,
      "mean_token_accuracy": 0.218867327272892,
      "num_tokens": 45768006.0,
      "step": 22055
    },
    {
      "entropy": 5.5945357322692875,
      "epoch": 2.001088534107402,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004599754708180809,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20445465445518493,
      "num_tokens": 45777981.0,
      "step": 22060
    },
    {
      "entropy": 5.567545747756958,
      "epoch": 2.0015420899854863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000459957047617323,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2058892861008644,
      "num_tokens": 45789083.0,
      "step": 22065
    },
    {
      "entropy": 5.577370500564575,
      "epoch": 2.0019956458635706,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045993862059160617,
      "loss": 4.8965,
      "mean_token_accuracy": 0.20997656732797623,
      "num_tokens": 45799750.0,
      "step": 22070
    },
    {
      "entropy": 5.539130353927613,
      "epoch": 2.0024492017416544,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004599201897413116,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21443042457103728,
      "num_tokens": 45810321.0,
      "step": 22075
    },
    {
      "entropy": 5.4724390506744385,
      "epoch": 2.0029027576197387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045990175506682024,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22037743180990219,
      "num_tokens": 45820581.0,
      "step": 22080
    },
    {
      "entropy": 5.513353395462036,
      "epoch": 2.003356313497823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045988331656851355,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20550288558006286,
      "num_tokens": 45831504.0,
      "step": 22085
    },
    {
      "entropy": 5.612738418579101,
      "epoch": 2.003809869375907,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045986487424677284,
      "loss": 4.991,
      "mean_token_accuracy": 0.2069953367114067,
      "num_tokens": 45841100.0,
      "step": 22090
    },
    {
      "entropy": 5.600549650192261,
      "epoch": 2.0042634252539915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004598464281019793,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21853204518556596,
      "num_tokens": 45851647.0,
      "step": 22095
    },
    {
      "entropy": 5.486761808395386,
      "epoch": 2.0047169811320753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045982797813451463,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22009832113981248,
      "num_tokens": 45862995.0,
      "step": 22100
    },
    {
      "entropy": 5.495511436462403,
      "epoch": 2.0051705370101596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045980952434476014,
      "loss": 4.8874,
      "mean_token_accuracy": 0.21044859141111374,
      "num_tokens": 45872360.0,
      "step": 22105
    },
    {
      "entropy": 5.545560312271118,
      "epoch": 2.005624092888244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045979106673309765,
      "loss": 4.9686,
      "mean_token_accuracy": 0.21053716242313386,
      "num_tokens": 45883440.0,
      "step": 22110
    },
    {
      "entropy": 5.516869068145752,
      "epoch": 2.006077648766328,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045977260529990883,
      "loss": 4.9391,
      "mean_token_accuracy": 0.20772037357091905,
      "num_tokens": 45894323.0,
      "step": 22115
    },
    {
      "entropy": 5.561671257019043,
      "epoch": 2.0065312046444124,
      "grad_norm": 1.125,
      "learning_rate": 0.0004597541400455755,
      "loss": 4.9718,
      "mean_token_accuracy": 0.20316219478845596,
      "num_tokens": 45904544.0,
      "step": 22120
    },
    {
      "entropy": 5.622522163391113,
      "epoch": 2.006984760522496,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045973567097047933,
      "loss": 4.8478,
      "mean_token_accuracy": 0.23042139708995818,
      "num_tokens": 45913225.0,
      "step": 22125
    },
    {
      "entropy": 5.55759711265564,
      "epoch": 2.0074383164005805,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045971719807500253,
      "loss": 4.9422,
      "mean_token_accuracy": 0.21341023445129395,
      "num_tokens": 45923306.0,
      "step": 22130
    },
    {
      "entropy": 5.517950534820557,
      "epoch": 2.0078918722786647,
      "grad_norm": 1.078125,
      "learning_rate": 0.000459698721359527,
      "loss": 4.921,
      "mean_token_accuracy": 0.2100364997982979,
      "num_tokens": 45933243.0,
      "step": 22135
    },
    {
      "entropy": 5.517447280883789,
      "epoch": 2.008345428156749,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004596802408244348,
      "loss": 4.9912,
      "mean_token_accuracy": 0.21013824194669722,
      "num_tokens": 45944498.0,
      "step": 22140
    },
    {
      "entropy": 5.495066976547241,
      "epoch": 2.0087989840348333,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004596617564701081,
      "loss": 4.8275,
      "mean_token_accuracy": 0.2200333520770073,
      "num_tokens": 45954550.0,
      "step": 22145
    },
    {
      "entropy": 5.603550148010254,
      "epoch": 2.009252539912917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045964326829692925,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20771233290433883,
      "num_tokens": 45965425.0,
      "step": 22150
    },
    {
      "entropy": 5.553044176101684,
      "epoch": 2.0097060957910013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045962477630528064,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22131365537643433,
      "num_tokens": 45975637.0,
      "step": 22155
    },
    {
      "entropy": 5.54318413734436,
      "epoch": 2.0101596516690856,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045960628049554453,
      "loss": 4.9658,
      "mean_token_accuracy": 0.2027548834681511,
      "num_tokens": 45985268.0,
      "step": 22160
    },
    {
      "entropy": 5.509607362747192,
      "epoch": 2.01061320754717,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045958778086810354,
      "loss": 4.8561,
      "mean_token_accuracy": 0.21449377685785292,
      "num_tokens": 45995335.0,
      "step": 22165
    },
    {
      "entropy": 5.540615558624268,
      "epoch": 2.011066763425254,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004595692774233402,
      "loss": 4.9489,
      "mean_token_accuracy": 0.20536104440689087,
      "num_tokens": 46006037.0,
      "step": 22170
    },
    {
      "entropy": 5.570735120773316,
      "epoch": 2.011520319303338,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045955077016163713,
      "loss": 4.8842,
      "mean_token_accuracy": 0.21552205085754395,
      "num_tokens": 46017016.0,
      "step": 22175
    },
    {
      "entropy": 5.614500570297241,
      "epoch": 2.0119738751814222,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045953225908337717,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20326677411794664,
      "num_tokens": 46026797.0,
      "step": 22180
    },
    {
      "entropy": 5.510417938232422,
      "epoch": 2.0124274310595065,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045951374418894304,
      "loss": 4.848,
      "mean_token_accuracy": 0.2150076538324356,
      "num_tokens": 46037799.0,
      "step": 22185
    },
    {
      "entropy": 5.550413417816162,
      "epoch": 2.0128809869375908,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045949522547871765,
      "loss": 4.8682,
      "mean_token_accuracy": 0.2123866006731987,
      "num_tokens": 46048617.0,
      "step": 22190
    },
    {
      "entropy": 5.554560995101928,
      "epoch": 2.013334542815675,
      "grad_norm": 1.0625,
      "learning_rate": 0.000459476702953084,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21581512242555617,
      "num_tokens": 46060122.0,
      "step": 22195
    },
    {
      "entropy": 5.525198078155517,
      "epoch": 2.013788098693759,
      "grad_norm": 1.046875,
      "learning_rate": 0.000459458176612425,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21504294276237487,
      "num_tokens": 46071488.0,
      "step": 22200
    },
    {
      "entropy": 5.601229619979859,
      "epoch": 2.014241654571843,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045943964645712407,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20384662449359894,
      "num_tokens": 46081954.0,
      "step": 22205
    },
    {
      "entropy": 5.561482715606689,
      "epoch": 2.0146952104499274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004594211124875642,
      "loss": 4.8906,
      "mean_token_accuracy": 0.21491264104843139,
      "num_tokens": 46092488.0,
      "step": 22210
    },
    {
      "entropy": 5.512816715240478,
      "epoch": 2.0151487663280117,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004594025747041287,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22213494777679443,
      "num_tokens": 46102451.0,
      "step": 22215
    },
    {
      "entropy": 5.5255286693573,
      "epoch": 2.015602322206096,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000459384033107201,
      "loss": 4.9278,
      "mean_token_accuracy": 0.21453291177749634,
      "num_tokens": 46113862.0,
      "step": 22220
    },
    {
      "entropy": 5.624523496627807,
      "epoch": 2.01605587808418,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004593654876971645,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20698038190603257,
      "num_tokens": 46124517.0,
      "step": 22225
    },
    {
      "entropy": 5.494170188903809,
      "epoch": 2.016509433962264,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004593469384744027,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21207363903522491,
      "num_tokens": 46134221.0,
      "step": 22230
    },
    {
      "entropy": 5.55197434425354,
      "epoch": 2.0169629898403483,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045932838543929924,
      "loss": 4.999,
      "mean_token_accuracy": 0.2088332384824753,
      "num_tokens": 46144512.0,
      "step": 22235
    },
    {
      "entropy": 5.592677450180053,
      "epoch": 2.0174165457184325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004593098285922378,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21223630905151367,
      "num_tokens": 46154373.0,
      "step": 22240
    },
    {
      "entropy": 5.619633913040161,
      "epoch": 2.017870101596517,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004592912679336021,
      "loss": 5.021,
      "mean_token_accuracy": 0.20819658786058426,
      "num_tokens": 46164834.0,
      "step": 22245
    },
    {
      "entropy": 5.5458661079406735,
      "epoch": 2.018323657474601,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004592727034637761,
      "loss": 4.8812,
      "mean_token_accuracy": 0.2142745301127434,
      "num_tokens": 46174815.0,
      "step": 22250
    },
    {
      "entropy": 5.522609376907349,
      "epoch": 2.018777213352685,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045925413518314363,
      "loss": 4.9128,
      "mean_token_accuracy": 0.2084043338894844,
      "num_tokens": 46185155.0,
      "step": 22255
    },
    {
      "entropy": 5.575503015518189,
      "epoch": 2.019230769230769,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004592355630920886,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21316245347261428,
      "num_tokens": 46194537.0,
      "step": 22260
    },
    {
      "entropy": 5.62617506980896,
      "epoch": 2.0196843251088534,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045921698719099527,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20964009761810304,
      "num_tokens": 46205494.0,
      "step": 22265
    },
    {
      "entropy": 5.507011365890503,
      "epoch": 2.0201378809869377,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045919840748024764,
      "loss": 4.8962,
      "mean_token_accuracy": 0.21459405422210692,
      "num_tokens": 46216399.0,
      "step": 22270
    },
    {
      "entropy": 5.4886754035949705,
      "epoch": 2.020591436865022,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045917982396023,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2169172540307045,
      "num_tokens": 46226515.0,
      "step": 22275
    },
    {
      "entropy": 5.583468389511109,
      "epoch": 2.021044992743106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045916123663132676,
      "loss": 4.9659,
      "mean_token_accuracy": 0.20524397492408752,
      "num_tokens": 46236688.0,
      "step": 22280
    },
    {
      "entropy": 5.549386024475098,
      "epoch": 2.02149854862119,
      "grad_norm": 1.125,
      "learning_rate": 0.00045914264549392215,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2174192637205124,
      "num_tokens": 46247166.0,
      "step": 22285
    },
    {
      "entropy": 5.583451938629151,
      "epoch": 2.0219521044992743,
      "grad_norm": 1.125,
      "learning_rate": 0.0004591240505484007,
      "loss": 4.9168,
      "mean_token_accuracy": 0.2175706371665001,
      "num_tokens": 46258902.0,
      "step": 22290
    },
    {
      "entropy": 5.581418418884278,
      "epoch": 2.0224056603773586,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045910545179514694,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20716777294874192,
      "num_tokens": 46269819.0,
      "step": 22295
    },
    {
      "entropy": 5.639160013198852,
      "epoch": 2.022859216255443,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004590868492345455,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2072971522808075,
      "num_tokens": 46280715.0,
      "step": 22300
    },
    {
      "entropy": 5.602681875228882,
      "epoch": 2.0233127721335267,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004590682428669813,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2026600271463394,
      "num_tokens": 46291079.0,
      "step": 22305
    },
    {
      "entropy": 5.590440082550049,
      "epoch": 2.023766328011611,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004590496326928387,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21221930533647537,
      "num_tokens": 46301362.0,
      "step": 22310
    },
    {
      "entropy": 5.5843523979187015,
      "epoch": 2.024219883889695,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045903101871250295,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21649962663650513,
      "num_tokens": 46311820.0,
      "step": 22315
    },
    {
      "entropy": 5.631623840332031,
      "epoch": 2.0246734397677795,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004590124009263587,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20982985496520995,
      "num_tokens": 46321992.0,
      "step": 22320
    },
    {
      "entropy": 5.550974369049072,
      "epoch": 2.0251269956458637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004589937793347912,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20854606479406357,
      "num_tokens": 46331537.0,
      "step": 22325
    },
    {
      "entropy": 5.574235677719116,
      "epoch": 2.0255805515239476,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045897515393818547,
      "loss": 4.9232,
      "mean_token_accuracy": 0.2099696949124336,
      "num_tokens": 46342214.0,
      "step": 22330
    },
    {
      "entropy": 5.5319356441497805,
      "epoch": 2.026034107402032,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004589565247369266,
      "loss": 4.9246,
      "mean_token_accuracy": 0.20641277283430098,
      "num_tokens": 46352471.0,
      "step": 22335
    },
    {
      "entropy": 5.530709600448608,
      "epoch": 2.026487663280116,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004589378917314,
      "loss": 4.949,
      "mean_token_accuracy": 0.21187900006771088,
      "num_tokens": 46363221.0,
      "step": 22340
    },
    {
      "entropy": 5.510725355148315,
      "epoch": 2.0269412191582004,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045891925492199087,
      "loss": 4.8199,
      "mean_token_accuracy": 0.22246098071336745,
      "num_tokens": 46373454.0,
      "step": 22345
    },
    {
      "entropy": 5.676409912109375,
      "epoch": 2.0273947750362846,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004589006143090847,
      "loss": 5.0993,
      "mean_token_accuracy": 0.2050787091255188,
      "num_tokens": 46384403.0,
      "step": 22350
    },
    {
      "entropy": 5.518602800369263,
      "epoch": 2.0278483309143684,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045888196989306697,
      "loss": 4.8028,
      "mean_token_accuracy": 0.22396107017993927,
      "num_tokens": 46393362.0,
      "step": 22355
    },
    {
      "entropy": 5.625599002838134,
      "epoch": 2.0283018867924527,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004588633216743232,
      "loss": 5.0268,
      "mean_token_accuracy": 0.21074841171503067,
      "num_tokens": 46405100.0,
      "step": 22360
    },
    {
      "entropy": 5.484832143783569,
      "epoch": 2.028755442670537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045884466965323914,
      "loss": 4.86,
      "mean_token_accuracy": 0.21490903645753862,
      "num_tokens": 46415058.0,
      "step": 22365
    },
    {
      "entropy": 5.550878620147705,
      "epoch": 2.0292089985486212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045882601383020045,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21324272751808165,
      "num_tokens": 46425863.0,
      "step": 22370
    },
    {
      "entropy": 5.562830114364624,
      "epoch": 2.0296625544267055,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004588073542055929,
      "loss": 4.9422,
      "mean_token_accuracy": 0.21125178188085555,
      "num_tokens": 46436172.0,
      "step": 22375
    },
    {
      "entropy": 5.621502637863159,
      "epoch": 2.0301161103047893,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045878869077980246,
      "loss": 4.9414,
      "mean_token_accuracy": 0.20946284085512162,
      "num_tokens": 46446717.0,
      "step": 22380
    },
    {
      "entropy": 5.596818637847901,
      "epoch": 2.0305696661828736,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045877002355321504,
      "loss": 4.9694,
      "mean_token_accuracy": 0.20303172022104263,
      "num_tokens": 46456944.0,
      "step": 22385
    },
    {
      "entropy": 5.57465672492981,
      "epoch": 2.031023222060958,
      "grad_norm": 1.125,
      "learning_rate": 0.0004587513525262168,
      "loss": 5.0337,
      "mean_token_accuracy": 0.2074401006102562,
      "num_tokens": 46467231.0,
      "step": 22390
    },
    {
      "entropy": 5.548603296279907,
      "epoch": 2.031476777939042,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045873267769919364,
      "loss": 4.8856,
      "mean_token_accuracy": 0.21052717566490173,
      "num_tokens": 46477107.0,
      "step": 22395
    },
    {
      "entropy": 5.531497287750244,
      "epoch": 2.0319303338171264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004587139990725319,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21285859644412994,
      "num_tokens": 46487351.0,
      "step": 22400
    },
    {
      "entropy": 5.566560363769531,
      "epoch": 2.0323838896952107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004586953166466179,
      "loss": 4.956,
      "mean_token_accuracy": 0.2106267601251602,
      "num_tokens": 46497685.0,
      "step": 22405
    },
    {
      "entropy": 5.556187582015991,
      "epoch": 2.0328374455732945,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004586766304218379,
      "loss": 4.9533,
      "mean_token_accuracy": 0.20468230098485946,
      "num_tokens": 46507866.0,
      "step": 22410
    },
    {
      "entropy": 5.490482902526855,
      "epoch": 2.0332910014513788,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045865794039857846,
      "loss": 4.8451,
      "mean_token_accuracy": 0.21900141686201097,
      "num_tokens": 46518797.0,
      "step": 22415
    },
    {
      "entropy": 5.539396524429321,
      "epoch": 2.033744557329463,
      "grad_norm": 1.125,
      "learning_rate": 0.0004586392465772259,
      "loss": 4.906,
      "mean_token_accuracy": 0.21572514176368712,
      "num_tokens": 46528759.0,
      "step": 22420
    },
    {
      "entropy": 5.615487670898437,
      "epoch": 2.0341981132075473,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045862054895816703,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20383737683296205,
      "num_tokens": 46539014.0,
      "step": 22425
    },
    {
      "entropy": 5.518242406845093,
      "epoch": 2.0346516690856316,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004586018475417884,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21100321561098098,
      "num_tokens": 46548856.0,
      "step": 22430
    },
    {
      "entropy": 5.5146091938018795,
      "epoch": 2.0351052249637154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004585831423284766,
      "loss": 4.8816,
      "mean_token_accuracy": 0.2158732146024704,
      "num_tokens": 46559164.0,
      "step": 22435
    },
    {
      "entropy": 5.5795269966125485,
      "epoch": 2.0355587808417996,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004585644333186188,
      "loss": 4.9857,
      "mean_token_accuracy": 0.20651134997606277,
      "num_tokens": 46569793.0,
      "step": 22440
    },
    {
      "entropy": 5.577578020095825,
      "epoch": 2.036012336719884,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045854572051260166,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21044086813926696,
      "num_tokens": 46580257.0,
      "step": 22445
    },
    {
      "entropy": 5.554258012771607,
      "epoch": 2.036465892597968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045852700391081234,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2066876173019409,
      "num_tokens": 46590297.0,
      "step": 22450
    },
    {
      "entropy": 5.534800577163696,
      "epoch": 2.0369194484760524,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004585082835136378,
      "loss": 4.8971,
      "mean_token_accuracy": 0.21658104062080383,
      "num_tokens": 46600647.0,
      "step": 22455
    },
    {
      "entropy": 5.5505164623260494,
      "epoch": 2.0373730043541363,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004584895593214652,
      "loss": 4.9035,
      "mean_token_accuracy": 0.20982402712106704,
      "num_tokens": 46610478.0,
      "step": 22460
    },
    {
      "entropy": 5.550243902206421,
      "epoch": 2.0378265602322205,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045847083133468164,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20916029214859008,
      "num_tokens": 46620159.0,
      "step": 22465
    },
    {
      "entropy": 5.575983047485352,
      "epoch": 2.038280116110305,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045845209955367464,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21891459673643113,
      "num_tokens": 46630555.0,
      "step": 22470
    },
    {
      "entropy": 5.559483385086059,
      "epoch": 2.038733671988389,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045843336397883146,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20618917942047119,
      "num_tokens": 46640188.0,
      "step": 22475
    },
    {
      "entropy": 5.5388209342956545,
      "epoch": 2.0391872278664733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004584146246105396,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2146706461906433,
      "num_tokens": 46650284.0,
      "step": 22480
    },
    {
      "entropy": 5.616133403778076,
      "epoch": 2.039640783744557,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045839588144918643,
      "loss": 4.9239,
      "mean_token_accuracy": 0.213815775513649,
      "num_tokens": 46661466.0,
      "step": 22485
    },
    {
      "entropy": 5.587265300750732,
      "epoch": 2.0400943396226414,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004583771344951598,
      "loss": 4.9139,
      "mean_token_accuracy": 0.2136480137705803,
      "num_tokens": 46671683.0,
      "step": 22490
    },
    {
      "entropy": 5.5348492622375485,
      "epoch": 2.0405478955007257,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045835838374884733,
      "loss": 4.967,
      "mean_token_accuracy": 0.20732931941747665,
      "num_tokens": 46681787.0,
      "step": 22495
    },
    {
      "entropy": 5.57278995513916,
      "epoch": 2.04100145137881,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045833962921063665,
      "loss": 4.9065,
      "mean_token_accuracy": 0.21543323397636413,
      "num_tokens": 46692752.0,
      "step": 22500
    },
    {
      "entropy": 5.590698528289795,
      "epoch": 2.041455007256894,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045832087088091574,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2086937367916107,
      "num_tokens": 46702980.0,
      "step": 22505
    },
    {
      "entropy": 5.530666971206665,
      "epoch": 2.041908563134978,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045830210876007257,
      "loss": 4.8675,
      "mean_token_accuracy": 0.21580173969268798,
      "num_tokens": 46712281.0,
      "step": 22510
    },
    {
      "entropy": 5.538294315338135,
      "epoch": 2.0423621190130623,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000458283342848495,
      "loss": 4.9045,
      "mean_token_accuracy": 0.21018820703029634,
      "num_tokens": 46722097.0,
      "step": 22515
    },
    {
      "entropy": 5.492975091934204,
      "epoch": 2.0428156748911466,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004582645731465713,
      "loss": 4.8293,
      "mean_token_accuracy": 0.21653183400630951,
      "num_tokens": 46733056.0,
      "step": 22520
    },
    {
      "entropy": 5.4882568359375,
      "epoch": 2.043269230769231,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045824579965468944,
      "loss": 4.884,
      "mean_token_accuracy": 0.21330133825540543,
      "num_tokens": 46742608.0,
      "step": 22525
    },
    {
      "entropy": 5.492119073867798,
      "epoch": 2.043722786647315,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045822702237323777,
      "loss": 4.8689,
      "mean_token_accuracy": 0.22310970425605775,
      "num_tokens": 46751762.0,
      "step": 22530
    },
    {
      "entropy": 5.488946914672852,
      "epoch": 2.044176342525399,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045820824130260457,
      "loss": 4.8463,
      "mean_token_accuracy": 0.2245909482240677,
      "num_tokens": 46762252.0,
      "step": 22535
    },
    {
      "entropy": 5.533684301376343,
      "epoch": 2.044629898403483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004581894564431783,
      "loss": 4.9454,
      "mean_token_accuracy": 0.20734742134809495,
      "num_tokens": 46772172.0,
      "step": 22540
    },
    {
      "entropy": 5.528283929824829,
      "epoch": 2.0450834542815675,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004581706677953473,
      "loss": 4.9283,
      "mean_token_accuracy": 0.21463873088359833,
      "num_tokens": 46782894.0,
      "step": 22545
    },
    {
      "entropy": 5.561836528778076,
      "epoch": 2.0455370101596517,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045815187535950034,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21210812479257585,
      "num_tokens": 46793097.0,
      "step": 22550
    },
    {
      "entropy": 5.611650180816651,
      "epoch": 2.045990566037736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045813307913602586,
      "loss": 5.0305,
      "mean_token_accuracy": 0.200439915060997,
      "num_tokens": 46804196.0,
      "step": 22555
    },
    {
      "entropy": 5.569736814498901,
      "epoch": 2.04644412191582,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004581142791253126,
      "loss": 4.9228,
      "mean_token_accuracy": 0.2087482213973999,
      "num_tokens": 46813985.0,
      "step": 22560
    },
    {
      "entropy": 5.493589544296265,
      "epoch": 2.046897677793904,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045809547532774943,
      "loss": 4.8426,
      "mean_token_accuracy": 0.20691220760345458,
      "num_tokens": 46824125.0,
      "step": 22565
    },
    {
      "entropy": 5.514066123962403,
      "epoch": 2.0473512336719883,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045807666774372514,
      "loss": 4.8747,
      "mean_token_accuracy": 0.2152392864227295,
      "num_tokens": 46834982.0,
      "step": 22570
    },
    {
      "entropy": 5.515033960342407,
      "epoch": 2.0478047895500726,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045805785637362874,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22229514569044112,
      "num_tokens": 46844669.0,
      "step": 22575
    },
    {
      "entropy": 5.5233604431152346,
      "epoch": 2.048258345428157,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004580390412178492,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21486603766679763,
      "num_tokens": 46855108.0,
      "step": 22580
    },
    {
      "entropy": 5.532042169570923,
      "epoch": 2.048711901306241,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045802022227677566,
      "loss": 4.885,
      "mean_token_accuracy": 0.2107235759496689,
      "num_tokens": 46864878.0,
      "step": 22585
    },
    {
      "entropy": 5.54194221496582,
      "epoch": 2.049165457184325,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045800139955079727,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2083554208278656,
      "num_tokens": 46874916.0,
      "step": 22590
    },
    {
      "entropy": 5.603750705718994,
      "epoch": 2.0496190130624092,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004579825730403033,
      "loss": 5.0055,
      "mean_token_accuracy": 0.20589129775762557,
      "num_tokens": 46885167.0,
      "step": 22595
    },
    {
      "entropy": 5.590600776672363,
      "epoch": 2.0500725689404935,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004579637427456831,
      "loss": 4.9481,
      "mean_token_accuracy": 0.20603059381246566,
      "num_tokens": 46895526.0,
      "step": 22600
    },
    {
      "entropy": 5.523321104049683,
      "epoch": 2.0505261248185778,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045794490866732607,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21477593183517457,
      "num_tokens": 46905426.0,
      "step": 22605
    },
    {
      "entropy": 5.528275632858277,
      "epoch": 2.050979680696662,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004579260708056217,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21035714596509933,
      "num_tokens": 46915728.0,
      "step": 22610
    },
    {
      "entropy": 5.515733766555786,
      "epoch": 2.051433236574746,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004579072291609596,
      "loss": 4.8976,
      "mean_token_accuracy": 0.211222904920578,
      "num_tokens": 46925972.0,
      "step": 22615
    },
    {
      "entropy": 5.529329919815064,
      "epoch": 2.05188679245283,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045788838373372935,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2146576702594757,
      "num_tokens": 46937108.0,
      "step": 22620
    },
    {
      "entropy": 5.509870147705078,
      "epoch": 2.0523403483309144,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045786953452432076,
      "loss": 4.8876,
      "mean_token_accuracy": 0.21134349703788757,
      "num_tokens": 46947684.0,
      "step": 22625
    },
    {
      "entropy": 5.615060329437256,
      "epoch": 2.0527939042089987,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004578506815331236,
      "loss": 4.999,
      "mean_token_accuracy": 0.20833959579467773,
      "num_tokens": 46958089.0,
      "step": 22630
    },
    {
      "entropy": 5.541087007522583,
      "epoch": 2.053247460087083,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004578318247605277,
      "loss": 4.8883,
      "mean_token_accuracy": 0.21928365379571915,
      "num_tokens": 46968618.0,
      "step": 22635
    },
    {
      "entropy": 5.5500030517578125,
      "epoch": 2.0537010159651667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004578129642069231,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21412350684404374,
      "num_tokens": 46978791.0,
      "step": 22640
    },
    {
      "entropy": 5.551610231399536,
      "epoch": 2.054154571843251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004577940998726998,
      "loss": 4.8697,
      "mean_token_accuracy": 0.21991531401872635,
      "num_tokens": 46988706.0,
      "step": 22645
    },
    {
      "entropy": 5.626473760604858,
      "epoch": 2.0546081277213353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045777523175824796,
      "loss": 5.0538,
      "mean_token_accuracy": 0.19763527065515518,
      "num_tokens": 47000140.0,
      "step": 22650
    },
    {
      "entropy": 5.540607929229736,
      "epoch": 2.0550616835994195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004577563598639577,
      "loss": 4.8813,
      "mean_token_accuracy": 0.21221189200878143,
      "num_tokens": 47009773.0,
      "step": 22655
    },
    {
      "entropy": 5.593835735321045,
      "epoch": 2.055515239477504,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045773748419021944,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20747489333152772,
      "num_tokens": 47020326.0,
      "step": 22660
    },
    {
      "entropy": 5.50357141494751,
      "epoch": 2.0559687953555876,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045771860473742347,
      "loss": 4.898,
      "mean_token_accuracy": 0.20945340842008592,
      "num_tokens": 47030410.0,
      "step": 22665
    },
    {
      "entropy": 5.569175815582275,
      "epoch": 2.056422351233672,
      "grad_norm": 1.125,
      "learning_rate": 0.00045769972150596006,
      "loss": 4.9419,
      "mean_token_accuracy": 0.211944343149662,
      "num_tokens": 47040546.0,
      "step": 22670
    },
    {
      "entropy": 5.5609148979187015,
      "epoch": 2.056875907111756,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045768083449621987,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22088221311569214,
      "num_tokens": 47050210.0,
      "step": 22675
    },
    {
      "entropy": 5.6539835929870605,
      "epoch": 2.0573294629898404,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045766194370859347,
      "loss": 5.0634,
      "mean_token_accuracy": 0.19811297506093978,
      "num_tokens": 47060641.0,
      "step": 22680
    },
    {
      "entropy": 5.566586685180664,
      "epoch": 2.0577830188679247,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004576430491434716,
      "loss": 4.9672,
      "mean_token_accuracy": 0.2138103500008583,
      "num_tokens": 47070737.0,
      "step": 22685
    },
    {
      "entropy": 5.616622066497802,
      "epoch": 2.0582365747460085,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045762415080124487,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21003361493349076,
      "num_tokens": 47080884.0,
      "step": 22690
    },
    {
      "entropy": 5.548874759674073,
      "epoch": 2.058690130624093,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045760524868230417,
      "loss": 4.9835,
      "mean_token_accuracy": 0.20848450362682341,
      "num_tokens": 47090865.0,
      "step": 22695
    },
    {
      "entropy": 5.560907363891602,
      "epoch": 2.059143686502177,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004575863427870404,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21426442712545396,
      "num_tokens": 47100570.0,
      "step": 22700
    },
    {
      "entropy": 5.5511678695678714,
      "epoch": 2.0595972423802613,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004575674331158446,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21494615375995635,
      "num_tokens": 47111199.0,
      "step": 22705
    },
    {
      "entropy": 5.543911790847778,
      "epoch": 2.0600507982583456,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004575485196691076,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21214178055524827,
      "num_tokens": 47121191.0,
      "step": 22710
    },
    {
      "entropy": 5.604065036773681,
      "epoch": 2.0605043541364294,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004575296024472209,
      "loss": 4.975,
      "mean_token_accuracy": 0.21015238910913467,
      "num_tokens": 47131608.0,
      "step": 22715
    },
    {
      "entropy": 5.618283414840699,
      "epoch": 2.0609579100145137,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004575106814505753,
      "loss": 4.9471,
      "mean_token_accuracy": 0.20580857396125793,
      "num_tokens": 47141492.0,
      "step": 22720
    },
    {
      "entropy": 5.503385400772094,
      "epoch": 2.061411465892598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004574917566795624,
      "loss": 4.8215,
      "mean_token_accuracy": 0.2223336100578308,
      "num_tokens": 47152576.0,
      "step": 22725
    },
    {
      "entropy": 5.550375938415527,
      "epoch": 2.061865021770682,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045747282813457345,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20989041775465012,
      "num_tokens": 47163863.0,
      "step": 22730
    },
    {
      "entropy": 5.534601259231567,
      "epoch": 2.0623185776487665,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000457453895816,
      "loss": 4.9076,
      "mean_token_accuracy": 0.2132399782538414,
      "num_tokens": 47174140.0,
      "step": 22735
    },
    {
      "entropy": 5.581269454956055,
      "epoch": 2.0627721335268503,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045743495972423346,
      "loss": 4.9705,
      "mean_token_accuracy": 0.20813938081264496,
      "num_tokens": 47184708.0,
      "step": 22740
    },
    {
      "entropy": 5.5300328731536865,
      "epoch": 2.0632256894049346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004574160198596653,
      "loss": 4.93,
      "mean_token_accuracy": 0.2126962959766388,
      "num_tokens": 47196286.0,
      "step": 22745
    },
    {
      "entropy": 5.450542259216308,
      "epoch": 2.063679245283019,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045739707622268757,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22261857241392136,
      "num_tokens": 47206774.0,
      "step": 22750
    },
    {
      "entropy": 5.497993469238281,
      "epoch": 2.064132801161103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004573781288136917,
      "loss": 4.8038,
      "mean_token_accuracy": 0.21777256727218627,
      "num_tokens": 47217009.0,
      "step": 22755
    },
    {
      "entropy": 5.552196025848389,
      "epoch": 2.0645863570391874,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045735917763306967,
      "loss": 4.944,
      "mean_token_accuracy": 0.21327149868011475,
      "num_tokens": 47226718.0,
      "step": 22760
    },
    {
      "entropy": 5.496644496917725,
      "epoch": 2.065039912917271,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004573402226812134,
      "loss": 4.826,
      "mean_token_accuracy": 0.2083118140697479,
      "num_tokens": 47237666.0,
      "step": 22765
    },
    {
      "entropy": 5.5309202671051025,
      "epoch": 2.0654934687953554,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045732126395851483,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20627781599760056,
      "num_tokens": 47249230.0,
      "step": 22770
    },
    {
      "entropy": 5.517955160140991,
      "epoch": 2.0659470246734397,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004573023014653661,
      "loss": 4.885,
      "mean_token_accuracy": 0.2155887544155121,
      "num_tokens": 47259232.0,
      "step": 22775
    },
    {
      "entropy": 5.521569013595581,
      "epoch": 2.066400580551524,
      "grad_norm": 1.125,
      "learning_rate": 0.00045728333520215933,
      "loss": 4.939,
      "mean_token_accuracy": 0.2105821266770363,
      "num_tokens": 47269006.0,
      "step": 22780
    },
    {
      "entropy": 5.5902228355407715,
      "epoch": 2.0668541364296082,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004572643651692867,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20120707750320435,
      "num_tokens": 47280048.0,
      "step": 22785
    },
    {
      "entropy": 5.550869750976562,
      "epoch": 2.0673076923076925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045724539136714063,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2146774187684059,
      "num_tokens": 47290019.0,
      "step": 22790
    },
    {
      "entropy": 5.550041007995605,
      "epoch": 2.0677612481857763,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004572264137961133,
      "loss": 4.8969,
      "mean_token_accuracy": 0.21964723169803618,
      "num_tokens": 47300573.0,
      "step": 22795
    },
    {
      "entropy": 5.49859585762024,
      "epoch": 2.0682148040638606,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045720743245659734,
      "loss": 4.905,
      "mean_token_accuracy": 0.21301880180835725,
      "num_tokens": 47311795.0,
      "step": 22800
    },
    {
      "entropy": 5.470848321914673,
      "epoch": 2.068668359941945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004571884473489853,
      "loss": 4.8148,
      "mean_token_accuracy": 0.2240484967827797,
      "num_tokens": 47321768.0,
      "step": 22805
    },
    {
      "entropy": 5.553780889511108,
      "epoch": 2.069121915820029,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004571694584736696,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2161708205938339,
      "num_tokens": 47332662.0,
      "step": 22810
    },
    {
      "entropy": 5.542087602615356,
      "epoch": 2.0695754716981134,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004571504658310432,
      "loss": 4.9035,
      "mean_token_accuracy": 0.2147950202226639,
      "num_tokens": 47344355.0,
      "step": 22815
    },
    {
      "entropy": 5.545966148376465,
      "epoch": 2.0700290275761972,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004571314694214987,
      "loss": 4.8736,
      "mean_token_accuracy": 0.21230216473340988,
      "num_tokens": 47354228.0,
      "step": 22820
    },
    {
      "entropy": 5.560914754867554,
      "epoch": 2.0704825834542815,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004571124692454291,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22051794081926346,
      "num_tokens": 47364477.0,
      "step": 22825
    },
    {
      "entropy": 5.556259107589722,
      "epoch": 2.0709361393323658,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045709346530322715,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21238750964403152,
      "num_tokens": 47375745.0,
      "step": 22830
    },
    {
      "entropy": 5.588777732849121,
      "epoch": 2.07138969521045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004570744575952859,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21455432325601578,
      "num_tokens": 47385341.0,
      "step": 22835
    },
    {
      "entropy": 5.580568599700928,
      "epoch": 2.0718432510885343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004570554461219985,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21542086750268935,
      "num_tokens": 47396266.0,
      "step": 22840
    },
    {
      "entropy": 5.57943058013916,
      "epoch": 2.072296806966618,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045703643088375806,
      "loss": 4.9378,
      "mean_token_accuracy": 0.20775123983621596,
      "num_tokens": 47406772.0,
      "step": 22845
    },
    {
      "entropy": 5.596522045135498,
      "epoch": 2.0727503628447024,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045701741188095793,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20698040276765822,
      "num_tokens": 47416833.0,
      "step": 22850
    },
    {
      "entropy": 5.519446086883545,
      "epoch": 2.0732039187227866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004569983891139912,
      "loss": 4.8985,
      "mean_token_accuracy": 0.21298438608646392,
      "num_tokens": 47426954.0,
      "step": 22855
    },
    {
      "entropy": 5.498295164108276,
      "epoch": 2.073657474600871,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045697936258325155,
      "loss": 4.869,
      "mean_token_accuracy": 0.22404805421829224,
      "num_tokens": 47437520.0,
      "step": 22860
    },
    {
      "entropy": 5.554725551605225,
      "epoch": 2.074111030478955,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045696033228913226,
      "loss": 4.8891,
      "mean_token_accuracy": 0.20843892097473143,
      "num_tokens": 47447977.0,
      "step": 22865
    },
    {
      "entropy": 5.582836627960205,
      "epoch": 2.074564586357039,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045694129823202684,
      "loss": 4.978,
      "mean_token_accuracy": 0.20482730567455293,
      "num_tokens": 47458458.0,
      "step": 22870
    },
    {
      "entropy": 5.559929847717285,
      "epoch": 2.0750181422351233,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004569222604123291,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2129221186041832,
      "num_tokens": 47469048.0,
      "step": 22875
    },
    {
      "entropy": 5.6153302669525145,
      "epoch": 2.0754716981132075,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004569032188304326,
      "loss": 4.9845,
      "mean_token_accuracy": 0.20533379912376404,
      "num_tokens": 47480353.0,
      "step": 22880
    },
    {
      "entropy": 5.511811351776123,
      "epoch": 2.075925253991292,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004568841734867313,
      "loss": 4.8144,
      "mean_token_accuracy": 0.21520548313856125,
      "num_tokens": 47489951.0,
      "step": 22885
    },
    {
      "entropy": 5.480315065383911,
      "epoch": 2.076378809869376,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004568651243816188,
      "loss": 4.8919,
      "mean_token_accuracy": 0.21955198049545288,
      "num_tokens": 47499015.0,
      "step": 22890
    },
    {
      "entropy": 5.561438846588135,
      "epoch": 2.07683236574746,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045684607151548923,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21274450570344924,
      "num_tokens": 47508694.0,
      "step": 22895
    },
    {
      "entropy": 5.6556556224823,
      "epoch": 2.077285921625544,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045682701488873655,
      "loss": 5.0717,
      "mean_token_accuracy": 0.2010905608534813,
      "num_tokens": 47521336.0,
      "step": 22900
    },
    {
      "entropy": 5.606842470169068,
      "epoch": 2.0777394775036284,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004568079545017549,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20524126589298247,
      "num_tokens": 47531857.0,
      "step": 22905
    },
    {
      "entropy": 5.6149121761322025,
      "epoch": 2.0781930333817127,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004567888903549385,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2053648605942726,
      "num_tokens": 47542224.0,
      "step": 22910
    },
    {
      "entropy": 5.579704284667969,
      "epoch": 2.078646589259797,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004567698224486813,
      "loss": 4.9643,
      "mean_token_accuracy": 0.21154584139585494,
      "num_tokens": 47552370.0,
      "step": 22915
    },
    {
      "entropy": 5.645527410507202,
      "epoch": 2.0791001451378808,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045675075078337803,
      "loss": 4.904,
      "mean_token_accuracy": 0.2167135238647461,
      "num_tokens": 47561997.0,
      "step": 22920
    },
    {
      "entropy": 5.5685585021972654,
      "epoch": 2.079553701015965,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004567316753594229,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21692643761634828,
      "num_tokens": 47571968.0,
      "step": 22925
    },
    {
      "entropy": 5.562157154083252,
      "epoch": 2.0800072568940493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045671259617721037,
      "loss": 4.9547,
      "mean_token_accuracy": 0.20979191064834596,
      "num_tokens": 47583459.0,
      "step": 22930
    },
    {
      "entropy": 5.519590950012207,
      "epoch": 2.0804608127721336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000456693513237135,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2219153270125389,
      "num_tokens": 47594026.0,
      "step": 22935
    },
    {
      "entropy": 5.617216777801514,
      "epoch": 2.080914368650218,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004566744265395916,
      "loss": 4.9486,
      "mean_token_accuracy": 0.21273531317710875,
      "num_tokens": 47605246.0,
      "step": 22940
    },
    {
      "entropy": 5.607788515090943,
      "epoch": 2.081367924528302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045665533608497467,
      "loss": 5.0235,
      "mean_token_accuracy": 0.19994515925645828,
      "num_tokens": 47616702.0,
      "step": 22945
    },
    {
      "entropy": 5.601177740097046,
      "epoch": 2.081821480406386,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045663624187367914,
      "loss": 4.9846,
      "mean_token_accuracy": 0.2063418596982956,
      "num_tokens": 47627616.0,
      "step": 22950
    },
    {
      "entropy": 5.563710880279541,
      "epoch": 2.08227503628447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004566171439060998,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21134505420923233,
      "num_tokens": 47637287.0,
      "step": 22955
    },
    {
      "entropy": 5.487324285507202,
      "epoch": 2.0827285921625545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004565980421826317,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21375772953033448,
      "num_tokens": 47647474.0,
      "step": 22960
    },
    {
      "entropy": 5.589184713363648,
      "epoch": 2.0831821480406387,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004565789367036697,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21286556273698806,
      "num_tokens": 47657036.0,
      "step": 22965
    },
    {
      "entropy": 5.522977304458618,
      "epoch": 2.083635703918723,
      "grad_norm": 1.125,
      "learning_rate": 0.0004565598274696091,
      "loss": 4.8472,
      "mean_token_accuracy": 0.21326051652431488,
      "num_tokens": 47668026.0,
      "step": 22970
    },
    {
      "entropy": 5.616479015350341,
      "epoch": 2.084089259796807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045654071448084506,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20693727433681489,
      "num_tokens": 47679508.0,
      "step": 22975
    },
    {
      "entropy": 5.580051517486572,
      "epoch": 2.084542815674891,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045652159773777273,
      "loss": 4.967,
      "mean_token_accuracy": 0.21123090833425523,
      "num_tokens": 47688474.0,
      "step": 22980
    },
    {
      "entropy": 5.688438034057617,
      "epoch": 2.0849963715529753,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004565024772407875,
      "loss": 5.094,
      "mean_token_accuracy": 0.19593742191791536,
      "num_tokens": 47698675.0,
      "step": 22985
    },
    {
      "entropy": 5.52439513206482,
      "epoch": 2.0854499274310596,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004564833529902848,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22284892052412034,
      "num_tokens": 47709214.0,
      "step": 22990
    },
    {
      "entropy": 5.5863415718078615,
      "epoch": 2.085903483309144,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045646422498666,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21298460066318511,
      "num_tokens": 47721000.0,
      "step": 22995
    },
    {
      "entropy": 5.508289813995361,
      "epoch": 2.0863570391872277,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045644509323030893,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21045969277620316,
      "num_tokens": 47731218.0,
      "step": 23000
    },
    {
      "entropy": 5.584042310714722,
      "epoch": 2.086810595065312,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045642595772162704,
      "loss": 4.918,
      "mean_token_accuracy": 0.21153358221054078,
      "num_tokens": 47742339.0,
      "step": 23005
    },
    {
      "entropy": 5.617961692810058,
      "epoch": 2.0872641509433962,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045640681846101014,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21020923554897308,
      "num_tokens": 47752858.0,
      "step": 23010
    },
    {
      "entropy": 5.517202377319336,
      "epoch": 2.0877177068214805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004563876754488539,
      "loss": 4.8849,
      "mean_token_accuracy": 0.21729450225830077,
      "num_tokens": 47763303.0,
      "step": 23015
    },
    {
      "entropy": 5.56376805305481,
      "epoch": 2.0881712626995648,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004563685286855544,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2116104394197464,
      "num_tokens": 47773212.0,
      "step": 23020
    },
    {
      "entropy": 5.612044334411621,
      "epoch": 2.0886248185776486,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004563493781715075,
      "loss": 5.0329,
      "mean_token_accuracy": 0.20824681967496872,
      "num_tokens": 47783376.0,
      "step": 23025
    },
    {
      "entropy": 5.500726747512817,
      "epoch": 2.089078374455733,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004563302239071093,
      "loss": 4.9278,
      "mean_token_accuracy": 0.21655123680830002,
      "num_tokens": 47794859.0,
      "step": 23030
    },
    {
      "entropy": 5.621681451797485,
      "epoch": 2.089531930333817,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045631106589275576,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21255283206701278,
      "num_tokens": 47804721.0,
      "step": 23035
    },
    {
      "entropy": 5.539343643188476,
      "epoch": 2.0899854862119014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045629190412884324,
      "loss": 4.8329,
      "mean_token_accuracy": 0.22014436423778533,
      "num_tokens": 47814701.0,
      "step": 23040
    },
    {
      "entropy": 5.550324535369873,
      "epoch": 2.0904390420899857,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045627273861576786,
      "loss": 4.897,
      "mean_token_accuracy": 0.214018777012825,
      "num_tokens": 47823941.0,
      "step": 23045
    },
    {
      "entropy": 5.532428026199341,
      "epoch": 2.0908925979680695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045625356935392614,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21294294595718383,
      "num_tokens": 47835793.0,
      "step": 23050
    },
    {
      "entropy": 5.62921404838562,
      "epoch": 2.0913461538461537,
      "grad_norm": 1.125,
      "learning_rate": 0.0004562343963437144,
      "loss": 4.989,
      "mean_token_accuracy": 0.21092495769262315,
      "num_tokens": 47845677.0,
      "step": 23055
    },
    {
      "entropy": 5.574399471282959,
      "epoch": 2.091799709724238,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004562152195855291,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22070108354091644,
      "num_tokens": 47855611.0,
      "step": 23060
    },
    {
      "entropy": 5.498503684997559,
      "epoch": 2.0922532656023223,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004561960390797669,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21368443369865417,
      "num_tokens": 47865115.0,
      "step": 23065
    },
    {
      "entropy": 5.627487516403198,
      "epoch": 2.0927068214804065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045617685482682454,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2100916847586632,
      "num_tokens": 47876690.0,
      "step": 23070
    },
    {
      "entropy": 5.541545295715332,
      "epoch": 2.0931603773584904,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004561576668270986,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2084503874182701,
      "num_tokens": 47887709.0,
      "step": 23075
    },
    {
      "entropy": 5.53898401260376,
      "epoch": 2.0936139332365746,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045613847508098587,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21091461032629014,
      "num_tokens": 47898209.0,
      "step": 23080
    },
    {
      "entropy": 5.537871313095093,
      "epoch": 2.094067489114659,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045611927958888335,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2170143648982048,
      "num_tokens": 47907494.0,
      "step": 23085
    },
    {
      "entropy": 5.56508584022522,
      "epoch": 2.094521044992743,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045610008035118797,
      "loss": 4.9727,
      "mean_token_accuracy": 0.20951011329889296,
      "num_tokens": 47916923.0,
      "step": 23090
    },
    {
      "entropy": 5.49456295967102,
      "epoch": 2.0949746008708274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045608087736829687,
      "loss": 4.8571,
      "mean_token_accuracy": 0.21550212651491166,
      "num_tokens": 47927043.0,
      "step": 23095
    },
    {
      "entropy": 5.5701109886169435,
      "epoch": 2.0954281567489113,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000456061670640607,
      "loss": 4.9238,
      "mean_token_accuracy": 0.20838372707366942,
      "num_tokens": 47937296.0,
      "step": 23100
    },
    {
      "entropy": 5.561228179931641,
      "epoch": 2.0958817126269955,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004560424601685157,
      "loss": 4.963,
      "mean_token_accuracy": 0.21140010803937911,
      "num_tokens": 47948268.0,
      "step": 23105
    },
    {
      "entropy": 5.587416791915894,
      "epoch": 2.09633526850508,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004560232459524201,
      "loss": 4.8873,
      "mean_token_accuracy": 0.20976995974779128,
      "num_tokens": 47958375.0,
      "step": 23110
    },
    {
      "entropy": 5.544656419754029,
      "epoch": 2.096788824383164,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045600402799271766,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21522742062807082,
      "num_tokens": 47968469.0,
      "step": 23115
    },
    {
      "entropy": 5.519022512435913,
      "epoch": 2.0972423802612483,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004559848062898058,
      "loss": 4.8586,
      "mean_token_accuracy": 0.2184354066848755,
      "num_tokens": 47979630.0,
      "step": 23120
    },
    {
      "entropy": 5.56188588142395,
      "epoch": 2.097695936139332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004559655808440821,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2151502102613449,
      "num_tokens": 47989828.0,
      "step": 23125
    },
    {
      "entropy": 5.618375301361084,
      "epoch": 2.0981494920174164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045594635165594404,
      "loss": 5.048,
      "mean_token_accuracy": 0.20327326655387878,
      "num_tokens": 47999219.0,
      "step": 23130
    },
    {
      "entropy": 5.575772428512574,
      "epoch": 2.0986030478955007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045592711872578925,
      "loss": 4.916,
      "mean_token_accuracy": 0.2209403082728386,
      "num_tokens": 48009244.0,
      "step": 23135
    },
    {
      "entropy": 5.583085727691651,
      "epoch": 2.099056603773585,
      "grad_norm": 1.125,
      "learning_rate": 0.00045590788205401557,
      "loss": 4.8919,
      "mean_token_accuracy": 0.21605250835418702,
      "num_tokens": 48019066.0,
      "step": 23140
    },
    {
      "entropy": 5.491763639450073,
      "epoch": 2.099510159651669,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045588864164102086,
      "loss": 4.8769,
      "mean_token_accuracy": 0.21540116965770723,
      "num_tokens": 48028704.0,
      "step": 23145
    },
    {
      "entropy": 5.520301723480225,
      "epoch": 2.0999637155297535,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045586939748720284,
      "loss": 4.9058,
      "mean_token_accuracy": 0.215412475168705,
      "num_tokens": 48039210.0,
      "step": 23150
    },
    {
      "entropy": 5.631437540054321,
      "epoch": 2.1004172714078373,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045585014959295965,
      "loss": 5.0383,
      "mean_token_accuracy": 0.206656414270401,
      "num_tokens": 48048446.0,
      "step": 23155
    },
    {
      "entropy": 5.579017972946167,
      "epoch": 2.1008708272859216,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045583089795868934,
      "loss": 4.935,
      "mean_token_accuracy": 0.2172452911734581,
      "num_tokens": 48058880.0,
      "step": 23160
    },
    {
      "entropy": 5.5533143997192385,
      "epoch": 2.101324383164006,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004558116425847899,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21692482084035875,
      "num_tokens": 48070224.0,
      "step": 23165
    },
    {
      "entropy": 5.5385700225830075,
      "epoch": 2.10177793904209,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004557923834716596,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20600900650024415,
      "num_tokens": 48080914.0,
      "step": 23170
    },
    {
      "entropy": 5.532124996185303,
      "epoch": 2.1022314949201744,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045577312061969677,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20475602895021439,
      "num_tokens": 48090873.0,
      "step": 23175
    },
    {
      "entropy": 5.679980802536011,
      "epoch": 2.102685050798258,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004557538540292997,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2085855036973953,
      "num_tokens": 48101885.0,
      "step": 23180
    },
    {
      "entropy": 5.607133913040161,
      "epoch": 2.1031386066763424,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004557345837008669,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20910215377807617,
      "num_tokens": 48111472.0,
      "step": 23185
    },
    {
      "entropy": 5.484545183181763,
      "epoch": 2.1035921625544267,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045571530963479684,
      "loss": 4.8316,
      "mean_token_accuracy": 0.22218698263168335,
      "num_tokens": 48121351.0,
      "step": 23190
    },
    {
      "entropy": 5.587153577804566,
      "epoch": 2.104045718432511,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045569603183148814,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2215775027871132,
      "num_tokens": 48131104.0,
      "step": 23195
    },
    {
      "entropy": 5.60333948135376,
      "epoch": 2.1044992743105952,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004556767502913395,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20779150277376174,
      "num_tokens": 48142895.0,
      "step": 23200
    },
    {
      "entropy": 5.4921972274780275,
      "epoch": 2.104952830188679,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045565746501474946,
      "loss": 4.8109,
      "mean_token_accuracy": 0.21743903458118438,
      "num_tokens": 48153294.0,
      "step": 23205
    },
    {
      "entropy": 5.5781573295593265,
      "epoch": 2.1054063860667633,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045563817600211714,
      "loss": 4.9845,
      "mean_token_accuracy": 0.206672765314579,
      "num_tokens": 48164319.0,
      "step": 23210
    },
    {
      "entropy": 5.580169200897217,
      "epoch": 2.1058599419448476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045561888325384124,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21305814683437346,
      "num_tokens": 48173918.0,
      "step": 23215
    },
    {
      "entropy": 5.615588712692261,
      "epoch": 2.106313497822932,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004555995867703209,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2084985077381134,
      "num_tokens": 48184127.0,
      "step": 23220
    },
    {
      "entropy": 5.520053958892822,
      "epoch": 2.106767053701016,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045558028655195504,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21235456466674804,
      "num_tokens": 48194198.0,
      "step": 23225
    },
    {
      "entropy": 5.468355131149292,
      "epoch": 2.1072206095791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045556098259914293,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21232093423604964,
      "num_tokens": 48204654.0,
      "step": 23230
    },
    {
      "entropy": 5.588650321960449,
      "epoch": 2.1076741654571842,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045554167491228357,
      "loss": 5.0538,
      "mean_token_accuracy": 0.19435366839170456,
      "num_tokens": 48215322.0,
      "step": 23235
    },
    {
      "entropy": 5.589576196670532,
      "epoch": 2.1081277213352685,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045552236349177637,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22067209631204604,
      "num_tokens": 48226142.0,
      "step": 23240
    },
    {
      "entropy": 5.512128734588623,
      "epoch": 2.1085812772133528,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045550304833802066,
      "loss": 4.8832,
      "mean_token_accuracy": 0.21464934051036835,
      "num_tokens": 48236896.0,
      "step": 23245
    },
    {
      "entropy": 5.46775712966919,
      "epoch": 2.109034833091437,
      "grad_norm": 1.09375,
      "learning_rate": 0.000455483729451416,
      "loss": 4.8861,
      "mean_token_accuracy": 0.21393498480319978,
      "num_tokens": 48247425.0,
      "step": 23250
    },
    {
      "entropy": 5.542385149002075,
      "epoch": 2.109488388969521,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045546440683236173,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20822241604328157,
      "num_tokens": 48258560.0,
      "step": 23255
    },
    {
      "entropy": 5.650876474380493,
      "epoch": 2.109941944847605,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045544508048125763,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2067191407084465,
      "num_tokens": 48269469.0,
      "step": 23260
    },
    {
      "entropy": 5.60484209060669,
      "epoch": 2.1103955007256894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004554257503985032,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2131854325532913,
      "num_tokens": 48278847.0,
      "step": 23265
    },
    {
      "entropy": 5.486921310424805,
      "epoch": 2.1108490566037736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045540641658449834,
      "loss": 4.8604,
      "mean_token_accuracy": 0.21766035705804826,
      "num_tokens": 48288091.0,
      "step": 23270
    },
    {
      "entropy": 5.487744569778442,
      "epoch": 2.111302612481858,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045538707903964275,
      "loss": 4.8656,
      "mean_token_accuracy": 0.21312690526247025,
      "num_tokens": 48297946.0,
      "step": 23275
    },
    {
      "entropy": 5.534839630126953,
      "epoch": 2.1117561683599417,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045536773776433647,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21512040346860886,
      "num_tokens": 48307650.0,
      "step": 23280
    },
    {
      "entropy": 5.582086658477783,
      "epoch": 2.112209724238026,
      "grad_norm": 1.125,
      "learning_rate": 0.00045534839275897937,
      "loss": 5.0283,
      "mean_token_accuracy": 0.20173343420028686,
      "num_tokens": 48318824.0,
      "step": 23285
    },
    {
      "entropy": 5.536384391784668,
      "epoch": 2.1126632801161103,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045532904402397157,
      "loss": 4.8831,
      "mean_token_accuracy": 0.21404924392700195,
      "num_tokens": 48329909.0,
      "step": 23290
    },
    {
      "entropy": 5.5862708568573,
      "epoch": 2.1131168359941945,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004553096915597132,
      "loss": 5.0504,
      "mean_token_accuracy": 0.20401450842618943,
      "num_tokens": 48340142.0,
      "step": 23295
    },
    {
      "entropy": 5.566872692108154,
      "epoch": 2.113570391872279,
      "grad_norm": 1.125,
      "learning_rate": 0.0004552903353666044,
      "loss": 4.892,
      "mean_token_accuracy": 0.21543981432914733,
      "num_tokens": 48350540.0,
      "step": 23300
    },
    {
      "entropy": 5.531789922714234,
      "epoch": 2.114023947750363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045527097544504556,
      "loss": 4.872,
      "mean_token_accuracy": 0.2165081039071083,
      "num_tokens": 48360153.0,
      "step": 23305
    },
    {
      "entropy": 5.497782611846924,
      "epoch": 2.114477503628447,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000455251611795437,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21449376791715621,
      "num_tokens": 48370088.0,
      "step": 23310
    },
    {
      "entropy": 5.582926845550537,
      "epoch": 2.114931059506531,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004552322444181792,
      "loss": 4.9555,
      "mean_token_accuracy": 0.20889388918876647,
      "num_tokens": 48381325.0,
      "step": 23315
    },
    {
      "entropy": 5.611449670791626,
      "epoch": 2.1153846153846154,
      "grad_norm": 1.0,
      "learning_rate": 0.0004552128733136726,
      "loss": 4.9383,
      "mean_token_accuracy": 0.20794569104909896,
      "num_tokens": 48393077.0,
      "step": 23320
    },
    {
      "entropy": 5.54364161491394,
      "epoch": 2.1158381712626997,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004551934984823179,
      "loss": 4.921,
      "mean_token_accuracy": 0.21392431557178498,
      "num_tokens": 48403260.0,
      "step": 23325
    },
    {
      "entropy": 5.47652792930603,
      "epoch": 2.116291727140784,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004551741199245157,
      "loss": 4.8837,
      "mean_token_accuracy": 0.21594650596380233,
      "num_tokens": 48413616.0,
      "step": 23330
    },
    {
      "entropy": 5.547682046890259,
      "epoch": 2.1167452830188678,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045515473764066685,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21258217990398406,
      "num_tokens": 48424189.0,
      "step": 23335
    },
    {
      "entropy": 5.576747989654541,
      "epoch": 2.117198838896952,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045513535163117205,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2100134789943695,
      "num_tokens": 48433603.0,
      "step": 23340
    },
    {
      "entropy": 5.584131240844727,
      "epoch": 2.1176523947750363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004551159618964323,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21004039496183396,
      "num_tokens": 48443043.0,
      "step": 23345
    },
    {
      "entropy": 5.494337797164917,
      "epoch": 2.1181059506531206,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004550965684368486,
      "loss": 4.8283,
      "mean_token_accuracy": 0.2251942738890648,
      "num_tokens": 48452765.0,
      "step": 23350
    },
    {
      "entropy": 5.531205320358277,
      "epoch": 2.118559506531205,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004550771712528219,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22011577785015107,
      "num_tokens": 48463191.0,
      "step": 23355
    },
    {
      "entropy": 5.548751163482666,
      "epoch": 2.1190130624092887,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004550577703447534,
      "loss": 4.8854,
      "mean_token_accuracy": 0.21713743656873702,
      "num_tokens": 48473280.0,
      "step": 23360
    },
    {
      "entropy": 5.517544317245483,
      "epoch": 2.119466618287373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045503836571304443,
      "loss": 4.829,
      "mean_token_accuracy": 0.2206776559352875,
      "num_tokens": 48483465.0,
      "step": 23365
    },
    {
      "entropy": 5.500027513504028,
      "epoch": 2.119920174165457,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004550189573580961,
      "loss": 4.8575,
      "mean_token_accuracy": 0.2173631399869919,
      "num_tokens": 48494497.0,
      "step": 23370
    },
    {
      "entropy": 5.540167760848999,
      "epoch": 2.1203737300435415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004549995452803098,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21660860776901245,
      "num_tokens": 48504106.0,
      "step": 23375
    },
    {
      "entropy": 5.61021842956543,
      "epoch": 2.1208272859216257,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004549801294800872,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21148616075515747,
      "num_tokens": 48515729.0,
      "step": 23380
    },
    {
      "entropy": 5.598976898193359,
      "epoch": 2.1212808417997095,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004549607099578295,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2067449375987053,
      "num_tokens": 48526611.0,
      "step": 23385
    },
    {
      "entropy": 5.519251966476441,
      "epoch": 2.121734397677794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004549412867139386,
      "loss": 4.8778,
      "mean_token_accuracy": 0.2188269853591919,
      "num_tokens": 48537016.0,
      "step": 23390
    },
    {
      "entropy": 5.564614009857178,
      "epoch": 2.122187953555878,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045492185974881587,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20828444957733155,
      "num_tokens": 48546786.0,
      "step": 23395
    },
    {
      "entropy": 5.59769926071167,
      "epoch": 2.1226415094339623,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004549024290628634,
      "loss": 4.9532,
      "mean_token_accuracy": 0.21260345876216888,
      "num_tokens": 48556306.0,
      "step": 23400
    },
    {
      "entropy": 5.582816505432129,
      "epoch": 2.1230950653120466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004548829946564827,
      "loss": 4.9756,
      "mean_token_accuracy": 0.21278643161058425,
      "num_tokens": 48566303.0,
      "step": 23405
    },
    {
      "entropy": 5.57006721496582,
      "epoch": 2.1235486211901304,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004548635565300759,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21467667669057847,
      "num_tokens": 48576313.0,
      "step": 23410
    },
    {
      "entropy": 5.523672008514405,
      "epoch": 2.1240021770682147,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004548441146840449,
      "loss": 4.935,
      "mean_token_accuracy": 0.21271440833806993,
      "num_tokens": 48587133.0,
      "step": 23415
    },
    {
      "entropy": 5.589933681488037,
      "epoch": 2.124455732946299,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045482466911879184,
      "loss": 5.0569,
      "mean_token_accuracy": 0.2045355036854744,
      "num_tokens": 48598524.0,
      "step": 23420
    },
    {
      "entropy": 5.527286243438721,
      "epoch": 2.1249092888243832,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045480521983471875,
      "loss": 4.8935,
      "mean_token_accuracy": 0.21761407554149628,
      "num_tokens": 48609870.0,
      "step": 23425
    },
    {
      "entropy": 5.547309160232544,
      "epoch": 2.1253628447024675,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004547857668322279,
      "loss": 4.9556,
      "mean_token_accuracy": 0.2139271765947342,
      "num_tokens": 48619828.0,
      "step": 23430
    },
    {
      "entropy": 5.548403358459472,
      "epoch": 2.1258164005805513,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004547663101117215,
      "loss": 4.8687,
      "mean_token_accuracy": 0.21930610984563828,
      "num_tokens": 48630361.0,
      "step": 23435
    },
    {
      "entropy": 5.569103813171386,
      "epoch": 2.1262699564586356,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045474684967360216,
      "loss": 4.9307,
      "mean_token_accuracy": 0.20342365950345992,
      "num_tokens": 48641113.0,
      "step": 23440
    },
    {
      "entropy": 5.58072943687439,
      "epoch": 2.12672351233672,
      "grad_norm": 1.078125,
      "learning_rate": 0.000454727385518272,
      "loss": 5.0156,
      "mean_token_accuracy": 0.19991778135299682,
      "num_tokens": 48651777.0,
      "step": 23445
    },
    {
      "entropy": 5.620652103424073,
      "epoch": 2.127177068214804,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004547079176461338,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2037887990474701,
      "num_tokens": 48663237.0,
      "step": 23450
    },
    {
      "entropy": 5.54860782623291,
      "epoch": 2.1276306240928884,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004546884460575901,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2187935620546341,
      "num_tokens": 48673932.0,
      "step": 23455
    },
    {
      "entropy": 5.490437364578247,
      "epoch": 2.1280841799709727,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004546689707530434,
      "loss": 4.8626,
      "mean_token_accuracy": 0.21403494775295256,
      "num_tokens": 48684304.0,
      "step": 23460
    },
    {
      "entropy": 5.485919523239136,
      "epoch": 2.1285377358490565,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045464949173289676,
      "loss": 4.8308,
      "mean_token_accuracy": 0.21719606965780258,
      "num_tokens": 48695569.0,
      "step": 23465
    },
    {
      "entropy": 5.472944831848144,
      "epoch": 2.1289912917271407,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004546300089975528,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21602316349744796,
      "num_tokens": 48706550.0,
      "step": 23470
    },
    {
      "entropy": 5.574724197387695,
      "epoch": 2.129444847605225,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045461052254741445,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20163908302783967,
      "num_tokens": 48716852.0,
      "step": 23475
    },
    {
      "entropy": 5.538242483139038,
      "epoch": 2.1298984034833093,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045459103238288474,
      "loss": 4.8617,
      "mean_token_accuracy": 0.21364195495843888,
      "num_tokens": 48726607.0,
      "step": 23480
    },
    {
      "entropy": 5.5888269424438475,
      "epoch": 2.130351959361393,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004545715385043667,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20001124292612077,
      "num_tokens": 48738011.0,
      "step": 23485
    },
    {
      "entropy": 5.56582202911377,
      "epoch": 2.1308055152394774,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004545520409122635,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2065620392560959,
      "num_tokens": 48747899.0,
      "step": 23490
    },
    {
      "entropy": 5.623044347763061,
      "epoch": 2.1312590711175616,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004545325396069783,
      "loss": 4.9612,
      "mean_token_accuracy": 0.20890477001667024,
      "num_tokens": 48758719.0,
      "step": 23495
    },
    {
      "entropy": 5.595681190490723,
      "epoch": 2.131712626995646,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004545130345889145,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21252372711896897,
      "num_tokens": 48769677.0,
      "step": 23500
    },
    {
      "entropy": 5.579863166809082,
      "epoch": 2.13216618287373,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004544935258584754,
      "loss": 4.9313,
      "mean_token_accuracy": 0.21335101127624512,
      "num_tokens": 48778440.0,
      "step": 23505
    },
    {
      "entropy": 5.610200357437134,
      "epoch": 2.1326197387518144,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004544740134160644,
      "loss": 5.0227,
      "mean_token_accuracy": 0.2070917695760727,
      "num_tokens": 48788328.0,
      "step": 23510
    },
    {
      "entropy": 5.606459283828736,
      "epoch": 2.1330732946298983,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004544544972620851,
      "loss": 4.9925,
      "mean_token_accuracy": 0.20363520532846452,
      "num_tokens": 48799218.0,
      "step": 23515
    },
    {
      "entropy": 5.616060352325439,
      "epoch": 2.1335268505079825,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004544349773969411,
      "loss": 4.9601,
      "mean_token_accuracy": 0.20226793438196183,
      "num_tokens": 48809648.0,
      "step": 23520
    },
    {
      "entropy": 5.560133981704712,
      "epoch": 2.133980406386067,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004544154538210359,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20985978841781616,
      "num_tokens": 48819965.0,
      "step": 23525
    },
    {
      "entropy": 5.6355616569519045,
      "epoch": 2.134433962264151,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004543959265347735,
      "loss": 5.0093,
      "mean_token_accuracy": 0.21467642337083817,
      "num_tokens": 48829862.0,
      "step": 23530
    },
    {
      "entropy": 5.609732627868652,
      "epoch": 2.1348875181422353,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045437639553855757,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21068114787340164,
      "num_tokens": 48840551.0,
      "step": 23535
    },
    {
      "entropy": 5.525649309158325,
      "epoch": 2.135341074020319,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004543568608327921,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21656380146741866,
      "num_tokens": 48851032.0,
      "step": 23540
    },
    {
      "entropy": 5.539538955688476,
      "epoch": 2.1357946298984034,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045433732241788105,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20535550713539125,
      "num_tokens": 48861773.0,
      "step": 23545
    },
    {
      "entropy": 5.617505979537964,
      "epoch": 2.1362481857764877,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045431778029422846,
      "loss": 5.0529,
      "mean_token_accuracy": 0.2073085844516754,
      "num_tokens": 48872430.0,
      "step": 23550
    },
    {
      "entropy": 5.575256824493408,
      "epoch": 2.136701741654572,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004542982344622385,
      "loss": 4.8317,
      "mean_token_accuracy": 0.21494340896606445,
      "num_tokens": 48882575.0,
      "step": 23555
    },
    {
      "entropy": 5.5381001949310305,
      "epoch": 2.137155297532656,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045427868492231533,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2106066793203354,
      "num_tokens": 48893337.0,
      "step": 23560
    },
    {
      "entropy": 5.649920129776001,
      "epoch": 2.13760885341074,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045425913167486333,
      "loss": 5.0761,
      "mean_token_accuracy": 0.20791659653186798,
      "num_tokens": 48904675.0,
      "step": 23565
    },
    {
      "entropy": 5.524318504333496,
      "epoch": 2.1380624092888243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045423957472028677,
      "loss": 4.8854,
      "mean_token_accuracy": 0.22435435503721238,
      "num_tokens": 48914831.0,
      "step": 23570
    },
    {
      "entropy": 5.476833629608154,
      "epoch": 2.1385159651669086,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004542200140589901,
      "loss": 4.8389,
      "mean_token_accuracy": 0.21513169705867768,
      "num_tokens": 48924574.0,
      "step": 23575
    },
    {
      "entropy": 5.561999893188476,
      "epoch": 2.138969521044993,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045420044969137787,
      "loss": 5.0027,
      "mean_token_accuracy": 0.2093830093741417,
      "num_tokens": 48933921.0,
      "step": 23580
    },
    {
      "entropy": 5.5500823020935055,
      "epoch": 2.139423076923077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004541808816178547,
      "loss": 4.9457,
      "mean_token_accuracy": 0.21145492494106294,
      "num_tokens": 48944571.0,
      "step": 23585
    },
    {
      "entropy": 5.566286087036133,
      "epoch": 2.139876632801161,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045416130983882524,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2070055663585663,
      "num_tokens": 48954569.0,
      "step": 23590
    },
    {
      "entropy": 5.570382690429687,
      "epoch": 2.140330188679245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045414173435469427,
      "loss": 4.924,
      "mean_token_accuracy": 0.21110861897468566,
      "num_tokens": 48965886.0,
      "step": 23595
    },
    {
      "entropy": 5.597574377059937,
      "epoch": 2.1407837445573294,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045412215516586654,
      "loss": 4.9425,
      "mean_token_accuracy": 0.21023617088794708,
      "num_tokens": 48976219.0,
      "step": 23600
    },
    {
      "entropy": 5.507729434967041,
      "epoch": 2.1412373004354137,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000454102572272747,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21244221925735474,
      "num_tokens": 48986322.0,
      "step": 23605
    },
    {
      "entropy": 5.546385765075684,
      "epoch": 2.141690856313498,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004540829856757406,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2178581491112709,
      "num_tokens": 48996297.0,
      "step": 23610
    },
    {
      "entropy": 5.5631684303283695,
      "epoch": 2.142144412191582,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004540633953752525,
      "loss": 4.9392,
      "mean_token_accuracy": 0.2150893673300743,
      "num_tokens": 49006165.0,
      "step": 23615
    },
    {
      "entropy": 5.526386499404907,
      "epoch": 2.142597968069666,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004540438013716876,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2047666236758232,
      "num_tokens": 49017278.0,
      "step": 23620
    },
    {
      "entropy": 5.5054583072662355,
      "epoch": 2.1430515239477503,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004540242036654514,
      "loss": 4.8807,
      "mean_token_accuracy": 0.21305353343486785,
      "num_tokens": 49027088.0,
      "step": 23625
    },
    {
      "entropy": 5.555301380157471,
      "epoch": 2.1435050798258346,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000454004602256949,
      "loss": 4.8754,
      "mean_token_accuracy": 0.21166100054979325,
      "num_tokens": 49038156.0,
      "step": 23630
    },
    {
      "entropy": 5.602395725250244,
      "epoch": 2.143958635703919,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004539849971465858,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20847307592630387,
      "num_tokens": 49047480.0,
      "step": 23635
    },
    {
      "entropy": 5.614607524871826,
      "epoch": 2.1444121915820027,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045396538833476725,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21530205011367798,
      "num_tokens": 49057373.0,
      "step": 23640
    },
    {
      "entropy": 5.558904933929443,
      "epoch": 2.144865747460087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004539457758218989,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2142976775765419,
      "num_tokens": 49068380.0,
      "step": 23645
    },
    {
      "entropy": 5.540809392929077,
      "epoch": 2.1453193033381712,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004539261596083863,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21224504560232163,
      "num_tokens": 49078646.0,
      "step": 23650
    },
    {
      "entropy": 5.550167655944824,
      "epoch": 2.1457728592162555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004539065396946351,
      "loss": 4.88,
      "mean_token_accuracy": 0.2214643895626068,
      "num_tokens": 49089281.0,
      "step": 23655
    },
    {
      "entropy": 5.573638916015625,
      "epoch": 2.1462264150943398,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004538869160810511,
      "loss": 4.9537,
      "mean_token_accuracy": 0.20815649181604384,
      "num_tokens": 49099514.0,
      "step": 23660
    },
    {
      "entropy": 5.573917293548584,
      "epoch": 2.146679970972424,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045386728876804,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20969745069742202,
      "num_tokens": 49110247.0,
      "step": 23665
    },
    {
      "entropy": 5.619055366516113,
      "epoch": 2.147133526850508,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045384765775600787,
      "loss": 5.0286,
      "mean_token_accuracy": 0.20521268248558044,
      "num_tokens": 49122492.0,
      "step": 23670
    },
    {
      "entropy": 5.594598436355591,
      "epoch": 2.147587082728592,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004538280230453606,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2154301866889,
      "num_tokens": 49132395.0,
      "step": 23675
    },
    {
      "entropy": 5.611228895187378,
      "epoch": 2.1480406386066764,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004538083846365042,
      "loss": 4.9674,
      "mean_token_accuracy": 0.20996081531047822,
      "num_tokens": 49142909.0,
      "step": 23680
    },
    {
      "entropy": 5.562023687362671,
      "epoch": 2.1484941944847606,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004537887425298448,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21113996356725692,
      "num_tokens": 49152933.0,
      "step": 23685
    },
    {
      "entropy": 5.5440997123718265,
      "epoch": 2.1489477503628445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045376909672578875,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2081219121813774,
      "num_tokens": 49163595.0,
      "step": 23690
    },
    {
      "entropy": 5.492967176437378,
      "epoch": 2.1494013062409287,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004537494472247422,
      "loss": 4.8984,
      "mean_token_accuracy": 0.21840045154094695,
      "num_tokens": 49174664.0,
      "step": 23695
    },
    {
      "entropy": 5.581111717224121,
      "epoch": 2.149854862119013,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004537297940271115,
      "loss": 4.9773,
      "mean_token_accuracy": 0.21579648852348327,
      "num_tokens": 49185105.0,
      "step": 23700
    },
    {
      "entropy": 5.547584915161133,
      "epoch": 2.1503084179970973,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004537101371333031,
      "loss": 4.8874,
      "mean_token_accuracy": 0.21742856353521348,
      "num_tokens": 49194679.0,
      "step": 23705
    },
    {
      "entropy": 5.57237753868103,
      "epoch": 2.1507619738751815,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045369047654372356,
      "loss": 4.8511,
      "mean_token_accuracy": 0.21498076766729354,
      "num_tokens": 49204563.0,
      "step": 23710
    },
    {
      "entropy": 5.598416185379028,
      "epoch": 2.151215529753266,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004536708122587795,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21903055608272554,
      "num_tokens": 49213759.0,
      "step": 23715
    },
    {
      "entropy": 5.504739713668823,
      "epoch": 2.1516690856313496,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004536511442788774,
      "loss": 4.8407,
      "mean_token_accuracy": 0.21756711155176162,
      "num_tokens": 49223604.0,
      "step": 23720
    },
    {
      "entropy": 5.545053482055664,
      "epoch": 2.152122641509434,
      "grad_norm": 1.125,
      "learning_rate": 0.0004536314726044241,
      "loss": 4.957,
      "mean_token_accuracy": 0.2128641873598099,
      "num_tokens": 49234000.0,
      "step": 23725
    },
    {
      "entropy": 5.592650985717773,
      "epoch": 2.152576197387518,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045361179723582654,
      "loss": 4.9375,
      "mean_token_accuracy": 0.20984700918197632,
      "num_tokens": 49244874.0,
      "step": 23730
    },
    {
      "entropy": 5.665120935440063,
      "epoch": 2.1530297532656024,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004535921181734914,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20604943633079528,
      "num_tokens": 49255495.0,
      "step": 23735
    },
    {
      "entropy": 5.5517720699310305,
      "epoch": 2.1534833091436867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004535724354178257,
      "loss": 4.8858,
      "mean_token_accuracy": 0.21245681643486022,
      "num_tokens": 49265357.0,
      "step": 23740
    },
    {
      "entropy": 5.604379177093506,
      "epoch": 2.1539368650217705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045355274896923664,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20246616750955582,
      "num_tokens": 49276396.0,
      "step": 23745
    },
    {
      "entropy": 5.5582743167877195,
      "epoch": 2.1543904208998548,
      "grad_norm": 1.125,
      "learning_rate": 0.0004535330588281312,
      "loss": 4.8679,
      "mean_token_accuracy": 0.21818099468946456,
      "num_tokens": 49285925.0,
      "step": 23750
    },
    {
      "entropy": 5.509740018844605,
      "epoch": 2.154843976777939,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004535133649949166,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2197282150387764,
      "num_tokens": 49296493.0,
      "step": 23755
    },
    {
      "entropy": 5.493279123306275,
      "epoch": 2.1552975326560233,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004534936674700001,
      "loss": 4.8728,
      "mean_token_accuracy": 0.21899235099554062,
      "num_tokens": 49307273.0,
      "step": 23760
    },
    {
      "entropy": 5.558998394012451,
      "epoch": 2.1557510885341076,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045347396625378914,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22031073719263078,
      "num_tokens": 49317055.0,
      "step": 23765
    },
    {
      "entropy": 5.583795595169067,
      "epoch": 2.1562046444121914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045345426134669103,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2076543837785721,
      "num_tokens": 49327420.0,
      "step": 23770
    },
    {
      "entropy": 5.564263010025025,
      "epoch": 2.1566582002902757,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045343455274911325,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20900900959968566,
      "num_tokens": 49338323.0,
      "step": 23775
    },
    {
      "entropy": 5.584456729888916,
      "epoch": 2.15711175616836,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004534148404614635,
      "loss": 4.966,
      "mean_token_accuracy": 0.2122632309794426,
      "num_tokens": 49348364.0,
      "step": 23780
    },
    {
      "entropy": 5.641426658630371,
      "epoch": 2.157565312046444,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045339512448414943,
      "loss": 4.9978,
      "mean_token_accuracy": 0.21000421792268753,
      "num_tokens": 49358274.0,
      "step": 23785
    },
    {
      "entropy": 5.5278363704681395,
      "epoch": 2.1580188679245285,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004533754048175787,
      "loss": 4.8781,
      "mean_token_accuracy": 0.21102921068668365,
      "num_tokens": 49368651.0,
      "step": 23790
    },
    {
      "entropy": 5.525886058807373,
      "epoch": 2.1584724238026123,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004533556814621591,
      "loss": 4.8891,
      "mean_token_accuracy": 0.20612184703350067,
      "num_tokens": 49378363.0,
      "step": 23795
    },
    {
      "entropy": 5.562664365768432,
      "epoch": 2.1589259796806965,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045333595441829865,
      "loss": 4.9131,
      "mean_token_accuracy": 0.2144285872578621,
      "num_tokens": 49388241.0,
      "step": 23800
    },
    {
      "entropy": 5.56508150100708,
      "epoch": 2.159379535558781,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004533162236864051,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2083830565214157,
      "num_tokens": 49397678.0,
      "step": 23805
    },
    {
      "entropy": 5.620884561538697,
      "epoch": 2.159833091436865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004532964892668867,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20939994752407073,
      "num_tokens": 49407368.0,
      "step": 23810
    },
    {
      "entropy": 5.5370481491088865,
      "epoch": 2.1602866473149493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045327675116015146,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21844192892313002,
      "num_tokens": 49417926.0,
      "step": 23815
    },
    {
      "entropy": 5.649501085281372,
      "epoch": 2.1607402031930336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004532570093666075,
      "loss": 5.0501,
      "mean_token_accuracy": 0.19813580811023712,
      "num_tokens": 49427899.0,
      "step": 23820
    },
    {
      "entropy": 5.547929573059082,
      "epoch": 2.1611937590711174,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004532372638866633,
      "loss": 4.8605,
      "mean_token_accuracy": 0.22209731489419937,
      "num_tokens": 49439102.0,
      "step": 23825
    },
    {
      "entropy": 5.551114797592163,
      "epoch": 2.1616473149492017,
      "grad_norm": 1.25,
      "learning_rate": 0.00045321751472072693,
      "loss": 4.8327,
      "mean_token_accuracy": 0.22399413585662842,
      "num_tokens": 49448013.0,
      "step": 23830
    },
    {
      "entropy": 5.476213884353638,
      "epoch": 2.162100870827286,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045319776186920695,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22196878045797347,
      "num_tokens": 49457620.0,
      "step": 23835
    },
    {
      "entropy": 5.486994743347168,
      "epoch": 2.1625544267053702,
      "grad_norm": 1.125,
      "learning_rate": 0.00045317800533251193,
      "loss": 4.9296,
      "mean_token_accuracy": 0.2092035233974457,
      "num_tokens": 49468038.0,
      "step": 23840
    },
    {
      "entropy": 5.511063575744629,
      "epoch": 2.163007982583454,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045315824511105027,
      "loss": 4.8703,
      "mean_token_accuracy": 0.21102674603462218,
      "num_tokens": 49477714.0,
      "step": 23845
    },
    {
      "entropy": 5.595931816101074,
      "epoch": 2.1634615384615383,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045313848120523073,
      "loss": 4.8559,
      "mean_token_accuracy": 0.219837549328804,
      "num_tokens": 49488214.0,
      "step": 23850
    },
    {
      "entropy": 5.57024393081665,
      "epoch": 2.1639150943396226,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045311871361546205,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21644212901592255,
      "num_tokens": 49498444.0,
      "step": 23855
    },
    {
      "entropy": 5.478400087356567,
      "epoch": 2.164368650217707,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004530989423421529,
      "loss": 4.8408,
      "mean_token_accuracy": 0.21539229452610015,
      "num_tokens": 49508759.0,
      "step": 23860
    },
    {
      "entropy": 5.462951278686523,
      "epoch": 2.164822206095791,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045307916738571223,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22646919190883635,
      "num_tokens": 49518787.0,
      "step": 23865
    },
    {
      "entropy": 5.5541462898254395,
      "epoch": 2.1652757619738754,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004530593887465491,
      "loss": 4.9194,
      "mean_token_accuracy": 0.21947762221097947,
      "num_tokens": 49528130.0,
      "step": 23870
    },
    {
      "entropy": 5.621574974060058,
      "epoch": 2.165729317851959,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045303960642507235,
      "loss": 5.0673,
      "mean_token_accuracy": 0.2061009705066681,
      "num_tokens": 49539128.0,
      "step": 23875
    },
    {
      "entropy": 5.498931550979615,
      "epoch": 2.1661828737300435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045301982042169123,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21534635871648788,
      "num_tokens": 49550571.0,
      "step": 23880
    },
    {
      "entropy": 5.532294464111328,
      "epoch": 2.1666364296081277,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045300003073681485,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2141397401690483,
      "num_tokens": 49561106.0,
      "step": 23885
    },
    {
      "entropy": 5.519245576858521,
      "epoch": 2.167089985486212,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045298023737085243,
      "loss": 4.8295,
      "mean_token_accuracy": 0.21391812115907669,
      "num_tokens": 49571389.0,
      "step": 23890
    },
    {
      "entropy": 5.542738008499145,
      "epoch": 2.1675435413642963,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004529604403242134,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2099479153752327,
      "num_tokens": 49582224.0,
      "step": 23895
    },
    {
      "entropy": 5.548014879226685,
      "epoch": 2.16799709724238,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045294063959730707,
      "loss": 4.9492,
      "mean_token_accuracy": 0.20937033742666245,
      "num_tokens": 49592413.0,
      "step": 23900
    },
    {
      "entropy": 5.5853612422943115,
      "epoch": 2.1684506531204644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045292083519054297,
      "loss": 5.0179,
      "mean_token_accuracy": 0.19779800474643708,
      "num_tokens": 49602796.0,
      "step": 23905
    },
    {
      "entropy": 5.516379737854004,
      "epoch": 2.1689042089985486,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004529010271043307,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21849001199007034,
      "num_tokens": 49613625.0,
      "step": 23910
    },
    {
      "entropy": 5.6649391651153564,
      "epoch": 2.169357764876633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045288121533907977,
      "loss": 5.0185,
      "mean_token_accuracy": 0.208993361890316,
      "num_tokens": 49624719.0,
      "step": 23915
    },
    {
      "entropy": 5.614285802841186,
      "epoch": 2.169811320754717,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045286139989520006,
      "loss": 4.988,
      "mean_token_accuracy": 0.20872919261455536,
      "num_tokens": 49634916.0,
      "step": 23920
    },
    {
      "entropy": 5.578924083709717,
      "epoch": 2.170264876632801,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004528415807731012,
      "loss": 4.8874,
      "mean_token_accuracy": 0.21143622994422911,
      "num_tokens": 49645215.0,
      "step": 23925
    },
    {
      "entropy": 5.541259670257569,
      "epoch": 2.1707184325108853,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045282175797319317,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21639792621135712,
      "num_tokens": 49655805.0,
      "step": 23930
    },
    {
      "entropy": 5.569958972930908,
      "epoch": 2.1711719883889695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004528019314958859,
      "loss": 4.9323,
      "mean_token_accuracy": 0.2147790864109993,
      "num_tokens": 49665264.0,
      "step": 23935
    },
    {
      "entropy": 5.5874114513397215,
      "epoch": 2.171625544267054,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045278210134158935,
      "loss": 4.989,
      "mean_token_accuracy": 0.20953418165445328,
      "num_tokens": 49676152.0,
      "step": 23940
    },
    {
      "entropy": 5.58998851776123,
      "epoch": 2.172079100145138,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004527622675107137,
      "loss": 4.9423,
      "mean_token_accuracy": 0.20832472294569016,
      "num_tokens": 49686762.0,
      "step": 23945
    },
    {
      "entropy": 5.613668918609619,
      "epoch": 2.172532656023222,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045274243000366897,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21186559349298478,
      "num_tokens": 49697504.0,
      "step": 23950
    },
    {
      "entropy": 5.5716845989227295,
      "epoch": 2.172986211901306,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004527225888208655,
      "loss": 5.0054,
      "mean_token_accuracy": 0.20693681091070176,
      "num_tokens": 49709197.0,
      "step": 23955
    },
    {
      "entropy": 5.57248010635376,
      "epoch": 2.1734397677793904,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004527027439627137,
      "loss": 4.9633,
      "mean_token_accuracy": 0.20444791167974471,
      "num_tokens": 49719764.0,
      "step": 23960
    },
    {
      "entropy": 5.5693206787109375,
      "epoch": 2.1738933236574747,
      "grad_norm": 1.125,
      "learning_rate": 0.00045268289542962377,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20790381133556365,
      "num_tokens": 49730363.0,
      "step": 23965
    },
    {
      "entropy": 5.575118541717529,
      "epoch": 2.174346879535559,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004526630432220063,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20046700537204742,
      "num_tokens": 49741037.0,
      "step": 23970
    },
    {
      "entropy": 5.541949987411499,
      "epoch": 2.1748004354136428,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004526431873402719,
      "loss": 4.8682,
      "mean_token_accuracy": 0.2153545692563057,
      "num_tokens": 49750127.0,
      "step": 23975
    },
    {
      "entropy": 5.653799819946289,
      "epoch": 2.175253991291727,
      "grad_norm": 1.109375,
      "learning_rate": 0.000452623327784831,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20731836259365083,
      "num_tokens": 49760298.0,
      "step": 23980
    },
    {
      "entropy": 5.575394868850708,
      "epoch": 2.1757075471698113,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004526034645560945,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20924535840749742,
      "num_tokens": 49770841.0,
      "step": 23985
    },
    {
      "entropy": 5.515012645721436,
      "epoch": 2.1761611030478956,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045258359765447306,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21334540694952012,
      "num_tokens": 49781000.0,
      "step": 23990
    },
    {
      "entropy": 5.541502332687378,
      "epoch": 2.17661465892598,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004525637270803776,
      "loss": 4.9125,
      "mean_token_accuracy": 0.215426667034626,
      "num_tokens": 49790886.0,
      "step": 23995
    },
    {
      "entropy": 5.552193069458008,
      "epoch": 2.1770682148040637,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045254385283421896,
      "loss": 4.9144,
      "mean_token_accuracy": 0.21511530727148057,
      "num_tokens": 49801477.0,
      "step": 24000
    },
    {
      "epoch": 2.1770682148040637,
      "eval_entropy": 5.3124205934515905,
      "eval_loss": 5.021059513092041,
      "eval_mean_token_accuracy": 0.21675180812193529,
      "eval_num_tokens": 49801477.0,
      "eval_runtime": 20.6759,
      "eval_samples_per_second": 1710.202,
      "eval_steps_per_second": 213.775,
      "step": 24000
    },
    {
      "entropy": 5.567465543746948,
      "epoch": 2.177521770682148,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004525239749164082,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21617337167263032,
      "num_tokens": 49812044.0,
      "step": 24005
    },
    {
      "entropy": 5.517443132400513,
      "epoch": 2.177975326560232,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004525040933273564,
      "loss": 4.866,
      "mean_token_accuracy": 0.21279383152723313,
      "num_tokens": 49821559.0,
      "step": 24010
    },
    {
      "entropy": 5.611775588989258,
      "epoch": 2.1784288824383164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045248420806747483,
      "loss": 5.035,
      "mean_token_accuracy": 0.2034580260515213,
      "num_tokens": 49832937.0,
      "step": 24015
    },
    {
      "entropy": 5.610374784469604,
      "epoch": 2.1788824383164007,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004524643191371745,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21035014986991882,
      "num_tokens": 49843176.0,
      "step": 24020
    },
    {
      "entropy": 5.558076333999634,
      "epoch": 2.179335994194485,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045244442653686684,
      "loss": 4.8717,
      "mean_token_accuracy": 0.22338457107543946,
      "num_tokens": 49852522.0,
      "step": 24025
    },
    {
      "entropy": 5.5005677223205565,
      "epoch": 2.179789550072569,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045242453026696325,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2199748158454895,
      "num_tokens": 49863143.0,
      "step": 24030
    },
    {
      "entropy": 5.598064994812011,
      "epoch": 2.180243105950653,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004524046303278752,
      "loss": 4.9663,
      "mean_token_accuracy": 0.211096827685833,
      "num_tokens": 49873465.0,
      "step": 24035
    },
    {
      "entropy": 5.659970378875732,
      "epoch": 2.1806966618287373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004523847267200141,
      "loss": 5.042,
      "mean_token_accuracy": 0.20355025976896285,
      "num_tokens": 49883690.0,
      "step": 24040
    },
    {
      "entropy": 5.613793897628784,
      "epoch": 2.1811502177068216,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045236481944379165,
      "loss": 5.0147,
      "mean_token_accuracy": 0.20588492155075072,
      "num_tokens": 49894418.0,
      "step": 24045
    },
    {
      "entropy": 5.574164962768554,
      "epoch": 2.1816037735849054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004523449084996196,
      "loss": 4.8774,
      "mean_token_accuracy": 0.21884071081876755,
      "num_tokens": 49903927.0,
      "step": 24050
    },
    {
      "entropy": 5.557644462585449,
      "epoch": 2.1820573294629897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004523249938879096,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21283381432294846,
      "num_tokens": 49914203.0,
      "step": 24055
    },
    {
      "entropy": 5.572479343414306,
      "epoch": 2.182510885341074,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045230507560907354,
      "loss": 4.9323,
      "mean_token_accuracy": 0.21755848675966263,
      "num_tokens": 49924320.0,
      "step": 24060
    },
    {
      "entropy": 5.510980892181396,
      "epoch": 2.1829644412191582,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045228515366352334,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2151745319366455,
      "num_tokens": 49934634.0,
      "step": 24065
    },
    {
      "entropy": 5.628230476379395,
      "epoch": 2.1834179970972425,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045226522805167095,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20996217131614686,
      "num_tokens": 49944586.0,
      "step": 24070
    },
    {
      "entropy": 5.573509883880615,
      "epoch": 2.1838715529753268,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004522452987739286,
      "loss": 4.939,
      "mean_token_accuracy": 0.21171447187662124,
      "num_tokens": 49955441.0,
      "step": 24075
    },
    {
      "entropy": 5.625335645675659,
      "epoch": 2.1843251088534106,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045222536583070823,
      "loss": 5.0501,
      "mean_token_accuracy": 0.21236481964588166,
      "num_tokens": 49966718.0,
      "step": 24080
    },
    {
      "entropy": 5.623920345306397,
      "epoch": 2.184778664731495,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045220542922242205,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20676001459360122,
      "num_tokens": 49979483.0,
      "step": 24085
    },
    {
      "entropy": 5.569506120681763,
      "epoch": 2.185232220609579,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045218548894948254,
      "loss": 4.9226,
      "mean_token_accuracy": 0.2171776846051216,
      "num_tokens": 49989976.0,
      "step": 24090
    },
    {
      "entropy": 5.5316712856292725,
      "epoch": 2.1856857764876634,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045216554501230187,
      "loss": 4.8031,
      "mean_token_accuracy": 0.21656409054994583,
      "num_tokens": 50000069.0,
      "step": 24095
    },
    {
      "entropy": 5.537426471710205,
      "epoch": 2.1861393323657476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004521455974112927,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21228833943605424,
      "num_tokens": 50009970.0,
      "step": 24100
    },
    {
      "entropy": 5.587401485443115,
      "epoch": 2.1865928882438315,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045212564614686735,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21030141711235045,
      "num_tokens": 50019590.0,
      "step": 24105
    },
    {
      "entropy": 5.57425594329834,
      "epoch": 2.1870464441219157,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004521056912194386,
      "loss": 4.941,
      "mean_token_accuracy": 0.2013477236032486,
      "num_tokens": 50030092.0,
      "step": 24110
    },
    {
      "entropy": 5.510309839248658,
      "epoch": 2.1875,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004520857326294189,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21099548786878586,
      "num_tokens": 50040318.0,
      "step": 24115
    },
    {
      "entropy": 5.581630992889404,
      "epoch": 2.1879535558780843,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045206577037722114,
      "loss": 5.0615,
      "mean_token_accuracy": 0.19968541115522384,
      "num_tokens": 50050507.0,
      "step": 24120
    },
    {
      "entropy": 5.713084602355957,
      "epoch": 2.1884071117561685,
      "grad_norm": 1.125,
      "learning_rate": 0.0004520458044632582,
      "loss": 5.0184,
      "mean_token_accuracy": 0.20956000536680222,
      "num_tokens": 50059727.0,
      "step": 24125
    },
    {
      "entropy": 5.605908966064453,
      "epoch": 2.1888606676342524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045202583488794285,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21035483330488206,
      "num_tokens": 50070812.0,
      "step": 24130
    },
    {
      "entropy": 5.614676666259766,
      "epoch": 2.1893142235123366,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004520058616516881,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2030799001455307,
      "num_tokens": 50081175.0,
      "step": 24135
    },
    {
      "entropy": 5.563994359970093,
      "epoch": 2.189767779390421,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045198588475490716,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21069299876689912,
      "num_tokens": 50092246.0,
      "step": 24140
    },
    {
      "entropy": 5.591595077514649,
      "epoch": 2.190221335268505,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004519659041980129,
      "loss": 4.987,
      "mean_token_accuracy": 0.2111479938030243,
      "num_tokens": 50101934.0,
      "step": 24145
    },
    {
      "entropy": 5.58695707321167,
      "epoch": 2.1906748911465894,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004519459199814187,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2108437404036522,
      "num_tokens": 50112159.0,
      "step": 24150
    },
    {
      "entropy": 5.547046375274658,
      "epoch": 2.1911284470246732,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004519259321055377,
      "loss": 4.97,
      "mean_token_accuracy": 0.207854525744915,
      "num_tokens": 50123079.0,
      "step": 24155
    },
    {
      "entropy": 5.56029863357544,
      "epoch": 2.1915820029027575,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004519059405707834,
      "loss": 4.9458,
      "mean_token_accuracy": 0.20935097485780715,
      "num_tokens": 50133097.0,
      "step": 24160
    },
    {
      "entropy": 5.640683174133301,
      "epoch": 2.1920355587808418,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045188594537756916,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21063667833805083,
      "num_tokens": 50143062.0,
      "step": 24165
    },
    {
      "entropy": 5.625964641571045,
      "epoch": 2.192489114658926,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045186594652630853,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21432217061519623,
      "num_tokens": 50154482.0,
      "step": 24170
    },
    {
      "entropy": 5.566651487350464,
      "epoch": 2.1929426705370103,
      "grad_norm": 1.125,
      "learning_rate": 0.000451845944017415,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21672665178775788,
      "num_tokens": 50164880.0,
      "step": 24175
    },
    {
      "entropy": 5.5295075416564945,
      "epoch": 2.1933962264150946,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045182593785130225,
      "loss": 4.8864,
      "mean_token_accuracy": 0.21461732685565948,
      "num_tokens": 50173896.0,
      "step": 24180
    },
    {
      "entropy": 5.587013864517212,
      "epoch": 2.1938497822931784,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004518059280283842,
      "loss": 4.9906,
      "mean_token_accuracy": 0.20583032071590424,
      "num_tokens": 50185078.0,
      "step": 24185
    },
    {
      "entropy": 5.597308826446533,
      "epoch": 2.1943033381712627,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045178591454907434,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21808431446552276,
      "num_tokens": 50195500.0,
      "step": 24190
    },
    {
      "entropy": 5.514723110198974,
      "epoch": 2.194756894049347,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045176589741378675,
      "loss": 4.8879,
      "mean_token_accuracy": 0.21505440771579742,
      "num_tokens": 50206103.0,
      "step": 24195
    },
    {
      "entropy": 5.494816303253174,
      "epoch": 2.195210449927431,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004517458766229354,
      "loss": 4.8877,
      "mean_token_accuracy": 0.21155305355787277,
      "num_tokens": 50216522.0,
      "step": 24200
    },
    {
      "entropy": 5.617796611785889,
      "epoch": 2.195664005805515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045172585217693426,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2140292927622795,
      "num_tokens": 50226721.0,
      "step": 24205
    },
    {
      "entropy": 5.660838317871094,
      "epoch": 2.1961175616835993,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045170582407619747,
      "loss": 4.9968,
      "mean_token_accuracy": 0.20557497888803483,
      "num_tokens": 50236327.0,
      "step": 24210
    },
    {
      "entropy": 5.515146970748901,
      "epoch": 2.1965711175616836,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004516857923211392,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2197105437517166,
      "num_tokens": 50245729.0,
      "step": 24215
    },
    {
      "entropy": 5.580048704147339,
      "epoch": 2.197024673439768,
      "grad_norm": 1.125,
      "learning_rate": 0.00045166575691217375,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20498782992362977,
      "num_tokens": 50257024.0,
      "step": 24220
    },
    {
      "entropy": 5.585898065567017,
      "epoch": 2.197478229317852,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045164571784971536,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21638111472129823,
      "num_tokens": 50266910.0,
      "step": 24225
    },
    {
      "entropy": 5.564283275604248,
      "epoch": 2.1979317851959363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045162567513417864,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21388950496912001,
      "num_tokens": 50278233.0,
      "step": 24230
    },
    {
      "entropy": 5.54777045249939,
      "epoch": 2.19838534107402,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045160562876597776,
      "loss": 4.9073,
      "mean_token_accuracy": 0.20627465546131135,
      "num_tokens": 50289276.0,
      "step": 24235
    },
    {
      "entropy": 5.553523683547974,
      "epoch": 2.1988388969521044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004515855787455276,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21545596271753312,
      "num_tokens": 50299336.0,
      "step": 24240
    },
    {
      "entropy": 5.560890579223633,
      "epoch": 2.1992924528301887,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004515655250732426,
      "loss": 4.8762,
      "mean_token_accuracy": 0.2169785499572754,
      "num_tokens": 50308308.0,
      "step": 24245
    },
    {
      "entropy": 5.527594089508057,
      "epoch": 2.199746008708273,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045154546774953764,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21240466088056564,
      "num_tokens": 50318295.0,
      "step": 24250
    },
    {
      "entropy": 5.504431247711182,
      "epoch": 2.2001995645863572,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004515254067748274,
      "loss": 4.9254,
      "mean_token_accuracy": 0.21872513592243195,
      "num_tokens": 50329296.0,
      "step": 24255
    },
    {
      "entropy": 5.55621371269226,
      "epoch": 2.200653120464441,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004515053421495267,
      "loss": 5.001,
      "mean_token_accuracy": 0.20430677384138107,
      "num_tokens": 50339389.0,
      "step": 24260
    },
    {
      "entropy": 5.650838375091553,
      "epoch": 2.2011066763425253,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045148527387405056,
      "loss": 5.0359,
      "mean_token_accuracy": 0.20117291063070297,
      "num_tokens": 50349647.0,
      "step": 24265
    },
    {
      "entropy": 5.642973470687866,
      "epoch": 2.2015602322206096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004514652019488141,
      "loss": 5.0029,
      "mean_token_accuracy": 0.2057658925652504,
      "num_tokens": 50361292.0,
      "step": 24270
    },
    {
      "entropy": 5.57130913734436,
      "epoch": 2.202013788098694,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004514451263742321,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21088439375162124,
      "num_tokens": 50372293.0,
      "step": 24275
    },
    {
      "entropy": 5.51547155380249,
      "epoch": 2.202467343976778,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004514250471507201,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21744245737791063,
      "num_tokens": 50382462.0,
      "step": 24280
    },
    {
      "entropy": 5.591727590560913,
      "epoch": 2.202920899854862,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045140496427869306,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2198596939444542,
      "num_tokens": 50393157.0,
      "step": 24285
    },
    {
      "entropy": 5.487828397750855,
      "epoch": 2.203374455732946,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004513848777585664,
      "loss": 4.847,
      "mean_token_accuracy": 0.21846867501735687,
      "num_tokens": 50403474.0,
      "step": 24290
    },
    {
      "entropy": 5.560375642776489,
      "epoch": 2.2038280116110305,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004513647875907556,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2093304619193077,
      "num_tokens": 50414337.0,
      "step": 24295
    },
    {
      "entropy": 5.637948083877563,
      "epoch": 2.2042815674891147,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045134469377567597,
      "loss": 5.0628,
      "mean_token_accuracy": 0.19925625026226043,
      "num_tokens": 50424863.0,
      "step": 24300
    },
    {
      "entropy": 5.61402997970581,
      "epoch": 2.204735123367199,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004513245963137432,
      "loss": 5.0093,
      "mean_token_accuracy": 0.20893265604972838,
      "num_tokens": 50435982.0,
      "step": 24305
    },
    {
      "entropy": 5.5586028575897215,
      "epoch": 2.205188679245283,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004513044952053728,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21000106036663055,
      "num_tokens": 50445112.0,
      "step": 24310
    },
    {
      "entropy": 5.49434266090393,
      "epoch": 2.205642235123367,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045128439045098045,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2223430946469307,
      "num_tokens": 50455266.0,
      "step": 24315
    },
    {
      "entropy": 5.593818855285645,
      "epoch": 2.2060957910014514,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045126428205098207,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20155433416366578,
      "num_tokens": 50466930.0,
      "step": 24320
    },
    {
      "entropy": 5.599624061584473,
      "epoch": 2.2065493468795356,
      "grad_norm": 1.125,
      "learning_rate": 0.0004512441700057934,
      "loss": 4.9551,
      "mean_token_accuracy": 0.21479129493236543,
      "num_tokens": 50477534.0,
      "step": 24325
    },
    {
      "entropy": 5.580485439300537,
      "epoch": 2.20700290275762,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004512240543158305,
      "loss": 4.9292,
      "mean_token_accuracy": 0.21126864701509476,
      "num_tokens": 50487352.0,
      "step": 24330
    },
    {
      "entropy": 5.515688037872314,
      "epoch": 2.2074564586357037,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045120393498150905,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22341122329235077,
      "num_tokens": 50496227.0,
      "step": 24335
    },
    {
      "entropy": 5.569360017776489,
      "epoch": 2.207910014513788,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045118381200324537,
      "loss": 4.9655,
      "mean_token_accuracy": 0.20855049341917037,
      "num_tokens": 50507536.0,
      "step": 24340
    },
    {
      "entropy": 5.61978759765625,
      "epoch": 2.2083635703918723,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004511636853814556,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2121044009923935,
      "num_tokens": 50517659.0,
      "step": 24345
    },
    {
      "entropy": 5.570859956741333,
      "epoch": 2.2088171262699565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045114355511655595,
      "loss": 4.935,
      "mean_token_accuracy": 0.2064822569489479,
      "num_tokens": 50528717.0,
      "step": 24350
    },
    {
      "entropy": 5.504991149902343,
      "epoch": 2.209270682148041,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045112342120896267,
      "loss": 4.8521,
      "mean_token_accuracy": 0.215097114443779,
      "num_tokens": 50538449.0,
      "step": 24355
    },
    {
      "entropy": 5.598603582382202,
      "epoch": 2.2097242380261246,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045110328365909215,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20462184697389602,
      "num_tokens": 50549098.0,
      "step": 24360
    },
    {
      "entropy": 5.603333568572998,
      "epoch": 2.210177793904209,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004510831424673609,
      "loss": 4.9632,
      "mean_token_accuracy": 0.20936270505189897,
      "num_tokens": 50560047.0,
      "step": 24365
    },
    {
      "entropy": 5.609273386001587,
      "epoch": 2.210631349782293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004510629976341853,
      "loss": 4.9174,
      "mean_token_accuracy": 0.20975878089666367,
      "num_tokens": 50569704.0,
      "step": 24370
    },
    {
      "entropy": 5.578531885147095,
      "epoch": 2.2110849056603774,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045104284915998215,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21189524382352828,
      "num_tokens": 50580006.0,
      "step": 24375
    },
    {
      "entropy": 5.522437000274659,
      "epoch": 2.2115384615384617,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045102269704516793,
      "loss": 4.9279,
      "mean_token_accuracy": 0.21835915595293046,
      "num_tokens": 50589593.0,
      "step": 24380
    },
    {
      "entropy": 5.56684250831604,
      "epoch": 2.211992017416546,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004510025412901595,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2102399706840515,
      "num_tokens": 50598774.0,
      "step": 24385
    },
    {
      "entropy": 5.5032249927520756,
      "epoch": 2.2124455732946298,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045098238189537367,
      "loss": 4.8785,
      "mean_token_accuracy": 0.21406116038560868,
      "num_tokens": 50608815.0,
      "step": 24390
    },
    {
      "entropy": 5.494785213470459,
      "epoch": 2.212899129172714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045096221886122733,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21701135933399202,
      "num_tokens": 50619255.0,
      "step": 24395
    },
    {
      "entropy": 5.5682470321655275,
      "epoch": 2.2133526850507983,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004509420521881375,
      "loss": 4.9253,
      "mean_token_accuracy": 0.21815359443426133,
      "num_tokens": 50629483.0,
      "step": 24400
    },
    {
      "entropy": 5.548216533660889,
      "epoch": 2.2138062409288826,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004509218818765211,
      "loss": 4.8681,
      "mean_token_accuracy": 0.2216430589556694,
      "num_tokens": 50638552.0,
      "step": 24405
    },
    {
      "entropy": 5.578045082092285,
      "epoch": 2.2142597968069664,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045090170792679544,
      "loss": 4.8847,
      "mean_token_accuracy": 0.21352774649858475,
      "num_tokens": 50648752.0,
      "step": 24410
    },
    {
      "entropy": 5.580213165283203,
      "epoch": 2.2147133526850507,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045088153033937765,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21593828201293946,
      "num_tokens": 50658373.0,
      "step": 24415
    },
    {
      "entropy": 5.544181299209595,
      "epoch": 2.215166908563135,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045086134911468496,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21816433668136598,
      "num_tokens": 50668529.0,
      "step": 24420
    },
    {
      "entropy": 5.570469760894776,
      "epoch": 2.215620464441219,
      "grad_norm": 1.25,
      "learning_rate": 0.0004508411642531347,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21682122796773912,
      "num_tokens": 50678038.0,
      "step": 24425
    },
    {
      "entropy": 5.5718841552734375,
      "epoch": 2.2160740203193035,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004508209757551445,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21935876458883286,
      "num_tokens": 50687788.0,
      "step": 24430
    },
    {
      "entropy": 5.594445133209229,
      "epoch": 2.2165275761973877,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045080078362113153,
      "loss": 4.9539,
      "mean_token_accuracy": 0.2059182718396187,
      "num_tokens": 50697748.0,
      "step": 24435
    },
    {
      "entropy": 5.504906702041626,
      "epoch": 2.2169811320754715,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045078058785151363,
      "loss": 4.8443,
      "mean_token_accuracy": 0.21902491003274918,
      "num_tokens": 50707776.0,
      "step": 24440
    },
    {
      "entropy": 5.534537315368652,
      "epoch": 2.217434687953556,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004507603884467085,
      "loss": 4.9992,
      "mean_token_accuracy": 0.20853210538625716,
      "num_tokens": 50717547.0,
      "step": 24445
    },
    {
      "entropy": 5.510666036605835,
      "epoch": 2.21788824383164,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004507401854071336,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22325698137283326,
      "num_tokens": 50728446.0,
      "step": 24450
    },
    {
      "entropy": 5.55709867477417,
      "epoch": 2.2183417997097243,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045071997873320687,
      "loss": 4.8459,
      "mean_token_accuracy": 0.21365610361099244,
      "num_tokens": 50737909.0,
      "step": 24455
    },
    {
      "entropy": 5.510608863830567,
      "epoch": 2.2187953555878086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045069976842534635,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21610202342271806,
      "num_tokens": 50749745.0,
      "step": 24460
    },
    {
      "entropy": 5.58241982460022,
      "epoch": 2.2192489114658924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004506795544839697,
      "loss": 4.9127,
      "mean_token_accuracy": 0.20932943969964982,
      "num_tokens": 50760073.0,
      "step": 24465
    },
    {
      "entropy": 5.608687257766723,
      "epoch": 2.2197024673439767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004506593369094951,
      "loss": 4.9105,
      "mean_token_accuracy": 0.20908024162054062,
      "num_tokens": 50769261.0,
      "step": 24470
    },
    {
      "entropy": 5.613238573074341,
      "epoch": 2.220156023222061,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045063911570234074,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20674648582935334,
      "num_tokens": 50779572.0,
      "step": 24475
    },
    {
      "entropy": 5.5163829803466795,
      "epoch": 2.2206095791001452,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004506188908629247,
      "loss": 4.8853,
      "mean_token_accuracy": 0.21855538785457612,
      "num_tokens": 50789893.0,
      "step": 24480
    },
    {
      "entropy": 5.482339859008789,
      "epoch": 2.2210631349782295,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004505986623916652,
      "loss": 4.8321,
      "mean_token_accuracy": 0.2160807654261589,
      "num_tokens": 50799177.0,
      "step": 24485
    },
    {
      "entropy": 5.515284681320191,
      "epoch": 2.2215166908563133,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004505784302889806,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22159485816955565,
      "num_tokens": 50810093.0,
      "step": 24490
    },
    {
      "entropy": 5.64695177078247,
      "epoch": 2.2219702467343976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045055819455528935,
      "loss": 5.0639,
      "mean_token_accuracy": 0.20666398406028746,
      "num_tokens": 50820272.0,
      "step": 24495
    },
    {
      "entropy": 5.598090410232544,
      "epoch": 2.222423802612482,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004505379551910099,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21202658116817474,
      "num_tokens": 50830301.0,
      "step": 24500
    },
    {
      "entropy": 5.6368327140808105,
      "epoch": 2.222877358490566,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004505177121965608,
      "loss": 4.9907,
      "mean_token_accuracy": 0.21438877284526825,
      "num_tokens": 50839937.0,
      "step": 24505
    },
    {
      "entropy": 5.54852991104126,
      "epoch": 2.2233309143686504,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045049746557236067,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22185829281806946,
      "num_tokens": 50850295.0,
      "step": 24510
    },
    {
      "entropy": 5.605092287063599,
      "epoch": 2.223784470246734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045047721531882827,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21104562729597093,
      "num_tokens": 50861889.0,
      "step": 24515
    },
    {
      "entropy": 5.591089344024658,
      "epoch": 2.2242380261248185,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004504569614363823,
      "loss": 4.917,
      "mean_token_accuracy": 0.21555061042308807,
      "num_tokens": 50872011.0,
      "step": 24520
    },
    {
      "entropy": 5.61662449836731,
      "epoch": 2.2246915820029027,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004504367039254416,
      "loss": 4.9594,
      "mean_token_accuracy": 0.20999687314033508,
      "num_tokens": 50882559.0,
      "step": 24525
    },
    {
      "entropy": 5.5419602394104,
      "epoch": 2.225145137880987,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045041644278642523,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20472318381071092,
      "num_tokens": 50894132.0,
      "step": 24530
    },
    {
      "entropy": 5.621619796752929,
      "epoch": 2.2255986937590713,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045039617801975214,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2130837544798851,
      "num_tokens": 50904337.0,
      "step": 24535
    },
    {
      "entropy": 5.63733286857605,
      "epoch": 2.2260522496371555,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045037590962584143,
      "loss": 4.9846,
      "mean_token_accuracy": 0.2069944381713867,
      "num_tokens": 50914961.0,
      "step": 24540
    },
    {
      "entropy": 5.613501644134521,
      "epoch": 2.2265058055152394,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004503556376051121,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21068182587623596,
      "num_tokens": 50925682.0,
      "step": 24545
    },
    {
      "entropy": 5.535657167434692,
      "epoch": 2.2269593613933236,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045033536195798355,
      "loss": 4.851,
      "mean_token_accuracy": 0.22736817598342896,
      "num_tokens": 50935809.0,
      "step": 24550
    },
    {
      "entropy": 5.619937372207642,
      "epoch": 2.227412917271408,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000450315082684875,
      "loss": 5.047,
      "mean_token_accuracy": 0.20100729316473007,
      "num_tokens": 50946098.0,
      "step": 24555
    },
    {
      "entropy": 5.615520238876343,
      "epoch": 2.227866473149492,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004502947997862059,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21025876849889755,
      "num_tokens": 50956240.0,
      "step": 24560
    },
    {
      "entropy": 5.579733180999756,
      "epoch": 2.228320029027576,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004502745132623957,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21890990436077118,
      "num_tokens": 50966870.0,
      "step": 24565
    },
    {
      "entropy": 5.560952377319336,
      "epoch": 2.2287735849056602,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004502542231138639,
      "loss": 4.9208,
      "mean_token_accuracy": 0.20925770401954652,
      "num_tokens": 50977528.0,
      "step": 24570
    },
    {
      "entropy": 5.581137466430664,
      "epoch": 2.2292271407837445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045023392934103005,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20734782069921492,
      "num_tokens": 50988565.0,
      "step": 24575
    },
    {
      "entropy": 5.579733753204346,
      "epoch": 2.2296806966618288,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045021363194431397,
      "loss": 4.992,
      "mean_token_accuracy": 0.21358343511819838,
      "num_tokens": 50998410.0,
      "step": 24580
    },
    {
      "entropy": 5.603824758529663,
      "epoch": 2.230134252539913,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004501933309241352,
      "loss": 4.9775,
      "mean_token_accuracy": 0.2056296244263649,
      "num_tokens": 51008890.0,
      "step": 24585
    },
    {
      "entropy": 5.516703844070435,
      "epoch": 2.2305878084179973,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004501730262809138,
      "loss": 4.9047,
      "mean_token_accuracy": 0.2169082432985306,
      "num_tokens": 51020264.0,
      "step": 24590
    },
    {
      "entropy": 5.5458611965179445,
      "epoch": 2.231041364296081,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004501527180150696,
      "loss": 4.8414,
      "mean_token_accuracy": 0.21861355155706405,
      "num_tokens": 51029892.0,
      "step": 24595
    },
    {
      "entropy": 5.640199327468872,
      "epoch": 2.2314949201741654,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004501324061270224,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20373291969299318,
      "num_tokens": 51040499.0,
      "step": 24600
    },
    {
      "entropy": 5.482412433624267,
      "epoch": 2.2319484760522497,
      "grad_norm": 1.125,
      "learning_rate": 0.0004501120906171926,
      "loss": 4.8561,
      "mean_token_accuracy": 0.21426650732755662,
      "num_tokens": 51050395.0,
      "step": 24605
    },
    {
      "entropy": 5.545130729675293,
      "epoch": 2.232402031930334,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045009177148600004,
      "loss": 4.8729,
      "mean_token_accuracy": 0.21900064647197723,
      "num_tokens": 51060316.0,
      "step": 24610
    },
    {
      "entropy": 5.520968961715698,
      "epoch": 2.232855587808418,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045007144873386514,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21584844291210176,
      "num_tokens": 51070032.0,
      "step": 24615
    },
    {
      "entropy": 5.623284912109375,
      "epoch": 2.233309143686502,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045005112236120795,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20193691551685333,
      "num_tokens": 51080738.0,
      "step": 24620
    },
    {
      "entropy": 5.575223064422607,
      "epoch": 2.2337626995645863,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000450030792368449,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2171100065112114,
      "num_tokens": 51090287.0,
      "step": 24625
    },
    {
      "entropy": 5.5425323963165285,
      "epoch": 2.2342162554426706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004500104587560086,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21029032915830612,
      "num_tokens": 51101188.0,
      "step": 24630
    },
    {
      "entropy": 5.523554563522339,
      "epoch": 2.234669811320755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004499901215243074,
      "loss": 4.8496,
      "mean_token_accuracy": 0.21467358469963074,
      "num_tokens": 51112310.0,
      "step": 24635
    },
    {
      "entropy": 5.584855222702027,
      "epoch": 2.235123367198839,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044996978067376575,
      "loss": 4.9789,
      "mean_token_accuracy": 0.20993858724832534,
      "num_tokens": 51122481.0,
      "step": 24640
    },
    {
      "entropy": 5.536005973815918,
      "epoch": 2.235576923076923,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044994943620480453,
      "loss": 4.9226,
      "mean_token_accuracy": 0.20727061331272126,
      "num_tokens": 51131918.0,
      "step": 24645
    },
    {
      "entropy": 5.561415958404541,
      "epoch": 2.236030478955007,
      "grad_norm": 1.0,
      "learning_rate": 0.00044992908811784446,
      "loss": 5.0119,
      "mean_token_accuracy": 0.19910293072462082,
      "num_tokens": 51142861.0,
      "step": 24650
    },
    {
      "entropy": 5.587594652175904,
      "epoch": 2.2364840348330914,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044990873641330615,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20912633687257767,
      "num_tokens": 51153956.0,
      "step": 24655
    },
    {
      "entropy": 5.579059219360351,
      "epoch": 2.2369375907111757,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004498883810916107,
      "loss": 4.8496,
      "mean_token_accuracy": 0.22160615921020507,
      "num_tokens": 51164499.0,
      "step": 24660
    },
    {
      "entropy": 5.60390830039978,
      "epoch": 2.23739114658926,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004498680221531789,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20653716325759888,
      "num_tokens": 51175722.0,
      "step": 24665
    },
    {
      "entropy": 5.558173799514771,
      "epoch": 2.237844702467344,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004498476595984319,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21457604616880416,
      "num_tokens": 51184597.0,
      "step": 24670
    },
    {
      "entropy": 5.50778603553772,
      "epoch": 2.238298258345428,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004498272934277906,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21343494057655335,
      "num_tokens": 51194751.0,
      "step": 24675
    },
    {
      "entropy": 5.573074197769165,
      "epoch": 2.2387518142235123,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044980692364167646,
      "loss": 4.923,
      "mean_token_accuracy": 0.2212190091609955,
      "num_tokens": 51204184.0,
      "step": 24680
    },
    {
      "entropy": 5.582958459854126,
      "epoch": 2.2392053701015966,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004497865502405105,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2145560324192047,
      "num_tokens": 51213761.0,
      "step": 24685
    },
    {
      "entropy": 5.572753095626831,
      "epoch": 2.239658925979681,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044976617322471424,
      "loss": 4.9087,
      "mean_token_accuracy": 0.2164181038737297,
      "num_tokens": 51224150.0,
      "step": 24690
    },
    {
      "entropy": 5.610567235946656,
      "epoch": 2.2401124818577647,
      "grad_norm": 1.171875,
      "learning_rate": 0.000449745792594709,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21368137151002883,
      "num_tokens": 51235133.0,
      "step": 24695
    },
    {
      "entropy": 5.604644346237182,
      "epoch": 2.240566037735849,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044972540835091617,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21150542795658112,
      "num_tokens": 51245632.0,
      "step": 24700
    },
    {
      "entropy": 5.56734790802002,
      "epoch": 2.241019593613933,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044970502049375736,
      "loss": 4.968,
      "mean_token_accuracy": 0.21566979140043258,
      "num_tokens": 51255747.0,
      "step": 24705
    },
    {
      "entropy": 5.56240406036377,
      "epoch": 2.2414731494920175,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004496846290236542,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21618969440460206,
      "num_tokens": 51265531.0,
      "step": 24710
    },
    {
      "entropy": 5.555978298187256,
      "epoch": 2.2419267053701017,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044966423394102845,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21604196578264237,
      "num_tokens": 51275319.0,
      "step": 24715
    },
    {
      "entropy": 5.575274419784546,
      "epoch": 2.2423802612481856,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044964383524630176,
      "loss": 4.9698,
      "mean_token_accuracy": 0.2166734054684639,
      "num_tokens": 51285655.0,
      "step": 24720
    },
    {
      "entropy": 5.512102842330933,
      "epoch": 2.24283381712627,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004496234329398961,
      "loss": 4.8978,
      "mean_token_accuracy": 0.20599962621927262,
      "num_tokens": 51295938.0,
      "step": 24725
    },
    {
      "entropy": 5.594706106185913,
      "epoch": 2.243287373004354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004496030270222333,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21997081786394118,
      "num_tokens": 51306483.0,
      "step": 24730
    },
    {
      "entropy": 5.577207040786743,
      "epoch": 2.2437409288824384,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044958261749373547,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2139873504638672,
      "num_tokens": 51316446.0,
      "step": 24735
    },
    {
      "entropy": 5.533952760696411,
      "epoch": 2.2441944847605226,
      "grad_norm": 1.40625,
      "learning_rate": 0.00044956220435482465,
      "loss": 4.9607,
      "mean_token_accuracy": 0.20779209733009338,
      "num_tokens": 51326840.0,
      "step": 24740
    },
    {
      "entropy": 5.570036602020264,
      "epoch": 2.244648040638607,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044954178760592283,
      "loss": 4.9931,
      "mean_token_accuracy": 0.2070079118013382,
      "num_tokens": 51337573.0,
      "step": 24745
    },
    {
      "entropy": 5.621497583389282,
      "epoch": 2.2451015965166907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044952136724745254,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2153195932507515,
      "num_tokens": 51347827.0,
      "step": 24750
    },
    {
      "entropy": 5.55088438987732,
      "epoch": 2.245555152394775,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004495009432798358,
      "loss": 4.925,
      "mean_token_accuracy": 0.2125857651233673,
      "num_tokens": 51358396.0,
      "step": 24755
    },
    {
      "entropy": 5.608096933364868,
      "epoch": 2.2460087082728593,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044948051570349503,
      "loss": 4.9478,
      "mean_token_accuracy": 0.20708179622888565,
      "num_tokens": 51368882.0,
      "step": 24760
    },
    {
      "entropy": 5.584758567810058,
      "epoch": 2.2464622641509435,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004494600845188528,
      "loss": 4.9998,
      "mean_token_accuracy": 0.2085038587450981,
      "num_tokens": 51378717.0,
      "step": 24765
    },
    {
      "entropy": 5.562956714630127,
      "epoch": 2.2469158200290273,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044943964972633154,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2069125398993492,
      "num_tokens": 51389270.0,
      "step": 24770
    },
    {
      "entropy": 5.55361909866333,
      "epoch": 2.2473693759071116,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004494192113263539,
      "loss": 4.952,
      "mean_token_accuracy": 0.21096873581409453,
      "num_tokens": 51400025.0,
      "step": 24775
    },
    {
      "entropy": 5.5834801197052,
      "epoch": 2.247822931785196,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044939876931934255,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2143678143620491,
      "num_tokens": 51410903.0,
      "step": 24780
    },
    {
      "entropy": 5.543683385848999,
      "epoch": 2.24827648766328,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004493783237057201,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2178712785243988,
      "num_tokens": 51420246.0,
      "step": 24785
    },
    {
      "entropy": 5.6401002407073975,
      "epoch": 2.2487300435413644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493578744859095,
      "loss": 5.0649,
      "mean_token_accuracy": 0.2057617723941803,
      "num_tokens": 51431309.0,
      "step": 24790
    },
    {
      "entropy": 5.633981227874756,
      "epoch": 2.2491835994194487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004493374216603338,
      "loss": 5.008,
      "mean_token_accuracy": 0.2052177146077156,
      "num_tokens": 51443142.0,
      "step": 24795
    },
    {
      "entropy": 5.68134036064148,
      "epoch": 2.2496371552975325,
      "grad_norm": 1.5,
      "learning_rate": 0.0004493169652294156,
      "loss": 5.158,
      "mean_token_accuracy": 0.19841213375329972,
      "num_tokens": 51454499.0,
      "step": 24800
    },
    {
      "entropy": 5.557585763931274,
      "epoch": 2.2500907111756168,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044929650519357823,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21476107835769653,
      "num_tokens": 51464667.0,
      "step": 24805
    },
    {
      "entropy": 5.621891021728516,
      "epoch": 2.250544267053701,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004492760415532447,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21079372763633727,
      "num_tokens": 51473627.0,
      "step": 24810
    },
    {
      "entropy": 5.571342897415161,
      "epoch": 2.2509978229317853,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004492555743088382,
      "loss": 4.815,
      "mean_token_accuracy": 0.21659711450338365,
      "num_tokens": 51483052.0,
      "step": 24815
    },
    {
      "entropy": 5.5308513164520265,
      "epoch": 2.2514513788098696,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000449235103460782,
      "loss": 4.9526,
      "mean_token_accuracy": 0.205110627412796,
      "num_tokens": 51492422.0,
      "step": 24820
    },
    {
      "entropy": 5.516646051406861,
      "epoch": 2.2519049346879534,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004492146290094995,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21152358651161193,
      "num_tokens": 51503566.0,
      "step": 24825
    },
    {
      "entropy": 5.539857292175293,
      "epoch": 2.2523584905660377,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044919415095541406,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21697582751512529,
      "num_tokens": 51513699.0,
      "step": 24830
    },
    {
      "entropy": 5.566490268707275,
      "epoch": 2.252812046444122,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044917366929894914,
      "loss": 4.9397,
      "mean_token_accuracy": 0.21021585911512375,
      "num_tokens": 51524745.0,
      "step": 24835
    },
    {
      "entropy": 5.557812213897705,
      "epoch": 2.253265602322206,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044915318404052846,
      "loss": 4.8814,
      "mean_token_accuracy": 0.2129119500517845,
      "num_tokens": 51536516.0,
      "step": 24840
    },
    {
      "entropy": 5.608424949645996,
      "epoch": 2.2537191582002905,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044913269518057547,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20357286930084229,
      "num_tokens": 51547063.0,
      "step": 24845
    },
    {
      "entropy": 5.580503749847412,
      "epoch": 2.2541727140783743,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000449112202719514,
      "loss": 4.9343,
      "mean_token_accuracy": 0.20828542709350586,
      "num_tokens": 51557244.0,
      "step": 24850
    },
    {
      "entropy": 5.6066258430480955,
      "epoch": 2.2546262699564585,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004490917066577678,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2110108882188797,
      "num_tokens": 51567630.0,
      "step": 24855
    },
    {
      "entropy": 5.545233631134034,
      "epoch": 2.255079825834543,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044907120699576075,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21206414699554443,
      "num_tokens": 51577951.0,
      "step": 24860
    },
    {
      "entropy": 5.606132650375367,
      "epoch": 2.255533381712627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044905070373391684,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21519958674907685,
      "num_tokens": 51587533.0,
      "step": 24865
    },
    {
      "entropy": 5.653391981124878,
      "epoch": 2.2559869375907113,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004490301968726599,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22210706323385238,
      "num_tokens": 51597212.0,
      "step": 24870
    },
    {
      "entropy": 5.57596230506897,
      "epoch": 2.256440493468795,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044900968641241426,
      "loss": 4.9688,
      "mean_token_accuracy": 0.21647840589284897,
      "num_tokens": 51608520.0,
      "step": 24875
    },
    {
      "entropy": 5.585930585861206,
      "epoch": 2.2568940493468794,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004489891723536039,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20756766498088836,
      "num_tokens": 51619651.0,
      "step": 24880
    },
    {
      "entropy": 5.539716196060181,
      "epoch": 2.2573476052249637,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004489686546966531,
      "loss": 4.849,
      "mean_token_accuracy": 0.21233907490968704,
      "num_tokens": 51630184.0,
      "step": 24885
    },
    {
      "entropy": 5.499330997467041,
      "epoch": 2.257801161103048,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044894813344198626,
      "loss": 4.8481,
      "mean_token_accuracy": 0.21994235813617707,
      "num_tokens": 51641293.0,
      "step": 24890
    },
    {
      "entropy": 5.581316947937012,
      "epoch": 2.2582547169811322,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004489276085900276,
      "loss": 4.945,
      "mean_token_accuracy": 0.21121911704540253,
      "num_tokens": 51652002.0,
      "step": 24895
    },
    {
      "entropy": 5.49250864982605,
      "epoch": 2.2587082728592165,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044890708014120174,
      "loss": 4.8225,
      "mean_token_accuracy": 0.227053464949131,
      "num_tokens": 51661668.0,
      "step": 24900
    },
    {
      "entropy": 5.619518613815307,
      "epoch": 2.2591618287373003,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004488865480959331,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20568533092737198,
      "num_tokens": 51672879.0,
      "step": 24905
    },
    {
      "entropy": 5.578426885604858,
      "epoch": 2.2596153846153846,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004488660124546464,
      "loss": 4.9223,
      "mean_token_accuracy": 0.21344517469406127,
      "num_tokens": 51683878.0,
      "step": 24910
    },
    {
      "entropy": 5.5278290748596195,
      "epoch": 2.260068940493469,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004488454732177662,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2105911359190941,
      "num_tokens": 51693971.0,
      "step": 24915
    },
    {
      "entropy": 5.551317501068115,
      "epoch": 2.260522496371553,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044882493038571737,
      "loss": 4.8475,
      "mean_token_accuracy": 0.22081339955329896,
      "num_tokens": 51703685.0,
      "step": 24920
    },
    {
      "entropy": 5.635507583618164,
      "epoch": 2.260976052249637,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044880438395892457,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2090151622891426,
      "num_tokens": 51714469.0,
      "step": 24925
    },
    {
      "entropy": 5.596588516235352,
      "epoch": 2.261429608127721,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004487838339378129,
      "loss": 5.0244,
      "mean_token_accuracy": 0.20091243386268615,
      "num_tokens": 51725283.0,
      "step": 24930
    },
    {
      "entropy": 5.5576558113098145,
      "epoch": 2.2618831640058055,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004487632803228072,
      "loss": 4.9371,
      "mean_token_accuracy": 0.20955396443605423,
      "num_tokens": 51735628.0,
      "step": 24935
    },
    {
      "entropy": 5.463111448287964,
      "epoch": 2.2623367198838897,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044874272311433265,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22551854848861694,
      "num_tokens": 51745991.0,
      "step": 24940
    },
    {
      "entropy": 5.517724275588989,
      "epoch": 2.262790275761974,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004487221623128143,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2159314587712288,
      "num_tokens": 51756306.0,
      "step": 24945
    },
    {
      "entropy": 5.593941879272461,
      "epoch": 2.2632438316400583,
      "grad_norm": 1.125,
      "learning_rate": 0.00044870159791867743,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21535757929086685,
      "num_tokens": 51765466.0,
      "step": 24950
    },
    {
      "entropy": 5.592212200164795,
      "epoch": 2.263697387518142,
      "grad_norm": 1.125,
      "learning_rate": 0.00044868102993234716,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21097905188798904,
      "num_tokens": 51775526.0,
      "step": 24955
    },
    {
      "entropy": 5.590959644317627,
      "epoch": 2.2641509433962264,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044866045835424896,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20720784664154052,
      "num_tokens": 51785434.0,
      "step": 24960
    },
    {
      "entropy": 5.5903346061706545,
      "epoch": 2.2646044992743106,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004486398831848082,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2029450610280037,
      "num_tokens": 51796047.0,
      "step": 24965
    },
    {
      "entropy": 5.624813604354858,
      "epoch": 2.265058055152395,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044861930442445037,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20542479157447815,
      "num_tokens": 51806173.0,
      "step": 24970
    },
    {
      "entropy": 5.6734966278076175,
      "epoch": 2.2655116110304787,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004485987220736012,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20142967998981476,
      "num_tokens": 51817644.0,
      "step": 24975
    },
    {
      "entropy": 5.6251873016357425,
      "epoch": 2.265965166908563,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004485781361326861,
      "loss": 5.0719,
      "mean_token_accuracy": 0.20615520179271699,
      "num_tokens": 51829746.0,
      "step": 24980
    },
    {
      "entropy": 5.575164556503296,
      "epoch": 2.2664187227866472,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000448557546602131,
      "loss": 5.0051,
      "mean_token_accuracy": 0.20462606251239776,
      "num_tokens": 51840868.0,
      "step": 24985
    },
    {
      "entropy": 5.597962856292725,
      "epoch": 2.2668722786647315,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044853695348236156,
      "loss": 4.8825,
      "mean_token_accuracy": 0.2149131953716278,
      "num_tokens": 51851385.0,
      "step": 24990
    },
    {
      "entropy": 5.585072469711304,
      "epoch": 2.2673258345428158,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044851635677380374,
      "loss": 4.915,
      "mean_token_accuracy": 0.20616335272789002,
      "num_tokens": 51861021.0,
      "step": 24995
    },
    {
      "entropy": 5.59707498550415,
      "epoch": 2.2677793904209,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004484957564768834,
      "loss": 5.0089,
      "mean_token_accuracy": 0.2096236154437065,
      "num_tokens": 51870688.0,
      "step": 25000
    },
    {
      "entropy": 5.583831310272217,
      "epoch": 2.268232946298984,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044847515259202657,
      "loss": 4.9627,
      "mean_token_accuracy": 0.20856474786996843,
      "num_tokens": 51880439.0,
      "step": 25005
    },
    {
      "entropy": 5.505197811126709,
      "epoch": 2.268686502177068,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044845454511965947,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2147160291671753,
      "num_tokens": 51890861.0,
      "step": 25010
    },
    {
      "entropy": 5.592146205902099,
      "epoch": 2.2691400580551524,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044843393406020823,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21316009610891343,
      "num_tokens": 51900687.0,
      "step": 25015
    },
    {
      "entropy": 5.636794376373291,
      "epoch": 2.2695936139332367,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044841331941409884,
      "loss": 5.0809,
      "mean_token_accuracy": 0.19589564800262452,
      "num_tokens": 51911805.0,
      "step": 25020
    },
    {
      "entropy": 5.581126880645752,
      "epoch": 2.270047169811321,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044839270118175793,
      "loss": 4.8645,
      "mean_token_accuracy": 0.2211558312177658,
      "num_tokens": 51922520.0,
      "step": 25025
    },
    {
      "entropy": 5.6436036109924315,
      "epoch": 2.2705007256894048,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004483720793636118,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2037653461098671,
      "num_tokens": 51934056.0,
      "step": 25030
    },
    {
      "entropy": 5.606632852554322,
      "epoch": 2.270954281567489,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044835145396008685,
      "loss": 5.0042,
      "mean_token_accuracy": 0.2117276072502136,
      "num_tokens": 51944696.0,
      "step": 25035
    },
    {
      "entropy": 5.610119152069092,
      "epoch": 2.2714078374455733,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004483308249716097,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21197018772363663,
      "num_tokens": 51955111.0,
      "step": 25040
    },
    {
      "entropy": 5.563672399520874,
      "epoch": 2.2718613933236576,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004483101923986068,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21111331433057784,
      "num_tokens": 51966036.0,
      "step": 25045
    },
    {
      "entropy": 5.5305970191955565,
      "epoch": 2.272314949201742,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044828955624150505,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21775335520505906,
      "num_tokens": 51976060.0,
      "step": 25050
    },
    {
      "entropy": 5.594537305831909,
      "epoch": 2.2727685050798256,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004482689165007311,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21135286688804628,
      "num_tokens": 51986016.0,
      "step": 25055
    },
    {
      "entropy": 5.552269363403321,
      "epoch": 2.27322206095791,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044824827317671177,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22400961071252823,
      "num_tokens": 51994794.0,
      "step": 25060
    },
    {
      "entropy": 5.499716854095459,
      "epoch": 2.273675616835994,
      "grad_norm": 1.109375,
      "learning_rate": 0.000448227626269874,
      "loss": 4.8706,
      "mean_token_accuracy": 0.2188450112938881,
      "num_tokens": 52004976.0,
      "step": 25065
    },
    {
      "entropy": 5.589347171783447,
      "epoch": 2.2741291727140784,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044820697578064485,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20299387872219085,
      "num_tokens": 52016380.0,
      "step": 25070
    },
    {
      "entropy": 5.6079377174377445,
      "epoch": 2.2745827285921627,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004481863217094513,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21439362019300462,
      "num_tokens": 52026439.0,
      "step": 25075
    },
    {
      "entropy": 5.564224910736084,
      "epoch": 2.2750362844702465,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004481656640567204,
      "loss": 4.8869,
      "mean_token_accuracy": 0.21509212106466294,
      "num_tokens": 52037111.0,
      "step": 25080
    },
    {
      "entropy": 5.531151485443115,
      "epoch": 2.275489840348331,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004481450028228795,
      "loss": 4.944,
      "mean_token_accuracy": 0.21160723865032197,
      "num_tokens": 52047463.0,
      "step": 25085
    },
    {
      "entropy": 5.566102647781372,
      "epoch": 2.275943396226415,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004481243380083558,
      "loss": 4.9656,
      "mean_token_accuracy": 0.20788409560918808,
      "num_tokens": 52058081.0,
      "step": 25090
    },
    {
      "entropy": 5.564450120925903,
      "epoch": 2.2763969521044993,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004481036696135767,
      "loss": 4.9644,
      "mean_token_accuracy": 0.21422482430934905,
      "num_tokens": 52069367.0,
      "step": 25095
    },
    {
      "entropy": 5.593443250656128,
      "epoch": 2.2768505079825836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004480829976389695,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21186336874961853,
      "num_tokens": 52080023.0,
      "step": 25100
    },
    {
      "entropy": 5.5127521514892575,
      "epoch": 2.277304063860668,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044806232208496194,
      "loss": 4.965,
      "mean_token_accuracy": 0.21391810327768326,
      "num_tokens": 52089184.0,
      "step": 25105
    },
    {
      "entropy": 5.615357303619385,
      "epoch": 2.2777576197387517,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004480416429519814,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20993676781654358,
      "num_tokens": 52098711.0,
      "step": 25110
    },
    {
      "entropy": 5.608211612701416,
      "epoch": 2.278211175616836,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004480209602404556,
      "loss": 4.9117,
      "mean_token_accuracy": 0.21644964814186096,
      "num_tokens": 52108826.0,
      "step": 25115
    },
    {
      "entropy": 5.5315389156341555,
      "epoch": 2.27866473149492,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004480002739508122,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22185250818729402,
      "num_tokens": 52118934.0,
      "step": 25120
    },
    {
      "entropy": 5.5315416812896725,
      "epoch": 2.2791182873730045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004479795840834791,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22009823471307755,
      "num_tokens": 52129803.0,
      "step": 25125
    },
    {
      "entropy": 5.482956600189209,
      "epoch": 2.2795718432510883,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044795889063888413,
      "loss": 4.8142,
      "mean_token_accuracy": 0.22065597474575044,
      "num_tokens": 52140070.0,
      "step": 25130
    },
    {
      "entropy": 5.682845640182495,
      "epoch": 2.2800253991291726,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004479381936174552,
      "loss": 5.0843,
      "mean_token_accuracy": 0.20483203381299972,
      "num_tokens": 52150462.0,
      "step": 25135
    },
    {
      "entropy": 5.65266170501709,
      "epoch": 2.280478955007257,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004479174930196204,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2097917303442955,
      "num_tokens": 52160552.0,
      "step": 25140
    },
    {
      "entropy": 5.487208604812622,
      "epoch": 2.280932510885341,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044789678884580767,
      "loss": 4.8523,
      "mean_token_accuracy": 0.21823084056377412,
      "num_tokens": 52169780.0,
      "step": 25145
    },
    {
      "entropy": 5.578747701644898,
      "epoch": 2.2813860667634254,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044787608109644536,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21678511202335357,
      "num_tokens": 52180055.0,
      "step": 25150
    },
    {
      "entropy": 5.577519845962525,
      "epoch": 2.2818396226415096,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004478553697719616,
      "loss": 4.9053,
      "mean_token_accuracy": 0.21233561933040618,
      "num_tokens": 52189772.0,
      "step": 25155
    },
    {
      "entropy": 5.558465003967285,
      "epoch": 2.2822931785195935,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004478346548727848,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22545824497938155,
      "num_tokens": 52199608.0,
      "step": 25160
    },
    {
      "entropy": 5.523789119720459,
      "epoch": 2.2827467343976777,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004478139363993433,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2116127207875252,
      "num_tokens": 52208236.0,
      "step": 25165
    },
    {
      "entropy": 5.589986658096313,
      "epoch": 2.283200290275762,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004477932143520655,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21422459483146666,
      "num_tokens": 52218403.0,
      "step": 25170
    },
    {
      "entropy": 5.614393663406372,
      "epoch": 2.2836538461538463,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044777248873138005,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21445739418268203,
      "num_tokens": 52228255.0,
      "step": 25175
    },
    {
      "entropy": 5.5787577629089355,
      "epoch": 2.2841074020319305,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044775175953771547,
      "loss": 4.9651,
      "mean_token_accuracy": 0.20945563316345214,
      "num_tokens": 52237732.0,
      "step": 25180
    },
    {
      "entropy": 5.664718961715698,
      "epoch": 2.2845609579100143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004477310267715005,
      "loss": 5.1079,
      "mean_token_accuracy": 0.2007945492863655,
      "num_tokens": 52249226.0,
      "step": 25185
    },
    {
      "entropy": 5.631059408187866,
      "epoch": 2.2850145137880986,
      "grad_norm": 1.125,
      "learning_rate": 0.0004477102904331639,
      "loss": 4.937,
      "mean_token_accuracy": 0.2161140561103821,
      "num_tokens": 52259550.0,
      "step": 25190
    },
    {
      "entropy": 5.632875394821167,
      "epoch": 2.285468069666183,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044768955052313443,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20975142866373062,
      "num_tokens": 52269670.0,
      "step": 25195
    },
    {
      "entropy": 5.618857431411743,
      "epoch": 2.285921625544267,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044766880704184104,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2161665678024292,
      "num_tokens": 52280135.0,
      "step": 25200
    },
    {
      "entropy": 5.545126342773438,
      "epoch": 2.2863751814223514,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004476480599897128,
      "loss": 4.8653,
      "mean_token_accuracy": 0.21941546797752381,
      "num_tokens": 52291806.0,
      "step": 25205
    },
    {
      "entropy": 5.500838422775269,
      "epoch": 2.2868287373004352,
      "grad_norm": 1.125,
      "learning_rate": 0.0004476273093671786,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2143415242433548,
      "num_tokens": 52302118.0,
      "step": 25210
    },
    {
      "entropy": 5.600105094909668,
      "epoch": 2.2872822931785195,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044760655517466766,
      "loss": 5.0065,
      "mean_token_accuracy": 0.2031088277697563,
      "num_tokens": 52312762.0,
      "step": 25215
    },
    {
      "entropy": 5.611731958389282,
      "epoch": 2.2877358490566038,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044758579741260916,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2034713312983513,
      "num_tokens": 52323367.0,
      "step": 25220
    },
    {
      "entropy": 5.64713225364685,
      "epoch": 2.288189404934688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044756503608143247,
      "loss": 5.0154,
      "mean_token_accuracy": 0.20669188350439072,
      "num_tokens": 52335231.0,
      "step": 25225
    },
    {
      "entropy": 5.631257104873657,
      "epoch": 2.2886429608127723,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044754427118156675,
      "loss": 4.895,
      "mean_token_accuracy": 0.21595149636268615,
      "num_tokens": 52345066.0,
      "step": 25230
    },
    {
      "entropy": 5.583748435974121,
      "epoch": 2.289096516690856,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004475235027134416,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21480559706687927,
      "num_tokens": 52356383.0,
      "step": 25235
    },
    {
      "entropy": 5.626114797592163,
      "epoch": 2.2895500725689404,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004475027306774865,
      "loss": 5.058,
      "mean_token_accuracy": 0.21025279760360718,
      "num_tokens": 52366013.0,
      "step": 25240
    },
    {
      "entropy": 5.597537517547607,
      "epoch": 2.2900036284470247,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044748195507413095,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21081256568431855,
      "num_tokens": 52376298.0,
      "step": 25245
    },
    {
      "entropy": 5.483092641830444,
      "epoch": 2.290457184325109,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044746117590380456,
      "loss": 4.8581,
      "mean_token_accuracy": 0.214636367559433,
      "num_tokens": 52387134.0,
      "step": 25250
    },
    {
      "entropy": 5.540257787704467,
      "epoch": 2.290910740203193,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004474403931669371,
      "loss": 4.8492,
      "mean_token_accuracy": 0.21721140146255494,
      "num_tokens": 52397389.0,
      "step": 25255
    },
    {
      "entropy": 5.5425032615661625,
      "epoch": 2.2913642960812775,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004474196068639584,
      "loss": 4.8719,
      "mean_token_accuracy": 0.2169995740056038,
      "num_tokens": 52407456.0,
      "step": 25260
    },
    {
      "entropy": 5.502102088928223,
      "epoch": 2.2918178519593613,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004473988169952983,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21231054067611693,
      "num_tokens": 52417967.0,
      "step": 25265
    },
    {
      "entropy": 5.617834186553955,
      "epoch": 2.2922714078374455,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044737802356138674,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21403928399085997,
      "num_tokens": 52428757.0,
      "step": 25270
    },
    {
      "entropy": 5.58148341178894,
      "epoch": 2.29272496371553,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004473572265626537,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2188520073890686,
      "num_tokens": 52438973.0,
      "step": 25275
    },
    {
      "entropy": 5.55060076713562,
      "epoch": 2.293178519593614,
      "grad_norm": 1.125,
      "learning_rate": 0.0004473364259995293,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21128974258899688,
      "num_tokens": 52449323.0,
      "step": 25280
    },
    {
      "entropy": 5.576638221740723,
      "epoch": 2.293632075471698,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004473156218724436,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21800605207681656,
      "num_tokens": 52459299.0,
      "step": 25285
    },
    {
      "entropy": 5.477514743804932,
      "epoch": 2.294085631349782,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044729481418182704,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21173993051052092,
      "num_tokens": 52470026.0,
      "step": 25290
    },
    {
      "entropy": 5.584178495407104,
      "epoch": 2.2945391872278664,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044727400292810985,
      "loss": 4.8852,
      "mean_token_accuracy": 0.21646443158388137,
      "num_tokens": 52479975.0,
      "step": 25295
    },
    {
      "entropy": 5.638037490844726,
      "epoch": 2.2949927431059507,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044725318811172227,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20517731010913848,
      "num_tokens": 52491067.0,
      "step": 25300
    },
    {
      "entropy": 5.529987907409668,
      "epoch": 2.295446298984035,
      "grad_norm": 1.109375,
      "learning_rate": 0.000447232369733095,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22039924412965775,
      "num_tokens": 52500671.0,
      "step": 25305
    },
    {
      "entropy": 5.567880916595459,
      "epoch": 2.2958998548621192,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004472115477926583,
      "loss": 4.8843,
      "mean_token_accuracy": 0.21178943365812303,
      "num_tokens": 52509910.0,
      "step": 25310
    },
    {
      "entropy": 5.56077823638916,
      "epoch": 2.296353410740203,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044719072229084295,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22036605179309846,
      "num_tokens": 52520517.0,
      "step": 25315
    },
    {
      "entropy": 5.552681875228882,
      "epoch": 2.2968069666182873,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004471698932280796,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21290714740753175,
      "num_tokens": 52530760.0,
      "step": 25320
    },
    {
      "entropy": 5.6255382061004635,
      "epoch": 2.2972605224963716,
      "grad_norm": 1.1875,
      "learning_rate": 0.000447149060604799,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20467558205127717,
      "num_tokens": 52541791.0,
      "step": 25325
    },
    {
      "entropy": 5.683467864990234,
      "epoch": 2.297714078374456,
      "grad_norm": 1.125,
      "learning_rate": 0.00044712822442143194,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2148816168308258,
      "num_tokens": 52551885.0,
      "step": 25330
    },
    {
      "entropy": 5.559752464294434,
      "epoch": 2.2981676342525397,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044710738467840927,
      "loss": 4.89,
      "mean_token_accuracy": 0.214259734749794,
      "num_tokens": 52563080.0,
      "step": 25335
    },
    {
      "entropy": 5.5173670768737795,
      "epoch": 2.298621190130624,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004470865413761621,
      "loss": 4.8557,
      "mean_token_accuracy": 0.22494282871484755,
      "num_tokens": 52573564.0,
      "step": 25340
    },
    {
      "entropy": 5.560221719741821,
      "epoch": 2.299074746008708,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044706569451512143,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21242703199386598,
      "num_tokens": 52583801.0,
      "step": 25345
    },
    {
      "entropy": 5.5933174133300785,
      "epoch": 2.2995283018867925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044704484409571825,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2079051613807678,
      "num_tokens": 52594908.0,
      "step": 25350
    },
    {
      "entropy": 5.555398559570312,
      "epoch": 2.2999818577648767,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044702399011838393,
      "loss": 4.8865,
      "mean_token_accuracy": 0.20889928340911865,
      "num_tokens": 52604861.0,
      "step": 25355
    },
    {
      "entropy": 5.5551429271698,
      "epoch": 2.300435413642961,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004470031325835497,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2133139193058014,
      "num_tokens": 52616203.0,
      "step": 25360
    },
    {
      "entropy": 5.527337265014649,
      "epoch": 2.300888969521045,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004469822714916468,
      "loss": 4.8687,
      "mean_token_accuracy": 0.21441444605588914,
      "num_tokens": 52625838.0,
      "step": 25365
    },
    {
      "entropy": 5.484132194519043,
      "epoch": 2.301342525399129,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004469614068431067,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2139444500207901,
      "num_tokens": 52635741.0,
      "step": 25370
    },
    {
      "entropy": 5.520980930328369,
      "epoch": 2.3017960812772134,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004469405386383608,
      "loss": 4.8594,
      "mean_token_accuracy": 0.21493380516767502,
      "num_tokens": 52645948.0,
      "step": 25375
    },
    {
      "entropy": 5.585295009613037,
      "epoch": 2.3022496371552976,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044691966687784087,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21457646936178207,
      "num_tokens": 52655073.0,
      "step": 25380
    },
    {
      "entropy": 5.547964763641358,
      "epoch": 2.302703193033382,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044689879156197833,
      "loss": 4.904,
      "mean_token_accuracy": 0.2276657536625862,
      "num_tokens": 52664429.0,
      "step": 25385
    },
    {
      "entropy": 5.62707109451294,
      "epoch": 2.3031567489114657,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044687791269120504,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2093101218342781,
      "num_tokens": 52674890.0,
      "step": 25390
    },
    {
      "entropy": 5.681874513626099,
      "epoch": 2.30361030478955,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044685703026595266,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2142328441143036,
      "num_tokens": 52685673.0,
      "step": 25395
    },
    {
      "entropy": 5.667135286331177,
      "epoch": 2.3040638606676342,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044683614428665314,
      "loss": 5.0367,
      "mean_token_accuracy": 0.20860655307769777,
      "num_tokens": 52696696.0,
      "step": 25400
    },
    {
      "entropy": 5.587316799163818,
      "epoch": 2.3045174165457185,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044681525475373843,
      "loss": 4.9836,
      "mean_token_accuracy": 0.20681025832891464,
      "num_tokens": 52707662.0,
      "step": 25405
    },
    {
      "entropy": 5.542032146453858,
      "epoch": 2.3049709724238028,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004467943616676403,
      "loss": 4.9186,
      "mean_token_accuracy": 0.217025026679039,
      "num_tokens": 52718438.0,
      "step": 25410
    },
    {
      "entropy": 5.538061571121216,
      "epoch": 2.3054245283018866,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004467734650287911,
      "loss": 4.9106,
      "mean_token_accuracy": 0.21316201984882355,
      "num_tokens": 52729267.0,
      "step": 25415
    },
    {
      "entropy": 5.536345815658569,
      "epoch": 2.305878084179971,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004467525648376229,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22299905866384506,
      "num_tokens": 52738754.0,
      "step": 25420
    },
    {
      "entropy": 5.513626670837402,
      "epoch": 2.306331640058055,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004467316610945679,
      "loss": 4.9016,
      "mean_token_accuracy": 0.21875788271427155,
      "num_tokens": 52749086.0,
      "step": 25425
    },
    {
      "entropy": 5.627924585342408,
      "epoch": 2.3067851959361394,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044671075380005825,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20999130755662918,
      "num_tokens": 52758961.0,
      "step": 25430
    },
    {
      "entropy": 5.613173151016236,
      "epoch": 2.3072387518142237,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004466898429545266,
      "loss": 4.9584,
      "mean_token_accuracy": 0.21569590419530868,
      "num_tokens": 52769447.0,
      "step": 25435
    },
    {
      "entropy": 5.646428394317627,
      "epoch": 2.3076923076923075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044666892855840513,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2048293337225914,
      "num_tokens": 52780306.0,
      "step": 25440
    },
    {
      "entropy": 5.551281499862671,
      "epoch": 2.3081458635703918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004466480106121266,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2122344419360161,
      "num_tokens": 52791263.0,
      "step": 25445
    },
    {
      "entropy": 5.594593191146851,
      "epoch": 2.308599419448476,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004466270891161234,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2077154964208603,
      "num_tokens": 52802194.0,
      "step": 25450
    },
    {
      "entropy": 5.5721752643585205,
      "epoch": 2.3090529753265603,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044660616407082827,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22245496213436128,
      "num_tokens": 52812391.0,
      "step": 25455
    },
    {
      "entropy": 5.605469846725464,
      "epoch": 2.3095065312046446,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044658523547667385,
      "loss": 4.9224,
      "mean_token_accuracy": 0.20990120321512223,
      "num_tokens": 52823540.0,
      "step": 25460
    },
    {
      "entropy": 5.529719972610474,
      "epoch": 2.309960087082729,
      "grad_norm": 1.125,
      "learning_rate": 0.0004465643033340931,
      "loss": 4.8675,
      "mean_token_accuracy": 0.21658777296543122,
      "num_tokens": 52833238.0,
      "step": 25465
    },
    {
      "entropy": 5.609346723556518,
      "epoch": 2.3104136429608126,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044654336764351886,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20375905334949493,
      "num_tokens": 52844510.0,
      "step": 25470
    },
    {
      "entropy": 5.580329370498657,
      "epoch": 2.310867198838897,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044652242840538406,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21358157694339752,
      "num_tokens": 52856311.0,
      "step": 25475
    },
    {
      "entropy": 5.528303480148315,
      "epoch": 2.311320754716981,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004465014856201218,
      "loss": 4.8753,
      "mean_token_accuracy": 0.22072956562042237,
      "num_tokens": 52866295.0,
      "step": 25480
    },
    {
      "entropy": 5.583780813217163,
      "epoch": 2.3117743105950654,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000446480539288165,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21221563816070557,
      "num_tokens": 52876818.0,
      "step": 25485
    },
    {
      "entropy": 5.623132562637329,
      "epoch": 2.3122278664731493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004464595894099469,
      "loss": 4.896,
      "mean_token_accuracy": 0.21328379809856415,
      "num_tokens": 52886173.0,
      "step": 25490
    },
    {
      "entropy": 5.517612457275391,
      "epoch": 2.3126814223512335,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044643863598590094,
      "loss": 4.9182,
      "mean_token_accuracy": 0.20874154418706894,
      "num_tokens": 52895875.0,
      "step": 25495
    },
    {
      "entropy": 5.586219549179077,
      "epoch": 2.313134978229318,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044641767901646025,
      "loss": 5.0206,
      "mean_token_accuracy": 0.20557665228843688,
      "num_tokens": 52906465.0,
      "step": 25500
    },
    {
      "entropy": 5.626094198226928,
      "epoch": 2.313588534107402,
      "grad_norm": 1.25,
      "learning_rate": 0.0004463967185020583,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21509349197149277,
      "num_tokens": 52916378.0,
      "step": 25505
    },
    {
      "entropy": 5.608964347839356,
      "epoch": 2.3140420899854863,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004463757544431285,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21702294945716857,
      "num_tokens": 52926698.0,
      "step": 25510
    },
    {
      "entropy": 5.637996673583984,
      "epoch": 2.3144956458635706,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004463547868401044,
      "loss": 4.9439,
      "mean_token_accuracy": 0.20812619775533675,
      "num_tokens": 52936469.0,
      "step": 25515
    },
    {
      "entropy": 5.615101003646851,
      "epoch": 2.3149492017416544,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044633381569341964,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21329115629196166,
      "num_tokens": 52947006.0,
      "step": 25520
    },
    {
      "entropy": 5.484480094909668,
      "epoch": 2.3154027576197387,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004463128410035079,
      "loss": 4.8253,
      "mean_token_accuracy": 0.2190695196390152,
      "num_tokens": 52956663.0,
      "step": 25525
    },
    {
      "entropy": 5.544159984588623,
      "epoch": 2.315856313497823,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044629186277080307,
      "loss": 4.8658,
      "mean_token_accuracy": 0.2122292175889015,
      "num_tokens": 52967383.0,
      "step": 25530
    },
    {
      "entropy": 5.564082908630371,
      "epoch": 2.316309869375907,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044627088099573884,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22225950062274932,
      "num_tokens": 52976567.0,
      "step": 25535
    },
    {
      "entropy": 5.544712543487549,
      "epoch": 2.3167634252539915,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044624989567874907,
      "loss": 4.938,
      "mean_token_accuracy": 0.21366299390792848,
      "num_tokens": 52986159.0,
      "step": 25540
    },
    {
      "entropy": 5.560599517822266,
      "epoch": 2.3172169811320753,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044622890682026787,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20261744856834413,
      "num_tokens": 52996028.0,
      "step": 25545
    },
    {
      "entropy": 5.6129392147064205,
      "epoch": 2.3176705370101596,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004462079144207292,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21037234514951705,
      "num_tokens": 53006446.0,
      "step": 25550
    },
    {
      "entropy": 5.630120658874512,
      "epoch": 2.318124092888244,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044618691848056726,
      "loss": 4.8909,
      "mean_token_accuracy": 0.21305860728025436,
      "num_tokens": 53016497.0,
      "step": 25555
    },
    {
      "entropy": 5.611555194854736,
      "epoch": 2.318577648766328,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004461659190002162,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21871512085199357,
      "num_tokens": 53027811.0,
      "step": 25560
    },
    {
      "entropy": 5.590132474899292,
      "epoch": 2.3190312046444124,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044614491598011027,
      "loss": 4.9358,
      "mean_token_accuracy": 0.20619187206029893,
      "num_tokens": 53038206.0,
      "step": 25565
    },
    {
      "entropy": 5.620518922805786,
      "epoch": 2.319484760522496,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004461239094206839,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21192284673452377,
      "num_tokens": 53047974.0,
      "step": 25570
    },
    {
      "entropy": 5.613166236877442,
      "epoch": 2.3199383164005805,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044610289932237154,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21265901774168014,
      "num_tokens": 53057904.0,
      "step": 25575
    },
    {
      "entropy": 5.526574945449829,
      "epoch": 2.3203918722786647,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044608188568560756,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21119341105222703,
      "num_tokens": 53068447.0,
      "step": 25580
    },
    {
      "entropy": 5.556252574920654,
      "epoch": 2.320845428156749,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004460608685108266,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2042188212275505,
      "num_tokens": 53079369.0,
      "step": 25585
    },
    {
      "entropy": 5.6675385475158695,
      "epoch": 2.3212989840348333,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044603984779846334,
      "loss": 4.9963,
      "mean_token_accuracy": 0.2049856588244438,
      "num_tokens": 53090794.0,
      "step": 25590
    },
    {
      "entropy": 5.603202867507934,
      "epoch": 2.321752539912917,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004460188235489524,
      "loss": 4.8846,
      "mean_token_accuracy": 0.21291725039482118,
      "num_tokens": 53100802.0,
      "step": 25595
    },
    {
      "entropy": 5.505622053146363,
      "epoch": 2.3222060957910013,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004459977957627285,
      "loss": 4.9722,
      "mean_token_accuracy": 0.2128172129392624,
      "num_tokens": 53111950.0,
      "step": 25600
    },
    {
      "entropy": 5.447553920745849,
      "epoch": 2.3226596516690856,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044597676444022683,
      "loss": 4.7931,
      "mean_token_accuracy": 0.22066829800605775,
      "num_tokens": 53121908.0,
      "step": 25605
    },
    {
      "entropy": 5.579293870925904,
      "epoch": 2.32311320754717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044595572958188194,
      "loss": 4.955,
      "mean_token_accuracy": 0.21176907867193223,
      "num_tokens": 53133065.0,
      "step": 25610
    },
    {
      "entropy": 5.490471887588501,
      "epoch": 2.323566763425254,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004459346911881291,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2105861097574234,
      "num_tokens": 53144135.0,
      "step": 25615
    },
    {
      "entropy": 5.559974908828735,
      "epoch": 2.3240203193033384,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004459136492594032,
      "loss": 4.857,
      "mean_token_accuracy": 0.2246379464864731,
      "num_tokens": 53153916.0,
      "step": 25620
    },
    {
      "entropy": 5.522878265380859,
      "epoch": 2.3244738751814222,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004458926037961394,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22637065798044204,
      "num_tokens": 53163937.0,
      "step": 25625
    },
    {
      "entropy": 5.617217588424682,
      "epoch": 2.3249274310595065,
      "grad_norm": 1.125,
      "learning_rate": 0.0004458715547987732,
      "loss": 5.0086,
      "mean_token_accuracy": 0.2043337643146515,
      "num_tokens": 53174327.0,
      "step": 25630
    },
    {
      "entropy": 5.584192562103271,
      "epoch": 2.3253809869375908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004458505022677396,
      "loss": 4.9324,
      "mean_token_accuracy": 0.21107761561870575,
      "num_tokens": 53185946.0,
      "step": 25635
    },
    {
      "entropy": 5.554040098190308,
      "epoch": 2.325834542815675,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004458294462034741,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21663629561662673,
      "num_tokens": 53195944.0,
      "step": 25640
    },
    {
      "entropy": 5.509247446060181,
      "epoch": 2.326288098693759,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004458083866064121,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21436946243047714,
      "num_tokens": 53206054.0,
      "step": 25645
    },
    {
      "entropy": 5.595161628723145,
      "epoch": 2.326741654571843,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004457873234769891,
      "loss": 4.9424,
      "mean_token_accuracy": 0.21010857969522476,
      "num_tokens": 53216923.0,
      "step": 25650
    },
    {
      "entropy": 5.568327903747559,
      "epoch": 2.3271952104499274,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044576625681564086,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21277860254049302,
      "num_tokens": 53226788.0,
      "step": 25655
    },
    {
      "entropy": 5.60441951751709,
      "epoch": 2.3276487663280117,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044574518662280276,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2090952455997467,
      "num_tokens": 53237573.0,
      "step": 25660
    },
    {
      "entropy": 5.585796451568603,
      "epoch": 2.328102322206096,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004457241128989108,
      "loss": 4.9884,
      "mean_token_accuracy": 0.20564925223588942,
      "num_tokens": 53247724.0,
      "step": 25665
    },
    {
      "entropy": 5.584307098388672,
      "epoch": 2.32855587808418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004457030356444006,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21314208954572678,
      "num_tokens": 53258774.0,
      "step": 25670
    },
    {
      "entropy": 5.581422567367554,
      "epoch": 2.329009433962264,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004456819548597082,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2158464729785919,
      "num_tokens": 53268550.0,
      "step": 25675
    },
    {
      "entropy": 5.497116804122925,
      "epoch": 2.3294629898403483,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044566087054526946,
      "loss": 4.8754,
      "mean_token_accuracy": 0.21785036027431487,
      "num_tokens": 53277975.0,
      "step": 25680
    },
    {
      "entropy": 5.621309804916382,
      "epoch": 2.3299165457184325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004456397827015204,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21554505378007888,
      "num_tokens": 53287939.0,
      "step": 25685
    },
    {
      "entropy": 5.613113832473755,
      "epoch": 2.330370101596517,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004456186913288971,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2131609320640564,
      "num_tokens": 53298355.0,
      "step": 25690
    },
    {
      "entropy": 5.664609336853028,
      "epoch": 2.3308236574746006,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004455975964278359,
      "loss": 5.036,
      "mean_token_accuracy": 0.21076706498861314,
      "num_tokens": 53308513.0,
      "step": 25695
    },
    {
      "entropy": 5.647288131713867,
      "epoch": 2.331277213352685,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044557649799877285,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21256288439035415,
      "num_tokens": 53318249.0,
      "step": 25700
    },
    {
      "entropy": 5.5144682884216305,
      "epoch": 2.331730769230769,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004455553960421443,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2187486171722412,
      "num_tokens": 53328460.0,
      "step": 25705
    },
    {
      "entropy": 5.5543145656585695,
      "epoch": 2.3321843251088534,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044553429055838685,
      "loss": 4.933,
      "mean_token_accuracy": 0.22022654116153717,
      "num_tokens": 53338984.0,
      "step": 25710
    },
    {
      "entropy": 5.611541986465454,
      "epoch": 2.3326378809869377,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044551318154793676,
      "loss": 4.9956,
      "mean_token_accuracy": 0.20399201065301895,
      "num_tokens": 53348092.0,
      "step": 25715
    },
    {
      "entropy": 5.538498067855835,
      "epoch": 2.333091436865022,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044549206901123056,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2126649484038353,
      "num_tokens": 53357343.0,
      "step": 25720
    },
    {
      "entropy": 5.53882098197937,
      "epoch": 2.333544992743106,
      "grad_norm": 1.078125,
      "learning_rate": 0.000445470952948705,
      "loss": 4.8499,
      "mean_token_accuracy": 0.21637628674507142,
      "num_tokens": 53367343.0,
      "step": 25725
    },
    {
      "entropy": 5.643903064727783,
      "epoch": 2.33399854862119,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044544983336079665,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21098848730325698,
      "num_tokens": 53379093.0,
      "step": 25730
    },
    {
      "entropy": 5.555362224578857,
      "epoch": 2.3344521044992743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004454287102479423,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21512557417154313,
      "num_tokens": 53388883.0,
      "step": 25735
    },
    {
      "entropy": 5.5754283428192135,
      "epoch": 2.3349056603773586,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004454075836105788,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21789999902248383,
      "num_tokens": 53399393.0,
      "step": 25740
    },
    {
      "entropy": 5.552889013290406,
      "epoch": 2.335359216255443,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044538645344914307,
      "loss": 4.8749,
      "mean_token_accuracy": 0.21663407981395721,
      "num_tokens": 53408545.0,
      "step": 25745
    },
    {
      "entropy": 5.5533253192901615,
      "epoch": 2.3358127721335267,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004453653197640721,
      "loss": 4.976,
      "mean_token_accuracy": 0.21037684231996537,
      "num_tokens": 53418325.0,
      "step": 25750
    },
    {
      "entropy": 5.573678302764892,
      "epoch": 2.336266328011611,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044534418255580286,
      "loss": 4.9431,
      "mean_token_accuracy": 0.20263057351112365,
      "num_tokens": 53427984.0,
      "step": 25755
    },
    {
      "entropy": 5.64089035987854,
      "epoch": 2.336719883889695,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044532304182477245,
      "loss": 5.0384,
      "mean_token_accuracy": 0.20502175986766816,
      "num_tokens": 53438765.0,
      "step": 25760
    },
    {
      "entropy": 5.646881151199341,
      "epoch": 2.3371734397677795,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004453018975714182,
      "loss": 4.9531,
      "mean_token_accuracy": 0.21376003324985504,
      "num_tokens": 53448792.0,
      "step": 25765
    },
    {
      "entropy": 5.648402404785156,
      "epoch": 2.3376269956458637,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004452807497961773,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20995502769947053,
      "num_tokens": 53459334.0,
      "step": 25770
    },
    {
      "entropy": 5.590705156326294,
      "epoch": 2.3380805515239476,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044525959849948704,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21176755279302598,
      "num_tokens": 53469841.0,
      "step": 25775
    },
    {
      "entropy": 5.569129753112793,
      "epoch": 2.338534107402032,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044523844368178503,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21363317519426345,
      "num_tokens": 53480660.0,
      "step": 25780
    },
    {
      "entropy": 5.604327297210693,
      "epoch": 2.338987663280116,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004452172853435085,
      "loss": 4.8892,
      "mean_token_accuracy": 0.21227259188890457,
      "num_tokens": 53490276.0,
      "step": 25785
    },
    {
      "entropy": 5.484948587417603,
      "epoch": 2.3394412191582004,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004451961234850953,
      "loss": 4.854,
      "mean_token_accuracy": 0.21953050494194032,
      "num_tokens": 53501100.0,
      "step": 25790
    },
    {
      "entropy": 5.518176174163818,
      "epoch": 2.3398947750362846,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004451749581069827,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21475230157375336,
      "num_tokens": 53511680.0,
      "step": 25795
    },
    {
      "entropy": 5.597425174713135,
      "epoch": 2.3403483309143684,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004451537892096087,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21020373851060867,
      "num_tokens": 53521298.0,
      "step": 25800
    },
    {
      "entropy": 5.572571229934693,
      "epoch": 2.3408018867924527,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044513261679341094,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22635883688926697,
      "num_tokens": 53530413.0,
      "step": 25805
    },
    {
      "entropy": 5.541699361801148,
      "epoch": 2.341255442670537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004451114408588273,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21247395426034926,
      "num_tokens": 53540959.0,
      "step": 25810
    },
    {
      "entropy": 5.57380428314209,
      "epoch": 2.3417089985486212,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044509026140629584,
      "loss": 4.9376,
      "mean_token_accuracy": 0.21315263956785202,
      "num_tokens": 53551265.0,
      "step": 25815
    },
    {
      "entropy": 5.615974426269531,
      "epoch": 2.3421625544267055,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044506907843625427,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20951335728168488,
      "num_tokens": 53562666.0,
      "step": 25820
    },
    {
      "entropy": 5.568785715103149,
      "epoch": 2.34261611030479,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044504789194914086,
      "loss": 5.0182,
      "mean_token_accuracy": 0.2091977760195732,
      "num_tokens": 53573688.0,
      "step": 25825
    },
    {
      "entropy": 5.594330787658691,
      "epoch": 2.3430696661828736,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044502670194539374,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21697243452072143,
      "num_tokens": 53584055.0,
      "step": 25830
    },
    {
      "entropy": 5.515743541717529,
      "epoch": 2.343523222060958,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044500550842545104,
      "loss": 4.8124,
      "mean_token_accuracy": 0.22413455843925476,
      "num_tokens": 53593765.0,
      "step": 25835
    },
    {
      "entropy": 5.498024892807007,
      "epoch": 2.343976777939042,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004449843113897512,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22074564099311828,
      "num_tokens": 53603480.0,
      "step": 25840
    },
    {
      "entropy": 5.560669851303101,
      "epoch": 2.3444303338171264,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004449631108387325,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21525682210922242,
      "num_tokens": 53613797.0,
      "step": 25845
    },
    {
      "entropy": 5.551551389694214,
      "epoch": 2.34488388969521,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004449419067728333,
      "loss": 4.875,
      "mean_token_accuracy": 0.217022804915905,
      "num_tokens": 53623646.0,
      "step": 25850
    },
    {
      "entropy": 5.54360671043396,
      "epoch": 2.3453374455732945,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044492069919249206,
      "loss": 4.917,
      "mean_token_accuracy": 0.2137106642127037,
      "num_tokens": 53634118.0,
      "step": 25855
    },
    {
      "entropy": 5.596779251098633,
      "epoch": 2.3457910014513788,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004448994880981476,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21488005816936492,
      "num_tokens": 53644370.0,
      "step": 25860
    },
    {
      "entropy": 5.608159160614013,
      "epoch": 2.346244557329463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044487827349023833,
      "loss": 4.9784,
      "mean_token_accuracy": 0.20647551417350768,
      "num_tokens": 53655560.0,
      "step": 25865
    },
    {
      "entropy": 5.576163816452026,
      "epoch": 2.3466981132075473,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044485705536920315,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21219031810760497,
      "num_tokens": 53666401.0,
      "step": 25870
    },
    {
      "entropy": 5.66574068069458,
      "epoch": 2.3471516690856316,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044483583373548075,
      "loss": 4.98,
      "mean_token_accuracy": 0.22362128794193267,
      "num_tokens": 53676133.0,
      "step": 25875
    },
    {
      "entropy": 5.569415044784546,
      "epoch": 2.3476052249637154,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004448146085895101,
      "loss": 4.9314,
      "mean_token_accuracy": 0.20970430672168733,
      "num_tokens": 53687683.0,
      "step": 25880
    },
    {
      "entropy": 5.5365321159362795,
      "epoch": 2.3480587808417996,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004447933799317299,
      "loss": 5.016,
      "mean_token_accuracy": 0.20889658778905867,
      "num_tokens": 53698343.0,
      "step": 25885
    },
    {
      "entropy": 5.577454376220703,
      "epoch": 2.348512336719884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004447721477625795,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21522369235754013,
      "num_tokens": 53709014.0,
      "step": 25890
    },
    {
      "entropy": 5.5477344512939455,
      "epoch": 2.348965892597968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044475091208249774,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21616510450839996,
      "num_tokens": 53718838.0,
      "step": 25895
    },
    {
      "entropy": 5.506769418716431,
      "epoch": 2.3494194484760524,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004447296728919239,
      "loss": 4.893,
      "mean_token_accuracy": 0.21399504840373992,
      "num_tokens": 53728526.0,
      "step": 25900
    },
    {
      "entropy": 5.607238864898681,
      "epoch": 2.3498730043541363,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004447084301912972,
      "loss": 5.0108,
      "mean_token_accuracy": 0.21239181607961655,
      "num_tokens": 53739261.0,
      "step": 25905
    },
    {
      "entropy": 5.605481147766113,
      "epoch": 2.3503265602322205,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004446871839810568,
      "loss": 4.9869,
      "mean_token_accuracy": 0.212431001663208,
      "num_tokens": 53749599.0,
      "step": 25910
    },
    {
      "entropy": 5.585039186477661,
      "epoch": 2.350780116110305,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044466593426164224,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20691626369953156,
      "num_tokens": 53759898.0,
      "step": 25915
    },
    {
      "entropy": 5.534364414215088,
      "epoch": 2.351233671988389,
      "grad_norm": 1.078125,
      "learning_rate": 0.000444644681033493,
      "loss": 4.9034,
      "mean_token_accuracy": 0.20887214094400405,
      "num_tokens": 53770561.0,
      "step": 25920
    },
    {
      "entropy": 5.546531629562378,
      "epoch": 2.3516872278664733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044462342429704845,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21344873458147048,
      "num_tokens": 53780408.0,
      "step": 25925
    },
    {
      "entropy": 5.485399580001831,
      "epoch": 2.352140783744557,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044460216405274827,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22735902070999145,
      "num_tokens": 53789125.0,
      "step": 25930
    },
    {
      "entropy": 5.536479234695435,
      "epoch": 2.3525943396226414,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044458090030103216,
      "loss": 4.8838,
      "mean_token_accuracy": 0.2195194274187088,
      "num_tokens": 53799164.0,
      "step": 25935
    },
    {
      "entropy": 5.549217414855957,
      "epoch": 2.3530478955007257,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044455963304233976,
      "loss": 4.871,
      "mean_token_accuracy": 0.2197946310043335,
      "num_tokens": 53810710.0,
      "step": 25940
    },
    {
      "entropy": 5.509092664718628,
      "epoch": 2.35350145137881,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000444538362277111,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2079112008213997,
      "num_tokens": 53820692.0,
      "step": 25945
    },
    {
      "entropy": 5.617864179611206,
      "epoch": 2.353955007256894,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004445170880057857,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2089620739221573,
      "num_tokens": 53831082.0,
      "step": 25950
    },
    {
      "entropy": 5.635778188705444,
      "epoch": 2.354408563134978,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044449581022880375,
      "loss": 5.0121,
      "mean_token_accuracy": 0.19856876879930496,
      "num_tokens": 53842602.0,
      "step": 25955
    },
    {
      "entropy": 5.548915004730224,
      "epoch": 2.3548621190130623,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004444745289466053,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21658931374549867,
      "num_tokens": 53853784.0,
      "step": 25960
    },
    {
      "entropy": 5.522101163864136,
      "epoch": 2.3553156748911466,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044445324415963043,
      "loss": 4.8197,
      "mean_token_accuracy": 0.22228266745805741,
      "num_tokens": 53863526.0,
      "step": 25965
    },
    {
      "entropy": 5.514787483215332,
      "epoch": 2.355769230769231,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004444319558683192,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21255785822868348,
      "num_tokens": 53874594.0,
      "step": 25970
    },
    {
      "entropy": 5.575469303131103,
      "epoch": 2.356222786647315,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044441066407311205,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21542281061410903,
      "num_tokens": 53884771.0,
      "step": 25975
    },
    {
      "entropy": 5.605532455444336,
      "epoch": 2.3566763425253994,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004443893687744492,
      "loss": 4.9253,
      "mean_token_accuracy": 0.21457685381174088,
      "num_tokens": 53895181.0,
      "step": 25980
    },
    {
      "entropy": 5.514532136917114,
      "epoch": 2.357129898403483,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044436806997277096,
      "loss": 4.8244,
      "mean_token_accuracy": 0.22391154766082763,
      "num_tokens": 53904153.0,
      "step": 25985
    },
    {
      "entropy": 5.499745178222656,
      "epoch": 2.3575834542815675,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044434676766851787,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22061283886432648,
      "num_tokens": 53914064.0,
      "step": 25990
    },
    {
      "entropy": 5.56377010345459,
      "epoch": 2.3580370101596517,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044432546186213054,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2109364777803421,
      "num_tokens": 53924240.0,
      "step": 25995
    },
    {
      "entropy": 5.593591117858887,
      "epoch": 2.358490566037736,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044430415255404944,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2133777320384979,
      "num_tokens": 53935395.0,
      "step": 26000
    },
    {
      "entropy": 5.577879428863525,
      "epoch": 2.35894412191582,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044428283974471533,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21419965773820876,
      "num_tokens": 53945768.0,
      "step": 26005
    },
    {
      "entropy": 5.498639488220215,
      "epoch": 2.359397677793904,
      "grad_norm": 1.15625,
      "learning_rate": 0.000444261523434569,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22037704586982726,
      "num_tokens": 53956467.0,
      "step": 26010
    },
    {
      "entropy": 5.513798713684082,
      "epoch": 2.3598512336719883,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044424020362405126,
      "loss": 4.8569,
      "mean_token_accuracy": 0.21677091717720032,
      "num_tokens": 53967587.0,
      "step": 26015
    },
    {
      "entropy": 5.534829187393188,
      "epoch": 2.3603047895500726,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004442188803136029,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2128376007080078,
      "num_tokens": 53977691.0,
      "step": 26020
    },
    {
      "entropy": 5.54969973564148,
      "epoch": 2.360758345428157,
      "grad_norm": 1.21875,
      "learning_rate": 0.000444197553503665,
      "loss": 4.9756,
      "mean_token_accuracy": 0.20649277418851852,
      "num_tokens": 53987608.0,
      "step": 26025
    },
    {
      "entropy": 5.599653005599976,
      "epoch": 2.361211901306241,
      "grad_norm": 1.125,
      "learning_rate": 0.00044417622319467857,
      "loss": 5.06,
      "mean_token_accuracy": 0.20780419409275055,
      "num_tokens": 53998010.0,
      "step": 26030
    },
    {
      "entropy": 5.565209770202637,
      "epoch": 2.361665457184325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044415488938708475,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2175135225057602,
      "num_tokens": 54009852.0,
      "step": 26035
    },
    {
      "entropy": 5.563492345809936,
      "epoch": 2.3621190130624092,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004441335520813247,
      "loss": 4.8692,
      "mean_token_accuracy": 0.21580897271633148,
      "num_tokens": 54019635.0,
      "step": 26040
    },
    {
      "entropy": 5.539868545532227,
      "epoch": 2.3625725689404935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044411221127783966,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2137252390384674,
      "num_tokens": 54030046.0,
      "step": 26045
    },
    {
      "entropy": 5.5821287631988525,
      "epoch": 2.3630261248185778,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004440908669770711,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21307014524936677,
      "num_tokens": 54041197.0,
      "step": 26050
    },
    {
      "entropy": 5.579737138748169,
      "epoch": 2.3634796806966616,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004440695191794602,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2196875348687172,
      "num_tokens": 54051592.0,
      "step": 26055
    },
    {
      "entropy": 5.5798334121704105,
      "epoch": 2.363933236574746,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004440481678854487,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2075364723801613,
      "num_tokens": 54062893.0,
      "step": 26060
    },
    {
      "entropy": 5.627664184570312,
      "epoch": 2.36438679245283,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044402681309547786,
      "loss": 4.9885,
      "mean_token_accuracy": 0.20873741209506988,
      "num_tokens": 54073651.0,
      "step": 26065
    },
    {
      "entropy": 5.5339478015899655,
      "epoch": 2.3648403483309144,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044400545480998956,
      "loss": 4.9127,
      "mean_token_accuracy": 0.22097087204456328,
      "num_tokens": 54084889.0,
      "step": 26070
    },
    {
      "entropy": 5.5441789627075195,
      "epoch": 2.3652939042089987,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044398409302942533,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2119978666305542,
      "num_tokens": 54095724.0,
      "step": 26075
    },
    {
      "entropy": 5.56972017288208,
      "epoch": 2.365747460087083,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000443962727754227,
      "loss": 4.8879,
      "mean_token_accuracy": 0.21552262902259828,
      "num_tokens": 54106581.0,
      "step": 26080
    },
    {
      "entropy": 5.6068250179290775,
      "epoch": 2.3662010159651667,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004439413589848364,
      "loss": 4.9464,
      "mean_token_accuracy": 0.20996494740247726,
      "num_tokens": 54116074.0,
      "step": 26085
    },
    {
      "entropy": 5.548000335693359,
      "epoch": 2.366654571843251,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004439199867216955,
      "loss": 4.9625,
      "mean_token_accuracy": 0.20807096362113953,
      "num_tokens": 54126902.0,
      "step": 26090
    },
    {
      "entropy": 5.5538010597229,
      "epoch": 2.3671081277213353,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004438986109652462,
      "loss": 4.9005,
      "mean_token_accuracy": 0.21629703044891357,
      "num_tokens": 54137554.0,
      "step": 26095
    },
    {
      "entropy": 5.5640634059906,
      "epoch": 2.3675616835994195,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004438772317159306,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21244510710239412,
      "num_tokens": 54148035.0,
      "step": 26100
    },
    {
      "entropy": 5.536005544662475,
      "epoch": 2.368015239477504,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044385584897419075,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22129922360181808,
      "num_tokens": 54157417.0,
      "step": 26105
    },
    {
      "entropy": 5.594117927551269,
      "epoch": 2.3684687953555876,
      "grad_norm": 1.125,
      "learning_rate": 0.000443834462740469,
      "loss": 4.9599,
      "mean_token_accuracy": 0.20927917808294297,
      "num_tokens": 54168353.0,
      "step": 26110
    },
    {
      "entropy": 5.554742050170899,
      "epoch": 2.368922351233672,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004438130730152074,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2028622180223465,
      "num_tokens": 54180182.0,
      "step": 26115
    },
    {
      "entropy": 5.606842660903931,
      "epoch": 2.369375907111756,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044379167979884847,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2098795086145401,
      "num_tokens": 54190973.0,
      "step": 26120
    },
    {
      "entropy": 5.609960985183716,
      "epoch": 2.3698294629898404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004437702830918347,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20531612187623977,
      "num_tokens": 54202077.0,
      "step": 26125
    },
    {
      "entropy": 5.572955703735351,
      "epoch": 2.3702830188679247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004437488828946084,
      "loss": 4.9428,
      "mean_token_accuracy": 0.20869537591934204,
      "num_tokens": 54211778.0,
      "step": 26130
    },
    {
      "entropy": 5.566894578933716,
      "epoch": 2.3707365747460085,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004437274792076122,
      "loss": 4.8134,
      "mean_token_accuracy": 0.23084428906440735,
      "num_tokens": 54220941.0,
      "step": 26135
    },
    {
      "entropy": 5.588488435745239,
      "epoch": 2.371190130624093,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044370607203128874,
      "loss": 5.0337,
      "mean_token_accuracy": 0.20585249811410905,
      "num_tokens": 54231602.0,
      "step": 26140
    },
    {
      "entropy": 5.5495302200317385,
      "epoch": 2.371643686502177,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004436846613660807,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2138413593173027,
      "num_tokens": 54241797.0,
      "step": 26145
    },
    {
      "entropy": 5.629802227020264,
      "epoch": 2.3720972423802613,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004436632472124308,
      "loss": 5.0088,
      "mean_token_accuracy": 0.21093651354312898,
      "num_tokens": 54253085.0,
      "step": 26150
    },
    {
      "entropy": 5.655661010742188,
      "epoch": 2.3725507982583456,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004436418295707821,
      "loss": 4.9764,
      "mean_token_accuracy": 0.2152244508266449,
      "num_tokens": 54262164.0,
      "step": 26155
    },
    {
      "entropy": 5.574744415283203,
      "epoch": 2.3730043541364294,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004436204084415774,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21065271645784378,
      "num_tokens": 54271922.0,
      "step": 26160
    },
    {
      "entropy": 5.567212820053101,
      "epoch": 2.3734579100145137,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004435989838252596,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22161255031824112,
      "num_tokens": 54281738.0,
      "step": 26165
    },
    {
      "entropy": 5.570029592514038,
      "epoch": 2.373911465892598,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004435775557222719,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21278369724750518,
      "num_tokens": 54292252.0,
      "step": 26170
    },
    {
      "entropy": 5.570502758026123,
      "epoch": 2.374365021770682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044355612413305736,
      "loss": 4.9866,
      "mean_token_accuracy": 0.2120476931333542,
      "num_tokens": 54302518.0,
      "step": 26175
    },
    {
      "entropy": 5.595395278930664,
      "epoch": 2.3748185776487665,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004435346890580593,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2197017937898636,
      "num_tokens": 54312917.0,
      "step": 26180
    },
    {
      "entropy": 5.613895893096924,
      "epoch": 2.3752721335268507,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044351325049772085,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2203422635793686,
      "num_tokens": 54322552.0,
      "step": 26185
    },
    {
      "entropy": 5.571087455749511,
      "epoch": 2.3757256894049346,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044349180845248536,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21754180192947387,
      "num_tokens": 54332917.0,
      "step": 26190
    },
    {
      "entropy": 5.502777910232544,
      "epoch": 2.376179245283019,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004434703629227964,
      "loss": 4.8812,
      "mean_token_accuracy": 0.21895778328180313,
      "num_tokens": 54342791.0,
      "step": 26195
    },
    {
      "entropy": 5.514581489562988,
      "epoch": 2.376632801161103,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004434489139090975,
      "loss": 4.8011,
      "mean_token_accuracy": 0.22824004888534546,
      "num_tokens": 54352821.0,
      "step": 26200
    },
    {
      "entropy": 5.634966659545898,
      "epoch": 2.3770863570391874,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044342746141183206,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20839511901140212,
      "num_tokens": 54363095.0,
      "step": 26205
    },
    {
      "entropy": 5.56240291595459,
      "epoch": 2.377539912917271,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004434060054314438,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2107112556695938,
      "num_tokens": 54373841.0,
      "step": 26210
    },
    {
      "entropy": 5.509205007553101,
      "epoch": 2.3779934687953554,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044338454596837637,
      "loss": 4.8809,
      "mean_token_accuracy": 0.216173654794693,
      "num_tokens": 54385012.0,
      "step": 26215
    },
    {
      "entropy": 5.5585497379302975,
      "epoch": 2.3784470246734397,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004433630830230737,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21192312687635423,
      "num_tokens": 54395298.0,
      "step": 26220
    },
    {
      "entropy": 5.568934106826783,
      "epoch": 2.378900580551524,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004433416165959795,
      "loss": 4.9314,
      "mean_token_accuracy": 0.21015683114528655,
      "num_tokens": 54405752.0,
      "step": 26225
    },
    {
      "entropy": 5.590658378601074,
      "epoch": 2.3793541364296082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004433201466875378,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21067356169223786,
      "num_tokens": 54416607.0,
      "step": 26230
    },
    {
      "entropy": 5.585775995254517,
      "epoch": 2.3798076923076925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004432986732981926,
      "loss": 4.914,
      "mean_token_accuracy": 0.2157020777463913,
      "num_tokens": 54426873.0,
      "step": 26235
    },
    {
      "entropy": 5.492963695526123,
      "epoch": 2.3802612481857763,
      "grad_norm": 1.0625,
      "learning_rate": 0.000443277196428388,
      "loss": 4.8394,
      "mean_token_accuracy": 0.22379397302865983,
      "num_tokens": 54437611.0,
      "step": 26240
    },
    {
      "entropy": 5.611445760726928,
      "epoch": 2.3807148040638606,
      "grad_norm": 1.046875,
      "learning_rate": 0.000443255716078568,
      "loss": 4.9384,
      "mean_token_accuracy": 0.21650339365005494,
      "num_tokens": 54448992.0,
      "step": 26245
    },
    {
      "entropy": 5.575338125228882,
      "epoch": 2.381168359941945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004432342322491769,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22124216109514236,
      "num_tokens": 54459172.0,
      "step": 26250
    },
    {
      "entropy": 5.499160575866699,
      "epoch": 2.381621915820029,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044321274494065903,
      "loss": 4.8903,
      "mean_token_accuracy": 0.21635547727346421,
      "num_tokens": 54469660.0,
      "step": 26255
    },
    {
      "entropy": 5.596979665756225,
      "epoch": 2.3820754716981134,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044319125415345886,
      "loss": 5.0191,
      "mean_token_accuracy": 0.2071585014462471,
      "num_tokens": 54479567.0,
      "step": 26260
    },
    {
      "entropy": 5.565261697769165,
      "epoch": 2.3825290275761972,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004431697598880205,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20945729017257692,
      "num_tokens": 54489690.0,
      "step": 26265
    },
    {
      "entropy": 5.585064554214478,
      "epoch": 2.3829825834542815,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044314826214478877,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2184809312224388,
      "num_tokens": 54500876.0,
      "step": 26270
    },
    {
      "entropy": 5.606135892868042,
      "epoch": 2.3834361393323658,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044312676092420804,
      "loss": 4.9813,
      "mean_token_accuracy": 0.2080484375357628,
      "num_tokens": 54510988.0,
      "step": 26275
    },
    {
      "entropy": 5.565968656539917,
      "epoch": 2.38388969521045,
      "grad_norm": 1.0,
      "learning_rate": 0.0004431052562267231,
      "loss": 4.8853,
      "mean_token_accuracy": 0.223616386950016,
      "num_tokens": 54522149.0,
      "step": 26280
    },
    {
      "entropy": 5.637773895263672,
      "epoch": 2.3843432510885343,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044308374805277867,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20029253214597703,
      "num_tokens": 54533482.0,
      "step": 26285
    },
    {
      "entropy": 5.562591981887818,
      "epoch": 2.384796806966618,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004430622364028194,
      "loss": 4.9738,
      "mean_token_accuracy": 0.20799656063318253,
      "num_tokens": 54543058.0,
      "step": 26290
    },
    {
      "entropy": 5.578580474853515,
      "epoch": 2.3852503628447024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044304072127729035,
      "loss": 4.9324,
      "mean_token_accuracy": 0.2148285612463951,
      "num_tokens": 54552655.0,
      "step": 26295
    },
    {
      "entropy": 5.569213390350342,
      "epoch": 2.3857039187227866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044301920267663633,
      "loss": 4.8874,
      "mean_token_accuracy": 0.20749802887439728,
      "num_tokens": 54562696.0,
      "step": 26300
    },
    {
      "entropy": 5.5724883556365965,
      "epoch": 2.386157474600871,
      "grad_norm": 1.125,
      "learning_rate": 0.0004429976806013024,
      "loss": 4.9138,
      "mean_token_accuracy": 0.21777765154838563,
      "num_tokens": 54572803.0,
      "step": 26305
    },
    {
      "entropy": 5.521192264556885,
      "epoch": 2.386611030478955,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004429761550517336,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21111124455928804,
      "num_tokens": 54582450.0,
      "step": 26310
    },
    {
      "entropy": 5.478409290313721,
      "epoch": 2.387064586357039,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044295462602837504,
      "loss": 4.8846,
      "mean_token_accuracy": 0.21607948392629622,
      "num_tokens": 54591868.0,
      "step": 26315
    },
    {
      "entropy": 5.5848370552062985,
      "epoch": 2.3875181422351233,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004429330935316721,
      "loss": 4.8984,
      "mean_token_accuracy": 0.21125050783157348,
      "num_tokens": 54601858.0,
      "step": 26320
    },
    {
      "entropy": 5.571154832839966,
      "epoch": 2.3879716981132075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044291155756206996,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21169865280389785,
      "num_tokens": 54612243.0,
      "step": 26325
    },
    {
      "entropy": 5.570634555816651,
      "epoch": 2.388425253991292,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044289001812001404,
      "loss": 4.8407,
      "mean_token_accuracy": 0.22116668671369552,
      "num_tokens": 54622365.0,
      "step": 26330
    },
    {
      "entropy": 5.552473688125611,
      "epoch": 2.388878809869376,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004428684752059497,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22066307663917542,
      "num_tokens": 54632356.0,
      "step": 26335
    },
    {
      "entropy": 5.6089372634887695,
      "epoch": 2.3893323657474603,
      "grad_norm": 1.125,
      "learning_rate": 0.0004428469288203226,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20489265620708466,
      "num_tokens": 54642976.0,
      "step": 26340
    },
    {
      "entropy": 5.657524251937867,
      "epoch": 2.389785921625544,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004428253789635783,
      "loss": 4.9738,
      "mean_token_accuracy": 0.21573778241872787,
      "num_tokens": 54652597.0,
      "step": 26345
    },
    {
      "entropy": 5.560487127304077,
      "epoch": 2.3902394775036284,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004428038256361622,
      "loss": 5.0598,
      "mean_token_accuracy": 0.207101708650589,
      "num_tokens": 54662930.0,
      "step": 26350
    },
    {
      "entropy": 5.689995527267456,
      "epoch": 2.3906930333817127,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004427822688385203,
      "loss": 5.1682,
      "mean_token_accuracy": 0.20526119619607924,
      "num_tokens": 54674562.0,
      "step": 26355
    },
    {
      "entropy": 5.630713653564453,
      "epoch": 2.391146589259797,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004427607085710984,
      "loss": 4.951,
      "mean_token_accuracy": 0.2108895167708397,
      "num_tokens": 54684432.0,
      "step": 26360
    },
    {
      "entropy": 5.632601737976074,
      "epoch": 2.3916001451378808,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044273914483434226,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20941840261220931,
      "num_tokens": 54694060.0,
      "step": 26365
    },
    {
      "entropy": 5.551342439651489,
      "epoch": 2.392053701015965,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044271757762869777,
      "loss": 4.9579,
      "mean_token_accuracy": 0.2113799899816513,
      "num_tokens": 54704411.0,
      "step": 26370
    },
    {
      "entropy": 5.565278053283691,
      "epoch": 2.3925072568940493,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004426960069546112,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21328798830509185,
      "num_tokens": 54715185.0,
      "step": 26375
    },
    {
      "entropy": 5.562286853790283,
      "epoch": 2.3929608127721336,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004426744328125284,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22069818377494813,
      "num_tokens": 54725529.0,
      "step": 26380
    },
    {
      "entropy": 5.648525619506836,
      "epoch": 2.393414368650218,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004426528552028955,
      "loss": 5.0024,
      "mean_token_accuracy": 0.2120383784174919,
      "num_tokens": 54736992.0,
      "step": 26385
    },
    {
      "entropy": 5.593166875839233,
      "epoch": 2.393867924528302,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044263127412615897,
      "loss": 4.9063,
      "mean_token_accuracy": 0.2161580428481102,
      "num_tokens": 54746951.0,
      "step": 26390
    },
    {
      "entropy": 5.52850022315979,
      "epoch": 2.394321480406386,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004426096895827649,
      "loss": 4.8532,
      "mean_token_accuracy": 0.22522672712802888,
      "num_tokens": 54757272.0,
      "step": 26395
    },
    {
      "entropy": 5.565531539916992,
      "epoch": 2.39477503628447,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044258810157315977,
      "loss": 5.0043,
      "mean_token_accuracy": 0.20992569923400878,
      "num_tokens": 54767835.0,
      "step": 26400
    },
    {
      "entropy": 5.588186407089234,
      "epoch": 2.3952285921625545,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044256651009779,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21384909749031067,
      "num_tokens": 54778432.0,
      "step": 26405
    },
    {
      "entropy": 5.637002038955688,
      "epoch": 2.3956821480406387,
      "grad_norm": 1.125,
      "learning_rate": 0.00044254491515710213,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20795969665050507,
      "num_tokens": 54788582.0,
      "step": 26410
    },
    {
      "entropy": 5.6249936580657955,
      "epoch": 2.3961357039187225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004425233167515426,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21718416959047318,
      "num_tokens": 54798848.0,
      "step": 26415
    },
    {
      "entropy": 5.541302633285523,
      "epoch": 2.396589259796807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004425017148815583,
      "loss": 4.8537,
      "mean_token_accuracy": 0.21915711760520934,
      "num_tokens": 54809867.0,
      "step": 26420
    },
    {
      "entropy": 5.5384095191955565,
      "epoch": 2.397042815674891,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044248010954759575,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2144401639699936,
      "num_tokens": 54819905.0,
      "step": 26425
    },
    {
      "entropy": 5.6020440578460695,
      "epoch": 2.3974963715529753,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044245850075010193,
      "loss": 4.9437,
      "mean_token_accuracy": 0.21297521591186525,
      "num_tokens": 54829723.0,
      "step": 26430
    },
    {
      "entropy": 5.551946640014648,
      "epoch": 2.3979499274310596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044243688848952364,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2183666169643402,
      "num_tokens": 54840008.0,
      "step": 26435
    },
    {
      "entropy": 5.564555835723877,
      "epoch": 2.398403483309144,
      "grad_norm": 1.125,
      "learning_rate": 0.0004424152727663078,
      "loss": 4.9441,
      "mean_token_accuracy": 0.20714806020259857,
      "num_tokens": 54851693.0,
      "step": 26440
    },
    {
      "entropy": 5.597462797164917,
      "epoch": 2.3988570391872277,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004423936535809015,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21012354642152786,
      "num_tokens": 54862086.0,
      "step": 26445
    },
    {
      "entropy": 5.687418413162232,
      "epoch": 2.399310595065312,
      "grad_norm": 1.125,
      "learning_rate": 0.0004423720309337517,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20929399728775025,
      "num_tokens": 54872883.0,
      "step": 26450
    },
    {
      "entropy": 5.622654771804809,
      "epoch": 2.3997641509433962,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044235040482530576,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2101919338107109,
      "num_tokens": 54882815.0,
      "step": 26455
    },
    {
      "entropy": 5.431966876983642,
      "epoch": 2.4002177068214805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004423287752560108,
      "loss": 4.7813,
      "mean_token_accuracy": 0.22615410685539244,
      "num_tokens": 54892480.0,
      "step": 26460
    },
    {
      "entropy": 5.460355615615844,
      "epoch": 2.4006712626995648,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044230714222631397,
      "loss": 4.8245,
      "mean_token_accuracy": 0.22623133063316345,
      "num_tokens": 54901607.0,
      "step": 26465
    },
    {
      "entropy": 5.678241014480591,
      "epoch": 2.4011248185776486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044228550573666294,
      "loss": 5.0553,
      "mean_token_accuracy": 0.21256614476442337,
      "num_tokens": 54913545.0,
      "step": 26470
    },
    {
      "entropy": 5.628347158432007,
      "epoch": 2.401578374455733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000442263865787505,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21294753700494767,
      "num_tokens": 54924392.0,
      "step": 26475
    },
    {
      "entropy": 5.581802845001221,
      "epoch": 2.402031930333817,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004422422223792877,
      "loss": 4.9217,
      "mean_token_accuracy": 0.21415944993495942,
      "num_tokens": 54934114.0,
      "step": 26480
    },
    {
      "entropy": 5.615803384780884,
      "epoch": 2.4024854862119014,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004422205755124586,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2141389861702919,
      "num_tokens": 54944268.0,
      "step": 26485
    },
    {
      "entropy": 5.608943605422974,
      "epoch": 2.4029390420899857,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044219892518746536,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21248486638069153,
      "num_tokens": 54954877.0,
      "step": 26490
    },
    {
      "entropy": 5.542152738571167,
      "epoch": 2.4033925979680695,
      "grad_norm": 1.125,
      "learning_rate": 0.00044217727140475585,
      "loss": 4.9016,
      "mean_token_accuracy": 0.21532075703144074,
      "num_tokens": 54964745.0,
      "step": 26495
    },
    {
      "entropy": 5.565544939041137,
      "epoch": 2.4038461538461537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044215561416477764,
      "loss": 4.8404,
      "mean_token_accuracy": 0.21421201080083846,
      "num_tokens": 54975256.0,
      "step": 26500
    },
    {
      "entropy": 5.649453830718994,
      "epoch": 2.404299709724238,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004421339534679788,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20731568038463594,
      "num_tokens": 54986226.0,
      "step": 26505
    },
    {
      "entropy": 5.620152568817138,
      "epoch": 2.4047532656023223,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004421122893148072,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20383986830711365,
      "num_tokens": 54997350.0,
      "step": 26510
    },
    {
      "entropy": 5.603748464584351,
      "epoch": 2.4052068214804065,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044209062170571086,
      "loss": 4.9188,
      "mean_token_accuracy": 0.21386370807886124,
      "num_tokens": 55007947.0,
      "step": 26515
    },
    {
      "entropy": 5.586966896057129,
      "epoch": 2.4056603773584904,
      "grad_norm": 1.125,
      "learning_rate": 0.0004420689506411379,
      "loss": 4.9634,
      "mean_token_accuracy": 0.21214287132024764,
      "num_tokens": 55018789.0,
      "step": 26520
    },
    {
      "entropy": 5.553619337081909,
      "epoch": 2.4061139332365746,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044204727612153643,
      "loss": 4.881,
      "mean_token_accuracy": 0.21062198877334595,
      "num_tokens": 55028934.0,
      "step": 26525
    },
    {
      "entropy": 5.553179025650024,
      "epoch": 2.406567489114659,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044202559814735475,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22172596901655198,
      "num_tokens": 55039395.0,
      "step": 26530
    },
    {
      "entropy": 5.515214729309082,
      "epoch": 2.407021044992743,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004420039167190411,
      "loss": 4.8549,
      "mean_token_accuracy": 0.21208732426166535,
      "num_tokens": 55049932.0,
      "step": 26535
    },
    {
      "entropy": 5.583847188949585,
      "epoch": 2.4074746008708274,
      "grad_norm": 1.296875,
      "learning_rate": 0.000441982231837044,
      "loss": 4.9732,
      "mean_token_accuracy": 0.20788389146327974,
      "num_tokens": 55061532.0,
      "step": 26540
    },
    {
      "entropy": 5.64016637802124,
      "epoch": 2.4079281567489117,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004419605435018117,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20780168771743773,
      "num_tokens": 55072368.0,
      "step": 26545
    },
    {
      "entropy": 5.589163970947266,
      "epoch": 2.4083817126269955,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004419388517137928,
      "loss": 4.9415,
      "mean_token_accuracy": 0.21165295392274858,
      "num_tokens": 55083128.0,
      "step": 26550
    },
    {
      "entropy": 5.5810487270355225,
      "epoch": 2.40883526850508,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044191715647343594,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2091229736804962,
      "num_tokens": 55094071.0,
      "step": 26555
    },
    {
      "entropy": 5.589597702026367,
      "epoch": 2.409288824383164,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044189545778118973,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21490462124347687,
      "num_tokens": 55104504.0,
      "step": 26560
    },
    {
      "entropy": 5.570869112014771,
      "epoch": 2.4097423802612483,
      "grad_norm": 1.109375,
      "learning_rate": 0.000441873755637503,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21151310801506043,
      "num_tokens": 55115722.0,
      "step": 26565
    },
    {
      "entropy": 5.536609172821045,
      "epoch": 2.410195936139332,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004418520500428244,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2122745156288147,
      "num_tokens": 55126628.0,
      "step": 26570
    },
    {
      "entropy": 5.540182781219483,
      "epoch": 2.4106494920174164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044183034099760287,
      "loss": 4.936,
      "mean_token_accuracy": 0.21147914379835128,
      "num_tokens": 55137352.0,
      "step": 26575
    },
    {
      "entropy": 5.6101806640625,
      "epoch": 2.4111030478955007,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004418086285022874,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22326893955469132,
      "num_tokens": 55147145.0,
      "step": 26580
    },
    {
      "entropy": 5.630547618865966,
      "epoch": 2.411556603773585,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044178691255732707,
      "loss": 4.9437,
      "mean_token_accuracy": 0.21231580674648284,
      "num_tokens": 55157386.0,
      "step": 26585
    },
    {
      "entropy": 5.61014232635498,
      "epoch": 2.412010159651669,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004417651931631708,
      "loss": 5.0169,
      "mean_token_accuracy": 0.20744775086641312,
      "num_tokens": 55168969.0,
      "step": 26590
    },
    {
      "entropy": 5.550246000289917,
      "epoch": 2.4124637155297535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004417434703202678,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2032379239797592,
      "num_tokens": 55180123.0,
      "step": 26595
    },
    {
      "entropy": 5.546901369094849,
      "epoch": 2.4129172714078373,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004417217440290673,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2159874141216278,
      "num_tokens": 55190668.0,
      "step": 26600
    },
    {
      "entropy": 5.527334308624267,
      "epoch": 2.4133708272859216,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004417000142900188,
      "loss": 4.8643,
      "mean_token_accuracy": 0.21732689291238785,
      "num_tokens": 55200852.0,
      "step": 26605
    },
    {
      "entropy": 5.70292558670044,
      "epoch": 2.413824383164006,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004416782811035715,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20385408997535706,
      "num_tokens": 55211174.0,
      "step": 26610
    },
    {
      "entropy": 5.5272753715515135,
      "epoch": 2.41427793904209,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044165654447017485,
      "loss": 4.8274,
      "mean_token_accuracy": 0.22203521877527238,
      "num_tokens": 55222042.0,
      "step": 26615
    },
    {
      "entropy": 5.579027938842773,
      "epoch": 2.4147314949201744,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044163480439027845,
      "loss": 4.967,
      "mean_token_accuracy": 0.20782914459705354,
      "num_tokens": 55231688.0,
      "step": 26620
    },
    {
      "entropy": 5.581716680526734,
      "epoch": 2.415185050798258,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004416130608643317,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21451929211616516,
      "num_tokens": 55241089.0,
      "step": 26625
    },
    {
      "entropy": 5.622543573379517,
      "epoch": 2.4156386066763424,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004415913138927844,
      "loss": 4.9023,
      "mean_token_accuracy": 0.21910626143217088,
      "num_tokens": 55251511.0,
      "step": 26630
    },
    {
      "entropy": 5.543213653564453,
      "epoch": 2.4160921625544267,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004415695634760864,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2101625144481659,
      "num_tokens": 55262855.0,
      "step": 26635
    },
    {
      "entropy": 5.511327409744263,
      "epoch": 2.416545718432511,
      "grad_norm": 1.125,
      "learning_rate": 0.0004415478096146873,
      "loss": 4.8254,
      "mean_token_accuracy": 0.2197219431400299,
      "num_tokens": 55273035.0,
      "step": 26640
    },
    {
      "entropy": 5.556900215148926,
      "epoch": 2.4169992743105952,
      "grad_norm": 1.078125,
      "learning_rate": 0.000441526052309037,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22204774171113967,
      "num_tokens": 55284033.0,
      "step": 26645
    },
    {
      "entropy": 5.592622947692871,
      "epoch": 2.417452830188679,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004415042915595856,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2190182015299797,
      "num_tokens": 55294290.0,
      "step": 26650
    },
    {
      "entropy": 5.625838899612427,
      "epoch": 2.4179063860667633,
      "grad_norm": 1.0625,
      "learning_rate": 0.000441482527366783,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21190671920776366,
      "num_tokens": 55305681.0,
      "step": 26655
    },
    {
      "entropy": 5.533696269989013,
      "epoch": 2.4183599419448476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004414607597310793,
      "loss": 4.907,
      "mean_token_accuracy": 0.2169225424528122,
      "num_tokens": 55316647.0,
      "step": 26660
    },
    {
      "entropy": 5.619781112670898,
      "epoch": 2.418813497822932,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004414389886529247,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20682642012834548,
      "num_tokens": 55328318.0,
      "step": 26665
    },
    {
      "entropy": 5.645016288757324,
      "epoch": 2.419267053701016,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004414172141327693,
      "loss": 4.9362,
      "mean_token_accuracy": 0.21090358942747117,
      "num_tokens": 55338483.0,
      "step": 26670
    },
    {
      "entropy": 5.60189962387085,
      "epoch": 2.4197206095791,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004413954361710636,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21040410697460174,
      "num_tokens": 55348648.0,
      "step": 26675
    },
    {
      "entropy": 5.540328168869019,
      "epoch": 2.4201741654571842,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044137365476825785,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21204569935798645,
      "num_tokens": 55358599.0,
      "step": 26680
    },
    {
      "entropy": 5.634864091873169,
      "epoch": 2.4206277213352685,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044135186992480253,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21380085200071336,
      "num_tokens": 55369847.0,
      "step": 26685
    },
    {
      "entropy": 5.568818044662476,
      "epoch": 2.4210812772133528,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044133008164114814,
      "loss": 4.941,
      "mean_token_accuracy": 0.2139912351965904,
      "num_tokens": 55381235.0,
      "step": 26690
    },
    {
      "entropy": 5.564904975891113,
      "epoch": 2.421534833091437,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044130828991774526,
      "loss": 4.9743,
      "mean_token_accuracy": 0.2082348197698593,
      "num_tokens": 55391527.0,
      "step": 26695
    },
    {
      "entropy": 5.620860910415649,
      "epoch": 2.4219883889695213,
      "grad_norm": 1.125,
      "learning_rate": 0.00044128649475504454,
      "loss": 5.0073,
      "mean_token_accuracy": 0.20783987045288085,
      "num_tokens": 55401528.0,
      "step": 26700
    },
    {
      "entropy": 5.638621282577515,
      "epoch": 2.422441944847605,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004412646961534968,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20500356405973436,
      "num_tokens": 55411411.0,
      "step": 26705
    },
    {
      "entropy": 5.496764421463013,
      "epoch": 2.4228955007256894,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004412428941135528,
      "loss": 4.7909,
      "mean_token_accuracy": 0.22079988569021225,
      "num_tokens": 55423164.0,
      "step": 26710
    },
    {
      "entropy": 5.562102031707764,
      "epoch": 2.4233490566037736,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004412210886356633,
      "loss": 4.8637,
      "mean_token_accuracy": 0.22214736938476562,
      "num_tokens": 55433397.0,
      "step": 26715
    },
    {
      "entropy": 5.551472425460815,
      "epoch": 2.423802612481858,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044119927972027936,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2118145391345024,
      "num_tokens": 55442898.0,
      "step": 26720
    },
    {
      "entropy": 5.5450035572052006,
      "epoch": 2.4242561683599417,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044117746736785195,
      "loss": 4.9276,
      "mean_token_accuracy": 0.21898261159658433,
      "num_tokens": 55452226.0,
      "step": 26725
    },
    {
      "entropy": 5.585405063629151,
      "epoch": 2.424709724238026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004411556515788322,
      "loss": 4.935,
      "mean_token_accuracy": 0.21154498159885407,
      "num_tokens": 55462503.0,
      "step": 26730
    },
    {
      "entropy": 5.639528274536133,
      "epoch": 2.4251632801161103,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004411338323536712,
      "loss": 5.0724,
      "mean_token_accuracy": 0.2096162259578705,
      "num_tokens": 55472789.0,
      "step": 26735
    },
    {
      "entropy": 5.54489369392395,
      "epoch": 2.4256168359941945,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004411120096928203,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2140708476305008,
      "num_tokens": 55482572.0,
      "step": 26740
    },
    {
      "entropy": 5.596988201141357,
      "epoch": 2.426070391872279,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004410901835967306,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21821867376565934,
      "num_tokens": 55493033.0,
      "step": 26745
    },
    {
      "entropy": 5.57719464302063,
      "epoch": 2.426523947750363,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044106835406585354,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21200233548879624,
      "num_tokens": 55503574.0,
      "step": 26750
    },
    {
      "entropy": 5.626631355285644,
      "epoch": 2.426977503628447,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044104652110064067,
      "loss": 4.9736,
      "mean_token_accuracy": 0.20684119313955307,
      "num_tokens": 55514119.0,
      "step": 26755
    },
    {
      "entropy": 5.5654267311096195,
      "epoch": 2.427431059506531,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004410246847015435,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21291468888521195,
      "num_tokens": 55524973.0,
      "step": 26760
    },
    {
      "entropy": 5.560578012466431,
      "epoch": 2.4278846153846154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004410028448690134,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21115875989198685,
      "num_tokens": 55534847.0,
      "step": 26765
    },
    {
      "entropy": 5.625048875808716,
      "epoch": 2.4283381712626997,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004409810016035023,
      "loss": 4.9555,
      "mean_token_accuracy": 0.20490000545978546,
      "num_tokens": 55544886.0,
      "step": 26770
    },
    {
      "entropy": 5.583855676651001,
      "epoch": 2.4287917271407835,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004409591549054617,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20886138081550598,
      "num_tokens": 55554534.0,
      "step": 26775
    },
    {
      "entropy": 5.600439214706421,
      "epoch": 2.4292452830188678,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044093730477534355,
      "loss": 4.9783,
      "mean_token_accuracy": 0.206478089094162,
      "num_tokens": 55564302.0,
      "step": 26780
    },
    {
      "entropy": 5.533636617660522,
      "epoch": 2.429698838896952,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004409154512135996,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21612302958965302,
      "num_tokens": 55574543.0,
      "step": 26785
    },
    {
      "entropy": 5.534697389602661,
      "epoch": 2.4301523947750363,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004408935942206818,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21744074523448945,
      "num_tokens": 55583849.0,
      "step": 26790
    },
    {
      "entropy": 5.5638738632202145,
      "epoch": 2.4306059506531206,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004408717337970423,
      "loss": 4.9681,
      "mean_token_accuracy": 0.20987699925899506,
      "num_tokens": 55594294.0,
      "step": 26795
    },
    {
      "entropy": 5.579269886016846,
      "epoch": 2.431059506531205,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044084986994313304,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21099089533090593,
      "num_tokens": 55604185.0,
      "step": 26800
    },
    {
      "entropy": 5.560730695724487,
      "epoch": 2.4315130624092887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004408280026594062,
      "loss": 4.9014,
      "mean_token_accuracy": 0.214804445207119,
      "num_tokens": 55615203.0,
      "step": 26805
    },
    {
      "entropy": 5.561976432800293,
      "epoch": 2.431966618287373,
      "grad_norm": 1.125,
      "learning_rate": 0.000440806131946314,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21126452535390855,
      "num_tokens": 55626157.0,
      "step": 26810
    },
    {
      "entropy": 5.628096294403076,
      "epoch": 2.432420174165457,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044078425780430877,
      "loss": 4.956,
      "mean_token_accuracy": 0.20642924904823304,
      "num_tokens": 55636862.0,
      "step": 26815
    },
    {
      "entropy": 5.566771030426025,
      "epoch": 2.4328737300435415,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004407623802338428,
      "loss": 4.8661,
      "mean_token_accuracy": 0.2200087621808052,
      "num_tokens": 55647646.0,
      "step": 26820
    },
    {
      "entropy": 5.568132448196411,
      "epoch": 2.4333272859216257,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044074049923536855,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2192446321249008,
      "num_tokens": 55659847.0,
      "step": 26825
    },
    {
      "entropy": 5.514598560333252,
      "epoch": 2.4337808417997095,
      "grad_norm": 1.125,
      "learning_rate": 0.0004407186148093386,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21441839039325714,
      "num_tokens": 55670626.0,
      "step": 26830
    },
    {
      "entropy": 5.5989141941070555,
      "epoch": 2.434234397677794,
      "grad_norm": 1.125,
      "learning_rate": 0.00044069672695620543,
      "loss": 4.992,
      "mean_token_accuracy": 0.21164102256298065,
      "num_tokens": 55681368.0,
      "step": 26835
    },
    {
      "entropy": 5.559159183502198,
      "epoch": 2.434687953555878,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004406748356764217,
      "loss": 4.8726,
      "mean_token_accuracy": 0.20950720459222794,
      "num_tokens": 55691606.0,
      "step": 26840
    },
    {
      "entropy": 5.6074470520019535,
      "epoch": 2.4351415094339623,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044065294097044017,
      "loss": 5.0474,
      "mean_token_accuracy": 0.21100991666316987,
      "num_tokens": 55702223.0,
      "step": 26845
    },
    {
      "entropy": 5.5065570831298825,
      "epoch": 2.4355950653120466,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004406310428387136,
      "loss": 4.8449,
      "mean_token_accuracy": 0.21619075089693068,
      "num_tokens": 55712729.0,
      "step": 26850
    },
    {
      "entropy": 5.5781244277954105,
      "epoch": 2.4360486211901304,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004406091412816949,
      "loss": 4.9549,
      "mean_token_accuracy": 0.21669239848852156,
      "num_tokens": 55722863.0,
      "step": 26855
    },
    {
      "entropy": 5.645908784866333,
      "epoch": 2.4365021770682147,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044058723629983694,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21438144147396088,
      "num_tokens": 55733124.0,
      "step": 26860
    },
    {
      "entropy": 5.6440221786499025,
      "epoch": 2.436955732946299,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004405653278935927,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21672249734401702,
      "num_tokens": 55743438.0,
      "step": 26865
    },
    {
      "entropy": 5.510346174240112,
      "epoch": 2.4374092888243832,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044054341606341533,
      "loss": 4.8227,
      "mean_token_accuracy": 0.21453460305929184,
      "num_tokens": 55753602.0,
      "step": 26870
    },
    {
      "entropy": 5.505764961242676,
      "epoch": 2.4378628447024675,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004405215008097579,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22253549844026566,
      "num_tokens": 55763473.0,
      "step": 26875
    },
    {
      "entropy": 5.59206075668335,
      "epoch": 2.4383164005805513,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044049958213307364,
      "loss": 4.8864,
      "mean_token_accuracy": 0.21952882260084153,
      "num_tokens": 55772848.0,
      "step": 26880
    },
    {
      "entropy": 5.527996444702149,
      "epoch": 2.4387699564586356,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044047766003381584,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22041210532188416,
      "num_tokens": 55783468.0,
      "step": 26885
    },
    {
      "entropy": 5.537233209609985,
      "epoch": 2.43922351233672,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044045573451243793,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21993866711854934,
      "num_tokens": 55793619.0,
      "step": 26890
    },
    {
      "entropy": 5.569816637039184,
      "epoch": 2.439677068214804,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044043380556939326,
      "loss": 4.9353,
      "mean_token_accuracy": 0.20997764617204667,
      "num_tokens": 55803869.0,
      "step": 26895
    },
    {
      "entropy": 5.627328300476075,
      "epoch": 2.4401306240928884,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004404118732051353,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21249611228704451,
      "num_tokens": 55814754.0,
      "step": 26900
    },
    {
      "entropy": 5.576613998413086,
      "epoch": 2.4405841799709727,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004403899374201176,
      "loss": 4.888,
      "mean_token_accuracy": 0.21150109469890593,
      "num_tokens": 55824862.0,
      "step": 26905
    },
    {
      "entropy": 5.489608192443848,
      "epoch": 2.4410377358490565,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044036799821479397,
      "loss": 4.8968,
      "mean_token_accuracy": 0.21324735581874849,
      "num_tokens": 55835506.0,
      "step": 26910
    },
    {
      "entropy": 5.601044845581055,
      "epoch": 2.4414912917271407,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044034605558961786,
      "loss": 4.9838,
      "mean_token_accuracy": 0.20867536216974258,
      "num_tokens": 55846244.0,
      "step": 26915
    },
    {
      "entropy": 5.527788400650024,
      "epoch": 2.441944847605225,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044032410954504325,
      "loss": 4.8862,
      "mean_token_accuracy": 0.222291462123394,
      "num_tokens": 55856192.0,
      "step": 26920
    },
    {
      "entropy": 5.570795631408691,
      "epoch": 2.4423984034833093,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044030216008152394,
      "loss": 4.9374,
      "mean_token_accuracy": 0.2087250292301178,
      "num_tokens": 55866509.0,
      "step": 26925
    },
    {
      "entropy": 5.595271015167237,
      "epoch": 2.442851959361393,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044028020719951376,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21611881256103516,
      "num_tokens": 55877321.0,
      "step": 26930
    },
    {
      "entropy": 5.596599817276001,
      "epoch": 2.4433055152394774,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004402582508994669,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2107618883252144,
      "num_tokens": 55887933.0,
      "step": 26935
    },
    {
      "entropy": 5.519869232177735,
      "epoch": 2.4437590711175616,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044023629118183726,
      "loss": 4.8762,
      "mean_token_accuracy": 0.21172595918178558,
      "num_tokens": 55897274.0,
      "step": 26940
    },
    {
      "entropy": 5.589644050598144,
      "epoch": 2.444212626995646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044021432804707895,
      "loss": 5.0299,
      "mean_token_accuracy": 0.20810270458459854,
      "num_tokens": 55907393.0,
      "step": 26945
    },
    {
      "entropy": 5.546515941619873,
      "epoch": 2.44466618287373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004401923614956463,
      "loss": 4.9148,
      "mean_token_accuracy": 0.20886614173650742,
      "num_tokens": 55918429.0,
      "step": 26950
    },
    {
      "entropy": 5.569817018508911,
      "epoch": 2.4451197387518144,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004401703915279935,
      "loss": 4.8892,
      "mean_token_accuracy": 0.21486484855413437,
      "num_tokens": 55928733.0,
      "step": 26955
    },
    {
      "entropy": 5.615135669708252,
      "epoch": 2.4455732946298983,
      "grad_norm": 1.125,
      "learning_rate": 0.00044014841814457486,
      "loss": 4.88,
      "mean_token_accuracy": 0.2158233180642128,
      "num_tokens": 55939128.0,
      "step": 26960
    },
    {
      "entropy": 5.612682914733886,
      "epoch": 2.4460268505079825,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004401264413458449,
      "loss": 5.045,
      "mean_token_accuracy": 0.2082000896334648,
      "num_tokens": 55950237.0,
      "step": 26965
    },
    {
      "entropy": 5.577176237106324,
      "epoch": 2.446480406386067,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004401044611322581,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22138510644435883,
      "num_tokens": 55960436.0,
      "step": 26970
    },
    {
      "entropy": 5.550209188461304,
      "epoch": 2.446933962264151,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004400824775042689,
      "loss": 4.8352,
      "mean_token_accuracy": 0.21578322649002074,
      "num_tokens": 55970446.0,
      "step": 26975
    },
    {
      "entropy": 5.563942766189575,
      "epoch": 2.447387518142235,
      "grad_norm": 1.125,
      "learning_rate": 0.000440060490462332,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21210048198699952,
      "num_tokens": 55980706.0,
      "step": 26980
    },
    {
      "entropy": 5.5506431579589846,
      "epoch": 2.447841074020319,
      "grad_norm": 1.125,
      "learning_rate": 0.0004400385000069021,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2054441064596176,
      "num_tokens": 55990669.0,
      "step": 26985
    },
    {
      "entropy": 5.665261888504029,
      "epoch": 2.4482946298984034,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044001650613843395,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2040278732776642,
      "num_tokens": 56000709.0,
      "step": 26990
    },
    {
      "entropy": 5.625793933868408,
      "epoch": 2.4487481857764877,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004399945088573824,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22189556807279587,
      "num_tokens": 56011435.0,
      "step": 26995
    },
    {
      "entropy": 5.586576795578003,
      "epoch": 2.449201741654572,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043997250816420246,
      "loss": 4.9081,
      "mean_token_accuracy": 0.21742316484451293,
      "num_tokens": 56021264.0,
      "step": 27000
    },
    {
      "epoch": 2.449201741654572,
      "eval_entropy": 5.310908531495349,
      "eval_loss": 5.009191989898682,
      "eval_mean_token_accuracy": 0.2188378827165964,
      "eval_num_tokens": 56021264.0,
      "eval_runtime": 20.6659,
      "eval_samples_per_second": 1711.035,
      "eval_steps_per_second": 213.879,
      "step": 27000
    },
    {
      "entropy": 5.570468521118164,
      "epoch": 2.449655297532656,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004399505040593489,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21505245566368103,
      "num_tokens": 56031463.0,
      "step": 27005
    },
    {
      "entropy": 5.518310689926148,
      "epoch": 2.45010885341074,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043992849654327695,
      "loss": 4.8361,
      "mean_token_accuracy": 0.22150550931692123,
      "num_tokens": 56041195.0,
      "step": 27010
    },
    {
      "entropy": 5.566996955871582,
      "epoch": 2.4505624092888243,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004399064856164416,
      "loss": 4.9708,
      "mean_token_accuracy": 0.20919194519519807,
      "num_tokens": 56051840.0,
      "step": 27015
    },
    {
      "entropy": 5.593413543701172,
      "epoch": 2.4510159651669086,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043988447127929817,
      "loss": 4.9754,
      "mean_token_accuracy": 0.21148500442504883,
      "num_tokens": 56062535.0,
      "step": 27020
    },
    {
      "entropy": 5.578488445281982,
      "epoch": 2.451469521044993,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043986245353230173,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22253292053937912,
      "num_tokens": 56072177.0,
      "step": 27025
    },
    {
      "entropy": 5.546461963653565,
      "epoch": 2.451923076923077,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043984043237590785,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2161324664950371,
      "num_tokens": 56081607.0,
      "step": 27030
    },
    {
      "entropy": 5.568353462219238,
      "epoch": 2.452376632801161,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043981840781057185,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20944924652576447,
      "num_tokens": 56093371.0,
      "step": 27035
    },
    {
      "entropy": 5.5949352264404295,
      "epoch": 2.452830188679245,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043979637983674913,
      "loss": 4.9431,
      "mean_token_accuracy": 0.20938770920038224,
      "num_tokens": 56103973.0,
      "step": 27040
    },
    {
      "entropy": 5.565538311004639,
      "epoch": 2.4532837445573294,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004397743484548953,
      "loss": 4.9291,
      "mean_token_accuracy": 0.20909886062145233,
      "num_tokens": 56113906.0,
      "step": 27045
    },
    {
      "entropy": 5.573693943023682,
      "epoch": 2.4537373004354137,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004397523136654659,
      "loss": 4.9105,
      "mean_token_accuracy": 0.210191810131073,
      "num_tokens": 56124172.0,
      "step": 27050
    },
    {
      "entropy": 5.5464723110198975,
      "epoch": 2.454190856313498,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043973027546891677,
      "loss": 4.8016,
      "mean_token_accuracy": 0.22593728750944136,
      "num_tokens": 56133088.0,
      "step": 27055
    },
    {
      "entropy": 5.549463558197021,
      "epoch": 2.4546444121915822,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004397082338657034,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21861016154289245,
      "num_tokens": 56143098.0,
      "step": 27060
    },
    {
      "entropy": 5.53235502243042,
      "epoch": 2.455097968069666,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043968618885628185,
      "loss": 4.8941,
      "mean_token_accuracy": 0.2166932627558708,
      "num_tokens": 56153889.0,
      "step": 27065
    },
    {
      "entropy": 5.540852499008179,
      "epoch": 2.4555515239477503,
      "grad_norm": 1.25,
      "learning_rate": 0.00043966414044110805,
      "loss": 4.8671,
      "mean_token_accuracy": 0.21665560603141784,
      "num_tokens": 56164164.0,
      "step": 27070
    },
    {
      "entropy": 5.6239914894104,
      "epoch": 2.4560050798258346,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043964208862063775,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20838364064693451,
      "num_tokens": 56173172.0,
      "step": 27075
    },
    {
      "entropy": 5.590054845809936,
      "epoch": 2.456458635703919,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043962003339532714,
      "loss": 4.925,
      "mean_token_accuracy": 0.21700985580682755,
      "num_tokens": 56183127.0,
      "step": 27080
    },
    {
      "entropy": 5.617047834396362,
      "epoch": 2.4569121915820027,
      "grad_norm": 1.125,
      "learning_rate": 0.00043959797476563226,
      "loss": 4.9626,
      "mean_token_accuracy": 0.20998311936855316,
      "num_tokens": 56192528.0,
      "step": 27085
    },
    {
      "entropy": 5.546414995193482,
      "epoch": 2.457365747460087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043957591273200943,
      "loss": 4.9151,
      "mean_token_accuracy": 0.21502474844455718,
      "num_tokens": 56203239.0,
      "step": 27090
    },
    {
      "entropy": 5.59856219291687,
      "epoch": 2.4578193033381712,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004395538472949146,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20955437868833543,
      "num_tokens": 56214068.0,
      "step": 27095
    },
    {
      "entropy": 5.61622805595398,
      "epoch": 2.4582728592162555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004395317784548045,
      "loss": 4.9749,
      "mean_token_accuracy": 0.20501068830490113,
      "num_tokens": 56224829.0,
      "step": 27100
    },
    {
      "entropy": 5.663437223434448,
      "epoch": 2.4587264150943398,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004395097062121351,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20961254239082336,
      "num_tokens": 56235412.0,
      "step": 27105
    },
    {
      "entropy": 5.640355491638184,
      "epoch": 2.459179970972424,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043948763056736315,
      "loss": 4.9944,
      "mean_token_accuracy": 0.2067007914185524,
      "num_tokens": 56246785.0,
      "step": 27110
    },
    {
      "entropy": 5.5982475757598875,
      "epoch": 2.459633526850508,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000439465551520945,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20904777944087982,
      "num_tokens": 56256701.0,
      "step": 27115
    },
    {
      "entropy": 5.563984632492065,
      "epoch": 2.460087082728592,
      "grad_norm": 1.125,
      "learning_rate": 0.00043944346907333744,
      "loss": 4.9462,
      "mean_token_accuracy": 0.21463871151208877,
      "num_tokens": 56267418.0,
      "step": 27120
    },
    {
      "entropy": 5.562194728851319,
      "epoch": 2.4605406386066764,
      "grad_norm": 1.015625,
      "learning_rate": 0.000439421383224997,
      "loss": 4.8816,
      "mean_token_accuracy": 0.21297574192285537,
      "num_tokens": 56278232.0,
      "step": 27125
    },
    {
      "entropy": 5.635591316223144,
      "epoch": 2.4609941944847606,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004393992939763804,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21307477504014968,
      "num_tokens": 56289866.0,
      "step": 27130
    },
    {
      "entropy": 5.593373823165893,
      "epoch": 2.4614477503628445,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043937720132794457,
      "loss": 4.9077,
      "mean_token_accuracy": 0.21929869055747986,
      "num_tokens": 56299595.0,
      "step": 27135
    },
    {
      "entropy": 5.610020589828491,
      "epoch": 2.4619013062409287,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043935510528014625,
      "loss": 4.9751,
      "mean_token_accuracy": 0.21973224878311157,
      "num_tokens": 56310152.0,
      "step": 27140
    },
    {
      "entropy": 5.591468572616577,
      "epoch": 2.462354862119013,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004393330058334426,
      "loss": 5.0402,
      "mean_token_accuracy": 0.20449508130550384,
      "num_tokens": 56320288.0,
      "step": 27145
    },
    {
      "entropy": 5.622226905822754,
      "epoch": 2.4628084179970973,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043931090298829047,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20896437019109726,
      "num_tokens": 56330345.0,
      "step": 27150
    },
    {
      "entropy": 5.662629175186157,
      "epoch": 2.4632619738751815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043928879674514694,
      "loss": 5.0846,
      "mean_token_accuracy": 0.21637462973594665,
      "num_tokens": 56340996.0,
      "step": 27155
    },
    {
      "entropy": 5.495919227600098,
      "epoch": 2.463715529753266,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004392666871044692,
      "loss": 4.8101,
      "mean_token_accuracy": 0.22434767037630082,
      "num_tokens": 56350326.0,
      "step": 27160
    },
    {
      "entropy": 5.587428092956543,
      "epoch": 2.4641690856313496,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043924457406671456,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2184249699115753,
      "num_tokens": 56360556.0,
      "step": 27165
    },
    {
      "entropy": 5.617318773269654,
      "epoch": 2.464622641509434,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004392224576323403,
      "loss": 4.9017,
      "mean_token_accuracy": 0.21992544382810592,
      "num_tokens": 56369980.0,
      "step": 27170
    },
    {
      "entropy": 5.587795972824097,
      "epoch": 2.465076197387518,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043920033780180367,
      "loss": 4.9503,
      "mean_token_accuracy": 0.20803798735141754,
      "num_tokens": 56379974.0,
      "step": 27175
    },
    {
      "entropy": 5.587786102294922,
      "epoch": 2.4655297532656024,
      "grad_norm": 1.125,
      "learning_rate": 0.0004391782145755623,
      "loss": 4.8843,
      "mean_token_accuracy": 0.21582925468683242,
      "num_tokens": 56390542.0,
      "step": 27180
    },
    {
      "entropy": 5.689225053787231,
      "epoch": 2.4659833091436867,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004391560879540736,
      "loss": 5.0431,
      "mean_token_accuracy": 0.20784496814012526,
      "num_tokens": 56401755.0,
      "step": 27185
    },
    {
      "entropy": 5.564447259902954,
      "epoch": 2.4664368650217705,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004391339579377952,
      "loss": 4.9341,
      "mean_token_accuracy": 0.2139232873916626,
      "num_tokens": 56412223.0,
      "step": 27190
    },
    {
      "entropy": 5.60780782699585,
      "epoch": 2.4668904208998548,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043911182452718476,
      "loss": 4.9794,
      "mean_token_accuracy": 0.21333203166723252,
      "num_tokens": 56422947.0,
      "step": 27195
    },
    {
      "entropy": 5.5757640361785885,
      "epoch": 2.467343976777939,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043908968772269984,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2183648630976677,
      "num_tokens": 56433189.0,
      "step": 27200
    },
    {
      "entropy": 5.620698690414429,
      "epoch": 2.4677975326560233,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043906754752479836,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21366307139396667,
      "num_tokens": 56444460.0,
      "step": 27205
    },
    {
      "entropy": 5.621269369125367,
      "epoch": 2.4682510885341076,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004390454039339383,
      "loss": 4.9845,
      "mean_token_accuracy": 0.2081169918179512,
      "num_tokens": 56454368.0,
      "step": 27210
    },
    {
      "entropy": 5.539461755752564,
      "epoch": 2.4687046444121914,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043902325695057743,
      "loss": 4.8793,
      "mean_token_accuracy": 0.21865448653697966,
      "num_tokens": 56464153.0,
      "step": 27215
    },
    {
      "entropy": 5.555964565277099,
      "epoch": 2.4691582002902757,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043900110657517377,
      "loss": 4.9059,
      "mean_token_accuracy": 0.21785708516836166,
      "num_tokens": 56473627.0,
      "step": 27220
    },
    {
      "entropy": 5.588455629348755,
      "epoch": 2.46961175616836,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004389789528081855,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2176984205842018,
      "num_tokens": 56483660.0,
      "step": 27225
    },
    {
      "entropy": 5.618128061294556,
      "epoch": 2.470065312046444,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004389567956500707,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21988930851221083,
      "num_tokens": 56492653.0,
      "step": 27230
    },
    {
      "entropy": 5.671018075942993,
      "epoch": 2.4705188679245285,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004389346351012876,
      "loss": 5.0523,
      "mean_token_accuracy": 0.2028290942311287,
      "num_tokens": 56503289.0,
      "step": 27235
    },
    {
      "entropy": 5.635839033126831,
      "epoch": 2.4709724238026123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043891247116229436,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20613661110401155,
      "num_tokens": 56513843.0,
      "step": 27240
    },
    {
      "entropy": 5.625729131698608,
      "epoch": 2.4714259796806965,
      "grad_norm": 1.125,
      "learning_rate": 0.00043889030383354957,
      "loss": 4.97,
      "mean_token_accuracy": 0.20706405639648437,
      "num_tokens": 56524175.0,
      "step": 27245
    },
    {
      "entropy": 5.5597631454467775,
      "epoch": 2.471879535558781,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004388681331155114,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2226969212293625,
      "num_tokens": 56534976.0,
      "step": 27250
    },
    {
      "entropy": 5.607913303375244,
      "epoch": 2.472333091436865,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004388459590086386,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21343802511692048,
      "num_tokens": 56545967.0,
      "step": 27255
    },
    {
      "entropy": 5.615893077850342,
      "epoch": 2.4727866473149493,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043882378151338956,
      "loss": 4.9931,
      "mean_token_accuracy": 0.20648104101419448,
      "num_tokens": 56555527.0,
      "step": 27260
    },
    {
      "entropy": 5.625734519958496,
      "epoch": 2.4732402031930336,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000438801600630223,
      "loss": 4.9561,
      "mean_token_accuracy": 0.2064664050936699,
      "num_tokens": 56565874.0,
      "step": 27265
    },
    {
      "entropy": 5.577193641662598,
      "epoch": 2.4736937590711174,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043877941635959755,
      "loss": 4.882,
      "mean_token_accuracy": 0.21218475699424744,
      "num_tokens": 56575452.0,
      "step": 27270
    },
    {
      "entropy": 5.497557878494263,
      "epoch": 2.4741473149492017,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043875722870197214,
      "loss": 4.8554,
      "mean_token_accuracy": 0.2242387428879738,
      "num_tokens": 56586176.0,
      "step": 27275
    },
    {
      "entropy": 5.629019355773925,
      "epoch": 2.474600870827286,
      "grad_norm": 1.25,
      "learning_rate": 0.0004387350376578054,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21332286596298217,
      "num_tokens": 56595473.0,
      "step": 27280
    },
    {
      "entropy": 5.600218296051025,
      "epoch": 2.4750544267053702,
      "grad_norm": 1.125,
      "learning_rate": 0.0004387128432275563,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21243224143981934,
      "num_tokens": 56606016.0,
      "step": 27285
    },
    {
      "entropy": 5.559001731872558,
      "epoch": 2.475507982583454,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000438690645411684,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2098272144794464,
      "num_tokens": 56617674.0,
      "step": 27290
    },
    {
      "entropy": 5.595889949798584,
      "epoch": 2.4759615384615383,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043866844421064745,
      "loss": 4.918,
      "mean_token_accuracy": 0.21643660962581635,
      "num_tokens": 56628077.0,
      "step": 27295
    },
    {
      "entropy": 5.619894218444824,
      "epoch": 2.4764150943396226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043864623962490575,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2159027114510536,
      "num_tokens": 56638691.0,
      "step": 27300
    },
    {
      "entropy": 5.590474224090576,
      "epoch": 2.476868650217707,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004386240316549181,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21426018923521042,
      "num_tokens": 56650338.0,
      "step": 27305
    },
    {
      "entropy": 5.562205076217651,
      "epoch": 2.477322206095791,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043860182030114374,
      "loss": 4.8651,
      "mean_token_accuracy": 0.2227303832769394,
      "num_tokens": 56660328.0,
      "step": 27310
    },
    {
      "entropy": 5.633257436752319,
      "epoch": 2.4777757619738754,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043857960556404214,
      "loss": 5.008,
      "mean_token_accuracy": 0.21100738793611526,
      "num_tokens": 56670674.0,
      "step": 27315
    },
    {
      "entropy": 5.5805905818939205,
      "epoch": 2.478229317851959,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004385573874440725,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2148086979985237,
      "num_tokens": 56680656.0,
      "step": 27320
    },
    {
      "entropy": 5.547960376739502,
      "epoch": 2.4786828737300435,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043853516594169446,
      "loss": 4.8495,
      "mean_token_accuracy": 0.22031059563159944,
      "num_tokens": 56691039.0,
      "step": 27325
    },
    {
      "entropy": 5.58705849647522,
      "epoch": 2.4791364296081277,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004385129410573675,
      "loss": 4.9222,
      "mean_token_accuracy": 0.20640294700860978,
      "num_tokens": 56703101.0,
      "step": 27330
    },
    {
      "entropy": 5.619821214675904,
      "epoch": 2.479589985486212,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043849071279155124,
      "loss": 4.9872,
      "mean_token_accuracy": 0.218429596722126,
      "num_tokens": 56712351.0,
      "step": 27335
    },
    {
      "entropy": 5.572011566162109,
      "epoch": 2.480043541364296,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043846848114470537,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22059279680252075,
      "num_tokens": 56722836.0,
      "step": 27340
    },
    {
      "entropy": 5.606530284881591,
      "epoch": 2.48049709724238,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043844624611728974,
      "loss": 4.9501,
      "mean_token_accuracy": 0.20923953354358674,
      "num_tokens": 56734032.0,
      "step": 27345
    },
    {
      "entropy": 5.61112380027771,
      "epoch": 2.4809506531204644,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043842400770976406,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21352135837078096,
      "num_tokens": 56743536.0,
      "step": 27350
    },
    {
      "entropy": 5.558936548233032,
      "epoch": 2.4814042089985486,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004384017659225882,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21404812783002852,
      "num_tokens": 56753475.0,
      "step": 27355
    },
    {
      "entropy": 5.512931776046753,
      "epoch": 2.481857764876633,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043837952075622224,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2228367283940315,
      "num_tokens": 56763602.0,
      "step": 27360
    },
    {
      "entropy": 5.564649868011474,
      "epoch": 2.482311320754717,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004383572722111261,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2213428407907486,
      "num_tokens": 56773859.0,
      "step": 27365
    },
    {
      "entropy": 5.554527521133423,
      "epoch": 2.482764876632801,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043833502028775995,
      "loss": 4.9266,
      "mean_token_accuracy": 0.21437181979417802,
      "num_tokens": 56783810.0,
      "step": 27370
    },
    {
      "entropy": 5.593163251876831,
      "epoch": 2.4832184325108853,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043831276498658395,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20869463980197905,
      "num_tokens": 56794088.0,
      "step": 27375
    },
    {
      "entropy": 5.510219144821167,
      "epoch": 2.4836719883889695,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004382905063080584,
      "loss": 4.8682,
      "mean_token_accuracy": 0.21661886125802993,
      "num_tokens": 56804623.0,
      "step": 27380
    },
    {
      "entropy": 5.548147821426392,
      "epoch": 2.484125544267054,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043826824425264357,
      "loss": 4.932,
      "mean_token_accuracy": 0.2082523375749588,
      "num_tokens": 56813842.0,
      "step": 27385
    },
    {
      "entropy": 5.619441175460816,
      "epoch": 2.484579100145138,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004382459788207997,
      "loss": 4.9312,
      "mean_token_accuracy": 0.20921426266431808,
      "num_tokens": 56823501.0,
      "step": 27390
    },
    {
      "entropy": 5.594589424133301,
      "epoch": 2.485032656023222,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043822371001298754,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21391973346471788,
      "num_tokens": 56834601.0,
      "step": 27395
    },
    {
      "entropy": 5.54300012588501,
      "epoch": 2.485486211901306,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004382014378296674,
      "loss": 4.9594,
      "mean_token_accuracy": 0.20829545557498932,
      "num_tokens": 56844701.0,
      "step": 27400
    },
    {
      "entropy": 5.532312822341919,
      "epoch": 2.4859397677793904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004381791622713,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21718046814203262,
      "num_tokens": 56854867.0,
      "step": 27405
    },
    {
      "entropy": 5.629356527328492,
      "epoch": 2.4863933236574747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000438156883338346,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21532835215330123,
      "num_tokens": 56865688.0,
      "step": 27410
    },
    {
      "entropy": 5.6419556617736815,
      "epoch": 2.486846879535559,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043813460103126595,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2146413117647171,
      "num_tokens": 56876504.0,
      "step": 27415
    },
    {
      "entropy": 5.575489521026611,
      "epoch": 2.487300435413643,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043811231535052076,
      "loss": 4.898,
      "mean_token_accuracy": 0.21808527261018754,
      "num_tokens": 56886083.0,
      "step": 27420
    },
    {
      "entropy": 5.528043127059936,
      "epoch": 2.487753991291727,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004380900262965714,
      "loss": 4.9758,
      "mean_token_accuracy": 0.21830589771270753,
      "num_tokens": 56897079.0,
      "step": 27425
    },
    {
      "entropy": 5.58110933303833,
      "epoch": 2.4882075471698113,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004380677338698787,
      "loss": 4.8889,
      "mean_token_accuracy": 0.21870796531438827,
      "num_tokens": 56906962.0,
      "step": 27430
    },
    {
      "entropy": 5.633279848098755,
      "epoch": 2.4886611030478956,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00043804543807090374,
      "loss": 5.0077,
      "mean_token_accuracy": 0.20929981619119645,
      "num_tokens": 56918156.0,
      "step": 27435
    },
    {
      "entropy": 5.578549575805664,
      "epoch": 2.48911465892598,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043802313890010757,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2116858184337616,
      "num_tokens": 56928888.0,
      "step": 27440
    },
    {
      "entropy": 5.604401922225952,
      "epoch": 2.4895682148040637,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004380008363579513,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21217681914567948,
      "num_tokens": 56939929.0,
      "step": 27445
    },
    {
      "entropy": 5.629007005691529,
      "epoch": 2.490021770682148,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043797853044489623,
      "loss": 4.8465,
      "mean_token_accuracy": 0.2207060232758522,
      "num_tokens": 56950992.0,
      "step": 27450
    },
    {
      "entropy": 5.669310283660889,
      "epoch": 2.490475326560232,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004379562211614036,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2106938138604164,
      "num_tokens": 56962118.0,
      "step": 27455
    },
    {
      "entropy": 5.563676357269287,
      "epoch": 2.4909288824383164,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004379339085079348,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21446316093206405,
      "num_tokens": 56972837.0,
      "step": 27460
    },
    {
      "entropy": 5.606867027282715,
      "epoch": 2.4913824383164007,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004379115924849512,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2118871331214905,
      "num_tokens": 56982794.0,
      "step": 27465
    },
    {
      "entropy": 5.67323784828186,
      "epoch": 2.491835994194485,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004378892730929145,
      "loss": 5.0805,
      "mean_token_accuracy": 0.2094934642314911,
      "num_tokens": 56993979.0,
      "step": 27470
    },
    {
      "entropy": 5.622083377838135,
      "epoch": 2.492289550072569,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043786695033228604,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21009259968996047,
      "num_tokens": 57004505.0,
      "step": 27475
    },
    {
      "entropy": 5.632649374008179,
      "epoch": 2.492743105950653,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043784462420352747,
      "loss": 4.97,
      "mean_token_accuracy": 0.20422477275133133,
      "num_tokens": 57014121.0,
      "step": 27480
    },
    {
      "entropy": 5.660454797744751,
      "epoch": 2.4931966618287373,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043782229470710067,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2153025597333908,
      "num_tokens": 57023748.0,
      "step": 27485
    },
    {
      "entropy": 5.545122814178467,
      "epoch": 2.4936502177068216,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004377999618434673,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21740242540836335,
      "num_tokens": 57034782.0,
      "step": 27490
    },
    {
      "entropy": 5.540149545669555,
      "epoch": 2.4941037735849054,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004377776256130892,
      "loss": 4.8597,
      "mean_token_accuracy": 0.2168115645647049,
      "num_tokens": 57045136.0,
      "step": 27495
    },
    {
      "entropy": 5.610564136505127,
      "epoch": 2.4945573294629897,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043775528601642833,
      "loss": 5.0736,
      "mean_token_accuracy": 0.20102916061878204,
      "num_tokens": 57055893.0,
      "step": 27500
    },
    {
      "entropy": 5.597240543365478,
      "epoch": 2.495010885341074,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043773294305394665,
      "loss": 4.981,
      "mean_token_accuracy": 0.211080926656723,
      "num_tokens": 57067189.0,
      "step": 27505
    },
    {
      "entropy": 5.6404197216033936,
      "epoch": 2.4954644412191582,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043771059672610636,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2158254489302635,
      "num_tokens": 57077374.0,
      "step": 27510
    },
    {
      "entropy": 5.595413112640381,
      "epoch": 2.4959179970972425,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043768824703336933,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21071065664291383,
      "num_tokens": 57088168.0,
      "step": 27515
    },
    {
      "entropy": 5.606458568572998,
      "epoch": 2.4963715529753268,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004376658939761979,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2151413932442665,
      "num_tokens": 57098743.0,
      "step": 27520
    },
    {
      "entropy": 5.577711009979248,
      "epoch": 2.4968251088534106,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004376435375550544,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2076060429215431,
      "num_tokens": 57109854.0,
      "step": 27525
    },
    {
      "entropy": 5.554155397415161,
      "epoch": 2.497278664731495,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043762117777040106,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22206318527460098,
      "num_tokens": 57120296.0,
      "step": 27530
    },
    {
      "entropy": 5.548946571350098,
      "epoch": 2.497732220609579,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004375988146227003,
      "loss": 4.9013,
      "mean_token_accuracy": 0.21424288153648377,
      "num_tokens": 57131295.0,
      "step": 27535
    },
    {
      "entropy": 5.577417039871216,
      "epoch": 2.4981857764876634,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004375764481124147,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21111445128917694,
      "num_tokens": 57141936.0,
      "step": 27540
    },
    {
      "entropy": 5.568736457824707,
      "epoch": 2.4986393323657476,
      "grad_norm": 1.125,
      "learning_rate": 0.00043755407824000664,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22450370639562606,
      "num_tokens": 57152510.0,
      "step": 27545
    },
    {
      "entropy": 5.604910373687744,
      "epoch": 2.4990928882438315,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043753170500593884,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21606001555919646,
      "num_tokens": 57162546.0,
      "step": 27550
    },
    {
      "entropy": 5.514680004119873,
      "epoch": 2.4995464441219157,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043750932841067397,
      "loss": 4.8462,
      "mean_token_accuracy": 0.21972191333770752,
      "num_tokens": 57174658.0,
      "step": 27555
    },
    {
      "entropy": 5.641879987716675,
      "epoch": 2.5,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043748694845467483,
      "loss": 5.0723,
      "mean_token_accuracy": 0.19958110600709916,
      "num_tokens": 57185157.0,
      "step": 27560
    },
    {
      "entropy": 5.584434175491333,
      "epoch": 2.5004535558780843,
      "grad_norm": 1.125,
      "learning_rate": 0.00043746456513840405,
      "loss": 4.8558,
      "mean_token_accuracy": 0.21656010299921036,
      "num_tokens": 57195277.0,
      "step": 27565
    },
    {
      "entropy": 5.62018780708313,
      "epoch": 2.5009071117561685,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004374421784623248,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21126658618450164,
      "num_tokens": 57205374.0,
      "step": 27570
    },
    {
      "entropy": 5.558691692352295,
      "epoch": 2.501360667634253,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043741978842689987,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2144061177968979,
      "num_tokens": 57215700.0,
      "step": 27575
    },
    {
      "entropy": 5.600464963912964,
      "epoch": 2.5018142235123366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004373973950325923,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21464745849370956,
      "num_tokens": 57225765.0,
      "step": 27580
    },
    {
      "entropy": 5.617864799499512,
      "epoch": 2.502267779390421,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043737499827986527,
      "loss": 4.9416,
      "mean_token_accuracy": 0.21197375357151033,
      "num_tokens": 57235913.0,
      "step": 27585
    },
    {
      "entropy": 5.5413017749786375,
      "epoch": 2.502721335268505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043735259816918184,
      "loss": 4.8642,
      "mean_token_accuracy": 0.21646612286567687,
      "num_tokens": 57246204.0,
      "step": 27590
    },
    {
      "entropy": 5.529222631454468,
      "epoch": 2.5031748911465894,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004373301947010053,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20813624560832977,
      "num_tokens": 57256167.0,
      "step": 27595
    },
    {
      "entropy": 5.540852451324463,
      "epoch": 2.5036284470246732,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043730778787579895,
      "loss": 4.893,
      "mean_token_accuracy": 0.21697479486465454,
      "num_tokens": 57266621.0,
      "step": 27600
    },
    {
      "entropy": 5.609730052947998,
      "epoch": 2.5040820029027575,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043728537769402627,
      "loss": 4.9262,
      "mean_token_accuracy": 0.213581383228302,
      "num_tokens": 57277282.0,
      "step": 27605
    },
    {
      "entropy": 5.599507427215576,
      "epoch": 2.5045355587808418,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004372629641561506,
      "loss": 5.0027,
      "mean_token_accuracy": 0.2103717416524887,
      "num_tokens": 57287541.0,
      "step": 27610
    },
    {
      "entropy": 5.592021560668945,
      "epoch": 2.504989114658926,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043724054726263547,
      "loss": 4.9394,
      "mean_token_accuracy": 0.21137694418430328,
      "num_tokens": 57298158.0,
      "step": 27615
    },
    {
      "entropy": 5.625513219833374,
      "epoch": 2.5054426705370103,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004372181270139444,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20922743231058122,
      "num_tokens": 57308356.0,
      "step": 27620
    },
    {
      "entropy": 5.589460182189941,
      "epoch": 2.5058962264150946,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043719570341054127,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22005687952041625,
      "num_tokens": 57317905.0,
      "step": 27625
    },
    {
      "entropy": 5.579386281967163,
      "epoch": 2.5063497822931784,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043717327645288956,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21698225289583206,
      "num_tokens": 57328277.0,
      "step": 27630
    },
    {
      "entropy": 5.638761806488037,
      "epoch": 2.5068033381712627,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043715084614145316,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2112189382314682,
      "num_tokens": 57338975.0,
      "step": 27635
    },
    {
      "entropy": 5.586807107925415,
      "epoch": 2.507256894049347,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000437128412476696,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21520341485738753,
      "num_tokens": 57349439.0,
      "step": 27640
    },
    {
      "entropy": 5.584844493865967,
      "epoch": 2.507710449927431,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004371059754590819,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2099724069237709,
      "num_tokens": 57359414.0,
      "step": 27645
    },
    {
      "entropy": 5.644027757644653,
      "epoch": 2.508164005805515,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043708353508907494,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2152398109436035,
      "num_tokens": 57369711.0,
      "step": 27650
    },
    {
      "entropy": 5.55789589881897,
      "epoch": 2.5086175616835993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004370610913671392,
      "loss": 4.8678,
      "mean_token_accuracy": 0.22408596128225328,
      "num_tokens": 57380001.0,
      "step": 27655
    },
    {
      "entropy": 5.617283725738526,
      "epoch": 2.5090711175616836,
      "grad_norm": 1.25,
      "learning_rate": 0.00043703864429373875,
      "loss": 5.0598,
      "mean_token_accuracy": 0.1973607823252678,
      "num_tokens": 57390794.0,
      "step": 27660
    },
    {
      "entropy": 5.579990482330322,
      "epoch": 2.509524673439768,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043701619386933793,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20638088434934615,
      "num_tokens": 57401975.0,
      "step": 27665
    },
    {
      "entropy": 5.545427894592285,
      "epoch": 2.509978229317852,
      "grad_norm": 1.125,
      "learning_rate": 0.00043699374009440085,
      "loss": 4.8311,
      "mean_token_accuracy": 0.22356573939323426,
      "num_tokens": 57411249.0,
      "step": 27670
    },
    {
      "entropy": 5.5337066650390625,
      "epoch": 2.5104317851959363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004369712829693919,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2189682364463806,
      "num_tokens": 57422679.0,
      "step": 27675
    },
    {
      "entropy": 5.539751482009888,
      "epoch": 2.51088534107402,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004369488224947757,
      "loss": 4.935,
      "mean_token_accuracy": 0.21954985707998276,
      "num_tokens": 57433265.0,
      "step": 27680
    },
    {
      "entropy": 5.615825271606445,
      "epoch": 2.5113388969521044,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004369263586710164,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2133643940091133,
      "num_tokens": 57444095.0,
      "step": 27685
    },
    {
      "entropy": 5.599020051956177,
      "epoch": 2.5117924528301887,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004369038914985789,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2126259282231331,
      "num_tokens": 57454059.0,
      "step": 27690
    },
    {
      "entropy": 5.581152868270874,
      "epoch": 2.512246008708273,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004368814209779276,
      "loss": 4.8842,
      "mean_token_accuracy": 0.21774860769510268,
      "num_tokens": 57463940.0,
      "step": 27695
    },
    {
      "entropy": 5.526418876647949,
      "epoch": 2.512699564586357,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043685894710952724,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21330235451459884,
      "num_tokens": 57474731.0,
      "step": 27700
    },
    {
      "entropy": 5.542140340805053,
      "epoch": 2.513153120464441,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004368364698938426,
      "loss": 4.8719,
      "mean_token_accuracy": 0.2170833706855774,
      "num_tokens": 57485096.0,
      "step": 27705
    },
    {
      "entropy": 5.57192440032959,
      "epoch": 2.5136066763425253,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004368139893313385,
      "loss": 4.941,
      "mean_token_accuracy": 0.21175574958324433,
      "num_tokens": 57496371.0,
      "step": 27710
    },
    {
      "entropy": 5.624359273910523,
      "epoch": 2.5140602322206096,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043679150542248,
      "loss": 4.9086,
      "mean_token_accuracy": 0.2158511221408844,
      "num_tokens": 57506012.0,
      "step": 27715
    },
    {
      "entropy": 5.60762300491333,
      "epoch": 2.514513788098694,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043676901816773174,
      "loss": 4.97,
      "mean_token_accuracy": 0.21026700586080552,
      "num_tokens": 57516536.0,
      "step": 27720
    },
    {
      "entropy": 5.545338296890259,
      "epoch": 2.514967343976778,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004367465275675591,
      "loss": 4.859,
      "mean_token_accuracy": 0.21712665408849716,
      "num_tokens": 57527450.0,
      "step": 27725
    },
    {
      "entropy": 5.614373064041137,
      "epoch": 2.5154208998548624,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043672403362242704,
      "loss": 4.9882,
      "mean_token_accuracy": 0.21164269000291824,
      "num_tokens": 57538234.0,
      "step": 27730
    },
    {
      "entropy": 5.5832830429077145,
      "epoch": 2.515874455732946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043670153633280066,
      "loss": 4.911,
      "mean_token_accuracy": 0.21933975219726562,
      "num_tokens": 57547938.0,
      "step": 27735
    },
    {
      "entropy": 5.60312852859497,
      "epoch": 2.5163280116110305,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004366790356991454,
      "loss": 4.957,
      "mean_token_accuracy": 0.2140236094594002,
      "num_tokens": 57560553.0,
      "step": 27740
    },
    {
      "entropy": 5.5407579898834225,
      "epoch": 2.5167815674891147,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004366565317219265,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21860406696796417,
      "num_tokens": 57570670.0,
      "step": 27745
    },
    {
      "entropy": 5.665750646591187,
      "epoch": 2.5172351233671986,
      "grad_norm": 1.0,
      "learning_rate": 0.00043663402440160924,
      "loss": 5.1212,
      "mean_token_accuracy": 0.20040595531463623,
      "num_tokens": 57582377.0,
      "step": 27750
    },
    {
      "entropy": 5.58102650642395,
      "epoch": 2.517688679245283,
      "grad_norm": 1.125,
      "learning_rate": 0.0004366115137386592,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22278811037540436,
      "num_tokens": 57592764.0,
      "step": 27755
    },
    {
      "entropy": 5.569661664962768,
      "epoch": 2.518142235123367,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004365889997335419,
      "loss": 4.9453,
      "mean_token_accuracy": 0.20830441564321517,
      "num_tokens": 57603768.0,
      "step": 27760
    },
    {
      "entropy": 5.552487564086914,
      "epoch": 2.5185957910014514,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043656648238672285,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21595267355442047,
      "num_tokens": 57614346.0,
      "step": 27765
    },
    {
      "entropy": 5.673894596099854,
      "epoch": 2.5190493468795356,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004365439616986678,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20825873613357543,
      "num_tokens": 57625015.0,
      "step": 27770
    },
    {
      "entropy": 5.608899831771851,
      "epoch": 2.51950290275762,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043652143766984247,
      "loss": 4.9988,
      "mean_token_accuracy": 0.20780922323465348,
      "num_tokens": 57635476.0,
      "step": 27775
    },
    {
      "entropy": 5.542952394485473,
      "epoch": 2.519956458635704,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043649891030071264,
      "loss": 4.8917,
      "mean_token_accuracy": 0.21576763242483138,
      "num_tokens": 57645512.0,
      "step": 27780
    },
    {
      "entropy": 5.569958066940307,
      "epoch": 2.520410014513788,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043647637959174425,
      "loss": 4.898,
      "mean_token_accuracy": 0.21391591429710388,
      "num_tokens": 57655766.0,
      "step": 27785
    },
    {
      "entropy": 5.5870044231414795,
      "epoch": 2.5208635703918723,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043645384554340315,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21954991668462753,
      "num_tokens": 57667325.0,
      "step": 27790
    },
    {
      "entropy": 5.545009422302246,
      "epoch": 2.5213171262699565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043643130815615536,
      "loss": 4.9286,
      "mean_token_accuracy": 0.20707884728908538,
      "num_tokens": 57677884.0,
      "step": 27795
    },
    {
      "entropy": 5.632546091079712,
      "epoch": 2.521770682148041,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043640876743046694,
      "loss": 4.9589,
      "mean_token_accuracy": 0.22201118171215056,
      "num_tokens": 57687765.0,
      "step": 27800
    },
    {
      "entropy": 5.569636011123658,
      "epoch": 2.5222242380261246,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004363862233668041,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21295838207006454,
      "num_tokens": 57697777.0,
      "step": 27805
    },
    {
      "entropy": 5.5323882579803465,
      "epoch": 2.522677793904209,
      "grad_norm": 1.0625,
      "learning_rate": 0.000436363675965633,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21258897781372071,
      "num_tokens": 57708765.0,
      "step": 27810
    },
    {
      "entropy": 5.626805973052979,
      "epoch": 2.523131349782293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004363411252274201,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21650134921073913,
      "num_tokens": 57719658.0,
      "step": 27815
    },
    {
      "entropy": 5.5963387966156,
      "epoch": 2.5235849056603774,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043631857115263153,
      "loss": 4.9095,
      "mean_token_accuracy": 0.22033073008060455,
      "num_tokens": 57730018.0,
      "step": 27820
    },
    {
      "entropy": 5.50908899307251,
      "epoch": 2.5240384615384617,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004362960137417338,
      "loss": 4.865,
      "mean_token_accuracy": 0.2187896952033043,
      "num_tokens": 57741251.0,
      "step": 27825
    },
    {
      "entropy": 5.596281671524048,
      "epoch": 2.524492017416546,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004362734529951933,
      "loss": 4.9687,
      "mean_token_accuracy": 0.21493223160505295,
      "num_tokens": 57751020.0,
      "step": 27830
    },
    {
      "entropy": 5.573217535018921,
      "epoch": 2.5249455732946298,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043625088891347685,
      "loss": 4.9086,
      "mean_token_accuracy": 0.2189878150820732,
      "num_tokens": 57760989.0,
      "step": 27835
    },
    {
      "entropy": 5.633127069473266,
      "epoch": 2.525399129172714,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043622832149705086,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2171213984489441,
      "num_tokens": 57770882.0,
      "step": 27840
    },
    {
      "entropy": 5.6543049812316895,
      "epoch": 2.5258526850507983,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043620575074638206,
      "loss": 4.9974,
      "mean_token_accuracy": 0.2138979196548462,
      "num_tokens": 57781812.0,
      "step": 27845
    },
    {
      "entropy": 5.59008378982544,
      "epoch": 2.5263062409288826,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043618317666193734,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22132632434368132,
      "num_tokens": 57792009.0,
      "step": 27850
    },
    {
      "entropy": 5.541980600357055,
      "epoch": 2.5267597968069664,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043616059924418346,
      "loss": 4.908,
      "mean_token_accuracy": 0.2208719477057457,
      "num_tokens": 57802109.0,
      "step": 27855
    },
    {
      "entropy": 5.57148928642273,
      "epoch": 2.5272133526850507,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004361380184935872,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21691748201847078,
      "num_tokens": 57812374.0,
      "step": 27860
    },
    {
      "entropy": 5.536352157592773,
      "epoch": 2.527666908563135,
      "grad_norm": 1.125,
      "learning_rate": 0.0004361154344106158,
      "loss": 4.8634,
      "mean_token_accuracy": 0.21948933154344558,
      "num_tokens": 57822936.0,
      "step": 27865
    },
    {
      "entropy": 5.669306802749634,
      "epoch": 2.528120464441219,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043609284699573604,
      "loss": 5.0946,
      "mean_token_accuracy": 0.20776402503252028,
      "num_tokens": 57832693.0,
      "step": 27870
    },
    {
      "entropy": 5.644558334350586,
      "epoch": 2.5285740203193035,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043607025624941523,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21231866776943206,
      "num_tokens": 57843161.0,
      "step": 27875
    },
    {
      "entropy": 5.610720205307007,
      "epoch": 2.5290275761973877,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004360476621721204,
      "loss": 4.8953,
      "mean_token_accuracy": 0.22512662410736084,
      "num_tokens": 57853027.0,
      "step": 27880
    },
    {
      "entropy": 5.586286020278931,
      "epoch": 2.5294811320754715,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004360250647643189,
      "loss": 4.9348,
      "mean_token_accuracy": 0.22143570929765702,
      "num_tokens": 57862821.0,
      "step": 27885
    },
    {
      "entropy": 5.540043497085572,
      "epoch": 2.529934687953556,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043600246402647805,
      "loss": 4.8941,
      "mean_token_accuracy": 0.2220374420285225,
      "num_tokens": 57874065.0,
      "step": 27890
    },
    {
      "entropy": 5.566473865509034,
      "epoch": 2.53038824383164,
      "grad_norm": 1.125,
      "learning_rate": 0.0004359798599590652,
      "loss": 4.9617,
      "mean_token_accuracy": 0.20926931351423264,
      "num_tokens": 57885190.0,
      "step": 27895
    },
    {
      "entropy": 5.5794103145599365,
      "epoch": 2.5308417997097243,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043595725256254786,
      "loss": 4.9048,
      "mean_token_accuracy": 0.2170964553952217,
      "num_tokens": 57894234.0,
      "step": 27900
    },
    {
      "entropy": 5.614807987213135,
      "epoch": 2.531295355587808,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004359346418373935,
      "loss": 5.063,
      "mean_token_accuracy": 0.2021944284439087,
      "num_tokens": 57905401.0,
      "step": 27905
    },
    {
      "entropy": 5.544233179092407,
      "epoch": 2.5317489114658924,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043591202778406975,
      "loss": 4.9064,
      "mean_token_accuracy": 0.21363851577043533,
      "num_tokens": 57916097.0,
      "step": 27910
    },
    {
      "entropy": 5.510739517211914,
      "epoch": 2.5322024673439767,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004358894104030442,
      "loss": 4.8145,
      "mean_token_accuracy": 0.22497545927762985,
      "num_tokens": 57926871.0,
      "step": 27915
    },
    {
      "entropy": 5.534551811218262,
      "epoch": 2.532656023222061,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043586678969478464,
      "loss": 4.8758,
      "mean_token_accuracy": 0.22367361187934875,
      "num_tokens": 57937344.0,
      "step": 27920
    },
    {
      "entropy": 5.617660713195801,
      "epoch": 2.5331095791001452,
      "grad_norm": 1.1875,
      "learning_rate": 0.000435844165659759,
      "loss": 4.9639,
      "mean_token_accuracy": 0.216617850959301,
      "num_tokens": 57947633.0,
      "step": 27925
    },
    {
      "entropy": 5.567021083831787,
      "epoch": 2.5335631349782295,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043582153829843493,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22023017555475236,
      "num_tokens": 57957948.0,
      "step": 27930
    },
    {
      "entropy": 5.596801614761352,
      "epoch": 2.5340166908563138,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004357989076112805,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21556100100278855,
      "num_tokens": 57967719.0,
      "step": 27935
    },
    {
      "entropy": 5.601662635803223,
      "epoch": 2.5344702467343976,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004357762735987637,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21218422204256057,
      "num_tokens": 57978218.0,
      "step": 27940
    },
    {
      "entropy": 5.6088603973388675,
      "epoch": 2.534923802612482,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004357536362613525,
      "loss": 4.8648,
      "mean_token_accuracy": 0.220647431910038,
      "num_tokens": 57987860.0,
      "step": 27945
    },
    {
      "entropy": 5.56914529800415,
      "epoch": 2.535377358490566,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043573099559951516,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21944055706262589,
      "num_tokens": 57997872.0,
      "step": 27950
    },
    {
      "entropy": 5.638042306900024,
      "epoch": 2.5358309143686504,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004357083516137199,
      "loss": 4.9991,
      "mean_token_accuracy": 0.20960499048233033,
      "num_tokens": 58007771.0,
      "step": 27955
    },
    {
      "entropy": 5.593334531784057,
      "epoch": 2.536284470246734,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000435685704304435,
      "loss": 4.9307,
      "mean_token_accuracy": 0.21017040610313414,
      "num_tokens": 58018972.0,
      "step": 27960
    },
    {
      "entropy": 5.549252510070801,
      "epoch": 2.5367380261248185,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043566305367212884,
      "loss": 4.8182,
      "mean_token_accuracy": 0.21416020691394805,
      "num_tokens": 58028581.0,
      "step": 27965
    },
    {
      "entropy": 5.5674216747283936,
      "epoch": 2.5371915820029027,
      "grad_norm": 1.125,
      "learning_rate": 0.00043564039971726964,
      "loss": 4.9777,
      "mean_token_accuracy": 0.21665931344032288,
      "num_tokens": 58038877.0,
      "step": 27970
    },
    {
      "entropy": 5.575457763671875,
      "epoch": 2.537645137880987,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043561774244032614,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2171279102563858,
      "num_tokens": 58048777.0,
      "step": 27975
    },
    {
      "entropy": 5.625828075408935,
      "epoch": 2.5380986937590713,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043559508184176675,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2120449960231781,
      "num_tokens": 58060270.0,
      "step": 27980
    },
    {
      "entropy": 5.568502759933471,
      "epoch": 2.5385522496371555,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043557241792206014,
      "loss": 4.9201,
      "mean_token_accuracy": 0.21771903932094575,
      "num_tokens": 58070838.0,
      "step": 27985
    },
    {
      "entropy": 5.5928129196167,
      "epoch": 2.5390058055152394,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043554975068167505,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21062657237052917,
      "num_tokens": 58079933.0,
      "step": 27990
    },
    {
      "entropy": 5.541003656387329,
      "epoch": 2.5394593613933236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004355270801210802,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21689495742321013,
      "num_tokens": 58090673.0,
      "step": 27995
    },
    {
      "entropy": 5.543783330917359,
      "epoch": 2.539912917271408,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004355044062407444,
      "loss": 4.7834,
      "mean_token_accuracy": 0.22975331395864487,
      "num_tokens": 58100172.0,
      "step": 28000
    },
    {
      "entropy": 5.555199098587036,
      "epoch": 2.540366473149492,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004354817290411366,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21375659555196763,
      "num_tokens": 58108965.0,
      "step": 28005
    },
    {
      "entropy": 5.5565770149230955,
      "epoch": 2.540820029027576,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043545904852272573,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22217636704444885,
      "num_tokens": 58120094.0,
      "step": 28010
    },
    {
      "entropy": 5.5980805397033695,
      "epoch": 2.5412735849056602,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043543636468598083,
      "loss": 4.9855,
      "mean_token_accuracy": 0.20712201595306395,
      "num_tokens": 58131525.0,
      "step": 28015
    },
    {
      "entropy": 5.61816987991333,
      "epoch": 2.5417271407837445,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043541367753137107,
      "loss": 4.8746,
      "mean_token_accuracy": 0.21651554405689238,
      "num_tokens": 58141986.0,
      "step": 28020
    },
    {
      "entropy": 5.602068138122559,
      "epoch": 2.5421806966618288,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004353909870593656,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20951160788536072,
      "num_tokens": 58154122.0,
      "step": 28025
    },
    {
      "entropy": 5.578133010864258,
      "epoch": 2.542634252539913,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004353682932704336,
      "loss": 4.874,
      "mean_token_accuracy": 0.2249719262123108,
      "num_tokens": 58164148.0,
      "step": 28030
    },
    {
      "entropy": 5.5288416862487795,
      "epoch": 2.5430878084179973,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004353455961650444,
      "loss": 4.8661,
      "mean_token_accuracy": 0.2150352865457535,
      "num_tokens": 58173836.0,
      "step": 28035
    },
    {
      "entropy": 5.552236127853393,
      "epoch": 2.543541364296081,
      "grad_norm": 1.125,
      "learning_rate": 0.0004353228957436675,
      "loss": 4.9254,
      "mean_token_accuracy": 0.21269828528165818,
      "num_tokens": 58184598.0,
      "step": 28040
    },
    {
      "entropy": 5.551817512512207,
      "epoch": 2.5439949201741654,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043530019200677215,
      "loss": 4.8922,
      "mean_token_accuracy": 0.21802981048822404,
      "num_tokens": 58194820.0,
      "step": 28045
    },
    {
      "entropy": 5.565774393081665,
      "epoch": 2.5444484760522497,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043527748495482806,
      "loss": 4.9878,
      "mean_token_accuracy": 0.21097971498966217,
      "num_tokens": 58206589.0,
      "step": 28050
    },
    {
      "entropy": 5.617750787734986,
      "epoch": 2.544902031930334,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043525477458830473,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2160607174038887,
      "num_tokens": 58216750.0,
      "step": 28055
    },
    {
      "entropy": 5.548269557952881,
      "epoch": 2.5453555878084178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004352320609076718,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2209152415394783,
      "num_tokens": 58228496.0,
      "step": 28060
    },
    {
      "entropy": 5.584446954727173,
      "epoch": 2.545809143686502,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043520934391339894,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2128905862569809,
      "num_tokens": 58237844.0,
      "step": 28065
    },
    {
      "entropy": 5.508859348297119,
      "epoch": 2.5462626995645863,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004351866236059561,
      "loss": 4.8683,
      "mean_token_accuracy": 0.21740521341562272,
      "num_tokens": 58248522.0,
      "step": 28070
    },
    {
      "entropy": 5.531892919540406,
      "epoch": 2.5467162554426706,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043516389998581305,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2148062378168106,
      "num_tokens": 58258492.0,
      "step": 28075
    },
    {
      "entropy": 5.530713367462158,
      "epoch": 2.547169811320755,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004351411730534398,
      "loss": 4.9086,
      "mean_token_accuracy": 0.20897070467472076,
      "num_tokens": 58268683.0,
      "step": 28080
    },
    {
      "entropy": 5.585899686813354,
      "epoch": 2.547623367198839,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004351184428093062,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20405336916446687,
      "num_tokens": 58279479.0,
      "step": 28085
    },
    {
      "entropy": 5.622782564163208,
      "epoch": 2.5480769230769234,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004350957092538825,
      "loss": 5.0001,
      "mean_token_accuracy": 0.20888267159461976,
      "num_tokens": 58290751.0,
      "step": 28090
    },
    {
      "entropy": 5.630240869522095,
      "epoch": 2.548530478955007,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004350729723876386,
      "loss": 5.016,
      "mean_token_accuracy": 0.2077210769057274,
      "num_tokens": 58302163.0,
      "step": 28095
    },
    {
      "entropy": 5.586409091949463,
      "epoch": 2.5489840348330914,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043505023221104497,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2068890079855919,
      "num_tokens": 58311488.0,
      "step": 28100
    },
    {
      "entropy": 5.551432752609253,
      "epoch": 2.5494375907111757,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043502748872457166,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21354300826787947,
      "num_tokens": 58321599.0,
      "step": 28105
    },
    {
      "entropy": 5.565755414962768,
      "epoch": 2.5498911465892595,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004350047419286892,
      "loss": 4.9043,
      "mean_token_accuracy": 0.21592913568019867,
      "num_tokens": 58330696.0,
      "step": 28110
    },
    {
      "entropy": 5.545326375961304,
      "epoch": 2.550344702467344,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043498199182386775,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21693719923496246,
      "num_tokens": 58341793.0,
      "step": 28115
    },
    {
      "entropy": 5.607199239730835,
      "epoch": 2.550798258345428,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043495923841057807,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20901485979557038,
      "num_tokens": 58352260.0,
      "step": 28120
    },
    {
      "entropy": 5.676895236968994,
      "epoch": 2.5512518142235123,
      "grad_norm": 1.125,
      "learning_rate": 0.00043493648168929053,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20567869544029235,
      "num_tokens": 58363052.0,
      "step": 28125
    },
    {
      "entropy": 5.665903997421265,
      "epoch": 2.5517053701015966,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004349137216604758,
      "loss": 4.9623,
      "mean_token_accuracy": 0.21507498174905776,
      "num_tokens": 58374039.0,
      "step": 28130
    },
    {
      "entropy": 5.5490013599395756,
      "epoch": 2.552158925979681,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004348909583246045,
      "loss": 4.8045,
      "mean_token_accuracy": 0.22377109080553054,
      "num_tokens": 58383441.0,
      "step": 28135
    },
    {
      "entropy": 5.522780227661133,
      "epoch": 2.552612481857765,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004348681916821475,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21670966148376464,
      "num_tokens": 58394164.0,
      "step": 28140
    },
    {
      "entropy": 5.536119890213013,
      "epoch": 2.553066037735849,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004348454217335755,
      "loss": 4.9327,
      "mean_token_accuracy": 0.21522781401872634,
      "num_tokens": 58404610.0,
      "step": 28145
    },
    {
      "entropy": 5.562320899963379,
      "epoch": 2.553519593613933,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043482264847935953,
      "loss": 4.8768,
      "mean_token_accuracy": 0.21156506538391112,
      "num_tokens": 58414512.0,
      "step": 28150
    },
    {
      "entropy": 5.6373857975006105,
      "epoch": 2.5539731494920175,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043479987191997045,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21469323486089706,
      "num_tokens": 58424669.0,
      "step": 28155
    },
    {
      "entropy": 5.511224222183228,
      "epoch": 2.5544267053701017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004347770920558792,
      "loss": 4.8332,
      "mean_token_accuracy": 0.21764508932828902,
      "num_tokens": 58435348.0,
      "step": 28160
    },
    {
      "entropy": 5.608684349060058,
      "epoch": 2.5548802612481856,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043475430888755706,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21417239904403687,
      "num_tokens": 58445534.0,
      "step": 28165
    },
    {
      "entropy": 5.591418504714966,
      "epoch": 2.55533381712627,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004347315224154751,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21402875781059266,
      "num_tokens": 58456718.0,
      "step": 28170
    },
    {
      "entropy": 5.660646438598633,
      "epoch": 2.555787373004354,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004347087326401045,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2094752609729767,
      "num_tokens": 58467176.0,
      "step": 28175
    },
    {
      "entropy": 5.582002782821656,
      "epoch": 2.5562409288824384,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043468593956191663,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2182113915681839,
      "num_tokens": 58477642.0,
      "step": 28180
    },
    {
      "entropy": 5.547383594512939,
      "epoch": 2.5566944847605226,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004346631431813829,
      "loss": 4.929,
      "mean_token_accuracy": 0.2082124650478363,
      "num_tokens": 58489055.0,
      "step": 28185
    },
    {
      "entropy": 5.635054254531861,
      "epoch": 2.557148040638607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004346403434989746,
      "loss": 4.9451,
      "mean_token_accuracy": 0.20965656787157058,
      "num_tokens": 58499062.0,
      "step": 28190
    },
    {
      "entropy": 5.605187797546387,
      "epoch": 2.5576015965166907,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004346175405151633,
      "loss": 4.9504,
      "mean_token_accuracy": 0.20859958678483964,
      "num_tokens": 58509154.0,
      "step": 28195
    },
    {
      "entropy": 5.539422559738159,
      "epoch": 2.558055152394775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345947342304207,
      "loss": 4.845,
      "mean_token_accuracy": 0.21875003278255462,
      "num_tokens": 58520056.0,
      "step": 28200
    },
    {
      "entropy": 5.5600813865661625,
      "epoch": 2.5585087082728593,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043457192464521827,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21735468059778212,
      "num_tokens": 58529934.0,
      "step": 28205
    },
    {
      "entropy": 5.522853803634644,
      "epoch": 2.5589622641509435,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004345491117600278,
      "loss": 4.8363,
      "mean_token_accuracy": 0.21823289692401887,
      "num_tokens": 58540013.0,
      "step": 28210
    },
    {
      "entropy": 5.6585124969482425,
      "epoch": 2.5594158200290273,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043452629557532097,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21542141437530518,
      "num_tokens": 58550282.0,
      "step": 28215
    },
    {
      "entropy": 5.650312662124634,
      "epoch": 2.5598693759071116,
      "grad_norm": 1.125,
      "learning_rate": 0.00043450347609156974,
      "loss": 4.964,
      "mean_token_accuracy": 0.20569372177124023,
      "num_tokens": 58560490.0,
      "step": 28220
    },
    {
      "entropy": 5.5479888916015625,
      "epoch": 2.560322931785196,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043448065330924595,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21200328916311265,
      "num_tokens": 58570371.0,
      "step": 28225
    },
    {
      "entropy": 5.589256954193115,
      "epoch": 2.56077648766328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004344578272288216,
      "loss": 4.8843,
      "mean_token_accuracy": 0.21767188757658004,
      "num_tokens": 58580701.0,
      "step": 28230
    },
    {
      "entropy": 5.623743343353271,
      "epoch": 2.5612300435413644,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043443499785076885,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2170988515019417,
      "num_tokens": 58590903.0,
      "step": 28235
    },
    {
      "entropy": 5.612763690948486,
      "epoch": 2.5616835994194487,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004344121651755596,
      "loss": 4.9054,
      "mean_token_accuracy": 0.21549489349126816,
      "num_tokens": 58600924.0,
      "step": 28240
    },
    {
      "entropy": 5.620186042785645,
      "epoch": 2.5621371552975325,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043438932920366615,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2145834818482399,
      "num_tokens": 58611369.0,
      "step": 28245
    },
    {
      "entropy": 5.6274457454681395,
      "epoch": 2.5625907111756168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004343664899355608,
      "loss": 5.0013,
      "mean_token_accuracy": 0.21357562839984895,
      "num_tokens": 58622725.0,
      "step": 28250
    },
    {
      "entropy": 5.580396556854248,
      "epoch": 2.563044267053701,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043434364737171587,
      "loss": 4.9171,
      "mean_token_accuracy": 0.2151700735092163,
      "num_tokens": 58634157.0,
      "step": 28255
    },
    {
      "entropy": 5.577631187438965,
      "epoch": 2.5634978229317853,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043432080151260366,
      "loss": 4.8506,
      "mean_token_accuracy": 0.21447116136550903,
      "num_tokens": 58643384.0,
      "step": 28260
    },
    {
      "entropy": 5.577665138244629,
      "epoch": 2.563951378809869,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043429795235869667,
      "loss": 4.8958,
      "mean_token_accuracy": 0.21589313298463822,
      "num_tokens": 58653748.0,
      "step": 28265
    },
    {
      "entropy": 5.612186479568481,
      "epoch": 2.5644049346879534,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043427509991046746,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2085652694106102,
      "num_tokens": 58664619.0,
      "step": 28270
    },
    {
      "entropy": 5.558182144165039,
      "epoch": 2.5648584905660377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043425224416838857,
      "loss": 4.8411,
      "mean_token_accuracy": 0.22760617434978486,
      "num_tokens": 58674506.0,
      "step": 28275
    },
    {
      "entropy": 5.610248851776123,
      "epoch": 2.565312046444122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004342293851329327,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20432157218456268,
      "num_tokens": 58686252.0,
      "step": 28280
    },
    {
      "entropy": 5.600450038909912,
      "epoch": 2.565765602322206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043420652280457255,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2097615897655487,
      "num_tokens": 58696477.0,
      "step": 28285
    },
    {
      "entropy": 5.494926166534424,
      "epoch": 2.5662191582002905,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043418365718378096,
      "loss": 4.8055,
      "mean_token_accuracy": 0.2248273029923439,
      "num_tokens": 58706664.0,
      "step": 28290
    },
    {
      "entropy": 5.499495935440064,
      "epoch": 2.5666727140783747,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004341607882710307,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2177201196551323,
      "num_tokens": 58716950.0,
      "step": 28295
    },
    {
      "entropy": 5.560776662826538,
      "epoch": 2.5671262699564585,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004341379160667949,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21756815910339355,
      "num_tokens": 58727723.0,
      "step": 28300
    },
    {
      "entropy": 5.570564651489258,
      "epoch": 2.567579825834543,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004341150405715464,
      "loss": 5.005,
      "mean_token_accuracy": 0.2085478648543358,
      "num_tokens": 58737956.0,
      "step": 28305
    },
    {
      "entropy": 5.572387790679931,
      "epoch": 2.568033381712627,
      "grad_norm": 1.125,
      "learning_rate": 0.00043409216178575826,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2219129577279091,
      "num_tokens": 58748476.0,
      "step": 28310
    },
    {
      "entropy": 5.620368814468383,
      "epoch": 2.5684869375907113,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004340692797099037,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21920992732048034,
      "num_tokens": 58759604.0,
      "step": 28315
    },
    {
      "entropy": 5.601776790618897,
      "epoch": 2.568940493468795,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004340463943444559,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21959961205720901,
      "num_tokens": 58769781.0,
      "step": 28320
    },
    {
      "entropy": 5.58840970993042,
      "epoch": 2.5693940493468794,
      "grad_norm": 1.25,
      "learning_rate": 0.0004340235056898882,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21749221235513688,
      "num_tokens": 58779384.0,
      "step": 28325
    },
    {
      "entropy": 5.6534069061279295,
      "epoch": 2.5698476052249637,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043400061374667374,
      "loss": 5.0013,
      "mean_token_accuracy": 0.21296125203371047,
      "num_tokens": 58791953.0,
      "step": 28330
    },
    {
      "entropy": 5.545256519317627,
      "epoch": 2.570301161103048,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004339777185152862,
      "loss": 4.8945,
      "mean_token_accuracy": 0.21814947873353957,
      "num_tokens": 58802415.0,
      "step": 28335
    },
    {
      "entropy": 5.634825372695923,
      "epoch": 2.5707547169811322,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004339548199961989,
      "loss": 4.93,
      "mean_token_accuracy": 0.21505463123321533,
      "num_tokens": 58812582.0,
      "step": 28340
    },
    {
      "entropy": 5.59175853729248,
      "epoch": 2.5712082728592165,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043393191818988543,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20902827382087708,
      "num_tokens": 58822987.0,
      "step": 28345
    },
    {
      "entropy": 5.553751277923584,
      "epoch": 2.5716618287373003,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004339090130968193,
      "loss": 4.8768,
      "mean_token_accuracy": 0.21788219213485718,
      "num_tokens": 58832651.0,
      "step": 28350
    },
    {
      "entropy": 5.586088085174561,
      "epoch": 2.5721153846153846,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004338861047174743,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2090076819062233,
      "num_tokens": 58842493.0,
      "step": 28355
    },
    {
      "entropy": 5.54956841468811,
      "epoch": 2.572568940493469,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004338631930523242,
      "loss": 4.8349,
      "mean_token_accuracy": 0.22456513494253158,
      "num_tokens": 58852428.0,
      "step": 28360
    },
    {
      "entropy": 5.521766948699951,
      "epoch": 2.573022496371553,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004338402781018429,
      "loss": 4.8497,
      "mean_token_accuracy": 0.22391195446252823,
      "num_tokens": 58863234.0,
      "step": 28365
    },
    {
      "entropy": 5.546092367172241,
      "epoch": 2.573476052249637,
      "grad_norm": 1.203125,
      "learning_rate": 0.000433817359866504,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21715689301490784,
      "num_tokens": 58872632.0,
      "step": 28370
    },
    {
      "entropy": 5.672575998306274,
      "epoch": 2.573929608127721,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004337944383467817,
      "loss": 5.0997,
      "mean_token_accuracy": 0.2020232558250427,
      "num_tokens": 58882774.0,
      "step": 28375
    },
    {
      "entropy": 5.5224446773529055,
      "epoch": 2.5743831640058055,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043377151354314996,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22404280304908752,
      "num_tokens": 58892380.0,
      "step": 28380
    },
    {
      "entropy": 5.600534152984619,
      "epoch": 2.5748367198838897,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004337485854560829,
      "loss": 4.9312,
      "mean_token_accuracy": 0.20759045481681823,
      "num_tokens": 58902659.0,
      "step": 28385
    },
    {
      "entropy": 5.546986436843872,
      "epoch": 2.575290275761974,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004337256540860545,
      "loss": 4.894,
      "mean_token_accuracy": 0.21621687859296798,
      "num_tokens": 58913979.0,
      "step": 28390
    },
    {
      "entropy": 5.619999170303345,
      "epoch": 2.5757438316400583,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004337027194335393,
      "loss": 4.9747,
      "mean_token_accuracy": 0.21403544396162033,
      "num_tokens": 58924481.0,
      "step": 28395
    },
    {
      "entropy": 5.594887161254883,
      "epoch": 2.576197387518142,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043367978149901125,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2066089317202568,
      "num_tokens": 58936287.0,
      "step": 28400
    },
    {
      "entropy": 5.6483711242675785,
      "epoch": 2.5766509433962264,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043365684028294493,
      "loss": 4.9416,
      "mean_token_accuracy": 0.21255273073911668,
      "num_tokens": 58946070.0,
      "step": 28405
    },
    {
      "entropy": 5.6377613067626955,
      "epoch": 2.5771044992743106,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043363389578581483,
      "loss": 5.0141,
      "mean_token_accuracy": 0.20137166529893874,
      "num_tokens": 58956083.0,
      "step": 28410
    },
    {
      "entropy": 5.606115674972534,
      "epoch": 2.577558055152395,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004336109480080953,
      "loss": 4.9259,
      "mean_token_accuracy": 0.20825573652982712,
      "num_tokens": 58967557.0,
      "step": 28415
    },
    {
      "entropy": 5.558809423446656,
      "epoch": 2.5780116110304787,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004335879969502608,
      "loss": 4.8882,
      "mean_token_accuracy": 0.21217737793922425,
      "num_tokens": 58977509.0,
      "step": 28420
    },
    {
      "entropy": 5.49219241142273,
      "epoch": 2.578465166908563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043356504261278625,
      "loss": 4.773,
      "mean_token_accuracy": 0.23292024582624435,
      "num_tokens": 58987498.0,
      "step": 28425
    },
    {
      "entropy": 5.665225219726563,
      "epoch": 2.5789187227866472,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004335420849961462,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20125290602445603,
      "num_tokens": 58998503.0,
      "step": 28430
    },
    {
      "entropy": 5.567272472381592,
      "epoch": 2.5793722786647315,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004335191241008154,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21268775016069413,
      "num_tokens": 59007384.0,
      "step": 28435
    },
    {
      "entropy": 5.6467200756073,
      "epoch": 2.5798258345428158,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004334961599272687,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21258178502321243,
      "num_tokens": 59017739.0,
      "step": 28440
    },
    {
      "entropy": 5.622543811798096,
      "epoch": 2.5802793904209,
      "grad_norm": 1.25,
      "learning_rate": 0.0004334731924759811,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21672325879335402,
      "num_tokens": 59028439.0,
      "step": 28445
    },
    {
      "entropy": 5.567311716079712,
      "epoch": 2.5807329462989843,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004334502217474274,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2108398988842964,
      "num_tokens": 59039250.0,
      "step": 28450
    },
    {
      "entropy": 5.650187587738037,
      "epoch": 2.581186502177068,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043342724774208277,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2137565240263939,
      "num_tokens": 59049375.0,
      "step": 28455
    },
    {
      "entropy": 5.646983861923218,
      "epoch": 2.5816400580551524,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043340427046042237,
      "loss": 5.0323,
      "mean_token_accuracy": 0.2114769548177719,
      "num_tokens": 59059873.0,
      "step": 28460
    },
    {
      "entropy": 5.586476039886475,
      "epoch": 2.5820936139332367,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004333812899029212,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21517833471298217,
      "num_tokens": 59069288.0,
      "step": 28465
    },
    {
      "entropy": 5.575258493423462,
      "epoch": 2.5825471698113205,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043335830607005456,
      "loss": 4.9423,
      "mean_token_accuracy": 0.20982772707939149,
      "num_tokens": 59079339.0,
      "step": 28470
    },
    {
      "entropy": 5.597243881225586,
      "epoch": 2.5830007256894048,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004333353189622979,
      "loss": 4.9451,
      "mean_token_accuracy": 0.21003630012273788,
      "num_tokens": 59089883.0,
      "step": 28475
    },
    {
      "entropy": 5.646843814849854,
      "epoch": 2.583454281567489,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004333123285801264,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21654506474733354,
      "num_tokens": 59100532.0,
      "step": 28480
    },
    {
      "entropy": 5.603966569900512,
      "epoch": 2.5839078374455733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004332893349240156,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21421366333961486,
      "num_tokens": 59111295.0,
      "step": 28485
    },
    {
      "entropy": 5.610342264175415,
      "epoch": 2.5843613933236576,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004332663379944411,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2104869604110718,
      "num_tokens": 59120808.0,
      "step": 28490
    },
    {
      "entropy": 5.659372758865357,
      "epoch": 2.584814949201742,
      "grad_norm": 1.25,
      "learning_rate": 0.00043324333779187844,
      "loss": 5.004,
      "mean_token_accuracy": 0.21949607878923416,
      "num_tokens": 59131438.0,
      "step": 28495
    },
    {
      "entropy": 5.568179035186768,
      "epoch": 2.585268505079826,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043322033431680314,
      "loss": 4.9399,
      "mean_token_accuracy": 0.21955933421850204,
      "num_tokens": 59141701.0,
      "step": 28500
    },
    {
      "entropy": 5.536624765396118,
      "epoch": 2.58572206095791,
      "grad_norm": 1.203125,
      "learning_rate": 0.000433197327569691,
      "loss": 4.8978,
      "mean_token_accuracy": 0.21859243512153625,
      "num_tokens": 59152184.0,
      "step": 28505
    },
    {
      "entropy": 5.55860276222229,
      "epoch": 2.586175616835994,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043317431755101785,
      "loss": 4.8723,
      "mean_token_accuracy": 0.22313813120126724,
      "num_tokens": 59162177.0,
      "step": 28510
    },
    {
      "entropy": 5.66088900566101,
      "epoch": 2.5866291727140784,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004331513042612594,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2112630531191826,
      "num_tokens": 59171964.0,
      "step": 28515
    },
    {
      "entropy": 5.587431955337524,
      "epoch": 2.5870827285921627,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004331282877008918,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21516778618097304,
      "num_tokens": 59182224.0,
      "step": 28520
    },
    {
      "entropy": 5.594990587234497,
      "epoch": 2.5875362844702465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004331052678703909,
      "loss": 4.922,
      "mean_token_accuracy": 0.21240060925483703,
      "num_tokens": 59191912.0,
      "step": 28525
    },
    {
      "entropy": 5.604145765304565,
      "epoch": 2.587989840348331,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004330822447702327,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21341202557086944,
      "num_tokens": 59200984.0,
      "step": 28530
    },
    {
      "entropy": 5.6295576095581055,
      "epoch": 2.588443396226415,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043305921840089347,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21466362923383714,
      "num_tokens": 59210672.0,
      "step": 28535
    },
    {
      "entropy": 5.612875938415527,
      "epoch": 2.5888969521044993,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043303618876284935,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21743858605623245,
      "num_tokens": 59220230.0,
      "step": 28540
    },
    {
      "entropy": 5.523725605010986,
      "epoch": 2.5893505079825836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043301315585657645,
      "loss": 4.8644,
      "mean_token_accuracy": 0.21540619283914567,
      "num_tokens": 59230615.0,
      "step": 28545
    },
    {
      "entropy": 5.5843648433685305,
      "epoch": 2.589804063860668,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004329901196825513,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21518293917179107,
      "num_tokens": 59241717.0,
      "step": 28550
    },
    {
      "entropy": 5.647132205963135,
      "epoch": 2.5902576197387517,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043296708024125025,
      "loss": 4.9947,
      "mean_token_accuracy": 0.2105607733130455,
      "num_tokens": 59251791.0,
      "step": 28555
    },
    {
      "entropy": 5.555968236923218,
      "epoch": 2.590711175616836,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043294403753314965,
      "loss": 4.8182,
      "mean_token_accuracy": 0.22235488146543503,
      "num_tokens": 59262434.0,
      "step": 28560
    },
    {
      "entropy": 5.564286708831787,
      "epoch": 2.59116473149492,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004329209915587261,
      "loss": 4.9898,
      "mean_token_accuracy": 0.21071625351905823,
      "num_tokens": 59272138.0,
      "step": 28565
    },
    {
      "entropy": 5.548455333709716,
      "epoch": 2.5916182873730045,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043289794231845624,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2155385732650757,
      "num_tokens": 59283470.0,
      "step": 28570
    },
    {
      "entropy": 5.547768402099609,
      "epoch": 2.5920718432510883,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004328748898128168,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21359224021434783,
      "num_tokens": 59293670.0,
      "step": 28575
    },
    {
      "entropy": 5.607025098800659,
      "epoch": 2.5925253991291726,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004328518340422842,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21235549300909043,
      "num_tokens": 59303728.0,
      "step": 28580
    },
    {
      "entropy": 5.57889757156372,
      "epoch": 2.592978955007257,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043282877500733555,
      "loss": 4.8998,
      "mean_token_accuracy": 0.21655260771512985,
      "num_tokens": 59314271.0,
      "step": 28585
    },
    {
      "entropy": 5.5851446151733395,
      "epoch": 2.593432510885341,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004328057127084476,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21399024426937102,
      "num_tokens": 59324581.0,
      "step": 28590
    },
    {
      "entropy": 5.596630477905274,
      "epoch": 2.5938860667634254,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043278264714609733,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21554073989391326,
      "num_tokens": 59333912.0,
      "step": 28595
    },
    {
      "entropy": 5.556984853744507,
      "epoch": 2.5943396226415096,
      "grad_norm": 1.125,
      "learning_rate": 0.00043275957832076173,
      "loss": 4.8783,
      "mean_token_accuracy": 0.2158809334039688,
      "num_tokens": 59344245.0,
      "step": 28600
    },
    {
      "entropy": 5.583969497680664,
      "epoch": 2.5947931785195935,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043273650623291775,
      "loss": 5.0088,
      "mean_token_accuracy": 0.21582089960575104,
      "num_tokens": 59353825.0,
      "step": 28605
    },
    {
      "entropy": 5.58726749420166,
      "epoch": 2.5952467343976777,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004327134308830427,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21076942086219788,
      "num_tokens": 59364901.0,
      "step": 28610
    },
    {
      "entropy": 5.564519739151001,
      "epoch": 2.595700290275762,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004326903522716137,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2169706791639328,
      "num_tokens": 59374772.0,
      "step": 28615
    },
    {
      "entropy": 5.528669786453247,
      "epoch": 2.5961538461538463,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000432667270399108,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22522260695695878,
      "num_tokens": 59383640.0,
      "step": 28620
    },
    {
      "entropy": 5.594297933578491,
      "epoch": 2.59660740203193,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043264418526600305,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21794225573539733,
      "num_tokens": 59392928.0,
      "step": 28625
    },
    {
      "entropy": 5.631557893753052,
      "epoch": 2.5970609579100143,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004326210968727761,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2088549554347992,
      "num_tokens": 59402969.0,
      "step": 28630
    },
    {
      "entropy": 5.6218935489654545,
      "epoch": 2.5975145137880986,
      "grad_norm": 1.125,
      "learning_rate": 0.0004325980052199047,
      "loss": 5.003,
      "mean_token_accuracy": 0.2038627967238426,
      "num_tokens": 59413439.0,
      "step": 28635
    },
    {
      "entropy": 5.629627132415772,
      "epoch": 2.597968069666183,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043257491030786644,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21281755566596985,
      "num_tokens": 59423403.0,
      "step": 28640
    },
    {
      "entropy": 5.610312032699585,
      "epoch": 2.598421625544267,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043255181213713883,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20955325216054915,
      "num_tokens": 59434519.0,
      "step": 28645
    },
    {
      "entropy": 5.609882116317749,
      "epoch": 2.5988751814223514,
      "grad_norm": 1.125,
      "learning_rate": 0.0004325287107081996,
      "loss": 4.948,
      "mean_token_accuracy": 0.20845523625612258,
      "num_tokens": 59445056.0,
      "step": 28650
    },
    {
      "entropy": 5.574291849136353,
      "epoch": 2.5993287373004357,
      "grad_norm": 1.25,
      "learning_rate": 0.0004325056060215265,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21606272011995314,
      "num_tokens": 59455037.0,
      "step": 28655
    },
    {
      "entropy": 5.568118667602539,
      "epoch": 2.5997822931785195,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004324824980775973,
      "loss": 4.9598,
      "mean_token_accuracy": 0.21121555864810942,
      "num_tokens": 59465093.0,
      "step": 28660
    },
    {
      "entropy": 5.576521301269532,
      "epoch": 2.6002358490566038,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043245938687688995,
      "loss": 4.9992,
      "mean_token_accuracy": 0.2153582751750946,
      "num_tokens": 59475449.0,
      "step": 28665
    },
    {
      "entropy": 5.653523874282837,
      "epoch": 2.600689404934688,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043243627241988235,
      "loss": 4.9495,
      "mean_token_accuracy": 0.2126238077878952,
      "num_tokens": 59484990.0,
      "step": 28670
    },
    {
      "entropy": 5.643709564208985,
      "epoch": 2.6011429608127723,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043241315470705254,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2096727728843689,
      "num_tokens": 59495739.0,
      "step": 28675
    },
    {
      "entropy": 5.595979118347168,
      "epoch": 2.601596516690856,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043239003373887855,
      "loss": 4.933,
      "mean_token_accuracy": 0.21496637016534806,
      "num_tokens": 59506432.0,
      "step": 28680
    },
    {
      "entropy": 5.590441560745239,
      "epoch": 2.6020500725689404,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004323669095158385,
      "loss": 4.9388,
      "mean_token_accuracy": 0.211165452003479,
      "num_tokens": 59516585.0,
      "step": 28685
    },
    {
      "entropy": 5.57189245223999,
      "epoch": 2.6025036284470247,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004323437820384107,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2089263454079628,
      "num_tokens": 59526824.0,
      "step": 28690
    },
    {
      "entropy": 5.641707992553711,
      "epoch": 2.602957184325109,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004323206513070734,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21190085262060165,
      "num_tokens": 59537749.0,
      "step": 28695
    },
    {
      "entropy": 5.624490976333618,
      "epoch": 2.603410740203193,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000432297517322305,
      "loss": 4.9262,
      "mean_token_accuracy": 0.20382642596960068,
      "num_tokens": 59548310.0,
      "step": 28700
    },
    {
      "entropy": 5.636884260177612,
      "epoch": 2.6038642960812775,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004322743800845838,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2139281451702118,
      "num_tokens": 59558191.0,
      "step": 28705
    },
    {
      "entropy": 5.602367115020752,
      "epoch": 2.6043178519593613,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043225123959438836,
      "loss": 5.0101,
      "mean_token_accuracy": 0.202468204498291,
      "num_tokens": 59568961.0,
      "step": 28710
    },
    {
      "entropy": 5.602570867538452,
      "epoch": 2.6047714078374455,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043222809585219716,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21022058576345443,
      "num_tokens": 59579137.0,
      "step": 28715
    },
    {
      "entropy": 5.450237035751343,
      "epoch": 2.60522496371553,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043220494885848896,
      "loss": 4.7484,
      "mean_token_accuracy": 0.22537932246923448,
      "num_tokens": 59589612.0,
      "step": 28720
    },
    {
      "entropy": 5.6324601650238035,
      "epoch": 2.605678519593614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004321817986137423,
      "loss": 4.9659,
      "mean_token_accuracy": 0.20888511687517167,
      "num_tokens": 59600332.0,
      "step": 28725
    },
    {
      "entropy": 5.5210652351379395,
      "epoch": 2.606132075471698,
      "grad_norm": 1.234375,
      "learning_rate": 0.000432158645118436,
      "loss": 4.8573,
      "mean_token_accuracy": 0.22621295005083084,
      "num_tokens": 59610491.0,
      "step": 28730
    },
    {
      "entropy": 5.707390689849854,
      "epoch": 2.606585631349782,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004321354883730488,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20002554804086686,
      "num_tokens": 59620729.0,
      "step": 28735
    },
    {
      "entropy": 5.561369371414185,
      "epoch": 2.6070391872278664,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043211232837805973,
      "loss": 4.8887,
      "mean_token_accuracy": 0.21987232565879822,
      "num_tokens": 59631243.0,
      "step": 28740
    },
    {
      "entropy": 5.604061460494995,
      "epoch": 2.6074927431059507,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043208916513394767,
      "loss": 4.9173,
      "mean_token_accuracy": 0.21859145760536194,
      "num_tokens": 59642055.0,
      "step": 28745
    },
    {
      "entropy": 5.664089727401733,
      "epoch": 2.607946298984035,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043206599864119164,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20990859270095824,
      "num_tokens": 59652144.0,
      "step": 28750
    },
    {
      "entropy": 5.606751394271851,
      "epoch": 2.6083998548621192,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043204282890027074,
      "loss": 4.8899,
      "mean_token_accuracy": 0.22807482182979583,
      "num_tokens": 59662367.0,
      "step": 28755
    },
    {
      "entropy": 5.64818639755249,
      "epoch": 2.608853410740203,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043201965591166413,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2180311530828476,
      "num_tokens": 59672729.0,
      "step": 28760
    },
    {
      "entropy": 5.649823617935181,
      "epoch": 2.6093069666182873,
      "grad_norm": 1.171875,
      "learning_rate": 0.000431996479675851,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21764556765556337,
      "num_tokens": 59683282.0,
      "step": 28765
    },
    {
      "entropy": 5.633039426803589,
      "epoch": 2.6097605224963716,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043197330019331074,
      "loss": 5.0466,
      "mean_token_accuracy": 0.204258731007576,
      "num_tokens": 59694297.0,
      "step": 28770
    },
    {
      "entropy": 5.540487909317017,
      "epoch": 2.610214078374456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043195011746452244,
      "loss": 4.8653,
      "mean_token_accuracy": 0.217235167324543,
      "num_tokens": 59704149.0,
      "step": 28775
    },
    {
      "entropy": 5.627522087097168,
      "epoch": 2.6106676342525397,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043192693148996585,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21168753057718276,
      "num_tokens": 59714811.0,
      "step": 28780
    },
    {
      "entropy": 5.535090017318725,
      "epoch": 2.611121190130624,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004319037422701203,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2220741555094719,
      "num_tokens": 59724950.0,
      "step": 28785
    },
    {
      "entropy": 5.614667510986328,
      "epoch": 2.611574746008708,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004318805498054654,
      "loss": 5.0572,
      "mean_token_accuracy": 0.19790161401033401,
      "num_tokens": 59735905.0,
      "step": 28790
    },
    {
      "entropy": 5.678648233413696,
      "epoch": 2.6120283018867925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043185735409648077,
      "loss": 4.9893,
      "mean_token_accuracy": 0.2086726114153862,
      "num_tokens": 59746743.0,
      "step": 28795
    },
    {
      "entropy": 5.645985221862793,
      "epoch": 2.6124818577648767,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043183415514364606,
      "loss": 4.9226,
      "mean_token_accuracy": 0.20773072242736818,
      "num_tokens": 59757205.0,
      "step": 28800
    },
    {
      "entropy": 5.563832569122314,
      "epoch": 2.612935413642961,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004318109529474411,
      "loss": 4.9307,
      "mean_token_accuracy": 0.21350997686386108,
      "num_tokens": 59767691.0,
      "step": 28805
    },
    {
      "entropy": 5.546957921981812,
      "epoch": 2.6133889695210453,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004317877475083456,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21252218782901763,
      "num_tokens": 59778618.0,
      "step": 28810
    },
    {
      "entropy": 5.612049341201782,
      "epoch": 2.613842525399129,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043176453882683955,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2077659174799919,
      "num_tokens": 59789423.0,
      "step": 28815
    },
    {
      "entropy": 5.678122425079346,
      "epoch": 2.6142960812772134,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043174132690340287,
      "loss": 5.0231,
      "mean_token_accuracy": 0.20145198255777358,
      "num_tokens": 59800021.0,
      "step": 28820
    },
    {
      "entropy": 5.561448335647583,
      "epoch": 2.6147496371552976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004317181117385156,
      "loss": 4.9133,
      "mean_token_accuracy": 0.20836929380893707,
      "num_tokens": 59810006.0,
      "step": 28825
    },
    {
      "entropy": 5.6187968254089355,
      "epoch": 2.6152031930333814,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043169489333265795,
      "loss": 4.97,
      "mean_token_accuracy": 0.21044576913118362,
      "num_tokens": 59819505.0,
      "step": 28830
    },
    {
      "entropy": 5.621313524246216,
      "epoch": 2.6156567489114657,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043167167168630977,
      "loss": 4.982,
      "mean_token_accuracy": 0.21647306084632872,
      "num_tokens": 59830439.0,
      "step": 28835
    },
    {
      "entropy": 5.603928470611573,
      "epoch": 2.61611030478955,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004316484467999516,
      "loss": 4.9398,
      "mean_token_accuracy": 0.21201158463954925,
      "num_tokens": 59841281.0,
      "step": 28840
    },
    {
      "entropy": 5.606333303451538,
      "epoch": 2.6165638606676342,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043162521867406376,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20431240797042846,
      "num_tokens": 59853581.0,
      "step": 28845
    },
    {
      "entropy": 5.680047702789307,
      "epoch": 2.6170174165457185,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004316019873091263,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21038201600313186,
      "num_tokens": 59864787.0,
      "step": 28850
    },
    {
      "entropy": 5.645145893096924,
      "epoch": 2.6174709724238028,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004315787527056199,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21263731718063356,
      "num_tokens": 59875244.0,
      "step": 28855
    },
    {
      "entropy": 5.569312000274659,
      "epoch": 2.617924528301887,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004315555148640251,
      "loss": 4.9173,
      "mean_token_accuracy": 0.21303060948848723,
      "num_tokens": 59886535.0,
      "step": 28860
    },
    {
      "entropy": 5.576078128814697,
      "epoch": 2.618378084179971,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004315322737848222,
      "loss": 4.9217,
      "mean_token_accuracy": 0.21538650393486022,
      "num_tokens": 59896298.0,
      "step": 28865
    },
    {
      "entropy": 5.500714635848999,
      "epoch": 2.618831640058055,
      "grad_norm": 1.171875,
      "learning_rate": 0.000431509029468492,
      "loss": 4.9095,
      "mean_token_accuracy": 0.21334036588668823,
      "num_tokens": 59907010.0,
      "step": 28870
    },
    {
      "entropy": 5.636637210845947,
      "epoch": 2.6192851959361394,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004314857819155153,
      "loss": 4.9108,
      "mean_token_accuracy": 0.2183952048420906,
      "num_tokens": 59916247.0,
      "step": 28875
    },
    {
      "entropy": 5.537689352035523,
      "epoch": 2.6197387518142237,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043146253112637267,
      "loss": 4.8363,
      "mean_token_accuracy": 0.22253036201000215,
      "num_tokens": 59926680.0,
      "step": 28880
    },
    {
      "entropy": 5.517590570449829,
      "epoch": 2.6201923076923075,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043143927710154504,
      "loss": 4.86,
      "mean_token_accuracy": 0.22130801826715468,
      "num_tokens": 59936346.0,
      "step": 28885
    },
    {
      "entropy": 5.585208892822266,
      "epoch": 2.6206458635703918,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004314160198415133,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2234189748764038,
      "num_tokens": 59947007.0,
      "step": 28890
    },
    {
      "entropy": 5.630993127822876,
      "epoch": 2.621099419448476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043139275934675844,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21387099623680114,
      "num_tokens": 59957123.0,
      "step": 28895
    },
    {
      "entropy": 5.602101898193359,
      "epoch": 2.6215529753265603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004313694956177613,
      "loss": 4.9489,
      "mean_token_accuracy": 0.21782744973897933,
      "num_tokens": 59966566.0,
      "step": 28900
    },
    {
      "entropy": 5.610617542266846,
      "epoch": 2.6220065312046446,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043134622865500334,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21037882417440415,
      "num_tokens": 59976065.0,
      "step": 28905
    },
    {
      "entropy": 5.619112396240235,
      "epoch": 2.622460087082729,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004313229584589655,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21438599228858948,
      "num_tokens": 59986201.0,
      "step": 28910
    },
    {
      "entropy": 5.566736698150635,
      "epoch": 2.6229136429608126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000431299685030129,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21046952307224273,
      "num_tokens": 59996574.0,
      "step": 28915
    },
    {
      "entropy": 5.608832836151123,
      "epoch": 2.623367198838897,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043127640836897516,
      "loss": 4.9444,
      "mean_token_accuracy": 0.218801411986351,
      "num_tokens": 60006082.0,
      "step": 28920
    },
    {
      "entropy": 5.612108373641968,
      "epoch": 2.623820754716981,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004312531284759854,
      "loss": 4.9428,
      "mean_token_accuracy": 0.2128320351243019,
      "num_tokens": 60015483.0,
      "step": 28925
    },
    {
      "entropy": 5.585036373138427,
      "epoch": 2.6242743105950654,
      "grad_norm": 1.484375,
      "learning_rate": 0.00043122984535164116,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22925111949443816,
      "num_tokens": 60025984.0,
      "step": 28930
    },
    {
      "entropy": 5.623856115341186,
      "epoch": 2.6247278664731493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043120655899642386,
      "loss": 4.9516,
      "mean_token_accuracy": 0.20588652789592743,
      "num_tokens": 60036109.0,
      "step": 28935
    },
    {
      "entropy": 5.673203086853027,
      "epoch": 2.6251814223512335,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043118326941081513,
      "loss": 5.015,
      "mean_token_accuracy": 0.20424046218395234,
      "num_tokens": 60046546.0,
      "step": 28940
    },
    {
      "entropy": 5.6346159934997555,
      "epoch": 2.625634978229318,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004311599765952966,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21607362776994704,
      "num_tokens": 60056768.0,
      "step": 28945
    },
    {
      "entropy": 5.637686634063721,
      "epoch": 2.626088534107402,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043113668055034995,
      "loss": 5.0634,
      "mean_token_accuracy": 0.20704057812690735,
      "num_tokens": 60066962.0,
      "step": 28950
    },
    {
      "entropy": 5.6153168201446535,
      "epoch": 2.6265420899854863,
      "grad_norm": 1.25,
      "learning_rate": 0.0004311133812764569,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21427322477102279,
      "num_tokens": 60076477.0,
      "step": 28955
    },
    {
      "entropy": 5.550201272964477,
      "epoch": 2.6269956458635706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043109007877409946,
      "loss": 4.8265,
      "mean_token_accuracy": 0.22151952981948853,
      "num_tokens": 60088060.0,
      "step": 28960
    },
    {
      "entropy": 5.597414588928222,
      "epoch": 2.6274492017416544,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043106677304375935,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21379705220460893,
      "num_tokens": 60098677.0,
      "step": 28965
    },
    {
      "entropy": 5.637823247909546,
      "epoch": 2.6279027576197387,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004310434640859186,
      "loss": 5.0191,
      "mean_token_accuracy": 0.20918598175048828,
      "num_tokens": 60109054.0,
      "step": 28970
    },
    {
      "entropy": 5.587820386886596,
      "epoch": 2.628356313497823,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004310201519010592,
      "loss": 4.8596,
      "mean_token_accuracy": 0.21929230242967607,
      "num_tokens": 60119951.0,
      "step": 28975
    },
    {
      "entropy": 5.499077987670899,
      "epoch": 2.628809869375907,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043099683648966335,
      "loss": 4.8327,
      "mean_token_accuracy": 0.21656721383333205,
      "num_tokens": 60130026.0,
      "step": 28980
    },
    {
      "entropy": 5.485948848724365,
      "epoch": 2.629263425253991,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004309735178522132,
      "loss": 4.8614,
      "mean_token_accuracy": 0.21893069446086882,
      "num_tokens": 60140748.0,
      "step": 28985
    },
    {
      "entropy": 5.570332002639771,
      "epoch": 2.6297169811320753,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004309501959891909,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21589876860380172,
      "num_tokens": 60151042.0,
      "step": 28990
    },
    {
      "entropy": 5.6394469261169435,
      "epoch": 2.6301705370101596,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043092687090107883,
      "loss": 4.9273,
      "mean_token_accuracy": 0.20734856873750687,
      "num_tokens": 60160758.0,
      "step": 28995
    },
    {
      "entropy": 5.531998872756958,
      "epoch": 2.630624092888244,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043090354258835934,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21061759293079377,
      "num_tokens": 60170249.0,
      "step": 29000
    },
    {
      "entropy": 5.531155824661255,
      "epoch": 2.631077648766328,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043088021105151483,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21220093667507173,
      "num_tokens": 60181216.0,
      "step": 29005
    },
    {
      "entropy": 5.546112728118897,
      "epoch": 2.6315312046444124,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004308568762910279,
      "loss": 4.8901,
      "mean_token_accuracy": 0.21996376067399978,
      "num_tokens": 60190559.0,
      "step": 29010
    },
    {
      "entropy": 5.6920863628387455,
      "epoch": 2.6319847605224966,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043083353830738097,
      "loss": 5.031,
      "mean_token_accuracy": 0.203672294318676,
      "num_tokens": 60201268.0,
      "step": 29015
    },
    {
      "entropy": 5.5402655601501465,
      "epoch": 2.6324383164005805,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043081019710105683,
      "loss": 4.7999,
      "mean_token_accuracy": 0.2213676854968071,
      "num_tokens": 60210907.0,
      "step": 29020
    },
    {
      "entropy": 5.521044445037842,
      "epoch": 2.6328918722786647,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004307868526725381,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22310011386871337,
      "num_tokens": 60220298.0,
      "step": 29025
    },
    {
      "entropy": 5.514516162872314,
      "epoch": 2.633345428156749,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004307635050223075,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22228290885686874,
      "num_tokens": 60230811.0,
      "step": 29030
    },
    {
      "entropy": 5.612066030502319,
      "epoch": 2.6337989840348333,
      "grad_norm": 1.203125,
      "learning_rate": 0.000430740154150848,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2183879643678665,
      "num_tokens": 60241211.0,
      "step": 29035
    },
    {
      "entropy": 5.522991895675659,
      "epoch": 2.634252539912917,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004307168000586423,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22052494883537294,
      "num_tokens": 60251506.0,
      "step": 29040
    },
    {
      "entropy": 5.524680042266846,
      "epoch": 2.6347060957910013,
      "grad_norm": 1.5,
      "learning_rate": 0.0004306934427461736,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2325665459036827,
      "num_tokens": 60261416.0,
      "step": 29045
    },
    {
      "entropy": 5.5910240650177006,
      "epoch": 2.6351596516690856,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004306700822139248,
      "loss": 4.9386,
      "mean_token_accuracy": 0.211593659222126,
      "num_tokens": 60272315.0,
      "step": 29050
    },
    {
      "entropy": 5.6002648830413815,
      "epoch": 2.63561320754717,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000430646718462379,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21055780500173568,
      "num_tokens": 60282722.0,
      "step": 29055
    },
    {
      "entropy": 5.595898580551148,
      "epoch": 2.636066763425254,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004306233514920194,
      "loss": 4.9219,
      "mean_token_accuracy": 0.21408590227365493,
      "num_tokens": 60292636.0,
      "step": 29060
    },
    {
      "entropy": 5.575620317459107,
      "epoch": 2.6365203193033384,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043059998130332933,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2143368050456047,
      "num_tokens": 60302979.0,
      "step": 29065
    },
    {
      "entropy": 5.599817514419556,
      "epoch": 2.6369738751814222,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043057660789679194,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21508259177207947,
      "num_tokens": 60313181.0,
      "step": 29070
    },
    {
      "entropy": 5.642970275878906,
      "epoch": 2.6374274310595065,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004305532312728906,
      "loss": 4.8985,
      "mean_token_accuracy": 0.21591168642044067,
      "num_tokens": 60323401.0,
      "step": 29075
    },
    {
      "entropy": 5.640989685058594,
      "epoch": 2.6378809869375908,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004305298514321088,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20930237025022508,
      "num_tokens": 60333478.0,
      "step": 29080
    },
    {
      "entropy": 5.56065936088562,
      "epoch": 2.638334542815675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043050646837493014,
      "loss": 4.898,
      "mean_token_accuracy": 0.2156144365668297,
      "num_tokens": 60343980.0,
      "step": 29085
    },
    {
      "entropy": 5.585715627670288,
      "epoch": 2.638788098693759,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004304830821018381,
      "loss": 4.902,
      "mean_token_accuracy": 0.2183808535337448,
      "num_tokens": 60354262.0,
      "step": 29090
    },
    {
      "entropy": 5.652287578582763,
      "epoch": 2.639241654571843,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004304596926133162,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21569766253232955,
      "num_tokens": 60364137.0,
      "step": 29095
    },
    {
      "entropy": 5.59980354309082,
      "epoch": 2.6396952104499274,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043043629990984823,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21163684576749803,
      "num_tokens": 60374528.0,
      "step": 29100
    },
    {
      "entropy": 5.580597925186157,
      "epoch": 2.6401487663280117,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043041290399191797,
      "loss": 4.9458,
      "mean_token_accuracy": 0.2219318628311157,
      "num_tokens": 60384880.0,
      "step": 29105
    },
    {
      "entropy": 5.564107131958008,
      "epoch": 2.640602322206096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004303895048600093,
      "loss": 4.8966,
      "mean_token_accuracy": 0.20896926820278167,
      "num_tokens": 60395543.0,
      "step": 29110
    },
    {
      "entropy": 5.609448099136353,
      "epoch": 2.64105587808418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004303661025146061,
      "loss": 4.9194,
      "mean_token_accuracy": 0.22017806619405747,
      "num_tokens": 60405432.0,
      "step": 29115
    },
    {
      "entropy": 5.663624382019043,
      "epoch": 2.641509433962264,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004303426969561922,
      "loss": 5.0049,
      "mean_token_accuracy": 0.20715552419424058,
      "num_tokens": 60416360.0,
      "step": 29120
    },
    {
      "entropy": 5.592259502410888,
      "epoch": 2.6419629898403483,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004303192881852518,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21244598925113678,
      "num_tokens": 60426371.0,
      "step": 29125
    },
    {
      "entropy": 5.588502359390259,
      "epoch": 2.6424165457184325,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043029587620226897,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21446119397878646,
      "num_tokens": 60437505.0,
      "step": 29130
    },
    {
      "entropy": 5.599785852432251,
      "epoch": 2.642870101596517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004302724610077278,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21085345447063447,
      "num_tokens": 60448170.0,
      "step": 29135
    },
    {
      "entropy": 5.661326694488525,
      "epoch": 2.6433236574746006,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004302490426021126,
      "loss": 5.022,
      "mean_token_accuracy": 0.20712724924087525,
      "num_tokens": 60458776.0,
      "step": 29140
    },
    {
      "entropy": 5.607163143157959,
      "epoch": 2.643777213352685,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004302256209859076,
      "loss": 4.9127,
      "mean_token_accuracy": 0.22149889469146727,
      "num_tokens": 60468279.0,
      "step": 29145
    },
    {
      "entropy": 5.566177225112915,
      "epoch": 2.644230769230769,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043020219615959727,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20772133022546768,
      "num_tokens": 60478507.0,
      "step": 29150
    },
    {
      "entropy": 5.610563278198242,
      "epoch": 2.6446843251088534,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004301787681236659,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20073592513799668,
      "num_tokens": 60489016.0,
      "step": 29155
    },
    {
      "entropy": 5.686238098144531,
      "epoch": 2.6451378809869377,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004301553368785981,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20334473699331285,
      "num_tokens": 60500169.0,
      "step": 29160
    },
    {
      "entropy": 5.592184114456177,
      "epoch": 2.645591436865022,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004301319024248784,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2178480789065361,
      "num_tokens": 60511109.0,
      "step": 29165
    },
    {
      "entropy": 5.5679283618927,
      "epoch": 2.6460449927431062,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043010846476299145,
      "loss": 4.91,
      "mean_token_accuracy": 0.2227904751896858,
      "num_tokens": 60521735.0,
      "step": 29170
    },
    {
      "entropy": 5.6478653907775875,
      "epoch": 2.64649854862119,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004300850238934219,
      "loss": 4.9822,
      "mean_token_accuracy": 0.20595175325870513,
      "num_tokens": 60532071.0,
      "step": 29175
    },
    {
      "entropy": 5.646704149246216,
      "epoch": 2.6469521044992743,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043006157981665446,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2191327765583992,
      "num_tokens": 60541909.0,
      "step": 29180
    },
    {
      "entropy": 5.538444995880127,
      "epoch": 2.6474056603773586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043003813253317415,
      "loss": 4.8718,
      "mean_token_accuracy": 0.21934362202882768,
      "num_tokens": 60551361.0,
      "step": 29185
    },
    {
      "entropy": 5.552004623413086,
      "epoch": 2.6478592162554424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043001468204346564,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21291962563991546,
      "num_tokens": 60562112.0,
      "step": 29190
    },
    {
      "entropy": 5.623826313018799,
      "epoch": 2.6483127721335267,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042999122834801406,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20764190554618836,
      "num_tokens": 60572827.0,
      "step": 29195
    },
    {
      "entropy": 5.5257162094116214,
      "epoch": 2.648766328011611,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004299677714473044,
      "loss": 4.8801,
      "mean_token_accuracy": 0.21866969019174576,
      "num_tokens": 60582222.0,
      "step": 29200
    },
    {
      "entropy": 5.593174457550049,
      "epoch": 2.649219883889695,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004299443113418217,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2189294531941414,
      "num_tokens": 60592943.0,
      "step": 29205
    },
    {
      "entropy": 5.625114297866821,
      "epoch": 2.6496734397677795,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042992084803205124,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21454366594552993,
      "num_tokens": 60603206.0,
      "step": 29210
    },
    {
      "entropy": 5.573830223083496,
      "epoch": 2.6501269956458637,
      "grad_norm": 1.203125,
      "learning_rate": 0.000429897381518478,
      "loss": 4.8816,
      "mean_token_accuracy": 0.2208836033940315,
      "num_tokens": 60614776.0,
      "step": 29215
    },
    {
      "entropy": 5.5716935157775875,
      "epoch": 2.650580551523948,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042987391180158757,
      "loss": 4.9209,
      "mean_token_accuracy": 0.208964903652668,
      "num_tokens": 60625914.0,
      "step": 29220
    },
    {
      "entropy": 5.587363529205322,
      "epoch": 2.651034107402032,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042985043888186516,
      "loss": 4.9458,
      "mean_token_accuracy": 0.21605506390333176,
      "num_tokens": 60636934.0,
      "step": 29225
    },
    {
      "entropy": 5.6113217830657955,
      "epoch": 2.651487663280116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004298269627597962,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20899191945791246,
      "num_tokens": 60647707.0,
      "step": 29230
    },
    {
      "entropy": 5.510380363464355,
      "epoch": 2.6519412191582004,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042980348343586624,
      "loss": 4.8837,
      "mean_token_accuracy": 0.2141405686736107,
      "num_tokens": 60658691.0,
      "step": 29235
    },
    {
      "entropy": 5.601118040084839,
      "epoch": 2.6523947750362846,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042978000091056074,
      "loss": 4.9884,
      "mean_token_accuracy": 0.21102332472801208,
      "num_tokens": 60668479.0,
      "step": 29240
    },
    {
      "entropy": 5.583303737640381,
      "epoch": 2.6528483309143684,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004297565151843653,
      "loss": 4.838,
      "mean_token_accuracy": 0.2179940462112427,
      "num_tokens": 60679158.0,
      "step": 29245
    },
    {
      "entropy": 5.646920108795166,
      "epoch": 2.6533018867924527,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004297330262577659,
      "loss": 5.0701,
      "mean_token_accuracy": 0.2052389770746231,
      "num_tokens": 60689363.0,
      "step": 29250
    },
    {
      "entropy": 5.564053058624268,
      "epoch": 2.653755442670537,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004297095341312479,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22076878547668458,
      "num_tokens": 60699650.0,
      "step": 29255
    },
    {
      "entropy": 5.533982181549073,
      "epoch": 2.6542089985486212,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004296860388052974,
      "loss": 4.8973,
      "mean_token_accuracy": 0.21476273685693742,
      "num_tokens": 60709921.0,
      "step": 29260
    },
    {
      "entropy": 5.613673686981201,
      "epoch": 2.6546625544267055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042966254028040015,
      "loss": 4.883,
      "mean_token_accuracy": 0.22746307998895646,
      "num_tokens": 60720112.0,
      "step": 29265
    },
    {
      "entropy": 5.569327735900879,
      "epoch": 2.65511611030479,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042963903855704216,
      "loss": 4.854,
      "mean_token_accuracy": 0.20987479984760285,
      "num_tokens": 60730604.0,
      "step": 29270
    },
    {
      "entropy": 5.57329797744751,
      "epoch": 2.6555696661828736,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042961553363570946,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2159927800297737,
      "num_tokens": 60740896.0,
      "step": 29275
    },
    {
      "entropy": 5.626656913757325,
      "epoch": 2.656023222060958,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004295920255168881,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21682820469141006,
      "num_tokens": 60751771.0,
      "step": 29280
    },
    {
      "entropy": 5.643973875045776,
      "epoch": 2.656476777939042,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042956851420106423,
      "loss": 4.9767,
      "mean_token_accuracy": 0.21327410638332367,
      "num_tokens": 60762802.0,
      "step": 29285
    },
    {
      "entropy": 5.534553098678589,
      "epoch": 2.6569303338171264,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004295449996887242,
      "loss": 4.8612,
      "mean_token_accuracy": 0.2147950768470764,
      "num_tokens": 60774102.0,
      "step": 29290
    },
    {
      "entropy": 5.554935693740845,
      "epoch": 2.65738388969521,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004295214819803541,
      "loss": 4.9539,
      "mean_token_accuracy": 0.2114003375172615,
      "num_tokens": 60785383.0,
      "step": 29295
    },
    {
      "entropy": 5.618823003768921,
      "epoch": 2.6578374455732945,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042949796107644053,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21727352291345597,
      "num_tokens": 60795173.0,
      "step": 29300
    },
    {
      "entropy": 5.595835161209107,
      "epoch": 2.6582910014513788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042947443697746956,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2198912486433983,
      "num_tokens": 60805621.0,
      "step": 29305
    },
    {
      "entropy": 5.615415000915528,
      "epoch": 2.658744557329463,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004294509096839279,
      "loss": 5.0421,
      "mean_token_accuracy": 0.2025065928697586,
      "num_tokens": 60816799.0,
      "step": 29310
    },
    {
      "entropy": 5.509220600128174,
      "epoch": 2.6591981132075473,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042942737919630215,
      "loss": 4.8702,
      "mean_token_accuracy": 0.2169080838561058,
      "num_tokens": 60826827.0,
      "step": 29315
    },
    {
      "entropy": 5.606493282318115,
      "epoch": 2.6596516690856316,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042940384551507876,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21344392150640487,
      "num_tokens": 60836567.0,
      "step": 29320
    },
    {
      "entropy": 5.6255066871643065,
      "epoch": 2.6601052249637154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004293803086407447,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21058397144079208,
      "num_tokens": 60846723.0,
      "step": 29325
    },
    {
      "entropy": 5.656845569610596,
      "epoch": 2.6605587808417996,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004293567685737863,
      "loss": 5.0333,
      "mean_token_accuracy": 0.2048659861087799,
      "num_tokens": 60857434.0,
      "step": 29330
    },
    {
      "entropy": 5.635298252105713,
      "epoch": 2.661012336719884,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042933322531469074,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21060677617788315,
      "num_tokens": 60868780.0,
      "step": 29335
    },
    {
      "entropy": 5.660285568237304,
      "epoch": 2.661465892597968,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042930967886394456,
      "loss": 4.988,
      "mean_token_accuracy": 0.2140889808535576,
      "num_tokens": 60879037.0,
      "step": 29340
    },
    {
      "entropy": 5.5364936828613285,
      "epoch": 2.661919448476052,
      "grad_norm": 1.03125,
      "learning_rate": 0.000429286129222035,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2132459357380867,
      "num_tokens": 60889543.0,
      "step": 29345
    },
    {
      "entropy": 5.554695510864258,
      "epoch": 2.6623730043541363,
      "grad_norm": 1.046875,
      "learning_rate": 0.000429262576389449,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2105930358171463,
      "num_tokens": 60900141.0,
      "step": 29350
    },
    {
      "entropy": 5.62404580116272,
      "epoch": 2.6628265602322205,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004292390203666736,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2165334090590477,
      "num_tokens": 60910244.0,
      "step": 29355
    },
    {
      "entropy": 5.534553909301758,
      "epoch": 2.663280116110305,
      "grad_norm": 1.1875,
      "learning_rate": 0.000429215461154196,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22304699718952178,
      "num_tokens": 60919908.0,
      "step": 29360
    },
    {
      "entropy": 5.6069643020629885,
      "epoch": 2.663733671988389,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042919189875250324,
      "loss": 4.9394,
      "mean_token_accuracy": 0.21423041969537734,
      "num_tokens": 60929096.0,
      "step": 29365
    },
    {
      "entropy": 5.618274927139282,
      "epoch": 2.6641872278664733,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042916833316208284,
      "loss": 4.9283,
      "mean_token_accuracy": 0.21121857315301895,
      "num_tokens": 60939346.0,
      "step": 29370
    },
    {
      "entropy": 5.611999607086181,
      "epoch": 2.6646407837445576,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004291447643834219,
      "loss": 4.9357,
      "mean_token_accuracy": 0.20974236279726027,
      "num_tokens": 60950364.0,
      "step": 29375
    },
    {
      "entropy": 5.561800813674926,
      "epoch": 2.6650943396226414,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000429121192417008,
      "loss": 4.8786,
      "mean_token_accuracy": 0.2221387267112732,
      "num_tokens": 60960905.0,
      "step": 29380
    },
    {
      "entropy": 5.613738965988159,
      "epoch": 2.6655478955007257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042909761726332864,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21911579966545106,
      "num_tokens": 60971281.0,
      "step": 29385
    },
    {
      "entropy": 5.579023218154907,
      "epoch": 2.66600145137881,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004290740389228712,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21611983478069305,
      "num_tokens": 60982522.0,
      "step": 29390
    },
    {
      "entropy": 5.575920820236206,
      "epoch": 2.6664550072568938,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042905045739612347,
      "loss": 4.9534,
      "mean_token_accuracy": 0.21015044897794724,
      "num_tokens": 60992738.0,
      "step": 29395
    },
    {
      "entropy": 5.674988174438477,
      "epoch": 2.666908563134978,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042902687268357297,
      "loss": 5.0483,
      "mean_token_accuracy": 0.21066773682832718,
      "num_tokens": 61003665.0,
      "step": 29400
    },
    {
      "entropy": 5.723544454574585,
      "epoch": 2.6673621190130623,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042900328478570745,
      "loss": 5.0558,
      "mean_token_accuracy": 0.2096869632601738,
      "num_tokens": 61014663.0,
      "step": 29405
    },
    {
      "entropy": 5.615775728225708,
      "epoch": 2.6678156748911466,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004289796937030148,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2180102750658989,
      "num_tokens": 61024916.0,
      "step": 29410
    },
    {
      "entropy": 5.522702074050903,
      "epoch": 2.668269230769231,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042895609943598286,
      "loss": 4.8565,
      "mean_token_accuracy": 0.21833383589982985,
      "num_tokens": 61035043.0,
      "step": 29415
    },
    {
      "entropy": 5.564690828323364,
      "epoch": 2.668722786647315,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004289325019850995,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21271272301673888,
      "num_tokens": 61046153.0,
      "step": 29420
    },
    {
      "entropy": 5.614724636077881,
      "epoch": 2.6691763425253994,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004289089013508528,
      "loss": 4.9201,
      "mean_token_accuracy": 0.21095340549945832,
      "num_tokens": 61057009.0,
      "step": 29425
    },
    {
      "entropy": 5.654237365722656,
      "epoch": 2.669629898403483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004288852975337308,
      "loss": 5.0517,
      "mean_token_accuracy": 0.20024534910917283,
      "num_tokens": 61068514.0,
      "step": 29430
    },
    {
      "entropy": 5.643145895004272,
      "epoch": 2.6700834542815675,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042886169053422165,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21623586118221283,
      "num_tokens": 61078902.0,
      "step": 29435
    },
    {
      "entropy": 5.6215770721435545,
      "epoch": 2.6705370101596517,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004288380803528135,
      "loss": 4.9556,
      "mean_token_accuracy": 0.2185447558760643,
      "num_tokens": 61089216.0,
      "step": 29440
    },
    {
      "entropy": 5.548191499710083,
      "epoch": 2.670990566037736,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042881446698999465,
      "loss": 4.8563,
      "mean_token_accuracy": 0.21845725029706956,
      "num_tokens": 61098470.0,
      "step": 29445
    },
    {
      "entropy": 5.686565208435058,
      "epoch": 2.67144412191582,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004287908504462534,
      "loss": 5.065,
      "mean_token_accuracy": 0.21650877594947815,
      "num_tokens": 61108694.0,
      "step": 29450
    },
    {
      "entropy": 5.560374593734741,
      "epoch": 2.671897677793904,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042876723072207825,
      "loss": 4.9367,
      "mean_token_accuracy": 0.21677561700344086,
      "num_tokens": 61119389.0,
      "step": 29455
    },
    {
      "entropy": 5.633248567581177,
      "epoch": 2.6723512336719883,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042874360781795757,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2091132178902626,
      "num_tokens": 61130491.0,
      "step": 29460
    },
    {
      "entropy": 5.604853534698487,
      "epoch": 2.6728047895500726,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042871998173437987,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21624602973461152,
      "num_tokens": 61140404.0,
      "step": 29465
    },
    {
      "entropy": 5.654856014251709,
      "epoch": 2.673258345428157,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042869635247183373,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2105361670255661,
      "num_tokens": 61150777.0,
      "step": 29470
    },
    {
      "entropy": 5.60025053024292,
      "epoch": 2.673711901306241,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042867272003080793,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21678118258714676,
      "num_tokens": 61160685.0,
      "step": 29475
    },
    {
      "entropy": 5.623936986923217,
      "epoch": 2.674165457184325,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004286490844117911,
      "loss": 4.9528,
      "mean_token_accuracy": 0.21411225348711013,
      "num_tokens": 61170553.0,
      "step": 29480
    },
    {
      "entropy": 5.65347204208374,
      "epoch": 2.6746190130624092,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042862544561527203,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21210510432720184,
      "num_tokens": 61181396.0,
      "step": 29485
    },
    {
      "entropy": 5.601888847351074,
      "epoch": 2.6750725689404935,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004286018036417396,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21526999771595,
      "num_tokens": 61192315.0,
      "step": 29490
    },
    {
      "entropy": 5.6078180313110355,
      "epoch": 2.6755261248185778,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042857815849168277,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21740586012601854,
      "num_tokens": 61202719.0,
      "step": 29495
    },
    {
      "entropy": 5.521963262557984,
      "epoch": 2.6759796806966616,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004285545101655904,
      "loss": 4.8249,
      "mean_token_accuracy": 0.22316931635141374,
      "num_tokens": 61212703.0,
      "step": 29500
    },
    {
      "entropy": 5.482142686843872,
      "epoch": 2.676433236574746,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004285308586639517,
      "loss": 4.8072,
      "mean_token_accuracy": 0.22762789130210875,
      "num_tokens": 61222406.0,
      "step": 29505
    },
    {
      "entropy": 5.568676710128784,
      "epoch": 2.67688679245283,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042850720398725575,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2032059445977211,
      "num_tokens": 61232758.0,
      "step": 29510
    },
    {
      "entropy": 5.55197582244873,
      "epoch": 2.6773403483309144,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042848354613599164,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22180147767066954,
      "num_tokens": 61243715.0,
      "step": 29515
    },
    {
      "entropy": 5.638426828384399,
      "epoch": 2.6777939042089987,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042845988511064863,
      "loss": 4.928,
      "mean_token_accuracy": 0.21806416511535645,
      "num_tokens": 61254669.0,
      "step": 29520
    },
    {
      "entropy": 5.595401859283447,
      "epoch": 2.678247460087083,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042843622091171625,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21941980272531508,
      "num_tokens": 61264377.0,
      "step": 29525
    },
    {
      "entropy": 5.541515111923218,
      "epoch": 2.6787010159651667,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004284125535396836,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21571475118398667,
      "num_tokens": 61275348.0,
      "step": 29530
    },
    {
      "entropy": 5.514275741577149,
      "epoch": 2.679154571843251,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042838888299504024,
      "loss": 4.8108,
      "mean_token_accuracy": 0.2235444962978363,
      "num_tokens": 61285579.0,
      "step": 29535
    },
    {
      "entropy": 5.548592805862427,
      "epoch": 2.6796081277213353,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042836520927827575,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21949210315942763,
      "num_tokens": 61296037.0,
      "step": 29540
    },
    {
      "entropy": 5.594270086288452,
      "epoch": 2.6800616835994195,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004283415323898797,
      "loss": 4.8702,
      "mean_token_accuracy": 0.22319923490285873,
      "num_tokens": 61305882.0,
      "step": 29545
    },
    {
      "entropy": 5.553998422622681,
      "epoch": 2.6805152394775034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042831785233034156,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22105389535427095,
      "num_tokens": 61317117.0,
      "step": 29550
    },
    {
      "entropy": 5.600816679000855,
      "epoch": 2.6809687953555876,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042829416910015116,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21059881448745726,
      "num_tokens": 61328190.0,
      "step": 29555
    },
    {
      "entropy": 5.572664546966553,
      "epoch": 2.681422351233672,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042827048269979834,
      "loss": 4.9347,
      "mean_token_accuracy": 0.2145954504609108,
      "num_tokens": 61337978.0,
      "step": 29560
    },
    {
      "entropy": 5.644948196411133,
      "epoch": 2.681875907111756,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004282467931297729,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2076480895280838,
      "num_tokens": 61348523.0,
      "step": 29565
    },
    {
      "entropy": 5.561984539031982,
      "epoch": 2.6823294629898404,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004282231003905647,
      "loss": 4.8244,
      "mean_token_accuracy": 0.22791755348443984,
      "num_tokens": 61359250.0,
      "step": 29570
    },
    {
      "entropy": 5.524264812469482,
      "epoch": 2.6827830188679247,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042819940448266376,
      "loss": 4.8421,
      "mean_token_accuracy": 0.21869919896125795,
      "num_tokens": 61369999.0,
      "step": 29575
    },
    {
      "entropy": 5.642303609848023,
      "epoch": 2.683236574746009,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042817570540656003,
      "loss": 4.9802,
      "mean_token_accuracy": 0.20877804309129716,
      "num_tokens": 61380279.0,
      "step": 29580
    },
    {
      "entropy": 5.564070177078247,
      "epoch": 2.683690130624093,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004281520031627438,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22326969504356384,
      "num_tokens": 61389794.0,
      "step": 29585
    },
    {
      "entropy": 5.6434056758880615,
      "epoch": 2.684143686502177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042812829775170497,
      "loss": 5.0361,
      "mean_token_accuracy": 0.21006974577903748,
      "num_tokens": 61401136.0,
      "step": 29590
    },
    {
      "entropy": 5.588241958618164,
      "epoch": 2.6845972423802613,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042810458917393404,
      "loss": 4.9787,
      "mean_token_accuracy": 0.215427365899086,
      "num_tokens": 61412830.0,
      "step": 29595
    },
    {
      "entropy": 5.6094013214111325,
      "epoch": 2.6850507982583456,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004280808774299211,
      "loss": 4.8711,
      "mean_token_accuracy": 0.2218219369649887,
      "num_tokens": 61422168.0,
      "step": 29600
    },
    {
      "entropy": 5.541056680679321,
      "epoch": 2.6855043541364294,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042805716252015665,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2150896817445755,
      "num_tokens": 61432471.0,
      "step": 29605
    },
    {
      "entropy": 5.564865398406982,
      "epoch": 2.6859579100145137,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042803344444513106,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21957983821630478,
      "num_tokens": 61443402.0,
      "step": 29610
    },
    {
      "entropy": 5.622091484069824,
      "epoch": 2.686411465892598,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004280097232053349,
      "loss": 4.915,
      "mean_token_accuracy": 0.2122504696249962,
      "num_tokens": 61454465.0,
      "step": 29615
    },
    {
      "entropy": 5.604333591461182,
      "epoch": 2.686865021770682,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004279859988012586,
      "loss": 4.9528,
      "mean_token_accuracy": 0.21692622750997542,
      "num_tokens": 61464671.0,
      "step": 29620
    },
    {
      "entropy": 5.5615818977355955,
      "epoch": 2.6873185776487665,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004279622712333929,
      "loss": 4.9085,
      "mean_token_accuracy": 0.22105324566364287,
      "num_tokens": 61474940.0,
      "step": 29625
    },
    {
      "entropy": 5.605070924758911,
      "epoch": 2.6877721335268507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004279385405022284,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21035200208425522,
      "num_tokens": 61485240.0,
      "step": 29630
    },
    {
      "entropy": 5.56203556060791,
      "epoch": 2.6882256894049346,
      "grad_norm": 1.140625,
      "learning_rate": 0.000427914806608256,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22004778534173966,
      "num_tokens": 61495382.0,
      "step": 29635
    },
    {
      "entropy": 5.551201391220093,
      "epoch": 2.688679245283019,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004278910695519663,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22094402760267257,
      "num_tokens": 61505372.0,
      "step": 29640
    },
    {
      "entropy": 5.597899532318115,
      "epoch": 2.689132801161103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042786732933385046,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21836409866809844,
      "num_tokens": 61516039.0,
      "step": 29645
    },
    {
      "entropy": 5.593144273757934,
      "epoch": 2.6895863570391874,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004278435859543992,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21281750351190568,
      "num_tokens": 61525435.0,
      "step": 29650
    },
    {
      "entropy": 5.640108060836792,
      "epoch": 2.690039912917271,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004278198394141037,
      "loss": 4.9286,
      "mean_token_accuracy": 0.20959557741880416,
      "num_tokens": 61534714.0,
      "step": 29655
    },
    {
      "entropy": 5.679019641876221,
      "epoch": 2.6904934687953554,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042779608971345496,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20783402323722838,
      "num_tokens": 61545841.0,
      "step": 29660
    },
    {
      "entropy": 5.5687628269195555,
      "epoch": 2.6909470246734397,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042777233685294407,
      "loss": 4.8382,
      "mean_token_accuracy": 0.227628718316555,
      "num_tokens": 61555355.0,
      "step": 29665
    },
    {
      "entropy": 5.580555009841919,
      "epoch": 2.691400580551524,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042774858083306246,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21662163287401198,
      "num_tokens": 61565474.0,
      "step": 29670
    },
    {
      "entropy": 5.609775495529175,
      "epoch": 2.6918541364296082,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042772482165430117,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2260458156466484,
      "num_tokens": 61575631.0,
      "step": 29675
    },
    {
      "entropy": 5.568985223770142,
      "epoch": 2.6923076923076925,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042770105931715167,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21264175325632095,
      "num_tokens": 61586398.0,
      "step": 29680
    },
    {
      "entropy": 5.581919479370117,
      "epoch": 2.6927612481857763,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004276772938221054,
      "loss": 4.8396,
      "mean_token_accuracy": 0.22089870870113373,
      "num_tokens": 61596630.0,
      "step": 29685
    },
    {
      "entropy": 5.603707885742187,
      "epoch": 2.6932148040638606,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004276535251696537,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21582499295473098,
      "num_tokens": 61605965.0,
      "step": 29690
    },
    {
      "entropy": 5.613928699493409,
      "epoch": 2.693668359941945,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004276297533602883,
      "loss": 4.993,
      "mean_token_accuracy": 0.21594161093235015,
      "num_tokens": 61616973.0,
      "step": 29695
    },
    {
      "entropy": 5.640426158905029,
      "epoch": 2.694121915820029,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042760597839450057,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20988346338272096,
      "num_tokens": 61627420.0,
      "step": 29700
    },
    {
      "entropy": 5.592198324203491,
      "epoch": 2.694575471698113,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004275822002727824,
      "loss": 4.9199,
      "mean_token_accuracy": 0.21896839737892151,
      "num_tokens": 61638315.0,
      "step": 29705
    },
    {
      "entropy": 5.547072124481201,
      "epoch": 2.6950290275761972,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042755841899562544,
      "loss": 4.9247,
      "mean_token_accuracy": 0.21824809312820434,
      "num_tokens": 61649294.0,
      "step": 29710
    },
    {
      "entropy": 5.625209617614746,
      "epoch": 2.6954825834542815,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042753463456352155,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2093016967177391,
      "num_tokens": 61660312.0,
      "step": 29715
    },
    {
      "entropy": 5.577877998352051,
      "epoch": 2.6959361393323658,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004275108469769625,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21825241595506667,
      "num_tokens": 61670539.0,
      "step": 29720
    },
    {
      "entropy": 5.631443214416504,
      "epoch": 2.69638969521045,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042748705623644025,
      "loss": 4.8748,
      "mean_token_accuracy": 0.2202342763543129,
      "num_tokens": 61679613.0,
      "step": 29725
    },
    {
      "entropy": 5.662536716461181,
      "epoch": 2.6968432510885343,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004274632623424468,
      "loss": 4.9447,
      "mean_token_accuracy": 0.21422010213136672,
      "num_tokens": 61689714.0,
      "step": 29730
    },
    {
      "entropy": 5.665488910675049,
      "epoch": 2.6972968069666186,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042743946529547435,
      "loss": 5.0085,
      "mean_token_accuracy": 0.21243639588356017,
      "num_tokens": 61700701.0,
      "step": 29735
    },
    {
      "entropy": 5.631139135360717,
      "epoch": 2.6977503628447024,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004274156650960148,
      "loss": 4.9756,
      "mean_token_accuracy": 0.21852733194828033,
      "num_tokens": 61711403.0,
      "step": 29740
    },
    {
      "entropy": 5.466707038879394,
      "epoch": 2.6982039187227866,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004273918617445605,
      "loss": 4.7851,
      "mean_token_accuracy": 0.2232792854309082,
      "num_tokens": 61721787.0,
      "step": 29745
    },
    {
      "entropy": 5.652282047271728,
      "epoch": 2.698657474600871,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042736805524160363,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20965745449066162,
      "num_tokens": 61732702.0,
      "step": 29750
    },
    {
      "entropy": 5.577242136001587,
      "epoch": 2.6991110304789547,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042734424558763656,
      "loss": 4.9309,
      "mean_token_accuracy": 0.2205286592245102,
      "num_tokens": 61743477.0,
      "step": 29755
    },
    {
      "entropy": 5.646869659423828,
      "epoch": 2.699564586357039,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004273204327831516,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20892202407121657,
      "num_tokens": 61753620.0,
      "step": 29760
    },
    {
      "entropy": 5.637120628356934,
      "epoch": 2.7000181422351233,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004272966168286414,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22123782932758332,
      "num_tokens": 61763678.0,
      "step": 29765
    },
    {
      "entropy": 5.6065515041351315,
      "epoch": 2.7004716981132075,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042727279772459837,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22268361002206802,
      "num_tokens": 61774419.0,
      "step": 29770
    },
    {
      "entropy": 5.638718748092652,
      "epoch": 2.700925253991292,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000427248975471515,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21097040474414824,
      "num_tokens": 61785546.0,
      "step": 29775
    },
    {
      "entropy": 5.622020864486695,
      "epoch": 2.701378809869376,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042722515006988397,
      "loss": 4.8901,
      "mean_token_accuracy": 0.21545098721981049,
      "num_tokens": 61795846.0,
      "step": 29780
    },
    {
      "entropy": 5.592202425003052,
      "epoch": 2.7018323657474603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004272013215201982,
      "loss": 4.964,
      "mean_token_accuracy": 0.21018354147672652,
      "num_tokens": 61806440.0,
      "step": 29785
    },
    {
      "entropy": 5.56593656539917,
      "epoch": 2.702285921625544,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042717748982295015,
      "loss": 4.9159,
      "mean_token_accuracy": 0.21833684891462327,
      "num_tokens": 61815371.0,
      "step": 29790
    },
    {
      "entropy": 5.647078561782837,
      "epoch": 2.7027394775036284,
      "grad_norm": 1.125,
      "learning_rate": 0.0004271536549786329,
      "loss": 4.8932,
      "mean_token_accuracy": 0.21756405532360076,
      "num_tokens": 61825225.0,
      "step": 29795
    },
    {
      "entropy": 5.572627830505371,
      "epoch": 2.7031930333817127,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004271298169877393,
      "loss": 4.9573,
      "mean_token_accuracy": 0.2187994599342346,
      "num_tokens": 61836705.0,
      "step": 29800
    },
    {
      "entropy": 5.564535331726074,
      "epoch": 2.703646589259797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004271059758507624,
      "loss": 4.902,
      "mean_token_accuracy": 0.22134292870759964,
      "num_tokens": 61847082.0,
      "step": 29805
    },
    {
      "entropy": 5.616335535049439,
      "epoch": 2.7041001451378808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042708213156819505,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21321527361869813,
      "num_tokens": 61857130.0,
      "step": 29810
    },
    {
      "entropy": 5.670545768737793,
      "epoch": 2.704553701015965,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004270582841405305,
      "loss": 5.019,
      "mean_token_accuracy": 0.20896028429269792,
      "num_tokens": 61866796.0,
      "step": 29815
    },
    {
      "entropy": 5.6275512218475345,
      "epoch": 2.7050072568940493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004270344335682618,
      "loss": 4.9693,
      "mean_token_accuracy": 0.20980193912982942,
      "num_tokens": 61876356.0,
      "step": 29820
    },
    {
      "entropy": 5.583295822143555,
      "epoch": 2.7054608127721336,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042701057985188235,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2077516570687294,
      "num_tokens": 61887374.0,
      "step": 29825
    },
    {
      "entropy": 5.654992198944091,
      "epoch": 2.705914368650218,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004269867229918854,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2114068567752838,
      "num_tokens": 61897376.0,
      "step": 29830
    },
    {
      "entropy": 5.663776206970215,
      "epoch": 2.706367924528302,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042696286298876424,
      "loss": 5.0167,
      "mean_token_accuracy": 0.2103147938847542,
      "num_tokens": 61908212.0,
      "step": 29835
    },
    {
      "entropy": 5.620198488235474,
      "epoch": 2.706821480406386,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004269389998430124,
      "loss": 5.0451,
      "mean_token_accuracy": 0.19946016520261764,
      "num_tokens": 61920699.0,
      "step": 29840
    },
    {
      "entropy": 5.582410192489624,
      "epoch": 2.70727503628447,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004269151335551233,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2113411009311676,
      "num_tokens": 61931556.0,
      "step": 29845
    },
    {
      "entropy": 5.7171485900878904,
      "epoch": 2.7077285921625545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004268912641255905,
      "loss": 5.0749,
      "mean_token_accuracy": 0.21553849875926973,
      "num_tokens": 61941483.0,
      "step": 29850
    },
    {
      "entropy": 5.64946403503418,
      "epoch": 2.7081821480406387,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042686739155490763,
      "loss": 4.9136,
      "mean_token_accuracy": 0.21388835608959197,
      "num_tokens": 61951262.0,
      "step": 29855
    },
    {
      "entropy": 5.658086347579956,
      "epoch": 2.7086357039187225,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004268435158435684,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2089886859059334,
      "num_tokens": 61962423.0,
      "step": 29860
    },
    {
      "entropy": 5.6905848503112795,
      "epoch": 2.709089259796807,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042681963699206667,
      "loss": 5.0591,
      "mean_token_accuracy": 0.20625297278165816,
      "num_tokens": 61973276.0,
      "step": 29865
    },
    {
      "entropy": 5.573126363754272,
      "epoch": 2.709542815674891,
      "grad_norm": 1.125,
      "learning_rate": 0.00042679575500089605,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22009770423173905,
      "num_tokens": 61984369.0,
      "step": 29870
    },
    {
      "entropy": 5.561296510696411,
      "epoch": 2.7099963715529753,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004267718698705506,
      "loss": 4.8842,
      "mean_token_accuracy": 0.21264982372522354,
      "num_tokens": 61994224.0,
      "step": 29875
    },
    {
      "entropy": 5.580608129501343,
      "epoch": 2.7104499274310596,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004267479816015242,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21732757687568666,
      "num_tokens": 62004546.0,
      "step": 29880
    },
    {
      "entropy": 5.640032386779785,
      "epoch": 2.710903483309144,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004267240901943109,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21724576652050018,
      "num_tokens": 62013924.0,
      "step": 29885
    },
    {
      "entropy": 5.58840913772583,
      "epoch": 2.7113570391872277,
      "grad_norm": 1.125,
      "learning_rate": 0.00042670019564940474,
      "loss": 4.927,
      "mean_token_accuracy": 0.21217675507068634,
      "num_tokens": 62026150.0,
      "step": 29890
    },
    {
      "entropy": 5.638037395477295,
      "epoch": 2.711810595065312,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004266762979672998,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21751690208911895,
      "num_tokens": 62036203.0,
      "step": 29895
    },
    {
      "entropy": 5.639289617538452,
      "epoch": 2.7122641509433962,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042665239714849044,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21079617738723755,
      "num_tokens": 62046961.0,
      "step": 29900
    },
    {
      "entropy": 5.556210947036743,
      "epoch": 2.7127177068214805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042662849319347073,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2123594179749489,
      "num_tokens": 62057365.0,
      "step": 29905
    },
    {
      "entropy": 5.5851380825042725,
      "epoch": 2.7131712626995643,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042660458610273527,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21034496277570724,
      "num_tokens": 62067934.0,
      "step": 29910
    },
    {
      "entropy": 5.669028568267822,
      "epoch": 2.7136248185776486,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004265806758767783,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2056246057152748,
      "num_tokens": 62077203.0,
      "step": 29915
    },
    {
      "entropy": 5.628414964675903,
      "epoch": 2.714078374455733,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042655676251609435,
      "loss": 4.9303,
      "mean_token_accuracy": 0.21692422777414322,
      "num_tokens": 62087262.0,
      "step": 29920
    },
    {
      "entropy": 5.580858945846558,
      "epoch": 2.714531930333817,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042653284602117795,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21459153443574905,
      "num_tokens": 62097931.0,
      "step": 29925
    },
    {
      "entropy": 5.594641256332397,
      "epoch": 2.7149854862119014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004265089263925236,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2076190382242203,
      "num_tokens": 62109478.0,
      "step": 29930
    },
    {
      "entropy": 5.676377820968628,
      "epoch": 2.7154390420899857,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000426485003630626,
      "loss": 5.024,
      "mean_token_accuracy": 0.2074105679988861,
      "num_tokens": 62120735.0,
      "step": 29935
    },
    {
      "entropy": 5.643840599060058,
      "epoch": 2.71589259796807,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004264610777359801,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2116520255804062,
      "num_tokens": 62130950.0,
      "step": 29940
    },
    {
      "entropy": 5.649484014511108,
      "epoch": 2.7163461538461537,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004264371487090804,
      "loss": 4.9731,
      "mean_token_accuracy": 0.20947919487953187,
      "num_tokens": 62141351.0,
      "step": 29945
    },
    {
      "entropy": 5.542219448089599,
      "epoch": 2.716799709724238,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042641321655042185,
      "loss": 4.8886,
      "mean_token_accuracy": 0.21609922051429747,
      "num_tokens": 62151388.0,
      "step": 29950
    },
    {
      "entropy": 5.669764995574951,
      "epoch": 2.7172532656023223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004263892812604995,
      "loss": 5.087,
      "mean_token_accuracy": 0.20519790947437286,
      "num_tokens": 62162809.0,
      "step": 29955
    },
    {
      "entropy": 5.601447343826294,
      "epoch": 2.7177068214804065,
      "grad_norm": 1.25,
      "learning_rate": 0.00042636534283980814,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21473787277936934,
      "num_tokens": 62172738.0,
      "step": 29960
    },
    {
      "entropy": 5.680136823654175,
      "epoch": 2.7181603773584904,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000426341401288843,
      "loss": 5.0706,
      "mean_token_accuracy": 0.2054516166448593,
      "num_tokens": 62182508.0,
      "step": 29965
    },
    {
      "entropy": 5.657821750640869,
      "epoch": 2.7186139332365746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042631745660809905,
      "loss": 5.0386,
      "mean_token_accuracy": 0.20377095639705659,
      "num_tokens": 62192873.0,
      "step": 29970
    },
    {
      "entropy": 5.680954551696777,
      "epoch": 2.719067489114659,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042629350879807155,
      "loss": 5.053,
      "mean_token_accuracy": 0.20653095990419387,
      "num_tokens": 62204727.0,
      "step": 29975
    },
    {
      "entropy": 5.627367162704468,
      "epoch": 2.719521044992743,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042626955785925576,
      "loss": 4.9416,
      "mean_token_accuracy": 0.20963971614837645,
      "num_tokens": 62214268.0,
      "step": 29980
    },
    {
      "entropy": 5.655849647521973,
      "epoch": 2.7199746008708274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042624560379214693,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21071461886167525,
      "num_tokens": 62224756.0,
      "step": 29985
    },
    {
      "entropy": 5.659814548492432,
      "epoch": 2.7204281567489117,
      "grad_norm": 1.25,
      "learning_rate": 0.0004262216465972405,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21357038319110871,
      "num_tokens": 62234065.0,
      "step": 29990
    },
    {
      "entropy": 5.636747360229492,
      "epoch": 2.7208817126269955,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004261976862750318,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21384130716323851,
      "num_tokens": 62243732.0,
      "step": 29995
    },
    {
      "entropy": 5.640603733062744,
      "epoch": 2.72133526850508,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004261737228260165,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2205175757408142,
      "num_tokens": 62253461.0,
      "step": 30000
    },
    {
      "epoch": 2.72133526850508,
      "eval_entropy": 5.327902805967029,
      "eval_loss": 4.993574619293213,
      "eval_mean_token_accuracy": 0.22100695479678084,
      "eval_num_tokens": 62253461.0,
      "eval_runtime": 20.622,
      "eval_samples_per_second": 1714.672,
      "eval_steps_per_second": 214.334,
      "step": 30000
    },
    {
      "entropy": 5.632916402816773,
      "epoch": 2.721788824383164,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004261497562506901,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21325458437204362,
      "num_tokens": 62263830.0,
      "step": 30005
    },
    {
      "entropy": 5.546909332275391,
      "epoch": 2.7222423802612483,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004261257865495483,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2191924974322319,
      "num_tokens": 62274383.0,
      "step": 30010
    },
    {
      "entropy": 5.668188095092773,
      "epoch": 2.722695936139332,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004261018137230865,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19490485787391662,
      "num_tokens": 62286476.0,
      "step": 30015
    },
    {
      "entropy": 5.721460390090942,
      "epoch": 2.7231494920174164,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042607783777180085,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21309360712766648,
      "num_tokens": 62296192.0,
      "step": 30020
    },
    {
      "entropy": 5.606363344192505,
      "epoch": 2.7236030478955007,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042605385869618693,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22656063735485077,
      "num_tokens": 62305300.0,
      "step": 30025
    },
    {
      "entropy": 5.610333442687988,
      "epoch": 2.724056603773585,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004260298764967408,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21354444324970245,
      "num_tokens": 62316802.0,
      "step": 30030
    },
    {
      "entropy": 5.508436489105224,
      "epoch": 2.724510159651669,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004260058911739583,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2136654421687126,
      "num_tokens": 62326444.0,
      "step": 30035
    },
    {
      "entropy": 5.587986135482788,
      "epoch": 2.7249637155297535,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042598190272833547,
      "loss": 4.8656,
      "mean_token_accuracy": 0.22007625699043273,
      "num_tokens": 62336872.0,
      "step": 30040
    },
    {
      "entropy": 5.524597644805908,
      "epoch": 2.7254172714078373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004259579111603685,
      "loss": 4.8361,
      "mean_token_accuracy": 0.21752707809209823,
      "num_tokens": 62346771.0,
      "step": 30045
    },
    {
      "entropy": 5.607333517074585,
      "epoch": 2.7258708272859216,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042593391647055334,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21466097235679626,
      "num_tokens": 62357503.0,
      "step": 30050
    },
    {
      "entropy": 5.54860200881958,
      "epoch": 2.726324383164006,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004259099186593864,
      "loss": 4.8421,
      "mean_token_accuracy": 0.2213379293680191,
      "num_tokens": 62367664.0,
      "step": 30055
    },
    {
      "entropy": 5.6246910572052,
      "epoch": 2.72677793904209,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004258859177273639,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2124890550971031,
      "num_tokens": 62376924.0,
      "step": 30060
    },
    {
      "entropy": 5.547257900238037,
      "epoch": 2.727231494920174,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004258619136749821,
      "loss": 4.8429,
      "mean_token_accuracy": 0.219496650993824,
      "num_tokens": 62387529.0,
      "step": 30065
    },
    {
      "entropy": 5.575240707397461,
      "epoch": 2.727685050798258,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004258379065027375,
      "loss": 4.9038,
      "mean_token_accuracy": 0.21730420738458633,
      "num_tokens": 62397833.0,
      "step": 30070
    },
    {
      "entropy": 5.673918008804321,
      "epoch": 2.7281386066763424,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004258138962111266,
      "loss": 4.9855,
      "mean_token_accuracy": 0.21244439482688904,
      "num_tokens": 62409022.0,
      "step": 30075
    },
    {
      "entropy": 5.627362155914307,
      "epoch": 2.7285921625544267,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004257898828006459,
      "loss": 4.9441,
      "mean_token_accuracy": 0.20647481828927994,
      "num_tokens": 62419105.0,
      "step": 30080
    },
    {
      "entropy": 5.642910432815552,
      "epoch": 2.729045718432511,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004257658662717919,
      "loss": 4.9366,
      "mean_token_accuracy": 0.2226934775710106,
      "num_tokens": 62428371.0,
      "step": 30085
    },
    {
      "entropy": 5.61530704498291,
      "epoch": 2.7294992743105952,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004257418466250615,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21682464480400085,
      "num_tokens": 62439301.0,
      "step": 30090
    },
    {
      "entropy": 5.61889533996582,
      "epoch": 2.7299528301886795,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042571782386095124,
      "loss": 4.9006,
      "mean_token_accuracy": 0.21412883698940277,
      "num_tokens": 62449450.0,
      "step": 30095
    },
    {
      "entropy": 5.619385862350464,
      "epoch": 2.7304063860667633,
      "grad_norm": 1.53125,
      "learning_rate": 0.00042569379797995793,
      "loss": 4.927,
      "mean_token_accuracy": 0.2092105895280838,
      "num_tokens": 62461051.0,
      "step": 30100
    },
    {
      "entropy": 5.5840870380401615,
      "epoch": 2.7308599419448476,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004256697689825785,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22211892753839493,
      "num_tokens": 62471755.0,
      "step": 30105
    },
    {
      "entropy": 5.558784008026123,
      "epoch": 2.731313497822932,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042564573686930987,
      "loss": 4.9025,
      "mean_token_accuracy": 0.21349219232797623,
      "num_tokens": 62482060.0,
      "step": 30110
    },
    {
      "entropy": 5.5939428329467775,
      "epoch": 2.7317670537010157,
      "grad_norm": 1.125,
      "learning_rate": 0.000425621701640649,
      "loss": 4.9578,
      "mean_token_accuracy": 0.21146677285432816,
      "num_tokens": 62493735.0,
      "step": 30115
    },
    {
      "entropy": 5.640731143951416,
      "epoch": 2.7322206095791,
      "grad_norm": 1.109375,
      "learning_rate": 0.000425597663297093,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21488242149353026,
      "num_tokens": 62504582.0,
      "step": 30120
    },
    {
      "entropy": 5.536716079711914,
      "epoch": 2.7326741654571842,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004255736218391389,
      "loss": 4.9014,
      "mean_token_accuracy": 0.20783267319202423,
      "num_tokens": 62516665.0,
      "step": 30125
    },
    {
      "entropy": 5.618150234222412,
      "epoch": 2.7331277213352685,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042554957726728394,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2149476155638695,
      "num_tokens": 62527745.0,
      "step": 30130
    },
    {
      "entropy": 5.617371129989624,
      "epoch": 2.7335812772133528,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004255255295820253,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21916296631097792,
      "num_tokens": 62538622.0,
      "step": 30135
    },
    {
      "entropy": 5.591380167007446,
      "epoch": 2.734034833091437,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042550147878386044,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21482049226760863,
      "num_tokens": 62548302.0,
      "step": 30140
    },
    {
      "entropy": 5.553401803970337,
      "epoch": 2.7344883889695213,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004254774248732867,
      "loss": 4.857,
      "mean_token_accuracy": 0.22325119376182556,
      "num_tokens": 62557150.0,
      "step": 30145
    },
    {
      "entropy": 5.7374327182769775,
      "epoch": 2.734941944847605,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042545336785080126,
      "loss": 5.1368,
      "mean_token_accuracy": 0.20294255018234253,
      "num_tokens": 62568291.0,
      "step": 30150
    },
    {
      "entropy": 5.6338053226470945,
      "epoch": 2.7353955007256894,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000425429307716902,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21749451905488967,
      "num_tokens": 62579351.0,
      "step": 30155
    },
    {
      "entropy": 5.516463851928711,
      "epoch": 2.7358490566037736,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042540524447208625,
      "loss": 4.8085,
      "mean_token_accuracy": 0.22037747055292128,
      "num_tokens": 62588614.0,
      "step": 30160
    },
    {
      "entropy": 5.62703948020935,
      "epoch": 2.736302612481858,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042538117811685174,
      "loss": 4.9388,
      "mean_token_accuracy": 0.21654695868492127,
      "num_tokens": 62598844.0,
      "step": 30165
    },
    {
      "entropy": 5.63707332611084,
      "epoch": 2.7367561683599417,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042535710865169616,
      "loss": 4.9567,
      "mean_token_accuracy": 0.20858405381441117,
      "num_tokens": 62608722.0,
      "step": 30170
    },
    {
      "entropy": 5.548538684844971,
      "epoch": 2.737209724238026,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004253330360771173,
      "loss": 4.8756,
      "mean_token_accuracy": 0.21380723863840104,
      "num_tokens": 62618712.0,
      "step": 30175
    },
    {
      "entropy": 5.604543399810791,
      "epoch": 2.7376632801161103,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042530896039361284,
      "loss": 5.0565,
      "mean_token_accuracy": 0.20018149465322493,
      "num_tokens": 62629329.0,
      "step": 30180
    },
    {
      "entropy": 5.6446315288543705,
      "epoch": 2.7381168359941945,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042528488160168077,
      "loss": 4.986,
      "mean_token_accuracy": 0.20789050310850143,
      "num_tokens": 62639644.0,
      "step": 30185
    },
    {
      "entropy": 5.659021329879761,
      "epoch": 2.738570391872279,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042526079970181915,
      "loss": 5.008,
      "mean_token_accuracy": 0.20486198961734772,
      "num_tokens": 62650979.0,
      "step": 30190
    },
    {
      "entropy": 5.676254081726074,
      "epoch": 2.739023947750363,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004252367146945259,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20933170914649962,
      "num_tokens": 62661163.0,
      "step": 30195
    },
    {
      "entropy": 5.6342103481292725,
      "epoch": 2.739477503628447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042521262658029905,
      "loss": 5.0285,
      "mean_token_accuracy": 0.20412251949310303,
      "num_tokens": 62672371.0,
      "step": 30200
    },
    {
      "entropy": 5.6026030540466305,
      "epoch": 2.739931059506531,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042518853535963673,
      "loss": 4.8968,
      "mean_token_accuracy": 0.21222202777862548,
      "num_tokens": 62682521.0,
      "step": 30205
    },
    {
      "entropy": 5.612781286239624,
      "epoch": 2.7403846153846154,
      "grad_norm": 1.484375,
      "learning_rate": 0.00042516444103303736,
      "loss": 4.8271,
      "mean_token_accuracy": 0.23078083395957946,
      "num_tokens": 62692420.0,
      "step": 30210
    },
    {
      "entropy": 5.55757327079773,
      "epoch": 2.7408381712626997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042514034360099907,
      "loss": 4.9079,
      "mean_token_accuracy": 0.21350128650665284,
      "num_tokens": 62702686.0,
      "step": 30215
    },
    {
      "entropy": 5.6044361114501955,
      "epoch": 2.7412917271407835,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042511624306402015,
      "loss": 4.9528,
      "mean_token_accuracy": 0.21285638809204102,
      "num_tokens": 62713262.0,
      "step": 30220
    },
    {
      "entropy": 5.6452392578125,
      "epoch": 2.7417452830188678,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004250921394225992,
      "loss": 4.9002,
      "mean_token_accuracy": 0.21436085402965546,
      "num_tokens": 62724460.0,
      "step": 30225
    },
    {
      "entropy": 5.6062828540802006,
      "epoch": 2.742198838896952,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004250680326772343,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21282104700803756,
      "num_tokens": 62734190.0,
      "step": 30230
    },
    {
      "entropy": 5.617833518981934,
      "epoch": 2.7426523947750363,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004250439228284244,
      "loss": 4.9192,
      "mean_token_accuracy": 0.2149022102355957,
      "num_tokens": 62744323.0,
      "step": 30235
    },
    {
      "entropy": 5.622685670852661,
      "epoch": 2.7431059506531206,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042501980987666807,
      "loss": 4.9481,
      "mean_token_accuracy": 0.21487101167440414,
      "num_tokens": 62753868.0,
      "step": 30240
    },
    {
      "entropy": 5.627379846572876,
      "epoch": 2.743559506531205,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004249956938224636,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21195608079433442,
      "num_tokens": 62763954.0,
      "step": 30245
    },
    {
      "entropy": 5.6044105052947994,
      "epoch": 2.7440130624092887,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042497157466631015,
      "loss": 4.918,
      "mean_token_accuracy": 0.22302944958209991,
      "num_tokens": 62774116.0,
      "step": 30250
    },
    {
      "entropy": 5.674649286270141,
      "epoch": 2.744466618287373,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042494745240870626,
      "loss": 4.99,
      "mean_token_accuracy": 0.2058614671230316,
      "num_tokens": 62784939.0,
      "step": 30255
    },
    {
      "entropy": 5.601968765258789,
      "epoch": 2.744920174165457,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042492332705015075,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21411143690347673,
      "num_tokens": 62794988.0,
      "step": 30260
    },
    {
      "entropy": 5.625520038604736,
      "epoch": 2.7453737300435415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042489919859114273,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21706822514533997,
      "num_tokens": 62804984.0,
      "step": 30265
    },
    {
      "entropy": 5.643641519546509,
      "epoch": 2.7458272859216253,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00042487506703218104,
      "loss": 4.9188,
      "mean_token_accuracy": 0.20847680121660234,
      "num_tokens": 62816401.0,
      "step": 30270
    },
    {
      "entropy": 5.560655736923218,
      "epoch": 2.7462808417997095,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004248509323737648,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22057805806398392,
      "num_tokens": 62826411.0,
      "step": 30275
    },
    {
      "entropy": 5.628651428222656,
      "epoch": 2.746734397677794,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004248267946163931,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20857589840888976,
      "num_tokens": 62837228.0,
      "step": 30280
    },
    {
      "entropy": 5.576494646072388,
      "epoch": 2.747187953555878,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042480265376056506,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22177619487047195,
      "num_tokens": 62847852.0,
      "step": 30285
    },
    {
      "entropy": 5.613132286071777,
      "epoch": 2.7476415094339623,
      "grad_norm": 1.125,
      "learning_rate": 0.00042477850980678,
      "loss": 4.9812,
      "mean_token_accuracy": 0.20616811960935594,
      "num_tokens": 62859516.0,
      "step": 30290
    },
    {
      "entropy": 5.62861328125,
      "epoch": 2.7480950653120466,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004247543627555372,
      "loss": 4.9579,
      "mean_token_accuracy": 0.2111853152513504,
      "num_tokens": 62869757.0,
      "step": 30295
    },
    {
      "entropy": 5.613959884643554,
      "epoch": 2.748548621190131,
      "grad_norm": 1.296875,
      "learning_rate": 0.000424730212607336,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2095622733235359,
      "num_tokens": 62879618.0,
      "step": 30300
    },
    {
      "entropy": 5.624938344955444,
      "epoch": 2.7490021770682147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004247060593626758,
      "loss": 4.9739,
      "mean_token_accuracy": 0.20830693542957307,
      "num_tokens": 62890491.0,
      "step": 30305
    },
    {
      "entropy": 5.697379207611084,
      "epoch": 2.749455732946299,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004246819030220562,
      "loss": 4.9849,
      "mean_token_accuracy": 0.22261755764484406,
      "num_tokens": 62899954.0,
      "step": 30310
    },
    {
      "entropy": 5.635549783706665,
      "epoch": 2.7499092888243832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004246577435859767,
      "loss": 4.9793,
      "mean_token_accuracy": 0.2149518683552742,
      "num_tokens": 62911857.0,
      "step": 30315
    },
    {
      "entropy": 5.574631595611573,
      "epoch": 2.7503628447024675,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004246335810549369,
      "loss": 4.9097,
      "mean_token_accuracy": 0.2194608524441719,
      "num_tokens": 62922023.0,
      "step": 30320
    },
    {
      "entropy": 5.647210216522216,
      "epoch": 2.7508164005805513,
      "grad_norm": 1.125,
      "learning_rate": 0.00042460941542943656,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21846277862787247,
      "num_tokens": 62932845.0,
      "step": 30325
    },
    {
      "entropy": 5.632871198654175,
      "epoch": 2.7512699564586356,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004245852467099753,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2101821258664131,
      "num_tokens": 62942881.0,
      "step": 30330
    },
    {
      "entropy": 5.5770667552947994,
      "epoch": 2.75172351233672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004245610748970531,
      "loss": 4.8896,
      "mean_token_accuracy": 0.21483128517866135,
      "num_tokens": 62953957.0,
      "step": 30335
    },
    {
      "entropy": 5.554941320419312,
      "epoch": 2.752177068214804,
      "grad_norm": 1.25,
      "learning_rate": 0.00042453689999116974,
      "loss": 4.8983,
      "mean_token_accuracy": 0.21520988047122955,
      "num_tokens": 62964803.0,
      "step": 30340
    },
    {
      "entropy": 5.593675518035889,
      "epoch": 2.7526306240928884,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004245127219928252,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21524002999067307,
      "num_tokens": 62975477.0,
      "step": 30345
    },
    {
      "entropy": 5.598050594329834,
      "epoch": 2.7530841799709727,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042448854090251944,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2135425627231598,
      "num_tokens": 62985430.0,
      "step": 30350
    },
    {
      "entropy": 5.601573753356933,
      "epoch": 2.7535377358490565,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042446435672075254,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2169000968337059,
      "num_tokens": 62997493.0,
      "step": 30355
    },
    {
      "entropy": 5.685353231430054,
      "epoch": 2.7539912917271407,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00042444016944802467,
      "loss": 5.0242,
      "mean_token_accuracy": 0.2181790977716446,
      "num_tokens": 63007573.0,
      "step": 30360
    },
    {
      "entropy": 5.66036901473999,
      "epoch": 2.754444847605225,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042441597908483603,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21657835245132445,
      "num_tokens": 63018372.0,
      "step": 30365
    },
    {
      "entropy": 5.59838342666626,
      "epoch": 2.7548984034833093,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004243917856316869,
      "loss": 4.9217,
      "mean_token_accuracy": 0.21629077792167664,
      "num_tokens": 63028632.0,
      "step": 30370
    },
    {
      "entropy": 5.6141571521759035,
      "epoch": 2.755351959361393,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042436758908907747,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2099437564611435,
      "num_tokens": 63038510.0,
      "step": 30375
    },
    {
      "entropy": 5.715930843353272,
      "epoch": 2.7558055152394774,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042434338945750833,
      "loss": 5.1083,
      "mean_token_accuracy": 0.19410379528999328,
      "num_tokens": 63049375.0,
      "step": 30380
    },
    {
      "entropy": 5.65672287940979,
      "epoch": 2.7562590711175616,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042431918673747974,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21619293242692947,
      "num_tokens": 63059422.0,
      "step": 30385
    },
    {
      "entropy": 5.6731163501739506,
      "epoch": 2.756712626995646,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042429498092949245,
      "loss": 5.0149,
      "mean_token_accuracy": 0.2059454470872879,
      "num_tokens": 63071101.0,
      "step": 30390
    },
    {
      "entropy": 5.546786546707153,
      "epoch": 2.75716618287373,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004242707720340468,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22282183915376663,
      "num_tokens": 63081104.0,
      "step": 30395
    },
    {
      "entropy": 5.600458192825317,
      "epoch": 2.7576197387518144,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004242465600516436,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2115418627858162,
      "num_tokens": 63091123.0,
      "step": 30400
    },
    {
      "entropy": 5.583108234405517,
      "epoch": 2.7580732946298983,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042422234498278353,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21760235726833344,
      "num_tokens": 63101541.0,
      "step": 30405
    },
    {
      "entropy": 5.652334785461425,
      "epoch": 2.7585268505079825,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042419812682796727,
      "loss": 5.0279,
      "mean_token_accuracy": 0.20402513295412064,
      "num_tokens": 63112012.0,
      "step": 30410
    },
    {
      "entropy": 5.6674047946929935,
      "epoch": 2.758980406386067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004241739055876957,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21688816398382188,
      "num_tokens": 63122021.0,
      "step": 30415
    },
    {
      "entropy": 5.68640022277832,
      "epoch": 2.759433962264151,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042414968126246985,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2107762411236763,
      "num_tokens": 63132286.0,
      "step": 30420
    },
    {
      "entropy": 5.572261667251587,
      "epoch": 2.759887518142235,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042412545385279055,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2180559977889061,
      "num_tokens": 63142159.0,
      "step": 30425
    },
    {
      "entropy": 5.56485390663147,
      "epoch": 2.760341074020319,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042410122335915883,
      "loss": 4.8849,
      "mean_token_accuracy": 0.217661452293396,
      "num_tokens": 63152225.0,
      "step": 30430
    },
    {
      "entropy": 5.591285991668701,
      "epoch": 2.7607946298984034,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004240769897820758,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2141361877322197,
      "num_tokens": 63162522.0,
      "step": 30435
    },
    {
      "entropy": 5.6189690113067625,
      "epoch": 2.7612481857764877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042405275312204266,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2210882067680359,
      "num_tokens": 63172094.0,
      "step": 30440
    },
    {
      "entropy": 5.633256101608277,
      "epoch": 2.761701741654572,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042402851337956066,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21410634964704514,
      "num_tokens": 63182122.0,
      "step": 30445
    },
    {
      "entropy": 5.582202386856079,
      "epoch": 2.762155297532656,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042400427055513094,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2122228890657425,
      "num_tokens": 63193174.0,
      "step": 30450
    },
    {
      "entropy": 5.609793853759766,
      "epoch": 2.7626088534107405,
      "grad_norm": 1.171875,
      "learning_rate": 0.000423980024649255,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22054361402988434,
      "num_tokens": 63203698.0,
      "step": 30455
    },
    {
      "entropy": 5.517379808425903,
      "epoch": 2.7630624092888243,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004239557756624342,
      "loss": 4.8349,
      "mean_token_accuracy": 0.21590106785297394,
      "num_tokens": 63213343.0,
      "step": 30460
    },
    {
      "entropy": 5.652347421646118,
      "epoch": 2.7635159651669086,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042393152359516994,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21349549740552903,
      "num_tokens": 63223962.0,
      "step": 30465
    },
    {
      "entropy": 5.594416332244873,
      "epoch": 2.763969521044993,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042390726844796384,
      "loss": 4.904,
      "mean_token_accuracy": 0.22045750319957733,
      "num_tokens": 63234279.0,
      "step": 30470
    },
    {
      "entropy": 5.644147968292236,
      "epoch": 2.7644230769230766,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004238830102213174,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2150461569428444,
      "num_tokens": 63245567.0,
      "step": 30475
    },
    {
      "entropy": 5.6111259937286375,
      "epoch": 2.764876632801161,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004238587489157325,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2222388952970505,
      "num_tokens": 63256227.0,
      "step": 30480
    },
    {
      "entropy": 5.554749441146851,
      "epoch": 2.765330188679245,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004238344845317107,
      "loss": 4.85,
      "mean_token_accuracy": 0.217688125371933,
      "num_tokens": 63266618.0,
      "step": 30485
    },
    {
      "entropy": 5.562039613723755,
      "epoch": 2.7657837445573294,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004238102170697539,
      "loss": 4.8811,
      "mean_token_accuracy": 0.21864323318004608,
      "num_tokens": 63277130.0,
      "step": 30490
    },
    {
      "entropy": 5.536762666702271,
      "epoch": 2.7662373004354137,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004237859465303637,
      "loss": 4.8083,
      "mean_token_accuracy": 0.22327447384595872,
      "num_tokens": 63287351.0,
      "step": 30495
    },
    {
      "entropy": 5.531674861907959,
      "epoch": 2.766690856313498,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042376167291404234,
      "loss": 4.8937,
      "mean_token_accuracy": 0.21843965947628022,
      "num_tokens": 63298573.0,
      "step": 30500
    },
    {
      "entropy": 5.598010063171387,
      "epoch": 2.7671444121915822,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042373739622129166,
      "loss": 4.9423,
      "mean_token_accuracy": 0.20653199702501296,
      "num_tokens": 63310740.0,
      "step": 30505
    },
    {
      "entropy": 5.5904233932495115,
      "epoch": 2.767597968069666,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042371311645261376,
      "loss": 4.9486,
      "mean_token_accuracy": 0.21120081692934037,
      "num_tokens": 63321323.0,
      "step": 30510
    },
    {
      "entropy": 5.614881372451782,
      "epoch": 2.7680515239477503,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004236888336085107,
      "loss": 4.923,
      "mean_token_accuracy": 0.21754363179206848,
      "num_tokens": 63332322.0,
      "step": 30515
    },
    {
      "entropy": 5.57806167602539,
      "epoch": 2.7685050798258346,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004236645476894846,
      "loss": 4.8441,
      "mean_token_accuracy": 0.22277876883745193,
      "num_tokens": 63342204.0,
      "step": 30520
    },
    {
      "entropy": 5.527768850326538,
      "epoch": 2.768958635703919,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004236402586960378,
      "loss": 4.9092,
      "mean_token_accuracy": 0.21968706101179122,
      "num_tokens": 63352460.0,
      "step": 30525
    },
    {
      "entropy": 5.522051906585693,
      "epoch": 2.7694121915820027,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004236159666286726,
      "loss": 4.8813,
      "mean_token_accuracy": 0.21861146092414857,
      "num_tokens": 63362474.0,
      "step": 30530
    },
    {
      "entropy": 5.640454578399658,
      "epoch": 2.769865747460087,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042359167148789135,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22466543614864348,
      "num_tokens": 63372537.0,
      "step": 30535
    },
    {
      "entropy": 5.555916166305542,
      "epoch": 2.7703193033381712,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004235673732741964,
      "loss": 4.8276,
      "mean_token_accuracy": 0.2260143905878067,
      "num_tokens": 63382985.0,
      "step": 30540
    },
    {
      "entropy": 5.570925855636597,
      "epoch": 2.7707728592162555,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042354307198809037,
      "loss": 4.9184,
      "mean_token_accuracy": 0.21783771067857743,
      "num_tokens": 63393424.0,
      "step": 30545
    },
    {
      "entropy": 5.533125305175782,
      "epoch": 2.7712264150943398,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004235187676300757,
      "loss": 4.8299,
      "mean_token_accuracy": 0.22464577108621597,
      "num_tokens": 63403160.0,
      "step": 30550
    },
    {
      "entropy": 5.636246109008789,
      "epoch": 2.771679970972424,
      "grad_norm": 1.125,
      "learning_rate": 0.0004234944602006552,
      "loss": 4.9583,
      "mean_token_accuracy": 0.2132783517241478,
      "num_tokens": 63413001.0,
      "step": 30555
    },
    {
      "entropy": 5.625688600540161,
      "epoch": 2.772133526850508,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042347014970033124,
      "loss": 4.9076,
      "mean_token_accuracy": 0.2218951016664505,
      "num_tokens": 63422735.0,
      "step": 30560
    },
    {
      "entropy": 5.531408596038818,
      "epoch": 2.772587082728592,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004234458361296069,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21520378440618515,
      "num_tokens": 63432566.0,
      "step": 30565
    },
    {
      "entropy": 5.589576768875122,
      "epoch": 2.7730406386066764,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042342151948898476,
      "loss": 4.9049,
      "mean_token_accuracy": 0.21632362753152848,
      "num_tokens": 63442412.0,
      "step": 30570
    },
    {
      "entropy": 5.59679684638977,
      "epoch": 2.7734941944847606,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042339719977896784,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21769312173128127,
      "num_tokens": 63453018.0,
      "step": 30575
    },
    {
      "entropy": 5.662491655349731,
      "epoch": 2.7739477503628445,
      "grad_norm": 1.1875,
      "learning_rate": 0.000423372877000059,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21131200045347215,
      "num_tokens": 63464286.0,
      "step": 30580
    },
    {
      "entropy": 5.55959038734436,
      "epoch": 2.7744013062409287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042334855115276117,
      "loss": 4.9431,
      "mean_token_accuracy": 0.213600817322731,
      "num_tokens": 63475138.0,
      "step": 30585
    },
    {
      "entropy": 5.623905515670776,
      "epoch": 2.774854862119013,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042332422223757756,
      "loss": 4.999,
      "mean_token_accuracy": 0.21151673048734665,
      "num_tokens": 63486104.0,
      "step": 30590
    },
    {
      "entropy": 5.556403493881225,
      "epoch": 2.7753084179970973,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042329989025501123,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2150347650051117,
      "num_tokens": 63496490.0,
      "step": 30595
    },
    {
      "entropy": 5.569782733917236,
      "epoch": 2.7757619738751815,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042327555520556536,
      "loss": 4.8906,
      "mean_token_accuracy": 0.21169639229774476,
      "num_tokens": 63506562.0,
      "step": 30600
    },
    {
      "entropy": 5.627491617202759,
      "epoch": 2.776215529753266,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042325121708974324,
      "loss": 4.9272,
      "mean_token_accuracy": 0.2239815905690193,
      "num_tokens": 63516641.0,
      "step": 30605
    },
    {
      "entropy": 5.622805070877075,
      "epoch": 2.7766690856313496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004232268759080482,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21638973504304887,
      "num_tokens": 63526864.0,
      "step": 30610
    },
    {
      "entropy": 5.586530923843384,
      "epoch": 2.777122641509434,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042320253166098354,
      "loss": 5.007,
      "mean_token_accuracy": 0.20974934697151185,
      "num_tokens": 63537160.0,
      "step": 30615
    },
    {
      "entropy": 5.639689874649048,
      "epoch": 2.777576197387518,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042317818434905275,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21598411500453948,
      "num_tokens": 63547485.0,
      "step": 30620
    },
    {
      "entropy": 5.61621241569519,
      "epoch": 2.7780297532656024,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004231538339727594,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21545920968055726,
      "num_tokens": 63559369.0,
      "step": 30625
    },
    {
      "entropy": 5.693338298797608,
      "epoch": 2.7784833091436862,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042312948053260704,
      "loss": 5.092,
      "mean_token_accuracy": 0.20301660895347595,
      "num_tokens": 63570033.0,
      "step": 30630
    },
    {
      "entropy": 5.5835034370422365,
      "epoch": 2.7789368650217705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004231051240290992,
      "loss": 4.8879,
      "mean_token_accuracy": 0.21866330057382583,
      "num_tokens": 63581274.0,
      "step": 30635
    },
    {
      "entropy": 5.638933849334717,
      "epoch": 2.7793904208998548,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042308076446273963,
      "loss": 4.9321,
      "mean_token_accuracy": 0.21271978169679642,
      "num_tokens": 63590903.0,
      "step": 30640
    },
    {
      "entropy": 5.56838641166687,
      "epoch": 2.779843976777939,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042305640183403214,
      "loss": 4.9276,
      "mean_token_accuracy": 0.20841511189937592,
      "num_tokens": 63600982.0,
      "step": 30645
    },
    {
      "entropy": 5.511862945556641,
      "epoch": 2.7802975326560233,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042303203614348057,
      "loss": 4.8421,
      "mean_token_accuracy": 0.2130862757563591,
      "num_tokens": 63612411.0,
      "step": 30650
    },
    {
      "entropy": 5.567126846313476,
      "epoch": 2.7807510885341076,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004230076673915888,
      "loss": 4.8886,
      "mean_token_accuracy": 0.22152767181396485,
      "num_tokens": 63622128.0,
      "step": 30655
    },
    {
      "entropy": 5.615833806991577,
      "epoch": 2.781204644412192,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042298329557886074,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21670996397733688,
      "num_tokens": 63632047.0,
      "step": 30660
    },
    {
      "entropy": 5.653778076171875,
      "epoch": 2.7816582002902757,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004229589207058003,
      "loss": 4.9648,
      "mean_token_accuracy": 0.21392653584480287,
      "num_tokens": 63642490.0,
      "step": 30665
    },
    {
      "entropy": 5.586833047866821,
      "epoch": 2.78211175616836,
      "grad_norm": 1.125,
      "learning_rate": 0.00042293454277291177,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2138791263103485,
      "num_tokens": 63653206.0,
      "step": 30670
    },
    {
      "entropy": 5.529301452636719,
      "epoch": 2.782565312046444,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042291016178069916,
      "loss": 4.924,
      "mean_token_accuracy": 0.21557315289974213,
      "num_tokens": 63663427.0,
      "step": 30675
    },
    {
      "entropy": 5.534003829956054,
      "epoch": 2.7830188679245285,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004228857777296668,
      "loss": 4.9117,
      "mean_token_accuracy": 0.21934759020805358,
      "num_tokens": 63674269.0,
      "step": 30680
    },
    {
      "entropy": 5.62264838218689,
      "epoch": 2.7834724238026123,
      "grad_norm": 1.125,
      "learning_rate": 0.0004228613906203188,
      "loss": 4.8938,
      "mean_token_accuracy": 0.21602804809808732,
      "num_tokens": 63683989.0,
      "step": 30685
    },
    {
      "entropy": 5.614567136764526,
      "epoch": 2.7839259796806965,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042283700045315955,
      "loss": 4.9184,
      "mean_token_accuracy": 0.21105228066444398,
      "num_tokens": 63694540.0,
      "step": 30690
    },
    {
      "entropy": 5.589615201950073,
      "epoch": 2.784379535558781,
      "grad_norm": 1.125,
      "learning_rate": 0.00042281260722869355,
      "loss": 4.899,
      "mean_token_accuracy": 0.21366845369338988,
      "num_tokens": 63705149.0,
      "step": 30695
    },
    {
      "entropy": 5.5503425121307375,
      "epoch": 2.784833091436865,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042278821094742495,
      "loss": 4.881,
      "mean_token_accuracy": 0.2166728511452675,
      "num_tokens": 63714061.0,
      "step": 30700
    },
    {
      "entropy": 5.6025643825531,
      "epoch": 2.7852866473149493,
      "grad_norm": 1.125,
      "learning_rate": 0.00042276381160985874,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21572295576334,
      "num_tokens": 63724732.0,
      "step": 30705
    },
    {
      "entropy": 5.596027994155884,
      "epoch": 2.7857402031930336,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004227394092164991,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21107783615589143,
      "num_tokens": 63734942.0,
      "step": 30710
    },
    {
      "entropy": 5.672230577468872,
      "epoch": 2.7861937590711174,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004227150037678509,
      "loss": 5.0425,
      "mean_token_accuracy": 0.21797527968883515,
      "num_tokens": 63745326.0,
      "step": 30715
    },
    {
      "entropy": 5.6225536346435545,
      "epoch": 2.7866473149492017,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004226905952644188,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2138058990240097,
      "num_tokens": 63755355.0,
      "step": 30720
    },
    {
      "entropy": 5.624581813812256,
      "epoch": 2.787100870827286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042266618370670743,
      "loss": 4.9212,
      "mean_token_accuracy": 0.212918521463871,
      "num_tokens": 63765112.0,
      "step": 30725
    },
    {
      "entropy": 5.571261072158814,
      "epoch": 2.7875544267053702,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042264176909522183,
      "loss": 4.8748,
      "mean_token_accuracy": 0.21993715316057205,
      "num_tokens": 63774680.0,
      "step": 30730
    },
    {
      "entropy": 5.621033334732056,
      "epoch": 2.788007982583454,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004226173514304669,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21756456643342972,
      "num_tokens": 63784939.0,
      "step": 30735
    },
    {
      "entropy": 5.629879808425903,
      "epoch": 2.7884615384615383,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004225929307129474,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20408451408147812,
      "num_tokens": 63795836.0,
      "step": 30740
    },
    {
      "entropy": 5.574771404266357,
      "epoch": 2.7889150943396226,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042256850694316844,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2149980679154396,
      "num_tokens": 63805607.0,
      "step": 30745
    },
    {
      "entropy": 5.53808741569519,
      "epoch": 2.789368650217707,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004225440801216353,
      "loss": 4.888,
      "mean_token_accuracy": 0.21340310722589492,
      "num_tokens": 63815081.0,
      "step": 30750
    },
    {
      "entropy": 5.556040334701538,
      "epoch": 2.789822206095791,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004225196502488529,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2129345253109932,
      "num_tokens": 63825932.0,
      "step": 30755
    },
    {
      "entropy": 5.579881572723389,
      "epoch": 2.7902757619738754,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042249521732532657,
      "loss": 4.8605,
      "mean_token_accuracy": 0.21878161281347275,
      "num_tokens": 63835653.0,
      "step": 30760
    },
    {
      "entropy": 5.674952793121338,
      "epoch": 2.790729317851959,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004224707813515615,
      "loss": 5.0583,
      "mean_token_accuracy": 0.20467241555452348,
      "num_tokens": 63847657.0,
      "step": 30765
    },
    {
      "entropy": 5.666828441619873,
      "epoch": 2.7911828737300435,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042244634232806324,
      "loss": 4.9602,
      "mean_token_accuracy": 0.2172686532139778,
      "num_tokens": 63858476.0,
      "step": 30770
    },
    {
      "entropy": 5.556519317626953,
      "epoch": 2.7916364296081277,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042242190025533694,
      "loss": 4.8279,
      "mean_token_accuracy": 0.22101556658744811,
      "num_tokens": 63867970.0,
      "step": 30775
    },
    {
      "entropy": 5.604297065734864,
      "epoch": 2.792089985486212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004223974551338881,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22202686071395875,
      "num_tokens": 63878294.0,
      "step": 30780
    },
    {
      "entropy": 5.545058012008667,
      "epoch": 2.792543541364296,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004223730069642224,
      "loss": 4.914,
      "mean_token_accuracy": 0.2130719855427742,
      "num_tokens": 63889315.0,
      "step": 30785
    },
    {
      "entropy": 5.520653200149536,
      "epoch": 2.79299709724238,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004223485557468454,
      "loss": 4.8923,
      "mean_token_accuracy": 0.21057825535535812,
      "num_tokens": 63899923.0,
      "step": 30790
    },
    {
      "entropy": 5.573180055618286,
      "epoch": 2.7934506531204644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042232410148226277,
      "loss": 4.8777,
      "mean_token_accuracy": 0.21802876740694047,
      "num_tokens": 63909813.0,
      "step": 30795
    },
    {
      "entropy": 5.607568979263306,
      "epoch": 2.7939042089985486,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042229964417098007,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21219711899757385,
      "num_tokens": 63920931.0,
      "step": 30800
    },
    {
      "entropy": 5.55732946395874,
      "epoch": 2.794357764876633,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042227518381350316,
      "loss": 4.868,
      "mean_token_accuracy": 0.22100864499807357,
      "num_tokens": 63930799.0,
      "step": 30805
    },
    {
      "entropy": 5.66492600440979,
      "epoch": 2.794811320754717,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042225072041033804,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2090522140264511,
      "num_tokens": 63941763.0,
      "step": 30810
    },
    {
      "entropy": 5.6471329689025875,
      "epoch": 2.7952648766328014,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042222625396199043,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20792181491851808,
      "num_tokens": 63952775.0,
      "step": 30815
    },
    {
      "entropy": 5.630888032913208,
      "epoch": 2.7957184325108853,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004222017844689663,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21161658465862274,
      "num_tokens": 63964223.0,
      "step": 30820
    },
    {
      "entropy": 5.6678611755371096,
      "epoch": 2.7961719883889695,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004221773119317719,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21187895834445952,
      "num_tokens": 63976413.0,
      "step": 30825
    },
    {
      "entropy": 5.673243284225464,
      "epoch": 2.796625544267054,
      "grad_norm": 1.375,
      "learning_rate": 0.00042215283635091313,
      "loss": 4.96,
      "mean_token_accuracy": 0.21708431094884872,
      "num_tokens": 63986578.0,
      "step": 30830
    },
    {
      "entropy": 5.615045881271362,
      "epoch": 2.7970791001451376,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004221283577268962,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21961487084627151,
      "num_tokens": 63996856.0,
      "step": 30835
    },
    {
      "entropy": 5.556055927276612,
      "epoch": 2.797532656023222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042210387606022724,
      "loss": 4.9003,
      "mean_token_accuracy": 0.21939728856086732,
      "num_tokens": 64007700.0,
      "step": 30840
    },
    {
      "entropy": 5.660103797912598,
      "epoch": 2.797986211901306,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004220793913514128,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20538849085569383,
      "num_tokens": 64018771.0,
      "step": 30845
    },
    {
      "entropy": 5.623592329025269,
      "epoch": 2.7984397677793904,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004220549036009589,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2099561020731926,
      "num_tokens": 64029381.0,
      "step": 30850
    },
    {
      "entropy": 5.5213452816009525,
      "epoch": 2.7988933236574747,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042203041280937224,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21454301476478577,
      "num_tokens": 64039596.0,
      "step": 30855
    },
    {
      "entropy": 5.6389189720153805,
      "epoch": 2.799346879535559,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004220059189771592,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22159379422664643,
      "num_tokens": 64049123.0,
      "step": 30860
    },
    {
      "entropy": 5.632424592971802,
      "epoch": 2.799800435413643,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042198142210482623,
      "loss": 4.9803,
      "mean_token_accuracy": 0.2122730866074562,
      "num_tokens": 64059426.0,
      "step": 30865
    },
    {
      "entropy": 5.575161647796631,
      "epoch": 2.800253991291727,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042195692219287997,
      "loss": 4.8447,
      "mean_token_accuracy": 0.22846995741128923,
      "num_tokens": 64069227.0,
      "step": 30870
    },
    {
      "entropy": 5.630750513076782,
      "epoch": 2.8007075471698113,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004219324192418272,
      "loss": 4.9382,
      "mean_token_accuracy": 0.22248561233282088,
      "num_tokens": 64078482.0,
      "step": 30875
    },
    {
      "entropy": 5.5800941467285154,
      "epoch": 2.8011611030478956,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042190791325217447,
      "loss": 4.9028,
      "mean_token_accuracy": 0.216207854449749,
      "num_tokens": 64088664.0,
      "step": 30880
    },
    {
      "entropy": 5.6033862113952635,
      "epoch": 2.80161465892598,
      "grad_norm": 1.25,
      "learning_rate": 0.0004218834042244287,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22169570922851561,
      "num_tokens": 64099316.0,
      "step": 30885
    },
    {
      "entropy": 5.567199945449829,
      "epoch": 2.8020682148040637,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042185889215909665,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21490092873573302,
      "num_tokens": 64109786.0,
      "step": 30890
    },
    {
      "entropy": 5.608392429351807,
      "epoch": 2.802521770682148,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042183437705668535,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22288060039281846,
      "num_tokens": 64120771.0,
      "step": 30895
    },
    {
      "entropy": 5.644888544082642,
      "epoch": 2.802975326560232,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042180985891770176,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21018804460763932,
      "num_tokens": 64130833.0,
      "step": 30900
    },
    {
      "entropy": 5.560075283050537,
      "epoch": 2.8034288824383164,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004217853377426528,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22247102856636047,
      "num_tokens": 64139672.0,
      "step": 30905
    },
    {
      "entropy": 5.7054849624633786,
      "epoch": 2.8038824383164007,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004217608135320456,
      "loss": 5.0529,
      "mean_token_accuracy": 0.21037883460521697,
      "num_tokens": 64150383.0,
      "step": 30910
    },
    {
      "entropy": 5.574859619140625,
      "epoch": 2.804335994194485,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004217362862863875,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22063290029764177,
      "num_tokens": 64160786.0,
      "step": 30915
    },
    {
      "entropy": 5.632054996490479,
      "epoch": 2.804789550072569,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042171175600618546,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21001256257295609,
      "num_tokens": 64172057.0,
      "step": 30920
    },
    {
      "entropy": 5.593688488006592,
      "epoch": 2.805243105950653,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042168722269194704,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21802769750356674,
      "num_tokens": 64182599.0,
      "step": 30925
    },
    {
      "entropy": 5.61622953414917,
      "epoch": 2.8056966618287373,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004216626863441794,
      "loss": 4.9647,
      "mean_token_accuracy": 0.2088227942585945,
      "num_tokens": 64193356.0,
      "step": 30930
    },
    {
      "entropy": 5.676929092407226,
      "epoch": 2.8061502177068216,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042163814696339014,
      "loss": 5.0388,
      "mean_token_accuracy": 0.2096481665968895,
      "num_tokens": 64203625.0,
      "step": 30935
    },
    {
      "entropy": 5.618952989578247,
      "epoch": 2.8066037735849054,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004216136045500866,
      "loss": 5.0406,
      "mean_token_accuracy": 0.21401924192905425,
      "num_tokens": 64214742.0,
      "step": 30940
    },
    {
      "entropy": 5.628974580764771,
      "epoch": 2.8070573294629897,
      "grad_norm": 1.25,
      "learning_rate": 0.0004215890591047763,
      "loss": 4.9425,
      "mean_token_accuracy": 0.21468819975852965,
      "num_tokens": 64223811.0,
      "step": 30945
    },
    {
      "entropy": 5.607697010040283,
      "epoch": 2.807510885341074,
      "grad_norm": 1.125,
      "learning_rate": 0.00042156451062796686,
      "loss": 4.8491,
      "mean_token_accuracy": 0.2281048372387886,
      "num_tokens": 64234854.0,
      "step": 30950
    },
    {
      "entropy": 5.597255277633667,
      "epoch": 2.8079644412191582,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042153995912016606,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21392620652914046,
      "num_tokens": 64245914.0,
      "step": 30955
    },
    {
      "entropy": 5.595199203491211,
      "epoch": 2.8084179970972425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004215154045818816,
      "loss": 5.0029,
      "mean_token_accuracy": 0.21000120639801026,
      "num_tokens": 64256217.0,
      "step": 30960
    },
    {
      "entropy": 5.631856536865234,
      "epoch": 2.8088715529753268,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004214908470136212,
      "loss": 4.9976,
      "mean_token_accuracy": 0.20288413763046265,
      "num_tokens": 64267211.0,
      "step": 30965
    },
    {
      "entropy": 5.686559963226318,
      "epoch": 2.8093251088534106,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004214662864158927,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21689907610416412,
      "num_tokens": 64277770.0,
      "step": 30970
    },
    {
      "entropy": 5.585747337341308,
      "epoch": 2.809778664731495,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042144172278920413,
      "loss": 4.8615,
      "mean_token_accuracy": 0.2203228995203972,
      "num_tokens": 64287973.0,
      "step": 30975
    },
    {
      "entropy": 5.566807556152344,
      "epoch": 2.810232220609579,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004214171561340634,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21033688634634018,
      "num_tokens": 64297898.0,
      "step": 30980
    },
    {
      "entropy": 5.590597438812256,
      "epoch": 2.8106857764876634,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042139258645097863,
      "loss": 4.8114,
      "mean_token_accuracy": 0.22951457649469376,
      "num_tokens": 64308114.0,
      "step": 30985
    },
    {
      "entropy": 5.567050170898438,
      "epoch": 2.811139332365747,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042136801374045775,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21565261632204055,
      "num_tokens": 64319586.0,
      "step": 30990
    },
    {
      "entropy": 5.6042194843292235,
      "epoch": 2.8115928882438315,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004213434380030091,
      "loss": 4.9978,
      "mean_token_accuracy": 0.21070887446403502,
      "num_tokens": 64330612.0,
      "step": 30995
    },
    {
      "entropy": 5.6452186584472654,
      "epoch": 2.8120464441219157,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004213188592391409,
      "loss": 5.015,
      "mean_token_accuracy": 0.21246811747550964,
      "num_tokens": 64341934.0,
      "step": 31000
    },
    {
      "entropy": 5.621395111083984,
      "epoch": 2.8125,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004212942774493614,
      "loss": 4.9173,
      "mean_token_accuracy": 0.21760687381029128,
      "num_tokens": 64352166.0,
      "step": 31005
    },
    {
      "entropy": 5.589327812194824,
      "epoch": 2.8129535558780843,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042126969263417885,
      "loss": 4.964,
      "mean_token_accuracy": 0.2079501509666443,
      "num_tokens": 64361995.0,
      "step": 31010
    },
    {
      "entropy": 5.610994625091553,
      "epoch": 2.8134071117561685,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042124510479410183,
      "loss": 4.8934,
      "mean_token_accuracy": 0.21078013777732849,
      "num_tokens": 64372336.0,
      "step": 31015
    },
    {
      "entropy": 5.679530954360962,
      "epoch": 2.813860667634253,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042122051392963885,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22433409094810486,
      "num_tokens": 64382309.0,
      "step": 31020
    },
    {
      "entropy": 5.62575511932373,
      "epoch": 2.8143142235123366,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004211959200412984,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2136935696005821,
      "num_tokens": 64393205.0,
      "step": 31025
    },
    {
      "entropy": 5.560979747772217,
      "epoch": 2.814767779390421,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004211713231295889,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21102684140205383,
      "num_tokens": 64403068.0,
      "step": 31030
    },
    {
      "entropy": 5.627737188339234,
      "epoch": 2.815221335268505,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004211467231950194,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21938664764165877,
      "num_tokens": 64414296.0,
      "step": 31035
    },
    {
      "entropy": 5.596076774597168,
      "epoch": 2.8156748911465894,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042112212023809824,
      "loss": 4.9661,
      "mean_token_accuracy": 0.20868293195962906,
      "num_tokens": 64424748.0,
      "step": 31040
    },
    {
      "entropy": 5.631599569320679,
      "epoch": 2.8161284470246732,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004210975142593345,
      "loss": 4.8872,
      "mean_token_accuracy": 0.2139439269900322,
      "num_tokens": 64434723.0,
      "step": 31045
    },
    {
      "entropy": 5.686302137374878,
      "epoch": 2.8165820029027575,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004210729052592369,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20711847096681596,
      "num_tokens": 64445864.0,
      "step": 31050
    },
    {
      "entropy": 5.538408374786377,
      "epoch": 2.8170355587808418,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004210482932383144,
      "loss": 4.8721,
      "mean_token_accuracy": 0.21520453244447707,
      "num_tokens": 64456704.0,
      "step": 31055
    },
    {
      "entropy": 5.518161869049072,
      "epoch": 2.817489114658926,
      "grad_norm": 1.234375,
      "learning_rate": 0.000421023678197076,
      "loss": 4.9433,
      "mean_token_accuracy": 0.219936366379261,
      "num_tokens": 64466870.0,
      "step": 31060
    },
    {
      "entropy": 5.6191342830657955,
      "epoch": 2.8179426705370103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004209990601360307,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21801260113716125,
      "num_tokens": 64477069.0,
      "step": 31065
    },
    {
      "entropy": 5.670992422103882,
      "epoch": 2.8183962264150946,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042097443905568765,
      "loss": 4.9477,
      "mean_token_accuracy": 0.20735147446393967,
      "num_tokens": 64487165.0,
      "step": 31070
    },
    {
      "entropy": 5.566104507446289,
      "epoch": 2.8188497822931784,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042094981495655596,
      "loss": 4.8265,
      "mean_token_accuracy": 0.2197411134839058,
      "num_tokens": 64497977.0,
      "step": 31075
    },
    {
      "entropy": 5.6225471019744875,
      "epoch": 2.8193033381712627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000420925187839145,
      "loss": 4.966,
      "mean_token_accuracy": 0.21211251765489578,
      "num_tokens": 64508767.0,
      "step": 31080
    },
    {
      "entropy": 5.55676589012146,
      "epoch": 2.819756894049347,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042090055770396383,
      "loss": 4.8432,
      "mean_token_accuracy": 0.21581576019525528,
      "num_tokens": 64519040.0,
      "step": 31085
    },
    {
      "entropy": 5.546733093261719,
      "epoch": 2.820210449927431,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042087592455152206,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22090413719415664,
      "num_tokens": 64529721.0,
      "step": 31090
    },
    {
      "entropy": 5.577895927429199,
      "epoch": 2.820664005805515,
      "grad_norm": 1.046875,
      "learning_rate": 0.000420851288382329,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20295707434415816,
      "num_tokens": 64541299.0,
      "step": 31095
    },
    {
      "entropy": 5.624560070037842,
      "epoch": 2.8211175616835993,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042082664919689406,
      "loss": 4.8865,
      "mean_token_accuracy": 0.21227426528930665,
      "num_tokens": 64550831.0,
      "step": 31100
    },
    {
      "entropy": 5.6072417259216305,
      "epoch": 2.8215711175616836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004208020069957269,
      "loss": 4.9273,
      "mean_token_accuracy": 0.20949650108814238,
      "num_tokens": 64562076.0,
      "step": 31105
    },
    {
      "entropy": 5.525624847412109,
      "epoch": 2.822024673439768,
      "grad_norm": 1.109375,
      "learning_rate": 0.000420777361779337,
      "loss": 4.9274,
      "mean_token_accuracy": 0.21800896972417833,
      "num_tokens": 64572955.0,
      "step": 31110
    },
    {
      "entropy": 5.642174816131591,
      "epoch": 2.822478229317852,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042075271354823427,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2185727134346962,
      "num_tokens": 64584376.0,
      "step": 31115
    },
    {
      "entropy": 5.6680704116821286,
      "epoch": 2.8229317851959363,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004207280623029281,
      "loss": 4.967,
      "mean_token_accuracy": 0.21231476217508316,
      "num_tokens": 64594522.0,
      "step": 31120
    },
    {
      "entropy": 5.633637762069702,
      "epoch": 2.82338534107402,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042070340804392853,
      "loss": 4.9713,
      "mean_token_accuracy": 0.22108199894428254,
      "num_tokens": 64604625.0,
      "step": 31125
    },
    {
      "entropy": 5.579609012603759,
      "epoch": 2.8238388969521044,
      "grad_norm": 1.25,
      "learning_rate": 0.0004206787507717454,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21223720610141755,
      "num_tokens": 64614184.0,
      "step": 31130
    },
    {
      "entropy": 5.68731722831726,
      "epoch": 2.8242924528301887,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042065409048688856,
      "loss": 5.0068,
      "mean_token_accuracy": 0.20688927471637725,
      "num_tokens": 64623019.0,
      "step": 31135
    },
    {
      "entropy": 5.616736316680909,
      "epoch": 2.824746008708273,
      "grad_norm": 1.21875,
      "learning_rate": 0.000420629427189868,
      "loss": 4.9311,
      "mean_token_accuracy": 0.212612546980381,
      "num_tokens": 64634017.0,
      "step": 31140
    },
    {
      "entropy": 5.579551267623901,
      "epoch": 2.825199564586357,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042060476088119376,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22012752294540405,
      "num_tokens": 64645007.0,
      "step": 31145
    },
    {
      "entropy": 5.660323905944824,
      "epoch": 2.825653120464441,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000420580091561376,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20943098068237304,
      "num_tokens": 64655705.0,
      "step": 31150
    },
    {
      "entropy": 5.705381870269775,
      "epoch": 2.8261066763425253,
      "grad_norm": 1.125,
      "learning_rate": 0.00042055541923092485,
      "loss": 5.0283,
      "mean_token_accuracy": 0.21066718846559523,
      "num_tokens": 64666094.0,
      "step": 31155
    },
    {
      "entropy": 5.672518348693847,
      "epoch": 2.8265602322206096,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004205307438903505,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20768718868494035,
      "num_tokens": 64676107.0,
      "step": 31160
    },
    {
      "entropy": 5.5879474639892575,
      "epoch": 2.827013788098694,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004205060655401633,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21730416864156724,
      "num_tokens": 64685846.0,
      "step": 31165
    },
    {
      "entropy": 5.548808813095093,
      "epoch": 2.827467343976778,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004204813841808736,
      "loss": 4.8909,
      "mean_token_accuracy": 0.21948938071727753,
      "num_tokens": 64695959.0,
      "step": 31170
    },
    {
      "entropy": 5.586927032470703,
      "epoch": 2.8279208998548624,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004204566998129917,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2189114898443222,
      "num_tokens": 64705594.0,
      "step": 31175
    },
    {
      "entropy": 5.636942720413208,
      "epoch": 2.828374455732946,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042043201243702834,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21315570473670958,
      "num_tokens": 64715974.0,
      "step": 31180
    },
    {
      "entropy": 5.584293985366822,
      "epoch": 2.8288280116110305,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004204073220534938,
      "loss": 4.9012,
      "mean_token_accuracy": 0.21684186160564423,
      "num_tokens": 64725846.0,
      "step": 31185
    },
    {
      "entropy": 5.567299652099609,
      "epoch": 2.8292815674891147,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004203826286628988,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2201531320810318,
      "num_tokens": 64735789.0,
      "step": 31190
    },
    {
      "entropy": 5.547651290893555,
      "epoch": 2.8297351233671986,
      "grad_norm": 1.171875,
      "learning_rate": 0.000420357932265754,
      "loss": 4.8443,
      "mean_token_accuracy": 0.22971422970294952,
      "num_tokens": 64745738.0,
      "step": 31195
    },
    {
      "entropy": 5.6054356575012205,
      "epoch": 2.830188679245283,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004203332328625701,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21167430728673936,
      "num_tokens": 64756523.0,
      "step": 31200
    },
    {
      "entropy": 5.6715209007263185,
      "epoch": 2.830642235123367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042030853045385796,
      "loss": 5.0277,
      "mean_token_accuracy": 0.20484247654676438,
      "num_tokens": 64767367.0,
      "step": 31205
    },
    {
      "entropy": 5.651865530014038,
      "epoch": 2.8310957910014514,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042028382504012837,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21473362445831298,
      "num_tokens": 64777729.0,
      "step": 31210
    },
    {
      "entropy": 5.709919738769531,
      "epoch": 2.8315493468795356,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042025911662189224,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2129921168088913,
      "num_tokens": 64788223.0,
      "step": 31215
    },
    {
      "entropy": 5.599450159072876,
      "epoch": 2.83200290275762,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042023440519966057,
      "loss": 4.928,
      "mean_token_accuracy": 0.21461839973926544,
      "num_tokens": 64797367.0,
      "step": 31220
    },
    {
      "entropy": 5.61387152671814,
      "epoch": 2.832456458635704,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042020969077394434,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2126474127173424,
      "num_tokens": 64807179.0,
      "step": 31225
    },
    {
      "entropy": 5.641021871566773,
      "epoch": 2.832910014513788,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042018497334525475,
      "loss": 4.862,
      "mean_token_accuracy": 0.20722534358501435,
      "num_tokens": 64817276.0,
      "step": 31230
    },
    {
      "entropy": 5.596399116516113,
      "epoch": 2.8333635703918723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004201602529141029,
      "loss": 4.945,
      "mean_token_accuracy": 0.21266809850931168,
      "num_tokens": 64827509.0,
      "step": 31235
    },
    {
      "entropy": 5.5908284187316895,
      "epoch": 2.8338171262699565,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042013552948100003,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2150601014494896,
      "num_tokens": 64836578.0,
      "step": 31240
    },
    {
      "entropy": 5.547596216201782,
      "epoch": 2.834270682148041,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004201108030464574,
      "loss": 4.8814,
      "mean_token_accuracy": 0.2227599397301674,
      "num_tokens": 64846518.0,
      "step": 31245
    },
    {
      "entropy": 5.682690858840942,
      "epoch": 2.8347242380261246,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042008607361098646,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2164894625544548,
      "num_tokens": 64856304.0,
      "step": 31250
    },
    {
      "entropy": 5.5925287246704105,
      "epoch": 2.835177793904209,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004200613411750985,
      "loss": 4.8492,
      "mean_token_accuracy": 0.2263615161180496,
      "num_tokens": 64867381.0,
      "step": 31255
    },
    {
      "entropy": 5.602374029159546,
      "epoch": 2.835631349782293,
      "grad_norm": 1.25,
      "learning_rate": 0.000420036605739305,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2157713457942009,
      "num_tokens": 64877306.0,
      "step": 31260
    },
    {
      "entropy": 5.516693449020385,
      "epoch": 2.8360849056603774,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004200118673041176,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2164098784327507,
      "num_tokens": 64886932.0,
      "step": 31265
    },
    {
      "entropy": 5.62765588760376,
      "epoch": 2.8365384615384617,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041998712587004776,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21025226265192032,
      "num_tokens": 64897387.0,
      "step": 31270
    },
    {
      "entropy": 5.675957298278808,
      "epoch": 2.836992017416546,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041996238143760726,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21875430792570114,
      "num_tokens": 64907251.0,
      "step": 31275
    },
    {
      "entropy": 5.583954572677612,
      "epoch": 2.8374455732946298,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041993763400730776,
      "loss": 4.913,
      "mean_token_accuracy": 0.21687745451927185,
      "num_tokens": 64917300.0,
      "step": 31280
    },
    {
      "entropy": 5.536740446090699,
      "epoch": 2.837899129172714,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004199128835796611,
      "loss": 4.8714,
      "mean_token_accuracy": 0.22533662021160125,
      "num_tokens": 64927133.0,
      "step": 31285
    },
    {
      "entropy": 5.578914594650269,
      "epoch": 2.8383526850507983,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000419888130155179,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21498076468706132,
      "num_tokens": 64937789.0,
      "step": 31290
    },
    {
      "entropy": 5.670873975753784,
      "epoch": 2.8388062409288826,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004198633737343735,
      "loss": 4.952,
      "mean_token_accuracy": 0.21623307466506958,
      "num_tokens": 64948127.0,
      "step": 31295
    },
    {
      "entropy": 5.650634765625,
      "epoch": 2.8392597968069664,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004198386143177565,
      "loss": 4.9908,
      "mean_token_accuracy": 0.21361104398965836,
      "num_tokens": 64959632.0,
      "step": 31300
    },
    {
      "entropy": 5.507171535491944,
      "epoch": 2.8397133526850507,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041981385190584003,
      "loss": 4.7715,
      "mean_token_accuracy": 0.2209300711750984,
      "num_tokens": 64970411.0,
      "step": 31305
    },
    {
      "entropy": 5.551486968994141,
      "epoch": 2.840166908563135,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004197890864991362,
      "loss": 4.9031,
      "mean_token_accuracy": 0.21426456719636916,
      "num_tokens": 64980748.0,
      "step": 31310
    },
    {
      "entropy": 5.561504077911377,
      "epoch": 2.840620464441219,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004197643180981572,
      "loss": 4.8534,
      "mean_token_accuracy": 0.21824093163013458,
      "num_tokens": 64990818.0,
      "step": 31315
    },
    {
      "entropy": 5.600249576568603,
      "epoch": 2.8410740203193035,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004197395467034152,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21965514421463012,
      "num_tokens": 65001993.0,
      "step": 31320
    },
    {
      "entropy": 5.558138227462768,
      "epoch": 2.8415275761973877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004197147723154225,
      "loss": 4.886,
      "mean_token_accuracy": 0.22362725287675858,
      "num_tokens": 65013007.0,
      "step": 31325
    },
    {
      "entropy": 5.664604043960571,
      "epoch": 2.8419811320754715,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004196899949346914,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2142934888601303,
      "num_tokens": 65023595.0,
      "step": 31330
    },
    {
      "entropy": 5.600962066650391,
      "epoch": 2.842434687953556,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004196652145617343,
      "loss": 4.8959,
      "mean_token_accuracy": 0.21897254139184952,
      "num_tokens": 65033631.0,
      "step": 31335
    },
    {
      "entropy": 5.672798156738281,
      "epoch": 2.84288824383164,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004196404311970638,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2126334637403488,
      "num_tokens": 65044059.0,
      "step": 31340
    },
    {
      "entropy": 5.59052209854126,
      "epoch": 2.8433417997097243,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004196156448411922,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21347975581884385,
      "num_tokens": 65053931.0,
      "step": 31345
    },
    {
      "entropy": 5.591944217681885,
      "epoch": 2.843795355587808,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041959085549463234,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21769054532051085,
      "num_tokens": 65064011.0,
      "step": 31350
    },
    {
      "entropy": 5.654287385940552,
      "epoch": 2.8442489114658924,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004195660631578966,
      "loss": 4.9885,
      "mean_token_accuracy": 0.21189521700143815,
      "num_tokens": 65074417.0,
      "step": 31355
    },
    {
      "entropy": 5.596863317489624,
      "epoch": 2.8447024673439767,
      "grad_norm": 1.1875,
      "learning_rate": 0.000419541267831498,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22106014043092728,
      "num_tokens": 65084593.0,
      "step": 31360
    },
    {
      "entropy": 5.57124547958374,
      "epoch": 2.845156023222061,
      "grad_norm": 1.171875,
      "learning_rate": 0.000419516469515949,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22277182787656785,
      "num_tokens": 65095047.0,
      "step": 31365
    },
    {
      "entropy": 5.631705808639526,
      "epoch": 2.8456095791001452,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041949166821176264,
      "loss": 5.0032,
      "mean_token_accuracy": 0.2164821818470955,
      "num_tokens": 65104434.0,
      "step": 31370
    },
    {
      "entropy": 5.678917980194091,
      "epoch": 2.8460631349782295,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041946686391945183,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21663412749767302,
      "num_tokens": 65114039.0,
      "step": 31375
    },
    {
      "entropy": 5.71366171836853,
      "epoch": 2.8465166908563138,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004194420566395294,
      "loss": 5.0491,
      "mean_token_accuracy": 0.20536773651838303,
      "num_tokens": 65125387.0,
      "step": 31380
    },
    {
      "entropy": 5.637423467636109,
      "epoch": 2.8469702467343976,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004194172463725084,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2191828489303589,
      "num_tokens": 65136351.0,
      "step": 31385
    },
    {
      "entropy": 5.585445976257324,
      "epoch": 2.847423802612482,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041939243311890195,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21333178132772446,
      "num_tokens": 65147343.0,
      "step": 31390
    },
    {
      "entropy": 5.546621465682984,
      "epoch": 2.847877358490566,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041936761687922324,
      "loss": 4.8875,
      "mean_token_accuracy": 0.222720368206501,
      "num_tokens": 65157097.0,
      "step": 31395
    },
    {
      "entropy": 5.605692672729492,
      "epoch": 2.8483309143686504,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041934279765398535,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21638421565294266,
      "num_tokens": 65167546.0,
      "step": 31400
    },
    {
      "entropy": 5.668427658081055,
      "epoch": 2.848784470246734,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004193179754437017,
      "loss": 4.9468,
      "mean_token_accuracy": 0.20696936845779418,
      "num_tokens": 65178497.0,
      "step": 31405
    },
    {
      "entropy": 5.586981344223022,
      "epoch": 2.8492380261248185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004192931502488855,
      "loss": 4.8505,
      "mean_token_accuracy": 0.22228772193193436,
      "num_tokens": 65189530.0,
      "step": 31410
    },
    {
      "entropy": 5.60063591003418,
      "epoch": 2.8496915820029027,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041926832207005025,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21276939660310745,
      "num_tokens": 65199836.0,
      "step": 31415
    },
    {
      "entropy": 5.507954788208008,
      "epoch": 2.850145137880987,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004192434909077092,
      "loss": 4.7683,
      "mean_token_accuracy": 0.22527859956026078,
      "num_tokens": 65210208.0,
      "step": 31420
    },
    {
      "entropy": 5.583266258239746,
      "epoch": 2.8505986937590713,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041921865676237605,
      "loss": 4.9268,
      "mean_token_accuracy": 0.21439330875873566,
      "num_tokens": 65220954.0,
      "step": 31425
    },
    {
      "entropy": 5.644137525558472,
      "epoch": 2.8510522496371555,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041919381963456437,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21211770176887512,
      "num_tokens": 65230870.0,
      "step": 31430
    },
    {
      "entropy": 5.635395431518555,
      "epoch": 2.8515058055152394,
      "grad_norm": 1.25,
      "learning_rate": 0.0004191689795247877,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22319147288799285,
      "num_tokens": 65241913.0,
      "step": 31435
    },
    {
      "entropy": 5.688692855834961,
      "epoch": 2.8519593613933236,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004191441364335598,
      "loss": 4.9308,
      "mean_token_accuracy": 0.21712172031402588,
      "num_tokens": 65252513.0,
      "step": 31440
    },
    {
      "entropy": 5.554673337936402,
      "epoch": 2.852412917271408,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041911929036139457,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21490330696105958,
      "num_tokens": 65262695.0,
      "step": 31445
    },
    {
      "entropy": 5.639480209350586,
      "epoch": 2.852866473149492,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041909444130880553,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2159951239824295,
      "num_tokens": 65274131.0,
      "step": 31450
    },
    {
      "entropy": 5.599239158630371,
      "epoch": 2.853320029027576,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041906958927630673,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2150556668639183,
      "num_tokens": 65285080.0,
      "step": 31455
    },
    {
      "entropy": 5.735603189468383,
      "epoch": 2.8537735849056602,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004190447342644122,
      "loss": 5.1101,
      "mean_token_accuracy": 0.2141365796327591,
      "num_tokens": 65296512.0,
      "step": 31460
    },
    {
      "entropy": 5.5775209903717045,
      "epoch": 2.8542271407837445,
      "grad_norm": 1.25,
      "learning_rate": 0.0004190198762736358,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2224300131201744,
      "num_tokens": 65306054.0,
      "step": 31465
    },
    {
      "entropy": 5.547127199172974,
      "epoch": 2.8546806966618288,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004189950153044917,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21349292397499084,
      "num_tokens": 65315046.0,
      "step": 31470
    },
    {
      "entropy": 5.581344699859619,
      "epoch": 2.855134252539913,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004189701513574939,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21237158179283142,
      "num_tokens": 65325482.0,
      "step": 31475
    },
    {
      "entropy": 5.564690732955933,
      "epoch": 2.8555878084179973,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004189452844331567,
      "loss": 4.8583,
      "mean_token_accuracy": 0.21760592311620713,
      "num_tokens": 65335518.0,
      "step": 31480
    },
    {
      "entropy": 5.601111650466919,
      "epoch": 2.856041364296081,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041892041453199443,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2173466458916664,
      "num_tokens": 65346079.0,
      "step": 31485
    },
    {
      "entropy": 5.560744714736939,
      "epoch": 2.8564949201741654,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004188955416545212,
      "loss": 4.8938,
      "mean_token_accuracy": 0.2206578239798546,
      "num_tokens": 65356719.0,
      "step": 31490
    },
    {
      "entropy": 5.647084045410156,
      "epoch": 2.8569484760522497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004188706658012516,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2081117421388626,
      "num_tokens": 65367877.0,
      "step": 31495
    },
    {
      "entropy": 5.680764293670654,
      "epoch": 2.857402031930334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041884578697269994,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20874454826116562,
      "num_tokens": 65379624.0,
      "step": 31500
    },
    {
      "entropy": 5.629879760742187,
      "epoch": 2.8578555878084178,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004188209051693807,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21204937994480133,
      "num_tokens": 65390417.0,
      "step": 31505
    },
    {
      "entropy": 5.501463937759399,
      "epoch": 2.858309143686502,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041879602039180856,
      "loss": 4.7681,
      "mean_token_accuracy": 0.2275952711701393,
      "num_tokens": 65401681.0,
      "step": 31510
    },
    {
      "entropy": 5.556026792526245,
      "epoch": 2.8587626995645863,
      "grad_norm": 1.25,
      "learning_rate": 0.00041877113264049805,
      "loss": 4.8873,
      "mean_token_accuracy": 0.21970682591199875,
      "num_tokens": 65412520.0,
      "step": 31515
    },
    {
      "entropy": 5.617603731155396,
      "epoch": 2.8592162554426706,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041874624191596383,
      "loss": 4.9313,
      "mean_token_accuracy": 0.21787261962890625,
      "num_tokens": 65423201.0,
      "step": 31520
    },
    {
      "entropy": 5.57971715927124,
      "epoch": 2.859669811320755,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004187213482187207,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21197264790534973,
      "num_tokens": 65432778.0,
      "step": 31525
    },
    {
      "entropy": 5.6699371337890625,
      "epoch": 2.860123367198839,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004186964515492835,
      "loss": 4.947,
      "mean_token_accuracy": 0.21519640684127808,
      "num_tokens": 65442323.0,
      "step": 31530
    },
    {
      "entropy": 5.60116810798645,
      "epoch": 2.8605769230769234,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004186715519081671,
      "loss": 4.9314,
      "mean_token_accuracy": 0.21230805218219756,
      "num_tokens": 65451980.0,
      "step": 31535
    },
    {
      "entropy": 5.5135551452636715,
      "epoch": 2.861030478955007,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004186466492958864,
      "loss": 4.8813,
      "mean_token_accuracy": 0.21719976216554643,
      "num_tokens": 65463359.0,
      "step": 31540
    },
    {
      "entropy": 5.607174587249756,
      "epoch": 2.8614840348330914,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004186217437129562,
      "loss": 4.9327,
      "mean_token_accuracy": 0.21512628197669983,
      "num_tokens": 65474155.0,
      "step": 31545
    },
    {
      "entropy": 5.631038093566895,
      "epoch": 2.8619375907111757,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041859683515989196,
      "loss": 4.8724,
      "mean_token_accuracy": 0.21913151890039445,
      "num_tokens": 65484884.0,
      "step": 31550
    },
    {
      "entropy": 5.630547952651978,
      "epoch": 2.8623911465892595,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041857192363720845,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21555111408233643,
      "num_tokens": 65494380.0,
      "step": 31555
    },
    {
      "entropy": 5.585837507247925,
      "epoch": 2.862844702467344,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004185470091454211,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21548750698566438,
      "num_tokens": 65505735.0,
      "step": 31560
    },
    {
      "entropy": 5.61636209487915,
      "epoch": 2.863298258345428,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004185220916850449,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2144456148147583,
      "num_tokens": 65515956.0,
      "step": 31565
    },
    {
      "entropy": 5.596684074401855,
      "epoch": 2.8637518142235123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041849717125659537,
      "loss": 4.8554,
      "mean_token_accuracy": 0.218240687251091,
      "num_tokens": 65526369.0,
      "step": 31570
    },
    {
      "entropy": 5.603471994400024,
      "epoch": 2.8642053701015966,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004184722478605877,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22364125698804854,
      "num_tokens": 65536032.0,
      "step": 31575
    },
    {
      "entropy": 5.567645168304443,
      "epoch": 2.864658925979681,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004184473214975374,
      "loss": 4.9614,
      "mean_token_accuracy": 0.20941196829080583,
      "num_tokens": 65546121.0,
      "step": 31580
    },
    {
      "entropy": 5.606190538406372,
      "epoch": 2.865112481857765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041842239216795994,
      "loss": 4.9045,
      "mean_token_accuracy": 0.21521912068128585,
      "num_tokens": 65555915.0,
      "step": 31585
    },
    {
      "entropy": 5.608984661102295,
      "epoch": 2.865566037735849,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041839745987237086,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20711799263954161,
      "num_tokens": 65566267.0,
      "step": 31590
    },
    {
      "entropy": 5.596514320373535,
      "epoch": 2.866019593613933,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004183725246112857,
      "loss": 4.914,
      "mean_token_accuracy": 0.22463058084249496,
      "num_tokens": 65577126.0,
      "step": 31595
    },
    {
      "entropy": 5.6551513671875,
      "epoch": 2.8664731494920175,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041834758638522034,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20194878727197646,
      "num_tokens": 65587158.0,
      "step": 31600
    },
    {
      "entropy": 5.6248420715332035,
      "epoch": 2.8669267053701017,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041832264519469026,
      "loss": 4.9044,
      "mean_token_accuracy": 0.2107012987136841,
      "num_tokens": 65597720.0,
      "step": 31605
    },
    {
      "entropy": 5.689760780334472,
      "epoch": 2.8673802612481856,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041829770104021143,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21968888491392136,
      "num_tokens": 65607409.0,
      "step": 31610
    },
    {
      "entropy": 5.618808603286743,
      "epoch": 2.86783381712627,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041827275392229956,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22368093729019164,
      "num_tokens": 65616785.0,
      "step": 31615
    },
    {
      "entropy": 5.604651355743409,
      "epoch": 2.868287373004354,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004182478038414706,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20952381938695908,
      "num_tokens": 65625794.0,
      "step": 31620
    },
    {
      "entropy": 5.526165246963501,
      "epoch": 2.8687409288824384,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041822285079824056,
      "loss": 4.8121,
      "mean_token_accuracy": 0.2223937079310417,
      "num_tokens": 65635789.0,
      "step": 31625
    },
    {
      "entropy": 5.644211196899414,
      "epoch": 2.8691944847605226,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004181978947931256,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20847029089927674,
      "num_tokens": 65646266.0,
      "step": 31630
    },
    {
      "entropy": 5.627982759475708,
      "epoch": 2.869648040638607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004181729358266415,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21238478124141694,
      "num_tokens": 65656538.0,
      "step": 31635
    },
    {
      "entropy": 5.680902433395386,
      "epoch": 2.8701015965166907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041814797389930474,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2185475692152977,
      "num_tokens": 65666926.0,
      "step": 31640
    },
    {
      "entropy": 5.621043920516968,
      "epoch": 2.870555152394775,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004181230090116313,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2103895977139473,
      "num_tokens": 65677219.0,
      "step": 31645
    },
    {
      "entropy": 5.589084482192993,
      "epoch": 2.8710087082728593,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041809804116413763,
      "loss": 4.9482,
      "mean_token_accuracy": 0.21239991039037703,
      "num_tokens": 65688281.0,
      "step": 31650
    },
    {
      "entropy": 5.735374021530151,
      "epoch": 2.8714622641509435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041807307035734,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20635735392570495,
      "num_tokens": 65698815.0,
      "step": 31655
    },
    {
      "entropy": 5.630136966705322,
      "epoch": 2.8719158200290273,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004180480965917548,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21364252865314484,
      "num_tokens": 65709481.0,
      "step": 31660
    },
    {
      "entropy": 5.66035270690918,
      "epoch": 2.8723693759071116,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041802311986789854,
      "loss": 4.9308,
      "mean_token_accuracy": 0.20607939064502717,
      "num_tokens": 65720561.0,
      "step": 31665
    },
    {
      "entropy": 5.653276872634888,
      "epoch": 2.872822931785196,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041799814018628765,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21045362800359727,
      "num_tokens": 65730384.0,
      "step": 31670
    },
    {
      "entropy": 5.612490224838257,
      "epoch": 2.87327648766328,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041797315754743883,
      "loss": 4.936,
      "mean_token_accuracy": 0.21761937886476518,
      "num_tokens": 65740711.0,
      "step": 31675
    },
    {
      "entropy": 5.704002571105957,
      "epoch": 2.8737300435413644,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041794817195186867,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20669930875301362,
      "num_tokens": 65751033.0,
      "step": 31680
    },
    {
      "entropy": 5.688816356658935,
      "epoch": 2.8741835994194487,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004179231834000939,
      "loss": 5.013,
      "mean_token_accuracy": 0.20879815518856049,
      "num_tokens": 65761500.0,
      "step": 31685
    },
    {
      "entropy": 5.560088014602661,
      "epoch": 2.8746371552975325,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041789819189263117,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22089462727308273,
      "num_tokens": 65771095.0,
      "step": 31690
    },
    {
      "entropy": 5.578502655029297,
      "epoch": 2.8750907111756168,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041787319742999756,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21886488795280457,
      "num_tokens": 65782058.0,
      "step": 31695
    },
    {
      "entropy": 5.593817472457886,
      "epoch": 2.875544267053701,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004178482000127098,
      "loss": 4.9596,
      "mean_token_accuracy": 0.2222698673605919,
      "num_tokens": 65791710.0,
      "step": 31700
    },
    {
      "entropy": 5.639909410476685,
      "epoch": 2.8759978229317853,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004178231996412848,
      "loss": 5.0005,
      "mean_token_accuracy": 0.20786140561103822,
      "num_tokens": 65802180.0,
      "step": 31705
    },
    {
      "entropy": 5.5874964714050295,
      "epoch": 2.876451378809869,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004177981963162396,
      "loss": 4.8637,
      "mean_token_accuracy": 0.22670624703168868,
      "num_tokens": 65812998.0,
      "step": 31710
    },
    {
      "entropy": 5.600241851806641,
      "epoch": 2.8769049346879534,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004177731900380914,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21553351879119872,
      "num_tokens": 65824210.0,
      "step": 31715
    },
    {
      "entropy": 5.5797172546386715,
      "epoch": 2.8773584905660377,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004177481808073572,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22095848470926285,
      "num_tokens": 65834322.0,
      "step": 31720
    },
    {
      "entropy": 5.594752216339112,
      "epoch": 2.877812046444122,
      "grad_norm": 1.125,
      "learning_rate": 0.00041772316862455426,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2244741439819336,
      "num_tokens": 65845347.0,
      "step": 31725
    },
    {
      "entropy": 5.640364313125611,
      "epoch": 2.878265602322206,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041769815349019986,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20914436727762223,
      "num_tokens": 65855616.0,
      "step": 31730
    },
    {
      "entropy": 5.650536251068115,
      "epoch": 2.8787191582002905,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00041767313540481125,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2161816403269768,
      "num_tokens": 65865542.0,
      "step": 31735
    },
    {
      "entropy": 5.592294454574585,
      "epoch": 2.8791727140783747,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004176481143689058,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21320167630910875,
      "num_tokens": 65874478.0,
      "step": 31740
    },
    {
      "entropy": 5.651920795440674,
      "epoch": 2.8796262699564585,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004176230903830011,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2180263802409172,
      "num_tokens": 65884609.0,
      "step": 31745
    },
    {
      "entropy": 5.642696285247803,
      "epoch": 2.880079825834543,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004175980634476145,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21794142872095107,
      "num_tokens": 65894486.0,
      "step": 31750
    },
    {
      "entropy": 5.576766109466552,
      "epoch": 2.880533381712627,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004175730335632635,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2173640325665474,
      "num_tokens": 65904643.0,
      "step": 31755
    },
    {
      "entropy": 5.522447299957276,
      "epoch": 2.8809869375907113,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041754800073046593,
      "loss": 4.8953,
      "mean_token_accuracy": 0.21958731561899186,
      "num_tokens": 65914107.0,
      "step": 31760
    },
    {
      "entropy": 5.579465913772583,
      "epoch": 2.881440493468795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004175229649497394,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2164247527718544,
      "num_tokens": 65925112.0,
      "step": 31765
    },
    {
      "entropy": 5.684026718139648,
      "epoch": 2.8818940493468794,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004174979262216016,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21679816395044327,
      "num_tokens": 65934624.0,
      "step": 31770
    },
    {
      "entropy": 5.6303047180175785,
      "epoch": 2.8823476052249637,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004174728845465704,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21469669044017792,
      "num_tokens": 65945951.0,
      "step": 31775
    },
    {
      "entropy": 5.533226585388183,
      "epoch": 2.882801161103048,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004174478399251636,
      "loss": 4.8376,
      "mean_token_accuracy": 0.22649696320295334,
      "num_tokens": 65956871.0,
      "step": 31780
    },
    {
      "entropy": 5.694487047195435,
      "epoch": 2.8832547169811322,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004174227923578992,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2127063751220703,
      "num_tokens": 65967814.0,
      "step": 31785
    },
    {
      "entropy": 5.6321759700775145,
      "epoch": 2.8837082728592165,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004173977418452952,
      "loss": 4.945,
      "mean_token_accuracy": 0.20828794687986374,
      "num_tokens": 65978118.0,
      "step": 31790
    },
    {
      "entropy": 5.570476722717285,
      "epoch": 2.8841618287373003,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041737268838786955,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2200121432542801,
      "num_tokens": 65988240.0,
      "step": 31795
    },
    {
      "entropy": 5.608204889297485,
      "epoch": 2.8846153846153846,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004173476319861404,
      "loss": 4.876,
      "mean_token_accuracy": 0.21797575950622558,
      "num_tokens": 65998126.0,
      "step": 31800
    },
    {
      "entropy": 5.619529390335083,
      "epoch": 2.885068940493469,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004173225726406261,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22170098423957824,
      "num_tokens": 66009118.0,
      "step": 31805
    },
    {
      "entropy": 5.641887760162353,
      "epoch": 2.885522496371553,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041729751035184457,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2094158411026001,
      "num_tokens": 66019340.0,
      "step": 31810
    },
    {
      "entropy": 5.611714315414429,
      "epoch": 2.885976052249637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041727244512031425,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2167496055364609,
      "num_tokens": 66029784.0,
      "step": 31815
    },
    {
      "entropy": 5.6632606983184814,
      "epoch": 2.886429608127721,
      "grad_norm": 1.25,
      "learning_rate": 0.0004172473769465536,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2013155400753021,
      "num_tokens": 66039987.0,
      "step": 31820
    },
    {
      "entropy": 5.6277772903442385,
      "epoch": 2.8868831640058055,
      "grad_norm": 1.25,
      "learning_rate": 0.00041722230583108097,
      "loss": 4.965,
      "mean_token_accuracy": 0.2157362937927246,
      "num_tokens": 66049150.0,
      "step": 31825
    },
    {
      "entropy": 5.64697790145874,
      "epoch": 2.8873367198838897,
      "grad_norm": 1.125,
      "learning_rate": 0.00041719723177441467,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21378494501113893,
      "num_tokens": 66060719.0,
      "step": 31830
    },
    {
      "entropy": 5.682665681838989,
      "epoch": 2.887790275761974,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004171721547770735,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21503500044345855,
      "num_tokens": 66071126.0,
      "step": 31835
    },
    {
      "entropy": 5.635784912109375,
      "epoch": 2.8882438316400583,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004171470748395759,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22183056324720382,
      "num_tokens": 66080928.0,
      "step": 31840
    },
    {
      "entropy": 5.652846622467041,
      "epoch": 2.888697387518142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004171219919624406,
      "loss": 4.916,
      "mean_token_accuracy": 0.21347257643938064,
      "num_tokens": 66090900.0,
      "step": 31845
    },
    {
      "entropy": 5.608264684677124,
      "epoch": 2.8891509433962264,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041709690614618624,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2088766023516655,
      "num_tokens": 66100720.0,
      "step": 31850
    },
    {
      "entropy": 5.587069129943847,
      "epoch": 2.8896044992743106,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041707181739133163,
      "loss": 4.9584,
      "mean_token_accuracy": 0.21069999933242797,
      "num_tokens": 66110780.0,
      "step": 31855
    },
    {
      "entropy": 5.633911228179931,
      "epoch": 2.890058055152395,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004170467256983957,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22332895696163177,
      "num_tokens": 66120976.0,
      "step": 31860
    },
    {
      "entropy": 5.641289472579956,
      "epoch": 2.8905116110304787,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004170216310678973,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22116368561983107,
      "num_tokens": 66130221.0,
      "step": 31865
    },
    {
      "entropy": 5.573739147186279,
      "epoch": 2.890965166908563,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041699653350035527,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2218844324350357,
      "num_tokens": 66140060.0,
      "step": 31870
    },
    {
      "entropy": 5.578892850875855,
      "epoch": 2.8914187227866472,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041697143299628886,
      "loss": 4.9294,
      "mean_token_accuracy": 0.2219777524471283,
      "num_tokens": 66151640.0,
      "step": 31875
    },
    {
      "entropy": 5.565376663208008,
      "epoch": 2.8918722786647315,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041694632955621697,
      "loss": 4.9001,
      "mean_token_accuracy": 0.22564877718687057,
      "num_tokens": 66162923.0,
      "step": 31880
    },
    {
      "entropy": 5.629379320144653,
      "epoch": 2.8923258345428158,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041692122318065885,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2127934455871582,
      "num_tokens": 66173108.0,
      "step": 31885
    },
    {
      "entropy": 5.608965492248535,
      "epoch": 2.8927793904209,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041689611387013363,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22055374085903168,
      "num_tokens": 66185280.0,
      "step": 31890
    },
    {
      "entropy": 5.562539577484131,
      "epoch": 2.8932329462989843,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004168710016251606,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2128838062286377,
      "num_tokens": 66196020.0,
      "step": 31895
    },
    {
      "entropy": 5.586726188659668,
      "epoch": 2.893686502177068,
      "grad_norm": 1.125,
      "learning_rate": 0.0004168458864462592,
      "loss": 4.848,
      "mean_token_accuracy": 0.21899206191301346,
      "num_tokens": 66205753.0,
      "step": 31900
    },
    {
      "entropy": 5.64402961730957,
      "epoch": 2.8941400580551524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041682076833394864,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21260420382022857,
      "num_tokens": 66218415.0,
      "step": 31905
    },
    {
      "entropy": 5.634943580627441,
      "epoch": 2.8945936139332367,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041679564728874845,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21050702184438705,
      "num_tokens": 66228090.0,
      "step": 31910
    },
    {
      "entropy": 5.563312005996704,
      "epoch": 2.8950471698113205,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004167705233111782,
      "loss": 4.865,
      "mean_token_accuracy": 0.22121773660182953,
      "num_tokens": 66237587.0,
      "step": 31915
    },
    {
      "entropy": 5.60017762184143,
      "epoch": 2.8955007256894048,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041674539640175733,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2185591533780098,
      "num_tokens": 66248162.0,
      "step": 31920
    },
    {
      "entropy": 5.585758066177368,
      "epoch": 2.895954281567489,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004167202665610055,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22525439858436586,
      "num_tokens": 66258370.0,
      "step": 31925
    },
    {
      "entropy": 5.600293588638306,
      "epoch": 2.8964078374455733,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041669513378944254,
      "loss": 4.9447,
      "mean_token_accuracy": 0.21419663578271866,
      "num_tokens": 66269267.0,
      "step": 31930
    },
    {
      "entropy": 5.6684362411499025,
      "epoch": 2.8968613933236576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041666999808758815,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21446811705827712,
      "num_tokens": 66280077.0,
      "step": 31935
    },
    {
      "entropy": 5.641676902770996,
      "epoch": 2.897314949201742,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000416644859455962,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22321052700281144,
      "num_tokens": 66291678.0,
      "step": 31940
    },
    {
      "entropy": 5.592736625671387,
      "epoch": 2.897768505079826,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004166197178950841,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21754908114671706,
      "num_tokens": 66302046.0,
      "step": 31945
    },
    {
      "entropy": 5.584465217590332,
      "epoch": 2.89822206095791,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041659457340547436,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21626962572336197,
      "num_tokens": 66311888.0,
      "step": 31950
    },
    {
      "entropy": 5.574025344848633,
      "epoch": 2.898675616835994,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041656942598765275,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2203891918063164,
      "num_tokens": 66321507.0,
      "step": 31955
    },
    {
      "entropy": 5.544738721847534,
      "epoch": 2.8991291727140784,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041654427564213924,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22370833903551102,
      "num_tokens": 66331458.0,
      "step": 31960
    },
    {
      "entropy": 5.6023890495300295,
      "epoch": 2.8995827285921627,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041651912236945416,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22036738842725753,
      "num_tokens": 66340875.0,
      "step": 31965
    },
    {
      "entropy": 5.642870759963989,
      "epoch": 2.9000362844702465,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041649396617011753,
      "loss": 4.9819,
      "mean_token_accuracy": 0.20874171257019042,
      "num_tokens": 66350784.0,
      "step": 31970
    },
    {
      "entropy": 5.6391205310821535,
      "epoch": 2.900489840348331,
      "grad_norm": 1.125,
      "learning_rate": 0.0004164688070446496,
      "loss": 4.9899,
      "mean_token_accuracy": 0.21138345450162888,
      "num_tokens": 66361591.0,
      "step": 31975
    },
    {
      "entropy": 5.576123046875,
      "epoch": 2.900943396226415,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004164436449935707,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22128966003656386,
      "num_tokens": 66371540.0,
      "step": 31980
    },
    {
      "entropy": 5.618195819854736,
      "epoch": 2.9013969521044993,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041641848001740126,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21928317099809647,
      "num_tokens": 66382053.0,
      "step": 31985
    },
    {
      "entropy": 5.593534517288208,
      "epoch": 2.9018505079825836,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004163933121166615,
      "loss": 4.9014,
      "mean_token_accuracy": 0.21862017661333083,
      "num_tokens": 66392563.0,
      "step": 31990
    },
    {
      "entropy": 5.583109045028687,
      "epoch": 2.902304063860668,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000416368141291872,
      "loss": 4.8749,
      "mean_token_accuracy": 0.2170826643705368,
      "num_tokens": 66403582.0,
      "step": 31995
    },
    {
      "entropy": 5.525620412826538,
      "epoch": 2.9027576197387517,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041634296754355346,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22272140830755233,
      "num_tokens": 66413713.0,
      "step": 32000
    },
    {
      "entropy": 5.582666110992432,
      "epoch": 2.903211175616836,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004163177908722262,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2152288094162941,
      "num_tokens": 66424802.0,
      "step": 32005
    },
    {
      "entropy": 5.600276708602905,
      "epoch": 2.90366473149492,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041629261127841105,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22269952297210693,
      "num_tokens": 66434716.0,
      "step": 32010
    },
    {
      "entropy": 5.5774510383605955,
      "epoch": 2.9041182873730045,
      "grad_norm": 1.25,
      "learning_rate": 0.00041626742876262876,
      "loss": 4.9347,
      "mean_token_accuracy": 0.2132781520485878,
      "num_tokens": 66445725.0,
      "step": 32015
    },
    {
      "entropy": 5.614554500579834,
      "epoch": 2.9045718432510883,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041624224332539997,
      "loss": 4.9759,
      "mean_token_accuracy": 0.20886536538600922,
      "num_tokens": 66457492.0,
      "step": 32020
    },
    {
      "entropy": 5.6291656494140625,
      "epoch": 2.9050253991291726,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004162170549672455,
      "loss": 4.8909,
      "mean_token_accuracy": 0.21748114973306656,
      "num_tokens": 66467490.0,
      "step": 32025
    },
    {
      "entropy": 5.626606416702271,
      "epoch": 2.905478955007257,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004161918636886865,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21209583431482315,
      "num_tokens": 66478160.0,
      "step": 32030
    },
    {
      "entropy": 5.669289064407349,
      "epoch": 2.905932510885341,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004161666694902437,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21435584425926207,
      "num_tokens": 66488181.0,
      "step": 32035
    },
    {
      "entropy": 5.653158140182495,
      "epoch": 2.9063860667634254,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004161414723724383,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22207406908273697,
      "num_tokens": 66498507.0,
      "step": 32040
    },
    {
      "entropy": 5.5459442138671875,
      "epoch": 2.9068396226415096,
      "grad_norm": 1.125,
      "learning_rate": 0.0004161162723357912,
      "loss": 4.8466,
      "mean_token_accuracy": 0.22229903638362886,
      "num_tokens": 66507752.0,
      "step": 32045
    },
    {
      "entropy": 5.5540379047393795,
      "epoch": 2.9072931785195935,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004160910693808236,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21004885584115982,
      "num_tokens": 66519360.0,
      "step": 32050
    },
    {
      "entropy": 5.659834241867065,
      "epoch": 2.9077467343976777,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041606586350805686,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2120181292295456,
      "num_tokens": 66529790.0,
      "step": 32055
    },
    {
      "entropy": 5.6144193649292,
      "epoch": 2.908200290275762,
      "grad_norm": 1.15625,
      "learning_rate": 0.000416040654718012,
      "loss": 4.8676,
      "mean_token_accuracy": 0.22535613626241685,
      "num_tokens": 66539617.0,
      "step": 32060
    },
    {
      "entropy": 5.585898780822754,
      "epoch": 2.9086538461538463,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041601544301121045,
      "loss": 4.8739,
      "mean_token_accuracy": 0.2196207672357559,
      "num_tokens": 66549861.0,
      "step": 32065
    },
    {
      "entropy": 5.672889518737793,
      "epoch": 2.90910740203193,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004159902283881737,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21636117845773697,
      "num_tokens": 66560388.0,
      "step": 32070
    },
    {
      "entropy": 5.690124416351319,
      "epoch": 2.9095609579100143,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000415965010849423,
      "loss": 4.9878,
      "mean_token_accuracy": 0.21009019166231155,
      "num_tokens": 66570795.0,
      "step": 32075
    },
    {
      "entropy": 5.645208120346069,
      "epoch": 2.9100145137880986,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041593979039548,
      "loss": 4.9451,
      "mean_token_accuracy": 0.20885093361139298,
      "num_tokens": 66580268.0,
      "step": 32080
    },
    {
      "entropy": 5.612254428863525,
      "epoch": 2.910468069666183,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004159145670268662,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21114560663700105,
      "num_tokens": 66591192.0,
      "step": 32085
    },
    {
      "entropy": 5.623955774307251,
      "epoch": 2.910921625544267,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004158893407441033,
      "loss": 4.9968,
      "mean_token_accuracy": 0.21587304323911666,
      "num_tokens": 66600768.0,
      "step": 32090
    },
    {
      "entropy": 5.716529178619385,
      "epoch": 2.9113751814223514,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041586411154771285,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21363837122917176,
      "num_tokens": 66612207.0,
      "step": 32095
    },
    {
      "entropy": 5.57863187789917,
      "epoch": 2.9118287373004357,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041583887943821667,
      "loss": 4.8709,
      "mean_token_accuracy": 0.22028097063302993,
      "num_tokens": 66622175.0,
      "step": 32100
    },
    {
      "entropy": 5.588039255142212,
      "epoch": 2.9122822931785195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004158136444161367,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2128696247935295,
      "num_tokens": 66633936.0,
      "step": 32105
    },
    {
      "entropy": 5.650552034378052,
      "epoch": 2.9127358490566038,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041578840648199455,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20419620871543884,
      "num_tokens": 66643502.0,
      "step": 32110
    },
    {
      "entropy": 5.614045572280884,
      "epoch": 2.913189404934688,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004157631656363124,
      "loss": 4.9208,
      "mean_token_accuracy": 0.21725158840417863,
      "num_tokens": 66654890.0,
      "step": 32115
    },
    {
      "entropy": 5.652681827545166,
      "epoch": 2.9136429608127723,
      "grad_norm": 1.09375,
      "learning_rate": 0.000415737921879612,
      "loss": 5.0162,
      "mean_token_accuracy": 0.2010793849825859,
      "num_tokens": 66667037.0,
      "step": 32120
    },
    {
      "entropy": 5.603060245513916,
      "epoch": 2.914096516690856,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041571267521241554,
      "loss": 4.8915,
      "mean_token_accuracy": 0.213776995241642,
      "num_tokens": 66677291.0,
      "step": 32125
    },
    {
      "entropy": 5.5318738460540775,
      "epoch": 2.9145500725689404,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004156874256352451,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2148596927523613,
      "num_tokens": 66687421.0,
      "step": 32130
    },
    {
      "entropy": 5.674749851226807,
      "epoch": 2.9150036284470247,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004156621731486228,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2059122547507286,
      "num_tokens": 66697190.0,
      "step": 32135
    },
    {
      "entropy": 5.644549131393433,
      "epoch": 2.915457184325109,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000415636917753071,
      "loss": 5.0034,
      "mean_token_accuracy": 0.2133711650967598,
      "num_tokens": 66707255.0,
      "step": 32140
    },
    {
      "entropy": 5.560888004302979,
      "epoch": 2.915910740203193,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004156116594491119,
      "loss": 4.84,
      "mean_token_accuracy": 0.22024239003658294,
      "num_tokens": 66716451.0,
      "step": 32145
    },
    {
      "entropy": 5.570637559890747,
      "epoch": 2.9163642960812775,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004155863982372678,
      "loss": 4.9395,
      "mean_token_accuracy": 0.21990731954574586,
      "num_tokens": 66727157.0,
      "step": 32150
    },
    {
      "entropy": 5.630880689620971,
      "epoch": 2.9168178519593613,
      "grad_norm": 1.125,
      "learning_rate": 0.00041556113411806117,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2164222851395607,
      "num_tokens": 66737523.0,
      "step": 32155
    },
    {
      "entropy": 5.629849147796631,
      "epoch": 2.9172714078374455,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041553586709201453,
      "loss": 4.9399,
      "mean_token_accuracy": 0.21538535058498381,
      "num_tokens": 66748183.0,
      "step": 32160
    },
    {
      "entropy": 5.650327491760254,
      "epoch": 2.91772496371553,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004155105971596503,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22188195288181306,
      "num_tokens": 66758794.0,
      "step": 32165
    },
    {
      "entropy": 5.681767988204956,
      "epoch": 2.918178519593614,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041548532432149105,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21754937320947648,
      "num_tokens": 66769916.0,
      "step": 32170
    },
    {
      "entropy": 5.639793348312378,
      "epoch": 2.918632075471698,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004154600485780595,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22241226732730865,
      "num_tokens": 66779658.0,
      "step": 32175
    },
    {
      "entropy": 5.579885864257813,
      "epoch": 2.919085631349782,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041543476992987843,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22080554962158203,
      "num_tokens": 66788909.0,
      "step": 32180
    },
    {
      "entropy": 5.685889005661011,
      "epoch": 2.9195391872278664,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041540948837747056,
      "loss": 4.971,
      "mean_token_accuracy": 0.21447289139032363,
      "num_tokens": 66799519.0,
      "step": 32185
    },
    {
      "entropy": 5.581276988983154,
      "epoch": 2.9199927431059507,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004153842039213586,
      "loss": 4.8906,
      "mean_token_accuracy": 0.21626971960067748,
      "num_tokens": 66808915.0,
      "step": 32190
    },
    {
      "entropy": 5.6084143161773685,
      "epoch": 2.920446298984035,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004153589165620656,
      "loss": 4.9692,
      "mean_token_accuracy": 0.20788655430078506,
      "num_tokens": 66818819.0,
      "step": 32195
    },
    {
      "entropy": 5.491002702713013,
      "epoch": 2.9208998548621192,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041533362630011435,
      "loss": 4.8135,
      "mean_token_accuracy": 0.22761358469724655,
      "num_tokens": 66829274.0,
      "step": 32200
    },
    {
      "entropy": 5.675210857391358,
      "epoch": 2.921353410740203,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000415308333136028,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21928514093160628,
      "num_tokens": 66839659.0,
      "step": 32205
    },
    {
      "entropy": 5.624256610870361,
      "epoch": 2.9218069666182873,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041528303707032957,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21585683524608612,
      "num_tokens": 66850415.0,
      "step": 32210
    },
    {
      "entropy": 5.600710535049439,
      "epoch": 2.9222605224963716,
      "grad_norm": 1.25,
      "learning_rate": 0.0004152577381035422,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22674308568239213,
      "num_tokens": 66860663.0,
      "step": 32215
    },
    {
      "entropy": 5.656810331344604,
      "epoch": 2.922714078374456,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041523243623618907,
      "loss": 4.9208,
      "mean_token_accuracy": 0.21767379641532897,
      "num_tokens": 66871473.0,
      "step": 32220
    },
    {
      "entropy": 5.5699444770812985,
      "epoch": 2.9231676342525397,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041520713146879333,
      "loss": 4.8867,
      "mean_token_accuracy": 0.22040941566228867,
      "num_tokens": 66881609.0,
      "step": 32225
    },
    {
      "entropy": 5.642283487319946,
      "epoch": 2.923621190130624,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041518182380187853,
      "loss": 4.9458,
      "mean_token_accuracy": 0.21329072564840318,
      "num_tokens": 66891780.0,
      "step": 32230
    },
    {
      "entropy": 5.662907314300537,
      "epoch": 2.924074746008708,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041515651323596774,
      "loss": 5.0123,
      "mean_token_accuracy": 0.20695139467716217,
      "num_tokens": 66902251.0,
      "step": 32235
    },
    {
      "entropy": 5.639466190338135,
      "epoch": 2.9245283018867925,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041513119977158464,
      "loss": 4.957,
      "mean_token_accuracy": 0.21719592362642287,
      "num_tokens": 66912538.0,
      "step": 32240
    },
    {
      "entropy": 5.620833015441894,
      "epoch": 2.9249818577648767,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004151058834092526,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21754180341959,
      "num_tokens": 66922648.0,
      "step": 32245
    },
    {
      "entropy": 5.588517904281616,
      "epoch": 2.925435413642961,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041508056414949515,
      "loss": 4.8638,
      "mean_token_accuracy": 0.21842678189277648,
      "num_tokens": 66933115.0,
      "step": 32250
    },
    {
      "entropy": 5.671231269836426,
      "epoch": 2.9258889695210453,
      "grad_norm": 1.125,
      "learning_rate": 0.000415055241992836,
      "loss": 5.069,
      "mean_token_accuracy": 0.20341045707464217,
      "num_tokens": 66943727.0,
      "step": 32255
    },
    {
      "entropy": 5.648931837081909,
      "epoch": 2.926342525399129,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004150299169397987,
      "loss": 4.9009,
      "mean_token_accuracy": 0.21552292704582215,
      "num_tokens": 66953488.0,
      "step": 32260
    },
    {
      "entropy": 5.7079362869262695,
      "epoch": 2.9267960812772134,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004150045889909071,
      "loss": 5.0431,
      "mean_token_accuracy": 0.20397693812847137,
      "num_tokens": 66964586.0,
      "step": 32265
    },
    {
      "entropy": 5.609491395950317,
      "epoch": 2.9272496371552976,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004149792581466848,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21083665490150452,
      "num_tokens": 66975032.0,
      "step": 32270
    },
    {
      "entropy": 5.591258382797241,
      "epoch": 2.9277031930333814,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041495392440765586,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21726235300302504,
      "num_tokens": 66984910.0,
      "step": 32275
    },
    {
      "entropy": 5.591516351699829,
      "epoch": 2.9281567489114657,
      "grad_norm": 1.25,
      "learning_rate": 0.00041492858777434407,
      "loss": 4.7341,
      "mean_token_accuracy": 0.2323855698108673,
      "num_tokens": 66995238.0,
      "step": 32280
    },
    {
      "entropy": 5.601317834854126,
      "epoch": 2.92861030478955,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004149032482472734,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2219078943133354,
      "num_tokens": 67005596.0,
      "step": 32285
    },
    {
      "entropy": 5.7226598262786865,
      "epoch": 2.9290638606676342,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041487790582696803,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19873730093240738,
      "num_tokens": 67017312.0,
      "step": 32290
    },
    {
      "entropy": 5.627328205108642,
      "epoch": 2.9295174165457185,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004148525605139518,
      "loss": 4.875,
      "mean_token_accuracy": 0.22271230667829514,
      "num_tokens": 67026692.0,
      "step": 32295
    },
    {
      "entropy": 5.554681205749512,
      "epoch": 2.9299709724238028,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041482721230874897,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21969154924154283,
      "num_tokens": 67036801.0,
      "step": 32300
    },
    {
      "entropy": 5.577628087997437,
      "epoch": 2.930424528301887,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041480186121188383,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22599717527627944,
      "num_tokens": 67047262.0,
      "step": 32305
    },
    {
      "entropy": 5.611681652069092,
      "epoch": 2.930878084179971,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004147765072238806,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21240981817245483,
      "num_tokens": 67058019.0,
      "step": 32310
    },
    {
      "entropy": 5.664731884002686,
      "epoch": 2.931331640058055,
      "grad_norm": 1.125,
      "learning_rate": 0.0004147511503452635,
      "loss": 5.0093,
      "mean_token_accuracy": 0.21111210733652114,
      "num_tokens": 67068844.0,
      "step": 32315
    },
    {
      "entropy": 5.6392110824584964,
      "epoch": 2.9317851959361394,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041472579057655703,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2270249679684639,
      "num_tokens": 67077400.0,
      "step": 32320
    },
    {
      "entropy": 5.633264112472534,
      "epoch": 2.9322387518142237,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004147004279182856,
      "loss": 5.0472,
      "mean_token_accuracy": 0.20062269568443297,
      "num_tokens": 67088352.0,
      "step": 32325
    },
    {
      "entropy": 5.69848895072937,
      "epoch": 2.9326923076923075,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004146750623709737,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21285239309072496,
      "num_tokens": 67097879.0,
      "step": 32330
    },
    {
      "entropy": 5.630833959579467,
      "epoch": 2.9331458635703918,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041464969393514595,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22104258835315704,
      "num_tokens": 67108355.0,
      "step": 32335
    },
    {
      "entropy": 5.625765562057495,
      "epoch": 2.933599419448476,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004146243226113269,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2136353299021721,
      "num_tokens": 67118348.0,
      "step": 32340
    },
    {
      "entropy": 5.527565765380859,
      "epoch": 2.9340529753265603,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041459894840004135,
      "loss": 4.8571,
      "mean_token_accuracy": 0.22264847457408904,
      "num_tokens": 67128944.0,
      "step": 32345
    },
    {
      "entropy": 5.657931852340698,
      "epoch": 2.9345065312046446,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041457357130181396,
      "loss": 4.9715,
      "mean_token_accuracy": 0.2092956617474556,
      "num_tokens": 67139365.0,
      "step": 32350
    },
    {
      "entropy": 5.620879411697388,
      "epoch": 2.934960087082729,
      "grad_norm": 1.125,
      "learning_rate": 0.00041454819131716945,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21241387724876404,
      "num_tokens": 67150406.0,
      "step": 32355
    },
    {
      "entropy": 5.684853267669678,
      "epoch": 2.9354136429608126,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041452280844663295,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20633440166711808,
      "num_tokens": 67161631.0,
      "step": 32360
    },
    {
      "entropy": 5.596579360961914,
      "epoch": 2.935867198838897,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004144974226907291,
      "loss": 4.947,
      "mean_token_accuracy": 0.21707209050655366,
      "num_tokens": 67171735.0,
      "step": 32365
    },
    {
      "entropy": 5.629430627822876,
      "epoch": 2.936320754716981,
      "grad_norm": 1.1875,
      "learning_rate": 0.000414472034049983,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21705561578273774,
      "num_tokens": 67181616.0,
      "step": 32370
    },
    {
      "entropy": 5.640116548538208,
      "epoch": 2.9367743105950654,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004144466425249197,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2170005351305008,
      "num_tokens": 67192519.0,
      "step": 32375
    },
    {
      "entropy": 5.654494857788086,
      "epoch": 2.9372278664731493,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004144212481160643,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21257491260766984,
      "num_tokens": 67202805.0,
      "step": 32380
    },
    {
      "entropy": 5.615114545822143,
      "epoch": 2.9376814223512335,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004143958508239419,
      "loss": 4.9113,
      "mean_token_accuracy": 0.2178141340613365,
      "num_tokens": 67212778.0,
      "step": 32385
    },
    {
      "entropy": 5.589913082122803,
      "epoch": 2.938134978229318,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004143704506490779,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21148374676704407,
      "num_tokens": 67222588.0,
      "step": 32390
    },
    {
      "entropy": 5.685904264450073,
      "epoch": 2.938588534107402,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004143450475919975,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21193605810403823,
      "num_tokens": 67233022.0,
      "step": 32395
    },
    {
      "entropy": 5.603614234924317,
      "epoch": 2.9390420899854863,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004143196416532259,
      "loss": 4.9442,
      "mean_token_accuracy": 0.2156660333275795,
      "num_tokens": 67242631.0,
      "step": 32400
    },
    {
      "entropy": 5.559706783294677,
      "epoch": 2.9394956458635706,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004142942328332887,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22285099625587462,
      "num_tokens": 67252376.0,
      "step": 32405
    },
    {
      "entropy": 5.632914686203003,
      "epoch": 2.9399492017416544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004142688211327111,
      "loss": 4.8714,
      "mean_token_accuracy": 0.21618175655603408,
      "num_tokens": 67262209.0,
      "step": 32410
    },
    {
      "entropy": 5.679757499694825,
      "epoch": 2.9404027576197387,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004142434065520189,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2097629651427269,
      "num_tokens": 67272893.0,
      "step": 32415
    },
    {
      "entropy": 5.590439653396606,
      "epoch": 2.940856313497823,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004142179890917376,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21640788316726683,
      "num_tokens": 67284017.0,
      "step": 32420
    },
    {
      "entropy": 5.569850015640259,
      "epoch": 2.941309869375907,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041419256875239275,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22171291708946228,
      "num_tokens": 67293674.0,
      "step": 32425
    },
    {
      "entropy": 5.57862958908081,
      "epoch": 2.941763425253991,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004141671455345101,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21450210213661194,
      "num_tokens": 67303759.0,
      "step": 32430
    },
    {
      "entropy": 5.676352024078369,
      "epoch": 2.9422169811320753,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041414171943861543,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2144734889268875,
      "num_tokens": 67314962.0,
      "step": 32435
    },
    {
      "entropy": 5.679864740371704,
      "epoch": 2.9426705370101596,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004141162904652346,
      "loss": 5.0533,
      "mean_token_accuracy": 0.20631122291088105,
      "num_tokens": 67325210.0,
      "step": 32440
    },
    {
      "entropy": 5.6194719791412355,
      "epoch": 2.943124092888244,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004140908586148934,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21291278451681137,
      "num_tokens": 67335110.0,
      "step": 32445
    },
    {
      "entropy": 5.638901090621948,
      "epoch": 2.943577648766328,
      "grad_norm": 1.125,
      "learning_rate": 0.0004140654238881178,
      "loss": 5.0435,
      "mean_token_accuracy": 0.2125014528632164,
      "num_tokens": 67345872.0,
      "step": 32450
    },
    {
      "entropy": 5.637533044815063,
      "epoch": 2.9440312046444124,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041403998628543375,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21328679472208023,
      "num_tokens": 67356532.0,
      "step": 32455
    },
    {
      "entropy": 5.622058010101318,
      "epoch": 2.9444847605224966,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041401454580736744,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2226173236966133,
      "num_tokens": 67366603.0,
      "step": 32460
    },
    {
      "entropy": 5.708301973342896,
      "epoch": 2.9449383164005805,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004139891024544448,
      "loss": 5.047,
      "mean_token_accuracy": 0.21689256876707078,
      "num_tokens": 67377487.0,
      "step": 32465
    },
    {
      "entropy": 5.688707780838013,
      "epoch": 2.9453918722786647,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004139636562271921,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21270985156297684,
      "num_tokens": 67387704.0,
      "step": 32470
    },
    {
      "entropy": 5.662493801116943,
      "epoch": 2.945845428156749,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041393820712613567,
      "loss": 4.9919,
      "mean_token_accuracy": 0.2091204598546028,
      "num_tokens": 67398409.0,
      "step": 32475
    },
    {
      "entropy": 5.63400411605835,
      "epoch": 2.9462989840348333,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041391275515180166,
      "loss": 4.8821,
      "mean_token_accuracy": 0.21090206652879714,
      "num_tokens": 67408905.0,
      "step": 32480
    },
    {
      "entropy": 5.625855731964111,
      "epoch": 2.946752539912917,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004138873003047165,
      "loss": 4.8921,
      "mean_token_accuracy": 0.22361928671598436,
      "num_tokens": 67418466.0,
      "step": 32485
    },
    {
      "entropy": 5.542302846908569,
      "epoch": 2.9472060957910013,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004138618425854066,
      "loss": 4.931,
      "mean_token_accuracy": 0.2166370004415512,
      "num_tokens": 67428597.0,
      "step": 32490
    },
    {
      "entropy": 5.594061994552613,
      "epoch": 2.9476596516690856,
      "grad_norm": 1.25,
      "learning_rate": 0.00041383638199439824,
      "loss": 4.9029,
      "mean_token_accuracy": 0.21714912801980973,
      "num_tokens": 67439158.0,
      "step": 32495
    },
    {
      "entropy": 5.665826368331909,
      "epoch": 2.94811320754717,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041381091853221835,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21099960058927536,
      "num_tokens": 67449540.0,
      "step": 32500
    },
    {
      "entropy": 5.611739921569824,
      "epoch": 2.948566763425254,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004137854521993932,
      "loss": 4.8953,
      "mean_token_accuracy": 0.21903390288352967,
      "num_tokens": 67459385.0,
      "step": 32505
    },
    {
      "entropy": 5.643102741241455,
      "epoch": 2.9490203193033384,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041375998299644946,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21498721092939377,
      "num_tokens": 67469597.0,
      "step": 32510
    },
    {
      "entropy": 5.61246542930603,
      "epoch": 2.9494738751814222,
      "grad_norm": 1.25,
      "learning_rate": 0.000413734510923914,
      "loss": 4.8602,
      "mean_token_accuracy": 0.21646612882614136,
      "num_tokens": 67479271.0,
      "step": 32515
    },
    {
      "entropy": 5.577236032485962,
      "epoch": 2.9499274310595065,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004137090359823135,
      "loss": 4.8633,
      "mean_token_accuracy": 0.2210137739777565,
      "num_tokens": 67488997.0,
      "step": 32520
    },
    {
      "entropy": 5.594775915145874,
      "epoch": 2.9503809869375908,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041368355817217474,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21761196702718735,
      "num_tokens": 67499507.0,
      "step": 32525
    },
    {
      "entropy": 5.623895502090454,
      "epoch": 2.950834542815675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041365807749402467,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22030483037233353,
      "num_tokens": 67509916.0,
      "step": 32530
    },
    {
      "entropy": 5.613157272338867,
      "epoch": 2.951288098693759,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004136325939483902,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2098008692264557,
      "num_tokens": 67519759.0,
      "step": 32535
    },
    {
      "entropy": 5.614022159576416,
      "epoch": 2.951741654571843,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004136071075357984,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21799196153879166,
      "num_tokens": 67529261.0,
      "step": 32540
    },
    {
      "entropy": 5.619625425338745,
      "epoch": 2.9521952104499274,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004135816182567763,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21682696789503098,
      "num_tokens": 67539084.0,
      "step": 32545
    },
    {
      "entropy": 5.662848567962646,
      "epoch": 2.9526487663280117,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000413556126111851,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2152765229344368,
      "num_tokens": 67548551.0,
      "step": 32550
    },
    {
      "entropy": 5.618746423721314,
      "epoch": 2.953102322206096,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004135306311015497,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21302825957536697,
      "num_tokens": 67558364.0,
      "step": 32555
    },
    {
      "entropy": 5.660374736785888,
      "epoch": 2.95355587808418,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004135051332263996,
      "loss": 4.9221,
      "mean_token_accuracy": 0.2231972724199295,
      "num_tokens": 67569257.0,
      "step": 32560
    },
    {
      "entropy": 5.613333082199096,
      "epoch": 2.954009433962264,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041347963248692814,
      "loss": 5.0083,
      "mean_token_accuracy": 0.20575937181711196,
      "num_tokens": 67580135.0,
      "step": 32565
    },
    {
      "entropy": 5.599939680099487,
      "epoch": 2.9544629898403483,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004134541288836626,
      "loss": 4.9353,
      "mean_token_accuracy": 0.21489370316267015,
      "num_tokens": 67589705.0,
      "step": 32570
    },
    {
      "entropy": 5.644237804412842,
      "epoch": 2.9549165457184325,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041342862241713034,
      "loss": 4.89,
      "mean_token_accuracy": 0.2166994646191597,
      "num_tokens": 67599648.0,
      "step": 32575
    },
    {
      "entropy": 5.662343168258667,
      "epoch": 2.955370101596517,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041340311308785886,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21581293493509293,
      "num_tokens": 67609624.0,
      "step": 32580
    },
    {
      "entropy": 5.5733386993408205,
      "epoch": 2.9558236574746006,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041337760089637577,
      "loss": 4.9252,
      "mean_token_accuracy": 0.2156374931335449,
      "num_tokens": 67619596.0,
      "step": 32585
    },
    {
      "entropy": 5.599661159515381,
      "epoch": 2.956277213352685,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041335208584320865,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2136589229106903,
      "num_tokens": 67629667.0,
      "step": 32590
    },
    {
      "entropy": 5.674688863754272,
      "epoch": 2.956730769230769,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004133265679288851,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2157885566353798,
      "num_tokens": 67640048.0,
      "step": 32595
    },
    {
      "entropy": 5.688122081756592,
      "epoch": 2.9571843251088534,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004133010471539329,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21059911847114562,
      "num_tokens": 67650731.0,
      "step": 32600
    },
    {
      "entropy": 5.597215557098389,
      "epoch": 2.9576378809869377,
      "grad_norm": 1.125,
      "learning_rate": 0.00041327552351887983,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2153029352426529,
      "num_tokens": 67661696.0,
      "step": 32605
    },
    {
      "entropy": 5.62748908996582,
      "epoch": 2.958091436865022,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041324999702425366,
      "loss": 4.9533,
      "mean_token_accuracy": 0.2140968382358551,
      "num_tokens": 67671758.0,
      "step": 32610
    },
    {
      "entropy": 5.585301160812378,
      "epoch": 2.9585449927431062,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004132244676705823,
      "loss": 4.9171,
      "mean_token_accuracy": 0.21776215583086014,
      "num_tokens": 67682488.0,
      "step": 32615
    },
    {
      "entropy": 5.675340986251831,
      "epoch": 2.95899854862119,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004131989354583938,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21965046972036362,
      "num_tokens": 67692945.0,
      "step": 32620
    },
    {
      "entropy": 5.641361427307129,
      "epoch": 2.9594521044992743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000413173400388216,
      "loss": 4.8613,
      "mean_token_accuracy": 0.21642902493476868,
      "num_tokens": 67702924.0,
      "step": 32625
    },
    {
      "entropy": 5.6127790927886965,
      "epoch": 2.9599056603773586,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004131478624605771,
      "loss": 4.9589,
      "mean_token_accuracy": 0.21458469033241273,
      "num_tokens": 67714284.0,
      "step": 32630
    },
    {
      "entropy": 5.637450075149536,
      "epoch": 2.9603592162554424,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004131223216760052,
      "loss": 5.0029,
      "mean_token_accuracy": 0.20886660665273665,
      "num_tokens": 67725902.0,
      "step": 32635
    },
    {
      "entropy": 5.65799617767334,
      "epoch": 2.9608127721335267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004130967780350286,
      "loss": 4.921,
      "mean_token_accuracy": 0.21364203691482545,
      "num_tokens": 67736268.0,
      "step": 32640
    },
    {
      "entropy": 5.710126209259033,
      "epoch": 2.961266328011611,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004130712315381753,
      "loss": 4.962,
      "mean_token_accuracy": 0.22410699874162673,
      "num_tokens": 67745709.0,
      "step": 32645
    },
    {
      "entropy": 5.6640753746032715,
      "epoch": 2.961719883889695,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041304568218597385,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21398856341838837,
      "num_tokens": 67756092.0,
      "step": 32650
    },
    {
      "entropy": 5.646423768997193,
      "epoch": 2.9621734397677795,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041302012997895246,
      "loss": 4.9561,
      "mean_token_accuracy": 0.21489520072937013,
      "num_tokens": 67766224.0,
      "step": 32655
    },
    {
      "entropy": 5.70157380104065,
      "epoch": 2.9626269956458637,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004129945749176395,
      "loss": 5.0903,
      "mean_token_accuracy": 0.21241241246461867,
      "num_tokens": 67776787.0,
      "step": 32660
    },
    {
      "entropy": 5.62306957244873,
      "epoch": 2.963080551523948,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041296901700256375,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21550153195858002,
      "num_tokens": 67787924.0,
      "step": 32665
    },
    {
      "entropy": 5.6585205554962155,
      "epoch": 2.963534107402032,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004129434562342535,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2023211821913719,
      "num_tokens": 67798544.0,
      "step": 32670
    },
    {
      "entropy": 5.589351844787598,
      "epoch": 2.963987663280116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041291789261323734,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21614918112754822,
      "num_tokens": 67809528.0,
      "step": 32675
    },
    {
      "entropy": 5.516956758499146,
      "epoch": 2.9644412191582004,
      "grad_norm": 1.125,
      "learning_rate": 0.00041289232614004416,
      "loss": 4.8051,
      "mean_token_accuracy": 0.23332008570432664,
      "num_tokens": 67820263.0,
      "step": 32680
    },
    {
      "entropy": 5.7073241710662845,
      "epoch": 2.9648947750362846,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004128667568152024,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20994754433631896,
      "num_tokens": 67830687.0,
      "step": 32685
    },
    {
      "entropy": 5.742996454238892,
      "epoch": 2.9653483309143684,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041284118463924107,
      "loss": 5.0012,
      "mean_token_accuracy": 0.20630993843078613,
      "num_tokens": 67841714.0,
      "step": 32690
    },
    {
      "entropy": 5.59776291847229,
      "epoch": 2.9658018867924527,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004128156096126889,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2160676062107086,
      "num_tokens": 67851952.0,
      "step": 32695
    },
    {
      "entropy": 5.5455316543579105,
      "epoch": 2.966255442670537,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041279003173607477,
      "loss": 4.8196,
      "mean_token_accuracy": 0.21863433271646499,
      "num_tokens": 67861958.0,
      "step": 32700
    },
    {
      "entropy": 5.569171380996704,
      "epoch": 2.9667089985486212,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004127644510099278,
      "loss": 4.8125,
      "mean_token_accuracy": 0.22394121438264847,
      "num_tokens": 67871654.0,
      "step": 32705
    },
    {
      "entropy": 5.570980644226074,
      "epoch": 2.9671625544267055,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004127388674347767,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22241017520427703,
      "num_tokens": 67881678.0,
      "step": 32710
    },
    {
      "entropy": 5.5784557342529295,
      "epoch": 2.96761611030479,
      "grad_norm": 1.125,
      "learning_rate": 0.00041271328101115075,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21070809960365294,
      "num_tokens": 67892663.0,
      "step": 32715
    },
    {
      "entropy": 5.742793178558349,
      "epoch": 2.9680696661828736,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041268769173957915,
      "loss": 5.0946,
      "mean_token_accuracy": 0.20620316565036773,
      "num_tokens": 67903308.0,
      "step": 32720
    },
    {
      "entropy": 5.677955484390258,
      "epoch": 2.968523222060958,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041266209962059097,
      "loss": 4.913,
      "mean_token_accuracy": 0.21964753568172454,
      "num_tokens": 67913092.0,
      "step": 32725
    },
    {
      "entropy": 5.618656778335572,
      "epoch": 2.968976777939042,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004126365046547154,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22545998841524123,
      "num_tokens": 67923708.0,
      "step": 32730
    },
    {
      "entropy": 5.556598711013794,
      "epoch": 2.9694303338171264,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004126109068424819,
      "loss": 4.9182,
      "mean_token_accuracy": 0.2105742424726486,
      "num_tokens": 67934164.0,
      "step": 32735
    },
    {
      "entropy": 5.669560670852661,
      "epoch": 2.96988388969521,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004125853061844198,
      "loss": 5.0164,
      "mean_token_accuracy": 0.20944148451089858,
      "num_tokens": 67943162.0,
      "step": 32740
    },
    {
      "entropy": 5.681047248840332,
      "epoch": 2.9703374455732945,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004125597026810585,
      "loss": 5.014,
      "mean_token_accuracy": 0.21345773637294768,
      "num_tokens": 67953943.0,
      "step": 32745
    },
    {
      "entropy": 5.580320644378662,
      "epoch": 2.9707910014513788,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041253409633292747,
      "loss": 4.907,
      "mean_token_accuracy": 0.21638807654380798,
      "num_tokens": 67964603.0,
      "step": 32750
    },
    {
      "entropy": 5.653650856018066,
      "epoch": 2.971244557329463,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041250848714055626,
      "loss": 4.9801,
      "mean_token_accuracy": 0.21489916741847992,
      "num_tokens": 67974541.0,
      "step": 32755
    },
    {
      "entropy": 5.652291536331177,
      "epoch": 2.9716981132075473,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004124828751044746,
      "loss": 4.9515,
      "mean_token_accuracy": 0.20700111091136933,
      "num_tokens": 67984968.0,
      "step": 32760
    },
    {
      "entropy": 5.621393728256225,
      "epoch": 2.9721516690856316,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004124572602252119,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2190841406583786,
      "num_tokens": 67995567.0,
      "step": 32765
    },
    {
      "entropy": 5.674803781509399,
      "epoch": 2.9726052249637154,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004124316425032981,
      "loss": 5.0553,
      "mean_token_accuracy": 0.20576776564121246,
      "num_tokens": 68008530.0,
      "step": 32770
    },
    {
      "entropy": 5.708988618850708,
      "epoch": 2.9730587808417996,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041240602193926294,
      "loss": 5.014,
      "mean_token_accuracy": 0.21019861698150635,
      "num_tokens": 68018393.0,
      "step": 32775
    },
    {
      "entropy": 5.693976020812988,
      "epoch": 2.973512336719884,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004123803985336362,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21103900372982026,
      "num_tokens": 68027677.0,
      "step": 32780
    },
    {
      "entropy": 5.616236925125122,
      "epoch": 2.973965892597968,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041235477228694786,
      "loss": 4.8312,
      "mean_token_accuracy": 0.22508990317583083,
      "num_tokens": 68037157.0,
      "step": 32785
    },
    {
      "entropy": 5.622634601593018,
      "epoch": 2.974419448476052,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041232914319972775,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22021604776382447,
      "num_tokens": 68047793.0,
      "step": 32790
    },
    {
      "entropy": 5.627812910079956,
      "epoch": 2.9748730043541363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041230351127250603,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21828629225492477,
      "num_tokens": 68058155.0,
      "step": 32795
    },
    {
      "entropy": 5.668702793121338,
      "epoch": 2.9753265602322205,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041227787650581266,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22094268053770066,
      "num_tokens": 68067440.0,
      "step": 32800
    },
    {
      "entropy": 5.63725643157959,
      "epoch": 2.975780116110305,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041225223890017785,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21871229112148285,
      "num_tokens": 68077815.0,
      "step": 32805
    },
    {
      "entropy": 5.656004333496094,
      "epoch": 2.976233671988389,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041222659845613173,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2228584349155426,
      "num_tokens": 68087692.0,
      "step": 32810
    },
    {
      "entropy": 5.583959674835205,
      "epoch": 2.9766872278664733,
      "grad_norm": 1.25,
      "learning_rate": 0.00041220095517420464,
      "loss": 4.8765,
      "mean_token_accuracy": 0.21812089830636977,
      "num_tokens": 68097774.0,
      "step": 32815
    },
    {
      "entropy": 5.606702470779419,
      "epoch": 2.9771407837445576,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004121753090549268,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2166294664144516,
      "num_tokens": 68107992.0,
      "step": 32820
    },
    {
      "entropy": 5.588299131393432,
      "epoch": 2.9775943396226414,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041214966009882857,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22164697498083114,
      "num_tokens": 68117103.0,
      "step": 32825
    },
    {
      "entropy": 5.689782953262329,
      "epoch": 2.9780478955007257,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004121240083064405,
      "loss": 4.9878,
      "mean_token_accuracy": 0.21295422017574311,
      "num_tokens": 68127951.0,
      "step": 32830
    },
    {
      "entropy": 5.694617986679077,
      "epoch": 2.97850145137881,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041209835367829293,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21347694098949432,
      "num_tokens": 68138229.0,
      "step": 32835
    },
    {
      "entropy": 5.555681800842285,
      "epoch": 2.9789550072568938,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041207269621491654,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21759978234767913,
      "num_tokens": 68148019.0,
      "step": 32840
    },
    {
      "entropy": 5.567492580413818,
      "epoch": 2.979408563134978,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004120470359168418,
      "loss": 4.9352,
      "mean_token_accuracy": 0.20564557015895843,
      "num_tokens": 68158704.0,
      "step": 32845
    },
    {
      "entropy": 5.6813959121704105,
      "epoch": 2.9798621190130623,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041202137278459946,
      "loss": 4.946,
      "mean_token_accuracy": 0.2196308732032776,
      "num_tokens": 68169105.0,
      "step": 32850
    },
    {
      "entropy": 5.6532128810882565,
      "epoch": 2.9803156748911466,
      "grad_norm": 1.125,
      "learning_rate": 0.0004119957068187202,
      "loss": 4.8721,
      "mean_token_accuracy": 0.21854404211044312,
      "num_tokens": 68179778.0,
      "step": 32855
    },
    {
      "entropy": 5.623121643066407,
      "epoch": 2.980769230769231,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004119700380197348,
      "loss": 4.8947,
      "mean_token_accuracy": 0.21520705074071883,
      "num_tokens": 68189405.0,
      "step": 32860
    },
    {
      "entropy": 5.511805057525635,
      "epoch": 2.981222786647315,
      "grad_norm": 1.25,
      "learning_rate": 0.00041194436638817417,
      "loss": 4.8594,
      "mean_token_accuracy": 0.2265425905585289,
      "num_tokens": 68199351.0,
      "step": 32865
    },
    {
      "entropy": 5.663418102264404,
      "epoch": 2.9816763425253994,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041191869192456907,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21237728297710418,
      "num_tokens": 68209904.0,
      "step": 32870
    },
    {
      "entropy": 5.685230398178101,
      "epoch": 2.982129898403483,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004118930146294506,
      "loss": 5.0164,
      "mean_token_accuracy": 0.2111487627029419,
      "num_tokens": 68221754.0,
      "step": 32875
    },
    {
      "entropy": 5.657987499237061,
      "epoch": 2.9825834542815675,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004118673345033496,
      "loss": 4.9369,
      "mean_token_accuracy": 0.2101563557982445,
      "num_tokens": 68232328.0,
      "step": 32880
    },
    {
      "entropy": 5.56987247467041,
      "epoch": 2.9830370101596517,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004118416515467974,
      "loss": 4.8615,
      "mean_token_accuracy": 0.2152936726808548,
      "num_tokens": 68242812.0,
      "step": 32885
    },
    {
      "entropy": 5.6841230392456055,
      "epoch": 2.983490566037736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004118159657603248,
      "loss": 5.0067,
      "mean_token_accuracy": 0.20781323462724685,
      "num_tokens": 68253888.0,
      "step": 32890
    },
    {
      "entropy": 5.678138589859008,
      "epoch": 2.98394412191582,
      "grad_norm": 1.25,
      "learning_rate": 0.0004117902771444633,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21844129264354706,
      "num_tokens": 68264877.0,
      "step": 32895
    },
    {
      "entropy": 5.731426858901978,
      "epoch": 2.984397677793904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000411764585699744,
      "loss": 5.0304,
      "mean_token_accuracy": 0.204776369035244,
      "num_tokens": 68275863.0,
      "step": 32900
    },
    {
      "entropy": 5.671324920654297,
      "epoch": 2.9848512336719883,
      "grad_norm": 1.125,
      "learning_rate": 0.00041173889142669806,
      "loss": 4.952,
      "mean_token_accuracy": 0.21256603598594664,
      "num_tokens": 68286768.0,
      "step": 32905
    },
    {
      "entropy": 5.598630094528199,
      "epoch": 2.9853047895500726,
      "grad_norm": 1.125,
      "learning_rate": 0.0004117131943258572,
      "loss": 4.889,
      "mean_token_accuracy": 0.22051786929368972,
      "num_tokens": 68296990.0,
      "step": 32910
    },
    {
      "entropy": 5.540438270568847,
      "epoch": 2.985758345428157,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041168749439775255,
      "loss": 4.8733,
      "mean_token_accuracy": 0.21655682027339934,
      "num_tokens": 68307646.0,
      "step": 32915
    },
    {
      "entropy": 5.670006847381591,
      "epoch": 2.986211901306241,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004116617916429156,
      "loss": 5.0244,
      "mean_token_accuracy": 0.2048196941614151,
      "num_tokens": 68319586.0,
      "step": 32920
    },
    {
      "entropy": 5.736089468002319,
      "epoch": 2.986665457184325,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004116360860618781,
      "loss": 4.9894,
      "mean_token_accuracy": 0.2112579584121704,
      "num_tokens": 68329692.0,
      "step": 32925
    },
    {
      "entropy": 5.69001579284668,
      "epoch": 2.9871190130624092,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004116103776551715,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2142190381884575,
      "num_tokens": 68339589.0,
      "step": 32930
    },
    {
      "entropy": 5.647755146026611,
      "epoch": 2.9875725689404935,
      "grad_norm": 1.125,
      "learning_rate": 0.00041158466642332743,
      "loss": 5.012,
      "mean_token_accuracy": 0.20411770343780516,
      "num_tokens": 68349485.0,
      "step": 32935
    },
    {
      "entropy": 5.647630310058593,
      "epoch": 2.9880261248185778,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004115589523668777,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22176818698644638,
      "num_tokens": 68359695.0,
      "step": 32940
    },
    {
      "entropy": 5.584457778930664,
      "epoch": 2.9884796806966616,
      "grad_norm": 1.125,
      "learning_rate": 0.000411533235486354,
      "loss": 4.8175,
      "mean_token_accuracy": 0.2219352290034294,
      "num_tokens": 68369591.0,
      "step": 32945
    },
    {
      "entropy": 5.638293361663818,
      "epoch": 2.988933236574746,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004115075157822883,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21667128652334214,
      "num_tokens": 68380382.0,
      "step": 32950
    },
    {
      "entropy": 5.5126505374908445,
      "epoch": 2.98938679245283,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004114817932552122,
      "loss": 4.8299,
      "mean_token_accuracy": 0.22157813012599945,
      "num_tokens": 68389865.0,
      "step": 32955
    },
    {
      "entropy": 5.587721443176269,
      "epoch": 2.9898403483309144,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041145606790565797,
      "loss": 4.848,
      "mean_token_accuracy": 0.22686622738838197,
      "num_tokens": 68399267.0,
      "step": 32960
    },
    {
      "entropy": 5.6585770606994625,
      "epoch": 2.9902939042089987,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004114303397341575,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21847444921731948,
      "num_tokens": 68409480.0,
      "step": 32965
    },
    {
      "entropy": 5.759487581253052,
      "epoch": 2.990747460087083,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004114046087412427,
      "loss": 5.0621,
      "mean_token_accuracy": 0.21091356724500657,
      "num_tokens": 68420019.0,
      "step": 32970
    },
    {
      "entropy": 5.5844038963317875,
      "epoch": 2.9912010159651667,
      "grad_norm": 1.25,
      "learning_rate": 0.000411378874927446,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22049353122711182,
      "num_tokens": 68429967.0,
      "step": 32975
    },
    {
      "entropy": 5.568133401870727,
      "epoch": 2.991654571843251,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004113531382932993,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2174044206738472,
      "num_tokens": 68441639.0,
      "step": 32980
    },
    {
      "entropy": 5.680133819580078,
      "epoch": 2.9921081277213353,
      "grad_norm": 1.09375,
      "learning_rate": 0.000411327398839335,
      "loss": 4.9494,
      "mean_token_accuracy": 0.20901804119348527,
      "num_tokens": 68452781.0,
      "step": 32985
    },
    {
      "entropy": 5.646609878540039,
      "epoch": 2.9925616835994195,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004113016565660854,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2187865272164345,
      "num_tokens": 68463353.0,
      "step": 32990
    },
    {
      "entropy": 5.55716233253479,
      "epoch": 2.9930152394775034,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041127591147408274,
      "loss": 4.907,
      "mean_token_accuracy": 0.20799251645803452,
      "num_tokens": 68473507.0,
      "step": 32995
    },
    {
      "entropy": 5.619699001312256,
      "epoch": 2.9934687953555876,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041125016356385945,
      "loss": 4.9235,
      "mean_token_accuracy": 0.21613216400146484,
      "num_tokens": 68483737.0,
      "step": 33000
    },
    {
      "epoch": 2.9934687953555876,
      "eval_entropy": 5.388790743167584,
      "eval_loss": 4.98230504989624,
      "eval_mean_token_accuracy": 0.22280678409257088,
      "eval_num_tokens": 68483737.0,
      "eval_runtime": 20.6588,
      "eval_samples_per_second": 1711.617,
      "eval_steps_per_second": 213.952,
      "step": 33000
    },
    {
      "entropy": 5.655807256698608,
      "epoch": 2.993922351233672,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041122441283594817,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2172561466693878,
      "num_tokens": 68494013.0,
      "step": 33005
    },
    {
      "entropy": 5.604366636276245,
      "epoch": 2.994375907111756,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004111986592908813,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22452555745840072,
      "num_tokens": 68504361.0,
      "step": 33010
    },
    {
      "entropy": 5.587920141220093,
      "epoch": 2.9948294629898404,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041117290292919134,
      "loss": 4.8609,
      "mean_token_accuracy": 0.221909761428833,
      "num_tokens": 68514825.0,
      "step": 33015
    },
    {
      "entropy": 5.6046061515808105,
      "epoch": 2.9952830188679247,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004111471437514111,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21499638706445695,
      "num_tokens": 68524858.0,
      "step": 33020
    },
    {
      "entropy": 5.599625158309936,
      "epoch": 2.995736574746009,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004111213817580733,
      "loss": 4.8385,
      "mean_token_accuracy": 0.2205299213528633,
      "num_tokens": 68534481.0,
      "step": 33025
    },
    {
      "entropy": 5.585771274566651,
      "epoch": 2.996190130624093,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004110956169497106,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21296273022890091,
      "num_tokens": 68545436.0,
      "step": 33030
    },
    {
      "entropy": 5.583476209640503,
      "epoch": 2.996643686502177,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004110698493268558,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22294741868972778,
      "num_tokens": 68555434.0,
      "step": 33035
    },
    {
      "entropy": 5.642462682723999,
      "epoch": 2.9970972423802613,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004110440788900419,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21189566403627397,
      "num_tokens": 68566295.0,
      "step": 33040
    },
    {
      "entropy": 5.577358722686768,
      "epoch": 2.9975507982583456,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041101830563980174,
      "loss": 4.8539,
      "mean_token_accuracy": 0.2244438037276268,
      "num_tokens": 68575927.0,
      "step": 33045
    },
    {
      "entropy": 5.668047857284546,
      "epoch": 2.9980043541364294,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041099252957666843,
      "loss": 4.9653,
      "mean_token_accuracy": 0.2177191987633705,
      "num_tokens": 68586254.0,
      "step": 33050
    },
    {
      "entropy": 5.565159559249878,
      "epoch": 2.9984579100145137,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041096675070117496,
      "loss": 4.9457,
      "mean_token_accuracy": 0.21760228872299195,
      "num_tokens": 68596881.0,
      "step": 33055
    },
    {
      "entropy": 5.556361150741577,
      "epoch": 2.998911465892598,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041094096901385434,
      "loss": 4.8287,
      "mean_token_accuracy": 0.22269561588764192,
      "num_tokens": 68606875.0,
      "step": 33060
    },
    {
      "entropy": 5.67962532043457,
      "epoch": 2.999365021770682,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004109151845152399,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21584342271089554,
      "num_tokens": 68617011.0,
      "step": 33065
    },
    {
      "entropy": 5.678834629058838,
      "epoch": 2.9998185776487665,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004108893972058649,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21373047977685927,
      "num_tokens": 68627764.0,
      "step": 33070
    },
    {
      "entropy": 5.660772180557251,
      "epoch": 3.0002721335268503,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041086360708626236,
      "loss": 4.9232,
      "mean_token_accuracy": 0.2192634165287018,
      "num_tokens": 68637272.0,
      "step": 33075
    },
    {
      "entropy": 5.565616655349731,
      "epoch": 3.0007256894049346,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041083781415696585,
      "loss": 4.8847,
      "mean_token_accuracy": 0.21277209371328354,
      "num_tokens": 68648058.0,
      "step": 33080
    },
    {
      "entropy": 5.5635254859924315,
      "epoch": 3.001179245283019,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004108120184185088,
      "loss": 4.805,
      "mean_token_accuracy": 0.2236357167363167,
      "num_tokens": 68658174.0,
      "step": 33085
    },
    {
      "entropy": 5.637840366363525,
      "epoch": 3.001632801161103,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041078621987142457,
      "loss": 4.8467,
      "mean_token_accuracy": 0.22610540241003035,
      "num_tokens": 68669281.0,
      "step": 33090
    },
    {
      "entropy": 5.590885257720947,
      "epoch": 3.0020863570391874,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004107604185162467,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21254901438951493,
      "num_tokens": 68679879.0,
      "step": 33095
    },
    {
      "entropy": 5.525386905670166,
      "epoch": 3.0025399129172716,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041073461435350877,
      "loss": 4.8058,
      "mean_token_accuracy": 0.2221756562590599,
      "num_tokens": 68689852.0,
      "step": 33100
    },
    {
      "entropy": 5.553326749801636,
      "epoch": 3.0029934687953554,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041070880738374447,
      "loss": 4.879,
      "mean_token_accuracy": 0.21886124163866044,
      "num_tokens": 68700149.0,
      "step": 33105
    },
    {
      "entropy": 5.597825431823731,
      "epoch": 3.0034470246734397,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041068299760748734,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22189686000347136,
      "num_tokens": 68710837.0,
      "step": 33110
    },
    {
      "entropy": 5.614534330368042,
      "epoch": 3.003900580551524,
      "grad_norm": 1.125,
      "learning_rate": 0.00041065718502527134,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2179928243160248,
      "num_tokens": 68721451.0,
      "step": 33115
    },
    {
      "entropy": 5.578591299057007,
      "epoch": 3.0043541364296082,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041063136963763013,
      "loss": 4.8614,
      "mean_token_accuracy": 0.22229890376329423,
      "num_tokens": 68732911.0,
      "step": 33120
    },
    {
      "entropy": 5.595605134963989,
      "epoch": 3.0048076923076925,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004106055514450977,
      "loss": 4.8231,
      "mean_token_accuracy": 0.22759455889463426,
      "num_tokens": 68742650.0,
      "step": 33125
    },
    {
      "entropy": 5.5638525009155275,
      "epoch": 3.0052612481857763,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004105797304482079,
      "loss": 4.8328,
      "mean_token_accuracy": 0.22094301879405975,
      "num_tokens": 68753660.0,
      "step": 33130
    },
    {
      "entropy": 5.568468713760376,
      "epoch": 3.0057148040638606,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004105539066474947,
      "loss": 4.762,
      "mean_token_accuracy": 0.22654962837696074,
      "num_tokens": 68763868.0,
      "step": 33135
    },
    {
      "entropy": 5.59303650856018,
      "epoch": 3.006168359941945,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041052808004349214,
      "loss": 4.8563,
      "mean_token_accuracy": 0.22007373422384263,
      "num_tokens": 68773911.0,
      "step": 33140
    },
    {
      "entropy": 5.642910814285278,
      "epoch": 3.006621915820029,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004105022506367344,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21345291584730147,
      "num_tokens": 68786423.0,
      "step": 33145
    },
    {
      "entropy": 5.635404300689697,
      "epoch": 3.0070754716981134,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004104764184277555,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21514613479375838,
      "num_tokens": 68797329.0,
      "step": 33150
    },
    {
      "entropy": 5.666480016708374,
      "epoch": 3.0075290275761972,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004104505834170898,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21784710884094238,
      "num_tokens": 68808447.0,
      "step": 33155
    },
    {
      "entropy": 5.636412954330444,
      "epoch": 3.0079825834542815,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041042474560527153,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2221631035208702,
      "num_tokens": 68819381.0,
      "step": 33160
    },
    {
      "entropy": 5.545153617858887,
      "epoch": 3.0084361393323658,
      "grad_norm": 1.1875,
      "learning_rate": 0.000410398904992835,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22535058856010437,
      "num_tokens": 68829319.0,
      "step": 33165
    },
    {
      "entropy": 5.599809694290161,
      "epoch": 3.00888969521045,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041037306158031464,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22253966927528382,
      "num_tokens": 68838640.0,
      "step": 33170
    },
    {
      "entropy": 5.646754407882691,
      "epoch": 3.0093432510885343,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004103472153682448,
      "loss": 4.8637,
      "mean_token_accuracy": 0.21811421811580659,
      "num_tokens": 68847960.0,
      "step": 33175
    },
    {
      "entropy": 5.618949508666992,
      "epoch": 3.009796806966618,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041032136635716016,
      "loss": 4.8798,
      "mean_token_accuracy": 0.21817819625139237,
      "num_tokens": 68858851.0,
      "step": 33180
    },
    {
      "entropy": 5.5537933826446535,
      "epoch": 3.0102503628447024,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004102955145475951,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2198275536298752,
      "num_tokens": 68869006.0,
      "step": 33185
    },
    {
      "entropy": 5.689778518676758,
      "epoch": 3.0107039187227866,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041026965994008435,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21148451864719392,
      "num_tokens": 68879606.0,
      "step": 33190
    },
    {
      "entropy": 5.561501836776733,
      "epoch": 3.011157474600871,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041024380253516253,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23179419040679933,
      "num_tokens": 68889862.0,
      "step": 33195
    },
    {
      "entropy": 5.719213008880615,
      "epoch": 3.011611030478955,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041021794233336444,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21639616787433624,
      "num_tokens": 68900512.0,
      "step": 33200
    },
    {
      "entropy": 5.622331953048706,
      "epoch": 3.012064586357039,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004101920793352249,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21384023427963256,
      "num_tokens": 68910475.0,
      "step": 33205
    },
    {
      "entropy": 5.673051500320435,
      "epoch": 3.0125181422351233,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004101662135412786,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21494946479797364,
      "num_tokens": 68920583.0,
      "step": 33210
    },
    {
      "entropy": 5.543615484237671,
      "epoch": 3.0129716981132075,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004101403449520606,
      "loss": 4.8092,
      "mean_token_accuracy": 0.2217329666018486,
      "num_tokens": 68931016.0,
      "step": 33215
    },
    {
      "entropy": 5.577709531784057,
      "epoch": 3.013425253991292,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041011447356810584,
      "loss": 4.8417,
      "mean_token_accuracy": 0.21917263716459273,
      "num_tokens": 68941489.0,
      "step": 33220
    },
    {
      "entropy": 5.656309318542481,
      "epoch": 3.013878809869376,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004100885993899493,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21367194205522538,
      "num_tokens": 68952322.0,
      "step": 33225
    },
    {
      "entropy": 5.638419771194458,
      "epoch": 3.01433236574746,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004100627224181261,
      "loss": 4.9124,
      "mean_token_accuracy": 0.21681924611330033,
      "num_tokens": 68962570.0,
      "step": 33230
    },
    {
      "entropy": 5.649652481079102,
      "epoch": 3.014785921625544,
      "grad_norm": 1.125,
      "learning_rate": 0.00041003684265317143,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22429651468992234,
      "num_tokens": 68973733.0,
      "step": 33235
    },
    {
      "entropy": 5.605364513397217,
      "epoch": 3.0152394775036284,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041001096009562045,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22386542409658433,
      "num_tokens": 68983526.0,
      "step": 33240
    },
    {
      "entropy": 5.704187202453613,
      "epoch": 3.0156930333817127,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004099850747460084,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22057290226221085,
      "num_tokens": 68993968.0,
      "step": 33245
    },
    {
      "entropy": 5.618980503082275,
      "epoch": 3.016146589259797,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040995918660487055,
      "loss": 4.8377,
      "mean_token_accuracy": 0.23325806260108947,
      "num_tokens": 69003599.0,
      "step": 33250
    },
    {
      "entropy": 5.577439785003662,
      "epoch": 3.0166001451378808,
      "grad_norm": 1.25,
      "learning_rate": 0.0004099332956727424,
      "loss": 4.8673,
      "mean_token_accuracy": 0.21647975891828536,
      "num_tokens": 69013686.0,
      "step": 33255
    },
    {
      "entropy": 5.620430660247803,
      "epoch": 3.017053701015965,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004099074019501593,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21302349269390106,
      "num_tokens": 69023803.0,
      "step": 33260
    },
    {
      "entropy": 5.591452693939209,
      "epoch": 3.0175072568940493,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004098815054376567,
      "loss": 4.7861,
      "mean_token_accuracy": 0.22871512919664383,
      "num_tokens": 69033885.0,
      "step": 33265
    },
    {
      "entropy": 5.589730310440063,
      "epoch": 3.0179608127721336,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004098556061357702,
      "loss": 4.8407,
      "mean_token_accuracy": 0.22262195646762847,
      "num_tokens": 69044471.0,
      "step": 33270
    },
    {
      "entropy": 5.543532085418701,
      "epoch": 3.018414368650218,
      "grad_norm": 1.125,
      "learning_rate": 0.00040982970404503544,
      "loss": 4.7505,
      "mean_token_accuracy": 0.22955181151628495,
      "num_tokens": 69055109.0,
      "step": 33275
    },
    {
      "entropy": 5.545413494110107,
      "epoch": 3.018867924528302,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040980379916598803,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22021372020244598,
      "num_tokens": 69064692.0,
      "step": 33280
    },
    {
      "entropy": 5.613268947601318,
      "epoch": 3.019321480406386,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004097778914991637,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22543514519929886,
      "num_tokens": 69075331.0,
      "step": 33285
    },
    {
      "entropy": 5.618462419509887,
      "epoch": 3.01977503628447,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040975198104509827,
      "loss": 4.8593,
      "mean_token_accuracy": 0.22036763429641723,
      "num_tokens": 69085017.0,
      "step": 33290
    },
    {
      "entropy": 5.548881959915161,
      "epoch": 3.0202285921625545,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004097260678043275,
      "loss": 4.8678,
      "mean_token_accuracy": 0.21486279517412185,
      "num_tokens": 69096128.0,
      "step": 33295
    },
    {
      "entropy": 5.6427265167236325,
      "epoch": 3.0206821480406387,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040970015177738725,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21123552620410918,
      "num_tokens": 69106479.0,
      "step": 33300
    },
    {
      "entropy": 5.710194063186646,
      "epoch": 3.021135703918723,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004096742329648136,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21697250306606292,
      "num_tokens": 69116346.0,
      "step": 33305
    },
    {
      "entropy": 5.629936933517456,
      "epoch": 3.021589259796807,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004096483113671425,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22117064744234086,
      "num_tokens": 69126489.0,
      "step": 33310
    },
    {
      "entropy": 5.57441349029541,
      "epoch": 3.022042815674891,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040962238698490994,
      "loss": 4.8652,
      "mean_token_accuracy": 0.2204676628112793,
      "num_tokens": 69136618.0,
      "step": 33315
    },
    {
      "entropy": 5.649873208999634,
      "epoch": 3.0224963715529753,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004095964598186522,
      "loss": 4.92,
      "mean_token_accuracy": 0.21439631134271622,
      "num_tokens": 69146242.0,
      "step": 33320
    },
    {
      "entropy": 5.623918104171753,
      "epoch": 3.0229499274310596,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004095705298689053,
      "loss": 4.8537,
      "mean_token_accuracy": 0.21990257501602173,
      "num_tokens": 69157269.0,
      "step": 33325
    },
    {
      "entropy": 5.7010791301727295,
      "epoch": 3.023403483309144,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040954459713620563,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21316420137882233,
      "num_tokens": 69169260.0,
      "step": 33330
    },
    {
      "entropy": 5.716715335845947,
      "epoch": 3.0238570391872277,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004095186616210893,
      "loss": 4.9893,
      "mean_token_accuracy": 0.2040004700422287,
      "num_tokens": 69179815.0,
      "step": 33335
    },
    {
      "entropy": 5.6676500797271725,
      "epoch": 3.024310595065312,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004094927233240928,
      "loss": 4.972,
      "mean_token_accuracy": 0.2176794484257698,
      "num_tokens": 69189959.0,
      "step": 33340
    },
    {
      "entropy": 5.628020620346069,
      "epoch": 3.0247641509433962,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040946678224575264,
      "loss": 4.8723,
      "mean_token_accuracy": 0.21981181353330612,
      "num_tokens": 69199485.0,
      "step": 33345
    },
    {
      "entropy": 5.574540615081787,
      "epoch": 3.0252177068214805,
      "grad_norm": 1.3125,
      "learning_rate": 0.000409440838386605,
      "loss": 4.8845,
      "mean_token_accuracy": 0.21889398992061615,
      "num_tokens": 69209180.0,
      "step": 33350
    },
    {
      "entropy": 5.635682249069214,
      "epoch": 3.0256712626995648,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004094148917471866,
      "loss": 4.9266,
      "mean_token_accuracy": 0.21798319220542908,
      "num_tokens": 69220648.0,
      "step": 33355
    },
    {
      "entropy": 5.553546714782715,
      "epoch": 3.0261248185776486,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000409388942328034,
      "loss": 4.7629,
      "mean_token_accuracy": 0.2267109766602516,
      "num_tokens": 69230685.0,
      "step": 33360
    },
    {
      "entropy": 5.553228282928467,
      "epoch": 3.026578374455733,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004093629901296838,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21376700401306153,
      "num_tokens": 69240407.0,
      "step": 33365
    },
    {
      "entropy": 5.718560123443604,
      "epoch": 3.027031930333817,
      "grad_norm": 1.25,
      "learning_rate": 0.0004093370351526727,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21289327889680862,
      "num_tokens": 69250624.0,
      "step": 33370
    },
    {
      "entropy": 5.717520999908447,
      "epoch": 3.0274854862119014,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004093110773975376,
      "loss": 4.9381,
      "mean_token_accuracy": 0.20824849009513854,
      "num_tokens": 69261131.0,
      "step": 33375
    },
    {
      "entropy": 5.592414999008179,
      "epoch": 3.0279390420899857,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004092851168648152,
      "loss": 4.9164,
      "mean_token_accuracy": 0.20945546329021453,
      "num_tokens": 69270668.0,
      "step": 33380
    },
    {
      "entropy": 5.582691621780396,
      "epoch": 3.0283925979680695,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004092591535550422,
      "loss": 4.8308,
      "mean_token_accuracy": 0.22511018961668014,
      "num_tokens": 69280528.0,
      "step": 33385
    },
    {
      "entropy": 5.57068829536438,
      "epoch": 3.0288461538461537,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004092331874687558,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22823573052883148,
      "num_tokens": 69290376.0,
      "step": 33390
    },
    {
      "entropy": 5.638965034484864,
      "epoch": 3.029299709724238,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040920721860649285,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22777412682771683,
      "num_tokens": 69301766.0,
      "step": 33395
    },
    {
      "entropy": 5.620216751098633,
      "epoch": 3.0297532656023223,
      "grad_norm": 1.125,
      "learning_rate": 0.0004091812469687904,
      "loss": 4.8536,
      "mean_token_accuracy": 0.21777494698762895,
      "num_tokens": 69312668.0,
      "step": 33400
    },
    {
      "entropy": 5.6081806182861325,
      "epoch": 3.0302068214804065,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004091552725561856,
      "loss": 4.9184,
      "mean_token_accuracy": 0.20867687910795213,
      "num_tokens": 69323195.0,
      "step": 33405
    },
    {
      "entropy": 5.602424287796021,
      "epoch": 3.0306603773584904,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004091292953692156,
      "loss": 4.8711,
      "mean_token_accuracy": 0.2210137277841568,
      "num_tokens": 69333812.0,
      "step": 33410
    },
    {
      "entropy": 5.621130561828613,
      "epoch": 3.0311139332365746,
      "grad_norm": 1.125,
      "learning_rate": 0.00040910331540841755,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22838663458824157,
      "num_tokens": 69344378.0,
      "step": 33415
    },
    {
      "entropy": 5.587288951873779,
      "epoch": 3.031567489114659,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040907733267432877,
      "loss": 4.8781,
      "mean_token_accuracy": 0.22115831971168518,
      "num_tokens": 69354592.0,
      "step": 33420
    },
    {
      "entropy": 5.628437328338623,
      "epoch": 3.032021044992743,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040905134716748656,
      "loss": 4.8804,
      "mean_token_accuracy": 0.21539197415113448,
      "num_tokens": 69365011.0,
      "step": 33425
    },
    {
      "entropy": 5.571248960494995,
      "epoch": 3.0324746008708274,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004090253588884284,
      "loss": 4.8867,
      "mean_token_accuracy": 0.21312977820634843,
      "num_tokens": 69377010.0,
      "step": 33430
    },
    {
      "entropy": 5.630991649627686,
      "epoch": 3.0329281567489113,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040899936783769157,
      "loss": 4.8894,
      "mean_token_accuracy": 0.2216227889060974,
      "num_tokens": 69387263.0,
      "step": 33435
    },
    {
      "entropy": 5.638337564468384,
      "epoch": 3.0333817126269955,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040897337401581374,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22455921322107314,
      "num_tokens": 69399188.0,
      "step": 33440
    },
    {
      "entropy": 5.621446037292481,
      "epoch": 3.03383526850508,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040894737742333244,
      "loss": 4.8899,
      "mean_token_accuracy": 0.21498363167047502,
      "num_tokens": 69408271.0,
      "step": 33445
    },
    {
      "entropy": 5.679814243316651,
      "epoch": 3.034288824383164,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004089213780607851,
      "loss": 4.9941,
      "mean_token_accuracy": 0.20780993551015853,
      "num_tokens": 69418424.0,
      "step": 33450
    },
    {
      "entropy": 5.655559444427491,
      "epoch": 3.0347423802612483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004088953759287096,
      "loss": 4.8907,
      "mean_token_accuracy": 0.2269052356481552,
      "num_tokens": 69427802.0,
      "step": 33455
    },
    {
      "entropy": 5.616131734848023,
      "epoch": 3.0351959361393326,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004088693710276436,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21780091226100923,
      "num_tokens": 69437581.0,
      "step": 33460
    },
    {
      "entropy": 5.6072611808776855,
      "epoch": 3.0356494920174164,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040884336335812494,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22064584642648696,
      "num_tokens": 69447711.0,
      "step": 33465
    },
    {
      "entropy": 5.674621725082398,
      "epoch": 3.0361030478955007,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004088173529206914,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2255007028579712,
      "num_tokens": 69457692.0,
      "step": 33470
    },
    {
      "entropy": 5.617612743377686,
      "epoch": 3.036556603773585,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004087913397158809,
      "loss": 4.8312,
      "mean_token_accuracy": 0.22081757634878157,
      "num_tokens": 69468034.0,
      "step": 33475
    },
    {
      "entropy": 5.568518733978271,
      "epoch": 3.037010159651669,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040876532374423135,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22349444180727004,
      "num_tokens": 69477719.0,
      "step": 33480
    },
    {
      "entropy": 5.603288698196411,
      "epoch": 3.0374637155297535,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004087393050062809,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21587074398994446,
      "num_tokens": 69487261.0,
      "step": 33485
    },
    {
      "entropy": 5.6113794326782225,
      "epoch": 3.0379172714078373,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040871328350256745,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2233819842338562,
      "num_tokens": 69496658.0,
      "step": 33490
    },
    {
      "entropy": 5.713962650299072,
      "epoch": 3.0383708272859216,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040868725923362927,
      "loss": 5.0183,
      "mean_token_accuracy": 0.2100777268409729,
      "num_tokens": 69507197.0,
      "step": 33495
    },
    {
      "entropy": 5.610123252868652,
      "epoch": 3.038824383164006,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040866123220000455,
      "loss": 4.9274,
      "mean_token_accuracy": 0.21952326893806456,
      "num_tokens": 69516905.0,
      "step": 33500
    },
    {
      "entropy": 5.629584217071534,
      "epoch": 3.03927793904209,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040863520240223144,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22321782410144805,
      "num_tokens": 69526869.0,
      "step": 33505
    },
    {
      "entropy": 5.662524604797364,
      "epoch": 3.0397314949201744,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004086091698408483,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21661264449357986,
      "num_tokens": 69537058.0,
      "step": 33510
    },
    {
      "entropy": 5.664655303955078,
      "epoch": 3.040185050798258,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040858313451639343,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20732590556144714,
      "num_tokens": 69547617.0,
      "step": 33515
    },
    {
      "entropy": 5.64209566116333,
      "epoch": 3.0406386066763424,
      "grad_norm": 1.125,
      "learning_rate": 0.0004085570964294053,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21068992018699645,
      "num_tokens": 69558733.0,
      "step": 33520
    },
    {
      "entropy": 5.717536878585816,
      "epoch": 3.0410921625544267,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004085310555804225,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21742648929357528,
      "num_tokens": 69570011.0,
      "step": 33525
    },
    {
      "entropy": 5.719327163696289,
      "epoch": 3.041545718432511,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040850501196998336,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21230290234088897,
      "num_tokens": 69580894.0,
      "step": 33530
    },
    {
      "entropy": 5.6406091213226315,
      "epoch": 3.0419992743105952,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040847896559862654,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21515671014785767,
      "num_tokens": 69591438.0,
      "step": 33535
    },
    {
      "entropy": 5.613695383071899,
      "epoch": 3.042452830188679,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004084529164668907,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21364197880029678,
      "num_tokens": 69601745.0,
      "step": 33540
    },
    {
      "entropy": 5.658150720596313,
      "epoch": 3.0429063860667633,
      "grad_norm": 1.25,
      "learning_rate": 0.00040842686457531453,
      "loss": 4.8821,
      "mean_token_accuracy": 0.21762502193450928,
      "num_tokens": 69611872.0,
      "step": 33545
    },
    {
      "entropy": 5.639253711700439,
      "epoch": 3.0433599419448476,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004084008099244368,
      "loss": 4.868,
      "mean_token_accuracy": 0.21429318338632583,
      "num_tokens": 69621673.0,
      "step": 33550
    },
    {
      "entropy": 5.562106704711914,
      "epoch": 3.043813497822932,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040837475251479633,
      "loss": 4.814,
      "mean_token_accuracy": 0.2233399495482445,
      "num_tokens": 69632090.0,
      "step": 33555
    },
    {
      "entropy": 5.561004781723023,
      "epoch": 3.044267053701016,
      "grad_norm": 1.15625,
      "learning_rate": 0.000408348692346932,
      "loss": 4.8554,
      "mean_token_accuracy": 0.22518489360809327,
      "num_tokens": 69642920.0,
      "step": 33560
    },
    {
      "entropy": 5.629204845428466,
      "epoch": 3.0447206095791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040832262942138277,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21594253033399582,
      "num_tokens": 69653936.0,
      "step": 33565
    },
    {
      "entropy": 5.595352268218994,
      "epoch": 3.0451741654571842,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004082965637386875,
      "loss": 4.8272,
      "mean_token_accuracy": 0.22017356157302856,
      "num_tokens": 69663362.0,
      "step": 33570
    },
    {
      "entropy": 5.6082963943481445,
      "epoch": 3.0456277213352685,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004082704952993854,
      "loss": 4.8409,
      "mean_token_accuracy": 0.2237624317407608,
      "num_tokens": 69673178.0,
      "step": 33575
    },
    {
      "entropy": 5.668405055999756,
      "epoch": 3.0460812772133528,
      "grad_norm": 1.125,
      "learning_rate": 0.0004082444241040155,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21675197184085845,
      "num_tokens": 69683627.0,
      "step": 33580
    },
    {
      "entropy": 5.593782234191894,
      "epoch": 3.046534833091437,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004082183501531169,
      "loss": 4.907,
      "mean_token_accuracy": 0.2181442439556122,
      "num_tokens": 69694010.0,
      "step": 33585
    },
    {
      "entropy": 5.603965950012207,
      "epoch": 3.046988388969521,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040819227344722893,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21518379151821138,
      "num_tokens": 69704888.0,
      "step": 33590
    },
    {
      "entropy": 5.58859658241272,
      "epoch": 3.047441944847605,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040816619398689074,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22560120820999147,
      "num_tokens": 69714718.0,
      "step": 33595
    },
    {
      "entropy": 5.608825016021728,
      "epoch": 3.0478955007256894,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040814011177264174,
      "loss": 4.8621,
      "mean_token_accuracy": 0.21948602199554443,
      "num_tokens": 69726076.0,
      "step": 33600
    },
    {
      "entropy": 5.568218040466308,
      "epoch": 3.0483490566037736,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040811402680502137,
      "loss": 4.841,
      "mean_token_accuracy": 0.2172714203596115,
      "num_tokens": 69736507.0,
      "step": 33605
    },
    {
      "entropy": 5.5609088897705075,
      "epoch": 3.048802612481858,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00040808793908456885,
      "loss": 4.852,
      "mean_token_accuracy": 0.21985357850790024,
      "num_tokens": 69746061.0,
      "step": 33610
    },
    {
      "entropy": 5.62335467338562,
      "epoch": 3.0492561683599417,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004080618486118241,
      "loss": 4.9231,
      "mean_token_accuracy": 0.21722546815872193,
      "num_tokens": 69756493.0,
      "step": 33615
    },
    {
      "entropy": 5.6466676712036135,
      "epoch": 3.049709724238026,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004080357553873261,
      "loss": 4.8543,
      "mean_token_accuracy": 0.2215881437063217,
      "num_tokens": 69766937.0,
      "step": 33620
    },
    {
      "entropy": 5.594765472412109,
      "epoch": 3.0501632801161103,
      "grad_norm": 1.1875,
      "learning_rate": 0.000408009659411615,
      "loss": 4.8752,
      "mean_token_accuracy": 0.22119349837303162,
      "num_tokens": 69776872.0,
      "step": 33625
    },
    {
      "entropy": 5.598317098617554,
      "epoch": 3.0506168359941945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004079835606852301,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2186738595366478,
      "num_tokens": 69788145.0,
      "step": 33630
    },
    {
      "entropy": 5.614964771270752,
      "epoch": 3.051070391872279,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004079574592087114,
      "loss": 4.8809,
      "mean_token_accuracy": 0.21911578327417375,
      "num_tokens": 69797353.0,
      "step": 33635
    },
    {
      "entropy": 5.58411922454834,
      "epoch": 3.051523947750363,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040793135498259854,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22167650908231734,
      "num_tokens": 69806915.0,
      "step": 33640
    },
    {
      "entropy": 5.590922689437866,
      "epoch": 3.051977503628447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040790524800743134,
      "loss": 4.8762,
      "mean_token_accuracy": 0.21981202960014343,
      "num_tokens": 69817695.0,
      "step": 33645
    },
    {
      "entropy": 5.553254222869873,
      "epoch": 3.052431059506531,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004078791382837498,
      "loss": 4.8028,
      "mean_token_accuracy": 0.2348508670926094,
      "num_tokens": 69827451.0,
      "step": 33650
    },
    {
      "entropy": 5.654099321365356,
      "epoch": 3.0528846153846154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040785302581209374,
      "loss": 4.8927,
      "mean_token_accuracy": 0.22072887271642685,
      "num_tokens": 69837897.0,
      "step": 33655
    },
    {
      "entropy": 5.615628433227539,
      "epoch": 3.0533381712626997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040782691059300326,
      "loss": 4.8573,
      "mean_token_accuracy": 0.22448501437902452,
      "num_tokens": 69847378.0,
      "step": 33660
    },
    {
      "entropy": 5.579566860198975,
      "epoch": 3.053791727140784,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004078007926270185,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2194240555167198,
      "num_tokens": 69857522.0,
      "step": 33665
    },
    {
      "entropy": 5.597421550750733,
      "epoch": 3.0542452830188678,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040777467191467946,
      "loss": 4.8598,
      "mean_token_accuracy": 0.2233879014849663,
      "num_tokens": 69868384.0,
      "step": 33670
    },
    {
      "entropy": 5.6122270107269285,
      "epoch": 3.054698838896952,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004077485484565264,
      "loss": 4.8865,
      "mean_token_accuracy": 0.21967986822128296,
      "num_tokens": 69878761.0,
      "step": 33675
    },
    {
      "entropy": 5.662349510192871,
      "epoch": 3.0551523947750363,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004077224222530994,
      "loss": 4.8609,
      "mean_token_accuracy": 0.21498922854661942,
      "num_tokens": 69888366.0,
      "step": 33680
    },
    {
      "entropy": 5.623615264892578,
      "epoch": 3.0556059506531206,
      "grad_norm": 1.09375,
      "learning_rate": 0.000407696293304939,
      "loss": 4.8847,
      "mean_token_accuracy": 0.21239088475704193,
      "num_tokens": 69898489.0,
      "step": 33685
    },
    {
      "entropy": 5.576485586166382,
      "epoch": 3.056059506531205,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004076701616125854,
      "loss": 4.8471,
      "mean_token_accuracy": 0.21782163083553313,
      "num_tokens": 69908425.0,
      "step": 33690
    },
    {
      "entropy": 5.661604166030884,
      "epoch": 3.0565130624092887,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004076440271765791,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21908647418022156,
      "num_tokens": 69918488.0,
      "step": 33695
    },
    {
      "entropy": 5.531662702560425,
      "epoch": 3.056966618287373,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004076178899974605,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23034947216510773,
      "num_tokens": 69928831.0,
      "step": 33700
    },
    {
      "entropy": 5.609535217285156,
      "epoch": 3.057420174165457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040759175007577017,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21517693400382995,
      "num_tokens": 69938974.0,
      "step": 33705
    },
    {
      "entropy": 5.577897357940674,
      "epoch": 3.0578737300435415,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004075656074120486,
      "loss": 4.8147,
      "mean_token_accuracy": 0.2197770282626152,
      "num_tokens": 69948804.0,
      "step": 33710
    },
    {
      "entropy": 5.640744161605835,
      "epoch": 3.0583272859216257,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004075394620068365,
      "loss": 4.8749,
      "mean_token_accuracy": 0.22303461283445358,
      "num_tokens": 69958945.0,
      "step": 33715
    },
    {
      "entropy": 5.598249816894532,
      "epoch": 3.0587808417997095,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040751331386067456,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22164223492145538,
      "num_tokens": 69968826.0,
      "step": 33720
    },
    {
      "entropy": 5.641958808898925,
      "epoch": 3.059234397677794,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004074871629741035,
      "loss": 4.9306,
      "mean_token_accuracy": 0.21732646077871323,
      "num_tokens": 69978961.0,
      "step": 33725
    },
    {
      "entropy": 5.559868907928466,
      "epoch": 3.059687953555878,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004074610093476641,
      "loss": 4.9092,
      "mean_token_accuracy": 0.218568617105484,
      "num_tokens": 69989848.0,
      "step": 33730
    },
    {
      "entropy": 5.595981121063232,
      "epoch": 3.0601415094339623,
      "grad_norm": 1.25,
      "learning_rate": 0.00040743485298189744,
      "loss": 4.8499,
      "mean_token_accuracy": 0.2210436075925827,
      "num_tokens": 69999607.0,
      "step": 33735
    },
    {
      "entropy": 5.605841207504272,
      "epoch": 3.0605950653120466,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040740869387734415,
      "loss": 4.8988,
      "mean_token_accuracy": 0.21666444838047028,
      "num_tokens": 70010446.0,
      "step": 33740
    },
    {
      "entropy": 5.571853971481323,
      "epoch": 3.0610486211901304,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004073825320345453,
      "loss": 4.8009,
      "mean_token_accuracy": 0.22282686531543733,
      "num_tokens": 70021148.0,
      "step": 33745
    },
    {
      "entropy": 5.654863500595093,
      "epoch": 3.0615021770682147,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040735636745404196,
      "loss": 4.96,
      "mean_token_accuracy": 0.2122356802225113,
      "num_tokens": 70030765.0,
      "step": 33750
    },
    {
      "entropy": 5.662096452713013,
      "epoch": 3.061955732946299,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040733020013637526,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21841022223234177,
      "num_tokens": 70041717.0,
      "step": 33755
    },
    {
      "entropy": 5.68151068687439,
      "epoch": 3.0624092888243832,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040730403008208633,
      "loss": 4.9477,
      "mean_token_accuracy": 0.22137918174266816,
      "num_tokens": 70052178.0,
      "step": 33760
    },
    {
      "entropy": 5.606624031066895,
      "epoch": 3.0628628447024675,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004072778572917163,
      "loss": 4.9038,
      "mean_token_accuracy": 0.21718788743019105,
      "num_tokens": 70062819.0,
      "step": 33765
    },
    {
      "entropy": 5.551634454727173,
      "epoch": 3.0633164005805513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004072516817658065,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21947057396173478,
      "num_tokens": 70073966.0,
      "step": 33770
    },
    {
      "entropy": 5.625006103515625,
      "epoch": 3.0637699564586356,
      "grad_norm": 1.125,
      "learning_rate": 0.00040722550350489825,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21627938151359558,
      "num_tokens": 70083884.0,
      "step": 33775
    },
    {
      "entropy": 5.674850797653198,
      "epoch": 3.06422351233672,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004071993225095327,
      "loss": 4.8703,
      "mean_token_accuracy": 0.21777383536100386,
      "num_tokens": 70094910.0,
      "step": 33780
    },
    {
      "entropy": 5.567288732528686,
      "epoch": 3.064677068214804,
      "grad_norm": 1.375,
      "learning_rate": 0.0004071731387802517,
      "loss": 4.7206,
      "mean_token_accuracy": 0.23919747024774551,
      "num_tokens": 70104480.0,
      "step": 33785
    },
    {
      "entropy": 5.570855808258057,
      "epoch": 3.0651306240928884,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004071469523175964,
      "loss": 4.8594,
      "mean_token_accuracy": 0.21891711354255677,
      "num_tokens": 70114003.0,
      "step": 33790
    },
    {
      "entropy": 5.630770921707153,
      "epoch": 3.065584179970972,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040712076312210844,
      "loss": 4.9218,
      "mean_token_accuracy": 0.21823698580265044,
      "num_tokens": 70123604.0,
      "step": 33795
    },
    {
      "entropy": 5.617650127410888,
      "epoch": 3.0660377358490565,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040709457119432945,
      "loss": 4.926,
      "mean_token_accuracy": 0.21489761173725128,
      "num_tokens": 70133895.0,
      "step": 33800
    },
    {
      "entropy": 5.6311156272888185,
      "epoch": 3.0664912917271407,
      "grad_norm": 1.125,
      "learning_rate": 0.00040706837653480113,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22623943090438842,
      "num_tokens": 70145063.0,
      "step": 33805
    },
    {
      "entropy": 5.623827266693115,
      "epoch": 3.066944847605225,
      "grad_norm": 1.296875,
      "learning_rate": 0.000407042179144065,
      "loss": 4.958,
      "mean_token_accuracy": 0.21199622601270676,
      "num_tokens": 70155456.0,
      "step": 33810
    },
    {
      "entropy": 5.523899555206299,
      "epoch": 3.0673984034833093,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040701597902266306,
      "loss": 4.8075,
      "mean_token_accuracy": 0.2228490799665451,
      "num_tokens": 70165427.0,
      "step": 33815
    },
    {
      "entropy": 5.6857884407043455,
      "epoch": 3.067851959361393,
      "grad_norm": 1.1875,
      "learning_rate": 0.000406989776171137,
      "loss": 4.941,
      "mean_token_accuracy": 0.21535684168338776,
      "num_tokens": 70175161.0,
      "step": 33820
    },
    {
      "entropy": 5.612390184402466,
      "epoch": 3.0683055152394774,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040696357059002877,
      "loss": 4.8945,
      "mean_token_accuracy": 0.223714242875576,
      "num_tokens": 70185026.0,
      "step": 33825
    },
    {
      "entropy": 5.668047618865967,
      "epoch": 3.0687590711175616,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004069373622798802,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22745070308446885,
      "num_tokens": 70195645.0,
      "step": 33830
    },
    {
      "entropy": 5.64789924621582,
      "epoch": 3.069212626995646,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004069111512412334,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21733848601579667,
      "num_tokens": 70206591.0,
      "step": 33835
    },
    {
      "entropy": 5.6326210498809814,
      "epoch": 3.06966618287373,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040688493747463036,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22249459475278854,
      "num_tokens": 70216945.0,
      "step": 33840
    },
    {
      "entropy": 5.555872583389283,
      "epoch": 3.0701197387518144,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004068587209806132,
      "loss": 4.7804,
      "mean_token_accuracy": 0.2328471839427948,
      "num_tokens": 70226983.0,
      "step": 33845
    },
    {
      "entropy": 5.599160385131836,
      "epoch": 3.0705732946298983,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040683250175972414,
      "loss": 4.8636,
      "mean_token_accuracy": 0.21896474659442902,
      "num_tokens": 70237730.0,
      "step": 33850
    },
    {
      "entropy": 5.651721715927124,
      "epoch": 3.0710268505079825,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004068062798125053,
      "loss": 4.9265,
      "mean_token_accuracy": 0.21341179311275482,
      "num_tokens": 70248499.0,
      "step": 33855
    },
    {
      "entropy": 5.599911260604858,
      "epoch": 3.071480406386067,
      "grad_norm": 1.203125,
      "learning_rate": 0.000406780055139499,
      "loss": 4.8763,
      "mean_token_accuracy": 0.21700489968061448,
      "num_tokens": 70259294.0,
      "step": 33860
    },
    {
      "entropy": 5.651598072052002,
      "epoch": 3.071933962264151,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004067538277412476,
      "loss": 4.9458,
      "mean_token_accuracy": 0.21679080724716188,
      "num_tokens": 70269400.0,
      "step": 33865
    },
    {
      "entropy": 5.652189302444458,
      "epoch": 3.0723875181422353,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004067275976182935,
      "loss": 4.9574,
      "mean_token_accuracy": 0.21994108855724334,
      "num_tokens": 70279380.0,
      "step": 33870
    },
    {
      "entropy": 5.640937185287475,
      "epoch": 3.072841074020319,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040670136477117906,
      "loss": 4.8866,
      "mean_token_accuracy": 0.2217855617403984,
      "num_tokens": 70289919.0,
      "step": 33875
    },
    {
      "entropy": 5.699266958236694,
      "epoch": 3.0732946298984034,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040667512920044684,
      "loss": 4.994,
      "mean_token_accuracy": 0.21905118376016616,
      "num_tokens": 70300136.0,
      "step": 33880
    },
    {
      "entropy": 5.597797870635986,
      "epoch": 3.0737481857764877,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040664889090663944,
      "loss": 4.796,
      "mean_token_accuracy": 0.22733922600746154,
      "num_tokens": 70310072.0,
      "step": 33885
    },
    {
      "entropy": 5.6391500473022464,
      "epoch": 3.074201741654572,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040662264989029933,
      "loss": 4.879,
      "mean_token_accuracy": 0.22217219769954683,
      "num_tokens": 70320185.0,
      "step": 33890
    },
    {
      "entropy": 5.617372179031372,
      "epoch": 3.074655297532656,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004065964061519694,
      "loss": 4.8779,
      "mean_token_accuracy": 0.2218205600976944,
      "num_tokens": 70330211.0,
      "step": 33895
    },
    {
      "entropy": 5.638485527038574,
      "epoch": 3.07510885341074,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004065701596921922,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21446337848901748,
      "num_tokens": 70341044.0,
      "step": 33900
    },
    {
      "entropy": 5.650462293624878,
      "epoch": 3.0755624092888243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004065439105115106,
      "loss": 4.9337,
      "mean_token_accuracy": 0.22086102813482283,
      "num_tokens": 70351320.0,
      "step": 33905
    },
    {
      "entropy": 5.602971076965332,
      "epoch": 3.0760159651669086,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040651765861046734,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22849898487329484,
      "num_tokens": 70361214.0,
      "step": 33910
    },
    {
      "entropy": 5.642775344848633,
      "epoch": 3.076469521044993,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040649140398960537,
      "loss": 4.921,
      "mean_token_accuracy": 0.22190209478139877,
      "num_tokens": 70371083.0,
      "step": 33915
    },
    {
      "entropy": 5.638441133499145,
      "epoch": 3.076923076923077,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040646514664946775,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21401370763778688,
      "num_tokens": 70381987.0,
      "step": 33920
    },
    {
      "entropy": 5.58901572227478,
      "epoch": 3.077376632801161,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004064388865905974,
      "loss": 4.8425,
      "mean_token_accuracy": 0.2244996890425682,
      "num_tokens": 70393462.0,
      "step": 33925
    },
    {
      "entropy": 5.579142665863037,
      "epoch": 3.077830188679245,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004064126238135373,
      "loss": 4.7486,
      "mean_token_accuracy": 0.22805921882390975,
      "num_tokens": 70403664.0,
      "step": 33930
    },
    {
      "entropy": 5.600571727752685,
      "epoch": 3.0782837445573294,
      "grad_norm": 1.125,
      "learning_rate": 0.0004063863583188308,
      "loss": 4.9067,
      "mean_token_accuracy": 0.21349358707666397,
      "num_tokens": 70414588.0,
      "step": 33935
    },
    {
      "entropy": 5.571170854568481,
      "epoch": 3.0787373004354137,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004063600901070207,
      "loss": 4.941,
      "mean_token_accuracy": 0.21524872928857802,
      "num_tokens": 70425572.0,
      "step": 33940
    },
    {
      "entropy": 5.604797840118408,
      "epoch": 3.079190856313498,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040633381917865057,
      "loss": 4.8442,
      "mean_token_accuracy": 0.21919606029987335,
      "num_tokens": 70436140.0,
      "step": 33945
    },
    {
      "entropy": 5.580093669891357,
      "epoch": 3.079644412191582,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004063075455342636,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22368747293949126,
      "num_tokens": 70447117.0,
      "step": 33950
    },
    {
      "entropy": 5.623998498916626,
      "epoch": 3.080097968069666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004062812691744032,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2238440290093422,
      "num_tokens": 70457756.0,
      "step": 33955
    },
    {
      "entropy": 5.63613920211792,
      "epoch": 3.0805515239477503,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004062549900996126,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21454330384731293,
      "num_tokens": 70468671.0,
      "step": 33960
    },
    {
      "entropy": 5.634016609191894,
      "epoch": 3.0810050798258346,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004062287083104354,
      "loss": 4.8338,
      "mean_token_accuracy": 0.22401514202356337,
      "num_tokens": 70478212.0,
      "step": 33965
    },
    {
      "entropy": 5.585205030441284,
      "epoch": 3.081458635703919,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040620242380741504,
      "loss": 4.8654,
      "mean_token_accuracy": 0.2190607324242592,
      "num_tokens": 70489062.0,
      "step": 33970
    },
    {
      "entropy": 5.625957870483399,
      "epoch": 3.0819121915820027,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004061761365910951,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22176443785429,
      "num_tokens": 70499732.0,
      "step": 33975
    },
    {
      "entropy": 5.6337282180786135,
      "epoch": 3.082365747460087,
      "grad_norm": 1.25,
      "learning_rate": 0.00040614984666201927,
      "loss": 4.8948,
      "mean_token_accuracy": 0.21288188844919204,
      "num_tokens": 70509777.0,
      "step": 33980
    },
    {
      "entropy": 5.63773512840271,
      "epoch": 3.0828193033381712,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004061235540207312,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21612088531255721,
      "num_tokens": 70520674.0,
      "step": 33985
    },
    {
      "entropy": 5.631599187850952,
      "epoch": 3.0832728592162555,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004060972586677746,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22063713520765305,
      "num_tokens": 70531115.0,
      "step": 33990
    },
    {
      "entropy": 5.584438419342041,
      "epoch": 3.0837264150943398,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004060709606036933,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22172345072031022,
      "num_tokens": 70540726.0,
      "step": 33995
    },
    {
      "entropy": 5.623329210281372,
      "epoch": 3.084179970972424,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004060446598290311,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2237810105085373,
      "num_tokens": 70551227.0,
      "step": 34000
    },
    {
      "entropy": 5.564810132980346,
      "epoch": 3.084633526850508,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000406018356344332,
      "loss": 4.875,
      "mean_token_accuracy": 0.21612243503332138,
      "num_tokens": 70560815.0,
      "step": 34005
    },
    {
      "entropy": 5.666827774047851,
      "epoch": 3.085087082728592,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004059920501501398,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2166752338409424,
      "num_tokens": 70570247.0,
      "step": 34010
    },
    {
      "entropy": 5.6672851085662845,
      "epoch": 3.0855406386066764,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004059657412469987,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21635723114013672,
      "num_tokens": 70581497.0,
      "step": 34015
    },
    {
      "entropy": 5.608234643936157,
      "epoch": 3.0859941944847606,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004059394296354527,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22011708319187165,
      "num_tokens": 70593133.0,
      "step": 34020
    },
    {
      "entropy": 5.609160947799682,
      "epoch": 3.086447750362845,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004059131153160459,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22407814413309096,
      "num_tokens": 70603953.0,
      "step": 34025
    },
    {
      "entropy": 5.560619926452636,
      "epoch": 3.0869013062409287,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004058867982893225,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22599049806594848,
      "num_tokens": 70615292.0,
      "step": 34030
    },
    {
      "entropy": 5.663981866836548,
      "epoch": 3.087354862119013,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004058604785558268,
      "loss": 5.0145,
      "mean_token_accuracy": 0.2074570193886757,
      "num_tokens": 70626681.0,
      "step": 34035
    },
    {
      "entropy": 5.662851715087891,
      "epoch": 3.0878084179970973,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040583415611610305,
      "loss": 4.935,
      "mean_token_accuracy": 0.2140041247010231,
      "num_tokens": 70637339.0,
      "step": 34040
    },
    {
      "entropy": 5.704662322998047,
      "epoch": 3.0882619738751815,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004058078309706956,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22234832793474196,
      "num_tokens": 70647980.0,
      "step": 34045
    },
    {
      "entropy": 5.586560487747192,
      "epoch": 3.088715529753266,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004057815031201488,
      "loss": 4.7292,
      "mean_token_accuracy": 0.23724641501903534,
      "num_tokens": 70657042.0,
      "step": 34050
    },
    {
      "entropy": 5.601404905319214,
      "epoch": 3.0891690856313496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004057551725650073,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21587942391633988,
      "num_tokens": 70668542.0,
      "step": 34055
    },
    {
      "entropy": 5.637851047515869,
      "epoch": 3.089622641509434,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040572883930581544,
      "loss": 4.9602,
      "mean_token_accuracy": 0.21750420480966567,
      "num_tokens": 70679332.0,
      "step": 34060
    },
    {
      "entropy": 5.7466113567352295,
      "epoch": 3.090076197387518,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004057025033431179,
      "loss": 5.0266,
      "mean_token_accuracy": 0.21392195224761962,
      "num_tokens": 70689897.0,
      "step": 34065
    },
    {
      "entropy": 5.632059240341187,
      "epoch": 3.0905297532656024,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040567616467745933,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22339710146188735,
      "num_tokens": 70699965.0,
      "step": 34070
    },
    {
      "entropy": 5.65618200302124,
      "epoch": 3.0909833091436867,
      "grad_norm": 1.25,
      "learning_rate": 0.0004056498233093843,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22152122110128403,
      "num_tokens": 70710583.0,
      "step": 34075
    },
    {
      "entropy": 5.557923030853272,
      "epoch": 3.0914368650217705,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040562347923943766,
      "loss": 4.8173,
      "mean_token_accuracy": 0.23478762060403824,
      "num_tokens": 70721800.0,
      "step": 34080
    },
    {
      "entropy": 5.592049312591553,
      "epoch": 3.0918904208998548,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004055971324681642,
      "loss": 4.9058,
      "mean_token_accuracy": 0.21415081173181533,
      "num_tokens": 70731527.0,
      "step": 34085
    },
    {
      "entropy": 5.653995752334595,
      "epoch": 3.092343976777939,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004055707829961087,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22178112268447875,
      "num_tokens": 70740821.0,
      "step": 34090
    },
    {
      "entropy": 5.630556106567383,
      "epoch": 3.0927975326560233,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040554443082381627,
      "loss": 4.8454,
      "mean_token_accuracy": 0.22050863802433013,
      "num_tokens": 70750506.0,
      "step": 34095
    },
    {
      "entropy": 5.607876491546631,
      "epoch": 3.0932510885341076,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004055180759518315,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22196728885173797,
      "num_tokens": 70762009.0,
      "step": 34100
    },
    {
      "entropy": 5.560287809371948,
      "epoch": 3.0937046444121914,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004054917183806999,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2206227406859398,
      "num_tokens": 70772030.0,
      "step": 34105
    },
    {
      "entropy": 5.648917388916016,
      "epoch": 3.0941582002902757,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040546535811096626,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21337200850248336,
      "num_tokens": 70782539.0,
      "step": 34110
    },
    {
      "entropy": 5.789413785934448,
      "epoch": 3.09461175616836,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040543899514317564,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20537374168634415,
      "num_tokens": 70793433.0,
      "step": 34115
    },
    {
      "entropy": 5.650471019744873,
      "epoch": 3.095065312046444,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004054126294778735,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21279655694961547,
      "num_tokens": 70802919.0,
      "step": 34120
    },
    {
      "entropy": 5.571983909606933,
      "epoch": 3.0955188679245285,
      "grad_norm": 1.125,
      "learning_rate": 0.00040538626111560487,
      "loss": 4.8412,
      "mean_token_accuracy": 0.2223627358675003,
      "num_tokens": 70813827.0,
      "step": 34125
    },
    {
      "entropy": 5.682165908813476,
      "epoch": 3.0959724238026123,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040535989005691515,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2172440230846405,
      "num_tokens": 70824117.0,
      "step": 34130
    },
    {
      "entropy": 5.556173610687256,
      "epoch": 3.0964259796806965,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040533351630234985,
      "loss": 4.8249,
      "mean_token_accuracy": 0.22764098197221755,
      "num_tokens": 70833387.0,
      "step": 34135
    },
    {
      "entropy": 5.600375604629517,
      "epoch": 3.096879535558781,
      "grad_norm": 1.203125,
      "learning_rate": 0.000405307139852454,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22222230583429337,
      "num_tokens": 70844363.0,
      "step": 34140
    },
    {
      "entropy": 5.64898681640625,
      "epoch": 3.097333091436865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040528076070777335,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21835437566041946,
      "num_tokens": 70854759.0,
      "step": 34145
    },
    {
      "entropy": 5.539999198913574,
      "epoch": 3.0977866473149493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004052543788688534,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2176368311047554,
      "num_tokens": 70865682.0,
      "step": 34150
    },
    {
      "entropy": 5.705310201644897,
      "epoch": 3.098240203193033,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004052279943362396,
      "loss": 4.9397,
      "mean_token_accuracy": 0.2154989242553711,
      "num_tokens": 70875449.0,
      "step": 34155
    },
    {
      "entropy": 5.652715063095092,
      "epoch": 3.0986937590711174,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004052016071104779,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2179457113146782,
      "num_tokens": 70885885.0,
      "step": 34160
    },
    {
      "entropy": 5.599583864212036,
      "epoch": 3.0991473149492017,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004051752171921136,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22145263701677323,
      "num_tokens": 70896320.0,
      "step": 34165
    },
    {
      "entropy": 5.568074798583984,
      "epoch": 3.099600870827286,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040514882458169267,
      "loss": 4.8132,
      "mean_token_accuracy": 0.2256337061524391,
      "num_tokens": 70907074.0,
      "step": 34170
    },
    {
      "entropy": 5.605743265151977,
      "epoch": 3.1000544267053702,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004051224292797609,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2163526952266693,
      "num_tokens": 70916567.0,
      "step": 34175
    },
    {
      "entropy": 5.629853200912476,
      "epoch": 3.100507982583454,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040509603128686413,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22729743719100953,
      "num_tokens": 70926163.0,
      "step": 34180
    },
    {
      "entropy": 5.6725012302398685,
      "epoch": 3.1009615384615383,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004050696306035483,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21403191089630128,
      "num_tokens": 70937358.0,
      "step": 34185
    },
    {
      "entropy": 5.5880731582641605,
      "epoch": 3.1014150943396226,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040504322723035937,
      "loss": 4.8772,
      "mean_token_accuracy": 0.2194350019097328,
      "num_tokens": 70949305.0,
      "step": 34190
    },
    {
      "entropy": 5.5348611831665036,
      "epoch": 3.101868650217707,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040501682116784326,
      "loss": 4.8482,
      "mean_token_accuracy": 0.2258891612291336,
      "num_tokens": 70959146.0,
      "step": 34195
    },
    {
      "entropy": 5.546608352661133,
      "epoch": 3.102322206095791,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040499041241654626,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21654050797224045,
      "num_tokens": 70968828.0,
      "step": 34200
    },
    {
      "entropy": 5.587368154525757,
      "epoch": 3.1027757619738754,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004049640009770144,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22193701118230819,
      "num_tokens": 70979733.0,
      "step": 34205
    },
    {
      "entropy": 5.672473382949829,
      "epoch": 3.103229317851959,
      "grad_norm": 1.25,
      "learning_rate": 0.0004049375868497939,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2205413907766342,
      "num_tokens": 70989726.0,
      "step": 34210
    },
    {
      "entropy": 5.656386852264404,
      "epoch": 3.1036828737300435,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040491117003543087,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2169051706790924,
      "num_tokens": 71000715.0,
      "step": 34215
    },
    {
      "entropy": 5.563777923583984,
      "epoch": 3.1041364296081277,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040488475053447186,
      "loss": 4.7298,
      "mean_token_accuracy": 0.23253817111253738,
      "num_tokens": 71011453.0,
      "step": 34220
    },
    {
      "entropy": 5.620495557785034,
      "epoch": 3.104589985486212,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040485832834746306,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21509634107351303,
      "num_tokens": 71021477.0,
      "step": 34225
    },
    {
      "entropy": 5.56983323097229,
      "epoch": 3.1050435413642963,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040483190347495097,
      "loss": 4.8432,
      "mean_token_accuracy": 0.21912508606910705,
      "num_tokens": 71031706.0,
      "step": 34230
    },
    {
      "entropy": 5.647357511520386,
      "epoch": 3.10549709724238,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000404805475917482,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22162728905677795,
      "num_tokens": 71042882.0,
      "step": 34235
    },
    {
      "entropy": 5.629679536819458,
      "epoch": 3.1059506531204644,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040477904567560275,
      "loss": 4.8634,
      "mean_token_accuracy": 0.22511965334415435,
      "num_tokens": 71052804.0,
      "step": 34240
    },
    {
      "entropy": 5.650507593154908,
      "epoch": 3.1064042089985486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004047526127498597,
      "loss": 4.9868,
      "mean_token_accuracy": 0.20449872314929962,
      "num_tokens": 71064349.0,
      "step": 34245
    },
    {
      "entropy": 5.590665483474732,
      "epoch": 3.106857764876633,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040472617714079964,
      "loss": 4.9086,
      "mean_token_accuracy": 0.218218632042408,
      "num_tokens": 71075754.0,
      "step": 34250
    },
    {
      "entropy": 5.642666578292847,
      "epoch": 3.107311320754717,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004046997388489691,
      "loss": 4.8416,
      "mean_token_accuracy": 0.2216945096850395,
      "num_tokens": 71085338.0,
      "step": 34255
    },
    {
      "entropy": 5.558362007141113,
      "epoch": 3.107764876632801,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004046732978749149,
      "loss": 4.8497,
      "mean_token_accuracy": 0.2154309943318367,
      "num_tokens": 71097089.0,
      "step": 34260
    },
    {
      "entropy": 5.532902479171753,
      "epoch": 3.1082184325108853,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000404646854219184,
      "loss": 4.7966,
      "mean_token_accuracy": 0.2322097212076187,
      "num_tokens": 71106990.0,
      "step": 34265
    },
    {
      "entropy": 5.5698882102966305,
      "epoch": 3.1086719883889695,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000404620407882323,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22159672081470488,
      "num_tokens": 71117634.0,
      "step": 34270
    },
    {
      "entropy": 5.54518461227417,
      "epoch": 3.109125544267054,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040459395886487885,
      "loss": 4.8049,
      "mean_token_accuracy": 0.2202684372663498,
      "num_tokens": 71127717.0,
      "step": 34275
    },
    {
      "entropy": 5.603554058074951,
      "epoch": 3.109579100145138,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004045675071673988,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22675131261348724,
      "num_tokens": 71137172.0,
      "step": 34280
    },
    {
      "entropy": 5.651407766342163,
      "epoch": 3.110032656023222,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004045410527904296,
      "loss": 4.9251,
      "mean_token_accuracy": 0.2083875447511673,
      "num_tokens": 71146843.0,
      "step": 34285
    },
    {
      "entropy": 5.554104900360107,
      "epoch": 3.110486211901306,
      "grad_norm": 1.25,
      "learning_rate": 0.0004045145957345184,
      "loss": 4.7769,
      "mean_token_accuracy": 0.23007616251707078,
      "num_tokens": 71156619.0,
      "step": 34290
    },
    {
      "entropy": 5.618606901168823,
      "epoch": 3.1109397677793904,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004044881360002123,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22491240054368972,
      "num_tokens": 71167234.0,
      "step": 34295
    },
    {
      "entropy": 5.615027379989624,
      "epoch": 3.1113933236574747,
      "grad_norm": 1.125,
      "learning_rate": 0.0004044616735880586,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22390307039022445,
      "num_tokens": 71178049.0,
      "step": 34300
    },
    {
      "entropy": 5.585771560668945,
      "epoch": 3.111846879535559,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040443520849860453,
      "loss": 4.81,
      "mean_token_accuracy": 0.22132199108600617,
      "num_tokens": 71188568.0,
      "step": 34305
    },
    {
      "entropy": 5.641154861450195,
      "epoch": 3.1123004354136428,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040440874073239733,
      "loss": 4.9085,
      "mean_token_accuracy": 0.21335671097040176,
      "num_tokens": 71199004.0,
      "step": 34310
    },
    {
      "entropy": 5.570372152328491,
      "epoch": 3.112753991291727,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040438227028998436,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22319533824920654,
      "num_tokens": 71209239.0,
      "step": 34315
    },
    {
      "entropy": 5.619079065322876,
      "epoch": 3.1132075471698113,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040435579717191313,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21719261556863784,
      "num_tokens": 71219469.0,
      "step": 34320
    },
    {
      "entropy": 5.609042549133301,
      "epoch": 3.1136611030478956,
      "grad_norm": 1.25,
      "learning_rate": 0.00040432932137873107,
      "loss": 4.9335,
      "mean_token_accuracy": 0.2149026244878769,
      "num_tokens": 71230670.0,
      "step": 34325
    },
    {
      "entropy": 5.679176712036133,
      "epoch": 3.11411465892598,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004043028429109857,
      "loss": 4.9925,
      "mean_token_accuracy": 0.2172279730439186,
      "num_tokens": 71241512.0,
      "step": 34330
    },
    {
      "entropy": 5.627719497680664,
      "epoch": 3.1145682148040637,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004042763617692245,
      "loss": 4.8422,
      "mean_token_accuracy": 0.22052338123321533,
      "num_tokens": 71250869.0,
      "step": 34335
    },
    {
      "entropy": 5.603387022018433,
      "epoch": 3.115021770682148,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040424987795399523,
      "loss": 4.8329,
      "mean_token_accuracy": 0.2186043456196785,
      "num_tokens": 71260577.0,
      "step": 34340
    },
    {
      "entropy": 5.623979616165161,
      "epoch": 3.115475326560232,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004042233914658455,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2218081012368202,
      "num_tokens": 71270185.0,
      "step": 34345
    },
    {
      "entropy": 5.539167213439941,
      "epoch": 3.1159288824383164,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040419690230532324,
      "loss": 4.8463,
      "mean_token_accuracy": 0.22802520543336868,
      "num_tokens": 71279427.0,
      "step": 34350
    },
    {
      "entropy": 5.584182500839233,
      "epoch": 3.1163824383164007,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040417041047297606,
      "loss": 4.8209,
      "mean_token_accuracy": 0.22426612675189972,
      "num_tokens": 71290857.0,
      "step": 34355
    },
    {
      "entropy": 5.637346792221069,
      "epoch": 3.116835994194485,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004041439159693518,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2221457377076149,
      "num_tokens": 71300971.0,
      "step": 34360
    },
    {
      "entropy": 5.659772205352783,
      "epoch": 3.117289550072569,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004041174187949985,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21609605699777604,
      "num_tokens": 71311017.0,
      "step": 34365
    },
    {
      "entropy": 5.649938869476318,
      "epoch": 3.117743105950653,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040409091895046413,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21190449893474578,
      "num_tokens": 71321546.0,
      "step": 34370
    },
    {
      "entropy": 5.594947814941406,
      "epoch": 3.1181966618287373,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040406441643629656,
      "loss": 4.8969,
      "mean_token_accuracy": 0.21915542632341384,
      "num_tokens": 71332345.0,
      "step": 34375
    },
    {
      "entropy": 5.6774999618530275,
      "epoch": 3.1186502177068216,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040403791125304405,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21980152428150176,
      "num_tokens": 71342670.0,
      "step": 34380
    },
    {
      "entropy": 5.61887640953064,
      "epoch": 3.119103773584906,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004040114034012546,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2188290163874626,
      "num_tokens": 71353440.0,
      "step": 34385
    },
    {
      "entropy": 5.626083898544311,
      "epoch": 3.1195573294629897,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040398489288147646,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21995634585618973,
      "num_tokens": 71363199.0,
      "step": 34390
    },
    {
      "entropy": 5.61260757446289,
      "epoch": 3.120010885341074,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040395837969425786,
      "loss": 4.8815,
      "mean_token_accuracy": 0.21791732609272002,
      "num_tokens": 71372925.0,
      "step": 34395
    },
    {
      "entropy": 5.650451803207398,
      "epoch": 3.1204644412191582,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004039318638401471,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21911511868238448,
      "num_tokens": 71383350.0,
      "step": 34400
    },
    {
      "entropy": 5.613069200515747,
      "epoch": 3.1209179970972425,
      "grad_norm": 1.25,
      "learning_rate": 0.0004039053453196926,
      "loss": 4.8233,
      "mean_token_accuracy": 0.23012498766183853,
      "num_tokens": 71394006.0,
      "step": 34405
    },
    {
      "entropy": 5.558915090560913,
      "epoch": 3.1213715529753268,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004038788241334427,
      "loss": 4.8659,
      "mean_token_accuracy": 0.21994834393262863,
      "num_tokens": 71403945.0,
      "step": 34410
    },
    {
      "entropy": 5.622850179672241,
      "epoch": 3.1218251088534106,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004038523002819458,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21319989562034608,
      "num_tokens": 71414238.0,
      "step": 34415
    },
    {
      "entropy": 5.595477724075318,
      "epoch": 3.122278664731495,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040382577376575055,
      "loss": 4.9167,
      "mean_token_accuracy": 0.21439329236745835,
      "num_tokens": 71423586.0,
      "step": 34420
    },
    {
      "entropy": 5.607227325439453,
      "epoch": 3.122732220609579,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004037992445854054,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22491985261440278,
      "num_tokens": 71434261.0,
      "step": 34425
    },
    {
      "entropy": 5.608119153976441,
      "epoch": 3.1231857764876634,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040377271274145916,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22074744552373887,
      "num_tokens": 71443952.0,
      "step": 34430
    },
    {
      "entropy": 5.66825647354126,
      "epoch": 3.1236393323657476,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004037461782344603,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22169191539287567,
      "num_tokens": 71454051.0,
      "step": 34435
    },
    {
      "entropy": 5.597895765304566,
      "epoch": 3.1240928882438315,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040371964106495777,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2259342685341835,
      "num_tokens": 71464579.0,
      "step": 34440
    },
    {
      "entropy": 5.598602676391602,
      "epoch": 3.1245464441219157,
      "grad_norm": 1.25,
      "learning_rate": 0.00040369310123350015,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22197497338056565,
      "num_tokens": 71474631.0,
      "step": 34445
    },
    {
      "entropy": 5.687271547317505,
      "epoch": 3.125,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040366655874063645,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22455458641052245,
      "num_tokens": 71485266.0,
      "step": 34450
    },
    {
      "entropy": 5.5737861633300785,
      "epoch": 3.1254535558780843,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004036400135869156,
      "loss": 4.7584,
      "mean_token_accuracy": 0.23174856156110762,
      "num_tokens": 71495248.0,
      "step": 34455
    },
    {
      "entropy": 5.585037755966186,
      "epoch": 3.1259071117561685,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040361346577288645,
      "loss": 4.8553,
      "mean_token_accuracy": 0.22350795120000838,
      "num_tokens": 71505600.0,
      "step": 34460
    },
    {
      "entropy": 5.6106870651245115,
      "epoch": 3.1263606676342524,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040358691529909796,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22294523566961288,
      "num_tokens": 71516634.0,
      "step": 34465
    },
    {
      "entropy": 5.68028244972229,
      "epoch": 3.1268142235123366,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040356036216609943,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2163221687078476,
      "num_tokens": 71526745.0,
      "step": 34470
    },
    {
      "entropy": 5.701898574829102,
      "epoch": 3.127267779390421,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040353380637443974,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20294490605592727,
      "num_tokens": 71537015.0,
      "step": 34475
    },
    {
      "entropy": 5.60182147026062,
      "epoch": 3.127721335268505,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040350724792466826,
      "loss": 4.7702,
      "mean_token_accuracy": 0.2262040987610817,
      "num_tokens": 71546660.0,
      "step": 34480
    },
    {
      "entropy": 5.578854274749756,
      "epoch": 3.1281748911465894,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004034806868173341,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22003268897533418,
      "num_tokens": 71557309.0,
      "step": 34485
    },
    {
      "entropy": 5.552659893035889,
      "epoch": 3.1286284470246732,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004034541230529866,
      "loss": 4.7871,
      "mean_token_accuracy": 0.22338122576475145,
      "num_tokens": 71567651.0,
      "step": 34490
    },
    {
      "entropy": 5.569345378875733,
      "epoch": 3.1290820029027575,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040342755663217507,
      "loss": 4.8246,
      "mean_token_accuracy": 0.2244902566075325,
      "num_tokens": 71577538.0,
      "step": 34495
    },
    {
      "entropy": 5.5653339385986325,
      "epoch": 3.1295355587808418,
      "grad_norm": 1.15625,
      "learning_rate": 0.000403400987555449,
      "loss": 4.8409,
      "mean_token_accuracy": 0.2266332596540451,
      "num_tokens": 71588299.0,
      "step": 34500
    },
    {
      "entropy": 5.623280048370361,
      "epoch": 3.129989114658926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004033744158233577,
      "loss": 4.937,
      "mean_token_accuracy": 0.21450992077589034,
      "num_tokens": 71598989.0,
      "step": 34505
    },
    {
      "entropy": 5.644419574737549,
      "epoch": 3.1304426705370103,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040334784143645087,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21343223452568055,
      "num_tokens": 71608596.0,
      "step": 34510
    },
    {
      "entropy": 5.553900718688965,
      "epoch": 3.1308962264150946,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004033212643952779,
      "loss": 4.8266,
      "mean_token_accuracy": 0.2250541105866432,
      "num_tokens": 71619182.0,
      "step": 34515
    },
    {
      "entropy": 5.612643527984619,
      "epoch": 3.1313497822931784,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040329468470038857,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21922181397676468,
      "num_tokens": 71629526.0,
      "step": 34520
    },
    {
      "entropy": 5.663424348831176,
      "epoch": 3.1318033381712627,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004032681023523324,
      "loss": 4.9644,
      "mean_token_accuracy": 0.21724731475114822,
      "num_tokens": 71640146.0,
      "step": 34525
    },
    {
      "entropy": 5.621704196929931,
      "epoch": 3.132256894049347,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004032415173516592,
      "loss": 4.8004,
      "mean_token_accuracy": 0.22204351276159287,
      "num_tokens": 71649652.0,
      "step": 34530
    },
    {
      "entropy": 5.61730318069458,
      "epoch": 3.132710449927431,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004032149296989187,
      "loss": 4.8867,
      "mean_token_accuracy": 0.22444941401481627,
      "num_tokens": 71659264.0,
      "step": 34535
    },
    {
      "entropy": 5.61833643913269,
      "epoch": 3.133164005805515,
      "grad_norm": 1.25,
      "learning_rate": 0.00040318833939466083,
      "loss": 4.8903,
      "mean_token_accuracy": 0.2178942382335663,
      "num_tokens": 71669175.0,
      "step": 34540
    },
    {
      "entropy": 5.572068166732788,
      "epoch": 3.1336175616835993,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004031617464394354,
      "loss": 4.8863,
      "mean_token_accuracy": 0.22315966784954072,
      "num_tokens": 71679537.0,
      "step": 34545
    },
    {
      "entropy": 5.625502586364746,
      "epoch": 3.1340711175616836,
      "grad_norm": 1.25,
      "learning_rate": 0.0004031351508337925,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2195585012435913,
      "num_tokens": 71690024.0,
      "step": 34550
    },
    {
      "entropy": 5.686118745803833,
      "epoch": 3.134524673439768,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004031085525782819,
      "loss": 4.8901,
      "mean_token_accuracy": 0.21952289789915086,
      "num_tokens": 71700019.0,
      "step": 34555
    },
    {
      "entropy": 5.644762468338013,
      "epoch": 3.134978229317852,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004030819516734539,
      "loss": 4.9301,
      "mean_token_accuracy": 0.20568177253007888,
      "num_tokens": 71710570.0,
      "step": 34560
    },
    {
      "entropy": 5.609179210662842,
      "epoch": 3.1354317851959363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040305534811985844,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22606145441532136,
      "num_tokens": 71720619.0,
      "step": 34565
    },
    {
      "entropy": 5.5872642993927,
      "epoch": 3.13588534107402,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004030287419180457,
      "loss": 4.8749,
      "mean_token_accuracy": 0.22151790261268617,
      "num_tokens": 71730201.0,
      "step": 34570
    },
    {
      "entropy": 5.643096542358398,
      "epoch": 3.1363388969521044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040300213306856606,
      "loss": 4.8998,
      "mean_token_accuracy": 0.22392575144767762,
      "num_tokens": 71741368.0,
      "step": 34575
    },
    {
      "entropy": 5.585963392257691,
      "epoch": 3.1367924528301887,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004029755215719696,
      "loss": 4.8117,
      "mean_token_accuracy": 0.22652245461940765,
      "num_tokens": 71751361.0,
      "step": 34580
    },
    {
      "entropy": 5.626019906997681,
      "epoch": 3.137246008708273,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040294890742880677,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2234893262386322,
      "num_tokens": 71761164.0,
      "step": 34585
    },
    {
      "entropy": 5.591586017608643,
      "epoch": 3.1376995645863572,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040292229063962787,
      "loss": 4.8618,
      "mean_token_accuracy": 0.22718488723039626,
      "num_tokens": 71770726.0,
      "step": 34590
    },
    {
      "entropy": 5.636309623718262,
      "epoch": 3.138153120464441,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004028956712049834,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22407835274934768,
      "num_tokens": 71781124.0,
      "step": 34595
    },
    {
      "entropy": 5.6022295475006105,
      "epoch": 3.1386066763425253,
      "grad_norm": 1.15625,
      "learning_rate": 0.000402869049125424,
      "loss": 4.8514,
      "mean_token_accuracy": 0.21854931563138963,
      "num_tokens": 71791218.0,
      "step": 34600
    },
    {
      "entropy": 5.610520124435425,
      "epoch": 3.1390602322206096,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004028424244015,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22201108783483506,
      "num_tokens": 71801013.0,
      "step": 34605
    },
    {
      "entropy": 5.637994575500488,
      "epoch": 3.139513788098694,
      "grad_norm": 1.171875,
      "learning_rate": 0.000402815797033762,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2094426929950714,
      "num_tokens": 71811742.0,
      "step": 34610
    },
    {
      "entropy": 5.560222387313843,
      "epoch": 3.139967343976778,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004027891670227609,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2140335962176323,
      "num_tokens": 71822265.0,
      "step": 34615
    },
    {
      "entropy": 5.627309799194336,
      "epoch": 3.140420899854862,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004027625343690472,
      "loss": 4.8382,
      "mean_token_accuracy": 0.219781094789505,
      "num_tokens": 71832090.0,
      "step": 34620
    },
    {
      "entropy": 5.603528928756714,
      "epoch": 3.140874455732946,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004027358990731717,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22248990088701248,
      "num_tokens": 71841381.0,
      "step": 34625
    },
    {
      "entropy": 5.588434362411499,
      "epoch": 3.1413280116110305,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040270926113568524,
      "loss": 4.8842,
      "mean_token_accuracy": 0.21590789556503295,
      "num_tokens": 71851383.0,
      "step": 34630
    },
    {
      "entropy": 5.592743062973023,
      "epoch": 3.1417815674891147,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040268262055713873,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22258922159671785,
      "num_tokens": 71862588.0,
      "step": 34635
    },
    {
      "entropy": 5.606044101715088,
      "epoch": 3.142235123367199,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004026559773380831,
      "loss": 4.848,
      "mean_token_accuracy": 0.22057845443487167,
      "num_tokens": 71873975.0,
      "step": 34640
    },
    {
      "entropy": 5.683474063873291,
      "epoch": 3.142688679245283,
      "grad_norm": 1.375,
      "learning_rate": 0.0004026293314790694,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2113919660449028,
      "num_tokens": 71884084.0,
      "step": 34645
    },
    {
      "entropy": 5.585837554931641,
      "epoch": 3.143142235123367,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004026026829806485,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21480174064636232,
      "num_tokens": 71894643.0,
      "step": 34650
    },
    {
      "entropy": 5.5260179996490475,
      "epoch": 3.1435957910014514,
      "grad_norm": 1.25,
      "learning_rate": 0.00040257603184337167,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22148592174053192,
      "num_tokens": 71905245.0,
      "step": 34655
    },
    {
      "entropy": 5.635767316818237,
      "epoch": 3.1440493468795356,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040254937806778995,
      "loss": 4.9362,
      "mean_token_accuracy": 0.2123051702976227,
      "num_tokens": 71916473.0,
      "step": 34660
    },
    {
      "entropy": 5.624259424209595,
      "epoch": 3.14450290275762,
      "grad_norm": 1.125,
      "learning_rate": 0.00040252272165445456,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22096159309148788,
      "num_tokens": 71927465.0,
      "step": 34665
    },
    {
      "entropy": 5.640947771072388,
      "epoch": 3.1449564586357037,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004024960626039168,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22290985137224198,
      "num_tokens": 71937380.0,
      "step": 34670
    },
    {
      "entropy": 5.600855922698974,
      "epoch": 3.145410014513788,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040246940091672806,
      "loss": 4.9119,
      "mean_token_accuracy": 0.2250762850046158,
      "num_tokens": 71948088.0,
      "step": 34675
    },
    {
      "entropy": 5.62983775138855,
      "epoch": 3.1458635703918723,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004024427365934395,
      "loss": 4.838,
      "mean_token_accuracy": 0.22383915185928344,
      "num_tokens": 71958284.0,
      "step": 34680
    },
    {
      "entropy": 5.612975645065307,
      "epoch": 3.1463171262699565,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040241606963460283,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21289132982492448,
      "num_tokens": 71968138.0,
      "step": 34685
    },
    {
      "entropy": 5.664534997940064,
      "epoch": 3.146770682148041,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040238940004076924,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21742864400148393,
      "num_tokens": 71978275.0,
      "step": 34690
    },
    {
      "entropy": 5.586671018600464,
      "epoch": 3.1472242380261246,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040236272781249047,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2201538622379303,
      "num_tokens": 71988875.0,
      "step": 34695
    },
    {
      "entropy": 5.638438892364502,
      "epoch": 3.147677793904209,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040233605295031795,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2229277342557907,
      "num_tokens": 71999560.0,
      "step": 34700
    },
    {
      "entropy": 5.638119268417358,
      "epoch": 3.148131349782293,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004023093754548034,
      "loss": 4.889,
      "mean_token_accuracy": 0.22454825639724732,
      "num_tokens": 72009970.0,
      "step": 34705
    },
    {
      "entropy": 5.624862957000732,
      "epoch": 3.1485849056603774,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004022826953264986,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22717690467834473,
      "num_tokens": 72020196.0,
      "step": 34710
    },
    {
      "entropy": 5.586294746398925,
      "epoch": 3.1490384615384617,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040225601256595525,
      "loss": 4.8148,
      "mean_token_accuracy": 0.22161322236061096,
      "num_tokens": 72031456.0,
      "step": 34715
    },
    {
      "entropy": 5.622655630111694,
      "epoch": 3.149492017416546,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000402229327173725,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2180787667632103,
      "num_tokens": 72042260.0,
      "step": 34720
    },
    {
      "entropy": 5.698824262619018,
      "epoch": 3.1499455732946298,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040220263915035997,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20166546255350112,
      "num_tokens": 72053463.0,
      "step": 34725
    },
    {
      "entropy": 5.749280071258545,
      "epoch": 3.150399129172714,
      "grad_norm": 1.25,
      "learning_rate": 0.00040217594849641195,
      "loss": 5.001,
      "mean_token_accuracy": 0.20983909219503402,
      "num_tokens": 72064508.0,
      "step": 34730
    },
    {
      "entropy": 5.576893901824951,
      "epoch": 3.1508526850507983,
      "grad_norm": 1.25,
      "learning_rate": 0.00040214925521243287,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21324467211961745,
      "num_tokens": 72074758.0,
      "step": 34735
    },
    {
      "entropy": 5.597273683547973,
      "epoch": 3.1513062409288826,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040212255929897474,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22307022362947465,
      "num_tokens": 72085548.0,
      "step": 34740
    },
    {
      "entropy": 5.614319133758545,
      "epoch": 3.1517597968069664,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004020958607565897,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2147376850247383,
      "num_tokens": 72096773.0,
      "step": 34745
    },
    {
      "entropy": 5.632926559448242,
      "epoch": 3.1522133526850507,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004020691595858299,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21896272301673889,
      "num_tokens": 72106832.0,
      "step": 34750
    },
    {
      "entropy": 5.633471155166626,
      "epoch": 3.152666908563135,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040204245578724754,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21699773967266084,
      "num_tokens": 72117758.0,
      "step": 34755
    },
    {
      "entropy": 5.638801765441895,
      "epoch": 3.153120464441219,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040201574936139475,
      "loss": 4.8491,
      "mean_token_accuracy": 0.2162176862359047,
      "num_tokens": 72127522.0,
      "step": 34760
    },
    {
      "entropy": 5.565290021896362,
      "epoch": 3.1535740203193035,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004019890403088239,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22387027591466904,
      "num_tokens": 72138310.0,
      "step": 34765
    },
    {
      "entropy": 5.589483213424683,
      "epoch": 3.1540275761973877,
      "grad_norm": 1.25,
      "learning_rate": 0.00040196232863008735,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22281512171030043,
      "num_tokens": 72148958.0,
      "step": 34770
    },
    {
      "entropy": 5.669883108139038,
      "epoch": 3.1544811320754715,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040193561432573754,
      "loss": 4.8814,
      "mean_token_accuracy": 0.22320373654365538,
      "num_tokens": 72158594.0,
      "step": 34775
    },
    {
      "entropy": 5.688465309143067,
      "epoch": 3.154934687953556,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004019088973963268,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2170100286602974,
      "num_tokens": 72168552.0,
      "step": 34780
    },
    {
      "entropy": 5.604195976257325,
      "epoch": 3.15538824383164,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040188217784240785,
      "loss": 4.8433,
      "mean_token_accuracy": 0.222317111492157,
      "num_tokens": 72178584.0,
      "step": 34785
    },
    {
      "entropy": 5.595452451705933,
      "epoch": 3.1558417997097243,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000401855455664533,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22452037483453752,
      "num_tokens": 72189144.0,
      "step": 34790
    },
    {
      "entropy": 5.6572870254516605,
      "epoch": 3.1562953555878086,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004018287308632551,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22017716318368913,
      "num_tokens": 72198326.0,
      "step": 34795
    },
    {
      "entropy": 5.655667352676391,
      "epoch": 3.1567489114658924,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004018020034391266,
      "loss": 4.8841,
      "mean_token_accuracy": 0.21748037934303283,
      "num_tokens": 72207916.0,
      "step": 34800
    },
    {
      "entropy": 5.657133388519287,
      "epoch": 3.1572024673439767,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004017752733927004,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22074204981327056,
      "num_tokens": 72219111.0,
      "step": 34805
    },
    {
      "entropy": 5.62120566368103,
      "epoch": 3.157656023222061,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004017485407245293,
      "loss": 4.91,
      "mean_token_accuracy": 0.22101642936468124,
      "num_tokens": 72230081.0,
      "step": 34810
    },
    {
      "entropy": 5.5638477325439455,
      "epoch": 3.1581095791001452,
      "grad_norm": 1.296875,
      "learning_rate": 0.000401721805435166,
      "loss": 4.8412,
      "mean_token_accuracy": 0.22440837621688842,
      "num_tokens": 72240693.0,
      "step": 34815
    },
    {
      "entropy": 5.526685953140259,
      "epoch": 3.1585631349782295,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040169506752516356,
      "loss": 4.8116,
      "mean_token_accuracy": 0.2326302409172058,
      "num_tokens": 72250925.0,
      "step": 34820
    },
    {
      "entropy": 5.547344970703125,
      "epoch": 3.1590166908563133,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004016683269950748,
      "loss": 4.8229,
      "mean_token_accuracy": 0.2293567731976509,
      "num_tokens": 72261027.0,
      "step": 34825
    },
    {
      "entropy": 5.63401837348938,
      "epoch": 3.1594702467343976,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040164158384545267,
      "loss": 4.7801,
      "mean_token_accuracy": 0.2347550079226494,
      "num_tokens": 72269430.0,
      "step": 34830
    },
    {
      "entropy": 5.541888332366943,
      "epoch": 3.159923802612482,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004016148380768504,
      "loss": 4.8173,
      "mean_token_accuracy": 0.22614311128854753,
      "num_tokens": 72280207.0,
      "step": 34835
    },
    {
      "entropy": 5.629393196105957,
      "epoch": 3.160377358490566,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040158808968982103,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22297317683696746,
      "num_tokens": 72290093.0,
      "step": 34840
    },
    {
      "entropy": 5.591180944442749,
      "epoch": 3.1608309143686504,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040156133868491766,
      "loss": 4.8083,
      "mean_token_accuracy": 0.22583741843700408,
      "num_tokens": 72300203.0,
      "step": 34845
    },
    {
      "entropy": 5.624714422225952,
      "epoch": 3.161284470246734,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040153458506269365,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21573832333087922,
      "num_tokens": 72311778.0,
      "step": 34850
    },
    {
      "entropy": 5.624976778030396,
      "epoch": 3.1617380261248185,
      "grad_norm": 1.296875,
      "learning_rate": 0.000401507828823702,
      "loss": 4.862,
      "mean_token_accuracy": 0.22003922164440154,
      "num_tokens": 72322142.0,
      "step": 34855
    },
    {
      "entropy": 5.652061033248901,
      "epoch": 3.1621915820029027,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040148106996849636,
      "loss": 4.939,
      "mean_token_accuracy": 0.21862932592630385,
      "num_tokens": 72332312.0,
      "step": 34860
    },
    {
      "entropy": 5.522206974029541,
      "epoch": 3.162645137880987,
      "grad_norm": 1.25,
      "learning_rate": 0.00040145430849762993,
      "loss": 4.8237,
      "mean_token_accuracy": 0.2223895251750946,
      "num_tokens": 72342157.0,
      "step": 34865
    },
    {
      "entropy": 5.633262014389038,
      "epoch": 3.1630986937590713,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040142754441165615,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21996026039123534,
      "num_tokens": 72353417.0,
      "step": 34870
    },
    {
      "entropy": 5.666251230239868,
      "epoch": 3.1635522496371555,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004014007777111285,
      "loss": 4.8583,
      "mean_token_accuracy": 0.22641475796699523,
      "num_tokens": 72363744.0,
      "step": 34875
    },
    {
      "entropy": 5.633161830902099,
      "epoch": 3.1640058055152394,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040137400839660066,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21772806495428085,
      "num_tokens": 72375241.0,
      "step": 34880
    },
    {
      "entropy": 5.6322822093963625,
      "epoch": 3.1644593613933236,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040134723646862607,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21977696418762208,
      "num_tokens": 72385475.0,
      "step": 34885
    },
    {
      "entropy": 5.613815355300903,
      "epoch": 3.164912917271408,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040132046192775835,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21981662660837173,
      "num_tokens": 72396417.0,
      "step": 34890
    },
    {
      "entropy": 5.6317102909088135,
      "epoch": 3.165366473149492,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004012936847745514,
      "loss": 4.9272,
      "mean_token_accuracy": 0.22171031087636947,
      "num_tokens": 72406169.0,
      "step": 34895
    },
    {
      "entropy": 5.636847257614136,
      "epoch": 3.165820029027576,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004012669050095589,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21598950028419495,
      "num_tokens": 72416522.0,
      "step": 34900
    },
    {
      "entropy": 5.657953691482544,
      "epoch": 3.1662735849056602,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004012401226333345,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2198248729109764,
      "num_tokens": 72426000.0,
      "step": 34905
    },
    {
      "entropy": 5.639194393157959,
      "epoch": 3.1667271407837445,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004012133376464322,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2110868737101555,
      "num_tokens": 72435952.0,
      "step": 34910
    },
    {
      "entropy": 5.650312232971191,
      "epoch": 3.1671806966618288,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040118655004940597,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2171285942196846,
      "num_tokens": 72449957.0,
      "step": 34915
    },
    {
      "entropy": 5.622080850601196,
      "epoch": 3.167634252539913,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004011597598428097,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2260361298918724,
      "num_tokens": 72459956.0,
      "step": 34920
    },
    {
      "entropy": 5.588940715789795,
      "epoch": 3.1680878084179973,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004011329670271975,
      "loss": 4.8308,
      "mean_token_accuracy": 0.2241721048951149,
      "num_tokens": 72471107.0,
      "step": 34925
    },
    {
      "entropy": 5.5678815841674805,
      "epoch": 3.168541364296081,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040110617160312334,
      "loss": 4.8459,
      "mean_token_accuracy": 0.22342223376035691,
      "num_tokens": 72481428.0,
      "step": 34930
    },
    {
      "entropy": 5.6213951587677,
      "epoch": 3.1689949201741654,
      "grad_norm": 1.25,
      "learning_rate": 0.0004010793735711415,
      "loss": 4.928,
      "mean_token_accuracy": 0.21212025880813598,
      "num_tokens": 72491133.0,
      "step": 34935
    },
    {
      "entropy": 5.549967670440674,
      "epoch": 3.1694484760522497,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040105257293180605,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22898245602846146,
      "num_tokens": 72502047.0,
      "step": 34940
    },
    {
      "entropy": 5.6256640434265135,
      "epoch": 3.169902031930334,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004010257696856712,
      "loss": 4.8881,
      "mean_token_accuracy": 0.21532421112060546,
      "num_tokens": 72513120.0,
      "step": 34945
    },
    {
      "entropy": 5.688839244842529,
      "epoch": 3.170355587808418,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040099896383329135,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22066273391246796,
      "num_tokens": 72523118.0,
      "step": 34950
    },
    {
      "entropy": 5.685703706741333,
      "epoch": 3.170809143686502,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040097215537522086,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21737915724515916,
      "num_tokens": 72533697.0,
      "step": 34955
    },
    {
      "entropy": 5.54258279800415,
      "epoch": 3.1712626995645863,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040094534431201406,
      "loss": 4.8351,
      "mean_token_accuracy": 0.23132589757442473,
      "num_tokens": 72543661.0,
      "step": 34960
    },
    {
      "entropy": 5.628453111648559,
      "epoch": 3.1717162554426706,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004009185306442255,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2180483266711235,
      "num_tokens": 72554551.0,
      "step": 34965
    },
    {
      "entropy": 5.649918031692505,
      "epoch": 3.172169811320755,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040089171437240954,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21000791490077972,
      "num_tokens": 72565134.0,
      "step": 34970
    },
    {
      "entropy": 5.615237665176392,
      "epoch": 3.172623367198839,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040086489549712097,
      "loss": 4.896,
      "mean_token_accuracy": 0.2244971662759781,
      "num_tokens": 72574981.0,
      "step": 34975
    },
    {
      "entropy": 5.63043999671936,
      "epoch": 3.173076923076923,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004008380740189142,
      "loss": 4.8965,
      "mean_token_accuracy": 0.2216053009033203,
      "num_tokens": 72585073.0,
      "step": 34980
    },
    {
      "entropy": 5.680956888198852,
      "epoch": 3.173530478955007,
      "grad_norm": 1.203125,
      "learning_rate": 0.000400811249938344,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2270319476723671,
      "num_tokens": 72595719.0,
      "step": 34985
    },
    {
      "entropy": 5.691639041900634,
      "epoch": 3.1739840348330914,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004007844232559652,
      "loss": 4.8925,
      "mean_token_accuracy": 0.21655526012182236,
      "num_tokens": 72605276.0,
      "step": 34990
    },
    {
      "entropy": 5.660261392593384,
      "epoch": 3.1744375907111757,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004007575939723324,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21795331686735153,
      "num_tokens": 72614964.0,
      "step": 34995
    },
    {
      "entropy": 5.660168313980103,
      "epoch": 3.17489114658926,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004007307620880004,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2194843575358391,
      "num_tokens": 72625461.0,
      "step": 35000
    },
    {
      "entropy": 5.631026840209961,
      "epoch": 3.175344702467344,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040070392760352424,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22143623232841492,
      "num_tokens": 72636192.0,
      "step": 35005
    },
    {
      "entropy": 5.669824457168579,
      "epoch": 3.175798258345428,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040067709051945895,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2148689240217209,
      "num_tokens": 72648432.0,
      "step": 35010
    },
    {
      "entropy": 5.632280921936035,
      "epoch": 3.1762518142235123,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004006502508363593,
      "loss": 4.8632,
      "mean_token_accuracy": 0.2156403437256813,
      "num_tokens": 72658678.0,
      "step": 35015
    },
    {
      "entropy": 5.712354135513306,
      "epoch": 3.1767053701015966,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040062340855478046,
      "loss": 5.0534,
      "mean_token_accuracy": 0.2102057456970215,
      "num_tokens": 72669191.0,
      "step": 35020
    },
    {
      "entropy": 5.638961601257324,
      "epoch": 3.177158925979681,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040059656367527754,
      "loss": 4.8681,
      "mean_token_accuracy": 0.2249057933688164,
      "num_tokens": 72679327.0,
      "step": 35025
    },
    {
      "entropy": 5.633044338226318,
      "epoch": 3.1776124818577647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004005697161984056,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21578340381383895,
      "num_tokens": 72689673.0,
      "step": 35030
    },
    {
      "entropy": 5.614214658737183,
      "epoch": 3.178066037735849,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040054286612472006,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20773992985486983,
      "num_tokens": 72700606.0,
      "step": 35035
    },
    {
      "entropy": 5.715998983383178,
      "epoch": 3.178519593613933,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040051601345477593,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2197418689727783,
      "num_tokens": 72711195.0,
      "step": 35040
    },
    {
      "entropy": 5.6628399848937985,
      "epoch": 3.1789731494920175,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040048915818912875,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2142451986670494,
      "num_tokens": 72721160.0,
      "step": 35045
    },
    {
      "entropy": 5.606712341308594,
      "epoch": 3.1794267053701017,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040046230032833374,
      "loss": 4.9254,
      "mean_token_accuracy": 0.21797827780246734,
      "num_tokens": 72730867.0,
      "step": 35050
    },
    {
      "entropy": 5.594057321548462,
      "epoch": 3.1798802612481856,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004004354398729464,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21410192549228668,
      "num_tokens": 72740453.0,
      "step": 35055
    },
    {
      "entropy": 5.670143795013428,
      "epoch": 3.18033381712627,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004004085768235222,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2144264131784439,
      "num_tokens": 72751395.0,
      "step": 35060
    },
    {
      "entropy": 5.630158948898315,
      "epoch": 3.180787373004354,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040038171118061673,
      "loss": 4.8692,
      "mean_token_accuracy": 0.22281871140003204,
      "num_tokens": 72761300.0,
      "step": 35065
    },
    {
      "entropy": 5.594234037399292,
      "epoch": 3.1812409288824384,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004003548429447855,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22644988000392913,
      "num_tokens": 72772011.0,
      "step": 35070
    },
    {
      "entropy": 5.634019613265991,
      "epoch": 3.1816944847605226,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004003279721165841,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21830470860004425,
      "num_tokens": 72782521.0,
      "step": 35075
    },
    {
      "entropy": 5.5902174949646,
      "epoch": 3.182148040638607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004003010986965683,
      "loss": 4.8103,
      "mean_token_accuracy": 0.22426186203956605,
      "num_tokens": 72793125.0,
      "step": 35080
    },
    {
      "entropy": 5.6961350440979,
      "epoch": 3.1826015965166907,
      "grad_norm": 1.234375,
      "learning_rate": 0.000400274222685294,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2188430055975914,
      "num_tokens": 72804541.0,
      "step": 35085
    },
    {
      "entropy": 5.7512688636779785,
      "epoch": 3.183055152394775,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040024734408331664,
      "loss": 4.9949,
      "mean_token_accuracy": 0.22102211117744447,
      "num_tokens": 72814447.0,
      "step": 35090
    },
    {
      "entropy": 5.653251791000367,
      "epoch": 3.1835087082728593,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040022046289119246,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22518065571784973,
      "num_tokens": 72823545.0,
      "step": 35095
    },
    {
      "entropy": 5.610453510284424,
      "epoch": 3.1839622641509435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040019357910947714,
      "loss": 4.7855,
      "mean_token_accuracy": 0.22406467497348787,
      "num_tokens": 72833803.0,
      "step": 35100
    },
    {
      "entropy": 5.680115413665772,
      "epoch": 3.1844158200290273,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040016669273872666,
      "loss": 4.9328,
      "mean_token_accuracy": 0.2164497271180153,
      "num_tokens": 72843383.0,
      "step": 35105
    },
    {
      "entropy": 5.7054911136627195,
      "epoch": 3.1848693759071116,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040013980377949716,
      "loss": 4.9762,
      "mean_token_accuracy": 0.21454016566276551,
      "num_tokens": 72853285.0,
      "step": 35110
    },
    {
      "entropy": 5.637636518478393,
      "epoch": 3.185322931785196,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040011291223234456,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21610147058963775,
      "num_tokens": 72863556.0,
      "step": 35115
    },
    {
      "entropy": 5.640329504013062,
      "epoch": 3.18577648766328,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040008601809782507,
      "loss": 4.8336,
      "mean_token_accuracy": 0.2251451939344406,
      "num_tokens": 72873296.0,
      "step": 35120
    },
    {
      "entropy": 5.668989276885986,
      "epoch": 3.1862300435413644,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004000591213764948,
      "loss": 4.9825,
      "mean_token_accuracy": 0.20847248286008835,
      "num_tokens": 72884738.0,
      "step": 35125
    },
    {
      "entropy": 5.691875886917114,
      "epoch": 3.1866835994194487,
      "grad_norm": 1.453125,
      "learning_rate": 0.00040003222206891017,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20698761194944382,
      "num_tokens": 72894573.0,
      "step": 35130
    },
    {
      "entropy": 5.567454862594604,
      "epoch": 3.1871371552975325,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004000053201756272,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22272568494081496,
      "num_tokens": 72904976.0,
      "step": 35135
    },
    {
      "entropy": 5.612062549591064,
      "epoch": 3.1875907111756168,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003999784156972024,
      "loss": 4.9129,
      "mean_token_accuracy": 0.21314848512411116,
      "num_tokens": 72915152.0,
      "step": 35140
    },
    {
      "entropy": 5.6872917175292965,
      "epoch": 3.188044267053701,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003999515086341921,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21276690661907197,
      "num_tokens": 72925450.0,
      "step": 35145
    },
    {
      "entropy": 5.622509717941284,
      "epoch": 3.1884978229317853,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003999245989871527,
      "loss": 4.8371,
      "mean_token_accuracy": 0.22074327170848845,
      "num_tokens": 72935990.0,
      "step": 35150
    },
    {
      "entropy": 5.610948801040649,
      "epoch": 3.1889513788098696,
      "grad_norm": 1.25,
      "learning_rate": 0.00039989768675664084,
      "loss": 4.7773,
      "mean_token_accuracy": 0.22453771084547042,
      "num_tokens": 72945662.0,
      "step": 35155
    },
    {
      "entropy": 5.562100028991699,
      "epoch": 3.1894049346879534,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039987077194321305,
      "loss": 4.7521,
      "mean_token_accuracy": 0.2287294700741768,
      "num_tokens": 72955313.0,
      "step": 35160
    },
    {
      "entropy": 5.677704954147339,
      "epoch": 3.1898584905660377,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039984385454742576,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20997219979763032,
      "num_tokens": 72965347.0,
      "step": 35165
    },
    {
      "entropy": 5.59677267074585,
      "epoch": 3.190312046444122,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039981693456983574,
      "loss": 4.8249,
      "mean_token_accuracy": 0.22215677350759505,
      "num_tokens": 72976794.0,
      "step": 35170
    },
    {
      "entropy": 5.597756052017212,
      "epoch": 3.190765602322206,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003997900120109998,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2243738904595375,
      "num_tokens": 72986837.0,
      "step": 35175
    },
    {
      "entropy": 5.600722742080689,
      "epoch": 3.1912191582002905,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003997630868714745,
      "loss": 4.9005,
      "mean_token_accuracy": 0.21610475033521653,
      "num_tokens": 72996819.0,
      "step": 35180
    },
    {
      "entropy": 5.581478214263916,
      "epoch": 3.1916727140783743,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039973615915181687,
      "loss": 4.7794,
      "mean_token_accuracy": 0.22965004295110703,
      "num_tokens": 73005963.0,
      "step": 35185
    },
    {
      "entropy": 5.610789489746094,
      "epoch": 3.1921262699564585,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039970922885258363,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2169812560081482,
      "num_tokens": 73015912.0,
      "step": 35190
    },
    {
      "entropy": 5.562548494338989,
      "epoch": 3.192579825834543,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003996822959743318,
      "loss": 4.8046,
      "mean_token_accuracy": 0.22984550297260284,
      "num_tokens": 73026528.0,
      "step": 35195
    },
    {
      "entropy": 5.629562854766846,
      "epoch": 3.193033381712627,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003996553605176183,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2149041324853897,
      "num_tokens": 73036255.0,
      "step": 35200
    },
    {
      "entropy": 5.655352020263672,
      "epoch": 3.1934869375907113,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003996284224830002,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21483247429132463,
      "num_tokens": 73046686.0,
      "step": 35205
    },
    {
      "entropy": 5.6188966751098635,
      "epoch": 3.193940493468795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039960148187103453,
      "loss": 4.8144,
      "mean_token_accuracy": 0.2261397659778595,
      "num_tokens": 73056070.0,
      "step": 35210
    },
    {
      "entropy": 5.70836615562439,
      "epoch": 3.1943940493468794,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039957453868227845,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22494465857744217,
      "num_tokens": 73066288.0,
      "step": 35215
    },
    {
      "entropy": 5.64421329498291,
      "epoch": 3.1948476052249637,
      "grad_norm": 1.25,
      "learning_rate": 0.00039954759291728915,
      "loss": 4.9006,
      "mean_token_accuracy": 0.21916361898183823,
      "num_tokens": 73076297.0,
      "step": 35220
    },
    {
      "entropy": 5.6523832321167,
      "epoch": 3.195301161103048,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039952064457662395,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21674781888723374,
      "num_tokens": 73086249.0,
      "step": 35225
    },
    {
      "entropy": 5.581540584564209,
      "epoch": 3.1957547169811322,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039949369366084,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22464638501405715,
      "num_tokens": 73096358.0,
      "step": 35230
    },
    {
      "entropy": 5.576098346710205,
      "epoch": 3.196208272859216,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039946674017049484,
      "loss": 4.8204,
      "mean_token_accuracy": 0.22194795459508895,
      "num_tokens": 73106891.0,
      "step": 35235
    },
    {
      "entropy": 5.5393165111541744,
      "epoch": 3.1966618287373003,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039943978410614565,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22040769308805466,
      "num_tokens": 73117106.0,
      "step": 35240
    },
    {
      "entropy": 5.629960393905639,
      "epoch": 3.1971153846153846,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039941282546835017,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22212623953819274,
      "num_tokens": 73127112.0,
      "step": 35245
    },
    {
      "entropy": 5.632509374618531,
      "epoch": 3.197568940493469,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039938586425766564,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22116705030202866,
      "num_tokens": 73137751.0,
      "step": 35250
    },
    {
      "entropy": 5.659935188293457,
      "epoch": 3.198022496371553,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003993589004746498,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22556164264678955,
      "num_tokens": 73149303.0,
      "step": 35255
    },
    {
      "entropy": 5.602596807479858,
      "epoch": 3.198476052249637,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039933193411986024,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2224455252289772,
      "num_tokens": 73160025.0,
      "step": 35260
    },
    {
      "entropy": 5.588270854949951,
      "epoch": 3.198929608127721,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039930496519385454,
      "loss": 4.869,
      "mean_token_accuracy": 0.22529269903898239,
      "num_tokens": 73170721.0,
      "step": 35265
    },
    {
      "entropy": 5.643705558776856,
      "epoch": 3.1993831640058055,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039927799369719047,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2158379837870598,
      "num_tokens": 73180902.0,
      "step": 35270
    },
    {
      "entropy": 5.537416172027588,
      "epoch": 3.1998367198838897,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003992510196304258,
      "loss": 4.7598,
      "mean_token_accuracy": 0.22765354961156845,
      "num_tokens": 73191259.0,
      "step": 35275
    },
    {
      "entropy": 5.614968538284302,
      "epoch": 3.200290275761974,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039922404299411846,
      "loss": 4.8887,
      "mean_token_accuracy": 0.22845399230718613,
      "num_tokens": 73201654.0,
      "step": 35280
    },
    {
      "entropy": 5.624163722991943,
      "epoch": 3.2007438316400583,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003991970637888262,
      "loss": 4.9304,
      "mean_token_accuracy": 0.2075518861413002,
      "num_tokens": 73212658.0,
      "step": 35285
    },
    {
      "entropy": 5.67769227027893,
      "epoch": 3.201197387518142,
      "grad_norm": 1.296875,
      "learning_rate": 0.000399170082015107,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2146735519170761,
      "num_tokens": 73222883.0,
      "step": 35290
    },
    {
      "entropy": 5.599152755737305,
      "epoch": 3.2016509433962264,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000399143097673519,
      "loss": 4.8653,
      "mean_token_accuracy": 0.22059713453054428,
      "num_tokens": 73233731.0,
      "step": 35295
    },
    {
      "entropy": 5.600179052352905,
      "epoch": 3.2021044992743106,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039911611076461995,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23347573727369308,
      "num_tokens": 73243794.0,
      "step": 35300
    },
    {
      "entropy": 5.625681829452515,
      "epoch": 3.202558055152395,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039908912128896817,
      "loss": 4.8797,
      "mean_token_accuracy": 0.21602460891008377,
      "num_tokens": 73254708.0,
      "step": 35305
    },
    {
      "entropy": 5.56280665397644,
      "epoch": 3.203011611030479,
      "grad_norm": 1.25,
      "learning_rate": 0.0003990621292471217,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2140805959701538,
      "num_tokens": 73265311.0,
      "step": 35310
    },
    {
      "entropy": 5.72505669593811,
      "epoch": 3.203465166908563,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039903513463963874,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21287549585103988,
      "num_tokens": 73275886.0,
      "step": 35315
    },
    {
      "entropy": 5.665661716461182,
      "epoch": 3.2039187227866472,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003990081374670777,
      "loss": 4.9035,
      "mean_token_accuracy": 0.2260631799697876,
      "num_tokens": 73285903.0,
      "step": 35320
    },
    {
      "entropy": 5.600179815292359,
      "epoch": 3.2043722786647315,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039898113772999665,
      "loss": 4.881,
      "mean_token_accuracy": 0.21989596933126448,
      "num_tokens": 73296571.0,
      "step": 35325
    },
    {
      "entropy": 5.630857992172241,
      "epoch": 3.2048258345428158,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039895413542895427,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21024704426527024,
      "num_tokens": 73307105.0,
      "step": 35330
    },
    {
      "entropy": 5.6180963039398195,
      "epoch": 3.2052793904209,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003989271305645086,
      "loss": 4.8411,
      "mean_token_accuracy": 0.22188646495342254,
      "num_tokens": 73316664.0,
      "step": 35335
    },
    {
      "entropy": 5.6249833583831785,
      "epoch": 3.205732946298984,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039890012313721835,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22378720790147782,
      "num_tokens": 73327438.0,
      "step": 35340
    },
    {
      "entropy": 5.629421186447144,
      "epoch": 3.206186502177068,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000398873113147642,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22626577019691468,
      "num_tokens": 73337635.0,
      "step": 35345
    },
    {
      "entropy": 5.589259052276612,
      "epoch": 3.2066400580551524,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039884610059633807,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22069051265716552,
      "num_tokens": 73347032.0,
      "step": 35350
    },
    {
      "entropy": 5.700584125518799,
      "epoch": 3.2070936139332367,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003988190854838652,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21592655628919602,
      "num_tokens": 73357780.0,
      "step": 35355
    },
    {
      "entropy": 5.67203049659729,
      "epoch": 3.207547169811321,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039879206781078214,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22255441993474961,
      "num_tokens": 73367831.0,
      "step": 35360
    },
    {
      "entropy": 5.642022132873535,
      "epoch": 3.2080007256894048,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003987650475776476,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2203782916069031,
      "num_tokens": 73378588.0,
      "step": 35365
    },
    {
      "entropy": 5.592491340637207,
      "epoch": 3.208454281567489,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039873802478502027,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22168018966913222,
      "num_tokens": 73388414.0,
      "step": 35370
    },
    {
      "entropy": 5.606138134002686,
      "epoch": 3.2089078374455733,
      "grad_norm": 1.25,
      "learning_rate": 0.0003987109994334591,
      "loss": 4.8733,
      "mean_token_accuracy": 0.22078590840101242,
      "num_tokens": 73398891.0,
      "step": 35375
    },
    {
      "entropy": 5.6265462875366214,
      "epoch": 3.2093613933236576,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003986839715235229,
      "loss": 4.938,
      "mean_token_accuracy": 0.21807720959186555,
      "num_tokens": 73409605.0,
      "step": 35380
    },
    {
      "entropy": 5.722667026519775,
      "epoch": 3.209814949201742,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039865694105577063,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21125759780406952,
      "num_tokens": 73419103.0,
      "step": 35385
    },
    {
      "entropy": 5.590012073516846,
      "epoch": 3.2102685050798256,
      "grad_norm": 1.25,
      "learning_rate": 0.0003986299080307614,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22253487408161163,
      "num_tokens": 73429134.0,
      "step": 35390
    },
    {
      "entropy": 5.630838298797608,
      "epoch": 3.21072206095791,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003986028724490541,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20802294313907624,
      "num_tokens": 73440382.0,
      "step": 35395
    },
    {
      "entropy": 5.631938171386719,
      "epoch": 3.211175616835994,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003985758343112079,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22155732810497283,
      "num_tokens": 73451300.0,
      "step": 35400
    },
    {
      "entropy": 5.6054424285888675,
      "epoch": 3.2116291727140784,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039854879361778193,
      "loss": 4.8064,
      "mean_token_accuracy": 0.2249677747488022,
      "num_tokens": 73461339.0,
      "step": 35405
    },
    {
      "entropy": 5.643403100967407,
      "epoch": 3.2120827285921627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039852175036933545,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21941620260477065,
      "num_tokens": 73470961.0,
      "step": 35410
    },
    {
      "entropy": 5.7341409683227536,
      "epoch": 3.2125362844702465,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039849470456642777,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21070910543203353,
      "num_tokens": 73481492.0,
      "step": 35415
    },
    {
      "entropy": 5.604830980300903,
      "epoch": 3.212989840348331,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039846765620961806,
      "loss": 4.8537,
      "mean_token_accuracy": 0.21573233157396315,
      "num_tokens": 73492351.0,
      "step": 35420
    },
    {
      "entropy": 5.639754962921143,
      "epoch": 3.213443396226415,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003984406052994658,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21697274297475816,
      "num_tokens": 73503265.0,
      "step": 35425
    },
    {
      "entropy": 5.643771028518676,
      "epoch": 3.2138969521044993,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003984135518365303,
      "loss": 4.9252,
      "mean_token_accuracy": 0.2145065635442734,
      "num_tokens": 73513804.0,
      "step": 35430
    },
    {
      "entropy": 5.677305603027344,
      "epoch": 3.2143505079825836,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039838649582137113,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21593131273984909,
      "num_tokens": 73523453.0,
      "step": 35435
    },
    {
      "entropy": 5.720432758331299,
      "epoch": 3.214804063860668,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039835943725454777,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20846251547336578,
      "num_tokens": 73533931.0,
      "step": 35440
    },
    {
      "entropy": 5.694847488403321,
      "epoch": 3.2152576197387517,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039833237613661984,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2214327409863472,
      "num_tokens": 73544562.0,
      "step": 35445
    },
    {
      "entropy": 5.640496635437012,
      "epoch": 3.215711175616836,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003983053124681469,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22305275797843932,
      "num_tokens": 73555149.0,
      "step": 35450
    },
    {
      "entropy": 5.564590978622436,
      "epoch": 3.21616473149492,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039827824624968873,
      "loss": 4.823,
      "mean_token_accuracy": 0.2190001577138901,
      "num_tokens": 73565131.0,
      "step": 35455
    },
    {
      "entropy": 5.592260074615479,
      "epoch": 3.2166182873730045,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039825117748180493,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22566003054380418,
      "num_tokens": 73574185.0,
      "step": 35460
    },
    {
      "entropy": 5.678895950317383,
      "epoch": 3.2170718432510883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039822410616505545,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2118372231721878,
      "num_tokens": 73585733.0,
      "step": 35465
    },
    {
      "entropy": 5.716606712341308,
      "epoch": 3.2175253991291726,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039819703230000006,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2108761712908745,
      "num_tokens": 73597077.0,
      "step": 35470
    },
    {
      "entropy": 5.56656494140625,
      "epoch": 3.217978955007257,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003981699558871986,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2193862110376358,
      "num_tokens": 73607885.0,
      "step": 35475
    },
    {
      "entropy": 5.622937774658203,
      "epoch": 3.218432510885341,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000398142876927211,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21663307696580886,
      "num_tokens": 73618769.0,
      "step": 35480
    },
    {
      "entropy": 5.715946435928345,
      "epoch": 3.2188860667634254,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039811579542059744,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21169230937957764,
      "num_tokens": 73629345.0,
      "step": 35485
    },
    {
      "entropy": 5.671459341049195,
      "epoch": 3.2193396226415096,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039808871136791786,
      "loss": 4.8654,
      "mean_token_accuracy": 0.2142948254942894,
      "num_tokens": 73639698.0,
      "step": 35490
    },
    {
      "entropy": 5.5969727516174315,
      "epoch": 3.2197931785195935,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039806162476973227,
      "loss": 4.9002,
      "mean_token_accuracy": 0.2204391047358513,
      "num_tokens": 73650954.0,
      "step": 35495
    },
    {
      "entropy": 5.670774650573731,
      "epoch": 3.2202467343976777,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000398034535626601,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2123030126094818,
      "num_tokens": 73661195.0,
      "step": 35500
    },
    {
      "entropy": 5.7414649486541744,
      "epoch": 3.220700290275762,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039800744393908414,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21257992535829545,
      "num_tokens": 73670512.0,
      "step": 35505
    },
    {
      "entropy": 5.658827781677246,
      "epoch": 3.2211538461538463,
      "grad_norm": 1.1875,
      "learning_rate": 0.000397980349707742,
      "loss": 4.9458,
      "mean_token_accuracy": 0.21366484761238097,
      "num_tokens": 73681927.0,
      "step": 35510
    },
    {
      "entropy": 5.701186227798462,
      "epoch": 3.2216074020319305,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039795325293313493,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21942930221557616,
      "num_tokens": 73692025.0,
      "step": 35515
    },
    {
      "entropy": 5.650922250747681,
      "epoch": 3.2220609579100143,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039792615361582324,
      "loss": 4.899,
      "mean_token_accuracy": 0.21642192751169204,
      "num_tokens": 73702278.0,
      "step": 35520
    },
    {
      "entropy": 5.7090349197387695,
      "epoch": 3.2225145137880986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039789905175636726,
      "loss": 5.0006,
      "mean_token_accuracy": 0.21112528294324875,
      "num_tokens": 73713511.0,
      "step": 35525
    },
    {
      "entropy": 5.57156548500061,
      "epoch": 3.222968069666183,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039787194735532765,
      "loss": 4.8076,
      "mean_token_accuracy": 0.22535914927721024,
      "num_tokens": 73723683.0,
      "step": 35530
    },
    {
      "entropy": 5.639390373229981,
      "epoch": 3.223421625544267,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003978448404132649,
      "loss": 4.8811,
      "mean_token_accuracy": 0.22362838089466094,
      "num_tokens": 73734191.0,
      "step": 35535
    },
    {
      "entropy": 5.680836391448975,
      "epoch": 3.2238751814223514,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039781773093073956,
      "loss": 4.922,
      "mean_token_accuracy": 0.21638190895318984,
      "num_tokens": 73743747.0,
      "step": 35540
    },
    {
      "entropy": 5.598128414154052,
      "epoch": 3.2243287373004352,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003977906189083122,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22042784988880157,
      "num_tokens": 73754466.0,
      "step": 35545
    },
    {
      "entropy": 5.641989135742188,
      "epoch": 3.2247822931785195,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003977635043465435,
      "loss": 4.8767,
      "mean_token_accuracy": 0.21699256002902984,
      "num_tokens": 73765948.0,
      "step": 35550
    },
    {
      "entropy": 5.5435771465301515,
      "epoch": 3.2252358490566038,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003977363872459943,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2227562829852104,
      "num_tokens": 73776003.0,
      "step": 35555
    },
    {
      "entropy": 5.708752870559692,
      "epoch": 3.225689404934688,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003977092676072254,
      "loss": 5.0667,
      "mean_token_accuracy": 0.21447840631008147,
      "num_tokens": 73787821.0,
      "step": 35560
    },
    {
      "entropy": 5.658638525009155,
      "epoch": 3.2261429608127723,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039768214543079746,
      "loss": 4.9121,
      "mean_token_accuracy": 0.21536696404218675,
      "num_tokens": 73799207.0,
      "step": 35565
    },
    {
      "entropy": 5.641917896270752,
      "epoch": 3.226596516690856,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003976550207172716,
      "loss": 4.8642,
      "mean_token_accuracy": 0.2239322379231453,
      "num_tokens": 73809987.0,
      "step": 35570
    },
    {
      "entropy": 5.56656551361084,
      "epoch": 3.2270500725689404,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039762789346720854,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2103703185915947,
      "num_tokens": 73819075.0,
      "step": 35575
    },
    {
      "entropy": 5.632872867584228,
      "epoch": 3.2275036284470247,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003976007636811694,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22548563927412033,
      "num_tokens": 73829831.0,
      "step": 35580
    },
    {
      "entropy": 5.677143001556397,
      "epoch": 3.227957184325109,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003975736313597153,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22132644355297088,
      "num_tokens": 73839826.0,
      "step": 35585
    },
    {
      "entropy": 5.621141004562378,
      "epoch": 3.228410740203193,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003975464965034072,
      "loss": 4.8521,
      "mean_token_accuracy": 0.22440076023340225,
      "num_tokens": 73850006.0,
      "step": 35590
    },
    {
      "entropy": 5.635663843154907,
      "epoch": 3.228864296081277,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003975193591128064,
      "loss": 4.901,
      "mean_token_accuracy": 0.21671372950077056,
      "num_tokens": 73860309.0,
      "step": 35595
    },
    {
      "entropy": 5.656742811203003,
      "epoch": 3.2293178519593613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039749221918847406,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20660825818777084,
      "num_tokens": 73871613.0,
      "step": 35600
    },
    {
      "entropy": 5.556329965591431,
      "epoch": 3.2297714078374455,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003974650767309713,
      "loss": 4.8159,
      "mean_token_accuracy": 0.22392415553331374,
      "num_tokens": 73882303.0,
      "step": 35605
    },
    {
      "entropy": 5.690486145019531,
      "epoch": 3.23022496371553,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039743793174085965,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2083287999033928,
      "num_tokens": 73893548.0,
      "step": 35610
    },
    {
      "entropy": 5.670873308181763,
      "epoch": 3.230678519593614,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039741078421870024,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21701249331235886,
      "num_tokens": 73903934.0,
      "step": 35615
    },
    {
      "entropy": 5.681013298034668,
      "epoch": 3.231132075471698,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039738363416505474,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21038107126951217,
      "num_tokens": 73913813.0,
      "step": 35620
    },
    {
      "entropy": 5.662494802474976,
      "epoch": 3.231585631349782,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039735648158048444,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2096915066242218,
      "num_tokens": 73924339.0,
      "step": 35625
    },
    {
      "entropy": 5.634679746627808,
      "epoch": 3.2320391872278664,
      "grad_norm": 1.25,
      "learning_rate": 0.0003973293264655509,
      "loss": 4.9101,
      "mean_token_accuracy": 0.2128859207034111,
      "num_tokens": 73935318.0,
      "step": 35630
    },
    {
      "entropy": 5.623164558410645,
      "epoch": 3.2324927431059507,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003973021688208158,
      "loss": 4.867,
      "mean_token_accuracy": 0.22606242448091507,
      "num_tokens": 73945547.0,
      "step": 35635
    },
    {
      "entropy": 5.6686169624328615,
      "epoch": 3.232946298984035,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003972750086468406,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2174498438835144,
      "num_tokens": 73956195.0,
      "step": 35640
    },
    {
      "entropy": 5.649539661407471,
      "epoch": 3.2333998548621192,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039724784594418714,
      "loss": 4.8446,
      "mean_token_accuracy": 0.22769755423069,
      "num_tokens": 73966843.0,
      "step": 35645
    },
    {
      "entropy": 5.61407151222229,
      "epoch": 3.233853410740203,
      "grad_norm": 1.140625,
      "learning_rate": 0.000397220680713417,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21844768673181533,
      "num_tokens": 73978516.0,
      "step": 35650
    },
    {
      "entropy": 5.5622883319854735,
      "epoch": 3.2343069666182873,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039719351295509203,
      "loss": 4.8353,
      "mean_token_accuracy": 0.22558247148990632,
      "num_tokens": 73988376.0,
      "step": 35655
    },
    {
      "entropy": 5.599919557571411,
      "epoch": 3.2347605224963716,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039716634266977413,
      "loss": 4.8701,
      "mean_token_accuracy": 0.2243556335568428,
      "num_tokens": 73998232.0,
      "step": 35660
    },
    {
      "entropy": 5.609225940704346,
      "epoch": 3.235214078374456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003971391698580252,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2171994373202324,
      "num_tokens": 74008726.0,
      "step": 35665
    },
    {
      "entropy": 5.644869995117188,
      "epoch": 3.23566763425254,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039711199452040696,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22001542896032333,
      "num_tokens": 74018746.0,
      "step": 35670
    },
    {
      "entropy": 5.640611839294434,
      "epoch": 3.236121190130624,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039708481665748157,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22133959829807281,
      "num_tokens": 74028615.0,
      "step": 35675
    },
    {
      "entropy": 5.649672031402588,
      "epoch": 3.236574746008708,
      "grad_norm": 1.125,
      "learning_rate": 0.0003970576362698111,
      "loss": 4.9401,
      "mean_token_accuracy": 0.2181209221482277,
      "num_tokens": 74039219.0,
      "step": 35680
    },
    {
      "entropy": 5.633511400222778,
      "epoch": 3.2370283018867925,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039703045335795766,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22398090362548828,
      "num_tokens": 74049385.0,
      "step": 35685
    },
    {
      "entropy": 5.711008930206299,
      "epoch": 3.2374818577648767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003970032679224833,
      "loss": 5.0416,
      "mean_token_accuracy": 0.20649878978729247,
      "num_tokens": 74059437.0,
      "step": 35690
    },
    {
      "entropy": 5.702699613571167,
      "epoch": 3.237935413642961,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039697607996395023,
      "loss": 4.9695,
      "mean_token_accuracy": 0.209164160490036,
      "num_tokens": 74068834.0,
      "step": 35695
    },
    {
      "entropy": 5.689772510528565,
      "epoch": 3.238388969521045,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003969488894829208,
      "loss": 4.9421,
      "mean_token_accuracy": 0.213767808675766,
      "num_tokens": 74079757.0,
      "step": 35700
    },
    {
      "entropy": 5.675665712356567,
      "epoch": 3.238842525399129,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003969216964799573,
      "loss": 4.8765,
      "mean_token_accuracy": 0.2203963503241539,
      "num_tokens": 74091092.0,
      "step": 35705
    },
    {
      "entropy": 5.587155389785766,
      "epoch": 3.2392960812772134,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003968945009556219,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22699668556451796,
      "num_tokens": 74102272.0,
      "step": 35710
    },
    {
      "entropy": 5.5951841354370115,
      "epoch": 3.2397496371552976,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003968673029104773,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22325626015663147,
      "num_tokens": 74113101.0,
      "step": 35715
    },
    {
      "entropy": 5.689749670028687,
      "epoch": 3.240203193033382,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003968401023450858,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22083511352539062,
      "num_tokens": 74123180.0,
      "step": 35720
    },
    {
      "entropy": 5.618878889083862,
      "epoch": 3.2406567489114657,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039681289926000985,
      "loss": 4.8847,
      "mean_token_accuracy": 0.2248257130384445,
      "num_tokens": 74133675.0,
      "step": 35725
    },
    {
      "entropy": 5.590866994857788,
      "epoch": 3.24111030478955,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003967856936558122,
      "loss": 4.8172,
      "mean_token_accuracy": 0.2213743433356285,
      "num_tokens": 74143472.0,
      "step": 35730
    },
    {
      "entropy": 5.635987854003906,
      "epoch": 3.2415638606676342,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039675848553305526,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21613086760044098,
      "num_tokens": 74154682.0,
      "step": 35735
    },
    {
      "entropy": 5.642532873153686,
      "epoch": 3.2420174165457185,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003967312748923019,
      "loss": 4.9155,
      "mean_token_accuracy": 0.2224576100707054,
      "num_tokens": 74164745.0,
      "step": 35740
    },
    {
      "entropy": 5.614141464233398,
      "epoch": 3.2424709724238028,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039670406173411473,
      "loss": 4.8122,
      "mean_token_accuracy": 0.22684427052736283,
      "num_tokens": 74175388.0,
      "step": 35745
    },
    {
      "entropy": 5.624486970901489,
      "epoch": 3.2429245283018866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003966768460590566,
      "loss": 4.8641,
      "mean_token_accuracy": 0.21987419575452805,
      "num_tokens": 74185929.0,
      "step": 35750
    },
    {
      "entropy": 5.641559648513794,
      "epoch": 3.243378084179971,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039664962786769013,
      "loss": 4.8714,
      "mean_token_accuracy": 0.22296553552150727,
      "num_tokens": 74196165.0,
      "step": 35755
    },
    {
      "entropy": 5.632019186019898,
      "epoch": 3.243831640058055,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039662240716057854,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22270023077726364,
      "num_tokens": 74206843.0,
      "step": 35760
    },
    {
      "entropy": 5.656398010253906,
      "epoch": 3.2442851959361394,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003965951839382845,
      "loss": 4.8596,
      "mean_token_accuracy": 0.2233958825469017,
      "num_tokens": 74216854.0,
      "step": 35765
    },
    {
      "entropy": 5.6067023277282715,
      "epoch": 3.2447387518142237,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000396567958201371,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22269334644079208,
      "num_tokens": 74226158.0,
      "step": 35770
    },
    {
      "entropy": 5.601041603088379,
      "epoch": 3.2451923076923075,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003965407299504013,
      "loss": 4.9146,
      "mean_token_accuracy": 0.222706201672554,
      "num_tokens": 74236860.0,
      "step": 35775
    },
    {
      "entropy": 5.638400554656982,
      "epoch": 3.2456458635703918,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003965134991859382,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22387139350175858,
      "num_tokens": 74247706.0,
      "step": 35780
    },
    {
      "entropy": 5.687844610214233,
      "epoch": 3.246099419448476,
      "grad_norm": 1.203125,
      "learning_rate": 0.000396486265908545,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2200131133198738,
      "num_tokens": 74259479.0,
      "step": 35785
    },
    {
      "entropy": 5.616486120223999,
      "epoch": 3.2465529753265603,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003964590301187849,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21905033588409423,
      "num_tokens": 74270371.0,
      "step": 35790
    },
    {
      "entropy": 5.623467445373535,
      "epoch": 3.2470065312046446,
      "grad_norm": 1.375,
      "learning_rate": 0.0003964317918172211,
      "loss": 4.8514,
      "mean_token_accuracy": 0.23185165226459503,
      "num_tokens": 74279637.0,
      "step": 35795
    },
    {
      "entropy": 5.66420350074768,
      "epoch": 3.247460087082729,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039640455100441684,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21282645016908647,
      "num_tokens": 74289889.0,
      "step": 35800
    },
    {
      "entropy": 5.673099660873413,
      "epoch": 3.2479136429608126,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039637730768093564,
      "loss": 4.923,
      "mean_token_accuracy": 0.2277629092335701,
      "num_tokens": 74300949.0,
      "step": 35805
    },
    {
      "entropy": 5.648768043518066,
      "epoch": 3.248367198838897,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003963500618473407,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22065519392490388,
      "num_tokens": 74312077.0,
      "step": 35810
    },
    {
      "entropy": 5.602660465240478,
      "epoch": 3.248820754716981,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003963228135041956,
      "loss": 4.9658,
      "mean_token_accuracy": 0.21205090433359147,
      "num_tokens": 74322610.0,
      "step": 35815
    },
    {
      "entropy": 5.5728880882263185,
      "epoch": 3.2492743105950654,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039629556265206373,
      "loss": 4.8184,
      "mean_token_accuracy": 0.21757723838090898,
      "num_tokens": 74331970.0,
      "step": 35820
    },
    {
      "entropy": 5.619731903076172,
      "epoch": 3.2497278664731493,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003962683092915088,
      "loss": 4.8552,
      "mean_token_accuracy": 0.2270527794957161,
      "num_tokens": 74342812.0,
      "step": 35825
    },
    {
      "entropy": 5.628123617172241,
      "epoch": 3.2501814223512335,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003962410534230943,
      "loss": 4.8321,
      "mean_token_accuracy": 0.2265948936343193,
      "num_tokens": 74353786.0,
      "step": 35830
    },
    {
      "entropy": 5.66396198272705,
      "epoch": 3.250634978229318,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039621379504738385,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21867869794368744,
      "num_tokens": 74363732.0,
      "step": 35835
    },
    {
      "entropy": 5.643849182128906,
      "epoch": 3.251088534107402,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039618653416494137,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2202475681900978,
      "num_tokens": 74374077.0,
      "step": 35840
    },
    {
      "entropy": 5.642728424072265,
      "epoch": 3.2515420899854863,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039615927077633043,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21897670477628708,
      "num_tokens": 74384758.0,
      "step": 35845
    },
    {
      "entropy": 5.708169603347779,
      "epoch": 3.2519956458635706,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039613200488211483,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22317821085453032,
      "num_tokens": 74395811.0,
      "step": 35850
    },
    {
      "entropy": 5.646434593200683,
      "epoch": 3.2524492017416544,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039610473648285857,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21638157069683076,
      "num_tokens": 74407830.0,
      "step": 35855
    },
    {
      "entropy": 5.589155197143555,
      "epoch": 3.2529027576197387,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003960774655791255,
      "loss": 4.8662,
      "mean_token_accuracy": 0.21954732090234758,
      "num_tokens": 74417565.0,
      "step": 35860
    },
    {
      "entropy": 5.678437662124634,
      "epoch": 3.253356313497823,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039605019217147947,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22099993377923965,
      "num_tokens": 74428015.0,
      "step": 35865
    },
    {
      "entropy": 5.597277975082397,
      "epoch": 3.253809869375907,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039602291626048475,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22099123448133468,
      "num_tokens": 74439523.0,
      "step": 35870
    },
    {
      "entropy": 5.691626119613647,
      "epoch": 3.2542634252539915,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039599563784670526,
      "loss": 4.9822,
      "mean_token_accuracy": 0.20558119863271712,
      "num_tokens": 74449897.0,
      "step": 35875
    },
    {
      "entropy": 5.621711540222168,
      "epoch": 3.2547169811320753,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003959683569307051,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22352193892002106,
      "num_tokens": 74459888.0,
      "step": 35880
    },
    {
      "entropy": 5.609740924835205,
      "epoch": 3.2551705370101596,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003959410735130485,
      "loss": 4.8138,
      "mean_token_accuracy": 0.21500419825315475,
      "num_tokens": 74470894.0,
      "step": 35885
    },
    {
      "entropy": 5.6058878898620605,
      "epoch": 3.255624092888244,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039591378759429974,
      "loss": 4.8099,
      "mean_token_accuracy": 0.23100608438253403,
      "num_tokens": 74481306.0,
      "step": 35890
    },
    {
      "entropy": 5.651317977905274,
      "epoch": 3.256077648766328,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000395886499175023,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2219666913151741,
      "num_tokens": 74491048.0,
      "step": 35895
    },
    {
      "entropy": 5.599463844299317,
      "epoch": 3.2565312046444124,
      "grad_norm": 1.125,
      "learning_rate": 0.00039585920825578266,
      "loss": 4.844,
      "mean_token_accuracy": 0.226677505671978,
      "num_tokens": 74501656.0,
      "step": 35900
    },
    {
      "entropy": 5.683964776992798,
      "epoch": 3.256984760522496,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003958319148371431,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21967142969369888,
      "num_tokens": 74513280.0,
      "step": 35905
    },
    {
      "entropy": 5.615320777893066,
      "epoch": 3.2574383164005805,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003958046189196687,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2172779306769371,
      "num_tokens": 74522898.0,
      "step": 35910
    },
    {
      "entropy": 5.697880649566651,
      "epoch": 3.2578918722786647,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039577732050392395,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21768889278173448,
      "num_tokens": 74532551.0,
      "step": 35915
    },
    {
      "entropy": 5.682796907424927,
      "epoch": 3.258345428156749,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003957500195904735,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22344327419996263,
      "num_tokens": 74542285.0,
      "step": 35920
    },
    {
      "entropy": 5.689100170135498,
      "epoch": 3.2587989840348333,
      "grad_norm": 1.328125,
      "learning_rate": 0.000395722716179882,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21521099656820297,
      "num_tokens": 74551982.0,
      "step": 35925
    },
    {
      "entropy": 5.661095523834229,
      "epoch": 3.259252539912917,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003956954102727138,
      "loss": 4.9128,
      "mean_token_accuracy": 0.21996078938245772,
      "num_tokens": 74562190.0,
      "step": 35930
    },
    {
      "entropy": 5.601325368881225,
      "epoch": 3.2597060957910013,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039566810186953376,
      "loss": 4.8368,
      "mean_token_accuracy": 0.2263374775648117,
      "num_tokens": 74572673.0,
      "step": 35935
    },
    {
      "entropy": 5.5878705978393555,
      "epoch": 3.2601596516690856,
      "grad_norm": 1.25,
      "learning_rate": 0.0003956407909709068,
      "loss": 4.8671,
      "mean_token_accuracy": 0.22473883330821992,
      "num_tokens": 74582686.0,
      "step": 35940
    },
    {
      "entropy": 5.693186950683594,
      "epoch": 3.26061320754717,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039561347757739734,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21870440244674683,
      "num_tokens": 74593658.0,
      "step": 35945
    },
    {
      "entropy": 5.697020244598389,
      "epoch": 3.261066763425254,
      "grad_norm": 1.390625,
      "learning_rate": 0.00039558616168957044,
      "loss": 4.9919,
      "mean_token_accuracy": 0.22000594884157182,
      "num_tokens": 74603585.0,
      "step": 35950
    },
    {
      "entropy": 5.626160335540772,
      "epoch": 3.2615203193033384,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003955588433079911,
      "loss": 4.8675,
      "mean_token_accuracy": 0.21849820017814636,
      "num_tokens": 74612806.0,
      "step": 35955
    },
    {
      "entropy": 5.648509740829468,
      "epoch": 3.2619738751814222,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039553152243322414,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2242448776960373,
      "num_tokens": 74623252.0,
      "step": 35960
    },
    {
      "entropy": 5.609339094161987,
      "epoch": 3.2624274310595065,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003955041990658346,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22297365814447404,
      "num_tokens": 74633555.0,
      "step": 35965
    },
    {
      "entropy": 5.628423833847046,
      "epoch": 3.2628809869375908,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003954768732063874,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22183336913585663,
      "num_tokens": 74643322.0,
      "step": 35970
    },
    {
      "entropy": 5.622592973709106,
      "epoch": 3.263334542815675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003954495448554478,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22827915698289872,
      "num_tokens": 74653461.0,
      "step": 35975
    },
    {
      "entropy": 5.6619940280914305,
      "epoch": 3.263788098693759,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039542221401358104,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2237296834588051,
      "num_tokens": 74663470.0,
      "step": 35980
    },
    {
      "entropy": 5.5964216709136965,
      "epoch": 3.264241654571843,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000395394880681352,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2268133893609047,
      "num_tokens": 74674074.0,
      "step": 35985
    },
    {
      "entropy": 5.696441793441773,
      "epoch": 3.2646952104499274,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039536754485932626,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21063876152038574,
      "num_tokens": 74684968.0,
      "step": 35990
    },
    {
      "entropy": 5.627312135696411,
      "epoch": 3.2651487663280117,
      "grad_norm": 1.25,
      "learning_rate": 0.00039534020654806906,
      "loss": 4.8134,
      "mean_token_accuracy": 0.23491248041391372,
      "num_tokens": 74694776.0,
      "step": 35995
    },
    {
      "entropy": 5.604242038726807,
      "epoch": 3.265602322206096,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039531286574814563,
      "loss": 4.8431,
      "mean_token_accuracy": 0.2202404573559761,
      "num_tokens": 74705506.0,
      "step": 36000
    },
    {
      "epoch": 3.265602322206096,
      "eval_entropy": 5.334039939043209,
      "eval_loss": 4.975577354431152,
      "eval_mean_token_accuracy": 0.22394091301003463,
      "eval_num_tokens": 74705506.0,
      "eval_runtime": 20.4906,
      "eval_samples_per_second": 1725.669,
      "eval_steps_per_second": 215.709,
      "step": 36000
    },
    {
      "entropy": 5.565581750869751,
      "epoch": 3.26605587808418,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003952855224601215,
      "loss": 4.8777,
      "mean_token_accuracy": 0.21996652632951735,
      "num_tokens": 74715677.0,
      "step": 36005
    },
    {
      "entropy": 5.679157495498657,
      "epoch": 3.266509433962264,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003952581766845622,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22630117535591127,
      "num_tokens": 74725296.0,
      "step": 36010
    },
    {
      "entropy": 5.629681348800659,
      "epoch": 3.2669629898403483,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003952308284220331,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22535340487957,
      "num_tokens": 74736124.0,
      "step": 36015
    },
    {
      "entropy": 5.534782409667969,
      "epoch": 3.2674165457184325,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039520347767309975,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21663733720779418,
      "num_tokens": 74746447.0,
      "step": 36020
    },
    {
      "entropy": 5.590958166122436,
      "epoch": 3.267870101596517,
      "grad_norm": 1.203125,
      "learning_rate": 0.000395176124438328,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2173424854874611,
      "num_tokens": 74756947.0,
      "step": 36025
    },
    {
      "entropy": 5.637294006347656,
      "epoch": 3.2683236574746006,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00039514876871828326,
      "loss": 4.8274,
      "mean_token_accuracy": 0.22575157433748244,
      "num_tokens": 74767856.0,
      "step": 36030
    },
    {
      "entropy": 5.636200904846191,
      "epoch": 3.268777213352685,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039512141051353144,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22309077829122542,
      "num_tokens": 74778335.0,
      "step": 36035
    },
    {
      "entropy": 5.678413057327271,
      "epoch": 3.269230769230769,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003950940498246382,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2241508588194847,
      "num_tokens": 74788430.0,
      "step": 36040
    },
    {
      "entropy": 5.668745613098144,
      "epoch": 3.2696843251088534,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039506668665216944,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21801635771989822,
      "num_tokens": 74800193.0,
      "step": 36045
    },
    {
      "entropy": 5.665507125854492,
      "epoch": 3.2701378809869377,
      "grad_norm": 1.234375,
      "learning_rate": 0.000395039320996691,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2133774757385254,
      "num_tokens": 74810368.0,
      "step": 36050
    },
    {
      "entropy": 5.698576593399048,
      "epoch": 3.270591436865022,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039501195285876874,
      "loss": 4.8717,
      "mean_token_accuracy": 0.2263675943017006,
      "num_tokens": 74820687.0,
      "step": 36055
    },
    {
      "entropy": 5.624888944625854,
      "epoch": 3.271044992743106,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003949845822389688,
      "loss": 4.8449,
      "mean_token_accuracy": 0.2258884862065315,
      "num_tokens": 74830489.0,
      "step": 36060
    },
    {
      "entropy": 5.526579141616821,
      "epoch": 3.27149854862119,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000394957209137857,
      "loss": 4.824,
      "mean_token_accuracy": 0.22220639288425445,
      "num_tokens": 74840638.0,
      "step": 36065
    },
    {
      "entropy": 5.669620084762573,
      "epoch": 3.2719521044992743,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003949298335559997,
      "loss": 5.0269,
      "mean_token_accuracy": 0.2081297218799591,
      "num_tokens": 74850600.0,
      "step": 36070
    },
    {
      "entropy": 5.664458608627319,
      "epoch": 3.2724056603773586,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003949024554939628,
      "loss": 4.901,
      "mean_token_accuracy": 0.22155162692070007,
      "num_tokens": 74861177.0,
      "step": 36075
    },
    {
      "entropy": 5.6895064353942875,
      "epoch": 3.272859216255443,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039487507495231263,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21687162518501282,
      "num_tokens": 74871747.0,
      "step": 36080
    },
    {
      "entropy": 5.610798788070679,
      "epoch": 3.2733127721335267,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003948476919316153,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21683374792337418,
      "num_tokens": 74882796.0,
      "step": 36085
    },
    {
      "entropy": 5.683705615997314,
      "epoch": 3.273766328011611,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003948203064324371,
      "loss": 4.8994,
      "mean_token_accuracy": 0.2152738094329834,
      "num_tokens": 74894123.0,
      "step": 36090
    },
    {
      "entropy": 5.676540040969849,
      "epoch": 3.274219883889695,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039479291845534453,
      "loss": 4.9413,
      "mean_token_accuracy": 0.2130412369966507,
      "num_tokens": 74904787.0,
      "step": 36095
    },
    {
      "entropy": 5.667060089111328,
      "epoch": 3.2746734397677795,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039476552800090384,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20901102274656297,
      "num_tokens": 74915041.0,
      "step": 36100
    },
    {
      "entropy": 5.675154781341552,
      "epoch": 3.2751269956458637,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039473813506968146,
      "loss": 4.87,
      "mean_token_accuracy": 0.21157475262880326,
      "num_tokens": 74926217.0,
      "step": 36105
    },
    {
      "entropy": 5.682069778442383,
      "epoch": 3.2755805515239476,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039471073966224404,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21418286114931107,
      "num_tokens": 74936616.0,
      "step": 36110
    },
    {
      "entropy": 5.6700982570648195,
      "epoch": 3.276034107402032,
      "grad_norm": 1.21875,
      "learning_rate": 0.000394683341779158,
      "loss": 4.9307,
      "mean_token_accuracy": 0.21193190813064575,
      "num_tokens": 74947443.0,
      "step": 36115
    },
    {
      "entropy": 5.658718061447144,
      "epoch": 3.276487663280116,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003946559414209899,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22059800773859023,
      "num_tokens": 74957618.0,
      "step": 36120
    },
    {
      "entropy": 5.628855991363525,
      "epoch": 3.2769412191582004,
      "grad_norm": 1.25,
      "learning_rate": 0.0003946285385883065,
      "loss": 4.836,
      "mean_token_accuracy": 0.2197231635451317,
      "num_tokens": 74967982.0,
      "step": 36125
    },
    {
      "entropy": 5.576105833053589,
      "epoch": 3.2773947750362846,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039460113328167434,
      "loss": 4.883,
      "mean_token_accuracy": 0.22566833794116975,
      "num_tokens": 74978801.0,
      "step": 36130
    },
    {
      "entropy": 5.58773193359375,
      "epoch": 3.2778483309143684,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003945737255016603,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22251618504524232,
      "num_tokens": 74989000.0,
      "step": 36135
    },
    {
      "entropy": 5.658890438079834,
      "epoch": 3.2783018867924527,
      "grad_norm": 1.125,
      "learning_rate": 0.00039454631524883123,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21813136637210845,
      "num_tokens": 74999664.0,
      "step": 36140
    },
    {
      "entropy": 5.649524211883545,
      "epoch": 3.278755442670537,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003945189025237538,
      "loss": 4.895,
      "mean_token_accuracy": 0.2238043650984764,
      "num_tokens": 75010187.0,
      "step": 36145
    },
    {
      "entropy": 5.734669208526611,
      "epoch": 3.2792089985486212,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000394491487326995,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2112666040658951,
      "num_tokens": 75021222.0,
      "step": 36150
    },
    {
      "entropy": 5.605030059814453,
      "epoch": 3.2796625544267055,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003944640696591219,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2221206784248352,
      "num_tokens": 75031136.0,
      "step": 36155
    },
    {
      "entropy": 5.577616596221924,
      "epoch": 3.28011611030479,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039443664952070135,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22218406498432158,
      "num_tokens": 75041545.0,
      "step": 36160
    },
    {
      "entropy": 5.646094274520874,
      "epoch": 3.2805696661828736,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039440922691230044,
      "loss": 4.8937,
      "mean_token_accuracy": 0.224629907310009,
      "num_tokens": 75051805.0,
      "step": 36165
    },
    {
      "entropy": 5.644822311401367,
      "epoch": 3.281023222060958,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039438180183448625,
      "loss": 4.8733,
      "mean_token_accuracy": 0.22180057913064957,
      "num_tokens": 75062551.0,
      "step": 36170
    },
    {
      "entropy": 5.616482400894165,
      "epoch": 3.281476777939042,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000394354374287826,
      "loss": 4.8904,
      "mean_token_accuracy": 0.21719923317432405,
      "num_tokens": 75073040.0,
      "step": 36175
    },
    {
      "entropy": 5.65125732421875,
      "epoch": 3.2819303338171264,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039432694427288695,
      "loss": 4.8806,
      "mean_token_accuracy": 0.2244493454694748,
      "num_tokens": 75082543.0,
      "step": 36180
    },
    {
      "entropy": 5.661353731155396,
      "epoch": 3.28238388969521,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039429951179023614,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21495041698217393,
      "num_tokens": 75092480.0,
      "step": 36185
    },
    {
      "entropy": 5.597170162200928,
      "epoch": 3.2828374455732945,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039427207684044116,
      "loss": 4.8071,
      "mean_token_accuracy": 0.22196868509054185,
      "num_tokens": 75102445.0,
      "step": 36190
    },
    {
      "entropy": 5.661404991149903,
      "epoch": 3.2832910014513788,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039424463942406917,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21748390942811965,
      "num_tokens": 75112793.0,
      "step": 36195
    },
    {
      "entropy": 5.7532491207122805,
      "epoch": 3.283744557329463,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003942171995416877,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2161255359649658,
      "num_tokens": 75123312.0,
      "step": 36200
    },
    {
      "entropy": 5.621792316436768,
      "epoch": 3.2841981132075473,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039418975719386406,
      "loss": 4.8509,
      "mean_token_accuracy": 0.21901890635490417,
      "num_tokens": 75133584.0,
      "step": 36205
    },
    {
      "entropy": 5.596186828613281,
      "epoch": 3.2846516690856316,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003941623123811659,
      "loss": 4.8903,
      "mean_token_accuracy": 0.2165917530655861,
      "num_tokens": 75144344.0,
      "step": 36210
    },
    {
      "entropy": 5.690426015853882,
      "epoch": 3.2851052249637154,
      "grad_norm": 1.25,
      "learning_rate": 0.00039413486510416087,
      "loss": 4.9362,
      "mean_token_accuracy": 0.21251555979251863,
      "num_tokens": 75155086.0,
      "step": 36215
    },
    {
      "entropy": 5.651762819290161,
      "epoch": 3.2855587808417996,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039410741536341634,
      "loss": 4.864,
      "mean_token_accuracy": 0.2203646719455719,
      "num_tokens": 75166046.0,
      "step": 36220
    },
    {
      "entropy": 5.593733596801758,
      "epoch": 3.286012336719884,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039407996315950016,
      "loss": 4.8205,
      "mean_token_accuracy": 0.22276679277420045,
      "num_tokens": 75175388.0,
      "step": 36225
    },
    {
      "entropy": 5.631497001647949,
      "epoch": 3.286465892597968,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039405250849297997,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21909729689359664,
      "num_tokens": 75186213.0,
      "step": 36230
    },
    {
      "entropy": 5.592648220062256,
      "epoch": 3.2869194484760524,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003940250513644236,
      "loss": 4.8538,
      "mean_token_accuracy": 0.22140010297298432,
      "num_tokens": 75196141.0,
      "step": 36235
    },
    {
      "entropy": 5.572011470794678,
      "epoch": 3.2873730043541363,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003939975917743988,
      "loss": 4.7585,
      "mean_token_accuracy": 0.23153959661722184,
      "num_tokens": 75205414.0,
      "step": 36240
    },
    {
      "entropy": 5.66672649383545,
      "epoch": 3.2878265602322205,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003939701297234735,
      "loss": 4.9682,
      "mean_token_accuracy": 0.21441813707351684,
      "num_tokens": 75215860.0,
      "step": 36245
    },
    {
      "entropy": 5.642694282531738,
      "epoch": 3.288280116110305,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039394266521221565,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2122296780347824,
      "num_tokens": 75226358.0,
      "step": 36250
    },
    {
      "entropy": 5.5598348617553714,
      "epoch": 3.288733671988389,
      "grad_norm": 1.46875,
      "learning_rate": 0.00039391519824119315,
      "loss": 4.7628,
      "mean_token_accuracy": 0.23325237333774568,
      "num_tokens": 75235779.0,
      "step": 36255
    },
    {
      "entropy": 5.686887407302857,
      "epoch": 3.2891872278664733,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039388772881097396,
      "loss": 5.0229,
      "mean_token_accuracy": 0.21225060522556305,
      "num_tokens": 75246417.0,
      "step": 36260
    },
    {
      "entropy": 5.665030384063721,
      "epoch": 3.289640783744557,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039386025692212637,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22235198467969894,
      "num_tokens": 75257094.0,
      "step": 36265
    },
    {
      "entropy": 5.617471361160279,
      "epoch": 3.2900943396226414,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003938327825752183,
      "loss": 4.8512,
      "mean_token_accuracy": 0.21995169371366502,
      "num_tokens": 75267374.0,
      "step": 36270
    },
    {
      "entropy": 5.5946238994598385,
      "epoch": 3.2905478955007257,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000393805305770818,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2270051583647728,
      "num_tokens": 75276936.0,
      "step": 36275
    },
    {
      "entropy": 5.5787183284759525,
      "epoch": 3.29100145137881,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003937778265094937,
      "loss": 4.8434,
      "mean_token_accuracy": 0.22402942776679993,
      "num_tokens": 75286403.0,
      "step": 36280
    },
    {
      "entropy": 5.665550279617309,
      "epoch": 3.291455007256894,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003937503447918137,
      "loss": 4.924,
      "mean_token_accuracy": 0.21691127717494965,
      "num_tokens": 75297421.0,
      "step": 36285
    },
    {
      "entropy": 5.678347826004028,
      "epoch": 3.291908563134978,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039372286061834633,
      "loss": 4.92,
      "mean_token_accuracy": 0.21643347889184952,
      "num_tokens": 75307955.0,
      "step": 36290
    },
    {
      "entropy": 5.6450341701507565,
      "epoch": 3.2923621190130623,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039369537398966,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21890365928411484,
      "num_tokens": 75318739.0,
      "step": 36295
    },
    {
      "entropy": 5.593415307998657,
      "epoch": 3.2928156748911466,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039366788490632286,
      "loss": 4.7794,
      "mean_token_accuracy": 0.2307997852563858,
      "num_tokens": 75328629.0,
      "step": 36300
    },
    {
      "entropy": 5.593215560913086,
      "epoch": 3.293269230769231,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039364039336890386,
      "loss": 4.8494,
      "mean_token_accuracy": 0.23059324324131011,
      "num_tokens": 75337986.0,
      "step": 36305
    },
    {
      "entropy": 5.574730110168457,
      "epoch": 3.293722786647315,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003936128993779712,
      "loss": 4.831,
      "mean_token_accuracy": 0.22380361407995225,
      "num_tokens": 75346169.0,
      "step": 36310
    },
    {
      "entropy": 5.653278160095215,
      "epoch": 3.2941763425253994,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003935854029340936,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22533187568187713,
      "num_tokens": 75356111.0,
      "step": 36315
    },
    {
      "entropy": 5.628912544250488,
      "epoch": 3.294629898403483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039355790403783954,
      "loss": 4.8498,
      "mean_token_accuracy": 0.23083717226982117,
      "num_tokens": 75366679.0,
      "step": 36320
    },
    {
      "entropy": 5.55464539527893,
      "epoch": 3.2950834542815675,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039353040268977785,
      "loss": 4.7684,
      "mean_token_accuracy": 0.22916725873947144,
      "num_tokens": 75377500.0,
      "step": 36325
    },
    {
      "entropy": 5.587055969238281,
      "epoch": 3.2955370101596517,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003935028988904773,
      "loss": 4.8407,
      "mean_token_accuracy": 0.2247241660952568,
      "num_tokens": 75387115.0,
      "step": 36330
    },
    {
      "entropy": 5.611231470108033,
      "epoch": 3.295990566037736,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00039347539264050654,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2118021786212921,
      "num_tokens": 75396330.0,
      "step": 36335
    },
    {
      "entropy": 5.604051113128662,
      "epoch": 3.29644412191582,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003934478839404345,
      "loss": 4.887,
      "mean_token_accuracy": 0.2271381363272667,
      "num_tokens": 75407168.0,
      "step": 36340
    },
    {
      "entropy": 5.657563877105713,
      "epoch": 3.296897677793904,
      "grad_norm": 1.25,
      "learning_rate": 0.00039342037279083006,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2176897868514061,
      "num_tokens": 75418117.0,
      "step": 36345
    },
    {
      "entropy": 5.691016626358032,
      "epoch": 3.2973512336719883,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003933928591922621,
      "loss": 5.0022,
      "mean_token_accuracy": 0.21178116798400878,
      "num_tokens": 75429026.0,
      "step": 36350
    },
    {
      "entropy": 5.675773763656617,
      "epoch": 3.2978047895500726,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003933653431452996,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22299687564373016,
      "num_tokens": 75439710.0,
      "step": 36355
    },
    {
      "entropy": 5.6629781246185305,
      "epoch": 3.298258345428157,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039333782465051173,
      "loss": 4.872,
      "mean_token_accuracy": 0.2245420053601265,
      "num_tokens": 75450086.0,
      "step": 36360
    },
    {
      "entropy": 5.61488094329834,
      "epoch": 3.298711901306241,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003933103037084674,
      "loss": 4.8808,
      "mean_token_accuracy": 0.21944738030433655,
      "num_tokens": 75460988.0,
      "step": 36365
    },
    {
      "entropy": 5.650631427764893,
      "epoch": 3.299165457184325,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039328278031973586,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2130212515592575,
      "num_tokens": 75471560.0,
      "step": 36370
    },
    {
      "entropy": 5.631032276153564,
      "epoch": 3.2996190130624092,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003932552544848863,
      "loss": 4.9299,
      "mean_token_accuracy": 0.214851413667202,
      "num_tokens": 75482124.0,
      "step": 36375
    },
    {
      "entropy": 5.634671640396118,
      "epoch": 3.3000725689404935,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000393227726204488,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21661461889743805,
      "num_tokens": 75492351.0,
      "step": 36380
    },
    {
      "entropy": 5.631858205795288,
      "epoch": 3.3005261248185778,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039320019547911013,
      "loss": 4.8086,
      "mean_token_accuracy": 0.2223750203847885,
      "num_tokens": 75502475.0,
      "step": 36385
    },
    {
      "entropy": 5.586136293411255,
      "epoch": 3.3009796806966616,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003931726623093221,
      "loss": 4.8145,
      "mean_token_accuracy": 0.227279132604599,
      "num_tokens": 75512345.0,
      "step": 36390
    },
    {
      "entropy": 5.58604416847229,
      "epoch": 3.301433236574746,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039314512669569323,
      "loss": 4.7896,
      "mean_token_accuracy": 0.2345356822013855,
      "num_tokens": 75523011.0,
      "step": 36395
    },
    {
      "entropy": 5.632163619995117,
      "epoch": 3.30188679245283,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039311758863879316,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2150120258331299,
      "num_tokens": 75533775.0,
      "step": 36400
    },
    {
      "entropy": 5.642969274520874,
      "epoch": 3.3023403483309144,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039309004813919117,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20664936155080796,
      "num_tokens": 75544558.0,
      "step": 36405
    },
    {
      "entropy": 5.7243523597717285,
      "epoch": 3.3027939042089987,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039306250519745696,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21097481548786162,
      "num_tokens": 75555599.0,
      "step": 36410
    },
    {
      "entropy": 5.6358414649963375,
      "epoch": 3.303247460087083,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039303495981416007,
      "loss": 4.8688,
      "mean_token_accuracy": 0.2211988389492035,
      "num_tokens": 75566320.0,
      "step": 36415
    },
    {
      "entropy": 5.636953592300415,
      "epoch": 3.3037010159651667,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003930074119898701,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21995309740304947,
      "num_tokens": 75576069.0,
      "step": 36420
    },
    {
      "entropy": 5.641560077667236,
      "epoch": 3.304154571843251,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003929798617251567,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21882313936948777,
      "num_tokens": 75586012.0,
      "step": 36425
    },
    {
      "entropy": 5.681649494171142,
      "epoch": 3.3046081277213353,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003929523090205898,
      "loss": 4.9705,
      "mean_token_accuracy": 0.211904639005661,
      "num_tokens": 75596499.0,
      "step": 36430
    },
    {
      "entropy": 5.583703279495239,
      "epoch": 3.3050616835994195,
      "grad_norm": 1.296875,
      "learning_rate": 0.000392924753876739,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22787392288446426,
      "num_tokens": 75606329.0,
      "step": 36435
    },
    {
      "entropy": 5.628923606872559,
      "epoch": 3.305515239477504,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003928971962941743,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21493573635816574,
      "num_tokens": 75617371.0,
      "step": 36440
    },
    {
      "entropy": 5.662423324584961,
      "epoch": 3.3059687953555876,
      "grad_norm": 1.25,
      "learning_rate": 0.0003928696362734655,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22077554017305373,
      "num_tokens": 75627949.0,
      "step": 36445
    },
    {
      "entropy": 5.643526458740235,
      "epoch": 3.306422351233672,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039284207381518266,
      "loss": 4.9447,
      "mean_token_accuracy": 0.21727335900068284,
      "num_tokens": 75638642.0,
      "step": 36450
    },
    {
      "entropy": 5.585008001327514,
      "epoch": 3.306875907111756,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039281450891989555,
      "loss": 4.8713,
      "mean_token_accuracy": 0.222797192633152,
      "num_tokens": 75649997.0,
      "step": 36455
    },
    {
      "entropy": 5.601866149902344,
      "epoch": 3.3073294629898404,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003927869415881745,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22872352302074433,
      "num_tokens": 75660223.0,
      "step": 36460
    },
    {
      "entropy": 5.61545729637146,
      "epoch": 3.3077830188679247,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003927593718205895,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22704134434461593,
      "num_tokens": 75670550.0,
      "step": 36465
    },
    {
      "entropy": 5.67523159980774,
      "epoch": 3.3082365747460085,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039273179961771065,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20890735536813737,
      "num_tokens": 75681631.0,
      "step": 36470
    },
    {
      "entropy": 5.622164487838745,
      "epoch": 3.308690130624093,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003927042249801082,
      "loss": 4.8493,
      "mean_token_accuracy": 0.22572583556175232,
      "num_tokens": 75691561.0,
      "step": 36475
    },
    {
      "entropy": 5.648086261749268,
      "epoch": 3.309143686502177,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039267664790835235,
      "loss": 4.8788,
      "mean_token_accuracy": 0.22560212314128875,
      "num_tokens": 75701639.0,
      "step": 36480
    },
    {
      "entropy": 5.650271844863892,
      "epoch": 3.3095972423802613,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003926490684030134,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2257583260536194,
      "num_tokens": 75712823.0,
      "step": 36485
    },
    {
      "entropy": 5.62116756439209,
      "epoch": 3.3100507982583456,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039262148646466185,
      "loss": 4.9272,
      "mean_token_accuracy": 0.21811418533325194,
      "num_tokens": 75723221.0,
      "step": 36490
    },
    {
      "entropy": 5.616157102584839,
      "epoch": 3.3105043541364294,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003925939020938679,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22351250499486924,
      "num_tokens": 75733565.0,
      "step": 36495
    },
    {
      "entropy": 5.68096957206726,
      "epoch": 3.3109579100145137,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039256631529120206,
      "loss": 4.8707,
      "mean_token_accuracy": 0.2291053056716919,
      "num_tokens": 75743511.0,
      "step": 36500
    },
    {
      "entropy": 5.728188228607178,
      "epoch": 3.311411465892598,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003925387260572349,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2139803558588028,
      "num_tokens": 75754593.0,
      "step": 36505
    },
    {
      "entropy": 5.559893608093262,
      "epoch": 3.311865021770682,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039251113439253693,
      "loss": 4.8422,
      "mean_token_accuracy": 0.23125759959220887,
      "num_tokens": 75764534.0,
      "step": 36510
    },
    {
      "entropy": 5.576159286499023,
      "epoch": 3.3123185776487665,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003924835402976788,
      "loss": 4.836,
      "mean_token_accuracy": 0.22600491642951964,
      "num_tokens": 75774707.0,
      "step": 36515
    },
    {
      "entropy": 5.54131326675415,
      "epoch": 3.3127721335268507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039245594377323106,
      "loss": 4.7895,
      "mean_token_accuracy": 0.22175977677106856,
      "num_tokens": 75784900.0,
      "step": 36520
    },
    {
      "entropy": 5.667115545272827,
      "epoch": 3.3132256894049346,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003924283448197645,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21477226465940474,
      "num_tokens": 75795711.0,
      "step": 36525
    },
    {
      "entropy": 5.639719295501709,
      "epoch": 3.313679245283019,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003924007434378498,
      "loss": 4.8893,
      "mean_token_accuracy": 0.2188007965683937,
      "num_tokens": 75805951.0,
      "step": 36530
    },
    {
      "entropy": 5.620043039321899,
      "epoch": 3.314132801161103,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003923731396280579,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22319262176752092,
      "num_tokens": 75816023.0,
      "step": 36535
    },
    {
      "entropy": 5.684797716140747,
      "epoch": 3.3145863570391874,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039234553339095944,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21676614433526992,
      "num_tokens": 75827395.0,
      "step": 36540
    },
    {
      "entropy": 5.671935987472534,
      "epoch": 3.315039912917271,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039231792472712554,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22910070717334746,
      "num_tokens": 75836663.0,
      "step": 36545
    },
    {
      "entropy": 5.632699489593506,
      "epoch": 3.3154934687953554,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000392290313637127,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21638139188289643,
      "num_tokens": 75846686.0,
      "step": 36550
    },
    {
      "entropy": 5.59012188911438,
      "epoch": 3.3159470246734397,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039226270012153485,
      "loss": 4.932,
      "mean_token_accuracy": 0.21944591253995896,
      "num_tokens": 75857386.0,
      "step": 36555
    },
    {
      "entropy": 5.63440637588501,
      "epoch": 3.316400580551524,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003922350841809203,
      "loss": 4.9085,
      "mean_token_accuracy": 0.21591539829969406,
      "num_tokens": 75867003.0,
      "step": 36560
    },
    {
      "entropy": 5.6424946784973145,
      "epoch": 3.3168541364296082,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003922074658158543,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2222453013062477,
      "num_tokens": 75878155.0,
      "step": 36565
    },
    {
      "entropy": 5.596266889572144,
      "epoch": 3.3173076923076925,
      "grad_norm": 1.25,
      "learning_rate": 0.00039217984502690797,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22051066756248475,
      "num_tokens": 75888478.0,
      "step": 36570
    },
    {
      "entropy": 5.559034824371338,
      "epoch": 3.3177612481857763,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039215222181465263,
      "loss": 4.835,
      "mean_token_accuracy": 0.2222893238067627,
      "num_tokens": 75899541.0,
      "step": 36575
    },
    {
      "entropy": 5.6551636219024655,
      "epoch": 3.3182148040638606,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039212459617965956,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2220456063747406,
      "num_tokens": 75910175.0,
      "step": 36580
    },
    {
      "entropy": 5.603334617614746,
      "epoch": 3.318668359941945,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003920969681224999,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21762137711048127,
      "num_tokens": 75919713.0,
      "step": 36585
    },
    {
      "entropy": 5.625803756713867,
      "epoch": 3.319121915820029,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039206933764374515,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21427001804113388,
      "num_tokens": 75929366.0,
      "step": 36590
    },
    {
      "entropy": 5.726399517059326,
      "epoch": 3.3195754716981134,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039204170474396656,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2068703591823578,
      "num_tokens": 75940284.0,
      "step": 36595
    },
    {
      "entropy": 5.614320516586304,
      "epoch": 3.3200290275761972,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003920140694237358,
      "loss": 4.8593,
      "mean_token_accuracy": 0.2233371004462242,
      "num_tokens": 75950752.0,
      "step": 36600
    },
    {
      "entropy": 5.6806182861328125,
      "epoch": 3.3204825834542815,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003919864316836242,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20822545737028123,
      "num_tokens": 75961308.0,
      "step": 36605
    },
    {
      "entropy": 5.6629071712493895,
      "epoch": 3.3209361393323658,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003919587915242034,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22195300310850144,
      "num_tokens": 75972317.0,
      "step": 36610
    },
    {
      "entropy": 5.610107803344727,
      "epoch": 3.32138969521045,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039193114894604504,
      "loss": 4.8837,
      "mean_token_accuracy": 0.2312440127134323,
      "num_tokens": 75982769.0,
      "step": 36615
    },
    {
      "entropy": 5.613437271118164,
      "epoch": 3.3218432510885343,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039190350394972074,
      "loss": 4.8568,
      "mean_token_accuracy": 0.2262508124113083,
      "num_tokens": 75992209.0,
      "step": 36620
    },
    {
      "entropy": 5.678586053848266,
      "epoch": 3.322296806966618,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003918758565358021,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22328823804855347,
      "num_tokens": 76001749.0,
      "step": 36625
    },
    {
      "entropy": 5.695167112350464,
      "epoch": 3.3227503628447024,
      "grad_norm": 1.25,
      "learning_rate": 0.000391848206704861,
      "loss": 4.8842,
      "mean_token_accuracy": 0.21765912771224977,
      "num_tokens": 76011326.0,
      "step": 36630
    },
    {
      "entropy": 5.573837375640869,
      "epoch": 3.3232039187227866,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039182055445746917,
      "loss": 4.8169,
      "mean_token_accuracy": 0.22764573246240616,
      "num_tokens": 76021527.0,
      "step": 36635
    },
    {
      "entropy": 5.582305240631103,
      "epoch": 3.323657474600871,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039179289979419853,
      "loss": 4.7488,
      "mean_token_accuracy": 0.23210876286029816,
      "num_tokens": 76031217.0,
      "step": 36640
    },
    {
      "entropy": 5.630798864364624,
      "epoch": 3.324111030478955,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039176524271562096,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21261018216609956,
      "num_tokens": 76041936.0,
      "step": 36645
    },
    {
      "entropy": 5.641689872741699,
      "epoch": 3.324564586357039,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039173758322230847,
      "loss": 4.946,
      "mean_token_accuracy": 0.21993618458509445,
      "num_tokens": 76052431.0,
      "step": 36650
    },
    {
      "entropy": 5.634128427505493,
      "epoch": 3.3250181422351233,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003917099213148329,
      "loss": 4.8116,
      "mean_token_accuracy": 0.22751404643058776,
      "num_tokens": 76061589.0,
      "step": 36655
    },
    {
      "entropy": 5.702914667129517,
      "epoch": 3.3254716981132075,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039168225699376646,
      "loss": 4.9286,
      "mean_token_accuracy": 0.22597684264183043,
      "num_tokens": 76071642.0,
      "step": 36660
    },
    {
      "entropy": 5.602454423904419,
      "epoch": 3.325925253991292,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039165459025968126,
      "loss": 4.8332,
      "mean_token_accuracy": 0.2252756506204605,
      "num_tokens": 76081084.0,
      "step": 36665
    },
    {
      "entropy": 5.6210531234741214,
      "epoch": 3.326378809869376,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003916269211131494,
      "loss": 4.8984,
      "mean_token_accuracy": 0.21773461997509003,
      "num_tokens": 76091125.0,
      "step": 36670
    },
    {
      "entropy": 5.716686582565307,
      "epoch": 3.3268323657474603,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003915992495547431,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21376671344041825,
      "num_tokens": 76101452.0,
      "step": 36675
    },
    {
      "entropy": 5.633708572387695,
      "epoch": 3.327285921625544,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039157157558503456,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21213718503713608,
      "num_tokens": 76112034.0,
      "step": 36680
    },
    {
      "entropy": 5.608516788482666,
      "epoch": 3.3277394775036284,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039154389920459615,
      "loss": 4.8846,
      "mean_token_accuracy": 0.222933891415596,
      "num_tokens": 76122395.0,
      "step": 36685
    },
    {
      "entropy": 5.673959445953369,
      "epoch": 3.3281930333817127,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003915162204140002,
      "loss": 4.8773,
      "mean_token_accuracy": 0.21718025356531143,
      "num_tokens": 76133622.0,
      "step": 36690
    },
    {
      "entropy": 5.685332250595093,
      "epoch": 3.328646589259797,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039148853921381916,
      "loss": 4.88,
      "mean_token_accuracy": 0.21922763735055922,
      "num_tokens": 76143757.0,
      "step": 36695
    },
    {
      "entropy": 5.780637454986572,
      "epoch": 3.3291001451378808,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003914608556046254,
      "loss": 5.1307,
      "mean_token_accuracy": 0.203074911236763,
      "num_tokens": 76154943.0,
      "step": 36700
    },
    {
      "entropy": 5.5843733787536625,
      "epoch": 3.329553701015965,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003914331695869914,
      "loss": 4.8975,
      "mean_token_accuracy": 0.21318686604499817,
      "num_tokens": 76165227.0,
      "step": 36705
    },
    {
      "entropy": 5.667214870452881,
      "epoch": 3.3300072568940493,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003914054811614899,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2168612614274025,
      "num_tokens": 76174127.0,
      "step": 36710
    },
    {
      "entropy": 5.710438919067383,
      "epoch": 3.3304608127721336,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003913777903286934,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21246084868907927,
      "num_tokens": 76183518.0,
      "step": 36715
    },
    {
      "entropy": 5.667507934570312,
      "epoch": 3.330914368650218,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003913500970891745,
      "loss": 4.934,
      "mean_token_accuracy": 0.21488931328058242,
      "num_tokens": 76193865.0,
      "step": 36720
    },
    {
      "entropy": 5.682605266571045,
      "epoch": 3.331367924528302,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003913224014435058,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21554100662469863,
      "num_tokens": 76204273.0,
      "step": 36725
    },
    {
      "entropy": 5.590853309631347,
      "epoch": 3.331821480406386,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039129470339226044,
      "loss": 4.8025,
      "mean_token_accuracy": 0.22912035286426544,
      "num_tokens": 76213706.0,
      "step": 36730
    },
    {
      "entropy": 5.658242321014404,
      "epoch": 3.33227503628447,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039126700293601086,
      "loss": 4.967,
      "mean_token_accuracy": 0.2194268748164177,
      "num_tokens": 76224484.0,
      "step": 36735
    },
    {
      "entropy": 5.638776445388794,
      "epoch": 3.3327285921625545,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003912393000753299,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2113253653049469,
      "num_tokens": 76236004.0,
      "step": 36740
    },
    {
      "entropy": 5.688573789596558,
      "epoch": 3.3331821480406387,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039121159481079077,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22338373214006424,
      "num_tokens": 76245109.0,
      "step": 36745
    },
    {
      "entropy": 5.672269201278686,
      "epoch": 3.3336357039187225,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039118388714296626,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22094995826482772,
      "num_tokens": 76255154.0,
      "step": 36750
    },
    {
      "entropy": 5.626294565200806,
      "epoch": 3.334089259796807,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003911561770724292,
      "loss": 4.8462,
      "mean_token_accuracy": 0.21998923122882844,
      "num_tokens": 76266014.0,
      "step": 36755
    },
    {
      "entropy": 5.529028224945068,
      "epoch": 3.334542815674891,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000391128464599753,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22455097436904908,
      "num_tokens": 76277593.0,
      "step": 36760
    },
    {
      "entropy": 5.557000541687012,
      "epoch": 3.3349963715529753,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003911007497255104,
      "loss": 4.7952,
      "mean_token_accuracy": 0.2335573747754097,
      "num_tokens": 76287816.0,
      "step": 36765
    },
    {
      "entropy": 5.657432889938354,
      "epoch": 3.3354499274310596,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003910730324502748,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21621690690517426,
      "num_tokens": 76298299.0,
      "step": 36770
    },
    {
      "entropy": 5.64734992980957,
      "epoch": 3.335903483309144,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003910453127746192,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21185123175382614,
      "num_tokens": 76308747.0,
      "step": 36775
    },
    {
      "entropy": 5.633251667022705,
      "epoch": 3.3363570391872277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003910175906991171,
      "loss": 4.8352,
      "mean_token_accuracy": 0.2219248116016388,
      "num_tokens": 76318923.0,
      "step": 36780
    },
    {
      "entropy": 5.666672945022583,
      "epoch": 3.336810595065312,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003909898662243415,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21398621648550034,
      "num_tokens": 76327985.0,
      "step": 36785
    },
    {
      "entropy": 5.648075580596924,
      "epoch": 3.3372641509433962,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039096213935086604,
      "loss": 5.001,
      "mean_token_accuracy": 0.21219079047441483,
      "num_tokens": 76338933.0,
      "step": 36790
    },
    {
      "entropy": 5.770057249069214,
      "epoch": 3.3377177068214805,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039093441007926396,
      "loss": 5.0038,
      "mean_token_accuracy": 0.2134322240948677,
      "num_tokens": 76349601.0,
      "step": 36795
    },
    {
      "entropy": 5.660004901885986,
      "epoch": 3.3381712626995648,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039090667841010874,
      "loss": 4.8662,
      "mean_token_accuracy": 0.2266321823000908,
      "num_tokens": 76359968.0,
      "step": 36800
    },
    {
      "entropy": 5.547734928131104,
      "epoch": 3.3386248185776486,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003908789443439737,
      "loss": 4.8116,
      "mean_token_accuracy": 0.22933948487043382,
      "num_tokens": 76369864.0,
      "step": 36805
    },
    {
      "entropy": 5.577858543395996,
      "epoch": 3.339078374455733,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003908512078814329,
      "loss": 4.8123,
      "mean_token_accuracy": 0.22675940841436387,
      "num_tokens": 76379158.0,
      "step": 36810
    },
    {
      "entropy": 5.62724461555481,
      "epoch": 3.339531930333817,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039082346902305937,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22314077764749526,
      "num_tokens": 76388798.0,
      "step": 36815
    },
    {
      "entropy": 5.570607900619507,
      "epoch": 3.3399854862119014,
      "grad_norm": 1.1875,
      "learning_rate": 0.000390795727769427,
      "loss": 4.8534,
      "mean_token_accuracy": 0.225554059445858,
      "num_tokens": 76399315.0,
      "step": 36820
    },
    {
      "entropy": 5.633226108551026,
      "epoch": 3.3404390420899857,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003907679841211095,
      "loss": 4.92,
      "mean_token_accuracy": 0.2147127464413643,
      "num_tokens": 76408841.0,
      "step": 36825
    },
    {
      "entropy": 5.6968847751617435,
      "epoch": 3.3408925979680695,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003907402380786806,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2105090230703354,
      "num_tokens": 76420055.0,
      "step": 36830
    },
    {
      "entropy": 5.60674901008606,
      "epoch": 3.3413461538461537,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003907124896427141,
      "loss": 4.8321,
      "mean_token_accuracy": 0.2286544471979141,
      "num_tokens": 76430921.0,
      "step": 36835
    },
    {
      "entropy": 5.626554727554321,
      "epoch": 3.341799709724238,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003906847388137837,
      "loss": 4.8723,
      "mean_token_accuracy": 0.21629425883293152,
      "num_tokens": 76441216.0,
      "step": 36840
    },
    {
      "entropy": 5.65028944015503,
      "epoch": 3.3422532656023223,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003906569855924636,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22278068959712982,
      "num_tokens": 76451940.0,
      "step": 36845
    },
    {
      "entropy": 5.672978687286377,
      "epoch": 3.3427068214804065,
      "grad_norm": 1.25,
      "learning_rate": 0.0003906292299793275,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22534059882164,
      "num_tokens": 76461701.0,
      "step": 36850
    },
    {
      "entropy": 5.674586582183838,
      "epoch": 3.3431603773584904,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039060147197494943,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20937626659870148,
      "num_tokens": 76473658.0,
      "step": 36855
    },
    {
      "entropy": 5.63720850944519,
      "epoch": 3.3436139332365746,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039057371157990347,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22110254019498826,
      "num_tokens": 76483821.0,
      "step": 36860
    },
    {
      "entropy": 5.616532707214356,
      "epoch": 3.344067489114659,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039054594879476374,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2158988744020462,
      "num_tokens": 76494699.0,
      "step": 36865
    },
    {
      "entropy": 5.653545284271241,
      "epoch": 3.344521044992743,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039051818362010433,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22793588489294053,
      "num_tokens": 76504299.0,
      "step": 36870
    },
    {
      "entropy": 5.703910684585571,
      "epoch": 3.3449746008708274,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003904904160564994,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21410917937755586,
      "num_tokens": 76514469.0,
      "step": 36875
    },
    {
      "entropy": 5.657998752593994,
      "epoch": 3.3454281567489117,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003904626461045232,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21513868421316146,
      "num_tokens": 76524666.0,
      "step": 36880
    },
    {
      "entropy": 5.553595781326294,
      "epoch": 3.3458817126269955,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039043487376475017,
      "loss": 4.8824,
      "mean_token_accuracy": 0.21571989506483077,
      "num_tokens": 76534787.0,
      "step": 36885
    },
    {
      "entropy": 5.626442337036133,
      "epoch": 3.34633526850508,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039040709903775446,
      "loss": 4.91,
      "mean_token_accuracy": 0.21576206535100936,
      "num_tokens": 76545055.0,
      "step": 36890
    },
    {
      "entropy": 5.584850931167603,
      "epoch": 3.346788824383164,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003903793219241106,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22780652195215226,
      "num_tokens": 76555458.0,
      "step": 36895
    },
    {
      "entropy": 5.597615909576416,
      "epoch": 3.3472423802612483,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039035154242439294,
      "loss": 4.8382,
      "mean_token_accuracy": 0.22727704793214798,
      "num_tokens": 76566272.0,
      "step": 36900
    },
    {
      "entropy": 5.70754919052124,
      "epoch": 3.347695936139332,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003903237605391759,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2175618126988411,
      "num_tokens": 76575656.0,
      "step": 36905
    },
    {
      "entropy": 5.712425422668457,
      "epoch": 3.3481494920174164,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003902959762690342,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2206042915582657,
      "num_tokens": 76586481.0,
      "step": 36910
    },
    {
      "entropy": 5.618427181243897,
      "epoch": 3.3486030478955007,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003902681896145423,
      "loss": 4.839,
      "mean_token_accuracy": 0.2296876609325409,
      "num_tokens": 76596023.0,
      "step": 36915
    },
    {
      "entropy": 5.53643536567688,
      "epoch": 3.349056603773585,
      "grad_norm": 1.125,
      "learning_rate": 0.00039024040057627487,
      "loss": 4.7993,
      "mean_token_accuracy": 0.22770035415887832,
      "num_tokens": 76605887.0,
      "step": 36920
    },
    {
      "entropy": 5.583228921890258,
      "epoch": 3.349510159651669,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003902126091548066,
      "loss": 4.8174,
      "mean_token_accuracy": 0.23167879432439803,
      "num_tokens": 76616007.0,
      "step": 36925
    },
    {
      "entropy": 5.664948844909668,
      "epoch": 3.3499637155297535,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039018481535071215,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21813850849866867,
      "num_tokens": 76626253.0,
      "step": 36930
    },
    {
      "entropy": 5.644758558273315,
      "epoch": 3.3504172714078373,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003901570191645665,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2226046070456505,
      "num_tokens": 76636341.0,
      "step": 36935
    },
    {
      "entropy": 5.6173115253448485,
      "epoch": 3.3508708272859216,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039012922059694416,
      "loss": 4.8527,
      "mean_token_accuracy": 0.22423526346683503,
      "num_tokens": 76647616.0,
      "step": 36940
    },
    {
      "entropy": 5.617476511001587,
      "epoch": 3.351324383164006,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003901014196484203,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22635423094034196,
      "num_tokens": 76658261.0,
      "step": 36945
    },
    {
      "entropy": 5.600188827514648,
      "epoch": 3.35177793904209,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039007361631956976,
      "loss": 4.888,
      "mean_token_accuracy": 0.2141212910413742,
      "num_tokens": 76667985.0,
      "step": 36950
    },
    {
      "entropy": 5.6251383304595945,
      "epoch": 3.3522314949201744,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039004581061096745,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2227220669388771,
      "num_tokens": 76678538.0,
      "step": 36955
    },
    {
      "entropy": 5.6550829887390135,
      "epoch": 3.352685050798258,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003900180025231886,
      "loss": 4.791,
      "mean_token_accuracy": 0.22498877644538878,
      "num_tokens": 76688052.0,
      "step": 36960
    },
    {
      "entropy": 5.590238904953003,
      "epoch": 3.3531386066763424,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038999019205680796,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2230977326631546,
      "num_tokens": 76699241.0,
      "step": 36965
    },
    {
      "entropy": 5.609111690521241,
      "epoch": 3.3535921625544267,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000389962379212401,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22016009092330932,
      "num_tokens": 76708979.0,
      "step": 36970
    },
    {
      "entropy": 5.651079273223877,
      "epoch": 3.354045718432511,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003899345639905426,
      "loss": 4.8535,
      "mean_token_accuracy": 0.2220097839832306,
      "num_tokens": 76719330.0,
      "step": 36975
    },
    {
      "entropy": 5.675118160247803,
      "epoch": 3.3544992743105952,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003899067463918083,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2108915165066719,
      "num_tokens": 76730210.0,
      "step": 36980
    },
    {
      "entropy": 5.72794623374939,
      "epoch": 3.354952830188679,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038987892641677307,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21876634955406188,
      "num_tokens": 76739977.0,
      "step": 36985
    },
    {
      "entropy": 5.606538152694702,
      "epoch": 3.3554063860667633,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003898511040660124,
      "loss": 4.8716,
      "mean_token_accuracy": 0.21951078474521638,
      "num_tokens": 76750200.0,
      "step": 36990
    },
    {
      "entropy": 5.6927406787872314,
      "epoch": 3.3558599419448476,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003898232793401017,
      "loss": 4.984,
      "mean_token_accuracy": 0.2083211973309517,
      "num_tokens": 76761420.0,
      "step": 36995
    },
    {
      "entropy": 5.610450220108032,
      "epoch": 3.356313497822932,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038979545223961627,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22148122191429137,
      "num_tokens": 76771872.0,
      "step": 37000
    },
    {
      "entropy": 5.57882080078125,
      "epoch": 3.356767053701016,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038976762276513163,
      "loss": 4.8524,
      "mean_token_accuracy": 0.22403264045715332,
      "num_tokens": 76782303.0,
      "step": 37005
    },
    {
      "entropy": 5.676229524612427,
      "epoch": 3.3572206095791,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003897397909172232,
      "loss": 4.975,
      "mean_token_accuracy": 0.21418311446905136,
      "num_tokens": 76792595.0,
      "step": 37010
    },
    {
      "entropy": 5.688700151443482,
      "epoch": 3.3576741654571842,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003897119566964669,
      "loss": 4.93,
      "mean_token_accuracy": 0.2173720806837082,
      "num_tokens": 76802577.0,
      "step": 37015
    },
    {
      "entropy": 5.648487854003906,
      "epoch": 3.3581277213352685,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003896841201034379,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22972912043333055,
      "num_tokens": 76812686.0,
      "step": 37020
    },
    {
      "entropy": 5.552531290054321,
      "epoch": 3.3585812772133528,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003896562811387121,
      "loss": 4.7747,
      "mean_token_accuracy": 0.23097454458475114,
      "num_tokens": 76823641.0,
      "step": 37025
    },
    {
      "entropy": 5.656320858001709,
      "epoch": 3.359034833091437,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038962843980286523,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21978622525930405,
      "num_tokens": 76834060.0,
      "step": 37030
    },
    {
      "entropy": 5.717332077026367,
      "epoch": 3.3594883889695213,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038960059609647306,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2201485112309456,
      "num_tokens": 76845078.0,
      "step": 37035
    },
    {
      "entropy": 5.687024450302124,
      "epoch": 3.359941944847605,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003895727500201113,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21406721323728561,
      "num_tokens": 76856539.0,
      "step": 37040
    },
    {
      "entropy": 5.726544904708862,
      "epoch": 3.3603955007256894,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038954490157435583,
      "loss": 4.993,
      "mean_token_accuracy": 0.2162439852952957,
      "num_tokens": 76866941.0,
      "step": 37045
    },
    {
      "entropy": 5.67036452293396,
      "epoch": 3.3608490566037736,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038951705075978264,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2171493262052536,
      "num_tokens": 76878205.0,
      "step": 37050
    },
    {
      "entropy": 5.674322271347046,
      "epoch": 3.361302612481858,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003894891975769677,
      "loss": 4.9326,
      "mean_token_accuracy": 0.20860136896371842,
      "num_tokens": 76888609.0,
      "step": 37055
    },
    {
      "entropy": 5.590348100662231,
      "epoch": 3.3617561683599417,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003894613420264869,
      "loss": 4.8955,
      "mean_token_accuracy": 0.2163400486111641,
      "num_tokens": 76899637.0,
      "step": 37060
    },
    {
      "entropy": 5.644905233383179,
      "epoch": 3.362209724238026,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038943348410891643,
      "loss": 4.9187,
      "mean_token_accuracy": 0.2158982515335083,
      "num_tokens": 76911186.0,
      "step": 37065
    },
    {
      "entropy": 5.672036266326904,
      "epoch": 3.3626632801161103,
      "grad_norm": 1.25,
      "learning_rate": 0.0003894056238248324,
      "loss": 4.8665,
      "mean_token_accuracy": 0.21881164312362672,
      "num_tokens": 76921704.0,
      "step": 37070
    },
    {
      "entropy": 5.587625932693482,
      "epoch": 3.3631168359941945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003893777611748108,
      "loss": 4.863,
      "mean_token_accuracy": 0.22683814764022828,
      "num_tokens": 76932362.0,
      "step": 37075
    },
    {
      "entropy": 5.661067008972168,
      "epoch": 3.363570391872279,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038934989615942807,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21478288173675536,
      "num_tokens": 76942581.0,
      "step": 37080
    },
    {
      "entropy": 5.655270290374756,
      "epoch": 3.364023947750363,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003893220287792603,
      "loss": 4.8944,
      "mean_token_accuracy": 0.2186492696404457,
      "num_tokens": 76952912.0,
      "step": 37085
    },
    {
      "entropy": 5.681910562515259,
      "epoch": 3.364477503628447,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038929415903488383,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2178400918841362,
      "num_tokens": 76963320.0,
      "step": 37090
    },
    {
      "entropy": 5.613359355926514,
      "epoch": 3.364931059506531,
      "grad_norm": 1.25,
      "learning_rate": 0.000389266286926875,
      "loss": 4.7885,
      "mean_token_accuracy": 0.22660157829523087,
      "num_tokens": 76973448.0,
      "step": 37095
    },
    {
      "entropy": 5.589280605316162,
      "epoch": 3.3653846153846154,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038923841245581026,
      "loss": 4.8424,
      "mean_token_accuracy": 0.22826192975044252,
      "num_tokens": 76983671.0,
      "step": 37100
    },
    {
      "entropy": 5.676054239273071,
      "epoch": 3.3658381712626997,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038921053562226607,
      "loss": 4.9991,
      "mean_token_accuracy": 0.20990085899829863,
      "num_tokens": 76994345.0,
      "step": 37105
    },
    {
      "entropy": 5.595503473281861,
      "epoch": 3.3662917271407835,
      "grad_norm": 1.125,
      "learning_rate": 0.00038918265642681884,
      "loss": 4.8031,
      "mean_token_accuracy": 0.22392067015171052,
      "num_tokens": 77004809.0,
      "step": 37110
    },
    {
      "entropy": 5.619469738006591,
      "epoch": 3.3667452830188678,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038915477487004525,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21928932517766953,
      "num_tokens": 77015505.0,
      "step": 37115
    },
    {
      "entropy": 5.554884624481201,
      "epoch": 3.367198838896952,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038912689095252183,
      "loss": 4.7275,
      "mean_token_accuracy": 0.2352317303419113,
      "num_tokens": 77025572.0,
      "step": 37120
    },
    {
      "entropy": 5.71528992652893,
      "epoch": 3.3676523947750363,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003890990046748251,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20950431376695633,
      "num_tokens": 77037273.0,
      "step": 37125
    },
    {
      "entropy": 5.713378381729126,
      "epoch": 3.3681059506531206,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038907111603753196,
      "loss": 4.966,
      "mean_token_accuracy": 0.21572712510824205,
      "num_tokens": 77047724.0,
      "step": 37130
    },
    {
      "entropy": 5.64531569480896,
      "epoch": 3.368559506531205,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038904322504121916,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22062236517667771,
      "num_tokens": 77057839.0,
      "step": 37135
    },
    {
      "entropy": 5.612880802154541,
      "epoch": 3.3690130624092887,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003890153316864632,
      "loss": 4.8565,
      "mean_token_accuracy": 0.2150793269276619,
      "num_tokens": 77068505.0,
      "step": 37140
    },
    {
      "entropy": 5.6744067668914795,
      "epoch": 3.369466618287373,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038898743597384126,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21884512454271315,
      "num_tokens": 77078126.0,
      "step": 37145
    },
    {
      "entropy": 5.630244588851928,
      "epoch": 3.369920174165457,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038895953790393014,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22241708636283875,
      "num_tokens": 77087802.0,
      "step": 37150
    },
    {
      "entropy": 5.657577037811279,
      "epoch": 3.3703737300435415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038893163747730664,
      "loss": 4.915,
      "mean_token_accuracy": 0.22626010477542877,
      "num_tokens": 77097901.0,
      "step": 37155
    },
    {
      "entropy": 5.631954956054687,
      "epoch": 3.3708272859216257,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038890373469454794,
      "loss": 4.8405,
      "mean_token_accuracy": 0.22109645754098892,
      "num_tokens": 77107604.0,
      "step": 37160
    },
    {
      "entropy": 5.645776796340942,
      "epoch": 3.3712808417997095,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038887582955623084,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22884132117033004,
      "num_tokens": 77117410.0,
      "step": 37165
    },
    {
      "entropy": 5.676944065093994,
      "epoch": 3.371734397677794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003888479220629326,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2230340287089348,
      "num_tokens": 77129546.0,
      "step": 37170
    },
    {
      "entropy": 5.606623029708862,
      "epoch": 3.372187953555878,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003888200122152304,
      "loss": 4.8502,
      "mean_token_accuracy": 0.22369149178266526,
      "num_tokens": 77139217.0,
      "step": 37175
    },
    {
      "entropy": 5.678354024887085,
      "epoch": 3.3726415094339623,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003887921000137013,
      "loss": 4.9946,
      "mean_token_accuracy": 0.21099012792110444,
      "num_tokens": 77149694.0,
      "step": 37180
    },
    {
      "entropy": 5.611155033111572,
      "epoch": 3.3730950653120466,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003887641854589226,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22674783021211625,
      "num_tokens": 77159291.0,
      "step": 37185
    },
    {
      "entropy": 5.594446086883545,
      "epoch": 3.3735486211901304,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003887362685514716,
      "loss": 4.8634,
      "mean_token_accuracy": 0.21796692460775374,
      "num_tokens": 77170507.0,
      "step": 37190
    },
    {
      "entropy": 5.64688925743103,
      "epoch": 3.3740021770682147,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003887083492919254,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2241587221622467,
      "num_tokens": 77179673.0,
      "step": 37195
    },
    {
      "entropy": 5.68870358467102,
      "epoch": 3.374455732946299,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003886804276808617,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22971894443035126,
      "num_tokens": 77190163.0,
      "step": 37200
    },
    {
      "entropy": 5.676418352127075,
      "epoch": 3.3749092888243832,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003886525037188578,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22358739078044892,
      "num_tokens": 77201498.0,
      "step": 37205
    },
    {
      "entropy": 5.625469636917114,
      "epoch": 3.3753628447024675,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003886245774064911,
      "loss": 4.8477,
      "mean_token_accuracy": 0.2226455807685852,
      "num_tokens": 77211475.0,
      "step": 37210
    },
    {
      "entropy": 5.625943946838379,
      "epoch": 3.3758164005805513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038859664874433923,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22737003713846207,
      "num_tokens": 77221972.0,
      "step": 37215
    },
    {
      "entropy": 5.6225871562957765,
      "epoch": 3.3762699564586356,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038856871773297965,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2190821498632431,
      "num_tokens": 77232701.0,
      "step": 37220
    },
    {
      "entropy": 5.6404424667358395,
      "epoch": 3.37672351233672,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038854078437299005,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2315544843673706,
      "num_tokens": 77243242.0,
      "step": 37225
    },
    {
      "entropy": 5.617395973205566,
      "epoch": 3.377177068214804,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003885128486649482,
      "loss": 4.8283,
      "mean_token_accuracy": 0.22328327894210814,
      "num_tokens": 77254279.0,
      "step": 37230
    },
    {
      "entropy": 5.627441310882569,
      "epoch": 3.3776306240928884,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038848491060943164,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22122459560632707,
      "num_tokens": 77265483.0,
      "step": 37235
    },
    {
      "entropy": 5.5777356147766115,
      "epoch": 3.3780841799709727,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003884569702070182,
      "loss": 4.8363,
      "mean_token_accuracy": 0.21845719665288926,
      "num_tokens": 77275848.0,
      "step": 37240
    },
    {
      "entropy": 5.584741640090942,
      "epoch": 3.3785377358490565,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038842902745828573,
      "loss": 4.8538,
      "mean_token_accuracy": 0.22069669812917708,
      "num_tokens": 77285943.0,
      "step": 37245
    },
    {
      "entropy": 5.608461809158325,
      "epoch": 3.3789912917271407,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003884010823638121,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22333627939224243,
      "num_tokens": 77295276.0,
      "step": 37250
    },
    {
      "entropy": 5.679275941848755,
      "epoch": 3.379444847605225,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003883731349241751,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22171399742364883,
      "num_tokens": 77305713.0,
      "step": 37255
    },
    {
      "entropy": 5.650550603866577,
      "epoch": 3.3798984034833093,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003883451851399528,
      "loss": 4.9611,
      "mean_token_accuracy": 0.2204166069626808,
      "num_tokens": 77316717.0,
      "step": 37260
    },
    {
      "entropy": 5.595538806915283,
      "epoch": 3.380351959361393,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038831723301172325,
      "loss": 4.8204,
      "mean_token_accuracy": 0.22316533625125884,
      "num_tokens": 77327763.0,
      "step": 37265
    },
    {
      "entropy": 5.670816946029663,
      "epoch": 3.3808055152394774,
      "grad_norm": 1.375,
      "learning_rate": 0.0003882892785400644,
      "loss": 4.9524,
      "mean_token_accuracy": 0.21523924469947814,
      "num_tokens": 77337495.0,
      "step": 37270
    },
    {
      "entropy": 5.6860452651977536,
      "epoch": 3.3812590711175616,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038826132172555444,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21573468148708344,
      "num_tokens": 77348849.0,
      "step": 37275
    },
    {
      "entropy": 5.623383569717407,
      "epoch": 3.381712626995646,
      "grad_norm": 1.25,
      "learning_rate": 0.0003882333625687714,
      "loss": 4.9399,
      "mean_token_accuracy": 0.21307836323976517,
      "num_tokens": 77359634.0,
      "step": 37280
    },
    {
      "entropy": 5.666058349609375,
      "epoch": 3.38216618287373,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038820540107029365,
      "loss": 4.8668,
      "mean_token_accuracy": 0.22210056483745574,
      "num_tokens": 77369342.0,
      "step": 37285
    },
    {
      "entropy": 5.697228813171387,
      "epoch": 3.3826197387518144,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003881774372306993,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22103191018104554,
      "num_tokens": 77379963.0,
      "step": 37290
    },
    {
      "entropy": 5.619308233261108,
      "epoch": 3.3830732946298983,
      "grad_norm": 1.375,
      "learning_rate": 0.0003881494710505667,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2210730493068695,
      "num_tokens": 77390135.0,
      "step": 37295
    },
    {
      "entropy": 5.685325288772583,
      "epoch": 3.3835268505079825,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038812150253047427,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21846597790718078,
      "num_tokens": 77400151.0,
      "step": 37300
    },
    {
      "entropy": 5.634002351760865,
      "epoch": 3.383980406386067,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003880935316710003,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22444234043359756,
      "num_tokens": 77410723.0,
      "step": 37305
    },
    {
      "entropy": 5.598892545700073,
      "epoch": 3.384433962264151,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003880655584727233,
      "loss": 4.8243,
      "mean_token_accuracy": 0.22736201137304307,
      "num_tokens": 77421213.0,
      "step": 37310
    },
    {
      "entropy": 5.646526718139649,
      "epoch": 3.384887518142235,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003880375829362217,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2224885955452919,
      "num_tokens": 77431705.0,
      "step": 37315
    },
    {
      "entropy": 5.599805164337158,
      "epoch": 3.385341074020319,
      "grad_norm": 1.25,
      "learning_rate": 0.0003880096050620741,
      "loss": 4.8609,
      "mean_token_accuracy": 0.21971948593854904,
      "num_tokens": 77442120.0,
      "step": 37320
    },
    {
      "entropy": 5.636885929107666,
      "epoch": 3.3857946298984034,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003879816248508591,
      "loss": 4.8762,
      "mean_token_accuracy": 0.22191519886255265,
      "num_tokens": 77451992.0,
      "step": 37325
    },
    {
      "entropy": 5.6995704650878904,
      "epoch": 3.3862481857764877,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038795364230315523,
      "loss": 5.0875,
      "mean_token_accuracy": 0.20254082083702088,
      "num_tokens": 77462830.0,
      "step": 37330
    },
    {
      "entropy": 5.675750303268432,
      "epoch": 3.386701741654572,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003879256574195414,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2236740231513977,
      "num_tokens": 77472617.0,
      "step": 37335
    },
    {
      "entropy": 5.562463808059692,
      "epoch": 3.387155297532656,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038789767020059614,
      "loss": 4.8201,
      "mean_token_accuracy": 0.2271183893084526,
      "num_tokens": 77483430.0,
      "step": 37340
    },
    {
      "entropy": 5.562888717651367,
      "epoch": 3.38760885341074,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038786968064689816,
      "loss": 4.8098,
      "mean_token_accuracy": 0.2263078898191452,
      "num_tokens": 77493038.0,
      "step": 37345
    },
    {
      "entropy": 5.659874010086059,
      "epoch": 3.3880624092888243,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003878416887590265,
      "loss": 4.9428,
      "mean_token_accuracy": 0.2145719274878502,
      "num_tokens": 77502816.0,
      "step": 37350
    },
    {
      "entropy": 5.581992673873901,
      "epoch": 3.3885159651669086,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038781369453756004,
      "loss": 4.8447,
      "mean_token_accuracy": 0.2255415290594101,
      "num_tokens": 77513026.0,
      "step": 37355
    },
    {
      "entropy": 5.652221059799194,
      "epoch": 3.388969521044993,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003877856979830775,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2155012905597687,
      "num_tokens": 77523217.0,
      "step": 37360
    },
    {
      "entropy": 5.615833377838134,
      "epoch": 3.389423076923077,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038775769909615805,
      "loss": 4.9712,
      "mean_token_accuracy": 0.22026692777872087,
      "num_tokens": 77533339.0,
      "step": 37365
    },
    {
      "entropy": 5.616430568695068,
      "epoch": 3.389876632801161,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038772969787738074,
      "loss": 4.856,
      "mean_token_accuracy": 0.2229970932006836,
      "num_tokens": 77543438.0,
      "step": 37370
    },
    {
      "entropy": 5.6887706279754635,
      "epoch": 3.390330188679245,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003877016943273244,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21426608860492707,
      "num_tokens": 77552825.0,
      "step": 37375
    },
    {
      "entropy": 5.635131931304931,
      "epoch": 3.3907837445573294,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038767368844656836,
      "loss": 4.9219,
      "mean_token_accuracy": 0.21278969645500184,
      "num_tokens": 77563107.0,
      "step": 37380
    },
    {
      "entropy": 5.702635431289673,
      "epoch": 3.3912373004354137,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038764568023569177,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21049367338418962,
      "num_tokens": 77573312.0,
      "step": 37385
    },
    {
      "entropy": 5.60852165222168,
      "epoch": 3.391690856313498,
      "grad_norm": 1.25,
      "learning_rate": 0.0003876176696952737,
      "loss": 4.9092,
      "mean_token_accuracy": 0.21554463654756545,
      "num_tokens": 77583742.0,
      "step": 37390
    },
    {
      "entropy": 5.733423137664795,
      "epoch": 3.3921444121915822,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038758965682589357,
      "loss": 4.9426,
      "mean_token_accuracy": 0.20895381271839142,
      "num_tokens": 77593667.0,
      "step": 37395
    },
    {
      "entropy": 5.642626810073852,
      "epoch": 3.392597968069666,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003875616416281307,
      "loss": 4.8424,
      "mean_token_accuracy": 0.21972652524709702,
      "num_tokens": 77604249.0,
      "step": 37400
    },
    {
      "entropy": 5.602638244628906,
      "epoch": 3.3930515239477503,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038753362410256435,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22382428199052812,
      "num_tokens": 77615095.0,
      "step": 37405
    },
    {
      "entropy": 5.588194799423218,
      "epoch": 3.3935050798258346,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003875056042497739,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22604676187038422,
      "num_tokens": 77625182.0,
      "step": 37410
    },
    {
      "entropy": 5.643997526168823,
      "epoch": 3.393958635703919,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038747758207033903,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22249070405960084,
      "num_tokens": 77635423.0,
      "step": 37415
    },
    {
      "entropy": 5.550887393951416,
      "epoch": 3.3944121915820027,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003874495575648391,
      "loss": 4.7797,
      "mean_token_accuracy": 0.2261244237422943,
      "num_tokens": 77645262.0,
      "step": 37420
    },
    {
      "entropy": 5.6124796867370605,
      "epoch": 3.394865747460087,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038742153073385356,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21230219751596452,
      "num_tokens": 77657647.0,
      "step": 37425
    },
    {
      "entropy": 5.629698801040649,
      "epoch": 3.3953193033381712,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038739350157796214,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2191387176513672,
      "num_tokens": 77668074.0,
      "step": 37430
    },
    {
      "entropy": 5.702228021621704,
      "epoch": 3.3957728592162555,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038736547009774445,
      "loss": 4.8502,
      "mean_token_accuracy": 0.21971208304166795,
      "num_tokens": 77678661.0,
      "step": 37435
    },
    {
      "entropy": 5.644606971740723,
      "epoch": 3.3962264150943398,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003873374362937803,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21349372565746308,
      "num_tokens": 77689206.0,
      "step": 37440
    },
    {
      "entropy": 5.646375417709351,
      "epoch": 3.396679970972424,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003873094001666493,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2182112842798233,
      "num_tokens": 77699254.0,
      "step": 37445
    },
    {
      "entropy": 5.642358350753784,
      "epoch": 3.397133526850508,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003872813617169312,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2223755970597267,
      "num_tokens": 77709595.0,
      "step": 37450
    },
    {
      "entropy": 5.599517297744751,
      "epoch": 3.397587082728592,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038725332094520607,
      "loss": 4.8364,
      "mean_token_accuracy": 0.22112639248371124,
      "num_tokens": 77719446.0,
      "step": 37455
    },
    {
      "entropy": 5.575517082214356,
      "epoch": 3.3980406386066764,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003872252778520536,
      "loss": 4.8563,
      "mean_token_accuracy": 0.2274940252304077,
      "num_tokens": 77729723.0,
      "step": 37460
    },
    {
      "entropy": 5.590863561630249,
      "epoch": 3.3984941944847606,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003871972324380537,
      "loss": 4.8618,
      "mean_token_accuracy": 0.21867963522672654,
      "num_tokens": 77739380.0,
      "step": 37465
    },
    {
      "entropy": 5.644940614700317,
      "epoch": 3.3989477503628445,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003871691847037865,
      "loss": 4.9072,
      "mean_token_accuracy": 0.22448876351118088,
      "num_tokens": 77750683.0,
      "step": 37470
    },
    {
      "entropy": 5.641706275939941,
      "epoch": 3.3994013062409287,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038714113464983205,
      "loss": 4.9579,
      "mean_token_accuracy": 0.2079791784286499,
      "num_tokens": 77761951.0,
      "step": 37475
    },
    {
      "entropy": 5.673253965377808,
      "epoch": 3.399854862119013,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003871130822767702,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21969587057828904,
      "num_tokens": 77772165.0,
      "step": 37480
    },
    {
      "entropy": 5.67796688079834,
      "epoch": 3.4003084179970973,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003870850275851814,
      "loss": 4.9728,
      "mean_token_accuracy": 0.20794787406921386,
      "num_tokens": 77781916.0,
      "step": 37485
    },
    {
      "entropy": 5.62520580291748,
      "epoch": 3.4007619738751815,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038705697057564556,
      "loss": 4.8782,
      "mean_token_accuracy": 0.21550885438919068,
      "num_tokens": 77791397.0,
      "step": 37490
    },
    {
      "entropy": 5.560829067230225,
      "epoch": 3.401215529753266,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000387028911248743,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22356773167848587,
      "num_tokens": 77801694.0,
      "step": 37495
    },
    {
      "entropy": 5.657276630401611,
      "epoch": 3.4016690856313496,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038700084960505414,
      "loss": 4.95,
      "mean_token_accuracy": 0.2195652574300766,
      "num_tokens": 77813199.0,
      "step": 37500
    },
    {
      "entropy": 5.709721088409424,
      "epoch": 3.402122641509434,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003869727856451591,
      "loss": 5.0204,
      "mean_token_accuracy": 0.2177110493183136,
      "num_tokens": 77823689.0,
      "step": 37505
    },
    {
      "entropy": 5.630706834793091,
      "epoch": 3.402576197387518,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038694471936963825,
      "loss": 4.858,
      "mean_token_accuracy": 0.2285211369395256,
      "num_tokens": 77834331.0,
      "step": 37510
    },
    {
      "entropy": 5.630411100387573,
      "epoch": 3.4030297532656024,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038691665077907213,
      "loss": 4.8143,
      "mean_token_accuracy": 0.22167232632637024,
      "num_tokens": 77843988.0,
      "step": 37515
    },
    {
      "entropy": 5.613130950927735,
      "epoch": 3.4034833091436867,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003868885798740412,
      "loss": 4.8687,
      "mean_token_accuracy": 0.2197393298149109,
      "num_tokens": 77854553.0,
      "step": 37520
    },
    {
      "entropy": 5.600230026245117,
      "epoch": 3.4039368650217705,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038686050665512583,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22118595987558365,
      "num_tokens": 77864629.0,
      "step": 37525
    },
    {
      "entropy": 5.651218271255493,
      "epoch": 3.4043904208998548,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038683243112290673,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21115792095661162,
      "num_tokens": 77875465.0,
      "step": 37530
    },
    {
      "entropy": 5.677199268341065,
      "epoch": 3.404843976777939,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003868043532779645,
      "loss": 4.9119,
      "mean_token_accuracy": 0.2210352823138237,
      "num_tokens": 77885337.0,
      "step": 37535
    },
    {
      "entropy": 5.660326862335205,
      "epoch": 3.4052975326560233,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038677627312087967,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22385770529508592,
      "num_tokens": 77896465.0,
      "step": 37540
    },
    {
      "entropy": 5.636050510406494,
      "epoch": 3.4057510885341076,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003867481906522331,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2208396539092064,
      "num_tokens": 77906214.0,
      "step": 37545
    },
    {
      "entropy": 5.662158632278443,
      "epoch": 3.4062046444121914,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003867201058726054,
      "loss": 5.0489,
      "mean_token_accuracy": 0.2059251293540001,
      "num_tokens": 77917466.0,
      "step": 37550
    },
    {
      "entropy": 5.71481351852417,
      "epoch": 3.4066582002902757,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003866920187825775,
      "loss": 4.9659,
      "mean_token_accuracy": 0.21619915813207627,
      "num_tokens": 77927648.0,
      "step": 37555
    },
    {
      "entropy": 5.605722093582154,
      "epoch": 3.40711175616836,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038666392938273015,
      "loss": 4.8082,
      "mean_token_accuracy": 0.22768684923648835,
      "num_tokens": 77938028.0,
      "step": 37560
    },
    {
      "entropy": 5.6339057922363285,
      "epoch": 3.407565312046444,
      "grad_norm": 1.25,
      "learning_rate": 0.0003866358376736443,
      "loss": 4.8477,
      "mean_token_accuracy": 0.22474474608898162,
      "num_tokens": 77947276.0,
      "step": 37565
    },
    {
      "entropy": 5.674513292312622,
      "epoch": 3.4080188679245285,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003866077436559009,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2203516498208046,
      "num_tokens": 77956791.0,
      "step": 37570
    },
    {
      "entropy": 5.6811309337615965,
      "epoch": 3.4084724238026123,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003865796473300809,
      "loss": 4.9219,
      "mean_token_accuracy": 0.21620651185512543,
      "num_tokens": 77967402.0,
      "step": 37575
    },
    {
      "entropy": 5.742505884170532,
      "epoch": 3.4089259796806965,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003865515486967653,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2149927243590355,
      "num_tokens": 77978000.0,
      "step": 37580
    },
    {
      "entropy": 5.5713791847229,
      "epoch": 3.409379535558781,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038652344775653535,
      "loss": 4.8047,
      "mean_token_accuracy": 0.23236246407032013,
      "num_tokens": 77987959.0,
      "step": 37585
    },
    {
      "entropy": 5.60419774055481,
      "epoch": 3.409833091436865,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003864953445099721,
      "loss": 4.8681,
      "mean_token_accuracy": 0.22378165721893312,
      "num_tokens": 77997990.0,
      "step": 37590
    },
    {
      "entropy": 5.662144613265991,
      "epoch": 3.4102866473149493,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038646723895765654,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21356898099184035,
      "num_tokens": 78008831.0,
      "step": 37595
    },
    {
      "entropy": 5.72682032585144,
      "epoch": 3.4107402031930336,
      "grad_norm": 1.25,
      "learning_rate": 0.0003864391311001702,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21466707438230515,
      "num_tokens": 78019338.0,
      "step": 37600
    },
    {
      "entropy": 5.706204509735107,
      "epoch": 3.4111937590711174,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003864110209380942,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2116099402308464,
      "num_tokens": 78030700.0,
      "step": 37605
    },
    {
      "entropy": 5.656766748428344,
      "epoch": 3.4116473149492017,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038638290847200997,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23068244606256486,
      "num_tokens": 78040190.0,
      "step": 37610
    },
    {
      "entropy": 5.62103419303894,
      "epoch": 3.412100870827286,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003863547937024987,
      "loss": 4.8721,
      "mean_token_accuracy": 0.23022119402885438,
      "num_tokens": 78050847.0,
      "step": 37615
    },
    {
      "entropy": 5.719326305389404,
      "epoch": 3.4125544267053702,
      "grad_norm": 1.34375,
      "learning_rate": 0.000386326676630142,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22499766647815705,
      "num_tokens": 78060584.0,
      "step": 37620
    },
    {
      "entropy": 5.591067981719971,
      "epoch": 3.413007982583454,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003862985572555212,
      "loss": 4.8011,
      "mean_token_accuracy": 0.22327739894390106,
      "num_tokens": 78071189.0,
      "step": 37625
    },
    {
      "entropy": 5.646454572677612,
      "epoch": 3.4134615384615383,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038627043557921785,
      "loss": 4.8788,
      "mean_token_accuracy": 0.22277014404535295,
      "num_tokens": 78082054.0,
      "step": 37630
    },
    {
      "entropy": 5.597133588790894,
      "epoch": 3.4139150943396226,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003862423116018136,
      "loss": 4.8614,
      "mean_token_accuracy": 0.21728608012199402,
      "num_tokens": 78094032.0,
      "step": 37635
    },
    {
      "entropy": 5.642258834838867,
      "epoch": 3.414368650217707,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038621418532389004,
      "loss": 4.896,
      "mean_token_accuracy": 0.2204824224114418,
      "num_tokens": 78104724.0,
      "step": 37640
    },
    {
      "entropy": 5.672085237503052,
      "epoch": 3.414822206095791,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038618605674602867,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21467514485120773,
      "num_tokens": 78115927.0,
      "step": 37645
    },
    {
      "entropy": 5.667862319946289,
      "epoch": 3.4152757619738754,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038615792586881145,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21859412938356398,
      "num_tokens": 78126744.0,
      "step": 37650
    },
    {
      "entropy": 5.593398332595825,
      "epoch": 3.415729317851959,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003861297926928199,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2184928685426712,
      "num_tokens": 78137625.0,
      "step": 37655
    },
    {
      "entropy": 5.5482312679290775,
      "epoch": 3.4161828737300435,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003861016572186359,
      "loss": 4.8356,
      "mean_token_accuracy": 0.22602216750383378,
      "num_tokens": 78147709.0,
      "step": 37660
    },
    {
      "entropy": 5.691802740097046,
      "epoch": 3.4166364296081277,
      "grad_norm": 1.25,
      "learning_rate": 0.00038607351944684135,
      "loss": 5.0243,
      "mean_token_accuracy": 0.21430206298828125,
      "num_tokens": 78158634.0,
      "step": 37665
    },
    {
      "entropy": 5.716847372055054,
      "epoch": 3.417089985486212,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003860453793780181,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2103051796555519,
      "num_tokens": 78170279.0,
      "step": 37670
    },
    {
      "entropy": 5.6639501571655275,
      "epoch": 3.417543541364296,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038601723701274813,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2191453233361244,
      "num_tokens": 78180209.0,
      "step": 37675
    },
    {
      "entropy": 5.652136945724488,
      "epoch": 3.41799709724238,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003859890923516134,
      "loss": 4.9572,
      "mean_token_accuracy": 0.2149360254406929,
      "num_tokens": 78189937.0,
      "step": 37680
    },
    {
      "entropy": 5.649466896057129,
      "epoch": 3.4184506531204644,
      "grad_norm": 1.265625,
      "learning_rate": 0.000385960945395196,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2210889995098114,
      "num_tokens": 78199224.0,
      "step": 37685
    },
    {
      "entropy": 5.642710781097412,
      "epoch": 3.4189042089985486,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003859327961440779,
      "loss": 4.8437,
      "mean_token_accuracy": 0.22446128427982331,
      "num_tokens": 78209912.0,
      "step": 37690
    },
    {
      "entropy": 5.630396986007691,
      "epoch": 3.419357764876633,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038590464459884133,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22270567715168,
      "num_tokens": 78220388.0,
      "step": 37695
    },
    {
      "entropy": 5.625854778289795,
      "epoch": 3.419811320754717,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038587649076006844,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21242413073778152,
      "num_tokens": 78231094.0,
      "step": 37700
    },
    {
      "entropy": 5.647361135482788,
      "epoch": 3.420264876632801,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003858483346283415,
      "loss": 4.8719,
      "mean_token_accuracy": 0.21633055359125136,
      "num_tokens": 78241111.0,
      "step": 37705
    },
    {
      "entropy": 5.706854677200317,
      "epoch": 3.4207184325108853,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003858201762042428,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22021221965551377,
      "num_tokens": 78251036.0,
      "step": 37710
    },
    {
      "entropy": 5.633130359649658,
      "epoch": 3.4211719883889695,
      "grad_norm": 1.125,
      "learning_rate": 0.0003857920154883545,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21122273206710815,
      "num_tokens": 78261840.0,
      "step": 37715
    },
    {
      "entropy": 5.622582292556762,
      "epoch": 3.421625544267054,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003857638524812591,
      "loss": 4.8648,
      "mean_token_accuracy": 0.21868070662021638,
      "num_tokens": 78272539.0,
      "step": 37720
    },
    {
      "entropy": 5.677495765686035,
      "epoch": 3.422079100145138,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038573568718353914,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22061379998922348,
      "num_tokens": 78283849.0,
      "step": 37725
    },
    {
      "entropy": 5.671757221221924,
      "epoch": 3.422532656023222,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003857075195957768,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21955160051584244,
      "num_tokens": 78293640.0,
      "step": 37730
    },
    {
      "entropy": 5.611025142669678,
      "epoch": 3.422986211901306,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003856793497185548,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22343799024820327,
      "num_tokens": 78303980.0,
      "step": 37735
    },
    {
      "entropy": 5.62617678642273,
      "epoch": 3.4234397677793904,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003856511775524556,
      "loss": 4.8798,
      "mean_token_accuracy": 0.21981031596660613,
      "num_tokens": 78314233.0,
      "step": 37740
    },
    {
      "entropy": 5.665864038467407,
      "epoch": 3.4238933236574747,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038562300309806193,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2266559988260269,
      "num_tokens": 78324449.0,
      "step": 37745
    },
    {
      "entropy": 5.68809814453125,
      "epoch": 3.424346879535559,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003855948263559563,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2170569509267807,
      "num_tokens": 78334357.0,
      "step": 37750
    },
    {
      "entropy": 5.653690385818481,
      "epoch": 3.424800435413643,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038556664732672147,
      "loss": 4.869,
      "mean_token_accuracy": 0.2198845401406288,
      "num_tokens": 78344221.0,
      "step": 37755
    },
    {
      "entropy": 5.611452341079712,
      "epoch": 3.425253991291727,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038553846601094033,
      "loss": 4.8419,
      "mean_token_accuracy": 0.22702255994081497,
      "num_tokens": 78354097.0,
      "step": 37760
    },
    {
      "entropy": 5.751009273529053,
      "epoch": 3.4257075471698113,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003855102824091954,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2186056852340698,
      "num_tokens": 78363797.0,
      "step": 37765
    },
    {
      "entropy": 5.64217963218689,
      "epoch": 3.4261611030478956,
      "grad_norm": 1.25,
      "learning_rate": 0.0003854820965220697,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2167496368288994,
      "num_tokens": 78374084.0,
      "step": 37770
    },
    {
      "entropy": 5.643146705627442,
      "epoch": 3.42661465892598,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038545390835014614,
      "loss": 4.8841,
      "mean_token_accuracy": 0.2159845620393753,
      "num_tokens": 78385346.0,
      "step": 37775
    },
    {
      "entropy": 5.630850076675415,
      "epoch": 3.4270682148040637,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038542571789400754,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22179464995861053,
      "num_tokens": 78395498.0,
      "step": 37780
    },
    {
      "entropy": 5.603688526153564,
      "epoch": 3.427521770682148,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038539752515423704,
      "loss": 4.8549,
      "mean_token_accuracy": 0.21432248651981353,
      "num_tokens": 78405792.0,
      "step": 37785
    },
    {
      "entropy": 5.6770118236541744,
      "epoch": 3.427975326560232,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003853693301314175,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2215485543012619,
      "num_tokens": 78415672.0,
      "step": 37790
    },
    {
      "entropy": 5.59925947189331,
      "epoch": 3.4284288824383164,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003853411328261321,
      "loss": 4.8308,
      "mean_token_accuracy": 0.22380381524562837,
      "num_tokens": 78426937.0,
      "step": 37795
    },
    {
      "entropy": 5.6452069759368895,
      "epoch": 3.4288824383164007,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000385312933238964,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2225162133574486,
      "num_tokens": 78436817.0,
      "step": 37800
    },
    {
      "entropy": 5.721516561508179,
      "epoch": 3.429335994194485,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003852847313704963,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2177321195602417,
      "num_tokens": 78447348.0,
      "step": 37805
    },
    {
      "entropy": 5.684627103805542,
      "epoch": 3.429789550072569,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003852565272213123,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22226367443799971,
      "num_tokens": 78457631.0,
      "step": 37810
    },
    {
      "entropy": 5.566251230239868,
      "epoch": 3.430243105950653,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003852283207919951,
      "loss": 4.83,
      "mean_token_accuracy": 0.23379361480474473,
      "num_tokens": 78468542.0,
      "step": 37815
    },
    {
      "entropy": 5.68494291305542,
      "epoch": 3.4306966618287373,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003852001120831282,
      "loss": 4.9314,
      "mean_token_accuracy": 0.21776785999536513,
      "num_tokens": 78478580.0,
      "step": 37820
    },
    {
      "entropy": 5.70031852722168,
      "epoch": 3.4311502177068216,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038517190109529497,
      "loss": 4.9092,
      "mean_token_accuracy": 0.21715081185102464,
      "num_tokens": 78488477.0,
      "step": 37825
    },
    {
      "entropy": 5.665575456619263,
      "epoch": 3.4316037735849054,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038514368782907864,
      "loss": 4.9245,
      "mean_token_accuracy": 0.22204332798719406,
      "num_tokens": 78498887.0,
      "step": 37830
    },
    {
      "entropy": 5.611326694488525,
      "epoch": 3.4320573294629897,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038511547228506284,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22409311830997466,
      "num_tokens": 78509030.0,
      "step": 37835
    },
    {
      "entropy": 5.543166065216065,
      "epoch": 3.432510885341074,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038508725446383097,
      "loss": 4.7634,
      "mean_token_accuracy": 0.22847527712583543,
      "num_tokens": 78518254.0,
      "step": 37840
    },
    {
      "entropy": 5.6645081520080565,
      "epoch": 3.4329644412191582,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003850590343659667,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22517662942409516,
      "num_tokens": 78527966.0,
      "step": 37845
    },
    {
      "entropy": 5.656971406936646,
      "epoch": 3.4334179970972425,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038503081199205353,
      "loss": 4.962,
      "mean_token_accuracy": 0.21889282315969466,
      "num_tokens": 78537932.0,
      "step": 37850
    },
    {
      "entropy": 5.703976011276245,
      "epoch": 3.4338715529753268,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038500258734267505,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21560388505458833,
      "num_tokens": 78549213.0,
      "step": 37855
    },
    {
      "entropy": 5.613211822509766,
      "epoch": 3.4343251088534106,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038497436041841517,
      "loss": 4.8276,
      "mean_token_accuracy": 0.21863780468702315,
      "num_tokens": 78558990.0,
      "step": 37860
    },
    {
      "entropy": 5.653682136535645,
      "epoch": 3.434778664731495,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038494613121985735,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21797598749399186,
      "num_tokens": 78568496.0,
      "step": 37865
    },
    {
      "entropy": 5.6276038646697994,
      "epoch": 3.435232220609579,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003849178997475856,
      "loss": 4.7791,
      "mean_token_accuracy": 0.23200920671224595,
      "num_tokens": 78578386.0,
      "step": 37870
    },
    {
      "entropy": 5.575590467453003,
      "epoch": 3.4356857764876634,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038488966600218365,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22493296712636948,
      "num_tokens": 78588855.0,
      "step": 37875
    },
    {
      "entropy": 5.571306371688843,
      "epoch": 3.4361393323657476,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038486142998423535,
      "loss": 4.7677,
      "mean_token_accuracy": 0.2306309998035431,
      "num_tokens": 78599573.0,
      "step": 37880
    },
    {
      "entropy": 5.71565842628479,
      "epoch": 3.4365928882438315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003848331916943248,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21993970721960068,
      "num_tokens": 78609793.0,
      "step": 37885
    },
    {
      "entropy": 5.5871048927307125,
      "epoch": 3.4370464441219157,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003848049511330358,
      "loss": 4.8464,
      "mean_token_accuracy": 0.22129549980163574,
      "num_tokens": 78619637.0,
      "step": 37890
    },
    {
      "entropy": 5.685829067230225,
      "epoch": 3.4375,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003847767083009525,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22472057044506072,
      "num_tokens": 78629889.0,
      "step": 37895
    },
    {
      "entropy": 5.587181615829468,
      "epoch": 3.4379535558780843,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038474846319865876,
      "loss": 4.8551,
      "mean_token_accuracy": 0.2252215027809143,
      "num_tokens": 78639853.0,
      "step": 37900
    },
    {
      "entropy": 5.6437427520751955,
      "epoch": 3.4384071117561685,
      "grad_norm": 1.25,
      "learning_rate": 0.0003847202158267389,
      "loss": 4.8366,
      "mean_token_accuracy": 0.22473836243152617,
      "num_tokens": 78649732.0,
      "step": 37905
    },
    {
      "entropy": 5.520084428787231,
      "epoch": 3.4388606676342524,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038469196618577694,
      "loss": 4.8011,
      "mean_token_accuracy": 0.22158783823251724,
      "num_tokens": 78659794.0,
      "step": 37910
    },
    {
      "entropy": 5.700807809829712,
      "epoch": 3.4393142235123366,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003846637142763573,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21339887082576753,
      "num_tokens": 78670526.0,
      "step": 37915
    },
    {
      "entropy": 5.652344036102295,
      "epoch": 3.439767779390421,
      "grad_norm": 1.234375,
      "learning_rate": 0.000384635460099064,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22571070194244386,
      "num_tokens": 78680177.0,
      "step": 37920
    },
    {
      "entropy": 5.694871664047241,
      "epoch": 3.440221335268505,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038460720365448154,
      "loss": 5.0473,
      "mean_token_accuracy": 0.20870906710624695,
      "num_tokens": 78690958.0,
      "step": 37925
    },
    {
      "entropy": 5.712219142913819,
      "epoch": 3.4406748911465894,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003845789449431941,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2156258851289749,
      "num_tokens": 78700904.0,
      "step": 37930
    },
    {
      "entropy": 5.7467200756073,
      "epoch": 3.4411284470246732,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003845506839657862,
      "loss": 4.9743,
      "mean_token_accuracy": 0.216835817694664,
      "num_tokens": 78712085.0,
      "step": 37935
    },
    {
      "entropy": 5.594562244415283,
      "epoch": 3.4415820029027575,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003845224207228422,
      "loss": 4.8619,
      "mean_token_accuracy": 0.2226394772529602,
      "num_tokens": 78722161.0,
      "step": 37940
    },
    {
      "entropy": 5.660646104812622,
      "epoch": 3.4420355587808418,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038449415521494667,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2206542745232582,
      "num_tokens": 78732485.0,
      "step": 37945
    },
    {
      "entropy": 5.646375036239624,
      "epoch": 3.442489114658926,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003844658874426841,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21391827762126922,
      "num_tokens": 78744564.0,
      "step": 37950
    },
    {
      "entropy": 5.672148942947388,
      "epoch": 3.4429426705370103,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000384437617406639,
      "loss": 4.8934,
      "mean_token_accuracy": 0.22410544902086257,
      "num_tokens": 78755227.0,
      "step": 37955
    },
    {
      "entropy": 5.642614650726318,
      "epoch": 3.4433962264150946,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003844093451073962,
      "loss": 4.8756,
      "mean_token_accuracy": 0.21557254791259767,
      "num_tokens": 78765393.0,
      "step": 37960
    },
    {
      "entropy": 5.648634147644043,
      "epoch": 3.4438497822931784,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038438107054554027,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22499302625656128,
      "num_tokens": 78775587.0,
      "step": 37965
    },
    {
      "entropy": 5.581699228286743,
      "epoch": 3.4443033381712627,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003843527937216558,
      "loss": 4.8086,
      "mean_token_accuracy": 0.23461911380290984,
      "num_tokens": 78785098.0,
      "step": 37970
    },
    {
      "entropy": 5.587944221496582,
      "epoch": 3.444756894049347,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003843245146363278,
      "loss": 4.8467,
      "mean_token_accuracy": 0.2272839516401291,
      "num_tokens": 78795348.0,
      "step": 37975
    },
    {
      "entropy": 5.7159576416015625,
      "epoch": 3.445210449927431,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038429623329014096,
      "loss": 4.9953,
      "mean_token_accuracy": 0.2142575204372406,
      "num_tokens": 78805348.0,
      "step": 37980
    },
    {
      "entropy": 5.5927667140960695,
      "epoch": 3.445664005805515,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038426794968368015,
      "loss": 4.8264,
      "mean_token_accuracy": 0.22185047566890717,
      "num_tokens": 78814765.0,
      "step": 37985
    },
    {
      "entropy": 5.663602828979492,
      "epoch": 3.4461175616835993,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003842396638175303,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22366278171539306,
      "num_tokens": 78824745.0,
      "step": 37990
    },
    {
      "entropy": 5.60273699760437,
      "epoch": 3.4465711175616836,
      "grad_norm": 1.25,
      "learning_rate": 0.0003842113756922763,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22094383835792542,
      "num_tokens": 78835101.0,
      "step": 37995
    },
    {
      "entropy": 5.701382303237915,
      "epoch": 3.447024673439768,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038418308530850333,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21216570138931273,
      "num_tokens": 78845881.0,
      "step": 38000
    },
    {
      "entropy": 5.66708722114563,
      "epoch": 3.447478229317852,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038415479266679624,
      "loss": 4.91,
      "mean_token_accuracy": 0.21548085063695907,
      "num_tokens": 78856536.0,
      "step": 38005
    },
    {
      "entropy": 5.622337532043457,
      "epoch": 3.4479317851959363,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003841264977677403,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22257553935050964,
      "num_tokens": 78866887.0,
      "step": 38010
    },
    {
      "entropy": 5.603978872299194,
      "epoch": 3.44838534107402,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038409820061192045,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22752683758735656,
      "num_tokens": 78877032.0,
      "step": 38015
    },
    {
      "entropy": 5.587260484695435,
      "epoch": 3.4488388969521044,
      "grad_norm": 1.4375,
      "learning_rate": 0.00038406990119992215,
      "loss": 4.8579,
      "mean_token_accuracy": 0.22937188297510147,
      "num_tokens": 78887237.0,
      "step": 38020
    },
    {
      "entropy": 5.536665725708008,
      "epoch": 3.4492924528301887,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038404159953233047,
      "loss": 4.7907,
      "mean_token_accuracy": 0.23267127424478531,
      "num_tokens": 78897436.0,
      "step": 38025
    },
    {
      "entropy": 5.649680042266846,
      "epoch": 3.449746008708273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003840132956097307,
      "loss": 4.883,
      "mean_token_accuracy": 0.22053813934326172,
      "num_tokens": 78907691.0,
      "step": 38030
    },
    {
      "entropy": 5.6220708847045895,
      "epoch": 3.450199564586357,
      "grad_norm": 1.25,
      "learning_rate": 0.00038398498943270816,
      "loss": 4.9182,
      "mean_token_accuracy": 0.22164131551980973,
      "num_tokens": 78917517.0,
      "step": 38035
    },
    {
      "entropy": 5.714004182815552,
      "epoch": 3.450653120464441,
      "grad_norm": 1.25,
      "learning_rate": 0.0003839566810018483,
      "loss": 4.8762,
      "mean_token_accuracy": 0.21921992003917695,
      "num_tokens": 78927835.0,
      "step": 38040
    },
    {
      "entropy": 5.589601755142212,
      "epoch": 3.4511066763425253,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038392837031773656,
      "loss": 4.8081,
      "mean_token_accuracy": 0.22518548220396042,
      "num_tokens": 78938853.0,
      "step": 38045
    },
    {
      "entropy": 5.601043081283569,
      "epoch": 3.4515602322206096,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003839000573809583,
      "loss": 4.8653,
      "mean_token_accuracy": 0.23010605573654175,
      "num_tokens": 78948508.0,
      "step": 38050
    },
    {
      "entropy": 5.674191284179687,
      "epoch": 3.452013788098694,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038387174219209916,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22717321664094925,
      "num_tokens": 78959002.0,
      "step": 38055
    },
    {
      "entropy": 5.660146427154541,
      "epoch": 3.452467343976778,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003838434247517446,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22319020330905914,
      "num_tokens": 78970358.0,
      "step": 38060
    },
    {
      "entropy": 5.666746377944946,
      "epoch": 3.452920899854862,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003838151050604803,
      "loss": 4.8127,
      "mean_token_accuracy": 0.227305606007576,
      "num_tokens": 78979408.0,
      "step": 38065
    },
    {
      "entropy": 5.527356481552124,
      "epoch": 3.453374455732946,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038378678311889184,
      "loss": 4.8247,
      "mean_token_accuracy": 0.22888706773519515,
      "num_tokens": 78990411.0,
      "step": 38070
    },
    {
      "entropy": 5.630104684829712,
      "epoch": 3.4538280116110305,
      "grad_norm": 1.234375,
      "learning_rate": 0.000383758458927565,
      "loss": 4.963,
      "mean_token_accuracy": 0.22190680354833603,
      "num_tokens": 79001734.0,
      "step": 38075
    },
    {
      "entropy": 5.685411882400513,
      "epoch": 3.4542815674891147,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003837301324870856,
      "loss": 4.8173,
      "mean_token_accuracy": 0.22820084542036057,
      "num_tokens": 79011381.0,
      "step": 38080
    },
    {
      "entropy": 5.667036437988282,
      "epoch": 3.454735123367199,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038370180379803927,
      "loss": 4.9524,
      "mean_token_accuracy": 0.2172626495361328,
      "num_tokens": 79021124.0,
      "step": 38085
    },
    {
      "entropy": 5.673640251159668,
      "epoch": 3.455188679245283,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038367347286101197,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20204024016857147,
      "num_tokens": 79033105.0,
      "step": 38090
    },
    {
      "entropy": 5.663362169265747,
      "epoch": 3.455642235123367,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003836451396765896,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21427410691976548,
      "num_tokens": 79043458.0,
      "step": 38095
    },
    {
      "entropy": 5.667144441604615,
      "epoch": 3.4560957910014514,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038361680424535795,
      "loss": 4.9018,
      "mean_token_accuracy": 0.2220745712518692,
      "num_tokens": 79055186.0,
      "step": 38100
    },
    {
      "entropy": 5.67755913734436,
      "epoch": 3.4565493468795356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003835884665679032,
      "loss": 4.9217,
      "mean_token_accuracy": 0.2195105254650116,
      "num_tokens": 79065743.0,
      "step": 38105
    },
    {
      "entropy": 5.67501859664917,
      "epoch": 3.45700290275762,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003835601266448112,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21161009669303893,
      "num_tokens": 79077285.0,
      "step": 38110
    },
    {
      "entropy": 5.6570065975189205,
      "epoch": 3.457456458635704,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003835317844766682,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22252228409051894,
      "num_tokens": 79087885.0,
      "step": 38115
    },
    {
      "entropy": 5.68416395187378,
      "epoch": 3.457910014513788,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038350344006406016,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22118688821792604,
      "num_tokens": 79097907.0,
      "step": 38120
    },
    {
      "entropy": 5.712332153320313,
      "epoch": 3.4583635703918723,
      "grad_norm": 1.25,
      "learning_rate": 0.00038347509340757343,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20990787148475648,
      "num_tokens": 79108478.0,
      "step": 38125
    },
    {
      "entropy": 5.711361885070801,
      "epoch": 3.4588171262699565,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003834467445077941,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21531004458665848,
      "num_tokens": 79117988.0,
      "step": 38130
    },
    {
      "entropy": 5.720006656646729,
      "epoch": 3.459270682148041,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003834183933653084,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22780102044343947,
      "num_tokens": 79127786.0,
      "step": 38135
    },
    {
      "entropy": 5.6404743671417235,
      "epoch": 3.4597242380261246,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003833900399807027,
      "loss": 4.8529,
      "mean_token_accuracy": 0.22621534764766693,
      "num_tokens": 79137775.0,
      "step": 38140
    },
    {
      "entropy": 5.547027111053467,
      "epoch": 3.460177793904209,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00038336168435456345,
      "loss": 4.8162,
      "mean_token_accuracy": 0.2237204611301422,
      "num_tokens": 79147803.0,
      "step": 38145
    },
    {
      "entropy": 5.5697732925415036,
      "epoch": 3.460631349782293,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038333332648747684,
      "loss": 4.8049,
      "mean_token_accuracy": 0.22736114859580994,
      "num_tokens": 79158377.0,
      "step": 38150
    },
    {
      "entropy": 5.634149217605591,
      "epoch": 3.4610849056603774,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003833049663800295,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21759562343358993,
      "num_tokens": 79168514.0,
      "step": 38155
    },
    {
      "entropy": 5.695762872695923,
      "epoch": 3.4615384615384617,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003832766040328078,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21513009667396546,
      "num_tokens": 79177588.0,
      "step": 38160
    },
    {
      "entropy": 5.643855905532837,
      "epoch": 3.461992017416546,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003832482394463984,
      "loss": 4.871,
      "mean_token_accuracy": 0.22111586779356002,
      "num_tokens": 79188128.0,
      "step": 38165
    },
    {
      "entropy": 5.580990886688232,
      "epoch": 3.4624455732946298,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003832198726213878,
      "loss": 4.8424,
      "mean_token_accuracy": 0.22148845195770264,
      "num_tokens": 79198928.0,
      "step": 38170
    },
    {
      "entropy": 5.557035636901856,
      "epoch": 3.462899129172714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038319150355836266,
      "loss": 4.812,
      "mean_token_accuracy": 0.23224353790283203,
      "num_tokens": 79209751.0,
      "step": 38175
    },
    {
      "entropy": 5.627704668045044,
      "epoch": 3.4633526850507983,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038316313225790965,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2158006548881531,
      "num_tokens": 79220210.0,
      "step": 38180
    },
    {
      "entropy": 5.638386583328247,
      "epoch": 3.4638062409288826,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003831347587206155,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2244991734623909,
      "num_tokens": 79231530.0,
      "step": 38185
    },
    {
      "entropy": 5.596593141555786,
      "epoch": 3.4642597968069664,
      "grad_norm": 1.203125,
      "learning_rate": 0.000383106382947067,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22113476693630219,
      "num_tokens": 79242771.0,
      "step": 38190
    },
    {
      "entropy": 5.664237976074219,
      "epoch": 3.4647133526850507,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038307800493785093,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22361240535974503,
      "num_tokens": 79253713.0,
      "step": 38195
    },
    {
      "entropy": 5.638049554824829,
      "epoch": 3.465166908563135,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038304962469355415,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21809325367212296,
      "num_tokens": 79264087.0,
      "step": 38200
    },
    {
      "entropy": 5.6401184558868405,
      "epoch": 3.465620464441219,
      "grad_norm": 1.25,
      "learning_rate": 0.0003830212422147637,
      "loss": 4.8877,
      "mean_token_accuracy": 0.2231224626302719,
      "num_tokens": 79274859.0,
      "step": 38205
    },
    {
      "entropy": 5.6572778701782225,
      "epoch": 3.4660740203193035,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003829928575020663,
      "loss": 4.92,
      "mean_token_accuracy": 0.22196309119462967,
      "num_tokens": 79284778.0,
      "step": 38210
    },
    {
      "entropy": 5.679198789596557,
      "epoch": 3.4665275761973877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038296447055604916,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2135196566581726,
      "num_tokens": 79295034.0,
      "step": 38215
    },
    {
      "entropy": 5.58616452217102,
      "epoch": 3.4669811320754715,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003829360813772992,
      "loss": 4.8246,
      "mean_token_accuracy": 0.23288164883852006,
      "num_tokens": 79306013.0,
      "step": 38220
    },
    {
      "entropy": 5.626765632629395,
      "epoch": 3.467434687953556,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038290768996640366,
      "loss": 4.985,
      "mean_token_accuracy": 0.21295868456363679,
      "num_tokens": 79317590.0,
      "step": 38225
    },
    {
      "entropy": 5.69479808807373,
      "epoch": 3.46788824383164,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038287929632394957,
      "loss": 4.9308,
      "mean_token_accuracy": 0.218781079351902,
      "num_tokens": 79328669.0,
      "step": 38230
    },
    {
      "entropy": 5.561502122879029,
      "epoch": 3.4683417997097243,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003828509004505241,
      "loss": 4.8252,
      "mean_token_accuracy": 0.23009409755468369,
      "num_tokens": 79339336.0,
      "step": 38235
    },
    {
      "entropy": 5.716666030883789,
      "epoch": 3.4687953555878086,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003828225023467146,
      "loss": 4.9482,
      "mean_token_accuracy": 0.21630705446004866,
      "num_tokens": 79350199.0,
      "step": 38240
    },
    {
      "entropy": 5.6421808242797855,
      "epoch": 3.4692489114658924,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003827941020131082,
      "loss": 4.873,
      "mean_token_accuracy": 0.23125556707382203,
      "num_tokens": 79359906.0,
      "step": 38245
    },
    {
      "entropy": 5.59387354850769,
      "epoch": 3.4697024673439767,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038276569945029225,
      "loss": 4.8239,
      "mean_token_accuracy": 0.21477842926979065,
      "num_tokens": 79371395.0,
      "step": 38250
    },
    {
      "entropy": 5.629295349121094,
      "epoch": 3.470156023222061,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003827372946588543,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22268614470958709,
      "num_tokens": 79381808.0,
      "step": 38255
    },
    {
      "entropy": 5.596413707733154,
      "epoch": 3.4706095791001452,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003827088876393816,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22246748208999634,
      "num_tokens": 79391906.0,
      "step": 38260
    },
    {
      "entropy": 5.610882520675659,
      "epoch": 3.4710631349782295,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038268047839246165,
      "loss": 4.8316,
      "mean_token_accuracy": 0.2280489668250084,
      "num_tokens": 79402522.0,
      "step": 38265
    },
    {
      "entropy": 5.5595237731933596,
      "epoch": 3.4715166908563133,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038265206691868195,
      "loss": 4.805,
      "mean_token_accuracy": 0.22312385439872742,
      "num_tokens": 79412751.0,
      "step": 38270
    },
    {
      "entropy": 5.716089820861816,
      "epoch": 3.4719702467343976,
      "grad_norm": 1.25,
      "learning_rate": 0.0003826236532186301,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21648291796445845,
      "num_tokens": 79424512.0,
      "step": 38275
    },
    {
      "entropy": 5.779439544677734,
      "epoch": 3.472423802612482,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003825952372928936,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20616847276687622,
      "num_tokens": 79434082.0,
      "step": 38280
    },
    {
      "entropy": 5.676422500610352,
      "epoch": 3.472877358490566,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038256681914206026,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22302726060152053,
      "num_tokens": 79444469.0,
      "step": 38285
    },
    {
      "entropy": 5.568297004699707,
      "epoch": 3.4733309143686504,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038253839876671765,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21630964428186417,
      "num_tokens": 79455036.0,
      "step": 38290
    },
    {
      "entropy": 5.640959596633911,
      "epoch": 3.473784470246734,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003825099761674535,
      "loss": 4.874,
      "mean_token_accuracy": 0.21984777450561524,
      "num_tokens": 79464600.0,
      "step": 38295
    },
    {
      "entropy": 5.655915832519531,
      "epoch": 3.4742380261248185,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003824815513448558,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22141298651695251,
      "num_tokens": 79474912.0,
      "step": 38300
    },
    {
      "entropy": 5.5332403659820555,
      "epoch": 3.4746915820029027,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038245312429951214,
      "loss": 4.7691,
      "mean_token_accuracy": 0.22960416227579117,
      "num_tokens": 79485283.0,
      "step": 38305
    },
    {
      "entropy": 5.501477861404419,
      "epoch": 3.475145137880987,
      "grad_norm": 1.25,
      "learning_rate": 0.0003824246950320105,
      "loss": 4.7585,
      "mean_token_accuracy": 0.22108541131019593,
      "num_tokens": 79495323.0,
      "step": 38310
    },
    {
      "entropy": 5.611756324768066,
      "epoch": 3.4755986937590713,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003823962635429388,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22048257440328597,
      "num_tokens": 79505144.0,
      "step": 38315
    },
    {
      "entropy": 5.695401239395141,
      "epoch": 3.4760522496371555,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003823678298328849,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2139294594526291,
      "num_tokens": 79515402.0,
      "step": 38320
    },
    {
      "entropy": 5.639161014556885,
      "epoch": 3.4765058055152394,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038233939390243697,
      "loss": 4.8169,
      "mean_token_accuracy": 0.23429738283157348,
      "num_tokens": 79525528.0,
      "step": 38325
    },
    {
      "entropy": 5.6395810604095455,
      "epoch": 3.4769593613933236,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038231095575218303,
      "loss": 4.9113,
      "mean_token_accuracy": 0.2202695742249489,
      "num_tokens": 79535991.0,
      "step": 38330
    },
    {
      "entropy": 5.69022421836853,
      "epoch": 3.477412917271408,
      "grad_norm": 1.25,
      "learning_rate": 0.0003822825153827112,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2145649865269661,
      "num_tokens": 79545890.0,
      "step": 38335
    },
    {
      "entropy": 5.689422178268432,
      "epoch": 3.477866473149492,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038225407279460964,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22076304703950883,
      "num_tokens": 79556363.0,
      "step": 38340
    },
    {
      "entropy": 5.641707849502564,
      "epoch": 3.478320029027576,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003822256279884665,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21994021832942962,
      "num_tokens": 79566924.0,
      "step": 38345
    },
    {
      "entropy": 5.601212692260742,
      "epoch": 3.4787735849056602,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038219718096487007,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22500191777944564,
      "num_tokens": 79577322.0,
      "step": 38350
    },
    {
      "entropy": 5.741864824295044,
      "epoch": 3.4792271407837445,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038216873172440855,
      "loss": 4.9917,
      "mean_token_accuracy": 0.21258788257837297,
      "num_tokens": 79588436.0,
      "step": 38355
    },
    {
      "entropy": 5.727066087722778,
      "epoch": 3.4796806966618288,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003821402802676704,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21798423528671265,
      "num_tokens": 79599015.0,
      "step": 38360
    },
    {
      "entropy": 5.618892860412598,
      "epoch": 3.480134252539913,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003821118265952439,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22861704528331755,
      "num_tokens": 79608443.0,
      "step": 38365
    },
    {
      "entropy": 5.582566738128662,
      "epoch": 3.4805878084179973,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003820833707077176,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22075275033712388,
      "num_tokens": 79619010.0,
      "step": 38370
    },
    {
      "entropy": 5.678500986099243,
      "epoch": 3.481041364296081,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003820549126056799,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21913174241781236,
      "num_tokens": 79629665.0,
      "step": 38375
    },
    {
      "entropy": 5.55227255821228,
      "epoch": 3.4814949201741654,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003820264522897193,
      "loss": 4.7838,
      "mean_token_accuracy": 0.2277619481086731,
      "num_tokens": 79640578.0,
      "step": 38380
    },
    {
      "entropy": 5.661253356933594,
      "epoch": 3.4819484760522497,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003819979897604243,
      "loss": 4.8551,
      "mean_token_accuracy": 0.21649368107318878,
      "num_tokens": 79649576.0,
      "step": 38385
    },
    {
      "entropy": 5.646604108810425,
      "epoch": 3.482402031930334,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038196952501838376,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21030827462673188,
      "num_tokens": 79659926.0,
      "step": 38390
    },
    {
      "entropy": 5.710169744491577,
      "epoch": 3.4828555878084178,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038194105806418617,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20706610530614852,
      "num_tokens": 79669969.0,
      "step": 38395
    },
    {
      "entropy": 5.640405654907227,
      "epoch": 3.483309143686502,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003819125888984202,
      "loss": 4.8421,
      "mean_token_accuracy": 0.2249169424176216,
      "num_tokens": 79679706.0,
      "step": 38400
    },
    {
      "entropy": 5.59892201423645,
      "epoch": 3.4837626995645863,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038188411752167466,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22400908172130585,
      "num_tokens": 79691543.0,
      "step": 38405
    },
    {
      "entropy": 5.636819362640381,
      "epoch": 3.4842162554426706,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003818556439345383,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2234833225607872,
      "num_tokens": 79701657.0,
      "step": 38410
    },
    {
      "entropy": 5.609713792800903,
      "epoch": 3.484669811320755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038182716813760004,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22530845254659654,
      "num_tokens": 79711615.0,
      "step": 38415
    },
    {
      "entropy": 5.646623134613037,
      "epoch": 3.485123367198839,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003817986901314487,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22565362751483917,
      "num_tokens": 79721427.0,
      "step": 38420
    },
    {
      "entropy": 5.665906095504761,
      "epoch": 3.485576923076923,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003817702099166732,
      "loss": 4.9213,
      "mean_token_accuracy": 0.22388988435268403,
      "num_tokens": 79732565.0,
      "step": 38425
    },
    {
      "entropy": 5.579575443267823,
      "epoch": 3.486030478955007,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003817417274938626,
      "loss": 4.7562,
      "mean_token_accuracy": 0.22740735560655595,
      "num_tokens": 79744496.0,
      "step": 38430
    },
    {
      "entropy": 5.6460710048675535,
      "epoch": 3.4864840348330914,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003817132428636058,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21806008368730545,
      "num_tokens": 79755626.0,
      "step": 38435
    },
    {
      "entropy": 5.616807222366333,
      "epoch": 3.4869375907111757,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038168475602649203,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21947404146194457,
      "num_tokens": 79765447.0,
      "step": 38440
    },
    {
      "entropy": 5.653316926956177,
      "epoch": 3.48739114658926,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003816562669831103,
      "loss": 4.8878,
      "mean_token_accuracy": 0.21821122616529465,
      "num_tokens": 79776047.0,
      "step": 38445
    },
    {
      "entropy": 5.6446298122406,
      "epoch": 3.487844702467344,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038162777573404975,
      "loss": 4.8493,
      "mean_token_accuracy": 0.22709202617406846,
      "num_tokens": 79786061.0,
      "step": 38450
    },
    {
      "entropy": 5.602183818817139,
      "epoch": 3.488298258345428,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038159928227989964,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21845054179430007,
      "num_tokens": 79796844.0,
      "step": 38455
    },
    {
      "entropy": 5.664546823501587,
      "epoch": 3.4887518142235123,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003815707866212493,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22813622355461122,
      "num_tokens": 79808120.0,
      "step": 38460
    },
    {
      "entropy": 5.685071992874145,
      "epoch": 3.4892053701015966,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003815422887586878,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22445545196533204,
      "num_tokens": 79818901.0,
      "step": 38465
    },
    {
      "entropy": 5.683806324005127,
      "epoch": 3.489658925979681,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003815137886928047,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2158046245574951,
      "num_tokens": 79829090.0,
      "step": 38470
    },
    {
      "entropy": 5.72820200920105,
      "epoch": 3.490112481857765,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003814852864241892,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21561352610588075,
      "num_tokens": 79838751.0,
      "step": 38475
    },
    {
      "entropy": 5.61675877571106,
      "epoch": 3.490566037735849,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038145678195343095,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21416785269975663,
      "num_tokens": 79849410.0,
      "step": 38480
    },
    {
      "entropy": 5.731597661972046,
      "epoch": 3.491019593613933,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003814282752811193,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21647255271673202,
      "num_tokens": 79859279.0,
      "step": 38485
    },
    {
      "entropy": 5.722654104232788,
      "epoch": 3.4914731494920175,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003813997664078438,
      "loss": 4.852,
      "mean_token_accuracy": 0.2274712398648262,
      "num_tokens": 79869535.0,
      "step": 38490
    },
    {
      "entropy": 5.68737759590149,
      "epoch": 3.4919267053701017,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038137125533419403,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2029233917593956,
      "num_tokens": 79881396.0,
      "step": 38495
    },
    {
      "entropy": 5.5814307689666744,
      "epoch": 3.4923802612481856,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003813427420607596,
      "loss": 4.8459,
      "mean_token_accuracy": 0.21493924409151077,
      "num_tokens": 79891250.0,
      "step": 38500
    },
    {
      "entropy": 5.601520586013794,
      "epoch": 3.49283381712627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038131422658813004,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22096751630306244,
      "num_tokens": 79901982.0,
      "step": 38505
    },
    {
      "entropy": 5.641486120223999,
      "epoch": 3.493287373004354,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038128570891689527,
      "loss": 4.9594,
      "mean_token_accuracy": 0.214323490858078,
      "num_tokens": 79912480.0,
      "step": 38510
    },
    {
      "entropy": 5.674976253509522,
      "epoch": 3.4937409288824384,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000381257189047645,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22532671242952346,
      "num_tokens": 79923793.0,
      "step": 38515
    },
    {
      "entropy": 5.661871242523193,
      "epoch": 3.4941944847605226,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038122866698096894,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21955492198467255,
      "num_tokens": 79933858.0,
      "step": 38520
    },
    {
      "entropy": 5.698384428024292,
      "epoch": 3.494648040638607,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038120014271745693,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21500428915023803,
      "num_tokens": 79944093.0,
      "step": 38525
    },
    {
      "entropy": 5.662602519989013,
      "epoch": 3.4951015965166907,
      "grad_norm": 1.28125,
      "learning_rate": 0.000381171616257699,
      "loss": 4.9229,
      "mean_token_accuracy": 0.21361628472805022,
      "num_tokens": 79955072.0,
      "step": 38530
    },
    {
      "entropy": 5.669543743133545,
      "epoch": 3.495555152394775,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038114308760228493,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22173127233982087,
      "num_tokens": 79966949.0,
      "step": 38535
    },
    {
      "entropy": 5.601763439178467,
      "epoch": 3.4960087082728593,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003811145567518048,
      "loss": 4.857,
      "mean_token_accuracy": 0.21989901810884477,
      "num_tokens": 79977109.0,
      "step": 38540
    },
    {
      "entropy": 5.684397029876709,
      "epoch": 3.4964622641509435,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038108602370684854,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21701878756284715,
      "num_tokens": 79987094.0,
      "step": 38545
    },
    {
      "entropy": 5.673355150222778,
      "epoch": 3.4969158200290273,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003810574884680063,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22414342314004898,
      "num_tokens": 79997503.0,
      "step": 38550
    },
    {
      "entropy": 5.63979287147522,
      "epoch": 3.4973693759071116,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038102895103586817,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2210952639579773,
      "num_tokens": 80007745.0,
      "step": 38555
    },
    {
      "entropy": 5.678734254837036,
      "epoch": 3.497822931785196,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003810004114110243,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21331811547279358,
      "num_tokens": 80019330.0,
      "step": 38560
    },
    {
      "entropy": 5.68154501914978,
      "epoch": 3.49827648766328,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003809718695940649,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21800312548875808,
      "num_tokens": 80029732.0,
      "step": 38565
    },
    {
      "entropy": 5.632097339630127,
      "epoch": 3.4987300435413644,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003809433255855803,
      "loss": 4.8033,
      "mean_token_accuracy": 0.22539930045604706,
      "num_tokens": 80041052.0,
      "step": 38570
    },
    {
      "entropy": 5.662900590896607,
      "epoch": 3.4991835994194487,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038091477938616075,
      "loss": 4.9228,
      "mean_token_accuracy": 0.2210698828101158,
      "num_tokens": 80051929.0,
      "step": 38575
    },
    {
      "entropy": 5.693867301940918,
      "epoch": 3.4996371552975325,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038088623099639653,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21416809111833573,
      "num_tokens": 80061153.0,
      "step": 38580
    },
    {
      "entropy": 5.686135196685791,
      "epoch": 3.5000907111756168,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000380857680416878,
      "loss": 4.8835,
      "mean_token_accuracy": 0.22171344608068466,
      "num_tokens": 80070560.0,
      "step": 38585
    },
    {
      "entropy": 5.8162181854248045,
      "epoch": 3.500544267053701,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003808291276481957,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2174808606505394,
      "num_tokens": 80081141.0,
      "step": 38590
    },
    {
      "entropy": 5.607828569412232,
      "epoch": 3.5009978229317853,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003808005726909401,
      "loss": 4.8023,
      "mean_token_accuracy": 0.23037008792161942,
      "num_tokens": 80092114.0,
      "step": 38595
    },
    {
      "entropy": 5.577740430831909,
      "epoch": 3.501451378809869,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038077201554570175,
      "loss": 4.837,
      "mean_token_accuracy": 0.22557417750358583,
      "num_tokens": 80101966.0,
      "step": 38600
    },
    {
      "entropy": 5.690110683441162,
      "epoch": 3.5019049346879534,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003807434562130711,
      "loss": 5.0054,
      "mean_token_accuracy": 0.22089681923389434,
      "num_tokens": 80113420.0,
      "step": 38605
    },
    {
      "entropy": 5.634019756317139,
      "epoch": 3.5023584905660377,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003807148946936387,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21413621455430984,
      "num_tokens": 80123398.0,
      "step": 38610
    },
    {
      "entropy": 5.657048320770263,
      "epoch": 3.502812046444122,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003806863309879956,
      "loss": 4.861,
      "mean_token_accuracy": 0.23065851628780365,
      "num_tokens": 80133885.0,
      "step": 38615
    },
    {
      "entropy": 5.625788354873658,
      "epoch": 3.503265602322206,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003806577650967321,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23959063440561296,
      "num_tokens": 80144168.0,
      "step": 38620
    },
    {
      "entropy": 5.592288446426392,
      "epoch": 3.5037191582002905,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003806291970204391,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21960345953702926,
      "num_tokens": 80153956.0,
      "step": 38625
    },
    {
      "entropy": 5.677044010162353,
      "epoch": 3.5041727140783747,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038060062675970743,
      "loss": 4.9372,
      "mean_token_accuracy": 0.20915284007787704,
      "num_tokens": 80163617.0,
      "step": 38630
    },
    {
      "entropy": 5.6456911087036135,
      "epoch": 3.5046262699564585,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003805720543151278,
      "loss": 4.8231,
      "mean_token_accuracy": 0.22625830322504042,
      "num_tokens": 80174049.0,
      "step": 38635
    },
    {
      "entropy": 5.7088581085205075,
      "epoch": 3.505079825834543,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003805434796872912,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22086495161056519,
      "num_tokens": 80185043.0,
      "step": 38640
    },
    {
      "entropy": 5.60680456161499,
      "epoch": 3.505533381712627,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038051490287678856,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22150878459215165,
      "num_tokens": 80194810.0,
      "step": 38645
    },
    {
      "entropy": 5.592621088027954,
      "epoch": 3.5059869375907113,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003804863238842109,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2285303995013237,
      "num_tokens": 80204861.0,
      "step": 38650
    },
    {
      "entropy": 5.63749737739563,
      "epoch": 3.506440493468795,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003804577427101491,
      "loss": 4.8046,
      "mean_token_accuracy": 0.23063274323940278,
      "num_tokens": 80214447.0,
      "step": 38655
    },
    {
      "entropy": 5.617111396789551,
      "epoch": 3.5068940493468794,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003804291593551944,
      "loss": 4.8245,
      "mean_token_accuracy": 0.22542436122894288,
      "num_tokens": 80224213.0,
      "step": 38660
    },
    {
      "entropy": 5.641775226593017,
      "epoch": 3.5073476052249637,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003804005738199377,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21927156448364257,
      "num_tokens": 80234161.0,
      "step": 38665
    },
    {
      "entropy": 5.638116979598999,
      "epoch": 3.507801161103048,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038037198610497037,
      "loss": 4.8778,
      "mean_token_accuracy": 0.2222333073616028,
      "num_tokens": 80243788.0,
      "step": 38670
    },
    {
      "entropy": 5.644446611404419,
      "epoch": 3.5082547169811322,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003803433962108835,
      "loss": 4.8976,
      "mean_token_accuracy": 0.21956930160522461,
      "num_tokens": 80254658.0,
      "step": 38675
    },
    {
      "entropy": 5.686088132858276,
      "epoch": 3.5087082728592165,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003803148041382683,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22225405722856523,
      "num_tokens": 80264820.0,
      "step": 38680
    },
    {
      "entropy": 5.57348403930664,
      "epoch": 3.5091618287373003,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038028620988771617,
      "loss": 4.8162,
      "mean_token_accuracy": 0.23009472638368605,
      "num_tokens": 80274639.0,
      "step": 38685
    },
    {
      "entropy": 5.588067197799683,
      "epoch": 3.5096153846153846,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003802576134598183,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22792890667915344,
      "num_tokens": 80284542.0,
      "step": 38690
    },
    {
      "entropy": 5.5946413516998295,
      "epoch": 3.510068940493469,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00038022901485516625,
      "loss": 4.8221,
      "mean_token_accuracy": 0.2267189532518387,
      "num_tokens": 80294405.0,
      "step": 38695
    },
    {
      "entropy": 5.627056837081909,
      "epoch": 3.510522496371553,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038020041407435133,
      "loss": 4.8573,
      "mean_token_accuracy": 0.21651632934808732,
      "num_tokens": 80305326.0,
      "step": 38700
    },
    {
      "entropy": 5.610388708114624,
      "epoch": 3.510976052249637,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038017181111796506,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2218819737434387,
      "num_tokens": 80315423.0,
      "step": 38705
    },
    {
      "entropy": 5.629005765914917,
      "epoch": 3.511429608127721,
      "grad_norm": 1.25,
      "learning_rate": 0.0003801432059865989,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21443607211112975,
      "num_tokens": 80325372.0,
      "step": 38710
    },
    {
      "entropy": 5.668954610824585,
      "epoch": 3.5118831640058055,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038011459868084446,
      "loss": 4.7988,
      "mean_token_accuracy": 0.22460764348506929,
      "num_tokens": 80335467.0,
      "step": 38715
    },
    {
      "entropy": 5.629906511306762,
      "epoch": 3.5123367198838897,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003800859892012933,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22232987880706787,
      "num_tokens": 80345420.0,
      "step": 38720
    },
    {
      "entropy": 5.599252700805664,
      "epoch": 3.512790275761974,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003800573775485372,
      "loss": 4.8779,
      "mean_token_accuracy": 0.23185952752828598,
      "num_tokens": 80354369.0,
      "step": 38725
    },
    {
      "entropy": 5.680689811706543,
      "epoch": 3.5132438316400583,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038002876372316774,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2172993987798691,
      "num_tokens": 80365292.0,
      "step": 38730
    },
    {
      "entropy": 5.709176874160766,
      "epoch": 3.513697387518142,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038000014772577657,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22354436218738555,
      "num_tokens": 80376340.0,
      "step": 38735
    },
    {
      "entropy": 5.565360164642334,
      "epoch": 3.5141509433962264,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003799715295569557,
      "loss": 4.8663,
      "mean_token_accuracy": 0.2194734290242195,
      "num_tokens": 80386193.0,
      "step": 38740
    },
    {
      "entropy": 5.725466680526734,
      "epoch": 3.5146044992743106,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003799429092172968,
      "loss": 5.0602,
      "mean_token_accuracy": 0.2148408219218254,
      "num_tokens": 80397055.0,
      "step": 38745
    },
    {
      "entropy": 5.669391775131226,
      "epoch": 3.515058055152395,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003799142867073918,
      "loss": 4.9593,
      "mean_token_accuracy": 0.2132392093539238,
      "num_tokens": 80407375.0,
      "step": 38750
    },
    {
      "entropy": 5.705404949188233,
      "epoch": 3.5155116110304787,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037988566202783256,
      "loss": 4.9458,
      "mean_token_accuracy": 0.21610867232084274,
      "num_tokens": 80417942.0,
      "step": 38755
    },
    {
      "entropy": 5.573472309112549,
      "epoch": 3.515965166908563,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003798570351792112,
      "loss": 4.7923,
      "mean_token_accuracy": 0.22393308877944945,
      "num_tokens": 80428716.0,
      "step": 38760
    },
    {
      "entropy": 5.609170150756836,
      "epoch": 3.5164187227866472,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037982840616211964,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21630744487047196,
      "num_tokens": 80439688.0,
      "step": 38765
    },
    {
      "entropy": 5.702987575531006,
      "epoch": 3.5168722786647315,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037979977497714984,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21656095534563063,
      "num_tokens": 80450434.0,
      "step": 38770
    },
    {
      "entropy": 5.6568663120269775,
      "epoch": 3.5173258345428158,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037977114162489404,
      "loss": 4.9633,
      "mean_token_accuracy": 0.215074360370636,
      "num_tokens": 80462187.0,
      "step": 38775
    },
    {
      "entropy": 5.640180349349976,
      "epoch": 3.5177793904209,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003797425061059443,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22144632190465927,
      "num_tokens": 80473049.0,
      "step": 38780
    },
    {
      "entropy": 5.601675081253052,
      "epoch": 3.5182329462989843,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037971386842089287,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2269955411553383,
      "num_tokens": 80484119.0,
      "step": 38785
    },
    {
      "entropy": 5.605462980270386,
      "epoch": 3.518686502177068,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003796852285703319,
      "loss": 4.7845,
      "mean_token_accuracy": 0.23113539516925813,
      "num_tokens": 80494882.0,
      "step": 38790
    },
    {
      "entropy": 5.674931955337525,
      "epoch": 3.5191400580551524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003796565865548538,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22131285667419434,
      "num_tokens": 80505362.0,
      "step": 38795
    },
    {
      "entropy": 5.6500050067901615,
      "epoch": 3.5195936139332367,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003796279423750508,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21475103944540025,
      "num_tokens": 80516027.0,
      "step": 38800
    },
    {
      "entropy": 5.69084734916687,
      "epoch": 3.5200471698113205,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003795992960315153,
      "loss": 4.9457,
      "mean_token_accuracy": 0.21867989450693132,
      "num_tokens": 80527045.0,
      "step": 38805
    },
    {
      "entropy": 5.63895173072815,
      "epoch": 3.5205007256894048,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037957064752483966,
      "loss": 4.8184,
      "mean_token_accuracy": 0.22899862825870515,
      "num_tokens": 80537373.0,
      "step": 38810
    },
    {
      "entropy": 5.601835632324219,
      "epoch": 3.520954281567489,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003795419968556164,
      "loss": 4.8911,
      "mean_token_accuracy": 0.21184316128492356,
      "num_tokens": 80546632.0,
      "step": 38815
    },
    {
      "entropy": 5.6124248027801515,
      "epoch": 3.5214078374455733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003795133440244381,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22231239527463914,
      "num_tokens": 80556694.0,
      "step": 38820
    },
    {
      "entropy": 5.683526468276978,
      "epoch": 3.5218613933236576,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003794846890318971,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2106945186853409,
      "num_tokens": 80567834.0,
      "step": 38825
    },
    {
      "entropy": 5.659314966201782,
      "epoch": 3.522314949201742,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003794560318785861,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22470719814300538,
      "num_tokens": 80578299.0,
      "step": 38830
    },
    {
      "entropy": 5.6747547626495365,
      "epoch": 3.522768505079826,
      "grad_norm": 1.25,
      "learning_rate": 0.00037942737256509786,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21745397597551347,
      "num_tokens": 80588620.0,
      "step": 38835
    },
    {
      "entropy": 5.798318910598755,
      "epoch": 3.52322206095791,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003793987110920248,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20557896494865419,
      "num_tokens": 80599076.0,
      "step": 38840
    },
    {
      "entropy": 5.731644535064698,
      "epoch": 3.523675616835994,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037937004745995987,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2126277893781662,
      "num_tokens": 80609440.0,
      "step": 38845
    },
    {
      "entropy": 5.76024899482727,
      "epoch": 3.5241291727140784,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003793413816694957,
      "loss": 4.983,
      "mean_token_accuracy": 0.21231929510831832,
      "num_tokens": 80619169.0,
      "step": 38850
    },
    {
      "entropy": 5.6157317638397215,
      "epoch": 3.5245827285921627,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003793127137212252,
      "loss": 4.8537,
      "mean_token_accuracy": 0.224270823597908,
      "num_tokens": 80628719.0,
      "step": 38855
    },
    {
      "entropy": 5.658299541473388,
      "epoch": 3.5250362844702465,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037928404361574125,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21427015513181685,
      "num_tokens": 80638391.0,
      "step": 38860
    },
    {
      "entropy": 5.655683851242065,
      "epoch": 3.525489840348331,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003792553713536367,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21779898703098297,
      "num_tokens": 80647720.0,
      "step": 38865
    },
    {
      "entropy": 5.5854710102081295,
      "epoch": 3.525943396226415,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003792266969355044,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22293582260608674,
      "num_tokens": 80658841.0,
      "step": 38870
    },
    {
      "entropy": 5.651483535766602,
      "epoch": 3.5263969521044993,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037919802036193753,
      "loss": 4.9098,
      "mean_token_accuracy": 0.221801295876503,
      "num_tokens": 80668735.0,
      "step": 38875
    },
    {
      "entropy": 5.685948181152344,
      "epoch": 3.5268505079825836,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037916934163352903,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22489866465330124,
      "num_tokens": 80680757.0,
      "step": 38880
    },
    {
      "entropy": 5.62222261428833,
      "epoch": 3.527304063860668,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000379140660750872,
      "loss": 4.8489,
      "mean_token_accuracy": 0.2200845703482628,
      "num_tokens": 80689833.0,
      "step": 38885
    },
    {
      "entropy": 5.570931053161621,
      "epoch": 3.5277576197387517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003791119777145595,
      "loss": 4.8311,
      "mean_token_accuracy": 0.22644873857498168,
      "num_tokens": 80700347.0,
      "step": 38890
    },
    {
      "entropy": 5.530778455734253,
      "epoch": 3.528211175616836,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003790832925251849,
      "loss": 4.7328,
      "mean_token_accuracy": 0.2385459452867508,
      "num_tokens": 80710559.0,
      "step": 38895
    },
    {
      "entropy": 5.658862543106079,
      "epoch": 3.52866473149492,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003790546051833412,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21817902177572251,
      "num_tokens": 80720805.0,
      "step": 38900
    },
    {
      "entropy": 5.662647724151611,
      "epoch": 3.5291182873730045,
      "grad_norm": 1.25,
      "learning_rate": 0.0003790259156896217,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2043817952275276,
      "num_tokens": 80730937.0,
      "step": 38905
    },
    {
      "entropy": 5.669439125061035,
      "epoch": 3.5295718432510883,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037899722404461983,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22262938022613527,
      "num_tokens": 80741597.0,
      "step": 38910
    },
    {
      "entropy": 5.6773701190948485,
      "epoch": 3.5300253991291726,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003789685302489289,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2141635835170746,
      "num_tokens": 80752496.0,
      "step": 38915
    },
    {
      "entropy": 5.675675678253174,
      "epoch": 3.530478955007257,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037893983430314226,
      "loss": 4.8775,
      "mean_token_accuracy": 0.21867239624261856,
      "num_tokens": 80762329.0,
      "step": 38920
    },
    {
      "entropy": 5.618783092498779,
      "epoch": 3.530932510885341,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037891113620785327,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21754271388053895,
      "num_tokens": 80773152.0,
      "step": 38925
    },
    {
      "entropy": 5.6123795986175535,
      "epoch": 3.5313860667634254,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037888243596365557,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22067504674196242,
      "num_tokens": 80783502.0,
      "step": 38930
    },
    {
      "entropy": 5.67082371711731,
      "epoch": 3.5318396226415096,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037885373357114265,
      "loss": 4.8219,
      "mean_token_accuracy": 0.22149303406476975,
      "num_tokens": 80792644.0,
      "step": 38935
    },
    {
      "entropy": 5.670018529891967,
      "epoch": 3.5322931785195935,
      "grad_norm": 1.203125,
      "learning_rate": 0.000378825029030908,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2167245164513588,
      "num_tokens": 80803184.0,
      "step": 38940
    },
    {
      "entropy": 5.623274326324463,
      "epoch": 3.5327467343976777,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003787963223435453,
      "loss": 4.8951,
      "mean_token_accuracy": 0.22082321494817733,
      "num_tokens": 80813353.0,
      "step": 38945
    },
    {
      "entropy": 5.654191780090332,
      "epoch": 3.533200290275762,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003787676135096482,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21453527063131334,
      "num_tokens": 80824663.0,
      "step": 38950
    },
    {
      "entropy": 5.766034555435181,
      "epoch": 3.5336538461538463,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003787389025298104,
      "loss": 5.0151,
      "mean_token_accuracy": 0.21007631421089173,
      "num_tokens": 80835507.0,
      "step": 38955
    },
    {
      "entropy": 5.666920852661133,
      "epoch": 3.53410740203193,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003787101894046257,
      "loss": 4.8238,
      "mean_token_accuracy": 0.2341285303235054,
      "num_tokens": 80845035.0,
      "step": 38960
    },
    {
      "entropy": 5.6574183940887455,
      "epoch": 3.5345609579100143,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037868147413468785,
      "loss": 4.8614,
      "mean_token_accuracy": 0.2206186130642891,
      "num_tokens": 80854992.0,
      "step": 38965
    },
    {
      "entropy": 5.663049173355103,
      "epoch": 3.5350145137880986,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003786527567205907,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21088607609272003,
      "num_tokens": 80864658.0,
      "step": 38970
    },
    {
      "entropy": 5.687146949768066,
      "epoch": 3.535468069666183,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037862403716292805,
      "loss": 4.8816,
      "mean_token_accuracy": 0.2201855942606926,
      "num_tokens": 80873964.0,
      "step": 38975
    },
    {
      "entropy": 5.673787450790405,
      "epoch": 3.535921625544267,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037859531546229405,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2163935199379921,
      "num_tokens": 80883752.0,
      "step": 38980
    },
    {
      "entropy": 5.684512710571289,
      "epoch": 3.5363751814223514,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003785665916192824,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21969943046569823,
      "num_tokens": 80894265.0,
      "step": 38985
    },
    {
      "entropy": 5.585261106491089,
      "epoch": 3.5368287373004357,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037853786563448724,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2253129854798317,
      "num_tokens": 80904246.0,
      "step": 38990
    },
    {
      "entropy": 5.6701438426971436,
      "epoch": 3.5372822931785195,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003785091375085027,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21751925051212312,
      "num_tokens": 80913595.0,
      "step": 38995
    },
    {
      "entropy": 5.628238582611084,
      "epoch": 3.5377358490566038,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003784804072419228,
      "loss": 4.898,
      "mean_token_accuracy": 0.21578398197889329,
      "num_tokens": 80924754.0,
      "step": 39000
    },
    {
      "epoch": 3.5377358490566038,
      "eval_entropy": 5.393534546632033,
      "eval_loss": 4.970121383666992,
      "eval_mean_token_accuracy": 0.2253137246330534,
      "eval_num_tokens": 80924754.0,
      "eval_runtime": 20.6395,
      "eval_samples_per_second": 1713.22,
      "eval_steps_per_second": 214.153,
      "step": 39000
    },
    {
      "entropy": 5.616163158416748,
      "epoch": 3.538189404934688,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003784516748353417,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23074238002300262,
      "num_tokens": 80935318.0,
      "step": 39005
    },
    {
      "entropy": 5.646154832839966,
      "epoch": 3.5386429608127723,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003784229402893536,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22214530706405639,
      "num_tokens": 80944886.0,
      "step": 39010
    },
    {
      "entropy": 5.6769355773925785,
      "epoch": 3.539096516690856,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003783942036045528,
      "loss": 4.9167,
      "mean_token_accuracy": 0.21518024057149887,
      "num_tokens": 80955566.0,
      "step": 39015
    },
    {
      "entropy": 5.695873641967774,
      "epoch": 3.5395500725689404,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037836546478153344,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21601938009262084,
      "num_tokens": 80965285.0,
      "step": 39020
    },
    {
      "entropy": 5.591855764389038,
      "epoch": 3.5400036284470247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003783367238208899,
      "loss": 4.8684,
      "mean_token_accuracy": 0.2274763986468315,
      "num_tokens": 80976402.0,
      "step": 39025
    },
    {
      "entropy": 5.690197992324829,
      "epoch": 3.540457184325109,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037830798072321665,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2182393953204155,
      "num_tokens": 80988408.0,
      "step": 39030
    },
    {
      "entropy": 5.676499557495117,
      "epoch": 3.540910740203193,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037827923548910806,
      "loss": 4.8945,
      "mean_token_accuracy": 0.21863435804843903,
      "num_tokens": 80999099.0,
      "step": 39035
    },
    {
      "entropy": 5.608412170410157,
      "epoch": 3.5413642960812775,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037825048811915854,
      "loss": 4.8043,
      "mean_token_accuracy": 0.22520106583833693,
      "num_tokens": 81009350.0,
      "step": 39040
    },
    {
      "entropy": 5.629689168930054,
      "epoch": 3.5418178519593613,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037822173861396256,
      "loss": 4.8056,
      "mean_token_accuracy": 0.2192219853401184,
      "num_tokens": 81019851.0,
      "step": 39045
    },
    {
      "entropy": 5.646512365341186,
      "epoch": 3.5422714078374455,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037819298697411475,
      "loss": 4.9358,
      "mean_token_accuracy": 0.21510945856571198,
      "num_tokens": 81028987.0,
      "step": 39050
    },
    {
      "entropy": 5.6185201644897464,
      "epoch": 3.54272496371553,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037816423320020964,
      "loss": 4.8399,
      "mean_token_accuracy": 0.2241588994860649,
      "num_tokens": 81038081.0,
      "step": 39055
    },
    {
      "entropy": 5.67468957901001,
      "epoch": 3.543178519593614,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037813547729284195,
      "loss": 4.8676,
      "mean_token_accuracy": 0.21889322996139526,
      "num_tokens": 81048066.0,
      "step": 39060
    },
    {
      "entropy": 5.519023180007935,
      "epoch": 3.543632075471698,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037810671925260633,
      "loss": 4.7136,
      "mean_token_accuracy": 0.24082041233778,
      "num_tokens": 81057421.0,
      "step": 39065
    },
    {
      "entropy": 5.582855892181397,
      "epoch": 3.544085631349782,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037807795908009745,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22361056208610536,
      "num_tokens": 81068435.0,
      "step": 39070
    },
    {
      "entropy": 5.671843242645264,
      "epoch": 3.5445391872278664,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037804919677591006,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22070489674806595,
      "num_tokens": 81079378.0,
      "step": 39075
    },
    {
      "entropy": 5.615456247329712,
      "epoch": 3.5449927431059507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003780204323406391,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21721493005752562,
      "num_tokens": 81089872.0,
      "step": 39080
    },
    {
      "entropy": 5.607994937896729,
      "epoch": 3.545446298984035,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003779916657748793,
      "loss": 4.8182,
      "mean_token_accuracy": 0.2268304780125618,
      "num_tokens": 81099653.0,
      "step": 39085
    },
    {
      "entropy": 5.625764989852906,
      "epoch": 3.5458998548621192,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003779628970792256,
      "loss": 4.8965,
      "mean_token_accuracy": 0.2192773386836052,
      "num_tokens": 81109840.0,
      "step": 39090
    },
    {
      "entropy": 5.593977785110473,
      "epoch": 3.546353410740203,
      "grad_norm": 1.296875,
      "learning_rate": 0.000377934126254273,
      "loss": 4.8158,
      "mean_token_accuracy": 0.23031169921159744,
      "num_tokens": 81120991.0,
      "step": 39095
    },
    {
      "entropy": 5.686555814743042,
      "epoch": 3.5468069666182873,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037790535330061635,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22146696597337723,
      "num_tokens": 81129831.0,
      "step": 39100
    },
    {
      "entropy": 5.645819234848022,
      "epoch": 3.5472605224963716,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003778765782188508,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22617004215717315,
      "num_tokens": 81140604.0,
      "step": 39105
    },
    {
      "entropy": 5.654625272750854,
      "epoch": 3.547714078374456,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037784780100957145,
      "loss": 4.9016,
      "mean_token_accuracy": 0.2243119955062866,
      "num_tokens": 81150591.0,
      "step": 39110
    },
    {
      "entropy": 5.722482585906983,
      "epoch": 3.5481676342525397,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003778190216733733,
      "loss": 4.962,
      "mean_token_accuracy": 0.2143990233540535,
      "num_tokens": 81161444.0,
      "step": 39115
    },
    {
      "entropy": 5.667954444885254,
      "epoch": 3.548621190130624,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037779024021085167,
      "loss": 4.8311,
      "mean_token_accuracy": 0.22865815758705138,
      "num_tokens": 81171216.0,
      "step": 39120
    },
    {
      "entropy": 5.659074115753174,
      "epoch": 3.549074746008708,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003777614566226016,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2257553979754448,
      "num_tokens": 81181665.0,
      "step": 39125
    },
    {
      "entropy": 5.543111944198609,
      "epoch": 3.5495283018867925,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003777326709092184,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22232834100723267,
      "num_tokens": 81191683.0,
      "step": 39130
    },
    {
      "entropy": 5.6455193042755125,
      "epoch": 3.5499818577648767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037770388307129747,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22266591638326644,
      "num_tokens": 81201211.0,
      "step": 39135
    },
    {
      "entropy": 5.657996082305909,
      "epoch": 3.550435413642961,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037767509310943407,
      "loss": 4.8664,
      "mean_token_accuracy": 0.2244148924946785,
      "num_tokens": 81210668.0,
      "step": 39140
    },
    {
      "entropy": 5.608302307128906,
      "epoch": 3.5508889695210453,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003776463010242235,
      "loss": 4.7573,
      "mean_token_accuracy": 0.22493915557861327,
      "num_tokens": 81221073.0,
      "step": 39145
    },
    {
      "entropy": 5.624004936218261,
      "epoch": 3.551342525399129,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037761750681626134,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21426354199647904,
      "num_tokens": 81231687.0,
      "step": 39150
    },
    {
      "entropy": 5.6993824481964115,
      "epoch": 3.5517960812772134,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037758871048614295,
      "loss": 4.9392,
      "mean_token_accuracy": 0.213796067237854,
      "num_tokens": 81242323.0,
      "step": 39155
    },
    {
      "entropy": 5.620159149169922,
      "epoch": 3.5522496371552976,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003775599120344638,
      "loss": 4.7702,
      "mean_token_accuracy": 0.23432543128728867,
      "num_tokens": 81252543.0,
      "step": 39160
    },
    {
      "entropy": 5.641052341461181,
      "epoch": 3.5527031930333814,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037753111146181966,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2142824798822403,
      "num_tokens": 81262180.0,
      "step": 39165
    },
    {
      "entropy": 5.758733129501342,
      "epoch": 3.5531567489114657,
      "grad_norm": 1.296875,
      "learning_rate": 0.000377502308768806,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20128827691078185,
      "num_tokens": 81272088.0,
      "step": 39170
    },
    {
      "entropy": 5.653423547744751,
      "epoch": 3.55361030478955,
      "grad_norm": 1.25,
      "learning_rate": 0.0003774735039560184,
      "loss": 4.8662,
      "mean_token_accuracy": 0.2229309692978859,
      "num_tokens": 81281723.0,
      "step": 39175
    },
    {
      "entropy": 5.686999320983887,
      "epoch": 3.5540638606676342,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037744469702405273,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22228230834007262,
      "num_tokens": 81292030.0,
      "step": 39180
    },
    {
      "entropy": 5.600500440597534,
      "epoch": 3.5545174165457185,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003774158879735045,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22498147338628768,
      "num_tokens": 81302283.0,
      "step": 39185
    },
    {
      "entropy": 5.672839069366455,
      "epoch": 3.5549709724238028,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003773870768049697,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21361616998910904,
      "num_tokens": 81313058.0,
      "step": 39190
    },
    {
      "entropy": 5.657902336120605,
      "epoch": 3.555424528301887,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000377358263519044,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22400048673152922,
      "num_tokens": 81323827.0,
      "step": 39195
    },
    {
      "entropy": 5.687113046646118,
      "epoch": 3.555878084179971,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003773294481163234,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21485089361667634,
      "num_tokens": 81333106.0,
      "step": 39200
    },
    {
      "entropy": 5.612260866165161,
      "epoch": 3.556331640058055,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003773006305974037,
      "loss": 4.873,
      "mean_token_accuracy": 0.22215779721736909,
      "num_tokens": 81343245.0,
      "step": 39205
    },
    {
      "entropy": 5.617501878738404,
      "epoch": 3.5567851959361394,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037727181096288077,
      "loss": 4.8802,
      "mean_token_accuracy": 0.2294653609395027,
      "num_tokens": 81352917.0,
      "step": 39210
    },
    {
      "entropy": 5.574107694625854,
      "epoch": 3.5572387518142237,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003772429892133509,
      "loss": 4.8353,
      "mean_token_accuracy": 0.21991948038339615,
      "num_tokens": 81362838.0,
      "step": 39215
    },
    {
      "entropy": 5.614162254333496,
      "epoch": 3.5576923076923075,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003772141653494099,
      "loss": 4.8148,
      "mean_token_accuracy": 0.23064998388290406,
      "num_tokens": 81373395.0,
      "step": 39220
    },
    {
      "entropy": 5.610518884658814,
      "epoch": 3.5581458635703918,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003771853393716538,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2208454817533493,
      "num_tokens": 81383976.0,
      "step": 39225
    },
    {
      "entropy": 5.629348039627075,
      "epoch": 3.558599419448476,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000377156511280679,
      "loss": 4.8791,
      "mean_token_accuracy": 0.21517060101032257,
      "num_tokens": 81394328.0,
      "step": 39230
    },
    {
      "entropy": 5.668675231933594,
      "epoch": 3.5590529753265603,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003771276810770815,
      "loss": 4.8948,
      "mean_token_accuracy": 0.2194511726498604,
      "num_tokens": 81405451.0,
      "step": 39235
    },
    {
      "entropy": 5.673462009429931,
      "epoch": 3.5595065312046446,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003770988487614575,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20736407488584518,
      "num_tokens": 81416589.0,
      "step": 39240
    },
    {
      "entropy": 5.561825180053711,
      "epoch": 3.559960087082729,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003770700143344033,
      "loss": 4.8358,
      "mean_token_accuracy": 0.22503251880407332,
      "num_tokens": 81428553.0,
      "step": 39245
    },
    {
      "entropy": 5.771281576156616,
      "epoch": 3.5604136429608126,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037704117779651514,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21511709988117217,
      "num_tokens": 81438720.0,
      "step": 39250
    },
    {
      "entropy": 5.612032556533814,
      "epoch": 3.560867198838897,
      "grad_norm": 1.484375,
      "learning_rate": 0.00037701233914838947,
      "loss": 4.9327,
      "mean_token_accuracy": 0.21700548082590104,
      "num_tokens": 81448142.0,
      "step": 39255
    },
    {
      "entropy": 5.654911708831787,
      "epoch": 3.561320754716981,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003769834983906226,
      "loss": 4.9557,
      "mean_token_accuracy": 0.218122760951519,
      "num_tokens": 81459448.0,
      "step": 39260
    },
    {
      "entropy": 5.653246259689331,
      "epoch": 3.5617743105950654,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037695465552381105,
      "loss": 4.8734,
      "mean_token_accuracy": 0.2210347518324852,
      "num_tokens": 81470356.0,
      "step": 39265
    },
    {
      "entropy": 5.641185569763183,
      "epoch": 3.5622278664731493,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037692581054855125,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2232051119208336,
      "num_tokens": 81480472.0,
      "step": 39270
    },
    {
      "entropy": 5.608519697189331,
      "epoch": 3.5626814223512335,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037689696346543964,
      "loss": 4.836,
      "mean_token_accuracy": 0.22760785669088363,
      "num_tokens": 81489949.0,
      "step": 39275
    },
    {
      "entropy": 5.577717256546021,
      "epoch": 3.563134978229318,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037686811427507284,
      "loss": 4.7935,
      "mean_token_accuracy": 0.22302671521902084,
      "num_tokens": 81500276.0,
      "step": 39280
    },
    {
      "entropy": 5.613933420181274,
      "epoch": 3.563588534107402,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037683926297804756,
      "loss": 4.793,
      "mean_token_accuracy": 0.2225412830710411,
      "num_tokens": 81510431.0,
      "step": 39285
    },
    {
      "entropy": 5.603700447082519,
      "epoch": 3.5640420899854863,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003768104095749603,
      "loss": 4.8877,
      "mean_token_accuracy": 0.21796686202287674,
      "num_tokens": 81520718.0,
      "step": 39290
    },
    {
      "entropy": 5.636259508132935,
      "epoch": 3.5644956458635706,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003767815540664079,
      "loss": 4.9678,
      "mean_token_accuracy": 0.2167618215084076,
      "num_tokens": 81531904.0,
      "step": 39295
    },
    {
      "entropy": 5.639812469482422,
      "epoch": 3.5649492017416544,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037675269645298697,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21602933704853058,
      "num_tokens": 81542093.0,
      "step": 39300
    },
    {
      "entropy": 5.633946657180786,
      "epoch": 3.5654027576197387,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003767238367352944,
      "loss": 4.823,
      "mean_token_accuracy": 0.22708417177200318,
      "num_tokens": 81552020.0,
      "step": 39305
    },
    {
      "entropy": 5.569300031661987,
      "epoch": 3.565856313497823,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037669497491392683,
      "loss": 4.7981,
      "mean_token_accuracy": 0.22968486398458482,
      "num_tokens": 81561644.0,
      "step": 39310
    },
    {
      "entropy": 5.619177913665771,
      "epoch": 3.566309869375907,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037666611098948144,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22486685812473298,
      "num_tokens": 81572158.0,
      "step": 39315
    },
    {
      "entropy": 5.718697118759155,
      "epoch": 3.566763425253991,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003766372449625549,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21216964423656465,
      "num_tokens": 81582335.0,
      "step": 39320
    },
    {
      "entropy": 5.743814134597779,
      "epoch": 3.5672169811320753,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003766083768337442,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2155804678797722,
      "num_tokens": 81593004.0,
      "step": 39325
    },
    {
      "entropy": 5.66798586845398,
      "epoch": 3.5676705370101596,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003765795066036464,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2164766550064087,
      "num_tokens": 81603656.0,
      "step": 39330
    },
    {
      "entropy": 5.622368669509887,
      "epoch": 3.568124092888244,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037655063427285854,
      "loss": 4.917,
      "mean_token_accuracy": 0.21528741270303725,
      "num_tokens": 81613039.0,
      "step": 39335
    },
    {
      "entropy": 5.5999902248382565,
      "epoch": 3.568577648766328,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037652175984197764,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22598878890275956,
      "num_tokens": 81624410.0,
      "step": 39340
    },
    {
      "entropy": 5.6626465797424315,
      "epoch": 3.5690312046444124,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037649288331160086,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22347572296857834,
      "num_tokens": 81635160.0,
      "step": 39345
    },
    {
      "entropy": 5.633468198776245,
      "epoch": 3.5694847605224966,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037646400468232546,
      "loss": 4.8654,
      "mean_token_accuracy": 0.21862493455410004,
      "num_tokens": 81645686.0,
      "step": 39350
    },
    {
      "entropy": 5.613604497909546,
      "epoch": 3.5699383164005805,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037643512395474853,
      "loss": 4.8254,
      "mean_token_accuracy": 0.22539474368095397,
      "num_tokens": 81656330.0,
      "step": 39355
    },
    {
      "entropy": 5.603452110290528,
      "epoch": 3.5703918722786647,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003764062411294674,
      "loss": 4.8814,
      "mean_token_accuracy": 0.2166259169578552,
      "num_tokens": 81665895.0,
      "step": 39360
    },
    {
      "entropy": 5.710248613357544,
      "epoch": 3.570845428156749,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037637735620707935,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2193511113524437,
      "num_tokens": 81676616.0,
      "step": 39365
    },
    {
      "entropy": 5.651619720458984,
      "epoch": 3.5712989840348333,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003763484691881818,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2194437175989151,
      "num_tokens": 81687169.0,
      "step": 39370
    },
    {
      "entropy": 5.664397573471069,
      "epoch": 3.571752539912917,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037631958007337195,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22131420522928238,
      "num_tokens": 81698291.0,
      "step": 39375
    },
    {
      "entropy": 5.685623979568481,
      "epoch": 3.5722060957910013,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003762906888632475,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21665120124816895,
      "num_tokens": 81708750.0,
      "step": 39380
    },
    {
      "entropy": 5.585685205459595,
      "epoch": 3.5726596516690856,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003762617955584058,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22218581289052963,
      "num_tokens": 81719474.0,
      "step": 39385
    },
    {
      "entropy": 5.735635280609131,
      "epoch": 3.57311320754717,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003762329001594443,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21766179352998732,
      "num_tokens": 81730092.0,
      "step": 39390
    },
    {
      "entropy": 5.6638270854949955,
      "epoch": 3.573566763425254,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003762040026669606,
      "loss": 4.8913,
      "mean_token_accuracy": 0.21569544225931167,
      "num_tokens": 81741235.0,
      "step": 39395
    },
    {
      "entropy": 5.594498157501221,
      "epoch": 3.5740203193033384,
      "grad_norm": 1.25,
      "learning_rate": 0.0003761751030815524,
      "loss": 4.8583,
      "mean_token_accuracy": 0.22641670405864717,
      "num_tokens": 81751434.0,
      "step": 39400
    },
    {
      "entropy": 5.646251058578491,
      "epoch": 3.5744738751814222,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003761462014038173,
      "loss": 4.9232,
      "mean_token_accuracy": 0.21071959733963014,
      "num_tokens": 81763935.0,
      "step": 39405
    },
    {
      "entropy": 5.635839939117432,
      "epoch": 3.5749274310595065,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003761172976343528,
      "loss": 4.8711,
      "mean_token_accuracy": 0.21976048350334168,
      "num_tokens": 81775007.0,
      "step": 39410
    },
    {
      "entropy": 5.639729356765747,
      "epoch": 3.5753809869375908,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037608839177375705,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21952951699495316,
      "num_tokens": 81785690.0,
      "step": 39415
    },
    {
      "entropy": 5.6427809715271,
      "epoch": 3.575834542815675,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003760594838226275,
      "loss": 4.8915,
      "mean_token_accuracy": 0.2263173833489418,
      "num_tokens": 81796143.0,
      "step": 39420
    },
    {
      "entropy": 5.7831244468688965,
      "epoch": 3.576288098693759,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003760305737815621,
      "loss": 4.9887,
      "mean_token_accuracy": 0.2179051697254181,
      "num_tokens": 81806607.0,
      "step": 39425
    },
    {
      "entropy": 5.615176486968994,
      "epoch": 3.576741654571843,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003760016616511586,
      "loss": 4.8235,
      "mean_token_accuracy": 0.2272084042429924,
      "num_tokens": 81817216.0,
      "step": 39430
    },
    {
      "entropy": 5.616209554672241,
      "epoch": 3.5771952104499274,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037597274743201507,
      "loss": 4.899,
      "mean_token_accuracy": 0.2211093619465828,
      "num_tokens": 81827211.0,
      "step": 39435
    },
    {
      "entropy": 5.635285043716431,
      "epoch": 3.5776487663280117,
      "grad_norm": 1.25,
      "learning_rate": 0.0003759438311247294,
      "loss": 4.8416,
      "mean_token_accuracy": 0.22919540703296662,
      "num_tokens": 81837365.0,
      "step": 39440
    },
    {
      "entropy": 5.632224273681641,
      "epoch": 3.578102322206096,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003759149127298995,
      "loss": 4.842,
      "mean_token_accuracy": 0.22837208211421967,
      "num_tokens": 81847803.0,
      "step": 39445
    },
    {
      "entropy": 5.606670904159546,
      "epoch": 3.57855587808418,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037588599224812363,
      "loss": 4.8328,
      "mean_token_accuracy": 0.22794439643621445,
      "num_tokens": 81858123.0,
      "step": 39450
    },
    {
      "entropy": 5.581077671051025,
      "epoch": 3.579009433962264,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003758570696799997,
      "loss": 4.8392,
      "mean_token_accuracy": 0.23114202916622162,
      "num_tokens": 81868668.0,
      "step": 39455
    },
    {
      "entropy": 5.608233213424683,
      "epoch": 3.5794629898403483,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037582814502612575,
      "loss": 4.8166,
      "mean_token_accuracy": 0.23261297196149827,
      "num_tokens": 81877482.0,
      "step": 39460
    },
    {
      "entropy": 5.554647159576416,
      "epoch": 3.5799165457184325,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003757992182871001,
      "loss": 4.7392,
      "mean_token_accuracy": 0.23605061024427415,
      "num_tokens": 81886700.0,
      "step": 39465
    },
    {
      "entropy": 5.631014251708985,
      "epoch": 3.580370101596517,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037577028946352107,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22622024714946748,
      "num_tokens": 81896446.0,
      "step": 39470
    },
    {
      "entropy": 5.662693023681641,
      "epoch": 3.5808236574746006,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003757413585559866,
      "loss": 4.887,
      "mean_token_accuracy": 0.23129415065050124,
      "num_tokens": 81907418.0,
      "step": 39475
    },
    {
      "entropy": 5.674208879470825,
      "epoch": 3.581277213352685,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037571242556509526,
      "loss": 4.8532,
      "mean_token_accuracy": 0.225468248128891,
      "num_tokens": 81917940.0,
      "step": 39480
    },
    {
      "entropy": 5.57749605178833,
      "epoch": 3.581730769230769,
      "grad_norm": 1.25,
      "learning_rate": 0.0003756834904914453,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22322502583265305,
      "num_tokens": 81928158.0,
      "step": 39485
    },
    {
      "entropy": 5.674011611938477,
      "epoch": 3.5821843251088534,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003756545533356351,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21963465213775635,
      "num_tokens": 81938111.0,
      "step": 39490
    },
    {
      "entropy": 5.67154631614685,
      "epoch": 3.5826378809869377,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000375625614098263,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2254747986793518,
      "num_tokens": 81948798.0,
      "step": 39495
    },
    {
      "entropy": 5.686433744430542,
      "epoch": 3.583091436865022,
      "grad_norm": 1.25,
      "learning_rate": 0.00037559667277992765,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22242577821016313,
      "num_tokens": 81959240.0,
      "step": 39500
    },
    {
      "entropy": 5.6343300342559814,
      "epoch": 3.5835449927431062,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003755677293812274,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22503917515277863,
      "num_tokens": 81969929.0,
      "step": 39505
    },
    {
      "entropy": 5.698253870010376,
      "epoch": 3.58399854862119,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003755387839027609,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2120912715792656,
      "num_tokens": 81981007.0,
      "step": 39510
    },
    {
      "entropy": 5.707571220397949,
      "epoch": 3.5844521044992743,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037550983634512683,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21395166218280792,
      "num_tokens": 81991172.0,
      "step": 39515
    },
    {
      "entropy": 5.613654708862304,
      "epoch": 3.5849056603773586,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003754808867089236,
      "loss": 4.7976,
      "mean_token_accuracy": 0.22859591394662857,
      "num_tokens": 82001282.0,
      "step": 39520
    },
    {
      "entropy": 5.630348110198975,
      "epoch": 3.5853592162554424,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037545193499475003,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22027432918548584,
      "num_tokens": 82011929.0,
      "step": 39525
    },
    {
      "entropy": 5.719523286819458,
      "epoch": 3.5858127721335267,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003754229812032049,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2286134496331215,
      "num_tokens": 82021663.0,
      "step": 39530
    },
    {
      "entropy": 5.676588821411133,
      "epoch": 3.586266328011611,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037539402533488694,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21083466708660126,
      "num_tokens": 82031547.0,
      "step": 39535
    },
    {
      "entropy": 5.696749544143676,
      "epoch": 3.586719883889695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003753650673903949,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22018466740846634,
      "num_tokens": 82041722.0,
      "step": 39540
    },
    {
      "entropy": 5.668599414825439,
      "epoch": 3.5871734397677795,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037533610737032773,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22211923450231552,
      "num_tokens": 82052715.0,
      "step": 39545
    },
    {
      "entropy": 5.65114541053772,
      "epoch": 3.5876269956458637,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037530714527528433,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22731748223304749,
      "num_tokens": 82062709.0,
      "step": 39550
    },
    {
      "entropy": 5.684697771072388,
      "epoch": 3.588080551523948,
      "grad_norm": 1.25,
      "learning_rate": 0.00037527818110586354,
      "loss": 4.911,
      "mean_token_accuracy": 0.2186839610338211,
      "num_tokens": 82072731.0,
      "step": 39555
    },
    {
      "entropy": 5.662290382385254,
      "epoch": 3.588534107402032,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003752492148626644,
      "loss": 4.8167,
      "mean_token_accuracy": 0.2284128338098526,
      "num_tokens": 82083216.0,
      "step": 39560
    },
    {
      "entropy": 5.676496124267578,
      "epoch": 3.588987663280116,
      "grad_norm": 1.171875,
      "learning_rate": 0.000375220246546286,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2155966728925705,
      "num_tokens": 82094482.0,
      "step": 39565
    },
    {
      "entropy": 5.659044170379639,
      "epoch": 3.5894412191582004,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003751912761573273,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2198419004678726,
      "num_tokens": 82104988.0,
      "step": 39570
    },
    {
      "entropy": 5.6291591167449955,
      "epoch": 3.5898947750362846,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037516230369638755,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21424392610788345,
      "num_tokens": 82114455.0,
      "step": 39575
    },
    {
      "entropy": 5.654451990127564,
      "epoch": 3.5903483309143684,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003751333291640658,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22112831622362136,
      "num_tokens": 82125001.0,
      "step": 39580
    },
    {
      "entropy": 5.662765979766846,
      "epoch": 3.5908018867924527,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037510435256096124,
      "loss": 4.8708,
      "mean_token_accuracy": 0.2205198049545288,
      "num_tokens": 82134586.0,
      "step": 39585
    },
    {
      "entropy": 5.602188396453857,
      "epoch": 3.591255442670537,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037507537388767326,
      "loss": 4.847,
      "mean_token_accuracy": 0.22516326755285263,
      "num_tokens": 82144612.0,
      "step": 39590
    },
    {
      "entropy": 5.585591554641724,
      "epoch": 3.5917089985486212,
      "grad_norm": 1.265625,
      "learning_rate": 0.000375046393144801,
      "loss": 4.8544,
      "mean_token_accuracy": 0.22928515672683716,
      "num_tokens": 82154913.0,
      "step": 39595
    },
    {
      "entropy": 5.655853939056397,
      "epoch": 3.5921625544267055,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003750174103329438,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21962901204824448,
      "num_tokens": 82164469.0,
      "step": 39600
    },
    {
      "entropy": 5.683607482910157,
      "epoch": 3.59261611030479,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037498842545270105,
      "loss": 4.9139,
      "mean_token_accuracy": 0.2225888878107071,
      "num_tokens": 82174575.0,
      "step": 39605
    },
    {
      "entropy": 5.692196655273437,
      "epoch": 3.5930696661828736,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037495943850467226,
      "loss": 4.893,
      "mean_token_accuracy": 0.2226075068116188,
      "num_tokens": 82184661.0,
      "step": 39610
    },
    {
      "entropy": 5.712615299224853,
      "epoch": 3.593523222060958,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037493044948945677,
      "loss": 4.91,
      "mean_token_accuracy": 0.21913286596536635,
      "num_tokens": 82194877.0,
      "step": 39615
    },
    {
      "entropy": 5.643481254577637,
      "epoch": 3.593976777939042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037490145840765414,
      "loss": 4.8929,
      "mean_token_accuracy": 0.2230779528617859,
      "num_tokens": 82206278.0,
      "step": 39620
    },
    {
      "entropy": 5.62686014175415,
      "epoch": 3.5944303338171264,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037487246525986384,
      "loss": 4.8935,
      "mean_token_accuracy": 0.2190272852778435,
      "num_tokens": 82216140.0,
      "step": 39625
    },
    {
      "entropy": 5.7096781730651855,
      "epoch": 3.59488388969521,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003748434700466856,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22008083909749984,
      "num_tokens": 82226239.0,
      "step": 39630
    },
    {
      "entropy": 5.674899291992188,
      "epoch": 3.5953374455732945,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003748144727687188,
      "loss": 4.868,
      "mean_token_accuracy": 0.2250571444630623,
      "num_tokens": 82235795.0,
      "step": 39635
    },
    {
      "entropy": 5.600372695922852,
      "epoch": 3.5957910014513788,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037478547342656345,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23095702081918718,
      "num_tokens": 82245644.0,
      "step": 39640
    },
    {
      "entropy": 5.620706701278687,
      "epoch": 3.596244557329463,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000374756472020819,
      "loss": 4.8631,
      "mean_token_accuracy": 0.2238653853535652,
      "num_tokens": 82255608.0,
      "step": 39645
    },
    {
      "entropy": 5.62995138168335,
      "epoch": 3.5966981132075473,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037472746855208525,
      "loss": 4.854,
      "mean_token_accuracy": 0.22334845811128617,
      "num_tokens": 82265488.0,
      "step": 39650
    },
    {
      "entropy": 5.63012580871582,
      "epoch": 3.5971516690856316,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003746984630209621,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2154064193367958,
      "num_tokens": 82276866.0,
      "step": 39655
    },
    {
      "entropy": 5.635176706314087,
      "epoch": 3.5976052249637154,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00037466945542804933,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22262948304414748,
      "num_tokens": 82287125.0,
      "step": 39660
    },
    {
      "entropy": 5.699474716186524,
      "epoch": 3.5980587808417996,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037464044577394676,
      "loss": 4.9151,
      "mean_token_accuracy": 0.21946818977594376,
      "num_tokens": 82296887.0,
      "step": 39665
    },
    {
      "entropy": 5.672021770477295,
      "epoch": 3.598512336719884,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003746114340592544,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21009893864393234,
      "num_tokens": 82306928.0,
      "step": 39670
    },
    {
      "entropy": 5.638008117675781,
      "epoch": 3.598965892597968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037458242028457216,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22398993670940398,
      "num_tokens": 82317873.0,
      "step": 39675
    },
    {
      "entropy": 5.581651830673218,
      "epoch": 3.599419448476052,
      "grad_norm": 1.125,
      "learning_rate": 0.00037455340445050023,
      "loss": 4.7702,
      "mean_token_accuracy": 0.23053000420331954,
      "num_tokens": 82328184.0,
      "step": 39680
    },
    {
      "entropy": 5.611350727081299,
      "epoch": 3.5998730043541363,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037452438655763843,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22242008596658708,
      "num_tokens": 82338191.0,
      "step": 39685
    },
    {
      "entropy": 5.628508281707764,
      "epoch": 3.6003265602322205,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037449536660658696,
      "loss": 4.9667,
      "mean_token_accuracy": 0.2164360001683235,
      "num_tokens": 82348095.0,
      "step": 39690
    },
    {
      "entropy": 5.626165437698364,
      "epoch": 3.600780116110305,
      "grad_norm": 1.265625,
      "learning_rate": 0.000374466344597946,
      "loss": 4.8003,
      "mean_token_accuracy": 0.2291667327284813,
      "num_tokens": 82358216.0,
      "step": 39695
    },
    {
      "entropy": 5.589766788482666,
      "epoch": 3.601233671988389,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003744373205323157,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22677585035562514,
      "num_tokens": 82368581.0,
      "step": 39700
    },
    {
      "entropy": 5.613586521148681,
      "epoch": 3.6016872278664733,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003744082944102962,
      "loss": 4.8805,
      "mean_token_accuracy": 0.21822649836540223,
      "num_tokens": 82379609.0,
      "step": 39705
    },
    {
      "entropy": 5.672139596939087,
      "epoch": 3.6021407837445576,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003743792662324878,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2171761378645897,
      "num_tokens": 82390589.0,
      "step": 39710
    },
    {
      "entropy": 5.582696962356567,
      "epoch": 3.6025943396226414,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037435023599949086,
      "loss": 4.8515,
      "mean_token_accuracy": 0.21978304535150528,
      "num_tokens": 82400915.0,
      "step": 39715
    },
    {
      "entropy": 5.622618770599365,
      "epoch": 3.6030478955007257,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037432120371190585,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2261260449886322,
      "num_tokens": 82411544.0,
      "step": 39720
    },
    {
      "entropy": 5.617326927185059,
      "epoch": 3.60350145137881,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037429216937033294,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2225077897310257,
      "num_tokens": 82422391.0,
      "step": 39725
    },
    {
      "entropy": 5.619288444519043,
      "epoch": 3.6039550072568938,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037426313297537263,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22646329551935196,
      "num_tokens": 82433564.0,
      "step": 39730
    },
    {
      "entropy": 5.6200549602508545,
      "epoch": 3.604408563134978,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003742340945276255,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2234766185283661,
      "num_tokens": 82445005.0,
      "step": 39735
    },
    {
      "entropy": 5.632007837295532,
      "epoch": 3.6048621190130623,
      "grad_norm": 1.28125,
      "learning_rate": 0.000374205054027692,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21421051472425462,
      "num_tokens": 82455998.0,
      "step": 39740
    },
    {
      "entropy": 5.645858478546143,
      "epoch": 3.6053156748911466,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037417601147617265,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22270803153514862,
      "num_tokens": 82466789.0,
      "step": 39745
    },
    {
      "entropy": 5.718984889984131,
      "epoch": 3.605769230769231,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037414696687366816,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21109584867954254,
      "num_tokens": 82478181.0,
      "step": 39750
    },
    {
      "entropy": 5.650729942321777,
      "epoch": 3.606222786647315,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037411792022077917,
      "loss": 4.8413,
      "mean_token_accuracy": 0.21976753324270248,
      "num_tokens": 82488452.0,
      "step": 39755
    },
    {
      "entropy": 5.690230798721314,
      "epoch": 3.6066763425253994,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003740888715181063,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22004005908966065,
      "num_tokens": 82498179.0,
      "step": 39760
    },
    {
      "entropy": 5.61546139717102,
      "epoch": 3.607129898403483,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037405982076625026,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22491095960140228,
      "num_tokens": 82508242.0,
      "step": 39765
    },
    {
      "entropy": 5.690432977676392,
      "epoch": 3.6075834542815675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037403076796581195,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21453270614147185,
      "num_tokens": 82519363.0,
      "step": 39770
    },
    {
      "entropy": 5.7164936542510985,
      "epoch": 3.6080370101596517,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003740017131173921,
      "loss": 4.921,
      "mean_token_accuracy": 0.21915996670722962,
      "num_tokens": 82529745.0,
      "step": 39775
    },
    {
      "entropy": 5.705504751205444,
      "epoch": 3.608490566037736,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003739726562215916,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21702400892972945,
      "num_tokens": 82539573.0,
      "step": 39780
    },
    {
      "entropy": 5.679179286956787,
      "epoch": 3.60894412191582,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037394359727901136,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22171331495046614,
      "num_tokens": 82551313.0,
      "step": 39785
    },
    {
      "entropy": 5.619750356674194,
      "epoch": 3.609397677793904,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003739145362902523,
      "loss": 4.8853,
      "mean_token_accuracy": 0.2150670364499092,
      "num_tokens": 82561283.0,
      "step": 39790
    },
    {
      "entropy": 5.618160009384155,
      "epoch": 3.6098512336719883,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003738854732559153,
      "loss": 4.8098,
      "mean_token_accuracy": 0.2296390116214752,
      "num_tokens": 82571118.0,
      "step": 39795
    },
    {
      "entropy": 5.6271984577178955,
      "epoch": 3.6103047895500726,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003738564081766017,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22267504632472992,
      "num_tokens": 82581663.0,
      "step": 39800
    },
    {
      "entropy": 5.610398149490356,
      "epoch": 3.610758345428157,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037382734105291234,
      "loss": 4.8495,
      "mean_token_accuracy": 0.2219679594039917,
      "num_tokens": 82592593.0,
      "step": 39805
    },
    {
      "entropy": 5.62231936454773,
      "epoch": 3.611211901306241,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037379827188544835,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21576057970523835,
      "num_tokens": 82603047.0,
      "step": 39810
    },
    {
      "entropy": 5.708199548721313,
      "epoch": 3.611665457184325,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037376920067481095,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2200032129883766,
      "num_tokens": 82612789.0,
      "step": 39815
    },
    {
      "entropy": 5.644432163238525,
      "epoch": 3.6121190130624092,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003737401274216012,
      "loss": 4.8689,
      "mean_token_accuracy": 0.22033663392066954,
      "num_tokens": 82622881.0,
      "step": 39820
    },
    {
      "entropy": 5.630075168609619,
      "epoch": 3.6125725689404935,
      "grad_norm": 1.375,
      "learning_rate": 0.0003737110521264206,
      "loss": 4.8365,
      "mean_token_accuracy": 0.21775075048208237,
      "num_tokens": 82633933.0,
      "step": 39825
    },
    {
      "entropy": 5.644597721099854,
      "epoch": 3.6130261248185778,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003736819747898702,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2244173228740692,
      "num_tokens": 82645069.0,
      "step": 39830
    },
    {
      "entropy": 5.686211824417114,
      "epoch": 3.6134796806966616,
      "grad_norm": 1.25,
      "learning_rate": 0.00037365289541255154,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2205202490091324,
      "num_tokens": 82655152.0,
      "step": 39835
    },
    {
      "entropy": 5.601044797897339,
      "epoch": 3.613933236574746,
      "grad_norm": 1.25,
      "learning_rate": 0.00037362381399506575,
      "loss": 4.8633,
      "mean_token_accuracy": 0.21666934192180634,
      "num_tokens": 82665260.0,
      "step": 39840
    },
    {
      "entropy": 5.623809194564819,
      "epoch": 3.61438679245283,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037359473053801435,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22947915643453598,
      "num_tokens": 82676287.0,
      "step": 39845
    },
    {
      "entropy": 5.657204723358154,
      "epoch": 3.6148403483309144,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003735656450419988,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21008455753326416,
      "num_tokens": 82686958.0,
      "step": 39850
    },
    {
      "entropy": 5.677321052551269,
      "epoch": 3.6152939042089987,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037353655750762063,
      "loss": 5.0309,
      "mean_token_accuracy": 0.20693982839584352,
      "num_tokens": 82698623.0,
      "step": 39855
    },
    {
      "entropy": 5.68940691947937,
      "epoch": 3.615747460087083,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037350746793548135,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21703476309776307,
      "num_tokens": 82709146.0,
      "step": 39860
    },
    {
      "entropy": 5.647206783294678,
      "epoch": 3.6162010159651667,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037347837632618246,
      "loss": 4.8297,
      "mean_token_accuracy": 0.22628388404846192,
      "num_tokens": 82719323.0,
      "step": 39865
    },
    {
      "entropy": 5.681375932693482,
      "epoch": 3.616654571843251,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037344928268032574,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2160400152206421,
      "num_tokens": 82730068.0,
      "step": 39870
    },
    {
      "entropy": 5.665633058547973,
      "epoch": 3.6171081277213353,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003734201869985127,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22275419533252716,
      "num_tokens": 82740460.0,
      "step": 39875
    },
    {
      "entropy": 5.627183532714843,
      "epoch": 3.6175616835994195,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003733910892813451,
      "loss": 4.8583,
      "mean_token_accuracy": 0.21603755205869674,
      "num_tokens": 82751488.0,
      "step": 39880
    },
    {
      "entropy": 5.540359878540039,
      "epoch": 3.6180152394775034,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037336198952942477,
      "loss": 4.7839,
      "mean_token_accuracy": 0.2355194568634033,
      "num_tokens": 82762061.0,
      "step": 39885
    },
    {
      "entropy": 5.6118522644042965,
      "epoch": 3.6184687953555876,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003733328877433534,
      "loss": 4.8433,
      "mean_token_accuracy": 0.23038543462753297,
      "num_tokens": 82773421.0,
      "step": 39890
    },
    {
      "entropy": 5.57828311920166,
      "epoch": 3.618922351233672,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037330378392373285,
      "loss": 4.7457,
      "mean_token_accuracy": 0.2281870111823082,
      "num_tokens": 82783342.0,
      "step": 39895
    },
    {
      "entropy": 5.63250994682312,
      "epoch": 3.619375907111756,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037327467807116507,
      "loss": 4.9027,
      "mean_token_accuracy": 0.21426529735326766,
      "num_tokens": 82793885.0,
      "step": 39900
    },
    {
      "entropy": 5.650369358062744,
      "epoch": 3.6198294629898404,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037324557018625186,
      "loss": 4.9345,
      "mean_token_accuracy": 0.21836500912904738,
      "num_tokens": 82803746.0,
      "step": 39905
    },
    {
      "entropy": 5.652484273910522,
      "epoch": 3.6202830188679247,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003732164602695952,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22193869054317475,
      "num_tokens": 82813858.0,
      "step": 39910
    },
    {
      "entropy": 5.707403230667114,
      "epoch": 3.620736574746009,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037318734832179705,
      "loss": 4.935,
      "mean_token_accuracy": 0.21966660618782044,
      "num_tokens": 82824358.0,
      "step": 39915
    },
    {
      "entropy": 5.676679849624634,
      "epoch": 3.621190130624093,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037315823434345963,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2315877005457878,
      "num_tokens": 82835801.0,
      "step": 39920
    },
    {
      "entropy": 5.567750787734985,
      "epoch": 3.621643686502177,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003731291183351848,
      "loss": 4.8228,
      "mean_token_accuracy": 0.2311795875430107,
      "num_tokens": 82845818.0,
      "step": 39925
    },
    {
      "entropy": 5.599418592453003,
      "epoch": 3.6220972423802613,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037310000029757486,
      "loss": 4.8589,
      "mean_token_accuracy": 0.22515048682689667,
      "num_tokens": 82856920.0,
      "step": 39930
    },
    {
      "entropy": 5.6821064949035645,
      "epoch": 3.6225507982583456,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003730708802312319,
      "loss": 4.961,
      "mean_token_accuracy": 0.21929134726524352,
      "num_tokens": 82867220.0,
      "step": 39935
    },
    {
      "entropy": 5.63896107673645,
      "epoch": 3.6230043541364294,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003730417581367582,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22628148794174194,
      "num_tokens": 82877541.0,
      "step": 39940
    },
    {
      "entropy": 5.690818405151367,
      "epoch": 3.6234579100145137,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003730126340147559,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22232334017753602,
      "num_tokens": 82888346.0,
      "step": 39945
    },
    {
      "entropy": 5.709954929351807,
      "epoch": 3.623911465892598,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003729835078658273,
      "loss": 4.9666,
      "mean_token_accuracy": 0.22023930847644807,
      "num_tokens": 82899721.0,
      "step": 39950
    },
    {
      "entropy": 5.662163877487183,
      "epoch": 3.624365021770682,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037295437969057493,
      "loss": 4.8809,
      "mean_token_accuracy": 0.22382118105888366,
      "num_tokens": 82910149.0,
      "step": 39955
    },
    {
      "entropy": 5.654334306716919,
      "epoch": 3.6248185776487665,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037292524948960093,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22447170913219452,
      "num_tokens": 82920883.0,
      "step": 39960
    },
    {
      "entropy": 5.646292591094971,
      "epoch": 3.6252721335268507,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003728961172635078,
      "loss": 4.9198,
      "mean_token_accuracy": 0.21913919001817703,
      "num_tokens": 82931091.0,
      "step": 39965
    },
    {
      "entropy": 5.651579666137695,
      "epoch": 3.6257256894049346,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037286698301289804,
      "loss": 4.8075,
      "mean_token_accuracy": 0.22505465894937515,
      "num_tokens": 82941108.0,
      "step": 39970
    },
    {
      "entropy": 5.613724613189698,
      "epoch": 3.626179245283019,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003728378467383742,
      "loss": 4.8716,
      "mean_token_accuracy": 0.22475457191467285,
      "num_tokens": 82951943.0,
      "step": 39975
    },
    {
      "entropy": 5.696922016143799,
      "epoch": 3.626632801161103,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003728087084405387,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22236319184303283,
      "num_tokens": 82961495.0,
      "step": 39980
    },
    {
      "entropy": 5.6964035987854,
      "epoch": 3.6270863570391874,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037277956811999414,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22365380823612213,
      "num_tokens": 82971858.0,
      "step": 39985
    },
    {
      "entropy": 5.661520957946777,
      "epoch": 3.627539912917271,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003727504257773433,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22139185965061187,
      "num_tokens": 82982805.0,
      "step": 39990
    },
    {
      "entropy": 5.646217393875122,
      "epoch": 3.6279934687953554,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037272128141318866,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21410252898931503,
      "num_tokens": 82993446.0,
      "step": 39995
    },
    {
      "entropy": 5.619251155853272,
      "epoch": 3.6284470246734397,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000372692135028133,
      "loss": 4.8053,
      "mean_token_accuracy": 0.22217184603214263,
      "num_tokens": 83003378.0,
      "step": 40000
    },
    {
      "entropy": 5.636310052871704,
      "epoch": 3.628900580551524,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037266298662277914,
      "loss": 4.8367,
      "mean_token_accuracy": 0.22047396302223204,
      "num_tokens": 83014223.0,
      "step": 40005
    },
    {
      "entropy": 5.6078626155853275,
      "epoch": 3.6293541364296082,
      "grad_norm": 1.25,
      "learning_rate": 0.00037263383619772986,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22089984714984895,
      "num_tokens": 83024691.0,
      "step": 40010
    },
    {
      "entropy": 5.651585388183594,
      "epoch": 3.6298076923076925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037260468375358794,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2201688289642334,
      "num_tokens": 83035704.0,
      "step": 40015
    },
    {
      "entropy": 5.693813705444336,
      "epoch": 3.6302612481857763,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003725755292909562,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21433429718017577,
      "num_tokens": 83047173.0,
      "step": 40020
    },
    {
      "entropy": 5.709642553329468,
      "epoch": 3.6307148040638606,
      "grad_norm": 1.375,
      "learning_rate": 0.0003725463728104378,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21617625653743744,
      "num_tokens": 83056399.0,
      "step": 40025
    },
    {
      "entropy": 5.623583221435547,
      "epoch": 3.631168359941945,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003725172143126354,
      "loss": 4.8387,
      "mean_token_accuracy": 0.21872634142637254,
      "num_tokens": 83066413.0,
      "step": 40030
    },
    {
      "entropy": 5.620371294021607,
      "epoch": 3.631621915820029,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037248805379815227,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21911440640687943,
      "num_tokens": 83077843.0,
      "step": 40035
    },
    {
      "entropy": 5.700660181045532,
      "epoch": 3.632075471698113,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003724588912675913,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21824240535497666,
      "num_tokens": 83088585.0,
      "step": 40040
    },
    {
      "entropy": 5.731334829330445,
      "epoch": 3.6325290275761972,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003724297267215557,
      "loss": 5.0125,
      "mean_token_accuracy": 0.22148914635181427,
      "num_tokens": 83099295.0,
      "step": 40045
    },
    {
      "entropy": 5.600619697570801,
      "epoch": 3.6329825834542815,
      "grad_norm": 1.375,
      "learning_rate": 0.00037240056016064844,
      "loss": 4.8074,
      "mean_token_accuracy": 0.2316656768321991,
      "num_tokens": 83108530.0,
      "step": 40050
    },
    {
      "entropy": 5.610688352584839,
      "epoch": 3.6334361393323658,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003723713915854728,
      "loss": 4.8933,
      "mean_token_accuracy": 0.21968873739242553,
      "num_tokens": 83119703.0,
      "step": 40055
    },
    {
      "entropy": 5.667976999282837,
      "epoch": 3.63388969521045,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003723422209966319,
      "loss": 4.873,
      "mean_token_accuracy": 0.22223049104213716,
      "num_tokens": 83129323.0,
      "step": 40060
    },
    {
      "entropy": 5.653559017181396,
      "epoch": 3.6343432510885343,
      "grad_norm": 1.25,
      "learning_rate": 0.0003723130483947292,
      "loss": 4.8419,
      "mean_token_accuracy": 0.2281960129737854,
      "num_tokens": 83139339.0,
      "step": 40065
    },
    {
      "entropy": 5.606348609924316,
      "epoch": 3.6347968069666186,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003722838737803678,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22200387716293335,
      "num_tokens": 83149845.0,
      "step": 40070
    },
    {
      "entropy": 5.68199348449707,
      "epoch": 3.6352503628447024,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037225469715415106,
      "loss": 4.9082,
      "mean_token_accuracy": 0.21887548118829728,
      "num_tokens": 83160332.0,
      "step": 40075
    },
    {
      "entropy": 5.687393379211426,
      "epoch": 3.6357039187227866,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037222551851668245,
      "loss": 4.9171,
      "mean_token_accuracy": 0.21995497047901152,
      "num_tokens": 83170892.0,
      "step": 40080
    },
    {
      "entropy": 5.6071295738220215,
      "epoch": 3.636157474600871,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003721963378685652,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22541280686855317,
      "num_tokens": 83181035.0,
      "step": 40085
    },
    {
      "entropy": 5.6039036273956295,
      "epoch": 3.6366110304789547,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037216715521040314,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22605569362640382,
      "num_tokens": 83191886.0,
      "step": 40090
    },
    {
      "entropy": 5.643513298034668,
      "epoch": 3.637064586357039,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003721379705427994,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22598879784345627,
      "num_tokens": 83202356.0,
      "step": 40095
    },
    {
      "entropy": 5.663507986068725,
      "epoch": 3.6375181422351233,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037210878386635774,
      "loss": 4.9216,
      "mean_token_accuracy": 0.2278437063097954,
      "num_tokens": 83212909.0,
      "step": 40100
    },
    {
      "entropy": 5.6779200553894045,
      "epoch": 3.6379716981132075,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037207959518168165,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2168870598077774,
      "num_tokens": 83223523.0,
      "step": 40105
    },
    {
      "entropy": 5.7612841606140135,
      "epoch": 3.638425253991292,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037205040448937485,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21286313831806183,
      "num_tokens": 83233665.0,
      "step": 40110
    },
    {
      "entropy": 5.625516033172607,
      "epoch": 3.638878809869376,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003720212117900409,
      "loss": 4.8119,
      "mean_token_accuracy": 0.23000577092170715,
      "num_tokens": 83243563.0,
      "step": 40115
    },
    {
      "entropy": 5.634356927871704,
      "epoch": 3.6393323657474603,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037199201708428356,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22218847423791885,
      "num_tokens": 83254246.0,
      "step": 40120
    },
    {
      "entropy": 5.66231861114502,
      "epoch": 3.639785921625544,
      "grad_norm": 1.25,
      "learning_rate": 0.0003719628203727066,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2198367714881897,
      "num_tokens": 83264361.0,
      "step": 40125
    },
    {
      "entropy": 5.692606019973755,
      "epoch": 3.6402394775036284,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003719336216559138,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22032452821731568,
      "num_tokens": 83274193.0,
      "step": 40130
    },
    {
      "entropy": 5.678310585021973,
      "epoch": 3.6406930333817127,
      "grad_norm": 1.34375,
      "learning_rate": 0.000371904420934509,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22529397606849672,
      "num_tokens": 83284593.0,
      "step": 40135
    },
    {
      "entropy": 5.596283674240112,
      "epoch": 3.641146589259797,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003718752182090961,
      "loss": 4.8619,
      "mean_token_accuracy": 0.2219424530863762,
      "num_tokens": 83294602.0,
      "step": 40140
    },
    {
      "entropy": 5.638093137741089,
      "epoch": 3.6416001451378808,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000371846013480279,
      "loss": 4.8765,
      "mean_token_accuracy": 0.2213722974061966,
      "num_tokens": 83305399.0,
      "step": 40145
    },
    {
      "entropy": 5.576687622070312,
      "epoch": 3.642053701015965,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037181680674866165,
      "loss": 4.8367,
      "mean_token_accuracy": 0.22188881933689117,
      "num_tokens": 83315696.0,
      "step": 40150
    },
    {
      "entropy": 5.712692785263061,
      "epoch": 3.6425072568940493,
      "grad_norm": 1.25,
      "learning_rate": 0.0003717875980148481,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21227151155471802,
      "num_tokens": 83326686.0,
      "step": 40155
    },
    {
      "entropy": 5.677354240417481,
      "epoch": 3.6429608127721336,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003717583872794423,
      "loss": 4.8554,
      "mean_token_accuracy": 0.2321374773979187,
      "num_tokens": 83336902.0,
      "step": 40160
    },
    {
      "entropy": 5.6812585353851315,
      "epoch": 3.643414368650218,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003717291745430484,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22088903337717056,
      "num_tokens": 83347324.0,
      "step": 40165
    },
    {
      "entropy": 5.57171950340271,
      "epoch": 3.643867924528302,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037169995980627056,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22425670623779298,
      "num_tokens": 83358458.0,
      "step": 40170
    },
    {
      "entropy": 5.580731678009033,
      "epoch": 3.644321480406386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037167074306971285,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21660740971565245,
      "num_tokens": 83369047.0,
      "step": 40175
    },
    {
      "entropy": 5.6755071640014645,
      "epoch": 3.64477503628447,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037164152433397964,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22275365889072418,
      "num_tokens": 83378891.0,
      "step": 40180
    },
    {
      "entropy": 5.674152517318726,
      "epoch": 3.6452285921625545,
      "grad_norm": 1.359375,
      "learning_rate": 0.000371612303599675,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2193396806716919,
      "num_tokens": 83389289.0,
      "step": 40185
    },
    {
      "entropy": 5.715043783187866,
      "epoch": 3.6456821480406387,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037158308086740336,
      "loss": 4.9654,
      "mean_token_accuracy": 0.222805917263031,
      "num_tokens": 83399316.0,
      "step": 40190
    },
    {
      "entropy": 5.6228443622589115,
      "epoch": 3.6461357039187225,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003715538561377689,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22524963915348054,
      "num_tokens": 83409507.0,
      "step": 40195
    },
    {
      "entropy": 5.710901165008545,
      "epoch": 3.646589259796807,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003715246294113762,
      "loss": 4.9905,
      "mean_token_accuracy": 0.20930583477020265,
      "num_tokens": 83419687.0,
      "step": 40200
    },
    {
      "entropy": 5.762218332290649,
      "epoch": 3.647042815674891,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037149540068882945,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21536059230566024,
      "num_tokens": 83429711.0,
      "step": 40205
    },
    {
      "entropy": 5.62012300491333,
      "epoch": 3.6474963715529753,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037146616997073327,
      "loss": 4.8466,
      "mean_token_accuracy": 0.21585071533918382,
      "num_tokens": 83440116.0,
      "step": 40210
    },
    {
      "entropy": 5.641993999481201,
      "epoch": 3.6479499274310596,
      "grad_norm": 1.25,
      "learning_rate": 0.00037143693725769214,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22562815397977828,
      "num_tokens": 83450877.0,
      "step": 40215
    },
    {
      "entropy": 5.590087413787842,
      "epoch": 3.648403483309144,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037140770255031066,
      "loss": 4.793,
      "mean_token_accuracy": 0.23063643127679825,
      "num_tokens": 83460080.0,
      "step": 40220
    },
    {
      "entropy": 5.63749942779541,
      "epoch": 3.6488570391872277,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037137846584919316,
      "loss": 4.945,
      "mean_token_accuracy": 0.2169843077659607,
      "num_tokens": 83469933.0,
      "step": 40225
    },
    {
      "entropy": 5.632030296325683,
      "epoch": 3.649310595065312,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003713492271549445,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22371985465288163,
      "num_tokens": 83480477.0,
      "step": 40230
    },
    {
      "entropy": 5.632258319854737,
      "epoch": 3.6497641509433962,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003713199864681693,
      "loss": 4.8867,
      "mean_token_accuracy": 0.221931654214859,
      "num_tokens": 83491022.0,
      "step": 40235
    },
    {
      "entropy": 5.662606525421142,
      "epoch": 3.6502177068214805,
      "grad_norm": 1.25,
      "learning_rate": 0.00037129074378947216,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22252005636692046,
      "num_tokens": 83501643.0,
      "step": 40240
    },
    {
      "entropy": 5.687870359420776,
      "epoch": 3.6506712626995643,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037126149911945796,
      "loss": 4.9414,
      "mean_token_accuracy": 0.2132761538028717,
      "num_tokens": 83511948.0,
      "step": 40245
    },
    {
      "entropy": 5.677402496337891,
      "epoch": 3.6511248185776486,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037123225245873137,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21967084407806398,
      "num_tokens": 83521939.0,
      "step": 40250
    },
    {
      "entropy": 5.723622798919678,
      "epoch": 3.651578374455733,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037120300380789727,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2197018086910248,
      "num_tokens": 83532098.0,
      "step": 40255
    },
    {
      "entropy": 5.658861398696899,
      "epoch": 3.652031930333817,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037117375316756063,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2199383035302162,
      "num_tokens": 83543238.0,
      "step": 40260
    },
    {
      "entropy": 5.6958314895629885,
      "epoch": 3.6524854862119014,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003711445005383263,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2177405059337616,
      "num_tokens": 83552969.0,
      "step": 40265
    },
    {
      "entropy": 5.697970724105835,
      "epoch": 3.6529390420899857,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037111524592079896,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2078731656074524,
      "num_tokens": 83564569.0,
      "step": 40270
    },
    {
      "entropy": 5.679282093048096,
      "epoch": 3.65339259796807,
      "grad_norm": 1.234375,
      "learning_rate": 0.000371085989315584,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21783456355333328,
      "num_tokens": 83575613.0,
      "step": 40275
    },
    {
      "entropy": 5.6941159725189205,
      "epoch": 3.6538461538461537,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003710567307232864,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21702758520841597,
      "num_tokens": 83586311.0,
      "step": 40280
    },
    {
      "entropy": 5.707441282272339,
      "epoch": 3.654299709724238,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037102747014451095,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2187309056520462,
      "num_tokens": 83596660.0,
      "step": 40285
    },
    {
      "entropy": 5.697196102142334,
      "epoch": 3.6547532656023223,
      "grad_norm": 1.359375,
      "learning_rate": 0.000370998207579863,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21857304722070695,
      "num_tokens": 83606882.0,
      "step": 40290
    },
    {
      "entropy": 5.704786157608032,
      "epoch": 3.6552068214804065,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003709689430299477,
      "loss": 4.9485,
      "mean_token_accuracy": 0.22320753186941147,
      "num_tokens": 83617380.0,
      "step": 40295
    },
    {
      "entropy": 5.661019563674927,
      "epoch": 3.6556603773584904,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003709396764953701,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21984029114246367,
      "num_tokens": 83627961.0,
      "step": 40300
    },
    {
      "entropy": 5.528314256668091,
      "epoch": 3.6561139332365746,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003709104079767357,
      "loss": 4.7984,
      "mean_token_accuracy": 0.2310820922255516,
      "num_tokens": 83637646.0,
      "step": 40305
    },
    {
      "entropy": 5.648510265350342,
      "epoch": 3.656567489114659,
      "grad_norm": 1.25,
      "learning_rate": 0.00037088113747464955,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22455067485570906,
      "num_tokens": 83647807.0,
      "step": 40310
    },
    {
      "entropy": 5.658698081970215,
      "epoch": 3.657021044992743,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000370851864989717,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21397348046302794,
      "num_tokens": 83658002.0,
      "step": 40315
    },
    {
      "entropy": 5.629558563232422,
      "epoch": 3.6574746008708274,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003708225905225435,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23341089189052583,
      "num_tokens": 83667875.0,
      "step": 40320
    },
    {
      "entropy": 5.6465513706207275,
      "epoch": 3.6579281567489117,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003707933140737345,
      "loss": 4.8806,
      "mean_token_accuracy": 0.2244977906346321,
      "num_tokens": 83678376.0,
      "step": 40325
    },
    {
      "entropy": 5.675090217590332,
      "epoch": 3.6583817126269955,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037076403564389516,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2119140088558197,
      "num_tokens": 83688318.0,
      "step": 40330
    },
    {
      "entropy": 5.6525876998901365,
      "epoch": 3.65883526850508,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003707347552336313,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2170547917485237,
      "num_tokens": 83698935.0,
      "step": 40335
    },
    {
      "entropy": 5.684643268585205,
      "epoch": 3.659288824383164,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003707054728435482,
      "loss": 4.9702,
      "mean_token_accuracy": 0.21953823864459993,
      "num_tokens": 83709382.0,
      "step": 40340
    },
    {
      "entropy": 5.627509021759034,
      "epoch": 3.6597423802612483,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003706761884742516,
      "loss": 4.81,
      "mean_token_accuracy": 0.23386919349431992,
      "num_tokens": 83719661.0,
      "step": 40345
    },
    {
      "entropy": 5.763510084152221,
      "epoch": 3.660195936139332,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037064690212634707,
      "loss": 4.9724,
      "mean_token_accuracy": 0.2139635980129242,
      "num_tokens": 83730679.0,
      "step": 40350
    },
    {
      "entropy": 5.58924617767334,
      "epoch": 3.6606494920174164,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003706176138004401,
      "loss": 4.8223,
      "mean_token_accuracy": 0.22291076332330703,
      "num_tokens": 83740595.0,
      "step": 40355
    },
    {
      "entropy": 5.609420347213745,
      "epoch": 3.6611030478955007,
      "grad_norm": 1.25,
      "learning_rate": 0.00037058832349713667,
      "loss": 4.958,
      "mean_token_accuracy": 0.21693361401557923,
      "num_tokens": 83751605.0,
      "step": 40360
    },
    {
      "entropy": 5.71450662612915,
      "epoch": 3.661556603773585,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003705590312170422,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21410494595766066,
      "num_tokens": 83763780.0,
      "step": 40365
    },
    {
      "entropy": 5.648506307601929,
      "epoch": 3.662010159651669,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037052973696076274,
      "loss": 4.8402,
      "mean_token_accuracy": 0.23000605702400206,
      "num_tokens": 83773800.0,
      "step": 40370
    },
    {
      "entropy": 5.716043519973755,
      "epoch": 3.6624637155297535,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037050044072890395,
      "loss": 4.9542,
      "mean_token_accuracy": 0.20661567896604538,
      "num_tokens": 83784402.0,
      "step": 40375
    },
    {
      "entropy": 5.627768611907959,
      "epoch": 3.6629172714078373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003704711425220717,
      "loss": 4.8629,
      "mean_token_accuracy": 0.21692775338888168,
      "num_tokens": 83794852.0,
      "step": 40380
    },
    {
      "entropy": 5.673155164718628,
      "epoch": 3.6633708272859216,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003704418423408718,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2184833139181137,
      "num_tokens": 83805673.0,
      "step": 40385
    },
    {
      "entropy": 5.649004077911377,
      "epoch": 3.663824383164006,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003704125401859104,
      "loss": 4.8999,
      "mean_token_accuracy": 0.2189710557460785,
      "num_tokens": 83815779.0,
      "step": 40390
    },
    {
      "entropy": 5.671184062957764,
      "epoch": 3.66427793904209,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037038323605779333,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22570633590221406,
      "num_tokens": 83826804.0,
      "step": 40395
    },
    {
      "entropy": 5.680573463439941,
      "epoch": 3.664731494920174,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003703539299571266,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22130713760852813,
      "num_tokens": 83836500.0,
      "step": 40400
    },
    {
      "entropy": 5.682557439804077,
      "epoch": 3.665185050798258,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003703246218845163,
      "loss": 4.8744,
      "mean_token_accuracy": 0.2224021315574646,
      "num_tokens": 83846647.0,
      "step": 40405
    },
    {
      "entropy": 5.595215797424316,
      "epoch": 3.6656386066763424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003702953118405686,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22574926912784576,
      "num_tokens": 83858088.0,
      "step": 40410
    },
    {
      "entropy": 5.673348903656006,
      "epoch": 3.6660921625544267,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003702659998258895,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2200544461607933,
      "num_tokens": 83868936.0,
      "step": 40415
    },
    {
      "entropy": 5.739882135391236,
      "epoch": 3.666545718432511,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037023668584108523,
      "loss": 5.0251,
      "mean_token_accuracy": 0.20719054192304612,
      "num_tokens": 83879813.0,
      "step": 40420
    },
    {
      "entropy": 5.606545877456665,
      "epoch": 3.6669992743105952,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037020736988676206,
      "loss": 4.8575,
      "mean_token_accuracy": 0.21960405111312867,
      "num_tokens": 83891619.0,
      "step": 40425
    },
    {
      "entropy": 5.722917604446411,
      "epoch": 3.6674528301886795,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037017805196352623,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21911616027355194,
      "num_tokens": 83901905.0,
      "step": 40430
    },
    {
      "entropy": 5.625254201889038,
      "epoch": 3.6679063860667633,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000370148732071984,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2237980470061302,
      "num_tokens": 83912827.0,
      "step": 40435
    },
    {
      "entropy": 5.5971778392791744,
      "epoch": 3.6683599419448476,
      "grad_norm": 1.25,
      "learning_rate": 0.0003701194102127418,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23318384289741517,
      "num_tokens": 83923719.0,
      "step": 40440
    },
    {
      "entropy": 5.539970636367798,
      "epoch": 3.668813497822932,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003700900863864059,
      "loss": 4.8664,
      "mean_token_accuracy": 0.22861595153808595,
      "num_tokens": 83933715.0,
      "step": 40445
    },
    {
      "entropy": 5.668402290344238,
      "epoch": 3.6692670537010157,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003700607605935828,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21911519765853882,
      "num_tokens": 83943946.0,
      "step": 40450
    },
    {
      "entropy": 5.710284471511841,
      "epoch": 3.6697206095791,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003700314328348789,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2219032183289528,
      "num_tokens": 83953414.0,
      "step": 40455
    },
    {
      "entropy": 5.631934976577758,
      "epoch": 3.6701741654571842,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003700021031109008,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2187546521425247,
      "num_tokens": 83963717.0,
      "step": 40460
    },
    {
      "entropy": 5.616801500320435,
      "epoch": 3.6706277213352685,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036997277142225506,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2211845651268959,
      "num_tokens": 83973467.0,
      "step": 40465
    },
    {
      "entropy": 5.637853670120239,
      "epoch": 3.6710812772133528,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036994343776954806,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22129099667072297,
      "num_tokens": 83984013.0,
      "step": 40470
    },
    {
      "entropy": 5.646801471710205,
      "epoch": 3.671534833091437,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003699141021533867,
      "loss": 4.8053,
      "mean_token_accuracy": 0.2273121953010559,
      "num_tokens": 83994349.0,
      "step": 40475
    },
    {
      "entropy": 5.622933959960937,
      "epoch": 3.6719883889695213,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003698847645743775,
      "loss": 4.8092,
      "mean_token_accuracy": 0.21950105279684068,
      "num_tokens": 84004525.0,
      "step": 40480
    },
    {
      "entropy": 5.612362957000732,
      "epoch": 3.672441944847605,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003698554250331271,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21936506181955337,
      "num_tokens": 84014286.0,
      "step": 40485
    },
    {
      "entropy": 5.599745082855224,
      "epoch": 3.6728955007256894,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003698260835302424,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22524648159742355,
      "num_tokens": 84024826.0,
      "step": 40490
    },
    {
      "entropy": 5.679998874664307,
      "epoch": 3.6733490566037736,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003697967400663302,
      "loss": 4.912,
      "mean_token_accuracy": 0.219159072637558,
      "num_tokens": 84035726.0,
      "step": 40495
    },
    {
      "entropy": 5.60289421081543,
      "epoch": 3.673802612481858,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036976739464199715,
      "loss": 4.7954,
      "mean_token_accuracy": 0.22538332641124725,
      "num_tokens": 84045573.0,
      "step": 40500
    },
    {
      "entropy": 5.6782999515533445,
      "epoch": 3.6742561683599417,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036973804725785024,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21584745347499848,
      "num_tokens": 84055983.0,
      "step": 40505
    },
    {
      "entropy": 5.703087425231933,
      "epoch": 3.674709724238026,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036970869791449644,
      "loss": 4.9807,
      "mean_token_accuracy": 0.219372983276844,
      "num_tokens": 84066905.0,
      "step": 40510
    },
    {
      "entropy": 5.609340000152588,
      "epoch": 3.6751632801161103,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036967934661254263,
      "loss": 4.8019,
      "mean_token_accuracy": 0.234575517475605,
      "num_tokens": 84076861.0,
      "step": 40515
    },
    {
      "entropy": 5.671474361419678,
      "epoch": 3.6756168359941945,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003696499933525957,
      "loss": 4.8625,
      "mean_token_accuracy": 0.2270957425236702,
      "num_tokens": 84086167.0,
      "step": 40520
    },
    {
      "entropy": 5.596030426025391,
      "epoch": 3.676070391872279,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036962063813526287,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2241722896695137,
      "num_tokens": 84095306.0,
      "step": 40525
    },
    {
      "entropy": 5.602426958084107,
      "epoch": 3.676523947750363,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003695912809611511,
      "loss": 4.8962,
      "mean_token_accuracy": 0.21810519695281982,
      "num_tokens": 84106505.0,
      "step": 40530
    },
    {
      "entropy": 5.69519681930542,
      "epoch": 3.676977503628447,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003695619218308676,
      "loss": 4.8571,
      "mean_token_accuracy": 0.2240533635020256,
      "num_tokens": 84116393.0,
      "step": 40535
    },
    {
      "entropy": 5.655912303924561,
      "epoch": 3.677431059506531,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036953256074501937,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2188106283545494,
      "num_tokens": 84125890.0,
      "step": 40540
    },
    {
      "entropy": 5.672696495056153,
      "epoch": 3.6778846153846154,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003695031977042138,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2242170676589012,
      "num_tokens": 84137736.0,
      "step": 40545
    },
    {
      "entropy": 5.63420524597168,
      "epoch": 3.6783381712626997,
      "grad_norm": 1.4375,
      "learning_rate": 0.00036947383270905795,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22295952886343,
      "num_tokens": 84148974.0,
      "step": 40550
    },
    {
      "entropy": 5.650039434432983,
      "epoch": 3.6787917271407835,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036944446576015914,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21645791977643966,
      "num_tokens": 84159041.0,
      "step": 40555
    },
    {
      "entropy": 5.616870880126953,
      "epoch": 3.6792452830188678,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036941509685812476,
      "loss": 4.8108,
      "mean_token_accuracy": 0.228526309132576,
      "num_tokens": 84169432.0,
      "step": 40560
    },
    {
      "entropy": 5.617069959640503,
      "epoch": 3.679698838896952,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036938572600356213,
      "loss": 4.8546,
      "mean_token_accuracy": 0.21958910375833512,
      "num_tokens": 84179964.0,
      "step": 40565
    },
    {
      "entropy": 5.640071535110474,
      "epoch": 3.6801523947750363,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036935635319707856,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2178154930472374,
      "num_tokens": 84189496.0,
      "step": 40570
    },
    {
      "entropy": 5.641722106933594,
      "epoch": 3.6806059506531206,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003693269784392816,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22978959679603578,
      "num_tokens": 84200396.0,
      "step": 40575
    },
    {
      "entropy": 5.590877389907837,
      "epoch": 3.681059506531205,
      "grad_norm": 1.5,
      "learning_rate": 0.00036929760173077863,
      "loss": 4.8264,
      "mean_token_accuracy": 0.22643833309412004,
      "num_tokens": 84210114.0,
      "step": 40580
    },
    {
      "entropy": 5.639663124084473,
      "epoch": 3.6815130624092887,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003692682230721773,
      "loss": 4.869,
      "mean_token_accuracy": 0.2270136922597885,
      "num_tokens": 84220472.0,
      "step": 40585
    },
    {
      "entropy": 5.664499092102051,
      "epoch": 3.681966618287373,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003692388424640851,
      "loss": 4.92,
      "mean_token_accuracy": 0.2198185846209526,
      "num_tokens": 84230718.0,
      "step": 40590
    },
    {
      "entropy": 5.685943126678467,
      "epoch": 3.682420174165457,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036920945990710964,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2216021552681923,
      "num_tokens": 84241736.0,
      "step": 40595
    },
    {
      "entropy": 5.546687650680542,
      "epoch": 3.6828737300435415,
      "grad_norm": 1.25,
      "learning_rate": 0.00036918007540185846,
      "loss": 4.8206,
      "mean_token_accuracy": 0.2299594685435295,
      "num_tokens": 84251770.0,
      "step": 40600
    },
    {
      "entropy": 5.652056455612183,
      "epoch": 3.6833272859216253,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003691506889489394,
      "loss": 4.9296,
      "mean_token_accuracy": 0.2200792133808136,
      "num_tokens": 84263355.0,
      "step": 40605
    },
    {
      "entropy": 5.6606934547424315,
      "epoch": 3.6837808417997095,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036912130054896005,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21096354573965073,
      "num_tokens": 84274835.0,
      "step": 40610
    },
    {
      "entropy": 5.7257383346557615,
      "epoch": 3.684234397677794,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036909191020252825,
      "loss": 4.9946,
      "mean_token_accuracy": 0.21024842858314513,
      "num_tokens": 84286099.0,
      "step": 40615
    },
    {
      "entropy": 5.701558589935303,
      "epoch": 3.684687953555878,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003690625179102517,
      "loss": 4.9598,
      "mean_token_accuracy": 0.222568841278553,
      "num_tokens": 84296168.0,
      "step": 40620
    },
    {
      "entropy": 5.767237567901612,
      "epoch": 3.6851415094339623,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003690331236727384,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20820657461881636,
      "num_tokens": 84307352.0,
      "step": 40625
    },
    {
      "entropy": 5.671647262573242,
      "epoch": 3.6855950653120466,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003690037274905961,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21867855489253998,
      "num_tokens": 84318496.0,
      "step": 40630
    },
    {
      "entropy": 5.59604959487915,
      "epoch": 3.686048621190131,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003689743293644327,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2231295496225357,
      "num_tokens": 84329101.0,
      "step": 40635
    },
    {
      "entropy": 5.60073299407959,
      "epoch": 3.6865021770682147,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003689449292948563,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22625523507595063,
      "num_tokens": 84339143.0,
      "step": 40640
    },
    {
      "entropy": 5.697188520431519,
      "epoch": 3.686955732946299,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036891552728247483,
      "loss": 4.8881,
      "mean_token_accuracy": 0.21939027756452562,
      "num_tokens": 84349693.0,
      "step": 40645
    },
    {
      "entropy": 5.692017459869385,
      "epoch": 3.6874092888243832,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036888612332789636,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21815599650144576,
      "num_tokens": 84359799.0,
      "step": 40650
    },
    {
      "entropy": 5.6600888729095455,
      "epoch": 3.6878628447024675,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036885671743172875,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22099415957927704,
      "num_tokens": 84370695.0,
      "step": 40655
    },
    {
      "entropy": 5.643673467636108,
      "epoch": 3.6883164005805513,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003688273095945805,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21444448083639145,
      "num_tokens": 84380205.0,
      "step": 40660
    },
    {
      "entropy": 5.670331954956055,
      "epoch": 3.6887699564586356,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036879789981705954,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22777147591114044,
      "num_tokens": 84391034.0,
      "step": 40665
    },
    {
      "entropy": 5.713764095306397,
      "epoch": 3.68922351233672,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036876848809977404,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2094938188791275,
      "num_tokens": 84401452.0,
      "step": 40670
    },
    {
      "entropy": 5.621707773208618,
      "epoch": 3.689677068214804,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036873907444333234,
      "loss": 4.8322,
      "mean_token_accuracy": 0.22599519193172454,
      "num_tokens": 84411126.0,
      "step": 40675
    },
    {
      "entropy": 5.659043836593628,
      "epoch": 3.6901306240928884,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036870965884834274,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22305284291505814,
      "num_tokens": 84422740.0,
      "step": 40680
    },
    {
      "entropy": 5.653793621063232,
      "epoch": 3.6905841799709727,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00036868024131541337,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2174427926540375,
      "num_tokens": 84432831.0,
      "step": 40685
    },
    {
      "entropy": 5.703464508056641,
      "epoch": 3.6910377358490565,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036865082184515296,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22220373898744583,
      "num_tokens": 84443176.0,
      "step": 40690
    },
    {
      "entropy": 5.716496276855469,
      "epoch": 3.6914912917271407,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036862140043816957,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21485623866319656,
      "num_tokens": 84453185.0,
      "step": 40695
    },
    {
      "entropy": 5.6939490795135494,
      "epoch": 3.691944847605225,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003685919770950718,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22202895432710648,
      "num_tokens": 84464303.0,
      "step": 40700
    },
    {
      "entropy": 5.671566581726074,
      "epoch": 3.6923984034833093,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036856255181646797,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2199340984225273,
      "num_tokens": 84474463.0,
      "step": 40705
    },
    {
      "entropy": 5.623949384689331,
      "epoch": 3.692851959361393,
      "grad_norm": 1.25,
      "learning_rate": 0.0003685331246029668,
      "loss": 4.8934,
      "mean_token_accuracy": 0.22054630070924758,
      "num_tokens": 84484682.0,
      "step": 40710
    },
    {
      "entropy": 5.663768720626831,
      "epoch": 3.6933055152394774,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036850369545517685,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2293693780899048,
      "num_tokens": 84494830.0,
      "step": 40715
    },
    {
      "entropy": 5.617914724349975,
      "epoch": 3.6937590711175616,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003684742643737065,
      "loss": 4.8379,
      "mean_token_accuracy": 0.2238822638988495,
      "num_tokens": 84505449.0,
      "step": 40720
    },
    {
      "entropy": 5.698879337310791,
      "epoch": 3.694212626995646,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036844483135916467,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21640323400497435,
      "num_tokens": 84515833.0,
      "step": 40725
    },
    {
      "entropy": 5.69665560722351,
      "epoch": 3.69466618287373,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003684153964121599,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21794192343950272,
      "num_tokens": 84526677.0,
      "step": 40730
    },
    {
      "entropy": 5.642237329483033,
      "epoch": 3.6951197387518144,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036838595953330095,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22644708156585694,
      "num_tokens": 84536656.0,
      "step": 40735
    },
    {
      "entropy": 5.6743004322052,
      "epoch": 3.6955732946298983,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036835652072319646,
      "loss": 4.8349,
      "mean_token_accuracy": 0.22813600599765776,
      "num_tokens": 84546506.0,
      "step": 40740
    },
    {
      "entropy": 5.677061223983765,
      "epoch": 3.6960268505079825,
      "grad_norm": 1.25,
      "learning_rate": 0.0003683270799824554,
      "loss": 4.9527,
      "mean_token_accuracy": 0.22106551826000215,
      "num_tokens": 84556921.0,
      "step": 40745
    },
    {
      "entropy": 5.6613450050354,
      "epoch": 3.696480406386067,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003682976373116865,
      "loss": 4.886,
      "mean_token_accuracy": 0.2261745363473892,
      "num_tokens": 84566325.0,
      "step": 40750
    },
    {
      "entropy": 5.607352590560913,
      "epoch": 3.696933962264151,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036826819271149877,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22172658294439315,
      "num_tokens": 84577167.0,
      "step": 40755
    },
    {
      "entropy": 5.617634248733521,
      "epoch": 3.697387518142235,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036823874618250105,
      "loss": 4.8654,
      "mean_token_accuracy": 0.21913614869117737,
      "num_tokens": 84586764.0,
      "step": 40760
    },
    {
      "entropy": 5.60825924873352,
      "epoch": 3.697841074020319,
      "grad_norm": 1.25,
      "learning_rate": 0.00036820929772530224,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2206513375043869,
      "num_tokens": 84598903.0,
      "step": 40765
    },
    {
      "entropy": 5.630033016204834,
      "epoch": 3.6982946298984034,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003681798473405114,
      "loss": 4.8388,
      "mean_token_accuracy": 0.22656822353601455,
      "num_tokens": 84608695.0,
      "step": 40770
    },
    {
      "entropy": 5.584740781784058,
      "epoch": 3.6987481857764877,
      "grad_norm": 1.25,
      "learning_rate": 0.00036815039502873756,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22289512306451797,
      "num_tokens": 84619377.0,
      "step": 40775
    },
    {
      "entropy": 5.61126298904419,
      "epoch": 3.699201741654572,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036812094079058985,
      "loss": 4.8394,
      "mean_token_accuracy": 0.21988830715417862,
      "num_tokens": 84629716.0,
      "step": 40780
    },
    {
      "entropy": 5.5905218601226805,
      "epoch": 3.699655297532656,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003680914846266773,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23450044095516204,
      "num_tokens": 84640214.0,
      "step": 40785
    },
    {
      "entropy": 5.663218450546265,
      "epoch": 3.7001088534107405,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003680620265376092,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21970319896936416,
      "num_tokens": 84650403.0,
      "step": 40790
    },
    {
      "entropy": 5.652617311477661,
      "epoch": 3.7005624092888243,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036803256652399456,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22251996397972107,
      "num_tokens": 84660912.0,
      "step": 40795
    },
    {
      "entropy": 5.661609745025634,
      "epoch": 3.7010159651669086,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036800310458644285,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21118397265672684,
      "num_tokens": 84672194.0,
      "step": 40800
    },
    {
      "entropy": 5.681054401397705,
      "epoch": 3.701469521044993,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036797364072556316,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21239921152591706,
      "num_tokens": 84681850.0,
      "step": 40805
    },
    {
      "entropy": 5.685785484313965,
      "epoch": 3.7019230769230766,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000367944174941965,
      "loss": 4.8975,
      "mean_token_accuracy": 0.2270340785384178,
      "num_tokens": 84691271.0,
      "step": 40810
    },
    {
      "entropy": 5.606936073303222,
      "epoch": 3.702376632801161,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036791470723625746,
      "loss": 4.8835,
      "mean_token_accuracy": 0.21239540725946426,
      "num_tokens": 84702189.0,
      "step": 40815
    },
    {
      "entropy": 5.6548752784729,
      "epoch": 3.702830188679245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003678852376090502,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23135502338409425,
      "num_tokens": 84712399.0,
      "step": 40820
    },
    {
      "entropy": 5.691664934158325,
      "epoch": 3.7032837445573294,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003678557660609525,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21444610953330995,
      "num_tokens": 84723174.0,
      "step": 40825
    },
    {
      "entropy": 5.6580400466918945,
      "epoch": 3.7037373004354137,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000367826292592574,
      "loss": 4.8737,
      "mean_token_accuracy": 0.230885548889637,
      "num_tokens": 84732498.0,
      "step": 40830
    },
    {
      "entropy": 5.679837846755982,
      "epoch": 3.704190856313498,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003677968172045241,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2171540766954422,
      "num_tokens": 84742362.0,
      "step": 40835
    },
    {
      "entropy": 5.642715358734131,
      "epoch": 3.7046444121915822,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003677673398974122,
      "loss": 4.904,
      "mean_token_accuracy": 0.22081730961799623,
      "num_tokens": 84753379.0,
      "step": 40840
    },
    {
      "entropy": 5.69362325668335,
      "epoch": 3.705097968069666,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003677378606718482,
      "loss": 4.9904,
      "mean_token_accuracy": 0.2137149155139923,
      "num_tokens": 84763245.0,
      "step": 40845
    },
    {
      "entropy": 5.667326545715332,
      "epoch": 3.7055515239477503,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003677083795284416,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21358240991830826,
      "num_tokens": 84773929.0,
      "step": 40850
    },
    {
      "entropy": 5.728064107894897,
      "epoch": 3.7060050798258346,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036767889646780204,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21249008625745774,
      "num_tokens": 84784733.0,
      "step": 40855
    },
    {
      "entropy": 5.663216829299927,
      "epoch": 3.706458635703919,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003676494114905394,
      "loss": 4.8878,
      "mean_token_accuracy": 0.21417641639709473,
      "num_tokens": 84794254.0,
      "step": 40860
    },
    {
      "entropy": 5.705105829238891,
      "epoch": 3.7069121915820027,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036761992459726325,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21380765736103058,
      "num_tokens": 84804361.0,
      "step": 40865
    },
    {
      "entropy": 5.648979663848877,
      "epoch": 3.707365747460087,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036759043578858344,
      "loss": 4.89,
      "mean_token_accuracy": 0.2239452138543129,
      "num_tokens": 84814497.0,
      "step": 40870
    },
    {
      "entropy": 5.688341093063355,
      "epoch": 3.7078193033381712,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036756094506510984,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22103255838155747,
      "num_tokens": 84826871.0,
      "step": 40875
    },
    {
      "entropy": 5.666487121582032,
      "epoch": 3.7082728592162555,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036753145242745226,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22331540137529374,
      "num_tokens": 84837053.0,
      "step": 40880
    },
    {
      "entropy": 5.564202928543091,
      "epoch": 3.7087264150943398,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003675019578762208,
      "loss": 4.8367,
      "mean_token_accuracy": 0.2255967691540718,
      "num_tokens": 84846681.0,
      "step": 40885
    },
    {
      "entropy": 5.595720052719116,
      "epoch": 3.709179970972424,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003674724614120251,
      "loss": 4.8211,
      "mean_token_accuracy": 0.2218451529741287,
      "num_tokens": 84856651.0,
      "step": 40890
    },
    {
      "entropy": 5.6585493087768555,
      "epoch": 3.709633526850508,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036744296303547547,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21268591433763503,
      "num_tokens": 84867531.0,
      "step": 40895
    },
    {
      "entropy": 5.624187850952149,
      "epoch": 3.710087082728592,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003674134627471818,
      "loss": 4.8777,
      "mean_token_accuracy": 0.2233106791973114,
      "num_tokens": 84879132.0,
      "step": 40900
    },
    {
      "entropy": 5.660195302963257,
      "epoch": 3.7105406386066764,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000367383960547754,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22991647124290465,
      "num_tokens": 84888873.0,
      "step": 40905
    },
    {
      "entropy": 5.641703701019287,
      "epoch": 3.7109941944847606,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036735445643780256,
      "loss": 4.928,
      "mean_token_accuracy": 0.21705437898635865,
      "num_tokens": 84898926.0,
      "step": 40910
    },
    {
      "entropy": 5.639666366577148,
      "epoch": 3.7114477503628445,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003673249504179374,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22094259560108184,
      "num_tokens": 84908502.0,
      "step": 40915
    },
    {
      "entropy": 5.652897644042969,
      "epoch": 3.7119013062409287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003672954424887687,
      "loss": 4.904,
      "mean_token_accuracy": 0.2173845425248146,
      "num_tokens": 84919681.0,
      "step": 40920
    },
    {
      "entropy": 5.5717450141906735,
      "epoch": 3.712354862119013,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003672659326509066,
      "loss": 4.8226,
      "mean_token_accuracy": 0.22530381232500077,
      "num_tokens": 84930149.0,
      "step": 40925
    },
    {
      "entropy": 5.5976927280426025,
      "epoch": 3.7128084179970973,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036723642090496166,
      "loss": 4.8561,
      "mean_token_accuracy": 0.2250242367386818,
      "num_tokens": 84940086.0,
      "step": 40930
    },
    {
      "entropy": 5.705440855026245,
      "epoch": 3.7132619738751815,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036720690725154404,
      "loss": 4.933,
      "mean_token_accuracy": 0.22080663591623306,
      "num_tokens": 84949248.0,
      "step": 40935
    },
    {
      "entropy": 5.633121633529663,
      "epoch": 3.713715529753266,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036717739169126395,
      "loss": 4.8604,
      "mean_token_accuracy": 0.22151263058185577,
      "num_tokens": 84960892.0,
      "step": 40940
    },
    {
      "entropy": 5.641955423355102,
      "epoch": 3.7141690856313496,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036714787422473206,
      "loss": 4.892,
      "mean_token_accuracy": 0.2165308937430382,
      "num_tokens": 84971024.0,
      "step": 40945
    },
    {
      "entropy": 5.598126220703125,
      "epoch": 3.714622641509434,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003671183548525586,
      "loss": 4.8162,
      "mean_token_accuracy": 0.22717070132493972,
      "num_tokens": 84980945.0,
      "step": 40950
    },
    {
      "entropy": 5.671842193603515,
      "epoch": 3.715076197387518,
      "grad_norm": 1.25,
      "learning_rate": 0.000367088833575354,
      "loss": 4.9469,
      "mean_token_accuracy": 0.21502626687288284,
      "num_tokens": 84992055.0,
      "step": 40955
    },
    {
      "entropy": 5.67706332206726,
      "epoch": 3.7155297532656024,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036705931039372904,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22320076823234558,
      "num_tokens": 85002619.0,
      "step": 40960
    },
    {
      "entropy": 5.673063468933106,
      "epoch": 3.7159833091436862,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003670297853082939,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22195317447185517,
      "num_tokens": 85013317.0,
      "step": 40965
    },
    {
      "entropy": 5.670144891738891,
      "epoch": 3.7164368650217705,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036700025831965937,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22089784741401672,
      "num_tokens": 85023457.0,
      "step": 40970
    },
    {
      "entropy": 5.651346683502197,
      "epoch": 3.7168904208998548,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036697072942843616,
      "loss": 4.8801,
      "mean_token_accuracy": 0.21573939472436904,
      "num_tokens": 85033498.0,
      "step": 40975
    },
    {
      "entropy": 5.76431360244751,
      "epoch": 3.717343976777939,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003669411986352348,
      "loss": 5.0889,
      "mean_token_accuracy": 0.21457834392786027,
      "num_tokens": 85043687.0,
      "step": 40980
    },
    {
      "entropy": 5.689324522018433,
      "epoch": 3.7177975326560233,
      "grad_norm": 1.0625,
      "learning_rate": 0.000366911665940666,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21958806365728378,
      "num_tokens": 85055045.0,
      "step": 40985
    },
    {
      "entropy": 5.653619813919067,
      "epoch": 3.7182510885341076,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036688213134534044,
      "loss": 4.8518,
      "mean_token_accuracy": 0.23805569857358932,
      "num_tokens": 85064625.0,
      "step": 40990
    },
    {
      "entropy": 5.739334964752198,
      "epoch": 3.718704644412192,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003668525948498691,
      "loss": 5.0565,
      "mean_token_accuracy": 0.20217333137989044,
      "num_tokens": 85075249.0,
      "step": 40995
    },
    {
      "entropy": 5.657452583312988,
      "epoch": 3.7191582002902757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036682305645486266,
      "loss": 4.896,
      "mean_token_accuracy": 0.22502252757549285,
      "num_tokens": 85085900.0,
      "step": 41000
    },
    {
      "entropy": 5.664230537414551,
      "epoch": 3.71961175616836,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036679351616093194,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22187916785478592,
      "num_tokens": 85095772.0,
      "step": 41005
    },
    {
      "entropy": 5.638325023651123,
      "epoch": 3.720065312046444,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036676397396868797,
      "loss": 4.8538,
      "mean_token_accuracy": 0.22622692584991455,
      "num_tokens": 85105653.0,
      "step": 41010
    },
    {
      "entropy": 5.716484880447387,
      "epoch": 3.7205188679245285,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003667344298787417,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2144452527165413,
      "num_tokens": 85115432.0,
      "step": 41015
    },
    {
      "entropy": 5.729081010818481,
      "epoch": 3.7209724238026123,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036670488389170395,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20987560600042343,
      "num_tokens": 85126572.0,
      "step": 41020
    },
    {
      "entropy": 5.587710952758789,
      "epoch": 3.7214259796806965,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036667533600818586,
      "loss": 4.844,
      "mean_token_accuracy": 0.22681321948766708,
      "num_tokens": 85136809.0,
      "step": 41025
    },
    {
      "entropy": 5.850367546081543,
      "epoch": 3.721879535558781,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003666457862287985,
      "loss": 5.0696,
      "mean_token_accuracy": 0.21007951498031616,
      "num_tokens": 85148346.0,
      "step": 41030
    },
    {
      "entropy": 5.661215829849243,
      "epoch": 3.722333091436865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036661623455415296,
      "loss": 4.8964,
      "mean_token_accuracy": 0.21990771293640138,
      "num_tokens": 85158807.0,
      "step": 41035
    },
    {
      "entropy": 5.688000917434692,
      "epoch": 3.7227866473149493,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003665866809848602,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20994648486375808,
      "num_tokens": 85170752.0,
      "step": 41040
    },
    {
      "entropy": 5.625526237487793,
      "epoch": 3.7232402031930336,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036655712552153164,
      "loss": 4.9025,
      "mean_token_accuracy": 0.21547161787748337,
      "num_tokens": 85180779.0,
      "step": 41045
    },
    {
      "entropy": 5.685053062438965,
      "epoch": 3.7236937590711174,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003665275681647784,
      "loss": 4.8748,
      "mean_token_accuracy": 0.22153379768133163,
      "num_tokens": 85191391.0,
      "step": 41050
    },
    {
      "entropy": 5.693622303009033,
      "epoch": 3.7241473149492017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003664980089152116,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2162279024720192,
      "num_tokens": 85201357.0,
      "step": 41055
    },
    {
      "entropy": 5.6578082084655765,
      "epoch": 3.724600870827286,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036646844777344284,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2227029785513878,
      "num_tokens": 85212193.0,
      "step": 41060
    },
    {
      "entropy": 5.664473676681519,
      "epoch": 3.7250544267053702,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036643888474008313,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2193174734711647,
      "num_tokens": 85223809.0,
      "step": 41065
    },
    {
      "entropy": 5.651267194747925,
      "epoch": 3.725507982583454,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003664093198157441,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22259458750486374,
      "num_tokens": 85234992.0,
      "step": 41070
    },
    {
      "entropy": 5.728595781326294,
      "epoch": 3.7259615384615383,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003663797530010369,
      "loss": 4.9597,
      "mean_token_accuracy": 0.2119491770863533,
      "num_tokens": 85244157.0,
      "step": 41075
    },
    {
      "entropy": 5.613103628158569,
      "epoch": 3.7264150943396226,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003663501842965732,
      "loss": 4.867,
      "mean_token_accuracy": 0.22966155111789704,
      "num_tokens": 85255946.0,
      "step": 41080
    },
    {
      "entropy": 5.610699987411499,
      "epoch": 3.726868650217707,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003663206137029645,
      "loss": 4.868,
      "mean_token_accuracy": 0.22856574654579162,
      "num_tokens": 85265273.0,
      "step": 41085
    },
    {
      "entropy": 5.573185014724731,
      "epoch": 3.727322206095791,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036629104122082215,
      "loss": 4.7211,
      "mean_token_accuracy": 0.2303519532084465,
      "num_tokens": 85275794.0,
      "step": 41090
    },
    {
      "entropy": 5.6412163257598875,
      "epoch": 3.7277757619738754,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036626146685075775,
      "loss": 4.9906,
      "mean_token_accuracy": 0.2108025535941124,
      "num_tokens": 85287334.0,
      "step": 41095
    },
    {
      "entropy": 5.67722520828247,
      "epoch": 3.728229317851959,
      "grad_norm": 1.3125,
      "learning_rate": 0.000366231890593383,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2216177687048912,
      "num_tokens": 85297320.0,
      "step": 41100
    },
    {
      "entropy": 5.601094150543213,
      "epoch": 3.7286828737300435,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003662023124493095,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2279417484998703,
      "num_tokens": 85308071.0,
      "step": 41105
    },
    {
      "entropy": 5.68747844696045,
      "epoch": 3.7291364296081277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000366172732419149,
      "loss": 4.9257,
      "mean_token_accuracy": 0.21675894409418106,
      "num_tokens": 85317790.0,
      "step": 41110
    },
    {
      "entropy": 5.7154566764831545,
      "epoch": 3.729589985486212,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036614315050351317,
      "loss": 4.984,
      "mean_token_accuracy": 0.21242806017398835,
      "num_tokens": 85329008.0,
      "step": 41115
    },
    {
      "entropy": 5.7604231357574465,
      "epoch": 3.730043541364296,
      "grad_norm": 1.25,
      "learning_rate": 0.0003661135667030137,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21173521876335144,
      "num_tokens": 85339095.0,
      "step": 41120
    },
    {
      "entropy": 5.699007987976074,
      "epoch": 3.73049709724238,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003660839810182624,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21794274002313613,
      "num_tokens": 85349833.0,
      "step": 41125
    },
    {
      "entropy": 5.721101188659668,
      "epoch": 3.7309506531204644,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036605439344987125,
      "loss": 4.9711,
      "mean_token_accuracy": 0.21819814294576645,
      "num_tokens": 85359512.0,
      "step": 41130
    },
    {
      "entropy": 5.70348858833313,
      "epoch": 3.7314042089985486,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000366024803998452,
      "loss": 4.9481,
      "mean_token_accuracy": 0.224931900203228,
      "num_tokens": 85370053.0,
      "step": 41135
    },
    {
      "entropy": 5.659979963302613,
      "epoch": 3.731857764876633,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036599521266461653,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22135833352804185,
      "num_tokens": 85380020.0,
      "step": 41140
    },
    {
      "entropy": 5.663234090805053,
      "epoch": 3.732311320754717,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036596561944897703,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22230454683303832,
      "num_tokens": 85390232.0,
      "step": 41145
    },
    {
      "entropy": 5.6056296825408936,
      "epoch": 3.7327648766328014,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003659360243521453,
      "loss": 4.8243,
      "mean_token_accuracy": 0.23131534904241563,
      "num_tokens": 85400401.0,
      "step": 41150
    },
    {
      "entropy": 5.6432029724121096,
      "epoch": 3.7332184325108853,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003659064273747334,
      "loss": 4.9306,
      "mean_token_accuracy": 0.21713484227657318,
      "num_tokens": 85411749.0,
      "step": 41155
    },
    {
      "entropy": 5.706897354125976,
      "epoch": 3.7336719883889695,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036587682851735336,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2157077744603157,
      "num_tokens": 85424065.0,
      "step": 41160
    },
    {
      "entropy": 5.707071161270141,
      "epoch": 3.734125544267054,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036584722778061743,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22047340571880342,
      "num_tokens": 85434412.0,
      "step": 41165
    },
    {
      "entropy": 5.6794806003570555,
      "epoch": 3.7345791001451376,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003658176251651376,
      "loss": 4.9031,
      "mean_token_accuracy": 0.2253032073378563,
      "num_tokens": 85444489.0,
      "step": 41170
    },
    {
      "entropy": 5.6987415790557865,
      "epoch": 3.735032656023222,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036578802067152615,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22232849448919295,
      "num_tokens": 85454556.0,
      "step": 41175
    },
    {
      "entropy": 5.59938268661499,
      "epoch": 3.735486211901306,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003657584143003954,
      "loss": 4.8317,
      "mean_token_accuracy": 0.2245876133441925,
      "num_tokens": 85464701.0,
      "step": 41180
    },
    {
      "entropy": 5.653997182846069,
      "epoch": 3.7359397677793904,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003657288060523574,
      "loss": 4.9656,
      "mean_token_accuracy": 0.22018109858036042,
      "num_tokens": 85475657.0,
      "step": 41185
    },
    {
      "entropy": 5.709797286987305,
      "epoch": 3.7363933236574747,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036569919592802466,
      "loss": 4.8963,
      "mean_token_accuracy": 0.22522761374711991,
      "num_tokens": 85486358.0,
      "step": 41190
    },
    {
      "entropy": 5.742954969406128,
      "epoch": 3.736846879535559,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003656695839280093,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21693163514137268,
      "num_tokens": 85497913.0,
      "step": 41195
    },
    {
      "entropy": 5.653684663772583,
      "epoch": 3.737300435413643,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036563997005292396,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22670360803604125,
      "num_tokens": 85508314.0,
      "step": 41200
    },
    {
      "entropy": 5.603851556777954,
      "epoch": 3.737753991291727,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036561035430338084,
      "loss": 4.8967,
      "mean_token_accuracy": 0.21857466250658036,
      "num_tokens": 85518142.0,
      "step": 41205
    },
    {
      "entropy": 5.656486845016479,
      "epoch": 3.7382075471698113,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036558073667999257,
      "loss": 4.784,
      "mean_token_accuracy": 0.22903891950845717,
      "num_tokens": 85526854.0,
      "step": 41210
    },
    {
      "entropy": 5.594369220733642,
      "epoch": 3.7386611030478956,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036555111718337166,
      "loss": 4.7946,
      "mean_token_accuracy": 0.2322963684797287,
      "num_tokens": 85537153.0,
      "step": 41215
    },
    {
      "entropy": 5.708690786361695,
      "epoch": 3.73911465892598,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003655214958141305,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21064232140779496,
      "num_tokens": 85548583.0,
      "step": 41220
    },
    {
      "entropy": 5.6763382911682125,
      "epoch": 3.7395682148040637,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036549187257288164,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2300586834549904,
      "num_tokens": 85559228.0,
      "step": 41225
    },
    {
      "entropy": 5.707980489730835,
      "epoch": 3.740021770682148,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036546224746023795,
      "loss": 4.8709,
      "mean_token_accuracy": 0.22223663926124573,
      "num_tokens": 85569757.0,
      "step": 41230
    },
    {
      "entropy": 5.649372529983521,
      "epoch": 3.740475326560232,
      "grad_norm": 1.125,
      "learning_rate": 0.00036543262047681186,
      "loss": 4.8983,
      "mean_token_accuracy": 0.23110807985067366,
      "num_tokens": 85582013.0,
      "step": 41235
    },
    {
      "entropy": 5.652613115310669,
      "epoch": 3.7409288824383164,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003654029916232161,
      "loss": 4.8869,
      "mean_token_accuracy": 0.2358803018927574,
      "num_tokens": 85594049.0,
      "step": 41240
    },
    {
      "entropy": 5.665416145324707,
      "epoch": 3.7413824383164007,
      "grad_norm": 1.421875,
      "learning_rate": 0.00036537336090006347,
      "loss": 4.922,
      "mean_token_accuracy": 0.21921512335538865,
      "num_tokens": 85605253.0,
      "step": 41245
    },
    {
      "entropy": 5.620599222183228,
      "epoch": 3.741835994194485,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036534372830796674,
      "loss": 4.8463,
      "mean_token_accuracy": 0.22598191499710082,
      "num_tokens": 85615032.0,
      "step": 41250
    },
    {
      "entropy": 5.65076756477356,
      "epoch": 3.742289550072569,
      "grad_norm": 1.25,
      "learning_rate": 0.0003653140938475386,
      "loss": 4.881,
      "mean_token_accuracy": 0.23063501119613647,
      "num_tokens": 85625497.0,
      "step": 41255
    },
    {
      "entropy": 5.659329271316528,
      "epoch": 3.742743105950653,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036528445751939206,
      "loss": 4.9308,
      "mean_token_accuracy": 0.21988602578639985,
      "num_tokens": 85636058.0,
      "step": 41260
    },
    {
      "entropy": 5.658809995651245,
      "epoch": 3.7431966618287373,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036525481932413997,
      "loss": 4.918,
      "mean_token_accuracy": 0.216049624979496,
      "num_tokens": 85646505.0,
      "step": 41265
    },
    {
      "entropy": 5.718615961074829,
      "epoch": 3.7436502177068216,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036522517926239513,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2189261719584465,
      "num_tokens": 85656864.0,
      "step": 41270
    },
    {
      "entropy": 5.6087257862091064,
      "epoch": 3.7441037735849054,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036519553733477055,
      "loss": 4.8361,
      "mean_token_accuracy": 0.2275415152311325,
      "num_tokens": 85666783.0,
      "step": 41275
    },
    {
      "entropy": 5.552579784393311,
      "epoch": 3.7445573294629897,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003651658935418794,
      "loss": 4.7818,
      "mean_token_accuracy": 0.2315801203250885,
      "num_tokens": 85677102.0,
      "step": 41280
    },
    {
      "entropy": 5.592778873443604,
      "epoch": 3.745010885341074,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036513624788433454,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23277545720338821,
      "num_tokens": 85687048.0,
      "step": 41285
    },
    {
      "entropy": 5.626360464096069,
      "epoch": 3.7454644412191582,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003651066003627491,
      "loss": 4.8936,
      "mean_token_accuracy": 0.21562793254852294,
      "num_tokens": 85698314.0,
      "step": 41290
    },
    {
      "entropy": 5.711394214630127,
      "epoch": 3.7459179970972425,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003650769509777362,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21696987748146057,
      "num_tokens": 85708039.0,
      "step": 41295
    },
    {
      "entropy": 5.645800971984864,
      "epoch": 3.7463715529753268,
      "grad_norm": 1.484375,
      "learning_rate": 0.00036504729972990904,
      "loss": 4.8097,
      "mean_token_accuracy": 0.23273114264011383,
      "num_tokens": 85717741.0,
      "step": 41300
    },
    {
      "entropy": 5.683948278427124,
      "epoch": 3.7468251088534106,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003650176466198807,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22189702838659286,
      "num_tokens": 85728286.0,
      "step": 41305
    },
    {
      "entropy": 5.649855327606201,
      "epoch": 3.747278664731495,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036498799164826453,
      "loss": 4.8684,
      "mean_token_accuracy": 0.224978768825531,
      "num_tokens": 85738970.0,
      "step": 41310
    },
    {
      "entropy": 5.607178974151611,
      "epoch": 3.747732220609579,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036495833481567375,
      "loss": 4.8389,
      "mean_token_accuracy": 0.22328358590602876,
      "num_tokens": 85749284.0,
      "step": 41315
    },
    {
      "entropy": 5.694136428833008,
      "epoch": 3.7481857764876634,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036492867612272176,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2162918210029602,
      "num_tokens": 85759595.0,
      "step": 41320
    },
    {
      "entropy": 5.728177928924561,
      "epoch": 3.748639332365747,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003648990155700217,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21251268386840821,
      "num_tokens": 85771049.0,
      "step": 41325
    },
    {
      "entropy": 5.670376825332641,
      "epoch": 3.7490928882438315,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036486935315818713,
      "loss": 4.8796,
      "mean_token_accuracy": 0.2176561549305916,
      "num_tokens": 85781656.0,
      "step": 41330
    },
    {
      "entropy": 5.579206848144532,
      "epoch": 3.7495464441219157,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036483968888783154,
      "loss": 4.7526,
      "mean_token_accuracy": 0.2277541771531105,
      "num_tokens": 85790922.0,
      "step": 41335
    },
    {
      "entropy": 5.634003162384033,
      "epoch": 3.75,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036481002275956813,
      "loss": 4.8398,
      "mean_token_accuracy": 0.22941983342170716,
      "num_tokens": 85799928.0,
      "step": 41340
    },
    {
      "entropy": 5.667712116241455,
      "epoch": 3.7504535558780843,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003647803547740106,
      "loss": 4.943,
      "mean_token_accuracy": 0.21469739377498626,
      "num_tokens": 85810594.0,
      "step": 41345
    },
    {
      "entropy": 5.636851453781128,
      "epoch": 3.7509071117561685,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036475068493177254,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21709463149309158,
      "num_tokens": 85820690.0,
      "step": 41350
    },
    {
      "entropy": 5.6454047679901125,
      "epoch": 3.751360667634253,
      "grad_norm": 1.25,
      "learning_rate": 0.0003647210132334674,
      "loss": 4.851,
      "mean_token_accuracy": 0.22203511595726014,
      "num_tokens": 85830314.0,
      "step": 41355
    },
    {
      "entropy": 5.652485418319702,
      "epoch": 3.7518142235123366,
      "grad_norm": 1.25,
      "learning_rate": 0.0003646913396797088,
      "loss": 4.9617,
      "mean_token_accuracy": 0.20690085887908935,
      "num_tokens": 85841099.0,
      "step": 41360
    },
    {
      "entropy": 5.707526826858521,
      "epoch": 3.752267779390421,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036466166427111046,
      "loss": 4.8923,
      "mean_token_accuracy": 0.2213337689638138,
      "num_tokens": 85852813.0,
      "step": 41365
    },
    {
      "entropy": 5.665508699417114,
      "epoch": 3.752721335268505,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036463198700828616,
      "loss": 4.8382,
      "mean_token_accuracy": 0.22168927043676376,
      "num_tokens": 85862778.0,
      "step": 41370
    },
    {
      "entropy": 5.616978883743286,
      "epoch": 3.7531748911465894,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003646023078918494,
      "loss": 4.921,
      "mean_token_accuracy": 0.222724349796772,
      "num_tokens": 85872729.0,
      "step": 41375
    },
    {
      "entropy": 5.659740447998047,
      "epoch": 3.7536284470246732,
      "grad_norm": 1.25,
      "learning_rate": 0.0003645726269224141,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22810726314783097,
      "num_tokens": 85883289.0,
      "step": 41380
    },
    {
      "entropy": 5.580375623703003,
      "epoch": 3.7540820029027575,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036454294410059407,
      "loss": 4.8119,
      "mean_token_accuracy": 0.22975561618804932,
      "num_tokens": 85892318.0,
      "step": 41385
    },
    {
      "entropy": 5.601224660873413,
      "epoch": 3.7545355587808418,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003645132594270032,
      "loss": 4.8727,
      "mean_token_accuracy": 0.22175518572330474,
      "num_tokens": 85902719.0,
      "step": 41390
    },
    {
      "entropy": 5.624606275558472,
      "epoch": 3.754989114658926,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036448357290225523,
      "loss": 4.8629,
      "mean_token_accuracy": 0.21994414627552034,
      "num_tokens": 85913271.0,
      "step": 41395
    },
    {
      "entropy": 5.588603734970093,
      "epoch": 3.7554426705370103,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003644538845269642,
      "loss": 4.7926,
      "mean_token_accuracy": 0.2242666721343994,
      "num_tokens": 85922397.0,
      "step": 41400
    },
    {
      "entropy": 5.67548303604126,
      "epoch": 3.7558962264150946,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000364424194301744,
      "loss": 4.928,
      "mean_token_accuracy": 0.2190377652645111,
      "num_tokens": 85933261.0,
      "step": 41405
    },
    {
      "entropy": 5.698658609390259,
      "epoch": 3.7563497822931784,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003643945022272087,
      "loss": 4.9563,
      "mean_token_accuracy": 0.221746763586998,
      "num_tokens": 85943860.0,
      "step": 41410
    },
    {
      "entropy": 5.687657260894776,
      "epoch": 3.7568033381712627,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003643648083039724,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2167433977127075,
      "num_tokens": 85953984.0,
      "step": 41415
    },
    {
      "entropy": 5.673469400405883,
      "epoch": 3.757256894049347,
      "grad_norm": 1.328125,
      "learning_rate": 0.000364335112532649,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2184562787413597,
      "num_tokens": 85964771.0,
      "step": 41420
    },
    {
      "entropy": 5.677586269378662,
      "epoch": 3.757710449927431,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003643054149138527,
      "loss": 4.9413,
      "mean_token_accuracy": 0.22036302238702773,
      "num_tokens": 85975859.0,
      "step": 41425
    },
    {
      "entropy": 5.755876684188843,
      "epoch": 3.758164005805515,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036427571544819765,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2161107450723648,
      "num_tokens": 85987022.0,
      "step": 41430
    },
    {
      "entropy": 5.611931705474854,
      "epoch": 3.7586175616835993,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003642460141362981,
      "loss": 4.838,
      "mean_token_accuracy": 0.22265738248825073,
      "num_tokens": 85997056.0,
      "step": 41435
    },
    {
      "entropy": 5.672608661651611,
      "epoch": 3.7590711175616836,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036421631097876823,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21579169929027558,
      "num_tokens": 86008853.0,
      "step": 41440
    },
    {
      "entropy": 5.573115348815918,
      "epoch": 3.759524673439768,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003641866059762223,
      "loss": 4.9138,
      "mean_token_accuracy": 0.21726031601428986,
      "num_tokens": 86020358.0,
      "step": 41445
    },
    {
      "entropy": 5.662569046020508,
      "epoch": 3.759978229317852,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003641568991292747,
      "loss": 4.9219,
      "mean_token_accuracy": 0.21406012922525405,
      "num_tokens": 86030706.0,
      "step": 41450
    },
    {
      "entropy": 5.730230045318604,
      "epoch": 3.7604317851959363,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036412719043853956,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21748934388160707,
      "num_tokens": 86041964.0,
      "step": 41455
    },
    {
      "entropy": 5.690208196640015,
      "epoch": 3.76088534107402,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036409747990463146,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22287604659795762,
      "num_tokens": 86052213.0,
      "step": 41460
    },
    {
      "entropy": 5.619532918930053,
      "epoch": 3.7613388969521044,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003640677675281649,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21440352350473404,
      "num_tokens": 86061419.0,
      "step": 41465
    },
    {
      "entropy": 5.715872955322266,
      "epoch": 3.7617924528301887,
      "grad_norm": 1.375,
      "learning_rate": 0.0003640380533097541,
      "loss": 4.91,
      "mean_token_accuracy": 0.2179046869277954,
      "num_tokens": 86071585.0,
      "step": 41470
    },
    {
      "entropy": 5.674550628662109,
      "epoch": 3.762246008708273,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036400833725001364,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21819409877061843,
      "num_tokens": 86081175.0,
      "step": 41475
    },
    {
      "entropy": 5.658315801620484,
      "epoch": 3.762699564586357,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003639786193495581,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21548353731632233,
      "num_tokens": 86092407.0,
      "step": 41480
    },
    {
      "entropy": 5.723626661300659,
      "epoch": 3.763153120464441,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036394889960900215,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21736511886119841,
      "num_tokens": 86102218.0,
      "step": 41485
    },
    {
      "entropy": 5.734164619445801,
      "epoch": 3.7636066763425253,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003639191780289602,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22272278666496276,
      "num_tokens": 86112631.0,
      "step": 41490
    },
    {
      "entropy": 5.700323009490967,
      "epoch": 3.7640602322206096,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003638894546100469,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20971276760101318,
      "num_tokens": 86123675.0,
      "step": 41495
    },
    {
      "entropy": 5.590743112564087,
      "epoch": 3.764513788098694,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036385972935287715,
      "loss": 4.8425,
      "mean_token_accuracy": 0.222859163582325,
      "num_tokens": 86133909.0,
      "step": 41500
    },
    {
      "entropy": 5.65754976272583,
      "epoch": 3.764967343976778,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003638300022580656,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21580233424901962,
      "num_tokens": 86144560.0,
      "step": 41505
    },
    {
      "entropy": 5.755287218093872,
      "epoch": 3.7654208998548624,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036380027332622684,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21797743290662766,
      "num_tokens": 86155140.0,
      "step": 41510
    },
    {
      "entropy": 5.713537216186523,
      "epoch": 3.765874455732946,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00036377054255797584,
      "loss": 4.9351,
      "mean_token_accuracy": 0.2162645310163498,
      "num_tokens": 86165555.0,
      "step": 41515
    },
    {
      "entropy": 5.670260858535767,
      "epoch": 3.7663280116110305,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036374080995392737,
      "loss": 4.897,
      "mean_token_accuracy": 0.22103025317192077,
      "num_tokens": 86176204.0,
      "step": 41520
    },
    {
      "entropy": 5.681659317016601,
      "epoch": 3.7667815674891147,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003637110755146964,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2189554899930954,
      "num_tokens": 86186888.0,
      "step": 41525
    },
    {
      "entropy": 5.6775728225708,
      "epoch": 3.7672351233671986,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036368133924089766,
      "loss": 4.8651,
      "mean_token_accuracy": 0.22728827893733977,
      "num_tokens": 86196654.0,
      "step": 41530
    },
    {
      "entropy": 5.649123334884644,
      "epoch": 3.767688679245283,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003636516011331464,
      "loss": 4.8632,
      "mean_token_accuracy": 0.2250275954604149,
      "num_tokens": 86207045.0,
      "step": 41535
    },
    {
      "entropy": 5.693004417419433,
      "epoch": 3.768142235123367,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036362186119205734,
      "loss": 4.9469,
      "mean_token_accuracy": 0.21175480037927627,
      "num_tokens": 86216906.0,
      "step": 41540
    },
    {
      "entropy": 5.618576431274414,
      "epoch": 3.7685957910014514,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003635921194182457,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2223455309867859,
      "num_tokens": 86227380.0,
      "step": 41545
    },
    {
      "entropy": 5.6646894931793215,
      "epoch": 3.7690493468795356,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003635623758123263,
      "loss": 5.0194,
      "mean_token_accuracy": 0.22061097025871276,
      "num_tokens": 86236948.0,
      "step": 41550
    },
    {
      "entropy": 5.676881885528564,
      "epoch": 3.76950290275762,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036353263037491445,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21786512285470963,
      "num_tokens": 86246784.0,
      "step": 41555
    },
    {
      "entropy": 5.631447172164917,
      "epoch": 3.769956458635704,
      "grad_norm": 1.25,
      "learning_rate": 0.0003635028831066253,
      "loss": 4.837,
      "mean_token_accuracy": 0.2246741160750389,
      "num_tokens": 86257381.0,
      "step": 41560
    },
    {
      "entropy": 5.611772966384888,
      "epoch": 3.770410014513788,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003634731340080739,
      "loss": 4.85,
      "mean_token_accuracy": 0.22241034358739853,
      "num_tokens": 86268117.0,
      "step": 41565
    },
    {
      "entropy": 5.7618732929229735,
      "epoch": 3.7708635703918723,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003634433830798756,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21439319401979445,
      "num_tokens": 86279195.0,
      "step": 41570
    },
    {
      "entropy": 5.574959754943848,
      "epoch": 3.7713171262699565,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003634136303226457,
      "loss": 4.7418,
      "mean_token_accuracy": 0.23419877290725707,
      "num_tokens": 86288154.0,
      "step": 41575
    },
    {
      "entropy": 5.659147024154663,
      "epoch": 3.771770682148041,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003633838757369992,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21908959299325942,
      "num_tokens": 86297524.0,
      "step": 41580
    },
    {
      "entropy": 5.652503728866577,
      "epoch": 3.7722242380261246,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003633541193235517,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21118232756853103,
      "num_tokens": 86307031.0,
      "step": 41585
    },
    {
      "entropy": 5.637276840209961,
      "epoch": 3.772677793904209,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003633243610829185,
      "loss": 4.9387,
      "mean_token_accuracy": 0.21906437277793883,
      "num_tokens": 86317629.0,
      "step": 41590
    },
    {
      "entropy": 5.6781778812408445,
      "epoch": 3.773131349782293,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036329460101571496,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21658796221017837,
      "num_tokens": 86328243.0,
      "step": 41595
    },
    {
      "entropy": 5.757309007644653,
      "epoch": 3.7735849056603774,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003632648391225568,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2148237019777298,
      "num_tokens": 86338905.0,
      "step": 41600
    },
    {
      "entropy": 5.680874967575074,
      "epoch": 3.7740384615384617,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036323507540405903,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22035324573516846,
      "num_tokens": 86350433.0,
      "step": 41605
    },
    {
      "entropy": 5.6957090377807615,
      "epoch": 3.774492017416546,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003632053098608374,
      "loss": 4.9458,
      "mean_token_accuracy": 0.2112775757908821,
      "num_tokens": 86360664.0,
      "step": 41610
    },
    {
      "entropy": 5.683890295028687,
      "epoch": 3.7749455732946298,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036317554249350767,
      "loss": 4.972,
      "mean_token_accuracy": 0.22332140058279037,
      "num_tokens": 86372101.0,
      "step": 41615
    },
    {
      "entropy": 5.66197657585144,
      "epoch": 3.775399129172714,
      "grad_norm": 1.125,
      "learning_rate": 0.0003631457733026851,
      "loss": 4.817,
      "mean_token_accuracy": 0.2249531030654907,
      "num_tokens": 86382494.0,
      "step": 41620
    },
    {
      "entropy": 5.764450597763061,
      "epoch": 3.7758526850507983,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036311600228898554,
      "loss": 5.0635,
      "mean_token_accuracy": 0.20831287354230882,
      "num_tokens": 86393292.0,
      "step": 41625
    },
    {
      "entropy": 5.5949243068695065,
      "epoch": 3.7763062409288826,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036308622945302455,
      "loss": 4.837,
      "mean_token_accuracy": 0.21961106061935426,
      "num_tokens": 86402940.0,
      "step": 41630
    },
    {
      "entropy": 5.595632410049438,
      "epoch": 3.7767597968069664,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000363056454795418,
      "loss": 4.8665,
      "mean_token_accuracy": 0.21814990639686585,
      "num_tokens": 86413382.0,
      "step": 41635
    },
    {
      "entropy": 5.69773497581482,
      "epoch": 3.7772133526850507,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003630266783167814,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21711234897375106,
      "num_tokens": 86423866.0,
      "step": 41640
    },
    {
      "entropy": 5.7516913414001465,
      "epoch": 3.777666908563135,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003629969000177307,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21043222546577453,
      "num_tokens": 86434908.0,
      "step": 41645
    },
    {
      "entropy": 5.560062265396118,
      "epoch": 3.778120464441219,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003629671198988817,
      "loss": 4.8308,
      "mean_token_accuracy": 0.23035694360733033,
      "num_tokens": 86445604.0,
      "step": 41650
    },
    {
      "entropy": 5.58459701538086,
      "epoch": 3.7785740203193035,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003629373379608502,
      "loss": 4.7759,
      "mean_token_accuracy": 0.23566643595695497,
      "num_tokens": 86455813.0,
      "step": 41655
    },
    {
      "entropy": 5.559374570846558,
      "epoch": 3.7790275761973877,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036290755420425214,
      "loss": 4.7502,
      "mean_token_accuracy": 0.23715393990278244,
      "num_tokens": 86465256.0,
      "step": 41660
    },
    {
      "entropy": 5.628467178344726,
      "epoch": 3.7794811320754715,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036287776862970356,
      "loss": 4.8824,
      "mean_token_accuracy": 0.22413792759180068,
      "num_tokens": 86474605.0,
      "step": 41665
    },
    {
      "entropy": 5.671857595443726,
      "epoch": 3.779934687953556,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036284798123782024,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22032371461391448,
      "num_tokens": 86484886.0,
      "step": 41670
    },
    {
      "entropy": 5.624969005584717,
      "epoch": 3.78038824383164,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003628181920292182,
      "loss": 4.9809,
      "mean_token_accuracy": 0.21418100744485855,
      "num_tokens": 86495816.0,
      "step": 41675
    },
    {
      "entropy": 5.632235431671143,
      "epoch": 3.7808417997097243,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003627884010045138,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22076003402471542,
      "num_tokens": 86506319.0,
      "step": 41680
    },
    {
      "entropy": 5.716420412063599,
      "epoch": 3.781295355587808,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003627586081643227,
      "loss": 4.9059,
      "mean_token_accuracy": 0.21986618638038635,
      "num_tokens": 86516952.0,
      "step": 41685
    },
    {
      "entropy": 5.71786470413208,
      "epoch": 3.7817489114658924,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036272881350926135,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21247045546770096,
      "num_tokens": 86527327.0,
      "step": 41690
    },
    {
      "entropy": 5.6497101306915285,
      "epoch": 3.7822024673439767,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003626990170399457,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22317089140415192,
      "num_tokens": 86538281.0,
      "step": 41695
    },
    {
      "entropy": 5.675183296203613,
      "epoch": 3.782656023222061,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036266921875699205,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21008704006671905,
      "num_tokens": 86549034.0,
      "step": 41700
    },
    {
      "entropy": 5.6818126201629635,
      "epoch": 3.7831095791001452,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003626394186610166,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21852248311042785,
      "num_tokens": 86559306.0,
      "step": 41705
    },
    {
      "entropy": 5.730512619018555,
      "epoch": 3.7835631349782295,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003626096167526356,
      "loss": 4.9323,
      "mean_token_accuracy": 0.21690935641527176,
      "num_tokens": 86569414.0,
      "step": 41710
    },
    {
      "entropy": 5.557330083847046,
      "epoch": 3.7840166908563138,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003625798130324655,
      "loss": 4.7823,
      "mean_token_accuracy": 0.2280845656991005,
      "num_tokens": 86578554.0,
      "step": 41715
    },
    {
      "entropy": 5.6679039478302,
      "epoch": 3.7844702467343976,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036255000750112256,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2222454845905304,
      "num_tokens": 86588624.0,
      "step": 41720
    },
    {
      "entropy": 5.717407751083374,
      "epoch": 3.784923802612482,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036252020015922304,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21759973019361495,
      "num_tokens": 86599266.0,
      "step": 41725
    },
    {
      "entropy": 5.726531982421875,
      "epoch": 3.785377358490566,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036249039100738357,
      "loss": 4.9639,
      "mean_token_accuracy": 0.20948874354362487,
      "num_tokens": 86609853.0,
      "step": 41730
    },
    {
      "entropy": 5.58441195487976,
      "epoch": 3.7858309143686504,
      "grad_norm": 1.25,
      "learning_rate": 0.0003624605800462205,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2207888185977936,
      "num_tokens": 86619927.0,
      "step": 41735
    },
    {
      "entropy": 5.710479736328125,
      "epoch": 3.786284470246734,
      "grad_norm": 1.375,
      "learning_rate": 0.00036243076727635027,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2182391494512558,
      "num_tokens": 86630079.0,
      "step": 41740
    },
    {
      "entropy": 5.670481777191162,
      "epoch": 3.7867380261248185,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036240095269838955,
      "loss": 4.9022,
      "mean_token_accuracy": 0.22344246208667756,
      "num_tokens": 86639744.0,
      "step": 41745
    },
    {
      "entropy": 5.626803636550903,
      "epoch": 3.7871915820029027,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036237113631295485,
      "loss": 4.9574,
      "mean_token_accuracy": 0.21240945160388947,
      "num_tokens": 86650866.0,
      "step": 41750
    },
    {
      "entropy": 5.632354879379273,
      "epoch": 3.787645137880987,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036234131812066286,
      "loss": 4.8102,
      "mean_token_accuracy": 0.2288187637925148,
      "num_tokens": 86659864.0,
      "step": 41755
    },
    {
      "entropy": 5.583402919769287,
      "epoch": 3.7880986937590713,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036231149812213006,
      "loss": 4.8275,
      "mean_token_accuracy": 0.22766820192337037,
      "num_tokens": 86670039.0,
      "step": 41760
    },
    {
      "entropy": 5.62681941986084,
      "epoch": 3.7885522496371555,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003622816763179732,
      "loss": 4.8762,
      "mean_token_accuracy": 0.22567262798547744,
      "num_tokens": 86680434.0,
      "step": 41765
    },
    {
      "entropy": 5.676084518432617,
      "epoch": 3.7890058055152394,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003622518527088091,
      "loss": 4.909,
      "mean_token_accuracy": 0.21579646021127702,
      "num_tokens": 86691026.0,
      "step": 41770
    },
    {
      "entropy": 5.701051568984985,
      "epoch": 3.7894593613933236,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036222202729525453,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21668500155210496,
      "num_tokens": 86701323.0,
      "step": 41775
    },
    {
      "entropy": 5.6415246486663815,
      "epoch": 3.789912917271408,
      "grad_norm": 1.203125,
      "learning_rate": 0.000362192200077926,
      "loss": 4.852,
      "mean_token_accuracy": 0.23028887510299684,
      "num_tokens": 86711152.0,
      "step": 41780
    },
    {
      "entropy": 5.700291538238526,
      "epoch": 3.790366473149492,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036216237105744076,
      "loss": 5.0067,
      "mean_token_accuracy": 0.220675265789032,
      "num_tokens": 86720814.0,
      "step": 41785
    },
    {
      "entropy": 5.67734055519104,
      "epoch": 3.790820029027576,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003621325402344154,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2294454827904701,
      "num_tokens": 86731229.0,
      "step": 41790
    },
    {
      "entropy": 5.644534921646118,
      "epoch": 3.7912735849056602,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036210270760946684,
      "loss": 4.8778,
      "mean_token_accuracy": 0.21905871629714965,
      "num_tokens": 86741440.0,
      "step": 41795
    },
    {
      "entropy": 5.730336475372314,
      "epoch": 3.7917271407837445,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003620728731832122,
      "loss": 4.9731,
      "mean_token_accuracy": 0.2130439653992653,
      "num_tokens": 86752022.0,
      "step": 41800
    },
    {
      "entropy": 5.710660171508789,
      "epoch": 3.7921806966618288,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003620430369562683,
      "loss": 4.8631,
      "mean_token_accuracy": 0.21816670000553132,
      "num_tokens": 86763090.0,
      "step": 41805
    },
    {
      "entropy": 5.650964021682739,
      "epoch": 3.792634252539913,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036201319892925214,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22272211015224458,
      "num_tokens": 86773357.0,
      "step": 41810
    },
    {
      "entropy": 5.626104068756104,
      "epoch": 3.7930878084179973,
      "grad_norm": 1.328125,
      "learning_rate": 0.000361983359102781,
      "loss": 4.8437,
      "mean_token_accuracy": 0.22403311729431152,
      "num_tokens": 86784402.0,
      "step": 41815
    },
    {
      "entropy": 5.678772974014282,
      "epoch": 3.793541364296081,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036195351747747177,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2217594116926193,
      "num_tokens": 86794945.0,
      "step": 41820
    },
    {
      "entropy": 5.651041793823242,
      "epoch": 3.7939949201741654,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036192367405394157,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2199644699692726,
      "num_tokens": 86806936.0,
      "step": 41825
    },
    {
      "entropy": 5.737751197814942,
      "epoch": 3.7944484760522497,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003618938288328077,
      "loss": 5.003,
      "mean_token_accuracy": 0.21363526582717896,
      "num_tokens": 86817776.0,
      "step": 41830
    },
    {
      "entropy": 5.613715744018554,
      "epoch": 3.794902031930334,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003618639818146873,
      "loss": 4.7739,
      "mean_token_accuracy": 0.22628486156463623,
      "num_tokens": 86828054.0,
      "step": 41835
    },
    {
      "entropy": 5.6177269458770756,
      "epoch": 3.7953555878084178,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036183413300019767,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22060526460409163,
      "num_tokens": 86838132.0,
      "step": 41840
    },
    {
      "entropy": 5.589739084243774,
      "epoch": 3.795809143686502,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036180428238995595,
      "loss": 4.9013,
      "mean_token_accuracy": 0.21964032500982283,
      "num_tokens": 86848219.0,
      "step": 41845
    },
    {
      "entropy": 5.607968997955322,
      "epoch": 3.7962626995645863,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003617744299845796,
      "loss": 4.8288,
      "mean_token_accuracy": 0.22354236543178557,
      "num_tokens": 86858559.0,
      "step": 41850
    },
    {
      "entropy": 5.620931577682495,
      "epoch": 3.7967162554426706,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003617445757846859,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22050397843122482,
      "num_tokens": 86868573.0,
      "step": 41855
    },
    {
      "entropy": 5.728419017791748,
      "epoch": 3.797169811320755,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036171471979089227,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2237102523446083,
      "num_tokens": 86878641.0,
      "step": 41860
    },
    {
      "entropy": 5.71133828163147,
      "epoch": 3.797623367198839,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003616848620038162,
      "loss": 4.948,
      "mean_token_accuracy": 0.2154051721096039,
      "num_tokens": 86888644.0,
      "step": 41865
    },
    {
      "entropy": 5.737832403182983,
      "epoch": 3.7980769230769234,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000361655002424075,
      "loss": 5.0507,
      "mean_token_accuracy": 0.2099604532122612,
      "num_tokens": 86899428.0,
      "step": 41870
    },
    {
      "entropy": 5.71644082069397,
      "epoch": 3.798530478955007,
      "grad_norm": 1.25,
      "learning_rate": 0.00036162514105228635,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21976384818553923,
      "num_tokens": 86910155.0,
      "step": 41875
    },
    {
      "entropy": 5.626809692382812,
      "epoch": 3.7989840348330914,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003615952778890676,
      "loss": 4.8824,
      "mean_token_accuracy": 0.22089843600988388,
      "num_tokens": 86919783.0,
      "step": 41880
    },
    {
      "entropy": 5.661923933029175,
      "epoch": 3.7994375907111757,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036156541293503653,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21564303934574128,
      "num_tokens": 86929647.0,
      "step": 41885
    },
    {
      "entropy": 5.676453304290772,
      "epoch": 3.7998911465892595,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036153554619081064,
      "loss": 4.865,
      "mean_token_accuracy": 0.22050565481185913,
      "num_tokens": 86940113.0,
      "step": 41890
    },
    {
      "entropy": 5.697487878799438,
      "epoch": 3.800344702467344,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003615056776570076,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22345710247755052,
      "num_tokens": 86949950.0,
      "step": 41895
    },
    {
      "entropy": 5.605305290222168,
      "epoch": 3.800798258345428,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036147580733424513,
      "loss": 4.8296,
      "mean_token_accuracy": 0.22983558177948,
      "num_tokens": 86959286.0,
      "step": 41900
    },
    {
      "entropy": 5.629342365264892,
      "epoch": 3.8012518142235123,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003614459352231409,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22396614104509355,
      "num_tokens": 86970165.0,
      "step": 41905
    },
    {
      "entropy": 5.638010835647583,
      "epoch": 3.8017053701015966,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003614160613243127,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22057184129953383,
      "num_tokens": 86980331.0,
      "step": 41910
    },
    {
      "entropy": 5.755900764465332,
      "epoch": 3.802158925979681,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003613861856383783,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21341018676757811,
      "num_tokens": 86991790.0,
      "step": 41915
    },
    {
      "entropy": 5.7179975509643555,
      "epoch": 3.802612481857765,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003613563081659556,
      "loss": 5.0289,
      "mean_token_accuracy": 0.22142690569162368,
      "num_tokens": 87001220.0,
      "step": 41920
    },
    {
      "entropy": 5.681937408447266,
      "epoch": 3.803066037735849,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003613264289076625,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22190347462892532,
      "num_tokens": 87012720.0,
      "step": 41925
    },
    {
      "entropy": 5.678753805160523,
      "epoch": 3.803519593613933,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003612965478641167,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21755171120166777,
      "num_tokens": 87021891.0,
      "step": 41930
    },
    {
      "entropy": 5.746618986129761,
      "epoch": 3.8039731494920175,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036126666503593643,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20899759232997894,
      "num_tokens": 87032546.0,
      "step": 41935
    },
    {
      "entropy": 5.675739765167236,
      "epoch": 3.8044267053701017,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003612367804237395,
      "loss": 4.868,
      "mean_token_accuracy": 0.21791332811117173,
      "num_tokens": 87042173.0,
      "step": 41940
    },
    {
      "entropy": 5.702556896209717,
      "epoch": 3.8048802612481856,
      "grad_norm": 1.25,
      "learning_rate": 0.0003612068940281439,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22178468108177185,
      "num_tokens": 87052475.0,
      "step": 41945
    },
    {
      "entropy": 5.666600227355957,
      "epoch": 3.80533381712627,
      "grad_norm": 1.25,
      "learning_rate": 0.0003611770058497679,
      "loss": 4.8782,
      "mean_token_accuracy": 0.21802892088890075,
      "num_tokens": 87062698.0,
      "step": 41950
    },
    {
      "entropy": 5.644330978393555,
      "epoch": 3.805787373004354,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036114711588922945,
      "loss": 4.8914,
      "mean_token_accuracy": 0.20887335687875747,
      "num_tokens": 87073168.0,
      "step": 41955
    },
    {
      "entropy": 5.706921482086182,
      "epoch": 3.8062409288824384,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003611172241471466,
      "loss": 4.981,
      "mean_token_accuracy": 0.21309708058834076,
      "num_tokens": 87084270.0,
      "step": 41960
    },
    {
      "entropy": 5.641605520248413,
      "epoch": 3.8066944847605226,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003610873306241376,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2200206845998764,
      "num_tokens": 87093869.0,
      "step": 41965
    },
    {
      "entropy": 5.635382652282715,
      "epoch": 3.807148040638607,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036105743532082075,
      "loss": 4.8631,
      "mean_token_accuracy": 0.21979954987764358,
      "num_tokens": 87103371.0,
      "step": 41970
    },
    {
      "entropy": 5.677100706100464,
      "epoch": 3.8076015965166907,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003610275382378142,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22230157405138015,
      "num_tokens": 87113384.0,
      "step": 41975
    },
    {
      "entropy": 5.748106241226196,
      "epoch": 3.808055152394775,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003609976393757362,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21906560957431792,
      "num_tokens": 87123925.0,
      "step": 41980
    },
    {
      "entropy": 5.659660196304321,
      "epoch": 3.8085087082728593,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036096773873520513,
      "loss": 4.873,
      "mean_token_accuracy": 0.21941274851560594,
      "num_tokens": 87134608.0,
      "step": 41985
    },
    {
      "entropy": 5.662461137771606,
      "epoch": 3.8089622641509435,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036093783631683927,
      "loss": 4.89,
      "mean_token_accuracy": 0.21935552656650542,
      "num_tokens": 87145184.0,
      "step": 41990
    },
    {
      "entropy": 5.639924192428589,
      "epoch": 3.8094158200290273,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003609079321212571,
      "loss": 4.7957,
      "mean_token_accuracy": 0.22748257219791412,
      "num_tokens": 87155483.0,
      "step": 41995
    },
    {
      "entropy": 5.656929588317871,
      "epoch": 3.8098693759071116,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036087802614907696,
      "loss": 4.989,
      "mean_token_accuracy": 0.21253972947597505,
      "num_tokens": 87166615.0,
      "step": 42000
    },
    {
      "epoch": 3.8098693759071116,
      "eval_entropy": 5.384581031950351,
      "eval_loss": 4.967734336853027,
      "eval_mean_token_accuracy": 0.22687854531528723,
      "eval_num_tokens": 87166615.0,
      "eval_runtime": 20.6387,
      "eval_samples_per_second": 1713.285,
      "eval_steps_per_second": 214.161,
      "step": 42000
    },
    {
      "entropy": 5.680204439163208,
      "epoch": 3.810322931785196,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036084811840091736,
      "loss": 4.8975,
      "mean_token_accuracy": 0.2192764014005661,
      "num_tokens": 87177151.0,
      "step": 42005
    },
    {
      "entropy": 5.660609245300293,
      "epoch": 3.81077648766328,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003608182088773969,
      "loss": 4.889,
      "mean_token_accuracy": 0.22212653309106828,
      "num_tokens": 87188085.0,
      "step": 42010
    },
    {
      "entropy": 5.658230924606324,
      "epoch": 3.8112300435413644,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003607882975791338,
      "loss": 4.9598,
      "mean_token_accuracy": 0.21465011090040206,
      "num_tokens": 87197240.0,
      "step": 42015
    },
    {
      "entropy": 5.644859695434571,
      "epoch": 3.8116835994194487,
      "grad_norm": 1.21875,
      "learning_rate": 0.000360758384506747,
      "loss": 4.8793,
      "mean_token_accuracy": 0.228286075592041,
      "num_tokens": 87208064.0,
      "step": 42020
    },
    {
      "entropy": 5.653703308105468,
      "epoch": 3.8121371552975325,
      "grad_norm": 1.25,
      "learning_rate": 0.00036072846966085496,
      "loss": 4.8079,
      "mean_token_accuracy": 0.22089341282844543,
      "num_tokens": 87218386.0,
      "step": 42025
    },
    {
      "entropy": 5.61584358215332,
      "epoch": 3.8125907111756168,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036069855304207634,
      "loss": 4.8014,
      "mean_token_accuracy": 0.22561967074871064,
      "num_tokens": 87228459.0,
      "step": 42030
    },
    {
      "entropy": 5.676889228820801,
      "epoch": 3.813044267053701,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003606686346510297,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21037752330303192,
      "num_tokens": 87240317.0,
      "step": 42035
    },
    {
      "entropy": 5.607076740264892,
      "epoch": 3.8134978229317853,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000360638714488334,
      "loss": 4.7968,
      "mean_token_accuracy": 0.22523918747901917,
      "num_tokens": 87250197.0,
      "step": 42040
    },
    {
      "entropy": 5.61420693397522,
      "epoch": 3.813951378809869,
      "grad_norm": 1.25,
      "learning_rate": 0.00036060879255460766,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2207457348704338,
      "num_tokens": 87261266.0,
      "step": 42045
    },
    {
      "entropy": 5.591328716278076,
      "epoch": 3.8144049346879534,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003605788688504699,
      "loss": 4.7762,
      "mean_token_accuracy": 0.23129039853811265,
      "num_tokens": 87270490.0,
      "step": 42050
    },
    {
      "entropy": 5.662823438644409,
      "epoch": 3.8148584905660377,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003605489433765392,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22233954071998596,
      "num_tokens": 87281260.0,
      "step": 42055
    },
    {
      "entropy": 5.630814409255981,
      "epoch": 3.815312046444122,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003605190161334345,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22971397191286086,
      "num_tokens": 87290577.0,
      "step": 42060
    },
    {
      "entropy": 5.596176338195801,
      "epoch": 3.815765602322206,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003604890871217748,
      "loss": 4.7494,
      "mean_token_accuracy": 0.23253820985555648,
      "num_tokens": 87300185.0,
      "step": 42065
    },
    {
      "entropy": 5.679486799240112,
      "epoch": 3.8162191582002905,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003604591563421791,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21951487958431243,
      "num_tokens": 87310088.0,
      "step": 42070
    },
    {
      "entropy": 5.665472793579101,
      "epoch": 3.8166727140783747,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003604292237952662,
      "loss": 4.9221,
      "mean_token_accuracy": 0.2151668593287468,
      "num_tokens": 87319942.0,
      "step": 42075
    },
    {
      "entropy": 5.676095628738404,
      "epoch": 3.8171262699564585,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003603992894816551,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22059175372123718,
      "num_tokens": 87329175.0,
      "step": 42080
    },
    {
      "entropy": 5.585857772827149,
      "epoch": 3.817579825834543,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000360369353401965,
      "loss": 4.8033,
      "mean_token_accuracy": 0.2286943018436432,
      "num_tokens": 87338555.0,
      "step": 42085
    },
    {
      "entropy": 5.6433518409729,
      "epoch": 3.818033381712627,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036033941555681486,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23093942701816558,
      "num_tokens": 87348424.0,
      "step": 42090
    },
    {
      "entropy": 5.626125717163086,
      "epoch": 3.8184869375907113,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036030947594682395,
      "loss": 4.8088,
      "mean_token_accuracy": 0.22642666101455688,
      "num_tokens": 87359724.0,
      "step": 42095
    },
    {
      "entropy": 5.644570446014404,
      "epoch": 3.818940493468795,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003602795345726113,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2167738288640976,
      "num_tokens": 87371324.0,
      "step": 42100
    },
    {
      "entropy": 5.666535758972168,
      "epoch": 3.8193940493468794,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003602495914347962,
      "loss": 4.8179,
      "mean_token_accuracy": 0.22643609046936036,
      "num_tokens": 87380770.0,
      "step": 42105
    },
    {
      "entropy": 5.637687110900879,
      "epoch": 3.8198476052249637,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036021964653399773,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21992146223783493,
      "num_tokens": 87391123.0,
      "step": 42110
    },
    {
      "entropy": 5.687300491333008,
      "epoch": 3.820301161103048,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036018969987083524,
      "loss": 4.9386,
      "mean_token_accuracy": 0.217915178835392,
      "num_tokens": 87401357.0,
      "step": 42115
    },
    {
      "entropy": 5.672944498062134,
      "epoch": 3.8207547169811322,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003601597514459281,
      "loss": 4.8788,
      "mean_token_accuracy": 0.22395591288805008,
      "num_tokens": 87411373.0,
      "step": 42120
    },
    {
      "entropy": 5.622548007965088,
      "epoch": 3.8212082728592165,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036012980125989567,
      "loss": 4.8599,
      "mean_token_accuracy": 0.23098938763141633,
      "num_tokens": 87421915.0,
      "step": 42125
    },
    {
      "entropy": 5.660698080062867,
      "epoch": 3.8216618287373003,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003600998493133571,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22224850803613663,
      "num_tokens": 87432148.0,
      "step": 42130
    },
    {
      "entropy": 5.650674295425415,
      "epoch": 3.8221153846153846,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003600698956069321,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21943789422512056,
      "num_tokens": 87442605.0,
      "step": 42135
    },
    {
      "entropy": 5.660288906097412,
      "epoch": 3.822568940493469,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003600399401412399,
      "loss": 4.8707,
      "mean_token_accuracy": 0.22572477757930756,
      "num_tokens": 87452482.0,
      "step": 42140
    },
    {
      "entropy": 5.573111867904663,
      "epoch": 3.823022496371553,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036000998291690005,
      "loss": 4.7588,
      "mean_token_accuracy": 0.22543646544218063,
      "num_tokens": 87462074.0,
      "step": 42145
    },
    {
      "entropy": 5.69090576171875,
      "epoch": 3.823476052249637,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003599800239345321,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22086234837770463,
      "num_tokens": 87472377.0,
      "step": 42150
    },
    {
      "entropy": 5.644957590103149,
      "epoch": 3.823929608127721,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003599500631947557,
      "loss": 4.8868,
      "mean_token_accuracy": 0.22008109092712402,
      "num_tokens": 87483436.0,
      "step": 42155
    },
    {
      "entropy": 5.709145784378052,
      "epoch": 3.8243831640058055,
      "grad_norm": 1.375,
      "learning_rate": 0.00035992010069819025,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22492895126342774,
      "num_tokens": 87492231.0,
      "step": 42160
    },
    {
      "entropy": 5.769653940200806,
      "epoch": 3.8248367198838897,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035989013644545555,
      "loss": 4.9738,
      "mean_token_accuracy": 0.2219487965106964,
      "num_tokens": 87503416.0,
      "step": 42165
    },
    {
      "entropy": 5.642581176757813,
      "epoch": 3.825290275761974,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035986017043717114,
      "loss": 4.8654,
      "mean_token_accuracy": 0.2276483952999115,
      "num_tokens": 87513737.0,
      "step": 42170
    },
    {
      "entropy": 5.695849800109864,
      "epoch": 3.8257438316400583,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003598302026739568,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21451593339443206,
      "num_tokens": 87523532.0,
      "step": 42175
    },
    {
      "entropy": 5.6309166431427,
      "epoch": 3.826197387518142,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035980023315643217,
      "loss": 4.7907,
      "mean_token_accuracy": 0.2308505430817604,
      "num_tokens": 87534066.0,
      "step": 42180
    },
    {
      "entropy": 5.718373584747314,
      "epoch": 3.8266509433962264,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003597702618852172,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20748225301504136,
      "num_tokens": 87544520.0,
      "step": 42185
    },
    {
      "entropy": 5.6563128471374515,
      "epoch": 3.8271044992743106,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003597402888609316,
      "loss": 4.9061,
      "mean_token_accuracy": 0.22385678589344024,
      "num_tokens": 87553913.0,
      "step": 42190
    },
    {
      "entropy": 5.679483699798584,
      "epoch": 3.827558055152395,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035971031408419524,
      "loss": 4.8956,
      "mean_token_accuracy": 0.22020772099494934,
      "num_tokens": 87563434.0,
      "step": 42195
    },
    {
      "entropy": 5.674253368377686,
      "epoch": 3.8280116110304787,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035968033755562795,
      "loss": 4.9175,
      "mean_token_accuracy": 0.22293412834405898,
      "num_tokens": 87573137.0,
      "step": 42200
    },
    {
      "entropy": 5.723353528976441,
      "epoch": 3.828465166908563,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003596503592758497,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22234390676021576,
      "num_tokens": 87582610.0,
      "step": 42205
    },
    {
      "entropy": 5.6488790035247805,
      "epoch": 3.8289187227866472,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003596203792454805,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22284764498472215,
      "num_tokens": 87593062.0,
      "step": 42210
    },
    {
      "entropy": 5.674526309967041,
      "epoch": 3.8293722786647315,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003595903974651403,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22136088013648986,
      "num_tokens": 87604250.0,
      "step": 42215
    },
    {
      "entropy": 5.7232660293579105,
      "epoch": 3.8298258345428158,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003595604139354491,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21748800873756408,
      "num_tokens": 87615434.0,
      "step": 42220
    },
    {
      "entropy": 5.675612068176269,
      "epoch": 3.8302793904209,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000359530428657027,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2279006376862526,
      "num_tokens": 87625005.0,
      "step": 42225
    },
    {
      "entropy": 5.623678255081177,
      "epoch": 3.8307329462989843,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003595004416304941,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22245027869939804,
      "num_tokens": 87635486.0,
      "step": 42230
    },
    {
      "entropy": 5.668291807174683,
      "epoch": 3.831186502177068,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035947045285647045,
      "loss": 4.8906,
      "mean_token_accuracy": 0.2254578799009323,
      "num_tokens": 87645771.0,
      "step": 42235
    },
    {
      "entropy": 5.754446744918823,
      "epoch": 3.8316400580551524,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003594404623355765,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2226011261343956,
      "num_tokens": 87656116.0,
      "step": 42240
    },
    {
      "entropy": 5.8198717594146725,
      "epoch": 3.8320936139332367,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003594104700684321,
      "loss": 5.093,
      "mean_token_accuracy": 0.21064186543226243,
      "num_tokens": 87666476.0,
      "step": 42245
    },
    {
      "entropy": 5.6665606021881105,
      "epoch": 3.8325471698113205,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035938047605565764,
      "loss": 4.9014,
      "mean_token_accuracy": 0.23099779933691025,
      "num_tokens": 87677406.0,
      "step": 42250
    },
    {
      "entropy": 5.614005374908447,
      "epoch": 3.8330007256894048,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035935048029787356,
      "loss": 4.8268,
      "mean_token_accuracy": 0.22654626965522767,
      "num_tokens": 87687461.0,
      "step": 42255
    },
    {
      "entropy": 5.650856590270996,
      "epoch": 3.833454281567489,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003593204827957,
      "loss": 4.8506,
      "mean_token_accuracy": 0.2269226223230362,
      "num_tokens": 87696470.0,
      "step": 42260
    },
    {
      "entropy": 5.659826469421387,
      "epoch": 3.8339078374455733,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003592904835497574,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21572404354810715,
      "num_tokens": 87706598.0,
      "step": 42265
    },
    {
      "entropy": 5.769496011734009,
      "epoch": 3.8343613933236576,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003592604825606661,
      "loss": 5.0127,
      "mean_token_accuracy": 0.21079034954309464,
      "num_tokens": 87717749.0,
      "step": 42270
    },
    {
      "entropy": 5.6562755584716795,
      "epoch": 3.834814949201742,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003592304798290465,
      "loss": 4.8361,
      "mean_token_accuracy": 0.22981930524110794,
      "num_tokens": 87727893.0,
      "step": 42275
    },
    {
      "entropy": 5.607091474533081,
      "epoch": 3.835268505079826,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003592004753555191,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22092060148715972,
      "num_tokens": 87737980.0,
      "step": 42280
    },
    {
      "entropy": 5.6526471138000485,
      "epoch": 3.83572206095791,
      "grad_norm": 1.25,
      "learning_rate": 0.0003591704691407045,
      "loss": 4.8936,
      "mean_token_accuracy": 0.223111492395401,
      "num_tokens": 87749271.0,
      "step": 42285
    },
    {
      "entropy": 5.649973440170288,
      "epoch": 3.836175616835994,
      "grad_norm": 1.375,
      "learning_rate": 0.00035914046118522313,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22665930539369583,
      "num_tokens": 87760523.0,
      "step": 42290
    },
    {
      "entropy": 5.648125171661377,
      "epoch": 3.8366291727140784,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035911045148969556,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22139427959918975,
      "num_tokens": 87770084.0,
      "step": 42295
    },
    {
      "entropy": 5.736617374420166,
      "epoch": 3.8370827285921627,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035908044005474244,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21465548723936081,
      "num_tokens": 87780163.0,
      "step": 42300
    },
    {
      "entropy": 5.579972600936889,
      "epoch": 3.8375362844702465,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003590504268809843,
      "loss": 4.8322,
      "mean_token_accuracy": 0.2264435574412346,
      "num_tokens": 87791382.0,
      "step": 42305
    },
    {
      "entropy": 5.541245746612549,
      "epoch": 3.837989840348331,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035902041196904204,
      "loss": 4.8544,
      "mean_token_accuracy": 0.22245441973209382,
      "num_tokens": 87802143.0,
      "step": 42310
    },
    {
      "entropy": 5.649176836013794,
      "epoch": 3.838443396226415,
      "grad_norm": 1.375,
      "learning_rate": 0.0003589903953195362,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22229018360376357,
      "num_tokens": 87812255.0,
      "step": 42315
    },
    {
      "entropy": 5.713312864303589,
      "epoch": 3.8388969521044993,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035896037693308757,
      "loss": 4.908,
      "mean_token_accuracy": 0.21780331134796144,
      "num_tokens": 87822744.0,
      "step": 42320
    },
    {
      "entropy": 5.758703374862671,
      "epoch": 3.8393505079825836,
      "grad_norm": 1.390625,
      "learning_rate": 0.000358930356810317,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21666617393493653,
      "num_tokens": 87833701.0,
      "step": 42325
    },
    {
      "entropy": 5.697555065155029,
      "epoch": 3.839804063860668,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003589003349518452,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21528303176164626,
      "num_tokens": 87844215.0,
      "step": 42330
    },
    {
      "entropy": 5.710298538208008,
      "epoch": 3.8402576197387517,
      "grad_norm": 1.25,
      "learning_rate": 0.00035887031135829305,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21628482341766359,
      "num_tokens": 87854281.0,
      "step": 42335
    },
    {
      "entropy": 5.710420608520508,
      "epoch": 3.840711175616836,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003588402860302816,
      "loss": 4.976,
      "mean_token_accuracy": 0.21514319777488708,
      "num_tokens": 87866003.0,
      "step": 42340
    },
    {
      "entropy": 5.681373405456543,
      "epoch": 3.84116473149492,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035881025896843165,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22183520942926407,
      "num_tokens": 87875854.0,
      "step": 42345
    },
    {
      "entropy": 5.726830339431762,
      "epoch": 3.8416182873730045,
      "grad_norm": 1.25,
      "learning_rate": 0.00035878023017336415,
      "loss": 5.0196,
      "mean_token_accuracy": 0.21679281890392305,
      "num_tokens": 87887242.0,
      "step": 42350
    },
    {
      "entropy": 5.711236095428466,
      "epoch": 3.8420718432510883,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003587501996457001,
      "loss": 4.912,
      "mean_token_accuracy": 0.2161334663629532,
      "num_tokens": 87897859.0,
      "step": 42355
    },
    {
      "entropy": 5.659146785736084,
      "epoch": 3.8425253991291726,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035872016738606067,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22292605489492417,
      "num_tokens": 87908945.0,
      "step": 42360
    },
    {
      "entropy": 5.668125963211059,
      "epoch": 3.842978955007257,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035869013339506674,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22772708535194397,
      "num_tokens": 87919572.0,
      "step": 42365
    },
    {
      "entropy": 5.69610104560852,
      "epoch": 3.843432510885341,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035866009767333964,
      "loss": 4.9561,
      "mean_token_accuracy": 0.21872315108776091,
      "num_tokens": 87929337.0,
      "step": 42370
    },
    {
      "entropy": 5.728446769714355,
      "epoch": 3.8438860667634254,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035863006022150034,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21361604034900666,
      "num_tokens": 87940513.0,
      "step": 42375
    },
    {
      "entropy": 5.659322309494018,
      "epoch": 3.8443396226415096,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035860002104017014,
      "loss": 4.8422,
      "mean_token_accuracy": 0.22852765172719955,
      "num_tokens": 87951181.0,
      "step": 42380
    },
    {
      "entropy": 5.671182870864868,
      "epoch": 3.8447931785195935,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003585699801299701,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22048140913248063,
      "num_tokens": 87961925.0,
      "step": 42385
    },
    {
      "entropy": 5.642671585083008,
      "epoch": 3.8452467343976777,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035853993749152165,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22414727210998536,
      "num_tokens": 87972233.0,
      "step": 42390
    },
    {
      "entropy": 5.685314893722534,
      "epoch": 3.845700290275762,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035850989312544596,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21532087177038192,
      "num_tokens": 87983131.0,
      "step": 42395
    },
    {
      "entropy": 5.648875999450683,
      "epoch": 3.8461538461538463,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003584798470323644,
      "loss": 4.8728,
      "mean_token_accuracy": 0.2225247800350189,
      "num_tokens": 87993926.0,
      "step": 42400
    },
    {
      "entropy": 5.745465564727783,
      "epoch": 3.84660740203193,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003584497992128984,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21157946884632112,
      "num_tokens": 88004403.0,
      "step": 42405
    },
    {
      "entropy": 5.661701297760009,
      "epoch": 3.8470609579100143,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003584197496676692,
      "loss": 4.8528,
      "mean_token_accuracy": 0.22661157548427582,
      "num_tokens": 88015289.0,
      "step": 42410
    },
    {
      "entropy": 5.581419181823731,
      "epoch": 3.8475145137880986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035838969839729846,
      "loss": 4.847,
      "mean_token_accuracy": 0.22180329710245134,
      "num_tokens": 88025724.0,
      "step": 42415
    },
    {
      "entropy": 5.611706209182739,
      "epoch": 3.847968069666183,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003583596454024073,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21425333470106125,
      "num_tokens": 88036444.0,
      "step": 42420
    },
    {
      "entropy": 5.755955171585083,
      "epoch": 3.848421625544267,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003583295906836176,
      "loss": 4.9257,
      "mean_token_accuracy": 0.22433597296476365,
      "num_tokens": 88047334.0,
      "step": 42425
    },
    {
      "entropy": 5.714289331436158,
      "epoch": 3.8488751814223514,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035829953424155065,
      "loss": 4.879,
      "mean_token_accuracy": 0.22144878804683685,
      "num_tokens": 88057268.0,
      "step": 42430
    },
    {
      "entropy": 5.589236450195313,
      "epoch": 3.8493287373004357,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003582694760768281,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22484455555677413,
      "num_tokens": 88067615.0,
      "step": 42435
    },
    {
      "entropy": 5.668517780303955,
      "epoch": 3.8497822931785195,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035823941619007155,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21915120780467987,
      "num_tokens": 88077754.0,
      "step": 42440
    },
    {
      "entropy": 5.637988471984864,
      "epoch": 3.8502358490566038,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035820935458190275,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21835225373506545,
      "num_tokens": 88088346.0,
      "step": 42445
    },
    {
      "entropy": 5.607458400726318,
      "epoch": 3.850689404934688,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035817929125294323,
      "loss": 4.8471,
      "mean_token_accuracy": 0.22551659047603606,
      "num_tokens": 88098534.0,
      "step": 42450
    },
    {
      "entropy": 5.629177379608154,
      "epoch": 3.8511429608127723,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003581492262038148,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21963482201099396,
      "num_tokens": 88109264.0,
      "step": 42455
    },
    {
      "entropy": 5.691736936569214,
      "epoch": 3.851596516690856,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003581191594351392,
      "loss": 4.9263,
      "mean_token_accuracy": 0.2220442771911621,
      "num_tokens": 88120407.0,
      "step": 42460
    },
    {
      "entropy": 5.71118311882019,
      "epoch": 3.8520500725689404,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003580890909475381,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21738689988851548,
      "num_tokens": 88131908.0,
      "step": 42465
    },
    {
      "entropy": 5.6135190486907955,
      "epoch": 3.8525036284470247,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035805902074163354,
      "loss": 4.8501,
      "mean_token_accuracy": 0.22048432677984237,
      "num_tokens": 88142559.0,
      "step": 42470
    },
    {
      "entropy": 5.638924598693848,
      "epoch": 3.852957184325109,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003580289488180472,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22537946105003356,
      "num_tokens": 88153441.0,
      "step": 42475
    },
    {
      "entropy": 5.669911479949951,
      "epoch": 3.853410740203193,
      "grad_norm": 1.890625,
      "learning_rate": 0.00035799887517740103,
      "loss": 4.8931,
      "mean_token_accuracy": 0.2249554842710495,
      "num_tokens": 88164153.0,
      "step": 42480
    },
    {
      "entropy": 5.690925455093383,
      "epoch": 3.8538642960812775,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000357968799820317,
      "loss": 4.9164,
      "mean_token_accuracy": 0.21847398728132247,
      "num_tokens": 88174138.0,
      "step": 42485
    },
    {
      "entropy": 5.676265048980713,
      "epoch": 3.8543178519593613,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003579387227474171,
      "loss": 4.8933,
      "mean_token_accuracy": 0.21850620657205583,
      "num_tokens": 88184704.0,
      "step": 42490
    },
    {
      "entropy": 5.758743667602539,
      "epoch": 3.8547714078374455,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003579086439593232,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21315994560718537,
      "num_tokens": 88194225.0,
      "step": 42495
    },
    {
      "entropy": 5.711723804473877,
      "epoch": 3.85522496371553,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003578785634566574,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21807560622692107,
      "num_tokens": 88206125.0,
      "step": 42500
    },
    {
      "entropy": 5.616908168792724,
      "epoch": 3.855678519593614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035784848124004187,
      "loss": 4.7939,
      "mean_token_accuracy": 0.22949125617742538,
      "num_tokens": 88217085.0,
      "step": 42505
    },
    {
      "entropy": 5.633609485626221,
      "epoch": 3.856132075471698,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003578183973100986,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2189400628209114,
      "num_tokens": 88228126.0,
      "step": 42510
    },
    {
      "entropy": 5.655173254013062,
      "epoch": 3.856585631349782,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035778831166744966,
      "loss": 4.99,
      "mean_token_accuracy": 0.2155231013894081,
      "num_tokens": 88238829.0,
      "step": 42515
    },
    {
      "entropy": 5.663262462615966,
      "epoch": 3.8570391872278664,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003577582243127175,
      "loss": 4.919,
      "mean_token_accuracy": 0.21680222153663636,
      "num_tokens": 88249821.0,
      "step": 42520
    },
    {
      "entropy": 5.665457344055175,
      "epoch": 3.8574927431059507,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003577281352465241,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22815054953098296,
      "num_tokens": 88260126.0,
      "step": 42525
    },
    {
      "entropy": 5.673862886428833,
      "epoch": 3.857946298984035,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035769804446949177,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22627848386764526,
      "num_tokens": 88271492.0,
      "step": 42530
    },
    {
      "entropy": 5.739221715927124,
      "epoch": 3.8583998548621192,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003576679519822428,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21619007140398025,
      "num_tokens": 88281966.0,
      "step": 42535
    },
    {
      "entropy": 5.568761157989502,
      "epoch": 3.858853410740203,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035763785778539963,
      "loss": 4.8066,
      "mean_token_accuracy": 0.23384695053100585,
      "num_tokens": 88292805.0,
      "step": 42540
    },
    {
      "entropy": 5.632377243041992,
      "epoch": 3.8593069666182873,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003576077618795844,
      "loss": 4.847,
      "mean_token_accuracy": 0.22488627582788467,
      "num_tokens": 88302727.0,
      "step": 42545
    },
    {
      "entropy": 5.546494102478027,
      "epoch": 3.8597605224963716,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003575776642654196,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22535971701145172,
      "num_tokens": 88313150.0,
      "step": 42550
    },
    {
      "entropy": 5.656926202774048,
      "epoch": 3.860214078374456,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035754756494352767,
      "loss": 4.8302,
      "mean_token_accuracy": 0.22973676770925522,
      "num_tokens": 88323060.0,
      "step": 42555
    },
    {
      "entropy": 5.748113536834717,
      "epoch": 3.8606676342525397,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003575174639145311,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22511588782072067,
      "num_tokens": 88332641.0,
      "step": 42560
    },
    {
      "entropy": 5.621601104736328,
      "epoch": 3.861121190130624,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003574873611790523,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22298148274421692,
      "num_tokens": 88342807.0,
      "step": 42565
    },
    {
      "entropy": 5.563461112976074,
      "epoch": 3.861574746008708,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035745725673771387,
      "loss": 4.7887,
      "mean_token_accuracy": 0.23116808235645295,
      "num_tokens": 88352958.0,
      "step": 42570
    },
    {
      "entropy": 5.685440301895142,
      "epoch": 3.8620283018867925,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035742715059113836,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22002824544906616,
      "num_tokens": 88363184.0,
      "step": 42575
    },
    {
      "entropy": 5.586659955978393,
      "epoch": 3.8624818577648767,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003573970427399484,
      "loss": 4.8427,
      "mean_token_accuracy": 0.2285648599267006,
      "num_tokens": 88373991.0,
      "step": 42580
    },
    {
      "entropy": 5.691594791412354,
      "epoch": 3.862935413642961,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035736693318476646,
      "loss": 4.9058,
      "mean_token_accuracy": 0.21916747093200684,
      "num_tokens": 88384518.0,
      "step": 42585
    },
    {
      "entropy": 5.57020583152771,
      "epoch": 3.8633889695210453,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035733682192621555,
      "loss": 4.7822,
      "mean_token_accuracy": 0.22705867737531663,
      "num_tokens": 88394506.0,
      "step": 42590
    },
    {
      "entropy": 5.663253259658814,
      "epoch": 3.863842525399129,
      "grad_norm": 1.375,
      "learning_rate": 0.0003573067089649181,
      "loss": 4.882,
      "mean_token_accuracy": 0.2241256892681122,
      "num_tokens": 88404625.0,
      "step": 42595
    },
    {
      "entropy": 5.701590538024902,
      "epoch": 3.8642960812772134,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00035727659430149686,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22270999997854232,
      "num_tokens": 88413754.0,
      "step": 42600
    },
    {
      "entropy": 5.729970026016235,
      "epoch": 3.8647496371552976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003572464779365748,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2104557454586029,
      "num_tokens": 88424613.0,
      "step": 42605
    },
    {
      "entropy": 5.682272052764892,
      "epoch": 3.8652031930333814,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035721635987077455,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22691932320594788,
      "num_tokens": 88435248.0,
      "step": 42610
    },
    {
      "entropy": 5.6208305835723875,
      "epoch": 3.8656567489114657,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035718624010471904,
      "loss": 4.7752,
      "mean_token_accuracy": 0.23039418011903762,
      "num_tokens": 88444802.0,
      "step": 42615
    },
    {
      "entropy": 5.6283266067504885,
      "epoch": 3.86611030478955,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003571561186390311,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2139541745185852,
      "num_tokens": 88455335.0,
      "step": 42620
    },
    {
      "entropy": 5.703613805770874,
      "epoch": 3.8665638606676342,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003571259954743338,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2262326166033745,
      "num_tokens": 88465857.0,
      "step": 42625
    },
    {
      "entropy": 5.664879179000854,
      "epoch": 3.8670174165457185,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035709587061124985,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21216832846403122,
      "num_tokens": 88475930.0,
      "step": 42630
    },
    {
      "entropy": 5.673430395126343,
      "epoch": 3.8674709724238028,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035706574405040245,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22543097287416458,
      "num_tokens": 88486732.0,
      "step": 42635
    },
    {
      "entropy": 5.698744487762451,
      "epoch": 3.867924528301887,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003570356157924145,
      "loss": 4.907,
      "mean_token_accuracy": 0.2204982176423073,
      "num_tokens": 88497017.0,
      "step": 42640
    },
    {
      "entropy": 5.63511905670166,
      "epoch": 3.868378084179971,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003570054858379092,
      "loss": 4.7828,
      "mean_token_accuracy": 0.235060678422451,
      "num_tokens": 88507423.0,
      "step": 42645
    },
    {
      "entropy": 5.6305540084838865,
      "epoch": 3.868831640058055,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035697535418750955,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2246344804763794,
      "num_tokens": 88517795.0,
      "step": 42650
    },
    {
      "entropy": 5.526465225219726,
      "epoch": 3.8692851959361394,
      "grad_norm": 1.25,
      "learning_rate": 0.00035694522084183864,
      "loss": 4.7891,
      "mean_token_accuracy": 0.22717912048101424,
      "num_tokens": 88527641.0,
      "step": 42655
    },
    {
      "entropy": 5.6489733219146725,
      "epoch": 3.8697387518142237,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00035691508580151966,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21360521465539933,
      "num_tokens": 88537874.0,
      "step": 42660
    },
    {
      "entropy": 5.659334278106689,
      "epoch": 3.8701923076923075,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003568849490671759,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22205095142126083,
      "num_tokens": 88548108.0,
      "step": 42665
    },
    {
      "entropy": 5.632326650619507,
      "epoch": 3.8706458635703918,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003568548106394305,
      "loss": 4.7962,
      "mean_token_accuracy": 0.22812678813934326,
      "num_tokens": 88558284.0,
      "step": 42670
    },
    {
      "entropy": 5.7024312019348145,
      "epoch": 3.871099419448476,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003568246705189067,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2200780212879181,
      "num_tokens": 88568635.0,
      "step": 42675
    },
    {
      "entropy": 5.660861492156982,
      "epoch": 3.8715529753265603,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003567945287062279,
      "loss": 4.9038,
      "mean_token_accuracy": 0.2169419765472412,
      "num_tokens": 88577968.0,
      "step": 42680
    },
    {
      "entropy": 5.722889089584351,
      "epoch": 3.8720065312046446,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035676438520201745,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22100013941526414,
      "num_tokens": 88590562.0,
      "step": 42685
    },
    {
      "entropy": 5.661634492874145,
      "epoch": 3.872460087082729,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035673424000689857,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2160722002387047,
      "num_tokens": 88601187.0,
      "step": 42690
    },
    {
      "entropy": 5.599388647079468,
      "epoch": 3.8729136429608126,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035670409312149495,
      "loss": 4.8119,
      "mean_token_accuracy": 0.22525851279497147,
      "num_tokens": 88611459.0,
      "step": 42695
    },
    {
      "entropy": 5.6537691116333,
      "epoch": 3.873367198838897,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035667394454642963,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22465277314186097,
      "num_tokens": 88621652.0,
      "step": 42700
    },
    {
      "entropy": 5.664369297027588,
      "epoch": 3.873820754716981,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003566437942823265,
      "loss": 4.7913,
      "mean_token_accuracy": 0.22261609137058258,
      "num_tokens": 88632425.0,
      "step": 42705
    },
    {
      "entropy": 5.586523389816284,
      "epoch": 3.8742743105950654,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035661364232980884,
      "loss": 4.8056,
      "mean_token_accuracy": 0.23191829174757003,
      "num_tokens": 88641263.0,
      "step": 42710
    },
    {
      "entropy": 5.605869436264038,
      "epoch": 3.8747278664731493,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035658348868950027,
      "loss": 4.8865,
      "mean_token_accuracy": 0.21858488917350768,
      "num_tokens": 88651665.0,
      "step": 42715
    },
    {
      "entropy": 5.633283805847168,
      "epoch": 3.8751814223512335,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035655333336202436,
      "loss": 4.8292,
      "mean_token_accuracy": 0.22767774611711503,
      "num_tokens": 88660791.0,
      "step": 42720
    },
    {
      "entropy": 5.575465106964112,
      "epoch": 3.875634978229318,
      "grad_norm": 1.375,
      "learning_rate": 0.0003565231763480047,
      "loss": 4.8078,
      "mean_token_accuracy": 0.2342938542366028,
      "num_tokens": 88671341.0,
      "step": 42725
    },
    {
      "entropy": 5.62753381729126,
      "epoch": 3.876088534107402,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003564930176480651,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21765490919351577,
      "num_tokens": 88682500.0,
      "step": 42730
    },
    {
      "entropy": 5.699588394165039,
      "epoch": 3.8765420899854863,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035646285726282896,
      "loss": 4.8847,
      "mean_token_accuracy": 0.21494409292936326,
      "num_tokens": 88693513.0,
      "step": 42735
    },
    {
      "entropy": 5.715431070327758,
      "epoch": 3.8769956458635706,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035643269519292027,
      "loss": 4.9589,
      "mean_token_accuracy": 0.21727383583784105,
      "num_tokens": 88703498.0,
      "step": 42740
    },
    {
      "entropy": 5.62911696434021,
      "epoch": 3.8774492017416544,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003564025314389627,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21947107315063477,
      "num_tokens": 88714217.0,
      "step": 42745
    },
    {
      "entropy": 5.644966650009155,
      "epoch": 3.8779027576197387,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035637236600157994,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21845815926790238,
      "num_tokens": 88725006.0,
      "step": 42750
    },
    {
      "entropy": 5.6078290939331055,
      "epoch": 3.878356313497823,
      "grad_norm": 1.296875,
      "learning_rate": 0.000356342198881396,
      "loss": 4.8256,
      "mean_token_accuracy": 0.22952368557453157,
      "num_tokens": 88735253.0,
      "step": 42755
    },
    {
      "entropy": 5.676779317855835,
      "epoch": 3.878809869375907,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035631203007903465,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21298106908798217,
      "num_tokens": 88746313.0,
      "step": 42760
    },
    {
      "entropy": 5.6253883838653564,
      "epoch": 3.879263425253991,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003562818595951198,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2282028838992119,
      "num_tokens": 88757448.0,
      "step": 42765
    },
    {
      "entropy": 5.6675474643707275,
      "epoch": 3.8797169811320753,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003562516874302754,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2274181440472603,
      "num_tokens": 88767426.0,
      "step": 42770
    },
    {
      "entropy": 5.667445850372315,
      "epoch": 3.8801705370101596,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003562215135851253,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2196241796016693,
      "num_tokens": 88778188.0,
      "step": 42775
    },
    {
      "entropy": 5.774208354949951,
      "epoch": 3.880624092888244,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035619133806029374,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22391708791255951,
      "num_tokens": 88788927.0,
      "step": 42780
    },
    {
      "entropy": 5.74346227645874,
      "epoch": 3.881077648766328,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035616116085640455,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2183104395866394,
      "num_tokens": 88798809.0,
      "step": 42785
    },
    {
      "entropy": 5.55405740737915,
      "epoch": 3.8815312046444124,
      "grad_norm": 1.25,
      "learning_rate": 0.0003561309819740819,
      "loss": 4.7812,
      "mean_token_accuracy": 0.23459322154521942,
      "num_tokens": 88808688.0,
      "step": 42790
    },
    {
      "entropy": 5.644283056259155,
      "epoch": 3.8819847605224966,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035610080141394983,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22634931802749633,
      "num_tokens": 88818400.0,
      "step": 42795
    },
    {
      "entropy": 5.626582670211792,
      "epoch": 3.8824383164005805,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035607061917663254,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2229297026991844,
      "num_tokens": 88828032.0,
      "step": 42800
    },
    {
      "entropy": 5.667214679718017,
      "epoch": 3.8828918722786647,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035604043526275413,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21971650868654252,
      "num_tokens": 88839856.0,
      "step": 42805
    },
    {
      "entropy": 5.693911695480347,
      "epoch": 3.883345428156749,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000356010249672939,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22619104236364365,
      "num_tokens": 88851016.0,
      "step": 42810
    },
    {
      "entropy": 5.647712564468383,
      "epoch": 3.8837989840348333,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003559800624078111,
      "loss": 4.9196,
      "mean_token_accuracy": 0.22074661105871202,
      "num_tokens": 88859795.0,
      "step": 42815
    },
    {
      "entropy": 5.6466803550720215,
      "epoch": 3.884252539912917,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003559498734679949,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2239233672618866,
      "num_tokens": 88869595.0,
      "step": 42820
    },
    {
      "entropy": 5.725682020187378,
      "epoch": 3.8847060957910013,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003559196828541148,
      "loss": 4.9351,
      "mean_token_accuracy": 0.21362386643886566,
      "num_tokens": 88880857.0,
      "step": 42825
    },
    {
      "entropy": 5.660222005844116,
      "epoch": 3.8851596516690856,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000355889490566795,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22948523759841918,
      "num_tokens": 88890687.0,
      "step": 42830
    },
    {
      "entropy": 5.650462865829468,
      "epoch": 3.88561320754717,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035585929660665984,
      "loss": 4.8593,
      "mean_token_accuracy": 0.22506752610206604,
      "num_tokens": 88899600.0,
      "step": 42835
    },
    {
      "entropy": 5.67701907157898,
      "epoch": 3.886066763425254,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003558291009743339,
      "loss": 4.8715,
      "mean_token_accuracy": 0.2272772118449211,
      "num_tokens": 88909521.0,
      "step": 42840
    },
    {
      "entropy": 5.650293779373169,
      "epoch": 3.8865203193033384,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00035579890367044153,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2181972533464432,
      "num_tokens": 88920061.0,
      "step": 42845
    },
    {
      "entropy": 5.588614845275879,
      "epoch": 3.8869738751814222,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003557687046956073,
      "loss": 4.7929,
      "mean_token_accuracy": 0.2298422113060951,
      "num_tokens": 88929504.0,
      "step": 42850
    },
    {
      "entropy": 5.759730815887451,
      "epoch": 3.8874274310595065,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035573850405045554,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20895724892616271,
      "num_tokens": 88940653.0,
      "step": 42855
    },
    {
      "entropy": 5.745896053314209,
      "epoch": 3.8878809869375908,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003557083017356111,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21899197846651078,
      "num_tokens": 88952133.0,
      "step": 42860
    },
    {
      "entropy": 5.62541823387146,
      "epoch": 3.888334542815675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003556780977516983,
      "loss": 4.8535,
      "mean_token_accuracy": 0.2242206260561943,
      "num_tokens": 88963190.0,
      "step": 42865
    },
    {
      "entropy": 5.565022134780884,
      "epoch": 3.888788098693759,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035564789209934183,
      "loss": 4.8013,
      "mean_token_accuracy": 0.22476990520954132,
      "num_tokens": 88973034.0,
      "step": 42870
    },
    {
      "entropy": 5.6414652347564695,
      "epoch": 3.889241654571843,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035561768477916655,
      "loss": 4.7931,
      "mean_token_accuracy": 0.23250749409198762,
      "num_tokens": 88982809.0,
      "step": 42875
    },
    {
      "entropy": 5.656826543807983,
      "epoch": 3.8896952104499274,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000355587475791797,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21689132004976272,
      "num_tokens": 88993563.0,
      "step": 42880
    },
    {
      "entropy": 5.5386251449584964,
      "epoch": 3.8901487663280117,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035555726513785785,
      "loss": 4.7928,
      "mean_token_accuracy": 0.23060047179460524,
      "num_tokens": 89002408.0,
      "step": 42885
    },
    {
      "entropy": 5.603125715255738,
      "epoch": 3.890602322206096,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003555270528179739,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2251296490430832,
      "num_tokens": 89013060.0,
      "step": 42890
    },
    {
      "entropy": 5.705453968048095,
      "epoch": 3.89105587808418,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035549683883277,
      "loss": 4.9341,
      "mean_token_accuracy": 0.2228753462433815,
      "num_tokens": 89023406.0,
      "step": 42895
    },
    {
      "entropy": 5.701925945281983,
      "epoch": 3.891509433962264,
      "grad_norm": 1.28125,
      "learning_rate": 0.000355466623182871,
      "loss": 4.9299,
      "mean_token_accuracy": 0.22259570062160491,
      "num_tokens": 89033800.0,
      "step": 42900
    },
    {
      "entropy": 5.751665115356445,
      "epoch": 3.8919629898403483,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003554364058689017,
      "loss": 4.909,
      "mean_token_accuracy": 0.2214192718267441,
      "num_tokens": 89044601.0,
      "step": 42905
    },
    {
      "entropy": 5.721828031539917,
      "epoch": 3.8924165457184325,
      "grad_norm": 1.203125,
      "learning_rate": 0.000355406186891487,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21884151101112365,
      "num_tokens": 89056138.0,
      "step": 42910
    },
    {
      "entropy": 5.640722942352295,
      "epoch": 3.892870101596517,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003553759662512519,
      "loss": 4.8614,
      "mean_token_accuracy": 0.2239087402820587,
      "num_tokens": 89065776.0,
      "step": 42915
    },
    {
      "entropy": 5.780185270309448,
      "epoch": 3.8933236574746006,
      "grad_norm": 1.25,
      "learning_rate": 0.00035534574394882135,
      "loss": 4.992,
      "mean_token_accuracy": 0.20792556554079056,
      "num_tokens": 89076926.0,
      "step": 42920
    },
    {
      "entropy": 5.657666540145874,
      "epoch": 3.893777213352685,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003553155199848203,
      "loss": 4.91,
      "mean_token_accuracy": 0.2116399422287941,
      "num_tokens": 89088036.0,
      "step": 42925
    },
    {
      "entropy": 5.668568706512451,
      "epoch": 3.894230769230769,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003552852943598739,
      "loss": 4.858,
      "mean_token_accuracy": 0.22347044944763184,
      "num_tokens": 89098606.0,
      "step": 42930
    },
    {
      "entropy": 5.635056400299073,
      "epoch": 3.8946843251088534,
      "grad_norm": 1.328125,
      "learning_rate": 0.000355255067074607,
      "loss": 4.8139,
      "mean_token_accuracy": 0.23013464957475663,
      "num_tokens": 89107733.0,
      "step": 42935
    },
    {
      "entropy": 5.635632848739624,
      "epoch": 3.8951378809869377,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035522483812964496,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2248557135462761,
      "num_tokens": 89119094.0,
      "step": 42940
    },
    {
      "entropy": 5.626153039932251,
      "epoch": 3.895591436865022,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035519460752561285,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2281900867819786,
      "num_tokens": 89129254.0,
      "step": 42945
    },
    {
      "entropy": 5.657587242126465,
      "epoch": 3.8960449927431062,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003551643752631358,
      "loss": 4.835,
      "mean_token_accuracy": 0.22820369005203248,
      "num_tokens": 89139609.0,
      "step": 42950
    },
    {
      "entropy": 5.665063571929932,
      "epoch": 3.89649854862119,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003551341413428391,
      "loss": 4.8604,
      "mean_token_accuracy": 0.21357677280902862,
      "num_tokens": 89149504.0,
      "step": 42955
    },
    {
      "entropy": 5.643418216705323,
      "epoch": 3.8969521044992743,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003551039057653478,
      "loss": 4.861,
      "mean_token_accuracy": 0.22563812881708145,
      "num_tokens": 89160036.0,
      "step": 42960
    },
    {
      "entropy": 5.656509733200073,
      "epoch": 3.8974056603773586,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003550736685312874,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2197621926665306,
      "num_tokens": 89170960.0,
      "step": 42965
    },
    {
      "entropy": 5.651523494720459,
      "epoch": 3.8978592162554424,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003550434296412831,
      "loss": 4.8525,
      "mean_token_accuracy": 0.23290955722332002,
      "num_tokens": 89179630.0,
      "step": 42970
    },
    {
      "entropy": 5.647391891479492,
      "epoch": 3.8983127721335267,
      "grad_norm": 1.25,
      "learning_rate": 0.0003550131890959603,
      "loss": 4.8833,
      "mean_token_accuracy": 0.2271190032362938,
      "num_tokens": 89188880.0,
      "step": 42975
    },
    {
      "entropy": 5.698987054824829,
      "epoch": 3.898766328011611,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035498294689594435,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21896457374095918,
      "num_tokens": 89199585.0,
      "step": 42980
    },
    {
      "entropy": 5.67002501487732,
      "epoch": 3.899219883889695,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035495270304186074,
      "loss": 4.8646,
      "mean_token_accuracy": 0.21500124335289,
      "num_tokens": 89209812.0,
      "step": 42985
    },
    {
      "entropy": 5.635092163085938,
      "epoch": 3.8996734397677795,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003549224575343347,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22597844451665877,
      "num_tokens": 89220244.0,
      "step": 42990
    },
    {
      "entropy": 5.643746662139892,
      "epoch": 3.9001269956458637,
      "grad_norm": 1.375,
      "learning_rate": 0.00035489221037399203,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21631388664245604,
      "num_tokens": 89230618.0,
      "step": 42995
    },
    {
      "entropy": 5.665724611282348,
      "epoch": 3.900580551523948,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035486196156145805,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22363721281290055,
      "num_tokens": 89242041.0,
      "step": 43000
    },
    {
      "entropy": 5.699419355392456,
      "epoch": 3.901034107402032,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003548317110973583,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22254988551139832,
      "num_tokens": 89252817.0,
      "step": 43005
    },
    {
      "entropy": 5.6780070781707765,
      "epoch": 3.901487663280116,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003548014589823185,
      "loss": 4.9337,
      "mean_token_accuracy": 0.22347624599933624,
      "num_tokens": 89263702.0,
      "step": 43010
    },
    {
      "entropy": 5.643161582946777,
      "epoch": 3.9019412191582004,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035477120521696417,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2279886230826378,
      "num_tokens": 89273822.0,
      "step": 43015
    },
    {
      "entropy": 5.725532913208008,
      "epoch": 3.9023947750362846,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003547409498019209,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21075901985168458,
      "num_tokens": 89283973.0,
      "step": 43020
    },
    {
      "entropy": 5.7172383785247805,
      "epoch": 3.9028483309143684,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035471069273781454,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22291269749403,
      "num_tokens": 89294059.0,
      "step": 43025
    },
    {
      "entropy": 5.664397382736206,
      "epoch": 3.9033018867924527,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003546804340252707,
      "loss": 4.8443,
      "mean_token_accuracy": 0.22493355721235275,
      "num_tokens": 89304049.0,
      "step": 43030
    },
    {
      "entropy": 5.6013750553131105,
      "epoch": 3.903755442670537,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035465017366491527,
      "loss": 4.78,
      "mean_token_accuracy": 0.23305300623178482,
      "num_tokens": 89314468.0,
      "step": 43035
    },
    {
      "entropy": 5.578039312362671,
      "epoch": 3.9042089985486212,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035461991165737386,
      "loss": 4.852,
      "mean_token_accuracy": 0.22753328382968901,
      "num_tokens": 89324580.0,
      "step": 43040
    },
    {
      "entropy": 5.721569490432739,
      "epoch": 3.9046625544267055,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035458964800327245,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2118929833173752,
      "num_tokens": 89334976.0,
      "step": 43045
    },
    {
      "entropy": 5.673843050003052,
      "epoch": 3.90511611030479,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035455938270323684,
      "loss": 4.9131,
      "mean_token_accuracy": 0.21832153350114822,
      "num_tokens": 89345210.0,
      "step": 43050
    },
    {
      "entropy": 5.638338375091553,
      "epoch": 3.9055696661828736,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003545291157578928,
      "loss": 4.7758,
      "mean_token_accuracy": 0.23181362450122833,
      "num_tokens": 89354623.0,
      "step": 43055
    },
    {
      "entropy": 5.645547008514404,
      "epoch": 3.906023222060958,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035449884716786653,
      "loss": 4.8549,
      "mean_token_accuracy": 0.22735803127288817,
      "num_tokens": 89364684.0,
      "step": 43060
    },
    {
      "entropy": 5.648725605010986,
      "epoch": 3.906476777939042,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035446857693378377,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21632350236177444,
      "num_tokens": 89375415.0,
      "step": 43065
    },
    {
      "entropy": 5.688320207595825,
      "epoch": 3.9069303338171264,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003544383050562706,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22451281994581224,
      "num_tokens": 89385791.0,
      "step": 43070
    },
    {
      "entropy": 5.691095590591431,
      "epoch": 3.90738388969521,
      "grad_norm": 1.28125,
      "learning_rate": 0.000354408031535953,
      "loss": 5.0067,
      "mean_token_accuracy": 0.22069910615682603,
      "num_tokens": 89395345.0,
      "step": 43075
    },
    {
      "entropy": 5.637933921813965,
      "epoch": 3.9078374455732945,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035437775637345715,
      "loss": 4.846,
      "mean_token_accuracy": 0.22864942252635956,
      "num_tokens": 89405413.0,
      "step": 43080
    },
    {
      "entropy": 5.668398857116699,
      "epoch": 3.9082910014513788,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00035434747956940906,
      "loss": 4.8723,
      "mean_token_accuracy": 0.22741466164588928,
      "num_tokens": 89415168.0,
      "step": 43085
    },
    {
      "entropy": 5.605725193023682,
      "epoch": 3.908744557329463,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035431720112443475,
      "loss": 4.8606,
      "mean_token_accuracy": 0.228644897043705,
      "num_tokens": 89426438.0,
      "step": 43090
    },
    {
      "entropy": 5.614774513244629,
      "epoch": 3.9091981132075473,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035428692103916067,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22236608415842057,
      "num_tokens": 89437454.0,
      "step": 43095
    },
    {
      "entropy": 5.678123807907104,
      "epoch": 3.9096516690856316,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003542566393142128,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22578066885471343,
      "num_tokens": 89448082.0,
      "step": 43100
    },
    {
      "entropy": 5.581694507598877,
      "epoch": 3.9101052249637154,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035422635595021745,
      "loss": 4.8434,
      "mean_token_accuracy": 0.2237669587135315,
      "num_tokens": 89458227.0,
      "step": 43105
    },
    {
      "entropy": 5.631004858016968,
      "epoch": 3.9105587808417996,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035419607094780084,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22126545161008834,
      "num_tokens": 89468374.0,
      "step": 43110
    },
    {
      "entropy": 5.649475145339966,
      "epoch": 3.911012336719884,
      "grad_norm": 1.375,
      "learning_rate": 0.00035416578430758934,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2207956552505493,
      "num_tokens": 89477728.0,
      "step": 43115
    },
    {
      "entropy": 5.700003051757813,
      "epoch": 3.911465892597968,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035413549603020915,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22394519299268723,
      "num_tokens": 89487107.0,
      "step": 43120
    },
    {
      "entropy": 5.716149759292603,
      "epoch": 3.911919448476052,
      "grad_norm": 1.375,
      "learning_rate": 0.00035410520611628676,
      "loss": 4.9136,
      "mean_token_accuracy": 0.2304343596100807,
      "num_tokens": 89497075.0,
      "step": 43125
    },
    {
      "entropy": 5.561764574050903,
      "epoch": 3.9123730043541363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003540749145664486,
      "loss": 4.8342,
      "mean_token_accuracy": 0.222140933573246,
      "num_tokens": 89508068.0,
      "step": 43130
    },
    {
      "entropy": 5.649159574508667,
      "epoch": 3.9128265602322205,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003540446213813211,
      "loss": 4.8288,
      "mean_token_accuracy": 0.2314961478114128,
      "num_tokens": 89518371.0,
      "step": 43135
    },
    {
      "entropy": 5.6770055294036865,
      "epoch": 3.913280116110305,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035401432656153057,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2243675783276558,
      "num_tokens": 89529024.0,
      "step": 43140
    },
    {
      "entropy": 5.598103141784668,
      "epoch": 3.913733671988389,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035398403010770377,
      "loss": 4.7841,
      "mean_token_accuracy": 0.2303077444434166,
      "num_tokens": 89540331.0,
      "step": 43145
    },
    {
      "entropy": 5.549307489395142,
      "epoch": 3.9141872278664733,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000353953732020467,
      "loss": 4.793,
      "mean_token_accuracy": 0.23020783811807632,
      "num_tokens": 89550385.0,
      "step": 43150
    },
    {
      "entropy": 5.679307174682617,
      "epoch": 3.9146407837445576,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035392343230044687,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21727267056703567,
      "num_tokens": 89561480.0,
      "step": 43155
    },
    {
      "entropy": 5.696958780288696,
      "epoch": 3.9150943396226414,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035389313094827017,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22030405551195145,
      "num_tokens": 89571897.0,
      "step": 43160
    },
    {
      "entropy": 5.656802701950073,
      "epoch": 3.9155478955007257,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035386282796456343,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2238266125321388,
      "num_tokens": 89582019.0,
      "step": 43165
    },
    {
      "entropy": 5.622116374969482,
      "epoch": 3.91600145137881,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003538325233499532,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22821482121944428,
      "num_tokens": 89591806.0,
      "step": 43170
    },
    {
      "entropy": 5.617962169647217,
      "epoch": 3.9164550072568938,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035380221710506636,
      "loss": 4.8681,
      "mean_token_accuracy": 0.22724343836307526,
      "num_tokens": 89601845.0,
      "step": 43175
    },
    {
      "entropy": 5.690812969207764,
      "epoch": 3.916908563134978,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035377190923052964,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22130796015262605,
      "num_tokens": 89612331.0,
      "step": 43180
    },
    {
      "entropy": 5.694004774093628,
      "epoch": 3.9173621190130623,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035374159972696966,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2182876229286194,
      "num_tokens": 89622631.0,
      "step": 43185
    },
    {
      "entropy": 5.684329986572266,
      "epoch": 3.9178156748911466,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003537112885950133,
      "loss": 4.9428,
      "mean_token_accuracy": 0.2195568785071373,
      "num_tokens": 89633065.0,
      "step": 43190
    },
    {
      "entropy": 5.696789741516113,
      "epoch": 3.918269230769231,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035368097583528756,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2251499831676483,
      "num_tokens": 89643451.0,
      "step": 43195
    },
    {
      "entropy": 5.771214723587036,
      "epoch": 3.918722786647315,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003536506614484192,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22323312610387802,
      "num_tokens": 89654786.0,
      "step": 43200
    },
    {
      "entropy": 5.7175980567932125,
      "epoch": 3.9191763425253994,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003536203454350351,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21938296854496003,
      "num_tokens": 89665459.0,
      "step": 43205
    },
    {
      "entropy": 5.628254461288452,
      "epoch": 3.919629898403483,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003535900277957621,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22421299070119857,
      "num_tokens": 89675149.0,
      "step": 43210
    },
    {
      "entropy": 5.638657760620117,
      "epoch": 3.9200834542815675,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035355970853122743,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2270113617181778,
      "num_tokens": 89684994.0,
      "step": 43215
    },
    {
      "entropy": 5.749697017669678,
      "epoch": 3.9205370101596517,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035352938764205793,
      "loss": 5.0245,
      "mean_token_accuracy": 0.2057173654437065,
      "num_tokens": 89694751.0,
      "step": 43220
    },
    {
      "entropy": 5.684262371063232,
      "epoch": 3.920990566037736,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003534990651288807,
      "loss": 4.8611,
      "mean_token_accuracy": 0.2194393590092659,
      "num_tokens": 89704976.0,
      "step": 43225
    },
    {
      "entropy": 5.644000768661499,
      "epoch": 3.92144412191582,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035346874099232274,
      "loss": 4.959,
      "mean_token_accuracy": 0.21563856154680253,
      "num_tokens": 89715863.0,
      "step": 43230
    },
    {
      "entropy": 5.663571834564209,
      "epoch": 3.921897677793904,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035343841523301125,
      "loss": 4.9104,
      "mean_token_accuracy": 0.2198469892144203,
      "num_tokens": 89725856.0,
      "step": 43235
    },
    {
      "entropy": 5.684086847305298,
      "epoch": 3.9223512336719883,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003534080878515733,
      "loss": 4.937,
      "mean_token_accuracy": 0.21983996629714966,
      "num_tokens": 89736345.0,
      "step": 43240
    },
    {
      "entropy": 5.7095537185668945,
      "epoch": 3.9228047895500726,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003533777588486362,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2216563880443573,
      "num_tokens": 89746960.0,
      "step": 43245
    },
    {
      "entropy": 5.665602207183838,
      "epoch": 3.923258345428157,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000353347428224827,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22010509073734283,
      "num_tokens": 89757373.0,
      "step": 43250
    },
    {
      "entropy": 5.673526763916016,
      "epoch": 3.923711901306241,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035331709598077304,
      "loss": 4.9132,
      "mean_token_accuracy": 0.21877120584249496,
      "num_tokens": 89767895.0,
      "step": 43255
    },
    {
      "entropy": 5.681426954269409,
      "epoch": 3.924165457184325,
      "grad_norm": 1.25,
      "learning_rate": 0.0003532867621171016,
      "loss": 4.8839,
      "mean_token_accuracy": 0.2201533630490303,
      "num_tokens": 89777512.0,
      "step": 43260
    },
    {
      "entropy": 5.67314715385437,
      "epoch": 3.9246190130624092,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035325642663443994,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2216028168797493,
      "num_tokens": 89787308.0,
      "step": 43265
    },
    {
      "entropy": 5.690020370483398,
      "epoch": 3.9250725689404935,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003532260895334155,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22531746327877045,
      "num_tokens": 89798384.0,
      "step": 43270
    },
    {
      "entropy": 5.710465955734253,
      "epoch": 3.9255261248185778,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035319575081465556,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22179538905620574,
      "num_tokens": 89808757.0,
      "step": 43275
    },
    {
      "entropy": 5.711591720581055,
      "epoch": 3.9259796806966616,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035316541047878753,
      "loss": 4.9239,
      "mean_token_accuracy": 0.21368564814329147,
      "num_tokens": 89818925.0,
      "step": 43280
    },
    {
      "entropy": 5.713443040847778,
      "epoch": 3.926433236574746,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003531350685264389,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22394015789031982,
      "num_tokens": 89828647.0,
      "step": 43285
    },
    {
      "entropy": 5.769444322586059,
      "epoch": 3.92688679245283,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035310472495823713,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21058460026979448,
      "num_tokens": 89838784.0,
      "step": 43290
    },
    {
      "entropy": 5.673107624053955,
      "epoch": 3.9273403483309144,
      "grad_norm": 1.25,
      "learning_rate": 0.0003530743797748098,
      "loss": 4.8812,
      "mean_token_accuracy": 0.21748734265565872,
      "num_tokens": 89848464.0,
      "step": 43295
    },
    {
      "entropy": 5.610161161422729,
      "epoch": 3.9277939042089987,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003530440329767844,
      "loss": 4.8669,
      "mean_token_accuracy": 0.21795052140951157,
      "num_tokens": 89859594.0,
      "step": 43300
    },
    {
      "entropy": 5.659009504318237,
      "epoch": 3.928247460087083,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003530136845647885,
      "loss": 4.8458,
      "mean_token_accuracy": 0.2207550436258316,
      "num_tokens": 89869869.0,
      "step": 43305
    },
    {
      "entropy": 5.746286869049072,
      "epoch": 3.9287010159651667,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035298333453944976,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2237633764743805,
      "num_tokens": 89879604.0,
      "step": 43310
    },
    {
      "entropy": 5.574654054641724,
      "epoch": 3.929154571843251,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003529529829013957,
      "loss": 4.7503,
      "mean_token_accuracy": 0.232356995344162,
      "num_tokens": 89889200.0,
      "step": 43315
    },
    {
      "entropy": 5.669319725036621,
      "epoch": 3.9296081277213353,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003529226296512541,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2268512010574341,
      "num_tokens": 89898525.0,
      "step": 43320
    },
    {
      "entropy": 5.622125482559204,
      "epoch": 3.9300616835994195,
      "grad_norm": 1.4375,
      "learning_rate": 0.00035289227478965267,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2150866523385048,
      "num_tokens": 89909053.0,
      "step": 43325
    },
    {
      "entropy": 5.593583869934082,
      "epoch": 3.9305152394775034,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003528619183172192,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22917849868535994,
      "num_tokens": 89918375.0,
      "step": 43330
    },
    {
      "entropy": 5.759978628158569,
      "epoch": 3.9309687953555876,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003528315602345814,
      "loss": 4.956,
      "mean_token_accuracy": 0.22752957791090012,
      "num_tokens": 89928252.0,
      "step": 43335
    },
    {
      "entropy": 5.651673460006714,
      "epoch": 3.931422351233672,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003528012005423671,
      "loss": 4.8816,
      "mean_token_accuracy": 0.21788123697042466,
      "num_tokens": 89938613.0,
      "step": 43340
    },
    {
      "entropy": 5.723757791519165,
      "epoch": 3.931875907111756,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035277083924120406,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22394309490919112,
      "num_tokens": 89948781.0,
      "step": 43345
    },
    {
      "entropy": 5.651911830902099,
      "epoch": 3.9323294629898404,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035274047633172037,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21476895809173585,
      "num_tokens": 89959928.0,
      "step": 43350
    },
    {
      "entropy": 5.717375087738037,
      "epoch": 3.9327830188679247,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003527101118145437,
      "loss": 4.941,
      "mean_token_accuracy": 0.21978844404220582,
      "num_tokens": 89970756.0,
      "step": 43355
    },
    {
      "entropy": 5.69220814704895,
      "epoch": 3.933236574746009,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003526797456903022,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2195417046546936,
      "num_tokens": 89981332.0,
      "step": 43360
    },
    {
      "entropy": 5.636444473266602,
      "epoch": 3.933690130624093,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035264937795962374,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22601353675127028,
      "num_tokens": 89991427.0,
      "step": 43365
    },
    {
      "entropy": 5.666295671463013,
      "epoch": 3.934143686502177,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003526190086231364,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22677917182445526,
      "num_tokens": 90001599.0,
      "step": 43370
    },
    {
      "entropy": 5.683712816238403,
      "epoch": 3.9345972423802613,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035258863768146816,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22446888238191604,
      "num_tokens": 90011932.0,
      "step": 43375
    },
    {
      "entropy": 5.6849061965942385,
      "epoch": 3.9350507982583456,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003525582651352471,
      "loss": 4.9835,
      "mean_token_accuracy": 0.22281498908996583,
      "num_tokens": 90022123.0,
      "step": 43380
    },
    {
      "entropy": 5.595370864868164,
      "epoch": 3.9355043541364294,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003525278909851013,
      "loss": 4.8297,
      "mean_token_accuracy": 0.23288505971431733,
      "num_tokens": 90033024.0,
      "step": 43385
    },
    {
      "entropy": 5.658297157287597,
      "epoch": 3.9359579100145137,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000352497515231659,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22497707605361938,
      "num_tokens": 90043742.0,
      "step": 43390
    },
    {
      "entropy": 5.640840816497803,
      "epoch": 3.936411465892598,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003524671378755484,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22248139083385468,
      "num_tokens": 90054508.0,
      "step": 43395
    },
    {
      "entropy": 5.677954006195068,
      "epoch": 3.936865021770682,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003524367589173976,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2262275755405426,
      "num_tokens": 90065058.0,
      "step": 43400
    },
    {
      "entropy": 5.664167881011963,
      "epoch": 3.9373185776487665,
      "grad_norm": 1.453125,
      "learning_rate": 0.00035240637835783485,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22082047611474992,
      "num_tokens": 90075813.0,
      "step": 43405
    },
    {
      "entropy": 5.687118291854858,
      "epoch": 3.9377721335268507,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035237599619748853,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2143964573740959,
      "num_tokens": 90087064.0,
      "step": 43410
    },
    {
      "entropy": 5.656747436523437,
      "epoch": 3.9382256894049346,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003523456124369868,
      "loss": 4.8342,
      "mean_token_accuracy": 0.23068393021821976,
      "num_tokens": 90096668.0,
      "step": 43415
    },
    {
      "entropy": 5.613543844223022,
      "epoch": 3.938679245283019,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003523152270769582,
      "loss": 4.7683,
      "mean_token_accuracy": 0.23296170085668563,
      "num_tokens": 90105569.0,
      "step": 43420
    },
    {
      "entropy": 5.647732496261597,
      "epoch": 3.939132801161103,
      "grad_norm": 1.296875,
      "learning_rate": 0.000352284840118031,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22127506881952286,
      "num_tokens": 90116040.0,
      "step": 43425
    },
    {
      "entropy": 5.67569146156311,
      "epoch": 3.9395863570391874,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003522544515608335,
      "loss": 4.9035,
      "mean_token_accuracy": 0.2222743347287178,
      "num_tokens": 90125929.0,
      "step": 43430
    },
    {
      "entropy": 5.78849425315857,
      "epoch": 3.940039912917271,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003522240614059944,
      "loss": 4.9747,
      "mean_token_accuracy": 0.22025911659002304,
      "num_tokens": 90136251.0,
      "step": 43435
    },
    {
      "entropy": 5.67856650352478,
      "epoch": 3.9404934687953554,
      "grad_norm": 1.28125,
      "learning_rate": 0.000352193669654142,
      "loss": 4.8473,
      "mean_token_accuracy": 0.22875205427408218,
      "num_tokens": 90146480.0,
      "step": 43440
    },
    {
      "entropy": 5.670795869827271,
      "epoch": 3.9409470246734397,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035216327630590476,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2215244874358177,
      "num_tokens": 90157192.0,
      "step": 43445
    },
    {
      "entropy": 5.684213590621948,
      "epoch": 3.941400580551524,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035213288136191136,
      "loss": 4.8804,
      "mean_token_accuracy": 0.21882251352071763,
      "num_tokens": 90167120.0,
      "step": 43450
    },
    {
      "entropy": 5.676723480224609,
      "epoch": 3.9418541364296082,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035210248482279034,
      "loss": 4.9294,
      "mean_token_accuracy": 0.2198939338326454,
      "num_tokens": 90176276.0,
      "step": 43455
    },
    {
      "entropy": 5.659224128723144,
      "epoch": 3.9423076923076925,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035207208668917026,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22431150525808335,
      "num_tokens": 90187150.0,
      "step": 43460
    },
    {
      "entropy": 5.602003812789917,
      "epoch": 3.9427612481857763,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035204168696167975,
      "loss": 4.8263,
      "mean_token_accuracy": 0.22026303112506868,
      "num_tokens": 90198112.0,
      "step": 43465
    },
    {
      "entropy": 5.671118688583374,
      "epoch": 3.9432148040638606,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035201128564094753,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21844212859869003,
      "num_tokens": 90208940.0,
      "step": 43470
    },
    {
      "entropy": 5.6670006275177,
      "epoch": 3.943668359941945,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003519808827276023,
      "loss": 4.89,
      "mean_token_accuracy": 0.2276556134223938,
      "num_tokens": 90220073.0,
      "step": 43475
    },
    {
      "entropy": 5.65039587020874,
      "epoch": 3.944121915820029,
      "grad_norm": 1.25,
      "learning_rate": 0.00035195047822227283,
      "loss": 4.9362,
      "mean_token_accuracy": 0.21646596640348434,
      "num_tokens": 90230059.0,
      "step": 43480
    },
    {
      "entropy": 5.617595767974853,
      "epoch": 3.944575471698113,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035192007212558787,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22388505637645723,
      "num_tokens": 90240486.0,
      "step": 43485
    },
    {
      "entropy": 5.744507598876953,
      "epoch": 3.9450290275761972,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003518896644381762,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21825908869504929,
      "num_tokens": 90250014.0,
      "step": 43490
    },
    {
      "entropy": 5.660885000228882,
      "epoch": 3.9454825834542815,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003518592551606666,
      "loss": 4.8715,
      "mean_token_accuracy": 0.21523140668869017,
      "num_tokens": 90261040.0,
      "step": 43495
    },
    {
      "entropy": 5.700621795654297,
      "epoch": 3.9459361393323658,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003518288442936881,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2216187447309494,
      "num_tokens": 90271243.0,
      "step": 43500
    },
    {
      "entropy": 5.673984575271606,
      "epoch": 3.94638969521045,
      "grad_norm": 1.375,
      "learning_rate": 0.0003517984318378695,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22099606990814208,
      "num_tokens": 90281176.0,
      "step": 43505
    },
    {
      "entropy": 5.754955577850342,
      "epoch": 3.9468432510885343,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003517680177938398,
      "loss": 4.976,
      "mean_token_accuracy": 0.21949897557497025,
      "num_tokens": 90292465.0,
      "step": 43510
    },
    {
      "entropy": 5.696611213684082,
      "epoch": 3.9472968069666186,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003517376021622279,
      "loss": 4.8325,
      "mean_token_accuracy": 0.22815789133310319,
      "num_tokens": 90303482.0,
      "step": 43515
    },
    {
      "entropy": 5.661256551742554,
      "epoch": 3.9477503628447024,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003517071849436629,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2200719103217125,
      "num_tokens": 90313696.0,
      "step": 43520
    },
    {
      "entropy": 5.665653133392334,
      "epoch": 3.9482039187227866,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035167676613877366,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2145412966609001,
      "num_tokens": 90324564.0,
      "step": 43525
    },
    {
      "entropy": 5.635372877120972,
      "epoch": 3.948657474600871,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003516463457481894,
      "loss": 4.793,
      "mean_token_accuracy": 0.23202825635671614,
      "num_tokens": 90334440.0,
      "step": 43530
    },
    {
      "entropy": 5.7076925277709964,
      "epoch": 3.9491110304789547,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003516159237725392,
      "loss": 4.9294,
      "mean_token_accuracy": 0.22238634526729584,
      "num_tokens": 90344029.0,
      "step": 43535
    },
    {
      "entropy": 5.718100929260254,
      "epoch": 3.949564586357039,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003515855002124521,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2175098955631256,
      "num_tokens": 90354628.0,
      "step": 43540
    },
    {
      "entropy": 5.672612094879151,
      "epoch": 3.9500181422351233,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003515550750685573,
      "loss": 4.8083,
      "mean_token_accuracy": 0.2323197677731514,
      "num_tokens": 90365091.0,
      "step": 43545
    },
    {
      "entropy": 5.668745040893555,
      "epoch": 3.9504716981132075,
      "grad_norm": 1.25,
      "learning_rate": 0.0003515246483414842,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21860220283269882,
      "num_tokens": 90374197.0,
      "step": 43550
    },
    {
      "entropy": 5.6756429195404055,
      "epoch": 3.950925253991292,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035149422003186176,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22352002710103988,
      "num_tokens": 90384686.0,
      "step": 43555
    },
    {
      "entropy": 5.725683355331421,
      "epoch": 3.951378809869376,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035146379014031943,
      "loss": 4.9734,
      "mean_token_accuracy": 0.22019983381032943,
      "num_tokens": 90396329.0,
      "step": 43560
    },
    {
      "entropy": 5.722560596466065,
      "epoch": 3.9518323657474603,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003514333586674864,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2203424468636513,
      "num_tokens": 90408957.0,
      "step": 43565
    },
    {
      "entropy": 5.620771789550782,
      "epoch": 3.952285921625544,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000351402925613992,
      "loss": 4.8629,
      "mean_token_accuracy": 0.21978174299001693,
      "num_tokens": 90420194.0,
      "step": 43570
    },
    {
      "entropy": 5.629699087142944,
      "epoch": 3.9527394775036284,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003513724909804656,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22595855444669724,
      "num_tokens": 90430881.0,
      "step": 43575
    },
    {
      "entropy": 5.629531669616699,
      "epoch": 3.9531930333817127,
      "grad_norm": 1.375,
      "learning_rate": 0.00035134205476753674,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21324550360441208,
      "num_tokens": 90440713.0,
      "step": 43580
    },
    {
      "entropy": 5.670223999023437,
      "epoch": 3.953646589259797,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003513116169758347,
      "loss": 4.9597,
      "mean_token_accuracy": 0.21602357625961305,
      "num_tokens": 90451274.0,
      "step": 43585
    },
    {
      "entropy": 5.744808530807495,
      "epoch": 3.9541001451378808,
      "grad_norm": 1.25,
      "learning_rate": 0.0003512811776059889,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2171168878674507,
      "num_tokens": 90461503.0,
      "step": 43590
    },
    {
      "entropy": 5.665510654449463,
      "epoch": 3.954553701015965,
      "grad_norm": 1.34375,
      "learning_rate": 0.000351250736658629,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23298123180866243,
      "num_tokens": 90471178.0,
      "step": 43595
    },
    {
      "entropy": 5.632181882858276,
      "epoch": 3.9550072568940493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003512202941343844,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22214230448007583,
      "num_tokens": 90481573.0,
      "step": 43600
    },
    {
      "entropy": 5.688739395141601,
      "epoch": 3.9554608127721336,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003511898500338847,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21614936739206314,
      "num_tokens": 90491853.0,
      "step": 43605
    },
    {
      "entropy": 5.676394081115722,
      "epoch": 3.955914368650218,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003511594043577594,
      "loss": 4.9399,
      "mean_token_accuracy": 0.21660227179527283,
      "num_tokens": 90502958.0,
      "step": 43610
    },
    {
      "entropy": 5.619556093215943,
      "epoch": 3.956367924528302,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035112895710663833,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22116641998291015,
      "num_tokens": 90512555.0,
      "step": 43615
    },
    {
      "entropy": 5.6964256286621096,
      "epoch": 3.956821480406386,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035109850828115105,
      "loss": 4.9282,
      "mean_token_accuracy": 0.2234891787171364,
      "num_tokens": 90522722.0,
      "step": 43620
    },
    {
      "entropy": 5.661873483657837,
      "epoch": 3.95727503628447,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035106805788192707,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2224666103720665,
      "num_tokens": 90533182.0,
      "step": 43625
    },
    {
      "entropy": 5.701808261871338,
      "epoch": 3.9577285921625545,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003510376059095963,
      "loss": 4.9659,
      "mean_token_accuracy": 0.2112626001238823,
      "num_tokens": 90543495.0,
      "step": 43630
    },
    {
      "entropy": 5.626921081542969,
      "epoch": 3.9581821480406387,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003510071523647886,
      "loss": 4.8225,
      "mean_token_accuracy": 0.22718972116708755,
      "num_tokens": 90553680.0,
      "step": 43635
    },
    {
      "entropy": 5.630295085906982,
      "epoch": 3.9586357039187225,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003509766972481336,
      "loss": 4.8209,
      "mean_token_accuracy": 0.22706011831760406,
      "num_tokens": 90563513.0,
      "step": 43640
    },
    {
      "entropy": 5.763727140426636,
      "epoch": 3.959089259796807,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000350946240560261,
      "loss": 4.9522,
      "mean_token_accuracy": 0.22226848006248473,
      "num_tokens": 90573337.0,
      "step": 43645
    },
    {
      "entropy": 5.641994285583496,
      "epoch": 3.959542815674891,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003509157823018009,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22536929994821547,
      "num_tokens": 90582979.0,
      "step": 43650
    },
    {
      "entropy": 5.731965208053589,
      "epoch": 3.9599963715529753,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035088532247338306,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21136672645807267,
      "num_tokens": 90593552.0,
      "step": 43655
    },
    {
      "entropy": 5.705495691299438,
      "epoch": 3.9604499274310596,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035085486107563733,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2216441199183464,
      "num_tokens": 90603813.0,
      "step": 43660
    },
    {
      "entropy": 5.7087073802948,
      "epoch": 3.960903483309144,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035082439810919385,
      "loss": 4.8547,
      "mean_token_accuracy": 0.23074931204319,
      "num_tokens": 90613885.0,
      "step": 43665
    },
    {
      "entropy": 5.677077341079712,
      "epoch": 3.9613570391872277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003507939335746824,
      "loss": 4.8061,
      "mean_token_accuracy": 0.2214651510119438,
      "num_tokens": 90624665.0,
      "step": 43670
    },
    {
      "entropy": 5.646004390716553,
      "epoch": 3.961810595065312,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003507634674727332,
      "loss": 4.8727,
      "mean_token_accuracy": 0.22370633184909822,
      "num_tokens": 90635496.0,
      "step": 43675
    },
    {
      "entropy": 5.697833728790283,
      "epoch": 3.9622641509433962,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003507329998039762,
      "loss": 4.8774,
      "mean_token_accuracy": 0.2209860160946846,
      "num_tokens": 90645502.0,
      "step": 43680
    },
    {
      "entropy": 5.580631494522095,
      "epoch": 3.9627177068214805,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035070253056904133,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23052534013986586,
      "num_tokens": 90654724.0,
      "step": 43685
    },
    {
      "entropy": 5.622163105010986,
      "epoch": 3.9631712626995643,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003506720597685589,
      "loss": 4.8671,
      "mean_token_accuracy": 0.227325701713562,
      "num_tokens": 90665280.0,
      "step": 43690
    },
    {
      "entropy": 5.665242767333984,
      "epoch": 3.9636248185776486,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035064158740315894,
      "loss": 4.9233,
      "mean_token_accuracy": 0.221587111055851,
      "num_tokens": 90675804.0,
      "step": 43695
    },
    {
      "entropy": 5.6679527759552,
      "epoch": 3.964078374455733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035061111347347174,
      "loss": 4.8509,
      "mean_token_accuracy": 0.2242763251066208,
      "num_tokens": 90686636.0,
      "step": 43700
    },
    {
      "entropy": 5.640151596069336,
      "epoch": 3.964531930333817,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035058063798012746,
      "loss": 4.885,
      "mean_token_accuracy": 0.22426643520593642,
      "num_tokens": 90697666.0,
      "step": 43705
    },
    {
      "entropy": 5.727765130996704,
      "epoch": 3.9649854862119014,
      "grad_norm": 1.828125,
      "learning_rate": 0.0003505501609237563,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21919141411781312,
      "num_tokens": 90708111.0,
      "step": 43710
    },
    {
      "entropy": 5.671432590484619,
      "epoch": 3.9654390420899857,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035051968230498853,
      "loss": 4.847,
      "mean_token_accuracy": 0.22903651744127274,
      "num_tokens": 90718711.0,
      "step": 43715
    },
    {
      "entropy": 5.616025638580322,
      "epoch": 3.96589259796807,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035048920212445454,
      "loss": 4.8329,
      "mean_token_accuracy": 0.231365367770195,
      "num_tokens": 90729585.0,
      "step": 43720
    },
    {
      "entropy": 5.5750213146209715,
      "epoch": 3.9663461538461537,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035045872038278467,
      "loss": 4.7708,
      "mean_token_accuracy": 0.23267118334770204,
      "num_tokens": 90739654.0,
      "step": 43725
    },
    {
      "entropy": 5.6397035121917725,
      "epoch": 3.966799709724238,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003504282370806091,
      "loss": 4.8447,
      "mean_token_accuracy": 0.22622293680906297,
      "num_tokens": 90749456.0,
      "step": 43730
    },
    {
      "entropy": 5.626939821243286,
      "epoch": 3.9672532656023223,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035039775221855845,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22617058604955673,
      "num_tokens": 90760051.0,
      "step": 43735
    },
    {
      "entropy": 5.60809326171875,
      "epoch": 3.9677068214804065,
      "grad_norm": 1.21875,
      "learning_rate": 0.000350367265797263,
      "loss": 4.826,
      "mean_token_accuracy": 0.22533978819847106,
      "num_tokens": 90770452.0,
      "step": 43740
    },
    {
      "entropy": 5.655235671997071,
      "epoch": 3.9681603773584904,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035033677781735336,
      "loss": 4.9061,
      "mean_token_accuracy": 0.22192625999450682,
      "num_tokens": 90780150.0,
      "step": 43745
    },
    {
      "entropy": 5.652624654769897,
      "epoch": 3.9686139332365746,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003503062882794599,
      "loss": 4.8216,
      "mean_token_accuracy": 0.23167807906866072,
      "num_tokens": 90790338.0,
      "step": 43750
    },
    {
      "entropy": 5.6471888542175295,
      "epoch": 3.969067489114659,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035027579718421327,
      "loss": 4.87,
      "mean_token_accuracy": 0.22395725697278976,
      "num_tokens": 90800195.0,
      "step": 43755
    },
    {
      "entropy": 5.678124904632568,
      "epoch": 3.969521044992743,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003502453045322439,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21838459521532058,
      "num_tokens": 90810872.0,
      "step": 43760
    },
    {
      "entropy": 5.674132204055786,
      "epoch": 3.9699746008708274,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035021481032418254,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22320411801338197,
      "num_tokens": 90821018.0,
      "step": 43765
    },
    {
      "entropy": 5.6932634830474855,
      "epoch": 3.9704281567489117,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003501843145606597,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21695664674043655,
      "num_tokens": 90830274.0,
      "step": 43770
    },
    {
      "entropy": 5.631274795532226,
      "epoch": 3.9708817126269955,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000350153817242306,
      "loss": 4.7934,
      "mean_token_accuracy": 0.23322552293539048,
      "num_tokens": 90839926.0,
      "step": 43775
    },
    {
      "entropy": 5.576013135910034,
      "epoch": 3.97133526850508,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003501233183697523,
      "loss": 4.7865,
      "mean_token_accuracy": 0.22268013060092925,
      "num_tokens": 90850263.0,
      "step": 43780
    },
    {
      "entropy": 5.658271551132202,
      "epoch": 3.971788824383164,
      "grad_norm": 1.25,
      "learning_rate": 0.00035009281794362926,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21503333896398544,
      "num_tokens": 90861747.0,
      "step": 43785
    },
    {
      "entropy": 5.680157852172852,
      "epoch": 3.9722423802612483,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035006231596456755,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2119024872779846,
      "num_tokens": 90872186.0,
      "step": 43790
    },
    {
      "entropy": 5.69822244644165,
      "epoch": 3.972695936139332,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035003181243319797,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22887516766786575,
      "num_tokens": 90881088.0,
      "step": 43795
    },
    {
      "entropy": 5.593884086608886,
      "epoch": 3.9731494920174164,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003500013073501514,
      "loss": 4.844,
      "mean_token_accuracy": 0.22143300473690034,
      "num_tokens": 90890589.0,
      "step": 43800
    },
    {
      "entropy": 5.641145515441894,
      "epoch": 3.9736030478955007,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003499708007160588,
      "loss": 4.931,
      "mean_token_accuracy": 0.2198511466383934,
      "num_tokens": 90901692.0,
      "step": 43805
    },
    {
      "entropy": 5.7143457412719725,
      "epoch": 3.974056603773585,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003499402925315509,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2204224780201912,
      "num_tokens": 90911811.0,
      "step": 43810
    },
    {
      "entropy": 5.69916877746582,
      "epoch": 3.974510159651669,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034990978279725846,
      "loss": 4.7904,
      "mean_token_accuracy": 0.2313077464699745,
      "num_tokens": 90922323.0,
      "step": 43815
    },
    {
      "entropy": 5.784437036514282,
      "epoch": 3.9749637155297535,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034987927151381283,
      "loss": 5.0496,
      "mean_token_accuracy": 0.2066980391740799,
      "num_tokens": 90932834.0,
      "step": 43820
    },
    {
      "entropy": 5.698661184310913,
      "epoch": 3.9754172714078373,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034984875868184466,
      "loss": 4.8901,
      "mean_token_accuracy": 0.22844714522361756,
      "num_tokens": 90943616.0,
      "step": 43825
    },
    {
      "entropy": 5.697509860992431,
      "epoch": 3.9758708272859216,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034981824430198517,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21706928163766862,
      "num_tokens": 90954468.0,
      "step": 43830
    },
    {
      "entropy": 5.677630853652954,
      "epoch": 3.976324383164006,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034978772837486536,
      "loss": 4.8233,
      "mean_token_accuracy": 0.22597233206033707,
      "num_tokens": 90965109.0,
      "step": 43835
    },
    {
      "entropy": 5.6912651538848875,
      "epoch": 3.97677793904209,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034975721090111625,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22189621776342391,
      "num_tokens": 90974789.0,
      "step": 43840
    },
    {
      "entropy": 5.668655872344971,
      "epoch": 3.977231494920174,
      "grad_norm": 1.25,
      "learning_rate": 0.0003497266918813689,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2170195236802101,
      "num_tokens": 90985154.0,
      "step": 43845
    },
    {
      "entropy": 5.664903926849365,
      "epoch": 3.977685050798258,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003496961713162545,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22212358862161635,
      "num_tokens": 90996007.0,
      "step": 43850
    },
    {
      "entropy": 5.705722808837891,
      "epoch": 3.9781386066763424,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003496656492064043,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2135677695274353,
      "num_tokens": 91006253.0,
      "step": 43855
    },
    {
      "entropy": 5.736240339279175,
      "epoch": 3.9785921625544267,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003496351255524494,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21396485716104507,
      "num_tokens": 91017169.0,
      "step": 43860
    },
    {
      "entropy": 5.667433547973633,
      "epoch": 3.979045718432511,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034960460035502117,
      "loss": 4.8738,
      "mean_token_accuracy": 0.21900187134742738,
      "num_tokens": 91026684.0,
      "step": 43865
    },
    {
      "entropy": 5.659739971160889,
      "epoch": 3.9794992743105952,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003495740736147508,
      "loss": 4.8345,
      "mean_token_accuracy": 0.2249004766345024,
      "num_tokens": 91037462.0,
      "step": 43870
    },
    {
      "entropy": 5.682798624038696,
      "epoch": 3.9799528301886795,
      "grad_norm": 1.125,
      "learning_rate": 0.00034954354533226953,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2138870894908905,
      "num_tokens": 91049418.0,
      "step": 43875
    },
    {
      "entropy": 5.625389623641968,
      "epoch": 3.9804063860667633,
      "grad_norm": 1.25,
      "learning_rate": 0.0003495130155082087,
      "loss": 4.834,
      "mean_token_accuracy": 0.22384393513202666,
      "num_tokens": 91061013.0,
      "step": 43880
    },
    {
      "entropy": 5.659660053253174,
      "epoch": 3.9808599419448476,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003494824841431998,
      "loss": 4.8886,
      "mean_token_accuracy": 0.22629764080047607,
      "num_tokens": 91070974.0,
      "step": 43885
    },
    {
      "entropy": 5.687782430648804,
      "epoch": 3.981313497822932,
      "grad_norm": 1.25,
      "learning_rate": 0.0003494519512378741,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2256431683897972,
      "num_tokens": 91082681.0,
      "step": 43890
    },
    {
      "entropy": 5.665985488891602,
      "epoch": 3.9817670537010157,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034942141679286307,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23574488162994384,
      "num_tokens": 91093552.0,
      "step": 43895
    },
    {
      "entropy": 5.687515687942505,
      "epoch": 3.9822206095791,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003493908808087982,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2167578235268593,
      "num_tokens": 91105470.0,
      "step": 43900
    },
    {
      "entropy": 5.668826484680176,
      "epoch": 3.9826741654571842,
      "grad_norm": 1.140625,
      "learning_rate": 0.000349360343286311,
      "loss": 4.8814,
      "mean_token_accuracy": 0.22682953923940657,
      "num_tokens": 91115822.0,
      "step": 43905
    },
    {
      "entropy": 5.635017395019531,
      "epoch": 3.9831277213352685,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034932980422603283,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23062579184770585,
      "num_tokens": 91125722.0,
      "step": 43910
    },
    {
      "entropy": 5.78600378036499,
      "epoch": 3.9835812772133528,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003492992636285955,
      "loss": 5.034,
      "mean_token_accuracy": 0.2230457440018654,
      "num_tokens": 91135783.0,
      "step": 43915
    },
    {
      "entropy": 5.683391761779785,
      "epoch": 3.984034833091437,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003492687214946304,
      "loss": 4.8693,
      "mean_token_accuracy": 0.22641801983118057,
      "num_tokens": 91146151.0,
      "step": 43920
    },
    {
      "entropy": 5.671380138397216,
      "epoch": 3.9844883889695213,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034923817782476927,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21648330688476564,
      "num_tokens": 91156711.0,
      "step": 43925
    },
    {
      "entropy": 5.644983911514283,
      "epoch": 3.984941944847605,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003492076326196436,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21649480909109114,
      "num_tokens": 91166601.0,
      "step": 43930
    },
    {
      "entropy": 5.730920743942261,
      "epoch": 3.9853955007256894,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034917708587988523,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2093385711312294,
      "num_tokens": 91176556.0,
      "step": 43935
    },
    {
      "entropy": 5.655743741989136,
      "epoch": 3.9858490566037736,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034914653760612584,
      "loss": 4.864,
      "mean_token_accuracy": 0.22366306483745574,
      "num_tokens": 91186499.0,
      "step": 43940
    },
    {
      "entropy": 5.643047571182251,
      "epoch": 3.986302612481858,
      "grad_norm": 1.25,
      "learning_rate": 0.0003491159877989971,
      "loss": 4.8642,
      "mean_token_accuracy": 0.21299372166395186,
      "num_tokens": 91197088.0,
      "step": 43945
    },
    {
      "entropy": 5.645539665222168,
      "epoch": 3.9867561683599417,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000349085436459131,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22560027092695237,
      "num_tokens": 91207188.0,
      "step": 43950
    },
    {
      "entropy": 5.643137979507446,
      "epoch": 3.987209724238026,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034905488358715907,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22218265235424042,
      "num_tokens": 91217349.0,
      "step": 43955
    },
    {
      "entropy": 5.699260997772217,
      "epoch": 3.9876632801161103,
      "grad_norm": 1.25,
      "learning_rate": 0.0003490243291837133,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22137006670236586,
      "num_tokens": 91227775.0,
      "step": 43960
    },
    {
      "entropy": 5.659594202041626,
      "epoch": 3.9881168359941945,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034899377324942556,
      "loss": 4.7973,
      "mean_token_accuracy": 0.232569183409214,
      "num_tokens": 91238315.0,
      "step": 43965
    },
    {
      "entropy": 5.673978233337403,
      "epoch": 3.988570391872279,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034896321578492767,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21239070743322372,
      "num_tokens": 91249178.0,
      "step": 43970
    },
    {
      "entropy": 5.634776592254639,
      "epoch": 3.989023947750363,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003489326567908516,
      "loss": 4.8647,
      "mean_token_accuracy": 0.22213993966579437,
      "num_tokens": 91257911.0,
      "step": 43975
    },
    {
      "entropy": 5.66892147064209,
      "epoch": 3.989477503628447,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034890209626782943,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22171754539012908,
      "num_tokens": 91267491.0,
      "step": 43980
    },
    {
      "entropy": 5.662269067764282,
      "epoch": 3.989931059506531,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003488715342164931,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22416286021471024,
      "num_tokens": 91278292.0,
      "step": 43985
    },
    {
      "entropy": 5.711904764175415,
      "epoch": 3.9903846153846154,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00034884097063747453,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21531758159399034,
      "num_tokens": 91287871.0,
      "step": 43990
    },
    {
      "entropy": 5.62492790222168,
      "epoch": 3.9908381712626997,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003488104055314059,
      "loss": 4.8237,
      "mean_token_accuracy": 0.2305402785539627,
      "num_tokens": 91297201.0,
      "step": 43995
    },
    {
      "entropy": 5.584606409072876,
      "epoch": 3.9912917271407835,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003487798388989193,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23055097311735154,
      "num_tokens": 91306449.0,
      "step": 44000
    },
    {
      "entropy": 5.57442045211792,
      "epoch": 3.9917452830188678,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034874927074064684,
      "loss": 4.7969,
      "mean_token_accuracy": 0.22127773016691207,
      "num_tokens": 91317516.0,
      "step": 44005
    },
    {
      "entropy": 5.664388370513916,
      "epoch": 3.992198838896952,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003487187010572206,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21966809183359146,
      "num_tokens": 91326850.0,
      "step": 44010
    },
    {
      "entropy": 5.669976854324341,
      "epoch": 3.9926523947750363,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003486881298492729,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22310503870248793,
      "num_tokens": 91338068.0,
      "step": 44015
    },
    {
      "entropy": 5.746202421188355,
      "epoch": 3.9931059506531206,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003486575571174359,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22514592111110687,
      "num_tokens": 91349627.0,
      "step": 44020
    },
    {
      "entropy": 5.686382913589478,
      "epoch": 3.993559506531205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034862698286234185,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21305015087127685,
      "num_tokens": 91361229.0,
      "step": 44025
    },
    {
      "entropy": 5.679896306991577,
      "epoch": 3.9940130624092887,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003485964070846229,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2128543496131897,
      "num_tokens": 91372348.0,
      "step": 44030
    },
    {
      "entropy": 5.627528333663941,
      "epoch": 3.994466618287373,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003485658297849117,
      "loss": 4.8377,
      "mean_token_accuracy": 0.22429872751235963,
      "num_tokens": 91383028.0,
      "step": 44035
    },
    {
      "entropy": 5.688759613037109,
      "epoch": 3.994920174165457,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034853525096384034,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2189745455980301,
      "num_tokens": 91393263.0,
      "step": 44040
    },
    {
      "entropy": 5.724278926849365,
      "epoch": 3.9953737300435415,
      "grad_norm": 1.375,
      "learning_rate": 0.0003485046706220411,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22322145402431487,
      "num_tokens": 91403881.0,
      "step": 44045
    },
    {
      "entropy": 5.705347490310669,
      "epoch": 3.9958272859216253,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003484740887601467,
      "loss": 4.9054,
      "mean_token_accuracy": 0.21835314631462097,
      "num_tokens": 91413723.0,
      "step": 44050
    },
    {
      "entropy": 5.694570159912109,
      "epoch": 3.9962808417997095,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034844350537878937,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22168629765510559,
      "num_tokens": 91424168.0,
      "step": 44055
    },
    {
      "entropy": 5.757430505752564,
      "epoch": 3.996734397677794,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034841292047860164,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21228594034910203,
      "num_tokens": 91433133.0,
      "step": 44060
    },
    {
      "entropy": 5.692979335784912,
      "epoch": 3.997187953555878,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000348382334060216,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20999932438135147,
      "num_tokens": 91444521.0,
      "step": 44065
    },
    {
      "entropy": 5.597850227355957,
      "epoch": 3.9976415094339623,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000348351746124265,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22766720056533812,
      "num_tokens": 91454870.0,
      "step": 44070
    },
    {
      "entropy": 5.679509258270263,
      "epoch": 3.9980950653120466,
      "grad_norm": 1.25,
      "learning_rate": 0.0003483211566713812,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22879617810249328,
      "num_tokens": 91465472.0,
      "step": 44075
    },
    {
      "entropy": 5.7269110679626465,
      "epoch": 3.998548621190131,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003482905657021971,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2221173882484436,
      "num_tokens": 91475467.0,
      "step": 44080
    },
    {
      "entropy": 5.728068685531616,
      "epoch": 3.9990021770682147,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003482599732173456,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22487736046314238,
      "num_tokens": 91485615.0,
      "step": 44085
    },
    {
      "entropy": 5.6725341796875,
      "epoch": 3.999455732946299,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003482293792174591,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22877063304185868,
      "num_tokens": 91495790.0,
      "step": 44090
    },
    {
      "entropy": 5.665362310409546,
      "epoch": 3.9999092888243832,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003481987837031703,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21953017264604568,
      "num_tokens": 91507164.0,
      "step": 44095
    },
    {
      "entropy": 5.62367000579834,
      "epoch": 4.000362844702467,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003481681866751121,
      "loss": 4.8182,
      "mean_token_accuracy": 0.23152355700731278,
      "num_tokens": 91516732.0,
      "step": 44100
    },
    {
      "entropy": 5.609641122817993,
      "epoch": 4.000816400580551,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034813758813391716,
      "loss": 4.7843,
      "mean_token_accuracy": 0.2339000418782234,
      "num_tokens": 91527073.0,
      "step": 44105
    },
    {
      "entropy": 5.639034366607666,
      "epoch": 4.001269956458636,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003481069880802183,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2266480579972267,
      "num_tokens": 91538314.0,
      "step": 44110
    },
    {
      "entropy": 5.668161392211914,
      "epoch": 4.00172351233672,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034807638651464815,
      "loss": 4.8761,
      "mean_token_accuracy": 0.2249499797821045,
      "num_tokens": 91549159.0,
      "step": 44115
    },
    {
      "entropy": 5.660141658782959,
      "epoch": 4.002177068214804,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003480457834378398,
      "loss": 4.8665,
      "mean_token_accuracy": 0.22389106452465057,
      "num_tokens": 91560901.0,
      "step": 44120
    },
    {
      "entropy": 5.65052752494812,
      "epoch": 4.002630624092888,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034801517885042597,
      "loss": 4.8311,
      "mean_token_accuracy": 0.22240670621395112,
      "num_tokens": 91570797.0,
      "step": 44125
    },
    {
      "entropy": 5.754239320755005,
      "epoch": 4.003084179970973,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003479845727530397,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21522759944200515,
      "num_tokens": 91580971.0,
      "step": 44130
    },
    {
      "entropy": 5.6647007942199705,
      "epoch": 4.003537735849057,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034795396514631386,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23024609982967376,
      "num_tokens": 91590834.0,
      "step": 44135
    },
    {
      "entropy": 5.658629846572876,
      "epoch": 4.003991291727141,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034792335603088135,
      "loss": 4.8981,
      "mean_token_accuracy": 0.21986954659223557,
      "num_tokens": 91602099.0,
      "step": 44140
    },
    {
      "entropy": 5.632980442047119,
      "epoch": 4.004444847605225,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034789274540737535,
      "loss": 4.7949,
      "mean_token_accuracy": 0.23817332834005356,
      "num_tokens": 91612269.0,
      "step": 44145
    },
    {
      "entropy": 5.715988397598267,
      "epoch": 4.004898403483309,
      "grad_norm": 1.25,
      "learning_rate": 0.00034786213327642874,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22462040036916733,
      "num_tokens": 91623644.0,
      "step": 44150
    },
    {
      "entropy": 5.682221460342407,
      "epoch": 4.005351959361393,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003478315196386747,
      "loss": 4.835,
      "mean_token_accuracy": 0.2328446999192238,
      "num_tokens": 91633563.0,
      "step": 44155
    },
    {
      "entropy": 5.629967498779297,
      "epoch": 4.005805515239477,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003478009044947463,
      "loss": 4.8112,
      "mean_token_accuracy": 0.2278311640024185,
      "num_tokens": 91644648.0,
      "step": 44160
    },
    {
      "entropy": 5.57977819442749,
      "epoch": 4.006259071117562,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003477702878452766,
      "loss": 4.6937,
      "mean_token_accuracy": 0.2477676525712013,
      "num_tokens": 91655366.0,
      "step": 44165
    },
    {
      "entropy": 5.607641983032226,
      "epoch": 4.006712626995646,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003477396696908988,
      "loss": 4.8221,
      "mean_token_accuracy": 0.2301895424723625,
      "num_tokens": 91665474.0,
      "step": 44170
    },
    {
      "entropy": 5.668443250656128,
      "epoch": 4.00716618287373,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034770905003224606,
      "loss": 4.8534,
      "mean_token_accuracy": 0.21977997571229935,
      "num_tokens": 91675291.0,
      "step": 44175
    },
    {
      "entropy": 5.586932802200318,
      "epoch": 4.007619738751814,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003476784288699517,
      "loss": 4.7646,
      "mean_token_accuracy": 0.23321352750062943,
      "num_tokens": 91686472.0,
      "step": 44180
    },
    {
      "entropy": 5.684142255783081,
      "epoch": 4.008073294629899,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034764780620464886,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22337005585432052,
      "num_tokens": 91696347.0,
      "step": 44185
    },
    {
      "entropy": 5.63988242149353,
      "epoch": 4.008526850507983,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003476171820369709,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2208269640803337,
      "num_tokens": 91706504.0,
      "step": 44190
    },
    {
      "entropy": 5.6283097743988035,
      "epoch": 4.008980406386066,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034758655636755103,
      "loss": 4.7959,
      "mean_token_accuracy": 0.2407778799533844,
      "num_tokens": 91716559.0,
      "step": 44195
    },
    {
      "entropy": 5.686963653564453,
      "epoch": 4.009433962264151,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034755592919702275,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22040775269269944,
      "num_tokens": 91726312.0,
      "step": 44200
    },
    {
      "entropy": 5.61539945602417,
      "epoch": 4.009887518142235,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034752530052601944,
      "loss": 4.7939,
      "mean_token_accuracy": 0.2281578853726387,
      "num_tokens": 91736882.0,
      "step": 44205
    },
    {
      "entropy": 5.643258762359619,
      "epoch": 4.010341074020319,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034749467035517437,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22000845968723298,
      "num_tokens": 91746905.0,
      "step": 44210
    },
    {
      "entropy": 5.6665655136108395,
      "epoch": 4.010794629898403,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034746403868512094,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22082189470529556,
      "num_tokens": 91757609.0,
      "step": 44215
    },
    {
      "entropy": 5.690264177322388,
      "epoch": 4.011248185776488,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034743340551649285,
      "loss": 4.8001,
      "mean_token_accuracy": 0.23034935891628266,
      "num_tokens": 91768017.0,
      "step": 44220
    },
    {
      "entropy": 5.5815088748931885,
      "epoch": 4.011701741654572,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034740277084992346,
      "loss": 4.7273,
      "mean_token_accuracy": 0.23843913972377778,
      "num_tokens": 91777820.0,
      "step": 44225
    },
    {
      "entropy": 5.593440008163452,
      "epoch": 4.012155297532656,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034737213468604626,
      "loss": 4.8411,
      "mean_token_accuracy": 0.21970656961202623,
      "num_tokens": 91789011.0,
      "step": 44230
    },
    {
      "entropy": 5.677101802825928,
      "epoch": 4.0126088534107405,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003473414970254949,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22159491628408431,
      "num_tokens": 91799165.0,
      "step": 44235
    },
    {
      "entropy": 5.646427488327026,
      "epoch": 4.013062409288825,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034731085786890306,
      "loss": 4.8722,
      "mean_token_accuracy": 0.227496874332428,
      "num_tokens": 91809602.0,
      "step": 44240
    },
    {
      "entropy": 5.690667772293091,
      "epoch": 4.013515965166909,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034728021721690407,
      "loss": 4.9247,
      "mean_token_accuracy": 0.221859672665596,
      "num_tokens": 91820092.0,
      "step": 44245
    },
    {
      "entropy": 5.6781834125518795,
      "epoch": 4.013969521044992,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003472495750701318,
      "loss": 4.8442,
      "mean_token_accuracy": 0.2285470813512802,
      "num_tokens": 91829730.0,
      "step": 44250
    },
    {
      "entropy": 5.674529409408569,
      "epoch": 4.014423076923077,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034721893142922,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22150059938430786,
      "num_tokens": 91840697.0,
      "step": 44255
    },
    {
      "entropy": 5.661016130447388,
      "epoch": 4.014876632801161,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003471882862948023,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22147967517375947,
      "num_tokens": 91852145.0,
      "step": 44260
    },
    {
      "entropy": 5.680552291870117,
      "epoch": 4.015330188679245,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034715763966751233,
      "loss": 4.8282,
      "mean_token_accuracy": 0.22101673036813735,
      "num_tokens": 91861940.0,
      "step": 44265
    },
    {
      "entropy": 5.726807165145874,
      "epoch": 4.0157837445573294,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034712699154798403,
      "loss": 4.8699,
      "mean_token_accuracy": 0.22533690333366393,
      "num_tokens": 91872179.0,
      "step": 44270
    },
    {
      "entropy": 5.68064169883728,
      "epoch": 4.016237300435414,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034709634193685126,
      "loss": 4.8029,
      "mean_token_accuracy": 0.22966966181993484,
      "num_tokens": 91882051.0,
      "step": 44275
    },
    {
      "entropy": 5.63207516670227,
      "epoch": 4.016690856313498,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003470656908347476,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22183433622121812,
      "num_tokens": 91892672.0,
      "step": 44280
    },
    {
      "entropy": 5.590099620819092,
      "epoch": 4.017144412191582,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003470350382423072,
      "loss": 4.7558,
      "mean_token_accuracy": 0.23339740931987762,
      "num_tokens": 91903418.0,
      "step": 44285
    },
    {
      "entropy": 5.630136013031006,
      "epoch": 4.0175979680696665,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003470043841601639,
      "loss": 4.8398,
      "mean_token_accuracy": 0.23290229886770247,
      "num_tokens": 91913076.0,
      "step": 44290
    },
    {
      "entropy": 5.61550121307373,
      "epoch": 4.018051523947751,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034697372858895147,
      "loss": 4.8284,
      "mean_token_accuracy": 0.23611175268888474,
      "num_tokens": 91923464.0,
      "step": 44295
    },
    {
      "entropy": 5.697745132446289,
      "epoch": 4.018505079825834,
      "grad_norm": 1.328125,
      "learning_rate": 0.000346943071529304,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22109244018793106,
      "num_tokens": 91934399.0,
      "step": 44300
    },
    {
      "entropy": 5.742668533325196,
      "epoch": 4.018958635703918,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00034691241298185553,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22678199261426926,
      "num_tokens": 91945330.0,
      "step": 44305
    },
    {
      "entropy": 5.696223402023316,
      "epoch": 4.019412191582003,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034688175294724004,
      "loss": 4.8893,
      "mean_token_accuracy": 0.21869195699691774,
      "num_tokens": 91956379.0,
      "step": 44310
    },
    {
      "entropy": 5.552106237411499,
      "epoch": 4.019865747460087,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034685109142609154,
      "loss": 4.7101,
      "mean_token_accuracy": 0.2322893977165222,
      "num_tokens": 91965969.0,
      "step": 44315
    },
    {
      "entropy": 5.719584178924561,
      "epoch": 4.020319303338171,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003468204284190442,
      "loss": 4.9806,
      "mean_token_accuracy": 0.2136175036430359,
      "num_tokens": 91977005.0,
      "step": 44320
    },
    {
      "entropy": 5.689169931411743,
      "epoch": 4.0207728592162555,
      "grad_norm": 1.25,
      "learning_rate": 0.00034678976392673206,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22458529472351074,
      "num_tokens": 91988735.0,
      "step": 44325
    },
    {
      "entropy": 5.688523006439209,
      "epoch": 4.02122641509434,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034675909794978926,
      "loss": 4.8546,
      "mean_token_accuracy": 0.22859515845775605,
      "num_tokens": 91998743.0,
      "step": 44330
    },
    {
      "entropy": 5.681196832656861,
      "epoch": 4.021679970972424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034672843048885017,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22578403055667878,
      "num_tokens": 92009465.0,
      "step": 44335
    },
    {
      "entropy": 5.601125860214234,
      "epoch": 4.022133526850508,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034669776154454875,
      "loss": 4.8119,
      "mean_token_accuracy": 0.23315410763025285,
      "num_tokens": 92020818.0,
      "step": 44340
    },
    {
      "entropy": 5.672318077087402,
      "epoch": 4.0225870827285926,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034666709111751945,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22158267945051194,
      "num_tokens": 92031528.0,
      "step": 44345
    },
    {
      "entropy": 5.7074450016021725,
      "epoch": 4.023040638606676,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003466364192083963,
      "loss": 4.9113,
      "mean_token_accuracy": 0.22390609830617905,
      "num_tokens": 92041856.0,
      "step": 44350
    },
    {
      "entropy": 5.695340156555176,
      "epoch": 4.02349419448476,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034660574581781386,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23131307661533357,
      "num_tokens": 92052117.0,
      "step": 44355
    },
    {
      "entropy": 5.691923713684082,
      "epoch": 4.0239477503628445,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003465750709464063,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22301280349493027,
      "num_tokens": 92062405.0,
      "step": 44360
    },
    {
      "entropy": 5.674297904968261,
      "epoch": 4.024401306240929,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000346544394594808,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22742346674203873,
      "num_tokens": 92073269.0,
      "step": 44365
    },
    {
      "entropy": 5.643460416793824,
      "epoch": 4.024854862119013,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034651371676365343,
      "loss": 4.7971,
      "mean_token_accuracy": 0.23453737646341324,
      "num_tokens": 92083645.0,
      "step": 44370
    },
    {
      "entropy": 5.756990957260132,
      "epoch": 4.025308417997097,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034648303745357705,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22420720756053925,
      "num_tokens": 92093696.0,
      "step": 44375
    },
    {
      "entropy": 5.662577962875366,
      "epoch": 4.0257619738751815,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003464523566652131,
      "loss": 4.874,
      "mean_token_accuracy": 0.22283466905355453,
      "num_tokens": 92103217.0,
      "step": 44380
    },
    {
      "entropy": 5.607229948043823,
      "epoch": 4.026215529753266,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034642167439919634,
      "loss": 4.7877,
      "mean_token_accuracy": 0.22635677754878997,
      "num_tokens": 92112934.0,
      "step": 44385
    },
    {
      "entropy": 5.612427234649658,
      "epoch": 4.02666908563135,
      "grad_norm": 1.25,
      "learning_rate": 0.0003463909906561611,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22943898290395737,
      "num_tokens": 92123066.0,
      "step": 44390
    },
    {
      "entropy": 5.683171558380127,
      "epoch": 4.027122641509434,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003463603054367419,
      "loss": 4.8563,
      "mean_token_accuracy": 0.22709472328424454,
      "num_tokens": 92133911.0,
      "step": 44395
    },
    {
      "entropy": 5.608241939544678,
      "epoch": 4.027576197387518,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003463296187415735,
      "loss": 4.7845,
      "mean_token_accuracy": 0.23208570182323457,
      "num_tokens": 92143941.0,
      "step": 44400
    },
    {
      "entropy": 5.624104022979736,
      "epoch": 4.028029753265602,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003462989305712904,
      "loss": 4.8016,
      "mean_token_accuracy": 0.2274986431002617,
      "num_tokens": 92153934.0,
      "step": 44405
    },
    {
      "entropy": 5.612405300140381,
      "epoch": 4.028483309143686,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003462682409265272,
      "loss": 4.7815,
      "mean_token_accuracy": 0.2356450840830803,
      "num_tokens": 92163642.0,
      "step": 44410
    },
    {
      "entropy": 5.63348970413208,
      "epoch": 4.0289368650217705,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003462375498079186,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22458656132221222,
      "num_tokens": 92174034.0,
      "step": 44415
    },
    {
      "entropy": 5.683361673355103,
      "epoch": 4.029390420899855,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034620685721609937,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22842602729797362,
      "num_tokens": 92183714.0,
      "step": 44420
    },
    {
      "entropy": 5.680710887908935,
      "epoch": 4.029843976777939,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034617616315170413,
      "loss": 4.9158,
      "mean_token_accuracy": 0.22030563801527023,
      "num_tokens": 92195291.0,
      "step": 44425
    },
    {
      "entropy": 5.663345384597778,
      "epoch": 4.030297532656023,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003461454676153678,
      "loss": 4.8682,
      "mean_token_accuracy": 0.2296669900417328,
      "num_tokens": 92206159.0,
      "step": 44430
    },
    {
      "entropy": 5.681922769546508,
      "epoch": 4.030751088534108,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000346114770607725,
      "loss": 4.894,
      "mean_token_accuracy": 0.2254640281200409,
      "num_tokens": 92216928.0,
      "step": 44435
    },
    {
      "entropy": 5.590759754180908,
      "epoch": 4.031204644412192,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003460840721294106,
      "loss": 4.8076,
      "mean_token_accuracy": 0.22976668179035187,
      "num_tokens": 92227029.0,
      "step": 44440
    },
    {
      "entropy": 5.658042335510254,
      "epoch": 4.031658200290276,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034605337218105955,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22761137336492537,
      "num_tokens": 92237310.0,
      "step": 44445
    },
    {
      "entropy": 5.631627368927002,
      "epoch": 4.03211175616836,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003460226707633066,
      "loss": 4.8298,
      "mean_token_accuracy": 0.22943200170993805,
      "num_tokens": 92247841.0,
      "step": 44450
    },
    {
      "entropy": 5.670345163345337,
      "epoch": 4.032565312046444,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034599196787678667,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22401938289403917,
      "num_tokens": 92258513.0,
      "step": 44455
    },
    {
      "entropy": 5.633939743041992,
      "epoch": 4.033018867924528,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003459612635221349,
      "loss": 4.8636,
      "mean_token_accuracy": 0.2253674179315567,
      "num_tokens": 92268659.0,
      "step": 44460
    },
    {
      "entropy": 5.634753799438476,
      "epoch": 4.033472423802612,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034593055769998595,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22036066949367522,
      "num_tokens": 92279810.0,
      "step": 44465
    },
    {
      "entropy": 5.664070272445679,
      "epoch": 4.0339259796806965,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003458998504109751,
      "loss": 4.8533,
      "mean_token_accuracy": 0.22927334308624267,
      "num_tokens": 92289625.0,
      "step": 44470
    },
    {
      "entropy": 5.6108012199401855,
      "epoch": 4.034379535558781,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003458691416557372,
      "loss": 4.8368,
      "mean_token_accuracy": 0.22815120667219163,
      "num_tokens": 92300669.0,
      "step": 44475
    },
    {
      "entropy": 5.656001996994019,
      "epoch": 4.034833091436865,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034583843143490744,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23028798550367355,
      "num_tokens": 92312160.0,
      "step": 44480
    },
    {
      "entropy": 5.708610582351684,
      "epoch": 4.035286647314949,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003458077197491208,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2161591589450836,
      "num_tokens": 92322677.0,
      "step": 44485
    },
    {
      "entropy": 5.641248512268066,
      "epoch": 4.035740203193034,
      "grad_norm": 1.25,
      "learning_rate": 0.0003457770065990126,
      "loss": 4.782,
      "mean_token_accuracy": 0.23111440986394882,
      "num_tokens": 92333148.0,
      "step": 44490
    },
    {
      "entropy": 5.6804444789886475,
      "epoch": 4.036193759071118,
      "grad_norm": 1.375,
      "learning_rate": 0.0003457462919852178,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22654978483915328,
      "num_tokens": 92344189.0,
      "step": 44495
    },
    {
      "entropy": 5.6912617683410645,
      "epoch": 4.036647314949202,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003457155759083717,
      "loss": 4.8436,
      "mean_token_accuracy": 0.23271348476409912,
      "num_tokens": 92356136.0,
      "step": 44500
    },
    {
      "entropy": 5.616119813919068,
      "epoch": 4.0371008708272855,
      "grad_norm": 1.25,
      "learning_rate": 0.0003456848583691094,
      "loss": 4.7463,
      "mean_token_accuracy": 0.2305932894349098,
      "num_tokens": 92366393.0,
      "step": 44505
    },
    {
      "entropy": 5.638161134719849,
      "epoch": 4.03755442670537,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003456541393680662,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2310309961438179,
      "num_tokens": 92376819.0,
      "step": 44510
    },
    {
      "entropy": 5.642312240600586,
      "epoch": 4.038007982583454,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034562341890587747,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22103004455566405,
      "num_tokens": 92386704.0,
      "step": 44515
    },
    {
      "entropy": 5.65023398399353,
      "epoch": 4.038461538461538,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003455926969831784,
      "loss": 4.9122,
      "mean_token_accuracy": 0.21993288546800613,
      "num_tokens": 92397412.0,
      "step": 44520
    },
    {
      "entropy": 5.725669908523559,
      "epoch": 4.038915094339623,
      "grad_norm": 1.25,
      "learning_rate": 0.00034556197360060436,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22121664732694626,
      "num_tokens": 92408641.0,
      "step": 44525
    },
    {
      "entropy": 5.649687671661377,
      "epoch": 4.039368650217707,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003455312487587908,
      "loss": 4.8457,
      "mean_token_accuracy": 0.2286076918244362,
      "num_tokens": 92419341.0,
      "step": 44530
    },
    {
      "entropy": 5.7088817119598385,
      "epoch": 4.039822206095791,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003455005224583729,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22246959805488586,
      "num_tokens": 92429297.0,
      "step": 44535
    },
    {
      "entropy": 5.739550065994263,
      "epoch": 4.040275761973875,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034546979469998637,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22285029739141465,
      "num_tokens": 92439851.0,
      "step": 44540
    },
    {
      "entropy": 5.590083408355713,
      "epoch": 4.04072931785196,
      "grad_norm": 1.25,
      "learning_rate": 0.00034543906548426646,
      "loss": 4.7981,
      "mean_token_accuracy": 0.23707572221755982,
      "num_tokens": 92451096.0,
      "step": 44545
    },
    {
      "entropy": 5.644199371337891,
      "epoch": 4.041182873730044,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034540833481184867,
      "loss": 4.8384,
      "mean_token_accuracy": 0.22562169134616852,
      "num_tokens": 92460516.0,
      "step": 44550
    },
    {
      "entropy": 5.734266376495361,
      "epoch": 4.041636429608127,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034537760268336866,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22014155089855195,
      "num_tokens": 92471429.0,
      "step": 44555
    },
    {
      "entropy": 5.672460412979126,
      "epoch": 4.042089985486212,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003453468690994619,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22560233771800994,
      "num_tokens": 92482047.0,
      "step": 44560
    },
    {
      "entropy": 5.609123849868775,
      "epoch": 4.042543541364296,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034531613406076387,
      "loss": 4.8033,
      "mean_token_accuracy": 0.23258839398622513,
      "num_tokens": 92492206.0,
      "step": 44565
    },
    {
      "entropy": 5.61236572265625,
      "epoch": 4.04299709724238,
      "grad_norm": 1.25,
      "learning_rate": 0.0003452853975679103,
      "loss": 4.7637,
      "mean_token_accuracy": 0.2292274311184883,
      "num_tokens": 92501979.0,
      "step": 44570
    },
    {
      "entropy": 5.765503072738648,
      "epoch": 4.043450653120464,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034525465962153683,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22355792969465255,
      "num_tokens": 92512388.0,
      "step": 44575
    },
    {
      "entropy": 5.713842105865479,
      "epoch": 4.043904208998549,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034522392022227915,
      "loss": 4.9083,
      "mean_token_accuracy": 0.21787252724170686,
      "num_tokens": 92522805.0,
      "step": 44580
    },
    {
      "entropy": 5.6409131526947025,
      "epoch": 4.044357764876633,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034519317937077283,
      "loss": 4.852,
      "mean_token_accuracy": 0.21904472261667252,
      "num_tokens": 92532957.0,
      "step": 44585
    },
    {
      "entropy": 5.6419586658477785,
      "epoch": 4.044811320754717,
      "grad_norm": 1.375,
      "learning_rate": 0.0003451624370676537,
      "loss": 4.8694,
      "mean_token_accuracy": 0.2216425746679306,
      "num_tokens": 92543451.0,
      "step": 44590
    },
    {
      "entropy": 5.759320878982544,
      "epoch": 4.045264876632801,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034513169331355754,
      "loss": 4.8583,
      "mean_token_accuracy": 0.2199556663632393,
      "num_tokens": 92553856.0,
      "step": 44595
    },
    {
      "entropy": 5.624948883056641,
      "epoch": 4.045718432510886,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034510094810911995,
      "loss": 4.7908,
      "mean_token_accuracy": 0.23299646228551865,
      "num_tokens": 92563816.0,
      "step": 44600
    },
    {
      "entropy": 5.5846916198730465,
      "epoch": 4.04617198838897,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034507020145497705,
      "loss": 4.8645,
      "mean_token_accuracy": 0.23018748313188553,
      "num_tokens": 92574062.0,
      "step": 44605
    },
    {
      "entropy": 5.595385599136352,
      "epoch": 4.046625544267053,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034503945335176444,
      "loss": 4.7991,
      "mean_token_accuracy": 0.2357534497976303,
      "num_tokens": 92583935.0,
      "step": 44610
    },
    {
      "entropy": 5.708527374267578,
      "epoch": 4.047079100145138,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034500870380011807,
      "loss": 4.8707,
      "mean_token_accuracy": 0.23073073029518126,
      "num_tokens": 92594037.0,
      "step": 44615
    },
    {
      "entropy": 5.632271385192871,
      "epoch": 4.047532656023222,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034497795280067386,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22318391799926757,
      "num_tokens": 92603336.0,
      "step": 44620
    },
    {
      "entropy": 5.587202072143555,
      "epoch": 4.047986211901306,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034494720035406777,
      "loss": 4.7643,
      "mean_token_accuracy": 0.23421523869037628,
      "num_tokens": 92614210.0,
      "step": 44625
    },
    {
      "entropy": 5.698281288146973,
      "epoch": 4.04843976777939,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003449164464609358,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2246902123093605,
      "num_tokens": 92624036.0,
      "step": 44630
    },
    {
      "entropy": 5.680367660522461,
      "epoch": 4.048893323657475,
      "grad_norm": 1.25,
      "learning_rate": 0.0003448856911219139,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2162610486149788,
      "num_tokens": 92634588.0,
      "step": 44635
    },
    {
      "entropy": 5.655884265899658,
      "epoch": 4.049346879535559,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034485493433763804,
      "loss": 4.8166,
      "mean_token_accuracy": 0.22789756804704667,
      "num_tokens": 92645027.0,
      "step": 44640
    },
    {
      "entropy": 5.792963933944702,
      "epoch": 4.049800435413643,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034482417610874437,
      "loss": 4.8999,
      "mean_token_accuracy": 0.2181426152586937,
      "num_tokens": 92656669.0,
      "step": 44645
    },
    {
      "entropy": 5.679650592803955,
      "epoch": 4.0502539912917275,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003447934164358689,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2216082990169525,
      "num_tokens": 92666558.0,
      "step": 44650
    },
    {
      "entropy": 5.637512874603272,
      "epoch": 4.050707547169812,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034476265531964786,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22565652132034303,
      "num_tokens": 92677384.0,
      "step": 44655
    },
    {
      "entropy": 5.614072370529175,
      "epoch": 4.051161103047895,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003447318927607173,
      "loss": 4.7897,
      "mean_token_accuracy": 0.2319969803094864,
      "num_tokens": 92687619.0,
      "step": 44660
    },
    {
      "entropy": 5.599501895904541,
      "epoch": 4.051614658925979,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003447011287597134,
      "loss": 4.749,
      "mean_token_accuracy": 0.23312059342861174,
      "num_tokens": 92697784.0,
      "step": 44665
    },
    {
      "entropy": 5.669119167327881,
      "epoch": 4.052068214804064,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003446703633172724,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21804513037204742,
      "num_tokens": 92709255.0,
      "step": 44670
    },
    {
      "entropy": 5.627715635299682,
      "epoch": 4.052521770682148,
      "grad_norm": 1.375,
      "learning_rate": 0.0003446395964340305,
      "loss": 4.803,
      "mean_token_accuracy": 0.23277949839830397,
      "num_tokens": 92720714.0,
      "step": 44675
    },
    {
      "entropy": 5.649722528457642,
      "epoch": 4.052975326560232,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034460882811062417,
      "loss": 4.8121,
      "mean_token_accuracy": 0.2265733599662781,
      "num_tokens": 92730783.0,
      "step": 44680
    },
    {
      "entropy": 5.697773218154907,
      "epoch": 4.0534288824383164,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034457805834768937,
      "loss": 4.9143,
      "mean_token_accuracy": 0.2200439915060997,
      "num_tokens": 92741943.0,
      "step": 44685
    },
    {
      "entropy": 5.627641344070435,
      "epoch": 4.053882438316401,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003445472871458627,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2277112364768982,
      "num_tokens": 92753510.0,
      "step": 44690
    },
    {
      "entropy": 5.629557943344116,
      "epoch": 4.054335994194485,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003445165145057803,
      "loss": 4.8052,
      "mean_token_accuracy": 0.23348532915115355,
      "num_tokens": 92763655.0,
      "step": 44695
    },
    {
      "entropy": 5.691075134277344,
      "epoch": 4.054789550072569,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034448574042807866,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22064407616853715,
      "num_tokens": 92773732.0,
      "step": 44700
    },
    {
      "entropy": 5.714195108413696,
      "epoch": 4.0552431059506535,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003444549649133943,
      "loss": 4.8145,
      "mean_token_accuracy": 0.22589239180088044,
      "num_tokens": 92784822.0,
      "step": 44705
    },
    {
      "entropy": 5.735225486755371,
      "epoch": 4.055696661828737,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003444241879623636,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21155412644147872,
      "num_tokens": 92795270.0,
      "step": 44710
    },
    {
      "entropy": 5.657004308700562,
      "epoch": 4.056150217706821,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003443934095756229,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22736946195364,
      "num_tokens": 92806578.0,
      "step": 44715
    },
    {
      "entropy": 5.684420204162597,
      "epoch": 4.056603773584905,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003443626297538088,
      "loss": 4.8532,
      "mean_token_accuracy": 0.2264147639274597,
      "num_tokens": 92815932.0,
      "step": 44720
    },
    {
      "entropy": 5.637942171096801,
      "epoch": 4.05705732946299,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003443318484975579,
      "loss": 4.8421,
      "mean_token_accuracy": 0.2338358834385872,
      "num_tokens": 92825536.0,
      "step": 44725
    },
    {
      "entropy": 5.65257158279419,
      "epoch": 4.057510885341074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003443010658075066,
      "loss": 4.8116,
      "mean_token_accuracy": 0.2279555842280388,
      "num_tokens": 92836451.0,
      "step": 44730
    },
    {
      "entropy": 5.706453847885132,
      "epoch": 4.057964441219158,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003442702816842917,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2217036709189415,
      "num_tokens": 92846923.0,
      "step": 44735
    },
    {
      "entropy": 5.642614412307739,
      "epoch": 4.0584179970972425,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034423949612854974,
      "loss": 4.7935,
      "mean_token_accuracy": 0.22996752411127092,
      "num_tokens": 92857549.0,
      "step": 44740
    },
    {
      "entropy": 5.683959913253784,
      "epoch": 4.058871552975327,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003442087091409173,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22380694895982742,
      "num_tokens": 92867059.0,
      "step": 44745
    },
    {
      "entropy": 5.647807455062866,
      "epoch": 4.059325108853411,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003441779207220311,
      "loss": 4.8657,
      "mean_token_accuracy": 0.21780441403388978,
      "num_tokens": 92878138.0,
      "step": 44750
    },
    {
      "entropy": 5.673136806488037,
      "epoch": 4.059778664731495,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003441471308725279,
      "loss": 4.879,
      "mean_token_accuracy": 0.2276317670941353,
      "num_tokens": 92888931.0,
      "step": 44755
    },
    {
      "entropy": 5.6840739250183105,
      "epoch": 4.060232220609579,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034411633959304444,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21783434301614762,
      "num_tokens": 92899718.0,
      "step": 44760
    },
    {
      "entropy": 5.683983373641968,
      "epoch": 4.060685776487663,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034408554688421743,
      "loss": 4.835,
      "mean_token_accuracy": 0.23158248215913774,
      "num_tokens": 92910381.0,
      "step": 44765
    },
    {
      "entropy": 5.638430881500244,
      "epoch": 4.061139332365747,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003440547527466837,
      "loss": 4.8237,
      "mean_token_accuracy": 0.22981788069009781,
      "num_tokens": 92919905.0,
      "step": 44770
    },
    {
      "entropy": 5.616773843765259,
      "epoch": 4.0615928882438315,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00034402395718108016,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23332347571849824,
      "num_tokens": 92930210.0,
      "step": 44775
    },
    {
      "entropy": 5.663282632827759,
      "epoch": 4.062046444121916,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034399316018804356,
      "loss": 4.8339,
      "mean_token_accuracy": 0.229477259516716,
      "num_tokens": 92940832.0,
      "step": 44780
    },
    {
      "entropy": 5.728418207168579,
      "epoch": 4.0625,
      "grad_norm": 1.375,
      "learning_rate": 0.0003439623617682107,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2175066888332367,
      "num_tokens": 92951349.0,
      "step": 44785
    },
    {
      "entropy": 5.656303358078003,
      "epoch": 4.062953555878084,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003439315619222188,
      "loss": 4.8524,
      "mean_token_accuracy": 0.2270846888422966,
      "num_tokens": 92961642.0,
      "step": 44790
    },
    {
      "entropy": 5.642505407333374,
      "epoch": 4.0634071117561685,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003439007606507046,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2269844591617584,
      "num_tokens": 92971061.0,
      "step": 44795
    },
    {
      "entropy": 5.721338081359863,
      "epoch": 4.063860667634253,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003438699579543051,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21646183878183364,
      "num_tokens": 92981732.0,
      "step": 44800
    },
    {
      "entropy": 5.720691967010498,
      "epoch": 4.064314223512337,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034383915383365726,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21890389025211335,
      "num_tokens": 92991570.0,
      "step": 44805
    },
    {
      "entropy": 5.688021612167359,
      "epoch": 4.064767779390421,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003438083482893982,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2271180510520935,
      "num_tokens": 93001623.0,
      "step": 44810
    },
    {
      "entropy": 5.700196599960327,
      "epoch": 4.065221335268505,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034377754132216496,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22592390924692154,
      "num_tokens": 93012167.0,
      "step": 44815
    },
    {
      "entropy": 5.633208990097046,
      "epoch": 4.065674891146589,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003437467329325947,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2338806062936783,
      "num_tokens": 93022090.0,
      "step": 44820
    },
    {
      "entropy": 5.645627832412719,
      "epoch": 4.066128447024673,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003437159231213244,
      "loss": 4.7339,
      "mean_token_accuracy": 0.23746891617774962,
      "num_tokens": 93030917.0,
      "step": 44825
    },
    {
      "entropy": 5.655427265167236,
      "epoch": 4.0665820029027575,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003436851118889914,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22182028144598007,
      "num_tokens": 93040113.0,
      "step": 44830
    },
    {
      "entropy": 5.664164066314697,
      "epoch": 4.067035558780842,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003436542992362326,
      "loss": 4.8728,
      "mean_token_accuracy": 0.230484339594841,
      "num_tokens": 93050053.0,
      "step": 44835
    },
    {
      "entropy": 5.749315977096558,
      "epoch": 4.067489114658926,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003436234851636856,
      "loss": 4.9139,
      "mean_token_accuracy": 0.21561993509531022,
      "num_tokens": 93060581.0,
      "step": 44840
    },
    {
      "entropy": 5.711426877975464,
      "epoch": 4.06794267053701,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00034359266967198734,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22212810218334197,
      "num_tokens": 93070158.0,
      "step": 44845
    },
    {
      "entropy": 5.590299463272094,
      "epoch": 4.068396226415095,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003435618527617752,
      "loss": 4.8155,
      "mean_token_accuracy": 0.22881777286529542,
      "num_tokens": 93080804.0,
      "step": 44850
    },
    {
      "entropy": 5.698328971862793,
      "epoch": 4.068849782293179,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003435310344336865,
      "loss": 4.8464,
      "mean_token_accuracy": 0.22443929463624954,
      "num_tokens": 93091346.0,
      "step": 44855
    },
    {
      "entropy": 5.620715570449829,
      "epoch": 4.069303338171263,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003435002146883585,
      "loss": 4.8175,
      "mean_token_accuracy": 0.23096765279769899,
      "num_tokens": 93100931.0,
      "step": 44860
    },
    {
      "entropy": 5.64532585144043,
      "epoch": 4.0697568940493465,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003434693935264286,
      "loss": 4.833,
      "mean_token_accuracy": 0.2314332216978073,
      "num_tokens": 93110181.0,
      "step": 44865
    },
    {
      "entropy": 5.699645280838013,
      "epoch": 4.070210449927431,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003434385709485343,
      "loss": 4.836,
      "mean_token_accuracy": 0.22298609912395478,
      "num_tokens": 93120140.0,
      "step": 44870
    },
    {
      "entropy": 5.721970176696777,
      "epoch": 4.070664005805515,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003434077469553128,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21836236864328384,
      "num_tokens": 93130853.0,
      "step": 44875
    },
    {
      "entropy": 5.672637224197388,
      "epoch": 4.071117561683599,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003433769215474017,
      "loss": 4.8253,
      "mean_token_accuracy": 0.22929693311452864,
      "num_tokens": 93141430.0,
      "step": 44880
    },
    {
      "entropy": 5.7178362846374515,
      "epoch": 4.0715711175616836,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034334609472543845,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21823622286319733,
      "num_tokens": 93152270.0,
      "step": 44885
    },
    {
      "entropy": 5.706017780303955,
      "epoch": 4.072024673439768,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003433152664900605,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22154334485530852,
      "num_tokens": 93162850.0,
      "step": 44890
    },
    {
      "entropy": 5.653424978256226,
      "epoch": 4.072478229317852,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003432844368419055,
      "loss": 4.7452,
      "mean_token_accuracy": 0.23382721096277237,
      "num_tokens": 93172265.0,
      "step": 44895
    },
    {
      "entropy": 5.63842363357544,
      "epoch": 4.072931785195936,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003432536057816108,
      "loss": 4.8034,
      "mean_token_accuracy": 0.23524775505065917,
      "num_tokens": 93182756.0,
      "step": 44900
    },
    {
      "entropy": 5.645920705795288,
      "epoch": 4.073385341074021,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034322277330981426,
      "loss": 4.8735,
      "mean_token_accuracy": 0.21888897716999053,
      "num_tokens": 93192710.0,
      "step": 44905
    },
    {
      "entropy": 5.644369411468506,
      "epoch": 4.073838896952105,
      "grad_norm": 1.375,
      "learning_rate": 0.00034319193942715343,
      "loss": 4.8283,
      "mean_token_accuracy": 0.22990211397409438,
      "num_tokens": 93202531.0,
      "step": 44910
    },
    {
      "entropy": 5.7184381008148195,
      "epoch": 4.074292452830188,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034316110413426586,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22301193028688432,
      "num_tokens": 93212208.0,
      "step": 44915
    },
    {
      "entropy": 5.77139983177185,
      "epoch": 4.0747460087082725,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003431302674317892,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21703479886054994,
      "num_tokens": 93222932.0,
      "step": 44920
    },
    {
      "entropy": 5.664581775665283,
      "epoch": 4.075199564586357,
      "grad_norm": 1.375,
      "learning_rate": 0.00034309942932036136,
      "loss": 4.7562,
      "mean_token_accuracy": 0.2378673493862152,
      "num_tokens": 93232049.0,
      "step": 44925
    },
    {
      "entropy": 5.649416732788086,
      "epoch": 4.075653120464441,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034306858980061994,
      "loss": 4.9001,
      "mean_token_accuracy": 0.22571207582950592,
      "num_tokens": 93244230.0,
      "step": 44930
    },
    {
      "entropy": 5.655362701416015,
      "epoch": 4.076106676342525,
      "grad_norm": 1.375,
      "learning_rate": 0.00034303774887320274,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22493933886289597,
      "num_tokens": 93254159.0,
      "step": 44935
    },
    {
      "entropy": 5.669178581237793,
      "epoch": 4.07656023222061,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034300690653874753,
      "loss": 4.7838,
      "mean_token_accuracy": 0.2320196285843849,
      "num_tokens": 93263742.0,
      "step": 44940
    },
    {
      "entropy": 5.664962720870972,
      "epoch": 4.077013788098694,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034297606279789215,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2174091726541519,
      "num_tokens": 93274446.0,
      "step": 44945
    },
    {
      "entropy": 5.674808597564697,
      "epoch": 4.077467343976778,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003429452176512745,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22100184857845306,
      "num_tokens": 93284102.0,
      "step": 44950
    },
    {
      "entropy": 5.673649883270263,
      "epoch": 4.077920899854862,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034291437109953244,
      "loss": 4.7899,
      "mean_token_accuracy": 0.23807226568460466,
      "num_tokens": 93295096.0,
      "step": 44955
    },
    {
      "entropy": 5.630706644058227,
      "epoch": 4.078374455732947,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003428835231433039,
      "loss": 4.8717,
      "mean_token_accuracy": 0.2250870794057846,
      "num_tokens": 93305377.0,
      "step": 44960
    },
    {
      "entropy": 5.732788848876953,
      "epoch": 4.07882801161103,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034285267378322674,
      "loss": 4.913,
      "mean_token_accuracy": 0.22364114671945573,
      "num_tokens": 93315859.0,
      "step": 44965
    },
    {
      "entropy": 5.652154445648193,
      "epoch": 4.079281567489114,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034282182301993903,
      "loss": 4.818,
      "mean_token_accuracy": 0.22866998463869095,
      "num_tokens": 93326519.0,
      "step": 44970
    },
    {
      "entropy": 5.677565002441407,
      "epoch": 4.079735123367199,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003427909708540787,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22327892780303954,
      "num_tokens": 93337055.0,
      "step": 44975
    },
    {
      "entropy": 5.691798877716065,
      "epoch": 4.080188679245283,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034276011728628393,
      "loss": 4.9124,
      "mean_token_accuracy": 0.21740656942129136,
      "num_tokens": 93346565.0,
      "step": 44980
    },
    {
      "entropy": 5.654440116882324,
      "epoch": 4.080642235123367,
      "grad_norm": 1.375,
      "learning_rate": 0.00034272926231719255,
      "loss": 4.8526,
      "mean_token_accuracy": 0.2255293011665344,
      "num_tokens": 93356366.0,
      "step": 44985
    },
    {
      "entropy": 5.677212429046631,
      "epoch": 4.081095791001451,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003426984059474428,
      "loss": 4.868,
      "mean_token_accuracy": 0.21957886815071107,
      "num_tokens": 93367317.0,
      "step": 44990
    },
    {
      "entropy": 5.6544046878814695,
      "epoch": 4.081549346879536,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003426675481776728,
      "loss": 4.8752,
      "mean_token_accuracy": 0.22476110458374024,
      "num_tokens": 93379601.0,
      "step": 44995
    },
    {
      "entropy": 5.6626802444458,
      "epoch": 4.08200290275762,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034263668900852065,
      "loss": 4.8233,
      "mean_token_accuracy": 0.2374037727713585,
      "num_tokens": 93389740.0,
      "step": 45000
    },
    {
      "epoch": 4.08200290275762,
      "eval_entropy": 5.415978300841146,
      "eval_loss": 4.963794708251953,
      "eval_mean_token_accuracy": 0.22841118181722736,
      "eval_num_tokens": 93389740.0,
      "eval_runtime": 20.7334,
      "eval_samples_per_second": 1705.463,
      "eval_steps_per_second": 213.183,
      "step": 45000
    },
    {
      "entropy": 5.76001386642456,
      "epoch": 4.082456458635704,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003426058284406244,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21899300515651704,
      "num_tokens": 93400232.0,
      "step": 45005
    },
    {
      "entropy": 5.722826671600342,
      "epoch": 4.082910014513788,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003425749664746226,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22981218695640565,
      "num_tokens": 93410662.0,
      "step": 45010
    },
    {
      "entropy": 5.659625625610351,
      "epoch": 4.083363570391873,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003425441031111532,
      "loss": 4.7516,
      "mean_token_accuracy": 0.23781661242246627,
      "num_tokens": 93420328.0,
      "step": 45015
    },
    {
      "entropy": 5.688152742385864,
      "epoch": 4.083817126269956,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034251323835085453,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21787895560264586,
      "num_tokens": 93430878.0,
      "step": 45020
    },
    {
      "entropy": 5.739926385879516,
      "epoch": 4.08427068214804,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003424823721943649,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2200404554605484,
      "num_tokens": 93441784.0,
      "step": 45025
    },
    {
      "entropy": 5.645525598526001,
      "epoch": 4.084724238026125,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003424515046423226,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22602445632219315,
      "num_tokens": 93453058.0,
      "step": 45030
    },
    {
      "entropy": 5.701102352142334,
      "epoch": 4.085177793904209,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003424206356953659,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22581003457307816,
      "num_tokens": 93463089.0,
      "step": 45035
    },
    {
      "entropy": 5.6917809009552,
      "epoch": 4.085631349782293,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034238976535413333,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2289608433842659,
      "num_tokens": 93473315.0,
      "step": 45040
    },
    {
      "entropy": 5.7374776840209964,
      "epoch": 4.086084905660377,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003423588936192633,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21516617387533188,
      "num_tokens": 93485113.0,
      "step": 45045
    },
    {
      "entropy": 5.622793197631836,
      "epoch": 4.086538461538462,
      "grad_norm": 1.375,
      "learning_rate": 0.0003423280204913941,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23042899817228318,
      "num_tokens": 93494269.0,
      "step": 45050
    },
    {
      "entropy": 5.606999731063842,
      "epoch": 4.086992017416546,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003422971459711644,
      "loss": 4.7908,
      "mean_token_accuracy": 0.2321794241666794,
      "num_tokens": 93505052.0,
      "step": 45055
    },
    {
      "entropy": 5.6597411155700685,
      "epoch": 4.08744557329463,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034226627005921244,
      "loss": 4.8458,
      "mean_token_accuracy": 0.22276150435209274,
      "num_tokens": 93514986.0,
      "step": 45060
    },
    {
      "entropy": 5.695908737182617,
      "epoch": 4.0878991291727145,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000342235392756177,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22508749216794968,
      "num_tokens": 93525286.0,
      "step": 45065
    },
    {
      "entropy": 5.703288841247558,
      "epoch": 4.088352685050798,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034220451406269636,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21571753174066544,
      "num_tokens": 93536034.0,
      "step": 45070
    },
    {
      "entropy": 5.598156547546386,
      "epoch": 4.088806240928882,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034217363397940937,
      "loss": 4.8378,
      "mean_token_accuracy": 0.2273579865694046,
      "num_tokens": 93546283.0,
      "step": 45075
    },
    {
      "entropy": 5.59804368019104,
      "epoch": 4.089259796806966,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034214275250695446,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22409984171390535,
      "num_tokens": 93555811.0,
      "step": 45080
    },
    {
      "entropy": 5.71259822845459,
      "epoch": 4.089713352685051,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003421118696459703,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21810808777809143,
      "num_tokens": 93567154.0,
      "step": 45085
    },
    {
      "entropy": 5.6417242050170895,
      "epoch": 4.090166908563135,
      "grad_norm": 1.25,
      "learning_rate": 0.00034208098539709566,
      "loss": 4.8301,
      "mean_token_accuracy": 0.2349393218755722,
      "num_tokens": 93577654.0,
      "step": 45090
    },
    {
      "entropy": 5.649134731292724,
      "epoch": 4.090620464441219,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034205009976096906,
      "loss": 4.8677,
      "mean_token_accuracy": 0.22095982879400253,
      "num_tokens": 93588712.0,
      "step": 45095
    },
    {
      "entropy": 5.67066798210144,
      "epoch": 4.0910740203193035,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003420192127382293,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2203037515282631,
      "num_tokens": 93599186.0,
      "step": 45100
    },
    {
      "entropy": 5.653915023803711,
      "epoch": 4.091527576197388,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034198832432951517,
      "loss": 4.7702,
      "mean_token_accuracy": 0.23333630859851837,
      "num_tokens": 93609439.0,
      "step": 45105
    },
    {
      "entropy": 5.677346897125244,
      "epoch": 4.091981132075472,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003419574345354654,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2196532353758812,
      "num_tokens": 93619903.0,
      "step": 45110
    },
    {
      "entropy": 5.653753089904785,
      "epoch": 4.092434687953556,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034192654335671893,
      "loss": 4.8123,
      "mean_token_accuracy": 0.23201809376478194,
      "num_tokens": 93629895.0,
      "step": 45115
    },
    {
      "entropy": 5.660627555847168,
      "epoch": 4.09288824383164,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034189565079391436,
      "loss": 4.9367,
      "mean_token_accuracy": 0.21982201486825942,
      "num_tokens": 93640677.0,
      "step": 45120
    },
    {
      "entropy": 5.692655467987061,
      "epoch": 4.093341799709724,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003418647568476907,
      "loss": 4.9539,
      "mean_token_accuracy": 0.2249619647860527,
      "num_tokens": 93651541.0,
      "step": 45125
    },
    {
      "entropy": 5.728185319900513,
      "epoch": 4.093795355587808,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003418338615186869,
      "loss": 4.9076,
      "mean_token_accuracy": 0.21813030689954757,
      "num_tokens": 93661901.0,
      "step": 45130
    },
    {
      "entropy": 5.682808351516724,
      "epoch": 4.094248911465892,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034180296480754174,
      "loss": 4.8495,
      "mean_token_accuracy": 0.2221786230802536,
      "num_tokens": 93672307.0,
      "step": 45135
    },
    {
      "entropy": 5.664837074279785,
      "epoch": 4.094702467343977,
      "grad_norm": 1.375,
      "learning_rate": 0.0003417720667148943,
      "loss": 4.8216,
      "mean_token_accuracy": 0.215610633790493,
      "num_tokens": 93683802.0,
      "step": 45140
    },
    {
      "entropy": 5.738400554656982,
      "epoch": 4.095156023222061,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003417411672413835,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21194143146276473,
      "num_tokens": 93693467.0,
      "step": 45145
    },
    {
      "entropy": 5.668906784057617,
      "epoch": 4.095609579100145,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034171026638764833,
      "loss": 4.8425,
      "mean_token_accuracy": 0.2330695942044258,
      "num_tokens": 93703858.0,
      "step": 45150
    },
    {
      "entropy": 5.656031799316406,
      "epoch": 4.0960631349782295,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034167936415432785,
      "loss": 4.853,
      "mean_token_accuracy": 0.23564156293869018,
      "num_tokens": 93714457.0,
      "step": 45155
    },
    {
      "entropy": 5.665589141845703,
      "epoch": 4.096516690856314,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003416484605420611,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22803847938776017,
      "num_tokens": 93725009.0,
      "step": 45160
    },
    {
      "entropy": 5.7171369075775145,
      "epoch": 4.096970246734398,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034161755555148725,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2169937163591385,
      "num_tokens": 93736357.0,
      "step": 45165
    },
    {
      "entropy": 5.641406297683716,
      "epoch": 4.097423802612482,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034158664918324536,
      "loss": 4.8185,
      "mean_token_accuracy": 0.22552237808704376,
      "num_tokens": 93746711.0,
      "step": 45170
    },
    {
      "entropy": 5.7034728050231935,
      "epoch": 4.097877358490566,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034155574143797455,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2179391413927078,
      "num_tokens": 93757732.0,
      "step": 45175
    },
    {
      "entropy": 5.66833610534668,
      "epoch": 4.09833091436865,
      "grad_norm": 1.375,
      "learning_rate": 0.00034152483231631413,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2288547307252884,
      "num_tokens": 93767048.0,
      "step": 45180
    },
    {
      "entropy": 5.824919176101685,
      "epoch": 4.098784470246734,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003414939218189032,
      "loss": 5.0682,
      "mean_token_accuracy": 0.21120611429214478,
      "num_tokens": 93778283.0,
      "step": 45185
    },
    {
      "entropy": 5.7421308040618895,
      "epoch": 4.0992380261248185,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000341463009946381,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21656214892864228,
      "num_tokens": 93788776.0,
      "step": 45190
    },
    {
      "entropy": 5.700044250488281,
      "epoch": 4.099691582002903,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034143209669938685,
      "loss": 4.883,
      "mean_token_accuracy": 0.21803147941827775,
      "num_tokens": 93798695.0,
      "step": 45195
    },
    {
      "entropy": 5.7308789730072025,
      "epoch": 4.100145137880987,
      "grad_norm": 1.25,
      "learning_rate": 0.00034140118207856,
      "loss": 4.9693,
      "mean_token_accuracy": 0.22296274751424788,
      "num_tokens": 93808993.0,
      "step": 45200
    },
    {
      "entropy": 5.725953722000122,
      "epoch": 4.100598693759071,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034137026608453973,
      "loss": 4.9766,
      "mean_token_accuracy": 0.21785318851470947,
      "num_tokens": 93818361.0,
      "step": 45205
    },
    {
      "entropy": 5.70720272064209,
      "epoch": 4.1010522496371555,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003413393487179655,
      "loss": 4.911,
      "mean_token_accuracy": 0.23547320812940598,
      "num_tokens": 93828988.0,
      "step": 45210
    },
    {
      "entropy": 5.791929578781128,
      "epoch": 4.10150580551524,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003413084299794767,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2215809315443039,
      "num_tokens": 93839350.0,
      "step": 45215
    },
    {
      "entropy": 5.703439235687256,
      "epoch": 4.101959361393324,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034127750986971255,
      "loss": 4.8394,
      "mean_token_accuracy": 0.22897980511188507,
      "num_tokens": 93849939.0,
      "step": 45220
    },
    {
      "entropy": 5.670909023284912,
      "epoch": 4.1024129172714074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003412465883893126,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22460082769393921,
      "num_tokens": 93861278.0,
      "step": 45225
    },
    {
      "entropy": 5.717130088806153,
      "epoch": 4.102866473149492,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003412156655389164,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22266440689563752,
      "num_tokens": 93871693.0,
      "step": 45230
    },
    {
      "entropy": 5.671990060806275,
      "epoch": 4.103320029027576,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003411847413191633,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2262684002518654,
      "num_tokens": 93883098.0,
      "step": 45235
    },
    {
      "entropy": 5.643094491958618,
      "epoch": 4.10377358490566,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034115381573069297,
      "loss": 4.7936,
      "mean_token_accuracy": 0.22806403338909148,
      "num_tokens": 93893347.0,
      "step": 45240
    },
    {
      "entropy": 5.738433074951172,
      "epoch": 4.1042271407837445,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034112288877414485,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21406890004873275,
      "num_tokens": 93903964.0,
      "step": 45245
    },
    {
      "entropy": 5.693863868713379,
      "epoch": 4.104680696661829,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003410919604501585,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2248979777097702,
      "num_tokens": 93914526.0,
      "step": 45250
    },
    {
      "entropy": 5.755918216705322,
      "epoch": 4.105134252539913,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003410610307593735,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2249869227409363,
      "num_tokens": 93924430.0,
      "step": 45255
    },
    {
      "entropy": 5.716763544082641,
      "epoch": 4.105587808417997,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003410300997024297,
      "loss": 4.8459,
      "mean_token_accuracy": 0.22883870154619218,
      "num_tokens": 93935150.0,
      "step": 45260
    },
    {
      "entropy": 5.633342456817627,
      "epoch": 4.106041364296082,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003409991672799665,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23588439375162124,
      "num_tokens": 93944496.0,
      "step": 45265
    },
    {
      "entropy": 5.684167146682739,
      "epoch": 4.106494920174166,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034096823349262376,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2192351371049881,
      "num_tokens": 93955745.0,
      "step": 45270
    },
    {
      "entropy": 5.613743114471435,
      "epoch": 4.106948476052249,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003409372983410411,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2272347167134285,
      "num_tokens": 93965930.0,
      "step": 45275
    },
    {
      "entropy": 5.643866682052613,
      "epoch": 4.1074020319303335,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003409063618258583,
      "loss": 4.7847,
      "mean_token_accuracy": 0.2328326880931854,
      "num_tokens": 93976435.0,
      "step": 45280
    },
    {
      "entropy": 5.545793581008911,
      "epoch": 4.107855587808418,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034087542394771524,
      "loss": 4.7311,
      "mean_token_accuracy": 0.23665671050548553,
      "num_tokens": 93987294.0,
      "step": 45285
    },
    {
      "entropy": 5.694854021072388,
      "epoch": 4.108309143686502,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034084448470725147,
      "loss": 4.8691,
      "mean_token_accuracy": 0.2259235665202141,
      "num_tokens": 93997177.0,
      "step": 45290
    },
    {
      "entropy": 5.752830266952515,
      "epoch": 4.108762699564586,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003408135441051071,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21901662945747374,
      "num_tokens": 94007663.0,
      "step": 45295
    },
    {
      "entropy": 5.677536821365356,
      "epoch": 4.1092162554426706,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003407826021419218,
      "loss": 4.7978,
      "mean_token_accuracy": 0.22582892924547196,
      "num_tokens": 94016899.0,
      "step": 45300
    },
    {
      "entropy": 5.67814884185791,
      "epoch": 4.109669811320755,
      "grad_norm": 1.375,
      "learning_rate": 0.0003407516588183356,
      "loss": 4.9185,
      "mean_token_accuracy": 0.21812294274568558,
      "num_tokens": 94027696.0,
      "step": 45305
    },
    {
      "entropy": 5.646271181106568,
      "epoch": 4.110123367198839,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003407207141349882,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22462472915649415,
      "num_tokens": 94038075.0,
      "step": 45310
    },
    {
      "entropy": 5.685809135437012,
      "epoch": 4.110576923076923,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003406897680925199,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21939867585897446,
      "num_tokens": 94049082.0,
      "step": 45315
    },
    {
      "entropy": 5.69429235458374,
      "epoch": 4.111030478955008,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034065882069157023,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22121205031871796,
      "num_tokens": 94060367.0,
      "step": 45320
    },
    {
      "entropy": 5.667037630081177,
      "epoch": 4.111484034833092,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034062787193277954,
      "loss": 4.8327,
      "mean_token_accuracy": 0.2274642691016197,
      "num_tokens": 94069415.0,
      "step": 45325
    },
    {
      "entropy": 5.722782325744629,
      "epoch": 4.111937590711175,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034059692181678775,
      "loss": 4.91,
      "mean_token_accuracy": 0.21705766022205353,
      "num_tokens": 94079047.0,
      "step": 45330
    },
    {
      "entropy": 5.658153343200683,
      "epoch": 4.1123911465892595,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034056597034423484,
      "loss": 4.8259,
      "mean_token_accuracy": 0.2258522093296051,
      "num_tokens": 94089631.0,
      "step": 45335
    },
    {
      "entropy": 5.637837314605713,
      "epoch": 4.112844702467344,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034053501751576097,
      "loss": 4.8327,
      "mean_token_accuracy": 0.22640466541051865,
      "num_tokens": 94099265.0,
      "step": 45340
    },
    {
      "entropy": 5.603198003768921,
      "epoch": 4.113298258345428,
      "grad_norm": 1.25,
      "learning_rate": 0.0003405040633320063,
      "loss": 4.8645,
      "mean_token_accuracy": 0.21789870262145997,
      "num_tokens": 94109673.0,
      "step": 45345
    },
    {
      "entropy": 5.643795442581177,
      "epoch": 4.113751814223512,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003404731077936108,
      "loss": 4.7719,
      "mean_token_accuracy": 0.2338087722659111,
      "num_tokens": 94120004.0,
      "step": 45350
    },
    {
      "entropy": 5.639846086502075,
      "epoch": 4.114205370101597,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003404421509012148,
      "loss": 4.8578,
      "mean_token_accuracy": 0.2245064362883568,
      "num_tokens": 94129935.0,
      "step": 45355
    },
    {
      "entropy": 5.678585243225098,
      "epoch": 4.114658925979681,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034041119265545847,
      "loss": 4.902,
      "mean_token_accuracy": 0.22027744352817535,
      "num_tokens": 94139035.0,
      "step": 45360
    },
    {
      "entropy": 5.70267424583435,
      "epoch": 4.115112481857765,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034038023305698196,
      "loss": 4.8405,
      "mean_token_accuracy": 0.22503584176301955,
      "num_tokens": 94149281.0,
      "step": 45365
    },
    {
      "entropy": 5.658447790145874,
      "epoch": 4.115566037735849,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003403492721064256,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22493139803409576,
      "num_tokens": 94160455.0,
      "step": 45370
    },
    {
      "entropy": 5.691318845748901,
      "epoch": 4.116019593613934,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034031830980442955,
      "loss": 4.898,
      "mean_token_accuracy": 0.21858693212270736,
      "num_tokens": 94170334.0,
      "step": 45375
    },
    {
      "entropy": 5.650800514221191,
      "epoch": 4.116473149492017,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034028734615163436,
      "loss": 4.8206,
      "mean_token_accuracy": 0.22225605845451354,
      "num_tokens": 94181055.0,
      "step": 45380
    },
    {
      "entropy": 5.6686070442199705,
      "epoch": 4.116926705370101,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003402563811486802,
      "loss": 4.8235,
      "mean_token_accuracy": 0.2321927949786186,
      "num_tokens": 94191182.0,
      "step": 45385
    },
    {
      "entropy": 5.696384477615356,
      "epoch": 4.117380261248186,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003402254147962073,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22072302550077438,
      "num_tokens": 94201229.0,
      "step": 45390
    },
    {
      "entropy": 5.623414182662964,
      "epoch": 4.11783381712627,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003401944470948563,
      "loss": 4.8389,
      "mean_token_accuracy": 0.2286711409687996,
      "num_tokens": 94210956.0,
      "step": 45395
    },
    {
      "entropy": 5.66583833694458,
      "epoch": 4.118287373004354,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034016347804526753,
      "loss": 4.9,
      "mean_token_accuracy": 0.22061336040496826,
      "num_tokens": 94220740.0,
      "step": 45400
    },
    {
      "entropy": 5.616199493408203,
      "epoch": 4.118740928882438,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034013250764808133,
      "loss": 4.7553,
      "mean_token_accuracy": 0.24380480200052262,
      "num_tokens": 94231497.0,
      "step": 45405
    },
    {
      "entropy": 5.67497501373291,
      "epoch": 4.119194484760523,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003401015359039384,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22862010300159455,
      "num_tokens": 94240492.0,
      "step": 45410
    },
    {
      "entropy": 5.625301551818848,
      "epoch": 4.119648040638607,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003400705628134791,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23075455278158188,
      "num_tokens": 94251586.0,
      "step": 45415
    },
    {
      "entropy": 5.657022476196289,
      "epoch": 4.120101596516691,
      "grad_norm": 1.34375,
      "learning_rate": 0.000340039588377344,
      "loss": 4.8579,
      "mean_token_accuracy": 0.23169631212949754,
      "num_tokens": 94261635.0,
      "step": 45420
    },
    {
      "entropy": 5.682074499130249,
      "epoch": 4.120555152394775,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003400086125961735,
      "loss": 4.8919,
      "mean_token_accuracy": 0.21863821297883987,
      "num_tokens": 94269881.0,
      "step": 45425
    },
    {
      "entropy": 5.6766022682189945,
      "epoch": 4.121008708272859,
      "grad_norm": 1.46875,
      "learning_rate": 0.00033997763547060847,
      "loss": 4.8213,
      "mean_token_accuracy": 0.22783233374357223,
      "num_tokens": 94279412.0,
      "step": 45430
    },
    {
      "entropy": 5.671648263931274,
      "epoch": 4.121462264150943,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003399466570012893,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22102775871753694,
      "num_tokens": 94290157.0,
      "step": 45435
    },
    {
      "entropy": 5.70859112739563,
      "epoch": 4.121915820029027,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033991567718885674,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22150469273328782,
      "num_tokens": 94299693.0,
      "step": 45440
    },
    {
      "entropy": 5.781943368911743,
      "epoch": 4.122369375907112,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003398846960339515,
      "loss": 4.9868,
      "mean_token_accuracy": 0.2201607570052147,
      "num_tokens": 94310245.0,
      "step": 45445
    },
    {
      "entropy": 5.668787813186645,
      "epoch": 4.122822931785196,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003398537135372141,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2243713304400444,
      "num_tokens": 94321190.0,
      "step": 45450
    },
    {
      "entropy": 5.724693059921265,
      "epoch": 4.12327648766328,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003398227296992854,
      "loss": 4.9206,
      "mean_token_accuracy": 0.22508213967084884,
      "num_tokens": 94330273.0,
      "step": 45455
    },
    {
      "entropy": 5.747559309005737,
      "epoch": 4.123730043541364,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033979174452080616,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21284834146499634,
      "num_tokens": 94341097.0,
      "step": 45460
    },
    {
      "entropy": 5.687551498413086,
      "epoch": 4.124183599419449,
      "grad_norm": 1.25,
      "learning_rate": 0.0003397607580024171,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22462472319602966,
      "num_tokens": 94352329.0,
      "step": 45465
    },
    {
      "entropy": 5.658494567871093,
      "epoch": 4.124637155297533,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003397297701447591,
      "loss": 4.817,
      "mean_token_accuracy": 0.22459701001644133,
      "num_tokens": 94362551.0,
      "step": 45470
    },
    {
      "entropy": 5.657405662536621,
      "epoch": 4.125090711175617,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000339698780948473,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22020042985677718,
      "num_tokens": 94372405.0,
      "step": 45475
    },
    {
      "entropy": 5.713830089569091,
      "epoch": 4.125544267053701,
      "grad_norm": 1.25,
      "learning_rate": 0.0003396677904141995,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22904844731092452,
      "num_tokens": 94382347.0,
      "step": 45480
    },
    {
      "entropy": 5.610005855560303,
      "epoch": 4.125997822931785,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033963679854257965,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22397020161151887,
      "num_tokens": 94393428.0,
      "step": 45485
    },
    {
      "entropy": 5.680051374435425,
      "epoch": 4.126451378809869,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003396058053342544,
      "loss": 4.887,
      "mean_token_accuracy": 0.23005108237266542,
      "num_tokens": 94403941.0,
      "step": 45490
    },
    {
      "entropy": 5.692630815505981,
      "epoch": 4.126904934687953,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033957481078986465,
      "loss": 4.8692,
      "mean_token_accuracy": 0.2214859664440155,
      "num_tokens": 94414206.0,
      "step": 45495
    },
    {
      "entropy": 5.717976760864258,
      "epoch": 4.127358490566038,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033954381491005126,
      "loss": 4.8471,
      "mean_token_accuracy": 0.23197571486234664,
      "num_tokens": 94424481.0,
      "step": 45500
    },
    {
      "entropy": 5.6614902973175045,
      "epoch": 4.127812046444122,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003395128176954554,
      "loss": 4.8317,
      "mean_token_accuracy": 0.22894318401813507,
      "num_tokens": 94434714.0,
      "step": 45505
    },
    {
      "entropy": 5.7151110649108885,
      "epoch": 4.128265602322206,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000339481819146718,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21963382065296172,
      "num_tokens": 94445267.0,
      "step": 45510
    },
    {
      "entropy": 5.707913446426391,
      "epoch": 4.1287191582002905,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033945081926448013,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22955146729946135,
      "num_tokens": 94455724.0,
      "step": 45515
    },
    {
      "entropy": 5.705229091644287,
      "epoch": 4.129172714078375,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033941981804938294,
      "loss": 4.9194,
      "mean_token_accuracy": 0.2216176763176918,
      "num_tokens": 94466655.0,
      "step": 45520
    },
    {
      "entropy": 5.615235090255737,
      "epoch": 4.129626269956459,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033938881550206744,
      "loss": 4.815,
      "mean_token_accuracy": 0.22803997248411179,
      "num_tokens": 94477699.0,
      "step": 45525
    },
    {
      "entropy": 5.661374616622925,
      "epoch": 4.130079825834542,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033935781162317485,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22512202560901642,
      "num_tokens": 94487639.0,
      "step": 45530
    },
    {
      "entropy": 5.801564645767212,
      "epoch": 4.130533381712627,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033932680641334635,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21225543767213823,
      "num_tokens": 94498687.0,
      "step": 45535
    },
    {
      "entropy": 5.704364824295044,
      "epoch": 4.130986937590711,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003392957998732231,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23066934794187546,
      "num_tokens": 94508706.0,
      "step": 45540
    },
    {
      "entropy": 5.618652629852295,
      "epoch": 4.131440493468795,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033926479200344626,
      "loss": 4.7776,
      "mean_token_accuracy": 0.23333349972963333,
      "num_tokens": 94517246.0,
      "step": 45545
    },
    {
      "entropy": 5.613811016082764,
      "epoch": 4.131894049346879,
      "grad_norm": 1.25,
      "learning_rate": 0.0003392337828046572,
      "loss": 4.7922,
      "mean_token_accuracy": 0.23112025260925292,
      "num_tokens": 94527956.0,
      "step": 45550
    },
    {
      "entropy": 5.714540481567383,
      "epoch": 4.132347605224964,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003392027722774971,
      "loss": 4.935,
      "mean_token_accuracy": 0.21406676471233368,
      "num_tokens": 94539101.0,
      "step": 45555
    },
    {
      "entropy": 5.665484237670898,
      "epoch": 4.132801161103048,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033917176042260735,
      "loss": 4.8096,
      "mean_token_accuracy": 0.2254219651222229,
      "num_tokens": 94549597.0,
      "step": 45560
    },
    {
      "entropy": 5.725054740905762,
      "epoch": 4.133254716981132,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003391407472406292,
      "loss": 4.94,
      "mean_token_accuracy": 0.22266182601451873,
      "num_tokens": 94559911.0,
      "step": 45565
    },
    {
      "entropy": 5.667895317077637,
      "epoch": 4.1337082728592165,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033910973273220416,
      "loss": 4.8054,
      "mean_token_accuracy": 0.22711390107870102,
      "num_tokens": 94569396.0,
      "step": 45570
    },
    {
      "entropy": 5.663702917098999,
      "epoch": 4.134161828737301,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033907871689797337,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2295814648270607,
      "num_tokens": 94579817.0,
      "step": 45575
    },
    {
      "entropy": 5.598019647598266,
      "epoch": 4.134615384615385,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003390476997385785,
      "loss": 4.8167,
      "mean_token_accuracy": 0.22663536965847014,
      "num_tokens": 94590398.0,
      "step": 45580
    },
    {
      "entropy": 5.62575478553772,
      "epoch": 4.135068940493468,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003390166812546608,
      "loss": 4.8398,
      "mean_token_accuracy": 0.2249739184975624,
      "num_tokens": 94600528.0,
      "step": 45585
    },
    {
      "entropy": 5.671099233627319,
      "epoch": 4.135522496371553,
      "grad_norm": 1.25,
      "learning_rate": 0.00033898566144686184,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22154261469841002,
      "num_tokens": 94611107.0,
      "step": 45590
    },
    {
      "entropy": 5.73763747215271,
      "epoch": 4.135976052249637,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003389546403158231,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22034264653921126,
      "num_tokens": 94620932.0,
      "step": 45595
    },
    {
      "entropy": 5.735232877731323,
      "epoch": 4.136429608127721,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033892361786218604,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2195568263530731,
      "num_tokens": 94630476.0,
      "step": 45600
    },
    {
      "entropy": 5.634406661987304,
      "epoch": 4.1368831640058055,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003388925940865924,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22063252180814744,
      "num_tokens": 94641487.0,
      "step": 45605
    },
    {
      "entropy": 5.703968477249146,
      "epoch": 4.13733671988389,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033886156898968353,
      "loss": 4.7808,
      "mean_token_accuracy": 0.23053215742111205,
      "num_tokens": 94651668.0,
      "step": 45610
    },
    {
      "entropy": 5.633303499221801,
      "epoch": 4.137790275761974,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033883054257210116,
      "loss": 4.8849,
      "mean_token_accuracy": 0.21612417250871657,
      "num_tokens": 94661218.0,
      "step": 45615
    },
    {
      "entropy": 5.623491001129151,
      "epoch": 4.138243831640058,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000338799514834487,
      "loss": 4.8252,
      "mean_token_accuracy": 0.22095274031162263,
      "num_tokens": 94672412.0,
      "step": 45620
    },
    {
      "entropy": 5.685295724868775,
      "epoch": 4.1386973875181425,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003387684857774824,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21403622925281524,
      "num_tokens": 94683102.0,
      "step": 45625
    },
    {
      "entropy": 5.672245645523072,
      "epoch": 4.139150943396227,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003387374554017294,
      "loss": 4.8613,
      "mean_token_accuracy": 0.23045235872268677,
      "num_tokens": 94692869.0,
      "step": 45630
    },
    {
      "entropy": 5.678238487243652,
      "epoch": 4.13960449927431,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033870642370786966,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23011452853679656,
      "num_tokens": 94702437.0,
      "step": 45635
    },
    {
      "entropy": 5.655371904373169,
      "epoch": 4.1400580551523944,
      "grad_norm": 1.25,
      "learning_rate": 0.00033867539069654476,
      "loss": 4.9,
      "mean_token_accuracy": 0.2237161949276924,
      "num_tokens": 94713130.0,
      "step": 45640
    },
    {
      "entropy": 5.750271606445312,
      "epoch": 4.140511611030479,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033864435636839647,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21517866253852844,
      "num_tokens": 94723251.0,
      "step": 45645
    },
    {
      "entropy": 5.701170539855957,
      "epoch": 4.140965166908563,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003386133207240668,
      "loss": 4.8755,
      "mean_token_accuracy": 0.23121597915887832,
      "num_tokens": 94733805.0,
      "step": 45650
    },
    {
      "entropy": 5.692129898071289,
      "epoch": 4.141418722786647,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003385822837641974,
      "loss": 4.813,
      "mean_token_accuracy": 0.22503754049539565,
      "num_tokens": 94743654.0,
      "step": 45655
    },
    {
      "entropy": 5.6265904903411865,
      "epoch": 4.1418722786647315,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033855124548943025,
      "loss": 4.8237,
      "mean_token_accuracy": 0.2254452720284462,
      "num_tokens": 94754635.0,
      "step": 45660
    },
    {
      "entropy": 5.691283750534057,
      "epoch": 4.142325834542816,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00033852020590040705,
      "loss": 4.8992,
      "mean_token_accuracy": 0.22603995949029923,
      "num_tokens": 94765371.0,
      "step": 45665
    },
    {
      "entropy": 5.653284215927124,
      "epoch": 4.1427793904209,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003384891649977698,
      "loss": 4.8118,
      "mean_token_accuracy": 0.23326751291751863,
      "num_tokens": 94775780.0,
      "step": 45670
    },
    {
      "entropy": 5.6912096500396725,
      "epoch": 4.143232946298984,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033845812278216045,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22231681942939757,
      "num_tokens": 94785345.0,
      "step": 45675
    },
    {
      "entropy": 5.641998767852783,
      "epoch": 4.143686502177069,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033842707925422095,
      "loss": 4.8237,
      "mean_token_accuracy": 0.22652334868907928,
      "num_tokens": 94796059.0,
      "step": 45680
    },
    {
      "entropy": 5.645629262924194,
      "epoch": 4.144140058055152,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033839603441459333,
      "loss": 4.845,
      "mean_token_accuracy": 0.22868782579898833,
      "num_tokens": 94805867.0,
      "step": 45685
    },
    {
      "entropy": 5.629306125640869,
      "epoch": 4.144593613933236,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003383649882639195,
      "loss": 4.854,
      "mean_token_accuracy": 0.22390590608119965,
      "num_tokens": 94817172.0,
      "step": 45690
    },
    {
      "entropy": 5.676585006713867,
      "epoch": 4.1450471698113205,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003383339408028416,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22290365248918534,
      "num_tokens": 94828748.0,
      "step": 45695
    },
    {
      "entropy": 5.641875982284546,
      "epoch": 4.145500725689405,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003383028920320017,
      "loss": 4.8563,
      "mean_token_accuracy": 0.22626406103372573,
      "num_tokens": 94838704.0,
      "step": 45700
    },
    {
      "entropy": 5.683857583999634,
      "epoch": 4.145954281567489,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033827184195204174,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2174295336008072,
      "num_tokens": 94849555.0,
      "step": 45705
    },
    {
      "entropy": 5.719347190856934,
      "epoch": 4.146407837445573,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003382407905636041,
      "loss": 4.9598,
      "mean_token_accuracy": 0.21727394759655,
      "num_tokens": 94862238.0,
      "step": 45710
    },
    {
      "entropy": 5.565261745452881,
      "epoch": 4.1468613933236576,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003382097378673308,
      "loss": 4.7433,
      "mean_token_accuracy": 0.24257777333259584,
      "num_tokens": 94872960.0,
      "step": 45715
    },
    {
      "entropy": 5.603706741333008,
      "epoch": 4.147314949201742,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000338178683863864,
      "loss": 4.8378,
      "mean_token_accuracy": 0.23169205635786055,
      "num_tokens": 94883612.0,
      "step": 45720
    },
    {
      "entropy": 5.629396867752075,
      "epoch": 4.147768505079826,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003381476285538458,
      "loss": 4.808,
      "mean_token_accuracy": 0.23540480583906173,
      "num_tokens": 94893176.0,
      "step": 45725
    },
    {
      "entropy": 5.664601755142212,
      "epoch": 4.14822206095791,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003381165719379187,
      "loss": 4.8227,
      "mean_token_accuracy": 0.2265253707766533,
      "num_tokens": 94903860.0,
      "step": 45730
    },
    {
      "entropy": 5.648194360733032,
      "epoch": 4.148675616835995,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003380855140167247,
      "loss": 4.8075,
      "mean_token_accuracy": 0.22897300720214844,
      "num_tokens": 94915681.0,
      "step": 45735
    },
    {
      "entropy": 5.643002367019653,
      "epoch": 4.149129172714078,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003380544547909061,
      "loss": 4.8389,
      "mean_token_accuracy": 0.2265712082386017,
      "num_tokens": 94927062.0,
      "step": 45740
    },
    {
      "entropy": 5.661525106430053,
      "epoch": 4.149582728592162,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003380233942611055,
      "loss": 4.8535,
      "mean_token_accuracy": 0.22940066307783127,
      "num_tokens": 94938838.0,
      "step": 45745
    },
    {
      "entropy": 5.602386140823365,
      "epoch": 4.1500362844702465,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000337992332427965,
      "loss": 4.8028,
      "mean_token_accuracy": 0.22933536618947983,
      "num_tokens": 94949678.0,
      "step": 45750
    },
    {
      "entropy": 5.668763065338135,
      "epoch": 4.150489840348331,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000337961269292127,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22446864545345308,
      "num_tokens": 94960314.0,
      "step": 45755
    },
    {
      "entropy": 5.675050973892212,
      "epoch": 4.150943396226415,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033793020485423385,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22197601497173308,
      "num_tokens": 94969831.0,
      "step": 45760
    },
    {
      "entropy": 5.681710052490234,
      "epoch": 4.151396952104499,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033789913911492814,
      "loss": 4.8561,
      "mean_token_accuracy": 0.2222761780023575,
      "num_tokens": 94980046.0,
      "step": 45765
    },
    {
      "entropy": 5.628504085540771,
      "epoch": 4.151850507982584,
      "grad_norm": 1.25,
      "learning_rate": 0.0003378680720748522,
      "loss": 4.875,
      "mean_token_accuracy": 0.2255491778254509,
      "num_tokens": 94990850.0,
      "step": 45770
    },
    {
      "entropy": 5.710575342178345,
      "epoch": 4.152304063860668,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003378370037346484,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21948318481445311,
      "num_tokens": 95001658.0,
      "step": 45775
    },
    {
      "entropy": 5.677415180206299,
      "epoch": 4.152757619738752,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00033780593409495936,
      "loss": 4.8381,
      "mean_token_accuracy": 0.22447699755430223,
      "num_tokens": 95011912.0,
      "step": 45780
    },
    {
      "entropy": 5.670226049423218,
      "epoch": 4.153211175616836,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003377748631564276,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22418673783540727,
      "num_tokens": 95021299.0,
      "step": 45785
    },
    {
      "entropy": 5.652838563919067,
      "epoch": 4.15366473149492,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033774379091969583,
      "loss": 4.8478,
      "mean_token_accuracy": 0.23108469396829606,
      "num_tokens": 95031520.0,
      "step": 45790
    },
    {
      "entropy": 5.599844932556152,
      "epoch": 4.154118287373004,
      "grad_norm": 1.25,
      "learning_rate": 0.00033771271738540634,
      "loss": 4.7757,
      "mean_token_accuracy": 0.2340543806552887,
      "num_tokens": 95042127.0,
      "step": 45795
    },
    {
      "entropy": 5.659517574310303,
      "epoch": 4.154571843251088,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033768164255420184,
      "loss": 4.8485,
      "mean_token_accuracy": 0.22394797950983047,
      "num_tokens": 95052181.0,
      "step": 45800
    },
    {
      "entropy": 5.745415258407593,
      "epoch": 4.155025399129173,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033765056642672507,
      "loss": 4.8994,
      "mean_token_accuracy": 0.223867204785347,
      "num_tokens": 95062583.0,
      "step": 45805
    },
    {
      "entropy": 5.710824298858642,
      "epoch": 4.155478955007257,
      "grad_norm": 1.25,
      "learning_rate": 0.00033761948900361857,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2287884086370468,
      "num_tokens": 95073250.0,
      "step": 45810
    },
    {
      "entropy": 5.700158548355103,
      "epoch": 4.155932510885341,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003375884102855251,
      "loss": 4.8609,
      "mean_token_accuracy": 0.22304863929748536,
      "num_tokens": 95083544.0,
      "step": 45815
    },
    {
      "entropy": 5.620473098754883,
      "epoch": 4.156386066763425,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033755733027308746,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22599489241838455,
      "num_tokens": 95093933.0,
      "step": 45820
    },
    {
      "entropy": 5.649737691879272,
      "epoch": 4.15683962264151,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003375262489669481,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22793227732181548,
      "num_tokens": 95104468.0,
      "step": 45825
    },
    {
      "entropy": 5.74627103805542,
      "epoch": 4.157293178519594,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033749516636775014,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22400702089071273,
      "num_tokens": 95114285.0,
      "step": 45830
    },
    {
      "entropy": 5.732165002822876,
      "epoch": 4.157746734397678,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033746408247613616,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2212749734520912,
      "num_tokens": 95123984.0,
      "step": 45835
    },
    {
      "entropy": 5.716455554962158,
      "epoch": 4.1582002902757615,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000337432997292749,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22392043173313142,
      "num_tokens": 95135466.0,
      "step": 45840
    },
    {
      "entropy": 5.637009811401367,
      "epoch": 4.158653846153846,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003374019108182315,
      "loss": 4.7715,
      "mean_token_accuracy": 0.23243832141160964,
      "num_tokens": 95145223.0,
      "step": 45845
    },
    {
      "entropy": 5.736983871459961,
      "epoch": 4.15910740203193,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003373708230532267,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2218468353152275,
      "num_tokens": 95154916.0,
      "step": 45850
    },
    {
      "entropy": 5.657283926010132,
      "epoch": 4.159560957910014,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033733973399837735,
      "loss": 4.8053,
      "mean_token_accuracy": 0.2262278601527214,
      "num_tokens": 95164717.0,
      "step": 45855
    },
    {
      "entropy": 5.738411140441895,
      "epoch": 4.160014513788099,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003373086436543263,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21771416515111924,
      "num_tokens": 95175544.0,
      "step": 45860
    },
    {
      "entropy": 5.658340501785278,
      "epoch": 4.160468069666183,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003372775520217168,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22239864319562913,
      "num_tokens": 95185997.0,
      "step": 45865
    },
    {
      "entropy": 5.632626390457153,
      "epoch": 4.160921625544267,
      "grad_norm": 1.25,
      "learning_rate": 0.0003372464591011916,
      "loss": 4.861,
      "mean_token_accuracy": 0.2191475123167038,
      "num_tokens": 95196967.0,
      "step": 45870
    },
    {
      "entropy": 5.62531361579895,
      "epoch": 4.161375181422351,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033721536489339374,
      "loss": 4.8122,
      "mean_token_accuracy": 0.2264287441968918,
      "num_tokens": 95206788.0,
      "step": 45875
    },
    {
      "entropy": 5.739284181594849,
      "epoch": 4.161828737300436,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033718426939896627,
      "loss": 4.896,
      "mean_token_accuracy": 0.22622128427028657,
      "num_tokens": 95217806.0,
      "step": 45880
    },
    {
      "entropy": 5.6841668605804445,
      "epoch": 4.16228229317852,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033715317261855235,
      "loss": 4.8142,
      "mean_token_accuracy": 0.23467088490724564,
      "num_tokens": 95228311.0,
      "step": 45885
    },
    {
      "entropy": 5.61879529953003,
      "epoch": 4.162735849056604,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033712207455279485,
      "loss": 4.8274,
      "mean_token_accuracy": 0.22531543672084808,
      "num_tokens": 95238872.0,
      "step": 45890
    },
    {
      "entropy": 5.568365049362183,
      "epoch": 4.163189404934688,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003370909752023371,
      "loss": 4.7998,
      "mean_token_accuracy": 0.23678479492664337,
      "num_tokens": 95248575.0,
      "step": 45895
    },
    {
      "entropy": 5.654595994949341,
      "epoch": 4.163642960812772,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003370598745678221,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2235287234187126,
      "num_tokens": 95259323.0,
      "step": 45900
    },
    {
      "entropy": 5.697019910812378,
      "epoch": 4.164096516690856,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003370287726498932,
      "loss": 4.8403,
      "mean_token_accuracy": 0.22598668783903123,
      "num_tokens": 95270018.0,
      "step": 45905
    },
    {
      "entropy": 5.6743532657623295,
      "epoch": 4.16455007256894,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033699766944919325,
      "loss": 4.8166,
      "mean_token_accuracy": 0.23115460723638534,
      "num_tokens": 95279783.0,
      "step": 45910
    },
    {
      "entropy": 5.6850687026977536,
      "epoch": 4.165003628447025,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033696656496636586,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22663683146238328,
      "num_tokens": 95291200.0,
      "step": 45915
    },
    {
      "entropy": 5.70444016456604,
      "epoch": 4.165457184325109,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033693545920205416,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2164785459637642,
      "num_tokens": 95301641.0,
      "step": 45920
    },
    {
      "entropy": 5.6081986904144285,
      "epoch": 4.165910740203193,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033690435215690125,
      "loss": 4.7917,
      "mean_token_accuracy": 0.22975433021783828,
      "num_tokens": 95312636.0,
      "step": 45925
    },
    {
      "entropy": 5.619971752166748,
      "epoch": 4.1663642960812775,
      "grad_norm": 1.375,
      "learning_rate": 0.0003368732438315506,
      "loss": 4.8197,
      "mean_token_accuracy": 0.23392310440540315,
      "num_tokens": 95322168.0,
      "step": 45930
    },
    {
      "entropy": 5.613124370574951,
      "epoch": 4.166817851959362,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033684213422664554,
      "loss": 4.804,
      "mean_token_accuracy": 0.2308719962835312,
      "num_tokens": 95333162.0,
      "step": 45935
    },
    {
      "entropy": 5.610457944869995,
      "epoch": 4.167271407837446,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003368110233428294,
      "loss": 4.8391,
      "mean_token_accuracy": 0.22784648090600967,
      "num_tokens": 95343806.0,
      "step": 45940
    },
    {
      "entropy": 5.655176162719727,
      "epoch": 4.167724963715529,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033677991118074554,
      "loss": 4.767,
      "mean_token_accuracy": 0.23597723990678787,
      "num_tokens": 95353906.0,
      "step": 45945
    },
    {
      "entropy": 5.7291559219360355,
      "epoch": 4.168178519593614,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003367487977410373,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21363867223262786,
      "num_tokens": 95363801.0,
      "step": 45950
    },
    {
      "entropy": 5.703226852416992,
      "epoch": 4.168632075471698,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033671768302434823,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22326904237270356,
      "num_tokens": 95375012.0,
      "step": 45955
    },
    {
      "entropy": 5.72789511680603,
      "epoch": 4.169085631349782,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033668656703132174,
      "loss": 4.8908,
      "mean_token_accuracy": 0.23269332200288773,
      "num_tokens": 95386429.0,
      "step": 45960
    },
    {
      "entropy": 5.634442090988159,
      "epoch": 4.169539187227866,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003366554497626014,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22063346654176713,
      "num_tokens": 95396534.0,
      "step": 45965
    },
    {
      "entropy": 5.65145354270935,
      "epoch": 4.169992743105951,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033662433121883057,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22448988556861876,
      "num_tokens": 95406621.0,
      "step": 45970
    },
    {
      "entropy": 5.681789922714233,
      "epoch": 4.170446298984035,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003365932114006528,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22154108583927154,
      "num_tokens": 95417587.0,
      "step": 45975
    },
    {
      "entropy": 5.605659246444702,
      "epoch": 4.170899854862119,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033656209030871184,
      "loss": 4.7498,
      "mean_token_accuracy": 0.23565778732299805,
      "num_tokens": 95428288.0,
      "step": 45980
    },
    {
      "entropy": 5.643091487884521,
      "epoch": 4.1713534107402035,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033653096794365115,
      "loss": 4.8341,
      "mean_token_accuracy": 0.2264254242181778,
      "num_tokens": 95438093.0,
      "step": 45985
    },
    {
      "entropy": 5.583966159820557,
      "epoch": 4.171806966618288,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033649984430611437,
      "loss": 4.7763,
      "mean_token_accuracy": 0.23059134930372238,
      "num_tokens": 95448397.0,
      "step": 45990
    },
    {
      "entropy": 5.745728158950806,
      "epoch": 4.172260522496371,
      "grad_norm": 1.25,
      "learning_rate": 0.00033646871939674506,
      "loss": 5.0002,
      "mean_token_accuracy": 0.21399605572223662,
      "num_tokens": 95459533.0,
      "step": 45995
    },
    {
      "entropy": 5.6715991497039795,
      "epoch": 4.172714078374455,
      "grad_norm": 1.328125,
      "learning_rate": 0.000336437593216187,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2279481530189514,
      "num_tokens": 95469507.0,
      "step": 46000
    },
    {
      "entropy": 5.7735827445983885,
      "epoch": 4.17316763425254,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003364064657650839,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21621053814888,
      "num_tokens": 95479742.0,
      "step": 46005
    },
    {
      "entropy": 5.619962692260742,
      "epoch": 4.173621190130624,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033637533704407936,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23762890696525574,
      "num_tokens": 95489751.0,
      "step": 46010
    },
    {
      "entropy": 5.716387367248535,
      "epoch": 4.174074746008708,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003363442070538173,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21937399208545685,
      "num_tokens": 95500580.0,
      "step": 46015
    },
    {
      "entropy": 5.741354894638062,
      "epoch": 4.1745283018867925,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033631307579494137,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22390652000904082,
      "num_tokens": 95511018.0,
      "step": 46020
    },
    {
      "entropy": 5.729398632049561,
      "epoch": 4.174981857764877,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033628194326809543,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22720793783664703,
      "num_tokens": 95521220.0,
      "step": 46025
    },
    {
      "entropy": 5.660093545913696,
      "epoch": 4.175435413642961,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033625080947392324,
      "loss": 4.854,
      "mean_token_accuracy": 0.22366981506347655,
      "num_tokens": 95531426.0,
      "step": 46030
    },
    {
      "entropy": 5.709395503997802,
      "epoch": 4.175888969521045,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003362196744130687,
      "loss": 4.8742,
      "mean_token_accuracy": 0.2330963507294655,
      "num_tokens": 95542698.0,
      "step": 46035
    },
    {
      "entropy": 5.709615802764892,
      "epoch": 4.1763425253991295,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033618853808617573,
      "loss": 4.8088,
      "mean_token_accuracy": 0.23700927942991257,
      "num_tokens": 95553545.0,
      "step": 46040
    },
    {
      "entropy": 5.744338417053223,
      "epoch": 4.176796081277214,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003361574004938882,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2242059513926506,
      "num_tokens": 95564452.0,
      "step": 46045
    },
    {
      "entropy": 5.6580479621887205,
      "epoch": 4.177249637155297,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033612626163685,
      "loss": 4.8874,
      "mean_token_accuracy": 0.233695150911808,
      "num_tokens": 95574756.0,
      "step": 46050
    },
    {
      "entropy": 5.74639253616333,
      "epoch": 4.1777031930333814,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033609512151570517,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21630845963954926,
      "num_tokens": 95585147.0,
      "step": 46055
    },
    {
      "entropy": 5.698428583145142,
      "epoch": 4.178156748911466,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033606398013109765,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2246038571000099,
      "num_tokens": 95596025.0,
      "step": 46060
    },
    {
      "entropy": 5.687728500366211,
      "epoch": 4.17861030478955,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003360328374836715,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22005927115678786,
      "num_tokens": 95607191.0,
      "step": 46065
    },
    {
      "entropy": 5.698869609832764,
      "epoch": 4.179063860667634,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003360016935740706,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22165920585393906,
      "num_tokens": 95617256.0,
      "step": 46070
    },
    {
      "entropy": 5.710460042953491,
      "epoch": 4.1795174165457185,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033597054840293927,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22560470551252365,
      "num_tokens": 95627638.0,
      "step": 46075
    },
    {
      "entropy": 5.676480531692505,
      "epoch": 4.179970972423803,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003359394019709214,
      "loss": 4.8217,
      "mean_token_accuracy": 0.22575396448373794,
      "num_tokens": 95637662.0,
      "step": 46080
    },
    {
      "entropy": 5.677263307571411,
      "epoch": 4.180424528301887,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033590825427866117,
      "loss": 4.8396,
      "mean_token_accuracy": 0.22200418263673782,
      "num_tokens": 95647505.0,
      "step": 46085
    },
    {
      "entropy": 5.648023796081543,
      "epoch": 4.180878084179971,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003358771053268027,
      "loss": 4.7962,
      "mean_token_accuracy": 0.21993348151445388,
      "num_tokens": 95657810.0,
      "step": 46090
    },
    {
      "entropy": 5.631044530868531,
      "epoch": 4.181331640058056,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033584595511599024,
      "loss": 4.8107,
      "mean_token_accuracy": 0.22664395570755005,
      "num_tokens": 95667838.0,
      "step": 46095
    },
    {
      "entropy": 5.69027624130249,
      "epoch": 4.181785195936139,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003358148036468679,
      "loss": 4.852,
      "mean_token_accuracy": 0.2323600947856903,
      "num_tokens": 95678497.0,
      "step": 46100
    },
    {
      "entropy": 5.630828619003296,
      "epoch": 4.182238751814223,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033578365092007986,
      "loss": 4.7747,
      "mean_token_accuracy": 0.22508976608514786,
      "num_tokens": 95688656.0,
      "step": 46105
    },
    {
      "entropy": 5.706089019775391,
      "epoch": 4.1826923076923075,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003357524969362705,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2185644045472145,
      "num_tokens": 95699156.0,
      "step": 46110
    },
    {
      "entropy": 5.709915685653686,
      "epoch": 4.183145863570392,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033572134169608395,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23125759065151213,
      "num_tokens": 95709680.0,
      "step": 46115
    },
    {
      "entropy": 5.748576498031616,
      "epoch": 4.183599419448476,
      "grad_norm": 1.375,
      "learning_rate": 0.00033569018520016465,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22202230393886566,
      "num_tokens": 95719142.0,
      "step": 46120
    },
    {
      "entropy": 5.683540487289429,
      "epoch": 4.18405297532656,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003356590274491568,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21833272874355317,
      "num_tokens": 95729415.0,
      "step": 46125
    },
    {
      "entropy": 5.6179084300994875,
      "epoch": 4.1845065312046446,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003356278684437049,
      "loss": 4.8696,
      "mean_token_accuracy": 0.2224570944905281,
      "num_tokens": 95739775.0,
      "step": 46130
    },
    {
      "entropy": 5.711679363250733,
      "epoch": 4.184960087082729,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033559670818445315,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22121995240449904,
      "num_tokens": 95751540.0,
      "step": 46135
    },
    {
      "entropy": 5.797618579864502,
      "epoch": 4.185413642960813,
      "grad_norm": 1.359375,
      "learning_rate": 0.000335565546672046,
      "loss": 5.0424,
      "mean_token_accuracy": 0.2182249128818512,
      "num_tokens": 95761885.0,
      "step": 46140
    },
    {
      "entropy": 5.724494981765747,
      "epoch": 4.185867198838897,
      "grad_norm": 1.21875,
      "learning_rate": 0.000335534383907128,
      "loss": 4.8802,
      "mean_token_accuracy": 0.2263200581073761,
      "num_tokens": 95772284.0,
      "step": 46145
    },
    {
      "entropy": 5.6371794700622555,
      "epoch": 4.186320754716981,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003355032198903434,
      "loss": 4.8135,
      "mean_token_accuracy": 0.22623331099748611,
      "num_tokens": 95783373.0,
      "step": 46150
    },
    {
      "entropy": 5.583518314361572,
      "epoch": 4.186774310595065,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033547205462233694,
      "loss": 4.7584,
      "mean_token_accuracy": 0.23460909575223923,
      "num_tokens": 95793226.0,
      "step": 46155
    },
    {
      "entropy": 5.550963544845581,
      "epoch": 4.187227866473149,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033544088810375287,
      "loss": 4.7725,
      "mean_token_accuracy": 0.22641163170337678,
      "num_tokens": 95803559.0,
      "step": 46160
    },
    {
      "entropy": 5.637325096130371,
      "epoch": 4.1876814223512335,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003354097203352358,
      "loss": 4.8488,
      "mean_token_accuracy": 0.2176286295056343,
      "num_tokens": 95814620.0,
      "step": 46165
    },
    {
      "entropy": 5.6432798385620115,
      "epoch": 4.188134978229318,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003353785513174305,
      "loss": 4.8461,
      "mean_token_accuracy": 0.2286339744925499,
      "num_tokens": 95825024.0,
      "step": 46170
    },
    {
      "entropy": 5.701894378662109,
      "epoch": 4.188588534107402,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033534738105098124,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22606592178344725,
      "num_tokens": 95837373.0,
      "step": 46175
    },
    {
      "entropy": 5.687940168380737,
      "epoch": 4.189042089985486,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00033531620953653276,
      "loss": 4.8707,
      "mean_token_accuracy": 0.22362521588802337,
      "num_tokens": 95846962.0,
      "step": 46180
    },
    {
      "entropy": 5.679882669448853,
      "epoch": 4.189495645863571,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003352850367747297,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22488296329975127,
      "num_tokens": 95857492.0,
      "step": 46185
    },
    {
      "entropy": 5.64149661064148,
      "epoch": 4.189949201741655,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003352538627662168,
      "loss": 4.8141,
      "mean_token_accuracy": 0.2241443157196045,
      "num_tokens": 95867903.0,
      "step": 46190
    },
    {
      "entropy": 5.761031198501587,
      "epoch": 4.190402757619739,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003352226875116386,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2197064623236656,
      "num_tokens": 95878720.0,
      "step": 46195
    },
    {
      "entropy": 5.6713916778564455,
      "epoch": 4.1908563134978225,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003351915110116399,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22056437879800797,
      "num_tokens": 95888406.0,
      "step": 46200
    },
    {
      "entropy": 5.655045127868652,
      "epoch": 4.191309869375907,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033516033326686537,
      "loss": 4.8866,
      "mean_token_accuracy": 0.21591058224439622,
      "num_tokens": 95898431.0,
      "step": 46205
    },
    {
      "entropy": 5.7061375141143795,
      "epoch": 4.191763425253991,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033512915427795975,
      "loss": 4.8522,
      "mean_token_accuracy": 0.23053670078516006,
      "num_tokens": 95909089.0,
      "step": 46210
    },
    {
      "entropy": 5.650598764419556,
      "epoch": 4.192216981132075,
      "grad_norm": 1.328125,
      "learning_rate": 0.000335097974045568,
      "loss": 4.8368,
      "mean_token_accuracy": 0.22529788613319396,
      "num_tokens": 95919787.0,
      "step": 46215
    },
    {
      "entropy": 5.683562994003296,
      "epoch": 4.19267053701016,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033506679257033485,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22522805631160736,
      "num_tokens": 95930620.0,
      "step": 46220
    },
    {
      "entropy": 5.679514217376709,
      "epoch": 4.193124092888244,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003350356098529051,
      "loss": 4.8315,
      "mean_token_accuracy": 0.2304411843419075,
      "num_tokens": 95940258.0,
      "step": 46225
    },
    {
      "entropy": 5.696340322494507,
      "epoch": 4.193577648766328,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003350044258939236,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22667087465524674,
      "num_tokens": 95950062.0,
      "step": 46230
    },
    {
      "entropy": 5.652636623382568,
      "epoch": 4.194031204644412,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033497324069403536,
      "loss": 4.7742,
      "mean_token_accuracy": 0.2318427234888077,
      "num_tokens": 95960658.0,
      "step": 46235
    },
    {
      "entropy": 5.701156949996948,
      "epoch": 4.194484760522497,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003349420542538851,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23048902302980423,
      "num_tokens": 95970497.0,
      "step": 46240
    },
    {
      "entropy": 5.696426820755005,
      "epoch": 4.194938316400581,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033491086657411797,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21871609836816788,
      "num_tokens": 95980405.0,
      "step": 46245
    },
    {
      "entropy": 5.74472074508667,
      "epoch": 4.195391872278664,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003348796776553788,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2173682987689972,
      "num_tokens": 95991111.0,
      "step": 46250
    },
    {
      "entropy": 5.7374592304229735,
      "epoch": 4.1958454281567485,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003348484874983127,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22624988704919816,
      "num_tokens": 96001962.0,
      "step": 46255
    },
    {
      "entropy": 5.670695638656616,
      "epoch": 4.196298984034833,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003348172961035646,
      "loss": 4.8389,
      "mean_token_accuracy": 0.2258397713303566,
      "num_tokens": 96011774.0,
      "step": 46260
    },
    {
      "entropy": 5.6405387878417965,
      "epoch": 4.196752539912917,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003347861034717795,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22053113281726838,
      "num_tokens": 96021915.0,
      "step": 46265
    },
    {
      "entropy": 5.614178800582886,
      "epoch": 4.197206095791001,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033475490960360266,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22341588735580445,
      "num_tokens": 96032851.0,
      "step": 46270
    },
    {
      "entropy": 5.628676700592041,
      "epoch": 4.197659651669086,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000334723714499679,
      "loss": 4.8591,
      "mean_token_accuracy": 0.23020168095827104,
      "num_tokens": 96042630.0,
      "step": 46275
    },
    {
      "entropy": 5.656966257095337,
      "epoch": 4.19811320754717,
      "grad_norm": 1.484375,
      "learning_rate": 0.00033469251816065375,
      "loss": 4.7998,
      "mean_token_accuracy": 0.23116175085306168,
      "num_tokens": 96052154.0,
      "step": 46280
    },
    {
      "entropy": 5.709416818618775,
      "epoch": 4.198566763425254,
      "grad_norm": 1.421875,
      "learning_rate": 0.000334661320587172,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22897300273180007,
      "num_tokens": 96062301.0,
      "step": 46285
    },
    {
      "entropy": 5.721240282058716,
      "epoch": 4.199020319303338,
      "grad_norm": 1.328125,
      "learning_rate": 0.000334630121779879,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2206738695502281,
      "num_tokens": 96072792.0,
      "step": 46290
    },
    {
      "entropy": 5.7347647190094,
      "epoch": 4.199473875181423,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033459892173941973,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21921680271625518,
      "num_tokens": 96082610.0,
      "step": 46295
    },
    {
      "entropy": 5.764332437515259,
      "epoch": 4.199927431059507,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003345677204664397,
      "loss": 4.9504,
      "mean_token_accuracy": 0.22139354050159454,
      "num_tokens": 96093095.0,
      "step": 46300
    },
    {
      "entropy": 5.653356266021729,
      "epoch": 4.20038098693759,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033453651796158397,
      "loss": 4.8444,
      "mean_token_accuracy": 0.23290420025587083,
      "num_tokens": 96103218.0,
      "step": 46305
    },
    {
      "entropy": 5.63716983795166,
      "epoch": 4.200834542815675,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033450531422549786,
      "loss": 4.8362,
      "mean_token_accuracy": 0.22758259177207946,
      "num_tokens": 96113937.0,
      "step": 46310
    },
    {
      "entropy": 5.778512382507325,
      "epoch": 4.201288098693759,
      "grad_norm": 1.59375,
      "learning_rate": 0.00033447410925882673,
      "loss": 5.0159,
      "mean_token_accuracy": 0.221172334253788,
      "num_tokens": 96124378.0,
      "step": 46315
    },
    {
      "entropy": 5.617116403579712,
      "epoch": 4.201741654571843,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003344429030622159,
      "loss": 4.7553,
      "mean_token_accuracy": 0.2324612870812416,
      "num_tokens": 96133657.0,
      "step": 46320
    },
    {
      "entropy": 5.634434843063355,
      "epoch": 4.202195210449927,
      "grad_norm": 1.46875,
      "learning_rate": 0.00033441169563631055,
      "loss": 4.8672,
      "mean_token_accuracy": 0.2298687994480133,
      "num_tokens": 96143669.0,
      "step": 46325
    },
    {
      "entropy": 5.6544252872467045,
      "epoch": 4.202648766328012,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033438048698175625,
      "loss": 4.8427,
      "mean_token_accuracy": 0.22578764855861663,
      "num_tokens": 96154782.0,
      "step": 46330
    },
    {
      "entropy": 5.678867053985596,
      "epoch": 4.203102322206096,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003343492770991984,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22567752599716187,
      "num_tokens": 96165128.0,
      "step": 46335
    },
    {
      "entropy": 5.663485908508301,
      "epoch": 4.20355587808418,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003343180659892824,
      "loss": 4.8845,
      "mean_token_accuracy": 0.22269556373357774,
      "num_tokens": 96176909.0,
      "step": 46340
    },
    {
      "entropy": 5.676246547698975,
      "epoch": 4.2040094339622645,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003342868536526536,
      "loss": 4.8365,
      "mean_token_accuracy": 0.23048171252012253,
      "num_tokens": 96188287.0,
      "step": 46345
    },
    {
      "entropy": 5.603690671920776,
      "epoch": 4.204462989840349,
      "grad_norm": 1.375,
      "learning_rate": 0.00033425564008995764,
      "loss": 4.8345,
      "mean_token_accuracy": 0.22889405637979507,
      "num_tokens": 96198852.0,
      "step": 46350
    },
    {
      "entropy": 5.646171998977661,
      "epoch": 4.204916545718432,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003342244253018399,
      "loss": 4.8626,
      "mean_token_accuracy": 0.2231426492333412,
      "num_tokens": 96208111.0,
      "step": 46355
    },
    {
      "entropy": 5.674980974197387,
      "epoch": 4.205370101596516,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003341932092889459,
      "loss": 4.8409,
      "mean_token_accuracy": 0.2309754431247711,
      "num_tokens": 96218925.0,
      "step": 46360
    },
    {
      "entropy": 5.69445948600769,
      "epoch": 4.205823657474601,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003341619920519213,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2319643661379814,
      "num_tokens": 96228573.0,
      "step": 46365
    },
    {
      "entropy": 5.694979572296143,
      "epoch": 4.206277213352685,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033413077359141166,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22321068197488786,
      "num_tokens": 96239109.0,
      "step": 46370
    },
    {
      "entropy": 5.694328022003174,
      "epoch": 4.206730769230769,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003340995539080625,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21522919684648514,
      "num_tokens": 96249286.0,
      "step": 46375
    },
    {
      "entropy": 5.620876121520996,
      "epoch": 4.207184325108853,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033406833300251947,
      "loss": 4.8452,
      "mean_token_accuracy": 0.22553667575120925,
      "num_tokens": 96259515.0,
      "step": 46380
    },
    {
      "entropy": 5.640081262588501,
      "epoch": 4.207637880986938,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003340371108754284,
      "loss": 4.7976,
      "mean_token_accuracy": 0.22967431098222732,
      "num_tokens": 96269136.0,
      "step": 46385
    },
    {
      "entropy": 5.70844874382019,
      "epoch": 4.208091436865022,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003340058875274347,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22910353690385818,
      "num_tokens": 96279792.0,
      "step": 46390
    },
    {
      "entropy": 5.66708517074585,
      "epoch": 4.208544992743106,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003339746629591842,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2255984514951706,
      "num_tokens": 96289606.0,
      "step": 46395
    },
    {
      "entropy": 5.626698398590088,
      "epoch": 4.2089985486211905,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033394343717132275,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21923031955957412,
      "num_tokens": 96299903.0,
      "step": 46400
    },
    {
      "entropy": 5.752191781997681,
      "epoch": 4.209452104499274,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033391221016449595,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21440760046243668,
      "num_tokens": 96310179.0,
      "step": 46405
    },
    {
      "entropy": 5.725928544998169,
      "epoch": 4.209905660377358,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033388098193934965,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2212996780872345,
      "num_tokens": 96320948.0,
      "step": 46410
    },
    {
      "entropy": 5.627597093582153,
      "epoch": 4.210359216255442,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003338497524965296,
      "loss": 4.8302,
      "mean_token_accuracy": 0.22239601463079453,
      "num_tokens": 96332046.0,
      "step": 46415
    },
    {
      "entropy": 5.6298182010650635,
      "epoch": 4.210812772133527,
      "grad_norm": 1.375,
      "learning_rate": 0.00033381852183668166,
      "loss": 4.7983,
      "mean_token_accuracy": 0.23208268731832504,
      "num_tokens": 96342327.0,
      "step": 46420
    },
    {
      "entropy": 5.641676425933838,
      "epoch": 4.211266328011611,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003337872899604517,
      "loss": 4.8766,
      "mean_token_accuracy": 0.21738712042570113,
      "num_tokens": 96352048.0,
      "step": 46425
    },
    {
      "entropy": 5.651961708068848,
      "epoch": 4.211719883889695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033375605686848566,
      "loss": 4.7863,
      "mean_token_accuracy": 0.2290177971124649,
      "num_tokens": 96362917.0,
      "step": 46430
    },
    {
      "entropy": 5.661551189422608,
      "epoch": 4.2121734397677795,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003337248225614293,
      "loss": 4.8789,
      "mean_token_accuracy": 0.2274763897061348,
      "num_tokens": 96373261.0,
      "step": 46435
    },
    {
      "entropy": 5.675552082061768,
      "epoch": 4.212626995645864,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003336935870399288,
      "loss": 4.9031,
      "mean_token_accuracy": 0.21607214212417603,
      "num_tokens": 96384279.0,
      "step": 46440
    },
    {
      "entropy": 5.742441177368164,
      "epoch": 4.213080551523948,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003336623503046299,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22264773398637772,
      "num_tokens": 96394071.0,
      "step": 46445
    },
    {
      "entropy": 5.658328247070313,
      "epoch": 4.213534107402032,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033363111235617854,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2297147184610367,
      "num_tokens": 96404233.0,
      "step": 46450
    },
    {
      "entropy": 5.6747157096862795,
      "epoch": 4.2139876632801165,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033359987319522095,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22754011899232865,
      "num_tokens": 96415388.0,
      "step": 46455
    },
    {
      "entropy": 5.732484245300293,
      "epoch": 4.2144412191582,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033356863282240295,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22248462736606597,
      "num_tokens": 96425744.0,
      "step": 46460
    },
    {
      "entropy": 5.695106267929077,
      "epoch": 4.214894775036284,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003335373912383707,
      "loss": 4.862,
      "mean_token_accuracy": 0.2353838324546814,
      "num_tokens": 96436340.0,
      "step": 46465
    },
    {
      "entropy": 5.747853851318359,
      "epoch": 4.2153483309143684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003335061484437704,
      "loss": 4.975,
      "mean_token_accuracy": 0.21487181186676024,
      "num_tokens": 96447384.0,
      "step": 46470
    },
    {
      "entropy": 5.647762203216553,
      "epoch": 4.215801886792453,
      "grad_norm": 1.296875,
      "learning_rate": 0.000333474904439248,
      "loss": 4.8439,
      "mean_token_accuracy": 0.23008871972560882,
      "num_tokens": 96456938.0,
      "step": 46475
    },
    {
      "entropy": 5.6212774276733395,
      "epoch": 4.216255442670537,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003334436592254496,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22531364113092422,
      "num_tokens": 96467343.0,
      "step": 46480
    },
    {
      "entropy": 5.652690887451172,
      "epoch": 4.216708998548621,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033341241280302153,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22551099061965943,
      "num_tokens": 96477443.0,
      "step": 46485
    },
    {
      "entropy": 5.660575437545776,
      "epoch": 4.2171625544267055,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033338116517260985,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21605146676301956,
      "num_tokens": 96487711.0,
      "step": 46490
    },
    {
      "entropy": 5.76071195602417,
      "epoch": 4.21761611030479,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003333499163348608,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2178516075015068,
      "num_tokens": 96497839.0,
      "step": 46495
    },
    {
      "entropy": 5.6869344234466555,
      "epoch": 4.218069666182874,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033331866629042055,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2209418907761574,
      "num_tokens": 96508237.0,
      "step": 46500
    },
    {
      "entropy": 5.625384140014648,
      "epoch": 4.218523222060958,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003332874150399355,
      "loss": 4.7795,
      "mean_token_accuracy": 0.22704704105854034,
      "num_tokens": 96518877.0,
      "step": 46505
    },
    {
      "entropy": 5.648231363296508,
      "epoch": 4.218976777939042,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003332561625840519,
      "loss": 4.8399,
      "mean_token_accuracy": 0.22353625744581224,
      "num_tokens": 96530734.0,
      "step": 46510
    },
    {
      "entropy": 5.732207155227661,
      "epoch": 4.219430333817126,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003332249089234159,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22388130128383638,
      "num_tokens": 96540983.0,
      "step": 46515
    },
    {
      "entropy": 5.735537195205689,
      "epoch": 4.21988388969521,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033319365405867394,
      "loss": 4.952,
      "mean_token_accuracy": 0.22156168520450592,
      "num_tokens": 96551866.0,
      "step": 46520
    },
    {
      "entropy": 5.681655502319336,
      "epoch": 4.2203374455732945,
      "grad_norm": 1.375,
      "learning_rate": 0.00033316239799047247,
      "loss": 4.8645,
      "mean_token_accuracy": 0.2233220160007477,
      "num_tokens": 96562515.0,
      "step": 46525
    },
    {
      "entropy": 5.630547428131104,
      "epoch": 4.220791001451379,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033313114071945764,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22379629909992219,
      "num_tokens": 96573266.0,
      "step": 46530
    },
    {
      "entropy": 5.650281715393066,
      "epoch": 4.221244557329463,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003330998822462762,
      "loss": 4.8226,
      "mean_token_accuracy": 0.23221779763698577,
      "num_tokens": 96583070.0,
      "step": 46535
    },
    {
      "entropy": 5.644886445999146,
      "epoch": 4.221698113207547,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033306862257157427,
      "loss": 4.8073,
      "mean_token_accuracy": 0.23255024403333663,
      "num_tokens": 96593582.0,
      "step": 46540
    },
    {
      "entropy": 5.6363588809967045,
      "epoch": 4.2221516690856316,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003330373616959984,
      "loss": 4.8488,
      "mean_token_accuracy": 0.23394003361463547,
      "num_tokens": 96603233.0,
      "step": 46545
    },
    {
      "entropy": 5.6279326438903805,
      "epoch": 4.222605224963716,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033300609962019523,
      "loss": 4.8285,
      "mean_token_accuracy": 0.22774043083190917,
      "num_tokens": 96614361.0,
      "step": 46550
    },
    {
      "entropy": 5.650464200973511,
      "epoch": 4.2230587808418,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000332974836344811,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22408775985240936,
      "num_tokens": 96624161.0,
      "step": 46555
    },
    {
      "entropy": 5.7006810188293455,
      "epoch": 4.2235123367198835,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033294357187049246,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22136003971099855,
      "num_tokens": 96634329.0,
      "step": 46560
    },
    {
      "entropy": 5.7343731880187985,
      "epoch": 4.223965892597968,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033291230619788604,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22745105028152465,
      "num_tokens": 96645575.0,
      "step": 46565
    },
    {
      "entropy": 5.710940170288086,
      "epoch": 4.224419448476052,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033288103932763846,
      "loss": 4.8895,
      "mean_token_accuracy": 0.21736026257276536,
      "num_tokens": 96656224.0,
      "step": 46570
    },
    {
      "entropy": 5.716402959823609,
      "epoch": 4.224873004354136,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003328497712603962,
      "loss": 4.8552,
      "mean_token_accuracy": 0.23283409625291823,
      "num_tokens": 96665950.0,
      "step": 46575
    },
    {
      "entropy": 5.611243534088135,
      "epoch": 4.2253265602322205,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033281850199680586,
      "loss": 4.7493,
      "mean_token_accuracy": 0.23403723090887069,
      "num_tokens": 96676916.0,
      "step": 46580
    },
    {
      "entropy": 5.680910301208496,
      "epoch": 4.225780116110305,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003327872315375142,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21581359058618546,
      "num_tokens": 96687289.0,
      "step": 46585
    },
    {
      "entropy": 5.694920206069947,
      "epoch": 4.226233671988389,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003327559598831679,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22803395241498947,
      "num_tokens": 96698149.0,
      "step": 46590
    },
    {
      "entropy": 5.655440187454223,
      "epoch": 4.226687227866473,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033272468703441363,
      "loss": 4.863,
      "mean_token_accuracy": 0.2261739954352379,
      "num_tokens": 96709466.0,
      "step": 46595
    },
    {
      "entropy": 5.692148208618164,
      "epoch": 4.227140783744558,
      "grad_norm": 1.3125,
      "learning_rate": 0.000332693412991898,
      "loss": 4.9163,
      "mean_token_accuracy": 0.22609069049358368,
      "num_tokens": 96720179.0,
      "step": 46600
    },
    {
      "entropy": 5.730257368087768,
      "epoch": 4.227594339622642,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003326621377562681,
      "loss": 4.921,
      "mean_token_accuracy": 0.21624848544597625,
      "num_tokens": 96730783.0,
      "step": 46605
    },
    {
      "entropy": 5.7256816864013675,
      "epoch": 4.228047895500726,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033263086132817035,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2196589559316635,
      "num_tokens": 96740903.0,
      "step": 46610
    },
    {
      "entropy": 5.677406930923462,
      "epoch": 4.2285014513788095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033259958370825165,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2334198921918869,
      "num_tokens": 96750600.0,
      "step": 46615
    },
    {
      "entropy": 5.657588529586792,
      "epoch": 4.228955007256894,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033256830489715905,
      "loss": 4.8106,
      "mean_token_accuracy": 0.22996080666780472,
      "num_tokens": 96760550.0,
      "step": 46620
    },
    {
      "entropy": 5.690332317352295,
      "epoch": 4.229408563134978,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003325370248955392,
      "loss": 4.8676,
      "mean_token_accuracy": 0.2272874429821968,
      "num_tokens": 96770784.0,
      "step": 46625
    },
    {
      "entropy": 5.643504905700683,
      "epoch": 4.229862119013062,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003325057437040389,
      "loss": 4.7928,
      "mean_token_accuracy": 0.23126513957977296,
      "num_tokens": 96780572.0,
      "step": 46630
    },
    {
      "entropy": 5.642734098434448,
      "epoch": 4.230315674891147,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003324744613233052,
      "loss": 4.7987,
      "mean_token_accuracy": 0.22768006026744841,
      "num_tokens": 96790504.0,
      "step": 46635
    },
    {
      "entropy": 5.6875114917755125,
      "epoch": 4.230769230769231,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033244317775398506,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22599518299102783,
      "num_tokens": 96800856.0,
      "step": 46640
    },
    {
      "entropy": 5.641908311843872,
      "epoch": 4.231222786647315,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003324118929967253,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22116409987211227,
      "num_tokens": 96811293.0,
      "step": 46645
    },
    {
      "entropy": 5.566002941131591,
      "epoch": 4.231676342525399,
      "grad_norm": 1.265625,
      "learning_rate": 0.000332380607052173,
      "loss": 4.7864,
      "mean_token_accuracy": 0.2353133425116539,
      "num_tokens": 96821844.0,
      "step": 46650
    },
    {
      "entropy": 5.67760157585144,
      "epoch": 4.232129898403484,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003323493199209751,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21882238537073134,
      "num_tokens": 96833438.0,
      "step": 46655
    },
    {
      "entropy": 5.632921171188355,
      "epoch": 4.232583454281568,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003323180316037787,
      "loss": 4.8167,
      "mean_token_accuracy": 0.22766333669424058,
      "num_tokens": 96843086.0,
      "step": 46660
    },
    {
      "entropy": 5.688339757919311,
      "epoch": 4.233037010159651,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003322867421012308,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22545151561498641,
      "num_tokens": 96853064.0,
      "step": 46665
    },
    {
      "entropy": 5.76229338645935,
      "epoch": 4.2334905660377355,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003322554514139784,
      "loss": 4.9517,
      "mean_token_accuracy": 0.2213694468140602,
      "num_tokens": 96864213.0,
      "step": 46670
    },
    {
      "entropy": 5.76208724975586,
      "epoch": 4.23394412191582,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033222415954266873,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21810267120599747,
      "num_tokens": 96875344.0,
      "step": 46675
    },
    {
      "entropy": 5.745862627029419,
      "epoch": 4.234397677793904,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033219286648794886,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22638072073459625,
      "num_tokens": 96885985.0,
      "step": 46680
    },
    {
      "entropy": 5.624972105026245,
      "epoch": 4.234851233671988,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003321615722504659,
      "loss": 4.7784,
      "mean_token_accuracy": 0.23417041301727295,
      "num_tokens": 96897512.0,
      "step": 46685
    },
    {
      "entropy": 5.655225849151611,
      "epoch": 4.235304789550073,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003321302768308671,
      "loss": 4.946,
      "mean_token_accuracy": 0.2111332595348358,
      "num_tokens": 96907630.0,
      "step": 46690
    },
    {
      "entropy": 5.615852022171021,
      "epoch": 4.235758345428157,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00033209898022979964,
      "loss": 4.7925,
      "mean_token_accuracy": 0.23420303910970688,
      "num_tokens": 96917400.0,
      "step": 46695
    },
    {
      "entropy": 5.77334303855896,
      "epoch": 4.236211901306241,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003320676824479106,
      "loss": 4.981,
      "mean_token_accuracy": 0.22056981325149536,
      "num_tokens": 96927700.0,
      "step": 46700
    },
    {
      "entropy": 5.718997192382813,
      "epoch": 4.236665457184325,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003320363834858474,
      "loss": 4.8295,
      "mean_token_accuracy": 0.22380704581737518,
      "num_tokens": 96938925.0,
      "step": 46705
    },
    {
      "entropy": 5.634799528121948,
      "epoch": 4.23711901306241,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003320050833442573,
      "loss": 4.8056,
      "mean_token_accuracy": 0.2273781418800354,
      "num_tokens": 96948956.0,
      "step": 46710
    },
    {
      "entropy": 5.637960147857666,
      "epoch": 4.237572568940493,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003319737820237875,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22205977588891984,
      "num_tokens": 96958565.0,
      "step": 46715
    },
    {
      "entropy": 5.636889505386352,
      "epoch": 4.238026124818577,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033194247952508544,
      "loss": 4.8854,
      "mean_token_accuracy": 0.22467081993818283,
      "num_tokens": 96968261.0,
      "step": 46720
    },
    {
      "entropy": 5.707002592086792,
      "epoch": 4.238479680696662,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033191117584879836,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2208097219467163,
      "num_tokens": 96978621.0,
      "step": 46725
    },
    {
      "entropy": 5.661589193344116,
      "epoch": 4.238933236574746,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003318798709955736,
      "loss": 4.8378,
      "mean_token_accuracy": 0.22556596994400024,
      "num_tokens": 96989081.0,
      "step": 46730
    },
    {
      "entropy": 5.76829948425293,
      "epoch": 4.23938679245283,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003318485649660586,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21397027522325515,
      "num_tokens": 97000518.0,
      "step": 46735
    },
    {
      "entropy": 5.673490381240844,
      "epoch": 4.239840348330914,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033181725776090093,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2255713611841202,
      "num_tokens": 97012038.0,
      "step": 46740
    },
    {
      "entropy": 5.68008508682251,
      "epoch": 4.240293904208999,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033178594938074787,
      "loss": 4.8475,
      "mean_token_accuracy": 0.22247387170791627,
      "num_tokens": 97022256.0,
      "step": 46745
    },
    {
      "entropy": 5.691211891174317,
      "epoch": 4.240747460087083,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003317546398262468,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22166019529104233,
      "num_tokens": 97033063.0,
      "step": 46750
    },
    {
      "entropy": 5.657290744781494,
      "epoch": 4.241201015965167,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003317233290980454,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22303344905376435,
      "num_tokens": 97043711.0,
      "step": 46755
    },
    {
      "entropy": 5.696803903579712,
      "epoch": 4.2416545718432515,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033169201719679105,
      "loss": 4.8268,
      "mean_token_accuracy": 0.2341694012284279,
      "num_tokens": 97053582.0,
      "step": 46760
    },
    {
      "entropy": 5.728871393203735,
      "epoch": 4.242108127721336,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033166070412313133,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2251521423459053,
      "num_tokens": 97064535.0,
      "step": 46765
    },
    {
      "entropy": 5.660004806518555,
      "epoch": 4.242561683599419,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003316293898777138,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2206759497523308,
      "num_tokens": 97074295.0,
      "step": 46770
    },
    {
      "entropy": 5.647121143341065,
      "epoch": 4.243015239477503,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033159807446118614,
      "loss": 4.8876,
      "mean_token_accuracy": 0.2193903759121895,
      "num_tokens": 97084781.0,
      "step": 46775
    },
    {
      "entropy": 5.75445065498352,
      "epoch": 4.243468795355588,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003315667578741958,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2187938004732132,
      "num_tokens": 97094848.0,
      "step": 46780
    },
    {
      "entropy": 5.716567182540894,
      "epoch": 4.243922351233672,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003315354401173905,
      "loss": 4.8687,
      "mean_token_accuracy": 0.2293837547302246,
      "num_tokens": 97104747.0,
      "step": 46785
    },
    {
      "entropy": 5.633616733551025,
      "epoch": 4.244375907111756,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003315041211914179,
      "loss": 4.8211,
      "mean_token_accuracy": 0.22519677132368088,
      "num_tokens": 97115599.0,
      "step": 46790
    },
    {
      "entropy": 5.657943773269653,
      "epoch": 4.24482946298984,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003314728010969257,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22055939733982086,
      "num_tokens": 97127492.0,
      "step": 46795
    },
    {
      "entropy": 5.715530729293823,
      "epoch": 4.245283018867925,
      "grad_norm": 1.375,
      "learning_rate": 0.00033144147983456154,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21817528158426286,
      "num_tokens": 97138336.0,
      "step": 46800
    },
    {
      "entropy": 5.676424264907837,
      "epoch": 4.245736574746009,
      "grad_norm": 1.375,
      "learning_rate": 0.0003314101574049732,
      "loss": 4.8414,
      "mean_token_accuracy": 0.2285957530140877,
      "num_tokens": 97149196.0,
      "step": 46805
    },
    {
      "entropy": 5.719177150726319,
      "epoch": 4.246190130624093,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00033137883380880847,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22124050855636596,
      "num_tokens": 97159622.0,
      "step": 46810
    },
    {
      "entropy": 5.732594442367554,
      "epoch": 4.2466436865021775,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000331347509046715,
      "loss": 4.97,
      "mean_token_accuracy": 0.21201612800359726,
      "num_tokens": 97170149.0,
      "step": 46815
    },
    {
      "entropy": 5.739738893508911,
      "epoch": 4.247097242380261,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003313161831193408,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2203262284398079,
      "num_tokens": 97180409.0,
      "step": 46820
    },
    {
      "entropy": 5.701783609390259,
      "epoch": 4.247550798258345,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033128485602733354,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22342330813407899,
      "num_tokens": 97190899.0,
      "step": 46825
    },
    {
      "entropy": 5.721866035461426,
      "epoch": 4.248004354136429,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003312535277713411,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22733433693647384,
      "num_tokens": 97200759.0,
      "step": 46830
    },
    {
      "entropy": 5.685166883468628,
      "epoch": 4.248457910014514,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033122219835201135,
      "loss": 4.8206,
      "mean_token_accuracy": 0.2319476678967476,
      "num_tokens": 97211476.0,
      "step": 46835
    },
    {
      "entropy": 5.7116588115692135,
      "epoch": 4.248911465892598,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003311908677699923,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21850743591785432,
      "num_tokens": 97222374.0,
      "step": 46840
    },
    {
      "entropy": 5.695628547668457,
      "epoch": 4.249365021770682,
      "grad_norm": 1.375,
      "learning_rate": 0.0003311595360259317,
      "loss": 4.8984,
      "mean_token_accuracy": 0.22292670607566833,
      "num_tokens": 97233079.0,
      "step": 46845
    },
    {
      "entropy": 5.598395824432373,
      "epoch": 4.2498185776487665,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033112820312047756,
      "loss": 4.7892,
      "mean_token_accuracy": 0.23457126766443254,
      "num_tokens": 97243796.0,
      "step": 46850
    },
    {
      "entropy": 5.626037979125977,
      "epoch": 4.250272133526851,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033109686905427794,
      "loss": 4.7605,
      "mean_token_accuracy": 0.23499299436807633,
      "num_tokens": 97253341.0,
      "step": 46855
    },
    {
      "entropy": 5.733515787124634,
      "epoch": 4.250725689404935,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033106553382798075,
      "loss": 4.942,
      "mean_token_accuracy": 0.21816972643136978,
      "num_tokens": 97264244.0,
      "step": 46860
    },
    {
      "entropy": 5.645431280136108,
      "epoch": 4.251179245283019,
      "grad_norm": 1.375,
      "learning_rate": 0.00033103419744223413,
      "loss": 4.8458,
      "mean_token_accuracy": 0.22767426669597626,
      "num_tokens": 97274239.0,
      "step": 46865
    },
    {
      "entropy": 5.613644599914551,
      "epoch": 4.251632801161103,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003310028598976859,
      "loss": 4.7696,
      "mean_token_accuracy": 0.2356210932135582,
      "num_tokens": 97284709.0,
      "step": 46870
    },
    {
      "entropy": 5.620578193664551,
      "epoch": 4.252086357039187,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033097152119498425,
      "loss": 4.799,
      "mean_token_accuracy": 0.22977874726057052,
      "num_tokens": 97295262.0,
      "step": 46875
    },
    {
      "entropy": 5.672809267044068,
      "epoch": 4.252539912917271,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003309401813347774,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2187335103750229,
      "num_tokens": 97306143.0,
      "step": 46880
    },
    {
      "entropy": 5.704710197448731,
      "epoch": 4.2529934687953554,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003309088403177132,
      "loss": 4.8117,
      "mean_token_accuracy": 0.23305461257696153,
      "num_tokens": 97316248.0,
      "step": 46885
    },
    {
      "entropy": 5.750253343582154,
      "epoch": 4.25344702467344,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033087749814444,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22110236883163453,
      "num_tokens": 97327704.0,
      "step": 46890
    },
    {
      "entropy": 5.744761657714844,
      "epoch": 4.253900580551524,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033084615481560585,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21289312839508057,
      "num_tokens": 97337817.0,
      "step": 46895
    },
    {
      "entropy": 5.731951761245727,
      "epoch": 4.254354136429608,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033081481033185903,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22055644243955613,
      "num_tokens": 97348315.0,
      "step": 46900
    },
    {
      "entropy": 5.659780740737915,
      "epoch": 4.2548076923076925,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003307834646938477,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22482480108737946,
      "num_tokens": 97358842.0,
      "step": 46905
    },
    {
      "entropy": 5.7696269989013675,
      "epoch": 4.255261248185777,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003307521179022201,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21794855296611787,
      "num_tokens": 97369247.0,
      "step": 46910
    },
    {
      "entropy": 5.68321213722229,
      "epoch": 4.255714804063861,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003307207699576246,
      "loss": 4.8448,
      "mean_token_accuracy": 0.22952199280261992,
      "num_tokens": 97379668.0,
      "step": 46915
    },
    {
      "entropy": 5.663760709762573,
      "epoch": 4.256168359941945,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033068942086070924,
      "loss": 4.8478,
      "mean_token_accuracy": 0.2255804255604744,
      "num_tokens": 97389131.0,
      "step": 46920
    },
    {
      "entropy": 5.730261564254761,
      "epoch": 4.256621915820029,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033065807061212254,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2156697317957878,
      "num_tokens": 97400583.0,
      "step": 46925
    },
    {
      "entropy": 5.7204207420349125,
      "epoch": 4.257075471698113,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003306267192125128,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2233791321516037,
      "num_tokens": 97410828.0,
      "step": 46930
    },
    {
      "entropy": 5.717368221282959,
      "epoch": 4.257529027576197,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003305953666625283,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2185799479484558,
      "num_tokens": 97421190.0,
      "step": 46935
    },
    {
      "entropy": 5.66599850654602,
      "epoch": 4.2579825834542815,
      "grad_norm": 1.25,
      "learning_rate": 0.00033056401296281733,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21589289009571075,
      "num_tokens": 97432257.0,
      "step": 46940
    },
    {
      "entropy": 5.674975824356079,
      "epoch": 4.258436139332366,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003305326581140286,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22064141631126405,
      "num_tokens": 97442272.0,
      "step": 46945
    },
    {
      "entropy": 5.667259407043457,
      "epoch": 4.25888969521045,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033050130211681017,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22867862433195113,
      "num_tokens": 97451462.0,
      "step": 46950
    },
    {
      "entropy": 5.651415634155273,
      "epoch": 4.259343251088534,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033046994497181077,
      "loss": 4.7692,
      "mean_token_accuracy": 0.22875206917524338,
      "num_tokens": 97461579.0,
      "step": 46955
    },
    {
      "entropy": 5.762288522720337,
      "epoch": 4.2597968069666186,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033043858667967875,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2201904386281967,
      "num_tokens": 97472892.0,
      "step": 46960
    },
    {
      "entropy": 5.644103956222534,
      "epoch": 4.260250362844703,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033040722724106257,
      "loss": 4.8386,
      "mean_token_accuracy": 0.22571358531713487,
      "num_tokens": 97484278.0,
      "step": 46965
    },
    {
      "entropy": 5.659617280960083,
      "epoch": 4.260703918722786,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033037586665661094,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2219989240169525,
      "num_tokens": 97494127.0,
      "step": 46970
    },
    {
      "entropy": 5.656411266326904,
      "epoch": 4.2611574746008705,
      "grad_norm": 1.25,
      "learning_rate": 0.0003303445049269722,
      "loss": 4.8025,
      "mean_token_accuracy": 0.23088260889053344,
      "num_tokens": 97504213.0,
      "step": 46975
    },
    {
      "entropy": 5.74541540145874,
      "epoch": 4.261611030478955,
      "grad_norm": 1.46875,
      "learning_rate": 0.00033031314205279503,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22348013669252395,
      "num_tokens": 97515242.0,
      "step": 46980
    },
    {
      "entropy": 5.757241582870483,
      "epoch": 4.262064586357039,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033028177803472793,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2247408375144005,
      "num_tokens": 97525956.0,
      "step": 46985
    },
    {
      "entropy": 5.6540446281433105,
      "epoch": 4.262518142235123,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003302504128734196,
      "loss": 4.8023,
      "mean_token_accuracy": 0.23293780237436296,
      "num_tokens": 97535973.0,
      "step": 46990
    },
    {
      "entropy": 5.630394220352173,
      "epoch": 4.2629716981132075,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033021904656951867,
      "loss": 4.8727,
      "mean_token_accuracy": 0.22867006212472915,
      "num_tokens": 97547580.0,
      "step": 46995
    },
    {
      "entropy": 5.666845893859863,
      "epoch": 4.263425253991292,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033018767912367376,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22088335007429122,
      "num_tokens": 97558050.0,
      "step": 47000
    },
    {
      "entropy": 5.710594606399536,
      "epoch": 4.263878809869376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033015631053653357,
      "loss": 4.8217,
      "mean_token_accuracy": 0.2318890631198883,
      "num_tokens": 97567997.0,
      "step": 47005
    },
    {
      "entropy": 5.675081968307495,
      "epoch": 4.26433236574746,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003301249408087469,
      "loss": 4.8451,
      "mean_token_accuracy": 0.2296413630247116,
      "num_tokens": 97577220.0,
      "step": 47010
    },
    {
      "entropy": 5.6979320526123045,
      "epoch": 4.264785921625545,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033009356994096237,
      "loss": 4.8885,
      "mean_token_accuracy": 0.2285613849759102,
      "num_tokens": 97588421.0,
      "step": 47015
    },
    {
      "entropy": 5.6226012229919435,
      "epoch": 4.265239477503629,
      "grad_norm": 1.25,
      "learning_rate": 0.00033006219793382867,
      "loss": 4.793,
      "mean_token_accuracy": 0.23029449880123137,
      "num_tokens": 97598624.0,
      "step": 47020
    },
    {
      "entropy": 5.691341543197632,
      "epoch": 4.265693033381712,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003300308247879948,
      "loss": 4.8814,
      "mean_token_accuracy": 0.2207088738679886,
      "num_tokens": 97609128.0,
      "step": 47025
    },
    {
      "entropy": 5.6923095703125,
      "epoch": 4.2661465892597965,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003299994505041094,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22919802665710448,
      "num_tokens": 97619549.0,
      "step": 47030
    },
    {
      "entropy": 5.6800089359283445,
      "epoch": 4.266600145137881,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032996807508282134,
      "loss": 4.8901,
      "mean_token_accuracy": 0.22280146628618241,
      "num_tokens": 97630184.0,
      "step": 47035
    },
    {
      "entropy": 5.660915565490723,
      "epoch": 4.267053701015965,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032993669852477954,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22722068428993225,
      "num_tokens": 97640569.0,
      "step": 47040
    },
    {
      "entropy": 5.659248495101929,
      "epoch": 4.267507256894049,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003299053208306328,
      "loss": 4.8312,
      "mean_token_accuracy": 0.2284248188138008,
      "num_tokens": 97651006.0,
      "step": 47045
    },
    {
      "entropy": 5.658032178878784,
      "epoch": 4.267960812772134,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003298739420010299,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22692545056343078,
      "num_tokens": 97661567.0,
      "step": 47050
    },
    {
      "entropy": 5.650281095504761,
      "epoch": 4.268414368650218,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003298425620366201,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22889568358659745,
      "num_tokens": 97673058.0,
      "step": 47055
    },
    {
      "entropy": 5.781052684783935,
      "epoch": 4.268867924528302,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003298111809380521,
      "loss": 5.0038,
      "mean_token_accuracy": 0.21487070620059967,
      "num_tokens": 97684509.0,
      "step": 47060
    },
    {
      "entropy": 5.764691877365112,
      "epoch": 4.269321480406386,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032977979870597477,
      "loss": 4.8948,
      "mean_token_accuracy": 0.2179829940199852,
      "num_tokens": 97693954.0,
      "step": 47065
    },
    {
      "entropy": 5.699521827697754,
      "epoch": 4.269775036284471,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003297484153410374,
      "loss": 4.8238,
      "mean_token_accuracy": 0.2289520174264908,
      "num_tokens": 97704345.0,
      "step": 47070
    },
    {
      "entropy": 5.611645889282227,
      "epoch": 4.270228592162554,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032971703084388885,
      "loss": 4.7461,
      "mean_token_accuracy": 0.2410870298743248,
      "num_tokens": 97714857.0,
      "step": 47075
    },
    {
      "entropy": 5.671855115890503,
      "epoch": 4.270682148040638,
      "grad_norm": 1.234375,
      "learning_rate": 0.000329685645215178,
      "loss": 4.892,
      "mean_token_accuracy": 0.2232367977499962,
      "num_tokens": 97725921.0,
      "step": 47080
    },
    {
      "entropy": 5.6664213180542,
      "epoch": 4.2711357039187225,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032965425845555413,
      "loss": 4.8581,
      "mean_token_accuracy": 0.2381445974111557,
      "num_tokens": 97735361.0,
      "step": 47085
    },
    {
      "entropy": 5.665806865692138,
      "epoch": 4.271589259796807,
      "grad_norm": 1.25,
      "learning_rate": 0.00032962287056566633,
      "loss": 4.8777,
      "mean_token_accuracy": 0.2219262659549713,
      "num_tokens": 97745915.0,
      "step": 47090
    },
    {
      "entropy": 5.698756551742553,
      "epoch": 4.272042815674891,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003295914815461637,
      "loss": 4.855,
      "mean_token_accuracy": 0.23077422380447388,
      "num_tokens": 97755663.0,
      "step": 47095
    },
    {
      "entropy": 5.738223314285278,
      "epoch": 4.272496371552975,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032956009139769514,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21545568406581878,
      "num_tokens": 97766300.0,
      "step": 47100
    },
    {
      "entropy": 5.7091155529022215,
      "epoch": 4.27294992743106,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032952870012091013,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2232059806585312,
      "num_tokens": 97776584.0,
      "step": 47105
    },
    {
      "entropy": 5.651587915420532,
      "epoch": 4.273403483309144,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032949730771645767,
      "loss": 4.8243,
      "mean_token_accuracy": 0.22760859429836272,
      "num_tokens": 97786230.0,
      "step": 47110
    },
    {
      "entropy": 5.718541860580444,
      "epoch": 4.273857039187228,
      "grad_norm": 1.3125,
      "learning_rate": 0.000329465914184987,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21965525448322296,
      "num_tokens": 97796671.0,
      "step": 47115
    },
    {
      "entropy": 5.640603303909302,
      "epoch": 4.274310595065312,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003294345195271473,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22892607748508453,
      "num_tokens": 97807569.0,
      "step": 47120
    },
    {
      "entropy": 5.742707204818726,
      "epoch": 4.274764150943396,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032940312374358794,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21365204751491546,
      "num_tokens": 97818680.0,
      "step": 47125
    },
    {
      "entropy": 5.7019205570220945,
      "epoch": 4.27521770682148,
      "grad_norm": 1.421875,
      "learning_rate": 0.000329371726834958,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2238304927945137,
      "num_tokens": 97828955.0,
      "step": 47130
    },
    {
      "entropy": 5.687495470046997,
      "epoch": 4.275671262699564,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000329340328801907,
      "loss": 4.874,
      "mean_token_accuracy": 0.23057658076286316,
      "num_tokens": 97839000.0,
      "step": 47135
    },
    {
      "entropy": 5.657153940200805,
      "epoch": 4.276124818577649,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003293089296450841,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2144649088382721,
      "num_tokens": 97850140.0,
      "step": 47140
    },
    {
      "entropy": 5.654101753234864,
      "epoch": 4.276578374455733,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003292775293651388,
      "loss": 4.7942,
      "mean_token_accuracy": 0.2294245958328247,
      "num_tokens": 97860887.0,
      "step": 47145
    },
    {
      "entropy": 5.698821640014648,
      "epoch": 4.277031930333817,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003292461279627203,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22793372720479965,
      "num_tokens": 97871190.0,
      "step": 47150
    },
    {
      "entropy": 5.663764429092407,
      "epoch": 4.277485486211901,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032921472543847804,
      "loss": 4.8041,
      "mean_token_accuracy": 0.2328104481101036,
      "num_tokens": 97881479.0,
      "step": 47155
    },
    {
      "entropy": 5.631919717788696,
      "epoch": 4.277939042089986,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003291833217930614,
      "loss": 4.878,
      "mean_token_accuracy": 0.22875399589538575,
      "num_tokens": 97892340.0,
      "step": 47160
    },
    {
      "entropy": 5.676173973083496,
      "epoch": 4.27839259796807,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032915191702711995,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22004313319921492,
      "num_tokens": 97902882.0,
      "step": 47165
    },
    {
      "entropy": 5.627889823913574,
      "epoch": 4.278846153846154,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032912051114130297,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23036182224750518,
      "num_tokens": 97913187.0,
      "step": 47170
    },
    {
      "entropy": 5.624239444732666,
      "epoch": 4.2792997097242385,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00032908910413626,
      "loss": 4.8221,
      "mean_token_accuracy": 0.23218010663986205,
      "num_tokens": 97922434.0,
      "step": 47175
    },
    {
      "entropy": 5.548560905456543,
      "epoch": 4.279753265602322,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003290576960126407,
      "loss": 4.7395,
      "mean_token_accuracy": 0.23581626266241074,
      "num_tokens": 97933020.0,
      "step": 47180
    },
    {
      "entropy": 5.6319740295410154,
      "epoch": 4.280206821480406,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032902628677109434,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2259765788912773,
      "num_tokens": 97944103.0,
      "step": 47185
    },
    {
      "entropy": 5.709651136398316,
      "epoch": 4.28066037735849,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003289948764122707,
      "loss": 4.8565,
      "mean_token_accuracy": 0.23097035884857178,
      "num_tokens": 97953692.0,
      "step": 47190
    },
    {
      "entropy": 5.72479887008667,
      "epoch": 4.281113933236575,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003289634649368193,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22914205342531205,
      "num_tokens": 97963592.0,
      "step": 47195
    },
    {
      "entropy": 5.751935958862305,
      "epoch": 4.281567489114659,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003289320523453895,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21587749272584916,
      "num_tokens": 97974092.0,
      "step": 47200
    },
    {
      "entropy": 5.6376060962677,
      "epoch": 4.282021044992743,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032890063863863126,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23120422214269637,
      "num_tokens": 97984175.0,
      "step": 47205
    },
    {
      "entropy": 5.636980056762695,
      "epoch": 4.282474600870827,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003288692238171941,
      "loss": 4.8874,
      "mean_token_accuracy": 0.23617047369480132,
      "num_tokens": 97994993.0,
      "step": 47210
    },
    {
      "entropy": 5.6741691589355465,
      "epoch": 4.282928156748912,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032883780788172756,
      "loss": 4.9002,
      "mean_token_accuracy": 0.218203404545784,
      "num_tokens": 98005892.0,
      "step": 47215
    },
    {
      "entropy": 5.623246669769287,
      "epoch": 4.283381712626996,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003288063908328814,
      "loss": 4.7963,
      "mean_token_accuracy": 0.23167288154363633,
      "num_tokens": 98016726.0,
      "step": 47220
    },
    {
      "entropy": 5.725933265686035,
      "epoch": 4.28383526850508,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003287749726713055,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22654304951429366,
      "num_tokens": 98028500.0,
      "step": 47225
    },
    {
      "entropy": 5.752040052413941,
      "epoch": 4.284288824383164,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003287435533976494,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21943883150815963,
      "num_tokens": 98039304.0,
      "step": 47230
    },
    {
      "entropy": 5.695298337936402,
      "epoch": 4.284742380261248,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003287121330125629,
      "loss": 4.8641,
      "mean_token_accuracy": 0.22220477908849717,
      "num_tokens": 98049448.0,
      "step": 47235
    },
    {
      "entropy": 5.701738929748535,
      "epoch": 4.285195936139332,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032868071151669585,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2332577019929886,
      "num_tokens": 98059532.0,
      "step": 47240
    },
    {
      "entropy": 5.61998610496521,
      "epoch": 4.285649492017416,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000328649288910698,
      "loss": 4.7823,
      "mean_token_accuracy": 0.22975722551345826,
      "num_tokens": 98069385.0,
      "step": 47245
    },
    {
      "entropy": 5.694291448593139,
      "epoch": 4.286103047895501,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003286178651952192,
      "loss": 4.8491,
      "mean_token_accuracy": 0.2296536386013031,
      "num_tokens": 98079354.0,
      "step": 47250
    },
    {
      "entropy": 5.711437082290649,
      "epoch": 4.286556603773585,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032858644037090924,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21793979853391648,
      "num_tokens": 98090864.0,
      "step": 47255
    },
    {
      "entropy": 5.66371111869812,
      "epoch": 4.287010159651669,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003285550144384181,
      "loss": 4.8395,
      "mean_token_accuracy": 0.22818129807710646,
      "num_tokens": 98101515.0,
      "step": 47260
    },
    {
      "entropy": 5.702941465377807,
      "epoch": 4.2874637155297535,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032852358739839555,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22469830214977266,
      "num_tokens": 98111786.0,
      "step": 47265
    },
    {
      "entropy": 5.683735179901123,
      "epoch": 4.287917271407838,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032849215925149155,
      "loss": 4.8073,
      "mean_token_accuracy": 0.23230572789907455,
      "num_tokens": 98121732.0,
      "step": 47270
    },
    {
      "entropy": 5.724025249481201,
      "epoch": 4.288370827285922,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003284607299983561,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21983799934387208,
      "num_tokens": 98131417.0,
      "step": 47275
    },
    {
      "entropy": 5.7351508140563965,
      "epoch": 4.288824383164005,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032842929963963915,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2256147414445877,
      "num_tokens": 98141664.0,
      "step": 47280
    },
    {
      "entropy": 5.666127300262451,
      "epoch": 4.28927793904209,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032839786817599065,
      "loss": 4.8204,
      "mean_token_accuracy": 0.22550419121980667,
      "num_tokens": 98150959.0,
      "step": 47285
    },
    {
      "entropy": 5.682355499267578,
      "epoch": 4.289731494920174,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003283664356080606,
      "loss": 4.966,
      "mean_token_accuracy": 0.2208641618490219,
      "num_tokens": 98161957.0,
      "step": 47290
    },
    {
      "entropy": 5.632346057891846,
      "epoch": 4.290185050798258,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032833500193649916,
      "loss": 4.8001,
      "mean_token_accuracy": 0.22397990375757218,
      "num_tokens": 98172053.0,
      "step": 47295
    },
    {
      "entropy": 5.67668514251709,
      "epoch": 4.2906386066763424,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032830356716195624,
      "loss": 4.8255,
      "mean_token_accuracy": 0.2245911180973053,
      "num_tokens": 98182074.0,
      "step": 47300
    },
    {
      "entropy": 5.725739431381226,
      "epoch": 4.291092162554427,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003282721312850819,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2233437180519104,
      "num_tokens": 98192591.0,
      "step": 47305
    },
    {
      "entropy": 5.674335956573486,
      "epoch": 4.291545718432511,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003282406943065264,
      "loss": 4.7824,
      "mean_token_accuracy": 0.23457320034503937,
      "num_tokens": 98203426.0,
      "step": 47310
    },
    {
      "entropy": 5.637891721725464,
      "epoch": 4.291999274310595,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032820925622693977,
      "loss": 4.8485,
      "mean_token_accuracy": 0.22813299000263215,
      "num_tokens": 98214342.0,
      "step": 47315
    },
    {
      "entropy": 5.680193567276001,
      "epoch": 4.2924528301886795,
      "grad_norm": 1.25,
      "learning_rate": 0.00032817781704697204,
      "loss": 4.8473,
      "mean_token_accuracy": 0.22287966310977936,
      "num_tokens": 98224155.0,
      "step": 47320
    },
    {
      "entropy": 5.663097906112671,
      "epoch": 4.292906386066764,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003281463767672736,
      "loss": 4.8274,
      "mean_token_accuracy": 0.23103382289409638,
      "num_tokens": 98233613.0,
      "step": 47325
    },
    {
      "entropy": 5.658789014816284,
      "epoch": 4.293359941944848,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003281149353884946,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2316870391368866,
      "num_tokens": 98243117.0,
      "step": 47330
    },
    {
      "entropy": 5.656659364700317,
      "epoch": 4.293813497822931,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003280834929112851,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22910133004188538,
      "num_tokens": 98253619.0,
      "step": 47335
    },
    {
      "entropy": 5.654382658004761,
      "epoch": 4.294267053701016,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003280520493362955,
      "loss": 4.8647,
      "mean_token_accuracy": 0.22564778476953506,
      "num_tokens": 98264221.0,
      "step": 47340
    },
    {
      "entropy": 5.663712453842163,
      "epoch": 4.2947206095791,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003280206046641759,
      "loss": 4.8117,
      "mean_token_accuracy": 0.2270164504647255,
      "num_tokens": 98275530.0,
      "step": 47345
    },
    {
      "entropy": 5.76324462890625,
      "epoch": 4.295174165457184,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032798915889557687,
      "loss": 4.9546,
      "mean_token_accuracy": 0.2223445236682892,
      "num_tokens": 98285937.0,
      "step": 47350
    },
    {
      "entropy": 5.722216606140137,
      "epoch": 4.2956277213352685,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003279577120311483,
      "loss": 4.8921,
      "mean_token_accuracy": 0.22277185022830964,
      "num_tokens": 98295690.0,
      "step": 47355
    },
    {
      "entropy": 5.6390468120574955,
      "epoch": 4.296081277213353,
      "grad_norm": 1.40625,
      "learning_rate": 0.000327926264071541,
      "loss": 4.8573,
      "mean_token_accuracy": 0.2283207058906555,
      "num_tokens": 98305073.0,
      "step": 47360
    },
    {
      "entropy": 5.696509647369385,
      "epoch": 4.296534833091437,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003278948150174049,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2187576100230217,
      "num_tokens": 98316008.0,
      "step": 47365
    },
    {
      "entropy": 5.651239633560181,
      "epoch": 4.296988388969521,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003278633648693906,
      "loss": 4.8309,
      "mean_token_accuracy": 0.22334285080432892,
      "num_tokens": 98327032.0,
      "step": 47370
    },
    {
      "entropy": 5.710759210586548,
      "epoch": 4.2974419448476056,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003278319136281484,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22898487001657486,
      "num_tokens": 98336085.0,
      "step": 47375
    },
    {
      "entropy": 5.698973941802978,
      "epoch": 4.297895500725689,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032780046129432885,
      "loss": 4.881,
      "mean_token_accuracy": 0.23336286693811417,
      "num_tokens": 98346180.0,
      "step": 47380
    },
    {
      "entropy": 5.6831762313842775,
      "epoch": 4.298349056603773,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032776900786858214,
      "loss": 4.8081,
      "mean_token_accuracy": 0.22490776032209397,
      "num_tokens": 98356001.0,
      "step": 47385
    },
    {
      "entropy": 5.735995531082153,
      "epoch": 4.2988026124818575,
      "grad_norm": 1.359375,
      "learning_rate": 0.000327737553351559,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22149443626403809,
      "num_tokens": 98366744.0,
      "step": 47390
    },
    {
      "entropy": 5.636945629119873,
      "epoch": 4.299256168359942,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032770609774390985,
      "loss": 4.8184,
      "mean_token_accuracy": 0.2351287066936493,
      "num_tokens": 98377350.0,
      "step": 47395
    },
    {
      "entropy": 5.64010705947876,
      "epoch": 4.299709724238026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003276746410462852,
      "loss": 4.849,
      "mean_token_accuracy": 0.23031098395586014,
      "num_tokens": 98387926.0,
      "step": 47400
    },
    {
      "entropy": 5.757110738754273,
      "epoch": 4.30016328011611,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00032764318325933546,
      "loss": 4.9338,
      "mean_token_accuracy": 0.22473542392253876,
      "num_tokens": 98397117.0,
      "step": 47405
    },
    {
      "entropy": 5.752656745910644,
      "epoch": 4.3006168359941945,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003276117243837113,
      "loss": 4.9051,
      "mean_token_accuracy": 0.22306031882762908,
      "num_tokens": 98407983.0,
      "step": 47410
    },
    {
      "entropy": 5.693604421615601,
      "epoch": 4.301070391872279,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003275802644200633,
      "loss": 4.908,
      "mean_token_accuracy": 0.2198258236050606,
      "num_tokens": 98419940.0,
      "step": 47415
    },
    {
      "entropy": 5.627506589889526,
      "epoch": 4.301523947750363,
      "grad_norm": 1.453125,
      "learning_rate": 0.000327548803369042,
      "loss": 4.8164,
      "mean_token_accuracy": 0.23481301814317704,
      "num_tokens": 98430487.0,
      "step": 47420
    },
    {
      "entropy": 5.664304542541504,
      "epoch": 4.301977503628447,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032751734123129814,
      "loss": 4.8511,
      "mean_token_accuracy": 0.231725250184536,
      "num_tokens": 98440231.0,
      "step": 47425
    },
    {
      "entropy": 5.7077515602111815,
      "epoch": 4.302431059506532,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003274858780074822,
      "loss": 4.9194,
      "mean_token_accuracy": 0.2230115592479706,
      "num_tokens": 98449741.0,
      "step": 47430
    },
    {
      "entropy": 5.6739737033844,
      "epoch": 4.302884615384615,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032745441369824485,
      "loss": 4.8016,
      "mean_token_accuracy": 0.2390468329191208,
      "num_tokens": 98461068.0,
      "step": 47435
    },
    {
      "entropy": 5.683884954452514,
      "epoch": 4.303338171262699,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032742294830423696,
      "loss": 4.8982,
      "mean_token_accuracy": 0.2266627088189125,
      "num_tokens": 98472585.0,
      "step": 47440
    },
    {
      "entropy": 5.716053915023804,
      "epoch": 4.3037917271407835,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003273914818261092,
      "loss": 4.901,
      "mean_token_accuracy": 0.22745440453290938,
      "num_tokens": 98482874.0,
      "step": 47445
    },
    {
      "entropy": 5.701595592498779,
      "epoch": 4.304245283018868,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032736001426451216,
      "loss": 4.8576,
      "mean_token_accuracy": 0.22890810072422027,
      "num_tokens": 98492995.0,
      "step": 47450
    },
    {
      "entropy": 5.7431660175323485,
      "epoch": 4.304698838896952,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003273285456200966,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22010931819677354,
      "num_tokens": 98503965.0,
      "step": 47455
    },
    {
      "entropy": 5.687218046188354,
      "epoch": 4.305152394775036,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003272970758935135,
      "loss": 4.884,
      "mean_token_accuracy": 0.21982184201478958,
      "num_tokens": 98515584.0,
      "step": 47460
    },
    {
      "entropy": 5.712123966217041,
      "epoch": 4.305605950653121,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032726560508541356,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22632251381874086,
      "num_tokens": 98527134.0,
      "step": 47465
    },
    {
      "entropy": 5.72504072189331,
      "epoch": 4.306059506531205,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032723413319644756,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22476772964000702,
      "num_tokens": 98537189.0,
      "step": 47470
    },
    {
      "entropy": 5.704430150985718,
      "epoch": 4.306513062409289,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003272026602272663,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22671725898981093,
      "num_tokens": 98547803.0,
      "step": 47475
    },
    {
      "entropy": 5.656298732757568,
      "epoch": 4.306966618287373,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032717118617852086,
      "loss": 4.8873,
      "mean_token_accuracy": 0.22278762012720107,
      "num_tokens": 98558332.0,
      "step": 47480
    },
    {
      "entropy": 5.723431205749511,
      "epoch": 4.307420174165458,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003271397110508619,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21766755729913712,
      "num_tokens": 98569046.0,
      "step": 47485
    },
    {
      "entropy": 5.6893257141113285,
      "epoch": 4.307873730043541,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003271082348449405,
      "loss": 4.8737,
      "mean_token_accuracy": 0.2260136842727661,
      "num_tokens": 98579358.0,
      "step": 47490
    },
    {
      "entropy": 5.681375789642334,
      "epoch": 4.308327285921625,
      "grad_norm": 1.25,
      "learning_rate": 0.0003270767575614074,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2232773020863533,
      "num_tokens": 98589224.0,
      "step": 47495
    },
    {
      "entropy": 5.725817108154297,
      "epoch": 4.3087808417997095,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003270452792009138,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2256453052163124,
      "num_tokens": 98598954.0,
      "step": 47500
    },
    {
      "entropy": 5.65271897315979,
      "epoch": 4.309234397677794,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032701379976411043,
      "loss": 4.8139,
      "mean_token_accuracy": 0.2240918129682541,
      "num_tokens": 98608813.0,
      "step": 47505
    },
    {
      "entropy": 5.645434856414795,
      "epoch": 4.309687953555878,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003269823192516486,
      "loss": 4.8496,
      "mean_token_accuracy": 0.2281813606619835,
      "num_tokens": 98618746.0,
      "step": 47510
    },
    {
      "entropy": 5.6741170406341555,
      "epoch": 4.310141509433962,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032695083766417913,
      "loss": 4.8491,
      "mean_token_accuracy": 0.23122671842575074,
      "num_tokens": 98628804.0,
      "step": 47515
    },
    {
      "entropy": 5.700046586990356,
      "epoch": 4.310595065312047,
      "grad_norm": 1.328125,
      "learning_rate": 0.000326919355002353,
      "loss": 4.8454,
      "mean_token_accuracy": 0.22894047796726227,
      "num_tokens": 98640056.0,
      "step": 47520
    },
    {
      "entropy": 5.660945177078247,
      "epoch": 4.311048621190131,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032688787126682143,
      "loss": 4.8241,
      "mean_token_accuracy": 0.23013663440942764,
      "num_tokens": 98651042.0,
      "step": 47525
    },
    {
      "entropy": 5.677336072921753,
      "epoch": 4.311502177068215,
      "grad_norm": 1.25,
      "learning_rate": 0.0003268563864582354,
      "loss": 4.843,
      "mean_token_accuracy": 0.22753780335187912,
      "num_tokens": 98661405.0,
      "step": 47530
    },
    {
      "entropy": 5.718926525115966,
      "epoch": 4.3119557329462985,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032682490057724616,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22202659100294114,
      "num_tokens": 98672323.0,
      "step": 47535
    },
    {
      "entropy": 5.735607481002807,
      "epoch": 4.312409288824383,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003267934136245047,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22524276077747346,
      "num_tokens": 98681431.0,
      "step": 47540
    },
    {
      "entropy": 5.571204090118409,
      "epoch": 4.312862844702467,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003267619256006623,
      "loss": 4.7991,
      "mean_token_accuracy": 0.24031769037246703,
      "num_tokens": 98692345.0,
      "step": 47545
    },
    {
      "entropy": 5.657980918884277,
      "epoch": 4.313316400580551,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003267304365063701,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22305618226528168,
      "num_tokens": 98703425.0,
      "step": 47550
    },
    {
      "entropy": 5.708440589904785,
      "epoch": 4.313769956458636,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003266989463422793,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22638604789972305,
      "num_tokens": 98714193.0,
      "step": 47555
    },
    {
      "entropy": 5.749283075332642,
      "epoch": 4.31422351233672,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032666745510904105,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23029360175132751,
      "num_tokens": 98723550.0,
      "step": 47560
    },
    {
      "entropy": 5.64726881980896,
      "epoch": 4.314677068214804,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032663596280730673,
      "loss": 4.8689,
      "mean_token_accuracy": 0.22294517308473588,
      "num_tokens": 98734075.0,
      "step": 47565
    },
    {
      "entropy": 5.657389545440674,
      "epoch": 4.315130624092888,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032660446943772755,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22976389676332473,
      "num_tokens": 98745447.0,
      "step": 47570
    },
    {
      "entropy": 5.707361030578613,
      "epoch": 4.315584179970973,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003265729750009547,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22419597804546357,
      "num_tokens": 98756169.0,
      "step": 47575
    },
    {
      "entropy": 5.703860235214234,
      "epoch": 4.316037735849057,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003265414794976397,
      "loss": 4.9223,
      "mean_token_accuracy": 0.22036930471658706,
      "num_tokens": 98767112.0,
      "step": 47580
    },
    {
      "entropy": 5.740336894989014,
      "epoch": 4.316491291727141,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003265099829284337,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22451334446668625,
      "num_tokens": 98777533.0,
      "step": 47585
    },
    {
      "entropy": 5.7530943870544435,
      "epoch": 4.316944847605225,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003264784852939881,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2183598831295967,
      "num_tokens": 98788101.0,
      "step": 47590
    },
    {
      "entropy": 5.773582029342651,
      "epoch": 4.317398403483309,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032644698659495437,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22269751280546188,
      "num_tokens": 98799438.0,
      "step": 47595
    },
    {
      "entropy": 5.65052433013916,
      "epoch": 4.317851959361393,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032641548683198384,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2212439477443695,
      "num_tokens": 98810325.0,
      "step": 47600
    },
    {
      "entropy": 5.616851425170898,
      "epoch": 4.318305515239477,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003263839860057279,
      "loss": 4.8387,
      "mean_token_accuracy": 0.22813035249710084,
      "num_tokens": 98821897.0,
      "step": 47605
    },
    {
      "entropy": 5.665419340133667,
      "epoch": 4.318759071117562,
      "grad_norm": 1.390625,
      "learning_rate": 0.000326352484116838,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22758509069681168,
      "num_tokens": 98833338.0,
      "step": 47610
    },
    {
      "entropy": 5.723418760299682,
      "epoch": 4.319212626995646,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032632098116596576,
      "loss": 4.8758,
      "mean_token_accuracy": 0.22723012864589692,
      "num_tokens": 98844308.0,
      "step": 47615
    },
    {
      "entropy": 5.703376817703247,
      "epoch": 4.31966618287373,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032628947715376247,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23022021502256393,
      "num_tokens": 98854491.0,
      "step": 47620
    },
    {
      "entropy": 5.687318134307861,
      "epoch": 4.320119738751814,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003262579720808797,
      "loss": 4.9122,
      "mean_token_accuracy": 0.21656761616468428,
      "num_tokens": 98864702.0,
      "step": 47625
    },
    {
      "entropy": 5.664725589752197,
      "epoch": 4.320573294629899,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032622646594796905,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22775158882141114,
      "num_tokens": 98874682.0,
      "step": 47630
    },
    {
      "entropy": 5.724459409713745,
      "epoch": 4.321026850507983,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003261949587556821,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21476083844900132,
      "num_tokens": 98885130.0,
      "step": 47635
    },
    {
      "entropy": 5.675551033020019,
      "epoch": 4.321480406386067,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032616345050467016,
      "loss": 4.886,
      "mean_token_accuracy": 0.22595305889844894,
      "num_tokens": 98896398.0,
      "step": 47640
    },
    {
      "entropy": 5.672812175750733,
      "epoch": 4.321933962264151,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003261319411955852,
      "loss": 4.8487,
      "mean_token_accuracy": 0.23337628692388535,
      "num_tokens": 98906571.0,
      "step": 47645
    },
    {
      "entropy": 5.651860952377319,
      "epoch": 4.322387518142235,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003261004308290786,
      "loss": 4.8147,
      "mean_token_accuracy": 0.2294571042060852,
      "num_tokens": 98916666.0,
      "step": 47650
    },
    {
      "entropy": 5.719958114624023,
      "epoch": 4.322841074020319,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003260689194058021,
      "loss": 4.8714,
      "mean_token_accuracy": 0.21983617842197417,
      "num_tokens": 98926599.0,
      "step": 47655
    },
    {
      "entropy": 5.635683155059814,
      "epoch": 4.323294629898403,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032603740692640733,
      "loss": 4.812,
      "mean_token_accuracy": 0.22517455071210862,
      "num_tokens": 98936512.0,
      "step": 47660
    },
    {
      "entropy": 5.709617042541504,
      "epoch": 4.323748185776488,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003260058933915459,
      "loss": 4.9934,
      "mean_token_accuracy": 0.2098014920949936,
      "num_tokens": 98946604.0,
      "step": 47665
    },
    {
      "entropy": 5.752908849716187,
      "epoch": 4.324201741654572,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003259743788018697,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2181127443909645,
      "num_tokens": 98956257.0,
      "step": 47670
    },
    {
      "entropy": 5.7294378757476805,
      "epoch": 4.324655297532656,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032594286315803026,
      "loss": 4.932,
      "mean_token_accuracy": 0.2204469084739685,
      "num_tokens": 98966346.0,
      "step": 47675
    },
    {
      "entropy": 5.674243783950805,
      "epoch": 4.3251088534107405,
      "grad_norm": 1.25,
      "learning_rate": 0.0003259113464606795,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22845391780138016,
      "num_tokens": 98977410.0,
      "step": 47680
    },
    {
      "entropy": 5.723272466659546,
      "epoch": 4.325562409288825,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003258798287104691,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22434550374746323,
      "num_tokens": 98987488.0,
      "step": 47685
    },
    {
      "entropy": 5.661989068984985,
      "epoch": 4.326015965166908,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032584830990805094,
      "loss": 4.8213,
      "mean_token_accuracy": 0.23168001770973207,
      "num_tokens": 98996792.0,
      "step": 47690
    },
    {
      "entropy": 5.649671030044556,
      "epoch": 4.326469521044992,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032581679005407667,
      "loss": 4.8039,
      "mean_token_accuracy": 0.22948745042085647,
      "num_tokens": 99007425.0,
      "step": 47695
    },
    {
      "entropy": 5.649986696243286,
      "epoch": 4.326923076923077,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003257852691491983,
      "loss": 4.8399,
      "mean_token_accuracy": 0.2244899556040764,
      "num_tokens": 99018275.0,
      "step": 47700
    },
    {
      "entropy": 5.712569665908814,
      "epoch": 4.327376632801161,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032575374719406755,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22437282502651215,
      "num_tokens": 99028853.0,
      "step": 47705
    },
    {
      "entropy": 5.714393901824951,
      "epoch": 4.327830188679245,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032572222418933644,
      "loss": 4.8995,
      "mean_token_accuracy": 0.21886498630046844,
      "num_tokens": 99039423.0,
      "step": 47710
    },
    {
      "entropy": 5.701907396316528,
      "epoch": 4.3282837445573294,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003256907001356568,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21604047417640687,
      "num_tokens": 99050041.0,
      "step": 47715
    },
    {
      "entropy": 5.687766027450562,
      "epoch": 4.328737300435414,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003256591750336806,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22190199047327042,
      "num_tokens": 99062050.0,
      "step": 47720
    },
    {
      "entropy": 5.602911329269409,
      "epoch": 4.329190856313498,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003256276488840596,
      "loss": 4.783,
      "mean_token_accuracy": 0.22968636453151703,
      "num_tokens": 99072337.0,
      "step": 47725
    },
    {
      "entropy": 5.582111358642578,
      "epoch": 4.329644412191582,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032559612168744613,
      "loss": 4.7989,
      "mean_token_accuracy": 0.22836442440748214,
      "num_tokens": 99082468.0,
      "step": 47730
    },
    {
      "entropy": 5.666590833663941,
      "epoch": 4.3300979680696665,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003255645934444918,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22091216295957566,
      "num_tokens": 99091757.0,
      "step": 47735
    },
    {
      "entropy": 5.728024005889893,
      "epoch": 4.330551523947751,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003255330641558489,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2189284458756447,
      "num_tokens": 99103192.0,
      "step": 47740
    },
    {
      "entropy": 5.626647424697876,
      "epoch": 4.331005079825834,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003255015338221693,
      "loss": 4.7799,
      "mean_token_accuracy": 0.2293917328119278,
      "num_tokens": 99112660.0,
      "step": 47745
    },
    {
      "entropy": 5.738275337219238,
      "epoch": 4.331458635703918,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032547000244410515,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2241392746567726,
      "num_tokens": 99123867.0,
      "step": 47750
    },
    {
      "entropy": 5.692056322097779,
      "epoch": 4.331912191582003,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032543847002230845,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22272821962833406,
      "num_tokens": 99134445.0,
      "step": 47755
    },
    {
      "entropy": 5.657172155380249,
      "epoch": 4.332365747460087,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003254069365574313,
      "loss": 4.8375,
      "mean_token_accuracy": 0.22691401988267898,
      "num_tokens": 99143483.0,
      "step": 47760
    },
    {
      "entropy": 5.6528760433197025,
      "epoch": 4.332819303338171,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003253754020501259,
      "loss": 4.8517,
      "mean_token_accuracy": 0.23200729936361314,
      "num_tokens": 99154514.0,
      "step": 47765
    },
    {
      "entropy": 5.661369800567627,
      "epoch": 4.3332728592162555,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032534386650104444,
      "loss": 4.9212,
      "mean_token_accuracy": 0.2137492299079895,
      "num_tokens": 99165969.0,
      "step": 47770
    },
    {
      "entropy": 5.593125247955323,
      "epoch": 4.33372641509434,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003253123299108388,
      "loss": 4.7876,
      "mean_token_accuracy": 0.23204695880413057,
      "num_tokens": 99176595.0,
      "step": 47775
    },
    {
      "entropy": 5.700669193267823,
      "epoch": 4.334179970972424,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003252807922801615,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22432172447443008,
      "num_tokens": 99186916.0,
      "step": 47780
    },
    {
      "entropy": 5.731165838241577,
      "epoch": 4.334633526850508,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003252492536096646,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2133137121796608,
      "num_tokens": 99197892.0,
      "step": 47785
    },
    {
      "entropy": 5.743923377990723,
      "epoch": 4.3350870827285926,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003252177139000003,
      "loss": 4.9272,
      "mean_token_accuracy": 0.22680503576993943,
      "num_tokens": 99208373.0,
      "step": 47790
    },
    {
      "entropy": 5.707012367248535,
      "epoch": 4.335540638606676,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003251861731518209,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22761063128709794,
      "num_tokens": 99218788.0,
      "step": 47795
    },
    {
      "entropy": 5.770860385894776,
      "epoch": 4.33599419448476,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032515463136577864,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22563234716653824,
      "num_tokens": 99228653.0,
      "step": 47800
    },
    {
      "entropy": 5.671164798736572,
      "epoch": 4.3364477503628445,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032512308854252577,
      "loss": 4.804,
      "mean_token_accuracy": 0.23209425061941147,
      "num_tokens": 99239925.0,
      "step": 47805
    },
    {
      "entropy": 5.632341098785401,
      "epoch": 4.336901306240929,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003250915446827147,
      "loss": 4.7608,
      "mean_token_accuracy": 0.23194602727890015,
      "num_tokens": 99250091.0,
      "step": 47810
    },
    {
      "entropy": 5.648854684829712,
      "epoch": 4.337354862119013,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003250599997869977,
      "loss": 4.9003,
      "mean_token_accuracy": 0.21233980655670165,
      "num_tokens": 99260799.0,
      "step": 47815
    },
    {
      "entropy": 5.630479717254639,
      "epoch": 4.337808417997097,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003250284538560272,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23726825714111327,
      "num_tokens": 99270338.0,
      "step": 47820
    },
    {
      "entropy": 5.674000692367554,
      "epoch": 4.3382619738751815,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003249969068904554,
      "loss": 4.8625,
      "mean_token_accuracy": 0.2235625520348549,
      "num_tokens": 99280419.0,
      "step": 47825
    },
    {
      "entropy": 5.737470293045044,
      "epoch": 4.338715529753266,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032496535889093496,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2193329319357872,
      "num_tokens": 99291590.0,
      "step": 47830
    },
    {
      "entropy": 5.705243444442749,
      "epoch": 4.33916908563135,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032493380985811814,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22330475896596907,
      "num_tokens": 99301987.0,
      "step": 47835
    },
    {
      "entropy": 5.733073616027832,
      "epoch": 4.339622641509434,
      "grad_norm": 1.46875,
      "learning_rate": 0.00032490225979265734,
      "loss": 4.9118,
      "mean_token_accuracy": 0.219923534989357,
      "num_tokens": 99312231.0,
      "step": 47840
    },
    {
      "entropy": 5.704366683959961,
      "epoch": 4.340076197387518,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032487070869520517,
      "loss": 4.83,
      "mean_token_accuracy": 0.22698791325092316,
      "num_tokens": 99322784.0,
      "step": 47845
    },
    {
      "entropy": 5.729310941696167,
      "epoch": 4.340529753265602,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000324839156566414,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22054585665464402,
      "num_tokens": 99333989.0,
      "step": 47850
    },
    {
      "entropy": 5.633415174484253,
      "epoch": 4.340983309143686,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003248076034069364,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22615868598222733,
      "num_tokens": 99345034.0,
      "step": 47855
    },
    {
      "entropy": 5.639495372772217,
      "epoch": 4.3414368650217705,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003247760492174248,
      "loss": 4.8084,
      "mean_token_accuracy": 0.22888606190681457,
      "num_tokens": 99356062.0,
      "step": 47860
    },
    {
      "entropy": 5.6990961074829105,
      "epoch": 4.341890420899855,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003247444939985319,
      "loss": 4.8434,
      "mean_token_accuracy": 0.23153524100780487,
      "num_tokens": 99365315.0,
      "step": 47865
    },
    {
      "entropy": 5.78768482208252,
      "epoch": 4.342343976777939,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00032471293775091013,
      "loss": 4.8902,
      "mean_token_accuracy": 0.23138579428195954,
      "num_tokens": 99374897.0,
      "step": 47870
    },
    {
      "entropy": 5.655116558074951,
      "epoch": 4.342797532656023,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032468138047521216,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2294129639863968,
      "num_tokens": 99383624.0,
      "step": 47875
    },
    {
      "entropy": 5.636184978485107,
      "epoch": 4.343251088534108,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003246498221720906,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22156410813331603,
      "num_tokens": 99393791.0,
      "step": 47880
    },
    {
      "entropy": 5.67339301109314,
      "epoch": 4.343704644412192,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000324618262842198,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2275726854801178,
      "num_tokens": 99403687.0,
      "step": 47885
    },
    {
      "entropy": 5.726601314544678,
      "epoch": 4.344158200290276,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032458670248618714,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22387943118810655,
      "num_tokens": 99413411.0,
      "step": 47890
    },
    {
      "entropy": 5.738475036621094,
      "epoch": 4.34461175616836,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003245551411047106,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22198912352323533,
      "num_tokens": 99425563.0,
      "step": 47895
    },
    {
      "entropy": 5.657603549957275,
      "epoch": 4.345065312046444,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032452357869842117,
      "loss": 4.8543,
      "mean_token_accuracy": 0.23643344938755034,
      "num_tokens": 99436297.0,
      "step": 47900
    },
    {
      "entropy": 5.69480676651001,
      "epoch": 4.345518867924528,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003244920152679715,
      "loss": 4.8709,
      "mean_token_accuracy": 0.22540836930274963,
      "num_tokens": 99445745.0,
      "step": 47905
    },
    {
      "entropy": 5.641480398178101,
      "epoch": 4.345972423802612,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032446045081401426,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2231060892343521,
      "num_tokens": 99456915.0,
      "step": 47910
    },
    {
      "entropy": 5.746298313140869,
      "epoch": 4.3464259796806965,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003244288853372024,
      "loss": 4.8887,
      "mean_token_accuracy": 0.21986974477767945,
      "num_tokens": 99466778.0,
      "step": 47915
    },
    {
      "entropy": 5.721002054214478,
      "epoch": 4.346879535558781,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003243973188381885,
      "loss": 4.8496,
      "mean_token_accuracy": 0.22375288605690002,
      "num_tokens": 99475984.0,
      "step": 47920
    },
    {
      "entropy": 5.611246776580811,
      "epoch": 4.347333091436865,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032436575131762553,
      "loss": 4.8513,
      "mean_token_accuracy": 0.2224792718887329,
      "num_tokens": 99486238.0,
      "step": 47925
    },
    {
      "entropy": 5.780407857894898,
      "epoch": 4.347786647314949,
      "grad_norm": 1.53125,
      "learning_rate": 0.00032433418277616623,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22479342371225358,
      "num_tokens": 99497806.0,
      "step": 47930
    },
    {
      "entropy": 5.687092351913452,
      "epoch": 4.348240203193034,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003243026132144635,
      "loss": 4.8745,
      "mean_token_accuracy": 0.21963958740234374,
      "num_tokens": 99507214.0,
      "step": 47935
    },
    {
      "entropy": 5.733822011947632,
      "epoch": 4.348693759071118,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003242710426331701,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2217227354645729,
      "num_tokens": 99517025.0,
      "step": 47940
    },
    {
      "entropy": 5.739341449737549,
      "epoch": 4.349147314949202,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032423947103293893,
      "loss": 4.914,
      "mean_token_accuracy": 0.2255868822336197,
      "num_tokens": 99526972.0,
      "step": 47945
    },
    {
      "entropy": 5.721365308761596,
      "epoch": 4.3496008708272855,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032420789841442306,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21642527878284454,
      "num_tokens": 99538583.0,
      "step": 47950
    },
    {
      "entropy": 5.613011074066162,
      "epoch": 4.35005442670537,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032417632477827526,
      "loss": 4.8119,
      "mean_token_accuracy": 0.22948300391435622,
      "num_tokens": 99549178.0,
      "step": 47955
    },
    {
      "entropy": 5.689499616622925,
      "epoch": 4.350507982583454,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032414475012514847,
      "loss": 4.852,
      "mean_token_accuracy": 0.23075646460056304,
      "num_tokens": 99558491.0,
      "step": 47960
    },
    {
      "entropy": 5.684060573577881,
      "epoch": 4.350961538461538,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003241131744556958,
      "loss": 4.918,
      "mean_token_accuracy": 0.21956392973661423,
      "num_tokens": 99568711.0,
      "step": 47965
    },
    {
      "entropy": 5.658336591720581,
      "epoch": 4.351415094339623,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003240815977705702,
      "loss": 4.882,
      "mean_token_accuracy": 0.22137337177991867,
      "num_tokens": 99578329.0,
      "step": 47970
    },
    {
      "entropy": 5.638819742202759,
      "epoch": 4.351868650217707,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032405002007042453,
      "loss": 4.7342,
      "mean_token_accuracy": 0.2339604079723358,
      "num_tokens": 99588646.0,
      "step": 47975
    },
    {
      "entropy": 5.726035690307617,
      "epoch": 4.352322206095791,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032401844135591213,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22094153463840485,
      "num_tokens": 99598780.0,
      "step": 47980
    },
    {
      "entropy": 5.709539604187012,
      "epoch": 4.352775761973875,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003239868616276857,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22480917423963548,
      "num_tokens": 99609528.0,
      "step": 47985
    },
    {
      "entropy": 5.613860988616944,
      "epoch": 4.35322931785196,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003239552808863986,
      "loss": 4.7971,
      "mean_token_accuracy": 0.22886189371347426,
      "num_tokens": 99619928.0,
      "step": 47990
    },
    {
      "entropy": 5.605508708953858,
      "epoch": 4.353682873730044,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032392369913270376,
      "loss": 4.7713,
      "mean_token_accuracy": 0.2326163098216057,
      "num_tokens": 99630135.0,
      "step": 47995
    },
    {
      "entropy": 5.713461303710938,
      "epoch": 4.354136429608127,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003238921163672544,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21180067509412764,
      "num_tokens": 99641557.0,
      "step": 48000
    },
    {
      "epoch": 4.354136429608127,
      "eval_entropy": 5.389565399976877,
      "eval_loss": 4.96226692199707,
      "eval_mean_token_accuracy": 0.22942079486687797,
      "eval_num_tokens": 99641557.0,
      "eval_runtime": 20.6495,
      "eval_samples_per_second": 1712.393,
      "eval_steps_per_second": 214.049,
      "step": 48000
    }
  ],
  "logging_steps": 5,
  "max_steps": 110230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.161275514380288e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}